Carbon Nanotubes: Methods and Protocols (Methods in Molecular Biology, 625)

Methods in Molecular Biology™ Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfi...

Author: Kannan Balasubramanian | Marko Burghard

70 downloads 787 Views 18MB Size Report

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

Report copyright / DMCA form

DOWNLOAD PDF

Methods

in

Molecular Biology™

Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK

For other titles published in this series, go to www.springer.com/series/7651

Carbon Nanotubes Methods and Protocols

Edited by

Kannan Balasubramanian Max-Planck-Institut für Festkörperforschung, Stuttgart, Germany

Marko Burghard Max-Planck-Institut für Festkörperforschung, Stuttgart, Germany

Editors Kannan Balasubramanian, Ph.D. Max-Planck-Institut für Festkörperforschung Stuttgart Germany [email protected]

Marko Burghard, Ph.D. Max-Planck-Institut für Festkörperforschung Stuttgart Germany [email protected]

ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-60761-577-4 e-ISBN 978-1-60761-579-8 DOI 10.1007/978-1-60761-579-8 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2010920245 © Humana Press, a part of Springer Science+Business Media, LLC 2010 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Printed on acid-free paper Humana Press is a part of Springer Science+Business Media (www.springer.com)

Preface Since their discovery in 1991, carbon nanotubes (CNTs) have had an enormous impact in materials science. More recently, CNTs have successfully entered the fields of molecular biology, biomedicine, and bioanalytical chemistry. Much of the increasing interest in CNTs is owed to their rare combination of high chemical stability and exceptional optical and electrical properties. Another major factor that has promoted the utilization of CNTs in biological sciences is their unique structure. In fact, their high surface-to-volume ratio and high aspect ratio render them close-to-ideal candidates as active components of biosensors, or as “nanosyringes” enabling the injection of drugs or biological markers into living cells. Over the last couple of years, a wide variety of high-quality CNTs have become commercially available, a fact that has strongly stimulated the recent development of biologyrelated CNT applications. However, the obtainable material often differs in purity, agglomeration state, as well as the length and diameter distribution of the tubes, all of which have a profound influence on important parameters like the tubes’ dispersability and surface properties. It is hence highly desirable to make reliable protocols, which include as many details on the used nanotubes as possible, available to a wide range of readers coming from different fields. We strongly hope that the present collection of protocols will contribute to the achievement of common standards and help to avoid discrepancies in future biology-related CNT studies. This book is organized into five parts. The first part focuses on CNT chemical functionalization approaches, which are required to tackle a major obstacle for using CNTs in biology and medicinal chemistry, namely their inherent hydrophobic character and the resulting lack of solubility in most solvents compatible with the biological milieu. CNT functionalization based upon covalent or non-covalent schemes has proven to be highly effective for enhancing the water solubility of the nanotubes and thus transforming their biocompatibility profile. At the same time, non-covalent functionalization often serves as the basis for further purification of the tubes via centrifugation or chromatography. The second part is devoted to toxicity studies of CNTs. In the meanwhile, it is well established that various types of functionalized CNTs exhibit a capacity to be taken up by a wide range of cells and are able to traffic through different cellular barriers. Recent studies have demonstrated that the cellular uptake of CNTs is largely independent of the nature and density of the appended functional groups, which paves the way for nanotube-tube delivery of a broad range of agents, including proteins, DNA, and synthetic polymers. The intracellular traffic of functionalized CNTs will be the topic of the third part, encompassing three different methodologies. Part 4 deals with modified CNT networks as scaffolds for cell growth, an area that attracts increasing attention due to the future perspective of designing therapies for CNS regeneration or the development of neurochip devices. Finally, Part 5 provides protocols related to CNT-based biosensors, with emphasis on amperometric detection principles. One central topic here is the use of tube coatings to enhance the selectivity of the sensor response toward specific analytes. Stuttgart, Germany

Kannan Balasubramanian Marko Burghard

v

Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v ix

Part I Functionalization 1 Non-covalent Attachment of Proteins to Single-Walled Carbon Nanotubes . . . . . Luís F.F. Neves, Ta-Wei Tsai, Naveen R. Palwai, David E. Martyn, Yongqiang Tan, David W. Schmidtke, Daniel E. Resasco, and Roger G. Harrison 2 Covalent Conjugation of Multi-walled Carbon Nanotubes with Proteins . . . . . . . Changqing Yi, Suijian Qi, Dawei Zhang, and Mengsu Yang 3 Covalently Linked Deoxyribonucleic Acid with Multi-walled Carbon Nanotubes: Synthesis and Characterization . . . . . . . . . . . . . . . . . . . . . . . Weiwei Chen, Changqing Yi, Tzang Chi-Hung, Shuit-Tong Lee, and Mengsu Yang 4 Temperature and pH-Responsive “Smart” Carbon Nanotube Dispersions . . . . . . Dan Wang and Liwei Chen

3

9

19

27

Part II Toxicity 5 Effects of Carbon Nanotubes on the Proliferation and Differentiation of Primary Osteoblasts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Dawei Zhang, Changqing Yi, Suijian Qi, Xinsheng Yao, and Mengsu Yang 6 Carbon Nanotube Uptake and Toxicity in the Brain . . . . . . . . . . . . . . . . . . . . . . . 55 Leying Zhang, Darya Alizadeh, and Behnam Badie 7 In Vitro and In Vivo Biocompatibility Testing of Functionalized Carbon Nanotubes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Gianni Ciofani, Vittoria Raffa, Orazio Vittorio, Alfred Cuschieri, Tommaso Pizzorusso, Mario Costa, and Giuseppe Bardi 8 Real-Time Monitoring of Cellular Responses to Carbon Nanotubes . . . . . . . . . . . 85 Qingxin Mu, Shumei Zhai, and Bing Yan 9 Reducing Nanotube Cytotoxicity Using a Nano-Combinatorial Library Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Qiu Zhang, Hongyu Zhou, and Bing Yan 10 DNA Damage by Carbon Nanotubes Using the Single Cell Gel Electrophoresis Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 Olga Zeni and Maria Rosaria Scarfì

Part III Trafficking 11 Assessment of Cellular Uptake and Cytotoxicity of Carbon Nanotubes Using Flow Cytometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 Khuloud T. Al-Jamal and Kostas Kostarelos

vii

viii

Contents

12 Cell Trafficking of Carbon Nanotubes Based on Fluorescence Detection . . . . . . . 135 Monica H. Lamm and Pu Chun Ke 13 Carbon Nanotubes as Intracellular Carriers for Multidrug Resistant Cells Studied by Capillary Electrophoresis–Laser-Induced Fluorescence . . . . . . . . 153 Ruibin Li, Hanfa Zou, Hua Xiao, and Renan Wu

Part IV Scaffolds 14 Carbon Nanotube-Based Neurochips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 Moshe David-Pur, Mark Shein, and Yael Hanein 15 Effect of Carbon Nanotubes on HepG2 Adhesion and Spreading . . . . . . . . . . . . . 179 Suijian Qi, Changqing Yi, Dawei Zhang, and Mengsu Yang

Part V Biosensors 16 Enzymatic Detection Based on Carbon Nanotubes . . . . . . . . . . . . . . . . . . . . . . . Martin Pumera 17 Carbon Nanotube Biosensors Based on Electrochemical Detection . . . . . . . . . . . Martin Pumera 18 Biosensors Based on Carbon Nanotube-Network Field-Effect Transistors . . . . . . . Cristina C. Cid, Jordi Riu, Alicia Maroto, and F. Xavier Rius 19 Detection of Biomarkers with Carbon Nanotube-Based Immunosensors . . . . . . . Samuel Sánchez, Esteve Fàbregas, and Martin Pumera 20 Carbon Nanotube Biosensors with Aptamers as Molecular Recognition Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hye-Mi So, Dong-Won Park, Hyunju Chang, and Jeong-O Lee Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

197 205 213 227

239 251

Contributors Khuloud T. Al-Jamal • Nanomedicine Laboratory, Centre for Drug Delivery Research, The School of Pharmacy, University of London, London, UK Darya Alizadeh • Division of Neurosurgery, City of Hope National Cancer Center, Duarte, CA, USA Behnam Badie • Division of Neurosurgery, City of Hope National Cancer Center, Duarte, CA, USA Giuseppe Bardi • Institute of Neurosciences, CNR, Via Moruzzi, Pisa, Italy Hyunju Chang • NanoBio Fusion Research Center, Korea Research Institute of Chemical Technology, Jang-dong 100, Eusung-gu, Daejeon 305–343, Korea Liwei Chen • Department of Chemistry and Biochemistry, Ohio University, Athens, OH, USA; Suzhou Institute of Nano Tech and Nano Bionics, Chinese Academy of Science, Suzhou, Jiangsu, P. R. China Weiwei Chen • Department of Physics & Materials Science, City University of Hong Kong, KLT, Hong Kong Tzang Chi-Hung • Department of Biology and Chemistry, City University of Hong Kong, KLT, Hong Kong Cristina C. Cid • Department of Analytical and Organic Chemistry, Universitat Rovira i Virgili, Tarragona, Catalonia, Spain Gianni Ciofani • Scuola Superiore Sant’Anna, Pisa, Italy Mario Costa • Institute of Neurosciences, CNR, Via Moruzzi, Pisa, Italy Alfred Cuschieri • Scuola Superiore Sant’Anna, Pisa, Italy Moshe David-Pur • School of Electrical Engineering, Tel-Aviv University, Tel-Aviv, Israel Esteve Fàbregas • Sensors and Biosensors Group, Analytical Chemistry Department, UAB.Edifici Cn, Bellaterra, Spain Yael Hanein • School of Electrical Engineering, Tel-Aviv University, Tel-Aviv, Israel Roger G. Harrison • School of Chemical, Biological and Materials Engineering, and Carbon Nanotube Technology Center, University of Oklahoma, Norman, OK, USA Pu Chun Ke • Department of Physics and Astronomy, Clemson University, Clemson, SC, USA Kostas Kostarelos • Nanomedicine Laboratory, Centre for Drug Delivery Research, The School of Pharmacy, University of London, London, UK Monica H. Lamm • Department of Chemical and Biological Engineering, Iowa State University, Ames, IA, USA Jeong-O Lee • NanoBio Fusion Research Center, Korea Research Institute of Chemical Technology, Jang-dong 100, Eusung-gu, Daejeon 305–343, Korea Shuit-Tong Lee • Physics & Materials Science, City University of Hong Kong, KLT, Hong Kong Ruibin Li • National Chromatographic R & A Center, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, P.R. China

ix

x

Contributors

Alicia Maroto • Department of Analytical and Organic Chemistry, Universitat Rovira i Virgili, Tarragona, Catalonia, Spain David E. Martyn • School of Chemical, Biological and Materials Engineering, and Carbon Nanotube Technology Center, University of Oklahoma, Norman, OK, USA Qingxin Mu • School of Chemistry and Chemical Engineering, Shandong University, Jinan, P.R. China; St. Jude Children’s Research Hospital, Memphis, TN, USA Luís F.F. Neves • School of Chemical, Biological and Materials Engineering, and Carbon Nanotube Technology Center, University of Oklahoma, Norman, OK, USA Naveen R. Palwai • School of Chemical, Biological and Materials Engineering, and Carbon Nanotube Technology Center, University of Oklahoma, Norman, OK, USA Dong-Won Park • NanoBio Fusion Research Center, Korea Research Institute of Chemical Technology, Jang-dong 100, Eusung-gu, Daejeon 305–343, Korea Tommaso Pizzorusso • Institute of Neurosciences, CNR, Via Moruzzi, Pisa, Italy Martin Pumera • Biomaterial Systems Group, Biomaterials Center and International Center for Materials Nanoarchitectonics (MANA), Tsukuba, Ibaraki, Japan Suijian Qi • Department of Biology and Chemistry, City University of Hong Kong, KLT, Hong Kong Vittoria Raffa • Scuola Superiore Sant’Anna, Pisa, Italy Daniel E. Resasco • School of Chemical, Biological and Materials Engineering, and Carbon Nanotube Technology Center, University of Oklahoma, Norman, OK, USA Jordi Riu • Department of Analytical and Organic Chemistry, Universitat Rovira i Virgili, Tarragona, Catalonia, Spain F. Xavier Rius • Department of Analytical and Organic Chemistry, Universitat Rovira i Virgili, Tarragona, Catalonia, Spain Samuel Sánchez • Sensors and Biosensors Group, Analytical Chemistry Department, UAB. Edifici Cn, Bellaterra, Spain Maria Rosaria Scarfì • CNR-Institute for Electromagnetic Sensing of Environment (IREA), Naples, Italy David W. Schmidtke • School of Chemical, Biological and Materials Engineering, and Carbon Nanotube Technology Center, University of Oklahoma, Norman, OK, USA Mark Shein • School of Electrical Engineering, Tel-Aviv University, Tel-Aviv, Israel Hye-Mi So • NEMS Bio Team, National NanoFab Center, 355 Gwahangno, Yuseong-gu Daejeon 304–806, Korea Yongqiang Tan • School of Chemical, Biological and Materials Engineering, and Carbon Nanotube Technology Center, University of Oklahoma, Norman, OK, USA Ta-Wei Tsai • School of Chemical, Biological and Materials Engineering, and Carbon Nanotube Technology Center, University of Oklahoma, Norman, OK, USA Orazio Vittorio • Scuola Superiore Sant’Anna, Pisa, Italy; Department of Oncology, Transplantation and Advanced Technologies in Medicine, University of Pisa, Pisa, Italy Dan Wang • Department of Chemistry and Biochemistry, Ohio University, Athens, OH, USA

Contributors

xi

Renan Wu • National Chromatographic R & A Center, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, P.R. China Hua Xiao • National Chromatographic R & A Center, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, P.R. China Bing Yan • School of Chemistry and Chemical Engineering, Shandong University, Jinan, P.R. China; St. Jude Children’s Research Hospital, Memphis, TN, USA Mengsu Yang • Department of Biology and Chemistry, City University of Hong Kong, KLT, Hong Kong Xinsheng Yao • Department of Natural Products Chemistry, Shenyang Pharmaceutical University, Shenyang, China Changqing Yi • Biotechnology & Health Centre, Shenzhen Virtual University Park, Shenzhen, China, City University of Hong Kong, KLT, Hong Kong Olga Zeni • CNR-Institute for Electromagnetic Sensing of Environment (IREA), Naples, Italy Shumei Zhai • School of Chemistry and Chemical Engineering, Shandong University, Jinan, P.R. China Dawei Zhang • Department of Biology and Chemistry, City University of Hong Kong, KLT, Hong Kong Leying Zhang • Division of Neurosurgery, City of Hope National Cancer Center, Duarte, CA, USA Qiu Zhang • School of Chemistry and Chemical Engineering, Shandong University, Jinan, Shandong, P. R. China Hongyu Zhou • School of Chemistry and Chemical Engineering, Shandong University, Jinan, P.R. China; St. Jude Children’s Research Hospital, Memphis, TN, USA Hanfa Zou • National Chromatographic R & A Center, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, P.R. China

Part I Functionalization

Chapter 1 Non-covalent Attachment of Proteins to Single-Walled Carbon Nanotubes Luís F.F. Neves, Ta-Wei Tsai, Naveen R. Palwai, David E. Martyn, Yongqiang Tan, David W. Schmidtke, Daniel E. Resasco, and Roger G. Harrison Abstract A method for the non-covalent attachment of proteins to single-walled carbon nanotubes (SWNTs) is described. In this method, the protein is adsorbed to SWNTs that are suspended using sodium cholate, a surfactant and bile salt. The sodium cholate is then removed by dialysis with retention of the protein on the SWNTs. This method has resulted in good protein loadings and good retention of protein activity. Key words: Non-covalent attachment, Adsorption, Dialysis, Proteins, Single-walled carbon nanotubes, Sodium cholate, Surfactant

1. Introduction Single-walled carbon nanotubes (SWNTs) have important optical, thermal, mechanical, and electronic properties (1), and are being developed for applications in various biological systems (2). For some of these biological applications, it is necessary to attach proteins to the SWNTs, for example, to target the SWNTs to specific cells or to construct a biosensor. In this chapter, we provide detailed information on a method that we have used successfully to attach proteins to SWNTs by adsorption, a non-covalent form of attachment (3). This method is advantageous in that important features in the UV-vis-NIR adsorption spectra of the SWNTs are preserved, which is critical in applications where it is desired, for example, that the SWNTs strongly absorb energy when NIR radiation is applied. One problem with direct covalent

K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_1, © Humana Press, a part of Springer Science + Business Media, LLC 2010

3

4

Neves et al.

attachment of molecules to the SWNTs is that this method results in the complete elimination of the UV-vis-NIR adsorption bands, which is thought to be due to the SWNT’s p system being disrupted (4, 5). The adsorption method we will describe was first used for proteins by Graff et al. (6). The first step in this method is to completely suspend the SWNTs in an aqueous solution of sodium cholate, which is a surfactant and a bile salt. After centrifugation, the protein is added, and the suspension is dialyzed using a dialysis membrane that will retain the protein but allow the sodium cholate to pass through. The suspension is centrifuged again, and the supernatant containing the suspended SWNTs with protein adsorbed is retained. This method for us has given good protein loading and good retention of biological activity for horseradish peroxidase (HRP, MW = 40 kDa) (3) and glucose oxidase (MW = 160 kDa) (unpublished data): HRP loading of 2 mg protein/mg SWNTs and 98% retention of native enzyme activity; glucose oxidase loading of 22 mg protein/mg SWNTs and 87% retention of native enzyme activity. Atomic force microscopy (AFM) analysis was performed on SWNTs with HRP adsorbed, and it is possible to visualize the HRP adsorbed on the nanotubes (Fig. 1) (3). In addition, for SWNTs with HRP adsorbed, there was retention of a substantial fraction of the NIR absorption at 980 nm (Fig. 2) (3). Other methods that have been used to adsorb proteins on SWNTs are the organic solvent displacement method (7) and the

Fig. 1. AFM images showing SWNT with sodium cholate and HRP protein. (a) SWNT in sodium cholate after sonication and centrifugation. Arrows indicate solid sodium cholate with a height of 1.0–1.5 nm (does not include 0.8 nm SWNT height). (b) SWNT/protein after dialysis and centrifugation. Arrows indicate protein associated with SWNT. The height is 3.8–6.0 nm (does not include 0.8 nm SWNT height). Reproduced from ref. (3) with permission from IOP Publishing Ltd.

Non-covalent Attachment of Proteins to Single-Walled Carbon Nanotubes

5

Fig. 2. UV-vis-NIR absorption spectra of the pristine SWNTs suspended in sodium cholate and the SWNT/HRP complex after the final centrifugation. Reproduced from ref. (3) with permission from IOP Publishing Ltd.

aqueous sonication method (8); however, these methods gave problems with complete dispersion of SWNTs and with the maintenance of protein activity or structure.

2. Materials 2.1. SWNT Preparation and Suspension

1. Purified and freeze dried CoMoCAT SG65 SWNTs, rich in (6,5) type with an average diameter of 0.75 nm, were provided by Southwest Nanotechnologies, Inc. (Norman, OK). 2. Sodium cholate (Sigma-Aldrich) was used as a dispersant. 3. A horn sonicator equipped with a microtip of 3.2 mm in diameter was used (CPX750, Cole Parmer). 4. An ultracentrifuge was used to centrifuge the SWNT suspension (Optima XL series preparative ultracentrifuge, Beckman Coulter).

2.2. Adsorption of Proteins on SWNTs

1. Sodium phosphate (20 mM, pH 7.4) for buffering. 2. Dialysis membranes (10 kDa and 100 kDa, Spectrum Laboratories). 3. Bradford protein assay kit (Quick Start, Bio-Rad).

6

Neves et al.

3. Methods 3.1. SWNT Preparation and Suspension

1. Add 3.0 mg of SWNTs to 7 mL of a 2 wt % sodium cholate solution and sonicate at a power of 7 W for 30 min. 2. Centrifuge the resulting suspension at 29,600 × g for 30 min (see Note 1).

3.2. Adsorption of Proteins on SWNTs

1. Mix sodium phosphate buffer with the SWNT suspension. 2. Add 20 mg of protein to the suspension at 4°C. 3. Dialyze the solution at 4°C for 12 h against 2 L of sodium phosphate buffer using a 10 kDa dialysis membrane to remove unadsorbed sodium cholate (see Note 2). 4. Transfer the solution to a 100 kDa dialysis membrane and then dialyze at 4°C against 2 L of sodium phosphate buffer, with a change of the 2 L of buffer at 2, 4, 16, and 24 h from the start of dialysis (see Note 3). Carry out the final dialysis for 4 h. 5. Centrifuge the solution at 29,600 × g for 1 h at 4°C. 6. Save the supernatant (see Note 4). 7. Perform the Bradford protein assay in order to quantify the protein concentration in the solution (see Note 5). Also, measure the protein concentration in the dialysate (the solution on the outside of the dialysis membrane). 8. Measure the absorbance at 800 nm of the initial suspension (SWNTs dispersed using sodium cholate) and the final suspension (SWNT/protein complex) for determination of the SWNT concentration (see Note 6).

4. Notes 1. The centrifugation step that is performed during the production of the suspension is important, since it eliminates nanotubes aggregates. The presence of aggregates in suspension is undesirable; the existence of aggregates can contribute to significant changes in the properties of the suspension. During the centrifugation step, it is important to use at least 29,600 × g centrifugal force when using single-walled carbon nanotubes, in order to assure a good nanotube dispersion of single-walled nanotubes, and to remove any aggregated nanotubes. 2. The reason for the use of a 10 kDa dialysis membrane is because the sodium cholate (MW = 431 Da) will be able to

Non-covalent Attachment of Proteins to Single-Walled Carbon Nanotubes

7

pass through the membrane pores freely and therefore will permit the removal of unadsorbed sodium cholate. Proteins with molecular weights higher than 10 kDa will be retained inside the membrane. 3. A membrane with a larger pore size (100 kDa molecular weight cutoff) is used in the second dialysis, in order to remove unadsorbed protein through the membrane. Extensive dialysis is used to ensure complete removal of unadsorbed protein; less dialysis can possibly be used for some proteins, especially low molecular weight proteins. The removal of unadsorbed protein can be evaluated by measuring the protein concentration in the dialysate. For low protein concentrations, it may be necessary to use a micro protein assay (for example, the micro BCA protein assay from Pierce). It is important to note that the membrane pore size should always be larger than the protein molecular size. Different membrane pore sizes are commercially available for the dialysis of proteins with a range between 10 and 300 kDa. It is also important to note that higher than a 100 kDa molecular weight cutoff of the membrane can lead to some loss of nanotubes through the membrane. 4. The initial suspension (SWNT dispersed using sodium cholate) and the final suspension (SWNT/protein complex) can be analyzed by atomic force microscopy (see the AFM images in Fig. 1 for the adsorption of HRP using this procedure). 5. The amount of protein used gives a relatively low coverage of the SWNT surface area. For example, using the weights of SWNTs and protein in Subheading 3 and assuming the protein is HRP, the ratio of the number of protein molecules to the number of six-carbon groups in the SWNTs is calculated to be 0.012. Assuming the protein is globular, the diameter of the protein is approximately 7.5 times the diameter of a SWNT. These calculations are verified by the AFM image of HRP adsorbed on SWNTs shown in Fig. 1.1, where the coverage by the protein on the SWNTs is relatively sparse and the protein is much larger than a SWNT. Seven and a half times the SWNT average height of 0.75 nm is 5.6 nm, which falls within the 3.8–6.0 nm height of the HRP protein measured. 6. A wavelength of 800 nm is used to measure the concentration of the SWNTs, since bands are not present on the absorption spectra at this wavelength (see Fig. 2 for SWNTs with HRP adsorbed). A calibration curve can be made from a plot of absorbance at 800 nm versus SWNT concentration.

8

Neves et al.

Acknowledgments This work was supported by the U.S. Department of Energy-Basic Energy Sciences (DE-FG02-06ER64239), the U.S. Department of Defense Breast Cancer Research Program (W81XWH07-1-0536), and the Foundation for Science and Technology of Portugal (Luis Neves).

References 1. Han J (2005) In: Meyyappan M (ed) Structures and properties of carbon nanotubes. CRC, Moffett Field, CA, pp 2–24 2. Lin Y, Taylor S, Li H, Fernando KAS, Qu L, Wang W, Gu L, Zhou B, Sun YP (2004) Advances toward bioapplications of carbon nanotubes. J Mater Chem 14:527–541 3. Palwai NR, Martyn DE, Neves LFF, Tan Y, Resasco DE, Harrison RG (2007) Retention of biological activity and near-infrared absorbance upon adsorption of horseradish peroxidase on single-walled carbon nanotubes. Nanotechnology 18:235601/1-235601/5 4. Buffa F, Hu H, Resasco DE (2005) Side-wall functionalization of single-walled carbon nanotubes with 4-hydroxymethylaniline followed by polymerization of e-caprolactone. Macromolecules 38:8258–8263

5. Bahr JL, Yang J, Kosynkin DV, Bronikowski MJ, Smalley RE, Tour JM (2001) Functionalization of carbon nanotubes by electrochemical reduction of aryl diazonium salts: a bucky paper electrode. J Am Chem Soc 123:6536–6542 6. Graff RA, Swanson JP, Barone PW, Baik S, Heller DA, Strano MS (2005) Achieving individual-nanotube dispersion at high loading in single-walled carbon nanotube composites. Adv Mater 17:980–984 7. Karajanagi SS, Vertegel AA, Kane RS, Dordick JS (2004) Structure and function of enzymes adsorbed onto single-walled carbon nanotubes. Langmuir 20:11594–11599 8. Matsuura K, Saito T, Okazaki T, Ohshima S, Yumura M, Iijima S (2006) Selectivity of watersoluble proteins in single-walled carbon nanotube dispersions. Chem Phys Lett 429:497–502

Chapter 2 Covalent Conjugation of Multi-walled Carbon Nanotubes with Proteins Changqing Yi, Suijian Qi, Dawei Zhang, and Mengsu Yang Abstract Linkage of proteins to carbon nanotubes (CNTs) is fundamentally important for applications of CNTs in medicinal and biological fields, as well as in biosensor or chemically modulated nanoelectronic devices. In this contribution, we provide a detailed protocol for the synthesis and characterization of covalent CNT-protein adducts. Functionalization of multiwalled carbon nanotubes (MWCNTs) with proteins has been achieved by the initial carboxylation of MWCNTs followed by amidation with the desired proteins. Attenuated total reflection Fourier transform infrared (ATR-FTIR) and X-ray photoelectron spectroscopy (XPS) measurements validated the presence of a covalent linkage between MWCNTs and proteins. The visualization of proteins on the surface of MWCNTs was furthermore achieved using atomic force microscopy (AFM). The protein-conjugated nanocomposites can also be assembled into multidimensional addressable heterostructures through highly specific biomolecular recognition system (e.g., antibody–antigen). Keywords: Carbon nanotubes, Covalent conjugation, Protein, Antibody, ATR-FTIR, XPS, AFM

1. Introduction The unique properties of nanomaterials (NMs) in combination with the biorecognition abilities of proteins offer particularly exciting opportunities in molecular imaging (1–3), therapy (4, 5), biomolecule delivery (6, 7), and design of functional nanodevices (8–12). The loss or retention of the native structure of proteins upon their conjugation onto NMs provides an additional variable for controlling NM assembly – intercomponent spacing (13). Though biomolecules, such as DNA and proteins, can be linked to nanotubes via noncovalent interactions (14, 15), the use of covalent chemistry is expected to provide better stability, accessibility, and selectivity (16, 17).

K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_2, © Humana Press, a part of Springer Science + Business Media, LLC 2010

9

10

Yi et al.

Herein, we report a protocol for covalent conjugation of proteins to carbon nanotube (CNT) surface, which has been achieved by the initial carboxylation of CNTs followed by amidation with the desired proteins. This two-step chemical method employs mild conditions and results in tethering of the organic functionality through a covalent bond. This protocol can also be employed to functionalize CNTs with various proteins and amines, including primary amines and secondary amines. The proteinconjugated nanocomposites can be further assembled into multidimensional addressable heterostructures through highly specific biomolecular recognition system, such as antibody– antigen interation.

2. Materials 2.1. Carboxylation of MWCNTs

1. Pristine MWCNTs, prepared by the chemical vaporization deposition (CVD) method (Nanotech Port, China). 2. 10-µm pore size PTFE filter paper (Advantec MFS, Inc.). 3. Ultrasonic cleaning bath (Electron Microscopy Sciences).

2.2. Amidation

1. Mouse monoclonal IgG (Santa Cruz Biotechnology). 2. 1-ethyl-3-(3-dimethyl-aminopropyl) carbodiimide (EDC) (Sigma-Aldrich). 3. N-hydroxysuccinimide (NHS) (Sigma-Aldrich). 4. Phosphate buffer solution (PBS): 0.1 mol/L, pH 7.4.

2.3. Equipment

1. Transmission electron microscope (TEM), Tecnai 12 (Philips). 2. Copper grids with formvar film (Electron Microscopy Sciences). 3. ULVAC-PHI 5802 XPS system (Kanagawa, Japan). 4. FTIR spectrometer Spectrum One (Perkin Elmer). 5. FTIR microscope equipped with a HgCdTe detector cooled with liquid nitrogen (i-Series). 6. Multimode atomic force microscope (Veeco Instruments).

3. Methods (See Notes 1 and 2) 3.1. Carboxylation

1. Pristine MWCNT powder was oxidized in a 3:1 mixture of concentrated H2SO4 (98%) and HNO3 (69%) at 70°C for 4 h (18–20) and filtered through a 10-µm pore size PTFE filter paper. Or

Covalent Conjugation of Multi-walled Carbon Nanotubes with Proteins

11

Pristine MWCNTs were refluxed in 4 M HNO3 for 24 h and filtered through a 10-µm pore size PTFE filter paper. 2. After filtration, the refluxed MWCNTs were exposed to 1 M HCl and sonicated for about 30 min. 3. The carboxylated MWCNTs were filtered, and washed with deionized water and dried in air. 3.2. Amidation

1. MWCNT-COOH and mouse monoclonal IgG was mixed in a tube at the ratio 1:10. 2. A mixture solution of 0.40 M 1-ethyl-3-(3-dimethyl-aminopropyl) carbodiimide (EDC) and 0.10 M N-hydroxy succinimide (NHS) was added to initiate cross-linking reactions between carboxyl groups on MWCNTs and amine groups in IgG. 3. The reaction tubes were rotated at room temperature for 1 h. 4. After centrifugation and decanting supernate, mouse IgG functionalized MWCNTs were re-suspended in PBS buffer.

3.3. TEM Characterization

1. 5 mL of dilute aqueous sample was spotted onto a grid and left for 90 s. 2. Lightly touch one edge of the grid with filter paper to wipe off moisture. 3. Grids were then allowed to air dry prior to analysis. 4. TEM: Observe the samples on grids in electron microscopy with an accelerating voltage of 120 kV. Figure 1 shows the results obtained.

3.4. XPS Measurements (See Notes 3 and 4)

1. Parameter setting: (a) Pressure ranges are as follows: 2 × 10−6 mbar (fast entry chamber), 4 × 10−8 mbar (preparation chamber), and 4 × 10−9 mbar (sample analysis chamber). (b) High transmission FAT mode, 14.12 keV, 25 mA, Al Ka (1,486.7 eV), was used for the analysis at 90° electron take off angle for normal non-charging samples (45° for the charging samples). (c) The analyzer slit width was set for 0.8 mm, and the resulting overall energy resolution was 0.35 eV. 2. The SCIENTA software was used for data acquisition and data analysis. 3. The binding energy of the C1s of graphite, 284.5 eV (±0.35 eV energy resolution of the spectrometer at the settings employed) was taken as the reference. 4. Prior to individual elemental scans, a survey scan was taken for all the samples in order to detect the elements present.

12

Yi et al.

Fig. 1. (a) and (b) TEM images of the carboxylated MWCNTs

5. The obtained XPS spectra of CNT-protein adduct are listed in Fig. 2. The C1s XPS of the oxidized nanotubes in Fig. 2a shows a large peak at 284.4 eV from the nanotubes, a smaller peak at 286.4 eV, and a well-separated peak at 288.8 eV. The peak at 288.8 eV is attributed to the carbonyl group in the carboxylic acid group. The peak at 286.4, 2 eV higher than the main peak, is attributed to C atoms in ether-like linkages. The corresponding N(1s) spectrum shows no signal above the detection limit of the instrument, even with extensive signal averaging. After conjugation with proteins, the modified tubes were characterized by XPS after briefly warming to 75°C in an ultrahigh vacuum to remove any residual physically adsorbed proteins. Compared to the main bulk C1s at 284.4 eV, the resulting C1s photoelectron spectrum shows some narrowing of the bulk peak (Fig. 2b). We notice that there is no significant intensity near 288.8 eV. The absence of intensity at 288.8 eV is important because the C1s binding energy of carboxylic groups is expected to decrease significantly when a carboxylic acid group is converted to a carbonyl amide, which is at 288.45 eV. Carbon atoms in carboxylic acid groups and in carbonyl amide groups typically have C1s binding energies ~4.0 eV and ~3.0 eV higher, respectively, than C atoms in alkanes (21). Thus, the changes we observe in the C1s spectrum support the formation of a carbonyl amide linkage to the nanotubes. The N1s spectrum shows a peak

Covalent Conjugation of Multi-walled Carbon Nanotubes with Proteins

13

Fig. 2. XPS of chemically modified CNTs. C1s spectrum of CNT-COOH (a); C1s spectrum of protein functionalized CNTs (b), showing elimination of the carboxylic peak and increased peak from the amide group. N1s spectrum of protein functionalized MWCNTs (c)

with a binding energy of 399.7 eV (Fig. 2c). Previous studies have shown that amides have binding energies in the range of ~399.5–400.2 eV (21). Therefore, the peak energies are consistent with the formation of the amide bonds. 3.5. ATR-FTIR Measurements (See Notes 5 and 6)

1. Sample preparation: MWCNT-protein adducts were grinded to a fine powder. 2. Sample loading: The ZnSe through top-plate of the horizontal ATR was covered by a fine layer of the grinded sample, avoiding air bubble formation on the crystal surface. 3. Parameters setting: A detector gain of 1 and a speed of the moving mirror of 0.6 cm−1 were employed.

14

Yi et al.

4. Spectra measurements: The spectra were obtained in the absorbance mode from 4,000 to 700 cm−1 by accumulating 100 scans, working with a spectral resolution of 2 cm−1. 5. Control: Absorbance spectra were corrected versus a spectrum of distilled water, obtained in the same instrumental conditions. 6. Figure 2.3 shows the typical ATR-FTIR spectra of MWCNTCOOH before and after functionalization by IgG. The absorption peak at 1,569 cm−1 indicates the presence of amide bond which comes from the covalent linkage between CNTs and IgG through the functional groups. 3.6. AFM Measurements

1. MWCNT-protein adducts suspension: Suspend MWCNTprotein adducts in PBS to a concentration of 1 mg/mL. This concentration provides convenient coverage for AFM imaging and may be used for a variety of similar size samples. 2. Prepare mica: Cleave a fresh mica surface by first pressing some adhesive tape against the top mica surface, then peeling off the tape. Glue mica to a small puck (e.g., using epoxy). 3. Deposit sample solution on mica: Deposit 50 mL of protein solution on the freshly cleaved mica. 4. Sample to bind to substrates: Allow 20–30 min for the sample to bind to the mica substrate. Binding time may vary with different samples (it can be up to 24 h). 5. Rinse unbound sample: Rinse the sample with a large quantity of buffer to remove unbound protein. 6. AFM measurements: AFM measurements were performed under ambient conditions using a NanoScope V Controller

Fig. 3. ATR-FTIR spectra of CNTs. The absorption peak at 1,569 cm−1 indicates the presence of amide linkage between MWCNTs and IgG

Covalent Conjugation of Multi-walled Carbon Nanotubes with Proteins

15

with a multimode atomic force microscope in tapping mode with standard 125 mm single-crystal silicon cantilevers. 7. Figure 4 shows the herein obtained AFM images: MWCNTs appear as bright lines in the image and the circle particles represent bound proteins. Proteins are connected to the sidewall of nanotubes, indicating that oxidization took place in the defect sites of sidewalls. From the observations of several samples, we conclude that protein attachment mainly occurred at nanotube sidewalls, because chemical functionalization occurred primarily at the sidewalls. This AFM image is similar to that of cytochrome c-functionalized purified SWCNTs which was taken by Davis and co-workers (14).

Fig. 4. (a) and (b) AFM (tapping mode) images of protein-MWCNT adducts. MWCNTs appear as bright lines and the circle particles represent bound proteins

16

Yi et al.

4 Notes 1. Unless stated otherwise, water used in this protocol was Milli-Q deionized water. 2. Unless stated otherwise, the CNT suspensions should be prepared in D.I water and followed by ultrasonication for 30 min. 3. Use only polyethylene gloves. Other gloves may contain silicones that can contaminate the surface. 4. Make sure everything, used to handle or store your samples, is clean (tweezers, etc.). It is recommended to have a dedicated set of clean tools for handling your samples. In particular, take care to avoid grease, oils, and silicone contaminants around your tools and work area. A general cleaning protocol that often works is to clean the utensils that will handle samples with the following solvents (in this order): Hexanes, Methylene chloride, Methanol, and Acetone. 5. The sample must be in direct contact with the ATR crystal, because the evanescent wave or bubble only extends beyond the crystal 0.5–5 mm. 6. The refractive index of the crystal must be significantly greater than that of the sample or else internal reflectance will not occur – the light will be transmitted rather than internally reflected in the crystal.

Acknowledgments The financial support of Key Laboratory Funding Scheme of Shenzhen Municipal Government, BTC operation fund (CityU project No. 9683001) and City University of Hong Kong (Project No. 7002100) are gratefully acknowledged.

References 1. Loo C, Hirsch L, Lee MH, Chang E, West J, Halas N, Drezek R (2005) Gold nanoshell bioconjugates for molecular imaging in living cells. Opt Lett 30:1012–1014 2. Medintz IL, Uyeda HT, Goldman ER, Mattoussi H (2005) Quantum dot bioconjugates for imaging, labelling and sensing. Nat Mater 4:435–446

3. Lewis JD, Destito G, Zijlstra A, Gonzalez MJ, Quigley JP, Manchester M, Stuhlmann H (2006) Viral nanoparticles as tools for intravital vascular imaging. Nat Med 12:354–360 4. Loo C, Lowery A, Halas N, West J, Drezek R (2005) Immunotargeted nanoshells for integrated cancer imaging and therapy. Nano Lett 5:709–711

Covalent Conjugation of Multi-walled Carbon Nanotubes with Proteins 5. Kam NWS, O’Connell M, Wisdom JA, Dai HJ (2005) Carbon nanotubes as multifunctional biological transporters and near-infrared agents for selective cancer cell destruction. Proc Natl Acad Sci USA 102:11600–11605 6. Pantarotto D, Briand JP, Prato M, Bianco A (2004) Translocation of bioactive peptides across cell membranes by carbon nanotubes. Chem Commun 16–17 7. Kam NWS, Jessop TC, Wender PA, Dai HJ (2004) Nanotube molecular transporters: internalization of carbon nanotube-protein conjugates into mammalian cells. J Am Chem Soc 126:6850–6851 8. Connolly S, Fitzmaurice D (1999) Programmed assembly of gold nanocrystals in aqueous solution. Adv Mater 11:1202–1205 9. Caswell KK, Wilson JN, Bunz UHF, Murphy CJ (2003) Preferential end-to-end assembly of gold nanorods by biotin-streptavidin connectors. J Am Chem Soc 125:13914–13915 10. Lee JA, Govorov AO, Dulka J, Kotov NA (2004) Bioconjugates of CdTe nanowires and Au nanoparticles: Plasmon-exciton interactions, luminescence enhancement, and collective effects. Nano Lett 4:2323–2330 11. Lee J, Govorov AO, Kotov NA (2005) Bioconjugated superstructures of CdTe nanowires and nanoparticles: multistep cascade Forster resonance energy transfer and energy channeling. Nano Lett 5:2063–2069 12. Wang S, Mamedova N, Kotov NA, Chen W, Studer J (2002) Antigen/antibody immunocomplex from CdTe nanoparticle bioconjugates. Nano Lett 2:817–822 13. Srivastava S, Verma A, Frankamp BL, Rotello VM (2005) Controlled assembly of protein-

14.

15.

16.

17.

18.

19.

20.

21.

17

nanoparticle composites through protein surface recognition. Adv Mater 17:617–621 Azamian BR, Davis JJ, Coleman KS, Bagshaw CB, Green MLH (2002) Bioelectrochemical single-walled carbon nanotubes. J Am Chem Soc 124:12664–12665 Shim M, Kan NWS, Chen RJ, Dai HJ (2002) Functionalization of carbon nanotubes for biocompatibility and biomolecular recognition. Nano Lett 2:285–288 Chen WW, Tzang CH, Tang JX, Yang MS, Lee ST (2005) Covalently linked deoxyribonucleic acid with multiwall carbon nanotubes: synthesis and characterization. Appl Phys Lett 86:103114 Sarah EB, Cai W, Lasseter TL, Weidkamp KP, Hamers RJ (2002) Covalently bonded adducts of deoxyribonucleic acid (DNA) oligonucleotides with single-wall carbon nanotubes: synthesis and hybridization. Nano Lett 2:1413–1417 Yi CQ, Fong CC, Zhang Q, Lee ST, Yang MS (2008) The structure and function of ribonuclease A upon interacting with carbon nanotubes. Nanotechnology 19:095102 Yi CQ, Fong CC, Chen WW, Qi SJ, Tzang CH, Lee ST, Yang MS (2007) Interactions between carbon nanotubes and DNA polymerase and restriction endonucleases. Nano technology 18:025102 Zhang DW, Yi CQ, Zhang JC, Chen Y, Yao XS, Yang MS (2007) Effects of carbon nanotubes on the proliferation and differentiation of primary osteoblasts. Nanotechnology 18: 475102 Moulder JF, Stickle WF, Sobol PE, Bomben KD (1992) Handbook of X-ray photoelectron spectroscopy. Perkin-Elmer, Eden Prairie, MN

Chapter 3 Covalently Linked Deoxyribonucleic Acid with Multi-walled Carbon Nanotubes: Synthesis and Characterization Weiwei Chen, Changqing Yi, Tzang Chi-Hung, Shuit-Tong Lee, and Mengsu Yang Abstract In this chapter, a multi-step protocol for covalently linking functionalized multi-walled carbon nanotubes (MWCNT) to deoxyribonucleic acid (DNA) oligonucleotides is provided. X-ray photoelectron spectroscopy (XPS) is used to characterize the initially formed amine-terminated MWCNTs, to which DNA is covalently anchored. Atomic force microscopy (AFM) investigation of the DNA–MWCNT conjugates reveals that the chemical functionalization occurs at both the ends and sidewalls of the nanotubes. The described methodology represents an important step toward the realization of DNA-guided self-assembly for carbon nanotubes. Key words: Carbon nanotubes, Covalent conjugation, DNA, ATR-FTIR, XPS, AFM

1. Introduction The discovery of carbon nanotubes in 1991 (1) and their subsequent production in bulk quantities (2) have paved the way to the exploration of the physical, chemical, and biological properties of single-walled carbon nanotubes (SWCNTs) and multiwalled carbon nanotubes (MWCNTs). Due to their unique electronic, chemical, and mechanical properties, carbon nanotubes (CNTs) have shown great promises in biosensing, tissue engineering, and biomedical applications (3–5). Many of the interesting and unique properties of nanoscale materials are realized when they are integrated into more complex assemblies (6, 7). In biology, the highly selective binding between complementary sequences of deoxyribonucleic acid (DNA) plays the central role in genetic replication. This selectivity can, in principle, K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_3, © Humana Press, a part of Springer Science + Business Media, LLC 2010

19

20

Chen et al.

be used to assemble a wide range of materials, by forming adducts between DNA and the materials of interest (8). While biomolecules like DNA can be linked to nanotubes via noncovalent interactions (9, 10), the use of covalent chemistry is expected to provide better stability, accessibility, and selectivity during competitive hybridization (11). In this chapter, a multi-step protocol to covalently link functionalized MWCNTs to deoxyribonucleic acid (DNA) oligonucleotides is demonstrated. And we combine the use of X-ray photoelectron spectroscopy (XPS) and atomic force microscopy (AFM) to achieve the chemical analysis of nanotube functionalization, as well as the direct visualization of the DNA–MWCNT adducts that were obtained from our developed multi-step method.

2. Materials 2.1. Carboxylation of MWCNTs

1. Pristine MWCNTs, prepared by the chemical vaporization deposition (CVD) method (Nanotech Port, China). 2. 10-µm pore size PTFE filter paper (Advantec MFS, Inc.). 3. Ultrasonic cleaning bath (Electron Microscopy Sciences). 4. Milli-Q system (Millipore).

2.2. CNT Functionalization with Poly-l-lysine

1. Poly-l-lysine (Sigma-Aldrich). 2. 1-[3-(dimethyl aminopropyl)]-3-ethylcarbodiimide hydrochloride (Sigma-Aldrich). 3. N-hydroxysuccinimide (NHS) (Sigma-Aldrich). 4. EtONH4 (Sigma-Aldrich). 5. Phosphate buffer solution (PBS): 0.1 M, pH 7.4.

2.3. Linkage of DNA Strands to MWCNTs

1. A pair of primers, HBB-1 (5¢ Amine-AGGGTTGGCCAAT CTACTCC-3¢) and HBB-2 (5¢ Amine-TCTCCCCTTCCTA TGACATGA-3¢) (Invitrogen). 2. PCR reaction solution containing genomic DNA, the pair of primers, reaction buffer, MgCl2, dNTPs, and Taq DNA polymerase (Roche, Basel Switzerland). 3. Biophotometer (Eppendorf, Germany).

2.4. Equipment

1. ULVAC-PHI 5802 XPS system (Kanagawa, Japan). 2. Multimode atomic force microscope (Veeco Instruments). 3. Mica substrates (Ted Pella, Inc.).

Covalently Linked Deoxyribonucleic Acid with Multi-walled Carbon Nanotubes

3. Methods (See Notes 1 and 2 and Fig. 1) 3.1. Carboxylation of MWCNTs

21

1. Pristine MWCNT powder was oxidized in a 3:1 mixture of concentrated H2SO4 (98%) and HNO3 (69%) at 70°C for 4 h and filtered through a 10-µm pore size PTFE filter paper. Or Pristine MWCNTs were refluxed in 4 M HNO3 for 24 h and filtered through a 10-µm pore size PTFE filter paper. 2. After filtration, the refluxed MWCNTs were exposed to 1 M HCl and sonicated for about 30 min. 3. The carboxylated MWCNTs were filtered, washed with deionized (D.I) water, and dried in air.

3.2. Functionalization of CNTs with Poly-l-lysine

1. The purified, oxidized MWCNTs were suspended in 0.4 mL 1-[3-(dimethyl aminopropyl)]-3-ethylcarbodiimide hydrochloride and 0.1 mL N-hydroxysuccinimide in an ultrasonic bath. 2. After centrifugation, 300 µL poly-l-lysine (0.1%) was added and the mixture was stirred for 1 h at 25°C. 3. This dispersion was centrifuged and washed two times with EtONH4 (pH = 8.5) to remove excess poly-l-lysine. 4. After rinsing with D.I water for three times, the mixture was placed in a 55°C oven for 15 min to form the amine-terminated nanotubes.

