BioMEMS (Microsystems)

BioMEMS MICROSYSTEMS Series Editor Stephen D. Senturia Massachusetts Institute of Technology Editorial Board Roger T...

Author: Gerald Urban

158 downloads 2937 Views 15MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

BioMEMS

MICROSYSTEMS Series Editor Stephen D. Senturia Massachusetts Institute of Technology Editorial Board Roger T. Howe, University of California, Berkeley D. Jed Harrison, University of Alberta Hiroyuki Fujita, University of Tokyo Jan-Ake Schweitz, Uppsala University OTHER BOOKS IN THE SERIES: x

x

x

x

x

x

x

x

x

Principles and Applications of NanoMEMS Ph... Series: Microsystems, Vol. 15 De Los Santos, Hector 2005, XV, 254 p., Hardcover, ISBN: 1-4020-3238-2 Optical Microscanners and Microspectrometers Using Thermal Bimorph Actuators Series: Microsystems, Vol. 14 Lammel, Gerhard, Schweizer, Sandra, Renaud, Philippe 2002, 280 p., Hardcover, ISBN: 0-7923-7655-2 Optimal Synthesis Methods for MEMS Series: Microsystems, Vol. 13 Ananthasuresh, S.G.K. (Ed.) 2003, 336 p., Hardcover, ISBN: 1-4020-7620-7 Micromachined Mirrors Series: Microsystems, Vol. 12 Conant, Robert 2003, XVII, 160 p., Hardcover, ISBN: 1-4020-7312-7 Heat Convection in Micro Ducts Series: Microsystems, Vol. 11 Zohar, Yitshak 2002, 224 p., Hardcover, ISBN: 1-4020-7256-2 Microfluidics and BioMEMS Applications Series: Microsystems, Vol. 10 Tay, Francis E.H. (Ed.) 2002, 300 p., Hardcover, ISBN: 1-4020-7237-6 Materials & Process Integration for MEMS Series: Microsystems, Vol. 9 Tay, Francis E.H. (Ed.) 2002, 300 p., Hardcover, ISBN: 1-4020-7175-2 Scanning Probe Lithography Series: Microsystems, Vol. 7 Soh, Hyongsok T., Guarini, Kathryn Wilder, Quate, Calvin F. 2001, 224 p., Hardcover ISBN: 0-7923-7361-8 Microscale Heat Conduction in Integrated Circuits and Their Constituent Films Series: Microsystems, Vol. 6 Sungtaek Ju, Y., Goodson, Kenneth E. 1999, 128 p., Hardcover, ISBN: 0-7923-8591-8

BioMEMS Edited by

Gerald A. Urban Albert-LudwigsGermany

Freiburg,

A C.I.P. Catalogue record for this book is available from the Library of Congress.

ISBN 10 ISBN 13 ISBN 10 ISBN 13

0-387-28731-0 (HB) 978-0-387-28731-7 (HB) 0-387-28732-9 ( e-book) 978-0-387-28732-4 (e-book)

Published by Springer, P.O. Box 17, 3300 AA Dordrecht, The Netherlands. www.springer.com

Printed on acid-free paper

All Rights Reserved © 2006 Springer No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Printed in the Netherlands.

This book is dedicated to Birgül, Gregor and Oliver

Contents

Contributing Authors

xv xvii

Preface EARLY BIOMEMS MULTI-SENSOR NEUROPROBES

1

1. 2.

1

INTRODUCTION EVOLUTION OF MICRO-SENSOR ARRAY DESIGNS FOR MEDICAL RESEARCH

2.1 2.2 2.3 3. 4. 5.

Electrical signal monitoring Sensor Design Evolution: from 2D to 3D Chamber Type of Electrochemical Oxygen Sensors OTHER APPLICATIONS—THE FIRST MICRO-FLUIDIC DEVICE CONCLUSION REFERENCES

MULTI-PARAMETER BIOMEMS FOR CLINICAL MONITORING 1. 2. 2.1 2.2 3. 3.1 3.2 3.3

INTRODUCTION BIOSENSORS Principle of Biosensors Amperometric Biosensors CLINICAL MONITORING Multi-analyte measurement Micro-dialysis BioMEMS for clinical monotoring

3 3 6 7 11 11 11 15 15 16 16 17 18 20 21 24

viii 3.4 3.5 3.5.1 3.5.2 4. 5.

Contents Multi-parameter monitoring Applications Monitoring of glucose and lactate with a micro-dialysis probe Ammonia monitoring CONCLUSIONS AND OUTLOOK REFERENCES

26 32 32 34 36 36

NEURAL IMPLANTS IN CLINICAL PRACTICE Interfacing neurons for neuro-modulation, limb control, and to restore vision–Part I

41

1. 2. 2.1 2.2 2.3 2.4 2.5 2.6 2.7 3. 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12 4.

41 45 45 47 48 49 50 52 54 55 56 57 59 60 61 61 63 65 66 67 68 69 70

INTRODUCTION TO NEURAL IMPLANTS ANATOMICAL AND BIOPHYSICAL FUNDAMENTALS Peripheral Nerve Anatomy Mechanisms of Peripheral Nerve Damage Excitability of Nerves Electrical Modelling of the Nerve Membrane Propagation of Action Potentials Extra–cellular Stimulation of Nerve Fibres Selective Activation of Nerve Fibres CLINICAL IMPLANTS Electrodes—The Key Component in Neural Prostheses Cardiac Pacemakers Implantable Defibrillators Cochlea Implants Phrenic Pacemakers Grasp Neuroprostheses Neuroprostheses for gait and posture Spinal Root Stimulator Drop Foot Stimulator Neuro-modulation Deep Brain Stimulation Vagal Nerve Stimulation REFERENCES

BIOMEDICAL MICRODEVICES FOR NEURAL IMPLANTS Interfacing neurons for neuromodulation, limb control, and to restore vision–Part II

71

1. 2. 2.1

71 75 77

THE CHALLENGE OF MICRO-IMPLANTS VISION PROSTHESES Cortical Vision Prostheses

Contents

ix

2.2 Optic Nerve Vision Prosthesis 2.3 Retinal Implants 2.3.1 Subretinal Vision Prostheses 2.3.2 Epiretinal Vision Prostheses 2.4 Conclusions on Vision Prostheses 3. PERIPHERERAL NERVE INTERFACES 3.1 Non-Invasive Nerve Interfaces 3.2 ‘Semi’-Invasive Interfaces 3.3 Invasive Interfaces 3.3.1 Intrafascicular Electrodes 3.3.2 Needle-Like Electrodes 3.3.3 Regenerative type of electrode 3.4 Biohybrid Approaches 4. FUTURE APPLICATIONS 4.1 Interfacing the Brain 4.2 Spinal Cord Implants 4.3 Multi-modal Neural Implants 5. CONCLUDING REMARKS 6. NEURAL IMPLANTS: BOON OR BANE? REFERENCES 7.

80 81 81 85 90 90 91 94 96 96 97 99 102 105 105 108 109 110 111 113

MICRO-FLUIDIC PLATFORMS

139

1. 2. 3. 3.1

139 141 142

3.2 3.3 3.3.1 3.3.2 3.3.3 3.4 3.4.1 3.4.2 4. 5.

INTRODUCTION WHAT IS A MICRO-FLUIDIC PLATFORM EXAMPLES OF MICRO-FLUIDIC PLATFORMS PDMS based Micro-fluidics for Large Scale Integration (‘Fluidigm platform’) Micro-fluidics on a Rotating Disk (‘Lab on a Disk’) Droplet based micro-fluidics (DBM) DBM based on electro-wetting DBM based on surface acoustic waves DBM based on two phase liquid flow Non-contact liquid dispensing ‘Dispensing Well Plate’ for ‘High Throughput Screening’ ‘TopSpot print heads’ for ‘High Throughput Fabrication of Microarrays’ CONCLUSION REFERENCES

142 146 149 149 151 153 155 158 160 161 162

x

Contents

DNA BASED BIO–MICRO-ELECTRONIC MECHANICAL SYSTEMS

167

1. INTRODUCTION 1.1 The unique features of nucleic acids 1.2 Lab on the Chip 1.2.1 Electrophoresis 1.2.2 Polymerase Chain Reaction (PCR) 1.3 Biochemical reaction chains for integration: biosensors and the ‘lab biochip’ 2. MICROARRAYS AND BIOCHIPS BASED ON DNA 2.1 The typical microarray experiment 2.2 Manufacturing of Microarrays 2.2.1 Synthesis on the chip 2.2.2 Spotting techniques 2.3 Transcription Analysis 2.4 Oligonucleotide Arrays for sequencing 2.5 Active arrays 2.5.1 Enzymes acting on immobilised DNA 2.5.2 PCR on the Chip 2.6 Integrated PCR 2.6.1 Micro-chamber Chips 2.6.2 Micro-fluidics Chips 3. NANO- BIOTECHNOLOGY: DNA AS MATERIAL 3.1 DNA directed immobilisation and nucleic acid tags 3.2 DNA for regular structures 3.3 DNA to structure surfaces 3.3.1 Stretching of DNA by fluidics 3.3.2 Stretching DNA by AC electric fields 3.4 Metallisation of DNA for electronic circuits 4. REFERENCES

171 172 173 173 174 175 175 176 176 177 178 182 182 183 184 185 187 189 189 190 191 192

SEPARATION AND DETECTION ON A CHIP

199

1. 2. 2.1 2.2 3. 3.1 3.2

199 201 201 203 206 208 209

INTRODUCTION THEORY OF CAPILLARY ELECTROPHORESIS ON A CE CHIP Mobility of ions Electro-osmotic flow JOULE HEATING IN MICROFABRICATED DEVICES Separation efficiency of a CE chip Separation of biomacromolecules and particles

167 168 168 169 169

Contents 4. 4.1 4.2 4.3 5. 6. 7. 8.

BUILDING BLOCKS OF CE CHIP DEVICES Wafer materials, micromachining and wafer bonding Power supplies, pumping, injection and channel geometries Detection strategies SELECTED EXAMPLES FOR CE ON A CHIP DIELECTROPHORESIS OUTLOOK K REFERENCES

xi 209 209 214 218 225 227 230 230

PROTEIN MICROARRAYS: TECHNOLOGIES AND APPLICATIONS

245

1. 2. 2.1 2.2 3. 4. 5.

245 249 249 255 257 260 261

INTRODUCTION FORWARD PHASE PROTEIN MICROARRAYS Protein Expression Analysis Using Protein Microarrays Protein Interaction Microarrays REVERSE MICROARRAYS OUTLOOK K REFERENCES

LAB-ON-A-CHIP SYSTEMS FOR CELLULAR ASSAYS 1. 2. 3. 4. 4.1 4.1.1 4.1.2 4.1.3 4.2 4.3 5. 6. 7.

INTRODUCTION DESIGN AND FABRICATION OF CHIPS FOR CELL BASED ASSAYS CELL CULTURE ON CHIPS AND MICRO-FLUIDIC SYSTEMS DETECTABLE CELLULAR OUTPUT SIGNALS Cell Metabolism Extra–cellular Acidification Cellular Oxygen Exchange Miscellaneous Metabolic Parameters Cell Morphology Electrical Patterns CELL MANIPULATION ON CHIPS CONCLUSIONS AND FUTURE PROSPECTS REFERENCES

269 269 274 278 280 282 282 283 285 286 288 292 295 298

xiii

Contents

NETWORK ON CHIPS Spatial and temporal activity dynamics of functional networks in brain slices and cardiac tissue

309

1. 2. 2.1 3. 4. 5. 6. 7. 7.1 7.2 7.3 8. 9. 10. 11.

309 312 314 317 319 320 323 326 327 327 331 333 335 337 338

INTRODUCTION TECHNICAL ASPECTS AND UNDERLYING ASSUMPTIONS System requirements ORIGIN OF THE SIGNAL RECORDED SPATIAL RESOLUTION LFP AND PLASTICITY NETWORK DYNAMICS AND EPILEPTIFORM ACTIVITY DRUG TESTING WITH MEAS Using Network Properties as Endpoints in Drug Assays Assessing Distributions of Neuronal Responses to Dopamine Cardiopharmacology DATA ANALYSIS OUTLOOK K ACKNOWLEDGEMENTS REFERENCES

BIO–NANO-SYSTEMS Overview and Outlook

351

1. INTRODUCTION 2. BASIC CONCEPTS AND EXPERIMENTAL METHODS 2.1 Self-assembly 2.2 Optical properties of semiconducting nanocrystals 2.3 Optical properties of metal nanocrystals 2.4 Magnetic nanoparticles 2.5 Conjugation of nanomaterials and biomolecules 2.6 Bioanalysis with bio-nano-systems 2.6.1 DNA detection 2.6.2 Immuno assays 2.6.3 Fluorescence resonance energy transfer (FRET) 2.7 Imaging 3. APPLICATIONS 3.1 DNA detection 3.1.1 DNA detection by spectral shift 3.1.2 DNA detection by Mie scattering

351 352 353 354 356 357 358 360 360 363 363 364 364 364 365 366

Contents

xiii

3.2 3.2.1 3.2.2 3.2.3 3.3 4. 5. 6.

366 367 367 369 370 371 372 372

Immuno assays Immuno assay on a microtiter plate Immuno assays on polymer beads FRET with nanocrystals Imaging CONCLUSION AND OUTLOOK ACKNOWLEDGEMENTS REFERENCES

Contributing Authors

Chapter 1: Gerald A. Urban, Otto Prohaska, Fethi Olcaytug Chapter 2: Isabella Moser Chapter 3: Thomas Stieglitz, Joerg-Uwe Meyer Chapter 4: Thomas Stieglitz, Joerg-Uwe Meyer Chapter 5: Peter Koltay, Jens Ducrée, Roland Zengerle Chapter 6: Frank F. Bier, Dennie Andresen, Antje Walter Chapter 7: Richard B.M. Schasfoort, Anna J. TüdĘs Chapter 8: Dieter Stoll, Markus F. Templin, Jutta Bachmann, Thomas O. Joos Chapter 9: Bernhard Wolf, Martin Brischwein, Helmut Grothe, Christoph Stepper, Johann Ressler, Thomas Weyh Chapter 10: Ulrich Egert Chapter 11: Thomas Nann, Jürgen Riegeler, Gerald A. Urban

Preface

Explosive growth in the field of micro system technology (MST) has introduced a variety of promising products in major disciplines, from microelectronics, automotive, telecommunications, process technology, to life sciences. Especially life science and the health care business was, and is, expected to be a major market for MST products. Undoubtedly the merging of biological with micro- and nano science will create a scientific and technological revolution in the future. Unfortunately such obvious facts were corrupted in the past by fancy visions, such as micro-submarines swimming in the arterial system repairing calcification of blood vessels. Major financial resources were dedicated to this new and exciting technology. However, the breakthrough, predicted in a ‘science fiction’-like prophecy, was significantly delayed as a result of the complexity and difficulties which MST developments faced in life science applications. This led to some misunderstandings of the real benefits of this technology and of the outcome for clinical science. MST on its own is now becoming established, and it will be valuable to look at the lessons of the past, the practical issues, and the future expectations for life science applications. Therefore the aim of this review is to gain a realistic view of this topic. In this book we shall display most of the important bio-related MST activities, and it will become obvious that micro miniaturization of devices, down to the nano-scale, approaching the size of biological structures, will be a prerequisite for the future success of life sciences. Microminiaturized analytical and therapeutic micro- and nano-systems will be mandatory for system biologists in the long run, in order to obtain insight into morphology and the interactive processes of the living system. This is the topic of system

xviii

Preface

biology, a new and very challenging research field investigating the whole complex functionality of a living system. With such a deeper understanding new and personalized drugs could be developed leading to a revolution in life science. However, the present tools for these scientific tasks are macro devices, such as MS or NMR equipment. The development of smart and sensitive micro analytical tools, ultimately with single molecule sensitivity, is obviously future MST tasks. Up to now, micro analytical devices are used in clinical analytics or in molecular biology as gene chips. In addition, standard micro biomedical products are employed in intensive care and the surgical theatre mainly for monitoring purposes. However, the gap between these two completely different scientific fields will be closed as soon as functional micro-devices can be produced allowing a deeper view into the function of cells, cell cultures, organs, and whole organisms. Despite that the word ‘micro system’ has become an established and commonly used term, it is still lacking a generally accepted definition. Some definitions emphasize more the micro miniaturization, others prefer a technology based view, while the third kind highlight the system aspect. Although the latter aspect is obviously a profound one, the others need still to be considered. Undoubtedly, miniaturization is a key feature of MST, yielding a significant benefit for applications in the biosciences and in space projects. In all other applications system specifications and costs, rather than miniaturization, have first priority. A new discipline evolved which focused on micro systems for living systems. Historically micro systems for automotive or computer periphery applications were called MicroElectroMechanical Systems (MEMS), since such micro-parts exhibited mechanical structures in acceleration and pressure sensor systems or micro-elements for actuating fluids. The combining of MEMS with biology and medicine created the common term ‘BIOMEMS’, although only in a few applications are mechanical movable structures used. Another important aspect of BIOMEMS devices is—despite the miniaturization—the use of innovative materials and biological substances in combination with micro-technological fabrication processes. The combination of these topics yielded completely new sensor devices, the so called ‘micro biosensors’. Examples of early realized miniaturized analytical BIOMEMS are glucose sensors, now well established in diabetology for home care measurement of diabetics, generating a fast growing market of several billion dollars. In the latter years another market pull for micro-biosensor devices was generated in the United States for measuring warfare agents, food and environmental safety. The ‘Homeland Security Program’ is providing

Preface

xix

substantial funds for the development of micro- and nano technology based handheld monitors for ubiquitous use. These BIOMEMS product successes point to the enormous potential of miniaturized devices which have historically always been of leading research interest. The initial stages of BIOMEMS developments can be found in the micro-sensor developments of the mid 20th century, whilst complete BIOMEMS based products just recently started to penetrate significant medical markets, illustrating the complexity and the difficulties of transferring a new technology into new products for large medical markets. The first chapter offers an overview of the basics and the historical background of BIOMEMS. Owing to the interdisciplinary character of BIOMEMS, different topics are presented in what follows in order to provide an overview of the broad BIOMEMS development field. The scientific topics range from implant devices to analytical micro-biosensors systems, complex DNA based micro systems, analytical protein arrays, and cell based systems. Some of the necessary technological prerequisites are micro-fluidic platforms and separation based tools on chips. As a view into the future of the emerging fields of whole tissue based arrays, as well as bio-nano-technology, research results are highlighted. These most interesting scientific research fields are expected to yield exciting research results and long awaited products for the next decade, particularly in the direct intra–cellular observation of metabolic pathways using nano-tools. The understanding of these complex functions will be a prerequisite for the developments in two emerging fields, system biology and BIONEMS—BIOlogical Nano-ElectroMechanical Systems. Subsequent book editions will undoubtedly deal with such topics in more detail. In this book the reader will be able to review at a glance the exciting field of bio microsystems, from their beginnings to indicators of future successes. This book will also show that a broad penetration of micro- and nanotechnologies into biology and medicine will be mandatory for future progress of scientific and new product development in life science.

Chapter 1 EARLY BIOMEMS MULTI-SENSOR NEUROPROBES G.A. Urban, O. Prohaska, F. Olcaytug IMTEK-Sensoren, Albert-Ludwigs-Universität Freiburg, Deutschland

Abstract:

The term ‘BIOMEMS’ can be traced back to the beginning of biomedical engineering sciences. Minimally invasive surgical tools, even implantable pacemakers, can be seen as predecessors of ‘BIOMEMS’. However, BIOMEMS type of devices, as we understand them today, were used first in neuroscience. One prominent and sophisticated example consisted of arraytype of neuroprobes and micro-fabricated, three-dimensional, chamber type of multiple sensors which were developed for brain research, measuring the extra-cellular electrical signals of nerve cells. Electrochemical measurements were performed within the electrolyte-filled miniaturized chambers by means of thin film electrodes. The chambers were formed by a specially developed silicon nitride insulation layer, which was deposited over a sacrificial layer of photo-resist, the first example of surface micro-machining in sensor technology. Also the first micro-fluidics structures on chips were introduced by this technology, leading to the first patent on capillary electrophoresis on a chip. The simple low cost formation of a variety of three-dimensional structures makes this technology an early example of a number of most valuable BIOMEMS device development and production processes.

Key words:

History of Biomems, Neuroprobes, chamber-type electrodes, needle sensor arrays, microfluidics

1.

INTRODUCTION

Despite that BIOMEMS are undoubtedly a very hot topic today, the roots of micro systems used in life science are surprisingly old. For the most important clinical disciplines, such as cardiology, neurology, and endoscopic surgery, many instruments were developed by means of delicate machinery and mechatronic technology. Endoscopic surgery had already been performed in the early 19th century [1]. The progress in the development of

G. Urban (ed.), BioMEMS , 1-13. © 2006 Springer. Printed in the Netherlands.

1

2

EARLY BIOMEMS MULTI-SENSOR NEUROPROBES E

effective surgical procedures and the pressure to reduce health care costs have led to significantly shorter surgical interventions on patients and to their shorter stays in the hospital. Minimally invasive surgery has evolved from this drive for patient treatment and improvement of cost efficiency [2]. Through the use of conventional miniaturization techniques, in combination with new materials and micro-optical measurement units, surgical tools and catheters can be expected to be equipped in the near future with complex optical and mechanical actuating micro-systems. An actual example of an advanced ‘BIOMEMS’ endoscope is an autonomous capsule endoscope for the video inspection of the human gastrointestinal system [3]. The next step involves also electrical components opening up the field of functional electro-stimulation (FES) [4]. The combination of micro-technologies and microelectronics led to autonomous, energy driven electronic devices with an electrical actuator, such as pacemakers. After the first developments in the early 20th century, Rune Elmquist and Åke Senning developed in 1958 the first fully implantable pacemaker. The first hybrid multi-channel cochlea implant was inserted in the year 1977 [5,6]. However, not only were micro-mechanical and microelectronic functionalities introduced in biology and medicine, but also in chemical sensors. Piet Bergveld, invented the first miniaturized ISFET pH sensor in the year 1972, produced by means of semiconductor and micro-fabrication techniques [7]. The very first micro-systems in biology and medicine, however, were developed for the advanced life science disciplines of the neurosciences and neurobiology for the investigation of functionalities, metabolic disorders, and biological processes down to the cellular level. One of the most challenging scientific areas is brain research. Miniaturized and micro-devices are needed to monitor and to influence neuronal activities, especially at the cellular level. In the first decades of the 20th century single glass or wire type of sensors were used to measure physiological parameters on the surface of or within human or animal tissue [8]. Such miniaturized devices immediately offered new insights into brain functions, information processing, and tissue abnormalities by measuring intra- and extra-cellular electrical events in the nervous tissue and in single neurons. Wires were also used to make sensor arrays for the analysis of the propagation of physiological events on the surface of organs. However, it was very cumbersome to form sensor arrays in an accurate and reproducible manner with wire or glass electrodes. In addition the array arrangement was not reproducible, nor did it yield the accuracy required for use at the level of cellular dimensions.

EARLY BIOMEMS MULTI-SENSOR NEUROPROBES

3

When the complexity of the brain demanded smaller and more precise electrodes, advanced photolithography and semiconductor micro-fabrication techniques opened the door for the development of miniaturized biosensors and instrumentation to fulfil the new needs of neurophysiologists. The first micro-machined devices were reported in the 60s and 70s [9,10]. They demonstrated the feasibility of producing miniaturized pressure sensors and gas chromatography columns. Grooves or indentations were etched into silicon which was then covered by a glass plate and sealed to the silicon to form micro-chambers, containers, and channels. The very first microfabricated structures invoked the interest of the medical community, although it took many years of product development before commercialisation of the first MEMS devices. Ken Wise used silicon etching and integrated circuit fabrication techniques to produce an array of thin bladelike microstructures with metal tips and insulated electrical connectors which mimicked conventional metal-micro electrodes [11]. A significant advantage of this device fabrication technique over conventional metal and glass electrode array formation was that reproducible detectors could be fabricated, and, at the same time, could be equipped with integrated data collecting and processing electronics. Additionally planar electrode arrays for measuring electrical activities of cultured neurons were investigated early on [12].

2.

EVOLUTION OF MICRO-SENSOR ARRAY DESIGNS FOR MEDICAL RESEARCH

Otto Prohaska and his team introduced in the 70s the first planar micro array sensor arrangements for the measurement of extra-cellular electrical nerve cell activity [13]. These probes yielded significant contributions to the resolution of a broad range of questions, concerning nerve cell interactions, nerve cell group activities, and the propagation of pathological cell activities in the cortical part of the brain [14–16].

2.1

Electrical signal monitoring

It was important to design low-contact impedance signal detectors in order to achieve a distortion-free measurement of the low frequency extracellular electrical signals in the brain in the range below 30 Hz. The impedance of the detectors is defined by the electrode–tissue interface and can be controlled through the proper choice of electrode material and surface area of the electrode. The electrode contact impedance, in turn, determines

4


the width’s and length’s sizes of the electrical connectors and the thickness of the insulation layers of these connectors, so that signal loss of the nerve cell’s activities at the electrode sites, as well as pick-up of signals generated by nerve cells situated along the conducting path, can be prevented. Finally, the electrode impedance also has an effect upon the crosstalk between two conductors which can be reduced to negligible values by proper spacing of the conductor paths and a specific quality of the insulation layer [17]. gold wire insulation layer synthetic resin

metal line t t 1 mm

electrode area

50 µm

glass substrate

electrode structure

1 µm 0.1mm

schematic representation

Figure 1-1. Schematic representation of a probe for electrical signal measurements with nine gold thin film sensors and SiNx insulated conducting lines which are leading to the bonding pads. The gold surface was covered electrochemically with silver which was then chlorided to increase the surface area and thus reduce the electrode impedance.

Fig. 1-1 shows the schematic drawing of a needle-shaped probe with nine electrical signal sensors. Each of the nine sensors consists of a planar gold thin film electrode area. The thin film electrical connector is insulated by an approximately 1–2 µm thick silicon nitride (SiNx) layer. Fethi Olcaytug


5

developed a low temperature inductively coupled cross-field double frequency plasma enhanced chemical vapour deposition (PECVD) process for the production of these high quality, pore-free SiNx insulation layers [18]. The electrode area was then plated with a 1–2 µm thick silver layer which was electrochemically converted on the surface into an Ag/AgCl layer with low electrode–electrolyte impedance in the low frequency range of electrical signals [19]. The actual electrode surface areas were, however, about 100 times larger, depending upon the formation process of the silver chloride surface [17]. Up to 16 electrodes with geometric recording areas of 50¯50 µm2 or 10¯10 µm2 were arranged in one or two rows on a probe. The probe consisted of a 100 µm thick glass plate which was cut with a diamond scriber to a needle-shaped form. Important for the success of these devices was the creation of a very fine tip, similar to that of glass microelectrodes, in order to prevent damage of the nerve cells during penetration of the probe into and its positioning within the brain.

Figure 1-2. Brain research probe consisting of a needle-shaped glass plate which is equipped with a fine tip and carries nine electrodes. Gold d wires connect the bonding pads with insulated copper wires which lead to the electronic measurement unit.

Fig. 1-2 shows a probe with one row of nine 50¯50 µm2 Ag/AgCl electrodes used to study the development and propagation of epileptic seizure activity in the cortical area of the brain [14,15]. The use of micro array probes has led to new neurological insights into brain function during epilepsy and Parkinson’s disease [20].

6

2.2


Sensor Design Evolution: from 2D to 3D

Otto Prohaska and his team invented and produced a three-dimensional (3D) sensor design in order to further reduce the effective recording sites of the detectors. The 3D thin film structure fabrication technique employs a sacrificial layer for the definition of the 3D configuration and is not restricted to silicon for its base material. A flat electrically insulating substrate, e.g., a 100 µm thick glass plate, was used for the fabrication of the medical probe. Gold thin film electrodes and electrical conductors were deposited onto the substrate for the formation of electrochemical sensors. The sensor areas were then surface-treated to obtain the desired performance characteristics. Subsequently photo-resist was used as a sacrificial layer and shaped by means of photo-lithography over the sensor areas to define the space which the miniaturized chambers would form around them. The photo-resist structures were then covered by the PECVD SiNx layer to form the walls of the desired three-dimensional micro-structures. During the single SiNx deposition process the electrical connectors were insulated at the same time as the chamber walls were formed. Subsequently a small opening was etched into the SiNx layer over the photo-resist, as shown schematically in Fig. 1-3, through which the photo-resist was dissolved. Upon insertion of the detector into an electrolyte the chamber was filled with electrolyte as a result of the capillary forces.

Figure 1-3. Cross-section of a chamber type of electrode with a large electrode and a small recording area. Principal advantage of this design: the size of the electrode area can be defined independently of the size of the recording area.

The electrolyte formed the electrical bridge between the tissue outside and the electrode inside the chamber. This opening in the SiNx layer served as the actual recording site, similar to the tip of a micro-glass electrode.


7

Figure 1-4. Scanning electron microscope photo of a row of chamber type of sensors: the insulated conducting lines are leading to the electrodes inside the chambers.

Fig. 1-4 shows a SEM photo of a chamber electrode array. The chambers are 100 µm wide, 150 µm long, and 2 µm high, generating an electrolyte volume of less than 10 pl around the electrical signal sensing electrode. The chamber type of sensors yielded exceptionally stable, low noise recordings of the low frequency brain signals [21,22]. This chamber type of sensor fabrication process represents the first application of surface micromachining in the development of medical sensors.

2.3

Chamber Type of Electrochemical Oxygen Sensors

The concept of a chamber type of sensor was used to design selfcontained electrochemical sensors, as shown schematically in Fig. 1-5. Initially these sensors were designed for the measurement of the partial pressure of oxygen in the tissue. The sensors were equipped with two openings for oxygen to diffuse to the gold sensing electrode located in the centre of the bottom of the 2 µm high chamber. The distance between the opening and the sensing electrode defines the response time of the detector. Fig. 1-6 shows an optical microscope photo of the miniaturized chamber type of oxygen sensor. The chamber’s design yields a thin layer electrochemical analysis device for stable, low noise recordings. Current– voltage measurements, performed with the chamber type of sensor, are shown in Fig. 1-7.

8


Figure 1-5. Schematic drawing of a chamber type of two-electrode electrochemical sensor. 1. SiNx chamber and insulation layer; 2. glass substrate; 3. thin film Ag electrode; 4. Ag/AgCl surface area of the reference electrode; 5. gold working electrode area; 6. electrical thin film conducting lines; 7. openings in the chamber through which the analyte can diffuse into the chamber and to the working electrode.

Figure 1-6. Optical microscope photo of the top view of the two-electrode chamber type electrochemical sensor which was used for measurement of the extra–cellular partial pressure of oxygen in tissue. SiNx is optically transparent and allows a clear view of the thin film electrode and conducting line structures.


9

Figure 1-7. Current–voltage graph generated with the two-electrode chamber type of microelectrochemical sensor for sensor calibration.

Gerald Urban developed a micro-thermistor, formed by amorphous Germanium, shown in Fig. 1-8. This device exhibits a temperature resolution of less than 1 mK and has a time constant of approximately 5 ms [23]. Here for the first time amorphous un-doped Germanium which is otherwise not commonly used in electronic integrated circuitry, was employed to create a superiorly working temperature sensor with a stable and reproducible temperature coefficient of 2 %/°K. The temperature sensor was integrated, together with the chamber type of electrical signal and oxygen sensors, onto a needle-shaped glass carrier, as shown in Fig.1-9. This probe was used for the investigations of interactions between the electrical and the metabolic activities in the brain. The simultaneous multiple measurements provided answers for a number of topics which had been discussed widely amongst neurophysiologists before the micro-machined chamber type of multiple sensor devices were available. In particular, the novel multiple sensor probes yielded results which allowed interpretation of the interdependence of nerve cell activities and oxygen supply to the tissue [24].

10 0


Figure 1-8. High resolution Germanium thin film temperature sensor, arranged in an array, with a fast response time.

Figure 1-9. Electrical signal, oxygen, and temperature sensors integrated on a needle-shaped glass probe with a fine tip which allows insertion of the probe into the tissue without destruction of the cells.


3.

11

OTHER APPLICATIONS—THE FIRST MICRO-FLUIDIC DEVICE

Micro-chambers were used to develop a variety of self-contained electrochemical cells. Depending on the cell design, two or three electrode arrangements of thin-layer electrochemical cells can be utilized to yield exceptional performance characteristics of a sensor. The addition of selective membranes inside the micro-chambers allows the reproducible fabrication of micro-enzyme and ion selective sensors. Micro-glucose and lactate sensors with enzyme layers in combination with the chamber type of sensor design were developed by Otto Prohaska and his team [25]. The low temperature SiNx formation in combination with photo-resist as a sacrificial layer provides a simple, versatile, and inexpensive mass production technology not limited to 3D micro-chambers. 3D channels can be formed for micro-fluidics applications, for separation analysis, or for gas or flow measurements in combination with integrated detectors and actuators, such as temperature sensors and heating elements. A modified channel design, equipped with conductivity and dielectric constant sensors was used for the fabrication of quality control sensors which could monitor the composition of gasoline and alcohol mixtures [26]. Also the first integrated gas chromatography module on a chip, designed by utilizing the capabilities of this surface micro-machining technology, was disclosed [27].

4.

CONCLUSION

Thin film and integrated circuit type of micro-fabrication techniques enabled the development of a large number of two- and three-dimensional micro-sensors and sensor arrays. In combination with selective and protective bio-structures, and with silicon micro-machining and integrated circuit fabrication and thin film techniques, the chamber type of structures were able to form the basis for cost-efficient, high-quality BIOMEMS devices with a broad range of medical and industrial applications.

5. [1] [2]

REFERENCES Rathet, P., Lutzeyer, W., Godwin, W., (1974), ‘Phillip Bossini and the Lichtleiter’, Urology, 3, 113–123. Bueb, G., Cuschieri, A., Perissat, J., (1994), ‘Operationslehre der Endoskopischen Chirurgie’, Springer, Berlin.

12 2


[3]

Appleyard, M., Fireman, Z., Glukhovsky, A., et al., (2000), ‘A randomized trial comparing wireless capsule endoscopy with push enteroscopy for the detection of small-bowel lesions’, Gastroenterology, 119, 1431–8. This volume, chapter 3. ‘History of cochlear implants’. http://www.dcig.de/ci-geschichte.htm. Hochmaier-Desoyer, I.J, Hochmair, E.S., Burian, K., Stiglbrunner, H.K., (1983), ‘Percepts from the Vienna Cochlear prosthesis’, Ann. N. Y. Acad. Sci 405, 295–306. Bergveld, P., (1972), ‘Development, operation, and application of the ion-sensitive field-effect transistor as a tool for electrophysiology’, IEEE Trans. on Biomed. Engin, BME-19, No. 5, 342. Bates, J.A.V., (1963), ‘Special investigations techniques-indwelling electrodes and electrocorticography in electroencephalography’, MacDonald, London. Fatt, P., Katz, B., (1952), ‘Spontaneous subthreshold activity at motor nerve endings’, Journal of Physiology, 117, 109. t silicon pressure transducer’, Gieles, A.CM., (1969), ‘Subminiature Digest IEEE ISSCC., Philadelphia, 108–109. Sander, C.S., Knutti, J.W., Meindl, J.M., (1980), ‘A monolithic capacitive pressure sensor with pulse-periodic output’, IEEE Trans. Electron. Devices ED-27 927–930. Petersen, K.E., (1982), ‘Silicon as a mechanical material’, IEEE Proc. 70, 420. Saadat, S., Terry, S.C., (1984), ‘A high-speed chromatographic gas analyzer’, American laboratory, 16, No. 3, 90. Wise, K., Angell, J., (1975), ‘A low-capacitance multi-electrode probe for use in extra–cellular neurophysiology’, IEEE Trans. Biomed. Eng. BME-22, No. 3, 212–219. Gross, G.W., Rieske, E., Kreutzberg, G.W., Meyer, A., (1977), ‘A new fixed-array multi-electrode system designed for long term monitoring of extra–cellular single unit neuronal activity in vitro’, Neuroscience letters, 6, 101–105. Prohaska, O., Olcaytug, F., Womastek, K., Petsche, H., (1977), ‘A multi-electrode for intracortical recordings produced by thin film technology’, Electroenceph. Clin. Neurophysio, 42, 421–423. Petsche, H., Prohaska, O., Rappelsberger, P., Vollmer R., Pockberger, H., (1977), ‘Simultaneous laminar intracortical recordings in seizures’, Electroenceph. Clin. Neurophysiol, 43, 414–417. Elger, C.E., Speckmann, E.J., Prohaska, O., Casper, H., (1981), ‘Pattern of intracortical potential distribution during focal interictal epileptiform discharges (FIED) and its relation to spinal field

[4] [5] [6] [7] [8] [9]

[10] [11] [12]

[13] [14] [15]


[16]

[17] [18] [19] [20]

[21] [22]

[23]

[24]

[25] [26] [27]

13

potentials in the rat’, Electroenceph. Clin. Neurophysiol, 51, 393–401. Buzsaki, G., Bickford, R.G., Ryan, L.J., Young, S., Prohaska, O., Mandel, R.J., Gage, F.H., (1989), ‘Multi-site recordings of brain filed potentials and unit activity in freely moving rats’, J. Neurosci. Methods, Vol 28, 209–217. Prohaska, O., Olcaytug, F., Pfundner, P., Dragaun, H., (1986), ‘Thin film electrode probes: Possibilities and limitations’, IEEE Trans. Biomed. Eng, BME-33, no.2, 223–230. G.Rieder, Olcaytug, F., (1982), ‘Preparation and dielectric properties of Si3N4 films’, Thin solid Films 89, 95–99. Geddes, L.A., (1972), ‘Electrodes and the measurement of bioelectric events’, Wiley-Interscience. Urban, G.A., Ganglberger, J., Olcaytug, F., Kohl, F., Schallauer, R., Trimmel, M., Schmid H., Prohaska, O., (1990), ‘Development of a multiple thin film semimicro-DC-probe for intracerebral recordings’, IEEE Trans. J. Biomed. Eng. 37, no.10. US patent 4,682,602, (1987), ‘Probe for medical application’. Prohaska, O., Kohl, F., Goiser, P., Olcaytug, F., Urban, G.A., Jachimovicz, A., Pirker, K., Chu, W., Patil, M., LaManna, J., Vollmer, R., (1987), ‘Multiple chamber type probe for biomedical application’, Proc. 4th Int. Conf. Solid State Sensors and Actuators, Tokyo, Japan, 812–815. Urban, G., Jachimowicz, A., Kohl, F., Kuttner, H., Olcaytug, F., Goiser, P., Prohaska, O., (1990), ‘High resolution thin film temperature sensor arrays for medical applications’, Sensors & Actuators A22, 650–654. Prohaska, O.J., Kohl, F., Goiser, P., Olcaytug, F., Urban, G.A., Jachimowicz, A., Pirker, K., Chu, W., Patil M., LaManna J., Vollmer R., (1987), ‘Multiple chamber type probe for biomedical application’, Digest of the Int. Conf. on Solid-State Sensors and Actuators, 812– 615. Schneider, B.H., Daroux, M.L., Prohaska, O., (1990), ‘Microminiature enzyme sensors for glucose and lactate based on chamber oxygen electrodes’, Sensors and Actuators, B1, 565–570. US patent 5,165,292, (1992), ‘Channel device and tube connection and their fabrication procedures’. US patent number 5,116,495, ‘Capillary Chromatography Device’, issued on May 1992, Application Number 455,502, filed on Dec 22, 1989.

Chapter 2 MULTI-PARAMETER BIOMEMS FOR CLINICAL MONITORING Dr. Isabella Moser IMTEK-Sensoren, Albert-Ludwigs-Universität, 79110 Freiburg; Dr. Isabella Moser Tel.: 49 761 203 7263, FAX: 49 761 203 7262 E-mail: [email protected]

Abstract:

Recent BioMEMS for clinical monitoring are mostly glucose sensing devices. One of the reasons is that for diabetes patients glucose monitoring means an important improvement of their life’s quality, and it is additionally a $4 billion a year business. Continuous glucose monitoring provides gapless glucose level control, an early warning of hypoglycaemia, and is intended to control insulin pumps. An upgrading to multi-parameter monitoring would not only benefit patients with severe metabolism defects but also the metabolism of diabetes patient could be better controlled by monitoring an additional parameter like lactate. Multi-parameter monitoring devices are not commercially available. One of the complications in the integration of different biosensors using the same detecting molecule for all analytes is chemical cross talk between adjacent amperometric biosensors. Recently some integrated biosensors were published, but either they were not mass producible or they were realized in an expensive silicon based technology. In addition to that most of them were not tested under monitoring conditions, but their integration principles will be discussed. As an example a low cost multi- parameter micro system, and some applications of it in clinical diagnosis, will be presented. Also an overview of non-invasive methods and (minimal) invasive methods will be given with a focus on micro-dialysis.

Key words:

Biosensors, BioMEMS, clinical monitoring, micro system

1.

INTRODUCTION

Clinical monitoring is the on-line measurement of clinical parameters over a period of interest. While in the intensive care unit the electrophysiological parameters such as ECG and pulse frequency can be measured non-invasively the magnitude of metabolic parameters like glucose can be


15

16

MULTI-PARAMETER BIOMEMS FOR CLINICAL MONITORING

measured accurately only after an in vivo assessment. After the collection of the sample it is analysed in the clinical laboratory or, to a smaller extent, with bedside clinical analysers. Especially in central labs of hospitals are the samples analysed with a delay of up to some hours. Over this period the concentration of the analyte in the sample can change by degradation owing to blood cells or chemical reactions. Even the bedside monitoring is not continuous because the withdrawal of the samples and the analysis needs time. So the sometimes important information of fast concentration chances cannot be detected. The most common measured parameter is glucose measured using a variety of electrochemical glucose biosensors. The reason for the outstanding importance of glucose biosensors or glucose sensing BioMEMS is that for diabetes patients glucose monitoring means an important improvement in their quality of life. The advantages of continuous glucose monitoring are obvious: gapless glucose level control, and therefore an early warning of hypoglycemia and better metabolic control, and, as a principal aim of the research, glucose biosensors should control insulin pumps to create an artificial pancreas.

2.

BIOSENSORS

2.1

Principle of Biosensors

Biosensors are analytical devices which consist of a biological molecular recognition element in close contact with a transducer. Biosensors measure selectively and reversibly the concentration of a chemical substance without addition of reagents. Combining biological recognition elements such as: a) enzymes; b) nucleic acids; c) antibodies; d) tissues; or even e) whole cells with 1) electrochemical; 2) optical; 3) thermometric; or 4) gravimetric transducers, a variety of different biosensors for its adequate application can be realized. Since 1962, since the description of the first biosensor [1] several reviews and books describing the progress in biosensor research were published [2,3,4,5,6].


2.2

17

Amperometric Biosensors

According to Roche [7] the ‘blood glucose monitoring market’ had a volume of 7.3 bn CHF in 2003. Most of the devices use an electrochemical transducer method. In the field of electrochemical biosensors [8] amperometric measurement is the most important method. Almost all commercially available biosensors are amperometric enzyme biosensors. By applying a proper potential between counter and working electrode immersed in an electrolyte containing an electro active species a current is induced which is related to the concentration of the electro active species in the sample solution which is oxidized or reduced in a heterogeneous electron transfer reaction. Consequently there are two possibilities of amperometric measurement. The first one is the cathodic reduction which is applied in, e.g., oxygen detection, the second method is the anodic oxidation which is the underlying principle of, e.g., H2O2 determination. In the strict sense oxygen sensors are chemical sensors because of the lack of a bio component. Total current is the sum of partial currents caused by all electro-active species for an applied potential. In order to achieve selectivity, it is important to block interfering responses. Different strategies can be applied: 1. Selection of an appropriate potential; 2. Selective transduction of the desired signal by using the selectivity of a biological active layer; 3. Controlled mass transport through perm selective membranes. The classical biosensor used in monitoring is an enzyme biosensor with an amperometric transduction principle. This enzyme catalyses the conversion of an analyte to a molecule which can be detected with the transducer. In the case of amperometry this means that it is oxidized or reduced at an electrode. The enzymatic reaction can be divided into two steps, recognition of the analyte, and lowering of the activation energy of a chemical reaction by bringing the reactants close and in a favourable orientation together. Beside this, distortion of the substrate caused by the enzyme has to be considered. Enzymes possess two advantages: they are extremely selective to a given substrate; and they are very effective in increasing the rate of a reaction, thus combining the recognition and amplification steps. Oxidases oxidize their substrates and need oxygen as a co-substrate, re-oxidizing the enzyme to the initial state. The hydrogen peroxide produced is again then oxidized at the electrode.

18

MULTI-PARAMETER BIOMEMS FOR CLINICAL MONITORING Glucose + O2

'

Gluconolactone + H2O2

Enzyme: glucose oxidase with its prosthetic group FAD Glucose + FAD FADH2 +O2

' '

Gluconolactone + FADH2 FAD + H2O2

Electrode reaction: Anodic reaction on platinum (+ 0.6V vs. Ag/AgCl; 3M): H2O2

'

2H+ +O2 + 2e-

The signal (current) is directly proportional to the concentration of the analyte in the sample:

Id

n A F DS ⋅

c0 , δN

A = area of the electrode, DS = diffusion coefficient of the analyte S., c ° = bulk solution concentration of the analyte, δN = thickness of the stagnant layer. There are different methods of immobilizing oxidases to the electrodes. Simple adsorption, covalent coupling [9], and gel-entrapment [10,11] can be used.

3.

CLINICAL MONITORING

Fully implanted systems consisting of an insulin pump and a glucose biosensor are very attractive since there is no transcutaneous access, avoiding infection paths and providing a high degree of wearing comfort for the patient as well as creating the painless procedure of making transcutaneous access. Sensors can be placed in the subcutaneous tissue or in blood vessels [12,13,17]. But long term reliability requirements are high and not yet met. The limited life time and inadequate stability of glucose biosensors are the major sources of unreliability. An interesting approach is the open microflow concept [18]. When biosensors are implanted under the skin the tissue environment still contains


19

significant quantities of proteins which may alter long term performance of the sensor. In the technique described a very slow flow (30 µl/ h) of a sterile fluid is passed over the implanted needle’s sensor surface. The needles were surrounded by a plastic cannula and supplied with open microflow perfusion fluid. Glucose was monitored for four hours in volunteers and gave good blood glucose concentrations (r2 = 0.97). Transcutaneous devices as, e.g., sensors and micro-dialysis or ultra-filtration probes coupled to a sensor have moderate stability requirements since replacement is easier and therefore can take place every 1–3 days. Currently three commercially available continuous glucose monitoring devices are on the market, the MiniMed-Medtronic CGMS [19,20,21], the GlucoWatch biographer from Cygnus [22,23], and the GlucoDay from Menarini Diagnostics [24]. The non-invasive GlucoWatch biographer uses reverse iontophoresis to extract glucose through intact skin. The disposable part of the device comprises two sets of biosensor and iontophoresis electrodes which are also used as biosensor counter electrodes. The glucose extracted at the iontophoretic cathode is sampled in hydrogel discs containing the adsorbed biosensor enzyme glucose oxidase. After a sampling time all the hydrogen peroxide produced is consumed (see also the glucose oxidase reaction in section 1.2). The glucose biosensor has to be calibrated using a finger prick glucose determination. The monitoring with intervals of ten minutes is not really continuous, in addition a considerable amount of the data is discarded by the instrument. The MiniMed continuous glucose monitoring system (CGMS) uses a needle type of amperometric glucose oxidase based biosensor coated with bio-compatible membranes to report glucose levels every 5 minutes in the subcutaneous tissue. It can be operated for 3 days. For calibration of the biosensor at least four blood glucose readings per day should be entered into the measurement system. Like the GlucoWatch, the CGMS is FDA approved, but the glucose readings are only stored and used by a health care professional for diabetes management for retrospective analysis [20]. Menarinis GlucoDay utilizes a subcutaneous hollow micro-dialysis fibre connected to an external amperometric glucose biosensor. Isotonic fluid is pumped through the micro-dialysis fibre where it picks up glucose from the surrounding interstitial fluid and finally passes the glucose sensor flow through cell. The device is recalibrated daily with a one point calibration. Those devices are already a big step forward in the management of diabetes as compared to the 3–6 times self testing per day. Glucose level dynamics, and accordingly the quality of the diabetes management practiced can be accessed, and also the severe threat of mainly nocturnal hypoglycaemic events is moderated. Nevertheless, all approaches still suffer

20


from a limited reliability of the reported glucose values, thus preventing the use of those values directly for the appropriate dosing of insulin or of glucose intake. The reasons are, e.g., sensor drift, sensor lag time, and generally a loss of correlation with glucose blood levels. Amongst a needle type of electrochemical glucose biosensor [12-15] Roche Diagnostics is also working on a micro-dialysis based system using a viscometric measurement principle [25]. Reliable non-invasive glucose measurement would be highly appreciated [4]. Unfortunately optical methods using infrared measurement are not very specific for glucose. Therefore much effort is made in signal processing using mathematical models to extract the glucose signals from the up to five orders of magnitude higher signals from, e.g., water, as described in [26]. In this paper Malin et al., used near-infrared diffuse reflectance spectroscopy at the arms of diabetic subjects with a mean error from 3% up to 17%, dependent on the individual.

3.1

Multi-analyte measurement

Simultaneous and continuous concentration measurement of more than one parameter provides additional information about the metabolic condition of patients or probands. In medical applications, but also for metabolism monitoring of cell cultures small sample consumption by the analytical system is an absolut requirement. One way of reducing the sample volume is miniaturization of the device, the other possibility is integration of several sensors in a common shared measurement chamber. It is obvious that combining both strategies would give the best results. On the other hand, an integration of small-dimensioned biosensors causes well known problems such as chemical cross talk between adjacent amperometric biosensors [14,15]. Suzuki et al. [14] compared the chemical cross talk between glucose and lactate biosensors prepared to give three different types of membrane biosensors. The first biosensor preparation was done by glutaraldehyde cross-linking of gelatine and enzymes, the second by enzyme entrapment during electro-polymerisation of pyrrole, and the third by using a photolithographic PVA–SbQ method. Palmisano et al. [15] describe integrated glucose/lactate biosensors on dual platinum disk electrodes modified by an electro polymerised polypyrrole film and covered by membranes of glucose oxidase or lactate oxidase cross-linked with glutaraldehyde. Both authors used relatively large commercial dual electrodes of comparable dimensions (7 mm²) and applied the enzyme solutions manually. Whilst Palmisano describes his system as being free from cross talk owing to whole design and the FIA (flow injection analysis) measurement method which would sweep


21

away generated H2O2, Suzuki finds with a similar measurement method cross talk for all bi-biosensors. In an other approach screen printed disposable ring and split-disk plastic film carbon electrodes were coated with osmium poly(vinylpyridine) redox polymer horseradish peroxidase (Os-gel-HRP) and used for immobilization of glucose oxidase and lactate oxidase by glutaraldehyde cross-linking [16]. Chemical cross talk between the biosensors, especially at low analyte concentration, is reported. Perdomo et al. [11] describe a system in which a glucose and a lactate biosensor are located in pyramid-like cavities micromachined in silicon and covered by anodic ally bonded glass. The biosensors showed low cross talk which is explained by the special geometry of so called containment sensors. And Sangodkar presents in his paper [27] a biosensor array consisting of the enzymes glucose oxidase, urease, and on polyaniline. The analyte concentrations were quantified by conductance measurement. But only in the case of glucose oxidase the authors did succeed in an immobilization of the enzyme by entrapment in the electropolymer, while lipase and urease lost their activities in the acid monomer solution. Therefore lipase and urease were immobilized onto the ready polyaniline by adsorption. Although the wafer was diced before polyaniline deposition and enzyme immobilization this approach has some potential for mass production after a modification of the equipment. It is obvious, but can also be seen in the references selected, that most applications require flow through devices. Therefore an integration of the liquid handling environment with the sensors should be an objective target because it reduces effort in assembly, minimizes internal volumes, and increases reliability of the measuring system. One way of realizing such small devices is the use of integrated chemoand biosensors implemented in a micro-flow system. Biosensor arrays for continuously monitoring glutamine, lactate, and glucose concentrations in mammalian cell cultivations or for ex vivo measurement in human metabolic monitoring were developed by the author (Moser 2002). This BioMEMS will be described later in more detail (chapter 3.4).

3.2

Micro-dialysis

Micro-dialysis is a technique for investigating extra cellular liquids in various tissues. It is based on a molecular diffusion process across a semipermeable membrane. In practical applications sterilized hollow fibres of a dialysis membrane are implanted, and, by tubing connections to a precision pump, perfused at a low flow rate with an isotonic fluid. The dialysate is

22


then transferred to a flow-through cell outside the body equipped with an appropriate sensor for the analyte (see Fig. 2-1).

Figure 2-1. Scheme of micro-dialysis probe construction and operation.

The gradient of the analyte concentration across the membrane is the driving force of the diffusion process. Provided that a constant flow rate of the dialysis fluid can be achieved, a linear relationship exists between the glucose concentration in the perfusate and the concentration gradient. The so called ‘recovery’ was defined as the analyte concentration in the dialysate divided by that in the undisturbed tissue [28]. The recovery naturally depends on the flow rate: the lower the flow rate, the higher the recovery. The in vitro flow dependence of the glucose biosensor signal with and without a micro-dialysis probe in shown in Fig. 2-2. It also depends on the size of the membrane and on the medium: for example, Lönnroth et al. [29] found a 50% decrease of the in vivo recovery compared with the in vitro recovery. To circumvent the problems associated with an unpredictable recovery, the ‘no net flux’ technique was developed. It is based on the following assumptions: if the probe is perfused with several solutions containing different concentrations of glucose, a linear relationship exists between the net increase of the glucose concentration in the perfusate over that in the inlet of the tubing. The concentration of glucose in the probe not resulting in any net influx of glucose in the perfusate can be calculated by regression analysis. At this concentration the gradient across the membrane is zero.


0.6

100%

0.5 I [nA/0.25mm²]

23

99% 90%

0.4

75%

0.3 0.2

52%

0.1 0 0

2

4

6

flow rate [µl/min] Figure 2-2. Flow rate dependence of sensor with (lower line) and without micro-dialysis probe (upper line).

According to Lönnroth et al. [29], even at large concentration gradients the recovery remained constant. They concluded that this indicates that no tissue depletion of glucose had occurred and that a steady state condition was maintained. Otherwise the glucose concentration would have been underestimated. Some contradicting observations, however, were attributed to local drainage of the analyte, which is still considered one of the major limitations of this technique: according to Benviste et al. [30], Bungay et al. [31] or Meyerhoff et al. [32], there is considerable local drainage of the analyte, leading to a concentration gradient in the surroundings of the microdialysis probe. This effect must then lead to the expected underestimation of the concentration in the tissue. The recirculating method by Sternberg et al. [33] is worth mentioning. They implanted two probes into the subcutaneous tissue. In one probe the buffer solution was circulated until the glucose concentration finally was equal to that of the surrounding tissue. Thereby the ‘absolute’ glucose concentration could be estimated by measurement in the ‘recirculate’. This value then served as a calibration aid for the other, openly driven, probe. The continuous ultra-slow micro-dialysis was developed for the following reasons: Two drawbacks still existing were: 1) time-consuming calibrations and calculations owed to unknown partial recovery of the

24


analyte; and 2) the abovementioned unknown underestimation owed to local analyte depletion. By reducing the flow rate considerably (down to 100–300 nl/min) a virtually 100% recovery was obtained, according to Kaptein et al. [34]. At this flow rate the drainage of the surrounding tissue was minimized to such an extent that a steady state could be assumed. The disadvantage of very low flow rates is the long time required for equilibration. It might be possible to overlook rapid changes and as a result to react too late to extreme glucose values if such a technique is relied upon [35]. An elegant calibration technique has been presented by Hoss et al. [36]. By taking advantage of the flow-dependency of the recovery, a perfusion with a physiological saline solution containing 5.5 mM glucose at a relatively high flow rate served as an internal standard. The same solution was also perfused at a low flow rate to perform the dialysis of glucose. Rosdahl et al. [37] used different flow rates to identify at what rate the interstitial fluid completely equilibrates with the perfusion fluid.

3.3

BioMEMS for clinical monotoring

Recently the miniaturization of sensors shows remarkable progress. Boehm et al. [38] described a device that involves the fabrication of a micro-machined micro-dialysis probe connector that incorporates a miniaturized sensor using standard silicon wafer processing. Rhemrev-Boom et al. [9] reports on a new biosensor which works at a sub-microliter level that can be coupled to a micro-dialysis probe. Petrou et al. [39] present the fabrication of a micro-device suitable for continuous sampling and analysis of glucose. The device is equipped with a micro-dialysis probe integrated on the flow through channel of a microfabricated electrochemical sensor. The aim of this work was to integrate a dialysis probe on a micro-fabricated sensor device in order to obtain a complete micro system suitable for continuous monitoring of glucose. For this purpose a specific dialysis probe was designed and a PCB was appropriately converted to support both the sensor chip and the dialysis probe. A polyacrilonitrile fiber of 0.34 mm diameter similar to those contained in an artificial kidney was used as dialysis probe. The membrane cut off value was 50 kda. The perfusion fluid probe is shown in Fig. 2-3. The perfusion fluid was pumped to the probe through a polyimide tube placed above the fluid inlet. In the dialysis probe the perfusion fluid exchanged substances with the surrounding medium and then the dialysate was driven by the stainless steel tube to the sensor flow channel where the fluid outlet hole was situated. The insertion of the polyimide tube between


25

the fluid inlet and the dialysis probe provided the ability to vary the length of the whole probe without changing the dialysis active surface. The polyacrilonitrile fiber is the part of the device that will be introduced in the human body and it was selected for reasons of biocompatibility.

Figure 2-3. Scheme of the microfabricted micro-dialysis probe with integrated biosensor array.

A printed circuit board (PCB) comprising the conductive pads, the plug for the potentiostat, and the gold counter electrode was used for the assembly of the biosensor device. The flow cell, as well as a 4 mm long and 0.2 mm wide channel extending from the left end of the flow channel to the fluid inlet hole, were created on the PCB using a dry film resist insulation layer. Another wider channel (approximately 0.5 mm) was milled onto the PCB from the fluid inlet hole to the PCB edge. Finally, the holes for fluid inlet and outlet were drilled on the PCB. A 4 cm long stainless steel tube was inserted through the 0.2 mm wide channel to the flow cell and fixed on the board with adhesive. Subsequently, the sensor chip was assembled using a conductive adhesive and the flow channel was sealed also with adhesive. The volume of the flow channel created by this way was 0.36 µl (5.5×0.7×0.094 mm3). A polyimide tube of about 2.5 cm length was then placed around the stainless steel tube and driven through the 0.5 mm wide channel to the fluid inlet hole. The polyimide tube was fixed in that position and the channel was sealed by an appropriately cut glass chip bonded with adhesive. The final step for the preparation of the device was the assembly of the micro-dialysis tube. The tube was placed around the stainless steel tube up to the end of the polyimide tube and was fixed there with adhesive.

26


The other end of the tube was closed with adhesive so that the distance between the end of the stainless steel and the micro-dialysis tube was about 0.5 mm. The total length of the active micro-dialysis probe was 10 mm. The miniaturized device was used in vitro. High sampling efficiency of the micro-dialysis probe was achieved by appropriate selection of the perfusion fluid flow rate. Response times varying from 1.5 to 3.0 min were determined for flow rates ranging between 1 and 0.2 µl/ min. The linear response range was up to 30 mM glucose and interference from other electro-active substances was almost negligible. The device showed excellent stability under continuous operation for at least 5 days and variation of sensitivity less than 3% over a period of 15 days.

3.4

Multi-parameter monitoring

Bioanalytical micro systems in hybrid- technology not only fulfil lab-ona-chip applications but also include the advantages of micro systems; namely fluidic and electric connections are at same level. Nano-liter volumes of flow through cells for on-line measurement, static mixers for bio-assays, and capillaries for heterogeneous quasi-continuous assays such as immuno assays can be realized. By using one raster a combination of two or more micro systems is possible. The flexibility of the design enables hybrid modular systems by combining modules of different life cycle times. So electronic parts can be used several times, whilst, e.g., parts which were in contact with body fluids of different patients can be changed. Although biocomponents on micro systems are said to be the most unstable part, very often the problems are caused by the opening of glued or other bonded connections. Thus a modular system gives the possibility to replace easily parts which became obsolete. The bioanalytical microsystem (BioMEMS) presented measures simultaneously and continuously glucose, lactate, glutamine, and glutamate in whole blood, blood serum, and fermentation broth [40,41,42,43]. The biosensor array consists of five amperometric enzyme membrane sensors, a temperature sensor, and one silver/silver chloride reference electrode (see Fig. 2-4). The immobilization of the biocomponents is either done in a photolithography process (see Fig. 2-5) or by nano-dispensing and subsequent UV exposure. One thin film device consists of five platinum working electrodes of a diameter of 400 µm and a Ag/AgCl reference electrode made on a 0.3 mm glass carrier. The Ti–Pt–Ti metallization deposited by means of high vacuum evaporation is patterned with a lift off process and insulated with silicon nitride. A five-micron thick silver layer is galvanically deposited from a bright silver bath onto the reference electrodes which are electrically


27

Figure 2-4. Micrograph of the biosensor array .

interconnected on the wafer level, and subsequently galvanically chlorininated in 0.1 M KCl. The working electrodes which are interconnected to electrically, are galvanically platinized in a 2% H2PtCl6 solution and subsequently covered with a semi-permeable membrane deposited by electro-polymerisation from a 3 mM 1,3-diaminobenzene solution in 0.1 M pH 7 phosphate buffer [42].

Figure 2-5. Scheme of biosensor integration by subsequent lithography of enzyme-hydrogel membranes

28


The respective oxidases like glucose oxidase, lactate oxidase, glutamate oxidase, and glutamate oxidase together with glutaminase were immobilized onto the working electrodes in a photo-cross-linked pHEMA membrane and additionally covered by a diffusion-limiting and a catalase-containing top layer (see Fig. 2-8). Deposition of these membranes was done by automated PC controlled dispensing of the photo-reactive membrane precursor solutions with a home made nano-liter dispenser attached to a CNC machine. Cross-linking of the precursors was done by UV flood light exposure under argon flushing. Although spin-coating and photo-patterning of the membranes is possible [10] and increases the packaging density of the sensors, owing to the additional costs caused by the labour intensive additional mask aligner exposures and the waste of expensive enzyme in the spin on process, the spin coating–photo-patterning route is justified only for special applications such as, for example, in vivo sensors. The dispensing route gains additional attractiveness by its high flexibility. Devices with different membrane configurations can be realized even on a single wafer. Together with the cheap masks used in the PCB manufacturing, small series of custom made devices can be produced cost effectively. The PCB which provides a mixed fluidic-electric platform for the assembly of the biosensor array, is made from standard materials with standard equipment. Subsequent lamination of dry film resist layers over previously patterned layers allows the creation of closed channels without the use of any sacrificial layer (Fig. 2-6). In this way, a PCB with a sensor flow chamber also comprising a large gold counter electrode, a mixing coil, and electrical leads and pads is created.

a

b

c

Figure 2-6. Scheme of dry film resist based process for micro-fluidics. a) lamination of dry film resist, b) exposure, c) development. Repetition of this process allows creating 3-D closed channels.


29

Assembling with the biosensor array results in the bioanalytical microsystem with a flow cell volume of 150 nl and a mixing coil volume of 1 µl. The advantage of the system is that small portions of reagents and samples can be brought together, mixed, and, finally, the biochemical response can be measured. The functional density of BioMEMS is increased by the addition of active fluidic components. The Printed Circuit Board (PCB) for the assembly of the thin film biosensor array comprises conducting pads for the thin film device and the plug connection to the potentiostat, a gold counter electrode, drilled through holes for liquid inlet and outlet, and a photo-patterned spacer made from the dry film resist used for insulation of the PCB (see Fig. 2-7). Another design has two inlets, and a static mixer for performing bio-assays on chip (see Fig. 2-12). Small capillaries are used for immobilizing biocomponents like antibodies for immuno assays. The entire processing was done with conventional PCB technology equipment.

Figure 2-7. Micrograph of the micro-fluidics PCB shown together with a biosensor array chip which is flapped down for assembly.

30


The biosensors with no crosstalking and high long term stability were produced by modifying electrochemical transducers and utilizing photopatternable enzyme membranes. The use of appropriate miniaturization technology leads to mass producible devices for in vivo and ex vivo applications such as in whole blood or in fermentation broth.

Figure 2-8. Scheme of biosensor multiple membrane set up. Main transport and reaction pathways indicated for a glucose sensor.

Because of a glutaminase with an activity optimum in the neutral pH range, direct and simultaneous monitoring of glutamine together with glucose, lactate, and glutamate could be performed [40]. The electrochemical measuring set up consisted of a home made SMD five working electrode–potentiostat operated in the ‘three electrode mode’ linked to a PC based data acquisition and actuating program with an ADC card. Since this device is made from two components made by means of well established mass production technologies (thin film and printed circuit board technology) and assembling of the parts is to some extent compatible to IC packaging techniques, cost effective mass fabrication of this device seems realistic.


31

Performance of the microsystem

I [nA]

100

50

0 0

20

40

substrate [mM] Figure 2-9. Calibration graphs of a glucose–lactate–glutamine and glutamate multi-analyte device. Circles: glucose; squares: L-lactate; crosses: glutamine; triangles: glutamate.

Fig. 2-9 shows the calibration graph of a biosensor array. Owing to the diffusion limiting spacer membranes the linear range of all biosensors fits the demands for on-line which means without sample dilution or changing pH steps. All biosensors within an array have with hydrogen peroxide oxidation the same detection molecule, therefore crosstalking between the electrodes should be expected. But signal independence is secured by the catalase top layer which destroys hydrogen peroxide within the membrane that would otherwise leak out into the solution. The second benefit of the catalase top layer is an extremely low flow rate dependence of the sensor signal. Fig. 2-2 shows as an example the unchanged response of a glucose biosensor under variation of flow rate. The lifetime of a biosensor array is always limited by the biosensor the enzyme of which loses its activity first. Compared with the stable enzymes glucose oxidase and glutamate oxidase, lactate oxidase is the Achilles‘heel of a biosensor array. But owing to the overloading of enzyme activity in the membranes and especially by the stabilizing effect of pHEMA even the lactate biosensor can be operated continuously in bovine serum for one month at 37°C [40].

32


3.5

Applications

3.5.1

Monitoring of glucose and lactate with a micro-dialysis probe

Simultaneous real time monitoring of glucose and lactate in the interstitial fluid of subcutaneous tissue gives excellent insight into tissue metabolism, which is of interest for metabolism research and for glucose monitoring of diabetes patients [44]. For the amperometric measurement a home made 4 channel potentiostat was used. The pulsing CMA 107 pump was selected because it is dedicated to clinical micro-dialysis. The inlet of the micro-dialysis catheter was connected to the micro-dialysis pump, while the outlet was connected to the microsystem (see Fig. 2-11). The pulsing of the pump had no influence on the sensor signals because owing to their especial membranes system the biosensors were not flow dependent. A laptop completed the measurement equipment. With this portable system we monitored 8 male volunteers at a flow rate of 0.3 µl min-1 for 6 hours. After application of the micro-dialysis catheter (abdominal) with each of the healthy volunteers a glucose tolerance test (2OGTT) was performed. The dialysate was collected from the outlet of the micro system for additional analysis with the CMA 600 analyzer which was used as the reference method with r 2 = 0.93 (8 experiments, glucose) and r2 = 0.84 (4 experiments, lactate simultaneous with glucose). Additionally this sampling procedure provided the possibility of a gravimetric flow measurement. In all experiments glucose and lactate blood concentrations were measured. Fig. 2-10 shows a typical result of a simultaneous glucose and lactate monitoring. The delay because of the internal volume of the measurement system was 15 min., the curves were corrected for it. The delay between the glucose maximum and the lactate maximum (40 min.) provides information about metabolic processes in the same way the comparison of blood glucose with s.c. glucose concentration does. Whilst the shape of the curves is comparable for all volunteers the relative recovery of glucose correlates negatively with the body mass index (BMI) of the volunteers.


33

Figure 2-10. Simultaneous glucose and lactate monitoring on healthy human volunteer using the micro-dialysis sensor coupling shown in Fig. 2-11.

Figure 2-11. Commercial CMA 70 micro-dialysis probe connected to the glucose–lactate monitoring device.


34

As it can be proved by checking the dialysate of 15 min. pools with the CMA 600 analyzer the continuous monitoring done with the bioanalytical microsystem is very reliable. Comparing blood glucose levels with the corresponding tissue concentration versus time and versus body mass index, significant differences between lean and adipose persons can be seen. This differences can be caused by differences in tissue perfusion and metabolism.

Figure 2-12. Photograph of bioMEMS with integrated static mixer. Right from the glucose– lactate–glutamte biosensor chip the mixing structure can be seen.

3.5.2

Ammonia monitoring

Hyperammonemia always indicates a severe damage of the liver parenchyma, except for patients with hereditary metabolic disorders associated with deficiency or inhibition of urea cycle enzymes who also have high concentrations of ammonia in their blood. For such metabolic disorders patients, mostly children, rapid and easy handling of the determination of the ammonia concentration is as important as the glucose measurement for diabetes patients.

Ammonia Assay For assays on chip a micro system with two inlets and a static mixer with an internal volume of 1 µl was used (see Fig. 2-12). Through the two inlets sample and reagent is pumped into the device and mixed by the mixer. According to schema 1 2-oxoglutarate, NADH, and ammonium of the sample are converted by glutamate dehydrogenase to glutamate. The formation of glutamate is monitored with the glutamate biosensor until the endpoint of reaction is reached. Pumping sample and substrate solution is


35

done by means of two computer controlled syringe pumps. Also on-chip calibration of biosensors was performed with the same configuration by automatically diluting one stock solution. Fig. 2-13 shows a calibration graph of an ammonium assay. The glutamate dehydrogenase activity in the reaction solution was 5 U/l. Because of the small reaction volume higher enzyme activities can be used cost effectively. In the future an immobilization of the enzyme is planned. Scheme 1: 2-oxoglutarate +NADH + H+ + NH4+ ' NAD+ + L-glutamate + H2O

Calibration of Ammonia with the Glutamate sensor 4

current [µA /[cm2]

3

2

1

0 0

50

100

150

200

250

300

concentration of Ammonia[ [µM]

Figure 2-13. Calibration graph of an amperometric ammonia assay with 5U/l glutamate dehydrogenase. Scheme 1 shows the reactions on which the assay is based.

36


4.

CONCLUSIONS AND OUTLOOK

Currently devices for glucose measurement allowing continuous monitoring of clinical parameters for the first time come into the market. At the company Therasense a miniaturized needle sensor measuring in the interstitial fluid is under development. Minimed–Medtronic is working on a long term (1–2 year) implanted sensor which measures glucose in a central vein of the heart. An implanted near infrared optical sensor measuring glucose directly in the blood of a vein is under research at the company Animas. Because this optical sensor is implanted the problems arising from the transcutaneous sensing should be avoided. Wireless sensor data transmission is now available because of the technological advance reached in this field, and will be inserted in monitoring BioMEMS in near future. There is a large gap between the stage of development of the glucose devices and the development of non-glucose biosensors. One reason is the unavailability of bio components as stable as glucose oxidase. Even in research real bianalyte or more analyte integrated devices are rarely described, although in metabolism research each additional parameter measured provides valuable information. Additionally these measurements should preferably be performed in vivo and in the tissues of interest.

5. [1] [2] [3] [4] [5] [6] [7] [8] [9]

REFERENCES Clark, L.C. Jnr., (1962), Ann. NY Acad. Sci. 102, 2945. Turner, A.P.F., Karube, I. and Wilson, G.S., (1987), ‘Biosensors: Fundamentals and Applications’, Oxford University Press, Oxford, 770. Eggins B.R., (2002), ‘Chemical Sensors and Biosensors’, Wiley. Narayanaswamy, Ramaier; Wolfbeis, Otto S. (ed.), (2004), ‘Optical Sensors, Industrial, Environmental and Diagnostic Applications’, Springer Series on Chemical Sensors and Biosensors. Nakamura H., Karube I., (2003), ‘Current research activity in biosensors’, Anal BioAnal. Chem., 377(3), 446-68. Wilson, G.S., Gifford, R. (2005). ‘Biosensors for real-time in vivo measurements’, Biosens. Bioelectron., 20, 2388. Ek S., Diabetes Care, (2004), ‘Beyond Meters and Strips’, R& D Day at www.roche.com. Bakker E., (2004), ‘Electrochemical Sensors’, Anal. Chem., 3285-3298. Rhemrev-Boom M.M., Jonker M.A., Venema K., Jobst G., Tiessen R., Korf J., (2001), ‘On-line continuous monitoring of glucose or lactate by ultraslow micro-dialysis combined with a flow-through nanoliter biosensor based on poly(m-phenylenediamine) ultra-thin polymer membrane as enzyme electrode’, Analyst, 126, 1073-1079.


37

[10] Jobst G., Moser I., Varahram M., Svasek P., Aschauer E., Trajanoski Z., Wach P., Kotanko P., Skrabal F., Urban G., (1996), ‘Thin film Microbiosensors for Glucose-Lactate Monitoring’, Anal. Chem., 68, 3173-3179. [11] Perdomo J., Hinkers H., Sundermeier C., Seifert W., Martinez Morell O., Knoll M., (2000), ‘Miniaturized real-time monitoring system for L-lactate and glucose using micro-fabricated multi-enzyme sensors’, Biosens. Bioelectron. 15, 515-522. [12] Gough D.A., Armour J.C., (1995), ‘Development of the implantable glucose sensor. What are the prospects and why is it taking so long?’, Diabetes, 44, 1005-1009. [13] Updike S.J., Shults M.C., Gilligan B.J., Rhodes R.K., (2000), ‘A subcutaneous glucose sensor with improved longevity, dynamic range and stability of calibration’, Diabetes Care, 23, 208-214. [14] Suzuki M., Akaguma H., (2000), ‘Chemical crosstalk in flow-type integrated enzyme sensors’, Sensors and Actuators B 64, 136-141. [15] Palmisano F., Rizzi R., Centonze D., Zambonin P.G., (2000), ‘Simultaneous monitoring of glucose and lactate by an interference and crosstalk free dual amperometric biosensor based on electropolymerized thin films’, Biosens. Bioelectron., 15, 531-539. [16] Osborne P.G., Niwa O., Yamamoto K., (1998), ‘Plastic film carbon electrodes: Enzymatic modification for On-line, contineous, and simultaneous measurement of lactate and glucose using micro-dialysis sampling’, Anal. Chem., 70, 1701-1706. [17] Robert J.J., (2002), ‘Continuous monitoring of blood glucose’, Hormone Research, 57, 81-84. [18] Rigby G.P., Ahmed S., Horsman G., Vadgama P., (1999), ‘In vivo glucose monitoring with open microflow influences of fluid composition and preliminary evaluation in man’, Analytica Chimica Acta, 385, 23. [19] Gross T.M., Bode B.W., Einhorn D., Kayne D.M., Reed J.H., White N.H., Mastrototaro J.J., (2000), ‘Performance and evaluation of the MiniMed Continuous Glucose Monitoring System during patient home use’, Diabetes Technol Ther 2, 49-56. [20] Mastrototaro J., (1999), ‘The MiniMed Continuous Glucose Monitoring System (CGMS)’, Journal of Pediatric Endocrinology & Metabolism, 12, 751-758. [21] Guerci B., Floriot M., Böhme P., Durain D., Benichou M., Jellimann S., Drouin P., (2003), ‘Performance of Continuous Glucose Monitoring System (CGMS)’, Diabetes Care, 26, 582 [22] Tierney M.J., Tamada J.A., Potts R.O., Jovanovich L., Garg S., and the Cygnus Research Team, (2001), ‘Clinical Evaluation of the GlucoWatch® Biographer: A Continual, Non-invasive Glucose Monitor for Patients with Diabetes’, Biosensors and Bioelectronics 16, 621-629. [23] Kulu E., Tamada J.A., Reach G., Potts R.O., Lesho M.J., (2003), ‘Physiological differences between interstitial glucose and blood glucose measured in human subjects’, Diabetes Care, 26, 2405.

38


[24] Maran A. et al., (2002), ‘Continuous subcutaneous glucose monitoring in diabetic patients, Diabetes Care, 25, 347. [25] Beyer U., Schafer D., Thomas A., Aulich H., Haueter U., Reihl B., Ehwald R., (2001), ‘Recording of subcutaneous glucose dynamics by a viscometric affinty sensor’, Diabetologia, 44, 416. [26] MaLin, S.F., Ruchti T.L., Blank T.B., Thennadil S.N., Monfre S.L., (1999), ‘Noninvasive Prediction of Glucose by Near-Infrared Diffuse Reflectance Spectroscopy’, Clinical Chemistry, 45, 1651-1658. [27] Sangodkar H., Sukeerthi S., Srinivasa R.S., Lai R., Contractor A.Q., (1996), ‘A biosensor array based on polyaniline’, Anal. Chem., 68, 779-783. [28] Ungerstedt U., Herrera-Marschitz J., Jungelins U., Stahle J., Tossman U., Zetterström U., (1982), ‘Dopamine synaptic mechanisms reflected in studies combining behavioural recordings and brain dialysis’, Adv. Dopamine Res., 37, 219-231. [29] Lönnroth P., Jansson P.A., Smith, U., (1987), ‘A micro-dialysis method allowing characterization of intercellular water space in humans’, Am. J. Physiol., 253, E228-E231. [30] Benviste H., (1989), ‘Brain micro-dialysis’, J. Neurochem., 52, 16671679. [31] Bungay P.M., Morrison P.F., Dedrick R.F., (1990), ‘Steady state theory for quantitative micro-dialysis of solutes and water in vivo and in vitro’, Life Sciences, 46, 105-119. [32] Meyerhoff C., Mennel F.J., Bischof F., Sternberg F., Pfeiffer E.F., (1994), ‘Combination of micro-dialysis and glucose sensor for continuous on line measurement of the subcutaneous glucose concentration: theory and practical application’, Horm. Metab. Res., 26, 538-543. [33] Sternberg F., Meyerhoff C., Mennel F.J., Hoß U., Mayer H., Bischof F., Pfeiffer E.F., (1994), ‘Calibration Problems of subcutaneous glucosensors when applied ‘in-situ‘ in man’, Horm. Metab. Res., 26, 523-525. [34] Kaptein W.A., Zwaagstra J.J., Venema K., Korf J., (1998), ‘Contiuous ultraslow micro-dialysis and ultrafiltration for subcutaneous sampling as demonstrated by glucose and lactate measurements in rats’, Anal. Chem., 70, 4696-4700. [35] Summers L.K.M., Clark M.L., Humphreys S.M., Bugler J., Frayn K.N., (1999), ‘The use of micro-dialysis to monitor rapid changes in glucose concentration’, Horm. Metab. Res, 31, 424-428. [36] Hoss U., Kalatz B., Gessler R., Pfleiderer H.J., Andreis E., Rutschmann M., Rinne H., Schoemaker M., Haug C., Fussgaenger R.D., (2001), ‘A novel method for continuous online glucose monitoring in humans: the comparative micro-dialysis technique’, Diabetes Technol. Ther., 3(2), 237-243. [37] Rosdahl H., Hamrin K., Ungerstedt U., Henriksson J., (1998), ‘Metabolite levels in human skeletal muscle and adipose tissue studied with micro-dialysis at low perfusion flow’, Endocrinol. Metab., 274 (5), E936-E945.


39

[38] Boehm S., Olthuis W., Bergveld, P., (2000), ‘A micro-machined double lumen micro-dialysis probe connector with incorporated sensor for on-line sampling’, Sens. Actuat. B., 63, 201/208. [39] Petrou P.S., Moser I., Jobst G., (2002), ‘BioMEMS device with integrated micro-dialysis probe and biosensor array’, Biosensors and Bioelectronics, 17, 859- 865. [40] Moser I., Jobst G., Urban G.A., (2002), ‘Biosensor Arrays for Simultaneous Measurement of Glucose, Lactate, Glutamate, and Glutamine’, Biosensors & Bioelectronics, 17/4, 297-302. [41] Moser I., Jobst G., Aschauer E., Svasek P., Varahram M., Urban G., Zanin V., Tjoutrina A., Zharikova A., Berezov T., (1995), ‘Miniaturized Thin film Glutamate and Glutamine Biosensors’, Biosensors & Bioelectronics 10, 527-532. [42] Jobst G., Svasek P., Moser I., Varahram M., Trajanoski Z., Wach P., Kotanko P., Skrabal F., Urban G., (1997), ‘Mass producible miniaturized flow through device with biosensor array’, Sensors & Actuators B., 43, 121-125. [43] Jobst G., Moser I., Urban G.A., (1999), ‘Simultaneous arterial-, venous-, and subcutaneous metabolite monitoring with highly stable biosensors’, Proceedings Transducers 99, Sendai, 2, 1712-13. [44] Freaney R., McShane A., Keavney T.V., McKenna M., Rabenstein K., Scheller F.W., Pfeiffer D., Urban G., Moser I., Jobst G., Manz, A., Verpoorte E., Widmer H.M., Diamond D., Dempsey E., Saez de Viteri F. J., Smyth M. R., (1997), ‘Novel instrumentation for real-time monitoring using miniaturized flow systems with integrated biosensors’, Ann. Clin. Biochem., 34, 291-302.

