Nanomedicine Design of Particles, Sensors, Motors, Implants, Robots, and Devices
Artech House Series Engineering in Medicine & Biology Series Editors Martin L. Yarmush, Harvard Medical School Christopher J. James, University of Southampton Advanced Methods and Tools for ECG Data Analysis, Gari D. Clifford, Francisco Azuaje, and Patrick E. McSharry, editors Advances in Photodynamic Therapy: Basic, Translational, and Clinical, Michael Hamblin and Pawel Mroz, editors Biological Database Modeling, Jake Chen and Amandeep S. Sidhu, editors Biomedical Informatics in Translational Research, Hai Hu, Michael Liebman, and Richard Mural Biomedical Surfaces, Jeremy Ramsden Genome Sequencing Technology and Algorithms, Sun Kim, Haixu Tang, and Elaine R. Mardis, editors Inorganic Nanoprobes for Biological Sensing and Imaging, Hedi Mattoussi and Jinwoo Cheon, editors Intelligent Systems Modeling and Decision Support in Bioengineering, Mahdi Mahfouf Life Science Automation Fundamentals and Applications, Mingjun Zhang, Bradley Nelson, and Robin Felder, editors Microscopic Image Analysis for Life Science Applications, Jens Rittscher, Stephen T. C. Wong, and Raghu Machiraju, editors Nanomedicine Design of Particles, Sensors, Motors, Implants, Robots, and Devices, Mark J. Schulz, Vesselin N. Shanov, and Yeoheung Yun, editors Nanoreactor Engineering for Life Sciences and Medicine, Agnes Ostafin and Katharina Landfester, editors Next Generation Artificial Vision Systems: Reverse Engineering the Human Visual System, Maria Petrou and Anil Bharath, editors Systems Bioinformatics: An Engineering Case-Based Approach, Gil Alterovitz and Marco F. Ramoni, editors Systems Engineering Approach to Medical Automation, Robin Felder Translational Approaches in Tissue Engineering and Regenerative Medicine, Jeremy Mao, Gordana Vunjak-Novakovic, Antonios G. Mikos, and Anthony Atala, editors
Nanomedicine Design of Particles, Sensors, Motors, Implants, Robots, and Devices Mark J. Schulz Vesselin N. Shanov Yeoheung Yun Editors
artechhouse.com
Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the U. S. Library of Congress.
British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library.
ISBN-13: 978-1-59693-279-1 Cover design by Igor Valdman
Cover images provided by Robert Freitas, Elena Heister, Thierry Lutz, Hui Zhang, Xuesong Feng, Yi Shu, Faqing Yuan, Peixuan Guo, Yin Yin Guo, Tomomi Nemoto, and Keith A. Crutcher.
© 2009 Artech House. All rights reserved. Printed and bound in the United States of America. No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from the publisher. All terms mentioned in this book that are known to be trademarks or service marks have been appropriately capitalized. Artech House cannot attest to the accuracy of this information. Use of a term in this book should not be regarded as affecting the validity of any trademark or service mark.
10 9 8 7 6 5 4 3 2 1
The coeditors and chapter authors hope that this book will help to free future generations of humanity from disease that mankind has endured for thousands of years. Mark J. Schulz also dedicates his contribution to the memory of his parents, Jeanne and Joseph Schulz, and sister, Margaret Schulz.
Contents Preface
xvii
Outline of the Book
xix
CHAPTER 1 A Nanotechnology Framework for Medical Innovation
1
1.1 Introduction 1.2 Descriptive Systems Modeling 1.2.1 Examples of Descriptive Systems Modeling 1.3 Instrumentation Needed to Develop DSM 1.4 Nanomaterials Made for Medicine 1.5 Implantable Nanomedical Devices 1.6 Nanorobots 1.6.1 Nanorobots for Revolutionizing Medicine 1.6.2 Nanorobot Factory 1.6.3 Biological Nanorobots 1.7 Biodegradable Metals for Temporary Implantable 1.7 Nanomedical Devices 1.8 Integration of Nanodevices in the Body 1.9 Safety and Ethical Implications of Nanomedicine 1.10 Efficiently Working Together Using Shared Resources 1.11 Chapter Summary and Conclusions Problems Acknowledgments References
17 18 19 20 21 22 22 22
PART I Nanoscale Materials and Particles
25
CHAPTER 2 Synthesis of Carbon Nanotube Materials for Biomedical Applications
27
2.1 Introduction to Nanoscale Materials 2.2 Synthesis of Long Carbon Nanotube Arrays 2.3 Characterization of CNT Arrays 2.3.1 Scanning Electron Microscopy and Transmission Electron 2.3.1 Microscopy 2.3.2 Raman Spectroscopy and Thermal Gravimetric Analysis 2.4 Patterned CNT Arrays 2.5 Production Scale Up of CNT Arrays at UC 2.5.1 Magnetron Sputtering for Substrate Preparation
1 2 3 4 6 8 10 12 16 17
27 29 31 31 31 32 32 32
vii
viii
Contents
2.6 Spinning Carbon Nanotubes into Thread 2.6.1 Mechanics of Array Spinning 2.6.2 Direct Spinning of Thread from Long CNT Arrays 2.6.3 Catalyst and Substrates for Growing of Spinable CNT Arrays 2.6.4 Spinning Thread from DWCNT Arrays 2.6.5 Pulling Ribbon from CNT Arrays 2.6.6 Post-Treatment of the CNT Thread 2.7 Mechanical and Electrical Characterization of CNT Thread 2.7.1 Tensile Testing of CNT Thread 2.7.2 Electrical Properties of CNT Thread 2.7.3 Temperature Dependence of the CNT Thread Resistance 2.7.4 Electrical Properties of CNT Ribbon 2.8 Nano-Handling of CNTs Using a Nanomanipulator Inside an ESEM 2.8.1 Instrumentation 2.8.2 Handling CNT Bundles 2.8.3 Building Nanomedical Devices Using the Nanomanipulator 2.9 Carbon Nanotube Threads in Wireless, Biomedical Sensor Applications 2.9.1 Wireless Communication and the Modern World 2.9.2 Development of CNT Thread-Based Antenna at UC 2.9.3 Future Medical Application of the CNT Thread Antenna 2.10 Applications of CNT Materials in Nanomedicine 2.10.1 Carbon Nanotube Array Immunosensor 2.10.2 Carbon Nanotube Actuators 2.10.3 Carbon Nanotube Materials as Scaffolds for Supporting 2.10.3 Directional Neurite Growth 2.11 Summary and Conclusions Problems Acknowledgments References CHAPTER 3 Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems for Targeted Cancer Therapy 3.1 Introduction to Targeted Cancer Therapy 3.1.1 Cancer Statistics 3.1.2 Present-Day Cancer Treatment and Associated Problems 3.1.3 A Brief Insight into Targeting Strategies 3.2 Carbon Nanotubes: A Versatile Material 3.2.1 Definition and Synthesis of Carbon Nanotubes 3.2.2 Characterization of Carbon Nanotubes 3.2.3 Purification of Carbon Nanotubes 3.2.4 Functionalization of Carbon Nanotubes for 3.2.4 Biomedical Applications 3.3 Carbon Nanotubes as Nanovectors for Multimodal Drug Delivery 3.3.1 Carbon Nanotube Drug Delivery Systems Based on 3.3.1 Surface Functionalization
35 36 36 38 38 40 42 43 43 43 44 46 46 46 47 47 49 49 49 50 51 51 52 53 54 54 55 55
61 61 61 62 63 65 65 67 69 70 71 71
Contents
ix
3.3.2 Carbon Nanotube Drug Delivery Systems Based on Filling 3.3.2 of the Inner Cavity 3.4 Challenges and Future Prospects 3.4.1 Toxicological Aspects 3.4.2 In Vivo Biodistribution of Carbon Nanotubes 3.5 Conclusion Problems Acknowledgments References
CHAPTER 4 Composite Nanoparticles for Cancer Imaging and Therapy: Engineering Surface, Composition, and Shape 4.1 Introduction 4.1.1 Nanoscience and Medicine: The Need and the Opportunity 4.1.2 Nanodevices 4.1.3 Principles of Nanodevice Design 4.2 Materials for Nanodevice Fabrication 4.2.1 Dendrimers 4.2.2 Engineering Size, Charge, and Surface Functionality of 4.2.2 PAMAM Dendrimers 4.2.3 Dendrimer Nanocomposites: Engineering Composition 4.2.4 Inorganic Nanoparticles: Engineering Shape 4.3 Application Examples of Nanodevices 4.3.1 Nanoparticles in Cancer Imaging 4.3.2 Application Examples of Dendrimer Nanodevices 4.3.3 Perspectives on Biomedical Applications of Shaped 4.3.3 Nanocrystals 4.4 Summary Problems References CHAPTER 5 Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices 5.1 5.2 5.3 5.4
Introduction Basics of Lithographic Fabrication The Need for Three-Dimensional Biomedical Devices Present Day Lithographically Structured Biomedical Devices 5.4.1 Drug Delivery Devices 5.4.2 Structural Devices 5.4.3 Implantable Organic /Electronic Devices 5.4.4 Microfluidic Devices for Diagnosis and Cell Growth 5.4.5 Soft and Wet 3-D Devices 5.4.6 Tissue Scaffolds—Growing Live Devices 5.4.7 Interactions with Body Components
77 80 80 84 85 85 86 86
95 95 95 97 98 98 98 99 105 109 113 113 116 117 118 118 120
127 127 130 132 134 135 135 137 137 138 139 139
x
Contents
5.5 Combination of Lithography and Self-Assembly to 5.5 Construct 3-D Devices 5.5.1 Three-Dimensional Self-Assembled Containers 5.5.2 Multilayer Thin Film Stress for 3-D Self Assembly 5.5.3 Three-Dimensional Constructs for Cell Culture 5.5.4 Microscale Tetherless Gripper (Chemically Triggered 5.5.4 Microsurgical Tools) 5.6 Conclusions 5.7 Future Directions Problems Acknowledgments References Selected Bibliography CHAPTER 6 Nanosized Magnetite for Biomedical Applications
140 140 145 145 147 148 148 148 150 150 155
157
6.1 Introduction 6.2 Crystalline Structure 6.2.1 Bulk Magnetite 6.2.2 Structural Characteristics of Nanoparticles 6.3 Nanosized Magnetism 6.3.1 Multidomain and Monodomain Particles: 6.3.1 Superparamagnetism 6.3.2 Experimental Data 6.4 Magnetic Particles and Biomedical Applications 6.4.1 Magnetite and Bioworld 6.4.2 Biomedical Applications of Magnetic Single-Domain Particles 6.5 Conclusions Problems References
163 171 175 175 177 183 185 185
CHAPTER 7 Progress in the Use of Aligned Carbon Nanotubes to Support Neuronal Attachment and Directional Neurite Growth
189
7.1 Background 7.1.1 CNS Regeneration Occurs Under Some Conditions 7.1.2 Factors That Inhibit or Stimulate Axonal Regeneration 7.1.3 The Geometry Hypothesis 7.1.4 Artificial Substrates Can Promote Axonal Regeneration 7.1.5 Carbon Nanotubes and Axonal Regeneration 7.1.6 Aligned Carbon Nanotubes as a Potential Scaffold for 7.1.6 Axonal Regeneration 7.1.7 CNT Cytotoxicity 7.2 Recent Progress 7.2.1 Preparation of CNTs 7.2.2 Neuronal Cultures
157 158 158 160 163
189 189 190 190 192 193 194 195 195 196 196
Contents
xi
7.2.3 Neuronal Attachment and Neurite Outgrowth on 7.2.3 Aligned CNTs 7.3 Future Directions and Challenges Problems Acknowledgments References
196 201 202 203 203
CHAPTER 8 RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor for Nanotechnology and Gene Delivery
211
8.1 Introduction 8.2 Components Related to the phi29 DNA Packaging Motor 8.2.1 phi29 pRNA 8.2.2 phi29 Procapsid 8.2.3 Gp16 8.2.4 DNA-gp3 8.2.5 Fiber (gp8.5), and Neck and Tail (gp9, gp11-12) Proteins 8.3 Construction of the Biomimetic phi29 DNA Packaging Motor 8.4 Structure of pRNA 8.5 Mechanism of the phi29 Motor Function 8.5.1 Symmetry Argument: Pentamer or Hexamer 8.5.2 ATP Hydrolysis Provides the Driving Force of the phi29 8.5.2 DNA Packaging Motor 8.5.3 Possible Models for phi29 Motor Function 8.5.4 Single Molecule Approaches to Elucidate Motor Mechanism 8.6 Potential Applications of the phi29 Motor in Nanotechnology and 8.6 Gene Therapy 8.6.1 A Nanomotor with the Potential to Be Incorporated 8.6.1 into Nanodevices 8.6.2 Connector Arrays for Nanotechnology 8.6.3 Polyvalent Gene Delivery System Using Phi29 pRNA 8.6.4 Engineered phi 29 Connectors as Therapeutic Tools 8.6.5 phi29 DNA Packaging Motors Act as Tools for Gene Therapy 8.6.6 The DNA-Packaging Motor as a DNA-Sequencing Apparatus 8.6.6 or Molecular Sorter 8.7 Prospectives Problems References
211 212 212 213 214 215 215 215 216 217 217 219 219 221 222 222 224 225 226 226 227 227 227 228
PART II Electronic Biomedical Devices
237
CHAPTER 9 Magnetic Nanomaterials, Nanotubes, and Nanomedicine
239
9.1 Introduction 9.1.1 Nanotechnology and Nanomedicine 9.1.2 Magnetic Nanomaterials
239 239 241
xii
Contents
9.2
9.3
9.4
9.5
9.6
9.7 9.8
9.1.3 Magnetic Nanomedicine 9.1.4 Status Physical Background for Magnetic Nanomedicine 9.2.2 Magnetic Manipulation 9.2.3 Fundamentals of Nanomagentism Magnetic Nanoparticles 9.3.1 Basics of Magnetic Nanoparticles 9.3.2 Synthesis Techniques 9.3.3 Functionalization Techniques 9.3.4 Biomedical Applications of Magnetic Nanoparticles Magnetic Nanowires 9.4.1 Typical Structures of Magnetic Nanowires 9.4.2 Synthesis of Magnetic Nanowires 9.4.3 Functionalization of Magnetic Nanowires 9.4.4 Biomecial Applications of Magnetic Nanowires Magnetic Nanotubes 9.5.1 Magnetism of Magnetic Nanotubes 9.5.2 Multifunctionality of Magnetic Nanotubes 9.5.3 Synthesis of Magnetic Nanotubes 9.5.4 Biomedical Applications Magnetic Biosensors 9.6.1 Typical Magnetic Biosensing Schemes 9.6.2 Magnetoresistance-Based Sensors 9.6.3 Hall-Effect Sensors 9.6.4 Sensors Detecting Magnetic Relaxations 9.6.5 Sensors Detecting Ferrofluid Susceptibility Magnetic Biochips Prospects Problems References
242 242 243 243 244 247 247 248 251 252 254 254 254 256 257 260 260 263 263 265 267 267 270 271 272 275 276 276 277 277
CHAPTER 10 Mobile Microscopic Sensors for In Vivo Diagnostics
285
10.1 Introduction 10.2 Robot Capabilities and Environment 10.2.1 Sensing 10.2.2 Communication 10.2.3 Motion: Passive and Active 10.2.4 Computation 10.2.5 Power 10.3 Using Microscopic Robots 10.4 Evaluating Robot Behaviors 10.4.1 Theoretical Studies 10.4.2 Modeling Multiple Physical Effects 10.4.3 Validation Experiments 10.5 Example Task: High-Resolution Diagnostics 10.5.1 Diagnostic Task Environment
285 287 287 289 290 291 292 292 294 294 296 300 300 301
Contents
xiii
10.5.2 Control 10.5.3 Detection Performance 10.5.4 Using the Diagnostic Information 10.6 Discussion Problems References
302 303 304 305 307 308
CHAPTER 11 Microcantilever Biomedical Sensors
313
The Microcantilever Platform Value of Biosensors in Cancer Diagnostics and Prognostication Cantilever Preparation Biomolecular Detection Assays 11.4.1 Detection of PSA 11.5 Implantable Sensors 11.6 Conclusion Problems References
313 315 316 316 317 318 319 321 322
CHAPTER 12 Nanoimaging and In-Body Nanostructured Devices for Diagnostics and Therapeutics
325
11.1 11.2 11.3 11.4
12.1 Introduction 12.2 Technology for In Vivo Sensing 12.2.1 Atomic Force Microscopy for Multimodal and 12.2.1 Multidimensional Imaging 12.2.2 Nanoimaging, Nanosensing and Intermolecular Interactions 12.2.3 Parallel Arrays of Sensors to Detect Complementary 12.2.3 Interactions 12.3 In-Body Nanosensors and Nanodevices 12.3.1 Edema Sensor 12.3.2 Remote Controlled, Magnetically Navigated Robot Capsule 12.3.3 Mobile Microscopic Robots 12.4 Conclusions Problems References
325 326
333 336 338 341 343 344 344 345
CHAPTER 13 Microfabricated Devices for Detecting Circulating Tumor Cells in Cancer Patient Blood Samples
347
13.1 Clinical Challenge 13.2 Technical Challenge 13.3 Techniques for Detecting CTC in Blood 13.3.1 Cell Enrichment Methods 13.3.2 Microfabricated Devices 13.4 Clinical Value of CTC Capture and Characterization 13.5 Cancer Stem Cells and Metastasis
347 348 349 349 352 356 358
326 328
xiv
Contents
13.6 Application of Nanotechnology in CTC Capture 13.7 Conclusion Problems References
358 359 359 360
PART III Tiny Machines
365
CHAPTER 14 Medical Nanorobotics: The Long-Term Goal for Nanomedicine
367
14.1 14.2 14.3 14.4 14.5 14.6 14.7 14.8
Introduction From Nanoparticles to Nanorobots Diamondoid Materials in Nanorobotics Early Steps Toward Diamondoid Molecular Manufacturing Massive Parallelism Enables Practical Molecular Manufacturing Examples of Diamondoid Medical Nanorobots An Ideal Nanorobotic Pharmaceutical Delivery Vehicle Conclusion Problems References
CHAPTER 15 Potential Strategies for Advanced Nanomedical Device Ingress and Egress, Natation, Mobility, and Navigation 15.1 Introduction 15.2 Potential Nanodevice Ingress Strategies 15.2.1 Hypodermic Injection and Dermal Burrowing 15.2.2 Aerosol Inhalation and Traversing the 15.2.2 Blood/Brain Barrier (BBB) 15.2.3 Transdermal Patch, Diffusive Gel, or Eye/Ear Drops 15.3 Molecular Motors 15.3.1 Powering Molecular Motors 15.3.2 Piezoelectric Elements 15.3.3 Molecular Propellers 15.4 Constraints on Molecular Motors 15.4.1 Brownian Motion 15.4.2 Brownian Shuttles 15.4.3 Viscous Forces 15.5 Traversing the Circulatory System 15.5.1 Whole Blood Composition and Viscosity 15.6 Traversing the Lymphatic System 15.7 Phagocyte Avoidance Strategies 15.8 Nanometric Biomimetic Analogs for Potential Nanomedical 15.8 Device Motility and Ambulatory Movement 15.8.1 Cilia and Flagella 15.8.2 Myosin and Actin 15.8.3 Kinesin and Dynein
367 368 371 373 378 379 382 387 387 388
393 393 394 394 396 397 398 399 399 399 400 401 402 402 403 403 406 406 407 408 410 410
Contents
xv
15.9 Nanodevice Aqueous Motility 15.9.1 Biomimetic Flagellar Propulsion Using Nanotubes 15.9.2 Nanoscale Earthworm Analog 15.9.3 External Magnetic Propulsion 15.9.4 Ultrasonic Peristaltic Propulsion 15.9.5 Nanofluidic Channels: Behavior and Potential for Propulsion 15.10 Ambulatory Nanomedical Devices 15.10.1 DNA Robot 15.10.2 Nanowalker 15.11 Hypothetical Concept for Clinically Localized GPS Navigation 15.11 Applied to Advanced Autonomous Nanomedical Devices 15.12 Nanodevice Egress Strategies 15.13 Conclusion Problems References
411 411 412 412 412 413 413 414 415 415 416 417 417 418
CHAPTER 16 Nanoscale Mechanics for Medicine
423
16.1 Introduction 16.2 Nanoinjection and Nanotube Biocompatibility 16.2.1 Nanoinjection 16.2.2 Nanotube Biocompatibility 16.3 Nanometer Propulsion 16.3.1 Rotational Nanomotors 16.3.2 Linear Nanomotors 16.3.3 Surface-Tension-Driven Nanomotors 16.4 Nanomechanical Radios and Sensors 16.4.1 Nanotube Radio Receiver 16.4.2 Nanotube Radio Transmitter 16.4.3 Nanomechanical Mass Sensing Problems Acknowledgments References
423 423 423 424 426 426 427 429 430 430 433 434 436 436 436
PART IV Biological Integration and Characterization
439
CHAPTER 17 Integration of Manmade Nanostructures with Biological Systems: Diagnosis of Cancer Using Semiconductor Quantum-Dot Biomolecule Complexes
441
17.1 17.2 17.2 17.3 17.3
Introduction Semiconductor Quantum Dots and Their Adaptation for Nanodiagnostics Semiconductor Quantum Dots as Applied to the Study of Cellular Properties
441 442 447
xvi
Contents
17.4 Semiconductor-Quantum-Dots—Biomolecule Complexes Used 17.4 in the Study of Carcinogenic Cells and in Cancer Diagnosis 17.5 Conclusion Problems Acknowledgments References
450 453 453 454 454
CHAPTER 18 Two-Photon Microscopy for In Vivo Analysis of Neural and Secretory Activities
459
18.1 The Need for Noninvasive Imaging 18.2 Features of Two-Photon Excited Fluorescence Microscopy 18.2.1 Deep and Benign Observations 18.2.2 Replacement for Ultraviolet Sources 18.2.3 Avoiding Self-Shielding Effects and Compensating for 18.2.3 Photobleaching 18.2.4 Precise, Simultaneous Multicolor Fluorescence Imaging 18.3 Overview of the Optical System 18.4 In Vivo Imaging of the Cerebral Neocortex 18.5 Imaging of Secretory Functions 18.6 Future Possibilities Problems Acknowledgments References
459 460 462 463 464 464 465 467 469 470 472 472 472
CHAPTER 19 Nanoscale Engineering of Electrodes, Biosensors, and Protein Surfaces
475
19.1 Introduction 19.2 Carbon Nanotube Electrode and Biosensor Development 19.2.1 Carbon Nanotube Array Synthesis 19.2.2 Carbon Nanotube Array Electrode 19.2.3 Individual Carbon Nanotube Electrode 19.2.4 Carbon Nanotube Array Biosensor 19.2.5 Initial Development of a CNT Array Electrode for 19.2.5 Prostate Cancer Cell Detection 19.3 New Polymer Synthesis for High Throughput Experimental Design 19.3.1 Biophotoresist Synthesis 19.3.2 Photolithography for Protein Patterning 19.3.3 Protein Patterning 19.4 Summary and Conclusions Problems Acknowledgments References
475 476 476 476 479 480 481 484 484 485 486 487 487 487 488
About the Editors List of Contributors
491 493
Index
497
Preface Our body is run by biological molecules interacting with a medley of biological cells. These molecules are nanoscale in size while the cells in our body are microscale in size. To prevent and treat disease, and to extend the human lifespan, it is obvious that we must be able to “work” at the nano- and microscales. Working at such small scales requires developing new tools and medical procedures that have extraordinary finesse and precision. Controlling life by manipulating biological materials at the nanoscale defines a new area of interdisciplinary research that we call nanomedicine. Nanomedicine is a rapidly growing field that is difficult to survey, especially in one book. The outline for this book went through three sets of reviews by experts. In the end, the authors have carefully assembled an integrated set of perspectives that describe where we are now and where we think we should be headed to put nanomedicine into applications as quickly as possible. The book is written by experienced and new researchers in the fields of medicine, science, and engineering. The book will help to focus research on promising areas that can have a tremendous payoff in terms of improving human health. Nanomedicine research currently appears segmented between biologists, chemists, and medical doctors on one hand, and physicists and engineers on the other. The biomedical cadre view nanomedicine as developmental or synthetic biology dealing with soft materials, while engineers and physicists generally view nanomedicine as developing hard abiotic devices and robots that can repair biological systems. This book seeks to bridge the two perspectives and apply engineering principles to design integrated biotic and abiotic devices. The book describes nanostructured materials and devices that are considered technically feasible to produce and that have high potential to produce major advances in medicine in the near future. In this book devices will be defined to mean something constructed to perform a biomedical function. Devices may be particles that image or deliver drugs or electromechanical systems that sense and respond to external stimuli. This definition matches the general use of the word devices in the nanomedicine community. A brief description of nanomedicine materials, devices, and systems is given next to set the stage for reading the book. An outline of the book follows the background. Nanomedicine Materials. Nanomedicine materials are the basic biotic and abiotic matter from which devices are built. In this book, development of responsive materials and devices is emphasized. A responsive material responds to external stimuli and converts energy from one form to another. Sensing, actuation, and energy harvesting are the usual functions of responsive or smart materials. Nanoscale smart materials including structural and electronic materials will be used for making devices and for reinforcing, supplementing, monitoring, controlling, and providing therapy to mitigate disease. Nanoscale materials must be designed to interface with biological materials. Nanoscale materials might be called biologically
xvii
xviii
Preface
responsive materials. Nanoscale materials in general include hard and soft, wet and dry, and biotic and abiotic materials that are either nanostructured, or contain nanoscale and nanophase components. These materials are formed with nanoscale structure in order to provide unique capabilities and multifunctional properties, and will be the gateway to building devices that can treat a formidable number of disease states at their own scale. Ultimately, the goal of nanomedicine is to produce new and responsive materials and devices to probe and measure nanoscale and microscale biological materials to prevent, detect, and cure disease and improve the quality of human life. The study of nanomedicine materials and devices may become a new primary focus of interdisciplinary research in the physical, engineering, and life sciences fields. Nanomedicine relies heavily on our ability to synthesize and build nano- and microscale materials and devices. Another objective of nanomedicine is life extension. This requires that disease be prevented or else detected early and treated, and that deterioration of the body be repaired. The development of preventative medicine, early diagnostics, therapy, and regenerative medicine can be greatly helped by nanoscale responsive materials and responsive biosensors that sense and respond to the measurement. Nanomedicine Devices. Nanomedicine devices are sensors, motors, mechanisms, machines, or transducers that perform work and measure or control the state of biological or mechanical systems. The design of nanomedicine devices may be the most technologically sophisticated engineering mankind has ever attempted. Devices are the end goal and what physicians want. Medical devices, in general, can be classified according to their invasiveness. Nanotechnology will allow devices to go inside a patient’s body without the harmful effects typically associated with invasive medicine. Nanoscale structural, electronic, and bionic materials will eventually be used to build electronically powered parts of the body and new devices that provide special capabilities needed to repair the body. Responsive materials and devices with nanostructures that can respond to external stimuli are especially promising. Techniques for the design of responsive materials and strategies for their application are discussed in the book. Nanoscale sensors will be used to understand cellular signaling, to monitor chemicals that affect neurodegenerative diseases, and to detect metastasis in cancer. Ion channel nanoscale membranes will be used to improve the sensitivity of biosensors. Nanoscale actuators will be used to promote growth of tissue, as biomimetic valves, and artificial muscle. Nanotube actuator materials can also be used in reverse to harvest power from blood flow in the body. Active membrane fuel cells are also possible. Carbon nanotubes will allow development of miniature antennas that can go inside the body to power tiny machines or operate nanoelectronic brain sensors. Active biosensors will overcome the problem of biofouling to allow sensors to work in the body for extended periods. Nanostructured robots will be able to perform surgery on living cells and on the beating heart. Improvements from nanotechnology will yield enhanced catheters, endoscopes, needles for electrostimulation, smart stents, gene or cell transfection systems, syringes for less traumatic sampling, local delivery of therapeutic agents, online monitoring sensors for detection of circulating molecules with low concentration, and untold other advances. By reducing the size of the components interacting
Outline of the Book
xix
with the biological samples, micro- and nanotechnologies are expected to bring real breakthroughs by introducing novel concepts for improved instrumental biocompatibility and sustainable power supply. Noninvasive medical devices like sensors for glucose monitoring, particles, magnetic, and surface electrodes could also benefit from miniaturization and integration of several functions on a chip or on a device. Data acquisition and processing from these devices will be thought of in an integrated approach including consideration of wireless nanotechnology. Nanomedicine Systems. Nanomedicine systems are considered as assemblies of sensors, actuators, nanostructured robots, smart particles for imaging and drug delivery (i.e., devices that can monitor and control processes and perform functions at the cellular level and above). Systems can have different levels of complexity and scale. For example, manmade nanosensors might be integrated with the human neural system as part of a hybrid personal diagnostic system. The human body is the most complex system known, and therefore the systems level perspective is important in the book. In particular, how nanoscale devices interface with the human, called nano-micro and meso-macro-scale interface design, is also discussed from a systems integration and characterization perspective. The book takes a comprehensive approach covering a spectrum of areas pertinent to nanomedicine and the interface of biology with non-biology. In February 2008, the inventor Ray Kurzweil said to BBC that “(by 2029) machines and humans will eventually merge through devices implanted in the body to boost intelligence and health.” He said that in regard of nanobots. How close are we to having nanobots? What are the advances in nanoengineering so far? What is coming in the future? What are the most promising projects and lines of investigation? This book provides current information to help in answering these questions.
Outline of the Book Nanomedicine Design of Particles, Sensors, Motors, Implants, Robots, and Devices is uniquely focused to serve as a primary reference book and a source of new ideas for the design and application of nanomedicine materials and devices. The innate interdisciplinary nature of nanomedicine makes the book applicable to readers from a wide array of fields. Faculty members, graduate students, postdoctoral researchers, undergraduates, industry researchers, venture capitalists, consulting engineers, and others working in the field of medicine and biology can benefit from the book. A solution manual accompanies the book for use in teaching and for reference by professionals, and to stimulate exploration down new paths of research. The supplementary materials and solutions manual is available through Artech (www. artechhouse.com). The problem set/solution manual is independent by chapter and will test understanding of the chapter materials and provide suggestions on where to attack barrier problems in the field. The book will be ideal for teaching graduate level survey courses that deal with fundamental nanomedicine devices. The book is broken down into an introduction chapter that gives a nanotechnology framework for medical innovation and then chapters that discuss different applications. Chapter 1 includes a clinical perspective and lays out the
xx
Preface
need for nanotechnology to meet the major challenges and provide what is lacking in medicine. The rest of the book is organized from a technological science point of view dealing with various devices and considerations for putting the devices into application. The complexity of nanomedicine is reduced by arranging the book into four general areas: Part I, Nanoscale Materials and Particles; Part II, Electronic Biomedical Devices; Part III, Tiny Machines; and Part IV, Biological Integration and Characterization. These sections present practical concepts, new experimental procedures, design of new bioactive devices, and chapter problems to stimulate interdisciplinary innovation and to acquaint the downstream potential user with new technologies. The book also examines integration, biocompatibility, and safety issues related to implementation of devices inside the body. Since the field of nanomedicine is expanding quickly, readers are strongly encouraged to investigate the references provided in this book and to survey the recent literature to obtain the most comprehensive view of nanomedicine devices in their area of interest. The literature survey will also allow readers to become aware of the extraordinary work that could not be included in this book due to space limitations or time limitations. The authors of this book are mechanical, electrical, biomedical, chemical, and materials engineers, chemists, physicists, and medical doctors. The idea for this book came from Mr. Wayne Yuhasz who is the Executive Acquisitions Editor at Artech House Publishers. It is his and the authors’ belief that tiny machines and devices can revolutionize the field of medicine. There is unlimited potential for neuro-electronic interfaces, cell repair machines, nanorobots, flesh welders, in vivo therapy, photonic, bionic, and nanoelectromechanical delivery devices for drugs, machines, shock wave assisted therapeutic devices, selective dissection of tissues by liquid jets, minimally intrusive drug and DNA delivery, controlling the molecular circuitries in cells, and control of the interactions of biomolecules to help doctors create cures for diseases based on how the body’s cells actually function. We need physicians, chemists, biologists, physicists, engineers, and materials scientists to work together to build nanodevices that can revolutionize medicine and translate scientific advances into new options for patients. A powerful example where nanomedicine can help is in cancer. In a tumor, we don’t see 80% of the life (cell divisions). Clinicians would like a digital sensor that can detect a single cancer cell early in a concentration of millions of normal cells in blood, or detect overexpression of growth factors that accompany cancer at picomolar concentration. It is our hope that this book helps to drive nanomedicine forward rapidly and beneficially by providing the ideas needed to engineer life itself and transcend biology so that humans will live happily to an old age.
CHAPTER 1
A Nanotechnology Framework for Medical Innovation Mark J. Schulz, Weifeng Li, Vesselin Shanov, Yeoheung Yun, Chaminda Jayasinghe, Pravahan Salunke, Ge Li, Wondong Cho, Douglas Hurd, Sergey Yarmolenko, and Svetlana Fialkova
1.1
Introduction Nanomedicine is the development of nanostructured materials and devices to improve human health and performance. Nanomedicine, or tiny medicine, has a big future because it is a way to design small devices such as sensors, particles, and microrobots that can go inside our body and do what we want them to do [1, 2; see also supplementary information]. This chapter concisely outlines a nanotechnology framework for creating these small devices and explains how these devices can fight disease and improve our health now and in the future. The discussion will cover hard nanotechnology, which deals with abiotic materials that have nanoscale structure, and soft nanotechnology, which deals with biological materials. A key aspect of nanotechnology is that it is highly interdisciplinary. The mix of disciplines is producing startling advances in fundamental and translational research and a richer understanding of biological processes. From the perspective of problems in our society, it can be argued that disease is at the top of the list because every natural death is caused by disease. This considers aging a disease that can be affected by nanotechnology. The disease that may be the most frightening, and where something radical needs to be done to stop it, is cancer. It was reported in the December 22, 2008, issue of Time Magazine that in the year 2010 cancer will overtake heart disease as the leading cause of death worldwide according to the World Health Organization. The problem is that cancer cells can survive when other cells cannot. And cancer cells adapt to resist therapy. Thus new and more sophisticated medicine is needed to alleviate cancer. Screening tests are important to detect cancer early when it is most effectively treated. Screening is done to check parameters of the body (chemical, protein, electrical) that are markers of cancer. A problem is that many parameters of the body are difficult to sense or monitor, especially in vivo. Thus an important advance for fighting cancer and other disease is to develop improved sensors and instrumentation that can measure desired chemical and biological parameters of the body. A descriptive systems modeling (DSM) approach that uses words to describe experiments is proposed here for modeling disease. Therapeutic inputs can be applied to DSM models and the outputs that describe the response of the body can be measured to determine the effect
1
2
A Nanotechnology Framework for Medical Innovation
of the therapy. New sensors and computer interpretation will reduce the need for animal and human testing. An area where record-keeping and simulation modeling are needed to make advances in nanomedicine is drug development. In 2008, Nature Materials magazine estimated that 130 nanotech-based drugs and delivery systems are currently being developed worldwide. Therefore, characterizing the efficacy of the large number of nanotechnology experiments will be an important aspect of nanomedicine research. A large database will be needed just to keep track of all the nanomedicine research results. A DSM approach that uses words to describe experiments can also be useful here. Looking ahead, we also expect that nanomedicine research will expand beyond the traditional research laboratory, and that a common standard for describing and cataloging results will be needed. It was reported in several newspapers [3] that “Hobbyists can now tinker with genes.” This provides the possibility to discover a cure for cancer “in the garage.” At-home nanomedicine is likely not far behind the grassroots gene science. Synthesis of various types of nanoparticles using chemical and electrochemical processes and functionalization of these particles can be done with standard facilities. However, characterization of the result requires more sophisticated instrumentation. A DSM database can help researchers around the world to model and predict the outcomes of using various therapeutic nanoparticles. Also, a database is needed for safety reasons to guide and regulate nanoparticle research. An overview of the DSM approach is given next, and is followed by a discussion of the various aspects of instrumentation, materials, and research that are important to make advances in medicine within the framework of nanotechnology.
1.2
Descriptive Systems Modeling Producing advances in nanotechnology involves complex interdisciplinary biological, mechanical, and electrical systems research. The complexity of nanomedicine might be handled more easily by using DSM, which uses words rather than equations to represent biological processes. A biological system to be described or modeled can be any biological process, organ, or the entire body, and different levels of detail can be used to describe the various systems. The physiological conditions or parameters of the system are called the states of the system. Inputs or stimuli are external or internal effects that cause the system to change state. The states that are measured are called the outputs. The descriptive systems representation thus describes or predicts in words rather than through equations the response and outputs of the system due to different inputs. The relationships between the inputs and outputs describe important physiological characteristics and the functioning of the system. This chapter proposes various techniques to gather information to develop DSM for nanomedicine. Nanomedicine systems are complex because there are a huge number of combinations of possible inputs and responses. Describing the behavior of nano- and biosystems will require an artificial reasoning algorithm to predict outputs due to inputs that have not been empirically tested. Then a physician can describe an input (such as a drug dose) in words, use DSM to predict the response of the system, and
1.2 Descriptive Systems Modeling
3
receive a description in words of the predicted effect of the therapy. Knowing how a system functions can in turn help us understand how to make advances in nanomedicine and allow us to test drugs on a computer descriptively before performing animal testing or prescribing medication to a patient. The descriptive systems approach will fill in the semiquantitative research and clinical area that falls in between highly analytical methods (such as biochemistry that applies to small systems and their interactions or systems biology that uses differential equations to describe the dynamic response of systems) and subjective, anecdotal, and empirical practice. A cancer biologist or physician, for instance, may not believe that systems biology equations can describe the complexity of cancer. Therefore, DSM may be an appropriate intermediary framework from which to model nanomedicine that will improve our health and lengthen our lifespans. DSM should be easy to accept by the medical community because it builds on techniques that physicians already use, which include differential diagnosis, which is mainly a systematic method to identify unknowns [4]; computer-aided differential diagnosis, which is the use of computer algorithms to assist in making a differential diagnosis; The Electronic Medical Textbook [5], which was developed as a medical knowledge base for physicians to optimize both their specialties and activities in clinical practice considering 3,500 diseases; and computer-aided medical expert systems for clinical laboratory diagnosis, which references electronic forms of human knowledge and imitates human thinking and decision-making. The use of expert systems and DSM can improve medical decisions made by physicians and patients [6]. 1.2.1
Examples of Descriptive Systems Modeling
The human body is a complex system whose parameters change on short and long time scales due to aging and other factors. Inputs affect the state or condition of the body and consist of almost any physical or chemical variable such as temperature, food, drugs, and mechanical force. The states of the body are all the variables that define the health and physiological condition of the body. It is obvious that all of these variables cannot easily be described using equations. But a description using words is possible. State variables that can be measured and described are called the outputs of the body. Any intervention by the person or a physician to affect the condition of the body based on the outputs is called feedback. There can be many inputs and outputs that are time-dependent. Certain outputs are automatically sensed by the human and purposely used in the feedback system. There are also internal feedback mechanisms in the body that operate automatically. As an example, suppose a virus is an input to a person’s body. The body will internally sense a foreign material and automatically generate a feedback reaction to generate antibodies to kill the virus. The person may also see a physician who could measure the body temperature and take a saliva sample (these are external sensing functions) and then prescribe antibiotics and medicine as external inputs. This is an example of combined internal and external sensing and feedback. Since the body is represented as a system with inputs and outputs, the inputs can be adjusted to drive the output to be a desired condition. In this way, a person can be monitored and tested to study disease. The steps to guarding the health of the body are to understand and model the behavior of the body, sense the response of
4
A Nanotechnology Framework for Medical Innovation
the body, and choose appropriate inputs to move the response of the body to a desirable condition. The biological processes for example might represent biological pathways involved in cancer spreading. As an example of DSM, suppose one output variable from the body is the number of circulating tumor cells (CTC) in blood. We want this state variable to be zero. But it is difficult to measure this variable to tell if it is zero. In fact, we would like to monitor this variable continuously to give early warning of cancer spreading called metastasis. Our hypothesis is: The body can be monitored using sensors to detect CTCs and to guide therapy. To test this hypothesis, new experimentation is needed to help develop multi-input, multioutput DSM to determine if the body can be successfully monitored to detect cancer early by measuring CTCs, and to direct therapy. Biology is a descriptive science while mathematics is an analytical discipline. A goal is to first develop DSM based on qualitative and descriptive medicine and continuously move toward being more quantitative and predictive. Instrumentation and new devices including sensors will be the enabling technology to develop accurate systems models. A proposed model for biomedical processes is illustrated in Figure 1.1. Using this model, the physician should be able to enter patient data into the reasoning algorithm and test different possible therapeutic inputs. The descriptive reasoning algorithm would provide a prediction of the outcome of the treatment for the individual patient. A database of input and output test results (e.g., measuring CTC concentration after different therapies with feedback) is needed to build enough information to make reasonable predictions. An area where DSM might do a good job is in risk-benefit analysis and developing individualized medicine. Software could be developed to stratify patients based on their individual genetic differences and other factors and determine the best particular technology for a niche patient population. As an example, there is often a dilemma in cardiovascular therapy choosing between stenting and bypass surgery. The possibility of a reaction to drug elution and the occurrence of in-stent restenosis for a specific individual patient must be weighed against the risk of mortality for bypass surgery. DSM would provide statistics to trade off the two therapies.
1.3
Instrumentation Needed to Develop DSM Instrumentation is an underlying enabling technology leading to advances in multiple fields. New instrumentation for nanomedicine research is briefly mentioned.
Inputs Feedback
Biological Process Model
Biological or Physician Based Feedback Rules Figure 1.1
Outputs
Internal and External Sensing
Proposed input-output descriptive systems model (DSM) of a biological system.
1.3 Instrumentation Needed to Develop DSM
5
Subcellular Nanotechnology
Subcellular nanotechnology is described as using tiny instrumentation to characterize cellular processes, which is in contrast to typical methods of fluorescence that are used to study chemical processes in cells. A nanomanipulator [7, 8] is an instrument that can move probes or tools with nanometer precision. Measurements that can be made using the nanomanipulators include intracellular sensing of pH, making electrochemical impedance measurements with nanoscale spatial resolution, probing, nanoinjection, patch clamping, optical manipulation using light to trap and move small biological particles, cellular nanosurgery, nanoetching, and manipulation and etching of single proteins. Biological Nanotomography
Biological nanotomography or bionanotomography is an exciting new high-throughput approach for obtaining 3-D, quantifiable information. Nanotomography [9] or computed tomography (CT) uses X-rays to create cross-sections from a 3-D object that later can be used to recreate a virtual model without destroying the original model. The term nano is used to indicate that the pixel sizes of the cross-sections are in the nanometer range. Nanotomography will be useful in studying how nanodevices operate in the body, and the outcomes of implantology, especially at the interfaces between tissue and biodegradable implants. Integrated AFM and FTIR for Materials Characterization
A new infrared spectroscopy technique for determining the molecular composition of the surfaces of materials at high spatial resolution is being developed by combining atomic force microscopy (AFM) with Fourier-transform infrared spectroscopy (FTIR). This new analytical technique involves the use of an AFM to detect the response of a material to the absorption of modulated infrared radiation from an FTIR spectrometer and is referred to as AFM/FTIR spectroscopy. When the technique of AFM/FTIR spectroscopy is completely developed, it may be used to probe the molecular structure of interphases in polymer materials and adhesive bonds. The instrument may also be used to study nanoparticles being developed for imaging and therapy [10]. Luminescent Imaging
Luminescent imaging is used to study functionalization (chemical modification of the surface) of biosensor electrodes, the structure of biological motors, ion channel activity in immunology, cancer pathways, and other applications. A single molecule imaging system has been developed that can view biological molecules such as DNA, RNA, and other molecules that are smaller than the 200-nm diffraction limit of standard optical microscopy systems [11, 12]. Also recently developed is a custom confocal microscope that allows individual pixel examination of protein interactions by detecting the spectrum and lifetime of fluorescent labels on proteins [11]. High-resolution luminescent imaging will be useful to study nonspecific binding and antibiofouling techniques for nanodevices.
6
A Nanotechnology Framework for Medical Innovation
Batch Synthesis of Nanoparticles
Mass producing nanoparticles and electrical components will be needed for creating nanomedicine devices. Synthesis of carbon nanotubes (CNT) can be scaled up using the ET6000 nanofurnace [13]. Another approach to scaling up production of nanoparticles is to use a vertical reactor and large wafers. A close coupled showerhead is used to introduce reagents into the reactor through a water-cooled showerhead surface over the entire area of deposition [14]. Base Flow CVD Synthesis of CNT
Improving synthesis of nanotubes is needed, which would be a breakthrough in nanotechnology because short CNT and defects in CNT have limited their applications. Our group [15] is investigating an alternate method that we call base flow CVD. The idea is to simplify the CVD process by reversing the way the gases are supplied to the CNT nucleation zone. This is one approach to improve conventional CVD in which a catalyst is deposited on an alumina buffer layer on a SiO2/Si wafer. During synthesis the catalyst may gradually change composition and this causes defects to occur in the CNT and eventual stopping of CNT growth. Base Flow CVD supplies the carbon atoms near to the nucleation site. Further Needs in Instrumentation for Nanomedicine
New instrumentation needed to accelerate nanotechnology and nanomedicine research includes a method of identifying and characterizing defects in CNT that is more sensitive than Raman spectroscopy and that is faster and less expensive than high-resolution transmission electron microscopy. Different types of individual defects must be identified and matched to the synthesis conditions that caused the defects. The defects must also be related to the changes in properties of the nanotube including the chemical, electrical, thermal, and mechanical and other properties. It would be useful to measure these properties during the synthesis process. Functionalization of nanotubes must also be characterized with a technique that does not use fluorescence.
1.4
Nanomaterials Made for Medicine Nanomaterials are defined as those materials whose structural elements such as clusters and crystallites have dimensions in the 100-nm range or smaller. Nanomaterials is a general term that includes; nanoscale materials where the material itself is at the nanoscale size; nanophase materials, which are hybrid materials that have a nanoscale phase or component; and nanostructured materials, where the material structure has nanoscale size or features. A characteristic of nanomaterials is their improved properties compared to their bulk states. The new properties and large surface areas per volume of material have been used to develop functional nanomaterials for cancer diagnosis and therapeutics ranging from delivering chemotherapy molecules in nano-sized capsules to functional nanomaterials that deliver thermal and radiotherapy at specific targeted sites. A review of various
1.4 Nanomaterials Made for Medicine
7
nanotechnologies in cancer therapy and diagnostics is given by Hede and Huilgol [16]. Four subareas of functional nanomaterials that will be important to creating nanomedicine are described below. Several of these ideas come from National Science Foundation initiatives [17]. Artificial Biomaterials and Systems
Artificial nanobiomaterials and enhanced systems are of interest to regenerate tissue, bone, and organs. Also of interest are implants that can grow and adapt to the body for applications like pediatric othopedics [15] Nanomaterials for Sustainability in the Body
Materials and biomanufacturing research should consider sustainability in the body in terms of wear, toxicity, and energy supply. Examples include fundamental research on synthesis, properties and mechanisms of biologically friendly nanocoatings, chemicals, and materials, materials for energy harvesting, thermoelectric conversion, new energy storage methods like biological fuel cells, biogalvanic batteries, and biodegradable metals for implants, sensors, and drug delivery vectors that can dissolve in the body without toxicity when they are no longer needed. Novel magnetic nanomaterials are described in Chapter 9.
Smart Nanomaterials for Biology and Medicine Smart nanostructured materials and biomedical devices have properties that can be changed by external stimuli. These stimuli include stress, temperature, moisture, pH, electric or magnetic fields, and biological stimuli for use in sensing and actuation. Applications of such adaptive materials and systems range from the ability to control artificial muscles to sense different physiological variables inside the body, possibly including temperature, pressure, acceleration, strain, chemicals, proteins, microbes, cells, axon potentials, and almost any physical variable. These smart or “responsive” materials might also have the potential to be self-healing and self-regenerating like the human body [18]. Nanoparticles for Cancer Therapy
Thermal, magnetic, chemical, and radioactive nanoparticles may be used to image and kill cancer cells. Several chapters in the book discuss this. However, the most near-term nanoparticle application might be gold nanoshells, which have already entered clinical trials. Jennifer West and Naomi Halas at Rice University [19] have constructed gold nanoshells that can absorb light or scatter it. The nanoshells are designed to absorb infrared light and image cancer cells or heat up. When they heat up, they will cauterize cancer cells. The nanoshells are injected into the patient’s bloodstream and they pass through the leaky vasculature and collect in the tumor. An optical fiber inserted into the tumor illuminates the nanoshells with infrared light. The light heats the nanoshells and kills the tumor. Clinical trials are now being conducted at three medical centers in Texas on patience with head and neck
8
A Nanotechnology Framework for Medical Innovation
cancers. Nanospectra Biosciences will commercialize the nanoshells. An advantage of photothermal ablation is that no drug is used. Drug-free therapy lessens the potential for toxicity. But toxicity could also come from the particle itself. Another type of thermal treatment of cancer is done using iron-oxide nanoparticles. These particles are injected directly into the tumor and are heated using alternating magnetic fields that are easily tolerated by patients. Clinical trials of the magnetic anticancer nanoparticles are being conducted by MagForce Nanotechnologies in Berlin, Germany. This technique reportedly is not toxic and is promising for use against glioblastoma brain cancer and prostate cancer. Thermal treatment may not be useful in all circumstances. An alternative is the use of nanoparticles that carry drugs. Multimodal nanoparticles are also being investigated that can deliver drugs, radiation, and thermal treatment. The particle might also be used for imaging. The multimode approach may have the advantage that if one mode fails, another might succeed.
1.5
Implantable Nanomedical Devices Implantable nanomedical devices science and technology is a pioneering integrative research area that seeks to realize Nobel laureate Richard Feynman’s vision to develop tiny machines that have exquisite finesse and can go inside the body and “do what we want them to do.” Richard Feynman believed that development of tiny machines could not be avoided because they would revolutionize health care. Recently there has been more interest in Feynman’s ideas. In February 2008, inventor Ray Kurzweil said to the BBC in regard to nanorobots, that “by 2029 machines and humans will eventually merge through devices implanted in the body to boost intelligence and health.” Also in 2008, Tad Hogg of Hewlett-Packard described plausible extensions of currently demonstrated nanoscale electronics, sensors, and motors relying on directed/automatic assembly of hard nanotechnology that can enable development of nonbiological robots that are stronger, faster, and with more operational programmability than is possible with biological organisms (see Chapter 10). Developing nanomedicine includes development of implantable nanomedical devices based on smart materials, small sensors and actuators, and nanoscale and nanostructured materials. Implantable nanomedical devices will provide revolutionary medical benefits by operating within the body. Implantable nanomedical devices can sense and continuously monitor their environment, perform simple tasks such as delivering drugs, sampling fluids, cleaning arteries, and killing cancer cells. The first approach to develop implantable medical devices was based on engineering of biological systems (e.g., bacteria executing simple programs). However, biological organisms have limited material properties and computational speed. Polymer particles and viruses are also being tested in the literature to deliver drugs or knock out genes and are sometimes referred to as nanomedicine devices. But biological/polymer materials are classified as soft nanotechnology and are limited in terms of sensing and actuation capability. Developing implantable medical devices based on nanostructured electronics, sensors, and actuators, and methods of directed and automatic assembly of these tiny components is transformative research that will not only have tremendous
1.5 Implantable Nanomedical Devices
9
impact on the early detection of diseases but also enhance our understanding of disease progression. With this knowledge, many novel and advanced smart implantable devices are feasible. This groundwork also holds promise for the development of future implants such as artificial organs. We strongly believe that just as novel drug and gene therapies have unlimited potential in the treatment and cure of human disease, so does the clinical application of implantable devices. As an example of the potential for implantable devices, it is revolutionary to be able to develop nanoparticles that provide a one-stop treatment to suppress cancer. But going farther, it would also be revolutionary to have a sensor continuously monitoring for the beginning of cancer and to provide therapy at the beginning to prevent the tumor in the first place. The initial focus of implantable nanomedical devices is on hard nanotechnology, which uses nonbiological nanoscale components to develop devices that are stronger, faster, and with more functionality than is possible with biological organisms. Tiny electromechanical components built using custom synthesized nanotubes, nanowires, and other nanoscale materials with desired properties will be self-assembled or assembled with manipulators in electron microscopes with nanoscale precision. The tiny devices will enable physicians to understand and control medical processes in the body that are presently inaccessible to investigation. A distinguishing feature of nanomedicine is that it is charting a new course by developing sophisticated devices that can explore inside the body, which has never been done before, and which should produce major advances in biology and medicine. The scope of this research spans from basic nanoscience through to solving interdisciplinary technological challenges in device design. Research that needs to be performed to create nanomedicine devices is organized within three basic science and technology areas described below. Nanomaterials and biological interfaces. Nanomedicine research depends on nanomaterials such as carbon nanotubes (CNT) and metal nanowires (NW). In most of nanomedicine research, building devices requires tailored nanomaterials. When it comes to nanodevices, the nanoparticle becomes a device and thus materials science matters. Technical areas that are important for nanomaterials development are nanotube synthesis, spinning of thread/ribbon, nanowire development, and interfaces and biocompatibility. Biosensor development. Biosensors are miniature sensors that measure biological media. Electronic biosensor platforms that can measure chemicals and proteins inside the body for extended periods should be developed. Technical areas that are important for biosensor development are sensor platform configuration, wireless communication and biofouling and particularly fibrosis at the tip of the sensor, and sensor functionalization chemistry. Bioactuator development. Bioactuators are miniature actuators that manipulate biological media. Nanotubes are used to manufacture miniature electronic components called carbotronics, which include microsolenoids and super R, L, C circuit elements. Technical areas important for bioactuator development are actuation, communication/control, and biocompatibility.
10
1.6
A Nanotechnology Framework for Medical Innovation
Nanorobots Nanorobots are conceptual microscopic devices that would work at the molecular and cellular level to perform tasks in medical and other fields. Nanorobots will be built using nanoscale materials like carbon nanotubes, metal nanowires, and perhaps diamondoid materials. Building devices using nanoscale materials and parts is called nanomanufacturing. Nanomanufacturing could revolutionize medicine, but the technical barriers and costs to develop nanorobots are large. On the other hand, as Richard Feynman said, “tiny machines” is an area that cannot be avoided. The field of nanorobotics will grow more rapidly now because of the new instrumentation and communication technologies becoming available. An overview of the state of the art of nanorobotics research is given next, and then research examples are described. How to build tiny machines is an open question. No one has been able to fabricate microscopic robots using nanoscale materials. One approach is to “engineer biology.” The alternate approach is new fabrication using nonbiological materials. Whichever approach is taken, it must be cheap to make large numbers of robots. Designs must consider the high surface area to volume ratio, high strength, and fast dynamics (electronics) of tiny devices. Inertial forces are small, viscous friction where force is proportional to velocity dominates (see Richard Feynman’s talk; “There’s Plenty of Room at the Bottom”). Brownian motion causes random movements of microscopic devices. This is explained by T. Hogg and P. Kuekes [20]. Powering robots is a challenge that might be met using biogalvanic batteries, glucose, or electromagnetic waves. Communication and control are barriers to robot design and are considered in [20]. Current robots are millimeter-size. Future nanorobots will be 100 to 1,000 times smaller. Some possible features of nanorobots are briefly discussed next (see also Chapter 14). Drug delivery directly to the site of injury or infection may be the first application of nanorobots because the robot can be very simple, actually just a particle that recognizes the target and is triggered to deliver drugs or do something else. An actuator single degree of freedom robot might be developed next. A robot that uses small tools to remove blockages and plaque and break clots into small pieces would be a revolutionary tool for physicians. Precautions are needed to prevent large clot fragments breaking free as they may cause blockage in other parts of the circulatory system. Electromagnetic radiation emission or ultrasonic wave generators in nanorobots may be used to destroy cancerous cells. A ruptured cancer cell might release proteins that could cause the cancer to spread. Microwaves might break the chemical bonds in the cancerous cell, killing it without rupturing the cell wall. Focused ultrasound might be used to produce high pressure and temperature at the cell killing it. The resonant frequencies of nanoscale circuits are high because the components are small. The high frequencies can be tuned to be in the ultrasound or microwave range as needed. Powering the nanorobot to produce radiation or waves might be done through inductive coupling and a transformer. A nanorobot might also heat the cancer cell to kill it. Heating might be done using microwaves, ultrasonic waves, or a laser. Tiny lasers could burn away harmful material like arterial plaque, cancerous cells, or blood clots. The nanorobot may need to be almost in contact with the cell or material to transfer enough energy
1.6 Nanorobots
11
because of the dissipating environment in the body. Electrical therapy using two nanotube electrodes protruding from the nanorobot could kill cancer cells by generating an electric current and heating the cell until it dies. On the other hand, electrical therapy has been shown to promote healing in wounds. Nanorobots may be beneficial to use in regenerative medicine, implantology, repair of injuries, and stimulation of bone growth. Nanorobot biosensors could use electrodes to measure pH, ion concentrations, or use receptors such as aptamers or antibodies to detect protein markers of disease. Telescoping may keep the electrodes clean from biofouling and allow the nanorobot to have a smaller profile for movement. The nanorobot could also sense electrical signals in nerves and in the brain. These signals could help to diagnose neurodegenerative diseases such as epilepsy. Overall, there are probably almost as many applications for nanorobots as there are diseases. At this point, getting some initial devices working in the field is important to stimulate further research and to overcome the skepticism and fear about having nanoscale robots operating in the body. Simulations will be very important to provide confidence in putting nanorobots into application. A graphics simulation approach has been developed for nano-assembly automation for medicine; see information on the Center for Automation in Nanobiotech [21]. Nanorobot control design for molecular manipulation and the use of evolutionary agents to enable the robustness of the design are described. Simulation of bloodborne nanosensors is described in Chapter 10. Multiphysics modeling of robots in the body environment is suggested. The robots may also be built using partially biodegradable materials including polymers and metals. Biodegradable Mg for example is being modeled in the body environment using COMSOL [22]. We believe that developing biodegradable robots is very important because they will help overcome some of the fears about the potential toxicity and self-replication of the robots. Robots that gradually dissolve and disappear from the body provide confidence in their safety. The subject of biodegradable metals is very new and important and recently the NSF is paying attention to it, see information on the ERC Center for RMB [23]. Surgery using nanorobots is a more long-term goal and could especially help in cancer therapy. Nanorobots may be able to remove malignant tissue more completely and reduce the chance of cancer reoccurring. Nanorobots for laparoscopic surgery are simulated using software described by Cavalcanti et al. Laparoscopic surgery is a minimally invasive surgery in which a fiber-optic light and video camera are inserted through a small incision in the body and are used to guide surgical instruments that perform the surgery. The da Vinci surgical system is one such surgical system in which four robot arms are remotely operated by the physician. Surgical teleoperation with nanorobots in theory would allow highly precise surgery and is proposed using a cell phone RF wireless or inductive wireless communication. A diamondoid surface of the robot is proposed to minimize fibrinogen. Three major challenges to be met in developing tiny machines and eventually nanorobots are (1) engineering them to be effective, (2) making them safe so that the nanorobot doesn’t hurt surrounding healthy tissue or cause an immune response or toxicity in the body, and (3) mass-producing the devices. Dedicated research teams around the world are working on creating practical medical nanorobots to treat everything from heart disease to cancer. Robots likely will work in swarms to diag-
12
A Nanotechnology Framework for Medical Innovation
nose and treat disease. Simple robots will come first with one or two functions and will make a large impact in many areas. Later, semiautonomous nanorobots could stay in the patient’s body for a limited time and be cleared or degrade and disappear. New robots could be used as needed. Temporary and permanent implantable sensors and devices will also be used in the body. In the future, engineering and biology will come together to make our bodies become resistant to disease, increase our strength, and improve our intelligence. 1.6.1
Nanorobots for Revolutionizing Medicine
Treating heart disease, cancer, and sensing using nanorobots is proposed here. Nanorobots to Maintain the Circulatory System
The most common disease of the coronary arteries is arteriosclerosis, commonly known as hardening of the arteries. Arteriosclerosis is the thickening and stiffening of the artery walls from too much pressure and usually refers to several diseases in which the arterial wall thickens and loses its elasticity. Arteriosclerosis can occur in the arteries in the brain, kidneys, heart, abdominal aorta, or legs. Atherosclerosis is the precise term for formation of plaques on the arterial walls. Atherosclerosis can lead to arteriosclerosis. Plaque is a combination of cholesterol and other fats, calcium, and other elements carried in the blood. Plaque builds up in the small blood vessels that feed the heart. This plaque buildup can eventually narrow the arteries so that blood flow to the heart is inadequate and symptoms of insufficient blood flow called angina develop. Angina is loosely used to describe chest pain caused by lack of oxygen to the heart due to poor blood supply. In addition to angina or chest pain, atherosclerosis and arteriosclerosis can produce fatigue, shortness of breath, and an abnormal heart beat or arrhythmia. Plaque also can tear the artery walls and form blood clots that can lead to a heart attack. Often, there are no symptoms of arteriosclerosis until a heart attack occurs. A famous and beloved newsman in the United States, Tim Russert, died from a heart attack in 2008 due to cholesterol plaque rupturing in an artery, causing sudden coronary thrombosis. To remove this type of plaque from arteries, the following procedures are performed: angioplasty, coronary artery bypass graft surgery, coronary stent, and rotational atherectomy. Proposed in this chapter is a nanorobot that will operate in the circulatory system as shown in Figure 1.2(a). The robot must work in moving blood crowded with cells and various chemicals and with small vessels. The robot can perform work on the circulatory system or exchange chemicals with tissue. The robot must be comparable to the size of the cells. Initial work on developing this primitive nanorobot is being performed at the University of Cincinnati. The nanorobot has one moving part that is a linear actuator called a nanosolenoid. This actuator might be used as a chisel or ultrasound generator to remove plaque or treat stenosis in arteries. The concept nanorobot is shown in Figure 1.3. The nanorobot would be 3 µm in diameter and 5 µm long, which is smaller than the red cells in our blood that are about 7 µm wide. The bloodborne medical nanorobot will travel through the bloodstream and pass through the arteries and possibly capillaries in the human body [24]. The
1.6 Nanorobots
13
(a)
(b)
Fascicle Axon capillary Nanoparticle receptor
Figure 1.2 (Color plate 1) The human circulatory and nervous systems. (a) Nanodevices might be used to clean and maintain the circulatory system. (b) Other nanodevices might act as artificial dendrite receptors to sense the condition of the body and use the nervous system for signal transmission. (this figure contains figures modified from: http://www.unithertechnology conference.com/downloads.html, presentation “Coordinating microscopic robots for nanomedicine” by Tad Hogg)
Polymer encapsulation Ni Nanobar Magnetic flux lines CNT electrical coil
Figure 1.3 Nanorobot concept, cross-section view. The electromagnetic solenoid nanorobot has one moving part, which is a rectangular nickel nanobar that is the core of a solenoid. Actuation is based on a nanosolenoid consisting of carbon nanotubes wound around the nickel nanobar. One coil at a time is energized to move the core back and forth generating acoustic waves in the blood (only the right coil is energized in the figure). The acoustic waves have high pressure and can abrade material and clean surfaces. The length of the nanorobot is five microns—small enough to pass through arteries and capillaries in the body. The robot could be oriented and positioned by a static magnetic field. Black circles with holes in the figure represent the CNT electrical coil.
nanorobots would be actuated by an alternating magnetic field from outside or inside the body. Analysis and design of the inductive coupling from a coil on the skin to the nanorobot is being performed and will define the performance of the device. The nanorobots would be actuated in select regions of the body where plaque occurs. A steady magnetic field possibly could be applied to hold the nanorobots longer in the area of interest. Many robots would be distributed in the blood and they would be actuated together in the region of interest and collectively they would wear away plaque and deposits on the inside of the blood vessels. The nanorobots could be injected into the body and they could be expelled from the body in different ways. Only a concept nanorobot design is available at this stage, but long carbon nanotubes and nickel nanowires (NW) needed to build the nanorobot are available.
14
A Nanotechnology Framework for Medical Innovation
Procedures to assemble a prototype robot using a nanomanipulator in an environmental scanning electron microscope are being developed. Mass production of many robots that would be needed to have a significant effect in the body might be done using self-assembly during synthesis of CNT. A bundle of long carbon nanotubes must be coiled around a Ni NW to form a solenoid that has a huge inductance for its small size. The first step of drawing CNT from an array is shown in Figure 1.4(a). The Ni NWs as produced have a strong attraction to a magnet. The CNT in the bundle are double wall with an outer diameter of about 10 nm and can carry a large current and be wound into coils with a large number of turns (see Figure 1.4(b)), which means the efficiency of nanodevices may be much greater than conventional large electronics. Because of the large surface area, CNT behave as supercapacitors in an electrolyte. Because CNT can be coiled in a large number of turns, they form a superinductor with inductance: L = μ 0 N 2 A / l where the constant μ is the permeability of free space, N is the number of turns, A is the cross-sectional area of the coil, and l is the length of the coil. Due to the N2 effect, a nanotube coil can have a huge inductance. Moreover, we are also developing straight nanotube thread that has a huge inductance. Thus, CNT and thread can have tailored R, L, and C properties depending on the electrolyte around the CNT. Heating may limit the current applied so that cells are not damaged, or the nanorobot could be used to kill cells by heating. The two coils would be switched on and off alternately. The frequency of oscillation can be designed based on the resistance, inductance, and capacitance properties of the CNT coil. The nanorobot solenoid may operate using a rectifier and capacitor to apply a DC field that is switched or directly coupled to the exciter coil at the resonant frequency of the circuit. Many of these nanorobots would be operating simultaneously chipping away at the plaque on the artery walls and then the robots would eventually be removed from the body. This solenoid can operate at high frequency and an ultrasound effect is expected, which means that high pressure and temperatures could be generated. Care will be needed not to break off large pieces of plaque or puncture the artery walls. The nanorobot itself may respond to the magnetic field excitation, but at a different frequency from the solenoid. A magnetic resonance imaging (MRI) system, for example, generates high magnetic fields that can magnetically guide the nanorobot in the three dimen-
(a)
(b)
Figure 1.4 Nanorobot being developed at the University of Cincinnati: (a) nanomanipulator arm pulling CNT from an array; and (b) closeup view of the nanotubes being pulled from the array and joining to form long fibers. The next step is to wind up the fiber around the Ni NW, flip, and wind up the second coil. See Chapter 2 for more details on producing CNT and joining them.
1.6 Nanorobots
15
sions and hence control its movement and orientation in the body. Imaging may also be done simultaneously with control of the nanorobot. But MRI is expensive (see the Center for Imaging Research [25]), and the precision of the control is not known at this time. Researchers [25] suggested that in-body micro-MRI can be developed. Nanorobots for Cancer Therapy
The nanorobot solenoid actuator may also kill cancer cells and do surgery. The nanosolenoid nanorobot is a possible new approach for cancer therapy that does not use nanoparticles, radiation, drugs, or heat. A disadvantage of nanoparticles that are used for cancer therapy is that the filtering system of the body eliminates many (up to 99%) of the particles before they can be used to deliver a drug or apply hyperthermia. It is particularly difficult to get the particles inside large tumors. And particles that contain drugs may be toxic to other parts of the body. Since there are larger vessels at tumors, the nanorobot might be able to penetrate into the tumor. The nanorobot could then apply focused ultrasound to break the cancer cells apart. This nanosolenoid nanorobot may be one of the first examples of Richard Feynman’s vision coming true of tiny machines that can go inside the body to repair things. One of the limitations of this research and the reason the research is taking a long time to produce devices is the cost. It is very expensive to obtain and maintain the microscopy, nanomanipulator, and electronics instrumentation needed to build nanodevices. The societal impacts of developing implantable devices lies in providing tiny medical devices based on biosensors and bioactuators to go inside the body to detect and cure disease, repair the body, slow aging, and extend the human lifespan. Most medical problems have a need for sensing or actuation. Physicians need and desire to get their hands on actual physical systems that they can begin experimenting with. A goal of nanomedicine research should be to develop device platforms that in five years could be delivered in sufficiently large quantities to enable therapeutic and diagnostic applications. The potential impact of tiny machines that operate inside our body is as wide open as our imagination. The Holy Grail is to develop implantable medical devices and robots based on nanostructured electronics, sensors, and actuators, and to develop methods of directed and automatic assembly of these tiny components. Neural Integrated Nanorobot Biosensor
Nanorobots might be used to act as artificial dendrite receptors to sense the condition of the body using the nervous system for signal transmission. Figure 1.2(b) shows the human nervous system. In this concept, nanorobot biosensors would go inside the body and attach to dendrites. The biosensor would transduce a physiological measurement into axon potentials that would be passed through the nervous system to the brain similar to a sensing of touch, temperature, or pressure. This biomimetic approach may solve the signal transduction problem for in-body communication. Fouling of the nanorobot biosensor will have to be prevented in order to make certain measurements. The telescoping solenoid nanorobot could provide a
16
A Nanotechnology Framework for Medical Innovation
retractable electrode that is self-cleaning. The sensors added to the nervous system would potentially measure temperature, pressure, acceleration, electrical impulses, chemicals, and proteins. These variables could be used to monitor our health. 1.6.2
Nanorobot Factory
The medical utility of nanodevices has been discussed in applications ranging from sensing to repair of arteries. Nanomanufacturing, also called nanoassembly or nanoconstruction, is discussed now because low-cost, mass production of nanodevices will be required. Miniaturization and biocompatibility are related chief technical goals in all the application thrusts and their specific goals. This section is to develop the tools/materials to be used to realize the applications. Developing a nanorobot factory provides an opportunity for synergy between different applications of nanorobots. The first objective of the nanofactory is to find a set of common materials and engineering standards that all applications can use, and then build a central fabrication/assembly facility that can be shared or duplicated. A nanorobotics factory is proposed to work inside an ESEM to make nanorobots and superelectronics parts. The nanomanufacturing factory will use tiny machines to mass produce electromechanical parts and then a variety of implantable nanomaterial scaffolds, sensors, and integrated active microelectromechanical-systems (MEMS), nanoelectromechannical systems (NEMS), and devices. Most of the nano- and microsensor technology needed to develop nanorobots is still under development and the initial goal is to get it working reliably in a lab environment and in an in vitro environment (with serum contact) before moving to the additional complexities of implantation in the body. There is a potential high relative sensitivity of NEMS for detecting cells/viruses/proteins/small molecules. A semipermeable membrane will keep out macrophages, but anything that lets protein through seems likely to lead to rapid fouling of the sensor electrodes by opsonized (opsonin is an antibody in blood serum that causes bacteria or other foreign cells to become more susceptible to the action of phagocytes) serum proteins. NEMS in vivo applications for small molecule detection (and to keep out proteins) are daunting from a signal to noise point of view. An integration plan is to engineer and build a MEMS assembler that works in low vacuum for assembly of devices. Figure 1.5 shows a nanorobot factory concept with two stages of manufacturing; in-vacuum
(a)
(b)
(+)
(-)
Figure 1.5 Nanorobot factory concept with two stages of manufacturing: (a) in-vacuum processing inside an environmental scanning electron microscope showing a CNT ribbon being drawn from the CNT array, electrically fused, and wound onto a roll; and (b) in-air processing under an optical microscope where the ribbon is unwound, coated, dried, and rewound onto a Teflon coated Ni NW to form the main body of the robot actuator.
1.7 Biodegradable Metals for Temporary Implantable Nanomedical Devices
17
processing inside an environmental scanning electron microscope; and in-air processing under an optical microscope. The nanorobots produced will be about 5 μm in size. Later, self-assembly approaches using chemical vapor deposition (CVD) and physical vapor deposition (PVD) will be explored. 1.6.3
Biological Nanorobots
Electromechanical devices and robots may be easier to design than biological devices and robots. However, a good source of motility strategies, navigation routes, and immune system avoidance strategies may be parasites that reliably navigate from organ to organ across their lifecycle (preferably incurring minimal tissue damage in the process). Most of these are not free swimming in the fast-moving circulation. Worm motility could be much more versatile and robust than that of a tiny submarine. While electromechanical devices will have ample specific power (relative to a biological entity) the worms may still adequately navigate with lower power. Use of biological organisms may be slow now, but modifying their metabolism rate may speed them up. The main attraction of bacterial organisms is that they can be precisely targeted. Bacteria may also be superior in terms of biocompatibility as compared to abiotic robots. In Chapter 8, Guo describes a bacteriophage phi29 DNA-packaging nanomotor for gene delivery. Tad Hogg proposed the idea of a hybrid robot. A combined bacteria and synthetic nanoparticle robot is described in [26]. The bacteria can swim and hence pull a particle load toward specific locations; using bacteria that orient toward magnetic fields allows external steering of the bacteria. This hybrid approach would be useful to reduce biofouling and energy requirements. Chapters 10 and 14 are oriented more toward nonbiological robots due to their potential for faster CPU and stronger materials—though they are more difficult to build. The bacteria-nanoparticle hybrid guided from outside is a step toward more autonomous robots.
1.7 Biodegradable Metals for Temporary Implantable Nanomedical Devices Certain nanomedical devices do not need to remain permanently in the body. Therefore, this section takes a different approach than the last section that considers long-term implantable nanomedical devices or removal of devices by the filtering systems in the body. Here we discuss biodegradable devices (sensors and actuators) that corrode and eventually disappear from the body. Iron (Fe) and magnesium (Mg) are two metals that are biodegradable. Iron may corrode too slowly in the body for many applications. There are also biodegradable polymers. But biodegradable polymers do not have the strength and stiffness of metals. Ideally, a biodegradable metal should degrade in the biological system at a desired rate and not be toxic or adversely affect the already existing metal-organic molecules and enzymes in the physiological environment. Also, the metal should not harmfully interact with the external environment and should not be antigenic (cause an adverse immune response). Mg alloys are good candidates for biomedical implants and devices. However, Fe is being considered for implants in certain applications (e.g.,
18
A Nanotechnology Framework for Medical Innovation
biocorrodible iron stents). Of late, Mg and its alloys are the most studied metal for use as biodegradable implants because of their potential nontoxicity, similar mechanical properties to bone, biocompatibility, and biodegradability [23]. The by-products of corrosion (chemicals, gas, and possibly particles) are released into the body as the implant corrodes. Assessment of the effect of these corrosion products on cells is needed. In addition, time dependant electrochemical characterization of Mg in different solutions coupled with in vivo testing is needed to understand the applicability of Mg as an implant material.
1.8
Integration of Nanodevices in the Body This section deals with the effectiveness, biocompatibility, and toxicity of nanodevices and particles in the body. This topic was briefly discussed in previous sections. Biofouling is the buildup of proteins and cells on implantable devices. Biofouling, clotting, and inflammation are a serious problem and can limit the lifetime of biosensors and devices that operate in the body. A nanoporous alumina membrane material and diamond-like films that reduce protein absorption and are compatible with cells are reported in [26]. Nanopatterned surfaces can reduce biofouling. The morphology of a surface with posts can be designed to allow or prevent cell adhesion. Various coatings are also being developed to prevent biofouling. Also, a telescoping electrode that is self-cleaning may be developed based on the solenoid robot described herein. Biofouling becomes less of a problem when nanoparticles/nanorobots are used temporarily in the body. Nanoparticles can be cleared from the body by the liver, spleen, and lymphatic system, or they can be eaten and remain in macrophages. Certain types of nanoparticles appear to have low toxicity (see Chapter 3). Others can enter cells and have a potential for toxicity or immune response. Toxicity depends on the size, shape, material, chemical functionalization, and concentration of the nanoparticles. This means that studying toxicity of nanoparticles is a significant task. Conflicting results are reported in the literature concerning the toxicity of different nanoparticles, and some experiments use unrealistic doses. The effect of combinations of nanoparticles and the long-term effects of nanoparticles also need to be investigated. Nanomedicine research in some cases has been delayed due to potential unknown toxicities of nanoparticles or nanomaterials. Regeneration of nerves using carbon nanotube thread and ribbon scaffolds may help the paralyzed walk again. Nanoparticles may be a breakthrough in fighting cancer. Some risk is justified to allow the lame to walk and to stop cancer. Different types of cells have different reactions to nanomaterials and the effect in one animal may not be the same in another animal or in humans. Thus, studying nanoparticle toxicity will require large effort and time. But research in critical areas where otherwise hope is small should not be stopped until toxicity is fully investigated. Determining the toxicity of CNT has been one of the most pressing questions. Nanoparticles certainly present possible risks, both medically and environmentally, due to unknown interactions and the effects nanotubes have on cells and living organisms. The reactive or catalytic prop-
1.9 Safety and Ethical Implications of Nanomedicine
19
erties that nanotubes may exhibit are also a concern. In general, any foreign body will elicit reactions from the immune system wherein biofouling, clotting, infection, and thrombosis are possible. Nanostructured membranes and coatings being developed provide hope for lessening these reactions and making implantable devices biocompatible [26]. Biofouling of devices is a problem that is being investigated by many research teams. A goal is to increase the circulation half-life of devices so they can do their job and then get out. There are 60,000 miles of blood vessels in the body. During circulation, particles are deactivated and filtered out of the body by proteins absorbed onto the particle, macrophages, phagocytes, and organs. Other approaches to increase the device half-life are mentioned next. Most proteins and cells in the body are negatively charged. Positively charged particles are removed from the body quickly. Thus, designing a particle or device that has a negative potential may be one way to increase the half-life of the device. Coating particles with poly-ethylene-glycol (PEG) or heparinizing the particle (to treat with heparin to prevent coagulation) may increase the circulation half-life of the particle. Nanopatterning the surface of the particle may also help, but the proteins that will attack the particle first are on the order of nanometers in size. A flexible nanotube forest coated onto a particle may reduce protein fouling but the possible toxicity of the particle must be considered [27]. An active biosensor that mechanically cleans the surface is an approach that would require precise design of the device. Creating nanomedicine hinges in part on solving possible toxicity problems. DSM would be useful to categorize toxicity effects and develop materials that are not toxic. There are also certain preclinical and regulatory issues that must be discussed and planned in order to get these sensors and devices quickly into the hands of physicians for use in healthcare and medicine.
1.9
Safety and Ethical Implications of Nanomedicine Safety is a concern when using artificial materials and nanodevices to repair and enhance the performance of the body. A nanodevice could pose a safety risk by altering the physical or mental performance of the body. Nanopills that are high-pressure cylinders that release oxygen are predicted to enhance athletic performance by allowing a person to run at full speed for 10 minutes without exhaustion. But the enhanced oxygen supply may overtax other parts of the body and this would be a safety concern. Also, what will the standard be for athletic competition when nanopills are available? Other types of nanopills may affect thinking and this is a safety concern. Biologically adaptive electronic (bioadaptronic) devices or smart materials can modify their behavior based on the state of the body. This is a feedback control process and must be designed to ensure stability and safety in the event of system degradation. This is also known as fault-tolerant control and it should be implemented on future in-body nanomedicine systems. Nanomedicine may be used to extend the human lifespan by curing the diseases of aging. The study of life extension is called biogerentology [28]. There will be safety concerns when trying to slow the aging process and trying to repair the body. What are the dangers of putting nanorobots in the body? Robots built using hard nanotechnology would not be able
20
A Nanotechnology Framework for Medical Innovation
to self-replicate. Biological types of nanorobots are of more concern for self-replicating or having an adverse effect on biological cells and organs. Ethical considerations also come up when repairing the aging body and trying to slow aging. What could happen if the average lifespan was increased considering there are currently 6 billion people on Earth? Looking ahead at safety and ethics is important to provide guidelines on how to produce beneficial nanomedicine. DSM could reduce the need for animal testing and shorten clinical testing bringing innovation to clinical applications quicker.1 A final note of caution is due now with reference to its source which is anonymous NSF reviewers. The toxicity hazard and pollution potential of nanomaterials (e.g. powdered nanotubes used for reinforcing polymers) were recognized long after they became commonplace in the laboratories. This has lead to many new programs of research studying their potential problems. We must not make a similar mistake in our quest for nanomedicine. Given the fact that world population is growing, and with it is a growing inequity in our access to healthcare and other resources, it is prudent to collaborate with a bioethicist to examine the unintended consequences of living much longer and healthier lives with the help of nanotechnology.
1.10
Efficiently Working Together Using Shared Resources This section suggests how people and organizations can most efficiently work together using shared intellectual and physical resources. At the highest level, nanomedicine international centers of excellence for research and education could be established to focus effort to solve the most pressing problems in medicine. International organizations cooperating on research can provide the greatest expertise and instrumentation to solve problems. Intellectual property becomes a barrier to collaborations in some cases but a clear policy put in place at the beginning can produce agreement. Cooperative research using shared resources can lead to more rapid advances in nanomedicine and more intellectual property than could be produced by individual institutions because resources are used more efficiently. It is important to make rapid progress in the area of nanomedicine to be competitive because this field of research progresses so quickly. This leads to the question of how to make the fastest progress in a research laboratory with a given amount of funding resources. There are many factors that go into answering this question. We have considered this question in a university setting where students are the researchers. We assume that students are all equally paid and efficient and their pay does not change over the period of the project. Also assumed is that the students can work together efficiently, the work can be done in parallel using more than one student if desired, and facilities are available and their cost does not depend on the number of students. These assumptions do not always apply of course, but they are a useful starting point. With these assumptions, two outcomes are apparent. First, the cost of performing a project is the same regardless of the number of students working on the project. This is so because if n students are working on a project at 1/n of the total effort each, the total effort is the same, given our assumptions. Second, the calendar time to complete the project decreases in inverse proportion to the number of students working on the project. The total man-hours of work is the same, but the pro-
1.11 Chapter Summary and Conclusions
21
ject is completed faster in terms of calendar time as the number of students increases. If a project takes one year of time for one student to complete, adding more students reduces the project duration. Thus two students can complete the project 50% faster than one student. Three students can complete the project 17% faster than two students and 67% faster than one student. Four students can complete the project 8% faster than three students, and so on. There is a decreasing advantage as more students are added. When organization and management of the project are considered, too many students would make the project less efficient. But it is significant that two or three students working on a project with individual but related tasks might finish the project in half the time or less at the same cost. In general, benefits of working in teams of two or more students are that the research may progress faster and be more competitive to secure funding especially from industry and small business innovative research (SBIR) type projects, results are faster to commercialization and applications, students learn more working in teams and it prepares them for working in industry, it is fun to work with others, and more joint publications and invention disclosures can result from working in teams. On the down side, working in teams means more sharing of ideas and helping others, which some people do not prefer.
1.11
Chapter Summary and Conclusions This chapter has provided a framework for nanomedicine research to enable development of devices that can be temporarily used or implanted in the body. The science to be investigated is described in the context of various important medical applications. A descriptive systems modeling approach was proposed to be used with new sensors and devices described in the book to treat diseases such as cancer in a systematic way. New instrumentation that can help in nanoscale research was discussed along with biocompatibility, toxicity, safety, and ethics of nanomedicine. International collaboration was emphasized to advance medicine. While there are significant barriers to be overcome, the vision of nanomedicine is transformative and will have significant impact on the early detection of diseases and also enhance our understanding of disease progression. Discoveries by working at the nanoscale can lead to new instrumentation and technologies that benefit human health. Knowledge gained will cut across many disciplines including physical, chemical, biological, and engineering sciences. Specific subgoals, such as implantable devices that, in a label-free manner, reversibly sense proteins in the body, and developing an efficient way to communicate these biomolecular signals to an external source outside the body could be transformational in their own right. The major contribution of this chapter is to stimulate development of new approaches for sensing, imaging, and therapy to fight cancer and other diseases. The ideas are extremely cutting edge and straddle many frontier areas of interdisciplinary research. Many of the anticipated breakthroughs of nanomedicine are mind-boggling, and give a feeling of science fiction come alive. If successful, they will provide revolutionary benefits in health care and can boost American competitiveness in the global economy across innumerable established and newly emergent areas in nanotechnology and nanomedicine.
22
A Nanotechnology Framework for Medical Innovation
Problems 1.1 Nanomedicine literature survey. Perform a literature search in your area of interest related to nanomedicine research. Compile a review paper in the area and submit it to a nanotechnology journal for possible publication. Review papers are useful because they describe the state of the art in the field of research. Review papers usually attract a lot of readers and journals like to publish review papers. A review of the literature is also a good way to begin a research thesis or dissertation. 1.2 Communication with in-body sensors. Determine the characteristics of near-field (magnetic) and far-field (RF) coupling for communicating with sensors imbedded inside the body. 1.3 Biocompatibility of nanodevices. Select a few new approaches (e.g., nanopatterning, applying voltage) to prevent biofouling of nanodevices and perform in vivo experiments to evaluate the approaches. Patent an approach if successful. Publish the results to help the nanomedicine community.
Acknowledgments This interdisciplinary work was sponsored by NSF grant 0727250 with technical monitor Shaochen Chen, by North Carolina A&T SU through the Office of Naval Research with technical monitor Ignacio Perez, NSF grant CMS-0510823 with technical monitors Shih-Chi Liu & K. Jimmy Hsia, by the UC Institute for Nanoscale Science and Technology, Captain John Bulmer and Mr. Kevin Yost of the AFRL and General Nano LLC through the SBIR project FA8605-08-M-2822, and by NSF ERC Revolutionizing Metallic Biomaterials, Lynn Preston and Leon Esterowitz program managers, EEC-0812348. Many people at the University of Cincinnati participated in or supported this research. They include Larry Schartman, Douglas Kohls, Luree Blythe, Henry Westheider, Dale Weber, Jeffrey Simpkins, Ronald Flennigen, and Rhonda Christman. Also, various excellent comments from National Science Foundation anonymous reviewers of the proposal have been put into this chapter. This support made this work possible.
References [1] http://www.aananomed.org/, American Academy of Nanomedicine; http://www.amsocnanomed.org/, American Society for Nanomedicine. [2] http://www.foresight.org/, Foresight Institute. [3] Cincinnati Enquirer article by Marcus Wohlsen of the Associated Press, “Hobbyists Now Can Tinker with Genes,” December 30, 2008. [4] http://en.wikipedia.org/wiki/Differential_diagnosis. [5] Search www.pubmed.gov, 42 references for electronic medical textbook, 2/09. [6] Torasso, P., “Knowledge Based Expert Systems for Medical Diagnosis,” Statistics in Medicine, Vol.4, No. 3, pp. 317–325, published online October 12, 2006.
References
23
[7] Homnatti, M., G. Hughes, and R. Draper, “Enabling Subcellular Nanotechnology: An Applications Overview,” Zyvex Application Note 9719, www.zyvex.com. [8] Kleindiek nanomanipulators, http://www.nanotechnik.com/, http://www. nanotechnik.com/nw-em.html. [9] http://www.phoenix-xray.com/en/applications/other_applications/index.html, Phoenix X-Ray, a division of GE. [10] Boerio, F. J., and M. J. Starr, “AFM/FTIR: A New Technique for Materials Characterization,” Journal of Adhesion, Vol. 84, No. 10, October 2008, pp. 874–897(24). [11] “Imaging Technologies, Lighting Up In Vitro and In Vivo Work, Drug Discovery News, November 2008, www.DrugDiscoveryNews.com. [12] Zhang, H., D. Shu, F. Huang, and P. Guo, “Instrumentation and Metrology for Single RNA Counting in Biological Complexes or Nanoparticles by a Single-Molecule Dual-View System, RNA, Vol. 13, 2007, pp. 1793–1802. [13] First Nano, http://www.firstnano.com/. [14] http://www.aixtron.com/index.php?id=1&L=1. [15] Nanoworld Laboratory, University of Cincinnati, http://www.min.uc.edu/ nanoworldsmart. [16] Hede, S., and N. Huilgol, “’Nano’: The New Nemesis of Cancer,” Review Article, J Can Res Ther, Vol. 2, No. 4, 2006, pp. 186–95. Available from http://www.cancerjournal.net/ text.asp?2006/2/4/186/29829. [17] http://www.nsf.gov/index.jsp, NSF Materials World Network (MWN), Partnerships for International Research and Education (PIRE). [18] Ball, P., Made to Measure, New Materials for the 21st Century, New Jersey: Princeton University Press, 1999. [19] http://www.ece.rice.edu/~halas/; http://bioe.rice.edu/FacultyDetail.cfm?RiceID=495; http://www.nanospectra.com/. [20] Hogg, T., and P. Kuekes, “Mobile Microscopic Sensors for High-Resolution In Vivo Diagnostics,” Nanomedicine: Nanotechnology, Biology, and Medicine, Vol. 2, No. 4, 2006, pp. 239. [21] http://www.nanorobotdesign.com/, Center for Automation in Nanobiotech. [22] http://www.comsol.com/, multi-physics simulation software. [23] NSF ERC Revolutionizing Metallic Biomaterials, http://erc.ncat.edu/. [24] Freitas, R. A. Jr., “Nanotechnology and Radically Extended Life Span,” Life Extension Magazine, January 2009, http://www.lef.org/. [25] http://www.cir.uc.edu/, information on 4.0 Tesla Varian UnityINOVA Whole Body MRI/MRS system. [26] Narayan, R. J., et al., “Mechanical and Biological Properties of Nanoporous Carbon Membranes,” Biomedical Materials, Vol. 3, 2008, p. 1–7. [27] Idea for possible coating on iron nanoparticles used to treat malignant brain tumors by Allan David, Technical Manager for Biomedical Applications, Industrial Science & Technology Network, Inc, York, PA, personal conversation, March 12, 2009. [28] http://www.methuselahfoundation.org/index.php?pagename=mj_sens_scientific, foundation for curing the diseases of aging.
Endnote 1. The possibility of improving human performance or intelligence should also be considered relative to current practice in the field of regenerative medicine. An expert in regenerative medicine was asked about a goal of repairing the body and making the body function better than new using medical implants and devices. His reply was: “better than new” is asking for trouble—what is “better”? A stronger muscle or bone would arguably shift
24
A Nanotechnology Framework for Medical Innovation stresses to surrounding tissues that might not be prepared for this and thus would then fail. Generally, the laws of unforeseen complications would apply here—across the board. To state this as a goal is to appear naive to the regenerative medicine community in my opinion. Instead, it should be all about “functional outcomes”. What is or are the functions that are to be restored? Once this question is answered, one can begin to be quantitative as to what is success. Mechanical function can be defined, relevant sensing function can be defined, etc. The related questions come down to how one will achieve this function. Are the structure-property-function relationships the same as for native tissue? Since we don’t completely understand these relationships, it is difficult to aim to recapitulate them - rather we just focus on function with some qualitative sense of the underlying mechanisms. In the vast majority of cases, restoring partial tissue function is a win. The other aspect is how the restoration changes with time. We can solve many problems temporarily, and to our best evaluation all looks pretty good—but failure occurs at a much accelerated rate. Why? Because we don’t understand enough about the true healthy state and our ability to mimic it. One also has to be careful about the term “new”. For congenital deformities and diseases, new doesn’t make sense. For trauma and acquired diseases, people understand what is meant by new, but it is still an ill-defined term. Addressing tissue insufficiency is the objective—enhancements are interesting to think about, but for those working in the area, we know that we are so far from just getting things to be functional or partially functional, the enhancement talk is not taken seriously.
PART I
Nanoscale Materials and Particles
CHAPTER 2
Synthesis of Carbon Nanotube Materials for Biomedical Applications Vesselin N. Shanov, David Mast, Weifeng Li, Chaminda Jayasinghe, Nilanjan Mallik, Wondong Cho, Pravahan Salunke, Ge Li, Yeoheung Yun, Sergey Yarmolenko, Supriya Chakrabarti, and Mark. J. Schulz
2.1
Introduction to Nanoscale Materials Nanotechnology deals with new materials such as zero dimensional nanoparticles and one-dimensional structures (nanotubes, nanowires, nanobelts, and nanorods) that are expected to revolutionize the fields of mechanical, electronic, and biomedical systems. Having 1- to 100-nm size and a high surface area to volume ratio, nanoscale materials can be combined with host materials to change dramatically the properties of the bulk material making them smart and applicable in many high-tech areas [1–25]. Smart materials and devices with advantages of high performance, small size, and low cost may be developed based on nanoscale particles [26–31]. In this chapter we will focus on one particular type of nanostructured material: long multiwall carbon nanotube (MWCNT) arrays and their processing to make threads and ribbons [32–34]. Carbon nanotube (CNT) arrays are parallel forests of aligned MWCNT up to centimeters long with good electrical and mechanical properties. MWCNT arrays as shown in Figure 2.1 may have the greatest mechanical properties among the different types of nanomaterials and are the focus of this chapter. For historical perspective, carbon nanotubes are long molecules. Sumio Ijima is credited with first observing carbon nanotubes in 1991. Carbon is one of the basic elements of matter and is nonmetallic with an atomic number of 6 and an atomic weight of 12. Although carbon is plentiful in the environment, producing CNT requires carefully controlled conditions and hence CNT are expensive. Nano is used as an adjective referring to the physical size range of the tubes. As a reference, the shortest wavelength of visible light is only 400 nanometers, so anything smaller is difficult to see without using an electron microscope. Thus, characterization of nanoparticles requires special instrumentation and comprises a large part of the research effort in developing nanostructured smart materials. General properties of single-wall carbon nanotubes (SWCNT) are discussed here and are considered better than the properties of MWCNT. SWCNT is the strongest and most flexible molecular material known because of the C-C covalent bonding and seamless hexagonal network architecture. SWCNT has a Youngs
27
28
Synthesis of Carbon Nanotube Materials for Biomedical Applications
(a)
(b)
(c)
Figure 2.1 ESEM images of aligned CNT arrays grown at UC by CVD on a multilayered Si substrate using a composite catalyst. (a) CNT array sitting on a Si substrate, (b) towers of CNT, (c) side view of a CNT array, growth conditions: 200 SCCM of H2, 200 SCCM of C2H4, 100 SCCM Ar flow through a bubbler with water, 750ºC temperature, 30 minutes growth time.
modulus of 1 TPa, which is above the modulus of 70 GPa for aluminum, and 700 GPa for the best carbon-fiber. The strength to weight ratio of a SWCNT nanocomposite could be four times the same ratio for graphite/epoxy [35, Chapter 15]. The maximum strain of SWCNT can be ~10%, and the thermal conductivity is ~ 3,000 W/mK in the axial direction. SWCNT have semiconducting or conducting electrical properties depending on the chirality of the nanotube. Electrically conductive nanotubes are called metallic tubes and have electrical conductivity like copper. Semiconducting tubes reveal semiconductor properties. Nanotubes have a high current carrying capacity, a high aspect ratio, and small tip radius of curvature ideal for field emission. SWCNT has piezoresistive properties, but negligible piezoelectric property. They also have electrochemical properties and a supercapacitance property in an electrolyte. They can form an electrochemical actuator and possibly artificial muscle. Properties of MWCNT are similar to SWCNT. Electrochemical actuation, a nano-bearing, structural reinforcement, a telescoping actuator, and coax cable are some of the envisioned applications of MWCNT. The electrical properties of MWCNTs and SWCNTs are complex. About 1/3 of SWCNTs are metallic and 2/3 are semiconducting. The quantum mechanical (quantized) resistance of an individual SWCNT is R=12.9 kOhms, and the electrical resistivity is about ρ = 3 *10 -7 W cm (copper resistivity is17 . *10 -6 W cm). The resistance of the ballistic nanotube is length independent up to about micron lengths. The current carrying density of SWCNT is about 100 A/cm2. MWCNTs are electrically conductive but may not have as high of current carrying density as metallic SWCNT. It has been shown that electrical conduction in CNTs is sensitive to atomic defects, especially holes in the tubes. Processing materials with nanoscale features requires specialized instrumentation. In the area of synthesis, researchers are continuously improving control over carbon nanotube length, diameter, and array density by careful design of the substrate and nanotube synthesis conditions [35, Chapter 5]. Dispersion of nanoparticles in polymers is another critical step that is continually being improved [36]. As the size of the particle decreases, the surface area to volume ratio increases, and the properties of the material improve.
2.2 Synthesis of Long Carbon Nanotube Arrays
2.2
29
Synthesis of Long Carbon Nanotube Arrays Chemical vapor deposition (CVD) is the most important commercial approach for manufacturing CNT. CVD is an irreversible deposition of a solid from a mixture of gases through a heterogeneous chemical reaction. This reaction takes place at the interface of a gas and a solid substrate. Depending on the deposition conditions, which may change during the synthesis period, the growth process is controlled by diffusion and/or surface kinetics. CVD can be easily scaled up to industrial production and is currently is the best-known technique for high yield and low impurity production of CNT at moderate temperatures. It offers better growth control because of the equilibrium nature of the chemical reactions involved. In addition, CVD has the capability to control the size, shape, and alignment of nanotubes. Details of the synthesis and characterization of CNT arrays are briefly described here. Oxidized silicon wafers are used as substrates (Figure 2.2). Electron beam (E-beam) deposition is employed to form an intermediate Al2O3 layer and a top catalytic iron film. The thickness of deposited Al2O3 ranges from 2 to 20 nm. The thickness of the alumina layer is important because it affects the catalyst size and nanotube diameter. The thickness of Fe ranges from 0.5 to 3 nm. The composition and thickness of the catalyst are critical as they affect the diameter of the nanotubes, their length and the self-assembly process that produces arrays of nanotubes. A composite catalyst developed at UC is used to produce long arrays [32–34]. CNTs are grown by thermal CVD from an H2-C2H4-H2O-Ar gas mixture at 750°C for up to 10 hours in an EasyTube furnace from First Nano Inc. The gas composition of the growth zone is monitored by a Quadrupole Mass Spectrometer, MKS, VISION 1000 P. A schematic of the furnace to grow CNT arrays is illustrated in Figure 2.3. Based on the process described, and use of a composite catalyst, centimeter-long CNT arrays were produced. Current efforts are to scale up the manufacturing of CNT arrays, especially the production of Black Cotton, which is described as a carpet of CNTs on large area substrate. Black Cotton is a trademark owned by General Nano LLC. These efforts are closing the gap between the laboratory and industrial environment related to the
(c)
Metal catalyst AI203 Thermal annealing
Si02
(b)
Catalyst nanoparticles CNT CVD growth
(a)
Figure 2.2 (Color plate 2) Substrate preparation for growing CNT. (a) Schematic of the substrate for growing CNT arrays, (b) centimeter square substrates with grown CNT arrays being removed from a platform in a reactor tube, and (c) atomic force microscopy (AFM) image of catalyst particles uniformly distributed on the substrate.
30
Synthesis of Carbon Nanotube Materials for Biomedical Applications
Figure 2.3
Schematic of the CVD reactor.
synthesis of CNT. Industrial-size facilities for producing large area CNT are already being developed [29]. Figure 2.4 displays different images of CNT arrays. They clearly reveal the morphology of the greatly aligned CNTs that grew perpendicular to the substrate. The MWCNT have from 6- to 30-nm outer diameter. The nanotubes grow in strands with each strand having perhaps a dozen individual MWCNTs that tend to entangle with each other. Among the nanomaterials available, CNTs are promising for developing unique and revolutionary smart materials and especially sensors due to their structural and electrical characteristics. CNTs have high strength as well as high thermal and electrical conductivities, and therefore can provide structural and functional capabilities simultaneously, including actuation [1] and sensing [2–4]. Several suppliers of nanoscale materials are given in [30, 31]. Our previous studies have already demonstrated that CNTs can be synthesized from a few microns long up to ~2 cm long with nanometer diameters [32–34]. For applications at the macro scale, CNT smart materials are usually based on composite materials [35–37]. Thanks to multifunctional material properties and many fabrication possibilities, CNTs are envisioned for many applications including structural reinforcement and various kinds of sensors from the nano to the macro scale in size [38, 39]. The small size of CNTs allows them to be used as a small sensor for medical and environmental monitoring. UC is growing superlong CNT arrays that can be utilized in different high-tech areas including nanomedicine [40, 41].
(a)
(b)
(c)
Figure 2.4 Images of CNT arrays. (a) Schematic of an array grown perpendicular to the substrate, (b) optical image of 1.5-mm CNT array grown using a composite catalyst, and (c) SEM image of a CNT array grown on an oxidized Si wafer with 10-nm Al2O3 and 2-nm metal catalyst on top of it.
2.3 Characterization of CNT Arrays
2.3
31
Characterization of CNT Arrays 2.3.1
Scanning Electron Microscopy and Transmission Electron Microscopy
Environmental SEM (FEI XL-30, Philips) was employed for morphological study of the CNT arrays. Figure 2.5(a) reveals that the CNTs in the array are well aligned. The long tubes grow in strands with each strand having ~20 individual MWCNT or small bundles that tend to wave or entangle each other. This entanglement possibly provides the support needed for centimeter-long growth. The internal structure of the CNT was examined by a High Resolution TEM, JEOL JEM-2010, operated at 200 KV. For TEM study the arrays were harvested from the substrate and dispersed in ethyl alcohol using ultrasonication. The samples were prepared by placing a droplet of the suspension onto Leci™ carbon coated grid and dried in air. Figure 2.5(b) and (c) show the multiwall structure of the CNT. The number of the walls as well as the diameter of the CNT can be controlled by the substrate preparation and by the size of the catalyst particle. The CVD growth conditions also play an important role in shaping the internal structure of the CNTs. 2.3.2
Raman Spectroscopy and Thermal Gravimetric Analysis
The CNT arrays were characterized by Raman spectroscopy using a Renishaw inVia Relex MicroRaman with 514-nm excitation and 25-mW power at 50X magnification. The exposure time was 10 seconds. Two samples were analyzed: one “as-grown” MWCNT and the same array after thermal annealing at 2,200°C in Ar for 1 hour. The Raman spectra are displayed in Figure 2.6(a). Both the 1,350-cm-1 (D band) and 1,586-cm-1 (G band) peaks appear in the spectrum. In addition, a strong band around 2,700-cm-1 (G´ band) is also shown. An as-grown CNT array revealed a G to D ratio=1.36 and after thermal annealing at 2200°C in Ar for 1 hour the spectrum changed showing a G to D ratio=3.6. It is obvious from this data that thermal annealing at high temperatures in an inert environment improves the crystallinity and degree of graphitization of CNT arrays.
(a)
(b)
(c)
Figure 2.5 Side view of array at different SEM magnifications. (a) Low magnification at the bottom and high magnification at the top, (b) TEM images of CNT synthesized at UC: high-resolution image of MWCNT with 24-nm outer diameter and 10-nm inner diameter, and (c) high-resolution TEM image of DWCNT about 10-nm outer diameter and 9-nm inner diameter.
32
Synthesis of Carbon Nanotube Materials for Biomedical Applications
-1
Raman shift (cm ) Temperature (°C) Figure 2.6 Effect of annealing a CNT array from one synthesis experiment. (a) Raman spectrum of CNT arrays after annealing at 2,200°C in Ar for 1 hour with G to D ratio=3.61, and (b); TGA curves of MWCNTs arrays indicating their high purity. (Thermal Treatment and Raman by John Bulmer, AFRL/ RZPG.)
This post-processing is expected to increase the electrical conductivity of CNTs and related products such as spun threads. Thermal gravimetric analysis (TGA) was performed from room temperature to 900°C in air using a TGA Instrument TA/TGA 2050. The results are displayed in Figure 2.6(b) and reveal that the CNTs have high purity due to little weight loss below 400°C in air, and the residual mass is less than 4%.
2.4
Patterned CNT Arrays By patterning the Si substrate with catalyst we were able to grow CNT arrays with different shape and size. Images of the features including letters, numbers, and posts are shown in Figure 2.7. The posts can be easily removed from the substrate and used for different applications, including for manufacturing devices or for reinforcing polymers in composites.
2.5
Production Scale Up of CNT Arrays at UC Current efforts are underway at UC to scale up the growth process based on a pending patent [34]. A newly installed CVD reactor made by CVD Equipment Corp., which is an improved version of our research grade CVD reactor ET 1000, allows simultaneous processing of several half of 4-inch substrates (Figure 2.8 and Figure 2.9(a)). In addition, successful growth in one run on both sides of the Si substrate was demonstrated in Figure 2.9(b). These results proved that successful scale up of the CNT arrays with centimeter length is possible. 2.5.1
Magnetron Sputtering for Substrate Preparation
Preparation of the substrate is the key step for growing long CNT arrays. Our substrate is a multilayered structure with a sophisticated design. Preparation of the substrate requires a clean room environment using thin film deposition techniques that
2.5 Production Scale Up of CNT Arrays at UC
33
(b)
(a)
(c)
Figure 2.7 Images of carbon nanotube features grown at UC on patterned with catalyst on a Si substrate. (a) ESEM image of the UC logo written with CNTs on a Si substrate, (b) optical image of 7-mm-long CNT posts, 1 mm in diameter and post spacing of 1 mm, and (c) ESEM image of selectively grown, 0.5-mm-long towers of MWCNT on a patterned substrate.
(c)
(a)
(b) Figure 2.8 Shifting the growth of CNT arrays at UC “from lab to fab”: (a) research grade CVD reactor ET 1000 used for developing of novel catalysts and process conditions for growing CNT arrays on 2-inch wafers; (b) patterned CNT array grown on 4-inch wafer; and (c) newly installed CVD reactor ET 3000 capable of scaling up the growth of CNT arrays on 4 inch wafers.
can be easily scaled up. The quality of the CNT arrays and their length depend strongly on the substrate preparation that also affects their cost. The e-beam technique for thin film preparation with uniform and controlled thickness is good for substrates up to 4 to 5 inches in size. Magnetron sputtering offers large area deposition (up to 12-inch wafers) of the layered structure. We are currently exploring the use of magnetron sputtering as an alternative to e-beam deposition for preparation of the alumina intermediate layer and of the metal catalyst on Si wafers [42]. The magnetron sputtering system AJA-1800F at NCA&T State University was employed to form both the intermediate Al2O3 layer and the top catalytic iron film, see Figure 2.10(a). The thickness of deposited Al2O3 range from 2–20 nm and that of Fe ranged from 0.5 to 2 nm. The magnetron sputtering system used in this study allowed processing of 4-inch wafers with different targets, thus fabricating the entire multilayered structure within one run without removing the wafer from the system. Initial results are shown in Figure 2.10(b) and (c).
34
Synthesis of Carbon Nanotube Materials for Biomedical Applications
(b)
(a)
Figure 2.9 Scale-up of CNT growth. (a) CNT array grown on 4-inch wafer, and (b) CNT arrays grown on both sides of a Si substrate (double growth).
(b)
(a)
(c)
Figure 2.10 CNT growth on 4-inch Si substrates prepared by magnetron sputtering. (a) AJA-1800F magnetron sputtering system at NCA&T State University used in this experiment, (b) 11-mm post of MWCNT arrays, and (c) optical images of a CNT “cake” grown on magnetron sputtered substrate prepared by a combinatorial approach used to optimize the thickness of the catalyst.
As seen from Figure 2.10(b), we were able to grow centimeter-long CNT arrays from substrates entirely prepared by magnetron sputtering. This approach will be further explored in order to apply it to wafers larger than 4 inches. Figure 2.10(c) reveals the combinatorial approach applied to this magnetron study where each segment of the CNT cake represents different substrate design and catalyst thickness. The obtained results confirm that magnetron sputtering can be successfully used as a tool for substrate preparation on which superlong, highly pure CNT arrays can be synthesized.
2.6 Spinning Carbon Nanotubes into Thread
2.6
35
Spinning Carbon Nanotubes into Thread A new approach for fabricating a novel carbon nanostructured material is to develop an intermediate product that can be used as a sensor and as a fiber reinforcing material in composites. The proposed material is CNT thread. The thread is made from spinning long CNT called Black Cotton. Properties of various fibers for comparison to CNT are given in [35, Chapter 15] and show that single wall carbon nanotubes (SWCNT) have higher strength, stiffness, and are lighter than any other fiber. If a thread could be made that has similar properties to the individual nanotubes, many new structural and electronic applications would open up. Spinning long nanotubes into threads for reinforcement, self-sensing, and self-repair may produce a variety of versatile new smart materials. Polymer fibers or metallic wires may also be combined with CNT thread to form a nanocomposite thread with unique properties, such as high electrical conductivity and high strength. The blended materials approach makes virtually all the capabilities of nanotechnology available for developing novel materials for advanced applications. Threads are necessary to provide strength to composites because CNT cannot be grown beyond ~2-cm length at the present time. Thus, centimeter-long nanotubes must be spun into thread to reinforce a strong bulk material. It is also possible that the CNT could be functionalized (chemically modified) to improve their adhesion or other properties before spinning thread. The goal is to provide a strong thread with novel material properties that cannot be achieved by any other material system on Earth. The spun thread will have a diameter in the micron range. Multiple threads will be woven together to form a fiber. The fibers can be used to form tows and unidirectional plies or woven into a fabric to provide two-directional properties. Smart fabric can be made by weaving the nanotube thread into cloth and should have several interesting applications such as wearable sensors embedded in clothing, reinforcing composites, and also for electromagnetic devices such as antennas for communication with sensors inside the body. Several excellent publications have already described state-of-the-art research on spinning nanoscale carbon fibers [43–53]. J. Jiang et al. describe drawing of thread from a bundle of nanotubes, 100μm high, held on an adhesive tape [44, 46, 47]. They estimated that an array area of 1 cm2 can produce about 10m of yarn that is 200 μm wide. K. R. Atkinson et al. [45] presented a method of spinning and drawing yarns from CVD synthesized CNT arrays. The yarns produced by the authors display moderate strength when compared with carbon fiber yarns but are twice as tough. Ya-Li Li, Ian A. Kintoch, and Alan H. Windle [43] reported a method for direct spinning of carbon nanotube fibers and ribbons from the CVD reaction zone of a furnace using a liquid source of carbon and floating catalyst. The best electrical conductivity measured along the axis was 8.3 ¥ 105 W -1 m -1 , which is higher than typical carbon fiber (2.85 ¥ 10 4 W -1 m -1 ). Mechanical measurements indicate that the fibers have a range of strength depending on process conditions, between 0.05N/Tex and 0.05N/Tex, which is equivalent to 0.1GPa and 1.0GPa, assuming a density of 2g/cm3; that is within the range of typical carbon fibers. Their recent studies demonstrated significant improvement of threads strength. The mechanical properties are greatly improving as the numbers of defects in the thread are reduced. The challenge is to produce yarns that are at the same time strong, creep resistant,
36
Synthesis of Carbon Nanotube Materials for Biomedical Applications
highly conducting and reversibly deformable over relatively large strains to absorb energy [44–54]. By introducing twist during spinning of multiwall carbon nanotubes from nanotube forests to make multiply torque stabilized yarns, Zhang, Atkinson and Baughman [48, 53] achieved yarn strength greater than 460 MPa. 2.6.1
Mechanics of Array Spinning
The basic steps of array spinning are described next. CNT arrays grown perpendicularly to the substrate are frequently called a forest. The method described in this section is to spin from a forest of CNT. There is no alignment problem for spinning the thread when the CNTs are grown as a forest. With the aligned CNT being very close to each other (100 nm and less) they are weakly held together by van der Waals intermolecular forces that work at very short distances. With the CNT in this configuration they can be harvested by pulling a small bundle of CNT away from an edge of the forest in the direction that keeps the centerlines parallel and maintains the close spacing. The CNT next to the first bundle moved will be pulled along also by the van der Waals forces. As the CNT bundles are pulled away from the forest they form a long line in which all of the CNT centerlines are aligned in parallel. This line has lost some of the density the material had in the forest. For this reason the line will need to be compacted before spinning so that it will not be torn apart before the CNTs are spun. Pretreatment of the bundle before or after spinning can improve the properties of the thread. A technique to compact the line of CNT bundles is to soak it in a solvent. When the CNT bundles are wetted in a solvent, surface tension and evaporation of the solvent will force the CNT to shrink together thus compacting the array for the next spinning operation. Forming the thread from a CNT forest pulling in the vertical direction is illustrated in Figure 2.11(a). Forming the thread from a CNT forest pulling in the horizontal direction is illustrated in Figure 2.11(b). In the horizontal approach, it is likely that the tops of CNTs and bottoms of CNTs attach to each other. 2.6.2
Direct Spinning of Thread from Long CNT Arrays
In the spinning process the fibers are held together by twisting them around neighboring fibers. This prevents the fibers from slipping along the length of their neighboring fiber when axial force is applied. As the fibers are twisted and pulled more fibers are added to form a long strong yarn. The twist is characterized by the direction of the rotation and the number of twist per linear distance. An experimental spinning machine was built with a configuration that reduces the strain to the bundles of CNT as they are pulled from the forest and before they are bound together by the spinning. It was important to incorporate a spool that winds up and stores the spun CNT yarn. The spool also applies the rotation to the compacted bundles of CNT that will be spun into yarn. To keep the line of CNT bundles straight as they are spun and wound, the axes of the spools were offset to one side of the spinning axes. This keeps the side of the spools circumference that the yarn winds around in alignment with the axes of the spinning. A direct current electric motor via a mechanical speed reduction rotates the spool. The motors rotational speed is controlled via a computer. The spool and motor are held in a frame that rotates. The
2.6 Spinning Carbon Nanotubes into Thread
(a)
Figure 2.11
37
(b)
Thread formation from a CNT forest: (a) vertical pulling, and (b) horizontal pulling.
frame is mounted to the end of a rotating shaft. A second direct current electric motor rotates the shaft. This motors speed is also controlled by a computer. The twist of the yarn is controlled by changing the speed of the two motors independent of each other. Figure 2.12 shows a setup for direct spinning of thread from CNT arrays. In this method CNTs are twisted and simultaneously drawn from CNT array and wound around a spool. The twisting and winding is performed with the help of two motors. The carbon nanotubes are held together by van der Waals forces in an array. Since these forces are intermolecular, they are best felt when the distance between molecules is 70 nm or less. Van der Waals forces are strong in a direction perpendicular to the CNT length but almost vanish parallel to the length. Thus it is possible to form threads by sliding nanotubes over one another from the array and simultaneously twisting. While van der Waals forces give weak bonding, twisting provides mechanical strength between individual CNTs to form long thread. It is clear that the spacing between individual nanotubes in an array needs to
Figure 2.12
Setup for direct spinning of CNTs into thread.
38
Synthesis of Carbon Nanotube Materials for Biomedical Applications
be 70 nm or less for effectively utilizing van der Waals forces to form thread. This can be achieved by improved substrate design for growing CNT arrays. 2.6.3
Catalyst and Substrates for Growing of Spinable CNT Arrays
It is observed that dense and aligned arrays are more spinable and the thread obtained from such array is stronger. Continuous thread can be drawn from dense aligned arrays of nanotubes. In order to achieve this goal, increased catalyst particle density on the substrate is required. There is consensus in the literature that double-wall carbon nanotubes (DWCNT) are very appropriate for spinning into threads. The UC standard procedure developed for CVD of CNT was modified for synthesis of well aligned and high purity DWCNT arrays. New catalyst based on an iron alloy with increased catalyst particle density was introduced. The catalyst alloy was deposited on Si/SiO2/Al2O3 substrate by e-beam deposition. After thermal annealing a uniform distribution of high-density catalyst particles was achieved, which was proved by AFM (Figure 2.13(a)). The growth was performed in the EasyTube 3000 reactor using gas system consisting of ethylene (C2H4), water vapor, hydrogen and argon with optimized concentrations, deposition temperature and flow rates. Critical for this study was to maintain low carbon partial pressure in the reaction zone. This resulted in growing of DWCNT and their internal structure studied by HRTEM is displayed in Figure 2.13(b). Two-hour growth with our catalyst produced a 1.1-mm-long array with excellent properties for spinning. 2.6.4
Spinning Thread from DWCNT Arrays
The length of the CNT arrays being tested for spinning ranges from 1 to 10 mm. As the quality of the CNT improves, longer and longer CNT will be used to spin thread. Figure 2.14(a) shows a uniform web of nanotubes being pulled and twisted into thread directly from the substrate. ESEM images of CNT threads at different magni-
(a)
(b)
Figure 2.13 Characterization study: (a) AFM image of catalyst particle distribution on the surface of the substrate (image before growth), and (b) high-resolution TEM image of DWCNT grown from the new catalyst.
2.6 Spinning Carbon Nanotubes into Thread
(a)
(b)
39
(c)
Figure 2.14 Spinning CNT into thread. (a) Uniform web formation while spinning thread directly from the substrate. (b) and (c) SEM of CNT threads at different magnifications.
fications are shown in Figure 2.14(b) and (c). The thread can be spun as fast as our spinning machine can rotate. We are improving spinning by varying the process parameters and initially using shorter arrays about 1 to 5 mm long to prevent aging of the CNT during the growth. Arrays produced using low carbon concentration were the easiest to spin into thread. The diameter of the thread can be controlled to a certain degree by the length of the CNT array and by the spinning parameters. It is observed that the spacing between nanotubes in an array can be reduced with the help of solvents like acetone or ethanol followed by drying it at room temperature. Scanning electron microscope images of untreated and acetone treated arrays reveal that the CNTs in the acetone treated array are densified and are more aligned than in the untreated array. The acetone treatment can also be done during the spinning operation as the thread is twisted and pulled from the array in a very small spray produced with the help of a syringe. The same solvent treatment can also be performed after the spinning operation. Treatment with acetone reduces the thread diameter due to surface tension between the CNTs within the thread. It is observed that the thread is well twisted and CNTs are aligned along the long direction, making it strong. The densified CNTs are seen to be aligned along the axial direction. This initial thread was formed with a twisting speed of ~60 rpm and a winding speed of ~1 rpm. The drawing speed of thread from the array is 0.66 mm/sec and there are 15 turns per centimeter length of the thread. After acetone treatment the diameter of the thread decreases by about 30%, which may improve the strength of the thread but may also make the thread brittle. The spool with coiled thread is shown in Figure 2.15(a) and a thread coiled on a quartz tube prepared for thermal post-treatment is displayed in Figure 2.15(b). The manufacturing approach for the above thread is based on spinning directly from the array attached to the Si substrate. This is a research technique used to evaluate the properties of CNT arrays and to optimize the spinning parameters. We envision an industrial approach to spinning will be the subject of subsequent research efforts. In order to increase the diameter, one can spin two or more threads to form a rope. This technique can also be used to blend CNT threads with other wire materials made of polymers or metals. ESEM images of a CNT rope at different magnifications made by twisting two threads together is shown in Figure 2.16. Defects in the rope may reduce the strength, making post-treatment critically important.
40
Synthesis of Carbon Nanotube Materials for Biomedical Applications
(a) (a)
(b) (b)
Figure 2.15 Spinning CNT thread. (a) Thread being spooled onto the mandrel after being drawn from the array, and (b) CNT thread wrapped onto a quartz tube for postprocessing.
Figure 2.16 ESEM images of two threads twisted together to form a rope. The magnification is increasing from (a) to (b) and (c).
2.6.5
Pulling Ribbon from CNT Arrays
A material with a new morphology was pulled from our arrays. The material is called carbon nanotube ribbon, which is about 100- to 300-nm thick and 5-mm wide. Winding CNT ribbon is shown in Figure 2.17. The process was easily performed and the width of the ribbon is limited only by the lateral size of the array. The ribbon tends to stick to surfaces and is somewhat difficult to handle as drawn. Functionalization of the ribbon will be investigated to control the hydrophobicity. A Teflon spool was used to overcome this problem during the drawing. Figure 2.18(a) shows winding ribbon onto a Teflon drum. The ribbon could be formed into textiles by building a nano-loom. Manual forming a mat has been already demonstrated in our lab (Figure 2.18(b)). ESEM images of the ribbon at different magnifications are shown in Figure 2.19. For convenience, the CNT ribbon is supported on a TEM grid. SEM observation of the CNT ribbon revealed the aligned structure of the drawn material. A better look at the structure and morphology of the CNT ribbon is displayed in Figure 2.20, showing TEM images at different magnifications. A special technique was developed to keep the tiny (almost two-dimensional) ribbon on the surface of the TEM grid.
2.6 Spinning Carbon Nanotubes into Thread
41
(a)
(b)
Figure 2.17
CNT array pulled into ribbon: (a) long ribbon, and (b) spool.
(b)
(a)
Figure 2.18 Ribbon manufacturing from CNT arrays. (a) Ribbon drawn in a spiral onto a 1.5-inch diameter Teflon drum, and (b) ribbon being formed as a mat by hand.
500µm
20µm Figure 2.19
50µm
10µm
1µm
900nm
SEM Images at different magnifications of CNT ribbon supported on a TEM grid.
SEM and TEM studies of the samples brought information about the geometry of the CNT ribbon. The aligned morphology of the ribbon is easily seen in Figures 2.19 and 2.20. The higher magnification TEM image (to the right bottom of Figure
42
Synthesis of Carbon Nanotube Materials for Biomedical Applications
Figure 2.20
100µm
5µm
1µm
20nm
TEM Images at different magnifications of CNT ribbon supported on a TEM grid.
G/
1.0 0.8
G
0.6
D 0.4 0.2 0.0 1000
Figure 2.21
1500
2000
2500
3000
Raman spectrum of CNT ribbon.
2.20) reveals DWCNTs sticking out of the main stream. Figure 2.21 displays the Raman spectrum of CNT ribbon showing typical D, G, and G¢ peaks. It is obvious from the ID/IG ratio that the quality of the ribbon is pretty good. 2.6.6
Post-Treatment of the CNT Thread
Several techniques to post-treat CNT yarn are being considered. Most of them consist of applying energy to the spun yarn in the form of heat or high energy electron flux. This has to be in a controlled atmosphere to prevent oxidation. The expected effect is to join the twisted CNTs together inside the thread or ribbon so they will resist slipping when a lateral force is applied to the yarn. Electrical conductivity of CNT arrays or spun threads can be improved by post-treatment and tailored ohmic
2.7 Mechanical and Electrical Characterization of CNT Thread
43
interface materials. The post-treatment includes thermal annealing in vacuum by passing current through the tread up to the point when the surface temperature of the carbon material exceeds 2,000°C. Another approach for decreasing the thread resistance counts on treatment of the CNT threads in Ar-O2 plasma in order to remove the nongraphitic carbon. Current research efforts are underway at UC to evaluate the mechanical properties of the threads with different diameters and spun from double wall or multiwall CNT arrays, and to improve their properties by post-treatment.
2.7
Mechanical and Electrical Characterization of CNT Thread There will be different applications of CNT threads taking advantage of their strength and electrical conductivity. The electrical and mechanical properties of CNT threads depend strongly on their diameter and morphology. 2.7.1
Tensile Testing of CNT Thread
For the first time we have performed electromechanical characterization of CNT thread. A sample of CNT thread was mounted on a tab and tested in a mini-Instron testing machine. Electrical leads were connected to the ends of the thread. This testing was done at North Carolina A&T State University. The load and deflection data and the simultaneous electrical impedance (EI) were measured. A Solartron EI system was used at one frequency which was 1,000 Hz. Resistivity data was graphed with stress-strain data as shown in Figure 2.22. In this case the as spun CNT thread with 8.6-μm diameter was tested at a strain rate of 0.2-mm/min using a 7.2-mm gauge length. This test shows that the electrical resistivity of the thread increases with increasing stress and strain. The thread failed at 4% strain, revealing maximum strength of 0.55 GPa. The jagged drops in the stress strain curve are due to bundles of CNT fracturing within the thread. We observed that the strength and strain to failure of the thread can be tuned for specific applications by changing the diameter of the thread and the spinning conditions. In some cases the thread failed at 11%, but the strength dropped to 150 MPa. The highest strength obtained at UC is 1 GPa. We expect to increase the strength of the CNT thread by different post-processing techniques including solvent and thermal treatment. The thread has a piezoresistive property that can be used for monitoring damage to the thread. This feature offers new applications for structural health monitoring. In this case the thread could be attached to the surface of a structure or embedded in a composite material to monitor the material for damage. The inset in Figure 2.22 is an ESEM image from the thread at the broken point after the mechanical test. Additional studies are in progress to evaluate the failure mechanism of the CNT thread exposed to tensile stress and strain. 2.7.2
Electrical Properties of CNT Thread
Four probe electrical measurements were performed on CNT threads at different positions of continuously spun thread from the same CNT forest. A 2-inch-long
44
Synthesis of Carbon Nanotube Materials for Biomedical Applications
Time (sec) 0
20
40
60
80
100
600 1.85x10-5
500 400
1.84x10-5 300 200 1.83x10-5 100 0 0.00
0.01
0.02
0.03
0.04
0.05
Strain (mm/mm) Figure 2.22 Resistivity data (red) combined with stress-strain data (black) for CNT thread with 8.6-μm diameter, tested at a strain rate of 0.2 mm/min and using 7.2-mm gauge length.
CNT thread was set on the probe station. Then the CNT thread was treated with acetone and silver paint was used to increase the contact between the leads and thread while being careful not to allow the paint to seep along the thread. The applied current and the voltage were measured using LABVIEW. Four probe measurements were performed at different positions along the CNT thread employing a Kiethley 6220 Delta Mode system. Figure 2.23 shows the I-V curve for the CNT thread. The resistance is computed using the I-V curve for different positions along the thread. Figure 2.23(a) and 2.23(b) show the I-V curve for an acetone treated thread measured at different positions. Resistivity was determined at different locations for two CNT threads spun from the same CNT forest. The resistivity was calculated using the four probe data described above. The distance between the leads for making the measurement was 2.4 mm. Calculated resistivity is in the range of 10-4 ohm cm. It is anticipated that the contact resistance between the electrode and the carbon thread can be reduced by tailoring the interface with intermediate layers of Ti/Au. Experiments with improved contact interface are in progress. Thermal annealing will be done in the future to further reduce the resistivity of the thread. 2.7.3
Temperature Dependence of the CNT Thread Resistance
The measured CNT thread is spun with 1.5 turns per mm (twisting rate of 262 rpm) and a pulling speed of 198 mm/minute. The sample was mounted on the gold leads and cooled down in a thermostat using liquid helium. The obtained results are shown in Figure 2.24. This measurement was done at Dr. David Masts laboratory in UC. The decrease of the thread resistance with temperature is similar to the known
2.7 Mechanical and Electrical Characterization of CNT Thread
45
3.0x10-5 7.0x10 -5
Acetone treated tread
Acetone treated tread
a
2.5x10-5
6.0x10 -5
y=131.5x + 1.14 y=133.98x + 3.5
Voltage (V)
5.0x10 -5
2.0x10-5
4.0x10 -5 3.0x10
-5
2.0x10
-5
1.0x10
-5
b a
1.5x10-5
1.0x10-5 5.0x10-6
0.0 0.0
-7
1.0x10
-7
2.0x10
-7
3.0x10
-7
4.0x10
-7
5.0x10
-7
6.0x10
0.0 0.0
-8
5.0x10
Figure 2.23 locations.
-7
1.0x10
-7
1.5x10
-7
2.0x10
Current (A)
Current (A)
(a, b) The I-V curves for one sample of CNT acetone treated thread, measured at two
3000
R 0 (300K) = 1081 ohm
2800
R esis tance (ohm)
2600 2400 2200 2000 1800 1600 1400 1200 0
50
100
150
200
250
Temperature (K) Figure 2.24 Electrical resistance of CNT thread at liquid He temperatures measured by 4 probe method. (Measurements performed at D. Mast’s lab at UC.)
behavior of graphite that reveals an increase of electrical conductivity with rise in temperature. The initial properties of CNT thread manufactured at UC were compiled in Table 2.1. Specific properties of Cu and CNT thread, normalized by the density of these materials, are compared. The electrical conductivity of Cu is higher than that of CNT thread. Lately, the properties of the CNT thread have been continuously improving. Strength is not optimized because the spinning method is in the early stage of development. When discussing electrical properties, AC conductivity should also be considered. Electrical conductivity is dependent on temperature. In metals, electrical conductivity decreases with increasing temperature. In CNTs and CNT thread, electrical conductivity may increase with increasing temperature and decrease at low temperature. This is the typical behavior of carbon. Resistance ver-
46
Synthesis of Carbon Nanotube Materials for Biomedical Applications
Table 2.1
Comparison of the Properties of Cu and Initial Data of CNT Thread Made at UC
Material
Density (103 kg/m3)
Resistivity (Ohm m)
Electrical Conduct. Specific (S/m) Conduct.
Strength (MPa)
Specific Strength
Elastic Modulus Specific (GPa) Modulus
Ultimate Strain %
Copper CNT thread
8.9 ~1 estimated
1.7 × 10-8 ~1 × 10-6
59 × 106 ~1 × 106
150 Y 1,000
16.9 1,000
110 ~20
3–60 ~10
6.6 × 106 ~1 × 106
12.4 ~20
sus the number of turns per unit length of threads is currently studied and will be determined at our lab. 2.7.4
Electrical Properties of CNT Ribbon
The resistance of the CNT ribbon was measured using four probe fixture as shown in Figure 2.25(a). The width of the ribbon is 4.7 mm and the distance between two electrical leads is 2.3 mm. The thickness of the ribbon was difficult to measure by TEM and for the shown ribbon in Figure 2.25(a) is about 100 nm. Conductive epoxy was used on the CNT ribbon for better contact with Au-coated four probe leads. The resistance of the ribbon is determined from the slope of the I-V curve shown in Figure 2.25(b) and is 214 ohms. The calculated resistivity is 4.4E-04 ohm cm.
2.8
Nano-Handling of CNTs Using a Nanomanipulator Inside an ESEM 2.8.1
Instrumentation
Nanomanipulators are powerful tools for property studies and handling of individual CNTs or of a small strand of them. These instruments are usually placed inside scanning electron microscopes, which allow visualization of CNT at high magnification during their handling such as translation, rotation, and testing. The Kleindiek MM3A-EM nanomanipulator is a unique robotic type device [55]. Our
CNT ribbon
(a)
(b)
Four probe leads Current (A) Figure 2.25 Four probe measurement of CNT ribbon. (a) Ribbon attached to gold coated leads using conductive epoxy, and (b) I vs. V curve of CNT ribbon.
2.8 Nano-Handling of CNTs Using a Nanomanipulator Inside an ESEM
47
nanomanipulator is installed inside the chamber of an FEI Environmental Scanning Electron Microscope (ESEM) XL-30 FEG. A picture of this nanomanipulator with 4 arms located at the UC Nanoworld Lab is shown in Figure 2.26. Each arm is highly mobile through two rotations and one translational axis that are individually controlled. Cooperative control is in development. The positioning accuracy is close to 1 nm. 2.8.2
Handling CNT Bundles
In order to make miniature circuits that can go into the human body, one has to be able to handle and build the individual components and devices that are expected to have nano and micro dimensions. The nanomicromanipulator is the right tool for such a task. Building tiny devices with the nanomanipulator requires skills to handle different nano-objects including picking up, relocating, dropping off, bending, twisting, stretching, cutting/breaking, and passing a current. Our initial experience is with CNT bundles, which we consider as important building blocks for future nanomedical devices. A single CNT is difficult to handle even under SEM. A bundle of 100 to 1,000 CNTs can be easily handled by a nanomanipulator. As shown in Figure 2.27, a small bundle of CNTs can be picked up by two probes (Figure 2.27(a)) or by a microgripper (Figure 2.27(b)). 2.8.3
Building Nanomedical Devices Using the Nanomanipulator
Our manipulator has great potential as a tool to build different nanomedical devices including miniature intracellular biosensors. Such nanodevices can help to better understand the genesis of cancer and to electronically diagnose devastating deceases at a biological cell level. We have designed a sensor with two carbon nanotube electrodes. The tiny electrodes will be inserted in the cell through the cell membrane. Such an arrangement will enable us to measure the electrical impedance of the intracellular fluid. The impedance depends on the concentration of different ions and a change in their concentration may indicate a disease at a very early stage. We have already manufactured tiny electrodes by welding small bundles of CNTs to the tungsten probes of the nanomanipulator using the ESEM and established procedures. The tips of these electrically conductive electrodes are about 50 nm in diame-
(a)
(b)
Figure 2.26 Robots for nanomanipulation: (a) Kleindiek nanomanipulator, and (b) two robotic arms inside an ESEM at UC.
48
Synthesis of Carbon Nanotube Materials for Biomedical Applications
(a)
(b)
(c)
(d)
Figure 2.27 ESEM images of the nanomanipulator in action, showing how different attachments can grasp strands of several CNTs and pull them away from the array. (a) Two probes picking up a CNT bundle. (b) Microgripper pulling a bundle from the CNT array. ESEM images of sharp CNT electrodes made from small bundles. (c) Low magnification. (d) High magnification.
ter (Figure 2.27). Such nanosharpness of the electrodes will allow easy penetration of the cellular membrane and entering a single cell without significantly disturbing the cell environment or damaging the membrane. These nanoprobes will act as electrodes and will be connected to an electrochemical impedance spectroscopy (EIS) system, which is available in our lab. The EIS system is capable of applying microvoltage power and measuring picoamp current. The EIS software will compute the impedance of the cell. Precancerous conditions related to changes in ion concentration and pH of the cell will be detected with this arrangement. There is hope that such a nanotool can be employed for early diagnostics of many diseases. For most applications individual CNTs are too small and fragile to form a sensor or related electrodes. The CNT bundles exhibit different collective properties compared to individual CNT. Our electrical measurements using the nanomanipulator showed that CNT arrays look like the perfect raw material since bundles harvested from them tolerate large current density (2.5 × 107 A/cm2), have low resistivity (10-4 ohm cm), and possess good strength. In order to find application for biosensor nanoelectrodes, the properties of the CNT bundles have to be well characterized and documented. Electrical properties of small CNT bundles and CNT threads are being investigated in our Nanoworld Lab. Nanotube threads will also be used to build sensors with desired size, strength, and electrical conductivity. The nanomanipulator and ESEM will be the tools of choice to assemble and characterize prototype sensors based on CNT materials. Different attachments to our Kleindiek nanorobot, such as a rotational arm and tweezers, will be employed to manufacture a variety of nanodevices. CNT arrays, bundles, and threads along with metal and semiconducting nanowires will be implemented to build biosensors and other nanodevices.
2.9 Carbon Nanotube Threads in Wireless, Biomedical Sensor Applications
49
2.9 Carbon Nanotube Threads in Wireless, Biomedical Sensor Applications 2.9.1
Wireless Communication and the Modern World
The way our society communicates and processes information is rapidly becoming totally wireless. For wireless sensor applications, the development of smaller and more versatile sensor elements has been driven, in part, by the use of novel, nanostructured materials, often with multiple functionalities engineered into their composition and structure. One of the most widely used functionalized nanostructured materials has been CNTs [56–58]. Carbon nanotubes can also be used as integral components for wireless communication. Wang et al. showed that CNTs can be used as antennae in receiving electromagnetic signals [59], while Jensen et al. have used a single, long carbon nanotube to construct a fully functional, fully integrated radio [60]. The development of ultraminiature wireless sensors made with individual CNT for use in vivo health monitoring, treatment, and diagnosis would have tremendous potential. There are, however, considerable technological challenges that need to be addressed in order to fabricate devices from CNT that are nanometers in size. An intermediate target is developing wireless sensors that use macroscopic-sized threads and yarns made with CNT for both the sensor element as well as for wireless communication. While these sensors would not be small enough to be used intravenously, they could be implanted or embedded in tissue. 2.9.2
Development of CNT Thread-Based Antenna at UC
We have used CNT threads, woven from multiwall carbon nanotubes (MWCNTs) as described above, for the construction of antennae for the transmission and reception for a variety of wireless communication signals as described above. The antennae were designed to operate in the frequency bands common to most current medical wireless devices [61]. We focused on wireless communication at frequencies close to the Wireless Medical Telemetry Service (WMTS) bands (608–614 MHz, 1,395–1,400 MHz, and 1,429–1,432 MHz). Additional measurements were carried out in the ISM (Industrial, Scientific, and Medical) 802.11 bands at 2.4–2.4835 MHz. This band overlaps the frequency bands used in the Bluetooth communication protocols. CNT thread antennae were tested both in transmission and in reception. A CNT thread that was about 20 μm in diameter was used to construct a simple dipole antenna. The design frequency of the antenna was 700 MHz (length of each arm was 4). Two identical antennae were also fabricated using 1-mm diameter copper wire. Initial measurements were made with separations between transmit and receive antennae of ~1.25 and ~2m. The antennae were visually aligned but no special care was taken to reduce the reception of reflected and other unwanted signals. The copper dipole antennae showed maximum transmitted signal intensity at a fundamental frequency of about 694 MHz. The CNT thread antenna was used in a number of initial measurements to determine its effectiveness in transmitting and receiving wireless communication signals as discussed above. These include: (1) the transmission and reception of a
50
Synthesis of Carbon Nanotube Materials for Biomedical Applications
continuous wave (CW) signal at fo = 694 and 1,388 MHz (2 × fo), (2) the transmission and detection of a CW signal plus sidebands at ± 100 kHz FM modulation, (3) an amplitude modulated (AM ) broadcast and/or reception of music (from an online classical music channel), (4) the transmission and reception of black and white (BW) composite video images, (5) the simultaneous broadcast or reception of two separate FM signals, and (6) the transmission of one FM signal while simultaneously receiving an FM signal at a different frequency (multiplexing). Finally, a separate CNT thread antenna (length ~2.5) was used as the antenna of a cell phone for demonstrating mobile phone communication. We have used the CNT thread antenna to broadcast and receive at multiple frequencies on the same antenna. We have also used the CNT thread antenna to simultaneously transmit and receive wireless signals on the same CNT thread antenna. Both of these measurements were carried out using FM modulation at audio frequencies. A CNT thread antenna was installed in a cell phone as shown in Figure 2.28. Several phone calls were made without degradation of signal clarity compared to calls made with the cell phones original antenna. The cell phone’s signal strength was one bar with no antenna and four bars with the CNT thread antenna installed. Finally, some additional transmission experiments were carried out using smaller single arm (half dipole) CNT antennae. These antennae could send and/or receive signals up to 8 GHz. The upper frequency range was limited by the upper range of the reference antenna used in these measurements. 2.9.3
Future Medical Application of the CNT Thread Antenna
The conducted investigation and related measurements clearly show that CNT threads can be used to construct antennae for wireless communication and that these antennae have suitable properties (size, mass, electrical conductivity, flexibility, and toughness) for potential use in wireless sensor applications. These research efforts are expected to result in the development of techniques to fabricate macroscopic wireless sensors using yarns and threads spun from carbon
Figure 2.28 Photograph of a cell phone with a CNT thread antenna on the back side. The thread is encapsulated in a polyimide film for handling. Inset shows an SEM picture of the thread used as an antenna.
2.10 Applications of CNT Materials in Nanomedicine
51
nanotubes. As far as we know, the use of CNT threads for high-frequency wireless communication is unique [62, 63]. The potential for applications of CNT thread antennas alone is quite broad, ranging from cell phones to wireless systems woven into clothes. For future medical application, the thread will be shielded with a thin film of insulating polymer and incorporated into an antenna that will be able to transmit and receive when immersed into a body fluid. Initial tests to show the feasibility of such biomedical sensors have been carried out using a simple loop antenna made from a 20-μm diameter CNT thread. The amplitude of the received signal was measured when the antenna was placed in sterile fetal bovine serum (FBS was used as phantom body fluid) and compared to that when the antenna was in air. A spiral antenna was used as the second antenna in these tests, with the separation between the antennae approximately 20 cm. These measurements were carried out using the CNT antenna to either receive or transmit; no difference was observed for the transmitted signal in either of these situations. At a transmit frequency of 1.750 MHz, the signal through the FBS was approximately 4.0 dB lower than through air, for a FBS depth of 2.67 cm. At a transmit frequency of 5.0 GHz the received signal decreased by approximately 4.7 dB for 2.8 cm of FBS. Estimated penetration depths at these frequencies are 2.9 cm at 1.750 GHz and 2.6 cm at 5.0 GHz. Since multiple refection could play a substantial role in these tests these values should be considered as rough estimates. These results do provide clear proof however that antennae made from these CNT threads can be used as implantable biosensors. Further in vitro experiments using CNT thread antenna in phosphate buffered saline (PBS) solution are in progress and will be published soon.
2.10
Applications of CNT Materials in Nanomedicine We have already mentioned how CNT arrays and CNT threads can be employed for medical applications. We envision sharp CNT bundles to be used as nanoprobes for interrogating biological cells, and CNT thread to transmit and emit wireless information for in-body imbedded biological sensor. A few more medical applications of carbon nanostructured materials are described in the following sections. 2.10.1
Carbon Nanotube Array Immunosensor
Biosensors that are small and label-free for direct detection of the analyte are expected to provide advances in the area of cancer detection by detecting cancer cells in solution, and in osteoporosis to monitor a number of bone markers. The use of biosensors also has enormous significance for measurement of neurotransmitters such as dopamine, epinephrine, norepinephrine, catecholamine, and serotonin for the examination of physiological functions and diagnosis of nervous diseases such as epilepsy, Parinsonian syndrome, and senile dementia. CNT arrays offer many advantages for manufacturing electrodes that are relatively easy to fabricate without employing any special procedures or instrumentation. Patterned silicon substrates allow repeatable carbon nanotube array towers to be synthesized to millimeter length, which is ideal for electrochemical biosensor development. A label-free immunosensor was developed based on carbon nanotube array electrodes
52
Synthesis of Carbon Nanotube Materials for Biomedical Applications
[64]. More details are presented in Chapter 19. Basically, highly aligned multiwall carbon nanotubes were grown on a Fe/Al2O3/SiO2/Si substrate by chemical vapor deposition. The substrate was patterned with 100-μm square blocks and 100-μm spacing between blocks. In one of the applications, carbon nanotube towers up to 6 mm in height grew from the blocks, and the towers were easy to peel off the silicon substrate. The harvested towers were cast in epoxy and both ends were polished. One end is used for electrical connection, and the other end for the biosensor electrode. The nanotube electrode was electrochemically activated to open the nanotube ends and to expose COOH groups on the surface. Antimouse IgG was then covalently immobilized on the nanotube array. Electrochemical impedance spectroscopy (EIS) was used to characterize the binding of mouse IgG to the antibody already immobilized on the nanotube electrode surface. A detection limit of 200 ng/mL and a dynamic range up to 100 mg/mL were obtained. The nanotube array immunosensor has good sensitivity and may find applications in the area of cancer detection, monitoring bone markers, and monitoring chemicals to understand neuron-degenerative diseases. This type of immunosensor might be particularly useful for rapid detection with moderate sensitivity as compared to optical immunoassay techniques. However, reducing the electrode size and improving the preparation of the surface of the electrode are needed to lower the detection limit. Overall, the carbon nanotube array immunosensor was mechanically strong and showed a fast response and good sensitivity based on electrochemical impedance spectroscopy analysis. The carbon nanotube array thus has potential for use in many electronic biosensing applications. 2.10.2
Carbon Nanotube Actuators
The actuation effect in CNT materials makes them attractive for muscle mimic research and artificial muscle development. Three basic approaches of actuation using nanotubes are being investigated. They are: (1) electrochemical bond expansion, (2) telescoping MWCNT, and (3) a CNT solenoid. They are described briefly. Only the first has been demonstrated. Electrochemical bond expansion actuation is described. Several CNT electrochemical based actuators are developed at UC and discussed in [65–68]. Three approaches are proposed to develop electrochemical actuators which are predicted to have about 1% strain based on the principle of electrochemical bond expansion. The first uses a solid polymer electrolyte with nanotubes or similar carbon nanoparticles dispersed into the polymer. Electrodes are put on the surfaces of the material. Applying a voltage causes the polymer to expand. This can be done using dry or liquid electrolyte. The dry actuator provides about 20% of the performance of the wet actuator. The solid polymer electrolyte actuators are slow and generate low force [66–68]. Recently, Baughman et al. demonstrated that CNT sheets can perform as mussels when applying a positive voltage of 3 to 5kV [69]. The second approach uses a nanotube array post with an electrical wire attached at one end and the post placed in an electrolyte with a counter electrode. Applying 1 volt to the nanotubes causes them to expand [65]. Since the nanotubes are long they easily buckle and the force generated would be small. The third approach is proposed but not tested. It is to use CNT thread in an electrolyte as an artificial muscle. The first
2.10 Applications of CNT Materials in Nanomedicine
53
experiment planned it to attach a spring to a long CNT thread in an electrolyte and apply voltage to the thread. A counterelectrode will be used for the return circuit. Since the CNT can only expand due to electrochemical bond expansion, the thread must be preloaded by the spring and only releases tension and cannot itself generate a tension load. If successful, this CNT thread actuator may have much higher forces than previous actuators described in the first two approaches. 2.10.3 Carbon Nanotube Materials as Scaffolds for Supporting Directional Neurite Growth
The devastating consequences of injury to the nervous system are well-known and there is currently little in the way of effective therapy. A research program at UC initiated by Dr. Keith Crutcher seeks to provide the foundation for the development of a prosthetic material based on carbon nanotubes, threads, and ribbons for promoting repair of the injured nervous system. More details can be found in K. Crutcher’s Chapter 7. The ongoing efforts at UC are focused on designing 3-D scaffolds for axonal nerve regeneration combining nanotechnology and neural regeneration by testing nerve track outgrowth on carbon nanotube materials that have been described previously in this chapter. This project takes advantage of recent developments at the University of Cincinnati in both nanotechnology and neural regeneration research to design a suitable three-dimensional scaffold to promote axonal regeneration within the injured brain, spinal cord, or peripheral nerve. In particular, findings from the Crutcher laboratory suggest that the spatial distribution of chemical cues in both peripheral nervous system (PNS) and central nervous system (CNS) white matter is important for the promotion of axonal regeneration. The recent success in fabricating and functionalizing aligned centimeter-long carbon nanotubes provided a unique opportunity to develop prosthetic material with suitable geometry and chemical composition that may promote axonal growth. The CNT are synthesized in different diameters (6–60 nm) and used individually or post-processed into intermediate forms, such as bundles, spun thread, or ribbons, and/or functionalized to provide the desired properties (diameter, length, roughness, surface chemistry) to promote neuronal adhesion and growth. CNT bundles were dissected from the arrays, placed horizontally in culture dishes, and secured with silicone grease. Embryonic chick sympathetic neurons (day 9–10) were dissociated and cultured with the CNT bundles in the presence or absence of exogenous nerve growth factor (NGF). For some cultures, the CNTs were functionalized with oxygen plasma prior to their use in culture. After 4 to 7 days in culture, the cells were labeled with a vital dye and images were collected using either an inverted or upright fluorescence microscope. Preliminary results indicate that both as-grown CNTs and CNTs that have been functionalized by plasma treatment are able to support neuronal attachment and neurite growth. Similar positive results have been obtained when employing CNT threads and ribbons. The majority of the neurites follow the longitudinal orientation of the CNT bundles and show almost no branching. These results demonstrate the feasibility of using aligned carbon nanotubes as a substrate for neurite growth in tissue culture, supporting the possibility of their use as a prosthetic substrate to support axonal regeneration in vivo [70].
54
2.11
Synthesis of Carbon Nanotube Materials for Biomedical Applications
Summary and Conclusions This chapter presents recent results in catalytic synthesis of centimeter-long CNT arrays by CVD and spinning them into threads and ribbons. The role of the catalyst design on the substrate surface and its impact on the length of the oriented CNTs was revealed. The length of the highly oriented CNT arrays depends on the substrate design and the nature of the catalyst. Growing centimeter-long nanotube arrays provides hope that continuous growth of CNTs in the meter length range is possible. CNT arrays open up new applications by mitigating the limitations of the powdered spaghetti type CNT. Additionally, characterization of the CNT materials by advanced instrumentation was discussed. Current research efforts to scale up the synthesis process and to develop manufacturing tools and methods that industry needs to mass-produce aligned CNTs were discussed. Special attention was given to manipulation and electrical measurements of CNT bundles using a nanomanipulator located inside an Environmental Scanning Electron Microscope. The chapter also illustrated exiting applications of the CNT materials related to their unique properties. Recent advances in the development of Black CottonTM, which is centimeter-long CNT arrays grown on large substrates, were discussed. High-quality nanotube arrays, threads, and ribbons are expected to become enabling materials for different high-technology applications. Nanoizing conventional materials will lead to producing a new generation of novel materials with applications from medicine to space. Biomedical applications of CNT arrays, threads, and ribbons were discussed with reference to other chapters in this book.
Problems 2.1 What is Nanotechnology? Please refer to the literature to describe the impact of nanotechnology on science and society. 2.2 What is Nanomedicine? Please search in the literature to support your answer. 2.3 How important are nanoscale materials for medicine? 2.4 What are the current and future applications of carbon nanotubes (CNT)? Describe each application briefly. 2.5 What are the major applications of CNT materials for medicine? Perform a literature search on (CNT) nanomaterials for medicine and reveal any new advances that have been made relative to the information provided in this chapter. 2.6 What is a CNT array? Describe the major advantages of using aligned CNTs over powdered (loose) CNT particles for manufacturing biosensors and devices for nanomedicine. Please study the literature and provide examples. 2.7 What is the chemical vapor deposition (CVD) technique? Why is it so important for fabrication of CNT as compared to other growth techniques? Please refer to the literature to elaborate your answer. 2.8 What are the major techniques used for characterization of CNT materials? Comment on the limitations and advantages of each technique.
Acknowledgments
55
2.9 What are the technical approaches used to prepare substrates for growing CNT arrays? Elaborate on the role of magnetron sputtering for production scale up of carbon nanotube materials. Please refer to this chapter and to the literature to form your explanation. 2.10 Can CNT arrays be spun to manufacture thread? Briefly describe the CNT spinning process for fabrication of threads and ribbons. 2.11 What are the characterization techniques mentioned in this chapter for studying the electrical and mechanical properties of CNT threads? 2.12 Can CNT thread replace copper in electrical applications, especially in power distribution? Please discuss the advantages and drawbacks of a carbon cable over a metal cable. 2.13 What is a nanomanipulator? Please provide adequate reasoning about the role of nanomanipulator for investigating CNT and building nanomedical devices. 2.14 What is the future of CNT threads in wireless communication and in nanomedicine?
Acknowledgments This interdisciplinary work was sponsored by NSF grant 0727250 with technical monitor Shaochen Chen, and by North Carolina A&T SU through the Office of Naval Research with technical monitor Ignacio Perez, NSF grant CMS-0510823 with technical monitors Shih-Chi Liu & K. Jimmy Hsia. The nanomedicine research was partially supported by the National Institute of Occupational Safety and Health and by the Health Pilot Research Project Training Program as part of the University of Cincinnati Education and Research Center Grant #T42/OH008432-03, and by the UC Institute for Nanoscale Science and Technology. Captain John Bulmer and Mr. Kevin Yost of the AFRL, and General Nano LLC provided support for electrical characterization and development of CNT thread for electrical wire through the SBIR project FA8605-08-M-2822. This support is gratefully acknowledged.
References [1] R. H. Baughman, C. Cui, A. Zakhidov, Z. Zafar Lqbal, J. N. Barisci, G. M. Spinks, G. G. Wallace, A. Mazzoldi, D. De Rossi, A. G. Rinzler, O. Jaschinski, S. Roth, and M. Kertesz, “Carbon Nanotube Actuators,” Science, Vol. 284, 1999, pp. 1340–1344. [2] S. Peng, J. OKeeffe, C. Wei, K. Cho, J. Kong, and R. Chen, “Carbon Nanotube Chemical and Mechanical Sensors,” 3rd International Workshop on Structural Health Monitoring, Stanford, CA, September 2003, pp. 15–17. [3] J. R. Wood, and H. D. Wagner, “Single-Wall Carbon Nanotubes as Molecular Pressure Sensors,” Applied Physics Letters, Vol. 76, No. 20, 2000, pp. 2883–2885. [4] J. Kong, N. R. Frankin, C. Zhou, M. G. Chapline, S. Peng, K. Cho, and H. Dai, “Nanotube Molecular Wires as Chemical Sensors,” Science, Vol. 287, 2000pp. 622–625. [5] http://altmine.mie.uc.edu/mschulz/public_html/smartlab/smartlab.html.
56
Synthesis of Carbon Nanotube Materials for Biomedical Applications [6] I. Kang, J. Y. Jung, G. R. Choi, H. Park, J. W. Lee, K. W. Yoon, Y. Yun, V. Shanov, and M. J. Schulz, “Developing Carbon Nanocomposite Smart Materials,” Solid State Phenomena, Vol. 119, 2007, pp. 207–210. [7] T. W. Tombler, C. Zhou, L. Alexseyev, J. Kong, H. Dal, L. Liu, C. S. Jayanthl, M. Tang, and S. Y. Wu, “Reversible Electromechanical Characteristics of Carbon Nanotubes Under Local-Probe Manipulation,” Nature, Vol. 405, 2000, pp. 769–772. [8] A. N. Watkins, J. L. Ingram, J. D. Jordan, R. A. Wincheski, J. M. Smits, and P. A. Williams, “Single Wall Carbon Nanotube-Based Structural Health Monitoring Sensing Materials,” Nanotech, Vol. 3, 2004. [9] I. Kang, M. J. Schulz, J. H. Kim, V. Shanov, and D. Shi, “A Carbon Nanotube Strain Sensor for Structural Health Monitoring, Smart Materials and Structures,” Vol. 15, No. 3, 2006, pp. 737–748. [10] Nader Jalili, Bhuvenesh C. Goswami, and Darren M. Dawson, “Distributed Sensors and Actuators via Electronic-Textiles,” NTC Project: M04-CL05, National Textile Center Research Briefs—Materials Competency, June 2005. [11] D. Abbott, B. Doyle, J. Dunlop, T. Farmer, M. Hedley, J. Herrmann, G. James, M. Johnson, B. Joshi, G. Poulton, D. Price, M. Prokopenko, T. Reda, D. Rees, A. Scott, P. Valencia, D. Ward, and J. Winter, “Development and Evaluation of Sensor Concepts for Ageless Aerospace Vehicles: Development of Concepts for an Intelligent Sensing System,” NASA/CR-2002-211773, 2002. [12] V. H. Ebron, Z. Yang, D. S. Seyer, M. Kozlov, J. Oh, H. Xie, J. Razal, L. J. Hall, J. P. Ferraris, A. G. MacDiarmid, and R. H. Baughman, “Fuel Powered Artificial Muscles,” Science, Vol. 311, No. 5767, 2006, pp. 1580–1583. [13] M. Zhang, F. Shaoli, A. A. Zakhidov, S. B. Lee, A. E. Aliev, C. D. Williams, K. R. Atkinson, and R. H. Baughman, “Strong, Transparent, Multifunctional, Carbon Nanotube Sheets,” Science, Vol. 309, No. 5738, 2005, pp. 1215–1219. [14] R. H. Baughman, “Materials Science, Playing Nature’s Game with Artificial Muscles,” Science, 308, No. 5718, 2005, pp. 63–65. [15] M. Zhang, K. R. Atkinson, and R. H. Baughman, “Multifunctional Carbon Nanotube Yarns by Downsizing an Ancient Technology,” Science, Vol. 306, No. 5700, 2004, pp. 1358–1361. [16] R. H. Baughman, “Materials Science, Muscles Made from Metal,” Science, Vol. 300, No. 5617, 2003, pp. 268–269. [17] H. Tsung-Chin, K. J. Loh, and J. P. Lynch, “Electrical Impedance Tomography of Carbon Nanotube Composite Materials,” Proceedings of the SPIE, Vol. 6529, 2007, p. 652926. [18] M. Meyyappan, Carbon Nanotubes: Science and Applications, Boca Raton, FL: CRC Press, 2005. [19] J. Soyoun, T. Ji, J. Xie, J. Jining, and V. K. Varadan, “A Novel Strain Sensor Using Carbon Nanotubes-Organic Semiconductor Matrix Composite on Polymeric Substrates,” Proceedings of the SPIE, Vol. 6414, 2007. [20] Y. Jin, and F. G. Yuan, “Simulation of Elastic Properties of Single-Walled Carbon Nanotubes,” Composites Science and Technology, Vol. 63, 2003, pp. 1507–1515. [21] R. Arun, J. Nader, and R. Himanshu, “Development of a Novel Strain Sensor Using Nanotube-Based Materials with Applications to Structural Vibration Control,” Proceedings of the SPIE, Vol. 5503, 2004, pp. 478–485. [22] S. Getty, SWCNT NanoCompass for Next-Generation Magnetometry, Materials Engineering Branch, NASA Goddard Space Flight Center, presented at SAMPE 2007, June 6, 2007. [23] G. Y. Lia, P. M. Wang, and X. Zhao, “Carbon; Mechanical Behavior and Microstructure of Cement Composites Incorporating Surface-Treated Multi-Walled Carbon Nanotubes,” Vol. 43 No. 6, May 2005, pp. 1239–1245. [24] J. Makar, J. Margeson, and J. Luh, “Carbon Nanotube/Cement Composites: Early Results and Potential Applications,” NRCC-47643, 3rd International Conference on Construction
References
[25]
[26]
[27]
[28] [29] [30] [31] [32]
[33]
[34]
[35] [36]
[37]
[38] [39]
[40] [41] [42]
57 Materials: Performance, Innovations and Structural Implications, Vancouver, B.C., Aug. 22–24, 2005. X. Xiao, J. W. Elam, S. Trasobares, O. Auciello, and J. A. Carlisle, “Synthesis of a Self-Assembled Hybrid of Ultrananocrystalline Diamond and Carbon Nanotubes,” Adv. Mater, Vol. 17, No. 12, June 2005, pp. 1451–1565. M. Schulz, “A Structural Neural System for Data Mining and Anomaly Detection,” Data Mining in Aeronautics, Sciences, and Exploration Systems Conference (DMASES), Computer History Museum, Mountain View, CA, June 26–27, 2007, http://ase.arc.nasa.gov/projects/dmases/2007/. M. Schulz, G. Kirikera, Y. Yun, V. Shanov, S. Mullapudi, G. Maheshwari, and R. Allemang, “A Structural Neural System with Multi-State Sensors for Integrated Systems Health Management,” Proceedings of the 2nd World Congress on Engineering Asset Management (EAM) and The 4th International Conference on Condition Monitoring, Harrogate, U.K., June 11–14, 2007, pp. 1739–1750. S. Ghosh, A. K. Sood, and N. Kumar, “Carbon Nanotube Flow Sensors,” Science, Vol. 299, 2003, pp. 1042–1044. The Easy Tube System, First Nano, Inc., 1860 Smithtown Ave, Ronkonkoma, NY 11779, Carbon Nanotube Synthesis, http://www.firstnano.com (2007). Applied Sciences Inc., and Pyrograf Products Inc., Cedarville, Ohio, www.apsci.com. Carbon Nanotechnologies Inc., 16200 Park Row, Houston, TX 77084-5159, http://cnanotech.com. Y. Yun,, V. Shanov, Y. Tu,, S. Subramaniam, and M. J. Schulz, “Growth Mechanism of Long Aligned Multiwall Carbon Nanotube Arrays by Water-Assisted Chemical Vapor Deposition,” J. Phys. Chem. B, Vol. 110, No. 47, 2006, pp. 23920–23925. V. Shanov, Y.-H. Yun, and M. J. Schulz, “Synthesis and Characterization of Carbon Nanotube Materials,” Journal of the University of Chemical Technology and Metallurgy, Vol. 41, No. 4, 2006, pp. 377–433. V. Shanov, A. Gorton, Y. Yun, M. Schulz, Invention Disclosure: UC 107-044 (patent pending), Catalyst and Method for Manufacturing Carbon Nanostructured Materials, October 17, 2006. M. J. Schulz, A. Kelkar, and M. Sundaresan, Nanoengineering of Structural, Functional and Smart Materials, CRC Press, 2006, pp. 125 and 347. R. Andrews, and M. Weisenberger, “Carbon Nanotube Polymer Composites: A Review of Recent Developments,” University of Kentucky Center for Applied Energy Research, 2004, www.isr.us/Spaceelevatorconference/. I. Kang, Y. Yun, J. H. Kim, J. W. Lee, R. Gollapudi, S. Subramaniam, S. Narsimhadevara, D. Hurd, G. R. Kirikera, V. Shanov, M. J. Schulz, D. Shi, F. J. Boerio, and S. Mall, “Introduction to Carbon Nanotube and Nanofiber Smart Materials,” Composites Part B: Engineering, Vol. 37, No. 6, 2006,pp. 382–394. J. Koo, Polymer Nanocomposites: Processing, Characterization, and Applications, New York: McGraw-Hill, 2006. V. N. Shanov, G. Choi, G. Maheshwari, G. Seth, S. Chopra, G. Li, Y. Yun, J. Abot, and M. J. Schulz, “Structural Nanoskin Based on Carbon Nanosphere Chains,” SPIE Smart Structures Conference, San Diego CA, March 2007. Press Release by NSF on Long Carbon Nanotube Growth Achieved at UC. http://www.nsf.gov/news/news_summ.jsp?cntn_id=108992&org=NSF&from=news. Press Release on Long CNT Growth, University of Cincinnati Researchers Grow Their Longest Carbon Nanotube Ever, http://www.uc.edu/news/NR.asp?id=4811. P. Salunke, S. Yarmolenko, S. Neralla, J. Sankar, K. Fischbach, G. Li, Y.-H. Yun, M. Schulz, and V. Shanov, “Substrate Preparation by Magnetron Sputtering and CVD Growth of Carbon Nanotube Arrays,” Proceedings of the 14th International Mechanical Engineering Congress IMECE 2007, IMECE2007-43190, Seattle, WA, Nov. 11–15, 2007.