Fig. 1. Scheme showing the steps involved in the fabrication of covalently linked DNAnanotube adducts. Reprinted with the permission from ref. (14), copyright 2005 American Institute of Physics

22

Chen et al.

3.3. Linkage of DNA Strands to MWCNTs

1. Linking DNA strands to the nanotube requires specially prepared DNA strands. An amine-terminal labeled 684 base pairs DNA was amplified from human b-hemoglobin (HBB) gene by polymerase chain reaction (PCR). 2. A pair of primer, HBB-1 (5¢ Amine-AGGGTTGGC CAATCTACTCC-3¢) and HBB-2 (5¢ Amine-TCTCCC CTTCCTATGACATGA-3¢), was designed to amplify the 5¢ amine labeled 684 base pairs PCR product from the template of exon 1 + 2 of the HBB gene. 3. Each 50 µL PCR reaction solution contained 100–200 ng of genomic DNA, 0.2–0.4 µM of the pair of primers, 1× reaction buffer, 1.5 mM MgCl2, 200 µM dNTPs, and 5.0 U Taq DNA polymerase. 4. After the reaction mixture was set up in a thin-walled PCR tube, the thermal cycling was carried out as follows: 95°C for 2 min, followed by 30 cycles of denaturation at 94°C for 30 s, annealing at 60°C for 30 s and extension at 72°C for 30 s, and a last extension step of 5 min at 72°C. 5. The PCR products were precipitated with 0.2 × volume of 3 M sodium acetate, pH = 4.5, and 2 × volume of 100% ethanol, and washed with 75% ethanol. 6. The purified PCR products were quantified by a photometer. 7. Poly-l-lysine-treated nanotubes were then mixed with 30 µL amine labeled DNA (100 ng/µL), followed by shaking at 26°C for 30 min. 8. After baking at 80°C, the product was rehydrated through water vapor for 30 s. 9. The sample was immersed into a mixture of methanol and acetic acid (3:1) for 5 min and baked again for 30 min. 10. Steps 7 and 8 were repeated twice to ensure that the incorporated DNA was removed from the DNA–MWCNTs adduct.

3.4. XPS Analysis (See Notes 3 and 4)

1. Parameter setting: (a) Pressure ranges are as follows: 2 × 10−6 mbar (fast entry chamber), 4 × 10−8 mbar (preparation chamber), and 4 × 10−9 mbar (sample analysis chamber). (b) High transmission FAT mode, 14.12 keV, 25 mA, Al Ka (1,486.7 eV) was used for the analysis at 90° electron take off angle for normal non-charging samples (45° for the charging samples). (c) The analyzer slit width was set for 0.8 mm and the resulting overall energy resolution was 0.35 eV. 2. The SCIENTA software was used for data acquisition and data analysis.

Covalently Linked Deoxyribonucleic Acid with Multi-walled Carbon Nanotubes

23

3. The binding energy of the C1s of graphite, 284.5 eV (±0.35 eV energy resolution of the spectrometer at the settings employed) was taken as the reference. 4. Prior to individual elemental scans, a survey scan was taken for all the samples in order to detect the elements present. 5. The obtained XPS spectra of oxidized nanotubes are listed in Fig. 2. The spectrum shows a large peak at 284.4 and 285 eV

Fig. 2. XPS of chemically modified nanotubes. (a) C (1s) spectrum of oxidized nanotubes (upper ), along with fit to three peaks. C (1s) spectrum of oxidized nanotubes functionalized with poly-l-lysine (lower ), showing elimination of the carboxylic peak and increased peak from the amide group. (b) N(1s) spectrum of oxidized nanotubes functionalized with poly-l-lysine (upper ); N (1s) spectrum of oxidized nanotubes (lower ), showing no detectable N (1s) signal. Reprinted with the permission from ref. (14), copyright 2005 American Institute of Physics

24

Chen et al.

from the nanotubes, a smaller peak at 286.4 eV, and a well-separated peak at 288.6 eV. The peak at 288.6 eV is attributed to the carbonyl group in the carboxylic acid group. The smaller peak at 285 eV is attributed to the satellite peaks arising from excitation of electronic transitions within the nanotubes. The peak at 286.4, 2 eV higher than the main peak, is attributed to C atoms in ether like linkages. The corresponding N(1s) spectrum (Fig. 2b) shows no signal above the detection limit of the instrument. 6. The amine-terminated nanotubes were characterized by XPS after briefly warming to ~75°C in an ultrahigh vacuum to remove any residual physically adsorbed amines. Compared to the main bulk C(1s) at 284.4 eV (as in Fig. 3.2a upper figure), the resulting C(1s) photoelectron spectrum (Fig. 3.2a lower one) shows some narrowing of the bulk peak and three shoulders at 285.05, 286.3, and 288.15 eV. There is no significant intensity near 288.6 eV. The absence of intensity at 288.6 eV is important because the C(1s) binding energy of carboxylic groups is expected to decrease significantly when a carboxylic acid group is converted to a carbonyl amide, which is at 288.15 eV. The shoulders at 285.05 and 286.3 eV are both attributed to the C atoms in alkanes; the higher 286.3 eV is due to electron donation from the adjacent N atom. Carbon atoms in carboxylic acid groups and in carbonyl amide groups typically have C(1s) binding energies, ~4.0 eV and ~3.0 eV higher, respectively, than C atoms in alkanes (12). Thus, the changes observed in the C(1s) spectrum support the formation of a carbonyl amide linkage to the nanotubes. The N(1s) spectrum (Fig. 3.2b lower figure) shows a peak with a binding energy of 399.8 eV. Previous studies have shown that amides have binding energies in the range of ~399.5–400.2 eV (12, 13). Therefore, the peak energies are consistent with the formation of the product depicted in Fig. 3.1a lower figure. 3.5. AFM Characterization

1. DNA–MWCNT adducts suspension: Suspend DNA–MWCNT adducts in PBS to a concentration of 1 mg/mL. This concentration provides convenient coverage for AFM imaging and may be used for a variety of similar size samples. 2. Prepare mica: Cleave a fresh mica surface by first pressing some adhesive tape against the top mica surface, then peeling off the tape. Glue mica to a small puck (e.g., using epoxy). 3. Deposit sample solution on mica: Deposit 50 mL of DNA– MWCNT adducts solution on the freshly cleaved mica with 10 mM MgCl2. 4. Sample to bind to substrates: Allow 20–30 min for the sample to bind to the mica substrate. Binding time may vary with different samples (it can be up to 24 h).

Covalently Linked Deoxyribonucleic Acid with Multi-walled Carbon Nanotubes

25

Fig. 3. (a) and (b) AFM (tapping mode) images of DNA–MWCNT adducts. MWCNTs appear as bright lines and the paler strands represent bound DNA. Reprinted with the permission from ref. (14), copyright 2005 American Institute of Physics

5. Rinse unbound sample: Rinse the sample with a large quantity of buffer to remove unbound protein. 6. AFM measurements: AFM measurements were performed under ambient conditions in tapping mode with standard 125 mm single-crystal silicon cantilevers. 7. As shown in Fig. 3a, at the tip site and the middle site of the carbon nanotube, there are two features, which can be attributed to DNA strands. Carbon nanotubes appear as bright lines in the image. While in Fig. 3b, DNA strands are connected to the sidewall of nanotubes, indicating that oxidization took place in the defect sites of sidewalls. From the observations of several samples, we conclude that DNA attachment occurred at nanotube ends and sidewalls, because chemical functionalization occurred primarily at the ends and sidewalls.

4 Notes 1. Unless stated otherwise, water used in this protocol was D.I water. 2. Unless stated otherwise, the CNT suspensions should be prepared in D.I water and followed by ultrasonication for 30 min. 3. Use only polyethylene gloves. Other gloves may contain silicones that can contaminate the surface. 4. Make sure everything used to handle or store your samples is clean (tweezers, etc.). It is recommended to have a dedicated

26

Chen et al.

set of clean tools for handling your samples. In particular, take care to avoid grease, oils, and silicone contaminants around your tools and work area. A general cleaning protocol that often works is to clean the utensils that will handle samples with the following solvents (in this order): Hexanes, Methylene chloride, Methanol, Acetone.

Acknowledgments The financial support of Key Laboratory Funding Scheme of Shenzhen Municipal Government, BTC operation fund (CityU Project No. 9683001) and City University of Hong Kong (Project No. 7002100) are gratefully acknowledged. References 1. Iijima S (1991) Helical microtubules of graphitic carbon. Nature 354:56–58 2. Ebbesen TW, Ajayan PM (1992) Large-scale synthesis of carbon nanotubes. Nature 358: 220–222 3. Chen RJ, Zhang YG, Wang DW, Dai HJ (2001) Noncovalent sidewall functionalization of single-walled carbon nanotubes for protein immobilization. J Am Chem Soc 123:3838–3839 4. Heller DA, Baik S, Eurell TE, Strano MS (2005) Single-walled carbon nanotube spectroscopy in live cells: towards long-term labels and optical sensors. Adv Mater 17: 2793–2799 5. Gao LZ, Nie L, Wang TH, Qin YJ, Guo ZX, Yang DL, Yan XY (2006) Carbon nanotube delivery of the GFP gene into mammalian cells. Chembiochem 7:239–242 6. Elghanian R, Storhoff JJ, Mucic RC, Letsinger RL, Mirkin CA (1997) Selective colorimetric detection of polynucleotides based on the distance-dependent optical properties of gold nanoparticles. Science 277:1078–1081 7. Perez JM, O’Loughin T, Simeone FJ, Weissleder R, Josephson L (2002) DNAbased magnetic nanoparticle assembly acts as a magnetic relaxation nanoswitch allowing screening of DNA-cleaving agents. J Am Chem Soc 124:2856–2857

8. Soto CM, Srinivasan A, Ratna BR (2002) Controlled assembly of mesoscale structures using DNA as molecular bridges. J Am Chem Soc 124:8508–8509 9. Guo Z, Sadler PJ, Tsang SC (1998) Immobilization and visualization of DNA and proteins on carbon nanotubes. Adv Mater 10:701–703 10. Shim M, Kan NWS, Chen RJ, Dai HJ (2002) Functionalization of carbon nanotubes for biocompatibility and biomolecular recognition. Nano Lett 2:285–288 11. Sarah EB, Cai W, Lasseter TL, Weidkamp KP, Hamers RJ (2002) Covalently bonded adducts of deoxyribonucleic acid (DNA) oligonucleotides with single-wall carbon nanotubes: synthesis and hybridization. Nano Lett 2: 1413–1417 12. Moulder JF, Stickle WF, Sobol PE, Bomben KD (1992) Handbook of X-ray photoelectron spectroscopy. Perkin-Elmer, Eden Prairie, MN 13. Lin Z, Strother T, Cai W, Cao X, Smith LM, Hamers RJ (2002) DNA attachment and hybridization at the silicon (100) surface. Langmuir 18:788–796 14. Chen WW, Tzang CH, Tang JX, Yang MS, Lee ST (2005) Covalently linked deoxyribonucleic acid with multiwall carbon nanotubes: synthesis and characterization. Appl Phys Lett 86:103114

Chapter 4 Temperature and pH-Responsive “Smart” Carbon Nanotube Dispersions Dan Wang and Liwei Chen Abstract Carbon nanotubes (CNTs) are a family of all-carbon quasi one-dimensional nanomaterials that are highly hydrophobic and typically aggregated in bundles. Recent accomplishments in dispersing CNTs in aqueous solutions open possibilities for their new applications in biomedicine. In many occasions, biological and biomedical applications demand an actuation mechanism; thus, it is highly desirable to control the dispersion and aggregation of CNTs in aqueous solutions with external stimuli. Here, we report two “smart” single-walled CNT (SWNT) aqueous dispersions that respond to temperature and pH changes through environment-responsive polymers, poly (N-isopropylacrylamide) (PNIPAAm) and Poly-l-lysine (PLL). Key words: Single-walled carbon nanotube, Dispersion, Aggregation, Temperature-responsive, pH-responsive, poly (N-isopropylacrylamide), Poly-l-lysine, Atomic force microscopy, Fluorescence

1. Introduction The unique all carbon quasi one-dimensional structure of SWNTs has intrigued much fundamental research on their mechanical, thermal, and electrical properties as well as potential applications in nanocomposites, field emission displays, and molecular electronics (1–5). Recent introduction of individually dispersed SWNTs to aqueous and biology compatible media has opened new frontiers of carbon nanotubes in biology and nanomedicine (6–8). Generally speaking, the surface of as-produced SWNTs is highly hydrophobic; and thus SWNTs exist in aggregated bundles. SWNT dispersions in organic or aqueous solvents can be obtained by either covalent functionalization or non-covalent solublization, with the latter being more favored for the preservation of nanotube structure and properties. Since non-covalent

K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_4, © Humana Press, a part of Springer Science + Business Media, LLC 2010

27

28

Wang and Chen

interactions with SWNT sidewalls (p–p stacking, van-der-Waals interactions, and hydrophobic interactions) do not require specificity and directionality, a surprisingly large variety of small molecule surfactants and water-soluble amphiphilic polymers are suitable for this purpose (9–17). Multi-functional polymer surfactants are also exploited for the separation (14, 16–18), alignment (19, 20), and hierarchical assembly of SWNTs (21), as well as for attaching fluorescent chromophore and biologically active cargos to SWNTs (22, 23). Many biomedical applications involve active processes, for example, drug and/or gene delivery involves active release of load at target locations; photothermal and photodynamic therapies involve light-induced actuation: heating and photochemistry, respectively. Utilization of amphiphilic polymer surfactants in SWNT dispersion enables potential actuation mechanisms via selected multi-functional polymers (24). Here, we discuss two “smart” SWNT dispersions that respond to temperature and pH changes using poly (N-isopropylacrylamide) (PNIPAAm) and poly-l-lysine (PLL). The action in response to temperature or pH stimuli is the aggregation of SWNTs in the dispersion. Controlled aggregation of SWNT may help clogging local blood vessel, “deliver” SWNT along with attached load by precipitating from circulation, or enhance photothermal effects. Therefore, the two environmentally responsive SWNT dispersions presented here are pioneering examples of SWNT actuation in biocompatible environments, and they may lead to more sophisticated biotechnologies in the future. PNIPAAm has long been known as a temperature responsive polymer (25). The structure of PNIPAAm contains balanced hydrophilicity from amide bonds and hydrophobicity from hydrocarbon main chain and isopropyl groups on side chains. At temperatures lower than the lower critical solution temperature (LCST, ~33°C for PNIPAAm), PNIPAAm assumes an extended chain conformation in which N and O atoms in amid bonds form H-bonding with surrounding water molecules. When the temperature is higher than the LCST, water molecules are released to maximize entropy while amid groups form intra-chain H-bonding among themselves; thus, the polymer chain takes a coiled and compact conformation (26–28). The extended conformation of PNIPAAm at low temperature allows the amphiphilic polymer chain to cover a large area of SWNT surface in dispersion. Raising temperature causes the chain conformation to shrink and thus exposes hydrophobic SWNT surface to water. This provides a driving force for SWNTs, which are dispersed at low temperatures, to aggregate into small bundles. PLL is an amphiphilic polyelectrolyte that contains both hydrophobic hydrocarbon moieties and partially protonated primary amine groups. SWNTs interact with PLL in aqueous

Temperature and pH-Responsive “Smart” Carbon Nanotube Dispersions

29

environment because of two interactions: one is hydrophobic interaction between PLL hydrocarbon linker moieties (–C4H8–) and SWNT sidewalls; the other is cation-p interaction between protonated amine groups and the p-electron system of SWNTs (26). The cation-p interaction is completely turned off at pH value equal to or higher than the isoelectric point of lysine (9.7) due to the deprotonation of –NH3+ groups. The hydrophobic interaction between –C4H8– linker and SWNT is also affected by pH via changes in the secondary structure of PLL. PLL chain adopts the a-helix conformation at high pH but changes to uncoiled conformations in acidic or neutral pH due to the electrostatic repulsion among side chain cations (29, 30). Since both interaction mechanisms between PLL and SWNT respond to the pH change of the solution, the PLL–SWNT dispersion is expected and indeed observed to be pH sensitive.

2. Materials 1. HiPCO SWNT powders, purified grade (Carbon Nano technologies, Inc., Houston, TX). 2. Poly (N-isopropylacrylamide), molecular weight 20,000– 25,000 (Aldrich). 3. 0.1 % (w/v) Poly-l-lysine solution in H2O (Sigma). 4. Sonicator VCX 130 (Sonix, Newtown, CT). 5. Atomic force microscopy (AFM) microscope MFP3D (Asylum Research, Santa Barbara, CA). 6. AFM probes NSC15/AlBS (MikroMasch USA, Wilsonville, OR) with resonance frequency around 325 KHz. 7. Fluorescence NS1 NanoSpectralyzer (Applied NanoFluo rescence, LLC, Houston, TX). 8. Spectropolarimeter Jasco-715 (Jasco Inc., Easton, MD).

3. Methods 3.1. SWNT Dispersion in PNIPAAm 3.1.1. Preparation of PNIPAAm–SWNT Dispersion

1. 10 mg/ml of PNIPAAm solution is prepared by dissolving 30 mg PNIPAAm in 2.4 ml water and then adding 0.6 ml of 0.1 M NaOH solution (see Note 1). 2. Add ~1 mg of HiPCO SWNT powder to 3 ml of 10 mg/ml PNIPAAm solution. This mixture is sonicated for 90 min in an ice-water bath at a power level of 130 W (Fig. 1).

30

Wang and Chen

Fig. 1. Experimental setup for SWNT sonication: (1) sonication probe; (2) polymer and SWNTs; (3) ice-water bath

Fig. 2. Absorption spectrum of PNIPAAm–SWNT dispersion. Inset: the structure of PNIPAAm

3. The solution is centrifuged for 5 min at 800 × g and 3 min at 2,300 × g to yield a PNIPAAm–SWNT dispersion in the supernatant (see Note 2). 3.1.2. Spectroscopic Characterization of PNIPAAm–SWNT Dispersion

The absorption spectrum (Fig. 2) showed resolved peaks in spectral ranges for the first interband transitions for metallic SWNTs (M11), the first (S11) and second (S22) interband transitions for semiconducting SWNTs (31, 32). The Raman spectrum (Fig. 3) showed features of SWNTs including the radial breathing mode (RBM), the tangential G-band, the disorder induced D-band,

Temperature and pH-Responsive “Smart” Carbon Nanotube Dispersions

31

Fig. 3. Raman spectrum of PNIPAAm–SWNT dispersion

and its second-order harmonic, the G¢-band (33). The van Hove peaks in absorption spectrum and the low intensity of D-band in Raman spectrum indicate that PNIPAAm molecules are noncovalently physisorbed on the sidewall of SWNTs (26). 3.1.3. Temperature Response of PNIPAAm– SWNT Complexes

To test the temperature response of the SWNT dispersion in PNIPAAm, we use atomic force microscopy to characterize the morphology of individual PNIPAAm–SWNT complexes and fluorescence spectra to characterize the response at the ensemble level. The PNIPAAm–SWNT complexes were first measured at room temperature, then heated in a 40°C water bath for 2 min and measured after cooling to room temperature, finally re-dispersed by 2 min sonication at 0°C and measured at room temperature. Photographs (Fig. 4) showed the solution becomes turbid after heating due to the aggregation of free PNIPAAm. When the solution is cooled down to room temperature, it became clear again. 1. AFM sample preparation: 10 µl of 0.1% w/v PLL solution was spin-coated onto a freshly cleaved mica and then 10 µl of the PNIPAAm–SWNT dispersion was spin coated on this substrate. The samples were then rinsed with deionized water and dried with argon gas. All AFM images were taken at room temperature and ambient conditions in AC mode. The images show that SWNTs aggregate into larger bundles (4–20 nm diameter) after heating (Fig. 5) (see Note 3). However, after 2 min sonication in 0°C ice-water bath, SWNTs are re-dispersed individually or in small bundles similar to those from the original dispersion (<5 nm) (Fig. 5). 2. Fluorescence spectra are taken at room temperature using an excitation wavelength of 785 nm. At this excitation energy, semiconducting SWNTs with different diameters fluoresce at their respective S11 band transition energies and form a

32

Wang and Chen

Fig. 4. Photographs of PNIPAAm–SWNT dispersions, from left to right: PNIPAAm–SWNT dispersions at 0°C, 40°C, cooled to room temperature, and redispersed by 2 min sonication at 0°C

Fig. 5. AFM images of PNIPAAm–SWNT complexes, from left to right: original dispersion, heated at 40°C for 2 min, redispersed by sonicating at 0°C for 2 min

spectrum with multiple peaks. Aggregation of SWNTs results in quenching of the fluorescence. Partial quenching of nanotube fluorescence after heating suggests bundling of SWNTs in the dispersion and the bundling effect is uniform across SWNTs with all different diameters (Fig. 6). The fluorescence intensity recovers to the level of the original dispersion after re-dispersion (Fig. 6), which indicates that the temperature stimulated switching is reversible (see Note 4). 3.2. SWNT Dispersion in PLL 3.2.1. Preparation and Spectroscopic Characterization of PLL–SWNT Dispersion

1. Add ~1 mg of purified HiPCO SWNT powders to 2 ml of 0.1% w/v Poly-l-lysine solution and sonicate for an hour in ice-water bath at a power level of 130 W. 2. Centrifuge the mixture for 10 min at 9,400 × g to yield the PLL–SWNT complexes in the supernatant. The pH of asprepared PLL–SWNT complexes solution is about 7.0.

Temperature and pH-Responsive “Smart” Carbon Nanotube Dispersions

33

Fig. 6. Fluorescence spectra of PNIPAAm–SWNT dispersions before (solid trace) and after (dash trace) heating in 40°C, and after redispersion 2 min sonication at 0°C (dot trace) (Excited at 785 nm)

Fig. 7. Absorption spectrum of PLL–SWNT dispersion. Inset: the structure of PLL

3. Take absorption (Fig. 7) and Raman spectra (Fig. 8) of PLL–SWNT dispersion. 3.2.2. pH Response of PLL–SWNT Complexes

Atomic force microscopy and fluorescence spectra are used to characterize the pH response of the SWNT dispersion in PLL solution. Circular dichroism (CD) spectroscopy is used to verify secondary structure changes of PLL at different pH values. 1. The pH of the PLL–SWNT solution was adjusted from 7.0 to 4.1, then increased to 9.7, decreased back to 8.3, and finally

34

Wang and Chen

Fig. 8. Raman spectrum of PLL–SWNT dispersion

Fig. 9. Photographs of PLL–SWNT dispersions at different pH values

decreased to 5.1 by adding 1.5 M HCl or NaOH solutions. Each change of pH is followed by 10 min sonication in ice-water bath. The photographs (Fig. 9) show that the solution becomes cloudy at pH 9.7 due to the aggregation of SWNTs. 2. AFM samples were prepared by spin coating 10 µl of the PLL–SWNT dispersion onto a fleshly cleaved mica substrate followed by rinsing with dionized water and drying with argon gas. Imaging conditions are the same as that in PNIPAAm–SWNT measurements. AFM images (Fig. 10) show SWNTs are individually dispersed or in small bundle (<5 nm) in acidic and neutral environments but aggregate into large bundles at pH 8.3. At pH 9.7, SWNTs coalesce into big bundles, precipitate out of the solution, and could not be imaged with AFM.

Temperature and pH-Responsive “Smart” Carbon Nanotube Dispersions

35

Fig. 10. AFM images of PLL–SWNT complexes at different pH values

3. Fluorescence spectra of PLL–SWNT dispersions at different pH values were taken and excited at 785 nm. Figure 11 shows that SWNT emission is partially quenched due to aggregation at pH 9.7 but remains largely unchanged in acidic and neutral environments. 4. Circular dichroism (CD) spectra of PLL–SWNT complexes were recorded at different pH values. The CD spectra of PLL–SWNT complexes in an acidic or neutral pH (Fig. 12) show a single negative peak around 200 nm, which is characteristic of uncoiled conformation. At pH 9.7, the spectrum displays two negative bands in the 200–230 nm range, characteristic of the a-helix conformation (29, 34).

4. Notes 1. All solutions are prepared in deionized water (18.2 MW cm, Millipore). 2. The PNIPAAm–SWNT dispersion is stable at room temperature for weeks with little or no precipitation, but it can not withstand high speed centrifugation. 3. Sonicating the dispersion while it is heated at 40°C does not prevent the complexes from aggregating.

36

Wang and Chen

Fig. 11. Fluorescence spectra of PLL–SWNT dispersions at different pH values excited at 785 nm

Fig. 12. Circular dichroism spectra of PLL–SWNT complexes at different pH values

4. Spectrum at 40°C could not be obtained because of the strong scattering of aggregated free PNIPAAm and PNIPAAm– SWNT complexes. We measure the fluorescence spectra when the solution cools down to room temperature.

Acknowledgments This work is partially supported by the Nanobiotechnology Initiative at Ohio University. D. W. thanks Ohio University Condensed Matter and Surface Sciences program for partial support.

Temperature and pH-Responsive “Smart” Carbon Nanotube Dispersions

37

References 1. Kim SN, Rusling JF, Papadimitrakopoulos F (2007) Carbon nanotubes for electronic and electrochemical detection of biomolecules. Adv Mater 19:3214–3228 2. Avouris P, Chen Z, Perebeinos V (2007) Carbon-based electronics. Nat Nanotechnol 2:605–615 3. Snow ES, Perkins FK, Robinson JA (2006) Chemical vapor detection using single-walled carbon nanotubes. Chem Soc Rev 35:790–798 4. Lu W, Lieber CM (2007) Nanoelectronics from the bottom up. Nat Mater 6:841–850 5. Prato M, Kostarelos K, Bianco A (2008) Functionalized carbon nanotubes in drug design and discovery. Acc Chem Res 41:60–68 6. Chen X, Tam UC, Czlapinski JL, Lee GS, Rabuka D, Zettl A, Bertozzi CR (2006) Interfacing carbon nanotubes with living cells. J Am Chem Soc 128:6292–6293 7. Cherukuri P, Gannon CJ, Leeuw TK, Schmidt HK, Smalley RE, Curley SA, Weisman RB (2006) Mammalian pharmacokinetics of carbon nanotubes using intrinsic near-infrared fluorescence. Proc Natl Acad Sci USA 103:18882–18886 8. Lacerda L, Bianco A, Prato M, Kostarelos K (2006) Carbon nanotubes as nanomedicines: from toxicology to pharmacology. Adv Drug Deliv Rev 58:1460–1470 9. Chatterjee T, Yurekli K, Hadjiev VG, Krishnamoorti R (2005) Single-walled carbon nanotube dispersions in poly(ethylene oxide). Adv Funct Mater 15:1832–1838 10. Dieckmann GR, Dalton AB, Johnson PA, Razal J, Chen J, Giordano GM, Munoz E, Musselman IH, Baughman RH, Draper RK (2003) Controlled assembly of carbon nanotubes by designed amphiphilic peptide helices. J Am Chem Soc 125:1770–1777 11. Karajanagi SS, Yang HC, Asuri P, Sellitto E, Dordick JS, Kane RS (2006) Protein-assisted solubilization of single-walled carbon nanotubes. Langmuir 22:1392–1395 12. Moore VC, Strano MS, Haroz EH, Hauge RH, Smalley RE, Schmidt J, Talmon Y (2003) Individually suspended single-walled carbon nanotubes in various surfactants. Nano Lett 3:1379–1382 13. Numata M, Asai M, Kaneko K, Bae AH, Hasegawa T, Sakurai K, Shinkai S (2005) Inclusion of cut and as-grown single-walled carbon nanotubes in the helical superstructure of schizophyllan and curdlan (ss-1, 3-glucans). J Am Chem Soc 127:5875–5884

14. Ortiz-Acevedo A, Xie H, Zorbas V, Sampson WM, Dalton AB, Baughman RH, Draper RK, Musselman IH, Dieckmann GR (2005) Diameter-selective solubilization of single-walled carbon nanotubes by reversible cyclic peptides. J Am Chem Soc 127:9512–9517 15. Wang D, Ji WX, Li ZC, Chen LW (2006) A biomimetic “polysoap” for single-walled carbon nanotube dispersion. J Am Chem Soc 128:6556–6557 16. Zheng M, Jagota A, Semke ED, Diner BA, McLean RS, Lustig SR, Richardson RE, Tassi NG (2003) DNA-assisted dispersion and separation of carbon nanotubes. Nat Mater 2:338–342 17. Zheng M, Jagota A, Strano MS, Santos AP, Barone P, Chou SG, Diner BA, Dresselhaus MS, McLean RS, Onoa GB, Samsonidze GG, Semke ED, Usrey M, Walls DJ (2003) Structure-based carbon nanotube sorting by sequence-dependent DNA assembly. Science 302:1545–1548 18. Huang X, McLean RS, Zheng M (2005) High-resolution length sorting and purification of DNA-wrapped carbon nanotubes by size-exclusion chromatography. Anal Chem 77:6225–6228 19. McLean RS, Huang X, Khripin C, Jagota A, Zheng M (2006) Controlled two-dimensional pattern of spontaneously aligned carbon nanotubes. Nano Lett 6:55–60 20. Shim BS, Kotov NA (2005) Single-walled carbon nanotube combing during layer-by-layer assembly: from random adsorption to aligned composites. Langmuir 21:9381–9385 21. Wang D, Li Z-C, Chen L (2006) Templated synthesis of single-walled carbon nanotube and metal nanoparticle assemblies in solution. J Am Chem Soc 128:15078–15079 22. Kam NWS, O’Connell M, Wisdom JA, Dai H (2005) Carbon nanotubes as multifunctional biological transporters and nearinfrared agents for selective cancer cell destruction. Proc Natl Acad Sci USA 102:11600–11605 23. Kam NWS, Liu Z, Dai H (2006) Carbon nanotubes as intracellular transporters for proteins and DNA: an investigation of the uptake mechanism and pathway. Angew Chem Int Ed Engl 45:577–581 24. Barone PW, Strano MS (2006) Reversible control of carbon nanotube aggregation for a glucose affinity sensor. Angew Chem Int Ed Engl 45:8138–8141

38

Wang and Chen

25. Schild HG (1992) Poly(N-isopropylacrylamide): experiment, theory and application. Prog Polym Sci 17:163–249 26. Wang D, Chen L (2007) Temperature and pH-responsive single-walled carbon nanotube dispersions. Nano Lett 7:1480–1484 27. Sun T, Wang G, Feng L, Liu B, Ma Y, Jiang L, Zhu D (2004) Reversible switching between superhydrophilicity and superhydrophobicity. Angew Chem Int Ed 43:357–360 28. Lin S-Y, Chen K-S, Run-Chu L (1999) Thermal micro ATR/FT-IR spectroscopic system for quantitative study of the molecular structure of poly(N-isopropylacrylamide) in water. Polymer 40:2619–2624 29. Wang Y, Chang YC (2003) Synthesis and conformational transition of surface-tethered polypeptide: poly(L-lysine). Macromolecules 36:6511–6518 30. Davidson B, Fasman GD (1967) Confor mational transitions of uncharged poly-L-lysine.

31.

32.

33.

34.

a Helix-random coil-b-structure. Biochemistry 6:1616–1629 Strano MS, Dyke CA, Usrey ML, Barone PW, Allen MJ, Shan H, Kittrell C, Hauge RH, Tour JM, Smalley RE (2003) Electronic structure control of single-walled carbon nanotube functionalization. Science 301:1519–1522 O’Connell MJ, Bachilo SM, Huffman CB, Moore VC, Strano MS, Haroz EH, Rialon KL, Boul PJ, Noon WH, Kittrell C, Ma J, Hauge RH, Weisman RB, Smalley RE (2002) Band gap fluorescence from individual single-walled carbon nanotubes. Science 297:593–596 Dresselhaus MS, Dresselhaus G, Jorio A, Souza Filho AG, Saito R (2002) Raman spectroscopy on isolated single wall carbon nanotubes. Carbon 40:2043–2061 Greenfield NJ, Fasman GD (1969) Computed circular dichroism spectra for the evaluation of protein conformation. Biochemistry 8:4108–4116

Part II Toxicity

Chapter 5 Effects of Carbon Nanotubes on the Proliferation and Differentiation of Primary Osteoblasts Dawei Zhang, Changqing Yi, Suijian Qi, Xinsheng Yao, and Mengsu Yang Abstract This chapter provides a detailed protocol for studying the effects of carbon nanotubes (CNTs) on the proliferation, differentiation, adipocytic transdifferentiation, and mineralization of primary osteoblasts. SWNTs, DWNTs, and MWNTs with the same mean length and various diameters were shown to reduce the viability of osteoblasts and inhibit the adipocytic transdifferentiation in both time- and dose-dependent manners. The order of inhibition effect is SWNTs > DWNTs > MWNTs. CNTs were found to inhibit the formation of mineralized nodules greatly and dose-dependently during the final stage of osteoblast differentiation, causing a 50% decrease in the formation of mineralized nodules at the concentration of 50 mg/mL. The expression of important proteins such as Runx-2 and Col-I in osteoblasts was also greatly inhibited by the CNTs. TEM results revealed that the effects on cellular behavior may be exerted by the CNTs from in- and outside of the cells. Key words: Carbon nanotubes, Primary osteoblast, Alkaline phosphatase, Adipocyte, TEM

1. Introduction Carbon nanotubes (CNTs) have potential applications in biosensors, tissue engineering, and biomedical devices, because of their unique electronic, chemical, and mechanical properties (1–4). Recent studies also suggested that carbon-based nanomaterials may be present in the atmospheric environment (5). Experimental evidence over the past few years showed that ultrafine nanomaterials with mean diameters <100 nm in the atmosphere are potentially toxic (6). Therefore, the envisaged broad applications of CNTs demand a thorough evaluation of the potential effects of CNTs on both the living system and the environment (7). Though considerable efforts have been devoted to the study of in vivo and in vitro toxicity of CNTs (8–13), no study has been carK. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_5, © Humana Press, a part of Springer Science + Business Media, LLC 2010

41

42

Zhang et al.

ried out to systematically investigate and compare the biological effects of CNTs on bone cells using primary osteoblasts model. Bone is a dynamic tissue that is constantly being reshaped by osteoblasts, which build bone, and osteoclasts, which resorb bone. Primary osteoblasts cell cultures offer several advantages, particularly for studying cell growth control mechanisms and differentiation in the context of a mineralizing matrix. An osteoblast is a mononucleate cell that is responsible for bone formation and mineralization of the osteoid matrix. Primary osteoblast cells have been shown to be a model system for revealing biological effects of nanomaterials because their susceptibility to nanomaterials is similar to that in vivo (14–16). Herein, we describe the detailed protocols for the study of the biological effects of CNTs, including SWNTs, DWNTs, and MWNTs, on the proliferation, differentiation, adipocytic transdifferentiation, and mineralization of primary osteoblasts.

2. Materials 2.1. CNTs

1. SWNTs (diameter <2 nm), DWNTs (diameter <5 nm), and MWNTs (diameter <10 nm) with the same mean length of 5–15 µm were prepared by the chemical vaporization deposition (CVD) method (Shenzhen Nanotech Port Co., Ltd , China) (see Note 1). 2. Quartz particles (Merk, Silica gel 60, 63–200 mm) were used as a positive control and dispersed in a-MEM.

2.2. Isolation and Culture of Primary Osteoblasts

1. NIH mice (Guangzhou University of Traditional Medicine, China). 2. Solution of trypsin (0.25%) and ethylenediamine tetraacetic acid (EDTA) (1 mM) (Gibco/BRL). 3. Minimum Essential Medium MEM Alpha Medium (Gibco/ BRL, Bethesda, MD) supplemented with 10% fetal bovine serum (FBS, HyClone, Ogden, UT), 100 U/mL penicillin, and 100 mg/mL streptomycin (Gibco/BRL). 4. Collagenase type II (Sigma) is dissolved at 1.0 g/L a-MEM and stored in single use aliquots at −20°C.

2.3. steoblast Viability and Proliferation Assay

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) dissolved at 5 mg/mL in PBS and stored in single use aliquots at −20°C (see Note 2).

2.4. Assay for Alkaline Phosphatase Activity

1. Alkaline phosphatase activity kits (Nanjing Jiancheng Biological Engineering Institute, China). 2. A micro-Bradford assay kit for protein content (Beyotime Biotechnology, China).

Effects of Carbon Nanotubes on the Proliferation and Differentiation

43

2.5. Assay for Adipocytic Transdifferentiation of Osteoblasts

1. Adipogenic supplement: Insulin (10 mg/L) and dexamethasone (1.0 × 10−7 mol/L) in a-MEM.

2.6. Assay for Mineralized Matrix Formation

1. Differentiation medium: b-glycerophosphate (BBI) (10 mM/L) and ascorbic acid (Sigma) (50 µg/mL) in a-MEM.

2. 0.6% Oil red O solution: 0.6 g oil red O in 100 mL 60% isopropyl alcohol and 40% water. Filtrate through 0.4 mm membrane filter before use.

2. Alizarin red-S (Sigma) solution: 40 mM/L in double-distilled (dd) H2O, pH adjusted to 4.2. Store at 4°C. 3. Cetylpyridium chloride (Sigma) solution: 10% in ddH2O. Store at room temperature.

2.7. Western Blotting Analysis

1. Buffer A: 48% 1 N HCl, 36.3% Tris, 0.23% TEMED, pH adjusted to 8.9 using 1 N HCl. 2. Buffer B: 48% 1 N HCl, 5.98% Tris, 0.46% TEMED, pH adjusted to 6.7 using 1 N HCl. 3. 30% Acrylamide/Bis solution: 37.5:1 mixture (30% T, 2.67% C) (Bio-rad). 4. Cell lysis buffer: 50 mM Tris–HCl (pH 7.4), 10 mM EDTA, 4.3 M Urea, and 1% Triton X-100. 5. SDS–PAGE running buffer (10×): 3.0% (w/v) Tris, 14.4% (w/v) glycine, 1.0% (w/v) SDS. Store at room temperature. 6. Transferring buffer (1×): 0.3% (w/v) Tris, 1.44% (w/v) glycine, 20% (v/v) methanol. Store at room temperature. 7. Prestained molecular weight markers: BenchMark™ PreStained Protein Ladder (Invitrogen). 8. Tris-buffered saline with Tween (TBST) solution: 10 mM Tris–HCl, 150 mM NaCl, 0.05% Tween, pH 8.0. 9. Supported nitrocellulose membrane from Millipore, Bedford, MA, and 3 mm chromatography paper (Whatman, Maidstone, UK). 10. Blocking buffer: 5% (w/v) nonfat dry milk in TBST. 11. Primary antibody dilution buffer: TBST supplemented with 5% (w/v) bovine serum albumen (BSA Sigma, St Louis, USA). 12. Anti-Runx-2, anti-COL-I, or anti-b-actin primary antibodies (Santa Cruz Biotechnology, Inc.) in the TBST solution at a 1:1,000 dilution. 13. Secondary antibody: Anti-rabbit and anti-goat IgG conjugated to horse radish peroxidase (Santa Cruz, Santa Cruz, CA). 14. Enhanced chemiluminescent (ECL) reagents and Hyperfilm (Amersham Biosciences, UK) (see Note 3).

44

Zhang et al.

2.8. Electron Microscopy

1. Washing buffer 0.1 M cacodylate: 21.4 g of sodium cacodylate trihydrate (Electron Microscopy Sciences, USA) in 400 mL ddH2O, adjust to pH 7.4 with 0.2 N HCl. Fill up with ddH2O to 500 mL. Store at room temperature. 2. Prefixing solution: Modified Karnovsky fixative (2% glutaraldehyde + 2% paraformaldehyde (Electron Microscopy Sciences, USA)) in 0.1 M cacodylate buffer (pH 7.4) with 0.05% CaCl2. 3. Postfixing solution: 1% osmium tetroxide (Electron Microscopy Sciences, USA) in 0.1 M cacodylate buffer (pH 7.4) (see Note 4). 4. Spurr’s resin: Prepare according to the manufacturer’s manual. Aliquot into small vials and store at −20°C. Thaw and return to room temperature before use.

3. Methods 3.1. Preparation of Water-dispersible Carbon Nanotubes

1. Oxidized CNTs were generated by refluxing in a mixture of concentrated sulfuric and nitric acids (3:1, 98%, 69%, respectively) at 70°C for 4 h (see Note 5). 2. Oxidized CNTs were washed several times with distilled water by ultracentrifugation (10,000 × g × 30 min) until the pH reaches 7.0, which could not affect the pH of the buffered cellular medium. 3. After this process, CNTs had been modified with carboxylic acid groups. 4. CNTs suspensions used in the experiments (100, 50, 10, 1, and 0.1 mg/mL) were prepared by suspending CNTs with minimum essential medium alpha (a-MEM) supplemented with trypsin and 10% fetal bovine serum. In addition, quartz particles dispersed in a-MEM were used as a positive control.

3.2. Isolation and Culture of Primary Osteoblasts

Primary osteoblasts were prepared from 3-day-old NIH mice calvarias following the sequential enzymatic digestion method. 1. Isolation: skull (frontal and perietal bones) was dissected; then endosteum and periosteum were stripped off, and the bone was cut into approximately 1–2 mm2 pieces. 2. Digestion: Bone slices were then sequentially digested with 2.5 g/L trypsin for 30 min and 1.0 g/L collagenase II twice for 1 h each time (see Note 6). 3. Collection and culture: The cells were collected and cultured in a-MEM with 10% heat-inactivated fetal bovine serum, 100 U/mL penicillin, and 100 mg/mL streptomycin, for

Effects of Carbon Nanotubes on the Proliferation and Differentiation

45

24 h in a humidified atmosphere of 5% CO2 in air at 37°C, then replaced with fresh medium. The culture medium was changed every 3 days during the experiments. 3.3. Osteoblast Viability and Proliferation Assay

1. First, osteoblast cells were seeded in 96-well tissue culture plates at the density of 2 × 103 cells per well and incubated overnight prior to the addition of CNTs suspensions. 2. The culture was then treated with different concentrations of CNTs, and cultured for another 24, 48, and 72 h. Cells with SiO2 and cells without CNTs treatment were used as positive control and negative control respectively, and wells without cells were set as blank. 3. After treatment, 20 µL of MTT was added to each well and incubated for another 4 h at 37°C. 4. Then, the supernatant was removed and DMSO was added, and absorbance at 570 nm was recorded on a microplate spectrophotometer. 5. The relative cytotoxicity was expressed as percentage of (ODsample − ODblank)/(ODcontrol − ODblank). Each experiment was performed in triplicate. 6. Results: MTT assay results were illustrated in Fig. 1. The treatment of osteoblasts with a series of dilutions of CNTs generated time- and dose-dependent decrease in cell viability with the maximum inhibition effect on day 3 at the concentration of 100 mg/mL. There is a tendency that the cytotoxic effects of SWNTs increased more greatly than DWNTs and MWNTs over time. Besides, CNTs exhibited more potent cytotoxic effects than SiO2 particles at the same concentration.

3.4. Assay for Alkaline Phosphatase Activity

1. Osteoblast cells (20,000 cells per well) were seeded in 48-well culture plates and cultured overnight at 37°C in a 5% CO2 humidified incubator.

Fig. 1. Time- and dose-dependence of SWNTs (a), DWNTs (b), MWNTs (c) on the viability and proliferation of osteoblasts. Results are mean ± SD of the triplicate experiments, p < 0.05. 10 and 50 mg/mL SiO2 were used as positive controls. Reproduced with the permission from (20), © 2007 Institute of Physics and IOP Publishing Limited

46

Zhang et al.

2. CNTs were added to culture medium at final concentrations of 1, 10, and 50 mg/mL and cultured for further 3 days. 3. The plates were washed twice with ice-cold PBS and lysed by two cycles of freezing and thaw (see Note 7). 4. Aliquots of supernatants were subjected to alkaline phosphatase activity and protein content measurement by an alkaline phosphatase activity kit and a micro-Bradford assay kit. 5. All results were normalized by protein content. Unit definition: One unit will convert 1 g of tissue protein to 1 mg p-nitrophenol and phosphate in 15 min at 37°C. 6. Results: ALP activity assay results were illustrated in Fig. 2, suggesting that CNTs inhibited the ALP activity of primary osteoblasts in a time-dependent manner without evident dose-dependence. A previous study also showed that there was no apparent correlation between nanotube dose and the change in expression of specific genes (17). 3.5. Assay for Adipocytic Transdifferentiation of Osteoblasts

1. Osteoblast cells (20,000 cells per well) were seeded in 48-well tissue culture plates, and were cultured for 10 days. 2. The adipogenic supplement and CNTs were added to the culture medium with final concentrations of 1, 10, and 50 mg/mL. 3. Fat droplets within trans-differentiated adipocytes from osteoclast cells were evaluated by oil red O staining.

Fig. 2. Effects of carbon nanotubes on the ALP activity of osteoblasts. Results are mean ± SD of the triplicate experiments, p < 0.05. 50 mg/mL SiO2 was used as a positive control. Reproduced with the permission from (20), ©2007 Institute of Physics and IOP Publishing Limited

Effects of Carbon Nanotubes on the Proliferation and Differentiation

47

4. Cell monolayers were washed by PBS twice, then stained by 0.6% (w/v) oil red O solution for 15 min at room temperature (see Note 8). 5. For quantification of oil red O content, the cells were washed with distilled H2O three times to remove background staining, and isopropyl alcohol was added to resolve oil red O. Absorbance at 510 nm was measured on a microplate spectrophotometer (Bio-rad Model 680, USA). 6. Results: The results of quantification of oil red O content were illustrated in Fig. 3, indicating that CNTs could reduce the adipocytic transdifferentiation of primary osteoblasts in a dose-dependent manner on day 10, and the effect was increased in the order of SWNT>DWNT>MWNT. 3.6 Assay for Mineralized Matrix Formation

1. Osteoblast cells (30,000 cells per well) were seeded in 24-well tissue culture plates and cultured overnight at 37°C in a 5% CO2 humidified incubator. 2. The medium was then changed to differentiation medium containing 10 mM b-glycerophosphate and 50 µg/mL ascorbic acid in the presence or absence of 0.1, 1, 10, and 50 mg/mL of CNTs for 21 days. 3. The formation of mineralized matrix nodules was determined by alizarin red-S staining. Briefly, the cells were fixed in 70%

Fig. 3. Dose-dependence of carbon nanotubes on the adipocytic transdifferentiation of osteoblasts. Results are mean ± SD of the triplicate experiments, *p < 0.05. 50 mg/mL SiO2 was used as a positive control. Reproduced with the permission from (20), ©2007 Institute of Physics and IOP Publishing Limited

48

Zhang et al.

ethanol for 1 h at room temperature. The fixed cells were washed with PBS and stained with 40 mM alizarin red S, pH 4.2, for 30 min at room temperature. 4. Quantification of alizarin red S staining was performed by elution with 10% (w/v) cetylpyridium chloride for 10 min at room temperature and measuring absorbance at 570 nm. Results are expressed as moles of alizarin red S per milligram of total cellular protein. 5. Results: Fig. 5.4 showed the quantification of alizarin red S staining results. CNTs inhibited the formation of mineralized nodules greatly and dose-dependently. Coupling the number count with quantification of ARS deposition revealed a greater than 50% decrease in the formation of mineralized nodules upon CNTs treatment at the concentration of 50 mg/mL, which was more significant than SiO2 particles treatment. The difference in diameter was not reflected in the inhibitory effects of CNTs at most concentrations. 3.7. Western Blotting Analysis

1. These instructions assume the use of but not restricted to a Bio-Rad Mini-protein 3 systems. It is very important to keep glass plates scrubbed clean with a rinsable detergent after use and rinsed extensively with distilled water. 2. Prepare a 1.5-mm thick, 10% gel by mixing 2.5 mL Buffer A with 3.5 mL Buffer C, 0.1 mL 10% SDS, 3.8 mL H2O, 20 mL TEMED, 0.15 mL APS. Pour the gel, leaving space for a stacking gel, and overlay with ddH2O water. The gel should polymerize more than 25 min. 3. Pour off the water and clean the droplet with absorbent paper.