‘

Chapter 3 NEURAL IMPLANTS IN CLINICAL PRACTICE Interfacing neurons for neuro-modulation, limb control, and to restore vision–Part I Thomas Stieglitz1, Joerg-Uwe Meyer2 1

Laboratory for Biomedical Microsystems, IMTEK-Institute for Microsystem Technology, University of Freiburg, Freiburg, Germany 2 Draegerwerk AG, Luebeck, Germany

Abstract:

Neural implants interface parts of the nervous system and technical systems to partially restore sensor and motor functions that have got lost due to trauma or diseases. Electrodes act as transducers to record neural signals or to excite neural cells with means of electrical stimulation. m The field of neural prostheses has grown over the last decades. In this chapter the reader is guided through some basics of nerve excitation and electrical stimulation. An overview of neural prostheses in clinical practice illustrates the opportunities and limitations of the implants and their performance in their current size and complexity.

Key words:

neural prostheses, vision prosthesis, retina implant, sieve electrode, nerve interface, functional electrical stimulation, recording; CNS, PNS, BCI

1.

INTRODUCTION TO NEURAL IMPLANTS

The use of electrical phenomena for curing diseases has been known for more than 2,000 years. In ancient times Roman physicians applied current by using electrical fish for alleviating pain caused by gout or rheumatism [1]. It took many centuries to obtain a scientific understanding of the bioelectric phenomena in the body, the physiological mechanisms behind our senses and movements, and the pathophysiological alterations in diseases and after trauma in spinal cord injury, in hearing or vision loss. Up to the present, most mechanisms can be scientifically described but there is still a challenge to transfer this knowledge into successful therapies and


41

42

NEURAL IMPLANTS IN CLINICAL PRACTICE C

rehabilitation devices which help disabled persons in the activities of their daily life. In Germany an estimated number of 80,000 persons is suffering from spinal cord injury (SCI) with 1,800 new injuries per year [2]. In the USA estimations number 40 million persons, with 10,600 new cases per year [3]. Only 100 years ago patients usually did not survive a spinal cord injury for more than 6 months in Europe. It was Sir Ludwig Guttmann who established modern rehabilitation medicine in the first quarter of the last century at Stoke Mandeville (England). With his tremendous contributions life expectancy of SCI patients increased, and improved social and job-related reintegration became feasible [4]. Nowadays life expectancy of SCI patients is similar to non–SCI subjects [3]. In diseases that affect sensory senses patient numbers are comparative levels. In Europe only, 600,000 persons suffer from incurable diseases of the retina, such as retinitis pigmentosa or macula degeneration, which finally lead to blindness. Numbers of other diseases are even higher. For some of them neurological rehabilitation offers possibilities of restoring lost functions to some limited extend by means of technical prostheses and implants [5,6]. However, many medical, biological, and technical considerations have to be taken into account. An interface has to be set up that allows exchanging the amount of information that is necessary for transforming the intention of the human subject into a correct technical command and to give back some information from the technical device to the user. During the last 35 years, many ‘success stories’ have been written, but implanted prostheses often remain in the shadows [7]. Existing clinical devices are often unknown to many physicians and patients, and health insurance companies have to be convinced to reimburse the costs for each and every case. Spinal Cord Injury (SCI) has devastating consequences for the person involved. He or she mustt learn a new way of life after the hospitalisation and rehabilitation phase. Activities of daily life (ADL) are more or less limited. Medical and social costs are high [3] but an adequate ‘expensive’ neural prosthesis might help the patient to participate in daily social and economic life again and decrease the long term costs in the end. But unobserved from most of the media, and owing to tremendous progress in microelectronics, micro-engineering and the material sciences within the last three decades, the so called neural prostheses have been transferred into clinical practice (Table 1). More than 200,000 devices have been implanted into patients [8].

NEURAL IMPLANTS IN CLINICAL PRACTICE

43

Table 3-1. Summary of fully implanted neural prostheses with more than 1,000 implanted systems [8] . Applications Number off implants Spinal cord stimulator to treat >130,000 intractable pain and motor disorders Auditory nerve stimulator to >55,000 restore hearing (cochlear implant) Deep brain stimulator for tremor, Parkinson‘s disease, and pain Vagal nerve stimulator to treat intractable epilepsy Sacral nerve stimulator for urinary urge incontinence, urinary retention, pelvic pain, and faecal incontinence Sacral nerve stimulator for bladder emptying Phrenic nerve stimulator for respiration

>20,000

Manufacturers Advanced Neuromodulation Systems, USA; Medtronic, USA Advanced Bionics, USA; AllHear, USA; Cochlear, Australia; MED-EL, Austria; MXM lab, France Medtronic, USA

>17,000

Cyberonics, USA

>10,000

Medtronic, USA

>2,500

Finetech Medical, UK

>1,600

Avery Laboratories, USA; Atrotech, Finland; MedImplant, Austria

The most prominent example is the spinal cord stimulator. It has been implanted in more than 130,000 cases to treat intractable chronic pain and to manage spasticity [9,10]. Cochlear implants restore hearing after loss of the outer hair cells when the acoustic nerve is still intact. They have been implanted in more than 55,000 cases and are very successful in clinical practice [11,12], as well for adults who lose hearing, and for children or babies born deaf. The youngest patient was implanted at the age of six months to support the development of the auditory cortex. The range of hearing understanding with the CI varies from patient to patient, but many persons are able to use the telephone again to communicate. An implant for deep brain stimulation to suppress tremor and to overcome dyskinesia in Parkinson‘s disease [13] has been implanted more than 20,000 times. Vagal nerve stimulation to treat intractable epilepsy is a relatively new application, but more than 17,000 systems have been implanted [14]. Other applications include stimulators for urinary urge incontinence [15,16,17] that are a quite important development, particularly with regard to the demographic development in the Western hemisphere in which 10% of all persons above the age of 60 develop incontinence. Sacral nerve stimulators for bladder management [18,19] are at the lower end of patient numbers, but have much

44


helped SCI persons to participate in social life again. Phrenic pacemakers for patients with high lesions of the spinal cord stimulate the phrenic nerve to restore ventilation and make people independent from an extra-corporal ventilation system [20,21]. The function of these neural prostheses is identical in all applications. The technical system interfaces the biological system by electrodes. By them bioelectrical signals can be recorded to obtain information from human sensors, and electrical stimulation can be used to activate the target organ. So far implants are mainly limited to the necessary condition of not harming the body, acting non–toxically and not degrading. The requirement of a functional, long term stable nerve interface is not yet accomplished completely. Non-invasive monitoring of a nerve, and an electrode that delivers sophisticated information is not available in clinical practice. However, information about the amount of growth of tissue encapsulating the implant during the healing phase, about nerve status, and the success and compatibility of the chosen stimulation parameters would help to optimise the performance of neural prostheses significantly. Clinically approved neural prostheses are characterized by their robustness for a long lifetime. Therefore they are made by means of precision mechanics which limits the functionality and nerve selectivity owing to its degree of miniaturization. Micro system technology opens new opportunities for complex systems with a high number of electrodes to interface the nervous system. In the following sections the fundamentals of nerve anatomy and excitation will be introduced and neural implants in clinical practice will be presented in detail. The opportunities and challenges of micro system technology will be discussed on the basis of several development pathways for miniaturized neural implants: interfaces for capturing regenerating nerves; substrates for establishing biohybrid assemblies; implants for restoring vision; and miniaturized implantable brain–machine interfaces give an insight into future applications within the emerging field of neural prostheses. For those who are interested in the details of neural prostheses’ theory and application, the comprehensive textbook ‘Neuroprosthetics’ [22] is recommended for further reading.


2.

45

ANATOMICAL AND BIOPHYSICAL FUNDAMENTALS

Since the time of Galvani und Volta1, physicians have been fascinated with curing diseases by means of electricity and with restoring functions of paralyzed persons. Many early experiments arose great public interest, but nowadays seem to have been close to charlatanism. The discipline of functional electrical stimulation (FES) has been developed on a scientific basis only for the last 45 years [23]. The following section introduces the basics of the origin and conduction mechanisms of nerve (action) potentials and describes the physical mechanisms during the extra–cellular electrical stimulation of nerves according to Frank Rattay [24]. Since the adaptation of micro-structures to the nerves is the major objective for a long term stable implant, anatomical and physiologic characteristics of (peripheral) nerves will be introduced with respect to the design and development of neural interfaces.

2.1

Peripheral Nerve Anatomy

The peripheral nerve is a complex biological micro system with a microanatomy that is composed of different tissues. The nerve fibres are long processes from the related nerve cell body which is located in the ganglion of the posterior spinal root for sensory neurons, and in the anterior horn of the spinal cord for motor neurons, respectively (Fig. 3-1). The nerve fibres are grouped in so called fascicles. Different layers of connective tissue (epineurium, perineurium, endoneurium) and the nerve fibres form the nerve stem (Fig. 3-2). The epineurium is a loose network of connective tissue that is located in between the fascicles and in superficial areas of the nerve. A mechanically stable sheath of connective tissue, the perineurium, encircles each fascicle. The connective tissue within the fascicles that surrounds the nerve fibres is the endoneurium. The number of fascicles and the amount of the epineural connective tissue varies between different nerves and along the course of each nerve. The amount of epineural tissue is increased if a nerve passes through a joint or bones, in order to protect them against compression or friction.

1

Alessandro Volta (1745–1827), Italian physicist; Luigi Galvani (1737–1798), Italian physician and natural scientist.

46

NEURAL IMPLANTS IN CLINICAL P PR RACTIC CE

Figure 3-1. Schematic view of sensory and motor components of a peripheral nerve [25].

Figure 3-2. Micro-anatomy of a peripheral nerve [25]. epi: epineurium; p: perineurium; end: endoneurium; ax: axon; my: myelin sheath; nR: node of Ranvier; Schw: Schwann cell.

The normal function of the nerve fibres is based on a continuous and sufficient supply of oxygen and nutrients. This supply is provided by a complex intra–neural microvascular system. Its structure allows the adaptation to the physiological movements of the nerve in the limbs. Oedema resulting from traumatic incidents might compress these blood vessels and finally cause acute or chronic nerve damage as a result of a lack in the supply of oxygen or nutrients. Every nerve fibre in a peripheral nerve is an appendage of a cell that is located in the central nervous system (CNS),


47

i.e., in the spinal ganglion or in the spinal horn. Larger fibres are normally myelinated, i.e., Schwann cells are wrapped around the fibres of the myelin sheath. The area between two Schwann cells is called the node of Ranvier (Fig. 3-2). The diameter of myelinated nerve fibres varies from 5–20 µm. Thick fibres are more sensitive to compression than thin fibres because of a higher deformation at a given pressure.

2.2

Mechanisms of Peripheral Nerve Damage

Protection mechanisms in peripheral nerves safeguard a normal structure and the functioning of nerves if a mechanical deformation does not exceed certain limits. If nerve damage occurs all components of the peripheral nerve—connective tissue, nerve fibres and blood vessels—might be affected. They react in different ways and play a different role in the pathophysiology of nerve damage. Nerve fibres react with local de-myelinization or Wallerian degeneration after a rupture. Damage of intra–neural vessels results in inadequate blood flow, leading to ischemia or edema. Chronic edemas finally cause intra–neural scar tissue. The fascicle topography, the amount of epineural connective tissue, the anatomical course of the nerve, and many other factors influence the amount of nerve damage that results in reversible or irreversible neurological defects, respectively (Table 2). Table 3-2. Classification of nerve damage according to Seddon and Sunderland [26]. classification according to Seddon Neuropraxia

classification according to Sunderland injury 1stt grade Axonotmesis injury 2nd grade Injury 3rdd grade Neurotmesis Injury 4th grade Injury 5th grade

nerve damage

Conduction block due to local compression; complete recovery within weeks to months Discontinuity of nerve fibres, integrity of Schwann cells and endoneural components; complete recovery is possible Loss of axonal continuity at intact perineurium; irreversible damage, neurological deficits Rupture of perineurium; irreversible damage, neurological deficits Complete injury of the nerve; irreversible damage, neurological deficits

A surgical preparation of a peripheral nerve, e.g., for electrode implantation, always endangers the normal structure and function. A nerve mobilization with dissection of segmental blood vessels over some centimeters seems feasible [25] and might be required for some electrode types. However, the invasiveness of electrodes should be taken into consideration with respect to the necessary degree of selectivity for the

48


application envisaged. Mechanical stretching of peripheral nerves is a further source of nerve damage. Normally nerves are able to adapt to a certain degree. In a physiological state a tensile strength can be observed that leads to a shortening of about 10–20% of their original length after dissection. Therefore a slight stretching of the nerve for cuff electrode implantation is tolerable. Another common source of nerve injury is some damage owed to pressure, e.g., after implantation of circum-neural electrodes. At an increased pressure of 20–50 mm Hg a reduced blood flow in epineural veins, an obstruction of the axonal transport mechanism, and epineural edemas alters the nerve conduction up to a complete block. After an irreversible damage Wallerian degeneration occurs, which leads to a physical disintegration and chemical degradation of the axon and the myelin sheath. The Schwann cells occupy the endoneural tube of the nerve stump. The fibres on the proximal nerve side undergo retrograde degeneration. The regeneration of dissected nerves varies. Neuroma formation on the proximal nerve stump often causes pain phenomena such as a phantom pain after an amputation trauma.

2.3

Excitability of Nerves

In 1868 Bernstein developed the hypothesis that living cells consist of an electrolytic interior that is surrounded by a membrane with relatively low permeability for ions. Different ionic concentrations on both sides of the membrane cause an electric potential difference (Table 3). In resting conditions the inside of the cell is more negative than the extra–cellular space. This state is called ‘polarized’. If nerve or muscle cells are activated from this resting condition they are ‘depolarised’, i.e., the potential rises by up to 50 mV and decreases again, with a short undershoot of the resting potential (‘hyper-polarization’), in general. Cell membranes consist of a bimolecular lipid layer with hydrophilic and lipophilic properties in their outer and inner regions, respectively. Ion channels are integrated in the membrane. They connect extra–cellular and intra–cellular space and are ion selective with a certain permeability. The resting potential of the cell can be calculated under these conditions by the Goldmann equation (Eq. 1). Additional ion pumps in the membrane transport ions, by the consumption of metabolic energy, against the electrochemical gradient from the intra– to the extra–cellular space, and vice versa.


49

Table 3-3. Ion concentration of a mammal muscle cell [27]. intra–cellular Na+ K+ Ca2+ ClHCO3large Anionsresting potential

Em =

12 mmol/l 155 mmol/l 10-8-10-7 mol/l 4 mmol/l 8 mmol/l 155 mmol/l - 90 mV

P [ RT ⋅ ln K F PK [

]e ]i

[ ]e ]i Na [ Na

extra-cellular Na+ K+ Ca2+ ClHCO3unrestricted mobility

[ ]i ]e Cl [ Cl

145 mmol/l 4 mmol/l 2 mmol/l 120 mmol/l 27 mmol/l

,

(Eq. 1)

PX : conductivity of ion X, [X]i/e : concentration of ion X intra / extra–cellular, R : gas constant, F : Faraday constant, T : temperature. The membrane resistance itself is quite high (1010 Ωcm). In the case of a nerve excitation the sodium channels open first, sodium ions flow into the cell and the electrical potential over the membrane decreases. In a secondary reaction the potassium channels open and cause a flux of potassium ions out of the cell. As a result the increase of the intra–cellular electrical potential is halted and the resting potential is re-established.

2.4

Electrical Modelling of the Nerve Membrane

Hodgkin and Huxley described the first complete mathematical model of nerve excitation in 1952 by four differential equations. It was the basis for the electrical model of the nerve membrane (Fig. 3-1). It described the conductivity of the ion channels as variable resistors depending on potential and time. The equilibrium potentials of the ions were modelled as voltage sources. Capacitors and insulation resistance represent the properties of the passive membrane elements. An additional voltage source considers the effects of the remaining ions. Different intra– and extra–cellular potentials Vi and Ve result in a transmembrane potential V = Vi - Ve that drives a current (Eq. 2) with a capacitive (IC) and a ionic portion (Iion),


50

I

Ic

I Ion

C⋅

dV V + dt R (

)

(Eq. 2)

Figure 3-3. Electrical model of a cell membrane element.

The remaining three Hodgkin-Huxley differential equations describe the gating process of the membrane channels as dynamic process depending on the different ion conductivities with respect to the membrane potential. The basic equations from Hodgkin and Huxley have been persistently refined over the last 40 years and adapted to the dynamics of mammalian nerves. The latest models allow detailed simulations, e.g., chronaxie-rheobase relationships and nerve blocks [28]. The details of passive nerve membrane properties are described in a comprehensive theoretical survey of J.J. Strujk [29].

2.5

Propagation of Action Potentials

The model introduced for nerve excitation not only describes the origin of the action potential but also matches the description of the propagation of the nerve action potential. The change of the membrane voltage can also be predicted if a current is ‘injected’ artificially, e.g., by a stimulation electrode. The segmentation into small cylindrical elements ∆x considers the real dimensions of the nerve fibre. Owing to leakage currents within nerve fibres, the voltage also becomes a function of the distance. The model matches unmyelinated as well as myelinated nerve fibres (Fig. 3-2) where it is the internodal distance [30]. In myelinated nerve fibres the nerve action propagation occurs saltatorically, i.e., the excitation jumps from one node of Ranvier to the next. Recharging of the nerve cell membrane only takes place at the nodes of Ranvier and as a result increases the velocity of nerve conduction with respect to unmyelinated nerve fibres.


51

Figure 3-4. Electrical network for simulating the currents within a nerve fibre [30].

The thickness of the nerve fibre has a major impact on the nerve conduction velocity. The intra–cellular conductivity Ga (Eq. 3), the membrane conductivity Gm (Eq. 4) and the membrane capacity Cm (Eq. 5) increase with the nerve fibre thickness and contribute to a higher nerve conduction velocity [23].

Ga =

π⋅d 2 4 ⋅ ρ i ⋅ ∆x

(Eq. 3)

Gm

gm

d L

(Eq. 4)

Cm

cm π d L

(Eq.5)

ρi : specific resitivity of the intra–cellular space,

cm: gm: d: L: ∆x:

membrane capacity per unit area, membrane conductivity per unit area, nerve fibre thickness, width of a node of Ranvier, internodal distance.


52

In clinical practice nerve fibres were classified according to their thickness and their conduction velocity (Table 4). Table 3-4. Nerve fibre classification [27]. Classification according to Erlanger/ r Gasser and Lloyd/ d/Hunt Aα I

II Aβ Aγγ Aδ

III

B C

IV

2.6

Function

Mean fibre diameter (µm)

Nerve conduction velocity (m/s)

primary muscle spindle afferents, motor fibres to skeletal muscles and tendon organ afferents skin mechanoreceptors skin afferents for touch, and pressure motor fibres for muscle spindles skin afferents for temperature, and nociception /deep pressure sensibility of the muscle sympathetic preganglionic skin afferents for nociception, sympathetic post-ganglionic efferents

13–15

70–120

9 8 5 2 functions) C = Communication D = Dual (P + S) (telemetry data exchange) R = Rate modulation (adaptation of space rate to activation signal) 0 = None 0 = None

0 = None

V Antiarrhythmia function P = Pacing (antitachy-antiarrhythmia) S = Shock

The international NBG code (NASPE/BPEG generic Pacemaker Code) describes the function of a pacemaker with the help of a maximum of five letters (Table 5) whereas the first three are mandatory and the fourth and fifth are optional [59].

3.3

Implantable Defibrillators

The most frequent reason for death in Western industrial countries is the sudden cardiac death (SCD); it accounts for 1,200 deaths per day in the USA. SCD is caused by ventricular tachycardia or fibrillation and death occurs within minutes. The only possibility for overcoming the tachycardia and to induce a regular heartbeat is to defibrillate the heart. Implantable Cardioverter-defibrillators (ICD) were developed in the 1980s, and more than 25,000 devices have been implanted worldwide up to the mid 1990s. The ICD are similar to a standard pacemaker but have more powerful batteries and circuitry for stimulation. They detect fibrillation of the myocard and defibrillate immediately. In the case of tachycardia in the atrium the method of choice is called cardioversion. The implant detects the R-spike of the electrocardiogram (ECG) and the stimulation is triggered with

60


a delay of 20 ms after the R-spike to prevent a short fibrillation of the ventricle myocard.

3.4

Cochlea Implants

Cochlea implants were developed to restore hearing after loss of the cochlea’s outer hair cells when the inner hair cells and the pathways to the auditory cortex are still intact [11]. The implants are well established in clinical practice [62] after a long development time of more than 25 years. They have a wide acceptance in adults with hearing experience and the latest implantations also report success in small children with no prior hearing experience. Worldwide, more than 60,000 persons use the implants to communicate [63]. The system consists of an extra-corporal microphone, a speech processor and a telemetry link that powers the implant and transmits the data for stimulation patterns and the choice of electrodes [12]. The implant’s electronics decodes the data to complex analogue stimulation patterns. A ceramic housing is implanted in the bone of the skull behind the ear. An electrode array with 2 to 22 platinum electrodes and an extracochlear counter electrode are connected via small wires to the implant. The electrode array and the cables are moulded into silicone rubber. No connectors between implant electronics and electrodes are present in this kind of implant. Different electrode designs were developed with the objective of coming as close as possible to the nerve cells for better spatial selectivity. These modiolus electrodes are subject of intensive research [64,65,66,67]. The power supply is realized by an inductive link because of the relatively high energy demand; the latest research aims at using rechargeable batteries for fully implantable systems without requiring external components. The housings have mostly been made of aluminium oxide ceramics to ensure a high data rate and a good efficacy, which is not possible with a titanium housing owing to its damping properties. If titanium housings are used the coil is placed outside the housing to have an optimal coupling efficiency for the inductive link. Normally the strategy of the extracorporal speech processor contributes to a major part to the understanding of free speech [68,69,70]. The major competitors in the field are MedEl (Austria), Cochlear Corp. (Australia), Advanced Bionics (USA) and MXM (France). Even though implants should last for decades, sometimes failure might occur or an upgrade seems to be necessary. Revisions in cochlear implants are feasible [71] and deliver the same quality of speech perception.


3.5

61

Phrenic Pacemakers

The induction of inspiration by means of electrical stimulation to resuscitate persons with a high level spinal cord injury is more than 100 years old. Stimuli were applied with body–surface electrodes for the ‘electroventilation’ technique [72]. These electrodes were placed bilaterally at the base of the neck, over the phrenic nerve motor points, or anteriorly to the axillae to stimulate the long thoracic and phrenic nerves. With any of these placements of electrodes inspired volumes in excess of spontaneous tidal volumes can be achieved in man, baboon, and dog. In the 1970ies the first implantable phrenic pacemakers for patients with high lesions of the spinal cord stimulate were described [73]. Stimulation of the phrenic nerve was used to restore ventilation and make people independent from an extra-corporal ventilation system [20,21]. The available systems for clinical application from Medimplant (Austria), Atrotech (Finland) and Avery (USA) differ in details of the surgical access of the phrenic nerve, the electrode design and stimulus protocols. Worldwide, more than 1,600 patients have been supplied with phrenic pacemaker implants up to 2004 [8]. Electrodes were placed around different branches of the phrenic nerve or several electrodes were randomly selected at the main branch, respectively, to prevent fatigue of the diaphragm [35] during long-term electrical stimulation.

3.6

Grasp Neuroprostheses

Neural prostheses of the upper limb have been developed to restore grasp in paraplegia after stroke and quadriplegia after spinal cord injury. Miniaturized single channel stimulators with a diameter of 2 mm and a length of 16 mm have recently been introduced for application in the upper limb. They were implanted by means of a biopsy needle. Inside of a glass tube, they have a rechargeable battery or an inductive link for the power supply, and data transfer for simulation data. Currently their application is proposed after paraplegia to prevent shoulder subluxation owed to muscle atrophy [74]. High lesions in spinal cord injury cause a loss of hand and arm function. If the lesion side is in the range of the 5th and 6th cervical vertebrae, elbow flexion and shoulder movement is feasible but the hand itself is paralysed. These patients may benefit from different neural prostheses. The Handmasterr (Ness, Israel) is an external orthosis-like system with integrated stimulation electrodes. It restores grasp in chronic stroke patients [75] as well as in quadriplegics [76]. The Bionic Glove has been developed at the


62

University of Alberta (Canada) as a garment-like device [77]. Clinical trials with quadriplegics led to an increased functionality of grasp [78,79]. Donning of the Handmaster is solved by a simple closure mechanism which patients can use on their own, but help is needed for the Bionik Glove. A notable system for paralysed patients was developed at the Case Western Reserve University nearly 30 years ago. Two different grasps for small and large objects have been implemented [80], respectively: the lateral and the palmar grasp (cylinder/pinch grip). The so called Freehand system (Fig. 3-7) has been commercialised by Neurocontrol, Inc with the FDA approval in 1997 and the CE mark in Europe in 1999. The implant is controlled and powered via an inductive link. The titanium housing with the implant electronics [81] is connected to a transmission coil and eight leads which lead to small connectors. Eight epimysial electrodes have to be placed on the motor points of arm muscles in order to induce lateral and palmar grasp in the patient. A quite selective stimulation is obtained by the neuro muscular approach, but tendon muscle transfers owed to the complex traumata of the patients often accompany a surgical intervention. Worldwide 270 patients have been implanted [2]. Despite clinical success and high consumer satisfaction, the system has been withdrawn from the market for economic reasons.

Shoulder Position Sensor

Electrode Leads Electrodes Implanted Stimulator Transmitting Coil

External Controller

Figure 3-7. Schematic view of the Freehand System (Neurocontrol, USA), a neural prosthesis for grasp in quadriplegic persons.

Approximately 12,200 Americans will be candidates for the FES neuroprosthetic hand grasp system under the current research protocols.


63

With both the expansion of current protocols to other diagnostic categories and further research and development, application of this neuroprosthesis to a considerable number of previously excluded subjects is likely to be possible [82]. However, a commercially available system is currently neither available nor to be seen on the horizon.

3.7

Neuroprostheses for gait and posture

The functional rehabilitation of stance and gait has been the subject of intensive research for many decades. However, no neural prosthesis has been commercialised so far. On one hand, the wheelchair is a socially well accepted device for increasing mobility; on the other hand, technical solutions have only given solutions with a poor performance. The reason for the limitations is the complexity of the biomechanics and interaction of the muscles allowing non-paralysed persons to stand stably and balanced over a long time without any signs of fatigue. During electrical stimulation of the lower limb patients have to use a walking frame to carry a major portion of the body weight and to stay in balance and walk a few steps. Different systems have been developed over the decades. Neither orthoses [83] nor surface stimulation will be discussed here. One of the leading centres in this field is the Case Western Reserve University and the Veterans Affairs Hospital in Cleveland, Ohio (USA). There the first implantable systems with percutaneous intra–muscular electrodes were stable, safe, and effective during long term use of functional electrical stimulation (FES) for exercise, standing, and walking in paraplegic persons. The follow up of two patients for 17 years showed that the electrodes’ stability was only limited. An average of two electrodes had to be replaced every 6 months [84]. Modified percutaneous double helix electrodes have been proved to be more stable [85,86]. The most common problems were the daily care of the electrodes at their exit sites, frequent irritation (inflammation and infection) of the skin around the electrodes, that had to be treated with antibiotics, and replacement of failed electrodes. It could be summarized as this percutaneous system has the potential for short-term rehabilitation in individuals with incomplete paraplegia or stroke, but only limited applicability for chronic use in complete paraplegics [84]. The development of a 16-channel implant with epimysial electrodes was implanted in a 39 year old patient with T10 paraplegia to restore sitting to standing, walking, and exercise functions [87]. The 16 electrodes were split into two implant electronics that were powered and controlled via a radio link. After a training phase the stimulation over the electrodes allowed controlled movements for standing and walking over a period of months.

64


Figure 3-8. System view of a lower limb neural prosthesis (By courtesy of D. Guiraud, Univ. Montpellier, France).

Within the European ‘SUAW’ project (Stand Up And Walk, Fig. 3-8), a technological solution was developed [88]. A combination of epineural and epimysial electrodes was chosen to elicit smooth movements by adequate control algorithms [88] without the use of the withdrawal reflex. A completely different approach was investigated in the United Kingdom. An implant for enabling paraplegics to exercise on a tricycle with electrical stimulation was developed to improve muscle status, bone density and circulation. Electrodes were placed around the spinal roots in the lumbar and sacral spinal cord. The LARSI system (Lumbar Sacral Anterior Root Stimulator Implant) requires a laminectomy and the direct access to the spinal canal. This surgical intervention is primarily done to implant an electrical stimulation system for urinary bladder management [89]. An Australian research group worked on similar developments. The Nucleus 22-channel implantable hearing prosthesis (Cochlear Pty. Ltd., Lane Cove, Sydney, Australia) has been modified by computer programming into a functional electrical stimulator. Individual or multiple channels can be sequenced and adjusted for the activation of single and multiple nerves. Activation of sciatic nerve branches via helical electrodes in anesthetized rabbits produced single contractions or co-contraction at the ankle and simultaneous bilateral joint movements [90]. The stimulation parameters were adjusted on human nerves [91] and the system worked quite well in pilot studies. The modified Nucleus-22 is now no longer manufactured. Therefore further implantations were stopped.


65

The fast muscle fatigue during stance and gait caused by neuro muscular or neural electrical stimulation, as well as the relatively low performance has been little attractive to possible users so far. If developments from spinal cord micro stimulation in cats in Vivian Mushawars group at the University of Alberta, Canada (see section 7, Future Developments) were to be transferred to humans there is a hope of performing a quantum leap in the functionality of neural prostheses for lower limb stimulation for the benefit of paralysed persons.

3.8

Spinal Root Stimulator

One of the oldest neural prosthesis was developed for urinary bladder (and bowel) management in spinal cord injured persons (Fig. 3-9a). The implantation is combined with a laminectomy and dorsal rhizotomy to open the reflex circles which lead to a spastic bladder. The intervention restores the storage capacity of the bladder and gives back continence to the patient. The prototype was introduced in 1972 by G.S. Brindley [92] contacting the anterior sacral spinal roots. The electrodes consist of a silicone rubber block with slots (Fig. 3-9 b). In each slot three platinum foils were embedded as electrodes. The middle electrode serves as cathode whilst the outer electrodes are used as anodes. In principle each electrode can be addressed independently, but in practice the two anodes were short circuited and serve, along with the intermediate cathode, as one stimulation channel.

Figure 3-9. The Sacral Anterior Root Stimulator Implant (SARSI), Finetech, Inc. (UK).

66


The spinal roots were placed in the slots comparable to a bookmark laid in between the pages of a book, thereby giving the device the name of ‘book electrode’. The slots were covered with a slit made of silicone rubber and fixed with silicone glue. The first micturition caused by electrical nerve root stimulation was reported in 1973 [93]. Cooper cables establish the connection between the electrodes and an implantable control unit. Their helical design of the cable gives good mechanical strength at high flexibility [94]. After more than 10 years of development the Brindley group were able to report the first 50 cases of a radio frequency controlled implant from batch production with medical approval called a sacral anterior root simulator (SARS) in humans [95]. The implant was commercialised as the Finetech–Brindley stimulator, and less than a decade later 500 had patients benefited from the system [18]. The system has been successfully transferred into clinical practice [96], even though improvements are desirable for overcoming the sphincter–detrusor dyssynergia that causes a post stimulus voiding. Nowadays there are 2,000 patients worldwide who carry an urinary bladder stimulator [8]. Besides the intradural surgical implantation, an epidural implantation procedure was developed [97] to enlarge the group of patients who might benefit from this kind of neural prosthesis, e.g., patients with unstable lumbar spinal cord, arachnoiditis, or oily myelography. The implantation was mostly combined with a deafferentation of the posterior sacral roots to interrupt reflex circles and spasms in order to restore a residual volume of the urinary bladder and give back continence. The disadvantages of this surgical procedure include the loss of reflectoric defecation and of reflectoric erection and ejaculation in male patients. It could be shown that the SARS could be also used to restore male sexual or reproductive function [98,99]. Even after a long clinical experience with excellent results, the electrodes seem to be very bulky, even though very robust from a technical point of view. There is a high demand for miniaturization, above all with regard to minimally invasive surgery, because operations on the open vertebral column are still highly risky.

3.9

Drop Foot Stimulator

The drop foot stimulator is one of the oldest neural prostheses in clinical practice. It was invented for patients left paraplegic after a stroke by Liberson in 1961 [100]. In his investigations surface electrical stimulation excited the peroneal nerve and induced foot flexion in the swing phase of the gait. The control signals were obtained from a switch in the heel of a shoe (Fig. 3-10 a). Today there are several approaches worldwide [101], e.g., from Finetech, Inc., in the United Kingdom and from Neurodan in Denmark (Fig. 3-10 b). One of the successful external devices is the Oddstock


67

Dropped Foot Stimulator (ODFS) from the Salisbury District Hospital that has been proved to improve paraplegic gait after a stroke [102]. Latest developments work with implanted nerve electrodes. The approach of the Center of Sensory Motor Control (SMI, Aalborg, Denmark) tries to rid devices of the external sensor component by recording nerve signals from natural sensors in the foot [103] to trigger the stimulation. The implants were powered via an inductive link to ensure a long lifetime at high stimulation frequencies. All control circuitry is placed extra-corporally.

Figure 3-10. The drop foot stimulator: from the first concept to an implantable neural prosthesis.

A second approach for the correction of drop foot is performed with injectable single channel stimulators, the so called BIONS. Compared with surface stimulation of the common peroneal nerve, they provide more selective activation of specific muscles [104].

3.10

Neuro-modulation

Electrical stimulation of ascending (afferent) pathways to modulate nerve activity is called neuro-modulation. It is based on the gate control theory for pain that was proposed from Melzack and Wall in 1965. It states that a mechanism in the dorsal horns of the spinal cord acts like a gate that inhibits

68


or facilitates transmission from the body to the brain on the basis of the action of peripheral fibres as well as the dynamic action of the brain processes. The gate control theory, however, is not able to explain several chronic pain problems, such as phantom limb pain which require a greater understanding of brain mechanisms. Therefore more sophisticated theories are under investigation [105]. However, since it was introduced in 1967 the stimulation of the spinal cord has been evolved to be the most successful application of neural prostheses. The methodology and technology of spinal cord stimulation for the management of chronic intractable pain have evolved continuously [106]. To overcome dissatisfactory pain paresthesia with single electrode implants new approaches were developed during the late 1990s that attempted to selectively cover one or more dermatomes with paresthesia as well as to provide sequential stimulation of different anatomical sites. These approaches have been applied both intra–spinally and extra-spinally by stimulating either the spinal nerves or the dorsal columns. Today more than 130,000 implants help patients to treat intractable chronic pain [8]. A second group of patients which benefits from spinal cord or sacral root stimulation suffer from urinary incontinence [15]. More than 10,000 patients have already been treated. The implant technology is directly derived from state of the art pacemakers, with Medtronic (USA) and Advanced Neuro-modulation Systems (USA) being the main manufacturers.

3.11

Deep Brain Stimulation

Deep brain stimulation has become a clinical method of curing some of the clinical syndromes of Parkinson’s disease. After a bilateral stereotactical implantation of a needle-like electrode in brain’s regions of the basal ganglia and thalamic or subthalamic nuclei, chronic tremor is suppressed and dyskinesia is overcome in many patients [13]. The application of deep brain stimulation is a good example for modifying already existing components to introduce an implant system for completely new applications. The implants were derived from cardiac pacemaker technology. They consist of a Titanium housing with stimulator electronics and a battery power supply to be implanted in the chest area. The electrodes are connected with the implant via subcutaneous cables with plugs, one in the brain region and another near the implant. The system can be programmed via a telemetry link and can be switched on and off by the patient via a magnet that is moved over the body region of the implant. Up to now more than 20,000 patients have been implanted [8].


69

Figure 3-11. Activa Deep Brain Stimulation Therapy, Medtronic.

After years of clinical experience the (psychiatric) side effects of this therapy have become obvious, including depression, mania, aggressions, and deficits in language [107]. For their own safety some patient groups should be excluded from the therapy [108]. The implantation of a DBS seems not to interfere with pre-implanted cardiac pacemakers if the receivers for telemetry programming had been placed at a distance of more than 15 cm (6 in) from each other [109]. For long term performance and the balance of benefits [110,111] and side effects, stimulus parameters have to be carefully adjusted [112].

3.12

Vagal Nerve Stimulation

Stimulation of the vagal nerve was introduced into clinical practice in 1997 with the neurocybernetic prosthesis system (NCP) from Cybernetics, Inc. (Houston, Texas, USA). An implantable multi-programmable pulse generator delivers constant current electrical signals. The signals are delivered on a predetermined schedule, or may be initiated by the patient by using an external magnet. The device is implanted in a subcutaneous pocket in the chest just below the clavicle, similarly to placement of a pacemaker. The stimulation signal is transmitted from the prosthesis to the (left) cervical vagus nerve through a lead connected to an electrode that is a multi-turn silicone helix, with a platinum band on the inside of one radical turn [113]. The system was grown out of pacemaker technology and does not represent

70


a technological breakthrough. However, the stimulation of the vagal nerve is targeted in many diseases because of many diffuse projections of the vagal nerve to the mid-brain and cortical regions. So far more than 17,000 NCPs have been implanted [8] to treat intractable epilepsy in adults [114,115] as well as in children [116]. Other proposed applications [117,118] include depression [119], pain, eating disorders (e.g., adipositas), migraine, dementia, and Parkinson’s disorders. Even though most results are tremendously promising, side effects have also been present. Daytime sleepiness owing to apneas and arousals associated with the stimulation [120]. Heart rate variability may occur after long term stimulation [121] and pain thresholds can be decreased [122]. However, the advantages mainly prevail as increasing numbers of implantated patients are showing.

4.

REFERENCES References are listed at the end of chapter 4.

Chapter 4 BIOMEDICAL MICRODEVICES FOR NEURAL IMPLANTS Interfacing neurons for neuromodulation, limb control and to restore vision–Part II Thomas Stieglitz1, Joerg-Uwe Meyer2 Affiliation 1 Laboratory for Biomedical Microsystems, IMTEK-Institute for Microsystem Technology, University of Freiburg, Freiburg, Germany 2 Draegerwerk AG, Luebeck, Germany

Abstract:

Neural implants interface parts of the nervous system and technical (micro-) systems to partially restore sensor and motor functions that have got lost owing to trauma or diseases. The implementation of micro-machining methods in neural implants opened new application fields but also new design paradigms and approaches with respect to the bio-stability of passivation and housing concepts and electrode interfaces. Micro system–specific applications like sieve electrodes for investigations into regenerating nerve stumps, vision prostheses, and central nervous system interfaces show the variety of applications and discuss the challenges in biomedical micro systems for chronic nerve interfaces in new and emerging research fields that bridge neuro–scientific disciplines with material science and engineering.

Key words:

neural prostheses, biomedical micro-devices, vision prosthesis, retina implant, sieve electrode, nerve interface, functional electrical stimulation, recording; CNS, PNS, BCI

1.

THE CHALLENGE OF MICRO-IMPLANTS

The most important biological requirement for a (micro-) implant is the proof of its bio-compatibility. This means in detail that the implant materials are non-toxic to biological cells and that the material surfaces must have no, or only slight, influence on cell growth and cell proliferation. The materials have to be biostable, i.e., no corrosion of metals and degradation of


71

72

BIOMEDICAL MICRODEVICES FOR NEURAL IMPLANTS T

anorganic or organic materials should occur during chronic implantation. Furthermore, the material properties and the shape of the implant should deliver good structural bio-compatibility, i.e., no damage of the surrounding tissue must be induced. Therefore the implants have to be mechanically compatible with to a low weight and high implant flexibility. Last but not least, neural implants have to be long term functional. They have to establish a bio-electrical interface for a bi-directional signal transmission, which means the recording of neural signals and a stimulation of nerve fibres or cell bodies. This interface is the transducer for selective and functional nerve and muscle excitation, and, ideally, allows the integration into a closed loop control with the patient in the loop. The design and development of miniaturized implants requires a detailed knowledge of the biological requirements and the possibilities and limitations of existing technologies. During the design phase of an ‘optimal’ neural implant, target specifications have to be harmonized with the biological requirements and anatomical constraints as well as with technical needs, material properties, and design rules from given process technologies. Implants in clinical practice must have FDA approval or the CE mark. The long experience over decades with helically wound cables, standardized plugs, titanium and ceramic housings for microelectronic components and the use of medically approved materials such as special silicone rubbers allows the speeding up of the time to market and of fixing the development costs at a reasonable cost by component combination without intensive material evaluation procedures. However, the approval procedure of a neural implant is in any case a long procedure. Therefore many new implants rely on the experience from pacemakers and neuro-modulation [8] and combine established components and materials for new products, e.g., silicone rubber based cuff electrodes and a ceramic housing. These devices are robust, stable, and reliable but suffer from a low spatial resolution and low selectivity of the electrodes, as well as from having a quite large size. If anatomy restricts the implant’s size and a higher number of electrodes seems to be mandatory for adequate performance, e.g., for vision prostheses that stimulate the retina with an implant in the eye [123], micro system technology has to be chosen. It offers a high degree of miniaturization and methods of integrating systems of high complexity. The established materials and process technologies for their deposition and structuring, as well as hybrid and monolithic assembling and packaging technologies, offer tools for fabricating miniaturized systems. The protection of microelectronic components requires a housing that ensures a sufficient electrical insulation and protection against moisture and ions. With respect to long lifetime of the implants, a telemetry energy supply or implantable batteries need systems

BIOMEDICAL MICRODEVICES FOR NEURAL IMPLANTS

73

with low power electronics. Additionally, implant specific challenges have to be taken into account [49]. While silicone rubber and metal wires are the materials of choice for precision mechanics, silicon is the most widespread substrate materials in micro-machining. It has shown good bio-compatibility, as well as have the metals gold, platinum, and iridium [124]. First experiments with flexible polymers such as polyimide have been reported for different applications. Polyimides have shown good results with respect to bio-compatibility [125,126,127]. Applications in cardiology [128], at the fixation of artificial intra–ocular lenses [129] and for cochlea implant electrodes [130] proved the long term stability and functionality of the implants. However, polyimide based ribbon cables in electrically active implants under a permanent voltage bias have shown stability problems with respect to insulation resistance [131,132] using polyimides with a water uptake above 5%. Today polyimide is established as substrate material for electrode carriers, interconnects, and flexible carriers for hybrid electronic assemblies. Their water uptake has been reduced to values below 0.5%. First products are already in clinical trials. Batch processing ensures the highest reproducibility and precision for the substrate perimeter and geometry and the active electrode sites. Since small changes in the design need a complete redesign cycle, modifications are time and cost intensive. Silicon is a brittle material with low flexibility, if mechanical bulk material properties are taken into account. However, if the material is thinned down to 10 to 15 microns, it becomes flexible and allows bending with small curvatures. The process technology to do this needs several process steps. The final probe geometry has to be formed by a masking step of the silicon followed by deep boron diffusion. In succeeding process steps interconnect lines, electrodes, insulation layers, or even microelectronic components, can be integrated [133,134]. In the last step all silicon except the doped areas is removed in an EDP (ethylene diamine pyrocatechol) anisotropic wet etching process. The fabrication of silicon based interconnection cables was shown to be technologically feasible [131,132] but costs, technology, and limited cable length prohibited a widespread use in applications to neuroscience and neural prostheses. The connection to neural micro systems, to cables, or to powerful telemetry devices is still one of the major problems for clinical use. One advantage of silicon based micro-devices compared to polymer substrates is the ability of monolithic integration of electronic circuits to record and carry out signal processing [134,135]. The passivation and packaging of silicon based devices should be carefully considered for neural implants in order to ensure bio-stability and long term stable behaviour of electrodes with integrated electronic devices. If standard CMOS electronic processes were directly transferred for implant

74


passivation, degradation during implantation deteriorated not only the insulation layer but also functional integrated structures in only one year [136]. Silicon nitride and silicon oxide insulation layers that have been deposited by chemical vapour deposition (CVD) were also not stable as monolayers in a physiological environment [137,138] but were altered in water-soluble materials such as silica or ammonia [139]. Better stability was achieved when combining the layers to multi-layers [139,140,141]. Polyimide as a flexible substrate and insulating material for electrodes, interconnection cables, and printed circuit board (PCB) carriers for neural implants [142] showed good in vitro and in vivo bio-compatibility [126,127] with respect to toxicity as well as with respect to bio-stability [143,144] as cuff and regenerative type of electrodes. Parylene C has been known as an insulation layer of electronic components for implants for a long time [145]. Polymer coatings with parylene C as insulation and encapsulation material are under investigation in combination with a silicone rubber coating [146,147]. Even though first chronic results from pilot experiments showed that these combination layers were stable over more than one year in vivo [148], statistically significant numbers of devices in long term tests have to be performed before micro systems will gain a remarkable segment with this implant concept in clinical applications. In general, silicone rubber is still the material of choice for encapsulation of cables and also as additional coating for hermetically sealed electronic circuits [50]. The second crucial subject of micro-implants is the stability of the electrode materials since, once implanted, they should remain within the body of the patient for many years, and remain functional in order to fulfil the task of connecting man to machines. Thus it has to be resistant to corrosion during stimulation and to the attack of biological fluids, enzymes, and macrophages produced during the initial foreign body reaction. The electrodes have to be composed of inert materials, both passively and when subjected to electrical stimulation, since deterioration of the device may result in implant failure and the release of toxic products. Materials typically used are platinum, iridium, tungsten, and stainless steel as conductors, and silicone rubber, polytetrafluoroethylene, and polyimide as insulating carriers [49,149]. The highest requirements for the electrode materials include minimal energy consumption while stimulating, stable electrochemical characteristics, good phase boundary behaviour for polarization and afterpotentials, adjustable and stable impedance and frequency response, stability against artefacts and noises are made [150]. A great deal of work has been done in fundamental and applied research to investigate the feasibility of different concepts of neural interfaces with biomedical micro-implants. The target specifications of an ‘optimal’ neural interface cannot be stated in general, but strongly depend on the application.


75

The function and the implantation site of the implant play an important role for the design and the choice of technology. Additionally, the envisaged number of patients who will use the implants (and pay for them) will determine whether hybrid assembling techniques will be favoured, as is done for small series production, or whether the patient numbers justify the costs of a monolithic integration approach. The following sections try to give a comprehensive overview of motor and sensor neural prostheses that have mainly been made by using micro system technology. A more comprehensive view of neural prostheses, and the theoretical background of the underlying mechanisms of motor and sensor neural prostheses can be found in the NEUROPROSTHETICS textbook [22].

2.

VISION PROSTHESES

The restoration of sight after visual impairment and loss of the sense of vision is a particular challenge in neural prostheses. More than a million sensory cells, the rods and cones, act as transducers in a healthy retina. The light that is projected onto the retina is transduced into electrical potentials. They are transferred into pulses that were processed over the different cell layers within the retina. These pulses are then conducted along the optic nerve and the lateral geniculate nuclei to the visual cortex. Depending on disease or trauma, visual impairment and blindness is caused for different reasons which have to be taken into account when an implantation site and a neural prosthesis have to be chosen. In 1929 Foerster showed for the first time that the electrical stimulation of the visual cortex led to the perception of light sensations in the shape of points, so called phosphenes [151]. Nowadays research focuses mainly on the development of vision prostheses for patients with retinitis pigmentosa, a hereditary disease in which the sensory cells of the retina degenerate. The visual field decreases successively from the periphery, leading to tunnel vision, and finally to complete visual impairment (blindness). Even though the photo–receptors degenerate in retinitis pigmentosa, the ganglion cell layer, and at least partially some structures from the amakrine and bipolar cell layer, are preserved.

76


Figure 4-1. Anatomical sites for the implantation of a vision prosthesis.

Worldwide, the number of 1.5 million individuals suffer from retinitis pigmentosa. Other diseases which cause visual impairment up to complete vision loss include retinal degeneration by glaucoma and loss of visual pathways and relais stations owed to hereditary or cancerous nerve destruction. Probably those patients might also benefit from vision prostheses in the long term run. Depending on the disease or trauma, three different implantation sites (Fig. 4-1) are currently under investigation [152]: the retina; the optical nerve; and the visual cortex. Comprehensive summaries from different points of view give a deep insight into the vision prosthesis field that would exceed the objective of this chapter [153, 152,154]. The long term goal of all the vision prostheses’ research might be the restoration of ‘normal’ vision but the short term goal must be a spatio-temporal resolution of the implant so that the patient is able to navigate in an unknown environment. Otherwise, the implant will not be accepted and commercialised for a wide number of patients. First, the implants have to prove their applicability. Accurate rehabilitation assessment should be addressed to having objective measures for the comparison of different concepts and their ability to restore vision to some extent.


2.1

77

Cortical Vision Prostheses

Research for interfacing the visual cortex with implants has been conducted for more than 30 years. Giles S. Brindes was the first to prove the feasibility of a cortical vision prosthesis in human pilot implantations [155] with a wireless energy supply and data transmission [156]. He implanted an array of 80 radio receiver stimulators beneath the scalp of a blind nurse, which was widely critized [157] but opened the field for serious investigations about the feasibility of cortical vision prostheses. Finally, he stopped his experiments. Many years later a second concept with an epicortical electrode array having 64 electrodes, which was based on a silicone rubber foil, was comparable to the grid arrays that were used during epilepsy monitoring to locate the sources of the seizures. Dobelle and his group developed this system (Fig. 4-2) in which a camera (100×100 pixels) and a signal processing system on a portable computer outside the body are connected with the electrode array via 64 cables through the skin of the patient. The visual acuity of the system was estimated to 20/1200 [158,47], which only allows the recognition of light and dark but no detailed recognition of unknown objects in unknown environment. Up to April 2002, eight patients have been implanted. The major disadvantage of this system is the percutaneous cable connection which is the origin of a high infection risk. Caused by the electrode sizes and the required current amplitudes for stimulation, interactions between phosphenes (flickerphosphenes) and multiple phosphenes occur at one electrode [159]. The high stimulation amplitude can also cause meningal pain and a theoretical tendency to focal epilepsy [160]. Miniaturized electrodes for intracortical focal stimulation have been developed to overcome the limitations of surface stimulation. Acute experiments on humans with needle electrodes (an electrode site of 200 µm²) led to phosphene perceptions by biphasic charge balanced stimulus pulses with a pulse width of 200 µs and amplitudes between 20 and 300 µA, whilst surface stimulation of the cortex needed 2000 µA to elicit phosphenes [161]. An array of 38 electrode sites reproducibly elicited phosphenes and allowed the recognition of simple geometric patterns [27]. In human trials it has been implanted in the right visual cortex for four months. The excitation thresholds were acceptable, and the resolution was about five times higher than with surface stimulation [155]. Even simple pattern recognition was feasible [27,161].

78


Figure 4-2. Cortical vision prosthesis designed by Dobelle and coworkers. Technical sketch (left) and x-ray scan of an implanted electrode array (right). (both images from: www.dobelle.com; Dobelle W.H., The Dobelle Institute Lda., Lisboa, Portugal)

Micro-technology offers many advantages in comparison with precision mechanics: Even though the impedance of the electrodes is higher, the current amplitude and thus the charge over the electrode site is lower than with surface electrodes, and the tissue can be stimulated more localized. Three-dimensional micro-machined electrode arrays have been developed [162,163,164] to interface the visual cortex at different depths within vision prostheses). The access to the visual cortex is quite simple but the implants have to be positioned pneumatically with a special insertion tool [165]. Electronic circuitry in CMOS technology can be directly integrated into the substrates of the silicon based electrodes [166]. An increase of electrode numbers is relatively simple, from a technological point of view, if the number of devices can be increased just by combining single implants at the telemetry communication level. The opportunity to combine small groups of needle electrodes to high channel implants has been taken by a consortium of the Illinois Institute of Technology and the University of Chicago (Chicago, IL, USA), the


79

Figure 4-3. Concept of a micro-machined cortical vision prosthesis with 128 intracortical stimulation sites. Source: European Cortivis consortium (http:/cortivis.umh.es).

Laboratory for Neural Control of the NIH (Bethesda, MD., USA), the EIC Laboratories (Norwood, MA, USA) and the Huntington Medical Research Institute (Pasadena, CA, USA). Several groups of 8 electrodes have been implanted resulting in 114 needle electrodes with functional contacts and 38 electrodes with various contact problems after the surgical intervention [167]. The mapping of phosphenes in the visual field of macaque monkeys was successful but varied from subject to subject. Even with electrode numbers around 900 the quality of vision perception might not be sufficient if inherent cortical signal processing strategies will not be taken into account in the stimulation paradigms [167]. Despite the advantage of cortical electrodes there are some disadvantages, as already stated above. In healthy individuals numerous steps of information processing are gone through before the visual information reaches the visual cortex. The complexity of this information processing increases with ascending levels of the visual pathway and is by now partially understood [152]. This means spatial organization is more extensive in the visual cortex than at the retinal level, for example. Therefore visuotopic organization in the cortex is highly complex including many modalities

80


(colour, contrast, motion) in adjacent areas. As a result neural prostheses on lower levels of the visual system have also been developed; probably to use the signal processing of the visual system, at these levels if the anatomical structures have not been affected by the trauma or disease.

2.2

Optic Nerve Vision Prosthesis

The optic nerve connects the eye globes with the lateral geniculate nuclei. A vision prosthesis to interface the optic nerve requires as well as a retina implant the presence of the inner retina and the ganglion cells.

Figure 4-4. Concept of an optical nerve vision prosthesis with cuff electrode. Source: European MIVIP/OPTIVIP consortium.

Otherwise the optic nerve undergoes degeneration and is no longer functional. The surgical access to the optic nerve is technically demanding and requires the opening of the dura mater cavity. This may include severe risks such as CNS infections and facial scars. Veraart and colleagues showed the feasibility of an optic nerve vision prosthesis [168] for producing visual sensations with a spatial resolution that correlates with the stimulation sites,


81

using superficial stimulation via a four polar nerve cuff electrode around the optic nerve [169]. These results are quite impressive, because of the high density of more than one million axons within the optic nerve having a diameter of only 2 mm. A female patient with retinitis pigmentosa obtained an implant. She was able to recognize a number of predefined objects in a predefined test environment after a training phase of several months [170]. The recognition score was 63% with a processing time of 60 s [171]. However, for the perception of simple patterns and objects a translation of external information, e.g., from a video camera, into comprehensive stimuli for the optic nerve is necessary [172]. Owing to the difficult and severe surgical intervention for implantation and the low resolution of the vision prosthesis, this approach is only applicable for selected patients. Fundamental technical innovations are especially welcome in this approach for a larger patient benefit.

2.3

Retinal Implants

Worldwide, scientific research groups and companies work on the development and commercialisation of retinal vision prostheses. The different approaches have to take into account the anatomical restrictions of the eye globe and the physiology of the human retina. The developments in which the electrode arrays have to be placed under the retina (subretinal) or on the retina (epiretinal) have yet to become feasible by the degree of miniaturization made by means of micro system technology. Retina prostheses simulate the physiological function of the deceased receptor cells and stimulate the visual system at a very proximal level. As a result they benefit from the retinotopic organization of the retina and the further information processing of the entire visual pathway [153].

2.3.1

Subretinal Vision Prostheses

The subretinal approach places micro-photodiode arrays (MPDA), which are only a few µm in thickness, with thousands of subunits, comprising a photovoltaic cell, the driver circuit, and a stimulation micro electrode under the retina to stimulate the remaining neuronal structures electrically, and to replace the physiological function of the degenerated sensory cells. No camera or image processing is required. This stimulation at a very low level of the visual system allows the use of the inherent signal processing opportunities of the retinal and the higher levels of the visual system up to the visual cortex. The visual prosthesis consists of a micro-photodiode arraywhich measures the light on the retina and electrodes which excite the nervous structures in the retina at these sites by electrical stimulus pulses. At

82


Figure 4-5. Light is passing through the eye and the retina before it is transduced into electrical potentials by the photoreceptors. These potentials will be preprocessed in the retinal layers before the signals leave the eye via the optic nerve.

the beginning of the research it was supposed that the incidental light would be sufficient to stimulate the retina via the implanted photodiodes. Feasibility studies have been conducted on an isolated retina [173,174], as well as bio-compatibility studies and long term side effect studies performed on animal models [175,176,177,178,179,173]. There were no foreign body reactions and only mild fibrosis, but owing to the blocked nutrition of the outer retina by the implant most of the outer retinal cells degenerated [175,178,180]. Designs with more open space to increase diffusion to the retina might solve this problem. The implantation procedure can be chosen through the sclera from the backside of the eye (ab externo) or after a vitrectomy and retinotomy through the retina (ab interno) to reach the subretinal space, respectively. The implant is attached by natural adherence forces between the retinal pigment epithelium and the neurosensory retina.


83

Figure 4-6. Epiretinal and subretinal approach of a vision prosthesis. The epiretinal implant is placed in the vitreous cavity on the surface of the inner retina. The subretinal implant is positioned in the subretinal space.

Figure 4-7. Subretinal implant of photodiodes in a retinitis pigmentosa patient (www. optobionics.com; A. Chow, Optobionics Corp., Naperville, IL, USA).

84


Both access paths deliver long term stable placement after implantation, but special implantation tools have been developed [173] to minimize the risk of retinal damage. In the meantime, Zrenner’s group has proved that the existing brightness is not adequate for stimulating the nerve cells. The energy supply is guaranteed by an external infrared light diode while the environmental light only selects the stimulation electrodes and the related stimulus amplitude. The temperature increase of adjacent tissue by this additional infrared light source is tolerable, but light scattering might limit the dynamics of the implants which are also sensitive to infrared light, not the visible range only [181]. The implant now consists of the micro-photodiode array, an infrared diode, and some electronics for control. After light stimulation of the microphotodiode arrays visually evoked potentials were recorded with spatial resolution from the visual cortex of the cat and the pig, respectively [175]. Chow and colleagues report on stable visual impressions with no side effects in voluntary patients with retinitis pigmentosa over an implantation time of 2.5 years [182]. However, no objective benefit of the implant itself has been reported even though the implants were electrically functional [183].

Figure 4-8. Subretinal implant with a subretinal vision chip with 1500 integrated receivers and stimulation electrodes (right), infrared receiver for energy supply (middle) and electrical charge storage (left) assembled on a thin polyimide foil which serves as printed circuit board. By courtesy of H. Haemmerle (NMI, Reutlingen, Germany).


85

The major advantages and shortcomings of subretinal implants can be summarized [152]: utilization of the physiological optics; small devices with high spatial resolution; surgery is comparatively easy; absence of foreign body reaction. But there are still major obstacles to overcome such as the energy supply and retinal cells degenerating as a result of blocked nutrition.

2.3.2

Epiretinal Vision Prostheses

Research groups and companies in Australia, the USA, and Germany are working on an epiretinal vision prosthesis. In all these approaches the systems consist of an extra-corporal camera, a retina encoder, a telemetry link and an implantable part with electronic components for decoding the data received, selecting electrodes, delivering stimulus pulses, and the electrodes on the retina (Fig. 4-9). Since 1995 a German consortium has been developing a micro system which can be completely implanted within the eye, and spectacles with an integrated camera and an inductive link which supplies the implant with energy and stimulation data. A retina encoder will simulate the signal processing properties of the retina and stimulate the ganglion cell layer with adequate patterns to improve the visual perception [184].

Figure 4-9. Sketch of epiretinal vision prosthesis (IIP Technologies GmbH, Bonn, Germany).

86


The implant is assembled on a flexible polyimide substrate with integrated electrodes [185,186]. The receiver for data and energy is assembled in an artificial intra–ocular lens; a thin ribbon cable interconnects the receiver with the stimulator chip and the stimulation electrodes [187]. The implantation procedure is demanding but technically feasible. After a training phase the implants were fixed and stayed in place [188]. The current implants have 25 electrodes [189] which are sufficient for proof of principle experiments [190] but not for a final implant for patients. However, a spatial resolution in the visual cortex was obtained [191] with a cable-bound stimulator (Fig. 4-10) at stimulation amplitudes below 100 µA and a pulse width of 250 µs (Fig. 4-11).

Figure 4-10. Micro electrode array with 24 stimulation sites for acute animal studies. The electrodes for epiretinal stimulation were integrated on a polyimide substrate with a thickness of 15 µm. (By courtesy of the Fraunhofer Institute for Biomedical Engineering, St. Ingbert, Germany) .

Epiretinal vision implants with telemetry energy and data supply via an inductive link (Fig. 4-12) also elicited spatially distributed cortically evoked potentials that were monitored by means of optical recording [192].


87

In Germany the company IIP Technologies (Bonn, Germany) works on the commercialisation of an epiretinal vision prosthesis. Functional demonstrators of the implantable system have been fabricated on flexible substrates. They include an energy and data receiver, an electrode selection and stimulus pulse unit, and integrated stimulation electrodes (Fig. 4-13).

Figure 4-11. Stimulation thresholds for supra-maximal excitation of the retina with epiretinal electrodes. The electrical response was recorded from the optical tract as compound potential. Signals from the optical tract and stimulation thresholds were traced back to the related stimulation electrode. The counter electrode was placed extraocularly. Pulse width of the monophasic stimulus pulses was 250 µs (By courtesy of Ulf Eysel, Ruhr-Universitaet Bochum, Germany).

88


Figure 4-12. Epiretinal vision prosthesis with inductive energy and data transmission of the German EPI-RET consortium. The system has been successfully been implanted and functionally tested in cats [192].

Test equipment for acute human epiretinal stimulation has been developed and polyimide film based stimulation electrodes have been tested during surgical interventions on 8 blind voluntary subjects [193]. Recently a multi-centre study with 20 patients was performed with iridium oxide electrodes on micro-contact foils. 19 of 20 patients reported pleasant and not flashy perceptions which were electrically evoked by the stimulation with stimulation charges in the range from 20 to 380 nC. US groups implanted voluntary individuals who suffered from retinitis pigmentosa and had complete vision loss with 25 channel implants which have been fabricated by precision methods. The individuals reported vision sensations and recognized simple test patterns in a defined test environment after a training phase [194,195]. The US company Second Sight (Valencia, CA, USA) implanted six subjects in the end stage of retinitis pigmentosa with the Second Sight, Model 1, intra–ocular retinal prosthesis [196]. This prosthesis consists of a 4×4 grid array of platinum


89

Figure 4-13. Demonstrator of an epiretinal vision prosthesis from IIP Technologies GmbH (Bonn, Germany).

electrodes which were embedded into silicone rubber connected by a multiwire cable to hermetically sealed implant electronics which was surgically placed in the temporal bone under the skin. The electrode was secured onto the retina with a tack. Input to the implant was given via telemetry with a picture acquired by a mini video camera. The patient study with implantation ranges up to more than three years, and with stimulation currents as low as 6 µA at the minimum up to 1.1 mA [197]. The Australian group chose a sheep animal model to test their approach of an epiretinal vision implant [198]. The implant consists of a 100 channel CMOS neuro–stimulation ASIC [199] which is hermetically sealed before being implanted into an artificial intra–ocular lens. The electrode array was first proposed to be made by precision mechanics [199]. The latest results show platinum sheet based electrode arrays, with integrated cables, that have been laser cut in arbitrary geometries [200] and with a silicone rubber encapsulation. There are even more groups worldwide that work on retinal vision prostheses. More detailed information about the latest results can be found via the Association for Research in Vision and Ophthalmology (www. arvo.org) from which research abstracts from the annual meetings can be downloaded without charge.

90

2.4


Conclusions on Vision Prostheses

10 years ago, nobody would have been promised to be in human clinical studies with vision prostheses. Nowadays, owing to the advances in the fabrication of flexible, highly integrated micro systems, the development of visual prostheses acting at various levels of the visual system have become a realistic option in the future treatment of currently untreatable conditions which lead to blindness. However, visual prostheses will not restore normal vision. They are expected to provide a better quality of life for blind persons, e.g., for navigation in unknown environment without the help of sticks or dogs. Additionally to the technical developments, other methods for rehabilitation and therapy have to be taken into account, including stem cell transplant and gene therapy. The time schedule for these developments is unknown today but will certainly exceed 5 to 10 years before therapies will be transferred into a broad clinical practice. A recent development in the field of combining biology and micro-devices to biohybrid systems is the ‘Artificial Synapse Chip’ [201]. It provides an interface for ganglion cells to spread on the chip’s surface, routed by growth factors. Five µm, individually addressable, micro-fluidics apertures deliver neurotransmitter to stimulate single cell depolarisation. A much higher resolution than with other prosthesis is theoretically possible, but up to now experiments are in a very early stage. Future developments and the feedback of the first patients with chronic vision prostheses will guide us towards bringing vision prostheses into clinical practice.

3.

PERIPHERERAL NERVE INTERFACES

For some time interfacing neurons in the peripheral nervous system (PNS) and neuromuscular electrical stimulation have been investigated for controlling body functions after paralysis with neural prostheses and to modulate pathways for therapeutic reasons. If additionally the recording of nerves extracts human sensor signals of motor command signals, control of natural limbs after paralysis, or artificial limb control after an amputation trauma could be done. There are different methods of coupling to the PNS depending upon the type of information that has to be transduced. The most common type of interfaces only uses the electrical coupling method. However, this coupling is associated with some degree of invasiveness for the PNS structures, although all of the electrodes have to implanted chronically. The term ‘invasiveness’ means in this context the degree of trauma these electrodes cause depending on their design and implantation


91

site. This degree of invasiveness and the selectivity of the interfaces are directly related. As a result detailed knowledge about the necessary information and the target specification of neural prosthesis’ performance help to select the most appropriate type of PNS interface. Whilst this section gives only an insight in the opportunities of micro system based approaches for implantable PNS interfaces, comprehensive and more detailed information will be found elsewhere [202,49,159,203].

3.1

Non-Invasive Nerve Interfaces

Various approaches have been investigated for placing electrodes in the vicinity of neuro muscular motor points [204] or whole intact nerves. Approaches that have reached clinical practice have been fabricated by precision mechanics. Epineural electrodes were applied in phrenic pacing [20] and helix electrodes were placed around the vagal nerve [7]. Cuff electrodes are widespread as a long term stable interface for intact nerves. They have been mainly implanted in animal studies with rats and cats, but have also been commercialised for the correction of foot drop in humans (see above). These cuff electrodes were implanted circumferentially around the nerve without any nerve trauma. The cuffs with integrated electrode sites have been fabricated bys of precision mechanics using silicone rubber sheets and small platinum foils for the electrodes [205,206]. The fabrication techniques limit the number of electrodes and the minimal diameter of the cuffs. Most chronic investigations were done on the cat sciatic nerve which has a diameter of about 4 mm [207]. The application of micro-technologies to develop cuff electrodes with multiple contact sites was introduced several years ago [208] but only few data have been reported [209] and no chronic results have been found. Early experiments with combinations of silicone rubber and polyimide (Kapton) sheets ranged from critical to negative comments for long term implantations. The edges of these cuff devices led to nerve compression and cracks developed at the interfaces of the material combination [210]. Latest developments of micro-machined cuff electrodes use a different type of polyimide for fabricating thin and small diameter multi-site cuff electrodes [142]. Chronic implantations of multi-channel cuff electrodes (Fig. 4-14) were performed on rat sciatic nerves which have diameters less than 2 mm. The chronic compatibility with respect to mechanical integrity of the material and functionality of the electrode sites, as well as the opportunity for selective stimulation, have been the main objectives in these investigations.

92


Figure 4-14. Micro-machined polyimide based cuff electrode to interface small nerves. Longitudinal electrode pitch: 5 mm.

Electro physiological evaluation of the implanted nerves over the whole implantation time and histological characterization of the nerves at the end of the study [211] proved excellent bio-compatibility of the micro-machined polyimide based cuff electrodes. No material incompatibility on the nerve, and no material degradation were observed, neither any functional change in the electro physiological behaviour. The results of the previous study [210] could not be confirmed. On the contrary, the results recommend the use of this interface type for chronic implantation on the peripheral nerve. For investigations of the selectivity of electrical stimulation on thin peripheral nerve, electro physiological data (latency, amplitude of elicited nerve compound action potentials) should be accompanied by non-invasive torque data during isometric muscle contraction at the foot to compare the data with experiments on large diameter nerves of cats [38,212,213,214]. In acute experiments cuff electrodes with less than 2.0 mm diameter and four tripoles on the perimeter were implanted around the sciatic nerve of rats. The plantar and dorsiflexion torque was measured [148] after electrical stimulation and recruitment curves, as well as compound action potentials of plantar and dorsiflexor muscles. The results of the detailed investigations showed excellent selectivity of the small cuff electrodes without any a priori alignment before implantation [215]. Tripolar stimulation led to selective activation of plantar and dorsiflexor muscles, respectively. The use of


93

transversal steering currents was comparable with studies on ‘thick’ nerves [45]. The success of this micro-machined cuff electrode approach was finally proved with these studies. Now they have to be integrated into implantable systems to prove their applicability in neural prostheses. A combination of the polyimide technology with silicone rubber sheets to ‘Hybrid Cuff Electrodes’ finally led to larger diameter cuffs of high flexibility with the opportunity to integrate electronic circuitry (Fig. 4-15), e.g., multiplexers to reduce the number of cables [216].

Figure 4-15. Hybrid Cuff Electrode with integrated multiplexer to reduce the number of cables [216].

A completely different approach has been developed with the ‘bionic neurons’, the so called BIONs [74]. These single channel stimulators were developed with integrated radio receiver (Fig. 4-16). Up to 255 of them can be combined to distributed neural systems and controlled via a single external coil. They have to be injected near the motor points by a hypodermic needle. First clinical trials run in Europe. Current applications include the prevention of shoulder subluxation in patients after a stroke, and bladder management for incontinence patients.

94


Figure 4-16. The BION. The bionic Neurons are one-channel stimulators which are injected near the motor points by a hypodermic needle. By courtesy of Advanced Bionics, Inc. (Sylmar, CA, USA).

3.2

‘Semi’-Invasive Interfaces

In a slight variation of the cuff type of interfaces there are circum–neural electrodes that vary the nerve geometry to penetrate the epineurium without any damage to the perineurium. They combine the simplicity of extraneural electrodes with the selectivity of intra–neural approaches. Both existing device types have been developed by D. Durand and coworkers at the Case Western Reserve University (Cleveland, OH, USA). One design variation is the flat interface nerve electrode (FINE) that reshapes peripheral nerves for selective stimulation [217]. By flattening the nerve into a more elliptical shape the fascicles were distributed next to each other and become more accessible to electrical stimulation in comparison with cylindrical cuffs (Fig. 4-17).


95

Figure 4-17. Flat interface nerve electrode (FINE).

Acute studies showed that different fascicles and populations of fibers within a fascicle can be selectively excited [217,218]. In chronic studies nerve damage occurred when a high reshaping force was necessary, whereas moderate reshaping forces did not cause any nerve damage. Moderately flattened nerves showed good selectivity in limb excitation throughout the implantation time up to three months [218,219]. The FINE seems to be an interesting approach for neural interfaces, although it has not yet proved its applicability in clinical use. The slowly penetrating inter-fascicular nerve electrode (SPINE) (Fig. 4-18) places electrode contacts in between the fascicles by blunt penetration of the epineurium without compromising the integrity of the perineurium [220].

96


t

Figure 4-18. Slowly penetrating inter-fascicular nerve electrode (SPINE).

The SPINE consists of a silicone tube with four paddle-like elements consisting of integrated electrodes that extend radially into the tube lumen. Experiments over one day showed selective activation of the nerve fascicles and no damage of the perineurium [221]. Up to now no chronic results have been reported.

3.3

Invasive Interfaces

Invasive miniaturized neural interfaces have to be implanted inside the nerve. The epineurium will be penetrated, and in most cases a penetration of the perineurium is taken into account to obtain a high spatial selectivity and probably record not only mass activity of the nerve but single action potentials from the axons. The most prominent approaches include intra– fascicular (wire) electrodes, needle arrays, and sieve-like implants for interfacing nerve stumps or regenerating fibres, respectively.

3.3.1

Intrafascicular Electrodes

Wire based electrodes have been developed for being directly placed into the nerve for recording and stimulation purposes. Owing to the vicinity to the nerve fibres one obtains a good signal–to–noise ratio and little cross talk. Several of those intra–fascicular electrodes may be implanted for access to multiple nerves or multiple access to a nerve. Longitudinally implanted inter-fascicular electrodes (LIFEs) interface with small subsets of fibres within peripheral nerves. They were developed and fabricated from thin Pt–Ir wires or metallized Kevlar fibers [22 22 ,2 23, 224,225.] The active zone of the LIFE electrodes is between 250 µm and 1.5 mm, the rest of the wire is insulated (Fig. 4-19).


97

Figure 4-19. Longitudinal inter-fascicular electrode (LIFE).

Polymer based LIFEs offers better flexibility at reasonable tensile strength [226]. In general the LIFE concept is long term stable and no nerve damage was observed over 6 months of implantation [227,228]. They offer good selectivity or stimulation [229,230] and recording [231,232]. Recently LIFEs have been implanted in 8 human subjects with limb amputation [233]. On two consecutive days after surgery it was possible to record volitional motor nerve activity that was associated with missing limb movements. Consequently the control of an artificial limb should be feasible with these nerve interfaces. Stimulation over the electrodes delivered graded sensations of touch, joint movement, and position. These results suggest that peripheral nerve interfaces might supply amputated subjects with natural feedback about the limb prosthesis. Probably this is a first step towards closed loop control of artificial limb prostheses with the human in the loop.

3.3.2

Needle-Like Electrodes

Intra–neural electrodes have been designed to obtain optimal selectivity in recording and stimulation. Ideally, single nerve fibres or only small groups of axons are to be addressed within a peripheral nerve fascicle. Micro neurography with small tungsten wire electrodes is a clinically established method for the diagnosis of nerve diseases and fundamental investigations about nerve (patho-) physiology (for reviews see, e.g., [234,235,236]. However, this method is only used for acute investigations. Multi-wire micro electrode arrays have been developed [237,203] that investigated recruitment properties and stability of recorded signals and excitation thresholds in the nerve (Fig. 4-20) in animal experiments.

98


Figure 4-20. Needle array studies to interface peripheral nerves (WLC Rutten, Univ. of Twente, Enschede, The Netherlands).

Micro-machining electrode arrays with one-, two-, and three-dimensional shank arrangements have been developed by many research groups [238,239,240,241,242,243]. Some approaches integrated microelectronic circuits into the devices, others did not. Most of the approaches have been designed for CNS applications, some have been also tested in the PNS, although implanting such devices is always associated with potential damage of the nerve. The Center for Neural Communication Technology at the University of Michigan (Ann Arbor, MI, USA) is one of the leading centres which have developed a large number of single and multi-shaft electrodes and three-dimensional assemblies [238,132,244,245,246]. Groups from the Universities of Twente and Utah focused on silicon and silicon–glass technology [247,248,249]. Rutten interfaces 350 alphamotor fibers [203] in a rat peroneal nerve for fundamental studies. A slanted electrode array from the University of Utah (Salt Lake City, UT, USA) with electrode lengths ranging from 0.5 mm to 1.5 mm with 0.1 mm differences has been developed for PNS implantation [250].


99

Figure 4-21. The Utah slanted electrode array.

Electrodes were capable of recording single unit responses from mechano–receptors and graded recruitment during stimulation with high selectivity and orderly recruitment was observed [251,250]. Electrodes with and without lead wires were implanted up to seven month in cats’ sciatic nerves [252]. The implantation procedure highly affects the performance, stability and tissue reaction of the implant. A case report from a human volunteer who had a ‘normal’ Utah array implanted which was originally not designed for the PNS (4-25), in the median nerve for three months [253]. The subject received feedback information from force and slip sensors in a prosthetic hand but apparently no nerve damage occurred. However, gradual degradation of the lead wires of the electrodes limited the experiment.

3.3.3

Regenerative type of electrode

The thought control of artificial limbs, after an amputation trauma, by neural signals from the remaining nerves in the amputation stump is an old dream of mankind. First experiments had already been done at the end of World War II on voluntary air force pilots after being shot down [254]. Early experiments with chronic implantations of micro technical systems to interface regenerating nerves started in the 1970s [255,256,141] in animal experiments using epoxy based devices with drilled via holes. The general idea of this approach is to place a sieve-like micro system on the proximal nerve stump after the amputation trauma. The nerves grow, at least to some extent, and pass holes with integrated electrodes in a micro system which transmits the signals to a signal processing unit. In a second step these

100


signals control an artificial limb. When technological sensors deliver information about grasp force or hand position this information might be given back as a feedback signal to the subject by electrical stimulation of afferent nerve fibres through the same nerve interface (Fig. 4-22).

Figure 4-22. System concept of a regenerative type of nerve interface to control artificial limbs after an amputation trauma.

Since micro-machining technologies first were developed to structure silicon, sieve electrodes with integrated electrodes (Fig. 4-23) have been developed from this material [257,258,259,260].

Figure 4-23. Example of a silicon based sieve electrode for interfacing regenerating nerves. By courtesy of the Fraunhofer Institute for Biomedical Engineering, St. Ingbert, Germany.


101

Recordings of electrical signals from regenerated nerve fibres have been made with this kind of neuro-technical interfaces in chronic animal experiments [260,261,262,263,264,265]. Depending on the thickness of the technical interface and the mechanical load on the nerve by the device, the amount of regeneration varied a lot. A major drawback of these approaches was the establishment of an electrical contact with the substrate-integrated electrodes. Some approaches only offered contact pads for needle probers directly on the sieve [262,264], or short silicon based cables had been monolithically integrated with complex technologies [258]. Neural activity recording was demonstrated in peripheral nerves of rat, frog, and fish [262,265,266,267,268,269]. The introduction of polyimide as flexible substrate material for sieve electrodes [142,270,271] offered the opportunity of integrating electrodes, quite long interconnection cables, and connection pads with relatively simple technologies (Fig. 4-24).

Figure 4-24. Polyimide based sieve electrode with integrated cable and nerve guidance chamber. By courtesy of the Fraunhofer Institute for Biomedical Engineering, St. Ingbert, Germany.

Chronic implantation experiments with polyimide based sieve electrodes delivered excellent results with respect to bio-compatibility and degree of regeneration. The sieve electrodes proved to be stable over months of in vivo implantation and showed much better regeneration than silicon dice [143,272,273]. In an amputation study the presence of a polyimide based sieve implant showed no chronic foreign body response, and the immuno-

102


histochemical measured patterns of proteins were comparable to those without an implant [274]. Compound nerve action potentials that were elicited by means of electrical stimulation were recorded over the sieve electrodes and electrical stimulation via the sieve electrodes resulted in muscle activation [143]. Implantation of a polyimide based sieve electrode as the interface between the optic nerve stump (CNS) directly at the eye globe and a peripheral nerve graft led to functional regeneration of CNS neurons, the ganglion cells into the nerve graft through the sieve electrode [275]. For the first time a functional regeneration of retinal ganglion cells, i.e., adult CNS neurons, was reported. Neuronal signals after flash light stimulation proved functional regeneration. The most logical and challenging application of these sieve electrodes is, of course, their implantation in severed nerves of amputated limbs for a bi-directional interface. So far, none of the approaches has been transferred into clinical trials. It is to be hoped that results from more fundamental studies will help to speed up the transfer to the thought control of artificial limbs by PNS interfaces.