58
Synthesis of Carbon Nanotube Materials for Biomedical Applications [43] Y. L. Li, I. A. Kintoch, and A. H. Windle, “Direct Spinning of Carbon Nanotube Fibers from Chemical Vapor Deposition Synthesis,” Science, Vol. 304, 2004, pp. 276–278. [44] K. Jiang, Q. Li, and S. Fan, “Spinning Continuous Carbon Nanotube Yarns,” Nature, Vol. 419, 2002, p. 801. [45] K. R. Atkinson, S. C. Hawkins, C. Huynh, C. Skourtis, J. Dai, M. Zhang, S. Fang, A. A. Zakhidov, S. B. Lee, A. E. Aliev, C. D. Williams, and R. H. Baughman, “Multifunctional Carbon Nanotube Yarns and Transparent Sheets: Fabrication, Properties and Applications,” Physics B, Vol. 394, 2007, pp. 339–343. [46] K. Jiang, S. Fan, and Q. Li, Method for Fabricating Carbon Nanotube Yarn, U.S. Patent No. 7,045,108 B2, 2006. [47] K. Jiang, S. Fan, and Q. Li, Method for Manufacturing a Light Filament from Carbon Nanotubes, U.S. Patent No. 6,957,993 B2, 2005. [48] M. Zhang, K. R. Atkinson, and R. H. Baughman, “Multifunctional Carbon Nanotube Yarns by Downsizing an Ancient Technology,” Science, Vol. 306, 2004, pp. 1358–1361. [49] B. Vigolo, A. Pénicaud, Claude Coulon, C. Sauder, R. Pailler, C. Journet, P. Bernier, and P. Poulin, “Macroscopic Fibers and Ribbons of Oriented Carbon Nanotubes,” Science, Vol. 290. No. 5495, 2000, pp. 1331–1334. [50] N. Jalili, “Functional Fabric with Embedded Nanotube Sensors/Actuators,” NTC Project: M03-CL07s, http://www.ces.clemson.edu/ssnems/ntc.htm. [51] H. Schreuder-Gibson, K. Senecal, M. Sennett, Z. Huang, J.-G. Wen, W. Li, D. Wang, S. Yang, Y. Tu, Z.Ren, and C. Sung, “Characteristics of Electrospun Carbon Nanotube Polymer Composites,” NanoLab Inc., Boston, MA, 2000, pp. 1–11. [52] L. Y. Yeo, and J. R. Friend, “Electrospinning Carbon Nanotube Polymer Composite Nanofibers,” Journal of Experimental Nanoscience, Vol. 1, No. 2, 2006, pp. 177–209. [53] R. H. Baughman, et al., “Fabrication and Mechanical Characterization of Carbon Nanotube Yarns, 3D Braids and Their Composites,” SAMPE Fall Technical Conference and Exhibition, 2006. [54] M. J. Schulz, G. Maheshwari, J. Abot, Y. Song, C. Jayasinghe, N. Mallik, V. Shanov, M. Dadhania, Y. Yun, S. Yarmolenko, and J. Sankar, “Responsive Nanomaterials for Engineering Asset Evaluation and Condition Monitoring,” BINDT, Insight, June 2008. [55] www.kleindiek.com. [56] D. Salisbury, “New Nanotube Sensor Can Continuously Monitor Minute Amounts of Insulin,” Science Daily, Apr. 17, 2008. [57] R. T. Kachoosangi, M. M. Musameh, I. Abu-Yousef, J. M. Yousef, S. M. Kanan, L. Xiao, S. G. Davies, A. Russell, and R. G. Compton, “Carbon Nanotube-Ionic Liquid Composite Sensors and Biosensors,” Anal. Chem., Vol. 81, No. 1, 2009, pp. 435–442. [58] M. Penza, G. Cassano, P. Aversa, A. Cusano, M. Consales, M. Giordano, and L. Nicolais, “Acoustic and Optical VOCs Sensors Incorporating Carbon Nanotubes,” IEEE Sensors Journal, Vol. 6, No. 4, 2006, pp. 867–875. [59] Y. Wang, K. Kempa, B. Kimball, J. B. Carlson, G. Benham, W. Z. Li, T. Kempa, J. Rybczynski, A. Herczynski, and Z. F. Ren, “Receiving and Transmitting Light-Like Radio Waves: Antenna Effect in Arrays of Aligned Carbon Nanotubes,” Applied Physics Letters, Vol. 85, No. 13, 2004, p. 2607. [60] K. Jensen, J. Weldon, H. Garcia, and A. Zettl, “Nanotube Radio,” Nano Letts., Vol. 7, 2007, pp. 3508–3511. [61] T. Moore, “Wireless Technology Choices Abound for Medical Monitoring,” RTC, 2005, pp1-3. [62] David Mast, Vesselin Shanov, Chaminda Jayasighe, and Mark Schulz, Use of Carbon Nanotube Thread, Ribbon and Arrays for Transmission and Reception of Electromagnetic Signals and Radiation, University of Cincinnati Invention Disclosure: UC 109-035, November 18, 2008. [63] UC Press Release, March 3, 2009, http://www.uc.edu/news/NR.aspx?id=9743
References
59
[64] Y. Yun, A. Bange, W. R. Heineman, H. B. Halsall, V. Shanov, Z. Dong, S. Pixley, M. Behbehani, A. Jazieh, Y. Tu, D. K. Y. Wong, and M. J. Schulz, “A Nanotube Array Immunosensor for Direct Electrochemical Detection of Antigen-Antibody Binding,” Sensors and Actuators B, Vol. 123, No. 1, 2007, pp. 177–182. [65] Y. Yun, V. Shanov, Y. Tu, M. J. Schulz, S. Yarmolenko, S. Neralla, J. Sankar, and S. Subramaniam, “A Multi-Wall Carbon Nanotube Tower Electrochemical Actuator,” Nanoletters, Vol. 6, No. 4, 2006. [66] Y. Yun, V. Shanov, M. J. Schulz, S. Narasimhadevara, S. Subramaniam, D. Hurd, and F. J. Boerio, “Development of Novel Single-Wall Carbon Nanotubeepoxy Composite Ply Actuators,” Smart Material and Structure, Vol. 14, 2005, pp. 1526–1532. [67] Y. Yun, A. Miskin, P. Kang, S. Jain, S. Narasimhadevara, D. Hurd, M. J. Schulz, V. Shanov, T. He, F. J. Boerio, D. Shi, and S. Srivinas, “Carbon Nanofiber Hybrid Actuators, Part I: Liquid Electrolyte-Based,” Journal of Intelligent Material Systems and Structures, Vol 17, No. 2, 2006, pp. 107–116. [68] Y. Yun, A. Miskin, P. Kang, S. Jain, S. Narasimhadevara, D. Hurd, M. J. Schulz; V. Shanov, T. He, F. J. Boerio, D. Shi, and S. Srivinas, “Carbon Nanofiber Hybrid Actuators, Part II: Solid Electrolyte-Based,” Journal of Intelligent Material Systems and Structures, Vol 17 No. 3, 2006, pp. 191–197. [69] A. E. Aliev, J. Oh, M. E. Kozlov, A. A. Kuznetsov, S. Fang, A. F. Fonseca, R. Ovalle, M. D. Lima, M. H. Haque, Y. N. Garstrein, M. Zhang, A. A. Zakhidov, and R. H. Baughman, “Giant-Stroke, Superelastic Carbon Nanotube Aerogel Muscles,” Science, Vol. 323, No. 5921, 2009, pp. 1575–1578. [70] K. A. Crutcher, M. J. Schulz, Y. H. Yun, and V. N. Shanov, “Aligned Carbon Nanotubes Support Directional Neurite Growth In Vitro,” Ninth International Conference on the Science and Application of Nanotubes (NT08), Montpellier, France, June 29–July 4, 2008.
CHAPTER 3
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems for Targeted Cancer Therapy Elena Heister and Thierry Lutz
3.1
Introduction to Targeted Cancer Therapy Approximately every fourth person in the world currently dies of cancer. Although more than 100 different types of cancers have been identified, half of the cancer-related deaths in men and women are attributed to tumors at four sites, namely lung, prostate, breast, and colorectal cancer. Many therapeutic approaches that have been developed and applied in the last decades still suffer from major drawbacks, such as insufficient effectiveness and severe side effects. Both problems have their origin in the lack of specificity of antineoplastic agents for their intended site of action in the body. The emerging field of nanomedicine provides a whole range of materials and techniques to develop customizable, drug delivery vehicles, which can assist the targeting of therapeutic agents to the desired site of action and hence increase the effectiveness of the treatment and reduce side effects. Among these, carbon nanotubes have emerged as promising candidates, as they are capable of penetrating mammalian cell membranes and at the same time allow for the attachment of high loads of drugs and targeting agents on their surface or in their inner cavity. This chapter will introduce functionalized carbon nanotubes as tools for the targeted delivery of anticancer drugs and will highlight the current state of research, as well as future prospects and challenges. The first part of this chapter will give an overview about the disease cancer and review present-day cancer treatments and the associated problems. Subsequently, the concept of drug targeting will be introduced as a possibility to increase therapeutic effectiveness. 3.1.1
Cancer Statistics
Cancer is among the top three killers in modern society, next to heart disease and stroke. According to the American Cancer Society, 7.6 million people died from cancer in the world during 2007 [1]. Although much progress has been made in reducing mortality rates, stabilizing incidence rates, and improving survival, cancer still accounts for more deaths than heart disease in persons under 85 years old [2]. Interestingly, the most common cancers of the past decades continue to be the ones
61
62
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems
that are most often lethal. This highlights the difficulties in treating these types of cancers, as well as the delays in early recognition. As illustrated in Figure 3.1, lung cancer, prostate cancer, breast cancer, and colorectal cancer are the four most deadly types of cancers for men and women. 3.1.2
Present-Day Cancer Treatment and Associated Problems
Cancer is the result of a long process, where a cell of an organ or tissue becomes damaged or altered in a way that causes it to break free from regulatory controls, resulting in uncontrolled growth, invasion of the surrounding tissue, and metastasis. The most commonly applied cancer therapies are surgery, chemotherapy, and radiation therapy, followed by hormone therapy and immune therapy. This chapter will primarily focus on chemotherapy and its remaining drawbacks, which are at the same time the challenges for newly developing anticancer treatments. One of the main difficulties in chemotherapy is the lack of specificity of antineoplastic agents for their intended site of action in the body, which is due to the fact that the biochemical differences between cancerous cells and healthy cells are sparse. Therefore, many anticancer drugs simply target fast-dividing cells, since cancer cells tend to divide and grow faster than most other cells in the body. However, fast-dividing cells can also be found in hair follicles, the bone marrow, the stomach, or the bowel lining, which explains some of the most common side effects of chemotherapy like nausea, bone marrow toxicity, hair loss, lack of appetite, and tiredness. In addition, this general therapeutic approach is only effective for tumors with reasonably high growth fractions, such as some types of leukemia and aggressive lymphomas, and can induce resistance over time. Hence, chemotherapies often comprise a combination of several drugs applied for varying lengths of time. Additionally, cancer is no longer considered as a single disease, but rather as a multitude of independent disorders that can result in malignant cell growth. Even though all
Figure 3.1 The 10 most deadly types of cancer for men and women (estimation in the United States for 2008), adapted from [2].
3.1 Introduction to Targeted Cancer Therapy
63
cancers have to overcome the same spectrum of regulatory controls in order to develop, the involved cell types and affected genes may differ, which results in a vast heterogeneity of more than a hundred different cancer types. In the light of these overwhelming complexities, the development of a single cure or a so-called magic bullet has been proven elusive. The challenge for oncology in the 21st century is to design targeted therapies that minimize the damage to noncancerous cells and their related side effects. The emerging field of nanomedicine provides a whole range of materials and techniques to develop customizable, targeted drug delivery vehicles including quantum dots, nanoparticles, dendrimers, and liposomes [3–6], which can minimize side effects and applied drug doses by targeting the therapeutic agents to the desired site of action. Among these, carbon nanotubes are probably one of the most striking discoveries in recent years. Carbon nanotubes do not only possess unique physical properties, such as tremendous strength, an extreme aspect ratio, and excellent thermal and electrical conductivity, but are also capable of bypassing biological barriers like cell membranes due to their nanoscale dimensions. Furthermore, their high surface area allows for the attachment of a whole spectrum of chemical and biological entities. The combination of these outstanding properties makes carbon nanotubes promising candidates as a new, targeted drug delivery system in the setting of oncology. 3.1.3
A Brief Insight into Targeting Strategies
As discussed above, current anticancer therapies suffer from their lack of effectiveness, which can either be due to insufficient targeting of the therapeutic agents to the desired site of action, or administration problems caused by poor bioavailability, inadequate in vivo stability, and unfavorable solubility of the drugs. In this context, different types of targeting strategies are currently under investigation, which can be vector-based and nonvector-based. The latter comprises the idea of molecular targeting, which describes the design of anticancer drugs that specifically attack fundamental molecular pathways involved in carcinogenesis without affecting normal cells and tissues. A prominent example for this new class of anticancer drugs is imatinib, currently marketed as Gleevec or Glivec, which was approved by the FDA in 2001 for the treatment of chronic myeloid leukemia (CML) and a rare form of stomach cancer called gastrointestinal stromal tumor (GIST). Imatinib’s mode of action involves the inhibition of a particular tyrosine kinase enzyme, which plays a role in tumor cell proliferation [7]. Another nonvector-based therapeutic approach to improve the selectivity of anticancer drugs is the so-called antibody-directed enzyme prodrug therapy (ADEPT). In ADEPT, selectivity for the target is achieved by an antibody in an antibody-enzyme conjugate, which binds to antigens that are preferentially expressed on the surface of tumor cells, or in the tumor interstitium. In the first step, this complex is administered and accumulates at the target site. In the second step, a nontoxic prodrug is injected, which is converted into a cytotoxic drug in situ by the enzyme in the conjugate at the tumor [8]. A major advantage of ADEPT is the amplification effect, as one molecule of enzyme catalyses the conversion of many prodrugs into their cytotoxic form. This leads to higher drug concentrations at the tumor site compared to direct injection of the drug alone.
64
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems
Although these novel, therapeutic strategies are quite fascinating, the thematic priority of this chapter shall lie on vector-based targeting strategies in the framework of nanomedicine. Nanovectors made for drug delivery often consist of three parts: a core constituent material, a therapeutic payload, and biological surface modifiers for targeting purposes. This multifunctional concept allows for the delivery of large amounts of therapeutic agents per targeted site of action, which is a major clinical improvement over simple immunotargeted drugs [9]. Indeed, it has been shown that when drugs are directly attached to antibodies, only a limited number can be attached before a loss in the inherent targeting function of the antibody occurs [10]. Targeting methods using nanovectors can either be based on active strategies employing covalently linked antibodies, aptamers, or other specific ligands, or passive strategies relying on mechanisms based on the size and physical properties of the nanovectors. A typical example for passive targeting is the enhanced permeation and retention (EPR) effect, which originates from the leaky vasculature of tumor-associated blood vessels and can be exploited by particle-mediated drug delivery systems to increase drug concentrations at tumor sites [11]. Liposomes are considered as the archetypal, simplest form of a particulate nanovector. Liposome-encapsulated formulations of the anthracycline anticancer drug doxorubicin (Caelyx®) were already approved by the FDA 10 years ago for the treatment of Kaposi’s sarcoma [12], and are now used against breast cancer and refractory ovarian cancer [6, 13]. However, inhomogeneous materials with high aspect ratios like carbon nanotubes cannot easily exploit the EPR effect and thus often rely on active targeting approaches that involve the attachment of antibodies, aptamers, or other ligands binding to specific target structures inherent of tumor cells. An example based on antigen-antibody interactions is the targeting of cancer types expressing carcinoembryonic antigen (CEA) by nanovector-CEA antibody-conjugates [14, 15]. CEA is part of the immunoglobulin superfamily and is overexpressed by gastrointestinal carcinomas and a number of lung, breast, and ovarian carcinomas. In another simple targeting approach based on the interaction of a ligand with its receptor, a nanovector is conjugated to a cyclic arginine-glycine-aspartic acid (RGD) peptide. RGD peptides impart a recognition moiety for integrin αvβ3 receptors; a class of transmembrane, cell adhesion receptors that are upregulated in a variety of solid tumors and are furthermore involved in many biological processes, such as angiogenesis, thrombosis, inflammation, and osteoporosis [16, 17]. A similar strategy exploits the binding of nanovector-folate conjugates to folate receptors, which are common tumor markers expressed at high levels by a broad spectrum of human cancers, facilitating cellular internalization of folate-conjugated nanovectors by receptor-mediated endocytosis [18, 19]. In the last decade, this approach has been widely investigated by conjugating folic acid to a variety of therapeutic agents, including imaging agents, chemotherapeutic agents, oligonucleotides, proteins, haptens, liposomes, nanoparticles, and gene transfer vectors with subsequent investigation of the respective targeting efficiency [20]. The targeting potential of drug delivery systems can be compared by building the ratio of drug uptake by the tumor and uptake by neighboring, healthy tissues, also called T/N ratio. According to Liu et al., a T/N ratio of 4–5 indicates that the candidate deserves to be further studied for its therapeutic potential [21]. In sum-
3.2 Carbon Nanotubes: A Versatile Material
65
mary, even though all previously mentioned targeting strategies are effective by themselves, the greatest gains in therapeutic selectivity will most likely be achieved by synergistic combinations of active and passive targeting strategies, which can only be achieved by means of a vector.
3.2
Carbon Nanotubes: A Versatile Material In this section, the reader will find out more about carbon nanotubes as a material, their characterization and purification, and about existing challenges related to their biomedical application. 3.2.1
Definition and Synthesis of Carbon Nanotubes
Carbon nanotubes (CNTs) are an allotrope of carbon, like diamond, graphite, or fullerenes, whose unique one-dimensional nanostructure confers fascinating physical properties. Although Radushkevich and coworkers reported about carbon nanotubes as early as 1952 [22, 23], it was not until the observation of multiwalled carbon nanotubes by Iijima in 1991 by means of high-resolution transmission electron microscopy (HRTEM) [24] that the carbon nanotube field was seriously launched. Structurally, carbon nanotubes can be regarded as segments of graphene sheets that have been rolled up to form seamless cylinders. Depending on the angle at which a sheet is rolled up (the nanotube’s chirality), the electrical and optical properties change drastically [25]. Especially notable is the fact that carbon nanotubes behave like metals or semiconductors, depending on their diameter and helicity [26]. Furthermore, carbon nanotubes can be made of more than one sheet of graphene (Figure 3.2). In contrast to single-walled carbon nanotubes (SWNTs), which are formed of a single sheet of graphene and were first described in the literature by Iijima in 1993 [27], multiwalled carbon nanotubes are assemblies of several, coaxially aligned concentric tubes. In the last 15 years, carbon nanotube research has expanded very rapidly. In the beginning, their extraordinary mechanical and electrical properties attracted the most attention [28, 29]. Indeed, the stiffness of carbon nanotubes measured in terms of the Young’s, modulus (ratio of stress over strain) can be as high as 1 TPa, which makes them 5 times stiffer than steel. Their tensile strength or breaking strain can reach values of up to 63 GPa, which is around 50 times higher than steel [30]. However, since these early days, the range of possible applications has greatly expanded and carbon nanotubes can be regarded as one of the most versatile materials existing at present. Carbon nanotubes are currently studied for various applications, including field emission [31, 32], molecular electronics [33, 34], solar cells and energy storage [35, 36], scanning probe microscopy [37, 38], or sensing [39, 40], to name only a few. Closer to the focus of this book, carbon nanotubes are also a subject of huge interest in the biomedical field [41–44]. Carbon nanotubes have originally been obtained in the soot of an arc discharge reactor [24, 45]. Since then, research has focused intensely on the synthesis of carbon nanotubes and various techniques have emerged that are able to produce different types of nanotubes, depending on the desired application [46]. In most cases, a
66
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems
Figure 3.2 Transmission electron microscopy (TEM) images and schematic drawings of single-, double-, and multiwalled carbon nanotubes. The nanotube coordinates were created with the nanotube coordinate generator by Dr. Shigeo Maruyama (http://www.photon.t.u-tokyo.ac.jp/ ~maruyama/wrapping3/wrapping. html).
catalyst is required to induce or modify growth, which is often composed of transition metals such as iron, cobalt, or nickel, and has a strong influence on the toxicological profile of the obtained nanotube material. The following paragraph will briefly elaborate on the two main types of CNT synthesis techniques, which are based on physical vapor deposition (PVD) or chemical vapor deposition (CVD) [47]. PVD techniques rely on the condensation of carbon evaporated at high temperatures (typically 3,000–4,000°C) from a solid carbon source using arc discharge or laser ablation. The obtained nanotubes are in the form of a powder and exhibit a broad length and diameter distribution, depending on the experimental conditions and the catalyst used. Common impurities obtained alongside carbon nanotubes are carbonaceous and include fullerenes, amorphous carbon, and graphite sheets/onions. It is noteworthy that the synthesis of multiwalled carbon nantubes by the arc discharge technique does not require a catalyst. CVD techniques are based on the decomposition of a carbon source—either over a heated substrate, on which the catalyst can be deposited or even patterned, or in a thermal reactor, where unsupported catalyst particles are created in situ. The process can be purely thermal or plasma-assisted and the applied temperatures are much lower compared to arc discharge or laser ablation (e. g., typically lower than 1,200°C.) The diameter and length of the tubes can be controlled by tuning the composition and size of the catalyst particles and the synthesis time. This method is especially useful to obtain arrays of vertically aligned nanotubes for applications in electronics or sensing. The main impurities in CVD carbon nanotubes are metallic catalyst nanoparticles attached to the tube, which has grown from it, next to
3.2 Carbon Nanotubes: A Versatile Material
67
graphitic nanostructures and amorphous carbon. A major disadvantage of this method is that the obtained nanotubes can be highly defective. 3.2.2
Characterization of Carbon Nanotubes
Although the general structure of carbon nanotubes is relatively simple, the material itself can differ significantly, depending on the synthesis technique. Furthermore, many types of purification treatments may alter the composition of the initial material. Thus, when working with carbon nanotubes, it is crucial to analyze their characteristics, such as diameter and length distribution, dispersion state, defect density, and type/composition/quantity of the different impurity species [48]. The main characterization techniques that are of interest in the framework of drug delivery by carbon nanotubes can be divided into microscopic techniques, spectroscopic techniques, and thermogravimetric techniques. Microscopic techniques are the only tools that allow for the direct observation of the morphology and dimensions of carbon nanotubes in a bulk sample. Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) are based on the interaction of electrons with matter. SEM has always been one of the most popular tools for the characterization of carbon nanotubes, as it provides a general idea of the morphology and purity of a carbon nanotube sample in connection with simple and straightforward sample preparation. TEM, on the other hand, affords very detailed information about a carbon nanotubes sample, such as the number of walls, length, diameter, filling state, and even chirality [49], but requires the laborious preparation of thin, solid specimen (< 200 nm). A major drawback of both techniques is the fact that the image frame visualizes less than 1 pg of material, which makes it impossible to achieve a quantitative and statistically significant assessment of the quality of carbon nanotube material in a reasonable amount of time [50]. The third one of the listed microscopic technique is atomic force microscopy (AFM) and works in a slightly different way. In fact, the term microscopy is not quite correct for this technique, as these microscopes do not use light or electrons, as in conventional microscopes, but scan a sharp probe over a surface and thus a create three-dimensional surface profile of a sample instead of a two-dimensional projection. Moreover, most AFM modes can work in ambient air or even liquid environments, which allows for the imaging of biological macromolecules and living organisms. With respect to the characterization of carbon nanotubes, AFM is used to obtain information about diameter and length distributions and can furthermore provide information about surface coatings. Figure 3.3 shows the characterization results of a sample of single-walled carbon nanotubes by means of the three introduced microscopic techniques; SEM, TEM, and AFM. Moving on to the spectroscopic methods, the major techniques for characterizing carbon nanotubes are UV/vis/NIR absorption spectroscopy and Raman spectroscopy. In UV/vis/NIR absorption spectroscopy, the intensity of a beam of light is compared before and after interaction with a sample. This provides information about the concentration of a sample solution, as, according to Lambert-Beer’s Law, absorbance and concentration are directly proportional. Furthermore, the vis/NIR part of a carbon nanotube spectrum (600–2,500 nm) exhibits characteristic absorption features due to the van Hove singularities, which originate from the one-dimen-
68
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems
Figure 3.3
SEM, TEM, and AFM image of a sample of single-walled carbon nanotubes.
sional nature of the nanotubes, inducing quantization of the density of states [51, 52]. According to Itkis et al., the intensity of these characteristic features compared to the featureless baseline provides a measure of the purity of the SWNT material [53]. However, these discrete absorption bands can only be observed for good dispersions and nonoxidized carbon nanotubes, since strong acid treatment alters their electronic properties [54]. Raman spectroscopy is another extremely powerful tool for the characterization of SWNTs, delivering detailed information on the diameter distribution, electronic structure, chirality, and purity of a carbon nanotube sample [55]. It is usually referred to as resonant Raman spectroscopy, since the strength of the obtained signal is greatly enhanced when the excitation wavelength of the laser is in resonance with the energy differences between the van Hove singularities. The characteristic bands that are found in a Raman spectrum of single-walled carbon nanotubes are the radial breathing mode (RBM), the D-band, and the G-band. The RBM results from low-energy radial vibrations of carbon atoms in the nanotube backbone and its frequency is inversely proportional to the tube diameter [56]. The disorder-induced D-band indicates the presence of defective sites on SWNTs, which are characterized by sp3 hybridization instead of the regular sp2 hybridization that is inherent of all graphitic materials. The G-band is a tangential vibrational mode characteristic to all graphitic materials. The intensity ratio of D-band to G-band is often used to estimate the defect density of carbon nanotubes [57]. The drawback of this resonance-based method is that a wide range of excitation wavelengths is required in order to obtain a full picture of a nanotube sample, which generally consists of a variety of tubes with different diameters and chiralities. Furthermore, Raman spectroscopy does not afford quantitative results. The last one of the characterization techniques presented in the framework of this chapter is thermogravimetric analysis (TGA). TGA determines changes in the weight of a sample when heated under a controlled atmosphere and records them as a weight-loss curve. The weight change that occurs when a carbon nanotube sample is burned under an oxidative atmosphere is a superposition of the weight loss due to oxidation of carbon into gaseous carbon dioxide and the weight gain due to oxidation of the residual metal catalyst. The relative amount of the different fractions of carbonaceous impurities in as-prepared carbon nanotubes can be identified if their combustion temperatures are well separated, since amorphous carbon burns in air at a lower temperature than graphitic carbon [58]. A sample containing a high
3.2 Carbon Nanotubes: A Versatile Material
69
amount of carbonaceous impurities would thus cause a noticeable weight loss below 400°C, followed by a second weight loss at higher temperatures [47]. Nevertheless, TGA is still primarily used for the determination of the amount of residual metal catalyst in carbon nanotube samples. In summary, it is important to underline that even though each of the discussed characterization techniques is very useful for determining certain characteristics of carbon nanotube bulk samples, they all have their limitations in the data produced and the conclusions that can be drawn. Hence, no single technique can give a complete description of a carbon nanotube sample. Moreover, due to the complexity of these samples, the whole range of techniques is usually needed in order to get a satisfactory idea of the investigated material. 3.2.3
Purification of Carbon Nanotubes
As mentioned earlier, as-prepared carbon nanotubes usually contain a range of impurities originating from the synthesis process, which can be divided into carbonaceous impurities and metallic impurities. These impurities can interfere with the desired properties of the nanotubes, whether for fundamental science or industrial use. Thus, many purification techniques have been developed to improve the quality of the raw carbon nanotube material [59–64]. The most commonly used purification methods are based on oxidative processes and can be carried out in liquid phase (acid oxidation) or gas phase. Acid oxidation of carbon nanotubes involves the treatment of carbon nanotubes with strong acids or acid mixtures, which gasify amorphous carbon and oxidize metallic catalyst particles, as long as these are not encapsulated by carbon shells. This process is known to reduce the length of the nanotubes and to introduce carboxylic groups, primarily at the tips and defective sites along the sidewalls, as these spots feature increased chemically reactivity [65]. The introduction of carboxylic groups by acid oxidation does not only open up the whole field of functionalizing carbon nanotubes by means of organic chemistry, but also renders the nanotubes soluble in common hydrophilic solvents like water. On the other hand, the generation of new defects can affect the electronic properties of the nanotubes in a negative way. In summary, acid oxidation is considered to be a very efficient, but harsh purification method for carbon nanotubes, which can sometimes result in the loss of 90% of the initial material [66]. Gas oxidation makes use of the fact that carbon nanotubes are less sensitive to oxidation than other carbon species produced in the synthesis process, which can thus be removed in preference to nanotubes. This method is primarily applied to remove carbonaceous impurities and to expose metallic catalyst particles encapsulated in graphite onions (see Figure 3.4). However, gas oxidation does not remove metallic impurity species. This can cause problems, since in the presence of oxidizing gases, metal particles catalyze low temperature oxidation of the nanotubes [67], which eventually leads to their destruction. It has been proven beneficial to apply a combination of acid oxidation and heat treatments, which can effectively remove both impurity species. Unwanted functionalities and defects on the nanotubes generated during the purification process can subsequently be removed by vacuum annealing, which restores the graphitic structure of the nanotubes [68].
70
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems
Figure 3.4
Metallic catalyst particles encapsulated in graphite onions.
Finally, an entirely different technique shall be mentioned, which has been developed in our group by Jeynes et al. [69]. Therein, carbon nanotubes are solubilized by wrapping with RNA, based on π-stacking interactions. Centrifugation then allows for separation of the RNA-wrapped carbon nanotubes from the impurities. In the last step, the enzyme RNase is used to remove the RNA in order to yield unmodified, highly pure single-walled carbon nanotubes. 3.2.4
Functionalization of Carbon Nanotubes for Biomedical Applications
In order to apply carbon nanotubes in the biomedical field, two major technical barriers have to be overcome. First, carbon nanotubes are not soluble in water and other water-based physiological fluids, and second, they are chemically inert. Thus, functionalization of carbon nanotubes is a key step for their integration into different biological and physiological environments. Besides, it has been shown that functionalization improves the biocompatibility of carbon nanotubes in vitro and in vivo [70]. Functionalization techniques can be divided into noncovalent and covalent approaches. In the noncovalent functionalization techniques, carbon nanotubes are wrapped or covered with molecules that act as surfactants. These can be common anionic, cationic, or nonionic surfactants, but also complexing agents, organic biopolymers, peptides, proteins, oligomers, or bile salts [71–77]. The lipophilic parts of these molecules generally attach to the sidewalls of the nanotubes via hydrophobic interactions or π-stacking, whereas the hydrophilic parts stick out into the solution and provide aqueous solubility. A particularly interesting approach is the wrapping of carbon nanotubes with single-stranded DNA or RNA, as this increases the biocompatibility of the nanotubes and enables them to interact with living tissue [69, 78, 79]. The advantage of the noncovalent approach lies in the preservation of the intrinsic sp2 structure of the nanotubes, which determines their electronic characteristics. However, noncovalent binding is dependent on environmental factors,
3.3 Carbon Nanotubes as Nanovectors for Multimodal Drug Delivery
71
such as pH and salt concentration, and is in general less stable than covalent binding. Covalent functionalization of carbon nanotubes mostly takes advantage of the reactivity of carboxylic acid moieties created by acid oxidation of carbon nanotubes, although various other approaches have been developed [80, 81]. The acid treatment has to be adjusted for every nanotube sample of a different origin with respect to acid concentration, time and temperature, as those samples with a high initial rate of defects and those containing tubes with small diameters are more sensitive to oxidation and thus are destroyed more easily [82, 83]. Li et al. have compared different acid oxidation methods for single-walled carbon nanotubes with respect to purity and found that presonication in diluted nitric acid, followed by refluxing in a 3:1 mixture of concentrated sulphuric and nitric acid yielded the best results [84]. Once oxidized, carbon nanotubes can be coupled covalently to a variety of different biomolecules bearing amino groups. However, the carboxylic groups of oxidized nanotubes are not reactive enough to undergo spontaneous reactions. A common solution to this problem is to convert the carboxylic groups into amine-reactive esters using sulfo-NHS (N-hydroxysulfosuccinimide) and EDC (1-Ethyl-3- [3-dimethylaminopropyl] carbodiimide hydrochloride); the latter serving as a dehydrating agent [85, 86]. Both EDC and sulfo-NHS are soluble in water and thus allow for coupling reactions in physiological environments. Another, slightly more time-consuming method involves the transformation of the carboxylic acid moieties into the respective acyl chlorides by means of thionyl chloride, which are significantly more reactive towards amines [87, 88]. Apart from covalent functionalization approaches based on oxidation of carbon nanotubes, other options are based on cycloadditions or radical reactions [89]. Within the different types of organic reactions, Georgakilas et al. have developed the 1,3-dipolar cycloaddition reaction of azomethine ylides, which creates pyrrolidine rings distributed along the nanotubes sidewalls and tips [90]. This method has been successfully applied for the attachment of peptides, fluorescent markers, and different types of drugs to carbon nanotubes [91] and even permits the generation of double functionalized carbon nanotubes [92].
3.3
Carbon Nanotubes as Nanovectors for Multimodal Drug Delivery The following paragraphs will elaborate on the functionalization of carbon nanotubes with drugs and other therapeutic agents for targeted cancer therapy in the framework of the current state of research. 3.3.1 Carbon Nanotube Drug Delivery Systems Based on Surface Functionalization
Functionalized carbon nanotubes can be conjugated to a wide variety of molecules, such as peptides, proteins, nucleic acids, or therapeutic agents, which renders them promising candidates as a platform for biosensors [93, 94]. Additionally, they are capable of penetrating mammalian cell membranes and can hence be utilized as shuttles for biomolecules, genes or drugs into mammalian cells. Interestingly, the
72
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems
uptake mechanism is still subject of intense discussion and seems to depend on the physicochemical properties of the nanotubes and on the type of functionalization. Pantarotto et al. describe the process of cell entry as passive and endocytosis-independent, suggesting a mechanism in which the nanotubes act as nanoneedles and pierce through the cell membrane in a mechanical way [95]. On the other hand, Kam and coworkers claim that carbon nanotubes are taken up by mammalian cells via clathrin-dependent endocytosis, since the addition of endocytosis inhibitors, such as sodium azide, clearly influenced nanotube uptake in their experiments [96]. In recent years, carbon nanotubes have been successfully applied to transport a broad spectrum of molecules into mammalian cells. The first approaches focused on the delivery of peptides and proteins attached to carbon nanotubes [95, 97, 98], whereas a few years later, carbon nanotubes were demonstrated to transport genes into cells without changing the properties of the cargo [99]. Recent studies focus intensely on the use of carbon nanotubes for multimodal, targeted drug delivery (Table 3.1). Figure 3.5 shows a schematic illustration of a multifunctionalized carbon nanotube loaded with therapeutic agents in the inner cavity and on the surface, antibodies for cellular targeting, a fluorescent marker for in vitro and in vivo visualization, and PEG to avoid recognition and uptake by the reticuloendothelial system. The first attempts to employ carbon nanotubes for drug delivery have mainly focused on the functionalization of the nanotubes with various types of drugs and the question, whether the therapeutic effect of the payload could be preserved. One of the first studies has been carried out by Wu et al., who conjugated the antibiotic amphotericin B to multiwalled carbon nanotubes [100]. Amphotericin B is considered as a problematic drug due to its narrow therapeutic index. Although it belongs to the most effective antibiotics for the treatment of chronic fungal infections, it is
Table 3.1
Recent Studies Investigating Carbon Nanotubes as Targeted Drug-Delivery Systems
Researchers
Year of Publication
Type of CNT
Delivered Agent
Targeting Strategy
Wu et al. [100] Pastorin et al. [92] Ali-Boucetta et al. [101] Feazell et al. [102] Liu et al. [103] McDevitt et al. [104]
2005 2006 2008 2007 2007 2007
Multiwalled Multiwalled Multiwalled Single-walled Single-walled Single-walled
None None None Prodrug
Chen et al. [105]
2008
Single-walled
Dhar et al. [106] Liu et al. [107] Villa et al. [108]
2008 2008 2008
Single-walled Single-walled Single-walled
Bhirde et al. [109]
2009
Single-walled
Amphotericin B Methotrexate Ddoxorubicin Cisplatin Doxorubicin Therapeutic radio-isotopes (111In, 90Y) Second-generation taxoid (SB-T-1214) Cisplatin Paclitaxel Radioisotope (111In) for labeling Cisplatin
Heister et al. [110]
2009
Single-walled
Doxorubicin
RGD peptide
Monoclonal antibodies (rituximab, lintuzumab) Prodrug + biotin Prodrug + folate EPR effect RGD peptide Epidermic growth factor (EGF) Monoclonal antibody (antiCEA)
3.3 Carbon Nanotubes as Nanovectors for Multimodal Drug Delivery
73
Figure 3.5 Simplified, schematic illustration of a carbon nanotube-based drug delivery system carrying anticancer drugs, antibodies as targeting moieties, a fluorescent marker, and PEG molecules to escape capture by the reticuloendothelial system.
highly toxic to mammalian cells, partly due to its low solubility in water that causes the formation of aggregates [111]. These aggregation phenomena and solubility problems could be solved by conjugation of amphotericin B to carbon nanotubes. Indeed, studies demonstrated a reduced toxic effect towards mammalian cells, while the antifungal activity of the drug was preserved. A study published a year later by the same group took a step forward and conjugated the first anticancer drug to multiwalled carbon nanotubes, again together with a fluorescent marker using a special double functionalization technique [92]. The investigated drug methotrexate is a widely used drug in oncology, but suffers from low cellular uptake and induction of drug resistance [112]. Conjugation to multiwalled carbon nanotubes facilitated the internalization of methotrexate by human Jurkat T lymphocytes. Both studies employed covalent coupling chemistry to attach the drug. However, in this case the efficiency of the therapy depends on the nature of the covalent bond between the nanotubes and the small therapeutic molecules. Thus, Ali-Boucetta and coworkers have developed a functionalization approach based on noncovalent interactions by attaching the anthracycline anticancer drug doxorubicin to solubilized multiwalled carbon nanotubes via π-stacking [101]. The doxorubicin-nanotube complex showed a significant enhancement of cytotoxicity compared to doxorubicin alone, indicating that mulitwalled carbon nanotubes can mediate the delivery of doxorubicin and hence enhance its cellular uptake. These results were a promising start for the development of carbon nanotube-based drug delivery systems; however, the aspect of drug targeting had not yet been considered. The first step in this direction was made by Feazell and colleagues,
74
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems
who prepared a construct of single-walled carbon nanotubes and a platinum(IV) complex, serving as a prodrug [102]. The platinum(IV) complex was attached to the amine-functionalized nanotubes through disulfide linkage of one of its axial ligands. This functionalization scheme for single-walled carbon nanotubes had been developed earlier by the same group to afford nanotube-biomolecule conjugates with cleavable bonds, since disulfide bonds are selectively split in the reducing environment of endosomes following cellular internalization [113]. To investigate the feasibility of this method, the testicular carcinoma cell line NTera-2 was treated with the free platinum(IV) prodrug and the nanotube-conjugate for 4 days. Compared to the active form of the drug, cisplatin, the cytotoxicity of the platinum(IV) prodrug was insignificant. The prodrug-nanotube conjugate, however, showed a substantial increase in cytotoxicity (IC50 = 0.02 μm) and surpassed that of cisplatin (IC50 = 0.05 μm) when compared on a per platinum basis. A different targeting approach has been followed up in the same lab, this time based on ligand-receptor interactions [103]. In a study by Liu et al., the anticancer drug doxorubicin was attached to single-walled carbon nanotubes via π-stacking, similar to the work of Ali-Boucetta et al. The nanotubes were solubilized by an amine-terminated phospholipid-PEG conjugate, to which a cyclic RGD peptide was attached. As described earlier, RGD peptides are ligands to integrin αvβ3 receptors, which are upregulated in a wide range of solid tumors. Confocal fluorescence imaging experiments indicated that the delivery of doxorubicin by RGD-tagged drug-nanotube conjugates to integrin αvβ3-positive U87MG human glioblastoma cancer cells was enhanced. To support this preliminary finding, U87MG cells were incubated with series concentrations of SWNT-doxorubicin, SWNT-RGDdoxorubicin and free doxorubicin for 24 hours. Cell viability was evaluated by means of an MTS assay and recorded in dose-response curves. SWNT-doxorubicin showed the highest IC50 value (~8 μm), followed by SWNT-RGD-doxorubicin (~3 µm) and finally free doxorubicin (~2 μm). This demonstrates that RGD-tagging is able to enhance the cytotoxic effect of the nanotube drug complexes, although free doxorubicin was still slightly more effective than the RGD-tagged drug conjugate. In the same year, McDevitt and coworkers conducted a study in which a CNT-antibody construct was designed and synthesized to specifically target the CD20 epitope on human Burkitt lymphoma cells and deliver a radionuclide to these cells [104]. In addition, the tumor uptake was quantified in vivo and results were obtained about blood clearance and the distribution of the complex to other key organs. The employed CD20 antibody rituximab was approved by the FDA in 1997 as a therapeutic antibody for the treatment of B-cell non-Hodgkin lymphoma resistant to other chemotherapy regimens [114]. In McDevitt’s study, it was shown that the covalent attachment of antibodies to the nanotube conjugates dramatically altered the kidney biodistribution and pharmacokinetics when comparing tumor-bearing and nontumor-bearing mice, indicating that the ability to specifically target CD20 expressing tumors was antibody-mediated and not determined by the CNT portion of the construct. In 2008, the investigation of carbon nanotube-based drug delivery systems was seriously launched. Five more studies have been carried out since then, using even more sophisticated nanotube constructs and targeting strategies. Chen and coworker, to start with, conjugated single-walled carbon nanotubes to a prodrug of
3.3 Carbon Nanotubes as Nanovectors for Multimodal Drug Delivery
75
a taxoid anticancer agent by a cleavable linker, which is activated to its cytotoxic form inside the tumor cells upon internalization and is thus released in situ [105]. This is similar to the idea of Feazell and coworkers, who attached a prodrug of cisplatin to carbon nanotubes. However, this study went a step further using nanotube-taxoid conjugates equipped with tumor-targeting ligands that specifically recognize cancer-specific receptors on the cell surface and induce receptor-mediated endocytosis. They chose biotin as the tumor-targeting module, based on the finding of Russell-Jones et al. [115] that biotin receptors on a wide range of tumor types could serve as a new tumor-specific target in a manner similar to the widely recognized folate receptors. The anticancer agent, simply referred to as taxoid, is a second-generation taxoid exhibiting a 100- to 1,000-fold higher potency against multidrug-resistant (MDR) cancer cell lines than paclitaxel, which is a widely used anticancer drug of the class of spindle poisons [116]. In order to evaluate the internalization efficacy, L1210FR leukemia cells were incubated with the nanotube-biotin complexes carrying fluorescently labeled taxoid molecules. Confocal microscopy imagines clearly showed that the fluorescent taxoid found the way to its target, the microtubules. Furthermore, L1210FR cells were incubated for 72 hours with the nanotube-biotin-taxoid complex and taxoid alone and cytotoxicity was evaluated by means of an MTT assay. The results indicated that the drug delivery by the nanotube-based, targeted delivery system (IC50 ~ 51 nm) was superior to simple exposure of the drug itself to the same cancer cell line (IC50 ~ 88 nm). Another study carried out in 2008 by Dhar et al. is a followup study of the one carried out by Feazell et al. a year before, in which a platinum(IV) complex serving as a prodrug of cisplatin was attached to amine-functionalized nanotubes through disulfide linkage of one if its axial ligands. In the Dhar-study, folic acid was attached additionally to the other axial ligand in order to enable cellular targeting [106]. As mentioned at the beginning of this chapter, this targeting strategy is based on the fact that folate receptors are expressed at high levels by a broad spectrum of human cancers and can thus facilitate cellular internalization of folate-conjugated nanovectors by receptor-mediated endocytosis [18]. The results of cytotoxicity studies showed that the complex exhibited high and specific binding to the folate receptor and upon reduction in the cellular environment after uptake forms the major cisplatin intrastrand crosslinks with nuclear DNA. The cell-killing properties of the nanotube-folate-platinum(IV) conjugates were enhanced by a factor of 8.6 compared to folate-platinum (IV) conjugates without nanotubes. This once more provides strong evidence that platinum-SWNT constructs containing a folate component are a valuable tool for the targeted delivery of platinum anticancer agents. The next three studies presented took another big step forward from in vitro experiments to in vivo animal studies. Liu and coworkers conducted the first successful study in which carbon nanotubes were used as drug delivery vehicles to achieve efficient in vivo tumor treatment in mice [107]. They conjugated paclitaxel, a widely used chemotherapeutical drug, to branched polyethylene glycol chains on single-walled carbon nanotubes via a cleavable ester bond to obtain a water-soluble nanotube-paclitaxel conjugate. The conjugate afforded higher efficacy in suppressing tumor growth than paclitaxel in a murine 4T1 breast cancer model, owing to prolonged blood circulation and 10-fold higher tumor uptake of paclitaxel; likely through enhanced permeability and retention (EPR). Paclitaxel molecules carried
76
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems
into the reticuloendothelial system were shown to be released from the nanotube carrier and excreted via the biliary pathway without causing obvious toxic effects to normal organs. This pilot study opens up the potential to translate carbon nanotube-based drug delivery into the clinic. However, the researchers pointed out that the efficacy of their experiments depended largely on the nanotube functionalization chemistry, since various other functionalization attempts had failed to give satisfactory treatment results. Furthermore, they expect to enhance treatment efficacy in the future by conjugating targeting ligands to their complexes, which would combine a passive and an active targeting strategy, leading to a synergistic effect. The second in vivo study carried out by Villa and coworkers employed covalently modified single-walled carbon nanotubes bearing single-stranded oligonucleotide analogs, radiotracing moieties, and a RGD peptide for targeting [108]. These constructs, however, were not designed for drug delivery, but as a prototype for targetable nanotube platforms capable of hybridizing cDNA addresses. Biodistribution studies of the radiolabeled SWNT-oligonucleotide constructs showed rapid clearance from the blood circulation with significant retention in kidney, liver, and spleen only, which is in good conformity with previous studies on this subject. Finally, the first study demonstrating in vivo drug delivery by an active targeting approach was published recently in January 2009 by Bhirde and colleagues in the U.S. [109]. In this study, single-walled carbon nanotubes were functionalized with the first-line anticancer drug cisplatin, epidermal growth factor (EGF), and quantum dots by covalent coupling chemistry. The targeting approach is based on the fact that many squamous cancer cells overexpress the EGF cell surface receptor and can hence be targeted by attaching the ligand EGF to the drug delivery system. Preliminary in vitro studies based on the MTT assay showed that SWNT-cisplatin-EGF dispersions with 1.3 μM cisplatin were more effective at cell killing than 10 μM free cisplatin. In the followup in vivo study, nude athymic mice bearing 7- to 10-mm tumor xenografts were injected with the nanotubebioconjugates three times at 48-hour intervals and tumor growth was monitored for 2 weeks. Mice treated with a nontargeted nanotube-cisplatin conjugated did not show tumor regression, whereas mice treated with the targeted conjugates showed a rapid decrease of tumor size. Recent work in our laboratory [110] follows up the work of Ali-Boucetta [101] and Liu [103], who attached the anthracycline anticancer drug doxorubicin to carbon nanotubes via π-π interactions. The stability of this noncovalent bond decreases with decreasing pH due to deprotonation of doxorubicins primary amino group. This affords specific release of the drug in endosomes upon cellular uptake, which are characterized by a slightly acidic pH. Ali-Boucetta and coworkers did not yet include a targeting strategy, whereas Liu and coworkers attached a cyclic RGD peptide as a targeting moiety to an amine-terminated phospholipid-PEG conjugate that was used to solubilize the nanotubes. However, in this case both the drug and the phosh-pholipid-PEG-RGD conjugate were attached to the nanotube via noncovalent interactions and thus had to share the nanotubes surface. We hence considered it advantageous to link the drug and the targeting agent to the nanotubes at separate binding sites. This was accomplished by attaching doxorubicin to oxi-
3.3 Carbon Nanotubes as Nanovectors for Multimodal Drug Delivery
77
dized carbon nanotubes in a noncovalent way to allow for its release after cellular uptake and then using a covalently attached protein as a linker for two additional agents: a fluorescent marker to enable tracking of the complex in cancer cells via confocal microscopy, and a monoclonal antibody for cellular targeting. Preliminary observations showed that the complexes are taken up by WiDr colon cancer cells and subsequently release doxorubicin, which then translocates to the nucleus, whereas the nanotubes remain in the cytoplasm. Figure 3.6 illustrates the designed conjugates (without antibodies). 3.3.2 Carbon Nanotube Drug Delivery Systems Based on Filling of the Inner Cavity
Another possibility to functionalize carbon nanotubes with therapeutic agents lies in the filling of the inner cavity (Figure 3.7) to allow for the protection of unstable drugs and controlled drug release at the desired site of action. A simple technique to fill carbon nanotubes is based on capillary forces. Since the interior of the nanotubes is very hydrophobic, it is advantageous to wet the inner surface of the nanotubes in order to facilitate filling [117, 118]. Furthermore, the surface tension of the filling fluid must be low enough. Many different types of molecules have already been successfully encapsulated in carbon nanotubes, including fullerenes, metals, DNA/RNA, or polymeric nanoparticles [119–122]. However, the filling of carbon nanotubes with therapeutic agents is still in its infancy. Hilder et al. developed a theoretical approach to the question, whether a nanotube drug carrier could be engineered in a way that makes
Figure 3.6 (Color plate 3) Carbon nanotube drug delivery shuttle functionalized with the anticancer drug doxorubicin (red), the linker protein BSA (purple) and a fluorescent marker attached to BSA (green). The complex further affords the attachment of monoclonal antibodies for targeting purposes at separate binding sites.
78
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems
Figure 3.7 (Color plate 4) Schematic illustration of a carbon nanotube filled with two molecules of the anticancer drug doxorubicin.
it energetically favorable for drug molecules to be encapsulated, and once taken up by cells, energetically favorable to be ejected. In their first study, they examined the suction behavior of cisplatin, a widely used, platinum-based anticancer drug. They concluded that a carbon nanotube must have a radius of at least 4.785 Å and preferably of 5.27 Å to accept cisplatin and take advantage of the maximum suction energy [123]. In another publication, the same group extended their previous theoretical work by investigating the encapsulation of two further anticancer drugs with more complicated molecular structures; paclitaxel and doxorubicin. Since these molecules may enter the tube at any orientation, the results are presented as probabilities of encapsulation. The highest probability of achieving both encapsulation and maximum uptake (or suction energy) for paclitaxel occurs in the radii range of 9.134 to 12.683 Å and for doxorubicin at approximately 8.855 to 10.511 Å [124]. This corresponds to diameters ranging from 1.8 to 2.5 nm, which are found in double-walled carbon nanotubes. To the best of our knowledge, only one practical study about the filling of carbon nanotubes with a therapeutic agent has been published so far, namely by one of our collaborating partners in the framework of the Marie Curie Research Training Network CARBIO. The group demonstrated successful filling of multiwalled carbon nanotubes with carboplatin using a wet chemical approach [125]. Carboplatin is a platinum-based anticancer drug similar to its parent drug cisplatin, but is more water-soluble and causes fewer side effects. The filling experiments showed that the drug concentration and the process temperature exert an important influence on the filling yield, which was about 30 wt% using optimal conditions (see Figure 3.8). In vitro studies with EJ28 human bladder cancer cells revealed a concentration-dependent cytotoxic effect of carboplatin-filled nanotubes on the cells, whereas unfilled, opened nanotubes did not have an effect on cell growth. However, the
3.3 Carbon Nanotubes as Nanovectors for Multimodal Drug Delivery
Figure 3.8
79
TEM images of multiwalled carbon nanotubes filled with carboplatin.
cytotoxic effect of the carboplatin-filled nanotubes could not be compared directly to the effect of the free drug due to the lack of quantitative data on the release of carboplatin from their nanotube carriers. Another study worth mentioning at this point was recently carried out by Ajima and coworkers [126]. Instead of carbon nanotubes, they envisaged another carbon-based nanomaterial, so called single-wall carbon nanohorns, as a new type of drug delivery system. Single-wall carbon nanohorns can be considered as nanoaggregates composed of single-walled carbon nanotubes with closed ends. In a previous study, the group had attached the anticancer-drug doxorubicin to oxidized single-walled carbon nanohorns via a PEG-linker. When these conjugates were administered intratumorally to mice bearing a human nonsmall cell lung cancer tumor (NCI-H460), they caused significant retardation of tumor growth associated with prolonged doxorubicin retention in the tumor [127]. Following up these promising results, they then incorporated the anticancer agent cisplatin inside oxidized single-wall carbon nanohorns through holes that had been opened by a nanoprecipitation method with a filling yield of 46% [126]. The total released quantity of the drug examined by means of atomic absorption spectroscopy was found to be 100%. The cisplatin-nanohorn complexes had a higher antitumor efficiency than free cisplatin in vitro and in vivo. According to the research group, this might be due to adherence of the drug delivery complexes to the cells, which would increase the local concentration of cisplatin and maintain this high concentration over a long period of time due to slow release. In fact, in vivo experiments using the same animal model than in their previous study (NCI-H460 nonsmall cell lung cancer tumors) showed that the drug-nanohorn complexes stayed in the tumor tissues for up to 25 days. Apart from that, this study demonstrated for the first time that single-wall carbon nanohorns themselves exhibited an anticancer effect in vivo, whereas no toxicity could be detected in relation with normal, non-cancerous tissues.
80
3.4
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems
Challenges and Future Prospects The previous sections demonstrated the great potential of carbon nanotubes and nanohorns for the development of multimodal drug delivery systems. The last part of this chapter will now discuss prospects and future challenges of the application of carbon nanotubes in nanomedicine with a particular focus on toxicity issues and biodistribution behavior. 3.4.1
Toxicological Aspects
With carbon nanotubes gradually playing a bigger role in bioapplications, it has become apparent that these nanostructures can have adverse health effects owing to their small size and extreme aspect ratio. Due to the growing public and commercial concern about the safety of nanomaterials, the field of nanotoxicology has been developing swiftly in the last years. With respect to carbon nanotubes, most of the studies undertaken can be differentiated into three areas: their impact on workers involved in their manufacturing and handling, the potential harmful effects of therapeutic carbon nanotubes on the patient, and the toxicity issues that arise through the incorporation of carbon nanotubes in commercially available products, leading to ecotoxicity upon disposal. The following paragraphs will attempt to give an overview of the recent situation in each of the mentioned areas in order to sensitize the reader to potential problems that can arise in the field of nanomedicine. The first area of carbon nanotube toxicology involves their impact on workers involved in their manufacturing and handling. A superficial resemblance of carbon nanotubes and asbestos fibers has led to the intense investigation of their pulmonary toxicity, since inhalation of asbestos fibers is known to induce a chronic inflammatory medical condition of the lungs named asbestosis, which eventually leads to the formation of malignant mesothelioma. A study by Poland and coworkers comparing carbon nanotubes with asbestos gained wide attention in the media in the beginning of 2008. The researchers exposed the mesothelial lining of the body cavity of mice (as a surrogate for the mesothelial lining of the chest cavity) to long, multiwalled carbon nanotubes and demonstrated their asbestos-like, length-dependent, pathogenic behavior, which was characterized by inflammation and the formation of lesions known as granulomas [128]. However, from a critical point of view one could argue that this study did not follow the natural route of exposure, inhalation, but introduced the nanomaterial directly into the body by intraperitoneal injection. Nevertheless, the toxicologic paradigm made by the same group, stating that a fiber thinner than 3 μm, longer than 20 μm, and biopersistent in the lungs can be hazardous, should be seriously considered [129]. This paradigm is based on the circumstance that long and rigid nanofibers cannot be engulfed by macrophages in contrast to short or entangled fibers (see Figure 3.9), which leads to their accumulation in the tissue and promotes inflammatory, mutagenic, and eventually carcinogenic effects [130]. A more natural route of exposure has been followed in a recent study by Shvedova and coworkers [131], who prepared stable and uniform carbon nanotube dispersions by a newly developed aerosolization technique. C57BL/6 mice were exposed to aerosolized carbon nanotubes in a whole body inhalation chamber for 5
3.4 Challenges and Future Prospects
81
Figure 3.9 Effect of carbon nanotube structure on phagocytosis: short and entangled nanotubes are easily engulfed by macrophages followed by lymphatic drainage, whereas long and rigid nanotubes cannot be phagocytosed and hence accumulate in the tissue, where they promote carcinogenesis. (Adapted from [130].)
hours per day on four consecutive days and were found to develop inflammatory responses, oxidative stress, collagen deposition, fibrosis, and mutations of a certain gene in their lungs. In combination with the first study, this clearly indicates that carbon nanotubes can pose an occupational inhalation exposure hazard and hence, strict safety precautions should be taken by workers to avoid inhalation. The second area concerned with carbon nanotube toxicology, namely their application as therapeutic agents, has become of increasing interest in the biomedical field due to the unique chemical and physical properties carbon nanotubes possess in comparison with other pharmaceutical agents and excipients. However, precisely these properties (i.e. their ability to cross biological barriers and their high aspect ratio) might be the cause for detrimental effects on human health [132]. While short-term effects like cytotoxicity are already widely studied, data about long-term effects like mutagenicity and carcinogenicity are scarce, since carbon nanotubes have only been actively investigated for therapeutic purposes for about a decade. According to various studies, the toxicity of carbon nanotubes in the human body seems to depend on a large number of parameters, such as their structure, length, aspect ratio, degree of aggregation, extent of oxidation, surface topology, bound functional groups, and spectrum of present impurities, as well as their concentration and the dose to cells or organisms are exposed [133]. Furthermore, the route of exposure plays a major role, as it determines the types of tissues that are first exposed to the nanotubes. Therapeutic carbon nanotubes are mostly introduced into the body via local or systemic injection. In the first case, the nanotubes get directly into contact with the target tissue and are taken up by cells before entering the bloodstream. To examine this scenario, several in vitro studies have been carried out in order to investigate the effect of carbon nanotube exposure on mammalian cells. In many cases (though not all), carbon nanotubes and aggregates did not show any immediate cytotoxicity [134, 135], but were found to cause acceler-
82
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems
ated oxidative stress, affect cell proliferation, provoke inflammatory reactions, and induce morphological alterations to the cellular structure [136–139]. These effects must not be underestimated, as they might result in more severe long-term toxic effects, although on the other hand they closely depend on the dimension of the nanotubes (aspect ratio/surface area), their agglomeration state and behavior and the type of functionalization [140–142]. In summary, most of the experimental results lack comparability and are even contradictory to each other, since different types of nanotubes with different amounts and kinds of impurities, different cell lines, different preparation procedures, and different detection methods were used. The case of systemic injection of carbon nanotubes for therapeutic purposes is an entirely different scenario. Here, the first aspect to consider is the interaction of carbon nanotubes with blood components (plasma proteins, blood cells, and platelets) and their recognition by the immune system. A study by Meng et al. in 2005 has investigated the effect of single-walled carbon nanotubes on plasma proteins and found that fibrinogen readily adsorbed to the nanotube surface, whereas albumin and platelets did not adhere [143]. These results were confirmed by Salvador-Morales and coworkers in a study in 2006, which explored protein adsorption and complement activation by purified, nonfunctionalized carbon nanotubes [144]. The binding of proteins to carbon nanotubes was shown to be highly selective, since out of many plasma proteins only few adsorbed to the nanotubes in a large quantity, including fibrinogen and apolipoproteins. The study further revealed that double-walled carbon nanotubes activate the human complement system via both the classical and the alternative pathway, whereas single-walled carbon nanotubes only activated the classical pathway. The activation of the complement system via the classical pathway usually leads to the generation of inflammatory peptides and might further induce the coating of the nanotubes by opsonins, the accumulation of neutrophils, and the adherence of phagocytic cells. In the case that the nanotubes are too large to be phagocytosed (see Figure 3.9), this might cause tissue damage and granuloma formation. However, the researchers hypothesize that the activation of the human complement system by carbon nanotubes might be diminished or eliminated by alteration of the surface chemistry via functionalization. One of the first systematic in vivo toxicity evaluations of functionalized singlewalled carbon nanotubes following IV injection has recently been carried out by Schipper et al. [145]. They examined the acute and chronic toxicity of functionalized single-walled carbon nanotubes when injected into the bloodstream of mice and found that single administrations of high doses did not lead to acute or chronic toxicity in nude mice, although the PEG-functionalized single-walled carbon nanotubes persisted in liver and spleen macrophages for 4 months. However, the researchers acknowledge that due to the small number of animals, the study must be considered as a pilot study, since small differences between the treatment groups could not be detected in this setup. A similar study carried out by Wang and coworkers tried to provide a general toxicological profile by focusing on the toxicity of single-walled carbon nanotubes to main organs [146]. The nanotubes remained in mice for 3 months after IV exposure and although serum and pulmonary inflammation were observed, no obvious cell apoptosis or changes of immunological indicators occurred. The authors propose the main toxicological mechanism to be oxidative stress in liver and lungs. An entirely different in vivo study investigating the impact
3.4 Challenges and Future Prospects
83
of carbon nanotubes on living organisms has been carried out by Leeuw and coworkers [147]. Drosophila (fruit fly) larvae were raised on food containing 10 ppm of disaggregated single-walled carbon nanotubes. Nanotubes were imaged in intact living larvae and dissected tissue specimen by near-infrared fluorescence spectroscopy. The study showed no short-term toxicity, impaired growth, or impaired fertility in mature individuals that have been fed dispersed carbon nanotubes. Long-term toxicity aspects like mutagenicity and carcinogenicity of systemically introduced carbon nanotubes have so far only been investigated in very preliminary studies, but will have to be assessed very carefully before taking carbon nanotube-based therapeutic systems to the stage of clinical trials. The third category of carbon nanotube toxicology deals with environmental aspects and is thus far the least investigated one. Considering that disposed carbon nanotube waste might lead to their ultimate release in the ground water, some studies have explored their impact on aquatic environments. For example, Mouchet and colleagues examined the ecotoxicological potential of carbon nanotubes in amphibian larvae at a wide range of concentrations in water by analyzing the toxicity and genotoxicity of double-walled carbon nanotubes in Xenopus laevis larvae after 12 days of static exposure under laboratory conditions [148]. The results showed no genotoxicity in erythrocytes of the larvae exposed to double-walled carbon nanotubes in water, but an acute toxicity at very high nanotube concentrations related to physical blockage of the gills and/or the digestive tract. In a similar study carried out by Smith and coworkers, rainbow trouts were exposed to single-walled carbon nanotubes dispersed in water with help of a surfactant for 10 days [149]. Nanotube exposure was shown to cause a dose-dependent rise in ventilation rate, gill pathology, and mucus secretion, with the apparition of carbon nanotube precipitates on the gill mucus. No major hematological or blood disturbances were observed; however, subtle cellular pathologies raised concern about cell cycle defects, neurotoxicity, and other yet unidentified factors that may mediate systemic pathologies. The last study that will be mentioned in the scope of ecotoxicity exposed zebra fish embryos to dispersed single- and double-walled carbon nanotubes and noticed a delay in hatching at higher nanotube concentrations [150]. The researchers assume that this effect was most likely induced by the cobalt and nickel catalyst impurities rather than the nanotubes. However, this nevertheless shows that raw, unpurified carbon nanotubes have the potential to affect aquatic life when released into the environment. The experimental evidence to date makes it difficult to draw a final conclusion about the safety of carbon nanotubes, especially about their use as therapeutic agents. Despite the amount of work already completed, many more studies are needed in order to obtain a clear understanding about the toxicity of carbon nanotubes. In particular, a standard material and standard methods for the characterization and functionalization of carbon nanotubes and for toxicity studies are needed in order to obtain comparable and undisputable results, especially when considering that some types of cytotoxicity assays may lead to false positive results due to adsorption of their reaction product to carbon nanotubes [151]. The last words in this matter shall be given to Professor Kostarelos from the Nanomedicine Lab at the Centre for Drug Delivery Research of the University of London, who believes that “if the unique clinical potential of carbon nanotubes is to be exploited,
84
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems
toxicological studies and pharmacological development must continue in parallel, before eventually converging to provide a clear framework acceptable to regulatory authorities and the public” [130]. 3.4.2
In Vivo Biodistribution of Carbon Nanotubes
For safe biomedical applications of carbon nanotubes in the human body, in particular drug delivery, it is very important to investigate pharmacokinetic aspects, starting from the administration of the nanovector-drug complexes to their distribution via the blood circulation, reaching of the target tissue, metabolism in the liver, and elimination via the kidneys or the biliary pathway. The next paragraph will present in vivo studies that have been carried out with regard to biodistribution of carbon nanotubes in the last years. The first biodistribution analysis was performed in 2004 by Wang and coworkers, who administered radiolabeled (125I), hydroxylated carbon nanotubes via intraperitoneal injection into mice [152]. Despite their high molecular weight, the nanotubes behaved similarly to small molecules and moved easily among the compartments and tissues of the body, accumulating in the stomach, kidney, and bone. Singh and colleagues investigated tissue biodistribution and blood clearance rates of radiolabeled, single-walled carbon nanotubes in mice by gamma scintigraphy after IV administration [153]. Their study showed that the nanotubes had a blood circulation half-life of about 3 hours, were not retained in any of the reticuloendothelial system organs like liver or spleen and were rapidly cleared via the renal excretion route. In a follow-up study, the researchers further investigated the shape and aggregation state of carbon nanotubes in relation to renal excretion by TEM imaging of ultrathin renal cortex sections [154]. Aggregated carbon nanotubes that were not able to translocate through the kidney filtration system were found in the glomerular capillaries. This indicates that dimension, shape, and structural characteristics are extremely important parameters for pharmacological and excretion profiles of carbon nanotubes. Furthermore, renal excretion seems to be the preferred mechanism of elimination as long as adequate individualization of the nanotubes can be achieved and maintained in vivo. Another study by McDevitt et al. that has already been discussed earlier in connection with specific tumor targeting by a soluble, carbon nanotube-antibody construct, also investigated the biodistribution behavior of their radiolabeled nanoconstructs [104]. The unspecific complexes were rapidly cleared from the blood compartment (~ 1 hour) and, similarly to the results of Wang et al., were observed to accumulate in the kidney, spleen, liver, and to a lesser extent, in the bone. However, biodistribution and pharmacokinetics were significantly altered by the covalent attachment of antibody molecules. Two further, very important and comprehensive studies were carried out by researchers at Stanford University in California. The first one in 2007 investigated the biodistribution of radiolabeled, single-walled carbon nanotubes in mice by in vivo positron emission tomography (PET), ex vivo biodistribution, and Raman spectroscopy [155]. The nanotubes were noncovalently functionalized with two different types of phospholipid-PEGs varying in the length of the PEG chain (2 kDa versus 5.4 kDa). An increased length of the PEG chain decreased the uptake of the nanotubes by the reticuloendothelial system and prolonged blood circulation times:
3.5 Conclusion
85
single-walled carbon nanotubes functionalized with 2 kDa PEG circulated in the blood for 0.5 hour, whereas the same nanotubes functionalized with 5.4 kDa PEG circulated in the blood for 2 hours. This study was also the first that took advantage of the intrinsic optical properties of carbon nanotubes by directly detecting them in various tissues of a mouse by Raman spectroscopy. Nanotubes were found in the tumor and liver sample, and to a slight extent in the kidney sample. The results were in reasonable agreement with the PET data based on radioactivity. The next study, published in the same year, used this new technique to investigate the long-term fate of carbon nanotubes in a mice model over 3 months [70]. The functionalization technique was slightly changed from using chain-like PEG conjugated to phospholipids to using branched PEG chains with a higher molecular weight. This enabled blood circulation of the nanotubes for up to 24 hours, relatively low uptake by the reticuloendothelial system, and near-complete clearance from the main organs in 2 months. Raman spectroscopy measurements at different time points after exposure detected nanotubes in the intestine, the feces, the kidney, and the bladder of mice, suggesting excretion and clearance of carbon nanotubes via both the renal and the biliary pathway. No toxic side effects to mice were observed in necropsy, histology, and blood chemistry measurements, which shows the major advantage of this functionalization scheme, since pristine carbon nanotubes have been shown to remain in main organs at relatively high accumulation levels over 28 days [156].
3.5
Conclusion This chapter has demonstrated that the ability of carbon nanotubes to penetrate cell membranes, carry multiple therapeutic agents as cargos, and selectively target tumor cells renders them promising candidates as novel, customizable, targeted drug delivery systems, which can increase the effectiveness and specificity of present-day anticancer therapies. Biodistribution studies have shown that a suitable functionalization scheme can increase the biocompatibility of carbon nanotubes, lower their uptake by the reticuloendothelial system and determine the excretion route and the blood circulation time. These promising results are a milestone for the development of carbon nanotubes as clinically successful therapeutic agents, although pharmacological and toxicological studies should always be run in parallel to obtain a clear picture of both the beneficial and the potential adverse health effects of this unique material.
Problems 3.1 List three characterization techniques for carbon nanotubes, the information they can provide about the sample, and their drawbacks. 3.2 Which different kinds of impurities can you find in commercially available samples and how can you eliminate them? 3.3 What are the major technical barriers for the use of carbon nanotubes in the biomedical field? Describe the different methods of functionalization
86
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems
3.4
3.5
3.6 3.7
in order to overcome these barriers, starting from the raw, as-produced material. The concentration of a carbon nanotube solution can be determined by UV/vis absorption spectroscopy. Assume that a sample consists of well-dispersed and entirely water-soluble carbon nanotubes and has a mass extinction coefficient of 13.5 L/g cm at a wavelength of 730 nm. Using Beer-Lamberts law, calculate the concentration of a carbon nanotube solution in mg/mL, with a measured absorbance value of 0.82 at a wavelength of 730 nm. The path length of the used cuvette is 1 cm. Anticancer drugs can be classified according to their mechanism of action. List the three main classes of anticancer drugs, including examples, and explain their mechanism of action. What are targeted cancer therapies and what impact will they have on cancer treatment? Elaborate on possible negative impacts of the widespread use of carbon nanotubes.
Acknowledgments The authors gratefully acknowledge funding and support from the EU FP6 Marie Curie Research Training Network CARBIO (multifunctional carbon nanotubes for biomedical applications). Furthermore, the authors thank Dr. Helen Coley for her invaluable advice and input.
References [1] “American Cancer Society Report Sees 7.6 Million Global 2007 Cancer Deaths,” Reuters, December 17, 2007. [2] Jemal, A., Siegel, R., Ward, E., et al., “Cancer Statistics 2008,” CA Cancer J Clin, Vol. 58, No. 2, 2008, pp. 71–96. [3] Qi, L. F., and Gao, X. H., “Emerging Application of Quantum Dots for Drug Delivery and Therapy,” Expert Opinion on Drug Delivery, Vol. 5, No. 3, 2008, pp. 263–267. [4] Okuda, T., Kawakami, S., and Akimoto, N., et al., “PEGylated Lysine Dendrimers for Tumor-Selective Targeting After Intravenous Injection in Tumor-Bearing Mice,” Journal of Controlled Release, Vol. 116, No. 3, 2006, pp. 330–336. [5] Panyam, J., and Labhasetwar, V., “Biodegradable Nanoparticles for Drug and Gene Delivery to Cells and Tissue,” Advanced Drug Delivery Reviews, Vol. 55, No. 3, 2003, pp. 329–347. [6] Park, J. W., “Liposome-Based Drug Delivery in Breast Cancer Treatment,” Breast Cancer Research, Vol. 4, No. 3, 2002, pp. 93–97. [7] Soverini, S., Martinelli, G., and Lacobucci, I., et al., “Imatinib Mesylate for The Treatment of Chronic Myeloid Leukemia,” Expert Review of Anticancer Therapy, Vol. 8, No. 6, 2008, pp. 853–864. [8] Niculescu-Duvaz, I., and Springer, C. J., “Antibody-Directed Enzyme Prodrug Therapy (ADEPT): A Review,” Advanced Drug Delivery Reviews, Vol. 26, No. 2-3, 1997, pp. 151–172.
References
87
[9] Ferrari, M., “Cancer Nanotechnology: Opportunities and Challenges,” Nature Reviews Cancer, Vol. 5, No. 3, 2005, pp. 161–171. [10] Hamblett, K. J., Senter, P. D., and Chace, D. F., et al., “Effects of Drug Loading on the Antitumor Activity of a Monoclonal Antibody Drug Conjugate,” Clinical Cancer Research, Vol. 10, No. 20, 2004, pp. 7063–7070. [11] Iyer, A. K., Khaled, G., Fang, J., et al., “Exploiting the Enhanced Permeability and Retention Effect for Tumor Targeting,” Drug Discovery Today, Vol. 11, No. 17–18, 2006, pp. 812–818. [12] Gottlieb, J. J., Washenik, K., Chachoua, A., et al., “Treatment of Classic Kaposis Sarcoma with Liposomal Encapsulated Doxorubicin,” Lancet, Vol. 350, No. 9088, 1997, pp. 1363–1364. [13] Muggia, F. M., Hainsworth, J. D., Jeffers, S., et al., “Phase II study of liposomal doxorubicin in refractory ovarian cancer: Antitumor Activity and Toxicity Modification by Liposomal Encapsulation,” Journal of Clinical Oncology, Vol. 15, No. 3, 1997, pp. 987–993. [14] Chester, K. A., Mayer, A., Bhatia, J., et al., “Recombinant Anti-Carcinoembryonic Antigen Antibodies for Targeting Cancer,” Cancer Chemotherapy and Pharmacology, Vol. 46, No. 2000, pp. S8–S12. [15] Xu, X. C., Clarke, P., Szalai, G., et al., “Targeting and Therapy of Carcinoembryonic Antigen-Expressing Tumors in Transgenic Mice with an Antibody-Interleukin 2 Fusion Protein,” Cancer Research, Vol. 60, No. 16, 2000, pp. 4475–4484. [16] Meyer, A., Auernheimer, J., Modlinger, A., et al., “Targeting RGD Recognizing Integrins: Drug Development, Biomaterial Research, Tumor Imaging and Targeting,” Curr Pharm Des, Vol. 12, No. 22, 2006, pp. 2723–2747. [17] Garanger, E., Boturyn, D., and Dumy, P., “Tumor Targeting with RGD Peptide Ligands-Design of New Molecular Conjugates for Imaging and Therapy of Cancers,” Anticancer Agents Med Chem, Vol. 7, No. 5, 2007, pp. 552–558. [18] Lu, Y., Sega, E., Leamon, C. P., et al., “Folate Receptor-Targeted Immunotherapy of Cancer: Mechanism and Therapeutic Potential,” Adv Drug Deliv Rev, Vol. 56, No. 8, 2004, pp. 1161–1176. [19] Russell-Jones, G., McTavish, K., McEwan, J., et al., “Vitamin-Mediated Targeting as a Potential Mechanism to Increase Drug Uptake by Tumours,” Journal of Inorganic Biochemistry, Vol. 98, No. 10, 2004, pp. 1625–1633. [20] Zhao, X., Li, H., and Lee, R. J., “Targeted Drug Delivery Via Folate Receptors,” Expert Opin Drug Deliv, Vol. 5, No. 3, 2008, pp. 309–319. [21] Liu, Y. F., and Wang, H. F., “Nanomedicine—Nanotechnology Tackles Tumours,” Nature Nanotechnology, Vol. 2, No. 1, 2007, pp. 20–21. [22] Radushkevich, L. V., and Lukyanovich, V. M., “O strukture ugleroda, obrazujucegosja pri termiceskom razlozenii okisi ugleroda na zeleznom kontakte,” Zurn Fis. Chim., Vol. 26, No. 1952, pp. 88–95. [23] Monthioux, M., and Kuznetsov, V. L., “Who Should Be Given the Credit for the Discovery of Carbon Nanotubes?” Carbon, Vol. 44, No. 9, 2006, pp. 1621–1623. [24] Iijima, S., “Helical Microtubules of Graphitic Carbon,” Nature, Vol. 354, No. 6348, 1991, pp. 56–58. [25] Odom, T. W., Huang, J. L., Kim, P., et al., “Structure and Electronic Properties of Carbon Nanotubes,” Journal of Physical Chemistry B, Vol. 104, No. 13, 2000, pp. 2794–2809. [26] Louie, S. G., “Electronic Properties, Junctions, and Defects of Carbon Nanotubes,” in Carbon Nanotubes: Synthesis, Structure, Properties and Applications, Dresselhaus, M. S., Dresselhaus, G., and Avouris, P. (eds.), Berlin/Heidelberg: Springer, 2001, pp. 113–145. [27] Iijima, S., and Ichihashi, T., “Single-Shell Carbon Nanotubes of 1-Nm Diameter,” Nature, Vol. 363, No. 6430, 1993, pp. 603–605. [28] Dresselhaus, M. S., Dresselhaus, G., and Avouris, P., Carbon Nanotubes: Synthesis, Structure, Properties and Applications, Berlin/Heidelberg: Springer, 2001.
88
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems [29] OConnell, M. J., Carbon Nanotubes: Properties and Applications, Boca Raton: CRC Press, Taylor & Francis Group, 2006. [30] Yakobson, B. I., and Avouris, P., “Mechanical Properties of Carbon Nanotubes,” in Carbon Nanotubes: Synthesis, Structure, Properties and Applications, Dresselhaus, M. S., Dresselhaus, G., and Avouris, P. (eds.), Berlin/Heidelberg: Springer, 2001, pp. 287–327, [31] Lee, N. S., Chung, D. S., Han, I. T., et al., “Application of Carbon Nanotubes to Field Emission Displays,” Diamond and Related Materials, Vol. 10, No. 2, 2001, pp. 265–270. [32] Carey, J. D., and Silva, S. R. P., “Nanostructured Materials for Field Emission Devices,” in Nanomaterials Handbook, Gogotsi, Y. (ed.), Boca Raton: CRC Press, 2006, pp. 665–684, [33] Feldman, A. K., Steigerwald, M. L., Guo, X., et al., “Molecular Electronic Devices Based on Single-Walled Carbon Nanotube Electrodes,” Accounts of Chemical Research, Vol. 41, No. 12, 2008, pp. 1731–1741. [34] Lyshevski, S. E., Nano and Molecular Electronics Handbook, Boca Raton: CRC Press, 2007. [35] Li, C., Chen, Y. H., Wang, Y. B., et al., “A Fullerene-Single Wall Carbon Nanotube Complex for Polymer Bulk Heterojunction Photovoltaic Cells,” Journal of Materials Chemistry, Vol. 17, No. 23, 2007, pp. 2406–2411. [36] Pushparaj, V. L., Shaijumon, M. M., Kumar, A., et al., “Flexible Energy Storage Devices Based on Nanocomposite Paper,” Proceedings of the National Academy of Sciences of the United States of America, Vol. 104, No. 34, 2007, pp. 13574–13577. [37] Dai, H. J., Hafner, J. H., Rinzler, A. G., et al., “Nanotubes as Nanoprobes in Scanning Probe Microscopy,” Nature, Vol. 384, No. 6605, 1996, pp. 147–150. [38] Schlaf, R., Emirov, Y., Bieber, J. A., et al., “Using Carbon Nanotube Cantilevers in Scanning Probe Metrology,” Metrology, Inspection, and Process Control for Microlithography XVI, Pts 1 & 2, Vol. 4689, No. 2002, pp. 53–57. [39] Roy, S., Vedala, H., and Choi, W., “Vertically Aligned Carbon Nanotube Probes for Monitoring Blood Cholesterol,” Nanotechnology, Vol. 17, No. 4, 2006, pp. S14–S18. [40] Star, A., Joshi, V., Skarupo, S., et al., “Gas Sensor Array Based on Metal-Decorated Carbon Nanotubes,” Journal of Physical Chemistry B, Vol. 110, No. 42, 2006, pp. 21014–21020. [41] Sethuraman, A., Stroscio, M., and Dutta, M., “Potential Applications of Carbon Nanotubes in Bioengineering,” in Biological Nanostructures and Applications of Nanostructures in Biology, Stroscio, M., and Dutta, M. (eds.), New York: Kluwer Academic Publishers, 2004, pp. 51–68. [42] Bianco, A., Kostarelos, K., Partidos, C. D., et al., “Biomedical Applications of Functionalised Carbon Nanotubes,” Chemical Communications, Vol. 5, 2005, pp. 571–577. [43] Yang, W. R., Thordarson, P., Gooding, J. J., et al., “Carbon Nanotubes for Biological and Biomedical Applications,” Nanotechnology, Vol. 18, No. 41, 2007, pp. 412001. [44] Harrison, B. S., and Atala, A., “Carbon Nanotube Applications for Tissue Engineering,” Biomaterials, Vol. 28, No. 2, 2007, pp. 344–353. [45] Ebbesen, T. W., and Ajayan, P. M., “Large-Scale Synthesis of Carbon Nanotubes,” Nature, Vol. 358, No. 6383, 1992, pp. 220–222. [46] Mann, D., “Synthesis of Carbon Nanotubes,” in Carbon Nanotubes: Properties and Applications, O’Connell, M. J. (ed.), Boca Raton: CRC Press, 2006, pp. 20–49. [47] Shanov, V., and Yun, Y. H., “Synthesis and Characterization of Carbon Nanotube Materials,” Journal of the University of Chemical Technology and Metallurgy, Vol. 41, No. 4, 2006, pp. 377–390. [48] McGuire, K., and Rao, A. M., “Characterization Techniques in Carbon Nanotube Research,” in Carbon Nanotubes: Science and Applications, Meyyappan, M., (ed.) pp. 117–136, Boca Raton: CRC Press, 2005. [49] Sato, Y., Yanagi, K., Miyata, Y., et al., “Chiral-Angle Distribution for Separated Single-Walled Carbon Nanotubes,” Nano Letters, Vol. 8, No. 10, 2008, pp. 3151–3154.
References
89
[50] Itkis, M. E., Perea, D. E., Jung, R., et al., “Comparison of Analytical Techniques for Purity Evaluation of Single-Walled Carbon Nanotubes,” Journal of the American Chemical Society, Vol. 127, No. 10, 2005, pp. 3439–3448. [51] Chiang, I. W., Brinson, B. E., Smalley, R. E., et al., “Purification and Characterization of Single-Wall Carbon Nanotubes,” Journal of Physical Chemistry B, Vol. 105, No. 6, 2001, pp. 1157–1161. [52] Kataura, H., Kumazawa, Y., Maniwa, Y., et al., “Optical Properties of Single-Wall Carbon Nanotubes,” Synthetic Metals, Vol. 103, No. 1–3, 1999, pp. 2555–2558. [53] Itkis, M. E., Perea, D. E., Niyogi, S., et al., “Purity Evaluation of As-Prepared Single-Walled Carbon Nanotube Soot by Use Of Solution-Phase Near-IR Spectroscopy,” Nano Letters, Vol. 3, No. 3, 2003, pp. 309–314. [54] Monthioux, M., Smith, B. W., Burteaux, B., et al., “Sensitivity of Single-Wall Carbon Nanotubes to Chemical Processing: An Electron Microscopy Investigation,” Carbon, Vol. 39, No. 8, 2001, pp. 1251–1272. [55] Jorio, A., Pimenta, M. A., Souza, A. G., et al., “Characterizing Carbon Nanotube Samples with Resonance Raman Scattering,” New Journal of Physics, Vol. 5, No. 2003, pp. 1–17. [56] Eklund, P. C., Holden, J. M., and Jishi, R. A., “Vibrational-Modes of Carbon Nanotubes—Spectroscopy and Theory,” Carbon, Vol. 33, No. 7, 1995, pp. 959–972. [57] Dillon, A. C., Yudasaka, M., and Dresselhaus, M. S., “Employing Raman Spectroscopy to Qualitatively Evaluate the Purity of Carbon Single-Wall Nanotube Materials,” J Nanosci Nanotechnol, Vol. 4, No. 7, 2004, pp. 691–703. [58] Pang, L. S. K., Saxby, J. D., and Chatfield, S. P., “Thermogravimetric Analysis of Carbon Nanotubes and Nanoparticles,” Journal of Physical Chemistry, Vol. 97, No. 27, 1993, pp. 6941–6942. [59] Hou, P. X., Liu, C., and Cheng, H. M., “Purification of Carbon Nanotubes,” Carbon, Vol. 46, No. 15, 2008, pp. 2003–2025. [60] Ebbesen, T. W., Ajayan, P. M., Hiura, H., et al., “Purification of Nanotubes,” Nature, Vol. 367, No. 6463, 1994, pp. 519. [61] Rinzler, A. G., Liu, J., Dai, H., et al., “Large-Scale Purification of Single-Wall Carbon Nanotubes: Process, Product, and Characterization,” Applied Physics a-Materials Science & Processing, Vol. 67, No. 1, 1998, pp. 29–37. [62] Tobias, G., Shao, L. D., Salzmann, C. G., et al., “Purification and Opening of Carbon Nanotubes Using Steam,” Journal of Physical Chemistry B, Vol. 110, No. 45, 2006, pp. 22318–22322. [63] Shelimov, K. B., Esenaliev, R. O., Rinzler, A. G., et al., “Purification of Single-Wall Carbon Nanotubes by Ultrasonically Assisted Filtration,” Chemical Physics Letters, Vol. 282, No. 5–6, 1998, pp. 429–434. [64] Harutyunyan, A. R., Pradhan, B. K., Chang, J. P., et al., “Purification of Single-Wall Carbon Nanotubes by Selective Microwave Heating of Catalyst Particles,” Journal of Physical Chemistry B, Vol. 106, No. 34, 2002, pp. 8671–8675. [65] Zhang, J., Zou, H. L., Qing, Q., et al., “Effect of Chemical Oxidation on the Structure of Single-Walled Carbon Nanotubes,” Journal of Physical Chemistry B, Vol. 107, No. 16, 2003, pp. 3712–3718. [66] Rosca, I. D., Watari, F., Uo, M., et al., “Oxidation of Multiwalled Carbon Nanotubes by Nitric Acid,” Carbon, Vol. 43, No. 15, 2005, pp. 3124–3131. [67] Vivekchand, S. R. C., Jayakanth, R., Govindaraj, A., et al., “The Problem of Purifying Single-Walled Carbon Nanotubes,” Small, Vol. 1, No. 10, 2005, pp. 920–923. [68] Geng, H. Z., Zhang, X. B., Mao, S. H., et al., “Opening and Closing of Single-Wall Carbon Nanotubes,” Chemical Physics Letters, Vol. 399, No. 1–3, 2004, pp. 109–113. [69] Jeynes, J. C. G., Mendoza, E., Chow, D. C. S., et al., “Generation of Chemically Unmodified Pure Single-Walled Carbon Nanotubes by Solubilizing with RNA and Treatment with Ribonuclease A,” Advanced Materials, Vol. 18, No. 12, 2006, pp. 1598–1602.
90
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems [70] Liu, Z., Davis, C., Cai, W. B., et al., “Circulation and Long-Term Fate of Functionalized, Biocompatible Single-Walled Carbon Nanotubes in Mice Probed by Raman Spectroscopy,” Proceedings of the National Academy of Sciences of the United States of America, Vol. 105, No. 5, 2008, pp. 1410–1415. [71] Haggenmueller, R., Rahatekar, S. S., Fagan, J. A., et al., “Comparison of the Quality of Aqueous Dispersions of Single Wall Carbon Nanotubes Using Surfactants and Biomolecules,” Langmuir, Vol. 24, No. 9, 2008, pp. 5070–5078. [72] Fu, K. F., and Sun, Y. P., “Dispersion And Solubilization of Carbon Nanotubes,” Journal of Nanoscience and Nanotechnology, Vol. 3, No. 5, 2003, pp. 351–364. [73] OConnell, M. J., Boul, P., Ericson, L. M., et al., “Reversible Water-Solubilization of Single-Walled Carbon Nanotubes by Polymer Wrapping,” Chemical Physics Letters, Vol. 342, No. 3–4, 2001, pp. 265–271. [74] Islam, M. F., Rojas, E., Bergey, D. M., et al., “High Weight Fraction Surfactant Solubilization of Single-Wall Carbon Nanotubes in Water,” Nano Letters, Vol. 3, No. 2, 2003, pp. 269–273. [75] Zorbas, V., Ortiz-Acevedo, A., Dalton, A. B., et al., “Preparation and Characterization of Individual Peptide-Wrapped Single-Walled Carbon Nanotubes,” Journal of the American Chemical Society, Vol. 126, No. 23, 2004, pp. 7222–7227. [76] Lin, Y., Allard, L. F., and Sun, Y. P., “Protein-Affinity of Single-Walled Carbon Nanotubes in Water,” Journal of Physical Chemistry B, Vol. 108, No. 12, 2004, pp. 3760–3764. [77] Wenseleers, W., Vlasov, I. I., Goovaerts, E., et al., “Efficient Isolation and Solubilization of Pristine Single-Walled Nanotubes in Bile Salt Micelles,” Advanced Functional Materials, Vol. 14, No. 11, 2004, pp. 1105–1112. [78] Zheng, M., Jagota, A., Semke, E. D., et al., “DNA-Assisted Dispersion and Separation of Carbon Nanotubes,” Nature Materials, Vol. 2, No. 5, 2003, pp. 338–342. [79] Nakashima, N., Okuzono, S., Murakami, H., et al., “DNA Dissolves Single-Walled Carbon Nanotubes in Water,” Chemistry Letters, Vol. 32, No. 5, 2003, pp. 456–457. [80] Bahr, J. L., and Tour, J. M., “Covalent Chemistry of Single-Wall Carbon Nanotubes, Journal of Materials Chemistry, Vol. 12, No. 7, 2002, pp. 1952–1958. [81] Hirsch, A., “Functionalization of Single-Walled Carbon Nanotubes,” Angewandte Chemie-International Edition, Vol. 41, No. 11, 2002, pp. 1853–1859. [82] Nagasawa, S., Yudasaka, M., Hirahara, K., et al., “Effect of Oxidation on Single-Wall Carbon Nanotubes,” Chemical Physics Letters, Vol. 328, No. 4–6, 2000, pp. 374–380. [83] Zhang, Y., Shi, Z., Gu, Z., et al., “Structure Modification of Single-Wall Carbon Nanotubes,” Carbon, Vol. 38, No. 15, 2000, pp. 2055–2059. [84] Li, Y., Zhang, X. B., Luo, J. H., et al., “Purification of CVD Synthesized Single-Wall Carbon Nanotubes by Different Acid Oxidation Treatments,” Nanotechnology, Vol. 15, No. 11, 2004, pp. 1645–1649. [85] Staros, J. V., Wright, R. W., and Swingle, D. M., “Enhancement by N-Hydroxysulfosuccinimide of Water-Soluble Carbodiimide-Mediated Coupling Reactions,” Analytical Biochemistry, Vol. 156, No. 1, 1986, pp. 220–222. [86] Huang, W. J., Taylor, S., Fu, K. F., et al., “Attaching Proteins to Carbon Nanotubes Via Diimide-Activated Amidation,” Nano Letters, Vol. 2, No. 4, 2002, pp. 311–314. [87] Gromov, A., Dittmer, S., Svensson, J., et al., “Covalent Amino-Functionalisation of Single-Wall Carbon Nanotubes,” Journal of Materials Chemistry, Vol. 15, No. 32, 2005, pp. 3334–3339. [88] DEste, M., De Nardi, M., and Menna, E., “A Co-Functionalization Approach to Soluble And Functional Single-Walled Carbon Nanotubes,” European Journal of Organic Chemistry, Vol. 11, 2006, pp. 2517–2522. [89] Tasis, D., Tagmatarchis, N., Bianco, A., et al., “Chemistry of Carbon Nanotubes,” Chemical Reviews, Vol. 106, No. 3, 2006, pp. 1105–1136.
References
91
[90] Georgakilas, V., Kordatos, K., Prato, M., et al., “Organic Functionalization of Carbon Nanotubes,” Journal of the American Chemical Society, Vol. 124, No. 5, 2002, pp. 760–761. [91] Prato, M., Kostarelos, K., and Bianco, A., “Functionalized Carbon Nanotubes in Drug Design and Discovery,” Accounts of Chemical Research, Vol. 41, No. 1, 2008, pp. 60–68. [92] Pastorin, G., Wu, W., Wieckowski, S., et al., “Double Functionalisation of Carbon Nanotubes for Multimodal Drug Delivery,” Chemical Communications, Vol. 11, 2006, pp. 1182–1184. [93] Shim, M., Kam, N. W. S., Chen, R. J., et al., “Functionalization of Carbon Nanotubes for Biocompatibility and Biomolecular Recognition,” Nano Letters, Vol. 2, No. 4, 2002, pp. 285–288. [94] He, P., and Dai, L., “Carbon Nanotube Biosensors,” in Biological and Biomedical Nanotechnology, Ferrari, M., Lee, A. P., and Lee, L. J. (eds.), Springer, 2006, pp. 171–202. [95] Pantarotto, D., Briand, J. P., Prato, M., et al., “Translocation of Bioactive Peptides Across Cell Membranes by Carbon Nanotubes,” Chemical Communications, Vol. 1, 2004, pp. 16–17. [96] Kam, N. W. S., Liu, Z. A., and Dai, H. J., “Carbon Nanotubes as Intracellular Transporters for Proteins and DNA: An Investigation of the Uptake Mechanism and Pathway,” Angewandte Chemie-International Edition, Vol. 45, No. 4, 2006, pp. 577–581. [97] Kam, N. W. S., Jessop, T. C., Wender, P. A., et al., “Nanotube Molecular Transporters: Internalization of Carbon Nanotube-Protein Conjugates into Mammalian Cells,” Journal of the American Chemical Society, Vol. 126, No. 22, 2004, pp. 6850–6851. [98] Kam, N. W. S., and Dai, H. J., “Single Walled Carbon Nanotubes for Transport and Delivery of Biological Cargos,” Physica Status Solidi B-Basic Solid State Physics, Vol. 243, No. 13, 2006, pp. 3561–3566. [99] Gao, L. Z., Nie, L., Wang, T. H., et al., “Carbon Nanotube Delivery of the GFP Gene into Mammalian Cells,” Chembiochem, Vol. 7, No. 2, 2006, pp. 239–242. [100] Wu, W., Wieckowski, S., Pastorin, G., et al., “Targeted Delivery of Amphotericin B to Cells by Using Functionalized Carbon Nanotubes,” Angewandte Chemie-International Edition, Vol. 44, No. 39, 2005, pp. 6358–6362. [101] Ali-Boucetta, H., Al-Jamal, K. T., McCarthy, D., et al., Multiwalled Carbon Nanotube-Doxorubicin Supramolecular Complexes for Cancer Therapeutics, Chemical Communications, Vol. 4, 2008, pp. 459–461. [102] Feazell, R. P., Nakayama-Ratchford, N., Dai, H., et al., “Soluble Single-Walled Carbon Nanotubes as Longboat Delivery Systems for Platinum(IV) Anticancer Drug Design,” Journal of the American Chemical Society, Vol. 129, No. 27, 2007, pp. 8438–8439. [103] Liu, Z., Sun, X. M., Nakayama-Ratchford, N., et al., “Supramolecular Chemistry on Water-Soluble Carbon Nanotubes for Drug Loading and Delivery,” ACS Nano, Vol. 1, No. 1, 2007, pp. 50–56. [104] McDevitt, M. R., Chattopadhyay, D., Kappel, B. J., et al., “Tumor Targeting with Antibody-Functionalized, Radiolabeled Carbon Nanotubes,” Journal of Nuclear Medicine, Vol. 48, No. 7, 2007, pp. 1180–1189. [105] Chen, J., Chen, S., Zhao, X., et al., “Functionalized Single-Walled Carbon Nanotubes as Rationally Designed Vehicles for Tumor-Targeted Drug Delivery,” Journal of the American Chemical Society, Vol. 130, No. 49, 2008, pp. 16778–16785. [106] Dhar, S., Liu, Z., Thomale, J., et al., “Targeted Single-Wall Carbon Nanotube-Mediated Pt(IV) Prodrug Delivery Using Folate as a Homing Device,” Journal of the American Chemical Society, Vol. 130, No. 34, 2008, pp. 11467–11476. [107] Liu, Z., Chen, K., Davis, C., et al., “Drug Delivery with Carbon Nanotubes for In Vivo Cancer Treatment,” Cancer Research, Vol. 68, No. 16, 2008, pp. 6652–6660. [108] Villa, C. H., McDevitt, M. R., Escorcia, F. E., et al., “Synthesis and Biodistribution of Oligonucleotide-Functionalized, Tumor-Targetable Carbon Nanotubes,” Nano Letters, Vol. 8, No. 12, 2008, pp. 4221–4228.
92
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems [109] Bhirde, A. A., Patel, V., Gavard, J., et al., “Targeted Killing of Cancer Cells In Vivo and In Vitro with EGF-Directed Carbon Nanotube-Based Drug Delivery,” ACS Nano, Vol. 3, No. 2, 2009, pp. 307–316. [110] Heister, E., Neves, V., Tîlmaciu, C., et al., “Triple Functionalisation of Single-Walled Carbon Nanotubes with Doxorubicin, a Monoclonal Antibody, and a Fluorescent Marker for Targeted Cancer Therapy,” Carbon, Vol. 47, No. 9, 2009, pp. 2152–2160. [111] Gallis, H. A., Drew, R. H., and Pickard, W. W., “Amphotericin B: 30 Years of Clinical Experience,” Rev Infect Dis, Vol. 12, No. 2, 1990, pp. 308–329. [112] Kessel, D., Hall, T. C., and Roberts, D., “Modes of Uptake of Methotrexate by Normal and Leukemic Human Leukocytes In Vitro and Their Relation to Drug Response,” Cancer Res, Vol. 28, No. 3, 1968, pp. 564–570. [113] Kam, N. W. S., Liu, Z., and Dai, H. J., “Functionalization of Carbon Nanotubes Via Cleavable Disulfide Bonds for Efficient Intracellular Delivery of siRNA and Potent Gene Silencing,” Journal of the American Chemical Society, Vol. 127, No. 36, 2005, pp. 12492–12493. [114] Scott, S. D., “Rituximab: A New Therapeutic Monoclonal Antibody for Non-Hodgkins Lymphoma,” Cancer Pract, Vol. 6, No. 3, 1998, pp. 195–197. [115] Russell-Jones, G., and McEwan, J., “Amplification of Biotin-Mediated Targeting,” US Patent Office, 2004, Pub. No. WO/2004/045647. [116] Ojima, I., Bounaud, P. Y., Takeuchi, C., et al., “New Taxanes as Highly Efficient Reversal Agents for Multi-Drug Resistance in Cancer Cells,” Bioorganic & Medicinal Chemistry Letters, Vol. 8, No. 2, 1998, pp. 189–194. [117] Ugarte, D., Chatelain, A., and deHeer, W. A., “Nanocapillarity and Chemistry in Carbon Nanotubes,” Science, Vol. 274, No. 5294, 1996, pp. 1897–1899. [118] Dujardin, E., Ebbesen, T. W., Hiura, H., et al., “Capillarity and Wetting of Carbon Nanotubes,” Science, Vol. 265, No. 5180, 1994, pp. 1850–1852. [119] Yudasaka, M., Ajima, K., Suenaga, K., et al., “Nano-Extraction and Nano-Condensation for C-60 Incorporation into Single-Wall Carbon Nanotubes in Liquid Phases,” Chemical Physics Letters, Vol. 380, No. 1–2, 2003, pp. 42–46. [120] Leonhardt, A., Hampel, S., Muller, C., et al., “Synthesis, Properties, and Applications of Ferromagnetic-Filled Carbon Nanotubes,” Chemical Vapor Deposition, Vol. 12, No. 6, 2006, pp. 380–387. [121] Gao, H. J., Kong, Y., Cui, D. X., et al., “Spontaneous Insertion of DNA Oligonucleotides into Carbon Nanotubes,” Nano Letters, Vol. 3, No. 4, 2003, pp. 471–473. [122] Kim, B. M., Qian, S., and Bau, H. H., “Filling Carbon Nanotubes with Particles,” Nano Letters, Vol. 5, No. 5, 2005, pp. 873–878. [123] Hilder, T. A., and Hill, J. M., “Carbon Nanotubes as Drug Delivery Nanocapsules,” Current Applied Physics, Vol. 8, No. 3–4, 2008, pp. 258–261. [124] Hilder, T. A., and Hill, J. M., “Probability of Encapsulation of Paclitaxel and Doxorubicin into Carbon Nanotubes,” Micro & Nano Letters, Vol. 3, No. 2, 2008, pp. 41–49. [125] Hampel, S., Kunze, D., Haase, D., et al., “Carbon Nanotubes Filled with a Chemotherapeutic Agent: A Nanocarrier Mediates Inhibition of Tumor Cell Growth,” Nanomed, Vol. 3, No. 2, 2008, pp. 175–182. [126] Ajima, K., Murakami, T., Mizoguchi, Y., et al., “Enhancement of In Vivo Anticancer Effects of Cisplatin by Incorporation Inside Single-Wall Carbon Nanohorns,” ACS Nano, Vol. 2, No. 10, 2008, pp. 2057–2064. [127] Murakami, T., Sawada, H., Tamura, G., et al., “Water-Dispersed Single-Wall Carbon Nanohorns as Drug Carriers for Local Cancer Chemotherapy,” Nanomed, Vol. 3, No. 4, 2008, pp. 453–463. [128] Poland, C. A., Duffin, R., Kinloch, I., et al., “Carbon Nanotubes Introduced into the Abdominal Cavity of Mice Show Asbestos-Like Pathogenicity in a Pilot Study,” Nature Nanotechnology, Vol. 3, No. 7, 2008, pp. 423–428.
References
93
[129] Donaldson, K., and Tran, C. L., “An Introduction to the Short-Term Toxicology of Respirable Industrial Fibres,” Mutation Research-Fundamental and Molecular Mechanisms of Mutagenesis, Vol. 553, No. 1–2, 2004, pp. 5–9. [130] Kostarelos, K., “The Long and Short of Carbon Nanotube Toxicity,” Nature Biotechnology, Vol. 26, No. 7, 2008, pp. 774–776. [131] Shvedova, A. A., Kisin, E., Murray, A. R., et al., “Inhalation vs. Aspiration of Single-Walled Carbon Nanotubes in C57BL/6 Mice: Inflammation, Fibrosis, Oxidative Stress, and Mutagenesis,” American Journal of Physiology-Lung Cellular and Molecular Physiology, Vol. 295, No. 4, 2008, pp. L552–L565. [132] Kagan, V. E., Bayir, H., and Shvedova, A. A., “Nanomedicine and Nanotoxicology: Two Sides of the Same Coin,” Nanomedicine, Vol. 1, No. 4, 2005, pp. 313–316. [133] Foldvari, M., and Bagonluri, M., “Carbon Nanotubes as Functional Excipients for Nanomedicines: II. Drug Delivery and Biocompatibility Issues,” Nanomedicine, Vol. 4, No. 3, 2008, pp. 183–200. [134] Flahaut, E., Durrieu, M. C., Remy-Zolghadri, M., et al., “Investigation of the Cytotoxicity of CCVD Carbon Nanotubes Towards Human Umbilical Vein Endothelial Cells,” Carbon, Vol. 44, No. 6, 2006, pp. 1093–1099. [135] Isobe, H., Tanaka, T., Maeda, R., et al., “Preparation, Purification, Characterization, and Cytotoxicity Assessment of Water-Soluble, Transition-Metal-Free Carbon Nanotube Aggregates,” Angewandte Chemie-International Edition, Vol. 45, No. 40, 2006, pp. 6676–6680. [136] Manna, S. K., Sarkar, S., Barr, J., et al., “Single-Walled Carbon Nanotube Induces Oxidative Stress and Activates Nuclear Transcription Factor-Kappa B in Human Keratinocytes,” Nano Letters, Vol. 5, No. 9, 2005, pp. 1676–1684. [137] De Nicola, M., Gattia, D. M., Bellucci, S., et al., “Effect of Different Carbon Nanotubes on Cell Viability and Proliferation,” Journal of Physics-Condensed Matter, Vol. 19, No. 39, 2007, pp. 1–7. [138] Brown, D. M., Kinloch, I. A., Bangert, U., et al., “An In Vitro Study of the Potential of Carbon Nanotubes and Nanofibres to Induce Inflammatory Mediators and Frustrated Phagocytosis,” Carbon, Vol. 45, No. 9, 2007, pp. 1743–1756. [139] Tian, F. R., Cui, D. X., Schwarz, H., et al., “Cytotoxicity of Single-Wall Carbon Nanotubes on Human Fibroblasts,” Toxicology in Vitro, Vol. 20, No. 7, 2006, pp. 1202–1212. [140] Wick, P., Manser, P., Limbach, L. K., et al., “The Degree and Kind of Agglomeration Affect Carbon Nanotube Cytotoxicity,” Toxicology Letters, Vol. 168, No. 2, 2007, pp. 121–131. [141] Sayes, C. M., Liang, F., Hudson, J. L., et al., “Functionalization Density Dependence of Single-Walled Carbon Nanotubes Cytotoxicity In Vitro,” Toxicology Letters, Vol. 161, No. 2, 2006, pp. 135–142. [142] Dumortier, H., Lacotte, S., Pastorin, G., et al., “Functionalized Carbon Nanotubes Are Non-Cytotoxic and Preserve the Functionality of Primary Immune Cells,” Nano Letters, Vol. 6, No. 12, 2006, pp. 1522–1528. [143] Meng, J., Song, L., Xu, H., et al., “Effects of Single-Walled Carbon Nanotubes on the Functions of Plasma Proteins and Potentials in Vascular Prostheses,” Nanomedicine: Nanotechnology, Biology and Medicine, Vol. 1, No. 2, 2005, pp. 136–142. [144] Salvador-Morales, C., Flahaut, E., Sim, E., et al., “Complement Activation and Protein Adsorption by Carbon Nanotubes,” Molecular Immunology, Vol. 43, No. 3, 2006, pp. 193–201. [145] Schipper, M. L., Nakayama-Ratchford, N., Davis, C. R., et al., “A Pilot Toxicology Study of Single-Walled Carbon Nanotubes in a Small Sample of Mice,” Nature Nanotechnology, Vol. 3, No. 4, 2008, pp. 216–221.
94
Functionalized Carbon Nanotubes as Multimodal Drug Delivery Systems [146] Yang, S. T., Wang, X., Jia, G., et al., “Long-Term Accumulation and Low Toxicity of Single-Walled Carbon Nanotubes In Intravenously Exposed Mice,” Toxicology Letters, Vol. 181, No. 3, 2008, pp. 182–189. [147] Leeuw, T. K., Reith, R. M., Simonette, R. A., et al., “Single-Walled Carbon Nanotubes in the Intact Organism: Near IR-Imaging and Biocompatibility Studies in Drosophila,” Nano Letters, Vol. 7, No. 9, 2007, pp. 2650–2654. [148] Mouchet, F., Landois, P., Sarremejean, E., et al., “Characterisation and In Vivo Ecotoxicity Evaluation of Double-Wall Carbon Nanotubes in Larvae of the Amphibian Xenopus laevis,” Aquatic Toxicology, Vol. 87, No. 2, 2008, pp. 127–137. [149] Smith, C. J., Shaw, B. J., and Handy, R. D., “Toxicity of Single Walled Carbon Nanotubes to Rainbow Trout (Oncorhynchus Mykiss): Respiratory Toxicity, Organ Pathologies, and Other Physiological Effects,” Aquatic Toxicology, Vol. 82, No. 2, 2007, pp. 94–109. [150] Cheng, J. P., Flahaut, E., and Cheng, S. H., “Effect of Carbon Nanotubes on Developing Zebrafish (Danio Rerio) Embryos,” Environmental Toxicology and Chemistry, Vol. 26, No. 4, 2007, pp. 708–716. [151] Wörle-Knirsch, J. M., Pulskamp, K., and Krug, H. F., “Oops They Did It Again! Carbon Nanotubes Hoax Scientists in Viability Assays,” Nano Letters, Vol. 6, No. 6, 2006, pp. 1261–1268. [152] Wang, H. F., Wang, J., Deng, X. Y., et al., “Biodistribution of Carbon Single-Wall Carbon Nanotubes in Mice,” Journal of Nanoscience and Nanotechnology, Vol. 4, No. 8, 2004, pp. 1019–1024. [153] Singh, R., Pantarotto, D., Lacerda, L., et al., “Tissue Biodistribution and Blood Clearance Rates of Intravenously Administered Carbon Nanotube Radiotracers,” Proceedings of the National Academy of Sciences of the United States of America, Vol. 103, No. 9, 2006, pp. 3357–3362. [154] Lacerda, L., Herrero, M. A., Venner, K., et al., “Carbon-Nanotube Shape and Individualization Critical for Renal Excretion,” Small, Vol. 4, No. 8, 2008, pp. 1130–1132. [155] Liu, Z., Cai, W. B., He, L. N., et al., “In Vivo Biodistribution and Highly Efficient Tumour Targeting of Carbon Nanotubes in Mice,” Nature Nanotechnology, Vol. 2, No. 1, 2007, pp. 47–52. [156] Yang, S. T., Guo, W., Lin, Y., et al., “Biodistribution of Pristine Single-Walled Carbon Nanotubes In Vivo,” Journal of Physical Chemistry C, Vol. 111, No. 48, 2007, pp. 17761–17764.
CHAPTER 4
Composite Nanoparticles for Cancer Imaging and Therapy: Engineering Surface, Composition, and Shape Lajos P. Balogh, Teyeb Ould Ely, and Wojciech G. Lesniak
4.1
Introduction In this chapter, we will discuss different strategies to synthesize multifunctional nanodevices, which may benefit medical imaging and therapy. Many early examples were generated as a result of the efforts of the National Institute of Health (primarily NCI and NIBIB) to fund development of nanomedicine approaches against cancer, but the general principles apply to other diseases as well. We will explain principles of nanodevice design to regulate surface (size, charge), composition and shape by using examples of synthesis, characterization, and biomedical applications of poly(amidoamine) (PAMAM) dendrimers, composite nanomaterials and shape engineered inorganic nanocrystals. The focus of this chapter is on two main types of multifunctional nanodevices: organic nanoparticles represented by dendrimers, and inorganic nanoparticles such as gold, metal oxides, and so forth. These multifunctional nanodevices might also be thought of as smart nanoparticles. 4.1.1
Nanoscience and Medicine: The Need and the Opportunity
Nanoscience refers to research and development at the atomic, molecular or macromolecular levels, where quantum-size and collective effects present new chemical, physical, or biological properties. This happens approximately in the 1to 100-nm size range and includes creating and using structures, devices, and systems that have novel properties and functions because of their nanoscopic size. Such properties of nanoscale components are distinct from bulk/macroscopic systems [1]. Nanotechnology is the application of nanoscience and engineering to develop, create, and use devices, tools, machines, methods, techniques, processes, and technology for manufacturing nanomaterials, nanostructures, and construct integrated systems that utilize these novel properties and new functions. Thus, the term “nanotechnology” refers to a wide range of technologies that measure, manipulate, or incorporate materials and/or features with at least one dimension between approximately 1 and 100 nm. Such applications exploit the properties of nanoscale components, distinct from bulk/macroscopic systems. The global revolution in
95
96
Composite Nanoparticles for Cancer Imaging and Therapy
nanoscience and nanotechnology has opened new avenues in medicine including a new, complementary field to conventional medicine: Nanomedicine. New journals have appeared, such as Nanomedicine: Nanotechnology, Biology and Medicine (Elsevier), International Journal of Nanomedicine (Dove Scientific), Nanomedicine (Future Medicine) that publish promising approaches and results. Nanoscience offers novel approaches to understand better how the body works, and offers new ways to solve a number of upcoming issues related to human health [2]. Cancer is one of the leading causes of deaths in the United States. It is a very complex disease, which involves many potential reasons and a large number of symptoms. As early interventions are the most successful, there is an increasing need for early identification of cancer/precancerous cells and tissues. Despite of the astonishing development in imaging technology, diagnostics are still lagging behind imaging, and this is where molecular imaging combined with nanotechnology can bring revolutionary improvements. Ideally, the whereabouts of any medicine should be detectable in the body, and an ideal medicine should identify and treat the disease without unwanted side effects, leaving healthy tissues and cells unaffected. Anticancer drugs must be cytotoxic—they do kill cells. Because our goal is to eliminate cancer cells without compromising the normal operation of healthy cells, unwanted side effects must be minimized. Nevertheless, the requirements towards chemotherapy drugs are contradictory: they need to be nontoxic while in the bloodstream or in healthy organs/tissues (i.e., have low systemic toxicity), but must be quite toxic against cancer cells to kill them effectively. They also should not bind to healthy cells, only target cancerous ones, and so on. Typically, a single molecule cannot simultaneously fulfill all these requirements, thus more complex, multifunctional medicines are needed. Multiple functions can exist only in objects of complex composition and structure, which are reproducibly made and practically identical (i.e., narrow polydispersity macromolecules and/or nanoparticles). Only these can have reproducible biodistribution and pharmacokinetics, which are fundamental requisites for any medicine. Properties of nanoparticles are the function of composition, structure, and architecture. Although there are a number of general requirements regarding nanomaterials used in nanomedicine (e.g., lack of toxicity, immunogenecity, and lack of inflammation, minimal side effects, low viscosity, appropriate osmotic pressure and pH), at the same time there are specific requirements as well: As an example, different biodistribution (therefore different composition and structure) is needed for imaging and for therapy. For successful imaging, a small percentage of the injected dose present in the target organ for a limited time may be sufficient to provide a detailed image (e.g., by positron emission tomography (PET)). Good quality imaging is the prerequisite for successful diagnosis. Observing irregularity in tissue structure does not mean that we can efficiently identify the cause and treat it. Successful diagnosis requires the identification of one or more known specific signature(s) different from healthy tissue. Contrast agents have to have a sufficient residence time to acquire the image and must not be toxic. For therapy, most of the dose has to be delivered to the target tissue or cells, and a therapeutic effect may need longer time. Many times therapeutic agents must be toxic, but in controlled ways. For
4.1 Introduction
97
the same reasons, determination of biosafety and development of the best dose-regimen require completely different experiments. 4.1.2
Nanodevices
Multiple methods and complex multifunctional materials can only address the apparent complexity of biology. Just like devices are designed to perform a specific task, nanodevices are built from different materials and connected by chemical and physical forces in a reproducible manner. We distinguish nanoparticulate materials used in biologic experiments from nanodevices. Nanodevices are designed to perform a specific task, and are made of molecular building blocks that have predetermined properties. Each component performs a specific function (e.g., binding, emitting energy, absorbing a specific wavelength), when the assembled nanodevice comes in contact with various components of the biologic system in vitro or in vivo. Nanomaterials comprise nanostructured materials and nanoscale materials. Nanostructured materials contain physically or chemically distinguishable components, at least one of which is nanoscale in one or more dimensions. Nanoscale (nanosized) materials have one or more dimensions from approximately 1 to 100 nm. It is common knowledge that properties of nanoscale particulate materials (nanoparticles) depend on their size. In reality, properties are the function of not only size, but also of composition, surface, volume, shape, structure, architecture, flexibility, conformations, and so forth. For a particular application, properties could be either relevant or irrelevant. In addition, properties and “uniformity” of nanoparticulate materials are characterized by their average individual properties and by the distribution of those relevant individual properties over a very large number of particles. These properties are the function of composition (chemical and physical), structure (linear measure, surface, volume, shape, and morphology), and architecture (i.e., how the individual components are connected) of the building blocks, as well as their structural hierarchy. So, what was known for small molecules as structure-property relationships (hydrophobic, electronic, steric) becomes a set of much more complex information for nanodevices. We observe properties by measuring interactions of objects with our characterization instruments. Therefore, “properties” inherently involve how the objects interact with their (chemical, physical, biologic, and so forth) environment. To ensure the existence of necessary properties, appropriate characterization methods must be used to provide information about relevant properties including their distributions. This is a cyclic and synergistic process: Improved characterization leads to a better understanding of composition-structure-property relationship, which results in better nanoparticle and nanostructure fabrication, and so on. Reproducible synthesis of practically uniform nanodevices with biologically relevant properties is necessary to observe a reproducible biodistribution, which is key to successful imaging and therapy. To design, synthesize, and characterize multifunctional nanodevices, it is necessary to clearly understand the relevant properties of (1) the material system, (2) the given biologic system, and (3) the mechanism of their interaction. In this chapter we describe examples of molecular building blocks and how they can be assembled into functional nanodevices.
98
Composite Nanoparticles for Cancer Imaging and Therapy
4.1.3
Principles of Nanodevice Design
Nanosized objects are much smaller than cells, and interact only with a small area of the cell surface. At the nanoscale the surface to volume ratios are very large, thus surface forces (Coulomb interactions, hydrogen bonding, van der Waals forces) greatly influence the characteristics of these devices and control the interactions of these devices with surfaces of biological systems. Most important parameters are size (as relates to “nano”-properties), charge (because Coulomb interactions are the strongest), specific binding motifs that are preferred by receptors (peptides, antibodies, vitamins, sugars), flexibility (e.g., crystalline inorganics versus liposomes), density (organics compared to metals), and so forth. In summary, properties of nanoscale organic and inorganic particles are a function of composition, size, structure, and architecture of all the components involved (including surface adsorbed molecules for nanocrystals and solvent molecules for organic nanoparticles). The major variables are (1) vehicle of defined composition, shape, and surface (size, and charge, as a result of surface composition at a specific pH), (2) binding motif(s) (peptides, antibodies, vitamins, sugars) that are covalently attached to the surface (active surface), (3) imaging component (absorbs or emits a measurable signal, e.g., radiation, electromagnetic field, light), or/and (4) therapeutic component (interacts with cell signaling, induces apoptosis, delivers radiation). These components first have to be synthesized and assembled by precise chemical and physical methods and characterized by various analytical techniques before undergoing in vitro and in vivo biologic testing. Dendrimers are one of the most promising multifunctional nanomaterials because of their narrow distribution, biofriendly composition, and chemically adjustable multiple functionality. Due to the available multifunctionality, one can design and assemble dendrimer-based nanodevices that after injection into the bloodstream allow detection, bind to specific target(s), deliver a drug, and indicate whether a therapeutic effect was achieved.
4.2
Materials for Nanodevice Fabrication Two groups of materials used for nanodevice fabrication will be highlighted here: dendritic materials with treelike structures, and shape-engineered inorganic nanoparticles. 4.2.1
Dendrimers
The term “dendrimer” may mean a perfectly branching geometry, a highly branched symmetric molecule, or a certain synthesized material. Dendrimer materials contain molecules with regularly and symmetrically branching treelike structures. Dendrimer molecules are synthesized through a sequence of iterative reactions resulting in a chemical structure emanating from a core and branching outward, producing “three-dimensional” molecules that have practically identical terminal groups after synthesis. Dendrimer materials are composed of highly branched small molecules, oligomers, or polymers. Dendritic polymers are usually contain one generation (named after the number of successive and repetitive synthesis steps) [3, 4] of
4.2 Materials for Nanodevice Fabrication
99
nearly monodisperse macromolecules with a well-defined structure, with a branch occurring at each monomer unit. Dendrimer materials have controlled size and shape characteristics, with narrow molecular weight and size distribution. Their molecules possess a persistent spherical structure in solution with well-defined molecular diameter and surface characteristics. Dendrimer families are synthesized using a variety of different synthetic routes; typically applying repeated sequences of reaction (and purification) steps. Some of the significant families are the poly(amidoamine) (PAMAM), [5] arborols, [6] poly(propyleneimine) (PPI), [7] polyether dendrimers,[8] and phosphorus-based dendrimers [9]. In addition to general properties, dendrimer molecules of various families have properties that are characteristic to their family. Members (generations) of a dendrimer family contain the same molecular motifs, but their size and mass are different and therefore may have very diverse physical properties (also depending on their termini) [10]. These symmetric, multifunctional, and nearly monodisperse dendrimers are attractive candidates for nanodevice construction by combining their multiple external and internal functionalities into nanoscale entities. One can construct nanodevices from dendrimers and inorganic materials by using dendrimers as templates to form composite nanoparticles/devices, or from prefabricated nanoparticles (NPs) and stabilize/modify their surface with dendrimers. There are inherent trade-offs when combining many functional groups into one nanostructure: for example, a limited number of attachment sites are available on the particle surface, making it difficult to couple several functional groups in sufficient concentration for each function; and some groups may interact sterically or chemically, or alter the activity of each other when combined in close proximity. Multiple functionalizations may also reduce nanoparticle stability or adversely change in vivo pharmacokinetics. With significant characterization and fine-tuning, both dendrimers and shape-engineered nanocrystals appear to have a great potential, since each of them has multiple site functionality. The combination of these two materials is expected to provide an exciting and rich platform for cancer imaging, targeting and therapy. Inorganic–organic hybrids combine inorganic nanostructures (nanoparticles, nanorods, shaped nanostructures) with organic molecules (polymers, dendrimers, proteins, DNAs). The resulting hybrid brings together the properties of both materials, for example the physical and spectroscopic characteristics of the nanocrystal and the biomolecular function of the surface-attached entities, without being restricted to covalent reactions. 4.2.2 Engineering Size, Charge, and Surface Functionality of PAMAM Dendrimers
PAMAM dendrimers were first synthesized by Tomalia in 1983 [11, 12]. This family of dendrimers is synthesized through Michael addition of methyl acrylate to primary amine functions leading to methyl ester, which is in the next step reacted with ethylenediamine (aminolysis). Completion of these reactions gives generation 0 (G0) dendrimer molecules. Sequential repetitions of these two reactions (coupled
100
Composite Nanoparticles for Cancer Imaging and Therapy
with purification steps) results in higher generation dendrimers of increased size and with an exponential increase in the number of terminal groups. Table 4.1 summarizes various theoretical characteristics of PAMAMs however in reality the values are different since their molecules always possess structures deviating from theory (e.g. incomplete substitutions, missing arms or intramolecular loops) [13-15], reducing ideality but in large macromolecules these deviations do not modify the fundamental properties (size, shape, charge, etc. PAMAM dendrimers are polyionic and flexible - their size depends on the protonation state of the interior tertiary amines and the nature and functionalization of their terminal groups. Controllable molecular weight, tunable surface properties like: charge, hydrophobicity, hydrophobicity, and highly modifiable terminal groups offer a variety of applications. PAMAMs can be prepared to be non-immunogenic, non-mutagenic, and nontoxic [16]. They contain β-alanine subunits and resemble in size and behavior fundamental blood proteins (generation 3: - insulin; generation 4: - cytochrome C; generation 5: - hemoglobin). They are not broken down by enzymes and are generally removed from the bloodstream by the filter organs [17]. The size of dissolved PAMAM molecules depends on the generation, and even though it is highly influenced by the surface groups, solvent, and pH, selecting the generation also defines the size. Figure 4.1 gives examples of commonly used reactions for manipulation of surface properties of PAMAM dendrimers. By partially or fully eliminating protonable or polarizable surface groups by acylation, succinamination, or glycidolation, their surface net charge (as expressed by their zeta-potential at a given pH) can be effectively controlled [18–21]. Terminal primary amino groups of PAMAM dendrimers can also be covalently modified with dyes, drugs, diagnostic/imaging modules, chelating agents [22], targeting moieties such as folate [23], peptides [24–26], antibodies [27], and apoptotic sensors [28–30], to combine multiple functionalities into a single nanodevice. In the following example, we briefly describe exactly how a nanodevice showing superior binding to specific tumor receptors can be fabricated and characterized [24, 26]. Every targeted delivery has the same philosophical problem: the more specific they are, the less general they get. Thus, we have decided to aim at a general target, the tumor microvasculature, which is part of every growing tumor. The αvβ3 integrin receptors are highly overexpressed on the angiogenic endothelial cells present in tumors, but are not available in normal blood vessels. RGD peptides selectively and strongly bind to the αvβ3 integrins, and they recently passed clinical trials as an Table 4.1
Features of Poly(Amidoamine) Dendrimers
G
M
# of NH2
D [nm]
0 1 2 3 4 5
517 1430 3256 6906 14216 28826
4 8 16 32 64 128
1.5 2.2 2.9 3.6 4.5 5.4
G=generation, # of NH2=number of primary amines, D=diameter (hydrodynamic diameter of the fully protonated single dendrimer molecule in water.
4.2 Materials for Nanodevice Fabrication
101
Figure 4.1 Schematic examples of commonly used reactions for manipulation of surface properties of PAMAM dendrimers: x=initial number of primary amines in starting dendrimer, z=number of modifying groups, y = (x - z) number of primary amines in the derivative.
antiangiogenic agent. Therefore, nanodevices (NDs) with covalently attached RGD peptides to their surface can be selectively bound to endothelial cells of the tumor microvasculature. In our example, four copies of a specific cyclic RGD peptide were conjugated per dendrimer macromolecule. Fabrication of this nanodevice involved the following four steps: partial acetylation (4.1), attachment of biotin (4.2), converting the rest of primary amine termini to succinamic acid groups (4.3), and conjugation with cRGD peptides (4.4): E5.(NH2)119 => E5.(Ac)72(NH2)47
(4.1)
E5.(Ac)72(NH2)47 => E5.(Ac)72(BT)8(NH2)39
(4.2)
E5.(Ac)72(BT)8(NH2)39 => E5.(Ac)72(BT)8(SAH)39
(4.3)
E5.(Ac)72(BT)8(SAH)39 => E5.(NHAc)72(NHBT)8 (SAH)35(SAcRGD)4
(4.4)
In these equations, Ac is acetamide, BT is biotin, SAH is succinamic acid, and cRGD is Phe(f)-Lys-Arg-Gly-Asp peptide. Subscripts correspond to number of functional groups measured by various analytical techniques. Acetylation was performed first (partial substitution of primary amines decreases the number of possible conjugation sites), to decrease toxicity and non-
102
Composite Nanoparticles for Cancer Imaging and Therapy
specific uptake (since positively charged, amino-terminated PAMAM dendrimers are known to interact with negatively charged cellular membranes leading to cytotoxicity) [31]. In this example biotin (BT) is selected as a reporter molecule for ND detection with fluorescent antibiotin antibody or avidin conjugated with horseradish peroxidase. Carboxyl groups of succinamic acid were used as linkers for conjugation of cRGD peptides to form an amide bond with the lysine side chain amino group of the peptide. Materials were purified after each synthetic step and analyzed by multiple techniques. Detailed chemical and physical characterization of multifunctional nanodevices plays a crucial role in understanding their biological properties. For dendrimer NDs methods include: size exclusion chromatography equipped with multiangle laser light scattering (MALLS), UV-vis diode-array detector (DAD) and differential refractive index (DRI) online detectors, potentiometric acid-base titration, matrix assisted laser desorption ionization–time-of-flight (MALDI-TOF) mass spectrometry, polyacrylamide gel electrophoresis (PAGE), and nuclear magnetic resonance (NMR), providing complementary results for comprehensive analysis. SEC is used to determine the average molecular mass, molecular mass distribution, and polydispersity of the nanodevices. Figure 4.2 demonstrates a SEC chromatogram of a technical grade PAMAM_E5.NH 2 dendrimer. This dendrimer contains (E5.NH2)2 dimers (16.42%), E5.NH2 generation five dendrimers (80.42%), and a small amount of trailing generation E4.NH2 (3.15%) PAMAMs. Average molecular mass of Mn, 27 280, Mw 30 280 with a polydispersity of Mw/Mn 1.110 were obtained for this material. The number of terminal functional groups can be calculated based on an increase of average molecular mass while comparison of polydispersity indexes before and after reaction indicates whether all components of the material were equally modified. Figure 4.3 illustrates MALDI-TOF of a generation 5 technical grade PAMAM.
B 1.067
A 0.1953
C 30
0.0226
.04179 35
40
Figure 4.2 Typical molecular mass distribution of a technical PAMAM_E5.NH2 dendrimer shown by the differential refraction index (DRI) detector-signal of its size exclusion chromatogram. The envelop of the DRI signal was deconvoluted using PeakFit software. (a) Dimers, (b) major component of generation 5, and (c) trailing generation (G4 molecules).
4.2 Materials for Nanodevice Fabrication
Figure 4.3 derivative.
103
MALDI-TOF spectra of (a) PAMAM_E5.(NH2)119 dendrimer, and (b) its acetylated
The shift towards higher molecular masses of all detected signals indicates successful modification of terminal amines of PAMAM_E5.(NH2)119. Using differences in molecular masses of the highest frequency fragments of the major peaks the acetylation degree can be approximately calculated. (Assuming that chemical changes were uniform on all molecules of the dendrimer material.) Elimination of primary amines can be fairly precisely measured by acid-base titration within the 3 to 12 pH range (Figure 4.4). The inflection point relates to deprotonation of primary amino groups and shifts from 7.74 micromoles (Figure 4.4(a)) to 3.2 mol equivalent (Figure 4.4(b)) of the titrant, indicating that 58.7% of initial amino groups were acetylated (in the case of titration presented in Figure 4.4(a) 0.00184g of dendrimer having MW 28270 was used; n=0.00184/28270=6.5×10-8 mole=>0.065 μmole, 7.74/0.065=119 ±5). The end points of titration for tertiary amines do not change (Figure 4.4(a) 6.83 μmol and Figure 4.4(b) 6.9 μmol; the observed 0.07 μmol difference is within the experimental error). The acetylation degree can also be calculated using integrals from proton NMR spectroscopy. After acetylation two new peaks related to protons of acetamide (Ac) and acetic acid (AA, byproduct of acetylation forming alkyl ammonium acetate) appear. Based on the theoretical number of protons of dendrimer (D, 2020) and integrals (see Figure 4.5) the molar ratio D : Ac and D: AA present in the sample can be obtained: 99.17/2020=0.049, (acetic group has 3 protons) 11.23/3=3.74 =>3.74/0.049=76.39 (which is in good agreement with other analytical data presented above for the E5.(Ac)72(NH2)47 derivative) and (1/3)/0.049=6.8. It must be pointed out that these ratios are somewhat overestimated since the practical number of 1H of dendrimer is usually less than ideal because of the structural imperfections in PAMAMs. Free acetic acid is removed from the sample during purifications (data not shown).
104
Composite Nanoparticles for Cancer Imaging and Therapy
Figure 4.4 Potentiometric titration of (a) the PAMAM_E5.NH2 amino-terminated dendrimer, and (b) its acetylated derivative, which shows a shift towards a lower number of primary amines in the product. (Reproduced from [26] with permission.)
Figure 4.5 1H NMR spectroscopy of (a) the PAMAM_E5.NH2 amino terminated dendrimer, and (b) its acetylated derivative.
Figure 4.6 illustrates PAGE electropherograms of E5.(NHAc)72(NHBT)8 (NHSAH)35(NHSAcRGD)4 and the four intermediates during its synthesis. All synthetic intermediates exhibit a different electrophoretic mobility, reflecting the decrease of net positive charge and increase of molecular mass due to the increasing substitution of free terminal amines. After succinamic acid conversion the resulting conjugates are negatively charged, and migrate towards the positive electrode (image B). Change of polarity indicates that E5.(NHAc)72(NHBT)8(NH2)39 has been modified with succinamic acid. The E5.(NHAc)72(NHBT)8(SAH)35(SAcRGD)4 nanodevice is nontoxic within physiologic concentration ranges (up to 2 μmol) and shows superior binding to the biologic target αvβ3 integrins when comparing with NDs without cRGD and with the free cRGD peptide [26]. Even a 400-fold excess of free peptide was unable to displace over 25% of the bound NDs in a competitive binding experiment in vitro,
4.2 Materials for Nanodevice Fabrication
105
Figure 4.6 PAGE of nanodevices with opposite polarities: (1A) E5.(NH2)119, (2A) E5.(NHAc)72(NH2)47, (3A) E5.(NHAc)72(NHB)8(NH2)39, (4A) E5.(NHAc)72(NHB)8(NHSAH)39, (5A) E5.(NHAc)72(NHB)8(NHSAH)35 (NHSAcRGD)4, 1B: E5.(NHAc)72(NHB)8(NHSAH)39. (Reproduced from [26] with permission.)
which is explained as the consequence of multivalent binding (i.e., simultaneous adhesion of multiple peptide ligands on the ND to multiple receptors on the plate or cell surface (Figure 4.7)). 4.2.3
Dendrimer Nanocomposites: Engineering Composition
The network of polyionic dendrimers can topologically trap (“encapsulate”) various guests, such as metal atoms, ions, anions, or drug molecules [32]. This guest-host concept circumvents limitations set forth by stoichiometry when the interior of the organic nanoparticle is modified. PAMAM dendrimers are also used as templates to fabricate composite nanoparticles for imaging and therapeutic applications [33–35]. The resulting composite nanoparticles are spherical, have a well-defined and variable size in addition to a specifically charged surface (positively/negatively charged and/or neutral surface). These properties permit size-related, charge-related, and/or surface recognition of the PAMAM-based nanodevices, which may contain encapsulated therapeutic materials, such as radioisotopes [36, 37]. Dendrimer nanocomposites [32] possess chemical and physical properties of the guest molecules or atoms, but physical interactions with the environment of the nanoparticle (solubility, compatibility) are dominated by the surface of the host molecule. As a consequence, an inorganic particle can be manipulated as if it was an organic one, and composite nanoparticles with either cationic, anionic, neutral, lipophilic, lipophobic, or mixed surfaces can be created [38, 39]. The principle of dendrimer nanocomposite preparation is illustrated in Figure 4.8. In the first step, the dendrimer template is equilibrated in solution with cations or anions, which is followed by in situ chemical reactions induced by a reducing agent (e.g., N2H4 - hydrazine) or physical treatment like irradiation (UV, light) that generates the composite nanoparticles. Both procedures yield small domains of
106
Composite Nanoparticles for Cancer Imaging and Therapy 0.4
0 BT-ND
ND
ND+4X
ND+40X
ND+400X
Figure 4.7 Nanodevice binding to integrin receptors in an in vitro plate binding assay. Plates coated with αvβ3 integrins were treated with the respective nanodevices with and without covalently bound cRGD targeting moieties in the absence (left two columns) and in the presence of the cRGD competitor peptide (right three columns). Biotin functionalities of the bound NDs were detected by avidin linked to horseradish peroxidase (the color was developed with o-phenylenediamine). The 4X, 40X, and 400X labels denote 4, 40, and 400 times of molar excess of competing cRGD mixed with the cRGD-BT-ND nanodevice, respectively, competing for the binding sites. (Reproduced from [26] with permission.)
Figure 4.8 Synthesis of PAMAM dendrimer nanocomposite devices via reactive encapsulation. (Reproduced from [120] with permission.)
metallic clusters dispersed in and integrated with the organic template without creating covalent bonds between the dendrimer and the entrapped inorganic matter. Considering the formation mechanism and specific chemistry involved, three basic structural types form [40] depending on generation, surface groups, and the inorganic/organic molar ratio: inorganic domains may be internal (I), mixed (M), and external (E) to the organic template molecules (“I”, “E”, and “M” structures). Stable internal composite structures of a single macromolecule as a host can be formed only from dendrimers and not from linear macromolecules. Reduction of
4.2 Materials for Nanodevice Fabrication
107
metals ions is accompanied with changes in characteristic spectroscopic features (Figure 4.9). A solution of {(Au0)7.71-cRGD-BT-ND} nanodevice is red in color and exhibits two peaks, one at 282 nm and the other at 520 nm. The signal at 520 nm can be assigned as a gold plasmon resonance resulting from aggregated nanoparticles. Absorbance peaks at 282 nm originate from single (not aggregated) CND particles. The organic template shows only a shoulder with maximum at 275 nm. Commercially available poly(amidoamine) PAMAM dendrimers and tecto-dendrimers are used to fabricate gold-PAMAM nanocomposites [40–44]. First, gold-dendrimer complexes are prepared by mixing dilute solutions of PAMAM dendrimer with a dilute solution of HAuCl4. In this step, a salt forms between the tertiary amines of the dendrimer interior and the tetrachloroaurate anions. Then, irradiating the sample with light completes the reduction. These composite nanoparticles are spherical; they have a nearly monodisperse distribution as well as well-defined and variable size, and a specific surface to permit both size related and/or surface recognition targeting of the encapsulated radioisotopes. The development of a specific nanodevice that will deliver β-radiation using a medium half-life (t1/2 = 2.69 d) Au-198 isotope in a form of {198Au} nanocomposites could be very useful in tumor therapy [37]. Radioactive gold nanocomposites are synthesized as monodisperse hybrid nanoparticles composed of radioactive guests immobilized by dendritic polymer hosts. The starting materials and the obtained products were carefully characterized by UV-vis, 1H and 13C NMR spectroscopy, size exclusion chromatography (SEC) and high-resolution transmission electron microscopy (HRTEM). Details of the experimental procedure and analytical techniques can be found in a previous study [40]. Activation of elemental gold in the nanocomposites into Au-198 was achieved by direct irradiation performed at the
Figure 4.9 UV-Vis spectra of PAMAM_E5.(NHAc)72(NHBT)8(NHSAH)35(NHSA) (cRGD)4 (blue dots) targeted dendrimer nanodevice used as template and its composite derivative: {(Au0)7.71PAMAM_E5. (NHAc)72(NHBT)8(NHSAH)35(NHSA) (cRGD)4}, for short: {(Au0)7.71-cRGD-BT-ND}, a gold composite nanodevice (CND).
108
Composite Nanoparticles for Cancer Imaging and Therapy
Ford Nuclear reactor of Phoenix Memorial Laboratories. Activation of gold nanocomposites into {Au-198} was successfully carried out both in solid form and in aqueous solutions. In this set of experiments a 5-nm gold nanocomposite {Au10-PAMAM_E5.NH2} was fabricated (gold content: 3.11 w/w%) and aliquot parts of an aqueous solution (2.86 w/w% for dendrimer) were irradiated in polypropylene containers. Direct activation of gold in the {Au(0)} nanocomposite to {198Au(0)} may be achieved by short-term neutron irradiation without damaging the organic matrix of the hybrid particle [37]. The particle size and the size distribution of the activated gold nanocomposites were studied by electron microscopy. At the lowest radiation level, the nanoparticles did not undergo structural changes. However, at the highest dose, the particle size considerably increased and radiation polymerization ensued. Composite nanodevices are characterized by multiple methods: UV-vis, fluorescence spectroscopy, dynamic light scattering, zeta potential measurements (to measure surface net charge), transmission electron microscopy, X-ray energy dispersive spectroscopy, and selected area electron diffraction (due to high electron density of metal domains to obtained their morphology, composition, and crystallites). Such a comprehensive characterization plays a critical role in understanding biological responses of multifunctional composite nanodevices and which responses are affected by size, surface charge, and the presence of terminal functionalities like: targeting moieties, drugs, and imaging molecules. The {(Au0)7.71-cRGD-BT-ND} composite nanodevice (CND) shows remarkable binding to the human dermal microvascular endothelial (HDMEC) cells (Figure 4.10(c)). A relatively low level of the {(Au0)7.71-BT-ND} could also be detected in specimen, the most likely due to nonspecific cellular internalization (Figure 4.10(b)). In summary, binding of metal clusters by encapsulation or chelation of multiple ions, including radioisotopes in a dendrimer molecule of defined size and engineered surface properties allows for the simple fabrication of dendrimer-based nanodevices. The radioactivity delivered to a tumor can be increased either by increasing the number and/or the activity of radioisotopes bound per nanodevice. Increasing the radioactivity is not possible with tumor-directed antibody and peptide therapy as only one radioactive molecule can usually be linked to a given anti-
(a)
(b)
(c)
Figure 4.10 (Color plate 5) Confocal fluorescence microscopy of αvβ3 receptor expressing HDMEC cells exposed to nanodevices, FITC labeled anti-biotin antibody (green) and DAPI (blue) staining only the nuclei. (a) No treatment, (b) {(Au0)7.71-BT-ND} - nanodevices without cRGD peptides, and (C) {(Au0)7.71-cRGD-BT-ND} treated.26 (BT: biotin).
4.2 Materials for Nanodevice Fabrication
109
body or peptide. These molecular nanocomposites are an exciting new class of agents with many potential medical applications. As we have previously discussed, their potential use in radiation treatment of cancer is especially intriguing. These nanodevices are synthesized as monodisperse hybrid nanoparticles composed of radioactive guests immobilized by dendritic polymer hosts. For example, delivery of beta-radiation may be achieved by encapsulating radioactive gold into the nanocomposites (i.e., by using {198Au} nanoparticles). Delivery of these radioactive gold CNDs was also achieved in an in vivo model by nonspecific interactions using intratumoral injections [37]. 4.2.4
Inorganic Nanoparticles: Engineering Shape
Shape-engineered nanostructures are comprised of crystalline domains engineered to provide a high degree of functionality. These nanostructures possess well-defined surfaces and morphologies because their nucleation and growth are controlled at the atomic level. Solution-based methods for producing shaped nanostructures require precise tuning of nucleation and growth to achieve crystallographic control. However, this is often not trivial to achieve since growth and nucleation are governed by a number of thermodynamic (e.g., temperature, reduction potential) and kinetic (e.g., reactant concentration, diffusion, solubility, reaction rate) parameters that are mutually dependant. During the early stages of nucleation, clusters with a small number of atoms usually adopt compact shapes. Among others, the icosahedron and the cubo-octahedron are the most common structures. Transition metal clusters with a small number of atoms (<150~200) crystallize in the form of icosahedra [45–108]. That structure becomes unstable for a large number of atoms and transforms to cubo-octahedra, which is just a patch of the face-centered-cubic lattice. The smallest cluster (order one) is formed by 13 atoms; one in the center (called zeroth shell) and 12 located at the vertices (first shell). The next cluster size (order 2) is formed by covering the surface of the previous cluster with a crust formed by 42 atoms. In the cubo-octahedron they distribute in three shells and in the icosahedron they occupy only two shells. By covering further this cluster with a crust of 92 atoms we obtain the next cluster size (order 3). The surface atoms are distributed in a number of shells that depends on the geometry of the cluster. In general, the total number of atoms in a cluster of order v is given by: NT = (10/3)υ3 + 5υ2 + (11/3)υ + 1, and the number of crust atoms Ns that form the cluster surface is given by: Ns = 10 υ2 + 2. Upon increasing the size, the excess of free energy per unit area for a particular crystallographic face, largely determines the faceting and crystal shaping. Usually, the most stable morphologies are those bound by the low-index crystal planes that exhibit closest atomic packing and minimal free surface energy. The growing crystal goes through various processes (twinning, ripening, oriented attachment) to minimize its surface energy and adopt the most energetically favorable morphology. A wide range of shaped nanostructures has been made in a wide range of geometries (sphere, cube, disks, rods, branched shapes). The synthesis of such anisotropic shapes can be achieved in solution by several strategies. The most effective approaches exploit: (1) reaction confined in micelles, (2) catalysis with metal particles, (3) seeded growth, (4) oriented attachment mechanisms, (5) surfactant or sol-
110
Composite Nanoparticles for Cancer Imaging and Therapy
vent induced anisotropy, and (6) application of external electric or magnetic fields. Details of these mechanisms are highlighted in several books and reviews [45]. These mechanisms could act separately or in concert to generate even more complex geometries (Figure 4.11) where the growth and shaping phenomena involved the development of microstructural defects generating a negative curvatures along (110) planes followed by solvent induced solvothermal etching processes [46]. The advantage of these nanostructures is that each shape represents a unique class of functionality, both in terms of geometry and degree of freedom (Figure 4.12). The increase of the degree of the particle complexity with regard to shape and composition will naturally enhance the particle functionality, as the chemical-physical properties of each material could be grouped or dispersed within the same nanostructures, and/or modulated as a consequence of mutual interactions among the components. Furthermore, multiple components hybrid multibranched nanostructures, made of different materials fused together in a single particle with or without bridging molecules have been reported [47]. These hybrid heteronanostructures are either based on a spherelike structure or on rodlike sections [45]. 4.2.4.1
Sphere-Based Shape-Engineered Heteronanostructures
This class refers to two or more nearly spherical inorganic particles connected epitaxially via a small junction area and so far they have been obtained for a limited number of material combinations [45]. These materials were first synthesized upon attempting to make alloyed nanoparticles of two compounds characterized by partial miscibility and large interfacial energy. In such condition, the two materials can phase-segregate into separate domains, respectively to form dimerlike structures. Several types of nanocrystal hetero-oligomers have been synthesized, such as Co-Pd
Figure 4.11 (Series 1) MnO nanoparticles. Progression of forms ranging from squares through partially “etched” (batch TOA/OA, 2:1; H2O/Mn, -4:1) squares (-132 nm) to fully formed cross forms and their derivatives. Part E (series 1) shows evolution of crystal growth conditions (arrow) kT values that distinguish and texture of the evolving nanocrystal. * and ** represent the limiting spiral, nucleation, and dendritic growth fields (series 2) progression of a hexapod nanoparticle to octahedral structures and derivatives thereof (batch TOA/OA, 2:1; H2O/Mn, -8:1). (Reproduced from [46] with permission.)
4.2 Materials for Nanodevice Fabrication
111
Figure 4.12 Nanohexapods and their fragments synthesized in the presence of TOA/OA (2:1) and H2O/Mn (8:1; A-C). (D) Dark field image of a hexapod showing its octahedral (trigonal antiprism) geometry where every other branch was found to be out of plane. Detail of the hexapod derivatives are obtained upon tilting each particle. (A, E) Transform upon tilting to octahedral (C) or square base pyramid (F); the tripod (G) transforms into trigonal base bipyramid (H). (Reproduced with permission from [46].)
and Cu-In sulfide dimmers [48, 49], FePt-CdS dimmers [50], and Fe2O3-MeS (where Me = Zn, Cd, Hg) [51]. A rather comprehensive approach to hybrid heteronanostructures comprising several new combinations of materials in a variety of heterostructure geometries (in terms of number of domains, their respective shapes and mutual spatial arrangement) has been recently reported [52]. Two interesting concepts have been introduced. First the heterodimers (as if they were small molecules) can react with each other via their functional moieties to form larger molecules. Second, the heteronanocrystals themselves can be used as a seeds to grow more complex nanostructures. This fact has been proven by the formation of ternary heteronanostructures (Fe3O4-Au-PbSe), in which a rodlike PbSe section grew out of a Fe3O4-Au dimer seed. It has to be noted that in such a hetero-oligomer growth, the most frequently observed coupled planes forming interface junctions are usually those that allow the best lattice fit between the respective crystal structures of the two domains. 4.2.4.2
Rod-Based Shape-Engineered Heteronanostructures
Most of rod-based shape-engineered heteronanostructures developed so far have been based on cadmium chalcogenide materials in the wurtzite phase, as these can be easily grown in rodlike and hyperbranched forms with a high control over their geometrical parameters [53, 54]. The rod sections, elongated in the c-axis direction, terminate into polar facets that are the most chemically reactive in such anisotropic nanostructures. This opens up the possibility of nucleating a second material exclusively at these locations. One additional peculiarity of the wurtzite structure is the absence of a plane of symmetry perpendicular to the c-axis, meaning that the two
112
Composite Nanoparticles for Cancer Imaging and Therapy
sides of the rods are chemically dissimilar. Combined with the facet selective surface adsorption, this fact translates into the possibility of controlling the growth mode of lateral sections of a second material onto the tips of starting nanorods [47]. An extended strategy to access several types of Au tipped dumbbell-like nanocrystal heterostructures has been developed recently, which involves the selective oxidation of either PbSe or CdTe sacrificial domains, initially grown on CdSe and CdS nanorods, with a Au(III):surfactant complex [47, 55]. This approach is especially advantageous in that Au domains can be grown via an indirect two-step procedure onto the tip of nanorods of materials like CdS for which direct metal deposition is impracticable. Interestingly, it was found that one-sided growth of a gold tip on a semiconductor quantum dot or rod is preceded by a two-sided growth. Experimental analysis and theoretical modeling show that a ripening process drives gold from one end to the other, leading to a phase-segregated structure and extending the realm of ripening phenomena and their importance in nanostructures. There have been many serious questions and concerns raised regarding the cytotoxicity of inorganic QDs containing Cd, Se, Zn, Te, Hg, and Pb [56]. These chemicals can be potent toxins, neurotoxins, and/or teratogens depending on the dosage, complexation, and accumulation in the liver and the nervous system. Therefore, chalcogenide-free shaped material involving noble metals and transition metal oxides seem to be the alternate solution. So far, there is a limited example of such hetero-nanostructures. Among these, one could cite Au-Fe3O4 dimers and flowerlike particles that have been obtained by heterogeneous growth of Au particle seed, followed by oxidation under air at room temperature. Similarly, CoPt3-Au heterodimers were formed by heterogeneous nucleation of Au on preformed CoPt3 [57]. A series of heterodimers involving Fe3O4-Au, FePt-Ag, Au-Ag has been achieved in a biphasic aqueous/organic systems using ultrasonic emulsification [58]. Recently, Ni-Au barcode arrays were made via on-wire lithography [59]; see Figure 4.13. These hetero-nanostructures represent another exciting class of materials that could have potential application in cancer diagnostic, imaging and therapy (Figure 4.14). Similar nanostructures (but involving CdS) were developed via wet chemistry [60]. 4.2.4.3
Surface Functionalization
Highly uniform shape-engineered nanoparticles of controlled composition and geometry have been made via wet chemistry, taking place in organic solutions. This nonaqueous method offers control over the particle size and shape with close to atomic layer precision—two important parameters that affect the chemical and physical properties of the nanostructures. In almost all of these colloidal systems, a layer of surfactant molecules is essential to prevent the nanoparticles from aggregation. The long hydrocarbon chains of the surfactants are responsible for the hydrophobicity. Biological applications of these nanoparticles are greatly restricted because of the poor solubility of these long aliphatic chains in aqueous media. To realize the potentials of shape-engineered nanoparticles with controlled composition and geometry for medical applications, it is important to develop a generic
4.3 Application Examples of Nanodevices
113
Figure 4.13 Scheme of the nanodisk code (NDC) method. (a) Synthesis and functionalization. (b) Thirteen possible 5-disk-pair NDCs with corresponding binary codes. (c) 2-D (top) and 3-D (middle) scanning Raman microscopy images of a 11111 NDC. Representative Raman spectrum of MB (bottom) taken from the center of the hot spot generated in the middle disk pair shown in the Raman maps above. (Reproduced with permission from [59].)
method for transferring the particles from the organic phase to aqueous solutions. Many approaches have been worked out to improve solubility and provide biofunctionalization, such as ligand exchange processes [51, 61–63], intercalation processes [64], or encapsulation in hydrophyilic matrices such as polymers [65], dendrimer bridging [64], and use of phospholipids [66]. These strategies led finally to the development of a new class of inorganic–organic nanohybrids [67].
4.3
Application Examples of Nanodevices 4.3.1
Nanoparticles in Cancer Imaging
Remarkable progress has been achieved during the past decades in cancer imaging technologies. Many different types of radiation have been exploited to provide images of the structure and function of tissues inside soft and solid tumors (see Table 4.2). Electromagnetic radiation, including radio waves (magnetic resonance imaging (MRI)), visible and near-infrared light (optical imaging), X-rays (X-ray computed tomography (CT)), gamma rays (single photon emission computed tomography (SPECT)), annihilation photons (positron emission tomography
114
Composite Nanoparticles for Cancer Imaging and Therapy
Figure 4.14 TEM images of superlattices formed through partial cation exchange. (a) The original 4.8- by 64-nm CdS nanorods. (b, c) Transformed CdS-Ag2S superlattices. (Inset) Histogram of Ag2S segment spacing (center-to-center). The average spacing is 13.8 ± 3.8 nm. The sample set for the histogram was greater than 250 spacings. (Reproduced with permission from [60].)
Table 4.2
Imaging Modalities Used in Cancer Imaging
Magnetic Resonance Imaging (MRI)
X-Ray Computed Tomography (CT)
Single Photon Emission Computed Tomography (SPECT)
Uses radiofrequency pulses to excite protons between two energy states that are created when a sample is placed in a high magnetic field and employs magnetic field gradients to encode position and produce cross-sectional images related to proton density and tissue relaxation properties.
Also known as CAT scans, in which X-rays are transmitted through the body to form tomographic images related to tissue density.
A nuclear imaging technique that utilizes a gamma camera and a collimator to record images of the distribution of radiolabeled molecules in vivo.
Positron Emission Tomography (PET) A nuclear imaging technique that measures the concentration of molecules Labeled with positron-emitting radionuclides in vivo.
Ultrasound A real-time imaging technique that measures reflections of high-frequency sound waves at tissue interfaces to provide structural information. Using the Doppler effect, ultrasound also can be used to measure blood flow in vessels.
(PET)), and high-frequency sound waves, or ultrasound, are all successfully employed to interrogate the structure and/or function of tissues [68].
4.3 Application Examples of Nanodevices
115
In some cases, endogenous contrast may be intrinsic to the body tissues (electron density for CT, proton density, and tissue relaxation times for MRI, acoustic properties of tissue for ultrasound, intrinsic optical properties of tissue for optical imaging). In many cases, contrast agents, designed to provide or augment the imaging signal, are introduced into the body. Depending on the modality, these may be radiolabeled probes (PET, SPECT), molecules laden with high Z nuclei (CT), paramagnetic agents (MRI), acoustically active microbubbles (ultrasound), or fluorescent molecules (optical imaging). By exploiting different types of radiation and using different contrast agents, an enormous variety of parameters can be imaged in vivo, ranging from basic tissue density with X-rays to specific molecular targets, gene expression, and protein-protein interactions [68]. Each imaging modality is characterized by differing resolutions on the spatial and temporal scales, and by a different sensitivity for measuring properties related to morphology or function. The use of nanomaterials as contrast agents has enabled improvements in cancer imaging by conventional imaging modalities, and has also established new techniques such as optical-based imaging for cancer detection. Magnetic nanoparticles has already demonstrated clinical efficacy in detecting liver cancer and staging lymph node metastasis noninvasively [69, 70]. Superparamagnetic nanoparticles disrupt local magnetic field gradients in tissues, causing a detectable signal void in MRI. Dextran coated iron-oxide nanoparticles administered intravenously get phagocytosed by normal macrophages of the liver and lymph and the failure of these tissues to darken after iron oxide administration identifies invading cancer cells. Direct targeting these nanoparticles to cancer cells has also been demonstrated [71, 72]. Other nanoparticle core including dendrimers, micelles, and liposomes modified with paramagnetic gadolinium have also been used for tumor targeted MRI contrast [72–74]. Gold nanoshells offer a promising alternative to MRI probes by providing contrast to optical imaging [75]. These nanoparticles are constructed from a dielectric core (silicon), and a metallic conducting shell (gold). By varying the dimension of the core and shell, the wavelength of the plasmon resonance can be altered to either absorb or scatter certain wavelengths of light from UV to infrared. Particles that are used to scatter light in the near-infrared, where tissues have minimal absorbance, have been used to enhance imaging modalities, such as reflectance confocal microscopy and optical coherence tomography (OCT) [76]. Although the penetration of optical techniques does not approach that of CT and MRI, imaging feature is possible at depth of few centimeters. Gold colloids have also been used for optical contrast agents, but these lack the inherent tunability of nanoshells. The conjugation of optical contrast agents to antibodies has been used for the molecular imaging of the EGFR receptor on early cervical precancers and for Her2+ breast carcinoma cells in mice [77]. Fluorescent nanoparticles offer another useful tool to enhance optical detection. Fluorescent semiconductor nanocrystals, quantum dots, have been used to show ligand-mediated nanoparticle targeting to distinct features in the tumor 6. Quantum dots have a distinct advantage over conventional fluorophores because of their size-tunable excitation and emission profiles, narrow bandwidths, and high
116
Composite Nanoparticles for Cancer Imaging and Therapy
photostability [78, 79]. Using nanocrystals that fluoresce in the near-infrared could extend their utility for clinical settings [80], though a key limitation has been their toxicity [81]. Efforts to make nontoxic quantum dots are ongoing. Alternative fluorescent nanoparticles probes have been developed including fluorescently tagged dendrimers and fluorophor-embedded silica nanoparticles. Nanoparticles formulations that provide contrast for other imaging modalities including ultrasound and CT have also been described. Perfluorocarbon emulsion nanoparticles composed of lipid-encapsulated perfluorocarbon liquid, about 250 nm in diameter are effective in giving echo-contrast [82]. Air trapping liposomes formulated for freeze-drying techniques have also been used to give ultrasound contrast [83, 84]. An interesting extension of the contrast agents described above is their combination into nanoparticles for multimodal imaging to integrate the strengths of two or more modalities. In this context contrast agents made of organic dendrimers selectively functionalized into homo- or hetero-nanostructured devices are presently the most promising materials. 4.3.2
Application Examples of Dendrimer Nanodevices
Dendrimers [10] have been used [85] as delivery vehicles for oligonucleotides, antisense oligonucleotides, oligonucleotide arrays, chemotherapeutic cancer drugs, and preparation of macromolecular contrast agents [86–90]. Year by year, an increasing number of biomedical applications are reported for dendrimer nanodevices. Dendrimers can be for drug delivery; (1) either by multiple covalent modifications of the termini, (which then may be exposed on the surface), or (2) initiate a physical binding in the dendrimer phase (encapsulation, encrustment, entrapment) [91, 92]. The basis of entrapment is that interior of medium (MGD) and high (HGD) generation dendrimer molecules in solutions can behave as a pseudo-phase of different polarity (“dendrimer phase”) in which a large number of solvent molecules reside [92, 93]. These nanoscopic domains can interact with guest molecules either by very strong electrostatic interactions, hydrogen bonding or hydrophobic interactions [94]. This dendrimer phase of solvated nonionic dendrimers—depending on its polar or apolar character—is able to dissolve/solubilize smaller molecules (e.g., apolar dyes or drugs) by hydrophobic interactions. Solubilization of apolar drugs is especially important for drug delivery, as an immense number of anticancer drugs have low solubility. Dendrimers with large internal dipole moments are able to solvate polar molecules [95]. Some of these examples of encapsulated drugs tried are doxorubicin [96, 97], 5-fluoroacil [98], niclosamide [99], cis-platin [100], nicotinic acid [101], and ketoprofen [102]. However, the directly used inclusion complex and the receptor specific nanodevices may have very different delivery profiles [96]. Receptor specific targeting of anticancer drugs may improve therapeutic response considerably. Targeting moieties include synthetically available molecules, such as folate [23, 24, 103], antibodies [27], and various peptides [26, 88]. Examples of covalently bound drugs are ibuprofen [104], methotrexate [24, 105], sialic acid [106], and taxol. Metal/dendrimer nanocomposites have been used for cell labeling [21] and for radiation therapy [37, 85].