Fig. 4. Effects of CNTs on the mineralized nodule formation of osteoblasts. (a) The number of alizarin red S staining nodules; (b) Mineralization determined by elution of alizarin red S from stained mineral deposits. Results are mean ± SE of the triplicate experiments, *p < 0.05. 50 mg/mL SiO2 was used as a positive control. Reproduced with the permission from (20), ©2007 Institute of Physics and IOP Publishing Limited

Effects of Carbon Nanotubes on the Proliferation and Differentiation

49

4. Prepare the stacking gel by mixing 1.25 mL Buffer B with 0.7 mL Buffer C, 50 mL 10% SDS, 2.85 mL H2O, 20 mL TEMED, and 75 mL APS. Pour the stack and insert the comb. The stacking gel should polymerize within 30 min. 5. Once the stacking gel has set, remove the Gel Cassette Assemblies from the Casting Stand. Prepare the inner chamber and lower the inner chamber into the Mini Tank. 6. Add the running buffer to the inner and lower buffer chambers of the Mini Tank. Carefully remove the comb and use a 2-mL syringe fitted with a 22-gauge needle to wash the wells with running buffer. 7. Wash the cultures with cold PBS and add the cell lysis buffer, then determine the protein concentration. 8. Load the samples into the wells with a pipette using gel loading tips. Include one well for prestained molecular weight markers. 9. Complete the assembly of the Mini Tank assembly and connect to a power supply. The gel can be run at 100 V. 10. The stacking gel is removed and the separating gel is then laid on top of the nitrocellulose membrane. 11. Two further sheets of 3 mm paper are wetted in the transfer buffer and carefully laid on top of the gel, ensuring that no bubbles are trapped in the resulting sandwich. Two wet foam sheets are laid on top and bottom, then the transfer cassette closed. 12. The cassette is placed into the Mini Tank. Add transfer buffer into the inner and lower chamber. Put the Mini Tank into an ice box. 13. Complete the assembly of the Mini Tank assembly and connect to a power supply. Transfers can be run at 70 V for 2 h. 14. Once the transfer is finished, the membrane was blocked overnight at 4°C in blocking buffer. 15. The blots were incubated with anti-Runx-2, anti-COL-I, or anti-b-actin primary antibodies in TBST solution at a 1:1,000 dilution for 2 h at room temperature. 16. Followed by 1 h incubation with anti-rabbit, anti-goat antibodies conjugated with horseradish peroxidase. 17. The blots were visualized with an enhanced chemiluminescence (ECL) kit. 18. Membranes were exposed for 5–15 min to Hyperfilm for detection of signals. b-actin protein expression was used as a loading control. 19. Results: Western blot results were illustrated in Fig. 5. The results clearly showed the marked reduction of Runx-2 and

50

Zhang et al.

Col-I protein expression after osteoblasts were exposed to 50 µg/mL. Runx-2 directly regulates osteoblast-specific genes including osteocalcin, osteopontin, and type I collagen, etc (18). Type I collagen is one of the markers for differentiation and maturation of the osteoblasts. As type I collagen is the major product of osteoblast occupying about 90% of the bone matrix, alternation of the synthesis may have a great effect on the bone quality (19). Western blot results also showed a tendency that the inhibitory effect was in an order of SWNTs > DWNTs > MWNTs, which was consistent with the observed inhibitory effect on the viability and mineralization of osteoblasts. 3.8. Electron Microscopic Observation

1. For this experiment, osteoblasts were incubated with 50 µg/mL SWNTs and MWNTs for 24 h. 2. Primary fixation: Cells were washed with PBS, and place the Thermanox tissue culture cover slip was placed into a glass vial containing primary fixative for 24 h (see Note 9). 3. Rinsing and washing in 0.1 M cacodylate buffer for 5 min and 10 min, respectively. 4. Secondary fixation: After being washed, cells were postfixed with 1% osmium tetroxide at 4°C for 1 h. 5. Rinsing and washing in 0.1 M cacodylate buffer for 5 min and 10 min, respectively. 6. After fixation, cells were observed with the TEM after dehydration, resin embedding, ultrathin sectioning, and staining with uranyl acetate and lead citrate. 7. The ultrastructural alterations and internalization of osteoblasts in the presence of CNTs were observed with a Phillips Tecnai 12 TEM operating at 80 kV. 8. Results: TEM images (Fig. 6) clearly showed that although most of SWNTs and MWNTs did remain in the extracellular space throughout the 24 h exposure time, there were a limited number of CNTs taken up by osteoblasts (a and d). Meanwhile, it can be observed that CNTs aggregated around

Fig. 5. Runx-2 and Type-I Collagen expressed were reduced in the presence of CNTs. b-actin protein expression was used as a loading control. Reproduced with the permission from (20), ©2007 Institute of Physics and IOP Publishing Limited

Effects of Carbon Nanotubes on the Proliferation and Differentiation

51

osteoblasts and interacted with the cell membrane of osteoblasts, which may have a further impact on the permeability and integrity of cell membrane (b and c).

4. Notes 1. Handle with care and wear a mask. 2. Keep in dark place and wrap it with aluminum foil. 3. Use different tips to prepare the fresh ECL working solution from bottle A and bottle B, respectively. Otherwise, the ECL stock solution will be contaminated and expired. 4. Postfixing solution should be freshly prepared in a fume hood and avoid the vapor. Before changing to postfixing solution, sample must be washed by washing buffer to remove glutaraldehyde completely. 5. Prepare carefully in a fume hood.

Fig. 6. TEM images of the internalization of carbon nanotubes. (a, b) Cells treated with 50 µg/mL SWNTs; (c, d) Cells treated with 50 µg/mL MWNTs. Dotted arrows indicate extrocytoplasmic surface and solid arrows show the CNTs. Reproduced with the permission from (20), ©2007 Institute of Physics and IOP Publishing Limited

52

Zhang et al.

6. It is possible to repeat and increase digestion time to collect more cells. 7. It is preferable to measure the ALP activity and protein content immediately. In our experiments, we found that the sample degraded even when it is kept in −80°C. 8. The staining time should less be than 20 min, or there will be a strong background. 9. Make sure that the tissue culture cover slip with cells face up. It is better to place the tissue culture cover slip into a glass vial in following steps, because acetone could do serious damage to plastics (e.g., tissue culture plate) during resin embedding.

Acknowledgements The authors would like to thank the National Natural Science Foundation of China (NSFC), the Key Laboratory Scheme of Science and Technology Bureau of Shenzhen Municipal Government, BTC operation fund (CityU project No. 9683001) and City University of Hong Kong (Project No.7002100) for financial support. References 1. Chen RJ, Zhan YG, Wang DW, Dai HJ (2001) Noncovalent sidewall functionalization of single-walled carbon nanotubes for protein immobilization. J Am Chem Soc 123:3838–3839 2. Heller DA, Baik S, Eurell TE, Strano MS (2005) Single-walled carbon nanotube spectroscopy in live cells: towards long-term labels and optical sensors. Adv Mater 17:2793–2799 3. Gao LZ, Nie L, Wang TH, Qin YJ, Guo ZX, Yang DL, Yan XY (2006) Carbon nanotube delivery of the GFP gene into mammalian cells. Chembiochem 7:239–242 4. Cui DX, Tian FR, Kong Y, Titushikin I, Gao HJ (2004) Effects of single-walled carbon nanotubes on the polymerase chain reaction. Nanotechnology 15:154–157 5. Murr LE, Bang JJ, Esquivel EV, Guerrero PA, Lopez A (2004) Carbon nanotubes, nanocrystal forms, and complex nanoparticle aggregates in common fuel-gas combustion sources and the ambient air. J Nanopart Res 6:241–251 6. Donaldson K, Li XY, Mac NW (1998) Ultrafine (nanometre) particle mediated lung injury. J Aerosol Sci 29(5/6):553–560

7. Nel A, Xia T, Madler L, Li N (2006) Toxic potential of materials at the nanolevel. Science 311:622–627 8. Wang HF, Wang J, Deng XY, Sun HF, Shi ZJ, Gu ZN, Liu YF, Zhao YL (2004) Biodistribution of carbon single-wall carbon nanotubes in mice. J Nanosci Nanotechnol 4(8):1019–1024 9. Shvedova AA, Castranova V, Kisin ER, Schwegler-Berry D, Murray AR, Gandelsman VZ, Maynard A, Baron P (2003) Exposure to carbon nanotube material: assessment of nanotube cytotoxicity using human keratinocyte cells. J Toxicol Environ Health 66:1909–1926 10. Jia G, Wang HF, Yan L, Wang X, Pei RJ, Yan T, Zhao YL, Guo XB (2005) Cytotoxicity of carbon nanomaterials: single-wall nanotube, multi-wall nanotube, and fullerene. Environ Sci Technol 39:1378–1383 11. Mattson MP, Haddon RC, Rao AM (2001) Molecular functionalization of carbon nanotubes and use as substrates for neuronal growth. J Mol Neurosci 14:175–182 12. Yi CQ, Fong CC, Chen WW, Qi SJ, Tzang CH, Lee ST, Yang MS (2007) Interactions between carbon nanotubes and DNA polymerase and restriction endonucleases. Nanotechnology 18:025102

Effects of Carbon Nanotubes on the Proliferation and Differentiation 13. Cui DX, Tian FR, Ozkan CS, Wang M, Gao HJ (2005) Effect of single wall carbon nanotubes on human HEK293 cells. Toxicol Lett 155:73–85 14. Gutwein LG, Webster TJ (2002) Osteoblast and chondrocyte proliferation in the presence of alumina and titania nanoparticles. J Nanopart Res 4:231–238 15. Zhang DW, Zhang JC, Chen Y, Yang MS, Yao XS (2007) Methods for anti-osteoporosis drug screening in vitro. Chin Pharm J 42(3):161–164 16. Zhang JC, Li XX, Xu SJ, Wang K, Yu SF, Lin Q (2004) Effects of rare earth ions on proliferation, differentiation and function expression of cultured osteoblasts in vitro. Prog Nat Sci 14(4):404–409 17. Ding LH, Stilwell J, Zhang TT, Elboudwarej O, Jiang HJ, Selegue JP, Cooke PA, Gray JW,

53

Chen FF (2005) Molecular characterization of the cytotoxic mechanism of multiwall carbon nanotubes and nano-onions on human skin fibroblast. Nano Lett 5:2448–2464 18. Yang S, Wie D, Wang D, Phimphilai M, Krebsbach PH, Franceschi RT (2003) In vitro and in vivo synergistic interactions between the Runx2/ Cbfa1 transcription factor and bone morphogenetic protein-2 in stimulating osteoblast differentiation. J Bone Miner Res 18:705–715 19. Park JH, Park BH, Kim HK, Park TS, Baek HS (2002) Hypoxia decreases Runx2/Cbfa1 expression in human osteoblast-like cells. Mol Cell Endocrinol 192:197–203 20. Zhang DW, Yi CQ, Zhang JC, Chen Y, Yao XS, Yang MS (2007) The effects of carbon nanotubes on the proliferation and differentiation of primary osteoblasts. Nanotechnology 18:475102

Chapter 6 Carbon Nanotube Uptake and Toxicity in the Brain Leying Zhang, Darya Alizadeh, and Behnam Badie Abstract The development of novel drug delivery systems is essential for the improvement of therapeutics for most human diseases. Currently used cellular delivery systems, such as viral vectors, liposomes, cationic lipids, and polymers, may have limited clinical efficacy because of safety issues, low gene transfer efficiency, or cytotoxicity. Carbon nanotubes (CNTs) have garnered much interest as possible biological vectors after the recent discovery of their capacity to penetrate cells. Inspite of the prominence of CNT studies in the nanotechnology literature, exploration of their application to central nervous system (CNS) therapeutics is at a very early stage. Before CNTs are used for treatment of brain and spinal cord disorders, however, several issues such as their CNS penetration and toxicity need to be addressed. Here, we discuss methods by which CNT uptake and toxicity can be assessed in animal models. Key words: Central nervous system, Macrophage, Microglia, Blood–brain barrier, Flow cytometry

1. Introduction The central nervous system (CNS) is considered a “privileged” organ because of the presence of capillary tight junctions that are integral components of a healthy blood–brain barrier (BBB) (1). Although this barrier is crucial in blocking the diffusion of harmful toxins and organisms into the CNS, it can also prevent the penetration of therapeutic drugs into diseased brain tissue. In fact, BBB is a major reason for lack of response of malignant brain tumors to systemically administered chemotherapies (2). The BBB, however, is not an absolute state (3). Analysis of the tracer experiments, which initially demonstrated lack of penetration of chemicals into the CNS, has revealed labeling of perivascular phagocytic cells suggesting active uptake of chemicals even through tight-junctions (3). Furthermore, in contrast to traditionally accepted

K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_6, © Humana Press, a part of Springer Science + Business Media, LLC 2010

55

56

Zhang, Alizadeh, and Badie

notion that the BBB restricts penetration of inflammatory cells into the CNS, it is now believed that continuous surveillance of CNS tissue is taking place by immune cells such as resident microglia and circulating lymphocytes (3). In addition, under pathological conditions, such as inflammation, autoimmune diseases, and cancers, immune-competent cells and monocytes can readily cross the disrupted BBB to participate in tissue repair or injury (4). Overall, these observations suggest that despite a strict and controlled mechanism for passage of chemicals through the BBB, the presence of cellular pathways may provide an alternate system for therapeutic drug delivery into the CNS. Use of nanoparticles (NPs) as drug carriers into the brain has received significant attention recently (5, 6). Polymer-based NPs have many advantages such as high systemic bioavailability, attenuation of early drug degradation, less toxicity, and high solubility. Complex macromolecules, for example, can be designed to carry not only therapeutic drugs, but also ligands that could specifically bind and target specific receptors on tumor cells. It is thought that extravasation of these molecules through an abnormal vasculature, in addition to specific surface ligands can potentially lead to higher local drug concentrations and uptake in pathological conditions. Although this hypothesis has been proven to be correct in a number of tumor models, penetration of these macromolecules through an intact BBB could hinder their uptake into the CNS. To overcome this limitation, nanoparticles are either covered by tumor specific ligands (such as EGFR antibody) or delivered directly into the CNS using convention-enhanced or bulk-flow delivery techniques (5). Another delivery technique that has not been as well studied is the uptake and transport of nanoparticles by tumor stromal cells such as tumor-associated macrophages (TAMs). Arising from both resident brain MG and circulating monocytes, TAMs represent a significant component of inflammatory response to gliomas (7–10). In CNS inflammation, activated MG and macrophages (MP) release proinflammatory cytokines and become capable of phagocytosis and antigen presentation (4). Several studies have also confirmed an intact TAM phagocytic function in gliomas (11, 12) and efficient internalization of carbon nanotubes (CNTs) (13). The ability of CNTs to penetrate the cellular membrane has promoted their application as potential nanovectors for drug delivery. Functionalized CNTs (f-CNT) have been used to deliver drugs, proteins, and nucleic acids into cells, both in vitro and in vivo (14, 15). Studies by Pantarotto et al. demonstrate a high antibody response with virus-neutralizing capacity in response to peptide-functionalized CNT immunization (16). Interestingly, CNT devoid of the peptide moiety were not immunogenic, as antibodies against CNT were not detected in this study.

Carbon Nanotube Uptake and Toxicity in the Brain

57

In addition to peptides, f-CNTs have been shown to effectively deliver genes and small DNAs into various models. Ammoniumfunctionalized CNT have been used to efficiently transfer plasmid reporter vectors into CHO cells in vitro, resulting in five- to tenfold higher levels of gene expression when compared with conventional DNA transfection techniques (17, 18). f-CNTs can also be used to deliver short oligodeoxynucleotides or short interfering RNAs (siRNA) into cells. Zhang et al. reported that using f-CNT to effectively deliver telomerase reverse transcriptase siRNA and inhibits lung tumor growth both in vitro and in vivo (19). Many safety concerns exist about in vivo utilization of CNTs. Before clinical applications are developed, environmental and biological safety and toxicity of CNTs must be carefully analyzed. Toxicity studies, conducted largely in mice have yielded several observations ranging from no toxicity or measurable inflammation (20), to granuloma formation in lungs (21), induced inflammation, including substantial lung neutrophil influx (22), and mortality at high doses (23), following intratracheal CNT administration. Similar discrepancies concerning cytotoxicity emerge when CNTs are administered intravenously and intraperitoneally. To the best of our knowledge, CNS toxicity of CNTs after intracranial administration has not yet been reported. Here, we detail flow cytometry and PCR methods by which CNT uptake and toxicity can be assessed in animal brains following their intracranial administration.

2. Materials 2.1. Carbon Nanotube Preparation

1. A sterile scalpel blade for removing CNT from the substrate. 2. Level II biosafety cabinet. 3. 1% Pluronic F108 (PF108) (BASF Corporation). 4. Mini Beadbeater-8 (Biospec Products) for homogenizing CNT–PF108. 5. VirSonic 300 (The Virus Company). 6. Centrifuge tubes (Beckman 50 Ultra-Clear Tubes ½ × 2 in.). 7. Ultracentrifuge.

2.2. Scanning and Transmission Electron Microscopy

1. 1.5% glutaraldehyde. 2. 0.1 M cacodylate buffer, pH 7.25. 3. 1% osmium tetroxide. 4. Alcohol. 5. Propylene oxide.

58

Zhang, Alizadeh, and Badie

6. Leica Ultracut UCT ultramicrotome. 7. 300 mesh uncoated copper grids. 8. 5% uranyl acetate solution. 9. Reynolds lead stain. 10. FEI Tecnai Twin 12 transmission electron microscope. 11. Gatan Ultrascan 1000 CCD camera. 2.3. PKH26 Labeling of CNTs

1. PKH26, a hydrophobic red fluorescent dye (Sigma). 2. PKH26 stock solution (10–3 M). 3. Diluent C. 4. Ultracentrifuge. 5. 1% Pluronic F108 (PF108) (BASF Corporation).

2.4. Confocal Microscopy

1. Zeiss LSM 510 Meta inverted 2-photon confocal microscope. 2. 33 mm2 cell culture dishes. 3. DMEM medium supplemented with 10% heat-inactivated FBS (BioWhittaker, Walkersville, MD), 100-U/ml penicillin–G, 100-mg/ml streptomycin, and 0.01-M Hepes (Life Technologies, Gaithersburg, MD). 4. CO2 incubator. 5. A Zeiss LSM 510 Meta inverted 2-photon confocal microscope with a 40×/0.6 objection for fluorescent imaging studies.

2.5. In Vivo Delivery of Labeled CNTs

1. Animals: C57BL6 mice (female) 6–8 week of age and weighing 16–17 g. 2. Stereotactic head frame. 3. Hamilton syringe. 4. 0.5 cc disposable syringe. 5. Ketamine (132 mg/kg)/ Xylazine (8.8 mg/kg). 6. Buprenex. 7. Scalpel. 8. Eye ointment. 9. Drill. 10. Incision glue. 11. Clipper. 12. Betadine. 13. 70% ethanol. 14. Gauze. 15. Cotton Tips.

Carbon Nanotube Uptake and Toxicity in the Brain

2.6. Flow Cytometry Analysis

59

1. Fresh tissue Samples. 2. Sterile scissors (with small and large blades), Scalpel blade, hemostat, and forceps. 3. Falcon centrifuge. 4. Gauze. 5. 70% ethanol. 6. Flow wash buffer, 1× PBS, 1%FBS, 2 mM EDTA. 7. 1× PBS. 8. Snaptop tubes (14 ml). 9. Falcon cell strainer (40 mm). 10. Syringe plunger. 11. Mouse serum. 12. Hemacytometer. 13. Pipette (5 ml). 14. FACSCalibur fluorescence cell sorter (BDIS, San Jose, CA).

2.7. RNA Isolation, cDNA Generation and Real-Time PCR

1. Trizol Reagent (Invitrogen Cat # 15596-026). 2. Chloroform. 3. Isopropyl alcohol. 4. 75% Ethanol (in DEPC-treated water). 5. RNase-free water. 6. Glass–Teflon homogenizer. 7. Qiagen RNeasy Mini Kit (Cat# 74106). 8. Nanodrop for RNA measurement. 9. Invitrogen Superscript III First-Strand Synthesis System catalog # 18080-051. 10. Bio-Rad Sybr Green. 11. Set of primer (6.25 mM).

3. Methods 3.1. CNT Preparation

1. Remove CNTs from the substrate by manual scraping with a sterile scalpel blade into a dish containing 1% Pluronic F108 (PF108) in a level II biosafety cabinet. 2. Transfer the CNTs solution to an eppendorf tube. 3. The CNT–PF108 suspension is homogenized by Mini Beadbeater-8 for 45 min. 4. The CNT–PF108 suspension is sonicated at 540 W for 30 min.

60

Zhang, Alizadeh, and Badie

5. CNTs are concentrated by ultracentrifugation at 122,000 × g for 4 h at room temperature. 6. The upper 55–60% of supernatant is then carefully decanted, leaving micelle-suspended nanotube solution (see Note 1). 7. The visible pellet is reused for additional sonication to increase yield (see Note 2). 3.2. Scanning and Transmission Electron Microscopy

1. Transmission electron microscopy (TEM) samples are fixed with 1.5% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.25 for 72 h at 4°C. 2. Wash in buffer and postfix with 1% osmium tetroxide for 50 min at 4°C. 3. Wash again in buffer, dehydrate through a graded series of alcohols, put through propylene oxide and embed in Eponate. 4. Following polymerization at 70°C, thin (80 nm) sections are cut with a diamond knife on a Leica Ultracut UCT ultramicrotome. 5. Pick up sections on 300 mesh uncoated copper grids, stain for 15 min with an aqueous 5% uranyl acetate solution followed by a 2 min stain with Reynolds lead stain. 6. The grids are then observed with an FEI Tecnai Twin 12 transmission electron microscope, and photographed with a Gatan Ultrascan 1000 CCD camera (see Note 3).

3.3. PKH26 Labeling of CNTs

1. CNTs are tagged with PKH26, a hydrophobic red fluorescent dye. 2. The PKH26 stock solution (10–3 M) is diluted to (2 × 10−6 M) using Diluent C and added to the CNT stock at a 1:1 ratio. 3. The CNT–PKH26 solution is incubated at room temperature for 5 min and diluted to (10−9 M). 4. CNT–PKH26 are washed at 122,000 × g for 4 h and 4°C. 5. Dilution of excess PKH26 and washes are repeated three times. 6. MWCNT–PKH26 are resuspended to 80 mg/ml using 1% PF108 and used within 48 h of labeling.

3.4. Confocal Microscopy

1. Cells are plated in 33 mm2 dishes in 2 ml media and CNTs-PKH26 are added at various concentrations. 2. Time-lapse experiments are conducted for 24 h with samples incubated at 37°C with 7% CO2. 3. The Argon 488 nm and HeNe 543 nm lasers are used to excite GFP and PKH or Cy3, respectively (Fig. 1). 4. Confocal images are taken every 10 min for 24 h for time-lapse experiments or at time-points otherwise noted.

Carbon Nanotube Uptake and Toxicity in the Brain

3.5. In Vivo Injection of CNTs-PKH

61

1. Sterilize procedure room with 70% ethanol. 2. Anesthetize mice with Ketamine (132 mg/kg)/ and Xylazine (8.8 mg/kg) solution. 3. Clip all hair from scalp and wipe with betadine solution using sterile cotton tips. 4. Make a small sagittal paramedian incision with a scalpel blade. 5. Drill injection site (1 mm right of midline and 1 mm forward of bregma). 6. Set time on 10 min and place mouse in a stereotactic head frame. 7. Pull 10 ml of labeled CNT–PKH (5 mg) into a Hamilton syringe. 8. Line up the syringe to the top of the hole, move the syringe 3 mm down into the brain and start the timer. 9. Every 2 min, pull up the syringe (1/2 mm) and inject material slowly into the brain (see Note 4). 10. Let syringe needle sit for one more min and then slowly pull it out of brain. 11. Wipe the area with sterile gauze or cotton tip. 12. Using wound glue, the incision area is closed and apply eye ointment. 13. For possible postop pain, inject 1 ml of Buprenex into peritoneum cavity.

Fig. 1. In vitro uptake of CNT by microglia. PKH-labeled CNTs (red particles) were incubated with GFP-expression GL261 glioma cells (green cells) and BV2 microglia (round cells) for 24 h and imaged using a Zeiss LSM 510 Meta inverted 2-photon. LSM 510 Meta confocal microscope. Microglia were more efficient in CNT uptake than glioma cells

62

Zhang, Alizadeh, and Badie

3.6. Flow Cytometry Analysis

1. Euthanize animals using CO2. 2. Sterilize neck and head area with 70% ethanol. 3. Disconnect spine at the neck area with sharp scissors. 4. Hold head on both sides and place the edge of short blade scissors into spinal column. 5. Cut skull in an axial plane on the left and right temporal area (move slowly to avoid cutting the brain tissue). 6. Use hemostat and pull skull cap back so that brain is exposed. 7. Use forceps to remove brain. 8. Dissect the section desired (tumor site) with scalpel blade. 9. Place the section in flow buffer on ice. 10. Mince tissue with scissors. 11. Triturate with pipette (5 ml). 12. Filter cells through a 40-mm filter. Use syringe plunger and 1× PBS to force cells through the filter. 13. Spin filtered cells at 1,200 rpm for 4 min. 14. Resuspend cells with flow buffer and add mouse serum. 15. Count cells with hemacytometer. 16. Each antibody should be optimized to best dilution depending on the type of antibody and tissue sample (see Note 4). 17. Samples are then analyzed using an FACSCalibur fluorescence cell sorter (see Note 5). 18. FlowJo 6.4.7 is used for data analysis (Fig. 6.2).

3.7. RNA Isolation, cDNA Generation and Real-Time PCR

1. Harvest and prepare brain tissue as discussed in the previous section (Subheading 3.6.1–3.6.10). 2. Use Invitrogen Trizol to isolate total RNA according to instructions. 3. Use Qiagen RNeasy Mini Kit to purify total RNA according to instructions. 4. Measure RNA concentration. 5. For each reaction, use 1–2 mg RNA for RT-PCR. 6. Use Invitrogen Superscript III First-Strand System for generating cDNA according to instructions. 7. Bring final reaction following to 120 ml for each initial 20 ml reaction volume and store at −80°C for real-time PCR. 8. Real-time PCR. 9. Each sample is run in triplicate. 10. For each reaction use: (a) Sybr Green 12.5 ml (b) Forward Primer 0.5 ml

Carbon Nanotube Uptake and Toxicity in the Brain

63

Fig. 2. CNTs are predominantly internalized by macrophages in a mouse glioma model. Intracranial GL261 tumors were directly injected with CNT–PKH (5 mg). Twenty-four hours later, animals were sacrificed and tumors harvested for analysis by flow cytometry. A representative histogram demonstrating that tumor-infiltrating macrophages (CD45high/ CD11b/chigh) phagocytosed CNTs (blue events) significantly more efficiently than other cell types. Tumor cells are represented as: CD45−/CD11b/c−, microglia as: CD45low/ CD11b/chigh, and lymphcytes as: CD45high/CD11b/c−

(c) Reverse Primer 0.5 ml (d) cDNA 4 ml (e) H2O to 25 ml final volume 11. PCR amplification condition: 1 cycle at 50°C for 2 min, 1 cycle at 95°C for 10 min, and 50 cycles at 95°C for 15 s and 60°C for 1–2 min (Fig. 3).

4. Notes 1. CNT preparation can differ from time to time. It is important to check if the 60% upper supernatant still contains CNTs by electron microscopy. If the supernatant still contains significant amount of CNTs, ultracentrifugation time needs to be increased. 2. CNT particles are stable for several months when stored at 4°C. However, even short periods of storage can lead to CNT aggregation. Repeat sonication is necessary before they can be reused. 3. Measuring the concentration of CNTs after the preparation is difficult. Techniques such as spectrophotometry can be used to estimate the concentration of CNT solutions based on their near-IR light absorption properties (15).

64

Zhang, Alizadeh, and Badie

Fig. 3. CNT toxicity in normal mouse brains. CNT–PKH (5 mg, dashed lines) or saline (control, solid lines) was injected into normal mouse brains. Tissue samples were harvested 0, 24, 48 and 72 h later and analyzed by quantitative RT-PCR. CNT injection did not significantly change the expression of inflammatory cytokines such as TNF-a or IL1-b suggesting minimal toxicity in this model. (n = 3 per time point)

4. Rapid injection on CNTs into the brain can lead to reflux of material through the needle tract. Slow injection over several min can prevent this. 5. Flow cytometry can be used for determination of various cell types in the tissue. The example given here is to characterize macrophage and lymphocytes in a brain tumor using specific surface markers.

Acknowledgments This work was supported by the American Cancer Society Research Scholar Grant (RSG-03-142-01-CNE) and James S. McDonnell Foundation. The City of Hope Flow Cytometry Core was equipped in part through funding provided by ONR N00014-02-1 0958, DOD 1435-04-03GT-73134, and NSF DBI-9970143.

Carbon Nanotube Uptake and Toxicity in the Brain

65

References 1. Bechmann I, Galea I, Perry VH (2007) What is the blood-brain barrier (not)? Trends Immunol 28:5–11 2. Muldoon LL, Soussain C, Jahnke K, Johanson C, Siegal T, Smith QR, Hall WA, Hynynen K, Senter PD, Peereboom DM, Neuwelt EA (2007) Chemotherapy delivery issues in central nervous system malignancy: a reality check. J Clin Oncol 25:2295–2305 3. Galea I, Bechmann I, Perry VH (2007) What is immune privilege (not)? Trends Immunol 28:12–18 4. Djukic M (2006) Circulating monocytes engraft in the brain, differentiate into microglia and contribute to the pathology following meningitis in mice. Brain 129:2394–2403 5. Jain KK (2007) Use of nanoparticles for drug delivery in glioblastoma multiforme. Expert Rev Neurother 7:363–372 6. Pardridge WM (2007) Drug targeting to the brain. Pharm Res 24:1733–1744 7. Badie B, Schartner JM (2000) Flow cytometric characterization of tumor-associated macrophages in experimental gliomas. Neuro surgery 46: 957–961; discussion 61–62 8. Roggendorf W, Strupp S, Paulus W (1996) Distribution and characterization of microglia/macrophages in human brain tumors. Acta Neuropathol 92:288–293 9. Shinonaga M, Chang CC, Suzuki N, Sato M, Kuwabara T (1988) Immunohistological evaluation of macrophage infiltrates in brain tumors. Correlation with peritumoral edema. J Neurosurg 68:259–265 10. Streit WJ (1994) Cellular immune response in brain tumors. Neuropathol Appl Neurobiol 20:205–206 11. Hussain SF, Yang D, Suki D, Grimm E, Heimberger A (2006) Innate immune functions of microglia isolated from human glioma patients. J Transl Med 4:15 12. Nickles D, Abschuetz A, Zimmer H, Kees T, Geibig R, Spiess E, Regnier-Vigouroux A (2008) End-stage dying glioma cells are engulfed by mouse microglia with a straindependent efficacy. J Neuroimmunol 197: 10–20 13. Kateb B, Van Handel M, Zhang L, Bronikowski MJ, Manohara H, Badie B (2007) Internalization of MWCNTs by microglia: possible application in immunotherapy of brain tumors. Neuroimage 37(Suppl 1): S9–S17 14. Klumpp C, Kostarelos K, Prato M, Bianco A (2006) Functionalized carbon nanotubes as

15.

16.

17.

18.

19.

20.

21.

22.

23.

emerging nanovectors for the delivery of therapeutics. Biochim Biophys Acta 1758: 404–412 Liu Z, Chen K, Davis C, Sherlock S, Cao Q, Chen X, Dai H (2008) Drug delivery with carbon nanotubes for in vivo cancer treatment. Cancer Res 68:6652–6660 Pantarotto D, Partidos CD, Hoebeke J, Brown F, Kramer E, Briand JP, Muller S, Prato M, Bianco A (2003) Immunization with peptide-functionalized carbon nanotubes enhances virus-specific neutralizing antibody responses. Chem Biol 10:961–966 Pantarotto D, Singh R, McCarthy D, Erhardt M, Briand JP, Prato M, Kostarelos K, Bianco A (2004) Functionalized carbon nanotubes for plasmid DNA gene delivery. Angew Chem Int Ed Engl 43:5242–5246 Singh R, Pantarotto D, Lacerda L, Pastorin G, Klumpp C, Prato M, Bianco A, Kostarelos K (2006) Tissue biodistribution and blood clearance rates of intravenously administered carbon nanotube radiotracers. Proc Natl Acad Sci U S A 103:3357–3362 Zhang Z, Yang X, Zhang Y, Zeng B, Wang S, Zhu T, Roden RB, Chen Y, Yang R (2006) Delivery of telomerase reverse transcriptase small interfering RNA in complex with positively charged single-walled carbon nanotubes suppresses tumor growth. Clin Cancer Res 12:4933–4939 Lacerda L, Bianco A, Prato M, Kostarelos K (2006) Carbon nanotubes as nanomedicines: from toxicology to pharmacology. Adv Drug Deliv Rev 58:1460–1470 Shvedova AA, Kisin ER, Mercer R, Murray AR, Johnson VJ, Potapovich AI, Tyurina YY, Gorelik O, Arepalli S, Schwegler-Berry D, Hubbs AF, Antonini J, Evans DE, Ku BK, Ramsey D, Maynard A, Kagan VE, Castranova V, Baron P (2005) Unusual inflammatory and fibrogenic pulmonary responses to singlewalled carbon nanotubes in mice. Am J Physiol Lung Cell Mol Physiol 289: L698–L708 Nemmar A, Hoet PH, Vandervoort P, Dinsdale D, Nemery B, Hoylaerts MF (2007) Enhanced peripheral thrombogenicity after lung inflammation is mediated by plateletleukocyte activation: role of P-selectin. J Thromb Haemost 5:1217–1226 Lam CW, James JT, McCluskey R, Hunter RL (2004) Pulmonary toxicity of single-wall carbon nanotubes in mice 7 and 90 days after intratracheal instillation. Toxicol Sci 77: 126–134

Chapter 7 In Vitro and In Vivo Biocompatibility Testing of Functionalized Carbon Nanotubes Gianni Ciofani, Vittoria Raffa, Orazio Vittorio, Alfred Cuschieri, Tommaso Pizzorusso, Mario Costa, and Giuseppe Bardi Abstract The explosive growth of nanotechnology in the last years has led to dramatic innovations in pharmacology, and it is revolutioning the development of biologically active compounds. Carbon nanotubes (CNTs) are widely explored for biomedical applications such as intracellular transporters for (bio)molecules, and represent promising future tools for efficient and safe cell therapy. Due to their nanoscale dimensions, the ability to interact with cells, and their easy functionalization, CNTs are close-to-ideal vectors for an efficient and safe cell therapy, obviating the risks associated with the use of viral vectors. Notwithstanding, conflicting data concerning the biocompatibility of CNTs have been reported in the literature; while some studies point toward very low toxicity of CNTs both in vitro and in vivo, others reveal various toxic effects such as oxidative stress, DNA damage, and cell apoptosis. Thus, standardized methods and independent test systems are urgently needed to verify cytotoxicity data in this research field. In this chapter, we summarize the used methods and the achieved main results in our laboratories concerning multiwalled carbon nanotubes (MWCNTs) biocompatibility studies. The in vitro response of human neuroblastoma cell line and primary mouse neurons was investigated following the exposure to different samples of MWCNTs in order to evaluate their effects on cell viability, oxidative stress, and apoptosis. Moreover, in vivo neurocompatibility tests were carried out through injections in mouse brains. Key words: Carbon nanotube biocompatibility, Pluronic F127, Cell lines, Primary neurons, Cerebral injections, Metabolic assessment, Apoptosis, Oxidative stress

1. Introduction Carbon nanotubes (CNTs) have recently entered the fields of biology and medicine (1). This has stirred increasing interest in biocompatibility issues. Concurrently, driven by the obvious need for toxicological evaluation of the health risks, the field

K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_7, © Humana Press, a part of Springer Science + Business Media, LLC 2010

67

68

Ciofani et al.

of nanotoxicology, defined as ‘‘science of engineered nanodevices and nanostructures that deals with their effects in living organisms’’, has emerged (2). The first review of the safety risk associated with CNTs has been published in 2003 (3). Previously, only few studies have addressed the toxicity of CNTs, one of which announced CNTs as the “next asbestos” (4). In the following years, studies on health risks related to nanoparticles (5) demonstrated that they enter the human body via the lungs or the intestines, depending on their size, on their surface properties, and where they are applied. The most extensive analysis of the research devoted to the toxicology of new CNTs has been collected in a Special Issue of the journal Carbon (Vol. 44) in 2006. During the last 5 years, conflicting data concerning the biocompatibility of CNTs have been reported; some demonstrate very low toxicity of CNTs in vitro or in vivo (6), while others indicate various toxic effects such as oxidative stress, DNA damage, and cell apoptosis (7). This discrepancy is largely due to differences in the CNT production, purity, and functionalization, as well as the used cell types. Some authors have suggested the interaction of nanotubes with the agents used in some standard cell proliferation assays as the possible source for the conflicting results, underscoring the need for standardized methods and independent test systems to verify cytotoxicity data (8). As mass-produced CNTs are strongly aggregated and highly hydrophobic, processes to render them water soluble are required for biological applications. To this end, suspensions in surfactant solutions are often exploited. Among these, the nonionic surfactant Pluronic F127 (PF127 polyoxyethylene–polyoxypropylene […] block copolymer with MW of 12,600) has been proven to be highly biocompatible if used at low concentrations. On this basis, we have recently developed an efficient and reproducible procedure to obtain highly concentrated and stable multiwalled carbon nanotube (MWCNT) dispersions in biocompatible aqueous solution (9). Briefly, these samples were prepared by mixing 5 mg of MWCNTs with 10 ml of a 0.1% PF127 solution (in PBS or growth medium), followed by sonication for 12 h with a Branson sonicator 2510, using an output power of 20 W. A pretreatment step involved placing the samples over a hot plate with stirring for 6 h, and subsequent centrifugation at 1,100 × g for 10 min in order to remove impurities. The resulting CNT concentration of the supernatant was determined by spectrophotometry at 270 nm (10). In this chapter, we describe the used methods and the achieved main results in our laboratories concerning the biocompatibility of MWCNTs toward human neuroblastoma cell line, mouse primary neurons, and in vivo in mouse brains. The in vitro response of human neuroblastoma cell line was investigated following exposure to different samples of MWCNTs in order to evaluate the effect of nanotube purity and surface oxidation on cytocompatibility. Three samples with varying purity were tested, namely, MWCNT 97%, MWCNT 99%, and MWCNT 97% oxidized by

In Vitro and In Vivo Biocompatibility Testing of Functionalized Carbon Nanotubes

69

HNO3, with the latter containing ~8% (w/w) of carboxylic groups on the nanotube surface. The in vitro primary cell culture and in vivo experiments were performed with MWCNT 97%.

2. Materials 2.1. Cell Line Culture

1. Human neuroblastoma cell line (SH-SY5Y; American Type Culture Collection, ATCC, Rockville, MD). 2. Dulbecco’s Modified Eagle’s Medium and Hams F-12 medium (DMEM; Lonza, Milan, Italy). 3. SH-SY5Y complete culture medium is a mixture 1:1 of Hams F12 and DMEM supplemented with 2-mM l-glutamine (Lonza), 100-IU/ml penicillin, 100-µg/ml streptomycin and 10% heat inactivated foetal bovine serum (FBS, Lonza). 4. The MWCNT-modified culture medium was obtained by mixing the MWCNT solution and the culture medium at a ratio 1:10 v/v, corresponding to a MWCNT concentration of 5 µg/ml and a PF-127 concentration of 0.01%. 5. Solution of trypsin (0.25%) and ethylenediamine tetraacetic acid (EDTA, 1 mM; Lonza). 6. 6-well and 24-well cell culture plate, 96-well microplate tissueculture treated in polystyrene (BD Biosciences, NJ, U.S.A.).

2.2. Metabolic Viability Assays: MTT and WST-1

1. PBS 10× stock solution (Phosphate buffered saline solution) cell culture tested (Sigma). Prepare working solution by dilution of one part with nine parts water and autoclave before storage at 4°C. 2. MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide; Sigma, St. Louis, MO). The lyophilized MTT should be stored at −20°C. 3. DMSO (Dimethyl sulfoxide, Sigma). Store at room temperature. 4. Quick cell proliferation assay kit contents: WST-1 2-(4-iodo phenyl)-3-(4-nitophenyl)-5-(2,4-disulfophenyl)-2H-tetrazoilium monosodium salt lyophilized; electro coupling solution (BioVision, CA, U.S.A.). Store at −20°C. 5. The absorbance was measured on a Versamax microplate reader (Molecular Devices, Sunnyvale, CA).

2.3. Hoechest Cell Staining for the Evaluation of Apoptotic Cell Death

1. Hoechst 33258, pentahydrate (bis-benzimide) 10 ml, 10 mg/ ml (16.0 mM) solution in water (Invitrogen, USA). Prepare a working solution 0.5 µg/ml by dilution in PBS 1×, protect from light and store at 2–6°C. 2. Wash buffer: PBS 1× (Sigma).

70

Ciofani et al.

2.4. Oxidative Stress

1. The Image-iT™ LIVE Green Reactive Oxygen Species (ROS) Detection Kit (Invitrogen, MD, U.S.A.) provides the key reagents necessary for the detection of ROS in live cells. This kit contains: • Component A: 5-(and-6)-carboxy-2¢,7¢-dichlorodihydro fluorescein diacetate (carboxy-H2DCFDA), five vials each containing 275 mg • Component B: Hoechst 33342, 400 ml of a 1-mM solution • Component C: tert-butyl hydroperoxide (TBHP) solution, 50 ml of a 7.78-M solution • Component D: Dimethylsulfoxide (DMSO), 500 ml 2. Upon receipt, store the kit at 2–6°C, desiccated, protected from light. It is important that the TBHP solution (Component C) is not frozen. Solutions of carboxy-H2DCFDA in DMSO can be divided into aliquots and stored at −20°C. 3. Hank’s balanced salt solution with calcium and magnesium (HBSS/Ca/Mg, Gibco, MD, USA).

2.5. Optical and Fluorescent Microscopy

1. Optical and fluorescent microscopy was performed with a Nikon TE2000U fluorescent microscope equipped with Nikon DS-5MC USB2 cooled CCD camera. 2. Microscope cover slips (Fisher, Pittsburgh, PA). 3. Paraformaldehyde (Fisher): prepare a 4% (w/v) solution in PBS fresh for each experiment. The solution may need to be carefully heated (use a stirring hot-plate) to dissolve, and then cooled to room temperature for use. 4. Cells counterstain: Evans Blue 0.5% w/v in PBS 1× (Sigma). 5. Mounting media: VECTASHIELD (Vector Laboratories, England).

2.6. In Vitro Study on Primary Neurons

1. Instruments: medium scissors, #2 forceps, iris scissor, glass Pasteur pipettes, sterile pipettes, tips, dishes, coverslips and cell media. 2. Dissection medium (DM): 16 mM glucose, 22 mM sucrose, 10 mM Hepes, 160 mM NaCl, 5 mM KCl, 1 mM Na2HPO4, 0.22 mM KH2PO4; pH 7.4; 320–330 mOsm. 3. Hanks’ Balanced Salt Solution (HBSS; Invitrogen, USA). 4. Solution of trypsin (0.25%) and DNAse (1:20 DNAse; stock: 3 mg DNAse + 5 mg MgSO4 in 2 ml, Sigma–Aldrich). 5. Dulbecco’s Modified Eagle’s Medium and heat inactivated foetal bovine serum (Lonza); Neurobasal (Invitrogen, USA); B27 (Invitrogen, USA); glutamine (Euroclone S.p.A., Italy); gentamicin (Euroclone S.p.A., Italy). 6. Polylysine for dish coating (Sigma, USA, #P2658).

In Vitro and In Vivo Biocompatibility Testing of Functionalized Carbon Nanotubes

71

7. Ara C (cytosine arabinoside) (Sigma, USA, #C1768). 8. Paraformaldehyde (Fisher): prepare a 4% (w/v) solution in PBS fresh for each experiment. 9. Hoechst 33342, 1 µg/ml (Clonetech Laboratories, Inc. CA94043, USA). 10. ApoAlert Kit (Clonetech Laboratories, Inc. CA94043, USA) was used to evaluate differences between normal and apoptotic cells after treatments. The kit contains: • Annexin V–FITC (20 mg/ml in Tris-NaCl) • 1× Binding Buffer • Propidium Iodide (50 mg/ml in 1× binding buffer) 2.7. In Vivo Study

1. Avertin solution for anaesthetization: (1.25 g tribromoethanol (Sigma–Aldrich)/2.5 ml tert–amyl–alcohol (J.T.Baker) in 10.5% saline solution; 0.1 ml/g mouse body weight). 2. Topic antibiotic: chlortetracycline (CTC) 3% (Wyeth Lederle SpA, Italy). 3. Paraformaldehyde (Fisher): prepare a 4% (w/v) solution in PBS for each perfusion. 4. Cresyl violet, 0.5% (w/v) in H2O (Sigma–Aldrich).

3. Methods 3.1. Cell Culture Preparation and MTT and WST-1, Assays

1. To determine the effect of CNTs on cell viability, MTT and WST-1 cell proliferation assays were used. 2. For MTT assay, 25 × 10³ cells were seeded into each well of a 96-well plate and then incubated with the culture media for 72 h. The culture medium is then replaced with 100 µl of medium containing 0.5 mg/ml of MTT (in PBS) and the plates are incubated for 2 h at 37°C and 5% CO2. 3. Mitochondrial dehydrogenases of viable cells reduce yellow water-soluble MTT to water-insoluble formazan crystals. The MTT-containing medium is again removed and replaced with 100 µl of DMSO and left for 10 min on a platform shaker to solubilize the converted formazan in the culture plates. 4. The absorbance was measured with a Versamax microplate reader at a wavelength of 570 nm with background subtracted at 690 nm. 5. As previous literature states that nanotube absorbance and its interaction with formazan could influence the results of the MTT assay at high nanotube concentration (11), WST-1 assay as double check was also performed.

72

Ciofani et al.