3.4

Biohybrid Approaches

Neural implants establish an interface between a technological device and a target structure in the nervous system. This target connection is more or less unspecific owing to a quite large technological device with respect to nerve cells. As a result some approaches work on the combination of cells and technical devices before implantation to establish a more biologically motivated interface to the nervous system. These approaches are often called bioelectronic or biohybrid interfaces. The objective is to precultivate and settle single nerve cells, nerve cell networks, or layers on technological surfaces with electrodes and to induce fibber growth to establish ‘biologic’ interface with a nerve-nerve contact in the target tissue of the body. Many investigations have been done in fundamental research to couple single nerve cells or nerve cell networks to technical recording equipment via electrodes or electronic components. A manifold of micro-machined neuron probes is described in the literature [203]. The structure’s topology, e.g., the width of grooves to guide axons, obtained an important impact on the growth of the cultivated cells [276,277,137]. Fromherz et al., were able to demonstrate the guided growth of nerve cells from leeches on a laminine coated substrate [278]. In further experiments they obtained a current-free measurement of the membrane potential of a single neuron of a leech, a so called retzius cell by direct coupling to the gate of a field effect transistor (FET) [279]. Using an array of FETs, a multi-electrode array for a network


103

of cultivated neurons will be possible. Usually capacitive titanium nitride or metal electrodes are used for a bidirectional information exchange with nerve cell cultures for recording and stimulation. FETs are not suitable for nerve cell stimulation; additionally, some groups report a significant higher amount of noise with the FET approach compared to ‘simple’ metal electrodes. The activity of embryonic spinal cord cells of mice has been recorded simultaneously over a period of several months on a planar glass substrate with 64 indium–tin–oxide (ITO) electrodes [280]. Autonomous signals generated from the cultivated neural networks without forcing them into predefined surface topologies were also able to be examined [281]. They were used for qualitative and quantitative drug detection. Having surface topologies on the substrate, axonal outgrowth of embryonic nerve cells from rats was observed and electrical activity was recorded in a microstructure with 16 electrodes which were connected via grooves [282,283]. These approaches were generally not intended to be transferred into implantable bioelectronic interfaces. If implantation of cells in microstructures for implantation is successful, the migration of the host cells out of the technical device must be prevented. In some investigations silicon micro-machining was used to create wells with gold electrodes on the bottom and a grillwork at the top [284]. The cells were seeded into the wells. The grillwork prevented the neurons from migrating. The group of Jerry Pine managed to plant single hippocampus cells of the rat into a threedimensional silicon structure which was termed a ‘cultured neuron probe’ [285]. They were able to demonstrate that the transplanted cells inside the micro-compartments retain their capacity for axonal outgrowth. Nerve action potentials were directly recorded extra–cellularly from the soma of the cells. However, the transfer of the cultured cells from the in vitro environment into the brain remained difficult. Another bioelectronic interface was introduced as a ‘cone electrode’ [286] (Kennedy, 1989). Here a piece of a rat sciatic nerve was placed into a cone of 1.5 mm length and 100–200 µm diameter. After implantation in the brain of a rat, neurite ingrowth was observed and stable recordings were obtained over six months. This approach has been commercialized recently (Neural Signals Inc, Atlanta, G.A., USA). Both approaches were not investigated in the peripheral nervous system. In the peripheral nervous system bioelectronic interfaces are in a very early and experimental stage. A traumatic lesion of a peripheral nerve causes paralysis and leads to degeneration of the distal nerve stump by Wallerian degeneration. Healing of such a lesion occurs only under favorable conditions and involves the outgrowth of axons from the proximal nerve stump, sprouting along the distal nerve stump as a guidance structure and reinnervation of the target muscles. Mostly healing is not successful, if the

104


lesion site is very proximal. As a consequence alteration on the muscles in the course of atrophy prevents a formation of new neuro muscular junctions at a later point of time when regenerating nerve fibres reach the muscle. As a result of all processes, a flaccid paralysis occurs with muscle atrophy and electrophysiological changes in muscle membrane excitability. So far, no clinical treatment or therapy is able to help those patients. Success has been reported in the development of an animal model for curing flaccid paralysis in a peripheral nerve with a biohybrid approach. Embryonic spinal cord cells have been transplanted into a containment adapted to the distal nerve stump to restore skeletal muscle function [287,288,289] in the animal model of the rat. The containment consisted of a 10 mm long piece of the autologous femoral vein [287] that has been sutured micro surgically to the distal stump. Fetal spinal cord was fragmented and injected into that vein containment. The proximally end of the vein containment was closed with surgical sutures. After 3–6 months, reinnervation of the gastrocnemius and tibialis muscle via the distal stump of the sciatic nerve was observed, although there was no communication with the central nervous system. A regeneration type of electrode was introduced as a bioelectronic interface [290] between the sprouting embryonic neurons and the peripheral nerve to allow neural stimulation for controlled muscle excitation, to prevent atrophy and restore some function, e.g., in grasp after upper limp flaccid paralysis. Instead of a vein a technological compartment was taken. The embryonic cells were purified and only motor neurons were injected into the containment. Functional regeneration was proved but the number of surviving cells decreased tremendously from the more than 10,000 that had been injected to some tens that finally innervated the muscle [291]. Further fundamental investigations are necessary to obtain long term survival of a sufficiently high number of transplanted cells. One approach could be a transient increase of stress resistivity by means of molecular biology or with more sophisticated transplantation concepts to survive the first 72 hours that seemed to be critical period according to our experience. Finally, the issue of using cells and their origin should be scientifically and ethically addressed at a very early point of time to develop strategies for transferring results from animal experiments into human patients. Owing to latest research results, autologous adult stem cells might be favourable for preventing foreign body reactions and for overcoming the crucial ethical issues of whether embryonic stem cells might be a solution or ethically far out of discussion.


4.

105

FUTURE APPLICATIONS

The research field of biomedical micro-devices for neural prostheses is still affected by preclinical and experimental studies, although more and more application are transferred into clinical studies. Many questions of interfacing neurons with artificial materials are not yet understood in their fundamentals, whereas the excitation of nerves can be simulated with many details. For some cases we are already lucky to have access to analytical solutions. New materials and technologies are under continuous evaluation for being adequate for neural interfaces. Research results change the state of the art from day to day. The latest technological developments and neuro– scientific findings should not be discussed here. The annual international conferences of the IEEE Engineering in Medicine and Biology Society (EMBS) and of the Society for Neurosciences are excellent places to update one's own knowledge and discuss new ideas. Here three scenarios from fundamental research up to clinical pilot experiments will be presented which are currently under investigation.

4.1

Interfacing the Brain

The implantation of penetrating micro-electrodes into areas of the cortex opens a new quality of interfaces to the brain. Miniaturized electrode arrays have been developed for more than 20 years. They proved their biological and technical stability and functionality, in acute and chronic animal trials for recording neural signals from different brain regions such as the hippocampus, the cerebellum and the cortex. So far only one array was implanted per trial. The connection between the electrodes of the array and the signal processing electronics and the data storage was cable bound. As a result chronic clinical implantations within humans have not been done and seem to be feasible only for subchronic pilot experiments.

106


Figure 4-25. The Utah electrode array.

Recently, in autumn of 2004, a 100 channel electrode array (Fig. 4-25) [240] was implanted in humans by Cybercinetics, Inc (Foxborough, MA, USA). It will allow locked-in patients who are completely paralysed and are not able to move any muscle, but have clear intellectual properties, to communicate with their environment via an implanted brain–computer interface (BCI). The step from one electrode array to distributed electrodes allows several approaches. The Illinois Institute of Technology (Chicago, IL, USA) in collaboration with the Huntington Medical Research Institute (Pasadena, CA, USA) developed implants with up to 1,024 electrodes which were grouped in assemblies of 8 precision mechanics needles. These groups were connected via cables to the central implant electronics. The University of Michigan (Ann Arbor, MI, USA) chose micro-fabrication technologies to develop silicon based shaft electrodes with single and multiple shanks and multiple electrode sites per device [245]. Comb-like electrode devices were assembled into three-dimensional arrays with up to 1,024 electrode sites.

Figure 4-26. Comb-like elements for a three-dimensional electrode array with integrated electronics for CNS applications (Univ. of Michigan, Ann Arbor, MI, USA).


107

Monolithic integration of microelectronic t circuits in the silicon based electrodes devices allows amplification, filtering, and multiplexing. The detection of action potentials [246], the so called spikes, allows data reduction on the devices and telemetry data transmission with reduced data depths. These arrays can be implanted in several regions of the brain and communicate like distributed systems. Application scenarios for these implants are in the (pre-) motor cortex region [292,293,294] and range from control of simple user interfaces for communication purposes up to direct motor control of artificial robotic limbs.

Figure 4-27. Arm control with an implanted brain–computer interface (M. Nicolelis, Duke Univ. Durham, NC., USA).

Visionary concepts discuss the use of distributed implants for data exchange between distant brain regions. While current investigations only use the electrode arrays for recording of neural signals, a bi-directional interface for recording from the neurons and electrically stimulating as actuator path within a closed loop control is feasible. Applications within the control of grasp could deliver information of grasp force or the spatial position of the hand to the somatosensory cortex, no matter whether signals from human receptors of a paralysed hand within a neural prosthesis or technical sensors of a robotic artefact are the origin of the data.

108

4.2


Spinal Cord Implants

Neural prostheses for restoring gait and posture were implanted in humans in only few numbers. A drawback of existing systems (see the section ‘neuroprostheses for gait and posture’ in this chapter) is the small opportunity of really giving back mobility to the patients. The patients fatigue very fast, need a walking frame, and have to train hard before they can stand up and walk some steps. The muscles fatigue fast under electrical stimulation because the fast fatiguing muscle fibres are excited first as a result of the so called inverse recruitment. Latest research results from the group of Vivian Mushahwar at the University of Alberta (Canada) showed complex movement patterns after spinal cord stimulation. Wire electrodes have been implanted in the lumbar spinal cord of cats to stimulate central pattern generator regions and to activate motor pools, which elicited movement patterns after stimulation via a single electrode [295]. The stimulation resulted in normal recruitment, i.e., no reverse recruitment occurred. A normal gait of the cat and weight bearing muscle force was obtained with only four electrodes per side. These results have been impossible with peripheral neural stimulation over longer stimulation times. No fatigue occurred when using spinal cord stimulation.

Figure 4-28. Concept of spinal cord stimulation with microwires (by courtesy of V. Mushahwar).

Many open questions have to be solved before a transfer to humans can be done. The presence of central pattern generators after complete spinal


109

cord injury and their use for spinal cord stimulation has to be investigated carefully. If they would be present and would deliver sufficiently complex movement patterns which could bear the patient‘s weight, the field of gait and posture neuroprostheses would be at a point of inflection. So far, prognosis should be done in a conservative manner because of a controversy discussion about central pattern generators in man [296]. However, the use of G. Colombo’s ‘LOKOMAT’, a machine which technically induces stepping movements, in group at the Balgrist University Hospital in Zurich (Switzerland) in incompletely spinal cord injured patients delivered good results in movement control of patients on the treadmill [297,298].

4.3

Multi-modal Neural Implants

eNural interfaces have mainly been used to obtain an insight into the electrical communication within the nervous system and to create bio-electrical sensor and actuator interfaces, respectively. However, the peripheral, and especially the central, nervous system communicate in a more complex way. eNuro transmitters modulate nerve activity in a very sophisticated way. Electrical stimulation in comparison excites large assemblies of nerve cells and elicits patterns that are useful in therapy, e.g., deep brain stimulation to alleviate the symptoms of Parkinson’s disease but do not finally lead to the release of substances such as dopamine. With respect to a patient oriented therapy some research groups have already started to investigate sensing and actuating principles other than recording and electrical stimulation in neuro science applications. Electro-active substances in the brain, such as dopamine, were monitored with tetrode arrangements of electrodes on ceramic based substrates [299]. Bioactive substances for attracting nerve fibres were integrated into flexible multi-channel shaft electrodes [300] for neural recording (Fig. 4-29).

110


Figure 4-29. Polyimide based electrode array for CNS applications with capability to store bioactive substances [300].

Figure 4-30. Cross-section of an electrode device with integrated micro-fluidics channels [301].

The combination of micro-fluidics channels and electrode sites for recording [301] and stimulation opens new opportunities for detecting pathophysiological neural excitation patterns and deliver specific pharmaceutical agents in small doses directly near the sites of the incidents. The combination of these measurement and actuation modalities to multimodal neural implants paves the way towards patient specific diagnosis and therapy and therapy control and hopefully helps to cure diseases and alleviate symptoms where currently no reliable methods and adequate therapies are present.

5.

CONCLUDING REMARKS

In the final concluding remarks we would like to give some flashlights on challenges and chances of micro system based neural interfaces that have been presented in this chapter. The technological developments of the last few decades in microelectronics, micro-machining, material sciences as well


111

as the gain of knowledge in clinical medicine and fundamental neurosciences drove many investigations which finally helped to bring some neural prostheses into clinical practice, and some ‘crazy’ ideas like the retina implant from the blackboard into first human clinical trials. However, the collaboration between technology and biology and medicine has to go onwards to learn the lessons of how to develop complex and long term stable biomedical micro-implants. Even though many investigations have reported no inflammation and little tissue reaction after implantation of interfaces to peripheral nerves, the retina and the brain, the implant is still a foreign body. Surface properties to mimic biological macro–molecular structures have not yet been taken into account. Further on, technological solutions have to be developed for basic requirements, especially when implants should become more and more complex to give more and more functionality back to the patient. One of the major challenges for implants is a reliable and powerful energy supply by a telemetry link or by (rechargeable) batteries. For a long implant lifetime and a small implant size low power circuitry and low impedance electrodes should be used. At high integration densities and large electrode numbers, even the resistance of connection cables should be taken into account. A reduction of the track resistance of a factor of 30 by electroplating the tracks [302] might help in systems with hundreds of electrodes to keep the dissipative energy at a reasonable level. The insertion sites of penetrating electrodes could be reduced by just a factor of 2 if double-sided electrode arrangements were to be introduced [242,302]. An increase of lateral spatial resolution and reduced tissue reaction would go hand in hand. The combination of active neural cells and technical neural prostheses to biohybrid systems opens opportunities for novel therapies for diseases that have been fatal or for better and more long term stable connections in the central nervous system. However, cells do not like the surfaces of the technological systems very much. As a result the ‘biologication’ of neural prostheses seems to be one of the major challenges for the next decade.

6.

NEURAL IMPLANTS: BOON OR BANE?

Neural implants offer a tremendous potential for new therapeutic approaches in many diseases. Their effect ranges from simple use of reflex circuitry (drop foot stimulator) to complex excitation of brain structures (cortical vision prostheses) in functional electrical stimulation for motor and sensory neural prostheses. Neural implants help to increase the abilities of daily living in disabled persons, even though neural prostheses for motor functions only have limited performance. However, in the case of a spinal cord lesion, phrenic pacing, and management of the urinary bladder or the

112


functional use of the hand as a result of electrical stimulation may help to gain an incredibly high amount of independence and self-determination. The efforts of neuro sciences in the last few years have pushed interfaces to central nervous system structures from fundamental research towards clinical practice. Deep brain stimulators for patients suffering from Parkinson’s disease are implanted in clinical routine interventions in many hospitals, not only in specialized centres. The latest success stories tell us that central electrical stimulation helps patients with psychiatric disorders which have been untreatable by means of psychiatric drugs. In autumn of 2004, the first needle arrays from Cyberonics (Foxborough, MA, USA) were implanted in patients to allow a very limited ‘thought control’ in locked-in patients to facilitate their ability to use a brain-computer-interface to communicate with their environment, e.g., to write emails, as has been spread in many press releases worldwide. Simple actuator input in neural prostheses triggers complex circuits in the brain by this kind of neuro– modulation. The path between the pure treatment of the disorder and a side effect of changing the patient’s behaviour or even the personality might be very narrow. Consequently we recommend a dialogue between neuroscientists, biomedical engineers, clinicians, and philosophers to address ethical issues and societal guidelines that finally will lead to ethical standards in parallel with the technical developments. An open information policy with a broad public discussion in different disciplines helps to create a broad societal acceptance and dissemination to all patients who might benefit from the latest technology with least side effects.


7.

113

REFERENCES

References cited in Chapter 3 and 4. [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

[11] [12] [13]

[14]

Hambrecht, F.T., Reswick, J.B. (Hg.), (1977), ‘Functional electrical stimulation’. Marcel Dekker, New York. Rupp, R., Gerner, H.J., (2004), ‘Neuroprosthetics of the upper extremity—clinical application in spinal cord injury and future perspectives’, Biomed. Techn., 49, no. 4, 93–98. Berkowitz, M., O’Leary, P.K, Kruse, D.L, Harvey, C., (1998) ‘Spinal cord injury: an analysis of medical and social costs’, Demos Medical Pubishing, New York. Guttmann, L., (1956), ‘The management of the paraplegic patient’, Practitioner 176(1052), 157–70. Brindley, G.S., Rushton D.N., (1995) ‘Neuroprotheses’, Clinical Neurology. 4(1). Makoschey, D., (1996), ‘Funktionelle Elektrostimulation und ihre Alternativen’, FTB, Wetter/Ruhr . Cavuoto, J., (2004) ‘Neural engineering’s image problem’, IEEE Spectrum, 41, no. 4, 20–25. Rijkhoff, NJM., (2004), ‘Neuroprostheses to treat neurogenic bladder dysfunction: current status and future perspectives’, Childs Nerv. Syst. 20, 75–86. North, R.B., Roark, G.L., (1995), ‘Spinal cord stimulation for chronic pain’, Neurosurg Clin North Am 6, 145–155. Waltz, J.M., (1997), ‘Spinal cord stimulation: a quarter century of development and investigation—a review of its development and effectiveness in 1,336 cases’, Stereotact Funct Neurosurg 69, 288– 299. Witte, R.J., Lane, J.I., Driscoll, C.L., Lundy, L.B., Bernstein, M.A., Kotsenas, A.L., Kocharian, A., (2003) ‘Pediatric and adult cochlear implantation’, Radiographics 23(5), 1185–1200. Federspil, P.A., Plinklert, P.K., (2004) ‘Restoring hearing with active hearing implants’, Biomed. Techn., vol 49, no. 4, 76–82. Deuschl, G., Wenzelburger, R., Kopper, F., Volkmann, J., (2003) ‘Deep brain stimulation of the subthalamic nucleus for Parkinson’s disease: a therapy approaching evidence based standards’, J. Neurol., 250 Suppl 1, I43–I46. Boon, P., Vonck, K., de Reuck, J., Caemaert, J., (2001), ‘Vagus nerve stimulation for refractory epilepsy’, Seizure 10, 448–455.

114


[15] Grunewald, V., Hofner, K., Thon, W.F., Kuczyk, M.A., Jonas, U., (1999), ‘Sacral electrical neuro-modulation as an alternative treatment option for lower urinary tract dysfunction. Restor’, Neurol. Neurosci. 14, 2–3, 189–193. [16] Siegel, S.W., Catanzaro, F., Dijkema, H.E., Elhilali, M.M., Fowler, C.J., Gajewski, J.B., Hassouna, M.M., Janknegt, R.A., Jonas, U., van Kerrebroeck, P.E., Lycklama a Nijeholt, A.A., Oleson, K.A., Schmidt, R.A.,(2000), ‘Long term results of a multi-center study on sacral nerve stimulation for treatment of urinary urge incontinence, urgencyfrequency, and retention’, Urology 56 [Suppl 6A], 87–91. [17] Bosch, J.L.H., Groen, J., (2000), ‘Sacral nerve neuro-modulation in the treatment of patients with refractory motor urge incontinence: long term results of a prospective longitudinal study’, J. Urol. 163, 1219–1222. [18] Brindley, G.S., (1994), ‘The first 500 patients with sacral anterior root stimulator implants: general description’, Paraplegia 32(12), 795–805. [19] van Kerrebroeck, P.E., Koldewijn, E.L, Debruyne, F.M., (1993), ‘World wide experience with the Finetech-Brindley sacral anterior root stimulator’, Neurourol. Urodyn., 12, 497–503. [20] Mayr, W., Bijak, M., Girsch, W., Holle, J., Lanmüller, H., Thoma, H., Zrunek, M., (1993), ‘Multi-channel stimulation of phrenic nerves by epineural electrodes’, ASAIO Journal, 39, 729–735. [21] Creasey, G., Elefteriades, J.A., DiMarco, A., Talonen, P., Bijak, M., Girsch, W., Kantor, C., (1996), ‘Electrical stimulation to restore respiration’, J. Rehabil. Res. Dev. 33, 123–133. [22] Horch, K., Dhillon G. (ed.), (2004), ‘NEUROPROSTHETICS: Theory and Practice’, (Series on Bioengineering and Biomedical Engineering, 2), World Scientific Publishing, River Edge, London, Singapore. [23] McNeal, D.R., (1977), ‘2000 Years of Electrical Stimulation’. In ‘Functional Electrical Stimulation. Applications in Neural Prostheses’ (eds.: Hambrecht, F.T., Reswick, J.B.), Marcel Dekker, Inc., New York, 3–35. [24] Rattay, F., (1990), ‘Electrical Nerve Stimulation’, SpringerVerlag, Wien. [25] Rydevik, B.L., Danielsen, N., Dahlin, L.B., Lundborg, G., (1990), ‘Pathophysiology of Peripheral Nerve Injury with Special Reference to Electrode Implantation’. In ‘Neural Prosteses’ (eds.:Agnew, W.F., McCreery, D.B.), Prentice Hall, Englewood Cliffs, New Jersey, 85– 105. [26] Kimura, J., (1989), ‘Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice’, F.A. Davis Company, Philadelphia.


115

[27] Schmidt, E.M., Bak, M.J., Hambrecht, F.T., et al., (1996), ‘Feasibility of a visual prosthesis for the blind based on intracortical microstimulation of the visual cortex’, Brain 119, 507–522. [28] Rattay, F., Aberham, M., (1993), ‘Modeling Axon Membranes for Functional Electrical Stimulation’, IEEE Trans. Biomed. Eng. 40, 1201–1209. [29] Struijk, J.J., (2004), ‘Passive models of excitable cells’ in: Horch, K.W., Dhillon, G.S. (eds.): Neuroprosthetics-Theory and Practice (Series on Bioengineering and Biomedical Engineering, 2, World Scientific Publishing Co., River Edge, NJ. [30] Rattay, F., (1986), ‘Analysis of Models for External Stimulation of Axons’, IEEE Trans. Biomed. Eng. 33, 974–977. [31] Grill, W.M., (2004), ‘Electrical stimulation af the peripheral nervous system: biophyics and excitaiton properties’ in: Horch, K.W., Dhillon, G.S. (eds.): Neuroprosthetics-Theory and Practice (Series on Bioengineering and Biomedical Engineering), 2, World Scientific Publishing Co., River Edge, NJ. [32] Gorman, P.H., Mortimer, J.T., (1983), ‘The Effect of Stimulus Parameters on the Recruitment Characteristics of Direct Nerve Stimulation’, IEEE Trans. Biomed. Eng. 30, 407–414. [33] Mortimer, J.T., (1990), ‘Electrical Excitation of Nerve’, In ‘Neural Prostheses’ (eds.: Agnew, W.F., McCreery, D.B.), Prentice Hall, Englewood Cliffs, New Jersey, 67–83. [34] Petrofsky, J.S., Phillips, C.A., (1981), ‘Impact of Recruitment Order on Electrode Design for Neural Prosthetics of Skeletal Muscle’. Am. J. Phys. Med. 60, 243–253. [35] Holle, J., Moritz, E., Thomas, H., Lischka, A., (1974), ‘Die Karusselstimulation, eine neue Methode zur elektrophrenischen Langzeitbeatmung’, Kin. Wochenzeitschrift, 86, 23. [36] Tai, C., Jiang, D. (1992), ‘A Modeling Study of Selective Stimulation of Nerve Fibres Using Single Cathode Electrode’, Proceedings, Ann. Int. Conf. IEEE Engineering in Medicine and Biology Society, 1370–1371. [37] Fitzpatrick, D.M., Struijk, J.J., Andrews, B.J., (1991), ‘A Nerve Cuff Design for the Selective Activation and Blocking of Myelinated Nerve Fibres’, Proceedings, Ann. Int. Conf. IEEE Engineering in Medicine and Biology Society, 906–907. [38] Veraart, C., Grill, W.M., Mortimer, J.T., (1993), ‘Selective Control of Muscle Activation with a Multi-polar Nerve Cuff Electrode’, IEEE Trans. Biomed. Eng. 40, 640–653.

116


[39] Rijkhoff, N.J.M., Koldewijn, E.L., van Kerrebroeck, P.E., Debruyne, F.M.J., Wijkstra, H., (1993), ‘Sacral Root Stimulation in the Dog: Reduction of Urethral Resistance’, Proceedings, Ann. Int. Conf. IEEE Engineering in Medicine and Biology Society, 1257–1258. [40] Goodall, E.V., Kosterman, M.L., Struijk, J.J., Holsheimer, J., (1993), ‘Simulation of Activation and Propagation Delay during Tripolar Neural Stimulation’. Proceedings, Ann. Int. Conf. IEEE Engineering in Medicine and Biology Society, 1203–1204. [41] Accornero, N., Bini, G., Lenzi, G.L., Manfredi, M., (1977), ‘Selective Activation of Peripheral Nerve Fibre Groups of Different Diameter by Triangular Shaped Stimulus Pulses’. J. Physiol. 277, 539–560. [42] Fang, Z.P., Mortimer, J.T., (1991), ‘Selective Activation of Small Motor Axons by Quasitrapezoidal Current Pulses’. IEEE Trans. Biomed. Eng. 38, 168–174. [43] Lertmanorat, Z., Durand, D.M., (2004), ‘A novel electrode array for diameter-dependent control of axonal excitability: a simulation study’, IEEE Trans. Biomed. Eng. 51(7), 1242–1250. [44] Baratta, R., Ichie, M., Hwang, S.K., Solomonow, M., (1989), ‘Orderly stimulation of skeletal muscle motor units with tripolar nerve cuff electrode’. IEEE Trans. Biomed. Eng. 36, 836–843. [45] Grill, W.M., Veraart, C., Mortimer, J.T., (1991), ‘Selective Activation of Peripheral Nerve Fascicles: Use of Field Steering Currents’. Proceedings, Ann. Int. Conf. IEEE Engineering in Medicine and Biology Society, 904–905. [46] Grill, W.M., Mortimer, J.T., (1993), ‘Selective Activation of Distant Nerve Fibres’. Proceedings, Ann. Int. Conf. IEEE Engineering in Medicine and Biology Society, 1249–1250. [47] Dobelle, W.H., (2000), ‘Artificial vision for the blind by connecting a television camera to the visual cortex’, ASAIO J., 46–49. [48] Suaning, G.J., Lovell, N.H., Kwok, C.Y., (2003), ‘Fabrication of platinum spherical electrodes in an intra–ocular prosthesis using highenergy electrical discharge’, Sens. Act. A 108, 155–161. [49] Heiduschka, P., Thanos, S., (1998), ‘Implantable bioelectric interfaces for lost nerve functions’, Prog. Neurobiol. 155, 433–461. [50] Stieglitz, T., Schüttler, M., Koch, K.P., (2004), ‘Neural prostheses in clinical applications - trends from precision mechanics towards biomedical micro systems in neurological rehabilitation’, Biomed. Tech. 49, 72–77. [51] Bronzino, J.D., (1995), ‘Biomedical Engineering’, Handbook; CRC Press, IEEE Press, Boca Raton.


117

[52] Geddes, L.A., Roeder, R., (2001), ‘Measurement of the direct-current (faradic) resistance of the electrode-electrolyte interface for commonly used electrode materials’, Ann. Biomed. Eng., 29, 181–186. [53] Geddes, L.A., Roeder, R., (2003), ‘Criteria for the selection of materials for implanted electrodes’, Ann. Biomed. Eng., 31, 879–890. [54] Janders, M., Egert, U., Stelzle, M., Nisch, W., (1996), ‘Novel thin film titanium nitride micro-electrodes with excellent charge transfer capability for cell stimulation ans sensing applications’, Proc. 19th Inte Conf. IEEE/EMBS., 1191–1193. [55] Weiland, J.D., Anderson D.J., Humayun M.S., (2002), ‘In vitro electrical properties for iridium oxide versus titanium nitride stimulation electrodes’, IEEE Trans. Biomed. Eng., 49, 1574–1579. [56] Robblee, L.S., Rose, T.L., (1990), ‘Electrochemical Guidelines for Selection of Protocols and Electrode Materials for Neural Stimulation‘, in ‘Neural Prostheses’ (eds.: Agnew, W.F., McCreery, D.B.), Prentice Hall, Englewood Cliffs, New Jersey, 25–66. [57] McCreery, D.B., Agnew, W.F., Yuen, T.G., Bullara, L.A., (1995), ‘Relationship between stimulus amplitude, stimulus frequency and neural damage during electrical stimulation of sciatic nerve of cat’, Med. Biol. Eng. Comput., 33(3 Spec No), 426–429. [58] Liu, X., McCreery, D.B., Carter, R.R., Bullara, L.A., Yuen, T.G., Agnew, W.F., (1999), ‘Stability of the interface between neural tissue and chronically implanted intracortical micro electrodes’, IEEE Trans. Rehabil. Eng. 7(3), 315–326. [59] Bolz, A., Urbaszek, W., (2002), ‘Technik in der Kardiologie: eine interdisziplinäre Darstellung für Ingenieure und Mediziner’, Springer Verlag, Berlin, Heidelberg, New York. [60] Schaldach, M., (1992), ‘Electrotherapy of the Heart’. Springer Verlag, Berlin. [61] Webster, J.G. (ed.), (1995), ‘Design of cardiac pacemakers’, IEEE Press, Piscataway, NJ. [62] van de Heyning, P.H., Vermeire, K., Cochet, E., Hofkens, A., De Bodt, M., (2004), ‘Cochlear implantation: a bionic system to restore hearing in severely hearing impaired patients’, Verh. K. Acad. Geneeskd. Belg. 66(4), 253–267. [63] Zeng, F.G., (2004), ‘Trends in cochlear implants’, Trends Amplif. 8(1), 1–34. [64] Ho, S.Y., Wiet, R.J., Richter, C.P., (2004), ‘Modifying cochlear implant design: advantages of placing a return electrode in the modiolus’, Otol Neurotol. 25(4), 497–503.

118


[65] Balkany, T.J., Eshraghi, A.A., Yang, N., (2002), ‘Modiolar proximity of three perimodiolar cochlear implant electrodes’, Acta Otolaryngol. 122(4), 363–369. [66] Pasanisi, E., Vincenti, V., Bacciu, A., Guida, M., Bacciu, S., (2002), ‘The nucleus contour electrode array: an electrophysiological study’, Laryngoscope, 112(9), 1653–1656. [67] Lenarz, T., Battmer, R.D., Goldring, J.E., Neuburger, J., Kuzma, J., Reuter, G., (2000), ‘New electrode concepts (modiolus-hugging electrodes)’, dv Otorhinolaryngol., 57, 347–353. [68] Moore, B.C., (2003), ‘Coding of sounds in the auditory system and its relevance to signal processing and coding in cochlear implants’, Otol Neurotol. 24(2), 243–254. [69] Rubinstein, J.T., Hong, R., (2003), ‘Signal coding in cochlear implants: exploiting stochastic effects of electrical stimulation’, Ann. Otol. Rhinol. Laryngol. Suppl., 191, 14–19. [70] Rubinstein, J.T., (2004), ‘How cochlear implants encode speech’, Curr. Opin. Otolaryngol. Head Neck Surg., 12(5), 444–448. [71] Fayad, J.N., Baino, T., Parisier, S.C., (2004), ‘Revision cochlear implant surgery: causes and outcome’, Otolaryngol. Head Neck Surg., 131(4), 429–432. [72] Geddes, L.A., Voorhees, W.D., Lagler, R., Riscili, C., Foster, K., Bourland, J.D., (1988), ‘Electrically produced artificial ventilation’, Med Instrum. 22(5), 263–71. [73] Potter, D., (1995), ‘Neuroprostheses for ventilatory support’, in: Brindley, G.S., Rushton DN. (eds.), Neuroprotheses. Clinical Neurology. 4 (1), 77–94. [74] Loeb, G.E., Peck, R.A., Moore, W.H., Hood, K., (2001), ‘BION System for Distributed Neural Prosthetic Interfaces’, Med. Eng. Phys., 23, 9–18. [75] Hendricks, H.T., Ijzerman, M.J., de Kroon, J.R., (2001), in: ’t Groen F.A., Zilvold, G., Functional electrical stimulation by means of the Ness Handmaster Orthosis in chronic stroke patients: an exploratory study’, Clin. Rehabil. 15(2), 217–220. [76] Snoek, G.J., Ijzerman M.J., (2000), in ’t Groen FA., Stoffers TS., Zilvold G. Use of the NESS handmaster to restore handfunction in quadriplegia: clinical experiences in ten patients’, Spinal Cord, 38(4), 244–249. [77] Prochazka, A., Gauthier, M., Wieler, M., Kenwell, Z., (1997), ‘The bionic glove: an electrical stimulator garment that provides controlled grasp and hand opening in quadriplegia’, Arch. Phys. Med. Rehabil. 78(6), 608–614.


119

[78] Popovic, D., Stojanovic, A., Pjanovic, A., Radosavljevic, S., Popovic, M., Jovic, S., Vulovic, D., (1999), ‘Clinical evaluation of the bionic glove’, Arch. Phys. Med. Rehabil. 80(3), 299–304. [79] Alon, G., McBride, K., (2003), ‘Persons with C5 or C6 quadriplegia achieve selected functional gains using a neuroprosthesis’, Arch. Phys. Med. Rehabil. 84(1), 119–124. [80] Kilgore, K. L., (1997), ‘An implanted upper-extremity neuroprosthesis’, J. Bone Joint. Surg. 79A, 533–541. [81] Smith, B., Peckham, P.H., Keith, M.W., Roscoe, D.D., (1987), ‘An externally powered, multi-channel, implantable stimulator for versatile control of paralyzed muscle’, IEEE Trans. Biomed. Eng. 34, 499–508. [82] Gorman, P.H., Wuolle, K.S., Peckham, P.H., Heydrick, D., (1997), ‘Patient selection for an upper extremity neuroprosthesis in quadriplegic individuals’, Spinal Cord. 35(9), 569–573. [83] Ferguson, K.A., Polando, G., Kobetic, R., Triolo, R.J., Marsolais, E.B., (1999), ‘Walking with a hybrid orthosis system’, Spinal Cord, 37(11), 800–804. [84] Agarwal, S., Kobetic, R., Nandurkar, r S., Marsolais, E.B., (2003), ‘Functional electrical stimulation for walking in paraplegia: 17-year follow-up of 2 cases’, J. Spinal. Cord. Med., 26(1), 86–91. [85] Kagaya, H., Sharma, M., Polando, G., Marsolais, E.B., (1998), ‘Reliability of closed double helix electrode for functional electrical stimulation’, Clin. Orthop. Jan, 346, 215–222. [86] Scheiner, A., Polando, G., Marsolais, E.B., (1994), ‘Design and clinical application of a double helix electrode for functional electrical stimulation’, IEEE Trans. Biomed. Eng. 41(5), 425–431. [87] Sharma, M., Marsolais, E.B., Polando, G., Triolo, R.J., Davis, J.A. Jr., Bhadra, N., Uhlir, J.P., (1998), ‘Implantation of a 16-channel functional electrical stimulation walking system’, Clin. Orthop., 347, 236–242. [88] Guiraud, D., Taroni, G., Denis, B., Couderc, P., Stieglitz, T., (2000), ‘Description of a Sixteen-Channel FES Implantable System’, Proc. 5th Ann. Int. Conf. IFESS., Aalborg/Denmark, 292–294. [89] Perkins, T.A., de N Donaldson, N., Hatcher, N.A., Swain, I.D., Wood, D.E., (2002), ‘Control of leg-powered paraplegic cycling using stimulation of the lumbo-sacral anterior spinal nerve roots’, IEEE Trans. Neural Syst. Rehabil. Eng. 10(3), 158–164. [90] Davis, R., Eckhouse, R., Patrick, J.F., Delehanty, A., (1987), ‘Computerized 22-channel stimulator for limb movement’, Appl. Neurophysiol. 50(1–6), 444–448.

120


[91] Davis, R., Eckhouse, R., Patrick, J.F., Delehanty, A., (1987), ‘Computer-controlled 22-channel stimulator for limb movement’, Acta Neurochir Suppl., Wien, 39, 117–120. [92] Brindley, G.S., (1972), ‘Electrode array for making long-lasting electrical connections to spinal roots’, J. Physiol., 222, 135–136. [93] Brindley, G.S., (1973), ‘Emptying the bladder by stimulating sacral ventral roots’, J. Physiol., 237, 15P–16P. [94] Donaldson, P.E.K., (1983), ‘The Cooper cable: an implantable multiconductor cable for neurological prostheses’, Med. Biol. Eng. Comp., 21, 371–374. [95] Brindley, G.S., Polkey, C.E., Rushton, D.N., Cardozo, L., (1986), 95Sacral anterior root stimulators for bladder control in paraplegia: the first 50 cases’, J. Neurol. Neurosurg. Psychol., 49, 1104–1114. [96] Jezernik, S., Craggs, M., Grill, W.M., Creasey, G., Rijkhoff, N.J., (2002), ‘Electrical stimulation for the treatment of bladder dysfunction: current status and future possibilities’, Neurol. Res., 24(5), 413–30. [97] Sauerwein, D., Ingunza, W., Fischer, J., Madersbacher, H., Polkey, C.E., Brindley, G.S., Colombel, P., Teddy, P., (1990), ‘Extradural implantation of sacral anterior root stimulators’, J. Neurol. Neurosurg. Psychiat., 53, 681–684. [98] Stief, C.G., Sauerwein, D., Thon, W.F., Allhoff, E.P., Jonas, U., (1992), ‘The influence of anterior root stimulation (S2) in deafferented spinal cord injury men on cavernous electrical activity’, J. Urol., 148, 107–110. [99] Brindley, G.S., (1995), ‘Neuroprostheses to restore male sexual or reproductive function’, in: Brindley, G.S., and Rushton DN. (eds) Neuroprostheses, Baillière Tindall, London Philadelphia Sydney Tokio Toronto, 15–20. [100] Liberson, W.T., Holmquest, H.J., Scott, D., Dow, A., (1961), ‘Functional Electrotherapy: Stimulation of the Peroneal Nerve Synchronized with the Swing Phase of Paraplegic Patients’, Arch. Phys. Med. Rehab. 42, 101–105. [101] Lyons, G.M., Sinkjaer, T., Burridge, J.H., Wilcox, D.J., (2002), ‘A review of portable FES based neural orthoses for the correction of drop foot’. IEEE Trans. Neural. Syst, Rehabil. Eng., 10(4), 260–279. [102] Burridge, J.H., Taylor, P.N., Hagan, S.A., Wood, D.E., Swain, I.D., (1997), ‘The effects of common peroneal stimulation on the effort and speed of walking: a randomized controlled trial with chronic paraplegic patients’, Clin. Rehabil., 11(3), 201–210. [103] Upshaw, B.J., Sinkjaer, T., (1997), ‘Natural versus artificial sensors applied in peroneal nerve stimulation’, Artif. Organs., 21(3), 227–231.


121

[104] Weber, D.J., Stein, R.B., Chan, K.M., Loeb, G.E., Richmond, F.J., Rolf, R., James, K., Chong, S.L., Thompson, A.K., Misiaszek, J., (2004), ‘Functional electrical stimulation using microstimulators to correct foot drop: a case study’, Can. J. Physiol. Pharmacol. 82(8–9), 784–792. [105] Melzack, R., (1993), ‘Pain: past, present and future’, Can. J. Exp. Psychol., 47(4), 615–629. [106] Alo, K.M., Holsheimer, J., (2002), ‘New trends in neuro-modulation for the management of neuropathic pain’, Neurosurgery, 50(4), 690– 704. [107] Piasecki, S.D., Jefferson, J.W., (2004), ‘Psychiatric complications of deep brain stimulation for Parkinson s disease’, J. Clin. Psychiatry, 65(6), 845–849. [108] Burkhard, P.R., Vingerhoets, F.J., Berney, A., Bogousslavsky, J., Villemure, J.G., Ghika, J., (2004), ‘Suicide after successful deep brain stimulation for movement disorders’, Neurology, 63(11), 2170–2172. [109] Senatus, P.B., McClelland, S., 3rd, Ferris, A.D., Ford, B., Winfield, L.M., Pullman, S.L., Yu, Q., McKhann, G.M., 2nd, Schneller, S.J., Goodman, R.R., (2004), ‘Implantation of bilateral deep brain stimulators in patients with Parkinson disease and preexisting cardiac pacemakers. Report of two cases’, J. Neurosurg., 101(6), 1073–1077. [110] Rodriguez-Oroz, M.C., Zamarbide, I., Guridi, J., Palmero, M.R., Obeso, J.,A., (2004), ‘Efficacy of deep brain stimulation of the subthalamic nucleus in Parkinson‘s disease 4 years after surgery: double blind and open label evaluation’, J. Neurol. Neurosurg. Psychiat. 75(10), 1382–1385. [111] Putzke, J.D., Wharen, R.E. Jr., Obwegeser, A.A., Wszolek, Z.K., Lucas, J.A., Turk, M.F., Uitti, R.J., (2004), ‘Thalamic deep brain stimulation for essential tremor: recommendations for long term outcome analysis’, Can. J. Neurol. Sci, 31(3), 333–342. [112] Kuncel, A.M., Grill, W.M., (2004), ‘Selection of stimulus parameters for deep brain stimulation’, Clin. Neurophysiol., 115(11), 2431–41. [113] Terry, R.S., Tarver, W.B., Zabara, J., (1991), ‘The implantable neurocybernetic prosthesis system’, Pacing Clin. Electrophysiol., 14(1), 86–93. [114] McLachlan, R.S., (1997), ‘Vagus nerve stimulation for intractable epilepsy: a review’, J. Clin. Neurophysiol., 14(5), 358–368. [115] Murphy, J.V., Patil, A., (2003), ‘Stimulation of the nervous system for the management of seizures: current and future developments’, CNS Drugs, 17(2), 101–115. ‘

122


[116] Murphy, J.V., (1999), ‘Left vagal nerve stimulation in children with medically refractory epilepsy. The Pediatric VNS Study Group’, J. Pediatr., 134(5), 563–566. [117] Hall, S.W., (2003), ‘Commerzializing Neuroprostheses: The business of putting the brain back in business’, BS Thesis in Molecular Biology, Princeton University, USA, 2003. [118] Hirshberg, L.M., Chiu, S., Frazier, J.A., (2005), ‘Emerging brain based interventions for children and adolescents: overview and clinical perspective’, Child Adolesc. Psychiatr. Clin. N. Am. 14(1), 1–19. [119] Macritchie, K.A., Young, A.H., (2001), ‘Emerging targets for the treatment of depressive disorder’, Expert Opin Ther Targets., 5(5), 601–612. [120] Holmes, M.D., Chang, M., Kapur, V., (2003), ‘Sleep apnea and excessive daytime somnolence induced by vagal nerve stimulation’ Neurology, 28;61(8), 1126–1129. [121] Galli, R., Limbruno, U., Pizzanelli, C., Giorgi, F.S., Lutzemberger, L., Strata, G., Pataleo, L., Mariani, M., Iudice, A., Murri, L., (2003), ‘Analysis of RR variability in drug-resistant epilepsy patients chronically treated with vagus nerve stimulation’, Auton. Neurosci., 29;107(1), 52–9. [122] Ness, T.J., Fillingim, R.B, Randich, A., Backensto, E.M., Faught, E., (2000), ‘Low intensity vagal nerve stimulation lowers human thermal pain thresholds’, Pain, 86(1–2), 81–85. [123] Meyer, J.U., (2002), ‘The retina implant—a BIOMEMS challenge’, Sens. Act. A., 97–98, 1–9. [124] Stensaas, S.S., Stensaas, L.J., (1978), ‘Histopathological evaluation of materials implanted in the cerebral cortex’, Acta Neuropath. (Berl.) 41, 145–155. [125] Loeb, G.E., Walker, A.E., Uematsu, S., Konigsmark, B.W., (1977), ‘Histological reaction to various conductive and dielectric films chronically implanted in the subdural space’, J. Biomed. Mat. Res. 11, 195–210. [126] Haggerty, H.S., Lusted, H.S., (1989), ‘Histological reaction to polyimide films in the cochlea’, Acta Otolaryngol. 107, 13–22. [127] Richardson, R.R., Miller, J.A., Reichert, W.M., (1993), ‘Polyimides as biomaterials: preliminary bio-compatibility testing’, J. Biomat. 14, 627–635. [128] Mastrototaro, J.J., Massoud, H.Z., Pilkington, T.C., Ideker, R.E., (1992), ‘Rigid and flexible thin film multi-electrode arrays for transmural cardiac recording’, IEEE Trans. Biomed. Eng. 39, 567–577.


123

[129] Kimura, W., Kimura, T., Sawada, T., Taguchi, H., Hirokane, C., Ohte, A., Hitani, H., (1996), ‘Post operative decentration of three-piece silicone intra–ocular lenses’, J. Cataract. Refract. Surg. 22 Suppl 2, 1277–1280. [130] van den Puije, P.D., Pon, C.R., Robillard, H., (1989), ‘Cylindrical Cochlear Electrode Arraly for Use in Humans’, Ann. Otol. Rhinol. Laryngol., 98, 466–471. [131] Hetke, J.F., Najafi, K., Wise, K.D., (1990), ‘Flexible miniature ribbon cables for long term connection to implantable sensors’, Sens. Act., A21–A23, 999–1002. [132] Hetke, J. F., Lund, J.L., Najafi, K., Wise, D., Anderson, D.J., (1994), ‘Silicon ribbon cables for chronically implantable micro electrode arrays’, IEEE Trans. Biomed. Eng. 41, 314–321. [133] Najafi, K., Wise, K.D., Mochizuki, T., (1985), ‘A high-yield ICcompatible multi-channel recording array’, IEEE Trans. Electron Devices, 32, 1206–1211. [134] Najafi, K., Wise, K.D., (1986), ‘An implantable multi-electrode array with on-chip signal processing’. IEEE J. Solid State Circuits, 21, 1035–1044. [135] Ji, J., Najafi, K., Wise, K.D., (1990). ‘A scaled electronicallyconfigurable multi-channel recording array’, Sens. Act., A21–A23, 589–591. [136] Haemmerle, H., Kobuch, K., Kohler, K., Nisch, W., Sachs, H., Stelzle, M., (2002), ‘Bio-stability of micro-photodiode arrays for subretinal implantation’, Biomat; 23, 797–804. [137] Connolly, P., Moores, G.R., Monaghan, W., Shen, J., Britland, S., Clark, P., (1992), ‘Microelectronic and nanoelectronic interfacing techniques for biological systems’, Sens and Act B., 6, 113–121. [138] Edell, D.J., Devaney, L., Durfee, W.K., Gleason, E.F., Gleason, K.K., Herndon, T.O., Thompson, C.V., Wrighton, M.S., (1993), ‘Bioresistant integrated circuits’, Quarterly Progress Report, NIH-NINDS NO1-NS-0-2399, Neural Prosthesis Program No. 9. [139] Vogt, M., Hauptmann, R., (1995), ‘Plasma-deposited passivation layers for moisture and water protection’, Surface Coatings and Technology, 74–75, 676–681. [140] Cogan, S.F., Liu, Y.-P., Mc Caffrey, M., Twardoch, U.M., (1993), ‘Thin film hermetic coatings’, Quarterly Progress Report, NIHNINDS N44-NS-2-2311, Neural Prosthesis Program No.5. [141] Edell, D.J., (1986), ‘A peripheral nerve information transducer for amputees: Long term multi-channel recordings from rabbit peripheral nerves’, IEEE Trans. Biomed. Eng. 33, 203–214.

124


[142] Stieglitz, T., Beutel, H., Schuettler, M., Meyer, J.H., (2000), ‘Micromachined, polyimide based devices for flexible neural interfaces’, Biomedical Micro-devices, 2 (4), 283–294. [143] Navarro, X., Calvet, S., Rodríguez, F.J., Stieglitz, T., Blau, C., Butí, M., Valderrama, E., Meyer, J.H., (1998), ‘Stimulation and recording from regenerated peripheral nerves through polyimide sieve electrodes’, J. Periph. Nerv. System 2, 91–101. [144] Rodríguez, FJ, Ceballos, D., Schuettler, M., Valero, A., Valderrama, E., Stieglitz, T., Navarro, X., (2000), ‘Polyimide cuff electrodes for peripheral nerve stimulation’, J. Neurosci. Meth., 98, 105–118. [145] Noordegraaf, J., (1997), ‘Conformal coating using parylene polymers’, Med. Device Technol., 8(1), 14–20. [146] Schneider, A., Stieglitz, T., (2004), ‘Implantable flexible electrodes for functional electrical stimulation’, Med. Device Technol., 15(1), 16–18. [147] Stieglitz, T., (2004), ‘Considerations on surface and structural bio-compatibility as prerequisite for long term stability of neural prostheses’, J. Nanosci. Nanotechnol. 4, 496–503. [148] Stieglitz, T., Schuettler, M., Schneider, A., Valderrama, E., Navarro, X., (2003), ‘Non-invasive measurement of torque development in the rat‘s foot during electrical stimulation: measurement setup and first results from acute stimulation of the sciatic nerve with polyimide based cuff electrodes’, IEEE Trans. on Neural Systems and Rehab. Eng. 11(4), 427–437. [149] Naples, G.G., Mortimer, J.T., Yuen T.G.H., (1990), ‘Overview of peripheral nerve electrode design and implantation’, in Agnew, W.F., McCreery, D.B., eds., Neural prostheses: Fundamental studies, Biophysics and Bioengineering Series, Prentice Hall, New Jersey, 5, 107–144. [150] Stieglitz, T., Meyer, J.H., (1998), ‘Microtechnical interfaces to neurons’, in Manz, A., Becker, H., eds. Topics in Current Chemistry Series, Micro system Technology in Chemistry and Life Science, Berlin, Heidelberg, Springer, 194, 131–162. [151] Foerster, O., (1992), ‘Beiträge zur Pathophysiologie der Sehbahn und der Sehsphäre’, J. Psychol. neurol., 39, 435–68. [152] Alteheld, N., Roessler, G., Vobig, M., Walter, P., (2004). ‘The retina implant—a new approach for a visual prosthesis’, Biomed. Techn., 49 (4), 99–103.


125

[153] Margalit, E., Maia, M., Weiland, J.D., Greenberg, R.J., Fujii, G.Y., Torres, G., Piyathaisere, D.V., O‘Hearn, T.M., Liu, W., Lazzi, G., Dagnelie, G., Scribner, D.A., de Juan, E., Humayun, M.S., (2002), ‘Retinal Prosthesis for the blind’, Survey of Ophthalmology, 47(4), 335–356. [154] Veraart, C., Duret, F., Brelén, M., Oozeer, M., Delbeke, J., (2004), ‘Vision rehabilitation in the case of blindness’, Expert Rev. Medical Devices, 1(1), 139–153. [155] Brindley, G.S., Lewin, W.S., (1968), ‘The sensations produced by electrical stimulation of the visual cortex’, J. Physiol. (Lond), 196(2), 479–493. [156] Brindley, G.S., Rushton, D., (1974), ‘Implanted stimulators of the visual cortex as visual prosthetic devices’, Trans. Am. Acad. Ophthalmol. Otolaryngol., 78, 741–745. [157] Cogan, D.G., (1969), ‘So the blind may ‘see’’. New England Journal of Medicine, 281, 215–216. [158] Dobelle, W.H., Mladejovsky, M.G., Evans, J.R., et al., (1976), ‘‘Braille’ reading by a blind volunteer by visual cortex stimulation’, Nature 259, 111–112. [159] Dobelle, W.H., Mladejovsky, M.G., (1974), ‘Phosphenes produced by electrical stimulation of human occipital cortex, and their application to the development of a prosthesis for the blind’, J. Physiol., 243, 553–76, 1974. [160] Pollen, D.A., (1977), ‘Responses of single neurons to electrical stimulation of the surface of the visual cortex’, Brain Behav. Evol., 14, 67–86. [161] Bak, M., Girvin, J.P., Hambrecht, F.T., Kufta, C.V., Loeb, G.E., Schmidt, E.M., (1990), ‘Visual sensations produced by intracortical microstimulation od the human occipital cortex’, Med. Biol. Eng. Comp. 28, 257–259. [162] Maynard, E.M., Nordhausen, C.T., Normann, R.A., (1997), ‘The Utah intracortical electrode array: a recording structure for potential brain-computer interfaces’, Electroencephal. Clin. Neurphysiol. 102, 228–239. [163] Normann, R.A., Maynard, E.M., Rousche, P.J., Warren, D.J., (1999), ‘A neural interface for a cortical vision prosthesis’, Vision Res., 39, 2577-2587. [164] Normann, R.A., Warren, D.J., Ammermuller, J., Fernandez, E., Guillory, S., (2001), ‘High-resolution spatio-temporal mapping of visual pathways using multi-electrode arrays’, Vision Res., 41(10–11), 1261–75.

126


[165] Fernandez, E., Ahnelt, P., Rabischong, P., Botella, C., Garcia-de Quiros, F., Bonomini, P., Marin, C., Climent, R., Tormos, J.M., Normann, R.A., (2002), ‘Towards a cortical visual neuroprosthesis for the blind’, IFMBE Proceedings Vienna 3(2), 1690–1691. [166] Tanghe, S., Wise, K., (1992), ‘A 16-channel CMOS neural stimulating array’, IEEE Solid –State Curcuits, 27, 1819–1825. [167] Troyk, P., Bak, M., Berg, J., Bradley, D., Cogan, S., Erickson, R., Kufta, C., McCreery, D., Schmidt, E., Towle, V., (2003), ‘A Model for intracortical visual prosthesis research’, Artif. Organs 27(11), 1005–1015. [168] Delbeke, J., Wanet-Defalque, M.C., Gerard, B., Troosters, M., Michaux, G., Veraart, C., (2002), ‘The micro systems based visual prosthesis for optic nerve stimulation’, Artif. Organs 26(3), 232–234. [169] Veraart, C., Raftopoulos, C., Mortimer, J.T., Delbeke, J., Pins, J., Michaux, G., Venlierde, A., Parrini, S., Wanet-Delfalque, M.C., (1998), ‘Visual sensations produced by optic nerve stimulation using an implanted self-sizing spiral cuff electrode’, Brain Res., 813, 181– 186. [170] Lambert, V., Laloyaux, C., Schmitt C., Gerard, B., Delbeke, J.J, Veraart, C., (2003), ‘Localisation, discrimination, and grasping of daily life objects with an implanted optic nerve prosthesis’, ARVO Confererence, Ft.Lauderdale, USA, PGM Nr. 4208. [171] Veraart, C., Wanet-Delfalque, M.C., Gérard, B., Venlierde, A., Delbeke, J., (2003), ‘Pattern recognition with the optical nerve visual prosthesis’, Artif. Organs, 27(11), 996–1004. [172] Delbeke, J., Oozeer, M., Veraart, C., (2003), ‘Position, size and luminosity of phosphenes generated by direct optic nerve stimulation’, Vision Res., 43(9), 1091–1102. [173] Zrenner, E., Stett, A., Weiss, S. et al., (1999), ‘Can micro-photodiode arrays replace degenerated retinal photoreceptors to restore vision?’, Vision Res., 39(15), 2555–2567. [174] Stett, A., Barth, W. , Weiss, S., Haemmerle, H., Zrenner, E., (2000), ‘Electrical multi-site stimulation of the isolated chicken retina’, Vision Res., 40(13), 1785–1795. [175] Schwahn, H.N., Gekeler, F., Kohler, K., Kobuch, K., Sachs, H.G., Schulmeyer, F., Jakob, W., Gabel, V.P., Zrenner, E., (2001), ‘Studies on the feasibility of a subretinal visual prosthesis: data from Yucatan micropig and rabbit’, Graefes Arch. Clin. Exp. Ophthalmol. 239(12), 961–967.


127

[176] Parowing, M.T., Stubbs, E.B. Jr., Perlman, J.I., et al., (2001), ‘Immunohistochemical studies of the retina following long term implantation with subretinal micro-photodiode arrays’, Exp. Eye Res., 73(3), 333–343. [177] Chow, A.Y., Parowing, M.T., Chow, V.Y., et al., (2001), ‘Implantation of silicon chip micro-photodiode arrays in the cat subdural space’, IEEE Trans Neural Syst. Rehab. Eng., 9(1), 86–95. [178] Chow, A.Y., Parowing, M.T., Perlman, J.I., Ball, S.L., Chow, V.Y., Hetling, J.R., Peyman, G.A., Liang, C., Stubbs, E.B., Peachey, N.S., (2002), ‘Subretinal implantation of semiconductor based photodiodes: durability of novel implant designs’, J. Rehabil. Res. Dev., 39(3), 313–321. [179] Guenther, E., Troger, B., Schlosshauer, B., Zrenner, E., (1994), ‘Long term survival of retinal cell cultures on retinal implant materials’, Vision Res., 39(24), 3988–3994. [180] Zrenner, E., Gekeler, F., Gabel, V.P., Graf, H.G., Graf, M., Guenther, E., Haemmerle, H., Hoefflinger, B., Kobuch, K., Kohler, K., Nisch, W., Sachs, H., Schlosshauer, B., Schubert, M., Schwahn, H., Stelzle, M., Stett, A., Troeger, B., Weiss, S., (2001), ‘Subretinal microphotodiode array as replacement for degenerated photoreceptors?’, Ophthalmologe, 98(4), 357–363. [181] Blatsios, G., Bondzio, I., Sailer, H., Shinoda, K., Kobuch, K., Sachs, H., Gabel, V.P., Zrenner, E., (2003), ‘Temperature change in the subretinal space during infrared irradiation’, ARVO Confererence, Ft.Lauderdale, USA, PGM Nr. 5077. [182] Chow, A.Y., Packo, K.H., Pollack, J.S., Schuchard, R.A., (2003), ‘Subretinal artificial silicon retina microchip implantation in retinitis pigmentosa patients: long term follow-up’, ARVO Confererence, Ft.Lauderdale, USA, PGM Nr.4205. [183] Chow, A.Y., Chow, V.Y., Packo, K.H. et al., (2004), ‘The artificial silicon retina microchip for the treatment of vision loss from retinitis pigmentosa’, Arch. Ophthalmol., 122(4), 460–469. [184] Eckmiller, R., (1997), ‘Learning retina implants with epiretinal contacts’, Ophthalmic Res., 29, 281–289. [185] Stieglitz, T., Beutel, H., Keller, R., Blau, C., Meyer, J.H., (1997), ‘Development of flexible stimulation devices for a retina implant system’, Proceedings of the 19th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Oct 30–Nov 2, 1997, Chicago (USA), 2307–2310.

128


[186] Stieglitz, T., Keller, R., Beutel, H., Ruf, H.H., Robitzki, A.A., Meyer, J.H., (1999), ‘Hybrid integration of a retina implant system for people suffering from retinitis pigmentosa’, Med. Biol. Eng.Comp., 37 Suppl. 2, 1102–1103. [187] Stieglitz, T., Keller, R., Beutel, H. Meyer, J.-U., (2000), ‘Micro system integration techniques for intra–ocular vision prostheses using flexible polyimide-foils’, Proceedings of the MICRO.tec 2000, 467–472. [188] Walter, P., Szurman, P., Vobig, M., Berk, H., Lüdtke-Handjery, H.C., Richter, H., Mittermayer, C., Heimann, K., Sellhaus, B., (1999), ‘Successful long term implantation of electrically inactive epiretinal micro electrode arrays in rabbits’, Retina, 19, 546–552. [189] Stieglitz, T., Habere, W. , Lau, C., Goertz, M., (2004), ‘Development of an inductively coupled epiretinal vision prosthesis’, 26th Ann. Int. Conf. of the IEEE Eng. in Med. and Biol. Soc., San Francisco, CA (USA), 4178–4181. [190] Wilms, M., Eger, M., Schanze, T., Eckhorn, R., (2003), ‘Visual resolution with epiretinal electrical stimulation estimated from activation profiles in the cat visual cortex’, Vis. Neurosci, 20(5), 543–555. [191] Eysel, U.T., Walter, P., Gekeler, F., Schwahn, H., Zrenner, E., Sachs, H.G., Gabel, V.P., Kisvárday, Z.F., (2002), ‘Optical imaging reveals 2-dimensional patterns of cortical activation after local retinal stimulation with sub- and epiretinal visual prostheses’, ARVO Confererence 2002, Ft.Lauderdale, USA, PGM Nr. 4486. [192] Walter, P., Kisvárday, Z.F., Görtz, M., Alteheld, N., Rössler, G., Stieglitz, T., Eysel, U.T., (2005), ‘Cortical activation with a completely implanted wireless retinal prosthesis’, Invest. Ophthal. Vis. Sci., (in press). [193] Hornig, R., Laube, T., Walter, P. et al., (2005), ‘A methodand technical equipment for an acute human trial to evaluate retinal implant technology’, J. Neural. Eng., 2, 139–134. [194] Humayun, M.S., de Juan, E., Jr., Weiland, J.D. et al., (1999), ‘Pattern electrical stimualtion of the human retina’, Vision Res. 39(15), 2569–2576. [195] Humayun, M.S., Weiland, J.D., Fujii, G.Y., Greenberg, R., Williamson, R., Little, J., Mech, B., Cimmarusti, V., van Boemel, G., Dagnelie, G., de Juan, E., (2003), ‘Visual perception in a blind subject with a chronic microelectronic retinal prosthesis’, Vision Res., 43(24), 2573–2581.


129

[196] Freda, R., Fujii, G., Agrawal, R.J., et al., (2005), ‘Epiretinal Prosthesis Implantations in humans’, ARVO Confererence, Ft.Lauderdale, USA, PGM Nr.1493, 2005. [197] Humayun, M.S., Freda, R., Fine, I., et al., (2005), ‘Implanted intra– ocular retinal prosthesis in six blind subjects’, ARVO Confererence, Ft.Lauderdale, USA, PGM Nr.1144, 2005. [198] Kerdraon, Y.A., Downie, J.A., Suaning, G.J., et al., (2002), ‘Development and surgical implantation of a vision prosthesis model into the ovine eye’, Clinical and Exp. Ophthalmol., 30, 36–40. [199] Suaning, G.J., Lovell, N.H., (2001), ‘CMOS neuro-stimulation ASIC with 100 channels, scalable output, and bidirectional radio-frequency telemetry’, IEEE Trans. Biomed. Eng., 48(2), 248–260. [200] Schuettler, M., Stiess, S., King, B.V., Suaning, G.J., (2005), ‘Fabrication of implantable micro electrode arrays by laser cutting of silicone rubber and platinum foil’, J. Neural. Eng. 2, S121–S128. [201] Peterman, M.C., Mehenti, N.Z., Bilbao, K.V., Lee, C.L., et al., (2003), ‘The artificial synapse chip: a flexible retinal interface on directed retinal cell growth and neurotransmitter stimulation’, Art. Organs, 27(11), 975–985. [202] Navarro, X., Krueger, T.B., Lago, N., Micera, S., Stieglitz, T., Dario, P., (2005), ‘A critical review of interfaces with the peripheral nervous system for the control of neuroprostheses and hybrid bionic systems’, J. Periph. Nervous Syst., (submitted 2005). [203] Rutten, W.L.C., (2002), ‘Selective electrical interfaces with the nervous system’, Annu. Rev. Biomed. Eng., 4, 407–452. [204] Akers, J.M., Peckham, H.P., Keith, M.W., Merritt, K., (1997), ‘Tissue response to chronically stimulated implanted epimysial and intramuscular electrodes’, IEEE Trans. Rehab. Eng., 5, 207–220. [205] Naples, G.G., Mortimer, J.T., Scheiner, A., Sweeney, J.D., (1988), ‘A spiral nerve cuff electrode for peripheral nerve stimulation’, IEEE Trans. Biomed. Eng., 35(11), 905–916. [206] Sweeney, J.D., Ksienski, D.A., Mortimer, J.T., (1990), ‘A nerve cuff technique for selective excitation of peripheral nerve trunk regions’, IEEE Trans. Biomed. Eng., 37(7), 706–715. [207] Sweeney, J.D., Crawford, N.R., Brandon, T.A., (1995), ‘Neuromuscular stimulation selectivity of multiple-contact nerve cuff electrode arrays’, Med. & Biol. Eng. & Comp., 33, 418–425. [208] Rozman, J., (1991), ‘Multi-electrode cuff for extraneural selective stimulation of serve fibers’, Proc. of the Ann. Int. Conf. IEEE EMBS., 914–915.

130


[209] Walter, J.S., McLane, J., Cai, W., Khan, T., Cogan, S., (1995), ‘Evaluation of a thin film peripheral nerve cuff electrode’, The Journal of Spinal Cord Medicine 18, 28–32. [210] Loeb, G.E., Peck, R. A., (1996), ‘Cuff electrodes for chronic stimulation and recording of peripheral nerve activity’, J. Neurosci. Meth., 64, 95–103. [211] Rodríguez, F.J., Ceballos, D., Schüttler, M., Valderrama, E., Stieglitz, T., Navarro, X., (2000), ‘Polyimide cuff electrodes for peripheral nerve stimulation’, J. Neurosci. Meth., 98, 105–118. [212] Grill, W.M., Mortimer, J.T., (1996), ‘Non-invasive measurement of the input-output properties of peripheral nerve stimulating electrodes’, J. Neurosci. Meth., 65, 43–50. [213] Grill, W.M., Mortimer, J.T., (1996), ‘Quantification of recruitment properties of multiple contact cuff electrode’, IEEE Trans. Rehab. Eng. 4, 49–62. [214] Grill, W.M., Mortimer, J.T., (1999), ‘Stability of the input-output properties of chronically implanted multiple contact nerve cuff stimulating electrodes’, IEEE Trans. Rehab. Eng., 6, 364–373. [215] Navarro, X., Valderrama, E., Stieglitz, T., Schuettler, M., (2001), ‘Selective fascicular stimulation of the rat sciatic nerve with multi-polar polyimide cuff electrodes’, Restor. Neuro Neurosurg., 18(1), 9–21. [216] Stieglitz, T., Koch, K.P., Schuettler, M., (2005), ‘Flexible, polyimide based modular implantable biomedical micro systems for neural prostheses’, IEEE Eng. Med. Biol. Mag., (in press). [217] Tyler, D.J., Durand, D.M., (2002), ‘Functionally selective peripheral nerve stimulation with a flat interface nerve electrode’, IEEE Trans. Neural Syst. Rehabil. Eng., 10, 294–303. [218] Leventhal, D.K., Durand, D.M., (2004), ‘Chronic measurement of the stimulation selectivity of the flat interface nerve electrode’, IEEE Trans. Biomed. Eng. 51, 1649–1658. [219] Tyler, D.J., Durand, D.M., (2003), ‘Chronic response of the rat sciatic nerve to the flat interface nerve electrode’, Ann. Biomed. Eng., 31, 633–642. [220] Tyler, D.J., Durand, D.M., (1994), ‘Interfascicular electrical stimulation for selectively activating axons’, IEEE EMBS Mag., 13, 575–583. [221] Tyler, D.J., Durand, D.M., (1997), ‘A slowly penetrating interfascicular nerve electrode for selective activation of peripheral nerves’, IEEE Trans Rehabil. Eng. 5, 51–61.


131

[222] Malagodi, M.S., Horch, K.W., Schoenberg, A.A., (1989), ‘An intrafascicular electrode for recording of action potentials in peripheral nerves’, Ann. Biomed. Eng., 17, 397–410. [223] McNaughton, T.G., Horch, K.W., (1996), ‘Metallized polymer fibers as leadwires and intrafascicular micro electrodes’, J. Neurosci. Meth., 70, 103–110. [224] Yoshida, K., Jovanovic, K., Stein, R.B., (2000), ‘Intrafascicular electrodes for stimulations and recording from mudpuppy spinal roots’, J. Neurosci. Meth., 96, 47–55. [225] Lawrence, S.M., Dhillon, G.S., Horch, K.W., (2003), ‘Fabrication and characteristics of an implantable, polymer based, intrafascicular electrode’, J. Neurosci. Meth., 131, 9–26. [226] Lawrence, S.M., Dhillon, G.S., Jensen, W., Yoshida, K., Horch, K.W., (2004), ‘Acute peripheral nerve recording characteristics of polymer based longitudinal intrafascicular electrodes’, IEEE Trans. Neural Syst. Rehabil. Eng., 12, 345–348. [227] Lefurge, T., Goodall, E., Horch, K., Stensaas, L., Schoenberg, A., (1991), ‘Chronically implanted intrafascicular recording electrodes’, Ann. Biomed. Eng., 19, 197–207. [228] Lawrence, S.M., Larsen, J.O., Horch, K.W., Riso, R., Sinkjaer, T., (2002), ‘Long term bio-compatibility of implanted polymer based intrafascicular electrodes’, J. Biomed. Mater. Res., 63, 501–506. [229] Nannini, N., Horch, K., (1991), ‘Muscle recruitment with intrafascicular electrodes’, IEEE Trans. Biomed. Eng., 38, 769–776. [230] Yoshida, K., Horch, K., (1993), ‘Selective stimulation of peripheral nerve fibers using dual intrafascicular electrodes’, IEEE Trans. Biomed. Eng., 40, 492–494. [231] Goodall, E.V., Lefurge, T.M., Horch, K.W., (1991), ‘Information contained in sensory nerve recordings made with intrafascicular electrodes’, IEEE Trans. Biomed. Eng., 38, 846–850. [232] McNaughton, T.G., Horch, K.W., (1994), ‘Action potential classification with dual channel intrafascicular electrodes’, IEEE Trans. Biomed. Eng., 41, 609–616. [233] Dhillon, G.S., Lawrence, S.M., Hutchinson, D.T., Horch, K., (2004), ‘Residual function in peripheral nerve stumps of amputees: implications for neural control of artificial limbs’, J. Hand Surg., 29A, 605–615. [234] Vallbo, A., Hagbarth, K., Torebjörk, H., Wallin, G., (1979), ‘Somatosensory, proprioceptive, and sympathetic activity in human peripheral nerves’, Physiol. Rev., 59, 919–957.

132


[235] Hagbarth, K.E., (2002), ‘Micro electrode recordings from human peripheral nerves (microneurography)’, Muscle Nerve, Suppl 11, S28–35. [236] Hagbarth, K.E., (1993), ‘Microneurography and applications to issues of motor control. Fifth Annual Stuart Reiner Memorial Lecture’, Muscle Nerve, 16, 693–705. [237] Smit, J.P., Rutten, W.L., Boom, H.B., (1999), ‘Endoneural selective stimulating using wire-micro electrode arrays’, IEEE Trans Rehabil. Eng., 7, 399–412. [238] Drake, K.L.,Wise, K.D., Farraye, J., Anderson, D.J., BeMent, S.L., (1988), ‘Performance of planar multi-site microprobes in recording extra–cellular single-unit intracortical activity’, IEEE Trans. Biomed. Eng., 35, 719–732. [239] Ehrfeld, W., Munchmeyer, D., (1991) ‘LIGA method, threedimensional micro-fabrication using synchrotron radiation’, Nucl. Instr. Methods Phys. Res., A 303, 523–531. [240] Normann, R.A., Maynard, E.M., Rousche, P.J., Warren, D.J., (1998), ‘A neural interface for a cortical vision prosthesis’, Vision Res., 12, 2577–2587. [241] Yoon, T.H., Hwang, E.J., Shin, D.Y., Park, S.I., Oh, S.J., Jung, S.C., Shin, H.C., Kim, S.J., (2000), ‘A micro-machined silicon depth probe for multi-channel neural recording’, IEEE Trans. Biomed. Eng., 47, 1082–1087. [242] Stieglitz, T., Gross, M., (2002), ‘Flexible BIOMEMS with electrode arrangements on front and back side as key component in neural prostheses and biohybrid systems’, Sens. Act., B 83, 8–14. [243] Takeuchi, S., Suzuki, T., Mabuchi, K., Fujita, H., (2004), ‘3D flexible multi-channel neural probe array’, J. Micromech. Microeng., 14, 104–107. [244] Hoogerwerf, A.C., Wise, K.D., (1994), ‘A 3D micro-electrode array for chronic neural recording’, IEEE Trans. Biomed. Eng. 41, 1136–1146. [245] Kim, C., Wise, K.D., (1996), ‘A 64-site multi-shank CMOS lowprofile neural stimulating probe’, IEEE J. Solid State Circuits 31, 1230–1238. [246] Kipke, D.R. Vetter, R.J., Williams, J.C., Hetke, J.F., (2003), ‘Siliconsubstrate intracortical micro electrode arrays for long term recording of neuronal spike activity in cerebral cortex’, IEEE Trans Rehabil. Eng., 11, 151–155. [247] Nordhausen, C.T., Maynard, E.M., Normann, R.A., (1996), ‘Single unit recording capabilities of a 100 microlectrode array’, Brain Res., 726, 129–140.


133

[248] Rutten, W.L.C., Smit, J.P.A., Rozijn, T.H., Meier, J.H., (1995), ‘3D neuro-electronic interface devices for neuromuscular control: design studies and realisation steps’, Biosens. Bioelectron., 10, 141–153. [249] Rutten, W.L.C., Smit, J.P.A., Frieswijk, T.A., Bielen, J.A., Brouwer, A.L.H., Buitenweg, J.R., Heida, C., (1999), ‘Neuro-electronic interfacing with multi-electrode arrays’, IEEE Eng. Med. Biol., 18, 47–55. [250] Branner, A., Stein, R.B., Normann, R.A., (2001), ‘Selective stimulation of cat sciatic nerve using an array of varying-length micro electrodes’, J. Neurophysiol., 85, 1585–1594. [251] Branner, A., Normann, R.A., (2000), ‘A multi-electrode array for intrafascicular recording and stimulation in sciatic nerve of cats’, Brain Res. Bull, 51, 293–306. [252] Branner, A., Stein, R.B., Fernandez, E., Aoyagi, Y., Normann, R.A., (2004), ‘Long term stimulation and recording with a penetrating micro electrode array in cat sciatic nerve’, IEEE Trans. Biomed. Eng. 51, 146–157. [253] Warwick, K., Gasson, M., Hutt, B., Goodhew, I., Kyberd, P., Andrews, B., Teddy, P., Shad, A., (2003), ‘The application of implant technology for cybernetic systems’, Arch. Neurol., 60, 1369–1373. [254] Keidel, W.D., (1989), ‘Biokybernetik des Menschen’, Wissenschaftliche Buchgesellschaft, Darmstadt. [255] Llinás, R., Nicholson, C., Johnson, K., (1973), ‘Implantable monolithic wafer recording electrodes for neurophysiology’, in: Phillips, M.I, ed., Brain unit activity during behaviour., Charles Thomas, Illinois, 105–110. [256] Mannard, A., Stein, R.B., Charles, D., (1974), ‘Regeneration electrode units: implants for recording from single peripheral nerve fibers in freely moving animals’, Science, 183, 547–549. [257] Akin, T., Najafi, K., (1991), ‘A micro-machined silicon sieve electrode for nerve regeneration applications’, Proc. Int. Conf. Solid State Sensors and Actuators, 128–131. [258] Akin, T., Najafi, K., Smoke, R.H., Bradley, R.M., (1994), ‘A micromachined silicon sieve electrode for nerve regeneration applications’, IEEE Trans. Biomed. Eng. 41, 305–313. [259] Kovacs, G.T.A., (1991), ‘Regeneration micro electrode arrays for direct interface to nerves’, Proc. Int. Conf. of Solid State Sensors and Actuators, 116–119. [260] Kovacs, G.T.A., Storment, C.W., Rosen, J.M., (1992), ‘Regeneration micro electrode array for peripheral nerve recording and stimulation’, IEEE Trans. Biomed. Eng., 39, 893–902.

134


[261] Bradley, R.M., Smoke, R.H., Akin, T., Najafi, K., (1992), ‘Functional regeneration of glossopharyngeal nerve through micro-machined sieve slectrode arrays’, Brain Res., 1–7. [262] Kovacs, G.T.A., Storment, C.W., Halks-Miller, M., Belczynski, C.R., Della Santina, C.C., Lewis, E.R., Maluf, N.I., (1994), ‘Siliconsubstrate micro electrode array for parallel recording of neural activity in peripheral and cranial nerves’, IEEE Trans. Biomed. Eng., 41, 567– 577. [263] Dario, P., Garzella, P., Toro, M., Micera, S., Alavi, M., Meyer, U., Valderrama, E., Sebastiani, L., Ghelarducci, B., Mazzoni, C., Pastacaldi, P., (1998), ‘Neural interfaces for regenerated nerve stimulation and recording’, IEEE Trans Rehabil. Eng., 6, 353–363. [264] Wallman, L., Levinsson, A., Schouenberg, J., Holmberg, H., Montelius, L., Danielsen, N., Laurell, T., (1999), ‘Perforated silicon nerve chips with doped registration electrodes: In vitro performance and In vivo operation’, IEEE Trans. Biomed. Eng. 46(9), 1065–1073. [265] Mensinger, A.F., Anderson, D.J., Buchko, C.J., Johnson, M.A., Martin, D.C., Tresco, P.A., Silver, R.B., Highstein, S.M., (2000), ‘Chronic recording of regenerating VIIIth nerve axons with a sieve electrode’, J. Neurophysiol., 83, 611–615. [266] Navarro, X., Calvet, S., Butí, M., Gómez, N., Cabruja, E., Garrido, P., Villa, R., Valderrama, E., (1996), ‘Peripheral nerve regeneration through micro electrode arrays based on silicon technology’, Restor. Neurol. Neurosci., 9, 151–160, 1996. [267] Bradley, R.M., Cao, X., Akin, T., Najafi, K., (1997), ‘Long term chronic recordings from peripheral sensory fibers using a sieve electrode array’, J. Neurosci. Meth., 73, 177–186. [268] Della Santina, C.C., Kovacs, G.T.A., Lewis, E.R., (1997), ‘Multi-unit recording from regenerated bullfrog eighth nerve using implantable silicon-substrate micro electrodes’, J. Neurosci. Meth., 72, 71–86. [269] Zhao, Q., Drott, J., Laurell, T., Wallman, L., Lindström, K., Bjursten, L.M., Lundborg, G., Montelius, L., Danielsen, N., (1997), ‘Rat sciatic nerve regeneration through a micro-machined silicon chip’, Biomaterials 18, 75–80. [270] González, C., Rodríguez, M., (1997), ‘A flexible perforated micro electrode array probe for action potential recording in nerve and muscles tissues’, J. Neurosci. Meth., 72, 189–195. [271] Stieglitz, T., Beutel, H., Meyer, J.-U., (1997), ‘A flexible, light-weight multi-channel sieve electrode with integrated cables for interfacing regenerating peripheral nerves’, Sens. Act. A60, 240–243.


135

[272] Ceballos, D., Valero-Cabré, A., Valderrama, E., Schuettler, M., Stieglitz, T., Navarro, X., (2002), ‘Morphological and functional evaluation of peripheral nerve fibers regenerated through polyimide sieve electrodes over long term implantation’, J. Biomed. Mat. Res., 60, 517–528. [273] Lago, N., Ceballos, D., Rodríguez, F.J., Stieglitz, T., Navarro, X., (2004), ‘Long term assessment of axonal regeneration through polyimide regenerative electrodes to interface the peripheral nerve’, Biomaterials, published online. [274] Klinge, P.M., Vafa, M.A., Brinker, T., Brandes, A., Walter, G.F., Stieglitz, T., Samii, M., Wewetzer, K., (2001), ‘Molecular characterization of axonal sprouting and tissue changes after long term implantation of a polyimide sieve electrode to the transsected adult rat sciatic nerve’, Biomaterials 22, 2333–2343. [275] Heiduschka, P., Romann, I., Stieglitz, T., Thanos, S., (2001), ‘Perforated micro electrode arrays implanted in the regenerating adult central nervous system’, Exp. Neurol., 171(1), 1–10. [276] Wilkinson, D.W., Dow, J.A.T., Curtis, A.S.G., Connolly, P., Clark, P., (1987), ‘Topographical Control of Cell Behaviour’, Development, 99, 439–448. [277] Clark, P., Connolly, P., Curtis, A.S.G., Dow, J.A.T., Wilkinson, C.D.W., (1990), ‘Topographical conrol of cell behaviour: II. Multiple grooved substrata’, Development, 108, 635–644. [278] Fromherz, P., Schaden, H., Vetter, T., (1991), ‘Guided Outgrowth of Leech Neurons in Culture’, Neuroscience Letters, 129, 77–80. [279] Fromherz, P., Offenhäuser, A., Vetter, T., Weis J., (1991), ‘A NeuronSilcon Junction: A Retzius Cell of the Leech on an Insulated-Gate Field-Effect Transistor’, Science, 252, 1290–1293. [280] Droge, M.H., Gross, G.W., Hightower, M.H., Czisny, L.E., (1986), ‘Multi-electrode analysis of coordinated, multi-site, rhythmic bursting in cultured CNS monolayer networks’, J. Neurosci., 6, 1583–1592. [281] Gross, G.W., Kowalski, J., (1991), ‘Experimental and theoretical analysis of random nerve cell network dynamics’, in: Antognetti, P. and Milutinovic, V. (eds), Neural Networks: Concepts, Applications, and Implementations, Englewood City, New Jersey, Prentice-Hall, 47–110. [282] Jimbo, Y., Kawana, A., (1992), ‘Electrical Stimulation and Recording from Cultured Neurons Using a Planar Electrode Array’, Bioelectrochemistry and Bioenergetics, 29, 193–204.