4.3 Application Examples of Nanodevices
4.3.3
117
Perspectives on Biomedical Applications of Shaped Nanocrystals
The ability to integrate shape-engineered nanoparticles into biological systems is expected to have the greatest impact on nanomedicine. Shaped noble metal nanostructures, for example, have already displayed an exciting potential in nanomedicine [107–110]. Unlike highly symmetric spherical particles that exhibit a single scattering peak, anisotropic shapes of noble metals such as rods, triangular prisms, and cubes exhibit multiple scattering peaks in the visible wavelengths due to highly localized charge polarizations at corners and edges. Interesting nanostructure geometries have been designed and show exciting properties. Among these anisotropic nanostructures; nanorods [109], nanoshells [110, 111] nanocages [112], and nanostars [113] showed promising applications in imaging, diagnostic and therapy, mostly due to their anisotropic shape. By manipulating nanoparticle shape, researchers were able to tune the optical resonance to any wavelength of interest. At wavelengths just beyond the visible spectrum in the near infrared, blood and tissue are maximally transmissive. When resonances are tuned to this region of the spectrum, shaped nanostructures become useful as contrast agents in the diagnostic imaging of tumors. When illuminated, they can serve as nanoscale heat sources, photothermally inducing cell death and tumor remission [114]. Shape engineering can also enhance plasmonic sensitivity in biomolecular recognition without any need of increasing particle size. The latter often leads to plasmon band broadening due to damping and retardation, reducing the sensitivity of determining band shifts [115]. New complex nanostructures are already synthesized and currently under intense investigations [116, 117]. Next generation shape-engineered nanostructures will incorporate in one nanostructure the virtues of sharp surface curvatures, (nanoshells [111]), tips (nanorods [118], nanoprisms [119, 120]), junctions (nanoheterostructures), and cavities (nanocapsules). Growing multibranched heteronanostrcutures with multiple domains (optical, magnetic, X-ray absorbers, radioactive) and high degree of curvature, tips, and inequivalent faces/facets, on a seed formed by nanocapsule or nanocages, will yield highly sophisticated nanodevices. One trivial application of such shape-engineered nanodevices is the possibility to use them for multimodal imaging and therapy. Preferential control over their properties and functionality could be achieved through careful tailoring of the size, shape, and composition of the constituent building units, as well as engineering of their relative spatial arrangement. As a result of their inherent structural complexity and of the electronic communication that is established across adjacent material portions, these nanodevices will represent exclusive platforms on which diverse properties can coexist and eventually exchange-couple with each other, leading to unique nanodevices with diversified, enhanced, and/or deliberately switchable responses, not otherwise achievable with any of the individual components alone or with their physical mixture counterparts. Functionalizing such nanostructures with functional dendrimers and selectively coating them will lead to a functionality that can be limited only by our imagination.
118
4.4
Composite Nanoparticles for Cancer Imaging and Therapy
Summary There is an ultimate need to develop better and more efficient ways to treat diseases. Nanoscience and nanotechnology offers a number of new opportunities to develop nanomedicines that differ from traditional approaches. These novel ways will not serve as a substitute for traditional medicine, but will add tremendous value to existing methods and treatments. Complexity of nanodevices and differing cellular uptake mechanisms allows the engineering of more efficient targeted nanomedicine agents. We have demonstrated the three most important variables of nanodevice design we have to regulate: surface (size, charge), composition, and shape by using examples of synthesis, characterization, and biomedical applications of poly(amidoamine) (PAMAM) dendrimers, composite nanomaterials and shape-engineered inorganic nanocrystals.
Problems 4.1 Calculate the number of tertiary and primary amines present in the PAMAM_E5.NH2 dendrimer based on the potentiometric titration data. (HCl was used to decrease pH of the sample to the start point and the resulting solution was titrated with 0.1M NaOH solution.) The amount of dendrimer was 1.6657 mg (average molecular mass = 29,680 D). Points: Experimental data; blue line: integral by summation, red line: differential of blue line. Numbers indicate microliters used at peak values. Refer to Figure 4-p1.
Figure 4-p1
Problems
119
4.2 Calculate the degree of acetylation of PAMAM_E5.(NH2)134 dendrimer (subscript denotes number of primary amine groups present in the starting dendrimer before acetylation) and the content of acetic acid in a sample based on the 1HNMR spectrum, assuming presence of 2020 alkyl protons in the dendrimer (theoretical number of CH2 protons in an ideal dendrimer). Refer to Figure 4-p2.
Figure 4-p2
4.3 MALDI-TOF spectra indicate shift of the highest frequency fragments in the major peaks of PAMAM_E5.(NH2)119 dendrimer and its glycidolated derivative from 26,572 m/z to 32,647. Calculate the average degree of glycidolation. 4.4 Assuming that you have synthesized gold nanoparticles with perfect icosahedral shape, and you wish to functionalize their surface by an equimolar amount of dihydrolipoic acid (C8H14O2S2, FW = 206.33). Calculate the number of atoms at the surface of each colloidal particle having a total number of 309 atoms. Calculate how many shells are composing this particle, and how many atoms are in the external shell. The total number of atoms is given as: NT = (10/3)υ3 + 5υ2 + (11/3)υ + 1. Atoms in the external shell are given as: Ns = 10 υ2 + 2 where υ is the symbol of the shell number, also called order of the cluster). 4.5 Which of these assertions (a–d) are correct: (a) MRI has better soft tissue contrast than CT; (b) negatively charged particles are less toxic than positively charged particles; (c) a dispersion containing only negatively charged particles is stable; (d) PET is an in vivo nuclear imaging technique that measures the concentration of molecules labeled with positron-emitting radionuclides?
120
Composite Nanoparticles for Cancer Imaging and Therapy
References [1] ASTM E 2456-06, Terminology for Nanotechnology, www.astm.org. [2] Kagan,V. E. Bayir, H., and Shvedova, A. A., “Nanomedicine and nanotoxicology: Two sides of the same coin,” Nanomedicine: Nanotechnology, Biology, and Medicine 1, 2005, 313–316. [3] Frechet J. M. J. and Tomalia, D.A., Dendrimers and Other Dendritic Polymers, John Wiley & Sons Ltd, New York, 2002. [4] Newkome, G. R., Moorefield, C. N., and Vögtle, F, Dendritic Macromolecules: Concepts, Syntheses, Perspectives, Wiley-VCH, New York, 1996. [5] Tomalia et al., US 4,558,120 A, 4,568,737A, US 4,587,329A, 4,631,337A etc. [6] Newkome, G. R., Yao, Z, Baker, G.R., and Gupta, V. K., “Cascade molecules: A new approach to micelles, A [27]-Arborol,” J. Org. Chem., 1985; 50: 2003–2004. [7] de Brabander-van den Berg, E. M. M., Meijer, E. W., “Poly(propyleneimine) dendrimers: large-scale synthesis by hetereogeneously catalyzed hydrogenations,” Angew. Chem. Int. Ed., 1993, 32:1308–1310. [8] Grayson S. M., and Frechet, J. M. J, “Convergent dendrons and dendrimers: from synthesis to applications,” Chem. Rev., 2001, 101: 3819–3867. [9] Launay, N, Caminade, A. M., Lahana, R, et al., “A general synthetic strategy for neutral phosphorus-containing dendrimers,” Angew. Chem. Int. Ed. Eng., 1994, 33: 1589–1592. [10] L. Balogh, “Dendrimer 101,” in Biological and Biomedical Applications of Engineered Nanostructures, Warren P. Chan (ed.), U. Toronto, Eurekah, 2007, pp. 136–155. [11] Tomalia, D. A., Baker, H, Dewald, J, et al., “A new class of polymers: Starburst®-dendritic macromolecules,” P. Polym. J. (Tokyo), 1985, 17:117–132. [12] Tomalia, D. A, Baker, H, Dewald, J, et al., “Dendritic macromolecules: Synthesis of starburst dendrimers,” Macromol, 1986, 19:2466–2468. [13] Ebber, A., Vaher, M., Peterson, J., and Lopp, M., “Application of capillary zone electrophoresis to the separation and characterization of poly(amidoamine) dendrimers with an ethylenediamine core,” Journal of Chromatography A, 2002, 949:351–358. [14] Peterson, J., Allikmaa, V., Subbi, J., Pehk, T., and Lopp, M. “Structural deviations in poly(amidoamine) dendrimers: a MALDI-TOF MS analysis,” Eur Polym J, 2003, 39:33–42. [15] Banyai, I., Lesniak, W., Islam, M. T., Davis, D. Z., Baker, J., and Balogh, L., “Generational, skeletal and substitutional diversities in generation one poly(amidoamine) dendrimers,” Polymer, 2005, 46: 3022, 30–34. [16] Roberts, J. C., Bhalgat, M. K., and Zera, R. T., “Preliminary biological evaluation of polyamidoamine (PAMAM) StarburstTM dendrimers,” J. Biomed. Mater. Res., 1996, 30: 53. [17] Malik, N., Wiwattanapatapee, R., Klopsch, R., Lorenz, K., Frey, H., Weener, J. W., Meijer, E. W., Paulus, W., and Duncan, R., “Dendrimers: Relationship between structure and biocompatibility in vitro and preliminary studies on the biodistribution of 125I-labelled PAMAM dendrimers in vivo,” J. Controlled Release, 2000, 65: 133–148. [18] Malik, N., Evagorou, E., and Duncan, R. (1999) Dendrimer platinate: a novel approach to cancer chemotherapy. Anticancer Drugs 10, 767-776. [19] El-Sayed, M., Kiani, M., Naimark, M., Hikal, A. H., and Ghandehari, H. (2001) “Extravasation of poly(amidoamine) (PAMAM) dendrimers across microvascular network endothelium.” Pharm. Res.18, 23–28. [20] Nigavekar, S. S., Sung, L. Y., Llanes, M., El-Jawahri, A., Lawrence, T. S., Becker, C. W., Balogh, L., and Khan, M. K. (2004) “3H dendrimer nanoparticle organ/tumor distribution,” Pharm. Res. 21, 476–483.
References
121
[21] Lesniak, W.; Shi, X.; Bielinska, A. U.; Janczak, K. W.; Sun, K.; Baker Jr. J. R.; and Balogh, L. P. “Silver/Dendrimer Nanocomposites as Biomarkers: Fabrication, Characterization, In Vitro Toxicity and Intracellular Detection,” Nano Letters 5(11): 2123–2130, 2005. [22] Vincent J. Venditto, Celeste Aida S. Regino, and Martin W. Brechbiel, “PAMAM Dendrimer Based Macromolecules as Improved Contrast Agents,” Molecular Pharmaceutics 2005, Vol 2, No. 4 302–311. [23] E.C. Wiener, S. Konda, A. Shadron, M. Brechbiel, O. Gansow, “Targeting dendrimer-chelates to tumors and tumor cells expressing the high-affinity folate receptor,” Invest. Radiol. 32(12), 748–754 (1997). [24] Kukowska-Latallo, Jolanta F. Kimberly A. Candido, Zhengyi Cao, Shraddha S. Nigavekar, Istvan J. Majoros, Thommey P. Thomas, Lajos P. Balogh, Mohamed K. Khan and James R. Baker, Jr., “Nanoparticle Targeting of Anticancer Drug Improves Therapeutic Response in Animal Model of Human Epithelial Cancer,” Cancer Research 2005, 65, 5317–5324. [25] A.K. Patri, I.J. Majoros, and J.R. Baker, Jr., “Dendritic polymer macromolecular carriers for drug delivery,” Current Opinion in Chemical Biology 6, 466–471 (2002). [26] W.G. Lesniak, M.S.T. Kariapper, B.M. Nair, L.P. Balogh, and M.K. Khan, “Synthesis and Characterization of PAMAM Dendrimer Based Multifunctional Nanodevices for Targeting ávb3 Integrins,” Bioconjugate Chemistry, 18, 1148–1154 (2007). [27] A.K. Patri, T. Thomas, J.R. Baker, Jr., and N.H. Bander, “Antibody-Dendrimer Conjugates for Targeted Prostate Cancer Therapy,” Polymeric Materials: Science & Engineering 86, 130 (2002). [28] Luisa Quinti, Ralph Weissleder, and Ching-Hsuan Tung, “A Fluorescent Nanosensor for Apoptotic Cells,” Nano Letters, 2006; 6(3); 488–490. [29] Kwangmeyung Kim, Misu Lee, Hyungkyu Park, Jong-Ho Kim, Sungwon Kim, Hesson Chung, Kuiwon Choi, In-San Kim, Baik L. Seong, and Ick Chan Kwon, “Cell-Permeable and Biocompatible Polymeric Nanoparticles for Apoptosis Imaging,” J. Am. Chem. Soc. 2006, 128, 3490–3491. [30] Andrzej Myc, Istvan J. Majoros, Thommey P. Thomas, and James R. Baker, Jr. “Dendrimer-Based Targeted Delivery of an Apoptotic Sensor in Cancer Cells,” Biomacromolecules 2007, 8, 13–18. [31] Hong, S. A. U. Bielinska, A. Mecke, B. Keszler, J. L. Beals, X. Shi, L. Balogh, B. G. Orr, J. R. Baker Jr., and M. M. Banaszak Holl, “Interaction of Poly(amidoamine) Dendrimers with Supported Lipid Bilayers and Cells: Hole Formation and the Relation to Transport,” Bioconjugate Chem. 2004, 15, 774–782. [32] Tomalia and Balogh, US 6,664,315 B2, US 6,224,898 B1, US 5,938,934. [33] Balogh, L., Bielinska, A., Eichman J. D., Valluzzi, R., Lee, I., Baker, J. R., Lawrence, T. E., Khan, M. K. (2002) “Dendrimer nanocomposites in medicine.” Chimica Oggi (Chemistry Today) 20, 35–40. [34] Balogh, L., Nigavekar S., Cook, A., Minc, L., Khan, M. K., (2003) “Dendrimer-Gold Radioactive Nanocomposites to Treat Cancer,” Microvasculature Pharma. Chem. 2, 94–99. [35] M.J. Zohdy, C. Tse, J.Y. Ye, L.P. Balogh, T.B. Norris, M. O’Donnell, “Optical and acoustical monitoring of femtosecond laser-induced intracellular contrast agents: Initial cell culture studies,” Proceedings of the 2005 SPIE Symposium in Biomedical Optics, Vol. 5697, 291–300 (2005). [36] Khan, Mohamed K., Shraddha S. Nigavekar, Leah D. Minc, Muhammed S. T. Kariapper, Bindu M. Nair, Wojciech G. Lesniak, and Lajos P. Balogh, “In Vivo Biodistribution of Dendrimers and Dendrimer Nanocomposites—Implications for Cancer Imaging and Therapy,” Technology in Cancer Research & Treatment, 2005, 4(6), 603–613. [37] Mohamed K. Khan, Leah D. Minc, Shraddha S. Nigavekar, Muhammed S. T. Kariapper, Bindu M. Nair, Matthew Schipper, Andrew C. Cook, Wojciech G. Lesniak, Lajos P. Balogh, “Fabrication of {198Au0} Radioactive Composite Nanodevices by Radiation
122
Composite Nanoparticles for Cancer Imaging and Therapy
[38]
[39] [40]
[41] [42] [43] [44] [45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53] [54]
[55]
Polymerization and their use for nanobrachytherapy,” Nanomedicine: Nanotechnology, Biology and Medicine (2007), 4(1), 57–69. Ottaviani, M. F., Valluzzi, R. and Balogh, L. “Internal Structure of Silver-Poly(amidoamine) Dendrimer Complexes and Nanocomposites.” Macromolecules 35, 5105–5115 (2002). Seo, Y.-S. et al. “Morphology of Amphiphilic Gold/Dendrimer Nanocomposite Monolayers.” Langmuir 18, 5927–5932 (2002). Balogh, L., Valluzzi, R., Laverdure, K. S., Gido, S. P.; Hagnauer, G. L., Tomalia, D. A., . “Formation of Silver and Gold Dendrimer Nanocomposites.” J. Nanoparticle Research 1, 353–368 (1999). Esumi, K., Suzuki, A., Aihara, N., Usui, K., Torigoe, K., “Preparation of Gold Colloids with UV Irradiation Using Dendrimers as Stabilizer,” Langmuir 1998, 14(12), 3157. Garcia, M. E., Baker, L. A., Crooks, R. M., Preparation and Characterization of Dendrimer-Gold Colloid Nanocomposites, Anal. Chem., 1999, 71, 256. Esumi, K., Suzuki, A., Yamahira, A., Torigoe, K., “Role of Poly(amidoamine) Dendrimers for Preparing Nanoparticles of Gold, Platinum, and Silver,” Langmuir, 2000, 16, 2604. Gröhn, F., Bauer, B. J., Akpalu, Y. A., Jackson, C. L., Amis, E. J., Macromolecules, 2000, 33(16), 6042. D. Cozzoli, L. Manna “Synthetic Strategies to Size and Shape Controlled Nanocrystals and Nanocrystals Heterostructures,” in Bio-Applications of Nanoparticles, W. C. W. Chan (ed.), Vol. 620, pp. 1-17, 2007. T. Ould-Ely, D. Prieto-Centurion, A. Kumar, W. Guo, W. V. Knowles, S. Asokan, M. S. Wong, I. Rusakova, A. Luettge, and K. H. Whitmire, “Manganese(II) Oxide Nanohexapods: Insight into Controlling the Form of Nanocrystals,” Chemistry of Materials, Vol. 18, pp. 1821–1829, 2006. D. J. Milliron, S. M. Hughes, Y. Cui, L. Manna, J. Li, L.-W. Wang, and A. P. Alivisatos, “Colloidal nanocrystal heterostructures with linear and branched topology,” Nature (London), Vol. 430, pp. 190–195, 2004. S.-H. Choi, E.-G. Kim, and T. Hyeon, “One-Pot Synthesis of Copper-Indium Sulfide Nanocrystal Heterostructures with Acorn, Bottle, and Larva Shapes,” Journal of the American Chemical Society, Vol. 128, pp. 2520–2521, 2006. T. Teranishi, Y. Inoue, M. Nakaya, Y. Oumi, and T. Sano, “Nanoacorns: Anisotropically Phase-Segregated CoPd Sulfide Nanoparticles,” Journal of the American Chemical Society, Vol. 126, pp. 9914–9915, 2004. H. Gu, R. Zheng, X. Zhang, and B. Xu, “Facile One-Pot Synthesis of Bifunctional Heterodimers of Nanoparticles: A Conjugate of Quantum Dot and Magnetic Nanoparticles,” Journal of the American Chemical Society, Vol. 126, pp. 5664–5665, 2004. M. Chen, Y. Xie, J. Lu, Y. Xiong, S. Zhang, Y. Qian, and X. Liu, “Synthesis of rod-, twinrod-, and tetrapod-shaped CdS nanocrystals using a highly oriented solvothermal recrystallization technique,” Journal of Materials Chemistry, Vol. 12, pp. 748–753, 2002. W. Shi, H. Zeng, Y. Sahoo, Y. Ohulchanskyy Tymish, Y. Ding, L. Wang Zhong, M. Swihart, and N. Prasad Paras, “A general approach to binary and ternary hybrid nanocrystals,” Nano letters, Vol. 6, pp. 875–881, 2006. X. Peng, U. Manna, W. Yang, J. Wickham, E. Scher, A. Kadavanich, and A. P. Alivisatos, “Shape control of CdSe nanocrystals,” Nature (London), Vol. 404, pp. 59–61, 2000. L. Manna, D. J. Milliron, A. Meisel, E. C. Scher, and A. P. Alivisatos, “Controlled growth of tetrapod-branched inorganic nanocrystals,” Nature Materials, Vol. 2, pp. 382–385, 2003. T. Mokari, C. G. Sztrum, A. Salant, E. Rabani, and U. Banin, “Formation of asymmetric one-sided metal-tipped semiconductor nanocrystal dots and rods,” Nature Materials, Vol. 4, pp. 855–863, 2005.
References
123
[56] W. Cai, A. R. Hsu, Z.-B. Li, and X. Chen, “Are quantum dots ready for in vivo imaging in human subjects?,” Nanoscale Research Letters, Vol. 2, pp. 265–281, 2007. [57] H. Yu, M. Chen, P. M. Rice, S. X. Wang, R. L. White, and S. Sun, “Dumbbell-like Bifunctional Au-Fe3O4 Nanoparticles,” Nano Letters, Vol. 5, pp. 379–382, 2005. [58] H. Gu, Z. Yang, J. Gao, C. K. Chang, and B. Xu, “Heterodimers of nanoparticles: formation at a liquid-liquid interface and particle-specific surface modification by functional molecules,” Journal of the American Chemical Society, Vol. 127, pp. 34–35, 2005. [59] L. Qin, M. J. Banholzer, J. E. Millstone, and C. A. Mirkin, “Nanodisk Codes,” Nano Letters, Vol. 7, pp. 3849–3853, 2007. [60] R. D. Robinson, B. Sadtler, D. O. Demchenko, C. K. Erdonmez, L.-W. Wang, and A. P. Alivisatos, “Spontaneous Superlattice Formation in Nanorods Through Partial Cation Exchange,” Science, Vol. 317, pp. 355–358, 2007. [61] T. Zhang, J. Ge, Y. Hu, and Y. Yin, “A General Approach for Transferring Hydrophobic Nanocrystals into Water,” Nano Letters, Vol. 7, pp. 3203–3207, 2007. [62] C. Zhang, S. O’Brien, and L. Balogh, “Comparison and Stability of CdSe Nanocrystals Covered with Amphiphilic Poly(Amidoamine) Dendrimers,” Journal of Physical Chemistry B, Vol. 106, pp. 10316–10321, 2002. [63] X.-S. Wang, T. E. Dykstra, M. R. Salvador, I. Manners, G. D. Scholes, and M. A. Winnik, “Surface Passivation of Luminescent Colloidal Quantum Dots with Poly(Dimethylaminoethyl methacrylate) through a Ligand Exchange Process,” Journal of the American Chemical Society, Vol. 126, pp. 7784–7785, 2004. [64] W. Guo, J. J. Li, Y. A. Wang, and X. Peng, “Conjugation chemistry and bioapplications of semiconductor box nanocrystals prepared via dendrimer bridging,” Chemistry of Materials, Vol. 15, pp. 3125–3133, 2003. [65] T. Pellegrino, L. Manna, S. Kudera, T. Liedl, D. Koktysh, A. L. Rogach, S. Keller, J. Raedler, G. Natile, and W. J. Parak, “Hydrophobic Nanocrystals Coated with an Amphiphilic Polymer Shell: A General Route to Water Soluble Nanocrystals,” Nano Letters, Vol. 4, pp. 703–707, 2004. [66] B. Dubertret, P. Skourides, J. Norris David, V. Noireaux, H. Brivanlou Ali, and A. Libchaber, “In vivo imaging of quantum dots encapsulated in phospholipid micelles,” Science, Vol. 298, pp. 1759–62, 2002. [67] L. Medintz, H. T. Uyeda, E. R. Goldman, and H. Mattoussi, “Quantum dot bioconjugates for imaging, labeling and sensing,” Nature Materials, Vol. 4, pp. 435–446, 2005. [68] S. R. Cherry, “Multimodality in vivo imaging systems: twice the power or double the trouble?,” Annual Review of Biomedical Engineering, Vol. 8, pp. 35–62, 2006. [69] G. Harisinghani Mukesh and R. Weissleder, “Sensitive, noninvasive detection of lymph node metastases,” PLoS Medicine, Vol. 1, pp. e66, 2004. [70] D. D. Stark, R. Weissleder, G. Elizondo, P. F. Hahn, S. Saini, L. E. Todd, J. Wittenberg, and J. T. Ferrucci, “Superparamagnetic iron oxide: clinical application as a contrast agent for MR imaging of the liver,” Radiology, Vol. 168, pp. 297–301, 1988. [71] H. Choi, R. Choi Seok, R. Zhou, F. Kung Hank, and I. W. Chen, “Iron oxide nanoparticles as magnetic resonance contrast agent for tumor imaging via folate receptor-targeted delivery,” Academic Radiology, Vol. 11, pp. 996–1004, 2004. [72] Y.-M. Huh, Y.-W. Jun, H.-T. Song, S. Kim, J.-S. Choi, J.-H. Lee, S. Yoon, K.-S. Kim, J.-S. Shin, J.-S. Suh, and J. Cheon, “In Vivo Magnetic Resonance Detection of Cancer by Using Multifunctional Magnetic Nanocrystals,” Journal of the American Chemical Society, Vol. 127, pp. 12387–12391, 2005. [73] S. J. Wang, M. Brechbiel, and E. C. Wiener, “Characteristics of a New MRI Contrast Agent Prepared From Polypropyleneimine Dendrimers, Generation 2,” Investigative Radiology, Vol. 38, pp. 662–668, 2003. [74] M. Lanza Gregory, R. Abendschein Dana, X. Yu, M. Winter Patrick, K. Karukstis Kerry, J. Scott Michael, W. Fuhrhop Ralph, E. Scherrer David, and A. Wickline Samuel, “Molecu-
124
Composite Nanoparticles for Cancer Imaging and Therapy
[75]
[76]
[77] [78]
[79]
[80]
[81] [82]
[83]
[84]
[85]
[86]
[87]
[88]
[89]
[90]
lar imaging and targeted drug delivery with a novel, ligand-directed paramagnetic nanoparticle technology,” Academic Radiology, Vol. 9 Suppl 2, pp. S330–1, 2002. J. L. West and N. J. Halas, “Engineered nanomaterials for biophotonics applications: Improving sensing, imaging, and therapeutics,” Annual Review of Biomedical Engineering, Vol. 5, pp. 285–292, 4 plates, 2003. C. Loo, A. Lin, L. Hirsch, M.-H. Lee, J. Barton, N. Halas, J. West, and R. Drezek, ”Nanoshell-enabled photonics-based imaging and therapy of cancer,” Technology in Cancer Research & Treatment, Vol. 3, pp. 33–40, 2004. C. Loo, A. Lowery, N. Halas, J. West, and R. Drezek, “Immunotargeted Nanoshells for Integrated Cancer Imaging and Therapy,” Nano Letters, Vol. 5, pp. 709,015–711, 2005. W. J. Parak, D. Gerion, T. Pellegrino, D. Zanchet, C. Micheel, S. C. Williams, R. Boudreau, M. A. Le Gros, C. A. Larabell, and A. P. Alivisatos, “Biological applications of colloidal nanocrystals,” Nanotechnology, Vol. 14, pp. R15–R27, 2003. X. Gao, Y. Cui, R. M. Levenson, L. W. K. Chung, and S. Nie, “In vivo cancer targeting and imaging with semiconductor quantum dots,” Nature Biotechnology, Vol. 22, pp. 969–976, 2004. S. Kim, Y. T. Lim, E. G. Soltesz, A. M. De Grand, J. Lee, A. Nakayama, J. A. Parker, T. Mihaljevic, R. G. Laurence, D. M. Dor, L. H. Cohn, M. G. Bawendi, and J. V. Frangioni, “Near-infrared fluorescent type II quantum dots for sentinel lymph node mapping,” Nature Biotechnology, Vol. 22, pp. 93–97, 2004. M. Derfus, W. C. W. Chan, and S. N. Bhatia, “Probing the Cytotoxicity of Semiconductor Quantum Dots,” Nano Letters, Vol. 4, pp. 11–18, 2004. G. M. Lanza, K. D. Wallace, M. J. Scott, W. P. Cacheris, D. R. Abendschein, D. H. Christy, A. M. Sharkey, J. G. Miller, P. J. Gaffney, and S. A. Wickline, “A novel site-targeted ultrasonic contrast agent with broad biomedical application,” Circulation, Vol. 94, pp. 3334–3340, 1996 H. Alkan-Onyuksel, S. M. Demos, G. M. Lanza, M. J. Vonesh, M. E. Klegerman, B. J. Kane, J. Kuszak, and D. D. McPherson, “Development of Inherently Echogenic Liposomes as an Ultrasonic Contrast Agent,” Journal of Pharmaceutical Sciences, Vol. 85, pp. 486–490, 1996. S.-L. Huang, A. J. Hamilton, A. Nagaraj, S. D. Tiukinhoy, M. E. Klegerman, D. D. McPherson, and R. C. MacDonald, “Improving ultrasound reflectivity and stability of echogenic liposomal dispersions for use as targeted ultrasound contrast agents,” Journal of Pharmaceutical Sciences, Vol. 90, pp. 1917–1926, 2001. L. P. Balogh and M. K Khan, “Dendrimer Nanocomposites for Cancer Therapy,” in Nanotechnology in Cancer Therapy, M. Amiji (ed.), CRC Press (Taylor and Francis Group), Boca Raton, London, New York, 2006. Bielinska, A., Kukowska-Latallo, J. F., Johnson, J., Tomalia, D. A., and Baker, J. R., Jr. (1996) “Regulation of in vitro gene expression using antisense oligonucleotides or antisense expression plasmids transfected using starburst PAMAM dendrimers.” Nucleic Acids Res. 24, 2176–2182. Majoros, I. J., Thommey, P. T., Chandan, B. M., and Baker, J. R., Jr. (2005) “Dendrimer-Based Multifunctional Engineered Nanodevice for Cancer Therapy.” J. Med. Chem. 48, 5892–5899. Shukla, R., Thomas, T. P., Peters, J., Kotlyar, A., Myc, A., Baker, J. R., Jr. (2005) “Tumor angiogenic vasculature targeting with PAMAM dendrimer-RGD conjugates.” Chem. Commun. 5739–5749. Eichman, J., Bielinska, A., Kukowska-Latallo, J., and Baker, J. R., Jr. (2000) “The Use of PAMAM Dendrimers in the Efficient Transfer of Genetic.” Pharm. Sci. Technol. Today 3, 232–245. V. J. Venditto, Celeste Aida S. Regino, and Martin W. Brechbiel. “PAMAM Dendrimer Based Macromolecules as Improved Contrast Agents,” Molecular Pharmaceutics 2005;2(4) 302–311.
References
125
[91] Balogh, L.P., Swanson, D.R., Spindler R., et al. “Formation and characterization of dendrimer-based water soluble inorganic nanocomposites,” Proc ACS PMSE, 1997; 77:118–119. [92] Beck-Tan, N., Balogh L., Trevino, S., “Structure of metallo-organic nanocomposites produced from dendrimer complexes,” Proc ACS PMSE 1997; 77:120. [93] Balogh L., Tomalia D.A., “Poly(amidoamine) dendrimer-templated nanocomposites. 1. Synthesis of zerovalent copper nanoclusters,” J. Am. Chem. Soc. 1998; 120:7355–7356. [94] Gupta, U., Agashe H..B, Jain N.K., et al. “Dendrimers: novel polymeric nanoarchitectures for solubility enhancement,” Biomacromol, 2006; 7(3):649–658. [95] Majoral J.P., Caminade, A.M., “Dendrimers containing heteroatoms (Si, P, B, Ge, or Bi),” Chem. Rev. 1999; 99: 845–880. [96] Patri, A.K., Kukowska-Latallo, J.F., Baker Jr. J.R., “Targeted drug delivery with dendrimers: Comparison of the release kinetics of covalently conjugated drug and non-covalent drug inclusion complex,” Adv Drug Deliv Rev 2005; 57:2203–2214. [97] Muggia, F. M., “Doxorubicin-polymer conjugates: further demonstration of the concept of enhanced permeability and retention.” Clin Canc Res 1999; 5(1):7–8. [98] Bhadra, D., Bhadra, S., Jain S. et al. “A PEGylated dendritic nanoparticulate carrier of fluorouracil,” Internat J of Pharmaceutics, 2003; 257(12):111–124. [99] Devarakonda B., Hill R.A., Liebenberg W., et al. “Comparison of the aqueous solubilization of practically insoluble niclosamide by polyamidoamine (PAMAM) dendrimers and cyclodextrins, Internat J Pharmaceutics 2005; 304:193–209. [100] Malik N., Evagorou E.G., Duncan R., “Dendrimer-platinate: a novel approach to cancer chemotherapy,” Anti-Cancer Drugs 1999; 10(8):767–775 [101] Yiyun C., Tongwen X., “Solubility of nicotinic acid in polyamidoamine dendrimer solutions,” Eur J Med Chem 2005; 40:1384–1389. [102] Yiyun C., Tongwen X., Rongqiang F., “Polyamidoamine dendrimers used as solubility enhancers of ketoprofen,” Eur J Med Chem 2005; 40:1390–1393. [103] Quintana, A., Raczka, E., Piehler, L., et al. “Design and function of a dendrimer-based therapeutic nanodevice targeted to tumor cells through the folate receptor,” Pharm Res 2002; 19(9):1310–1316. [104] Kolhe, P., Khandare, J., Pillai, O., et al. “Preparation, cellular transport, and activity of polyamidoamine-based dendritic nanodevices with a high drug payload,” Biomaterials 2006; 27(4):660–669. [105] Gurdag, S., Khandare, J., Sarah Stapels, S., et al., “Activity of dendrimer-methotrexate conjugates on methotrexate-sensitive and-resistant cell lines,” Bioconjugate Chem. 2006; 17:275–283. [106] Reuter, J.D, Myc, A, Hayes, M.M., et al. “Inhibition of viral adhesion and infection by sialic-acid-conjugated dendritic polymers,” Bioconj Chem. 1999; 10:271–278. [107] Majoros, I, Myc, A, Thomas, T, et al. “PAMAM dendrimer-based multifunctional conjugate for cancer therapy: synthesis, characterization, and functionality,” Biomacromolecules 2006; 7:572–579. [108] Montejano-Carrizales J. M. and Moran-Lopez, J.L., Nanostructured Materials, Vol. 1, 1992, pp. 397–409. [109] Huang, X.; El-Sayed, I. H.; Qian, W.; El-Sayed, M. A. “Cancer Cell Imaging and Photothermal Therapy in the Near-Infrared Region by Using Gold Nanorods,” J. Am. Chem. Soc. 2006, 128, 2115–2120. [110] Loo, C., A.; Lowery, A.; Halas, N.; West, J.; Drezek, R. “Immunotargeted Nanoshells for Integrated Cancer Imaging and Therapy,” Nano Lett., 2005, 5, 709–711. [111] Hirsch, L. R.; Stafford, R. J.; Bankson, J. A.; Sershen, S. R.; Rivera, B.; Price, R. E.; Hazle, J. D.; Halas, N. J.; West, J. L. “Nanoshell-Assisted Tumor Ablation Using near Infrared Light under Magnetic Resonance Guidance,” Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 13549–13554.
126
Composite Nanoparticles for Cancer Imaging and Therapy [112] Hu, M.; Chen, J.; Li, Z.-Y.; Au, L.; Hartland, G. V.; Li, X.; Marquez, M.; Xia, Y. Gold “Nanostructures: Engineering their Plasmonic Properties for Biomedical Applications,”Chem. Soc. Rev. 2006, 35, 1084–1094. [113] C. L. Nehl, H. Liao, and J. H. Hafner, Nano Lett. 2006, Vol. 6, No. 4, pp. 683–688. [114] S. Lal, S. E. Clare, N. J. Halas, Accounts of Chemical Research, Vol. 41, No. 12, December 2008, pp. 1842–1851. [115] P. K. Jain, X. A Huang, I. H. El-sayed, and M. A. El-Sayed, Accounts of Chemical Research, Vol. 41, No. 12 December 2008 1578-1586. [116] Sihai Chen, Zhong Lin Wang, John Ballato, Stephen H. Foulger, and David L. Carroll, “Monopod, Bipod, Tripod, and Tetrapod Gold Nanocrystals,” J. Am. Chem. Soc., 2003, 125 (52), 16186–16187. [117] Jianping Xie, Qingbo Zhang, Jim Yang Lee, and Daniel I. C. Wang, “The Synthesis of SERS-Active Gold Nanoflower Tags for In Vivo Applications,” ACS Nano, 2008, 2 (12), 2473–2480. [118] Lee, K.-S.; El-Sayed, M. A., “Gold and Silver Nanoparticles in Sensing and Imaging: Sensitivity of Plasmon Response to Size, Shape, and Metal Composition,” J. Phys. Chem. B, 2006, 110: 19220–19225. [119] Haes, A. J.; Van Duyne, R. P., “A Nanoscale Optical Biosensor: Sensitivity and Selectivity of an Approach Based on the Localized Surface Plasmon Resonance Spectroscopy of Triangular Silver Nanoparticles,” J. Am. Chem. Soc., 2002, 124: 10596–10604. [120] Bielinska, J. D. Eichman, I. Lee, J. R. Baker, Jr., and L. Balogh, “Imaging {Au0-PAMAM} Gold-Dendrimer Nanocomposites in Cells,” J. Nanoparticle Research, 2002, 4: 395–403.
CHAPTER 5
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices Noy Bassik and David H. Gracias
5.1
Introduction We live in an era of miniaturization in medicine that seeks to transform the way disease is diagnosed and treated. One example of such a transformative miniaturization strategy is the development of laboratories on chips for high throughput diagnostics [1]. Here, the idea has been put forth that it should be possible to diagnose disease using a single drop of blood or serum, with rapid turnaround times. It is expected that minimally instrumented point of care devices will be widely accessible to enable testing at home and in developing nations. Another example involves the development of novel probes such as magnetic capsules, nanocrystals, and nanoparticles [2, 3] for magnetic resonance (MR), computed tomography (CT), and optical imaging. These probes seek to detect diseases such as cancer at the cellular level, thereby enabling an early diagnosis [4]. At the present time, cancer is normally detected only after the cells have multiplied to form larger clinically detectable tumors. Drug delivery has also seen a huge burst of research in the development of “smart” pills, and the idea of a “pharmacy on a chip” has been put forward [5]. The dream of scientists in the field is to create pills that are multifunctional, with on-board sensors and actuators, for precise, autonomous dosing of therapeutic drugs. Moreover, conventional systemic chemotherapies, currently used to treat localized diseases such as solid tumors, are expected to be replaced by spatially localized drug delivery. This local targeting of the powerful drugs specifically to the diseased site will minimize side effects [6–10]. Finally, there is the possibility that miniaturization could lead to the development of minimally invasive, microscale surgical tools. The miniaturization vision described above hinges on the ability to construct tiny replicas of macroscale instruments and devices, as well as new ones with novel functionality. Researchers are developing strategies to fabricate small structures with sizes ranging from tens of nanometers (on the range of the size of protein molecules) to tens of microns (cell dimensions). This scale is of special relevance to medicine; a number of pathogens such as viruses and bacteria have sizes within this range. The challenge however is that it is not easy to fabricate devices on such a small scale. Engineering operations that seem trivial on the macroscale such as cut-
127
128
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices
ting, sawing, welding, soldering, grinding, and molding are extremely challenging to miniaturize. Moreover, structures with sizes below 100 µm cannot be seen easily with the naked eye; tools such as scissors, knives, drill bits, and saws are hard to come by with these dimensional constraints. Furthermore, how could one hold and utilize miniaturized tools such as scissors, whose sizes are a fraction of the width of a human hair? It is true that pick-and-place robotic tools can enable some of these functions. However, new challenges arise for cost-effective large scale manufacturing and manipulation at these small sizes, and especially in three dimensions (3-D). In the early 1960s, engineers developed a scheme for fabricating small structures using optical lithography [11]. Lithography, or the art of printing using chemical processes such as etching on surfaces, has been around since the late 1700s. Engineers devised a strategy to project an image with microscale features onto an optically sensitive polymeric material whose solubility was altered on exposure to light. This technology has been widely used to enable the era of very large scale integration (VLSI) of electronic devices [12]; semiconductor chips such as logic processors or flash memory devices are fabricated using this process. Optical lithography has also enabled the fabrication of microelectromechanical systems (MEMS) [13] such as resonators, gyroscopes, and microbatteries. Although incredibly successful, the lithographic microfabrication paradigm is limited in that it is inherently a two-dimensional (2-D) process [14]. The image to be transferred is projected onto a 2-D surface like the projection of a transparency slide onto a screen. Lithography can be used to fabricate quasi-3-D multilayer devices by sequentially patterning one layer at a time. As an example, all 8 to 12 layers of the insulated wiring in microprocessors are lithographically fabricated, one after another. This strategy of patterning multilayers sequentially works well when the number of layers are small (<10), but becomes increasingly difficult and cost-prohibitive as the numbers of layers increase (>10). Hence, several alternate strategies have been suggested to address this challenge of constructing more 3-D-like devices; these include holographic 3-D projection, stereotactic machining using laser, ion, or electron beams and molding. Some of these strategies work in limited applications while many of them are cost-prohibitive in mass production. On the whole, it is fair to say that 3-D fabrication is still very challenging on the micro- and nanoscale. One could ask why do we even need micro- and nanofabrication in 3-D? The immediate answer that comes to mind is that humans live in a 3-D macroworld, and if we are to envision miniaturized analogs of 3-D macrodevices then we would need the ability to fabricate in 3-D on the micro- and nanoscale. A simple thought experiment is to imagine life in a 2-D flatland, or one in which only simple 3-D shapes such as spheres, cylinders, and cubes could be fabricated. Imagine also that these 3-D objects could only be fabricated with very limited surface patterns. This world would be severely limiting. We also know that for applications in a 3-D environment, devices that approximate a 3-D profile are better suited than planar objects. Two-dimensional devices have limited interaction with the surrounding medium through only one or at most two interfaces. Three-dimensional devices also have a greater surface area to volume ratio, and may be able to mimic the high surface areas observed in critical organs such as the kidneys and the lungs. A greater surface area to volume ratio inherent in 3-D devices permits more usable surface, while at the
5.1 Introduction
129
same time allowing for small form factors. This small size may be especially critical in medicine, to permit introduction into the body with minimally invasive methods. Hence, a shift in engineering strategies is required on the small scale to allow for more versatile 3-D fabrication and patterning. One place to look for inspiration is towards biological fabrication. Nature builds miniaturized structures, devices, and automata with exquisite precision, and in a highly parallel and cost-effective manner, using a self-assembly approach [15–17]. Viruses provide a wonderful inspiration of how micro- and nanomedicine can achieve therapeutics or diagnostics at a single cell level. These mass produced, self-assembled biological machines can find specific cells and inject their genetic cargo through cell membranes. However, the self-assembling manufacturing paradigm has been largely unexplored in human engineering since the process is generally perceived to be indeterministic and uncontrollable. As an example of the self-assembling paradigm that is ubiquitous in nature, let us examine a simple sodium chloride salt crystal (Figure 5.1). In what is perhaps the ultimate subnanoscale engineering feat, sodium and chlorine ions floating randomly in solution arrange themselves alternatively in a three-dimensional pattern with subnanometer scale precision. How did the ions get to their precise positions? In attempting to build such a structure, our conventional human engineering approach would be to pick each ion, one at a time, and place it precisely in its specified 3-D location, an impossible task! We however can easily rationalize that “smart” components like the sodium ions with their one positive charge and the chlorine ions with their one negative charge would form this three-dimensional construct spontaneously out of a saturated salt solution, due to the delicate balance of coulomb forces. This self-assembly occurs with the aid of thermal energy that allows the components to overcome weak energy minima and coalesce into a global energy minimum (i.e., that of a perfectly arranged lattice). Sodium chloride crystals have defects, but nature has learned to live with them, and given enough time these defects would also be eradicated. In reality, one could state that the sodium chloride crystal builds itself or self-assembles.
Figure 5.1 A 3-D schematic diagram of a salt crystal. (Illustration prepared by Aasiyeh Zarafshar and Anum Azam in the Gracias Laboratory.)
130
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices
Human engineering is now attempting to embrace this self-assembling paradigm. For example, a search of the term “self-assembly” in a database for scientific publications (SciFinder Scholar®) reveals approximately 7,300 publications in 2008 as opposed to only 1,300 in 1998 and 90 in 1988. Self-assembly relies on the fabrication of “smart” components that are capable of interacting with each other under prespecified conditions to spontaneously form complex 3-D structures. The idea is that components patterned with interactions will find their global energy minimum, at equilibrium, if they are agitated with sufficient energy to overcome weak metastable binding events. Ordered structures can also form out of equilibrium, but only when energy is dissipated. These assemblies are usually much harder to design, but are ubiquitous in living structures. One discerning feature of these nonequilibrium assemblies is that the order or assembly falls apart if energy is not continuously provided. A salt crystal is an example of a structure formed by equilibrium self-assembly whereas a bacterium is one formed out of equilibrium. As long as the interactions and the energy of agitation are tuned, one can in principle, as nature does, scale this engineering concept across length scales from the nanometer to the micrometer and far beyond—after all, galaxies too are self-assembled. In order to make the kinetics of self-assembly practical in the laboratory, the system may be given a head start on the assembly [18]; an example would be tethering different amino acids together in a protein before assembling them into a 3-D configuration. When attached in a linear fashion, the number of possible 3-D conformations is limited, assembly is expedited and yield and defect tolerance are increased. This chapter discusses the fabrication of complex 3-D self-assembled devices with a focus on applications in medicine. The components for the devices are lithographically fabricated; self-assembly is an additional step that transforms them into 3-D devices. Our strategy for fabricating miniaturized medical devices is based on lithographic fabrication strategies developed in the microelectronics industry. These strategies allow multilayer construction and highly precise patterning over large areas with submicron scale resolution. However, as current lithographic fabrication processes are inherently two-dimensional (2-D), this two-dimensionality is especially limiting for the construction of biomedical devices for the 3-D human body. Therefore, this chapter surveys different solution paths that address the challenge of fabricating complex lithographically patterned 3-D biomedical devices. We also discuss the need for such devices in specific 3-D biomedical applications as well as future challenges.
5.2
Basics of Lithographic Fabrication Lithographic micro- and nanofabrication has been the tour de force behind the incredible success of the microelectronics industry. Here, lithographic fabrication enables wafer-level processing, which allows large numbers of devices to be fabricated in a cost-effective and highly parallel manner. Microprocessor chips such as the Pentium are fabricated en masse on large 12-inch-wide silicon wafer substrates. Although the wafers are subjected to over a hundred sequential chemical processes, individual chips from different parts of the wafers have similar device properties; an incredible engineering feat, considering that some of the features in these devices have dimensions on the order of nanometers.
5.2 Basics of Lithographic Fabrication
131
We live in an era where any pattern that can be drawn on a piece of paper can be fabricated lithographically in a wide range of materials. Briefly, the process of fabricating a miniature object is reduced to breaking it up into a number of thin film layers; each layer is patterned with registry to underlying layers. The patterns are laid out using a software program such as AutoCAD. The patterns are then transferred to a photosensitive polymer (called a photoresist), which is then used as a mask to either add or remove material from constituent thin films (Figure 5.2). The key steps, including ones relevant to biomedical engineering, are described in Section 5 in the supplementary materials and solution manual. The methods described in the supplementary section represent the most important processes utilized to fabricate most structures in 2-D. Once a 2-D pattern is generated, the process is repeated to build quasi-3-D structures. Lithographic fabrication is arguably one of the greatest inventions of human engineering. Its numerous advantages for constructing miniaturized devices include (a) precise,
(a)
(b)
(d)
(c)
Figure 5.2 Schematic diagram of the basic concept behind optical lithography with associated process flow steps. Clockwise: (a) In thermal evaporation, a boat filled with metal is heated until the metal boils, and a thin film condenses on the substrate above. (b) Photoresist, a photosensitive polymer, is deposited on the substrate and spun at high speed to generate an even film. (c) An ultraviolet source illuminates a mask with transparent and opaque regions. Light passes only through the transparent regions and irradiates the photoresist selectively. (d) After developing, only certain regions of the substrate are protected by photoresist. The substrate is now placed into a wet etchant that removes exposed metal film deposited earlier in step (a). (Illustration prepared by Anum Azam and Aasiyeh Zarafshar in the Gracias Laboratory.)
132
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices
high-resolution patterning on the micro- and nanoscale, (b) cost-effective and highly parallel fabrication due to a wafer scale methodology, and (c) versatile processing that is compatible with a wide range of materials. The fabrication paradigm is widely used to fabricate nano-micro scale structures, but due to the inherent two-dimensionality of the patterning process, the structures formed are quasi-3-D.
5.3
The Need for Three-Dimensional Biomedical Devices The human body is a complex three-dimensional machine that is exquisitely structured and functions on length scales ranging from the subnanometer to the meter scale. While we are a long way off from constructing an entire organism, engineers are trying to mimic architectures in the body and the processes used to fabricate them. At the tissue or organ level, there is a need to control the structural, physical, and chemical environments in which cells multiply. The transition from 2-D to 3-D tissue constructs can be engineered in several ways: folded or branched pathways, parallel networks, and nested structures. Brain tissue is a highly connected network that achieves complexity via layering and folding. The key to its performance is a connected set of neurons organized in precise spatial configurations [19]. The cerebral cortex has several well-defined layers with different neuronal populations that span 3-D space, generating cellular networks with very high levels of interconnectivity (Figure 5.3(a)). For example, it is estimated that the 25 billion neurons in the neocortex (the portion of the brain that separates us from other species) are connected on an average to about 7,000 other cells [20], allowing a staggeringly large number of interconnections. These large numbers of interconnections are critical to achieving complex function and defect tolerance (such as a recovery from brain injury or stroke). Folding of cellular layers also features prominently in the heart. Specific cellular regions must bend and involute in appropriate patterns, and at precise time periods during development, to form the complete asymmetric four chambered heart with valves and conduits. Many of the mechanisms used in the heart are common to limb development, cell division, and patterning across the organism [21, 22]. Many tissues also utilize a 3-D architecture to pack a maximum surface area into a minimum volume. The lungs function as a gas exchange medium and the kidneys as a filtration device, both enabled by their finely divided branches and networks. The similarities of these biological architectures to ion exchange media, porous catalysts, and other high surface area to volume structures used in engineering [23] are striking. As opposed to these 3-D engineered constructs, the lungs and kidneys also need to grow and enlarge while maintaining this 3-D architecture (Figure 5.3(b)). This growth occurs in a dendritic fashion (splitting tubes), solving the problem of transportation of molecules from macroscopic conduits to tiny micron-sized vessels. Bones consist of vessels and marrow within a strong mineralized supportive tissue. This architecture allows a strong skeletal support and facilitates regenerative functions such as blood cell production. Each mature bone cell or osteocyte lives locked in a framework of minerals. Here, the body utilizes the third dimension to minimize space and maximize strength and connectivity—a bone is stronger if the
5.3 The Need for Three-Dimensional Biomedical Devices
133
(c)
(a)
Liver lobule
Neuron
(b)
Lung alveoli Figure 5.3 (Color plate 6) The 3-D aspects of human microanatomy. (a) Reconstructed 3-D representation of a neuron. Neuronal morphologies vary widely, with many branching patterns of dendtrites and axons allowing for complex interconnections. Neurons are the longest cells in the body, often reaching 1-2 meters in length. (Reprinted from Wearne, S. L., et al., “New Techniques for Imaging, Digitization and Analysis of Three-Dimensional Neural Morphology on Multiple Scales,” Neuroscience, Vol. 136, pp. 661–680, copyright 2005, with permission from Elsevier.) (b) Anatomy of the human lung. Branching conduits subdivide airflow to end in a specialized area of exchange called the alveolus. A 3-D structure is vital to ensure a high surface area/volume ratio for gas exchange. (Illustration by Patrick J. Lynch and C. Carl Jaffe via Creative Commons Attribution 2.5 License 2006.) (c) Anatomy of the human liver. Each liver lobule mixes nutrient-rich blood from the digestive system with oxygen-rich arterial blood, separates out toxins and filters nutrients, while also producing bile for use in digesting food. A well-defined 3-D architecture involving separate blood and bile handling faces for each liver cell connects all lobules in a hexagonal packed array. (From Cunningham, C. C., and Van Horn, C .G., “Energy Availability and Alcohol-Related Liver Pathology,” Alcohol Research & Health, Vol. 27, No. 4, 2003, pp. 281–299, via National Institutes of Health.)
same amount of mineral mass is distributed in a cylindrical profile to bear the compressive load (imagine how hard it is to axially compress a paper towel roll); the internal space may be used for other tasks. The liver, which among its many metabolic functions filters all blood emerging from the digestive tract, uses a parallel network of lobules. A unique three-dimensional layout allows deoxygenated blood to mix with oxygenated blood in contact with a single interface of liver cells (hepatocytes). The reverse cell interfaces secrete bile into a separate collection of vessels that connect to the digestive system. Therefore, the liver contains a parallel set of plumbing that is continuous from the level of an individual cell to a centimeter-scale vein or duct, never allowing blood and bile to mix (Figure 5.3(c)) [24].
134
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices
All of the above organs serve as examples of the kinds of 3-D biomedical devices that will be needed to replace, augment, or emulate biological function. While these are not the only paradigms that could be used to engineer biomedical devices, they serve as an excellent inspiration. Additionally, if we wish to harness the power of living cells to perform filtration, metabolic functions, or self-repair, we will either need to engineer the cells to function in new physical environments, or alternatively design environments well suited to the needs of different cell types.
5.4
Present Day Lithographically Structured Biomedical Devices On the macroscale, current 3-D devices include everyday items such as gauze sponges to complex technological wonders such as artificial hearts and brain implants. In each of these, the 3-D structure is critical to its operation, from absorbing liquids to pumping blood to interacting with large numbers of neurons. Many biomedical devices have a 2-D topography in a 3-D housing. Devices such as dialysis membranes interface between two different environments to allow for nutrient/waste exchange or purification and are inherently two-dimensional. These membranes operate on the principle of size or ion selective exclusion and therefore need a 2-D barrier with differential pressures, concentrations, or ion gradients to assist in their operation. Often, the key is packaging—squeezing a surface area the size of a tennis court into a volume of 5L, such as in the human lungs. Any artificial device with the same performance will therefore need a very high surface area to volume ratio. As with any technological innovation, device construction is also a major factor in shifting from laboratory research to commercially available products. Several tissue engineering reactors have been built as one-of-a-kind lab instruments [25, 26]. Implantable sensing devices have also been constructed with macroscale serial processing [27], but these are not easily mass producible. Recently, several serial lithographic approaches have been developed to increase the three-dimensionality of lithographically fabricated structures and devices. One approach involves wafer stacking [28, 29] using dielectric [30] and metal bonding [31]. Wafer stacking has disadvantages, however; the process is a serial one and sequential⎯this greatly increases costs and limits the number of layers that can be stacked. In the complementary manufacturing arena of MEMS and nanoelectromechanical systems (NEMS) most of the techniques for fabricating 3-D structures, such as electroforming [32], micromachining [33], and stereolithography [34] include layer-by-layer serial fabrication. For example, polysilicon micromachining, which is widely used at multiuser facilities such as the Cronos foundry [35] and Sandia National Laboratory [36], exploits differences in properties of deposited materials to form quasi-3-D structures, and the cycle of thin film deposition, pattern transfer, etching, and electrodeposition of each material is repeated to build multilayer structures. It is impossible, however, to fabricate fully 3-D structures in a layer-by-layer manner since registry with underlying layers is challenging, and the serial manufacturing process is cost-prohibitive. Some of the quasi-3-D biomedical devices developed to date are discussed below.
5.4 Present Day Lithographically Structured Biomedical Devices
5.4.1
135
Drug Delivery Devices
Drug delivery involves the science of transporting specific therapeutic molecules from the outside world into the patient. The oldest (and most straightforward) example of delivering a drug is by ingestion. Here, the specialized absorption system that is our gut takes over, breaking up the ingested object and absorbing therapeutic molecules. The harsh nature of the digestive pathway, however, is not compatible with several drugs, especially proteins [37]. Therefore drugs such as antibodies must be injected or infused into the bloodstream directly, greatly raising costs and limiting their use [38]. Implantable drug-release devices utilize materials that facilitate the controlled release of a drug over specified predetermined periods of time. In these cases, the drug is usually noningestible and its therapeutic ratio or mechanism of action necessitates its delivery over longer periods of time. Recent drug release devices utilize precisely engineered polymeric matrices that allow high control over the dosing profiles [39]. Spatial control over drug distribution has also been suggested as a means of controlling cell development and growth [40, 41]. Temporal control has been achieved by mixing polymers of differing molecular weights; limited spatial control can be facilitated by introducing these polymers at specified locations [42, 43]. Recently, it has also been demonstrated that spatiotemporal control can be achieved by microelectronic control over drug delivery devices [40, 43–46]. New drug delivery mechanisms seek to harness the small size scales and parallel nature of lithography to achieve predictable and controllable dosing characteristics [44, 47, 48]. One such demonstration is the microwell array-based “pharmacy-on-a-chip” approach being developed by MicroCHIPS. Here, many small wells are lithographically patterned on a pacemaker-sized device, and the lids on these wells can be dissolved either individually or all at once, using on-board electronic circuitry (Figure 5.4(a)). Limitations of this design, however, include the large size of the overall chips that require surgical implantation and the small volume of each well, limiting delivery only to potent agents such as peptides [49]. Other lithographic approaches have been used to transport drugs from a macroscopic reservoir through microfabricated needles. Using anisotropic etching, engineers have fabricated microneedles that penetrate the skin’s natural barrier layer (the stratum corneum), without causing trauma to tissue below (Figure 5.4(b)). These microneedles facilitate pain-free injections since they penetrate only deep enough to deliver medication without affecting nerve cells below [50]. 5.4.2
Structural Devices
Structural biomedical devices seek to replace load-bearing components of the body that have been damaged or removed due to disease or trauma. Specialty coatings are used on dental and orthopedic implants, and these are often applied via plasma, oxidation, or sputtering. It has been proposed that surface roughening aids in bone implant retention and toughness; this roughening has been demonstrated on the macroscale using methods such as laser micromachining [51]. However, there are major challenges [52] that need to be overcome in using essentially 2-D techniques
136
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices
(a)
(b)
(c)
Figure 5.4 Lithographic devices used for drug delivery. (a) Electronic programmable reservoir array that may be implanted and used to deliver medications at prespecified times. Lithography is used to fabricate multiple wells connected by circuitry; the wells are filled with drug and then sealed. When signaled, the onboard electronics dissolve the capping layer and release the drug. (From Maloney, J. M., et al., “Electrothermally Activated Microchips for Implantable Drug Delivery and Biosensing,” Journal of Controlled Release, Vol. 109, pp. 244–255, copyright 2005, with permission from Elsevier.) (b) Microneedles are produced by selective etching of crystalline materials in different directions (left). Subsequent molding steps allow for fabrication of microneedles from polymeric materials. The needles penetrate only the first few layers of skin (right), allowing for delivery of medication under the barrier layers with no pain. (From Park J., Allen, M. G., and Prausnitz, M. R., “Biodegradable Polymer Microneedles: Fabrication, Mechanics and Transdermal Drug Delivery,” Journal of Controlled Release, Vol. 104, pp. 51–66, copyright 2005, with permission from Elsevier.) (c) Stent hardware inside a vessel with directional release. Microfabricated reservoirs in the internal and external surfaces of the stent allow for one drug’s release into the vessel sidewall to prevent growth and occlusion, and another drug to be directed into the vasculature. (Illustration created by Noy Bassik in the Gracias Laboratory.)
to pattern these 3-D surfaces on the microscale. The most successful bone models to date have utilized polymeric foams [53] or metal etching [54] instead to create these 3-D patterns. Stents are used to prop open vessels that have lost the structural strength needed to keep them open. Stents are also used to keep vessels with blockages such as coronary arteries open after plaque removal. Many stents are made from exotic metal alloys (nitinol or superelastic copper alloys) via macroscopic processes such as laser
5.4 Present Day Lithographically Structured Biomedical Devices
137
machining; some utilize lithographically patterned features [55]. Lithography has allowed for directional drug release in stents (Figure 5.4(c)). A common problem after using stents to open cardiac vessels is reocclusion of the artery with scar tissue, months after stent placement [56]. Drugs can now be delivered selectively into the artery wall to inhibit growth [57]. These drugs are somewhat toxic in the bulk and would not be effective if released systemically into the blood stream at large. 5.4.3
Implantable Organic /Electronic Devices
A goal of modern biomedical devices is to harness the advances of the microchip revolution and apply sensing, communication, and logic capabilities to medical devices. Lithographic devices can readily be integrated with electronic modules. Portable pacemakers and small hearing aids are only possible due to circuit miniaturization. The core of their processing and decision logic has been miniaturized from the size of suitcases or fanny packs that patients carried in the past to the size of a dime [58, 59]. Following the widespread use of pacemakers and defibrillators to control heart rhythm, similar techniques have been used to deliver precise electrical energy signals to parts of the brain as a treatment for seizures and depression [60]. Additionally, flexible polyimide films with thin metal wiring can stimulate different anatomic locations, allowing for more specific tuning of a brain or cochlear implant and reducing the need for reoperation when adjustments are required (Figure 5.5(a) [60]. Sensors are envisioned as a future use of onboard electronics, as it often takes processing to interpret the raw signal from an electrochemical sensor. Logical processing can allow for autonomous control of insulin delivery. For example, a stent-based glucose sensor has been developed that allows for measurements inside a vessel (Figure 5.5(b)) [27]. This device was lithographically patterned in 2-D and folded into 3-D manually. 5.4.4
Microfluidic Devices for Diagnosis and Cell Growth
Photolithography has also been used to construct microfluidic devices to manipulate chemistry on a small scale [61–64]. These devices have allowed for a reduction in sample volumes necessary for diagnostics and are especially relevant for the testing with the tiny blood volumes available from neonates (several milliliters) or lab animals (150 μL from a mouse) [65] and in high throughput and array-based screening. Additionally, engineers have succeeded in organizing cells into in vivo like cultures within these microfluidic devices, allowing drug screening in a more realistic and cost-effective manner [66]. Due to laminar flow of liquids in microfluidic devices, varying concentrations of reagents can even be applied to different regions of the same cell [64]. To date, microfluidics devices with quasi-3-D architectures have used layered 2-D channels joined with vertical pipes called vias that were manually aligned [67–69]. Cylinders have been created via microfluidic techniques [70–73] to address the large clinical problem, especially in tissue and organ transplants, of getting a
138
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices
Figure 5.5 (Color plate 7) 3-D biomedical devices with integrated electronics. (a) Implantable neural electrode. Microfabrication has allowed for miniaturization of electronics to allow for on-board processing and decision making in implanted devices. In this case, a manually assembled electrode patterned via lithography can be used to interact with a larger portion of brain tissue. (From Stieglitz, T., Schuetter, M., and Koch, K.P., “Implantable Biomedical Microsystems for Neural Prostheses,” IEEE Engineering in Medicine & Biology Magazine, Vol. 24, pp. 58–65, copyright 2005, with permission from IEEE.) (b) Helical glucose sensor uses microfabrication in 2-D to pattern areas with an enzyme that allows electrical detection of glucose. The sensor is manually rolled into a 3-D shape. (From [27]. Used with permission.)
proper blood supply to the tissue [74]. During transplant or surgery, vessels must be torn or cut, and new vessels will not grow into tissue in a short enough time to ensure graft viability. Compared to other geometries, a cylinder is straightforward and several sacrificial and rolling manufacturing techniques have been used [71, 75–77]. More on the use of microfabrication for tissue scaffolds can be found below. 5.4.5
Soft and Wet 3-D Devices
Biomedical devices are also constructed using “soft and wet” materials such as hydrogels [39, 78, 79]. Recently, advances in photopatterning polymers have allowed for construction of small hydrogel microparticles for drug delivery, including claws that grip the small intestine for targeted delivery [80]. Some hydrogels have interesting transition properties, allowing for smart behavior under varying
5.4 Present Day Lithographically Structured Biomedical Devices
139
temperature or pH [78]. Cells have also been reversibly patterned by functionalized photopatterened polymers [81]. The field of hydrogel polymer lithography is immature as compared to the remainder of photolithography and microfabrication. As more experience is gained with patterning of soft, smart materials, these attributes will be grafted onto electronic modules such as sensors and actuators. For example, it has been demonstrated that growth factors can be embedded in the photocrosslinkable hydrogel vehicle to modify interactions with tissue [82]. 5.4.6
Tissue Scaffolds—Growing Live Devices
A large promise of microfabrication has been the development of cell scaffolds to create living functional structures. These scaffolds afford the ability to place cells in precise spatial relationships while nourishing them to achieve a desired state of health and interdependence. Tissue scaffolds seek to provide a framework for cells to grow and organize into properly functioning tissue. Rather than growing cells on a simple flat glass or plastic Petri dish, more realistic models of cell growth and interactions will allow for clinically relevant cell culture. A Holy Grail of modern research is replacement of the laboratory animal with a nonsentient experimental apparatus [83]. An ideal scaffold will replicate the final 3-D geometry of a desired organ, which may represent branches, folds, and void spaces. Growing live functional tissue is still a far-off goal, and most efforts to date have focused on exploring cell behavior in artificial environments. Tissue scaffolds have been constructed using lithographic techniques and reveal much about the behavior of cultured cells [63]. Many of these have been adapted microfluidic technologies and resulted in quasi-3-D approaches [23, 26]. Nonetheless they have revealed dependence of cell behavior on microtopography, allowed for examinations of cell growth on varying surfaces and shapes, and allowed coculture of different cell types to investigate extracellular signaling [61, 63, 84]. Adhesion and traction forces of individual cells have been measured via microfabricated techniques [85, 86]. In the absence of true control over the intrinsic developmental pathway that begins with an embryo, scientists have templated cells onto existing macrostructures in order to cue development. For example, a decellularized heart from a donor animal was used as a scaffold for seeding embryonic heart cells, which then grew into physiologically normal shape and form, and was able to pump at several percent of normal output [87]. Clinical success has already been achieved with artificial bladders constructed using the patient’s own cells seeded on a collagen and polymer matrix that was then implanted into the body [88]. Further developments will undoubtedly use engineered 3-D scaffolds, possibly made though lithographic fabrication. 5.4.7
Interactions with Body Components
As seen above, materials used in microfabrication are varied and interact with the body in many ways. In many cases it may be desired to tailor interaction with the body by functionalizing surfaces or modifying materials themselves to control
140
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices
adhesion and differentiation of cells [43, 63]. Understanding and tailoring these interactions is critical to safety and performance. Under normal conditions, foreign bodies entering an organism are invaders and must be destroyed or walled off. The immune system is programmed to attack these foreign elements or surround them with fibrous tissue [89]. It is important to consider immune responses when designing an implantable device, as a device surrounded by a fibrous capsule will perform differently than intended. Some devices may be biodegradable over decades, such as nickel alloy implants once used in orthopedic and dental applications [90]. A range of materials used today such as metal alloys and ultrahigh molecular weight polyethylene are functionally permanent over a human lifespan, but are subject to mechanical wear in high load environments such as joints [91]. An ideal passive device will tend to be inert or invisible to the body. Three-dimensional microfabrication of biomedical devices can learn from drug delivery, where tuning interaction parameters has allowed for novel functions. For use in drug delivery, many biodegradable polymers have been developed that break down over days to months [79]. These are tailored with specific functional groups that are hydrolyzed when in contact with water or specific groups that degrade due to enzymes in the body.
5.5 Combination of Lithography and Self-Assembly to Construct 3-D Devices The devices discussed above are quasi-3-D or they are fabricated in a serial manner in a layer-by-layer methodology. Recently, there has been an attempt to utilize lithography to pattern “smart” components in 2-D that self-assemble into a 3-D structure when released from the substrate or triggered by heating [92]. Basically, the strategy is similar to hands-free origami. Structures are fabricated with rigid segments separated by hinge joints. In order to create smart structures that self-assemble via folding, the emphasis is reduced to creating smart hinge joints (Figure 5.6(a)). 5.5.1
Three-Dimensional Self-Assembled Containers
Inspired by surface tension based microstructure fabrication that was developed to rotate micromirrors and mechanical structures tethered to substrates [93], we developed a strategy to fabricate free-standing, mobile containers for cell encapsulation therapy and drug delivery [92, 94]. Here, the hinge was made of a material that can be liquefied. We utilized solders that have high surface tensions as the hinge materials. The polyhedral containers were first laid out in the form of 2-D patterns of frames and hinges. For example, 11 different 2-D patterns of six square faces (connected by hinges) can be folded (without tearing) into a cube. We utilized one of these patterns (a cruciform); and fabricated the 2-D template using conventional photolithography. We used two layers of photopatterning, one to fabricate the six square faces, and another one to construct solder hinges that were placed in between the six faces [92, 95]. The templates were fabricated on a wafer with a sacrificial thin
5.5 Combination of Lithography and Self-Assembly to Construct 3-D Devices
141
Figure 5.6 Combination of lithography and self-assembly to construct 3-D devices. (a) A schematic diagram showing the basic concept behind the self-assembly of a structure by self-folding. The structure is composed of rigid frames and a smart hinge. (Illustration created by Paul Lester in the Gracias Laboratory.) (b) Schematic of the wafer scale self-assembly of lithographically patterned cruciforms with solder hinges into hollow cubes. (Illustration captured from an animation created by David Filipiak in the Gracias Laboratory.)(c) Remote release of a chemical from a container by heating an encapsulated gel soaked with the chemical. (Images were obtained by Timothy Leong, Daniel Slanac and Hongke Ye in the Gracias Laboratory)
film sandwiched between the substrate and the template. After fabrication, the sacrificial layer was dissolved to release the template. On heating, above the melting point of the hinges, the templates folded spontaneously into hollow cubic containers (Figure 5.6(b)). The driving force for folding was derived from the minimization of the interfacial free energy of the liquid solder at the hinges. The hinges however were “smart” in that when designed with the right geometry (e.g., height, width, and depth), the hinge caused adjacent features of the template to fold precisely at right angles. Hinges at the periphery of the template helped lock the faces on folding and correct defects. On cooling, the containers became rigid as the solder solidified. Although solders are toxic, the containers were coated with inert materials such as gold or platinum after assembly, using electrodeposition, to render them bioinert. We have fabricated containers with volumes in the nano- to attoliter range and have explored applications of the containers in cell encapsulation therapy and drug delivery. Presently, we are fabricating such self-assembling containers with other
142
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices
polymeric and gel based materials. The real highlight of these containers is that size, shape, and wall porosity are precisely structured in all three dimensions. Three-Dimensional Self-Assembled Containers for Cell Encapsulation Therapy
Cell encapsulation therapy (CET) provides an attractive means to transplant cells (allo- or xenotransplantation) without the need for immunosuppression. Typically, the cell encapsulant protects the cells from immune rejection by surrounding them with an artificial, semipermeable nanoporous size exclusion membrane that allows selective permeation of nutrients and therapeutic molecules to and from cells while preventing elements of the immune system from attacking the encapsulated cells (Figure 5.7) [89, 96–98]. Despite considerable interest and several clinical trials, the technology is limited by a range of challenges including a lack of reproducibility; the inability to fabricate uniform capsules in terms of shape, size, morphology, and porosity; biofouling of implanted encapsulants due to tortuous porosity; the lack of chemical and mechanical stability of the encapsulants; and the inability to image transplanted cells to monitor efficacy. The result is that progress in the field has not lived up to expectations. We have fabricated containers with nanoporous surfaces for immunoisolation. In order to provide proper immunoisolation, the surface of the containers must have a pore size smaller than all potentially immunotoxic substances (typically below 25 nm). As a result, small molecules such as insulin and oxygen can diffuse freely between the inside and the outside of the container, while larger elements of the immune system such as leukocytes and immunoglobulins are blocked from entering the containers.
Figure 5.7 Lithographically fabricated capsule for immunoisolation. Cells live inside a nanoporous container where nutrients and oxygen diffuse in through pores too small to allow interaction with the immune system. In the illustrated case, cells produce insulin as a treatment for diabetes. (From Leoni, L., and Desai, T. A., “Micromachined Biocapsules for Cell-Based Sensing and Delivery,” Advanced Drug Delivery Reviews, Vol. 56, pp. 211–229, copyright 2004, with permission from Elsevier.)
5.5 Combination of Lithography and Self-Assembly to Construct 3-D Devices
143
Regarding the overall size of the containers, we have developed containers for this application in the size range of 100 to 200 μm. The justification for this size range is as follows: We are developing devices that can be introduced using minimally invasive surgery (e.g., injection), while at the same time being large enough to have a finite encapsulation volume for cells. Moreover, in order for the cells to receive enough nutrition from the surrounding medium, they cannot be too far away from the nutrient source (about 200 μm [99]). Hence, although we can fabricate containers larger than 200 μm, they would be difficult to implant and the large size would mean that cells in the interior of the container would have limited access to nutrients from the surrounding media. Sizes smaller than 100 μm will hold too few cells, thereby limiting the therapeutic effect. We have chosen a size of 200 μm as our optimum size, since this has 8x volume of a 100-μm container to hold more cells, and is small enough to be implanted by injection. Additionally, a 200-μm container has a size allowing adequate diffusion (the distance from the side face of the container to the center is one-half times the length of the container or 100 μm). The typical volume of a mammalian cell is in a range of 2.5 pL (diameter 16.8 μm) to 7.5 pL (diameter 24.2 μm) [100]. A typical beta cell for example is approximately 2.5 pL in volume (diameter 17 μm). [101]. From the above data on the volume of a β-cell, we could encapsulate a maximum of approximately 2,800 cells in a 200 mm container (since a β-cell can swell to 2.85 pL). However this number is not realistic, since these 2,800 cells would densely pack the containers and would not allow adequate diffusion of nutrients and waste to/from cells in the center of the container. Therefore, realistic numbers are approximately 1,200 to 1,600 cells per 200-μm container. The number of cells required for treatment depends on the disease. We have developed containers that encapsulate cells during assembly [102]. Hence the containers self-load cells en masse. This feature coupled with the fact that containers can be patterned with nanopores points to a straightforward cell packaging strategy. Additionally, since the containers are patterned in 3-D, they function as 3-D microwells. We have verified that cells and embryos packaged within the container are viable (Figure 5. 8). Three-Dimensional Self-Assembled Containers for Drug Delivery
Since the containers have a finite volume, they can also be used for drug delivery. Additionally, the containers can be made out of metal, so they can be coupled remotely to external electromagnetic fields. When a container is placed above a coil through which an alternating current is passed, an electromotive force (EMF) is induced in the container due to an inductive effect. This induced EMF causes eddy currents to flow within the metallic frame of the container, resulting in Joule or resistive heating. One can witness this heating in metallic pots on induction cookers that are powered by radio frequency (RF) signals. Hence, when the container is placed in an RF field, it can be heated precisely. We utilized this effect to remotely trigger the release of chemicals from the container (Figure 5.6(c)). The chemical to be released was soaked in a gel that was inserted into the container. We utilized a variety of gels such as n-isopropylacrylamide or Pluronics® block copolymer. These gels collapse or soften
144
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices
(a)
(b)
(c)
Figure 5.8 Self assembling 3-D microcontainers capturing cells. (a) Patterned 2-D sheets spontaneously fold into 3-D microcontainers in the presence of cells, encapsulating them. (b) Photomicrograph of a container with living cells (fluorescing, visible as lighter gray) visible through windows. The polymeric hinge material appears dark. (c) Photomicrograph of a captured Triops embryo inside a windowed container. (These experiments were done by Timothy Leong, Bryan Benson, Christina Randall and Jillian Epstein in the Gracias Laboratory.)
on heating. Hence when the container was heated with the RF field, the gel released its chemical contents. Since the containers could be heated remotely, we were able to release chemicals to live cells and in tissue from distances as far away as 5 mm at the push of a button. We believe remote on-demand release of drugs may be of therapeutic value in medical conditions whose exact time of occurrence cannot be predicted (e.g., therapies for alleviation of seizures and on-demand vaccines against hostile chemical attacks). This application highlights one of the major needs in medicine; that is, remote telemetry or remote sensing, actuation, and recording of information from within the body. This remote monitoring remains a major challenge at the sub-millimeter scale. One limitation is that although conventional radio frequency identification chips have been miniaturized, they usually operate in the gigahertz range. Power coupling remains challenging on sizes far below the wavelength of RF radiation used, therefore powering miniaturized implanted electronic devices is difficult with the 30-cm wavelength of a gigahertz wave. We have done both in vitro and in vivo biocompatibility studies with the gold-coated devices fabricated to date. In vitro studies were done using the ISO protocol ISO 10993-5. These experiments were carried out with L929 mouse fibroblast cells. For the direct contact test, the cells were grown to a confluent layer in a 24-well glass-bottom dish. In addition to the encapsulated cell plates, positive controls (media exposure only) and negative controls (media with toxic latex exposure) were used for each test. After incubating, media were removed from the wells, the cells were rinsed with saline and then stained with the LIVE/DEAD® viability assay (Invitrogen). We observed no difference between the amount of dead cells in the positive controls and the amount of dead cells in the presence of devices. The latex
5.5 Combination of Lithography and Self-Assembly to Construct 3-D Devices
145
negative control caused increased cell death. Similar tests will be carried out for all microdevices. Additionally, we have implanted the devices subcutaneously in a severe combined immunodeficiency (SCID) mouse and have observed no necrotic death of encapsulated cells over a period of 48 hours. Clearly, more long-term studies need to be carried out to test for immune responses, inflammation, and fibrosis. Additionally, it is necessary to explore possible surface coatings of the devices including gold, platinum, silicon, carbon, titanium, and polymeric coatings such as polyethylene glycol. Three-Dimensional Self-Assembled Containers for Sampling
As the containers may be designed with openings of various sizes, they may be used for size-selective sampling of the environment, in analogy to a butterfly net. Their small size allows motion and maneuverability via gravity and liquid forces, and the containers may be moved with magnetic forces or platinum catalyzed gas decomposition [103]. Using such techniques, we are exploring applications in separations, by moving containers with specific porosity to entrap cells and other biologically relevant objects and molecules based on size [104]. 5.5.2
Multilayer Thin Film Stress for 3-D Self Assembly
Alternative forces may be used to assist the assembly from 2-D to 3-D structures. For example, thin films of metals deposited by thermal evaporation or sputtering exhibit residual tension similar to a rubber band that is stretched. If the rubber band were to relax alone, it would shrink to a smaller equilibrium length. However, when adhered to another unstressed thin film, the stressed layer (rubber band) will cause the bilayer to roll up. By this mechanism it is possible to construct stacks of thin metal films with different stresses and cause a hinge to fold out of the plane, enabling spontaneous 3-D self-assembly. As described below, we have used the residual tension in a thermally evaporated Cr film to drive assembly of lithographically patterned shapes into 3-D structures [105]. For manufacture of cubic containers, the frames were made of rigid electrodeposited Ni. These containers closed in warm water and were compatible with cells. We have extended this approach to 3-D cell culture on large constructs and the manufacture of microsurgical tools. 5.5.3
Three-Dimensional Constructs for Cell Culture
As noted earlier, cells organize in 3-D geometries in the human body. Hence, there is a need for 3-D cell culture constructs with complex architectures. We have utilized self-assembly to transform 2-D patterned substrates into complex 3-D geometries including spirals, coils, arches, and cylinders [106]. The strategy used large patterns of alternating rigid segments and flexible regions using lithography. By altering the geometry and patterns of the two parts, and varying mechanical properties and stress along these features, different structures formed on release from the substrate (Figure 5.9(a)).
146
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices
(a)
(b)
(c)
Figure 5.9 (Color plate 8) Multilayer thin films for 3-D assembly. (a) Patterned 2-D sheets spontaneously assemble into 3-D structures. The left panel shows an array of rigid metallic segments (coated with gold) on a surface of patterned thin metallic films (dark background). The center panel shows the resulting cylinder after the structure is released from an underlying substrate, where the thin metallic films release stress and self-assemble into a 3-D structure. The right panel shows a collection of cylinders that self-assembled from sheets patterned simultaneously on a single wafer. (b) The cylinders from (a) were used as a support on which cells could be cultured in 3-D. The left panel shows an optical/fluorescent overlay image of cells (green) on the 3-D structure. The right panel shows a fluorescence image only. With the stain used, living cells fluoresce green and dead cells red. (This experiment was done by Noy Bassik, George Stern, and Mustapha Jamal in the Gracias Laboratory.) (c) Snapshots of the capture of cells from a cell culture mass using a tetherless gripper. The top-left panel shows a microgripper fabricated via lithography made of metallic and polymeric components. The next panel to the right shows a glass capillary (approximately 1.5–2 mm in diameter) in a liquid bath with packed cells at one end (red mass). Above the capillary is a microfabricated gripper that can be manipulated using a magnet. In this sequence of still images taken from a video, the gripper enters the capillary, where a thermal trigger is applied and the gripper closes. The bottom panels show the gripper grabbing cells as it is pulled through the capillary. In the final panel at the lower right, the gripper has completed an in vitro cell retrieval and rests outside the capillary. (This experiment was done by Bryan Benson, Christina Randall, and Timothy Leong in the Gracias Laboratory. For further information, see reference 107.)
The flexible hinges are thin bilayer metal films and the stiff elements are either electrodeposited metal or photolithographically patterned polymer. The hinges operate using the thin film bending mechanism discussed above, and the rigid
5.5 Combination of Lithography and Self-Assembly to Construct 3-D Devices
147
segments are made of thicker material to resist bending and provide support. The use of lithographic fabrication allows for arbitrary patterning of rigid segments and stressed hinges, providing for a variety of flat precursors to spontaneously assemble into 3-D structures. Since the process is compatible with polymeric and metallic materials, surfaces may be functionalized to yield precise cell patterning in 2-D before folding to 3-D. In preliminary work, cells were seeded into folded 3-D cylinders and allowed to grow (Figure 5.9(b)). The cells were able to localize within the structure, providing preliminary evidence that these lithographically patterned 3-D self-assembled structures could be used as cell culture supports or scaffolds. 5.5.4 Microscale Tetherless Gripper (Chemically Triggered Microsurgical Tools)
Inspired by biological articulations in insects, we have also fabricated a gripper that can be remotely controlled from a distance [107]. The gripper was utilized to grab cells from a cell culture mass within a 2-mm tube that was accessible only from one end (Figure 5.9(c)). The grippers were designed in the shape of hands with a flat palm and alternating rigid and flexible joints, and were lithographically fabricated in 2-D. Grippers were designed with spans ranging from 700 μm to 2 mm when open and 190 μm to 0.5 mm when closed. The grippers contained joints composed of a bimetallic layer and a polymer; actuation was based on the thin film self-assembly mechanism described above. In this setting, however, self-assembly was arrested by the presence of a thick polymer layer atop the stressed thin metal films. The polymer acted as a trigger and controlled the closing of the gripper. Only when the polymer was softened (by heating) or degraded/dissolved by exposure to specific chemical environments did the gripper close. In Figure 5.9(c), we demonstrate the use of the mobile gripper in the retrieval of cells from a cell culture mass at the end of a tube. This procedure can be achieved with the use of present day conventional tethered endoscopic tools. However, it is difficult to manipulate these conventional tethered endoscopic tools in tortuous conduits; hence our mobile 3-D tools may be useful in some microsurgical procedures. This gripper also represents a new class of micromachines that can be lithographically fabricated and are functional in 3-D. These recently developed devices can be triggered en masse since they do not have any wiring or tethers, and can be designed to respond to either thermal or chemical signals as a trigger. Such chemically triggered machinery is ubiquitous in the human body; for example phagocytosis by macrophages is triggered by highly specific chemical triggers such as antigens or foreign matter. Both closing and subsequent opening of the microgrippers in response to chemicals has been demonstrated [108]; however multiuse reversible operation has yet to be achieved. An important future milestone will be the demonstration of such controllable microtool actions in an in vivo environment in response to specific chemical and biochemical sources.
148
5.6
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices
Conclusions In summary, lithographic micro- and nanoscale patterning is a very attractive strategy for defining structures in 2-D. When combined with additional paradigms such as self-assembly, it is possible to construct 3-D biomedical devices with precise patterning. This three-dimensionality and precise patterning is required in certain devices for diagnostics, imaging, and therapeutics.
5.7
Future Directions The use of self-assembly and lithography to assemble complex architectures is new and there are several challenges that need to be addressed. There is the need to extend this strategy to other materials such as biodegradable gels, biocompatible polymers, and oxides so that one can incorporate a high degree of biological function while maintaining precise surface patterning and device geometry. It is necessary to explore the size limitations of this paradigm. Size is critical in biomedical devices, since this parameter determines the extent of invasiveness, passage through different conduits in the circulatory system, as well as immune responses. It is also necessary to understand the kinds of three-dimensional structures that are needed for different applications in medicine. Self-assembly is an relatively new concept and its application to dynamical structures and microsurgical tools is entirely unexplored. Nature is replete with examples of small machines that perform a variety of functions such as capture of pathogens, crawling, pick and place, and rotation. As scientists striving to duplicate, master, and exceed nature we can only draw inspiration from its endless beautiful and functional designs [109]. In his seminal lecture in 1959 titled “There’s Plenty of Room at the Bottom,” Nobel Prize winning physicist Richard Feynman noted, “A friend of mine (Albert R. Hibbs) [110] suggests a very interesting possibility for relatively small machines. He says that, although it is a very wild idea, it would be interesting in surgery if you could swallow the surgeon. You put the mechanical surgeon inside the blood vessel and it goes into the heart and “looks” around. (Of course the information has to be fed out.) It finds out which valve is the faulty one and takes a little knife and slices it out. Other small machines might be permanently incorporated in the body to assist some inadequately-functioning organ.” In 1959, these advanced micromachines may have seemed far removed, but perhaps some day soon, these will become a reality.
Problems 5.1 Using wafer-scale microfabrication and conventional photolithography, sketch a process flow to fabricate several hundred (1 × 1 × 10 μm) micromagnetic stir bars in the shape of rectangular parallelepipeds. 5.2 Posts on substrates can be used to study interaction of cells with topography, create gecko feet like structures, interconnects, and so forth.
Problems
149
5.3
5.4
5.5
5.6
Starting with a silicon wafer substrate, draw process flows to fabricate pillar-like structures made of copper, by conventional microfabrication. (a) 100-μm tall × 100-μm wide posts separated by 10 μm. (b) 100-nm tall × 100-nm wide posts separated by 100 nm. (c) 3-nm tall × 1-mm wide posts separated by 1 μm. Serial versus parallel patterning: In optical lithography an entire 3-inch wafer can be patterned at once. Imagine instead that you were writing this wafer (by scanning its surface) with an electron or ion beam with a spot size of 1 μm. Calculate the time it would take to pattern a 3-inch wafer with 1-μm diameter circles arrayed uniformly on the wafer in a square lattice with a 3-μm spacing. Assume the beam is moving at the average rate of 10 μm/second. Imagine you have 5 μl of a solution with 1- and 5-μm balls. Draw a process flow for a polydimethylsiloxane (PDMS) based fluidic chip that can extract the solution, and separate (for later retrieval) the 1-μm balls from the 5-μm balls using size filtration. (This is a simple example that could be extended to a complex real fluid like blood, wherein you could separate plasma, red blood cells, and platelets. The separation in blood could also be done by other methods such as centrifugation; however, if you have a small volume of blood or you want to add in situ analysis, a centrifuge may not be the best way to go.) A kidney normally receives 20% of the total cardiac output of 5L/min, but this amount may be greatly reduced in times of stress. We will design an artificial kidney to supplement failing natural kidneys that will operate with 10% of cardiac output. (a) We will design this organ to filter 1L of waste output per day. What percent of filtered water must be reabsorbed? (b) Kidneys primarily filter dissolved substances like ions and small molecules into the waste stream. The organs then reabsorb some back into the body, and excrete others from the body to the waste stream. Simple filtration in our engineered device is used to remove creatinine with a blood concentration of 0.07 mmol/L, we wish to keep the blood concentration constant for a diet that produces 2 grams/day. Assume a mean arterial pressure of 100 mmHg into the device and a venous pressure of 30 mmHg postdevice. For what permeance of creatinine (in mmol/mmHg/m2/s) is the required surface area practical? (c) Unfortunately, glucose and sodium ions will leak through this membrane as well, with permeance values of 15% and 50% of creatinine. Normally, the kidney will reabsorb 100% of glucose and 99% of the sodium, but we must replace these intravenously. Calculate the replacement rate for glucose and sodium; assume a plasma concentration of 5 mmol/L glucose and 140 mmol/L for sodium. A drug delivery platform is proposed for sustained release of a cardiac drug in the gut. The drug should take action within 30 minutes, which requires a loading dose of 200 μg, and then maintain a serum level of 4 μg/L. Assume that the average patient will have 5L of blood volume and that 100 μg of the drug will be retained in noncirculating areas.
150
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices
Unfortunately, the therapeutic ratio of the drug is low and serum concentrations must stay below 10 μg/L. Design a multilayer device that will meet these criteria, using photolithographed polymers. 5.7 Independent research question: You are a venture capitalist. An enterprising M.D./Ph.D. engineer proposes to solve the clinical problem of degenerating knees by introducing a lithographically patterned scaffold seeded with patients own cells into the knee joint. The device will be custom fit for every patient, or a set of sizes may be developed. Examine this engineer’s proposal. What is the size of the expected market? Are there competing technologies for this therapy? Estimate costs for prototype fabrication in a clean room. Estimate costs for production in a cell culture environment. Would you fund this proposal?
Acknowledgments We acknowledge funds from the Arnold and Mabel Beckman Foundation and by the NIH Director’s New Innovator Award Program, part of the NIH Roadmap for Medical Research, through grant number 1-DP2-OD004346-01. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the funding agencies. We also acknowledge the assistance of Aniruddha Rajan for background research and referencing and Aasiyeh Zarafshar, Anum Azam, and David Filipiak for illustrations.
References [1] Linder, V., “Microfluidics at the crossroad with point-of-care diagnostics,” Analyst, 2007. 132 (12): pp. 1186–92. [2] Kobayashi, H. and M. W. Brechbiel, “Nano-sized MRI contrast agents with dendrimer cores,” Adv Drug Deliv Rev, 2005. 57 (15): pp. 2271–86. [3] Mulder, W. J., et al., “Magnetic and fluorescent nanoparticles for multimodality imaging,” Nanomed, 2007. 2 (3): pp. 307–24. [4] Jain, K. K., “Recent advances in nanooncology,” Technol Cancer Res Treat, 2008. 7 (1): pp. 1–14. [5] Santini, J. T., Jr., et al., “Microchip technology in drug delivery,” Ann Med, 2000. 32 (6): pp. 377–9. [6] Brannon-Peppas, L. and J. O. Blanchette, “Nanoparticle and targeted systems for cancer therapy,” Advanced Drug Delivery Reviews, 2004. 56 (11): pp. 1649–1659. [7] Ferrari, M., “Cancer nanotechnology: opportunities and challenges,” Nat Rev Cancer, 2005. 5 (3): pp. 161–71. [8] Har-el, Y.-e. and Y. Kato, “Intracellular delivery of nanocarriers for cancer therapy,” Current Nanoscience, 2007. 3: pp. 329–338. [9] Sengupta, S. and R. Sasisekharan, “Exploiting nanotechnology to target cancer,” Br J Cancer, 2007. 96 (9): pp. 1315–9. [10] Kwon, I. K., et al., “Nanoparticulate drug delivery systems for cancer therapy,” in Cancer Nanotechnology, H.S. Nalwa and T. Webster (eds.), American Scientific Publishers, 2007, pp. 333–344.
References
151
[11] Mack, C., Fundamental Principles of Optical Lithography: The Science of Microfabrication, Wiley, 2008, p. 534. [12] Taur, Y. and T. H. Ning, Fundamentals of Modern VLSI Devices, Cambridge University Press, 1998. [13] Gad-el-Hak, M., The MEMS Handbook, CRC Press, 2005. [14] Madou, M. J., Fundamentals of Microfabrication: The Science of Miniaturization, 2nd ed., Boca Raton: CRC Press, 2002. [15] Jones, R., “What can biology teach us?” Nat Nano, 2006. 1 (2): pp. 85–86. [16] Lehn, J. M., “Toward self-organization and complex matter,” Science, 2002. 295 (5564): pp. 2400–3. [17] Whitesides, G. M., and B. Grzybowski, “Self-assembly at all scales,” Science, 2002. 295 (5564): pp. 2418–2421. [18] Azam, A., et al., “Compactness Determines the Success of Cube and Octahedron Self-assembly,” PLoS One, 2009. 4(2): pp e4451. [19] Toro, R., et al., Brain Size and Folding of the Human Cerebral Cortex. Cereb Cortex, 2008. [20] Tang, Y., et al., “Total regional and global number of synapses in the human brain neocortex.” Synapse, 2001. 41 (3): pp. 258–273. [21] Taber, L. A., “Biomechanics of cardiovascular development,” Annu Rev Biomed Eng, 2001. 3: pp. 1–25. [22] Newman, S. A., “‘Generic’physical mechanisms of morphogenesis and pattern formation,” Development, 110: 1990. pp. 1–18. [23] Borenstein, J. T., et al., “Microfabrication of three-dimensional engineered scaffolds,” Tissue Eng, 2007. 13 (8): pp. 1837–44. [24] Allen, J. W. and S. N. Bhatia, “Engineering liver therapies for the future,” Tissue Eng, 2002. 8 (5): pp. 725–37. [25] Radisic, M., et al., “Functional assembly of engineered myocardium by electrical stimulation of cardiac myocytes cultured on scaffolds,” 2004, National Acad Sciences. pp. 18129–18134. [26] Tsang, V. L., et al., “Fabrication of 3D hepatic tissues by additive photopatterning of cellular hydrogels,” 2007, FASEB, p. 790. [27] Li, C., J. Han, and C. H. Ahn, Flexible Biosensors on Spirally Rolled Micro Tube for Cardiovascular In Vivo Monitoring, 2007, Elsevier. pp. 1988–1993. [28] Miki, N., et al., “Multi-stack silicon-direct wafer bonding for 3D MEMS manufacturing,” Sensors and Actuators a-Physical, 2003. 103 (1–2): pp. 194–201. [29] Gosele, U. and M. Alexe, “Wafer bonding. A flexible approach to materials integration,” Electrochemical Society Interface, 2000. 9 (2): pp. 20–25. [30] Lu, J. Q., et al., “A wafer-scale 3D IC technology platform using dielectric bonding glues and copper damascene patterned inter-wafer interconnects,” Interconnect Technology Conference, 2002. Proceedings of the IEEE 2002 International, 2002. pp. 78–80. [31] Tadepalli, R. and C.V. Thompson, “Quantitative characterization and process optimization of low-temperature bonded copper interconnects for 3-D integrated circuits.” Interconnect Technology Conference, 2003. Proceedings of the IEEE 2003 International, 2003. pp. 36–38. [32] Hruby, J., “Liga Technologies.” Journal of Micromech. Microeng, 2001. p. 337. [33] Howe, R.T., Recent advances in surfacemicromachining, Technical Digest: 13th Sensor Symposium, Tokyo, Japan, 1995, pp. 1–8 Varadan, K. and V. Vasundara, “Micro stereo lithography and fabrication of 3D MEMS [34] and their applications.” Proceedings of SPIE-The International Society for Optical Engineering, 2001. 4592: pp. 9–20.
152
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices [35] Markus, K.W., et al., “MEMS infrastructure: the multiuser MEMS processes (MUMPs).” 1995, SPIE. pp. 54. [36] Rodgers, M.S., et al., “Advanced micromechanisms in a multilevel polysilicon technology.” 1997, SPIE. p. 120. [37] Friend, D.R., “Drug delivery to the small intestine.” Curr Gastroenterol Rep, 2004. 6(5): pp. 371–6. [38] Gombotz, W.R. and D.K. Pettit, “Biodegradable Polymers for Protein and Peptide Drug Delivery.” Bioconjugate Chemistry, 1995. 6(4): pp. 332–351. [39] Peppas, N.A., et al., Hydrogels in Biology and Medicine: From Molecular Principles to Bionanotechnology. 2006, Wiley. pp. 1345–1360. [40] Saltzman, W.M. and W.L. Olbricht, “Building drug delivery into tissue engineering. Nat Rev Drug Discov, 2002. 1(3): pp. 177–86. [41] Chen, R.R., et al., “Spatio-temporal VEGF and PDGF delivery patterns blood vessel formation and maturation.” Pharm Res, 2007. 24(2): pp. 258–64. [42] Langer, R., “Drug delivery. Drugs on target.” Science, 2001. 293(5527): pp. 58–9. [43] Park, H., et al., “Nanofabrication and microfabrication of functional materials for tissue engineering.” Tissue Eng, 2007. 13(8): pp. 1867–77. [44] Randall, C.L., et al., “3D lithographically fabricated nanoliter containers for drug delivery.” Advanced Drug Delivery Reviews, 2007. 59(15): pp. 1547–1561. [45] Santini Jr, J.T., et al., Mikrochips als Systeme zur kontrollierten Freisetzung von Medikamenten. 2000. pp. 2486–2498. [46] Santini Jr, J.T., et al., “Microchips as Controlled Drug-Delivery Devices.” Angew Chem Int Ed Engl, 2000. 39(14): pp. 2396–2407. [47] Tao, S.L. and T.A. Desai, “Microfabricated drug delivery systems: from particles to pores.” Adv Drug Deliv Rev, 2003. 55(3): pp. 315–28. [48] Bhatia, S.N., et al., “Effect of cell-cell interactions in preservation of cellular phenotype: cocultivation of hepatocytes and nonparenchymal cells.” Faseb J, 1999. 13(14): pp. 1883–900. [49] Prescott, J.H., et al., “Chronic, programmed polypeptide delivery from an implanted, multireservoir microchip device.” Nature Biotechnology, 2006. 24(4): pp. 437–8. [50] McAllister, D.V., M.G. Allen, and M.R. Prausnitz, “Microfabricated microneedles for gene and drug delivery.” Annu Rev Biomed Eng, 2000. 2: pp. 289–313. [51] Hallgren, C., et al., “An in vivo study of bone response to implants topographically modified by laser micromachining.” Biomaterials, 2003. 24(5): pp. 701–10. [52] Hallgren, C., et al., “The importance of surface texture for bone integration of screw shaped implants: an in vivo study of implants patterned by photolithography.” J Biomed Mater Res, 2001. 57(4): pp. 485–96. [53] Holy, C.E., et al., “Use of a biomimetic strategy to engineer bone.” J Biomed Mater Res A, 2003. 65(4): pp. 447–53. [54] Brunette, D.M. and B. Chehroudi, “The effects of the surface topography of micromachined titanium substrata on cell behavior in vitro and in vivo.” J Biomech Eng, 1999. 121(1): pp. 49–57. [55] Caves, J.M. and E.L. Chaikof, “The evolving impact of microfabrication and nanotechnology on stent design.” Journal of Vascular Surgery, 2006. 44(6): pp. 1363–8. [56] Sousa, J.E., P.W. Serruys, and M.A. Costa, “New Frontiers in Cardiology Drug-Eluting Stents: Part I.” 2003, Am Heart Assoc. pp. 2274–2279. [57] Reed, M.L., et al., “Micromechanical devices for intravascular drug delivery.” J Pharm Sci, 1998. 87(11): pp. 1387–94. [58] Clark, G.M., “The multiple-channel cochlear implant: the interface between sound and the central nervous system for hearing, speech, and language in deaf people - a personal per-
References
[59]
[60]
[61] [62] [63] [64] [65] [66] [67] [68] [69] [70]
[71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81]
153 spective.” Philosophical Transactions of the Royal Society B-Biological Sciences, 2006. 361(1469): pp. 791–810. Patrick, J.F., P.A. Busby, and P.J. Gibson, “The Development of the Nucleus(R) FreedomTM Cochlear Implant System.” Trends in Amplification, 2006. 10(4): pp. 175–200. Stieglitz, T., M. Schuettler, and K.P. Koch, “Implantable biomedical microsystems for neural prostheses.” IEEE Engineering in Medicine and Biology Magazine, 2005. 24(5): pp. 58–65. Weibel, D.B., W.R. Diluzio, and G.M. Whitesides, “Microfabrication meets microbiology.” Nat Rev Microbiol, 2007. 5(3): pp. 209–18. Voldman, J., “BioMEMS: Building with cells.” Nat Mater, 2003. 2(7): pp. 433–4. Li, N., A. Tourovskaia, and A. Folch, “Biology on a chip: microfabrication for studying the behavior of cultured cells.” Crit Rev Biomed Eng, 2003. 31(5–6): pp. 423–88. Andersson, H. and A. van den Berg, “Microfabrication and microfluidics for tissue engineering: state of the art and future opportunities.” Lab Chip, 2004. 4(2): pp. 98–103. Argmann, C.A. and J. Auwerx, “Collection of blood and plasma from the mouse.” Curr Protoc Mol Biol., 2006. Bhadriraju, K. and C.S. Chen, “Engineering cellular microenvironments to improve cell-based drug testing.” Drug Discov Today, 2002. 7(11): pp. 612–20. Borenstein, J.T., et al., “Microfabrication Technology for Vascularized Tissue Engineering.” Biomedical Microdevices, 2002. 4: pp 167–175. King, K.R., et al., “Biodegradable Microfluidics.” Advanced Materials 2007. 16: pp 2007–2012. Bettinger, C.J., et al., “Microfabrication of Poly (Glycerol–Sebacate) for Contact Guidance,” Biomaterials, 2006, 27: pp 2558–2565. Sarkar, S., et al., “Vascular tissue engineering: microtextured scaffold templates to control organization of vascular smooth muscle cells and extracellular matrix.” Acta Biomaterialia, 2005. 1(1): pp. 93–100. Golden, A.P. and J. Tien, “Fabrication of microfluidic hydrogels using molded gelatin as a sacrificial element.” Lab on a Chip, 2007. 7(6): pp. 720–725. Shin, M., et al., “Endothelialized Networks with a Vascular Geometry in Microfabricated Poly (dimethyl siloxane).” Biomedical Microdevices, 2004. 6(4): pp. 269–278. Kaihara, S., et al., “Silicon micromachining to tissue engineer branched vascular channels for liver fabrication.” Tissue Eng, 2000. 6(2): pp. 105–17. Edelman, E.R., “Vascular Tissue Engineering Designer Arteries.” 1999, Am Heart Assoc. pp. 1115–1117. Nerem, R.M. and D. Seliktar, “Vascular Tissue Engineering.” Annual Review of Biomedical Engineering, 2001. 3: pp. 225–243. Amiel, G.E., et al., “Engineering of Blood Vessels from Acellular Collagen Matrices Coated with Human Endothelial Cells.” Tissue Eng, 2006. 12: pp. 2355. Uchida, T., et al., “Development of biodegradable scaffolds based on patient-specific arterial configuration.” J Biotechnol, 2007. Li, Y. and T. Tanaka, “Phase Transitions of Gels.” 1992, Annual Reviews. Mater. Sci. 22: pp. 243–277. Peppas, N.A., Hydrogel in Medicine and Pharmacy II and III, 1987, CRC Press, Boca Raton, FL. Guan, J., et al., “Fabrication of polymeric microparticles for drug delivery by soft lithography.” Biomaterials, 2006. 27(21): pp. 4034–41. Liu, H. and Y. Ito, “Cell attachment and detachment on micropattern-immobilized poly (N-isopropylacrylamide) with gelatin.” Lab on a Chip, 2002. 2(3): pp. 175–178.
154
Three-Dimensional Lithographically Structured Self-Assembled Biomedical Devices [82] Bourke, S.L., et al., “A photo-crosslinked poly(vinyl alcohol) hydrogel growth factor release vehicle for wound healing applications.” AAPS PharmSci, 2003. 5(4): p. E33. [83] Balls, M., “Replacement of animal procedures: alternatives in research, education and testing.” Laboratory Animals, 1994. 28(3): pp. 193–211. [84] Sharma, S., et al., “Controlled-release microchips.” Expert Opin Drug Deliv, 2006. 3(3): pp. 379–94. [85] Parker, K.K., et al., “Directional control of lamellipodia extension by constraining cell shape and orienting cell tractional forces.” Faseb J, 2002. 16(10): pp. 1195–204. [86] McBeath, R., et al., “Cell shape, cytoskeletal tension, and RhoA regulate stem cell lineage commitment.” Dev Cell, 2004. 6(4): pp. 483–95. [87] Ott, H.C., et al., “Perfusion-decellularized matrix: using nature’s platform to engineer a bioartificial heart.” Nat Med, 2008. 14(2): pp. 213–21. [88] Atala, A., et al., “Tissue-engineered autologous bladders for patients needing cystoplasty.” Lancet, 2006. 367(9518): pp. 1241–1246. [89] Desai, T.A., D.J. Hansford, and M. Ferrari, “Micromachined interfaces: new approaches in cell immunoisolation and biomolecular separation.” Biomol Eng, 2000. 17(1): pp. 23–36. [90] Lucas, L.C. and J.E. Lemons, “Biodegradation of restorative metallic systems.” Advances in Dental Research, 1992. 6(1): pp. 32–37. [91] Dowson, D., “New joints for the Millennium: wear control in total replacement hip joints.” Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine, 2001. 215(4): pp. 335–358. [92] Leong, T.G., et al., “Surface Tension-Driven Self-Folding Polyhedra. Langmuir, 2007. 23(17): pp. 8747 –8751. [93] Syms, R.R.A., et al., “Surface tension-powered self-assembly of micro structures - The state-of-the-art.” Journal of Microelectromechanical Systems, 2003. 12(4): pp. 387–417. [94] Gimi, B., et al., “Self-assembled three dimensional radio frequency (RF) shielded containers for cell encapsulation.” Biomedical Microdevices, 2005. 7(4): pp. 341–345. [95] Leong, T., et al., “Spatially controlled chemistry using remotely guided nanoliter scale containers.” Journal of the American Chemical Society, 2006. 128(35): pp. 11336–11337. [96] Desai, T.A., et al., “Nanopore Technology for Biomedical Applications.” Biomed. Microdevices, 1999. 2(1): pp. 11–40. [97] Desai, T.A., et al., “Microfabricated immunoisolating biocapsules.” Biotechnology and Bioengineering, 1998. 57(1): pp. 118–120. [98] Desai, T.A., “Microfabrication technology for pancreatic cell encapsulation.” Expert Opin Biol Ther, 2002. 2(6):p p. 633–46. [99] Thomlinson, R.H. and L.H. Gray, “The Histological Structure of Some Human Lung Cancers and the Possible Implications for Radiotherapy.” British Journal of Cancer, 1955. 9(4): pp. 539–49. [100] Korchev, Y.E., et al., “Cell Volume Measurement Using Scanning Ion Conductance Microscopy.” 2000, Biophysical Soc., pp. 451–457. [101] Miley, H.E., et al., “Glucose-induced swelling in rat pancreatic beta-cells.” Journal of Physiology-London, 1997. 504(1): pp. 191–198. [102] Leong, T.G., et al., Self-loading lithographically structured microcontainers: 3D patterned, mobile microwells.” Lab Chip, 2008. 8(10): pp. 1621–4. [103] Paxton, W.F., et al., “Chemical locomotion.” Angewandte Chemie-International Edition, 2006. 45(33): pp. 5420–5429. [104] Randall, C.L., et al., “Size selective sampling using mobile, three-dimensional nanoporous membranes,” Analytical and Bioanalytical Chemistry, 2009. 393(4): pp 1217-1224. [105] Leong, T.G., et al., “Thin film stress driven self-folding of microstructured containers.” Small, 2008. 4(10): pp. 1605–9.
References
155
[106] Bassik, N., et al. “Patterning thin film mechanical properties to drive assembly of complex 3D structures,” Advanced Materials, 2008. 20: pp. 4760–4764. [107] Leong, T.G., et al., “Tetherless thermo-biochemically actuated microgrippers.” PNAS, 2009. 106: pp. 703–708. [108] Randhawa, J.S., et al., “Pick-and-place using chemically actuated microgrippers.” Journal of the American Chemical Society (JACS), 2008. 130(51): pp. 7238–7239. [109] Sanchez, C., H. Arribart, and M.M. Guille, “Biomimetism and bioinspiration as tools for the design of innovative materials and systems.” Nature Materials. 2005. pp. 277–88. [110] Feynmans lecture was reprinted in: Feynman, R. P., “There’s plenty of room at the bottom”, Journal of Microelectromechanical Sytems, 1992, 1 (1) pp. 60–66.
Selected Bibliography Madou, M.J., Fundamentals of Microfabrication: the science of miniaturization. 2 ed. 2002, Boca Raton: CRC Press. Gilbert, S. F. 1997. Developmental Biology. 5th ed. Sinauer Associates Inc., Sunderland. Lanza, R.; Langer, L.; Vacanti, J.; Principles of Tissue Engineering. 3rd ed. 2008, Academic Press Morgan, J. R., Yarmush, M. L. eds; Tissue Engineering Methods and Protocols. Vol. 18, Methods in Molecular Medicine, 1998; Humana Press Inc.; Totowa, NJ Tay, F.E.H.; Materials & process integration for MEMs, 2002, Kluwer Academic Publishers, Boston Öberg, P. A. , Togawa, T., Spelman, F.A. eds; Sensors in Medicine and Health Care, 2004, Wiley-VCH Xing, W., Cheng, J.; Biochips: Technology and Applications, 2003, Springer-Verlag, Heidelberg Baker, R.W.; Membrane technology and applications, 2004, Wiley Schäfer, A.I., Fane, A.G., and Waite, T.D. eds; Nanofiltration: principles and applications, 2004, Elsevier Advanced Technology, New York
CHAPTER 6
Nanosized Magnetite for Biomedical Applications I. Nedkov, R. E. Vandenberghe, and Ph. Tailhades
6.1
Introduction Magnetic materials and magnetism have stimulated man’s imagination since the most ancient of times. Magnetite was the first magnetic material to be described [1]. In antiquity, its peculiar property of being attracted by iron objects and of moving when driven by no apparent physical force was attributed to divine will. In ancient Egypt, it was believed that magnetite originated from the bones of god Hor (or Horus), son of Osiris and grandson of Geb, the god of Earth [2]. In ancient Greece, Thales of Miletus bestowed a soul on magnetite [1]. In the Middle Ages and up to the 18th century, the phenomenon of magnetism remained unexplained and shrouded in mysticism and superstition. In the Old Slavonic languages the words magnetism and magic have a common root, expressing the notion of a supernatural magic behavior. The mutual attraction of stones containing a magnetic material has in the past often evoked associations of a love relationship and a kiss. The Chinese word for magnetite, chu-shih, can be translated as stones in love, while in French aimant means magnet, but also loving, affectionate, fond. Powders of magnetite and magnetic stones have always been preferred materials in the panoply of metaphysicists and sorcerers. They have been often prescribed for contradictory applications, such as charms for love or for separation. This chapter discusses the properties and biological applications of magnetite nanomagnetic ferroxide ionic crystals. Nanosized particles of inorganic substances can be considered as being a borderline state of matter between an atomic cluster formation and an ordered crystalline structure. Early studies in field of magnetic clusters and small particles have mainly been focused on simpler 3-D metal structures in view of clarifying the basic mechanisms of magnetic interaction within the individual particle and between the particle and the surrounding medium. In view of the practical necessity for materials having suitable magnetic anisotropy, a blocking temperature close to the room temperature, and superparamagnetic behavior, recent research has been directed to 3-D metal alloys possessing large magnetic moments while being magnetically soft materials [3, 4]. Modern magnetic nanoscience has concentrated its attention on nanostructured oxides (spinels, perovskites, hexaferrites), since their magnetic properties are directly associated with the near and far structural orders and optimization of their magnetic characteristics can be achieved via cation substitution or variations in
157
158
Nanosized Magnetite for Biomedical Applications
stoichiometry. In certain oxide ionic compounds, such as the transition group ferrites and rare-earth iron garnets, the antiferromagnetically coupled sublattices are not equivalent. In these materials there appears one exchange interactions of the magnetic ions through a nanomagnetic oxide ion [5], which Neel determined as the magnetic “superexchange” interaction [6]. The critical magnetic energy barrier of this interaction is responsible for the different physical behavior of materials and depends on the crystalline anisotropy length and on the exchange constant of interatomic spin interactions in ionic crystals. From the point of view of biomagnetic applications, the oxides have the important advantage of being biocompatible. In the 4 billion years of its existence, nature has created some astounding solutions to the problems it has encountered. Life structures its matter down to the finest detail. The scale of the living cell for example is the micrometer but a cell is equipped with highly complex nanoscale machinery. There is now substantial evidence that all living organisms, including animals and humans, contain magnetic nanoformations that play the role of bioreceptors [7]. Hemoglobin for example is an iron complex in blood with magnetic behavior. Magnetism is profoundly related with living organisms because it is an intrinsic property of every atom. Nanostructured magnetite is one of the first ferroxide materials with proven vital functions in the bioworld. The bio-origin (the so-called biomineralization of ferromagnetic materials) of a large part of the magnetite on the Earth’s surface is one of the first proofs for the existence of living organisms on our planet [8]. In 1960 [9], with the development of the first magnetic nanoparticles based on a stable magnetic fluid (ferrofluids) and of the nanotechnologies, new hopes for medical applications arose, such as targeted drug delivery in vivo, contrast agents in MRI tomography, biomagnetic separation of DNA and proteins, and hyperthermia treatment. These applications are based on the biocompatibility of magnetite and on the nanoparticle’s magnetic characteristics, namely, superparamagnetic behavior and high saturation magnetization.
6.2
Crystalline Structure 6.2.1
Bulk Magnetite
Magnetite is a model system for ferromagnetic materials with spinel crystalline structure and exhibits one of the largest magnetic moments among oxides. In an original and useful way it combines magnetic properties that are brought about by the presence of the Fe3+ located in both magnetic sublattices of the spinel, and electric properties that arise from the electron exchange between Fe2+ and Fe3+. At temperatures near 130 K, magnetite undergoes a phase transition from semiconductor to metal, known as the Verwey transition. Magnetite is abundant in nature and has a cubic crystalline structure that is schematically represented in Figure 6.1 [10]. Magnetite (Fe3O4) crystallizes in a spinel structure. The large oxygen ions are closely packed in a cubic order, while the smaller Fe ions fill the voids, the lattice being of two types: 1. Tetrahedral sites: the Fe ion is surrounded by four oxygen ions. 2. Octahedral sites: the Fe ion is surrounded by six oxygen ions.
6.2 Crystalline Structure
159
Figure 6.1 (Color plate 9) Structure of magnetite. (a) Natural crystal, and (b) crystalline structure of magnetite with O- oxygen, A, the tetrahedral Fe3+ cation, and B, the octahedral Fe3+ cation.
The tetrahedral and the octahedral sites form two magnetic sublattices called A and B, respectively. The spins in the A sublattice are antiparallel to those in the B sublattice. These two crystallographic positions are quite different, the result of which is the complicated exchange interaction between the two Fe3+ ions in the two types of sites (see Figure 6.2).The term “superexchange interaction” was proposed for the first time by Kramers [5] and it imeans the interaction between the noncompensated spins of the magnetic cations through the oxygen ions. As it is known, when materials where superexchange interaction arises (antiferromagnetic materials, and especially ferrimagnetic materials) are subjected to magnetic fields with intensity exceeding that necessary for the technical saturation of the respective material, this interaction causes a noncollinearity of the magnetic moments of the sublattices, so that the material reaches saturation at much higher magnetic fields [11]. Magnetite’s structural formula is [Fe3+]A [Fe3+, Fe2+]B O2-4. This manner of cation ordering in the A and B sublattices is known as the inverse spinel structure. The A-B exchange interaction is negative and the general magnetic moment is MΣ = MAMB; therefore, the overall magnetic moment of magnetite is due to the Fe2+ ions in
B
A BIJ Figure 6.2
Superexchange interaction in a spinel structure.
I
160
Nanosized Magnetite for Biomedical Applications
the B sublattice. Figure 6.3 represents the magnetic moment of the A and B sublattices as a function of temperature and the full magnetic moment of magnetite as a function of temperature and some magneto-structural parameters. 6.2.2
Structural Characteristics of Nanoparticles
Ms [emu/g] orc*T [cm3K/mol]
The X-ray systematic study of Fe3O4 nanostructured powders given in more detail in work [12] revealed a decrease of the crystal lattice parameter a at partcile size in the order of 80 nm; above this limit, the data for the powders and bulk material practically did not differ. The sizes of the tetrahedral and octahedral voids can be calculated by using the Gorter relations [13], RA = (u - ¼)a. 3 - r(O2-) and RB = (5/8 - u)a - r(O2-) respectively, where r(O2-) = 1.32 Å is the oxygen radius and u is the oxygen parameter of lattice cell; for pure Fe3O4 it is usually assumed a = 8.394±0.001 Å and u = 0.379±0.001 Å, so that RA = 0.0550 nm and RB = 0.0745 nm [13]. It was observed experimentally the variation of both types of voids in the magnetite studied as a function of the particles size. Figure 6.4 demonstrates that in the zone where particles with spherical shape predominate, one observes an increase of the size of tetrahedral voids, while that of octahedral voids (which possess higher symmetry) is somewhat decreased, so that can assume they are practically the same as in the bulk material. As one can see, the lattice distortion has led to a significant increase in the size of the void with lower degree of symmetry. In a spherical nanoparticle, which consists of 500 to 103 atoms, the number of atoms on the surface is larger than or commensurate with that of the atoms inside. The atoms on the surface have fewer neighbors as compared with those inside, which disturbs the equilibrium and the symmetry in the distribution of the forces of attraction; as a result, the surface exerts a pressure on the interior known as Laplace pressure [14]. For the sake
240 220 200 180 160 140 120 100 80 60 40 20 0 -20 -40 -60 -80 -100 -120 -140
(P1)critic = kT/|Jab| = 25.67 P2 = Jaa/|Jab| = 0.70 P3 = Jbb/|Jab| = 0.50 with (Tc)exp =858K Jab = - 23.23 cm-1 Jaa = 16.26 cm-1 Jbb = 11.62 cm-1
% (Ms result.[Oe*cm3/g]) % (Ms_aa [Oe*cm3/g]) % (Ms_bb [Oe*cm3/g]) % (cT [cm3*K/mol]) 0
10
20
30
kT/|Jab|
Figure 6.3
Structure and magnetic properties of Fe3O4.
40
50
RA and RB, nm
6.2 Crystalline Structure
161
0.100
0.080
0.095
0.072
0.090
0.064
RB-bulk
0.085
RB RA
0.056
0.080 0.075
RB-bulk
RB
RA-bulk
0
20
40
60
80 100
0.070 0.065 0.060
RA
0.055
RA-bulk
0
100
200
300
400
500
600
700
d, nm Variation of the tetrahedral-A and octahedral-B voids size-R in a model spinel structure Fe3O4 as a function of the particles size (insert shows particles up to 100 nm).
Figure 6.4
of clarity, we will refer to the spherical particles as Laplace particles (LPs). A characteristic feature of the nanodisperse particle is the fact that, under the action of Laplace pressure, the average interatomic distance changes, giving rise to the relative volume change: DV/V = K 2 /r
(6.1)
where K is the coefficient of volume compression, r is the radius of the particle, and 2 /r is the Laplace pressure. The numerical estimates show that for a particle size of about 10 nm, ≈ 1 Jm-2 and the volume change will be approximately 10-2, corresponding to a change of the interatomic distance Da = 3.10-3 (i.e., approximately 0.3%). The Da calculations can be performed using the empirical relation: Da =
a bulk - a nanoparticle a bulk
100%
(6.2)
The data, obtained by us, for spherical Fe3O4 particles with diameter 12±5 nm showed a = 0.45 %. These changes are reflected in the variation of the magnetite’s oxygen parameter u, which increased by 0.31%. It seems that in nanoscale powders with spherical shape and average particle size below 30-nm crystal structures can arise, which are metastable in bulk materials of the same composition from the viewpoint of the temperature range of their existence, or which are not typical at all for the bulk material. This can be related to a change in the thermodynamic equilibrium conditions in the nanosized particle as compared with the bulk material with the same chemical composition, as well as to the conditions of formation of the particles. The progress achieved in soft chemical processes for ferroxide production allowed the development of techniques for the formation at room temperature of nanostructured particles of quasi-spherical shape and with minimal defects on the surface.
162
Nanosized Magnetite for Biomedical Applications
Under the influence of the surface energy, in a nanosized particle the thermodynamic conditions for phase equilibrium undergo a change. New phases can appear, which are not characteristic for the respective bulk material; in the cases when a polymorphic transformation takes place in the bulk material, its temperature (or content) can also change. As the size of a sample of a given material decreases to the nanosized scale, the contribution of the surface increases: Fs = (S/V)
(6.3)
F = Fv + Fs,
(6.4)
where F, Fv, and Fs are, respectively, the free energy and the volume and surface contribution to the free energy, and is the surface energy. The case could therefore arise where, if phase 1 is stable in the bulk samples at a given temperature (i.e., Fv (1)< Fv(2)), as the size decreases, and accounting for Fs, the condition could be fulfilled: Fv2 + ( 2S2/V2)
Fv1 + ( 1S1/V1)
(6.5)
so that phase 2 would be stable. Using the condition of phase equilibrium (6.5), one can obtain an expression for the relative change of the phase transition temperature T12 of the nanosized particle with respect to that of a bulk sample T12 . Bearing in mind that Fv2 – Fv1 = (1 – T12/ T12 ) where
(6.6)
is the heat of phase transition, one arrives at T12 – T12 / T12 = 1/ ( 1S1/V1 –
S2/V2)
2
(6.7)
To determine the phase transition temperature in the nanosized particle, one should also take into account the surface energy at Laplace pressure (for particles size 10– 1 nm, P = 108 – 1010 Pa), the effect of the various defects and additives, as well as the electric and magnetic fields. The equilibrium condition (6.6) and (6.7) means that, as the particles size decreases, phases will be preferably formed with lower surface energy (i.e., more closely packed phases). For example, in the case of the two of the most common metal crystal lattices, namely, the face-centered and volume-centered ones, the former could prove to be energetically more advantageous, since its spherical volume and surface energy are lower. Thus, if in the bulk state the closely packed face-centered phase is stable, as the particles volume decreases, this phase will exist until it is transformed into amorphous or liquid phase. If, however, the relatively “loose” volume-centered phase is stable in the bulk state at some temperature, then, as the particles size decreases, a phase transition would become possible leading to the formation of the more densely packed face-centered lattice modification. The above considerations have been confirmed by some experimental results [15].
The stabilization in the nanosized particle of structural modifications, different from those observed in the same compounds in bulk state, the change in the inter-
6.3 Nanosized Magnetism
163
atomic distances (and, therefore, in the density), have their thermodynamic explanation. If the condition F1< F2, is fulfilled for the microsamples; that is, the free energy corresponding to structural modification 1 is lower (indicating thermodynamic stability of this modification), then when the contribution of the surface energy s is taken into account, condition (6.7) could hold true for the nanodisperse particles. This takes place whenever the addition to the volume contribution due to the excess of surface energy for modification 1 exceeds that for structure 2. The use of thermodynamic relations to determine the stability regions of various structural modifications is accompanied by several difficulties. For example, it is relatively difficult to find experimentally, or estimate theoretically, the surface energy. Moreover, in the region of PS below 10 nm, the notion of surface energy loses its strict meaning, inasmuch as its definition requires a transitional zone at the phase interface, whose dimension is in the order of the particles radii. As one can see in Figure 6.4, there exists a change in the LP’s structural features having to do with the changes in the radii of the two sublattices; this can be the result of the changes in the magnetic properties of the nanostructured ferrospinel. Most probably under the action of the Laplace pressure, antiferromagnetic superstructure arise in the spherical particle The structure at different temperatures would represent a compromise between a ferrimagnet of the ordered Néel’s type and independent sublattices with elements of antiferromagnetic ordering (the sublattice is divided into smaller substructures). This is confirmed by the difficulty in attaining saturation magnetization in these particles. Moreover, this assumption correlates well with the distinct paraprocess observed in the LP with deviation in the octahedral sublattice (e.g., Jahn-Teller cooperative effect or increased concentration of Fe3+ in the octahedral voids).
6.3
Nanosized Magnetism 6.3.1
Multidomain and Monodomain Particles: Superparamagnetism
Free currents generate fields classified in terms of magnetic field strength or intensity (H). The relationship between magnetic induction (B) of a material is a function of H and the magnetization or polarization (M) in SI system is given by the equation: B = μo(H + M) [Tesla, T]
(6.8)
The B unit is called the Tesla and the total B field is the sum of the H field and the magnetization M of the medium. The constant μo is called the permeability of free space. In SI it is equal to 4π × 10-7 Henry/m. In GSM μo is set equal to unity, which makes B and H, and M numerically equal to one another, but each have different unit names—Gauss, Oersted, and emu/cm3. The CGS equation (6.8) become B = H + 4πM. The 4π factor originates from the unit field created by a unit polar on the surface of a sphere of 1-cm radius, which encloses the pole with a surface area of 4π cm2. The magnetization (M) is the density of net magnetic dipole moments in a material and is a fundamental parameter.
164
Nanosized Magnetite for Biomedical Applications
The classical magnetism is defined on the fundamental concepts of magnetic dipoles consisting of equal and opposite moments. In an unmagnetized ferromagnetic material the collections of moments are randomly oriented through the material, resulting in a minimal to zero net magnetization. In the presence of a sufficiently large external magnetic field, the moments rotate parallel to the direction of applied magnetic field unit all the dipoles are aligned. The plateau region of the magnetization curve is the saturation magnetization (Ms). Modeling and experimentation have shown that the physical properties of the free cluster and the small particle depend on their size and differ from the properties of both the isolated atom and of the respective bulk material. When the cluster grows in size, some atoms become completely surrounded by neighboring atoms. The nanoparticle of finite size is frequently described by the ratio of the number of surface atoms to the total number of atoms. When particles are nanoscopically small, the number of atoms on the surface increases greatly in proportion to those inside. Surface atoms, however, frequently have different properties compared to those in the core of the particle and usually become much more ready and prone to react. Gold for instance becomes a good catalyst for fuel cells at nanoscopic sizes. Nanoparticles can also be coated with other substances, thus allowing materials of such composite (hybrid) particles to combine several properties. Hybrid nanoparticles of iron oxide in oil create a ferrofluid, a liquid that can be shaped magnetically. The interaction with the adjacent or surrounding medium alters substantially the properties of the small particle due to such effects as changes in the geometry and hybridization of the chemical bonds, diffusion, and, at sufficiently high concentrations, interaction between the separate particles. Following the classical concept of magnetism, in the absence of a magnetic field the magnetic material exhibits a “domain structure.” Under the critical Curie temperature, the material is divided into small uniformly magnetized volumes characterized by its saturation magnetization Ìs, (a fundamental constant that depends on the type of material); thus, the magnetic state of the system is concentrated in a finite region at a minimum of the system’s energy. In the absence of a magnetic field, magnetization vectors of the individual domains are oriented in such a way that the overall magnetization of the sample is zero. When the material is magnetized to saturation, the domain structure disappears and the magnetostatic energy density per unit volume is described by the expression Es = 1/2DMs2, where D is a demagnetization factor that depends on size and shape of the object being magnetized. The domain structure is determined by the magnetic, crystallographic, and microstructural characteristics of the particular material. The domain concept helps explain the hysteretic properties of magnetic materials and the magnetization processes taking place when an external magnetic field is applied. An important role in the magnetic interactions is played by the border layers between adjacent domains called domain walls (DWs), where reversal takes place of the spin magnetic moments of the two domains. Figure 6.5(a) illustrates a simplified model of a domain wall proposed by Kittel and Galt (1956), which shows N spin magnetic moments along the wall thickness. The exchange energy density transversely to the wall can be described as: Eexch = A
2
/ N = A( d /dx)
2
(6.9)
6.3 Nanosized Magnetism
165
where is the angle of full magnetic moment reversal within the wall (see Figure 6.5(a)), and A is an exchange interaction constant (A=nJS2/a). In an oxide ionic crystal À depends on Si—the spin magnetic moments of the cations in the separate sublattices (the factor n) and the exchange integral J, with a being the interatomic distance. The crystal anisotropy of the particular material determines the direction of the spin magnetic moments orientation; the energy of anisotropy is represented by the function K( . Thus, the total energy per unit surface is given by: •
E DW =
Ú [A(dθ / dx )
2
+ K( θ )dx
]
(6.10)
-•
Equilibrium conditions for
are reached when A( dθ / dx ) 2 = K( θ )
(6.11)
In a magnetically uniaxial crystal, K( ) can be expressed sufficiently accurately by the expressions K( )=K1 sin2 , where K1 is the first anisotropy constant. The substitution in (6.10) leads to E DW = 2(AK1 )
0.5
π
Ú sin θdθ = 4(AK ) 1
0.5
(6.12)
0
Kittel (1946) showed that in the case of the simple two-domain structure illustrated in Figure 6.5(a)), the space occupied by a domain can be calculated by the expression: d = 2EDW /D Ms2
(6.13)
The Globus model proposes the sample idea for the magnetization process. In a two-domain particle, the domain wall can be considered as being a membrane pinned to the particle’s wall [16]. On the initial magnetization curve plotted in Figure 6.5(b)), point 2 corresponds to the magnetization value due to the joint effects of the displacement and the bulging. When decreasing the applied field, the wall again becomes flat, and at point 3 the magnetization M is only due to the displacement (remanence). When the field changes its sign, the wall moves in the inverse direction. The magnetic field applied in the negative direction required to return the magnetization to zero is the coercive force (Hc) and is related with nonreversible processes of break of the domain wall from the pinning center. As the grain size decreases, a critical size will be reached where the grain can no longer accommodate a domain wall as the domain wall energy exceeds the value of all other energies related to the magnetic structures combined. Below this dcr, the grain contains a single domain. Assuming a spherically shaped particle where D = 1/3, the critical size for a monodomain particle can be estimated by using the following expression: 0.5
dcr = 24 (AK1) / Ms2
(6.14)
166
Nanosized Magnetite for Biomedical Applications
(a) initial curve
magnetization
(b) Figure 6.5 (a) Spin structure in a Bloch domain wall for a two-domain particle; Na is the wall thickness, with a being the interatomic distance and N, the number of spin magnetic moments crossing the wall, is the angle between Ms and the direction of the spin magnetic moments along the y axis. (b) Globus model for magnetization where Ms, is the saturation magnetization and Hc is the coercivity [16].
Aharoni and Jakubovics [17] calculated the critical diameter for the appearance of a monodomain structure in some oxides; they thus demonstrated that the particle’s size must been the nanoscale; for example, for magnetite (Fe3O4) the theoretical critical size is 54 nm (the fundamental constants used for magnetite were A=1011 J/m and K1=1.35Í104 J/m3). The drastic changes in the magnetic properties of nanosized materials were motivated the intensive scientific efforts on studying and fabricating such materials. In a monodomain particle, the magnetic interactions are determined by the following characteristics correlation lengths: Of the crystal anisotropy LK = (A/K)1/2; Of the magnetic field applied LH = (2A/HMs)1/2; Of the magnetostatic interactions Ls = (A/2 Ms2)1/2. As the size of a monodomain spherical (Ls = 0) particle is decreased, the particle’s size may become commensurate with LK so that the material becomes anisotropic; all spins are then oriented along the easy magnetization axis and the medium exhibits a collective paramagnetic behavior, which Bean described for the fist time as superparamagnetic[18]. The multidomain particles are magnetically soft with low values of the coercivity and remanence. To change the magnetization of a multidomain particle, one only needs to translate the domain wall, which is an energetically easy process and can be accomplished in relatively low fields. In contrast, a single domain particle
6.3 Nanosized Magnetism
167
is uniformly magnetized to its saturation magnetization. The only way to change the magnetization of a single domain is to rotate the magnetization, an energetically difficult process. Variation of a monodomain particle’s energy upon application of an external magnetic field is illustrated in Figure 6.6. The monodomain particle may be considered as being a bistable magnetic system, with its bistability being determined by the crystal anisotropy. In an applied external magnetic field, the energy of the system is given by: E = KV sin 2 θ - M s HV cos θ
(6.15)
where θ is the angle between the particle’s magnetic moment and the vector of the external magnetic field H; the two boundary states are determined by the ratio between H and the system’s anisotropy field HA=2K/Ms, namely, MsH/2K=1 MsH/2K=0 and depend on the crystal’s easy magnetization axis. When the external field is such that MsH/2K>1, the system has only one direction of orientation and is not relaxational. This state can vary with the temperature H=f(T) and H, which leads to values of MsH/2K<1. When placed in an external magnetic field the magnetic moment of a single domain particle align in the direction of the field via moment and particle rotation. When the field is removed the frequency of thermally activated reversals is given by: -E/kT
f = f0 e
(6.16)
where f0 is the “attempt frequency” that is approximately 109 s-1. For relaxation times of about 100 sec the critical energy barrier is: DE = ln(τ f0)kB T
(6.17)
Model of Small Particle magnetism Single Domain Particle H
Energy Diagram
M E
q” q ’
M
Metastable M
M Stable
DE(H) Easy access of Crystal Figure 6.6 field.
q’
q”
Variation of a monodomain particle’s energy upon application of an external magnetic
168
Nanosized Magnetite for Biomedical Applications
The condition for superparamagnetism is observed and the field acts independently on each single domain particle. The particle with uniaxial anisotropy displays zero coercive force as mathematically defined by KV = kB T
(6.18)
where K is the effective magnetic anisotropy energy constant (a function of the magnetocrystalline, shape and surface anisotropy), V is the volume of the particle, and kBT, kB being the Boltzmann constant is the thermal energy. The blocking temperature, TB, is an important characteristic of a nanosized particle; above it an isolated single domain magnetic particle acquires superparamagnetic properties. If the particle’s volume V is sufficiently small, its energy of anisotropy (E = KVsin2 ) can become commensurate with the thermal energy (E = kBT); the particle’s magnetic moment starts to fluctuate randomly between MsH/2K=1 and MsH/2K=0. This phenomenon is triggered by the ambient temperature being raised above the blocking temperature and the system becomes superparamagnetic (SPM). An SPM particle exhibits a “hysteresis-less” magnetization curve. Figure 6.7 shows experimental magnetization curves of nanostructured magnetite in a multidomain (300 nm), a monodomain (30 nm), and SPM (5 nm) state. The single-domain particle has a smaller saturation magnetization. The magnetic moments of the cations in the oxide interact through the oxygen anion (i.e., one observes a superexchange interaction, so that the oxide exhibits an antiferromagnetic structure). In the presence of defects in the oxygen packing, which
Figure 6.7 Magnetization curves of nanostructured magnetite in a multidomain (300 nm), a monodomain (30 nm), and SPM (5 nm) state.
6.3 Nanosized Magnetism
169
is always the case on the particle’s surface, the exchange bridge metal cation-oxygen-metal cation (Me-O-Me), will be destroyed. Thus, a magnetically disordered layer may appear on the surface (see Figure 6.8). The superexchange interactions are very sensitive to changes in the length of the Me-O bonds and in the angle between the arms Me-O-Me, both of which are very likely to change on the surface. Based on these considerations, Kodama and Berkowitz [19] developed a numerical model for calculating the spin magnetic moments distribution in small particles, where the total energy is expressed as
[
(
) ]
E = - Â g i μ B S i ◊ H = 1 / 2 Â 2 Ji ij S j / g i μ B S j
(6.19)
where gi BSj is the magnitude of the ionic magnetic moments, with their directions being determined by the unit vector Sj. The model assumes that some of the surface spins are pointing in a direction inverse to their normal orientation, thus forming a “canted” surface layer exhibiting magnetic disorder. The relaxation time of an SPM particle follows an Arrhenius-type relation: τ = τoe
KV kT
(6.20)
is the time between attempts to escape over the barrier (equal to 10-9–1011 seconds). The critical size below which a particle would become superparamagnetic corresponds to E = KV ≈ 25kBT. o
Figure 6.8
Surface model when H=0 in <111> of nanoparticles with D=2.5 nm [19].
170
Nanosized Magnetite for Biomedical Applications
The experimental approach is of crucial importance in order for the particle’s superparamagnetic behavior to be revealed, as one has to account for the relaxation processes taking place when a nanosized particle interacts with a magnetic field. The various possible situations in this respect are listed below as distinguished by the correlation between the measurement time, m, and the relaxation time, : •
•
•
When m << , the properties revealed will be similar to those of the respective bulk material, with the exception of the specific cases of antiferromagnetic interactions. When m >> , the properties will depend on the time for a randomly chosen magnetization direction with respect to the particle’s crystallographic axes. When m is commensurate with , one encounters a situation in between the two listed above.
Table 6.1 shows the time of measurement of different parameters summarized on the basis of experimental studies [20]. Mössbauer spectroscopy has serious advantages in comparison with other techniques. This is a localized measurement yielding data on the superposition of local effects, rather than averaged data. Although relaxation phenomena are also involved, a particular spectrum can be interpreted sufficiently accurately so as to produce a considerable amount of information. Moreover, when m , the evolution of the spectrum as a function of can be obtained with very high precision. The spectrum dependence on the temperature can be used to determine the particles size. This is the reason for the considerable interest in these measurements as applied to the study of nanostructured materials. The results obtained by Mössbauer spectroscopy being dependent on the particles relaxation time, the spectra produced can be of different nature. When m << , the spectrum is of magnetic character; for m , the spectrum is distorted; and for m >> , the spectrum is superparamagnetic. While in the first and the second cases the analysis of the spectra is well mastered, there still exist problems related to the second case. The studies of the superparamagnetic properties of powders are performed in the mode m >> . The characteristic superparamagnetic doublet is observed for particles with size in the order of 10 nm. As the particles size increases, the symmetric lines appear that are typical for a ferrimagnetic phase At low temperatures the spectrum is of typical magnetic nature. When T is raised, the particles with smaller diameter exhibit relaxation time falling within the critical interval m due to which a symmetric broadening of the magnetic lines is observed. This broadening continues until at a certain temperature the relaxation Table 6.1 Type of Measurement Hysteresis, magnetic resistance Magnetic susceptibility Mössbauer effect Ferromagnetic resonance Neutronography
m
, sec
From 1 to 100 From 0.1 to 107 depending on f 2 × 108 10 From 3 to 0.5 × 10 depending on f -14 10
6.3 Nanosized Magnetism
171
time enters the interval m >> and one observes a superparamagnetic spectrum. In the case of existence of a certain volume distribution, asymmetric lines will be observed in the overall spectrum. Based on the proportion of particles whose m , this techniques allows one to estimate the degree of homogeneity of nanosized particles; it also makes it possible to analyze different magnetic phases and to determine the percentage content of the superparamagnetic phase. In practice, measurements based on Mössbauer spectroscopy allows one to detect with certainty the presence of particles with superparamagnetic behavior; the absence, however, of the characteristic doublet cannot be interpreted as pointing to the fact that the particles are not monodomain. This requires that this technique be combined with another in order to establish whether the particles are monodomain. 6.3.2
Experimental Data
Magnetite (Fe3O4) is a particularly suitable material for such studies, as it represents a basic structure of the large group of ferrospinels which, undoubtedly, will play a major role in the development of magnetic materials with practical applications. The Mössbauer spectroscopy (MöS) measurements of the selected powders at room temperature also pointed to the presence of maghemite. Figure 6.9. shows spectra of powders with different average particles size. The larger particles with size 300 nm and above 300 nm exhibit typical magnetite spectra, while decreasing the size leads to changes in the spectra structure related to oxidation processes. The MöS reveals superparamagnetic behavior of the powders with particle size of less than 5 nm. The double peak in the spectrum of superparamagnetic magnetite contradicts Mørup theory [21] and gives reason to expect oxidation of the particles’ surface and changes in the magnetic behavior due to the increased contribution of the surface in small particles. The clarification of the processes of oxidation in nanostructured magnetite was deepened by using integrated low-energy electron Mössbauer spectroscopy (ILEEMS), which allows one to examine the particle’s surface [22]. Figure 6.10 illustrates a typical spectrum of particles with size of less than 10 nm. The analysis of the changes in the surface content of Fe3+/Fe2+ in a magnetite particle as a function of its size allowed us to explore the oxidation processes. The latter are expressed in changes in the Fe2+ content in the octahedral sublattice of magnetite at the expense of the appearance of cation vacancies. The surface structure approaches that of maghemite (γ-Fe2O3). Our experiments showed that the changed surface stoichiometry can be described by the structural formula: (Fe3+)A[Fe3+5xFe2.5+2-6x x]BO4, where denotes vacancies in the octahedral sublattice. In the samples studied x varies from 0 to 0.22 but does not reach the value typical for pure maghemite (x = 0.33). This is why we refer to this surface as a quasi-maghemite structure. The ILEEMS demonstrated that these changes in the surface defects occur smoothly from the particle core to its surface, starting from a depth of approximately 3 nm, with the particle core remaining pure magnetite. This depth of oxidation, formed immediately following the powder preparation is retained regardless
172
Nanosized Magnetite for Biomedical Applications 102
A
100
B
98 96 94
300 nm
290 K
92 -10
-8
-6
-4 -2 0 2 Velocity (mm/s)
4
6
8
10
(a) 101 100 99 98 97 96 95 94 -10
30 nm -5
290 K
0 Velocity (mm/s)
5
10
(b) 101 100 99 98 97
10 nm
96 95 -10
-8
-6
290 K
-4 -2 0 2 Velocity (mm/s) (c)
4
6
8
10
(d)
Figure 6.9 MöS of magnetite powders with average particles size (à) 300±10 nm; (b) 30±5 nm; (c) 10±3 nm; and (d) 5±2 nm.
of the particles size. The TEM studies of the same powders confirm the ILEEMS data [22]. Figure 6.11 presents SEM data for the size and shape of particles with diameter of up to 30 nm. The XRD data for the same particles indicated single-phase magnetite, while the Mössbauer spectra (Figure 6.10(b)) unambiguously revealed the presence of a second magnetic phase. The investigations by means of a tunneling microscopy—Magnetic force microscope (MFM) confirmed the spherical shape of the particles and the relatively high degree of homogeneity—average diameter of 30 nm, with the particles of this size being about 78 %. Figure 6.12 is an illustration of the changes in the structure observed by means of high-resolution SEM, and presents also summarized ILEEMS and transmission MöS data on the surface and core of a magnetite particle with size 30 nm. One can see from HRSEM that the changes take place to a depth of about 3 nm. The studies explain the changes in the MöS spectra as the particle’s diameter is reduced, which
6.3 Nanosized Magnetism
173
103.0 102.5 102.0 101.5 101.0 100.5 100.0 99.5 -10 Figure 6.10
-8
-6
-4
-2 0 2 4 Velocity (mm/s)
6
8
10
Study of the surface of a magnetite particle (10 nm) by ILEEMS.
Figure 6.11 Data from high-resolution SEM measurements. The inset presents typical oxidation of the particle’s surface layer.
are obviously related to the increased contribution of the surface (i.e., as the particle’s size diminishes, the effective surface augments its share and exerts increasing influence on the nanoparticles magnetic properties. The experimental data given here is not in contradiction with the most widely known models concerning the magnetic interactions in a nanoparticle, which so far have been focused on explaining the influence of surface effects on the nanoparticle’s magnetic properties. The interpretation gravitates around the notions of “finite size effect” and surface spin disorder and crystalline defects. As was discussed in the previous section, in 1996 Kodama and Berkowitz [19] proposed a model of a ferrimagnetic particle’s core surrounded by a canted surface layer of a second magnetic phase having spin-glass properties and of the anisotropic exchange interactions arising between the two phases.
174
Nanosized Magnetite for Biomedical Applications
Figure 6.12 (a) Picture of the surface structure of magnetite particle with size 30 nm and initial formulas that summarized MöS data on the changes in the cation distribution and in ion vacancies in the surface and the core of the particle. (b) Model representation of the surface oxidation variation depending on the depth of magnetite particles with grain size 10±2 nm.
These zones affect the magnetic moments density (respectively, the magnetization process), and depend on the temperature. The structure at different temperatures would represent a compromise between a ferrimagnet of the ordered Néel’s type and independent sublattices with elements of antiferromagnetic ordering (the sublattice is divided into smaller substructures). This is confirmed by the difficulty in attaining saturation magnetization in these particles. Moreover, this assumption correlates well with the distinct paraprocess observed in the small spherical particles with deviation in the octahedral sublattic (e.g., Jahn-Teller cooperative effect or increased concentration of Fe3+ in the octahedral voids) and with the strong dependence of M(H) on the temperature [12].
6.4 Magnetic Particles and Biomedical Applications
6.4
175
Magnetic Particles and Biomedical Applications 6.4.1
Magnetite and Bioworld
Blakemore in 1975 [23] discovered bacteria—magnetotactic bacteria (see Figure 6.13(a)), which synthesize nanostructured magnetite particles that are covered with an intercellular phospholipids membrane vacuole, forming structures called magnetosomes (see Figure 6.13(b) [24]. Figure 6.13 shows that the magnetite nanoparticles in magnetotactic bacteria form the chain that acts as compass needles and passively torques the bacteria cells into alignment with the Earth’s magnetic field. Such kinds of structures are observed in many organisms as well as the honeybee [25], termites [26], Atlantic salmon [27], bobolink [28], homing pigeon [29], and the pacific dolphin [30]. Investigation of magnetosomes probably is a key to explaining the migration processes of some animals in the nature. In 1996, J. Dobson and P. Grassi [31] reported nanosized magnetite discovered in the human brain. Thus nanostructured magnetite is one of the first ferroxide materials with proven vital functions in the bio-world. The bio-origin so-called biomineralization of ferromagnetic materials of a large part of the magnetite on the Earth’s surface is one of the first proofs for the existence of living organisms on our planet and have been produced for a diverse range of organisms. Obviously, the biomagnetic systems are the result of many years of evolution and form an inevitable part of the bioworld. In meteorites studied in NASA, magnetic nanoparticles have been found that were very similar to those present in magnetotactic bacteria [32, 33].
(a)
(b)
Figure 6.13 (a) Nanotechnology in nature (magnetotactic bacteria in water and electron micrograph of a chain in a magnetic field) [24], and (b) magnetosome in Fe3O4 nanoparticles.
176
Nanosized Magnetite for Biomedical Applications
Some authors [32] postulated that there are six criteria that characterize biologically produced magnetite crystals: (1) definite size range; (2) chemical purity; (3) Crystallographic performance; (4) chains linear arrangement of the crystals; (5) elongation of crystals in the <111> crystallographic direction; and (6) crystal morphology. Where is the place of these crystals from the point of view of magnetism? In various organisms receptor cells containing single-domain magnetic nanoparticles form a discrete sublayer of the olfactory lamellae that responds physiologically to the surrounding magnetic fields. In Oncorhynchus mykiss (rainbow trout) it was confirmed [34] that magnetite nanoparticles are arranged in chains within the receptors as well as in a multilobed single cell. Davila et al. [35] for the first time reported magnetoreception based on superparamagnetic clusters of magnetite in homing pigeons. Having briefly considered the phenomenon in the bioworld of magnetism, it is interesting to note that magnetite particles with size in large range of 10 to 100 nm have been found in numerous living organisms. Figure 6.14 shows the domain characterization based on the particle size as a function of coercivity field Hc, which indirectly marks how ferromagnets can retain a memory of an applied field once it is removed. The sizes shown correspond to a particle’s single domain state at room temperature. One is bound to ask why nature has exceeded the superparamagnetism size limit (for the magnetite is about 10 nm). It is evident that biomineralized magnetic nanoparticles are a product of bioevolution and were functionally designed to interact with the geomagnetic field. They are the sensory system used to monitor information on field intensity and direction. Several hypotheses could be developed to interpret the mechanisms of magnetoreception. It would seem that a particle’s hysteresis is also of importance in what concerns its function in a living organism; besides orientation in an external magnetic intensity, the particle may fulfill tasks having to do with storing information related to surrounding magnetic intensity. The natural interest for application of nanotechnologies in biology and medicine is related to the fact that the vital cell is a small world when the size scale is lim-
Figure 6.14 Coercivity dependence from particle diameter for nanostructured magnetite. SPM=superparamagnetic, SD=single domain, and MD=multidomain particles.
6.4 Magnetic Particles and Biomedical Applications
177
ited from nano- to micrometer dimensions (see illustration in Figure 6.15). Investigations on magnetic nanoparticles are of great interest for medicine. Many neurodegenerative diseases for example are connected with the disruption of normal iron homeostasis in the brain [36]. New drugs and mechanisms of medical interventions based on magnetic interactions will be developed. Magnetite is one of the most extensively investigated nanomaterial for medical applications because of its biocompatibility. Magnetite nanoparticles display the highest magnetization and high magnetic susceptibility that are of great advantage from the point of view of magnetic treatment in some medical applications. They could be designed with biocompatible surface stabilizers and polymer’s drug transporters for new biomedical applications [37, 38]. In the past decade or so the research efforts related to the development of bioapplicable nanotechnologies have been concentrated in involving superparamagnetic particles. As we attempted to make clear above, modern technologies still cannot produce analogs of the natural nanostructured magnetite with a surface free of defects. This is one of the reasons why technologies have been recently under development aimed at preparing hybrid complexes that combine magnetic nano-objects with organic molecules, thus opening up novel ways for biomedical applications, and possibly bringing us close to fulfilling the dream of building a biocomputer. We will now discuss two of most promising experiments with magnetic nanoparticles. 6.4.2
Biomedical Applications of Magnetic Single-Domain Particles
The development in 1960 of the first magnetite-nanoparticles-based stable magnetic fluids (ferrofluids) provided new hope for medical applications such as targeted drug delivery in vivo, contrast agents in MRI tomography, biomagnetic The cell: a world between micrometer and nanometer 0.1 nm
1 nm
10 nm
100 nm 1 μm
10 μm 100 μm
1 nm
lipids
proteins small molecules
bacterium virus cell
DNA
Figure 6.15
The bioworld in nanometer scale.
10 nm
178
Nanosized Magnetite for Biomedical Applications
separation of DNA, and proteins, hyperthermia, and other applications. These applications are based on the biocompatibility of magnetite and on the nanoparticle’s magnetic characteristics, namely, superparamagnetic behavior and high saturation magnetization (magnetite exhibits one of the largest magnetic moments among oxides). An important characteristic in view of possible bioapplications is the particle’s total magnetic moment. The existence is well known in nanoparticles of the boundary effect, which leads to a decrease in the particle’s magnetization in comparison with that of bulk magnetite [19]. This is why comprehensive studies are necessary of the particle’s surface and its contribution to its overall magnetic behavior. Magnetic iron oxide nanoparticles are among the most promising nanomaterials being developed as targeted imaging and therapeutic agents for use in detecting and treating cancer. In [40] the use was reported of magnetic iron oxide particles coated with luteinizing hormone releasing hormone (LHRH), a peptide that binds to a receptor found on most breast cancer cells. After determining that these nanoparticles did not aggregate under physiological conditions—aggregation would prevent cells from taking up the nanoparticles—the investigators injected the LHRH-labeled nanoparticles into mice with implanted breast cancer tumors. Twenty hours later, close to 60% of the injected nanoparticles had accumulated in the primary tumor and another 20% had accumulated within metastatic lesions in the animals’ lungs. In contrast, only 9% of similar nanoparticles lacking the LHRH targeting agent had accumulated in tumor over the same time period. The research also detected nanoparticles in the cell nucleus, suggesting that these nanoparticles could serve as a tool for delivering anticancer genes and anticancer agents that interact with a tumor cell’s genes into a cancer cell’s nucleus. When the magnetic particle interacts with a magnetic field, two types of loss mechanisms were found—hysteresis losses and relaxation losses—and there exists optimum grain sizes that are different for both loss mechanisms. In a narrow transition region to superparamagnetic behavior, remanance, and coercivity decrease abruptly. At the same region considerable Neel losses arise that may be nearly of the same order of magnitude as hysteresis losses and are restricted to a limited region of particle sizes [41]. The particle’s superparamagnetic properties at temperatures typical for vital processes are the first necessary condition for the applications discussed, since they determine manifestation of the particles’ magnetic behavior only in the presence of an external magnetic field. When the particle size reaches a critical threshold where the remnant magnetization and the coercive force tend to zero, the particles become superparamagnetic (SPM). For SPM particles, the total magnetic moment in the absence of an external magnetic field and at T>0 K is zero. When an external magnetic field is applied, one observes alignment of the magnetic moments. The magnetic state of such particles is similar to the paramagnetic state, with the exception that in SPM particles the magnetic moment is a collective effect due to all atoms in the particles, rather than of a single atom (as in paramagnetism). In a superparamagnetic state, therefore, one observes a much higher value of the magnetic susceptibility in comparison with the paramagnetic state. Since SPM particles exhibit magnetic properties only in the presence of a magnetic field, in biofluids, the particles can be removed from suspension by applying a magnetic field and it is easy
6.4 Magnetic Particles and Biomedical Applications
179
to redisperse them in a homogeneous mixture in the absence of a magnetic field. As a result of variation of the applied field or of the temperature, SPM particles reach an equilibrium magnetization value after a characteristic relaxation time. The relaxation processes in monodomain particles takes place via two different mechanisms. The first one consists of rotation of the particle due to Brownian motion (in the case of ferrofluids). The second one corresponds to rotation of the magnetization vector, if one ignores the Brownian motion and considers a motionless particle. It is known as Néel relaxation [42] for single-domain magnetic particles. The relaxation time for Brownian motion is: τB =
4πηr 3 kB T
(6.21)
while the Néel relaxation is [42]: τ N = τ 0 exp
KV kB T
(6.22)
where η is the viscosity of the liquid, r is the hydrodynamic radius of the particle, kB is Boltzmann constant, τ0 is a time constant, τ0~10-9 s, and V is the particle volume. When an external ac magnetic field supplies energy, it assists the magnetic moments in overcoming the energy barrier and the energy could dissipate when the particle moment relaxes to its equilibrium orientation. The losses caused by the Néel relaxation lead to heating of the particle ensemble (i.e., the effect of hyperthermia is manifested, and the fluid heats up) [43]. Experiments on animals have shown that hyperthermia is a promising technique for treating tumor formations [38, 44–46]. Considering the relaxation time as a function of V and T makes it possible to define a blocking temperature, TB, (at V=const) or VB (at T=const), where the magnetization passes from an unstable state ( <
>t) (see Figure 6.16). Figure 6.17 shows that maximum DT when the radiation is about 12 seconds is 12°C, for a SPM particles and 3.5°C. The specific absorption rate (SAR) can be calculated from the experimental curve using the relation SAR = C (ΔT/Δt) [47], where C is the sample specific heat capacity, which is calculated as a mass weighted mean value of the magnetite. The first step is to verify that the applied electromagnetic field does not affect the temperature of a solution free of magnetic nanoparticles. For magnetite, the specific heat capacity is Cmag = 0.937 J/g K, whereas ΔT/Δt is the initial slope of the time-dependent temperature curve. The SAR values of the SPM sample were calculated to be about 3.47 W/g. The results show that the ferrofluid containing SPM magnetite particles covered with β-cyclodextrin can be successfully used in experiments related with hyperthermia. The heating of magnetic nanoparticles is due to the movement of the magnetic moment away from the crystal axis called the Néel mode (τN= o exp KV/kBT) and the oscillation of the whole nanoparticle is called the Brownian mode (τB=4πηr2/kBT) and is a strong function of the size of the nanoparticles [23]. When the observation time of the electromagnetic measurements is comparable to the resonance frequency of the magnetic particles in the fluid – fr, then the M relax during the measurement and the phenomenon known as “magnetic viscosity” appears. The
180
Nanosized Magnetite for Biomedical Applications Cryst. Anysotropy
Magn. Anys
Figure 6.16
Relaxation in SPM particles—the Néel mode.
36 34 32 30 28 26 24 22 20
A
B
0
2
4
6
8
10
12 14
Time (minutes) Figure 6.17 Effect of hyperthermia in ferrofluids based on (a) nonsuperparamagnetic monodomain hybrid particles Fe3O4/β-CD, and (b) superparamagnetic particles.
magnetization as a function of time after ac applied field interaction is the following [48]: Ê E τ ˆ M( τ ) = M 0 Á1 - Ú e f n( E)dE˜ Ë ¯ 0
(6.23)
where n(E)dE is the fraction of particles having an energy between E and E-dE. The existence of “boundary effect” [19] in the magnetic nanoparticles leads to a decrease in the particles’ magnetization in comparison with that of bulk magnetite. Ferrofluids consisting of particles of magnetite with different sizes are used for hyperthermia applications in tumor treatments [49–52]. When fluids that contain single-domain particles [53] are injected in tumor, these particles are easily dispersed into the cells (diameters are 1–100 nm). In [51, 52] was demonstrated that SPM nanoparticles absorb more of the AC magnetic field’s power using Brownian and Neel relaxation and provide a more homogenous heat distribution. Some tumor cells as well as malignant tumors take up to nine times more magnetic nanoparticles than normal cells [54]. The coating of SPM magnetic particles by liposomes makes
6.4 Magnetic Particles and Biomedical Applications
181
them more dispersible. Magnetoliposomes are biocompatible and open the possibility for using a large number of ferri- and ferromagnetic particles [55]. This therapy is called magnetic intercellular hyperthermia and there are a number of reports of experiments that are close to clinical trials [56–58]. It is evident that by choosing appropriate technical parameters for the external AC electromagnetic field in combination with particle size, a very small amount of magnetic material may easily be used to raise the temperature of biological tissue locally up to the level to cause cell necrosis. Development of technologies for preparation of more and more controlled hybrid magnetic nanoparticles open the possibility for large medical applications. The magnetic separation of cells and biomolecules due to superparamagnetic particles are among the different bioseparation techniques that are the most promising. The cells or biomolecules that are nonmagnetic in nature can be modified by attachment of magnetic responsive objects. In the direct treatment, magnetic nanoparticles are used to immobilize ligands that bind the target cells so that the complex formed can be separated by a magnetic field. For the indirect treatment approach, the target cell initially interacts with the primary antibody which is then immobilized on magnetic particles and added to the medium containing the cells. A simple magnetic separator can move and separate the magnetic complex. Superparamagnetic magnetite particles coated or encapsulated with polymers or liposomes can be used for magnetic labeling [59]. The isolation of mRNA, genomic DNA, and proteins using superparamagnetic particles is described in [60]. Tumor cell separation from peripheral blood has been successfully performed by immobilization of antibody on silica coated iron oxide [61]. Therapeutic applications of drug targeting are under investigation. The drug can be encapsulated in nanosized magnetic hybrid complexes and intravenously can be accumulated via magnetic field in the location of the disease. The drug can be transported via a magnetic field at the target site providing the required therapeutic concentration. Drug localization is one approach for regional antitumor treatment. The effectiveness of chemotherapy treatment may be enhanced to a great extent by magnetically assisted delivery of the cytotoxic agent to the disease site. Various novel biodegradable magnetic drug carriers are synthesized and their targeting to tumor was evaluated in vitro and in animal models [38]. In 2001, an MRI contrast agent called Resovist® was commercialized on the European market. Resovist consists of superparamagnetic iron oxide (SPIO) nanoparticles coated with carboxydextranes. These particles are accumulated by phagocytosis in cells of the reticuloendothelial system (RES) of the liver. The uptake of Resovist injection in the reticuloendothelial cells results in a decrease of the signal intensity of normal liver parenchyma on both T2- and T1-weighted images. Most malignant liver tumors do not contain RES cells and therefore do not uptake the iron particles. The resulting imaging effect is an improved contrast between the tumor (bright) and the surrounding tissue (dark). Some others attempts to create contrast agents based on iron nanoparticles used lipid bilayer liposome coatings to encapsulate them (hence the term “magnetosomes”), nonsymmetrical bilayers made of fatty acids and amphiphilic block copolymers, and so forth [62, 63]. A newer approach is to use a carbohydrate coating.