6. Again 25 × 103 cells were seeded into each well of a 96-well plate and then incubated with the culture media. After 72 h of cell incubation, 10 µl of WST-1 solution was added (as described in quick cell proliferation assay kit, BioVision, USA) and incubated for 2 h in standard conditions. 7. The mitochondrial dehydrogenases reduced WST-1 in tetrazolium salt which is water-soluble, and therefore no DMSO treatment is necessary. The absorbance of treated and untreated samples was measured at a wavelength of 450 nm with the reference wavelength 650 nm. 8. The results of the both assays are given as percentage of the control assay. 9. After 3 days of incubation, the MTT test detects about 8–10% loss of viability, while WST-1 assay does not indicate a decrease of viability in cell cultures treated with the MWCNT-modified culture media when compared with the control cultures. An explanation of this divergence is provided in the work of Wörle-Knirsch and colleagues (11). Their experimental data coming from A549 cells incubated with SWCNTs revealed a strong cytotoxic effect within the MTT assay, whereas the same treatment with the same nanotubes, but detection with WST-1, reveals no cytotoxicity. Nanotubes appear to interact with some tetrazolium salts such as MTT but not with others (such as WST-1, INT, XTT). This interference does not seem to affect the enzymatic reaction but lies rather in the insoluble nature of MTT-formazan. Based on this experimental evidence and on the suggestions of these authors, viability of cells incubated with culture medium supplemented with different MWCNTs and Pluronic 0.01% for 2 weeks was monitored every week with WST-1 assay. 10. 2 × 105 neuroblastoma cells were seeded into each well of a 6-well plate and then incubated with the culture media. After 3 days of cells incubation, cells were detached with trypsin/ EDTA and reseeded at 1:4 ratio into each well of a 6-well plate. The same procedure was repeated again after 3 days, but seeding, at the same time, aliquots of cell samples in six wells of 96-well microplate (used for WST-1 absorbance measuring). 11. After 24 h, the WST-1 assay was performed showing the viability of cells treated with MWCNTs solution after 1 week. The same procedure was repeated after 2 weeks (Fig. 1). 12. Experiments showed that after 1 week of treatment with MWCNTs 97% and 97% HNO3 the viability of cells decreases, (this effect is more evident in the second week of cell propagation in the presence of nanotubes). Moreover, it is clear that cell treated with MWCNT 99% maintained the highest

In Vitro and In Vivo Biocompatibility Testing of Functionalized Carbon Nanotubes

73

viability, confirming the importance of the purity of MWCNTs (Fig. 2). 13. Results are shown as the mean ± S.E.M. of six experiments, each one carried out in triplicate. Statistical significance was assessed by one way analysis of variance (ANOVA) followed by posthoc comparison test (Tukey) at *p < 0.05, **p < 0.01, ***p < 0.001 (see Note 1). 3.2. Preparation of Hoechst Cell Staining for the Evaluation of Apoptotic Cell Death

1. 25 × 103 SH-SY5Y cultured on coverslips were exposed for 24, 48, and 72 h to MWCNT-modified culture media; control cells were treated with the culture medium and a PF-127 0.01%-modified culture medium. 2. SH-SY5Y cell death was evaluated by Hoechst 33258 0.5 µg/ ml staining of cell nuclei (excitation 346 nm; emission 460 nm), which enables determination of the presence of DNA condensates, a characteristic feature of apoptosis. 3. After washing with PBS, cells were fixed with 40 mg/ml paraformaldehyde in PBS and incubated for 30 min at 4°C. Cells were then washed three times with PBS and incubated for 20 min at 37°C with 0.5 µg/ml Hoechst 33258 in PBS. Cells were then washed again with PBS and dried. 4. Each coverslip was inverted onto a slide containing 10 µl of mounting media, removing the excess mounting media with fibre-free paper; thereafter, each coverslip was sealed with regular transparent nail polish and allowed to dry for 3 min. 5. The picnotic cells were observed with the fluorescence microscope (Fig. 3). The number of apoptotic nuclei was determined on at least five randomly selected areas from three

Fig. 1. Gantt diagram of the 2-weeks viability assay on SH-SY5Y cells

74

Ciofani et al.

Fig. 2. WST-1 assay results on SH-SY5Y cells after 1 and 2 weeks of incubation with MWCNTs. Results are the mean ± S.E.M. of six experiments each carried out in triplicate. Statistical significance was assessed by one way analysis of variance (ANOVA) followed by posthoc comparison test (Tukey) at *p < 0.05, **p < 0.01, ***p < 0.001

coverslips of each experimental group, each area containing approximately 100 cells. (see Note 2). 6. These images show that all three MWCNTs different samples dispersed in PF-127 0.01% did not induce apoptosis after 72 h of incubation on neuroblastoma cell line. In Fig. 3f, we observed 98.5% of apoptosis cells following incubation with a strong apoptosis inducer (TBHP, see Subheadings 2.4 and 3.3 for the protocols). 3.3. Cell and Solution Preparation for Oxidative Stress Assay

Many authors described the ability of CNTs to induce the formation of intracellular ROS in different cell lines (12, 13); this effect is due to the CNT concentration and their metal impurities content. We used the carboxy-H2DCFDA fluorescence as a reporter of intracellular oxidant production to study whether different purity CNTs could influence the induction of intracellular ROS in SH-SY5Y cells (Fig. 4).

In Vitro and In Vivo Biocompatibility Testing of Functionalized Carbon Nanotubes

75

Fig. 3. Evaluation of apoptotic cells rate after 72 h of incubation with MWCNTs: (a) Control sample; (b) SH-SY5Y treated with MWCNT 99%; (c) SH-SY5Y treated with MWCNT 97%; (d) SH-SY5Y treated with MWCNT 97% HNO3; (e) SH-SY5Y treated with PF-127 0.01%; (f) SH-SY5Y treated with 50 µM TBHP for 12 h

Fig. 4. Detection of intracellular reactive oxygen species in SH-SY5Y treated with MWCNTs. Red fluorescent Evans-blue was used for cell membrane staining. The induction of oxidative stress is represented by green fluorescent cells. All three samples of MWCNTs do not induce oxidative stress in neuroblastoma cell line after 72 h incubation. (a, b) SH-SY5Y treated with TBHP, a common inducer of ROS (positive control); (c, d) SH-SY5Y treated with PF-127 0.01% modified culture medium for 72 h; (e, f) SH-SY5Y treated with MWCNT 97% modified culture medium for 72 h; (g, h) SH-SY5Y treated with MWCNT HNO3 97% modified culture medium for 72 h; (i, j) SH-SY5Y treated MWCNT 99% modified culture medium for 72 h

76

Ciofani et al.

3.3.1. Oxidative Stress Assay

For oxidative stress assay, 1 × 105 SH-SY5Y cells were seeded into each well of a 6-well plate and then incubated with the culture media for 72 h. The following experimental protocols from Image-iT™ LIVE Green Reactive Oxygen Species Detection Kit, have been modified for SH-SY5Y.

3.3.2. Labelling with Carboxy-H2DCFDA; (see Note 3)

1. Add 50 ml of DMSO (Component D) to one vial of carboxyH2DCFDA (Component A, 275 mg) to make a 10-mM stock solution.

3.3.2.1. Preparation of 10-mM Carboxy-H2DCFDA Stock Solution

2. Vortex the vial until the powder is completely dissolved.

3.3.2.2. Preparation of 25-mM Carboxy-H2DCFDA Working Solution

1. Add 5.0 ml of the 10-mM carboxy-H2DCFDA stock solution (prepared using the steps given in “Preparation of 10 mM Carboxy-H2DCFDA Stock Solution” under Subheading 3.3.2) to 2.0 ml of warm HBSS/Ca/Mg.

3.3.2.3. Labelling of Cells

1. Gently wash cells once with warm HBSS/Ca/Mg. 2. Apply sufficient amount of the 25-mM carboxy-H2DCFDA working solution (prepared using the steps given in “Preparation of 25 mM Carboxy-H2DCFDA Working Solution” under Subheading 3.3.2) to cover the cells adhering to the coverslip(s). Incubate for 30 min at 37°C, protect from light.

3.3.2.4. (Optional) Counterstain with Hoechst 33342 or Evans Blue

1. Hoechst 33342 should be added at a final concentration of 1.0-mM to the carboxy-H2DCFDA staining solution during the last 5 min of the incubation in “Labelling of Cells” under Subheading 3.3.2. The Hoechst 33342 stain (Component B) is supplied as a 1.0 mM solution, so for each 1.0 ml of carboxy-H2DCFDA working solution used in “Labelling of Cells” under Subheading 3.3.2 add 1.0 ml of Hoechst 333342 stain. If counterstaining of the cell membrane is preferred, Evans blue 0.5% w/v in PBS during the last 2 min of the incubation in step 2 of subheading 3.3.2.3 can be used.

3.3.2.5. Mounting in Warm Buffer and Immediate Imaging

1. Gently wash the coverslips three times in warm HBSS/Ca/Mg.

3.3.3. Induction of Cellular ROS Production with TBHP

1. Add 1.0 ml of TBHP (Component C, 7.78 M) to 77 ml of high-purity water to make a 100-mM stock solution. Slow pipetting of the viscous TBHP solution is recommended.

3.3.3.1. Preparation of 100-mM Stock Solution of TBHP

2. Best results are obtained when imaging takes place immediately after washing and mounting the sample.

In Vitro and In Vivo Biocompatibility Testing of Functionalized Carbon Nanotubes 3.3.3.2. Make 100-mM Working Solution of TBHP

77

1. Dilute the 100-mM TBHP stock (prepared using the steps given in “Preparation of 100 mM Stock Solution of TBHP” under Subheading 3.3.3) 1:1,000 in appropriate complete growth media to produce a 100 mM working solution. 2. As example, to make 1.0 ml of 100-mM TBHP working solution, add 1.0-ml 100-mM TBHP to 1.0 ml of complete media.

3.3.3.3. Cell Treatment

1. Apply a sufficient amount of the 100-mM TBHP working solution (prepared using the steps given in “Make 100 mM Working Solution of TBHP” under Subheading 3.3.3) to the cells adhering to the coverslip(s). Incubate the coverslip(s) at 37°C and 5% CO2 for 30 min. Appropriate incubation periods for ROS production in other cell lines should be empirically determined. 2. Gently wash the coverslips twice in warm HBSS/Ca/Mg. After washing, label the cells with carboxy-H2DCFDA as described in subheading 3.3.2 (see Note 4).

3.4. Primary Mouse Neuronal Cultures

1. P0-2 mice are decapitated and brains removed in sterile conditions (all instruments must be sterilized and placed in 95% ethanol before using them). 2. Brains must be placed in dissection medium or HBSS. 3. Cortices (or other brain areas of interest) are dissected and meninges removed. Chopped small pieces of cortex are triturated with a Pasteur pipette in the presence of 0.25% trypsin and DNAse to reach a single-cell suspension. 4. After 5–10 min, a small volume of suspension at the top is placed in equal volume of DMEM supplemented with 10% horse serum. 5. Cells are then counted and directly cultured on coverslips (8 × 104 cells/coverslip) or Petri dishes (1 × 106/35 mm dish) previously coated with polylysine (100 µg/ml). Polylysine coated coverslips or dishes need to be washed twice with H2O and dried before plating. 6. After 24 h, DMEM/horse serum is replaced with Neurobasal A medium supplemented with 1% B27, 2 mM-glutamine and 40-µg/ml gentamicin (complete Neurobasal A). 7. After 7–9 days, fully differentiated neurons lay on glia cells growing in a layer underneath the neurons. 8. To increase the fraction of neurons, glia cell growth can be blocked by adding 10-µM Ara C, (cytosine arabinoside) within 24 h of cell plating.

3.5. Cytotoxicity Test (by Cytochemistry) on Primary Neurons

1. Primary mixed (neurons + glia) cultures from P2 mice have been plated at 8 × 104 cells/coverslip in complete Neurobasal A for nine divisions.

78

Ciofani et al.

2. Groups: (a) Neurobasal/B27, PBS (control); (b) Neurobasal/ B27, 0.01% PF-127/PBS; (c) Neurobasal/B27, 0.01% PF-127/PBS + MWCNTs (3.5 µg/ml final); (d) Neurobasal/ B27, 0.05% PF-127/PBS; (e) Neurobasal/B27, 0.05% PF-127/PBS + MWCNTs (17.5 µg/ml final). 3. Cells maintained at 37°C, 5% CO2, 24 h or 48 h. 4. Fixation of coverslips in 4% PF (paraformaldehyde): Add a 4% PF to culture medium (1:1 vol/vol) and leave it for 10 min, remove the 2% PF solution and replace with fresh 4% PF. Leave it for 20 min, wash 3× (5–10 min each) with PBS, and keep fixed cells in the cold room until used. 5. Staining of nuclei for apoptosis with Hoechst 33342 (1 µg/ ml) (Figs. 5 and 6): 15 min, dark, RT, wash twice 5 min with PBS. 6. Staining with Annexin V/PI (Fig. 7): Rinse 1 × 105–1 × 106 cells with 1× Binding Buffer, add to the cells 200 ml of 1× Binding Buffer, add 5 ml of Annexin V and 10 ml of propidium iodide, finally incubate at room temperature for 5–15 min in the dark. 7. Mount, seal with nail polish. 8. Microscopy: phase contrast and fluorescence cell images were acquired by a camera mounted Zeiss Axioskop microscope and by Leica TCS NT Confocal Microscope. 9. Statistics: to evaluate the statistical significance of all the described in vitro experiments, ten microscopic fields per coverslip were counted and three coverslips/treatment were used for each experiment. Three independent experiments have been performed in triplicate. One way statistical analysis of variance (ANOVA) followed by analysis Student–Newman–Keuls method has been performed (see Note 5). 10. PF-127 induces apoptosis of primary neurons; on the contrary, the presence of MWCNTs significantly avoids Pluronic-induced apoptosis of the neurons, probably by a simple mechanism of surfactant sequestration from the solution (14). 3.6. In Vivo Study (Intracerebral Injections)

1. Animals were used in accordance with protocols approved by the Italian Minister for Scientific Research. 2. Mice were anaesthetized with avertin (0.5 ml/hg) and mounted on a stereotaxic apparatus. Injections are made at specific stereotaxic locations in the visual cortex by means of a glass pipette (30-µm tip diameter) mounted on a motorized (0.1 µm step) three-axis micromanipulator connected to an injector (Sutter instruments, USA). 3. 350 nl of CNT suspension were released at 700 µm and another 350 nl were released 400 µm below cortical surface to allow homogenous dispersion of CNTs along the cortical depth.

In Vitro and In Vivo Biocompatibility Testing of Functionalized Carbon Nanotubes

79

Fig. 5. Pluronic F-127 induces apoptosis of mouse primary neurons in vitro. (a) Pictures of primary mixed cortical cultures are taken in phase contrast and fluorescence after 24 h at 37°C without or with an increasing concentration of PF-127 (from 0.001% to 0.05%). Cell nuclei have been previously stained using Hoechst 33342 (1 µg/ml). (b) and (c) show the percentage of neurons and glia respectively, after treatments. The percentage of cell numbers as compared to control cells is shown (mean ± SEM) (treatments versus control, ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001). (d) Fluorescence (nuclei) and phase contrast (whole cell body) pictures of mouse cortical neurons and glia treated for 24 h at 37°C with PF-127 (0.05% treatment as an example) are shown. In the lower panel, it is shown the difference of picnotic cell nuclei and nuclei of healthy cells. (e) Quantification and statistical analysis of apoptotic nuclei without or with increasing concentration of PF-127 (from 0.001% to 0.05%). The percentage of apoptotic nuclei as compared to control cells is shown (mean ± SEM; treatments versus control, *p < 0.05; **p < 0.01; ***p < 0.001). Reprinted from (14) with permission from Elsevier

80

Ciofani et al.

Fig. 6. Dispersed MWCNTs avoid cell death induced by PF-127. Treatments for 24 or 48 h at 37°C without (white bars) or with 0.01% of PF-127 (grey bars) have been performed in the presence of 3.5-µg/ml MWCNTs (dark stripes filled white bars). Panels (a) and (b) show the percentage of neurons and glia respectively, under the different treatments. (c) apoptotic neurons following 24 h treatments in the presence of 0.01% of PF-127 alone (grey bars) or with 3.5 mg/ml MWCNTs (dark stripes filled white bars); 0.05% of PF-127 alone (dark grey bar ) or with 17.5 µg/ml MWCNTs (dark stripes filled grey bar ); untreated cells (white bars). Percentages of cell numbers are shown compared to control as mean ± SEM (treatments versus control, *p < 0.05; **p < 0.01; ***p < 0.001) (PF-127 + CNTs vs. PF-127 treatment alone, #p < 0.01). Reprinted from (14) with permission from Elsevier

In Vitro and In Vivo Biocompatibility Testing of Functionalized Carbon Nanotubes

81

Fig. 7. Confocal microscopy. Images of primary neurons stained with Annexin V-FITC (green) and Propidium Iodide (red) after treatment with 0.05% PF-127 in the presence or absence of 17.5-µg/ml MWCNTs. Yellow arrows indicate visible aggregates of MWCNTs. Reprinted from (14) with permission from Elsevier

4. During injections, animals were oxygenated and heated by means of a thermostated blanket ensuring a 37°C rectal temperature. 5. After surgery, antibiotics (Gentamycin) were topically administered to prevent infections. In these conditions, the whole procedure should last about 20 min and recovery from anaesthesia occurs after 60–90 min. 6. After recovery, animals were returned to their home cages. Injected mice were transcardially perfused with 4% paraformaldehyde in phosphate buffer solution (0.1 M) after 3 or 18 days. 7. Brains were sectioned on a sliding microtome in 40 µm sections and Cresyl violet staining was performed. 8. Images were acquired on a camera mounted Zeiss Axioskop microscope and lesion area was measured using the Metamorph software. 9. Injected PF127–MWCNTs provoked no damage to the overall organization of the mouse brain and the organization in layers of the tissue surrounding the lesion site was unaffected (14). Close to the lesion site, an area of small injury area was present, but comparable to the lesions of the control injections (Fig. 8).

4. Notes 1. The results obtained with MTT and WST-1 assay are directly correlated to the cell number in each well. In order to have

82

Ciofani et al.

Fig. 8. Sections of mouse brain at the site of CNT injection. (a) The injection site is labelled with a star. cc cerebral cortex, wm white matter. Calibration bar is 1.2 mm. (b) The injection site in mice treated for 3 days at higher magnification. Note the black precipitate at the centre of the injection site and the area with reduced density of large cells corresponding to the lesioned area. Normal neuronal density and tissue layering is present outside the lesioned site. Lesioned area is outlined with a dashed line. Calibration bar is 400 mm. (c) Higher magnification of the transition between the lesioned and the unlesioned area (on the right). Note the clear difference in the size and the density of the cells between the unlesioned and the lesioned area. The transition between the lesioned and the unlesioned area is outlined with a dashed line. Control mice (d, e) and MWCNT-injected mice (f, g) brain cortices present a scar of glial cells surrounding the injection site. Higher magnifications of (e) the control and (g) the MWCNT-injected mice are shown. Reprinted from (14) with permission from Elsevier

good and reliable results, the same cell number in each well for all the samples has to be seeded. 2. Protect each coverslip from the light to minimize photobleaching of Hoechst stained cells. 3. Carboxy-H2DCFDA is air sensitive; activity of this reagent (in powder form) is best preserved by storage at 2–6°C with minimum exposure to air. 4. We recommend the use of neutral density filters to overcome limitations traditionally observed with reduced fluorescein dyes, which are susceptible to photooxidation as well as photobleaching. If the cells were labelled with Hoechst, we

In Vitro and In Vivo Biocompatibility Testing of Functionalized Carbon Nanotubes

83

recommend to use a Hoechst (or DAPI) filter set and neutral density filter(s) to assist with locating the cells on the coverslip, followed by a fluorescein filter set for imaging. This strategy minimizes photobleaching of fluorescein because of prolonged exposure to blue wavelengths of light. 5. In order to achieve satisfactory results in terms of statistic significance, it is important to obtain dishes with well homogeneous cultures. Several images should be taken for a correct statistical analysis.

Acknowledgments The work described in this paper was supported by the NINIVE (Non Invasive Nanotransducer for In Vivo gene therapy, STRP 033378) project, cofinanced by the 6FP of the European Commission. References 1. Lacerda L, Raffa V, Prato M, Bianco A, Kostarelos K (2007) Cell-penetrating CNTs for delivery of therapeutics. Nano Today 2:38–43 2. Smart SK, Cassady AI, Lu GQ, Martin DJ (2006) The biocompatibility of carbon nanotubes. Carbon 44:1028–1033 3. Gogotsi Y (2003) How safe are nanotubes and other nanofilaments. Mater Res Innov 7:192–1949 4. Service RF (1998) Superstrong nanotubes show they are smart, too. Science 281:940–942. 5. Hoet PHM, Brüske-Hohlfeld I, Salata OV (2004) Nanoparticles – known and unknown health risks. J Nanobiotechnology 2:12 6. Muller J, Huaux F, Lison D (2006) Respiratory toxicity of carbon nanotubes: how worried should we be? Carbon 44:1028–1033 7. Zhu L, Chang DW, Dai L, Hong Y (2007) DNA damage induced by multiwalled carbon nanotubes in mouse embryonic stem cells. Nano Lett 7:3592–3597 8. Hurt RH, Monthioux M, Kane A (2006) Toxicology of carbon nanomaterials: status, trends, and perspectives on the special issue. Carbon 44:1028–1033 9. Ciofani G, Raffa V, Pensabene V, Meciassi A, Dario P (2009) Dispersion of multi-wall carbon nanotubes in aqueous Pluronic F127 solutions

10.

11.

12.

13.

14.

for biological applications. Fullerenes Nanotubes Carbon Nanostruct 17:11–25 Yu J, Grossiord N, Koning CE, Loos J (2007) Controlling the dispersion of multi-wall carbon nanotubes in aqueous surfactant solution. Carbon 45:618–623 Woerle-Knirsch JM, Pulskamp K, Krug HF (2006) Oops they did it again! Carbon nanotubes hoax scientists in viability assays. Nano Lett 6:1261–1268 Pulskamp K, Diabatè S, Krug HF (2007) Carbon nanotubes show no sign of acute toxicity but induce intracellular reactive oxygen species in dependence on contaminants. Toxicol Lett 168:58–74 Shvedova AA, Castranova V, Kisin ER, Schwegler-Berry D, Murray AR, Gandelsman VZ, Maynard A, Baron P (2003) Exposure to carbon nanotube material: assessment of nanotube cytotoxicity using human keratinocyte cells. J Toxicol Environ Health A66: 1909–1926 Bardi G, Tognini P, Ciofani G, Raffa V, Costa M, Pizzorusso T (2009) Pluronic-coated carbon nanotubes do not induce degeneration of cortical neurons in vivo and in vitro. Nanomedicine: Nanotechnology, Medicine and Biology 5:96–104

Chapter 8 Real-Time Monitoring of Cellular Responses to Carbon Nanotubes Qingxin Mu, Shumei Zhai, and Bing Yan Abstract Dynamic cellular responses to carbon nanotubes were monitored by a real-time cell electronic sensing assay. This approach is based on the parallel impedance measurement of attached cells using electronic sensors integrated in wells of 96-well E-plate. It measures the real-time multiparameter index of cell growth named cell index (CI), which reflects the cell proliferation, morphology, attachment, and spreading. The label-free, real-time, and high-throughput assay overcomes many drawbacks in current optical based cytotoxicity assays in carbon nanotubes research, and enables dynamic monitoring of cellular responses to carbon nanotubes. Key words: Carbon nanotubes, Cell electronic sensing, Nanotoxicity, Real-time assay, Cellular dynamics

1. Introduction Carbon nanotubes (CNTs) have been widely applied as tumor imaging agents, tissue engineering scaffolds, and drug delivery carriers (1–4). In this context, the biological activities of CNTs have become crucial for both application and safety concerns. Restricted by capacity issue in animal study and the animal’s welfare concerns, in vitro cellular assays have been widely used to evaluate the potential nanomedicine applications and toxicity of nanomaterials. In response to CNTs, cells undergo time-dependent particle uptake, with subsequent morphological and pathological changes. These changes are dynamic in nature and depend on the nanomaterials, cell types, and the underlying molecular mechanisms of their interactions (5). They cannot be effectively studied by measuring cell properties at a single time point. Unfortunately,

K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_8, © Humana Press, a part of Springer Science + Business Media, LLC 2010

85

86

Mu, Zhai, and Yan

current assays for cytotoxicity of nanomaterials such as MTT, MTS, LDH, WST-1, MMP, oxidative stress assay (GSH), and flow cytometry are all single time point assays (6–12). At least, three drawbacks exist in the current portfolio of cellular assays. First, these assays provide static information at one time point only, such that important time-dependent effects may be missed. Second, CNTs interact with dyes that interfere with assays based on optical measurements. For example, CNTs have been found to interact with MTT-formazan complex causing false positive results in the MTT assay (13). Third, cellular assay results depend on the health of cells, which is sensitive to many environmental factors. None of the above assays can monitor both the cell health status and cellular effects of nanoparticle exposure in a single experiment. Therefore, nonoptical, dynamic nanocytotoxicity assays are needed for the efficient development of CNT-based biocompatible nanomaterials and nanomedicine. In this chapter, we describe a real-time cell electronic sensing (RT-CES) assay for exploring the dynamics of cellular interactions between human cells (HEK293) and carboxylated single-walled CNTs (SWCNT–COOH). This assay addresses the issues mentioned above. It is based on the parallel impedance measurement of attached cells using electronic sensors integrated in the bottom of standard 96-well microplate. The electronic impedance of sensor electrodes reflects changes in cells on the electrodes. In general, the total electrode impedance comprises three components, namely the resistance of the electrolyte solution, the impedance of the cell, and the impedance at the electrode/solution and the electrode/cell interfaces. The resistance of the electrolyte solution is usually small and can hence be ignored. The presence of the cells affects the local ionic environment at the electrode/solution interface, leading to an increase in the electrode impedance that is proportional to the number of cells. Furthermore, the impedance change depends on the extent of cell attachment onto the electrodes. For example, cell spreading increases cell/electrode contact area and further increases impedance. The cell’s biological status, including cell viability, cell number, cell morphology, and adhesion degree affect the measurement of electrode impedance (14). When CNTs influence the cell status, such as cell number, morphology, or adhesion, the Cell Index (CI) will be changed and such change is dynamically monitored. The principle of RT–CES is shown in Fig. 1. The CI is calculated using formula (1) (15).

 R (fi)  − 1 CI = max  cell i =1,...,N  R (fi)  b

(1)

Herein, Rb(fi) and Rcell(fi) are the frequency-dependent electrode resistances, N is the number of the frequency points at

Real-Time Monitoring of Cellular Responses to Carbon Nanotubes

87

Fig. 1. Schematic illustration of using RT-CES technology for study of carbon nanotubes cellular responses

which the impedance is measured. The RT–CES assay is label free, real-time, quantitative, and high throughput by measuring 576 wells simultaneously. The dynamic cell growth curves allow us to analyze time-dependent IC50s, time-dependent dose responses, slopes of cell growth curves, and area under curves, such analyses cannot be acquired through end point or single point assays. This assay has been applied in studying cell proliferation, compound-mediated cytotoxicity, cell-induced cytotoxicity, IgE receptor function, GPCR signaling, and environmental pollute particles (16–21). Its application in CNTs cytotoxicity study will promote the investigation of modified CNTs (22) in the fields of nanomedicine and nanotoxicity.

2. Materials 2.1. Preparation of Cell Culture Medium Suspension of SWCNT–COOH

1. Carboxylated single-walled CNTs (SWCNT–COOH) (Sigma Aldrich, St. Louis, MO). 2. Dulbecco’s Minimum Essential Medium (DMEM) (Gibco, Grand Island, NY). 3. Fetal bovine serum (Invitrogen, Carlsbad, CA). 4. Ultrasonic bath FS60 (Fisher Scientific, Pittsburgh, PA).

2.2. Cell Maintenance

1. HEK293 (ATCC, Rockville, MD). 2. Dulbecco’s Minimum Essential Medium (DMEM) (Gibco, Grand Island, NY). 3. Fetal bovine serum (Invitrogen, Carlsbad, CA). 4. Phosphate Buffer Saline (PBS) (Lonza, Walkersville, MD). 5. Trypsin EDTA (Mediatech, Herndon, VA). 6. L-glutamine solution 100× (Invitrogen, Carlsbad, CA).

88

Mu, Zhai, and Yan

7. Pen–Strep solution 100× (Invitrogen, Carlsbad, CA). 8. T75 cell culture flask (Corning Inc., Corning, NY). 9. Cellometer® automatic cell counter (Nexcelom Bioscience, Lawrence, MA). 10. HERAcell 150 CO2 Incubator (Thermo Scientific, Waltham, MA). 2.3. Real-Time Monitoring Dynamic Cellular Responses of SWCNT–COOH

1. Multi E-Plate (MP) System (Roche Applied Science, Indianapolis, IN). This system contains four parts: MP software, Analyzer, 96-well E-plate Station, and E-Plates. 2. Slides Warmer XH-2002 (C&A Scientific, Manassas, VA). 3. Multichannel pipette (300 mL) (Eppendorf, Westbury, NY). 4. Reagent reservoir (50 mL) (Corning Inc., Corning, NY). 5. 96-well Microplate (Corning Inc., Corning, NY).

2.4. Data Analysis

1. ACEA CellOffice® v1.0 (ACEA Biosciences, San Diego, CA). 2. SigmaPlot v10.0 (Systat Software Inc. Richmond, CA).

3. Methods 3.1 Preparation of Stock Solution of SWCNT-COOH

Solubility of CNTs is very low because of hydrophobic carbon atoms, organic solvents such as THF can dissolve CNTs at high concentration, but such solvents are harmful to cells. Considering CNTs can only exist in protein-bound form in blood and biological environment and protein helps CNTs solubilize in aqueous solutions, we use cell culture medium to suspend CNTs. 1. Prepare DMEM containing 10% FBS. 2. Weigh 2.0 mg SWCNT–COOH and add into 4 mL DMEM (10% FBS). 3. Suspend SWCNT–COOH in medium using ultrasonic bath to prepare 500 mg/mL stock solution. Sonicate the mixture for 3 min. Pause one time every 10 sec to minimize damage to medium nutrients. Gently shake the suspension during each pause. Use medium sonicated as control (see Note 1). 4. Store SWCNT–COOH medium suspension and medium control in refrigerator (4°C) for up to 1 month.

3.2. Cell Maintenance

1. Prepare complete DMEM cell culture medium. DMEM cell culture medium contains 10% FBS, 2 mM L-glutamine, 100 mg/mL penicillin and 100 U/mL streptomycin. Warm up medium to 37°C before use.

Real-Time Monitoring of Cellular Responses to Carbon Nanotubes

89

2. Maintain HEK293 cells in T75 flask and put flask into humidified incubator at 37°C (95% humidity, 5% CO2). Replace cell culture medium twice one week (see Note 2). 3. Trypsinize and subculture cells after cells are 80% confluenced. Drain all cell culture medium and wash with PBS one time. Drain PBS and add 2 mL Trypsin-EDTA, put flask into incubator. Trypsinize cells for 2 ~ 3 min. Use light microscope to observe cell morphology. When cells roundup, add 7 ~ 8 mL medium into flask to stop trypsinization. Use pipette to aspirate cells. Centrifuge cell suspension at 1,000 rpm for 3 min. Collect cell pellet and resuspend into 5 mL medium. Count cell number using Cellometer® cell counter. Split cells into new flasks (1 mL per flask) and add fresh medium. 3.3. Real-Time Monitoring Dynamic Cellular Responses of Carbon Nanotubes

1. Trypsinize and aspirate cells in T75 flask as described in Subheading 3.2.3. Centrifuge cell suspension at 1,000 rpm for 3 min. Resuspend cells in 5 mL complete DMEM cell culture medium. Count cell number using Cellometer® cell counter. Dilute cells to get the concentration 200,000 cells/mL for 5 mL. 2. Add 50 mL cell-free culture medium into 96-well E-plate using multichannel pipette. 3. Launch MP System and ACEA RT-CES® MP software and insert the 96-well E-plate into MP System. 4. Set up ExpNote. Input information of Exp Name, Exp Purpose, and Exp Procedure. 5. Design Experiment Layout. Experiment Layout includes detailed information of each well. Future data analysis is based on the information of experiment layout. Select all wells and input information as follows: Cell type: 293; Cell number: 10,000 (see Note 3). Select specific wells and input different concentrations (0, 12.5, 50, and 200 mg/ mL), triplicate each concentration, avoid using edge wells (see Note 4). 6. Scan plate and check connectivity. After experiment layout was setup, click “Scan Plate” to check the connectivity of each well. Set up test time when “Connections OK” is shown in Message page (see Note 5). 7. Setup Test Time. Test time allows setting up experiment time and measurement intervals. Intervals can be as short as 1 min. Before experiment start, set 1 min background scanning step, the electric resistance measured will be used as Rb in formula (1). Set up test time as follows (see Note 6): Step 1-1: Interval: 1 min; Sweeps: 1 Step 1-2: Interval: 20 min; Sweeps: 60

90

Mu, Zhai, and Yan

Step 2-1: Interval: 10 min; Sweeps: 24 Step 2-2: Interval: 20 min; Sweeps: 460 8. Take out 96-well E-plate from MP System. Slightly vortex cell suspension, transfer cell suspension into a reservoir and add 50 mL for each well into 96-well E-plate using multichannel pipette. Pipette cells to mix cell culture medium with cell suspension gently (see Note 7). 9. Insert the 96-well E-plate into MP System. Start the experiment. This step will run around 20 h when CI reaches to 1.0. 10. Prepare SWCNT–COOH cell culture medium suspensions in 96-well microplate. Considering adding CNTs suspensions into cells dilutes concentrations of CNTs, dilute stock solution of CNTs to two times of final concentrations, 25, 100, and 400 mg/mL, respectively. Volume is 1 mL for all concentrations. Transfer SWCNT–COOH suspensions into a normal 96-well microplate. The layout is the same as designed in Subheading 3.3.5. The volume of each well is 100 mL. Warm up the microplate to 37°C using slides warmer (see Note 8). 11. Stop step manually when CI reaches to ~1.0, the time is ~20 h. Take out 96-well E-plate and put onto slides warmer (37°C). Transfer CNTs suspensions from 96-well microplate into E-plate, mix well with gently pipetting (see Note 9). 12. Insert E-plate into instrument and start new step. 13. Take out 96-well E-plate after experiment stopped. 3.4 Data Analysis

The advantages of real-time monitoring cellular response curves include obtaining of time-dependent IC50s, time-dependent dose response curves, slopes of cell growth curves, area under curves (AUC), etc. Here, we demonstrate the plotting of cell growth curves, time-dependent IC50s, and curve slopes for examples. 1. To plot dynamic cell growth curves using MP software, select experimental wells and add into plot. The real time CI is plotted and averaged for triplicated wells. Normalize CI to 1.0 at the last time point before CNTs were added (see Note 10). The real-time normalized CI curves reflect dynamic cellular responses to SWCNT–COOH in terms of cell morphological changes, attachment, spreading and cell killing events. Plot can be displayed in CellOffice window and also can be plotted with labels or adjustable colors through exporting data into SigmaPlot. A typical growth inhibition process is shown in Fig. 2 (see Note 11). 2. Launch CellOffice® v1.0. software and load data file. Select wells to be plotted, normalize CI and set time period to be calculated (40 h–140 h), the software calculates time-dependent IC50s between 40 h and 140 h. A desensitized process happens in later phase of exposure (Fig. 3, plotted in SigmaPlot).

Real-Time Monitoring of Cellular Responses to Carbon Nanotubes

91

Fig. 2. Real-time monitoring of dynamic cellular responses of carboxylated single-walled carbon nanotubes. Light gray regions on both sides of the curves show error bars of all time points

Fig. 3. Time-dependent IC50s of carboxylated single-walled carbon nanotubes to HEK293 cells

Such process cannot be obtained through conventional endpoint assays. 3. To plot slopes of curves, select relevant wells, normalize CI, set time period from 0 to 155 h, and add into plot. A typical dose dependent slope decrease is displayed in Fig. 4 (plotted in SigmaPlot), which reflects inhibition of cell growing speed by SWCNT–COOH. The SWCNT–COOH and human embryonic kidney cells are selected as models for this approach. Since different CNTs may have different cytotoxicity profiles, and such cytotoxicity is also

92

Mu, Zhai, and Yan

Fig. 4. Slopes of cell growth curves under exposure of various doses of carboxylated single-walled carbon nanotubes

cell type specific, the dynamic cell growth curves may vary when using a different type of CNTs or cell line. Furthermore, the approach uses real-time electronic sensing to overcome current drawbacks in optical based end point cytotoxicity assays. We expect that this assay will play an important role in nanotoxicity and nanomedicine research.

4. Notes 1. To prepare SWCNT–COOH cell culture medium suspensions, ultrasonication is needed to help debundle them. Sonicate mixture with pauses is necessary to minimize damage to medium nutrient contents. It’s also necessary to use medium control which is sonicated and treated in the same way. 2. RT-CES assay can only be used for attached cells such as endothelial cell or epithelial cells. It can not monitor suspend cells such as blood cells. 3. The analyzer measures electric resistance of cells attached onto 96-well E-plate bottom. Because different cell types have different shapes and growing rate, they have different electric resistance profiles, thus the cell number should be optimized before launching an experiment. Usually, the CI reaches to 1.0 when bottoms of wells are 50–60% covered. To optimize the cell number for this experiment, select several numbers range from 1,000 to 50,000 and monitor cell growth.

Real-Time Monitoring of Cellular Responses to Carbon Nanotubes

93

Use a cell number, of which the CI reaches to 1.0 around 24 h after cell seeding. The cell number 10,000 was selected in this work. 4. Avoid using edge wells to do experiment since the cells in those wells usually grows slower than inner wells. 5. Plate connectivity checking is necessary and it must be done before experiment. This step checks the connectivity of the E-plate with analyzer. Do next step when connectivity of all wells is okay. 6. The test time may vary depending experiment. It is suggested to set up a relative longer time to monitor cell growth. Such monitoring can be manually stopped between two scheduled scannings. 7. Mix cells well when seeding cells into the wells. Cells would settle down to the bottom of the cell-suspension tubes. In order to have consistent cell number added into the E-plates, it is essential to have cells uniformly suspended. 8. HEK293 cells are sensitive to temperature change. Before adding CNTs suspensions into E-plate, put the plate onto a 37°C slides warmer. Warming up CNTs suspensions is also necessary. 9. Use a “mother plate” for SWCNT–COOH suspensions. This is to minimize the time required between E-plate operations. It ensures continuous cell growth and also minimizes environmental influences such as temperature change. 10. To plot dynamic cell growth curves or other secondary plots such as time dependent IC50s calculation, normalization of CI is necessary since cells growth has small difference in different wells and the electronic sensing is very sensitive. Normalization makes data comparable between wells. 11. SWCNT–COOH do not generate any electric resistance in wells through monitoring cell-free CNTs suspensions. Thus, the addition of SWCNT–COOH does not interfere the monitoring of CI.

References 1. Son YW, Ihm J, Cohen ML, Louie SG, Choi HJ (2005) Electrical switching in metallic carbon nanotubes. Phys Rev Lett 95:4 2. Zavaleta C, de la Zerda A, Liu Z, Keren S, Cheng Z, Schipper M, Chen X, Dai H, Gambhir SS (2008) Noninvasive Raman spectroscopy in living mice for evaluation of tumor targeting with carbon nanotubes. Nano Lett 8:2800–2805

3. Harrison BS, Atala A (2007) Carbon nanotube applications for tissue engineering. Biomaterials 28:344–353 4. Wu W, Wieckowski S, Pastorin G, Benincasa M, Klumpp C, Briand JP, Gennaro R, Prato M, Bianco A (2005) Targeted delivery of amphotericin B to cells by using functionalized carbon nanotubes. Angew Chem Int Ed 44:6358–6362

94

Mu, Zhai, and Yan

5. Ding LH, Stilwell J, Zhang TT, Elboudwarej O, Jiang HJ, Selegue JP, Cooke PA, Gray JW, Chen FQF (2005) Molecular characterization of the cytotoxic mechanism of multiwall carbon nanotubes and nano-onions on human skin fibroblast. Nano Lett 5:2448–2464 6. Mosmann T (1983) Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods 65:55–63 7. Cory AH, Owen TC, Barltrop JA, Cory JG (1991) Use of an aqueous soluble tetrazolium/formazan assay for cell growth assays in culture. Cancer Commun 3:207–212 8. Korzeniewski C, Callewaert DM (1983) An enzyme-release assay for natural cytotoxicity. J Immunol Methods 64:313–320 9. Ishiyama M, Tominaga H, Shiga M, Sasamoto K, Ohkura Y, Ueno K, Watanabe M (1995) Novel cell proliferation and cytotoxicity assays using a tetrazolium salt that produces a water-soluble formazan dye. In Vitro Toxicol 8:187–190 10. Smiley ST, Reers M, Mottola-Hartshorn C, Lin M, Chen A, Smith TW, Steele GD Jr, Chen LB (1991) Intracellular heterogeneity in mitochondrial membrane potentials revealed by a J-aggregate-forming lipophilic cation JC-1. Proc Nati Acad Sci USA 88:3671–3675 11. Guntherberg H, Rost J (1966) The true oxidized glutathione content of red blood cells obtained by new enzymic and paper chromatographic methods. Anal Biochem 15:205–210 12. Nicoletti I, Migliorati G, Pagliacci MC, Grignani F, Riccardi C (1991) A rapid and simple method for measuring thymocyte apoptosis by propidium iodide staining and flow cytometry. J Immunol Methods 139:271–279 13. Worle-Knirsch JM, Pulskamp K, Krug HF (2006) Oops they did it again! Carbon nanotubes hoax scientists in viability assays. Nano Lett 6:1261–1268

14. Giaever I, Keese CR (1993) A morphological biosensor for mammalian cells. Nature 366:591–592 15. Solly K, Wang XB, Xu X, Strulovici B, Zheng W (2004) Application of real-time cell electronic sensing (RT-CES) technology to cellbased assays. Assay Drug Dev Technol 2:363–372 16. Atienza JM, Zhu J, Wang XB, Xu X, Abassi Y (2005) Dynamic monitoring of cell adhesion and spreading on microelectronic sensor arrays. J Biomol Screen 10:795–805 17. Xing JZ, Zhu LJ, Jackson JA, Gabos S, Sun XJ, Wang XB, Xu X (2005) Dynamic monitoring of cytotoxicity on microelectronic sensors. Chem Res Toxicol 18:154–161 18. Zhu J, Wang XB, Xu X, Abassi YA (2006) Dynamic and label-free monitoring of natural killer cell cytotoxic activity using electronic cell sensor arrays. J Immunol Methods 309:25–33 19. Abassi YA, Jackson JA, Zhu J, O’Connell J, Wang XB, Xu X (2004) Label-free, real-time monitoring of IgE-mediated mast cell activation on microelectronic cell sensor arrays. J Immunol Methods 292:195–205 20. Yu N, Atienza JM, Bernard J, Blanc S, Zhu J, Wang X, Xu X, Abassi YA (2006) Real-time monitoring of morphological changes in living cells by electronic cell sensor arrays: an approach to study G protein-coupled receptors. Anal Chem 78:35–43 21. Huang L, Xie L, Boyd JM, Li XF (2008) Cellelectronic sensing of particle-induced cellular responses. Analyst 133:643–648 22. Zhou H, Mu Q, Gao N, Liu A, Xing Y, Gao S, Zhang Q, Qu G, Chen Y, Liu G, Zhang B, Yan B (2008) A nano-combinatorial library strategy for the discovery of nanotubes with reduced protein-binding, cytotoxicity, and immune response. Nano Lett 8:859–865

Chapter 9 Reducing Nanotube Cytotoxicity Using a Nano-Combinatorial Library Approach Qiu Zhang, Hongyu Zhou, and Bing Yan Abstract Carbon nanotubes (CNTs) have a great potential for applications in medicine. However, their biocompatibility and toxicity cause a great concern. Due to the large surface area of CNTs, chemical modification can dramatically alter their physiochemical properties and hence biological activity. Using a combinatorial chemistry approach, we report the synthesis of an 80-member surface-modified nanotube library. Based upon screening of this library with respect to protein-binding capacity, cytotoxicity, and immune response, we were able to select highly biocompatible nanotubes with reduced protein-binding cytotoxicity and immune response. Key words: Nanotube, Nano-combinatorial library, Cytotoxicity, Immune response, Protein-binding

1. Introduction Carbon nanotubes (CNTs) have become a focus of research because of their potential applications as intracellular probes, drug carriers, imaging agents, and DNA modulators (1–8). For these aims, they need to fulfill the requirements of biocompatibility, easy blood circulation, low clearance rate, low acute and chronic cytotoxicity, and low immune responses. Owing to the large surface area of CNTs, chemical surface modifications strongly affect their physiochemical properties. Chemical functionalization is a suitable means to confer a wide variety of interesting properties such as stealth characteristics, reduced toxicity, and site-specific delivery onto CNTs (9). Combinatorial chemistry has been successfully used in drug discovery by creating diverse drug-like small molecules. A combinatorial approach to modify the surface of CNTs

K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_9, © Humana Press, a part of Springer Science + Business Media, LLC 2010

95

96

Zhang, Zhou, and Yan

enables further mapping of the chemical parameter space. In general, CNTs have a strong tendency to bind proteins (7, 9–13). Protein binding increases a nanoparticle’s propensity to be opsonized (14) for clearance or to create cryptic epitopes in cellular signaling proteins causing toxic responses. CNTs can also interact with immune cells and immunoproteins in blood and tissues, thus stimulating immune responses. The following protocol describes an effective approach to rapidly synthesize a nanotube library and identify nanotube candidates that possess low protein binding, reduced cytotoxicity, and reduced immune responses (15). This is demonstrated by a novel surface-modified MWNT library containing 80 members, prepared using in silico design and combinatorial synthesis. Multiple biological screenings including protein binding, cytotoxicity, and immune responses yielded useful structure–activity relationships.

2. Materials 2.1. Equipment 2.1.1. Synthesis, Purification and Characterization of f-MWNT Library

1. Nicolet 380 FTIR spectrophotometer (Madison, WI). 2. VarioEL III analyzer (Elementar Analysensysteme GmbH, Germany). 3. Varian 500 spectrometer equipped with a 4-mm gHX Nanoprobe (Variannmr Inc, Palo Alto, CA). 4. Accelrys and SciTegic software packages (Accelrys, San Diego, CA). 5. Zhicheng 63 positions heating parallel synthesizer (Shanghai, China). 6. JEOL 1200 EX Transmission Electron Microscope (JEOL, Tokyo, Japan). 7. UV–Vis spectrometer (Shimadzu, Japan). 8. LC/MS (Shimadzu, Japan). 9. HPLC (SPD–M20A) (Shimadzu, Japan).

2.1.2. Protein Binding Assays

1. Hitachi F-4500 spectrofluorometer (Hitachi Co. Ltd., Tokyo, Japan). 2. Model 4700 Proteomics Analyzer (Applied Biosystems, Foster City, CA).

2.1.3. Cytotoxicity Assay and Quantification of NO Release

1. HF90 CO2 cell incubator (Heal Force, Hong Kong). 2. Heal Force safe-1200 (Heal Force, Hong Kong). 3. Inverted fluorescence microscope IX71 (Olympus, Japan). 4. Inverted microscope CKX31 (Olympus, Japan).