136


[283] Jimbo, Y., Robinson, H.P.C., Kawana, A., (1993), ‘Simultaneous measurement of intra–cellular calcium and electrical activity from patterened neural networks in culture’, IEEE Trans. Biomed. Eng., 40(8), 804–810. [284] Tatic-Lucic, S., Tai, Y.-C., Wright, J.A., Pine, J., Denison, T., (1993), ‘Silicon-Micro-machined Neurochips for in vitro Studies of Cultured Neural Networks’, Proceedings, Int. Conf. Solid State Sensors and Actuators, 943–946. [285] Pine, J., Tai, Y.-C., Buzsaki, G., Bragin, A., Carpi, D., (1994), ‘The cultured neuron probe’, Quarterly Progress Report, NIH-NINDS NO1NS-3-2393, Neural Prosthesis Program No. 4. [286] Kennedy, P.R., (1989), ‘The cone electrode: a long-term electrode that records from neurites grown onto its recording surface’, J. Neurosci. Meth., 29, 181–193. [287] Katsuki, M., Atsuka, Y., Hirayama, T., (1997), ‘Reinnervation of denervated muscle by transplantation of fetal spinal cord to transected sciatic nerve in the rat’, Brain Res., 771, 31–36. [288] Thomas, C.K., Erb, D.E., Grumbles, r M., Bunge, R.P., (2000), ‘Embryonic cord transplants in peripheral nerve restore skeletal muscle function’, J. Neurophysiol., 84, 591–595. [289] Klinge, P.M., Groos, S., Wewetzer, K., Haastert, K., Rosahl, S.K., Vafa, M.A., Hosseini, H., Samii, M., Brinker, T., (2001), ‘Regeneration of a transsected nerve by transplantation of spinal cord encapsulated in a Vein’, Neuroreport, 12, 1271–1275. [290] Stieglitz, T., Ruf, H.H., Gross, M., Schuettler, M., Meyer, J.-U., (2002), ‘A biohybrid system to interface peripheral nerves after traumatic lesions: design of a high channel sieve electrode’, Biosens. Bioelectron., 17, 685–696. [291] Stieglitz, T., (2004), ‘The neuron microprobe project, personal communication’. [292] Patil, P.G., Carmena, J.M., Nicolelis, M.A., Turner, D.A., (2004), ‘Ensemble recordings of human subcortical neurons as a source of motor control signals for a brain-machine interface’, Neurosurgery, 55(1), 27–38. [293] Schwartz, A.B., (2004), ‘Cortical neural prosthetics’, Annu. Rev. Neurosci., 27, 487–507. [294] Serruya, M.D., Hatsopoulos, N.G., Paninski, L., Fellows, M.R., Donoghue, J.P., (2002), ‘Instant neural control of a movement signal’, Nature, 416, 141–142. [295] Saigal, R., Rwnzi, C., Mushahwar, V.K., (2004), ‘Intra–spinal Microstimulation generates functional movements after spinal-cord injury’, IEEE Trans Neural Syst. Rehab. Eng. 12(4), 430–440.


137

[296] Burke, R.E., Degtyarenko, A.M., Simon, E.S., (2001), ‘Patterns of locomotor drive to motoneurons and last-order interneurons: clues to the structure of the CPG’, J. Neurophysiol., 86(1), 447–462. [297] Dietz, V., Muller, R., Colombo, G., (2002),‘Locomotor activity in spinal man: significance of afferent input from joint and load receptors’, Brain, 125(Pt12), 2626–2634. [298] Hesse, S., Schmidt, H., Werner, C., Bardeleben, A., (2003),‘Upper and lower extremity robotic devices for rehabilitation and for studying motor control’, Curr. Opin. Neurol., 16(6), 705–710. [299] Burmeister, J.J., Moxon, K., Gerhardt, G.A., (2000), ‘Ceramic based multi-site micro electrodes for electrochemical recordings’, Anal. Chem., 72(1), 187–192. [300] Rousche, P.J., Pellinen, D.S., Pivin, D.P., Jr., et al., (2001), ‘Flexible polyimide based intracorticalrecording arrays with bioactive capability’, IEEE Trans. Biomed. Eng. 48, 361–371. [301] Metz, S., Holzer, R., Renaud, P., (2001), ‘Fabrication of flexible, implantable micro electrodes with embedded fluidic microchannels’, Transducers 01 Eurosensors XV: The 11th International Conference on Solid State Sensors and Actuators, 1210–1213. [302] Gross, M., Altpeter, D., Stieglitz, T., Schuettler, M., Meyer, J.H., (2002), ‘Micro-machining of flexible neural implants with low-ohmic wire traces using electroplating’, Sens. Act., A96, 105–110.

Chapter 5 MICRO-FLUIDIC PLATFORMS Koltay, P., Ducrée, J., Zengerle, R. IMTEK - University of Freiburg; Laboratory for MEMS Applications; Georges-Koehler-A - llee 106; 79110 Freiburg, Germany

Abstract:

A paramount factor for the commercial success of micro-fluidics technologies is the bundling of RTD efforts. In this chapter, we outline how this critical issue can be addressed by micro-fluidic platforms. These platforms comprise a set of unit operations for pumping, valving, metering, mixing and detecting which can be readily combined to form more complex applications. A platform also includes backend technologies, e.g., for the fabrication, simulation and testing of the devices. Different applications, e.g., for liquid handling and analytical applications, can n be realized on one common platform, thus appreciably saving RTD efforts by synergies and technological knowhow. This chapter reviews various examples of emerging micro-fluidic platforms.

Key words: microfluidics, rotating platforms, electrowetting, non-contact dispensing, dispensing well plate

1.

INTRODUCTION

Nowadays most MEMS related technology roadmaps and market studies point out the significant technological and scientific impact micro-fluidics will have on the scientific community and various industries, especially within the Life Sciences. A dedicated ‘micro-fluidics roadmap’ was prepared recently by the authors and others within the EU–sponsored ‘FlowMap’ project [1]. Within this study the economic development related to microfluidics technologies for the life sciences has been estimated and important market drivers and road blocks have been pinpointed. Furthermore, the paramount technology drivers which will determine the present and expected capabilities have been identified.


139

MICRO-FLUIDIC PLATFORMS M

140 0

Regarding the economic impact of micro-fluidics the FlowMap study anticipates an overall growth rate for micro-fluidics technologies and products in the life sciences of more than 30% per annum with

• • •

drug discovery, medical diagnostics, and therapeutic devices

representing the most promising fields. The overall global market of microfluidics in the life sciences has been estimated to be worth approximately 500 M€ in 2002, increasing with an assumed annual growth rate of 19% to 1.4 billion € in 2008. Besides the economic impact of micro-fluidics, technological trends have also been clearly identified within the roadmap. Amongst those the most relevant are:

• •

the need for micro-fluidic platforms equipped with a basic set of validated fluidic base operations to arrive rapidly at application specific micro-fluidic systems, the use of low-cost technologies such as injection moulding or low-cost materials such as PDMS, PCB or Polyimide for the prototyping and fabrication of micro-fluidic devices

These two major trends highlight the importance of decreasing costs and development times in order to make micro-fluidics applications commercially attractive. The costs are directly related to the costs of the base materials such as silicon, glass, or polymer materials, and of the applied fabrication technologies. Obviously replication technologies such as injection moulding are much more cost efficient than traditional lithographic MEMS technologies for high production numbers. In this chapter however, we want to focus on the other aspect, the need to reduce development costs and times. This necessity is reflected in the trend to use validated sets of compatible fluidic component to realise application specific micro-fluidic solutions in short time at low risk. This system oriented, so called ‘platform concept’ contrasts the frequent approaches in which individual components such as pumps [2] or valves [3] have been optimised at an individual level. The power of the ‘platform concept’ will be highlighted by discussing common criteria for successful micro-fluidic platforms. Owing to the limited space the discussion will be confined to four examples. This is by far from being a complete overview of the micro-fluidic platforms currently discussed in literature. They serve rather to substantiate the importance of the platform concept in general.


2.

141

WHAT IS A MICRO-FLUIDIC PLATFORM

Very similar to the ASIC industry in microelectronics which provides validated elements (transistors, etc.) and processes (e.g., CMOS) to make electronic circuits, a dedicated ‘micro-fluidic platform’ comprises a reduced set of validated components. These components have to be able to perform the basic fluidic operations required within a given application area. Such basic fluidic operations are, for example, fluid transport, fluid metering, fluid mixing, valving, and separation or concentration of molecules or particles. The collection of fluidic operations needed for diagnostic applications may have only little overlap with the collection needed for pharmaceutical applications or for applications in micro-reaction technology. In some cases detection methods will also belong to the basic set of micro-fluidics operations, and in other cases not (see Table 1). Nevertheless in all cases the user of a platform has to be able to readily combine the basic components within a given platform in order to implement the desired application. Often an efficient development is intimately linked to the availability of (standard) test setups and simulation tools. More important than providing a totally complete set of fluidic operations or components within a platform is the fact that all components have to be amenable to a well established fabrication technology. Furthermore all components or modules of a platform have to be connectible, ideally in a monolithic way or at least by a well defined, ready-to-use interconnection and packaging process. If a platform allows a seamless and simple integration of different fluidic components in a monolithic way, e.g., without sophisticated additional packaging techniques, this provides a significant advantage compared to other platforms. Thus speaking about micro-fluidic platforms involves also at least one validated fabrication technology to realize complete systems out of the components. Table 5-1. Common features of micro-fluidic platforms Micro-fluidics Operations validated components for basic micro-fluidics operations such as • fluid transport • fluid metering • fluid valving • fluid mixing • separation • concentration • detection …

Fabrication Technology validated manufacturing technology for the whole set of fluidic components (prototyping and mass fabrication)

seamless integration of different components • ideally in a monolithic way • or by a well defined and easy packaging technique


142 2

3.

EXAMPLES OF MICRO-FLUIDIC PLATFORMS

It is outside the scope of this article to give a complete survey of micro-fluidic platforms, which can be deduced from the literature. It is, furthermore, not intended to assess the different platforms by their value to the industry or the research community. The following examples are just selected to highlight the power of the platform concept in general, as well as to indicate their diversity regarding:

• •

the fabrication technologies and underlying materials; as well as the different micro-fluidics principles used to accomplish the basic micro-fluidics operations within different platforms.

On the one hand, the diversity and co-existence of various micro-fluidic platforms and related fabrication technologies could be regarded as a drawback because R&D efforts go in many directions and cannot be focused exclusively on one technology (as it is, for example, the case for microelectronic devices, where CMOS is the predominant technology). On the other hand, the diversity of approaches and technologies can be considered as an advantage for their successful adoption in different application fields. Owing to the diversity, specific advantages of certain platforms are likely to succeed in different application areas. It is, however, essential that a specific platform provides all required components along with a dedicated fabrication technology, as outlined before.

3.1

PDMS based Micro-fluidics for Large Scale Integration (‘Fluidigm platform’)

PDMS (Polydimethylsiloxane) is an inexpensive and still powerful material offering several advantages compared to silicon or glass. It is rubber-like and optically transparent down to wavelengths of 230 nm. Moreover PDMS can be easily micro-structured at low cost by various rapid manufacturing methods within days instead of weeks, as it is still the case for micro-structures in silicon. We briefly summarize the most convenient fabrication: After designing the micro-structures the CAD patterns are printed on blank transparencies. Lateral resolutions of 25 µm can be routinely achieved with image setters operating at 5,080 dpi. The resolution can be further enhanced to 8 µm using photo–plotters operating at 20,000 dpi. The transparencies are then used as photo–masks in UV lithography, and structures are transferred into SU-8 resist which serves as a master for fabricating PDMS moulds. A liquid


143

prepolymer is poured onto that master. After curing for 1h at 70°C the PDMS layer can be peeled off the master. Several layers of PDMS can be hermetically bonded on top of each other after plasma activation processes [4] or by using a technique called ‘multi-layer soft lithography’ [5]. The use of PDMS as a material for micro-fluidic applications originates mostly form the pioneering work of Georges Whitesides of Harvard University [4,6,7] and Stephen Quake from Caltech [5,8]. A detailed up to date view of the use of PDMS can be found in [9]. The micro-fluidic platform described in the following has been pushed forward by the company Fluidigm Corporation [10].

Figure 5-1. Setup of a micro-valve within the ‘Fluidigm platform’ [10]. Air pressure in the upper channel controls the liquid flow in the lower channel via displacement of a thin membrane at the crossing of the two fluid channels.

Figure 5-2. Micro-fluidic modules within the ‘Fluidigm platform’; basic layout for the different micro-fluidic operations: valving; pumping; mixing.

144 4


The Fluidigm platform mainly relies on the elastomer properties of PDMS. It is made of a planar glass substrate and two layers of PDMS on top of each other. The lower elastomer layer contains the fluidic ducts and the upper elastomer layer features pneumatic control channels. The basic functional module of the Fluidigm platform is a micro-valve made simply from a pneumatic control channel crossing a fluidic duct (Fig. 5-1). A pressure applied to the control channel squeezes the elastomer into the lower layer, where it blocks the liquid flow. By combining several micro-valves and actuating them in a peristaltic sequence more basic fluidic operations such as • fluid transport (pumping), • fluid mixing (pumping in cycle) can be realized (Fig. 5-2). Fluid metering can be implemented by defining a liquid compartment between two neighbouring control valves. Metering of volumes down to 250 pl has been demonstrated. The unique advantage of Fluidigm’s micro-fluidic platform is its amenability to large scale micro-fluidic integration similarly to integrated circuits in the electronics industry. The key principle is the multiplexing technique of micro-fluidic operations which allows for the control of N fluid channels by only 2 log2 N control channels (Fig. 5-3). This, for instance, means that 256 fluid channels can be controlled individually by 33 pneumatic control channels. Thorsen et al., demonstrated a comparator chip containing 256 different reaction sites on a single chip displaying a surface area of 25 x 25 mm2 [11] (Fig. 5-4). All 256 chambers on that chip can be individually addressed, pairwise mixed, and individually purged with 18 connections to the outside.


145

Figure 5-3. Illustration of multiplexing from [11]: 8 vertical fluid channels can be individually controlled by 6 horizontal control channels. The pressurized control channels block the fluid channels only at positions where the width of the control channel is ‘wide’ enough.

146 6


To summarise, the PDMS based ‘Fluidigm platform’ has certainly the potential to become one of the foremost micro-fluidic platforms for highly integrated micro-fluidic applications. It is a flexible and configurable technology which stands out owing to its suitability for large scale integration. The PDMS fabrication technology is comparably cheap and robust and it can be used to fabricate disposables. Reconfigured layouts can be assembled from a small set of validated modules, design iteration periods for new chips are of the order of days. A set of functions is programmable by the control software. The approach of building up an entire processor from a very limited set of elementary units resembles the microelectronics industry in which functional steps are configured by a circuit of transistors and capacitors. Limitations of the Fluidigm platform are related to the material properties of PDMS: for example, chemicals which are not inert to PDMS cannot be processed, or elevated temperatures such as in micro-reaction technology are not feasible.

3.2

Micro-fluidics on a Rotating Disk (‘Lab on a Disk’)

The modular ‘Lab on a Disk’ platform is inspired from the conventional compact disk (CD) known from audio applications or data storage [12,13,14,15,16]. It is constituted of a disposable disk integrating various passive micro-fluidic structures like channels, metering and mixing chambers, as well as a drive unit similar to an ordinary CD player controlling its spinning frequency. The platform has proved to allow a seamless and cost effective integration of the various micro-fluidic operations needed in lab-on-a-chip applications [15]. Liquid samples are applied to the micro-fluidic channels near the centre of the disk. During rotation the liquid is transported by the centrifugal force towards the outer periphery of the disk. The pulse free centrifugal flow can be stopped by hydrophobically patterned constrictions without moving actuation elements [12]. At these so called capillary burst valves the capillary depression at the barrier counteracts the centrifugal force acting on the liquid plug. The burst frequency for a particular valve structure, at which the centrifugal pressure exceeds the capillary pressure, is defined by channel geometry, the radial plug length and the contact angle which can be controlled at the disk’s fabrication. Metering of liquids can easily be done by chambers containing a well defined volume (Fig. 5-5). The outlet of such a chamber is blocked by a capillary burst valve and the liquid exceeding the volume of the chamber is purged via drain channels into a waste chamber. Mixing and switching of different reagents have been demonstrated in various ways [17,18]. Last but not least, the Lab on a Disk platform can be


147

regarded as a centrifuge which can be used in an elegant way for separating particles from solvents, e.g., red blood cells from blood plasma (Fig. 5-6) [19].

Figure 5-5. Metering structure on a Lab on a Disk platform: the structure is made of a chamber having a well defined volume, a hydrophobic constriction blocking the outlet of the metering chamber, and an overflow channel to the waste chamber.

On a single disk numerous samples can be processed in parallel becausethe format offers a rather big area compared to microchips which makes it very efficient. So far there are several companies which have already developed or are in the process of developing, Lab on a Disk systems. The most visible companies are Gyros AB in Uppsala, Sweden, performing an on-disk preparation of protein samples for matrix-assisted laser desorption and ionization mass spectrometry (MALDI-MS). Their ‘Gyrolab’ disk processes 96 protein samples simultaneously. The proteins are concentrated and purified on the disk. The disk is then transferred into an ordinary mass spectrometer for detection. A second type of disk is dedicated to quantifying nano-liters of protein sample with a fluorescence reader which is integrated into the CD player-like ‘Gyrolab Workstation’. The LabCD of Tecan is used for screening in drug discovery, testing the functionality of drug candidates. To this end blood serum is mixed with fluorescent probe molecules which bind to target proteins. Upon adding drug molecules they compete with the protein-bound fluorescent probe molecules, which can be detected by a change in the fluorescence. This technology allows for investigation of the interactions between different drugs as well as serum protein bindings.


148

Figure 5-6. Separation of red blood cells from whole human blood within the ‘Bio-Disk’ [16], [19]: (a) photo of the Bio-Disk; (b) capillary filling of 5 µl blood from a finger to the Bio-Disk; (c) details from the separation structure.

Within the Bio-Disk project coordinated by IMTEK in Freiburg, Germany, immuno assays for medical diagnostics on whole blood are being developed. Two exemplary applications are particularly considered:

• •

The determination of the immuno status for several important pathogens such as Tetanus, Diphtheria, Measles, and Hepatitis from a small droplet of blood; The time-critical recording of important blood parameters for emergency medical where often only a few minutes are decisive for the patient’s convalescence.

Functionalised beads which are polymer spheres of about 50 µm in diameter, are used as sensors. Compared to flat surfaces, an ensemble of beads offers extraordinarily large surfaces and short diffusion lengths. Thus drastically reduced times for affinity reactions with target molecules can be achieved. The affinity is detected by a final fluorescence readout step.


149

A basic advantage of the Bio-Disk platform is the control and automation of all fluidic processes by an easily accessible experimental parameter, the rotational frequency of the disk. Furthermore, this platforms offers a great potential for cost effective miniaturisation of portable analysis tools.

3.3

Droplet based micro-fluidics (DBM)

Droplet based micro-fluidics (DBM) systems are characterized by the fact of assays or reactions taking place within a single droplet. This is in contrast to the approaches discussed before, where the fluidic operations are performed in a ‘flow through’ manner in capillary channels. All DBM platforms make use of the high interfacial forces of small liquid droplets which hold the reagents within a well confined round shape (droplet). The droplets are kept in an immiscible surrounding fluid (air or oil is typically used). This way a lateral dispersion can be avoided while moving the droplets to different locations. The elimination of dispersion is a specific characteristic of this platform and can turn out to be advantageous for certain applications. DBM devices provide in some sense more flexibility, because assay protocols are not preconditioned by given micro-fluidic structures but can be defined by the sequence in which droplets are transported, mixed and separated. There are at least three different droplet based micro-fluidics (DBM) platforms which are intensively discused in the literature. Each of them provides different mechanisms to transporting and manipulating droplets, as will be detailed in the following.

3.3.1

DBM based on electro-wetting

The method for moving droplets enclosed between two parallel electrodes by means of electro-wetting was published by Pollack et al. [22] and Kim et al. [23] in 2000 and is currently commercialised, for example, by the company Nanolytics from Raleigh, NC., USA. The basic principle of this platform relies on electro-wetting which is the possibility of changing the contact angle of the liquid/solid interface of a droplet on a hydrophobic dielectric substrate by applying an electrical potential. This principle is called electro-wetting on dielectric substrates (EWOD), and if it is applied to one side of a droplet or liquid column the liquid volume will move according to the established difference in the contact angle. Since the first experiments various groups have started to exploit this effect to realize micro-fluidic components and systems. For example, Kim et al. developed a freely programmable platform [26] for handling liquid droplets in the µl-volume range based on electro-wetting (Figs. 7-9). In their ‘digital micro-fluidic


150 0

platform’, droplets are moved within a planar two-dimensional array of individually addressable electrodes [21,24]. Because the path of the droplets is determined by the pattern of electric potentials, the platform is easily programmable and reconfigurable. No channels have to be carved into a surface because the motion takes place between two flat surfaces only. EWOD systems typically consist of two glass plates (Fig. 5-7). The top plate contains a reference electrode typically realized as a transparent conductive layer. The bottom plate supports linear arrays of electrodes at right angles to each other which are separated by a dielectric layer. The electrode system is coated with a hydrophobic layer of Teflon which causes droplets to bunch into discrete units which can be shunted across the chip by a sequence of voltage signals.

Figure 5-7. a) Sketch of working principle of EWOD devices from [21]; b) top view of EOWD device with droplet on inter–digitated electrodes according to [22].

The devices proposed by Kim et al., can be fabricated easily by standard lithographic techniques on glass substrates. The technology is quite simple because only electrodes as active elements and a dielectric passivation layer are required. All fluidic operations are established by suitable switching of the electrodes exclusively. Typical operations such as have been demonstrated (cf. Fig. 5-8) and tested in biological assays, as reported in [25,26]. • • • •

droplet transport, droplet dispensing, droplet mixing, droplet splitting,


151

Figure 5-8. Basic operations of the EWOD platform according to [21].

Digital micro-fluidics based on EWOD can be considered a promising tool for droplet based chemical and biochemical analytics. However, droplet based chemistry is not an established technology, and it remains to be seen whether the industry will adopt this new paradigm of liquid handling. Currently EWOD is typically performed with aqueous solutions only.

3.3.2

DBM based on surface acoustic waves

Another method of realizing a ‘digital micro-fluidic platform’ was put forward by Wixforth et al. [27], [28], [29], [30]. Actually the company Advalytix from Brunnthal, Germany, is commercialising these type of devices. Therefore the platform is also referred to as ‘Advalytix platform’. The ‘Advalytix approach’ is based on surface acoustic waves (SAW) which were originally applied in Material Science and Solid State Physics [31]. Surface acoustic waves are mechanical waves with amplitudes of typically only a few nano-meters. In contrast to ordinary sound waves, which are of longitudinal type, SAWs are transversal waves. Their amplitude can either be parallel to the surface (Love waves), or perpendicular to the surface (Rayleigh waves), or of mixed type. For manipulating droplets SAW based devices typically apply Rayleigh waves.

152 2


In the ‘Advalytix platform’ the SAWs are generated by a piezoelectric transducer chip (e.g., quartz) which can be controlled by external electronics. Liquid droplets are placed onto the hydrophobic surface of the transducer chip and can be moved by the SAWs arbitrarily on the surface. To create the SAWs appropriate electrical AC signals are applied via electrodes to certain regions of the piezoelectric layer only. Thus the SAWs are exclusively generated at the desired location and can be focused onto the droplet. If the acoustic pressure exerted on the liquid is high enough the droplet can be deformed or even be moved (Fig. 5-9a). The ‘Advalytix platform’ is sometimes also referred to as ‘flat fluidics’, because no cover or slit like that in the EOWD approach is required, but a flat surface only where the SAWs are generated.

Figure 5-9. Single droplet a) deformed (and subsequently moved) by a SAW; b) subjected to a low energy SAW to induce internal streaming (mixing) according to [30].

Similarly to the EWOD platform also, the SAW platform provides most of the fundamental fluidic operations: Transport, merging, separating, and mixing have been successfully demonstrated [30]. Especially mixing seems to be very efficient by a rotating convection induced within a single droplet (Fig. 5-9b). In contrast to the EWOD approach the SAW platform allows for the use of ‘arbitrary’ liquids. This is because mechanical actuation is applied for manipulation instead of electrical actuation which requires specific electrical properties of the liquids. The Advalytix platform is also comparable to the EWOD platform in terms of technology. Apart from the control electronics only electrode structures on a Quartz substrate are required for building the actuator. These can be fabricated by standard lithographic techniques. Furthermore, for both of the DBM devices a hydrophobic coating is necessary on the surface to make droplets form. Thus both of the approaches discussed provide the opportunity to design and fabricate devices which are able to transport individual droplets arbitrarily on a flat surface cost effectively.


153

Figure 5-10. Processing of droplets on the Advalytix platform a) to d) represent a series of snapshots visualizing the movement and merging of the droplets. In the centre of the device sensors and heaters are integrated [30].

3.3.3

DBM based on two phase liquid flow

In contrast to the DBM platforms mentioned before which enable the transport of droplets in two dimensions on a plane surface, the method put forwards by Ismagilov et al. [32,33,34,35,36,37] is only one–dimensional. Droplets of the liquids under study are injected successively into a microchannel forming a column of sample plugs. The individual liquid droplets are thereby immersed in a carrier liquid which is immiscible with the sample liquids (e.g., oil and aqueous solutions). All liquids together are transported by a pressure driven flow (PDF) along the micro-channel. Basically the situation described is always given when liquid emulsions are streamed through fluidic tubings. Ismagilov and his co-workers, however, found means of producing precisely defined pl sized volumes and compact samples separated by defined amounts of carrier liquid. The sample plugs are arranged chain-like inside the channel and stay separated even for considerable distance of transport. Thus each plug can be considered as a closed container containing a specific chemical composition. The carrier

154 4


fluid typically prevents cross-contamination and evaporation which can be considered as a generic advantage of this platform. Another important feature of the platform is that mixing can be enhanced by convection. As reported in [36], recirculation within the droplet can assist the mixing effectively (Fig. 5-12). However, the efficiency of mixing depends on various parameters.

Figure 5-11. Alternating formation of droplets in oil a) sketch of principal set up; b) micrograph of corresponding experiment from [37]; c) photograph of the device studied in [33].

Amazingly, this approach does not require sophisticated means of creating the droplets. If the fluidic channels are appropriately chosen, droplets with defined volumes form spontaneously. Typically, different samples and carrier liquid are supplied by external syringe pumps through different supply channels and merged at a junction. For a given set up the pressure ratio (or flow rate) between the various sample liquids and the carrier liquid laminated together at a junction leads to a regular pattern of sample liquids being separated by plugs of oil as depicted in ( Fig. 5-11a) and 12. With this simple set up interesting applications such as protein crystallization and measurements of millisecond reaction kinetics have been demonstrated [33,37]. The devices studied by Ismagilov and co-workers are mainly fabricated by PDMS technology similar to the Fluidigm platform. In contrast to the Fludigm platform the actuators are in this case however, external which means connected by macroscopic lines (Fig. 5-11c). Possibly pumping and valving elements of the Fludigm platform could be integrated with the DBM approach of Ismagilov et. al in the future. Thus extremely miniaturized systems could be realized, including the pumps required for liquid transport.


155

Figure 5-12. a) Principle sketch of mixing by recirculation; b) Micrographs of mixing at different flow rates according to [36].

3.4

Non-contact liquid dispensing

A very special case of a DBM platform is given in the case in which liquids are delivered as free flying droplets. Thus assays can be performed on flat substrates as well as in conventional containers such as micro-titer plates (MTP) or on any other target. This approach is closest to the traditional lab routine which is based on assembling assays via successive pipetting steps, manually or by automated lab equipment. It provides greatest flexibility with respect to media and assays protocols, and therefore justifies a separate treatment. We focus the discussion here on a platform for highly parallel noncontact liquid dispensing based on components put forward by the authors in the past [38,39,40,41,42,43,44,45]. The platform presented seamlessly covers the volume range from several tens of picoliters to several microliters. A key feature of the platform is that a multitude of individual dispensing units can be easily arranged on a substrate in parallel. It is therefore possible to handle hundreds and even thousands of different liquids simultaneously (Fig. 5-13) at a pitch ranging from several hundreds of microns to several millimetres. The liquids are delivered as free flying droplets or jets in a non-contact manner by three basic dispensing techniques. These techniques described in the following are referred to as DWP, TopSpot®, and TopSpot Vario. For all of the techniques a single dispensing unit simply consists of a reservoir for holding the liquid, a nozzle (nozzle chamber and orifice) from which the liquid is dispensed, and a connection channel connecting reservoir and nozzle. These dispensing units can be arranged arbitrarily on a so called dosage chip which is driven by an external actuation device. In Fig. 5-14 the working principles of all three techniques are illustrated for a single dispensing unit.


156 6

number off different i parallel ll l reagents in

10 000 10.000 1 000 1.000 100 10 1

DWP

pipetting workstations

manual pipettes

100 00 10 0 1 microliter croliter

TopSpot

TopSpot Vario

microarrays NanoJet / PipeJet

InkJet

aerosol

lower ower volumes olumes

100 10 1 100 10 1 100 10 picoliter femtoliter nanoliter

Figure 5-13. Map of low volume dispensing. The platform relies on three different techniques for creating well defined liquid volumes which can be dispensed in a contact-free manner (DWP, TopSpot, TopSpot Vario). The technologies seamlessly cover the volume range from several tens of picoliters to several microliters. Several hundreds or thousands of dispensing units can be easily arranged in parallel to handle different liquids simultaneously.

Figure 5-14. (a) Pressure based actuation principle for dispensing from 10 nl to several µl based on the DWP principle; (b) pressure based actuation for dispensing volumes in the lower nl volume range based on the TopSpot principle; (c) direct displacement based actuation via an elastomer for dispensing volumes in the 50 pl to 1.000 pl range based on the TopSpot Vario principle.


157

Dispensing Well Plate (DWP) Technology [38, 39, 40] Within the so called Dispensing Well Plate (DWP) technology the dosage volume is defined by the volume of a micro-machined nozzle chamber. When liquid is filled into the reservoir it is transported by capillary forces to the micro-nozzle, which comprises a defined geometrical volume. The total liquid contained in the nozzle can then be dispensed by applying a pneumatic pressure. Amplitude and duration of that pressure pulse are of the order of 0.3 to 0.8 bar and 3 to10 ms respectively. However, the dispensed volume is defined essentially by the geometrical volume of the nozzle, and it is to a great extent independent of the duration and amplitude of the pressure pulse. Furthermore, the dosage volume is hardly affected by the liquid properties such as viscosity, density, and surface tension, which makes the method very robust.

TopSpot Dispensing Technology [41,42, 43] The TopSpot® principle also relies on a pneumatic pressure pulse which in this case is on the sub-ms time scale. The process of ejecting the liquid is similar to the ‘drop on demand’ (DoD) principle used in inkjet printing [46]. In the case of TopSpot the pressure pulse is created by a piezo-stack actuator driving a piston. The piston movement compresses the air in a closed cavity in the back of the nozzles and generates the required pneumatic pressure pulse. The pressure pulse acts equally upon all the nozzles, causing (with sufficient dynamics) each nozzle to eject a single droplet. The volume of the ejected droplets is typically of the order of 1 nl. In contrast to the DWP technology only a small fraction of the volume contained in a nozzle is ejected. The exact amount of the dosage volume is determined by a complex interplay between liquid properties and actuation parameters which is well known for DoD devices. Different volumes can be achieved by using different nozzle diameters and different liquids.

TopSpot Vario Dispensing Technology [44, 45] In contrast to the pneumatic technologies described so far, the TopSpot® Vario principle applies the direct displacement of an incompressible, but easily deformable, elastomer to drive the liquid. The movement of a piston causes the elastomer to be displaced into the nozzles. Hereby a kind of micro-syringe is formed over the nozzle. A well defined volume of liquid in the nozzle chamber is displaced and the corresponding liquid volume is ejected from the orifice. The direct displacement allows for adjusting the

158 8


ejection volume and speed by the stroke and speed of the piezo actuator. The achievable volume range is between 100pl and 1400pl (1.4nl). It can easily be adapted within one given device just by varying the control voltage of the piezo actuator. All three dispensing techniques have in common that a multitude of dispensing channels can be arranged in parallel easily. This will be highlighted by the following examples adopted for different applications. Another important point is that the pitch of the reservoirs can be designed independently from the pitch of the nozzles. This allows for choosing the pitch of the reservoirs to enable an easy filling by pipetting robots whilst the pitch of the nozzles can be set according to the application (e.g., a few hundreds of µm for the fabrication of microarrays). Furthermore, the basic structures (reservoirs, capillary channel, nozzle chamber, and nozzle orifice) of the dispensers can be fabricated by different technologies in different materials. This has been demonstrated for DRIE etching of silicon, lithographic fabrication in SU-8 or PDMS [38,40,44] as well as by injection moulding of plastics [47]. In the following examples we describe two typical applications realized so far.

3.4.1

‘Dispensing Well Plate’ for ‘High Throughput Screening’

The liquid handling in High Throughput Screening (HTS)—which is an integral part of the drug discovery process—demands that liquids are transferred between the wells of standard well plates. The pitch of the reservoirs is typically 9.0 mm, 4.5 mm, or 2.25 mm according to the 96,384, or 1,536 SBS well plate standard. Fig 15 shows a dosage chip with 384 dispensing units which was especially designed for HTS applications. The chip size is 80×120 mm and 16×24 dispensing channels are arranged in parallel at a pitch of 4.5 mm corresponding to the SBS 384 well plate standard (hence the name ‘Dispensing Well Plate’). The dosage volume of every dispensing unit was designed to be 50 nl. However, if required the nozzle volume can even be raised to several microliters. By applying a pressure pulse to the whole upper surface of the Dispensing Well Plate the liquid volumes contained in the individual nozzle chambers are driven out completely. After switching off the driving pressure the nozzles refill again from the reservoirs by capillary forces.


159

Figure 5-15. a) Dispensing Well Plate in 384 format micro-machined in silicon; b) array of 50 nl droplets simultaneously dispensed by the 384 channel dispensing well plate.

Figure 5-16. a) Dispensing Well Plate in 384 format micro-machined in PMMA; b) reproducibility of the dosage volume for different f liquids and actuation parameters.

Fluorescence measurements with a plate reader have shown that the homogeneity over all 384 dispensing units of a DWP micro-machined in silicon is within a coefficient of variation (CV) of 5%. Gravimetric measurements revealed that the volume accuracy can be kept constant and independent of the liquid level in the reservoir until the reservoir is nearly completely depleted (a rest volume of about 700 nl). Because of the geometric definition of the dosage volume it has been proved to be independent of liquid properties and actuation parameters within a certain range.

160 0

3.4.2

MICRO-FLUIDIC PLATFORMS M ‘TopSpot print heads’ for ‘High Throughput Fabrication of Microarrays’

Microarrays are highly parallel biosensors. Their sensor effect is based on a chemical reaction between molecules with a lock and key principle. When producing a microarray the different f catching molecules have to be put on at a defined pitch (typically 500 µm) on the substrate. The reaction between the catching molecules and a complex mixture of molecules is the actual measurement. Molecules caught can be detected after washing off any surplus material. Microarrays are usually analysed using fluorescence techniques. The molecules in the complex mixture are equipped with fluorescence molecules for later detection. In the detection equipment (a scanner) a picture of the fluorescence signal intensity is taken which is translated in a characteristic fluorescence image. The position of the catching molecule in its array provides information about the identity of the caught molecule.

Figure 5-17. Illustration of the working principle of a microarray: (a) probe molecules are immobilized at a regular pitch on a flat substrate; (b) sample molecules hybridise to probe molecules if chemical structures are complementary; (c) fluorescent readout of the microarray.

The key to enhance the speed and throughput in microarray fabrication is an integrated format change in the dispensing system. In a TopSpot device the reservoirs are arranged at the pitch of 2.25 mm amenable to filling by standard lab automation. They are connected to nozzles typically at a pitch of 500 µm. The liquid is transported simply by capillary forces to the central nozzle area. Liquid volumes of several µl can be loaded into the print head, enough for dispensing several thousand times without the need for refilling. Fig. 5-18 shows a 96 channel TopSpot print head and Fig. 5-19 shows a detail from a 384 channel TopSpot print head illustrating the change of format between the pitch of reservoirs and nozzles.


161

So far the TopSpot technique has been validated for various kinds of liquids in many experiments. The coefficient of variation (CV) of generated spot diameters is typically measured as being smaller than 1% within one single dispensing nozzle and smaller than 1.5% within all nozzles of a print head for all printing buffers used [43]. No carry over and no crosstalk were found in extensive experiments with oligonucleotides and optimized printing buffer compositions.

Figure 5-18. TopSpot print head with 96 dispensing channels in parallel.

Figure 5-19. Details of a 384 channel TopSpot print head showing the different pitch of the outer reservoirs (2.25 mm) and the nozzles (1,000 mm).

4.

CONCLUSION

It is most likely that micro-fluidic platforms will have a huge impact on the development of application specific, integrated micro-fluidic systems in the near future. The power of the micro-fluidic platform concept consists in the combination of a set of validated fluidic components combined with a


162 2

proved low cost fabrication technology. Full systems can be built easily, based on a given platform, without the need for ever new developments of either components or fabrication technologies as is unfortunately still often the case today. The focusing on micro-fluidic platforms will enable the micro-fluidics community to leave the device and technology oriented research of today behind in order to enter into the next challenge: the flexible and cost efficient development of thousands of different applications accessible by using the full potential of micro-fluidics and by following a system oriented approach. The collection of the examples given in this article shows that the platform idea has already been taken up by many research groups. These groups do not only try to work on individual components and specific technology developments, but they are focusing on the combination of a few validated components by simple proved technologies to implement more and more application specific devices. This paradigm shift from a component and technology based research to a system oriented approach is only possible if there are proved platforms which can be relied on. Owing to the benefits of the platform approach we highly recommend researchers engaged in micro-fluidics to identify available micro-fluidic platforms and design new micro-fluidic components around existing platforms. This will make new components and design rules easily accessible to others. Too many components reported in the literature have not been adopted in micro-fluidic systems because they were not compatible with other components required to built the system. However, if a component fits into a given platform it can be used as an important cornerstone to highly integrated micro-fluidic systems. The research community as a whole will profit most from this approach.

5. [1] [2] [3] [4] [5]

REFERENCES Ducrée, J., Zengerle, R., (eds.), (2004), ‘FlowMap—Micro-fluidics Roadmap for the Life Sciences’. Laser, D.J., Santiago, J.G., (2004), ‘A review of micropumps’, J. Micromech. Sys., 14(6), R35–R64. Ducrée, J., Zengerle, R., (2005), ‘Micro-fluidics’, Springer-Press, (to be published). Duffy, D.C., McDonald J.C., Schueller, O.J.A., Whitesides, G.M., (1998), Anal. Chem., 70, 4974–4984. Unger, M.A., Chou, H.-P., Thorsen, T., Scherer, Quake, A.S.R., (2000), ‘Monolithic Micro-fabricated Valves and Pumps by Multilayer Soft Lithography’, Science, 288, 113–116.

MICRO-FLUIDIC PLATFORMS [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]

[16] [17]

[18] [19] [20] [21]

[22] [23]

163

McDonald, J.C., Duffy, D.C., Anderson, J.R., Chiu, D.T., Wu, H., Schueller, O.J.A., Whitesides, G.M., (2000), Electrophoresis, 21, 27–40. Ng, J.M.K., Gitlin, G.I., Stroock, A. D., Whitesides, G.M., (2002), Electrophoresis, 23, 3461–3473. Quake, S.R., Scherer, A., (2000), ‘From Micro-to Nanofabrication With Soft Materials’, Science, 290, 1536–1540. Sia, S.K., Whitesides, G.M., (2003), ‘Micro-fluidic devices fabricated in poly(dimethylsiloxane) for biological studies’, Electrophoresis, 24, 3563–3576. Fluidigm Corporation, San Francisco, CA, www.fluidigm.com Thorsen, T., Maerkl, S.J., Quake, S.R., (2002), ‘Micro-fluidics Large Scale Integration’, Science, 298, 580–584. Madou, M., Kellogg, G., (1998), ‘The LabCD., A centrifuge based micro-fluidic platform p for Diagnostics’, g , SPIE Proceedings. g Ekstrand, G., Holmquist, C., Edman Örlefors, A., Hellman, B., Larsson, A., Andersson, P., (2000), ‘Micro-fluidics in a Rotating CD’, µTAS 2000, 311–314. Madou, M., et al., (2001), ‘A novel design on a CD disc for 2-point calibration measurement’, Sensors and Actuators, A 91. Madou, M., Lee, L.J., Daunert, S., Lai, S., Shih, C.-H., (2001), ‘Design and Fabrication of CD-like Micro-fluidic platforms for Diagnostics: Micro-fluidics Functions’, Biomedical Microdevices, 3:3, 245–254. www.bio-disk.com Ducrée, J., Brenner, T., Glatzel, T., Zengerle, R., (2003), ‘A Coriolis based Split-and-Recombine Laminator for Ultrafast Mixing on Rotating Disks’, Proc. µTAS 2003, Squaw Valley, California, USA, 603–606. Brenner, T., Glatzel, T., Zengerle, R., Ducrée, J., (2003), ‘A Flow Switch Based on Coriolis Force’, Proc. µTAS 2003, Squaw Valley, California, USA, 903–906. Brenner, T., Haeberle, S., Zengerle, R., Ducrée, J., (2004), ‘Continuous Centrifugal Separation of Whole Blood on a Disk’, MicroTAS 2004, 566–568. Lee, J., Kim C.-J., (2000), ‘Surface-tension-driven microactuation based on continuous electro-wetting’, Journal of Microelectromechanical Systems, 9(2), 171–180. Cho, S.K., Moon, H., Kim, C.-J., (2003) ‘Creating, Transporting, Cutting, and Merging Liquid Droplets by Electro-wetting based Actuation for Digital Micro-fluidics Circuits’, Journal of Microelectromechanical Systems, 12(1), 70–80. Pollack, M.G., Fair, R.B., and Shenderov, A.D., (2000), ‘Electrowetting based actuation of liquid droplets for micro-fluidics applications’, Applied Physics Letters, 77(11), 1725–1726. Lee, J., and Kim, C.J., (2000), ‘Surface-tension-driven microactuation based on continuous electro-wetting’, Journal of Microelectromechanical Systems, 9(2), 171–180.

1644


[24]

Paik, P., Pamula, V.K., and Fair, R.B., (2003), ‘Rapid droplet mixers for digital micro-fluidic systems’, Lab-on-a-chip, 3, 253–2590. Srinivasan, V., Pamula, V.K., Pollack, M.G., Fair, R.B., (2003), ‘Clinical diagnostics on human whole blood, plasma, serum, urine, saliva, sweat, and tears on a digital micro-fluidic platform’, Proceedings of µTAS 2003, MESA Monographs, 1287–1290. Fan, S.-K., Hashi, C., Kim, C.-J., (2003), ‘Manipulation of Multiple Droplets in NxM Grid by Cross-reference EOWD driving scheme and pressure-contact packaging’, Proc. IEEE-MEMS 2003, Kyoto, 694 –697. Wixforth, A., (2003), ‘Acoustically driven planar micro-fluidics’, Superlattices and Micro-structures, 33(5–6), 389–396. Wixforth, A., Gauer, C., Scriba, J., Wassermeier, M., Kirchner, R., (2003), ‘Flat fluidics: A new route toward programmable biochips’, Proceedings of SPIE 2003. Wixforth, A., Strobl, C., Gauer, C., Toegl, A., Scriba, J., von Guttenberg, Z., (2004), ‘Acoustic manipulation of small droplets’, Analytical and Bioanalytical Chemistry, 379(7–8), 982–991. Hess, P., (2002), acoustic waves in materials science’, Physics Today, 55(3), 42–47. Ismagilov, R.F., (2003), ‘Integrated micro-fluidic systems’, Angewandte Chemie-International Edition, 42(35), 4130–4132. Song, H. and Ismagilov, R.F., (2003), ‘Millisecond kinetics on a micro-fluidics chip using nanoliters of reagents, Journal of the American Chemical Society, 125(47), 14613–14619. Ismagilov, R.F., (2003), ‘Integrated micro-fluidic systems’, Angewandte Chemie–International Edition, 42(35), 4130–4132. Shestopalov, I.A., Tice, J.D., and Ismagilov, R.F., (2004), ‘MultiStep Chemical Reactions Performed On Millisecond Time Scale In a Micro-fluidics Droplet based System’, Lab-on-a-chip, 4, 316–321. Tice, J.D., Song, H., Lyon, A.D. and Ismagilov, R.F., (2003), ‘Formation of Droplets and Mixing in Multiphase Micro-fluidics at Low Values of the Reynolds and the Capillary Numbers’, Langmuir 19, 9127. Zeng, B., Roach, L.S., and Ismagilov, R.F., (2003), ‘Screening of protein crystallization conditions on a micro-fluidics chip using nanoliter-size droplets’, Journal of the American Chemical Society, 125(37), 11170–11171. Koltay, P., Steger, R., Bohl, B., Zengerle, R., (2004), ‘The Dispensing Well Plate: A Novel Nano-dispenser for the MultiParallel Delivery of Liquids (DWP Part I)’, Sensors & Actuators, A116, 483–491. Koltay, P., Kalix, J., Moosmann, C., Zengerle, R., (2004), ‘Theoretical Evaluation of the Dispensing Well Plate Method (DWP part II)’, Sensors & Actuators, A116, 472–482. Steger, R., Bohl, B., Zengerle, R., Koltay, P., (2004), ‘The dispensing well plate: a novel device for nanoliter liquid handling in ultra high throughput screening’, Journal of the Association for Laboratory Automation (JALA), 9(5), 291–299.

[25]

[26]

[27] [28] [29] [30] [31] [32] [33] [34] [35]

[36]

[37]

[38] [39]

MICRO-FLUIDIC PLATFORMS [40]

[41]

[42]

[43]

[44] [45] [46]

165

Gutmann, O., Kuehlewein, R., Reinbold, S., Niekrawietz, R., Steinert, C.P., de Heij, B., Zengerle, R., Daub, M., (2004), ‘A Highly Parallel Nanoliter Dispenser for Microarray Fabrication’, Biomedical Microdevices, 6:2, 131–137. Gutmann, O., Niekrawietz, R., Kuehlewein, R., Steinert, C.P., de Heij, B., Zengerle, R., Daub, M., (2004), ‘Impact of Medium Properties on Droplet Release in a Highly Parallel Nanoliter Dispenser’, Sensors and Actuators, A116/2, 187–194. Gutmann, O., Niekrawietz, R., Kuehlewein, R., Steinert, C.P., Reinbold, S., de Heij, B., Daub, M., Zengerle, R., (2004), ‘NonContact Production of Oligonucleatide Microarrays Using the Highly Parallel TopSpot Nanoliter Dispenser’, The Analyst, 129, 835–840. Steinert, C.P., Goutier, I., Gutmann, O., Sandmaier, H., Daub, M., de Heij, B., Zengerle, R., (2004), ‘A Highly Parallel Picoliter Dispenser with an Integrated, Novel Capillary Channel Structure’, Sensors and Actuators, A116/1, 171–177. de Heij, B., Steinert, C., Sandmaier, H., Zengerle, R., (2003), ‘A Tuneable and Highly Parallel Picolitre-Dispenser Based on Direct Liquid Displacement’, Sensors & Actuators, A 103, 88–92. Le, Hue P., (1998), ‘Progress and Trends in Inkjet Printing Technology’, Journal of Imaging Science and Technology, no. 42, 49–62. Steger, R., unpublished results.

Chapter 6 DNA BASED BIO–MICRO-ELECTRONIC MECHANICAL SYSTEMS Frank F. Bier, Dennie Andresen, Antje Walter Fraunhofer-Institute for Biomedical Engineering, Department of Molecular Bioanalysis and Bioelectronics, Potsdam-Nuthetal, Germany

Abstract:

DNA is not only the most interesting biomolecule because of its role as information storage but is also subject in various analytical and technical applications. It can be used as probes for DNA arrays analyzing genetic defects and SNP’s using lab-on-chip approaches as PCR on chip and electrophoresis. Furthermore thestability of DNA can be used to create bioinspired technical systems. Nucleicacids in general, may be viewed as a material with very special physical andphysico-chemical features, becoming a versatile tool for nano-construction. Examples areDNA directed immobilisation and nucleic acid tags and stretching andmetallization of DNA for electronic purposes. In such a way mechanical andelectronically building blocks for nanotechnology can be created.

Key words:

DNA chips, lab-on-chip, PCR on chip, DNA stretching, DNA as device

1.

INTRODUCTION

This chapter will cover two different aspects of DNA with respect to BioMEMS: First, DNA is the most important biomolecule because of its role in genetics as information storage, and therefore of highest interest in many analytical applications1. Therefore many recent inventions have been made to create analytical tools for the analysis of a huge variety of applications.

1

This is also true for RNA especially in its role as information carrier, the messenger RNA. However, using reverse transcription it is possible to get a DNA copy and thus all that is said for DNA analysis in this chapter will be valid also for most RNA analysis.


167

DNA BASED BIO–M –MEM MS

168 8

Second, DNA, or nucleic acids in general, may be viewed as a material with very special physical and physico-chemical features, becoming a versatile tool for nano-construction.

1.1

The unique features of nucleic acids

The most important feature of DNA is its ambivalence: On the one hand, DNA is chemically a homogeneous material with repeating units, the sugar phosphate backbone. On the other hand, the structure is not identically repeating, but owing to the four different bases, adenin, cytosin, guanin, and thymidin (resp., uracil), the sequence is unique depending on composition.

Figure 6-1. The building blocks of DNA, base-pairing, the double-helix, hybridisation.

1.2

Lab on the Chip

It is the dream of analytics for daily life—e.g., in point of care situations to have an analytical device, that takes up the sample and presents the desired data. All preparation and processing is done internally. The user does not see any of the complex preparation and measuring steps inside the device. This notion is condensed in what has been called ‘lab on the chip’ or ‘micro-total analysis system—µTAS’[1]. Both have the aim of integrating

DNA BASED BIO–MEMS

169

sample pre-treatment such as cleaning, separation, and amplification, as well as signal generation and transduction in a self-concealed manner. A second aim of integrating chemical and biochemical analysis in miniaturised devices is to achieve higher sensitivity and lower consumption of sample and reagent. Sample volume, all reagents needed for the analysis, and also dead volumes and loss of material that are typical for all fluid handling, are scaled down by miniaturisation. Up to now only a few of the steps mentioned are already completely integrated. Some successful examples have already been commercialised for laboratory use (e.g., Agilent Bioanalyzer 2100; HPLC Chip).

Figure 6-2. The Lab on the Chip concept.

1.2.1

Electrophoresis

The most advanced lab on a chip devices in connection with nucleic acids are examples of miniaturised electrophoresis. Capillary electrophoresis uses capillaries of 50 µm inner diameter. These can be condensed easily by winding channels without loss of separation capability. Also gel electrophoresis may be miniaturised with benefit by reduction of sample volume and processing time, and also better resolution with sharpened bands [2].

1.2.2

Polymerase Chain Reaction (PCR)

Another field of vivid R&D in the context of nucleic acids in lab on a chip devices is the miniaturisation and automation of the polymerase chain

170 0


reaction (PCR). Since amplification of the sample material is necessary in most DNA analysis, PCR is as fundamental to analysis related to nucleic acid as electrophoresis. The basic principle of PCR is described in Fig. 6-3. Since PCR has the power to amplify DNA starting from only a single molecule, miniaturising the device is quite feasible since it only has to be ensured that at least one molecule is within the volume applied to the analysis device (µPCR device). Even more, a miniaturised device may have the potential to more accurately present the desired temperature profile of the three PCR steps increasing by that way efficiency as well as stringency of the assay. Various aspects of the PCR integration into biochips are discussed in more detail below (Sec. 2.5.2 and 2.6).

Figure 6-3. The Polymerase Chain Reaction (PCR).

The Polymerase Chain Reaction is divided into three different steps during one amplification cycle: during the denaturation step at 95°C (1) the hydrogen bonds of the DNA template are broken down. The double-stranded


171

DNA divides into two single-strands. In the annealing step the specific primers hybridise to their corresponding sense and anti sense template strands (2). The annealing temperature depends on the TM of the primer (typically between 55 and 70°C). After the annealing step each primer is elongated in the 5’–3’direction by the DNA polymerase; typically at 72°C. At the end of the first PCR Cycle one DNA copy of the original DNA template is obtained (4). Each further cycle leads to exponential amplifycation of the available DNA templates. 30–35 cycles are usually sufficient to obtain enough DNA copies for further analysis.

1.3

Biochemical reaction chains for integration: biosensors and the ‘lab biochip’

While device integration does require a high level of engineering and microsystem technology, the biotechnological contribution may come with less effort, at least with regard to the mechanical properties. The biosensor approach makes use of the unique feature of biomolecules to be able to act even in turbid environment. The biomolecular recognition is highly specific and non-specific interference is either low or may be separated easily by the character of the binding.

measuring solution

signal electronics

receptor transducer

The biosensor concept.

Biosensors are usually divided into enzymatic (or metabolic) and affinity type. Whilst the first produces an observable change of mass of a specific substance, other physical properties have to be employed for the detection of binding events. DNA hybridisation is an example for binding, labels are often needed to make binding events visible.

172 2


More complex biochemical reactions have been introduced into biosensors by the use of coupled enzyme reactions or coupled binding and enzymatic reactions. Whilst this is a theme of it is own in biosensor research for metabolic biosensors, only few reports have been given on DNA modifying enzymes (see section 2.5). An example of both integration by micro-fluidics and complex biocomponents is the Biacore instrument that is based on surface plasmon resonance (SPR) as the transducing mechanism. However, this device is far from being a ‘BioMEMS’ because it still uses a macroscopic cartridge for fluidic manipulation and needs a heavy bench top machine to run the device.

2.

MICROARRAYS AND BIOCHIPS BASED ON DNA

Microarray technology allows massive parallel determination and multiple measurements of a variety of binding events to be carried out simultaneously. In addition it has the advantage of requiring a small amount of material and a modest investment of labour; moreover, it might save a great deal of time and may easily be automated. Microarrays in general consist of many microscopic spots each containing identical molecules, i.e., receptors, probes or targets. The numbers of spots may vary from less than 100 to several 100,000. The molecules are attached to a solid support which can be made from glass, silicon or a polymer. In the case of nucleic acids, the receptors are usually oligonucleotides or cDNA, and the binding event is simply the hybridisation of complementary strands. Biochip technology (especially DNA chips) is a rapidly developing field with a high potential for commercial success. Introductions to the principles of the technology as well as to various applications are available in many reviews [3,4] and in the well written practical approach by Schena [5]. Thus the combination of molecular biology, micro-fabrication, and bioinformatics has generated novel tools and has the strong potential to bring more products onto the market for genetic analysis purposes; applications of which can be used in all branches of the life sciences.


2.1

173

The typical microarray experiment

The basic assay principle of all biochips or microarrays is a binding event such as that in immuno assays. One binding partner, either the receptor or the ligand, the probe or the target, is immobilised within a small but well defined area on a flat solid support of glass or polymer; referred to as the ‘spot’ or ‘feature’ of the microarray. There are also examples of prestructured slides, structured, e.g., by micro-cavities (nano-titerplates) or by chemical structuring as well as by electronic features, such as micro electrodes. The features contain identical molecules, whilst the whole array contains a manifold of arbitrarily chosen or systematically varied compounds. Once the microarray is produced, the sample that contains the ligands or targets to be investigated is added and binding occurs at several features on the chip which results in a characteristic pattern representing the sample. The procedure of a typical biochip experiment is schematically represented in Fig. 6-5. Fluorescence labels are by far the most frequently used labels to detect the binding event. The sample itself is labelled in a step prior to incubation. After incubation the chip is read out by means of a scanning or imaging device, usually in a dry state, giving a snap shot picture of fluorescence intensities. Internal standards for comparison purposes have to be incorporated into this procedure.

2.2

Manufacturing of Microarrays

Array technology benefits from intensive industrial development at both ends of array handling, namely array production with specifically constructed spotting robots and microarray readout technology. In the area of chip production, two different approaches are used: (1) Synthesis on the chip; and (2) separate synthesis with subsequent deposition on the chip. The first method is applicable to generate chemical libraries, e.g., of short oligonucleotides or peptides; the second method may also be adapted to long polynucleotides or proteins, or any other manifold of receptors. Detailed description of these methods may be found in [6,7,8].


174

Figure 6-5. The biochip experiment. Step 1: preparation of sample including labelling of the sample (1a), and microarray production (1b); Step 2: the incubation step for initiating the specific binding event, hybridisation of the labelled sample to the immobilised oligonucleotide probes; Step 3: detection; Step 4: data analysis and bioinformatic processing.

2.2.1

Synthesis on the chip

Fodor et al., reported for the first time, as early as 1991, the notion of synthesising an array of oligomers on a chip surface by use of stepwise local photo-deprotection [9]. In his first paper peptide synthesis was in the focus; however, it was soon discovered that oligonucleotide arrays would be needed much more and, owing to only 4 bases instead of 20 amino acids, would be easier to implement. The process of production is similar to wafer production in microelectronics: A chemical activated silicon wafer surface is covered with photo-labile protecting groups. After UV irradiation through a mask with high spatial resolution the activated groups are reactive at certain


175

localities and a nucleotide which again bears the same photo-labile protecting group at its 3’-end, may couple with its 5’-end. Repetition of this procedure with all 4 bases will end up with defined sequences in each feature of the array, 4 steps (and masks) are needed for each nucleotide. Standard microarrays have features of 0.1 mm×0.1 mm, but the process is optimised down to a feature size of 10 µm×10 µm enabling several thousands up to hundred thousands features on one chip. With this type of arrays whole genome screening is possible.

2.2.2

Spotting techniques

An alternative way of building up a microarray is to deposit small amounts of pre-activated oligonucleotides by micro dispensing. An advantage compared to on-chip synthesis is the possibility of cleaning up the material that will be deposited. While on-chip synthesis is limited in sequence length by the efficiency of each coupling step during the synthesis, pre-synthesised material may be cleaned up even from a rather mediocre chemical environment. Moreover, the dispensing methods are not limited to any chemical species, but may be used for any kind of arrays including PCR products or complete genomes. Two different techniques are currently in use, (1) contact, and (2) noncontact printing. The first technique uses pin tools, or needles, that are dipped into the probe solution and dispense a certain amount of material by contacting the support material. Reproducibility of this method is limited, but non-contact techniques rely on the piezo effect. As in an inkjet printer a micro-fluidics nozzle is activated to release droplets of less than one nanolitre. The distance of the features can be minimised down to 200 µm. This type of dispensing allows for high accurate manufacturing of low and medium dense microarrays (up to a few thousand features).

2.3

Transcription Analysis

The application used by far most frequently of DNA microarrays is transcription analysis. Owing to the basic dogma of molecular biology the information flow runs from the storage medium, the genomic DNA, via transcription to the messenger RNA (mRNA) and by translation in the ribosome machinery to the protein. Many details of the regulation of gene activity have been discovered in recent years. It is of great scientific and also medical interest to learn more about this mechanism and the analysis of transcribed genes, the first step of gene activity is currently in the focus of all life sciences. The ‘transcriptum’, i.e., the manifold of all genes transcribed in a specific cell (cell type) at a given time under defined


176 6

conditions, can effectively be analysed using DNA microarrays. Usually a comparison is made between a cell type in state A and an altered state, state B., which might be linked for instance to a disease. Both samples, material extracted from cells of states A and B, are labelled with two different fluorochromes. The difference in gene activity in both cell types can easily be recognised wherever one colour exceeds the other. The comparison method compensates for many short comings of the array method, since each feature of the array is referenced in itself. The only information drawn is with regard to the binding of species A or B to the same probe on the surface. Different hybridisation conditions for different features on the array result in various absolute signal values (fluorescence intensities); however, the relative amount of the two compared samples are still valid. Therefore by far the most applications of microarray technology reported today are in the realm of transcription anaylsis.

2.4

Oligonucleotide Arrays for sequencing

The first driving force for microarray development was the Human Genome Project (HUGO) conducted as an international exercise to decode the whole human genome for a first example [10]. The microarray approach of sequencing was the so called method of sequencing by hybridisation (SBH) [11]. It is based on the notion of combining the complete sequence of a sample by presenting all possible sequences as a complement on the chip, these are for 12 bases 412 (= 16,777,216) oligomers have to be presented on the chip. This goal was not only difficult to achieve technically in the early state of microarray fabrication technology, but also some basic features of natural sequences were neglected. For instance, it would be impossible to reach homogeneous hybridisation conditions for all sequences on one single chip since the melting temperature depends strongly on the ratio of GC to AT pairs within this sequence. But even worse there was no way of overcoming the problem of biologically meaningful redundancies within the genome originating from similar proteins, so it turned out to be impossible to get all the short probed sequences unambiguously linked together. Today SBH can be used in all those contexts where the sequence of interest is already known to a certain extent and deviations are sought.

2.5

Active arrays

Recently the concept of biosensors for measuring binding events and other biomolecular activities using surface bound molecules has been


177

adapted to microarray technology [12]. By use of time and spatially resolved measurement the kinetics of biomolecular interactions may be disclosed for a set of immobilised receptors in just one single experiment. A few examples are given in the next two sections.

2.5.1

Enzymes acting on immobilised DNA

In the early 1990s it was shown by several authors that enzymes modifying DNA are capable to act on immobilised DNA templates or primers [13,14]. Primers, oligonucleotides of less than 30 bases, are especially of interest in the context of microarrays. In our laboratory we achieved the parallel measurement of enzyme activities on several templates by virtue of a micro-fluidics chamber mounted on top of the microarray slide. For an example the restriction endonuclease (EcoRI) was chosen acting on different templates simultaneously. In this latter case a usual microarray covered with immobilised oligonucleotides is exposed to a variety of complementary sequences each of which is labelled with a fluorochrome (FITC in the actual case). By hybridisation in various spots the endonuclease’s cleaving site is formed, forcing the applied enzyme to bind. By addition of the cofactor Mg2+ the enzymes start to cleave and release the short oligonucleotide with the fluorochrome. Already in 1996 Buck and Buckle et al., published first results on the observation of polymerase acting on an immobilised template by use of an SPR device (Biacore) in a single channel. This approach could be extended to the enzyme telomerase that is responsible for the elongation of chromosomal ends during cell proliferation. Since its activity is downregulated in differentiated cells, the amount of activity has been established as a significant tumour marker in recent years [15]. Schmidt et al., succeeded to demonstrate that telomerase acts also on artificial telomers immobilised on a sensor surface [16,17]. The use of immobilised templates for polymerase processing was also succesful with long DNA including whole genomes. The method has been applied to the production of complete and translationable messenger RNA (on-chip transcription) [18,19]. Also for the analysis of single nucleotide polymorphisms (SNPs) enzymes acting on immobilised templates have been employed. This was first demonstrated by Erdogan et al., who used DNA polymerases to elongate surface bound primers [20]. These authors also showed the reaction on an oligonucleotide microarray. They were using immobilised allelespecific oligonucleotide primers on a glass slide. Single stranded PCR products serve as template that hybridise to the corresponding oligonucleotide probes on the microarray. The match and mismatch primer differ


178 8

at their 3’-end by a variable base which is discriminated by the DNA polymerase during the elongation process and the incorporation of Cy3labelled dUTP owing to the corresponding signal intensity [20]. In this work a limit for the template length was stated at 5.7 kb. Recent work by von Nickisch-Rosenegk et al., could not find such a limit and showed on-chip elongation up to 16.5 kb, i.e., the whole mitochondrial genome [21].

2.5.2

PCR on the Chip

Whilst all reactions described up to now have been one-step reactions, a complete PCR on a chip needs to reproduce the polymerisation step several times, and to heat up the whole chip to 95°C for denaturation between every polymerisation step. A complete microarray based amplification (on-chip PCR) was first described by Adessi et al., [22]. For the amplification the authors proposed the specific covalent attachment of oligonucleotide primers via their 5’-end to the glass slide, allowing the free 3’-end to prime the DNA synthesis. Template DNA can hybridise to these free 3’-ends and will be available for elongation by a DNA polymerase. After elongation, the resulting amplification product is covalently attached to the surface via the primer. Detection of the immobilised amplicon can be achieved by direct labelling with fluorescently labelled primers or dUTPs. Another possibility may be the use of intercalators such as SybrGreen, but because of resulting background problems this option has not been described as being successful so far. The same is true for fluorescent dendrimer labels, e.g., 3DNA Labels from Genisphere which are available for microarray hybridisations. For further sequence verification the amplificates can be denatured with alkali and subsequently hybridised with a sequence specific fluorescent probe. Primers used for On-Chip PCR have to fulfil three criteria besides the usual properties such as specificity, G/C-content, and melting temperature. They have to be immobilised in a density that allows the detection of the resulting amplicons. Adessi et al., state that for the immobilised primer a concentration of 50 µM is best suited for On-Chip PCR experiments [22],


179

however, they fail to determine the actual primer density on the glass slide they could achieve by their immobilisation procedure. The linkage of the primer to the surface must be stable enough to withstand the extreme temperature conditions during the cycling process. Therefore covalent attachment (e.g., EDC coupling) is the best option. The last important criterion to be considered is the attachment of the primer at the 5’-end and the introduction of a spacer sequence for superior hybridisation efficiency. The length and composition of the spacer sequence can be variable. An optimal length is described to be between 10 and 18 nucleotides, preferentially as polyT-spacer [22,23]. The amplification process takes place in two different phases. In the liquid phase and on the solid phase. Fig. 6-6 and 6-7 represent a schematic overview of the general principle of On-Chip PCR.

Figure 6-6. Principle of the On-Chip PCR Liquid phase amplification.

180


Liquid phase PCR is working in the same way as a common PCR. The Template DNA is denatured in the first step and free primers in solution hybridise during the annealing step. After the annealing the primers are elongated at their 3’-ends by the DNA polymerase to produce a new copy of the template. To prevent liquid phase amplification becoming predominant over the solid phase amplification the free reverse primer is used in a ratio of 1:8 compared to the Cy5-labelled forward Primer (ratios may differ from experiment to experiment and should be determined individually). This should allow primary synthesis of PCR products which will be used in the on–going solid phase PCR as template.

Figure 6-7. Principle of the On-Chip PCR: Solid phase amplification.

In the solid phase PCR the template DNA, which can be either from the liquid phase or the original template, hybridises to the specific 5’-bound primer (1) during the annealing step (2). After the annealing the elongation step by the polymerase follows (3). In the next cycle, after denaturation and annealing of the Cy5-labelled Primer (4), we shall find the final Cy5-labelled PCR amplificate 5’-bound to the surface (5) and ready for subsequent analysis. This process is repeated during every cycle of the On-Chip PCR.


181

Because the cycling takes place directly on the slide surface a special containment for the PCR-Mix is needed. To prevent vaporisation of the PCR-Mix this containment has to be absolutely leak proof even at elevated temperatures. For this purpose there are two different options available at the moment. The first option is the usage of adhesive reaction chambers, such as SealFrame (MJ Research, USA) or HybriWells (Sigma Aldrich, USA). Another option is the use of the SelfSeal reagent (MJ Research, USA) which is added directly to the PCR-Mix and builds a tight seal at the edges of a coverslip by polymerisation during the first denaturation cycle. Both options have been successfully applied for the On-Chip PCR [20,22,24,25,26]. The thermocycling of the On-Chip PCR is carried out in in situ PCR blocks. Usually the 16×16 Twin Tower in situ block for the PTC 200 thermocycler from MJ.R.esearch is used for this purpose [20,22,24,25,26]. Sensitivities for the On-Chip PCR are reported to be in the range of picoto nano-molar DNA template concentrations [22] or between 30–100 ng genomic DNA [20,24,25]. Because On-Chip PCR has been developed quite recently, only a few reports have been given about applications. Huber et al., were focussing on the detection of single base alterations in genomic DNA (SNP analysis). In advancing the method the authors focus on sequence specific genotyping by solid phase amplification [24,25]. Because there is no need of template preparation, advantages such as time and cost savings are obvious when using the On-Chip PCR. Huber et al., were able to show that their experimental setup is suitable for the detection of SNP s in the human tumour suppressor gene p53 [24]. Further integration might be achieved if the single On-Chip PCR is extended for parallel detection of several SNP s in one reaction: a Multiplex On-Chip PCR. Huber et al., were able to show the Multiplex On-Chip PCR for accessing SNP s in genomic DNA [25]. In this paper the genotyping of 10 different polymorphic sites within 7 human genes by direct On-Chip Multiplex PCR has been shown. Mitterer et al., reported On-Chip PCR in a combination of universal primer pairs targeting for the Helix 43 and 69 region of the 23s rDNA and species specific primers immobilised on the chip surface were used to detect 22 common bacteria causing infertility and abortion in mares [26]. Although false positive signals were obtained on some rare occasions, this work shows possible future perspectives for the On-Chip PCR in clinical diagnostics.


182 2

Recently we demonstrated the Multiplex On-Chip PCR with a variety of markers for food ingredients [27,28]. The area of diagnostic applications is still under development in the field of chip based assays. A technique that could give fast and reliable answers about, e.g., the current state of an individual‘s viral or pathogen load, would be of great benefit. With further optimisation the sensitivity of the PCR and the miniaturisation and parallelism of microarrays could join in form of the On-Chip PCR and build a core part in diagnostic BioMEMS for future point of care devices.

2.6

Integrated PCR

The alternative route to miniaturised PCR is based on microsystems technology. As described above, the Lab on a chip (or µTAS) approach brings together novel micro-fluidic systems with biochemistry. Micro capillaries, micro cavities, and micro pumps may be fused with heating and detection units. The advantages of this approach compared to conventional PCR are, besides smaller sample volume and reduced consumption of reagents, that heat transport can be made faster, speeding up the entire analysis time. Moreover, a smaller sample volume should help to make analysis more sensitive [29]. Recently a new generation of PCR machines has been launched by several companies, that make use of miniaturising sample volume and optimised heat transfer, consequently significantly speeding up the processing time. These themocyclers, however, are not integrating fluidics or any kind of sample handling, and therefore are not a platform for future automation and integration.

2.6.1

Micro-chamber Chips

The Lab on chip route for PCR was pioneered by Northup et al., who developed a micro-fabricated reaction chamber, the fabrication technology being compatible with more elaborate integration steps [30]. There are two different development lines for miniaturised PCR: Microchambers and micro-fluidics. The first creates reaction chambers in the range from 1–20 µl. The reagents may be applied from the outside by micro dispensing tools (as described in section 2.2.2), or added in advance and freeze dried to be activated by sample addition [31]. The chambers are heated and cooled actively by Peltier elements achieving heating and cooling rates which are similar to or even faster than the above mentioned advanced conventional machines [32,33].

DNA BASED BIO–MEMS 2.6.2

183

Micro-fluidics Chips

Another option of micro-chamber chips is generating a micro-chamber divided into three different temperature compartments for denaturation, annealing, and elongation. By usage of micro pumps the sample is moved from one compartment to the other [34], simulating the cycling of conventional machines. This approach is leading to the second group of miniaturised PCR systems, the micro-fluidics PCR chips. Micro-fluidic chips typically employ meander like channels where the sample, ranging from 12 nl to 10 µl, is moved by micropumps over three differently tempered zones [35,36,37,38,39]. In these devices cycling times are determined independently from heating and cooling rates, by channel dimension and flow rate. By injection of mineral oil [38] or air bubbles filled with washing solution it is possible to divide different samples and amplify them continuously in a row [39]. Curcio et al., recently showed a new approach for dividing 300 nl samples in microfluidic chips by using organic solvents such as decan [40]. Avoidance of cross-contamination is achieved by using washing solution in the organic solvent phase. Very little volumes as of 12 nl PCR-Mix could be successfully amplified by Liu et al. [41] in a nano-liter rotary PCR chip (2.5 mm radius). By using this circular channel very small dimensions of the chip layout is possible. Cycling is achieved by pumping the reaction mix via the circular channel over the three incorporated tempered zones. Manufacturing of miniaturised PCR modules is done using different materials like glass [37,42,43,44], silicon [41,45,46,47,48], ceramics [35], and teflon [40], but also various polymers are used for this purpose. When choosing one of the above mentioned materials it is important to ensure that there will not be any inhibitory effect or absorption causing inefficiency of the PCR [49,50]. In the past silicon and silicon nitride have been identified as causing inhibition of the PCR [46]. For minimizing these inhibitory effects, silicon oxide and different polymers as well as silanisation of the chip material, are employed [48]. It should be mentioned that there are some drawbacks, given that there is a high surface/volume ratio silicon oxide will not be able to avoid the inhibitory effect, and silanisation shows a decrease in stability with ongoing PCR cycling [29]. BSA is widely used to passivate the surface of glass/silicone chips against absorption of PCR components, especially the DNA polymerase [44]. Polymers have two advantages: they are compatible with most of the chemical and biochemical assays. Moreover they are cheap materials which can be structured easily by imprinting. For manufacturing of PCR Chips different materials such as Polydimethylsiloxane (PDMS) [36,41,51], Polyimide [52], Polycarbonate [35], and acrylic glass [53] are used. Multiple

184 4


use of SU-8 is observed in the field of microsystem fabrication [32,54]. One reason is that fluidic components with superb chemical and thermal resistance can easily be structured by means of photo–lithography, making them well suited to PCR experiments [55]. However there could be still some need for surface passivation, e.g., silanisation of SU-8 [32] or PDMS [36], to enhance PCR recovery. Tempering of the miniaturised PCR modules is carried out in two different ways. First, the most obvious way of heating the module is by direct contact with the heating device. For this purpose copper blocks or peltier elements are commonly used [39,44,48,51,53]. Considering the further integration towards µTAS thin film heating elements of materials such as platinum [32,34,45], indium–tin–oxide [36,37], wolfram [41], nickel silica [32], or polysilica [49] are structured by photo–lithographic methods on the chip surface and contacted by aluminium pads. The second way is a contact-free heating by infrared radiation. Because of the specific absorption spectrum it has the advantage of selectively heating the reaction mix but not the chip itself [42,52]. Measurement of the temperature on a chip surface is carried out mainly by use of contact thermometers which can be integrated into the chip as well. Integrated temperature sensors are commonly platinum thermometer [36,37,38,39,54]. Temperature control occurs either via PID control, programmed in LabView [32,49,51,56], or as an integrated micro-control unit [33,38,39,53]. Miniaturised PCR modules are needed in the course of further Labon-a-chip development, because they play a crucial role in the sample analyzing. In combination with other compartments, such as sample preparation (e.g., cell separation, DNA isolation) and analyzing compartments such as capillary electrophoresis [43,51] or detection via laser induced fluorescence respectively electrochemical signals [39,44,47,56], highly integrated BioMEMS can be designed. However, the compartments described above are available; a fully integrated BioMEMS has still to be developed.

3.

NANO-BIOTECHNOLOGY: DNA AS MATERIAL

The self-assembly of complex structures at molecular level in the living cell has inspired the scientific world. Bio-engineers aim to mimic this


185

process. DNA is a promising candidate for this task since it offers excellent specificity in base-pairing combined with addressability at the nano-meter scale. Therefore DNA based construction has been investigated intensively in recent years. In the following section we will give a brief introductory overview of this emerging field. Several conferences have been devoted to this topic [57,58,59], and some aspects of this topic are reviewed in the book by Niemeyer and Mirkin [60]. The applications envisioned for such constructs reach from bioanalysis to nano-electronics and DNA computing. Physically, DNA may be viewed as a structuring and constructing material: The sequence complementarity may be used to direct the arrangement of DNA molecules to build complex and non linear structures, different from the usually given picture of DNA as a mere information string. This notion has first been raised by Seeman [61,62]. He and his coworkers aimed at the generation of regular arrangements in two and three dimensions.

3.1

DNA directed immobilisation and nucleic acid tags

The most simple application of DNA’s high specific recognition capability is its use as a tag to direct and attach a functional unit to a solid phase, e.g., a surface area or a particle. The DNA tag has two features different from any other tag. First, it is highly specific and addressable, so a great variety of tags may be used simultaneously and, second, the DNA tag is reversible, which means it may be switched, e.g., by extreme pH or heat. The first feature allows using any DNA microarray to switch into another type of chemical array, e.g., a protein array. The method of DNA directed immobilisation (DDI) was first introduced by Niemeyer [63] and has since then been applied to various biochemical and bioanalytical examples [60]. The principle of DDI is depicted in Fig. 8.

186


Figure 6-8. Scheme of site specific immobilisation of functional entities by use of nucleic acid tags.

DNA oligomers fixed on a surface guide proteins or any other entities which bear a complementary DNA oligomer as a tag. Alternatively, other nucleic acids may also be used, peptide nucleic acids (PNA) are especially discussed in this context. PNA has been designed to have same spacing between neighbouring bases in the double strand with natural DNA (i.e., 0.34 nm), but has a chemically different backbone, coupled by peptide bonds [64]. Owing to its synthetic nature PNA has the advantage of not being degraded by enzymes, it also produces very stable hybrids both with DNA and RNA [65]. The second feature specific for DNA tethered surface coupling is the potential of regeneration of surfaces and re-use with an altered application. The DNA double strand formed by the DNA tag with its complement attached to the solid phase may be released into its complementary single strands by heat or alkaline treatment separating the functional unit from the solid phase. After removing the tagged functional unit the DNA on the solid phase is ready to hybridize again with any complementary nucleic acid oligomer. This feature has been demonstrated for immuno assays [66]. In this example a hapten (A), a small molecule to which antibodies have been gained, was linked covalently to a DNA oligomer and thus forms an A-DNA conjugate. The complementary DNA oligomer was immobilised on a sensor


187

surface. Hapten A was thus immobilised by hybridisation, an immuno assay was performed on the sensor surface. Afterwards the tagged hapten A was removed by alkaline washing, and another reagent, hapten B., was immobilised using a B-DNA conjugate with the same tag, i.e., the same sequence of DNA. Again a complete immuno assay was performed, demonstrating that hapten B was immobilised with the same efficacy as hapten A. In the case reported the haptens were pesticides from different chemical classes (atrazin (A) and 2,4-dichlorophenoxy acetic acid (B). Oligomer tags may thus be used to switch any surface, especially arrays of DNA oligomers to a protein chip, or any kind of chemical library. Alternatively, DNA may also be replaced by other nucleic acids, PNA should again be mentioned here as an example because of its high binding strength and independence from bio-active interference, such as enzymatic activity in physiological samples.

3.2

DNA for regular structures

In nature DNA occurs as a huge linear macromolecule. However, base complementarity may also occur in distinct DNA strands, leading to the possibility of cross-over from one strand to the other. Whilst in the living cell this happens only exceptionally, the phenomenon may serve as a fundamental building step for supra-molecular structures. Since the early work of Seeman [61] many attempts have been made to generate regular structures in the form of DNA grids. Owing to the high flexibility of long DNA strands, the construction of grids has to introduce special elements to stiffen the structure. This problem has been successfully addressed by the design of double-double stranded DNA with four cohesive ends, so called DX molecules (in fact, these are supra-molecules). The length of all components may be chosen in that way of initiating a flat grid, or, if wanted, a curve or even a wave form. Rigid and flat surface structures were first calculated by Winfree et al., and demonstrated also in the Seeman lab [41,67]. Atomic force microscopy (AFM) was used to validate a regular structure where markers (e.g., hairpin structures) were found at the distance and regularity predicted by calculation (65 nm, whilst 64 nm was calculated from the grid formation assuming classical B-form DNA). In Fig. 6-9 an example is given which shows how a series of DNA molecules may be used to fill a plane. Generation of all such constructs is guided by self assembling through base recognition.

188 8


Figure 6-9. A regular grid generated by DNA basic building blocks which form the regular structure by self-assembly.

This technique, sometimes called tiling, allows a positioning, particles for instance, in a regular way. The primary step in this approach is to form the basic building blocks, i.e., DNA molecules that assemble by base pairing of so called ‘sticky ends’. These ends are overlaps of several nucleotides. They guide the self-assembling process which is finalised by forming a covalent bond in the backbone by an enzymatic step, the ligation, a process carried out in the living world by enzymes called ligases. Many different kinds of super-molecular structures have been generated by this method: in two-dimensional ribbons surface coverage by tiling to form grids has been shown [68]. In three dimensions cubes and octahedra have been demonstrated. The process route of all these examples is: forming linear DNA oligomers, assembling by base-pairing, covalent bond formation by ligation; briefly, first, self-assembling, and second, covalent coupling. Von Kiedrowski demonstrated an alternative way of creating structures based on nucleic acids. He proposed reversing the fabrication order: first, create branched, covalently linked oligonucleotides, then assemble by base pairing. The basic notion is to form chemically coupled oligonucleotides, so called tris-oligomers which have distinct sequences in each branch. The complementary oligonucleotides couple by base pairing and give access to other functionalities. This approach is capable of forming similar constructs such as cubes, as in the abovementioned Seeman approach [69]. Von Kiedrowski also showed that such molecules are capable of copying and self-replication in a manner as described by his own group a few years earlier [70].