182
Nanosized Magnetite for Biomedical Applications
Recently [54], a new contrast agent for liver magnetic resonance imaging (MRI) was investigated. The task was to study the effect and to determine the degree of change in the rabbit liver MRI signal by using the experimental contrast substance: SPIO nanosized particles entrapped in â-cyclodextrin. As mentioned, it basically acts as a negative T2 contrast. At the specified size of the nanoparticles these accumulate in the Kupfer cells and lead to a substantial lowering of the liver parenchyma signal intensity during T2 imaging. This effect can be used to differentiate between healthy and pathological liver tissue (negative contrasting). Figure 6.18 demonstrates the healthy liver morphology without and with contrast. One can see that liver tissue appears darker and more detailed after a contrast agent application. Figure 6.18(b) illustrates the dependence of signal intensity versus scan number at precontrast (control) and postcontrast scans. On the ordinate is the pixels’ intensity in MR units or in percentages. On the abscissa the scan number from the corresponding series of images or the image time in seconds could be set. The number of pixels in ROI depends on the chosen matrix at scanning. Signal intensity is relatively constant at control scans while a gradual decrease of the same parameter is evident for the postcontrast scans. This finding is a definite demonstration of the applicability of the contrast agent and its proper localization within the investigated liver parenchyma. The initial experiments demonstrate that liver tissue difference of signals before and after contrasting is sufficient for the detection of focal liver lesions and is potentially usable for their diagnosis.
(a)
(b)
Figure 6.18 (a) A contrast agent for liver MRI, image of transversal slice trough liver tissue of a rabbit after contrasting, and (b) a graph showing magnetic resonance intensity in a selected region of interest as a function of the number of successive scans [63].
6.5 Conclusions
181
From the above mentioned it is clear that the rapid development of the nanotechnologies and especially of the magnetic materials opens new possibilities for medical applications. However, excited by the promising results recently achieved, we may happen to ignore the existing risks connected with such applications. For example, many companies offer nanostructured magnetite powders on the market, but the investigations of the structural peculiarities in very small particles (discussed in Section 6.2.2) show that at oxidation of the powder, or due to technological inaccuracies, the boundaries between magnetite and maghemite are very thin. Presence of needle-like crystals of maghemite in a commercial product labeled as magnetite is illustrated in Figure 6.19. The needle-like crystals themselves are not a poison, but they may turn into cancer centers if they get stuck in tissue. A strict control of the particles shape, chemical stoichiometry, and magnetic structure in powders intended for medical applications should be imposed in this respect. As mentioned before, the X-ray diffraction data does not yield a clear picture of the magnetic phases in the powder. A necessary condition for good characterization is additional Moessbauer spectroscopy investigation. The relations between the nanoparticles and the cell structure of the living organism are not yet well understood. The penetration of in vivo introduced nanosized magnetite in tumor cells of a sheep being directed by a magnetic field is illustrated in Figure 6.20. It was proven that about 40% of the magnetic particles penetrated in the cancerous cells. This is exactly the aim of this experiment, but it requires a very careful performance. This demonstrates that the investigations on the risk of medical applications of nanomaterials must be continued.
6.5
Conclusions Nanomedicine and nanobiotechnology are characterized by their highly interdisciplinary nature. Developing successful nano-based magnetic techniques for such applications in the near future would require a close collaboration between life scientists, physicists, and engineers. In the light of the above presentation,
Figure 6.19 Presence of two magnetic phases with different particles shape in a commercial product—magnetite; TEM image.
184
Nanosized Magnetite for Biomedical Applications
Figure 6.20
Proliferation of nanosized magnetite in a living cell.
nanobiotechnologies with magnetic materials could be defined as the processing, fabrication, and packaging of organic or biomaterial devices or assemblies where the dimension of at least one functional component lies between 1 and 100 nm. Our world is immersed in a sea of magnetic fields created by natural and man-made sources. Evolution has resulted in a multitude of remarkable examples of living organisms adapting to this magnetic world, which were unknown until very recently. The onset of nanotechnologies opened the lid of a new Pandora ’s Box with tools for controlling processes in the sacrosanct territory within the cells of living organisms. Extensive efforts are focused at present on exploring such phenomena as magnetic separation with the cell, drug delivery, and hyperthermia in vivo and in vitro. However, the physics of single-domain particles is still not well understood so that the real challenge remains still in the world of science. The unique property of hysteresis and the fine temperature boundary between the singled-domain magnetism and superparamagnetism are phenomena that suggest that in the nano-world it might be possible to create memorizing matter that not only could be controlled by magnetic fields, but could also change its behavior by itself depending on the surrounding medium. Leaving for now the distant future to the dreamers, it should not be too farfetched to predict that nanomagnetic therapy will soon be successfully applied to the treatment of cancer and Parkinson’s disease, among others. It is very possible that the new contrast magnetic materials will turn the nuclear magnetic resonance technique into a commonplace clinical method, like the good old X-ray checkup. On the other hand, the insufficient understanding of how nanoparticles interact with the cell is fraught with risks. It turned out that having once penetrated the living
Problems
185
organism, the nanoparticles proliferate quickly and, although it is possible to achieve accumulation on targeted tissues by way of applying an external magnetic field, it is far from clear whether risks are not thus being created for the surrounding tissues, and if so, how these tissues can be protected from the nanoparticles’ invasion. Some pessimists in the scientific community are inclined to compare the euphoria around the nanotechnologies with the early stages of studying radioactive materials, when many researchers fell victim to their insidious radiation. Only by exploring closely the balance between the positive and the negative contributions of these technologies will one be able to form an opinion on their development. We are among the optimists who believe in their successful future.
Problems 6.1 6.2 6.3 6.4 6.5
What is a single-domain particle? Explain superparamagnetism. What is magnetic nanobiotechnology? Which criterion characterize biologically produced magnetite crystals? What is magnetic intercellular hyperthermia?
References [1] Power, M., Lectures on History of Theories of Electricity and Magnetism, Oxford University Press, 1995. [2] Pinch, G., Egyptian Mythology: A Guide to Gods, Goddesses and Traditions of Ancient Egypt, Oxford University Press, 2004. [3] Bansmann, S. et al. “Magnetic and structural properties of isolated and assembled clusters,” Surface. Sci. Reports, Vol. 56, 2005, pp. 189-275. [4] Krezhov, K., “Neutronographical studies of ferroxides,” Thesis, Bulg. Acad. Sci., 2006. [5] Kramers, H.A. L’interaction entre les atoms magnetogenes dans un cristal paramagnetique, Physica, vol.1, 1934, pp. 182-193 [6] Neel, L., “Magnetism and the local molecular field,” Nobel Lecturers, 1970. [7] Häfeli, U. and Shütt, W., In Scientific and Clinical Applications of Magnetic Carriers, U. Häfeli, W. Shütt, J.Teller, M.Zborowski (eds) New York, London: Plenum, 1997. [8] Dong, H., “Biomineralization and ferroxides,” Chemical Geology, Vol. 3, 2000, pp. 299-302. [9] Rosensweig, R.E., Ferrofluides, Scientific American., Vol. 247, 1960, pp.136-141. [10] Banerjee, S. K., and B. M. Moskowitz, In Ferrimagnetic properties of Magnetite, in Magnetite Biomineralization and Magnetoreception in Organisms: A New Magnetism, Kirschvink, J. L. et al. (eds.), Plenum Publishing Corporation, 1985. [11] Belov, K.P., Ferrites in the High Magnetic Fields, Ed. Science (in Russian), 1972. [12] Nedkov, I., et al., “Phase and structural particularities of nanosized granular inverse spinels,” Phys. Stat. Sol. (A), Vol. 201, No. 5, 2004, pp. 1001–1010. [13] Krupicka, S., Physic der Ferrite und der Verwandten Magnetischen Oxide, Akademie der Wissenschaften, Prag, 1972. [14] Klabunde, K.J., Free Atoms, Clusters and Nanoscale Particles, Academic Press, San Diego, CA, 1944.
186
Nanosized Magnetite for Biomedical Applications [15] Vissokov, G.P. and Pirgov, P.S., Ultrafine Powders Obtained by Plasma Sputtering, Pollyprint, Sofia, 1988, 359. [16] Globus, A., “Magnetic hysteresis in policrystalline ferrites,” Proc. Intern. Conf. Magnetism, Moscow, 1973, pp. 192-195. [17] Aharoni, A. and Jakubovics J.P., “Cylindrical domains in small ferromagnetic spheres with cubic anisotropy,” IEEE Trans. Magn., Vol.24, 198, pp. 1892-1897. [18] Bean, C.P., “Hysteresis Loops of Mixtures of Ferromagnetic Micropowders,” Journ. Appl. Phys., Vol. 26, 1955, pp. 1381-1390. [19] Kodama, R.H. and Berkowitz, A.E., “Atomic-scale magnetic modeling of oxide nanoparticles,” Phys. Rev., Vol. 59, 1999, pp. 6321-6328. [20] Dorman, J-L., “Le phenomene de superparamagnetisme,” Revue Phys. Appl.,Vol.16, 1981, pp. 275-291. [21] Mørup, S. et al., “A new interpretation of Mössbauer spectra of microcrystalline goethite: ’Super-ferromagnetism’ or ‘super-spin-glass‘ behaviour,” Journal of Magnetism and Magnetic Materials, Vol. 40, 1983,pp. 163. [22] Nedkov, I., “Nanosized magnetite for biomedical applications,” Journal of Optoelectronics and advanced Materials, Vol. 9, No. 1, 2007, pp. 24 – 29. [23] Blakemore, B., “Magnetotactic bacteria,” Science, Vol. 190, 1975, pp. 377-379. [24] Schuler, D. and Francel, R.B., “Bacterial magnetosomes: microbiology, biomineralization and biotechnological applications,” Appl.Microbiol.Biotechnol.,Vol. 52, 1999, pp. 464-473. [25] El-Jaick, L.J. et al., “Electron paramagnetic resonance study of honeybee Apis mellifera abdomens,” Eur. Biophys. J., Vol. 29, 2001, pp. 579-586. [26] Winklhofer, M. et al., “Clusters of superparamagnetic magnetite particles in the upper-beak skin of homing pigeons: evidence of a magnetoreceptor,” Eur. J. Mineral., Vol.13, 2001, pp. 659-669. [27] Moore, A., Freake, S.M. and Thomas I.M., “Magnetic particles in the lateral line of the atlantic salmon,” Philos.Trans.R.Soc.Lond.,B, Biol.Sci.,Vol. 329, 1990, pp.11-15. [28] Beason, R.C., Dussourd, N. and Deutschlander, M.E., “Behavioural evidence for the use of magnetic material in magnetoreception by a migratory bird,” J.Exp.Biol., Vol. 198, 1995, pp.141-146. [29] Hanzlik, M. et al., “Superparamagnetic magnetite in the upper beak tissue of homing pigeons,” Biometaals, Vol. 13, 2000, pp. 325-331. [30] Zoeger, J., Dunn, J.R. and Fuller, M., “Magnetic material in the head of the common Pacific dolphin,” Science, Vol. 213, 1981, pp. 892-894. [31] Dobson, J. and Grassi, P., “Magnetic properties of human hippocampal tissue—Evaluation of artefact and contamination sources,” Brain Res.Bull., Vol. 39, 1996, pp. 255-259. [32] Thomas-Keprta, K.l. at al., “Elongated prismatic magnetite crystals in ALH84001 carbonate globules: Potential Martian magnetofossils,” Geochim.Cosmochim.Acta, Vol. 64, 2000, pp. 4049-4081. [33] Friedmann, E.I. et al., “Special feature: Chains of magnetite crystals in the meteorite alh84001: Evidence of biological origin,” Proc.Natl.Acad.Sci., USA, 2001, pp. 2176-2182. [34] Diebel, C.E. et al., “Magnetite defines a vertebrate magnetoreceptor,” Nature, Vol. 4-6, 2000, pp. 299-302. [35] Davila, A.F. et al., “A new model for a magnetoreceptor in homing pigeons based on interacting clusters of superparamagnetic magnetite,” Phys.Chem. Earth, Vol. 28, 2003, pp. 647-652. [36] Safarik, I. and Safarikova, M., “Magnetic nano- and microparticles in biosciences and biotechnology,” In Naoscale Magnetic Oxides and Bio-World, I. Nedkov and Ph. Taihlades (eds.) Heron Press, Sofia, 2004. [37] Harris, L., “Polymer Stabilized Magnetic Nanoparticles,” Ph.D. Thesis,Virginia Polytechnic Institute, 2002.
References
187
[38] Bahadur, D. and Giri, J., “Biomaterials and magnetism,” Sadhana ( printed in India), Vol. 28, parts 3&4, 2003, pp. 639-656. [39] Rosensweig, R.E., “Magnetic fluids,” Scientific American, Vol. 247, 1982, pp. 136-141. [40] Leushner, C., “Sub-cellular accumulation of magnetic nanoparticles in breast tumors and metastases,” Raport 2007 Louisiana State University in Baton Rouge (abstract is available through PubMed). [41] Hergt, R. et al., “Physical limits of hyperthermia using magnetite fine particles,” IEEE Trans.Magn., Vol. 34, No. 5, 1998, pp. 3745-3754. [42] Pajic, D. et al., “Superparamagnetic relaxation in CuFe2O4 nanoparticles,” Journal of Magnetism and Magnetic Materials, Vol. 281, 2004, pp. 353-363. [43] Hrianca, I. and Malaescu, I., “The rf magnetic permeability of statically magnetized ferrofluids,” Journ. Magn. Magn. Mater., Vol. 150, 1995, pp.131-14. [44] Hergt, R. et al., “Physical limits of hyperthermia using magnetite fine particles, IEEE Trans. Magn., Vol. 34, No. 5, 1998, pp. 3745-52. [45] Yagi, K. et al.,” Interferon-â endogenously produced by intratumoral injection of cationic liposome- encapsulated gene: cytocidal effect on glioma transplanted into nude mouse brain,” Biochem.Mol.Biol.Int., Vol .32, 1994, pp. 167-171. [46] Le, B. et al., “Preparation of tumor-specific magnetoliposomes and their application for hyperthermia,” J. Chem.Eng.Jpn., Vol. 34, 2001, pp. 66-75. [47] Babincova, M., Leszczynska, D. and Sourivong, P., “Superparamagnetic gel as a novel material for electromagnetically induced hyperthermia,” J. Magn. Magn. Mater, Vol. 225, 2001, pp.109-118. [48] Dormann, J.L., Fiorani, D. and Tronc, E., “Magnetic relaxation in fine- particle systems,” Adv. Chem. Phys., Vol. 98, 1997, pp. 283-291. [49] Gilchrist, R. et al., “Effects of electromagnetic heating on internal viscera: a preliminary to the treatment of human tumors,” Ann. Surg., Vol. 161, 1965, pp. 890–896. [50] Ramachand, C. N. et al., “Application of magnetic fluids in medicine and biotechnology,” Indian J. Pure Appl. Phys., Vol. 39, 2001, pp. 683–686. [51] Mosbach, K., Andersson, L., “Magnetic ferrofluids for preparation of magnetic polymers and their application in affinity chromatography,” Nature, Vol. 270, 1977, pp. 259–261. [52] Andreas, J. et al., “Presentation of a new magnetic field therapy system for the treatment of human solid tumors with magnetic fluid hyperthermia,” J. Magn. Magn. Mater., Vol. 225, 2001, pp. 118–126. [53] Gilchrist, R .K. et al., “Selective inductive heating of lymph nodes,” Ann. Surg., Vol. 146, 1957, pp. 596–606. [54] Jianhua, C. and Naru, Y., Chin. Ceram. Soc., Vol. 29, 2001, pp. 238-244. [55] Chan, D.C., Kirpotin, D.B. and Bunn, P.A., Jr., In: Scientific and Clinical Applications of Magnetic Carriers (eds) Häfeli, U., Schütt, W., Teller, J., Zaborowski, M. (New York, London: Plenum) 1997, p. 607. [56] Jordan, A. et al., “Magnetic fluid hyperthermia (MFH): Cancer treatment with AC magnetic field induced excitation of biocompatible superparamagnetic nanoparticles,” J. Magn. Magn. Mater., Vol. 201, 1999, pp. 413–419. [57] Hilger, I. et al., “Electromagnetic heating of breast tumors in interventional radiology,” Vitro and in Vivo Studies in Human Cadavers and Mice, Radiology, Vol. 218, 2001, pp. 570–575. [58] Hiergeist, R. et al., “Application of magnetite ferrofluids for hyperthermia,” J. Magn. Magn. Mater., Vol. 201, 1999, pp. 420–422. [59] Nakamura, N. and Matsunaga, T., “Highly sensitive detection of allergen using bacterial magnetic particles,” Anal. Chim. Acta, Vol. 281, 1993, pp. 585-589. [60] Cell Separation and Protein Purification, 1996 Technical handbook, Dynal, Oslo.
188
Nanosized Magnetite for Biomedical Applications [61] Kutushov, M. N., Kuznetsov, A. A. and Filipov, V. I., Scientific and clinical Applications of Magnetic Carriers (eds) U. Häfeli, W. Shütt, J. Teller, M. Zborowski (New York, London:Plenum) 1997. [62] La Conte, L., Nintin, N. and Bao, G., “Magnetic nanoparticle probes,” Materials Today, suppl. Nanotoday, Vol. 8, No. 5, 2005, pp. 32-38. [63] Kalionsky, T. et al., “Magnetic resonance imaging by specially formulated iron oxid nanoparticles,” Comptes rendus de l’Academie bulgare des Sciences, Vol. 60, No. 8, 2007, pp. 893-898.
CHAPTER 7
Progress in the Use of Aligned Carbon Nanotubes to Support Neuronal Attachment and Directional Neurite Growth Keith A. Crutcher, Chaminda Jayasinghe, Yeoheung Yun, and Vesselin N. Shanov
7.1
Background Injury to the peripheral nervous system (PNS) and to the central nervous system (CNS) represents a significant cause of chronic and debilitating neurological impairment. Spinal cord injury is perhaps the most dramatic example, with approximately 11,000 new injuries being reported each year, but traumatic brain injury (TBI) is increasingly recognized as a significant problem in its own right. Strategies for sparing or restoring neural function following injury must address both acute and chronic changes in the tissue. There are two general strategies: (1) minimizing the initial consequences of injury, and (2) promoting repair of the damaged tissue. Although significant advances have been made in defining the consequences of injury to the nervous system, there has been less progress in the development of therapeutic strategies to treat such injuries. As a rule, recovery from damage to the peripheral nervous system is much greater than that following damage to the CNS, but even so, regeneration is usually imperfect and there remains significant interest in the development of methods to enhance the accuracy and rate of repair of peripheral nerves as well. 7.1.1
CNS Regeneration Occurs Under Some Conditions
The bleak conclusion reached by Ramon y Cajal [1] regarding the possibility of axonal regeneration in the CNS following injury has now been tempered by numerous examples of axonal sprouting and regrowth documented under various experimental conditions, a topic that has been comprehensively reviewed by several investigators in recent years [2–29]. In short, there is increasing optimism that refinement of current strategies may, in fact, give rise to practical methods for enhancing functional recovery in the CNS that may rival the best recovery obtained following PNS injury. These strategies include stimulation of intrinsic mechanisms of neuronal plasticity through physical rehabilitation, delivery of factors that stimulate new growth, implantation of tissues and cells to replace damaged tissue, devel-
189
190
Progress in the Use of Aligned Carbon Nanotubes
opment of engineered scaffolds to support axonal growth, or some combination of these approaches. The success of any of these strategies depends on elucidation of the mechanisms that normally promote or inhibit neural repair. 7.1.2
Factors That Inhibit or Stimulate Axonal Regeneration
Two broad classes of factors can be defined that influence axonal regeneration: positive and negative regulators of axonal growth. The positive regulators include the neurotrophins, such as nerve growth factor (NGF) and matrix factors that promote growth (e.g., laminin). The negative regulators include putative inhibitors associated with myelin, such as Nogo, MAG, and OMgp [30–32], molecules associated with glial scars, such as various proteoglycans [33–35], some of which have been identified and include brevican [36], phosphacan [37], aggrecan [38], NG2 [39] and versican V2 [40], and other molecules that cause growth cone collapse [41–44]. The relative failure of axonal growth in the CNS as compared with the PNS has been ascribed to the presence of scars and/or myelin-associated inhibitors in CNS white matter. For example, CNS myelin, and the oligodendrocytes that produce it, inhibit neurite growth in a variety of in vitro assays [31, 32, 45, 46]. Myelin-associated glycoprotein (MAG) was also found to inhibit neurite outgrowth [30, 47–51] in tissue culture studies and a third factor associated with myelin, OMgp, has been added to this list of putative inhibitors [52, 53]. All of these inhibitors are thought to mediate inhibition via a membrane receptor complex that includes NGr [54] and p75 [55, 56]. The identification of specific molecules and/or signaling pathways that mediate these effects is perhaps less important than answering the question of whether myelin-associated inhibitors account for regeneration failure in the CNS. Certainly there is evidence that axons can grow within CNS white matter, which contains abundant myelin. This phenomenon was first documented using grafts of embryonic tissue [57–59] but transplants of adult neurons were found to give similar results [34, 60, 61]. Axonal regeneration into myelinated tracts must presumably take place in the presence of the various putative inhibitors that have been identified [34, 61–63]. In cases where tissue was carefully grafted directly into myelinated fiber tracts with minimal injury, one possible interpretation is that axonal regeneration is possible because no glial scar has formed [34, 59, 64]. In fact, such results have bolstered the view held by some that it is the scar that is the primary barrier to new growth, not the presence of myelin-associated inhibitory factors. Whatever the relative contribution of myelin and scars may be, one striking feature of the majority of axonal growth that occurs in white matter tracts (and in peripheral nerves) is that it almost invariably occurs in parallel with the long axis of the fiber tract. This common observation has taken on greater significance in light of results suggesting that the geometry of the neural tissue may be an important determinant of regeneration success. 7.1.3
The Geometry Hypothesis
Tissue culture studies in which neurons are grown on fresh frozen sections of nervous tissue [65–68] demonstrated that successful axonal growth on white matter
7.1 Background
191
depends on the plane of section [68]. This led to the hypothesis that myelin and its associated “inhibitors” normally constrain and guide axonal regeneration by suppressing local sprouting and restricting growth to the longitudinal axis of the fiber tract [20, 68–70]. In other words, the spatial distribution of the “inhibitors,” or other aspects of the tissue geometry, plays an important role in promoting and directing axonal regeneration. This geometry hypothesis is shown schematically in Figure 7.1. Only the myelinated fibers are shown in the diagram. The other tissue elements (astrocytes, oligodendrocytes, blood vessels, and unmyelinated fibers) are omitted for clarity. However, these other tissue elements almost certainly contribute substrates that participate in promoting axonal growth. Constrained, directed growth likely requires spatial choices between permissive and relatively less permissive substrates
CNS White Matter
Cross-section
Oblique-section
Longitudinal-section
Elongating axons
Myelin (Neurite inhibitors)
Figure 7.1 Diagram of the organization of myelinated axons in CNS white matter. The densely packed bundles of nerve fibers are shown as they would look when cut in different planes of section (i.e., transversely, obliquely, or longitudinally). The geometry hypothesis posits that regenerating axons are more likely to find a navigable pathway through the “inhibitory” factors associated with myelin by growing along the longitudinal axis of the fiber tract. If so, this provides a rationale for developing a longitudinally-oriented prosthetic material to support axonal regeneration.
192
Progress in the Use of Aligned Carbon Nanotubes
in parallel alignment in order to promote net growth along the axis of the tract. Otherwise, axons should be able to grow in any direction, which is not what is observed either in tissue culture or in organisms. Successful cases of axonal regeneration within CNS white matter or in peripheral nerves appear to involve conditions in which the geometry of the fiber tract is relatively unaltered or reconstructed. In such cases, the axons extend in a direction that is almost exclusively in parallel with the longitudinal axis of the tract. The geometry hypothesis proposes that parallel alignment of tissue elements is needed to promote longitudinal axonal growth and to reduce local branching. If this hypothesis is correct, it has potential implications for the development of prosthetic substrates for optimizing neural regeneration. 7.1.4
Artificial Substrates Can Promote Axonal Regeneration
There is a long-standing history of attempts to promote axonal regeneration within both the PNS and the CNS through the use of both natural and synthetic substrates. Grafts of peripheral nerve segments or other tissues were used in early studies and are still commonplace in the treatment of peripheral nerve injury. Other approaches involve the use of synthetic materials [71], some of which have been shown to enhance regeneration, especially in the PNS. Poly-l-lactide (PLLA) microfilaments, for example, promote peripheral nerve regeneration [72]. Previous work on synthetic nerve guides (primarily in peripheral nerve regeneration) has demonstrated the general concept of using such materials to promote axonal growth in the CNS as well. In fact, early work by Khan and coworkers provided intriguing evidence that carbon filaments could promote astrocyte migration and axonal regeneration when implanted into the injured rat spinal cord [73, 74]. More recently, Yoshii and colleagues have reported regeneration and functional recovery with the use of implanted collagen filaments [75–79]. A variety of other materials have been tested including silicon, polyethylene, polytetrafluoroethylene, poly-L-lactic acid/caprolactone, poly (phosphoester), collagen-polyglycolide, and poly (L-lactide-co-glycolide) polymer (PLGA). Most of these materials are biodegradable and relatively inert. Braided PLGA fibers, for example, are biodegradable and support peripheral nerve regeneration [80]. Electrospun PLLA fibers have also been shown to promote directed neurite outgrowth from neural stem cells [81]. A more recent study used electrospun polycaprolactone (PCL) and PCL-collagen nanofibers to demonstrate directed neurite growth [82]. The diameter of fibers was also shown to affect the length of fibroblasts grown on PGA/collagen nanofibers [83]. Functional regeneration has been reported with a meshwork of nanofibers produced from self-assembling peptides using the injured optic nerve as a model [84]. This effect may be partly related to the hemostatic effect of the material [85]. The same group recently used this nanofiber matrix in conjunction with stem cells or glial cells to demonstrate integration into the injured rat spinal cord with support of some axonal growth [86]. Another group recently reported that electrospun aligned poly (acrylonitrile-co-methylacrylate) (PAN-MA) fibers are a better substrate for peripheral nerve regeneration than the same material provided with random orien-
7.1 Background
193
tation [87]. These results are consistent with the geometry hypothesis outlined above. 7.1.5
Carbon Nanotubes and Axonal Regeneration
The discovery of carbon nanotubes (CNTs) in 1991 led to the rapid development of various methods for producing and modifying this unique form of carbon. These materials can be manufactured at cellular and subcellular scales, thus affording the opportunity to make substrates that may be more relevant to cell function. The nanoengineering group at the University of Cincinnati has made significant advances in the development of CNTs for a variety of biomedical applications such as nanosensors [88, 89]. CNTs offer unique physical properties such as electrical conductivity, flexibility, and chemical modifiability. Their conductivity permits applications such as development of electrochemical sensors [90] and conducting prosthetics [91]. Chemical modification includes substances such as glycopolymers [92] and trophic factors [93]. A recent study by Dubin et al. [94] reports that carbon nanotube fibers are not toxic to neurons and will support attachment although the fibers were not aligned. A variety of cells and tissues have been studied for interactions with CNTs but most relevant here are studies in which such materials are used to support neuronal growth. This literature has been reviewed by others [95, 96] and will only be briefly summarized here. The first study to examine the interaction of living cells with CNTs involved their use as a substrate for cultures of embryonic rat hippocampal neurons [97]. These investigators demonstrated that as-produced (AP) or modified CNTs would support neuronal attachment and neurite outgrowth. The CNTs that had been modified (with 4-hydroxynonenal) supported more growth and branching of neurites than the AP-CNTs. They also noted that the CNTs did not appear to provide any directional guidance to the growing neurites. A subsequent study by another group used a similar approach and noted that different chemical modifications of the CNTs altered the pattern of neurite growth [98]. One technical advance in the latter study was the use of a vital dye to visualize the neurons and their processes. They did not mention whether there was any influence of the CNTs on guidance of neurite growth but such effects were not apparent in the examples provided in the publication. The same group also reported that a CNT-polyethyleneimine copolymer will support neuronal attachment and growth [99], not unlike the results obtained with a polycarbonate urethane (PU)-carbon nanofiber matrix [100]. A substrate made from polyelectrolyte poly(ethyleneimine) and CNTs was found to promote differentiation of neural stem cells [101] and functionalized CNT mats will support attachment and neurite growth of rat DRG neurons [102]. CNTs have also been used to create patterns of preferred substrates for use as a positioning method to determine the location of clusters of interconnected neurons [103, 104] or to pattern the growth of neurons [105]. A recent study demonstrated the feasibility of using CNTs as a substrate for both the growth and stimulation of hippocampal neurons [91]. The ability to use carbon nanofibers as a substrate for PC12 cells has also been demonstrated [106, 107].
194
Progress in the Use of Aligned Carbon Nanotubes
CNTs offer unique morphology and properties that can be exploited for use in neural regeneration. However, most of the preparations that have been studied involve formulations that are presented as a 2-dimensional substrate. One exception to this is the development of free-standing thin-film membranes [108]. Although they used a cell line for their studies, the differentiated cells showed good neurite extension along the films. All of these CNT preparations involve the use of dispersed nanotubes that are randomly oriented. The possibility that aligned CNTs (10 nm × 300 μm) can be used to promote directed axonal growth is supported by recent studies involving CNT sheets [109]. Attachment and growth of both PNS and CNS neurons occurred on these materials and in some cases there was apparent orientation of neurites with the underlying substrate. In the same study, both Schwann cell migration and directed neurite growth were observed along CNT “yarns.” The advances that have been made in the ability to manufacture aligned CNTs [110, 111] and studies demonstrating that 2-D aligned surfaces will promote parallel axonal growth encourage pursuit of aligned 3-D CNTs to support axonal regeneration over distances that could support neural repair in the damaged PNS or CNS. Previous studies involving CNTs have been limited to the use of nanotubes that are at most a few microns in length. The ability to now extend CNTs to millimeter and even centimeter lengths provides the opportunity to develop scaffolds of aligned CNTs that could theoretically bridge significant distances in both the PNS and CNS. 7.1.6 Aligned Carbon Nanotubes as a Potential Scaffold for Axonal Regeneration
Whether chemical modification is required for CNTs to promote neuronal attachment and neurite growth is not clear. There appears to be some evidence that unmodified CNTs are sufficient to support neural growth but little is known about what factors might be adsorbed to the nanotubes under the various culture conditions that have been used. Various methods have been developed to functionalize CNTs [112]. BSA-nanotube conjugates have been produced using diimide-activated amidation with no major loss of protein function [113]. Recently, a number of “biofibers” have been developed through the use of different molecular dispersants [114]. A recent study found that CNTs can be modified with neurotrophins that retain their biological activity and stimulate neurite growth from embryonic chick sensory neurons [93]. However, it should be noted that this effect was produced by adding the modified CNTs to the medium. The CNTs were not used as a substrate, which in this study was laminin. Nor is it clear from this study whether the neurotrophins acted as bound or soluble factors. The question of direct interest to work described here is whether the geometry of the CNTs is relevant to their suitability as a substrate for neural growth. Based on the geometry hypothesis outlined above, it seemed that nanoscale materials that mimic the spatial features of neural tissue, particularly the structure of CNS white matter and peripheral nerves, would be more suitable than materials that do not mimic those features. This is not to say that the chemical properties are unimportant
7.2 Recent Progress
193
but instead that they might be insufficient in designing optimized materials to support neural regeneration. 7.1.7
CNT Cytotoxicity
The design of material supports for neural repair must also take into consideration possible cytotoxic effects. One study suggests that some nanoparticles can affect brain function [115]. However, it is likely that the effects obtained with dispersed metallic nanoparticles will be different from that of CNTs, which do not undergo appreciable degradation, are mostly inert, and would unlikely have wide systemic effects. There are also some reports that AP-CNTs exhibit cytotoxic effects when tested with certain cells such as alveolar macrophages [116], HaCaT cells [117], and HEK293 cells [118]. However, there is very little evidence for toxicity of the CNT materials that have been used for neuronal cultures and the CNT arrays produced by the group at UC are almost catalyst metal-free because of the unique process used to manufacture them. Furthermore, a recent study suggests that there are few if any long-term consequences of injecting functionalized (PEG) CNTs in mice [119]. The prosthetic material used in the work described here (bundles, threads, and sheets) is made of millimeter-long CNTs that are chemically inert and strongly interconnected by van der Waals forces. The latter prevents them from easily separating, which should limit their circulation in the body. In addition, biocompatible coatings could be used to create a thin insulating film on the material to encapsulate it and minimize potential toxicity without affecting its morphology. Even so, the development of materials intended for chronic residence in the nervous system must include assessment of possible cytotoxicity.
7.2
Recent Progress The nanoengineering group at the University of Cincinnati has been working on carbon nanotube array growth and its functionalization for several years [89, 120–123]. Different types of nanotubes, including single-wall and multiwall nanotubes, as well as carbon nanofibers, have different geometric, mechanical, electrical, and chemical properties. As far as we are aware, there are no studies that have taken advantage of the geometrical properties of millimeter-long and aligned CNTs to study axonal regeneration. The ability to pattern catalysts using electron-beam lithography provides the opportunity to manipulate parameters of vertically aligned carbon nanotube such as spacing, length, and bundle diameter to approximate the required cellular and subcellular scale. In addition, CNTs are electrically conductive, a property that might be used to both stimulate and record such prosthetic material ultimately placed within the nervous system. Several lines of evidence (reviewed above) point to the importance of tissue geometry in permitting axonal regeneration. In order to determine the feasibility of using CNTs as a substrate for neural growth, we have sought to determine whether aligned CNTs can support attachment and growth of primary neurons in tissue culture. The results demonstrate that such material not only supports neuronal attach-
196
Progress in the Use of Aligned Carbon Nanotubes
ment and neurite growth but is able to direct growth along the direction of alignment. 7.2.1
Preparation of CNTs
AP-CNTs were prepared in the Nanoworld Laboratory at UC that includes a furnace with a multichannel gas delivery system, a Quadrupole Mass Spectrometer, and computer control of all process variables. Multiwall CNTs (MWCNT) were produced using oxidized silicon wafers as substrates. Electron beam (E-beam) deposition was used to form both the intermediate Al2O3 layer and the top catalytic iron film. The CNTs were grown by thermal CVD from a H2-C2H4-H2O-Ar gas mixture at 750° C for 5 to 10 hours in an EasyTubeTM furnace from First Nano Inc. The resulting CNT arrays consisted of aligned MWCNTs with a length of several millimeters. For some studies, CNT “threads” were produced by “spinning” the material as it was being pulled from a CNT array. For yet other experiments, CNT “sheets” were “drawn” according to the method described by Zhang et al. [124]. 7.2.2
Neuronal Cultures
The CNT arrays were mechanically separated into bundles that were mounted directly on to the bottom of 35-mm dishes using silicone grease or deposited on the poly-ornithine substrate that is normally used to promote neuronal attachment and neurite outgrowth. Threads were mounted in a similar fashion. CNT sheets were placed directly on glass coverslips and secured by beads of silicone grease. The coverslips were then secured to the bottom of 35-mm dishes using additional spots of silicone grease. Primary neuron cultures were prepared from embryonic chick sympathetic ganglia, which were dissociated before plating them on the CNT material. Most of the cultures were also supplemented with nerve growth factor (NGF) and grown for 2 to 8 days. Living cells and neurites were visualized by treating with a vital dye that is taken up and converted to a fluorescent derivate by living cells. This permits visualization of the neurons and their processes using fluorescence microscopy. In some cases, both phase contrast and fluorescence illumination were used at the same time to reveal the relationship of the growing neurons to the substrate material. Due to the optical density of the nanotubes, the surface of CNT bundles was illuminated by incident light for some images. 7.2.3
Neuronal Attachment and Neurite Outgrowth on Aligned CNTs
Neurons cultured on poly-ornithine in the presence of CNTs showed good attachment and neurite growth on the substrate (Figure 7.2 (a, b)) as normally observed with cultures of this type. Although there appeared to be some examples of neurites growing in the vicinity of the CNTs, there was no preferential association with the nanotubes, nor did the neurites appear to be specifically following or avoiding the CNT bundles. It is possible that the poly-l-ornithine prevented detection of the nanotubes by the neurons or that the topography established by the nanotubes was not reliably detected by the cells.
7.2 Recent Progress
197
Figure 7.2 (Color plate 10) Embryonic chick sympathetic neurons cultured in the presence of carbon nanotube bundles on a substrate of poly-l-ornithine show no specific association with the nanotubes (a, b). However, when grown in the presence of suspended CNT bundles but without any poly-l-ornithine, neurons attach and grow on the nanotubes (c, d). Two neurons are shown clustered together and extending neurites in both directions along the nanotubes (inset). The neurons and their neurites have been labeled with a fluorescent vital dye. (a) and (c) are phase-contrast images of the same fields shown in fluorescence in (b) and (d).
In contrast, in cultures in which no poly-l-ornithine was provided (the untreated plastic of the dishes does not normally support attachment and growth of neurons), neurons were found to attach to suspended CNT bundles. In some cases, neurons also extended short neurites along the long axis of the bundles. Examples are shown in Figure 7.2 (c, d) and Figure 7.3. Quite often more than one cell was attached to the same region of the CNT bundle. When neurites were present they always extended along the longitudinal axis of the bundle and showed virtually no evidence of branching. Some of the neurites showed a beaded appearance, such as the example shown in Figure 7.3 (b, c). Usually this is evidence of degenerative changes; however, this appearance was more the exception than the rule and it is not uncommon to see this in vital dye-stained preparations under prolonged fluorescence illumination. We saw no evidence of cytotoxicity that could be attributed to the presence of the nanotubes. Additional cultures were established in which higher densities of neurons were plated. In such cultures there were often clusters of cells that gave rise to thick bundles of neurites such as those shown in Figure 7 (a, b). Where CNT bundles had pulled away from the main group, the neurites were often found to follow the tortuous path of the separated tubes. When CNT “threads” were used as the substrate, neurons that attached to the threads gave rise to neurites that wrapped around the thread, following the spiral pattern established by the twisted CNTs as shown in figure 7.4 (c, d). Finally, we also examined the potential growth of neurons on CNT sheets, prepared according to the method of Zhang et al. [124]. As in the case of the CNT bun-
198
Progress in the Use of Aligned Carbon Nanotubes
Figure 7.3 Several examples of embryonic chick sympathetic neurons attached and extending neurites on CNT bundles are shown using fluorescence illumination following labeling with a vital dye. Some cells occur in clusters (a) and (b), but others appear to be solitary (e). Neurites extend along the longitudinal axis of the CNT bundles. Part (c) shows a higher magnification of the neurite extending from the cell in the lower right corner of the field in part (b). The neurite initially extended along one bundle and then appeared to cross to another bundle to continue along its longitudinal axis. The “beading” of the neurite is often observed after exposure to ultraviolet radiation and is unlikely the result of a toxic effect of the CNTs.
dles and threads, neuronal attachment occurred in the presence of the CNT sheets. However, there were two different patterns of neurite growth. Some of the cells extended neurites without any apparent influence of the nanotubes, as shown in Figure 7.5(a). Neurites extending from other cells, however, showed preferential alignment with the sheet. Closer examination revealed that the main difference between the two types of neurite growth was whether the neurites were growing on the underlying glass coverslip or on the CNT sheet. When growing between the sheet and the coverslip, the neurites showed no preferential orientation. When growing on the sheets, however, the neurites were highly aligned. These findings suggest that AP-CNTs are able to support both the attachment and growth of primary neurons in tissue culture. This was a bit surprising in light of the strong hydrophobicity of the nanotubes. Since no attempt was made to “functionalize” the nanotubes for these experiments, the ability of this material to support attachment and growth might be due to adsorption of components from the medium. Although the majority of cultures contained NGF, the presence of this growth factor was not absolutely required for at least some attachment and growth. Perhaps the most encouraging feature of the growth is that the neurites grow along the longitudinal axis of the nanotube bundles with little, if any, branching. In fact, it was rare to see anything other than a straight fiber extending from a cell or a cluster of cells. This outgrowth morphology is reminiscent of that observed from neurons cultured on longitudinal sections of the mammalian spinal cord or peripheral nerve. The straight, unbranched pattern of growth is also characteristic of the dense bundles of nerve fibers that make up both peripheral nerves and CNS white
7.2 Recent Progress
197
Figure 7.4 (Color plate 11) An example of a thick CNT bundle to which clusters of neurons have attached is shown in (a) using phase contrast illumination. The same field is shown in (b), illuminated with fluorescence to show the location of the living cell clusters and their neurites. Due to the optical density of the CNT bundle, only the neurites and cells along the edges are visible. Where portions of the bundle have separated from the main trunk, neurites have followed their contours, indicating their affinity for the substrate and the directional guidance provided by the CNTs. Parts (c) and (d) show an example of neurons that have attached to a CNT thread and extended neurites that follow the spiral path of the nanotubes. The surface of the thread has been illuminated using incident light in part (d) to show the spiral topography.
Figure 7.5 Examples of neurons growing either below (a) or above (b) CNT sheets that were placed on glass coverslips. Only when neurons were attached and growing on the surface of the sheets did they exhibit aligned growth of neurites.
matter. If, as we have proposed, the geometry of the tissue is important to the guidance of regenerating axons, the linear growth observed with the CNT bundles is promising.
200
Progress in the Use of Aligned Carbon Nanotubes
These results are broadly consistent with previous studies demonstrating that CNTs can support neuronal attachment and neurite outgrowth [91, 94, 97, 98, 109, 125]. Consistent with these experiments, we found that neurons will attach and extend neurites without the requirement of chemical modification. Whether adsorption of factors from the medium onto the CNTs contributes to the attachment and growth of the neurons is unknown. Furthermore, we saw no evidence for cytotoxicity induced by the CNTs. Thus, the general conclusion from existing studies is that carbon nanotubes exhibit biocompatibility with neurons and might be suited to serve as an artificial substrate to promote neural regeneration. However, in addition to biocompatibility, there is evidence that any suitable substrate must also provide the appropriate geometry to support directional growth of regenerating axons. In particular, there is reason to believe that a highly aligned substrate is needed to provide suitable guidance. The extent to which artificial substrates have been able to accomplish this goal is not clear. Alignment of collagen and laminin presented in silicone tubes was reported to enhance peripheral nerve regeneration [126] and a recent study reported that aligned fiber films made by electrospinnning poly (acrylonitrile-co-methylacrylate) supported directional neurite growth in vitro and greater nerve regeneration in vivo [87]. Aligned fibers spun from poly(l-lactic acid) were also reported to support directional outgrowth of neural stem cells [81]. The possibility that aligned carbon nanotubes will support directional neurite growth is not clear from the available literature. Galvan-Garcia et al. [109] observed cell migration and neurite growth with CNT sheets and yarns that had been produced by drawing from a sidewall of MWCNT forests [124]. The CNT sheets used by Galvan-Garcia et al. were found to support growth of various cell types, including mouse cortical and cerebellar neurons. However, the extent to which neurite outgrowth was influenced by the alignment of the CNTs is a bit difficult to tell. They reported a faster rate of fibroblast migration (that did not reach statistical significance) and some of the figures suggested cellular orientation along the axis of the nanotube sheet. On the other hand, there did not appear to be a dominant parallel orientation of neurites with the underlying substrate in the examples shown. In fact, their results are strikingly similar to the pattern of neurite growth we observed when the neurites extended beneath the CNT sheets on the glass coverslip. These results suggest that suspended bundles of long, aligned CNTs will support attachment and directional growth of primary neurons in tissue culture. The straight, unbranched growth of neurites likely reflects the ability of the neurons to detect the topography of the underlying substrate. It is possible that such aligned growth is due to the three-dimensional nature of the CNT bundles because the cultures in which CNTs were deposited on the poly-l-ornithine substrate did not result in clear contact guidance. However, the results obtained with the CNT sheets suggest that even a thin layer of aligned CNTs can orient neurites. The lack of such orientation in the study by Galvan-Garcia et al. [109] might be due to growth on the underlying substrate (they did not specify how the sheets were secured) or to the fact that the sheets they used were condensed. Therefore, it is possible that the topographical features were not reliably detected by the neurites growing on what approximates a two dimensional surface.
7.3 Future Directions and Challenges
7.3
201
Future Directions and Challenges The fact that long-aligned CNT bundles are able to support the attachment and growth of primary neurons in tissue culture encourages additional studies to determine the suitability of this material for neural prosthesis. If, as we suspect, the three-dimensional nature of the material is important for promoting directional growth, it should be possible to design and manufacture material of appropriate dimensions that could be used to bridge injured peripheral nerves or damaged areas of CNS white matter. The ability to produce long-parallel arrays of CNTs provides the opportunity to theoretically produce sufficiently long aligned material to bridge distances ranging from a few millimeters to several centimeters. The biocompatibility of the CNTs with neural tissue also suggests that there is unlikely to be significant toxicity with this material. Obviously, this possibility will need to be explored in more detail. It is somewhat surprising that the CNTs are able to support neuronal growth without chemical modification since some studies have suggested that functionalization may be important [97, 98]. As noted above, it is possible that factors from the medium adsorb onto the CNTs, thereby making them a suitable substrate for the neurons. We have also noticed that the most extensive growth occurs in cases where clusters of nerve cells have attached to the material. This is consistent with the fact that cultured neurons grow better when seeded at higher density, presumably due to the exchange of factors that are mutually supportive. Thus, it is possible that the neurons produce factors that contribute to survival and outgrowth. The presence of NGF no doubt contributes to the observed growth although its presence does not seem to be absolutely required. The ability to chemically modify CNTs also provides the opportunity to examine the effect of different molecules that could promote more extensive growth. The possibility of coating carbon nanotubes with neurotrophic factors, for example, has been reported by Matsumoto and colleagues [93]. Whether similar modification of the aligned CNTs used here will enhance the attachment and outgrowth of neurons remains to be examined. In addition, the electrical properties of nanotubes might be useful for recording and/or stimulating associated neurons as recently reported by others [91]. It is intriguing to consider the possibility of using a material that not only supports regeneration but might also exhibit electrical properties that can be used to advantage. Although these results are encouraging, there are significant challenges to the study and development of any neural prosthetic material based on CNTs. For example, the handling and manipulation of the arrays is not trivial. For the results reported here, we manually separated CNT bundles from dense arrays and secured them through the use of silicone grease to the bottom of the culture dish or coverslips. This approach offers very little control over the bundle size and placement. Some improvement in standardizing the dimensions of the CNTs can be expected from the use of patterned substrates that permit growth of CNT posts or the use of methods that permit the production of CNT “threads” and “yarns.” One potential limitation of threads and yarns, however, is the resulting chirality, which appears to result in spiraling of neurites as shown in Figure 7.4(c).
202
Progress in the Use of Aligned Carbon Nanotubes
Another technical limitation that thwarts the study of three-dimensional CNT arrays is their optical density. Neurons and neurites attached to the surface of the arrays can be visualized through the use of a vital dye illuminated by epifluorescence, as was done for the examples shown here. Unless the cells are attached along the edge, however, it is not possible to detect them using standard brightfield or phase-contrast optics. This presents a special problem when trying to detect fibers or cells that may grow within a CNT array designed to serve as a 3-D scaffold. It might be possible to cut thin sections through the material to accomplish this goal. It is also not yet clear what geometric and/or chemical properties are optimal for any neural prosthetic material, including carbon nanotubes. We believe that three-dimensional parallel alignment is a critical variable but this needs to be examined systematically. It is also not obvious whether the attachment of factors that stimulate growth would necessarily result in more productive growth (meaning elongation in the desired direction). The geometry hypothesis asserts that one of the roles of myelin-associated inhibitors in CNS white matter is to limit collateral sprouting and constrain axons to grow along the length of the fiber tract [20, 68]. The presence of a uniformly positive substrate might encourage sprouting and secondarily reduce the longitudinal orientation of the fibers. In that regard, it is interesting to consider the possibility that the reason that aligned CNTs are able to promote longitudinal growth is because the material is not a uniformly permissive substrate. In spite of these remaining challenges, the availability of a material that is biocompatible and that can be produced with the appropriate dimensions to support neurite growth in tissue culture encourages ongoing studies. The results should provide information not only regarding the suitability of specific materials but also about the normal constraints on axonal regeneration, thus paving the way for the development of neural prosthetics that can ameliorate the devastating consequences of injury to the nervous system.
Problems 7.1 One concern with the development of any material intended for use in the body is undesired effects such as cytotoxicity or immunoreactivity. In the case of a nanotube scaffold for neural repair, how might you determine whether or not there is a risk of adverse effects with the use of such material? 7.2 What are the optimal properties of a material intended to be used for promoting neural repair? 7.3 How do you determine whether a synthetic material is effective in promoting neural repair?
Acknowledgments
203
Acknowledgments The contributions of Sarah Rosile, Erik Sass, and Samih El-akkad to the work described here are gratefully acknowledged. These studies were supported, in part, by the NIH (NS049972) and by a grant from the Nanoscale Institute at the University of Cincinnati College of Medicine.
References [1] Ramon y Cajal S.. Degeneration and Regeneration of the Nervous System. London: Oxford University Press, 1928. [2] Behar O., Mizuno K., Neumann S., Woolf C.J. “Putting the spinal cord together again.” Neuron 2000;26 (2):291–293. [3] Bethea J.R. “Spinal cord injury-induced inflammation: a dual-edged sword.” Prog Brain Res 2000;128:33–42. [4] Condic M.L., Lemons M.L. “Extracellular matrix in spinal cord regeneration: getting beyond attraction and inhibition.” Neuroreport 2002;13 (3):A37–48. [5] Domeniconi M., Filbin M.T. “Overcoming inhibitors in myelin to promote axonal regeneration.” J Neurol Sci 2005;233 (1–2):43–47. [6] Emery D.L., Royo N.C., Fischer I., Saatman K.E., McIntosh T.K. “Plasticity following injury to the adult central nervous system: is recapitulation of a developmental state worth promoting?” J Neurotrauma 2003;20 (12):1271–1292. [7] Fawcett J.W., Geller H.M. “Regeneration in the CNS: optimism mounts.” Trends Neurosci 1998;21 (5):179–180. [8] Fawcett J.W., Asher R.A. “The glial scar and central nervous system repair.” Brain Res Bull 1999;49 (6):377–391. [9] Filbin M.T. “Myelin-associated inhibitors of axonal regeneration in the adult mammalian CNS.” Nat Rev Neurosci 2003;4 (9):703–713. [10] Fouad K., Dietz V., Schwab M.E. “Improving axonal growth and functional recovery after experimental spinal cord injury by neutralizing myelin associated inhibitors.” Brain Res Brain Res Rev 2001;36 (2–3):204–212. [11] Goldberg J.L. “How does an axon grow?” Genes Dev 2003;17 (8):941–958. [12] Hu F., Strittmatter S.M. “Regulating axon growth within the postnatal central nervous system.” Semin Perinatol 2004;28 (6):371–378. [13] Hunt D., Coffin R.S., Anderson P.N. “The Nogo receptor, its ligands and axonal regeneration in the spinal cord; a review.” J Neurocytol 2002;31 (2):93–120. [14] Jacobs W.B., Fehlings M.G. “The molecular basis of neural regeneration.” Neurosurgery 2003;53 (4):943–948; discussion 948–950. [15] McDonald J.W. “Repairing the damaged spinal cord.” Sci Am 1999;281 (3):64–73. [16] McKerracher L., Winton M.J. “Nogo on the go.” Neuron 2002;36 (3):345–348. [17] Moon L., Bunge M.B. “From animal models to humans: strategies for promoting CNS axon regeneration and recovery of limb function after spinal cord injury.” J Neurol Phys Ther 2005;29 (2):55–69. [18] Oudega M., Moon L.D., de Almeida Leme R.J. “Schwann cells for spinal cord repair.” Braz J Med Biol Res 2005;38 (6):825–835. [19] Priestley J.V., Ramer M.S., King V.R., McMahon S.B., Brown R.A. “Stimulating regeneration in the damaged spinal cord.” J Physiol Paris 2002;96 (1–2):123–133. [20] Raisman G. “Myelin inhibitors: does NO mean GO?” Nat Rev Neurosci 2004;5 (2):157–161.
204
Progress in the Use of Aligned Carbon Nanotubes [21] Ramer M.S., McMahon S.B., Priestley J.V. “Axon regeneration across the dorsal root entry zone.” Prog Brain Res 2001;132:621–639. [22] Ramer L.M., Ramer M.S., Steeves J.D. “Setting the stage for functional repair of spinal cord injuries: a cast of thousands.” Spinal Cord 2005;43 (3):134–161. [23] Rosenzweig E.S., McDonald J.W. “Rodent models for treatment of spinal cord injury: research trends and progress toward useful repair.” Curr Opin Neurol 2004;17 (2):121–131. [24] Sandvig A., Berry M., Barrett L.B., Butt A., Logan A. “Myelin-, reactive glia-, and scar-derived CNS axon growth inhibitors: expression, receptor signaling, and correlation with axon regeneration.” Glia 2004;46 (3):225–251. [25] Schwab M.E. “Repairing the injured spinal cord.” Science 2002;295 (5557):1029–1031. [26] Silver J., Miller J.H. “Regeneration beyond the glial scar.” Nat Rev Neurosci 2004;5 (2):146–156. [27] Stichel C.C., Muller H.W. “Experimental strategies to promote axonal regeneration after traumatic central nervous system injury.” Prog Neurobiol 1998;56 (2):119–148. [28] Tuszynski M.H., Conner J., Blesch A., Smith D., Merrill D.A., Vahlsing H.L. “New strategies in neural repair.” Prog Brain Res 2002;138:401–409. [29] Zhang N., Yan H., Wen X. “Tissue-engineering approaches for axonal guidance.” Brain Res Brain Res Rev 2005;49 (1):48–64. [30] Filbin M.T. “Myelin-associated glycoprotein: a role in myelination and in the inhibition of axonal regeneration?” Curr Opin Neurobiol 1995;5 (5):588–595. [31] Schwab M.E., Kapfhammer J.P., Bandtlow C.E. “Inhibitors of neurite growth.” Annu Rev Neurosci 1993;16:565–595. [32] Schwab M.E. “Molecules inhibiting neurite growth: a minireview.” Neurochem Res 1996;21 (7):755–761. [33] Bovolenta P., Fernaud-Espinosa I., Mendez-Otero R., Nieto-Sampedro M. “Neurite outgrowth inhibitor of gliotic brain tissue. Mode of action and cellular localization, studied with specific monoclonal antibodies.” Eur J Neurosci 1997;9 (5):977–989. [34] Davies S.J., Fitch M.T., Memberg S.P., Hall A.K., Raisman G., Silver J. “Regeneration of adult axons in white matter tracts of the central nervous system.” Nature 1997;390 (6661):680–683. [35] Stichel C.C., Muller H.W. “The CNS lesion scar: new vistas on an old regeneration barrier.” Cell Tissue Res 1998;294 (1):1–9. [36] Yamada H., Fredette B., Shitara K., Hagihara K., Miura R., Ranscht B., Stallcup W.B., Yamaguchi Y. “The brain chondroitin sulfate proteoglycan brevican associates with astrocytes ensheathing cerebellar glomeruli and inhibits neurite outgrowth from granule neurons.” J Neurosci 1997;17 (20):7784–7795. [37] Milev P., Friedlander D.R., Sakurai T., Karthikeyan L., Flad M., Margolis R.K., Grumet M., Margolis R.U. “Interactions of the chondroitin sulfate proteoglycan phosphacan, the extracellular domain of a receptor-type protein tyrosine phosphatase, with neurons, glia, and neural cell adhesion molecules.” J Cell Biol 1994;127 (6 Pt 1):1703–1715. [38] Snow D.M., Lemmon V., Carrino D.A., Caplan A.I., Silver J. “Sulfated proteoglycans in astroglial barriers inhibit neurite outgrowth in vitro.” Exp Neurol 1990;109 (1):111–130. [39] Jones L.L., Yamaguchi Y., Stallcup W.B., Tuszynski M.H. “NG2 is a major chondroitin sulfate proteoglycan produced after spinal cord injury and is expressed by macrophages and oligodendrocyte progenitors.” J Neurosci 2002;22 (7):2792–2803. [40] Schmalfeldt M., Bandtlow C.E., Dours-Zimmermann M.T., Winterhalter K.H., Zimmermann D.R. “Brain derived versican V2 is a potent inhibitor of axonal growth.” J Cell Sci 2000;113 ( Pt 5):807–816. [41] Hanbali M., Bernard F., Berton C., Gatineau G., Perraut M., Aunis D., Luu B., Bagnard D. “Counteraction of axonal growth inhibitory properties of Semaphorin 3A and
References
[42] [43]
[44] [45]
[46] [47]
[48]
[49]
[50]
[51]
[52]
[53]
[54] [55] [56]
[57]
[58]
[59] [60]
205 myelin-associated proteins by a synthetic neurotrophic compound.” J Neurochem 2004;90 (6):1423–1431. Pasterkamp R.J., Verhaagen J. “Emerging roles for semaphorins in neural regeneration.” Brain Res Brain Res Rev 2001;35 (1):36–54. Shibata A, Wright MV, David S, McKerracher L, Braun PE, Kater SB. “Unique responses of differentiating neuronal growth cones to inhibitory cues presented by oligodendrocytes.” J Cell Biol 1998;142 (1):191–202. Raper J.A., Kapfhammer J.P. “The enrichment of a neuronal growth cone collapsing activity from embryonic chick brain.” Neuron 1990;4 (1):21–29. Caroni P., Schwab M.E. “Two membrane protein fractions from rat central myelin with inhibitory properties for neurite growth and fibroblast spreading.” Journal of Cell Biology 1988;106:1281–1288. Savio T., Schwab M.E. “Rat CNS white matter, but not gray matter, is nonpermissive for neuronal cell adhesion and fiber outgrowth.” J Neurosci 1989;9 (4):1126–1133. Li M., Shibata A., Li C., Braun P.E., McKerracher L., Roder J., Kater S.B., David S. “Myelin-associated glycoprotein inhibits neurite/axon growth and causes growth cone collapse.” J Neurosci Res 1996;46 (4):404–414. McKerracher L., David S., Jackson D.L., Kottis V., Dunn R.J., Braun P.E. “Identification of myelin–associated glycoprotein as a major myelin- derived inhibitor of neurite growth.” Neuron 1994;13 (4):805–811. Mukhopadhyay G., Doherty P., Walsh F.S., Crocker P.R., Filbin M.T. “A novel role for myelin-associated glycoprotein as an inhibitor of axonal regeneration.” Neuron 1994;13 (3):757–767. Shen Y.J., DeBellard M.E., Salzer J.L., Roder J., Filbin M.T. “Myelin-associated glycoprotein in myelin and expressed by Schwann cells inhibits axonal regeneration and branching.” Mol Cell Neurosci 1998;12 (1–2):79–91. Tang S., Woodhall R.W., Shen Y.J., deBellard M.E., Saffell J.L., Doherty P., Walsh F.S., Filbin M.T. “Soluble myelin-associated glycoprotein (MAG) found in vivo inhibits axonal regeneration.” Mol Cell Neurosci 1997;9 (5–6):333–346. Wang K.C., Koprivica V., Kim J.A., Sivasankaran R., Guo Y., Neve R.L., He Z. “Oligodendrocyte-myelin glycoprotein is a Nogo receptor ligand that inhibits neurite outgrowth.” Nature 2002;417 (6892):941–944. Kottis V., Thibault P., Mikol D., Xiao Z.C., Zhang R., Dergham P., Braun P.E. “Oligodendrocyte-myelin glycoprotein (OMgp) is an inhibitor of neurite outgrowth.” J Neurochem 2002;82 (6):1566–1569. Fournier A.E., GrandPre T., Strittmatter S.M. “Identification of a receptor mediating Nogo-66 inhibition of axonal regeneration.” Nature 2001;409 (6818):341–346. Wang K.C., Kim J.A., Sivasankaran R., Segal R., He Z. “P75 interacts with the Nogo receptor as a co-receptor for Nogo, MAG and OMgp.” Nature 2002;420 (6911):74–78. Wong S.T., Henley J.R., Kanning K.C., Huang K.H., Bothwell M., Poo M.M. “A p75(NTR) and Nogo receptor complex mediates repulsive signaling by myelin-associated glycoprotein.” Nat Neurosci 2002;5 (12):1302–1308. Wictorin K, Brundin P, Gustavii B, Lindvall O, Bjorklund A. “Reformation of long axon pathways in adult rat central nervous system by human forebrain neuroblasts.” Nature 1990;347 (6293):556–558. Wictorin K, Brundin P, Sauer H, Lindvall O, Bjorklund A. “Long distance directed axonal growth from human dopaminergic mesencephalic neuroblasts implanted along the nigrostriatal pathway in 6-hydroxydopamine lesioned adult rats.” J Comp Neurol 1992;323 (4):475–494. Davies SJ, Field PM, Raisman G. “Long interfascicular axon growth from embryonic neurons transplanted into adult myelinated tracts.” J Neurosci 1994;14 (3 Pt 2):1596–1612. Davies SJ, Silver J. “Adult axon regeneration in adult CNS white matter.” Trends Neurosci 1998;21 (12):515.
206
Progress in the Use of Aligned Carbon Nanotubes [61] Davies SJ, Goucher DR, Doller C, Silver J. “Robust regeneration of adult sensory axons in degenerating white matter of the adult rat spinal cord.” J Neurosci 1999;19 (14):5810–5822. [62] Fischer D, Heiduschka P, Thanos S. “Lens-Injury-Stimulated Axonal Regeneration throughout the Optic Pathway of Adult Rats.” Exp Neurol 2001;172 (2):257–272. [63] Leon S, Yin Y, Nguyen J, Irwin N, Benowitz LI. “Lens injury stimulates axon regeneration in the mature rat optic nerve.” J Neurosci 2000;20 (12):4615–4626. [64] Davies SJ, Field PM, Raisman G. “Long fibre growth by axons of embryonic mouse hippocampal neurons microtransplanted into the adult rat fimbria.” Eur J Neurosci 1993;5 (2):95–106. [65] Carpenter MK, Hassinger TD, Whalen LR, Kater SB. “CNS white matter can be altered to support neuronal outgrowth.” J Neurosci Res 1994;37 (1):1–14. [66] Crutcher KA. “Tissue sections as culture substrates: overview and critique.” Hippocampus 1993;3 (Spec No):157–163. [67] Brewer GJ, Torricelli JR, Evege EK, Price PJ. “Optimized survival of hippocampal neurons in B27-supplemented Neurobasal, a new serum-free medium combination.” J Neurosci Res 1993;35 (5):567–576. [68] Pettigrew DB, Crutcher KA. “White matter of the CNS supports or inhibits neurite outgrowth in vitro depending on geometry.” J Neurosci 1999;19 (19):8358–8366. [69] Pettigrew DB, Crutcher KA. “Myelin contributes to the parallel orientation of axonal growth on white matter in vitro.” BMC Neurosci 2001;2 (1):9. [70] Pettigrew DB, Shockley KP, Crutcher KA. “Disruption of spinal cord white matter and sciatic nerve geometry inhibits axonal growth in vitro in the absence of glial scarring.” BMC Neurosci 2001;2 (1):8. [71] Nomura H, Tator CH, Shoichet MS. “Bioengineered strategies for spinal cord repair.” J Neurotrauma 2006;23 (3–4):496–507. [72] Ngo TT, Waggoner PJ, Romero AA, Nelson KD, Eberhart RC, Smith GM. “Poly(L-Lactide) microfilaments enhance peripheral nerve regeneration across extended nerve lesions.” J Neurosci Res 2003;72 (2):227–238. [73] Khan T, Dauzvardis M, Sayers S. “Carbon filament implants promote axonal growth across the transected rat spinal cord.” Brain Res 1991;541 (1):139–145. [74] Chauhan NB, Figlewicz HM, Khan T. “Carbon filaments direct the growth of postlesional plastic axons after spinal cord injury.” Int J Dev Neurosci 1999;17 (3):255–264. [75] Yoshii S, Oka M. “Peripheral nerve regeneration along collagen filaments.” Brain Res 2001;888 (1):158–162. [76] Yoshii S, Oka M. “Collagen filaments as a scaffold for nerve regeneration.” J Biomed Mater Res 2001;56 (3):400–405. [77] Yoshii S, Oka M, Shima M, Akagi M, Taniguchi A. “Bridging a spinal cord defect using collagen filament.” Spine 2003;28 (20):2346–2351. [78] Yoshii S, Oka M, Shima M, Taniguchi A, Taki Y, Akagi M. “Restoration of function after spinal cord transection using a collagen bridge.” J Biomed Mater Res A 2004;70 (4):569–575. [79] Yoshii S, Shima M, Oka M, Taniguchi A, Taki Y, Akagi M. “Nerve regeneration along collagen filament and the presence of distal nerve stump.” Neurol Res 2004;26 (2):145–150. [80] Bini TB, Gao S, Xu X, Wang S, Ramakrishna S, Leong KW. “Peripheral nerve regeneration by microbraided poly(L-lactide-co-glycolide) biodegradable polymer fibers.” J Biomed Mater Res A 2004;68 (2):286–295. [81] Yang F, Murugan R, Wang S, Ramakrishna S. “Electrospinning of nano/micro scale poly(L-lactic acid) aligned fibers and their potential in neural tissue engineering.” Biomaterials 2005;26:2603–2610.
References
207
[82] Schnell E, Klinkhammer K, Balzer S, Brook G, Klee D, Dalton P, Mey J. “Guidance of glial cell migration and axonal growth on electrospun nanofibers of poly-epsilon-caprolactone and a collagen/poly-epsilon-caprolactone blend.” Biomaterials 2007;28 (19):3012–3025. [83] Tian F, Hosseinkhani H, Hosseinkhani M, Khademhosseini A, Yokoyama Y, Estrada GG, Kobayashi H. “Quantitative analysis of cell adhesion on aligned micro- and nanofibers.” Journal of Biomedical Materials Research Part A 2008;84A:291–299. [84] Ellis-Behnke RG, Liang YX, You SW, Tay DK, Zhang S, So KF, Schneider GE. “Nano neuro knitting: peptide nanofiber scaffold for brain repair and axon regeneration with functional return of vision.” Proc Natl Acad Sci U S A 2006;103 (13):5054–5059. [85] Ellis-Behnke RG, Liang YX, Tay DK, Kau PW, Schneider GE, Zhang S, Wu W, So KF. “Nano hemostat solution: immediate hemostasis at the nanoscale.” Nanomedicine 2006;2 (4):207–215. [86] Guo J, Su H, Zeng Y, Liang YX, Wong WM, Ellis-Behnke RG, So KF, Wu W. “Reknitting the injured spinal cord by self-assembling peptide nanofiber scaffold.” Nanomedicine 2007;3 (4):311–321. [87] Kim YT, Haftel VK, Kumar S, Bellamkonda RV. “The role of aligned polymer fiber-based constructs in the bridging of long peripheral nerve gaps.” Biomaterials 2008;29 (21):3117–3127. [88] Yun Y, Dong Z, Shanov V, Heineman WR, Halsall HB, Bhattacharya A, Conforti L, Narayan RJ, Ball WS, Schulz MJ. “Nanotube electrodes and biosensors.” Nanotoday 2007;2 (6):1–8. [89] Yun Y, Bange A, Shanov VN, Heineman WR, Halsall HB, Dong Z, Jazieh A, Tu Y, Wong D, Pixley S, Behbehani M, Schulz MJ. “A carbon nanotube needle biosensor.” J Nanosci Nanotechnol 2007;7 (7):2293–2300. [90] Gao M, Dai L, Wallace GG. “Biosensors based on aligned carbon nanotubes coated with inherently conducting polymers.” Electroanalysis 2003;15 (13):1089–1094. [91] Mazzatenta A, Giugliano M, Campidelli S, Gambazzi L, Businaro L, Markram H, Prato M, Ballerini L. “Interfacing neurons with carbon nanotubes: electrical signal transfer and synaptic stimulation in cultured brain circuits.” J Neurosci 2007;27 (26):6931–6936. [92] Chen X, Tam UC, Czlapinski JL, Lee GS, Rabuka D, Zettl A, Bertozzi CR. “Interfacing carbon nanotubes with living cells.” J Am Chem Soc 2006;128 (19):6292–6293. [93] Matsumoto K, Sato C, Naka Y, Kitazawa A, Whitby RL, Shimizu N. “Neurite outgrowths of neurons with neurotrophin-coated carbon nanotubes.” J Biosci Bioeng 2007;103 (3):216–220. [94] Dubin RA, Callegari G, Kohn J, Neimark A. “Carbon nanotube fibers are compatible with Mammalian cells and neurons.” IEEE Trans Nanobioscience 2008;7 (1):11–14. [95] Malarkey EB, Parpura V. “Applications of carbon nanotubes in neurobiology.” Neurodegener Dis 2007;4 (4):292–299. [96] Bekyarova E, Ni Y, Malarkey EB, Montana V, McWilliams JL, Haddon RC, Parpura V. “Applications of carbon nanotubes in biotechnology and biomedicine.” Journal of Biomedical Nanotechnology 2005;1:3–17. [97] Mattson MP, Haddon RC, Rao AM. “Molecular functionalization of carbon nanotubes and use as substrates for neuronal growth.” J Mol Neurosci 2000;14 (3):175–182. [98] Hu H, Ni Y, Montana V, Haddon RC, Parpura V. “Chemically functionalized carbon nanotubes as substrates for neuronal growth.” Nano Lett 2004;4 (3):507–511. [99] Hu H, Ni Y, Mandal SK, Montana V, Zhao B, Haddon RC, Parpura V. “Polyethyleneimine functionalized single-walled carbon nanotubes as a substrate for neuronal growth.” J Phys Chem B 2005;109 (10):4285–4289. [100] Webster TJ, Waid MC, McKenzie JL, Price RL, Ejiofor JU. “Nano-biotechnology: carbon nanofibers as improved neural and orthopaedic implants.” Nanotechnology 2004;15:48–54. [101] Jan E, Kotov NA. “Successful differentiation of mouse neural stem cells on layer-by-layer assembled single-walled carbon nanotube composite.” Nano Lett 2007;7 (5):1123–1128.
208
Progress in the Use of Aligned Carbon Nanotubes [102] Xie J, Chen L, Aatre KR, Srivatsan M, Varadan VK. “Somatosensory neurons grown on functionalized carbon nanotube mats.” Smart Materials and Structures 2006;15:N85–N88. [103] Gabay T, Jakobs E, Ben-Jacob E, Hanein Y. “Engineered self-organization of neural networks using carbon nanotube clusters.” Physica A 2005;350:611–621. [104] Sorkin R, Gabay T, Blinder P, Baranes D, Ben-Jacob E, Hanein Y. “Compact self-wiring in cultured neural networks.” J Neural Eng 2006;3 (2):95–101. [105] Zhang X, Prasad S, Niyogi S, Morgan A, Ozkan M, Ozkan CS. “Guided neurite growth on patterned carbon nanotubes.” Sensors and Actuators B 2005;106:843–850. [106] Nguyen-Vu TD, Chen H, Cassell AM, Andrews R, Meyyappan M, Li J. “Vertically aligned carbon nanofiber arrays: an advance toward electrical-neural interfaces.” Small 2006;2 (1):89–94. [107] Nguyen-Vu TD, Chen H, Cassell AM, Andrews RJ, Meyyappan M, Li J. “Vertically aligned carbon nanofiber architecture as a multifunctional 3-D neural electrical interface.” IEEE Trans Biomed Eng 2007;54 (6 Pt 1):1121–1128. [108] Gheith MK, Sinani VA, Wicksted JP, Matts RL, Kotov NA. “Single-walled carbon nanotube polyelectrolyte multilayers and freestanding films as a biocompatible platform for neuroprosthetic implants.” Advanced Materials 2005;17:2663–2670. [109] Galvan-Garcia P, Keefer EW, Yang F, Zhang M, Fang S, Zakhidov AA, Baughman RH, Romero MI. “Robust cell migration and neuronal growth on pristine carbon nanotube sheets and yarns.” J Biomater Sci Polym Ed 2007;18 (10):1245–1261. [110] Kang SJ, Kocabas C, Kim HS, Cao Q, Meitl MA, Khang DY, Rogers JA. “Printed multilayer superstructures of aligned single-walled carbon nanotubes for electronic applications.” Nano Lett 2007;7 (11):3343–3348. [111] Vijayaraghavan A, Blatt S, Weissenberger D, Oron-Carl M, Hennrich F, Gerthsen D, Hahn H, Krupke R. “Ultra-large-scale directed assembly of single-walled carbon nanotube devices.” Nano Lett 2007;7 (6):1556–1560. [112] Sun YP, Fu K, Lin Y, Huang W. “Functionalized carbon nanotubes: properties and applications.” Acc Chem Res 2002;35 (12):1096–1104. [113] Huang W, Taylor S, Fu K, Zhang D, Hanks TW, Rao AM, Sun Y. “Attaching proteins to carbon nanotubes via diimide-activated amidation.” Nano Lett 2002;2 (4):311–314. [114] Lynam C, Moulton SE, Wallace GG. “Carbon-nanotube biofibers.” Advanced Materials 2007;19:1244–1248. [115] Sharma HS, Sharma A. “Nanoparticles aggravate heat stress induced cognitive deficits, blood-brain barrier disruption, edema formation and brain pathology.” Prog Brain Res 2007;162:245–273. [116] Jia G, Wang H, Yan L, Wang X, Pei R, Yan T, Zhao Y, Guo X. “Cytotoxicity of carbon nanomaterials: single-wall nanotube, multi-wall nanotube, and fullerene.” Environ Sci Technol 2005;39 (5):1378–1383. [117] Shvedova AA, Castranova V, Kisin ER, Schwegler-Berry D, Murray AR, Gandelsman VZ, Maynard A, Baron P. “Exposure to carbon nanotube material: assessment of nanotube cytotoxicity using human keratinocyte cells.” J Toxicol Environ Health A 2003;66 (20):1909–1926. [118] Cui D, Tian F, Ozkan CS, Wang M, Gao H. “Effect of single wall carbon nanotubes on human HEK293 cells.” Toxicol Lett 2005;155 (1):73–85. [119] Liu Z, Davis C, Cai W, He L, Chen X, Dai H. “Circulation and long-term fate of functionalized, biocompatible single-walled carbon nanotubes in mice probed by Raman spectroscopy.” Proc Natl Acad Sci U S A 2008. [120] Shanov V, Yun Y, Schulz MJ. “Synthesis and characterization of carbon nanotube materials (review).” J Univ Chem Technol Metal 2006;41 (4):377–390. [121] Yun Y, Gollapudi R, Shanov V, Schulz MJ, Dong Z, Jazieh A, Heineman WR, Halsall HB, Wong DK, Bange A, Tu Y, Subramaniam S. “Carbon nanotubes grown on stainless steel to
References
[122]
[123]
[124]
[125]
[126]
209 form plate and probe electrodes for chemical/biological sensing.” J Nanosci Nanotechnol 2007;7 (3):891–897. Yun Y, Shanov V, Tu Y, Schulz MJ, Yarmolenko S, Neralla S, Sankar J, Subramaniam S. “A multi-wall carbon nanotube tower electrochemical actuator.” Nano Lett 2006;6 (4):689–693. Yun Y, Shanov V, Tu Y, Subramaniam S, Schulz MJ. “Growth mechanism of long aligned multiwall carbon nanotube arrays by water-assisted chemical vapor deposition.” J Phys Chem B 2006;110 (47):23920–23925. Zhang M, Fang S, Zakhidov AA, Lee SB, Aliev AE, Williams CD, Atkinson KR, Baughman RH. “Strong, transparent, multifunctional, carbon nanotube sheets.” Science 2005;309 (5738):1215–1219. Lovat V, Pantarotto D, Lagostena L, Cacciari B, Grandolfo M, Righi M, Spalluto G, Prato M, Ballerini L. “Carbon nanotube substrates boost neuronal electrical signaling.” Nano Lett 2005;5 (6):1107–1110. Verdu E, Labrador RO, Rodriguez FJ, Ceballos D, Fores J, Navarro X. “Alignment of collagen and laminin-containing gels improve nerve regeneration within silicone tubes.” Restor Neurol Neurosci 2002;20 (5):169–179.
CHAPTER 8
RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor for Nanotechnology and Gene Delivery Hui Zhang, Xuesong Feng, Yi Shu, Faqing Yuan, and Peixuan Guo
8.1
Introduction Viruses usually consist of the genomic DNA or RNA, and protein shells (capsid) that protect the genomes inside. The replication of a virus is achieved by the synthesis of genome and viral proteins in the host cells and afterward the virus assembly to virions [1]. An essential process in the assembly of many linear double-stranded DNA (dsDNA) viruses is that the viral genome is inserted into a preformed protein shell, called procapsid, during maturation (Figure 8.1) [2–7]. Such translocation of DNA is energetically unfavorable as the DNA is packaged to near-crystalline density (~400 mg/ml) inside the capsid [1], and thus is powered by the energy converted from ATP hydrolysis [8].
Scaffold protein
DNA gp16
pRNA gp16
Prohead
pRNA
DNA-filled head
Scaffold protein
ATP
ADP + Pi
Empty capsid pRNA
Figure 8.1 Pathway of bacteriophage phi29 DNA packaging [29]. (Adapted with permission from the American Veterinary Medical Association.)
211
212
RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor
In 1987, a small viral RNA was discovered to be involved in the DNA packaging of bacteriophage phi29 [9]. This small viral RNA was found indispensable for DNA packaging and is called packaging RNA (pRNA). The pRNA was first discovered as the packaging of phi29 DNA was sensitive to treatment with RNase A or T1. A DNase-resistant 120-base nucleic acid was found in the electrophoresis gel of the procapsids [9]. This nucleic acid can be isolated from the procapsids by incubation with EDTA, and can bind to RNA-free procapsids in the presence of Mg++. The pRNA-free procapsids were found inactive in DNA packaging, and restored the activities in packaging and virion assembly after addition of pRNA. Bacteriophage phi29 has been widely studied as a model system to understand the mechanical and physical behavior of motor components during viral packaging. It has also been explored for potential applications in nanotechnology and gene delivery. This chapter will focus on the studies of phi29 pRNA and its functions in the motor system, as well as applications of the phi29 motor in nanotechnology and gene delivery.
8.2
Components Related to the phi29 DNA Packaging Motor 8.2.1
phi29 pRNA
Although phi29 bears the common characteristics of dsDNA viruses, it is unique as a small packaging RNA (pRNA) was found to be indispensable for the DNA packaging motor. The pRNA is not present in the mature virion [9]. However, it binds to the connector on the procapsid for the motor to function [10, 11] and remains associated with DNA-filled capsids [12], but leaves the capsid upon completion of packaging and after the addition of the tail protein gp9 [12]. This pRNA also binds ATP [13] and stimulate the ATPase activity of gp16 [14]. Similar but distinct pRNAs have been found through phylogenetic analyses of other phages. Their observed secondary structures are quite similar despite that few conserved bases were found [15, 16]. Nonetheless, phi29 pRNA is not replaceable by other phages’ pRNA for its DNA packaging [15]. The pRNA was transcribed primarily as a 174-base molecule in the phi29-infected cells. However, nuclease cleavage during procapsid purification eliminated 54 bases from its 3´-end [17]. Both the 174- and 120-base RNA molecules were found have the same activity in DNA packaging in vitro [17, 18]. This chapter will focus on studies of the 120-base pRNA molecules. Following the method developed to construct circular tRNA for new termini [19, 20], a series of pRNA molecules with circular permutations (cp-pRNA) have been constructed, as pRNA’s 5´ and 3´- ends are in close proximity [21, 22]. Two tandem pRNA-coding sequences separated by a 3-base or 17-base loop sequence have been cloned into a plasmid. The cp-pRNAs with different terminis could be transcribed from the PCR fragment using different primer pairs. It has been shown that neither a small nor a large linker loop interferes with the biological activity of such transcribed pRNA molecules. Most of the cp-pRNAs with various termini designs were found biologically active [21]. The designs of the cp-pRNAs are especially helpful in labeling the pRNAs at desired base positions along its sequence and in the subsequent studies.
8.2 Components Related to the phi29 DNA Packaging Motor
8.2.2
213
phi29 Procapsid
Phage procapsids are largely composed of three structural proteins: (1) the capsid protein, (2). the scaffolding protein, and (3). the connector protein [3, 5, 23]. The three-dimensional structure of phi29’s empty procapsid has been obtained by reconstruction of Cryo-Electron Microscopy images (Cryo-EM) [24]. The icosahedral procapsid of phi29 consists of 235 copies of the major capsid protein (gp8), 180 copies of scaffolding protein (gp7), and 12 copies of the head-tail connector protein (gp10) [24]. The procapsid is assembled by the three proteins together in a way that is different from the single assembly pathway. The coexistence of scaffolding protein, major capsid protein, and connector over a threshold concentration level is required to form the active procapsid. During such a process, scaffolding protein links major capsid protein and connector [25]. The three components interact so quickly that no clear intermediates are produced. 8.2.2.1
Capsid Protein (gp8)
Capsid proteins are the major structural components for the formation of viral protein shells, such as gpE for bacteriophage lambda [26, 27], gene 23 (major capsid subunit) or gene 24 (pentamer subunit) product for bacteriophage T4 [28], and gp8 for bacteriophage phi29 [25]. The protein shells protect the viral genomes from the degradation by the ubiquitous nuclease. 8.2.2.2
Scaffolding Protein (gp7)
Scaffolding proteins are the structural proteins required for correct procapsid assembly. It is known that scaffolding protein is released from procapsids either at the initiation of or during DNA packaging, and it is not part of the mature virion. For phages phi29 and lambda, the exact function of the scaffolding protein in DNA-packaging is not yet clear, and it is not required for the procapsids function [5]. The scaffolding protein may form a core structure around which capsid proteins assemble [30–35]. It may promote the correct folding of capsid proteins, be involved in processes such as mediating a putative capsid protein/connector protein interaction, excluding cellular proteins from the inside of the procapsid, or facilitating the early stages of DNA entrance into the procapsid [36]. 8.2.2.3
Connector Protein (gp10)
The procapsid contains a single portal vertex, which is also called a connector due to its role in binding. Studies on bacteriophages such as T4, Lambda, T3, P22, and phi29 showed that the connector protein is not only involved in the formation of the procapsid, but also in DNA packaging, and in binding the tail proteins upon maturation [3, 6, 37–43]. It was found interacting with other viral components during DNA translocation [2, 44, 45]. The connectors of T4, Lambda, P22, and phi29 share a common 3-D structure even though they lack of similarity in their amino acid sequences. They all contain a penetrating region at the center, a 12-fold domain composed of 12 identical morphological units perpendicular to the axis of the virion particle (the axis direction is parallel to the direction of DNA packaging into the
214
RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor
procapsid), and a narrower domain with a cylindrical shape along the same axis [46–48]. Several techniques such as atomic force microscopy [49], TEM [50], Cryo-EM [51], immunoelectron microscopy [52], and X-ray crystallography [54] have revealed the three-dimensional structure of the phi29 connector. An immunoelectron microscopy study showed that the phi29 connector is divided into three regions: the narrow end, the central area, and the wide end [52]. The narrow end protrudes from the portal vertex of the procapsid, with the wide end of the connector buried inside [51, 56, 57]. Four domains were found in the phi29 connector: the procapsid domain, which is essential for correct procapsid formation [58, 59]; the tail domain which, is responsible for collar and tail proteins binding [60; 61]; the DNA-binding domain, which is important for the recognition of DNA at the initial stage of packaging [45]; and the RNA-binding domain for specific pRNA binding, which is essential in DNA packaging [9, 62, 63]. Cosedimentation of the pRNA with purified connector in sucrose gradients, and retention of radioactive-labeled pRNA with the connector protein in nitrocellulose binding assays, indicate that the connector is responsible for procapsid binding of pRNA [11, 18, 64]. The data concludes that the affinity of pRNA to the connector is much stronger than to capsid protein, and the connector is the foothold for pRNA. The possibility of transient interaction between pRNA and capsid protein, however, cannot be excluded. 8.2.3
Gp16
In many bacteriophages, two nonstructural DNA packaging proteins usually work together to link the procapsid with the DNA packaging substrate, cut monomers of DNA from concatemers, or serve as a DNA translocation enzyme and/or ATPase [65]. It was found that gp16 of phi29 was such one of these proteins. It contains both A-type and B-type consensus ATP-binding sequences and the predicted secondary structure for ATP binding [66]. The A-type sequence of gp16 is the “basic-hydrophobic region-G-X2-G-X-G-K-S-X7 hydrophobic.” It has been shown that gp16 can bind and hydrolyze ATP [8, 67, 68]. It has also been found that gp16 binds to the pRNA-containing procapsid more strongly than to the pRNA-free procapsid and the pRNA’s 5´/3´ paired region is the domain for such binding [14, 69]. The hydrophobicity, low solubility, and self-aggregation of phi29 gp16 have for a long time hindered the further refinement of the current understanding of the packaging mechanism. Contradictory results regarding ATPase activity, binding location, and the stoichiometry of gp16 have been published [63, 66, 70]. To avoid aggregation from overexpression, gp16 was purified in a denatured condition, and was renatured afterwards [8]. Nonetheless the renatured gp16 aggregated again within 15 minutes after renaturation. It was reported that gp16 was made soluble in the cell by coexpression with groE. However, the method could not solve the problem of self-aggregation after purification [70]. Another approach has been developed to obtain the active and soluble gp16 by fusing thioredoxin to the N-terminus of gp16 [67, 71]. Although the role of gp16 in phi29 DNA-packaging is unclear, studies showed that gp16 could work similarly as AAA+ protein and bind pRNA to form an ATPase complex, which hydrolyzes ATP and translocates phi29 DNA [14].
8.3 Construction of the Biomimetic phi29 DNA Packaging Motor
8.2.4
215
DNA-gp3
The phi29 genome is a 19,285 bp linear dsDNA. It has a viral-encoded terminal protein, gp3, covalently bonded to the 5´ ends. The linkage is between the serine residue 232 of gp3 and the 5’ dAMP of the DNA [72]. The gp3 functions as a “primer” in the initiation of DNA replication [72], and as an enhancer for DNA packaging [73]. Full-length phi29 genomic DNA can be synthesized in vitro [72, 74–78]. The DNA-gp3 is treated as one component of phi29 motor and only one copy of procapsid and one copy of DNA-gp3 is required for the assembly of one virion [78–80]. 8.2.5
Fiber (gp8.5), and Neck and Tail (gp9, gp11-12) Proteins
The head fibers (gp8.5) branch out from the apical regions of the viral head [81–83] and may work as a stabilizer for the procapsids. The tail proteins are usually the ones that penetrate the host cell wall for viral DNA delivery. Fiber and tail proteins are not involved in viral DNA packaging in the phi29 system. After DNA packaging, 6 copies of lower collar (gp11), 12 appendages (gp12), and 10 copies of tail knob protein (gp9) are assembled onto the DNA-filled procapsid in a single morphogenetic pathway to yield one mature virion [58, 84–86]. For neck and tail (gp11-12) assembly, a morphogenetic factor (gp13) is also required [58, 78]. The morphogenetic factor gp13 may also assist holding the genome inside the capsid against the high pressure [87].
8.3
Construction of the Biomimetic phi29 DNA Packaging Motor There are two major approaches for viral genome packaging. For the helical and icosahedral ssRNA viruses, capsid proteins are usually assembled around the viral genome. For dsDNA viruses, which is the most common approach for packaging, their genomes are delivered into the preformed procapsids with the aid of certain packaging motors, which will be discussed in detail here. Packaging of dsDNA serves as an exciting model to understand the fundamental biological mechanisms of bionanomotors that transduce energy from ATP hydrolysis into mechanical motion of DNA. Fully defined in vitro DNA packaging systems have been developed and constructed for various bacteriophages, such as T3 [88], T4 [89], λ [90], and phi29 [8]. The phi29 motor has been found to be by far the strongest molecular motor, packaging DNA against an external force of about 57pN [56]. The infectious phi29 virions can be assembled by the purified gene coded components. To perform such assembly, the pRNA is first allowed to bind to procapsid and then mixed with gp16, DNA-gp3, and a reaction buffer (containing ATP as the energy source) to complete the DNA packaging. The tail protein gp9, neck proteins gp11 and gp12, and morphogenic factor gp13 are added sequentially to convert the DNA-filled procapsids into infectious virions. The yield is then assayed by standard plaque formation. Up to 90% of the added DNA-gp3 can be packaged into the procapsid and up to 109 pfu/ml of the infectious phi29 virions can be assembled in vitro with the exclusive use of these components. Omitting any of the components would
216
RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor
result in zero plaque formation. Electron microscopy and genome restriction mapping have confirmed the identity of the infectious phi29 virions synthesized using this system. With this highly sensitive virion assembly system, it has been possible to detect 5 the activity of mutant pRNAs with 10 -fold reductions in DNA packaging efficiency [91], which could be too low for detection using other assay methods available [8, 73, 92]. The system has also been used extensively to study the structure and function of pRNA, including functional domains [21, 22, 93] and stoichiometry [80, 94].
8.4
Structure of pRNA The phi29 pRNA is vital to the phi29 motor for DNA packaging and virion assembly. The pRNA binds to the connector of procapsid and forms a hexameric ring through the intermolecular hand-in-hand loop interactions (Figure 8.2). The structure of pRNA has been probed by several approaches (see the Supplementary Materials and Solution Manual), and the information was applied in 3-D computer modeling of pRNA (Figure 8.3).
Figure 8.2 (A) Predicted secondary structure of pRNA. The connector binding domain and the DNA translocation domain are marked in bold lines. Sequences responsible for loop-loop interactions are boxed (94). (B) Structure of a pRNA hexameric ring by loop-loop interaction (94). (C) Structure of phi29 DNA packaging motor. (Adapted with permission from Cell Press.)
8.5 Mechanism of the phi29 Motor Function
217
Figure 8.3 (Color plate 12) (I) AFM images showing monomer (A), dimer (B), trimer (C) and an array of engineered pRNA (D) [53]. (II) Computer 3-D models of pRNA (A-C), connector (D), pRNA/connector complex (E) and pRNA/procapsid complex (F) (62). (Adapted with permission from the American Chemical Society and American Society for Biochemistry and Molecular Biology from the respective citation.)
8.5
Mechanism of the phi29 Motor Function 8.5.1
Symmetry Argument: Pentamer or Hexamer
The phi29 DNA packaging motor contains a six-fold (12-subunit) connector surrounded by a five-fold symmetrical capsid shell. A pRNA ring is participating in the motor motion. The symmetry of the ring is thus very important for elucidating the motor mechanism. However, whether pRNA stoichiometry is six-fold (a hexamer) or five-fold (a pentamer) is under fervent debate. The hexamer hypothesis is supported by the following. In 1998, two labs [94, 95] independently demonstrated that pRNAs formed hexamers as part of the phi29 motor. Crosslinking experiments revealed that pRNA did not bind to the capsid proteins, but to the connector [64], which had six-fold (12-subunit) symmetry. Subsequent to the biochemical, mathematical, and genetic approach, a hexameric pRNA ring was reported in 2002 using Cryo-EM [57]. Further study provided solid evidence that the pRNA interacted with the N-terminus of the connector protein gp10, specifically, the three basic amino acids at its N-terminus [11]. Mutation of two of these three amino acids resulted in completely abolishing pRNA binding to the DNA packaging motor.
218
RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor
Recent single molecule counting using the very sensitive dual view total internal reflection fluorescence microscopy revealed that each DNA packaging motor contained six copies of pRNA (Figure 8.4) [97, 98]. The pentamer hypothesis is mainly based on the Cryo-EM reconstruction of a pRNA pentamer ring on procapsid [99–101]. It was explained that pRNA hexamers could be formed initially, but after binding to procapsids, one of the pRNAs was dissociated from the procapsids leaving five pRNAs still bound [99]. We prefer the pRNA hexamers hypothesis. First, Cryo-EM image reconstruction using the averaging approach for RNA remains difficult due to the potential degradation of RNA. Second, crosslinking experiments revealed that pRNA did not bind directly to the capsid protein (five-fold symmetry), but to the connector (six-fold symmetry) [64]. The argument of one pRNA molecule leaving after procapsid binding was not supported by the finding that covalently linked pRNA dimers are active in DNA packaging [102]. In addition, single molecule fluorescence microscopy studies on purified DNA packaging intermediates revealed that the active motor during DNA translocation contained six copies of pRNA [98]. Results from the aforementioned approaches for stoichiometry studies support that only the
Figure 8.4 Stoichiometric study of pRNA by single molecule counting (98). (A) Schematic drawing of the experimental design. (B) A typical time trajectory of fluorescence intensity indicating six pRNA molecules bound to a single motor. (C) Comparison of experimental distribution in photobleaching steps with the theoretical one based on 70% lableing of pRNA. (D) A typical overlaid fluorescence image of DNA packaging intermediate. (Red: Cy5-DNA; Green: Cy3-pRNA; Yellow: intermediate.)
8.5 Mechanism of the phi29 Motor Function
219
hexameric but not the pentameric pRNA ring is active in DNA packaging [12, 94, 98, 102, 103]. 8.5.2 ATP Hydrolysis Provides the Driving Force of the phi29 DNA Packaging Motor
Most molecular motors use chemical energy derived from the hydrolysis of ATP. This is also the case for the phi29 DNA packaging motor. The energy drives conformational changes in the motor and results in motor movement. This process repeats with hydrolysis of another ATP molecule and the motor movement can continue. The packaged DNA undergoes about a 30- to 100-fold decrease in DNA volume compared with its volume before packaging [104, 105]. Thus, the packaging process is very thermodynamically unfavorable. Studies involving bacteriophages lambda [106–111], phi29 [8, 66, 112], T3/T7 [88, 113–118], T4 [89, 119, 120], and P2 [121, 122] prove that ATP hydrolysis provides the driving force for viral DNA packaging motors. The first quantification of ATP consumption in DNA packaging was performed in the phi29 system with purified components [66]. In phi29, recent results indicated that pRNA stimulated the ATPase activity of gp16 [68]. Procapsid could enhance the stimulation but it alone could not produce ATPase activity [63]. These results suggested that ATP effects in the DNA packaging system were very complicated. In phi29, all components in the packaging system, which includes pRNA, procapsid, gp16, and DNA-gp3, may be involved in the generation of maximal ATPase activity. 8.5.3
Possible Models for phi29 Motor Function
Several models for DNA translocation into the phi29 procapsid have been proposed [12, 99, 123–129]. 8.5.3.1
Model 1: Brownian Motion
In this model, it proposed that Brownian motion causes the movement of DNA [126–128]. The motor is a ratchet that rectifies oscillating forces [126] simply to ensure that DNA only moves in one direction, but not the opposite. There is no ATP-driven power stroke. With this model it is difficult to explain the final stage of the DNA packaging process: When the procapsid is almost filled up and the internal pressure has increased, Brownian motion as a force to drive the DNA will be unfavorable. Contrary to ATP-driven power strokes, the Brownian motion as a driving force in this hypothesis will not be reduced at low ATP concentration. A slower rate of DNA packaging at limiting ATP concentrations would therefore be associated with a “leaky” ratchet mechanism rather than reduced frequency of power strokes. This model is not favored. 8.5.3.2
Model 2: Sequencial Action of Motor Components
DNA is translocated through the axial hole of the portal vertex much like a threaded rod moving through a nut and thus DNA packaging could be achieved by
220
RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor
utilizing the threaded helical nature of dsDNA [12, 129]. The motor composed of the pRNA, gp16, and connector works like a hex nut to drive a bolt [12]. The model is based on biochemical data indicating that pRNA is a hexamer [62, 80, 94, 102, 130], and that it binds to the connector but not to the procapsid protein gp8 [64]. A sequential action of pRNA is required for DNA translocation [12]. The symmetry mismatch between pRNA or a pRNA-gp16 complex (six-fold) and the vertex of the procapsid in which the connector is located (five-fold) gives rise to ATP-driven 6 or 30 (5 × 6) steps of the connector-pRNA-(gp16) complex which in turn translocates DNA in a similar way to a hex nut driving a bolt. The pRNA and gp16 undergo a conformational change during the ATP hydrolysis cycle and act sequentially by pushing against the static procapsid vertex. After each circle of the sequential steps the hexameric pRNA ring, and/or the gp16, aligns with an equivalent point of the environment composed of the procapsid vertex and the connector. Although the sequential action model clearly implied the sequential shifting nature as long as the related motion of DNA and connector, as well as the action style of the components, is concerned, this model did not imply the continuous turning of the connector, pRNA, or gp16 with a 360° circle. The motion can be a sequential action with each step in a back-and-forth shifting manner, with the dislocation of any two components less than 60° for the entire DNA packaging process [12]. 8.5.3.3
Model 3: DNA Translocation Outside the Central Pore
Supercoiled DNA wraps around the portal vertex and rotation of the portal vertex allows DNA to pass into the procapsid via the outside of the portal vertex as proposed by Anderson et al. [125]. The model is unique in suggesting that DNA enters the procapsid via the outside of the connector rather than the central channel. However, this model is not favored by recent structural data on the DNA packaging motor of phi29 [24, 57, 99, 131]. In particular, Simpson et al. found DNA in the central connector channel when they analyzed partially packaged procapsids by cryo-electron microscopy [99]. Although data showing pRNA as a hexamer and direct observation of rotation would be consistent with the model, proof of the hypothesis would likely require high resolution structural analysis of packaging intermediates. 8.5.3.4
Model 4: Lysine-Phosphate Alignment Rotation
The five-fold/six-fold symmetry mismatch and sequential contraction and relaxation of the motor components generate a force for motor rotation [12, 129]. The link between connector rotation and DNA translocation is given by the alignment of lysine residues in the connector channel with phosphate residues of the DNA backbone [131]. The model suggests that a 6° rotation of the connector correlates with a DNA translocation by one base pair, so that lysine-phosphate pairs are reestablished one base-pair further on. Hydrolysis of one ATP leads to a 12°/2 bp step, consistent with a five-fold/six-fold symmetry mismatch and with experimental data showing that one molecule of ATP is required to package 2 bp of DNA [66]. This model leaves open whether the ATP-generated force acts on the connector to drive DNA translocation, or on the DNA, so that the connector rotates in a passive way.
8.5 Mechanism of the phi29 Motor Function
8.5.3.5
221
Model 5: Connector Contraction
DNA located in the central channel interacts with one subunit of the portal [99]. A 12° rotation of the narrow end of the connector leads to lengthwise expansion of the connector via a slight change in the angle of the long helices, and the wide end of the connector follows the narrow end. Such a “following” allows the structure to relax and contract while translating two base pairs of DNA into the capsid [99]. If DNA moves irregularly relative to the rotation circle, this model will be excluded. This model implies that one complete turn of the connector transfers 60 bp of DNA into the procapsid. A cornerstone of the model is that pRNA is a component of the static part of the motor with five-fold symmetry [99, 100]. The symmetry mismatch between pRNA (five-fold) and connector (six-n-fold) gives rise to discrete counterclockwise (view down the connector axis toward the procapsid) 12°-rotational steps of the connector, with one molecule of ATP consumed and two base pairs of DNA translocated per step. If this model is correct, pRNA will crosslink to capsid protein gp8 with higher affinity than to connector protein gp10. However, crosslinking between procapsid and pRNA showed that pRNA was crosslinked to connector gp10, not the capsid protein gp8 [64]. 8.5.4
Single Molecule Approaches to Elucidate Motor Mechanism
Taking advantage of the highly efficient in vitro DNA packaging system, it is possible to reveal the motor mechanism with more details. 8.5.4.1
The Use of Optical Tweezers in Studies of Motor Mechanism
Optical tweezers offer a direct way to measure packaging force and speed [56, 132, 133]. A partially packaged motor complex had a microsphere attached via unpackaged DNA. This microsphere was caught in an optical trap and tethered to a second bead. Upon the introduction of ATP and initiation of packaging, the two beads moved closer together. The amount of DNA tension can be monitored, and from this, the bead displacement can be calculated; various measurements of packaging dynamics are then possible, including an examination of the presence of packaging “slips” and “pauses,” where irregularities occur in packaging speed. With the addition of laser radiation pressure, the force needed to prevent the DNA from being inserted was found to be about 57 piconewtons, indicating that the phi29 DNA-packaging motor was the strongest biomotor studied to date. Using this method, it was possible to determine the speed of phi29 DNA packaging, which was initially around 100 bases per second, gradually slowing to a halt as the procapsid was filled [56, 132, 133]. The optical tweezers have also been applied to reveal the step size of ATP hydrolysis during phi29 DNA packaging at high resolution [134]. It was found that the phi29 motor packages DNA at increments of 10 bp, with four 2.5-bp substeps for each ATP hydrolysis. A coordinative mechanism in phi29 DNA packaging was suggested according to the results [134].
222
RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor
8.5.4.2
Advanced Imaging Technologies
The sizes of various components of the motor span from subnanometer range to the hundred-nanometer range. They have been studied by various advanced imaging technologies such as NMR, X-ray crystallography, AFM and, Cryo-EM depending on the size suitability. In recent years, fluorescence imaging has also been applied to study phi29 motors [97, 98, 136–138]. 8.5.4.2.1 Single Molecule Study of Motor Function
Recent advances in single molecule fluorescence microscopy have provided a new way to understand motor mechanisms by directly counting motor components and observing motion events [97, 98]. It allows the analysis of individual motor components, as opposed to the ensemble averaging of the measurements from a massive population of homogenous molecules that are motionless or requiring synchronous motion [139–142]. Single molecule approaches also allow the direct observation of physical behaviors to answer many questions, including: (1) how the chemical energy is converted into the physical motion [143–145], (2) how the force is generated [146], (3) how the molecular structure is involved in chemical reactions [147, 148], (4) how the motion starts and continues without interruptions, (5) how each motor component responds to the applied force [132], and (6) how the conformational change of each motor component is correlated to the generation of the force [132, 149–152]. Unveiling the clues to these questions will make it possible to elucidate the properties of bionanomotors, and will also help to design novel nanodevices as well as imitate natural organs [153–157]. 8.5.4.2.2 Direct Observation of Motor Motion
Attaching a fluorescent bead at the tip of DNA has been employed to track the motion of DNA during translocation (Figure 8.5, part IA) [98]. The attached fluorescent microsphere amplifies the signal to be directly observed by fluorescence microscopy. Its motion can be profiled in a three-dimensional space (x, y, and z axis). The DNA packaging intermediates were stalled by a nonhydrolysable γ-S-ATP, restarted by addition of ATP, and observed in real time by fluorescence microscopy. DNA migration caused the attached microsphere to show a gradual reduction in swing range. Finally, the motion stopped due to the physical restriction of the DNA being completely packaged, and appeared under the CCD camera as a zero distance change from the reference origin (Figure 8.5, part IB). Optical imaging of a magnetic bead attached to one end of DNA-gp3 was also developed for direct observation and dynamic analysis of packaging (Figure 8.5, part II) [137].
8.6 Potential Applications of the phi29 Motor in Nanotechnology and Gene Therapy 8.6.1
A Nanomotor with the Potential to Be Incorporated into Nanodevices
Nanomotors have the potentials to be utilized in nanotechnology [158–160]. One viable option that has been pursued for the development of mechanical parts for nanotechnology is to incorporate the phi29 motor and its constituent parts into nanodevices. The techniques generally involve the characterization, manipulation,
8.6 Potential Applications of the phi29 Motor in Nanotechnology and Gene Therapy
223
Figure 8.5 (Color plate 13) Direct observation of phi29 DNA translocation. (I) Observation of a fluorescent microsphere attached to phi29 DNA (98). (A) Design for the observation. (B) Displacement of the DNA-tethered fluorescent bead, in X and Y direction respectively. The measurements are in pixel. (II) Observation of a magnetic microsphere attached to phi29 DNA (137). (A) Design of the observation. (B) Tether length of DNA versus time for two observed motors. (C) The calculated DNA packaging rate from (B) against percentage of packaged DNA. (Adapted with permission from American Institute of Physics from the respective citation.)
modification, control, creation, and/or assembly of organized materials on the nanoscale level [161, 162], thereby helping to form supramolecular structures. These materials can then be used as building blocks for the construction of larger devices and systems. Nanotechnological endeavors play critical roles in many scientific disciplines, including chemistry, physics, biology, medicine, material science, engineering, computer technology, and many other fields. The phi29 DNA packaging motor is ideal for applications in nanotechnology and other fields, as it has been well studied regarding its structures and functions. It can be readily incorporated into nanodevices. The motor particles can be attached to the nanopores of nanoporous anodic aluminum oxide (AAO) membranes with controllable pore sizes (Figure 8.6, part I). The attachment is either to the aldehyde-silanized inner surface, or by centrifugation. The empty procapsids were found separated from DNA-filled procapsids by the membrane with 40-nm pores. The nanoporous AAO membranes could be applied for future interfacing of the phi29 motor with artificial nanostructures [163].
224
RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor
Figure 8.6 (I) Field-emission scanning electron microscope (FE-SEM) images of the anodic aluminum oxide (AAO) membrane with pore sizes of (A, B) 75±5 nm as anodized nanoporous membranes, (C, D) 39±4 nm, (E, F) 25± 3 nm, and (G, H) 15±1 nm (163). (II) Construction of the phi29 connector array (165). (A) Side view of a multiple layer connector array. (B) Side view of a single layer connector array, with the reengineered connector carrying an N-terminal tag. (C) Schematic illustration of single layer reengineered connector array formed with the assistance of lipid. (1) Streptavidin/N-Strep dodecamer solution is placed in a Teflon well; (2) Biotinylated lipids DPPE and helper lipids Egg PC were spread at the air-water interface to attract dodecamers to the surface of the liquid; (3) Dodecamers bound to the lipid via specific biotin-streptavidin interactions. (D) Negative-stain TEM image of the single-layer array. (E) Fourier transforms and (F) corresponding Fourier projection maps of lipid directed array of re-engineered connectors carrying an N-terminal tag. (G) AFM image of tetragonal arrays of reengineered connectors carrying an N-terminal tag. (H) and (I) Cross-sections along the axes of the two-dimensional array in (G). The unit cell is a parallelogram with cell dimensions of 16 × 13 nm. (J) AFM image of tetragonal array of a reengineered connector carrying a C-terminal tag. The unit cell is rectangular with a lattice constant of ~18 nm. (Adapted with permission from Springer Science + Business Media, LLC, and the American Chemical Society from the respective citation.)
8.6.2
Connector Arrays for Nanotechnology
Approaches for bottom-up assembly of nanopatterned materials as disease diagnostic chips, or as ultrahigh-density data storages have been developed. The in vitro construction of large-scale carpet-like phi29 connector arrays were reported (Figure 8.6 part II) [49, 96, 164, 165]. A uniform and highly ordered single layer array of connectors has been successfully constructed using a supporting lipid monolayer [165]. The addition of pRNA into the array caused a dramatic shift in array structure, and resulted in the conversion of tetragonal arrays into larger decagonal structures. RNase digestion confirmed that the conformational shift was caused by pRNA and that pRNA was present in the decagons. A conformational shift of motor components can generate force for biomotor motion. The conformational shift reported here may be utilized as a potential force-generating mechanism for the nanomachine construction. Three-dimensional computer models of the constructed arrays were also produced using a variety of connector building blocks with or without the N- or C-terminal sequence, which was absent from the current published crystal structures
8.6 Potential Applications of the phi29 Motor in Nanotechnology and Gene Therapy
225
[96]. Both the connector array and the decagon are ideal candidates to be used as templates to build patterned suprastructures in nanotechnology. 8.6.3
Polyvalent Gene Delivery System Using Phi29 pRNA
The phi29 pRNA contains two functional domains: the procapsid binding domain and the DNA translocation domain that is located at the 5´/3´ paired ends. The two domains fold independently of each other. The complementary loop sequences located at the procapsid binding domain allow for the formation of predefined pRNA multimers, modifiable at the earliest possible stage of the construction of larger structures and systems. Replacement of or insertion into the 5’/3’ helical domain does not interfere with multimer formation. Thus, end conjugation of pRNA with the chemical moiety of fusing pRNA with a receptor-binding RNA aptamer, small interfering RNA (siRNA), or ribozyme may not disturb multimer formation or interfere with the function of inserted moieties. Using the circular permutation approach [19–21], almost any nucleotide of the entire pRNA can serve as either the new 5´ or 3´ end of the RNA monomer. These unique features make pRNA a maneuverable and controllable polyvalent vehicle for nanomachine fabrication, pathogen detection, and therapeutics delivery. 8.6.3.1 Construction of pRNA Chimera Harboring siRNA or Ribozyme for Therapeutic Purposes
Complementary modification studies have revealed that altering the primary sequences of any nucleotide in the helical region does not affect the pRNA structure and folding as long as the two strands are paired [102]. Extensive studies reveal that siRNA is a double-stranded RNA helix [166–169]. Thus, it is possible to replace the helical region in pRNA with double-stranded siRNA. A variety of chimeric pRNAs with different targets were constructed to carry siRNA connected to bases 29 and 91 of the pRNA. This pRNA/siRNA chimera was proven to be the building block which successfully inhibited target gene expression [170, 171]. In addition, connecting the pRNA 5’/3’ ends with variable sequences did not disturb its folding and function. These unique features, which help prevent two common problems—exonuclease degradation and misfolding in the cell—make pRNA an ideal vector to carry therapeutic RNAs. A pRNA-based vector was designed to carry hammerhead ribozymes that cleaved the hepatitis B virus (HBV) polyA signal [172]. Another pRNA/ribozyme (survivin) chimera that targeted the antiapoptosis factor survivin was shown to suppress survivin expression and initiate apoptosis then cell death [173]. 8.6.3.2
Construction of pRNA Chimera for Cell Targeting
The application of in vitro SELEX [174, 175] to screen RNA aptamers that bind to specific targets has become a powerful tool for selecting RNA molecules specific to cell surface receptors. Aptamers are linked to the 3´ and 5´ end of the pRNA. To facilitate independent folding, poly U or poly A linker might be placed between the pRNA and the aptamer. The nascent 5´/3´ end of the pRNA will be relocated to
226
RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor
nucleotide (nt) 71 and 75 using the circular permutation approach [21, 176]. The tightly folded nt 71 and 75 region protects the 5’/3’ end from exonuclease digestion. In addition to aptamer targeting, chemical targeting is also performed. For example, to target folate receptors on the cell surface, folate was conjugated to pRNA for the specific targeting [177–181]. The use of small RNA in gene therapy was significantly hampered by the difficulties involved in producing a safe and efficient system by which specific cells could be targeted. The strength of using phi29 pRNA as a delivery vehicle relies on its ability to easily form stable dimers, trimers, and hexamers, which can be manipulated and sequence-controlled [94, 102, 182]. The polyvalent pRNA can carry up to six kinds of therapeutics and targeting reagents [183]. Thus, this particular system provides an unprecedented versatility in constructing polyvalent delivery vehicles by separately constructing individual pRNA subunits with various cargos and mixing them together in any desired combination. One or more subunits of the deliverable pRNA complex can be altered to carry an RNA aptamer or folic acid that binds to the surface receptors of cancer cells, thereby inducing receptor-mediated endocytosis. Another one or two subunits can be altered to carry components that will be utilized to enhance endosome disruption and release therapeutic molecules from the endosome. Other subunits may carry therapeutic siRNA, ribozyme RNA, and antisense RNA for gene silencing. Anticancer drugs could also be attached to a pRNA subunit to enhance the therapeutic effect or overcome the drug resistance by combination therapy [172]. The detection of therapeutics efficacy could also be combined into one nanoparticle, making simultaneous therapy and detection of the therapeutic outcome possible with only one administration. 8.6.4
Engineered phi 29 Connectors as Therapeutic Tools
One of the essential components of the phi29 DNA packaging motor is the connector complex, a dodecameric cylindrical structure with a 3.6-nm central channel compromised of 12 gp10 subunits. The N-terminal and C-terminal of the connector ring can be fused with cell membrane penetrating peptide such as TAT, polyarginine or penetratin, which will help the connector punch into the cell membrane and release the cell contents, and subsequently cause cell death. Thus, the connector can be designed to punch into the cancer cells or viral infected cells and be used for cancer or viral disease therapy. The specificity can be obtained by fusing the N-terminal or C-terminal of gp10 with single chain antibody. An alternative method is to passively translocate the specific DNA fragments into the cells for therapeutic purpose after the connector punches a hole in the cell membrane. 8.6.5
phi29 DNA Packaging Motors Act as Tools for Gene Therapy
The double-stranded DNA can be translocated by the phi29 DNA packaging motor with remarkable velocity by ATP hydrolysis. It is possible that by fusing an antibody labeled connector and pRNA to the targeted cell membrane, an intact active motor can be constructed. By using the energy generated from ATP hydrolysis, the motor may insert functional genes into the cells that could rectify the genetic deficiency of
8.7 Prospectives
227
some genetic diseases or translocate therapeutic siRNAs into cancer cells as well as viral infected cells to induce apoptosis. 8.6.6 The DNA-Packaging Motor as a DNA-Sequencing Apparatus or Molecular Sorter
Another growing area of interest in biology is the development of efficient and sensitive analytical tools that can be used to probe and manipulate single molecules. Recent projects of mapping bacterial, human, and other mammalian genomes require new types of arrays and methodologies to analyze, interpret, and utilize genomic information efficiently and inexpensively. Scientists are working to develop a nanopore-based DNA sequencing device [135]. Such a device will recognize a single base pair, based on the electrical signals generated through the interaction of the bases of the DNA with a pore. The phi29 DNA packaging motor has the potential to be developed into a DNA sequencing apparatus since the DNA packaging process involves movement of the DNA through the pRNA bound connector channel that can be modified to accept chemical or electrical signals.
8.7
Prospectives The development of a defined in vitro packaging system utilizing purified phi29 components helps to provide insights in the mechanisms of phi29 and other DNA packaging systems. The phi29 system also possesses a number of potential technological advances, among which are the targeted delivery of therapeutic agents to cells, precise single-molecule sequencing techniques, and direct incorporation of various motor parts to nanodevices. In addition, the strong tendency of phi29 pRNA to form dimers, trimers, and hexamers can potentially be utilized to construct self-assembling nanostructures. Such nanostructures may have far-reaching technological possibilities, such as development of ultrahigh-density memory storage systems, and the isolation and separation of multiple pathogens in medical diagnoses. Progresses in ongoing research into specific motor parts and overall motor function are making these possibilities increasingly realistic.
Problems 8.1 How many and what kinds of components are involved in the DNA packaging motor of bacteriophage phi29? What are the functions and stoichiometries of each component? 8.2 What are the roles of each phi29 motor component relevant to motor motion? What are their analog parts in mechanical motors or machine pumps? 8.3 How would you evaluate the models for phi29 motor function described in this chapter? Could you propose a new model for the motor, besides the ones already mentioned in this chapter?
228
RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor
References [1] Earnshaw, W.C. and Casjens, S.R. (1980) “DNA packaging by the double-stranded DNA bacteriophages.” Cell, 21, 319-331. [2] Bazinet, C. and King, J. (1985) “The DNA translocation vertex of dsDNA bacteriophages.” Ann. Rev. Microbiol., 39, 109-129. [3] Casjens, S. and Hendrix, R. (1988) “Control mechanisms in dsDNA bacteriophage assembly.” In The Bacteriophages Vol.1., Calendar, R. (ed.). Plenum Pubishing Corp., New York, pp. 15-92. [4] Guo, P. (1994) “Introduction: Principles, perspectives, and potential applications in viral assembly.” Seminars in Virology (Editor’s Introduction), 5(1), 1-3. [5] Anderson, D.L. and Reilly, B. (1993) “Morphogenesis of bacteriophage phi29.” In Bacillus Subtilis and Other Gram-Positive Bacteria: Biochemistry, Physiology, and Molecular Genetics. Sonenshein, A.L., Hoch, J.A. and Losick, R. (eds.), American Society for Microbiology, Washington, D.C., pp. 859-867. [6] Black, L.W. (1989) “DNA packaging in dsDNA bacteriophages.” Ann Rev Microbiol, 43, 267-292. [7] Guo, P. and Trottier, M. (1994) “Biological and biochemical properties of the small viral RNA (pRNA) essential for the packaging of the double-stranded DNA of phage phi29.” Seminars in Virology, 5, 27-37. [8] Guo, P., Grimes, S. and Anderson, D. (1986) “A defined system for in vitro packaging of DNA-gp3 of the Bacillus subtilis bacteriophage phi29.” Proc. Natl. Acad. Sci. USA, 83, 3505-3509. [9] Guo, P., Erickson, S. and Anderson, D. (1987) “A small viral RNA is required for in vitro packaging of bacteriophage phi29 DNA.” Science, 236, 690-694. [10] Wichitwechkarn, J., Johnson, D. and Anderson, D. (1992) “Mutant prohead RNAs in the in vitro packaging of bacteriophage phi 29 DNA-gp3.” J Mol. Biol, 223, 991-998. [11] Xiao, F., Moll, D., Guo, S. and Guo, P. (2005) “Binding of pRNA to the N-terminal 14 amino acids of connector protein of bacterial phage phi29.” Nucleic Acids Res, 33, 2640-2649. [12] Chen, C. and Guo, P. (1997) “Sequential action of six virus-encoded DNA-packaging RNAs during phage phi29 genomic DNA translocation.” J. Virol., 71, 3864-3871. [13] Shu, D. and Guo, P. (2003) “A viral RNA that binds ATP and contains an motif similar to an ATP-binding aptamer from SELEX.” J. Biol. Chem., 278(9), 7119-7125. [14] Lee, T.J. and Guo, P. (2006) “Interaction of gp16 with pRNA and DNA for genome packaging by the motor of bacterial virus phi29.” J. Mol Biol., 356, 589-599. [15] Bailey, S., Wichitwechkarn, J., Johnson, D., Reilly, B., Anderson, D. and Bodley, J.W. (1990) “Phylogenetic analysis and secondary structure of the Bacillus subtilis bacteriophage RNA required for DNA packaging.” J Biol Chem, 265, 22365-22370. [16] Pecenkova, T., Benes, V., Paces, J., Vlcek, C. and Paces, V. (1997) “Bacteriophage B103: complete DNA sequence of its genome and relationship to other Bacillus phages.” Gene, 199, 157-163. [17] Wichitwechkarn, J., Bailey, S., Bodley, J.W. and Anderson, D. (1989) “Prohead RNA of bacteriophage phi29: size, stoichiometry and biological activity.” Nucelic Acids Research, 17(9), 3459-3468. [18] Guo, P., Bailey, S., Bodley, J.W. and Anderson, D. (1987) “Characterization of the small RNA of the bacteriophage phi29 DNA packaging machine.” Nucleic Acids Res., 15, 7081-7090. [19] Pan, T., Gutell, R.R. and Uhlenbeck, O.C. (1991) “Folding of circularly permuted transfer RNAs.” Science, 254, 1361-1364. [20] Nolan, J.M., Burke, D.H. and Pace, N.R. (1993) “Circularly permuted tRNAs as specific photoaffinity probes of ribonuclease P RNA structure.” Science, 261, 762-765.
References
229
[21] Zhang, C.L., Trottier, M. and Guo, P.X. (1995) “Circularly permuted viral pRNA active and specific in the packaging of bacteriophage phi29 DNA.” Virology, 207, 442-451. [22] Zhang, C.L., Lee, C.-S. and Guo, P. (1994) “The proximate 5’ and 3’ ends of the 120-base viral RNA (pRNA) are crucial for the packaging of bacteriophage phi29 DNA.” Virology, 201, 77-85. [23] Kellenberger, E. and Wunderli-Allenspach, H. (1995) “Electron microscopic studies on intracellular phage development - History and perspectives.” Micron, 26, 213-245. [24] Tao, Y., Olson, N.H., Xu, W., Anderson, D.L., Rossmann, M.G. and Baker, T.S. (1998) “Assembly of a Tailed Bacterial Virus and its Genome Release Studied in Three Dimensions.” Cell, 95, 431-437. [25] Lee, C.S. and Guo, P. (1995) “Sequential interactions of structural proteins in phage phi29 procapsid assembly.” J. Virol., 69, 5024-5032. [26] Murialdo, H. and Becker, A. (1977) “Assembly of biologically active proheads of bacteriophage lambda in vitro.” Proc Natl Acad Sci U. S. A, 74, 906-910. [27] Shaw, J.E. and Murialdo, H. (1980) “Morphogenetic genes C and Nu3 overlap in bacteriophage lambda.” Nature, 283, 30-35. [28] Black, L.W. and Showe, M.K. (1983) “Morphogenesis of the T4 head.” In Bacteriophage T4. Mathews,C.K., Kutter,E.M., Mosig,G. and Berget, P.B. (eds.). American Society for Microbiology, Washington, D.C., pp. 219-245. [29] Guo, P., Scholz, E., Turek, J., Nordgren, R. and Maloney, B. (1993) “Assembly pathway of avian infectious laryngotracheitis virus.” Am J Vet Res, 54, 2031-2039. [30] Traub, F. and Maeder, M. (1984) “Formation of the prohead core of bacteriophage T4 in vivo.” J. Virol., 49, 892-901. [31] Traub, F., Keller, B., Kuhn, A. and Maeder, M. (1984) “Isolation of the prohead core of bacteriophage T4 after cross-linking and determination of protein composition.” J. Virol., 49, 902-908. [32] Kuhn, A., Keller, B., Maeder, M. and Traub, F. (1987) “Prohead core of bacteriophage T4 can act as an intermediate in the T4 head assembly pathway.” J. Virol., 61, 113-118. [33] Thomsen, D.R., Roof, L.L. and Homa, F.L. (1994) “Assembly of herpes simples virus (HSV) intermediate capsids in insect cells infeced with recombinant baculoviruses expressing HSV capsid proteins.” J. Virol., 68, 2442-2457. [34] Tatman, J.D., Preston, V.G., Nicholson, R.M., Elliott, R.M. and Rixon, F.J. (1994) “Assembly of herpes simplex virus type 1 capsids using a panel of recombinant baculoviruses.” J Gen Virol, 75, 1101-1113. [35] Newcomb, W., Homa, F.L., Thomsen, D.L., Ye, Z. and Brown, J. (1994) “Cell-free assembly of the herpes simplex capsid.” J. Virol., 68, 6059-6063. [36] Hendrix, R.W. and Garcea, R.L. (1994) “Capsid assembly of dsDNA viruses.” Seminars in Virology, 5, 15-26. [37] Hsiao, C.L. and Black, L.W. (1977) “DNA packaging and the pathway of bacteriophage T4 head assembly.” Proc Natl Acad Sci USA, 74, 3652-3656. [38] Kochan, J., Carrascosa, J.L. and Murialdo, H. (1984) “Bacteriophage lambda preconnectors: Purification and structure.” J Mol Biol, 174, 433-447. [39] Nakasu, S., Fujisawa, H. and Minagawa, T. (1985) “Purification of characterization of gene 8 product of bacteriophage T3.” Virology, 143, 422-434. [40] Raj, A.S., Raj, A.Y. and Schmieger, H. (1974) “Phage genes involved in the formation generalized transducing particles in Salmonella-Phage P22.” Mol. Gen. Genet., 135, 175-184. [41] Herranz, L., Salas, M. and Carrascosa, J.L. (1986) “Interaction of the bacteriophage phi29 connector protein with the viral DNA.” Virology, 155, 289-292. [42] Becker, A. and Murialdo, H. (1990) “Bacteriophage λ DNA: The begining of the end.” J Bact, 172, 2819-2824. [43] Valpuesta, J.M. and Carrascosa, J. (1994) “Structure of viral connectors and their funciton in bacteriophage assembly and DNA packaging.” Quart. Rev. Biophys., 27, 107-155.
230
RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor [44] Carazo, J.M., Donate, L.E., Herranz, L., Secilla, J.P. and Carrascosa, J.L. (1986) “Three-dimensional reconstruction of the connector of bacteriophage phi29 at 1.8 nm resolution.” J Mol Biol, 192, 853-867. [45] Donate, L.E., Valpuesta, J.M., Rocher, A., Mendez, E., Rojo, F., Salas, M. and Carrascosa, J.L. (1992) “Role of the amino-terminal domain of bacteriophage phi 29 connector in DNA binding and packaging”. J Biol Chem, 267, 10919-10924. [46] Driedonks, R.A., Engel, A., tenHeggeler, B. and van Driel, R. (1981) “Gene 20 product of bacteriophage T4: Its purification and structure.” J Mol Biol, 152, 641-662. [47] van den Worm, S.H.E., Stonehouse, N.J,.and Valegard, K. (1998) “Crystal structure of MS2 coat protein mutants in complex with wide-type RNA operator fragment.” Nucelic Acids Research, 26(5), 1345-1351. [48] Carazo, J.M., Fujisawa, H., Nakasu, S. and Carrascosa, J.L. (1986) “Bacteriophage T3 gene 8 product oligomer structure.” Journal of Ultrastructure & Molecular Structure Research, 94, 105-113. [49] Muller, D.J., Engel, A., Carrascosa, J.L. and Vélez, M. (1997) “The bacteriophage phi29 head-tail connector imaged at high resolution with the atomic force micrescope in buffer solution.” The EMBO Journal, 10, 2547-2553. [50] Jimenez, J., Santisteban, A., Carazo, J.M. and Carrascosa, J.L. (1986) “Computer graphic display method for visualizing three-dimensional biological structures.” Science, 232, 1113-1115. [51] Valpuesta, J.M., Fernandez, J.J., Carazo, J.M. and Carrascosa, J.L. (1999) “The three-dimensional structure of a DNA translocating machine at 10 A resolution.” Structure, 7, 289-296. [52] Valle, M., Kremer, L., Martinez, A., Roncal, F., Valpuesta, J.M., Albar, J.P. and Carrascosa, J.L. (1999) “Domain architecture of the bacteriophage phi29 connector protein.” J Mol. Biol, 288, 899-909. [53] Shu, D., Moll, D., Deng, Z., Mao, C. and Guo, P. (2004) “Bottom-up assembly of RNA arrays and superstructures as potential parts in nanotechnology.” Nano Lett., 4, 1717-1724. [54] Guasch, A., Pous, J., Parraga, A., Valpuesta, J.M., Carrascosa, J.L. and Coll, M. (1998) “Crystallographic analysis reveals the 12-fold symmetry of the bacteriophage phi29 connector particle.” J Mol. Biol, 281, 219-225. [55] Chernicky, C.L., Yi, L., Tan, H., Gan, S.U. and Ilan, J. (2000) “Treatment of human breast cancer cells with antisense RNA to the type I insulin-like growth factor receptor inhibits cell growth, suppresses tumorigenesis, alters the metastatic potential, and prolongs survival in vivo.” Cancer Gene Ther., 7, 384-395. [56] Smith, D.E., Tans, S.J., Smith, S.B., Grimes, S., Anderson, D.L. and Bustamante, C. (2001) “The bacteriophage phi29 portal motor can package DNA against a large internal force.” Nature, 413, 748-752. [57] Ibarra, B, Caston, J.R., Llorca, O., Valle, M., Valpuesta, J.M. and Carrascosa, J.L. (2000) “Topology of the components of the DNA packaging machinery in the phage phi29 prohead.” J. Mol. Biol., 298, 807-815. [58] Hagen, E.W., Reilly, B.E., Tosi, M.E. and Anderson, D.L. (1976) “Analysis of gene function of bacteriophage phi29 of Bacillus subtilis: identification of cistrons essential for viral assembly.” J. Virol., 19(2), 501-517. [59] Camacho, A. and Salas, M. (1977) “Assembly of Bacillus Subtillis phage phi29, mutants in the cistrons coding for the structural proteins.” Eur. J Biochem., 73, 39-55. [60] Carazo, J., Santisteban, A. and Carrascosa, J. (1985) “Three-dimensional reconstruction of bacteriophage ?29 neck particles at 2.2 nm resolution.” J Mol Biol, 183, 79-88. [61] Carrascosa, J., Carazo, J., Herranz, L., Donate, L.E. and Secilla, J.P. (1999) “Study of two related configurations of the neck of bacteriophage phi29.” Comput. Math. Applic., 20, 57-69.
References
231
[62] Hoeprich, S. and Guo, P. (2002) “Computer modeling of three-dimensional structure of DNA-packaging RNA(pRNA) monomer, dimer, and hexamer of phi29 DNA packaging motor.” J Biol. Chem., 277(23), 20794-20803. [63] Grimes, S. and Anderson, D. (1990) “RNA dependence of the bateriophage phi29 DNA packaging ATPase.” J. Mol. Biol., 215, 559-566. [64] Garver, K. and Guo, P. (1997) “Boundary of pRNA functional domains and minimum pRNA sequence requirement for specific connector binding and DNA packaging of phage phi29.” RNA., 3, 1068-1079. [65] Guo, P.X. and Lee, T.J. (2007) “Viral nanomotors for packaging of dsDNA and dsRNA.” Mol. Microbiol., 64, 886-903. [66] Guo, P., Peterson, C. and Anderson, D. (1987) “Prohead and DNA-gp3-dependent ATPase activity of the DNA packaging protein gp16 of bacteriophage phi29.” J Mol Biol, 197, 229-236. [67] Huang, L.P. and Guo, P. (2003) “Use of acetone to attain highly active and soluble DNA packaging protein gp16 of phi29 for ATPase assay.” Virology, 312(2), 449-457. [68] Lee, T.J., Zhang, H., Liang, D. and Guo, P. (2008) “Strand and nucleotide-dependent ATPase activity of gp16 of bacterial virus phi29 DNA packaging motor.” Virology, 380, 69-74. [69] Koti, J.S., Morais, M.C., Rajagopal, R., Owen, B.A., McMurray, C.T. and Anderson, D. (2008) “DNA packaging motor assembly intermediate of bacteriophage phi29.” J Mol. Biol., 381, 1114-1132. [70] Ibarra, B., Valpuesta, J.M. and Carrascosa, J.L. (2001) “Purification and functional characterization of p16, the ATPase of the bacteriophage phi29 packaging machinary.” Nucleic Acids Res., 29(21), 4264-4273. [71] Huang, L.P. and Guo, P. (2003) “Use of PEG to acquire highly soluble DNA-packaging enzyme gp16 of bacterial virus phi29 for stoichiometry quantification.” J Virol Methods, 109, 235-244. [72] Salas, M. (1991) “Protein-priming of DNA replication.” Ann. Rev. Biochem., 60, 39-71. [73] Grimes, S. and Anderson, D. (1989) “In Vitro Packaging of Bacteriophage phi29 DNA Restriction Fragments and the Role of the Terminal Protein gp3.” J. Mol. Biol., 209, 91-100. [74] Blanco, L. and Salas, M. (1985) “Replication of phage phi 29 DNA with purified terminal protein and DNA polymerase: synthesis of full-length phi 29 DNA.” Proc. Natl. Acad. Sci. U. S. A, 82, 6404-6408. [75] Blanco, L., Bernad, A., Lázaro, J.M., Martin, G., Garmendia, C. and Salas, M. (1989) “Highly efficient DNA synthesis by the phage phi29 DNA polymerase symmetrical mode of DNA replication.” J. Biol. Chem., 264, 8935-8940. [76] Gutierrez, C., Martin, G., Sogo, J.M. and Salas, M. (1991) “Mechanism of Stimulation of DNA Replication by Bacteriophage phi29 Single-stranded DNA-binding Protein p5.” J Biol Chem, 266(4), 2104-2111. [77] Salas, M., Mendez, J., Esteban, J.A., Bernad, A., Soengas, M.S., Lazaro, J.M., Blasco, M.A. and Blanco, L. (1992) “Protein-primed replication of bacteriophage phi29 DNA.” DNA Replication and Cell Cycle, 27-34. [78] Lee, C.S. and Guo, P. (1995) “In vitro assembly of infectious virions of ds-DNA phage phi29 from cloned gene products and synthetic nucleic acids.” J. Virol., 69, 5018-5023. [79] Strobel, E., Behnish, W. and Schmieger, H. (1984) “In vitro packaging of mature phage DNA by Salmonella phage P22.” Virology, 133, 158-165. [80] Trottier, M. and Guo, P. (1997) “Approaches to determine stoichiometry of viral assembly components.” J. Virol., 71, 487-494. [81] Anderson, D.L., Hickman, H.H. and Reilly, B.E. (1966) “Structure of Bacillus subtilis bacteriophage phi29 and the length of phi29 deoxyribonucleic acid.” J Bact, 91, 2081-2089.
232
RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor [82] Reilly, B.E., Nelson, R.A. and Anderson, D.L. (1977) “Morphogenesis of bacteriophage phi29 of Bacillus subtilis: mapping and functional analysis of the head fiber gene.” J. Virol., 24, 363-377. [83] Tosi, M. and Anderson, D.L. (1973) “Antigenic properties of bacteriophage ?29 structural proteins.” J. Virol., 12, 1548-1559. [84] Carrascosa, J.L., Mendez, E., Corral, J., Rubio, V., Ramirez, G., Salas, M. and Vinuela, E. (1981) “Structural organization of Bacilus subtilis phage phi29. A model. ” Virology, 111, 401-413. [85] Nelson, R.A., Reilly, B.E. and Anderson, D.L. (1976) “Morphogenesis of bacteriophage phi29 of Bacillus subtilis: preliminary isolation and characterization of intermediate particles of the assembly pathway.” J. Virol., 19, 518-532. [86] Tosi, M.E., Reilly, B.E. and Anderson, D.L. (1975) “Morphogenesis of bacteriophage phi29 of Bacillus subtilis: cleavage and assembly of the neck appendage protein.” J. Virol., 16, 1282-1295. [87] Cohen, D.N., Erickson, S.E., Xiang, Y., Rossmann, M.G. and Anderson, D.L. (2008) “Multifunctional roles of a bacteriophage phi 29 morphogenetic factor in assembly and infection.” J Mol Biol, 378, 804-817. [88] Hamada, K., Fujisawa, H. and Minagawa, T. (1986) “A defined in vitro system for packaging of bacteriophage T3 DNA.” Virology, 151, 119-123. [89] Rao, V.B. and Black, L.W. (1988) “Cloning, overexpression and purification of the terminase proteins gp16 and gp17 of bacteriophage T4: construction of a defined in vitro DNA packaging system using purified terminase proteins.” J Mol Biol, 200, 475-488. [90] Hwang, Y. and Feiss, M. (1997) “A defined system for in vitro lambda DNA packaging.” Virology, 211, 367-376. [91] Lee, C.S. and Guo, P. (1994) “A highly sensitive system for the in vitro assembly of bacteriophage phi29 of Bacillus subtilis.” Virology, 202, 1039-1042. [92] Bjornsti, M.A., Reilly, B.E. and Anderson, D.L. (1981) “In vitro assembly of the Bacillus subtilis bacteriophage phi29.” Proc. Natl. Acad. Sci. USA, 78, 5861-5865. [93] Zhang, C.L., Garver, K. and Guo, P. (1995) “Inhibition of phage phi29 assembly by antisense oligonucleotides targeting viral pRNA essential for DNA packaging.” Virology, 211, 568-576. [94] Guo, P., Zhang, C., Chen, C., Trottier, M. and Garver, K. (1998) “Inter-RNA interaction of phage phi29 pRNA to form a hexameric complex for viral DNA transportation.” Mol. Cell., 2, 149-155. [95] Zhang, F., Lemieux, S., Wu, X., St.-Arnaud, S., McMurray, C.T., Major, F. and Anderson, D. (1998) “Function of hexameric RNA in packaging of bacteriophage phi29 DNA in vitro.” Mol. Cell., 2, 141-147. [96] Guo, Y., Blocker, F., Xiao, F., and Guo, P. (2005) “Construction and 3-D computer modeling of connector arrays with tetragonal to decagonal transition induced by pRNA of phi29 DNA-packaging motor.” J. Nanosci. Nanotechnol., 5, 856-863. [97] Xiao, F., Zhang, H., and Guo, P. (2008) “Novel mechanism of hexamer ring assembly in protein/RNA interactions revealed by single molecule imaging.” Nucleic Acids Res, 36 (20), 6620-6632. [98] Shu, D., Zhang, H., Jin, J. and Guo, P. (2007) “Counting of six pRNAs of phi29 DNA-packaging motor with customized single molecule dual-view system.” EMBO J., 26, 527-537. [99] Simpson, A.A., Tao, Y., Leiman, P.G., Badasso, M.O., He, Y., Jardine, P.J., Olson, N.H., Morais, M.C., Grimes, S., Anderson, D.L., et al. (2000) “Structure of the bacteriophage phi29 DNA packaging motor.” Nature, 408, 745-750. [100] Morais, M.C., Tao Y, Olsen, N.H., Grimes, S., Jardine, P.J., Anderson, D., Baker, T.S. and Rossmann, M.G. (2001) “Cryoelectron-Microscopy Image Reconstruction of Symmetry Mismatches in Bacteriophage phi29.” J Struct Biol, 135, 38-46.
References
233
[101] Morais, M.C., Choi, K.H., Koti, J.S., Chipman, P.R., Anderson, D.L. and Rossmann, M.G. (2005) “Conservation of the capsid structure in tailed dsDNA bacteriophages: the pseudoatomic structure of phi29.” Mol Cell, 18, 149-159. [102] Chen, C., Sheng, S., Shao, Z. and Guo, P. (2000) “A dimer as a building block in assembling RNA: A hexamer that gears bacterial virus phi29 DNA-translocating machinery.” J Biol Chem, 275(23), 17510-17516. [103] Chen, C., Trottier, M. and Guo, P. (1997) “New approaches to stoichiometry determination and mechanism investigation on RNA involved in intermediate reactions.” Nucleic Acids Symposium Series, 36, 190-193. [104] Hohn, T. (1976) “Packaging of genomes in bacteriophages: a comparison of ssRNA bacteriophages and dsDNA bacteriophages.” Philos Trans R Soc London Ser B, 276, 143-150. [105] Hohn, T., Wurtz, M. and Hohn, B. (1976) “Capsid transformation during packaging of bacteriophage λ DNA.” Philos Trans R Soc London Ser B, 276, 51-61. [106] Kaiser, A.D., Syvanen, M. and Masuda, T. (1975) “DNA packaging steps in bacteriophage lambda head assembly.” J Mol Biol, 91, 175-186. [107] Higgins, R.R., Lucko, H.J. and Becker, A. (1988) “Mechanism of cos DNA cleavage by bacteriophage lambda terminase-multiple roles of ATP.” Cell, 54, 765-775. [108] Parris, W., Davidson, A., Keeler, C.L., Jr. and Gold, M. (1988) “The Nu1 subunit of bacteriophage λ terminase.” J Biol Chem, 263, 8413-8419. [109] Davidson, A. and Gold, M. (1987) “A novel in vitro DNA packaging system demonstrating a direct role for the bacteriophage λ FI gene product.” Virology, 161, 305-314. [110] Gold, M. and Becker, A. (1983) “The bacteriophage λ terminase: partial purification and preliminary characterizaton of properties.” J Biol Chem, 258, 14619-14625. [111] Becker, A., Murialdo, H., Lucko, H. and Morell, J. (1988) “Bacteriophage lambda DNA packaging. The product of the F1 gene promotes the incorporation of the prohead to the DNA-terminase complex.” J Mol Biol, 199, 597-607. [112] Bjornsti, M.A., Reilly, B.E. and Anderson, D.L. (1983) “Morphogenesis of bacteriophage phi29 of Bacillus subtilis: oriented and quantized in vitro packaging of DNA protein gp3.” J. Virol., 45, 383-396. [113] Hamada, K., Fujisawa, H. and Minagawa, T. (1987) “Characterization of ATPase activity of a defined in vitro system for packaging of bacteriophage T3 DNA.” Virology, 159, 244-249. [114] Serwer, P. (1988) “The source of energy for bacteriophage DNA packaging: an osmotic pump explains the data.” Biopolymers, 27, 165-169. [115] Shibata, H., Fujisawa, H. and Minagawa, T. (1987) “Early events in a defined in vitro system for packaging of bacteriophage T3 DNA.” Virology, 159, 250-258. [116] Shibata, H., Fujisawa, H. and Minagawa, T. (1987) “Charactization of the bacteriophage T3 DNA packaging reaction in vitro in a defined system.” J Mol Biol, 196, 845-851. [117] White, J.H. and Richardson, C.C. (1987) “Gene 18 protein of bacteriophage T7. Overproduction, purification, and characterization.” J Biol Chem, 262, 8845-8850. [118] Morita, M., Tasaka, M. and Fujisawa, H. (1993) “DNA packaging ATPase of bacteriophage T3.” Virology, 193, 748-752. [119] Franklin, J. (1992) “A DNA packaging protein of phage T4, the product of gene 17, binds cooperatively to single-stranded DNA.” Ph.D. Dissertation. Vanderbilt University, Nashville, TN. [120] Manne, V.S., Rao, V.B. and Black, L.W. (1982) “A bacteriophage T4 DNA packaging related DNA-dependent ATPase-endonuclease.” J Biol Chem, 257, 13223-13232. [121] Bowden, D.W. and Calendar, R. (1979) “Maturation of bacteriophage P2 DNA in vitro: a complex, site-specific system for DNA cleavage.” J Mol Biol, 129, 1-18. [122] Bowden, D.W. and Modrich, P. (1985) “In vitro maturation of circular bacteriophage P2 DNA: purification of ter components and characterization of the reaction.” J Biol Chem, 260, 6999-7007.
234
RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor [123] Hendrix, R.W. (1998) “Bacteriophage DNA packaging: RNA gears in a DNA transport machine” (Minireview). Cell, 94, 147-150. [124] Dubé, P., Tavares, P., Lurz, R. and van Heel, M. (1993) “The portal protein of bacteriophage SPP1: a DNA pump with 13-fold symmetry.” EMBO J., 12, 1303-1309. [125] Turnquist, S., Simon, M., Egelman, E. and Anderson, D. (1992) “Supercoiled DNA wraps around the bacteriophage phi29 head-tail connector.” Proc. Natl. Acad. Sci. USA, 89, 10479-10483. [126] Serwer, P. (2003) “Models of bacteriophage DNA packaging motors.” J Struct. Biol, 141, 179-188. [127] Astumian, R.D. (1997) “Thermodynamics and kinetics of a Brownian motor.” Science, 276, 917-922. [128] Leibler, S. (1994) “Brownian motion. Moving forward noisily.” Nature, 370, 412-413. [129] Hendrix, R.W. (1978) “Symmetry mismatch and DNA packaging in large bacteriophages.” Proc. Natl. Acad. Sci. USA, 75, 4779-4783. [130] Trottier, M., Zhang, C.L. and Guo, P. (1996) “Complete inhibition of virion assembly in vivo with mutant pRNA essential for phage phi29 DNA packaging.” J. Virol., 70, 55-61. [131] Guasch, A., Pous, J., Ibarra, B., Gomis-Ruth, F.X., Valpuesta, J.M., Sousa, N., Carrascosa, J.L. and Coll, M. (2002) “Detailed architecture of a DNA translocating machine: the highresolution structure of the bacteriophage phi29 connector particle.” J. Mol. Biol., 315, 663-676. [132] Chemla, Y.R., Aathavan, K., Michaelis, J., Grimes, S., Jardine, P.J., Anderson, D.L. and Bustamante, C. (2005) “Mechanism of force generation of a viral DNA packaging motor.” Cell., 122, 683-692. [133] Fuller, D.N., Rickgauer, J.P., Jardine, P.J., Grimes, S., Anderson, D.L. and Smith, D.E. (2007) “Ionic effects on viral DNA packaging and portal motor function in bacteriophage phi 29.” Proc. Natl. Acad. Sci. U. S. A., 104, 11245-11250. [134] Moffitt, J.R., Chemla, Y.R., Aathavan, K., Grimes, S., Jardine, P.J., Anderson, D.L. and Bustamante, C. (2009) “Intersubunit coordination in a homomeric ring ATPase 1.” Nature, 457, 446-4U2. [135] Deamer, D.W. and Akeson, M. (2000) “Nanopores and nucleic acids: prospects for ultrarapid sequencing.” Trends Biotechnol., 18, 147-151. [136] Zhang, H., Shu, D., Huang, F. and Guo, P. (2007) “Instrumentation and metrology for single RNA counting in biological complexes or nanoparticles by a single molecule dual-view system.” RNA, 13, 1793-1802. [137] Chang CL, Zhang H, Shu D, Guo P and Savran CA (2008) “Bright-field analysis of phi29 DNA packaging motor using a magnetomechanical system.” Appl. Phys. Lett., 93, 153902-153902-3. [138] Hugel, T., Michaelis, J., Hetherington, C.L., Jardine, P.J., Grimes, S., Walter, J.M., Faik, W., Anderson, D.L. and Bustamante, C. (2007) “Experimental test of connector rotation during DNA packaging into bacteriophage phi 29 capsids.” Plos Biology, 5, 558-567. [139] Rueda, D., Bokinsky, G., Rhodes, M.M., Rust, M.J., Zhuang, X. and Walter, N.G. (2004) “Single-molecule enzymology of RNA: essential functional groups impact catalysis from a distance.” Proc. Natl. Acad. Sci. U. S. A., 101, 10066-10071. [140] Ha,T. (2001) “Single-molecule fluorescence resonance energy transfer.” Methods, 25, 78-86. [141] Ishijima, A., Kojima, H., Funatsu, T., Tokunaga, M., Higuchi, H., Tanaka, H. and Yanagida, T. (1998) Simultaneous observation of individual ATPase and mechanical events by a single myosin molecule during interaction with actin. Cell, 92, 161-171. [142] Vale, R.D., Funatsu, T., Pierce, D.W., Romberg, L., Harada, Y. and Yanagida, T. (1996) “Direct observation of single kinesin molecules moving along microtubules.” Nature, 380, 451-453.
References
235
[143] Hirono-Hara, Y., Noji, H., Nishiura, M., Muneyuki, E., Hara, K.Y., Yasuda, R., Kinosita, K., Jr. and Yoshida, M. (2001) “Pause and rotation of F(1)-ATPase during catalysis.” Proc. Natl. Acad. Sci. U. S. A, 98, 13649-13654. [144] Yasuda, R., Noji, H., Yoshida, M., Kinosita, K., Jr. and Itoh, H. (2001) “Resolution of distinct rotational substeps by submillisecond kinetic analysis of F1-ATPase.” Nature, 410, 898-904. [145] Adachi, K., Yasuda, R., Noji, H., Itoh, H., Harada, Y., Yoshida, M. and Kinosita, K., Jr. (2000) “Stepping rotation of F1-ATPase visualized through angle-resolved singlefluorophore imaging.” Proc. Natl. Acad. Sci. U. S. A, 97, 7243-7247. [146] Svoboda, K. and Block, S.M. (1994) “Force and velocity measured for single kinesin molecules.” Cell, 77, 773-784. [147] Zhuang, X., Bartley, L.E., Babcock, H.P., Russell, R., Ha, T., Herschlag,D. and Chu, S. (2000) “A single-molecule study of RNA catalysis and folding.” Science, 288, 2048-2051. [148] Ha, T., Rasnik, I., Cheng, W., Babcock, H.P., Gauss, G.H., Lohman, T.M. and Chu, S. (2002) “Initiation and re-initiation of DNA unwinding by the Escherichia coli Rep helicase.” Nature, 419, 638-641. [149] Jankowsky, E. (2005) “Biophysics: helicase snaps back.” Nature., 437, 1245. [150] Seidel, R., Bloom, J.G., van, N.J., Dutta, C.F., Dekker, N.H., Firman, K., Szczelkun, M.D. and Dekker, C. (2005) “Dynamics of initiation, termination and reinitiation of DNA translocation by the motor protein EcoR124I.” EMBO J., 24, 4188-4197. [151] Myong, S., Rasnik, I., Joo, C., Lohman, T.M. and Ha, T. (2005) “Repetitive shuttling of a motor protein on DNA.” Nature., 437, 1321-1325. [152] Rondelez, Y., Tresset, G., Nakashima, T., Kato-Yamada, Y., Fujita, H., Takeuchi, S. and Noji, H. (2005) “Highly coupled ATP synthesis by F1-ATPase single molecules.” Nature., 433, 773-777. [153] Pomerantz, R.T., Ramjit, R., Gueroui, Z., Place, C., Anikin, M., Leuba, S., Zlatanova, J. and McAllister, W.T. (2005) “A tightly regulated molecular motor based upon T7 RNA polymerase.” Nano Lett., 5, 1698-1703. [154] Fletcher, S.P., Dumur, F., Pollard, M.M. and Feringa, B.L. (2005) “A reversible, unidirectional molecular rotary motor driven by chemical energy.” Science., 310, 80-82. [155] Bringmann, H., Skiniotis, G., Spilker, A., Kandels-Lewis, S., Vernos, I. and Surrey, T. (2004) “A kinesin-like motor inhibits microtubule dynamic instability.” Science., 303, 1519-1522. [156] Tomishige, M., Klopfenstein, D.R. and Vale, R.D. (2002) “Conversion of Unc104/KIF1A kinesin into a processive motor after dimerization.” Science., 297, 2263-2267. [157] Soong, R.K., Bachand, G.D., Neves, H.P., Olkhovets, A.G., Craighead, H.G. and Montemagno, C.D. (2000) “Powering an inorganic nanodevice with a biomolecular motor.” Science, 290, 1555-1558. [158] Grigoriev, D.N., Moll, W., Hall, J. and and Guo, P. (2003) “Bionanomotor.” Encyclopedia of Nanoscience and Nanotechnology, 1, 361-374. [159] Baneyx, G. and Vogel, V. (1999) “Self-assembly of fibronectin into fibrillar networks underneath dipalmitoyl phosphatidylcholine monolayers: role of lipid matrix and tensile forces.” Proc. Natl. Acad. Sci. U. S. A, 96, 12518-12523. [160] Hess, H. and Vogel, V. (2001) “Molecular shuttles based on motor proteins: Active transport in synthetic environments.” J Biotechnol., 82, 67-85. [161] Niemeyer, C.M. (2002) “The developments of semisynthetic DNA-protein conjugates.” Trends Biotechnol., 20, 395-401. [162] Schmidt, O.G. and Eberl, K. (2001) “Nanotechnology. Thin solid films roll up into nanotubes.” Nature, 410, 168. [163] Moon, J.M., Akin, D., Xuan, Y., Ye, P.D., Guo, P. and Bashir, R. (2009) “Capture and alignment of phi29 viral particles in sub-40 nanometer porous alumina membranes.” Biomed. Microdevices., 11, 135-142.
236
RNA Ring-Geared Bacteriophage phi29 DNA Packaging Nanomotor [164] Carazo, J.M., Garcia, N., Santisteban, A. and Carrascosa, J.L. (1984) “Structural study of tetragonal-ordered aggregates of phage phi 29 necks.” Journal of Ultrastructure Research, 89, 79-88. [165] Xiao, F., Sun, J., Coban, O., Schoen, P., Wang, J.C., Cheng, R.H. and Guo, P. (2009) “Fabrication of massive sheets of single layer patterned arrays using reengineered phi29 motor dodecamer.” ACS Nano, 3, 100-107. [166] Li, H., Li, W.X. and Ding, S.W. (2002) “Induction and suppression of RNA silencing by an animal virus.” Science, 296, 1319-1321. [167] Brummelkamp, T.R., Bernards, R. and Agami, R. (2002) “A system for stable expression of short interfering RNAs in mammalian cells.” Science, 296, 550-553. [168] Carmichael, G.G. (2002) “Medicine: Silencing viruses with RNA.” Nature, 418, 379-380. [169] Elbashir, S.M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K. and Tuschl, T. (2001) “Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells.” Nature, 411, 494-498. [170] Khaled, A., Guo, S., Li, F. and Guo, P. (2005) “Controllable self-assembly of nanoparticles for specific delivery of multiple therapeutic molecules to cancer cells using RNA nanotechnology.” Nano Letters, 5, 1797-1808. [171] Guo, S., Tschammer, N., Mohammed, S. and Guo, P. (2005) “Specific delivery of therapeutic RNAs to cancer cells via the dimerization mechanism of phi29 motor pRNA.” Hum Gene Ther., 16, 1097-1109. [172] Hoeprich, S., ZHou, Q., Guo, S., Qi, G., Wang, Y. and Guo, P. (2003) “Bacterial virus phi29 pRNA as a hammerhead ribozyme escort to destroy hepatitis B virus.” Gene Ther., 10, 1258-1267. [173] Liu, H., Guo, S., Roll, R., Li, J., Diao, Z., Shao, N., Riley, M.R., Cole, A.M., Robinson, J.P., Snead, N.M. et al. (2007) “phi29 pRNA vector for efficient escort of hammerhead ribozyme targeting survivin in multiple cancer cells.” Cancer Biol. Ther., 6, 697-704. [174] Ellington, A.D. and Szostak, J.W. (1990) “In vitro selection of RNA molecules that bind specific ligands.” Nature, 346, 818-822. [175] Tuerk, G. and Gold, L. (1990) “Systematic evolution of ligands by exponential enrichment: RNA ligands to bacteriophage T4 DNA ploymerase.” Science, 249, 505-510. [176] Zhang, C.L., Tellinghuisen, T. and Guo, P. (1997) “Use of circular permutation to assess six bulges and four loops of DNA-Packaging pRNA of bacteriophage phi29.” RNA, 3, 315-322. [177] Lu, Y. and Low, P.S. (2003) “Immunotherapy of folate receptor-expressing tumors: review of recent advances and future prospects.” J. Control Release, 91, 17-29. [178] Sudimack, J. and Lee, R.J. (2000) “Targeted drug delivery via the folate receptor.” Adv. Drug Deliv. Rev., 41, 147-162. [179] Lee, R.J. and Low, P.S. (1994) “Delivery of liposomes into cultured KB cells via folate receptor-mediated endocytosis.” J Biol Chem, 269, 3198-3204. [180] Mathias, C.J., Wang, S., Lee, R.J., Waters, D.J., Low, P.S. and Green, M.A. (1996) “Tumor-selective radiopharmaceutical targeting via receptor-mediated endocytosis of gallium-67-deferoxamine-folate.” J Nucl. Med, 37, 1003-1008. [181] Guo, S., Huang, F. and Guo, P. (2006) “Construction of folate-conjugated pRNA of bacteriophage phi29 DNA packaging motor for delivery of chimeric siRNA to nasopharyngeal carcinoma cells.” Gene Ther, 13, 814-820. [182] Shu, D., Huang, L., Hoeprich, S. and Guo, P. (2003) “Construction of phi29 DNA-packaging RNA (pRNA) monomers, dimers and trimers with variable sizes and shapes as potential parts for nano-devices.” J. Nanosci. Nanotechnol., 3, 295-302. [183] Guo, P. (2005) “RNA nanotechnology: Engineering, assembly and applications in detection, gene delivery and therapy.” Journal of Nanoscience and Nanotechnology, 5(12), 1964-1982.
PART II
Electronic Biomedical Devices
CHAPTER 9
Magnetic Nanomaterials, Nanotubes, and Nanomedicine Vijay K. Varadan, Linfeng Chen, and Jining Xie
9.1
Introduction Progress in nanotechnology [1–41] has led to the development of a new subfield, nanomedicine, which is generally defined as the biomedical applications of nanomaterials [37, 38, 42]. Nanomedicine stands at the boundaries between physical, chemical, biological, and medical sciences, and the advances in nanomedicine [43–100] have made it possible to analyze and treat biological systems at the cell and subcell levels, providing revolutionary approaches for the diagnosis, prevention, and treatment of some fatal diseases [36, 59, 101]. For example, the U.S. National Cancer Institute expects that nanotechnology will be harnessed for the purposes of eliminating death and suffering from cancer. Many nanomedicine approaches are already quite close to fruition [102–128], and the U.S. Food and Drug Administration has started to consider the complex issues related to the approval of nanomaterials, nanodevices and nanosystems, for human betterment. Magnetic nanomaterials are among the most promising nanomaterials for clinical diagnostic and therapeutic applications [115]. Due to the great market potential of magnetic nanomaterials for biomedical applications, many research institutions, commercial companies, hospitals, and government organizations have spent a great deal of resources in the research of magnetic nanomaterials for biomedical applications, and amazing progress has been made in this field. Some biomedical applications of magnetic nanoparticles and biosensors are under clinical trials, and encouraging results have been reported [58]. 9.1.1
Nanotechnology and Nanomedicine
Nanotechnology involves the creation, manipulation, and application of materials, devices, and systems at the nanometer scale, and is expected to revolutionize almost every discipline of science and engineering. As most of the human cells, virus, antibodies, and proteins have dimensions in the nanometer range, many efforts have been made to utilize nanotechnology for healthcare. The marriage of nanotechnology and medicine has yielded an offspring, nanomedicine, which is set to bring essential advances in fighting against a range of diseases. It should be indicated that, as the definitions of nanotechnology and nanomedicine may change with the rapid progress in these two areas, there exist
239
240
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
strong arguments about the definitions of nanotechnology and nanomedicine. Here we try to introduce the definitions reflecting the latest achievements in these areas. 9.1.1.1
What Is Nanotechnology?
The National Nanotechnology Initiative (NNI) defines nanotechnology as the research and development at the atomic, molecular, or macromolecular levels in the sub-100-nm range to create structures, devices, and systems that have novel functional properties. At this scale, atoms can be manipulated to create stronger, lighter, and more efficient materials (nanomaterials) with tailored properties. In addition to the numerous advantages provided by this scale of miniaturization, quantum effects provide additional novel properties of nanomaterials. However, some experts consider the NNI definition of nanotechnology too rigid, emphasizing instead the continuum of scale from nanometer to micrometer [8]. According to them, the NNI definition excludes many devices and materials with micrometer dimensions, a scale that is included within the definition of nanotechnology by many nanoscientists, and they further point out that nanotechnology is not a new technology. For example, nanoscale carbon particles have been used as a reinforcing additive in tires for more than a century. Moreover, some experts indicate that most of molecular medicine and biotechnology may be considered nanotechnology [83]. For example, peptides are similar in size to quantum-dots (about 10 nm), and some viruses have the same size as drug-delivery nanoparticles (about 100 nm). Bawa et al. (2005) proposed a more practical definition of nanotechnology: “The design, characterization, production, and application of structures, devices, and systems by controlled manipulation of size and shape at the nanometer scale (atomic, molecular, and macromolecular scale) that produces structures, devices, and systems with at least one novel/superior characteristic or property.” Research in nanotechnology began with discoveries of novel physical and chemical properties that only appear in structures at the nanometer scale. Also, it is at this scale that biological molecules and structures inside living cells operate. Understanding the properties at the nanometer scale enables engineers to use materials in new ways and build new structures. The same holds true for the biological structures inside living cells of the body. Researchers have developed powerful tools to extensively investigate the parts of cells in vivid detail, and to find out how these intracellular structures operate. This will lead to better diagnostic tools and engineered nanostructures for more specific treatments of diseases [83]. 9.1.1.2
What Is Nanomedicine?