Reducing Nanotube Cytotoxicity Using a Nano-Combinatorial Library Approach

97

5. Biological microscope CX21 (Olympus, Japan). 6. Microplate Reader (Bio-Rad, America). 7. Refrigerator (Thermo electron, America). 8. MiniSpin centrifuge (Eppendorf, Germany). 9. Biofuge stratos centrifuge (Heraeus Ltd., Germany). 2.2. Chemicals 2.2.1. Synthesis, Purification and Characterization of f-MWNT Library 2.2.2. Protein Binding Assays

1. Pristine multiwalled carbon nanotubes (p-MWNTs, grown by chemical vapor deposition with average diameter of 40 nm, Chengdu Carbon Nanomaterials R&D Center, China). 2. Tyrosine (GL Biochem Ltd., China). 3. All other chemicals (Acros Organics, Belgium). 1. BSA, carbonic anhydrase (CA), chymotrypsin, and hemoglobin (Worthington Biochemical Corporation, USA). 2. Minimum essential medium eagle (GIBCO, USA). 3. Horse serum (Invitrogen, Carlsbad, CA): inactivated at 56°C for 30 min before use. 4. Human plasma (Innovative Research Inc., Southfield, MI). 5. PBS: 0.01 M, pH 7.2

2.2.3. Cytotoxicity Assay and Quantification of NO Release

1. RPMI 1640 (GIBCO, USA). 2. Fetal bovine serum (FBS, Hangzhou Sijiqing Biological Engineering Materials Co., Ltd., China): inactivated at 56°C for 30 min before it was used in cell culture. 3. Penicillin stock solution: 100 mg/mL in saline. Store at −20°C. 4. Streptomycin stock solution: 10,000 U/mL in saline. Store at −20°C. 5. LPS (Lipopolysaccharide, Sigma–Aldrich, St.) stock solution: 10 µg/mL in RMPI1640. Store at −20°C. 6. PMA (Phorbol 12-myristate 13-acetate, Promega) stock solution: 5 µg/mL in RMPI1640. Store at −20°C. 7. WST-1 Cell Proliferation and Cytotoxicity Assay Kit (Beyotime, China): Store at −20°C. 8. Nitric oxide (NO) Assay Kit (Beyotime, China): Store at −20°C. 9. Cell culture 96-well plate (Costar, America).

2.3. Cell Lines

1. THP-1: human acute monocytic leukemia cell line (American Type Culture Collection (ATCC), USA). THP-1 was cultivated in RPMI 1640 with 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine, 100 µg/mL penicillin and 100 U/ mL streptomycin, and grown in a humidified incubator at 37°C (95% room air, 5% CO2).

98

Zhang, Zhou, and Yan

3. Methods To identify biocompatible nanotubes with reduced protein-binding, cytotoxicity, and immune response, we first carried out molecular diversity design, combinatorial library synthesis, and characterization (LC/MS, FTIR, MAS 1H NMR, TEM). We then investigated the protein binding with a representative subset of proteins, the acute cytotoxicity of the functionalized MWNT (f-MWNT) library in macrophages by the WST-1 assay (16), and NO production after incubation of the f-MWNTs with THP-1 derived macrophage (17, 18). Based on the above results, the sums of each f-MWNT’s rankings (designated its multiassay score) were compared. Suitable candidates with smaller scores, indicative of lower binding, higher cell viability, and lower NO generation, were selected for further optimization. 3.1. Synthesis, Purification, and Characterization of f-MWNT Library 3.1.1. Selecting Building Blocks for the f-MWNT Library

3.1.2. Preparation of the f-MWNT Library 3.1.2.1. Preparation of MWNT–COOH

1. Based on commercial availability, highly diverse building blocks containing 31 amines and 26 acylators were first selected to build a virtual library. 2. Estimated molecular properties such as structure similarities, hydrophobicity, solubility, topological, and stereochemical properties were calculated by Accord and Pipeline Pilot from Accelrys (San Diego, CA) (see Note 1). 3. Eight amines and nine acylators (Figs. 1 and 2) were selected for library synthesis containing 80 f-MWNTs (8 × 10; eight amines, nine acylators, and one without acylation) based on the most diverse molecular and physicochemical properties. 1. The p-MWNT (10.0 g), suspended in ca. 150 mL H2O2/ H2SO4 (v/v = 1:3), was placed in an ultrasonic bath at 60°C for 2 h. 2. The mixture was poured into 500 mL deionized water and centrifugated at 3,220 × g for 30 min. The pallet was washed five times by deionized water and then dried at 65°C under vacuum overnight. 3. The solid was again added to 150 mL HNO3/H2SO4 (v/v = 1:3) slowly with stirring in an ice-bath. The temperature rose slowly to 60°C and the mixture was sonicated for 3 h (see Note 2). 4. The mixture was poured into 500 mL deionized water and centrifugated at 3,220 × g for 30 min, washed five times by deionized water until the pH of the filtrate was ~7.0. The pellet was dried at 65°C under vacuum for 24 h.

Reducing Nanotube Cytotoxicity Using a Nano-Combinatorial Library Approach

99

Fig. 1. Synthesis route of the f-MWNT library. Reprinted from (15). Copyright 2008 American Chemical Society

Fig. 2. Surface molecular compositions of combinatorial MWNT library members. Reprinted from (15). Copyright 2008 American Chemical Society

100

Zhang, Zhou, and Yan

3.1.2.2. Preparation of MWNT–Fmoc–Tyr–COOH

1. MWNT–COOH (~7.0 g) was suspended in 120 mL SOCl2 with catalytical amount of DMF. The mixture was heated to reflux for 18 h under nitrogen atmosphere. 2. After removal of the solvent by rotavapor under vacuum, the mixture was suspended in anhydrous THF and evaporated under vacuum again to remove trace amount of SOCl2. The purified f-MWNT were subsequently dried at 65°C under vacuum overnight. 3. The acid chloride functionalized MWNT (~7.0 g) was suspended in 70 mL anhydrous DMAC. Fmoc protected tyrosine solid (7.0 g) and 5 mL pyridine was added to the suspension. The suspension was stirred at 50°C for 36 h under nitrogen atmosphere. 4. The mixture was centrifuged at 3,220 × g for 2 h (see Note 3) and washed with ethanol and water alternatively to remove the unreacted tyrosine and byproducts. The resulting product solid was dried at 65°C in a vacuum oven overnight.

3.1.2.3. Preparation of MWNT-Fmoc-Amines

1. MWNT–Fmoc–Tyr–COOH (~6.0 g) was suspended in 100 mL SOCl2 with 2.5 mL of DMF as catalyst. The mixture was heated to reflux for 18 h under nitrogen atmosphere. 2. SOCl2 was removed by centrifugation and the solid washed five times by anhydrous THF. The resulting product was dried at 65°C under vacuum for 24 h. 3. The solid was dispersed in 50 mL of anhydrous DMAC and sonicated for 30 min at room temperature to form a homogeneous suspension. The mixture was divided into eight parts. To each part, 1.0 mL of pyridine and 2.5 g of the corresponding amine was added. The reaction was shaken at 50°C for 24 h using the parallel synthesizer. 4. The mixture was centrifuged at 3,220 × g for 2 h, and alternatively washed with ethanol and deionized water several times to remove unreacted amine and byproducts. The resulting product was dried at 65°C under vacuum for 24 h.

3.1.2.4. Parallel Preparation of Final MWNT Library

1. MWNT–Fmoc–Tyr–amine (~1.00 g) was suspended in 30 mL piperidine/DMF (v/v, 1:4). The mixture was shaken on an orbital shake at 50°C for 2 h to clave the Fmoc protecting group. 2. The mixture was centrifuged at 3,220 × g for 30 min and washed with alternative ethanol and deionized water for five times. The resulting solid was dried at 60°C in a vacuum oven overnight. 3. Each of MWNT–Tyr–amine was suspended in 50 mL of anhydrous THF and sonicated for 15 min at 25°C to form a

Reducing Nanotube Cytotoxicity Using a Nano-Combinatorial Library Approach

101

homogenous suspension. Each suspension was divided into ten parts. To each part, 500 µL of related acyl chloride or sulfonyl chloride and 300 µL of pyridine were added. 4. The reaction mixtures were shaken at 60°C for 24 h on the parallel synthesizer before ethanol was added to each sample. 5. The mixture was washed thoroughly with ethanol and water and purified by wash/centrifugation cycles five times. The resulting product was dried at 65°C under vacuum for 24 h. 3.1.3. Characterization of the f-MWNT Library 3.1.3.1. Characterization of the Morphological Integrity by TEM

To confirm the morphological integrity of f-MWNTs after multiple reaction steps, selected intermediates and the final f-MWNTs were characterized by TEM. The results indicated that there was no difference among the starting material, the intermediates and the final products. So the size and shape of f-MWNTs was not affected by the multiple steps combinatorial modification.

3.1.3.2. Characterization of the Purity by LC/MS Analysis

To confirm there were no small molecules or catalysts left on the surface of the f-MWNTs through noncovalent absorption, LC/MS method was developed for monitoring the effectiveness of each purification process. After the reaction mixture was centrifuged at 3,220 × g for 2 h, the supernatant was checked by LC/MS. Water and methanol were used alternatively to remove the salt and the organic residue.

3.1.3.3. Structural Identity of the Surface Modification by FTIR

FTIR has been used widely to characterize the structure of compounds on the resin and to study the kinetic properties of solid phase reaction. Here, it was adapted for characterization of special functional groups on the surface of MWNTs, such as carbonyl group. 1. When p-MWNT was oxidized, an IR band at 1,713 cm−1 appeared indicating the formation of carboxylic acid groups. 2. MWNT–COOH reacted with Fmoc-Tyr and the IR bands of phenol ester carbonyl and carboxylic acid of tyrosine overlapped at ~1,710 cm−1. 3. The amidation of MWNT–Fmoc–Tyr–COOH was characterized by disappearance of an acid carbonyl band at 1,710 cm−1 and the formation of methyl and methane bands at 2,843 and 2,924 cm−1. 4. Final product synthesis induced no further IR changes. But in some case, an ester carbonyl band at 1,787 cm−1 was formed because of an ester group on its building block.

3.1.3.4. Structural Identity of the Surface Modification by MAS 1H NMR

The broad peaks of the solution NMR of nanoparticles, caused by the bad mobility of the individual molecules on nanoparticles and the magnetic susceptibility differences, showed no detectable signals. More structural information could be obtained when high

102

Zhang, Zhou, and Yan

resolution MAS 1H NMR was used for CNTs. A 2.0-mg sample was transferred into a Nano NMR tube and 40 µL DMSO-d6 was then added. The NMR tube was sonicated for 30 min before data acquirement. The spectra were acquired with spectral widths of 4,000–5,000 Hz and 32,000 data points at 25°C. The spinning rate was 2,000 Hz (see Note 4). 1H NMR spectra were recorded using a rotor-synchronized Carr–Purcell–Meibom–Gill (CPMG) pulse sequence with water suppression. 1. The MAS 1H NMR spectrum of MWNT–COOH showed signals of NMR solvent DMSO-d6 at 2.5 ppm, water at 3.33 ppm and a number of small NMR signals, possibly from hydroxyl groups and some defect sites. 2. Through analysis of chemical shifts and signal ratios, we positively confirmed molecules bound to nanotubes combining with the reaction progression monitoring by FTIR. 3.1.3.5. Determination of Surface Coverage by UV–Vis Spectroscopy

The surface coverage was firstly quantified by the released Fmoc groups. Using MWNT-Fmoc-Tyr-COOH as an example (see Note 5), the loading of the carboxyl groups was 0.41 ± 0.03 mmol/g. 1. 20.0 mg MWNT–Fmoc–Tyr–COOH was suspended in 4.0 mL piperidine/DMF (v/v, 1:4) and stirred at 50°C for 2 h. 2. The mixture was filtered through 0.22 µm membrane (pore size, Teflon). The solid was washed and the filtrate was then adjusted to 10 mL by DMF. 3. The UV–Vis spectrum was recorded for the filtrate. The loading of the carboxyl group was calculated by the absorption at 300 nm (see Note 6).

3.1.4. Determination of Surface Coverage by Elemental Analysis

The surface coverage for the intermediates and final products were also quantified by elemental analysis for their nitrogen content. For MWNT–Fmoc–Tyr–COOH, the loading of the carboxyl group was 0.45 ± 0.01 mmol/g, consistent with the UV–Vis Spectroscopy method.

3.2. Protein Binding Assays

f-MWNTs were solved in deionized water at 500 mg/mL. Soni cation after immersing f-MWNTs in water for 48 h was necessary to ensure the homogenous solution of f-CNTs. Suspensions were resonicated before use.

3.2.1. Single Protein Binding 3.2.1.1. Preparation of MWNTs Stock Solution 3.2.1.2. Preparation of Protein Solution

All proteins (BSA, carbonic anhydrase, chymotrypsin, and hemoglobin) were solved in PBS (pH 7.2, 0.01 M) at a final concentration of 50 mg/mL.

Reducing Nanotube Cytotoxicity Using a Nano-Combinatorial Library Approach 3.2.1.3. Steady-State Fluorescence Spectra

103

Steady-state fluorescence spectra were acquired on a Hitachi F-4500 spectrofluorometer at 22°C and this assay was also amenable to 96-well microplate. The protein solution (50 mg/mL) was titrated with f-MWNTs at three different concentrations (0.0, 7.5, and 15 mg/mL). The excitation wavelength was set at 280 nm. Emission fluorescence spectra were recorded between 300 and 400 nm (see Note 7). The scanning speed was 1,200 nm/min and both the excitation and emission slits were set at 5.0 nm. PMT voltage was set to 700 V. Fluorescence intensities at 340 nm were used for calculation in Stern–Volmer equation: F0 / F = 1 + Ksv[Q] where F0 and F represent initial and modified fluorescence intensity, respectively. Ksv was the Stern–Volmer constant and [Q] was the quencher concentration (see Note 8).

3.2.2. Total Protein Binding

f-MWNTs were suspended in cell culture medium with 10% heat-inactivated horse serum or human plasma to form 500 mg/ mL suspensions after 3 min sonication. Photographs were taken 24 h afterwards.

3.3. Cytotoxicity Assay and Quantification of NO Release

Pristine MWNT and other f-MWNTs were sterilized at 121°C for 3 h and then were suspended in RMPI1640 with 10% heat-inactivated fetal bovine serum to get the mother liquor concentration of 1 mg/mL. After careful sonication and 48 h soak at room temperature, the stock solution was stored at 4°C. Sonication was needed before use (see Note 9).

3.3.1. Preparation of MWNT Suspension 3.3.2. Differentiation of THP-1 into Macrophage

3.3.3. Cytotoxicity Assay

1 × 106/well THP-1 cells in logarithmic growth phase in 100 µL cell culture were seeded into 96-well plates and PMA was added into each well at final concentration of 50 ng/mL. After 48 h incubation, the nonadherent THP-1 cells were carefully removed and the adherent macrophages were washed twice with RPMI 1640 (see Note 10) for later cytotoxicity assay and NO release assay. 1. Cells were incubated with MWNT suspensions (pristine MWNT and 83 kinds of f-MWNTs) at the concentration of 50 and 200 µg/mL respectively. 2. After 24 h incubation, cell proliferation (WST-1) assay was used to determine the cell viability according to the instruction. The supernatant medium was replaced by 100 µL/well WST-1 diluted 1:10 (v/v) with complete culture medium under dark conditions. In parallel, the blank control was set in triplicate (100 µL/well WST-1 diluted 1:10 (v/v) with complete culture medium). 3. After 4 h incubation, the supernatants without particles were transferred into another 96-well plate (see Note 11) and measured at 450 nm.

104

Zhang, Zhou, and Yan

3.3.4. Quantification of NO Release

NO production of the macrophages was detected by analyzing nitrite in the culture supernatants by a colorimetric method using the Griess reaction. 1. Cells were incubated with MWNT suspensions (pristine MWNT and 83 kinds of f-MWNTs) at the concentration of 50 and 200 µg/mL respectively. Lipopolysaccharide was added to the cultures at final concentration of 100 ng/mL. At the same time, control was set in triplicate (cells were only incubated with 100 ng/mL LPS) (see Note 12). 2. After 24 h incubation, the supernatant medium from each well was collected into 1.5 mL Eppendorf tubes respectively. 3. To remove the remaining MWNTs in supernatants, Eppendorf tubes were centrifuged at 9,000 ´ g, 4°C, for 10 min. 4. 50 µL cell supernatant from each tube was transferred to another 96-well plate. At the same time, standard samples were diluted into a series of concentration (0, 1, 2, 5, 10, 20, 40, 60, 100 µmol/L) in the same 96-well plate. 50 µL Griess Reagent I and 50 µL Griess Reagent II were then subsequently/in turn added into cell supernatants and diluted standard samples on the plate. 5. 96-well plate was shaked for 1 min on a rocking platform to mix thoroughly and then detected with microreader at 540 nm. Nitrite concentrations were calculated from a sodium nitrite standard curve using a linear curve fit.

3.4. Multiassay Score of the f-MWNT Library

To select suitable candidates for further lead optimization, we analyzed results from all assays (Fig. 3). The results from four protein binding assays, nitric oxide generation, and the cytotoxicity assay were each ranked from 1 to 80 for all f-MWNTs (smaller number means lower binding, higher cell viability and lower NO generation). The sum of an f-MWNT’s rankings was designated its multiassay score. A smaller multiassay score was deemed superior in terms of overall biocompatibility. The f-MWNTs #40, #57, #49, and #65 had scores less than 100 in a possible score range of 6–480, showing that they were suitable candidates for further optimization. A remarkable structure–activity relationship was also revealed in this study. Three out of the top five lead candidates contained the building block AC005, and all eight AC005-containing f-MWNTs in the library were in the top 26 multiassay scores. Examining the least biocompatible (or higher score) f-MWNTs, AC001, AC003, AM005, AM007, and AM008 each generated 4–5 f-MWNTs that have multiassay scores at the bottom 30, suggesting that these building blocks generated surface molecules that made less biocompatible nanotubes.

Reducing Nanotube Cytotoxicity Using a Nano-Combinatorial Library Approach

105

Fig. 3. Multiassay score of the f-MWNT library. Findings from four protein binding assays, cytotoxicity, and immune response assays (at 50 mg/mL) were ranked for all library members. The sum of their ranks was designated the multiassay score and is shown as vertical bars in the graph. Reprinted from (15). Copyright 2008 American Chemical Society

In summary, this approach helped identify novel f-MWNTs with reduced protein binding, cytotoxicity, and immune responses. At the same time, structure–activity and structure–toxicity relationship were also obtained. AC005 was required for generating surface molecules that confer better biocompatibility in nanotubes.

4. Notes 1. In order to obtain f-MWNTs with reduced toxicity and explore the structure–activity relationship without a priori knowledge, we decided to modify the MWNTs through combinatorial chemistry approach with the maximum surface structural diversity. 2. The preparation of MWNT–COOH was performed in HNO3/H2SO4 followed by H2O2/H2SO4. 3. Sometimes, the f-MWNTs cannot be collected by centrifugation at 3,220 × g. In that case, we either use ultra-speed centrifuge for higher centrifugal force or add same volume of water into the mixture to reduce the dispersability of f-MWNTs. 4. Unlike the solution NMR, the spinning sideband which is spinning speed dependent may affect the assignment of the

106

Zhang, Zhou, and Yan

MAS NMR signals. However, very high spinning speed is detrimental for MAS NMR since the huge centrifugal force can firmly pack the originally well suspended nanoparticles against the rotor wall. It is important to select a proper spinning speed for HAS NMR. 5. Since there is no method to determine the loading of carboxyl group for MWNT–COOH directly, an alternative method is developed. MWNT–COOH is converted to MWNT–Fmoc– Tyr–COOH. The loading of the latter is determined by both UV–Vis spectroscopy and elemental analysis. However, as the conversion efficiency is less than 100%, the value represents a lower limit of the loading. 6. Fmoc group has three diagnostic UV peaks at 278, 289, and 300 nm. Since f-MWNTs have UV absorbance at 240 nm, we use the absorbance at 300 nm for quantification. 7. Experiments proved that the influence of carbon nanotubes light scattering is insignificant (1/200 intensity of protein fluorescence) at the experimental concentrations. 8. Because the molecular weight of multiwalled carbon nanotubes is not known, the mass concentration was used in the Stern– Volmer plots. 9. To ensure the homogenous solution of f-MWNTs, the materials were soaked in culture medium for 48 h before sonication. 10. Differentiated cells are characterized by adhering to the plastic surface. Care must be taken to avoid damaging and losing cells. In addition, RPMI 1640 without 10% heat-inactivated fetal bovine serum was used to avoid the production of bubble. 11. Since CNT has absorption in UV region, it is necessary to transfer the supernatants to new microplate for measurement. 12. LPS were used here as a positive control. It is well known that LPS could activate the macrophage and stimulate the production of many inflammation factors including NO. It was also possible to visualize any further cell reactivity following interaction with f-MWNTs. References 1. Mattson MP, Haddon RC, Rao AM (2000) Molecular functionalization of carbon nanotubes and use as substrates for neuronal growth. J Mol Neurosci 14:175–182 2. Liu Z, Cai WB, He LN, Nakayama N, Chen K, Sun XM et al (2007) In vivo biodistribution and highly efficient tumour targeting of carbon nanotubes in mice. Nat Nanotechnol 2:47–52

3. Singh R, Pantarotto D, Lacerda L, Pastorin G, Klumpp C, Prato M et al (2006) Tissue biodistribution and blood clearance rates of intravenously administered carbon nanotube radiotracers. Proc Natl Acad Sci U S A 103: 3357–3362 4. Wang HF, Wang J, Deng XY, Sun HF, Shi Z J , G u Z N e t a l ( 2 0 0 4 ) B i o d i s t r i b ut ion of carbon single-wall carbon

Reducing Nanotube Cytotoxicity Using a Nano-Combinatorial Library Approach nanotubes in mice. J Nanosci Nanotechnol 4:1019–1024 5. Cherukuri P, Bachilo SM, Litovsky SH, Weisman RB (2004) Near-infrared fluorescence microscopy of single-walled carbon nanotubes in phagocytic cells. J Am Chem Soc 126:15638–15639 6. Kam NWS, Dai HJ (2005) Carbon nanotubes as intracellular protein transporters: generality and biological functionality. J Am Chem Soc 127:6021–6026 7. Karajanagi SS, Vertegel AA, Kane RS, Dordick JS (2004) Structure and function of enzymes adsorbed onto single-walled carbon nanotubes. Langmuir 20:11594–11599 8. Zheng M, Jagota A, Semke ED, Diner BA, McLean RS, Lustig SR et al (2003) DNAassisted dispersion and separation of carbon nanotubes. Nat Mater 2:338–342 9. Nahar M, Dutta T, Murugesan S, Asthana A, Mishra D, Rajkumar V et al (2006) Functional polymeric nanoparticles: an efficient and promising tool for active delivery of bioactives. Crit Rev Ther Drug Carrier Syst 23:259–318 10. Colvin VL, Kulinowski KM (2007) Nanopar ticles as catalysts for protein fibrillation. Proc Natl Acad Sci U S A 104:8679–8680 11. Cedervall T, Lynch I, Lindman S, Berggard T, Thulin E, Nilsson H et al (2007) Understanding the nanoparticle-protein corona using methods to quantify exchange rates and affinities of

107

proteins for nanoparticles. Proc Natl Acad Sci U S A 104:2050–2055 12. Vertegel AA, Siegel RW, Dordick JS (2004) Silica nanoparticle size influences the structure and enzymatic activity of adsorbed lysozyme. Langmuir 20:6800–6807 13. Lynch I, Dawson KA, Linse S (2006) Detecting cryptic epitopes created by nanoparticles. Sci STKE 2006:p14 14. Owens DE, Peppas NA (2006) Opsonization, biodistribution, and pharmacokinetics of polymeric nanoparticles. Int J Pharm 307:93–102 15. Zhou H, Mu Q, Gao N, Liu A, Xing Y, Gao S et al (2008) A nano-combinatorial library strategy for the discovery of nanotubes with reduced protein-binding, cytotoxicity, and immune response. Nano Lett 8:859–865 16. Worle-Knirsch JM, Pulskamp K, Krug HF (2006) Oops they did it again! Carbon nanotubes hoax scientists in viability assays. Nano Lett 6:1261–1268 17. Pulskamp K, Diabate S, Krug HF (2007) Carbon nanotubes show no sign of acute toxicity but induce intracellular reactive oxygen species in dependence on contaminants. Toxicol Lett 168:58–74 18. Green LC, Wagner DA, Glogowski J, Skipper PL, Wishnok JS, Tannenbaum SR (1982) Analysis of nitrate, nitrite, and [15 N] nitrate in biological fluids. Anal Biochem 126:131–138

Chapter 10 DNA Damage by Carbon Nanotubes Using the Single Cell Gel Electrophoresis Technique Olga Zeni and Maria Rosaria Scarfì Abstract The single-cell gel electrophoresis (SCGE) or comet assay is a simple and sensitive method for quantitatively measuring DNA breakage and repair in individual cells. It can be applied to proliferating and nonproliferating cells and cells of those tissues, which are the first contact sites for mutagenic/carcinogenic substances. In this technique, cells are embedded in agarose, lysed, subjected to electrophoresis, and stained with a fluorescent DNA-binding dye. Cells with increased DNA damage display increased DNA migration from the nucleus toward the anode, which resembles the shape of a comet. The migration is observed by fluorescence microscopy after staining with a fluorescent DNA-binding dye, and the intensity of the comet tail reflects the number of DNA breaks. The assay is performed in almost all eukaryotic cells and has applications in many fields, including genetic toxicology, biomonitoring, ecotoxicology, medical, and nutritional research. The assay is a very sensitive tool to investigate the effect of carbon nanotubes on DNA of human cells in vitro. This chapter describes a procedure to perform the comet assay, in its alkaline version, on cell cultures treated with carbon nanotubes. Key words: Comet assay, Single cell analysis, DNA damage, Human biomonitoring, Occupational exposure, Carbon nanotubes

1. Introduction Carbon nanotubes (CNTs) are one of the new materials of emerging technologies, and are becoming increasingly studied for possible applications in electronics, optics, and biology. Nevertheless, there is a paucity of information on their toxicological properties and on potential human health risk. Among the cellular targets potentially affected by CNTs, DNA molecule deserves particular attention since its integrity plays a key role in the toxicological evaluation. As a matter of fact, alterations in the normal DNA structure have many direct and indirect effects on cells and organisms K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_10, © Humana Press, a part of Springer Science + Business Media, LLC 2010

109

110

Zeni and Scarfì

(mutations, genetic recombination, the inhibition or alteration of cellular processes, chromosomal aberration, tumorigenesis, and cell death). Moreover, DNA damage is always present in the first step of cancerogenesis. A useful technique for analysing and quantifying DNA damage and repair in individual cells is the single cell gel electrophoresis (SCGE) or comet assay. It was first introduced by Ostling and Johanson in 1984 (1) as an assay performed under neutral conditions, while the more versatile alkaline method was developed in 1988 by Singh and co-workers (2). The assay measures DNA strand breaks. In its alkaline version (pH > 13), a broad spectrum of DNA lesions, such as single strand breaks, double strand breaks, alkali-labile sites, and DNA–DNA and DNA-proteins cross-links, can be detected. Moreover, transient DNA strand breaks arising from DNA repair processes can be measured (2). The strength of the assay is in its high sensitivity to detect DNA lesions that, in most cases, could be repaired before being fixed as mutations. In this respect, the assay is a marker of effect rather than a marker of genotoxicity. It combines the simplicity of biochemical techniques for detecting DNA damage with the single cell approach typical of cytogenetic assays. Several reviews have been published in recent years on the potentialities of the assay in genotoxicological, ecotoxicological, and biomonitoring studies (3–6). As the comet assay measures DNA fragmentation in single cells, cytotoxicity-induced DNA double strand breaks may confound the interpretation of data from this test. In this respect, it is recommended that viability measurements should always be made in parallel. However, the determination of cell survival, whenever possible, seems to be useful with respect to the biological significance of test results since DNA damage is only relevant if it occurs in cells capable of surviving the damage (7). For the assay, cells are embedded in agarose on a microscope slide, lysed with detergent, subjected to electrophoresis, stained with a fluorescent DNA-binding dye, and subsequently observed by fluorescence microscopy. Negatively charged loops/fragments of DNA migrate out of the nuclei forming a tail in the direction of the anode, giving the nuclei the appearance of a comet (DNA migration pattern). DNA migration is determined on the bases of the shape and intensity of comet tail relative to the head, and can be quantified by visual or computerized image analysis, by measuring several parameters of the comets, including the percentage of migrated DNA, tail length and tail moment. Apart from the high sensitivity for detecting DNA damage, the comet assay presents several advantages, including low cost, relatively simple laboratory procedure, individual cell data collection (allowing for more robust statistics), and a small amount of cells per sample needed (<10,000). In addition, the analysis can

DNA Damage by Carbon Nanotubes Using the Single Cell Gel Electrophoresis Technique

111

be carried out in any organism, tissue or culture from which single cells can be isolated. This technique allows evaluating DNA damage of virtually any physical, chemical or industrial compound. It can be applied to proliferating and non-proliferating cells and to cells of those tissues, which are the first sites of contact with mutagenic/carcinogenic substances. In this chapter, a laboratory protocol is provided to apply the alkaline comet assay on cell cultures treated with CNTs.

2. Materials 2.1. Cell Culture Preparation

1. Phosphate-buffered saline (PBS) Ca2+, Mg2+ free: Add 8 g of NaCl, 1.15 g of Na2HPO4 and 0.2 g of KCl to 900 ml water (see Note 1); adjust pH to 7.4 with HCl. Add water to achieve 1 L. Sterilize by autoclaving (121°C for 30 min) and store refrigerate (4°C). 2. Cell culture medium (see Note 2). 3. Complete cell culture medium: medium supplemented with 15% foetal bovine serum (FBS) and 2 mM l-glutamine. 4. Burker hemocytometer with cover slip (24 × 24 mm). 5. Inverted microscope. 6. Autoclave.

2.2. Carbon Nanotubes Treatment

1. CNTs purified (see Note 3). 2. Autoclave. 3. Ultrasound bath. 4. Cell culture medium.

2.3. Cell Viability

1. Burker hemocytometer with cover slip (24 × 24 mm). 2. Trypan blue stain 0.5% (Euroclone). 3. Inverted microscope.

2.4. Alkaline Comet Assay

1. Microscope slides (frosted end). 2. Microscope cover slips (24 × 60 mm). 3. Normal melting point agarose (NMA): 1% in sterile PBS (1 g/100 ml PBS) freshly prepared (see Note 4). 4. Low melting point agarose (LMA): 0.5% in sterile PBS (0.1 g/20 ml PBS). 5. Thermostatic bath (37°C). 6. Heating magnetic stirrer. 7. Minicentrifuge.

112

Zeni and Scarfì

8. Horizontal gel electrophoresis apparatus. 9. Electrophoresis power supply. 10. Fluorescence microscope equipped with an excitation filter of 515–560 nm and a 40x objective. 11. Phosphate-buffer saline (PBS) Ca2+, Mg2+ free: (see Subheading 2.1). 12. Lysis solution: 0.1 M Na2EDTA, 2.5 M NaCl, 100 mM Tris. Adjust pH to 10 with concentrated NaOH solution (10 N). Store at room temperature. Working solution (prepared before use; reagents for 200 ml): add 2 ml of Triton X-100 and 20 ml of DMSO to 178 ml of stock lysis solution (see Notes 5 and 6). Refrigerate for at least 1 h before use. 13. Unwinding/electrophoresis buffer (reagents for 1 L freshly prepared): add 30 ml of 10 N NaOH to 4 ml of 0.25 M Na2EDTA, adjust the volume to about 900 ml ensuring that pH> 13 is reached and refrigerate at 4°C for at least 1 h (see Note 5). 14. Neutralization solution: 0.3 M sodium acetate in absolute ethanol (see Note 5). 15. Staining solution: prepare 1 mg/ml stock solution ethidium bromide in PBS and store at 4°C. Prior to stain, mix 12 µl stock solution in 988 µl PBS (working solution 12 µg/ml). This step must be performed wearing gloves (ethidium bromide is toxic).

3. Methods 3.1. Carbon Nanotubes Preparation

1. Suspend CNTs in cell culture medium at concentration of 0.5 mg/ml. 2. Sterilize by autoclaving (121°C for 30 min). 3. Disperse CNTs by 3 h treatment in an ultrasound bath prior to administer them to cell cultures (see Note 7).

3.2. Treatment of Cell Cultures

This procedure is performed in sterile conditions, using gloves and a laminar flow cabinet. Both floating and adherent cells can be used (see Note 8). 1. Seed cells at approximately 1 × 106/ml. 2. Treat with CNTs at the desired concentration and duration (see Note 9). 3. At the end of CNTs treatments wash cells by centrifugation (10 min at 400 × g) in culture medium and perform cell counting and viability determination (see Notes 10 and 11).

DNA Damage by Carbon Nanotubes Using the Single Cell Gel Electrophoresis Technique

3.3. Cell Counting and Viability

113

1. Transfer 10 µl of cell suspension in a microcentrifuge tube, add 90 µl of 0.5% trypan blue (w/v) and mix. 2. Transfer 10 µl of cell mixture to both chambers of the hemocytometer. 3. Count at the inverted microscope the viable (shiny) and dead (blue) cells on four squares from each chamber. 4. Determine % viability as (total number of viable cells/total number of cells) × 100, and cell number/ml as (number of viable cells/8) × 10,000 and × 10 (dilution factor). This procedure is performed at room temperature using gloves (Trypan blue is toxic).

3.4. Alkaline Comet Assay (see Note 12)

1. Prepare 1% NMA in PBS and melt at high temperature on a heating magnetic stirrer (avoid boiling).

3.4.1. Slides Preparation

2. Dip clean slides in 1% NMA, remove the agarose from the slide bottom by using a towel and leave the slides to dry horizontally overnight. Slides can be used for up to 1 week when stored at room temperature. 3. Prepare 0.5% LMA in PBS and dissolve at high temperature on a heating magnetic stirrer (avoid boiling). 4. Prepare aliquots of LMA (about 500 µl) in microcentrifuge tubes and keep them at 37°C in a water bath. 5. Pellet 100,000 cells for each sample with a microcentrifuge, discard supernatant and suspend cells (avoiding bubbles) in 200 µl of pre-warmed LMA (see Note 13). 6. Gently stratify 100 µl of cell suspension by a tip on a microscope slide pre-coated with NMA and pre-labelled (two slides are prepared for each sample with 50,000 cells each, to avoid the scoring bias due to excessive cell density). 7. Cover quickly (to avoid agarose condensation) with a cleaned cover slip to obtain an homogeneous stratification (see Note 14) and keep slides horizontally at 4°C for about 10 min to ensure the agarose to solidify. 8. Gently remove cover slip. 9. Add 100 µl pre-warmed LMA, cover with a cover slip (see Note 14) and leave at 4°C for additional 10 min. 10. Remove the cover slip and dip slides into cold lysis solution for at least 1 h. 11. Wash slides in a Coplin jar filled with cold electrophoresis buffer. 12. Transfer slides in an horizontal gel electrophoresis tank filled with cold electrophoresis buffer and leave for 1 h at 4°C to allow DNA unwinding and expression of alkali labile sites.

114

Zeni and Scarfì

13. Turn on power supply to 30 V (1.2 V/cm depending on the electrophoresis box size) and adjust the current to 300 mA by raising or lowering the buffer volume. Perform the electrophoresis for 50 min at 4°C (see Note 15). 14. Turn off the power, transfer the slides in a tray filled with neutralization buffer, and leave for 20 min in the dark at room temperature. 15. Dry slides at room temperature and stain in the dark with 60 µl of 12 µg/ml ethidium bromide, cover with a cover slip, gently press the cover slip to remove excess staining, and immediately observe at the fluorescent microscope (see Note 16). If slides need to be stored and observed later, after neutralization dehydrate leaving them for 2 h in absolute ethanol. Before microscope observation re-hydrate them in 70% ethanol in water for 5 min. Steps 12–14 have to be performed under yellow/dimmed light. This is to prevent any DNA damage that may arise from fluorescent white light. A schematic representation of the main steps of the protocol is depicted in Fig. 1. A list of troubleshooting is given in Note 17. 3.4.2. Evaluation of DNA Migration

A fluorescence microscope at 400× magnification is used to visualize DNA migration. Slides must be scored without knowledge of the treatment groups. Use an image analysis system to quantify DNA migration (see Note 18) by considering the following

Fig. 1. Schematic representation of the main steps to perform the comet assay

DNA Damage by Carbon Nanotubes Using the Single Cell Gel Electrophoresis Technique

115

Fig. 2. Alkaline single cell gel electrophoresis performed on human lymphocytes. (a) Untreated cells (control); (b) Cells with severe DNA damage after treatment with 50 µM methyl–methane sulphonate for 2 h

parameters: percent DNA in the tail, tail moment and tail length on ethidium bromide-stained nuclei (see Note 19). For an accurate analysis, 400 nuclei per sample should be examined (200 for each duplicate slide). An example of nuclei appearance is shown in Fig. 2. Concerning statistical analysis, it is important to highlight that the sample rather than the cell is the experimental unit, and statistical analyses based upon measures from the individual cells can lead to serious misinterpretation of results (8). On the other hand, one of the most important advantages of the comet assay is the ability to detect increased levels of damage among the subsets of cells within a larger population (7). The knowledge of the data distribution has a key role to choose the appropriate statistical analysis.

4. Notes 1. Unless otherwise stated, all solutions should be prepared in distilled water. This is referred to as “water” in the text. 2. Culture medium varies on the basis of cell type investigated. 3. An accurate characterization of the commercial CNTs employed can be useful in the interpretation of the results. In fact, metal residues derived from the production of nanomaterial that can vary in terms of amount, form, and encapsulation state (9), in some cases, can be responsible for the biological effect (10). 4. Sterile PBS is preferred to reduce fluorescence background due to impurities. 5. Volumes needed depend on the size of the container employed.

116

Zeni and Scarfì

6. DMSO is added in the lysis solution to scavenge free radicals. It is needed when slides are kept in lysis solution for a long period or when whole blood is employed. 7. CNTs are water-insoluble if not functionalized (11). 8. In the toxicity evaluation of CNTs with respect to human health, the response of human blood cells is interesting if they are to be used for drug delivery or bioimaging in disease and health, since they would be among the first exposed cell types upon intravenous administration. If isolated leukocytes are employed, separate cells from whole blood as follows: take 3 ml whole blood (WB) from a healthy donor by venipuncture; dilute 3 ml WB in 3 ml RPMI medium (usually 1 × 106 leukocytes are obtained from 1 ml WB); add 3 ml pre-warmed lymphoprep (LymphoprepTM, Axis–Shield, Oslo, Norway) in a 15 ml sterile tube, and gently stratify the diluted blood avoiding the mixture with lymphoprep; centrifuge at 800 × g for 30 min at room temperature; remove carefully the leukocytes and re-suspend them in 10 ml pre-warmed (37°C) sterile PBS; centrifuge at 500 × g for 10 min; remove supernatant and resuspend in 10 ml pre-warmed sterile PBS; centrifuge at 500 × g for 10 min and resuspend leukocytes in 5 ml complete RPMI medium. Count cells in Burker hemocytometer. 9. It is suggested to test more than a CNTs concentration, avoiding high doses. In fact, due to the strong tendency of CNTs to aggregate, the use of high concentrations can affect the results (12). Positive control has to be included in the experimental design to determine the sensitivity of the cells investigated and ensure that the technique is carried out properly. Usually methyl–methane sulphonate (10–50 µM final concentration for 2 h depending on cell type investigated) is employed as positive control, but other agents can also be used, depending on cell type. Alkylating agents, such as Mytomicin-C, should be avoided because they cause DNA cross-linking lesions that inhibit DNA unwinding (13). 10. If adherent cells are used, prior to count detach cells from the culture dish by trypsin treatment as follows: discard culture medium; add 0.005% trypsin to the cells; allow the cells to detach for 5 min at 37°C; add an equal volume of complete medium to neutralize trypsin. 11. As the comet assay measures DNA fragmentation in single cells, cytotoxicity induced DNA double strand breaks may confound the interpretation of data from this test. It has been recommended that viability measurements should always be made parallel to the comet assay. However, the determination of cell survival whenever possible seems to be useful with respect to the biological significance of test results

DNA Damage by Carbon Nanotubes Using the Single Cell Gel Electrophoresis Technique

117

since genotoxicity is only relevant if it occurs in cells capable of surviving the damage (7). 12. For each cell type and for each laboratory, the method should be adjusted to obtain valid and reproducible results. Main technical variables include concentration of LMA and NMA, composition of the lysis solution and lysis time, DNA unwinding and electrophoresis conditions. 13. This procedure should be performed as quickly as possible to avoid altered levels of damage. Increasing the time between the end of treatments and processing could reduce DNA damage due to DNA repair and/or the loss of heavily damaged cells. 14. Air bubbles are undesirable and careful application of the top layer minimizes their appearance. 15. On the basis of cell type investigation and the desired DNA migration in control cells, electrophoresis duration varies. The longer the electrophoresis duration, the higher the extent of DNA migration. However, after optimization, the same electrophoresis conditions must be used through the study. It is strongly suggested that untreated cells give comets with a background level of breaks of about 10% of DNA in the tail (3). 16. Ethidium bromide (EB) is most commonly used to stain the DNA on Comet Assay slides. Propidium iodide, 6-diamidino2-phenylindole (DAPI), YOYO dye, and Hoechst 33258 can be used as alternative stains for the visualization of comet DNA. The lower the stain concentration, the lower the fluorescence background. 17. Troubleshooting

• L arge comets on control and treated slides: (a) reduce time between cell harvest and slide preparation; (b) check for LMA temperature (37°C); (c) reduce electrophoresis voltage and/or duration.

• N o comets on control or treated slides: (a) test alternative lysis buffer or extend incubation time; (b) increase unwinding time; (c) increase electrophoresis voltage and/or duration.

• A garose detached from slides: (a) handle slides with care during incubation steps; (b) check for electrophoresis temperature; (c) verify the LMA percentage.

• C omets near the edges of a slide do not match comets of the remaining portion: do not score comets along the edge of slides.

• C omets on a slide differ widely for tail length and direction: check for the agarose layer uniformity.

118

Zeni and Scarfì

18. Several commercially available or public domain specific image analysis software packages can be employed, but they differ for the calculation of comet parameters. Visual scoring is also performed by several investigators. Collins and coworkers (14) developed a scheme for visual scoring based on five recognizable classes of comet, from class 0 (undamaged, no discernible tail) to class 4 (almost all DNA in tail). 19. Computed parameters generally considered to describe DNA migration are: (a) the percentage of migrated DNA (calculated as the integrated intensity of DNA in the tail, divided by the integrated intensity of DNA for the total image and multiplied by 100), (b) the tail length (in µm; the distance from the middle or estimated leading edge of the head region to leading edge of the tail) and (c) the tail moment, i.e. the product of the amount of DNA in the tail and the mean distance of migration in the tail (15). It is important to note that some parameters (e.g. tail moment) may be calculated differently among image analysis systems, and this can lead to quantitative differences, which can be problematic when comparing inter-laboratory data. If comparisons among laboratories are required, the DNA percentage in the tail is the most suitable parameter since it bears a linear relationship to break frequency and is relatively unaffected by threshold setting. Moreover, it allows discrimination of damage over the widest possible range (in theory, from 0% to 100% DNA in the tail) and gives a very clear indication of what the comets looked like (3, 16). References 1. Ostling O, Johanson KJ (1984) Microelectrophoretic study of radiation–induced DNA damages in individual mammalian cells. Biochem Biophys Res Commun 123:291–298 2. Singh NP, McCoy MT, Tice RR, Schneider EL (1988) A simple technique for quantitation of low level of DNA damage in individual cells. Exp Cell Res 175:184–187 3. Collins AR (2004) The comet assay for DNA damage and repair. Mol Biotechnol 26:249–254 4. Dixon DR, Pruski AM, Dixon LRJ, Jha AN (2002) Marine invertebrate eco-genotoxicology: a methodological overview. Mutagenesis 17:495–507 5. Fairbairn DW, Olive PL, O’Neill KL (1995) The comet assay: a comprehensive review. Mutat Res 339:37–59 6. Lee RF, Steinert S (2003) Use of the single cell gel electrophoresis/Comet Assay for detecting DNA damage in aquatic (marine and freshwater) animals. Mutat Res 544:43–64

7. Albertini RJ, Anderson D, Douglas RG, Hagmar L, Hemminki K, Merlo F, Natarajan AT, Norppa H, Shuker DE, Tice RR, Waters MD, Aitio A (2000) IPCS guidelines for the monitoring of genotoxic effects of carcinogens in humans. International Programme on Chemical Safety. Mutat Res 463: 111–172 8. Lowell DP, Thomas G, Dobow R (1999) Issues related to the experimental design and subsequent statistical analysis of in vivo and in vitro comet studies. Teratog Carcinog Mutagen 19:109–119 9. Hurt RH, Monthioux M, Kane A (2006) Toxicology of carbon nanomaterials: status, trends, and perspectives on the special issue. Carbon 44:1028–1033 10. Pulskamp K, Diabate S, Krug HF (2007) Carbon nanotubes show no sign of acute toxicity but induce intracellular reactive oxygen species in dependence on contaminants. Toxicol Lett 168:58–74

DNA Damage by Carbon Nanotubes Using the Single Cell Gel Electrophoresis Technique 11. Klumpp C, Kostarelos K, Prato M, Bianco A (2006) Functionalized carbon nanotubes as emerging nanovectors for the delivery of therapeutics. Biochim Biophys Acta 1758:404–412 12. Wick P, Manser P, Limbach LK, DettlaffWeglikowska U, Krumeich F, Roth S, Stark WJ, Bruinink A (2007) The degree and kind of agglomeration affect carbon nanotube cytotoxicity. Toxicol Lett 168:121–131 13. Miyamae Y, Zaizen K, Ohara K, Mine Y, Sasaki YF (1998) Detection of DNA lesions induced by chemical mutagens by the single cell gel electrophoresis (Comet) assay. 1. Relationship between the onset of DNA damage and the characteristics of mutagens. Mutat Res 415:229–235

119

14. Collins A, Dusinska M, Franklin M, Somarovska M, Petrovska H, Duthie S, Fillion L, Panayiotidis M, Raslova K, Vaughan N (1997) Comet assay in human biomonitoring studies: reliability, validation and applications. Environ Mol Mutagen 30:139–146 15. Olive PL (1999) DNA damage and repair in individual cells: applications of the comet assay in radiobiology. Int J Radiat Biol 65:395–405 16. Burlinson B, Tice RR, Speit G, Agurell E, Brendler-Schwaab SY, Collins AR et al (2007) In vivo Comet Assay workgroup, part of the Fourth International Workgroup on Genotoxicity Testing: results of the in vivo Comet Assay workgroup. Mutat Res 627:31–35

Part III Trafficking

Chapter 11 Assessment of Cellular Uptake and Cytotoxicity of Carbon Nanotubes Using Flow Cytometry Khuloud T. Al-Jamal and Kostas Kostarelos Abstract The field of carbon nanotube (CNT) functionalisation is increasingly growing for the purpose of enhancing the biocompatibility of CNT for medical and biological applications. Properties of CNT such as the type of functionalisation, charge density, and the dispersibility profile are expected to modulate CNT cellular uptake and toxicity profile in vitro. The assay described here allows for rapid screening of CNT cellular uptake in vitro and assessing the acute cytotoxicity simultaneously. CNT cellular uptake is measured qualitatively by light scattering analysis without differentiating between cell binding and internalisation of the CNT by the cells. In addition, flow cytometry is used to combine light scattering analysis with flow cytometry-based Annexin V/propidium iodide assay to measure the cytotoxicity. This assay is rapid, reliable, and allows for comparative analysis between various types of CNT studied. Key words: Carbon nanotubes, CNT, Association, Binding, Internalisation, Toxicity, Flow cytometry, Light scatter, Annexin V/PI, Apoptosis, Necrosis

1. Introduction The emergence of carbon nanotubes (CNT) as advanced nanomaterials, and in particular towards biomedical and biotechnological applications is of great interest (1–4). Due to advancements in the available functionalisation chemistries of CNT and the development of new constructs of polymer-CNT assembly, a range of surface functionalised carbon nanotubes (f-CNT) of various types, charge densities, and dispersibility profiles have been generated (3, 5–10). In terms of the biologically relevant features of CNT, one of the most attractive properties described is the capacity to translocate cellular barriers (such as the plasma membrane) by

K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_11, © Humana Press, a part of Springer Science + Business Media, LLC 2010

123

124

Al-Jamal and Kostarelos

mechanisms that are novel and seem to be reminiscent of a nano-needle piercing the cells (11, 12). Numerous laboratories using various types of CNT have now reported their cellular uptake by a wide range of cells (11, 13–16). Therefore, a facile and rapid screening method to study the interaction of non-fluorescently labelled CNT with cells is needed. Some of the most popular techniques employed to study the interaction of CNT with mammalian cells in vitro include confocal laser scanning microscopy (17, 18), light microscopy, and transmission electron microscopy (19). Confocal laser scanning microscopy requires the tagging of CNT with fluorescent probes while transmission electron microscopy is a laborious process that not many laboratories may have access to and is not intended for routine use. Moreover, almost all microscopy-based techniques will always provide qualitative information with regards to the interaction between nanomaterials, including CNT, and cellular surfaces and compartments, with statistical analysis almost impossible to infer. Flow cytometry-based assays have been proposed to assess cell-nanoparticle associations (both cell bound or internalised) either by quantitatively measuring the absolute number of fluorescent particles associated with cells or by qualitatively measuring the increase in the sideward scattering of cells incubated with non-fluorescent particles. Qualitative measurements are based on the fact that as nanoparticles bind to the cells, the granularity of the cells increases, which concomitantly increases the sideward scattering intensity. Although internalisation of a variety of CNT into various types of cells can be seen as an attractive feature of these nanomaterials, particularly in relation to the delivery or detection of molecules intracellularly, there may be cytotoxic side-effects associated with such property that should be considered. In all cases that interaction between CNT and the biological milieu takes place, the ensuing cytotoxicity needs to be addressed. The new subfield of nanotoxicology has emerged in the last few years to specifically address and obtain a better understanding of the impact of novel nanoparticles and their health hazards. The cytotoxicity of CNT has been assessed in vitro using methods that are well described in the literature (20, 21). They include direct counting of cell numbers through Trypan Blue exclusion assay (22), colorimetric assays such as the MTT (23, 24) or the LDH assay (23), measurement of protein concentrations by the Bradford assay (25), or by clonogenic assay (26). FACS and confocal microscopy cytotoxicity assays, such as the mitochondrial membrane potential determination and Annexin V/propidium iodide (PI) staining, have also been reported as methods to assess the cytotoxicity of CNT (22, 23, 27). There have been conflicting conclusions from the assessment of the various types of CNT by different groups. One reason for

Assessment of Cellular Uptake and Cytotoxicity of Carbon

125

the discrepancies resulting from cytotoxicity assays is due to the variation in the impurities contained in different CNT preparations (i.e. metals such as iron and nickel in addition to the amorphous carbon content). Moreover, some studies have shown contradicting results when assessing the cytotoxicity of CNT by colorimetry-based assays such as MTT, WST-1, XTT or LDH assays (23). It has been already reported that the presence of CNT may lead to false positive results with the MTT assay due to the interaction between the water-insoluble MTT formazan crystals and the CNT backbone after MTT reduction by mitochondrial dehydrogenase of physiologically active cells. On the contrary, no interaction seems to occur between the CNT and the watersoluble formazan products produced by reduction of tetrazolium salts such as WST-1, XTT or INT which are the main components of the WST-1, XTT or LDH assays (23). Besides interactions reported in colorimetry-based assay, one has to be careful in fluorescence-based cytotoxicity assays, as interaction between CNT and fluorescent probes may occur based on the properties on the CNT studied (surface, length, and dispersibility), the fluorescent molecules used and the concentration range of the CNT tested. Another common issue arises from the fact that the total surface area available from the CNT is enough to adsorb reagent or fluorescent molecules (28), particularly those with many aromatic rings, therefore leading to false negative cytotoxicity results. There have been reports that CNT may also be able to quench the fluorescence of quantum dots (QDs), that are strongly and intrinsically fluorescent nanoparticles, through the formation of QD supramolecular assemblies around the CNT (29). The assay presented in this method is designed to measure qualitatively by light scattering analysis, the CNT association with cells without differentiating between those nanotubes that bind on the cells to those that are internalised by the cells. In addition, flow cytometry is used to combine light scattering analysis with flow cytometry-based Annexin V/PI cytotoxicity assay to allow for simultaneous and rapid screening for CNT-cell association and cytotoxicity assessment.