3.3

189

DNA to structure surfaces

A quite different approach aims at the arrangement of arbitrary structures completely directed by the sequence. One first task is to stretch DNA that is naturally (entropically) coiled. Two techniques are reported in the literature: fluid streams to ‘comb’ long DNA stretches by virtue of mechanical forces; second, forces exerted by an alternating electric field stretch DNA. New applications in bio sensing, e.g., interaction analysis at the single molecule level, or nano-biotechnology, e.g., ultradense DNA microarrays, have been devised that rely on stretched DNA. Also the construction of nano-wires for nano-meterized electronics may benefit from such techniques (see below). The basic technology required is to deposit spatially defined, stretched DNA bridges between anchoring structures on surfaces. For this latter purpose in our lab we developed the method of entrapment of macro molecules in a polymer network generated by electro polymerisation [71]. A non-conducting polymer is generated by electro polymerisation in the presence of streptavidin molecules that are saturated with biotinylated oligomers. This immobilisation procedure restricts the attachment of nucleic acid oligomers for any other biotinylated species to one electrode that has been activated during the polymerisation procedure and may be repeated several times on various electrodes within one setting. A prerequisite for this approach is a well defined electrode structure, and spatial resolution depends strongly on the quality of the edge forming parameters of electrode production. More freedom in construction and surface chemistry is promised by the recently developed method of dip pen nano-lithography (DPN), which has recently been introduced by Mirkin [72]. Based on an AFM stage, an AFM-tip is used as a pin tool to deposit a low amount of activated oligomers with a high spatial accuracy in the nm range.

3.3.1

Stretching of DNA by fluidics

At the meniscus of a water/air interface the forces of flow, e.g., generated by evaporation, act perpendicular to the interface. Long molecules are aligned along the direction of the force vector. This method applied to polymer molecules is called molecular combing. The method of molecular combing was primarily developed to stretch DNA to investigate molecular forces [73]. These forces may be estimated from comparison of all forces acting on the air/water interface. The technique has been advanced recently by Fritzsche and co-workers to bridge long dsDNA over nano- electrodes, so that only a few, or, optimally, only a single, DNA molecule is involved. Combining this method with nano-meter-sized electrodes, single molecule electronics become accesible [58].


190 0

3.3.2

Stretching DNA by AC electric fields

It was Washizu who first reported on the stretching of long DNA-double strands in alternating electric fields [74]. He and his co-workers showed that at a modest field strength of several V/m and a frequency of around 1 MHz long dsDNA may be stretched to full length expected from B-form DNA. The method has been developed further by Hölzel [75,76,77], who also showed that even single stranded DNA may be stretched to a linear line applying high frequency AC fields [76]. The mechanism was described as an effect similar to dielectrophoresis which is widely used for cell manipulation [78,79]. A polarisable particle, cell, or macromolecule forms a dipole in an electric field. In a homogeneous field the dipole will not be moved and no force will be effective, since the charges on the particle compensate each other. In a non-homogeneous field a dipole feels a force along the gradient of the field. The gradient is still acting when the field switches polarity. While at high frequences (~1 MHz) the net movement of charges will compensate, a dipole will be attracted along the field lines of the gradient. However, for a polymer that is polarisable in its segments, a gradient might not be necessary. Local dipoles are formed that constrain the polymer to minimal energy configuration: this is, under the action of an alternating field, the completely stretched form (Fig. 10).

Figure 6-10. Schematic representation of a segmented polymer, forming local dipoles that are aligned by virtue of an AC field.


191

However, all experiments described so far are made with a gradient; in fact, it is difficult to avoid it in small dimensions in the micrometer range. Even at inter-digitated electrodes there is a strong gradient at the edges of the electrode structure in the z-direction. Whilst the elucidation of the mechanismen might be mainly of academic interest, the method may well be employed to provide linear DNA for arranging tagged entities in the nano-meter range. An example has been given by our group with fluorescent particles that have been tagged with PNA-oligomers that were capable of formin triple-helix or invade dsDNA. The PNA-oligomers were complementary to two sites on a M13-DNA (7.2Kb) with 1 µm spacing. The particles could well be distinguished with a standard fluorescence microscope (Fig. 6-11).

Figure 6-11. Double stranded DNA of 7.2 kb = 2 µm length is stretched by an AC electric field. Two sites have been addressed by two fluorescent nanoparticles of different size. The particles bearing a PNA tag specific for two sites within the DNA are positioned precisely at a distance of 1 µm corresponding to the distances of the recognition sites on the completely stretched DNA.

3.4

Metallisation of DNA for electronic circuits

A most important and also promising field of application of DNA nanostructures, is nano- electronics. Apart from the dispute about the conductivity and the mechanism of charge transport within a double stranded DNA molecule, the self-assembling potential of DNA has inspired engineers to think about the molecular bottom-up construction of nanoelectronic circuits. One route in this field follows the direction of starting with DNA as a blueprint for electronic circuity and then guiding metallic particles along a DNA structure to form nanowires that are already connected. Pompe and coworkers used nucleic acids as condensation spots for metallic adsorbants to model ultra thin current lines for microelectronic application [80].

192 2


The examples given demonstrate that the structures formed are mainly created by chance, and thus the need for a ‘rational approach’ to direct the DNA on the artificial surface in a biohybrid system. To gain full access to the pattern formation and self-assembling capability of DNA it is necessary to fix the DNA molecule precisely on a surface.

4. [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11]

[12] [13] [14] [15]

REFERENCES Manz, A., Eijkel, J.C., van den Berg, A., (2004), ‘Cyclic electrophoretic and chromatographic separation methods’, Electrophoresis, 25, 243–52. Manz, A., Vilkner, T., Janasek, D., (2004) ‘Micro-total analysis systems. Recent developments’, Anal. Chem., 76, 3373–85. Nature Supplement, (1999), ‘The Chipping Forecast 1’. Nature Supplement, (2002), ‘The Chipping Forecast 2’. Schena, M., Microarray Analysis 2003 Cold Spring Harbor Press. Eisen, M.B., Brown, P.O., (1999), ‘DNA Arrays for Analysis of Gene Expression’, Methods Enzymol, 303, 179–205. Schena, M., (1999), ‘DNA Microarrays: A Practical Approach’, Oxford Universitiy Press, Oxford. Sambrook, J., (2003), DNA Microarrays: ‘A Molecular Cloning Manual’, Cold Spring Harbor Press. Fodor, S.P.A., Read, J.L., Pirrung, M.C., Stryer, L.A.T., Solas, D., (1991), ‘Light-directed, spatially adressable parallel chemical synthesis’, Science, 251, 767–773. International Human Genome Sequencing Consortium, (2001), ‘Initial sequencing and analysis of the human genome’, Nature, 409, 860–921. Drmanac, R., Drmanac, S., Chui, G., Diaz, R., Hou, A., Jin, H., Jin, P., Kwon, S., Lacy, S., Moeur, B., Shafto, J., Swanson, D., Ukrainczyk, T., Xu, C., Little, D., (2002), ‘Sequencing by hybridization (SBH): Advantages, Achievments, and Opportunities’, Advances in Biochemical Engineering/Biotechnology, 75–101. Bier, F. F., Kleinjung, F., (2002). ‘Determination of the turnover number of the restriction endonuclease EcoRI using evanescent wave technology’, Anal. Bioanal. Chem., 372, 308–313. Buckle, M., Williams, R.M., Negroni, M., Buc, H., (1996), ‘Real time measurements of elongation by a reverse transcriptase using surface plasmon resonance’, Proc. Natl. Acad. Sci., USA., 93, 889–894. Bier, F.F., Scheller, F.W., (1996), ‘Label-free observation of DNA hybridisation and endonuclease activity on a wave guide surface using a grating coupler’, Biosensors & Bioelectronics, 11, 669–674. Granger, M.P., Wright, W.E., Shay, J.W., (2002), ‘Telomerase in cancer and aging’, Crit. Rev. Oncol. Hematol., 41(1), 29–40.

DNA BASED BIO–MEMS [16] [17]

[18] [19] [20] [21]

[22]

[23] [24]

[25]

[26]

[27] [28] [29]

193

Schmidt, P.M., Lehmann, C., Matthes, E., Bier, F.F., (2002), ‘Detection of activity of telomerase in tumor cells using fiber optical biosensors’, Biosensors & Bioelectronics, 17, 1081–1087. Schmidt, P.M., Matthes, E., Scheller, F.W., Bienert, M., Lehmann, C., Ehrlich, A., Bier, F.F., (2002), ‘Real-time determination of telomerase activity in cell extracts using an optical biosensor’, Biol. Chem., 383, 1659–1666. Steffen, J., von Nickisch-Rosenegk, M., Bier, F.F., (2004), ‘Transcription of Reporter Genes with Immobilized Templates’, Book of Abstracts: Dechema Chip-Status-Seminar 2004, Frankfurt a.M. Steffen, J., von Nickisch-Rosenegk, M., Bier, F.F., (2005), ‘In vitro transcription of a whole gene on a surface-coupled template’, Lab on Chip (in press). Erdogan, F., (2001), ‘Detection of mitochondrial single nucleotide polymorphisms using a primer elongation reaction on oligonucleotide microarrays’, Nucleic Acids Research, 29, 36. Nickisch-Rosenegk, M., Marschan, X., Andresen, D., Abraham, A., Heise, C., Bier, F.F., (2005), ‘On-chip PCR amplification of very long templates using immobilized primers on glassy surfaces’, Biosens. Bioelectron., 20, 1491–1498. Adessi, C., Matton, G., Ayala, G., Turcatti, G., Mermod, J.J., Mayer, P., Kawashima, E., (2000), ‘Solid phase DNA amplification: characterisation of primer attachment and amplification mechanisms’, Nucleic Acid Research, 28(20), e87. Southern, E., Mir, K., Shchepinov, M., (1999) ‘Molecular interactions on microarrays’, Nature Genetics supplement, 21, 5–9. Huber, M., Losert, D., Hiller, R., Harwanegg, C., Mueller, M.W., Schmidt, W.M., (2001), ‘Detection of single base alterations in genomic DNA by solid phase polymerase chain reaction on oligonucleotide microarrays’, Analytical Biochemistry, 299, 24–30. Huber, M., Mündlein, A., Dornstauder, E., Schneeberger, C., Tempfer, C.B., Mueller, M.W., Schmidt, W.M., (2002), ‘Accessing single nucleotide polymorphisms in genomic DNA by direct multiplex polymerase chain reaction amplification on oligonucleotide microarrays’, Analytical Biochemistry, 303, 25–33. Mitterer, G., Huber, M., Leidinger, E., Kirisits, C., Lubitz, W., Mueller, M.W., Schmidt, W.M., (2004), ‘Microarray based identification of bacteria in clinical samples by solid-phase PCR amplification of 234S ribosomal DNA sequences’, Journal of Clinical Microbiology, 42(3), 1048–1057. Andresen, D., Ehrentreich-Förster, E., Bier, F.F., Kuhn, M., (2004), ‘Development of a DNA Chip for Detection of Food Allergens’, Poster at Biosensor Worldcongress Granada (Spain). Andresen, D., Ehrentreich-Förster, E., Bier, F.F., Kuhn, M., (2005), ‘Multi-plex OnChip-PCR for the analysis of food ingredients’, Poster at German Biosensor Symposium Regensburg (Germany). Felbel, J., Bieber, I., Pipper, J., Köhler, J.M., (2004), ‘Investigations on the compatibility of chemically oxidized silicon (SiOx)-surfaces

194 4

[30]

[31]

[32]

[33] [34]

[35]

[36] [37] [38] [39]

[40] [41] [42]

DNA BASED BIO–M –MEM MS for applications towards chip based polymerase chain reaction’, Chemical Engineering Journal, 101, 333–338. Northrup, M.A., Ching, M.T., White, R.M., Watson, R.T., (1993), ‘DNA amplification with a micro-fabricated reaction chamber’, Tranducers ’93: Proceedings of the 7th international conference on solid-state Sensors and Actuators, Yokohama, Japan. Deshpande, M., Greiner, K.B., Romanowicz, B.F., Gilbert, J.R., St John, P.M., Woudenberg, T., (1999), ‘CAD analysis of PCR well containment’, Technical Proceedings of the 1999 International Conference on Modeling and Simulation of Micro systems, 2, 350–354. El-Ali, J., Perch-Nielsen, I.R., Poulsen, C.R., Jensen, M., Telleman, P., Wolff, A., (2003), ‘Micro-fabricated DNA amplification device monolithically integrated with advanced sample pre-treatment’, TRANSDUCERS ’03: Proceedings of the 12th International Conference on Solid State Sensors, Actuators and Micro systems, Boston, USA. Zhao, Z., Cui, Z., Cui, D., Xia, S., (2003), ‘Monolithically integrated PCR biochip for DNA amplification’, Sensors and Actuators, A108, 162–167. Bu, M., Melvin, T., Ensell, G., Wilkinson, J.S., Evans, A.G.R., (2003), ‘Design and theoretical evaluation of a novel micro-fluidic device to be used for PCR’, Journal of Micromechanics and Microengineering, 13, 125–130. Chou, C.F., Changrani, R., Roberts, P., Sadler, D., Burdon, J., Zenhauern, F., Lin, S., Mulholland, A., Swami, N., Terbrueggen, R., (2002), ‘A miniaturized cycling PCR device—modeling and experiments’, Microelectronic Engineering, 61–62, 921–925. Fukuba, T., Yamamoto, T., Naganuma, T., Fujii, T., (2004), ‘Microfabricated flow-through device for DNA amplification—towards in situ gene analysis’, Chemical Engineering Journal, 101, 151–156. Sun, K., Yamaguchi, A., Ishida, Y., Matsuo, S., Misawa, H., (2002), ‘A heater-integrated transparent microchannel chip for continuousflow PCR’, Sensors and Actuators B84, 283–289. Schneegaß, I., Bräutigam, R., Köhler, J.M., (2001), ‘Miniaturized flow-through PCR with different template types in a silicon chip Thermocycler’, Lab-on-a-chip, 1, 42–49. Obeid, P.J., Christopoulos, T.K., Crabtree, H.J., Backhouse, C.J., (2003), ‘Micro-fabricated device for DNA and RNA amplification by continuous-flow polymerase chain reaction and reverse transcriptionpolymerase chain reaction with cycle number selection’, Analytical Chemistry, 75, 288–295. Curcio, M., Roeraade, J., (2003), ‘Continuous segmented-flow polymerase chain reaction for high throughput miniaturized DNA amplification’, Analytical Chemistry, 75(1), 1–7. Liu, J., Enzelberger, M., Quake, S., (2002), ‘A nanoliter rotary device for polymerase chain reaction’, Electrophoresis, 23, 1531–1536. Oda, R.P., Strausbauch, M.A., Huhmer, A.F.R., Borson, N., Jurrens, S.R., Craighead, J., Wettstein, P.J., Eckloff, B., Kline, B., Landers,


[43] [44]

[45] [46] [47]

[48] [49]

[50] [51]

[52]

[53]

[54]

[55]

195

J.P., (1998), ‘Infrared-mediated thermocycling for ultrafast polymerase chain reaction amplification of DNA’, Analytical Chemistry, 70, 4361–4368. Lagally, E.T., Emrich, C.A., Mathies, R.A., (2001), ‘Fully integrated PCR-capillary electrophoresis micro system for DNA analysis’, Labon-a-chip, 1, 102–107. Khandurina, J., McKnight, T.E., Jacobson, S.C., Waters, L.C., Foote, R.S., Ramsey, J.M., (2000), ‘Integrated system for rapid PCR based DNA analysis in micro-fluidic devices’, Analytical Chemistry, 72, 2995–3000. Lee, T.M.H., Hsing, I.M., Lao, A.I.K., Carles, M.C., (2000), ‘A miniaturized DNA aplifier: Its application in traditional Chinese medicine’, Analytical Chemistry, 72, 4242–4247. Poser, S., Schulz, T., Dillner, U., Baier, V., Köhler, J.M., Schimkat, D., Mayer, G., Siebert, A., (1997), ‘Chip elements for fast thermocycling.’, Sensors and Actuators A62, 672–675. Cheng, J., Waters, L.C., Fortina, P., Hvichia, G., Jacobson, S.C., Ramsey, J.M., Kricka, L.J., Wilding, P., (1998), ‘Degenerate oligonucleotide primed-polymerase chain reaction and capillary electrophoretic analysis of human DNA on microchip based devices’, Analytical Biochemistry, 257, 101–106. Shoffner MA., Cheng, J., Hvichia, G.E., Kricka, L.J., Wilding, P., (1996), ‘Chip PCR. I. Surface passivation of micro-fabricated siliconglass chips for PCR’, Nucleic Acids Res.,earch, 24, 375–379. Erill, I., Campoy, S., Erill, N., Barbé, J., Aguiló, J., (2003), ‘Biochemical analysis and optimization of inhibition and adsorption phenomena in glass-silicon PCR-chip’, Sensors and Actuators, B96, 685–692. Panaro, N.J., Lou, X.J., Fortina, P., Kricka, L.J., Wilding, P., (2004), ‘Surface effects on PCR reactions in multi-chip micro-fluidic platforms’, Biomedical Microdevices, 6, 75–80. Hong, J.W., Fujii, T., Seki, M., Yamamoto, T., Endo, I., (2001), ‘Integration of gene amplification and capillary gel electrophoresis on a polydimethylsiloxane.glass hybrid microchip’, Electrophoresis, 22, 328–333. Giordano, B.C., Ferrance, J., Swedberg, S., Hühmer, A.F.R., Landers, J.P., (2001), ‘Polymerase chain reaction in polymeric microchips: DNA amplification in less than 240 seconds’, Analytical Biochemistry, 291(1), 124–132. Yuen, P.K., Kricka, L.J., Fortina, P., Panaro, N.J., Sakazume, T., Wilding, P., (2001), ‘Micro-fluidics module for blood sample preparation and nucleic acid amplification reaction’, Genome Research, 11, 405–412. El-Ali, J., Perch-Nielsen, I.R., Poulsen, C.R., Bang, D.D., Telleman, P., Wolff, A., (2004), ‘Simulation and experimental validation of a SU-8 based PCR thermocycler chip with integrated heaters and temperature sensor’, Sensors and Actuators, A110, 3–10. Ghantasala, M.K., Hayes, J.P., Harvey, E.C., Sood, D.K., (2001), ‘Pattering, electroplating and removal of SU-8 moulds by excimer

196 6

[56] [57] [58] [59] [60] [61] [62] [63]

[64]

[65] [66] [67] [68] [69] [70] [71]

DNA BASED BIO–M –MEM MS laser micro-machining’, Journal of Micromechanics and Microengineering, 11, 133–139. Lee, T.M.H., Carles, M.C., Hsing, I.M., (2003), ‘Micro-fabricated PCR-electrochemical device for simultaneous DNA amplification an detection’, Lab-on-a-chip, 3, 100–105. Fritzsche, W., ed., (2002), ‘DNA based molecular construction’, AIP Conf. Proc., 640. Fritzsche, W., ed., (2004), ‘DNA based molecular electronics’, AIP Conf. Proc., 725. Reif, J., ed., (2004) ‘Foundations of Nano-science—Self-Assembled Architectures and Devices’, Book of Abstracts, Snowbird Cliff Lodge, Utah. Niemeyer, C.M., Mirkin, C.A., (2004), ‘Nanobiotechnology— Concepts, Applications and Perspectives’, Wiley-VCH, Weinheim. Seeman, N.C., (1982), ‘Nucleic acid junctions and lattices’, J. Theor. Biol., 99, 237–247. Seeman, N.C., (1999), “DNA engineering and its application to nanotechnology”, Trends Biotechnol., 17, 437–443. Niemeyer, C. M., Sano, T., et al., (1994), ‘Oligonucleotide-directed self-assembly of proteins: semisynthetic DNA streptavidin hybrid molecules as connectors for the generation of macroscopic arrays and the construction of supra-molecular bioconjugates’, Nucleic Acids Research, 22, 5530–5539. Egholm, M., Buchardt, O., Christensen, L., Behrens, C., Freier, S.M., Driver, D.A., Berg, R.H., Kim, S.K., Norden, B., Nielsen, P.E., (1993), ‘PNA hybridizes to complementary oligonucleotides obeying the Watson-Crick hydrogen-bonding rules’, Nature, 365, 566–568. Nielsen, P. E., (1999), ‘Applications of peptide nucleic acids’, Current Opinion in Biotechnology, 10, 71–75. Bier, F.F., Kleinjung, F., et al., (1999), ‘Changing Functionality of Surfaces by Directed Self-Assembly Using Oligonucleotides— The Oligo-Tag. Biotechniques, 27, 752–760. Winfree, E., Liu, F., Wenzler, L.A., Seeman, N.C., (1998), ‘Design and self-assembly of two-dimensional DNA crystals’, Nature, 394, 539–544. Yan, H., LaBean, T.H., Feng, L., Reif, J.H., (2003), ‘Directed nucleation assembly of DNA tile complexes for barcode-patterned lattices’, Proc. Natl. Acad. Sci. USA, 100, 8103–8108. Eckardt L.H., Naumann K., Pankau W.M., Rein M., Schweitzer M., Windhab, N., von Kiedrowski, G., (2002), ‘DNA nanotechnology: chemical copying of connectivity’, Nature, 420, 286. Luther, A., Brandsch, R., et al., (1998), ‘Surface-promoted replication and exponential amplification of DNA analogues’, Nature, 396, 245–248. Gajovic-Eichelmann, N., Ehrentreich-Förster, E., Bier, F.F., (2003), ‘Directed immobilization of nucleic acids at ultramicroelectrodes using a novel electro-deposited polymer’, Biosensors & Bioelectronics, 19, 417–422.

DNA BASED BIO–MEMS [72] [73]

[74] [75] [76]

[77]

[78] [79] [80]

197

Piner R.D., Zhu, J., Xu F., Hong S., Mirkin C.A., (1999), ‘“Dip-pen” nanolithography’, Science, 283, 661–663. Michalet, X., Ekong, R., Fougerousse, F., Rousseaux, S., Schurra, C., Hornigold, N., van Slegtenhorst, M., Wolfe, J., Povey, S., Beckmann, J.S., Bensimon, A., (1997), ‘Dynamic molecular combing: Stretching the whole human genome for high-resolution studies’, Science, 277, 1518–1523. Washizu, M., (1990), ‘Handling of biological molecules and membranes in microfabricated structures’, in Karube, I., (ed.), Automation in Biotechnology, Elsevier, Amsterdam, 113–125. Hölzel, R., (2002), ‘Single particle characterization and manipulation by opposite field dielectrophoresis’, Journal of Electrostatics, 56, 435–447. Hölzel, R., Gajovic-Eichelmann N., Bier F.F., (2003), ‘Oriented and vectorial immobilization of linear M13 dsDNA between interdigitated electrodes—towards single molecule DNA nanostructures’, Biosensors and Bioelectronics 5–6, 555–564. Hölzel, R., Bier, F.F., (2004), ‘Monitoring dielectrophoretic collection of DNA by impedance measurement’, in Fritzsche, W., (ed.) ‘DNA-based Molecular Electronics’, AIP Conference Proceedings 2004, 725, 77–83. Zimmermann, U., Arnold, W.M., (1983), ‘The interpretation and use of the rotation of biological cells’, in: Fröhlich, H., Kremer, F., (eds.), Coherent excitations in biological systems, Springer, Berlin. Green, N. G., Ramos, A., Morgan, H., (2000), ‘AC electro-kinetics: a survey of sub-micrometre particle dynamics’, Journal of Physics D–Applied Physics, 33, 632–641. Richter J., Mertig, M., Pompe, W., Monch, I., Schackert, H.K., (2001), ‘Construction of highly conductive nanowires on a DNA template’, Applied Physics Letters, 78, 536–538.

Chapter 7 SEPARATION AND DETECTION ON A CHIP To Stellan Hjertén, a pioneer of capillary electrophoresis

Richard B.M. Schasfoort and Anna J. TüdĘs University of Twente, MESA+ Research Institute, Biochip Group, P.O.Box 217, 7500 AE Enschede, Netherlands, email: [email protected]

Abstract:

In this chapter capillary electrophoresis (CE) in micro-fabricated devices is described. Firstly, those general aspects of the CE theoretical background are revisited which are necessary to understand the typical events on a chip. The theoretical introduction is followed by an overview of the building blocks of lab-on-a-chip devices and the requirements for their application in a CE chip device. The chapter concludes with the most relevant applications and an outlook describing future development of this area.

Key words: Micro-machining technology; bonding off wafers; cross or double-T channel injection; joule heating; electro-osmotic flow (EOF); electrochemical; optical and acoustic detectors in micro-fluidic devices; dielectrophoresis.

1.

INTRODUCTION

Capillary electrophoresis (CE) is a widely applied separation technique in analytical chemistry. CE separation is based on the differences in mobility of particles in an electrolyte when an electric field is applied. The basic technique, capillary zone electrophoresis (CZE), is only suited for the separation of charged particles, but the technique has been extended towards the separation of neutral molecules (e.g., micellar electrokinetic chromatography MEKC). Various capillary electrophoretic separation modes can be distinguished [1,2]. The different separation modes can be performed using basically the same equipment but different buffer systems. Only CZE and capillary gel electrophoresis (CGE) will be treated in this chapter


199

200

SEPARATION AND DETECTION ON A CHIP I

because these techniques are most relevant and most frequently applied in lab-on-a-chip. A trend for miniaturization of electrophoresis started during the past decade with the introduction of lab-on-a-chip devices or miniaturized total analysis systems (µ-TAS) [3,4,5]. Introduced by Manz et al., in 1990 [6], the term µ-TAS has been used for systems with strongly reduced size, capable of performing all analysis steps including sample preparation. Nowadays lab-ona-chip instead of µ-TAS is the widely used term, because it includes not only analysis but also synthesis and other chemical processes on a chip [7]. Microscale electrophoresis systems have distinctive properties as a result of their small dimensions. First of all, liquid flow is generally laminar. Owing to the lack of turbulence, diffusion is the only process for mixing fluids in narrow channels. Miniaturization offers flexibility in design towards a certain application concerning, e.g., separation efficiency and speed. Detection in miniaturized systems is a challenge, owing to the small physical dimensions of the detector and the minuscule detectable amounts. The choice of material, geometry of the device, and the fabrication technology are vital for optimal system performance. Table 7-1. Lab-on-a-chip building blocks Lab-on-a-chip building block Channel Pump Dialysis Separator Extractor Mixer Flow cell Filter Valve Detector

Description/example Formed e.g., by dry or wet etching Enables fluid transport Separation of compounds in solution through membrane Based on electrophoresis or chromatography (CE, CEC) Two-phase systems Splitting or coiling of laminar flow Address flow, confined manifold Porous silicon or membrane controlled devices Mechanical, flowFET, electro-wetting T, P, electrochemical, optical, acoustic

Research in miniaturization is primarily driven by the need to reduce costs by reducing the consumption of reagents, decreasing analysis times, increasing (separation) efficiency and to enable automation. Miniaturization of devices will offer advantages when rapid and selective monitoring is required, especially where simultaneous parallel operation is possible. Because microchips can be inexpensively manufactured in large quantities, disposable chips can eliminate the problem of carry over in analysis. Owing to their small size, portable systems become available for environmental monitoring and clinical point of care applications.


201

Widely accepted analytical techniques such as chromatography, flow injection analysis and electrophoresis have been carried out at a reduced scale in terms of size [8] and time [9]. The growth of microfluidic applications has created a need for building blocks from which fluidic systems can be constructed [10]. Table 1 contains a compilation of the presently available building blocks for unit operations in lab-on-a-chip devices. In this chapter the theoretical and practical requirements are discussed for the application of the lab-on-a-chip building blocks in CE chip devices, together with representative examples. The chapter is structured as follows: Theory of capillary electrophoresis on a CE chip – Mobility of ions, – Electro-osmotic flow, – Joule heating in microfabricated devices, – Separation efficiency of a CE chip, – Separation of bio macro molecules and particles; Building blocks of CE chip devices – Wafer materials, micromachining and wafer bonding, – Power supplies, pumping, injection and channel geometries, – Detection strategies; Examples; Dielectrophoresis; Outlook.

2.

THEORY OF CAPILLARY ELECTROPHORESIS ON A CE CHIP

2.1

Mobility of ions

On applying an electric field to an electrolyte solution the ions will migrate towards the oppositely charged electrode with an electrophoretic velocity (vi), determined by the electric field strength (E) and by the electrophoretic mobility of ion i (mi): vi = mi E The electrophoretic mobilities of ions are often tabulated as the limiting ionic mobility, i.e., the mobility of ions at infinite dilution [11]. The determining factor of the electrophoretic mobility of weak electrolytes is the degree of dissociation. The pH of the solution has a major influence on the degree of dissociation and on the effective electrophoretic mobility of weak ions or buffer ions. In Fig. 7-1 the effective electrophoretic mobilities of several buffer ions are shown as a function of the pH.


202

Figure 7-1. Effective electrophoretic mobility of buffer ions as a function of pH: (A) balanine; (B) HEPES (4-(2-hydroxyethyl)-1-piperazine ethanesulphonic acid); (C) glutamic acid; (D) phosphoric acid. With permission from [12].

A charged particle is accelerated in an electric field by an electric force (Fi) as follows: Fi = q i E where qi is the charge of the particle. The velocity of the particle increases until the electric force is counterbalanced by the frictional force given by Stokes’ law: Ffrr = -6 πriηvi where ri is the particle (ion) radius and η the dynamic viscosity of the medium. Combining the above equations (where Fi+Ffr = 0) the electrophoretic velocity of ions can be calculated: vi = qiE /6πriη The electrophoretic separation of particles will primarily be determined by the charge/size ratio (qi/ri) at a given field strength. A higher charge and smaller size increases electrophoretic velocity, whereas neutral particles


203

cannot be separated with the basic technique. The speed of separation can be increased by increasing the field strength or, if possible, reducing the viscosity of the electrolyte.

2.2

Electro-osmotic flow

The wall of glass capillaries contains weak acidic silanol groups which cause a negative surface potential if the pH of the electrolyte in the capillary is above the point of zero charge of the material [13]. As shown in Fig. 7-2, the silanol groups of the glass wall can be considered as weak fixed negative charges counterbalanced by cations of the electrolyte. The first layer of cations close to the silanol groups of the capillary surface form a rigid plane, the so called Stern layer. Further away the cations become more diffuse until they can move freely in the bulk of the solution. As also illustrated in the figure, the composition of the diffuse layer is different from that of the bulk solution. The distance at which the wall potential (φ) drops to 1/e or 36.79 % of the initial maximum value is the Debye length (κ−1 ).

Figure 7-2. Charge and potential distribution at a charged wall in an electrolyte [14]. The Stern layer, the diffuse layer including Debye length, is illustrated in the figure. Note that the wall potential (-f) is different from the zeta potential (-z-potential).

The Debye length, the distance over which the distribution of free charge carriers is affected, is inversely proportional to the square root of the concentration (in an electrolyte where both the cation and anion have a single charge). A few examples for the Debye length in biologically relevant media are included in table 2.

204


Table 7-2. Illustration of the Debye length in biologically relevant media Solution 0.15 M physiological solution 10 mM KCl 1 mM KCl

Debye length (nm) 0.8 ~3 10

Davies and Rideal [15] found that the pH near a charged surface (pHi) depends on the potential (φ) of the double layer and the pH in the bulk of the solution (pHs) as follows: pHi = pHs+ F φ / 2.3 RT where F is the Faraday constant, R is the gas constant and T is the absolute temperature. This equation corresponds to the Nernst equation [11] describing the potential shift as a function of the pH measured by a pH electrode (ideally 59 mV/pH) [16]. When an electric field is applied in the longitudinal direction of a glass capillary, the cations close to the capillary wall move towards the cathode (Fig. 7-3). Wrapped in the layer of cations, the bulk solution is transported in the direction of the cathode. The only plane of friction is between the stationary layer at the capillary wall and the layer of cations in motion. As illustrated in the figure, unlike in pressure-driven systems with parabolic flow profile (Poiseuille flow), the velocity of the bulk is constant resulting in a flat flow profile (plug flow). The velocity of the electro-osmotic flow is given by [17]: vEOF = ( ε0ε /4ʌη ) ζ E where ε0 and ε stand for the dielectric constants of vacuum and of the buffer, respectively, whereas ζ is the ζ-potential, the potential at the first moving layer at the capillary wall. In order to modify the magnitude and direction of the EOF either the electric field E or the ζ-potential should be modified. The principle of modifying or even reversing the EOF inside a capillary has been presented before [18,19,20]. In these investigations fused silica capillaries were used and the purpose was the enhancement of the electrophoretic separation efficiency, rather than the control of flows in integrated devices [21,22].


205

Figure 7-3. The flow profile (EOF) of electro-osmotic flow under influence of an electric field (E). Liquid motion is towards the cathode (negative electrode).

As the EOF is generated at the capillary wall, the viscosity η at the wall is one of the determining factors of the flow velocity. Additives can adsorb and change the viscosity and/or charge at the wall [23]. This way it is possible to influence, even reverse, the direction of the EOF [24]. If the viscosity of the bulk solution remains unchanged the electrophoretic mobility of ions in the electrolyte remains unaffected. The overall velocity of a particle in capillary electrophoresis is not only determined by the electrophoretic velocity of the ion [25] but also by the electro-osmotic flow. As illustrated in Fig. 7-4, the overall velocity of cations will be faster than the electro-osmotic flow. Neutral particles have no electrophoretic mobility, hence their overall velocity will be equal to the EOF. Negatively charged species will move in opposite direction to the EOF. The difference between the absolute values of their velocity (v and vEOF) will determine whether the negative species will reach the detection zone. In etched uncoated glass capillaries at pH > 4 the electro-osmotic flow overrules the electrophoretic mobility of most anions. Therefore, in capillary zone electrophoresis even most anions can be separated next to cations and neutral species.


206

V+

V tot+

Cation VEOF

V tot EOF

Electro-osmotic flow V-

V tot-

Anion V tot 0

Neutral species Electrophoretic mobility

Overall velocity

Figure 7-4. Illustration of the electrophoretic velocity of a positive, neutral, and negatively charged particle and the electro-osmotic flow in capillary electrophoresis.

3.

JOULE HEATING IN MICROFABRICATED DEVICES

Field strengths in the order of 500 V/cm are commonly used in CE as high electric fields are favorable with respect to migration time and efficiency of the separation [26]. In a capillary filled with a conductive medium (electrolyte) the applied electric field will generate current and heat (Joule heat). Heat dissipation is only possible through the capillary wall, inducing a parabolic temperature profile in capillaries with a round crosssection. Joule heating can induce adverse effects for the CE separation, e.g., by introducing conductivity changes and diffusion related peak broadening [27,28,29]. In conventional systems dissipation of Joule heat is the limiting factor in increasing the field strength and separation efficiency [30]. The generated Joule heat per time unit (Q/t) can be expressed as the function of the applied potential (V), the conductivity of the medium (Ȝ) and the radius of the capillary (r): Q/t= V2 r2 ʌ Ȝ /L The peak broadening owing to Joule heating increases rapidly with the field strength (E5) and with the capillary bore (r6) [31,32]. Joule heating can be limited by reducing the current density by limiting the applied voltage, or,


207

more often, by avoiding high conductivity buffers. The most efficient method, however, is by decreasing the channel diameter. Efficient and reproducible heat dissipation appears to be of major importance even in lab-on-a-chip devices for measurement reproducibility and system stability. When Joule heating is negligible the current is a linear function of the potential. At a high level of power input, the current vs. voltage increases more rapidly owing to the temperature induced increase of the conductivity. As an estimate, the tabulated values of the conductivity at a temperature of 298°K K (θ298) increase by 2% per degree. The temperature increase due to Joule heating in a lab-on-a-chip device was experimentally demonstrated in our laboratory as shown in Fig. 7-5 (unpublished results). The point where excessive Joule heating starts can be determined from the voltage where deviation from the straight (ohmic) line occurs. The rise in temperature can be estimated from the increase of the specific conductivity of the fluid.

Figure 7-5. Theoretical and experimental diagrams of the current vs. field strength relationship to investigate Joule heating in a lab-on-a-chip device (3 cm × 25 µm × 25 µm nitride coated silicon, filled d with 1:1 50 mM KCl/ethanol).

208

3.1


Separation efficiency of a CE chip

Separation efficiency can be expressed as the number of theoretical plates (N) of the separation system: N = (L/ı)2. Where L is the effective length of the capillary (to the detector), and ı stands for the zone broadening. The efficiency of the separation will be influenced by all factors contributing to zone broadening, e.g., axial diffusion, temperature gradients, sample-related effects (differences in conductivity, overload conditions, interaction with channel wall) and instrumental effects (injection volume, detection volume, race track effect): ı2tot = ı2difff + ı2temp + ı2sample + ı2instr In conventional systems zone broadening owing to longitudinal diffusion along the capillary is the most dominant factor, and separation efficiency can be expressed by the diffusion term ı2 = 2Dt, where D is the diffusion coefficient of the solute, and t stands for the migration time. Inserting this into the equation of the plate number gives (where length of the capillary to the detector is equal to the total length of the capillary): N = miV/2D. The separation efficiency will be determined by the voltage in a system where the diffusion coefficient and ionic mobility are constant. This means that separation channels can be short (a few cm) without loss of performance, leading to significantly faster analysis times. Under ideal conditions this equation can be further simplified to N = 20 qiV [33]. However, when the length of the capillary becomes shorter eventually the peak broadening contribution of instrumental effects such as injection and detection is no longer negligible. When the peak broadening owing to injection and detection need to be taken into consideration, the plate number can be approximated by [34]: N = 12 L2/ (ı2inj+ ı2det).


209

In microfabricated devices the injection zone broadening can be reduced considerably by pinching the sample at the injection intersection [35] or applying dynamic loading [36].

3.2

Separation of biomacromolecules and particles

Large biomolecules such as DNA and proteins are usually separated in gel-filled capillaries (capillary gel electrophoresis), because their intrinsic mobility differences would not allow sufficient resolution without the sieving effect of the gel. Biomacromolecules can have diffusion coefficients 100 times lower than small solutes, therefore not the diffusion but the zone broadening owing to injection (ı2inj) will often dominate separation efficiency. This is typically the case for DNA separations. Bousse et al., demonstrated [34] that plate numbers >105 can be achieved, allowing single base resolution of fragment sequencing in various applications. In order to separate or trap large molecules, viruses, or bacteria dielectrophoresis is a preferred technique. Dielectrophoresis (DEP) is a technique for the manipulation of uncharged polarizable particles in inhomogeneous fields induced by alternating voltages [37]. Electrophoresis and dielectrophoresis differ in modes of operation (DC or AC), homogeneous fields versus non-uniform fields, electrode geometries, and application areas. Dielectrophoresis is optimal for large particles, e.g., bacteria, whereas electrophoresis is optimal for small ions, (e.g., K+, Na+). A common aspect is that both electrophoresis and dielectrophoresis benefit from miniaturization.

4.

BUILDING BLOCKS OF CE CHIP DEVICES

4.1

Wafer materials, micromachining and wafer bonding

With the increasing number of fabrication methods available [38,39], the diversity of substrate materials has been expanding. For example, next to the existing materials of silicon and glass, new fabrication methods such as moulding and hot embossing enabled the use of polymers as cost effective alternatives for silicon and glass. Requirements for materials and methods used for CE chips include well defined channel geometry (sharp corners, parallel channel walls) and smooth surface to avoid peak broadening. The materials used in CE chips need to be suited for high voltages. When


210

disposable structures are required the price per chip is a critical parameter. If chips are reused, cleaning and regenerating the surface is a prerequisite. Silicon is traditionally the material for chip fabrication. A typical sequence of silicon chip fabrication developed at MESA+ Research Institute is shown in Fig. 7-6. The advantages of silicon include excellent structure definition with photolithography and etching processes, good mechanical and thermal properties. For example, low thermal expansion and excellent thermal conductivity are ideal characteristics in CE chips. Disadvantages of silicon are that it is brittle, not transparent, and that it is the most expensive CE chip material. Until recently the major obstacle to the widespread use of silicon structures in CE was that being a semiconductor, application of high voltages appeared to be virtually impossible. With an insulating top layer, e.g. of silicon oxide or nitride, silicon has been shown compatible with the high voltages of CE. 1,2 Dry etching

3,4

Anodic bonding

Anisotropic back etching/polymer deposition

5,6

Silicon

glass

thin (glass) layer

polymer, glue

Figure 7-6. Illustration of the fabrication sequence of silicon based micro channels.

Glass is an attractive alternative to silicon for CE chip fabrication because it combines many of the optimal characteristics of silicon (good structural definition with photolithography and etching processes, low thermal expansion, constant shape, good mechanical properties) with other desirable ones. An attractive feature of glass is that being transparent, direct optical detection on the chip is possible. Unlike silicon, glass is an insulator providing an optimal chip material for high voltage methods like CE. Also an advantage of glass is the relative low cost of the starting material. The


211

disadvantages include that glass is not compatible with all microfabrication techniques. Owing to the laborious machining requiring clean room processes for photolithography and bonding no mass production of glass wafers is available. The high production costs currently prevent the use of glass chips in disposable applications. Polymeric materials (polydimethyl siloxane or PDMS., polymethyl methacrylate or PMMA and polycarbonates) provide a new horizon in labon-a-chip applications. Chip fabrication of these materials is entirely different from the established clean room techniques used for glass and silicon. Once a mold is prepared, mass production of cheap polymeric devices is possible. An additional advantage is that most polymeric materials used are transparent. The excellent insulator properties allow the use of high voltages. Mechanical properties of polymeric materials are usually less advantageous, for example, shape changes may occur as a result of absorption of organics and water. Additionally, Joule heating of the structure can be a problem. Owing to their low heat conductance, structures can undergo deformation from an excessive increase of temperature. Most polymeric materials are not compatible with standard microfabrication techniques. Finally, owing to their hydrophobic characteristics filling polymeric devices is often difficult. Table 3 comprises the suitability of wafer material and micromachining technology combinations for CE. As seen in the table the only universally suited method is laser cutting. Molding and hot embossing are only suited for polymers, whereas etching is used for glass and silicon. Table 7-3. Suitability of wafer materials aand material micromachining technology combinations for CE chips Method Silicon (+/-) +

Wet chemical etching Radiation-induced etching Deep reactive ion + etching (DRIE) + Laser cutting +/Powder blast Injection molding NA Hot embossing NA NA not applicable + suited for CE chips not suited for CE chips +/limited applicability for CE chips

Material Glass (+) + +

Plastic (+) NA +

+

-

+ +/NA NA

+ NA + +

212


Wet chemical etching g is a traditional technology, suited for anisotropic materials such as silicon in which precision engineering features are possible. In glass wet chemical etching usually generates channels wider at the top than at the base [40]. In order to fabricate channels in glass the top plate is coated with a 1 µm amorphous silicon (aSi) film. After conventional photolithography the channel pattern is etched into the aSi film. The aSi film acts as the mask layer for the following glass etching in 10% hydrogen fluoride. Radiation-induced etching g methods produce deep channels with parallel walls. Various radiation sources are available, including X-ray synchrotron radiation [41] and a range of lasers, mostly operating in the UV region [42]. A two-dimensional pattern is placed between the substrate and the radiation source, or direct write laser micromachining such as a UV excimer laser can be used to cut channels into various polymeric materials with excellent geometrical characteristics [43]. Microfluidic cards can be processed from CAD drawing to finished microfluidic design in less than 30 minutes (Micronics, Redmond, WA, USA) through this low cost manufacturing process, based on laser cutting and lamination of plastic sheets. Powder-blasting technology is an alternative approach for producing channels and other structures in a solid insulating material. By bombarding a patterned photoresist foil covered wafer with 30 µm Al2O3 parabolic microstructures can be blasted in glass providing a rough surface [44]. The advantages of this technique include the possibility of using commercially available photoresist foil and the lack of any under-etch of the mask. Although examples exist for powder blasted channels used for electrophoresis purposes [45], the roughness of the channel walls imposes severe disadvantages with respect to separation efficiency, as it introduces additional peak broadening [46]. The technology is especially suited for fabrication of the in- and outlet holes for CE chips [47,48]. Deep reactive ion etching (DRIE) is a promising technique for CE applications. It provides well-defined structures in silicon with high structure density per unit area and because of the availability of a silicon nitride coating providing insulating channel walls. Thin-wall vertical channels with an insulating layer of silicon nitride coating were made in silicon template [49]. After patterning, the silicon was totally sacrificed, leaving hollow, transparent, insulated microchannels (micro-transparent insulated channels, µTIC) [50]. The micromachined structure is shown in Fig. 7-7. The novel fabrication method allows high flexibility in the design of insulated channels with respect to dimensions and cross-sectional shapes [51]. The thin walls are optimally cooled and optically transparent, enabling the use of high electric field strengths as well as optical detection.


213

Figure 7-7. Vertical channels with parallel walls in silicon (mTIC) made with DRIE. Reproduced with permission [48].

Molded polymer chips made of PDMS were used in separations [52] in clinical samples. The channels were molded from a silicon master wafer with a positive surface relief and the cured slab was peeled off. The ease of producing large numbers with the associated low fabrication costs can be considered a major advantage of polymer molding. From a PDMS slab individual devices were fabricated in our laboratory with channels 50 µm wide and approximately 100 µm high. Holes for fluid reservoirs and electrode access were punched through the PDMS using a blunt needle [53]. Hot embossingg of polymeric material can offer the advantage of low cost for embossing tools, simple replication, and high volume mass fabrication. It involves heating a preformed block (the embossing tool) to just above the glass temperature for the polymer, and then pressing the tool into the polymer substrate. Once cooled, the device can be separated from the embossing tool. Hot embossing has been demonstrated for the preparation of microfabricated PMMA chips [54], and for a 96-microchannel electrophoretic separation device [55]. The last fabrication step is mostly the attachment of the cover plate, and the wafer bonding. Glass structures are usually heated up to ca. 680ºC in order to bond the cover plate onto the chip [56], although temperatures as low as 90ºC for 1 hour have also been reported [57]. If bioactive material is immobilized the plate cannot be heated. Sometimes bonding the cover plate is carried out with glue. Special care needs to be taken to prevent the glue from entering and block the channels [58]. PDMS, a flexible material, provides a leak-free reversible seal by pressing the material to the glass substrate [59]. An alternative approach is to oxidize the cover plate and the channel section (both made of PDMS) in oxygen plasma [60]. This method

214


was suited to bond channel networks to other substrates, such as glass, silicon and polystyrene. A possible advantage owing to oxidation of the channel walls is improvement in EOF characteristics [61,62,63].

4.2

Power supplies, pumping, injection and channel geometries

A high voltage power supply is essential for CE chip operation. The power supply needs to have the required number of electrodes and the capacity to independently adjust the potential, and thus the electro-osmotic flow in each channel or section. In Table 4 an overview is given of power supplies suited for lab-on-a-chip devices. If the power supply has multiple electrodes it enables controlled liquid handling in various compartments on a chip. A source, sink, and floating capability of the output voltages ensures reliable liquid handling and the prevention of liquid ‘leakage’ into a compartment. By combination of more instruments the number of electrodes can be extended accordingly. An additional useful feature is separate display of the output voltage and output current of each electrode. Table 7-4. Power supplies for lab-on-a-chip instrumentation Micralyne (Edmonton, Canada) Number of electrodes Switching time Maximum voltage Step size Maximum current Voltage monitoring Current monitoring Noise Labview control

2–8 2–5 ms 6 kV 5V 85 µA yes yes ? yes

IBIS Technologies (Hengelo, Netherlands) 4 15 ms 1 kV 1V 200 µA yes yes 0.3 V yes

Micronit (Eschede, Netherlands) 8 < 2 ms 3 kV ? 1 mA yes yes < 1.5 V ?

Various pumping devices have been developed for microfluidic systems [64] based on inducing pressure differences with valves [65,66] or with valveless elements [67,68]. EOF pumping is the liquid handling method of choice in CE chip devices not primarily because of the flexibility in handling of small volumes but because the power supply is already present in each CE system.


215

An inventive novel pumping and flow controlling device, the flowFET, was introduced in 1999 [69]. The electrical counterpart of the Field Effect Transistor (FET) is based on a method of controlling the ζ-potential of a channel physically. In the flowFET an additional electric field perpendicular to the channel wall is used to adjust the charges at the wall, thus the magnitude and the direction of the electro-osmotic flow inside the microfluidics channel. The flowFET as a controlling and switching element in microfluidics networks is illustrated in Fig. 7-8.

Figure 7-8. Photograph showing two flowFETs integrated with a channel junction to generate opposite flows. A perpendicular electric field of 1.5 megavolts per centimeter was generated at the ‘gate’ area by a voltage as low as 50 V. [Reprinted with permission from 48].

In CE chips, where picoliter–sized plugs are injected into the separation channel, the geometrical definition of the plug is vital for optimal separation efficiency. Samples can be injected using the electro-osmotic flow (electrokinetic injection) or applying pressure (hydrodynamic injection). If the sample plug is injected electrokinetically under the influence of the applied field the separation starts already during injection in the sample plug. This effect can give rise to discriminative injection as well as sharpening of the injected band (focusing). An advantage of electro-osmotic injectors is that they contain no moving parts, guaranteeing a high degree of reliability, and ease of fabrication, since only two electrodes placed at the ends of the channel network are needed. Hydrodynamic injection is non-discriminative and the reproducibility of the injected volume is usually sufficient. However, hydrodynamic injection superimposes a parabolic profile on the flat profile of the electro-osmotic flow. This effect can increase the band broadening reducing the overall separation efficiency.

216


Figure 7-9. Example of T-injector on a chip. The bottom and top circles represent the buffer reservoir, left and/or right circles indicate the sample reservoirs. With permission from Micronit B.V., Enschede, the Netherlands).

A schematic view of a simple cross injector is given in Fig. 7-9. The buffer reservoirs (top and bottom circles) are filled with the separation buffer, whereas the sample reservoirs (left and/or right) contain the sample. The injection sequence starts with applying a potential between the sample reservoirs. The induced electro-osmotic flow fills the horizontal channel with sample solution. Next, potential is applied between the two buffer reservoirs and the electro-osmotic flow introduces an upward motion. The sample band in the intersection of the two channels moves upward sandwiched between the separation buffer plugs, and the separation commences. The definition of the sample plug volume can be improved by applying a potential across the other channels where fluid motion is not required. The applied potential can prevent leakage of solution into these channels. Improvement of the injector was achieved by the introduction of the double-T channel junction [70]. With this system plugs can also be injected sequentially. The experimental setup of an auto-sampler type of injector is depicted in Fig. 7-10. The analysis sequence starts with filling the cuvettes and the capillary with the separation buffer followed by draining both cuvettes. Next, the sample is filled in the cuvette and the sample plug is injected. During the injection, a hydrodynamic injection component is also established owing to the difference of height between sample and sensor cuvette. Then the sample cuvette is drained. Finally, the sample and sensor cuvettes are filled with equal amounts of buffer solution and the separation is started.


217

Drain buffers

2 Fill system

Sample Injection

3 Electrophoresis separation

5

Drain D i sample l

4

Figure 7-10. Experimental set up of an autosampler type of injector. The inlet–outlet holes of the chip connected by a single serpentine channel of 6 cm are automatically filled with buffer (1). After draining the inlet and outlet (2) the sample is exposed to the inlet channel during 5 seconds (3). Then the sample is removed (4) and the inlet hole is washed and filled with buffer (5). Separation takes place at 500 V for 500 seconds, and the cycle starts again. [71]

Studies have shown that serpentine microchannels intended to increase the separation efficiency of electrophoresis on microchips, can decrease efficiency [72,73] owing to extended band broadening. As illustrated in Fig. 7-11A, the sample plug migrates through the inner corner faster, not only because of the shorter distance but also because of the higher electric field at the inner wall. This so-called ‘race-track’ effect can be alleviated using compensating corners [74] or tapered turns [75,76]. When a corner correction is applied (Fig. 7-11B) the sample is forced to travel along the outer curve and the peak broadening is minimized. Although tapering can give rise to Joule heating, experimental results show overall improvement of the separation efficiency [77].

218


Figure 7-11. Simulations showing (A) the ‘race track’ effect in curved microchannels and (B) compensation with tapered turns in CE chips. The diffusion is neglected in the simulation. Adapted from [78].

4.3

Detection strategies

Sensitive detection schemes are essential in microfabricated CE devices owing to the extremely small detection volume. Some electrochemical and optical detectors have been integrated directly in the chip. Most commonly only the detection cell is integrated in the chip device, other elements of the detector are not miniaturized and require external placement. Highest sensitivities have been obtained with optical (laser induced fluorescence) and electrochemical detection. In this section a brief outline is provided of the theory and practice of detection schemes that have been used in combination with CE on a chip. UV absorbance detection is the most commonly used technique in commercial CE systems. This detection method is based on the UV absorption of the compounds analyzed: light passes through the separation channel, the extent of absorption will be dependent on the pathlength (Ȝ) of the detector cell, the concentration (C) as well as the molar extinction coefficient (İ) of the UV absorbing compound. Absorption (A) is described


219

by the Lambert–Beer law, describing the relationship between the intensity of the incident (I0) and emergent light (I): A = Log I0/I = İ Ȝ C Adapted in the early 90s for lab-on-a-chip devices [79,80], UV absorption detection has recently been improved [81,82]. Although this detection method is very simple and quite universal for biomolecules, the reduced path length imposes a proportional loss of sensitivity in miniaturized systems, therefore its application is limited in lab-on-a-chip devices. Thermo-optical detection is a detection scheme based on absorbance which is not hampered by miniaturization. With this technique single molecule [83,84] and single particle detection has been demonstrated [85]. The principle is based on two laser beams which intersect at the detection section of the channel. The wavelength of the light of one laser (pump beam) is matching the absorbance spectrum of the target component. The second laser (probe beam) is monitoring the changes in refractive index in the detection area. When the target component reaches the detection area, increased absorbance of the pump beam causes thermal changes resulting in a change of the refraction index [1,86]. Thermal lens microscopy detection has been demonstrated in combination with CE [87]. Laser-induced fluorescence (LIF) detection has gained in popularity in lab-on-a-chip devices because of its extreme sensitivity [88,89]. The technique is based on excitation of fluorescent target molecules [90,91] or fluorescently labeled molecules [92,93,94] by laser light upon which the excited molecule emits light while returning to the ground state. The emitted light is collected within an angle or in the confocal mode using a dichroic mirror followed by signal amplification in a photomultiplier. A good fluorophore absorbs UV, visible, IR, or near IR radiation with a high efficiency at the exciting wavelength and emits light with a high (i.e., close to 1) fluorescence quantum yield. Photostability of the fluorophore is an important factor expressed by the number of excitation/emission cycles a molecule can undergo before quenching occurs. Quenching agents (e.g., O2, halogens or amines) can deactivate fluorophores. Degassing of solutions helps avoiding poor detection limits. The most prominent advantage of LIF detection is its sensitivity, down to single-molecule [95]. A disadvantage is that non-fluorescent molecules require labeling [96,97]. The integration of optical input and output is an important issue in fully miniaturizing chip based analytical systems [98,99,100,101]. For example, a solid polymer dye laser has been integrated in polymer devices [102], whereas another integrated optical detector contains Rhodamine 6G in ethanol [103].

220


Chemiluminescence (CL) offers a promising method of detection in micro-total analysis systems where the ultimate purpose is to integrate the whole analytical process on a micro-scale device, as CL does not require excitation with a light source. Chemiluminescence is based on a chemical reaction yielding electronically excited species which emit light while returning to the ground state. CL detection combined with separation methods can offer excellent analytical selectivity and sensitivity; however, a chemical reaction yielding a chemiluminescent product needs to precede the detection step. Various CL reagents, such as luminol [104,105], acridinium [106], peroxyoxalate [107,108], and a Ru(II) complex [109], have been utilized. In practice, CE separation conditions impose limitations on the choice, composition, and pH of the running buffer. Chemiluminescence and electrochemiluminescence detection have been demonstrated in micro-flow injection analysis [110,111] and microchip capillary electrophoresis [112]. The ability to detect and measure biospecific interactions in real time has been greatly improved by the appearance of surface plasmon resonance (SPR), a detection technique based on refractive index measurement in the vicinity of a metal surface. SPR provides at present the most versatile and widely used bioaffinity sensing technique [113,114]. The detector design for SPR imaging developed in our laboratory is shown in Figures 7-12 and 13. A thin gold film coated onto the glass structure is exposed to polarized light at constantly changing incident angle. Most of the light will be reflected by the gold surface. At a specific incident angle excess light absorption occurs, owing to an interaction with the free electrons of the gold film. The incident angle where this phenomenon appears is dependent on the optical density (i.e., refraction index) in the solution, immediately at the gold film.

Figure 7-12. Illustration of SPR imaging (iSPR).


221

Immobilization of selective entities such as antibodies or enzymes can be very useful in order to enhance the selectivity of a detector [115] or array [116,117,118]. When the patterned gold surfaces of SPR are coated with selective binding entities, adsorption of the target compounds, e.g. proteins, will be quantitatively detected based on the change in the refraction index. SPR can be applied for analysis of protein–protein, DNA-DNA, cell to surface, and cell to protein interactions at the detector surface. A major advantage of SPR detection is that it does not require labeling or staining, it is (bio)specific and provides sensitivity below 0.1 nm layer thickness. A disadvantage lies in the necessity of regeneration of the sensor surface.

A

B

C

Figure 7-13. Images of a 1 mm wide microfluidics channel with eight gold patches (100 mm×800 mm). The dark gold patches in full resonance (B) turn bright when out of resonance owing to small (A) and large incidence light angle (C). Unpublished results (Biochip Group, MESA+, University of Twente).

Electrochemical detection schemes are mainly based on concentrationdependent signals and therefore generally not hampered by miniaturization. Under special circumstances the performance of electrochemical detectors can even be improved by miniaturization. Electrochemical detection can be tuned to measure general, bulk features of the solution (conductivity and sometimes potentiometry) or selectively aimed at detection of certain compounds (potentiometry, amperometry). Total conductivity (ț ț [S/cm]) is determined by the concentration (Ci), charge number (zi) and individual equivalent conductance (Ȝi) of each ionic species present in the solution: Ȝi κ = 1 / 1000

n

¦C

i

zi λi

i=1

The measured electric conductance G [S] of the solution is dependent on the geometric parameters of the detector cell as follows:

222


G= ț A/l, where A is the surface area and l is the distance between the electrodes. Miniaturization of conductivity detectors can enhance performance if the distance between the electrodes decreases more rapidly than their surface area. Amperometric detection is based on the electron transfer (oxidation or reduction) between a solid electrode and an electroactive compound under an applied DC voltage. The redox reaction gives rise to a current signal directly related to the number of electrons transferred at the electrode surface. The measured current is determined by the conversion yield which is in conventionally sized flow cells usually 15 %, depending on the cell geometry. The current (I) is a function of the diffusion coefficient (D), the electrode surface area (A) the geometrical height of the cell (b) and the flow velocity of the liquid (v): I = const. nFC (D A/b)2/3 v1/3 where n is the number of electrons involved in the redox reaction and F is the Faraday constant. Miniaturization will have a complex effect on the conversion yield, based on changes in the flow, electrode and cell geometry. Because the noise level will also decrease with the decreasing electrode surface area, the signal–to–noise ratio may even improve. Potentiometric detection is based on selective transfer of an ion from solution onto a sensor surface generating a potential difference [1]. The ideal potentiometric detector yields a strong signal for all ions except the ions of the background electrolyte. These response characteristics are defined by the selectivity coefficient (KiJ) as expressed by the Nicolsky–Eisenman equation [119,120] for two identically charged monovalent background ion i and target ion J: E = E0 + RT/zF ln (ai + KiJaJ). In this equation E is the electromotive force of the cell assembly: reference electrode/sample solution/ionselective membrane/(3 M KCl, AgCl/Ag electrode), E0 contains the junction potential and a constant, R is the universal gas constant, T the absolute temperature and z is the charge of the ions (here z = 1). The response characteristics of potentiometric measurements are based on the concentration difference and are therefore not affected by miniaturizing. A well known chip based potentiometric sensor, the ion sensitive field effect transistor (ISFET), was introduced by Bergveld [121] in the 1970s.


223

These semiconductor based electrodes have an ion-sensitive surface: the surface electrical potential is changed (relative to an external reference electrode) by interaction of the semiconductor with ions. For example, a material that bears hydroxyl groups at the surface, such as silicon oxide, can be protonated or deprotonated, thus making that surface a pH (H+ ion) sensor. According to the Nernst equation one should observe a 59 mV shift in the potential per ten-fold change in ion concentration. ISFETs are made for the selective detection of ions by covering the sensor surface with a polymer layer designed to be selectively permeable to the target ion. In these so called CHEMFETs, schematically shown in Fig. 7-14, the target ions diffuse into and across the polymer layer and change the semiconductor’s surface potential [122].

Figure 7-14. Schematic representation of a CHEMFET.

Attempts were made to sense the charge redistribution of proteins directly as a result of immunochemical reactions at ISFETs. It was found that when the ISFET shows a Nernstian response and the double layer of the gate area overlaps with the charge distribution of a protein an exact compensation takes place [123]. A direct potentiometric static detection of proteins would only be possible when proteins replace the surface charges of highly electrically insulating membranes. Charge detection of amphoteric molecules is, in principle, possible in this case, otherwise charge

224


compensation may ruin the effect. However, to our knowledge this effect has not been demonstrated in the scientific community yet. Besides the intrinsic and not fully solved drift effects of ISFETs and in relation to the weak charge distribution shift that is expected as a result of the protein interaction at the gate surface, this method can hardly compete with more advanced direct sensing techniques such as surface plasmon resonance. A dynamic potentiometric approach was demonstrated for the successful detection of protein charge distribution at ISFETs [124]. This so called ‘ion step’ method allowed detection of charge distribution of proteins adsorbed on the gate area of modified ISFETs. The surface of the ISFET was covered with a monolayer of amino beads (Fig. 7-15) followed by immobilization of protein ligands onto the beads. The bead technology was successfully applied for stimulus response or dynamic potentiometric protein charge sensing using the ion step method (Fig. 7-16) [125]. Integrated lab-on-a-chip technology with automatic injection, transport, and regeneration is expected to improve system performance significantly in ion step measurements.

Figure 7-15. A monolayer of beads (diameter: 0.9 µm) covering the surface of the ISFET, the lighter section is the gate area.

Experimental challenges faced in all forms of electrochemical detection are posed by the high electric field used for the CE separation, because the output signal has to be separated from the influence of the DC power supply. Several amperometric [126,127], potentiometric [128], and conductivity detectors [129,130] have been described for chip based devices. An integrated electrochemical detector was presented by Mathies et al. [131]. Of all detection methods in CE the combination with mass spectrometry (MS) has great promise. The popularity of MS in separations is based on its potential for providing information on the molecular weight and structure of the analytes enabling their identification. The technical difficulties posed by on-line coupling of the CE separation system and the detector, as well as the available low detectable amounts in miniaturized systems, were overcome in


225

40

1

potential (mV)

30 20

2

10 0

3

-10

4

-20

0.00

0.25

0.50

0.75

1.00

tim e (s) Figure 7-16. Ion step responses of amino bead covered ISFETs before (curve 1) and after incubation with poly L-Lysine (curve 4) and washing with Tween® 20 solution (curve 2: non-activated, and curve 3: glutardialdehyde pre-activated). Ion stepping was performed at pH 7.42.

1997 by different groups [132,133,134]. The most popular systems use electrospray ionization interfaces between the separation capillary outlet and the atmospheric pressure inlet of the MS [135]. Matrix assisted laser desorption ionization (MALDI) MS carried out off-line in the solid phase is another relevant MS detection method. The separated fractions of the sample are deposited as spots on a solid substrate target and are detected off-line. Recent examples are provided for multidimensional separation and off-line MALDI analysis for proteome profiling [136] as well as for CE separation with enzymatic digestion and MALDI-TOF spectrometry [137]. The first microchip atmospheric pressure chemical ionization (APCI) MS compatible with microfluidics devices was reported by Östman et al. [138].

5.

SELECTED EXAMPLES FOR CE ON A CHIP

The automated procedure for sample injection shown in Fig. 7-10 enabled the optimization of injection volumes, and a high sample throughput. A µTIC design with an injection channel and a separation channel containing a 100 µm long injection loop was realized. Conductivity detection using an electrically floating lock–in amplifier was integrated on the chip which improved the signal–to–noise ratio and detection limits. The chip format allowed fast and parallel analysis, important for high throughput screening. The system was tested with two channel fabrication methods. Reproducible separation (both in time and peak height) of a mixture of 500 µΜ Μ potassium,

226


sodium, and lithium was obtained with smooth DRIE etched channels (Fig. 7-17). Separations carried out in powder–blasted channels were of inferior quality, probably as a result of the high surface roughness of the channels [70].

Figure 7-17. Separation of three cations in a micro-CE system. Sample: 500 µM K+, Na+, Li+ (injection 5 s, 500 V). Separation conditions: 6 cm channel, 50×27 µm Tris/MES buffer (25 mM., pH 7.4), 500 V, 0.6 µA, conductivity detection (7 kHz, lock-in amplifier) from [70].

The first direct measurement of lithium in whole blood was described using microchip capillary electrophoresis (CE). In serum a detection limit for lithium of 0.4 mmol/l was reached. In addition potassium and sodium were also detected in the same run. The electrokinetic transport of red blood cells inside the channels was studied to find optimal sample loading conditions for the analysis of lithium without injecting cells into the separation channel. Blood collected from a finger stick was mixed with anticoagulant and transferred onto the chip without extraction or removal of components [139]. Another prominent example of small molecule determination in plastic devices is provided by the CE based bioassay of small metabolites [140]. The authors demonstrate simultaneous bioassay of glucose, uric and ascorbic acid, and acetaminophen. Improvements in the separation were achieved by a microchip with on-line sample stacking and preconcentration using reverse migrating micelles and a water plug [141]. DNA based microsystems and protein chip arrays [142] are covered elsewhere in this book. Therefore, we provide only a few historical examples


227

and commercially available systems here. Pioneering work was carried out in DNA sequencing by establishing a microfabricated 384–lane capillary array for ultra high throughput genetic analysis [143,144]. Concentration and sequence specific separation of single-base substituted mutant DNA has been demonstrated [96]. A microchip has been developed for continuous online analysis of PCR products [145]. Recent examples of on-chip peptide and protein separation include a label-free native fluorescence detection scheme [91]. A complete system has been developed for rapid determination of bacillus spores using a thermolyser with automated sample preparation and analysis [146]. A novel device is suited for the evaluation of cell surface charge based on its electrophoretic mobility in an electrophoresis chip [147]. The first fully commercialized successful implementation of microfluidics is the Agilent 2100 bioanalyser which offers electrophoretic DNA, RNA and protein analysis (Fig. 7-18) next to the most recently introduced cell assays. Similarly, a Caliper chip is used in an instrument called Experion launched by BioRAD. A similar product, the Hitachi Cosmo, is equipped with disposable plastic chips for DNA analysis. Shimadzu has also developed a microchip electrophoresis system (MCE-2010) with UV detection for DNA analysis.

6.

DIELECTROPHORESIS

Electrophoresis is the movement of particles caused by a constant DC electric field. The movement of particles caused by a non-uniform, alternating field (AC) is called dielectrophoresis [149]. An alternating field induces dipoles in particles such as biological cells, large molecules, or bacteria. The dielectrophoretic forces are proportional to the size of the particle and allow their manipulation in fluids. The method is primarily suited for large particles, usually at the cellular scale (10 µm) [150]. Also smaller particles, including viruses [151,152], beads of >10 nm, and DNA fragments [153] have been successfully manipulated. At field strengths of several hundred kV/m dielectrophoretic forces allow the trapping of particles without loss. Arrays of planar electrodes positioned in parallel, quadrupole, or octupole configurations as well as three-dimensional field cages were found suited to position or trap cells efficiently. Dielectrophoresis appears to be an excellent tool for sorting particles [154,155] and separating [156] or trapping cells [157,158] on the basis of their electrical properties. Examples have been published for the separation of bacteria from water [159] red blood cells, viable and non-viable yeast cells [160], and normal and cancerous cells [161]. Fig. 7-19 shows an example of a quadrupole field trap for the collection of viruses [162].

228


Figure 7-18. Microfluidics channel system of Caliper Life Sciences for protein analysis. Reprinted with permission [148].

One of the first implementations of dielectrophoretic phenomena in microfluidics devices was the alignment and deformation of DNA strands [163]. Dielectrophoretic forces are often used in microfluidic devices for the manipulation of particles [164,165,166]. Other AC electrokinetic manipulation techniques also based on polarization in an inhomogeneous field are suited to levitate, capture or concentrate particles (see Fig. 7-20).

Figure 7-19. Field ‘funnel‘ created by a planar quadrupole electric field in which particles (viruses) are trapped by negative dielectrophoretic forces [161].


229

Figure 7-20. Schematics of concentration (a–c) and separation (d–g) steps in dielectrophoresis: large circles represent 50.1 µm beads, small circles represent yeast cells, dotted circles represent the DEP cages. Reprinted with permission from [167].

230

7.


OUTLOOK

The body of literature on lab-on-a-chip is expanding at a rapid pace, yet the field is far from mature. CE chip technology has clearly been the most advanced area of all lab-on-a-chip methods described in the literature. One of the underlying reasons could be that this area was the most easy to adapt to microscale. Another reason might be that miniaturization in CE directly exploits the benefits of the smaller scale. It can be anticipated that the trend of explosive growth will even accelerate, owing to the versatility of CE as a separation technique. The number of applications for the analysis of complex real life samples is relatively limited. With the introduction of commercial products, e.g., for point of care testing, it can be anticipated that real life applications will also grow in number. Broadening the options for sample pretreatment on a chip would certainly also contribute to the increase of the field of applications. A severe limitation of miniaturization is imposed by the amount of sample that can be introduced without overloading the system. With inventive sample loading, continuous sample injection can be achieved, potentially improving the limits of detection by orders of magnitude. At its introduction the lab-on-a-chip principle may have seemed to be a utopian dream. The authors hope to have convinced the reader that our dreams in this field can hardly catch up with the fast unfolding reality: lab-on-a-chip technology offers the key to reveal the secrets of biological systems.

8. [1] [2] [3] [4] [5] [6]

REFERENCES Landers, J.P., (1997), ‘Handbook of Capillary Electrophoresis’, 2nd Edition, CRC Press, New York. Altria, K.D., (1995), ‘Capillary Electrophoresis Guidebook, Principles, Operation, and Applications’, Humana Press, New Jersey. Kricka, L.J., (2001), Clinica Chimica Acta, 307, 219–223. Leung, Y.F., Pang, C.P., (2001), Trends in Biotechnology, 19, 430. Vo-Dinh, T., Cullum, B.M., Stokes, D.L., (2001), Sensors and Actuator, B74, 2–11. Manz, A., Graber, N., Widmer, H.M., (1990), Sensors and Actuators, B1, 244–248.

SEPARATION AND DETECTION ON A CHIP [7] [8] [9] [10] [11] [12] [13] [14]

[15] [16]

[17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28]

231

Ehrfeld, W., Hessel, V., Löwe, H., (2000), ‘Microreactors: New technology for modern chemistry’, Wiley-VCH, Weinheim. Jacobson, S.C., Hergenröder, R., Koutny, L.B., Ramsey, J.M., (1994), Anal. Chem., 66, 2369–2373. TüdĘs, A., Besselink, G.A.J., Schasfoort, f R.B.M., (2001), ‘Review: Trends in miniaturized total analysis systems for point–of–care testing in clinical chemistry’, Lab-on-a-chip, 1, 83–95. Gardeniers, J.G.E., Schasfoort, R.B.M., van den Berg, A., (2000), Technologies and devices for micro-chemical systems, BIOD., 10–13. Moore, W.J., (1972), ‘Physical Chemistry’, (Longman, London). Reinhoud, N.J., (1995), ‘Bioanalytical Capillary Electrophoresis, Thesis’, Leiden, The Netherlands. Schasfoort, R.B.M., Bergveld, P., Kooyman, R.P.H., Greve, J., (1990), ‘Possibilities and limitations of direct detection of protein charges by means of an ImmunoFET’, Analytica Chimica Acta, 238, 323–329. van Theemsche, A., Deconinck, J., van den Bossche, B., Bortels, L., (2002), ‘Numerical Solution of a Multi-Ion One-Potential Model for Electro-osmotic Flow in Two-Dimensional Rectangular Microchannels’, Analytical Chemistry, 74, 4919–4926. Davies, J.T. , Rideal, E.K. Interfacial Phenomena’, (1963), Academic Press Inc., London. Bergveld, P., A. Sibbald, (1988), ‘Analytical and biomedical applications of ion-selective field-effectt transistors’, Elsevier, Amsterdam. Jorgenson, J.W., Lukacs, K.D., J., (1981), High Resol. Chromatogr., 4, 230. Wu, C.T., Lopes, T., Lee, B.-C.S., (1992), Anal. Chem., 64, 886. Hayes, M.A., Ewing, A.G., (1992), Anal. Chem., 64, 512. Lee, C.S., Blanchard, W.C., Wu, C.T., (1990), Anal. Chem., 62, 1550. Hayes, M.A., Khetarpal, I., Ewing, A.G., (1993), Anal. Chem., 65, 27. Hayes, M.A., Khetarpal, I., Ewing, A.G., (1993), Anal. Chem., 65, 2010. Wätzig, H., Kaupp, S., Graf, M., (2003), ‘Inner surface properties of capillaries for electrophoresis’, Tr. Anal. Chem., 22, 588–604. Baryla, N.E., Lucy, C.A., (2002), ‘Semi-permanent surfactant coatings for inorganic anion analysis in capillary electrophoresis’, J. Chromatogr., 956, 271–277. Landers, J.P., (ed.), (1997), Handbook of capillary electrophoresis, 2nd Edition, CRC Press, Boca Raton. Knox, J.H., (1988), Chromatographia, 26, 329. Davis, J.M., (1990), J. Chromatogr., 517, 521–547. Gobie, W.A., Ivory, C.F., (1990), J.Chromatogr., 516, 191–210.

232


[29] Gaš, B., (1993), J. Chromatogr., 644, 161–174. [30] Hjerten, S., (1990), Electrophoresis, 11, 665–690. [31] Tudos, A.J., (1993), ‘Selectivity enhancement and miniaturization of electrochemical detection in high performance separation techniques’, Thesis, University of Amsterdam, 25–26. [32] Petersen, N.J., Nikolajsen, R.P., Mogensen, K.B., Kutter, J.P., (2004), ‘Effect of Joule heating on efficiency and performance for microchip based and capillary based electrophoretic separation systems: a closer look’, Electrophoresis, 25, 253–69. [33] Gaš, B., Kenndler, E., (2002), ‘Peak broadening in microchip electrophoresis: A discussion of the theoretical background’, Electrophoresis, 23, 3817–3826. [34] L. Bousse, B. Dubrow, K. Ulfelder, (1998), ‘High performance DNA separations in microchip electrophoresis systems’, in Harrison, D.J., van den Berg, A. (ed.), Micro-Total Analysis Systems ’98, Kluwer Academic Publishers, Dordrecht. [35] Jacobson, S.C., Hergenröder, R., Koutny, L.B., R.J. Warmack, Ramsey, J.M., (1994), ‘Effects of injection schemes and column geometry on the performance of microchip electrophoresis devices’, Anal. Chem., 66, 1107–1113. [36] Sinton, D., Ren, L., Li, D., (2003), ‘A dynamic loading method for controlling on-chip microfluidics sample injection’, J. Coll. Interface Sci., 266, 448-456. [37] Pohl, H.A., (1978), Dielectrophoresis, Cambridge University Press, Cambridge. [38] Gardeniers, J.G.E., Oosterbroek, R.E., van den Berg, A., (2003), ‘Silicon and glass micromachining for µTAS’, in Oosterbroek, R.E., van den Berg, A. (ed.), Lab-on-a-chip Miniaturized systems for (Bio)chemical analysis and synthesis, Elsevier, Amsterdam, 37–64. [39] Becker, H., Gärtner, C., (2003), ‘Microreplication technologies for polymer based µTAS applications’, in Oosterbroek, R.E., van den Berg, A., (ed.), Lab-on-a-chip Miniaturized systems for (Bio) chemical analysis and synthesis, Elsevier, Amsterdam, 21–35. [40] Fisher, J.C., Jacobson, S.C., Davis, L.M., Ramsey, J.M., (1998), Anal. Chem., 70, 431. [41] Tolfree, D.W.L., (1998), Reports on Progress in Physics, 61, 313. [42] www.micronics.net. [43] Roberts, M.A., Rossier, J.S., Bercier, P., Girault, H., (1997), Anal. Chem., 69, 2035–2042. [44] Wensink, H., Elwenspoek, M.C., (2002), ‘Reduction of sidewall inclination and blast lag of powder blasted channels’, Sensors and Actuators, A102, 157–164.


233

[45] Schlautmann, S., Wensink, H., Schasfoort, R.B.M., Elwenspoek, M.C., van den Berg, A., (2000), ‘Powder-Blasting Technology as a Powerful Tool for Microfabrication of CE chips with Integrated Conductivity Sensors’, 11th Workshop on Micromachining, Micromechanics and Microsystems (MME00), Uppsala, Sweden. [46] Solignac, D., A. Sayah, S. Constantin, R. Freitag, M.A.Gijs, M., (2001), Sens. Actuators A 92, 388–393. [47] Wensink, H., Berenschot, J.W., Jansen, H.V., Elwenspoek, M.C., (2000), ‘High Resolution Powder Blast Micromachining’, 13th Int. Workshop on Micro-Electro Mechanical Systems (MEMS 2000), Miyazaki, Japan, 769–774. [48] Schlautmann, S., Wensink, H., Schasfoort, R.B.M., Elwenspoek, M.C., van den Berg, A., (2001), J. Micromech. Microeng., 11, 386. [49] R.B.M.Schasfoort, Schlautmann, S., Hendrikse, J., van den Berg, A., (1999), Science, 286, 942–945. [50] Tjerkstra, R.W., Gardeniers, J.G.E., Elwenspoek, M.C., van den Berg, A., (1997), Electrochimica Acta, 42(20–22), 3399. [51] Fintschenko, Y., Fowler, P., Spiering, V., Burger, G.J., van den Berg, A., (1998), ‘Characterization of Silicon based Insulated Channels for Capillary Electrophoresis’, in Harrison, D.J., van den Berg, A. (ed.), Micro-Total Analysis Systems ‘98, Kluwer Academic Publishers, Dordrecht,) 327–331. [52] Effenhauser, C.S., Bruin, G.J.M., Paulus, A., Ehrat M., (1997), 69 3451. [53] Beusink, B., Besselink, G.A.J., Schasfoort, R.B.M., unpublished results. [54] Lee, G.-W., Chen, S.-H., Huang, G.-R., Sung, W.-C., Lin, Y.-H., (2001), Sensors and Actuators, B75, 142–148. [55] Gerlach, A., Knebel, G., Guber, A.E., Heckele, M., Hermann, D., Muslija, A., Schaller, T., (2002), Microsystem Technologies, 7, 265. [56] Greenway, G.M., Haswell, S.J., Petsul, P.H., (1999) Analytica Chimica Acta, 387, 1. [57] Wang, H.Y., Foote, R.S., Jacobson, S.C., Schneibel, J.H., Ramsey, J.M., (1997), Sensors and Actuators, B 45, 199–207. [58] Schlautmann, S., Besselink, G.A.J., Radhakrishna Prabhu, G., Schasfoort, R.B.M., (2003), ‘Fabrication of a microfluidics chip by UV bonding at room temperature for integration of temperature sensitive layers’, J. Micromech. Microeng, 13, S81–S84. [59] Duffy, D., et al., Rapid prototyping of microfluidics switches in poly(dimethylsiloxane) and their actuation by electro-osmotic flow. J. Micromech. Microeng. 9, (1999) 211–217.