Nanomedicine can be taken as a subfield of nanotechnology. It is defined as application of nanomaterials to achieve breakthroughs in healthcare, by exploiting the novel physical, chemical, and biological properties of materials at the nanometer scale. It refers to highly specific medical interventions at the molecular scale for curing diseases or repairing damaged tissues, such as bone, muscle and nerve, and the medical advances that may be possible through nanomedicine range from diagnostic to therapeutic, and everything in between. Due to its powerful analytical capacity, nanomedicine allows an earlier and more personalized treatment for many diseases,
9.1 Introduction
241
exploiting the in-depth understanding of diseases at a molecular level. Nanomedicine holds the promise to greatly improve the efficacy of pharmaceutical therapy, reduce side effects, and make drug administration more convenient and efficient. Therefore, by enabling earlier diagnosis, better therapy and improved followup care, nanomedicine has the potential to make the healthcare process more effective in terms of clinical outcome for patients, and more affordable for society as discussed by the European Technology Platform; National Institute of Health, and the National Cancer Institute. 9.1.2
Magnetic Nanomaterials
Magnetic nanomaterials are quite different from other nanomaterials, because the fundamental properties of magnetic materials are defined at nanometer length scales. The magnetic nanomaterials often used in nanomedicine generally fall into four categories: nanospheres, nanowires, nanotubes, and thin films. It should be indicated that, in some literature, nanospheres, nanowires, and nanotubes are all called nanoparticles, while in some literature nanoparticles are mainly nanospheres. Magnetic nanospheres are the most widely used magnetic nanomaterials in nanomedicine. To realize their biomedical applications, magnetic nanospheres should be stably suspended in carrier liquids, and they should also carry out certain biomedical functions. The magnetic nanomaterials most often used are iron oxides, mainly in the form of magnetite (Fe3O4) or maghemite (γ-Fe2O3), and usually the carrier liquids are water. Nanowires are straight solid one-dimensional high aspect-ratio nanomaterials. In most cases they are cylindrical in shape with a radius in the range from 5 to 500 nm, and length up to about 100 μm. The elongated structure of nanowires may result in inherent chemical, electrical, magnetic, and optical anisotropy that can be exploited for interactions with cells and biomolecules in fundamentally new ways [6]. Although a majority of the magnetic carriers currently used in nanomedicine are magnetic nanospheres, nanowires are an alternative type of nanoparticle with considerable potential [53]. A magnetic nanotube is the hollow counterpart of a magnetic nanowire. Similar to magnetic nanowires, magnetic nanotubes have high aspect ratio, and usually have much stronger magnetization than magnetic nanospheres. As a nanotube has distinctive inner and outer surfaces, the inner surface and the outer surface of a nanotube can be functionalized to perform different biomedical functions. Depending on its inner diameter, the inner empty space of a nanotube can be used to capture, concentrate, and release biological entities ranging in size from small molecules to large proteins. The outer surface of a nanotube can be functionalized with environmentally-friendly molecules or probing molecules to a specific target [106]. Magnetic thin films are actually sheets of magnetic materials with thicknesses usually less than 100 nm. Magnetic thin films can have single-layer or multilayer structures, and they can be single-crystal, polycrystalline, or amorphous. One important property of a magnetic thin film is that its electrical resistance may change when an external magnetic field is applied, and this phenomenon is called
242
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
magnetoresistance. For some multilayer structures composed of alternating ferromagnetic and nonmagnetic layers, their resistances drop dramatically as an external magnetic field is applied, and this phenomenon is called giant magnetoresistance [2, 17]. 9.1.3
Magnetic Nanomedicine
Magnetic nanomedicine is a special branch of nanomedicine, dealing with the biomedical applications of magnetic nanomaterials. It is growing rapidly, and there is already a broad range of applications including cell separation, drug delivery, biosensing, studies of cellular functions, as well as a variety of other potential diagnostic and therapeutic uses. Among the four types of magnetic nanomaterials discussed above, magnetic nanoparticles, including nanospheres, nanowires, and nanotubes, are mainly used for labeling biological molecules and delivering drugs. Magnetic nanoparticles have attractive advantages in biomedical applications [89, 120]. First, because almost all the biological entities are nonmagnetic, magnetic nanoparticles in biological systems can be easily detected and traced. One typical example is the enhancement of the signal from magnetic resonance imaging (MRI) using magnetic nanoparticles. Second, magnetic nanoparticles may rotate under an external uniform magnetic field, and may make translational movements under an external magnetic field gradient. Therefore, magnetic nanoparticles, or magnetically tagged biological molecules, can be manipulated by applying an external magnetic field. This is important for transporting magnetically tagged drug molecules to diseased sites. Third, magnetic nanoparticles can resonantly respond to a time-varying magnetic field, transferring energy from the exciting magnetic field to the nanoparticles and the tagged biological molecules. This property has been used in hyperthermia treatment of cancer tumors [89]. Magnetic thin films are often used in the development of magnetic biosensors and biochips, for detecting biological entities tagged by magnetic nanoparticles. Due to their extremely high sensitivity and high accuracy, such biosensors and biochips can detect the existence of single pieces of target biological entities. This is crucial for the early diagnosis of some fatal diseases, such as cancer. 9.1.4
Status
Nanomedicine, including magnetic nanomedicine, has past its infancy, and advancements in this field are being made every day. Nanomedicine has become a large industry, with sales reaching 6.8 billion dollars in 2004. The United States and the European Union are investing billions of dollars, and plan to invest more in the future. Japan also invests heavily in this field. As nanomedicine continues to grow, it is expected to have significant impacts on healthcare and the economy. However, many challenges exist in the pathway for nanomedicine practice [35, 36, 83]. The major challenges are the strict requirements from the Food and Drug Administration (FDA), which embody its regulatory responsibilities, and are crucial for the safety of nanomedicine practice and the long-term development of this field [103].
9.2 Physical Background for Magnetic Nanomedicine
9.2
243
Physical Background for Magnetic Nanomedicine 9.2.1
Nanoparticles and Ferrofluids
Magnetic particles used in nanomedicine can be in the form of discrete nanoparticles, or in the form of magnetic microbeads made by embedding magnetic nanoparticles in a suitable matrix. In the design and fabrication of magnetic nanoparticles and microbeads, the biocompatibility and stability should be taken into full consideration [40]. In the synthesis of magnetic nanoparticles, three aspects should be considered [90]. First, magnetic nanoparticles should be crystalline, and each nanoparticle consists of only one domain. Second, the size distribution of the nanoparticles should be as narrow as possible. Third, all the magnetic nanoparticles in a particular sample should have a unique and uniform shape. In synthesizing magnetic microbeads, usually biocompatible matrices are used, and the concentration of the magnetic nanoparticles should be controlled to obtain the desired magnetic properties. In most of the biomedical applications, magnetic particles, including nanoparticles and microbeads, are dispersed in an aqueous or organic carrier medium, resulting in a ferrofluid. To obtain a stable biocompatible ferrofluid in physiological media at neutral pH and appropriate ionic strength, the particle surface should be functionalized. Usually these particles are coated with dextran, albumin, or synthetic polymers such as methacrylates and organosilanes. The effector, usually an antibody, is attached to the particle through a covalent bond to the coating polymer, and the stability of the coating determines the stability of the effector-particle complex [25, 46]. 9.2.2
Magnetic Manipulation
The manipulation of magnetic particles by external magnetic fields is required by most of the biomedical applications, such as magnetic separation and drug delivery. In the manipulation of magnetic particles, it is important to recognize that a magnetic field gradient is required to exert a translation force; while a uniform field gives rise to a torque, but no translational action [89]. The magnetic force acting on a magnetic particle is given by [89]: Ê B2 ˆ Ê1 ˆ Fm = Vm Dχ— Á B ◊ H˜ = Vm Dχ— Á ˜ Ë2 ¯ Ë2 μ 0 ¯
(9.1)
where Vm is the volume of the magnetic particle, Dχ = χ m - χ w is the effective susceptibility of the magnetic particle relative to the water, H is the strength of magnetic field, and B is the magnetic inductance. For the case of a dilute suspension of particles in pure water, we assume that B = μ0H. Equation (9.1) indicates that the magnetic force on a magnetic particle is related to the differential of the magnetostatic field energy density, (1/2)B⋅H. If Δχ > 0, the magnetic force is in the direction of the steepest ascent of the energy density scalar field. To effectively manipulate ferrofluids for desired biomedical functions, besides the magnetic forces on magnetic particles inside a ferrofluid, the physical properties
244
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
of the ferrofluid should also be taken into consideration. Foe example, the effects of magnetic field on ferrofluid viscosity should be studied [19]. 9.2.3
Fundamentals of Nanomagentism
Nanomagnetism has been under intensive investigation for decades [27, 48]. Here we discuss the fundamental magnetic properties of magnetic nanomaterials, with emphasis laid on the fundamentals of nanomagnetism related to their biomedical applications. We concentrate on the magnetism of magnetic nanospheres, and the effects of shape variations, such as triangular, square, and pentagonal, on the magnetic properties. A description of the variation of nanoparticle properties with shape can be found in Cowburn (2000) [26]. 9.2.3.1
Superparamagnetism
Because the thermal fluctuations of a very small particle prevent the existence of a stable magnetization, the coercity Hc of a magnetic particle approaches zero when the particle becomes very small. This phenomenon is related to superparamagnetism. There are two experimental criteria for superparamagnetism [13]. First, the magnetization curve exhibits no hysteresis, and second the magnetization curves at different temperatures must superpose in a plot of M versus H/T. Figure 9.1 shows the magnetization curves of iron amalgam on H/T bases. The magnetization curves at 77K and 200K superpose each other. The imperfect H/T superposition may be due to a broad distribution of particle sizes, and the changes in the spontaneous magnetization of the particle as a function of temperature or anisotropy effects. 20 15 10 5
30
20
10
10
20
30
5 10
77° K 200° K
15 20
Figure 9.1 Magnetization curves on H/T bases, as a demonstration of the superparamegnetism of iron amalgam (Bean and Jacobs 1956) [13].
9.2 Physical Background for Magnetic Nanomedicine
245
The fundamental mechanism of superparamagnetism is based on the relaxation time τ of the net magnetization of a magnetic particle [20]: Ê DE ˆ τ = τ 0 exp Á ˜ Ë kB T ¯
(9.2)
where ΔE is the energy barrier to moment reversal, and kBT is the thermal energy. For noninteracting particles the pre-exponential factor τ0 is on the order of 10-10–10-12 s and only weakly dependent on temperature. The energy barrier has several origins, including both intrinsic and extrinsic effects such as the magnetocrystalline and shape anisotropies, respectively [89]. However, in the simplest cases, it is given by DE = KV, where K is the anisotropy energy density and V is the particle volume. For small particles, DE is comparable to kBT at room temperature, therefore they may exhibit superparamagnetic properties at room temperature. 9.2.3.2
Nanoparticle Assemblies
Stoner-Wohlfarth theory describes the behavior of an assembly of single-domain, noninteracting particles with uniaxial anisotropy [108]. This theory analyzes how different types of anisotropies affect the magnetic properties, including the coercivity and remanence, of fine particles. One important conclusion about the remanence behavior can be derived from the Stoner-Wohlfarth theory [122]: M DCD ( H ) Ê M IRM ( H )ˆ = Á1 - 2 ˜ M( • ) M( • ) ¯ Ë
(9.3)
where MDCD is the dc demagnetization remanence, MIRM is the isothermal remanent magnetization, and M(∞) = MDCD(H=∞) = MIRM(H=∞). As most real materials could not strictly satisfy the noninteracting assumption required by (9.3), the deviation from (9.3) can be used to investigate the interactions between the particles in real materials. Kelly et al. (1989) suggested the difference term, DM(H), defined by [62]: DM( H ) =
M DCD ( H ) Ê M IRM ( H )ˆ = Á1 - 2 ˜ M( • ) M( • ) ¯ Ë
(9.4)
Positive values of DM(H) indicate the presence of stabilizing (ferromagnetic) interactions, while negative values indicate demagnetizing interactions [69]. 9.2.3.3
Core-Shell Nanoparticles
The magnetic particles in ferrofluids are usually covered with a surfactant layer, forming a core-shell structure. A core-shell nanoparticle usually consists of a magnetic core encapsulated in a protective shell that is usually biocompatible. The applications of magnetic nanoparticles can be significantly extended by the introduction of core-shell structure [27]. There are two types of core-shell structures. In
246
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
the first type, the core is magnetic while the shell is nonmagnetic; and in the second type, the core and the shell are made of magnetic materials with different hardness. A magnetic core-shell nanoparticle in the first type can be formed by surrounding the core with an oxide shell, which is a natural result of exposure to environmental oxygen. Such core-shell nanoparticles can also be synthesized chemically through tuning the dimensions of both layers and the interface between them. In such a structure, the core material can be ferromagnetic while the shell is antiferromagnetic, such as colloidal Co/CoO and CoNi/(CoNi)O. As shown in Figure 9.2(a), a shifted hysteresis loop can be observed due to the exchange anisotropy caused by the interfacial couplings. This effect has been used to control the magnetization of devices, such as a spin valve sensor, via the giant magnetoresistance effect. Though the exchange bias was discovered a long time ago [80], the microscopic mechanisms of exchange bias are yet to be fully understood. Most of the functionalized magnetic nanoparticles for biomedical applications are developed in this approach. Biomedical applications require rigorous surface functionalizations to make particles invisible to the reticulo-endothelial system of the body, and to prevent aggregation that would inhibit the transport of particles through the body. For example, by coating a magnetic core with a thin layer of gold, ligands with various functionalities can be introduced through Au-thiol chemistry, meanwhile maintaining the magnetic utility of the core [27]. The second type of core-shell magnetic nanoparticles can be synthesized by growing another ferromagnetic material on a ferromagnetic core. Figure 9.2(b) shows a core-shell structure consisting of a soft ferromagnetic core covered by a hard ferromagnetic shell. Due to the magnetic exchange coupling between the core and the shell, such a system combines a large coercive field and large magnetization. In such a system, the interphase coupling between the soft and hard ferromagnetic materials can be tuned [125, 126, 127], and extremely strong permanent magnets can be achieved by optimizing the interphase coupling [27]. Though there are many problems to be addressed for single-component and basic core-shell systems, and many applications could be developed based on sin-
(a)
(b)
Figure 9.2 (a) M-H hysteresis loops of an exchange bias system. The center of the loop is shifted to the side after field cooling from above the Neel temperature due to coupling between antiferromagnetic and ferromagnetic layers. (b) M–H hysteresis loops of an exchange spring system. Coupling of the soft and hard ferromagnets leads to large coercivity and large magnetization. (Darling and Bader 2005) [27].
9.3
Magnetic Nanoparticles
247
gle-component and basic core-shell systems, researchers are becoming interested in the synthesis and characterization of “onion” particles, which are composed of many layers [27]. The properties of multilayer core-shell particles could be further tailored to meet special application requirements.
9.3
Magnetic Nanoparticles 9.3.1
Basics of Magnetic Nanoparticles
The magnetic nanoparticles discussed in this section are mainly magnetic nanospheres. It has been well known that the magnetic properties of magnetic nanoparticles are correlated with their nanostructure. Hence, a classification of nanostructured magnetic morphologies was desired. Leslie-Pelecky and Rieke proposed a classification, which was designed to emphasize the magnetic behavior-related physical mechanisms [69]. The classification is illustrated in Figure 9.3. Type A represents systems consisting of isolated particles with nanoscale diameters. Since the interparticle interactions cannot be ignored for these systems, their unique magnetic properties are completely determined by the isolated components with their reduced sizes. Another type, type-D, is denoted as bulk materials with nanoscale structure, which has a significant fraction (up to 50%) of the sample volume composed of grain boundaries and interfaces. Unlike type-A systems, the interparticle interaction in type-D systems can not be ignored and the bulk magnetic properties for type-D are indeed dominated by the interactions. Because of these existing interactions and grain boundaries, the magnetic behavior of type-D nanostructures cannot be predicted theoretically. Other than type-A and type-D systems, intermediate forms such as core-shell nanoparticles (type-B) and nanoparticle-based nanocomposites (type-C) are classified. For type-B systems, the shells on magnetic nanoparticles are usually used to reduce the interparticle interactions. The magnetic properties of type-C systems, nanocomposites, are determined
Figure 9.3 Schematic of different types of magnetic nanostructured materials. (Leslie-Pelecky and Rieke 1996) [69].
248
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
by the fraction of magnetic nanoparticles and the characteristics of the matrix materials. As the ultrafine magnetic particles are considered, single-domain and multidomain are important for their magnetic property. With their characteristic widths and energies, domain walls separate domains-groups of spins all pointing in the same direction and acting cooperatively. The motion of domain walls induced reversing magnetization. Multidomains are formed particularly for large particles where domain walls form energy favorably. As the particle size decreases below a critical diameter, Dc, single domain particles form and the formation of domain walls becomes energetically unfavorable. In this case, magnetization reversal can not be obtained readily leading to larger coercivities because of the lack of nucleation and motion of the domain walls. If the particle size continues to decrease, the thermal fluctuations will significantly affect the spin and this phenomenon is called superparamagnetism. Theoretical simulation has demonstrated the existence of a stable magnetization with zero coercivity Hc for very small magnetic particles. Superparamagnetic nanoparticles are featured by no hysteresis for the magnetization curve and overlapping of the magnetization curves at different temperatures. However, in practical measurements, imperfect superposition was always observed. Possible reasons include anisotropy effects, a wide distribution of particle sizes, and changes of spontaneous particle magnetization based on temperature. 9.3.2
Synthesis Techniques
Efforts have been made to synthesize uniform nanoparticles for both technological and fundamental scientific importance. Particularly, magnetic nano-particles/ nanocrystals with 2- to 20-nm dimensions are attractive [56]. These ultrafine magnetic nanoparticles have broad applications in magnetic ferrofluids, contrast enhancement in magnetic resonance imaging (MRI), highly active catalysts, magnetic refrigeration systems, magnetic carriers for drug targeting, hyperthermia treatment, and so forth. Colloidal chemical synthesis is a powerful approach for oxide nanomaterial synthesis. Generally, in a colloidal system, separated nanoparticles with ultrafine dimensions are stabilized by adding surfactant reagents resulting in a uniform suspension in a solvent. Magnetic colloidal systems consist of magnetic nanoparticles/nanocrystals and they are examples of type-A, where contributions from interparticle interactions are negligible. Hence their unique magnetic behaviors are primarily attributed to their reduced dimension. The role of surfactants playing in colloidal systems is to prevent magnetic nanoparticles clustering. Consequently, these surfactant-attached magnetic nanoparticles can be suspended in various solvents uniformly. Several synthetic procedures have been reported for synthesizing monodisperse magnetic nanoparticles. Generally, all of the chemical methods share the same growth procedure that involves a sudden burst of nucleation process followed by a slow growth in a controlled manner. The hot-injection solvothermal method represents a promising technique, in which neutral organometallic precursors are used in a coordinating alkyl solution with a high boiling point [28]. Figure 9.4 depicts a typical hot injection synthetic procedure, where the rapid injection of the reagents (often organometallic compounds) into a hot solution with surfactants causes a
9.3
Magnetic Nanoparticles
249
Figure 9.4 Schematic of synthesis procedures of monodisperse magnetic nanoparticles by injecting precursors into a solution containing surfactants at high temperature followed by a post-treatment process such as aging and a size control process. (Hyeon 2003) [56].
simultaneous formation of many nuclei [56]. Alternatively, chemical precursors can be stirred at room temperature followed by slowly heating in a controlled manner to generate nuclei. Then, reactive species are added for particle growth. Oswalt ripening, described as smaller nanocrystals dissolving and precipitating on nanoparticles with larger sizes, is generally the process for larger particle formation. The growth termination is usually performed by promptly dropping the temperature. These synthetic procedures usually produce nanoparticles with a narrow particle size distribution (σ ~ 10%). And the particle sizes can be controlled by carefully changing the reaction conditions including reaction temperature, reaction time, and different surfactants, and reagent concentrations. In general, larger particles are obtained with longer reaction times and a higher reaction temperature. Additionally, it is possible to further narrow the particle size distribution (σ < 5%) by additional size-selection processes, such as adding a relatively poor solvent to deposit the particles in larger diameters. When adding a poor solvent to a suspension of nanoparticles with various dimensions, the larger particles aggregate and precipitate first due to the strong van der Waals attraction between each other. Based on elemental compositions, magnetic nanoparticles can be divided into monometallic, metal alloy, metal oxide, and core-shell nanoparticles. As an example, Hyeon et al. (2001) [57] reported their synthesis of highly crystalline and monodisperse γ-Fe2O3 nanocrystallites. The synthetic strategy is to prepare monodisperse iron nanoparticles followed by further oxidation to iron oxide. Figure 9.5 a is a TEM image showing the hexagonal arrangement of monodisperse γ-Fe2O3 nanoparticles with 11-nm diameter in a closed packed way. The high-resolution transmission electron micrograph shown in the inset not only demonstrates the uniformity of the nanoparticle size but also reveals the highly crystalline nature of the γ-Fe2O3 nanoparticles. The temperature-dependent magnetization of γ-Fe2O3 nanoparticles was measured using a superconducting quantum interference device following the zero-field cooling (ZFC) and field cooling (FC) procedures in an applied magnetic field of 100 Oe at 5 to 300K. Figure 9.5(b) illustrates the relation between temperatures and magnetization of 4-, 13-, and 16-nm γ-Fe2O3
250
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
Figure 9.5 (a) TEM image of a two-dimensional assembly consisting of γ-Fe2O3 nanocrystals. The inset shows a high-resolution image. (b) Plot of the temperature-dependent magnetization of γ-Fe2O3 nanocrystallites with different diameters measured with zero-field cooling at the applied magnetic field of 100 Oe. (Hyeon et al. 2001) [57].
nanocrystals. From the plots, it is obvious that the magnetizations reduce at certain temperatures for all samples. These blocking temperatures were found to be around 23, 185, and 290K for the γ-Fe2O3 nanocrystals of 4-, 13-, and 16-nm diameters, respectively. Another promising synthesis technique, the so-called reverse micelle method, can be used to prepare magnetic nanoparticles. In this method, reverse micelles serve as templates for nanoparticle formation. Usually, the micelles were obtained by adding surfactants. For instance, for champagne cork-shaped surfactants consisting of branched hydrocarbon chains and small polar heads, reverse micelles with spherical shapes are formed. As illustrated in Figure 9.6, the inner core is created by the head groups and the chains form the outer surface [92]. This spherical reverse micelle is also termed a water-in-oil droplet. It was found that the dimension of reverse micelles increases linearly with the amount of water added to the system; hence the diameter of the synthetic nanoparticles can be controlled. It is believed that this
Figure 9.6 Schematic of generalized synthesis of nanoparticles in aqueous solutions by a reverse micelles method. (Pileni 2003) [92].
9.3
Magnetic Nanoparticles
251
synthesis technique is rather general and that a broad range of spherical nanomaterials can be prepared such as semiconductors, metals, oxides, and alloys. Particularly, this method is ideal for the synthesis of some alloy nanoparticles that cannot be produced readily by other synthesis methods. Another merit of this reverse micelle method is its high purity. Biorelated synthesis is a relatively new technique used for magnetic nanoparticle synthesis. In biological methods, biological entities usually serve as templates for nanoparticle formation, and assembly of crystalline inorganic materials can be regulated by biological organisms under environmentally benign conditions. For example, the use of porous protein crystals, manipulation of bacteria to produce oxide nanoparticles, and selection of metal-specific polypeptides from combinatorial libraries [95]. To prepare magnetic nanoparticles, 24-subunit ferritin can be used. Self-assembly of ferritin will form a spherical cage with a ~8.0 nm-diameter cavity, which can be used for the biological storage of iron in the form of ferrihydrite, an iron (III) oxy-hydroxide. Protein, named Apoferritin, has been used as a generic reaction container for the synthesis of a variety of magnetic materials, such as iron and manganese oxides. Other than regular magnetic nanoparticles, this protein has also been used for the production of superparamagnetic magnetite and semiconducting cadmium sulphide nanoparticles. A good example of using apoferritin protein for nanoparticle synthesis is Co–Pt nanoparticle dispersion [79]. This two-step synthesis process needs to be repeated multiple times so that nanoparticles continually grow and fill the cavity of proteins resulting in a high yield. Then nanoparticles inside the protein cavity can be obtained from the dispersion by filtration and L1 phase nanoparticles formed after thermal annealing. 9.3.3
Functionalization Techniques
Functionalization of nanomaterials is an effective approach to modify their surface properties. For certain applications, it is essential to tune the surface characteristics of nanomaterials. Similarly, for magnetic nanoparticles, functionalization has been used particularly for: (1) agglomeration prevention and (2) biointeraction enhancement. The first functionalization technique is based on organic molecules. Magnetic nanoparticles modified with organic molecules have been widely used for biomedical and biotechnological applications, such as cell separation, drug delivery, hyperthermia, automated DNA extraction, gene targeting, and magnetic resonance imaging. Organic molecules including chemically functional groups and polymer chains can be linked to magnetic nanoparticles by bindings. Covalent bonding is a widely used functionalization technique. Alternatively, biofunctionalization with biological entities has its own unique characteristics. Biofunctionalization is a surface modification technique that attaches biological entities such as proteins, ligands, antibodies, and enzyme, to synthetic materials. Biofunctionalized magnetic nanoparticles have unique interaction with biological entities. And assisted by biological recognition elements attached to their surface, magnetic nanoparticles are able to access to the target specifically. For example, magnetic nanoparticles coated with antibodies can be used for sensitive immunoassays. Also, magnetic nanoparticles modified with single-stranded DNA can be used to identify organisms and analyze single-nucleotide polymorphism due to the DNA hybridization capa-
252
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
bility. This type of specific binding offers several advantages in biomolecule detection over conventional radioactive, electrochemical, optical, and other techniques [86]. However, surface modification chemistries of magnetic nanoparticles are left far behind their synthesis achievements. Effective biofunctionalization of magnetic nanoparticles is still in demand for various biological applications [50]. 9.3.4
Biomedical Applications of Magnetic Nanoparticles
Understanding biological processes and hence developing biomedical applications has been continuously pursued. The interest in magnetic nanoparticles for bioapplications has come from their comparable dimensions to biological entities coupled with their unique magnetic behavior. Though common living organisms are composed of cells with size of about 10 μm, the cell components are much smaller and generally in the nanosize dimension. Synthetic magnetic nanoparticles have adjustable dimensions and very fine nanoparticles can be synthesized by carefully designing experimental procedures and controlling experimental conditions. With such a nanoscale dimension, it would be possible for magnetic nanoparticles to get close to a biological entity of interest. Moreover, the interaction between magnetic nanoparticles and biological entities can be tuned by biofunctionalization. This offers a controllable means of tagging or addressing the binding at the nanoscale. It is their comparable dimension and magnetic properties that has prompted an idea of using magnetic nanoparticles as very small probes to spy on biological processes at the cellular scale without introducing too much interference [98]. The diagnostic applications of magnetic nanoparticles include processes to detect malignant tissues or pathogenic bioaggregates by using magnetic nanoparticles. Common applications include enhancement of magnetic resonant imaging, magnetic labeling, magnetic separation and purification, spatially resolved magnetorelaxometry, biological assay systems, and magnetic nanosensors. Magnetic nanoparticles are able to enhance the image contrast in MRI, which has attracted enormous attention. Magnetic resonance imaging is one of the most powerful medical tools for diagnostic purposes owing to its noninvasive process, high spatial resolution, and multidimensional tomographic capabilities. However, this technique suffers by low-signal sensitivity. To overcome the weakness of current MRI techniques, signal enhancers need to be used. Magnetic nanoparticles/ nanocrystals have been demonstrated for their signal-enhancing capabilities. Magnetite (Fe3O4), as a member of clinically benign iron-oxide-based nanoparticles, has been widely explored for signal-enhancing purposes. Jun et al. (2005) demonstrated that magnetite (Fe3O4) nanoparticles are a signal enhancer in MRI [60]. In modern biology or biomedicine, separation of particular biological entities (e.g. DNA, proteins, ions, and molecules) from their native environment is often required. Magnetically labeled biological entities exhibit magnetic properties so that they can be separated by applying an external magnetic field. Prompted by this idea, magnetic separation and purification have been widely used to prepare concentrated biosamples in biomedical applications. Spatially resolved magnetorelaxometry is a novel technique to detect magnetic labeled biological entities by measuring the relaxation of their magnetization when the magnetizing field is removed. This technique also deals with magnetic labeling of biomolecules. Research investigation has
9.3
Magnetic Nanoparticles
253
demonstrated its feasibility of detecting and locating immobilized magnetic nanoparticles in vivo. More and more interest in spatially resolved magnetorelaxometry has been apparent, especially in medical diagnostics, with respect to the developing technology in diseases at the molecular level. A biological assay system is widely used for biomaterial detection. Analyses of DNA, protein, and other biotargets can be realized by using an array-based bioassay for advanced medical care and environmental measurements. Recently, using a combination of magnetic nanoparticles and a patterned substrate covered with a self-assembled monolayer provides a promising technique for magnetic detection of biomolecular interactions [88]. The novel concept of using magnetic nanoparticles in biosensors, stemming from the magnetic control of bioelectrocatalysis, was also developed by [49]. Dual biosensing of two analytes, glucose and lactate, by magneto-controlled bioelectrocatalysis has been demonstrated [61]. The unique magnetic properties of magnetic nanoparticles can be used in therapeutic applications as well. Described by Coulomb’s law, magnetic nanoparticles can be manipulated by an external magnetic field gradient. Magnetic nanoparticles are able to transport into human tissue due to the intrinsic penetrability of magnetic fields into the human body. This “action at a distance” opens up many potential bioapplications. Another important property of magnetic nanoparticles is their resonant response related to a time-varying magnetic field [89]. Hence energy transfer from the exciting field to the magnetic nanoparticles is able to be achieved. In this way, certain amounts of thermal energy can be delivered via magnetic nanoparticles to the targeted tumor cells and induce malignant cell destruction. This process is called hyperthermia. Gilchrist et al. (1957) did these experimental investigations for the first time when they heated various tissue samples with γ-Fe2O3 particles 20-100 nm in diameter by a 1.2 MHz magnetic field [49]. Since then, studies have shown the feasibility of using the hyperthermic effect generated from magnetic nanoparticles by applying a high-frequency AC magnetic field as an alternate therapeutic approach for cancer treatment. Briefly speaking, the hyperthermic effect is generated from the relaxation of magnetic energy of the magnetic nanoparticles which is able to destroy tumor cells effectively [70]. Interestingly, it was found that magnetic nanoparticles themselves exhibited antitumor effects [105]. Previous experiments on mammary adenocarcinoma revealed that injected biocompatible magnetic fluid determined the lysis of the tumor cells. An interesting phenomenon was observed regarding to the significant difference of magnetic nanoparticles endocytosis by tumor cells and normal cells. Tumor cells are able to take up a high quantity of magnetic nanoparticles from extracellular matrix while nanoparticles in normal cells are not present. Controlled drug delivery is of great importance for therapeutic treatments. Various materials have been investigated as the drug carrier to bring drugs, genes, and proteins to target specific areas in the living body. The potential of magnetic nanoparticles for drug delivery applications stems from their inherent magnetic properties coupled with other common properties shared with other nanomaterials. Their unique magnetic properties include superparamagnetism, high saturation magnetization, and high magnetic susceptibility. Also, chemical and biofunctionalization techniques have been developed to improve both the stability and the biocompatibility of magnetic nanoparticles. Researchers can select suitable coating materials on magnetic nanoparticles for
254
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
particular drugs and certain targets. The most attractive characteristic of using magnetic nanoparticles in drug delivery is their capability of controlled delivery to a specific area by an external magnetic field. Selectively delivering drug molecules to the diseased site while not increasing their level in the healthy tissues of the organism are always desirable in drug delivery for cancer treatment [128]. In cancer treatment, magnetically controlled drug targeting is based on binding anticancer drugs with ferrofluid and desorbing them from the ferrofluid after reaching the area of interest by means of an external magnetic field [74]. For magnetic nanoparticles, most bioapplications depend upon the relationship between the external magnetic field and the biological system. Magnetic fields should not be deleterious to either biological tissues or biotic environments. On the other hand, magnetic nanoparticles are usually injected intravenously into a human body and are transported to the targeted region via blood circulation for biomedical diagnostics or treatment. Therefore a desirable magnetic medium should not contain nanoparticle aggregation, which will block its own spread. The stability of this magnetic colloidal suspension depends on two parameters: an ultrasmall dimension and surface chemistry. The particle size should be sufficiently small to avoid precipitation due to gravitation forces while the charge and surface groups should create both steric and coulombic repulsions that stabilize the colloidal suspensions. For many biomedical applications, magnetic nanoparticles presenting superparamagnetic behavior (no remanence along with a rapidly changing magnetic state) at room temperature is desirable.
9.4
Magnetic Nanowires 9.4.1
Typical Structures of Magnetic Nanowires
Many efforts have been made in optimizing the composition, shape, size, and surface chemistry of magnetic nanoparticles for desired biomedical applications. In this respect, nanowires have attractive advantages. Magnetic nanowires possess unique properties that are quite different from those of bulk ferromagnetic materials, spherical particles and thin films. As schematically shown in Figure 9.7, the architecture and composition of a nanowire along its axis can be modulated precisely, and this could be used to precisely control the magnetic properties of nanowires for special biomedical applications [110]. Furthermore, as shown in Figure 9.7(d), by using ligands that selectively bind to different segments of a multisegment wire, spatially modulated multiple functionalization can be realized in these wires, and this feature can be used to improve the performances of magnetic nanowires in biomagnetic applications [94]. The properties of both arrayed and dispersed magnetic nanowires have been extensively investigated (Maurice et al., 1998). However, in most of the biomedical applications, discrete magnetic nanowires are used, and they are usually suspended in ferrofluids. 9.4.2
Synthesis of Magnetic Nanowires
Many methods have been developed for the synthesis of magnetic nanowires, such as ion-beam and electron-beam nanolithography, evaporation condensation,
9.4 Magnetic Nanowires
255
Figure 9.7 Inherent shape anisotropy and functionalization of nanowires. (a) Single-component nanowire; (b) two-component nanowire in which the aspect ratio of each segment is greater than one; (c) two-component multilayer nanowire, in which the aspect ratio of each segment is less than one; and (d) functionalization of a two-component nanowire. In the figure, ligands L and L’ selectively bind to the two components, and thus the functional groups R and R’, corresponding to L and L’, respectively, are spatially separated. (Sun et al. 2005) [110].
vapor–liquid–solid (VLS) growth, hydrothermal synthesis, and chemical synthesis [55, 102]. Among these methods, a template-assisted electrodeposition method is the most widely used. As shown in Figure 9.8(a), electrodeposition of nanowires is usually done in a three-electrode arrangement, consisting of a reference electrode, a specially designed cathode and an anode or counter electrode [14]. An electrical current passes through an electrolyte of metallic ions, and a reduction takes place when the ion encounters the cathode (working electrode). Because of the existence of the nanoporous membrane, electrodeposition takes place inside the channels of the membrane [7]. Figure 9.8(b) shows the time dependence of the electrodeposition current in the fabrication of nanowires. As the material is electrodeposited, the nanowires grow from the bottom. At Region I, materials are deposited in the pores of the membrane until the pores are fully filled. At Region II, the material grows out of the pores, forming hemisphere caps on the ends of the nanowires. At Region III, the hemisphere caps formed at Region II form a contiguous surface over the surface of the membrane. To obtain discrete nanowires, the growth should be stopped somewhere within Region I, and discrete magnetic nanowires can be obtained by chemically dissolving the membrane. An advantage of electrodeposition is that different materials can be sequentially deposited in the templates, yielding nanowires comprised of different segments [55, 93, 121]. This is usually achieved by altering the applied potential of a solution with more than one precursor, or by changing the deposition solution [6, 7]. Fert and Piraux (1999) developed a pulse-plating method in which two metals are deposited from a single solution by switching between the deposition potentials of the two constituents [34]. Such a single-bath method can be used to make various multilayer nanowires, such as Co/Cu, NiFe/Cu, Ni/Cu, and CoFe/Cu [18, 23, 93]. Using different electrolytic solutions and depositing at different potentials, a variety of alloys
256
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
(a)
(b)
Figure 9.8 (a) Three-electrode arrangement for electrodeposition of nanowires (Bera et al. 2004) [14]. (b) Time dependence of the deposition current in the synthesis of nickel nanowires (Bauer et al. 2004) [6].
with controlled composition can be fabricated, such as NiCu, NiFe, CoNiFe, CoFe, CoPd and CoPt [31, 76]. 9.4.3
Functionalization of Magnetic Nanowires
Surface functionalization of nanowires is necessary for realizing their biocompatibility and functionality. It is often used to tailor surface properties (such as hydrophilicity, hydrophobicity, surface charge), impart other properties (such as fluorescence), and introduce molecular recognition for small molecules (such as drugs), biopolymers (such as peptides, proteins, and DNA), protein assemblies (such as viruses) and other nanoparticles (such as particle-DNA conjugates) [121]. Multisegment nanowires represent a unique platform for engineering multifunctional nanoparticles for varieties of biomedical applications [85] and bring exciting new perspectives for surface functionalization that can modify the chemical and biological properties of nanowires [77]. Generally speaking, there are two approaches for surface functionalization: chemical functionalization [7] and biological functionalization [121]. In the following, we discuss chemical functionalization. The most direct approach for functionalizing multisegment nanowires is to transfer the well-developed surface chemistry at planar metal interfaces to the nanowire geometry. It has been shown that porphyrins with terminal carboxylic acid groups bind selectively to the native oxide on nickel in single-component nickel nanowires (Tanase et al. 2001), and in two-segment gold–nickel nanowires [7]. Bauer et al. (2003) used HemIX as a fluorescent probe to quantify the coordination of carboxylic acids to the native oxide on nickel [7]. HemIX has two carboxylic acid groups linked to the porphyrin ring by a flexible ethane spacer. When two-segment nickel-gold nanowires are reacted with HemIX, the nickel segment shows uniform fluorescence, while the gold segment shows weak and nonuniform fluorescence. The gold segment could be made nonfluorescent by adding a long-chain thiol to the reaction mixture, and the thiol likely displaces any weakly bound HemIX. Figure 9.9 shows optical images of a nickel-gold nanowire with a nickel-gold segment ratio of
9.4 Magnetic Nanowires
257
Figure 9.9 Nickel-gold nanowire functionalized with HemIX and nonylmercaptan. The diameter of the nanowire is about 350 nm, and the length of the nanowire is about 22 μm. (a) Reflection image of the nanowire. (b) Fluorescent image of the nanowire. (Bauer et al. 2003) [7].
2:3, functionalized with HemIX and nonylmercaptan. Figure 9.9(a) is a reflection image of the nanowire, and Figure 9.9(b) is a fluorescence image of the same nanowire. 9.4.4
Biomecial Applications of Magnetic Nanowires
The special properties of magnetic nanowires make them attractive for biological applications. Due to their large magnetic moments and shape anisotropy, strong forces, and torques can be applied by external magnetic fields. As the diameter and length of magnetic nanowires can be independently controlled, magnetic nanowires with suitable dimensions for biological entities with different length-scales can be synthesized. Furthermore, magnetic nanowires usually do not disrupt the growth cycle of cells, and biologically active molecules can be functionalized on them so that the defined biomedical functions can be performed [7]. The typical applications of magnetic nanowires mainly include manipulation of biomolecules, suspended biosensing systems, drug and gene delivery, and hybrid devices with magnetic nanowires. 9.4.4.1
Manipulation of Biomolecules
Manipulation of biomolecules using magnetic nanowires is based on the bindings between magnetic biomolecules and magnetic nanowires, and it is the basis of many biomedical applications of magnetic nanowires. For example, magnetic nanowires can be used in performing high-yield separations of biomolecules. Generally speaking, magnetic nanowires outperform magnetic spherical beads in biomolecules separations. Hultgren et al. (2004) studied the applications of Ni nanowires in cell separation [54]. It was found that high-purity separations can be achieved for nanowires over a wide range of sizes, while the optimum separation yield is achieved when the average length of the nanowires matches the average diameter of the cells. Figure 9.10(a) shows transmitted light images of trypsinized 3T3 cells attached to nickel nanowires. If the diameter of the cell is less than the length of the nanowire, the nanowire protrudes from the cell, whereas if the diameter of the cell is larger than
258
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
(a)
(b)
Figure 9.10 (a) Optical images of trypsinized 3T3 cells. Top row: normal cells (average diameter 15 μm) with (i) 15-μm nanowire and (ii) 22-μm nanowire. Bottom row: treated cells (average diameter 23 μm) with (iii) 10-μm nanowire and (iv) 22-μm nanowire [54]. (b) Percent yield as a function of nanowire length for separations of 3T3 cell populations with average diameters d = 15 μm and 23 μm [54].
the length of the nanowire, the nanowire is enclosed by the cell. Usually, the cells with larger diameter are able to engulf longer nanowires. Two parameters, purity and yield, can be used to compare the effectiveness of cell separations using beads and nanowires [54]. Purity refers to the percentage of cells that have a magnetic particle attached in all the captured populations. For the beads and the 5-μm nanowires, the purity is about 40 %, while for nanowires longer than 15 μm, the purity is increased up to 80%. The percent yield is the number of captured cells tagged with magnetic particles normalized by the initial number of cells tagged with magnetic particles. Figure 9.10(b) shows the percent yields of nanowires and beads for the separations of 15- and 23-μm diameter cell populations. For both populations, when the length of the nanowires is equal to the diameter of the cells, the percent yield reaches its maximum value, and the maximum yield is about four times larger than the percent yield when the beads are used. This effect suggests the potential to magnetically separate cell populations based on their sizes. Both single-segment and multisegment magnetic nanowires have been used for cell separations. Lee et al. (2004) demonstrated that Au/Ni/Au multisegment nanowires have rapid binding kinetics in the separation of His-tagged proteins [68]. When the Au/Ni/Au multisegment nanowires are introduced to a solution containing both His-tagged and untagged proteins, the His-tagged proteins attach to the nickel segment of the nanowire, and can be removed from solution by applying an external magnetic field. In a similar way, the Au/Ni/Au multisegment nanowires functionalized with poly-His can be used to efficiently separate mixtures of anti-His proteins from other antibodies. 9.4.4.2
Suspended Biosensing System
As schematically shown Figure 9.11, multilayer nanowires can be used as a substrate in a biosensing platform for sandwich immunoassays. The multilayer
9.4 Magnetic Nanowires
259
Figure 9.11 (a) Analogy between a conventional barcode and a metallic segment-encoded nanowire. Ni segments (50 nm) are deposited at both ends of the magnetic nanowire (not drawn to scale). (b) Schematic of the sandwich immunoassay performed on a nanowire. (Tok et al. 2006) [112].
nanowires consist of submicrometer layers of different metals, and are usually synthesized by electrodeposition within a porous alumina template. As a lot of variations can be realized in the synthesis of nanowires, a large number of unique yet easily identifiable encoded nanowires can be included in a multiplex array format. Tok et al. (2006) studied the applications of multilayer metallic nanowires in a suspended format for rapid and sensitive immunoassays [112]. The basic working principle of a suspended biosensing system is illustrated in Figure 9.11(b). In a suspended biosensing system, the target analytes are captured and hybridized in solution. To ensure that the nanowires can be manipulated by external magnetic fields, an appropriate ferromagnetic metallic component, for example, nickel, is incorporated. Usually the nickel segments integrated at both ends of the nanowires have a length of 25 to 150 nm. The easy magnetization axis for these disk-shaped magnetic segments is perpendicular to the nanowire axis, and thus these nanowires align perpendicular to the externally applied magnetic field. 9.4.4.3
Gene Delivery
Gene delivery using multisegment magnetic nanowires exhibits obvious advantages. As the properties of conventional gene delivery systems could not be controlled at the nanoscale, they are limited by their relatively low transfection efficiency, which limits the ability of the system to incorporate foreign DNA inside a target cell [55]. However, in the fabrication of multisegment nanowires, the materials of each segment and their properties can be precisely controlled to nanometer dimensions. Furthermore, multisegment nanowires can be endowed with different functionalities in spatially defined regions, and so the precise control of antigen placement and the stimulation of multiple immune responses can be achieved. Salem et al. (2003) investigated the application of electrochemically synthesized Au/Ni nanowires for therapeutic purposes, and confirmed that multisegment nanowires are more efficient in transfection than transferring-modified single-component nanowires [99].
260
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
9.4.4.4
Hybrid Devices
Hybrid devices, based on molecular biology and micro/nanofabrication, can be used in the development of advanced biosensors and force bioactuators for medical and therapeutic applications. Among various hybrid systems, special attention is paid to motor proteins, such as Adenosine Tri-Phosphate synthase (ATPase), which can convert the chemical energy derived from the ATP hydrolysis into mechanical work [87]. Ren et al. (2006) integrated nanowires with F1-ATPase motors [96]. The multisegment (Ni/Au/Ni) nanowires are synthesized by electrochemical deposition. The thiol group modified ssDNA and the biotinylated peptide are selectively bound to the gold and nickel segments, respectively. The F1-ATPase motor only attaches to the nickel segment of the nanowire by biotin-streptavidin linkage. Figures 9.12(a) and (b) show the schematic of the motor device and the TEM pictures of the multisegment nanowires, respectively. The rotations of the multicomponent nanowires driven by F1-ATPase motors can observed.
9.5
Magnetic Nanotubes 9.5.1
Magnetism of Magnetic Nanotubes
Nanotubes have become a fast developing research area of nanotechnology. Theoretical simulation indicated that a narrow tube might be less energetically favored than a finite strip because of the strain energy coupled with bending. However, when the diameter is reduced to a critical value (in nanometers), the strain in the nanotubes becomes smaller than the energy associated with the edges in the layered strips and the self-closed cylindrical geometry, which is free of dangling bonds, becomes the most stable structure [104]. In general, the advantage of nanotubes with hollow structures can be used as pipes, cavities, or capsules at the nano scale. Additionally, nanotube-imbedded nanostructured hybrid systems with extremely large surfaces have considerable advantages over nanoparticle-based systems in many applications, including catalysis and sensor technology. Magnetic nanotubes exhibit unique magnetic properties. The magnetic properties of magnetic nanotubes depend upon several factors such as the elemental com-
(a)
(b)
Figure 9.12 Schematic of rotation of multicomponent nanowire driven by an F1-ATPase motor. (a) The nickel segment on the nanowire is used as the region attached to the rotating shaft on the F1-ATPase motor. (b) TEM picture of three segment nanowires. (Ren et al. 2006) [96].
9.5 Magnetic Nanotubes
261
position, crystallinity, shape, diameter, wall thickness, and orientation. For instance, for nanotubes with small and large diameters, the preferential arrangements of the magnetic moments for magnetic nanotubes are the parallel configuration and vortex state, respectively [32]. As the inner diameter of a nanotube decreases, its coercivity and remanence increase accordingly. In the case of closely spaced oriented magnetic nanotube arrays, the preferential magnetization direction is perpendicular to the nanotube axis due to the strong interactions among individual nanotubes and the alignment [3]. Because of their shape anisotropy, magnetic nanotubes exhibit enhanced anisotropic magnetic behavior. Compared with solid magnetic nanowires, magnetic nanotubes have lower weight density that gives them higher stability in solutions with less chance of precipitation. Also, magnetic nanotubes and magnetic nanowires display differences in the remanent state, thus showing different magnetic or magnetoresistive behavior. To study the static distribution of magnetic moments, a theory of magnetic charge was used to calculate the demagnetization factor of the magnetite nanotubes (Fe3O4) [118]. The demagnetization factor is defined as the ratio of the negative of the demagnetizing field to the magnetization of a magnetic sample. Figure 9.13(a) illustrates a schematic nanotube with a three-dimensional orthogonal coordinate set up in the center of the nanotube. In the calculation, the length of the nanotube was defined as 10 µm and the values of the outer and inner radius were 100 and 88 nm, respectively. An external magnetic field along the Y axis was assumed to be applied to magnetize the nanotube to saturation. Calculations based on the theory of magnetic charge gave the demagnetization factors in different areas of the nanotube wall in the X-Y plane. The result is plotted in Figure 9.13(b). The X axis is β, which is the angle between the Y axis and the line connecting the original point O
Figure 9.13 (a) Schematic structure of a magnetic nanotube; and (b) the relationship between the demagnetization factor and angle β. (Wang et al. 2006) [118].
262
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
and the point on the wall. From Figure 5.1(b), it can be seen that the demagnetization factor reaches the highest value of 0.94 when β is 0 (point A in the nanotube wall) while lowest value of 0.06 is achieved at β = 90° (point B in the nanotube wall). In the range of 0-90°, the factor decreases with the increase of the angle. In magnetic nanotubes, the mechanism of magnetization reversal needs to be clarified. Several reversal mechanisms, induced by a homogeneous external magnetic field, are illustrated in Figure 9.14 [109]. Flux closure during magnetization reversal induces magnetization curling, which can be seen in Figure 9.14 (a, b). Figure 9.14(c) depicts a coherent rotation that leads to surface poles. For magnetic nanotubes, as the radius R is larger than a critical value, a transition takes place from coherent rotation to curling. Particularly, this transition only occurs for nanotubes with very small diameters. For the curling mode in tubes having a wall thicknesses much smaller than the tube radius (t«R), the exchange energy Hn can be expressed as: Hn = Ha +
A μ 0 MsR 2
(9.5)
where Ha is the anisotropy field, A is the exchange stiffness, and Ms is the spontaneous magnetization. From this formula, the value of radii for curling to occur is 2ζ0, and ζ0, the exchange length of the system is (A/μ0Ms2)1/2. Magnetic nanotubes have a lower exchange energy (a factor of 5 smaller value) compared with other magnetic nanomaterials. The reason is believed to be the absence of curling related vortices. In the case of polycrystalline nanotubes, the curling character of the reversal mode may be distorted by polycrystallinity. A mode without curling-type flux closure is shown in Figure 9.14(d). In this mode, the curling mode converts to a localized mode at radius Rrand, which is determined by the magnetocrystalline anisotropy. The reversal mode shown in Figure 9.14(d) is favorable for anisotropy, but unfavorable for both
(a)
(b)
(c)
(d)
(e)
Figure 9.14 Schematic reversal modes in magnetic nanotubes. (a) and (b) curling, (c) coherent rotation, (d) perturbed curling, and (e) low-lying noncurling mode. (Sui et al. 2004) [109].
9.5 Magnetic Nanotubes
263
exchange and magnetostatics. The Rrand can be expressed by applying standard random anisotropy analysis: R rand ª δ B2 t 1 / 2 / a 3 / 2
(9.6)
where a is the polycrystalline grain size, and δB, the Bloch-wall thickness of the corresponding bulk material, is equal to (A/|K1|)1/2. With estimated values of a, t, and δB, the value of Rrand can be obtained for different magnetic nanotubes, which shows if the nanotube is curling-like or not. It needs to pointed out that derivation of this equation does not include magnetostatic self-interaction. When this is important, the random-anisotropy analysis shows that internal poles, such as around the dashed lines in Figure 9.14(e), enhance Rrand. 9.5.2
Multifunctionality of Magnetic Nanotubes
Magnetic nanotubes have the capability of multifunctionality due to their distinct inner and outer surfaces. For nanospheres/nanoparticles, there is only one surface available for modification; hence it is difficult to generate multifunctionalities [30]. By contrast, nanotubes with different functionalizations on both surfaces could be a type of multifunctional nanomaterial for several research problems. Magnetic nanomaterials have been extensively explored for their extraordinary properties as well as their potential applications, especially in biomedical and biotechnological applications. In most applications, magnetic nanoparticles have been used, mainly because of the ease of their synthesis and dimensional control. However, the structural limitation of spherical nanoparticles causes problems when multifunctionality is needed for certain bioapplications. Magnetic nanotubes possessing a number of attributes are potential candidates for certain bioapplications under such situations. There are two typical aspects regarding to the multifunctionality of magnetic nanotubes. First, nanotubes have inner voids, which can be filled and immobilized with biological species ranging in size from large proteins to small molecules [82]. Second, the distinct inner and outer surfaces of nanotubes render them a novel type of nanomaterial with multifunctional properties after being functionalized differently. Functionalization of multifunctional magnetic nanotubes is illustrated in Figure 9.15. For instance, the inner surface of nanotubes can be chemically functionalized to obtain a hydrophilic surface while the outer surface can be modified to exhibit hydrophobic properties. 9.5.3
Synthesis of Magnetic Nanotubes
To date, a number of chemical and physical methods have been used for magnetic nanotube synthesis, such as thermal decomposition of precursors, hydrothermal synthesis, galvanic displacement reactions, and so forth [111]. Without any exception, every technique employed in magnetic nanotube synthesis aims to obtained magnetic nanotubes with controlled elemental composition, dimension, morphology, crystallinity, and tailored magnetic properties. Among various approaches, template-assisted synthesis is a versatile and inexpensive technique [4]. Martin’s group first demonstrated template synthesis of
264
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
Different functional groups
Different binding ligands
Figure 9.15 Functionalization of a magnetic nanotube for which L and L‘ represent ligands that bind selectively to inner and outer surfaces, and R and R‘ represent spatially separated functional groups.
various nanotubes using membranes with aligned nanochannels [52]. By varying the physical and chemical properties of the template used and controlling the preparation conditions, the size, shape, and structural properties of nanotubes can be tailored. Importantly, this method is a general synthesis strategy that can be used to prepare metallic, polymeric, semiconducting, and carbon nanotubes [113]. A template-assisted synthesis of ferromagnetic Fe3O4 nanotubes was developed [109]. FePt nanotubes were also synthesized by a similar approach. In the synthesis, commercially available AAO membranes, as the template, were first treated by thermal annealing at 600°C. After cooling down, they were wetted with alcohol. Meanwhile, Fe(NO3)3⋅9H2O in ethanol solution was prepared and filtered through the membranes. In this step, the solution passed through the nanochannels inside the AAO membrane and some substances in the solution bound to the inner wall of the nanochannels. In the next step, the solution-loaded AAO membranes were placed in an oven vertically and underwent a thermal decomposition of Fe(NO3)3 at 250°C, giving iron oxide solid materials. Another thermal reduction was conducted by introducing hydrogen flow into the oven at the same temperature for two and half hours. The final step involved an etching process to remove the AAO template by 0.3 M NaOH aqueous solution. A bundle of released Fe3O4 nanotubes are shown in Figure 9.16(a), a TEM image. Hollow structure and uniform nanotube diameter were observed. The XRD pattern of nanotubes reveals a cubic crystal structure of Fe3O4, as indicated in Figure 9.16(b). Figure 9.16(c) shows the hysteresis loops of nanotubes measured at room temperature. Measurements along the parallel and perpendicular directions, respective to the magnetic field, resulted in pronounced difference. Synthesis of single crystalline Fe3O4 nanotubes was reported by Liu et al. (2004) [73]. Their synthesis strategy is coating Fe3O4 on the surface of MgO to form MgO/Fe3O4 core-shell nanowires followed by etching the inner cores. The three-step process is depicted in Figure 9.17(a). The merit of this synthesis is that it is able to control the length, diameter, and wall thickness of the homogeneous magnetite nanotubes. Single crystalline MgO nanowires were grown on Si/SiO2 substrates.
9.5 Magnetic Nanotubes
265
Figure 9.16 (a) TEM image of magnetite (Fe3O4) nanotubes; (b) X-ray diffraction pattern of as-prepared Fe3O4 nanotubes; and (c) hysteresis loops of the Fe3O4 nanotubes under the external fields applied in parallel and perpendicular to the nanotubes. (Sui et al. 2004) [109].
Then a pulsed laser deposition (PLD) technique was applied to deposit a conformal layer of Fe3O4 on the nanowire surfaces. The final step included a selective etching of the MgO inner cores of the MgO/Fe3O4 core-shell nanowires at an elevated temperature. As shown in Figure 9.17(b), a typical TEM image, the nanotubule structure is verified by the phase contrast between the tube wall and the inside hollow region. A Fe3O4 nanotube is very straight with a smooth and uniform sidewall along the whole length. The TEM image also shows an open end of the nanotube from which the etchant entered the nanotube for the etching purpose. The high-resolution TEM image of the sidewall of the nanotube is shown in the lower inset. The top inset of Figure 9.17(b) indicates a single-crystal nature of the magnetite nanotubes. 9.5.4
Biomedical Applications
Magnetic nanotubes inherit the paramagnetic nature of magnetic nanoparticles and they even display much higher saturation magnetization than their bulk counterparts. Generally speaking, magnetic nanotubes could be used in almost all applications for magnetic nanoparticles and nanowires. Additionally, the higher surface area, lower density, and multifunctionality of magnetic nanotubes render them a promising candidate for many other applications in special situations. Magnetic nanotubes may have possible applications in ultrahigh-density magnetic storage devices, nanoelectromechanical system (NEMS), sensors, catalysts, and so forth [71]. Recently, magnetic nanotubes have been envisioned to be used in biomedical
266
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
Figure 9.17 (a) Schematic of the Fe3O4 nanotubes fabrication process; and (b) low-magnification TEM image of a single Fe3O4 nanotube, with the SEAD shown in the upper inset and a high-resolution image showing the multilayer structure in the lower inset. (Liu et al. 2004) [73].
and biological fields [109]. Major bioapplications of magnetic nanotubes include bioseparation, cell manipulation, targeted drug or gene delivery, neuron regulation, and tips for magnetic force microscopes. Magnetic beads have been widely used in cell and cell membrane manipulations. Magnetic nanotubes could be a better material candidate for enhanced cell manipulation because of their high surface area. It is expected that magnetic nanotubes can be used as the magnetic handle to manipulate cells efficiently by an applied magnetic field. To prove this concept, work was conducted by Gao and others (2006) demonstrating the manipulation of sheep red blood cells using magnetic carbon nanotubes [110]. For magnetic drug delivery, unlike magnetic nanoparticles, magnetic nanotubes have an inner surface and outer surface for drug incorporation and drug immobilization, respectively. Most importantly, magnetic nanotubes have the capability to enhance biointeractions between the outer surface of nanotubes and the target biospecies. For this reason, using magnetic nanotubes could enhance drug delivery efficiency. A proof-of-concept experiment was conducted to demonstrate the enhanced drug delivery performance [106]. The inner surface of magnetic nanotubes 60 nm in diameter was functionalized while their outer surface was modified and rabbit IgG antibodies attached. These multifunctional magnetic nanotubes were added to an antirabbit IgG-modified glass slide and incubation was performed with an applied magnetic field from the bottom of the glass slide. To check the effect of the magnetic field for biointeraction, a fluorescence microscope was used to image the number of bound nanotubes after washing the unbound ones. Figure 9.18(a) and (b) are fluorescence microscopy images derived from averaging
9.6 Magnetic Biosensors
(a)
267
(b)
(c)100
5-FU 4-NO2-Ph ibuprofen
50
0
20
40 60 Time (h)
80
Figure 9.18 (Color plate 14) Fluorescence microscope images showing the binding between magnetic nanotubes and the surface of anti-rabbit IgG-modified glass after antigen-antibody interaction with (a) and without (b) a magnetic field under the glass substrate. (c) comparison of the in vitro release of ibuprofen, 4-nitrophenol (4-NO2Ph), and 5-fluorouracil (5-FU) from magnetic nanotubes. (Son et al. 2005) [106].
at five different areas on the glass slide. The comparison between the two images indicated that the antibody-antigen interaction was enhanced about 4.2-fold with an applied magnetic field. Thus the efficiency of the biointeraction can be controlled spatially by means of an external field. The performance of magnetic nanotubes in drug loading and drug release was tested, as shown in Figure 9.18(c). Application of magnetic nanotubes in neural science and technology has stemmed from the carbon nanotubes’ substratum behavior for neuron growth. Carbon nanotubes have been demonstrated as a promising substrate for neuron growth, as they boosted neurite growth and neuronal electrical signaling [51, 78]. Investigation was conducted to explore the effect of magnetic nanotubes in PC12 cell differentiation and neurite growth [123]. NGF-incorporated magnetic nanotubes were present in the PC12 cell culture. Two goals for this research are (1) to test the accessibility of NGF from magnetic nanotubes in PC12 cell differentiation into neurons, and (2) to study the neurite growth possibly regulated by magnetic nanotubes. Both differentiation of PC12 cells and the neuritic process growth in the presence of functional magnetic nanotubes with incorporated NGF were observed. Figure 9.19(a) reveals a random distribution of nanotubes on the glass substrate and an appreciable neurite outgrowth. Major components of a growing neuron including soma, neurites, and a growth cone can be discerned, suggesting that the functional magnetic nanotubes are not toxic for neuron survival and neither do they impair neurite outgrowth. It also confirms that the bioactive nanotube-bound NGF was available to the PC12 cells to induce neuronal differentiation. The enlarged SEM image (inset of Figure 9.19(a) shows the growth cone area, located on the tip of the axon, where slender extensions, the filopodia, were formed towards the nearby magnetic nanotubes. Figure 9.19(b) shows a nanotube split longitudinally contacted by the tip of a growing neurite, a filopodium, and it seems the neuritic process growing into the nanotube.
9.6
Magnetic Biosensors 9.6.1
Typical Magnetic Biosensing Schemes
Biosensors utilize biological reactions for detecting target analytes. A biosensor mainly consists of a biological recognition element (or bioreceptor) and a signal
268
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
Figure 9.19 (a) SEM micrograph of a typical PC12 cell after culture showing the cell body and neurites (the inset is a magnified image of the growth cone area); (b) high-resolution SEM micrograph of a NGF-incorporated magnetic nanotube and its interaction with a filopodium. (Xie et al. 2008) [123].
transducer [97, 114, 119]. When an analyte interacts with the bioreceptor, the resulting complex produces a change that can be converted into a measurable effect by the transducer. Common types of bioreceptor/analyte complexes are based on antibody/antigen interactions, nucleic acid interactions, enzymatic interactions, cellular interactions, or the interactions using biomimetic materials. The most prevalent signal transduction methods include optical measurements, electrochemical, and mass-sensitive measurements [116]. For practical applications, biosensors should have high sensitivity, small size, low power-consumption, stability of operating parameters, quick response, resistance to aggressive medium, and low price [64]. Two sensing schemes are often used for detecting magnetic particles. A solid-state based sensor detects the stray magnetic fields of the magnetic particles bound to the biomolecule-functionalized surface of the sensor, while a substrate-free sensor is based on the change in the Brownian relaxation time of the particles suspended in liquids upon their binding to the target molecules [81]. 9.6.1.1
Solid-State Based Sensors
In a solid-state based sensor, the analyte to be detected is usually attached to a substrate, and detects the stray magnetic fields of magnetic nanoparticles or magnetic microbeads tagged to targets [65]. The idea of using a magnetic field sensor in combination with magnetic particles working as magnetic labels for detecting molecular recognition events was first reported in [5]. Magnetic labels have obvious advantages in biosensing [24, 65, 66, 97]. First, the size of magnetic nanoparticles can range from a few nanometers to several micrometers and thus is compatible with biological entities ranging from proteins (a few nm) to cells and bacteria (several μm). By coating the magnetic nanoparticles with specific ligands, the nanoparticles can selectively bind to a target material of interest. Second, the magnetic properties of nanoparticles are very stable, and they are not affected by reagent chemistry or subject to photo-bleaching. Third, magnetic fields are not screened by aqueous reagents or biomaterials. Since most biological systems do not exhibit ferromagnetism and typically have only small magnetic susceptibilities revealing dia- or para-
9.6 Magnetic Biosensors
269
magnetism, the magnetic moment from a magnetic nanoparticle can be detected with little noise from the biological environment. Fourth, magnetic nanoparticles can be remotely manipulated using field gradients even when they are embedded in a biological environment. Finally, there are many highly sensitive magnetic field detection devices that have been developed and that could be used for biosensing applications, such as giant magnetoresistive (GMR) sensors, anisotropic magnetoresistive (AMR) sensors, inductive sensors, miniature Hall crosses, and superconducting quantum interference devices (SQUIDs). In order to achieve single molecule detection, the dimension of magnetic particle labels should be comparable to that of biomolecules [72]. For example, in detection of DNA fragments, it is ideal to have the particle labels at 20 nm or smaller in diameter. In this way, one nanoparticle label may be conjugated with one or at most a few DNA fragments, and this is important for establishing a quantitative relationship with sufficient accuracy between the number of captured particle labels and the actual biorecognition events. The detection schemes of solid-state based biosensors can be generally classified into two categories: the direct detection scheme and secondary detection scheme. As shown in Figure 9.20(a), in a direct detection scheme, an array of probe molecules is immobilized over a magnetic field sensor, and the biomolecules (target analytes) to be detected are magnetically labeled and pass over the probe array. The target analytes are bound to the array, and the unbound biomolecules are then washed away [43]. In this manner the target material acts as a link between the nanoparticles and the magnetic field sensor, and the presence of the target is confirmed by detecting the stray fields from the nanoparticles [24]. The secondary detection scheme is based on a secondary detection step, performed after the interrogation of the probe array with the target molecules [43]. As shown in Figure 9.20(b), this scheme consists of two steps. In the first step, a small biochemical label, such as biotin, is tagged to the target molecules, and the tagged
(a)
(b)
Figure 9.20 Schematic illustrations two detection schemes. (a) Direct detection scheme, and (b) secondary detection scheme (Graham et al. 2004) [43].
270
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
target molecules are bound to complementary probe molecules. In the second step, magnetic labels functionalized with streptavidin, which is complementary to biotin, are introduced, and the magnetic fields produced by the magnetic labels are detected. 9.6.1.2
Substrate-Free Sensors
In a substrate-free sensor, the target analytes labeled by magnetic particles are not immobilized to a substrate; instead, they are suspended in liquid, forming ferrofluids. The detection of the magnetic particles, or the target analytes, is deduced from the properties of the ferrofluids, such as relaxation time and magnetic susceptibility. This scheme offers an excellent method for distinguishing between multiple potential target molecules in the solution based on their size difference, but is limited in terms of the size of the magnetic particles that can be used for detection and also requires relatively high concentration of magnetic particles for a measurable signal. In the following sections, we discuss two kinds of solid-state based sensors: magnetoresistance-based sensors and Hall-effect sensors, and two types of substrate-free sensors: sensors detecting magnetic relaxations and sensors detecting ferrofluid susceptibility. 9.6.2
Magnetoresistance-Based Sensors
Magnetoresistance (MR)-based sensors provide a highly sensitive sensing technology with wide dynamic range [43]. MR sensors mainly include giant magnetoresistance (GMR) sensors, anisotropic magnetoresistance ring (AMR) sensors, spin valves, and magnetic tunnel junction sensors. GMR biosensors are promising for sensitive, large-scale, inexpensive, and portable biomolecular identification [72]. A basic GMR structure mainly consists of a pair of magnetic thin films separated by a non-magnetic conducting layer. When the magnetizations of the magnetic layers are aligned by an external magnetic field, the electrical resistance of the structure decreases due to the reduction of spin-dependent electron scattering within the structure. A GMR sensor can be of microscopic size, and can sensitively detect the presence of a magnetic particle with micron or smaller size in close proximity [97]. Rife et al. (2003) developed a biosensor system, the Bead ARray Counter (BARC) [97]. As illustrated in Figure 9.21, a BARC sensor is made from a multilayer GMR film with a large saturation field and GMR effect. The GMR film consists of four ferromagnetic layers separated by three non-ferromagnetic layers. Each ferromagnetic layer has three sublayers, consisting of a layer of NiFeCo, sandwiched between two thin films of CoFe. The thicknesses of the films are optimized to ensure antiparallel exchange coupling across the CuAgAu layers and meanwhile maintaining the high sensitivity and linearity. Due to the shape anisotropy, the magnetization of each GMR trace naturally lies in the plane of the film, therefore only the planar components of the induced microbead field cause appreciable magnetoresistance [97]. To detect magnetic microbeads using a BARC system, an ac magnetic field, H0z, is applied normal to the chip (the z direction). As schematically shown in Figure 9.21(a), when a single bead resting above the GMR sensor is magnetized by an
9.6 Magnetic Biosensors
271
Figure 9.21 Illustration of a BARC sensor. (a) The arrangement of a bead and a sensor. (b) The multilayer structure of the sensor. (c) Detailed structure of the GMR stack (Rife et al. 2003) [97].
external magnetic field, it will generate a local dipole field, B, whose planar components are strong enough to cause a magnetoresistance effect. The overall GMR signal, ΔR/R, is mainly determined by the sensor geometry and the cumulative local magnetoresistance changes associated with individual microbeads. Usually a Wheatstone bridge is used in the measurement of the GRM signal, ΔR/R. As shown in Figure 9.22, the changes of resistance due to individual beads are independent, and additive until a saturation level is achieved. 9.6.3
Hall-Effect Sensors
Various types of Hall effect sensors have been developed for the detection of magnetic nanoparticles based on different Hall effects. The ordinary Hall effect is due to the Lorentz force acting on charge carriers in metals, semi-metals and semiconduc-
Figure 9.22 The relationship between the Wheatstone bridge signal and the number of beads deposited onto a BARC sensor. The open circles, filled circles, and filled triangles, indicate the measurement data of three different chips. The dashed line is a linear fit to the triangles corresponding to 15 nV per bead. The inset is an optical micrograph a sensor with 14 beads (Rife et al. 2003) [97].
272
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
tors. Magnetic materials show additional Hall phenomena generated by spin-orbit interactions: extraordinary, and planar Hall effects. Some heterostructures show quantum-well Hall effect. Silicon Hall-effect sensors are among the most popular Hall-effect sensors. A single magnetic microbead can be detected and characterized using a silicon Hall sensor fabricated with standard metal-oxide-semiconductor (CMOS) technology [15]. As shown in Figure 9.23, a single microbead is manually placed in the center of a cross-shaped silicon Hall sensor. To determine the magnetic characteristics of the microbead, the Hall sensor and the bead are placed in a static field H0 in the z direction. A small ac field H2 is applied parallel to H0 at frequency f0 to measure the small signal behavior of the magnetization. The Hall voltage VH at frequency f0 is measured using a low-noise preamplifier and lock-in detector. The voltage VH measures the magnetic induction at the location of the sensor as a function of the externally applied magnetic field. Since the magnetization of the bead saturates, the Hall voltage at high magnetic field is used to normalize the measured values. Here we define the apparent susceptibility χapp by [15]: χ app (H 0 ) + 1 ∫
VH (H 0 ) VH (H 0 Æ •)
(9.7)
The value of χapp depends on the geometry of the bead and on the distance between the bead and the sensor. It tends to vanish at high fields. Figure 9.24 shows the measured and modeled apparent susceptibility of a single bead. 9.6.4
Sensors Detecting Magnetic Relaxations
Magnetorelaxometry (MRX) is a powerful analytical tool for the specific detection of biological molecules, such as proteins, bacteria, or viruses. The basic idea of MRX is that the moments of magnetic nanoparticles are aligned by a magnetic field and then, after switching off the field, the decay of the net magnetic signal as a func-
(a)
(b)
Figure 9.23 Schematic view of the measurement setup. A single superparamagnetic bead is centered on a Hall sensor, biased with the current Ib. A dc magnetic field H0 is applied perpendicular to the sensor. An ac field is added either in the z direction (H2) or in the x direction (H1). The Hall voltage VH, containing a component proportional to the magnetic induction produced by the bead, is measured with a lock-in amplifier. (Besse et al. 2002) [15].
9.6 Magnetic Biosensors
273
Figure 9.24 Measurement of the apparent susceptibility. (Ib=0.3 mA, H2=0.18 kA/m, τ=2 s, Δf=0.06 Hz, and f0=520 Hz.) (Besse et al. 2002) [15].
tion of time is analyzed [75]. Mobile magnetic nanoparticles relax via the Brownian mechanism on a time scale of microseconds, whereas nanoparticles that are immobilized, for example, by binding to a large biomolecule, relax via the Néel mechanism. Therefore, bound and unbound magnetic nanoparticles can be distinguished by their different relaxation times and time dependencies. Measurement times typically amount to a few seconds. Usually MRX is performed using superconducting quantum interference devices (SQUIDs) known to be the most sensitive solid-state magnetic field sensors. Grossman et al. (2004) developed a technique for the detection of magnetically labeled Listeria monocytogenes and for the measurement of the binding rate between antibody-linked magnetic particles and bacteria [44]. Using this technique to quantify specific bacteria, the bacteria do not need to be immobilized, and the unbound magnetic particles do not need to be washed away. In the measurement, a pulsed magnetic field is applied to align the magnetic moments, and when the pulsed magnetic field is turned off, a SQUID is used to detect the magnetic relaxation signal. Brownian rotation of unbound particles is too quick to be detected. On the contrary, the particles bound to L. monocytogenes relax in about one second by rotation of the internal dipole moment. Such a Néel relaxation process can be detected by SQUID, and the binding rate between the particles and bacteria can be obtained by time-resolved measurements. Figure 9.25 shows the measurement configuration and the basic structure of the SQUID. The voltage across the current-biased SQUID oscillates quasi-sinusoidally as a function of the magnetic flux Φ threading the loop with a period of the magnetic flux quantum, Φ0 = h/2e ≈ 2 × 10–15 T⋅m–2. To linearize the flux-to-voltage conversion, the SQUID is operated in a flux-locked loop that maintains the flux through it at a constant value; the output voltage of this feedback circuit is proportional to Φ. Figure 9.26 shows typical time traces for an L. monocytogenes sample and associated controls. These data were fit to a sum of logarithmic and exponential functions. The logarithmic decay is characteristic of Neel relaxation for particles with a wide distribution of sizes, and therefore of relaxation times. The exponential decay comes from particle aggregates, which are large enough to relax via Brownian rota-
274
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
(a)
(b)
Figure 9.25 (a) Top portion of the SQUID microscope. The SQUID, inside a vacuum enclosure, is mounted on a sapphire rod thermally connected to a liquid nitrogen reservoir (not shown). A 75-μm-thick sapphire window separates the vacuum chamber from atmosphere. The sample is contained in a Lucite holder, with a 3-μm-thick Mylar base, aligned against a positioning element. (b) Configuration of the YBCO SQUID. The slit is 4-μm wide. (Grossman et al. 2004) [44].
Figure 9.26 Example of magnetic decay signals. For the traces shown, the concentration of bacteria was 108 per ml, and the concentration of particles was 0.13 relative to the stock suspension. A 0.4-mT field was pulsed on for 1s and off for 1s, and data was recorded each time when the field was turned off; 100 averages were taken. (Grossman et al. 2004) [44].
tion on a measurable timescale without being bound to targets. The fitting function is:
(
)
(
F(t ) = F offset + F s ln 1 + τ mag / t + F exp exp -t / τ exp
)
(9.8)
where Φoffset is an offset caused by the fact that the SQUID measures relative, rather than absolute, magnetic flux; Φs, the logarithmic decay amplitude, is proportional to the number of bound particles; τmag = 1 s is the magnetization time; Φexp, the exponential decay amplitude, depends on the number of unbound particle aggregates; and τexp is the exponential decay time constant. Fitting (9.8) to the measurement results gives τexp≈ 15 ms, corresponding to a hydrodynamic diameter of ≈340 nm for a sphere [44].
9.6 Magnetic Biosensors
9.6.5
275
Sensors Detecting Ferrofluid Susceptibility
Biomolecules could be detected by using ac magnetic susceptibility measurements to monitor the binding-induced modification of Brownian relaxation of magnetic nanoparticles suspended in liquids. This method is based on the fact that binding with biomolecules increases the hydrodynamic radii of the microbeads, and thus the frequency-dependence of the magnetic susceptibility of the fluid is shifted. This substrate-free detection scheme has several advantages [24, 25]: (1) it generates a useful signal in both the presence and absence of the target, thus providing an inherent check for integrity, (2) it permits discrimination between several potential targets, since beside the binding affinity additional information about the target size can be obtained, and (3) it can generate quantitative information about the target concentration. Kriz et al. (1996) developed a transducer concept in biosensors, utilizing measurements of magnetic permeability [63]. As the magnetic permeability of a material inside a coil influences the inductance of the coil, it is possible to detect changes in magnetic permeability using inductance measurements. The inductance L for a coil with a magnetic material inside is described by [63]: L = ( μ r μ 0 A / l)N 2
(9.9)
where μr is the relative magnetic permeability of the material in the coil, μ0 is the permeability of a vacuum, A is the cross section area, l is the coil length, and N is the turn number on the coil. To measure the inductance, and thus indirectly the relative magnetic permeability, the coil can be placed in a Maxwell bridge [63]. As shown in Figure 9.27, the transducer measures the changes in the magnetic permeability of materials and comprises a coil that is a part of a balanced Maxwell bridge, with two variable resistances. The voltage difference measured over the Maxwell bridge is further processed by a differential operational amplifier circuit and rectified. The introduction of ferromagnetic materials inside the coil causes an increase in the voltage difference over the Maxwell bridge.
Figure 9.27 The measurement system. Measuring coil L4 (transducer) is part of a balanced Maxwell bridge. A sinusoidal wave is fed into the bridge. The voltage difference measured over the bridge is further processed by a differential operational amplifier circuit, rectified, and finally recorded. (Kriz et al. 1996) [63].
276
Magnetic Nanomaterials, Nanotubes, and Nanomedicine
Figure 9.28 shows the response obtained from the Maxwell bridge as a function of the initial concentration of the ferromagnetic model analyte (dextran ferrofluid) in a sample solution. As expected, saturation is achieved for higher concentrations. To investigate whether the observed increase in response was due to the specific interaction between Con A Sepharose and dextran ferrofluid, the Con A Sepharose was exchanged for Sepharose. In this case, no response could be observed, as shown by the lower curve in Figure 9.28.
9.7
Magnetic Biochips A biochip is generally defined as a material or a device that has an array of probes used for biochemical assays. In general, any device or component incorporating a two-dimensional array of reaction sites and having biological materials on a solid substrate is referred to as a biochip. Research on biochips involves both miniaturization, usually in microarray formats, and the possibility of low-cost mass production. Magnetic biochips are discussed in Chapter 9 of the Supplementary Materials and Solution Manual.
9.8
Prospects There are two major trends in the research of magnetic nanomaterials and nanomedicine. One is the development of magnetic nanoparticles that can perform more than one function. For example, it is favorable that a multifunctional nanoparticle for cancer therapy has a cancer cell targeting component guiding the nanoparticle to reach the cancer cells, a drug delivery indicator controlling and reporting the drug delivery, a cell death sensor reporting whether the cancer cell has
Figure 9.28 Response obtained from a Maxwell bridge as a function of various initial concentrations of dextran ferrofluid incubated with Con A Sepharose (upper curve) and with Sepharose (lower curve). (Kriz et al. 1996) [63].
Problems
277
been killed, and a contrast agent checking the therapeutic effects. However, it should be noted that there is a long way to go to develop such multifunctional nanoparticles, and now it is still in an initial stage as viewed by National Institute of Health and National Cancer Institute. The other major trend is the development of magnetoelectronic tools that can precisely detect and manipulate individual cells and biomolecules. The development of such magnetoelectronic tools is based on microfluidics. In technical literature, many words have been coined to describe such devices: lab-on-a-chip device, micro total analysis system (μTAS), miniaturized analysis system, microfluidic system, and nanofluidic system. Such a device is a combination and integration of fluidic, sensor, and detection elements to perform a complete sequence of chemical reactions or analyses, including sample preparation, mixing, reaction, separation, and detection. This technology can be regarded as an interface between the nano-world and the macro-world [1, 103]. For the responsible development of nanomedicine, including magnetic nanomedicine, it is necessary to discuss the related ethical and social issues [21, 100]. As with other biotechnological advances before it, some of the important ethical and social concerns initially focus on risk assessment and environmental management. As it moves from proof-of-concept to clinical trials, and further to clinics, nanomedicine will be facing significant challenges. Nanomedicine may launch a new epoch in healthcare where pharmaceuticals will be more effective and less toxic, where disease monitoring can be done on a highly sensitive and specific level, and where injections, surgical procedures, and a host of other interventions will be made less painful, less toxic, and with fewer side effects than their current counterparts. However, such advances should not come at the expense of fairness, safety, or basic understanding of what it means to be a healthy human being [8]. The future of nanomedicine practice depends on whether such ethical and social issues can be satisfactorily addressed.
Problems 9.1 Describe the superparamagnetism principle. Why is it related to nanoparticles? 9.2 What is hyperthermia? How can magnetic nanoparticles transfer energy to the body? What is the usefulness of this?
References [1] Alper, J. (2005). Chips ahoy - the lab-on-a-chip revolution is near, Monthly Feather, NCI Alliance for Nanotechnology in Cancer, April 2005. [2] Baibich, M. N., Broto, J. M., Fert, A., van Dau, F. N., Petroff, F., Eitenne, P., Creuzet, G., Friederich, A. and Chazelas, J. (1988). Giant magnetoresistance of (001) Fe / (001) Cr magnetic superlattices, Physical Review Letters, 61, 2472-2475. [3] Bao, J., Xu, Z., Hong, J., Ma, X. and Lu, Z. (2004). Fabrication of cobalt nanostructures with different shapes in alumina template, Scripta Materialia, 50, 19-23.
278
Magnetic Nanomaterials, Nanotubes, and Nanomedicine [4] Bao, J., Tie, C., Xu, Z., Zhou, Q., Shen, D. and Ma, Q. (2001). Template synthesis of an array of nickel nanotubules and its magnetic behavior, Advanced Materials, 13, 1631-1633. [5] Baselt, D. R., Lee, G. U., Natesan, M., Metzger, S. W., Sheehan, P. E. and Colton, R. (1998). A biosensor based on magnetoresistance technology, Biosensors and Bioelectronics, 13, 731-739. [6] Bauer, L. A., Birenbaum, N. S. and Meyer, G. J. (2004). Biological applications of high aspect ratio nanoparticles, Journal of Materials Chemistry, 14, 517-526. [7] Bauer, L. A., Reich, D. H. and Meyer, G. J. (2003). Selective functionalization of two-component magnetic nanowires, Langmuir, 19, 7043-7048 [8] Bawa, R. and Johnson, S. (2007). The ethical dimensions of nanomedicine, The Medical Clinics of North America, 91, 881-887. [9] Bawa, R., Bawa, S. R., Maebius, S., et al. (2006). Bionanotechnology patents: challenges and opportunities, in Bronzino, J. D. (ed.). The Biomedical Engineering Handbook. 3rd edition. Boca Raton (FL): CRC Press. [10] Bawa, R. (2005). Will the nanomedicine “patent land grab” thwart commercialization?, Nanomedicine, 1, 346-350. [11] Bawa, R. and Bawa, S. R. (2005). Protecting new inventions in nanomedicine, in Wagner, C. G. (ed.). Foresight, Innovation, and Strategy: Towards a Wiser Future, World Future Society, Bethesda. [12] Bawa, R., Bawa, S. R., Maebius, S. B., et al. (2005). Protecting new ideas and inventions in nanomedicine via patents, Nanomedicine, 1 (2), 150-158. [13] Bean, C. P. and Jacobs, I. S. (1956). Magnetic granulometry and super-paramagnetism, Journal of Applied Physics, 27, 1448-1452. [14] Bera, D., Kuiry, S. C. and Seal, S. (2004). Synthesis of nanostructured materials using template-assisted electrodeposition, JOM, 56 (1), 49-53. [15] Besse, P. A., Boero, G., Demierre, M., Pott, V. and Popovic, R. (2002). Detection of a single magnetic microbead using a miniaturized silicon Hall sensor, Applied Physics Letters, 80 (22), 4199-4201. [16] Bhatia, S. N. and Chen, C. S. (1999). Tissue engineering at the microscale, Biomedical Microdevices, 2, 131–144. [17] Binasch, G., Grunberg, P., Saurenbach, F. and Zinn, W. (1989). Enhanced magnetoresistance in layered magnetic structures with antiferromagnetic interlayer exchange, Physical Review B, 39, 4828-4830. [18] Blondel, A., Meier, J. P., Doudin, B. and Ansermet, J. P. (1994). Giant magnetoresistance of nanowires of multilayers, Applied Physics Letters, 65, 3019-3021. [19] Bossis, G., Volkova, O., Lacis, S. and Meunier, A. (2002). Magnetorheology: fluids, structures and rheology, Lecture Notes in Physics, 594, 202-230. [20] Brown, W. F. Jr. (1963). Thermal fluctuations of a single-domain particle, Physical Review, 130, 1677–1686. [21] Bruce, D. (2006). Ethical and social issues in nanobiotechnologies, EMBO Reports, 7 (8), 754-758. [22] Cardoso, F. A., Ferreira, H. A., Conde, J. P., Chu, V., Freitas, P. P., Vidal, D., Germano, J., Sousa, L., Piedade, M. S., Costa, B. A., Lemos, J. M. (2006). Diode/magnetic tunnel junction cell for fully scalable matrix-based biochip, Journal of Applied Physics, 99, 08B307. [23] Chen, M., Sun, L., Bonevich, J. E., Reich, D. H., Chien, C. L. and Searson, P. C. (2003). Tuning the response of magnetic suspensions, Applied Physics Letters, 82 (19), 3310-3312. [24] Chung, S. H., Hoffmann, A., Guslienko, K., Bader S. D., Liu, C., Kay, B., Makowski, L. and Chen, L. (2005). Biological sensing with magnetic nanoparticles using Brownian relaxation (invited), Journal of Applied Physics, 97, 10R101.
References
279
[25] Connolly, J. and St Pierre T. G. (2001). Proposed biosensors based on time-dependent properties of magnetic fluids, Journal of Magnetism and Magnetic Materials, 225, 156-160. [26] Cowburn, R. P. (2000). Property variation with shape in magnetic nanoelements, Journal of Physics D: Applied Physics, 33, R1–R16. [27] Darling, S. B. and Bader, S. D. (2005). A materials chemistry perspective on nanomagnetism, Journal of Materials Chemistry, 15, 4189–4195. [28] Donegá, C. de M., Liljeroth, P. and Vanmaekelbergh, D. (2005). Physicochemical evaluation of the hot-injection method, a synthesis route for monodisperse nanocrystals, Small, 1, 1152-1162. [29] Edelstein, R. L., Tamanaha, C. R., Sheehan, P. E., Miller, M. M., Baselt, D. R., Whitman, L. J. and Colton, R. J. (2000). The BARC biosensor applied to the detection of biological warfare agents, Biosensors & Bioelectronics, 14, 805–813. [30] Eisenstein, M. (2005). An attractive alternative, Nature Methods, 2, 484. [31] Encinas, A., Demand, M., Vila, L., Huynen, I. and Piraux, L. (2002). Tunable remanent state resonance frequency in arrays of magnetic nanowires, Applied Physics Letters, 81, 2032-2034. [32] Escrig, J., Landeros, P., Altbir, D., Vogel, E. E. and Vargas, P. (2007). Phase diagrams of magnetic nanotubes, Journal of Magnetism and Magnetic Materials, 308, 233-237. [33] European Technology Platform, (2006), Nanomedicine: Nanotechnology for Health, Office for Official Publications of the European Communities, Belgium. [34] Fert, A. and Piraux, L. (1999). Magnetic nanowires, Journal of Magnetism and Magnetic Materials, 200, 338-358. [35] Flynn, T. and Wei, C. (2005). The pathway to commercialization for nanomedicine, Nanomedicine: Nanotechnology, Biology, and Medicine, 1, 47-51. [36] Freitas, R. A. (2005). What is nanomedicine?, Nanomedicine: Nanotechnology, Biology, and Medicine, 1, 2-9. [37] Freitas, R. A. (2003). Nanomedicine, Vol. IIA: Biocompatibility, Landes Bioscience, Georgetown. [38] Freitas, R. A. (1999). Nanomedicine, Vol. I: Basic Capabilities, Landes Bioscience, Georgetown. [39] Gao, C., Li, W., Morimoto, H., Nagaoka, Y. and Maekawa, T. (2006). Magnetic carbon nanotubes: synthesis by electrostatic self-assembly approach and application in biomanipulations, Journal of Physical Chemistry B, 110, 7213-7220. [40] Gijs, M. A. M. (2004). Magnetic bead handling on-chip: new opportunities for analytical applications, Microfluid and Nanofluid, 1, 22–40. [41] Gilchrist, R. K., Medal, R., Shorey, W. D., Hanselman, R. C., Parrott, J. C. and Taylor, C. B. (1957). Selective inductive heating of lymph nodes, Annals of Surgery, 146, 596-606. [42] Gordon, N. and Sagman, U. (2003). Nanomedicine Taxonomy, Canadian NanoBusiness Alliance, Toronto. [43] Graham, D. L., Ferreira, H. A. and Freitas, P. P. (2004). Magnetoresistive-based biosensors and biochips, Trends in Biotechnology, 22, 455-462. [44] Grossman, H. L., Myers, W. R., Vreeland, V. J., Bruehl, R., Alper, M. D., Bertozzi, C. R. and Clarke, J. (2004). Detection of bacteria in suspension by using a superconducting quantum interference device, Proceedings of the National Sciences of the United States of America, 101, 129-134. [45] Hafeli, U. (1997). Scientific and Clinical Applications of Magnetic Carriers, Plenum Press, New York. [46] Halbreich, A., Roger, J., Pons, J. N., Geldwerth, D., Da Silva, M. F., Roudier, M. and Bacri, J. C. (1998). Biomedical applications of maghemite ferrofluid, Biochimie, 80, 369-390.
280
Magnetic Nanomaterials, Nanotubes, and Nanomedicine [47] Han, S. J., Xu, L., Yu, H., Wilson, R.J., White, R.L., Pourmand, N. and Wang S.X. (2006). CMOS integrated DNA microarray sased on GMR sensors, IEEE International Electron Devices Meeting, 2006, 1–4. [48] Himpsel, F. J., Ortega, J. E., Mankey, G. J. and Willis, R. F. (1998). Magnetic nanostructures, Advances in Physics, 47 (4), 511-597. [49] Hirsch, R., Katz, E. and Willner, I. (2000). Magneto-switchable bioelectrocatalysis, Journal of American Chemical Society, 122, 12053-12054. [50] Hong, R., Fischer, N. O., Emrick, T. and Rotello, V. M. (2005). Surface PEGylation and ligand exchange chemistry of FePt nanoparticles for biological applications, Chemistry of Materials, 17, 4617-4621. [51] Hu, H., Ni, Y., Montana, V., Haddon, R. C. and Parpura, V. (2004). Chemically functionalized carbon nanotubes as substrates for neuronal growth, Nano Letters, 4, 507-511. [52] Hulteen, J. C. and Martin, C. R. (1997). A general template-based method for the preparation of nanomaterials, Journal of Materials Chemistry, 7, 1075-1087. [53] Hultgren, A., Tanase, M., Felton, E. J., Bhadriraju, K., Salem, A. K., Chen, C. S. and Reich, D. H. (2005). Optimization of yield in magnetic cell separations using nickel nanowires of different lengths, Biotechnology Progress, 21, 509-515. [54] Hultgren, A., Tanase, M., Chen, C. S. and Reich, D. H. (2004). High-yield cell separations using magnetic nanowires, IEEE Transactions on Magnetics, 40 (4), 2988-2900. [55] Hurst, M. J. Payne, E. K. Qin, L. and Mirkin, C. A. (2006). Multisegmented one-dimensional nanorods prepared by hard-template synthetic methods, Angewandte Chemie, International Edition, 45, 2672 –2692. [56] Hyeon, T. (2003). Chemical synthesis of magnetic nanoparticles, Chemical Communications, issue 8, 927-934. [57] Hyeon, T., Lee, S. S., Park, J., Chung, Y. and Na, H. B. (2001). Synthesis of highly crystalline and monodisperse maghemite nanocrystallites without a size-selection process, Journal of the American Chemical Society, 123, 12798-12801. [58] Johannsen, M., Gneveckow, U., Taymoorian, K. et al. (2007). Morbidity and quality of life during thermotherapy using magnetic nanoparticles in locally recurrent prostate cancer: Results of a prospective phase I trial, International Journal of Hyperthermia, 23 (3), 315-323. [59] Jotterand, F. (2007). Nanomedicine: how it could reshape clinical practice, Nanomedicine, 2 (4), 401-405. [60] Jun, Y., Huh, Y., Choi, J., Lee, J., Song, H., Kim, S., Yoon, S., Kim, K., Shin, J., Suh, J. and Cheon, J. (2005). Nanoscale size effect of magnetic nanocrystals and their utilization for cancer diagnosis via magnetic resonance imaging, Journal of the American Chemical Society, 127, 5732-5733. [61] Katz, E., Sheeney-Haj-Ichia, L., Bückmann, A. F. and Willner, I. (2002). Dual Biosensing by magneto-controlled bioelectrocatalysis, Angewandte Chemie International Edition, 41, 1343-1346. [62] Kelly, P. E., O’Grady, K., Mayo, P. I. Chantrell, R. W. (1989). Switching mechanism in cobalt-phosphorus thin-films, IEEE Transactions on Magnetics, 25, 3880-3883. [63] Kriz, C. B., Radevik, K. and Kriz, D. (1996). Magnetic permeability measurements in bioanalysis and biosensors, Analytical Chemistry, 68, 1966-1970. [64] Kurlyandskaya, G. and Levit, V. (2005). Magnetic Dynabeads® detection by sensitive element based on giant magnetoimpedance, Biosensors and Bioelectronics, 20, 1611-1616. [65] Landry, G., Miller, M. M., Bennett, B. R., Johnson, M. and Smolyaninova, V. (2004). Characterization of single magnetic particles with InAs quantum-well Hall devices, Applied Physics Letters, 85 (20), 4693-4695. [66] Lany, M., Boero, G. and Popovic, R. S. (2005). Superparamagnetic microbead inductive detector, Review of Scientific Instruments, 76, 084301.
References
281
[67] Lee, H., Liu, Y., Westervelt, R.M. and Ham, D. (2006). IC/microfluidic hybrid system for magnetic manipulation of biological cells, IEEE Journal of Solid-State Circuits, 41, 1471–1480. [68] Lee, K. B., Park, S. and Mirkin, C. A. (2004). Multicomponent magnetic nanorods for biomolecular separations, Angewandte Chemie, International Edition, 43, 3048-3050. [69] Leslie-Pelecky, D. L. and Rieke, R. D. (1996). Magnetic properties of nanostructured materials, Chemistry of Materials, 8, 1770-1783. [70] Levy, L., Sahoo, Y., Kim, K. S., Bergey, E. J. and Prasad, P. N. (2002). Nanochemistry: synthesis and characterization of multifunctional nanoclinics for biological applications, Chemistry of Materials, 14, 3715-3721. [71] Li, F. S., Zhou, D., Wang, T., Wang, Y., Song, L. J. and Xu, C. T. (2007). Fabrication and magnetic properties of FeCo alloy nanotube array, Journal of Applied Physics, 101, 014309-1-3. [72] Li, G. X., Sun, S. H., Wilson, R. J., White, R. L., Pourmand, N. and Wang, S. X. (2006). Spin valve sensors for ultrasensitive detection of superparamagnetic nanoparticles for biological applications, Sensors and Actuators A, 126, 98-106. [73] Liu, Z., Zhang, D., Han, S., Li, C., Lei, B., Lu, W., Fang, J. and Zhou, C. (2004). Single crystalline magnetite nanotubes, Journal of the American Chemical Society, 127, 6-7. [74] Lübbe, A. S., Alexiou, C. and Bergemann, C. (2001). Clinical applications of magnetic drug targeting, Journal of Surgical Research, 95, 200-206. [75] Ludwig, F., Mauselein, S., Heim, E. and Schilling, M. (2005). Magnetorelaxometry of magnetic nanoparticles in magnetically unshielded environment utilizing a differential fluxgate arrangement, Review of Scientific Instruments, 76, 106102. [76] Mallet, J., Yu-Zhang, K., Chien, C. L., Eagleton, T. S and Searson, P. C. (2004). Fabrication and magnetic properties of fcc CoXPt1-X nanowires, Applied Physics Letters, 84 (19), 3900-3902. [77] Martin, B. R., Dermody, D. J., Reiss, B. D., Fang, M., Lyon, L. A., Natan, M. J. and Mallouk, T. E. (1999). Orthogonal self-assembly on colloidal gold-platinum nanorods, Advanced Materials, 11 (12), 1021-1025. [78] Mattson, M. P., Haddon, R. C. and Rao, A. M. (2000). Molecular functionalization of carbon nanotubes and use as substrates for neuronal growth, Journal of Molecular Neuroscience, 14, 175-182. [79] Mayes, E., Bewick, A., Gleeson, D., Hoinville, J., Jones, R., Kasyutich, O., Nartowski, A, Warne, B., Wiggins, J. and Wong, K. K. W. (2003). Biologically derived nanomagnets in self-organized patterned media, IEEE Transactions on Magnetics, 39, 624-627. [80] Meiklejohn, W. H. and Bean, C. P. (1957). New magnetic anisotropy, Physical Review, 105 (3), 904-913. [81] Mihajlovic, G., Xiong, P., von Molnar, S., Ohtani, K., Ohno, H., Field, M. and Sullivan, G. J. (2005). Detection of single magnetic bead for biological applications using an InAs quantum-well micro-Hall sensor, Applied Physics Letters, 87, 112502. [82] Mitchell, D. T., Lee, S. B., Trofin, L., Li, N., Bevanen, T. K., Söderlund, H. and Martin, C. R. (2002). Smart nanotubes for bioseparations and biocatalysis, Journal of the American Chemical Society, 124, 11864-11865. [83] Morrow, K. J., Bawa, R. and Wei, C. M. (2007). Recent advances in basic and clinical nanomedicine, Medical Clinics of North America, 91 (5), 805-843. [84] National Institute of Health, and National Cancer Institute. (2004). Cancer Nanotechnology, US Department of Health and Human Services. [85] Nicewarner-Pena, S. R., Freeman, R. G., Reiss, B. D., He, L., Pena, D. J., Walton, I. D., Cromer, R., Keating, C. D. and Natan, M. J. (2001). Submicrometer metallic barcodes, Science, 294 (5540), 137-141.
282
Magnetic Nanomaterials, Nanotubes, and Nanomedicine [86] Nidumolu, B. G., Urbina, M. C., Hormes, J., Kumar, C. S. S. R. and Monroe, W. T. (2006). Functionalization of gold and glass surfaces with magnetic nanoparticles using biomolecular interactions, Biotechnology Progress, 22, 91-95. [87] Noji, H., Yasuda, R., Yoshida, M. and Kinosita, K. Jr. (1997). Direct observation of the rotation of F-1-ATPase, Nature, 386, 299-302. [88] Osaka, T., Matsunaga, T., Nakanishi, T., Arakaki, A., Niwa, D. and Iida, H. (2006). Synthesis of magnetic nanoparticles and their application to bioassays, Analytical and Bioanalytical Chemistry, 384, 593-600. [89] Pankhurst, Q. A., Connolly, J., Jones, S. K. and Dobson, J. (2003). Applications of magnetic nanoparticles in biomedicine, Journal of Physics D: Applied Physics, 36, R167-R181. [90] Parak, W. J., Gerion, D., Pellegrino, T., Zanchet, D., Micheel, C., Williams, S. C., Boudreau, R., Le Gros, M. A., Larabell, C. A. and Alivisatos, A. P. (2003). Biological applications of colloidal nanocrystals, Nanotechnology, 14, R15–R27. [91] Piedade, M., Sousa, L. A., de Almeida, T. M., et al. (2006). A new hand-held microsystem architecture for biological analysis, IEEE Transactions on Circuits and Systems I, 53 (11), 2384-2395. [92] Pileni, M. P. (2003). The role of soft colloidal templates in controlling the size and shape of inorganic nanocrystals, Nature Materials, 2, 145-150. [93] Piraux, L., Encinas, A., Vila, L., Mátéfi-Tempfli, S., Mátéfi-Tempfli, M., Darques, M., Elhoussine, F. and Michotte, S. (2005). Magnetic and superconducting nanowires, Journal of Nanoscience and Nanotechnology, 5, 372–389. [94] Reich, D. H., Tanase, M., Hultgren, A., Bauer, L. A., Chen, C. S. and Meyer, G. J. (2003). Biological applications of multifunctional magnetic nanowires, Journal of Applied Physics, 93 (10), 7275-7280. [95] Reiss, B. D., Mao, C., Solis, D. J., Ryan, K. S., Thomson, T. and Belcher, A. M. (2004). Biological routes to metal alloy ferromagnetic nanostructures, Nano Letters, 4, 1127-1132. [96] Ren, Q., Zhao, Y. P., Yue, J. C. and Cui Y. B. (2006). Biological application of multi-component nanowires in hybrid devices powered by F1-ATPase motors, Biomedical Microdevices, 8, 201–208. [97] Rife, J. C., Miller, M. M., Sheehan, P. E., Tamanaha, C. R., Tondra, M. and Whitman, L. J. (2003). Design and performance of GMR sensors for the detection of magnetic microbeads in biosensors, Sensors and Actuators A, 107, 209-218. [98] Salata, O. V. (2004). Applications of nanoparticles in biology and medicine, Journal of Nanobiotechnology, 2, 1-6. [99] Salem, A. K., Searson, P. C. and Leong, K.W. (2003). Multifunctional nanorods for gene delivery, Nature Materials, 2, 668-671. [100] Sandler, R. (2009). Nanotechnology: The Social and Ethical Issues, Woodrow Wilson International Center for Scholars, Washington, DC. [101] Sanhai, W. R., Sakamoto, J. H., Canady, R., and Ferrari, M., (2008). Seven challenges for nanomedicine, Nature Nanotechnology, 3, 242-244. [102] Sarkar, J., Khan, G. G. and Basumallick, A. (2007). Nanowires: properties, applications and synthesis via porous anodic aluminium oxide template, Bulletin of Materials Science, 30 (3), 271-290. [103] Satyanarayana, M. (2005). Microfluidics: the flow of innovation continues, Monthly Feather, NCI Alliance for Nanotechnology in Cancer, August 2005. [104] Seifert, G., Köhler, T. and Tenne, R. (2002). Stability of Metal Chalcogenide Nanotubes, Journal of Physical Chemistry B, 106, 2497-2501. [105] Sincai, M., Ganga, D., Ganga, M., Argherie, D. and Bica, D. (2005). Antitumor effect of magnetite nanoparticles in cat mammary adenocarcinoma, Journal of Magnetism and Magnetic Materials, 293, 438-441.
References
283
[106] Son, S. J., Reichel, J., He, B., Schuchman, M. and Lee, S. B. (2005). Magnetic nanotubes for magnetic-field-assisted bioseparation, biointeraction, and drug delivery, Journal of the American Chemical Society, 127, 7316-7317. [107] Stemme, E. and Stemme, G. (1993). A valveless diffuser:nozzle-based fluid pump. Sensors and Actuators A, 39, 159–167. [108] Stoner, E. C. and Wohlfarth, E. P. (1948). A mechanism of magnetic hysteresis in heterogeneous alloys, Philosophical Transactions of the Royal Society of London Series A – Mathematical and Physical Sciences, 240, 599-642. [109] Sui, Y. C., Skomski, R., Sorge, K. D. and Sellmyer, D. J. (2004). Nanotube magnetism, Applied Physics Letters, 84, 1525-1527. [110] Sun, L., Hao, Y., Chien, C. L. and Searson, P. C. (2005). Tuning the properties of magnetic nanowires, IBM Journal of Research and Development, 49 (1), 79-102. [111] Tao, F., Guan, M., Jiang, Y., Zhu, J., Xu, Z. and Xue, Z. (2006). An easy way to construct an ordered array of nickel nanotubes: the triblock-copolymer-assisted hard-template method, Advanced Materials, 18, 2161-2164. [112] Tok, J. B. H., Chuang, F. Y. S., Kao, M. C., Rose, K. A., Pannu, S. S., Sha, M. Y., Chakarova, G., Penn, S. G. and Dougherty, G. M. (2006). Metallic striped nanowires as multiplexed immunoassay platforms for pathogen detection, Angewandte Chemie, International Edition, 45, 6900-6904. [113] Tourillon, G., Pontonnier, L., Levy, J. P. and Langlais, V. (2000). Electrochemically synthesized Co and Fe nanowires and nanotubes, Electrochemical and Solid-State Letters, 3, 20-23. [114] Turner, A.P. (2000). Biosensors – sense and sensitivity, Science, 290, 1315-1317. [115] Varadan, V. K., Chen, L. F. and Xie, J. N. (2008). Nanomedicine: Design and Applications of Magnetic Nanomaterials, Nanosensors and Nanosystems, Wiley, Chichester. [116] Vo-Dinh, T., Cullum, B. M. and Stokes, D. L. (2001). Nanosensors and biochips: frontiers in biomolecular diagnostics, Sensors and Actuators B, 74, 2-11. [117] Wang, S. X. and Li, G. (2008). Advances in giant magnetoresistance biosensors with magnetic nanoparticle tags: review and outlook, IEEE Transactions on Magnetics, 44 (7), 1687-1702. [118] Wang, T., Wang, Y., Li, F., Xu, C. and Zhou, D. (2006). Morphology and magnetic behavior of an Fe3O4 nanotube array, Journal of Physics: Condensed Matters, 18, 10545-10551. [119] Wang, J. (2000). From DNA biosensors to gene chips, Nucleic Acids Research, 28 (16), 3011-3016. [120] Whitesides, G. M. and Wong, A. P. (2006). The Intersection of Biology and Materials Science, MRS Bulletin, 31 (1), 19-27. [121] Wildt, B., Mali, P. and Searson, P. C. (2006). Electrochemical template synthesis of multisegment nanowires: fabrication and protein functionalization, Langmuir 22, 10528-10534. [122] Wohlfarth, E. P. (1958). Relations between different modes of acquisition of the remanant magnetization of ferromagnetic particles, Journal of Applied Physics, 29, 595-595. [123] Xie, J., Chen, L., Varadan, V. K., Yancey, J. and Srivatsan, M. (2008). The effects of functional magnetic nanotubes with incorporated nerve growth factor in neuronal differentiation of PC12 cells, Nanotechnology, 19, 105101-1-7. [124] Yih, T. C. and Wei, C. (2005). Nanomedicine in cancer treatment, Nanomedicine: Nanotechnology, Biology, and Medicine, 1, 191-192. [125] Zeng, H., Li, J., Wang, Z. L., Liu J. P. and Sun, S. (2004a). Bimagnetic core/shell FePt/Fe3O4 nanoparticles, Nano Letters, 4, 187-190. [126] Zeng, H., Sun, S., Li, J., Wang Z. L. and Liu, J. P. (2004b). Tailoring magnetic properties of core/shell nanoparticles, Applied Physics Letters, 85 (5), 792-794. [127] Zeng, H., Li, J., Liu, J. P., Wang Z. L. and Sun, S. (2002). Exchange-coupled nanocomposite magnets by nanoparticle self-assembly, Nature, 420, 395-398.
284
Magnetic Nanomaterials, Nanotubes, and Nanomedicine [128] Zhou, W., Gao, P., Shao, L., Caruntu, D., Yu, M., Chen, J. and O’Connor, C. J. (2005). Drug-loaded, magnetic, hollow silica nanocomposites for nanomedicine, Nanomedicine: Nanotechnology, Biology, and Medicine, 1, 233-237.
CHAPTER 10
Mobile Microscopic Sensors for In Vivo Diagnostics Tad Hogg
10.1
Introduction Nanotechnology has the potential to revolutionize health care [1–3]. A current example is the use of nanoscale particles for enhanced medical imaging [4]. Future possibilities include programmable machines comparable in size to cells and able to sense and modify their environments. Such microscopic robots (“nanorobots”) could provide significant medical benefits by operating within the body [1, 5] to gather information continually over a period of time, in contrast with the more limited monitoring possible with a series of conventional laboratory tests. The limitations of such tests have already led to clinical use of implanted or ingested medical devices, such as pacemakers and pill-sized cameras to view the digestive tract. Extending this capability to much smaller devices can provide additional benefits. In particular, a large number of microscopic devices could monitor individual cells in many locations simultaneously, providing more comprehensive monitoring than is possible with today’s larger implanted devices. Realizing these benefits requires fabricating the robots cheaply and in large numbers. Such fabrication is beyond current technology. Nevertheless, ongoing progress in developing nanoscale devices could eventually enable production of such robots. One approach is engineering biological systems (e.g., bacteria executing simple programs) [6]. For creating microscopic programmable devices in large numbers, engineering biological organisms have the advantage of using existing functioning systems so there is no need, for instance, to fabricate power sources de novo. Instead, desired modifications to behavior are introduced in the context of an existing metabolic system, for example, through introduction of additional genes. However, biological organisms have limited material properties and computational speed. Moreover, any modifications must maintain viability of the organisms. For in vivo applications, such organisms may trigger immune responses, limiting their operation time or adversely affecting the host. Thus the focus in this chapter is on machines based on plausible extensions of currently demonstrated nanoscale electronics, sensors, and motors [7–13] and relying on directed assembly [14]. These demonstrations include programmable electronics suitable for memories and logic operations as needed for storing and executing programs. Another example is indi-
285
286
Mobile Microscopic Sensors for In Vivo Diagnostics
vidual chemical sensors, based on functionalized nanowires that exhibit highly specific chemical detection. These components enable nonbiological robots that are stronger, faster, and with more operational programmability than is possible with biological organisms. Current technology does not allow combining these demonstrated components into complete robots (i.e., machines combining sensing, power generation, computation, and other capabilities such as communication or locomotion). Our long-term vision is that molecular electronics and nanoscale chemical sensors could enable constructing microscopic robots suitable for a variety of biomedical tasks, and currently demonstrated components provide sufficient insight into the likely robot capabilities to form the basis for evaluating robot performance for various tasks, including biomedical applications. Thus, even at this early stage of technology development, it is both feasible and worthwhile to evaluate generic tasks for these future robots to identify their capabilities and aid selecting among various hardware designs possibilities. A major challenge for nanorobots arises from the physics of their microenvironments and the hardware limitations of the robots, which differ in several significant respects from today’s larger robots. First, the robots will often operate in fluids containing many moving objects, such as cells, and dominated by viscous forces. Second, thermal noise is a significant source of sensor error and Brownian motion limits the ability to follow precisely specified paths. Third, relevant objects are often recognizable via chemical signatures rather than visual markings or shape. Fourth, the tasks involve large numbers of robots, each with limited abilities. Moreover, a task will generally require only a modest fraction of the robots to respond appropriately, not for all, or even most, robots to do so. This observation contrasts with teams of larger robots with relatively few members, such as robot soccer or surveillance: incorrect behavior by even a single robot can significantly decrease team performance. The focus in this chapter is on biomedical applications requiring only modest hardware capabilities. This focus is relevant for exploring likely medical applications for relatively early development of nanorobots whose capabilities will be more limited than with a mature technology. Designing control protocols for microscopic robots is a key challenge, not only enabling useful performance but also compensating for their limited computation, locomotion, or communication abilities. Distributed control is well-suited to these capabilities by emphasizing locally available information and achieving overall objectives through self-organization of the collection of robots. Theoretical studies allow developing such controls and estimating their performance prior to fabrication, thereby indicating design trade-offs among hardware capabilities, control methods, and task performance. Such studies of microscopic robots complement analyses of individual nanoscale devices [13, 15], and indicate even modest capabilities enable a range of novel applications. To illustrate the potential of microscopic robots for medicine, this chapter describes plausible robot capabilities from early nanotechnology based on extrapolations from current laboratory demonstrations of nanoscale devices. A significant feature of these robots is their small size, which allows them to pass through even the smallest blood vessels. This capability makes the robots well-suited for monitoring for specific patterns of chemicals, enabling a prototypical diagnostic task of finding a small chemical source in a multicellular organism via its circulatory system. Infor-
10.2 Robot Capabilities and Environment
287
mation from a large number of microscopic robots allows inferring properties of tiny chemical sources in a macroscopic tissue volume. Although such robots cannot yet be fabricated, estimates of their capabilities allow evaluating sensing performance for typical medically relevant chemicals released by tissues in response to localized injury or infection. The circulation could also be used to precisely aggregate robots at chemically distinguished sites within the body (e.g., to aid microsurgery). In this case, the primary operation of the robots would be in relatively fixed locations and their mobility would be used as a technique to reach those locations. Alternatively, a group of robots could be directly implanted within a macroscopic surgery area by a physician, after which the devices make small adjustments to their position based on signals from their environment. Theoretical studies of medical applications, such as those described in this chapter, suggest the robots can give significantly better performance than current medical technology, for diagnostics and, to an even greater extent, for interventions such as drug delivery and aiding microsurgery. From this flexibility of application, we can expect significant benefits even from relatively early developments of nanotechnology, which will pave the way for more complex applications as the technology matures.
10.2
Robot Capabilities and Environment A robot is a machine able to sense and act on its environment, and with computational ability to determine actions in response to sensor inputs. A typical size for a nanorobot is about one micron, comparable to the size of bacteria, and small enough to fit inside even the smallest blood vessels. Significant events take place on millisecond time scales, and motions involve speeds comparable to flow rates in small blood vessels: a few millimeters per second. This section describes major physical properties of the robots and their environments. Order of magnitude estimates of these properties are sufficient to identify tasks for which the robots could perform well. Minimal robot capabilities for biomedical tasks include chemical sensing, computation, and power. Additional capabilities, enabling more sophisticated applications, include abilities to stick to specific cell surfaces by altering surface properties [16], communicate, move, and alter the robot’s environment (e.g., by releasing chemicals). 10.2.1
Sensing
Large-scale robots often use sonar or cameras to sense their environment. These sensors locate objects from a distance, and involve sophisticated interpretation algorithms. In contrast, microscopic robots for biological applications will mainly use chemical sensors. For example, nanoscale sensors can convert molecular recognition to a measurable electric current [17]. Microscopic robots and bacteria face similar physical constraints in detecting chemicals [18]. Current molecular electronics [13] and nanoscale sensors [19–21] indicate plausible sensor capabilities. At low concentrations, sensor performance is
288
Mobile Microscopic Sensors for In Vivo Diagnostics
primarily limited by the time for molecules to diffuse to the sensor. Statistical fluctuations in the number of molecules encountered are a major source of noise. Even at the scale of these robots, individual molecules are tiny so for evaluating robot behaviors the chemicals can be approximated as a continuous concentration C of molecules per unit volume. For the molecules, and to a lesser extent the robots themselves, thermal noise gives a continual, random change in position. A diffusion coefficient D characterizes this Brownian motion: the root-mean-square displacement x in at time t is x = 6Dt
(10.1)
Thus the typical distance an object in the fluid travels grows only as the square root of the time rather than linearly. This means diffusion is slow at large scales but much faster at the small scales relevant for microscopic robots. A simple estimate of sensor performance treats the robots as absorbing spheres. The diffusive capture rate r for a sphere of radius a in a region with concentration C is [22] r = 4πDaC
(10.2)
For chemicals in the fluid, the concentration C is governed by the diffusion equation [22], ∂C = -— ∑ F ∂t
(10.3)
where F = -D—C + vC is the chemical flux; that is, the rate at which molecules pass through a unit area, v is the fluid velocity vector and — ∑ F is the sum of derivatives of the flux in each of three mutually perpendicular directions (conventionally denoted as the x, y, and z coordinate directions). The first term in the flux is diffusion, which acts to reduce concentration gradients, and the second term is motion of the chemical due to the movement of the fluid in which the chemical is dissolved. Figure 10.1 compares a solution to this diffusion equation with underlying random walks of individual molecules due to Brownian motion. Because individual molecules are small compared to the robots, the continuum approximation leading to the diffusion equation works well even for microscopic robots. Thus there is no need for models of robot behavior to follow paths of individual molecules. For estimating the behavior of chemicals, the simplest approximation is that the vessel is occupied only by fluid in which the chemical is dissolved [23, 24]. A better approximation includes effects of objects, such as blood cells, in the fluid. Specifically, the hydrodynamic effect of rotating blood cells moving in the fluid can be approximated as increasing the diffusion constant of objects in the fluid. A quantitative estimate for the effect of blood cells in small vessels is to increase the diffusion coefficient by an amount D rotate ª 10 -9 m 2 / s [25]. In addition to chemicals, robots could sense other properties of their environment. For example, nanoscale sensors for fluid motion can measure flow rates at speeds relevant for biomedical tasks [26], allowing robots to examine microfluidic behavior in small vessels. Since boundaries significantly alter fluid behavior far into
10.2 Robot Capabilities and Environment
289
Figure 10.1 (Color plate 15) Two-dimensional diffusion of molecules in a box, starting from the left wall and ending at an absorbing circle in the center, with diffusion coefficient D = 10-10 m2/s and flux from the wall of magnitude F = 1013 molecule/s/m2. The circle has radius of one micron and the area is a square 20 μm on a side. Left: three random walks, starting from the gray points on the left wall, representing paths of individual molecules. Right: the continuum approximation to this diffusion, averaged over many molecules to give the concentration value, ranging from 1018 molecule/m3 at the left wall (red) to zero at the circle (dark blue). The curves show the average motion of molecules from the wall to the circle.
the vessel [27], several such sensors, extending a small distance from the robot surface in various directions, could detect changes in the vessel geometry. Such estimates of local geometry might, for example, help distinguish normal vessels from leaky new vessels formed within tumors [28]. As another example, robots could sense local electric fields [29, 30] or mechanical stiffness of cells which can be a useful diagnostic [31]. Combining several sensor modalities provides a wide range of options for nanorobot use. 10.2.2
Communication
Several forms of communication could be useful for nanorobots. These include: • • •
Broadcast from a large external source to all robots in a macroscopic volume; Two-way communication among neighboring robots; Aggregated, slow communication from robots to an external receiver.
The simplest form of communication is receiving electromagnetic or acoustic signals broadcast from outside the body. This only requires the robots to receive signals, thereby avoiding more difficult fabrication and increased power consumption required to enable the robots to transmit signals [5]. Such signals could be used to activate robots only within certain regions of the body at, say, centimeter-length scales. Furthermore, the communication could be selective for particular robots (e.g., those with specified identifiers or detecting certain events) while others receiving the signal just ignore it. Communicating among nearby robots and sending information to detectors outside the body increases the range of tasks for the robots. For instance, acoustic
290
Mobile Microscopic Sensors for In Vivo Diagnostics
communication among nearby robots (e.g., within about 100 μm of each other [5]) allows coordinating their activities (e.g., to form structures using simple controls proposed for larger reconfigurable robots [32–34]). For limited communication with the attending physician, the robots could carry nanoscale materials with high response to an external signal. For example, such materials could respond to light of particular wavelengths when near the skin [35], or give enhanced external imaging [36]. Such visualization mechanisms combined with a selective ability to stick to vessel walls allows detecting aggregations of devices at specified locations near the surface of the body [37]. Robots could use various areas near the skin (e.g., marked with various light or ultrasound frequencies) at centimeter scales as readout regions during operation. For example, robots that have detected certain chemicals could aggregate at the corresponding readout location, which would then be visible externally. Robots could choose how long to remain at the aggregation points based on how high a concentration of the chemical pattern they detected. Robots with local communication capabilities could compare observations while in these aggregation regions, allowing further computation to influence the communicated result, for example, by changing how long the devices remain at the readout location or whether they aggregate in other locations at a later time. This indication of whether, and (at a coarse level) what, the devices have found could help decide how long to continue circulating to improve statistics for weak chemical signatures. These aggregation points could also be used to signal to the devices (e.g., instructing them to switch among a few preprogrammed modes of operation). The robots could also communicate with nearby biological cells. For such communication, molecules on the surface of the robot could mimic existing signaling molecules to bind to receptors on the cell surface, for example to activate nerve cells [38]. Such signaling could be useful both for diagnostics and treatment. 10.2.3
Motion: Passive and Active
Large robots typically need active locomotion to find suitable locations for their task (e.g., foraging). Microscopic robots, on the other hand, can be introduced into circulating fluids (e.g., in blood vessels) and rely on the fluid to move them around the body. Biomedical applications will often involve robots operating in fluids. Viscosity dominates the motion of microscopic objects in fluids, with different physical behaviors than for larger organisms and robots [27, 39–42]. The ratio of inertial to viscous forces for an object of size s moving with velocity v through a fluid with viscosity η and density ρ is the Reynolds number Re=s ρv/η. Using typical values for density, viscosity (e.g., of water or blood plasma) and fluid speeds, motion of a micron-sized robot has Re << 1, so viscous forces dominate. By contrast, a swimming person has Reynolds number about a billion times larger, and viscous forces are minor. Details of the flow are governed by the Navier-Stokes equation which, in all but the simplest cases, must be solved numerically [42]. Passive motion in fluid is useful to disperse robots throughout a macroscopic tissue volume, with no need for powered locomotion or navigation. An approximation for behavior of robots without active locomotion is to assume they move with the
10.2 Robot Capabilities and Environment
291
Figure 10.2 Diffusion of molecules from a 20 μm source on the wall of a two-dimensional channel (indicated by the black line at the bottom of the plot), of width 10 μm. Chemical diffusion coefficient and flux magnitude from the source are the same as in Figure 10.1. The fluid flows from left to right with average speed of 1 μm/s. Concentration ranges from about 1017 molecule/m3 near the source (white) to zero (black).
same velocity as fluid would have at the center of the robot if the robot were not there. This approximation neglects the change in fluid flow due to the robots, and is reasonable when the robots are separated by many times their size, and each robot is small compared to the vessel containing the fluid. Closer packing leads to more complex motion due to hydrodynamic interactions [43, 44]. Active locomotion could be useful over short distances for robots to home in on specific locations once passive motion brings them close enough to detect those locations (e.g., due to chemical signatures). At a fine scale, locomotion requires motors to move through the fluid, analogous to biological molecular motors [45]. For coarser motion control, with simpler technological requirements, external fields can push suitably engineered devices toward specific regions [46]. An additional consideration for motion planning of microscopic robots is Brownian motion, which randomly changes the location and orientation of microscopic robots, thereby limiting the time over which they can reliably compare different locations or directions. This behavior contrasts with long range path planning with maps of the environment often used for larger robots. As with chemicals, a simple approximation for the effect of objects in the fluid, such as cells, is increasing the diffusion coefficient of the robots by Drotate, ignoring systematic hydrodynamic interactions that could bias location of robots (e.g., toward walls). 10.2.4
Computation
Computation is a key difference between robots and simpler nanoparticles with selective binding as currently used for imaging or drug delivery. A control program, even a fairly simple one, operating in each robot allows a wider range of activities and responses than relying on physical or chemical interactions of nanoparticles. For instance, the robot response could be programmed to depend on the number of other nearby robots or the specifics of the local environment. To estimate computational requirements, high resolution diagnosis involves chemical sources as small as a single cell (i.e., about 10 μm in size). Flow in the small vessels has a range of speeds up to about a millimeter per second. Thus, robots
292
Mobile Microscopic Sensors for In Vivo Diagnostics
would pass near the source, where concentration would be highest and easiest to detect, on millisecond time scales. Recognizing a chemical detection involves at least a few arithmetic operations to compare sensor counts to prespecified threshold values. An estimate on the required computational capability is about 100 elementary logic operations within a 10-ms measurement time. This gives about 104 logic operations per second. While modest compared to current computers, this rate is significantly faster than demonstrated for programmable bacteria [6] but well within the capabilities of molecular electronics. 10.2.5
Power
For biomedical tasks, computation generally uses much less power than communication or locomotion. A micron-scale robot moving through a water-like fluid at 1 μm/s dissipates about a picowatt [22] to overcome fluid drag and the inefficiencies of locomotion. Communication power requirements are of a similar order of magnitude, depending on the distance and desired communication rate [5]. This amount of power is comparable to that used by a single biological cell. For tasks of limited duration, onboard fuel created during robot manufacture could suffice. Otherwise, the robots could use energy available in their environment, such as converting externally generated vibrations to electrical energy [47] or chemical generators (e.g., a fuel cell using glucose and oxygen in the bloodstream [5]). In some applications, power and a coarse level of control can be combined by using an external source (e.g., light, to activate chemicals in the fluid to power the machines in specific locations [48]).
10.3
Using Microscopic Robots The properties of microscopic robots require different approaches to task planning than is the case for large robots. For example, instead of detailed planning for individual robots, the statistics of random variations can ensure, with high probability, at least some robots complete the task. For example, passive motion in the fluid of the circulatory system is a simple approach to bring robots close to all parts of a tissue, without need for active locomotion or long-range path planning. With a sufficient number of robots, at least some of them are likely to pass through a given small vessel in a reasonable time, allowing robots to disperse throughout the tissue volume without need for coordinated path planning. Figure 10.3 illustrates the minimal components for the robots: sensors, computing, and power. The ability to execute a program stored in memory distinguishes these devices from currently used nanoparticles, which are also somewhat smaller than the robot. Depending on the task, robots could also carry drugs to release as determined by their programs or include the communication or locomotion capabilities described in Section 10.2. Individual robots have limited capabilities, both physically and computationally, but can monitor and act in their local environments. While individual robots are limited in their effects on the environment, a large number of them could perform complex tasks over macroscopic scales. By contrast, larger scale computers and people can make more complex decisions, using information aggregated
10.3 Using Microscopic Robots
Figure 10.3
293
Schematic illustration of a microscopic robot with minimal capabilities.
from many of the robots, but at coarser spatial resolution and at slower time scales. Thus a natural approach to nanorobot task planning combines slow, coarse, global control (to evaluate aggregated information, set overall constraints and make complex decisions) with simple, fast, reactive local control for individual robots. Organisms contain many microenvironments, with distinct physical, chemical, and biological properties. Often, precise quantitative values of properties relevant for robot control will not be known a priori. This observation suggests a multistage protocol for using the robots. First, an information-gathering stage with passive robots placed into the organism, for example through the circulatory system, to measure relevant properties [24]. The robots could also use additional sensing modalities and communication to identify correlations among multiple chemical sources. Actions based on the information from the robots would form a second stage of activity, perhaps with a smaller number of specialized robots (e.g., containing drugs to deliver near cells), with controls based on the calibration information retrieved earlier. An example of an action the robots could take is releasing drugs only to cells matching a prespecified chemical profile [5, 49] as an extension of a recent in vitro demonstration of this capability using DNA computers [50]. The detection thresholds for these actions could be determined with the information retrieved from the first stage of operation. For more sophisticated actions, robots could follow chemical gradients to aggregate at the chemical sources [23] or manipulate biological structures based on surface chemical patterns on cells. Robots aggregated at chemically identified targets could perform precise microsurgery at the scale of individual cells, extending surgical capabilities of simpler nanoscale devices [51]. Since biological processes often involve activities at molecular, cell, tissue, and organ levels, such microsurgery could complement conventional surgery at larger scales. For instance, a few millimeter-scale manipulators, built from micromachine (MEMS) technology, and a population of microscopic devices could act simultaneously at tissue and cellular size scales. An example involving microsurgery for nerve repair with plausible bio-
294
Mobile Microscopic Sensors for In Vivo Diagnostics
physical parameters indicates the potential for significant improvement in both speed and accuracy compared to using the larger-scale machines alone [52, 53]. These active scenarios require more advanced robot capabilities, such as locomotion and intrarobot communication, than needed for passive sensing. These scenarios also require more complex controls; for example, microscopic active swimmers [54] could exploit the hydrodynamic interactions among swimming objects as they aggregate so the distance between devices becomes only a few times their size [43]. A more subtle from of manipulating their environment is if robots release chemical signals that affect behavior of nearby cells, which could then amplify the response by recruiting other cells (e.g., for an immune response). Such an approach would be a small-scale analog of robots affecting self-organized behavior of groups of organisms [55]. The robots could monitor environmental changes due to their actions, thereby documenting the progress of the treatment. Thus the physician could monitor the robots’ progress and decide whether and when they should continue to the next step of the procedure. Using a series of steps, with robots continuing with the next step only when instructed by the supervising person, maintains overall control of the robots, and simplifies the control computations each robot must perform itself.
10.4
Evaluating Robot Behaviors Because it is not yet possible to fabricate nanorobots, studies of their behavior rely on theory and simulations. As technology develops to fabricate early versions of the robots, simple experiments will help validate the simulations. In particular, one challenge for theoretical studies is the poorly characterized physical parameters of the microenvironments the robots will operate in. Early nanorobots, with limited capabilities, could help quantify these properties, thereby leading to more accurate simulations and improved designs for the robots. This section describes some of these evaluation possibilities. 10.4.1
Theoretical Studies
A variety of theoretical approaches allow estimating the task performance of nanorobots. The simplest approach relies on estimates of individual capabilities to indicate the plausible range of tasks the robots could perform [5]. More detailed studies consider the combination of robot capabilities and the physical properties of the task environment. Cellular automata are one approach to evaluating collective robot behavior. These automata are a set of simple machines that are laid out on a regular lattice. Each machine is capable of sending and receiving messages with its neighbors on the lattice and updates its internal state based on a simple rule. For example, a two-dimensional scenario shows how robots could assemble structures [56] using local rules. Such models can help understand structures formed at various scales through simple local rules and some random motions [57, 58]. However, cellular automata models either ignore or greatly simplify physical behaviors such as fluid
10.4 Evaluating Robot Behaviors
295
flow. A related analysis technique considers swarms [59] (i.e., large groups of simple machines or biological organisms such as ants). In these systems, individuals are assumed to use simple rules to determine their behavior from information about a limited portion of their environment and neighboring individuals. Typically, individuals in swarms are not constrained to have a fixed set of neighbors but instead continually change their neighbors as they move. Swarm models are well-suited to microscopic robots with their limited physical and computational capabilities and large numbers. Most swarm studies focus on macroscopic robots or behaviors in abstract spaces [60], which do not specifically include physical properties unique to microscopic robots. In spite of the simplified physics, these studies show how local interactions among robots lead to various collective behaviors and provide broad design guidelines. Simulations including physical properties of microscopic robots and their environments can evaluate robot performance. Simple models, such as a two-dimensional simulation of chemotaxis [61], indicate how robots find microscopic chemical sources. A more elaborate simulator [62] includes three-dimensional motions in viscous fluids, Brownian motion and environments with numerous cellsized objects, though without accounting for how they change the fluid flow. Studies of hydrodynamic interactions [43] among moving devices include more accurate fluid effects. Another approach estimates typical robot behaviors using a stochastic approximation [63]. This method directly evaluates average behaviors of many robots through differential equations determined from the state transitions used in the robot control programs. Direct evaluation of average behavior avoids the numerous repeated runs of a simulation needed to obtain the same result. This approach successfully describes behaviors of teams of small numbers of large robots. Microscopic robots, with limited computational capabilities, will likely use simple controls, with minimal dependencies on events in individual robot histories, for which this stochastic approximation is ideally suited. The approach can also readily incorporate spatial variations such as fluid speeds and chemical concentrations [64] relevant for nanorobot tasks. Even at micron scales, the molecular nature of these quantities can be approximated as continuous fields governed by partial differential equations. For application to microscopic robots, this approximation extends to the robots themselves, treating their locations as a continuous concentration field, and their various control states as corresponding to different fields, much as multiple reacting chemicals are described by separate concentration fields. This continuum approximation for average behavior of the robots will not be as accurate as when applied to chemicals or fluids, but nevertheless gives a simple approach to average behaviors for large numbers of microscopic robots. Cellular automata, swarms, physically-based simulations and stochastic analysis are all useful tools for evaluating the behaviors of microscopic robots. One example is evaluating the feasibility of rapid, fine-scale response to chemical events too small for detection with conventional methods, including sensor noise inherent in the discrete molecular nature of low concentrations. The stochastic analysis approach allows incorporating more realistic physics than used with cellular automata studies, and is computationally simpler than repeated simulations to obtain average behaviors. This technique is limited in requiring approximations for
296
Mobile Microscopic Sensors for In Vivo Diagnostics
dependencies introduced by the robot history, but readily incorporates physically realistic models of sensor noise and consequent mistakes in the robot control decisions. The stochastic analysis indicates plausible performance, on average, and thereby suggests scenarios suited for further, more detailed, simulation studies. For a more detailed look at the robot behaviors, simulations can include various levels of detail, giving a trade-off between physical accuracy and computation required to simulate large numbers of robots over relevant time scales. Such simulations can readily include individual robot histories and correlations in behavior that are not easily treated with the stochastic approximation. In addition to evaluating performance of hypothetical nanorobots, theoretical studies identifying tradeoffs among control complexity and hardware capabilities can aid future fabrication. Specifically, control can compensate for limited hardware (e.g., sensor errors or power limitations), providing design freedom to simplify the hardware through additional control programs. Thus the studies can help determine minimum hardware performance capabilities needed to provide robust systems-level behavior. 10.4.2
Modeling Multiple Physical Effects
One challenge for theoretical evaluation of robot behaviors is the relevance of a variety of physical effects. Depending on the specific scenario, these effects can include fluid flow, diffusion of chemicals, chemical reactions, acoustics for communication and heat generated by robot operation. These effects can interact: for example, fluid flow and diffusion bring chemicals used as fuel, the devices heat their environments as they use the fuel to generate power, and the change in temperature alters the flow and diffusion rates. Thus one role for theoretical studies is quantifying these physical effects and identifying which interactions among them are significant. In developing scenarios to study, computational requirements are an important practical issue. Except for the simplest geometries, physical properties such as chemical concentration must be computed numerically, with the finite element method [65] a popular approach. Section 10.6 describes some available software packages for this approach as well as other nanoscale simulations. Using these methods for general three-dimensional geometries is computationally intensive. Fortunately, for preliminary quantitative studies of robot behaviors, simplified scenarios can give useful insight with significantly reduced computational requirements. Such simplifications include two-dimensional and axially symmetric three- dimensional geometries. The latter case, appropriate for behavior within circular vessels, has physical behavior independent of angle of rotation within the vessel that allows modeling with just a two-dimensional slice through the vessel, extending from the center of the vessel to its wall. Another choice in designing scenarios is whether to include variation in time or focus on steady-state behavior. Finally, computational feasibility requires a choice between level of detail of modeling individual devices and the scale of the simulation, both in number of devices and physical size of the environment considered. Suitable boundary conditions can account, approximately, for behavior of the physical environment beyond the portion included directly in the simulation. For example, in models of robot behaviors in small vessels, boundary conditions characterize how the fluid flow in small vessels connects to that of larger
10.4 Evaluating Robot Behaviors
297
vessels [66] and specify the background concentration of chemicals entering the vessel. To illustrate these computational issues, we consider interacting physical effects for robots using oxygen and glucose available in blood plasma as their power source. For simplicity we consider an axially symmetric geometry illustrated in Figure 10.4: a segment of a small vessel with a group of robots attached in a ring around the vessel wall. To ensure axial symmetry, we model the interior of the robots with uniform physical properties and take their shapes conforming to the vessel wall with no gaps between neighboring robots, as shown in Figure 10.5. In this example, we take the robot and vessel diameters to be 2 and 20 μm, respectively, and model a section of the vessel 50 μm long. Glucose and oxygen continually enter at the left of the vessel with the fluid flow, at average speed vavg = 10-3m/s. We examine steady-state behavior to determine the power available to the robots and consequences on the fluid flow, chemical concentrations and tissue heating due to these robots.
Figure 10.4 Schematic illustration of flow in one vessel, of radius R. Fluid flows from left to right with average velocity vavg. The gray area indicates a region wrapped around the surface of the vessel used for models with axial symmetry.
Figure 10.5 Schematic placement of devices on the inner wall of a small vessel. To allow viewing the inside of the vessel, only one half of the vessel is shown. This example has five groups of robots, each group consisting of 30 robots forming a ring around the vessel. This choice of robot placement gives an axially-symmetric geometry. The dashed line marks a half-plane within the vessel that is sufficient for the analysis: behavior at all other locations in this model is the same as on the corresponding location on this half-plane.
298
Mobile Microscopic Sensors for In Vivo Diagnostics
This simulation involves coupling among fluid flow, chemical diffusion, power generation from reacting chemicals and waste production, of both chemicals and heat. In this case, the chemical reaction is combining glucose and oxygen to produce water and carbon dioxide: C6H12O6 + 6O2 Æ 6CO2 + 6H2O. Each such reaction releases 4×10-18J of energy. Concentrations in blood plasma are in the millimolar range (about 1024molecule/m3) for glucose and CO2 but only about 1022molecule/m3 for oxygen, which is the limiting factor in obtaining power from this reaction. This illustration considers only chemicals dissolved in the blood plasma, so does not include any additional oxygen released by blood cells as they pass the group of robots in about 10 milliseconds. For the heat response, we take thermal properties to be similar to that of water and treat the vessel as embedded in a larger tissue environment whose boundary is at body temperature, as is the fluid entering the vessel. Numerically solving this coupled model gives the fluid flow, chemical concentrations, power generation, and temperature throughout the modeled region. For example, Figure 10.6 shows the resulting power generation. The robots at the upstream edge of the group receive more oxygen than those downstream and hence produce more power. Interestingly, power generation does not decrease monotonically: robots at the downstream edge have somewhat more available oxygen than those in the middle of the group. The fraction of this power available for useful activity within the robot (e.g., computation) depends on the efficiency of the glucose engine design [5]. The devices in this example have volume less than 10 μm3 so the values in Figure 10.6 correspond to power densities around 107 W/m3. Such large values raise concerns of significantly heating the surrounding tissue. However, for the group of 150 robots used in this scenario, heat is rapidly removed resulting in negligible temperature increases, as shown in Figure 10.7. Corresponding to their larger power generation, the upstream robots (on the left) have the highest temperature.
Figure 10.6 Steady-state power generation for robots as a function of their position along the vessel wall, starting from those at the upstream end of the group (position 1) and continuing to those at the downstream end (position 5).
10.4 Evaluating Robot Behaviors
299
Figure 10.7 (Color plate 16) Temperature distribution in the robots, vessel and surrounding tissue. The five small squares are the locations of the robots along the vessel wall, corresponding to the cross section through the robots of the half-plane shown in Figure 10.5. The lower portion of the diagram, below the robots, is tissue outside the vessel. The full three-dimensional solution from this model arises from rotating the plane shown here around the line at the top marking the center of the vessel. The temperature ranges from equal to the surrounding body temperature (dark blue) to an increase of only 10-4 °C above body temperature (red). The lines show the average flow of oxygen (similar to the diffusion shown in Figure 10.1) through the vessel or absorbed by the robots.
This scenario also indicates how the robots affect the chemical concentrations. Their effect on oxygen is largest, since the reaction is oxygen limited: the robots consume about 80% of the dissolved oxygen. On the other hand, glucose concentration decreases by less than 1% and CO2 increases by 3%. The robots change the fluid flow by constricting the vessel, which reduces fluid speed by about 30% if the pressure does not change to compensate. In summary, this example illustrates how various physical properties interact with robot behaviors. Solving the model with different parameter choices can show how behavior depends on the properties of the environment (e.g., vessel size, fluid speed, chemical concentrations, and changes in diffusion due to Drotate). Such parametric studies can identify robust behaviors and suitable control programs for the robots to enable their effective operation in the range of environments they are likely to encounter. In practice, the robots need not generate power as fast as they receive fuel, but could instead operate in bursts of activity as they detect events of interest. In that case, they could store chemicals received over time to enable higher bursts of power (e.g., for communication). Moreover, with active control over their absorption of chemicals, the upstream robots could absorb less oxygen and thereby make more available for downstream robots. Thus the results from this simple scenario can suggest suitable controls for the robots to distribute power when they aggregate.
300
Mobile Microscopic Sensors for In Vivo Diagnostics
10.4.3
Validation Experiments
As technology advances to constructing early versions of microscopic robots, experimental evaluations will supplement theoretical studies. One such experiment is embedding the devices in bacterial biofilms to monitor chemical signals exchanged among the bacteria. In this case, the robots could be fabricated on a surface and the film grown over them, greatly simplifying constructing the robots. The surface could provide power and communication during operation. This experiment would test the ability of the chemical sensors and the onboard computation to detect patterns of chemical activity, as well as the durability of the robots. Another early validation experiment is operating the robots in manufactured microfluidic channels [27]. This would test the robots’ ability for independent operation without direct connections to external devices for power or communication. Such studies would allow testing the robots’ ability to infer properties of their microenvironments, such as vessel branching, based on fluid flow nanoscale sensors, and calibrating the chemical sensors with known concentrations introduced in the fluid. The robots could also demonstrate the ability to aggregate at chemically defined locations. After such in vitro experiments, early in vivo tests could involve robots acting as passive sensors in the circulatory system. The chemical patterns found would quantify properties of microenvironments in the body. Such nanorobots will be useful not only as diagnostic tools and sophisticated extensions to drug delivery capabilities [67], but also as an aid to develop robot designs and control methods for more active tasks.
10.5
Example Task: High-Resolution Diagnostics A prototypical medical task for microscopic robots is identifying a cell-sized source releasing chemicals into a small blood vessel. This scenario illustrates a basic capability for the robots: identifying small chemically distinctive regions, with high sensitivity due to the robots’ ability to get close (within a few cell-diameters) to a source. Microscopic robots acting independently to detect specific patterns of chemicals are analogous to swarms [59] used in foraging, surveillance or search tasks. Large numbers of such devices could move through tissues by flowing passively in moving fluids (e.g., through blood vessels or with lymph fluid). Chemicals with high concentrations are readily detected with the simple procedure of analyzing a blood sample in a medical laboratory. Thus the chemicals of interest for microscopic robot applications are those with low concentrations. With sufficiently low concentrations and small sources, the devices are likely to only encounter a few molecules while passing near the source, leading to significant statistical fluctuations in number of detections. The robots could detect localized high concentrations that are too low to distinguish from background concentrations when diluted in the whole blood volume as obtained with a sample. Moreover, if the detection consists of the joint expression of several chemicals, each of which also occurs from separate sources, the robot’s pattern recognition capability could identify the spatial locality, which would not be apparent when the chemicals are mixed throughout the blood volume. The pattern
10.5 Example Task: High-Resolution Diagnostics
301
could consist of chemicals expressed over some spatial extent, which the devices could determine by comparing their detections with previous events stored in their memories. By recording events over time (e.g., minutes to days), the sensors could collect information on changes (e.g., in response to an external stimulus such as introduction of a drug) that would be impractical to obtain from repeated blood samples. To examine average behavior of large numbers of robots, a natural approach for evaluating robot motion is via the stochastic approximation method discussed in Section 10.4.1. In this approximation, the same numerical software used to compute chemical concentration also provides estimates for the robot motion [23]. 10.5.1
Diagnostic Task Environment
As an example diagnostic task, consider a macroscopic tissue volume V containing a single microscopic source producing a particular chemical while the rest of the tissue has this chemical at much lower concentrations. This tissue volume contains a large number of blood vessels as illustrated in Figure 10.8. We focus on chemical detection in the small vessels, since they allow exchange of chemicals with surrounding tissue. Localization to volume V could be due, for example, to a distinctive chemical environment (e.g., high oxygen concentrations in the lungs) or an externally supplied signal (e.g., ultrasound) detectable by sensors passing through vessels within the volume. The devices are active only when they detect they are in the specified volume. Robots moving with fluid in the vessels will, for the most part, be in vessels containing only the background concentration, providing numerous opportunities for incorrectly interpreting background concentration as source signals. These false positives are spurious detections due to statistical fluctuations from the background concentration of the chemical. Although such detections can be rare for individual devices, when applied to tasks involving small sources in a large tissue volume, the
Figure 10.8 ing vessels.
Schematic illustration of the task geometry of a tissue volume with multiple branch-
302
Mobile Microscopic Sensors for In Vivo Diagnostics
number of opportunities for false positive responses can be orders of magnitude larger than the opportunities for true positive detections. Thus even a low false positive rate can lead to many more false positive detections than true positives. Figure 10.4 shows the task geometry with the gray area indicating a source on the vessel wall emitting a chemical into the moving fluid. In this example the robots have no locomotion capability and thus move passively with the fluid, continually entering the left side of the vessel with the fluid flow, at average speed vavg = 10–3m/s. This diagnostic scenario corresponds to detecting small areas of infection or injury. The chemicals arise from the initial immunological response at the injured area and enter nearby small blood vessels to recruit white blood cells [68]. A typical protein produced in response to injury has concentration near the injured tissue of about 30 ng/ml and background concentration in the bloodstream about 300 times smaller. These chemical concentrations are well above the demonstrated sensitivity of nanoscale chemical sensors [20, 21]. This example incorporates features relevant for medical applications: a chemical indicating a region of interest, diffusion into flowing fluid, and a background level of the chemical limiting sensor discrimination. As a quantitative example, we consider a chemical source of length Lsource = 30 μm producing the chemical with flux magnitude Fsource = 5.6×1013 molecule/s/m2 which gives concentration near source of Csource = 1.8×1018 molecule/m3. We take the background concentration of the chemical as Cbackground = 6×1015 molecule/m3. These parameters correspond to the typical concentrations mentioned above and give a challenging diagnostic task with relatively low concentration from a single cell-sized source. We consider 109 robots in the entire 5-liter blood volume of a typical adult, an example of medical applications using a huge number of microscopic robots [5]. These robots occupy only about 10-6 of the vessel volume, far less than the 20% to 40% occupied by blood cells. The total mass of all the robots is about 4 mg. 10.5.2
Control
The limited capabilities of the robots and the need to react on millisecond time scales require simple controls based on local information. For the chemical detection task, the key control design issue is a rule whereby a robot determines whether it passed a source. Ideally, this rule would be simple to compute (due to limited computational capability of the robot), effectively distinguish passing a source (true positive) from detection due to background concentration (false positive) and allow decision while the robot is still relatively near the source (e.g., for subsequent tasks requiring the robot to move to the source). The key observational distinction between background and the much higher source concentration is the difference in count rates from molecules of the target chemical reaching the robot. Due to statistical variation in the counts, the robot needs to accumulate counts over enough time to reliably distinguish true from false positives, but a short enough time that decisions are made while still close to the source. The choice of measurement time must balance having enough time to receive adequate counts, thereby reduce errors due to statistical fluctuations, while still responding before the robot has moved far downstream of the source where response would give poor localization of the source or, if robots are to take some action near the source, require moving upstream against the fluid flow.
10.5 Example Task: High-Resolution Diagnostics
303
Moreover, far downstream of the source the concentration from the target is small so additional measurement time is less useful. A simple decision criterion is if a sufficient number of detections occur in a short time interval. Quantitatively, such a rule considers a source detected if the robot observes at least Cthreshold counts in a measurement time interval Tmeasure. A device passing through a vessel with a source will have about Tmeasure = Lsource/vavg = 30ms
(10.4)
with high concentration, so a measurement time of roughly this magnitude or smaller allows making a decision while still fairly near the source. A low value for Cthreshold will produce many false positives, while a high value means many robots will pass the source without detecting it (i.e., false negatives). False negatives increase the time required for detection while false positives could lead to inappropriate subsequent activities (e.g., releasing a drug to treat the injury or infection at a location where it will not be effective). 10.5.3
Detection Performance
To quantify the performance of the robots, the detections are a Poisson random process. That is, during a time interval and chemical concentration where the average detection rate is r (e.g., as given by (10.2)), the probability to detect k molecules is the Poisson distribution e-rrk/k!. For example, with the parameters described above, the detection rate near the source is about 2,000 molecule/s while at the background concentration the rate is only about 8. Using a measurement time Tmeasure =10 ms, Figure 10.9 shows the probabilities of various counts. In this case, picking a threshold Cthreshold between 4 and 8 counts gives a tiny probability for detection at the background concentration and a substantial detection likelihood when passing near the source. This comparison has two caveats. First, as seen in Figure 10.2, the highest concentrations are only close to the vessel wall so robots passing near the center of the vessel will have fewer counts, suggesting a lower choice for Cthreshold. Second,
Figure 10.9 Probability distribution of number of detections during Tmeasure = 10 ms near the source and at the background concentration.
304
Mobile Microscopic Sensors for In Vivo Diagnostics
while the detection probability at the background concentration is tiny, there are many more opportunities for false positive detections than for true positives. False positives, in which a statistical fluctuation from background concentration causes a robot to incorrectly conclude it has passed a high-concentration source, could lead to a misdiagnosis of concluding a chemical source is present when it is not. False positives can be reduced by increasing the detection threshold Cthreshold. But if raised too high, robots will likely miss detecting the source when it does exist, a false negative error. Thus good diagnostic performance is only possible if there is a range of threshold choices high enough to avoid false positives but still low enough to avoid false negatives, with acceptably high probabilities. As long as the detection distributions between the source and background concentrations are well separated, as shown in Figure 10.9, suitable choices of Cthreshold give high probability for detecting a source if it exists without undue false positives [24]. Specifically, such control choices can have about 98% of the robots passing the source detect it, for a detection rate of about 10–2/s. By contrast, these control parameters give false positives at a rate of only about 10–11/s, in spite of the much larger number of opportunities for false positives compared to the single vessel with the source in a macroscopic tissue volume of about a cubic centimeter. This difference is due to the ability of robots to pass close to the source, where concentration is significantly higher than background. The difference in rates gives a wide range for the total time to run the diagnostic task with a high chance of true positive detection and low false positives. For example, operating for 300s gives 99% chance at least one robot detects the source if it is there (true positive) while the probability any robot makes a detection when there is no source is less than 10–8. In summary, simple control allows fast and accurate detection of even a single cell-sized source within a macroscopic tissue volume. The key feature enabling this performance is the robots’ ability to pass close to individual cells, where concentration from released chemicals is much higher than in fluid far from the cell. For comparison, instead of using microscopic sensors, one could use a blood sample extracted from the body, which is a routine medical procedure. Conventional laboratory analysis outside the body could attempt to detect the chemical in the sample. However, such a sample will represent the average concentration of the chemical in the full blood volume, leading to much lower concentration from the source, about 10–4 of the background concentration for this scenario [24]. Thus the additional chemical released by the source would be difficult to detect against small variations in background concentration. 10.5.4
Using the Diagnostic Information
For information gathering, each robot notes in its memory whenever chemicals matching a prespecified pattern are found. In this respect the robot behaves just as larger scale robots running programs and saving results in nonvolatile memory for later use, though with computation and memory based on nanoscale electronics [13]. The simplest application for this information is determining whether a chemical source with the pattern exists in the tissue. This provides diagnostic information for determining further treatments, either conventional therapies or actions taken with the nanorobots.
10.6 Discussion
305
As a more elaborate application of this information gathering task, the devices are retrieved and information in their memories extracted for further analysis in a conventional computer with far more computational resources than available to any individual microscopic robot. This computer would have access to information from many robots, allowing evaluation of aggregate properties of the population of cells that individual robots would not have access to, for example, the number of cells presenting a specific combination of chemicals. This information allows estimating spatial structure and strength of the chemical sources. Estimating the structure of the chemical sources from the microscopic sensor data is analogous to computerized tomography [69]. In tomography, the data consists of integrals of the quantity of interest (e.g., tissue density) over a large set of lines with known geometry selected by the experimenter. The microscopic sensors, on the other hand, record data points throughout the tissue, providing more information than just an aggregate value such as the total number of events. However, the precise path of each sensor through the tissue (i.e., which vessel branches it took and the precise locations of those vessels), will not be known. Figure 10.10 illustrates the challenge of determining source structure. In this case, the concentration profile with two sources is significantly different from that with a single source, shown in Figure 10.2. However, a robot moving with the fluid will pass the high concentration regions in a few tens of milliseconds. During this time the robot will detect, on average, only a few molecules, as estimated by (10.2). Thus statistical variations in the actual number of detections will blur the distinction between these two cases, requiring information from many robots passing through the vessels to distinguish these cases. Better performance requires either higher chemical concentrations or larger separation between the sources.
10.6
Discussion Plausible capabilities for microscopic robots suggest a range of novel applications in biomedical research and medicine. Sensing and acting with micron spatial resolu-
Figure 10.10 Diffusion of molecules from two 5-μm source regions on the wall of a two-dimensional channel, of width 10 μm. The black lines at the bottom of the plot indicate the two source regions. Parameters are as in Figure 10.2 but with twice the flux from each source so the total rate of production from the two smaller sources is the same as for the single larger source in Figure 10.2. Concentration ranges from about 1017 molecule/m3 near the sources (white) to zero (black).
306
Mobile Microscopic Sensors for In Vivo Diagnostics
tion and millisecond timing allows access to activities of individual cells. The large numbers of robots enable such activities simultaneously on a large population of cells in multicellular organisms. In particular, the small size of these robots allows access to tissue through blood vessels. Thus a device passes within a few tens of microns of essentially every cell in the tissue in a time ranging from seconds to hours, depending on the number of devices used. This access to cells allows rapid chemical sensing with much higher resolution than possible from conventional laboratory analysis of blood samples. Furthermore, the precision of localization and the robots’ programmability gives them a degree of flexibility to alter microenvironments (e.g., by releasing drugs), well beyond that possible with either large scale surgery or nonprogrammable chemically targeted drug delivery. The devices could also detect properties of cells outside but near the small vessels (e.g., electrical activity of nerves [30]). The full range of biomedical situations that could benefit from this flexibility remains to be seen. The diagnostic task described in this chapter highlights key control principles for microscopic robots. Specifically, by considering the overall task in a series of stages, the person deploying the robots remains in the decision loop, especially for the key decision of whether to proceed with manipulation (e.g., release a drug) based on diagnostic information reported by the devices. Information retrieved during treatment can also indicate how well the procedure is performing and provide high-resolution documentation of what was done to help improve future treatments. More generally, this hybrid control illustrates an important approach to using self-organized systems: use local, distributed control to achieve robust self-organized behaviors on small scales in space and time, combined with feedback from a slower, larger central control (e.g., a person) to verify performance and consider global constraints not easily incorporated within local controllers. Safety is important for medical applications of microscopic robots. Thus, evaluating a control protocol should consider its accuracy allowing for errors, failures of individual devices, or variations in environmental parameters. For the simple distributed sensing discussed in this chapter, statistical aggregation of many devices’ measurements provides robustness against these variations, a technique recently illustrated using DNA computing to respond to patterns of chemicals [50]. Furthermore, the devices must be compatible with their biological environment [70] for enough time to complete their task. Appropriate surface coatings should allow sufficient biocompatibility during robot operation [71, 72]. However, even if individual devices are inert, too many in the circulation could be harmful. Micron-sized particles occupying up to 10-3 of the vessel volume have been experimentally demonstrated to be safely tolerated in the circulatory system of at least some mammals [71]. This fraction is much larger than the 10-6 used with the example discussed above. Thus there is likely to be wide latitude in using enough robots for rapid diagnosis while remaining in safely tolerated limits. A further design challenge is the eventual removal or safe degradation of the robots after they complete their tasks. Various simulation tools are useful for modeling nanorobots at different levels of detail, from individual nanoscale components to behavior of groups of robots operating in microscopic environments. Specific tools allow examining behavior of electrical devices, sensors, and fluids [73, 74]. Simulating behavior of groups of robots, averaging over details of device behavior, can benefit from tools for solving
Problems
307
partial differential equations (e.g., for fluid flow and chemical diffusion) that readily incorporate interactions among different physical effects [75]. The latter approach was used to produce the figures in this chapter. Despite the simplifications used to model robot behavior, the estimates obtained in this chapter with plausible biophysical parameters show high-resolution sensing is possible with passive device motion in the circulatory system, even without communication capabilities. Thus relatively modest hardware capabilities could provide useful in vivo sensing capabilities far more flexible and specific than current technology. Research studies of tissue microenvironments with such robots will improve knowledge of their biophysical parameters, and hence enable better inferences from the data collected by these devices. The improved understanding will, in turn, indicate distributed controls suitable for more capable devices and appropriate trade-offs between scale and capability for hybrid systems combining coarse centralized control with the flexibility of self-organization within the biological microenvironments.
Problems 10.1 How rapidly will an absorbing sphere of radius 1μm collect molecules with diffusion coefficient D = 10–10 m2/s in a region with concentration C = 1018 molecule/m3? 10.2 If such a sphere has 10 ms to collect molecules, what is the probability it will detect none? 10.3 If a chemical with diffusion coefficient D = 10-10 m2/s is released, how long will it take to diffuse one micron? 100 microns? 10.4 Equation (10.1) shows why diffusion is slow over long distances. On the other hand, this equation indicates diffusion speed, x/t = 6D/x, grows arbitrarily large as distance and time decrease toward zero. How is this prediction reconciled with the limit on how fast physical objects can move? 10.5 The speed of sound in fluids such as water is about a kilometer per second. For robots 100-μm apart using sound to coordinate activities and moving through the fluid at about 1 mm/s, how far will robots move in the time they can communicate using ultrasound? How does this compare with the time required if they were communicating by releasing a chemical with diffusion coefficient D = 10–10 m2/s? 10.6 Demonstrated nanoscale electronics can perform basic computation operations using about 103kT per operation, where k = 1.4 × 10–23J/K is the Boltzmann constant and T is the temperature. How much power does this require to perform 104 operations per second? 10.7 A person has a few billion small blood vessels and the time required to complete one trip through the circulation (from the heart, through the body, and back to the heart) is about a minute. For a diagnostic task with 109 robots, how much time will the task require to have 99% probability for at least one robot to pass through each vessel?
308
Mobile Microscopic Sensors for In Vivo Diagnostics
References [1] K. Morris, “Macrodoctor, come meet the nanodoctors,” The Lancet, vol. 357, p. 778, March 10 2001. [2] NIH, “National Institutes of Health roadmap: Nanomedicine,” 2003. Available at nihroadmap.nih.gov/nanomedicine/index.asp. [3] B. L. Keszler, I. J. Majoros, and J. R. Baker Jr., “Molecular engineering in nanotechnology: Structure and composition of multifunctional devices for medical application,” in Proc. of the Ninth Foresight Conference on Molecular Nanotechnology, 2001. [4] T. Vo-Dinh, P. Kasili, and M. Wabuyele, “Nanoprobes and nanobiosensors for monitoring and imaging individual living cells,” Nanomedicine: Nanotechnology, Biology, and Medicine, vol. 2, pp. 22-30, 2006. [5] R. A. Freitas Jr., Nanomedicine, vol. I: Basic Capabilities. Georgetown, TX: Landes Bioscience, 1999. Available at www.nanomedicine.com/NMI.htm. [6] E. Andrianantoandro, S. Basu, D. K. Karig, and R. Weiss, “Synthetic biology: New engineering rules for an emerging discipline,” Molecular Systems Biology, vol. 2, no. msb4100073, pp. E1-E14, 2006. [7] A. Barreiro et al., “Subnanometer motion of cargos driven by thermal gradients along carbon nanotubes,” Science, vol. 320, pp. 775-778, 2008. [8] C. P. Collier et al., “Electronically configurable molecular-based logic gates,” Science, vol. 285, pp. 391- 394, 1999. [9] H. G. Craighead, “Nanoelectromechanical systems,” Science, vol. 290, pp. 1532-1535, 2000. [10] J. Howard, “Molecular motors: Structural adaptations to cellular functions,” Nature, vol. 389, pp. 561- 567, 1997. [11] J. Fritz et al., “Translating biomolecular recognition into nanomechanics,” Science, vol. 288, pp. 316- 318, 2000. [12] C. Montemagno and G. Bachand, “Constructing nanomechanical devices powered by biomolecular motors,” Nanotechnology, vol. 10, pp. 225-231, 1999. [13] S.-Y. Wang and R. S. Williams, eds., Nanoelectronics, vol. 80, (New York), Springer, March 2005. Special issue of Applied Physics A. [14] S. K. Kufer et al., “Single-molecule cut-and-paste surface assembly,” Science, vol. 319, pp. 594-596, 2008. [15] C. W. McCurdy et al., “Theory and modeling in nanoscience,” workshop report, www.science.doe.gov/bes/reports/files/tmn rpt.pdf, US Dept. of Energy, 2002. [16] J. Lahann and R. Langer, “Smart materials with dynamically controllable surfaces,” MRS Bulletin, vol. 30, pp. 185-188, March 2005. [17] C. R. Martin and Z. S. Siwy, “Learning nature’s way: Biosensing with synthetic nanopores,” Science, vol. 317, pp. 331-332, 2007. [18] H. C. Berg and E. M. Purcell, “Physics of chemoreception,” Biophysical Journal, vol. 20, pp. 193-219, 1977. [19] Z. Li et al., “Silicon nanowires for sequence-specific DNA sensing: Device fabrication and simulation,” Applied Physics A, vol. 80, pp. 1257-1263, 2005. [20] F. Patolsky and C. M. Lieber, “Nanowire nanosensors,” Materials Today, vol. 8, pp. 20-28, April 2005. [21] P. E. Sheehan and L. J. Whitman, “Detection limits for nanoscale biosensors,” Nano Letters, vol. 5, pp. 803-807, 2005. [22] H. C. Berg, Random Walks in Biology. Princeton Univ. Press, 2nd ed., 1993. [23] T. Hogg, “Coordinating microscopic robots in viscous fluids,” Autonomous Agents and Multi-Agent Systems, vol. 14, no. 3, pp. 271-305, 2007. [24] T. Hogg and P. J. Kuekes, “Mobile microscopic sensors for high-resolution in vivo diagnostics,” Nanomedicine: Nanotechnology, Biology, and Medicine, vol. 2, pp. 239-247, 2006.