2. Materials 1. Nanoparticle cell uptake experiments can be performed using adherent cells such as A549 lung epithelial cell line (CCL-185, ATCC, UK) or other cell lines. Cells should be removed from the culture dish before analysis with flow cytometry by trypsinisation. 2. Trypsin-EDTA (Gibco, Invitrogen, UK).

126

Al-Jamal and Kostarelos

3. Culture media appropriate for the cell line studied. When A549 monolayers were used, the media used contained Dulbecco’sModifiedEagleMedium(DMEM)(Gibco,Invitrogen, UK), supplemented with 10% foetal bovine serum (FBS) (Gibco, Invitrogen, UK), 50 U/ml penicillin, and 50 µg/ml streptomycin (Gibco, Invitrogen, UK). 4. 24 Well Clear TC-treated microplates (Costar®, USA). 5. 1.5-ml microcentrifuge tubes. 6. 1, 5, and 25 ml serological pipettes (VWR, UK). 7. 10 µl, 200 µl, and 1 ml pipette tips (Starlab Ltd, UK). 8. 37°C and CO2 incubator for maintaining the cells. 9. Centrifuge (350×g for 5 min) for pelleting cells. 10. Bath sonicator (Ultrasonic cleaner, VWR). 11. Annexin-V-Fluos staining kit (Roche Diagnostics GmbH, Germany), if toxicity experiments are needed. The kit contains ready-to-use Annexin V-FITC solution, PI solution, and HEPES incubation buffer. 12. CNT preparation to be tested, CNT functionalisation and dispersibility properties may vary. Pristine CNT can be used as an example of non-functionalised CNT. 13. Pluronic F127 co-polymer (Sigma) as dispersing agent for pristine CNT.

3. Methods 3.1. Preparation of CNT Dispersions

1. Disperse CNT powder in 5% dextrose, PBS or water, up to a maximum concentration of 1 mg/ml, by bath sonication for 15 min. 2. Store CNT dispersion in the fridge when not in use, and sonicate each time immediately before use. 3. Pristine CNT can be dispersed in 1% Pluronic F127 in water, aided by bath sonication for 15 min. Final concentration of Pluronic F127 when incubated with the cells should not exceed 1%.

3.2. Cell Culture

The following protocol describes the incubation of CNT with A549 lung epithelial cell lines. However, the type of cells can be changed based on the experimental design. The cells can be treated with various inhibitors, while the CNT can be modified by surface functionalisation or by changing the dispersing agent. The time of incubation with cells and CNT concentration can also be varied.

Assessment of Cellular Uptake and Cytotoxicity of Carbon

127

1. A549 cells should be passaged when they reach 80% confluence in order to maintain exponential growth and used for a maximum of 10 passages. 2. To trypsinise the monolayer, rinse it with PBS then incubate with trypsin-EDTA at 37°C for 5 min; the cells are then detached by vigorous up and down pipetting. 3. Centrifuge cells at 350×g for 5 min at 4°C and resuspend in complete media. 4. Count cells and determine cell viability by Trypan blue dye exclusion assay. 5. Adjust cell suspension to 100,000 cells/ml in complete media. 6. Seed 50,000 cells per each well of 24-well plates and incubate for 24 h at 37°C in a humidified atmosphere (5% CO2) incubator. 7. Add CNT to 500 µl complete media in 1.5-ml microcentrifuge tubes, and vortex the suspension to mix. Use CNT concentrations between 1 and 100 µg/ml. 8. Allow the cells to interact with the CNT for 15 min, 60 min, 4 h, 24 h and 72 h, at 37°C in a humidified atmosphere (5% CO2) incubator. 9. After incubation period is finished, aspirate media containing the CNT. 10. Add 500 µl of PBS to rinse the cells and remove unbound CNT (see Note 1). 11. Remove the adherent cells by adding 100 µl Trypsin-EDTA per well and incubate the cells at 37°C in a humidified atmosphere (5% CO2) incubator. 12. Add 500 µl of tissue culture media, and detach the cells by vigorous up and down pipetting. 13. Transfer cells to 1.5-ml microcentrifuge tubes. 14. Centrifuge cells at 350×g for 5 min at 4°C and resuspend in PBS. 15. Keep cell suspensions on ice, and analyse immediately by flow cytometry. 3.3. Instrumentation and Gating

The scatter plots and gating have to be performed first. In the following methods described, the commands used are specific for the Summit version 4.3 for use with the CyAn™ ADP HighPerformance Research Flow Cytometer (DakoCytomation, USA).

3.3.1. Setting a Bivariate Scatter Histogram

1. Set a bivariate sideward scattering (SS Lin) vs forward scattering histogram (FS Lin). The SS Lin should be displayed on the ordinate and FS Lin displayed on the abscissa (Fig. 1a).

128

Al-Jamal and Kostarelos

Fig. 1. A bivariate scatter histogram of the sideward scattering vs. forward scattering signals recorded with FACS for (a) cells incubated with media without CNT, (b) cells incubated with 10-µg/ml cationic f-CNT for 24 h, and (c) 10-µg/ml cationic f-CNT without cells. Each recorded event is presented as a point in the diagram. (d) A univariate scatter histogram showing the cell number (counts) vs. sideward scattering (SS Lin) for cells incubated with varying concentrations of the f-CNT for 24 h. (e) and (f) are light micrographs of untreated A549 cells and cells treated with 10-µg/ml cationic f-CNT for 24 h, respectively

Assessment of Cellular Uptake and Cytotoxicity of Carbon

129

2. The photomultiplier tube (PMT) voltage has to be set in a way that both the negative control (a sample containing cells without CNT) (Fig. 1a) and the positive control (a sample containing cells incubated with cationic CNT for 24 h) (Fig. 1b) have channel number from 0 to 265 for both ordinate and abscissa. If one cell type is used, only one cluster of cells will be observed (R1) and selected for setting a univariate scatter histogram (see Subheading 3.3.3). 3. In order not to include the CNT in the cell population group, a suspension containing the highest concentration of CNT without any cells should be run first to exclude any free CNT from R1 group (Fig. 1c). 3.3.2. Gating Cells from Bivariate Scatter Plot

1. Gating means an electronic gate (R1) should be selected inside the bivariate plot (Fig. 1). In this protocol, the gate R1 is selected to include all the cells being studied and exclude any cell debris (Fig. 1b) or free CNT (Fig. 1c). 2. Gate the cells to include cells over the entire range of the SS Lin channel (0–265). Cells interacting with CNT will have high SS Lin channel number, therefore, histograms of a positive and a negative control cells should be used when selecting the gate in order not to underestimate the SS Lin after incubation with CNT. 3. Adjust the optimum FS Lin (width) of the gate to remove the free unbound CNT and cell debris. Cell debris samples appear as a population with small FS Lin and small SS Lin (Fig. 1a). CNT appear as a population with small FS Lin and medium SS Lin (Fig. 1c). This step is important to minimise any interference from the CNT particles on both light scatter and toxicity data.

3.3.3. Setting a Univariate Scatter Histogram

3.3.4. Measuring SS Lin of the Samples

Set a univariate histogram that plots cell number (counts) on the ordinate and SS Lin on the abscissa (Fig. 1d). The abscissa channel number should be set from 0 to 265. SS Lin should be gated from the cell population R1 in the bivariate plot (see Subheading 3.3.2). 1. Measure the SS Lin of a sample containing the CNT alone at the highest concentration used in the experiment. Verify that gate R1 set in the bivariate sideward scattering (SS Lin) vs. forward scattering histogram (FS Lin) exclude all the unbound CNT. 2. Measure the SS Lin of the negative control samples (cells without CNT) and set the SS Lin channel number to around 50 (0–256 scale). 3. Measure the SS Lin of the positive control samples (cells incubated with cationic CNT for 24 h). Verify that gate R1

130

Al-Jamal and Kostarelos

set in the bivariate sideward scattering (SS Lin) vs. forward scattering histogram (FS Lin) include all the cells and exclude the free CNT. 4. Analyse at least 20,000 cells per sample. Record the median SS Lin from the univariate sideward scattering (SS Lin) set in Subheading 3.3.3. 5. Cell-association data will be analysed as in Subheading 3.3.5 (see Notes 2 and 3). 3.3.5. Data Analysis

Typical result from the 2D density plot of the forward scattering and sideward scattering is shown in Fig. 1a. High forward scattering events correspond to large particles such as cells (Fig. 1a and b) whereas smaller forward scattering events correspond to smaller particles such as CNT (Fig. 1c) or cell debris (Fig. 1b). Taking into account the events with high forward scattering events (cells), high sideward scattering events correspond to cells that are associated with CNT which can either be bound to the cell membrane or internalised by the cell (Fig. 1b). On the other hand, low sideward scattering events correspond to cells without associated CNT (Fig. 1a). Data analysis is done by comparing the sideward scattering intensity of control cells (no CNT) to that of cells incubated with CNT. The median sideward scattering intensity from the univariate sideward scattering histogram can be used for comparison (Fig. 1d). Data are generally expressed as fold changes in median sideward scattering intensity for a given CNT/ cell sample as compared to that of control cells.

3.4. CNT Cytotoxicity by Flow Cytometry

Cell death is known to occur by two distinct modes: necrosis and apoptosis (30, 31). Although morphological, biological, and molecular differences between necrosis and apoptosis are evident, the boundaries between necrosis and apoptosis are not always very clear as the patterns of biochemical or morphological changes are not always of typical necrosis or apoptosis. The light scattering properties of cells during death can change due to morphological changes such as cell swelling, cell shrinkage, rupture of the plasma membrane, chromatin condensation, nucleus fragmentation, and shedding of apoptotic bodies. Necrotic death is characterised by rapid initial increase in forward and sideward scattering due to cell swelling. Apoptotic death is characterised by a decrease in both forward and sideward scattering, however, an initial increase in sideward scattering parallel with a decrease in forward scattering has been observed in some cell types (32, 33). In general, broken cells, isolated nuclei, cell debris, and apoptotic bodies have low light scatter properties. Since light scatter analysis is specific to neither apoptosis nor necrosis, more mechanistic data can be obtained by combining light scatter analysis to another cytofluorimetric analysis such as Annexin V/PI staining.

Assessment of Cellular Uptake and Cytotoxicity of Carbon 3.4.1. Annexin V/PI Staining

131

In this assay, we combine analysis of CNT-cell association by light scatter changes with another cytofluorometric cytotoxicity assay for the detection of phosphatidylserine with Annexin V-FITC conjugate. In early stages of apoptosis, the plasma membrane phospholipid, phosphatidylserine (PS), is exposed to the outside of the plasma membrane (34). Annexin V is a Ca++-dependent phospholipid-binding protein, which binds to PS residues. Annexin-V-FITC conjugate can be used to detect apoptosis. Since PS externalisation may also happen during cell necrosis, including membrane impermeable dye such as PI can distinguish apoptotic cells from necrotic cells. Non-apoptotic non-necrotic cells are Annexin V-negative (FITC−) and PI-negative (PI−), early apoptotic cells are Annexin V-positive and (FITC+) and PI-negative (PI−), and late apoptotic cells and necrotic cells are intensely stained with PI. If cytotoxicity assessment is required, follow steps 1–13 (see Subheading 3.2) and continue as below: 1. Aspirate PBS from cell pellets. 2. Add 500 µl of PBS to remove any serum traces. 3. Centrifuge cells at 350×g for 5 min at 4°C. 4. Prepare Annexin V-FITC/PI labelling solution as instructed by the manufacturer. Mix 98 µl of HEPES incubation buffer with 1 µl of ready-to-use Annexin V-FITC and 1 µl of ready-to-use PI solution. 5. Keep three tubes with cells that are unlabelled, labelled with Annexin V-FITC only or PI only, to set the photomultiplier tube (PMT) voltage and compensation settings. 6. Aspirate the PBS from each 1.5 ml microcentrifuge tubes and add 100 µl of the above solution to each tube. 7. Resuspend the cell pellet in 100 ml of Annexin V-FITC/PI labelling solution, and incubate for 10–15 min at 15–25°C, then keep cell suspensions on ice and analyse immediately by flow cytometry. 8. Just before analysis, dilute the cell suspension with 0.4 ml of HEPES incubation buffer and transfer to a test tube for analysis directly on the flow cytometer.

3.4.2. Flow Cytometry

1. Samples are analysed on a flow cytometer using 488 nm excitation and a 515 nm bandpass filter for fluorescein detection and a filter of 615 nm for PI detection. 2. Electronic compensation of the instrument is performed to exclude overlapping of the two emission spectra. 3. At least 20,000 cells per sample are analysed. 4. Cytotoxicity data will be analysed as in Subheading 3.4.3.

132

Al-Jamal and Kostarelos

3.4.3. Data Analysis

Cell death was expressed as percentage cell population stained with Annexin V-FITC or with PI staining (see Note 4).

4. Notes 1. CNT may bundle and if the size of the bundle is not small enough (i.e. less than 2 µm), they will be counted as unstained cells leading to false negative results. Therefore, it is highly recommended that the cytotoxicity of CNT needs has to be verified using at least two independent cytotoxicity assays. 2. This is a qualitative rather than a quantitative assay that measures CNT-cell association, in contrast to assays using fluorescent CNT where precise quantification of the percen tage of internalised CNT may be feasible. 3. From Fig. 1b, it is difficult to distinguish between the CNT that are bound to the cell surface or internalised by the cells because sideward scattering intensity is an indication of surface roughness of the cell. However, this technique can be combined with other techniques such as light microscopy, transmission electron microscopy, or confocal laser scanning microscopy to distinguish CNT cell binding from internalisation. Light scattering analysis is being proposed here as a screening tool to study the effect of varying CNT characteristics on their interaction properties with cells. CNT of different surface charge, charge density, and dispersibility profile can be tested comparatively. Data generated using light microscopy established a good correlation between the increase in sideward scattering intensity and the increase in CNT intracellular accumulation (Fig. 1e and f), which suggested that adsorption of the CNT onto the cell membrane will eventually lead to intracellular uptake. 4. A 2D density plot of the green (apoptosis) and red fluorescence (necrosis) is shown in Fig. 2. The events shown in Fig. 2 are gated to the cell events (R1) chosen in Fig. 1. The diagram can be divided into four populations. Events with low red and low green fluorescence (bottom left quadrant) correspond to non-apoptotic non-necrotic cells, events with low red and high green fluorescence (bottom right quadrant) correspond to early apoptotic cells, events with high red and high green fluorescence (top right quadrant) correspond to late apoptotic or necrotic cells, and events with high red and low green fluorescence (top left quadrant) correspond to nuclear fragments. Integration over all events in each quadrant yields the total number of events.

Assessment of Cellular Uptake and Cytotoxicity of Carbon

133

Fig. 2. Evaluation of cell viability using Annexin V/PI staining. All events are gated to R1 population obtained in Fig. 1. The diagram can be divided into four populations. Events with low red and low green fluorescence (bottom left quadrant) correspond to non-apoptotic non-necrotic cells, events with low red and high green fluorescence (bottom right quadrant) correspond to early apoptotic cells, events with high red and high green fluorescence (top right quadrant) correspond to late apoptotic or necrotic cells, and events with high red and low green fluorescence (top left quadrant) correspond to nuclear fragments

References 1. Bianco A, Kostarelos K, Prato M (2005) Applications of carbon nanotubes in drug delivery. Curr Opin Chem Biol 9:674–679 2. Bianco A, Kostarelos K, Prato M (2008) Opportunities and challenges of carbon-based nanomaterials for cancer therapy. Expert Opin Drug Deliv 5:331–342 3. Karousis N, Ali-Boucetta H, Kostarelos K, Tagmatarchis N (2008) Aryl-derivatized, water-soluble functionalized carbon nanotubes for biomedical applications. Mater Sci Eng B 152:8–11 4. Lacerda L, Bianco A, Prato M, Kostarelos K (2006) Carbon nanotubes as nanomedicines: from toxicology to pharmacology. Adv Drug Deliv Rev 58:1460–1470 5. Campidelli S, Sooambar C, Lozano DE, Ehli C, Guldi DM, Prato M (2006) Dendrimer-functionalized single-wall carbon nanotubes: synthesis, characterization, and photoinduced electron transfer. J Am Chem Soc 128:12544–12552 6. Garcia A, Herrero MA, Frein S, Deschenaux R, Munoz R, Bustero I, Toma F, Prato M (2008) Synthesis of dendrimer-carbon nanotube conjugates. Phys Stat Sol A 205:1402–1407

7. Georgakilas V, Tagmatarchis N, Pantarotto D, Bianco A, Briand JP, Prato M (2002) Amino acid functionalisation of water soluble carbon nanotubes. Chem Commun (24):3050–3051 8. Herrero MA, Prato M (2008) Recent advances in the covalent functionalization of carbon nanotubes. Mol Cryst Liq Cryst 483:21–32 9. Pastorin G, Wu W, Wieckowski S, Briand JP, Kostarelos K, Prato M, Bianco A (2006) Double functionalisation of carbon nanotubes for multimodal drug delivery. Chem Commun (11):1182–1184 10. Tasis D, Tagmatarchis N, Bianco A, Prato M (2006) Chemistry of carbon nanotubes. Chem Rev 106:1105–1136 11. Kostarelos K, Lacerda L, Pastorin G, Wu W, Wieckowski S, Luangsivilay J, Godefroy S, Pantarotto D, Briand JP, Muller S, Prato M, Bianco A (2007) Cellular uptake of functiona lized carbon nanotubes is independent of func tional group and cell type. Nat Nanotechnol 2:108–113 12. Lacerda L, Raffa S, Prato M, Bianco A, Kostarelos K (2007) Cell-penetrating CNTs for delivery of therapeutics. Nano Today 2:38–43

134

Al-Jamal and Kostarelos

13. Becker ML, Fagan JA, Gallant ND, Bauer BJ, Bajpai V, Hobbie EK, Lacerda SH, Migler KB, Jakupciak JP (2007) Length-dependent uptake of DNA-wrapped single-walled carbon nanotubes. Adv Mater 19:939 14. Kam NW, Liu Z, Dai H (2006) Carbon nanotubes as intracellular transporters for proteins and DNA: an investigation of the uptake mechanism and pathway. Angew Chem Int Ed Engl 45:577–581 15. Kang B, Yu DC, Chang SQ, Chen D, Dai YD, Ding YT (2008) Intracellular uptake, trafficking and subcellular distribution of folate conjugated single walled carbon nanotubes within living cells. Nanotechnology 19:375103–375111 16. Shi Kam NW, Jessop TC, Wender PA, Dai H (2004) Nanotube molecular transporters: internalization of carbon nanotube-protein conjugates into Mammalian cells. J Am Chem Soc 126:6850–6851 17. Pantarotto D, Briand JP, Prato M, Bianco A (2004) Translocation of bioactive peptides across cell membranes by carbon nanotubes. Chem Commun (1):16–17 18. Lacerda L, Pastorin G, Gathercole D, Buddle J, Prato M, Bianco A, Kostarelos K (2007) Intracellular trafficking of carbon nanotubes by Confocal laser scanning microscopy. Adv Mater 19:1789 19. Porter AE, Gass M, Muller K, Skepper JN, Midgley PA, Welland M (2007) Direct imaging of single-walled carbon nanotubes in cells. Nat Nanotechnol 2:713–717 20. Smart SK, Cassady AI, Lu GQ, Martin DJ (2006) The biocompatibility of carbon nanotubes. Carbon 44:1034–1047 21. Ying Z, WenXin L (2008) Cytotoxicity of carbon nanotubes. Sci China B 51:1021–1029 22. Bottini M, Bruckner S, Nika K, Bottini N, Bellucci S, Magrini A, Bergamaschi A, Mustelin T (2006) Multi-walled carbon nanotubes induce T lymphocyte apoptosis, Toxicol Lett 160:121–126 23. Worle-Knirsch JM, Pulskamp K, Krug HF (2006) Oops they did it again! Carbon nanotubes hoax scientists in viability assays. Nano Lett 6:1261–1268 24. Davoren M, Herzog E, Casey A, Cottineau B, Chambers G, Byrne HJ, Lyng FM (2007) In vitro toxicity evaluation of single walled carbon nanotubes on human A549 lung cells. Toxicol In Vitro 21:438–448

2 5. Isobe H, Tanaka T, Maeda R, Noiri E, Solin N, Yudasaka M, Iijima S, Nakamura E (2006) Preparation, purification, characterization, and cytotoxicity assessment of watersoluble, transition-metal-free carbon nanotube aggregates. Angew Chem Int Ed Engl 45: 6676–6680 26. Herzog E, Casey A, Lyng FM, Chambers G, Byrne HJ, Davoren M (2007) A new approach to the toxicity testing of carbon-based nanomaterials–the clonogenic assay. Toxicol Lett 174:49–60 27. Dumortier H, Lacotte S, Pastorin G, Marega R, Wu W, Bonifazi D, Briand JP, Prato M, Muller S, Bianco A (2006) Functiona lized carbon nanotubes are non-cytotoxic and preserve the functionality of primary immune cells. Nano Lett 6:1522–1528 28. Hurt RH, Monthioux M, Kane A (2006) Toxicology of carbon nanomaterials: status, trends, and perspectives on the special issue. Carbon 44:1028–1033 29. Cui D, Pan B, Zhang H, Gao F, Wu R, Wang J, He R, Asahi T (2008) Self-assembly of quantum dots and carbon nanotubes for ultrasensitive DNA and antigen detection. Anal Chem 80:7996–8001 30. Compton MM (1992) A biochemical hallmark of apoptosis: internucleosomal degradation of the genome. Cancer Metastasis Rev 11: 105–119 31. Dive C, Gregory CD, Phipps DJ, Evans DL, Milner AE, Wyllie AH (1992) Analysis and discrimination of necrosis and apoptosis (programmed cell death) by multiparameter flow cytometry. Biochim Biophys Acta 1133: 275–285 32. Darzynkiewicz Z, Bruno S, Del BG, Gorczyca W, Hotz MA, Lassota P, Traganos F (1992) Features of apoptotic cells measured by flow cytometry. Cytometry 13:795–808 33. Ormerod MG, Paul F, Cheetham M, Sun XM (1995) Discrimination of apoptotic thymocytes by forward light scatter. Cytometry 21:300–304 34. Vermes I, Haanen C, Steffens-Nakken H, Reutelingsperger C (1995) A novel assay for apoptosis. Flow cytometric detection of phosphatidylserine expression on early apoptotic cells using fluorescein labelled Annexin V. J Immunol Methods 184: 39–51

Chapter 12 Cell Trafficking of Carbon Nanotubes Based on Fluorescence Detection Monica H. Lamm and Pu Chun Ke Abstract Cell trafficking of carbon nanotubes (CNTs) is an area of scientific inquiry that has great implications in medicine, biosensing, and environmental science and engineering. The essence of this inquiry resides in the interaction of carbon nanostructures and cell membranes, regulated by the laws of molecular cell biology and the physiochemical properties of the nanostructures. Of equal importance to this inquiry is a description of cellular responses to the integration of man-made materials; yet, how cellular responses may invoke whole-organism level reaction remains unclear. In this chapter, we show three experimental studies, which may be beneficial to obtaining such an understanding. Among the reservoir of methodologies, which have proved of merit, we focus our attention on fluorescence microscopy, one of the most powerful and yet least invasive means of probing nanoparticles in biological systems. Especially, we present the method of fluorescence energy transfer induced between a lysophospholipid molecule and a single-walled CNT upon cellular uptake, and describe coating nanotubes with RNA and suspending fullerenes with phenolic acids for facilitating their translocation across cell membranes and shuttling between cell organelles. Finally, we comment on the perspective of using molecular simulations for facilitating and guiding such experiments. Key words: Cell membrane, Uptake, CNT, Fullerene, Lipid, Gallic acid, Energy transfer

1. Introduction Much progress has been made since Richard Feynman’s famous lecture “There is plenty of room at the bottom”. Nanotechnology today enjoys a great presence at the frontiers of electronics, molecular assembly, tissue engineering, biosensing, and nanomedicine. Carbon nanotubes (CNTs), a major class of tubular nanostructures, have been utilized as catalysts for facilitating biochemical reactions and biological processes (1), as substrates for detecting single antibody binding and gene sequencing (2),

K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_12, © Humana Press, a part of Springer Science + Business Media, LLC 2010

135

136

Lamm and Ke

and as carriers for gene and drug delivery (3). Since the first in vitro gene and protein delivery using CNTs by Dai’s group at Stanford (4), Prato and Bianco’s groups at the University of Trieste and CNRS (5), and our group at Clemson (6), all in the exciting year of 2004 have now come to the realization that the uptake of CNTs may occur irrespectively of (mammalian) cell types (7). Consequently, research in this field has now focused on describing in vivo responses and genotoxicity (8–12), for biological and environmental systems that may be under the threat of nanomaterials exposure. All the complexity associated with CNT cell trafficking may arise from this simple fact: CNTs are inherently hydrophobic but can become water soluble through covalent (i.e., organic chemistry) or noncovalent surface functionalization (i.e., supramolecular assembly). In addition to increased solubility, functionalized CNTs also invoke minimal immunogenicity and toxicity, and show no apparent tissue/organ accumulation. Technically, when trespassing upon cell membranes and transporting within cellular space, CNTs may be followed with single molecule resolution by monitoring the near infrared photoluminescence of individual single-walled carbon nanotubes (SWNTs). This approach requires the use and adaptation of an infrared CCD camera and is spearheaded by Wiseman’s group at Rice (13) and Strano’s group at MIT (14). One obvious advantage of this method is its minimal v to cells and high signal-to-noise ratio due to suppressed cell autofluorescence in near infrared. Alternatively, as we shall see in the following, CNTs and fullerenes in extra- and intracellular spaces may be detected using the technique of fluorescence energy transfer, or, by tracing the fluorescence of RNA polymer or fullerenegallic acid. Furthermore, in silico simulations – the “third way” of doing science, may provide complementary information on the subject of CNT cell trafficking.

2. Materials 2.1. Lysophospholipid and Hydrolytic Enzyme

1. 1,2-Dipalmitoyl-sn-Glycero-3-Phosphoethanolamine-N(Lissamine Rhodamine B Sulfonyl), Rd-PE (Avanti Polar Lipids), Ex/Em: 571/583 nm. 2. Phospholipase A2 (Sigma).

2.2. Carbon Nanoparticles and Coating Agents

1. HiPco SWNTs, from Nano-C, Inc. (Prepurified, 70 or 97% preferred. see Note 1) or grown using the chemical vapor deposition technique as follows. Add 1.5 mg of iron nitrate to 10 mL of isopropanol and sonicate for 10 min. Dip silicon substrates in this solution for 10 s, rinse in n-hexane and then

Cell Trafficking of Carbon Nanotubes Based on Fluorescence Detection

137

bake in a glass oven at 200°C for 15 min before placing them inside a quartz tube furnace. Anneal the substrates at 750°C for 30 min under the flow of 600 standard cubic centimeters (sccm) argon and 200~400 sccm hydrogen. Inject xylene at the rate of 1 mL/h for 2 h. Finally, cool the furnace under flowing argon. 2. RNA polymer poly(rU), 500–2,600 nucleotides (Midland). 3. Fullerene C70, 99% purity (Nano-C, Inc. or SES Research). 4. Gallic acid (C7H6O5, FW: 170; Sigma). 2.3. Cell Lines, Buffer, Radioactive, and Fluorescence Labeling

1. MCF7 breast cancer cell line (ATCC). 2. HT-29 colon cancer cell line (ATCC). 3. TE buffer: 10 mM Tris–HCl, 1 mM EDTA. 4. Radioactive thymidine [methyl-3H] (MP Biomedicals). 5. Lipophilic dye 1,1¢ dioctadecyl-3,3,3¢,3¢-tetramethylindo carbocyanine perchlorate, DiIC18 (Molecular Probes), Ex/ Em: 644/665 nm, for membrane labeling. 6. Propidium iodide dye, PI (Sigma), Ex/Em: 535/617 nm, for labeling RNA. PI is known to intercalate between the base pairs of nucleic acids. PI is also membrane impermeable, and is therefore generally excluded by viable cells. 7. Filter, MWCO 3,000 Da (Millipore). 8. DispoDialyzer filters, MWCO 500 Da (Spectrumlabs). 9. Cassette Slide-A-Lyzer, MWCO, 10,000 (Pierce).

2.4. Instruments

1. Mass spectrometer (Q-Tof micro™). 2. Spectrofluorometer (QM-8/2005, PTI, resolution: 1 nm). 3. Confocal fluorescence microscope LSM510 (Zeiss). 4. Raman spectrometer Triax 550 (ISA). 5. UV-vis spectrophotometer (Biomate 3). 6. Sonicator VC 130 PB (Sonics & Materials). 7. Scintillation counter LS6500 (Beckman).

3. Methods 3.1. Detection of SWNT Uptake Using Fluorescence Energy Transfer

The mechanism of fluorescence energy transfer was first deciphered by Stryer and Haugland in the sixties of the twentieth century (15). The principle of this physical phenomenon is the induced-dipole and induced–dipole interaction between two fluorophores, i.e., a donor, and its neighboring fluorophore, an acceptor. Since this energy transfer has an inverse sixth power

138

Lamm and Ke

distance dependence, its effect can only be felt with a spatial range of 10 nm. Techniques based on this mechanism, much thrived since 1996 (16), have been utilized for characterizing intermolecular interaction and intramolecular fluctuations. Isolated SWNTs have recently been reported as robust fluorescent tags for cellular imaging and sensing (13, 14). Semiconducting SWNTs have very large spectral shifts between excitation and emission wavelengths when they are excited at second van Hove absorption transitions (500–900 nm) and detected through first van Hove emission (800–1,600 nm) (17). Photophysical studies have found that SWNTs can act collectively as quenchers for their covalently tethered and/or p-stacked pyrenes and chromophores (18, 19). These phenomena were attributed to electron transfer or energy transfer from the photoactive tethers to the SWNTs, resulting from their tight binding as well as the broad absorption spectrum of the SWNTs. We have recently introduced the use of lysophospholipids, or single-tailed phospholipids, for the effective suspension of SWNTs in the aqueous phase (20, 21). We have further observed the energy transfer between a fluorescently labeled lysophospholipid, Rd-LPE, and its substrate SWNT upon binding (Fig. 1) (22). Although the binding mode of amphiphiles and SWNT is still under debate, this physical method offers a sensitive tool for gene and drug delivery, as well as for sensing the uptake of CNTs by cell membranes. 3.1.1. Cleavage of Phospholipids

1. To obtain Rd-LPE, cleave off one fatty acyl tail of the Rd-PE using enzyme phospholipase A2, which hydrolyzes the acyl group attached to the SN2 carbon of Rd-PE. 2. At the conclusion of the digestion, expect four products to remain: Rd-LPE, fatty acyl tails, undigested Rd-PE, and free PE head groups. These groups can be separated by thin layer chromatography (TLC).

Fig. 1. Scheme of the energy transfer between Rd-LPE molecules (Rd small spheres, LPE large spheres with single fatty acyl tails) and an SWNT (gray cylinder). The laser excited rhodamine fluorescence is quenched by the SWNT, which in turn fluoresces in near infrared. Also shown are free Rd-LPE molecules fluorescing under the excitation of the laser beam. No energy transfer occurs in the latter case

Cell Trafficking of Carbon Nanotubes Based on Fluorescence Detection 3.1.2. Thin Layer Chromatography

139

1. Scrap the pink band containing the Rd-LPE off the TLC plate, place it into an Eppendorf tube. 2. Dissolve the pink band with chloroform/methanol solvent (50:50), vortex, and centrifuge for 5 min at 9,030 RCF. 3. Extract the supernatant and place it in an Eppendorf tube to dry under inert gas. Repeat this process until the silica is no longer pink, indicating the successful removal of most Rd-LPE lipids from silica scrapings. 4. Remove a small sample from each of the four bands observed from TLC and run through a mass spectrometer to confirm substances present.

3.1.3. Formation of LPE-SWNT Complex

1. After complete drying, add SWNTs to the Rd-LPE at approximately a weight ratio of 1:5. Then add 300 mL of water and tip-sonicate the solution at 8 W for 15 min (see Note 2). 2. Centrifuge the solution for 5 min at 11,739 RCF and dispose of the supernatant to remove free lipids and SWNT catalysts. A concomitant color change of the solution from pink to gray shall occur when SWNTs are added to Rd-LPE. The mixing of SWNTs and Rd-LPE shall yield a stable and homogeneous solution. The Rd-LPE–SWNT complexes are formed by the interaction between the hydrophobic SWNT surfaces and the hydrophobic tails of the Rd-LPE.

3.1.4. Confirmation of Fluorescence Energy Transfer

3.1.5. Cell Culture

Measure the fluorescence emission of Rd-LPE and Rd-LPE– SWNT (see Note 3). The absorption and emission peaks for Rd-LPE occur at 571 and 583 nm, while the absorption peak for Rd-LPE–SWNT is blue-shifted to 557 nm and the emission peak red-shifted to 590 nm. The quenching of rhodamine fluorescence for Rd-LPE–SWNT indicates effective binding of Rd-LPE and SWNT. 1. Culture breast cancer MCF7 cell line in DMEM with 1% penicillin streptomycin, 1% sodium pyruvate, and 10% fetal bovine serum. 2. Seed approximately 10,000 MCF7 cells in each (200 mL) of an eight-chambered glass slide and allow the cells to attach overnight at 37°C with 5% CO2. 3. Add approximately 10 mL of 2 mg/mL of Rd-LPE–SWNT to each chamber and incubate for 0.5, 1, 2, and 3 h, respectively. 4. After incubation, wash the chambers three times with 1×PBS and keep the cells in PBS for imaging.

140

Lamm and Ke

3.1.6. Confocal Fluorescence Imaging of LPE-SWNT Uptake

3.1.7. Micro-Raman Spectroscopy to Confirm SWNT Uptake

Image control samples of Rd-LPE incubated with MCF7 cells for 3 h and Rd-LPE–SWNT without MCF7 cells using a confocal fluorescence microscope. The first control shall show no uptake of Rd-LPE without SWNTs. The second control without the addition of cells to chamber glass shall show no fluorescence from the Rd-LPE–SWNT complexes. Over time increased rhodamine fluorescence shall be observed for the cells incubated with Rd-LPE–SWNT, suggesting some Rd-LPE molecules are stripped off SWNTs when crossing cell membranes. The cells with 3 h incubation shall show strong fluorescence at various axial focal depths, indicating high Rd-LPE translocation efficiency and the physical separation of SWNTs and Rd-LPE inside the cells (see Note 4). The lack of rhodamine fluorescence in extracellular space suggests the stable binding of Rd-LPE–SWNT complexes prior to their translocation. In contrast, the strong rhodamine fluorescence in intracellular space implies the efficient release of Rd-LPE from SWNT transporters, a feature desirable for gene and drug delivery. The nonuniform distribution of rhodamine fluorescence inside cells further suggests that SWNTs in these areas are beyond proximity (>10 nm) from the Rd-LPE molecules. Since lysophospholipids offer great versatility in biofunctionalization and cell signaling, this method offers an optical indicator for detecting the real-time interaction of SWNTs and biomolecules. 1. Use lysophosphotidylcholine (LPC) instead of LPE to avoid fluorescence overwhelming the much weaker Raman signal. Incubate SWNTs coated by LPC with MCF7 cells for 2 h, followed by thoroughly washing with 1×PBS prior to drying on a silicon substrate. 2. Use a Leica microscope equipped with a 50× dry objective (NA = 0.75) to focus the laser beam onto the sample. Excite the Raman spectra of SWNTs with a 514.5 nm laser line, while maintaining the laser power at 1 mW to avoid sample heating. 3. Collect Raman spectra for pristine SWNTs, SWNT coated by LPC, and SWNT-LPC incubated with MCF7 cells. Search for G-bands at ~1,590/cm for all samples. These bands indicate the presence of isolated SWNTs and confirm the uptake of SWNTs by MCF7 cells (see Note 5).

3.2. Detection of Fullerene Uptake Based on Membrane Compaction

Recently, we have developed a novel technique to monitor the real-time uptake of fullerene C70 across cell membranes. This method relied on imaging cell membranes labeled with lipo philic dye DiIC18 (23). The fullerene molecules were suspended in aqueous solutions by phenolic gallic acid (GA, Fig. 2a), a compound present in red wine, tea, and all plant species and fruits. The binding of C70 and GA is believed to be a result of hydrophobic

Cell Trafficking of Carbon Nanotubes Based on Fluorescence Detection

141

Fig. 2. (a) Chemical structure of a GA molecule. (b) MD simulations of the coating of GA (planar molecules) on a fullerene C70 cluster (spherical molecules) (23)

interaction, enhanced by the pi-stacking between the SP2 electrons of the fullerene and the aromatic rings of the acid (Fig. 2b) (23). The strong fluorescence of DiIC18 at 665 nm, when excited at 644 nm, revealed a remarkable compaction of the cell membranes upon the exposure of the latter to C70–GA, a possible mechanism of explaining cytotoxicity induced by nanomaterials. The complex of C70–GA also gave off green fluorescence at 540 nm when excited with blue light at 488 nm, which enabled our monitoring of the fullerene particles in the intracellular space, in real time. This method circumvents the diffraction limit and offers a robust means of detecting the transport, aggregation, and transformation of nanomaterials in the cellular environment. 3.2.1. Formation of C70–GA Complexes

1. Mix C70 and GA of weight ratio 1:10 in Milli-Q water and tip-sonicate the mixture at 8 W for 30 min. 2. After 3 days of incubation at room temperature, centrifuge the mixture of C70–GA at 6,772 RCF for 3 min to remove large C70 aggregates. 3. Filter the sample through Microcon filters to remove large clumps of C70 particles and dialyze for 12 h using Dispo Dialyzer filters to remove free GA (see Note 6). 4. To determine the concentration of the suspensions, dry each sample using SpeedVac. 5. Dissolve the pellet in aromatic solvent 1,2-dichlorobenzene, which can suspend fullerenes up to 20 mg/mL. The unsuspended nanoparticles separate from the suspended ones and enter the organic phase. 6. Determine the quantity of the unsuspended nanoparticles using a precalibrated absorbance curve for C70 in dichlorobenzene. Obtain this calibration curve at 639 nm.

142

Lamm and Ke

3.2.2. Fluorescence of C70–GA

1. Measure the fluorescence of C70–GA using the emission scan function of the spectrofluorometer (see Note 7). Excite the green fluorescence at 540 nm of C70–GA using blue light at 488 nm. 2. Measure the control of GA dissolved in Milli-Q water. Expect to see minimal fluorescence over the wavelength range of 515–800 nm.

3.2.3. Cell Culture

1. Culture HT-29 colon cancer cells in DMEM with 1% penicillin streptomycin, 1% sodium pyruvate, and 10% fetal bovine serum. 2. Seed approximately 1,000 HT-29 cells in each (200 mL) of an eight-chambered well glass slide and allow the cells to attach overnight at 37°C with 5% CO2. 3. After incubation, wash the chambers three times with 1×PBS and keep the cells in PBS for imaging.

3.2.4. Confocal Fluorescence Imaging of C70 Cellular Distribution

1. After 1 h of incubation with C70–GA, mount the HT-29 cells on a fluorescence microscope. Excite the cells with 488 nm blue light. 2. Scan the cross sections of the cells and stack the images along the z-axis. The green fluorescence indicates the appearance of C70–GA and shall appear on the cell membranes and inside the cells.

3.2.5. Confocal Fluorescence Imaging of Real-Time Uptake of C70

1. Label HT-29 cells with lipophilic DiIC18 which turn the cell membranes into red fluorescent “rings” surrounding the cell interiors (Fig. 3). 2. Add approximately 100 mL of 5% w/v DiIC18 to each well of a chamber glass slide and incubate with cells for 1 h at 37°C with 5% CO2. 3. Perform imaging with confocal fluorescence microscope, now with a He–Ne laser of 633 nm as the excitation source. The final concentration of C70–GA in each chamber glass is ~0.3 mg/ mL (pH = 7.4). 4. Set the time lapse between each consecutive image at 1 or 2 min, depending on the experimenter’s proficiency with sample focusing and image recording. 5. Switching between the fluorescence and bright field mode helps identifying the locations of C70 or C70–GA aggregates with respect to the cell membranes and the z-axis. At the recommended dye concentration, this experiment may last for 1 h without experiencing the issue of significant photobleaching.

3.3 RNA Coated SWNT Cell Trafficking

In the year of 2004, we first examined the diffusion of SWNTs coated by RNA polymer poly(rU) (24). Based on this study,

Cell Trafficking of Carbon Nanotubes Based on Fluorescence Detection

143

Fig. 3. HT-29 cells labeled with DiIC18, with only the membranes shown as “rings”

we further demonstrated the delivery of RNA polymer to MCF7 cells using an SWNT transporter (6). This proof-of-concept experiment fully utilized the sectioning property of confocal fluorescence microscopy to illustrate the delivery of fluorescently labeled poly(rU) onto the cell membranes, in the cytoplasm, and in the cell nuclei. 3.3.1. Cell Culture

1. Deposit approximately 10,000 MCF7 cells in each well of an eight-chambered slide to form a sparsely distributed layer of cells. This ensures good exposure of cell membranes to SWNT-poly(rU) hybrids. 2. Culture cells directly on the chamber glass slide using RPMI-1640 growth medium at 37°C under 5% CO2. Incubate all the cells for 24 h until approximately 60% confluence is achieved. 3. Add 20 mL of PI, PI+poly(rU), and PI labeled SWNTpoly(rU) hybrids (0.05 mL/mL) to the chamber slide. 4. Incubate the cells with the added reagents for 2 h. 5. After incubation, wash the cells twice with 200 mL growth medium and keep cells in 400 mL PBS prior to imaging.