234


[60] Duffy, D.C., McDonald, J.C., Schueller, O.J.A., Whitesides, G.M., (1998), Ana. Chem., 70, 4974. [61] Erill, I., Villa, R., Goudignon, P., Fonseca, L., Plaza, J.A., (2000), Microelectronics Reliability, 40, 787–789. [62] Liu, Y., Fanguy, J.C., Bledsoe, J.M., Henry, C.S., (2000), Anal. Chem., 72, 5939–5944. [63] Ren, X., Bachman, M., Sims, C., Li, G.P., Allbritton, N., (2001), J. Chromatogr. B762, 117–125. [64] van der Wijngaart, W., Andersson, H., Stemme, G., (2003), ‘Handling of beads in microfluidics devices for biotech applications’, in Oosterbroek, R.E., van den Berg, A. (ed.), Lab-on-a-chip Miniaturized systems for (Bio)chemical analysis and synthesis, Elsevier, Amsterdam. [65] Veenstra, T.T., Berenschot, J.W., Gardeniers, J.G.E., Sanders, R.G.P., Elwenspoek, M.C., van den Berg, A., (2001), J. Electrochem. Soc., 184, G68–G72. [66] Gravesen, P., Branebjerg, J., Sondergard Jensen, O., (1993), ‘Microfluidics-a review’, J. Micromech. and Microeng., 3, 168–182. [67] Stemme, E., Stemme, G., A valveless diffuser/nozzle based fluid pump, Sensors and Actuators A39 (1993) 159–167. [68] Olsson, A., Enoksson, P., Stemme, G., Stemme, E., (1997) ‘Micromachined flat-walled valveless diffuser pumps’, J. Microelectromech Syst., 6, 161–166. [69] Schasfoort, R.B.M., Hendrikse, J., van den Berg, A., (1999), ‘µTransparent insulating channels as components for miniaturized chemical separation Devices’, Proceedings of Transducers ’99, Sendai, Japan, 872 –875. [70] Shultz-Lockyear, L.L., Colyer, C.L., Roy, K.I., Harrison, D.J., (1998), ‘Sample effects on injection and sample loading in integrated capillary electrophoresis devices’, in Harrison, D.J., van den Berg, A. (ed.), Micro-Total Analysis Systems ’98, Kluwer Academic Publishers, Dordrecht, 169–172. [71] Srichaiyo, T., Hjertén, S., (1992), J. Chromatogr., 604, 85–89. [72] Griffiths, S.K., Nilson, R.H., (2000), Anal. Chem., 72, 5473–5482. [73] Molho, J.I., Herr, A.E., Mosier, B.P., Santiago, J.G., Kenny, T.W., Brennen, R.A., Gordon, G.B., Mohammadi, B., (2001), ‘Designing Corner Compensation Electrophoresis in Compact Geometries’, Anal. Chem., 73, 1350–1360. [74] Paegel, B.M., Hutt, L.D., Simpson, P.C., Mathies, R.A., (2000), Anal. Chem, 72, 3030–3037. [75] Griffiths, S.K., Nilson, R.H., (2001), Anal. Chem., 73, 272–278.


235

[76] Molho, J.I., Herr, A.E., Mosier, B.P., Santiago, J. G., Kenny, T.W., Brennen, R.A., Gordon, G.B., (2000), ‘Designing Corner Compensation Electrophoresis in Compact Geometries’, in van den Berg, A., Olthuis, W., Bergveld, P., (eds.), Micro-Total Analysis Systems 2000, Kluwer Publishing, Amsterdam, 287–290. [77] Baidya, B., Mukherjee, T., Hoburg, J.F., (2002), ‘Dispersion Modeling in Microfluidics Channels for System-level Optimization’, Proceedings of the Fifth International Conference on Modeling and Simulation of Microsystems (MSM 2002), San Juan, Puerto Rico, 182–5. [78] Harrison, D.J., Fan, Z.H., Seiler, K., Manz, A., Widmer, H.M., (1993), Anal. Chim. Acta, 283, 361–366. [79] Burggraf, N., Manz, A., Effenhauser, C.S., Verpoorte, E., de Rooij, N.F., Widmer, H.M., (1993), J. High Resol. Chromatogr., 16, 594– 596. [80] Mogensen, K.B., F. Eriksson, Nikolaysen, R.P.H., Gustafsson, O., Kutter, J.P., (2004), ‘Single-mask step integration of high-aspect ratio sub-micrometer channels, pure silica waveguides and fiber couplers for separation chips’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 1, 39–41. [81] Hulme, J.P., Miyahara, Y., (2004), ‘Parallel absorbance detection using hot embossed elevated optical elements’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 2, 467–469. [82] Tokeshi, M., Uchida, M., Uchiyama, K., Sawada, T., Kitamori, T., (1999), J. Luminescence, 261, 83–84. [83] Tokeshi, M., Uchida, M., Hibara, A., Sawada, T., Kitamori, T., (2001), Anal. Chem., 73, 2112–2116. [84] Mawatari, K., Kitamori, T., Sawada, T., (1998), Anal. Chem., 70, 5037–5041. [85] Kikutani, Y., Hibara, A., Hisamoto, H., Tokeshi, M., Kitamori, T., (2003), ‘Micro-integrated systems for general use’, in Oosterbroek, R.E., van den Berg, A. (eds.), Lab-on-a-chip Miniaturized systems for (Bio)chemical analysis and synthesis, Elsevier, Amsterdam. [86] Kitagawa, F., Sueyoshi, K., Mizuno, J., Wada, Y., Shoji, S., Otsuka, K., (2004), ‘Application of thermal lens microscopy and on-line sample preconcentration for high sensitive detection in microchip electrophoresis’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J.,

236

[87] [88] [89]

[90]

[91]

[92] [93] [94]

[95]

[96]

SEPARATION AND DETECTION ON A CHIP I Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 1, 67–69. Mathies, R.A., Huang, X.C., (1992), Nature, 359, 167–169. Dolnik, V., Liu, S., Jovanovich, S., (2000), ‘Capillary electrophoresis on microchip’, Electrophoresis, 21, 41–54. Qin, J., Fung, Y., Zhu, D., Lin, B., (2004), ‘Native fluorescence detection of flavin derivatives by microchip capillary electrophoresis with LIF intensified charge-coupled device detection’, J. Chromatogr., 1027, 223–229. Misir, K., Harrison, D.J., (2004), ‘Development of a laser induced native fluorescence detection scheme for peptides and proteins on microchip’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 2, 449–451. Jiang, G., Attiya, S., Ocvirk, G., Lee, W.E., Harrison, D.J., (2000), ‘Red diode induced fluorescence detection with a confocal microscope on a microchip for capillary electrophoresis’, Biosens. Bioelectronics, 14, 861–869. Webster, J.R., Burns, M.A., Burke, D.T., Mastrangelo, C.H., (2001), Anal. Chem, 73, 1622–1626. Lin, C.H., Fu, L.M., Chen, S.H., (2005), ‘Integrated optical-fiber capillary electrophoresis microchips with novel spin-on-glass surface modification’, Biosens. Bioelectronics, in press. Jahn, A., Vreeland, W.N., Gatain, M., Locascio, L.E., (2004), ‘Multichannel continuous flow microfluidic system for single molecule vials’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 1, 345–347. Ito, T., Inoue, A., Sato, K., Hosokawa, K., Maeda, M., (2004), ‘Single-step concentration and sequence-specific separation of DNA by affinity microchip electrophoresis’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 1, 159–161. Petersen, D., Foote, R.S., Geschke, O., Ramsey, J.M., (2004), ‘An integrated micro-device for on-chip preconcentration, separation and labeling of proteins’, in Laurell, T., Nilsson, J., Jensen, K., Harrison,


237

D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 1, 117–119. [97] Hübner, J., Zauner, D., Anhoj, T.A., Jorgensen, A.M., (2004), ‘Multimethode integrated optical components for µTAS-A rigorous approach’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 1, 357–359. [98] Hoffmann, O., Miller, P., de Mello, J.C., Bradley, D.D.C., de Mello, A.J., (2004), ‘Integrated optical detection for microfluidic systems using thinfilm polymer light emitting diodes and organic photodiodes’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 2, 506–508. [99] Balslev, S., Roxhed, N., Griss, P., Stemme, G., Kristensen, A., (2004), ‘Microfluidics dye laser with compact, low-cost liquid dye dispenser’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 2, 375–377. [100] Kim, J.-H., Shin, K.-S., Paek, K.-K., Kim, Y.-H., Kim, Y.-M., Kim, Y.-K., Kim, T.-S., Kang, J.-Y., Yang, E.-G., Kim, S.-S., Ju, B.-K., (2004), ‘An advanced microchip with organic light emitting diode integrated on a microchannel for applications in the fluorescence detection’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 2, 428–430. [101] Nillson, D., Balslev, S., Kristensen, A., (2004), ‘Solid polymer dye laser based on a single mode SU-8 planar waveguide’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 2, 369–371.

238


[102] Balslev, S., Bilenberg, B., Geschke, O., Jorgensen, A.M., Kristensen, A., Kutter, J.P., Morgensen, K.B., Snakenborg, D., (2004), ‘µTAS with integrated optical transducers’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 2, 348–350. [103] Tsukagoshi, K., Fujimura, S. and Nakajima, R., (1997), Anal. Sci., 13, 279. [104] Daddo, R., Colon, L.A. and Zare, R.N., (1994), Anal. Chem., 66, 303. [105] Ruberto, M.A. and Grayski, M.L., (1992), Anal. Chem., 64, 2758. [106] Tsukagoshi, K., Tanaka, A., Nakajima, R. and Hara, T., (1996), Anal. Sci., 12, 525. [107] Wu, N. and Huie, C.W., (1993), J. Chromatogr., 634, 309. [108] Tsukagoshi, K., Okuzono, N. and Nakajima, R., (2002). J. Chromatogr., A958, 283. [109] Nozaki, O. and Kawamoto, H., (2000), Luminescence, 15, 137–142. [110] Michel, P.E., Fiaccabrino, G.C., de Rooij, N.F., Koudelka-Hep, M., (1999), Anal. Chim. Acta, 392, 95–103. [111] Hashimoto, M., Tsukagoshi, K., Nakajima, R., Kondo, K., and Arai, A., (2000), J. Chromatogr., A867, 271–279. [112] Whelan, R.J., Zare, R.N., (2003), ‘Surface plasmon resonance detection for capillary electrophoresis separations’, Anal. Chem., 75, 1542–1547. [113] Besselink, G.A.J., Kooyman, R.P.H., van Os, P.J.H.J., Engbers, G.H.M., Schasfoort, R.B.M., (2004), ‘Signal amplification on planar and gel-type sensor surfaces in surface plasmon resonance based detection of prostate-specific antigen’, Analytical Biochemistry. [114] Xiong, L., Regnier, F.E., (2001), ‘Channel-specific coatings on microfabricated chips’, J. Chromatogr., A924, 165–176. [115] Blazej, R.G., Paegel, B.M., Emrich, C.A., Mathies, (2003), ‘R.A., Microfabricated capillary array electrophoresis: implementation and applications’, in Oosterbroek, R.E., van den Berg, A. (ed.), Lab-on-achip Miniaturized systems for (Bio)chemical analysis and synthesis, Elsevier, Amsterdam. [116] Kricka, L.J., (2001), ‘Microchips, micro-arrays, biochips and nanochips: personal laboratories for the 21st century’, Clin. Chim. Acta, 307, 219–223. [117] Wegener, G.J., Wark, A.W., Lee, H., J., Codner, E., Saeki, T., Fang, S., Corn, R.M., (2004), ‘Real-time, surface plasmon resonance imaging measurements for the multiplexed determination of protein adsorption/desorption kinetics and surface enzymatic reactions on peptide micro-arrays’, Anal. Chem, 76, 5677–5684. [118] Eisenman, G., (1967), ‘Glass electrodes for hydrogen and other cations’, Marcel Dekker, New York, 1967.


239

[119] Durst, R.A., (1969), Ion selective electrodes, Special Publications No. 314, National Bureau of Standards, Washington. [120] Bergveld, P., (1970), ‘IEEE Development of an ion sensitive solid state device for neurophysiological measurements’, Trans. Biomed. Eng. BME-17, 70. [121] van den Berg, A., Bergveld, P., Reinhoudt, D.N., Sudholter, E.J.R., (1985), ‘Sensitivity control of ISFETs by chemical surface modification’, Sens. Actuators, 8, 129. [122] Schasfoort, R.B.M., P.Bergveld, Kooyman, R.P.H., Greve, J., (1990), ‘Possibilities and limitations of direct detection of protein charges by means of an ImmunoFET’, Analytica Chimica Acta, 238, 323–329 [123] Schasfoort, R.B.M., Kooyman, R.P.H., Bergveld, P., and Greve, J., (1990), ‘A new approach to ImmunoFET operation’, Biosensors and Bioelectronics, 5(2), 103–125. [124] Besselink, G.A.J., Schasfoort, R.B.M., & Bergveld, P., (2003), ‘Modification of ISFETs with a monolayer of latex beads for specific detection of proteins’, Biosensors and Bioelectronics, 18, 109–1114. [125] Wu, C.C., Wu, R.G., Huang, J.G., Lin, Y.C., Bang, H.C., (2003), ‘Three-electrode electrochemical detector and platinum film decoupler integrated with a capillary electrophoresis microchip for amperometric detection’, Anal. Chem., 75, 947–952. [126] Emrich, C.A., Ertl, P., Singhal, P., Mathies, R.A., (2004), ‘Sheathflow support for high-sensitivity end-column electrochemical detection in micro-devices’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 2, 527–529. [127] L.J. Nagels, Poels, I., (2000), ‘Solid state potentiometric detection systems for LC., CE., and µTAS methods’, Tr. in Anal. Chem., 19, 410–417. [128] Reay, R.J., and Kovacs, G.T.A., (1994), ‘Microfabricated EC detector for CE’, Proceedings of the Solid State S&A Workshop, Hilton Head, South Carolina, 61–64. [129] Schasfoort, R.B.M., Guit-van Duin, R., Schlautmann, S., Frank, H., Billiet, H., van Dedem, G., van den Berg, A., (2000), ‘Miniaturized capillary electrophoresis system with integrated conductivity detector’, in van den Berg, A., Olthuis, W., Bergveld, P., (eds.) Micro-Total Analysis Systems 2000, Kluwer Publishing, Amsterdam, 391–394. [130] Wolley, A.T., Lao, K., Glazer, A.N., Mathies, R.A., (1998), ‘Capillary Electrophoresis Chips with Integrated Electrochemical Detection’, Anal. Chem., 70(4), 684–688. [131] Xue, Q.F., Foret, F., Dunayevskiy, Y.M., Zavracky, P.M., Mcgruer, N.E., Karger, B.L., (1997), Anal. Chem., 69, 426–430.

240


[132] Figeys, D., Ning, Y.B., Aebersolc, R., (1997), Anal. Chem., 69, 3153–3160. [133] Ramsey, R.S., Ramsey, J.M., (1997), Anal. Chem., 69, 1174–1178. [134] Kameoka, J., Craighead, H.G., H. Zhang, Henion, J., (2001), Anal. Chem., 73, 1935–1941. [135] Buch, J.S., Li, Y., Wang, Y.-X., Cooper, J.W., DeVoe, D.L., Lee, C.S., (2004), ‘Multi-dimensional microfluidics based comprehensive proteome profiling’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 1, 384–386. [136] Brunet, E., Dodge, A., Chen, S., Goulpeau, J., Labas, V., Royer, N., Vinh, J., Tabeling, P., (2004), ‘Coupling u of separation and digestion of proteins in a PDMS device for mass spectrometry analysis’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 2, 545–547. [137] Östman, P., Marttila, S., Koiaho, T., Franssila, S., Kostiainen, R., (2004), ‘Microchip atmospheric pressure chemical ionisation-mass spectrometry’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 2, 437–439. [138] Vrouwe, E.X., Luttge, R., van den Berg, A., (2004), ‘Direct measurement of lithium in whole blood using microchip capillary electrophoresis with integrated conductivity detection’, Electrophoresis, 25,10–11 , 1660–1667. [139] Wang, J., Charathi, M.P., Tian, B., Polsky, R., (2000), ‘Microfabricated electrophoresis chips for simultaneous bio-assays of glucose, uric acid, ascorbic acid and acetaminophen’, Anal. Chem 72 (2000) 2514–2518. [140] Sueyoshi, K., Nagai, H., Wakida, S., Nishii, J., Kitagawa, F., Otsuka, K., (2004), ‘Designing multi-functional microchips for electrophoretic analysis‘, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 1, 84–86. [141] Schasfoort, R .B.M., (2004), ‘ Proteomics-on-a-chip: the challenge to couple lab-on-a-chip unit operations’, Expert Rev. Proteomics, 1(1), 123–132. [142] Emrich, C.A., Tian, H., Medintz, I.L., Mathies, R.A., (2002), Anal. Chem., 74 5076–5083. [143] C.N. Liu, Toriello, N.M., Maboudian, R., Mathies, R.A., (2004), ‘High throughput polymerase chain reaction-capillary array electrophoresis (PCR-CAE) microchip, in Laurell, T., Nilsson, J., Jensen,


[144]

[145]

[146]

[147] [148] [149] [150] [151] [152]

[153]

[154]

241

K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 2, 297–299. Kim, H., Suk, S., N. Park, Hahn, J.H., (2004), ‘Microchip for continuous on-line PCR product analysis’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 1, 273–275. Hukar, K.W., Patel, K., Renzi, R.F., West, J.A.A., (2004), in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), MicroTotal Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 2, 321–323. Omasu, F., Nakano, Y., Ichiki, T., (2004), ‘Evaluation of cell electrophoretic mobility using microcapillary electrophoresis chips’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 1, 64–66. Agilent technologies, Lab-on-a-chip Technology Application Literature (Publication Number 5988- 3035EN), (2001), reprinted from Pharmaceutical Technology Europe, PDF Reprint No. 0468. Crane, J.S., Pohl, H.A., (1968), ‘A study of living and dead yeast cells using dielectrophoresis’, J. Electrochem. Soc.: Electrochem. Sci. 115, 584–586. Wang, X.B., Cheng, J., (2001), ‘Electronic Manipulation of Cells on Microchip based Devices’, in Cheng, J., Kricka, L.J. (ed.) Biochip technology, Harwood Academic Publishers, Amsterdam, 135–159. Hughes, M.P., Morgan, H., (1998), J. Phys. D.: Appl. Phys, 31, 2205–2210. Hughes, M.P., Morgan, H., Rixon, F.J., Burt, J.P.H., Pethig, R., (1998), Biochim. Biophys. Acta, 1425, 119–126. Miles, R., Belgrader, P., Bettencourt, K., Hamilton, J., Nasarabadi, S., (1999), ‘Dielectrophoretic Manipulation of Particles for Use in Microfluidic Devices’, Microelectromechanical Systems, ASME Publication, Nashville, Tennessee, 1, 497–501, Nov. 14–19. Homes, D., Sandison, M.E., Green, N.G., Morgan, H., (2004), ‘Highspeed particle sorting combining dielectrophoresis and fluid flow’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 1, 6–8. Kostner, S., Nieuwenhuis, J.H., Svasek, E., Svasek, P., Jachimowicz, A., Vellekoop, M.J., 2004), Continuous particle separator based on periodical DEP elements’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 1, 9–11.

242


[155] Chen, T.-I., Liu, C.-H., (2004), ‘A dielectrophoretic cell separaton microchip with size filtering’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 1, 108–110. [156] Frénéa, M., Chiral, M., Le Pioufle, B., Melaine, N., Pineau, C., Jégou, B., (2004), ‘On-chip spermatozoa trapping by dielectrophoresis’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 1, 111–113. [157] Rosenthal, A., Voldman, J., (2004), ‘Simple, strong, size-selective dielectrophoretic traps for single-cell patterning’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 2, 228–230. [158] Simmons, B., Lapizco-Encinas, B., Shediac, R., Hachman, J., Chames, J., Brazzle, J., Ceremuga, J., Fiechtner, G., Cummings, E., Flintschenko, Y., (2004), ‘Polymeric insulator based (electrodeless) dielectrophoresis (iDEP) for the monitoring of water-borne pathogens’, in Laurell, T., Nilsson, J., Jensen, K., Harrison, D.J., Kutter, J.P. (eds.), Micro-Total Analysis Systems 2004, Proceedings of µTAS 2004, Malmö, Sweden, Royal Society of Chemistry, Cambrigde, 2, 171–173. [159] Wang, X.B., Huang, Y., Hölzel, R., Burt, J.P.H., and Pethig, R., (1993), ‘Theoretical and experimental investigations of the interdependence of the dielectric dielectrophoretic and electrorotational behaviour of colloidal particles’, J. Phys. D: Appl. Phys. 26, 312– 322. [160] Gascoyne, P.R.C., Pethig, R., Burt, J.P.H., Becker, F.F., (1993), ‘Membrane changes accompanying the induced differentiation of Friend murine erythroleukemia cells studies by dielectrophoresis’, Biochim. Biophys. Acta 1149, 119–126. [161] Schnelle, T., Müller, T., Fiedler, S., Shirley, S.G., Ludwig, K., Herrmann, A.M., Fuhr, G., Wagner, B., Zimmermann, U., (2004), ‘Trapping of viruses in high-frequencey electric field cages’, Naturwissenschaften 83, 172–176. [162] M. Washizu, Kurosawa, O., (1990), IEEE Trans. Ind. Appl., 26, 1165–1172. [163] Huang, Y., Pethig, R., (1991), Meas. Sci. Techn. 2,1142–1146. [164] Pethig, R., Huang, Y., X-B.Wang, Burt, J.P.H., (1992), J. Phys. D: Appl. Phys, 25, 881–888.


243

[165] Gascoyne, P.R.C., Huang, Y., Pethig, R., Vykoukal, J., Becker, F.F., (1992), Meas. Sci Techn., 3, 439–445. [166] Medoro, G., Manaresi, N., Altomare, L., Leonardi, A., Tartagni, M., Guerrieri, R., (2002), ‘A Lab-on-a-chip for Cell Separation Based on the Moving-Cages Approach’, Proceedings of Eurosensors XVI, 15–18 Sept. 2002, Prague, Czech Republic.

Chapter 8 PROTEIN MICROARRAYS: TECHNOLOGIES AND APPLICATIONS Dieter Stoll, Markus F. Templin, Jutta Bachmann§, Thomas O. Joos* D. Stoll, M.F. Templin, T.O. Joos: NMI Natural and Medical Sciences Institute at the University of Tuebingen, Markwiesenstr. 55, 72770 Reutlingen, Germany § Nøkkefaret 12, M [email protected] 1450 Nesoddtangen, Norway, Phone: +47-66912809, E-Mail: * corresponding author: E-Mail: [email protected]

Abstract:

Protein microarray technology has been successfully applied for the identification, quantification, and functional analysis of proteins in basic and applied proteome research. It can be shown that these miniaturized and parallelized assay systems have the potential of replacing the singleplex analysis systems. However, robustness and automation need to be demonstrated before this technology can be used reliably for high throughput and routine applications. In this review we summarize the current stage of protein microarray technology. Recent applications used for the simultaneous determination of a variety of parameters from a minute amount of sample will be described and future challenges of this cutting–edge technology will be discussed.

Key words:

protein microarray, parallelized assay, Proteomics

1.

INTRODUCTION

Although the basic principles of protein microarray technology had already been described in the early 1980s by Roger Ekins’ambient analyte theory [1], the interest in microarray based assays has increased enormously owing to the growing demand of genomics and Proteomics which focus on the analysis of gene and protein function in a global perspective. The possibility of determining thousands of parameters in one single experiment in a parallel fashion provided the perfect solution for these kinds of analyses.


245

246


In the meantime, DNA microarrays, some of them built from tens of thousands of different oligonucleotide probes per square centimeter, are well established high throughput hybridization systems which are able to generate huge sets of mRNA expression data within a single experiment [2]. However, there is no absolute correlation between mRNA level and corresponding protein level [3]. Additional high throughput technologies deciphering direct protein profiling are needed for overcoming the gap between genomics and Proteomics. Currently microarray technology has been expanded beyond DNA chips. Within the last few years examples of protein–array based approaches have been published. These protein microarray based assays can be grouped according to different formats and different types of applications. These will also be discussed in greater detail. Forward

Reverse

Fractionate Label

...

2

1

1

B

D

2

...

C

Figure 8-1. Model of microarray based forward and reverse protein capture assays. Forward assays involve specific protein capture molecules immobilized on distinct micro-spots. The entire array is incubated with a protein mixture t (A) which can be labeled (B). Analyte proteins are bound to their specific capture agents and can be detected either using specific secondary antibodies in a sandwich format (A) or by labeling them directly (B). Reverse arrays rely on the sample itself. Fractions of a certain sample (C) or different samples (D) are immobilized on a large set of arrays. These arrays can be probed using different specific antibodies that bind to their immobilized antigen targets. Each array is probed with only one antibody. Reproduced with permission from The Thomson Corporation and Joos, T.O., Stoll, D., Templin, M.F., Bachmann, J., ‘Protein microarrays: Applications and future challenges’, (2005), Curr. Opin. Drug. Discovery. Dev., 8(2), 239-252. Copyright 2005, The Thomson Corporation.


247

Currently, forward phase protein microarrays (Fig. 8-1) are the most frequently used microarray assay formats. They can be used for the simultaneous analysis of different parameters from distinct samples which are incubated on arrays of numerous microspots, each containing one single, well defined immobilized capture molecule. Examples of forward phase protein microarrays include antibody microarrays that are used to identify and quantify target proteins of interest and protein affinity assays that are used to study the interactions between proteins and immobilized binding molecules such as proteins, peptides, low molecular weight compounds, oligosaccharides, or DNA (Fig. 8-2) [4,5].

Figure 8-2. Different types of forward protein interaction and protein capture microarrays and of reverse microarrays: Specific interaction microarrays have been described for receptor– ligand (A); enzyme–substrate (B); protein–protein (C); sugar–protein (D); and protein–DNA (E) interactions. Specific protein capture microarrays used for the identification and quantification of proteins are based on highly specific protein capture molecules such as affibodies, aptamers, or antibodies (F) and can be used as direct capture (G) or sandwich assay (H). Reverse assays are based on immobilized cell lysates (I,L) , cells (J) or tissue slices (K,M). The markers were detected with specific capture molecules (J,K) or MALDI-TOF mass spectrometry (L,M). Adapted from Stoll et al., (2004), ‘Microarray technology: An increasing variety of screening tools for proteomic research’, DRUG DISCOVERY TODAY Y, 3, No 1, 2004, 24–31, with permission from Elsevier.

248


Reverse phase arrays (Fig. 8-1) are one of the most recent developments of protein microarrays. They can be used to determine a distinct set of parameters in a large collection of tissue or cell samples [6,7], or sample fractions that are immobilized in an array format on a solid support. Single soluble probes such as antibodies are used to screen these tiny spots simultaneously for the presence or absence of distinct target proteins [8–10]. The design and generation of specific capture microarrays depends on the target analytes and the availability of the corresponding highly specific capture agents. These prerequisites can be met easily for all types of capture assays involving DNA chips However, proteins need a completely different treatment. DNA is a rather uniform molecule and binds its complementary DNA targets according to the well defined principle of base–pairing. Based on the primary sequence of the target DNA, highly selective and specific DNA capture sequences can easily be predicted. In addition, high throughput oligonucleotide synthesis facilities or PCR based approaches enable the fast and inexpensive generation of DNA capture agents. Therefore the generation of DNA arrays and the design of simple assays is a very straightforward approach. In contrast to DNA, proteins are not as easy to handle. It is impossible to predict high affinity capture molecules for proteins from their primary amino acid sequence alone. This is owed to very diverse tertiary structures of proteins and the manifold possibilities of interacting with each other. Interaction depends on strong electrostatic forces, hydrogen bonds, or weak hydrophobic Van der Waals interactions, but most often all in combination. In addition, proteins can also interact with different binding partners simultaneously and often appear in complexes which makes their handling even more difficult. Steady or dynamic post-translational modifications such as glycosylation or phosphorylation have also an enormous additional influence on protein interactions. Each protein capture molecule must be generated individually after large sets of candidate capture molecules have been screened against their individual target proteins. No PCR equivalents are available for proteins, either. Therefore the development of methods for the cost effective, fast high throughput generation of highly specific, high affinity protein capture molecules and protein targets is the first obstacle on the cumbersome path to establishing protein microarrays for proteomic research. Not least difficult is the necessity of retaining their functionality. Proteins must be immobilized without damaging their tertiary structure. This is, of course, much more difficult than attaching oligonucleotides or PCR fragments to a solid support. However, the first obstacles have been overcome: microarray surfaces can be optimized to accept functional proteins, protein labeling technologies have been improved, and our


249

knowledge of additives for stabilizing proteins has considerably increased. These achievements have already led to successful applications of protein microarray technology in a large variety of assay systems In the following sections we will discuss the different types of protein microarray assays and their applications in protein profiling, interaction studies, and reverse screening (Fig. 8-2).

2.

FORWARD PHASE PROTEIN MICROARRAYS

2.1

Protein Expression Analysis Using Protein Microarrays

Similarly to the dual color labeling approach used for the visualization of differential mRNA expression, first studies have been undertaken to analyze the differential display of proteins. Antibody–antigen microarray assays were performed with two different protein samples which were labeled with two different fluorophores. Equal amounts of total protein were mixed and incubated on the antibody microarray (Fig. 8-3) [11]. The differences in the concentrations of the target proteins in each capture spot were detected via dual wavelength fluorescence. Results gained with this kind of assay must be handled carefully since proteins are often assembled in multi-protein complexes. Therefore a strong signal can either result from a large amount of target protein or from the capture of a huge complex of different proteins bound to the captured target (Fig. 8-3). Nevertheless, field applications proved the general applicability of this technology. Sreekumar et al., demonstrated the use of antibody microarrays in the detection of altered protein levels of LoVo Colon carcinoma cells after treatment with ionizing radiation [12]. Knecevic et al., were able to show cancer specific alterations in protein levels in specific cellular populations from squamous cell carcinoma of the oral cavity [13]. Both research groups identified more than ten proteins which altered their expression levels in response to the ionizing radiation treatment or in correlation to tumor progression. In the meantime the first commercial antibody microarrays consisting of several hundred monoclonal antibodies have entered the market (Table 1).

250


Sample A (control)

+

Sample B (sample)

+

Figure 8-3. Principle of differential capture protein microarray assays: Proteins from two different samples are isolated, labeled with two different colors (fluorophores), mixed and incubated on the same protein capture microarray. The dual color detection (green /red) of bound proteins reveals the differences in protein expression between control and sample directly. Complications might arise if not only single proteins but whole protein complexes are captured on the micro-spot since the detected signal does not represent the true concentration of the analyte in the sample. Adapted with permission from The Thomson Corporation and Joos, T.O., Stoll, D., Templin, M.F., Bachmann, J., (2005), ‘Protein microarrays: Applications and future challenges’, Curr. Opin. Drug. Discovery. Dev., 8(2), 239-252. Copyright 2005, The Thomson Corporation.


251

Table 8-1. Examples of commercially available tools, kits, and services within the field of microarray technology. Planar arrays are offered for comparative protein profiling, reverse screening, cytokine profiling, substrate–enzyme assays, and protein interaction studies. Bead based systems are used for the focused protein expression profiling of cytokines and cell signaling molecules. Currently the majority of commercially available bead based protein microarrays are based on the Luminex™100 platform. Planar Protein Arrays Company BD Biosciences Hypromatrix, Inc.

Grace Bio-Labs Inc Grace Bio-Labs Inc

Molecular Staging Inc. Pepscan Systems Pierce Biotechnology Inc. Protometrix, Inc.

Product BD Clontech™Ab Microarray Signal Transduction AntibodyArrayTM

ProPlate™ Multi-array Slide System PepStar™, ProteaseSpots™, PepStar™, PhosphoSiteDetector™ Rolling circle amplification technology (RCAT™) PepChip® Kinase SearchLightTM Arrays

Applications Comparative protein analysis Protein-protein Interactions and Protein Phosphorylations Integration of microplate and biochip peptide arrays for enzyme profiling and mapping of protein– protein interaction.

Web link www.bdbiosciences.com

Multiplexed protein profiling

www.molecularstaging.com

kinase substrate assays Cytokine profiling

www.pepscan.com www.searchlightonline.com

Services protein interaction studies Cytokine and protein profiling

www.protometrix.com

www.hypromatrix.com/

www.gracebio.com www.jerini.com

RayBiotech Inc.

The Yeast ProtoArrayTM RayBio TM Cytokine Arrays, Custom Ab Arrays

Schleicher & Schuell Bioscience

FAST® Frame Multi- Throughput Processing www.schleicherof Protein Microarrays schuell.de/bioscience Slide Plate,

SIGMA-ALDRICH Co. Zeptosens AG

Zyomyx Inc.

ProvisionTM HCA Panorama™ Ab Microarray Cell Signalling Kit™ ZeptoMARKTM CeLyA Cell Lysate Arrays Zyomyx Protein Profiling Biochip System

www.raybiotech.com

Cytokine profiling Comparative protein analysis

www.sigmaaldrich.com

Reverse Screening

www.zeptosens.com

Cytokine profiling

www.zyomyx.com

252 PROTEIN MICROARRAYS: TECHNOLOGIES AND APPLICATIONS

Planar Protein Arrays Company Bead Based Systems Luminex Corporation Bender Medsystems Bio-Rad Laboratories GmbH BioSource International INOVA Diagnostics Inc. LINCO Research, Inc. Qiagen Radix BioSolutions R&D Systems Rules Based Medicine Inc. Upstate Biotechnology Cytometric Bead Array (CBA)

Product

Applications

Web link

Luminex 100™

xMAP® Fluorescent bead immuno assays (FBIs) Bio-Plex™ Array

www.luminexcorp.com

Multiplexed antibody bead kits QUANTA Plex™

www.biosource.com

LINCOplex™ cytokine arrays Liquichip Customized assay products for research Fluorokine® MAP arrays Species–specific multianalyte profiles Beadlyte® protein profiling, kinase assays Protein profiling

www.bendermedsystems.com

www.bio-rad.com

www.inovadx.com www.lincoresearch.com www.qiagen.com www.radixbiosolutions.com R&D Systems www.rulesbasedmedicine.com

www.beadlyte.com www.bdbiosciences.com

Belov et al., [14] developed antibody microarrays for the analysis of the composition of the cluster of differentiation (CD) antigen level on different leukemia cells. Microarrays composed of up to 90 different anti-CD antibodies were incubated with a cell suspension to allow the binding of cells which expressed the corresponding CD antigen. Different patterns of cell binding were obtained for normal peripheral blood leukocytes and different types of leukemia. These types of cell capture antibody microarrays do not require fluorescent tags or other sophisticated detection systems, and might therefore be an interesting alternative to flow cytometry assays in immunophenotyping or chip based mRNA expression profiling of leukemia. In addition, intact cells which are captured on antibody microspots can be further characterized applying different soluble, fluorescencelabeled antibodies. Recently, Soen et al., demonstrated the use of microarrays containing peptide–MHC complexes for the identification and characterization of multiple antigen specific T cell populations. Cytotoxic (CD8+ CTL) or helper (CD4+) T cells were captured onto such arrays according to their ligand specificity. These authors demonstrated a method that was suitable for the detection of a rare population of antigen specific T


253

cells after vaccination. This approach might prove very useful for identifying the antigenic epitopes during the immune response after bacterial or viral infection and for analyzing different types of T cell populations binding to different epitopes [15]. There is a growing demand in research and clinical applications for profiling the expression of proteins assays which can possibly be fulfilled with the development of highly sensitive miniaturized sandwich immuno assays [16-27]. One of the very first microarray based analytical assays was developed by Ekins and colleagues [1,17] Using a microspot system the authors succeeded in quantifying analytes such as thyroid stimulating hormone (TSH) or Hepatitis–B surface antigen (HbsAG) accurately quantified (even down to the femtomolar concentration range). Shao et al., developed the most complex multiplexed sandwich immuno assay so far. This assay allowed the quantification of 150 different cytokines in parallel [18]. The rolling–circle signal amplification (RCA) technology was used for the highly sensitive detection of bound analytes. More than half of the detected proteins had detection sensitivities in the pg/ml range.In principle, it can be assumed that the RCA detection principle is suitable for the detection of single binding events on microspots [34]. Miniaturized and multiplexed immuno assays can be regarded as an excellent solution for applications in which several parameters of a single sample must be analyzed in parallel. For large scale Proteomics research it can further be assumed that high content protein microarrays will accelerate the identification of new biomarkers. In contrast to such high content multiplexed analyses, diagnostic assays usually focus on a clearly defined, small set of analytes. For example, in immune or allergy diagnostics it is necessary to screen patient sera for the presence or absence of a relatively small number of different types of auto–antibodies or antibodies directed against parasitic and viral antigens or allergens [35–40]. Recently, Knight et al., demonstrated the versatility of multiplexed cytokine assays for the profiling of 16 cytokines in human blood [19]. The authors compared the results of already established, well accepted ELISA tests with their microarray assay: Virtually identical cytokine concentrations could be measured in endotoxin stimulated human blood with either method. The complex network of the cell signaling system defines the need for parallel quantification of different proteins at different phosphorylation states. Nielsen et al., [20] have described an Ab microarray integrated within 96-well microtiter plates for the quantification of the amount and modification state of ErbB receptors in crude cell lysates. The assay showed significant differences in ErbB expression among different cell lines and different responses to EGF stimulation. In principle, this assay could be

254


scaled up to 10–20 elements and the systematic analysis of interactions in signaling networks in a rapid and parallel fashion would certainly be within reach. Apart from planar microarrays, bead based systems provide an alternative when the number of parameters to be determined in parallel is rather low. Bead based assay systems employ different color- or size-coded microspheres (beads) as solid support for the immobilization of capture molecules. Different bead types can be identified in a flow cytometer, at the same time the amount of captured target protein is quantified with adequate reporter systems. Sensitivity, reliability, and accuracy are similar to those observed with well established ELISA procedures. A set of a hundred different color coded beads is commercially available to perform miniaturized multiplexed ligand binding assays (xMAP system, www. luminex. com, Table 1). Such systems have been used to determine the concentration of cytokines or antibodies in biological samples such as patient serum or cell culture supernatant [21–27,41,42]. Owing to the increasing number of commercially available kits and the uncomplicated use, bead based systems have a great potential to become an established tool in multiplexed assays within the near future (Table 1). Opalka and colleagues used a bead based assay system to quantify neutralizing antibodies directed against the papilloma virus. Using a competitive immuno assay these researchers were able to measure the titers of polyclonal antibodies in serum directed against conformational sensitive, neutralizing epitopes on the respective virus like particles accurately. The competitive bead based immuno assay proved to be as sensitive, accurate, and precise as a competitive radio immuno assay [43]. McBride, et al., [26] have developed a bead based multiplexed sandwich immuno assay to detect a broad range of pathogens including virus, protein toxins, and bacterial spores. The assay performance of the multiplexed assay revealed excellent specificity and resulted in a dynamic range and sensitivity similar to that observed for the singleplex enzyme linked immuno sorbent assay. The same group also described the integration of the bead based system into a fully autonomous pathogen detection system that is capable of monitoring the environment for airborne biological threat agents continuously. The system performs aerosol collection, multiplexed immuno assays, sample archiving, data reporting, and alarming. The complete system was successfully evaluated in the detection of virulent biological threat agents such as Bacillus antracis and Yersinia pestis [27]. Based on the bead based Luminex technology several companies offer a steadily growing list of ready–to–use multiplexed sandwich immuno assays to quantify cytokines and cell signaling molecules and to analyze kinase activities (Table 1).


255

Rules Based Medicine Inc. (www.rulesbasedmedicine.com) provides a special service for screening mouse or human serum samples for the presence or absence of approximately 90 mouse serum parameters, and approximately 170 human serum parameters from minute amounts of sample volumes. Using the bead based technology, thousands of samples can be screened within a very short time. These examples show that bead based systems are perfectly suitable for the high throughput analysis of smallest amounts of sample. Recently, the BD™ FACSArray Bioanalyzer (www.bdbiosciences.com) was launched as an alternative bead based system (Cytometric Bead Array, BD™ CBA). This bead based system can discriminate between different bead sizes and rely on a two color detection. Therefore it allows the design of more complex assays. Nevertheless, before multiplexed systems will be able to enter the highly competitive diagnostic market, sound validations of these miniaturized assay systems need be performed. Currently immuno assays are highly automated in clinical diagnostics. Any new assay format intended for use in diagnostic applications has to compete with the current highly robust technology—not only with regard to its performance but also with regard to its low costs. Microarrays are not competitive if only a few parameters have to be analyzed from a sample. It will be much more economical to increase the throughput of the currently highly robust and automated diagnostic analyzers. However, with a continuously growing number of diagnostic parameters, as are available for allergy diagnostics, for example, and with the increasing knowledge in system biology and cellular networks, multiplexing analyses will become a very demanding issue. Planar microarrays or bead based systems offer promising solutions.

2.2

Protein Interaction Microarrays

Protein microarray based interaction analysis has been described for the analysis of protein–protein, enzyme–substrate, protein–DNA, protein– oligosaccharide, and protein–drug interactions (Fig. 8-2). Low and high density protein arrays have been used to investigate the binding of DNA, RNA, small chemical ligands, and proteins [44–51]. Enzyme–substrate assays have been performed for restriction enzymes, phosphatases, peroxidases, and phosphokinases. Such microarrays have the potential of providing functional data on a global scale and are required for the functional analysis of complex protein networks within biological systems. So far the first and only global protein interaction studies have been performed using a yeast proteome chip involving recombinant protein probes of 5,800 open reading frames [52]. To test for protein–protein

256


interactions the yeast proteome was probed with biotinylated calmodulin. Many known CamKinases and calcineurins were identified. In addition, 33 new potential binding partners of calmodulin with a potential binding motif were found. In addition, high density protein microarrays have also been used to screen antibodies for their specificity and cross-reactivity which is a prerequisite for the generation of highly multiplex antibody arrays [53]. Lueking et al., generated a glass chip based high density protein microarray from 2,413 non-redundant purified human fusion proteins on a polymer surface at a spot density of 1,600 proteins/cm2 [54]. These protein microarrays have been use to characterize antibody binding, specificity, and cross-reactivity; and to profile the antibody repertoire in body fluids, such as serum from patients with autoimmune diseases. In another approach a protein domain microarray was prepared containing GST–fusion proteins with peptide specific binding motifs such as SH3, SH2 (Src homology 2 and 3) or PDZ domains (a domain originally identified in PSD-95, DLG and ZO-1 proteins). These domains serve as recognition modules for the assembly of multi-protein complexes and might be able to contribute to our understanding of complex protein–protein interactions in cell signaling pathways. The mapping involving short peptides showed that the arrayed protein domains retained their specific binding integrity. In addition, immobilized protein domains were able to ‘fish’ specific interaction partners from the complex protein mixture of a total cell lysate [55]. Miniaturized multiplexed assay systems are perfectly suited to measuring kinase activity and specificity within a single experiment. These types of enzyme–substrate arrays have been described by different groups using recombinant proteins or peptides immobilized onto a solid support [46,49,51] (Table 1). Each substrate is only phosphorylated by its specific kinase. The phosphorylation detection methods employed either radioisotopes (gamma(32/33)P)ATP labeling) or phosphoamino acid-selective antibodies. On microarrays, protein kinase activity can be measured using a novel fluorescent phosphorylation sensor dye. The Pro-Q Diamond phosphoprotein dye technology, involving a small fluorescent probe, has been proved suitable for the fluorescent detection of phosphoserine-, phosphothreonine-, and phosphotyrosine-containing proteins. Steinberg and colleagues [56] demonstrated that this dye allows a very sensitive detection and quantification of phosphorylated amino acid resiowings in peptides and proteins after protein kinase reactions on peptide and protein microarrays. The small fluorescent Pro-Q Diamond probe is a universal sensor of the phosphorylation status and can be used to replace phosphoamino acid-selective antibodies. Carbohydrate microarrays are another emerging application. Carbohydrates are the key components of glycoproteins, glycolipids, and proteo-


257

glycans. They are involved in recognition processes such as cell adhesion, migration, and signaling. Protein–carbohydrate interactions are essential for many biological processes including normal tissue growth and repair, cell– cell adhesion and inflammation, cell growth, fertilization, viral replications, parasitic infection, as well as tumor cell motility and progression. Alterations of glycosylation events are also involved in a number of diseases. Proof of concept experiments using carbohydrate binding proteins of the immune system (antibodies, selectins, a cytokine, and a chemokine) and several plant lectins demonstrated that carbohydrate microarrays can help analyze the complex interactions between the glycome and proteome. In addition, carbohydrate arrays can elucidate host–pathogen interactions and might be an interesting approach to determine different kinds of infections [57-59]. Microarrays have the potential to be integrated into drug screening processes. The identification of potential drugs relies on the immobilization of small organic compounds and the screening for receptor ligand interactions. Although first results have shown that this technology enables parallel high throughput screening for ligand receptor interactions and only requires very small quantities of small molecule ligands and target proteins, this approach has not yet been established in a high throughput fashion [60,61]. An alternative approach was developed was developed by Gosalia and Diamond [62]. Chemical libraries were printed on microarrays for fluid phase nano-liter reactions. The researchers immobilized the chemical compounds of interest within individual nano-liter droplets of glycerol in a microarray format onto glass slides and added reacting reagents by aerosol deposition. This approach allows the rapid addition of multi-component reactions without cross-contamination or the need of separating walls. The feasibility of this approach was demonstrated using kinetic profiling of protease mixtures and protease–substrate interactions. An inhibitor of caspases 2, 4, and 6 was identified by using a 352-compound combinatorial library microarrayed and screened against caspases 2, 4, and 6, as well as thrombin and chymotrypsin. This technology might certainly be of interest to improve procedures aimed at screening for active substances employed by the pharmaceutical industry.

3.

REVERSE MICROARRAYS

An alternative microarray set up is the reverse screening approach, a method in which tiny amounts of a tissue or cell sample are immobilized in a microarray format onto a solid support [7,63]. Tissue or cell lysates are prepared from differentially treated cultured cells or micro dissected tissues (e.g., procured by LCM–Laser capture microdissection), followed by the

258


generation of microarrays from the crude protein mixtures or fractions of them. The immobilized sample proteins thus represent the whole repertoire of cellular proteins at a distinct state. These microarrays can be screened for the presence or absence of distinct target proteins using different specific antibodies. This set up allows the screening of a large collection of tissue or cell lysates with a large numbers of antibodies or patient sera. Only very low amounts of sample are required [7,64–67]. Recently Nishizuka [68] used the reverse phase approach to analyze protein expression within 60 different human cancer cell lines (NCI-60). Each cell line was printed in 10 two-fold serial dilutions. Fifty–two different mouse monoclonal antibodies were used in their study. Signals were clustered and patterns of protein expression were obtained. The researchers were able to identify two promising pathological markers that distinguish colon from ovarian adenocarcinomas. A detailed comparison of protein expression with mRNA expression derived from DNA microarrays revealed a high correlation between mRNA and protein levels of cell structure related proteins across all NCI-60 cell lines. In general, non-cell structure related proteins exhibit no correlation. A drawback of the reverse phase approach is the low concentration of target molecules of interest. High abundance proteins in a protein mixture can be easily detected, whereas rare protein species are only limited in a microspot. As Western blot technology shows, the separation of proteins prior to detection by specific antibodies greatly enhances detection sensitivity, even for low abundant proteins, the pre-fractionation of crude protein extracts prior to the generation of reverse microarrays is one way of increasing sensitivity and screen for rare biomarkers. The group of Samir Hanash [10,69] has shown the benefits of pre-fractionation of cell lysates by chromatography prior to arraying. Solubilized proteins from the LoVo colon adenocarcinoma cell line were separated into 1760 fractions. The same was done for the prostate cancer cell line LNCaP. Subsequently these fractions were arrayed onto nitrocellulose–coated slides and auto–antibodies which are present in the serum of patients suffering from cancer, detected. A distinct pattern of reactivity was observed with sera derived from patients suffering from colon cancer relative to patients diagnosed with lung cancer. This approach allows the characterization of the auto–antigen by further analysis of the fraction of interest using standard mass spectrometric approaches. The authors identified ubiquitin C-terminal hydrolase isozyme 3 (UCH-L3) as the auto–antigen and validated the occurrence of auto– antibodies to UCH-L3 using Western blots. Such a reverse microarray set up allows the rapid screening of large number of patient samples with the subsequent identification of the antigenic proteins.


259

One of the major advantages of reverse microarray screening is that the sample proteins do not need to be labeled. The reverse screening approach hence provides a new way of identifying biomarkers and could be useful in the early profiling of drug candidates with regard to their efficacy and toxicity. Zeptosens AG (www.zeptosens.com) offers a commercialized reverse screening platform (CeLyA Cell Lysate Arrays) which is able to detect a defined set of proteins or protein modifications of interest using multiplexed, direct affinity assays. Compared to classical Western blotting approaches, the Zeptosens system enables a much higher throughput. Ciphergen Biosystems Inc. (www.ciphergen.com) has established another reverse screening technology. The SELDI (Surface Enhanced Laser Desorption Ionization) approach (Fig. 8-3E) is based on mass spectrometry as the readout system [70,71]. Cell extracts derived from different sources are incubated on different spots of the same adsorptive surface chemistry (e.g., pseudo-specific chromatographic surfaces like cation/anion exchange material or hydrophobic surfaces, immobilized metal affinity surfaces, and biospecific interactive surfaces). Unbound or weakly bound proteins can be washed away, whereas the whole variety of non-specifically captured target proteins can be analyzed by mass spectrometry. The mass spectrum mirrors the different molecular weights of the captured proteins. The comparison of two MS data sets generated from two different samples immediately identifies the differentially expressed proteins by their molecular mass. In some cases the differentially displayed proteins might be identified immediately on the basis of their molecular weights. However, in most cases it is necessary to identify such markers separately. Proteins are enriched by affinity chromatography. This can be achieved easily using the same adsorptive material as used for the SELDI experiment. The enriched proteins can then be identified by standard methods (e.g., Edman sequencing, Western blot, digest mass fingerprinting). The SELDI technology is easy to handle and suitable for the fast detection of differences in total protein content of different samples. As the detector sensitivity of time of flight mass analyzers decreases with increasing molecular weights, SELDI is perfectly suited for the detection of small proteins and peptides but exhibits limitations with respect to high molecular weight proteins or membrane proteins. Although assay sensitivity in SELDI experiments is much lower in comparison to sandwich immuno assays [72], the SELDI approach is still a quick screening platform for any unknown protein biomarker [73]. Another interesting reverse phase approach was recently published by Wang (2004). An antibody microarray based immuno staining method was developed which allows the simultaneous detection of a large number of different proteins sin one immobilized cell type. A ‘dissociable’ antibody microarray was generated and brought into close contact with the cells

260


which were to be analyzed for the presence of the antigens. These cells were attached to a second support onto which the ‘dissociable’ antibody microarray was placed. The antibodies dissociated from the array support and bound to the cellular antigens if present. Bound antibodies were visualized by standard immunohistochemistry techniques. Such ‘dissociable’ antibody arrays can be used to analyze the expression and the subcellular localization of different proteins in various cultured cell types. Currently, tissue microarrays represent the most prominent reverse screening approach (Fig. 8-3G). This technology was first described by Juha Kononen and Olli Kalloiniemi. These two researchers generated a microarray of tissue samples which contained hundreds of tissue specimens [6]. The microarrays were then screened for the presence or absence of DNA of RNA molecules or proteins such as p53 or the tumor markers Her-2 and EGFR. Standard analytical methods such as immuno histochemistry, fluorescence in situ hybridization (FISH), or other molecular detection methods were used for detection [6]. Tissue microarrays have a considerable advantage over conventional histological approaches: a large number of specimens can be treated simultaneously in an identical manner. The traditional histological analysis of tissue specimens is a rather slow and labor intensive process. In contrast, the parallel processing of a large number of histological samples increases the throughput dramatically. Therefore, tissue microarrays are the method of choice for screening a large number of samples for well defined parameters. However, one disadvantage of tissue microarray analysis remains. It is difficult to say whether biopsy material represents the whole specimen since it only represents a tiny fraction of the whole tumor.

4.

OUTLOOK

The full potential of protein microarray can only be envisaged now, although the principles of protein microarray technology were already described and established several years ago. Protein microarrays are powerful tools which can be used for the identification and quantification of proteins. In addition they have also proved useful for the analysis of interactions between proteins with other proteins, peptides, low molecular weight compounds, oligosaccharides or DNA (Fig. 8-2). The bottleneck associated with protein profiling is the limited availability of highly-specific and selective capture molecules. In the future the bottleneck can certainly be widened through improvements in the generation of large sets of recombinant proteins and the high throughput generation of capture molecules. Protein microarrays can be used to study protein function within


261

biological systems and can be applied broadly in discovery and quantitative analysis. Proteomic research, high throughput drug compound screening, and diagnostic applications envision a growing demand for protein microarray technologies. In medical research protein microarrays can significantly accelerate immune diagnostics by providing additional options for the simultaneous analysis of relevant diagnostic parameters. In addition, efforts must be undertaken to reduce sample volume, in particular in analytical applications relying on limited amounts of sample as in the analysis of multiple tumor markers from biopsy material, for example. Microarray technology will no doubt help uncover new possibilities with respect to patient monitoring during disease treatment and therapy.

5. [1] [2] [3] [4] [5] [6]

[7] [8] [9]

REFERENCES Ekins, R.P., (1989), ‘Multi-analyte immuno assay’, J. Pharm. Biomed. Anal., 7, 155-168. Anonymous, (2002), ‘Chipping Forecast II’, Nature Genetics, 32, 461552. Gygi, S.P., Rochon, Y., Franza, B.R., Aebersold, R., (1999), ‘Correlation between protein and mRNA abundance in yeast’, Mol. Cell. Biol., 19, 1720-1730. MacBeath, G., (2002), ‘Protein microarrays and Proteomics’, Nat. Genet., 32 Suppl, 526-532. Templin, M.F., Stoll, D., Schrenk, M., Traub, P.C., Vohringer, C.F., Joos, T.O., (2002), ‘Protein microarray technology’, Trends Biotechnol., 20, 160-166. Kononen, J., Bubendorf, L., Kallioniemi, A., Barlund, M., Schraml, P., Leighton, S., Torhorst, J., Mihatsch, M.J., Sauter, G., Kallioniemi, O.P., (1998), ‘Tissue microarrays for high throughput molecular profiling of tumor specimens’, Nat. Med., 4, 844-847. Petricoin, E.F., Liotta, L.A., (2002), ‘Proteomic analysis at the bedside: early detection of cancer’, Trends Biotechnol., 20, S30-34. Yan, F., Sreekumar, A., Laxman, B., Chinnaiyan, A.M., Lubman, D.M., Barder, T.J., (2003), ‘Protein microarrays using liquid phase fractionation of cell lysates’, Proteomics, 3, 1228-1235. Paweletz, C.P., Charboneau, L., Bichsel, V.E., Simone, N.L., Chen, T., Gillespie, J.W., Emmert-Buck, M.R., Roth, M.J., Petricoin III, E.F., Liotta, L.A., (2001), ‘Reverse phase protein microarrays which capture disease progression show activation of pro-survival pathways at the cancer invasion front’, Oncogene, 20, 1981-1989.

262


[10] Bouwman, K., Qiu, J., Zhou, H., Schotanus, M., Mangold, L.A., Vogt, R., Erlandson, E., Trenkle, J., Partin, A.W., Misek, D., et al., (2003), ‘Microarrays of tumor cell derived proteins uncover a distinct pattern of prostate cancer serum immunoreactivity’, Proteomics, 3, 22002207. [11] Haab, B.B., (2001), ‘Advances in protein microarray technology for protein expression and interaction profiling’, Curr. Opin. Drug. Discov. Devel., 4, 116-123. [12] Sreekumar, A., Nyati, M.K., Varambally, S., Barrette, T.R., Ghosh, D., Lawrence, T.S., Chinnaiyan, A.M., (2001), ‘Profiling of cancer cells using protein microarrays: discovery of novel radiation-regulated proteins’, Cancer Res., 61, 7585-7593. [13] Knezevic, V., Leethanakul, C., Bichsel, V.E., Worth, J.M., Prabhu, V.V., Gutkind, J.S., Liotta, L.A., Munson, P.J., Petricoin, E.F., 3rd, Krizman, D.B., (2001), ‘Proteomic profiling of the cancer microenvironment by antibody arrays’, Proteomics, 1, 1271-1278. [14] Belov, L., Huang, P., Barber, N., Mulligan, S.P., Christopherson, R.I., (2003), ‘Identification of repertoires of surface antigens on leukemias using an antibody microarray’, Proteomics, 3, 2147-2154. [15] Soen, Y., Chen, D.S., Kraft, D.L., Davis, M.M., Brown, PO, (2003), Detection and Characterizationof Cellular Immune Responses Using Peptide-MHC Microarrays. P. Los. Bio., 1, E65. [16] Moody, M.D., van Arsdell, S.W., Murphy KP., Orencole, S.F., Burns, C., (2001), ‘Array based ELISAs for high throughput analysis of human cytokines’, Biotechniques, 31, 186-190, 192-184. [17] Huang, R.P., (2001), ‘Simultaneous detection of multiple proteins with an array based enzyme-linked immunosorbent assay (ELISA) and enhanced chemi-luminescence (ECL)’, Clin. Chem. Lab. Med., 39, 209-214. [18] Tam, S.W., Wiese, R., Lee, S., Gilmore, J., Kumble, K.D., (2002), ‘Simultaneous analysis of eight human Th1/Th2 cytokines using microarrays’, J. Immunol. Methods., 261, 157-165. [19] Jenison, R., La, H., Haeberli, A., Ostroff, R., Polisky, B., (2001), ‘Silicon based biosensors for rapid detection of protein or nucleic acid targets’, Clin. Chem., 47, 1894-1900. [20] Wiese, R., Belosludtsev, Y., Powdrill T., Thompson, P., Hogan, M., (2001), ‘Simultaneous multi-analyte ELISA performed on a microarray platform’, Clin. Chem., 47, 1451-1457. [21] Schweitzer, B., Roberts, S., Grimwade, B., Shao, W., Wang, M., Fu, Q., Shu, Q., Laroche, I., Zhou, Z., Tchernev, V.T., et al., (2002), ‘Multiplexed protein profiling on microarrays by rolling-circle amplification’, Nat. Biotechnol., 20, 359-365.


263

[22] Finckh, P., Berger, H., Karl, J., Eichenlaub, U., Weindel, K., Hornauer, H., Lenz, H., Sluka, P., Weinreich, G.E., Chu, F., et al., (1998), ‘Microspot—an ultrasensitive microarray based ligand assay system. A practical application of ambient analyte assay theory’, Proc. UK NEQAS Meeting, 3, 155-165. [23] Shao, W., Zhou, Z., Laroche, I., Lu, H., Zong, Q., Patel, D.D., Kingsmore, S., Piccoli, S.P., (2003), ‘Optimization of Rolling-Circle Amplified Protein Microarrays for Multiplexed Protein Profiling’, J. Biomed. Biotechnol., 299-307. [24] Schweitzer, B., Wiltshire, S., Lambert, J., O’Malley, S., Kukanskis, K., Zhu, Z., Kingsmore, S.,F., Lizardi, P.M., Ward, D.C., (2000), ‘Immuno assays with rolling circle DNA amplification: a versatile platform for ultrasensitive antigen detection’, Proc. Natl. Acad. Sci. USA, 97, 10113-10119. [25] Robinson, W.H., DiGennaro, C., Hueber, W., Haab, B.B., Kamachi, M., Dean, E.J., Fournel, S., Fong, D., Genovese, M.C., de Vegvar, H.E., et al., (2002), ‘Auto–antigen microarrays for multiplex characterization of auto–antibody responses’, Nat. Med., 8, 295-301. [26] Joos, T.O., Schrenk, M., Hopfl, P., Kroger, K., Chowdhury, U., Stoll, D., Schorner, D., Durr, M., Herick, K., Rupp, S., et al., (2000), ‘A microarray enzyme-linked immunosorbent assay for autoimmune diagnostics. Electrophoresis, 21, 2641-2650. [27] Bacarese-Hamilton, T., Mezzasoma, L., Ingham, C., Ardizzoni, A., Rossi, R., Bistoni, F., Crisanti, A., (2002), ‘Detection of allergenspecific IgE on microarrays by use of signal amplification techniques’, Clin. Chem., 48, 1367-1370. [28] Harwanegg, C., Laffer, S., Hiller, R., Mueller, M.W., Kraft, D., Spitzauer, S., Valenta, R., (2003), ‘Microarrayed recombinant allergens for diagnosis of allergy’, Clin. Exp. Allergy, 33, 7-13. [29] Hiller, R., Laffer, S., Harwanegg, C., Huber, M., Schmidt, W.M., Twardosz, A., Barletta, B., Becker, W.M., Blaser, K., Breiteneder, H., et al., (2002), ‘Microarrayed allergen molecules: diagnostic gatekeepers for allergy treatment’, Faseb. J., 16, 414-416. [30] Wiltshire, S., O’Malley, S., Lambert, J., Kukanskis, K., Edgar, D., Kingsmore, S.,F., Schweitzer, B., (2000), ‘Detection of multiple allergen-specific IgEs on microarrays by immuno assay with rolling circle amplification’, Clin. Chem., 46, 1990-1993. [31] Knight, P.R., Sreekumar, A., Siddiqui, J., Laxman, B., Copeland, S., Chinnaiyan, A., Remick, D.G., (2004), ‘Development of a sensitive microarray immuno assay and comparison with standard enzymelinked immuno assay for cytokine analysis’, Shock, 21, 26-30.

264


[32] Nielsen, U.B., Cardone, M.H., Sinskey, A.J., MacBeath, G., Sorger, P.K., (2003), ‘Profiling receptor tyrosine kinase activation by using Ab microarrays’, Proc. Natl. Acad. Sci. USA, 100, 9330-9335. [33] Carson, R.T., Vignali, D.A., (1999), ‘Simultaneous quantitation of 15 cytokines using a multiplexed flow cytometric assay’, J. Immunol. Methods., 227, 41-52. [34] Dunbar, S.A., van der Zee, C.A., Oliver, K,G., Karem, K.L., Jacobson, J.W., (2003), ‘Quantitative, multiplexed detection of bacterial pathogens: DNA and protein applications of the Luminex LabMAP system’, J. Microbiol. Methods, 53, 245-252. [35] Prabhakar, U., Eirikis, E., Davis, H.M., (2002), ‘Simultaneous quantification of proinflammatory cytokines in human plasma using the LabMAP assay’, J. Immunol. Methods., 260, 207-218. [36] De Jager, W., Te Velthuis, H., Prakken, B.J., Kuis, W., Rijkers, G.T., (2003), ‘Simultaneous detection of 15 human cytokines in a single sample of stimulated peripheral blood mononuclear cells’, Clin. Diagn. Lab. Immunol., 10, 133-139. [37] Chen, R., Lowe, L., Wilson, J.D., Crowther, E., Tzeggai, K., Bishop, J.E., Varro, R, (1999), ‘Simultaneous Quantification of Six Human Cytokines in a Single Sample Using Microparticle based Flow Cytometric Technology’, Clin. Chem., 45, 1693-1694. [38] Fulton, R.J., McDade, R.L., Smith, P.L., Kienker, L.J., Kettman, J.R., Jr., (1997), ‘Advanced multiplexed analysis with the FlowMetrix system’, Clin. Chem., 43, 1749-1756. [39] Bellisario, R., Colinas, R.J., Pass, K.A., (2001), ‘Simultaneous measurement of antibodies to three HIV-1 antigens in newborn dried blood-spot specimens using a multiplexed microsphere based immuno assay’, Early Hum. Dev., 64, 21-25. [40] McBride, M.T., Gammon, S., Pitesky, M., O’Brien T.W., Smith, T., Aldrich, J., Langlois, R.G., Colston, B., Venkateswaran, K.S. (2003), ‘Multiplexed liquid arrays for simultaneous detection of simulants of biological warfare agents’, Anal. Chem., 75, 1924-1930. [41] McBride, M.T., Masquelier, D., Hindson, B.J., Makarewicz, A.J., Brown, S., Burris, K., Metz, T., Langlois, R.G., Tsang, K.W., Bryan, R., et al., (2003), ‘Autonomous detection of aerosolized Bacillus anthracis and Yersinia pestis’, Anal. Chem., 75, 5293-5299. [42] Ge, H., (2000), ‘UPA, a universal protein array system for quantitative detection of protein-protein, protein-DNA, protein-RNA and proteinligand interactions’, Nucleic Acids Res., 28, e3. [43] Bulyk, M.L., Gentalen, E., Lockhart, D.J., Church, G.M., (1999), ‘Quantifying DNA protein interactions by double-stranded DNA arrays’, Nat. Biotechnol., 17, 573-577.


265

[44] Zhu, H., Klemic, J.F., Chang, S., Bertone, P., Casamayor, A., Klemic, K.G., Smith, D., Gerstein, M., Reed, M.A., Snyder, M., (2000), ‘Analysis of yeast protein kinases using protein chips’, Nat. Genet., 26, 283-289. [45] Arenkov P., Kukhtin A., Gemmell A., Voloshchuk. S., Chupeeva. V., Mirzabekov. A., (2000), ‘Protein microchips: use for immuno assay and enzymatic reactions’, Anal. Biochem., 278, 123-131. [46] MacBeath, G., S.L., (2000), ‘Printing proteins as microarrays for high throughput function determination’, Science, 289, 1760-1763. [47] Houseman, B.T., Huh, J.H., Kron, S.J., Mrksich, M., (2002), ‘Peptide chips for the quantitative evaluation of protein kinase activity. Nat. Biotechnol. 2002, 20, 270-274. [48] Lizcano, J.M., Deak, M., Morrice, N., Kieloch, A., Hastie, C.J., Dong, L., Schutkowski, M., Reimer, U., Alessi, D.R., (2002), ‘Molecular basis for the substrate specificity of NIMA-related kinase-6 (NEK6). Evidence that NEK6 does not phosphorylate the hydrophobic motif of ribosomal S6 protein kinase and serum- and glucocorticoid-induced protein kinase in vivo’, J. Biol. Chem., 277, 27839-27849. [49] Reineke, U., Volkmer-Engert, R., Schneider-Mergener, J., (2001), ‘Applications of peptide arrays prepared by the SPOT-technology’, Curr. Opin. Biotechnol., 12, 59-64. [50] Zhu, H., Bilgin, M., Bangham, R., Hall, D., Casamayor, A., Bertone, P., Lan, N., Jansen, R., Bidlingmaier, S., Houfek, T., et al., (2001), Global Analysis of Protein Activities Using Proteome Chips. Science, 2001, 26. [51] Michaud, G.A., Salcius, M., Zhou, F., Bangham, R., Bonin, J., Guo, H., Snyder, M., Predki, P.F., Schweitzer, B.I., (2003), ‘Analyzing antibody specificity with whole proteome microarrays’, Nat. Biotechnol., 21, 1509-1512. [52] Lueking, A., Possling A., Huber O., Beveridge, A., Horn, M., Eickhoff, H., Schuchardt, J., Lehrach, H., Cahill, D.J., (2003), ‘A Nonredundant Human Protein Chip for Antibody Screening and Serum Profiling’, Mol. Cell. Proteomics, 2, 1342-1349. [53] Espejo, A., Cote, J., Bednarek, A., Richard, S., Bedford, M.T., (2002), ‘A protein-domain microarray identifies novel protein-protein interactions’, Biochem. J., 367, 697-702. [54] Martin, K., Steinberg, T.H., Cooley, L.A., Gee, K.R., Beechem, J.M., Patton, W.F., (2003), ‘Quantitative analysis of protein phosphorylation status and protein kinase activity on microarrays using a novel fluorescent phosphorylation sensor dye’, Proteomics, 3, 1244-1255.

266 PROTEIN MICROARRAYS: TECHNOLOGIES AND APPLICATIONS [55] Wang H., Hanash, S.M., (2002), ‘Contributions of proteome profiling to the molecular analysis of cancer’, Technol. Cancer Res. Treat., 1, 237-246. [56] Fukui S., Feizi, T., Galustian, C., Lawson, A.M., Chai, W., (2002), ‘Oligosaccharide microarrays for high throughput detection and specificity assignments of carbohydrate-protein interactions’, Nat. Biotechnol., 20, 1011-1017. [57] Feizi, T., Fazio, F., Chai, W., Wong, C.H., (2003), ‘Carbohydrate microarrays—a new set of technologies at the frontiers of glycomics’, Curr. Opin. Struct. Biol., 13, 637-645. [58] MacBeath, G., Koehler, A., Schreiber, S., (1999), ‘Printing small molecules as microarrays and detecting protein-ligand interactions en masse’, J. Am. Chem. Soc., 121, 7967-7968. [59] Winssinger, N., Ficarro, S., Schultz, P.G., Harris, J.L., (2002), ‘Profiling protein function with small molecule microarrays’, Proc. Natl. Acad. Sci. USA, 99, 11139-11144. [60] Gosalia, D.N., Diamond, S.L., (2003), ‘Printing chemical libraries on microarrays for fluid phase nano-liter reactions’, Proc. Natl. Acad. Sci. USA, 100, 8721-8726. [61] Hanash, S.M., Bobek, M.P., Rickman, D.S., Williams, T., Rouillard, J.M., Kuick, R., Puravs, E., (2002), ‘Integrating cancer genomics and Proteomics in the post-genome era’, Proteomics, 2, 69-75. [62] Petricoin, E.F., Zoon KC., Kohn EC., Barrett JC., Liotta, L.A., (2002), Clinical Proteomics: translating benchside promise into bedside reality’, Nat. Rev. Drug. Discov., 1, 683-695. [63] Paweletz, C.P., Charboneau, L., Bichsel, V.E., Simone, N.L., Chen, T., Gillespie, J.W., Emmert-Buck, M.R., Roth, M.J., Petricoin, I.E., Liotta, L.A., (2001), ‘Reverse phase protein microarrays which capture disease progression show activation of pro-survival pathways at the cancer invasion front’, Oncogene, 20, 1981-1989. [64] Espina, V., Dettloff, K.A., Cowherd, S., Petricoin, E.R.E., Liotta, L.A., (2004), ‘Use of proteomic analysis to monitor responses to biological therapies’, Expert Opin. Biol. Ther., 4, 83-93. [65] Espina, V., Mehta, A.I., Winters, M.E., Calvert, V., Wulfkuhle, J., Petricoin, E.F., 3rd, Liotta, L.A., (2003), ‘Protein microarrays: molecular profiling technologies for clinical specimens’, Proteomics, 3, 2091-2100. [66] Nishizuka S., Charboneau, L., Young, L., Major, S., Reinhold, W.C., Waltham, M., Kouros-Mehr, H., Bussey, K.J., Lee, J.K., Espina, V., et al., (2003), ‘Proteomic profiling of the NCI-60 cancer cell lines using new high density reverse-phase lysate microarrays’, Proc. Natl. Acad. Sci. USA, 100, 14229-14234.

PROTEIN MICROARRAYS: TECHNOLOGIES AND APPLICATIONS 267 [67] Nam, M.J, Madoz-Gurpide, J., Wang, H., Lescure, P., Schmalbach, C.E., Zhao, R., Misek, D.,E., Kuick, R., Brenner, D.E., Hanash, S.M., (2003), ‘Molecular profiling of the immune response in colon cancer using protein microarrays: occurrence of auto–antibodies to ubiquitin C-terminal hydrolase L3’, Proteomics, 3, 2108-2115. [68] Davies, H.A., (2000), ‘The ProteinChip System from Ciphergen: a new technique for rapid, micro-scale protein biology’, J. Mol. Med., 78, B29. [69] von Eggeling, F., Davies, H., Lomas, L., Fiedler, W., Junker, K., Claussen, U., Ernst, G., (2000), ‘Tissue-specific microdissection coupled with ProteinChip array technologies: applications in Cancer Research’, Biotechniques, 29, 1066-1070. [70] Xiao, Z., Jiang, X., Beckett, M.L., Wright, G.L., Jr., (2000), ‘Generation of a baculovirus recombinant prostate-specific membrane antigen and its use in the development of a novel protein biochip quantitative immuno assay’, Protein Expr. Puriff , 19, 12-21. [71] Tang, N., Tornatore, P., Weinberger, S.R., (2004), ‘Current developments in SELDI affinity technology’, Mass. Spectrom. Rev., 23, 34-44. [72] Wang, Y., (2004), ‘Immunostaining with dissociable antibody microarrays’, Proteomics, 4, 20-26. [73] Kallioniemi, O.P., Wagner, U., Kononen, J., Sauter, G., (2001), ‘Tissue microarray technology for high throughput molecular profiling of cancer’, Hum. Mol. Genet., 10, 657-662.

Chapter 9 LAB-ON-A-CHIP SYSTEMS FOR CELLULAR ASSAYS Bernhard Wolf, Martin Brischwein*, Helmut Grothe, Christoph Stepper, Johann Ressler, Thomas Weyh Heinz Nixdorf Lehrstuhl für Medizinische Elektronik Technische Universität München Theresienstrasse 90, Geb. N 3 D-80333 München, FR.G.* Contact: Dr. Martin Brischwein Tel.: (089) 289 22344 Fax: (089) 289 22950 E-mail: [email protected]

Abstract:

Cells are able to respond in an extremely sensitive way to changes in their environment. Electronic microstructures on sensor chips can be applied to analyse such responses by recording parameters of cell metabolism, cell morphology and electrical patterns. These parameters are linked intimately to the cellular signaling network and can thus be used as indicators for any perturbation of cellular physiology like toxic insults or receptor activation events. “Cell-on-a-chip” systems involving both cell culturing and functional analysis might become routine tools in pharmaceutical drug discovery, clinical diagnostics and whole-cell biosensor applications. However, the success of such devices will depend on robust and smart technologies in chip packaging and electric connections, electronic sensor control and automated liquid handling. A presentation of solutions developed in the author’s laboratory is given together with a survey of current approaches in the whole field of sensor based cellular analysis.

Key words:

cellular assay, microsensor, cell-on-a-chip, extra–cellular recording, cell metabolism, cell morphology

1.

INTRODUCTION

Animal and plant cells are dynamic nano-structured microsystems which have evolved during billions of years. They contain functional subunits located in distinct compartments and interconnected by a complex signaling network. Some of these compartments can be made visible by electron microscopy (Fig. 9-1).


269

270

LAB-ON-A-CHIP SYSTEMS FOR CELLULAR ASSAYS Y

Figure 9-1. Transmission electron microscopy photograph of an explanted human tumor cell. Some of the different cellular nano-sized structures with specific functions are visible.

For a selective sensing of input signals from the outside the cell uses specific receptor proteins. These transmit and amplify the signals to elecit an immediate response or to change gene activities in the cell nucleus. Fig. 9-2 shows schematically parts of the complex intra–cellular signaling network. A differentiated eukaryotic cell can perform 103–104 different chemical reactions simultaneously [1] in a volume of a few picoliters. In order to better understand the dynamic properties of this network and its input–output relationship, approaches of system analysis have to be employed adequate for a description of the massively parallel structure of cellular signal computing [2,3,4]. From these studies we have learned that multi-parameter planar micro-sensor arrays might be ideal tools for an on-line and real time acquisition of complex cellular reaction patterns and for the realization of biohybrid components [2].

LAB-ON-A-CHIP SYSTEMS FOR CELLULAR ASSAYS

271

Figure 9-2. Schematic representation of the complexity of interacting signal transduction and metabolic pathways. Right panel adapted from [4].

Not only neuronal cells have computing capabilities. In a typical cellular reaction pattern a weak biochemical stimulation of the cellular signal perception apparatus is amplified by intra–cellular signaling cascades to perform an appropriate functional reaction. The functional analogy in data processing capabilities between a technical neuron and a eukaryotic cell is illustrated in Fig. 9-3. The state of the art in the functional analysis of living cells is the use of micro test plates in combination with various optical readout systems [5]. Components such as optical microscopes, fiber optics, or CCD cameras are used to detect visible, fluorescent, or luminescent signals from cells or tissues. The advent of cellular engineering to include reporter elements such as green fluorescent protein has allowed us to detect specific cellular signaling events. Although the ability to array and to screen cells in high density formats meets the demands of many users, optical screening seems less suited for long term monitoring of cultured cells and tissues, mostly owing to toxic or phototoxic properties of many dyes. In contrast to micro electrode structures, optic setups with light sources, light guides and detectors are less easily integrated into small portable lab-on-a-chip systems. Moreover, no truly appropriate optical technique is available for studies on cell adhesion, cell metabolic activity or electric activity.

272


Figure 9-3. Analogy of the signal transmission in a technical neuron and in a eukaryotic cell. Both calculate an ‘output signal’ by weighting of different incoming signals.

The response to input signals from the environment, i.e., the cellular ‘output signal’, is usually accompanied by a change in the rates of energy metabolism, by changes in electric activity, and/or by cell morphological alterations. To investigate cellular signaling properties for diagnostic or screening approaches these changes can be monitored in real time and noninvasively by allowing cells and tissues to make contact with adequate micro-sensors. This coupling is usually achieved by growing cells directly on the surface of silicon or glass based sensor chips which are inert and nontoxic materials and accepted by many cells as substrate for stable adhesion and growth. The chips may integrate different types of sensor structures that have, in the case of electric microsensors, a potentiometric, amperometric or impedimetric principle of function. The setup may be regarded as a celltransducer hybrid that can be used in different applications such as pharmacology, neurobiology, research in cell biology, medical diagnostic tests, and in cell based biosensors. After having produced a sensor chip (which may integrate micro-fluidic structures or not), there are additional challenges in the development of practical analytical tools. For assays and screening applications, the sensor chips have to be arranged in high density arrays. It requires substantial technological efforts in chip–integrated electronic sensor control and in data


273

pre-evaluation to reduce the number of necessary electric connections to the external electronic equipment. Chip packaging, such as the incorporation of chips into a multi-well plate format with all the electric connections shielded safely against liquids, must be optimized in order to come up with an inexpensive and convenient device. Micro-fluidics must be characterized properly in terms of flow dynamics and the physics of the interactions between aqueous samples and the different surfaces, regardless of what kind of micro-fluidic system is chosen. Finally, the importance of computation of final results from sensor raw data, statistical design and extraction of information from response libraries will increase along with increasing array densities. Most of the cited work on lab-on-a-chip concepts for cell sampling, cell trapping and sorting, cell treatment, and cell analysis derives from the past five years. The majority of assays on intact cells seems to be suited to a short term analysis of cell suspensions without extended cell culturing. An exception may be those micro-fluidics carrier devices which allow the culturing, inspection, and characterization of cells on a microscope stage [6,7]. On the other hand, existing sensor chips for a dynamic monitoring of cell parameters such as electric activity or metabolic rates usually do not contain micro-fluidic components [8,9,10]. It appears that only very few labon-a-chip systems for a simultaneous cultivation and monitoring of cells with integrated sensor functions (i.e., µTAS systems with micro bioreactors) have been described so far [11,12]. While in many cases cultured cell populations are examined as a whole, there are various approaches to manipulate and to characterize single cells which have been reviewed recently [13]. The field of application for cell based bioanalytical systems is as large as the variety of different cell types to investigate. The source and preparation of cells and tissues is most important for the quality of the analysis. Whilst primary cell cultures often provide the greatest approximation to the in vivo phenotype, they also present potential sources of variability. The discovery of different types of stem cells and the possibility of inducing differentiation in vitro presents a new opportunity to obtain a source of cells with fewer functional defects and alterations than found in ordinary cell lines. On the other hand, a range of applications requires cellular material freshly derived from individuals. As an example, predictive individualized drug screening in oncology is performed on tumor tissue biopsies. It appears that miniaturization and the possibility of directly following up the response profiles of short term cultures in the course of the drug treatments makes chip technologies particularly suitable for this kind of cell analysis since the amount of the available tissue material is very limited [14,15,16]. Again, however, the preparation of the tissue (e.g., consideration of functional interactions between

274


malignant cell and stroma cells) and the composition of the test medium (e.g., presence of serum components) have to be considered carefully. In cell based biosensors cellular recognition and signaling capabilities can also be exploited to detect a physiological response to possibly hazardous substances. In this case cells are used for providing the appropriate receptor proteins and for the generation and amplification of an output signal which is detectable by a physical transducer. In fact, one of the most important advantages of using living cells as biological signal discriminators and amplifiers is their ability to regenerate the signal receptors and amplifying signaling cascades permanently in a native state. Such sensors which could be specific for toxins might be used, for example, in deterring biological warfare weapons, or they could serve as a biomonitor of waste water effluents [17,18,19,20]. For such devices additional criteria must be met. Apart from the selection of the target cell/organism (which may be genetically manipulated) appropriate for the specific range of compounds to detect, there are numerous issues to be considered. These include energy management for small, portable devices, sample preparation (filtration, bioavailability, sterilization), response time, possible effects of physiological adaptation by sub-toxic concentrations of chemical compounds, life support systems for robust and invariable sensitivities and well controlled abiotic factors such as temperature and light exposition. The question of the minimal detectable concentration is linked not only to the inherent sensitivity of the cellular target, but also to the determination of any sources of noise in the system. The probability of false alarm needs to be minimized by correlation analysis of parallel sensors. The aim of this contribution is to give a general overview of the state of the art in cellular assays on chips. The emphasis will be on technologies and applications related to the author’s research objectives. These include not only the current microsensor strategies for cellular assays but also approaches for manipulation of cells on chips.

2.

DESIGN AND FABRICATION OF CHIPS FOR CELL BASED ASSAYS

In the following section a brief description of the basics of sensor chip design and processing is provided. It is confined to the fabrication of ISFETs, LAPS, and various electrode structures on silicon and glass substrates as performed in the available clean room facilities. For the silicon chip a layout was developed which includes different types of sensors for multi-parameter readout (Fig. 9-4). Four ISFETs, two


275

pO2-FETs, one amperometric pO2 sensor, an inerdigitated electrode structure (IDES), a temperature diode and a reference transistor are processed.