References
309
[25] K. H. Keller, “Effect of fluid shear on mass transport in flowing blood,” in Proc. of Federation of American Societies for Experimental Biology, pp. 1591-1599, Sept.-Oct. 1971. [26] S. Ghosh et al., “Carbon nanotube flow sensors,” Science, vol. 299, pp. 1042-1044, 2003. [27] T. M. Squires and S. R. Quake, “Microfluidics: Fluid physics at the nanoliter scale,” Reviews of Modern Physics, vol. 77, pp. 977-1026, 2005. [28] H. F. Dvorak, “How tumors make bad blood vessels and stroma,” American J. of Pathology, vol. 162, pp. 1747-1757, 2003. [29] K. M. Tyner, R. Kopelman, and M. A. Philbert, “Nano-sized voltmeter enables cellular-wide electric field mapping,” Biophysical Journal, vol. 93, pp. 1-12, 2007. [30] R. R. Llinas et al., “Neuro-vascular central nervous recording/stimulating system: Using nanotechnology probes,” J. of Nanoparticle Research, vol. 7, pp. 111-127, 2005. [31] S. E. Cross, Y.-S. Jin, J. Rao, and J. K. Gimzewski, “Nanomechanical analysis of cells from cancer patients,” Nature Nanotechnology, vol. 2, pp. 780-783, 2007. [32] H. Bojinov, A. Casal, and T. Hogg, “Multiagent control of modular self-reconfigurable robots,” Artificial Intelligence, vol. 142, pp. 99-120, 2002. arxiv.org preprint cs.RO/0006030. [33] D. Rus and M. Vona, “Self-reconfiguration planning with compressible unit modules,” in Proc. of the Conference on Robotics and Automation (ICRA99), vol. 4, pp. 2513-2520, IEEE, 1999. [34] B. Salemi, W.-M. Shen, and P. Will, “Hormone controlled metamorphic robots,” in Proc. of the Intl. Conf. on Robotics and Automation (ICRA2001), vol. 4, pp. 4194-4199, IEEE, 2001. [35] H. Wang et al., “In vitro and in vivo two-photon luminescence imaging of single gold nanorods,” Proc. of the Natl. Acad. Sci. USA, vol. 102, pp. 15752-15756, 2005. [36] J. Liu et al., “Nanoparticles as image enhancing agents for ultrasonography,” Physics in Medicine and Biology, vol. 51, pp. 2179-2189, 2006. [37] R. F. Service, “Nanotechnology takes aim at cancer,”Science, vol. 310, pp. 1132-1134, 2005. [38] T. Q. Vu et al., “Activation of membrane receptors by a neurotransmitter conjugate designed for surface attachment,” Biomaterials, vol. 26, pp. 1895-1903, 2005. [39] E. M. Purcell, “Life at low Reynolds number,” American Journal of Physics, vol. 45, pp. 3-11, 1977. [40] S. Vogel, Life in Moving Fluids. Princeton Univ. Press, 2nd ed., 1994. [41] Y. C. Fung, Biomechanics: Circulation. NY: Springer, 2nd ed., 1997. [42] G. Karniadakis, A. Beskok, and N. Aluru, Microflows and Nanoflows: Fundamentals and Simulation. Berlin: Springer, 2005. [43] J. P. Hernandez-Ortiz, C. G. Stoltz, and M. D. Graham, “Transport and collective dynamics in suspensions of confined swimming particles,” Physical Review Letters, vol. 95, p. 204501, 2005. [44] I. H. Riedel et al., “A self-organized vortex array of hydrodynamically entrained sperm cells,” Science, vol. 309, pp. 300-303, 2005. [45] M. G. L. van den Heuvel and C. Dekker, “Motor proteins at work for nanotechnology,” Science, vol. 317, pp. 333-336, 2007. [46] S. Martel, “The coming invasion of the medical nanorobots,” Nanotechnology Perceptions, vol. 3, pp. 165-173, 2007. [47] X. Wang et al., “Direct-current nanogenerator driven by ultrasonic waves,” Science, vol. 316, pp. 102- 105, 2007. [48] R. Tucker, P. Katira, and H. Hess, “Hearding nanotransporters: Localized activation via release and sequestration of control molecules,” Nano Letters, vol. 8, pp. 221-226, 2008. [49] R. A. Freitas Jr., “Pharmacytes: An ideal vehicle for targeted drug delivery,” Journal of Nanoscience and Nanotechnology, vol. 6, pp. 2769-2775, 2006.
310
Mobile Microscopic Sensors for In Vivo Diagnostics [50] Y. Benenson, B. Gil, U. Ben-Dor, R. Adar, and E. Shapiro, “An autonomous molecular computer for logical control of gene expression,” Nature, vol. 429, pp. 423-429, 2004. [51] S. P. Leary, C. Y. Liu, and M. L. J. Apuzzo, “Toward the emergence of nanoneurosurgery: Part iii-nanomedicine: Targeted nanotherapy, nanosurgery, and progress toward the realization of nanoneurosurgery,” Neurosurgery, vol. 58, pp. 1009-1026, 2006. [52] D. Sretavan, W. Chang, C. Keller, and M. Kliot, “Microscale surgery on axons for nerve injury treatment,” Neurosurgery, vol. 57, no. 4, pp. 635-646, 2005. [53] T. Hogg and D. W. Sretavan, “Controlling tiny multi-scale robots for nerve repair,” in Proc. of the 20th Natl. Conf. on Artificial Intelligence (AAAI2005), pp. 1286-1291, AAAI Press, 2005. [54] R. Dreyfus et al., “Microscopic artificial swimmers,” Nature, vol. 437, pp. 862-865, 2005. [55] J. Halloy et al., “Social integration of robots into groups of cockroaches to control self-organized choices,” Science, vol. 318, pp. 1155-1158, 2007. [56] D. Arbuckle and A. A. G. Requicha, “Active self-assembly,” in Proc. of the IEEE Intl. Conf. on Robotics and Automation, pp. 896-901, 2004. [57] G. M. Whitesides and B. Grzybowski, “Self-assembly at all scales,” Science, vol. 295, pp. 2418-2421, 2002. [58] S. Griffith, D. Goldwater, and J. M. Jacobson, “Robotics: Self-replication from random parts,” Nature, vol. 437, p. 636, 2005. [59] E. Bonabeau, M. Dorigo, and G. Theraulaz, Swarm Intelligence: From Natural to Artificial Systems. Oxford: Oxford University Press, 1999. [60] V. Gazi and K. M. Passino, “Stability analysis of social foraging swarms,” IEEE Trans. on Systems, Man and Cybernetics, vol. B34, pp. 539-557, 2004. [61] A. Dhariwal, G. S. Sukhatme, and A. A. G. Requicha, “Bacterium-inspired robots for environmental monitoring,” in Proc. of the IEEE Intl. Conf. on Robotics and Automation, pp. 1436-1443, 2004. [62] A. Cavalcanti and R. A. Freitas Jr., “Autonomous multi-robot sensor-based cooperation for nanomedicine,” Intl. J. of Nonlinear Sciences and Numerical Simulation, vol. 3, pp. 743-746, 2002. [63] K. Lerman et al., “A macroscopic analytical model of collaboration in distributed robotic systems,” Artificial Life, vol. 7, pp. 375-393, 2001. [64] A. Galstyan, T. Hogg, and K. Lerman, “Modeling and mathematical analysis of swarms of microscopic robots,” in Proc. of the IEEE Swarm Intelligence Symposium (SIS2005) (P. Arabshahi and A. Martinoli, eds.), pp. 201-208, 2005. [65] W. G. Strang and G. J. Fix, An Analysis of the Finite Element Method. Englewood Cliffs: Prentice-Hall, 1973. [66] I. E. Vignon and C. A. Taylor, “Outflow boundary conditions for one-dimensional finite-element modeling of blood flow and pressure waves in arteries,” Wave Motion, vol. 39, pp. 361-374, 2004. [67] T. M. Allen and P. R. Cullis, “Drug delivery systems: Entering the mainstream,” Science, vol. 303, pp. 1818-1822, 2004. [68] C. A. Janeway et al., Immunobiology: The Immune System in Health and Disease. New York: Garland, 5th ed., 2001. [69] F. Natterer, The Mathematics of Computerized Tomography. Philadelphia: Soc. for Industrial and Applied Math (SIAM), 2001. [70] A. Nel et al., “Toxic potential of materials at the nanolevel,” Science, vol. 311, pp. 622-627, 2006. [71] R. A. Freitas Jr., Nanomedicine, vol. IIA: Biocompatibility. Georgetown, TX: Landes Bioscience, 2003. Available at www.nanomedicine.com/NMIIA.htm. [72] A. M. Schrand et al., “Are diamond nanoparticles cytotoxic?,”Journal of Physical Chemistry B, vol. 111, pp. 2-7, 2007.
References
311
[73] Network for Computational Nanotechnology, “Nanohub: Online simulation and more.”www.nanohub.org, 2008. [74] “National nanotechnology infrastructure network.” www.nnin.org, 2008. [75] “Comsol multiphysics.” comsol.com, 2008.
CHAPTER 11
Microcantilever Biomedical Sensors Ram H. Datar, Thomas L. Ferrell, and Thomas Thundat
11.1
The Microcantilever Platform Microcantilever sensors have attracted considerable attention over the last decade for their potential use as exquisitely sensitive chemical and biological sensors [1, 2]. Microcantilever sensors also offer unprecedented opportunity for developing miniature biosensors [3–8]. The microcantilever platform integrates nanoscale science and microfabrication technology for label-free biochemical detection, allowing for miniaturization and low power consumption. When confined to a single side, biomolecular adsorption on the surface of deformable micromechanical structures such as microfabricated cantilever beams causes cantilever bending through nanometer deflection (see Figure 11.1). This nanoscale deflection occurs due to variation in cantilever surface stress caused by biomolecular interactions, and can be measured by optical or electrical means, thereby reporting on the presence of biomolecules. Chemical and biological specificity in detection is achieved by immobilizing a selective layer of specific capture ligands on one side of the cantilever, via a chemical linking process called surface functionalization. When exposed to the target analyte in a fluid, the functionalized cantilever deflects due to selective adsorption of target molecules.
Target Molecule Molecu Probe Molecule Coating Substrate Substrate
Cantilever
Target Binding Substrate Substrate
Deflection, Dh
Figure 11.1 (Color plate 17) Specific biomolecular interactions between target and probe molecules alter intermolecular interactions within a self-assembled monolayer on one side of a cantilever beam. This can produce a sufficiently large surface stress to bend the cantilever beam and generate motion.
313
314
Microcantilever Biomedical Sensors
The signal transduction process between the interacting biomolecules and the sensor elements is critical in designing an ideal biosensor. For example, glucose sensors typically use an electrochemical signal transduction mechanism involving the enzyme glucose oxidase and a suitable electrode for detecting either oxygen or hydrogen peroxide. These enzymatically catalyzed reaction products are directly proportional to glucose levels. Similar schemes can be developed for detecting reactions involving antibody-antigens or complementary nucleic acids. Currently available detection systems like ELISA for protein detection or hybridization for nucleic acid detection require the use of specifically labeled expensive reagents such as fluorescently labeled antibodies or probe molecules. Direct interpretation, without the use of additional labels, of any biochemical interaction by optical or electrical detection using the cantilever would be the ideal approach for biosensors. Microcantilever-based biosensing can be implemented in several formats. It can be accomplished by exploiting adsorption-induced stress as alluded to above, or by measuring the resonance frequency variation of the mechanical structures modified with chemically selective molecules. It should be noted, however, that resonance frequency based biosensors operating under solution have poor sensitivity due to liquid damping. Adsorption-induced cantilever bending is ideal for liquid-based applications, with sensitivity that is orders of magnitude higher than resonance frequency vibration-based cantilevers. In another format, these structures can be coated with a metal layer or integrated with a temperature-sensitive device to create a calorimetric sensor. Here we will limit our discussion to the adsorption-induced bending mode of cantilever operation. While the cantilevers are primarily mechanical microstructures, they can be imparted with electrical read-out capability by embedding a piezoelectric or piezoresistive element within their surface, or an optical read-out capability by exploiting a laser beam reflected off the free end of the cantilever and measured using a position sensitive detector. Another readout option is to measure the variation in cantilever capacitance, but this mode is unsuitable for liquid-based applications. The piezoelectric approach is more suited for the resonance frequency based detection method. The most common readout technique for cantilever deflection is the optical beam deflection technique. Interferometry optics has also been adapted to read out cantilever motion, where the deflection is detected by movement of a sensor cantilever relative to a reference cantilever or the chip substrate. Another attractive readout technique is based on piezoresistivity, whereby the bulk electrical resistivity varies with applied stress. Thaysen et al. [9] developed piezoresistive cantilever sensors with integrated differential readout. Each cantilever has a thin fully encapsulated resistor made of doped Si fabricated on top, the resistance of which changes due to any load on the cantilever. Each sensor element is comprised of a measurement cantilever and a built-in reference cantilever, which enables differential signal readout. The two cantilevers are connected in a Wheatstone bridge and the surface-stress change on the measurement cantilever is detected as the output voltage from the Wheatstone bridge. The researchers later applied the sensor for DNA sensing [10]. The adsorption-induced bending of the cantilever is often converted into surface stress using Stoney’s equation [11], σ = [hE / (1 - ν)](d / L) 2 , where h is cantilever deflection, E and ν are the elastic modulus and Poisson’s ratio
11.2 Value of Biosensors in Cancer Diagnostics and Prognostication
315
of the cantilever material, and d and L are the cantilever thickness and length respectively. The sensor was determined to have approximately 5 mJ/m2 minimum detectable surface-stress change. The electrical readout technique has several advantages over the optical beam deflection method. For example, the optical beam deflection method probes the bending of the free end of the cantilever. It is assumed that the cantilever bending is uniform along the length of the cantilever. The piezoresistive method, on the other hand, measures the integrated bending of the cantilever. Piezoresistive cantilevers can be encapsulated in silicon nitride for operation under solution thus avoiding the longstanding problems associated with optical path lengths and refractive index variation. In addition, because no external optics components are required, the electronic readout is more amiable for miniaturization, and ideal for portable devices. Electronic readout is compatible with array arrangements since both cantilevers and readout circuits can be fabricated simultaneously on the same chip. However, currently available piezoresistive cantilever sensors are an order of magnitude less sensitive than those using optical readout techniques. This discrepancy in sensitivity, however, is vanishing fast due to the recent progress in piezoresistive cantilever development.
11.2
Value of Biosensors in Cancer Diagnostics and Prognostication Although many cancers can be treated and cured if they are diagnosed while the tumors are still localized, a significant proportion of cancers are not detected until after they have invaded the surrounding tissue or metastasized to distant sites. For example, only about 50% of breast cancers, 56% of prostate cancers, and 35% of colorectal cancers are localized at the time of diagnosis [12]. Serum is a key source of putative protein and nucleic acid biomarkers, and, by its nature, can reflect organ-confined events. The development of tumor-associated serum protein markers has greatly facilitated the clinical management of some types of cancer, but very few biomarkers are effective for diagnosing primary cancer singly (for example, NMP22 for bladder cancer and PSA for prostate cancer). Some other serum markers have been approved by the FDA for cancer diagnosis, such as alpha-fetoprotein for hepatocellular carcinoma and testicular cancer, catecholamines for neuroblastoma, and immunoglobulins for multiple myeloma. Because the current assays for these proteins are neither sensitive nor specific enough for use as the sole screening method for early cancer detection, all are used as an adjunct to other direct detection and diagnostic methods. Other tumor-associated serum proteins are used clinically as markers to monitor therapy or recurrence, notably, carcinoembryonic antigen (CEA) for colorectal cancer; HER–2/ neu, CA 15-3, and CA 27-29 for advanced breast cancer; and CA 125 for ovarian cancer [13–15]. The complex nature of the molecular, cellular, and clinical information needed for better cancer diagnosis and therapeutic monitoring suggests that a single parameter may not answer all of the critical questions [16, 17]. Instead, a panel of individual tests and measurements that could be used in sum could provide the needed information. Cancer diagnostics has therefore begun to shift from traditional single markers to molecular “signatures.” This is based on the premise that the more dis-
316
Microcantilever Biomedical Sensors
crete the data points used in sum, the greater the ability for the collection of data to describe the biology of a tumor [18]. Ideally, these measurements would be done in a single readout and on material that is easily accessible, such as serum. As mentioned above, single biomarkers are shown to have limited value in cancer screening due to their lack of sensitivity and specificity. This issue becomes particularly relevant for screening of cancers of low prevalence. Extending the same logic further for sensing technology, the ability to sense multiple biomarkers sensitively while achieving savings in both sample biosource as well as reagent costs, will be highly desirable. Biosensing technologies such as cantilever array platforms offer an effective resource in this regard.
11.3
Cantilever Preparation Generally cantilevers are coated on one side with 2 to 3 nm of chromium then with 25 to 30 nm of gold using an e-beam evaporator. Chromium acts as an adhesion layer for the gold. Many approaches can be used to immobilize the molecular recognition agents to the microcantilever sensor, depending upon the final application. We have utilized both the silicon side of the cantilever (using silane chemistry) and the gold side of the cantilever (using thiol chemistry) depending on the final application for the molecular recognition assay. For thiol-self-assembled monolayers (SAM) and organosilane modification, dip coating is the preferred method for functionalization to allow for high-density immobilization on the cantilever surface; all reactive surfaces of the cantilever and substrate that are exposed to the modifying solution(s) will have a coating. Thiol SAMs are self-limited to coverages of a monolayer or less of the thiol on a gold film. Organosilane coatings also are of the order of a monolayer but can become multilayered upon extended exposure to the solution. Regardless of coating chemistry employed, typically all experimental surfaces are freshly prepared no more than 48 hours prior to assay. Stability studies to determine the effects of aging on the prepared surfaces remain to be done.
11.4
Biomolecular Detection Assays The microcantilever arrays enable multiplexed label-free analysis for various biomolecular reactions. In this section, we summarize research by us and others on detecting specific biomolecular interactions such as DNA hybridization and antibody-antigen bindings [4, 19, 20]. Detection of specific DNA sequences using the microcantilever platform has been demonstrated by several groups [4–7]. In these experiments, single-stranded DNA (ssDNA) molecules are immobilized on cantilevers with a thin layer (25 nm) of gold on one side using gold-thiol chemistry. Single stranded DNA molecules with a thiol linker at one end serve as the probe (or receptor) molecule for the target complementary strands. When the functionalized cantilevers were exposed to a solution containing single stranded DNA molecules of complementary sequence, the cantilever underwent deflection due to DNA hybridization. Such specific deflection was not seen when the incoming DNA strands were noncomplementary.
11.4 Biomolecular Detection Assays
317
Although both DNA hybridization and protein-protein (antigen-antibody) binding can be detected using cantilever deflections, what remained unclear for a while was whether this technique has sufficient specificity and sensitivity to be used for the detection of disease-related proteins at clinically relevant conditions and concentrations. To address this technologically critical issue, sensitive and specific detection of a prostate cancer marker, prostate specific antigen (PSA), was conducted as an example of both protein-protein binding in general, and of cancer diagnostic tumor marker detection in particular. We developed a sensitive and specific assay for prostate-specific antigen (PSA) using a gold-covered silicon microcantilever coated with a polyclonal anti-PSA antibody. Changes in the surface stress caused by the binding of PSA and its monoclonal antibody used to functionalize the cantilever surface resulted in a deflection of the cantilever, which could be detected optically using a low-power laser. Furthermore, changes in surface stress were related quantitatively to the concentration of PSA. The technique is simpler and potentially more cost-effective than ELISA, the current “gold standard” assay for PSA detection, because it does not require labeling and can be performed in a single reaction without additional reagents. The following section briefly describes development of this assay. 11.4.1
Detection of PSA
Prostate cancer has emerged as the most common nonskin cancer and the second leading cause of cancer death in men in North America and Europe. While transrectal ultrasonography (TRUS) and digital rectal examination (DRE) are the common clinical examinations, the most widely used biochemical test involves analyzing the presence of prostate specific antigen (PSA). PSA is a 33-kD serine protease secreted by prostatic luminal epithelial cells. Use in population screening of the detection of elevated serum PSA is credited with dramatic advances in the early diagnosis and management of men with prostatic carcinoma. The majority of the recently marketed assays are based on the commonly used reference range (< 4 ng/ml), and almost all of them employ some variation of the technique of enzyme-linked immunosorbent assay (ELISA). Antigen-antibody interactions are a class of highly specific protein-protein binding that play a critical role in molecular biology. When antibody molecules were immobilized to one surface of a cantilever, specific binding between antigens produced cantilever deflection. Using a model solvent system prepared with phosphate buffered saline, we demonstrated specific cantilever deflection as a function of time for quantitative detection of dissolved PSA against a much higher background of bovine serum albumin (BSA). Similar tests were performed against high backgrounds of human serum albumin (HSA) and human plasminogen, both of which are found abundantly in human sera. Of note was the finding that PSA concentrations can be detected below 4 ng/ml, the clinical threshold for prostate cancer. In fact, we could detect concentrations down to 0.2 ng/ml. Since for the same PSA concentrations, cantilever deflections varied with their geometry, it is important to standardize these measurements in terms of surface stress using Stoney’s formula rather than cantilever deflections.
318
Microcantilever Biomedical Sensors
Since cantilever motion originates from the free energy change induced by specific biomolecular binding, this technique offers a common platform for high-throughput, label-free analysis of protein-protein binding, DNA hybridization, and DNA-protein interactions, as well as drug discovery. Hence, one can build an array of cantilevers such that each array is specialized to detect either one specific molecule from multiple biosamples microfluidically flowed into each of the array cells, or a variety of different molecules from the same biosample. Similarly, one can design and fabricate an array of microcantilevers that will enable simultaneous experiments with hundreds of DNA sequences. Such multiplexed detection capability can permit the examination of intricate biological machinery. The potential advantages of a label-free assay, which can measure multiple analytes in a single step, are enormous and could ultimately translate into much lower cost per test. For example, this technique could enable simultaneous detection of multiple serum tumor markers from the same patient sample, making population screening, which is currently cost-prohibitive, more generally available. The ability to vary the length and thickness of cantilevers to increase their sensitivity can enable both high resolution and a high dynamic range, as exemplified in our study of PSA detection [4]. We note here that despite being able to detect PSA at the current limit of ELISA (0.2 ng/ml), there is room for further improvement in sensitivity of cantilevers, either through controlling the roughness of the gold surface or by controlling the surface density of probe molecules. The label-free approach makes microcantilever sensors particularly attractive for drug discovery, which requires one to detect specific binding between numerous candidate small molecules with proteins. Similar detection studies have been carried out for cancer biomarkers CEA (carcinoembryonic antigen) and b-hCG (human chorionic gonadotropin beta subunit). Figure 11.2 shows the surface stress-concentration plot for b-hCG. The response shows a Langmuir type adsorption behavior for lower concentration and a linear response for higher concentrations. The observed linear behavior is most probably an artifact due to high concentration. The response at lower concentration can be used for quantitative analysis.
11.5
Implantable Sensors Compared to lab-based assays based on cantilevers, development of implantable sensors still remains a challenge due to many current technological limitations. For example, at present there are no highly selective receptors that can work reliably and reversibly in vivo. Therefore, regenerating the sensor device after detection, and keeping the efficacy of the receptors for extended periods of time in vivo are also major issues that need to be addressed. Also, potential biofouling of the sensor surface and the biocompatibility of the chemical interface for target analyte adsorption are formidable problems. Despite all these challenges, short term implantable sensors using microcantilever platforms have been demonstrated by Cheney et al. [21]. These sensors are developed for detection of much simpler analytes such as blood gases such as alcohol, and physical variables such as temperature. Piezoresistive microcantilever arrays are an ideal platform for developing implantable sensors due to their miniature size, low-power consumption, and high
11.6 Conclusion
319
18 16 14 12 10 8 6 4 2 0
0
100 200 300 400 500 600 700 800
Figure 11.2 Surface stress (Y axis) versus concentration of b-hCG in ng/ml (X axis). The variation is surface stress as a function of concentration (lower concentrations) can be used for quantitative analysis of b-hCG adsoprtion on cantilever surface.
sensitivity. The signal of the cantilever bending can be measured electronically by a Wheatstone bridge and transmitted in real-time by telemetry. Cheney et al. have demonstrated such an implanted sensor system for detecting blood alcohol levels in rats [21]. The cantilevers were coated with methyl phenyl mercapto propyl silicone (OV17 MCP20) as a reversible chemical interface for ethanol detection. The cantilever chip, associated electronic components, and batteries were placed in a capsule (see Figure 11.3), with a hydrophobic vapor permeable membrane covering a small opening over cantilevers for vapor detection. The capsule implanted subcutaneously allows blood gases to reach the cantilever array through the hydrophobic membrane. The polymer OV17 MCP20 has a high partition coefficient for alcohol vapors in the presence of high humidity. The presence of alcohol in the bloodstream causes the cantilevers to bend in proportion to the alcohol concentration. In the actual experiments the capsule was surgically implanted subcutaneously in Wistar rats, which after recovery were injected with alcohol solution. Figure 11.4 shows the differential cantilever response as a function of time when the Wistar rat was injected intravenously with alcohol solution.
11.6
Conclusion Experimental and theoretical research has shown that when biomolecular reactions are confined to a single side of a cantilever, they result in cantilever deflection. The cantilever deflection is attributed to an increase in strain energy that is needed to compensate the decrease in free energy due to molecular adsorption. Therefore, the bending of a cantilever provides a quantitative measure of the free energy density of a surface reaction. Since free energy reduction is the common driving force for all biochemical reactions, cantilevers form a universal sensor platform for biomolecular interactions without needing any label for signal detection. Bimaterial microcantilever surfaces can be functionalized with ease using a number of different
320
Microcantilever Biomedical Sensors
Figure 11.3 Views of closed and open implantable telesensor capsule shown alongside a U.S. quarter coin. A titanium frit is seen in the upper portion as the circular membrane on the left side of the closed capsule. Under the frit is a membrane filter for partial separation of the analyte from the interstitial fluid. This covers two arrays of four piezoresistive microcantilever sensors that are connected in a set of Wheatstone bridges. The signals are delivered to a sigma-delta ADC and thence to a very low-power µprocessor. This is programmed to allow modulation of the transmissions to a nearby receiver/computer combination where the data are displayed graphically. In the present embodiment, the telesensor capsule can transmit both chemical data and body temperature every five minutes for six weeks (longer for lesser duty cycles).
chemistries, enhancing their ability to serve as universal sensor surfaces for different biological macromolecules as well as small molecules. In addition, a variety of detection modes like optical, piezoresistive, piezoelectric, capacitance, and calorimetry are available. Such detection has been successfully shown for biomolecules-like proteins and DNA sequences. This label-free approach can be multiplexed to study many reactions simultaneously. Since the readout of cantilever bending using piezoresitive technique is compact, it is possible to develop miniature sensor systems that are implantable. Such a subcutaneously implanted sensor system for detecting vapor phase analytes such as alcohol in the blood of laboratory rats has already been demonstrated. With advancements in sensor interface design, it will be possible in the future to detect many other anlytes such as cancer and cardiac markers using implanted sensors. However, more work is needed in developing reversible and robust chemical and biological interfaces that can be immobilized on these cantilever sensors. In summary, microcantilever array sensors detect biomolecular interactions with high specificity and sensitivity in a multiplexed, label-free fashion, by virtue of
Problems
321
200
1.00
150
0.95
100 0.90 50 0.85 0 0
1
2
4 3 Time (h)
5
6
7
Figure 11.4 Continuous line shows differential piezoresistive cantilever (implanted sensor) response to alcohol administered intravenously to a Wistar rat. Discrete points show blood alcohol concentration (BEC) values obtained with traditional methods (Y-axis right side). Arrows show two separate intraperitoneal injection of 20% alcohol at a dose of 2.5g ethanol/kg of body weight. (Adapted from [21].)
nanoscale deflections. The cantilever deflections are measurable in a number of different optional ways including electrical and optical readouts. Finally, the cantilever technology offers an attractive, low-cost alternative to the current assay methods of biomolecular detection, making population screening for disease susceptibility a potential reality as well as enhancing development of implantable sensor devices.
Problems 11.1 What is the basic “sensing” mechanism in a microcantilever sensor? State some of the advantages of the microcantilever-based sensing approach. 11.2 Outline various alternative readout options for microcantilever sensors, and the merits and demerits of some of these approaches. 11.3 Briefly describe the principle and the design of piezoresistive cantilever sensor. 11.4 What are some of the challenges in development of implantable sensors?
References [1] Thundat, T., Oden, P. I., and Warmack, R. J., “Microcantilever Sensors,” Microscale Thermophys. Eng., Vol. 1, 1997, p. 185.
322
Microcantilever Biomedical Sensors [2] Dutta, P., Chapman, P. J., Datskos, P. G., and Sepaniak, M. J., Characterization of Ligand-Functionalized Microcantilevers for Metal Ion Sensing,” Anal Chem, Vol. 77, No. 20, 2005, pp. 6601–6608. [3] Raiteri, R., Nelles, G., Butt, H.-J., Knoll, W., and Skládal, P., “Sensing Biological Substances Based on The Bending of Microfabricated Cantilevers,” Sensors Actuators B, Vol. 61, 1999, p. 213. [4] Fritz, J., Baller, M .K., Lang, H. P., Rothuizen, H., Vettiger, P., Meyer, E., Güntherodt, H. J., Gerber, C., and Gimzewski, J. K., “Translating Biomolecular Recognition into Nanomechanics,” Science, Vol. 288, 2001, p. 316. Hansen, K., Ji, H., Wu, G., Datar, R., Cote, R., and Majumdar, A., “Cantilever-Based [5] Optical Deflection Assay for Discrimination of DNA Single-Nucleotide Mismatches,” Analytical Chemistry, Vol. 73, 2001, p. 1567. [6] Wu, G., Ji, H., Hansen, K., Thundat, T., Datar, R., Cote, R., Hagan, M. F., Chakraborty, A. K., and Majumdar, A, “Origin of nanomechanical Cantilever Motion Generated from Biomolecular Interactions,” Proceedings of National Academy of Science,Vol. 98, 2001, p. 1560. [7] Wu, G., Datar, R., Hansen, K., Thundat, T., Cote, R., and Majumdar, A, “Bioassay of Prostate Specific Antigen (PSA) Using Microcantilevers,” Nature Biotechnology, Vol. 19, 2001, p. 856. [8] McKendry, R., Zhang, J., Arntz, Y., Strunz, T., Hegner, M., Lang, H. P., Baller, M. K., Certa, U., Meyer, E., Guntherodt, H. –J., and Gerber, C., “Multiple label-Free Biodetection and Quantitative DNA-Binding Assays on a Nanomechanical Cantilever Array,” Proceedings of National Academy of Science, Vol. 99, 2002, p. 9783. [9] Thaysen, J., Boisen, A., Hansen, O., and Bouwstra, S., “Atomic Force Microscopy Probe with Piezoresistive Read-Out and a Highly Symmetrical Wheatstone Bridge Arrangement,” Sensors and Actuators A, Vol. 83, 2000, p. 47. [10] Marie, R., Jensenius, H., Thaysen, J., Christensen, C. B., and Boisen, A., “Adsorption Kinetics And Mechanical Properties of Thiol-Modified DNA-Oligos on Gold Investigated by Microcantilever Sensors,” Ultramicroscopy, Vol. 91, 2002, p. 29. [11] Stoney, G. G., “The Tension of Metallic Films Deposited by Electrolysis,” Proc. Roy. Soc. Lond. A, Vol. 82, 1909, p. 172. PDQ Cancer Information Summaries, http://www.can[12] Cancer-Net cer.gov/cancertopics/pdq/screening/overview/healthprofessional#Section_27. [13] Diamandis, E. P., “Prognostic Markers in Breast Cancer,” Clin Lab News, Vol. 22, 1996, pp. 235–239. [14] Stein, J. P, Grossfeld, D.G., Ginsberg, D. A., et al., “Prognostic Markers in Bladder Cancer: A Contemporary Review of the Literature,” J Urol, Vol. 160, No.3, Part 1, 1998, pp. 645–659. [15] Buzdar, A. U., and Hortobagyi, G. N., “Breast Cancer,” in Cancer Chemotherapy and Biological Response Modifiers Annual 18, Pinedo, H. M., Longo, D. L., and Chabner, B. A., (eds.), Amsterdam: Elsevier Science BV, 1999, pp. 435–469. [16] Etzioni, R., Urban, N., Ramsey, S., McIntosh, M., Schwartz, S., Reid, B., Radich, J., Anderson, G., and Hartwell, L., “The Case for Early Detection,” Nat. Rev. Cancer, Vol. 3, 2003, pp. 243–252. [17] Veenstra, T.D., Conrads, T. P., Hood, B. L., Avellino, A.M., Ellenbogen, R.G., and Morrison, R.S., “Biomarkers: Mining the Biofluid Proteome,” Mol Cell Proteomics, Vol. 4, 2005, pp. 409–418. [18] Villanueva, J., Philip, J., Entenberg, D., Chaparro, C. A., Tanwar, M. K., Holland, E. C., Tempst, P., “Serum Peptide Profiling by Magnetic Particle-Assisted, Automated Sample
Problems
323
Processing and MALDI-TOF Mass Spectrometry,” Anal. Chem., Vol. 76, 2004, pp. 1560–1570. [19] Yue, M., Lin, H., Dedrick, D.E., Satyanarayana, S., Majumdar, A., Bedekar, A. S., Jenkins, J. W., and Sundaram, S., “A 2-D Microcantilever Array for Multiplexed Biomolecular Analysis”, J. MEMS Vol. 13, 2004, p. 290. Yue, M., Stachowiak, J. C., Lin, H., Datar, R., Cote, R, and Majumdar, A., “Label-Free [20] Protein Recognition Two-Dimensional Array Using Nanomechanical Sensors,” NanoLetters, Vol. 8, No. 2, 2008, pp. 520–524. [21] Cheney, C. P., Wig, A., Farahi, R. H., Gehl, A., Hedden, D. L., Ferrell, T. L., Ji. D., Bell, R., McBride, W. J., and O’Cornor, S., “In Vivo Real-Time Ethanol Vapor Detection in the Interstitial Fluid of a Wistar Rat Using Piezoresistive Cantilevers, ” App. Phys. Lett., Vol. 90, 2007, p. 013901.
CHAPTER 12
Nanoimaging and In-Body Nanostructured Devices for Diagnostics and Therapeutics Jai Raman, Neeraj Jolly, Sungho Jin, and Ratnesh Lal
12.1
Introduction Recent advances in nanoscience and technology are of particular relevance to a wide range of disciplines that constitute nanomedicine. Their applications encompass a range of capabilities that assist in: (1) understanding the basic mechanisms of disease, (2) examining the roles of cell and tissue interactions and environmental perturbations, both internal as well as external, (3) the diagnosis of disease using nanoscale sensors and devices, (4) designing therapeutics and drugs, (5) designing efficient delivery of drugs and therapeutics, and (6) designing nano/microelectrical and mechanical stimulations for maintaining various body activities. This chapter discusses how nanoscale imaging and intermolecular interactions can be used as biosensors and devices for a wide array of diagnostics and therapeutics. The platform technology for these sensors and devices includes integrated cantilevered and optical scanning probe techniques, nanofluidics, nanochips, nano-MEMS, nanoelectronics, and wireless technology. Potential uses of nanosensors and nanodevices include measurement in real time of tissue physiological or pathological states that underlie many cardiovascular and other systemic diseases. Sensors within the body have always provided a snapshot of the inner workings of the human body. The earliest sensor that predates written history is probably the use of visual cues such as the color of the skin reflecting tissue perfusion. The palpable measurement of the pulses, which has in some traditional medical texts allowed extensive diagnostic capabilities, is another example of simple sensor technology. Interestingly, one of the most fundamental problems in clinical medicine that is edema or retention of fluid is still measured in an age-old fashion—by looking at pitting edema! This calls for pushing on the shin to see how much the skin pits to give the examiner an idea of the extent of peripheral edema. The utility of in vivo sensors have been limited to date, because implanting sensors in the body has been associated with a variety of side-effects. Most sensors within the body have not used wireless technology or significant miniaturization. Our hope in this chapter is to show you how a combination of nanotechnology and wireless applications can revolutionize this arena. The next few pages will out-
325
326
Nanoimaging and In-Body Nanostructured Devices for Diagnostics and Therapeutics
line the role of nanotechnology in the development of some very basic but important sensors.
12.2
Technology for In Vivo Sensing Since the earliest sensing of diseases as described above, the essence of all these sensors is to detect and estimate the energy emanating from body—the energy whether it is heat, mechanical pulsation, surface deformations (like bumps), or body color is primarily sensed by perceived changes in receptors whether they are our eyes, hands, or some other gauge that we use. Significantly, the simplest and most effective of all these sensors incorporate some sort of mechanical movement/deflection similar to what has now been the basic theme behind the emerging technologies of various scanning probe microscopes, including atomic force microscopy (AFM). 12.2.1 Atomic Force Microscopy for Multimodal and Multidimensional Imaging 12.2.1.1
Operating Principle
The AFM works on the principle that a very sharp tip can passively sense the localized forces between the atoms and molecules of the scanning tip and the specimen surface, meaning that molecular and even atomic resolution can be achieved. AFM technology is based on raster scanning a cantilevered tip in the x-y plane over a sample surface, either by attaching the sample or the cantilevered tip itself to a scanner. The z (vertical) position of the sample (or tip) is monitored simultaneously. The deflection of the cantilever is proportional to the interaction force between the sample and the AFM probe. Most commonly, the cantilever deflection is monitored by bouncing a laser beam of the back of the cantilever to a set of photodetectors, which converts the light signals into electrical signals (Figure 12.1). The two main imaging modes areI: (1) the contact mode, in which an electronic feedback circuit maintains a constant deflection, ensuring a constant force of interaction between cantilever tip and sample. The amount of z variation needed to maintain a constant interaction force is plotted versus the x and y coordinates, hence producing a topographic image, although the images may not be purely topographical (see Section 12.2.1.2), and (2) the tapping or vibration mode, in which the amplitude of vibration of an oscillating cantilever is maintained constant during scanning. In the tapping mode, the phase lag between the driving circuit and the actual tip vibration is also measured. 12.2.1.2
Forces in AFM
The deflection of the cantilever in the contact mode and the damping of vibration amplitude in tapping mode are caused by a sum of attractive and repulsive forces. The dominant repulsive force sensed by the AFM cantilever results from the overlapping of electron orbitals between the atoms of the tip and of the sample. The dominant attractive force is a van der Waals interaction, which is primarily due to nonlocalized dipole-dipole interactions. Another strong attractive force component
12.2 Technology for In Vivo Sensing
327
Photodiode Laser Piezo
Cantilever Sample
Feedback and x,y,z scan control
x,y,z Piezo tube scanner
Apply V, measure I
Figure 12.1 Operating principle of AFM. A cantilevered tip is scanned over the sample, maintaining constant force using feedback control to keep cantilever deflection amplitude (measured using an optical beam deflection system) constant. Topography is recorded, and simultaneous additional channels can show phase, deflection, amplitude, or current flow between the tip and sample. (Reproduced from [34].)
that exists while imaging in air is the meniscus-surface force due to adsorbed water layers. In fluids, consideration should be given to electrostatic interactions between charges and the sample and tip, and structural forces such as hydration force, solvation forces, and adhesion forces [1, 2]. In the tapping mode, conductive/magnetically coated cantilevers can sense electrostatic and magnetic forces, and image for instance magnetic domains, surface charge distributions, local surface capacitance, and local conductance. These forces can be used to generate images that provide valuable information on the differences in local surface chemistry, like separate lipid and protein clusters in a membrane, and as such can be used as effective sensors of energy and the functional states of a specimen. 12.2.1.3 AFM and Fluid Flow Properties: Measuring Nanoscale Flow Rate and Viscosity
Microscale fluid velocity, viscosity, and shear stress play significant roles in tissue sustenance and several pathologies (e.g., atherosclerosis, thrombosis). Also, local fluid mechanics would affect interaction of any therapeutics with their targets and would even be indicative of local pathologies, such as reduced flow rate around an occluded vein or artery. Yet we know very little about them. Liquid viscosity is hard to measure with high precision and in small volumes. Traditionally, ultrasonic devices are used to measure viscosity [3]. They operate at MHz frequency at which the viscosity of non-Newtonian fluids can be different versus its low-frequency value which is of greater interest. Flexural-mode resonance devices, such as microfabricated cantilevers, may be more reliable since they allow for measurement at lower frequencies. Using an optical detection approach typical in standard AFM
328
Nanoimaging and In-Body Nanostructured Devices for Diagnostics and Therapeutics
equipment, viscous drag has been measured using a piezoelectric actuator to vibrate an AFM silicon cantilever [4]. Other ways of using AFM to measure liquid viscosity include measuring the torsion in an AFM cantilever while scanning a whisker tip inside the liquid [5]. 12.2.1.4
AFM and Ion Conductance/Permeability Assay Tools
Combining an AFM with other tools is important for obtaining simultaneous information about sample functional properties and activity. In its earliest form of a combined AFM and an ionic conductance measurement system, a nanometer inner diameter glass electrode served two purposes; it acted like an AFM cantilevered tip and also an electrode for recording ionic conductance [6]. In a later version, using appropriate voltage drop across a bilayer/cell membrane, conductance through pores in a synthetic nuclear filter was measured [7]. As an extension of the combined AFM and ion conductance measuring system, AFM has recently been combined with advanced nanochip supported double chamber permeability and a transport assay system (Figure 12.2). This combined system allows study of the activity of ion channels and pores and that will be useful for high throughput sensing of pathogens and toxic signals that modify channel activity and in turn can also be used to design antidotes (potential therapeutics). Figure 12.2 shows an example of one such design in which ionic conductance through gramicidin channels is reconstituted in a lipid bilayer supported over a silicon chip with nanopores ~ 70 nm in diameter. Gramicidin is a gram-negative bacteria that induces ion channel-like activity when in contact with the cell membrane, and is a good test structure for defining membrane permeability and transport. Significantly, this study also shows that the AFM imaging force is soft enough to image delicate and fragile biological membranes that could be used for screening membrane modifiers (e.g., pathogens and toxicants). Lal and his colleagues [8] were able to implement a new conducting AFM tip that can be used for direct study of the conformational changes in ion channels as would occur in response to pathogens/toxicants. It can also be used for characterizing many advanced materials with wide biomedical applications. Ionescu et al. [8] studied the direct structure-function relation in conjugated polymer blends. Polymer blends have wide biological applications, including microactuators [9], chemical sensors [10], and light emitting diodes [11]. Conjugated optically and electrically sensitive polymer blends are being tested as sensors for pathological biological markers. 12.2.2 12.2.2.1
Nanoimaging, Nanosensing and Intermolecular Interactions Structure and Changes in Membrane Effectors
Defining cell membrane structures and their changes could be used as markers of diseased versus normal cell behavior. The identification membrane-associated targets, such as ion channels and receptors, is complex and mostly indirect since these channels, receptors, and other nanoscale structures are smaller than the current resolution of light microscopic imaging. In one of the earliest examples of AFM mapping of cell membrane receptors, the AFM was used to identify the individual
12.2 Technology for In Vivo Sensing
(a)
(d)
329 (c)
(b)
(e)
(f)
Figure 12.2 AFM integrated with a nanochip-separated double chamber electrical recording system. (Adapted from [32].) (a) AFM image of arrays of nanopores. The pores are produced in a silicon nitride chip using electron beam lithography and range in size from 50 (top rows) to 100 nm (rows on the right). (b) After deposition of lipid bilayers, AFM imaging in PBS shows a bilayer covering the pores. The bilayer is not fully contiguous and has several holes (red arrows); the cross section at a hole in the bilayer (inset) indicates the thickness of a lipid bilayer ~5.3 nm. Blue arrows indicate 500-nm wide corner alignment marks, and red arrows indicate defects in the bilayer. (c) Schematic of the liquid cell AFM setup for imaging the silicon nitride chip and bilayer. For details, please see [32]. AFM images (panel d) and conductance maps (panel e) over a pore in a microfabricated silicon chip when the lipid bilayer is formed. (f) When gramicidin is added to the bilayer, the overall current across the pores increases and conductance increases from picoSiemens to nanoSiemens suggesting the formation of hundreds of gramicidin ion channels. For details, see [32] and [34].
nicotinic acetyl choline receptor (nAchR) that was expressed in xenopus oocytes. Using ion conductance microscopy [6, 7, 12] and intermolecular force mapping (see below) capability, AFM can provide density, distribution, clustering, and functional viability of most of the cell membrane macromolecules. 12.2.2.2
Intermolecular Interactions and Defining Effectors
AFM has developed into a valuable tool for measuring molecular interactions. The possibility of linking molecules to the probing tip provides a range of options to probe molecular interactions between the probe on the AFM tip and the sample surface [13–17]. Using such modified and functionalized cantilevers, adhesion forces have been measured and mapped between ligands and receptors on the surface of living cells (Figure 12.3). Using force volume imaging [18], the regional distribution
330
Nanoimaging and In-Body Nanostructured Devices for Diagnostics and Therapeutics
Figure 12.3 Left panels: Adhesion forces between VEGF receptor and antibody (a), with distribution of unbinding forces (b). Blocking peptide prevents the interaction (c), and (d) shows tapping mode image of VEGF receptors on mica. (e) and (f) show adhesion forces between the VEGF receptor on endothelial cells and antibody on the tip without (e) and with (f) the presence of a blocking peptide. Right panels: AFM images of endothelial cells before (a) and after (b) addition of VEGF, showing cytoskeletal reorganization. Fluorescence images showing the presence of Flk-1 receptors (c, d) and control (e) with nonspecific antibody. (From [34].)
as well as ligand or antibody induced clustering of VEGF receptors have been reported (Figure 12.4) [19]. Using combined AFM with fluorescence microscopy, Quist et al. [18] have examined cell volume regulation in response to external perturbations. The added benefit of force mapping is the simultaneous mapping of the stiffness of the cell membrane (Figure 12.4), thus AFM cantilevers can be used as “stiffness sensors.” There are active efforts now to use stiffness sensors to distinguish cancerous versus normal cells [20]. Intermolecular interaction force can also be used to sense normal versus abnormally folded proteins that underlie many diseases. As an example, Liu et al. [17] used an AFM tip conjugated with antibodies specific to different regions of a hemichannel peptide, connexin 43 (Cx43), to map the specific Cx43 epitopes that open and close the hemichannel in response to changing calcium concentration (Figure 12.5). The force required to unwind the peptide (for channel opening) was correlated with the mobility of specific portions of Cx43. The precise estimate of the energy to unfold and stretch (contract) peptides using the AFM shows the promise of using AFM for more accurate indication of the therapeutic efficacy of drugs and pharmacological agents with single molecule resolution. 12.2.2.3
Cell Mechanics
The proper functioning of cells depends on controlled biochemical processes as well as regulated cytoskeletal structural reorganization. Reorganization of the
12.2 Technology for In Vivo Sensing
331
Figure 12.4 (Color plate 18) Top histograms show distribution of adhesion forces between the anti-VEGFR antibody-conjugated AFM tip and VEGFR in the cell plasma membrane. Middle and bottom panels show simultaneous acquisition of topography (middle panel) and elasticity images (bottom panels) before, and 10 and 45 minutes after adding VEGF antibody. Clustering of receptors can be observed both in topography, and elasticity maps (spots labeled 1–4). For details, see [19]. (From [34].)
cytoskeleton and changes in its mechanical properties play key roles in cell growth, migration, and development [21]. The local mechanical properties of a cell are closely associated with biochemical gradients across the membrane, but most techniques that have been used to study single cells average over the entire cell [22]. AFM gives the opportunity to measure the mechanical properties of cells with high spatial resolution. For instance, rat atrial myocytes were imaged clearly showing the cytoskeletal network beneath the cell membrane and myofibrillar structure (Figure 12.6, left panel [23]). Using a constant cantilever deflection maintained by feedback, contractile activity and the change in contractile activity using perturbations in the buffer environment (Figure 12.6, right panel) was examined. This demonstrates the possibility to quantify the coupling between subcellular substrates to cellular functions such as contraction, migration, growth, and differentiation. Such monitoring of cell mechanics can be used effectively to diagnose pathologies associated with abnormal cell mechanics as well as to monitor the efficacy of
332
Nanoimaging and In-Body Nanostructured Devices for Diagnostics and Therapeutics
Figure 12.5 (Color plate 19) AFM antibody conjugated tip pulling experiment: Sensing intermolecular interaction for mapping conformational epitopes. Left side: Model of a gap junction plaque with opposing cell membranes with hexagonally packed hemichannels (left). Schematic of force spectroscopy measurement showing binding of antibody connected to the AFM tip with flexible PEG spacers to Cx43 hemichannels reconstituted in the lipid bilayer (middle) and the schematic of Cx43 membrane topology with two extracellular loops, one cytoplasmic loop, and the cytoplasmic carboxyl-terminal domain and locations of antigenic binding sites for anti-CT252–270 and anti-CT360–382 (right) are shown. (From [34].) Right side: Probability histograms of the rupture forces of the measured anti-CT252–270-Cx43 (A), anti-CT360–382-Cx43 (b), and GAP26-Cx43 interactions (c). (d): A representative force-extension curve showing specific avidin-biotin interaction in the PEG spacer extension test system. (e) The average extension of PEG spacer stretching (~28 nm). (f) Histograms of measured tether extensions in anti-CT252–270-Cx43, anti-CT360–382-Cx43, and GAP26-Cx43, respectively. (For details, see [17].)
therapeutics that are expected to correct abnormal cell mechanics. Indeed, this is one of the underlying principles behind proposed stent endothelialization sensors. 12.2.2.4
Tissue Nanoelasticity and Nanopatterning
Biological fibers have nanomechanical properties that depend on their morphology as well as the chemical heterogeneity of their constituent subunits. A detailed understanding of tissue elasticity can be used for efficient and early diagnosis and for monitoring the progression of diseases and their treatments, especially diseases of bones and calcified tissues. Indeed, there are some new diagnostic tools being tested to diagnose bone diseases [24, 25], that resulted from our understanding of the tissue mechanics at the molecular level [26]. The correlation between the mechanical properties and the heterogeneous subunits on a nano scale has been limited. Using AFM, such correlation studies are possible [27]. An example of such studies is summarized in Figure 12.7. AFM force mapping was used to study the mechanical properties of the different constituents of wool fibers. It was shown that the exocuticle part has the highest elastic modulus while the endocuticle and cortical regions of the fiber have significantly lower modulus. The indentations made by a diamond-tipped cantilever look indeed distinctly different for the different regions (Figure 12.7, right panel). Furthermore, the AFM study could give conclusive evidence of the role of
12.2 Technology for In Vivo Sensing
333
Figure 12.6 Left panels: AFM image of live myocytes (a); inset shows myofibril structure, deflection image of cells (b), and fluorescence image after staining for F actin (c). Right panels show local contractile activity of a beating atrial cell recorded with AFM tip, under 1.8-mM (a), and 5-mM (b) calcium, and 4-mM butanedione monoxime with high calcium.
disulfide bonds in the fiber stiffness. Reduction of such bonds, abundant in the exocuticle, using DTT, resulted in a reduction of modulus. Using the unique feature of defined nanoindentation, one can create specific patterned structures, like nanocavities, nanopores, nanowells, and nanochannels and use them for array sensors of pathogens and toxicants. 12.2.3
Parallel Arrays of Sensors to Detect Complementary Interactions
Emerging technologies are generating a myriad of array sensors ranging from high throughput screening of genetic materials to individual protein and peptides that rely on specific interactions using antibodies, peptides, and fluorescence labeling. For assaying drugs for “channelopathies” and for identifying biomarkers of diseases, two powerful avenues involve patch clamp(s) on chips and microfluidics. Significantly, AFM can be combined with both techniques easily and thus provide powerful techniques.
334
Nanoimaging and In-Body Nanostructured Devices for Diagnostics and Therapeutics
Figure 12.7 Left panels: schematic diagram of wool fiber (a). TEM image (b); higher magnification in (d) and AFM image (c); higher magnification in (e) of same wool fiber region. AFM and TEM images correlate well with respect to identified subcellular structures. Center panels: force curves on cortex (top), exocuticle (middle), and embedding resin (bottom). Solid line is collected on hard glass surface. Larger shaded area indicates more elastic surface. Right panels: Top right image shows corresponding AFM image. Bottom right image shows surface after indentation with diamond tip to create specific patterns on wool fiber. For details, see [27]. (From [34].)
12.2.3.1
Patch Clamp on a Chip
Patch clamp recording is one of the main techniques employed in electrophysiological studies and commonly used in the study of channelopathies. Generally, a precisely pulled glass pipette is brought in the vicinity of a cell of interest under an optical microscope using a micromanipulator [28]. This results in the technique being slow and having a low throughput, making it not very suitable for use in for instance proteomics and drug discovery development. For these fields of science, an automated patch clamp that uses disposable devices would be an important outcome [29, 30]. An example of such an on-chip microchannel planar patch clamp is shown in Figure 12.8. The middle panels show a HeLa cell being introduced in the microfluidic system and being trapped at one of the channels. Cell deformation can be observed, and current traces show a seal resistance of 144 MOhm [29]. Such devices using a cell reservoir linked to patch channels allows for simultaneous optical and electrical recording that will facilitate studying the role of ion channels with respect to cellular functions. Future devices can be designed with an improved patch pore geometry and surface treatment to obtain a better seal necessary for single channel conductance studies. Since the cell reservoir is linked to many patch clamp pores, a parallel readout is possible analyzing multiple cells simultaneously. Also,
12.2 Technology for In Vivo Sensing
335
Figure 12.8 Schematic drawing of conventional patch clamp and an on-chip patch clamp (left). Cells can be pulled into one of the on-chip patch pores (middle top panel), resulting in membrane deformation and patch formation (middle bottom panel). Resulting current measured before (top right top panel), and after cell attachment (right bottom panel). For details, see [29]. (From [34].)
there are many new variations of this technique commonly being developed, the most important being an array of patch clamping on nanochips with integrated microfluidics. 12.2.3.2
Microfluidic Viscosity, Velocity, and Molecular Affinity Sensors
Parallel readout techniques based on AFM include cantilever arrays that allow for analysis of large areas with high resolution and exquisite intermolecular interaction sensitivity. When combined with micro- (and nano-) fluidic chambers, these array nanosensors are high throughput screening devices for biomarkers and therapeutic agents. They can probe for instance multiple live cells for their elastic properties or the presence of receptors in the cell membrane. Similarly, coating a parallel array of cantilever with a different material or reagent on each lever results in a “chemical nose” that can sense a variety of chemicals or toxins in very small volumes. Figure 12.9 shows schematics of cantilevered microfluidic sensing tools in which a piezoelectric cantilever senses nanoscale viscosity and velocity of fluids with different viscosity and ionic composition and can be easily adapted for screening biomarkers in blood and body fluids (Figure 12.9, top panels). When combined with local fluorescence sensors (e.g., TIRF, FRET, photo-sensitive nanoparticles), electrical properties (e.g., piezoelectric circuitry) and mechanical properties (e.g., cantilever deflection), an array of cantilevers with specific complements can provide a powerful and highly sensitive tool for high throughput screening of pathologies and therapeutics from a very small (micro to milliliter) amount of biofluids (Figure 12.10).
336
Nanoimaging and In-Body Nanostructured Devices for Diagnostics and Therapeutics
Figure 12.9 Nanosensors for biomarkers and therapeutic agents. Microfluidic viscosity, velocity, molecular affinity sensors are shown. (a) Schematics of cantilevered microfluidic sensor. (b) Setup of the stainless steel needle (top) and silicon channels (bottom). The cantilever is inserted and aligned using a micromanipulator and held with the base nearly touching the upper edge of the channel opening. The left side shows side view, and the right side shows a front view looking into the channel. (c) FIB-milled cantilevers serve as flow and viscosity sensors for biological fluids. (d) Comparison of voltage readout at different flow speeds. To guide the eye, points are connected by lines. High-viscosity fluids such as ethylene glycol saturate the amplifier at relatively low flow speeds. At higher flow speeds, the sensor distinguished between DMEM buffer with 5% and 50% fetal bovine serum. The protein content of blood serum is a major contributor to the viscosity of biological fluids. For details, see [33]. (From [34].)
12.3
In-Body Nanosensors and Nanodevices There are a variety of diagnostic and therapeutic capabilities using nanotechnology that can be envisaged. For example, in vivo diagnostics would include improved contrast imaging agents for PET, CT, MR imaging, nanoparticles and conjugated imaging agents (e.g., quantum dots), sensors for cellular and/or tissue growth, and tissue sensors looking at fluid and nutrient levels. In vitro diagnostics technologies are more advanced; there are sensors for biomarkers—using nanoarrays (chips, microfluidics, etc.) and for high throughput screening of biomarkers and pathogens. Since this chapter deals with in body sensors, let us look at them in greater detail. Some common and simple characteristics of these sensors, by virtue of their size and location, make them easy to implant with little chance of adverse reactions. These common features, which are so desirable, include (1) nanoscale sensors that allow safe implantation with little chance of tissue reaction, (2) remote actuation and remote control (which utilizes wireless technology), (3) easily implantable in the target tissue or organ, and (4) a complex interplay of biomedicine, electronics,
12.3 In-Body Nanosensors and Nanodevices
337
(a)
(b)
(c)
Figure 12.10 Cantilever-based array detectors for complements and biomarkers assay in a microfluidic chamber. (a) Arrays of AFM cantilevers that are used for sensing intermolecular interaction force. (b) Schematics of cantilevers conjugated with receptors and channels (top) and their complements (e.g., ligand, agonists/antagonists, antibodies). (c) Schematics of a microfluidic chamber and cover slips containing arrays of functionalized cantilevers that sense various biomarkers in a fluid (e.g., blood sample). (From [34].)
materials science, nanotechnology, which allows the fabrication and use of these sensors. Some examples of these sensors are all potential applications that we are working on. We have conceptualized and designed components of remotely controlled and read sensors for edema (body fluid accumulation) that is often life-threatening if not detected in a timely manner and for stent endothelialization that governs the duration of drug treatment after insertion of a stent, which is very expensive and also induces blood loss in case of any injury or operation. Schematics of these devices are shown in Figures 12.11 and 12.12. Other sensors include sensors to assess muscular contraction and relaxation—in a variety of organs, most importantly the heart, and electrical sensors looking at electrical activity of a particular organ such as the heart, which can then couple with a pacemaker; alternative iterations could be used to monitor brain activity in epileptic patients. None of these sensors are in actual clinical use. However, the promises of such small, nanoscale devices that can be implanted have opened vast horizons in this field. We will use
338
Nanoimaging and In-Body Nanostructured Devices for Diagnostics and Therapeutics
Figure 12.11 Remotely controlled in vivo endothelialization sensors. (a) Schematics of components that comprise a potential remotely controlled sensing mechanism. (b) Placement of sensor with regard to blood flow. (c) Schematics of a sensor attached to a stent. (d) Key components of a stent endothelialization sensor, in this case a piezoresistive or piezoelectric pressure sensor. (e) Cartoon of sensor and RF transmitted in an integrated microchip that is placed in an intracoronary stent strut.
the edema sensor as a broad example of nanoscale in-body sensors, because they use a variety of sensing mechanisms to assess fluid buildup in tissue. 12.3.1
Edema Sensor
Tissue edema is a fundamental response to injury. Tissue edema that occurs in the heart and lung has devastating consequences and is often diagnosed in advanced stages because of the fairly imprecise nature of existing imaging technologies. Edema is the accumulation of fluid within tissues, either inside or around cells in the tissue. Typically, edema affecting the lungs, as seen in heart failure, is a very common cause of morbidity and the single most common cause for hospital admissions in the developed world. Edema is difficult to detect in internal organs until the degree of edema is advanced and even then, the detection is based on subtle changes in soft tissue imaging. Edema in peripheral tissues such as the legs is evident only when it is very advanced. The adverse consequences of edema are dramatic in almost every part of the body (e.g., reduction in tissue oxygenation, reduction in function, and delayed healing). Understanding its pathogenesis and its prevention, therefore, is of prime importance to management of patients with edema and fluid accumulation. Current practice guidelines recommend admission of patients, monitoring their body weights,
12.3 In-Body Nanosensors and Nanodevices
A
339
Brain edema sensor
RF, magnetic or thermal input signal and power
Computer analysis and control
Controller module for RF magnetic or thermal power and signal transmitter and signal receiver
Lung edema sensor Kidney edema sensor Arm edema sensor Remote RF output signal from implanted edema sensor
Leg edema sensor Implanted edema sensor(s )
B
R i
Electrical resistance measured to differentiate the high resistivity through air or low resistivity through edema fluid Current flow between two electrodes
V
C
Acoustic wave (piezo electric or magnetostrictive)
Vibration Acoustic wave generator electrode
V
Acoustic wave receiver/ detector electrode
Figure 12.12 Remotely controlled in vivo edema sensors. (a) Schematics of components that comprise remotely controlled edema sensing mechanism. Edema sensors in various body organs are shown. (b) Schematics of an electrical resistance sensor. (c) Schematics of an acoustic sensor.
measuring urine output and assessing their degree of “lung water” using chest x-rays. All of these techniques are indirect and very imprecise. Currently, there is no convenient and economically viable noninvasive method to evaluate the amount of edema in a given tissue. The buildup of fluid within a cell and its surrounding alters the tissue in a variety of ways—increased weight of the tissue, increased volume of the cells, increased space between cells, altered electrical conductivity of the tissue, changed tissue impedance on ultrasonography, and altered chemical composition of the tissue fluid. Our aim is to have a composite sensor that has multiple methods of assessing tissue edema and then use that to measure the buildup or reduction of fluid. This can then be used to monitor therapy. Edema is a condition characterized by the accumulation of fluid in the interstices of tissues. This fluid is typically low in protein content and has the same concentration of physiological extra-cellular fluid. Detection of increased extracellular tissue fluid is often a clinical diagnosis, based on clinical examination or indirect measurement utilizing imaging modalities. In our conceptualized design of in vivo sensors, many sensors are typically arranged in an array of multiple sensors that are stacked together, allowing easy
340
Nanoimaging and In-Body Nanostructured Devices for Diagnostics and Therapeutics
implantation in the tissues of interest. These sensor arrays are minimally reactive to surrounding tissue and at the same time have the ability to sense the changes in the organ of interest. Remotely applied actuating signals in high-frequency RF radiation, magnetic field radiation, optical illumination, or thermal energy radiation in multiple train pulse or single pulse configurations are transmitted from outside the human or animal body towards an earlier implanted sensor or sensors. The remote sensor measures a change in the physical or biological properties of the environment near the edema sensor, thus providing information as to how much fluid has built up or edema has developed. For example, a change in acoustic behavior, electrical resistance, capacitance, heat dissipation rate, local pressure change, local displacement of a portion of the sensors, or a detection of the presence or absence of specific biomolecules can be measured by the sensor. The implanted sensor, in response to the interrogating input signal, then transmits a remote output signal that is picked up by the circuits in the remote (outside the human or animal body) controller. Significantly, such remotely controlled edema sensors can proactively monitor edema formation or its progression in one organ, multiple organs, and/or the lungs. In order to make edema and endothelialization sensors foolproof and to yield a high level of confidence, these sensors are designed to sense various physicochemical properties and fluid macromolecules and as such are referred to as acoustic implant edema sensors, electrical implant edema sensors, heat dissipation based edema sensors, tissue displacement implant sensors, biological analysis implant sensors, pressure sensors, and capacitance sensors. Other remote sensors could detect immune mediated edema in organ rejection and in vivo tissue culture speed as edema sensors. 12.3.1.1
Principles of In Vivo or In Body Sensors
Figures 12.11 and 12.12 schematically illustrate the mechanism for remotely detecting the degree of (or the absence versus presence of) stent endothelialization and edema progress, respectively. The remote operatively associated endothelialization sensors are strategically positioned near or attached to the stent system implanted in the subject. Figure 12.11(a) schematically illustrates the components that comprise a potential remotely controlled sensing mechanism. The sensors can transmit an RF or DC signal to be detectable by an external RF receiver system that remotely and wirelessly retrieves such signal for data acquisition and analysis to proactively measure the degree or the absence/presence of endothelialization. The sensors can detect the presence or absence of stent exposure to flowing blood. For example, the sensors can use heat dissipation rate, electrical resistance, magnetostrictive forces, pressure or capacitance, biomolecular sensors to measure sugar level, ionic contents, oxygen contents, or other intraluminal biomolecule presence, concentration, or absence, or to measure change of electrical, magnetic, or molecular circuit connection. Infrared-radiation-based thermoelectric sensors can also be used. One or more sensors can be attached and operatively associated with the stent device. The sensor can be attached to any portion of the stent device, for example, on a stent wire surface. The stent sensor device can be placed either flat (as illustrated in Figure 12.11(b) and (e) or positioned to protrude from the stent structure into the lumen of the vessel in
12.3 In-Body Nanosensors and Nanodevices
341
which it is implanted (as illustrated in Figure 12.11(c). The key components of a stent endothelialization sensor can also be made of a piezoresistive or piezoelectric pressure sensor (Figure 12.11(d)). Shown in Figure 12.11(e) is a cartoon of sensor and RF transmission circuit in an integrated microchip that is placed in an intracoronary stent strut. The edema sensor system, shown in Figure 12.12, basically comprises two components—a controller module outside the human or animal body (as illustrated in the left portion of Figure 1212.(a)) and at least one sensor within a human or animal body attached onto or into the tissue in question (as illustrated on the right-side portion of Figure 12.12(a)). The outside module has both a signal/power transmitter and a signal receiver. The transmitter part of the module sends an interrogating signal periodically to the implanted sensor. The transmitted signal can be an RF, magnetic, or thermal signal at an appropriate frequency. For example, 100-kHz to 500-MHz RF signals or DC 300-MHz magnetic signals can be utilized. The thermal signal can be in the form of laser pulses or broad-spectrum light, preferably close to the infrared regime for the purpose of enhanced penetration of intended heat through the human tissue toward the sensor or a power storage device on the implanted sensor. An example sensor based on electrical resistance change is described as Figure 12.12(b) while a sensor based on a principle of acoustic signal change is shown as Figure 12.12(c). The micro- or nanoscale edema sensors should preferably be remotely operated without their own power source rather than having an awkward configuration of a battery power source implanted and connected to the sensor. One possible scenario is to utilize the outside module as a remote source of power for the implanted sensor. For example, magnetic induction can be employed and the energy so captured by a portion of the sensor circuit can be stored as a magnetic or capacitive energy that the sensor can use for sending back the measurement data to the outside control module. Alternately, a micro scale thermoelectric device can be attached as a part of the implanted sensor to capture a temperature differential near the sensor in the tissue created by infrared light heating from outside through the human tissue, and use the generated voltage for transmitting sensor signals to the controller module. The implanted sensor portion of the system contains a circuit or circuits, which send the measured data by the sensor as a remote RF or magnetic signal to the outside controller module. One or more edema sensors can be attached or coated onto the surface or parenchyma of the organ or tissue in question. 12.3.2
Remote Controlled, Magnetically Navigated Robot Capsule
A small capsule device that can be swallowed has been available in the medical community for the past several years. For further improved functionality, a remote-controllable capsule device as described below can be designed for guided navigation through the GI tract of the intestine as depicted in Figure 12.13. The capsule would contain a small camera, light source, battery, circuit chips for control and wireless transmission of video and other signals, and distributed magnetic materials. The location and movement of the capsule in the GI tract can then be remotely controlled by a three-dimensional magnetic field. The guided capsule will perform two functions (i.e., diagnosis and therapeutics). When a patient swallows the capsule,
342
Nanoimaging and In-Body Nanostructured Devices for Diagnostics and Therapeutics
Remote electromagnet array 1
2
3
4
5
GI Tract Magnetizable stripes Intestine wall Swallowed camera
Motion in gradient field
1 Figure 12.13 drug release.
2
3
4
Drug delivery Tumor capsule (MEMS or ulcer or magnetically actuated)
5
Remote magnetically navigated robot capsule for imaging, biopsy, and on-demand
the natural muscular waves of the digestive tract propel it downward and as it moves the camera takes pictures of the small intestine wall for video transmission to detect tumors, ulcers, or causes of bleeding. This procedure permits the visual diagnosis of the small intestine, which is difficult to access by colonoscopy. For larger regions, such as the stomach or the large intestine, the remote magnetic field guidance can be utilized to control the capsule orientation so that no portion of the GI tract surface would be missed. Movement of the magnetic object is induced if there is a magnetic field gradient near the magnetic object. A programmable or sequential change of magnetic field strength and gradient near the magnetically tagged capsule will guide and move the capsule at a programmed speed as illustrated in Figure 12.13. The capsule is dimensioned and shaped to move within a tract duct or cavity of the body. Advantageously, the outer surface of the capsule comprises biocompatible material. The exemplary capsule is a camera for video imaging that can also include other navigatable device and control systems such as a MEMS device for biopsy to take tissue samples and store the sample in the capsule. A drug-release device, or diagnostic or treatment device, or any combination of these may also be considered. The navigatable diagnostic and/or therapeutic treatment devices can also include, for example, a local ultrasonic wave source and/or detector for localized acoustic diagnosis or treatment (e.g., to damage or disrupt unwanted cell structure), a localized x-ray source and/or detector, a source of intense heat radiation to disrupt or damage tumor growth, or a mechanical vibration source and/or detector to stimulate or slow down cell growth. For therapeutics, the magnetically navigated capsule will be positioned at a fixed location near the site of a tumor for remote-actuated, on-demand release of cancer drugs or other therapeutic agents. If there is more than one tumor location, the capsule can be magnetically moved up or down the tract for repeated therapeutic release of drugs with programmable doses and release intervals. The navigating capsule can be combined with a remote, on-off switchable drug release capsule using an RF magnetic field. The capsule can be powered with a battery, for example, typically with several hours of life. Alternatively, the system can be provided with an AC magnetic field
12.3 In-Body Nanosensors and Nanodevices
343
source (e.g., 60 Hz) and the navigating biocapsule can be equipped with a transducer device for receiving energy through the external AC magnetic field. Using a magnetic AC induction coupling, a transformer solenoid in the capsule can receive power from outside, and then convert the AC power to DC power and store the energy in the battery or a capacitor in the capsule. Such a system can indefinitely extend the battery life, and the useful life of the in vivo device. 12.3.3
Mobile Microscopic Robots
This section proposes ideas on the application of technology that is described in Chapters 10, 14, and 15. It is anticipated that the authors of these chapters, Hogg, Freitas, and Boehm, respectively, can design an ensemble of mobile microscopic robots/sensors that will circulate in the bloodstream going throughout the body. When these sensors arrive at the diseased area, the concentration of the chemical or marker will be higher than the average concentration in the blood, which will make detection easier. Also, when the sensors trigger their signal, the location of the disease will be known. Thus the disease will be detected earlier when treatment can be more effective, and the disease can be treated locally, which has fewer side effects for the patient. The sensors will have the capacity for multiple applications and might be administered to the patient via injection, pill, or aerosol form. Externally controlled acoustic, photonic, or magnetic stimuli would activate specific diagnostic or therapeutic operations. We propose that circulating sensors/robots be used in conjunction with the types of fixed and large for arterial use implantable sensors described in this chapter. In one instance, the fixed implantable sensors would monitor signals from the circulating sensors serving as an intermediate sensor station that would subsequently transmit information obtained from the micron-sized circulating sensors outside the body by near field electromagnetic radiation (batteryless inductive coupling). Micron-sized circulating sensors will have limited functionality due to their small size; communication capability in particular will be limited. The substation approach can overcome to some degree the limited functionality by interrogate the circulating sensors and transmitting information outside the body. Level-one communication between the circulating sensors and substation will be over short distance and through magnetic, thermal, or optical communication. Details of the design of this two-stage sensing approach have not been worked out yet. But it is obvious that a highly interdisciplinary team must be assembled to handle this challenging open-ended design problem. A second idea stems from the problems of having sensor that are foreign objects circulating in the body. W. Wagner [31] notes that for the indwelling blood sensors, a major challenge will be avoiding thrombosis (blood clotting) and fibrosis (scarring over the sensor). Infection is also an ongoing risk in the bloodstream for foreign materials. These challenges have not been effectively addressed to date, and would limit the duration a fully functional sensor could be operational. Therefore, some activities must focus on trying to protect metallic surfaces from these adverse physiologic responses. For example, we envision the mobile sensors may have nanopatterned surfaces to reduce fouling or the sensors may biodegrade in the body thus avoiding these problems. Another possibility is that mobile microscopic sen-
344
Nanoimaging and In-Body Nanostructured Devices for Diagnostics and Therapeutics
sors themselves are biological materials or are imbedded within biological materials such as biological cells that normally circulate in blood. The integration of biotic and abiotic materials to form sensors and actuators seems futuristic, but is within the scope of investigations that can now be conducted using nanotechnology tools and devices. It is speculated that different types of sensors will be used in nanomedicine and that parallel research investigating many approaches is needed at this early stage in order to downselect to the most promising approaches for near-term turnover of practical devices to physicians for pre-clinical evaluation.
12.4
Conclusions This chapter has described nanoscale imaging and a range of feasible implantable sensors and nanodevices that will take medicine to a new higher level of sophistication. Our goal is to provide physicians the tools that they need to work at the nanoscale to diagnose and fight disease. These may be the most complicated tools that mankind has ever developed because the tools integrate biology, nanotechnology, electronics, and mechanical systems. Thus, nanomedicine and development of bionanoelectromechanical systems is a new small science that will have large benefits for all of us.
Problems 12.1 An implantable device will be identified as a foreign body and the human body will react by trying to remove or degrade the device. What are the biocompatibility issues related to implantable devices? 12.2 In your opinion, what are the most important possible clinical aplications of implantable devices? 12.3 Thrombosis (blood clotting) and fibrosis (scarring over the sensor) may occur for implantable sensors. Infection is also an ongoing risk in the blood stream for foreign materials. These problems may limit the duration a fully functional sensor could be operational. Investigate how these problems can be overcome. Provide different approaches. 12.4 Protecting metallic surfaces from adverse physiologic responses is important to allow implantable devices to operate for extended periods. Investigate the use of nanopatterned surfaces to reduce fouling and to allow devices to operate in the body. 12.5 To make protein sensing possible in vivo, the problems of nonspecific binding, limited life of antibodies and receptors, and protein fouling of the sensor electrode must be overcome. Perform a literature survey and suggest ways these problems might be overcome.
References
345
References [1] Mechler, A., Kokavecz, J., Heszler, P., and Lal, R. “Surface Energy Maps of Nanostructures: Atomic Force Microscopy and Numerical Simulation Study.” Applied Physics Letters, 82:3740–3703, 2003. [2] Thimm, J., Mechler, A., Lin, H., Rhee, S.K., and Lal, R., “Calcium-Dependent Open/Closed Conformations and Interfacial Energy Maps of Reconstituted Hemichannels,” Journal of Biological Chemistry, 280: 10646–10654, 2005. [3] Jensenius, H., Thaysen, J., Rasmussen, A.A., Veje, L.H., Hansen, O., and Boisen, A., “A Microcantilever-Based Alcohol Vapor Sensor-Application and Response Model,” Applied Physics Letters, 76: 2615–2617, 2000. [4] Hauptmann, P., Lucklum, R., Puttmer, A, and Henning, B., “Ultrasonic Sensors for Process Monitoring and Chemical Analysis: State-of-the-Art and Trends,” Sensors and Actuators A-Physical, 67:32–48, 1998. [5] Mechler, A., Piorek, B., Lal, R., and Banerjee, S., “Nanoscale Velocity-Drag Force Relationship in Thin Liquid Layers Measured by Atomic Force Microscopy,” Applied Physics Letters, 85:3881–3883, 2004. [6] Hansma, P.K., Drake, B., Marti, O., Gould, S.A., and Prater, C.B. “The Scanning Ion-Conductance Microscope,” Science, 243:641–643, 1989. [7] Proksch, R., Lal, R., Hansma, P.K., Morse , D., and Stucky, G., “Imaging the Internal and External Pore Structure of Membranes in Fluid: Tapping Mode Scanning Ion Conductance Microscopy,” Biophys J., 71:2155–2157, 1996. [8] Ionescu-Zanetti, C., Mechler, A., Carter, S.A., and Lal, R., “Semiconductive Polymer Blends: Correlating Structure with Transport Properties at the Nanoscale,” Advanced Materials, 16: 385–9, 2004. [9] Jager, E.W.H., Smela, E., and Inganas, O. ,“Microfabricating conjugated Polymer Actuators,” Science, 290:1540–1545, 2000. [10] Hagleitner, C., Hierlemann, A., Lange, D., Kummer, A., Kerness, N., Brand, O., and Baltes, H., “Smart Single-Chip Gas Sensor Microsystem,” Nature, 414: 293–296, 2001. [11] Carter, S.A., Angelopoulos, M., Karg, S., Brock, P.J., and Scott, J.C., “Polymeric Anodes for Improved Polymer Light-Emitting Diode Performance,” Applied Physics Letters, 70: 2067–2069, 1997. [12] Shevchuk, A.I., Hobson, P., Lab, M.J., Klenerman, D., Krauzewicz, N., and Korchev, Y.E., “Endocytic Pathways: Combined Scanning Ion Conductance and Surface Confocal Microscopy Study,” Pflugers Arch., 456(1):227–235, 2008. [13] Baumgartner, W., Hinterdorfer, P., Ness, W., Raab, A., Vestweber, D., Schindler, H., and Drenckhahn, D., “Cadherin Interaction Probed by Atomic Force Microscopy,” Proc. Natl. Acad. Sci. USA, 97: 4005–4010, 2000. [14] Hinterdorfer, H., Baumgartner, W., Gruber, J.H., Schilcher, K., and Schindler, H., “Detection and Localization of Individual Antibody-Antigen Recognition Events by Atomic Force Microscopy,” Proc. Natl. Acad. Sci. USA, 93: 3477–3481, 1996. [15] Yuan, C.B., Chen, A., Kolb, A., and Moy, V.T., “Energy Landscape of Streptavidin-Biotin Complexes Measured by Atomic Force Microscopy,” Biochemistry, 39: 10219–10223, 2000. [16] Zhang, X.H., Wojcikiewicz, E., Moy, V.T., “Force Spectroscopy of the Leukocyte Function-Associated Antigen-1/Intercellular Adhesion Molecule-1 Interaction,” Biophysical Journal 83:2270–2279, 2000. [17] Liu, F., Arce, F.T., Ramachandran, S., and Lal, R., “Nanomechanics of Hemichannel Conformations: Connexin Flexibility Underlying Channel Opening and Closing.” J Biol Chem., 281(32):23207–23217, 2006.
346
Nanoimaging and In-Body Nanostructured Devices for Diagnostics and Therapeutics [18] Quist, A.P., Rhee, S.K., Lin, H., and Lal, R., “Physiological Role of Gap-Junctional Hemichannels: Extracellular Calcium-Dependent Isosmotic Volume Regulation,” Journal of Cell Biology, 148: 1063–1074, 2000. [19] Almqvist, N., Bhatia, R., Primbs, G., Desai, N., Banerjee, S., and Lal, R. “Elasticity and Adhesion Force Mapping Reveals Real-Time Clustering of Growth Factor Receptors and Associated Changes in Local Cellular Rheological Properties,” Biophysical Journal, 86: 1753–1762, 2004. [20] Cross, S.E., Jin, Y.S., Rao, J., and Gimzewski, J.K., Nanomechanical Analysis of Cells from Cancer Patients. Nat Nanotechnol., 2(12):780–783, Epub 2007. [21] Lal, R., Drake, B., Blumberg, D., Saner, D.R., Hansma, P.K., and Feinstein, S.C., “Imaging Real-Time Neurite Outgrowth and Cytoskeletal Reorganization with sn Atomic-Force Microscope,” American Journal of Physiology-Cell Physiology, 38: C275–C285, 1995. [22] Brady, A.J,. “Mechanical-Properties of Isolated Cardiac Myocytes,” Physiological Reviews, 71: 413–428, 1991. [23] Shroff, S.G., Saner, D.R., and Lal, R., “Dynamic Micromechanical Properties of Cultured Rat Atrial Myocytes Measured by Atomic-Force Microscopy,” American Journal of Physiology-Cell Physiology, 38: C286–C292, 1995. [24] Fantner, G.E., Birkedal, H., Kindt, J.H., Hassenkam, J., Weaver, J.C., Cutroni, J.A., Bosma, B.L., Bawazer, L., Finch, M.M., Stucky, G.D., and Hansma, P.K., “Influence of the Degradation of the Organic Matrix on the Microscopic Fracture Behavior of Trabecular Bone,” Bone, 35: 1013–1022, 2004. [25] Fantner, G.E., Birkedal, H., Kindt, J.H., Hassenkam, J., Weaver, J.C., Cutroni, J.A., Bosma, B.L., Bawazer, L., Finch, M.M., Stucky, G.D., and Hansma, P.K., “Sacrificial Bonds and Hidden Length Dissipate Energy as Mineralized Fibrils Separate During Bone Fracture,” Nature Materials, 4: 612–616, 2005. [26] Thompson, J.B., Kindt, J.H., Drake, B., Hansma, H.G., Morse, D.E., and Hansma, P.K., Bone Indentation Recovery Time Correlates with Bond Reforming Time, Nature, 414: 773–776, 2001. [27] Parbhu, A.N., Bryson, W.G., and Lal, R., “Disulfide Bonds in the Outer Layer of Keratin Fibers Confer Higher Mechanical Rigidity: Correlative Nano-Indentation and Elasticity Measurement with an AFM,” Biochemistry, 38: 11755–11761, 1999. [28] Sakmann, B., and Neher, E., Single Channel Recording, Plenum, New York, 1983. [29] Seo, J., Ionescu-Zanetti, C., Diamond, J., Lal, R., and Lee, L.P., “Integrated Multiple Patch-Clamp Array Chip Via Lateral Cell Trapping Junctions,” Applied Physics Letters, 84: 1973–1975, 2004. [30] Xu, J., Wang, X.B., Ensign, B., Li, M., Wu, L., Guia, A., Xu, and J.Q., “Ion-Channel Assay Technologies: Quo Vadis?” Drug Discovery Today, 6: 1278–1287, 2001. [31] Personal Communication, 10/08, Dr. William R. Wagner, Deputy Director of the McGowan Institute for Regenerative Medicine, Deputy Director for the Engineering Research Center Revolutionizing Metallic Biomaterials, http://www.mirm.pitt.edu/people/bios/Wagner1.asp. [32] Quist, A.P., Chand, A., Ramachandran, S., Daraio, C., Jin, S., and Lal, R., “Atomic Force Microscopy Imaging and Electrical Recording of Lipid Bilayers Supported over Microfabricated Silicon Chip Nanopores: Lab-on-a-Chip System for Lipid Membranes and Ion Channels.” Langmuir,. 23(3):1375–1380, 2007. [33] Quist, A., Chand, A., Ramachandran, S., Cohen, D., and Lal, R., “Piezoresistive Cantilever Based Nanoflow and Viscosity Sensor for Microchannels.” Lab Chip. 6(11):1450–1454, 2006. [34] Lal, R., Arnsdorf, M., “Multi-Dimensional Atomic Force Microscopy for Drug Discovery: A Versatile Tool for Defining Targets, Designing Therapeutics and Monitoring Their Efficacy.” Life Sciences, in press, 2009.
CHAPTER 13
Microfabricated Devices for Detecting Circulating Tumor Cells in Cancer Patient Blood Samples *
*
Henry Lin , Siyang Zheng , Marija Balic, Richard Cote, Yu Chong Tai, and Ram H. Datar
13.1
Clinical Challenge The clinical outcome of cancer patients is largely influenced by the incidence of distal spread of the disease or metastasis. In patients with primary tumors, this disease relapse is mostly due to clinically occult metastasis present in secondary organs that goes undetected at primary diagnosis even with high-resolution imaging procedures. Early indicators of therapeutic response or resistance are an important issue in oncological practice, but the present imaging techniques are inadequate in that regard. Detection of circulating tumor cells (CTCs) in peripheral blood (PB) or disseminated tumor cells (DTCs) at the single-cell level in bone marrow (BM) could be useful tools in early detection of relapse and response to systemic chemotherapy. Uncontrolled growth of cells is the underlining mechanism leading to the formation of primary tumors; moreover, tumor cells are capable of spreading throughout the body by the process of metastasis, where tumor cells penetrate into lymphatic and blood vessels, circulate through the bloodstream, and then invade normal tissues such as lung, bone and liver. Tumor cells become mobile, a fundamental asset of metastatic cells, by decreasing intercellular adhesion at the primary tissue site, and invade surrounding extracellular stroma, eventually entering into blood vessels and lymphatics. Once in the circulation, they develop physical and biochemical features that enable them to withstand the changes in environment from the primary site, such as the shear forces and the surveillance of immune cells within the circulatory system. Those tumor cells that survive their transit in blood or lymph must extravasate to exit the circulation and invade (enter) distant organs, either by physical means through intravascular cell mass increase resulting in disruption of small capillaries [1] or in a regulated manner via intricate invasive biochemical properties that the tumor has acquired [2]. Once attached within distant organs, these tumor cells develop into overt metastases that begin to successfully adapt to the microenvironment at the new location and grow rapidly, ultimately causing patient death. Since the circulatory system is the organ system frequently employed by the tumor cells for their transit regardless of the origin of the primary
*
Equal contribution by authors.
347
348
Microfabricated Devices for Detecting Circulating Tumor Cells in Cancer Patient Blood Samples
malignancy, developing a reliable technology for enumerating and characterizing CTC in peripheral blood can be greatly beneficial. Sensitive and specific immunocytochemical and molecular assays enable the detection and characterization of DTC and CTC. Immunocytochemical studies have shown statistically significant correlations between DTC detection in bone marrow and blood but bone marrow was more frequently positive than blood [3]. One possible explanation is that bone marrow is a homing organ for DTC, whereas blood analyses allow only a “snapshot” of tumor cell dissemination. Detection of DTC by BM aspiration involves an invasive procedure (and hence far more likely to cause complications and lack of compliance from both patients and physicians), in comparison, PB sampling offers a relatively less invasive option. Therefore, although the rates of DTC detection in the BM far exceed those for CTC detection in the PB [4], the investigation of PB as a target compartment is receiving more attention in recent years, and it is felt that compared to repeated bone marrow aspirations, CTC detection in PB, by sequential analysis if necessary, will be more acceptable to both the patients and the treating clinicians. In metastatic breast cancer patients, for example, a repeated sampling of bone marrow is almost outdated, and the detection of CTC has provided significant prognostic information [5, 6] and seems to be better than the conventional imaging methods to evaluate response [7]. In contrast, the prognostic relevance of CTC in the blood of patients with early-stage disease without overt metastasis is under investigation, with encouraging results from smaller single-center studies [8–11]. Hence, development of technology platforms which improve CTC capture and therefore CTC detection will crucially help enhance the value of CTC as biomarkers of cancer progression.
13.2
Technical Challenge One of the reasons for the lower rate of detection of CTC in blood could be the limitation of the current methodologies, which are suboptimal to capture CTC from blood. Development of methods such as flow-cytometry, magnetic cell separation, and di-electrophoresis has been proposed to increase the yield in the PB [12]. Although in development for many years, most methods are still available essentially only in the research setting, their clinical applicability largely restricted due to lack of efficiency and/or consistency in PB. Recently, the Food and Drug Administration (FDA) approved the CellSearch™ System, which is the only example to be introduced to clinics for monitoring metastatic breast cancer. Therefore, it is clear that technology development plays an important role in disease management in the field of metastasis. The main technical challenge for the detection of CTCs is their extremely low concentration in blood coupled with the task of correctly identifying the “event” as a tumor cell. Human blood normally consists of white blood cells (WBCs) (5–10×106 ml-1), red blood cells (RBCs) (5–9×109 ml-1), and platelets (2.5–4×108 ml-1). The number of CTCs in blood from a healthy person ranges from 0 to 1 ml-1 [13]. Due to the rarity of the CTCs (on the order of only a few cells in every 10 ml of blood), existing techniques lack the sensitivity or efficiency to isolate CTC for further analysis. After successfully isolating the cells of interest, the second
13.3 Techniques for Detecting CTC in Blood
349
challenge is to correctly identify the tumor cells from nontarget cells in order to minimize the risk of false positives, which could generate poor clinical and therapeutic choices having a negative impact on the quality and/or expectancy of life in patients with cancer. The present chapter focuses upon approaches to more effectively detect circulating tumor cells in blood, with particular emphasis on a membrane microfilter device that we have developed for single stage capture. We compare this technology with other optional approaches and further elaborate on the potential of this device to allow genomic analysis.
13.3
Techniques for Detecting CTC in Blood 13.3.1
Cell Enrichment Methods
Enrichment Based on Physical Characteristics
The most established method for tumor cell enrichment from BM and PB samples is performed by cell density gradient centrifugation on the ficoll-hypaque solution [14]. The basis of this cell separation assay is the differential migration of the cells during centrifugation according to their buoyant density, which results in the separation of different cell types in distinct layers. This method has been employed in most key clinical trials evaluating the BM aspirates for the presence of single tumor cells [15]. Some efforts have been made to enhance the efficacy of the enrichment, such as utilization of a commercially available OncoQuick assay. The main advance here is the porous barrier that, prior to centrifugation, separates the lower compartment with the separation medium from the PB sample. Following buoyant density gradient centrifugation, the buff coat containing enriched tumor cells along with the mononuclear lymphocytes can be easily aspirated, keeping the RBCs and the granulocytes partitioned below the porous barrier plug. Dielectrophoresis. Dielectrophoresis (DEP) force is a force exerted on a dielectric particle in a non-uniform electrical field when the particle and the surrounding medium have different polarizabilities. The nonuniform electrical field induces a dipole moment of the particle. For a homogeneous spherical particle with radius r and complex permittivity ε*p in a medium with complex permittivity ε*m and electric r field strength E, the time-dependant DEP force can be expressed as [16]: ÏÔ ε*p - ε*m ¸Ô r 2 F DEP = 2 πr 3 ε m Re Ì * —E * ˝ ÓÔ ε p + 2 ε m ˛Ô
(13.1)
The complex dielectric constant is defined as ε* = ε + jσω , where ε is the dielectric constant, is the electrical conductivity, and ω is the frequency of the electrical field. The terms in the parentheses defines the complex Clausius-Mossotti function, which contains all the frequency dependence of the DEP force. If the electric permittivity of the particles is higher than the surrounding medium, the particles are attracted toward the field maxima (termed positive DEP, or pDEP). On the other hand, if the electric permittivity of the particles is lower than the surrounding
350
Microfabricated Devices for Detecting Circulating Tumor Cells in Cancer Patient Blood Samples
medium, the particles are propelled toward the field minima (termed negative DEP, or nDEP). The direction of the DEP force is determined by the gradient of the electrical field, not the polarity of the field. So both DC and AC fields can be used, although an AC field is commonly used to employ frequency as a discriminating parameter and to prevent electrochemical reaction on the electrode surface. When the AC field is used, the direction and the strength of the force experienced by a particle is frequency dependent and there can be crossover frequencies where the particle experiences no force. DEP forces have been used for cell separation and enrichment. Different cell types are morphologically distinct and have different dielectric phenotypes [17, 18]. The DEP force is sensitive to the size, the shape, cell membrane permeability, and also the internal structure of the cell. The dielectric property of a single cell varies with frequency, and its frequency spectra can be measured with electrorotation [19–22], computerized analysis of DEP motion [23], or multifrequency electrochemical impedance spectroscopy [24, 25]. The AC frequencies used in DEP separation can thus be chosen wisely according to the frequency spectra measurement. For cases where the frequency spectra information of the target cells are unknown or the target cell population is highly heterogeneous, a wide frequency range can be applied spatially or temporally. Typical applications include separation of different cell types (e.g., leukocytes from erythrocytes and different types of leukocytes) [26] and separation of viable cells and dead cells [27]. Size-Based Separation Isolation of CTC based on cell size has been demonstrated to be an efficient, inexpensive, and user-friendly way for enrichment of CTCs [28–32], dating back to 1960s [33, 34]. Circulating epithelial tumor cells are significantly larger than the surrounding blood cells [31, 35, 36], where a pore diameter of ~8 μm has been shown to be optimal for CTC retention [37]. However, the commercially available polycarbonate filters are fabricated using track etching [38], which results in random placement of pores with relatively low density that often results in fusion of two or more pores, which reduces CTC capture efficiency to 50% to 60% [35, 39]. The design and performance of our microdevice will be discussed in detail later in this chapter. Antibody-Based Enrichment
CTC enrichment by antibody-based methods relies on the expression of specific antigens on the surface of epithelial tumor cells or mononuclear hematopoietic cells. Magnetic separation involves either positive selection via direct epithelial tumor cell capture or negative selection by hematopoietic cell depletion [40, 41]. The antibodies employed in the positive selection methods target the epithelial tumor cell surface markers, while those used in negative selection assays are directed against the surface markers expressed abundantly in hematopoietic cells of different lineages. Currently, antibodies are coupled to magnetic particles by either direct functionalization or through DNA linkers, and there are multiple different ways to enrich target tumor cells, such as a column format, where the target or nontarget
13.3 Techniques for Detecting CTC in Blood
351
cells are attached to magnetized particles that are further separated using magnetic automated cell sorter (MACS) systems (Miltenyi Biotec, Germany), or immunomagentic particle-based separations, like Dynabeads (Dynal, Norway), or RosetteSep (Stem Cell Technologies, Canada). When interacting with samples for the positive selection assays, the antibody-coated ferromagnetic beads enable magnetic capture of tumor cells [42]. In the literature, the most targeted marker described on cell surface of tumor cells is epithelial cell adhesion molecule (EpCAM, also known by various other names such as HEA, GA733.2, CO17-1A, KSA, KS1-4, and Ber-Ep4), which is expressed in most carcinomas of epithelial origin. There are some recent reports on the use of a panel of monoclonal antibodies [43, 44], striving for better recovery rates in model systems, and therefore detecting CTC in more patients. This may also help make the magnetic separation fail-safe by alleviating the dependence on one or a few markers (such as EpCAM) which are known to be variably expressed among some primary tumor cells [45, 46]. More recently, Nagrath et al. engineered a microfluidic chip with increased surface area for cells to interact with EpCAM by placing functionalized microposts that are 100-μm tall and 100 μm in diameter along the fluidic path [47]. Using such a device, the numbers of CTCs recovered from patients of various malignancies are 100- to 1,000-fold higher than the current literature reports with 50% purity. As alluded to above, another way of immunomagnetic enrichment is depletion of nontumor cells using their specific markers (negative selection). Most commonly used markers in this setting are CD45 for lymphocytes and glycophorin for erythrocytes [48]. Often, a cocktail of antibodies to multiple antigens is used, as in the RosetteSep assay (Stem Cell Technologies). Immunomagnetic enrichment strategies are nevertheless limited by the fact that carcinoma cells with absent or low target antigen expression may be missed by these methods. Furthermore, subpopulations of hematopoietic cells also express epithelial markers and copurify with tumor cells in the sample. Although there is substantial data on different antibody-based enrichment techniques, none of the methods described above is considered very efficient. Many trials have shown that the recovery rates in model systems (consisting of cultured tumor cells spiked into normal donor blood) range widely between 10% to 90%, with the results being frequently inconsistent. Recently, the development of an automated flow-based immunomagnetic detection system has gained popularity. The automated system developed by Immunicon scientists uses antibody-coated ferrous particles to separate EpCAM-expressing epithelial cells from whole blood specimens. Confirmation that these are indeed epithelial cells is provided by automated fluorescent microscopy after staining with fluorescently labeled monoclonal antibodies against cytokeratins and CD45 (a pan-leukocyte marker) as well as DAPI. A trained observer reviews computer-generated composite images of the detected events. Cytokeratin-positive, DAPI-positive events are tallied by the computer as epithelial cells and expressed as “CTC/mL of blood collected” (usually 7.5 mL in a 10-mL Vacutainer), while “events” that fail to stain for cytokeratin or DAPI, or that stain for CD45, are not. With this system, an FDA-approval was successfully obtained for the application of monitoring metastatic breast cancer patients.
352
Microfabricated Devices for Detecting Circulating Tumor Cells in Cancer Patient Blood Samples
13.3.2
Microfabricated Devices
Affinity-Based Chip
An affinity-based microfluidic chip (CTC chip) was developed for CTC enrichment from cancer patient blood samples [47]. The microfluidic system includes a microfluidic chip to perform the enrichment, a manifold to enclose the chip, and a pneumatic pump to establish the flow. The main part of the microfluidic device is a chamber filled with an equilateral triangular array of microposts fabricated with deep reactive ion etching (DRIE) in silicon. The dimensions of the chip are 25 mm by 66 mm, with an active capture area of 19 mm by 51 mm. The array incorporates 78,000 microposts within a surface area of 970 mm2. The microposts are 100-μm tall and 100 μm in diameter with an average 50-μm gap between them. Special considerations were made to limit the flow velocity to allow enough cell–micropost interaction time, also to reduce the shear stress to ensure maximum cell–micropost attachment. The microchip is functionalized with EpCAM antibodies for CTC capture. The device is reported to have >60% capture efficiency with sample flow rate less than 2 mL/hour and the maximal capture efficiency is around 65%, which is compatible with the capture efficiency reported for immunomagnetic based enrichment. Also in a small pool of metastatic cancer patients under systemic treatment, the CTC numbers identified after the microfluidic chip enrichment correlates well with the clinical course of the disease measured by standard radiographic methods. Later the same system was used to study 27 cases of metastaic nonsmall-cell lung cancer [49]. After enrichment, DNA is recovered from the CTCs and the epidermal growth factor receptor (EGFR) mutational analysis using allele-specific polymerase-chain-reaction amplification is performed to detect the drug resistant T790 mutation. The results are compared with concurrently isolated free plasma DNA and the original tumor-biopsy specimens. The existence of the mutation correlates well with reduced progression-free survival. Also an increase of CTCs is associated with tumor progression and in some cases the emergence of additional EGFR mutations in serial analysis; while a reduction of CTCs is associated with successful radiographic tumor response. Deterministic Flow
A novel type of microfluidic device based on deterministic hydrodynamic flow has successfully demonstrated continuous, precise separation of particles based on size [50–53]. The device is composed of a micropost array inside a microfluidic chamber. The shape of the micropost is normally circular. The diameter of the micropost, the distance between the micropost in each individual row, and the row-to-row shift are important design parameters. For each design parameter set, there is a corresponding critical hydrodynamic diameter and thus the separation is binary. The operating principle of the device depends on establishment of a stable hydrodynamic field inside the device. Due to the scale of the device (in the micrometer range), the Reynolds number of the fluid flow under normal device operation is very low. Thus the flow is laminar and a particle inside the device follows the streamline of the flow until it interacts with the device wall or other particles. All the particles are focused (mainly hydrodynamically) to a central stream before entering the separation region of the device. For particles below the critical hydrodynamic diameter of the device, it
13.3 Techniques for Detecting CTC in Blood
353
follows streamlines cyclically through the gaps, moving in an average downward flow direction. Particles above the critical hydrodynamic diameter do not fit into the streamline and are bumped during the interaction with the micropost into the sequential streamline. Consequently, particles below and above the critical hydrodynamic diameter follow different trajectories and are separated after a certain distance inside the device. Regions of microposts with different design parameter sets can be cascaded to allow separation of particles of various sizes. The device can be fabricated with different material (e.g., silicon, PDMS), although the surface properties of the material may influence the interaction between the particles and the device wall, thus affecting the separation efficiency and the lifetime of the device. Soft lithography using PDMS seems to be a cost effective way for batch fabrication and the PDMS material itself is compatible with the majority of biological samples. These type of devices have been used to separate blood plasma from blood cells, different types of blood cells (erythrocytes, leukocytes, and platelets), and DNA fragments of different size. One limitation preventing this device from being used in CTC enrichment of clinical samples is that the volume flow rates are reported up to 1 μl/min, corresponding to a flow velocity of 1 mm/sec for a single device. This flow velocity is roughly in the range of the flow velocity of physiological blood. So the shear generated by the flow is expected to be less likely to cause cell damage and lysis. Parallel processing and reducing the size of each individual processing unit might enable this technology to be used for CTC enrichment of clinical samples. DEP Chip
Gascoyne’s group has been using microfabricated dielectrophoresis field-flow fractionation (DEP-FFF) devices for cell separation since the 1990s [26, 54–56]. The separations were accomplished using thin, flat chambers having microelectrode arrays on the bottom wall. The devices are made by bonding two 2-mm-thick glass plates to a 100-μm-thick Teflon gasket by UV-curing Epoxy glue. The top surface of the bottom glass plate is patterned with interdigitated electrodes by microfabrication. The DEP forces are generated by applying AC fields to the electrodes. An AC frequency is chosen so that the cells that experience nDEP forces are kept away from the bottom electrodes. The nDEP force and the hydrodynamic lift force exerted on the cell are balanced by the sedimentation force. At different distances from the bottom electrode, the nDEP force and the hydrodynamic force vary. Cells with similar properties will find their steady-state position at a specific distance from the bottom plate electrodes. Since the flow profile of the cross section along the flow direction is almost a perfect parabola, particles flowing in fluid layers at different distances from the bottom plate will have different flow velocities. Consequently, cells of different types are separated without interfering with their viability according to their dielectric properties and hydrodynamic flow properties. This technique has demonstrated 100% separation efficiency when human breast cancer MDA-231 cells were spiked in blood with a tumor cell to blood cell ratio of 1:105 [18, 57]. The DEP mean crossover frequencies of MDA-231 cells were found to be 15, 58, and 95 kHz by electrorotation measurements in low conductivity media, which is significantly different from those of T lymphocytes and erythrocytes. The
354
Microfabricated Devices for Detecting Circulating Tumor Cells in Cancer Patient Blood Samples
rates of separation are reported to be at least 103 cells/s. The limitations of the approach for actual patient sample applications include small sample volume, and the requirement of separating the cellular components from plasma and their resuspension in low conductivity isotonic sucrose and dextrose based medium. Size-Based Microfluidic Separation
To address the current limitations in CTC isolation, our group has developed a novel and simple microdevice with a size-based filter for isolation of CTC that has the potential for integration with downstream RNA, DNA, and protein characterization functionalities to truly achieve single station analysis with minimal intervention. Most disseminated tumor cells of epithelial origin are larger in comparison to hematopoietic cells in the background. Thus, a size-based filter can exploit this physical difference to isolate CTC in a simple, fast, and antibody-independent manner. Our initial results using a size-based filter show recovery rates of >85% for capturing cultured prostate cancer cells (LNCaP) spiked in 1 and 5 ml of neat, undiluted blood requiring only 10 minutes for processing each sample. Whereas the idea of exploiting the tumor cell size for enrichment of tumor cells from the background cells has already been shown as a potentially good parameter [58], it is the ability to fabricate high-density pores that enabled us to enhance the recovery rate and enrichment. This novel technology may represent an important advance as it may result in a portable, cost-effective, and simple device to capture and characterize CTC. Details of the Microfilter Device. Isolation of CTC based on cell size using polycarbonate filters has been demonstrated to be an efficient, inexpensive and user-friendly way for enrichment of CTCs [29–32, 59] by exploiting the fact that circulating epithelial tumor cells are significantly larger than the surrounding blood cells [35]. These polycarbonate filters are fabricated with track etching [60], which results in random placement of pores with relatively low density. Track etching often results in fusion of two or more pores which reduces CTC capture efficiency to 50% to 60% [39]. There are two main challenges in building a microfabricated system for CTC isolation. First is the requirement for high efficiency of CTC recovery and the effective separation from blood cells. The detection sensitivity required is high, with the ability to capture as few as one CTC in 7.5 mL of whole blood, which contains about 10 billion blood cells. Second, the sample volume required to be processed is in the milliliter range, while microdevices are normally used to process nanoliter or even femtoliter volumes of sample. Such a challenge is further exacerbated when dilution of blood is required. Filters made with microfabrication technologies have several advantages for CTCs capture. Unlike the track-etched polycarbonate filters, the size, geometry, and density of the pores can be precisely controlled. With batch fabrication, this technology can be very cost effective, which makes it suitable to develop a device for routine testing in clinics. The filter with uniformly spaced pores of identical diameter can allow maximum parallel processing, which reduces processing time and filter clogging due to back-pressure. Several distinct properties make it one of the best candidates for this application. First, as the highest USP class IV biocompatible polymer
13.3 Techniques for Detecting CTC in Blood
355
for implementation, biofouling is expected to be minimal for parylene-C. This strong but flexible material has excellent mechanical properties. It has a Young’s modulus of 4GPa and high malleability that can withstand up to 200% elongation. It also has desirable electrical properties with a low dielectric constant and high resistivity, which make it a good isolating material for electronics. Unlike the opaque polycarbonate filters, parylene is transparent in the UV and visible range, which enables staining and observation of captured CTCs directly on the membrane without transferring captured CTCs to glass slides; this translates into minimal cell loss. Finally, we have established processing technologies to fabricate the filters. With room temperature, conformal and pinhole-free deposition, high-quality parylene-C film can be routinely obtained. Metal deposition and oxygen plasma etching in a reactive ion etching (RIE) system makes it possible to be integrated with further downstream analysis directly on a single chip. During microfabrication, photoresist is spin-coated onto a silicon wafer. Next, parylene-C is conformally deposited to 10-μm thickness and patterned with oxygen plasma etching using a reactive ion etching (RIE) technique, where either AZ9260 or Cr/Au is used as a mask layer. Finally, the whole film is released in acetone or photoresist stripper at 80° overnight. Within a 6×6 mm surface, 40,000 pores are evenly distributed. Several different pore sizes and shapes were initially tested. We concluded that the round pore shape with a 7- to 8-μm diameter was ideal for capture and characterization of CTC on the membrane. The final product of this process is a wafer with a set of 40 to 50 filters as illustrated in Figure 13.1(a) and the device is assembled by sandwiching the individual microfilter between two PDMS jigs as shown in Figure 13.1(c). We have demonstrated employment of a novel parylene membrane for CTC capture as shown in Figure 13.1(b) with recovery rates higher or comparable to existing technologies [61]. Incorporation of On-Chip Cell Lysis Functionality for Captured CTC. Cell lysis in microdevices has been demonstrated based on various principles. Chemical lysis was achieved by mixing with lysis buffer [62–64] or local hydroxide electro-generation [65]. Flowing cells through nanostructured barbs [66] or spinning them with beads can affect mechanical cell lysis [67]. Finally bacterial, yeast, and mammalian cells have been electrolyzed with either DC or AC signals on microchips [68–73]. Electrolysis has the advantages of not requiring additional chemicals or mixing, no additional moving structures need to be introduced, and the electrodes are prefabricated. Micro-electrical cell lysis devices also have the advantage of lower applied electrical voltage relative to macroscopic electrical cell lysis instruments, which minimizes electrode damage and water hydrolysis of target cells, while also reducing operator-risk. But even for microdevices, the working potential is much higher than the voltage threshold for water hydrolysis (~1 V), so a rapid alternating signal is normally preferred to minimize gas bubble formation inside the device and the extreme pH conditions close to the electrodes. Unlike previous microcell electrolysis devices, in our device the cells are lysed in situ on the membrane, as shown in Figure 13.2, instead of lysing them inside physically distinct fluidic channels or chambers, which typically can result in biomaterial loss. Our approach also has the potential of improving the lysis efficiency due to minimal cell movement and a lower working voltage. Highly efficient on-chip electrical lysis
356
Microfabricated Devices for Detecting Circulating Tumor Cells in Cancer Patient Blood Samples
(a)
(b) PDMS top chamber
Acrylic jig top
Cr/Au electrode Parylene membrane filter Clamp (c)
PDMS bottom chamber
PEEK jig bottom
Figure 13.1 Illustration of device assembly. (a) Bright field image of an optically transparent parylene filter with uniformly shaped and spaced 8-μm pores. (b) SEM picture of single cultured tumor cell captured on the membrane. (c) Schematic drawing of a functional microdevice consists of parylene membrane filter sandwiched between rectangular PDMS slabs and clamped in between acrylic jigs with inlet and outlet for syringes.
after cell capture was confirmed optically with a microscope and biochemically by successful genomic DNA PCR and RT-PCR for captured tumor cells in a model system to prove the feasibility of downstream genomic analysis. A microfabricated device such as this can become a single-station assay system with an ability to capture, concentrate, quantify, and characterize CTC; and render these latter crucial biomarkers of cancer progression available for clinicians to monitor disease progression as well as therapeutic response in a cancer patient, a facility very much needed for effective management of this dreadful disease.
13.4
Clinical Value of CTC Capture and Characterization The hematogenous spread of cells is among the most important factors affecting the outcome of patients with invasive cancer [74, 75]. A proportion of patients with no evidence of tumor dissemination by standard clinical, radiographic, physical, or pathological assessment will develop recurrent disease after primary curative therapy. It is the most accepted theory that disseminated cells remain dormant at distant sites, and are reactivated through accumulating genetic changes and interactions with the surrounding microenvironment. The recently published clinical trial from the ABCSG group (ABCSG 12) has gained a lot of attention [76]. It was shown that zoledronic acid given at adjuvant stage in breast cancer patients has a positive impact on overall survival. This result may be interpreted as a potential interaction
13.4 Clinical Value of CTC Capture and Characterization
357
Electrodes
Figure 13.2 (Color plate 20) On-chip electrolysis. Bottom panel shows bright field image of parylene filter with integrated electrodes, dark stripes across the image, for electrolysis of captured tumor cells, stained with dark blue hematoxylin nuclear dye. Top panel is a schematic of a single hole with electrodes.
of zoledronic acid with the microenvironment or systemic environment and disseminated tumor cells. While this in only a preliminary observation, together with recent progress of the research on micrometastases, it finally demonstrates the clinical importance to further characterize CTC, their molecular profiles, and correlate them in the context of the micro/ systemic environment [77]. Several recent studies have shown the prognostic relevance of CTC in metastatic breast cancer patients as well as their usefulness for monitoring therapeutic success [5]. After the CellSearch technique was introduced and the first clinical trial demonstrating the clinical significance of CTC in metastatic breast cancer patients published, several additional trials in other cancer entities followed. Among those, prognostic significance of CTC was shown in prostate and colorectal cancer [78, 79]. An ability to clinically use a technology platform for effective CTC enrichment and enumeration, including a capability of further characterization of CTC, will therefore have profound impact in the practice of oncology by enabling prognostication as well as monitoring the therapeutic responses of patients to currently available and novel anticancer drugs.
358
13.5
Microfabricated Devices for Detecting Circulating Tumor Cells in Cancer Patient Blood Samples
Cancer Stem Cells and Metastasis Increasing evidence in the past decade has alerted us to existence of cancer stem cells or cancer initiating cells. Such cells have been identified in many hematological malignancies and some solid tumors [80–83]. Additionally, their presence has increasingly been associated with the resistance of cancer to conventional therapeutic strategies [84]. Putative breast cancer stem cells were shown to be associated with the CD44+CD24-/low phenotype in the study performed by Al Hajj et al. [81]. After obtaining single cell solutions from primary tumors or metastatic formations and sorting the cells with flow-cytometry based on expression of superficial markers, the CD44+CD24-/low subpopulation was found to be highly tumorigenic and capable of self-renewal and differentiation in animal models [81]. More recently a putative breast cancer stem cell subpopulation was associated with the presence of the enzyme aldehyde dehydrogenase [85], providing a novel possibility to identify a more potent subpopulation within the CTC population. In addition to the findings by Al Hajj et al., primary breast carcinoma-derived cell cultures were shown to encompass undifferentiated cells capable of self-renewal, extensive proliferation as clonal nonadherent spherical clusters, and differentiation along different mammary epithelial lineages (ductal and myoepithelial). These cultured cells were also shown to have the CD44+CD24– phenotype [86]. DTC in bone marrow, as representatives of micrometastases, are considered “virtual” targets for adjuvant treatment strategies [87]. On the other hand, cancer stem cells among the DTCs may be responsible for resistance of the cancer to therapy. Hypothesizing that cancer stem cells may also be the cells that metastasize from primary breast cancer to distant locations, we performed the analysis of DTC for the described putative breast cancer stem cell phenotype [88]. In an analysis of 50 DTC positive bone marrow specimens from early breast cancer patients, in addition to pancytokeratin we performed immunohistochemistry for CD44 and CD24. Surprisingly, we detected the putative stem cell phenotype in all cytokeratin expressing cells, with the mean prevalence of 72% among the overall disseminated tumor cells per patient. This finding suggests that DTC may be selectively enriched for breast cancer stem cells. The DTC, and conversely the CTC, should in the future be analyzed for cancer stem cell features, which may facilitate identification of novel therapeutic targets. It is important to emphasize here that such positive identification of cancer stem cells among CTC will only be feasible by means of novel CTC capture technologies such as the ones described here.
13.6
Application of Nanotechnology in CTC Capture With the ability of controlled fabrication in the nanometer range, a plethora of tools described in other chapters in this book are available for the detection and characterization of CTC. While it is likely that nanotechnological advances may be useful in enhancing the CTC capture surfaces (in the microfilter platform for example, patterning a surface with a series of nanoscale bumps can reduce biofouling of surfaces and allow selective CTC adhesion), it is in the downstream analytical steps that nanotechnology will probably have most applicability. Quantum dots (QDs) are
13.7 Conclusion
359
semiconducting inorganic nanocrystals that can be tuned by varying the size or the material composition to exhibit unique fluorescent properties that allow single excitation bandwidth with multiple color QDs. Moreover, employing QDs with different colors, each with a narrow symmetric emission spectrum, maximizes the number of biomolecules to be interrogated when the QDs are coupled with antibodies. Due to heterogeneous nature of the cancer cells, the use of QDs for multimarker staining of CTC can further characterize the subpopulation of cells that are responsible for causing overt metastasis such as cancer stem cells. Magnetic nanoparticles that are less than 30 to 40 nm in diameter exhibit superparamagnetism, where particles are strongly magnetized only in the presence of an external magnetic field. Although such magnetic nanoparticles are currently employed for cell capture as described earlier, advances in the synthesis of uniform nanoparticles may be exploited to enable more efficient CTC separation. In addition, a variety of nanosensors with unprecedented sensitivities such as nanowires can be integrated downstream of a CTC capture chip for further single cellular molecular analysis. With the advancement in sensing and separation methods using nanotechnology, biologists are empowered with tools to gain new insights to understand the process of metastasis, with the goal of better patient management by oncologists subsequently.
13.7
Conclusion Metastasis is the primary reason of treatment failure and eventual mortality in the majority of advanced cancer patients. CTC detection systems could be useful tools in early detection of disease relapse and response to systemic chemotherapy for cancer. Even though serum protein or RNA tumor markers have been used for the early diagnosis of metastases, their systematic determination has not had an effect on survival. Methods that are more reliable are needed to detect occult metastases earlier than with the common clinical methods. Enhancing the CTC detection by their cellular and molecular characterization will permit understanding of such important phenomena as existence of cancer stem cells among the CTC, which can be hoped to allow oncologists to define better targets for therapy and start appropriate treatment even before an overt relapse. A technology platform that combines effective detection and characterization of CTC in patient blood samples will thus go a long way in cancer management.
Problems 13.1 This multipart question refers to the antibody method of capturing CTC. The question might be answered by surveying the literature or by performing experiments. (a) What is the optimal spatial density of antibodies to capture CTC? (b) What is the optimal surface shape for capturing CTC, flat, curved, and so forth? (c) How can functionalization of the surface with antibody be characterized? (d) Does the length of the
360
Microfabricated Devices for Detecting Circulating Tumor Cells in Cancer Patient Blood Samples
linker molecule that connects the antibody to the surface have an effect on capture of CTC? 13.2 Discuss the feasibility of electronic detection of CTC. Assume that a surface (posts or flat) is functionalized with antibody to capture CTC and the surface is electrically conductive. Could CTC be captured and counted electronically without the need for an optical step thus simplifying the sensor? 13.3 Try to think of ways that nanotechnology can play a role to improve CTC detection and characterization. 13.4 What could be an advantage of an implantable CTC sensor that stays in the body for a certain length of time?
References [1] Al-Mehdi, A.B., et al., “Intravascular origin of metastasis from the proliferation of endothelium-attached tumor cells: A new model for metastasis.” Nat Med, 2000. 6(1): p. 100-2. [2] Kim, Y.J., et al., “P-selectin deficiency attenuates tumor growth and metastasis.” Proc Natl Acad Sci U S A, 1998. 95(16): p. 9325-30. [3] Muller, V., et al., “Circulating tumor cells in breast cancer: correlation to bone marrow micrometastases, heterogeneous response to systemic therapy and low proliferative activity.” Clin Cancer Res, 2005. 11(10): p. 3678-85. [4] Pierga, J.Y., et al., “Clinical significance of immunocytochemical detection of tumor cells using digital microscopy in peripheral blood and bone marrow of breast cancer patients.” Clin Cancer Res, 2004. 10(4): p. 1392-400. [5] Cristofanilli, M., et al., “Circulating tumor cells, disease progression, and survival in metastatic breast cancer.” N Engl J Med, 2004. 351(8): p. 781-91. [6] Hayes, D.F., et al., “Circulating tumor cells at each follow-up time point during therapy of metastatic breast cancer patients predict progression-free and overall survival.” Clin Cancer Res, 2006. 12(14 Pt 1): p. 4218-24. [7] Braun, S. and B. Naume, “Circulating and disseminated tumor cells.” J Clin Oncol, 2005. 23(8): p. 1623-6. [8] Cristofanilli, M. and J. Mendelsohn, “Circulating tumor cells in breast cancer: Advanced tools for ‘tailored’ therapy?” Proc Natl Acad Sci U S A, 2006. 103(46): p. 17073-4. [9] Ntoulia, M., et al., “Detection of Mammaglobin A-mRNA-positive circulating tumor cells in peripheral blood of patients with operable breast cancer with nested RT-PCR.” Clin Biochem, 2006. 39(9): p. 879-87. [10] Quintela-Fandino, M., et al., “Breast cancer-specific mRNA transcripts presence in peripheral blood after adjuvant chemotherapy predicts poor survival among high-risk breast cancer patients treated with high-dose chemotherapy with peripheral blood stem cell support.” J Clin Oncol, 2006. 24(22): p. 3611-8. Xenidis, N., et al., “Predictive and prognostic value of peripheral blood cytokeratin-19 [11] mRNA-positive cells detected by real-time polymerase chain reaction in node-negative breast cancer patients.” J Clin Oncol, 2006. 24(23): p. 3756-62. Gilbey, A.M., et al., “The detection of circulating breast cancer cells in blood.” J Clin [12] Pathol, 2004. 57(9): p. 903-11. [13] Allard, W.J., et al., “Tumor cells circulate in the peripheral blood of all major carcinomas but not in healthy subjects or patients with nonmalignant diseases.” Clin Cancer Res, 2004. 10(20): p. 6897-904.
References
361
[14] Rosenberg, R., et al., “Comparison of two density gradient centrifugation systems for the enrichment of disseminated tumor cells in blood.” Cytometry, 2002. 49(4): p. 150-8. [15] Braun, S., et al., “Cytokeratin-positive cells in the bone marrow and survival of patients with stage I, II, or III breast cancer.” N Engl J Med, 2000. 342(8): p. 525-33. [16] Jones, T.B., Electromechanics ofParticles 1995, Cambridge: Cambridge University Press. [17] Gascoyne, P.R.C. and J.V. Vykoukal, “Dielectrophoresis-based sample handling in general-purpose programmable diagnostic instruments.” Proceedings of the IEEE, 2004. 92(1): p. 22-42. [18] Gascoyne, P.R.C., et al., “Dielectrophoretic separation of cancer cells from blood.” IEEE Transactions on Industry Applications, 1997. 33(3): p. 670-678. [19] H. Ziervogel, R.G., D. Schadow, and S. Heymann, “Electrorotation of lymphocytes-The influence of membrane events and nucleus.” Biosci. Rep., 1986. 6(11): p. 973-982. Archer, S., H. Morgan, and F.J. Rixon, “Electrorotation studies of baby hamster kidney [20] fibroblasts infected with herpes simplex virus type 1.” Biophysical Journal, 1999. 76(5): p. 2833-2842. [21] Griffith, A.W. and J.M. Cooper, “Single-cell measurements of human neutrophil activation using electrorotation.” Anal Chem, 1998. 70(13): p. 2607-2612. [22] Yang, J., et al., “Dielectric properties of human leukocyte subpopulations determined by electrorotation as a cell separation criterion.” Biophysical Journal, 1999. 76(6): p. 3307-3314. [23] Pethig, R., et al., “Dielectrophoretic studies of the activation of human T lymphocytes using a newly developed cell profiling system.” Electrophoresis, 2002. 23(13): p. 2057-2063. [24] Gawad, S., L. Schild, and P.H. Renaud, “Micromachined impedance spectroscopy flow cytometer for cell analysis and particle sizing.” Lab Chip, 2001. 1(1): p. 76-82. [25] Han, A., L. Yang, and A.B. Frazier, “Quantification of the heterogeneity in breast cancer cell lines using whole-cell impedance spectroscopy.” Clin Cancer Res, 2007. 13(1): p. 139-43. [26] Yang, J., et al., “Differential analysis of human leukocytes by dielectrophoretic field-flow-fractionation.” Biophysical Journal, 2000. 78(5): p. 2680-2689. [27] Markx, G.H., M.S. Talary, and R. Pethig, “Separation of Viable and Nonviable Yeast Using Dielectrophoresis.” Journal of Biotechnology, 1994. 32(1): p. 29-37. Pinzani, P., et al., “Isolation by size of epithelial tumor cells in peripheral blood of patients [28] with breast cancer: correlation with real-time reverse transcriptase-polymerase chain reaction results and feasibility of molecular analysis by laser microdissection.” Human Pathology, 2006. 37(6): p. 711-718. [29] Vona, G., et al., “Enrichment, immunomorphological, and genetic characterization of fetal cells circulating in maternal blood.” American Journal of Pathology, 2002. 160(1): p. 51-58. [30] Vona, G., et al., “Impact of cytomorphological detection of circulating tumor cells in patients with liver cancer.” Hepatology, 2004. 39(3): p. 792-797. [31] Vona, G., et al., “Isolation by size of epithelial tumor cells—A new method for the immunomorphological and molecular characterization of circulating tumor cells.” American Journal of Pathology, 2000. 156(1): p. 57-63. [32] Kahn, H.J., et al., “Enumeration of circulating tumor cells in the blood of breast cancer patients after filtration enrichment: correlation with disease stage.” Breast Cancer Research and Treatment, 2004. 86(3): p. 237-247. Seal, S.H., “A sieve for the isolation of cancer cells and other large cells from the blood.” [33] Cancer, 1964. 17(5): p. 637-642.
362
Microfabricated Devices for Detecting Circulating Tumor Cells in Cancer Patient Blood Samples [34] Fleischer, R.L., P.B. Price, and R.M. Walker, “Tracks of charged particles in solids.” Science, 1965. 149(3682): p. 383-393. [35] Lara, O., et al., “Enrichment of rare cancer cells through depletion of normal cells using density and flow-through, immunomagnetic cell separation.” Experimental Hematology, 2004. 32(10): p. 891-904. [36] Lin, H.K., et al., “Portable filter-based microdevice for detection of circulating tumor cells.” submitted, 2008. [37] Zabaglo, L., et al., “Cell filtration-laser scanning cytometry for the characterisation of circulating breast cancer cells.” Cytometry Part A, 2003. 55A(2): p. 102-108. [38] Fleischer, R.L., et al., “Particle Track Etching.” Science, 1972. 178(4058): p. 255-&. [39] Rostagno, P., et al., “Detection of rare circulating breast cancer cells by filtration cytometry and identification by DNA content: Sensitivity in an experimental model.” Anticancer Research, 1997. 17(4A): p. 2481-2485. [40] Iinuma, H., et al., “Detection of tumor cells in blood using CD45 magnetic cell separation followed by nested mutant allele-specific amplification of p53 and K-ras genes in patients with colorectal cancer.” Int J Cancer, 2000. 89(4): p. 337-44. [41] Siewert, C., et al., “Rapid enrichment and detection of melanoma cells from peripheral blood mononuclear cells by a new assay combining immunomagnetic cell sorting and immunocytochemical staining.” Recent Results Cancer Res, 2001. 158: p. 51-60. [42] Naume, B., et al., “Immunomagnetic techniques for the enrichment and detection of isolated breast carcinoma cells in bone marrow and peripheral blood.” J Hematother, 1997. 6(2): p. 103-14. [43] Deng, G., et al., “Enrichment with anti-cytokeratin alone or combined with anti-EpCAM antibodies significantly increases the sensitivity for circulating tumor cell detection in metastatic breast cancer patients.” Breast Cancer Res, 2008. 10(4): p. R69. [44] Woelfle, U., et al., “Bi-specific immunomagnetic enrichment of micrometastatic tumour cell clusters from bone marrow of cancer patients.” J Immunol Methods, 2005. 300(1-2): p. 136-45. [45] Went, P.T., et al., “Frequent EpCam protein expression in human carcinomas.” Hum Pathol, 2004. 35(1): p. 122-8. Gottschalk, J., et al., “HEA 125 and Ber EP4: two monoclonal anti-epithelial, [46] non-cytokeratin antibodies distinguishing metastatic carcinomas from glial tumors.” Clin Neuropathol, 1993. 12(2): p. 68-72. [47] Nagrath, S., et al., “Isolation of rare circulating tumour cells in cancer patients by microchip technology.” Nature, 2007. 450(7173): p. 1235-9. [48] de Cremoux, P., et al., “Detection of MUC1-expressing mammary carcinoma cells in the peripheral blood of breast cancer patients by real-time polymerase chain reaction.” Clin Cancer Res, 2000. 6(8): p. 3117-22. [49] Maheswaran, S., et al., “Detection of mutations in EGFR in circulating lung-cancer cells.” N Engl J Med, 2008. 359(4): p. 366-77. [50] Huang, L.R., et al., “Continuous particle separation through deterministic lateral displacement.” Science, 2004. 304(5673): p. 987-990. [51] Davis, J.A., et al., “Deterministic hydrodynamics: Taking blood apart.” Proc Natl Acad Sci U S A, 2006. 103(40): p. 14779-14784. Inglis, D.W., et al., “Critical particle size for fractionation by deterministic lateral displace[52] ment.” Lab on a Chip, 2006. 6(5): p. 655-658. [53] S. Zheng, R.Y., Y. C. Tai, and H. Kasdan. “Deterministic lateral displacement MEMS device for continuous blood cell separation,” 18th IEEE International Conference on Micro Electro Mechanical Systems (MEMS ‘05), 2005. Miami Beach, FL.
References
363
[54] Huang, Y., et al., “Introducing dielectrophoresis as a new force field for field-flow fractionation.” Biophysical Journal, 1997. 73(2): p. 1118-1129. [55] Markx, G.H., J. Rousselet, and R. Pethig, “DEP-FFF: Field-flow fractionation using non-uniform electric fields.” Journal of Liquid Chromatography & Related Technologies, 1997. 20(16-17): p. 2857-2872. [56] Wang, X.B., et al., “Cell separation by dielectrophoretic field-flow-fractionation.” Anal Chem, 2000. 72(4): p. 832-839. [57] Becker, F.F., et al., “Separation of Human Breast-Cancer Cells from Blood by Differential Dielectric Affinity,” Proc Natl Acad Sci U S A, 1995. 92(3): p. 860-864. [58] Fleischer, R.L., “Cancer filter deja vu.” Science, 2007. 318(5858): p. 1864. [59] Pinzani, P., et al., “Isolation by size of epithelial tumor cells in peripheral blood of patients with breast cancer: correlation with real-time reverse transcriptase-polymerase chain reaction results and feasibility of molecular analysis by laser microdissection.” Hum Pathol, 2006. 37(6): p. 711-8. [60] Fleischer, R.L., et al., “Particle track etching.” Science, 1972. 178(58): p. 255-63. [61] Zheng, S., et al., “Membrane microfilter device for selective capture, electrolysis and genomic analysis of human circulating tumor cells.” J Chromatogr A, 2007. 1162(2): p. 154-61. [62] El-Ali, J., et al., “Cell stimulus and lysis in a microfluidic device with segmented gas-liquid flow.” Anal Chem, 2005. 77(11): p. 3629-36. Irimia, D., R.G. Tompkins, and M. Toner, “Single-cell chemical lysis in picoliter-scale [63] closed volumes using a microfabricated device.” Anal Chem, 2004. 76(20): p. 6137-43. [64] Schilling, E.A., A.E. Kamholz, and P. Yager, “Cell lysis and protein extraction in a microfluidic device with detection by a fluorogenic enzyme assay.” Anal Chem, 2002. 74(8): p. 1798-804. [65] Di Carlo, D., et al., “On-chip cell lysis by local hydroxide generation.” Lab Chip, 2005. 5(2): p. 171-8. [66] Di Carlo, D., K.H. Jeong, and L.P. Lee, “Reagentless mechanical cell lysis by nanoscale barbs in microchannels for sample preparation.” Lab Chip, 2003. 3(4): p. 287-91. [67] Kim, J., et al., “Cell lysis on a microfluidic CD (compact disc).” Lab Chip, 2004. 4(5): p. 516-22. [68] Cheng, J., et al., “Isolation of cultured cervical carcinoma cells mixed with peripheral blood cells on a bioelectronic chip.” Anal Chem, 1998. 70(11): p. 2321-6. [69] Cheng, J., et al., “Preparation and hybridization analysis of DNA/RNA from E. coli on microfabricated bioelectronic chips.” Nat Biotechnol, 1998. 16(6): p. 541-6. [70] Lee, S.W. and Y.C. Tai, “A micro cell lysis device.” Sensors and Actuators a-Physical, 1999. 73(1-2): p. 74-79. [71] Han, F., et al., “Fast electrical lysis of cells for capillary electrophoresis.” Anal Chem, 2003. 75(15): p. 3688-96. [72] Lu, H., M.A. Schmidt, and K.F. Jensen, “A microfluidic electroporation device for cell lysis.” Lab Chip, 2005. 5(1): p. 23-9. [73] Ramadan, Q., et al., “Simultaneous cell lysis and bead trapping in a continuous flow microfluidic device.” Sensors and Actuators B-Chemical, 2006. 113(2): p. 944-955. T.G., et al., Detection and measurement of occult disease for the prognosis of solid Lugo, [74] tumors. J Clin Oncol, 2003. 21(13): p. 2609-15. [75] Cote, R.J., “Occult metastases: real harm or false alarm?” J Thorac Cardiovasc Surg, 2003. 126(2): p. 332-3.
364
Microfabricated Devices for Detecting Circulating Tumor Cells in Cancer Patient Blood Samples [76] Gnant, M., et al., “Adjuvant endocrine therapy plus zoledronic acid in premenopausal women with early-stage breast cancer: 5-year follow-up of the ABCSG-12 bone-mineral density substudy.” Lancet Oncol, 2008. 9(9): p. 840-9. [77] Bidard, F.C., et al., “A ‘class action’ against the microenvironment: Do cancer cells cooperate in metastasis?” Cancer Metastasis Rev, 2008. 27(1): p. 5-10. [78] Cohen, S.J., et al., “Relationship of circulating tumor cells to tumor response, progression-free survival, and overall survival in patients with metastatic colorectal cancer.” J Clin Oncol, 2008. 26(19): p. 3213-21. [79] de Bono, J.S., et al., “Circulating tumor cells predict survival benefit from treatment in metastatic castration-resistant prostate cancer.” Clin Cancer Res, 2008. 14(19): p. 6302-9. [80] Dick, J.E., “Stem cells: Self-renewal writ in blood.” Nature, 2003. 423(6937): p. 231-3. [81] Al-Hajj, M., et al., “Prospective identification of tumorigenic breast cancer cells.” Proc Natl Acad Sci U S A, 2003. 100(7): p. 3983-8. Al-Hajj, M. and M.F. Clarke, “Self-renewal and solid tumor stem cells.” Oncogene, 2004. [82] 23(43): p. 7274-82. [83] O’Brien, C.A., et al., “A human colon cancer cell capable of initiating tumour growth in immunodeficient mice.” Nature, 2007. 445(7123): p. 106-10. [84] Al-Hajj, M., et al., “Therapeutic implications of cancer stem cells.” Curr Opin Genet Dev, 2004. 14(1): p. 43-7. [85] Ginestier, C., et al., “ALDH1 is a marker of normal and malignant human mammary stem cells and a predictor of poor clinical outcome.” Cell Stem Cell, 2007. 1(5): p. 555-67. [86] Ponti, D., et al., “Isolation and in vitro propagation of tumorigenic breast cancer cells with stem/progenitor cell properties.” Cancer Res, 2005. 65(13): p. 5506-11. [87] Schabel, F.M., Jr., “Rationale for adjuvant chemotherapy.” Cancer, 1977. 39(6 Suppl): p. 2875-82. [88] Balic, M., et al., “Most early disseminated cancer cells detected in bone marrow of breast cancer patients have a putative breast cancer stem cell phenotype.” Clin Cancer Res, 2006. 12(19): p. 5615-21.
PART III
Tiny Machines
CHAPTER 14
Medical Nanorobotics: The Long-Term Goal for Nanomedicine Robert A. Freitas Jr.
14.1
Introduction Nanotechnology involves the engineering of molecularly precise structures and, ultimately, molecular machines. BCC Research [1] estimated the global market for nanotools and nanodevices was $1.5 billion in 2006 and projected to reach $8.6 billion by 2011, rapidly gaining on the slower-growing nanomaterials market, which is estimated at $9.0 billion (2006) and $16.6 billion (2011). As distinct from nanoscale materials and today’s simple nanotools and nanodevices having nanoscale features, molecular nanotechnology encompasses the concept of engineering functional machine systems at the molecular scale—including mechanical systems designed and built to atomic precision. Molecular manufacturing (Section 14.4) would make use of positionally controlled mechanosynthesis (mechanically-mediated chemistry) guided by molecular machine systems to build complex products, including additional nanomachines. Nanomedicine [2, 3] is the application of nanotechnology to medicine: the preservation and improvement of human health, using molecular tools and molecular knowledge of the human body. Nanomedicine encompasses at least three types of molecularly precise structures [4]—nonbiological nanomaterials, biotechnology materials and engineered organisms, and nonbiological devices including diamondoid nanorobotics. In the near term, the molecular tools of nanomedicine will employ biologically active nanomaterials and nanoparticles having well-defined nanoscale features. In the midterm (5–10 years), knowledge gained from genomics and proteomics will make possible new treatments tailored to specific individuals, new drugs targeting pathogens whose genomes have been decoded, and stem cell treatments. Genetic therapies, tissue engineering, and many other offshoots of biotechnology will become more common in therapeutic medical practice. We also may see biological robots derived from bacteria or other motile cells that have had their genomes reengineered and reprogrammed, along with artificial organic devices that incorporate biological motors or self-assembled DNA-based structures for a variety of useful medical purposes. In the farther term (2020s and beyond), the first fruits of medical nanorobotics—the most powerful of the three classes of nanomedicine technology, though clinically the most distant and still mostly theoretical today–should begin to appear in the medical field. Nanotechnologists will learn how to build nanoscale
367
368
Medical Nanorobotics: The Long-Term Goal for Nanomedicine
molecular parts like gears, bearings, and ratchets. Each nanopart may comprise a few thousand precisely placed atoms. These mechanical nanoparts will then be assembled into larger working machines such as nanosensors, nanomanipulators, nanopumps, nanocomputers, and even complete nanorobots which may be micron-scale or larger. The presence of onboard computers is essential because in vivo medical nanorobots will be called upon to perform numerous complex behaviors that must be conditionally executed on at least a semiautonomous basis, guided by receipt of local sensor data and constrained by preprogrammed settings, activity scripts and event clocking, and further limited by a variety of simultaneously executing real-time control protocols and by external instructions sent into the body by the physician during the course of treatment. With medical nanorobots in hand, doctors should be able to quickly cure most diseases that hobble and kill people today, rapidly repair most physical injuries our bodies can suffer, and significantly extend the human healthspan [5]. The early genesis of the concept of medical nanorobotics sprang from the visionary idea that tiny nanomachines could be designed, manufactured, and introduced into the human body to perform cellular repairs at the molecular level. Although the medical application of nanotechnology was later championed in the popular writings of Drexler [6] in the 1980s and 1990s and in the technical writings of Freitas [2, 3] in the 1990s and 2000s, the first scientist to voice this possibility was the late Nobel physicist Richard P. Feynman, who worked on the Manhattan Project at Los Alamos during World War II and later taught at CalTech for most of his professorial career. In his prescient 1959 talk “There’s Plenty of Room at the Bottom,” Feynman proposed employing machine tools to make smaller machine tools, these to be used in turn to make still smaller machine tools, and so on all the way down to the atomic level [7]. He prophetically concluded that this is “a development which I think cannot be avoided.” After discussing his ideas with a colleague, Feynman offered the first known proposal for a medical nanorobotic procedure of any kind–in this instance, to cure heart disease: “A friend of mine (Albert R. Hibbs) suggests a very interesting possibility for relatively small machines. He says that, although it is a very wild idea, it would be interesting in surgery if you could swallow the surgeon. You put the mechanical surgeon inside the blood vessel and it goes into the heart and looks around. (Of course the information has to be fed out.) It finds out which valve is the faulty one and takes a little knife and slices it out. Other small machines might be permanently incorporated in the body to assist some inadequately functioning organ.” Later in his historic 1959 lecture, Feynman urges us to consider the possibility, in connection with microscopic biological cells, “that we can manufacture an object that maneuvers at that level!”
14.2
From Nanoparticles to Nanorobots The greatest power of nanomedicine will emerge when we can design and construct complete artificial medical nanorobots using rigid diamondoid nanometer-scale parts such as molecular gears and bearings. Diamondoid nanorobots may be constructed using future molecular manufacturing technologies such as diamond
14.2 From Nanoparticles to Nanorobots
369
mechanosynthesis that are currently being investigated theoretically using quantum ab initio and density-functional computational methods. Complete artificial nanorobots may possess subsystems including onboard sensors, pumps, motors, manipulators, clocks, power supplies, communication systems, navigation systems, and molecular computers. Conceptual designs for diamondoid nanorobots that can mimic important natural biological cells (e.g., erythrocytes and leukocytes) have been published, and other nanorobots could perform medical tasks not found in nature such as drug delivery or chromosome replacement in individual living cells in vivo. To bridge the gap in our knowledge between present-day nanoparticle-based technologies and future nanorobotic technologies, a great deal of research remains to be done. In the relatively near term, over the next 5 years, pre-nanorobotic nanomedicine can address many important medical problems by using for drug delivery nanoscale-structured materials and basic nanodevices that can already be manufactured today—most notably organic polymer or lipid-based systems such as polymeric micelles, liposomes and solid lipid nanoparticles, and various nanocrystal-based systems, many of which have already advanced to marketed products. Surveys of these technologies are available elsewhere [4, 8], so here we report just a few selected examples of nanoparticle-related work that may exemplify early steps toward the more sophisticated capabilities that nanorobots will ultimately possess. Kopelman’s group at the University of Michigan has developed dye-tagged nanoparticles to be inserted into living cells as biosensors. This quickly led to more complex nanoparticle platforms incorporating a variety of plug-in modules, creating molecular nanodevices for the early detection and therapy of brain cancer [9]. In this instance, one type of nanoparticle is attached to a cancer cell antibody that adheres to cancer cells, but is also affixed with a contrast agent to make the particle highly visible during MRI while also enhancing the selective cancer-killing effect during subsequent laser irradiation of the treated brain tissue. Baker’s group at the University of Michigan works with dendrimers, treeshaped synthetic molecules with a regular branching structure emanating outward from a core. The outermost layer can be functionalized with other useful molecules such as genetic therapy agents, decoys for viruses, or anti-HIV agents. The next step is to create dendrimer cluster agents, multicomponent nanodevices called tecto-dendrimers built up from a number of single-dendrimer modules [10, 11]. These modules may perform specialized functions such as diseased cell recognition, diagnosis of disease state, therapeutic drug delivery, location reporting, and therapy outcome reporting. The framework can be customized to fight a particular cancer simply by substituting any one of many possible distinct cancer recognition or “targeting” dendrimers. The larger trend in medical nanomaterials is to migrate from single-function molecules to multilayer or multimodule entities that can do many things but only at certain times, or under certain conditions, or in a particular sequence. This exemplifies a continuing and inevitable technological evolution toward a device-oriented nanomedicine, working from the bottom up. On the top-down pathway, there are ongoing attempts to build microrobots for in vivo medical use. In 2002, Ishiyama et al. at Tohoku University developed tiny magnetically driven spinning screws intended to swim along veins and carry drugs
370
Medical Nanorobotics: The Long-Term Goal for Nanomedicine
to infected tissues or even to burrow into tumors and kill them with heat [12]. In 2003, the “MR-Sub” project of Martel’s group at the NanoRobotics Laboratory of Ecole Polytechnique in Montreal tested using variable MRI magnetic fields to generate forces on an untethered microrobot containing ferromagnetic particles, developing sufficient propulsive power to direct the small device through the human body [13]. Brad Nelson’s team at the Swiss Federal Institute of Technology in Zurich has continued this approach. In 2005 they reported [14] the fabrication of a microscopic robot small enough (~200 μm) to be injected into the body through a syringe. They hope this device or its descendants might someday be used to deliver drugs or perform minimally invasive eye surgery. Nelson’s simple microrobot has successfully maneuvered through a watery maze using external energy from magnetic fields, with different frequencies able to vibrate different mechanical parts on the device to maintain selective control of various functions. Gordon’s group at the University of Manitoba has also proposed magnetically controlled “cytobots” and “karyobots” for performing wireless intracellular and intranuclear surgery [15]. These approaches illustrate the first steps toward developing the ability to externally control microscopic objects after they have been placed inside the human body, an important capability for future medical nanorobots. Other methods for controlling the activity of the tiniest robotic devices—or even individual macromolecules—are being investigated in the laboratory. Most interestingly, Jacobson and colleagues [16] have attached tiny radio-frequency antennas—1.4-nm gold nanocrystals of less than 100 atoms—to DNA. When a ~1-GHz radio-frequency magnetic field is transmitted into the tiny antennas, alternating eddy currents induced in the nanocrystals produce highly localized inductive heating, causing the double-stranded DNA to separate into two strands in a matter of seconds in a fully reversible dehybridization process that leaves neighboring molecules untouched. The long-term goal is to apply the antennas to living systems and control DNA (e.g., gene expression, giving the ability to turn genes on or off) via remote electronic switching. Such a tool could give pharmaceutical researchers a way to simulate the effects of potential drugs, which also turn genes on and off. The gold nanocrystals can be attached to proteins as well as DNA, opening up the possibility of future radio frequency biology electronically controlling more complex biological processes such as enzymatic activity, protein folding and biomolecular assembly [17]. Motors and bearings for nanoscale machines have received a great deal of experimental attention, including the 78-atom chemically-powered rotating nanomotor synthesized in 1999 by Kelly [18], a chemically-powered rotaxane-based linear motor exerting ~100 pN of force with a 1.9 nm throw and a ~250-sec contraction cycle by Stoddart’s group [19], a UV-driven catenane-based ring motor by Wong and Leigh [20], and an artificial 58-atom motor molecule that spins when illuminated by solar energy by Feringa [21]. Zettl’s group at U.C. Berkeley has experimentally demonstrated an essentially frictionless bearing made from two corotating nested nanotubes [22], which can also serve as a mechanical spring because the inner nanotube “piston” feels a restoring force as it is extracted from the outer nanotube “jacket.” Zettl’s group then fabricated a nanomotor mounted on two of these nanotube bearings, demonstrating the first electrically powered nanoscale motor [23]. Deshpande and coworkers [24] have demonstrated a simple electrostatic
14.3 Diamondoid Materials in Nanorobotics
371
motor in which an inside nanotube “piston” is forced to slide out of its outer nanotube jacket by increasing the applied electrical voltage from 4.5 to 10 volts. In 2005, Tour’s group at Rice University reported [25] constructing a tiny molecular “nanocar” measuring 3 to 4 nm across that consists of a chassis, two freely rotating axles made of well-defined rodlike acetylenic structures with a pivoting suspension, and wheels made of C60 buckyball molecules that can turn independently because the bond between them and the axle is freely rotatable. Placed on a gold surface at 170oC, the nanocar spontaneously rolls on all four wheels, but only along its long axis in a direction perpendicular to its axles (a symmetrical three-wheeled variant just spins in place). When pulled with an STM tip, the nanocar cannot be towed sideways—the wheels dig in, rather than rolling. A larger, more functionalized version of the nanocar might carry other molecules along and dump them at will. Indeed, the Rice team has apparently “already followed up the nanocar work by designing a light-driven nanocar and a nanotruck that’s capable of carrying a payload” [26].
14.3
Diamondoid Materials in Nanorobotics Many theorists believe that the most reliable, durable, and efficacious medical nanorobots will be built using “diamondoid” materials [2–6, 27] that combine the key properties of high bond strength, high bond density, simplicity (hence predictability) of 3-D bonding chemistry, and maximum mechanical stiffness. What is diamondoid? First and foremost, diamondoid materials include pure diamond, the crystalline allotrope of carbon. Among other exceptional properties, diamond has extreme hardness, high thermal conductivity, low frictional coefficient, chemical inertness, a wide electronic bandgap, and (along with carbon nanotubes and fullerenes) is the strongest and stiffest material presently known at ordinary pressures. Diamondoid materials also may include any stiff covalent solid that is similar to diamond in strength, chemical inertness, or other important material properties, and possesses a dense three-dimensional network of bonds. Examples of such materials are single-crystal silicon and strong covalent ceramics such as silicon carbide, silicon nitride, and boron nitride, plus a few very stiff ionic ceramics such as sapphire (monocrystalline aluminum oxide). Many of these can be covalently or nanomechanically bonded to pure covalent structures such as diamond—for example, as in silicon-on-sapphire [28] and diamond-on-sapphire [29, 30] devices, diamond-Al composites [31], and both van der Waals [32] and mechanical [33] diamond-sapphire bonding. Of course, large pure crystals of diamond are brittle and easily fractured. The intricate molecular structure of a nanofactory-built diamondoid medical nanomachine will more closely resemble a complex composite material, not a brittle solid crystal. These products, and the nanofactories that build them, should be extremely durable in normal use. Complex diamondoid medical nanorobots probably cannot be manufactured using the conventional techniques of self-assembly. As noted in the final report [34] of the 2006 congressionally mandated review of the U.S. National Nanotechnology Initiative by the National Research Council (NRC) of the National Academies and the National Materials Advisory Board (NMAB): “For the manufacture of more
372
Medical Nanorobotics: The Long-Term Goal for Nanomedicine
sophisticated materials and devices, including complex objects produced in large quantities, it is unlikely that simple self-assembly processes will yield the desired results. The reason is that the probability of an error occurring at some point in the process will increase with the complexity of the system and the number of parts that must interoperate.” The opposite of self-assembly processes is positionally controlled processes, in which the positions and trajectories of all components of intermediate and final product objects are controlled at every moment during fabrication and assembly. Positional processes should allow more complex products to be built with high quality and should enable rapid prototyping during product development. Positional assembly is the norm in conventional macroscale manufacturing (e.g., cars, appliances, houses) but is only recently [35, 36] starting to be seriously investigated experimentally for nanoscale manufacturing. Of course, we already know that positional fabrication will work in the nanoscale realm. This is demonstrated in the biological world by ribosomes, which positionally assemble proteins in living cells by following a sequence of digitally encoded instructions (even though ribosomes themselves are self-assembled). Lacking this positional fabrication of proteins controlled by DNA-based software, large, complex, digitally specified organisms would probably not be possible and biology as we know it could not exist. The most important materials for positional assembly may be the rigid covalent or diamondoid solids, since these could potentially be used to build the most reliable and complex nanoscale machinery [45]. Preliminary theoretical studies have suggested great promise for these materials in molecular manufacturing [37–55]. The NMAB/NRC Review Committee recommended [34] that experimental work aimed at establishing the technical feasibility of positional molecular manufacturing should be pursued and supported: “Experimentation leading to demonstrations supplying ground truth for abstract models is appropriate to better characterize the potential for use of bottom-up or molecular manufacturing systems that utilize processes more complex than self-assembly.” Making complex nanorobotic systems requires manufacturing techniques that can build a molecular structure by positional assembly [37]. This will involve picking and placing molecular parts one by one, moving them along controlled trajectories much like the robot arms that manufacture cars on automobile assembly lines. The procedure is then repeated over and over with all the different parts until the final product, such as a medical nanorobot, is fully assembled using, say, a desktop nanofactory. The development pathway for diamondoid medical nanorobots will be long and arduous. First, theoretical scaling studies [38–44] and basic experimental efforts are used to assess basic concept feasibility. These initial studies must then be followed by more detailed computational simulations of specific nanorobot components and assemblies, and ultimately full systems simulations, all thoroughly integrated with additional simulations of massively parallel manufacturing processes from start to finish consistent with a design-for-assembly engineering philosophy. Once molecular manufacturing capabilities become available, experimental efforts may progress from fabrication and testing of components (built from small-molecule or atomic precursors) to the assembly of components into nanomechanical devices and nanomachine systems, and finally to prototypes and mass manufacture of medical nanorobots, ultimately leading to clinical trials. As noted earlier there has been some
14.4 Early Steps Toward Diamondoid Molecular Manufacturing
373
limited experimental work with microscale-component microscopic microrobots [12–15] but progress on nanoscale-component microscopic nanorobots today is largely at the concept feasibility and preliminary design stages and will remain so until experimentalists develop the capabilities required for molecular manufacturing, as reviewed below.
14.4
Early Steps Toward Diamondoid Molecular Manufacturing It is worth taking a brief look at the earliest steps along the development pathway leading to diamondoid nanorobotics. The key threshold technology that must be mastered and demonstrated is called “mechanosynthesis.” Mechanosynthesis, which involves molecular positional fabrication, is the formation of covalent chemical bonds using precisely applied mechanical forces to build, for example, diamondoid structures. Mechanosynthesis employs chemical reactions driven by the mechanically precise placement of extremely reactive chemical species in an ultra-high vacuum (UHV) environment. Mechanosynthesis can subsequently be automated via computer control, enabling programmable molecular positional fabrication. Molecularly precise fabrication involves holding feedstock atoms or molecules, and a growing nanoscale workpiece, in the proper relative positions and orientations so that when they touch they will chemically bond in the desired manner (because the reaction is arranged to be thermodynamically preferred). In this process, a mechanosynthetic tool is brought up to the surface of a workpiece. One or more transfer atoms are added to, or removed from, the workpiece by the tool. Then the tool is withdrawn and recharged. This process is repeated until the workpiece (e.g., a growing nanopart) is completely fabricated to molecular precision with each atom in exactly the right place. Note that the transfer atoms are under positional control at all times, in UHV, to prevent unwanted side reactions from occurring. Side reactions are also prevented using proper reaction design so that the reaction energetics help us avoid undesired pathological intermediate structures. The positional assembly of diamondoid structures, some almost atom by atom, using molecular feedstock has been examined theoretically [45–55] via computational models of diamond mechanosynthesis (DMS). DMS is the controlled addition of carbon dimers (C2), single methyl groups (CH3), or other small molecular groups to the growth surface of a diamond crystal lattice workpiece in a vacuum manufacturing environment. Covalent chemical bonds are formed one by one as the result of positionally constrained mechanical forces applied at the tip of a scanning probe microscope (SPM) apparatus. For example, programmed sequences of carbon dimer placement on growing diamond surfaces in vacuo appear feasible in theory [51, 55], as illustrated by the hypothetical DCB6Ge tooltip which is shown depositing the first two carbon atoms on a clean diamond C(110) surface in Figure 14.1. The first experimental proof that individual atoms could be manipulated was obtained by IBM scientists in 1989 when they used a scanning tunneling microscope to precisely position 35 xenon atoms on a nickel surface to spell out the corporate logo “IBM” (Figure 14.2). However, this feat did not involve the formation of cova-
374
Medical Nanorobotics: The Long-Term Goal for Nanomedicine
Figure 14.1 DCB6Ge tooltip shown depositing two carbon atoms on a diamond surface. (© 2004 Robert A. Freitas Jr. All Rights Reserved [29].)
Figure 14.2 IBM logo spelled out using 35 xenon atoms arranged on a nickel surface by an STM. (Courtesy of IBM Research Division.)
lent chemical bonds. One important step toward the practical realization of DMS was achieved in 1999 by Ho and Lee [56], who achieved the first site-repeatable site-specific covalent bonding operation of a two diatomic carbon-containing molecules (CO), one after the other, to the same atom of iron on a crystal surface, using an SPM. SPM-mediated single-molecule chemistry is now an active research area. The first experimental demonstration of pure mechanosynthesis, establishing covalent bonds using only mechanical forces—albeit on silicon atoms, not carbon atoms—was reported in 2003 by Oyabu and colleagues [57] in the Custance group. In this landmark experiment, the researchers vertically manipulated single silicon atoms from the Si(111)–(7×7) surface, using a low-temperature near-contact atomic force microscope to demonstrate: (1) removal of a selected silicon atom from its
14.4 Early Steps Toward Diamondoid Molecular Manufacturing
375
equilibrium position without perturbing the (7×7) unit cell, and (2) the deposition of a single Si atom on a created vacancy, both via purely mechanical processes. The same group later repeated this feat with germanium atoms [58], and in 2008 progressed to more complex 2-D structures fabricated entirely via mechanosynthesis [59]; the mechanosynthesis of carbon nanostructures is now being pursued by other groups [60]. To achieve molecularly precise fabrication, the first challenge is to make sure that all chemical reactions will occur at precisely specified places on the surface. A second problem is how to make the diamond surface reactive at the particular spots where we want to add another atom or molecule. A diamond surface is normally covered with a layer of hydrogen atoms. Without this layer, the raw diamond surface would be highly reactive because it would be studded with unused (or “dangling”) bonds from the topmost plane of carbon atoms. While hydrogenation prevents unwanted reactions, it also renders the entire surface inert, making it difficult to add carbon (or anything else) to it. To overcome these problems, we are developing a set of molecular-scale tools that would, in a series of well-defined steps, prepare the surface and create hydrocarbon structures on a layer of diamond, atom by atom and molecule by molecule. A mechanosynthetic tool typically has two principal components: a chemically active tooltip and a chemically inert handle to which the tooltip is covalently bonded. The tooltip is the part of the tool where chemical reactions are forced to occur. The much larger handle structure is big enough to be grasped and positionally manipulated using an SPM or similar macroscale instrumentality. At least three types of basic mechanosynthetic tools (Figure 14.3) have already received considerable theoretical (and some experimental) study [61] and are likely among those required to build molecularly precise diamond via positional control: Hydrogen Abstraction Tools. The first step in the process of mechanosynthetic fabrication of diamond might be to remove a hydrogen atom from each of one or two specific adjacent spots on the diamond surface, leaving behind one or two reactive dangling bonds or a penetrable C=C double bond. This could be done using a
Hydrogen abstraction tool
Hydrogen donation tool
Carbon placement tool
(a)
(b)
(c)
Figure 14.3 Examples of three basic mechanosynthetic tooltypes that are required to build molecularly precise diamond via positional control (black = C atoms, gray = Ge atoms, white = H atoms) [48]. (© 2007 Robert A. Freitas Jr. All Rights Reserved.)
376
Medical Nanorobotics: The Long-Term Goal for Nanomedicine
hydrogen abstraction tool [52] that has a high chemical affinity for hydrogen at one end but is elsewhere inert (Figure 14.3(a)). The tool’s unreactive region serves as a handle or handle attachment point. The tool would be held by a molecular positional device, initially perhaps a scanning probe microscope tip but ultimately a molecular robotic arm, and moved directly over particular hydrogen atoms on the surface. One suitable molecule for a hydrogen abstraction tool is the acetylene or “ethynyl” radical, comprised of two carbon atoms triply bonded together. One carbon of the two serves as the handle connection, and would bond to a nanoscale positioning tool through a much larger handle structure perhaps consisting of a lattice of adamantane cages as shown in Figure 14.4. The other carbon of the two has a dangling bond where a hydrogen atom would normally be present in a molecule of ordinary acetylene (C2H2). The working environment around the tool would be inert (e.g., vacuum or a noble gas such as neon). Hydrogen Donation Tools. After a molecularly precise structure has been fabricated by a succession of hydrogen abstractions and carbon depositions, the fabricated structure must be hydrogen-terminated to prevent additional unplanned reactions or structural rearrangements. While the hydrogen abstraction tool is intended to make an inert structure reactive by creating a dangling bond, the hydrogen donation tool [54] does the opposite. It makes a reactive structure inert by terminating a dangling bond. Such a tool would be used to stabilize reactive surfaces and help prevent the surface atoms from rearranging in unexpected and undesired ways. The key requirement for a hydrogen donation tool is that it includes a weakly attached hydrogen atom. Many molecules fit that description, but the bond between hydrogen and germanium is sufficiently weak so that a Ge-based hydrogen donation tool (Figure 14.3(b)) should be effective. Build next-generation recyclable mechanosynthetic tool
C2 dimer © 2004 Rober A. Freitas Jr. (www.rfreitas.com)
6,194 atoms C 1,452 atoms H 2 atomes Ge Total: 7,648 atoms
5.16 nm
2.78 nm 2.73 nm
DCB6Ge-Xtip
Figure 14.4 Recyclable DCB6Ge tooltip with crossbar handle motif [44]. (© 2004 Robert A. Freitas Jr. All Rights Reserved.)