3.3.2. Radioisotope Labeling Assay to Confirm SWNT Uptake

1. Probe-sonicate SWNTs in RPMI-1640 growth medium and incubate with radioactive thymidine [methyl-3H] overnight.

144

Lamm and Ke

2. Collect the radioactively labeled SWNTs by centrifugation and re-suspend the SWNTs in PBS prior to incubation with MCF7 cells. 3. After the incubation, release the cells by trypsin-EDTA and recollect the pellet by centrifugation. 4. Wash the pellet of MCF7 cells thoroughly with PBS buffer twice to remove excess SWNTs and thymidine bound on the cell surfaces. 5. Gather translocation efficiency of radioactively labeled SWNTs based on their scintillation counts read from the MCF7 cells, as opposed to those from the supernatants and the PBS washing buffer, for incubation times of 0.5, 1, 2, and 4 h, respectively. This practice confirms that SWNTs alone can penetrate through cellular membranes with an increased efficiency over time. 3.3.3. Poly(rU) Fluorescence Labeling

1. To visualize SWNT–poly(rU) hybrids, label poly(rU) with propidium iodide (PI). PI is membrane impermeable, and therefore is generally excluded by viable cells. Specifically, incubate PI at a concentration of 0.05 mL/mL in TE buffer with SWNT–poly(rU) hybrids at a volume ratio of 1:20 for 30 min. 2. Remove the excess PI and unbound poly(rU) by cassette dialysis for a total of 36 h with three changes of buffer solution.

3.3.4. Formation of SWNT–Poly(rU) Hybrids

1. Probe-sonicate a mixture of 0.5 mg/mL poly(rU) and 0.125 mg/mL SWNTs in TE buffer (see Note 8). As with the binding of SWNTs and DNA, the hydrophobic nitrogenous bases of poly(rU) are expected to bind to the hydrophobic surface of the SWNTs. The bases of poly(rU) also form p-stacking with the carbon rings on the SWNT surface. This nonspecific binding scheme results in the exposure of the charged phosphate backbone of poly(rU), which aids in the suspension of SWNTs in aqueous solutions. 2. Scanning electron microscopy (SEM) can be used to further confirm the binding of SWNTs and poly(rU). The resolution of the SEM is not too critical since poly(rU) is a fairly large molecule and tends to form loops on its own.

3.3.5. Confocal Fluorescence Imaging of Poly(rU) Delivery by SWNTs

1. Collect the fluorescence from PI in the TRITC channel, and record the background in the bright-field channel of the microscope. 2. Record confocal fluorescence images of the two controls, PI and PI+poly(rU), and expect to see minimal fluorescence for healthy MCF7 cells (appear elongated). Intense fluorescence from the controls indicates unhealthy cells (appear circular).

Cell Trafficking of Carbon Nanotubes Based on Fluorescence Detection

145

Images of MCF7 cells incubated with the SWNT–poly(rU) hybrids for 2 h shall indicate RNA polymer translocation. 3. Scan the z-axis of the microscope over 20 mm and stack the images to locate fluorescence on the cell surfaces and within. The observed fluorescence could also be from PI-labeled poly(rU) released from SWNTs due to the dissociation kinetics and the change of pH between cytosol and nucleus. The released nucleic acids may be used as templates for transfection at a later stage. It shall be pointed out that from the energetic viewpoint, the fluorescence in the nucleus is unlikely to be from free PI or PI dissociated from poly(rU) and then re-intercalated with the host genome (6). The uptake of the SWNTs–poly(rU) is hypothesized to be a result of amphipathic interaction between the cell membranes and the SWNT–poly(rU) hybrids. The lateral diffusion of phospholipids in the cell membranes may mediate this interaction by exposing the hydrocarbon lipid chains to the SWNT–poly(rU) hybrids. Cell mitosis and endocytosis may also play significant roles in the internalization of the SWNT–poly(rU) hybrids. Once the SWNT–poly(rU) hybrids are taken up by the MCF7 cells, they may be stored in the endosomes and/or lysosomes in the cytoplasm. The disposal of these tubes by cellular machinery is unclear. 3.4. Molecular Simulations of CNT and Fullerene Cell Trafficking

Investigation of cell trafficking by computer simulation is an appealing strategy for gaining fundamental, molecular-based insight that complements the experimental approaches described above. There are a few recent examples of molecular simulation studies that model the penetration of CNTs or fullerenes through cell membranes with molecular dynamics simulations. The main pursuit in these modeling studies has been a microscopic description of the mechanism by which carbon nanostructures are taken up by lipid bilayers. Atomistic molecular dynamics simulations have addressed several issues surrounding C60 fullerene interactions with lipid bilayers, including the preferred location of a single fullerene inside the bilayer (24, 25), the effective interactions between two fullerenes within the bilayer (24) and the transport of fullerenes across the bilayer (25, 26). Both Qiao et al. (25) and Bedrov et al. (24, 26) found that there is effectively no energetic barrier for the penetration of pristine C60 fullerenes into the lipid bilayer and that once inside the membrane ,a fullerene’s preferred location is 7–11 Å from the center of the bilayer. This off-center positioning reflects a balance between the fullerene–bilayer interaction energy and the minimal distortion of the lipid structure (24). Qiao and coworkers (25) have also examined the interactions of a functionalized fullerene (C60(OH)20) and found that its penetration is limited to adsorption on the lipid head groups at the surface of the membrane.

146

Lamm and Ke

The simulations with atomistic models discussed above are limited to modeling phenomena that occurs over timescales on the order of tens of nanoseconds. This limitation is due to the computational cost of keeping track of all the atoms in the system. Routinely, system sizes are on the order of 105 atoms. To alleviate the limitations posed by the large system sizes, coarse-grained molecular models have been developed and implemented for modeling lipid bilayers (27). The degrees of freedom (i.e., number of atoms) are reduced in a coarse-grained strategy by grouping selected atoms of the macromolecule into a single effective interaction site. For example, a single chain of a dipalmitoylphosphatidylcholine (DPPC) lipid with 130 atoms has been reduced to 12 coarse-grained sites with the Marrink model (27). Once the degrees of freedom are reduced, the coarse-grained models are capable of modeling phenomena over larger timescales, on the order of 1,000 ns. Using this approach, molecular dynamics simulations with coarse-grained molecular models have examined important aspects of fullerene and CNT penetration in lipid bilayers, including the mechanisms for fullerene and CNT translocation through a cell membrane (28–30), the effect of high fullerene concentration (28), and the effect of CNT functionalization (29). Lopez et al. (29) and Wallace and Sansom (30) found that lipids block the entrance to CNTs, which suggests that attaching the payload to the exterior of the CNT is a more prudent strategy for delivery than encapsulating it within the tube. These studies also demonstrate that the bilayer restores its structure, through the so-called “self-sealing” (30), after the CNT penetrates the membrane. Wong-Ekkabut et al. (28) and Lopez et al. (29) simulated the spontaneous insertion of fullerenes and CNTs, respectively, into a lipid bilayer. These calculations show that penetration by the carbon nanostructures causes little disruption to the bilayer structure. Wong-Ekkabut et al. (28) have further demonstrated that once inserted in the bilayer, an aggregate of fullerenes breaks apart on the time scale of microseconds. The protocols below will describe how to: (1) select a starting structure and prepare the initial configuration, (2) select a force field, (3) equilibrate the system, and (4) conduct postsimulation analysis. For a brief introduction to the theory behind molecular dynamics simulation and an explanation of the relevant terms in force fields, we refer readers to ref. (31). 3.4.1. Preparing the Initial Configuration

There are a limited number of model lipid bilayer structures available for download. For the structures listed here, the coordinates are provided in the Protein Data Bank (PDB) format, which is accepted by widely available molecular dynamics simulation packages, such as LAMMPS (32), NAMD (33), and GROMACS (34). Kartturen (35) provides atomic coordinates for DPPC bilayers, while Tieleman (36) and Feller (37) provide

Cell Trafficking of Carbon Nanotubes Based on Fluorescence Detection

147

united atom coordinates for DPPC, dodecylphosphocholine (DPC), dimyristoylphosphatidylcholine (DMPC), and others. In general, these structures are hydrated and equilibrated, but users should conduct an independent assessment of the suitability of the structure before using it. Protocols have been developed for using an existing model lipid bilayer configuration to build another lipid bilayer with a related chemistry (38). Coordinates for atomistic models of CNTs and fullerenes can be constructed by the user or with web-based software (39). Coordinates for coarse-grained models should be constructed by the user to comply with the chosen coarse-grained force field. As model CNTs or fullerenes are placed in the lipid bilayer structure, care should be taken to avoid atom overlaps between the carbon nanostructure and the bilayer. 3.4.2. Selecting a Force Field

The first choice one must make is whether the problem under investigation is better suited to an atomistic or a coarse-grained model. An estimate of the length and time scales of the phenomena to be modeled will dictate the class of force field that is suitable. Generally, simulations with atomistic force fields can model processes on the order of tens of nanoseconds and they are best used for situations where the CNT or fullerene is placed within the lipid bilayer to collect information on position-dependent interactions. Simulations with coarse-grained force fields can model processes on the order of microseconds, and thus, are capable of addressing slow or rare events, such as the spontaneous insertion of a CNT in a lipid bilayer. The disadvantage of the coarse-grained approach is that the resulting acceleration of the dynamics can obscure the true time scale covered by the simulation (40). Consulting recent examples from the literature (24–30) can give guidance for force field selection. Atomistic and coarse-grained force fields for lipid chains, CNTs, and fullerenes are available in the literature (see examples cited above and references therein).

3.4.3. Equilibration Simulation

The starting configuration generated above should be brought to an equilibrium state before collecting trajectories for analysis. This is achieved by running molecular dynamics for a set number of time steps before trajectory data are saved for analysis. Depending on the number of high-energy contacts in the starting configuration, the time step may need to be decreased at the beginning of the equilibration simulation in order to “ease” the system into a more favorable state. The progress of the system toward equilibration can be assessed by monitoring quantities such as the potential energy or constraint force (26) for convergence during the simulation.

3.4.4. Production Simulation and Analysis of Results

After the system has equilibrated, a production simulation can be started. The length of this production run will be necessarily longer than the length of the equilibration run. The user must decide

148

Lamm and Ke

how often to collect configurations from the trajectory as most properties of the system are calculated postsimulation. Typical quantities that can be obtained from the simulation include the free energy profile (26, 28), time-dependent position of the carbon nanostructure (24, 28), position-dependent diffusion coefficient of the carbon nanostructure (26, 28), and the potential of mean force (25). 3.4.5. Multiscale Simulation Methods

There is an increasing trend toward developing multiscale molecular models and simulation algorithms that allow for the representation of the system in both the atomistic and coarsegrained framework (41, 42). These advanced techniques should be especially appealing for modeling CNT and fullerene cell trafficking, because the close coupling between the atomistic and coarse-grained descriptions would provide direct access to the real-time scales involved in spontaneous insertion, aggregation, and dispersion of carbon nanostructures within a lipid bilayer.

4. Notes 1. SWNTs may be directly purchased from commercial sources such as Nano-C, Inc. or Carbon Nanotube Solutions. Avoid the use of “As Produced SWNTs” since these tubes have a low purity of ~25%. For best solubility with nucleic acid use “Pre-Purified SWNTs” which have ~75% purity. The “Purified SWNTs” typically do not bind as well to nucleic acid due to their fully functionalized/acid treated surfaces. 2. The stability of the LPE–SWNT complex not only depends on the binding affinity of the Rd-LPE for the SWNTs, but the free Rd-LPE in the suspension also plays a role in maintaining the equilibrium of the system. In order to protect the rhodamine label on the LPE excess probe sonication at high power should be avoided. 3. In addition to the spectrofluorometer listed in Subheading 12.2.4, such measurement may also be conducted using a Varian Cary Eclipse Fluorescence Photometer. The measurement of Rd-LPE fluorescence is straightforward. The measurement of Rd-LPE–SWNT fluorescence needs to look for anticorrelation between the emission at 583 and 590 nm. 4. For this imaging experiment, it is recommended to switch on brightfield mode, fluorescence mode for Rd-LPE, and overlaid brightfield and fluorescence mode. Use the highest numerical aperture of your objective set to maximize image resolution.

Cell Trafficking of Carbon Nanotubes Based on Fluorescence Detection

149

5. Nonfluorescent LPC instead of Rd-LPE is used to avoid overwhelming the Raman signals with fluorescence. Depending on the sensitivity of the Raman spectrometer, the G-bands at ~1,590/cm might shift slightly due to the tight binding between the weakly negatively charged LPE and the SWNTs, similar to that reported for the binding of single-stranded DNA to SWNTs. 6. This step needs to be taken with caution. Depending on how well the C70 particles are suspended by the GA, Microcon filters with different MWCOs may be attempted. Also the removal of free GA can be performed with DispoDialyzer filters multiple times to ensure better results. 7. The fluorescence of C70–GA is rather weak and its origin is unclear. The emission scans for C70–GA need to be examined close to 540 nm, say between 500 and 600 nm. A green color can be seen from GA alone in water over a period of 1–2 days. However, such rather misleading green color does not give off a green spectrum as confirmed by the spectrofluorometer. 8. For the purpose of concept demonstration, the probe sonication may be done for 20–30 min, at 8 W. For the practice of gene delivery with an SWNT transporter where transfection is desired at a later stage such sonication must be avoided and replaced with gentle shaking of SWNTs and nucleic acids overnight. Also the SWNTs used must be autoclaved prior to their mixing with the nucleic acids.

Acknowledgments Ke thanks Drs. Emppu Salonen, Ilpo Vattulainen, Xi Wang, Yonnie Wu, and his students Sijie Lin and Jessica Moore for contributions instrumental to this presentation. The authors acknowledge financial support from NSF grants CBET-0736037 (PCK), CAREER CBET-0744040 (PCK), CBET-0403864 (MHL), and OCI-0749156 (MHL). References 1. Mitchell DT, Lee SB, Trofin L, Li N, Nevanen TK, Söderlund H, Martin CR (2002) Smart nanotubes for bioseparations and biocatalysis. J Am Chem Soc 124:11864–11865 2. Wang J, Liu G, Jan MR (2004) Ultrasensitive electrical biosensing of proteins and DNA: carbon-nanotube derived amplification of the

recognition and transduction events. J Am Chem Soc 126:3010–3011 3. Ke PC, Qiao R (2007) Carbon nanomaterials in biological systems. J Phys Condens Matter 19(373101):1–25 4. Kam NWS, Jessop TC, Wender PA, Dai HJ (2004) Nanotube molecular transporters:

150

Lamm and Ke

internalization of carbon nanotube-protein conjugates into mammalian cells. J Am Chem Soc 126:6850–6851 5. Pantarotto D, Singh R, McCarthy D, Erhardt M, Braind J-P, Prato M, Kostarelos K, Bianco A (2004) Functionalized carbon nanotubes for plasmid DNA gene delivery. Angew Chem Int Ed Engl 43:5242–5246 6. Lu Q, Moore JM, Huang G, Mount AS, Rao AM, Larcom LL, Ke PC (2004) RNA polymer translocation with single-walled carbon nanotubes. Nano Lett 4:2473–2477 7. Kostarelos K, Lacerda L, Pastorin G, Wu W, Wieckowski S, Luangsivilay J, Godefroy S, Pantarotto D, Briand J-P, Muller S, Prato M, Bianco A (2007) Cellular uptake of functionalized carbon nanotubes is independent of functional group and cell type. Nat Nanotechnol 2:108–113 8. Liu Z, Davis C, Cai W, He L, Chen X, Dai H (2008) Circulation and long-term fate of functionalized biocompatible single-walled carbon nanotubes in mice probed by Raman spectroscopy. Proc Natl Acad Sci USA 105: 1410–1415 9. Poland CA, Duffin R, Kinloch I, Maynard A, Wallace WAH, Seaton A, Stone V, Brown S, MacNee W, Donaldson K (2008) Carbon nanotube introduced into the abdominal cavity of mice show asbestos-like pathogenicity in a pilot study. Nat Nanotechnol 3(7):423–428 10. Zhu L, Chang DW, Dai L, Hong Y (2007) DNA damage induced by multiwalled carbon nanotubes in mouse embryonic stem cells. Nano Lett 7:3592–3597 11. Lewinski N, Colvin V, Drezek R (2008) Cytotoxicity of nanoparticles. Small 4:26–49 12. Scheringer M (2008) Environmental risks of nanomaterials. Nat Nanotechnol 3:322–323 13. Tsyboulski DA, Bachilo SM, Weisman RB (2005) Versatile visualization of individual single-walled carbon nanotubes with nearinfrared fluorescence microscopy. Nano Lett 5:975–979 14. Heller DA, Baik S, Eurell TE, Strano MS (2005) Single-walled carbon nanotube spectroscopy in live cells: toward long-term labels and optical sensors. Adv Mater 17: 2793–2799 15. Stryer L, Haugland RP (1967) Energy transfer: a spectroscopic ruler. Proc Natl Acad Sci USA 58:719–726 16. Ha T, Enderle Th, Ogletree DF, Chemla DS, Selvin PR, Weiss S (1996) Probing the interaction between two single molecules: fluorescence resonance energy transfer between a single

donor and a single acceptor. Proc Natl Acad Sci USA 93:6264–6268 17. Bachilo SM, Strano MS, Kittrell C, Hauge RH, Smalley RE, Weisman RB (2002) Structureassisted optical spectra of single-walled carbon nanotubes. Science 298:2361–2366 18. Guldi DM, Marcaccio M, Paolucci D, Paolucci F, Tagmatarchis N, Tasis D, Vazquez E, Prato M (2003) Single-wall carbon nanotube–ferrocene nanohybrids: observing intramolecular electron transfer in functionalized SWNTs. Angew Chem Int Ed Engl 42:4206–4209 19. Zhu W, Minami N, Kazaoui S, Kim YJ (2003) Fluorescent chromophore functionalized single-wall carbon nanotubes with minimal alteration to their characteristic one-dimensional electronic states. Mater Chem 13:2196–2201 20. Wu Y, Lu Q, Hudson JS, Mount AS, Moore JM, Rao AM, Alexov E, Ke PC (2006) Coating single-walled carbon nanotubes with lysophospholipids. J Phys Chem B 110:2475–2478 21. Qiao R, Ke PC (2006) Lipid-carbon nanotube self-assembly in aqueous solution. J Am Chem Soc 128:13656–13657 22. Lin S, Keskar G, Wu Y, Wang X, Mount AS, Klaine SJ, Moore JM, Rao AM, Ke PC (2006) Detection of phospholipid-carbon nanotube translocation using fluorescence energy transfer. Appl Phys Lett 89(143118):1–3 23. Salonen E, Lin S, Reid ML, Allegood MS, Wang X, Rao AM, Vattulainen I, Ke PC (2008) Real-time translocation of fullerene reveals cell contraction. Small 4(11):1986–1992 24. Li L, Davande H, Bedrov D, Smith GD (2007) A molecular dynamics simulation study of C60 fullerenes inside a dimyristoylphosphatidylcholine lipid bilayer. J Phys Chem B 111: 4067–4072 25. Qiao R, Roberts AP, Mount AS, Klaine SJ, Ke PC (2007) Translocation of C60 and its derivatives across a lipid bilayer. Nano Lett 7:614–619 26. Bedrov D, Smith GD, Davande H, Li L (2008) Passive transport of C60 fullerenes through a lipid membrane: a molecular dynamics simulation study. J Phys Chem B 112:2078–2084 27. Marrink SJ, de Vries AH, Mark AE (2004) Coarse grained model for semiquantitative lipid simulations. J Phys Chem B 108: 750–760 28. Wong-Ekkabut J, Baoukina S, Wannapong T, Tang IM, Tieleman DP, Monticelli L (2008) Computer simulation study of fullerene

Cell Trafficking of Carbon Nanotubes Based on Fluorescence Detection translocation through lipid membranes. Nat Nanotechnol 3:363–368 29. Lopez CF, Nielsen SO, Moore PB, Klein ML (2004) Understanding nature’s design for a nanosyringe. Proc Natl Acad Sci USA 101: 4431–4434 30. Wallace EJ, Sansom MSP (2008) Blocking of carbon nanotube based nanoinjectors by lipids: a simulation study. Nano Lett 8:2751–2756 31. Lindahl ER (2008) Molecular Dynamics. Methods Mol Biol 443:3–23 32. http://www.ks.uiuc.edu/Research/namd/ 33. http://lammps.sandia.gov/ 34. http://www.gromacs.org/ 3 5. Kartturen, http://www.apmaths.uwo.ca/ ~mkarttu//downloads.shtml 36. Tieleman, http://moose.bio.ucalgary.ca/

151

37. Feller, http://persweb/wabash.edu/facstaff/ fellers/ 38. Dickey A, Faller R (2008) Examining the contributions of lipid shape and headgroup charge on bilayer behavior. Biophys J 95:2636–2646 39. http://www.jcrystal.com/products/wincnt/ index.htm (September, 2008) 40. Aksimentiev A, Schulten K (2004) Extending molecular modeling methodology to study insertion of membrane nanopores. Proc Natl Acad Sci USA 101:4337–4338 41. Ross RB, Mohanty S (2008) Multiscale simulation methods for nanomaterials. Wiley, Hoboken, NJ 42. Izvekov S, Voth G (2005) A multiscale coarse-graining method for biomolecular systems. J Phys Chem B 109:2469–2473

Chapter 13 Carbon Nanotubes as Intracellular Carriers for Multidrug Resistant Cells Studied by Capillary Electrophoresis–Laser-Induced Fluorescence Ruibin Li, Hanfa Zou, Hua Xiao, and Renan Wu Abstract Fluorescently labeled carbon nanotube probes (CNTP) are prepared by derivatizing oxidized (o)-MWNTs with a fluorescein dye. Capillary electrophoresis coupled with laser-induced fluorescence (CE–LIF) detection is used to separate and detect CNTP in multidrug-resistant cells (K562A) and the parent cells (K562S). CE–LIF and flow cytometry investigation reveal that the CNTP can traverse the membranes in both cell lines without being pumped out by P-glycoprotein. The CE-LIF method is also useful for quantitative analysis of CNT in single cells, enabling drug delivery and multidrug resistance (MDR) studies. Moreover, toward quantifying the intracellular uptake of oxidized (o)-SWNTs with anchored Rhodamine123 (Rho123), fluorescence-quenching of Rho123 is measured by micellar electrokinetic chromatography coupled with LIF detection. Enhanced uptake of Rho123 in multidrug-resistant leukemia cells can be achieved with the aid of the o-SWNTs carriers. Besides being able to overcome MDR, o-SWNTs are shown to be excellent intracellular carriers possessing large adsorption capacity and prolonged release ability. Finally, it is demonstrated that o-SWNTs are safe for biological applications at concentrations of up to 40 µg/mL. Key words: Multidrug resistance, Capillary electrophoresis, Fluorescence detection, Drug deliver, Cell analysis

1. Introduction Multidrug resistance (MDR), the ability to withstand a wide variety of structural and functional distinct drugs or chemicals, has become a major hurdle in cancer chemotherapy (1). The most common mechanism of MDR is the overexpression of membrane proteins that pump the anticancer agents out of the cells, thus decreasing the intracellular drug concentration (2). One of the best-known MDR-related membrane proteins is the P-glycoprotein (P-gp), which is more strongly expressed in leukemia cells with MDR than K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_13, © Humana Press, a part of Springer Science + Business Media, LLC 2010

153

154

Li et al.

in sensitive leukemia cells (3). To overcome the MDR, a great diversity of experimental and clinical efforts have been directed toward inhibiting the drug efflux activity of P-gp in cancer cells, such as combining surfactants and P-gp inhibitors with anticancer drugs (4, 5). However, these methods did not find clinical application due to damages resulting from the high applied drug concentrations. Functionalized carbon nanotubes (f-CNTs) have sparked enormous enthusiasm in recent years (6). They have been utilized to transport various types of molecules including peptides, amino acids, proteins, nucleic acids, and other drugs into living cells. In most cases, covalent bonding of macro-biomolecules was employed in order to firmly attach the molecular cargos to f-CNTs (7–12). However, major problems arise due to the loss of biological activity of the cargo, and the difficulty to release it from the carrier. In these respects, the noncovalent adsorption approach offers advantages for preparing drug-carrier conjugates, in particular by ensuring a more favorable drug release from the carrier. Carbon nanotubes (CNTs) have emerged as promising drug carriers that are able to overcome MDR. However, the so far investigated small molecules attached to f-CNTs are usually fluorescent, and cannot be directly detected because of fluorescence quenching on CNTs (13, 14). Accordingly, alternative detection methods are needed. One promising analytical tool is capillary electrophoresis coupled with laser-induced fluorescence (CE–LIF) (15, 16). In the past decades, CE–LIF was frequently applied in bioanalysis (17) and clinical analysis (18). It offers a range of advantages, including (1) small sample volumes, (2) high separation efficiency and low detection limits, (3) biocompatible environment for cell analysis, (4) various separation modes, etc. In the present work, carbon nanotube probes (CNTP) are prepared by conjugating short oxidized multi-walled carbon nanotubes (o-MWNTs) with fluorescein isothiocyanate (FITC), which enables determining their distribution in multidrug resistant (K562R) and sensitive (K562S) leukemia cells. The intracellular uptake of CNTP in K562R and K562S cells is measured by CE–LIF in both, single cells and cell lysate. Moreover, oxidized single-walled carbon nanotubes (o-SWNTs) are employed as intracellular carriers for delivery of Rhodamine123 (Rho123) to K562R cells. The intracellular uptake of Rho123 carried by o-SWNTs is quantified via CE–LIF.

2. Materials 2.1. Carbon Nanotubes and Functionalizing Agents

1. MWNTs (Professor F. Wei, Tsinghua University, Beijing, China). 2. SWNTs (Sino-nano Company, China).

Carbon Nanotubes as Intracellular Carriers

155

3. 1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC) (Alfa Aesar, England). 4. N-hydroxysuccinimide (NHS) (Sigma-Aldrich, USA). 5. Fluorescein isothiocyanate (FITC) (Fluka, Switzerland). 6. 2-Morpholin-4-yl-ethanesulfonic acid (MES) buffer: 0.1 M, pH 6.0. 7. Na2CO3–NaHCO3 buffer: 50 mM, pH 9.4. 8. Rhodamine123 (Rho123) (Sigma-Aldrich, USA). 2.2. Cell Culture and Cell Sample Preparation

1. Culture medium: RPMI-1640 and 10% fetal calf serum. 2. TBS buffer for cell lysis: 50 mM Tris–boric acid, 10 mM SDS, pH adjusted to 7.4. 3. PBS buffer: 0.2 g KCl, 0.2 g KH2PO4, 8 g NaCl, and 1.14 g Na2HPO4 in 1 L H2O, pH adjusted to 7.4.

2.3. CE–LIF Analysis

1. Running buffer for CNTP analysis: 50 mM Tris–boric acid, 0.5% SDS, pH 9.4. 2. Running buffer for Rho123 analysis: 50 mM Tris–boric acid, 0.5% SDS, pH 7.4. 3. Running buffer for CZE mode: 50 mM Tris–boric acid, pH 7.4. 4. Fused-silica capillary for CNTP analysis: 365 mm o.d. × 25 mm i.d. × 50/65 cm long (Polymicro, USA). 5. Capillary for Rho123 analysis: 365 mm o.d. × 50 mm i.d. × 15/65 cm long (Yongnian Optic Fiber Plant, Hebei, China).

2.4. Flow Cytometry

1. Annexin V-FITC (BD Biosciences, NJ, USA). 2. Propidium iodide (BD Biosciences, NJ, USA).

2.5. Adsorption and Release Experiments In Vitro

1. Fetal bovine serum (FBS) and PBS (1/1 v/v). 2. Fluorescence spectrophotometer (FS) F-2500 (Hitachi, USA).

3. Methods o-MWNTs were derivatized with FITC to detect their intracellular distribution in K562R and K562S cells. In order to completely analyze CNTP distribution differences between K562R and K562S cells, CE–LIF and FCM were employed to detect the intracellular uptake of CNTP. Because CNTs could overcome MDR and cross cell membrane into K562R cells, o-SWNTs were used as intracellular carriers to load Rho123 by adsorption. Additionally based on the detection

156

Li et al.

method developments on fluorescence-quenched fluorophores, the uptake of Rho123 in K562 cells carried by o-SWNTs was measured by CE–LIF. Besides the loading capacity, release ability and biology safety are important characters for drug carriers. All of them were tested, and it was found o-SWNTs were excellent intracellular carriers for overcoming MDR. 3.1. Preparation of CNTP

1. o-MWNTs were prepared by the literature method (19). It involved a reaction of CNTs with mixed acid by stirring under reflux at 120°C for 30 min (see Note 1). The o-MWNTs in aqueous solution were centrifuged at 4,500 × g to remove any large unreacted CNT from the solution. MWNTs and o-MWNTs were characterized by transmission electron microscopy (TEM, JEM-2000EX) at 120 kV. Figs. 1a and b show the TEM images of MWNTs and o-MWNTs, respectively. 2. o-MWNTs were then reacted with hexamethylenediamine in the presence of EDC to afford a linker between the CNT and the subsequent fluorescent probe. Briefly, 2 mg o-MWNTs were mixed with 5 mg hexamethylenediamine and 1 mg EDC in MES buffer. The mixed solution was stirred at room temperature for 2 h. The CNTs conjugated to hexamethylenediamine (CNT–CONH(CH2)6NH2) were obtained by centrifugation at 44,000 × g (see Note 2). 3. Subsequent reaction of CNT–CONH (CH2)6NH2 with FITC was performed in Na2CO3–NaHCO3 buffer for 24 h at room temperature, and it resulted in the formation of fluorescently labeled CNTP. Excess FITC was removed by rinsing with PBS and centrifuging three times at 25,000 rpm (see Note 3). The TEM image of CNTP is shown in Fig. 1c.

Fig. 1. TEM images of (a) original MWNTs, (b) o-MWNTs, and (c) fluorescently labeled CNTP. The scale bar corresponds to 200 nm. The arrow in (c) indicates a short oxidized multi-walled carbon nanotube with approximately 20 nm in diameter and 150 nm long. The o-MWNTs seem to be thinner and shorter because the exterior walls of the MWNT were oxidized, resulting in shortened CNT, especially for those after centrifugation. Reproduced with permission from ref. (22)

Carbon Nanotubes as Intracellular Carriers

3.2. Preparation of Rho123–SWNTs Conjugates

157

1. o-SWNTs were prepared following the same protocol as for the o-MWNTs. 2. o-SWNTs solution was centrifuged at 30,000 × g to remove any large unreacted SWNTs, the supernatant liquid was collected and dried. 3. The Rho123–SWNTs conjugates were prepared by ultrasonically mixing 2 mg o-SWNTs and 20 µg Rho123 into 1 mL H2O at room temperature for 1 h. The prepared conjugates solution was stored at 4°C in dark before use.

3.3. Analysis of CNTP via CE–LIF

1. A laboratory-built CE–LIF system described elsewhere (20) was used in this study, with an Ar-ion laser (l = 488 nm) operated at 6 mW for excitation, and fluorescence detected at l = 515 nm. A valve connected to a hydraulic differential at the capillary outlet serves to generate a pulse of siphoning force for injection of individual cells while under obser vation using an inverted microscope (see Note 4). A micromanipulator was used to control and move the injection capillary. 2. To reduce electro-osmotic flow and capillary wall adsorption, a polyacrylamide-coated capillary was used in all the experiments. Because CNTP was negatively charged, micellar electrokinetic chromatographic analysis was conducted under negative polarity. For the capillary wall modification, 50% g-methacryloxypropyltrimethoxysilane (g-MAPS) in methanol was used (see Note 5). Four percent acrylamide solution containing 0.1% ammonium persulfate and 0.1% N,N,N ¢, N ¢-tetramethylethylenediamine (TEMED) was employed as coating reagent for capillary – see Note 6), in accordance with the procedures described by Hjertén (21). 3. CNTP standard solution (5 µg/mL) was injected into capillar y by applying a pressure with 10.8 kPa for 5 s. Then the sample was analyzed in running buffer under 153 V/cm separation potential (see Note 7). The electropherogram obtained from CNTP is compared with that obtained from FITC in Fig. 2.

3.4. Analysis of CNTP in Cells via CE–LIF

1. K562R and K562S cells were incubated in 175-cm2 vented flasks at 37°C and 5% CO2 and split every 2–3 days. In the whole experiments, the cells were treated in the logarithmic growth phase and the passage number of leukemia cells was less than 15. 2. Both kinds of cells were washed three times with PBS to remove residues from the medium. The cell number of the two samples was measured by a cell counting plate and controlled with equal number.

158

Li et al.

Fig. 2. Electropherograms obtained from (a) 5-mg/mL CNTP and (b) 10−7-M FITC. The electropherogram obtained from CNTP is compared with that obtained from FITC. The pattern of CNTP is quite different from that of FITC, which shows that CNTP has quite unique electrophoretic behavior and a broad separation window under the electrophoresis conditions used. The SDS in the separation buffer and samples improved the analysis reproducibility. Relative standard deviations (RSD) of the highest peak of CNTP from three consecutive injections were 1.20% for migration time and 7.14% for peak area measurements. Reproduced with permission from ref. (22)

3. After the two kinds of cells were incubated with 25 µg/mL CNTP in PBS for 1 h, the cell samples were rinsed with PBS for three times. 4. Single cell injection for CE–LIF has been described elsewhere (20). The injection end of the capillary was etched. While under observation by inverted microscopy, individual cells were injected into the capillary by applying a pulse of suction at its detection end. Inside the capillary, the single cell was lysed by the CE running buffer and analyzed under separation potential. The electropherograms obtained from a single K562A cell and K562S cell are shown in Fig. 3a. 5. Approximately 2 × 105 cells that had been incubated with 25 mg/mL CNTP at 37°C for 1 h were lysed by treatment with 50 mL TBS buffer. An aliquot of the lysate was injected by siphoning and analyzed by CE-LIF (10.8 kPa for 5 s). The electropherograms obtained from both K562S and K562A cells incubated with CNTP and from K562A cells not incubated with CNTP are shown in Fig 3b.

Carbon Nanotubes as Intracellular Carriers

159

Fig. 3. Electropherograms obtained from CE-LIF analysis of (a) CNTP in a single cell and (b) CNTP in lysate (cell blank was lysate from K562A). The distribution of CNTP in K562 cells was also analyzed at the single-cell level to characterize their unique influx. The total CNTP peak area is similar for the two cells although the intensities of individual peaks vary. For cell lyzate analysis, the ratio of average total peak area for K562S to that for K562A is 1.05 ± 0.06 (n = 3, RSD = 6%). These results imply that uptake levels of CNTP by both K562S and K562A cells are similar, even though different levels of P-glycoprotein are present in the two cell lines. Reproduced with permission from ref. (22)

3.5. Measurement of CNTP in Cells via FCM

1. As described before, both kinds of cells were incubated with CNTP (25 µg/mL) at 4°C and 37°C for 1 h, respectively. Then the cells were rinsed with PBS at 4°C for three times.

160

Li et al.

2. The prepared cell samples were re-suspended in 500 mL PBS, and the cell number was controlled to ca. 1.0 × 105. After that, both K562A and K562S cells were characte rized by FCM for the presence of fluorescent CNTP (Fig. 4a and b). The corresponding statistical histograms are shown in Fig. 4c.

Fig. 4. Measurement of CNTP in K562A and K562S cells (a) incubated at 4°C, (b) incubated at 37°C, and (c) the corresponding histograms comparing relative fluorescence intensity in the K562A and K562S cells. There is no significant difference between the distributions of CNTP in the two cell lines after incubation at 4°C. The patterns also show that a smaller amount of CNTP is present in both cell lines than after incubation at 37°C. After incubation at 37°C the difference is only small. These results indicate that the incubation temperature affects CNTP distribution in both cell lines. The result suggests CNTP may traverse cellular membrane by endocytosis and penetration and this is consistent

Carbon Nanotubes as Intracellular Carriers

161

Fig. 4. (continued) with results reported elsewhere (10). The ratio of the average fluorescence intensity of CNTP in K562S and K562A is 1.22 ± 0.02 (n = 3, RSD = 2%), which is consistent with the results (ratio = 1.05 ± 0.06, n = 3, RSD = 6%) obtained by CE–LIF analysis for cell lyzate. Considering that the multidrug-resistant K562A cells expressed much more P-glycoprotein (P-gp) than the parent K562S cells, the similar amounts of CNTP observed in both cell lines suggest that CNTP is not pumped out by the P-gp expressed in K562A cells. Reproduced with permission from ref. (22)

3.6. Method Set-Up for FluorescenceQuenched Fluorophores Detection

1. For the fluorescence-quenched Rho123 detection, the fluorescence-quenched solution samples were prepared by adding Rho123 into 0, 0.5, 2, 5, and 7 µg/mL o-SWNTs solution, respectively. And the Rho123 concentrations for all of the five solutions were kept 10 ng/mL, respectively. Then the emission spectrum of Rho123 in five samples was measured (Ex 488 nm) by fluorescence spectrophotometer (FS). The emission spectra of the samples are shown in Fig. 5a. 2. Different CE modes were used to analyze fluorescencequenched Rho123. The fluorescence-quenched Rho123 sample containing 7 µg/mL o-SWNTs and 10 ng/mL Rho123 was analyzed by capillary zone electrophoresis (CZE) and micellar electrokinetic capillary chromatography (MEKC) with LIF detection. The running buffers for CZE and MEKC modes are TB and TBS, respectively. Both electropherograms obtained from the two modes are compared (Fig. 5b). 3. For setting up analytical method for detection of fluorescencequenched Rho123, Rho123 was added into, 5, 10, 20, 30, 40, and 60 µg/mL o-SWNTs solutions, respectively and the concentrations of Rho123 in all samples kept 10 ng/mL. The fluorescence intensity of Rho123 in the seven samples was monitored by MEKC-LIF (Fig. 6a), and each of the seven solution samples was analyzed three times. Then the calibration curves were obtained by analyzing Rho123

162

Li et al.

Fig. 5. Detection of fluorescence-quenched Rho123 by (a) FS and (b) CE–LIF with CZE and MEKC modes. As shown in (a), the fluorescence intensity of Rho123 dramatically decreases with the addition of o-SWNTs. The fluorescence of Rho123 is almost completely suppressed while the concentration of o-SWNTs exceeding 7 µg/mL. The fluorescence quenching phenomenon of Rho123 on o-SWNTs will result in the difficulty of monitoring Rho123 adsorbed on o-SWNTs. CE–LIF was attempted to resolve that problem, two separation modes including CZE and MEKC were chosen to achieve the goal. As can be seen in (b), the conjugates of Rho123 and o-SWNTs have not been destroyed only by the differences of electrophoretic mobility in CZE mode. However, the fluorescence of Rho123 can be well detected during the MEKC mode. The destruction of the conjugates by SDS in running buffer and the hydrophobic difference between Rho123 and o-SWNTs are the key factors to the successful analysis, both of them cooperate with each other to resume the quenched fluorescence completely during MEKC-LIF analysis

standards by MEKC-LIF. In detail, three replicates were run for each of five Rho123 solutions (20 µg/mL o-SWNTs) with different concentration ranging from 0.5 to 50 ng/mL. (Fig. 6b). 3.7. Cell Sample Preparation and CE–LIF Analysis of Rho123

1. At first, for detecting the intracellular uptake of Rho123, Rho123 was added into the culture media (200 ng/mL) of K562R and K562S cells, respectively and incubated for 1 h. Then Rho123-treated cells were separated from the medium by centrifugation (Allegra 64R centrifuge, Benckman Coulter, Fllerton, CA, USA) on 1,000 × g at 4°C for 5 min and followed by three rinses with PBS (4°C). 2. The two cell samples were resuspended in 2 mL PBS. The cell number in the two samples was measured by a cell counting plate and was controlled to ca. 4.0 × 106. 3. Both kinds of cells were collected by centrifugation, and followed by an addition of 500 mL TBS buffer into the two samples. After the cells were completely lysed via a freezethaw method for three times, the lysis was centrifuged at 1,000 × g for 5 min.

Carbon Nanotubes as Intracellular Carriers

163

Fig. 6. (a) The electropherogram of Rho123 at various concentrations of o-SWNTs and (b) the standard curve of Rho123 in the presence of o-SWNTs. The intra-RSD values for migration time and peak area analysis in each one of seven samples are less than 2 and 7%, respectively. The influence of o-SWNTs to Rho123 detection was evaluated by the Student t-test. And it was found that the quantification of Rho123 via MEKC-LIF would not be affected by the concentration of o-SWNTs (p > 0.05). The standard curve of Rho123 in the presence of o-SWNTs was also measured, and the linear range of the standard curve is from 0.5 to 50 ng/mL

164

Li et al.

Fig. 7. Electropherograms of Rho123 by cell lyzate analysis in (a) K562R cells and (b) K562S cells. Due to the pumping out function of P-gp in K562R cells, the uptake of Rho123 in K562R cells is significantly lower (ca. 4.3-fold low, p < 0.001) than that in K562S cells. While o-SWNTs are employed as the carriers of Rho123, there is no significant difference in intracellular uptake of Rho123 between K562S and K562R cells (p > 0.05). The results indicate that Rho123 carried by o-SWNTs can be well transported into multidrug-resistant cells

4. For intracellular quantification of Rho123 in the existence of o-SWNTs as its carriers, Rho123–SWNTs conjugates were added into the culture media (200 ng/mL Rho123, 20 µg/mL o-SWNTs) instead of Rho123 as described above (steps 1–3). 5. Fresh cell samples were done on the day of the analysis and the cell lysate was not stored for reuse after more than 12 h. The lysate was injected and analyzed in capillary just as before. Fig. 7a and b show the electrophoregrams. Each one of the five replicates was analyzed for three times. 6. The four cell samples were prepared simultaneously with same passage cells, and all of these samples were replicated five times with different passage cells. 3.8. Adsorption and Release Experiments In Vitro

1. Batch adsorption experiments were carried out by adding 2 mg/mL of o-SWNTs solution (125 mL) into 1.875 mL H2O with varied concentration of Rho123 from 10 ng/ mL to 5 mg/mL. The mixture was then shaken with a rate of 150 rpm/min at 25°C for 1 h. After that, the content of Rho123 in solution is quantified, and the isotherm adsorption curve is shown in Fig. 8a. 2. For release measurements in vitro, 1 mg o-SWNTs were dispersed in 5 mL H2O containing 100 µg/mL Rho123. After shaken for 1 h at 25°C, the solution was filtrated. The conjugates of o-SWNTs and Rho123 were redissolved in 2 mL dialyzate. The suspension was dialyzed in 5 mL of FBS/PBS mixture at 37°C for 4 days (see Note 8).

Carbon Nanotubes as Intracellular Carriers

165

Fig. 8. Adsorption and release of Rho123 on o-SWNTs. (a) Adsorption isotherm and (b) release profile in vitro. The saturated adsorption capacity of Rho123 on o-SWNTs reaches 30 µg/mg. Furthermore, the adsorbed molecules can be slowly released from the o-SWNTs. Within the first 24 h, nearly 20% of Rho123 is released. The results by this experiment indicate that o-SWNTs have great potential to be used as the carriers of small molecular drugs with desirable loading capacity and prolonged release ability

The release of Rho123 was measured and the dialyzate was refreshed every 2 h. The cumulative release amounts of Rho123 divided by the saturated adsorption capacity is the cumulative release ratio. Fig. 8b is the release profile of Rho123 on o-SWNTs. 3.9. Evaluation of Biological Safety

1. K562S cells were incubated with 0, 10, 20, and 40 mg/mL o-SWNTs for 1 h, respectively. And then followed by a washing procedure, the cells were incubated for an additional 60 min and diluted to ca. 1.0 × 105 cells per tube. 2. After the addition of AnnexinV-FITC and propidium iodide, each tube was incubated in dark for another 15 min, and the resulting mixture was diluted to a final volume of 500 mL for the further apoptosis test by FCM. The result of cytofluorimetric analysis for K562S cells treated with o-SWNTs is shown in Fig. 9.

4. Notes 1. After oxidization, water was added into the products till upto ten times of dilution. Then the mixture was filtrated and rinsed. 2. Before reacting with FITC, CNT–CONH(CH2)6NH2 should be rinsed with 50 mM MES buffer containing 2 mM NHS.

166

Li et al.

Fig. 9. Flow cytometry analysis of K562S cells incubated with (a) 0 mg /mL o-SWNTs (94.69% survival), (b) 10-mg/mL o-SWNTs (94.24% survival), (c) 20-mg/mL o-SWNTs (94.85% survival), and (d) 40-mg/mL o-SWNTs (91.93% survival). It was observed that the cytotoxicity of o-SWNTs is unremarkable by calculating the ratio of living cells even though the concentration of o-SWCNT in culture media was up to 40 mg/mL

3. The FITC in the supernatant liquid was detected by fluorescence spectrophotometer after each centrifugation. And CNTP was rinsed with water until there was no FITC that can be detected. 4. In order to inject single cells easily, the slides were coated. For coating with sigmacote (sigma, St. Louis, MO, USA), the slides were immersed into 5% sigmacate solution for 1 min, thoroughly washed with ethanol, and air-dried. 5. The capillary must be activated by rinsing with 0.1 M NaOH (more than 24 h) and 0.1 M HCl (more than 24 h) orderly. Before the reaction of the capillary with g-MAPS, the capillary needs to be dried completely. 6. The polymerized solution should be filled into the capillary within 10 min. Then the reaction time must be controlled (48–96 h). 7. Polyacryamide coating was important for the detection of fluorophores. The peak shape, detection limit, and retention

Carbon Nanotubes as Intracellular Carriers

167

time were determined by the factor. After each analysis, the capillary should first be rinsed by water for about 10 min and then filled with running buffer. The capillary must be rinsed by water and dried with N2 to protect the coating after the analysis is finished everyday. 8. The release experiment was carried in dialysis tubing. The dialyzate was added into the dialysis tubing with mole cular weight cut-off of 3,000.