(A)

(B)

Figure 9-4. (A) Silicon chip layout. The outer dimensions are 7.5×7.5 mm. Four pH-ISFETs, one inter-digitated electrode structure (electrode width and distance: 50 µm), one amperometric sensor structure and a temperature diode are integrated. The chip is bonded to a printed wire board (24×24 mm) and packaged to form a free circular opening (diameter: 6 mm) for sensors and cell culturing. (B) Packaged chip, on PLCC-68 socket, with fluidic setup.

Most of the process steps are standard MOS. Owing to the particular application for cell culture however, special materials and a different order of process steps must be applied. Silicon chip fabrication is started with diffusion of the source and drain areas (channel–MOSFETs). Then dielectric material for the gate area is processed by dry oxidation and additional deposition of a thin silicon nitride layer (10 nm SiO2, 20–30 nm Si3N4). This non-standard step is included to obtain better cell adhesion on the nitride layer. Other materials with different physical parameters (dielectric constant, proton binding characteristics, surface charge) such as Al2O3 and TaO5 are currently being tested. After defining the active areas by local oxidation and deposition of a polysilicon protection layer, a titanium layer of a few nm thickness is deposited to ensure better adhesion of the following sputtered platinum or palladium metalization (200–300 nm). The metalization layer yields the electrode structures for IDES, pO2-FET and amperometric pO2-sensor by a lift off photoresist step. The use of these metals implements another non-standard step in the fabrication process but it is necessary for catalyzing electrochemical reactions and for ensuring bio-compatibility. On the ISFETs two passivation layers (silicon nitride and/or silicon oxide) are deposited and

276


opened on the gate areas. It is important that these passivation layers provide good adhesion and sealing against liquids to prevent any leakage current. The completed silicon chips are mounted and bonded on printed wire boards fitting into standard PLCC 68 sockets. They are encapsulated using a mold which filled with a bio-compatible epoxy resin. Prior to cell culture the packaged chips are sterilized either with 70% ethanol or by autoclaving. Another silicon sensor for pH measurements is the light addressable potentiometric sensor (LAPS). As in ISFET fabrication, all process steps are standard MOS, but fewer steps are required. The fabrication starts with an nor p-doped silicon substrate, on which a dielectric layer (SiO2) is deposited by dry oxidation. For better proton exchange characteristics at the sensor surface, an additional silicon nitride layer (Si3N4) can be deposited. For an ohmic contact on the back of the silicon substrate, a titanium adhesive layer and a gold metalization layer are sputtered. This metalization layer is structured by a lift off photoresist step to get an open area illuminated by an infrared light emitting diode (IR–LED). Like the ISFET, the LAPS is an electrolyte–insulator–semiconductor structure. In the case of the LAPS, a DC bias voltage is applied to this structure. The width of the depletion layer which is formed at the insulator– semiconductor interface and thus the capacitance of this interface is affected by the pH-dependent surface potential. The variation of the capacitance is read out as a photo current, induced by the modulated light of the IR-LED. All potentiometric methods aimed at the measurement of absolute concentrations (e.g., of protons) are dependent on the availability of a high quality reference potential. Obviously the integration of Ag/AgCl half cells on chips with electrolyte solutions saturated with Ag+ ions is problematic, and, indeed, the advantages of planar potentiometric sensors are often attenuated when planar sensor structures have to be combined with conventional ‘tube and wire’ reference electrodes. Fortunately some applications tolerate a certain degree of signal drift which may be caused by simplified, so called ‘quasi-reference’ electrodes (e.g., micro-planar Ag/AgCl structures, directly immersed into the cell culture medium). An example for the use of such chips is the monitoring of changes in pH owed to cell metabolism within short time intervals (e.g., some minutes). However, it should be noted that the determination of absolute values of pH is usually not the objective of such configurations. Its excellent optic properties make glass an attractive substrate material for cell chips. It provides the option that cells are studied with optic microscopes or other photonic techniques in addition to measurements with electric sensors. An economic advantage in comparison to silicon based sensor technologies are the lower fabrication costs, especially at small and medium production rates. By using photo sensitive glass materials (e.g.,


277

Foturan, Schott), several process steps can be facilitated. The chip made in the author’s laboratory (Fig. 9-5) includes two amperometric three-electrode structures and two IDES. The metal structures are fabricated using the same platinum lift off process as described above. Afterwards a single passivation layer is deposited (Si3N4, thickness of about 500 nm) to prevent electric shortcuts in liquids. Sensors for pH are either integrated by a glass–silicon hybrid technology (insertion of ISFET-silicon chips into the glass chip) or by deposition of proton sensitive layers of metal oxides such as Al2O3, TaO5 or RuO2. A special passivation (hard mask) is deposited on the glass prior to etching in order to protect the surface.

Figure 9-5. Glass chip layout. The outer dimensions are 24.0×33.8 mm, the cell culture area has a diameter of 12 mm. Two inter-digitated electrode structures and two amperometric electrode structures are integrated. For pH-measurement a 3.5×7.0 mm silicon chip can be inserted to form a glass–silicon hybrid. Alternatively, pH sensitive metal oxide structures such as ruthenium oxide or iridium oxide can be deposited using thin film technology. Electric connections are made with needle contacts.

278


Based on such glass chips, the fabrication of chip arrays with densities approaching those necessary for screening purposes can be started. In a first step 24 IDES were integrated on a common glass plate forming the bottom of a 24 well plate. The structures were processed on a glass plate covered with an ITO layer (Indium Tin Oxide, transparent and conductive) by photoresist–masked etching. As the number of electric pathways and contacts for sensors rapidly increases with the density of such plates, an electronic readout system was developed which allowed a successive impedance measurement in each well [21].

Figure 9-6. 24-well plate. In this layout sensors for electric impedance are combined with optic sensors for pH and oxygen (PreSens GmbH, Regensburg, Germany, blue and red spots). This plate allows functional assays on cell metabolic activity and cell morphological alterations.

3.

CELL CULTURE ON CHIPS AND MICROFLUIDIC SYSTEMS

Cells require a defined environment in order to survive and to respond reproducibly to signals from the outside. As a general rule it is desirable to mimic the physiological conditions of the cell. The major conditions which have to be maintained in culturing cells are (1) physicochemical properties of the culture medium, (2) temperature, and (3) sterility. Physicochemical properties of the cell culture medium include pH, oxygen partial pressure, osmolarity, and defined concentrations of nutrients. In the course of cell culturing, metabolic waste products and secreted


279

compounds (e.g., proteins) become additional constituents. For medium- and long term assays on chips, a liquid handling system is necessary for adjusting and to maintain the required physico-chemical parameters and, eventually, drug concentrations of the medium. To ensure sensitive measurements of metabolic rates small closed micro volumes should surround the cell cultures: only when a sufficiently high ratio of cell number to medium volume is achieved can those rates be recorded precisely [22]. This brings about an additional requirement, namely, the repeated exchange of cell culture media (with cycle times of some minutes) while retaining the cells in the micro-chambers. Significant advances have recently been made in the development of micro-scaled devices for biomedical purposes. One category out of these ‘lab-on-a-chip’ systems aims at the fabrication of micro-containments for a miniaturized cell culture. The surfaces of the resulting cell microarrays can be patterned with molecular structures with selective cell adhesion properties [23,24,25]. The ability to decrease the consumption of valuable reagent compounds and (primary) cells is obviously the most attractive feature of cellular lab-ona-chip devices. Some basic design considerations for micro-fluidic systems to meet micro-environmental requirements of cells have been summarized recently [26]. In micro-fluidic systems however, there is a limit to the miniaturization of micro channels since beyond a critical size aggregation of cells and debris may clog up the system. With flows driven by electroosmotic forces, possible side effects on cellular physiology exerted at high electric field strength must be considered, since electric fields in the range of several hundreds of volts per centimeter are required [27,28]. Another difficulty which often moderates the advantages of micro-fluidic systems is the necessary interfacing between microchip and ‘macro world’. Samples and media usually have to be introduced manually or with laboratory pumps. A more practicable way for interfacing cell chips with ‘macro–fluidic’ components, suitable for high throughput purposes, may be the use of pipetting robots: this allows the exchange of culture media, addition of drug solutions, the adjustment of microvolumes, and the aspiration of cell suspensions in combination with cell chip arrays [29]. There are also efforts to stabilize and to conserve mammalian cells on solid substrates for the preparation of cell based biosensors with extended shelf lives. Such efforts may become meaningful for applications such as environmental monitoring, where robust and portable devices are most important [30].


280

(A)

(B)

Figure 9-7. (A): Front view of a screening device equipped with six channels (modules), each with cell culture vials on silicon sensor chips connected to a fluid perfusion system. An internet port allows for online transfer of data. Fig. 9-7 (B): Single modules can be operated independently as portable cell signal analyzers. For remote control applications, sensor data can be transmitted by wireless communication to a mobile phone.

Figure 9-8. Experimental setup for a 24-well plate. The glass bottom of the cell culture plate is equipped with an array of 24 sensor units. Each unit consists of an indium–tin–oxide (ITO) structure for electric impedance measurement and optic sensor spots for pH and oxygen measurement (read out with optic interfaces at the bottom of the plate). A pipetting robot (not shown) coordinates microvolume adjustment and medium exchange.

4.

DETECTABLE CELLULAR OUTPUT SIGNALS

Currently chip based sensors are capable of detecting four major cellular output signals (Fig. 9-9): (1) Changes in the rate of proton extrusion are detected with pH micro-sensors, namely ion sensitive field effect transistors


281

(ISFETs) or light activated potentiometric sensors (LAPS). (2) Changes in the rate of cellular oxygen exchange are detected with planar oxygen sensors such as amperometric oxygen sensors or so called pO2-FETs (an ISFETelectrode combination, see below). (3) Changes in cell morphology and membrane associated electrical effects are monitored with impedance sensors such as interdigitated electrode structures (IDES). (4) Electrophysiological activity of cells, i.e., rapid changes in the membrane potential of electrically active cells or currents of ion channels are detected with micro electrode arrays or with patch–clamp chips. With most lab-on-a-chip systems only one parameter is recorded. However, there are efforts to develop multi-parameter chips combining several complementary sensors for a more comprehensive understanding of dynamic cellular behavior [31,32].

Figure 9-9. Parameters of living cells which are currently monitored by chip based sensors.

282


4.1

Cell Metabolism

4.1.1

Extra–cellular Acidification

For the analysis of proton fluxes across the cell membrane pH-ISFETs with gate regions of about 10×100 µm2 (L×W) are used. In such transistors a low current passing through a semiconductor channel between two electrodes (‘source and drain’) is controlled by the voltage of a third, nonmetallic, ‘gate-area’ on top of this channel. In pH-ISFETs the gate is covered with an isolating material which binds selectively protons from the ambient analyte solution, resulting in a pH dependent potential (potential between gate and source connection, UGS). With silicon nitride as gate isolator material, the pH sensitivity is between 40 and 55 mV/pH. Owing to the pH sensitivity of the structures the rate of extra–cellular acidification of cells growing directly on the sensors can be detected [33,34]. Since the typical noise level of pH recordings by ISFETs is about 0.5 mV, pH variations of about 0.01 pH can be detected. Most important for a low noise level is a fluid connection to the reference electrode maintaining the electric potential of the analyte solution. In a cell culture setting a simple and practicable method is to use the culture medium itself as the electrolyte solution for an Ag/AgCl reference electrode, if precautions are made to avoid intoxication with silver ions.

Figure 9-10. A measurement sequence reflecting a cyclic pattern of extra–cellular acidification of LS 174 T cells (a cell line derived from human colon adenocarcinoma) growing on an ISFET on a silicon sensor chip. Addition of Triton X-100 leaves a residual signal which is subtracted upon raw data evaluation. For evaluation the slope of the graph during the flow off intervals is calculated.


283

Monitoring extra–cellular acidification rates involves a stop–and–flow mode of a fluid perfusion: Fresh culture medium is conducted on the cell chip during a first flow interval, followed by a stop interval for acid accumulation. A series of such stop–and–flow cycles affecting the measured UGS of an ISFET is shown in Fig. 9-10. Values were measured every six seconds. For the transformation of ISFET raw data into quantitative information about extra–cellular acidification the slope of the graph (voltage vs. time) during the stop phase is calculated by linear regression [30]. With LAPS technology the same fluidic mode has been employed successfully for microphysiometry. A general description of the physical chemistry and cell biology underlying the measurement of extra–cellular acidification was given in [35]. In a comparative study LAPS and ISFETs showed similar features with respect to pH sensitivity and drift [36]. Further studies aimed at the evaluation of a surface potential imaging (with the surface potential influenced by the cell metabolic activity) with highly integrated sensors [37]. Since cell cultures or tissue specimen (e.g., slices of explanted tumor biopsies) are often heterogeneous, the possibility of resolving measured pH values spatially would improve the assay. A reasonable setup for obtaining multi-site records can be an ISFET array which monitors pH values directly beneath the cells or a CMOS camera [38]. An optical technique for monitoring extra–cellular acidification rates called ‘Bead Injection Spectrophotometry’ has also been developed [39] and realized as a semi-automated ‘lab on a valve’ system [40]. This approach uses micro beads loaded with pH indicators. The absorbance of the indicator changes upon pH variations caused by the extra–cellular acidification of cells is detected with a spectral photometer.

4.1.2

Cellular Oxygen Exchange

The measurement of dissolved oxygen in a cell biological context is a convenient way of analyzing respiratory activity. Usually amperometric Clark–type of sensor electrodes are employed. The technical conditions necessary for such recordings can be estimated from the typical rate of oxygen consumption in mammalian cell cultures which is roughly between 10-16 and 10-17 mol/cell*sec [41,42,43]. Despite the development of technological solutions for planar miniaturized amperometric oxygen sensors [44] and despite the potential advantages of planar chip integrated sensors for monitoring in cell cultures, only a few published reports exist with an application in cell biology [30,45]. One of the reasons is probably the poor long term stability of most of these sensors. Moreover, the fabrication

284


technology which includes the deposition of reference electrode and membrane structures, is not compatible with a standard CMOS process. Generally, three-electrode systems controlled by a potentiostatic electronic circuit and two electrode systems are feasible. By using a threeelectrode system a net consumption of oxygen in the vicinity of the cathode and a consumption of silver in reference electrodes can be avoided. However, this would not be serious points if the cathodic current is limited to a few nano-amperes and the lifetime of the reference electrode is at least equivalent to the lifetime of the whole chip. Amongst the different technological solutions for planar amperometric oxygen sensors tested in our laboratory, a simplified palladium (Pd) electrode structure without a gas permeable membrane (but with a poly-HEMA hydrogel cover to prevent cells from directly growing on the electrodes) yielded promising results. As a (quasi-) reference electrode, a bare Pd electrode is connected. The constant concentration of chloride ions in cell culture media is able to sufficiently stabilize the reference electrode’s potential. However, such a sensor configuration cannot be used for the determination of absolute pO2 values but only for the analysis of relative changes within short time periods. A measurement with cells growing on the bare electrode structures is shown in Fig. 9-11. The advantage of avoiding additional membrane structures would clearly be the feasibility of a low cost, CMOS-compatible fabrication of those sensors.

Figure 9-11. A measurement sequence reflecting a cyclic pattern of oxygen depletion owing to cell respiration in a cell culture of LS 174 T cells which has been grown to confluence directly on an amperometric sensor structure. During the stop intervals of the fluid systems the cellular oxygen consumption causes decreasing current values. Addition of Triton X-100 disrupts cell membranes and cell respiration ceases. For evaluation of sensor raw data an algorithm similar to that used for ISFET evaluation (see Fig. 9-10) was used.


285

Another method of analyzing cellular oxygen consumption is based on the formation of OH- ions during oxygen reduction and the corresponding pH shift [46,47]. If the pH sensitive gate area of an ISFET is surrounded by an electrode, the ISFET will detect an increase of pH when the palladium (or platinum) electrode catalyzes oxygen reduction at a cathodic potential (–750 mV vs. Ag/AgCl). In order to prevent deterioration of electro-catalytic activity intermittent anodic potentials are applied. In a recent work [48] a cyto-sensor micro physiometer was modified in order to measure cellular oxygen consumption rates. However, the voltammetric measurement was not performed with planar electrodes on the (LAPS-) silicon chip but with nafion-coated platinum wires inserted into the plunger head of the micro physiometer. Oxygen determination based on quenched luminescence was used by another group for a respirometric assay [49]. Since this assay is performed in sealed micro-chambers with 3 µl volume, the sensitivity is reported to exceed considerably the sensitivity of former optical respiration assays.

4.1.3

Miscellaneous Metabolic Parameters

A further detection concept for assaying cellular metabolism is directed to the measurement of metabolic heat with micro-machined planar calorimetric sensors. The transducer principle is based on the Seebeck effect. The material combinations employed (e.g., p+ polysilicon and aluminium) have a high Seebeck coefficient. The sensitivity is further increased by the arrangement of a high number of thermocouples in series to form a thermopile structure and by limiting the heat conduction between thermopile junctions which is achieved by etching thin membraneous conductors. If a cell culture is grown on one side, metabolic heat will cause a temperature difference and thus a measurable voltage difference [50,51]. Obviously there are other approaches employing (amperometric) sensors for metabolic species such as glucose for cell based assays [52]. Such efforts resulted in the fabrication of a micro-fluidic system with an integrated microbioreactor which is reported to yield glucose consumption and lactate release rates of hepatocyte cell lines. For detection a fiber optic equipment was connected to the chip and a NAD+-coupled assay was performed [53]. Multi-analyte sensor systems with planar electrochemical glucose and lactate detection have also been developed which could be employed in vitro for cell based assays [54].

286

4.2


Cell Morphology

Further critical parameters in cell culture that can be assessed with sensors on chips are the number and the growth of viable cells as well as changes in cell adhesion or cell morphology. In the past impedance measurements proved to be applicable well for the analysis of a variety of such parameters, including cell growth, cell attachment and spreading, cell motility and cell migration, barrier function of confluent layers, cell substratum spacing and cell death [55,56,57,58,59,60,61,62]. In a recent report it was even shown that a glucose sensing device with a rather linear response up to 14 mM glucose could be realized based on an impedance sensor structure grown with fibroblasts [63]. In all these configurations adherent cells were cultured directly on electrode structures and a low amplitude alternating current in a frequency range of about 10–20 kHz was applied. The measured electric impedance is the ratio of a sinusoidal voltage applied to a pair of electrodes to the sinusoidal component of the current. Unless the system is purely resistive, the impedance is a complex value because the current and voltage have different phase angles. The cytoplasmic membranes of living cells are effective insulators for alternating current at the frequency used. The measured impedance values reflect the process of cell spreading/cell adhesion and subtle rearrangements of the cytoskeleton which is linked to cell–cell and cell–matrix junctions. For example, changes in intra–cellular calcium ion concentration are known to alter the structure of the cell cytoskeleton, resulting in morphological changes that can be detected by impedance sensors. Fig. 9-13 illustrates the experimental setup whilst Fig. 914 shows an exemplary experiment with the recordings of two sensors (inter-digitated electrode structures) on the same glass chip, detecting the morphological response of a confluent monolayer of Hela cells upon stimulation of the histamine (H1-) receptor. As the most simple equivalent model for the system composed of electrode, cell layer, and cell culture medium a circuit with a resistor and a capacitor in parallel can be selected. The results can be represented either as a pair of Capacitance Cparr and Resistance Rparr or as a pair of the absolute value of the impedance |Z| and the phase angle. Although refined equivalent


287

Figure 9-12. Geometry of IDES (left side) and IDES based impedance measurement on adherent cell cultures (right side). The insulating properties of cell membranes increase the measured electric impedance in the frequency range used (10 kHz). If the cell morphology changes upon drug stimulation this change is usually reflected by a change in impedance.

Figure 9-13. Effect of histamine on the capacitive component of the IDES impedance. The glass chip has two identical IDES, both were grown with a confluent monolayer of Hela tumor cells. Although no distinct morphological alteration can be detected with the microscope, a reversible effect occurs, possibly on cell adhesion. The insert diagram shows the bi–phasic response of impedance in detail. Histamine does not cause measurable effects neither on extra–cellular acidification nor on cell respiration (data not shown). At the end of the experiment the cells were killed with 0.1% Triton X-100.

288


circuits of the cell–electrode interface can be designed, the value of such models for the interpretation of measured impedance data is limited and depends on additional experimental information [64]. A completely different transducer type which has been used for the detection of morphological changes of adherent cells is a piezoelectric quartz crystal micro balance: Perturbations of the cytoskeleton of endothelial cell layers are reported to cause a shift in resonance frequency and resonance admittance [65,66].

4.3

Electrical Patterns

Inspired by the pioneering work of Gross, G. [67], several groups started to use multi-electrode arrays on planar substrates for the multi-site analysis of activity patterns of explanted networks of neuron cells or muscle cell cultures [68,69,71,71,72,73,74]. For the first time it was feasible to address experimentally questions such as how the nature of a neuron’s connections with its neighbours affects its ability to generate action potentials. Arrays with different size, spacing and geometry of electrodes are meanwhile offered commercially (even arrays with three-dimensional protruding electrodes are available). The arrays are produced on glass using varying photo-lithographic techniques. Cultures of explanted primary cells, as well as whole brain tissue slices [75], can be placed on the arrays to monitor spontaneous or micro-electrically stimulated electro-physiological activity with remarkable sensitivity. Sampling rates of at least 10 kHz are necessary for a sufficient time resolution. However, the network of connections between neighbouring neurons is typically so complex that it is difficult to assign the inputs of any given cell. In order to overcome this problem attempts are made to guide neurons on the chip surface by patterning organic compounds attracting or repelling the cells [76]. Micro electrode arrays were also utilized to set up a differential cell based biosensor with both genetically engineered and wild type of mammalian cells of the same type [77]. Starting with the work of Fromherz, who successfully recorded electrical activity of a leech neuron by placing it on top of a silicon field effect transistor [78], the use of ISFETs for electropyhsiological measurements is now emerging [79,80,81]. The noise level of measurements, however, seems to be consistently higher compared to micro electrodes. More recently the fabrication of an array of 16,384 sensors on a single silicon chip was published, allowing records of cellular electric activity with high spatiotemporal resolution (www.infineon.com/bioscience).


289

Figure 9-14. ‘Neurochip’ layout with FETs and micro electrodes. This layout was designed to test both transducers simultaneously with a single neuron network for their ability to detect cellular electric activity.

The method preferred for many electrophysiological experiments is patch clamp. For many years this technique involving highly sophisticated mechanical setups was left to trained personal. Recently the invention of planar glass chips (patch clamp chips) with µm-apertures paved the way for a simplified instrumentation, for automation and thus for considerable increases of experimental throughput in electro physiological screening [82, 83,84]. Cells are drawn to the small apertures in glass chips (e.g., formed with laser ablation) by applying a controlled negative pressure from beneath the glass. This is followed by the development of a very high shunt


290

Figure 9-15. Living neuron (phase contrast microscopy) near the terminal of a cruciform, thin film recording electrode. Indium–tin oxide conductors are 8 µm wide (picture with friendly permission from G.W. Gross).

resistance (‘gigaseal’), the prerequisite for successful measurements. Using a more advanced and costly technology, micro nozzles can be structured in silicon substrates for the immobilization and electrical characterization of cells [85]. The focus of patch clamp technology is the characterization of ion channels. Malfunctions of ion channels are involved in the molecular pathophysiology of many diseases. These channels are, therefore, most important targets of pharmaceutical drug screening. Table 9-1. Current methods for cell based assays on chips and a selection of relevant publications Parameter extra–cellular acidification

Technology ISFETs

LAPS

optic sensors

Examples, selected References (1) Baumann, W. H., et al., (1999), ‘Microelectronic sensor system for micro physiological application on living cells’, Sens. Act. B 55, 77-89 (2) Martinoia S., Rosso, N., Grattarola, M., Lorenzelli, L., Margesin B., Zen, M., (2001), ‘Development of ISFET array based microsystems for bioelectrochemical measurements of cell populations’, Biosens Bioelectron 16, 1043–50 (1) McConnell, H. M., Owicki, J. C., Parce, J. W., Miller, D. L., Baxter, G.T., Wada, H. G., Pitchford, S., (1992), ‘The cytosensor micro physiometer: biological applications of silicon technology’, Science, 257, 1906–12 (2) Metzger, R., et al., (2001), ‘Towards in vitro prediction of an in vivo cytostatic response of human tumor cells with a fast chemo sensitivity assay’, Toxicology, 166, 97–108 (1) Erxleben, H.A., Manion, M.K., Hockenbery, D.M., Scampavia, L., Ruzicka, J., (2004), ‘A novel approach for monitoring extra–cellular acidification f rates: based on bead injection spectro photometry and the lab on valve system’, The Analyst, 129, 205–12


291

Parameter

Technology

Examples, selected References

oxygen consumption

amperometric sensors

(1) Amano, Y., et al., (1998), ‘Measuring Respiration of Cultured Cell with Oxygen Electrode as a Metabolic Indicator for Drug Screening’, Hum. Cell 12, 3–10 (2) Brischwein, M., Motrescu, E.R., Otto, A.M., Cabala, E., Grothe, H., Wolf, B., (2003), ‘Functional Cellular Assays with Multi-parametric Silicon Sensor Chips’, Lab-on-a-chip, 3, 234– 240 (1) Alderman, J., Hynes, J., Floyd, S.M., Krüger, J., O’Connor, R., Papkovsky, D.B., (2004), ‘A low-volume platform for cellrespirometric screening based on quenched–luminescence oxygen sensing’, Biosensors and Bioelectronics, 19, 1529–35 (1) Schulz, C.M., Scampavia, L., Ruzicka, J., (2002), ‘Real time monitoring of lactate extrusion and glucose consumption of cultured cells using a lab on valve system’, The Analyst, 127, 1583–88 (1) Ehret, R., et al., (1997), ‘Monitoring of cellular behaviour by impedance measurements on interdigitated electrode structures’, Biosensors & Bioelectronics, 12, 29–41 (2) Wegener, J., et al., (2000), ‘Electric Cell-Substrate Impedance Sensing (ECIS) as a Non invasive Means to Monitor the Kinetics of Cell Spreading to Artificial Surfaces’, Exp. Cell. Res., 259, 158–166 (1) Fertig, N., et al., (2002), ‘Whole Cell Patch Clamp Recording Performed on a Planar Glass Chip’, Biophysical Journal 82, 3056–62 (2) Stett, A., Bucher, V., Burkhardt, C., Weber, U., Nisch, W., (2003), ‘Patch–clamping of primary cardiac cells with microopenings in polyimide films’, Medical & Biological Engineering & Computing, 41, 233–240 (1) Gross, G.W., et al., (1995), ‘The use of neuronal netwoks on multi-electrode arrays as biosensors’, Biosensors & Bioelectronics, 10, 553–567 (2) Egert, U., et al., (1998): A novel organotypic long term culture of the rat hippocampus on substrate–integrated multi-electrode arrays. Brain Res Prot 2, 229–242 (1) Ingebrandt, S., et al., (2001): Cardiomyocyte-transistor hybrids for sensor application. Biosensors & Bioelectronics, 16, 565–570 (2) www.infineon.com/bioscience (3) Stepper, C., Wolf, B., Wiest, J.., Loeser, M., Brischwein, M., Grothe, H., Hansch, W., Schmitt-Landsiedel, D., (2003), ‘Technological Pre-Investigations on the Realization of a Local Resolution Microphysiologic Cell Chip for Medical Diagnostics and Pharmaceutical Screening’, Proceedings of Sensor 2003 11th International Conference, 13–15 May 2003, Nürnberg, Germany. Part II, 335–338

optic sensors

glucose/lactate exchange

optic sensors

changes in cellular morphology

electric impedance sensors

electric activity

patch–clamp chips

micro electrode arrays

FET arrays

292

5.


CELL MANIPULATION ON CHIPS

Engineering advances in micro-fluidics have allowed the construction of versatile devices transporting substances through micro channels of a glass or plastic chip by electro-osmosis, capillary forces, or pressure. Both bacteria and eukaryotic cells have been successfully transported in microfluidic systems [25]. The number of publications describing different types of cell manipulations, including complex multi-step procedures, such as cell separation [86,87,88], cell fusion [89,90], electroporation [91], cell lysis [88,92], and incubation in different media performed [7] in such devices is rapidly increasing. A recent review discusses the use of micro-technologies as a very useful tool for cell manipulation and analysis [93]. Micro-fluidic structures on chips for cell manipulations are frequently combined with analytical functions in order to realize cellular assays. These assays include both a chemical analysis of cell constituents after cell lysis [90,94,95] and the analysis of intact cells, the latter most often based on labeling with specific fluorescent dyes and read out with a fluorescence detector [7,96]. For cell counting the principle of cytometry, focusing cells into a single line for individual analysis has been adapted to micro-fluidic devices by different groups [97,98,99,100]. Commercially available products are already utilized for cell staining and cell analysis on chip systems [101]. Versatile micro-fabricated flow cells as a basis for applications in cell analysis have been described by the group of Vellekoop [102]. Unlike the Coulter Counter principle which relies on measuring the volume displaced by passing cells, the polarization response of cells passing through the channel of an integrated micro-fluidic device in an electric field region was reported to give information about the DNA content of the cell [103]. A typical application of cell chips is the detection and quantification of viable bacteria in aqueous solutions with electrode structures. In the present chip setup (Fig. 9-16) an inter-digitated electrode structure is used both to accumulate the bacteria in their vicinity using dielectrophoretic forces between the cells and the electrodes (prior to the measurement) and to detect the bacteria by impedance measurement. The grade of accumulation can be changed by the voltage applied to the sensor electrodes. Fig. 9-17 illustrates the accumulation of E. coli cells in regions with high electric field gradients at 2 V and 10 kHz. Varying the voltage causes a change in impedance which depends also on the concentration of the bacteria.


293

Figure 9-16. E.coli bacteria, stained with a fluorescent dye, are attracted to the electrodes by dielectrophoretic forces. Typical values for voltage and frequency are 1–2 Volts and 10–100 kHz.

There are numerous examples for the use of dielectrophoretic forces in lab-on-a-chip devices for cell separations and cell trapping [104]. Using various types of electric fields it is possible to move, separate, fuse, perforate, or deform cells. Such methods are broadly applied in biotechnology. In contrast to the effects of weak electric and electromagnetic fields, in these techniques rather strong fields are used, with an energy input significantly larger than the energy of thermal noise. For example, dielectrophoretic field cages can be created with electrode structures for the handling of single cells on chips. In addition, such field cages are useful for a dielectrical characterization of cells by electrorotation [105]. To avoid various drawbacks associated with dielectrophoresis (such as unwanted electrochemical effects), ‘electrodeless’ devices have been described [106, 107], creating a constriction and thus a high gradient of an electric field in a conductive solution. Another non-contact method for cell manipulation is based on superimposed magnetic AC-fields, acting on cells labeled with magnetic nanoparticles and independent of any material constants of the extra–cellular liquid [108]. For many applications the use of mathematical methods is necessary in order to obtain optimal electrode geometries. For example, a glass chip was

294


constructed with a novel sensor structure for measuring the concentration of bacteria in an aqueous solution using dielectrophoretic forces. The four sensor electrodes have a sandwich structure with two similarly designed electrode layers separated by an isolating layer. In each layer two similarly modeled electrodes are arranged in two insulated layers. The electrodes of one single layer are active for the collection mode of the sensor. In the measuring mode, two of the back–to–back lying electrodes are used to measure the concentration dependent impedance. The collecting property can be achieved by a combination of high electric field strength and a large field gradient. A minimal capacity is necessary for a high sensitivity of the sensor. The electrode structure was therefore optimized by a threedimensional numerical field calculation. With this program the dielectric properties of the glass substrate, the electrodes, and the aqueous solution were simulated. The result of optimization was an electrode arrangement consisting of one even and two zigzag electrodes (Fig. 9-17).

Figure 9-17. Electric field distribution in a distance of 20 µm above the electrode surface. These data were used to optimize the electrode geometry.


6.

295

CONCLUSIONS AND FUTURE PROSPECTS

Micro- and nano-technologies are rapidly expanding into biomedical applications. Amongst these applications live cell studies are becoming increasingly important not only in pharmaceutical drug discovery but also in clinical diagnostics and monitoring of environmental toxicants and pathogens. There is a considerable demand for practical and low cost tools for cell studies in all these areas. Cells are the minimum functional and communicating unit of any living system and the ultimate target of any drug. Although cellular signal amplification mechanisms often result in impressive cellular responses, such ‘output signals’ are not easily detected with conventional analytical methods, at least in small cell numbers. A substantial aim of micro- and nano-technology is to integrate the ‘microsystem’ of the living cell with technical microsystems providing the necessary physio-logical environment, the micro-fluidic systems for controlled supply of media and drug solutions, and transducers for a sensitive recording of subtle changes in cellular behavior. If not only a short term cell culture is involved, it will be important for the lab-on-a-chip system to create an in vitro cellular micro environment as close as possible to in vivo conditions. Groups working with cell based assays are frequently facing problems with gradual cell culture deterioration in the time course of the studies. Therefore conditions including the protection of cell culture media from evaporation, adjustment of the gas composition and measures against contamination must be guaranteed. The structural versatility of micro-fluidic systems should help to meet all these requirements. Chip devices integrating micro bioreactors, micro-fluidics, and transducer functions are expected to emerge within the next few years. Research and development in cell based assays strongly focuses on detection principles based on fluorescence optics. This is owed to a steady improvement of live cell stains directed to a great variety of cellular targets and goes along with a similar progress in instrumental technology adapted to a very versatile readout of fluorescence parameters. On the other hand, the equipment for optical detection (optic fibers, microscope lenses) sometimes appears to attenuate the advantages of small and cheap cell chips. Compromises are obviously necessary. Moreover, photobleaching and (photo-)toxic properties of virtually any dye counteract a monitoring of cells on a time scale longer than a few minutes.

296


With microelectric transducers special attention has to be paid to a practical solution for electric contacting and to a condensation and preprocessing of the generated data, particularly with respect to multi-parameter sensor chips and high density plates. To achieve this aim substantial technological efforts in chip processing are necessary: one of the most important benefits of CMOS technology is the possibility of on-chip circuitry for signal amplification, data analysis and sensor self-testing. Onchip sensor multiplexing is a precondition for the construction of 96- or 384multi-well arrays since the number of necessary electric connections becomes unmanageable. Although single-parameter assays are reasonable for many purposes, it would increase the efficacy of cell studies if it were not necessary to combine data from different assays which might have been performed under slightly different experimental conditions. Multi-parameter cell chips would help to reveal various aspects of cellular events within a single assay, and they would do so in a dynamic and real time mode. With glass chips the combination of electric sensors with optic sensors, e.g., for pH and oxygen, is obvious. If required, even high resolution light microscopy providing imaging information can be combined with on-line sensor monitoring of the cells. Although miniaturization of cell chips has its advantages by saving valuable (primary) cells and drug compounds, and although the size of single sensor elements on chips can be as small as about 100 µm2 (e.g., the sensitive gate area of a pH-ISFET or a micro electrode for the extra–cellular detection of action potentials), it is mostly not intended to analyze single cells with electric transducers. It should be emphasized that single-cell measurements which do not take into account the typical social context of cells must sometimes be assessed critically. This fact is highlighted by neuronal networks on micro electrode arrays which begin to develop after explantation and which can only be reasonably assessed as whole functional units. Nevertheless, a direct analysis of single cells on chips which does not involve long term cell culturing is possible, and even a direct chemical analysis of single cells (e.g., cell lysis followed by PCR) is expected to emerge in the next years [13].


297

In metabolic assays efforts dedicated to metabolic imaging using highly integrated two-dimensional chips arrays are only in their beginnings. G enerally, the strength of metabolic (and morphological) assays is their unspecificity, allowing us to detect a great variety of cellular responses. This is because cell metabolism and cell morphology itself are closely coupled to the signaling apparatus and can thus be regarded as first step signal transducers. F or the interpretation of results a multi-parameter and kinetic analysis may provide first indications on a drug’s mode of action. F or interpretations at a molecular level, methods such as the comparison of different transformed cell lines or the use of specifically acting biochemical inhibitors or antibodies have to be applied. Thus it is likely that cell chip systems will find their place in the screening of pharmaceutical agents for activity or toxicity, as an additional source of information about cellular behavior. An example for a multi-parameter record of tumor cells treated with a cytotoxic drug is given in Fig. 9-18.

Figure 9-18. Effect of chloroacetaldehyde (50 µM) on LS 174 T cell cultures (human colon adenocarcinoma cell line, two cultures have been run in parallel), monitored with pH-, oxygen and impedance sensors. The drug was added twice for one hour to test for reversibility of drug effects. A strong, but reversible, effect on cell respiration is observed. Impedance measurements, however, reveal that cell death is not the predominant effect (cell death would be accompanied by early morphologic changes and cell attachment, detectable with IDES). At the end of the experiment the cells were killed with 0.1% Triton X-100.

298

LAB-ON-A-CHIP SYSTEMS FOR CELLULAR ASSAYS Y REFERENCES

[1] [2] [3] [4] [5] [6] [7]

[8] [9]

[10] [11] [12] [13] [14]

Davidson, M., Karlsson, M., Sinclair J., Scott, K., Orwar, O., (2002), ‘Nanotubevesicle networks with functionalized membranes and interiors’, Journal of the American Chemical Society 125, 374–78. Kraus, M., Wolf, B., (1995), ‘Structured Biological Modelling: A New Approach to Biophysical Cell Biology’, CRC Press, Inc., Boca Raton. Hartwell, L.H., Hopfield, J.J., Leibler, S., Murray, A.W., (1999) From molecular to modular cell biology’, Nature 402, C47–C52. Dumont, E.D., Pécasse, F., Maenhaut, C., (2001), ‘Crosstalk and specificity in signalling. Are we crosstalking ourselves into general confusion?’, Cellular Signalling 13, 4457–463. Taylor, D.L., Woo, E.S., Giuliano, K.A., (2001), ‘Real–time molecular and cellular analysis: the new frontier of drug discovery’, Current Opinion in Biotechnology 12, 75–81. Rädler, U., ‘The µ-slide for live cell imaging—combining cell biology with optical microscopy . American Biotechnology Laboratory 2004, accepted. Yang, M., Li, C., Yang, J., (2002), ‘Cell Docking and On-Chip Monitoring of Cellular Reactions with a Controlled Concentration Gradient on a Micro-fluidic device’, Analytical Chemistry 74, 3991–4001. Gross, G.W., Schwalm, F.U., (1994), ‘A closed flow chamber for long term multi-channel recording and optical monitoring’, Journal of Neuroscience Methods 52, 73–85. Wolf, B.W., Brischwein, M., Baumann, W., Ehret, R., Kraus, M., (1998), ‘Monitoring of cellular signalling and metabolism with modular sensor-technique: The PhysioControl-Micro system (PCM)’, Biosensors and Bioelectronics 13, 501–09. Cooper, J.M., (1999), ‘Toward electronic Petri Dishes and picolitrescale single cell technologies’, Trends in Biotechnology 17, 226–230. Hediger, S., Sayah, A., Horisberger, J.D., Gijs, M.A.M., (2001), ‘Modular micro system for epithelial cell culture and electrical characterisation’, Biosensors and Bioelectronics 16, 689–94. Park, T.H., Shuler, M.L., (2003), ‘Integration of Cell Culture and Micro-fabrication Technology’. Biotechnology Progress 19, 243–53. Anderson, H., van den Berg, A., (2004), ‘Micro-technologies and nanotechnologies for single-cell analysis’, Current Opinion in Biotechnology 15, 44–49. Ekelund, S., Nygren, P., Larsson, R., (1998), ‘Microphysiolmetry: new technology for evaluation of anticancer drug activity in human tumor cells in vitro’, Anti-Cancer Drugs 9, 531–38.

LAB-ON-A-CHIP SYSTEMS FOR CELLULAR ASSAYS [15] [16]

[17]

[18] [19] [20]

[21]

[22]

[23] [24] [25] [26]

299

Hafner, F., (2000), ‘Cytosensor Microphysiometer: technology and recent applications’, Biosensors and Bioelectronics, 15, 149–58. Metzger, R., Deglmann, C.J., Hoerrlein, S., Zapf, S., Hilfrich, J., (2001), ‘Towards in vitro prediction of an in vivo cytostatic response of human tumor cells with a fast chemosensitivity assay’, Toxicology, 166, 97–108. Schubnell, D., Lehmann, M., Baumann, W. , Rott, FG, Wolf, B., Beck, C.F., (1999), ‘An ISFET-algal (Chlamydomonas) hybrid provides a system for eco-toxicological tests’, Biosensors and Bioelectronics, 14, 465–73. DeBusschere, B.D., Kovacs, G.T.A., (2001), ‘Portable cell based biosensor system using integrated CMOS cell-cartridges’, Biosensors and Bioelectronics 16, 543–56. Rudolph, A.S., Reasor, J., (2001), ‘Cell and tissue based technologies for environmental detection and medical diagnostics’, Biosensors and Bioelectronics 16, 429–31. Pancrazio, J.J., Gray, S., Shubin, Y.S., Kulagina, N., Cuttino, D.S., Shaffer, K.M., Eisemann, K., Curran, A., Zim, B., Gross, G.W., O’Shaughnessy, T.J., (2003), ‘A portable micro electrode array recording system incorporating neuronal networks for neurotoxin detection’, Biosensors and Bioelectronics 18, 1339–47. Ressler, J., Grothe, H., Motrescu, E.R., Wolf, B., (2004), ‘New concepts for chip-supported multi-well-plates: Realization of a 24well-plate with integrated impedance-sensors for functional cellular screening applications and automated microscope aided cell based assays’, Proceedings of the 26th Annual International Conference of the IEEE EMBS., SanFrancisco, , 2074–2077. Owicki, JC., Bousse, L.J., Hafeman, D.G., Kirk, G.L., Olson, J.D., Wada, H.G., Parce, J.W., (1994), ‘The Light-Addressable Potentiometric Sensor: Principles and Biological Applications’, Annual Reviews of Biophysics and Biomolecular Structures 23, 87–114. Curtis, A., Wilkinson, C., (2001), ‘Nanotechniques and approaches in biotechnology’, Trends in Biotechnology 19, 97–101. Koh, W., Itle, L.J., Pishko, M.V., (2003), ‘Molding of hydrogel micro-structures to create multi-phenotype cell microarrays’, Analytical Chemistry 75, 5783–89. Weigl, B.H., Bardell, R.L., Cabrera, C.R., (2003), ‘Lab-on-a-chip for drug development’, Advanced Drug Delivery Reviews 55, 349–77. Walker, G.M., Zeringue, H.C., Beebe, D.J., (2004), ‘Microenvironment design considerations for cellular scale studies’, Lab Chip 4, 91–97.

300


[27]

Bousse, L., Cohen, C., Nikiforov, T., Chow, A., Kopf-Sill, A.R., Dubrow, R., Parce, J.W., (2000), ‘Electro-kinetically Controlled Micro-fluidics Analysis Systems’, Annuals Reviews of Biophysics and Biomolecular Structure 29, 155–181. Wheeler, A.R., Throndset, W.R., Whelan, R.J., Leach, A.M., Zare, R.N., Liao, Y, Farrell, K., Manger, I.D., Daridon, A., (2003), ‘Microfluidic device for Single-Cell Analysis’, Analytical Chemistry 75, 3581–86. Farinas, J., Chow, A.W., Wada, H.G., (2001), ‘A Micro-fluidic device for Measuring Cellular Membrane Potential’, Analytical Biochemistry 295, 138–42. Bloom, FR, Price, P., Lao, G., Xia, J.L., Crowe, J.H., Battista, J.R., Helm, R.F., Slaughter, S., Potts, M., (2001), ‘Engineering mammalian cells for solid-state sensor applications’, Biosensors and Bioelectronics 16, 603–08. Brischwein, M., Motrescu, E.R., Otto, A.M., Cabala, E., Grothe, H., Wolf, B. (2003) Functional Cellular Assays with Multi-parametric Silicon Sensor Chips’, Lab Chip 3, 234–240. Lorenzelli, L., Margesin, B., Martinoia, S., Tedesco, M.T., Valle, M., (2003), ‘Bioelectrochemical signal monitoring of in vitro cultured cells by means of an automated micro system based on solid state sensor-array’, Biosensors and Bioelectronics 18, 621–26. Baumann, W., Lehmann, M., Schwinde, A., Ehret, R., Brischwein, M., Wolf, B., (1999), ‘Microelectronic sensor system for micro physiological application on living cells’, Sensors and Actuators, B55, 77–89. Martinoia, S., Rosso, N., Grattarola, M., Lorenzelli, L., Margesin, B., Zen, M., (2001), ‘Development of ISFET array based micro systems for bioelectrochemical measurements of cell populations’, Biosensors and Bioelectronics 16, 1043–50. Owicki, J.C., Parce, J.W., (1992), ‘Biosensors based on the energy metabolism of living cells: the physical chemistry and cell biology of extra–cellular acidification’, Biosensors and Bioelectronics 7, 255–72. Fanigliulo, A., Accossato, P., Adami, M., Lanzi, M., Martinoia, S., Paddeu, S., Parodi, M.T., Rossi, A., Sartore, M., Grattarola, M., Nicolini, C., (1996), ‘Comparison between a LAPS and a FET based sensor for cell-metabolism detection’, Sensors and Actuators, B32, 41–48. George M., Parak W.J., Gaub H.E., (2000), ‘Highly integrated surface potential sensors’, Sensors and Actuators B69, 266–75.

[28]

[29] [30]

[31] [32]

[33]

[34]

[35]

[36]

[37]

LAB-ON-A-CHIP SYSTEMS FOR CELLULAR ASSAYS [38]

[39] [40]

[41] [42]

[43]

[44] [45]

[46] [47]

301

Stepper, C., Wolf, B., Wiest, J., Loeser, M., Brischwein, M., Grothe, H., Hansch, W., Schmitt-Landsiedel, D., (2003), ‘Technological Pre-Investigations on the Realization of a Local Resolution Microphysiologic Cell Chip for Medical Diagnostics and Pharmaceutical Screening’, Proceedings of Sensor 2003 11th International Conference, Nürnberg, Germany. Part II, 335–338. Lähdesmäki, I., Ruzicka, J., Ivaska, A., (2000), ‘Novel flow injection methods for drug-receptor interaction studies, based on probing cell metabolism’, The Analyst 125, 1889–95. Erxleben, H.A., Manion, M.K., Hockenbery, D.M., Scampavia, L., Ruzicka, J., (2004), ‘A novel approach for monitoring extra–cellular acidification rates: based on bead injection spectrophotometry and the lab-on-valve system’, The Analyst 129, 205–12. Schlage W. K., Bereiter-Hahn, J., (1983), ‘A microscope perfusion respirometer for continuous respiration measurement of cultured cells during microscopic observation’, Microscopica Acta 87, 19–34. Ruffieux, P.A., von Stockar, U., Marison, I.W., (1988), ‘measurement of volumetric (OUR) and dedermination of specific (qO2) oxygen uptake rates in animal cell cultures’, Journal of Biotechnology 63, 85–95. Casciari, J.J., Sotirchos, S.V., Sutherland, R.M., (1992), ‘Variations in tumor cell growth rates and metabolism with oxygen concentration, glucose concentration and extra–cellular pH’, Journal of Cellular Physiology 151, 386–94. Lambrechts, M., Sansen, W., (1992), ‘Biosensors: Microelectrochemical Devices’, IOP Publishing Ltd., Bristol. Amano, Y., Okumura, C., Yoshida, M., Katayama, H., Unten, S., Arai, J., Tagawa, T., Hoshina, S., Hashimoto, H., Ishikawa, H. (1999), ‘Measuring Respiration of Cultured Cell with Oxygen Electrode as a Metabolic Indicator for Drug Screening’, Human Cell 12, 3–10. Sohn, B., Kim, C., (1996), ‘A new pH-ISFET based dissolved oxygen sensor by employing electrolysis of oxygen’, Sensors and Actuators B34, 435–40. Lehmann, M., Baumann, W., Brischwein, M., Gahle, H.J., Freund, I., Ehret, R., Drechsler, S., Palzer, H., Kleintges, M., Sieben, U., Wolf, B., (2001), ‘Simultaneous measurement of cellular respiration and acidification with single CMOS ISFET’, Biosensors and Bioelectronics 16, 195–203.

302


[48]

Ekelund, S., Cliffel, D.E., Kozlov, E., Prokop, A., Wikswo, J., Baudenbacher, F., (2003), ‘Modification of the Cytosensor™ microphysiometer to simultaneously measure extra–cellular acidification and oxygen consumption rates’, Analytica Chimica Acta 496, 93–101. Alderman, J., Hynes, J., Floyd, S.M., Krüger, J., O’Connor, R., Papkovsky, D.B., (2004), ‘A low-volume platform for cellrespirometric screening based on quenched-luminescence oxygen sensing’, Biosensors and Bioelectronics 19, 1529–35. Verhaegen, K., Simaels, J., van Driessche, W., Baert, K., Sansen, W., Puers, B., Hermans, L., Mertens, R., (1999), ‘A Biomedical Microphysiometer’, Biomedical Micro-devices 2, 93–98. Johannessen, E.A., Weaver, J.M.R., Bourova, L., Svoboda, P., Cobbold, P.H., Cooper, J., (2002), ‘Micro-machined Nanocalorimetric Sensor for Ultra-Low-Volume Cell based Assays’, Analytical Chemistry 74, 2190–97. von Woedtke, T., Jülich, W.D., Alhitari, N., Hanschke, R., Abel, P.U., (2002), ‘Biosensor-controlled perfusion culture to estimate the viability of cells’, Medical & Biological Engineering & Computing 40, 704–11. Schulz, C.M., Scampavia, L., Ruzicka, J., (2002), ‘Real-time monitoring of lactate extrusion and glucose consumption of cultured cells using a lab-on-valve system’, The Analyst 127, 1583–88. Petrou, P.S., Moser, I., Jobst, G., (2003), ‘Micro-device with integrated dialysis probe and biosensor array for continuous multianalyte monitoring’, Biosensors and Bioelectronics 18, 613–19. Giaever, I., Keese C.R., (1993), ‘A morphological biosensor for mammalian cells’, Nature 366, 591–92. Ehret, R., Baumann, W., Brischwein, M., Schwinde, A., Stegbauer, K., Wolf, B., (1997), ‘Monitoring of cellular behaviour by impedance measurements on interdigitated electrode structures’, Biosensors and Bioelectronics 12, 29–41. Ehret R., Baumann W., Brischwein M., Schwinde, A., Wolf, B., (1998), ‘On-line control of cellular adhesion with impedance measurements using interdigitated electrode structures’, Medical & Biological Engineering & Computing 36, 365–70. Wegener, J., Keese, C.R., Giaever, I., (2000), ‘Electric Cell–Substrate Impedance Sensing (ECIS) as a Noninvasive Means to Monitor the Kinetics of Cell Spreading to Artificial Surfaces’, Experimental Cell Research 259, 158–66.

[49]

[50] [51]

[52]

[53] [54] [55] [56]

[57]

[58]

LAB-ON-A-CHIP SYSTEMS FOR CELLULAR ASSAYS [59]

[60]

[61] [62]

[63]

[64]

[65] [66]

[67] [68] [69] [70]

303

Jager, E.W.H., Immerstrand, C., Peterson, K.H., Magnusson, K., Lundström, I., Inganäs, O., (2002), ‘The Cell Clinic: Closable microvials for Single Cell Studies’, Biomedical Micro-devices 4, 177–87. Luong, J.H., Habibi-Rezaei, M., Meghrous, J., Xiao, C., Male, K.B., Kamen, A., (2001), ‘Monitoring Motility, Spreading, and Mortality of Adherent Insect Cells Using an Impedance Sensor’, Analytical Chemistry 73, 1844–48. Arndt, S., Seebach, J., Psathaki, K., Galla, H.-J., Wegener, J., (2004), ‘Bioelectrical impedance assay to monitor changes in cell shape during apoptosis’, Biosensors and Bioelectronics 19, 583–94. Yang, L., Li, Y., Griffis, C.L., Johnson, M.G., (2004), ‘Interdigitated micro electrode (IME) impedance sensor for the detection of viable Salmonella typhimurium’, Biosensors and Bioelectronics 19, 1139–47. Tlili, C., Reybier, K., Geloen, A. Ponsonnet, L., Martelet, C., Ben Ouada, H., Lagarde, M., Jaffrezic-Renault, N., (2003), ‘Fibroblast Cells: A Sensing Bioelement for Glucose Detection by Impedance Spectroscopy’, Analytical Chemistry 75, 3340–44. Wegener, J., Zink, S., Rösen, P., Galla, H.J., (1999), ‘Use of electrochemical impedance measurements to monitor ȕ-adrenergic stimulation of bovine aortic endothelial cells’, European Journal of Physiology 437, 925–34. Wegener, J., Seebach, J., Janshoff, A., Galla, H.J., (2000), ‘Analysis of the Composite Response of Shear Wave Resonators to the Attachment of Mammalian Cells’, Biophysical Journal 78, 2821–33. Marx, K.A., Zhou, T., Montrone, A., Schulze, H., Braunhut, S., (2001), ‘A quartz crystal microbalance cell biosensor: detection of microtubule alterations in living cells at nM nocodazole concentrations’, Biosensors and Bioelectronics 16, 773–82. Gross, G.W., (1979), ‘Simultaneous single unit recording in vitro with a photoetched deinsulated gold micro electrode surface’, IEEE Transactions on Biomedical Engineering 26, 273–77. Pine, J., (1980), ‘Recording action potentials from cultured neurons with extra–cellular microcircuit electrodes’, Journal of Neuroscience Methods 2, 19–28. Hämmerle, H., Egert, U., Mohr, A., Nisch, W., (1994), ‘Extra– cellular recording in neuronal networks with substrate integrated micro electrode arrays’, Biosensors and Bioelectronics 9, 691–6. Hofer, E., Urban, G, Spach, M.S., Schafferhofer, I., Mohr, G., Platzer, D., (1994), ‘Measuring activation patterns of the heart at a microscopic size scale with thin film sensors’, American Journal of Physiology 266, H2136–45.

304


[71]

Blau, A., Ziegler, C., Heyer, M., Endres, F., Schwitzgebel, G., Matthies, T., Stieglitz, T., Meyer, J.H., Göpel, W., (1997), ‘Characterization and optimization of micro electrode arrays for in vivo nerve signal recording and stimulation’, Biosensors and Bioelectronics 12, 883–92. Bove, M., Martinoia, S., Verreschi, G., Guigliano, M., Grattarola, M., (1998), ‘Analysis of the signals generated by networks of neurons coupled to planar arrays of microtransducers in simulated experiments’, Biosensors and Bioelectronics 15, 601–12. Stenger, D.A., Gross, G.W., Keefer, E.W., Shaffer, K.M., Andreadis, J.D., Ma, W., Pancrazio, J.J., (2001), ‘Detection of physiologically active compounds using cell based biosensors’, Trends in Biotechnology 19, 304–09. Stett, A., Egert, U., Guenther, E., Hofmann, F., Meyer, T., Nisch, W., Haemmerle, H., (2003), ‘Biological application of micro electrode arrays in drug discovery and basic research’, Analytical and BioAnalytical Chemistry 377, 486–95. Egert, U., Heck, D., Aertsen, A., (2002), ‘Two-dimensional monitoring of spiking networks in acute brain slices’, Experimental Brain Research 142, 268–274. Stenger, D.A., Hickman, J.J., Bateman, K.E., Ravenscroft, M.S., Ma, W., Pancrazio, J.J., Shaffer, K., Schaffner, A.E., Cribbs, D.H., Cotman, C.W., (1998), ‘Microlithographic determination of axonal/ dendritic polarity in cultured hippocampal neurons’, Journal of Neuroscience Methods 82, 167–73. Aravantis, A.M., DeBusschere, B.D., Chruscinski, A.J., Gilchrist, K.H., Kobilka, B.K., Kovacs, G.T.A., (2001), ‘A genetically engineered cell based biosensor for functional classification of agents’, Biosensors and Bioelectronics 16, 571–77. Fromherz, P., Offenhäuser, A., Vetter, T., Weis, J., (1991), ‘A neuron-silicon junction: a Retzius cell of the leech on an insulatedgate field-effect transistor’, Science, 252, 1290–93. Offenhäuser, A., Sprössler, C., Matsuzawa, M., Knoll, W., (1997), ‘Field-effect transistor array for monitoring electrical activity from mammalian neurons in culture’, Biosensors and Bioelectronics 12, 819–26. Offenhäuser, A., Knoll, W., (2001), ‘Cell-transistor hybrid systems and their potential application’, Trends in Biotechnology 19, 62–66. Besl, B., Fromherz, P., (2002), ‘Transistor array with an organotypic brain slice: field potential records and synaptic currents’, European Journal of Neuroscience 15, 999–1005.

[72]

[73]

[74]

[75] [76]

[77]

[78] [79]

[80] [81]

LAB-ON-A-CHIP SYSTEMS FOR CELLULAR ASSAYS [82] [83]

[84]

[85] [86] [87] [88]

[89] [90] [91]

[92] [93] [94]

305

Fertig, N., Blick, R.H., Behrends, J.C., (2002), ‘Whole cell patch clamp recording performed on a planar glass chip’, Biophysical Journal 82, 3056–62. Stett, A., Bucher, V., Burkhardt, C., Weber, U., Nisch, W., (2003), ‘Patch-clamping of primary cardiac cells with micro-openings in polyimide films’, Medical & Biological Engineering & Computing, 41, 233–40. Brueggemann, A., George, M., Klau, M., Beckler, M., Steindl, J., Behrends, J.C., (2004), ‘Ion Channel Drug Discovery and Research: The Automated Nano-Patch-Clamp Technology’, Current Drug Discovery Technologies 1, 91–96. Lehnert, T., Gijs, M.A.M., Netzer, R., Bischoff, U., (2002), ‘Realization of SiO2 micronozzles for electrical measurements on living cells’, Applied Physics Letters 81, 5063–65. Fiedler, S., Shirley, S.G., Schnelle, T., Fuhr, G., (1998), ‘Dielectrophoretic Sorting of Particles and Cells in a Micro system’, Analytical Chemistry 70, 1909–15. Fu, A.Y., Spence, C., Scherer, A., Arnold, F.H., Quake, S.R., (1999) Nature Biotechnology 17, 1109–11. Huang, Y., Joo, S., Duhon, M., Heller, M., Wallace, B., Xu, X., (2002), ‘Dielelctrophoretic cell separation and gene expression profiling on microelectronic chip arrays’, Analytical Chemistry 74, 3362–71. Stromberg, A., Karlsson, A., Ryttsen, F., Davidson, M., Chiu, D.T., Orwar, O., (2001), ‘Micro-fluidic device for Combinatorial Fusion of Liposomes and Cells’, Analytical Chemistry 73, 126–30. Chiu, D.T., (2001), ‘A micro-fluidic platform for cell fusion’, Current Opinion in Chemical Biology 5, 609–12. Huang, Y., Sekhon, N., Borninsky, J., Chen, N., Rubinsky, B., (2003), ‘Instantaneous, quantitative single-cell viability assessment by electrical evaluation of cell membrane integrity with microfabricated devices’, Sensors and Actuators A105, 31–39. Gao, J., Yin, X., Fang, Z., (2004), ‘Integration of single cell injection, cell lysis, separation and detection of intra–cellular constituents on a micro-fluidics chip’, Lab Chip 4, 47–52. Anderson, H., van den Berg, A., (2003), ‘Micro-fluidic devices for cellomics: a review’, Sensors and Actuators B92, 315–25. Schilling, E.A., Kamholz, A.E., Yager, P., (2002), ‘Cell Lysis and Protein Extraction in a Micro-fluidic device with Detection by a Fluorogenic Enzyme Assay’, Analytical Chemistry 74, 1798–1804.

306


[95]

McClain, M.A., Culbertson, C.T., Jacobson, S.C., Allbritton, N.L., Sims, C.E., Ramsey, J.M., (2003), ‘Micro-fluidic devices for the High throughput Chemical Analysis of Cells’, Analytical Chemistry 75, 5646–55. Matsubara, Y., Murakami, Y., Kobayashi, M., Morita, Y., Tamiya, E., (2004), ‘Application of on-chip cell cultures for the detection of allergic response’, Biosensors and Bioelectronics 19, 741–47. Gawad, S., Schild, L., Renaud, P., (2004), ‘Micro-machined impedance spectroscopy flow cytomyter for cell analysis and particle sizing’, Lab Chip 4, 47–52. McClain, M.A., Culbertson, C.T., Jacobson, S.C., Ramsey, J.M., (2001), ‘Flow Cytometry of Escherichia colii on Micro-fluidic devices’, Analytical Chemistry 73, 5334–38. Fu, A.Y., Chou, H., Spence, C., Arnold, F.H., Quake, S.R., (2002), ‘An integrated micro-fabricated cell sorter’, Analytical Chemistry 74, 2451–57. Wolff, A., Perch-Nielsen, I.R., Larsen, U.D., Friis, P., Goranovic, G., Poulsen, C.R., Kutter, J.P., Telleman, P., (2003), ‘Integrating advanced functionality in a micro-fabricated high throughput fluorescent-activated cell sorter’, Lab Chip 3, 22–27. Buhlmann, C., Preckel, T., Chan, S., Luedke, G., Valer, M., (2003), ‘A New Tool for Routine Testing of Cellular Protein Expression: Integration of Cell Staining and Analysis of Protein Expression on a Micro-fluidics Chip based System’, Journal of Biomolecular Techniques 14, 119–27. Nieuwenhuis, J.H., Bastemeijer, J., Sarro, P.M., Vellekoop, M.J., (2003), ‘Integrated flow-cells for novel adjustable sheath flows’, Lab Chip 3, 56–61. Sohn, L.L., Saleh, O.A., Facer, G.R., Beavis, A.J., Allan, R.S., Notterman, D.A., (2000), ‘Capacitance cytometry: measuring biological cells one by one’, Proceedings of the National Academy of Sciences of the United States of America 97, 10687–90. Gambari, R., Borgatti, M., Altomare, L., Manaresi, N., Medoro, G., Romani, A., Tartagni, M., Guerrieri, R., (2003): ‘Applications to Cancer Research of ‘lab-on-a-chip’ devices based on dielectrophoresis (DEP)’, Technology in Cancer Research & Treatment 2, 31–40. Gimsa, J., (2001), ‘A comprehensive approach to electro-orientation, electrodeformation, dielectro-phoresis, and electrorotation of ellipsoidal particles and biological cells’, Bioelectrochemistry 54, 23–31.

[96] [97] [98] [99] [100]

[101]

[102] [103]

[104]

[105]


307

[106] Chou, C.-F., Zenhausern, F., (2003), ‘Electrodeless Dielectrophoresis for Micro-Total Analysis Systems’, IEEE Engineering in Medicine and Biology Magazine 22, 62–67. [107] Cummings, E.B., (2003), ‘Streaming Dielectrophoresis for Continuous-Flow Micro-fluidic devices’, IEEE Engineering in Medicine and Biology Magazine 22, 75–84. [108] Koch, M., Weyh, T., Wolf, B., (2004), ‘Field-controlled motion of sub-nano, nano- and macroparticles’, Proceedings of the Fourth IEEE Conference on Nanotechnology, Munich.

Chapter 10 NETWORKS ON CHIPS Spatial and temporal activity dynamics of functional networks in brain slices and cardiac tissue Ulrich Egert Neurobiology & Biophysics, Institute for Biology III, University of Freiburg, Germany

Abstract:

Cells of the same or different type in biological tissue closely interact and influence each other. Individual cells should therefore be studied in the context of the embedding cellular network. This particularly applies to the highly dynamic electrical activity of neuronal and myocardial networks which continuously shapes the properties of these networks. Substrate-integrated microelectrode arrays allow the monitoring of the network activity by extra– cellular recording in vitro with multiple electrodes simultaneously, thus providing access to the spatio-temporal patterns of electrical activity. This chapter gives an outline of approaches that allow investigations of the properties of neuronal networks in brain slices and cell cultures of cardiac myocytes. It presents and discusses examples for the application of these techniques in biomedical and pharmaceutical research.

Key words:

micro electrode arrays, brain slices, neuronal networks, cardiac myocytes, analysis of local field potentials, drug screening,

1.

INTRODUCTION

In many biological tissues the reciprocal interaction between individual cells and cell groups, together with the individual cellular properties, define the function of the network they form. These tissues and their interconnections are highly structured into networks over a wide range of spatial scales. Brain tissue, for example, is organized in functional pathways consisting of, e.g., layers and nuclei with specific cell type composition and local structure [1,2], glial networks and glia/neuron complexes around synapses on dendritic spines [3–5]. In cardiac tissue specialized transition structures form atrial and sinus nodes, conducting pathways, and working


309

310 0

NETWORK ON CHIPS

muscle tissue [6–9]. The structural organization is essential for proper neuronal and cardiac network function. Pathological processes which affect this structure and the network composition of the tissue may, for instance, lead to epilepsy in neuronal tissue [10], to arrhythmia in cardiac tissue, and to other disorders. These tissue architectures are brought about by developmental processes and are modified continuously by interactions of local network activity and cellular properties [11–15]. On a molecular level, gene expression, the availability of receptors, their phosphorylation, and distribution may be changed by the activity in the network, directly or via its structure [16–18]. Arrays evaluating such changes, though not their fast dynamics are addressed in Chapter 5. Besides the obvious modulation of the overall action potential (AP) activity, excitation and inhibition in the network, the release of various neuromodulators, modulation of the extra–cellular K+-concentration, etc., and the spatiotemporal distribution of activity are important, e.g., for short term synaptic plasticity and the balance of inhibition and excitation. An increasing body of data collected in recent years indicates that the organization and activity of the local network continuously modifies the properties and the behavior of individual cells on time scales as short as minutes or seconds. A network of excitatory and inhibitory neurons, for example, shapes each neuron which, in turn, is an active part of this network. The properties of this neuron, its ion channel configuration, the distribution of synaptic strengths and locations, the fine structure of the dendritic tree, input impedance, etc., are continuously modified by the activity of the surrounding network [19–25]. The spatial and temporal structure of activity of the network as a whole ultimately represents information and its processing in the nervous system. Single-cell studies are thus obviously unsuitable to correctly assess the properties and activity dynamics in a network, or even of individual cells in this network. Similar influences of activity dynamics on cellular properties were observed in networks of cardiac myocytes [7,29–31], with their specialized sino-atrial networks, activity dependent embryonic development, and complex interactions of multiple pacemakers under pathological conditions. Cardiomyocytes are interconnected through gap junctions, effectively forming a large syncytium. The electrical resistance of these pores can also be modified. As a consequence the persistence or decay of a cellular response to a sensory or chemical stimulus will depend on the dynamics of the network activity. In addition the uptake and metabolization of a molecule by nonr It is therefore excitable cells in the vicinity will modify the effect of a drug. necessary to evaluate the response of an individual cell to manipulations affecting the network, e.g., by drug assays, electrical stimuli, and sensory

NETWORK ON CHIPS

311

stimuli in the context of the activity dynamics of that network. Whilst the optimal situation with respect to network integrity would be in vivo studies, these are often precluded when it is unacceptably difficult or timeconsuming to implement multi-electrode recordings, the desired control of the stimulus and environment, or to ensure access of the drug to the region of interest. In addition, the high complexity of the in vivo situation may hamper the interpretation of the results. Moreover, in vitro studies can be conducted with a higher throughput than in vivo experiments. Using in vitro preparations simplifies these experiments and allows better control of the cellular environment. Quite early on the recognition of these relationships led to the request for and the development of micro electrode arrays (MEAs) for extra–cellular recordings from neuronal networks and cardiac myocytes [32–52]. These pioneer studies contributed valuable insights, but were hampered by the limitations of the computing technology available at the time. The advent of inexpensive computing power suitable for acquireing, storing and analyzing the enormous data flow in such recordings renewed the interest in MEA techniques and spurred the development of suitable biological preparations, creating new perspectives for industrial applications. The use of MEAs for studying local networks of excitable cells in vitro, in particular in brain slices and networks of cardiac myocytes will be the focus of this chapter. Investigating the interaction within networks improves to our understanding of network function in general and of tissue architectures in particular. It helps our comprehension of the interplay between individual cells and network dynamics, and of the role of neuro modulatory transmitter systems within neuronal networks. From an application oriented view, the including of small yet complex in vitro networks in, for example, a drug assay, should provide a more realistic picture of the overall effect of a tested substance. This, in turn, can be expected to enhance the predictability of a drug’s effect on the intact organism and would help to identify the mechanisms by which this effect is mediated. Networks, for instance, readjust synaptic weights, adapting to modulations of ion channels or compensate changes otherwise. Moreover, multi-electrode recordings accelerate the collection of sample sizes necessary for statistical analyses1. The possibility to apply complex electrical stimuli through the MEA electrodes further expands the range of

1

For some applications optical recording using voltage–sensitive dyes [53–59] could be an alternative to this technique. The trade off between signal–to–noise ratio inherent to the technique, bleaching and the potential for phototoxicity, the lack of stimulation capability and the restricted recording time, however, limit their use.

312 2

NETWORK ON CHIPS

applications and bio-assays, and contributes to the evaluation of pharmacological test substances in industrial research.

2.

TECHNICAL ASPECTS AND UNDERLYING ASSUMPTIONS

Technically, MEAs can be classified into active and passive arrays. The development of active arrays is guided by the idea which an array could be designed most compactly with integrated circuits that at the same time incorporate a supporting or growth substrate for the tissue and a recording device, including amplifiers, filters and a readout system. Whilst several successful components and prototypes have been developed, with their function demonstrated in recordings [33,50,60–63], for a long time the approach has been limited by the complexity of the technology required and the need for bio-compatibility. Currently these devices are still too expensive to produce for routine applications in the life sciences, and hence are not commercially available. Passive arrays, on the other hand, in principle simply consist of insulated, thin film metal electrodes (Fig. 10-1) or in some cases of doped and therefore conducting silicon. These electrodes are arranged on a carrier (Fig. 10-1), e.g., a glass slide, and are connected to separate, dedicated amplifiers, AD converters and data acquisition setup. Electrode arrays based on silicon bulk designs have also been developed as 3D needle arrays, similar to wire bundles, mainly intended for in vivo work, e.g., for long term recording, stimulation and as neuroprosthetic devices in the brain or peripheral nerves [64–66]. The first approaches towards substrate integrated electrode arrays used cultures of dissociated neuronal or cardiac tissue [34,36,37], following the idea that the general properties of a network can be restored from individual cells dissociated from native tissue and cultured to grow connective structures between them [27,67–76]. Application oriented studies evaluated such cultures for chemical sensors based on neuronal networks [77–79]. To investigate the electro-physiological properties of cellular interactions in intact tissue, on the other hand, preparations of, e.g., tissue cultures, cardiac papillary muscle, freshly prepared (‘acute’) brain slices, retinal tissue, etc., are well established. For acute slice preparations thin sections of the tissue are cut such that the propagation pathways and receiving cells are

NETWORK ON CHIPS

313

sufficiently intact to model the necessary functionality of the tissue [80,81]. Such slices are also maintained in tissue culture [82–88] (Fig. 10-8A) to allow some regeneration of damaged cells, removal of cell debris, and the rebalancing of excitatory and inhibitory synaptic weights. Today MEAs suitable for routine electro-physiological recording to monitor the activity of neuronal populations in vitro are commercially available2. MEAs are used extra–cellularly to record low frequency local field potentials (LFP) and spike activity with substrate integrated electrodes at up to 70 sites in the tissue, i.e., the extra–cellular correlates of neuronal or cardiac APs. Such arrays are produced by thin film photo-lithography with conducting leads of gold, indium–tin oxide or similar materials, and insulated with silicon dioxide, polyimide, or silicone polymers which thus constitute the main contact surface for the cells. The actual electrodes are formed by an opening in the insulator 5–50 µm in diameter, exposing the tip of the leads underneath. During recording, a Helmholtz double layer builds up at the metal–electrolyte interface, effectively forming a high–pass filter [89–91]. The exposed electrode surface is usually coated with a porous material to lower the cut off frequency of this high–pass filter by increasing the capacitance of the Helmholtz double layer. This can be achieved, for instance, with galvanically deposited platinum or sputtered titanium nitride (TiN) [87,92] which forms a mechanically stable columnar structure (Fig. 10-1C). These modifications also increase the charge transfer capacity of the electrode, facilitating electrical stimulation via the MEA. The recording conditions for MEAs are defined by several biophysical and biological factors (Fig. 10-2). The decision about which specific MEA design to choose with respect to dimensions and materials, the data acquisition electronics and production techniques necessarily has to follow the biological question to be addressed. The design of an MEA used to record LFPs, for instance, should allow the coverage of an appropriate area of the tissue and should have electrodes with low impedance to reduce the lower cut off frequency produced by the capacitance of the electrode. Such electrodes also allow the application off electrical current stimuli with low voltages, avoiding electrolysis. From a technical perspective, the metal– electrolyte interface and the type of contact to the cells define the interface between cell and electrode [91]. The stray capacitance across the insulator

2

Links to developers of substrate-integrated arrays and corresponding equipment: Multi- Channel Systems, http://www.multi-channelsystem.com/; AlphaMed, http://www.amedsci.com/english/; Ayanda Biosystems, http://www.ayanda-biosys.com/; BioCell-Interface, http://www.biocell-interface.com/.

NETWORK ON CHIPS

314 4

largely determines the frequency gain function, influences the signal–to–noise ratio of the recording, and thus limits the range of detectable biological signals, and therefore the questions that can be targeted in an experiment. The Helmholtz capacitance at the electrode/electrolyte interface defines the electrode impedance and the lower cut off frequency. To record LFPs (1–300 Hz) or other low frequency components, the electrode impedance and the lower cut off frequency should be as low as possible. The MEAs used in our laboratories have 60 electrodes of 10–30 µm diameter on a 200 µm grid, with a coating of columnar TiN to minimize the impedance (Fig. 9-1). A

B

C

Figure 10-1. Design of a substrate integrated microelectrode array as used in our laboratory. (A) This version is on a 5×5 cm glass plate with conducting gold leads. A glass ring forms the recording chamber. (B) SEM micrograph of the recording area. The leads are insulated except for the center of the circles at the end of each lead. (C) SEM micrograph of a break-away of the electrode surface. Titanium nitride sputtered onto the electrode forms a columnar structure with a large inner surface, increasing the capacitance of the electrode area. MEAs from other producers are similar in their basic layout (Images by courtesy of Multi Channel Systems, Reutlingen and NMI Reutlingen, Germany).

2.1

System requirements

Recording single-unit spike activity3 with a dominant signal frequency of about 1 kHz in cultures of dissociated neuronal cells allows smaller

3

In contrast to multi-unit and population activity the term single-unit activity is mostly used when the spikes in a train recorded extra–cellularly can be attributed to individual neurons. This may mean that (i) only spikes from one neuron were recorded, (ii) that the spikes from different neurons were separated using spike sorting, or (iii) that spikes from one neuron were be separated from the rest. Multi-unit activity includes spikes detected with one electrode from several neurons that either were or cannot be separated, or in which individual spikes cannot be resolved at all. Population activity refers to contributions from

NETWORK ON CHIPS

315

electrodes of relatively high impedance. Such preparations are now used for drug and toxin screening [77,93], research on pattern and rhythm generation in networks [94,95], and lately to control the movement of objects in virtual space based on the activity in neuronal networks [96]. Besides ensuring tissue vitality (see below) and adequate conditions for cell culture4, the probably most critical aspect for successful recording from acute slice preparations with MEAs is establishing and maintaining very close contact between tissue and electrode. From the perspective of the equivalent circuit, decreasing this distance increases the seal resistance between the electrode tip and the reference electrode (Fig. 10-2), and thus the signal–to–noise ratio in the recording. This can be implemented by either carefully pressing the tissue onto the array [98,99] or, probably less stressful for the tissue, gluing the slice onto the MEA using an adhesive coating, e.g., polyethylenimine (PEI) or cellulose nitrate, as used in our lab [100]. Both techniques mediate close and flat adhesion of the slice to the insulating surface of the MEA, allowing rapid superfusion of the tissue. After positioning the sliced tissue, the arrays are mounted into a setup holding the amplifiers, filters, and the data acquisition system. Since acute slices have a high oxygen demand they rapidly decay without a good superfusion system. This needs to ensure rapid flow of buffer saturated with 95% O2/5% CO2 across the tissue, exchanging the bath volume 6–8 times per minute. Under these conditions, the O2 partial pressure decreases rapidly with depth in the tissue (Fig. 10-3), but remains well above arterial O2 pressure. Reducing the exchange rate results in suboptimal O2 supply.

large populations of neurons overlapping in time that consist of spikes and /or subthreshold membrane potentials. 4

Culture conditions vary considerably for different cell types. An MEA specific technical improvement was published by Potter [97] using a gas permeable, but water vapor impermeable, membrane which drastically reduces evaporation and thus the increase in osmolarity in the small culture chambers, which was a persistent problem even in hightech incubators.

NETWORK ON CHIPS

316 6

Figure 10-2. Equivalent circuit illustrating the main components of the recording configuration (modified from [101]). For extra–cellular recording the cell is considered a current source producing a voltage drop across the voltage divider formed by the resistance Rge through the buffer between cell membrane and electrode tip, the seal resistance RS between electrode tip and reference electrode and the impedance of the Helmholtz double layer (Re, Ce) in series with the resistances of the conducting leads Rm. A band pass filter is created by the impedance of the Helmholtz double layer at the metal–electrolyte interface and the stray capacitances across the insulator in the electrolyte solution (C st-e) and cables (Cst-cc). Increasing the electrode surface increases Ce and thus reduces the electrode impedance, improving its recording properties.

A

MEA

flow guide

brain slice

B

pO2 [hPa]

700 600 500 400 0

100 200 300 depth in slice [µm]

400

(A) Superfusion arrangement for MEA recording. A glass ring forms the recording chamber. The flow guide inserted into this ring reduces the volume to 0.5 ml and ensures a smooth flow of buffer across the tissue. (B) Oxygen partial pressure at increasing depth within the slice. O2 was measured with a fiber optic system (PreSens Precision Sensing, Regensburg, Germany) with a 30 µm sensor. When superfused with a buffer saturated with 95% O2/5% CO2 at 6–8 exchanges/minute, the O2 supply is well above arterial O2 partial pressure (approx. 120 hPa) at any depth in the tissue at room temperature. The slope of the O2 partial pressure depends on temperature and superfusion rate.

NETWORK ON CHIPS

3.

317

ORIGIN OF THE SIGNAL RECORDED

As described by Rall [102,103], the time course of the potential recorded extra–cellularly is related to the time course of the first derivative of the membrane potential of the cell studied. This corresponds to the time course of the depolarizing and hyperpolarizing currents across the membrane. Na+, K+, and Cl- flowing through ion channels in the membrane and compensating capacitive currents flowing simultaneously in the opposite directions carry the major part of these currents. The spatial distribution of the latter very much depends on the distribution of the internal and leak resistances of the cell which in turn are influenced by its morphology. As a result of the complex and ramified neuronal dendritic tree, it is essentially impossible to fully reconstruct the trans–membrane potential from an extra– cellular recording [104], neither in its amplitude nor its time course. This also indicates that the extra–cellular signal from thin axons will be minor, as only weak currents flow across their small membrane area. In addition, at any one site in the extra–cellular space the local potential will be the sum of contributions from all these current sources within the recording horizon of the electrode [100,105]. In native tissue this encompasses a large number of cells. The situation is somewhat simplified if the cell or tissue under study can be considered large with respect to the electrode, homogeneous in its morphology and if the seal resistance is high. This is achieved when cardiac myocytes are cultured on MEAs. These cells form extensive layers of cells coupled through gap junctions. Under suitable culture conditions they develop spontaneous rhythmic contractions with large APs propagating across the array (Fig. 10-4). This configuration allows intra–cellular recordings simultaneously with extra–cellular MEA recordings [105]. Analysis of the intra– and extra–cellular time course indicates that the current components mentioned above can in part be identified in the MEA recording. In the regions initiating the activity a sharp negative peak in the MEA recording accompanies the depolarizing upstroke of the intra–cellular AP that is mainly driven by Na+. Conversely, a positive peak parallels the repolarization phase carried by K+. At recording sites along the propagation pathway a positive peak, reflecting passive outward currents, preceded the negative peak. These experiments thus confirm the prediction that AP properties can be estimated from FPs and enable analyses of the cardiac AP duration based on MEA recordings.