14.4 Early Steps Toward Diamondoid Molecular Manufacturing
377
Carbon Placement Tools. After the abstraction tool has created adjacent reactive spots by selectively removing hydrogen atoms from the diamond surface but before the surface is reterminated by hydrogen, carbon placement tools may be used to deposit carbon atoms at the desired reactive surface sites. In this way a diamond structure would be built on the surface, molecule by molecule, according to plan. The first complete tool ever proposed for this carbon deposition function is the “DCB6Ge” dimer placement tool [47]; in this example, a carbon (C2) dimer having two carbon atoms connected by a triple bond, with each carbon in the dimer connected to a larger unreactive handle structure through two germanium atoms (Figure 14.3(c)). This dimer placement tool, also held by a nanoscale positioning device, is brought close to the reactive spots along a particular trajectory, causing the two dangling surface bonds to react with the ends of the carbon dimer. The dimer placement tool would then withdraw, breaking the relatively weaker bonds between it and the C2 dimer and transferring the carbon dimer from the tool to the surface, as illustrated in Figure 14.1. A positionally controlled dimer could be bonded at many different sites on a growing diamondoid workpiece, in principle allowing the construction of a wide variety of useful nanopart shapes. As of 2009, the DCB6Ge dimer placement tool remains the most intensively studied of any mechanosynthetic tooltip to date [47, 48, 50, 51, 53, 55], having had more than 150,000 CPU-hours of computation invested thus far in its analysis, and it remains the only DMS tooltip motif that has been successfully simulated and validated for its intended function on a full 200-atom diamond surface model [51]. Other proposed dimer (and related carbon transfer) tooltip motifs [45–47, 49, 53, 55] have received less extensive study but are also expected to perform well. In 2008, Freitas and Merkle [55] published the results of a three-year project to computationally analyze a comprehensive set of DMS reactions and an associated minimal set of tooltips that could be used to build basic diamond, graphene (e.g., carbon nanotubes and fullerenes), and all of the tools themselves including all necessary tool recharging reactions. The research defined 65 DMS reaction sequences incorporating 328 reaction steps, with 354 pathological side reactions analyzed and with 1,321 unique individual density functional theory (DFT)-based quantum chemistry reaction energies reported. (These mechanosynthetic reaction sequences range in length from 1–13 reaction steps (typically 4) with 0–10 possible pathological side reactions or rearrangements (typically 3) reported per reaction step.) For the first time, this toolset provides clear developmental targets for a comprehensive near-term DMS implementation program. Although the first practical proposal for building a DMS tool experimentally was published by Freitas in 2005 and was the subject of the first mechanosynthesis patent ever filed [50], the 2008 Freitas–Merkle study [55] provides even simpler practical proposals for building several DMS tools experimentally, also using only experimental methods that are already available today. Processes are identified for the experimental fabrication of a hydrogen abstraction tool, a hydrogen donation tool, and two alternative carbon placement tools (other than DCB6Ge), and these processes and tools are part of the second mechanosynthesis patent ever filed. Direct experimental tests of these proposals were underway by early 2009 [60].
378
14.5
Medical Nanorobotics: The Long-Term Goal for Nanomedicine
Massive Parallelism Enables Practical Molecular Manufacturing The ultimate goal of molecular nanotechnology is to develop a manufacturing technology able to inexpensively manufacture most arrangements of atoms that can be specified in molecular detail, including complex arrangements involving millions or billions of atoms per product object, as in the hypothesized medical nanorobots (Section 14.6). This will provide the ultimate manufacturing technology in terms of atomic precision, system flexibility, and low cost. But to be practical, molecular manufacturing must also be able to assemble very large numbers of identical medical nanorobots very quickly. Two central technical objectives thus form the core of our current strategy for diamondoid molecular manufacturing: (1) programmable positional assembly including fabrication of diamondoid structures using molecular feedstock, as discussed above, and (2) massive parallelization of all fabrication and assembly processes, briefly described below. Molecular manufacturing systems capable of massively parallel fabrication [62] might employ, at the lowest level, large arrays of DMS-enabled scanning probe tips all building similar diamondoid product structures in unison. Analogous approaches are found in present-day larger-scale systems. For example, simple mechanical ciliary arrays consisting of 10,000 independent microactuators on a 1-cm2 chip have been made at the Cornell National Nanofabrication Laboratory for microscale parts transport applications, and similarly at IBM for mechanical data storage applications [63]. Active probe arrays of 10,000 independently actuated microscope tips have been developed by Mirkin’s group at Northwestern University for dip-pen nanolithography [64] using DNA-based “ink.” Almost any desired 2-D shape can be drawn using 10 tips in concert. Another microcantilever array manufactured by Protiveris Corp. has millions of interdigitated cantilevers on a single chip [65]. Martel’s group has investigated using fleets of independently mobile wireless instrumented microrobot manipulators called NanoWalkers to collectively form a nanofactory system that might be used for positional manufacturing operations [66]. Zyvex Corp. (www.zyvex.com) of Richardson TX received a $25 million, five-year, National Institute of Standards and Technology (NIST) contract to develop prototype microscale assemblers using microelectromechanical systems [62]. Eventually this research should lead to the design of production lines in a nanofactory, both for diamondoid mechanosynthesis and for component assembly operations. Making complex nanorobotic systems will require manufacturing techniques that can build a molecular structure via positional assembly. This will involve picking and placing molecular parts one by one, moving them along controlled trajectories much like the robot arms that manufacture cars on automobile assembly lines. The procedure is then repeated over and over with all the different parts until the final product, such as a medical nanorobot, is fully assembled. Ultimately, medical nanorobots will be manufactured in nanofactories efficiently designed for this purpose. The nanofactory system will likely include a progression of fabrication and assembly mechanisms at several different physical scales. At the smallest scale, molecular mills could manipulate individual molecules to fabricate successively larger submicron-scale building blocks. These could be passed to larger block assemblers that assemble still larger microblocks, which are themselves passed to
14.6 Examples of Diamondoid Medical Nanorobots
379
even larger product assemblers that put together the final product. The microblocks would be placed in a specific pattern and sequence following construction blueprints created using a modern “design for assembly” philosophy.
14.6
Examples of Diamondoid Medical Nanorobots The greatest power of nanomedicine will emerge, perhaps in the 2020s, when we can design and construct complete artificial medical nanorobots using rigid diamondoid nanometer-scale parts such as molecular gears and bearings [45]. Complete artificial nanorobots may possess a full range of autonomous subsystems including onboard sensors, pumps, motors, manipulators, clocks, power supplies, communication systems, navigation systems, and molecular computers, as has been extensively described elsewhere [2]. Several conceptual designs of medical nanorobots have been published [38–44]. The first theoretical design study [38] of a complete medical nanorobot ever published in a peer-reviewed journal (in 1998) described a hypothetical artificial mechanical red blood cell or “respirocyte” made of 18 billion precisely arranged structural atoms. The respirocyte (Figure 14.5) is a bloodborne spherical 1-μm diamondoid 1,000-atmosphere pressure vessel with reversible molecule-selective pumps powered by endogenous serum glucose. This nanorobot would deliver 236 times more oxygen to body tissues per unit volume than natural red cells and would
Figure 14.5 (Color plate 21) The respirocyte [31], an artificial mechanical red cell. (Designer Robert A. Freitas Jr. ©2000 E-spaces (3danimation.e-spaces.com) and Robert A. Freitas Jr. (www.rfreitas.com).
380
Medical Nanorobotics: The Long-Term Goal for Nanomedicine
manage carbonic acidity, controlled by gas concentration sensors and an onboard nanocomputer. A mere 5-cc therapeutic dose of 50% respirocyte saline suspension containing 5 trillion nanorobots could exactly replace the gas carrying capacity of the patient’s entire 5.4 liters of blood. Respirocytes could provide a universal blood substitute, quick treatment for asphyxia (e.g., monoxide poisoning), backup tissue oxygenation for heart and surgical patients, site-specific deoxygenation of tumors, and support for other nanorobots (e.g., augmenting local O2, the limiting resource for oxyglucose nanorobot power). The supervising physician can transmit control commands via ultrasound signals that are received by acoustic sensors on the nanorobot hull. Another conceptual design exists for the nanorobotic artificial phagocytes called “microbivores” [42] that could patrol the bloodstream, seeking out unwanted pathogens including bacteria, viruses, or fungi and then digesting them using a combination of onboard mechanical and artificial enzymatic systems (Figure 14.6). Nanorobots recognize a target bacterium by direct contact with its foreign surface coat antigen markers. Microbivores (2–3 μm oblate-shaped nanorobots with a mouth at one end) could achieve complete clearance of even the most severe septicemic infections in hours or less if a sufficient number of devices are employed [42]. This is far better than the weeks or months needed for antibiotic-assisted natural phagocytic defenses. Microbivores don’t increase the risk of sepsis or septic shock because the pathogens are completely digested into harmless sugars, amino acids and the like, which are the only effluents from the nanorobot. The biocompatibility (including immunoreactivity, thrombogenicity, phagocytosis, and granulomatous reaction) of diamondoid medical nanorobots such as respirocytes and microbivores has been extensively reviewed in a book-length treatment elsewhere [3]. Note that such devices will likely require some form of biocompatible
Figure 14.6 (Color plate 22) An artificial white cell—the microbivore [35]. (Designer Robert A. Freitas Jr., additional design by Forrest Bishop. ©2001 Zyvex Corp. Used with permission.)
14.6 Examples of Diamondoid Medical Nanorobots
381
coating [3] (depending on mission type and duration) to give them a high level of biocompatibility closer to that of lipid-based systems [67] than to that of uncoated fullerene- or carbon nanotube-based systems [68]. A conceptual design has also been published [44] for a very advanced medical nanorobot called the chromallocyte. This is a hypothetical mobile cell-repair nanorobot whose primary purpose is to perform chromosome replacement therapy (CRT). In CRT, the entire chromatin content of the nucleus in a living cell is extracted and promptly replaced with a new set of prefabricated chromosomes which have been artificially manufactured as defect-free copies of the originals. The chromallocyte (Figure 14.7) will be capable of limited vascular surface travel into the capillary bed of the targeted tissue or organ, followed by diapedesis (exiting a blood vessel into the tissues) [2], histonatation (locomotion through tissues) [2], cytopenetration (entry into the cell interior) [2], and complete chromatin replacement in the nucleus of the target cell [44]. The CRT mission ends with a return to the bloodstream and subsequent extraction of the device from the body at the original infusion site. Replacement chromosomes are manufactured in a desktop ex vivo chromosome sequencing and manufacturing facility, then loaded into the nanorobots for delivery to specific targeted cells during CRT. A single lozenge-shaped 69 micron3 chromallocyte measures 4.18 microns and 3.28 microns along cross-sectional diameters and 5.05 microns in length, typically consuming 50 to 200 pW in normal operation and a maximum of 1,000 pW in bursts during outmessaging, the most energy-intensive task. Treatment of an entire large human organ such as a liver, involving simultaneous CRT on all 250 billion hepatic tissue cells, might require the localized infusion of a ~1 terabot (1012 devices) ~69 cm3 chromallocyte dose in a 1-liter 7% saline suspension during a ~7-hour course of therapy. The chromallocyte includes an extensible primary manipulator 4 microns long and 0.55 microns in diameter called the Proboscis that is used to spool up chromatin strands via slow rotation when inserted into the cell nucleus. After spooling, a segmented funnel assembly is extended around the spooled bolus of DNA, fully enclosing and sequestering the old genetic material. The new chromatin is then discharged into the nucleus through the center of the Proboscis by pistoning from internal storage vaults, while the old chromatin that is sequestered inside the sealed watertight
(a)
(b)
Figure 14.7 (Color plate 23) Artist’s conceptions of the basic chromallocyte [37] design: (a) devices walking along luminal wall of blood vessel ; and (b) schematic of telescoping funnel assembly and proboscis operation. (Image © 2006 Stimulacra LLC (www.stimulacra.net) and Robert A. Freitas Jr. (www.rfreitas.com).)
382
Medical Nanorobotics: The Long-Term Goal for Nanomedicine
funnel assembly is forced into the storage vaults as space is vacated by the new chromatin that is simultaneously being pumped out. The chromallocyte includes a mobility system similar to the microbivore grapple system [42], along with a solvation wave drive [2] that is used to ensure smooth passage through cell plasma and nuclear membranes.
14.7
An Ideal Nanorobotic Pharmaceutical Delivery Vehicle What would an ideal drug delivery vehicle look like? To start with, it would be targetable not just to specific tissues or organs, but to individual cellular addresses within a tissue or organ. Alternatively, it would be targetable to all individual cells within a given tissue or organ that possessed a particular characteristic (e.g., all cancer cells, or all bacterial cells of a defined species). This ideal vehicle would be biocompatible and virtually 100% reliable, with all drug molecules being delivered only to the desired target cells and none being delivered elsewhere so that unwanted side effects are eliminated. The ideal vehicle would remain under the continuous control of the supervising physician, including post-administration. Even after the vehicles had been injected into the body, the doctor would still be able to activate or inactivate them remotely, or alter their mode of action or operational parameters. Once treatment was completed, all of the vehicles could be removed intact from the body, leaving no lingering evidence of their passage. This hypothetical ideal drug delivery vehicle may be called a “pharmacyte” [43]. Pharmacytes will be self-powered, computer-controlled nanorobotic systems capable of digitally precise transport, timing, and targeted delivery of pharmaceutical agents to specific cellular and intracellular destinations within the human body. Drug molecules could be purposely delivered to one cell, but not to an adjacent cell, in the same tissue. The exemplar pharmacyte would not be a relatively passive nanoparticle but rather would be an active medical nanorobot 1–2 μm in size, similar to the respirocyte but slightly larger. It would be capable of carrying up to ~1 μm3 of pharmaceutical payload stored in onboard tanks that are mechanically offloaded using molecular sorting pumps [2, 45] mounted in the hull, operated under the control of an onboard nanocomputer. Depending on mission requirements, the payload could be discharged into the proximate extracellular fluid or delivered directly into the cytosol using a transmembrane injector mechanism [2, 3]. The sorting pumps are typically envisioned as ~1,000 nm3-size devices that can transfer ~106 molecules/sec [2, 45]. Each pump employs reversible binding sites mounted on a rotating structure that cycles between the interior and exterior of the nanorobot, allowing transport of a specific molecule even against a considerable concentration gradient (Figure 14.8). Other reversible binding sites comprise sensors on the surface of the nanorobot in order to recognize the unique biochemical signature of specific vascular and cellular addresses [2], simultaneously testing encountered biological surfaces for a sufficiently reliable combination (at least 5–10 in number) of positive-pass and negative-pass molecular markers to ensure virtually 100% targeting accuracy. Onboard power may be provided by glucose and oxygen drawn from the local environment (e.g., circulating blood, interstitial fluid, or cytosol) that is metabolized using fuel
14.7 An Ideal Nanorobotic Pharmaceutical Delivery Vehicle
383
Diamondoid wall Molecule in binding site
Blood plasma
Target molecules
Sorting rotor Cam Diamondoid wall
0 Figure 14.8
5 10 15 Scale (nanometers)
20
Individual sorting rotor. (Redrawn from Drexler [38]).
cells or other methods for biochemical energy conversion [2]. If needed for a particular application, deployable mechanical cilia and other locomotive systems can be added to the pharmacyte to permit transvascular and transcellular mobility [2], thus allowing delivery of pharmaceutical molecules to specific cellular and even intracellular addresses with negligible error. Because sorting pumps can be operated reversibly, pharmacytes could also be used to selectively extract specific molecules from targeted locations as well as deposit them. Pharmacytes, once depleted of their payloads or having completed their mission, would be recovered from the patient via centrifuge nanapheresis [2] or by conventional excretory pathways. The nanorobots might then be recharged, reprogrammed and recycled for use in a subsequent patient who may need a different pharmaceutical agent targeted to different tissues or cells than in the first patient. Phagocytosis and foreign-body granulomatous reaction are major issues for all medical nanorobots intended to remain in the body for extended durations [3], though short-duration pharmacytes that can quickly be extracted from the body may face somewhat fewer difficulties. In either case, bloodborne 1 to 2 μm pharmacytes can avoid clearance by the RES (whether via geometrical trapping or phagocytic uptake [3]) and techniques have been proposed for phagocyte avoidance and escape at each step in the phagocytic process [3]. Modest concentrations of pharmacytes will not embolize small blood vessels because the minimum viable human capillary that allows passage of intact erythrocytes and white cells is 3 to 4 μm in diameter, which is larger than the largest proposed pharmacyte. Pharmacytes can also be equipped with mobility systems [2] to allow mechanically-assisted passage through partially occluded vessels or unusually narrow spaces such as the interendothelial slits of the spleen [3]. Targeting ligands or receptors in the cell membrane exterior can be recognized by chemotactic sensors [2] on the nanorobot surface, but note that the pharmacyte (as distinguished from conventional nanoparticles) need not always be endocytosed. For example, in some cases
384
Medical Nanorobotics: The Long-Term Goal for Nanomedicine
nanorobots may use transmembrane mechanical nanoinjectors [3] to avoid having to enter a target cell. Alternatively, if the mission requires cytopenetration then endocytosis of the nanorobot may be purposely stimulated using biomimetic or completely artificial (including powered mechanical) methods [2]; after payload delivery, indigestible diamondoid nanorobots will require exocytosis by similar means. Nanorobot volume is only 1 to 10 μm3 compared to 103 to 104 μm3 for most human tissue cells so pharmacytes could be targeted to intracellular organelles, though nanorobots would have insufficient room to enter one (excepting perhaps the ER and nucleus) and would have to rely on nanoinjection in those cases. There are many potential uses of pharmacytes but it will suffice to briefly mention just two general classes of applications. First, it is often desired to deliver cytocidal agents to tumor cells. Current methods involve introducing large quantities of chemotherapy agents into the body in an effort to kill a relatively few cancerous cells, with numerous unwanted side effects on healthy cells. Precise targeting using pharmacytes can ensure delivery only to the correct cellular addresses, with presentation of cytocidal chemical agents literally on a cell-by-cell basis. In one trivial scenario, the targeted killing of 1 billion (109) cancer cells with each cell capable of being killed by ~106 precisely delivered ~1,000-dalton cytocidal molecules (i.e., lethality similar to bufagin toxin) would require a total whole-body treatment dose of just ~1015 cytocidal molecules or ~0.001 mm3 (~2 μg) of delivered material. This dose could be carried and dispensed by one trillion pharmacyte nanorobots (total injected volume of therapeutic nanorobots ~2 cm3) assuming that only 0.1% of the nanorobots encounter an acceptable target and are allowed to release a 0.001 μm3 cytocidal payload into the targeted cell, while the remaining 99.9% of the nanorobots release nothing. After initiating cell death, unmetabolized free cytocidal molecules can be locally reacquired by the pharmacyte and subsequently transported out of the patient, thus minimizing any post-treatment collateral damage. Note that the strict size requirements for macromolecules to reach the leaky vasculature of a tumor and convectively enter its pores [69] may apply to passively-diffusing payload molecules that might be conveyed and released by pharmacytes, but these limits do not apply to the motorized active nanorobots themselves. Upon arriving in the vicinity of a tumor, the pharmacyte may deliver its payload either via direct nanoinjection [3] (for tumor cells adjoining the vasculature) or by progressive cytopenetration [2] through adjacent cells until the targeted tumor cell that awaits payload delivery is reached. It is well-known that apoptotic cellular “death receptors” can be expressed on both normal and cancerous cells in the human body, so one challenge for conventional drug-based therapy is to find some way to activate death receptors selectively on cancer cells only [70]. With pharmacytes, such selectivity should be simple and routine using multiple chemosensors [2], a benefit that may be characteristic of most future nanorobot-based therapeutics. For example, if caspase cascade amplification is sufficient to permit single-site activation of the cascade, then in principle an extracellular nanorobot intending cytocide of a detected cancerous cell could press onto the outer surface of the target cell an appropriate ligand display tool. This tool might contain suitably exposed trimeric CD95L (aka FasL) ligand (binds to the extracellular domains of three CD95 death receptors), TNF or lymphotoxin alpha (binds to CD120a), Apo3L ligand aka TWEAK (binds to DR3), or Apo2L ligand
14.7 An Ideal Nanorobotic Pharmaceutical Delivery Vehicle
385
aka TRAIL (binds to DR4 and DR5) [70, 71]. The binding event would then activate a single death receptor complex, potentially triggering the entire irreversible cytocidal cascade. If necessary, multiple such display tools could be employed. This technique avoids much of the storage requirement for bulky consumables aboard the medical nanorobot. As yet another approach, molecular sorting pumps on the pharmacyte surface could be used to selectively extract from the cytoplasm of a target cell specific crucial molecular species of inhibitors of apoptosis (IAPs) that normally hold the apoptotic process in check. Examples include survivin, commonly found in human cancer cells [72], the transcription factor NF-κB, and Akt, which delivers a survival signal that inhibits the apoptosis induced by growth factor withdrawal in neurons, fibroblasts, and lymphoid cells. Conversely, decoy receptors (DcRs) [73] that compete with DR4 and DR5 for binding to Apo2L could be saturated with intrinsically harmless but precisely engineered intracellular “chaff” ligands. With IAPs removed or DcRs blockaded, apoptosis may be free to proceed. Pharmacytes could also tag target cells with biochemical substances capable of triggering a reaction by the body’s natural defensive or scavenging systems, a strategy called “phagocytic flagging” [2]. For example, novel recognition molecules are expressed on the surface of apoptotic cells. In the case of T lymphocytes, one such molecule is phosphatidylserine, a lipid that is normally restricted to the inner side of the plasma membrane [2] but, after the induction of apoptosis, appears on the outside [74]. Cells bearing this molecule on their surface can then be recognized and removed by phagocytic cells. Seeding the outer wall of a target cell with phosphatidylserine or other molecules with similar action could activate phagocytic behavior by natural macrophages that had mistakenly identified the target cell as apoptotic. Loading the target cell membrane surface with B7 costimulator molecules also permits T-cell recognition, allowing an immunologic response via the immunological synapse [75]. These tagging operations should work well against cells that have an apoptotic response that can be triggered by cytotoxic T cells, such as human cancer cells and cysts. A second major application area of pharmacytes would be the control of cell signaling processes. As a trivial example, Ca++ serves as an intracellular mediator in a wide variety of cell responses including secretion, cell proliferation, neurotransmission, cellular metabolism (when complexed to calmodulin), and signal cascade events that are regulated by calcium-calmodulin-dependent protein kinases and adenylate cyclases. The concentration of free Ca++ in the extracellular fluid or in the cell’s internal calcium sequestering compartment (which is loaded with a binding protein called calsequestrin) is ~10-3 ions/nm3. However, in the cytosol, free Ca++ concentration varies from 6 × 10-8 ions/nm3 for a resting cell up to 3 x 10-6 ions/nm3 when the cell is activated by an extracellular signal; cytosolic levels >10-5 ions/nm3 may be toxic (e.g., via apoptosis). To transmit an artificial Ca++ activation signal to a typical (20 μm)3 tissue cell in ~1 msec, a single pharmacyte stationed in the cytoplasm must promptly raise the cytosolic ion count from 480,000 Ca++ ions to 24 million Ca++ ions. This is a transfer rate of ~2.4 × 1010 ions/sec that may be accomplished using ~24,000 hull-mounted molecular sorting pumps [2, 45] across a total nanorobot emission surface area of ~2.4 μm2. Onboard storage volume of ~1 μm3 can hold up to ~20 billion calcium atoms, enough to transmit up to ~1,000 arti-
386
Medical Nanorobotics: The Long-Term Goal for Nanomedicine
ficial Ca++ signals into the cell even assuming no reabsorption and recycling of the ions. Properly configured in cyto pharmacytes could also modify natural intracellular message traffic according to preprogrammed rules or by following external commands issued by the supervising physician. In the case of steroids and thyroid hormones, this may involve the direct manipulation of the signaling molecules themselves (after they have passed through the cell membrane) or their bound receptor complexes. However, most signaling molecules are absorbed at the cell surface, initiating a signal cascade which may be modulated by manipulating second-messenger molecules or other components of the in cyto signal cascade. A few basic examples [2] of signal modifying action involving cAMP would include: Amplification. A single epinephrine molecule received by a beta adrenergic receptor at a cell surface transduces the activation of dozens of G-protein alpha subunits, each of which in turn activates a single adenylate cyclase enzyme which cyclizes hundreds of ATP molecules into cAMP molecules. The intracellular population of cAMP (in muscle or liver target cells) is normally <10-6 M or ~5 million molecules for a typical (20 μm)3 tissue cell. Stimulation by epinephrine raises the cAMP population to ~25 million molecules in a few seconds. However, upon detecting this rising tide of cAMP during the first few msec, each in cyto pharmacyte could quickly amplify this existing chemical signal by releasing 20 million cAMP molecules (occupying a storage volume of ~0.01 ìm3) from onboard inventories in ~1 msec—thus accelerating cellular response time by several orders of magnitude. Suppression. Similarly, upon detection of rising cAMP levels in target cells, resident pharmacytes could use molecular pumps to rapidly remove cAMP from the cytosol as quickly as it is formed, even under maximum adrenal stimulation. The diffusion-limited intake current at the basal concentration (~6 x 10-7 molecules/nm3) for a cAMP-absorbing spherical nanodevice 1 μm in radius is ~4 million molecules/sec [2], so a single such device could probably keep up with natural cAMP production rates and thus completely extinguish the response by preserving a flat basal concentration even in the face of a maximum stimulus. (As a practical matter, it may be more efficient to control epinephrine generation at its glandular source unless it is desired to interface with just a single tissue type.) Simultaneously, the cAMPabsorbing nanorobot may hydrolyze the stored cAMP in the manner of the cAMP phosphodiesterases, then excrete these deactivated AMP messenger molecules back into the cytosol. Similar methods might be useful in ligand-gated ion channel desensitization or in disease symptom suppression, as, for example, in suppressing the prolonged elevation of cAMP in intestinal epithelial cells associated with the cholera toxin, which produces severe diarrhea by causing a large influx of water into the gut. Replacement. Combining suppression and amplification, an existing chemical signal could be eliminated and replaced by a different—even an opposite—message pathway using resident pharmacyte mediators. Alternative pathways may be natural or wholly synthetic. Novel responses to existing signals may be established within the cell to enhance functionality or to improve stability or controllability. For
14.8 Conclusion
387
instance, detection of one species of cytokine by a pharmacyte could trigger rapid specific absorption of that cytokine and a simultaneous fast release of another (different) species of cytokine in its place. Such procedures must of course take into account the many redundant signaling pathways and backup systems (e.g., developmental signals, immune system, blood clotting) that exist within the body. Medical nanorobots can allow the replacement of many redundant pathways with more refined and specific responses. Linkage. Previously unlinked signal cascades may be artificially linked using in cyto nanorobots. As a fanciful example, the receipt of epinephrine by pharmacytes located in the capillaries of the brain could trigger these devices to suppress the adrenalin response while simultaneously releasing chemical messengers producing message cascades that stimulate production of enkephalins or other opioids, thus encouraging a subjective state of psychological relaxation rather than the “fight or flight” response to certain stressful conditions.
14.8
Conclusion Nanomedicine is the application of nanotechnology to medicine: the preservation and improvement of human health, using molecular tools and molecular knowledge of the human body. The greatest power of nanomedicine will emerge when we can design and construct complete artificial medical nanorobots using rigid diamondoid nanometer-scale parts such as molecular gears and bearings. These diamondoid nanorobots may be constructed using future molecular manufacturing technologies such as diamond mechanosynthesis which are currently being investigated theoretically using quantum ab initio and density-functional computational methods to design useful molecular toolsets, and early steps toward experimental validation of the basic principles of positionally controlled diamondoid mechanosynthesis are now in progress. Complete artificial nanorobots may possess a full panoply of autonomous subsystems including onboard sensors, pumps, motors, manipulators, clocks, power supplies, communication systems, navigation systems, and molecular computers. Conceptual designs for diamondoid nanorobots that can mimic important natural biological cells (e.g., erythrocytes and leukocytes) have been published, along with designs for other nanorobots that could perform medical tasks not found in nature such as drug delivery or chromosome replacement in individual living cells in vivo. Medical nanorobotics will become the foundational medical technology of 21st century medicine, allowing future physicians to quickly cure most diseases that hobble and kill people today, rapidly repair most physical injuries our bodies can suffer, and significantly extend the human healthspan.
Problems 14.1 What are the remaining technical challenges to achieving positionally controlled atomically precise diamondoid mechanosynthesis?
388
Medical Nanorobotics: The Long-Term Goal for Nanomedicine
14.2 Unlike nanoparticles, nonbiological nanorobots are small machines with lots of mechanical moving parts. What mechanical subsystems would be required to ensure safe and correct functioning of a medical nanorobot placed inside a living human body? 14.3 Suggest and analyze possible methods by which nonbiological medical nanorobots could be safely removed from the human body.
References [1] McWilliams, A., “Nanotechnology: A Realistic Market Assessment,” BCC Research, July 2006; http://www.bccresearch.com/nan/NAN031B.asp. [2] Freitas, R. A. Jr., Nanomedicine, Volume I: Basic Capabilities, Georgetown, TX: Landes Bioscience, 1999; http://www.nanomedicine.com/NMI.htm. [3] Freitas, R. A. Jr., Nanomedicine, Volume IIA: Biocompatibility, Georgetown, TX: Landes Bioscience, 2003; http://www.nanomedicine.com/NMIIA.htm. [4] Freitas, R. A. Jr., “What is Nanomedicine?” Nanomedicine: Nanotech. Biol. Med., Vol. 1, 2005, pp. 2–9; http://www.nanomedicine.com/Papers/WhatIsNMMar05.pdf. [5] Freitas, R. A. Jr., “Comprehensive Nanorobotic Control of Human Morbidity and Aging.” In The Future of Aging: Pathways to Human Life Extension, G. M. Fahy, et al. (eds.), New York: Springer, 2009. [6] Drexler, K. E., Engines of Creation: The Coming Era of Nanotechnology, New York, NY: Anchor Press/Doubleday, 1986. Drexler, K. E., C. Peterson, and G. Pergamit, Unbounding the Future: The Nanotechnology Revolution, New York: William Morrow/Quill Books, 1991. [7] Feynman, R. P., “There’s plenty of room at the bottom,” Eng Sci (CalTech), Vol. 23, 1960, pp. 22–36. [8] Jain, K. K., “Nanomedicine: application of nanobiotechnology in medical practice,” Med Princ Pract., Vol. 17, 2008, pp. 89–101. [9] Tang, W., et al., “Photodynamic characterization and in vitro application of methylene blue-containing nanoparticle platforms,” Photochem Photobiol., Vol. 81, No. 2, 2005, pp. 242–249. [10] Shi, X., I. J. Majoros, and J. R. Baker Jr., “Capillary electrophoresis of poly(amidoamine) dendrimers: from simple derivatives to complex multifunctional medical nanodevices,” Mol Pharm., Vol. 2, No. 4, 2005, pp. 278–294. [11] “Tecto-Dendrimers,” Michigan Nanotechnology Institute for Medicine and Biological Sciences, University of Michigan, 1 November 2005; http://nano.med.umich.edu/projects/tecto-dendrimers.htm. [12] Ishiyama, K., M. Sendoh, and K. I. Arai, “Magnetic micromachines for medical applications,” J Magnetism Magnetic Mater., Vol. 242–245, 2002, pp. 1163–1165. [13] Mathieu, J. B., et al., “MRI systems as a mean of propulsion for a microdevice in blood vessels,” Biomed Mater Eng., Vol. 15, 2005, p. 367. [14] Yesin, K. B., et al., “Biomedical micro-robotic system,” Proc. 8th Intl. Conf. on Medical Image Computing and Computer Assisted Intervention (MICCAI 2005 / www.miccai2005.org), Palm Springs, CA, 26–29 October 2005, p. 819. [15] Chrusch, D. D., B. W. Podaima, and R. Gordon, “Cytobots: intracellular robotic micromanipulators,” Conference Proceedings, 2002 IEEE Canadian Conference on Electrical and Computer Engineering, W. Kinsner and A. Sebak, eds., Winnipeg, Canada, 12–15 May 2002; http://www.umanitoba.ca/faculties/medicine/radiology/ Dick_Gordon_papers/Chrusch,Podaima,Gordon2002.
References
389
[16] Hamad-Schifferli, K., et al. “Remote electronic control of DNA hybridization through inductive coupling to an attached metal nanocrystal antenna,” Nature, Vol. 415, 2002, pp. 152–156; http://www.media.mit.edu/molecular/HamadNature.pdf. See also: Ball, P., “Remote-control for bacteria. Radio waves switch proteins on and off,” Nature ScienceUpdate, 6 December 2002; http://www.nature.com/nsu/021202/021202–12.html [17] Service, R. F., “Nanotechnology. Biology offers nanotechs a helping hand,” Science, Vol. 298, 2002, 2322–2323; http://www.sciencemag.org/cgi/content/full/298/5602/2322?etoc [18] Kelly, T. R., H. De Silva, and R. A. Silva, “Unidirectional rotary motion in a molecular system,” Nature, Vol. 401, 1999, pp. 150–152. [19] Huang, T. J., et al., “Molecular shuttle switching in closely packed Langmuir films,” 11th Foresight Conf. Mol. Nanotech., San Francisco, CA, 10–12 October 2003. [20] Leigh, D. A., et al., “Unidirectional rotation in a mechanically interlocked molecular rotor,” Nature, Vol. 424, 2003, pp. 174–179. [21] Koumura, N., et al., “Light-driven monodirectional molecular rotor,” Nature, Vol. 401, 1999, pp. 152–155. [22] Cumings, J., and A. Zettl, “Low-friction nanoscale linear bearing realized from multiwall carbon nanotubes,” Science, Vol. 289, 2000, pp. 602–604. [23] Fennimore, A. M., et al., “Rotational actuators based on carbon nanotubes,” Nature, Vol. 424, 2003, pp. 408–410. [24] Deshpande, V. V., et al., “Carbon Nanotube Linear Bearing Nanoswitches,” Nano Letters, Vol. 6, 2006, pp. 1092–1095; http://nanoscale.caltech.edu/publications/pdf/ deshpandeetal.pdf. [25] Shirai, Y., et al., “Directional control in thermally driven single-molecule nanocars,” Nano Letters, Vol. 5, 2005, pp. 2330–2334. [26] “Rice scientists build world’s first single-molecule car,” 20 October 2005; http://www.sciencedaily.com/releases/2005/10/051021021040.htm [27] Merkle, R. C., “Nanotechnology and Medicine,” Advances in Anti-Aging Medicine, Vol. I, R. M. Klatz (ed.), Liebert Press, 1996, pp. 277–286. [28] Ponce, F. A., and J. Aranovich, “Imaging of the silicon on sapphire interface by high-resolution transmission electron microscopy,” Appl Phys Lett., Vol. 38, 1981, pp. 439–441. [29] Yoshimoto, M., et al., “Growth of Diamond on Sapphire by Pulsed Laser Ablation under Oxygen Atmosphere,” Mat Res Soc Proc., Vol. 555, 1999, pp. 383–386. [30] Ravariu, C. et al., “Simulations results of some Diamond On Insulator nano-MISFETs,” Diam Rel Mater., Vol. 15, 2006, pp. 777–782. [31] Rucha, P.W., et al., “Selective interfacial bonding in Al(Si)-diamond composites and its effect on thermal conductivity,” Composites Sci Technol., Vol. 66, 2006, pp. 2677–2685. [32] Lee, H., H. E. Meissner, and O. R. Meissner, “Adhesive-free bond (AFB) CVD diamond/sapphire and CVD diamond/YAG crystal composites,” Proc SPIE, Vol. 6216, 2006, pp. 62160O. [33] Ternyak, O., R. Akhvlediani, and A. Hoffman, “Study on diamond films with ultra high nucleation density deposited onto alumina, sapphire and quartz,” Diam Rel Mater., Vol. 14, 2005, pp. 323–327. [34] Committee to Review the National Nanotechnology Initiative, National Materials Advisory Board (NMAB), National Research Council (NRC), A Matter of Size: Triennial Review of the National Nanotechnology Initiative, The National Academies Press, Washington, DC, 2006; http://www.nap.edu/catalog/11752.html#toc. [35] Kenny, T., “Tip-Based Nanofabrication (TBN),” Defense Advanced Research Projects Agency (DARPA)/Microsystems Technology Office (MTO), Broad Agency Announcement BAA 07-59, 2007; http://www.fbo.gov/spg/ODA/DARPA/CMO/BAA07-59/listing.html. [36] Nanofactory Collaboration website, 2008; http://www.MolecularAssembler.com/ Nanofactory.
390
Medical Nanorobotics: The Long-Term Goal for Nanomedicine [37] Freitas, R. A. Jr., “Current status of nanomedicine and medical nanorobotics,” J Comput Theor Nanosci., Vol. 2, 2005, pp. 1–25; http://www.nanomedicine.com/ Papers/NMRevMar05.pdf. [38] Freitas, R. A. Jr., “Exploratory design in medical nanotechnology: A mechanical artificial red cell,” Artif Cells Blood Subst Immobil Biotech., Vol. 26, 1998, pp. 411–430; http://www.foresight.org/Nanomedicine/Respirocytes.html. [39] Freitas, R. A. Jr., “Clottocytes: artificial mechanical platelets,” IMM Report No. 18, Foresight Update No. 41, 2000, pp. 9–11; http://www.imm.org/Reports/Rep018.html [40] Freitas, R. A. Jr., “Nanodentistry,” J Amer Dent Assoc., Vol. 131, 2000, pp. 1559–1566; http://www.rfreitas.com/Nano/Nanodentistry.htm. [41] Freitas, R. A. Jr., and C. J. Phoenix, “Vasculoid: A personal nanomedical appliance to replace human blood,” J Evol Technol., Vol. 11, 2002, pp. 1–139; http://www.jetpress.org/ volume11/vasculoid.pdf. [42] Freitas, R. A. Jr., “Microbivores: Artificial mechanical phagocytes using digest and discharge protocol,” J Evol Technol., Vol. 14, 2005, pp. 1–52; http://www.jetpress.org/ volume14/freitas.html. [43] Freitas, R. A. Jr., “Pharmacytes: an ideal vehicle for targeted drug delivery,” J Nanosci Nanotechnol., Vol. 6, 2006, pp. 2769–2775; http://www.nanomedicine.com/ Papers/JNNPharm06.pdf. [44] Freitas, R. A. Jr., “The ideal gene delivery vector: Chromallocytes, cell repair nanorobots for chromosome replacement therapy,” J Evol Technol., Vol. 16, 2007, pp. 1–97; http://jetpress.org/v16/freitas.pdf. [45] Drexler, K. E., Nanosystems: Molecular Machinery, Manufacturing, and Computation, New York: John Wiley & Sons, 1992. [46] Merkle, R. C., “A proposed ‘metabolism’ for a hydrocarbon assembler,” Nanotechnology, Vol. 8, 1997, pp. 149–162; http://www.zyvex.com/nanotech/hydroCarbonMetabolism. html. [47] Merkle, R. C., and R. A. Freitas Jr., “Theoretical analysis of a carbon-carbon dimer placement tool for diamond mechanosynthesis,” J Nanosci Nanotechnol., Vol. 3, 2003, pp. 319–324; http://www.rfreitas.com/Nano/JNNDimerTool.pdf. [48] Mann, D. J., et al., “Theoretical analysis of diamond mechanosynthesis. Part II. C2 mediated growth of diamond C(110) surface via Si/Ge-triadamantane dimer placement tools,” J Comput Theor Nanosci., Vol. 1, 2004, pp. 71–80; http://www.MolecularAssembler.com/ JCTNMannMar04.pdf. [49] Allis, D. G., and K. E. Drexler, “Design and analysis of a molecular tool for carbon transfer in mechanosynthesis,” J Comput Theor Nanosci., Vol. 2, 2005, pp. 45–55; http://e-drexler.com/d/05/00/DC10C-mechanosynthesis.pdf. [50] Freitas, R. A. Jr., “A simple tool for positional diamond mechanosynthesis, and its method of manufacture,” U.S. Provisional Patent Application No. 60/543,802, filed 11 February 2004; U.S. Patent Pending, 11 February 2005; http://www.MolecularAssembler.com/ Papers/DMSToolbuildProvPat.htm. [51] Peng, J., et al., “Theoretical analysis of diamond mechanosynthesis. Part III. Positional C2 deposition on diamond C(110) surface using Si/Ge/Sn-based dimer placement tools,” J Comput Theor Nanosci., Vol. 3, 2006, pp. 28–41; http://www.MolecularAssembler.com/ Papers/JCTNPengFeb06.pdf. [52] Temelso, B., et al., “High-level ab initio studies of hydrogen abstraction from prototype hydrocarbon systems. J Phys Chem A, Vol. 110, 2006, pp. 11160–11173; http://www.MolecularAssembler.com/Papers/TemelsoHAbst.pdf. [53] Freitas, R. A. Jr., D. G. Allis, and R. C. Merkle, “Horizontal Ge-substituted polymantane-based C2 dimer placement tooltip motifs for diamond mechanosynthesis,” J Comput Theor Nanosci., Vol. 4, 2007, pp. 433–442; http://www. MolecularAssembler.com/Papers/DPTMotifs.pdf.
References
391
[54] Temelso, B., et al., “Ab initio thermochemistry of the hydrogenation of hydrocarbon radicals using silicon, germanium, tin and lead substituted methane and isobutane,” J Phys Chem A, Vol. 111, 2007, pp. 8677–8688; http://www.MolecularAssembler.com/ Papers/TemelsoHDon.pdf. [55] Freitas, R. A. Jr., and R. C. Merkle, “A minimal toolset for positional diamond mechanosynthesis,” J Comput Theor Nanosci., Vol. 5, 2008, pp. 760–861; http://www.MolecularAssembler.com/MinTool/MinToolset.pdf. [56] Lee, H. J., and W. Ho, “Single bond formation and characterization with a scanning tunneling microscope,” Science, Vol. 286, 1999, pp. 1719–1722; http://www.physics.uci.edu/%7Ewilsonho/stm-iets.html. [57] Oyabu, N., et al., “Mechanical vertical manipulation of selected single atoms by soft nanoindentation using near contact atomic force microscopy,” Phys Rev Lett., Vol. 90, 2003, p. 176102; http://link.aps.org/abstract/PRL/v90/e176102. [58] Oyabu, N., et al., “Mechanical Vertical Manipulation of Single Atoms on the Ge(111)-c(2x8) Surface by Noncontact Atomic Force Microscopy,” Abstracts of Seventh International Conference on Non-Contact Atomic Force Microscopy, Seattle, Washington, USA, 12–15 September, 2004, p. 34; http://www.engr.washington.edu/epp/afm/ abstracts/15Oyabu2.pdf. [59] Sugimoto, Y., et al., “Complex Patterning by Vertical Interchange Atom Manipulation Using Atomic Force Microscopy,” Science, Vol. 322, 2008, pp. 413–417. [60] “Nanofactory Collaboration Colleague Awarded $3M to Conduct First Diamond Mechanosynthesis Experiments,” 11 August 2008; http://www.MolecularAssembler.com/ Nanofactory/Media/PressReleaseAug08.htm. [61] “Annotated Bibliography on Diamond Mechanosynthesis,” Nanofactory Collaboration website, 2008; http://www.MolecularAssembler.com/Nanofactory/AnnBibDMS.htm. [62] Freitas, R. A. Jr., and R. C. Merkle, Kinematic Self-Replicating Machines, Georgetown, TX: Landes Bioscience, 2004; http://www.MolecularAssembler.com/KSRM.htm. [63] Vettiger, P., et al., “The Millipede—nanotechnology entering data storage,” IEEE Trans Nanotechnol., Vol. 1, 2002, pp. 39–55. [64] Bullen, D., et al., “Development of parallel dip pen nanolithography probe arrays for high throughput nanolithography,” Symposium LL: Rapid Prototyping Technologies, Materials Research Society Fall Meeting, Boston, MA, 2–6 Dec 2002, Proc. MRS, Vol. 758, 2002; http://mass.micro.uiuc.edu/publications/papers/84.pdf. [65] Protiveris Corp., “Microcantilever arrays,” 2003; http://www.protiveris.com/cantilever_tech/microcantileverarrays.html. [66] Martel, S., and I. Hunter, “Nanofactories based on a fleet of scientific instruments configured as miniature autonomous robots,” Proc. 3rd Intl Workshop on Microfactories, Minneapolis, MN, 16–18 September 2002, pp. 97–100. [67] Koziara, J. M., et al., “Blood compatibility of cetyl alcohol/polysorbate-based nanoparticles,” Pharm. Res., Vol. 22, 2005, pp. 1821–1828. [68] Radomski, A., et al., “Nanoparticle-induced platelet aggregation and vascular thrombosis,” J. Pharmacol., Vol. 146, 2005, pp. 882–893. [69] Brown, E. B., et al., “Measurement of macromolecular diffusion coefficients in human tumors,” Microvasc. Res., Vol. 67, 2004, pp. 231–236. [70] Ashkenazi, A., and V. M. Dixit, “Death Receptors: Signaling and Modulation,” Science, Vol. 281, 1998, pp. 1305–1308. [71] Pan, G., et al., “The Receptor for the Cytotoxic Ligand TRAIL,” Science, Vol. 276, 1997, pp. 111–113. [72] Ambrosini, G., C. Adida, and D. C. Altieri, “A novel anti-apoptosis gene, survivin, expressed in cancer and lymphoma,” Nature Medicine, Vol. 3, 1997, pp. 917–921. [73] Sheridan, J. P., et al., “Control of TRAIL-Induced Apoptosis by a Family of Signaling and Decoy Receptors,” Science, Vol. 277, 1997, pp. 818–821.
392
Medical Nanorobotics: The Long-Term Goal for Nanomedicine [74] Fadok, V.A., et al., “Exposure of phosphatidylserine on the surface of apoptotic lymphocytes triggers specific recognition and removal by macrophages,” J. Immunol., Vol. 148, 1992, pp. 2207–2216. [75] Grakoui, A., et al., “The Immunological Synapse: A Molecular Machine Controlling T Cell Activation,” Science, Vol. 285, 1999, pp. 221–227.
CHAPTER 15
Potential Strategies for Advanced Nanomedical Device Ingress and Egress, Natation, Mobility, and Navigation F. Boehm
15.1
Introduction Nanomedicine is rapidly emerging as one of the most important facets of applied nanotechnology. Envisaged nanomedical devices hold immense potential for initially enhancing, and eventually superseding many conventional medical technologies and procedures. Paradigm shifts are poised to occur in virtually every sector of the medical field, encompassing preventative medicine, diagnostics, therapeutics, tissue engineering, genetics, regenerative medicine, and patient health monitoring/surveillance. (Note: For the sake for brevity in this chapter, nanodevice will be synonymous with nanomedical device.) A significant departure from the status quo will also be apparent toward addressing the cumulative degenerative processes that are involved in the seemingly unavoidable disease state that we presently call aging. Within the next several decades we may well witness the advent of a virtually limitless array of beneficial nanomedical applications related to practically every major human disease state, and injurious condition. When conceptualizing future nanomedical devices and systems, nanoengineers and designers will be required to consider and address an extensive range of challenges. Of these, innovative strategies will have to be devised for approaching an array of problems related to how nanodevices might safely and efficiently enter the human body (ingress); how they will propel themselves and be precisely guided in vivo once internal access has been achieved; and finally, how they will exit the patient (egress) once their assigned medical tasks have been completed. A critical component for any advanced nanomedical procedure that may involve from several hundred to perhaps millions of nanodevices working together in parallel would be a sophisticated and powerful “outbody” navigational control and tracking capability. Contingent upon the particular species of nanodevice that a physician may select to perform a specific medical task, dynamic modes of propulsion would be required to traverse the complex and relatively harsh conditions (from the perspective of a nanodevice) present within the human vasculature, various internal organ structures, as well as among and within the myriad types of tissues and cells.
393
394
Potential Strategies for Advanced Nanomedical Device Ingress and Egress
If a particular “nanomed” assignment should entail the targeting and destruction of cancer cells or the incremental treatment of a tumor that is resident within a patient, for instance, a prescribed “unit” of nanodevices might be administered in the most appropriate manner so as to expedite the procedure. Once internalized, the nanodevices might initially make several circuits of the vasculature, for power up, preliminary orientation, and self-organization. This may also allow time for the initialization, calibration, and lock-in of the navigation system. Ideally, an operating room scale GPS system might be invaluable for facilitating precision in vivo nanodevice navigation. Based on specific coordinates and trajectories provided by previously obtained 3-D diagnostic maps, nanodevices would be guided precisely to targeted treatment sites. Once the treatment is verified as complete, patient egress commands would be issued to guide these entities to the most appropriate exit site. Once gathered at this egress site, all nanodevices in the unit would be accounted for (e.g., via scanning of individual onboard identification tags) and extracted. There may be instances whereby nanodevices might congregate, embed themselves, and power down to become dormant within fingernail or toenail beds, or within the roots of hair follicles for eventual egress via these routes.
15.2
Potential Nanodevice Ingress Strategies One of the many technical challenges to confront nanomedical conceptualists, engineers and designers will involve the formulation of strategies for how future medical nanodevices might be administered and introduced into the human body. Innovative logistics will be required when considering various approaches for the ingress of functionalized microscopic and nanoscopic entities into the patient. Each potential method of entry will be associated with its own attendant set of physiological hurdles. Primary modes for transferring nanomedical devices into the human in vivo environment may include hypodermic injection, aerosol inhalation, ingestion via a pill, or conveyance using a transdermal patch or topical gel. Injection may be the most straightforward, albeit, the most invasive of these ingress methods. Various concentrations of nanodevices, dependant on the specific treatment prescribed for administration, could be infused within an appropriate fluid as a colloidal suspension. 15.2.1
Hypodermic Injection and Dermal Burrowing
There are currently a variety of injection options that might be employed (e.g., intravenous, intramuscular, intradermal, subcutaneous, intraperitoneal, and intraosseous), and these modes may still be utilized for the deployment of early generations of nanomedical devices. Future versions of existing MED-JET [1] or ultrasonic SonoPrep [2] injection systems might be used in similar fashion, but may be damaging to cells and tissues that lay in the direct path of the ballistic pressures and sonic vibrations that are imparted by these devices, in addition to any attendant pain. However, in 10 to 20 years, with the arrival of advanced nanodevices that are endowed with the capacity for automatically burrowing through the skin into
15.2 Potential Nanodevice Ingress Strategies
395
capillaries, (Figure 15.1) and finally into the bloodstream, hypodermic injection might have long been relegated to the annals of medical history. Nanometric telescoping manipulator arms that are envisaged to enable micron-sized ice-burrowing nanorobots to progress at ~1 μm/s might be adapted to traverse the extracellular matrix within the various layers of skin tissue. As to possible negative effects imparted by such motility, some level of discomfort might be caused by nanodevices that may have rough edges or sharp tips that happen to mechanically disturb free nerve endings. They also pose the risk of engaging of the immune system, which may result in itching or a rash. Hence, smooth surfaces are likely to be preferred for the prevention of these types of irritation [3]. Exceptions may be those instances where there are requirements that correspond to particular applications (e.g., vascular plaque removal or the selective “lancing” of undesired cells or pathogens). However, even under these circumstances, any sharp-edged nanoscale instrumentation would likely be designed to be retractable and thus would remain internalized within nanodevices until deployed for a specific function. Spiral-type magnetically driven micromachines were developed by Ishiyama et al. at Tohoku University in 2002, with the envisaged application of burrowing into tumors and killing them via hyperthermia. In experiments, these devices were induced to propel themselves through viscous media, such as gels, by the spinning motion imparted by a rotating magnetic field [4].
Hair shaft Sweat pore Dermal papila Sensory nerve ending for touch Stratum corneum Pigment layer Stratum germinativum
Epidermis
Stratum spinosum Stratum basale
Dermis
Arrector pili muscle Sebaceous (oil) gland Hair folicle
Subcutaneous fatty tissue (hypodermis)
Papilla of hair Nerve fiber
Vein Artery
Blood and lymph vessels
Sweat gland Pacinian corpuscle
Figure 15.1 (Color plate 24)
Cross-section of skin layers.
396
Potential Strategies for Advanced Nanomedical Device Ingress and Egress
The lateral extension of nanometric crampons with each incremental forward stroke of a dermal burrowing nanodevice, which is designed to be unobtrusive and harmless to the patient by virtue of their infinitesimal size, may facilitate their movement through the skin layers, until onboard sensors indicate that they have arrived at a capillary. Dermatologist Adnan Nasir, at Duke University and UNC Chapel Hill, speculates as to whether nanoparticles might penetrate skin and organs through the use of chemical, electrical, or magnetic gradients, and if motility might be enhanced by employing nanoparticles that are fashioned like lances or barbs, allowing for exclusive one-way transit. He cautions, however, that potential tissue layer damage may be caused by this mode of nanodevice migration, as well as the possibility for the initiation of cysts, and injury to the epithelial lining via squamous metaplasia dysplasia, and neoplasia [5, 6]. Because of their diminutive physical dimensions (e.g., ~1 micron in diameter), the envisaged beneficial activities of nanomedical devices should be designed to proceed painlessly and to be completely undetectable by the patient. Verification of the safety of future nanomedical technologies will, of course, will be a critical issue to address concurrently with the progress made in their efficacy. It is likely, in view of the current rapidly escalating surge in nanomedical research, coupled with the growing number of clinical trials involving functional nanoparticles that the positive health effects imparted by even first generation nanodevices may be significant and extensive. 15.2.2
Aerosol Inhalation and Traversing the Blood/Brain Barrier (BBB)
Nanomedical devices might be inhaled as an aerosol via the use of a nebulizer and may enter the bloodstream via the pulmonary capillaries. They would be required (in this scenario) to migrate through the three layers of the respiratory membrane to access the bloodstream. Detailed permeability studies of the respiratory membrane (0.5–1.0 μm thick), comprised of alveolar epithelial membrane, capillary endothelial membrane, and fused basement membrane that separate the two, would elucidate whether ~1 μm in diameter nanodevices might be small enough to diffuse across this membrane, or if an alternate method of vascular ingress from the lungs may be required [7]. Additional investigations would aim to specifically elucidate what effects that lung resident cilia may have on nanodevices (in that the cilia will attempt to “sweep” them out of the lungs, handling them as they would foreign particulates). Nanodevices might also become trapped within the sacs of the alveoli en-route to a vascular ingress site. In regard to neurological maladies, it has been estimated that approximately 99% of drugs that may be of potential benefit are not capable of traversing the blood/brain barrier (BBB) [8]. Nanodevice administration via a nasal spray might be advantageous if, for example, applying nanomedical therapeutics for the purpose of dissolving amyloid plaque material in Alzheimer’s patients, due to the close proximity for the rapid transit of nanodevices through the BBB via the nasal mucosa. Intranasal delivery can apparently circumvent the BBB using pathways through the olfactory epithelium, olfactory and trigeminal nerves [9].
15.2 Potential Nanodevice Ingress Strategies
397
The BBB can be accessed by osmosis or via biochemical or pharmacological methods through the administration of hyperosmotic mannitol and arabinose compounds. They have the effect of reversibly widening the tight junctions of the BBB, thereby increasing the permeability of endothelial cells via temporary dehydrative shrinkage. This transitory state has the disadvantage, however, of increasing the likelihood that undesirable neurotoxic elements may be allowed to pass into the brain [10]. It has been discovered by Krueter et al. [8, 11] and Schroeder, et al. [12–15] that poly(butylcyanoacrylate) nanoparticles (200–300 nm) coated with hydrophilic surfactants (e.g., polysorbate 80) have the capacity for adsorbing a variety of drugs, in unaltered form, to effectively circumvent the BBB to target the brain. It has been determined that the most likely mechanism for the transit of these nontoxic coated nanoparticles into the brain is not by opening the BBB, and thereby making it vulnerable, but rather by receptor-mediated endocytosis [10]. 15.2.3
Transdermal Patch, Diffusive Gel, or Eye/Ear Drops
An adhesive patch or the application of a viscous diffusive gel suspension (e.g., similar to the consistency of aloe vera) that contains a predetermined population of nanodevices may provide a preferable method for administration, as the devices would unnoticeably diffuse through the various epidermal cell layers assisted by paracellular fluid movement (e.g., passage of water between cells), or through the pores, and into the bloodstream [16]. A self-contained “smart” transdermal patch may impart a low voltage into the skin surface that could potentially facilitate nanodevice ingress into the patient [17]. Alternately, a physician may select eye or ear drops for the localized administration of nanomedical devices if specifically treating these areas. Advances to increase the permeability of the skin using patches that are currently under development include iontophoresis, ultrasound, gels, microneedles (Figure 15.2) sonophoresis, lasers, and electroporatic techniques. Iontophoresis
(a)
(b)
Figure 15.2 (a) AdminPatch Microneedles (From nanoBioSciences, LLC, http://www.nanobiosciences.com/), and (b) artistic depiction of ultrasharp nanoscale needles.
398
Potential Strategies for Advanced Nanomedical Device Ingress and Egress
makes use of repulsive electromotive forces that employ a small applied voltage to positively and negatively charged chambers that contain solvents and active ingredients that have an opposite charge. Thus, the chambers act to strongly repel their contents, and as a result, force their expulsion into to skin [18]. Antares Pharma has developed a product called CombiGel that includes a hydro alcoholic gel mixed with enhancers. This transparent gel is formulated for quick transfer through the skin. The company has adapted this technology to produce the first transdermal testosterone gel for men. Though a number of hurdles still exist toward the maturation of this technology, it seems inevitable that many other drugs may soon be administered by employing this technique. Conceptual tissue migrating nanodevices might have a much less cumbersome route to negotiate through epidermal layers before finally accessing the capillaries if they enter via the sweat glands. This may be an appropriate venue for the ingress of nanodevices that are applied as a topical gel suspension or ejected from a transdermal patch. The soporiferous or sweat glands are located in almost every area of the skin, positioned in small pits slightly beneath the corium (a deep sensitive layer under the epidermis), or more commonly, in the subcutaneous areola (small spaces in between fibrous tissue), surrounded by a mass of fatty tissue. Each sweat gland is comprised of a single tube with a deeper section that is shaped like an oval or sphere, which is the body of the gland. The shallow part, or duct, opens at the exterior of the skin as a funnel-shaped orifice [19]. On the palm of the hand there are ~370 pores per/cm2, back of the hand ~200 per/cm2, forehead ~175 per/cm2, breast, abdomen, and forearm ~155 per/cm2, and on the leg and back from ~60–80 per/cm2. A typical skin pore diameter is ~50 microns, and the estimated pore population for the epidermis for the entire human body is ~2 million [19]. Entry through the pores of the palm of the hand might prove to be a good strategy for nanodevice ingress due to the optimal number of potential entry points, and hence allowung for expeditious device diffusion.
15.3
Molecular Motors Various categories of molecular motors, whether they are manifest as completely synthetic, bio-based, or hybrid constructs, will likely form the hearts of nanomedical device propulsion systems. Their critical task will be to efficiently convert chemical or thermal energy, which is harvested from the in vivo environment of a patient, into molecular level mechanical torque, providing lateral, radial, or linear thrust. Alternatively, motor components might be activated and manipulated by outbody systems that broadcast photonic, acoustic, magnetic, or radio frequency stimuli to induce the generation of useable voltage within nanodevices. As relates to the activation and kinetics of molecular motors, some type of reversible switch may be incorporated into the system. Switches are not capable of utilizing chemical energy to sustainably drive a system away from equilibrium, whereas motors can have this ability. The majority of molecular machines to date (2008) may be classified as switches that toggle between “on” and “off” states, rather than motors, which impart force to travel along a certain trajectory. The chemistries of these systems must, at a fundamental level, be capable of constraining
15.3 Molecular Motors
399
the motion of its components whose changes of position in three-dimensional space are stimulated by an external energy contribution [20]. 15.3.1
Powering Molecular Motors
There are a range of stimuli, as mentioned above, which might be employed for the activation and control of molecular motors within nanodevices. These may consist of pulsed signals that emanate from dedicated beacons that are situated externally to the patient. Alternatively, or working in conjunction with these signals, the provision of energy for powering molecular motors may be derived from the in vivo environment, via controlled chemical reactions or the extraction of energy from thermal fluctuations. The second law of thermodynamics will forbid the conversion of heat into useful work if there is no temperature differential that exists between the ambient in vivo environment and the components residing within nanodevices. It should also be noted here that temperature gradients cannot be sustained over nanoscale distances. There will be a requirement for the continuous movement of systems away from equilibrium via the continuous input of external energy. The preservation of a thermally initiated stepping down process will bias Brownian motion toward equilibrium [21]. By managing to adapt to and compensate for Brownian motion, an important step will have been taken toward the powering and controlled manipulation of nanomedical devices. A subsequent and critical issue, however, will concern how harvested or generated power might be conveyed to perform mechanical tasks at the nanoscale. 15.3.2
Piezoelectric Elements
Piezoelectricity is a unique mode of voltage generation that might exhibit utility for the powering of nanomedical devices. Piezoelectric elements (e.g., quartz, zinc oxide nanowires, lead-zirconate-titante, or PZT), are comprised of crystalline structures that will create a voltage when mechanically stressed or deformed. Conversely, these materials will physically deform if a voltage is applied to them (e.g., enabling nanoscale actuators or artificial muscles). Perhaps supplemental energy might be harvested from the action of Brownian motion on the external surfaces of nanodevices if they were to be studded, for instance, with arrays of ultra thin, yet stiff piezoelectric zinc oxide nanowires or nanoribbons (Figure 15.3). Hence, they may prove to have significant utility when designed into such diminutive mechanisms. 15.3.3
Molecular Propellers
Molecular scale propellers might be devised (Figures 15.4 and 15.5) where tight steric (atomic level packing conditions) prevail. This may initiate the formation of molecular scale blades derived from helically configured structures [22–24]. The rotation of molecular elements can be made to occur around C-C single bonds by interfacing triptycene (an aromatic hydrocarbon) with other molecules that have
400
Potential Strategies for Advanced Nanomedical Device Ingress and Egress
Conductive plate Piezoelectric nanoribbons (static)
Piezoelectric nanoribbons (mechanically stressed) Conductive substrate Cumulative voltage generation
Figure 15.3
Conceptual piezo power generation.
Streptavidin
Actin filament Peripheral stalk
ATP ADP + P1 Central stalk
Proton half-channel
Phospholipid bilayer membrane Proton half-channel C-ring rotation (only orange components rotate) C-ring
Figure 15.4 (Color plate 25) ATP synthase-based nanopropeller. (Adapted from W. Junge, et al., TIBS, 22 (1997) and Duncan et al. Proc. Natl. Acad. Sci., 92 (1995).)
complementary affinities [25]. Induced conformational changes or the altered orientation of molecular elements (e.g., ions) can be employed as braking mechanisms.
15.4
Constraints on Molecular Motors There are several constraints, as described below, which will be imposed on the functionality of molecular scale motors. They will act to impede their operation
15.4 Constraints on Molecular Motors
401
N Me2N N02
Si
S S S Figure 15. 5
Dipolar rotor. (From [9].)
unless the molecular motors can be cleverly conceived and designed to harness these forces to perform useful work. 15.4.1
Brownian Motion
The inception and foundation of the possibility of synthetic molecular machines can be ascribed to Robert Brown, the Scottish botanist, who observed the relentless random motion of particulates within pollen grains that were suspended in water [26]. This ceaseless molecular activity, commonly known as Brownian motion, will have a major influence over, and impacts on, any mechanical apparatus that is intended to operate at the nanoscale. This is a critical issue that will have to be reckoned with, for no matter how exquisitely fabricated or efficiently operating nanodevices are envisaged to be, they will have to somehow accommodate, or ideally exploit, these fluctuations. Despite the effects of these thermal vibrations, we are compelled to recognize that useful and prolific work may indeed be accomplished at nanometric scales. This is definitively evidenced by the elegant and precise functionality of natural biological molecular machines such as the multisubunit ribosome. This amazing molecular assemblage rapidly churns out nanometric linear polypeptide chains that automatically, and for the most part, flawlessly fold into the myriad of proteins that are vital for the functionality of living organisms. Biological motor proteins devour ATP (adenosine triphosphate) molecules at a rate of 100 to 1,000 per second. This correlates to an energy output of 10–16 to 10–17 W per molecule. When we consider that the constant and random collisions mole-
402
Potential Strategies for Advanced Nanomedical Device Ingress and Egress
cules are subject to in aqueous media are equivalent to 10-8 W, it is indeed extraordinary that such precise and organized operations can be accomplished at all [20, 27, 28]. 15.4.2
Brownian Shuttles
Catenanes and rotaxanes are interlocking mechanical systems whose degrees of freedom are highly inhibited by mechanical bonds. However, the remaining allowable direction of movement can be exploited to serve as linear shuttles (Figure 15.6). Rotaxane-based shuttles can linearly relocate in response to external stimuli (e.g., fluctuations in temperature) [29] that will disrupt the systems equilibrium (e.g., destabilization of a favored binding site or increasing the binding strength of a less attractive site). The resolution of positional control can occur for distances in the 15Å range over 100-μs timelines [20]. The first authentic switchable Brownian motion-driven molecular shuttle that utilized a dual station design was reported in 1994. This shuttle could change position by the addition or removal of electrons [30]. 15.4.3
Viscous Forces
There is another critical issue to address that will compound the pervasive presence of Brownian motion at the nanoscale, and which will place further constraints on the motility and general functionality of nanomedical devices. Macroscopic level motion is directed by inertial forces. When we shrink to the mesoscopic and nanometric domains, however, inertia no longer has effect and viscous forces come
Figure 15.6
Molecular shuttle. (From [9].)
15.5 Traversing the Circulatory System
403
into play. Bacteria, at the size range of about 0.2 to 5 μm, and viruses, with dimensions of from 0.005 to 0.1 μm are already under its influence. Viscosity at the molecular level is quantified by the Reynolds number (e.g., ratio of inertia to viscous forces). At very low Reynolds numbers nanoscale entities will be devoid of momentum and hence cannot be set in motion once pushed, as is experienced in the macroworld. Together with Brownian motion, the ambient viscosity of blood plasma, and that which exists within the cytoplasmic interiors of cells, will influence the motion and operation of nanomedical mechanisms. Therefore, any nanodevice design should include strategies for the compensation or exploitation of these molecular forces. The high ratio of surface area to volume will make nanodevices inherently sticky (e.g., via surface charges) and will therefore have a significant impact on how they interact with each other and when engaged with the cells and tissues they are designed to diagnose and treat. We cannot presume that there will be any linear correlation insofar as the form and function of macroscale mechanisms and those that will operate at the nanoscale. Motors and machines do, however, operate quite efficiently at the nanoscale as is validated by biology [27]. It is quite likely that once a deeper elucidation of critical natural processes has been achieved, the resulting insights will serve as a guide to assist with envisaging useful synthetic nanomedical constructs. A variety of apparatus within biological systems have evolved over millennia that utilize membrane encapsulated partitions that reside and function within cells and organelles. Disparate gradients are thus established and maintained between discrete cellular compartments and those which exist within organelles, which are mediated by membrane bound channels and gates. The transit of charged ions and other carrier groups facilitate movement between segregated compartments [31]. These systems typically exhibit optimal performance when they are kept far from equilibrium and operate through a transitory relaxation from a higher to a lower gradient, which moves the system toward thermodynamic equilibrium, but never achieves it. This sophisticated strategy allows natural systems at the nanoscale to perform useful work despite the significant constraints imposed by both Brownian motion and aqueous viscosity. Biomembrane embedded ion channels operate via the coupling of structural alterations with charge mediated binding affinities at particular transmembrane sites. These transformations direct the availability and access to these sites from either side of the membrane, and hence control ion direction and flow [32].
15.5
Traversing the Circulatory System 15.5.1
Whole Blood Composition and Viscosity
At a fundamental level, whole human blood might be considered as a colloidal suspension. This non-Newtonian fluid is comprised primarily of water, and is combined with red cells, white cells, proteins, platelets, nutrients, electrolytes, individual hormones, gases, and metabolic waste products suspended in an aqueous plasma media. The viscosity of the blood plasma (1.2 centiPoise (cP) at 37ºC), which is itself a Newtonian fluid, is slightly higher than that for water (~1 centiPoise (cP) at 20ºC) [33].
404
Potential Strategies for Advanced Nanomedical Device Ingress and Egress
Formidable technical obstacles will confront nanomedical engineers toward the development of advanced autonomous nanomedical devices when evolving strategies for potentially traversing the nearly 19,000 km of human vasculature under outbody computer control. Some questions for consideration may include: 1. What might the optimal mode/s of propulsion be for nanodevices deployed within the aqueous whole blood environment with its varying levels of viscosity, blood velocity, and shear flow? Investigative simulations of symmetrical propulsive power strokes, imparted by nanoscale fins or blades in viscous aqueous fluids with low Reynolds numbers, have revealed that an initial power stroke will allow an incremental movement forward. However, on the second power stroke, the device will return to its original position. It is these types of insights that may facilitate the development of asymmetrical propulsive elements that may successfully propel nanodevices through the viscous media of the human vasculature [34]. 2. How will nanomedical devices manage to successfully navigate through the myriad of arterioles and arteries (~0.1–24.0 mm in diameter), venules and veins (~0.15–30.0 mm in diameter), and capillaries (~4.0–8.0 μm in diameter) [35]? An additional challenge when navigating the circulatory system will involve the development of intuitive algorithms to assist with correctly directing autonomous nanomedical devices as they approach innumerable bifurcations (divisions of the vasculature into smaller branches). On this topic, innate logic would dictate that for all intents and purposes nanomedical devices that are destined to traverse the human vasculature should always “go with the flow.” Although there may be contingencies whereby nanoscale devices would be required to venture upstream against prevailing vascular currents (e.g., for the almost instantaneous repair of serious injuries or the stabilization of sudden physical trauma), it seems likely that the majority of diagnostic and therapeutic operations might be effectively conducted within reasonable timelines when nanodevices travel with the blood flow. To rectify those instances where nanodevices should happen to miss their mark, or be somehow blocked from arriving at their intended exit sites, they might reposition themselves and take corrective measures during subsequent rounds through the circulatory system (typically ~60 seconds in duration) [35]. If nanodevices are deployed to work in massively parallel fashion, they would likely be communicating with each other as well as with external sources. It is thus conceivable that at least one nanodevice will arrive at an assigned in vivo diagnostic or treatment site. 3. How might nanomedical devices avoid, or deal with, inevitable collision events with red blood cells (RBCs) and other blood resident constituents? When one considers that there are approximately 5 billion RBCs per milliliter of blood, exclusive of all other suspended blood-borne elements, this fact will definitely give new meaning to the term “collision avoidance” for nanomedical devices [33]. As relates to nanodevices that are designed to
15.8 Nanometric Biomimetic Analogs for Potential Nanomedical Device Motility
405
traverse the vascular system, it is inevitable that multiple collision events will occur during a diagnostic scan, or a particular course of therapeutic treatment. To maintain precision in the case of the acquisition of in vivo spatial data, appropriate compensatory measures (e.g., dedicated algorithms) might be implemented. It may turn out to be far less problematic to integrate ruggedized features into the designs of nanodevices to counteract or tolerate the constant pummeling that they will undoubtedly encounter from a diverse array of biomaterials suspended in blood or lymph, than to risk overloading primary systems. Corrective measures, outside the scope of normal operating procedures using propulsive and navigation systems, might be implemented only in extreme instances, where nanodevices have been knocked far off course, or have somehow become trapped. For earlier generations of nanomedical devices, if enough identical entities are deployed, the perceived complexity of a given task might be reduced to a statistical probability issue. In one hypothetical scenario, for example, if only 700,000 out of a million injected nanodevices manage to successfully accomplish a task at hand with enough efficacy, and within an appropriate timeline (e.g., targeted delivery of potent drug molecules, photothermal therapy at a tumor site, or molecular materials transport to osteoporosis ravaged bones), this percentage may be deemed as sufficient to qualify the treatment as a success. With each additional circuit through the vasculature this percentage might also be likely to increase. Therefore, optimal treatment exposure times might be established and standardized for particular classes of nanomedical devices intended to address specific conditions. A high degree of inherent redundancy for all critical components and systems should be included as a matter of course when considering any advanced nanodevice design. Ideally the goal would be to have all nanodevices complete their assigned tasks with negligible error levels. There have been investigations into collisions of rigid and pliable spheres, hydrodynamics of individual swimming cells, interactions between two parallel swimming cells, and solid wall effects [35–43]. When flagellar entities approach a boundary there is a reduction in velocity of ~5% at a distance of 10 object radii from the boundary [44]. In addition, when two cells are traveling side by side they will be attracted to each other as opposed to a repulsive action that will occur when they are swimming one in front of the other [45]. An observation made by Purcell states “Turn anything—if it isn’t perfectly symmetrical, you’ll swim.” Within the viscosity-dominated in vivo domain, locomotion designs based solely on reciprocal deformation or thrusting will not make forward progress. This is based on the so-called “scallop theorem.” The device will move forward as the result of an initial power stroke; however, it will then revert back to its original position [34, 46]. Nanomedical devices might utilize mechanisms that are analogous to ciliary propulsion, as exhibited in Paramecium caudatum. This ciliate uses the ~2,500 cilia on its outer surface to propel itself, and can adjust its beating wave patterns to traverse a range of viscous environments [47–49]. An appropriate velocity for an in vivo nanomedical device might be set at ~4–8 cm/sec, having a shear force of ~2–6N/m2 . This is within the normal range of human blood and most likely nonthrombogenic to platelets. However, Freitas pro-
406
Potential Strategies for Advanced Nanomedical Device Ingress and Egress
poses a conservatively safe nanodevice speed limit of 1cm/sec based on potential red cell impact pressures. A higher velocities, device/cell impacts may cause changes in red cell surface area and possible damage. RBCs are subject to deformation at ~>2m/s and may rupture at ~>20m/s [3]. Swimming speeds of certain species of sperm cells approach ~100 to 200 μm/s [50]. Some other examples of in vivo locomotion might include devices that employ screw and corkscrew drives [51]. Bacterial flagellum is activated by a ~0.0001 pW motor that can rotate up to 300 Hz at 310K (~15 Hz under load). It can reverse direction in ~1 millisec, and burns up 0.1% of its metabolic energy (under growth conditions) to rotate the flagellum [45–49]. The E-colibacterium can swim at a velocity of 30 μm/sec having a thrust force of 0.5 pN at less than 1% efficiency [52].
15.6
Traversing the Lymphatic System The lymphatic system is comprised of a network of organs, tissues, nodes, vessels, and capillaries that serve a number of important functions: •
•
•
Collects and drains protein containing interstitial fluid from the intracellular spaces of tissues, which has leaked from blood capillaries; Assists with the transfer of fats from the gastrointestinal tract back into the bloodstream; Serves as a sentinel infrastructure for the immune system to provide protection of the human body from “non-self” cells, microbes, and cancer cells through the use of lymphocytes working in cooperation with macrophages.
Lymphatic capillaries have microscopic openings that exist between the endothelial cells that make up its walls. Fluid flow can proceed into these capillaries but it is blocked from exiting. The entire lymphatic system is equipped with a series of one-way valves to ensure delivery of the lymph to the thoracic duct and into the bloodstream through the right heart [33]. Future nanomedical devices may be directed to traverse the lymphatic system subsequent to any given diagnostic or therapeutic procedure. This strategy may be advantageous in that it might be much easier to locate egress sites utilizing this slower, low pressure aqueous environment, which will be relatively clear of circulating elements in comparison whole blood. Hence, there would be a far lower energy expenditure required for the engagement of collision avoidance maneuvers within the lymph. One issue to be mindful of when traversing the lymphatic system might be the requirement to compensate for changes in viscosity that will likely be encountered as nanodevices make their way through the body.
15.7
Phagocyte Avoidance Strategies Ideally, for many diagnostic and therapeutic applications, administered nanomedical devices would arrive at their targets, complete their assigned tasks, and proceed to exit the patient before eliciting detection and response by the immune
15.8 Nanometric Biomimetic Analogs for Potential Nanomedical Device Motility
407
system. From a practical standpoint, virtually all classes of motile nanodevices that are prescribed to remain in vivo for longer periods should be enabled with a suite of capabilities for the avoidance of ingestion by circulating phagocytes. The most direct and successful strategy for circumventing an immune response might be to sheath the exteriors of nanodevices with inert biocompatible nanomaterials (e.g., diamondoid, sapphire) or compounds (e.g., polyethylene glycol – PEG). With the advent of advanced nanodevices, integrated immune response circumventing strategies might include the initiation of evasive maneuvers in order to avoid physical contact with white cells via some form of dedicated identification and proximity detection. If nanodevices do happen to collide with phagocytes, which may indeed occur quite frequently while they inhabit the vasculature, protocols for the prevention of binding to their surfaces might be instituted. It has been suggested that the surfactant sodium dodecylsulfate might be released by nanodevices in these cases to prevent antigen-antibody binding [35]. As estimated by Freitas, the velocity of blood within a 1 mm in diameter artery is about 100 mm/s [35]. If there are a total of 1012 nanodevices in the bloodstream, the probability of a collision with a white cell for each nanodevice of 2 μm size might be once every ~3 seconds along the inner surface of the lumen, and about once every ~300 seconds along the central luminal axis. The process of macrophage ingestion may take from 10 or 20 seconds, up to a half-hour to conclude, depending on the size of the particulate that is being internalized. Therefore, nanodevices should have ample time to identify approaching macrophages, to escape from those that are in their pursuit, and to take appropriate actions for avoiding them [35].
15.8 Nanometric Biomimetic Analogs for Potential Nanomedical Device Motility and Ambulatory Movement As alluded to above, nature will undoubtedly provide invaluable inspiration for the conceptualization and design of a wide range of nanometric propulsive mechanisms for use within the in vivo aqueous environments of the human body. It is probable that important lessons will be gleaned from the natural world, which will assist with the endowment of therapeutic nanodevices with the capacity for penetrating cell and organelle membranes, and to enable their ambulatory traversal of internal and external biosubstrates. Some useful elucidations as to the primary characteristics of molecular biological systems that might be taken into account when considering the design of potential nanomedical biomimetic analogs have been listed by Kay et al. [20]: • •
•
Biomechanisms are labile. Nanometric biosystems operate at close to ambient temperatures, and so any generated heat will be dissipated almost immediately. Temperature differentials are not available to them for exploitation. Biomotors utilize chemical energy (e.g., breaking of covalent bonds, formation of high energy compounds such as ATP, NADH, NADPH, and the use of concentration gradients).
408
Potential Strategies for Advanced Nanomedical Device Ingress and Egress
• •
•
•
•
•
•
Biomachines typically function in solution, or at highly viscous surfaces. Brownian motion is exploited rather than opposed. (a) Biomolecular entities need not employ chemical energy exclusively in order to attain mobility. Because of these thermal perturbations their components are never at rest. Hence, this activity can be harnessed via ratcheting. (b) Continuous thermal motion within minute reaction vessels such as organelles and cells guarantee an exceptionally quick and thorough mixing of bio molecules, regardless of the highly viscous environment. Low friction, smooth surfaces are not required by bioentities that are continually subject to thermal activity in the aqueous, high viscosity environment that exists within the human body. Molecular scale, biological mechanisms such as kinesin and other entities utilize infrastructures (e.g., tracks) to constrain their degrees of freedom to accurately perform critical operations. Ion pumps maintain functional integrity through the use of compartmentalization, which negate the chances of ions interacting where they shouldn’t. Biomachines exploit aqueous media, and are controlled in this environment via the utilization of non-covalent interfaces. Most bioentities are comprised of a surprisingly small set of constituents (e.g., amino acids, nucleic acids, lipids, and saccharides). Biological life forms function far from equilibrium. This condition is initiated and sustained by the separation of processes via the use of discrete compartments (e.g., vesicles, organelles, and cells).
15.8.1
Cilia and Flagella
Microorganisms manage to convey motility through the use of functional components such as bacterial flagellar motors. Cilia exist in the groups of protozoa, sponges, coelenterates (hydras, jellyfish, sea anemone, corals), ctenophores (comb jellies), turbellarians (flatworms), rotifers (small ~1,000-celled animal), annelids (worms and leeches), echinoderms (sea stars), ectoprocts (moss animals, filter feeders), tunicates (sea squirts), and vertebrates. The common function of cilia and flagella organelles relates to the transfer of fluids relative to their attached orientation. If the entity bearing the cilia or flagella is diminutive enough, these organelles will move the body, conversely fluid will be moved over the surface of a static body. The difference between cilia and flagella is essentially a functional one. Cilium fluid movement is at right angles to its long axis, whereas for flagella, fluid movement is along the length of the central axis. Cilium moves fluid only during part of its beat cycle, but flagella moves fluid constantly. Therefore, the flagella are more efficient as they work cumulatively to propel a body forward. In bacterial flagella, protons are transited through the motor and their energy is extracted as they pass through the electrically charged cell wall [52, 53]. In flame cells (found in some invertebrates), the flagella run along the interior lumen of tubular channels such that their activation propels fluid toward an external opening, from base to tip. This might be accomplished with cilia; however, their
15.8 Nanometric Biomimetic Analogs for Potential Nanomedical Device Motility
409
bodies would have to be oriented at right angles to the intended direction of motion. Flagella is not in all cases longer than cilia, as cilia may be compounded to form structures much longer than a flagellum. Flagella incorporate several bending waves within their length. Length may be limited by factors such as inefficient metabolite transfer to furnish energy for contraction [54]. When many cilia move in concentrated groups they exhibit the ciliary beating pattern, which may be more efficient than the flagellar pattern, and more adaptable (e.g., in instances where a reversal of direction is required). A consistently even flow of fluid is ensured via rhythmic beating (which is synchronized metachronically) of the cilia. At any given time there will be cilia in various phases of their stroke, some will be in active phase and others will be in recovery and preparing to initiate another. Cilia moves a shallow layer of fluid over the cell surface, but can produce further reaching currents when longer and sweep through a more substantial liquid volume. The comb-plates of ctenophores, comprised of compound cilia, are perhaps the extreme depiction of these propulsive entities. Their dimensions are such that a succession of plates beat and move through water analogous to a paddle wheel [52, 54, 55]. Coordination of flagella beating patterns may be accomplished by mechanical interactions imparted and transferred through water [52]. Flagella attached to small bodies displace water from base to tip initiating a forward thrust, but in the process cause gyration and rotation. Cilia, flagella, and sperm tails share a common structural design (Figure 15.7). Cilia are made up of a number of longitudinal fibrils, geometrically arranged as a ring of nine duplets surrounding a central two. A membrane that is continuous with Inner dynien arm Outer dynien arm
Plasma membrane
Nexin
Central pair of singlet microtubules Radial spoke Spokehead
Inner sheath
A tubule Figure 15.7
B tubule
Flagellar doublet. (From B. Huang, et al., Cell, 29 (1982).)
410
Potential Strategies for Advanced Nanomedical Device Ingress and Egress
the cell membrane sheaths the entire bundle (axoneme). The diameter of cilia is ~0.15-0.3 μm and ~5 to 20 μm long. They can be combined to attain lengths of ~2,000 μm. (flagella length is ~5–150 μm, with sperm tails at ~200 μm). The outer membrane of the main shaft is a three-layered lamination, two dense layers ~20- to 30Å thick, sandwiching a less dense layer that is 30Å thick, giving a total thickness of ~70 to 90Å [54]. The two central tubular fibrils run parallel for the entire length of the shaft, conferring a bilateral symmetry. These are positioned at ~300 to 350Å from center to center, with a total diameter of ~150 to 250Å and a ~40 to 50Å wall thickness. A sheath can link or surround the two central fibrils. Around the central duplet are nine longitudinal peripheral fibrils that form a cylindrical diameter of ~1,600Å. Each of these is a double set of tubules ~200 to 250Å in diameter with ~60Å thick walls. One tubule (A) is larger than the other and is comprised of 13 protofilaments, and the other (B) has 10, and the connection between them is strengthened by a 2 nm in diameter by ~48 nm long protein called tektin (α-helical structure). This protein runs longitudinally along the joining wall between A and B [54, 56]. Attached to the A tubule of each duplet are inner and outer dynein arms that reach out to adjacent B tubules. These may assist in the sliding of the tubules past one another during the beating motion. Three sets of cross-linked proteins bind the axoneme together. The central pair is tethered by periodic bridges, and the outer doublet tubules are joined by the nexin protein, spaced at 86-nm intervals and are most likely elastic to accommodate the sliding duplets as well. A third linkage is comprised of radial spokes. These emanate from the central pair and connect to each A tubule of peripheral duplets, are arranged in pairs that have a 96-nm periodic distance. The diameter of the cilia gradually decreases to the tip and individual tubule lengths end at varying intervals as they elongate toward the tip. Where the cilia attaches to the cell, the axoneme links with the basal body containing nine triplets of microtubules [54, 56]. 15.8.2
Myosin and Actin
In muscle tissues, myosins are thick filaments (~12–18 nm), whereas actins are thin filaments (~5–8 nm). Myosins interpenetrate actins and utilize ATP to temporarily attach the cross-bridges present on its surfaces to the actin filament. Thus, myosins traverse along actin filaments via sequences of binding events to initiate muscle contraction [30, 57]. 15.8.3
Kinesin and Dynein
Biomimetic versions of kinesin and dynein walking may be employed to facilitate the design of surface-roving nanodevices for nanomedical applications. The transportation of molecular scale materials within cells is facilitated by entities such as kinesin and dynein. Kinesins are two-legged molecular motors that transfer vital molecular-scale payloads within cells by traversing hollow cylindrical microtubules. They employ a head over head “walking” motion that is powered by the cleaving of ATP molecules at its “heads.” They are responsible for the separation of chromosomes during cell division, and the transport of nerve cell neurotransmitters.
15.11 Hypothetical Concept for Clinically Localized GPS Navigation
15.9
411
Nanodevice Aqueous Motility A diverse range of strategies might be employed for efficiently propelling nanomedical devices through various types of aqueous media within the human body. Described below are a number of potentially feasible techniques toward the eventual realization of this capability. 15.9.1
Biomimetic Flagellar Propulsion Using Nanotubes
A microrobot has been conceptualized and designed to swim inside the human ureter for the purpose of destroying kidney stones noninvasively. The device uses multiwall carbon nanotubes that serve as biomimetic synthetic flagella, driven into rotating helical profiles by micro motors (Figure 15.8). Estimated swimming speeds of 1 mm/sec were deemed possible using 1 nW of power. Two orthogonal comb drives per motor (e.g., the aim is high efficiency) are used to convert electric potential into mechanical work, and imparts rotating motion to the nanotubes. A thin wire or radio receiver allows for external communications exchange [59]. The microrobot is 1 mm3, nanotube radius is ~30 nm, and the rotation translating substrate is about ~100 μm long. A swimming speed of 0.5 mm/sec was projected at a displacement of 10 μm at 100 Hz with an efficiency of 2%. Assuming that the rotary comb drive could achieve a torque of 130 μN·nm in operation with an efficiency of 0.1%, the total power draw would be 1 nW [59].
Multiwall carbon nanotubes
Rotating base
Figure 15.8
Nanotube flagella. (From [59].)
412
Potential Strategies for Advanced Nanomedical Device Ingress and Egress
15.9.2
Nanoscale Earthworm Analog
One team has devised a propulsive method that consists of three spheres connected by two rodlike structures. One actuation cycle involves first shortening the left arm, and then shortening the right arm. Next, the left arm is lengthened and then the right. The resulting movement is akin to an earthworm pushing through soil (Figure 15.9) [60]. A swimming robot only a few centimeters long has also been developed that uses polymer actuators, sensors, and a power source [54–63]. 15.9.3
External Magnetic Propulsion
A magnetic means for propelling a microrobot having an embedded ferromagnetic core uses a strong and variable magnetic field emanating from a MRI machine. The MRI magnetic field is capable of exerting forces in three dimensions and its inherent imaging attributes can assist in tracking device displacements and the acquisition of positional feedback. The proviso is that external magnetic forces must dominate over the drag force of the blood flow impacting on the microrobot [64]. 15.9.4
Ultrasonic Peristaltic Propulsion
The smallest ultrasonic rotary motors fabricated to date are ~1.4 mm in diameter by 5 mm long [65]. In the linear category, German company Physik Instrumente has developed the world’s smallest ultrasonic piezomotor linear drive. This device has dimensions of 9 × 5.7 × 2.2 mm, has velocities of up to 80 mm/sec, and operates at 3V [66]. Progress has been made in the fabrication of ultrasonically activated pumps having no moving parts (Figure 15.10). A unique volume displacing mechanism using flexural traveling waves operates peristaltically to displace fluid. This innovation
Figure 15.9
Biomimetic earthworm. (From [60].)
15.10 Ambulatory Nanomedical Devices
413
Piezoelectric polarization Piezoelectric actuator
Flexural traveling waves Induced chamber
Figure 15.10
Drive elements
Peristaltic propulsion. (From [68].)
allows the induction of the pumping effect while eliminating of any valves or moving parts. The pumping action is accomplished by utilizing multiple chambers formed between the peaks and valleys of a traveling wave [67, 69]. Propulsion through the human vasculature using ultrasonic propulsion may be possible contingent on investigations into, and the quantification of, safe operating parameters for such devices in vivo. An in-depth study may be warranted to elucidate the specific effects of sustained ultrasonic emanation from nanodevices on human whole blood components and tissues. 15.9.5
Nanofluidic Channels: Behavior and Potential for Propulsion
Nanofluidic propulsion systems would most likely be dealing with quite different environments than their microfluidic counterparts. There are totally different fluidic behaviors apparent within this domain in that the physical device dimensions are on par with relative length scales of elements within the fluids themselves. The distances involved are so small that diffusion processes prevail in mass and heat transfer at very small timescales [71]. Studies have been conducted involving fluid flow through micro/nano channels with diameters ranging between 20 nm and 20 μm with applied voltages of –0.4 × +0.4V. Under conditions of partial double layer overlap, asymmetrical I-V behavior was observed. The primary transport mechanism driving fluid through the orifice was via electro-osmosis. This shows potential for the design of nanochannels having rectified eletro-osmotic flow properties [71–73].
15.10
Ambulatory Nanomedical Devices Ambulatory classes of nanomedical devices might be deployed in vivo to traverse cellular interfaces within various tissues in order to access diseased sites, or to
414
Potential Strategies for Advanced Nanomedical Device Ingress and Egress
deliver bone rebuilding materials specifically to bone resident void sites for repair in osteoporosis patients. These same nanodevices might also be prescribed over the course of multiple treatments to add bone mass or to “top dress” particularly fragile areas on certain bones. This capability might be envisaged, as well, for facilitating the maintenance of bone mass for astronauts during long space missions, as an enhancement in addition to their regularly scheduled exercise regimes. Nanomedical devices might also be programmed to either penetrate or traverse the entire external surfaces of various organs; certain sections of the vasculature; traverse the blood/brain barrier to locate and dissolve beta amyloid plaque material; make their way to specific tissue areas anywhere within the human body via the extracellular matrix to deliver drugs; scan for precursors of disease; or to perform molecular level biopsies on site. In addition, they might be deployed for more long-term “body security” tasks. For instance, they could serve as mobile sentinels to potentially enhance weakened immune systems, or to make healthy immune systems even more robust, enabling very rapid responses for the eradication of any threatening contagion, toxin, or other biochemical threat. Within air-exposed cavities (nasal, oral, ear canal) and external human tissues, ambulatory nanomedical devices might “patrol” dermal surfaces and subdermal layers for signs of disease, perform on-site treatments for melanomas, or to effect cosmetic repairs. In the field of nanodentistry, dedicated, or ideally, multifunctional nanodevices might be administered by a mouthwash to survey all tooth enamel surfaces, and the interfaces between the teeth and gums to eradicate accumulated plaque material and bacteria, and to perform repairs on cavities. Cell-sized entities under external computer control might be deployed to provide painless, yet highly effective anesthesia. Multitudes of nanodevices would migrate, via ambulation, to the pulp chambers of all teeth within several minutes. They would then stand by, poised to disrupt local nerve impulses, and hence would instantly numb any tooth on command, as issued by the dentist. This operation would be completely reversible, and on orders from the dentist’s computer, normal nerve impulse flow would be restored [74]. 15.10.1
DNA Robot
Ned Seeman’s group at New York University were the first to succeed in creating a nanoscale biped. This DNA robot uses 10-nm long segments of DNA as its legs, which can walk along a track that is also comprised of DNA, by performing sequential attachment and detachment operations. Each leg is 36 base pairs in length and is comprised of two oligonucleotides that combine to form a duplex, which is connected at the top. At the lower “foot” portion of the assembly, an extra length of single-stranded sticky DNA protrudes from each of the duplex legs. These entities are immersed in a nondenaturing buffer solution, to prevent DNA degradation. The DNA track has segments of unpaired bases studding its surface to serve as footholds, and that are designed to bind separately with either the left or right foot. Single strands called anchors bind to a foot on one end and a foothold on the other. To walk, a free section of DNA called an unset strand is added that preferentially binds to the anchor strand and thus strips it away, which liberates the foot [75].
15.11 Hypothetical Concept for Clinically Localized GPS Navigation
15.10.2
415
Nanowalker
A research group lead by Ludwig Bartel at the University of California, Riverside, attached walking linkers serving as feet for a 9, 10-Dithioanthracene, (DTA) molecule. The resulting nanowalker molecule exhibits bipedal motion to traverse a flat copper substrate. It is supplied with heat energy, which it utilizes to alternately lift only one linker at a time to traverse a flat plane, in a straight line, without the use of guidance rails or grooves. It was demonstrated that this walking molecule could make 10,000 steps flawlessly [76]. In a further development, the group managed to induce a 9, 10-dioxanthracene (anthraquinone) molecule to transport a payload of two CO2 molecules [77].
15.11 Hypothetical Concept for Clinically Localized GPS Navigation Applied to Advanced Autonomous Nanomedical Devices Investigations into the potential for precise nanomedical device positional determination and effective guidance via triangulation in a clinical setting may be a worthy endeavor. Navigational control of singular and multiple in vivo nanodevices from external outbody sources may perhaps be accomplished by using beacons as analogues of arrayed satellite configurations that are typically employed as components of commercial and military GPS systems. Might it be possible to use near-infrared light (e.g., full body coverage, programmable lasers) to accurately guide and track multitudes of ~1 micron nanomedical devices inside the human body? Could directed ultrasound be used in conjunction with specific types of lasers as a nanomedical navigational tool? Radio frequencies in the 0.1-MHz range can traverse human tissues to a depth of 20 cm without being dissipated, and may consequently prove useful for accessing and even powering in vivo nanodevices [3]. The near infrared light spectrum resides in wavelengths that range from 800 to 2,500 nm, and in one study tissue penetration depths were obtained for the neonatal head that ranged from between 6.3 and 8.5 mm [78]. In a light therapy investigation by NASA, the combination of three optimal wavelengths for LEDs in the near-infrared reached depths of 23 cm through surface tissue and muscles [79]. Standard global positioning system (GPS) frequencies and associated wavelengths would most likely be inappropriate for in vivo nanomedical applications. To achieve the required resolution, a scaled down analogous system might utilize frequencies that are transmitted and received at much higher ranges, with wavelengths in the nanometer domain. One scenario for such a system might employ three or four beacons referenced to each other in a dedicated spatial metrics room (Figure 15.11). The self-referencing of the beacons might be accomplished when they transmit a particular signal that intersects at a specific point in space relative to the patient. This process might spatially demarcate with high precision a reference set point that is defined by the confluence of the beams from three or four beacons, which would be registered in the outbody computer system. The system would then lock on to and be calibrated to this single sustained reference point.
416
Potential Strategies for Advanced Nanomedical Device Ingress and Egress Scanning volume
Confluence fiducial Navigation beacon
Figure 15.11
Conceptual spatial metrics room.
All subsequent movements made by internalized nanomedical devices might be directed, tracked, and visualized in relation to this “confluence fiducial.” Each nanodevice would be tagged and tracked via embedded metallic nanoshells or other nanometric entity, each with its own unique identifying frequency signature. The outbody navigational computers would calculate the time lapse of signals sent from the external beacons to each nanodevice (based on the speed of light) to determine its real-time orientation in 3-D space (cross-referenced to the intersect node mentioned above) and velocity (perhaps using a fourth beacon). The system might be equipped with atomic clocks to ensure highly localized time/distance calculations. Positions of all nanodevices in 3-D space would be triangulated by the three beacons set against the established intersect node. Steering each individual nanodevice through the vasculature may well present very significant and complex challenges, as they would be required to be endowed with the capacity for traveling in multiple directions (perhaps following the axes of virtuallyall vascular entities) at prescribed velocities.
15.12
Nanodevice Egress Strategies Strategies having an equal importance to nanomedical device ingress would be the formulation of techniques for efficient egress out of the patient, post-treatment. One hypothetical (albeit quite protracted) method would be to instruct the nanodevices to migrate to the germinal matrix (behind the nail beds of the fingers and toes), where they would embed themselves in the matrix that forms the nail material (com-
15.13 Conclusion
417
prised of proteins, keratin, and sulphur) and to then permanently shut down. They would remain encapsulated there to eventually “grow” out of the body naturally with the nails, at a rate ~.05 to 1.2 mm/week. Another, more rapid method of egress might be to direct the devices to embed themselves within the matrix of forming hair follicles (comprised of keratin, trichohyalin granules, melanin) to subsequently exit with the hair at a growth rate of ~1 cm per month. The diameter of hair follicles range from ~17 to 181 μm, thereby providing a substantial mass of material in which a 1-μm nanomedical device may embed itself [80, 81]. Even more rapid egress scenarios (within ~24 hours) may include exiting the body naturally via migration to, and self-embedding within the biomass of the intestinal tract, or by natural diffusion or flushing from the system with bodily fluids (e.g., urine, sweat) after primary systems shut down. Yet another scenario might induce all nanodevices, subsequent to the completion of a procedure, to respond to and gravitate toward a homing signal that would emanate from a dedicated retrieval patch adhered to the skin. Upon arriving at the source site, the devices would diffuse, or burrow up through the epidermal layers and adsorb to a specifically designed patch undersurface for subsequent removal. Dependant on the level of individual nanodevice and infrastructure sophistication, this approach might take place within tens of minutes. Primary considerations for any of these egress options would be that all devices, first of all, are accounted for via all-clear protocols, and secondly, that the selected mode of egress should proceed in a completely discreet and innocuous manner. This process should be as unnoticeable as the low-level perspiration that is continuously emanating through the pores of the skin.
15.13
Conclusion It is hoped that this brief, and by no means comprehensive, survey of several potential propulsive, ambulatory, and navigational strategies that might be employed by future nanomedical devices may provide some modicum of what may be possible toward their development. The author has the humble and hopeful aim of serving to possibly inspire the design and development of a myriad of innovative and highly effective nanomedical tools. The careful, deliberate, and thoughtful creation of safe yet robust nanoscale instruments of health may have strong potential in leading to many positive health impacts for all of humankind.
Problems 15.1 Describe five potential techniques whereby nanomedical devices may perform ingress into the human body. 15.2 How might more sophisticated nanodevices gain access to the bloodstream? 15.3 What is one way by which nanodevices might circumvent the blood/brain barrier to diagnose or treat the brain?
418
Potential Strategies for Advanced Nanomedical Device Ingress and Egress
15.4 What is iontophoresis? 15.5 What are two significant physiological obstacles that nanodevices will have to overcome in order to effectively propel themselves within the aqueous in vivo environment of the human body? 15.6 Name several types of molecular mechanisms that might be integrated into future nanomedical devices. 15.7 What type of power stroke might enable nanodevices to traverse the viscous whole blood/plasma media of the human vasculature, and which type of power stroke will be ineffectual at the nanoscale? 15.8 Which class of nanomaterials might serve as biomimetic analogs of flagella or cilia for the potential propulsion of nanodevices through the human circulatory system?
References [1] MED-JET, Medical International Technology, http://www.mitcanada.ca/products/ med.html. [2] SonoPrep® Facilitated Transdermal Vaccination, Sontra Medical Corporation, 2004, http://sontra.icorpsdev.com/products/drugdelivery/, and http://sontra.icorpsdev.com/ products/vaccinedelivery/. [3] Robert A. Freitas Jr., Nanomedicine, Volume I, Basic Capabilities, Landes Bioscience, 1999. [4] K. Ishiyama, K.I. Arai, M. Sendoh, A. Yamazaki, “Spiral-type micro-machine for medical applications,” Journal of Micromechatronics, Volume 2, (1), 77-86, (10), 2002. [5] A. Nasir, “Nanotechnology and Dermatology,” Presentation at the American Academy of Dermatology, San Antonio, Texas, Feb. 2008. [6] P.R. Lockman, J.M. Koziara, R.J. Mumper, D.D. Allen, “Nanoparticle surface charges alter blood-brain barrier integrity and permeability,” J. Drug Target, 12, 635-641, 2004. [7] H.M. Mansour, A.J. Hickey, “Raman Characterization and Chemical Imaging of Biocolloidal Self-Assemblies, Drug Delivery Systems, and Pulmonary Inhalation Aerosols: A Review,” AAPS PharmSciTech, 8.4, E99: 1-16:140-155, 2007. [8] J. Kreuter, R. N. Alyautin, D. A. Kharkevich, and A. A. Ivanov, “Passage of peptides through the blood-brain barrier with colloiddal polymer particles (nanoparticles) ,” Brain Res., 674, l7l, 1995. [9] S. Talegaonkar, P.R. Mishra, “Intranasal delivery: An approach to bypass the blood brain barrier,” Indian Journal of Pharmacology, 36, 3, 140-147, 2004. [10] K. Ringe,C . M. Walz,B . A. Sabel, “Nanoparticle Drug Delivery to the Brain,” Encyclopedia of Nanoscience and Nanotechnology, Ed. Hari Singh Nalwa, American Scientific Publishers, 2004. [11] J. Kreuter, V E. Petrov, D. A. Kharkevich, and R. N. Alyautdin, “Influence of the type of surfactant on the analgesic effects induced by the peptide dalargin after its delivery across the blood–brain barrier using surfactant-coated nanoparticles,” J. Controlled Release, 49, 81-87, 1997. [12] U. Schroedear, B. A. Sabel, “Nanoparticles, a drug carrier system to pass the blood-brain barrier, permit central analgesic effects of i.v. dalargin injections,” Brain Res., 7l0, l2l–124, 1996. [13] U. Schroeder, P. Sommerfeld, B. A. Sabel, “Efficacy of Oral Dalargin-loaded Nanoparticle Delivery across the Blood–Brain Barrier,” Peptides, 19, 777-780, 1998.
References
419
[14] U. Schroeder, P. Sommerfeld, S. Ulrich, B. A. Sabel “Nanoparticle technology for delivery of drugs across the blood-brain barrier,” J. Pharm. Sci., 87, 1305-1307, 1998. [15] U. Schroeder, B. A. Sabel, H. Schroeder, “Body distribution of 3HH-labelled dalargin bound to poly(butyl cyanoacrylate) nanoparticles after I.V. injections to mice, ” Life Sci., 66,495-502, 2000. [16] J.P. Ryman-Rasmussen, J. E. Riviere, N. A. Monteiro-Riviere, “Penetration of Intact Skin by Quantum Dots with Diverse Physicochemical Properties, ” Toxicological Sciences, 91(1), 159-165, 2006. [17] D. V. McAllister, P. M. Wang, S. P. Davis, J-H. Park, P. J. Canatella, M. G. Allen, M. R. Prausunitz, “Microfabricated needles for transdermal delivery of macromolecules and nanoparticles: fabrication methods and transport studies,” Proc. Natl. Acad. Sci. USA, 100, 13755-13760, 2003. [18] T. Morrow, “Transdermal Patches are More Than Skin Deep,” Managed Care Magazine, Vol. 13, No. 4, April 2004, pp. 50–51, http://www.managedcaremag.com/archives/0404/ 0404.biotech.html. [19] H. Gray, Anatomy of the Human Body, 1918, from bartleby.com, http://www.bartleby.com/107/pages/page1070.html [20] E. R. Kay, D. A. Leigh, F. Zerbetto, “Synthetic Molecular Motors and Mechanical Machines,” Angew. Chem. Int. Ed, 46, 72-191, 2007. [21] P. A. Skordos, W. H. Zurek, Am. J. Phys., “Maxwell’s demon, rectifiers, and the second law: Computer simulation of Smoluchowski’s trapdoor,” 60, 876–882, 1992. [22] Z. Rappoport, S. E. Biali, “Threshold Rotational Mechanisms and Enantiomerization Barriers of Polyarylvinyl Propellers,” Acc. Chem. Res. 30, 307–314, 1997. [23] M. Oki, “Ungewöhnlich hohe Rotationsbarrieren am tetraedrischen Kohlenstoffatom,” Angew. Chem., 88, 67–74, 1976. [24] Y. Tabe, H. Yokoyama, “Coherent collective precession of molecular rotors with chiral propellers,” Nat. Mater. 2, 806–809, 2003. [25] J. P. Sestelo, T. R. Kelly, “A prototype of a rationally designed chemically powered Brownian motor,” Appl. Phys. A, 75, 337–343, 2002. [26] R. Brown, The Miscellaneous Botanical Works of Robert Brown, Vol. 1 (Ed.: J. J. Bennett), Ray Society, London, pp. 463–486, 1866. [27] M. Schliwa, Molecular Motors, Wiley-VCH, Weinheim, 2003. [28] R. D. Astumian, P. H. Xnggi, “Brownian Motors,” Phys. Today, 55 (11), 33-39, 2002. [29] G. Bottari, F. Dehez, D. A. Leigh, P. J. Nash, E.M. Perez, J. K.Y. Wong, F. Zerbetto, “Entropy-Driven Translational Isomerism: A Tristable Molecular Shuttle,” Angew. Chem., 2003, 115, 6066–6069; Angew. Chem. Int. Ed., 42, 5886-5889, 2003. [30] R. A. Bissell, E. Cordova, A. E. Kaifer, J . F. Stoddart, “A chemically and electrochemically switchable molecular shuttle,” Nature, 369, 133-137, 1994. [31] R. B. Gennis, Biomembranes—Molecular Structure and Function, Springer, New York, 1989. [32] P. Lauger, Electrogenic Ion Pumps, Sinauer Associates, Sunderland, MA, 1991. [33] C.R. Ethier, C.A. Simmons, Introductory Biomechanics, From Cells to Organisms, Cambridge University Press, 2007. [34] E.M. Purcell, “Life at low Reynolds number,” Am. J. Physics 45, 3-11, 1977. [35] Robert A. Freitas Jr., “How Nanorobots Can Avoid Phagocytosis by White Cells – Part I,” Foresight Update, 45, 10-12, 2001, http://www.imm.org/Reports/Rep027.php. [36] H.L. Goldsmith, S.G. Mason, “The flow of suspensions through tubes. III. Collisions of small uniform spheres,” Proc. R. Soc. A 282, 569-591, 1964. [37] J. Lighthill, “Flagellar hydrodynamics,” SIAM Review 18, 161-230, 1976. [38] J.J.L. Higdon, “The hydrodynamics of flagellar propulsion: Helical waves,” J. Fluid Mech. 94, 331-351, 1979.
420
Potential Strategies for Advanced Nanomedical Device Ingress and Egress [39] M. Ramia, D.L. Tullock, N. Phan-Thien, “The role of hydrodynamic interaction in the locomotion of microorganisms,” Biophys. J. 65, 755-778, 1993. [40] H. Winet, “Wall drag on free-moving ciliated micro-organisms,” J. Exp. Biol. 59, 753-766, 1973: [41] M.D. Levin, C.J. Morton-Firth, W.N. Abouhamad, R.B. Bourret, D. Bray, “Origins of individual swimming behavior in bacteria,” Biophys. J. 74, 175-181, 1998. [42] M. Ramia, D.L. Tullock, N. Phan-Thien, “The role of hydrodynamic interaction in the locomotion of microorganisms,” Biophys. J. 65, 755-778, 1993. [43] K.C. Chen, R.M. Ford, P.T. Cummings, “Mathematical models for motile bacterial transport in cylindrical tubes,” J. Theoret. Biol. 195, 481-504, 1998. [44] S.H. Lee, L.G. Leal, “Motion of a sphere in the presence of a plane interface. Part 2: An exact solution in bipolar coordinates,” J. Fluid Mech. 98, 193-24, 1980. [45] Robert Dillon, Lisa Fauci, Donald Gaver III, “A Microscale Model of Bacterial Swimming, Chemotaxis and Substrate Transport,” J. Theoret. Biology, 177, 325- 340, 1995. [46] S. Childress, Mechanics of Swimming and Flying, Cambridge University Press, Cambridge, UK, 1981. [47] James A. Wilson, Principles of Animal Physiology, Macmillan Publishing Company, New York, 1972. [48] H. Machemer, “Ciliary activity and the origin of metachrony in Paramecium: Effects of increased viscosity,” J. Exp. Biol., 57, 239-259, 1972. [49] S. Gueron, K. Levit-Gurevich, “Computation of the internal forces in cilia: application to ciliary motion, the effects of viscosity, and cilia interactions,” Biophys. J., 74, 1658-1676, 1998. [50] Michael E.J. Holwill, Peter Satir, “II.4 Generation of Propulsive Forces by Cilia and Flagella,” in J. Bereiter-Hahn, O.R. Anderson, W.-E. Reif, eds., Cytomechanics: The Mechanical Basis of Cell Form and Structure, Springer-Verlag, Berlin, pp. 120-130, 1987 [51] S.F. Goldstein, N.W. Charon, “Motility of the spirochete Leptospira,” Cell Motil. Cytoskeleton, 9, 101-110, 1988. [52] D.R Pitelka, C.N Schooley, “The fine structure of the flagellar apparatus in Trichonympha,” J. Morph., 102, 199-246, 1958. [53] D. Bray, Cell Movements, 2nd ed., Garland Publishing, New York, 259-261, 2001. [54] J. Gray, Ciliary Movement, Cambridge University Press, 1928. [55] M.A. Sleigh, The Biology of the Cilia and Flagella, Pergamon Press, 1962. [56] H. Lodish, et al, Molecular Cell Biology, Scientific American Books, 1998. [57] S. M. Block, “Fifty Ways to Love Your Lever: Myosin Motors,” Cell, 87, 151-157, 1996. [58] M. Shwartz, “Walking’ proteins need to rock and roll, study finds,” Stanford Report, July 11, 2001, http://news-service.stanford.edu/news/2001/july11/kinesin-711.html. [59] J. Edd, S. Payen, B. Rubinsky, M.L. Stoller, and M. Sitti, “Biomimetic Propulsion for a Swimming Surgical Micro-Robot,” IEEE/RSJ Intelligent Robotics and Systems Conference, Las Vegas, USA, Oct., 2003, http://www.me.cmu.edu/faculty1/sitti/nano/publications/iros03_last.PDF. [60] A. Najafi, R. Golestanian, “Simple Swimmer at Low Reynolds Number: Three Linked Spheres, ” Phys. Rev. E, 69, 062901, 2004. [61] S. Guo, Y. Hasegaw, T. Fukuda, and K. Asaka, “Fish-like underwater microrobot with multi DOF,” International Symposium on Micromechatronics and Human Science, Piscataway, NJ, USA, 2001. [62] S. Guo, T. Fukuda, and K. Asaka, “A new type of fish-like underwater microrobot,” IEEE/Asme Transactions on Mechatronics, 8, 136-141, 2003. [63] G. Laurent and E. Piat, “Efficiency of Swimming Microrobots using Polymer Metal Composite Actuators,” IEEE Robotics and Automation Conference, Seoul, Korea, 2001.
References
421
[64] J-B. Mathieu, S. Martel, L. H. Yahia, G. Soulez, G. Beaudoin, “Preliminary studies for using magnetic resonance imaging systems as a means of propulsion for Microrobots in blood vessels and evaluation of ferromagnetic artefacts,” CCGEI 2003, Montréal, May, 2003, http://www.nano.polymtl.ca/Articles/2003/MR-Sub%20CCECE%2003% 20v7.pdf. [65] T. Morita, M. K. Kurosawa, and T. Higuchi, “Cylindrical shaped micro ultrasonic motor utilizing PZT thin film (1.4 mm in diameter and 5.0 mm long stator transducer),” Sensors and Actuators, A:Physical: The 10th International Conference on Solid-State Sensors and Actuators TRANSDUCERS ‘99, Jun 7-Jun 10 1999, vol. 83, no.1. May, pp. 225-230, 2000. [66] O. Vyshnevskyy, S. Kovalev, J. Mehner, “Coupled tangential-axial resonant modes of piezoelectric hollow cylinders and their application in ultrasonic motors,” IEEE Trans Ultrason Feroelectr Freq Control, Jan, 52 (1), 31-6, 2005. [67] Z. Chang, Y. Bar-Cohen, “Piezopumps using no physically moving parts,” NDEAA Technologies, Jet Propulsion Laboratory, 1999, http://eis.jpl.nasa.gov/ndeaa/ndeaa-pub/ pumps/Piezopump-99-workshop.pdf. [68] Z. Chang, Y. Bar-Cohen, “Piezoelectrically Actuated Miniature Peristaltic Pump,” Proceedings of SPIE’s 7th Annual International Symposium on Smart Structures and Materials, Newport, 1-5 March, 2000.CA. Paper No. 3992-103 SPIE, http:// eis.jpl.nasa.gov/ndeaa/ndeaa-pub/SPIE-2000/paper-3992-102-Piezopump.pdf. [69] R. Liang et al, “A novel piezo vibration platform for probe dynamic performance calibration,” Meas. Sci. Technol., 12, 1509-1514, 2001, http://www.iop.org/EJ/abstract/ 0957-0233/12/9/318. [70] W. Xudong, S. Jinhui, L. Jin, L.W. Zhong, “Direct-Current Nanogenerator Driven by Ultrasonic Waves,” Science, 316 (5821), 102-105, 2007. [71] P. Mela, N. R. Tas, J. E. ten Elshof, A. van den Berg, “Nanofluidics”, Encyclopedia of Nanoscience and Nanotechnology, American Scientific Publishers, 2004. [72] R. Qiao and N.R. Aluru, “Transient Analysis of Electroosmotic Flow in Nano-diameter Channels,” Beckman Institute for Advanced Science and Technology, http://www.cr.org/publications/MSM2002/pdf/196.pdf. [73] J. Alam, J.C. Bowman, “Energy-Conserving Simulation of Incompressible Electro-Osmotic and Pressure-Driven Flow,” Theoret. Comput. Fluid Dynamics, 16,133–150, 2002. [74] R.A. Freitas, “Nanodentistry”, J Am Dent Assoc, 131, 1559-1565, 2000, http://jada.ada.org/cgi/reprint/131/11/1559.pdf. [75] W.B. Sherman, N.C. Seeman, “A Precisely Controlled DNA Biped Walking Device,” Nano Lett., 4 (9), 1801-1801, 2004. [76] K.-Y. Kwon, K. L. Wong, G. Pawin, L. Bartels, S. Stolbov, T. S. Rahman, “Unidirectional Adsorbate Motion on a High-Symmetry Surface: ‘Walking’ Molecules Can Stay the Course,” PRL 95, 166101, 2005 [77] K.L. Wong, G. Pawin, K.-Y. Kwon, X. Lin, T. Jiao, U. Solanki, R.H.J. Fawcett, L. Bartels, S. Stolbov, T.S. Rahman, “A Molecule Carrier,” Science, 315, 1391, 2007. [78] F. Faris, et al., “Non-invasive in vivo near-infrared optical measurement of the penetration depth in the neonatal head,” Clin. Phys. Physiol. Meas., 12, 353-358, 1991. [79] Chance B., Nioka S., Kent J., McCully K., Fountain M., Greenfield R., Holtom G., “Time-Resolved Spectroscopy of Hemoglobin and Myoglobin in Resting and Ischemic Muscle,” Analytical Biochemistry, 174, 698-707, 1988. [80] J. Lademanna, H. Richtera, U.F. Schaeferb, U. Blume-Peytavia, A. Teichmanna, N. Otberga, W. Sterrya, “Hair Follicles - A Long-Term Reservoir for Drug Delivery,” Skin Pharmacol. Physiol., 19(4), 232-236, 2006. [81] R. Alvarez-Román, A. Naik, Y.N. Kalia, R.H. Guy, H. Fessi, “Skin penetration and distribution of polymeric nanoparticles,” Journal of Controlled Release, 99, 53-62, 2004.
CHAPTER 16
Nanoscale Mechanics for Medicine J. Weldon, T. Yuzvinsky, and A. Zettl
16.1
Introduction New nanomanipulation tools and new nanoscale materials have facilitated important advances in nanomechanical systems with relevance to biology and medicine. This chapter discusses the use of carbon nanotubes for applications including nanoinjectors, biocompatible structural materials, nanomotors, nanoscale radio transmission and reception, and compact, atomic-resolution mass sensors. The unique mechanical characteristics and chemical makeup of carbon nanotubes suggest diverse biomedical applications. For example, the high stiffness and large aspect ratio of nanotubes could be useful for injection needles, sensing probes, or general molecule manipulation arms. Nanotube-based composites might be utilized in bone replacement materials or heart valves.
16.2
Nanoinjection and Nanotube Biocompatibility 16.2.1
Nanoinjection
Intelligent drug design promises to revolutionize pharmaceutical research, but it requires extensive knowledge of the biochemical interactions that govern cellular behavior. Technologies for introducing molecules into living cells are vital for probing this molecular basis of disease, but conventional methods can cause collateral damage that may confound experimental results. The major challenge is to overcome the barrier imposed by the plasma membrane, which has been accomplished in a variety of ways, including membrane permeabilization (with lipids, electric currents, or pore-forming toxins) and physical penetration of the membrane (microinjection or microprojectile bombardment). Each method has its advantages and disadvantages, but one common liability is physical damage to the cell membrane. An alternative method of intracellular delivery combines the microinjection concept with emerging tools from nanotechnology. Figure 16.1 shows in schematic form a “nanoinjection” system, whereby biologically-relevant cargo is delivered to a specific location in a living cell [1]. In this particular implementation, the cargo is first fastened to the exterior of the nanotube by means of tailored functionalization and linker bonding. The tip of the nanotube with cargo is then pushed through the cell membrane, where the chemical environment within the cell cleaves the bonds linking the cargo to the nanotube and the cargo is released within the cell. The
423
424
Nanoscale Mechanics for Medicine Functionalization
Insertion
Release
Microtip Nanotube Living cell
Functional group
Figure 16.1 (Color plate 26) Nanoinjection technique for transmembrane transport of cargo. (Image courtesy C.R. Bertozzi group.)
nanotube is then withdrawn. The small diameter of the nanotube insures that the cell membrane is not irreparably damaged upon nanoinjection, and the fact that no additional fluid is injected to the cell interior (in contrast to microinjection) eliminates the chance of cellular disruption through buildup of excessive internal fluid pressure. Although carbon nanotubes are hollow, the hollow core is not here used as a transport channel. The key to effective nanoinjection is the ability to control the cargo attachment and release chemistry, and to reliably target and mechanically puncture a specific cell. Several approaches have been used to address these challenges. An atomic force microscope (AFM), as shown in Figure 16.1, with suitable optics and biocompatible sample carriers (i.e., compatible with aqueous solutions) allows precise cell identification and provides a means to controllably pierce a selected cell membrane with the functionalized nanotube. Indeed, the motion control and force feedback signal of the AFM tip allows the operator to electronically “tell” when the nanotube is just touching the cell membrane, when it has pierced it, and how far it has been inserted into the cell. Three-dimensional positioning within the cell is thus excellent. Nanoinjection experiments have released various types of cargo into living cells, including protein-coated quantum dots. Using fluorescence methods, thus-inserted quantum dots have been tracked optically in real time, and from their motion within the cell, the diffusion dynamics in the cytosol have been extracted. One effective method for cargo attachment and release employs the chemistry of Figure 16.2. A chain compound having a pyrene moiety foot binds strongly to the nanotube surface via p-p stacking. Partway down the chain is a disulfide linker bond. This linker bond remains intact in aqueous environments exterior to the cell, but it is readily cleaved in the oxidizing environment within the cell. The head of the chain molecule is endowed with a biotin moiety. Biotin has a strong affinity for streptavidin, which allows for secure attachment of selected cargo such as streptavidin-coated nanocrystals. 16.2.2
Nanotube Biocompatibility
The nanoinjection process exploits intentional coating or functionalization of the nanotube. Many different functionalization schemes for different nanotube applica-
16.2 Nanoinjection and Nanotube Biocompatibility
425
O O N H
S
S
HN NH H H
H N
=
SS
S O
Figure 16.2 Linker for attaching cargo to a carbon nanotube. The foot of the linker molecule has a pyrene moiety (square shape on right schematic) that attaches to the nanotube, while the head of the molecule has a biotin moiety (arrow shape on right schematic) that attaches to the cargo. The disulfide bond (ss on right schematic) can be cleaved to release the cargo. (Adapted from [1].)
tions have been proposed. For example, nanotubes coated with fullerenes may serve as critical components in organic solar cells. In the biological/medical arena, carbon nanotubes are a mixed blessing. In some studies nanotubes have been used as templates on which living organisms are readily grown; in other studies nanotubes have been found to be extremely toxic to living cells. Although the precise mechanism of cell poisoning is not yet understood, and in some cases the toxicity may have resulted not from the nanotubes themselves but from residual transition metal catalyst particles, it is clear that the interaction between nanotube and cell can be dramatically altered by nanotube functionalization. Indeed, suitably functionalized carbon nanotubes are quite biocompatible. A particularly attractive nanotube biofunctionalization scheme utilizes carbohydrate-functionalized polymers designed to mimic the structures of mucin glycopolymers [2]. Natural mucin glycopolymers protect and lubricate cells. Mucin mimics have been developed [3] that display similar functionality but are easier to manipulate than their natural relatives. The polymers comprise a poly(methyl vinyl ketone) backbone decorated with α-N-acetylgalactosamine (α-GalNAc) residues; the residues are similar to the O-linked lycans that decorate mucin glycoprotiens. The H3C-(CH2)17 lipid tail of the polymers provides a hydrophobic anchor for attachment to the nanotube wall. Figure 16.3 shows a schematic representation of the polymers attached to a carbon nanotube. The functionalized nanotubes are soluble in water, and can be readily interfaced to other biologically relevant structures, such as cells, via further chemistry including carbohydrate-receptor binding [4]. For example, the polymer coated nanotubes can be first bound to HPA, a hexavalent α-GalNAc binding lectin; thereafter the complex can be bound to cell surface glycoconjugates using available HPA binding sites presented on the nanotubes. An alternate cell binding method involves first attaching HPA to the cell surface glycoconjugates, whereupon the HPA binding sites on the cell surface are then bound to α-GalNAc residues on the polymer coated nanotubes. Importantly, glycodendrimer-coating of carbon nanotubes mitigates their cytotoxicity. Living cells including HEK293 cells reproduce normally in the presence of such functionalized nanotubes, whereas in the presence of unfunctionalized carbon nanotubes severe cytotoxicity is encountered. The synthetic methods used to construct glycodendrimer-functionalized nanotubes can be directly adapted to ligands for other receptor interactions. It is also worthwhile noting that boron nitride nanotubes, structures with geometrical and mechanical properties similar to those of carbon nanotubes, appear nontoxic to living cells even in their native,
426
Nanoscale Mechanics for Medicine
Figure 16.3 Schematic of carbon nanotube functionalized with biocompatible molecules, such as glycodendrimers. (Adapted from [2].)
unfunctionalized state [5]. Boron nitride nanotubes can be functionalized to enhance their application potential [6].
16.3
Nanometer Propulsion While the high strength, small size, and extreme sharpness of carbon nanotubes make them the ideal passive scaffold for nanoinjection applications, their extraordinary mechanical properties also open the door for their use as active elements in nanoscale biomedical devices. We here discuss several nanoelectromechanical motors based on carbon nanotubes that could in principle be used to provide mechanical forces for tissue penetration, manipulation, and propulsion for in vivo biomedical probes. 16.3.1
Rotational Nanomotors
As demonstrated by bacterial flagella [7], the availability of rotational drive about an axis can be efficiently harnessed to provide controlled propulsion in liquid environments. At its base, each flagellum contains a protein-based motor powered by protomotive force that drives the flagellar filament in a circular motion, propelling the bacterium in the direction away from the rotating flagellum. While this protein-based motor does not lend itself to easy incorporation into a synthetic device, a lithographically fabricated nanotube motor would be easy to interface without concern for protein stability or denaturation during the assembly process. An attractive synthetic candidate is the rotational nanotube motor [8] pictured in Figure 16.4. The key element in this device is the multiwalled carbon nanotube
16.3 Nanometer Propulsion
427
Figure 16.4 Rotational nanotube motor schematic. A gold rotor (R) is attached to and supported by the outer wall of a multiwalled carbon nanotube that has been suspended between two anchors (A1, A2). The rotor can be rotated by the sequential application of voltages to the three stators (S1, S2, S3). The in-plane stators S1 and S2 are fabricated in the same step as the anchors and rotors; stator S3 is the degenerately doped back-gate beneath the oxide surface. (Adapted from [8].)
(MWCNT) positioned along its center, which acts as both a support and axle for the attached gold rotor to spin about. The concentric, atomically smooth walls of the MWCNT comprise an ultralow-friction bearing for the motor [9–11], increasing the amount of force delivered and providing a very long lifetime for the device. Rotational nanotube motors can be fabricated by lithographic means in any standard silicon processing facility. The primary requirement necessary for successful operation is the use of high-quality MWCNTs with no interwall defects, such as those grown by arc-discharge (other synthesis methods usually yield more defective multiwalled carbon nanotubes). These MWCNTs are deposited out of isopropyl alcohol onto a silicon wafer coated with 1 micron of thermally grown silicon oxide. The surface is then covered with resist and patterned by electron-beam lithography, opening windows above either end of the nanotube, with a smaller window in the center. Evaporation of ~100 nm of gold (with a thin chromium sticking layer) and subsequent liftoff yields a nanotube pinned to the silicon oxide surface by two anchors, with a smaller, centrally positioned rotor attached to the outer wall. Next, a brief hydrofluoric acid etch completely undercuts the rotor (but not the anchors), releasing the rotor from the surface. Finally, the outer wall of the MWCNT must be removed or severed on either side of the rotor to allow free rotation. While electrical breakdown [12] and reactive ion etching [13] can both be used, simple torsion of the device is the easiest method to sever the outer wall. Once free to rotate, the rotor can be driven by alternately applying voltages to the three surrounding stators. Capacitive attraction pulls the rotor towards each stator in turn, rotating it about the MWCNT axis, as shown in Figure 16.5. Even after thousands of iterations no wear is observed at the atomic scale, suggesting that the motor can be driven indefinitely without damage. 16.3.2
Linear Nanomotors
Linear motors or actuators can be made with even smaller dimensions and with less complexity than for rotational motors, and as such can thus provide extremely high
428
Nanoscale Mechanics for Medicine
(a)
0°
(e)
1800°
(b)
45°
(f)
205°
(c)
(d)
90°
14°
(g)
(h)
270°
340°
Figure 16.5 (a–h) Rotational nanotube motor operation. Scanning electron micrographs of the rotational nanotube motor moving through one full rotation. The motor was paused in each position to accommodate image acquisition; it can also be driven at high speed. The scale bar is 300 nm. (Adapted from [8].)
power densities for ultrasmall probes. For example, the solid-state nanocrystalpowered nanomotor shown in Figure 16.6 [14] has an available power density of ~8 GW/m3, far exceeding the power density of either biological systems or macroscopic internal combustion engines (see Table 16.1). The nanocrystal motor operates by electrically shuttling individual indium atoms from a reservoir to a nanocrystal (the indium atoms are driven by an electrical current-induced chemical potential gradient). As the nanocrystal grows, it acts as a ram, pushing apart the two supporting arms. The process is reversible: by reversing the current and shuttling atoms in the opposite direction, the nanocrystal shrinks, allowing the arms to release and move closer together. An important feature of the nanocrystal motor is its ability to easily lock in place at any position: once the applied current is turned off, the nanocrystal cannot grow or shrink; it is truly “set and forget.” Another feature is the available variation in size; for if a suitably large reservoir is available, the nanocrystal can in principle be grown to any size, potentially serving as part of a stent assembly to hold open constricted blood vessels. Finally, since the driving element literally regrows during each actuation, operational damage and wear are not a concern as they are automatically repaired during use.
16.3 Nanometer Propulsion
(a)
429
(b)
(c)
(d)
Figure 16.6 Nanocrystal-powered nanomotor. (a) Schematic: indium atoms are shuttled between a reservoir (not shown) and a nanocrystal positioned between two arms. Atoms added to the crystal cause it to grow, pushing apart the arms. (b–d) Transmission electron micrographs of the nanomotor in action. The atom reservoir is the dominant dark shape in (b). In (d), the nanocrystal ram is over 100 nm long, and the atom reservoir is approximately half depleted. The scale bar is 100 nm. (Adapted from [14]).
Table 16.1 Order of Magnitude Comparison of Maximum Output Force and Available Power Density Among Different Motors
Maximum force Power density
f1-ATPase Biomotors
BMW 740i Internal Com- Nanocrystal-Powered bustion Engine Nanomotor
50 pN 50 MW/m3
105 N 3 50 MW/m
10 nN 3 10 GW/m
Interestingly, the nanoscopic biomotor and the refined macroscopic internal combustion engine have comparable power densities; the nanocrystal-powered motor has a power density 200 times larger.
16.3.3
Surface-Tension-Driven Nanomotors
The rotational and linear actuators described above require an alternating current source to power oscillatory motion. While such power sources are widely available, a direct current source (i.e., a battery or electrochemical fuel cell) may be much easier to incorporate into a small-probe geometry. Such a power source could power the surface-tension-driven nanoelectromechanical relaxation oscillator [7] shown in Figure 16.7 [15]. Equally important, surface tension is a very powerful force that scales favorably into the nanometer regime. The relaxation oscillator motor operates in similar manner to the nanocrystal-powered motor above, but instead of growing a crystal, indium atoms merge into a liquid droplet of indium (indium has a relatively low melting point, so the droplets are not very hot). As more indium is added to the droplet, it grows in diameter and eventually touches the liquid indium reservoir. At that moment, indium is free to flow between the droplet and the reservoir via a new, high-throughput hydrodynamic channel, and mass flows to minimize overall surface tension energy. The energy is reduced by an immediate reduction in the size of the droplet, the atoms of which are largely absorbed by the reservoir. If electrical
430
Nanoscale Mechanics for Medicine
Resevoir Droplet
MWNT
Figure 16.7 Surface-tension-driven nanoelectromechanical relaxation oscillator. Indium atoms are shuttled along a multiwalled nanotube (MWNT) from liquid atom reservoir to liquid droplet (black horizontal arrow), causing the droplet to grow. Once the droplet touches the reservoir, surface tension energy minimization induces a large transport of indium from droplet to reservoir (grey arrow), resetting the system and beginning a new cycle of the oscillation. (Adapted from [15].)
current is continually applied, the process repeats itself and the motor exhibits oscillatory motion, which can be coupled to moveable lever arms for linear motion. Just like rotational motion, oscillatory linear motion can be harnessed for propulsion, as in the case of cilia or eukaryotic flagella, which whip back and forth to propel cells through liquid.
16.4
Nanomechanical Radios and Sensors The ability to transfer information is fundamental to most electrical or biological systems. Wires are often used to send information electronically and within an organism a variety of electrical and chemical methods are used. One of the challenges associated with studying biological systems is transferring information both into and out of the system. An especially promising technique is through radio waves, and at appropriate power levels these signals are not harmful to biological systems. However, radio transceivers are typically too large and require significant power for operation. Amazingly, fully functional radio receivers and transmitters can be constructed from only a single carbon nanotube. The same geometry can be implemented for single-atom mass detection. 16.4.1
Nanotube Radio Receiver
Electrically, carbon nanotubes are capable of both metallic and semiconducting behavior and they can have a very high current density. The mechanical behavior includes very high elastic modulus, high tensile strength, and sharp mechanical resonance peaks. The nanotube radio design [16] leverages both the electrical and mechanical properties to yield a fully functional radio system many orders of magnitude smaller than any previous radio designs (most based on silicon-chip technology).
16.4 Nanomechanical Radios and Sensors
431
There are four critical parts vital to any radio receiver: the antenna, tuner, amplifier, and demodulator, as shown in Figure 16.8(a). The nanotube radio implements the four elements in a method very different compared to a conventional radio receiver. The antenna is needed to convert the radio signal into an electrical signal. The tuner then chooses a signal at a specific frequency and the amplifier increases the amplitude of this signal, which is typically very small. Finally, the demodulator extracts the desired information from the radio signal. Conventional radio receivers operate only with electrical signals but the nanotube radio uses both the electrical and mechanical behavior of nanotubes to operate. The nanotube is arranged, in vacuum, in a single-clamped, cantilever configuration as shown in Figure 16.8(b). The nanotube has a natural mechanical resonance that is determined by the shape of the nanotube and the applied DC voltage (which tensions the nanotube axially). The DC bias voltage, applied between the nanotube and the counter electrode, causes negative charge to accumulate at the tip of the nanotube. When the frequency of an incoming radio wave matches the natural resonant frequency, the nanotube vibrates mechanically because of the forces applied to the electrons in the tip of the nanotube. In this way, the nanotube is acting like a mechanical antenna: it converts the radio signal to a mechanical vibration. Furthermore, the resonant frequency of the nanotube can be externally controlled through the DC bias. As a result, it can filter only the desired signal, which implements the tuner function.
Figure 16.8 Nanotube radio. (a) A radio requires the functionality of four components: antenna, tuner, amplifier, and demodulator. The nanotube radio performs the operation of these components with a single nanotube. (b) Schematic of the nanotube radio. Radio frequency electromagnetic waves cause the charged tip of the nanotube to vibrate. Field emission from the tip the counter electrode is used to demodulate and amplify the modulated transmission.
432
Nanoscale Mechanics for Medicine
The transduction from mechanical antenna motion, and amplifier and demodulation functions, rely on field emission from the tip of the nanotube. Field emission occurs because the DC bias causes large electrical field enhancement at the tip of the nanotube, due to the incredibly sharp tip. The field emission current in nominally constant but it is very dependent on the position of the nanotube. The relationship between the position of the nanotube and the field emission current is nonlinear and therefore mechanical vibrations of the nanotube can demodulate the radio signal. This demodulated signal is powered by an external DC source and not the incoming radio signal. As a result, in addition to demodulation, the signal is also amplified. Nanotube radio operation has been confirmed by a number of methods. Due to the mechanical motion of the nanotube, a high-resolution transmission electron microscope (TEM) can be used to visualize the nanotube during operation. A nanotube is mounted to a wire and a modulated RF signal is applied with an external antenna. When the frequency of the transmitted RF signal matches the resonant frequency of the nanotube, it vibrates and blurs the image observed in the TEM. The top of Figure 16.9 is a TEM image of the nanotube when the frequency of the transmitted RF signal does not match the resonant frequency of the nanotube. The lower portion of Figure 16.9 shows the same nanotube when the frequencies match. The image is blurred due to the vibration of the nanotube; in this state the radio is locked onto a transmitting radio station. Another important function performed by a radio receiver is tuning. The nanotube radio can be fine-tuned via bias voltage tensioning, as described above. Importantly, the general operation band of the radio can be burned in using a postproduction coarse tuning approach. This is accomplished by irreversibly shortening the length of the nanotube, which in turn increases the resonant frequency. The shortening is performed by a controlled field emission current that is much higher than that typically used for radio operation. This high current removes carbon atoms from the nanotube tip. Coarse tuning of a typical nanotube is shown in the upper section of Figure 16.10 where the resonant frequency has been reduced from 350 MHz to approximately 100 MHz. Figure 16.10(b) shows typical reversible fine-tuning achieved via bias voltage tensioning. The nanotube radio conveniently operates in the commercial radio frequency band (kHz to hundreds of MHz). It has been tested using a transmitted RF signal
Figure 16.9 TEM images of a nanotube constituting a nanoscale radio receiver. The nanotube is shown when the transmitted signal matches the resonant frequency (bottom) and when it is unmatched (top). (Adapted from [16].)
16.4 Nanomechanical Radios and Sensors
433
(a)
(b)
Figure 16.10 Two frequency-tuning methods for a nanoradio receiver. (a) Coarse tuning. The nanotube is irreversibly shortened through field emission atom ejection, raising the resonant frequency. (b) Fine tuning. The resonant frequency is reversibly controlled through tension on the nanotube by varying the DC voltage on the nanotube.
modulated with an audio recording. The demodulated signal is easily identified by ear and the waveforms match the transmitted signal. 16.4.2
Nanotube Radio Transmitter
The nanotube radio demonstrates a functional radio receiver that is orders of magnitude smaller than conventional radios. The size allows for applications in biological systems and other environments that were previously not possible. However, in addition to receiving information, many biomedical applications would benefit from the ability to transmit as well. The concept of a radio transmitter based on a similar field-emitting nanotube has been proposed [17]. To enable a nanometer scale transmitter require RF oscillations that are not driven by a large external source. It has been demonstrated that during field-emission, a single nanowire will mechanically resonate in the proper configuration [18]. This same concept can be
434
Nanoscale Mechanics for Medicine
applied to nanotubes. However, a transmitter requires the ability modulate the RF signal. Figure 16.11 shows a schematic of the proposed nanotube transmitter in which the magnitude or frequency of the self-oscillations can be modulated by an external control voltage. In this configuration a DC bias applied to the nanotube draws electrons to the tip. Field enhancement at the tip induces field emission to the counter electrode, which in turn can cause the nanotube to vibrate at the natural resonant frequency. The vibrating charges in the tip radiate an RF electromagnetic wave that can be frequency modulated by controlling the tension, and therefore the resonant frequency, on the nanotube. The tension on the nanotube is controlled by a third electrode. Simulations of the vibrating nanotube with excess charge drawn to the tip are shown in Figure 16.12. 16.4.3
Nanomechanical Mass Sensing
There are numerous medical and biological applications that require the ability to sense and measure small mass units in an accurate, reliable way. Ideally this measurement would have single-molecule resolution without damaging a potentially fragile sample. The singly clamped nanotube cantilever device, used in both the nanotube radio receiver and transmitter, can also be utilized for sensing [19]. This nanomechanical resonator approach has been used to successfully measure the mass of individual gold atoms. The method could be adapted to biological molecular samples. The nanotube resonator operates as a mass sensor because changes in the mass of a resonator shift the resonant frequency. Therefore, by measuring the resonant frequency, the mass of adsorbed materials can be determined. A schematic of the device is shown in Figure 16.13, along with TEM images of an actual nanotube mass sensor element. A difficult problem with most nanoscale resonators designs is in the determination of the resonant frequency. For the nanotube-based mass spectrometer, this is accomplished by effectively using the nanotube as radio receiver, as Eradsin(wct+F (t))
IFE
Vbias
Vtension
Figure 16.11 Nanotube radio transmitter schematic. Electron field emission from the tip causes oscillations in the nanotube, which results in a radiated RF signal. A third electrode modulates the tension on the nanotube, which effectively modulates the transmitted signal.
16.4 Nanomechanical Radios and Sensors
435
Figure 16.12 (Color plate 27) Simulation of the electric field of a vibrating, field-emitting nanotube. (a) Field enhancement is evident in the tip of the nanotube and is at a maximum when the tube is straight and (b) field enhancement is reduced as the tube bends.
(a)
(b)
Figure 16.13 Nanomechanical mass spectrometer schematics and images. (a) TEM images of double-walled carbon nanotube used as the resonant element in the mass spectrometer. (b) To detect the changes in frequency of the nanomechanical element in the mass spectrometer, the nanotube is configured like a nanotube radio. The additional equipment is used to detect the changes in frequency of the vibrating nanotube. (Adapted from [19].)
described earlier. A high-frequency, modulate signal is transmitted to the nanotube and the demodulated signal is used to easily determine the resonance frequency.
436
Nanoscale Mechanics for Medicine
Due to the small size and high resonant frequency of the nanotube resonator, this sensor is able to detect single atoms adsorbing on the surface of the nanotube. Traditional mass spectrometers require the ionization of samples, which is potentially destructive, especially for sensitive biological specimens. The nanomechanical mass spectrometer does not require this ionization and thus is well suited for ionization-sensitive samples. Surface functionalization of the nanotube to select specific atoms or molecules could also be done.
Problems 16.1 Wireless communication to and from sensors and robots within the body is needed for many applications of nanodevices. Model a CNT antenna operating inside the body and estimate the size of the antenna and frequency range needed. 16.2 How can a larger antenna be made from short CNT? 16.3 What is nano about the nano-radio? 16.4 Can functionalization be used with the nanomechanical mass balances? 16.5 What are the opportunities for using carbon nanotubes electronics in fabricating devices?
Acknowledgments The research described in this chapter was performed together with A. Fennimore, B.C. Regan, C. Bertozzi, X. Chen, K. Jensen, A. Kis, K. Kim, A. Kis, T. Sainsbury, H. Garcia, and K. Erikson. Support from the U.S. DOE and NSF are gratefully acknowledged.
References [1] X. Chen, A. Kis, A. Zettl, C.R. Bertozzi, “A cell nanoinjector based on carbon nanotubes,” PNAS, 104, 8218–8222 (2007). [2] X. Chen, U. C. Tam, J. L. Czlapinski, G. S. Lee, D. Rabuka, A. Zettl and C. R. Bertozzi, “Interfacing Carbon Nanotubes with Living Cells,” J. Am. Chem. Soc,. 128, 6292–6293 (2006). [3] X. Chen, G. S. Lee, A. Zettl, and C.R. Bertozzi, “Biomimetic engineering of carbon nanotubes by using cell surface mucim mimics,” Angewandte Chemie, Int. Ed. 43, 6111–6116 (2004); G.S. Lee, Y. Shin, I. Choi, H. Hann, and C.R. Bertozzi (unpublished). [4] P. Wu, X. Chen, N. Hu, U.C. Tam, O. Blixt, A. Zettl, and C.R. Bertozzi, “Biocompatible carbon nanotubes generated by functionalization with glycodendrimers,” Angewandte Chemie-International Edition, 47 (27), 5022–5025 (2008). [5] X. Chen, C.R. Bertozzi, and A. Zettl (unpublished). [6] T. Sainsbury, T. Ikuno, D. Okawa, D. Pacil, J.M.J. Frechet, and A. Zettl, “Self-Assembly of Gold Nanoparticles at the Surface of Amine- and Thiol-Functionalized Born Nitride Nanotubes,” J. Phys. Chem., C 111, 12992–12999, (2007). [7] G. Meister and H. Berg, “Rapid rotation of flagellar bundles in swimming bacteria,” Nature, 325, 637–640 (1987).
References
437
[8] A.M. Fennimore, T.D. Yuzvinsky, Wei-Qiang Han, M.S. Fuhrer, J. Cumings and A. Zettl, “Rotational actuators based on carbon nanotubes,” Nature, 424, 408–410 (2003). [9] Charlier, J.-C. & Michenaud, J.-P., “Energetics of multilayered carbon tubules,” Phys. Rev. Lett., 70, 1858–1861 (1993). [10] Kolmogorov, A. N. & Crespi, V. H., “Smoothest bearings: Interlayer sliding in multiwalled carbon nanotubes,” Phys. Rev. Lett., 85, 4727–4730 (2000). [11] A. Kis, K. Jensen, S. Aloni, W. Mickelson, A. Zettl, “Interlayer Forces and Ultralow Sliding Friction in Multiwalled Carbon Nanotubes,” Phys. Rev. Lett., 97, 025501 (2006). [12] T.D. Yuzvinsky, W. Mickelson, S. Aloni, S.L. Konsek, A.M. Fennimore, G.E. Begtrup, A. Kis, B.C. Regan, and A. Zettl, “Imaging the life story of nanotube devices,” Appl. Phys. Lett., 87, 083103 (2005). [13] T.D. Yuzvinsky, A.M. Fennimore and A. Zettl, “Engineering Nanomotor Components from Multi-Walled Carbon Nanotubes via Reactive Ion Etching In Electronic Properties of Synthetic Nanostructures,” Kuzmany, Fink, Mehring, Roth eds. AIP Conference Proceedings 723, 512–515 (2004). [14] B.C. Regan, S. Aloni, K. Jensen, R.O. Ritchie and A. Zettl, “Nanocrystal-Powered Nanomotor,” Nano Letters, 5, 1730–1733 (2005). [15] B.C. Regan, S. Aloni, K. Jensen, and A. Zettl, “Surface-tension-driven nanoelectromechanical relaxation oscillator,” Appl. Phys. Lett., 86, 123119 (2005). [16] K. Jensen, J. Weldon, H. Garcia, and A. Zettl, “Nanotube Radio,” Nano Letters, 7 (11), 3508–3511 (2007). [17] J. Weldon, K. Jensen, and A. Zettl, “Nanomechanical radio transmitter,” Phys. Stat. Sol. (b), 245 (10), 2323–2325 (2008). [18] A. Ayari, P. Vincent, S. Perisanu, M. Choueib, V. Gouttenoire, M. Bechelany, D. Cornu, and S.T. Purcell, “Self-oscillations in field emission nanowire mechanical resonators: a nanometric dc-ac conversion,” Nano Letters, 7 (8), 2252–22577 (2007). [19] K. Jensen, K. Kim, and A. Zettl, “An atomic-resolution nanomechanical mass sensor,” Nature Nanotechnology, 3, (9) 533–537 (2008)
PART IV
Biological Integration and Characterization
CHAPTER 17
Integration of Manmade Nanostructures with Biological Systems: Diagnosis of Cancer Using Semiconductor Quantum-Dot Biomolecule Complexes Michael A. Stroscio and Mitra Dutta
17.1
Introduction During the last decade, there has been growing diversity in the types of semiconductor quantum-dot biomolecule complexes used in the diagnosis of disease. This chapter provides information on the types and properties of semiconductor quantum dots used in such applications and highlights a selection of applications of semiconductor quantum-dot biomolecule complexes in the study of carcinogenic cells as well as in the diagnosis of cancer. The rapid rate of increase in the use and the number of applications of semiconductor quantum dots has been facilitated by the increasing types and the quality of available semiconductor quantum dots. One of the key attributes of semiconductor quantum dots that facilitates disease diagnosis is the emission of light from these quantum dots at well defined frequencies upon illumination of the quantum dots with light with energy greater than or roughly equal to the bandgap of each particular type of quantum dot. In some respects, these semiconductor quantum dots are similar to fluorescent dyes that have long been used by the biological and medical communities to study biological phenomena including diagnosis of diseases. In other respects, such as the brightness, isotopic light emission capability, controllable optical emission wavelength, more flexible illumination options, and the long lifetimes of quantum dots, semiconductor quantum dots are superior to the traditional fluorescent dyes. Moreover, the improved synthesis of these semiconductor quantum dots has resulted in more predictable and controllable optical emission properties of quantum dots as well as a broader range of optical properties. Indeed, emission linewidths of high-quality quantum dots are about 5% or less, and the available emission wavelengths of tested quantum dots now span parts of the ultraviolet and infrared spectra and all of the visible spectrum. This chapter surveys information on the properties of the principal types of semiconductor quantum dots that have been used in biological studies, including disease diagnosis, and describes the integration of these semiconductor quantum dots with biomolecules through the processing of binding biomolecules to the quantum dots. This process of binding biomolecules to quantum dots is referred to as functionalizing the quantum dots with biomolecules, similar to the binding of fluo-
441
442
Integration of Manmade Nanostructures with Biological Systems
rescent dyes to biomolecules. As will be discussed, the advent of a growing variety of such semiconductor quantum dot biomolecule complexes has been motivated, in many cases, by the potential of such complexes to provide efficient means of diagnosing diseases. Accordingly, this chapter provides information on the types and properties of semiconductor quantum dots used in such applications and highlights a selection of applications of semiconductor quantum-dot biomolecule complexes in the diagnosis of disease. After the technology underlying semiconductor quantum dots and integrated semiconductor quantum-dot biomolecule complexes are described, this chapter describes a selection of applications of semiconductor quantum-dot biomolecule complexes in the study of carcinogenic cells and in the diagnosis of cancer.
17.2 Semiconductor Quantum Dots and Their Adaptation for Nanodiagnostics The applications of semiconductor quantum dots in biology have been highlighted [1, 2] in the study of subcellular and cellular processes of fundamental importance in biology. In recent years, semiconductor quantum dots have been used as well for imaging tumors. This chapter deals with all of these imaging domains with emphasis on the diagnosis of diseases. Quantum dots are a relatively new type of fluorescent probe with a number of advantages over fluorescent dyes that include brighter emission, isotropic emission, less stringent illumination conditions, and much longer lifetimes. Fluorescence microscopy coupled with image recording devices—usually with a charge coupled device (CCD)—may be used for the sensitive, high-resolution observations of fluorescent dots in vitro and in vivo. Prior to the widespread availability of semiconductor quantum dots, fluorophores were used for many years in biological imaging applications. In contract, semiconductor quantum dots have great advantages [3, 4] over conventional fluorophores. Such advantages include narrow symmetric emission spectra, the option for continuously and precisely tuning the emission wavelength of quantum dots by changing the size of each quantum dot, a single light source can be used for simultaneous excitation of multiple semiconductor quantum dots with different emission spectra of longer wavelengths than the source, ability to function after storage times of months, and extreme stability of coated quantum dots against photobleaching as well as changes in the pH of the biological electrolytes that are ubiquitous in biological environments. Indeed, semiconductor quantum dots are now viewed as efficient fluorophores for ultrasenstive, multicolor, and multiplexing biological imaging applications. The application of semiconductor of quantum dots as fluorescent probes in biological environments has been made possible by synthesizing semiconductor quantum dots that remain fluorescent in aqueous environments, and by conjugating semiconductor quantum dots with molecules that have affinities for binding to specific biological structures. To render semiconductor quantum dots water-soluable, early workers in the field functionalized quantum dots with organic compound such as mercaptoacetic acid (see [5]) or coated them with a hydrophilic layer such as silica (see [6]). There are many examples of conjugating semiconductor quantum dots with biomolecules for selective binding; many of these techniques have been adapted
17.2 Semiconductor Quantum Dots and Their Adaptation for Nanodiagnostics
443
from earlier work on binding organic dyes to biomolecules with the desired binding properties. These same molecules have been used widely in functionalizing semiconductor quantum dots with molecules that have affinities for binding to specific biological target molecules. As examples, biotin and avidin are known to bind to each other with high binding affinity; likewise, amine and carboxyl groups bind together in a wide range of biomolecular reactions. In another widely used approach, the side group of cysteine is used to form a strong thiol bond. As is well known, many of the important types of molecules—peptides, proteins, and antibodies—contain carboxyl and amine groups as end groups. Indeed, it is the binding of many amino acids together in chains through their terminal end groups (COOH, carboxyl groups, and NH2, amine groups) that leads to the formation of peptides (short chains of amino acids), proteins (long chains of amino acids), and antibodies (subset of protein having highly selective binding properties). Antibodies are known to bind with high selectivity to specific molecular targets, or antigens, and are of great utility in binding semiconductor quantum dots to specific targets. They have been used widely in semiconductor quantum dot studies, just as they were used widely in earlier studies with organic fluorophores. Cross-linking reagents used widely in catalyzing the binding of biomolecules containing carboxyl or amine groups to semiconductor quantum dots functionalized with amine or carboxyl groups are EDC [1-Ethyl-3-(3-Dimethylaminopropyl) carbodiimide Hydrochloride] and Sulfo-NHS (N-Hydroxysuccinimide). For semiconductor quantum dots functionalized with carboxyl groups, EDC reacts with the carboxylic acid group and activates the carboxyl group to form an active O-acylisourea intermediate, allowing it to be coupled to the amino group in the reaction mixture. An EDC by-product is released as a soluble urea derivative. The O-acylisourea intermediate is unstable in aqueous solutions, making it ineffective in two-step conjugation procedures without increasing the stability of the intermediate using N-hydroxysuccinimide; this intermediate reacts with a primary amine to form an amide derivative [7]. Using such cross-linking procedures, semiconductor quantum dots functionalized with carboxyl or amine groups may be readily bound to peptides, proteins, and antibodies. As is well-known, peptide conjugated structures find many uses in binding to a variety of cellular structures because different peptides have binding affinities to selected amino-acid-based biomolecules known as integrins, receptors, and ion channels. These integrins, receptors, and ion channels are also known as transmembrane proteins because they generally span the bilipid layer that forms the cellular membrane. Integrins are heterodimers composed of α-coil and β-sheet protein subunits. Integrins perform many functions such as the binding of cells to extracellular matrices (ECMs) as well as the adhesion of cells to cells. The 23 different integrin heterodimers are formed by 17 different α and 8 β subunits. As is well known, these transmembrane proteins form contacts with ECM-ligands including collagens, entactin, fibronectin, fibrinogen, laminin, thrombospondin, virtronectin, and intercellular adhesion molecules [8, 9]. Clearly, binding peptides to semiconductor quantum dots results in a biomolecule quantum-dot complex that has many applications in selective binding to specific biostructures. In the same way, quantum-dot antibody complexes may be used to label specific antigens with fluorescent quantum dots. Such means of functionalizing semiconductor quantum dots have been used widely in the study of
444
Integration of Manmade Nanostructures with Biological Systems
biological processes on the nanoscale; see for example [10–13] and references therein. In this chapter, these quantum-dot biomolecule complexes will be considered in terms of their utility for diagnosing diseases. Semicondcutor quantum dots are nanocrystalline structures with diameters generally ranging from approximately 2 to 20 nm. Because they have a periodic crystalline structure, they form electronic bands similar to those of bulk semiconductors. However, as a result of their small sizes, their electronic band energies change with size due to quantum-mechanical confinement effects. These dimensional size effects cause semiconductor quantum dots to have extraordinary electronic, optical, and mechanical properties; these properties have been investigated by a broad community of scientists and engineers for over a decade [13–26]. As discussed previously, semiconductor quantum dots have proven to be highly efficient nanoscale fluorescent markers. From semiconductor physics, it is known that a pure, defect-free semiconductor has a gap in energy, Egap, where no quantum (electronic) states exist. For energies greater than this energy gap there is a band of energy states known as the conduction band and for energies less than the energy gap there is a band of energy states known as the valence band. The lowest energy in the conduction band in a semiconductor is denoted by Ec and the highest energy in the valence band is denoted by Ev. Many of the basic properties of semiconductor quantum dots may be understood by treating the electrons and holes as being trapped within the diameter of the quantum dot. In this model, the electron (or hole) is treated quantum-mechanically as a wave using the de Broglie relation, λ = h/m*e, hv, where λ is the wavelength associated with the particle, h is Planck’s constant, 6.62 × 1034 Joule-seconds, and m*e,h is the effective mass of the electron (hole). Alternatively, the product of m*e,h and v may be written as pe,h , the momentum of the electron (hole). In this simple but useful model, it is assumed that the electron (or hole) wave is confined within the diameter, d, of the quantum dot by forming traditional standing waves. That is, the electron (or hole) wave forms wave patterns such that integral multiples of half-wavelengths fit into the quantum dot; that is, n λn/2 = d, where n = 1, 2, 3, …, and where a subscript n has been associated with λn to distinguish the different confined modes [27, 28]. To calculate the energy associated with the electron (or hole) wave, the standard formula, *
2
*
E = (1/2) m e,h v = pe,h 2/ 2 m e,h
(17.1)
is used where p = h/λ and λ satisfies n λn/2 = d. It follows straightforwardly, for electrons that,
( ) = h / (2m (2d / n) ), n = 1, 2, 3, K,
E n( e ) = pe2, n / 2 m*e = h 2 / 2 m*e λ2n
(
= n π h / 2m d 2
2
2
* e
2
* e
2
2
) = n h / (8m d ) 2
2
* e
2
(17.2)
where h is defined by h = h / 2π and is referred to as h-bar. Likewise, for holes a similar derivation gives,
(
)
E n( h ) = n ¢ 2 π 2 h 2 / 2 m*h d 2 , n ¢ = 1, 2, 3, K ,
(17.3)
17.2 Semiconductor Quantum Dots and Their Adaptation for Nanodiagnostics
445
In these equations, n is known as the quantum number for the conduction band states and n´ is the quantum number for the valence states. For a bulk semiconductor (i.e., a semiconductor that has dimensional sizes much larger than the electron wavelength), the energy states just above the conduction band edge are treated as continuous. In contrast, for a quantum dot the states are discrete and are given (17.2) and (17.3). In the limit that d becomes large, the discrete states become very close to each other and the continuum states of the bulk semiconductor are recovered. This is verified easily by considering (17.2) and (17.3) for the case of a bulk semiconductor where d is equal to say 1 cm (107 nm) and a quantum dot where d is taken as 10 nm; thus, the energy levels for the quantum dots are separated by a factor of 1012 more than for the bulk semiconductor. Since energy level separations in quantum dots are roughly on the order of 0.01 to 0.1 eV (electron volt), it follows that the energy level separations in the bulk material are of the magnitude of only 10-13 eV! This is an exceedingly small energy level separation and the continuum approximation works quite well for the bulk semiconductor but not at all for the quantum dot. In the case of the quantum dot, it is clear that dimensional confinement (d becoming small) leads to a separating of the many states in the so-called bulk semiconductor continuum into a series of discrete states that are clearly evident in optical studies of quantum dots. As discussed previously, the lowest electron energy level in the quantum dot corresponds to a standing wave of one-half wavelength confined in the quantum dot (n = 1), the second electron energy level in the quantum dot corresponds to a standing wave of one wavelength (two half wavelengths) confined in the quantum dot (n = 2), the third electron energy level in the quantum dot corresponds to a standing wave of one-and-half wavelengths (three half wavelengths) confined in the quantum dot (n = 3), and so forth. The analysis used to derive (17.2) and (17.3) is based on confinement in only one linear dimension. It can be shown that in the case of a spherical quantum dot with three-dimensional confinement that the ground state energy, E1, of a particle of mass m in an infinitely deep spherical well is given by simply replacing the width of the one-dimensional well, d, with the radius, a, of the quantum dot. [29]. When an electron in the highest occupied state quantum dot—generally the highest state in the valence band (n´ = 1) for an undoped, defect-free semiconductor at low temperature—absorbs a photon of energy h 1,1´, where 1,1’ is the frequency of the photon, the electron is elevated to the lowest state in the conduction band (n = 1). Taking into account that the conduction band and valence band are separated in energy by the band gap, Egap, it follows that:
(
hν 1 ,1 ¢ = π 2 h 2 / 2 m*h a 2
)+π
2
(
h 2 / 2 m*e a 2
)+E
gap
+ E exciton
(17.4)
where the first two terms on the right side of this equation represent the ground state energies of an electron of mass m*e and the hole of mass m*h in the quantum dot of radius, a, Egap is the energy difference between the conduction and valence band edges, and Eexciton is a small correction due to the Coulomb attraction between the electron and the hole. The exciton energy is due to the attraction of the electron and hole for each other (through a Coulomb attraction) and it lowers the energy of the electron-hole pair by an amount that is generally 10s of millielectron volts; thus, it is usually the case that E gap >> E exciton since Egap is usually of the order of an eV or so.