NETWORK ON CHIPS

318 1

B MEA rows

2

e

4

d

5 6

c

7

2 mV

b

8 5 ms

C

10 10

a

1

b c d

4 6 8 10 12 14 16 18 20

20

3

delay to first spike [ms]

A

2

3 4 5 6 7 MEA columns

8

D

0

FPdur 140 #14

400

100 [ms]

[µm]

60

800

FPpre

FPMAX

20 #57

1200

200 µV 0

400

800 [µm]

1200

500 ms

20 µV FPMIN

50 ms

Figure 10-4. Spread of excitation in a cell culture of neonatal cardiac myocytes after 6 days in cultures. Based on recordings from 60 electrodes in this confluent cell layer, we reconstructed the spread of excitation through the tissue. (A) The delays and changes of the spike shape were readily visible in these sample traces recorded at 5 positions along the path of propagation (a–e in B). (B) In this case the origin of excitation lay in the bottom right corner of the MEA (approx. at the asterisk). Excitation spread towards the top left corner, as indicated by the arrow. The isochrones (interpolated based on 60 recording positions) give the delay of the spike minimum at each position with respect to the earliest peak. Only a thin layer of cells that did not visibly contract covered the right side of the array (columns 7 & 8). The minima in A were clipped graphically. (C) Overlay of the FP recording and false color rendering of the delays of the first negative peak on each electrode to the earliest occurrence on the MEA [105]. ( D) Identification of the FP minimum (FP M I N ) marking the onset of the AP and the maximum (FP M AX) indicating the downslope of the AP.

NETWORK ON CHIPS

4.

319

SPATIAL RESOLUTION

Because of the extensive and overlapping dendritic trees of many types of neurons, it is not straightforward to identify where the neuronal signal recorded at an electrode was generated. Obviously the amplitude of neuronal spikes will decrease with distance. A second scaling factor is the size of a neuron, or rather its membrane, as can be deduced from the relation of the extra–cellular signal to the currents flowing across the membrane. Since the specific membrane resistance is constant (unless ion channels open), small neurons have a lower capacitance and a higher input resistance. The charge required to maximally depolarize their membrane is thus smaller than for large neurons. Their extra–cellular signal is therefore smaller than that of larger cells. This ambiguity impairs the identification of the spatial location of signal source in single unit recordings. We investigated this question in MEA recordings from acute parasagittal slices from the rat cerebellum (Fig. 10-5). The structure of the cerebellum is such that in these slices the large, essentially two-dimensional, dendritic tree of the Purkinje cells lies flat on the MEA, extending into the molecular layer only. Their cell bodies are aligned in the Purkinje cell layer. These cells are spontaneously active at high rates. To determine the recording horizon of MEA electrodes we probed the region around one MEA electrode that picked up spikes with a single, separate micropipette electrode. We then measured the amplitude distribution of the potential recorded with the micro pipette when a spike was detected on the MEA electrode as a function of the distance between the two electrodes (Fig. 10-6). The resulting potential surface indicated that the recording horizon of such an has a radius of approx. 60 µm [100]. The fields of view of the MEA electrodes spaced on a 200 µm grid therefore did not overlap. This footprint is obviously valid for unit spikes only, and the electrode horizon scales with the size of the cell for reasons explained above. Extensive dendritic trees in some cases will allow the recording of spikes at greater distances. The lack of conspicuous correlation between spikes recorded at neighboring electrodes further supports that there was no detectable overlap between the recordings at neighboring electrodes (Fig. 10-6D) [100].

NETWORK ON CHIPS

320 B

A

0.4 mm

0.5 mV 10 ms

Figure 10-5. (A) Cerebellar slice from a juvenile rat in the recordings situation. The layers of the cerebellar cortex and the white matter are clearly distinguishable (dark: myelinated fibers in the white matter; next layers: granule cell layer, Purkinje cell layer (only one cell body in thickness) and molecular layer with the dendritic tree of the Purkinje and Golgi cells). Fissures separate the lobes of the cerebellum. (B) Raw signal recorded at the electrodes marked in A. Spikes can be readily detected. The signal–to–noise ratio is typical for this preparation, and typically even higher in organotypic cultures of such slices.

5.

LFP AND PLASTICITY

The discussion in the previous section has already indicated that, given the number of neurons and their neurites located within the recording horizon, the FPs created by these neurons can overlap in time and add up linearly. Because of this summation, contributions by single neurons occurring more or less synchronously may therefore become indiscernible. On the other hand, because of this spatial integration signals from synchronized events which would normally be too small to be identified may rise above the noise level and become detectable. This is particularly interesting for excitatory and inhibitory post synaptic potentials (EPSPs and IPSPs). These events, depolarizing the membrane by less than 10 mV and lasting several tens of milliseconds, overlap in time in this situation and become visible as population PSPs (pPSPs). Mixed with population spikes they constitute LFPs. The long time constants of the individual EPSPs and IPSPs and their temporal distribution lead to dominating frequencies below approx. 300 Hz in the LFP. These signal components can be separated from spike activity consisting mainly of components above 500 Hz by filtering. The detailed frequency composition of LFPs is quite variable, ranging from low frequency response components, with time constants up to hundreds of milliseconds, to approxamtely 300 Hz. Spikes may contain components up to 3.5 kHz.

NETWORK ON CHIPS

321

A

B

y-distance [mm]

100

0

-100 -10

m] 0

C

0.2 0.4 0.6 ratio of minima

0.8

20 ms

30*/80 µV

D

0 500 ms

Figure 10-6. Spike activity is the basic single-neuron activity recorded with extra–cellular electrodes. In the cerebellum, for example, Purkinje cells are firing spontaneously. (A) The micrograph illustrates the structure and dimensions of an acute parasagittal slice of a rat cerebellum on an MEA. (B) Relative strength of the peak recorded with a micro pipette at the positions marked with white dots. From the voltage measured with the micro pipette when a spike was detected at the reference MEA electrode (square), we calculated the ratio of this voltage and the voltage at the MEA electrode. The greyscale image shows the interpolated distribution of these values, indicating that background level is reached at approx. 60 µm from the maximum. The bright region approx. 20 µm identifies the likely position of the cell body. The dashed lines indicate the Purkinje cell layer. (C) Single (gray) and spike-triggered average (black) recordings at nine neighboring MEA electrodes, centered on the spike time of a neuron at the central electrode. There were no detectable spikes at the surrounding electrodes that could correspond to the spike att the center electrode. Even the average traces show at best only small peaks that are not detectable at this noise level. (D) These small peaks were not suff fficient to resultt in peaks in the cross correlograms [100].

322 2

NETWORK ON CHIPS

Common fields of research are the investigation of synaptic plasticity, the role and origin of specific frequency bands such as the > 30 Hz gamma band, the 4–10 Hz theta rhythm found in the cerebellum and brain stem, and the acetylcholine modulated beta rhythm (10–30 Hz) [106]. Further applications are studies on epileptogenesis and on synaptic plasticity, e.g., long term potentiation (LTP) and long term depression (LTD). With slice preparations it is possible to address several of these topics. A classical preparation for investigations on (LTP) is the hippocampal slice, a region involved in memory formation and spatial orientation. It can be cut so that there are three major populations of synapses and local connections of, e.g., inhibitory interneurons (Fig. 10-7A). The strength of these synapses depends on the history of the incoming activity and the efficacy with which the activation of the synapses contributed to the firing of the post synaptic cells. The pyramidal cells in the CA3 subregion of the hippocampus, for example, project onto the dendrites of the pyramidal cells in the CA1 region along the Schaffer collateral pathway. The efficacy of the synapses formed by these axons can be enhanced by high frequency stimulation delivered through either MEA or separate wire electrodes. The distribution and dynamics of the responses to such stimuli, as well as of the synaptic plasticity determined from changes of the response magnitude (Fig. 10-7B) can be determined from LFP responses in MEA recordings. With the interpretation of the origin of negative and positive voltages given above, one can, with owing caution, now describe some of the dynamics of the voltage pattern as a sequence of the propagation of axonal spikes, post synaptic activation in the dendritic layers of the CA region, and post synaptic population spike activity. Pre- and post synaptic elements of the waveform can be identified by subtracting the response obtained after blocking synaptic transmission, e.g., by removal of Ca2+ from the buffer, from the response obtained in normal recording buffer. The outstanding mechanical stability of the recording configuration makes these analyses highly reproducible and easy to do. To determine the location of the synaptic currents, current source density (CSD) analysis can be a suitable tool if the tissue has a laminar structure and the electrode spacing is sufficiently small to distinguish all layers [107–113]. CSD analysis was originally used to analyze LFP responses sampled sequentially at regular intervals along the

NETWORK ON CHIPS

323

track of an electrode pushed into the brain. It relies on the assumption that the general structure of the tissue changes along one dimension only, e.g., as in layered tissues such as the neocortex or the lateral geniculate nucleus. Novak and Wheeler extended this analysis to two dimensions in MEA recordings of the hippocampus [108,109]. They approximated the procedure by applying a two-dimensional Laplacian kernel to the voltage map gained from MEA recordings. As any procedure working on sampled data, this assumes that the spatial frequencies of the current maps are sampled at sufficiently high spatial intervals. Most MEAs used, however, do not meet this requirement. Filtering for low spatial frequencies alleviates this problem, but restricts the interpretation to relatively large current sinks and sources [106].

6.

NETWORK DYNAMICS AND EPILEPTIFORM ACTIVITY

For long term studies brain slices can be maintained in tissue cultures on MEAs, enabling experiments which require either intermittent or continuous recording and stimulation of the tissue [87,114]. Tissue cultures on MEAs were used to investigate neuronal regeneration [114], neuronal development [98,99,115,116], rhythmogenesis [75,76,117], circadian rhythms [118], and the initiation of epileptiform activity [88,108,109]. Cultures of hippocampal slices maintain their general layout with the principal areas, layers and connections outlined above (Fig. 10-8A). In these cultures spontaneous spike activity develops even though the input pathways are severed. A network of inhibitory interneurons contributes to the stability of the electrical activity. Interrupting this inhibitory transmission by blocking GABAA receptors unbalances the network and leads to synchronous, epileptiform activity (Fig. 10-8). With respect to the treatment of epilepsy, this transition from desynchronized balanced spike activity to highly synchronized propagating population activity is particularly interesting because it marks the onset phase of the epileptiform activity dynamics. This transition could potentially be more easily treated than fully fledged epileptic seizures.

324 4

NETWORK ON CHIPS

Figure 10-7. Analysis of long term potentiation (LTP) in an acute slice of the hippocampus. The distribution of LFP properties and changes in hippocampal slices (A) can be assessed with MEAs. (B) Population spike amplitudes induced by electrical stimulation (at the white dot in A) can reach more than 2 mVpeak to peak , with pronounced minima (Vminn) in the pyramidal cell layer and the proximal apical dendrites. (C) Brief bursts of stimuli (*, three repetitions) induced LTP., drastically increasing the response amplitude. The insert shows a typical response profile (thin line: before LTP induction, thick line: recorded at 3100 s) after high frequency stimulation. The responses recorded at the asterisk in A increased temporarily by up to 140% (post tetanic potentiation) and persistently by more than 50%. (D) The distribution of the increase of the response amplitude is visualized as an interpolated pseudo– color image. Significant increases of the response are mainly found in areas activated by orthodromic or antidromic activation of synapses a in the CA region and the dentate gyrus [119].

This preparation thus provides a model in which the synchronization of activity in time and space can be monitored and related to the structure of the network. This requires mapping of the LFP data onto the dynamics of activity in the network, i.e., the spatial and temporal distribution of synaptic and spike activity . Extracting these distributions from the LFP by separating the average contribution of a neuron and the temporal distribution of firing

NETWORK ON CHIPS

325

A

D

C proximal CA3

distal CA3

B

CA1

= pSpike

+ syn.

+ uSpk

5 ms

10 spikes/ms 5 ms

Figure 10-8. Population spikes and firing time distributions. (A) Organotypic culture of a hippocampal slice on an MEA. Inter-electrode distance is 200 µm. (B) Population spikes (pSpike) detected in LFPs can be seen as a convolution of the firing time distribution (FTD) of the local neuron population, convoluted with a unit spike (uSpk) plus contributions of synaptic events (syn). Based on this model, the FTD can be estimated by deconvolving the LFPs (C) with the average single-unit spike as uSpk. The synaptic event residual is then estimated from negative values of the firing time distribution. After subtraction of this residual the deconvolution is repeated resulting in n the FTDs in (D). In this recording of an epileptiform spike the population activity synchronized as it propagated from the initiation site in CA3 towards CA1, resulting in sharper FTDs with higher peak spike rates (see also Fig. 10-9) (modified from [88]).

times can create this link. The contribution of a cell within the recording horizon of an electrode can be estimated by averaging the waveforms of all spikes detected at an electrode. Behind this approach stands the assumption that the composition of the neuronal population contributing these spikes reflects the population participating in generating the LFP. The population spike can then be considered as the convolution of the average neuronal spike and the distribution of the times at which the cells within the recorded volume fire (firing time distribution, FTD) (Fig. 10-8). The structure of the FTDs at the MEA electrodes was compared to the dynamics in computational network models built with hippocampus-like neurons and architecture. The simulated activity quite faithfully reproduced the dynamics observed in the culture, indicating that only a few cellular parameters need to be changed to induce the transition (Fig. 10-9) from normal to epileptiform activity. Such computational models could be used to identify the mechanisms and factors influencing the spatio-temporal dynamics of the

NETWORK ON CHIPS

326 6

FTD during the transition from irregular activity to synchronized firing. This could lead to new targets for the development of anti-epileptic drugs [88,121].

Figure 10-9. LFP and corresponding FTD as false color representation. (A) Grey-scale rendering of the LFPs recorded at successive electrodes along the pyramidal cell layer in an organotypic culture of the rat hippocampus on an MEA. The LFPs were epileptiform spikes provoked by disinhibiting the network with the GABAA receptor blocker bicuculline. As the population spike propagates along the pyramidal cell layer from CA3 (top) to CA1 (bottom), its amplitude increases and its temporal width decreases. (B) FTDs were calculated from the LFPs as described above. They now represent the spike activity in the underlying network. The dynamics of this activity and the conditions for the initiation and propagation of this type of epileptiform activity was further analyzed in the simulations of networks with hippocampus-like properties. The network reproduced most of the phenomena observed in MEA recordings [88,120].

7.

DRUG TESTING WITH MEAS

One of the incentives for developing the MEA technique and the necessary biological preparations was that they might be useful tools for drug evaluation in pharmaceutical industry. Several aspects contributed to

NETWORK ON CHIPS

327

this expectation: one concept is that the insights gained from network studies will facilitate the search for new pharmaceutical drugs and the evaluation of their potential benefits and risks in the pre-clinical screening phase. In addition to the approach to evaluating network function to improve the predictive value of in vitro experiments, MEA studies through parallel recording could decrease the time needed to collect statistically relevant sample sizes, increase the throughput of test substances and reduce the associated costs per data point. Because of the mechanical design of the setup, parallel, semi-automatic recordings from multiple setups become feasible. MEAs allow multiple sterile measurements from the same preparation where needed, e.g., in regeneration studies, and they can be used to simultaneously assess the response distribution in a population, given that the responses are generated independently.

7.1

Using Network Properties as Endpoints in Drug Assays

Some properties of electrical activity in the nervous system are a direct consequence of neuronal interactions that cannot be detected in single cell studies. Amongst these are spatially distributed oscillations of spike activity, observable, for example, in correlation analyses. These patterns have gained renewed interest since it was discovered that in vivo the γ frequency band (30–70 Hz) is associated with the detection of contiguous objects in visual stimuli [122]. The mechanisms mediating and controlling these and other network patterns are, however, not always clear. β wave (10–30 Hz) rhythms are known to be modulated by connections using acetylcholine as a transmitter [106]. This type of activity was found in acute slices of the hippocampus and was successfully mapped with MEA recordings [106]. The dynamics and spatio-temporal extent of these oscillations as well as mechanisms controlling them can be rapidly investigated in vitro using MEAs. Carbachol, a cholinergic agonist, induced such oscillations which may serve as indicators of cholinergic action of a test substance (Fig. 10-10).

7.2

Assessing Distributions of Neuronal Responses to Dopamine

Dopamine is an important neuro–modulating transmitter known to be involved in a wide spectrum of physiological and pathological processes. The receptors are classified into various subtypes which can be linked to the different functions of dopamine. D1 receptors are, for example, linked to learning and reward, D2 receptors to Parkinson’s disease, and D3 receptors to

328 8

NETWO ORK ON CHIPS

schizophrenia. Identifying sites of action and responses to the presence of an agonist at these receptors in the network is therefore an important topic in industrial drug screening. Most neurons, however, express several receptor subtypes in combination, hampering the analysis of a particular substance s selectivity and specificity. Only a few areas in the brain have separable populations of neurons with D2 and D3 receptors, interspersed with neurons lacking them. An electro physiological study therefore faces a heterogeneous response profile of the population and needs to record from a large number of cells to collect a representative sample. In the vermis, the central ridge of the rat cerebellum, it was discovered that lobes 9 and 10, the hindmost ventral lobes, have a high density of D3 receptors but lack D2 receptors [123]. ‘

Figure 10-10. Current source density analysis of carbachol r induced beta activity in a rat hippocampus slice. (A) This high density array covers parts of CA3 and CA1 in the slice. (B) On this snapshot taken during β wave activity, the sources are located in the pyramidal cell layer, sinks in the dendritic tree, indicating excitatory synaptic activity. Though induced by carbachol, the activity depends on glutamatergic transmission, probably involving several local networks which are involved in the mechanisms inducing β wave activity. The pattern propagated with a 23 Hz cycle (modified from [106]).

NETWORK ON CHIPS

329

Figure 10-11. Using spontaneous spike activity for dopamine assays. (A) Firing rate of spontaneous activity in a cerebellar slice. Activity fluctuates in a burst pattern. Intra–burst activity increases when the dopamine D 2 /D 3 receptor agonist quinpirole is applied in increasing concentrations. Dashed lines indicate the evaluated time window and application of different concentrations of the drug. (B) The overall activity is summarized in the mean rate and the Fano factor as a measure of the local variability of the spike count for each concentration. (C) The dynamics of the autocorrelation during the burst illustrates a pattern of regular firing. The mean firing frequency within the burst increased with the drug concentration. (D) Autocorrelograms for two intervals resp. drug concentrations [124,125].

Other lobes in the cerebellum in turn do express the D2 but not the D3 receptor. We therefore investigate the modulation by dopaminergic drugs of spontaneous spike activity in acute slice preparations from the cerebellum of neonatal rats. Since, as described above, spike activity detected in MEA recordings from the cerebellum reflects electrically and physiologically independent spike sources, each experiment provides a sample from a population of cells. We therefore analyze the modulation by dopaminergic drugs of spontaneous spike activity in acute slice preparations from the cerebellum of neonatal rats. The geometry of the MEAs used in our laboratory allows the

NETWORK ON CHIPS

330 0

experimenter to record from cells in up to three lobes simultaneously, with several electrodes in each of them. This approach thus provides a sample of the populations with different receptor sets. Spikes are readily detected in these preparations (Fig. 10-5). Fig. 10-11 illustrates the changes of the spontaneous activity found at an electrode in one such experiment. The temporal structure of the spike activity has a burst component with regular intra–burst intervals between spikes and intermittent pauses of up to several seconds. The autocorrelation, reflecting the periodicity of the spike activity, reveals this structure with distinct side bands. The dynamics of the changes induced by a drug at increasing concentrations is suitably captured by a false color representation of the autocorrelation calculated for spikes within a moving time window and summarized in correlograms of the relevant phases of the experiment. From such experiments we determined the typical response profile of the neuron population in a cerebellar lobe to drugs with known receptor selectivity. These ‘finger prints’ summarize the distributions of the directions of change of the above parameters (Fig. 10-12). We expect that this technique will allow the rapid identification of the response profile for a given drug by comparison with the respective fingerprints of known substances. Quinpirole D2/D3 agonist

frequency of a response type [%]

N

97

100

71

97

PD128907 semi-selective D 3 agonist 95

38

22

0

25

87

100 79

35

24

9

50

38

100 76

0

26 32

30

0

0 rate

coeff. of autocorr. variation

burst incid.

rate

response parameters

0

14

burst coeff. of autocorr. variation incid. down or wider up or tighter

Figure 10-12. The distribution of the responses which are highly variable across the population recorded forms a fingerprint-like pattern which is different for each of the drugs tested. These patterns could help to identify related response types [124,125].

NETWORK ON CHIPS

7.3

331

Cardiopharmacology

In cardiology the spread of activity, its embryonic development and its pathological conditions, such as arrhythmias, are highly important issues (Fig. 10-3). Here the continuously contracting tissue poses quite a challenge for recordings with conventional micro electrodes. A comparatively slow pacemaker current depolarizes the cell to a threshold at which a fast AP is elicited. The AP is dominated by an inward Na+-current and repolarized by outward K+-currents through various ion channels. The characteristic plateau phase of the cardiac AP is sustained by a Ca2+ inward current, whose time course varies across cells at different positions within the cardiac tissue, i.e., their functional differentiation. The relative strengths of these currents are highly critical to ensuring the correct succession of excitation and propagation in the cardiac muscle and thus its physiological function and performance. The electro physiological measurement and analysis of these currents, reflected in the time course of the membrane potential, is therefore an important field in basic research and routine analysis in pharmacological research and development. The AP duration and its variability is a highly important parameter in safety pharmacology. In vivo this property of the AP is reflected in the interval between the Q and the T wave of the electro cardiogram (duration of the QT period). Under certain conditions prolongation of QT duration has been linked to lethal side effects of drugs [126–128]. Test for QT prolongation have therefore become a mandatory requirement in safety pharmacology [127]. This necessitates tests on a large number of drugs, either under development or already on the market. Using in vivo tests is inconvenient because of the associated high costs and low throughput. Since several ion channels and receptors are involved in the dynamics determining the AP duration, with the HERG channel being the most important one, data obtained from cell lines expressing a single ion channel type can be misleading, resulting in false negatives or false positives. The interplay of the various cellular mechanisms is better captured in native cells. In recordings with conventional electrodes, motion of the tissue with respect to the electrode may not only damage the cells but also produces motion artifacts in the electrical signal that prevent a detailed analysis of the FP. It is therefore generally very difficult, and often impossible, to determine spatial dynamics of excitation, details of the waveform and long term changes, let alone perform multiple simultaneous measurements in cardiac tissue. The combination of MEA recording with cultures of cardiac myocytes thus marks significant progress, supported by the fortuitous situation that the contraction is essentially isometric at the tissue surface adhering to the substrate. The signal amplitude can be very high and motion

332 2

NETWORK ON CHIPS

artifacts are not observed. Substrate-integrated electrodes thus create a unique mechanical situation forming a highly stable recording situation. This allows the observation and analysis of activity in contracting cardiac tissue over periods of several days, enabling a new type of experiments on cardiac development [129,130]. Numerous studies have exploited this advantage to study signal to electrode coupling, arrhythmia or signal propagation in cardiac myocyte cultures. Substrate integrated MEAs also allow the assessment of the duration of the cardiac AP. As shown above, changes of the AP duration can be estimated in cell cultures by analyzing the waveform of the FP recorded with MEAs (Fig. 10-4). Considering the FP as a function of the inverse of the first derivative of the membrane voltage with respect to time, the strong depolarizing Na+ current and the repolarizing K+ current can be identified as leading negative and trailing positive peaks in the FP. The interval between these peaks is defined as FPdurr and is proportional to the QT interval [105]. MEAs thus provide a simple tool for implementing semi-automatic assays on such cultures. Recent experiments validated the approach using drugs with well known effects on the QT duration (Fig. 10-14). The concentrationdependent response profiles found in cardiomyocyte cultures on MEAs were very similar to those found in vivo [126,131,132]. These results lead to the development of simplified, and therefore inexpensive, single usage recording arrays on printed circuit boards. Such electrodes with 100 µm diameter can be integrated in the 96-well plate format commonly used with pipetting robots for routine applications. Such arrays are currently tested for medium throughput QT screening in industrial drug safety pharmacology (Fig. 10-13).

Figure 10-13. Prototype of a system with substrate integrated micro electrodes produced on printed circuit boards with standard 96-well plate format. In this case cardiac myocytes will be cultured on the arrays for tests of QT prolongation. The standard 96-well layout and robot compatibility facilitate the adaptation to common laboratory procedures and increases the throughput of the assay. The data acquisition electronics is integrated onto the base plate (by courtesy of Multi-Channel Systems, Reutlingen, Germany).

NETWORK ON CHIPS

Control 2 µM 5 µM 15 µM 50 µM

B

Quinidine Sotalol

E-4031

2 normalized FP duration

A

333

Cisapride Astemizol Terfenadine

150 µM

Amiodarone

1 Verapamil 100 ms

-1

10

0

10

1

10

2

10 10 concentration [nM]

3

4

10

5

10

Figure 10-14. MEA recordings allow in vitro prescreening for QT prolongation in cardiac myocytes. (A) Individual field potentials (FP) recorded in a chicken cardiomyocyte culture on an MEA during treatment with the sodium channel blocker quinidine. Quinidine increases the duration of the cardiac AP and of the QT interval in the in vivo electrocardiogram. This in vitro FP waveform shows conspicuous peaks the timing of which correlates with onset and end of the AP (see Fig. 10-4). The AP prolongation elicited by this drug is clearly visible in the FP. (B) All drugs tested thus far changed the FP duration as expected from ECG recordings [126,131].

8.

DATA ANALYSIS

A critical aspect in all multi-electrode recording is the processing of the large amount of data acquired in each experiment. For a 60-electrode array, for example, continuous recording at a sampling rate of 25 kHz with a 12 bit AD converter amounts to approximately 3.2 MB per second. Today it is simple to store continuously these and higher data rates to hard disc, nonetheless preprocessing, visualizing, analyzing, and searching these data for suitable descriptive parameters becomes time consuming, in particular when interactions between neurons are of interest. It exceeds the scope of this chapter to discuss such data analysis in detail. We shall, however, give a short outline of the demands appearing in MEA experiments and provide links to literature and software available for this purpose. The analysis process can be segregated into online and offline analyses. Online analyses are needed to monitor the experiment and for quick analysis

334 4

NETWORK ON CHIPS

of important parameters, such as spike rate, peri-stimulus time histograms (PSTH), extraction of minima and maxima of a waveform and their timing, etc. Offline analyses may either require more processing time or more of the experimenter’s attention than is available during an experiment. Other analyses may require preprocessing, e.g., to remove artifacts, sort spikes recorded from different neurons at the same electrode by more sophisticated methods, detrending, and the like. In particular, analyses of correlations or inter–dependencies of activity across recording sites and time is generally not possible online. Some visualizations might be feasible but are not useful in real time, e.g., animated displays of the spatial dynamics of voltage distributions or of spike rates lasting only several tens of milliseconds, for which slow motion movies are more adequate, obviously precluding online analysis. Finally, the spatial distribution of activity or its properties may be combined with histological data, such as cell type distributions or receptor distributions. Performing these analyses electrode by electrode on a routine basis is impractical, necessitating automated processing for electrodes within and across experiments. For each scientific question the data will be more or less similar, allowing their processing for specific features in recurring stages. Given the multitude of aspects to analyze and questions to ask, there is, unfortunately, no unique answer as to which tools should be used. Numerous authors have published individual analysis techniques, programmed implementations of tools, or analyzed in detail the pitfalls associated with the interpretation of some of their results or their combinations [133–136]. A description of this work would exceed the scope of this chapter. The collection of such techniques will probably grow as new recording techniques and data acquisition programs appear on the market. Besides a few commercial software tools (see below), many laboratories have developed their own routines for specific applications and experimental conditions, some of which can be extended as needed which often requires some programming knowledge. Researchers seeking to identify new aspects of network behavior might be better advised to resort to highly flexible, i.e., programmable analysis tools which allow more control but are less easy to use and sometimes slower in their performance, such as MATLAB (The Mathworks, Natick, Ms., USA) which is widspread in the neuro science community,

NETWORK ON CHIPS

335

NeuroExplorer (Nex Technologies) with its scripting language, or IGOR Pro (WaveMetrics)5. The software interfaces of the most widespread data acquisition systems usually allow some online analyses to monitor the experiment and some more complex offline analysis that may require more processing power. This multitude of complex data acquisition hardware and data analysis tools led to the idea that a platform should be available that allows the applicability and the exchange of analysis tools written for specific acquisition systems and data file formats. For this purpose, a number of hardware developers recently formed the NEUROSHARE forum to design and agree on a common interchange format and public domain standards6. The forum also provides a platform for software for neurophysiology in general.

9.

OUTLOOK

The applications and techniques presented above illustrate the wide range of questions which can be approached with substrate integrated micro electrode arrays. Beyond the applications mentioned above, substrate integrated arrays have also been used to study acute and cultured brain slices from various brain regions [76,114,137], retinal function and development [98,99], embryonic stem cells[129,138–140] and even whole chicken embryos to study early cardiac development. The option to combine MEAs with conventional electrophysiological techniques, optical measurements, with Ca2+-sensitive dyes and transparent Indium–Tin Oxide (ITO) electrodes, and additional integrated sensors promises further uses for this technique. The spatio-temporal dynamics of electrical network activity can thus be matched with the spatial distribution and, in some cases, the dynamics of other processes, as well as with anatomical maps of, e.g., receptor distributions gained from immunohistochemical or in situ hybridization studies.

5

Links to data analysis tools: NeuroExplorer, Nex Technologies, http://www.neuroexplorer.com; IGOR Pro, WaveMetrics, http://www.wavemetrics.com/Products/IGORPro/IgorPro.html; Matlab, The Mathworks, http://www.mathworks.com/products/matlab; Plexon Inc., http://www.plexoninc.com.

6

www.neuroshare.org.

336 6

NETWORK ON CHIPS

The technical similarity with implantable devices with respect to the materials used and the interface to the tissue facilitates preparatory in vitro studies towards, e.g., neuroprosthetic implants [141,142]. Nonetheless, several challenges remain and ask for new developments. On the technical side the viability of the slices should be further improved, in particular for preparations which need to be recorded at physiological temperature or under conditions that elicit high activity. These cells have a high metabolic demand that needs to be maintained by diffusion alone. Perforated flexible arrays have been proposed and were successfully used for this purpose [143], but these are not yet available commercially. Recent experiments by W. Stein (NMI Reutlingen, personal communication) and in our laboratory indicate that using such arrays with some of the perfusion buffer guided to flow beneath the slice and the perforated array indeed improves the oxygen availability and the signal–to–noise ratio (SNR) in hippocampal and cerebellar slices. Using such arrays should also shorten the time until a drug fully permeates the tissue and thus the response time of the tissue.

Figure 10-15. 8-well multiple MEA recording device. This type of developments is intended either for simultaneous recording from cell cultures or for acute slices (by courtesy of Ayanda Biosystems, Lausanne, CH).

NETWORK ON CHIPS

337

To increase the throughput and cut the costs of experiments using slices in drug research it is desirable to run several preparations simultaneously. This requires that the systems can run largely unsupervised which necessitates highly stable recording, stimulation, perfusion- and drug-application configurations. MEAs potentially enable this approach, in particular with the perspective for new, m multi-well plate arrays (Figs. 10-15, 13). To further simplify routine experiments, tissue specific electrode arrangements have been developed that facilitate the reproduction of a tissue–to–electrode configuration while giving the necessary spatial coverage and resolution. One such design is an arrangement that provides a line of electrodes to stimulate the Schaffer Collaterals in the hippocampus, e.g., to investigate LTP induction, while monitoring the CA1 region with optimal spatial resolution [144]. This trend towards parallelizing the recording creates the need for further efforts for automated, unsupervised data analysis with tools whose results should be robust against differences in SNR, frequency composition and artifacts in the signal. It would further be desirable to reduce the need for single unit analyses which often require extensive pre-processing, e.g., spike sorting, and safeguarding against nonstationarity of the spike generating process. We expect that the availability of these techniques, of the corresponding data acquisition and data analysis tools, and an increasing number of experimental protocols will contribute to our understanding of signal propagation and information processing in neuronal and cardiac networks, as well as pathological conditions thereof and approaches for their treatment.

10.

ACKNOWLEDGEMENTS

The authors would like to thank the German BMBF (FKZ 0310964D and 16SV1743), the Land Baden-Württemberg, and the Deutsche Forschungsgemeinschaft (SFB 505) for their support. We would also like to thank C. Boucsein and M. Nawrot for helpful discussions on the manuscript.

NETWORK ON CHIPS

338 8

11. [1]

REFERENCES

Petersen, C.C.H., Sakmann, B., (2000), ‘The excitatory neuronal network of rat layer 4 barrel cortex’, J. Neurosci. 20, 7539–7546. [2] Lübke, J., Egger, V., Sakmann, B., Feldmeyer, D., (2000), ‘Columnar organization of dendrites and axons of single and synaptically coupled excitatory spiny neurons in layer 4 of the rat barrel cortex’, J. Neurosci. 20, 5300–5311. [3] Chang, F.L.F., Hawrylak, N., Greenough, W.T., (1993), ‘Astrocytic and synaptic response to kindling in hippocampal subfield CA1.1. Synaptogenesis in response to kindling in vitro’, Brain Res. 603, 302–308. [4] Hawrylak, N., Chang, F.L.F., Greenough, W.T., (1993), ‘Astrocytic and synaptic response to kindling in hippocampal subfield CA1.II. synaptogenesis and astrocytic process increases to in vivo kindling’, Brain Res. 603, 309–316. [5] Derouiche, A., Frotscher, M., (1991), ‘Astroglial processes around identified glutamatergic synapses contain glutamine synthetase– evidence for transmitter degradation’, Brain Res. 552, 346–350. [6] Yamamoto, M., Honjo, H., Niwa, R., Kodama, I., (1998), ‘Low frequency extra–cellular potentials recorded from the sinoatrial node’, Cardiovasc. Res. 39, 360–372. [7] Spach, M.S., Boineau, J.P., (1997), ‘Microfibrosis produces electrical load variations owing to loss of side-to-side cell connections: a major mechanism of structural heart disease arrhythmias’, Pacing & Clinical Electrophysiology 20, 397–413. [8] Spach, M.S., Kootsey, J.M., (1983), ‘The nature of electrical propagation in cardiac muscle’, Am. J. Physiol. 244, H3–22. [9] Spach, M.S., (1983), ‘The role of cell-to-cell coupling in cardiac conduction disturbances’, Advances in Experimental Medicine & Biology 161, 61–77. [10] Haas, C.A., Frotscher, M., (2004), ‘Migration disorders and epilepsy’, in: Brain Damage and Repair. From molecular research to clinical therapy, edited by J. M. Delgado-Garcia and T. Herdegen, Kluwer Academic Publishers,. [11] Haas, C.A., Dudeck, O., Kirsch, M., Huszka, C., Kann, G., Pollak, S., Zentner, J., Frotscher, M., (2002), ‘Role for reelin in the development of granule cell dispersion in temporal lobe epilepsy’, J. Neurosci. 22, 5797–5802. [12] Frotscher, M., Haas, C.A., Forster, E., (2003), ‘Reelin controls granule cell migration in the dentate gyrus by acting on the radial glial scaffold’, Cereb. Cortex 13, 634–640.

NETWORK ON CHIPS

339

[13] Shatz, C.J., (1990), ‘Impulse activity and patterning of connections during CNS development’, Neuron 5, 745–756. [14] Goodman, C.S., Shatz, C.J., (1993), ‘Developmental mechanisms that generate precise patterns of neuronal connectivity’, Cell 72, 77–98. [15] Shatz, C.J., (1996), ‘Emergence of order in visual system development’, Proc. Natl. Acad. Sci. USA, 93, 602–608. [16] Grosshans, D.R., Clayton, D.A., Coultrap, S.J., Browning, M.D., (2002), ‘LTP leads to rapid surface expression of NMDA but not AMPA receptors in adult rat CA1’, Nat. Neurosci. 5, 27–33. [17] Beatti, E.C., Stellwagen, D., Morishita, W., Bresnahan, J.C., Ha, B.K., von Zastrow, M., Beattie, M.S., Malenka, R.C., (2002), ‘Control of synaptic strength by glial TNFalpha’, Science 295, 2282–2285. [18] Daw, M. I., Bortolotto, Z.A., Saulle, E., Zaman, S., Collingridge, G.L., Isaac, J.T., (2002), ‘Phosphatidylinositol 3 kinase regulates synapse specificity of hippocampal long term depression’, Nat. Neurosci. 5, 835–836. [19] Nitsch, R., Frotscher, M., (1991), ‘Maintenance of peripheral dendrites of GABAergic neurons requires specific input’, Brain Res. 554, 304–307. [20] Markram, H., Lübke, J., Frotscher, M., Sakmann, B., (1997), ‘Regulation of synaptic efficacy by coincidence of post synaptic APs and EPSPs’, Science 275, 213–215. [21] Maletic-Savatic, M., Malinow, R., Svoboda, K., (1999), ‘Rapid dendritic morphogenesis in CA1 hippocampal dendrites induced by synaptic activity’, Science 283, 1923–1927. [22] Lendvai, B., Stern, E.A., Chen, B., Svoboda, K., (2000), ‘Experiencedependent plasticity of dendritic spines in the developing rat barrel cortex in vivo’, Nature 404, 876–881. [23] Stern, E.A., Maravall, M., Svoboda, K., (2001), ‘Rapid development and plasticity of layer 2/3 maps in rat barrel cortex in vivo’, Neuron 31, 305–315. [24] Engert, F., Bonhoeffer, T., (1999), ‘Dendritic spine changes associated with hippocampal long term synaptic plasticity’, Nature 399, 66–70. [25] Ahissar, E., Vaadia, E., Ahissar, M., Bergman, H., Arieli, A., Abeles, M., (1992), ‘Dependence of cortical plasticity on correlated activity of single neurons and on behavioral context’, Science 257, 1412–1415. [26] Noble, D., (2002), ‘Modeling the heart—from genes to cells to the whole organ’, Science, 295, 1678–1682. [27] Feld, Y., Melamed-Frank, M., Kehat, I., Tal, D., Marom, S., Gepstein, L., (2002), ‘Electrophysiological modulation of cardiomyocytic tissue by transfected fibroblasts expressing potassium channels: a novel strategy to manipulate excitability’, Circulation 105, 522–529.

340 0

NETWORK ON CHIPS

[28] Rohr, S., Kucera, J.P., Kleber, A.G., (1997), ‘Form and function: Impulse propagation in designer cultures of cardiomyocytes’, News in Physiological Sciences 12, 171–177. [29] Spach, M.S., Josephson, M.E., Initiating reentry: The role of nonuniform anisotropy in small circuits’, J. Cardiovasc. Electrophysiol. 5, 182–209 (1994). [30] Spach, M.S., Kootsey, J.M., Sloan, J.D., (1982), ‘Active modulation of electrical coupling between cardiac cells of the dog. A mechanism for transient and steady state variations in conduction velocity’, Circ. Res. 51, 347–362. [31] Guevara, M.R., Glass, L., Shrier, A., (1981), ‘Phase locking, perioddoubling bifurcations, and irregular dynamics in periodically stimulated cardiac cells’, Science, 214, 1350–1353. [32] Wise, K.D., Angell, J.B., Starr, A., (1970), ‘An integrated circuit approach to extra–cellular micr oelectrodes’, IEEE Trans. Biomed. Eng. BME-17, 238–246. [33] Bergveld, P., (1972), ‘Development, operation and application of the ISFET as a tool for electrophysiology’, IEEE Trans. Biomed. Eng. BME-19, 342–351. [34] Thomas, C.A., Springer, P.A., Loeb, G.W., Berwald-Netter, Y., Okun, L.M., (1972), ‘A miniature micro electrode array to monitor the bioelectric activity of cultured cells’, Exp. Cell Res. 74, 61–66. [35] Bergveld, P., Wiersma, J., Maertens, H., (1976), ‘Extracellullar potential recording by means of a Field effect transistor without gate metal, called OSFET’, IEEE Trans. Biomed. Eng. BME-23, 136–144. [36] Gross, G.W., Rieske, E., Kreutzberg, G.W., Meyer, A., (1977), ‘A new fixed-array multi-micro electrode system designed for long-term monitoring of extra–cellular single unit neuronal activity in vitro’, Neurosci. Lett. 6, 101–105. [37] Gross, G., (1979), ‘Simultaneous single unit recording in vitro with a photoetched laser deinsulated gold multi-electrode surface’, IEEE Trans. Biomed. Eng. BME-26, 273–279. [38] Israel, D.A., Barry, W.H., Edell, D.J., Mark, R.G., (1984), ‘An array of micr oelectrodes to stimulate and record from cardiac cells in culture’, Am. J. Physiol. 247, H669–H674. [39] Najafi, K., Wise, K.D., Mochizuki, T., (1985), ‘A high-yield ICcompatible multi-channel recording array’, IEEE Transactions on Electron Devices ED-32, 1206–1211. [40] Gross, G.W., Wen, W., Lin, J.W., (1985), ‘Transparent indium–tin oxide electrode patterns for extra–cellular, multi-site recording in neuronal cultures’, J. Neurosci. Meth. 15, 243–252.

NETWORK ON CHIPS

341

[41] Droge, M.H., Gross, G.W., Hightower, M.H., Czisny, L.E., (1986), ‘Multi-electrode analysis of coordinated, multi-site, rythmic bursting in cultured CNS monolayer networks’, J. Neurosci. 6, 1583–1592. [42] Najafi, K., Wise, K.D., (1986), ‘An implantable multi-electrode array with on-chip signal processing’, IEEE J. Solid State Circ. SC-21, 1035–1044. [43] Rohr, S., (1990), ‘A computerized device for long term measurements of the contraction frequency of cultured rat heart cells under stable incubating conditions, Pflug. Arch. Eur. J. Phy. 416, 201–206. [44] Rohr, S., Scholly, D.M., Kleber, A.G., (1991), ‘Patterned growth of neonatal rat heart cells in culture morphological and electrophysiological characterization’, Circ. Res. 68, 114–130. [45] Mastrototaro, J.J., Massoud, H.Z., Pilkington, T.C., Ideker, R.E., (1992), ‘Rigid and flexible thin film multi-electrode arrays for transmural cardiac recording’, IEEE Trans. Biomed. Eng. 39, 271– 379. [46] Eckhorn, R., Thomas, U., (1993), ‘A new method for the insertion of multiple microprobes into neural and muscular tissue, including fiber electrodes, fine wires, needles and microsensors’, J. Neurosci. Meth. 49, 175–179. [47] Hoogerwerf, A.C., Wise, K.D., (1994), ‘A three-dimensional micro electrode array for chronic neural recording’, IEEE Trans. Biomed. Eng. 41, 1136–1146. [48] Nádasny, Z., Csicsvari, J., Penttonen, M., Hetke, J., Wise, K., Buzsáki, G., (1998), ‘Extra–cellular recording and analysis of neuronal activity: from single cells to ensembles, in: Neuronal Ensembles: Startegies for recording and decoding’, edited by H. B. Eichenbaum and J.L. Davis, (Wiley-Liss, New York), 17–55. [49] Mohr, G., Hofer, E., Plank, G., (1999), ‘A new real-time mapping system to detect microscopic cardiac excitation patterns’, Biomed. Instrum. Technol. 33, 455–461. [50] Sprössler, C., Denyer, M., Britland, S., Knoll, W., Offenhäusser, A., (1999), ‘Electrical recordings from rat cardiac muscle cells using field-effect transistors’, Phys. Rev. E 60, 2171–2176. [51] Bai, Q., Wise, K.D., (2001), ‘Single-unit neural recording with active micro electrode arrays’, IEEE Trans. Biomed. Eng. 48, 911–920. [52] Nicolelis, M.A., Dimitrov, D., Carmena, J.M., Crist, R., Lehew, G., Kralik, J.D., Wise, S.P., (2003), ‘Chronic, multi-site, multi-electrode recordings in macaque monkeys’, Proc. Natl. Acad. Sci. USA, 100, 11041–11046. [53] Tominaga, T., Tominaga, Y., Ichikawa, M., (2001), ‘Simultaneous multi-site recordings of neural activity with an inline multielectrode

342 2

[54] [55] [56]

[57] [58] [59] [60] [61] [62]

[63] [64] [65] [66]

NETWORK ON CHIPS array and optical measurement in rat hippocampal slices’, Pflug. Arch. Eur. J. Phy. 443, 317–322. Cohen, D., Yarom, Y., (1999), ‘Optical measurements of synchronized activity in isolated mammalian cerebellum’, Neuroscience 94, 859–866. Ebner, T.J., Chen, G., (1995), ‘Use of voltage-sensitive dyes and optical recordings in the central nervous system’, Prog. Neurobiol. 46, 463–506. Rohr, S., Salzberg, B.M., (1994), ‘Multiple site optical recording of transmembrane voltage (MSORTV) in patterned growth heart cell cultures: Assessing electrical behavior, with microsecond resolution, on a cellular and subcellular scale’, Biophys. J. 67, 1301–1315. Grinvald, A., Frostig, R.D., Siegel, R.M., Bartfeld, E., (1991), ‘HighResolution Optical Imaging of Functional Brain Architecture in the Awake Monkey’, Proc. Natl. Acad. Sci. USA, 88, 11559–11563. Grinvald, A., (1985), ‘Real time optical mapping of neuronal activity: From single growth cones to intact mammalian brain’, Ann. Rev. Neurosci. 8, 263. Windisch, H., Mueller, W., Tritthart, H.A., (1985), ‘Fluorescence monitoring of rapid changes in membrane potential in heart muscle’, Biophys. J. 48, 877–884. Vassanelli, S., Fromherz, P., (1997), ‘Neurons from rat brain coupled to transistors’, Applied Physics A–Solids & Surfaces 65, 85–88. Fromherz, P., Offenhäusser, A., Vetter, T., Weis, J., (1991), ‘A neuron-silicon junction: a Retzius cell of the leech on an insulatedgate field-effect transistor’, Science, 252, 1290. Fromherz, P., (2003), ‘Neuroelectronic interfacing: semiconductor chips with ion channels, nerve cells and brain’, in: Nano-electronics and Information Technology, Waser, R., (ed.), Wiley-VCH, Berlin, 781–810. Zeck, G., Fromherz, P., (2001), ‘Noninvasive neuroelectronic interfacing with synaptically connected snail neurons immobilized on a semiconductor chip’, Proc. Natl. Acad. Sci. USA, 98, 10457–10462. Hetke, J., Ferson, D.J., (2004), ‘Silicon electrodes for extra–cellular recording, in: Handbook of neuroprosthetic methods’, Peterson, J.K., Finn, W.E., Lopresti, P.G., (eds.), CRC Press, Boca Raton, 163–191. Schwartz, A.B., (2004), ‘Cortical neural prosthetics’, Annual Review of Neuroscience 27, 487–507. Rutten, W.L., Smit, J.P., Frieswijk, T.A., Bielen, J.A., Brouwer, A.L., Buitenweg, J.R., Heida, C., (1999), ‘Neuro-electronic interfacing with multi-electrode arrays’, IEEE Eng. Med. Biol. Mag. 18, 47–55.

NETWORK ON CHIPS

343

[67] Bi, G.Q., Poo, M.M., (1998), ‘Synaptic modifications in cultured hippocampal neurons: dependence on spike timing, synaptic strength, and post synaptic cell type’, J. Neurosci. 18, 10464–10472. [68] Eytan, D., Brenner, N., Marom, S., (2003), ‘Selective adaptation in networks of cortical neurons’, J. Neurosci. 23, 9349–9356. [69] Marom, S., Shahaf, G., (2002), ‘Development, learning and memory in large random networks of cortical neurons: lessons beyond anatomy’, Q. Rev. Biophys. 35, 63–87. [70] Shahaf, G., Marom, S., (2001), ‘Learning in networks of cortical neurons’, J. Neurosci. 21, 8782–8788. [71] Tal, D., Jacobson, E., Lyakhov, V., Marom, S., (2001), ‘Frequency tuning of input-output relation in a rat cortical neuron in vitro’, Neurosci. Lett. 300, 21–24. [72] Jimbo, Y., Robinson, H.P., (2000), ‘Propagation of spontaneous synchronized activity in cortical slice cultures recorded by planar electrode arrays’, Bioelectrochemistry 51, 107–115. [73] Jimbo, Y., Kawana, A., Parodi, P., Torre, V., (2000), ‘The dynamics of a neuronal culture of dissociated cortical neurons of neonatal rats’, Biol. Cybern. 83, 1–20. [74] van Ooyen, A., (ed.), (2003), ‘Modeling Neural Development’, The MIT Press, Cambridge MA, London. [75] Streit, J., Tscherter, A., Heuschkel, M.O., Renaud, P., (2001), ‘The generation of rhythmic activity in dissociated cultures of rat spinal cord’, Eur. J. Neurosci. 14, 191–202. [76] Tscherter, A., Heuschkel, M.O., Renaud, P., Streit, J., (2001), ‘Spatiotemporal characterization of rhythmic activity in rat spinal cord slice cultures’, Eur. J. Neurosci. 14, 179–190. [77] Gross, G.W., Harsch, A., Rhoades, B.K., Göpel, W., (1997), ‘Odor, drug and toxin analysis with neuronal networks in vitro: extra–cellular array recording of network responses’, Biosens. Bioelectron. 12, 373– 393. [78] Gross, G.W., Rhoades, B.K., Azzazy, H.M., Wu, M.C., (1995), ‘The use of neuronal networks on multi-electrode arrays as biosensors’, Biosens. Bioelectron. 10, 553–567. [79] Gramowski, A., Schiffmann, D., Gross, G.W., (2000), ‘Quantification of acute neurotoxic effects of trimethyltin using neuronal networks cultured on micro electrode arrays’, Neurotoxicology, 21, 331–342.

344 4

NETWORK ON CHIPS

[80] Lipton, P., Aitken, P.G., Dudek, F.E., Eskessen, K., Espanol, M.T., Ferchmin, P.A., Kelly, J.B., Kreisman, N.R., Landfield, P.W., Larkman, P.M., (1995), ‘Making the best of brain slices: comparing preparative methods’, J. Neurosci. Methods 59, 151–156. [81] Alger, B.E., Dhanjal, S.S., Dingledine, R., Garthwaite, J., Henderson, G., King, G.L., Lipton, P., North, A., Schwartzkroin, P.A., Sears, T.A., Segal, M., Whittingham, T.S., Williams, J., (1984), ‘Brain Slice Methods’, in: Brain Slices, Dingledine, R., (ed.), Plenum Press, New York, 406. [82] Gaehwiler, B.H., (1988), ‘Organotypic cultures of neural tissue’, Trends Neurosci. 11, 484–489. [83] Robert, F., Corrèges, P., Duport, S., Stoppini, L., (2001), ‘Combined electrophysiology and micro-dialysis on hippocampal slice cultures using the physiocard®system’, Current Separations 16, 3–10. [84] Stoppini, L., Duport, S., Corrèges, P., (1997), ‘A new extra–cellular multi-recording system for electrophysiological studies: application to hippocampal organotypic cultures. J. Neurosci. Meth. 72, 23–33. [85] Thiébaud, P., de Rooij, N.F., Koudelka-Hep, M., Stoppini, L., (1997), ‘Micro electrode arrays for electrophysiological monitoring of hippocampal organotypic slice cultures’, IEEE Trans. Biomed. Eng. 44, 1159–1163. [86] Stoppini, L., Buchs, P.A., Muller, D., (1991), ‘A simple method for organotypic cultures of nervous tissue’, J. Neurosci. Meth. 37, 173–182. [87] Egert, U., Schlosshauer, B., Fennrich, S., Nisch, W., Fejtl, M., Knott, Th., Müller, T., Hämmerle, H., (1998), ‘A novel organotypic long term culture of the rat hippocampus on substrate-integrated multielectrode arrays’, Brain Res. Prot. 2, 229–242. [88] Knott, Th., (2001), ‘Population synchronization during propagation of epileptiform activity in organotypic hippocampal slices—a micro electrode array study’, Der Andere Verlag, Osnabrueck. [89] Malmivuo, J., Plonsey, R., (1995), ‘Bioelectromagnetism: Principles and applications of bioelectric and biomagnetic fields’, Oxford University Press, USA. [90] Heuschkel, M.O., Fejtl, M., Raggenbass, M., Bertrand, D., Renaud, P., (2002), ‘A three-dimensional multi-electrode array for multi-site stimulation and recording in acute brain slices’, J. Neurosci. Methods 114, 135–148. [91] Heuschkel, M.O., (2001), ‘Fabrication of multi-electrode array devices for electrophysiological monitoring of in vitro cell/tissue cultures’, in Series in Microsystems, Besse, P.A., Gijs, M., Popvic, R.S., Renaud, P., Hartung-Gorre Verlag, Konstanz.

NETWORK ON CHIPS

345

[92] Janders, M., Egert, U., Stelzle, M., Nisch, W., (1996), ‘Novel thin film titanium nitride micro-electrodes with excellent charge transfer capability for cell stimulation and sensing applications’, Bridging Disciplines for Biomedicine, Proceedings of the 18th Annual Conference of the IEEE Engineering in Medicine and Biology Society, 48–49. [93] Rosenberg, L.J., Jordan, R.S., Gross, G.W., Emery, D.G., Lucas, J.H., (1996), ‘Effects of methylprednisolone on lesioned and uninjured mammalian spinal neurons: viability, ultrastructure, and network electrophysiology’, J. Neurotrauma 13, 417–437. [94] Jimbo, Y., Tateno, T., Robinson, H.P., (1999), ‘Simultaneous induction of pathway-specific potentiation and depression in networks of cortical neurons’, Biophys. J. 76, 670–678. [95] Kamioka, H., Maeda, E., Jimbo, Y., Robinson, H.P.C., Kawana, A., (1996), ‘Spontaneous periodic synchronized bursting during formation of patterns of connections in cortical cultures’, Neurosci. Lett. 206, 109–112. [96] Potter, S.M., (2001), ‘Distributed processing in cultured neuronal networks’, Prog. Brain Res. 130, 49–62. [97] Potter, S.M., DeMarse, T.B., (2001), ‘A new approach to neural cell culture for long term studies’, J. Neurosci. Meth. 110, 17–24. [98] Wong, R.O.L., Meister, M., Shatz, C.J., (1993), ‘Transient period of correlated bursting activity during development of the mammalian retina’, Neuron 11, 923–938. [99] Meister, M., Wong, R.O.L., Baylor, D.A., Shatz, C.J., (1991), ‘Synchronous bursts of action potentials in ganglion cells of the developing mammalian retina’, Science 252, 939–943. [100] Egert, U., Heck, D., Aertsen, A., (2002), ‘2-Dimensional monitoring of spiking networks in acute brain slices’, Exp. Brain Res. 142, 268–274. [101] Grattarola, M., Martinoia, S., (1993), ‘Modeling the neuron-microtransducer junction—from extra–cellular to patch recording’, IEEE Trans. Biomed. Eng. 40, 35–41. [102] Rall, W., (1969), ‘Distribution of potential in cylindrical coordinates and time constants for a membrane cylinder’, Biophys. J. 9, 1509– 1541. [103] Plonsey, R., (1977), ‘Action potential sources and their volume conductor fields’, Proc. IEEE 65, 601–611. [104] Henze, D.A., Borhegyi, Z., Csicsvari, J., Mamiya, A., Harris, K.D., Buzsáki, G., (2000), ‘Intra–cellular features predicted by extra– cellular recordings in the hippocampus in vivo’, J. Neurophysiol. 84, 390–400.

346 6

NETWORK ON CHIPS

[105] Halbach, M.D., Egert, U., Hescheler, J., Banach, K., (2003), ‘Estimation of action potential changes from field potential recordings in multi-cellular mouse cardiac myocyte cultures’, Cellular Physiology and Biochemistry 13, 271–284. [106] Shimono, K., Brucher, F., Granger, R., Lynch, G., Taketani, M., (2000), ‘Origins and distribution of cholinergically induced beta rhythms in hippocampal slices’, J. Neurosci. 20, 8462–8473. [107] Mitzdorf, U., Singer, W., (1977), ‘Laminar segregation of afferents to lateral geniculate nucleus of the cat: an analysis of current source density’, J. Neurophysiol. 40, 1227–1244. [108] Novak, J.L., Wheeler, B.C., (1989), ‘Two-dimensional current source density analysis of propagation delays for components of epileptiform bursts in rat hippocampal slices’, Brain Res. 497, 223–230. [109] Wheeler, B.C., Novak, J.L., (1986), ‘Current source density estimation using micro electrode array data from the hippocampal slice preparation’, IEEE Trans. Biomed. Eng. BME-33, 1204–1213. [110] Freeman, J.A., Nicholson, C., (1975), ‘Experimental optimization of current source-density technique for anuran cerebellum’, J. Neurophysiol. 38, 369–382. [111] Nicholson, C., Llinas, R., (1975), ‘Real time current source-density analysis using multi-electrode array in cat cerebellum’, Brain Res. 100, 418–424. [112] Haberly, L.B., Shepherd, G.M., (1973), ‘Current density analysis of summed evoked potentials in opossum prepyriform cortex’, J. Physiol. (Lond.) 36, 789–803. [113] Nicholson, C., Llinas, R., (1971), ‘Field potentials in the alligator cerebellum and theory of their relationship to Purkinje cell dendritic spikes’, J. Neurophysiol. 34, 509–531. [114] Stett, A., Egert, U., Guenther, E., Hofmann, F., Meyer, Th., Nisch, W., Hämmerle, H., (2003), ‘Biological applications of micro-electrode arrays in drug discovery and basic research’, Analytical and Bioanalytical Chemistry 377, 486–495. [115] Tian, N., Copenhagen, D.R., (2001), ‘Visual deprivation alters development of synaptic function in inner retina after eye opening’, Neuron 32, 439–449. [116] Tian, N., Copenhagen, D.R., (2003), ‘Visual stimulation is required for refinement of ON and OFF pathways in postnatal retina’, Neuron 39, 85–96. [117] Streit, J., (1993), ‘Regular oscillations of synaptic activity in spinal networks invitro’, J. Neurophysiol. 70, 871–878.

NETWORK ON CHIPS

347

[118] Tousson E., Meissl, H., (2004), ‘Supra-chiasmatic nuclei grafts restore the circadian rhythm in the paraventricular nucleus of the hypothalamus’, J. Neurosci. 24, 2983–2988. [119] Egert, U., Hämmerle, H., (2002), ‘Application of the micro electrodearray (MEA) technology in pharmaceutical drug research, in: Sensoren im Fokus neuer Anwendungen’, edited by Baselt, J.P., Gerlach, G., w.e.b. Universitätsverlag, Dresden, 51–54. [120] Knott, Th., Fejtl, M., Schlosshauer, B., Egert, U., Aertsen, A., Hämmerle, H., (1999), ‘Long term monitoring of in vitro epileptogenesis: study of e/s-potentiation in organotypic hippocampal slices using a micro-electrode-array’, Soc. Neurosci. Abstr. 25, 217.5. [121] Knott, Th., Egert, U., Hehl, U., Hämmerle, H., Aertsen, A., (2000), ‘Synchronization in propagating epileptiform population spikes in organotypic hippocampal slices—experiments and models’, 22. Reutlingen. International meeting on substrate-integrated micro electrode arrays: hard-, soft-, and wetware. [122] Gray, C.M., König, P., Engel, A.K., Singer, W., (1989), ‘Oscillatory responses in cat visual cortex exhibit inter- columnar synchronization which reflects global stimulus properties’, Nature 338, 334–337. [123] Bouthenet, M.L., Souil, E., Martres, M.P., Sokoloff, P., Giros, B., Schwartz, J.C., (1991), ‘Localization of dopamine D3 receptor mRNA in the rat brain using in situ hybridization histochemistry: comparison with dopamine D2 receptor mRNA’, Brain Res. 564, 203–219. [124] Jost, B., Aertsen, A., Egert, U., (2002), ‘Dopamine D2/D3 receptor mediated modulation of cerebellar spike activity’, FENS Abstr. 1, A012.11. [125] Jost, B., (2001), ‘Dopamine receptor mediated modulation of cerebellar spike activity’, Diploma, Universität Freiburg. [126] Redfern, W.S., Carlsson, L., Davis, A.S., Lynch, W., MacKenzie, G.I., Palethorpe, S., Siegl, P.K., Strang, I., Sullivan, A. T., Wallis, R., Camm, A.J., Hammond, T.G., (2003), ‘Relationships between preclinical cardiac electrophysiology, clinical QT interval prolongation and torsade de pointes for a broad range of drugs: evidence for a provisional safety margin in drug development’, Cardiovasc. Res. 58, 32–45. [127] Fermini, B., Fossa, A.A., (2003), ‘The impact of drug-induced QT interval prolongation on drug discovery and development’, Nat. Rev. Drug Discov. 2, 439–447. [128] Vieweg, W.V.R., (2002), ‘Mechanisms and risks of electrocardiographic QT interval prolongation when using antipsychotic drugs’, J. Clin. Psychiatry 63, 18–24.

348 8

NETWORK ON CHIPS

[129] Igelmund, P., Fleischmann, B.K., Fischer, I.V., Soest, J., Gryshchenko, O., Sauer, H., Liu, Q., Hescheler, J., (1999), ‘Action potential propagation failures in long term recordings from embryonic stem cell-derived cardiomyocytes in tissue-culture’, Plug. Arch. Eur. J. Phy. 437, 669–679. [130] Rohr, S., Kucera, J.P., Fast, V.G., Kleber, A.G., (1997), ‘Paradoxical improvement of impulse conduction in cardiac tissue by partial cellular uncoupling’, Science, 275, 841–844. [131] Lazarra, R., (1993), ‘Antiarrhythmic drugs and torsade de pointes’, Eur. Heart J. 44, 88–92. [132] Meyer, T., Boven, K.-H., Guenther, E., Fejtl, M., (2004), ‘MicroElectrode Arrays (MEA)—a novel tool in cardiac safety pharmacology to study QT-prolongation’, Drug Safetyy . [133] Bar-Gad, I., Ritov, Y., Vaadia, E., Bergman, H., (2001), ‘Failure in identification of overlapping spikes from multiple neuron activity causes artificial correlations’, J. Neurosci. Meth. 107, 1–13. [134] Wheeler, B.C., Heetderks, W.J., (1982), ‘A comparison of techniques for classification of multiple neural signals’, IEEE Trans. Biomed. Eng. 29, 752–759. [135] Ben Shaul, Y., Bergman, H., Ritov, Y., Abeles, M., (2001), ‘Trial to trial variability in either stimulus or action causes apparent correlation and synchrony in neuronal activity’, J. Neurosci. Methods 111, 99– 110. [136] Abeles, M., (1982), ‘Quantification, smoothing, and confidence limits for single-unitshistograms’, J. Neurosci. Methods 5, 317–325. [137] Beggs, J.M., Plenz, D., (2004), ‘Neuronal avalanches in neocortical circuits’, J. Neurosci. 23, 11167–11177. [138] Banach, K., Halbach, M., Hu, P., Hescheler, J., Egert, U., (2003), ‘Development of electrical activity in cardiac myocyte aggregates derived from mouse embryonic stem cells’, Am. J. Physiol. 284, H2114–H2123. [139] Kehat, I., Gepstein, A., Spira, A., Itskovitz-Eldor, J., Gepstein, L., (2002), ‘High-resolution electrophysiological assessment of human embryonic stem cell-derived cardiomyocytes: a novel in vitro model for the study of conduction’, Circ. Res. 91, 659–661. [140] Kehat, I., Karsenti, D., Amit, M., Druckmann, M., Feld, Y., ItskovitzEldor, J., Gepstein, L., (2001), ‘Long term, high-resolution, electrophysiological assessment of human embryonic stem cell derived cardiomyocytes: a novel in vitro model for the human heart’. [141] Stett, A., Barth, W., Weiss, S., Hämmerle, H., Zrenner, E., (2000), ‘Electrical multi-site stimulation of the isolated chicken retina’, Vision Res. 40, 1785–1795.

NETWORK ON CHIPS

349

[142] Zrenner, E., Stett, A., Weiss, S., Aramant, R.B., Guenther, E., Kohler, K., Miliczek, K.D., Seiler, M.J., Hämmerle, H., (1999), ‘Can subretinal micro-photodiodes successfully replace degenerated photoreceptors?’, Vision Res. 39, 2555–2567. [143] Boppart, S.A., Wheeler, B.C., Wallace, C.S., (1992), ‘A flexible perforated micro electrode array for extended neural recordings’, IEEE Trans. Biomed. Eng. 39, 37–42. [144] Gholmieh, G., Soussou, W., Courellis, S., Marmarelis, V., Berger, T., Baudry, M., (2001), ‘A biosensor for detecting changes in cognitive processing based on nonlinear systems analysis’, Biosens. Bioelectron. 16, 491–501.

Chapter 11 BIO–NANO-SYSTEMS Overview and Outlook PD Dr. Thomas Nann, Dipl. Chem. Jürgen Riegeler, Prof. G. Urban Freiburg Materials Research Center (FMF), Stefan-Meier-Str. 21, D-79104 Freiburg

Abstract:

This chapter describes basic concepts and experimental methods for the application of nano materials to biology in general and bioanalytics in particular. Furthermore, some examples for applications are given. Optical properties of metal and semiconducting nanocrystals are discussed from the viewpoint of bioanalytical applications. A picture of the conjugation of nanocrystals to biomolecules is given and the fundamental methods for DNA detection and immuno assays are described. Moreover, the possibilities for the application of nano crystals to imaging purposes in biology are presented together with selected examples. Finally, an outlook on future methods is suggested.

Key words:

Quantum Dots, luminescent nanocrystals, colloidal metals, immuno assay, DNA chip, gene chip.

1.

INTRODUCTION

The popular novel ‘Prey’, written by Michael Crichton[1], describes a scenario in which modified bacteria produce an autonomous (almost ‘living’) swarm of artificial nanoparticles. Highly integrated electromechanical nano-systems are produced with biotechnological methods. This intriguing idea summarizes the vision and the content of bio-nanotechnology (perhaps it must be mentioned that the current real and potential applications for bio-nano-systems under-achieve these expectations to a large extent). Bio-nano-system is a compound word combining biology and nanotechnology. In other words, it covers the field of application of nano technology in biology or vice versa. Furthermore, nano technology can be defined as a technology concerning the application of nanomaterials.


351

352 2

BIO–NANO-SYSTEM MS

Therefore, bio-nano-systems are applications of nanomaterials in the field of biology. Nanomaterials are materials which are characterized by a confinement of characteristic elements (e.g., charge carriers) in at least one dimension on an atomistic or molecular level, e.g., gold colloids, with their characteristic size-dependent plasmon resonance, are nanomaterials. Even though proteins and DNA have dimensions in the lower nano-meter range, they are—according to this definition—not nanomaterials, because their properties are not based on a spatial confinement. Thus a gold nanocrystal coupled to a protein is an example for a bio-nano-system. The application potential of nanomaterials in biology relies largely on the advantageous properties of nanomaterials, and that their size corresponds to that of large proteins. Mostly these advantageous properties are optical or magnetic, and the bio-nano-system is a conjugate between a (usually inorganic) nanoparticle and a biomolecule (oligonucleotide or protein). Thus the field of applications for bio-nano-systems becomes restricted to the bioanalytics and medical diagnostics. Actually almost all publications concerning this field are related to bioanalytics, although some interesting efforts were undertaken to use mobile proteins—e.g., ATPase—for mechanical actuation or carbon nano-tubes (CNTs) for different purposes. This chapter is divided in two major parts: first, the basic concepts and experimental methods for the preparation and application of bionano-systems are introduced in section 2. Second, some exemplary applications are discussed in section 3. Finally, a look beyond the state of the art is undertaken on potential future research and application concepts.

2.

BASIC CONCEPTS AND EXPERIMENTAL METHODS

In general the application of nanomaterials to bioanalytics is currently a replacement of organic dyes, radioactive or metal labels and contrast agents by metal, oxide or luminescent nanocrystals. Therefore the bioanalytical and imaging methods (which are briefly described in the sections 2.6 and 2.7) remain mostly unchanged, whereas the tagged or labelled biomolecule is replaced by a bio-nano-system. The conjugation between biomolecule and nanocrystal—which is discussed in section 2.5—is crucial for every bionano-system as it determines the overall biological properties of the conjugate. The self-assembly of monomolecular layers—which underlies the conjugation to a large extend—is discussed in section 2.1. The optical properties of the bio-nano-system (described in the sections 2.2 and 2.3) are determined by the nanocrystals itself.

BIO–NANO-SYSTEMS

353

Nanocrystal–biomolecule conjugates dominate in the literature to a large extent. Nevertheless, there are numerous interesting publications describing singular experiments or applications and which are not included here. The ‘nano-barcode’ is such an example [2].

2.1

Self-assembly

Self-assembly is an important tool for the build-up of nano-systems. Usually two kinds of materials are self-assembled: First, molecules are selfassembled onto surfaces by chemisorption. Second, nanomaterials are selfassembled to larger units (e.g., quantum-aggregates). For bio-nano-systems the self-assembly of molecules to self-assembled monolayers (SAMs) plays an important role. SAMs are used to modify the surface properties of nanomaterials: They can passivate surfaces or provide functional groups to maintain solubility or coupling groups for biomolecules. A self-assembled monolayer consists of a surface and a layer of molecules that are attached to the surface by means of a coupling group. Usually, these monolayers are prepared by treatment of the surface with a solution of the molecules in question. Fig. 11-1 describes schematically the exchange of a physisorbed SAM (green triangles) with a chemisorbed SAM (red rods). This process is the common method to modify surface properties of nanocrystals.

A

B

C

Figure 11-1. Schematic description of the chemisorption of a molecular species (red rods) onto the surface of spherical nanocrystals. A second monolayer (green triangles) is therefore desorbed.

There is a difference between physisorption and chemisorption. In the case of physisorption the molecules are bound weakly to the surface (usually

354 4


by van-der-Waals forces). Chemisorption has an almost covalent character. The strength of a surface bond is expressed with the adsorption enthalpy. In the case of physisorption, the adsorption enthalpy is in the range of a condensation enthalpy, in case of chemisorption of a reaction enthalpy. In most cases the Langmuir Adsorption Isotherm describes the adsorption process sufficiently3. As a result, in order to exchange a surface it is advantageous to add the adsorbing molecules in excess and conduct the exchange at low temperatures.

2.2

Optical properties of semiconducting nanocrystals

The increasing interest in semiconducting nanocrystals is primarily caused by the opto-electronic properties of these materials. For some applications their optical properties have significant advantages compared with the frequently used organic fluorophores: nanocrystals show a high brilliance (product of extinction coefficient and quantum yield), the emission maxima are continuously tuneable owing to quantum size effects (Fig. 11-2 depicts CdSe nanocrystals of different sizes), they have a broad absorption characteristic (displayed in Fig. 11-3), the emission has a narrow spectral linewidth (also shown in Fig. 11-3; typical full width half maxima (fwhm) of 25 nm are reached), and they do not show photobleaching.

Figure 11-2. Photo luminescence of CdSe-nano rods of different sizes (from [4]).

Nevertheless, luminescent nanocrystals have some limiting properties for certain applications. Initially they show an increasing luminescence under excitation. This effect is well known in solid state physics and can be attributed to traps (surface, intrinsic, impurities, etc.) in the nanocrystals. A competition between band edge luminescence and alternative decay paths over traps occurs under excitation. Gradually the traps are saturated, and band edge luminescence is preferred. When all traps are saturated or there is

BIO–NANO-SYSTEMS

355

Figure 11-3. Photoluminescence (PL) and absorption spectrum of CdSe-nanocrystals.

an equilibrium between band edge luminescence and other forms of recombination, the luminescence remains constant. This effect is not necessarily disadvantageous, especially as long as the traps can be saturated quickly by high power excitation (e.g., by means of a laser pulse), but it has to be taken into account in the case of quantitative analysis. Thermoquenching can be another disadvantage for some biological applications. The luminescence of semiconductor nanocrystals decreases with increasing temperature—most probably because of increased electronphonon coupling at higher temperatures. As long as most bioanalytical experiments are performed under isothermal conditions, the thermoquenching can be usually neglected [5]. Spherical semiconductor nanocrystals have typical sizes between 3 and 10 nm, depending primarily on their surface derivatization. Assumed mean sizes of 4 nm, these crystals have dimensions in the same range as large proteins. Therefore, they can affect the biological functionality of the process under study and transport, e.g., through cell membranes could become a crucial parameter. Size matters mostly in living and/or dynamic systems, but is usually uncritical with static situations, e.g., biochip applications. Despite these problems, luminescent nanocrystals are promising fluorescent labels for many bio-analytical applications. Currently such nanocrystals are commercially available as streptavidin conjugates or in native, water-soluble form (to the best of our knowledge three companies sell luminescent nanocrystals for bio-analytical applications: Quantum Dot

356 6


Corp., Hayward, CA, USA; Evident Technologies, Troy, NY, USA and Biocrystal, Westerville, OH, USA).

2.3

Optical properties of metal nanocrystals

Whereas the optical properties of semiconducting nanocrystals are dominated by their luminescence, metal nanocrystals can only be detected by means of optical absorbance, (Mie) scattering, spectral shift or surface plasmon resonance (SPR) imaging [6]. Metal nanocrystals have extinction coefficients between 106 and 1012 (Mcm)-1—which is many orders of magnitude larger than those of organic dyes. The optical absorption of metal nanocrystals is caused by resonance of electrons in the conducting band (so called ‘plasmons’) and depends on the metal, the morphology and the size of the nanocrystals. Fig. 11-4 shows a photographic image of colloidal silver and gold nanocrystals. The corresponding absorbance spectra are displayed in Fig. 11-5. Quantitative analysis can therefore be done by detection of the absorbance, because— like semiconducting nanocrystals—metal nanocrystals do not show photobleaching.

Figure 11-4. From left to right: Colloidal silver, gold and coagulated gold nanocrystals in aqueous suspension.

Beside their large extinction coefficients, many metal nanocrystals have very large scattering coefficients. For instance, silver nanocrystals have scattering cross-sections of up to 10 -10 cm-2. Therefore scattered light from metal nanocrystal-labelled biomolecules can be detected quantitatively and

BIO–NANO-SYSTEMS

357

Figure 11-5. Absortion spectra of silver (a) and gold (b) colloids. Red line: coagulated gold nanocrystals (cf. Fig. 11-4).

sensitively. Similar to absorption, the wavelength and intensity of the scattered light is caused by plasmon resonance and thus depends strongly on the kind, size and morphology of the nanocrystals. Two companies are selling nanocrystal scatterers as DNA array labels (Seashell Technology, La Jolla , CA, USA and Genicon Sciences, San Diego, CA, USA). Plasmon resonance of metal nanocrystals is crucially influenced by interparticle spacing, e.g., a red gold colloid turns to blue on coagulation of the nanocrystals (cf. Fig. 10-5b). This effect can be used to detect the attachment of gold-labelled biomolecules onto surfaces[7]. SPR is a well known technique in biochip technology: Changes of the refractive index of a thin gold film are detected by measuring the deflection angle of a totally reflected laser beam at the ‘bottom’ plane of the film. It has been shown that nanocrystal-labelled biomolecules deflect the laser beam much more drastically then unlabelled targets [8].

2.4

Magnetic nanoparticles

Magnetic resonance imaging (MRI) is an important tool for medical diagnostics. Usual MRI contrast agents are various gadolinium chelates. The

358 8


use of these chelates is not satisfactory, because gadolinium is highly toxic and the improved contrast is weak. Therefore research on new nanoparticle contrast agents such as Fe3O4- or FePt-nanoparticles is being conducted [22]. The application of such magnetic nanoparticles as MRI contrast agents has been successfully shown [23].

2.5

Conjugation of nanomaterials and biomolecules

Whereas the preparation and derivatization of nanoparticles affects the optical and chemical properties, the conjugation affects the biological functionality and the stability of the bio-nano-system. The different methods of conjugation between nanocrystals and biomolecules can be divided into three categories. The first category includes methods in which biomolecules are bound non-covalently to nanoparticles. Therefore nanoparticles are first derivatized with a chemisorbed monolayer or the capping agent from synthesis to have hydrophobic surfaces (resp., properties). In a second step these hydrophobic nanoparticles are precipitated and re-dissolved in water within tensidic micelles. In principle this method works with all common micelle building agents such as phospholipids, sodium dodecylsulfate (SDS), Triton X100 or CHAPS. In a final step biomolecules are coupled covalently to functional groups at the outer sphere of the micelles [9]. Fig. 11-6A depicts schematically this kind of conjugation. A major advantage of this method is that the whole process—from non-polar/polar solvent transfer to the coupling—is relatively easy to be performed. The bond between nanoparticles and biomolecules is based on hydrophobic interactions within the micelles. Therefore the conjugate disintegrates relatively easily. Furthermore, many biological experiments include detergents which leads to destruction (fusion) of the micelles and coagulation of the nanoparticles. Within an improvement of this method the nanoparticles are derivatized with charged surface molecules and biomolecules are attached to polyelectrolytes of charge opposite as that of the nanoparticles. These polyelectrolytes are subsequently linked by ‘Coulomb attraction to the nanoparticles’ surface [10]. The second category contains methods, in which biomolecules are chemisorbed onto nanoparticles by means of a ‘linker’. This can be realized in two variations: first, the biomolecules contain surface active groups such as, e.g., thiols, and are directly chemisorbed onto the nanoparticles (this case

BIO–NANO-SYSTEMS

A

359

B

W

C

O

R S

R

S S R NC C

NC C

R

O

S

NC C

Si S

S R

O

S

Si

S S

R R

W

O

Si O Si O Si

S

O Si O

O O

O

O Si

R

O O O Si O Si

O O

Si

R

O O

Si

R

Figure 11-6. Schematic description of different types of coupling between nanocrystals and biomolecules. W are functional groups to provide water solubility and R biomolecules of coupling groups. A) Coupling by hydrophobic interactions, B) coupling by chemisorption and C) covalent coupling onto a cross-linked surface shell.

is illustrated in Fig. 11-6B). Second, a bifunctional molecule is chemisorbed onto the nanoparticles and biomolecules are coupled to these molecules in a second step[11]—likewise to the micelles from the first category. Chemisorption of thiols onto gold surfaces is well-known and as long as the adsorption energy is

BioMEMS (Microsystems)

BioMEMS and Biomedical Nanotechnology

BioMEMS and Biomedical Nanotechnology

BioMEMS and Biomedical Nanotechnology (Biomems and Biological Nanotechnology)

BioMEMS and Biomedical Nanotechnology

BioMEMS and Biomedical Nanotechnology (Biomems and Biological Nanotechnology)

Laser Diode Microsystems

Sensors and Microsystems

Fundamentals of Microsystems Packaging

Microelectronics, Microsystems and Nanotechnology

Ambient Intelligence with Microsystems: Augmented Materials and Smart Objects (Microsystems)

Fundamentals of Microsystems Packaging

Fundamentals of Microsystems Packaging

Optical Inspection of Microsystems

Microelectronics, Microsystems and Nanotechnology

Electrokinetics and Electrohydrodynamics in Microsystems

Comprehensive Microsystems, Three-Volume Set

Advanced Microsystems Automotive Applications 2004

Advanced microsystems for automotive applications

Sensors and Microsystems: AISEM 2010 Proceedings

CMOS Circuits for Passive Wireless Microsystems

Microsystems Technology: Fabrication, Test & Reliability

Silicon Carbide Microsystems for Harsh Environments

BioMEMS and Biomedical Nanotechnology: Biomolecular Sensing, Processing and Analysis

Laser Diode Microsystems (Microtechnology and MEMS)

Integrated Microsystems: Electronics, Photonics, and Biotechnology

Advanced Microsystems for Automotive Applications 2008

Heteromagnetic Microelectronics: Microsystems of Active Type

Optical Nano- and Microsystems for Bioanalytics

Design and Manufacturing of Active Microsystems

Principles of Bacterial Detection: Biosensors, Recognition Receptors and Microsystems

BioMEMS (Microsystems)

BioMEMS and Biomedical Nanotechnology

BioMEMS and Biomedical Nanotechnology

BioMEMS and Biomedical Nanotechnology (Biomems and Biological Nanotechnology)

BioMEMS and Biomedical Nanotechnology

BioMEMS and Biomedical Nanotechnology (Biomems and Biological Nanotechnology)

Laser Diode Microsystems

Sensors and Microsystems

Fundamentals of Microsystems Packaging

Microelectronics, Microsystems and Nanotechnology

Ambient Intelligence with Microsystems: Augmented Materials and Smart Objects (Microsystems)

Fundamentals of Microsystems Packaging

Fundamentals of Microsystems Packaging

Optical Inspection of Microsystems

Microelectronics, Microsystems and Nanotechnology

Electrokinetics and Electrohydrodynamics in Microsystems

Comprehensive Microsystems, Three-Volume Set

Advanced Microsystems Automotive Applications 2004

Advanced microsystems for automotive applications

Sensors and Microsystems: AISEM 2010 Proceedings

CMOS Circuits for Passive Wireless Microsystems

Microsystems Technology: Fabrication, Test & Reliability

Silicon Carbide Microsystems for Harsh Environments

BioMEMS and Biomedical Nanotechnology: Biomolecular Sensing, Processing and Analysis

Laser Diode Microsystems (Microtechnology and MEMS)

Integrated Microsystems: Electronics, Photonics, and Biotechnology

Advanced Microsystems for Automotive Applications 2008

Heteromagnetic Microelectronics: Microsystems of Active Type

Optical Nano- and Microsystems for Bioanalytics

Design and Manufacturing of Active Microsystems

Principles of Bacterial Detection: Biosensors, Recognition Receptors and Microsystems

Recommend Documents