446
Integration of Manmade Nanostructures with Biological Systems
There are important exceptions to this case; indeed, for semiconductors with very narrow bandgaps, such as HgxCd1-xTe with bandgaps corresponding to infrared frequencies, E gap ª E exciton . For the general case describing a transition between states n and n ¢, (17.4) may be modified using the general forms of (17.2) and (17.3) to yield,
(
hν n , n ¢ = n 2 π 2 h 2 / 2 m*e a 2
) + n¢
2
(
π 2 h 2 / 2 m*h a 2
)+E
gap
+ E exciton
(17.5)
The simple models discussed previously are based on the approximation that the potential wells are infinitely deep. While this approximation generally works well for the states near the band edge, it is not strictly valid. For finite tapping energies of the electrons (or holes) in a semiconductor quantum dot, it is necessary to use numerical techniques to solve the transcendental equations of quantum mechanics describing the energy levels [29, 30]. In addition, dielectric screening must be taken into account when the quantum dot is functionalized or coated with materials of different dielectric constants or immersed in a liquid of a specific dielectric constant; this is especially important in the case of immersion in water since ε = 80 for water. For quantum dots with noncubic crystal symmetries, notably wurtzite crystals with hexagonal units and with polar binding, spontaneous polarization may be very large and it follows that the associated internal electric fields lead to bandbending that causes shifts in the quantum dot energy levels. Semiconductor quantum dots that may form in a wurtzite crystal structure include GaN, AlN, InN, ZnO, ZnS, CdS, and CdSe; under appropriate growth conditions some of these materials form quantum dots with cubic crystal symmetries. For the wurtize structures with polar binding—such as GaN, for example—built-in spontaneous polarizations are of the order of –0.029 C/m2 [31], which corresponds to the very large field of –3.615 MV/cm. Two additional phenomena affecting the energy levels of quantum dots are in electrolytic suspensions and ligand-induced dipoles: (1) when colloidal quantum dots are in electrolytic suspensions they have surface charges that depend on the pH and result in double layers that perturb the energy states of the quantum dots; and (2) binding of a ligand to the surface of a quantum dot produces a dipole moment that may be 10s of Debyes [32], which may produce electric fields large enough to shift the energy levels of the quantum dot. Notwithstanding these limitations, (17.4) describes the n = 1 to n´ = 1 transition in many quantum dots with some accuracy provided d be replaced with the quantum dot radius, a, to account for three-dimensional confinement, as discussed previously. As depicted in Figure 17.1, a photon may be emitted when an electron in the lowest conduction band energy level, E(e)1 , makes a transition to the level, E(h)1’, in the valence band. In most semiconductor that emit light, the n = 1 to n´ = 1 transition is the most likely transition, and the emitted radiation has a strong peak at the energy of that transition. However, transitions between other levels may also occur (generally with lower probabilities) as long as the change in the energy of the electron equals the energy of the emitted photon. In addition to emitting photons (quanta of light), an electron (or hole) may also emit phonons (quanta of vibrational energy). For discussions of energy loss by phonon emission see [14]. From Figure 17.2 and (17.5), it is clear that the fluorescent spectrum of a particular quantum dot depends on the diameter of the quantum dot, the exciton energy, and the energy bandgap of the semiconductor. As a result, quantum dots may be engineered to
17.3 Semiconductor Quantum Dots as Applied to the Study of Cellular Properties
447
d Diameter of Quantum Dot = d Figure 17.1 Energy band structure and energy levels, E(e)n. and E(h)n, with a radius a = d/2. The lowest three electronic states—E(e)1, E(e)2, E(e)3, and the lowest two hole state - E(h)1, and E(h)2—are depicted. An emitted photon is depicted by the wavy line to the right.
tailor the emission spectrum. As an example, the peak emission wavelength—and therefore the color—of CdSe quantum dots may be engineered to be at any wavelength in the visible spectrum simply by changing the radius of the quantum dot. This is an example of quantum engineering. Clearly, if several CdSe quantum-dot radii are selected, the color (emission wavelength) of each dot will be different from the colors of the other quantum dots. This implies that quantum dots emitting at different wavelengths may be functionalized with different biomolecules, say with different antibodies or with different peptides, to simultaneously label different antigens. Moreover, unlike the case of organic dyes, all of the quantum dots may be excited with the same light source (laser, light-emitting diode, lamp, and so forth) so long as the photons have enough energy to overcome the bandgap and confinement energies illustrated in Figure 17.1. Alexson et al., [33], Stroscio and Dutta [12], and Gao et al. [34] have reviewed the properties of colloidal quantum dots synthesized in electrolytic environments; among other key properties, these colloidal suspensions may contain very roughly about 1016 quantum dots per cm3. Cross linking techniques may be used to functionalize these colloidal quantum dots with a variety of ligands including peptides, proteins, antibodies, and other desired biomolecules. The vast number of possible quantum-dot complexes that may be realized is evident. Not only is there a large number of ligands, there is also considerable variability in the types of quantum dots available for synthesizing these quantum dot complexes. Table 17.1 summarizes some of the properties of quantum dots that may be synthesized as water-based colloids.
17.3 Semiconductor Quantum Dots as Applied to the Study of Cellular Properties As an example of the labeling of breast cancer cells, Wu et al. [41] used ZnS-coated CdSe semiconductor quantum dots functionalized with IgG antibody and streptavidin to label the breast cancer marker Her2 on the surfaces of both fixed and
448
Integration of Manmade Nanostructures with Biological Systems
Table 17.1
Selected Quantum Dots with Energy Bandgaps and Values of the Spontaneous Polarizations
Compound Semiconductor AlN CdMnTe/Hg CdS hexagonal
CdS cubic CdSe hexagonal
CdSe cubic CdSeTe CdSe/CdTe CdTe GaN PbS PbSe
TiO2 ZnS ZnO
Bandgap (eV) 6.2 ca. 1.6; broad peak at 770 nm 2.4 Eg (A) 2.5 Eg (B) 2.55 Eg (C) 2.5 1.75 Eg (A) 1.771 Eg (B) 2.17 Eg (C) 1.9 ca. 1.55; emission peak near 800 nm ca. 1.45; Near-IR ca. 850 nm 1.49 3.36 0.41 0.27 3.2 3.68 3.35
Spontaneous Polarization, 2 (c/m ), and References –0.081 [35] [24] 0.002 [36]
0.006 [17, 19]
[25] [37] [38] –0.029 [13, 22] [21, 22] [23] [39] [40] –0.07 [36]
live cancer cells. The technique of electrostatic self-assembled CdSe-ZnS quantum dots was used by Jaiswal et al. [42] to demonstrate long-term multiple-color imaging of live cells for periods of over a week; see earlier work of Mattoussi et al. [43] for foundational aspects of this approach. Jaiswal et al. [42] used have a synthetically engineered protein G-zb (leucine zipper-containing peptide fused to the B2 binding domain of streptococcal protein G) or avidin was used to conjugate antibodies to colloidal quantum dots; moreover, the noninvasive labeling of mammalian HeLa cells through endocytosis of DHLA-capped quantum dots was observed. In still other early studies of tumors, Akerman et al. [44] used tri-n-octylphosphine oxide-coated ZnS-capped CdSe QDs coated with mercaptoacetic acid to render them water-soluable. The studies of Akerman et al. [44] included functionalizing the quantum dots with KDEPQRRSARLSAKPAPPKPEPKPKKAPAKK peptide (F3 peptide), which is known to preferentially bind to the blood vessels in various tumors and tumor cells; and functionalizing quantum dots with CGNKRTRGC (LyP-1), which recognizes the lymphatic vessels in certain tumors and the tumor cells. It was verified that these peptide-functionalized quantum dots did indeed bind to the indicated biological structures, by injecting suspensions of these quantum dots into the blood stream of a mouse. Semiconductor quantum dots functionalized with LDV-based peptides have been used [33, 45] to study HT1080 human fibrosarcoma cells. This approach is similar to the earlier work of Winter et al. [46].
17.4 Semiconductor-Quantum-Dots
449
In this work, different peptides containing LDV and RGD were conjugated to both CdS-mercaptoacetic (CdS/M) quantum dots and carboxylic-group-functionaled CdSe-ZnS quantum dots. The cysteine (C) amino acid on one end of the CGGGRGDS and CGGGLDV peptides was used to bind these peptides for binding to quantum dots through the relatively strong thiol linkage. For the CdSe-ZnS quantum dots, the previously described EDC (1-ethy-3-(3-dimethylamino propyl) carbodiimide) cross-linking protocol was used to bind these peptides to the quantum dots. HT1080 human fibrosarcoma cells were labeled with these quantum-dot complexes and observed using a fluorescent microscope. The selection of RGD- and LDV-based peptides was based on a reported analysis of integrin expression pattern on HT1080 fibrosarcoma cells. LDV and RGD both bind to α5β1, and RGD binds to α3β1 as well [45]; in addition, Schifferli and Henrich [47] show that different types of integrins including α2β1, α3β1, and α5β1 are expressed on HT1080 cells. Shi et al. [45] demonstrated by the labeling of HT1080 cells with quantum dots functionalized with both of these peptides, and they found that the LDV receptors are highly localized over selected regions of the cells; in contrast, the RGD receptors are distributed over the cellular membrane. Figure 17.2 illustrates the case where green-yellow quantum dots are bound to integrins that have affinities for binding to LDV.
(a)
(b)
(c)
Figure 17.2 (Color plate 28) Fibroblast cell imaged with white light (a) and with the light from green-yellow quantum dots—emitting at a wavelength of 565 nm—that are bound to the cellular membrane (b and c); these quantum dots are functionalized with the peptide, GGGGLDV. It is known that the LDV peptide has an affinity for binding to integrins that are commonly found on carcinogenic cells. The image on the left (a) is the white-light image of the fibroblast, while the central image (b), and the right image (c), reveal the green-yellow quantum-dot luminescence emitted from the quantum dots bound to transmembrane integrins that bind to the LDV peptide. In this case, the transmembrane integrins are bound to the quantum dots through the GGGGLDV linker. All three images are of the same cell but image (a) is a white-light image of the fibroblast, and images (b) and (c) are of the luminescent quantum dots that are illuminated with light with wavelengths in the range of 350 to 450 nm. As explained in the text, these quantum dots absorb in one band of wavelengths (350–450 nm in this case) and they emit at a different wavelength (565 nm in this case). In recording images (b) and (c), two different focal points were used, one (b) to focus in the quantum dots bound to integrins located near the central region of the cell, and the other (c) to focus on the quantum dots outside of the central region of the image. In this way, these quantum-dot tags reveal the locations of the integrins that are bound to the GGGGLDV-functionalized quantum dots. A Nikon E800 microscope with a 100x objective and a numerical aperture of 1.45 was used in conjunction with a charge-coupled device to record these images.
450
Integration of Manmade Nanostructures with Biological Systems
17.4 Semiconductor-Quantum-Dots—Biomolecule Complexes Used in the Study of Carcinogenic Cells and in Cancer Diagnosis Among the uses of semiconductor quantum dots for in vivo imaging applications are: imaging of tumor-specific membrane antigens [34, 48]; imaging of tumor vasculature [24, 49–51]; and imaging of sentinel lymph nodes (SLNs) [26, 52, 53]. Gao et al. [48] have developed quantum-dot-based multifunctional nanoparticle probes for cancer targeting and imaging in live animals. These nanoparticle probes are designed by encapsulating luminescent QDs with an ABC triblock copolymer; the amphiphilic polymer is linked to tumor-targeting ligands as well as drug-delivery structures. These nanoparticle probes have been applied to in vivo studies of human prostate cancer growing in nude mice. The findings show that the functionalized probes accumulate at tumors and bind, via antibodies, to cancer-related biomarkers on cell surfaces. Multicolor fluorescence imaging of cancer cells under in vivo conditions was accomplished using (a) subcutaneous injection of nanoprobe-tagged cancer cells and (b) systemic injection of multifunctional nanoprobes. Gao et al. [34, 48] have also reported on an integrated a whole-body macroillumination system with wavelength-resolved spectral imaging for efficient background removal and precise delineation of weak spectral signatures. These results portend future capabilities of multiplexed imaging of molecular targets in vivo with high sensitivity. Gao et al. [34, 48] have used core/shell CdSe-ZnS semiconductor quantum dots that are protected by two means: (1) a coordinating ligand, tri-n-octylphosphine oxide (TOPO); and (2) an amphiphilic polymer coating. This novel design is effective as a result of the strong hydrophobic interactions between TOPO and the polymer hydrocarbon; indeed, these layers bond to each other to form a hydrophobic protection structure that resists hydrolysis and enzymatic degradation even under complex in vivo conditions. More specifically, Gao et al. [34, 48] have used a high-molecular-weight (100-kDa) copolymer with an ABC triblock structure and a grafted 8-carbon (C-8) alkyl side chain. The triblock polymer consists of (a) a hydrophobic hydrocarbon side chain, (b) a hydrophobic hydrocarbon side chain, (c) a polybutylacrylate segment (hydrophobic), and (d) a polyethylacrylate segment (hydrophobic). Upon linking to polyethylene glycol (PEG) molecules. Gao et al. [34, 48] find that the polymer-coated quantum-dot complexes did not change their absorption spectra, emission spectra and fluorescence quantum yields for the following conditions: (1) over a very broad range of pH spanning the region from 1 to 14, (2) over a wide range of salt conditions varying from 0.01 to 1M, and (3) upon treatment with 1.0 M hydrochloric acid. The engineering of these robust and effective quantum-dot-based structures illustrates the future potential for the even wider use of quantum-dot-based nanocomplexes in nanomedicine. Morgan et al. [24] have examined the use of near-infrared semiconductor quantum dots in deep-tissue in vivo optical imaging. The penetration depth of near-infrared radiation is greater than that of visible radiation and, therefore, quantum dots emitting in the near-infrared region have considerable promise in deep-tissue studies. In these studies, 5-nm diameter CdMnTe/Hg semiconductor quantum dots, with a broad fluorescence peak at 770 nm, coated with bovine serum albumin (BSA) were used as the near-infrared labels. These quantum dots were injected subcutaneously or intravenously into mice and then excited with a spatially broad 633-nm source.
17.4 Semiconductor-Quantum-Dots
451
The fluorescence image was recorded with a sensitive charge-coupled device (CCD) camera. These quantum-dot complexes were shown to be useful as an angiographic contrast agent for vessels surrounding and penetrating a murine squamous cell carcinoma in a C3H mouse. Morgan et al. [24] also performed a preliminary assessment of the depth of penetration for excitation and emission for experiments done by imaging a beating mouse heart, both after a thoracotomy and through an intact thorax. Morgan et al. [24] found that there is no significant degradation or photobleaching of the quantum-dot complexes even after an hour of continuous excitation. They also found that the stability of the quantum-dot complexes as well as the time resolution of the optical signal make them attractive candidates for future pharmacokinetic imaging studies. Ballou et al. [52] have used quantum dots having four different surface coatings to assess their utility for in vivo imaging. The quantum dots used in this study were core/shell zinc-sulfide-coated cadmium selenide quantum dots (Quantum Dot Corp.) with emission peaks at 606, 635, 645, and 655 nm; there, quantum dots were coated with an amphiphilic poly(acrylic acid) polymer (denoted as amp QDs) or the same coating conjugated with methoxy- or carboxy-terminated poly(ethylene glycol) amine (varieties denoted as mPEG-750 QDs, mPEG-5000 QDs, and COOH-PEG-3400 QDs). The PEG (poly(ethylene glycol)) formulations used in this study were obtained from Shearwater Polymers Corp (Huntsville; COOH-PEG3400-amine, mPEG-5000-amine) and from Sigma-Aldrich (Milwaukee, WI; mPEG-750-amine). These quantum dots were dispersed at 1.0 μM in 50-mM sodium borate, pH 8.0, with a molar ratio of 2,000:1 of PEG-amine (7.5 mg). In addition, these materials were then reacted with a 1500:1 molar ratio of N-dimethylaminopropyl-N-ethylcarbodiimide (EDC, 1.4 mg) for 2 hours. Purification was accomplished using a 100-kDa centrifugal ultrafiltration device (Vivascience, Edgewood, NY). As is well-established, Bollou et al. [52] demonstrated that these quantum dots may be excited for wavelengths shorter than the emission maximum. (In theory, it is necessary to use an excitation energy that is equal to or greater than the semiconductor bandgap of the quantum dot which, in general, is a few tens of meV greater than the energy of the emitted light due to the negative binding energy of the excitons, as discussed previously. This exciton binding energy results in small shift in the emission energy to energies below the bandgap of the semiconductor quantum dots.) Using only quantum dots for detection, imaging of live mice was preformed using (1) fluorescence imaging, (2) necropsy, (3) frozen tissue sections for optical microscopy, and (d) electron microscopy, on scales ranging from centimeters to nanometers. For the amphiphilic poly(acrylic acid), short-chain (750-Da) methoxy-PEG, and long-chain (3,400-Da) carboxy-PEG quantum dots, circulating half-lives were less than 12 minutes. For long-chain (5,000-Da) methoxy-PEG quantum dots the half-life was about 77 minutes. These quantum dots were found to remain fluorescent for at least 4 months in vivo. Cai et al. [49] have reported the in vivo targeting and imaging of tumor vasculature using semiconductor quantum dots functionalized with arginine-glycine-aspartic acid (RGD) peptides. In this study, mice bearing subcutaneous U87MG human glioblastoma tumors were administered quantum-dot-RGD complexes intravenously. It was observed that the tumor fluorescence intensity
452
Integration of Manmade Nanostructures with Biological Systems
when labeled with cadmium telluride quantum dots—with a principal emission line at the wavelength of 705 nm—reached a maximum at 6 hours after injection. The results reported provide encouraging findings portending the use of semiconductor quantum dots functionalized with peptides in near-infrared optical imaging of tumor vasculature in cancer detection as well as in cancer management, possibly including imaging-guided surgery. The live animal imaging system used by Cai et al. is described in some detail by Zhang et al. [53]. Parungo et al. [53] have used studied the near-infrared fluorescence signatures produced by both dyes and quantum dots of esophageal sentinel lymph nodes of six Yorkshire pigs. Their analysis indicated that such a procedure is a novel and reliable intraoperative technique and that it is useful in assisting in the identification and resection of esophageal sentinel lymph nodes. The lymph tracers used in this study fluoresce in the near-infrared and they facilitate visualization of migration to sentinel lymph nodes using a custom intraoperative imaging system. It was found that the injection of the near-infrared fluorescent lymph tracers into the esophagus revealed communicating lymph nodes within 5 minutes of injection. It was further found that in all six pigs given the quantum dot injection, only a single sentinel lymph node was identified; among pigs given a fluorophore-conjugated albumin injection, 5 of 12 single-sentinel lymph nodes were identified, while 7 of 12 two-sentinel lymph nodes were identified. In these studies, Type II core/shell semiconductor quantum dots [26] were used that contained an inorganic core of cadmium telluride, an inorganic shell of cadmium selenide, and an outer organic coating of solubilizing oligomeric phosphines; the hydrodynamic diameter measured using gel filtration techniques was found to be 15 to 20 nm; the dimensional sizes of the shell and the core were selected so that these quantum dots fluoresced in the near-infrared portion of the spectrum with a peak emission wavelength of 840 nm. (Type II semiconductor structures have higher values of both Ec and Ev in one of the two semiconductors making up the structure. This is in distinction to Type I semiconductor structures which band alignments as shown in Figure 17.1.) A stock solution of 0.2-μmol/L of quantum dots in phosphate-buffered saline solution (pH 7.4) was used. The dye-based tracers were synthesized using human serum albumin (HSA) that was covalently conjugated to the near-infrared fluorophore CW800 by means of an amide bond (HSA800); the peak absorbance and emission of the HSA800 complexes were 778 and 795 nm, respectively. The HSA800 was used in a phosphate buffered saline solution, pH 7.4; the stock solution of 0.8-mg/mL HSA800 in phosphate-buffered saline solution was used throughout the studies of Parungo et al. [53]. It was found that these fluorescent biologic labels were easily visible deep within tissue. Diagaradjane et al. [25] have used peptide-labeled semiconductor quantum dots for in vivo imaging of the epidermal growth factor (EGF) receptor (EGFR) that is overexpressed on tumors as compared with expression on adjacent normal tissues. Diagaradjane et al. have developed and characterized EGF-conjugated quantum dots (EGF-QD) with an emission peak near 800 nm (near-infrared spectral region) to image EGFR expression in human colon cancer xenografts. The quantum dots used in this study are ZnS-coated CdSeTe core/shell quantum dots; these quantum dots are conjugated with EGF as nanoprobes of the EGF receptors (EGRFs). Diagaradjane et al. note that these nanoprobes could potentially be used for quantifiable and repetitive imaging of EGFR during and after a therapeutic intervention.
17.5 Conclusion
453
These studies have emphasized EGFR since it is a transmembrane protein that controls biological phenomena, including proliferation, angiogenesis, tissue invasion, and metastasis. EGFR is ubiquitously expressed in normal tissues, but it is overexpressed on the surface of many tumors. Diagaradjane et al. conclude that EGF-QD nano-probes may permit quantifiable imaging of EGFR expression since they observed measurable contrast enhancement of tumors 4 hours after the administration of EGF-QD and its subsequent normalization at 24 hours.
17.5
Conclusion In just over a decade, colloidal semiconductor quantum dots have made the transition from early studies on synthesizing and functionalizing semiconductor quantum dots for tagging specific biomolecules to the study of diseased cells in vitro as well as the study of diseases in vivo. This progress is due in large measure to the increasing variety of available water-based colloids of semiconductor quantum dots that have strong optical emission spectra spanning the ultraviolet, visible, and infrared spectra. As discussed in this chapter, these quantum dots have a number of attractive features including bright emission spectra and long shelf lives. Accompanying the progress in the synthesis of luminescent quantum-dot probes, is the progress made in imaging quantum dots functionalized for biomolecule tagging; these imaging techniques have evolved from the use of tradition fluorescence microscopy to imaging of quantum-dot luminescence in studies of live animals. This chapter has focused on providing illustrative examples of the application of these techniques to the study of cancer since this easily represents one of the potentially most important applications of semiconductor quantum dots.
Problems 8
17.1 The speed of light in vacuum is approximately 3 × 10 m/sec. (a) Using the relationship λf = c, find the frequency of blue light having a wavelength of 0.48 μm. (b) What is the wavelength of a visible 5 × 1014-Hz photon? (c) What is the wavelength of a radiowave photon with frequency 107 Hz? 17.2 The energy of a photon is given by E = hf where h is Planck’s constant and f is the frequency of the light. Using the fact that λf = c, derive a relationship between E and λ. 17.3 Consider a CdSe quantum-dot biotag encased in the high-bandgap material ZnS. Assume that the n = 2 to n = 1 electronic intersubband transition is in the infrared portion of the spectrum at λ = 10 μm. Assume that the energy levels of this semiconductor quantum dot are described by (17.4) and (17.5). If it is desired to change the 10-μm luminescence to 8-μm luminescence, how much must be the fractional change in the radius of the quantum dot? 17.4 What properties of the luminescent spectra of ZnS-coated CdSe quantum dots make them especially attractive for use as biological tags?
454
Integration of Manmade Nanostructures with Biological Systems
17.5 Assume that you have at your disposal two different CdSe quantum dots with different photoluminescent (PL) spectra; say one emits at 600 nm and the other at 500 nm. Assume that you have at your disposal GaN quantum dots with PL spectra in the ultraviolet portion of the spectrum with a peak emission at 325 nm. If a laser emitting at 400 nm is used to illuminate all three types of quantum dots, which dots will be luminescent? (b) If these different quantum dots are functionalized with different molecules to facilitate the simultaneous multicolor imaging of three different biostructures, what wavelengths are allowed for the illuminating laser?
Acknowledgments The authors acknowledge the support of their studies of semiconductor quantum dots from ARO, AFOSR, DTRA, DoE, SRC, NSF, and Richard and Loan Hill.
References [1] Stroscio, Michael A. and Mitra Dutta, editors, Biological Nanostructures and Applications of Nanostructures in Biology: Electrical, Mechanical & Optical Properties Kluwer Academic Publishers, New York, 2004. [2] Jovin, T.M. “Quantum Dots Finally Come of Age,” Nature Biotechnology, Vol. 21, 2003, 32–33. [3] Alivisatos, A.P., “Perspectives on the Physical Chemistry of Semiconductor Nanocrystals,” J. Phys. Chem., Vol. 100, No. 31, 1996, pp. 13226–13239. [4] Klarreich, E. 2001 “Biologists Join the Dots,” Nature, Vol. 413, 2001, pp. 450–452. [5] Chan, W.C., and S. Nie, “Quantum Dot Bioconjugates for Ultrasensitive Nonisotropic Detection,” Science, Vol. 281, 1998, pp. 2013–2016. [6] Bruchez, M. Jr., M. Moronne, P. Gin, S. Weiss, and A. P. Alivisatos, “Semiconductor Nanocrystals as Fluorescent Biological Labels,” Science, Vol. 281, No. 5385, 1998, pp. 2013–2016. [7] Hermanson, G.T., Bioconjugate Techniques, San Diego, Academic Press, 1996 [8] Alberts, B., A. Johnson, J. Lewis, M. Raff, K. Roberts, and P. Walter, Molecular Biology of the Cell, 4th Ed. New York, Garland Science, 2002. [9] Mizejewski, G. J., “Role of Integrins in Cancer: Survey of Expression Pattern,” Exp. Bio. Med., Vol. 222, 1999, pp. 124–138. [10] Chan, W.C., D. J. Maxwell, X. Gao, R. E. Bailey, M. Han, and S. Nie, “Luminescent Quantum Dots for Multiplexed Biological Detection and Imaging,” Curr. Opin. Biotechnol., Vol. 13, No. 1, 2002, 40–46. [11] Dahan, M., S. Levi, C. Luccardini, P. Rostaing, B. Riveau, and A. Triller, “Diffusion Dynamics of Glycine Receptors revealed by Single Quantum Dot Tracking,” Science, Vol. 302, No. 5644, 2003, pp. 442–445. [12] Stroscio, Michael A., and Mitra Dutta, “Integrated Biological-Semiconductor Devices,” Proceedings of the IEEE, Vol. 93, No. 10, 2005, pp. 1772–1783. [13] Zhang, Christina J., Howard W. H. Lee, Ian M. Kennedy,. and Subhash H. Risbud, “Observation of quantum confined excited states in GaN nanocrystals,” Appl. Phys. Lett., Vol. 72, No. 23, 1998, pp. 3035–3037.
References
455
[14] Stroscio, Michael A., and Mitra Dutta, Phonons in Nanostructures, Cambridge University Press, Cambridge, 2001. [15] Chen, H.M., X. F. Huang, L. Xu, K. J. Chen, and D. Feng, “Self-assembly and Photoluminescence of CdS-mercaptoacetic Clusters with Internal Structures,” Superlattices and Microstructures, Vol. 27, No. 1, 2000, pp. 1–5. [16] Empedocles, E., D. J. Norris, and M. G. Bawendi, “Photoluminescence Spectroscopy of Single CdSe Nanocrystallite Quantum Dots,” Phy. Rev. Lett., Vol. 77, No. 18, 1996, pp. 3873–3876. [17] Hines, Margaret and Philippe Guyot-Sionnest, “Synthesis and characterization of strongly luminescing ZnS-capped CdSe nanocrystals,” Journal of Physical Chemistry, Vol. 100, No. 2, 1996, pp. 468–471. [18] Kuno, M., D. P. Fromm, H. F. Hamann, A. Gallagher, and D. J. Nesbitt, “Nonexponential “Blinking” Kinetics of Single CdSe Quantum Dots: A Universal Power Law Behavior,” J. Chem. Phys., Vol. 112, No. 7, 2000, pp. 3117–3120. [19] Schmidt, Mark E., Sean A. Blanton, Margaret A. Hines, and PhilippeGuyot-Sionnest, “Polar CdSe Nan ocrystals: Implications for Electronic Structure,” Journal of Chemical Physics, Vol. 106, No. 12, 1997, pp. 5254–5259. [20] Micic, O. I., S. P. Ahrenkiel, D. Bertram, and A. J. Nozik, “Synthesis, Structure, and Optical Properties of Colloidal GaN Quantum Dots,” Appl. Phys. Lett., Vol. 75, No. 4, 1999, pp. 478–480. [21] Nenadovic, Milica T., Mirjana I. Comor, Vesna Vasic, and Olga I. Micic, “Transient Bleaching of Small PbS Colloids,” J. Phys. Chem., Vol. 94, No. 16, 1990, pp. 6390–6396. [22] Machol, Janet L., Frank W. Wise, Ramesh C. Patel, and David B. Tanner, “Vibronic Quantum Beats in PbS Microcrystals,” Phys. Rev. B, Vol. 48, No. 4, 1993, 2819–2822. [23] Wehrenberg, Bryan L., Congjun Wang, and Phillip Guyot-Sionnest, “Interband and intraband optical studies of PbSe colloidal quantum dots,” J. Phys. Chem. B, Vol. 106, No. 41, 2002, pp. 10634–10640. [24] Morgan, N.Y., S. English, W. Chen, V. Chernomordik, A. Russo, P. D. Smith, “Real time in vivo non-invasive optical imaging using near-infrared fluorescent quantum dots,” Academic Radiology, Vol. 12, No. 3, 2005, pp. 313–323. [25] Diagaradjane, P., J. M. Orenstein-Cardona, N. E. Colon-Casasnovas, A. Shentu Deorukhkar, S., N. Kuno, D. L. Schwartz, J. G. Gelovani, and S. Krishnan, “Imaging epidermal growth factor receptor expression in vivo: pharmacokinetic and biodistribution characterization of a bioconjugated quantum dot nanoprobe,” Clin. Cancer Res., Vol. 14, No. 3, 2008, pp. 731–741. [26] Kim, S., Y. T. Lim, E. G. Soltesz, A. M. De Grand, J. Lee, A. Nakayama, Anthony J. Parker, Tomislav Mihaljevic, Rita G. Laurence, Delphine M. Dor, Lawrence H. Cohn, Mounge G. Bawendi, John V. Frangioni, “Near-infrared fluorescent type II quantum dots for sentinel lymph node mapping,” Nat Biotechnol. , Vol. 22, No. 1, 2003, pp. 93–97. [27] Mitin, V., V. Kochelap, and Michael A. Stroscio, Introduction to Nanoelectronics, Cambridge University Press, Cambridge, 2007. [28] Mitin, V., V. Kochelap, and Michael A. Stroscio, Quantum Heterostructures, Cambridge University Press, Cambridge, 1999. [29] Liboff, Richard, Introductory Quantum Mechanics, San Francisco, Holden-Day, Inc., 1980. [30] Banyai, L., and S. W. Koch, Semiconductor Quantum Dots, Singapore World Scientific Publishing Co., 1993. [31] Bernardini, F., V. Fiorentini, and D. Vanderbilt, “Spontaneous polarization and piezoelectric constants of III-V nitrides,” Phys. Rev. B, Vol. 56, No. 16, 1997, pp. R10024–10027. [32] Rajh, T., L. X. Chen, K. Lukas, T. Liu, M. C. Thurnauer, and D. M. Tiede, “Surface Restructuring of Nanoparticles: An Efficient Route for Ligand-Metal Oxide Crosstalk,” J. Phys. Chem. B, Vol. 106, No. 41, 2002, pp. 10543–10552.
456
Integration of Manmade Nanostructures with Biological Systems [33] Alexson, D., Y. Li, D. Ramadurai, P. Shi, L. George, M. Uddin, P. Thomas, S. Rufo, M. Dutta, M., and M. A. Stroscio, “Binding of Semiconductor Quantum Dots to Cellular Integrins,” IEEE Tran. on Nanotechnology, Vol. 3, No. 1, 2004, 86–92. [34] Gao, X.,L. Yang, J. A. Petros, F. F. Marshall, J. W. Simons, S. Nie,. “In Vivo Molecular and Cellular Imaging with Quantum Dots,” Current Opinion Biotechnology, Vol. 16, No. 1, 2005, pp. 63–72. [35] Andreev, A. D., and E. P. O’Reilly, “Theory of the Electric Structure of GaN/AlN Hexagonal Quantum Dots,” Phys. Rev. B, Vol. 62, No. 23, 2001, pp. 15851–15870. [36] Jerphagnon, J., “Invariants of the Third-Rank Cartesian-Tensor: Optical Nonlinear Susceptibilities,” Phys. Rev. B, Vol. 2, No. 4, 1970, pp. 1091–1098. [37] Kim, S., Y. T. Lim, E. G. Soltesz, A. M. De Grand, J. Lee, A. Nakayama,Anthony J. Parker, Tomislav Mihaljevic, Rita G. Laurence, Delphine M. Dor, Kawrence H. Cohn, Moungi G. Bawendi, and John V. Frangioni, “Near-Infrared Fluorescent Type II Quantum Dots for Sentinel Lymph Node Mapping,” Nat Biotechnol., Vol. 22, No. 1, 2003, pp. 93–97. [38] de Paula, A.M., and L. C. Barbosa, C. B. H. Cruz, O. L. Alves, J. A. Sanjurjo, and C. L. Cesar, “Quantum Confinement Effects on the Optical Phonons of CdTe Quantum Dots,” Superlattices and Microstructures, Vol. 23, No. 5, 1998, pp. 1104–1106. [39] Rajh, T., Z. Saponjic, J. Liu, N. M. Dimitrijevic, N. F. Scherer, M. Vega-Arroyo, P. Zapol, L. A. Curtis, and M. Thurnauer,“Charge Transfer Across the Nanocrystal-DNA Interface: Probing DNA Recognition,” Nano Lett., Vol. 4, No. 6, 2004, pp. 1017–1023, 2004. [40] Xu, Yong and M.A. A. Schoonen, “The Absolute Energy Positions of Conduction and Valence Bands of Selected Semiconducting Minerals,” American Mineralogist, Vol. 85, No. 3–4, 2000, pp. 543–556. [41] Wu, X., H. Liu, J. Liu, K. N. Haley, J. A. Treadway, J. P. Larson, N. Ge, F. Peale, and M. P. Bruchez, “Immunofluorescent Labeling of Cancer Marker Her2 and other Cellular Targets with Semiconductor Quantum Dots,” Nat. Biotechnol., Vol. 21, 2003, pp. 41–46. [42] Jaiswal, J.K., H. Mattoussi, J. M. Mauro, and S. M. Simon, “Long-Term Multiple-Color Imaging of Live Cells Using Quantum Dot Bioconjugates,” Nat. Biotechnol., Vol. 21, 2003, pp. 47–51. [43] Mattoussi, H., J. M. Mauro, E. R. Goldman, G. P. Anderson, V. C. Sundar, F. V. Mikulec, and M. G. Bawendi, “Self-Assembly of CdSe-ZnS Quantum Dot Bioconjugates Using an Engineered Recombinant Protein,” J. Amer. Chem. Soc., Vol. 122, No. 49, 2000, pp. 12142–12150. [44] Akerman, M.E., W. C. Chan, C. P. Laakkonen, S. N. Bhatia, and E. Ruoslahti, “Nanocrystal Targeting In Vivo,” Proc Natl. Acad. Sci. USA , Vol. 99, No. 20, 2002, pp. 12617–12621. [45] Shi, Peng, Hongfeng Chen, Michael R. Cho, and Michael A. Stroscio, “Peptide-Directed Binding of Quantum Dots to Cellular Integrins,” IEEE Transactions on Nanobioscience, Vol. 5, No. 1, 2006, pp. 15–19. [46] Winter, J. O., T. Y. Liu, B. A. Korgel, and C. E. Schmidt, “Recognition Molecule Directed Interfacing Between Semiconductor Quantum Dots and Nerve Cells,” Advanced Materials, Vol. 13, No. 22, 2001, pp. 1673–1677. [47] Schifferli, K.P., and C. J. Henrich, “Analysis of Integrin Expression and Function in HT1080 Cells Using Inhibitory Anti-Integrin Antibodies,” Focus, Vol. 18, No. 1, 1996, pp. 13–15. [48] Gao, X., Y. Cui, R. M. Levenson, L. W. Chung, and S. Nie, “In Vivo Cancer Targeting and Imaging with Semiconductor Quantum Dots,” Nature Biotechnology, Vol. 22, No. 8, 2004, pp. 969–976. [49] Cai, W., D.-W. Shin, K. Chen, O. Gheysens, Q. Cao, S. X. Wang, S. S. Gambhir, and X. Chen, “Peptide-Labeled Near-Infrared QuantumDots for Imaging Tumor Vasculature in Living Subjects,” Nano Letters, Vol. 6, No. 4, 2006, pp. 669–676.
References
457
[50] Akerman, M.E., W. C. Chan, P. Laakkonen, S. N. Bhatia, and E. Ruoslahti, E., “Nanocrystal Targeting In Vivo,” Proc. Natl. Acad. Sci., USA, Vol. 99, No. 20, 2002, pp. 12617–12621. [51] Larson, D.R., W. R. Zipfel, R. M. Williams, S. W. Clark, M. P. Bruchez, F. W. Wise, and W. W. Webb, “Watersoluble Quantum Dots for Multiphoton Fluorescence Imaging In Vivo,” Science, Vol. 300, 2003, pp. 1434–1436. [52] Ballou, B., B. C. Lagerholm, L. A. Ernst, M. P. Bruchez, and A. S. Waggoner, “Noninvasive Imaging of Quantum Dots in Mice,” Bioconjugate Chemistry, Vol. 15, No. 1, 2004, 79–86. [53] Parungo, C., S. Ohnishi, S.-W. Kim, S. Kim, R. G. Laurence, E. G. Soltesz, F. Y. Chen, Y. L. Colson, L. H. Cohn, M. G. Bawendi, and J. V. Frangioni, “Intraoperative Identification of Esophageal Sentinel Lymph Nodes with Near-Infrared Fluorescence Imaging,” J Thorac Cardiovasc Surg, Vol. 129, No. 4, 2005, 844–850. [54] Zhang, G.-J., T. B. Chen, R. Hargreaves, C. Sur , and D. L. Williams, “Bioluminescence Imaging of Hollow Fibers in Living Animals: Its Application in Monitoring Molecular Pathways,” Nature Protocols, Vol. 3, No. 5, 2008, 891–899.
CHAPTER 18
Two-Photon Microscopy for In Vivo Analysis of Neural and Secretory Activities Tomomi Nemoto
18.1
The Need for Noninvasive Imaging Optical microscopy enables us to observe various simultaneously occurring events in living cells. This capability is important for monitoring the spatio-temporal patterns of events in living cells. Two-photon excited fluorescence (TPEF) microscopy, a technology based on multiphoton excitation (MPE), is one of the most promising candidates for such imaging. This is because TPEF microscopy has the demonstrated capability to obtain cross-sectional images from deep within nearly intact tissue samples over long observation times with excellent spatial resolution. These advantages have spurred wider adoption of the method, especially in neurological studies [1, 2]. This chapter describes both the current and anticipated capabilities of TPEF microscopy, based on a discussion of previous publications and recently obtained data. One of the aims of recent advancements in clinical and diagnostic technologies has been to reduce the burden on the patient by minimizing surgical procedures and manipulations during testing and treatment. This is critical for relieving the physical and mental strain on a patient, reducing treatment time, and improving the patient’s quality of life. In this sense, noninvasive imaging techniques such as magnetic resonance imaging and positron emission tomography are currently the most important testing technologies. The development of endoscopy and its application to surgical procedures have enabled many patients to avoid undergoing major operations. Furthermore, imaging technologies will play an increasingly important role in research in fields such as basic medicine and life sciences, since it is desirable to, as much as possible, observe samples in their natural state. After the completion of the Human Genome Project, one of the goals of life science research is to understand vital functions as systems. Now that modern life science, by adopting a reductionism approach, has successfully obtained the database of the human genome, we need to develop an integrated understanding of how the network of interactions between biological molecules performs the functions of cells and of the organism. The development of modern clinical medicine and of pharmacological research cannot be explained without reference to the simultaneous development of the life sciences, which, to a great extent, was accomplished through advances in molecular biology. Thus, one of the current aims in modern clinical medicine and pharmaco-
459
460
Two-Photon Microscopy for In Vivo Analysis of Neural and Secretory Activities
logical research is to examine how physiologically and pharmacologically active substances spatially and temporally affect the dynamic states of the functional networks of interactions between biological molecules. In addition, an urgent issue in pharmaceutical research is to determine what kinds of individual traits affect drug efficacy. Another pressing need is to determine the site specificity of the effects of drugs in the body in order to prevent systemic side effects. In order to satisfy these demands, it is necessary to confirm whether a drug will be delivered to the target tissues and organs, and whether it will combine with the intended molecules in the cells. To demonstrate such real behaviors of a drug in a living body, it is critical to be able to observe them noninvasively. Thus, techniques are required for monitoring the dynamic states of individual molecules in the living body over long periods of time.
18.2
Features of Two-Photon Excited Fluorescence Microscopy In conventional single-photon absorption, a molecule absorbs a photon that has an energy equal to the energy difference between the ground state and the first excited state of the molecule. Although it is an extremely rare occurrence, the quantum mechanics for the interaction between light and materials also allows for the possibility of a molecule absorbing two photons simultaneously that both have approximately twice the wavelength (i.e., half the excitation energy) of the above-mentioned single photon (see Table 18.1). This is referred to as the two-photon excitation process (see Figure 18.1(a)). In turn, it is possible for a molecule to simultaneously absorb three photons that each have approximately a third of the excitation energy; this is called three-photon excitation or, more generally, multiphoton excitation (MPE). Maria Göppert-Mayer predicted MPE in 1931, at the dawn of the quantum mechanical era [3], and after she had immigrated to the United States, she was awarded a Nobel Prize in physics for her contributions to nuclear theory. However,
Table 18.1
TPEF Absorption Cross-Sections of Key Fluorochromes
Fluorochrome
TPEF Absorption Cross-Section s2 (cm4sphoton)
Emission Wavelength (nm)
Two-Photon Excitation Wavelength (nm)
Indo-1 Ca2+-bound form
1.5 × 10
405
590
423
750
450
700
490
700
-50
510
700
-50
-50
-50
Cascade blue
1 × 10
DAPI
0.16 × 10
Indo-1 free
3.5 × 10
-50
2+
Fura -2 Ca -bound form
-50
12 × 10
510
700
-49
512
920
-48
520
782
-50
533
860
-48
565
700
-48
600
840
Fura-2 free form
11 × 10
lparBodipy
1 × 10
Fluorescein
1 × 10
Lucifer yellow
1 × 10
Dil
1 × 10
Rhodamine B
2 × 10
Source: [11, 34, 35].
18.2 Features of Two-Photon Excited Fluorescence Microscopy
Figure 18.1
461
Process of multiple-photon excitation.
the photon densities required to experimentally demonstrate this theory are so high, that it took thirty years and the invention of the laser to be able to do so; MPE was experimentally demonstrated by Kaiser and Garrett in 1961 [4]. About thirty years after that, a report by Dr. Webb’s group in 1990 inspired extensive application of MPE to fluorescence microscopy in biological fields [5]. The reason for this time span is thought to be the need for near-infrared ultrashort-pulse lasers to become sufficiently stable. Comparatively high laser powers are required to induce the process of MPE. However, at such laser powers, the associated one-photon absorption causes phototoxicity and thermal damage in living cells and tissue. It is therefore essential to suppress one-photon absorption as much as possible during MPE, and this requires using ultrashort laser pulses. This is the reason for the long delay noted above. Currently, the most commonly used light sources for MPE are diode-pumped solid-state (DPSS) Ti:sapphire lasers. They supply stable and very short pulses enabling TPEF to be observed easily. Fluorescence is generated only in the tiny volume at the beam’s focus that has a volume of less than one femtoliter, because the highest density of the excitation laser light is achieved only in the focus (see Figure 18.1(b)). Hence, the excitation region is extremely localized, which is one of the most important advantages of TPEF microscopy. When a laser beam is scanned horizontally, the tissue will be excited in layers, enabling tomographic imaging to be performed. This implementation for tomographic imaging is in marked contrast to that for confocal microscopy. In confocal microscopy, tomographic images at the focal plane are obtained by using a “confocal” pinhole to select only fluorescence photons originating from the focus. This blocks both fluorescence light generated in regions outside the focal plane and light scattered by the tissue. In contrast, a TPEF microscope employs the fluorescence photons scattered deep within the sample; it does not require a confocal pinhole due to the highly localized nature of MPE. Theoretically, a confocal microscope offers better spatial resolution (0.13, 0.3 μm in the focal plane and in the axial direction, respectively) than a TPEF microscope (0.3, 1 μm in the focal plane and in the axial direction, respectively), but in actual use, a confocal microscope must sometimes be used at a lower resolution
462
Two-Photon Microscopy for In Vivo Analysis of Neural and Secretory Activities
than its theoretical limit. This is because the number of the fluorescence photons that pass through the pinhole drops off rapidly as the focal plane becomes deeper within the sample; a larger pinhole must then be used in order to obtain a sufficiently strong fluorescent signal. Thus, there is a trade-off between pinhole size and spatial resolution in confocal microscopy; in other words, enlarging the pinhole sacrifices resolution. The intensity of the excitation laser beam cannot be increased, since this would cause serious damage to the cell or the tissue. In contrast, the photophysical and chemical properties of the MPE process indicate that TPEF offers a number of potential advantages for analyzing physiological functions (Figure 18.2). 18.2.1
Deep and Benign Observations
The excitation lasers used (Ti:sapphire, Nd:YLF, and so forth.) emit in the near-infrared region, their laser beam is absorbed and scattered in in vivo tissues less than visible light. This means that the focus of laser beam in the deep area is not so faded or defocused in the comparison to that in confocal microscopy. This enables the excitation lasers to illuminate deeper portions of the tissue. Of course the emitted or returning light is visible and scatted in the specimen, but such scattered visible emitted light is captured by PMT in TPEF: TPEF is not required to employ a confocal pinhole to obtain a cross-sectional image as discussed above. This leads superior S/N ratio of fluorescence images in TPEF. Actually, in the case of in vivo imaging of living mouse brain, TPEF can visualize finer structures of neurons within deeper layers than 1 mm from the surface (see 18.4). They also cause less heat damage to cells due to absorption of radiation, allowing longer observations to be made; this consideration is important for in vivo imaging, which is discussed below.
Figure 18.2
Characteristics of two-photon excitation microscopy.
18.2 Features of Two-Photon Excited Fluorescence Microscopy
463
Another advantage is that there is less scattering of light in the background, which manifests itself as noise in the signal, and the emission wavelengths are well outside those for visible fluorescence, so that the entire signal spectrum can be used to image the tissue sample, improving the signal-to-noise (S/N) ratio. 18.2.2
Replacement for Ultraviolet Sources
Usually, BK7 and other glasses are used for manufacturing microscope objective lenses, but these glasses have low transmissions in the ultraviolet. Ultraviolet confocal microscopes were developed at one stage by replacing glass lenses with artificial quartz lenses, but, in addition to being expensive, they were plagued with performance issues. There was also a high risk of damaging the cells under the intense ultraviolet illumination. In contrast, TPEF microscopy is much easier to use and does not damage tissue while still allowing the use of the same ultraviolet indicators, because the excitation light is in the range of NIR, not in UV. This means that a fluorescent calcium indicator, fura-2, can be used. This indicator has been extensively studied and has excellent physical properties. In addition, TPEF can perform tomographic imaging using cellular autofluorescence. For example, a major component of the cellular autofluorescence is that of NAD(P)H, which is usually monitored by irradiating with UV light. We have attempted to visualize and analyze the activity of individual mitochondria using the two-photon excitation of NAD(P)H (Figure 18.3) and we have demonstrated that the rapid oxygen consumption occurring after activation of cultured CNS neurons is mediated by Ca2+-dependent activation of the mitochondria [6]. We have also succeeded in visualizing monoamine neurons (unpublished) and secretory granules by using UV autofluorescence excited by three-photon excitation [7]. Molecules that can be photoactivated by UV light can also be activated by two-photon excitation. This property is utilized in cellular and biological studies using caged compounds, photoactivated fluorescent proteins, and photoreceptor channels. Since, as noted above, two-photon excitation is confined to the focal point of the objective lens, the excitation of these photoactive molecules is similarly local-
Three-photon Zymogen granule (peptide?)
Two-photon Mitochondria (NADH)
Figure 18.3 (Color plate 29) Autofluorescent images of intact pancreatic acini obtained simultaneously by two- and three-photon excitation.
464
Two-Photon Microscopy for In Vivo Analysis of Neural and Secretory Activities
ized at the focal point. In other words, this method enables the user to localize the photoactivation to a tiny, subfemtoliter volume on the surface or in the interior of an in vivo tissue or cell sample. If this technique is employed to “uncage” a caged substance, a physiologically active substance can be administered to a very tiny region of interest at any desired time. We have also succeeded in performing functional mapping [8] and in determining the relationship between the glutamate receptor distribution and spine morphology. A variant of the green fluorescent protein (GFP) has been developed that changes from being nonfluorescing to fluorescing on exposure to UV radiation; it is an example of a photoactivable GFP (PA-GFP) [8]. This protein can also be excited by two-photon pumping at a wavelength of about 750 nm [9], enabling molecules to be labeled within a subfemtoliter volume [10]. 18.2.3
Avoiding Self-Shielding Effects and Compensating for Photobleaching
Under single-photon excitation in confocal microscopes, if the concentration of fluorescent dyes is increased greatly in order to obtain more intense fluorescent signals, more radiation is absorbed before it reaches the focal plane. The wavelengths used for two-photon excitation are, however, not absorbed to such an extent so that the beam penetrates deep into the tissue. In addition, photobleaching of dyes is restricted to the immediate vicinity of the focal point and it is readily compensated by diffusion of the dyes around the focal point. In confocal microscopes, on the other hand, compensation by dye diffusion is not very effective since photobleaching occurs in the extremely large three-dimensional volume that the excitation laser light passes through. 18.2.4
Precise, Simultaneous Multicolor Fluorescence Imaging
As discussed, a several advantages of being less invasive for intact living samples and being able to avoid self-shielding effects and to compensate for photobleaching is quite effective for thick samples or in vivo imaging. These advantages would be thought not so effective for thin living samples and in vitro cultured cells. However, it should be emphasized that TPEF’s ability to perform precise, simultaneous, multicolor fluorescence imaging is a further advantage of TPEF, which is beneficial not only for real-time in vivo imaging of thick living samples, but also for thin living samples and in vitro cultured cells. The spectrum for two-photon excitation frequently involves more than simply doubling the wavelength. It generally broadens and shifts the excitation spectrum due to the effect of factors such as blue shifting, which might be caused by the symmetry of fluorescent molecules or by contamination of three or more photons excitation process. This increases the overlap with the excitation spectra of different fluorochromes [11]. Consequently, two-photon microscopy can use a wider range of fluorochrome combinations that can be simultaneously excited by a single wavelength than can single-photon excitation processes such as those used in confocal microscopy. This expands the number of effective combinations of fluorochromes that can be used in simultaneous double-staining experiments, compared to those that can be used in confocal microscopy. For example, if a central wavelength of 830
18.3 Overview of the Optical System
465
nm is used for excitation, both the UV-excitable Ca2+ indicator fura-2 and the green-excitable dye Sulforhodamin B can be simultaneously used by broadening of those excitation spectra for tomographic imaging [7]. This means that TPEF can precisely investigate causality among morphological changes, and intracellular signaling molecules like Ca2+. By exploiting this advantage, we have successfully confirmed Ca2+-dependent cell functions such as exocytosis, electrolyte transport in exocrine glands, and mitochondrial activation in the hippocampal neuron [6, 7, 12–14]. Here, we must emphasize that the increased number of dyes that can be used simultaneously releases the researcher from problems associated with parallax and chromatic aberration. This represents an even more significant advantage over confocal microscopy. As discussed above, confocal microscopes employ pinholes to improve their spatial resolution by eliminating the signal due to light scattering in the sample and to fluorescence originating from areas outside the focal point of the objective lens. As a result, their confocal resolution inevitably depends on the emission wavelength, due to chromatic aberration of the objective lens and other optical components. By contrast, fluorescence in two-photon microscopes is limited to a single point, the focal point, so that the use of pinholes is not required. It follows that fluorescence photons of different wavelengths emitted by different fluorescent molecules originate from essentially the same spatial point, so they provide information about a single spatial location. Since there are no concerns about chromatic aberration or parallax, it is possible to have a rigorous discussion about the spatial distribution of multiple kinds of fluorescent molecules at a single location. This also applies to thin samples visualized in vitro and to cultured cells. It represents the most important benefit of TPEF.
18.3
Overview of the Optical System Our optical system consists of three DPSS near-infrared ultrashort-pulse lasers with different tunable ranges, and three microscopes (an Olympus IX70 inverted microscope and two Olympus BX50/WI upright microscopes) that can be used in any desired combination (Figure 18.4(a)). However, to simplify the explanation, in this study just one of the lasers and one of the microscopes are used (Figure 18.4(b)). We describe the system by starting with light generation. At first, the pump light is provided by a diode-pumped solid-state green laser (Spectra Physics, Millenia). In this laser, near-infrared laser light is generated using a semiconductor laser diode (LD). This light is used to pump Nd:YVO4 to generate a coherent laser beam with a wavelength of 1,064 nm. A second harmonic of this beam is generated using a beta barium borate (BBO) crystal, which is a nonlinear optical material, providing a 532-nm beam with a power in the range 5 to 10W. This green pumping beam is directed into an actively mode-locked titanium-sapphire (Ti:Al2O3) laser (Spectra Physics, Tsunami), which generates a femtosecond near-infrared pulse. A nonlinear optical element inserted in a resonant cavity is driven in an appropriate way by an exterior source so as to modify the optical path. This sets up a standing wave with a desired spectrum inside the oscillator, and by the principle of superposition, results in femtosecond pulses being output as
466
Two-Photon Microscopy for In Vivo Analysis of Neural and Secretory Activities
(a)
(b)
Figure 18.4
TPEF microscope system: (a) photo, and (b) overview.
a pulse train with a high repetition rate. Next, the near-infrared pulse is fed through a chirp compensator and then through a neutral density filter to modify the power of the beam. The beam diameter is then adjusted and the beam enters the laser scanning unit. We now consider the individual components of the system. Employing an optical system that has chirp compensation is critical for in vivo TPEF imaging as illustrated in Figure 18.4(b). To carry out long-time observations in living tissue, it is important to reduce the intensity of the excitation laser beam. As mentioned above, it is essential for the illumination pulses to be as short as possible in order to maximize the probability of simultaneous absorption of two photons. Femtosecond
18.4 In Vivo Imaging of the Cerebral Neocortex
467
pulses tend to disperse when they pass through the glass of the microscope lenses and other elements due to the nonlinear effect of group velocity dispersion in the optical materials. In such positively chirped materials, components with shorter wavelengths propagate faster than those with longer wavelengths (an effect known as positive chirping), thus increasing the pulse width. The chirp compensating optical system shown in Figure 18.4(b) was used to apply a negative chirp (i.e., delay the arrival of the shorter-wavelength components relative to that of the longer-wavelength components) to produce a pulse that is as narrow as possible when it enters the tissue sample. The compensator is tuned to cause the shorter-wavelength components to “catch up” with the longer-wavelength components just when they reach the focal plane inside the specimen. This ensures that the final incoming packet is chirp-free and is as narrow as possible. The beam expander is designed so that the beam completely fills the back aperture of the objective lens so that the highest possible numerical aperture is obtained. As mentioned earlier, this condition must be fulfilled in order to obtain the tightest focus of the laser beam in the sample; it is thus a requisite condition for extracting the highest possible resolution from the microscope. In order to achieve this condition, the objective lens was replaced with a water-immersion lens that had a high transmissivity for near-infrared radiation, a low chromatic aberration, and superior performance in focusing off-axis objects. Also, the dichroic mirror (DM) in the scanning unit was replaced with a specially designed reverse DM, that, in the opposite manner to conventional DM, reflects long wavelengths and passes visible wavelengths; this was done to collect MPE fluorescence photons. We have two sets of reverse DMs, one for MPE and the other for confocal microscopy, allowing us to use the confocal microscope in combination with another microscope. We also used an external photomultiplier detector, since it is more efficient at detecting weak fluorescence than an internal photomultiplier detector.
18.4
In Vivo Imaging of the Cerebral Neocortex In vivo TPEF microscopy is used for imaging processes in individual living bodies; it takes full advantage of the ability of TPEF to image deep within the sample without damaging the tissue. Here, “in vivo imaging” refers to observing the interior of intact, living organs of animals. There have been previous attempts at imaging individual living bodies, but conventional TPEF systems only image to depths of several hundreds of microns and their spatial resolution is not very good. A previous report has described an attempt to image deep layers of a living cerebrum about 1,000 mm beneath the surface using a laser regenerative amplifier [15]. Even if the infrared light is rarely scattered or absorbed by the tissue, propagation over such long distances could reduce the probability of MPE occurring. Thus, the authors attempted to increase the height of pulse to compensate for the height reduction by using the regenerative amplifier. However, this reduced the repetition rate of the pulse train from the regenerative amplifier, prolonging the time required to obtain the image. Another problem was that it reduced the wavelength selectivity. We now have used a very stable Ti:sapphire laser, which we recently received from the manufacturer,
468
Two-Photon Microscopy for In Vivo Analysis of Neural and Secretory Activities
in the hope of developing a system that is capable of performing in vivo imaging without using a regenerative amplifier. The TPEF microscopy laboratory at the National Institute for Physiological Sciences (NIPS) in Okazaki, Japan, has a functional TPEF system, consisting of three ultrashort-pulse lasers having different characteristics, one upright and one inverted microscope for TPEF tomographic imaging, and one inverted microscope for activating photo-activatable molecules (see Figure 18.5). Recently, we succeeded in imaging neurons deep within the cerebral cortex of an anesthetized mouse by using the upright microscope (Figure 18.5). Currently, it is possible to observe neurons at depths of over 0.9 mm from the surface of the cerebral cortex. This newly developed upright microscope enables us to obtain a complete picture of a living animal’s neuron, which includes widely spread branches, dendrites, and axon fibers, in a three-dimensional pattern. This development also enables fine structures such as submicrometer as dendritic spines, boutons, and axon terminals to be imaged at high spatial resolutions; thus, it provides the foundation for developing a method to monitor the long-term changes in a neuron of a living body over a month. We have been informed by some optical microscope manufacturers that this system is probably the only one of its kind in the world. Unlike the typical specially built systems in overseas research laboratories, this was assembled by simply arranging the components, but it is nonetheless capable of obtaining high-resolution images deep within a sample. One of the reasons why optical manufacturers have assessed this system as being world-class is that it has already taken some of the deepest images in the world. Another significant point is that, whereas other research groups (both inside and outside Japan) are developing TPEF microscopes
Figure 18.5 (Color plate 30)
In vivo images of the cerebral neocortex of an anesthetized mouse.
18.5 Imaging of Secretory Functions
469
for in vivo observations require high laser powers of over 500 mW, our images were obtained using a laser power of only 30 mW, which is less than a tenth the output power of the lasers used by other groups. Our new TPEF microscope thus has quite a high sensitivity, suggesting that it has low invasiveness for in vivo imaging. We are currently using this system to study the dynamic states of neurons and glial cells. Our current findings are only preliminary, but it has become increasingly more apparent that living samples undergo far different reactions from those observed in dead tissues. It would be appropriate to discuss the factors that have made it possible to obtain these high-quality images deep within samples. First of all, the optical system was designed to be very simple, for the specific purpose of observing tissues in vivo. A specialized objective lens that was optimal for that purpose was selected and the system was adjusted to illuminate the sample to a deep level and to reach the focal point with the narrowest beam possible. Another key component was the MaiTai HP laser (Spectra Physics), which has a long output wavelength of over 900 nm. This laser is almost completely automatic; it is also stable, and provides a high output power. As Dr. Takeharu Nagai of Hokkaido University and Dr. Atsushi Miyawaki of Riken [16] have shown, GFP-altered fluorescence protein is anticipated to be widely used in the future; its fluorescence is brighter than that of EGFP and its emission wavelength is longer, improving the S/N ratio of the image [17]. Advances in laser technology will be the key to improving TPEF imaging. Lastly, we have succeeded in developing a set of special adaptors for operating on mice and immobilizing them on the microscope stage. These adaptors have enabled us to perform continuous observations and to obtain good images over long periods of time. The methods for obtaining these samples were developed on the basis of the excellent suggestions by a collaborator, Dr. Hiroaki Wake of NIPS, and the observations were the result of his patient efforts. His diligence is an inspiration to believe in one’s observations and to persevere with one’s efforts. So far, we have confirmed that this in vivo imaging method can be applied not just to neurons but also to neuroscience objects and to microglial cells. Researchers at NIPS are combining TPEF imaging with a noninvasive method to produce infarction in the living brain using oxygen radicals released by fluorochromes, for investigating the recovery processes from a cerebral infarction, or ischemia. Some researchers are attempting new methods to image blood vessels in the brain and to measure the blood flow rate [18]. Others are experimenting with ways to apply this system to other tissue, including lymph nodes and tumors [19]. Most of these efforts are still at the animal testing level, but some human trials have already been initiated, and there is great interest in the advent of clinical applications for TPEF imaging.
18.5
Imaging of Secretory Functions We now turn to look at other examples of applications to cellular biology performed by our research team. Exocytosis is one of the processes by which a biosynthesized substance in a cell is excreted outside the cell. It is used by a wide variety of tissues, including the release of neurotransmitters by neurons and the
470
Two-Photon Microscopy for In Vivo Analysis of Neural and Secretory Activities
release of digestive enzymes. In exocytosis, a granule or vesicle is formed around the substance to be excreted; the vesicle then fuses with the cell membrane, and the contents of the vesicle are ejected through the membrane. We have taken noninvasive living images of the pancreatic exocrine gland (see Figure 18.3). These enabled us to perform a qualitative analysis of the spatial distribution of changes in the Ca2+ concentration in the cells and to visualize a single phenomenon that occurred in the cell. In addition, we were the first in the world to demonstrate the existence of the sequential compound exocytosis in the pancreatic exocrine gland [7]. Sequential compound exocytosis has subsequently been confirmed in a variety of systems having excretory functions [12, 13, 20, 21], and appears to be a very common phenomenon [22]. We have also succeeded in developing a method to analyze the dynamic state of vesicles during endocytosis [23]. The diameter of vesicles in the endocytotic process has been estimated as being less than 100 nm, making it impossible in principle to resolve individual vesicles, since they are smaller than the diffraction limit of the light. Taking advantage of the possibility for simultaneous double staining and the almost complete absence of parallax and aberration in TPEF, a method has been developed to use this technology to accurately measure vesicle diameters. This method works on the fact that the ratio of two different color fluorescent signals from the vesicle surface and from the vesicle inner space is directly proportional to the vesicle radius. This method allows the observer to estimate whether the diameter of a vesicle in a pixel is 50 or 100 nm. We have shown that endocytotic vesicles having different diameters perform different physiological functions [24]. This first demonstration that the diversity in tiny vesicles observed under the electron microscope corresponds to the diversity in their actual physiological functions was enabled by the ability of TPEF microscopy to image events in living cells. This is a particularly compelling example of the importance of imaging of live cells. We have also successfully imaged Ca2+-dependent water and electrolyte transport and it is expected that this method can be used to investigate allergic rhinitis [14]. In a recent extension of this method to developmental biology, we have begun to obtain some new findings in the cellular biology of embryonic differentiation.
18.6
Future Possibilities It is anticipated that many research institutes, universities, and other institutions will acquire TPEF microscopes as they gain a more important role in the medical and life science fields. The NIH has allocated a sizeable budget for bio-imaging for several years and hundreds of ultrashort pulse lasers have been sold in the United States for use in TPEF microscopes. The main issue in the adoption of these microscopes is their expense; the ultrashort pulse laser itself accounts for over half of the purchase price of the system. This may be quickly ameliorated by the imminent advent of inexpensive replacements for currently used titanium-sapphire lasers. This brings us to the subject of future developments in TPEF microscope technology. Reports have already begun to describe supercontinuum lasers, which can deliver several-hundred-nanometer ultrabroadband femtosecond pulses, and applications of these lasers to TPEF microscopy. They have the potential to greatly
18.6 Future Possibilities
471
expand the range of phenomena observable with these microscopes [25]. More recently, microscopes that are based not on MPE but on nonlinear optical processes such as second harmonic generation (SHG) and stimulated emission depletion (STED) have begun to creep into medical and biological laboratories [26, 27]. The shock waves associated with coherent ultrashort pulses are also being developed into a tool for cell manipulation under the microscope, such as transferring genes into target cells [28]. The next stage of in vivo TPEF microscopy will be the application to conscious animals. A published study describes a system with optical fibers and a miniaturized laser scanner [29]. This technology will directly contribute to connecting TPEF techniques with endoscopy, and has the potential to introduce dramatic changes in surgical methods, which now depend on optical fibers. A report described an in vivo experiment with a conscious animal using miniaturized laser scanning mirrors that had been created with microelectromechanical systems (MEMS) technology [29]. The development of ultraminiaturized, lightweight lasers with fiber laser technology has also been reported, and it is now realistic to think in terms of bedside use of TPEF microscopy. Researchers are reporting about TPEF microscopes that incorporate these technologies. To mention just one example of a clinical application: medical oncologists are very interested in their potential for treating tumors. What we believe is even more important is the potential for noninvasive methods of stimulating cells with coherent light, as is currently done with the above-mentioned caged reagents. Beyond handling of organic compounds such as caged substances, there have recently been reports of cloning of the light receptor channels of Chlamydomonas. These channels are non-selective cation channels and are activated by irradiation with blue light. This discovery of how to cause this phenomenon in neurons has enabled researchers to artificially stimulate the neurons to the excited state using light [30]. There are also great expectations of applications for photoactivated cAMP synthetase and photoactivated adenylyl cyclase obtained from Euglena proxima [31]. At the time of writing, no publications have described successful activation of these channels using the TPEF process, but when that becomes a reality, it will have great promise as a tool for researching neural networks and the cell function. Thus, there is widespread anticipation for noninvasive optical methods for analyzing cellular and physiological functions in both basic and clinical medicine. Much work remains to be done before TPEF microscopes will be ready for everyday use. For example, even “completely automated” ultrashort pulse lasers require some fine adjustments every time they are powered up. Laser manufacturers and microscope manufacturers need to establish a serious cooperative program in order to resolve this issue. Optical microscopes are essential in biological research. Are the basic techniques of using optical microscopes being sufficiently conveyed to the next generation? I am troubled by the tendency for young researchers to think that the only instruments they require skills for are pipettes and genetic engineering kits. NIPS conducts lectures in its summer training course in the fundamentals of optical microscopy and uses images obtained with TPEF microscope systems. Readers who are interested in these matters are cordially invited to contact the author.
472
Two-Photon Microscopy for In Vivo Analysis of Neural and Secretory Activities
Problems 18.1 What determines the spatial resolutions of TPEF microscopes and confocal microscopes? Discuss this, referring to appropriate mathematical equations. 18.2 What is the reason for the superior images obtained by TPEF microscopes of living tissue? 18.3 What are some possible ways TPEF microscopes can be used to study the interaction of nanoparticles with living tissue? Search the current literature to get some ideas.
Acknowledgments I would like to express my deep gratitude to Professor Haruo Kasai for his many suggestions about Ca2+ imaging with the TPEF microscope, sequential compound exocytosis, and countless other matters. I am also deeply indebted to Faculty Head Prof. Yasunobu Okada, Prof. Junichi Nabekura and many other members of NIPS, without whose moral and financial support I would never have been able to assemble a world-class TPEF microscope. These studies were supported by Grants-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan, by the Japan Science and Technology Agency, and by the Naito Foundation.
References [1] Denk, W. and K. Svoboda, “Photon upmanship: why multiphoton imaging is more than a gimmick,” Neuron, 1997. 18(3): p. 351-7. [2] Matsuzaki, M., et al., “Structural basis of long-term potentiation in single dendritic spines” Nature, 2004. 429(6993): p. 761. [3] Goppert-Mayer, V.M., “Uber Elementarrakte mit zwei Quantensprungen,” Ann Phys, 1931. 9: p. 273-294. [4] Kaiser, W. and C.G.B. Garret, “Two-photon excitation in CaF2:Eu2+,” Phys Rev Lett, 1961. 7: p. 229-231. [5] Denk, W., J.H. Strickler, and W.W. Webb, “Two-photon laser scanning fluorescence microscopy,” Science, 1990. 248(4951): p. 73-6. [6] Hayakawa, Y., et al., “Rapid Ca2+-dependent increase in oxygen consumption by mitochondria in single mammalian central neurons,” Cell Calcium, 2005. 37: p. 359-370. [7] Nemoto, T., et al., “Sequential-replenishment mechanism of exocytosis in pancreatic acini,” Nat. Cell Biol., 2001. 3: p. 253-258. [8] Patterson, G.H. and J. Lippincott-Schwartz, “A photoactivatable GFP for selective photolabeling of proteins and cells,” Science, 2002. 297: p. 1873-1877. [9] Schneider, M., et al., “Two-Photon Activation and Excitation Properties of PA-GFP in the 720-920-nm Region,” Biophys. J., 2005. 89(2): p. 1346-1352. [10] Hess, S.T., T.P.K. Girirajan, and M.D. Mason, “Ultra-High Resolution Imaging by Fluorescence Photoactivation Localization Microscopy,” Biophys. J., 2006. 91(11): p. 4258-4272.
References
473
[11] Xu, C., et al., “Multiphoton fluorescence excitation: New spectral windows for biological nonlinear microscopy,” Proc Natl Acad Sci (USA), 1996. 93(20): p. 10763-10768. [12] Kishimoto, T., et al., “Vacuolar sequential exocytosis of large dense-core vesicles in adrenal medulla,” Embo J, 2006. 25(4): p. 673-82. [13] Kishimoto, T., et al., “Sequential compound exocytosis of large dense-core vesicles in PC12 cells studied with TEPIQ (two-photon extracellular polar-tracer imaging-based quantification) analysis,” J Physiol (Lond), 2005. 568(3): p. 905-915. [14] Oshima, A., et al., “Two-photon microscopic analysis of acetylcholine-induced mucus secretion in guinea pig nasal glands,” Cell Calcium, 2005. 37: p. 359-370. [15] Theer, P., M.T. Hasan, and W. Denk, “Two-photon imaging to a depth of 1000 micron in living brains by use of a Ti:Al2O3 regenerative amplifier,” Opt Lett, 2003. 28(12): p. 1022-4. [16] Nagai, T., et al., “A variant of yellow fluorescent protein with fast and efficient maturation for cell-biological applications,” Nat Biotechnol, 2002. 20(1): p. 87-90. [17] Miyawaki, A., “Innovations in the Imaging of Brain Functions using Fluorescent Proteins,” Neuron, 2005. 48(2): p. 189-199. [18] Nishimura, N., et al., “Targeted insult to subsurface cortical blood vessels using ultrashort laser pulses: three models of stroke,” Nat Methods, 2006. 3(2): p. 99-108. [19] Miller, M.J., et al., “Two-Photon Imaging of Lymphocyte Motility and Antigen Response in Intact Lymph Node,” Science, 2002. 296(5574): p. 1869-1873. [20] Leung, Y.M., et al., “Visualization of sequential exocytosis of rat pancreatitc islet beta Cells,” Biochem Biophys Res Com, 2002. 292: p. 980-986. [21] Hafez, I., A. Stolpe, and M. Lindau, “Compound exocytosis and cumulative fusion in eosinophils,” J. Biol. Chem., 2003. 278(45): p. 44921-44928. [22] Pickett, J.A. and J.M. Edwardson, “Compound Exocytosis: Mechanisms and Functional Significance,” Traffic, 2006. 7: p. 109-116. [23] Kasai, H., et al., “A new quantitative (two-photon extracellular polar-tracer imaging-based quantification (TEPIQ)) analysis for diameters of exocytic vesicles and its application to mouse pancreatic islets,” J Physiol, %R 10.1113/jphysiol.2005.093047, 2005. 568(3): p. 891-903. [24] Liu, T.-T., et al., “Exocytosis and endocytosis of small vesicles in PC12 cells studied with TEPIQ (two-photon extracellular polar-tracer imaging-based quantification) analysis,” J Physiol (Lond), 2005. 568(3): p. 917-929. [25] Kano, H. and H. Hamaguchi, “Supercontinuum Dynamically Visualizes a Dividing Single Cell,” Anal. Chem., 2007. 79(23): p. 8967-8973. [26] Campagnola, P.J., et al., “High-resolution nonolinear optical imaging of live cells by second harmonic generation,” Biophys J, 1999. 77: p. 3341-3349. [27] Willig, K.I., et al., “Nanoscale resolution in GFP-based microscopy.” Nat Methods, 2006. 3(9): p. 721-3. [28] Tirlapur, U.K. and K. Konig, “Cell biology: Targeted transfection by femtosecond laser,” Nature, 2002. 418(6895): p. 290-291. [29] Piyawattanametha, W., et al., “Fast-scanning two-photon fluorescence imaging based on a microelectromechanical systems two-dimensional scanning mirror” Opt Lett, 2006. 31(13): p. 2018-20. [30] Boyden, E.S., et al., “Millisecond-timescale, genetically targeted optical control of neural activity,” Nat Neurosci, 2005. 8(9): p. 1263-1268. [31] Iseki, M., et al., “A blue-light-activated adenylyl cyclase mediates photoavoidance in Euglena gracilis,” Nature, 2002. 415(6875): p. 1047-1051.
474
Two-Photon Microscopy for In Vivo Analysis of Neural and Secretory Activities [32] Born, M. and E. Wolf, Principle of Optics, 1974: Pergamon Press. [33] Wilson, T. and C. Sheppard, Theory and Practice of Scanning Optical Microscopy. 1984, London: Academic Press. [34] Lakowicz, J.R. and I. Gryczynski, Topics in Fluorescence Spectroscopy, Vol. 5. 1997. [35] Xu, C. and W.W. Webb, “Measurement of two-photon excitation cross sections of molecular fluorophores with data from 690 to 1050 nm,” J. Opt. Soc. Am., 1996. B. 13: p. 481.
CHAPTER 19
Nanoscale Engineering of Electrodes, Biosensors, and Protein Surfaces Yeoheung Yun, Laura Conforti, Zhongyun Dong, Vesselin N. Shanov, Joon Sub Shin, William Heineman, H. Brian Halsall, Wei Feng, Chong H. Ahn, and Junsang Doh
19.1
Introduction Nanotechnology has attracted tremendous interest from scientists in the fields of physics, chemistry, and materials science [1, 2]. But biomedical applications of nanotechnology (called bionanotechnology) have an especially large potential to benefit society. Three applications of bionanotechnology can be categorized as: (1) biomedical diagnostic techniques, (2) new drug delivery systems, and (3) prostheses and implants. Biomedical applications for use outside of the body, such as diagnostic sensors like DNA, RNA chips, lab-on-a-chip, and micrototal analysis systems (μTASs), are suitable for analyzing biomolecular samples. For inside the body, anticancer imaging and drug and gene therapies are major growing areas. Other researchers are working on prostheses and implants that include nanostructured materials and surfaces that can improve attachment of tissue to implants and improve biocompatibility of prostheses. This chapter describes novel nanomaterials synthesis through to surface preparation and electrode development, including ways nanotechnology can support medical research by providing new nanotools such as biosensors. One of the advanced nanomaterials, carbon nanotube (CNT) arrays, is discussed in detail since CNT have high electrical conductivity, high thermal conductivity, low density, high elastic modulus, and a high length/diameter aspect ratio. Thus, electroanalytical method-driven biosensing using CNT arrays is described, which has the potential to provide a simple sensor design with a low detection limit. These highly sensitive nanoelectrodes are under development to detect not only biomolecules but also specific cells such as prostate cancer cells. Nanofabrication techniques combined with surface chemistry and materials science are providing new tools to further explore the interactions of cells with different environment conditions. Cellular behavior such as proliferation, differentiation, migration, and apoptosis can be controlled by multiple proteins on a surface, which is a new type of culture assay. This chapter includes organic nanotechnology by synthesizing novel polymers which allow spatial distribution of different proteins on glass that enable the study of cellular biology and physiology. The ability to spatiotemporally control the cell-adhering substrate provides new knowledge of
475
476
Nanoscale Engineering of Electrodes, Biosensors, and Protein Surfaces
cell-surface interactions and enables high throughput screening and will be useful for further dynamic modeling of pathways and interactions in cells.
19.2
Carbon Nanotube Electrode and Biosensor Development Different methods and fabrication techniques to make carbon nanotube electrodes and biosensors are discussed. The electrodes are designed to be able to replace conventional electrodes used in electrochemical analyses. The biosensors provide a continuous signal that can be monitored with time to determine the concentration of the analyte. 19.2.1
Carbon Nanotube Array Synthesis
Synthesis, processing, and device fabrication techniques for carbon nanotube (CNT)-based biosensor applications have recently been improved with the help of intensive research and material characterization [3, 4]. Vertically aligned arrays of CNT that have well-defined properties with uniform length and diameter are being prepared using thermally driven chemical vapor deposition (CVD) on catalytically patterned surfaces. The thermally driven CVD method uses hydrocarbon precursor molecules (CH4, C2H2, C2H4) for the carbon source, and Fe, Ni, and Co as catalysts to grow high-density arrays on Si substrates. Figure 19.1 shows the typical preparation steps for CNT growth using CVD. Briefly, P-type Si wafers with a typical resistivity of 1–20 Ohm-cm are used with a 500-nm SiO2 layer. Then, an e-beam evaporator is used to deposit an Al thin film about 10-nm thick, and the Al is oxidized to convert to Al2O3. Finally, catalytic iron or composite films of controlled thickness of 1–2 nm are deposited on the surface of the Al2O3 surfaces. The efficient CVD synthesis of CNT where the activity of the catalyst is enhanced by water during the synthesis is demonstrated since the water-stimulus oxidizes amorphous carbon without significantly damaging the nanotubes [3]. Growth parameters including flow rates, water vapor delivery, temperature, time, catalyst type, and film thickness on the Si wafer should be carefully considered. Environmental scanning electron microscopy (ESEM), high-resolution transmission electron microscopy (HRTEM), and atomic force microscopy (AFM) before and after MWCNT growth provide understanding of the growth kinetics and relationships between different catalysts and the CNT array morphology including nanotube diameter and length. Synthesis of CNT was described in Chapter 2 in detail. 19.2.2
Carbon Nanotube Array Electrode
Here, we present a novel fabrication method for CNT array electrodes, followed by their electrochemical characterization for electroanalytical purposes [6–10]. The most difficult problems faced in the electrode fabrication were achieving repeatability and low contact resistance. In this work, we report a direct soldering method of attaching the nanotube array to a prepatterned printed circuit board (PCB) to overcome contact limitations. Figure 19.2(a) shows the schematic diagram and fabrica-
19.2 Carbon Nanotube Electrode and Biosensor Development
477
Si/SiO 2
Step 1 Deposition of Al or Al2O3
Al2O3
Step 2 Deposition of Fe
Fe Step 3 Thermal treatment 400°C for 5 hours Fe Oxide Step 4 CNT growth Step 5
Step 6
Figure 19.1
CNT
Semiconductor processing from substrate preparation to CNT synthesis.
tion process for the nanotube electrode (tower 1-mm diameter). This procedure is simple and mass production is possible. First, the CNT tower was peeled off from the Si substrate. The CNT tower was placed on a Cu patterned printed circuited board (PCB). To assemble the CNT tower onto the PCB, heat was applied to the CNT tower and to a patterned Cu film causing the solder to melt and be drawn into the CNT array by wetting action. The 8-mm long CNT tower was successfully soldered onto the PCB. After cooling to room temperature, the whole component was immersed in epoxy. After casting in epoxy, the top section of a CNT tower was polished using a Vibromat polisher (Buehler). After polishing the nanotube array–epoxy composite, reactive ion etching (RIE) was used to remove excess epoxy and open the nanotube tips. An example of an electrode fabricated using soldering and plasma treatment is shown in Figure 19.2(b). Conductive epoxy (Epotek H20E) can be used as an alternative to solder. In order to evaluate the electrochemical properties of the CNT electrode, cyclic voltammetry (CV) was carried out as shown in Figure 19.3. The CV peak-separation (ΔEp) of Fe3(CN)63-/4- is very sensitive to the electronic properties and surface chemistry (i.e., hydrogen or oxygen functional groups on the carbon material-based electrodes). We can characterize the nanotube as a “rolled-up” structure of graphite with basal and edge planes. Therefore, Fe(CN)63-/4- is useful for probing the open-ended nanotube electrode kinetics. Figure 19.3 shows the CVs for
478
Nanoscale Engineering of Electrodes, Biosensors, and Protein Surfaces Patterning Cu using PCBs
Peeling off from substrate
Soldering nanotube onto Cu
Casting into epoxy
Polishing the top
(a)
(b)
Figure 19.2 (a) Nanoelectrode fabrication process using a CNT tower electrode. (b) Nanotube electrode: three CNT array electrode posts are soldered onto a PCB,and ESEM image of the CNT electrode after encasing in epoxy and treating with RF plasma. Tips of the CNT are protruding from the epoxy [6].
(a)
(b)
Figure 19.3 Cyclic voltammetry of 6 mM K3Fe(CN)6 in 1.0 M KNO3 using the nanotube tower electrode with various scan rates: (a) 100, 50, 20, 5 mV/s, and (b) 0.5, 1, 2, 5 V/s. A Bioanalytical Systems (BAS, West Lafayette, IN) electrochemical analyzer was operated by an Epsilon system. A platinum wire and an Ag|AgCl wire were used as the auxiliary and reference electrodes, respectively [5–7].
the reduction of 6.0 mM Fe3(CN)6 (in 1.0M KNO3 as a supporting electrolyte) at a nanotube tower electrode obtained at different scan rates. A sigmoidal-shaped
19.2 Carbon Nanotube Electrode and Biosensor Development
479
voltammogram was observed when a scan rate up to 1 V/s was used, indicating that the mass transport was dominated by radial diffusion. This high speed voltammetry might be useful to evaluate the larger rate constants of rapid heterogeneous/homogeneous reactions. Also, the sigmoidal feature of these CV results suggests that the individual diffusion fields of the electrodes remained isolated at this scan rate, and a state of radial diffusion was attained. The edge plane of the open-ended nanotube structure provides effective electrocatalytic sites since electron transport through the edge plane is primarily responsible for the conductivity. However, the current did not vary linearly with the square root of the scan rate and the absolute anodic and cathodic peak current ratios Ipc/Ipa were approximately unity for scan rates between 5 mV/s and 5 V/s. This phenomenon explains the radial diffusion of nanoelectrodes. Radial diffusion of the nanotube tower electrode gives rise to advantages over macroelectrodes such as the increase of sensitivity and lower detection limit. The array nanotube electrode uses the tips of thousands of nanotubes as a sensing surface [6–10]. The tips of the nanotubes provide radial diffusion characteristics and can be functionalized to be selective to desired analytes. 19.2.3
Individual Carbon Nanotube Electrode
Even though there have been many of reports about the excellence of CNTs in terms of electrical, thermal, and mechanical respects, the difficulty in manipulation of a nanotube due to the tiny size is a major hurdle for achieving its potential [11]. Also, a controllable method for assembling individual nanotubes between electrodes is required to achieve a highly sensitive nano-biosensor with precise measurements. There have been various attempts to align CNTs such as utilizing dielectrophoretic force (DEP), and chemical templates. However, DEP alignment is limited because metallic nanotubes are attracted by much higher DEP force than semiconducting nanotubes. Also, the chemical template method can change the electrical properties of the CNT. To overcome these problems, fluidic self-assembly of individual CNT’s was achieved by magnetically attracting the catalyst particle inside one end of the CNT to a desired location. As described in Chapter 2, the synthesis of CNT by thermal CVD requires a metal catalyst for the seeding site of the nanotube. During the CVD process, the metal catalyst is a small island on which the carbon nanotube grows by dissolving carbon atoms in the catalyst and precipitating the carbon atoms to form a nanotube. As a result of the base or tip growth, the catalyst is located at the one end of a long cylindrical nanotube. The commonly used catalysts for CVD synthesis are Iron (Fe), Nickel (Ni), and Cobalt (Co), which have a ferromagnetic property. Due to the ferromagnetic property of these metals, the catalysts can be magnetized and manipulated by applying a magnetic force. This common structure of a long cylindrical shape with a magnetizable head motivates a way for handling the CNT by attracting the metal catalyst and aligning the CNT parallel with a surrounding flow. The suggested technique is illustrated in Figure 19.4(a). First, Ni patterns on a substrate to induce a magnetic force were fabricated using electron beam (e-beam) lithography. A PDMS microchannel fabricated by soft lithography was attached to the Si/SiO2 substrate to guide the solution
480
Nanoscale Engineering of Electrodes, Biosensors, and Protein Surfaces
(a)
(b)
Figure 19.4 (a) Fluidic assembly of carbon nanotubes by magnetic attraction of a catalyst in the end of the nanotube. (b) Individual carbon nanotube alignment on Ni/gold electrodes. The bar is 1 μm.
containing the nanotubes to the Ni patterns. The induced magnetic force by the Ni pattern attracts the metal catalyst, and the nanotube is finally aligned by fluidic shear force. Sequential assembly of individual nanotubes on a Ni pattern on a gold electrode with a gap of 1 μm between electrodes was achieved as shown in Figure 19.4(b). This provides the feasibility of aligning CNT with small intervals between each electrode pair, leading to increased circuit density and complexity. This method can be used for biosensor development to achieve a low detection limit. An advantage of magnetic alignment as compared to the DEP method is that the electrical connection at each electrode to apply the voltage is already provided, and thereby improves design flexibility. The individual nanotube electrode uses the side of the nanotube as a sensing surface. Parallel nanotubes can provide many sensors in a small planar surface area. A biosensor using this electrode is currently being fabricated. 19.2.4
Carbon Nanotube Array Biosensor
Compared to traditional electrochemistry methods such as cyclic votammetry, electrochemical impedance spectroscopy (EIS) is a more recent engineering method used to develop biosensors. Based on the CNT array electrode, a label-free immunosensor is described here that uses EIS for sensing. The nanotube array electrode was electrochemically activated to open the nanotube ends and to expose COOH groups on the surface. Antimouse IgG was then covalently immobilized on the nanotube array as shown schematically in Figure 19.5(a). EIS were used to characterize the binding of mouse IgG to its specific antibody already immobilized on the nanotube electrode surface. Figure 19.5(b) shows the sequential impedance plots obtained for the Fe(CN)6 3– /Fe(CN)6 4- reaction at a bare nanotube electrode, after an antimouse antibody was immobilized at the bare electrode, and after an antimouse IgG was bound to the immobilized antibody. In Figure 19.7(b), the diameters of the respective semicircles were observed to increase, indicating an increasing electron transfer resistance for the Fe(CN)6 3-/Fe(CN)6 4- redox reaction following the two immobilization steps. These EIS results are also typical Randles’s circuit responses [12]. The diameter of the semicircle corresponds to the electron transfer resistance, Ret, which is directly related to the concentration of antigen. Thus EIS can be used for a new type of
19.2 Carbon Nanotube Electrode and Biosensor Development
(a)
481
(b)
Figure 19.5 CNT Array immunosensor. (a) schematic of biofunctionalized electrode, and (b) EIS results for the carbon nanotube array. In (b) (a) Bare electrode, (b) with immobilization of antibody, and (c) with antigen.
biosensor by measuring electron-transfer resistance of a CNT electrode. Improvement of the functionalization of the electrode is underway to improve the sensitivity of the sensor. How the size of the analyte affects the sensitivity is also being investigated. The potential challenge of CNT immunosensor is to prevent electrodes from nonspecific binding or adsorption of multiple proteins. Quantitative study for cross interference with similar structural proteins and further test with blood is needed. As a long-term use as implanted device, the evaluation of thrombosis, coagulation, platelets and platelet function, hematology and immunology is critical important for the interaction with blood. 19.2.5 Initial Development of a CNT Array Electrode for Prostate Cancer Cell Detection
There are many important possible applications for biosensors including detecting markers of disease, microbes, virus, biomolecules, and cells. Perhaps the most important socioeconomic goal is to fight cancer. In 2007 there were 1.4 million new cases of cancer and 550,000 deaths from cancer in the United States. Thus cancer is the second deadliest disease, after heart diseases, in the United States [13]. These statistics are similar around the world. Mortality rates for heart diseases have dropped by more than half from 1950 to 2004, and other major disease categories show similar trends, but cancer death rates have stayed pretty much the same [13]. The economic cost associated with cancer reached $74 billion in the United States in 2005, while the overall economic costs (including loss of economic output due to days off and premature death) were estimated to be over $200 billion per year (2005 data). Taking advantage of nanotechnology to develop biosensors may help to relieve suffering and death due to cancer. Fighting cancer involves four phases: (1) prevention, (2) detection, (3) treatment, and (4) monitoring. Biosensing can come into play in phases (2) through (4). This is discussed next. Detection at the earliest stage is critical because cancer has a logarithmic growth rate. Early detection provides more options for treatment and the greatest chance of survival. A tumor that is one cubic centimeter in size may have 40 to 50 cell divi-
482
Nanoscale Engineering of Electrodes, Biosensors, and Protein Surfaces
sions and typically we don’t see 80% of the life of a tumor. Detection can be done using a number of techniques—including standard immunoassays and biopsies such as colonsoscopy, mammography, CA-125, and PSA—but most detection occurs at an advanced stage of growth. The more techniques are available and approved by FDA, the more chance to survive. Nanotechnology is offering new detection approaches including targeted contrast agents, nanoscale cantilevers coated with antibodies against tumor markers, and magnetic nanoparticles coated with DNA labeling. Several advances are described in other chapters in this book. But overall, the problem is daunting because there are 50 common types of cancer and in practice it is difficult to ask people to come to the clinic on a regular basis and to screen for many types of cancer. Early detection may require identifying 10 cancer cells within one billion normal cells in a blood sample. Physicians ideally would like a quick portable device that can tell whether cancer is there or not. A nanotube electronic biosensor might help in the detection of cancer by providing a low-cost test that can be done at regular intervals. One approach for a cancer sensor is described next. This electronic approach is more complex than the filtering approach discussed in Chapter 3. The electronic approach may allow an at-home sensor to be developed. Recently, cancer detection techniques are improving. In the area of prostate cancer, more patients are being initially diagnosed with localized prostate cancer and fewer patients are being diagnosed with disseminated disease. A strategy to increase the likelihood of detection before cancer spreads is to use the biosensor to detect the concentration of cancer cells in the bloodstream and hence infer the metastatic potential of the cancer. Since the type of cancer is often known, the biosensor can be functionalized to capture a particular type of cell. As a preliminary study, the nanotube array electrode described is used to measure impedance of prostate cancer cells. Figure 19.6 shows a cross section of the final fabricated device for electrochemical impedance measurement to detect cells in solution. The upper layer of a fluid channel was made using PDMS (Sylgard 184 PDMS, Dow Corning). A Pt electrode (0.5 mm diameter) was inserted through a hole in the PDMS and sealed using
(a)
(b)
Figure 19.6 Schematic representation of a microfabricated flow cell with a carbon nanotube array electrode: (a) Cross section view of fluidic channel with nanotube electrode, and (b) overall setup for the flow cell experiment.
19.2 Carbon Nanotube Electrode and Biosensor Development
483
UV curing adhesive. A syringe pump is used to flow the DI, PBS, and prostate cancer cells in the fluidic channel. Figure 19.7 shows a bode plot of HBSS buffer solution and LNCaP cancer cells in solution for different incubation times. When LNCaP cells attach and spread on the surface of individual electrodes they behave essentially like insulating particles that hinder unrestricted current flow from the electrode into the bulk electrolyte. However, when the applied frequency is high enough, current can penetrate the plasma membrane and cross the cell layer. In general, we can separate three frequency regions called α, β, γ dispersions. As shown in Figure 19.7, the impedance of LNCaP prostate cells increased especially at 100 to 1,000 Hz compared to the HBSS. Probably the capacitance of the LNCaP cells blocks the current from 100 to 1,000 Hz. Another difference between the HBSS and LNCaP prostate cells is the phase angles that have a more capacitive behavior with the increase of incubation time. Probably LNCaP prostate cells form a tight contact and cover the individual electrodes, which blocks ion movement. Therefore, the increase of resistance and capacitance on the individual electrode surfaces changes the impedance compared to the HBSS solution. This phenomenon can be seen clearly on the Nyquist plot in Figure 19.7, showing the diameter of the semicircle at high frequency increasing with incubation time. This experiment showed that cancer cells in solution and settling on the electrode could be detected by monitoring the impedance signature the CNT electrode. Further experiments are underway to use antibodies to selectively capture the cancer cells on the electrode and to count the number of cells based on the impedance signature. An adjustable active electrode is also being developed to improve the sensitivity of the sensor [14]. Even though the result in this section is promising, we should not ignore a realistic situation. This biosensor should able to identify 10 cancer cells within one billion or normal cells for the early diagnosis of cancer. Thus device should be modified, and optimized for practical use. The test in medium with cancer cell is also not enough since biofouling by nonspecific protein absorption mask original signal and
1
108
-10
6
10
104
2
2
10
0
10
2
4
10
10 (a)
10
6
-10
10
0
10
2
10
4
(b)
Figure 19.7 Sensor response to cancer cells in solution. (A) Magnitude, and (B) phase plot of electrochemical impedance in (a) HBSS; and then HBSS with LNCaP with different incubation times, (b) 5 minutes, and (c) 2 hours [5].
10
6
484
Nanoscale Engineering of Electrodes, Biosensors, and Protein Surfaces
further immobilization of platelets will even encapsulate of device. Controlled release of anticlotting agent such as heparin should be coupled to minimize fouling effect.
19.3 New Polymer Synthesis for High Throughput Experimental Design The ability to pattern multiple proteins on a surface to control cellular behavior has promoted the development of new cellular bioassays such as cell-cell interaction, and cell-adhering surfaces assays. These new assays allow the study of cell spreading, adhesion, migration, differentiation, and molecular signaling pathways. This cellular array provides not only new insight for cellular biology but also novel cell-based sensors for drug screening. Thus this section describes (1) material synthesis of a new biophotoresist, (2) protein patterning, and (3) medical applications. 19.3.1
Biophotoresist Synthesis
Surface immobilization of biomolecules such as antibodies, enzymes, and nucleic acid chains in microscale patterns is very important to understand cell biology and develop electrodes for biosensors. Several techniques for surface immobilization of biological materials including soft lithography, microcontact printing, dip-pen lithography, and e-beam or photolithography have been suggested. But the problem with using conventional photolithography is that the alkaline developer solution and the organic solvent denature most proteins. One research group has recently reported the use of photoresists requiring only low temperature hardening and a mild developing solution such as phosphate buffered saline (PBS) [15, 16]. The basic idea of using the new biophotoresist is to synthesize a polymer that has multifunctional abilities based on pH and UV light. We have synthesized the terpolymer poly(O-nitrobenzyl methacrylate-r-methyl methacrylate-r-poly(ethylene glycol) methacrylate) (PNMP, Figure 19.8), using O-nitrobenzyl methacrylate, poly(ethylene glycol) methacrylate, and methyl methacrylate. Further carboxylation
UV
No soluble
(a)
Soluble only above pH 6.5 PBS
(b)
Figure 19.8 Biocompatible photoresist: (a) the structure of PNMP, and (b) when UV irradiated, the o-nitrobenzyl group is cleaved off of the polymer leaving behind a pH-sensitive carboxyl group.
19.3 New Polymer Synthesis for High Throughput Experimental Design
485
and biotinylation was done using succinic anhydride and amine-PEO-biotin. Biotinylation of PNMP was confirmed by FITC labeled streptavidin (SAv). 19.3.2
Photolithography for Protein Patterning
Protein patterning can be done using biophotolithography as shown in Figure 19.9. Briefly, biotinylated PNMP was spin-coated onto the APTES conjugated glass surface (Figure 19.9(a)) and UV irradiated at 254 nm using a mask to create exposed and unexposed regions of polymer (Figure 19.9(b)). The exposed regions were then developed, using a neutral pH PBS solution (pH 7.4, Figure 19.9(c)). From here, the entire glass surface was irradiated again turning any remaining PNMP into a simple polyanion (Figure 19.9(d)). The first biotinylated protein is then attached to the PNMP by incubation in a pH 6.0 buffer; this creates two tiers of biotinylated protein 1 (Figure 19.9(e)). The top tier is then removed by developing in pH 7.4 PBS. This leaves behind a tier of biotinylated protein 1 in a specific pattern and exposed biotinylated PMNP that is suitable for binding a second set of proteins (Figure 19.9(f–g)). This procedure uses a developer at a neutral pH of 7.4, a protein incubating buffer at pH 6.0 (little to no PNMP dissolves in this acidic pH 6.0 buffer). These three features allow patterning of intact proteins. Thus we have established that the PNMP we synthesized displays the required pH and UV sensitivities.
Glass
UV
(a) (d) APTS SAv with anti-CD3 in pH 6.0 sol. Biotinylated PNMP
(e)
(b)
UV and development in pH 7.4 Sol.
(c)
Development in pH 7.4 Sol.
(f)
SAv with ICAM1 (g)
Figure 19.9 Diagram of methods of protein patterning. The important things to note about this procedure are that the developer used is at a neutral pH of 7.4, little to no PNMP dissolves in the acidic, pH 6.0, buffer that incubates the proteins, and no proteins are ever exposed to UV light. These three factors allow for two protein patterning of healthy, natural proteins.
486
Nanoscale Engineering of Electrodes, Biosensors, and Protein Surfaces
19.3.3
Protein Patterning
Figure 19.10 shows the patterning of streptavidin labeled with a FITC fluorophore (SAv-FITC) in the initially irradiated and developed regions of the polymer while streptavidin labeled with Cy5 fluorophore (SAv-Cy5) surrounds it. This patterning was done for demonstration using a coarse TedPella TEM Grid Mask. Based on this protein patterning method, we show one application in immunology. If we can pattern two proteins, ICAM1 (Inter-Cellular Adhesion Molecule 1) and anti-CD3. T lymphocyte cells will recognize this surface as an antigen presenting cell and be activated [15, 16]. As shown in Figure 19.11, the anti-CD3 (blue circle) was patterned with ICAM around and resting T cells were seeded on this surface for 15 minutes. Figure 19.11(a) shows the artificial antigen presenting cell surface with activation sites. Figure 19.11(b) shows T cells on the artificial antigen presenting cell (APC)-like surface. The T cell receptor (TCR) is stained as shown in Figure 19.11(c). The overlapped image is shown in Figure 19.11(d). The T cell was observed crawling and then being activated after seeing the activation site. Further
(a)
(b)
(c)
Figure 19.10 (Color plate 31) Protein patterning using a Ted Pella TEM Grid Mask. Proteins attached to the PNMP are (a) the green SAv-FITC was the first protein attached to the PNMP, (b) the red SAv-Cy5 was the second; and (c) when combined, the complete dual protein pattern is seen.
Figure 19.11 (Color plate 32) T cell activation on an artificial antigen presenting surface: (a) activation sites (anti-CD3) is stained with cy5, (b) DIC image of T cells on the APC surface, (c) TCR is stained, and (d) the overlapped image.
19.4 Summary and Conclusions
487
activation of T cell should be proved by immunostaining or intracellular calcium imaging.
19.4
Summary and Conclusions This chapter discussed how nanoscale electrodes and patterned surfaces can be used to investigate the interactions between nanomaterials and biological systems and provides a new platform and nanotools for the study of biology in medical science. Among the various nanomaterials and possible applications, this chapter described development of label-free immunobiosensors using carbon nanotube arrays. Carbon nanotube arrays were synthesized using the chemical vapor deposition method. A novel array nanoelectrode was then fabricated and characterized for biosensor development. A label-free immunosensor and an initial design of a prostate cancer cell sensor were discussed. Inorganic nanoparticles were also discussed. A novel ter-polymer that has pH and UV sensitivity was synthesized using radical polymerization for use as a biophotoresist. Photolithography was then used to pattern different proteins using this novel polymer as a biophotoresist. Biologically inspired protein patterns were created using this novel polymer surface to allow the study of live cell interactions on two-dimensional protein surfaces, which helps to understand immunology in medical science. Further development of biosensors is expected to provide a method for detecting metastasis in cancer. Patterning of proteins will help to understand ion channel mobility and understanding diseases such as lupus.
Problems 19.1 Describe the mechanism of operation of the label-free immunosensor. What will be the limitations for the label-free immunosensor to be used in blood solution. 19.2 Describe a method to conjugate an antibody on the nanotube electrode. 19.3 What is an antigen presenting cell (APC)? How does the APC interact with the T lymphocyte?
Acknowledgments This interdisciplinary research was sponsored by several grants, including the University of Cincinnati Research Council Interdisciplinary Faculty Research Grant, membrane signaling in geometrically patterned immunological synapses; NSF grant CMMI 0727250, Nanomanufacturing and Production Scale-up of Long Carbon Nanotube Arrays for Advanced Applications; NSF grant CMMI 0700747 Improving the Mechanical Integrity of Biomaterials Using Carbon Nanotubes; CMMI 0510823 Telescoping Nanotube Arrays; and NSF EEC 0812348 grant Engineering Research Center (ERC) for Revolutionizing Metallic Biomaterials. This support is gratefully acknowledged.
488
Nanoscale Engineering of Electrodes, Biosensors, and Protein Surfaces
References [1] Sumio Iijima, “Helical Microtubules of Graphitic Carbon,” Nature, Vol. 354, No. 56, 1991. [2] YeoHeung Yun, Zhongyun Dong, Vesselin N. Shanov, William R. Heineman, H. Brian Halsall, Amit Bhattacharya, and Mark J. Schulz, “Carbon Nanotube Electrodes and Biosensors: Review,” Nanotoday, December 2007, Vol. 2, No. 6. pp. 30–38. [3] Kenji Hata, Don N. Futaba, Kohei Mizuno, Tatsunori Namai, Motoo Yumura, and Sumio Iijima, “Water-Assisted Highly Efficient Synthesis of Impurity-Free Single-Walled Carbon Nanotubes,” Science, Vol. 306. No. 5700, November 19, 2004, pp. 1362–1364. [4] YeoHeung Yun, Vesselin Shanov, Yi Tu, Srinivas Subramaniam, and Mark J. Schulz, “Synthesis of Long Aligned Multi-Wall Carbon Nanotube Arrays by Water-Assisted Chemical Vapor Deposition,” Journal of Physical Chemistry B, Vol. 110 No. 47, 2006, pp. 23920–23925. [5] YeoHeung Yun, Zhongyun Dong, Vesselin N. Shanov, and Mark J. Schulz, “Carbon Nanotube Array for Electrochemical Impedance Measurement of Prostate Cancer Cells under Microfludic channel,” Monthly Featured Article, Nanotechnology, Vol. 18, 2007, p. 465505. [6] YeoHeung Yun, Adam Bange, Vesselin N. Shanov, William R. Heineman, H. Brian Halsall, Zhongyun Dong, Abdul Jazieh, Yi Tu, Danny Wong, Sarah Pixley, Michael Behbehani, and Mark J. Schulz, “Highly Sensitive Carbon Nanotube Needle Biosensors,” Journal of Nanoscience and Nanotechnology, Vol. 7, 2007, pp. 2293–2300. [7] YeoHeung Yun, Zhongyun Dong, Vesselin N. Shanov, Amos Doepke, William R. Heineman, H. Brian Halsall, Amit Bhattacharya, Danny K.Y. Wong and Mark J. Schulz, “Fabrication and Characterization of Carbon Nanotube Array Electrodes with Gold Nanoparticle Tips,” Sensors and Actuators B: Chemical, Vol. 133, No. 1, July 28, 2008, pp. 208–212. [8] YeoHeung Yun, Vesselin Shanov, Mark J. Schulz, Zhongyun Dong, Abdul Jazieh, William R. Heineman, H. Brian Halsall, Danny K. Y. Wong, Adam Bange, Yi Tu, and Srinivas Subramaniam, “High Sensitivity Carbon Nanotube Tower Electrodes,” Sensors and Actuators B, 2006, Vol. 120, pp.298–304. [9] YeoHeung Yun, Adam Bange, Vesslin N. Shanov, William W. Heineman., H. Brain Halsall, Danny K. Wang, M. Behbehani, S. Pixley, and M. J. Schulz, “A Nanotube Composite Microelectrode for Monitoring Dopamine Levels using Cyclic Voltammetry and Differential Pulse Voltammetry,” Proceedings of the Institution of Mechanical Engineers, Part N, Journal of Nanoengineering and Nanosystems. Vol. 220, 2006, pp. 53–60. [10] YeoHeung Yun, Adam Bange, William R. Heineman, H. Brian Halsall, Vesselin N. Shanov, Zhongyun Dong, Sarah Pixley, Michael Behbehani, Abdul Jazieh, Yi Tu, Danny K.Y. Wong, and Mark J. Schulz, “A Nanotube Array Immunosensor for Direct Electrochemical Detection of Antigen-Antibody Binding,” Sensors and Actuators B, 2007, Vol. 123, pp. 177–182. [11] Joon S. Shim, Yeo-Heung Yun, Michael J. Rust, Jaephil Do, Vesselin Shanov, Mark J. Schulz, and Chong H. Ahn, “High Precision Fluidic Alignment of Carbon Nanotubes Using Magnetic Attraction on a Metal Catalyst,” MEMS 2008, Tucson, AZ, January 13–17, 2008. [12] Sverre Grimnes, and Orjan Martinsen, Bioimpedance and Bioelectricity Basics, Academic Press Inc. (London) Ltd., 2000. [13] http://www.cancer.gov/. [14] NANOWORLD Lab, http://www.min.uc.edu/nanoworldsmart. [15] Junsang Doh, and Darrell J. Irvine, “Immunological Synapse Arrays: Patterned Protein Surfaces that Modulate Immunological Synapse Structure Formation in T Cells,” PNAS, Vol. 103, No. 15, 2006, pp. 5700–5705.
References
489
[16] Junsang Doh, and Darrell J. Irvine, “Photogenerated Polyelectrolyte Bilayers from an Aqueous-Probcessible Photoresist for Multicomponent Protein Patterning,” Journal of the American Chemical Society: JACS, Vol. 126, No. 30, 2004, pp. 9170–9171.
About the Editors Mark J. Schulz is an associate professor of mechanical engineering and codirector of the Nanoworld and Smart Materials and Devices Laboratories at the University of Cincinnati. He is also a deputy director of the National Science Foundation’s Engineering Research Center for Revolutionizing Metallic Biomaterials. The center is led by North Carolina A&T State University partnering with the University of Pittsburg and Hannover Medical Institute. Mark is also managing editor of Structural Health Monitoring, an international journal. His expertise is in the areas of smart materials and nanotechnology. Students in his laboratory develop nanostructured sensors and machines—including structural health monitoring sensors, carbon nanotube materials, and carbon electronics; and smart medical devices —including biodegradable magnesium implants, implantable nanomedical devices, active biosensors, and implants that expand and adapt to the human body. He received a BT degree from Buffalo State College, and M.S. and Ph.D. degrees in mechanical engineering from the State University of New York at Buffalo. He is also chief technology officer of General Nano LLC, in Cincinnati, OH. Vesselin N. Shanov is professor of chemical and materials engineering at the University of Cincinnati. He has international academic and industrial experience in development of facilities and technologies for processing nanostructured materials and thin films. Dr. Shanov has received several prestigious awards, among them the Fulbright Award for Research and Teaching in the United States, and the German Academic Foundation (DAAD) Award. His current research is focused on synthesis, processing, characterization, and application of nanostructured materials with emphasis on carbon nanotubes. He is codirector of the UC Nanoworld Laboratories. Dr. Shanov has published more than 150 papers, 14 patents, 30 invention disclosures, and 4 books. He received his M.S. in electronic materials from the University of Chemical Technology and Metallurgy, Sofia, Bulgaria. Dr. Shanov completed his Ph.D. in solid state chemistry at the University of Regensburg, Germany, and at the University of Chemical Technology and Metallurgy, Sofia, Bulgaria. He developed “Black CottonTM” and he is the Chief Scientific Officer of General Nano LLC, Cincinnati, OH. Yeoheung Yun is a research associate in the Department of Mechanical Engineering at the University of Cincinnati. He leads the nanomedicine and biomedical research performed in the Nanoworld Laboratory at the university. His interest is to apply engineering methods such as nanotechnology, microfluidics, and nanoparticles to solve biomedical problems. The main focus of his research is to design nano-devices for the study of cancer biology, immunology, bone biology, and systems biology. He developed a smart material using carbon nanotubes or carbon nanofibers, several biosensors, assays, and a method for patterning biological cells. He received an
491
492
About the Editors
M.S. degree in Mechanical Engineering from Chonbuk National University in South Korea and a Ph.D. degree in mechanical engineering from the University of Cincinnati.
List of Contributors
493
List of Contributors Chong H. Ahn Electrical and Computer Engineering University of Cincinnati Cincinnati, OH, 45221 USA e-mail: [email protected]
Laura Conforti College of Medicine University of Cincinnati USA e-mail: [email protected]
Marija Balic Department of Oncology Medical University of Graz Graz, Austria e-mail: [email protected]
Richard Cote Department of Pathology University of Miami Miller School of Medicine Miami, FL USA e-mail: [email protected]
Lajos P. Balogh Roswell Park Cancer Institute Buffalo, NY USA e-mail: [email protected]; [email protected] website: http://www.roswellpark.org/baloghl
Keith A. Crutcher Department of Neurosurgery College of Medicine University of Cincinnati Cincinnati, OH 45267 USA e-mail: [email protected]
Noy Bassik Department of Chemical and Biomolecular Engineering, and School of Medicine Johns Hopkins University Baltimore, Maryland USA e-mail: [email protected]
Ram H. Datar Oak Ridge National Laboratory, Oak Ridge Tennessee Miller School of Medicine, University of Miami USA e-mail: [email protected]
Frank Boehm NanoApps Medical, Inc. 324 North Court St. Thunder Bay, ON, P7A 4W7 Canada e-mail: [email protected] Supriya Chakrabarti Department of Chemical and Materials Engineering University of Cincinnati Cincinnati, OH 45221-0012 USA e-mail: [email protected] Linfeng Chen Center of Excellence for Nano/Neuro Electronics, Sensors and Systems Department of Electrical Engineering University of Arkansas Fayetteville, Arkansas 72701 USA e-mail: [email protected] Ying Chen Department of Chemical and Materials Engineering University of Cincinnati Cincinnati, OH 45221-0012 USA e-mail: [email protected] Wondong Cho Department of Chemical and Materials Engineering University of Cincinnati Cincinnati, OH 45221-0012 USA e-mail: [email protected]
Junsang Doh Department of Mechanical Engineering POSTECH South Korea e-mail: [email protected] Zhongyun Dong College of Medicine University of Cincinnati Cincinnati, Ohio 45221 USA e-mail: [email protected] Mitra Dutta Dept of Electrical and Computer Engineering, MC 154 Dept. of Physics 851 S. Morgan Street University of Illinois at Chicago Chicago, IL 60607 USA e-mail: [email protected] Teyeb Ould Ely Roswell Park Cancer Institute Buffalo, NY USA e-mail: [email protected] Xuesong Feng Nanyang Technological University School of Chemical and Biomedical Engineering Block N1.2 B3-15 637459 Singapore e-mail: [email protected]
494
List of Contributors Svetlana Fialkova North Carolina A&T State University Greensboro, NC USA e-mail: [email protected] Thomas L. Ferrell Department of Physics, University of Tennessee USA e-mail: [email protected] Robert A. Freitas Jr. Institute for Molecular Manufacturing 555 Bryant Street, Suite 354 Palo Alto, CA 94301 USA e-mail: [email protected] web site: http://www.rfreitas.com David H. Gracias Department of Chemical and Biomolecular Engineering and Department of Chemistry Johns Hopkins University Baltimore, Maryland USA e-mail: [email protected] Peixuan Guo Department of Biomedical Engineering University of Cincinnati/College of Engineering and College of Medicine Cincinnati, OH 45221 USA e-mail: [email protected]; [email protected] websites: http://www.eng.uc.edu/nanomedicine/ peixuanguo.html http://www.eng.uc.edu/nanomedicine/ http://nanomedicinecenter.org/centers/phi29 H. Brian Halsall Department of Chemistry University of Cincinnati Cincinnati, OH 45221 USA e-mail: [email protected] William Heineman Department of Chemistry University of Cincinnati Cincinnati, OH 45221 USA e-mail: [email protected] Elena Heister Faculty of Health and Medical Sciences University of Surrey Guildford GU2 7XH UK e-mail: [email protected] web site: http://www.surrey.ac.uk/fhms/research/themes/ matnano/elena.htm Tad Hogg Hewlett-Packard Laboratories Palo Alto, CA USA e-mail: [email protected]
Douglas Hurd College of Engineering University of Cincinnati Cincinnati, OH 45221-0072 USA e-mail: [email protected] Chaminda Jayasinghe Department of Chemical and Materials Engineering University of Cincinnati Cincinnati, OH 45221-0012 USA e-mail: [email protected] Sungho Jin UCSD Materials Science & Engineering Program Department of Mechanical & Aerospace Engineering University of California San Diego, La Jolla, CA 92093 USA e-mail: [email protected] website: http://maeweb.ucsd.edu/~jin/ Neeraj Jolly Section of Cardiology Departments of Medicine University of Chicago Chicago, IL 60637 USA e-mail: [email protected] Ratnesh Lal Center for Nanomedicine Departments of Medicine University of Chicago Chicago, IL 60637 USA e-mail: [email protected] Wojciech G. Lesniak Roswell Park Cancer Institute Buffalo, NY USA e-mail: [email protected] Ge Li Department of Chemical and Materials Engineering University of Cincinnati Cincinnati, OH 45221-0012 USA e-mail: [email protected] Weifeng Li Department of Mechanical Engineering University of Cincinnati Cincinnati, OH 45221-0072 USA e-mail: [email protected] Henry Lin Biosciences Division Oak Ridge National Laboratory Oak Ridge, Tennessee USA e-mail: [email protected]
List of Contributors Thierry Lutz Department of Chemistry Imperial College London South Kensington Campus London SW7 2AZ UK e-mail: [email protected] Nilanjan Mallik Institute of Technology Banaras Hindu University India e-mail: [email protected] David Mast Department of Physics University of Cincinnati USA e-mail: [email protected] I. Nedkov Institute of Electronics Bulgarian Academy of Sciences 72 Tzarigradsko Chausee, 1784 Sofia, Bulgaria Bulgaria e-mail: [email protected] Tomomi Nemoto Supportive Center for Brain Research National Institute for Physiological Science The Graduate University for Advanced Studies (SOKENDAI), and JST, CREST 38 Myodaiji, Nishigonaka, Okazaki, Aichi 444-8585 Japan e-mail: [email protected] Jai Raman Section of Cardiothoracic Surgery Departments of Surgery University of Chicago Chicago, IL 60637 USA e-mail: [email protected] Pravahan Salunke Department of Chemical and Materials Engineering University of Cincinnati Cincinnati, OH 45221-0012 USA e-mail: [email protected] Mark J. Schulz Department of Mechanical Engineering University of Cincinnati Cincinnati, OH 45221-0072 USA e-mail: [email protected] website: http://www.min.uc.edu/nanoworldsmart Vesselin N. Shanov Department of Chemical and Materials Engineering University of Cincinnati Cincinnati, OH 45221-0012 USA e-mail: [email protected] website: http://www.min.uc.edu/nanoworldsmart
495 Joon Sub Shim Electrical and Computer Engineering University of Cincinnati Cincinnati, OH, 45221 USA e-mail: [email protected] Yi Shu Department of Biomedical Engineering University of Cincinnati/College of Engineering and College of Medicine Cincinnati, OH 45221 USA e-mail: [email protected] Michael A. Stroscio Dept. of Bioengineering, Dept of Electrical and Computer Engineering, MC 154 Dept. of Physics 851 S. Morgan Street University of Illinois at Chicago Chicago, IL 60607 USA e-mail: [email protected] website: http://tigger.uic.edu/depts/nanotechcenter/index.htm Yu chong Tai Department of Electrical Engineering California Institute of Technology Pasadena, CA USA e-mail: [email protected] Ph. Tailhades CIRIMAT, CNRS-UPS-INPT 5085 University “P. Sabatie” 31062 Toulouse Cedex 4 France e-mail: Thomas Thundat Biosciences Division, Oak Ridge National Laboratory, Oak Ridge TN Department of Physics, University of Tennessee Knoxville, TN USA e-mail: [email protected] R. E. Vandenberghe Dept. Subatomic and Radiation Physics Gent University 86 Proeftuinstraat, B-9000 Gent Belgium e-mail: Sergey Yarmolenko North Carolina A&T State University Greensboro, NC USA e-mail: [email protected] Faqing Yuan Department of Biomedical Engineering University of Cincinnati/College of Engineering and College of Medicine Cincinnati, OH 45221 USA e-mail: [email protected]
496
List of Contributors Yeoheung Yun Department of Mechanical Engineering University of Cincinnati Cincinnati, OH 45221-0072 USA e-mail: [email protected]
T. Yuzvinsky Condensed Matter Physics Department of Physics University of California at Berkeley USA e-mail: [email protected]
Vijay K. Varadan Center of Excellence for Nano/Neuro Electronics, Sensors and Systems Department of Electrical Engineering, Department of Neurosurgery University of Arkansas, Fayetteville, Arkansas 72701 Department of Neurosurgery Pennsylvania State University, Hershey, PA USA e-mail: [email protected]
Alex Zettl Condensed Matter Physics Department of Physics University of California at Berkeley USA e-mail: [email protected]
J. Weldon Condensed Matter Physics Department of Physics University of California at Berkeley USA e-mail: [email protected] Jining Xie Center of Excellence for Nano/Neuro Electronics, Sensors and Systems Department of Electrical Engineering University of Arkansas, Fayetteville, Arkansas 72701 USA e-mail: [email protected]
Hui Zhang Department of Biomedical Engineering University of Cincinnati/College of Engineering and College of Medicine Cincinnati, OH 45221 USA e-mail: [email protected] Siyang Zheng Department of Bioengineering The Pennsylvania State University University Park, PA USA e-mail: [email protected]
Index A Acetylation, 101–2 Actin, 410 Actuators, CNT, 52–53 Aerosol inhalation, 396–97 Aerosolization technique, 80 Affinity-based chip, 352 Aligned CNTs for axonal regeneration, 194–95 geometric properties, 195 longitudinal growth, 202 manufacture of, 194 neurite outgrowth, 196–200 neuronal attachment, 196–200 neuronal cultures, 196 preparation of, 196 thin layer, 200 3-D, 194 See also Carbon nanotubes (CNT) Ambulatory nanodevices, 413–15 display of, 413–14 DNA robot, 414–15 programming, 414 Amplification, 386 Anisotropic magnetoresistive (AMR) sensors, 269 Annihilation photons, 113 Anodic aluminum oxide (AAO), 223 Antibody-based enrichment, 350–51 Antibody-directed enzyme prodrug therapy (ADEPT), 63 Antiferromagnetic ordering, 174 Artificial biomaterials, 7 Atomic force microscopy (AFM), 5, 67, 214, 326–28, 424, 476 antibody conjugated tip pulling experiment, 332 cantilever results, 326 cantilevers, arrays of, 337 fluid flow properties and, 327–28 force mapping, 332 forces in, 326–27 integrated with nanochip-separated recording system, 329
ion conductance/permeability assay tools, 328 live myocytes image, 333 mapping, 328 in measuring molecular interactions, 329 for multimodal/multidimensional imaging, 326–28 operating principle, 326 parallel readout techniques based on, 335 AutoCAD, 131 Axonal regeneration, 189, 190, 192 aligned CNTs for, 194–95 artificial substrates promoting, 192–93 carbon nanotubes and, 193–94
B Batch synthesis, 6 Bead ARray Counter (BARC), 270 Beam expander, 465 Bearings, nanotube, 370–71 Bioactuators, 9 Biochips, magnetic, 276 Biocompatible photoresist, 484 Bioentities, 408 Biofibers, 194 Biofouling, 19 Biological nanorobots, 17 Biological nanotomography (bionanotomography), 5 Biomachines, 408 Biomechanisms, 407 Biomedical applications magnetic nanoparticles, 252–54 magnetic nanotubes, 263–65 magnetic nanowires, 257–60 magnetic particles and, 175–83 magnetic single-domain particles, 177–83 Biomedical devices drug delivery, 135 implantable organic/electronic, 137 interactions with body components, 139–40 lithographically structured, 134–40 live, growing, 139 microfluidic, 137–38
497
498
size and, 148 structural, 135–37 3-D, 132–34, 138–39 Biomolecular detection assays, 316–18 DNA hybridization and, 316, 317 PSA detection, 317–18 See also Microcantilever sensors Biomolecule manipulation, 257–58 Biomotors, 407 Biophotoresist synthesis, 484–85 Biorelated synthesis, 251 Biosensors CNT array, 480–81 defined, 267 development, 9 elements, 267–68 magnetic, 267–76 nanorobot, 11, 15–16 neural-integrated, 15–16 sensitivity, 268 Bioworld of magnetism, 176 magnetite and, 175–77 in nanometer scale, 177 nanotechnology in, 175 superparamagnetism and, 176 Black Cotton, 35 Blocking temperature, 168 Blood/brain barrier (BBB), traversing, 396–97 Boltzmann constant, 168 Bones, 132–33 Bovine serum albumin (BSA), 317 Brownian mode, 179 Brownian motion, 219, 288 exploitation, 408 molecular motors and, 401–2 Brownian shuttles, 402
C Cancer biomarkers, 318 complexity, 96 diagnostics, 315–16 imaging, 113–16 present-day treatment, 62–63 prostate, cell detection, 481–84 statistics, 61–62 stem cells, 358 Cancer therapy composite nanoparticles for, 95–118 nanoparticles for, 7–8 nanorobots for, 15
Index
targeted, 61–85 Cantilevered-based array detectors, 337 Cantilevered microfluidic sensing tools, 335 Cantilever preparation, 316 Capsid proteins, 213 Carbon nanotube drug delivery systems, 71–79 based on filling the inner cavity, 77–79 based on surface functionalization, 71–77 investigation launch, 74–75 shuttle functionalization, 77 Carbon nanotubes (CNT), 9 acid oxidation, 69 actuators, 52–53 aligned, 189–202 applications for, 30 axonal regeneration and, 193–94 biosensors, 479–80 bundles, handling of, 47 bundles, neuronal attachment and, 198, 199 characterization of, 29, 67–69 coil, 14 cytotoxicity, 195 defects in, 6 defined, 65 double-wall (DWCNT), 38–40 fluidic assembly by magnetic attraction, 480 functionalization for biomedical applications, 70–71 functionalized with biocompatible molecules, 426 glycodendrimer-coating of, 425 materials as scaffolds, 53–54 multiwall (MWCNT), 27, 28, 79, 200, 427 nano-handling of, 46–48 as nanovectors for multimodal drug delivery, 71–79 neurite growth and, 193 purification of, 69–70 RNA-wrapped, 70 sheets, 194 sheets, neuronal attachment and, 198 single-wall (SWCNT), 27–28 spinning into thread, 35–43 as substrate for neural growth, 195 substrate preparation for growing, 29 superficial resemblance, 80 synthesis of, 6, 14, 27–54, 65–67 tower, 477 toxicological aspects, 80–84 as versatile material, 65–71 in vivo biodistribution, 84–85
Index
See also CNT arrays; CNT electrodes; CNT growth; CNT threads Carbon placement tools, 377 Carcinoembryonic antigen (CEA), 315 Carcinogenicity, 83 Cell encapsulation therapy (CET) defined, 142 3-D self-assembled containers, 142–43 Cell enrichment methods, 349–51 antibody-based, 350–51 based on physical characteristics, 349–50 See also Circulating tumor cell (CTC) detection Cells magnetic separation, 181 mechanical properties measurement, 331 mechanics, 330–32 seeding, 385 signaling processes, 385 T, 385, 486 3-D constructs for culture, 145–47 Cellular automata, 294, 295 Central nervous system (CNS), 53, 189 axonal regeneration, 189–90, 192–95 functional recovery in, 189 white matter, 190, 202 Cerebral neocortex, in vivo imaging, 467–69 Charge-coupled devices (CCDs), 442, 451 Chemical vapor deposition (CVD), 17, 66 CNT, 66 defined, 29 techniques, 66 thermal, 479 Chirp compensation, 466 Chromallocytes, 381–82 defined, 381 mobility system, 382 primary manipulator, 381 See also Diamondoid nanorobots Chromosome replacement therapy (CRT), 381 Chronic myeloid leukemia (CML), 63 Cilia, 408–9, 430 Circulating tumor cell (CTC) detection, 347–59 cell enrichment methods, 349–51 DTC detection and, 348 enhancing, 359 lower rate of, 348 microfabricated devices, 352–56 technical challenge, 348–49 techniques, 349–56 Circulating tumor cells (CTCs), 4, 347
499
capture, nanotechnology, application of, 358–59 captured, 355–56 characterization, 348 characterization in blood samples, 359 clinical significance of, 357 isolation of, 350 number of, 348 prognostic relevance of, 348, 357 See also Circulating tumor cell (CTC) detection Circulatory system illustrated, 13 nanorobots for maintenance, 12–15 CNT arrays centimeter-long, 54 characterization, 31–32 CVD synthesized, 35 defined, 27 direct spinning from, 36–38 electroanalytical method-driven biosensing with, 475 highly pure, 34 images of, 30 immunosensor, 51–52 length of, 38 long, synthesis of, 29–30 manufacturing of, 29–30 patterned, 32 production scale up of, 32–34 pulling ribbon from, 40–42 quality of, 33 spinning, mechanics, 36 synthesis, 476 See also Carbon nanotubes (CNT) CNT electrodes, 476–84 electron-transfer resistance, 481 fabrication, 476, 478 individual, 479–80 for prostate cancer cell detection, 481–84 See also Carbon nanotubes (CNT) CNT growth, 6, 29 on 4-inch Si substrates, 34 scale-up of, 34 See also Carbon nanotubes (CNT) CNT ribbon electrical properties of, 46 four-probe measurement of, 46 manufacturing, 41 pulling from arrays, 40–42 Raman spectrum of, 42 SEM images of, 41
500
supported on TEM grid, 40 TEM images of, 42 CNT thread-based antenna, 49–51 cell phone photograph, 50 development of, 49–50 future medical application of, 50–51 CNT threads actuator, 52–53 blending, 39 characterization of, 43–46 electrical conductivity, 45 electrical properties of, 43–44 ESEM images of, 38–39 need for, 35 post-treatment of, 42–43 resistance, 44–46 spinning, 35–43 as substrate for neuronal attachment, 197 tensile testing of, 43 in wireless, biomedical sensor applications, 49–51 See also Carbon nanotubes (CNT) Colloidal chemical synthesis, 248 CombiGel, 398 Composite nanodevices (CNDs), 108 biomedical applications, 118 delivery of, 109 Composite nanoparticles, 95–118 Computed tomography (CT), 113, 115, 127 Confocal fluorescence microscopy, 108 Connector arrays, 224–25 Connector proteins, 213–14 Core-shell nanoparticles, 245–47 formation, 246 properties, 247 types of, 245–46 See also Nanoparticles Coulomb attraction, 445 Covalent coupling chemistry, 73 Covalent functionalization, 71 Cross interference, 481 Crystalline structure, 158–63 Cyclic voltammetry (CV), 477–78 Cytobots, 370 Cytopenetration, 384 Cytoskeletal structural reorganization, 330–31 Cytotoxicity, 81, 195
D DAPI, 351 DC bias, 434 Death receptors, 384
Index
Decision criterion, 303 Deep reactive ion etching (DRIE), 352 Dendrimers, 98–99, 115 application examples, 116 defined, 98 for drug delivery, 116 families, 99 materials, 99 molecules, 98 nanocomposites, 105–9 PAMAM, 99–105 Dendritic polymers, 98 Dermal burrowing, 394–96 Descriptive systems modeling (DSM), 1, 2–4 acceptance, 3 approach, 2 database, 2 defined, 2 examples, 3–4 in risk-benefit analysis, 4 See also DSM models Detection performance, 303–4 Detection thresholds, 293 Deterministic flow, 352–53 Diagnostic information, 304–5 Diagnostic task environment, 301–2 Diamond mechanosynthesis (DMS), 373 Diamondoid nanorobots, 371–73 carbon placement tools and, 377 chromallocytes, 381 examples of, 379–82 hydrogen abstraction tools and, 375–76 hydrogen donation tools and, 376 manufacture of, 371–72 materials, 371 microbivore, 380 molecular manufacturing, 373–77, 378–79 positionally controlled processes, 372 respirocyte, 379 self-assembly processes, 371–72 See also Nanorobots Diamond surfaces, 375 Dichroic mirror (DM), 467 Dielectrophoresis (DEP) force, 349–50, 479 for cell separation/enrichment, 350 defined, 349 generation, 353 negative (nDEP), 350 positive (pDEP), 349–50 Dielectrophoresis field-flow fractionation (DEP-FFF), 353 Differential piezoresistive cantilever, 321
Index
Diffusive gels, 397–98 Diode-pumped solid-state (DPSS), 461 Dipolar rotor, 401 Disseminated tumor cells (DTCs), 347 in bone marrow, 358 characterization of, 348 detection, 348 Distributions, 97 molecular mass, 102 Poisson, 303 spin magnetic moments, 169 DNA computers, 293 double-stranded (dsDNA), 211 hybridization, 316, 317 hybridization capability, 251–52 packaging, 212, 213, 214 robot, 414 sequence detection, 316 single-stranded (ssDNA), 251, 316 translocation of, 211 DNA-gp3, 215 DNA-packaging motor, 227 Domain walls (DWs) defined, 164 spin structure in, 166 Double-stranded DNA (dsDNA), 211 Double-wall CNT (DWCNT), 38 growing of, 38 spinning thread from, 38–40 Drug delivery controlled, 253 devices, 135, 136 ideal nanorobot vehicle, 382–87 magnetic nanoparticles and, 254 magnetic nanotubes and, 266 3-D self-assembled containers for, 143–45 Drug targeting, 181 DSM models illustrated, 4 instrumentation for development, 4–6 See also Descriptive systems modeling (DSM) Dynein, 410
E Edema, 338 characterization, 339 defined, 338 Edema sensors, 338–41 foolproof, 340
501
microscale, 341 nanoscale, 341 principles of, 340–41 remotely controlled in vivo, 339 system, 341 EGF-conjugated quantum dots (EGF-QD), 452 EGF receptors (EGRFs), 452–53 Egress strategies, nanodevice, 416–17 Electroanalytical method-driven biosensing, 475 Electrochemical impedance spectroscopy (EIS), 480 Electrodeposition, 141 Electromotive force (EMF), 143 ELISA, 314, 317, 318 Endothelialization sensors, 340 Enhanced permeation and retention (EPR), 64 Environmental SEM (ESEM), 31, 47, 476 images of nanomanipulator, 48 for morphological study, 31 EpCAM, 351 Epidermal growth factor (EGF), 76 Epidermal growth factor receptor (EGFR), 352 Exocytosis, 470 External magnetic propulsion, 412 Extracellular matrices (ECMs), 443 Eye/ear drops, 397–98
F Ferrimagnet, 174 Ferrofluids, 180, 243 Ferrospinels, 171 Field cooling (FC), 249 Flagella, 408–10, 430 Flagellar doublet, 409 Fluid flow properties, 327–28 Fluorescent nanoparticles, 115 Fourier-transform infrared spectroscopy (FTIR), 5 Free energy reduction, 319 Functionalization covalent, 71 magnetic nanoparticles, 251–52 magnetic nanotubes, 264 nanowires, 255, 256–57 surface, 71–77, 112–13 Functionalized CNTs conjugation, 71 neurite growth and, 193 schematic illustration, 73
502
G Gamma rays, 113 Gas oxidation, 69 Gastrointestinal stromal tumor (GIST), 63 Gene delivery, 259 Gene therapy, phi29 DNA packaging motors, 226–27 Giant magnetoresistive (GMR) sensors, 269 Global positioning system (GPS) frequencies, 415 Globus model, 165 Gold colloids, 115
H Hall-Effect sensors, 271–72 Hematopoietic cell depletion, 350 HemIX, 256, 257 Hemoglobin, 158 Hepatocytes, 133 Heterodimers, 112 Heteronanostructures, 110–12 hybrid, 111 multibranched, 117 rod-based shape-engineered, 111–12 sphere-based shape-engineered, 110–11 High-resolution diagnostics, 300–305 control, 302–3 decision criterion, 303 detection performance, 303–4 diagnostic information, 304–5 localized high concentrations and, 300 measurement choice, 302 probability distribution, 303 task environment, 301–2 task geometry schematic illustration, 301 See also Microscopic robots High-resolution TEM (HRTEM), 107, 476 Hot-injection solvothermal method, 248 Human engineering, 130 Human microanatomy, 133 Human serum albumin (HSA), 317, 452 Hybrid nanorobots, 17 Hydrogen abstraction tools, 375–76 Hydrogen donation tools, 376 Hypodermic injection, 394–96 Hysteresis, 178, 184
I Implantable nanomedical devices, 8–9 biodegradable metals for, 17–18 biofouling, 19
Index
development of, 8–9 focus of, 9 integration in the body, 18–19 science and technology areas, 9 Implantable organic/electronic devices, 137 Implantable sensors, 318–19 In-body nanosensors, 336–44 edema, 338–41 magnetically navigated robot capsule, 341–43 mobile microscopic robots, 343–44 principles of, 340–41 Ingress strategies, 394–98 aerosol inhalation, 396–97 dermal burrowing, 394–96 diffusive gel, 397–98 eye/ear drops, 397–98 hypodermic injection, 394–96 transdermal patch, 397–98 See also Nanodevices Inhibitors of apoptosis (IAPs), 385 Inorganic nanoparticles, 109–13 advantage of, 110 rod-based shape-engineered heteronanostructures, 111–12 sphere-based shape-engineered heteronanostructures, 110–11 surface functionalization, 112–13 See also Nanoparticles Integrated low-energy electron Mössbauer spectroscopy (ILEEMS), 171 defined, 171 magnetite particle surface study, 173 surface defect changes, 171 Intermolecular interaction, 330 Intracellular delivery, 423 In vivo endothelialization sensors, 338 In vivo imaging, cerebral neocortex, 467–69 In vivo sensing technology, 326–36 AFM, 326–28 nanoimaging and nanosensing, 328–33 parallel sensor arrays, 333–36 Ion conductance/permeability assay tools, 328
K Karyobots, 370 Kinesin, 410
L Lambert-Beer’s Law, 67 Laplace pressure, 163
Index
Linear nanomotors, 427–29 Linkage, 387 Lipid nanoparticles, 369 Liposomes, 115, 369 Liquid viscosity, 327 Lithographically structured biomedical devices, 134–40 drug delivery, 135, 136 implantable organic/electronic, 137 interactions with body components, 139–40 live, growing, 139 microfluidic, 137–38 soft 3-D, 138–39 structural, 135–37 wet 3-D, 138–39 Lithographic fabrication, 127–48 basics, 130–32 paradigm, 132 processes, 130 self-assembly and, 130 wafer-level processing, 130 Lithography optical, schematic diagram, 131 in quasi 3-D multilayer fabrication, 128 self-assembly with, 140–47 soft, 353 Liver, parallel network of lobules, 133 Lorentz force, 271 Luminescent imaging, 5 Lungs as gas exchange medium, 132 Luteinizing hormone releasing hormone (LHRH), 178 Lymphatic capillaries, 406 Lymphatic system, 406 Lysine-phosphate alignment rotation, 220
M Magnetically navigated robot capsule, 341–43 defined, 341–42 illustrated, 342 power, 342–43 for therapeutics, 342 Magnetic automated cell sorter (MACS), 351 Magnetic beads, 266 Magnetic biochips, 276 Magnetic biosensors, 267–76 detecting ferrofluid susceptibility, 275–76 Hall-effect, 271–72 magnetoresistance-based, 270–71 schemes, 267–70 solid-state based, 268–70 substrate-free, 270
503
Magnetic colloidal systems, 248 Magnetic force microscope (MFM), 172 Magnetic induction, 163 Magnetic interactions, 164 Magnetic intercellular hyperthermia, 181 Magnetic iron oxide nanoparticles, 178 Magnetic labels, 268, 270 Magnetic manipulation, 243–44 Magnetic moments, static distribution, 261 Magnetic nanomaterials, 239, 241–42 characteristics, 241 prospects, 239, 276–77 types of, 241–42 Magnetic nanomedicine, 242 Magnetic nanoparticles, 242, 247–54 basics, 247–48 biomedical applications of, 252–54 core-shell, 245–47 in drug delivery, 254 ferrofluids and, 243 functionalization techniques, 251–52 manipulation, 243–44 monodisperse, 249 physical background, 243–47 synthesis of, 243, 248–51 See also Nanoparticles Magnetic nanoscience, 157 Magnetic nanospheres, 241 Magnetic nanotubes, 260–67 biomedical applications, 265–67 defined, 241 drug delivery, 266 fabrication process, 266 functionalization, 264 magnetism of, 260–63 magnetization reversal, 262 multifunctionality of, 263 paramagnetic nature, 265 schematic structure of, 261 synthesis of, 263–65 See also Nanotubes Magnetic nanowires, 254–60 biomedical applications of, 257–60 biomolecule manipulation, 257–58 functionalization of, 256–57 gene delivery, 259 hybrid devices, 260 properties, 254 structures, 254 suspended biosensing system, 258–59 synthesis, 254–56 Magnetic permeability, 275
504
Magnetic resonance imaging (MRI), 14, 113 contrast agents, 181, 182 liver, 182 tomography, 158 Magnetic resonance (MR), 127, 182 Magnetic separation, 181 Magnetic thin films, 241, 242 Magnetic viscosity, 179 Magnetism, 157, 158 bioworld of, 176 classical, 164 of magnetic nanotubes, 260–63 nanosized, 163–74 properties, 260–61 superparamagnetism, 163–71 Magnetite biocompatibility, 178 bioworld and, 175–77 bulk, 158–60 crystalline structure, 158 defined, 158 ferrospinel representation, 171 particle, surface structure, 174 particle core, 171 penetration of, 183 proliferation in living cell, 184 structural formula, 159 structure and magnetic properties, 160 structure illustration, 159 surface content, 171 TEM image, 183, 265 Magnetite nanoparticles heating of, 179 magnetization, 177 stable magnetic fluids, 177–78 Magnetization bead saturates, 272 changing, 166 curves, 168, 244 defined, 163 equilibrium, 179 magnetite nanoparticles, 177 reversal, 262 saturation, 168 value, 165 zero net, 164 Magnetocrystalline anisotropy, 262 Magnetoelectronic tool development, 277 Magnetorelaxometry (MRX), 272–73 defined, 272 performance, 273 Magnetoresistance-based sensors, 270–71
Index
BARC, 270–71 types of, 270 Magnetosomes, 175 Magnetron sputtering, 32–34 Maxwell bridge, 275, 276 Mechanosynthesis, 373 Medical nanorobots. See Nanorobots MED-JET, 394 Micelles, 115, 369 Microbivores, 380 Microblocks, 378–79 Microcantilever sensors, 313–21 biomolecular detection assays, 316–18 in cancer diagnostics, 315–16 cantilever preparation, 316 electrical readout technique, 315 functions, 313 implantable, 318–19 implementation, 314 platform, 313–15 summary, 320–21 Microelectromechanical-systems (MEMS), 16, 128 Microfabricated devices, 352–56 affinity-based chip, 352 DEP chip, 353–54 deterministic flow, 352–53 size-based microfluidic separation, 354–56 See also Circulating tumor cell (CTC) detection Microfluidic devices, 137–38 Microfluidic viscosity, 335 Microscale fluid velocity, 327 Microscale tetherless gripper, 147 Microscopic robots, 285–307 active locomotion, 291 bacteria and, 287 behavior models, 296–97 behaviors, 294–300 for biological applications, 287 capabilities and environment, 287–92 communication, 289–90 computation, 291–92 constructing, 286 environmental change monitoring, 294 high-resolution diagnostics, 300–305 microfluidic examination, 288 motion, 290–91 multiple physical effect modeling, 296–99 passive motion, 290–91 performance evaluation, 295 physical properties of, 295
Index
plausible capabilities for, 305 potential of, 286 power, 292 properties, 292 schematic illustration, 293 sensing, 287–89 simulation tools, 306–7 steady-state power generation for, 298 temperature distribution in, 299 theoretical studies, 294–96 using, 292–94 validation experiments, 300 for in vivo diagnostics, 285–307 for in vivo medical use, 369–70 See also Nanorobots Millimeter-scale manipulators, 293 Mobile microscopic robots, 343–44 MOhm, 334 Molecular manufacturing, 378–79 Molecular motors, 398–400 Brownian motion and, 401–2 Brownian shuttles and, 402 constraints on, 400–403 piezoelectric elements, 399 powering, 399 propellers, 399–400 viscous forces and, 402–3 Molecular propellers, 399–400 Molecular shuttle, 402 Molecular sorter, 227 Monodomain particles, 163–71 Mössbauer spectroscopy, 170, 171, 183 Multiangle laser light scattering (MALLS), 102 Multibranched heteronanostructures, 117 Multicolor fluorescence imaging, 450 Multidomain particles, 163–71 Multidomains, 248 Multidrug-resistant (MDR), 75 Multilayer thin films, for 3-D self assembly, 145, 146 Multiphoton excitation (MPE), 460 light sources for, 461 probability of occurring, 467 process, 461 Multisegment nanowires, 256 Multiwall CNTs (MWCNT), 27 axis, 427 defined, 27 forests, 200 high-quality, 427 properties, 28 TEM images, 79
505
See also Carbon nanotubes (CNT); CNT arrays Mutagenicity, 83 Myelin-associated glycoprotein (MAG), 190 Myosin, 410
N Nanocrystals powered nanomotor, 429 shaped, biomedical applications of, 117–18 Nanodentistry, 414 Nanodevices, 97 ambulatory, 413–15 application examples, 113–17 aqueous motility, 411–13 binding to integrin receptors, 106 composite (CND), 108 conceptual tissue migrating, 398 with covalently attached RGD peptides, 101 dendrimer, application examples, 116 design principles, 98 egress strategies, 416–17 fabrication materials, 98–113 ingress strategies, 394–98 molecular motors, 398–403 nanomotor in, 222–24 PAMAM-based, 105 reproducible synthesis, 97 size, 98 summary, 118 in 3-D space, 416 variables, 98 Nanodisk code (NDC) method, 113 Nanoelectromechanical systems (NEMS), 134, 265 Nanofabrication techniques, 475, 476 Nanofactory system, 378 Nanofluidic channels, 413 Nanohexapods, 111 Nanoimaging, 328–33 cell mechanics, 330–32 defining effectors, 329–30 intermolecular interactions, 329–30 structure/changes in membrane effectors, 328–29 tissue nanoelasticity and nanopatterning, 332–33 Nanoinjection, 423–24 Nanomagnetism, 244–47 core-shell nanoparticles, 245–47 nanoparticle assemblies, 245 superparamagnetism, 244–45
506
Nanomanipulators, 46–48 building nanomedical devices with, 47–48 defined, 46 ESEM images of, 48 nano-handling of CNTs with, 46–48 Nanomaterials artificial, 7 characteristics of, 6–7 defined, 6, 97 magnetic, 239, 241–42 for medicine, 6–8 smart, 7 for sustainability in the body, 7 Nanomechanical mass sensing, 434–36 Nanomedicine advanced procedures, 393 creating, 19 defined, 1, 240–41, 367 DSM for, 2 emergence, 393 magnetic, 242, 243–47 pharmaceutical therapy and, 241 power of, 368 research, 18–19 safety and ethical implications, 19–20 status, 242 Nano-MEMS, 325 Nanometer propulsion, 426–30 linear nanomotors, 427–29 rotational nanomotors, 426–27 surface-tension-driven, 429–30 See also Propulsion Nanometric biomimetic analogs, 407–10 actin, 410 characteristics, 407–8 cilia, 408–9 dynein, 410 flagella, 408–10 kinesin, 410 myosin, 410 Nanometric crampons, 396 Nanometric telescoping, 395 Nanomotors, 426–30 linear, 427–29 nanocrystal-powered, 429 rotational, 426–27 surface-tension-driven, 429–30 Nanoparticles alloyed, 110 assemblies, 245 batch synthesis, 6 in cancer imaging, 113–16
Index
for cancer therapy, 7–8 cleared by the body, 18 composite, 95–118 core-shell, 245–47 direct targeting, 115 fluorescent, 115 generalized synthesis, 250 inorganic, 109–13 iron-oxide, 8 lipid, 369 magnetic, 242, 247–54 magnetic iron oxide, 178 magnetite, 177 perfluorocarbon emulsion, 116 prefabricated, 99 properties of, 96 quasi-spherical shape, 161 shape, manipulating, 117 structural characteristics, 160–63 Nanopatterned surfaces, 18, 19 Nanopatterning, 332–33 Nanopropellers, 399–400 Nanorobot drug delivery vehicle, 382–87 Nanorobots, 10–17 bearings, 370 biological, 17 biosensors, 11, 15–16 for cancer therapy, 15 in circulatory system maintenance, 12–15 communication, 289 concept, 368 cross-section view, 13 defined, 10 development challenges, 11–12 diamondoid materials in, 371–73 factories, 16–17 future technology bridge, 369 hybrid, 17 motors, 370 neural-integrated biosensors, 15–16 parts, 368 performance evaluation, 296 for revolutionizing medicine, 12–16 self-replication and, 19–20 surgery with, 11 Nanoscale earthworm analog, 412 Nanoscale mechanics, 423–54 Nanoscale powders, 161 Nanoscience defined, 95 global revolution in, 95–96 magnetic, 157
Index
medicine and, 95–97 Nanosensing, 328–33 cell mechanics, 330–32 defining effectors, 329–30 intermolecular interactions, 329–30 structure/changes in membrane effectors, 328–29 tissue nanoelasticity/nanopatterning, 332–33 Nanosensors for biomarkers, 336 in-body, 336–44 Nanosized magnetite, 157–85 equilibrium condition, 162 stabilization in, 162–63 thermodynamic conditions and, 162 See also Magnetite Nanospheres, magnetic, 241 Nanostructured magnetite, 158 Nanostructures fabrication, 97 geometry, 117 heteronanostructures, 110–12 magnetite, 158 Nanotechnology connector arrays for, 224–25 defined, 95, 239, 240 global revolution in, 95–96 in nature, 175 Nanotomography, 5 Nanotube bearings, 370–71 Nanotube flagella, 411 Nanotube radio design, 430 illustrated, 433 operation frequency, 432 transmitters, 433–34 Nanotube radio receivers, 430–33 frequency-tuning methods, 433 parts of, 431 Nanotube resonators, 434 operation as mass sensors, 434 resonant frequency, 436 size, 436 Nanotubes biocompatibility, 424–26 biofunctionalization, 425 biomimetic flagellar propulsion with, 411 coarse tuning, 432 defined, 241 magnetic, 260–67 vibrating, field-emitting, 435
507
vibration of, 432 Nanowalker, 415 Nanowires, 9 defined, 241 magnetic, 254–60 multisegment, 256 nickel, 13, 14 National Institute for Physiological Sciences (NIPS), 468, 469 National Nanotechnology Initiative (NNI), 240 Neel relaxation, 180, 273 Nerve growth factor (NGF), 53, 190 Nervous system, 13 Neurite growth aligned CNTs, 196–200 CNTs as substrate for, 195 functionalized CNTs and, 193 Noninvasive imaging, 459–60
O On-chip electrolysis, 357 Optical coherence tomography (OCT), 115 Optical imaging, 127 Optical system, 465–67 beam expander, 467 chirp compensation, 466 dichroic mirror (DM), 467 elements, 465 light generation, 465 See also Two-photon excited fluorescence (TPEF) microscopy Optical tweezers, 221 Ordered structures, 130
P Packaging RNA (pRNA) binding, 217 chimera, 225–26 cosedimentation of, 214 defined, 212 phi29, 212 polyvalent, 226 ring, 217, 218, 219 stoichiometric study, 218 structure of, 216–17 PAMAM dendrimers, 99–105 amino-terminated, 100, 102 biomedical applications of, 118 features of, 100 MALDI-TOF spectra of, 103
508
molecular mass distribution, 102 NMR spectroscopy of, 104 potentiometric titration of, 104 surface properties, 100, 101 synthesis, 99, 106 as templates, 105 See also Dendrimers Parallel sensor arrays, 333–36 multifluidic viscosity, 335 patch clamp on a chip, 334–35 Patch clamp on a chip, 334–35 Patterned CNT arrays, 32 PEG chains, 84, 85 Perfluorocarbon emulsion nanoparticles, 116 Peripheral nervous system (PNS), 53, 189 Peristatic propulsion, 412–13 Permeability and retention (EPR), 75 Phagocyte avoidance strategies, 406–7 Phagocytosis, 383 carbon nanotube structure, 81 medical robots and, 383 Pharmaceutical therapy, 241 Pharmacytes as active medical nanorobots, 382 in cell signaling control, 385 characteristics, 382 defined, 382 mobility systems, 383 potential uses, 384 selectivity with, 384 tagging target cells, 385 Phase equilibrium, 162 Phi29 DNA packaging nanomotor, 211–27 applications, 222–27 ATP hydrolysis and, 219 Brownian motion model, 219 connector arrays, 224–25 connector contraction, 221 construction of, 215–16 DNA-gp3, 215 as DNA-sequencing apparatus, 227 DNA translocation outside central pore, 220 engineered connectors, 226 fiber protein, 215 gp16, 214 lysine-phosphate alignment rotation, 220 mechanism, 217–22 models, 219–21 motor function, direct observation, 222 motor function, single molecule study, 222 in nanodevices, 222–24
Index
neck protein, 215 optical tweezers, 221 polyvalent gene delivery with, 225–26 pRNA, 212, 216–17 procapsid, 213–14 sensitive virion assembly system, 216 sequential action of components, 219–20 single molecule approaches, 221–22 six-fold connector, 217 strength, 215 symmetry argument, 217–19 tail protein, 215 as tools for gene therapy, 226–27 Phi29 pRNA, 212 Phi29 procapsid, 213–14 capsid protein, 213 connector protein, 213–14 scaffolding protein, 213 Phosphate buffered saline (PBS), 484 Photoactivable GFP (PA-GFP), 464 Photobleaching, 268, 464 Photolithography, 485 Physical vapor deposition (PVD), 17, 66 Piezoelectric elements, 399 Piezoresistive microcantilever arrays, 318 Poisson distribution, 303 Polyethylene glycol (PEG) molecules, 450 Poly-l-lactide (PLLA) microfilaments, 192 Polymer synthesis, 484–87 Polyvalent gene delivery, 225 Positionally controlled processes, 372 Positron emission tomography (PET), 84–85, 96, 113–14 data based on radioactivity, 85 in vivo, 84 Prepatterned printed circuit boards (PCBs), 476 Propulsion behavior and potential for, 413 external magnetic, 412 nanometer, 426–30 ultrasonic peristatic, 412–13 Prostate cancer cell detection CNT array electrode for, 481–84 importance, 481–82 technique improvement, 482 Prostate-specific antigen (PSA), 317 concentration of, 317 detection, 317–18 dissolved, 317 Protein patterning, 486–87 diagram of methods of, 485
Index
509
See also Nanomotors
photolithography for, 485 Protein-protein binding, 317
S Q Quantum dots (QDs), 358–59 diameter of, 446 EGF-conjugated (EGF-QD), 452 emission linewidths of, 441 emission spectrum and, 446–47 with energy bandgaps, 448 energy levels of, 446 lifetimes of, 441 multifunctional nanoparticle probes, 450 RGD complexes, 451 semiconductor, 441–54 wavelength emission and, 447
R Radial breathing mode (RBM), 68 Radio frequency (RF) signals, 143 Raman spectroscopy, 31–32 CNT array characterization, 31 SWCNT characterization, 68 Reactive ion etching (RIE), 355 Red blood cells (RBCs), 404 Relaxation losses, 178 Relaxation oscillator motor, 429 Remote controlled, magnetically navigated robot capsule, 341–43 Replacement, 386–87 Resovist, 181 Respirocyte, 379 Reticuloendothelial system (RES), 181 Reverse micelle method, 250 RNA-wrapped carbon nanotubes, 70 Robots communication capabilities, 290 computation and, 291 fabricating, 285 large-scale, 287 locomotion, 290 for nanomanipulation, 47 steady-state power generation for, 298 temperature distribution in, 299 See also Microscopic robots; Nanorobots Rod-based shape-engineered heteronanostructures, 111–12 RosetteSep assay, 351 Rotational nanomotors, 426–27 operation, 428 schematic, 427
Sampling, 3-D self-assembled containers for, 145 Scaffolding proteins, 213 Scanning electron microscopy (SEM), 67 environmental, 31 high-resolution measurements, 173 images, 41 Secretory function imaging, 469–70 Self-assembly diamondoid nanorobots, 371–72 lithography with, 140–47 paradigm, 129 3-D containers, 140–45 Self-shielding effects, 464 Semiconductor quantum dots, 441–54 adaptation for nanodiagnostics and, 442–47 applied to study of cellular properties, 447–49 with biomolecules, 441–42 with energy bandgaps, 448 fluorescent dyes and, 441 integrated, 442 as nanocrystalline structures, 444 properties, 444 See also Quantum dots (QDs) Severe combined immunodeficiency (SCID), 145 Shaped nanocrystals, 117–18 Shape engineered inorganic nanocrystals, 95 Shape engineering, 117 Signal modifying actions, 386–87 amplification, 386 linkage, 387 replacement, 386–87 suppression, 386 Signal-to-noise (S/N) ratio, 463 Simultaneous multicolor fluorescence imaging, 464–65 Single-domain nanoparticles, 176 Single photon emission computed tomography (SPECT), 113 Single-wall nanotubes (SWCNT), 27–28 characterization, 27–28 current carrying density, 28 properties, 28 Size-based microfluidic separation, 354–56 cell lysis functionality, 355–56 microfilter device, 354–55 Size-based separation, 350
510
Skin layers, 395 Small business innovative research (SBIR) projects, 21 Small interfering RNA (siRNA), 225 Smart nanomaterials, 7 Soft 3-D devices, 138–39 Soft lithography, 353 Solid-state based sensors, 268–70 SonoPrep, 394 Spatial metrics room, 416 Sphere-based shape-engineered heteronanostructures, 110–11 Spin magnetic moments distribution, 169 Spinning thread, 35–43 direct, 36–38 from DWCNT arrays, 38–40 illustrated, 39 mechanics, 36 See also CNT threads SQUID current-biased, 273 measurements, 274 microscope, 274 structure, 273 uses of, 273 Stem cells, cancer, 358 Stents, 136–37 use problem, 137 uses, 136 Stochastic analysis approach, 295–96 Stoner-Wohlfarth theory, 245 Structural devices, 135–37 Subcellular nanotechnology, 5 Substrate-free sensors, 270 Substrate preparation for growing CNT, 29 magnetron sputtering for, 32–34 Superconducting quantum interference devices (SQUIDs), 269 Superparamagnetic iron oxide (SPIO), 181, 182 Superparamagnetic (SPM) particles coating, 180–81 equilibrium magnetization, 179 magnetic properties, 178 relaxation in, 180 total magnetic moment, 178 Superparamagnetism, 244–45 condition for, 168 critical particle size below, 169 nature exceeding limit of, 176 Suppression, 386
Index
Surface coatings, 306 Surface functionalization, 74–77, 112–13 Surface-tension-driven nanomotors, 429–30 defined, 429 illustrated, 430 See also Nanomotors Suspending biosensing system, 258–59
T Targeted cancer therapy brief insight into, 63–65 CNTs for, 61–85 introduction to, 61–65 See also Cancer; Cancer therapy T cells, 385, 486 TEM images of CNT ribbon, 40, 41–42 magnetite, 183, 265 of MWCNT, 79 of nanoscale radio receiver, 432 renal excretion, 84 of superlattices, 114 of 2-D assembly, 250 Therapeutic inputs, 1 Thermal gravimetric analysis (TGA), 32 Thermal noise, 286 Thermal treatment, cancer, 8 Thermodynamic relations, 163 Thermogravimetric analysis (TGA), 68–69 3-D devices, 128 with integrated electronics, 138 lithography and self-assembly in constructing, 140–47 soft, 138–39 wet, 138–39 3-D geometries, 145 3-D self-assembled containers, 140–45 for cell encapsulation therapy, 142–43 for drug delivery, 143–45 for sampling, 145 3-D self-assembly, multilayer thin film stress for, 145 Tissue edema, 338 Tissue migrating nanodevices, 398 Tissue nanoelasticity, 332–33 Tissue scaffolds, 139 construction of, 139 defined, 139 microfabrication for, 138 T lymphocytes, 385 Total analysis system (TAS), 277 Toxicity, 18
Index
Toxicology, 80–84 categories, 82–83 See also Carbon nanotubes (CNT) Transdermal patch, 397–98 Transmission electron microscopy (TEM), 31 basis, 67 grid, 40 high-resolution (HRTEM), 107, 476 in nanotube visualization, 432 See also TEM images Traumatic brain injury, 189 Two-dimensional devices, 128 Two-photon excited fluorescence (TPEF) microscopy, 459–71 absorption cross-sections, 460 advantages, 461 application to conscious animals, 471 capabilities, 459 deep and benign observations, 462–63 future possibilities, 470–71 illustrated, 466 improving, 469 microscopes, 468–69 optical system, 465–67 overview, 466 replacement for ultraviolet sources, 463–64 secretory function imaging, 469–70 self-shielding effects and, 464 simultaneous multicolor fluorescence imaging, 464–65
511
tomographic imaging, 468 in vivo, 467–69
U Ultra-high vacuum (UHV) environment, 373 Ultrasonic peristatic propulsion, 412–13 Unset strand, 414
V Very large scale integration (VLSI), 128 Viscosity, 403 Viscous forces, 402–3
W Wafer stacking, 134 Wet 3-D devices, 138–39 Wheatstone bridge, 314 Wireless communication carbon nanotube threads and, 49, 51 modern world and, 49
X X-rays, 113
Z Zero-field cooling (ZFC), 249
(a)
(b)
Fascicle Axon
capillary
Nanoparticle receptor Color plate 1 (Figure 1.2)
(c)
Metal catalyst Thermal annealing
AI2O3 SIO2 Catalyst nanoparticles CNT
CVD growth
Color plate 2 (Figure 2.2)
(a)
(b)
Color plate 3
(Figure 3.6)
Color plate 4
(Figure 3.7)
(b)
(a)
Color plate 5
(Figure 4.10)
(c)
(a)
Liver lobule
Neuron
(b)
Lung alveoli Color plate 6
(c)
(Figure 5.3)
(b) In plane fabrication
(a)
© 2005 IEEE Kapton film
Microsensor 1
Microsensor 2
Electrical leads
Spirally-rolling
Assembling Spirally-rolled tube
Sensor 1
Sensor 2
Enzyme layer Au WE Ag/AgCI RE Au CE Electrical leads
Glucose sensor inside
Color plate 7
(Figure 5.5)
Overlapped region part
Electroplated coil outside
(a)
(b)
(c)
Color plate 8
(Figure 5.9)
oxygen tetrahedral Fe A-site octahedral Fe B-site
(a) Color plate 9
(Figure 6.1)
(b)
Color plate 10
(Figure 7.2)
Color plate 11
(Figure 7.4)
Color plate 12
(Figure 8.3)
Color plate 13
(Figure 8.5)
(a)
(b)
(c)
100
5-FU
50
4-NO2-Ph ibuprofen
0
20
40
Time (h) Color plate 14
(Figure 9.18)
Color plate 15
(Figure 10.1)
Color plate 16
(Figure 10.7)
60
80
Molecule Target Molecu Probe Molecule Coating Substrate
Cantilever
Target Binding
Substrate Color plate 17
(Figure 11.1)
Color plate 18
(Figure 12.4)
Deflection, Δh
Color plate 19
(Figure 12.5)
Electrodes
Color plate 20
(Figure 13.2)
Color plate 21
(Figure 14.5)
Color plate 22
(Figure 14.6)
(a)
Color plate 23
(b)
(Figure 14.7)
Hair shaft Sweat pore Dermal papila Sensory nerve ending for touch Stratum corneum Pigment layer Stratum germinativum
Epidermis
Stratum spinosum Stratum basale
Dermis
Arrector pili muscle Sebaceous (oil) gland Hair folicle
Subcutaneous fatty tissue (hypodermis)
Papilla of hair Nerve fiber
Vein Artery
Blood and lymph vessels
Sweat gland Pacinian corpuscle
Color plate 24
(Figure 15.1)
Streptavidin
Actin filament Peripheral stalk
ATP ADP + P1 Central stalk
Proton half-channel
Phospholipid bilayer membrane Proton half-channel C-ring rotation (only orange components rotate) C-ring
Color plate 25
(Figure 15.4)
Functionalization Microtip Nanotube Living cell
Color plate 26
Functional group
(Figure 16.1)
Insertion
Release
(a) Color plate 27
(Figure 16.12)
(a) Color plate 28
(b)
(Figure 17.2)
(b)
(c)
Three-photon Zymogen granule (peptide?)
Color plate 29
(Figure 18.3)
Color plate 30
(Figure 18.5)
Two-photon Mitochondria (NADH)
(a) Color plate 31
(a)
(b)
(c)
(Figure 19.10)
Activation sites
(b)
T cells
(d)
(c)
T cells in contact with activation sites
T cell crawling to activation site Color plate 32
(Figure 19.11)