Acknowledgments Financial support from the National Natural Sciences Foundation of China (No. 20520120220), the China State Key Basic Research Program Grants (2003CB716002), and the Knowledge Inno vation Program of DICP to H.Z. are gratefully acknowledged. References 1. Gottesman MM, Pastan I (1993) Biochemistry of multidrug-resistance mediated by the multidrug-resistant gene. Annu Rev Biochem 62:385–427 2. Higgins CF (2007) Multiple molecular mechanisms for multidrug resistance transporters. Nature 446:749–757 3. Hadjeri M, Barbier M, Ronot X, Mariotte AM, Boumendjel A, Boutonnat J (2003) Modu lation of P-glycoprotein-mediated multidrug resistance by flavonoid derivatives and analo gues. J Med Chem 46:2125–2131 4. Mazel M, Clair P, Rousselle C, Vidal P, Scherrmann JM, Mathieu D, Temsamani J (2001) Doxorubicin-peptide conjugates over come multidrug resistance. Anticancer Drugs 12:107–116 5. Koizumi S, Konishi M, Ichihara T, Wada H, Matsukawa H, Goi K, Mizutani S (1995) Flow cytometric functional analysis of multidrug resistance by Fluo-3: a comparison with rhodamine-123. Eur J Cancer 31A:1682–1688 6. Lin Y, Taylor S, Li HP, Fernando KAS, Qu LW, Wang W, Gu LR, Zhou B, Sun YP (2004) Advances toward bioapplications of carbon nanotubes. J Mater Chem 14:527–541 7. Singh R, Pantarotto D, McCarthy D, Chaloin O, Hoebeke J, Partidos CD, Briand JP, Prato M, Bianco A, Kostarelos K (2005) Binding and condensation of plasmid DNA onto functionalized carbon nanotubes: toward the construction of nanotube-based gene delivery vectors. J Am Chem Soc 127:4388–4396

8. Kam NWS, Dai HJ (2005) Functionalization of carbon nanotubes via cleavable disulfide bonds for efficient intracellular delivery of siRNA and potent gene silencing. J Am Chem Soc 127:6021–6026 9. Pantarotto D, Singh R, McCarthy D, Erhardt M, Briand JP, Prato M, Kostarelos K, Bianco A (2004) Functionalized carbon nanotubes for plasmid DNA gene delivery. Angew Chem Intl Ed Engl 43:5242–5246 10. Pantarotto D, Partidos CD, Hoebeke J, Brown F, Kramer E, Briand JP, Muller S, Prato M, Bianco A (2003) Immunization with peptidefunctionalized carbon nanotubes enhances virus-specific neutralizing antibody responses. Chem Biol 10:961–966 11. Liu Z, Winters M, Holodniy M, Dai H (2007) siRNA delivery into human T cells and primary cells with carbon-nanotube transporters. Angew Chem Intl Ed Engl 46:2023–2027 12. Pantarotto D, Briand JP, Prato M, Bianco A (2004) Translocation of bioactive peptides across cell membranes by carbon nanotubes. Chem Commun (1):16–17 13. Pagona G, Fan J, Maigne A, Yudasaka M, Iijima S, Tagmatarchis N (2007) Photoinduced electron transfer on aqueous carbon nanohornpyrene-tetrathiafulvalene architectures. Diam Relat Mater 16:1150–1153 14. Qu LW, Martin RB, Huang WJ, Fu KF, Zweifel D, Lin Y, Sun YP, Bunker CE, Harruff BA, Gord JR, Allard LF (2002) Single-walled

168

Li et al.

carbon nanotubes tethered with porphyrins: synthesis and photophysical properties. J Chem Phys 117:8089–8094 15. Pappas TJ, Gayton-Ely M, Holland LA (2005) Recent advances in micellar electrokinetic chromatography. Electrophoresis 26:719–734 16. Beale SC (1998) Capillary electrophoresis. Anal Chem 70:279–300 17. Hu S, Dovichi NJ (2002) Capillary electrophoresis for the analysis of biopolymers. Anal Chem 74:2833–2850 18. Sniehotta M, Schiffer E, Zurbig P, Novak J, Mischak H (2007) CE – a multifunctional application for clinical diagnosis. Electropho resis 28:1407–1417 19. Pan CS, Xu SY, Hu LG, Ou JJ, Zou HF, Guo Z, Zhang Y, Guo BC (2005) Using oxidized

carbon nanotubes as matrix for analysis of small molecules by MALDI-TOF MS. J Am Soc Mass Spectrom 16:883–892 20. Xiao H, Li X, Zou HF, Yang L, Wang YL, Wang HL, Le XC (2006) CE-LIF coupled with flow cytometry for high-throughput quantitation of fluorophores in single intact cells. Electrophoresis 27:3452–3459 21. Hjertén S (1985) High-performance electrophoresis: elimination of electroendosmosis and solute adsorption. J Chromatogr 347: 191–198 22. Xiao H, Yang LS, Zou HF, Yang L, Le XC (2007) Analysis of oxidized multi-walled carbon nanotubes in single K562 cells by capillary electrophoresis with laser-induced fluorescence. Anal Bioanal Chem 387:119–126

Part IV Scaffolds

Chapter 14 Carbon Nanotube-Based Neurochips Moshe David-Pur, Mark Shein, and Yael Hanein Abstract High-density carbon nanotube (CNT)-coated surfaces are highly neuro-adhesive. When shaped into regular arrays of isolated islands on a non-adhesive support substrate (such as a clean glass), CNTs can function as effective encoring sites for neurons and glia cells for in-vitro applications. Primarily, patterned CNT islands provide a means to form complex, engineered, interconnected neuronal networks with pre-designed geometry via utilizing the self-assembly process of neurons. Depositing these CNT islands onto multielectrode array chip can facilitate both cell anchoring but also electrical interfacing between the electrodes and the neurons. Key words: Neuro-chip, Cell-patterning, Neurons, Glia cells, Cell-adhesion, Nanotopography

1. Introduction Surface topography is an important parameter affecting cell adhesion. Even strictly non-adhesive substrates such as SiO2 can become adhesive through roughening (1). Cell sensitivity to roughness has many interesting manifestations such as conspicuous dependence between cell-adhesion and the degree of surfaceroughness (1), preferred cell adhesion to ridges rather than to the bottom of grooves (2, 3), and sensitivity to small irregularities in ordered nano-pit arrangements (4). While the nature of the underlying mechanism remains unclear, it appears that intertwinement of cellular prolongations with underlying rough surfaces such as CNTs is a likely mechanism. Moreover, it was suggested that neurite outgrowth of neurons cultured on PLL-functionalized CNTs is dependent on the CNT mechanical properties, as ascertained using CNTs of different lengths (5). It was argued that short CNTs are too rigid, whereas long CNTs are flexible and thus can deform in shape when grasped by the K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_14, © Humana Press, a part of Springer Science + Business Media, LLC 2010

171

172

David-Pur, Shein, and Hanein

filopodia of the growth cone. This deformation enables the elongation of the growth cone. High-density three-dimensional CNT coatings can be utilized to create extremely rough surfaces and are thus effective in supporting neuronal adhesion (6–19). Being cell-adhesive, biologically inert, and mechanically very strong, CNT is a good material for neuro-chip applications in which there is a need to stabilize cells, particularly neurons, to specific locations on the chip. Owing to the fact that CNTs are also electrically conductive, they can be effectively used as a cell-adhesive coating material for electrodes. High-density CNT-coated surfaces with wellcontrolled tube parameters (length and diameter) can therefore be effectively used to anchor neuronal cells to specific locations on surfaces and to facilitate direct electrical interfacing with them (recording and stimulation). Cell anchoring and effective electrical interfacing with neurons is useful for various applications, and in particular for neuro-chip applications, in which cultured neuronal network are investigated in-vitro for prolonged periods of time (weeks). The approach can be used to build advanced neuro-chips for bio-sensing applications (e.g., drug and toxin detection) where the structure, stability, and reproducibility of the networks are of great relevance. The method described here outlines the fabrication of isolated, CNT island arrays, and the consecutive cell culturing. The array preparation method is based on photo-lithography, and CNT chemical vapor deposition techniques. The formation of welldefined CNT islands for cell culturing applications requires the micro-fabrication of nickel (Ni) islands on top of titanium nitride (TiN) support islands. The Ni islands later act as catalyst seeds for

Fig. 1. A Self-assembled patterned neuronal network on an array of CNT islands. (a) After several days of plating, the cells on the CNT neuro-chip interconnect to form networks with cells preferentially aggregating at the CNT islands. Image was obtained using a confocal microscope after fixation and drying of the sample. (b) The arrangement of neurons and glia cell clusters on a single CNT island. Image was obtained using confocal fluorescence microscopy with glia cells (green) and neuronal synapses (red) specifically stained after fixation

Carbon Nanotube-Based Neurochips

173

the chemical vapor deposition (CVD) of CNTs. Once the CNT islands are formed, cell patterning is achieved by plating the dissociated neurons and glia cells on top of the CNT arrays (Fig. 1). The process described here can be expanded by patterning the TiN as conducting traces and by insulating these traces with a passivation layer. These additions facilitate the formation of full-fledged CNT-based MEA electrodes (18, 19).

2 Materials 2.1. Micro Fabrication Equipment

1. Sputter system for TiN sputtering (MRC 8620). 2. E-beam evaporator for Ni evaporation (VST). 3. RIE system with SF6 gases for TiN etch (Nextral NE860). 4. Mask alignment system MA6 (Karl Suss). 5. Oxygen plasma system picoRF-PC (Diener electronic).

2.2. Micro Fabrication Materials

1. Quartz substrate 0.5–1 mm thick (United Silica Inc.). 2. Positive photo resist S1818 (Shipley). 3. Developer MIF-319 (Shipley).

2.3. Cell Culturing

1. Dissociated cortical cells from 18 day Sprague Dawley embryos. 2. Poly-D-lysine (PDL) (Sigma). 3. Modified essential medium with Eagle’s salts (Eagle MEM) (Beith HaEmek). 4. Horse serum (Beith Hahemek). 5. Gentamycin (Beith Hahemek).

3. Methods Substrates are prepared using optical photolithography and conventional thermal CVD system. 3.1. Substrate Fabrication

1. Quartz substrates (see Note 1) (0.5–1 mm thick) are cleaned with argon plasma (inside the sputter chamber) for 30 s. 100 nm TiN layer (see Notes 2–4) is then sputtered on the quartz substrate (the TiN layer acts as a diffusion barrier for the Ni catalyst in the CVD process, see Note 5).

174

David-Pur, Shein, and Hanein

2. Samples are spin coated with positive photo resist (see Note 6) followed by standard lithography process to pattern small TiN islands. 3. The exposed TiN regions are etched by reactive ion etch (RIE). 4. The photoresist is then removed with acetone or N-Methyl2-pyrrolidone (NMP) by sonication for 10 min at 40°C. The samples are then cleaned and washed with isopropanol (ISO) and dried with nitrogen. If following sonication there are still photo resist leftovers, one can use oxygen plasma to remove them. 5. A second lithography step is conducted to pattern Ni islands directly onto the TiN islands. 6. A 7–10 nm Ni layer is then deposited by e-beam evaporation. 7. Finally the photo resist is removed as in step 4. The above process results with quartz substrates with Ni islands on top of TiN islands. The Ni islands are the substrate onto which the CNTs are later grown. 3.2. CNT Growth

1. Substrates are put into the CVD system (thermal CVD system based on a tube furnace with 1 inch dia. quartz tube). 2. The system is purged at room temperature in a hydrogen atmosphere (200–400 sccm) for 5–10 min. 3. Furnace temperature is ramped to the growth temperature (900°C at a rate of 40°C min−1 at a constant hydrogen flow of 1,000 sccm (see Notes 7 and 8). 4. At the target temperature, ethylene gas (at 20 sccm) is added (see Note 9) to the hydrogen gas flow for 10 additional minutes. 5. The furnace and ethylene gas flow are switched off, allowing the samples to cool down to 300°C under hydrogen atmosphere (200 sccm). 6. At 300°C the hydrogen flow is stopped, the furnace is opened and the samples are removed. This process results with quarts substrates with multi-walled CNTs on TiN islands (see Notes 10 and 11).

3.3. Cell Culturing

1. Prior to culturing, substrates are cleaned using O2 plasma. 2. The CNT patterned substrates are placed in 35 mm petri dishes coated with PDL. 3. Dissociated cortical cells are suspended in 3 ml Eagle MEM containing 5%-horse serum, 5 mg/ml gentamycin, 50 mM glutamine, and 0.02 mM glucose.

Carbon Nanotube-Based Neurochips

175

4. The suspended cells are plated onto the Petri dishes with the CNT patterned substrates. Cell density of 1,000 cells/mm2 is commonly used (see Note 12). 5. The cultures are maintained in an incubator at 37°C with 5% CO2 and 95% humidity. 6. One third of the growth medium is replaced every 3–4 days.

4. Notes 1. The substrates have to be compatible with the high temperature of the CVD growth, and with the lithography fabrication steps. 2. The TiN layer is fundamental to the success of the fabrication process as it facilitates several important requirements: (1) The TiN is an excellent barrier layer, which facilitates effective CNT growth on any substrate including quartz. Ni deposited directly on quartz resulted with poor CNT growth. (2) The high melting temperature of the TiN ensures its survival in the CVD growth, thus its suitability as an underlying electrode material. (3) TiN is durable in biological solutions, and thus can survive for very long time in biological medium. 3. The thickness of the TiN layer is important as some of the Ni diffuses into the TiN layer during the CVD growth. If it is too thin, the majority of the layer is consumed into the TiN and there is not enough catalyst for the CNT growth. If the layer thickness is too high, the resulting CNTs are very thick and may not be very effective for cell culturing purposes. 4. Instead of using TiN target, one can use Ti target with the addition of nitrogen gas flow (in addition to argon) during the sputter process. 5. The diffusion barrier layer efficiency is determined by the sputter parameters. The deposition rate determines the quality of the layer, including its density. 6. Different photo resists can be used, provided all residues are removed prior to the CVD growth. For S1818, the following process parameters are used: spin at 2,400 rpm for 40 s, 5 s exposure, 1 min development with MIF-319. These parameters should be adjusted to obtain best results. 7. The CNT growth depends on the relationship between three major parameters: TiN layer thickness, Ni layer thickness, and the temperature ramp rate. For example, in case the TiN layer thickness is fixed, lowering the Ni layer thickness will require an increase in the temperature ramp rate or vice versa.

176

David-Pur, Shein, and Hanein

Alternatively, in case the temperature ramp is constant, reducing the Ni layer thickness will require an increase in the TiN layer or vice versa. 8. CNTs are grown at a temperature between 860°C and 1,000°C (upper limit of the furnace). 9. CNTs can be grown with methane instead of the ethylene gas. 10. A similar process can be applied to form CNT-based MEA. The MEA fabrication requires additional fabrications steps to form the passivation layer. This step is introduced after the TiN layer etch. Sputtered Si3N4 layer is a good choice as it is durable in the CVD process. 11. CNT diameter is 50 ± 20 nm. 12. CNT arrays have to cover large parts of the substrate to form as much adhesive area as possible so that no excess cells remain loose. Loose cells tend to aggregate into giant clusters and interfere with the patterning scheme.

Acknowledgment The authors thank Inna Brains for technical assistance. References 1. Fan YW, Cui FZ, Hou SP, Xu QY, Chen LN, Lee IS (2002) Culture of neural cells on silicon wafers with nano-scale surface topography. J Neurosci Methods 120:17–23 2. Goldner JS, Bruder JM, Li G, Gazzola D, Hoffman-Kim D (2006) Neurite bridging across micropatterned grooves. Biomaterials 27:460–472 3. Johansson F, Carlberg P, Danielsen N, Montelius L, Kanje M (2006) Axonal outgrowth on nano-imprinted patterns. Biomate rials 27:1251–1258 4. Curtis AS (2004) Small is beautiful but smaller is the aim: review of a life of research. Eur Cell Mater 8:27–36 5. Zhang X, Prasad S, Niyogi S, Morgan A, Ozkan M, Ozkan CS (2005) Guided neurite growth on patterned carbon nanotubes. Sensor Actuators B 106:843–850 6. Harrison BS, Atala A (2007) Carbon nanotube applications for tissue engineering. Biomate rials 28:344–353 7. Malarkey EB, Parpura V (2007) Applications of carbon nanotubes in neurobiology. Neuro degener Dis 4:292–299

8. Mattson MP, Haddon RC, Rao AM (2000) Molecular functionalization of carbon nanotubes and use as substrates for neuronal growth. J Mol Neurosci 14:175–182 9. Hu H, Ni YC, Mandal SK, Montana V, Zhao N, Haddon RC et al (2005) Polyethyleneimine functionalized single-walled carbon nanotubes as a substrate for neuronal growth. J Phys Chem B 109:4285–4289 10. Hu H, Ni YC, Montana V, Haddon RC, Parpura V (2004) Chemically functionalized carbon nanotubes as substrates for neuronal growth. Nano Lett 4:507–511 11. Gabay T, Ben-David M, Kalifa I, Sorkin R, Abrams ZR, Ben-Jacob E et al (2007) Electro-chemical and biological properties of carbon nanotube based multi-electrode arrays. Nanotechnology 18:035201–035206 12. Gabay T, Jakobs E, Ben-Jacob E, Hanein Y (2005) Engineered self-organization of neural networks using carbon nanotube clusters. Physica A 350:611–621 13. Wang K, Fishman HA, Dai HJ, Harris JS (2006) Neural stimulation with a carbon nanotube microelectrode array. Nano Lett 6:2043–2048

Carbon Nanotube-Based Neurochips 14. Mazzatenta A, Giugliano M, Campidelli S, Gambazzi L, Businaro L, Markram H et al (2007) Interfacing neurons with carbon nanotubes. Electrical signal transfer and synaptic stimulation in cultured brain circuits. J Neurosci 27:6931–6936 1 5. Lovat V, Pantarotto D, Lagostena L, Cacciari B, Grandolfo M, Righi M et al (2005) Carbon nanotube substrates boost neuronal electrical signaling. Nano Lett 5: 1107–1110 16. Nguyen-Vu TDB, Chen H, Cassell AM, Andrews R, Meyyappan M, Li J (2006) Vertically aligned carbon nanofiber arrays:

177

an advance toward electrical-neural interfaces. Small 2:89–94 17. Sorkin R, Gabay T, Blinder P, Baranes D, Ben-Jacob E, Hanein Y (2006) Compact selfwiring in cultured neural networks. J Neural Eng 3:95–101 18. Gabay T, Ben-David M, Kalifa I, Sorkin R, Abrams ZR, Ben-Jacob E, Hanein Y (2007) Electro-chemical and biological properties of carbon nanotube based multi-electrode arrays. Nanotechnology 18:035201–035206 19. Ben-Jacob E, Hanein Y (2008) Carbon nanotube micro-electrodes for neuronal interfacing. J Mater Chem 18:5181

Chapter 15 Effect of Carbon Nanotubes on HepG2 Adhesion and Spreading Suijian Qi, Changqing Yi, Dawei Zhang, and Mengsu Yang Abstract There are several in vitro cell models for studying the interactions of carbon nanotubes (CNTs) with biological systems. This chapter provides a detailed protocol for studying the effects of CNTs on cell adhesion and spreading. The protocol is a combination of methods of electron microscopy, cell biology, and molecular biology, focusing on the detection of the cell morphology and cellular responses related to cell adhesion and spreading process, including the observation of cellular skeletal changes upon cell adhesion, the measurement of the expression level changes of cell adhesion, and the spreading specific genes and proteins upon exposure to CNTs. Key words: Carbon nanotubes, HepG2, Cell adhesion, Cell spreading, Cell viability, Transmission electron microscopy, Scanning electron microscopy reverse transcription polymerase chain reaction, Western immunoblotting

1. Introduction Carbon nanotubes (CNTs) have attracted a lot of research interests due to their unique properties, and several in vitro cell models have already been established for studying the interactions of CNTs with biological systems, such as cytotoxicity of CNTs (1– 4), CNTs and tissue engineering (5, 6), CNTs and gene delivery (7), CNTs and cell growth guidance (8). In this chapter, the methods developed for studying the effects of CNTs on cell adhesion and spreading are described. The initial cell adhesion, including cell–cell adhesion and cell–ECM (extracellular matrix) adhesion, plays a major role in cellular communication and regulation, and is of fundamental importance in the development and maintenance of tissues (9–11). The study of the effects of CNTs on cell adhesion and spreading is K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_15, © Humana Press, a part of Springer Science + Business Media, LLC 2010

179

180

Qi et al.

very important to get a better knowledge of the interactions between CNTs and biological systems. We demonstrate here that by combining the techniques of electron microscopy (transmission electron microscopy and scanning electron microscopy), cell biology, and molecular biology (reverse transcription polymerase chain reaction and western immunoblotting), one can obtain the information of the effects of CNTs on cell behaviors, such as adhesion and spreading.

2. Materials 2.1. Cell Line and Cell Culture

1. HepG2 cell line (hepatocellular carcinoma, liver) (American Type Culture Collection, VA). 2. Complete cell culture medium: RPMI 1,640 Medium (Gibco, CA) supplemented with 10% fetal bovine serum (FBS) (Gibco, CA), 1% L-glutamine, penicillin, and streptomycin (Gibco, CA). FBS was heated to 56°C for 30 min before use. 3. Trypsin–EDTA: 0.25% trypsin, 1 mM EDTA. 4. 75 cm2 Tissue culture flasks (TPP, Switzerland). 5. 6-well, 48-well cell culture plates (TPP, Switzerland). 6. Plastic coverslips (Chemicon, CA).

2.2. Cell Viability Test

1. 0.4% Trypan Blue solution (Sigma). 2. 1 × Phosphate buffer solution (PBS). 3. Hemocytometer.

2.3. Transmission Electron Microscopy

1. Fixation solution A: 2% glutaraldehyde and 2% paraformaldehyde in 0.1 M cacodylate buffer. 2. 1% OsO4 in 0.1 M cacodylate buffer. 3. Staining solution: 2.5 ml glutaraldehyde plus 19 ml 4% paraformaldehyde and 2 ml 0.2 M cacodylate buffer.

2.4. Scanning Electron Microscopy

1. Fixation solution A: 2% glutaraldehyde and 2% paraformaldehyde in 0.1 M cacodylate buffer. 2. 1% OsO4 in 0.1 M cacodylate buffer. 3. Ethanol, hexamethyldisilazane (HMDS).

2.5. Reverse Transcription Polymerase Chain Reaction

1. 1 × Phosphate buffer solution (PBS). 2. TRIZOL reagent (Gibco, CA). 3. RNA secure solution (Invitrogen, Gaithersburg, MD). 4. 10 × DNase I buffer (Invitrogen, Gaithersburg, MD).

Effect of Carbon Nanotubes on HepG2 Adhesion and Spreading

181

5. 0.5 mg/ml Oligo (dT) (Invitrogen, Gaithersburg, MD). 6. 10 mM dNTP mix (Invitrogen, Gaithersburg, MD). 7. 5× First Strand buffer (Invitrogen, Gaithersburg, MD). 8. RNase Inhibitor (Invitrogen, Gaithersburg, MD). 9. Superscript III RNase Reverse Transcriptase (Invitrogen, Gaithersburg, MD). 10. Taq DNA polymerase (Invitrogen, Gaithersburg, MD). 11. QiAquick PCR Purification Kit (Qiagen) 2.6. Western Immunoblotting for Col-I and a-actin

1. Cell lysis buffer: 1 M Tris-HCl, pH 7.4, 25 ml; 0.5 M EDTA, 10 ml; 20% Triton X-100, 25 ml; Urea, 130 g; add water to a final volume of 500 ml. Store at 4°C. 2. Bradford reagent (GE Healthcare). Store at 4°C in a refrigerator. 3. Bovine serum albumin (BSA): Prepare 0.02 mg/ml solution in water right before use. 4. 5 × loading buffer: Tris, 0.375 g; Sodium dodecyl sulfate (SDS), 1.0 g; Glycerol, 5.0 ml; b-mercaptoethaol (Sigma), 2.5 ml; Bromephenol blue (Sigma), 50 mg; H2O, 2.5 ml. Store at −20°C in a refrigerator. Thaw before use. 5. SDS-PAGE: 30% Acrylamide/Bis solution (37.5:1 with 2.6% C) (GE Healthcare); N, N, N, N¢-Tetramethyl-ethylenediamine (TEMED) (GE Healthcare); Ammonium persulfate (APS): Prepare 10% solution in water and store at 4°C refrigerator. Recommend to be used up within 1 week; 0.5 M Tris-HCl, pH 6.8; 1.5 M Tris-HCl, pH 8.8; 10% SDS solution. 6. SDS-PAGE buffer: Tris, 3 g; Glycine, 14.4 g; SDS, 1 g; add water to a final volume of 1,000 ml. 7. Transferring buffer: Tris, 3.03 g; Glycine, 14.42 g; Methanol, 200 ml; add water to a final volume of 1,000 ml. 8. Nitrocellulose membrane (Millipore). 9. TBS-T (Washing buffer): 1 M Tris-HCl (pH 8.0), 10 ml; 1 M NaCl, 150 ml; Tween-20, 0.5 ml; add water to 1,000 ml. 10. Blocking buffer: 5% (w/v) nonfat milk powder in TBS-T. 11. Antibody dilution buffer: 1% (w/v) nonfat milk powder in TBS-T. 12. Primary antibodies: Mouse antibodies specific to type I collage (CalBioChem, CP17L), a-actin (DakoCytomation M 0851); Rabbit antibody specific to b-tubulin (Santa Cruz, 9,104).

182

Qi et al.

Dilute the antibodies with antibody dilution buffer at 1:100. Store at −20°C in a refrigerator. Thaw before use. 13. Secondary antibody: Anti-mouse or anti-rabbit IgG conjugated to horse radish peroxidase (Santa Cruz). Dilute the antibodies with antibody dilution buffer at 1:1,000. Store at −20°C in a refrigerator. Thaw before use. 14. Enhanced chemiluminescent (ECL) reagents (GE Healthcare). Store at 4°C refrigerator. Place at room temperature before use. 15. Film, X-ray film cassette (Kodak); film developing reagent and film fixing reagent. To be used in a dark room.

3. Methods (see Note 1) 3.1. Cell Viability Test: Trypan Blue Exclusion Method

1. Cells were cultured in complete medium at 37°C in a humidified 5% CO2 incubator. 2. Trypan Blue will stain dead or dying cells, while it will not stain live cells with intact cell membranes. Hence, dead cells are shown as a distinctive blue color under a microscope, while live cells are not. Therefore, the ability of Trypan Blue to distinguish live and dead cells can be applied to test the viability of the cells. (see Note 2) 3. Seed 1 × 105 HepG2 cells in 48-well cell culture plate and incubate overnight; add CNTs suspensions of different concentrations to the cells and incubate for the desired hours (24, 48, or 72 h). (see Notes 3 and 4). 4. At the conclusion of cell culture, remove the culture medium and wash the cells with PBS twice. 5. Harvest the cells: add 0.3 ml of Trypsin-EDTA to each well, incubate at 37°C in a humidified atmosphere of 5% CO2 in air for about 5 min; then, add 0.5 ml of complete culture medium to terminate the trypsinization; later, centrifuge at 680g for 5 min to collect the cell pellet. 6. Remove the supernatant; resuspend the cell pellet in PBS; pipet gently to get even cell suspension. 7. Transfer the following into a 1.5 ml Eppendorf centrifuge tube: a. 0.5 ml of 0.4% Trypan Blue solution b. 0.3 ml of PBS c. 0.2 ml of cell suspension in PBS (= dilution 1:5) (see Note 5) 8. Allow the tube to stand for 10 min, but do not exceed 30 min. 9. Pipet 10 ml of this mixture into coverslipped chambers of hemocytometer, avoiding cell clusters by pipetting up and down.

Effect of Carbon Nanotubes on HepG2 Adhesion and Spreading

183

10. Count viable and nonviable cells under microscope. 11. Calculate the viability of the cells: a. Cells/ml: the number of cells per quadrant equals 10 4 cells/ml b. Total cells: cells/ml × original volume c. Cell viability (%): total viable cells (unstained)/total cells (stained and unstained) × 100 12. Plot the cell viability curve according to the incubation time with carbon nanotubes suspension or according to the concentrations of carbon nanotubes suspensions. 3.2. Transmission Electron Microscopy

1. Cell culture: seed 1 × 105 HepG2 cells on sterilized plastic coverslip and incubate overnight; then add CNTs suspension to the cells and incubate for another 48 h. 2. Fixation: after cell culture, wash the cells with PBS buffer twice. Perform primary fixation with fixation solution A in darkness for 2 h at room temperature. Wash the cells with 0.1 M cacodylate buffer for 5 min twice. Postfix in 1% OsO4 in 0.1 M cacodylate buffer at room temperature in darkness for 2 h. Rinse in 0.1 M cacodylate buffer for 10 min; then, wash in 0.1 M cacodylate buffer for 10 min twice; later, wash in H2O for 10 min twice. Keep in mind that the side of the plastic coverslip with the cells should always be facing up. 3. Dehydration: dehydrate the cells following the steps: (see Note 6) 10% ethanol

15 min

30% ethanol

15 min

50% ethanol

15 min

70% ethanol

15 min

80% ethanol

15 min

90% ethanol

15 min

100% ethanol

15 min

100% ethanol

15 min

100% ethanol–acetone (3:1)

15 min

100% ethanol–acetone (1:1)

15 min

100% ethanol–acetone (1:3)

15 min

100% acetone

15 min

100% acetone

15 min

4. Infiltration: infiltrate the cells following the steps; evenly rotate the samples while processing.

184

Qi et al. 100% acetone–Spurr’s resin (4:1)

1 h

100% acetone–Spurr’s resin (3:2)

overnight

100% acetone–Spurr’s resin (2:3)

half day

100% acetone–Spurr’s resin (1:4)

half day

Pure Spurr’s resin

overnight

Pure Spurr’s resin

half day

Pure Spurr’s resin

half day

Pure Spurr’s resin

overnight

Pure Spurr’s resin

half day

5. Embedding: embed each plastic coverslip in a plastic mold at room temperature (1–2 hr). Put in a 70°C oven for 2 days. 6. Block trimming and sectioning: carefully trim the resin blocks to expose the underlying blocks. Cut the sections of various thickness (200, 300, and 500 nm) using Leica Ultracut UCT microtome and transfer to 300 mesh copper grids. 7. Staining: stain the sections on copper grids for 1 h with staining solution in darkness at room temperature. Wash with water thoroughly. 8. Transmission Electron Microscopy: observe the sections on grids in Jeol (JEM 2,000 FX) electron microscope at 160 kV. Figure 1 shows the results obtained. 3.3. Scanning Electron Microscopy

1. Step 1 and step 2 are the same as those described in subheading 3.2 Transmission Electron Microscopy. 2. Dehydration: dehydrate the cells following the steps: (see Note 6) 10% ethanol

15 min

30% ethanol

15 min

50% ethanol

15 min

70% ethanol

15 min

80% ethanol

15 min

95% ethanol

15 min

95% ethanol

15 min

100% ethanol

15 min

100% ethanol

15 min

100% ethanol

15 min

HMDS

1 min

HMDS

1 min

Effect of Carbon Nanotubes on HepG2 Adhesion and Spreading

185

Fig. 1. TEM images of the interaction of carbon nanotubes with HepG2 cells. (a) Control, HepG2 cell without carbon nanotube treatment; (b) Single-wall carbon nanotubes were seen in the cytoplasm of HepG2 cells after the cells were treated with single-wall carbon nanotubes; (c) Multiwall carbon nanotubes were also seen in the cytoplasm of HepG2 cell after the cells were treated with multiwall carbon nanotubes; (d) The higher magnified images of the red box in (b) showed the intact morphology of the single-wall carbon nanotubes in the cells

3. Dry the sample in 60°C oven; then mount the plastic coverslip on an aluminum stub with the cells facing up. 4. Sputter coat a lay of gold to the surface of the cell layer; observe the samples by SEM. Figure 2 shows the results obtained. 3.4. Reverse Transcription Polymerase Chain Reaction (RT-PCR)

1. Several genes specific to cell adhesion and spreading are selected for illustration. The protocol described here can apply to the similar study of the gene expression detection of other genes of interest. 2. Cell culture: culture 1 × 105 HepG2 cells with 100 mg/ml of CNTs in a 6-well cell culture plate at 37°C in a humidified atmosphere of 5% CO2 in air for 48 h. Cells without treatment are used as control.

186

Qi et al.

Fig. 2. SEM images of the spreading morphology of HepG2 cells upon interaction of carbon nanotubes. (a) Control, HepG2 cell without carbon nanotube treatment; (b) Cell morphology after multiwall carbon nanotubes treatment; (c) The higher magnified image of (b) showed that the cell spreading behavior was affected by carbon nanotubes treatment

Effect of Carbon Nanotubes on HepG2 Adhesion and Spreading

187

3. At the conclusion of treatment, remove the medium. 4. Wash cells throughly with PBS. (see Note 7) 5. Lysis the cell pellets in each well with 1 ml TRIZOL reagent. Collect each of the cell lysis into an Eppendorf centrifuge tube. 6. Add 0.2 ml of chloroform to each 1 ml of lysis and then incubate at room temperature for 2–3 min. 7. Centrifuge at 12,000 × g for 15 min at 4°C. 8. Transfer the upper aqueous phase to a fresh tube. 9. Add 0.5 ml of isopropanol to each tube containing the aqueous phase, mix by hand, and incubate the tubes at room temperature for 10 min. 10. Centrifuge at 12,000 × g for 10 min at 4°C to precipitate RNA. 11. Remove the supernatant, wash the RNA pellets twice with 75% ethanol. Mix the sample by vortexing and centrifuge at 7,500 × g for 5 min at 4°C. 12. Remove nearly every drop of 75% ethanol, briefly dry the pellet at room temperature for 5–10 min. 13. Dissolve RNA by adding proper amount of RNA secure solution. 14. Quantify the RNA samples by photometer and check the quality by agarose gel electrophoresis to ensure the absence of degradation. 15. Reverse transcription: use about two micrograms of each total RNA sample for reverse transcription under standard conditions. Assume that the concentration of the RNA sample measured at step 12 is 0.4 mg/ml, so 5 ml of the RNA sample is needed to carry out the reverse transcription. In a 0.2 ml Eppendorf centrifuge tube, add 1 ml of 10 × DNase I buffer, 2 ml of DNase I, and 2 ml of H2O to the 5 ml of the RNA sample; incubate at room temperature for 15 min; then, add 1 ml of 25 mM EDTA to the tube; incubate at 65°C for 10 min; place on ice for 1 min; then, add 1 ml of 0.5 mg/ml Oligo (dT), 1 ml of 10 mM dNTP to the tube and incubate at 65°C for 10 min; later, add 4 ml of 5 × First Strand buffer, 1 ml of 0.1 M DTT, 1 ml of RNase Inhibitor, 1 ml of Superscript III RNase Reverse Transcriptase to the tube; incubate at 50°C for 60 min and then 70°C for 15 min. 16. Purify the resulting cDNA with QiAquick PCR Purification Kit (Qiagen, USA) or other commercialized purification kit according to their handbooks. 17. Use the purified cDNA as template in subsequent PCR. 18. Sequences of interest are amplified using the following primer pairs: a-actin (5¢-ATC TGG CAC CAC ACC TTC TA-3¢,

188

Qi et al.

5¢-AGC TCG TAG CTC TTC TCC AG), integrin (5¢-GAC CTG CCT TGG TGT CTG TGC-3¢, 5¢- AGC AAC CAC ACC AGC TAC AAT-3¢), FAK (5¢-GAA GTC TTC AGG GTC CGA TTG-3¢, 5¢- CAT TCT CGT ACA CCT TAT CAT TCG-3’), type I collagen (5¢-AAC ATG ACC AAA AAC CAA AAG TG-3¢, 5¢-CAT TGT TTC CTG TGT CTT CTG G-3¢). GAPDH (5¢- GAC TTC AAC AGC AAC TCC CAC -3¢, 5¢- TCC ACC ACC CTG TTG CTG TA-3¢) is used as endogenous reference housekeeping gene. 19. In a 0.2 ml Eppendorf centrifuge tube, mix the following reagents to prepare for PCR reaction: Reagent

Final concentration

cDNA template

10 pg–1 mg

Primer I

0.1–1 mM

Primer II

0.1–1 mM

Taq DNA polymerase

1.25 U/50 ml

10 mM dNTP mix

0.2 mM of each

25 mM MgCl2

1–4 mM

10 × PCR buffer

1×

H2O

-- (to make the final volume to 20 ml)

20. PCR conditions are as follows: 30 s 94°C, 30 s 58°C, 30 s 72°C, (5 min 94°C, 5 min 72°C) × 35 cycles. (a-actin, b-actin, integrin); 30 s 94°C, 30 s 56°C, 30 s 72°C, (5 min 94°C, 5 min 72°C) × 45 cycles. (FAK, type I collagen). 21. Subject the PCR product to agarose gel electrophoresis. Figure 3 shows the results obtained. 3.5. Western Immunoblotting for Col-I and a-actin

1. Two proteins that have strong relationship with cell adhesion and spreading are selected for illustration. The protocol described here can apply to the similar study of other proteins of interest. 2. Cell culture: culture 1 × 105 HepG2 cells with 100 mg/ml of CNTs in a 6-well cell culture plate at 37°C in a humidified atmosphere of 5% CO2 in air for 48 h. Cells without CNTs treatment are used as control. At the conclusion of treatment, remove the medium and wash the cells with PBS three times (see Note 7). 3. Cell lysis: add 400 ml cell lysis buffer to each well; pipet gently; collect the cell lysis into a 1.5 ml Eppendorf centrifuge tube. Place the tube in an ice-box. Determine the protein concentration immediately (see Note 8).

Effect of Carbon Nanotubes on HepG2 Adhesion and Spreading

189

Fig. 3. Images of 1.5% agarose gel electrophoresis for RT-PCR products to view the expression profile of adhesion-associated genes. Lane 1: Control; Lane 2: Treatment. (a) FAK; (b) type I collagen; (c) integrin; (d) a-actin; (e) GAPDH, endogenous reference housekeeping gene

4. Bradford protein assay: prepare the standard solution in 1.5 ml Eppendorf centrifuge tubes according to the following table: Standard

0.02 mg/ml BSA (ml)

Bradford reagent (ml)

H2O (ml)

0

0

200

800

1

50

200

750

2

100

200

700

4

200

200

600

6

300

200

500

8

400

200

400

10

500

200

300

Mix well and vortex; incubate at room temperature for 10 min; Measure absorbance using Biophotometer (OD595) ((Eppendorf, Germany)); Plot standard curve. Dilute sample to final volume as 800 ml, add 200 ml Bradford reagent, mix well and vortex; incubate at room temperature for 10 min, then measure absorbance using Biophotometer (OD595). Find out the protein concentration of the sample according to the standard curve. Adjust equal protein concentration of treatment sample and control sample with cell lysis buffer. Add 8 ml 5 × loading buffer to

190

Qi et al.

every 30 ml sample, mix; denature in 100°C water bathe for 3 min, then place at room temperature. 5. SDS-PAGE: The protocols are based on using the Bio-Rad’s Modular Mini Electrophoresis System. We will use the Mini-PROTEAN® II dual slab cell for running SDS-PAGE electrophoresis gels, and use the Mini Trans-Blot cell for transferring gel to the membrane. The unique feature of this electrophoresis system is that the electrode modules are interchangeable. After finishing one task, remove the electrode module from the buffer tank, insert a new electrode module, add a new buffer, and the next electrophoresis application can be performed. Make sure the gel system is water tight before the following experiments. Prepare a 1.5mm thick, 10% separating gel: 30% Acrylamid/ Bis, 3.33 ml; 1.5 M Tris-HCl (pH 8.8), 2.5 ml; 10% SDS, 100 ml; H2O, 4.02 ml; TEMED, 5 ml; 10% APS, 50 ml. Mix well. Pour the gel (about 7 ml), leaving space for a stacking gel, and overlay with ethanol. Wait for about 30 min till the gel gets polymerized. Pour off the ethanol. Prepare 4% stacking gel: 30% Acrylamid/Bis, 0.66 ml; 0.5 M Tris-HCl (pH 6.8), 1.26 ml; 10% SDS, 50 ml; H2O, 3 ml; TEMED, 5 ml; 10% APS, 50 ml. Mix well. Pour the gel to fill the upper top of the separating gel, insert the comb immediately. Wait for about 30 min till the gel gets polymerized. Carefully remove the comb from the stacking gel and wash the wells with SDS-PAGE buffer. Place the gel into the electrophoresis tank filled with SDSPAGE buffer. Load the 38 ml of each sample in a well. Include one well for protein molecular weight markers. Run the gel electrophoresis at 100 V for about 2 h 6. Western immunoblotting: Following the steps to transfer the separated gel to the nitrocellulose membrane: a. Pour out the SDS-PAGE buffer and pour the transferring buffer into the electrophoresis tank; b. Remove the stacking gel and keep the separating gel after SDS-PAGE. Cut the membrane and the filter paper to the dimensions of the gel. Always wear gloves when handling membranes to prevent contamination. Equilibrate the gel and soak the membrane, filter paper, and fiber pads in transfer buffer. c. Prepare the gel sandwich as shown in Fig. 4: Place the cassette, with the gray side down, on a clean surface. Place one prewetted fiber pad on the gray side of the cassette.

Effect of Carbon Nanotubes on HepG2 Adhesion and Spreading

191

Fig. 4. The illustration for preparing the gel sandwich

Place a sheet of filter paper on the fiber pad. Place the equilibrated gel on the filter paper. Place the prewetted membrane on the gel. Complete the sandwich by placing a piece of filter paper on the membrane. Add the last fiber pad. Remove any air bubbles during the assembly. d. Close the cassette firmly, being careful not to move the gel and filter paper sandwich. Lock the cassette closed with the white latch. e. Place the cassette in module, then place in the tank filled with transferring buffer. f. Add a standard stir bar to help maintain even buffer tempe rature and ion distribution in the tank. Set the speed as fast as possible to keep ion distribution even. g. Adjust the power supply at 70 V, transfer for 2 h at 4°C cool room with stirring. h. Upon completion of the run, disassemble the blotting sandwich and keep the membrane for development. Then, put the membrane into a small box. Block the membrane with blocking buffer at room temperature for 1 h with shaking. Pour out the blocking buffer. Incubate the membrane with primary antibody at 4°C cool room overnight on a rocking platform. Make sure the volume of the antibody can cover the membrane. Pour out the primary antibody. Wash the membrane three times with TBS-T at room temperature, 20 min each time. Shake while washing.

192

Qi et al.

Fig. 5. Results of western immunoblotting. Lane 1: Control; Lane 2: Treatment. (a) type I collagen; (b) a-actin; (c) b-tubulin was used as endogenous reference

Pour out the final TBS-T buffer. Incubate the membrane with secondary antibody at room temperature for 1 h on a rocking platform. Make sure the volume of the antibody can cover the membrane. Pour out the secondary antibody. Wash the membrane three times with TBS-T at room temperature, 20 min each time. Shake while washing. Perform the following steps in a dark room: Pour out the final TBS-T buffer. Mix equal volume of the ECL reagents 1 and 2, immediately add to the membrane. Gently shake by hand for 1 min to ensure even coverage. Remove the membrane, suck the redundant liquid with Kim-Wipes; Wrap the membrane in thin plastic membrane; Cut the film according to the size of membrane and place the film on top of the membrane. Place the membrane and film into an X-ray film cassette; expose for a suitable time. Develop and fix the film. Figure 5 shows the results obtained.

4. Notes 1. The methods described above can be applied to study the effects of carbon nanotubes to most anchorage-dependent cell lines, but not to the suspension cell lines. It is a general protocol to study the effects of CNTs on cell adhesion and spreading. Different observations and results will be generated based on the different kinds and sources of CNT samples being tested. 2. There are many other cell viability assays, such as MTT assay, Alamar Blue assay, but in the case of detecting the viability of cells after being treated with CNTs, we recommend the use

Effect of Carbon Nanotubes on HepG2 Adhesion and Spreading

193

of the Trypan Blue exclusion assay, because the graphitic materials will reduce the coloration of the aromatic dye and give out incorrect results (12, 13). 3. Unless stated otherwise, the CNTs suspensions should be prepared in water and followed by ultrasonication for half an hour. 4. Unless stated otherwise, the H2O used in the protocol is the autoclaved deionized water. 5. For optimal results, adjust cell density to 20–50 cells/square. 6. Tissue samples should never be allowed to dry. 7. After treatment of the cells with CNTs, there will be many CNTs remaining at the bottom of the cell culture plates and the surrounding of the cell body. The cells should be washed with PBS thoroughly to wash out as much CNTs as possible. Otherwise, the remaining CNTs will greatly affect the following experiments. 8. The cell lysis can be stored at −20°C for further use. But keep in mind that the protein concentration will change during the storage. Therefore, the protein concentration should be determined just before performing SDS-PAGE to get reliable results, even if the protein concentration was determined prior to putting the cell lysis into the refrigerator.

Acknowledgments This work was supported by City University of Hong Kong (Project No.7002100), BTC operation fund (CityU project No. 9683001) and the Key Laboratory Fund of Shenzhen Municipal Government, China.

References 1. Jia G, Wang HF, Yan L, Wang X, Pei RJ, Yan T, Zhao YL, Guo XB (2005) Cytotoxicity of carbon nanomaterials: single-wall nanotube, multi-wall nanotube, and fullerene. Environ Sci Technol 39:1378–1383 2. Cui DX, Tian FR, Ozkan CS, Wang M, Gao HJ (2005) Effect of single wall carbon nanotubes on human HEK293 cells. Toxicol Lett 155:73–85 3. Monteiro-Riviere NA, Nemanich RJ, Inman AO, Wang YY, Riviere JE (2005) Multi-walled carbon nanotube interactions with human epidermal keratinocytes. Toxicol Lett 155: 377–384

4. Zhu Y, Ran T, Li YG, Guo JX, Li WX (2006) Dependence of the cytotoxicity of multi-walled carbon nanotubes on the culture medium. Nanotechnology 17:4668–4674 5. Correa-Duarte MA, Wagner N, Rojas-Chapana J, Morsczeck C, Thie M, Giersig M (2004) Fabrication and biocompatibility of carbon nanotube-based 3D networks as scaffolds for cell seeding and growth. Nano Lett 4:2233–2236 6. Harrison BS, Atala A (2007) Carbon nanotube applications for tissue engineering. Biomaterials 28:344–353 7. Singh R, Pantarotto D, McCarthy D, Chaloin O, Hoebeke J, Partidos CD,

194

Qi et al.

Briand JP, Prato M, Bianco A, Kostarelos K (2005) Binding and condensation of plasmid DNA onto functionalized carbon nanotubes: toward the construction of nanotube –based gene delivery vectors. J Am Chem Soc 127: 4388–4396 8. Park SY, Namgung S, Kim B, Im J, Kim JY, Sun K, Lee KB, Nam JM, Park Y, Hong S (2007) Carbon nanotube monolayer patterns for directed growth of mesenchymal stem cells. Adv Mater 19:2530–2534 9. Ben-Ze’ev A, Robinson GS, Bucher NLR, Farmer SR (1988) Cell-cell and cell-matrix interactions differentially regulate the expression of hepatic and cytoskeletal genes in primary cultures of rat hepatocytes. Proc Natl Acad Sci 85:2161–2165 10. Hansen LK, Mooney DJ, Vacanti JP, Ingber DE (1994) Integrin binding and cell spreading on

extracellular matrix act at different points in the cell cycle to promote hepatocyte growth. Mol Biol Cell 5:967–975 11. Fassett J, Toblot D, Hansen LK (2006) Type I collagen structure regulates cell morphology and EGF signaling in primary rat hepatocytes through cAMP-dependent protein kinase A. Mol Biol Cell 17:345–356 12. Isobe H, Tanaka T, Maeda R, Noiri E, Solin N, Yudasaka M, Iijima S, Nakamura E (2006) Preparation, purification, characterization, and cytotoxicity assessment of watersoluble, transition-metal-free carbon nanotube aggregates. Angew Chem Int Ed 45:6676–6680 13. Worle-Knirsch JM, Pulskamp K, Krug HF (2006) Oops they did it again! Carbon nanotubes hoax scientists in viability assays. Nano Lett 6:1261–1268

Part V Biosensors

Chapter 16 Enzymatic Detection Based on Carbon Nanotubes Martin Pumera Abstract We describe in this chapter the preparation of simple and cheap carbon nanotube-based biosensor for sensing of glucose. Such biosensor is based on coupling carbon nanotubes and glucose oxidase. Key words: Carbon nanotubes, Enzyme, Biosensing, Electrochemistry

1. Introduction The carbon nanotube (CNT) based enzymatic biosensor has attracted considerable attention recently. The CNT can be used as nanowires connecting the redox active center of enzyme with electrode surface, or as electrode material bringing advantages of high conductivity, high surface area, and 3D structure allowing careful nanoarchitectonics of electrodes. Single-walled carbon nanotubes (SWCNT) can be employed as long-range nanowires, which connect the surface of electrode with the redox center of enzyme (1, 2). As one of the examples, SWCNTs were linked to gold electrode surface by covalent bond. On the other end, they were covalently linked to glucose oxidase redox center (2). The electron turnover rate transferred via SWCNT to the electrode surface was about 4,100 s−1, which is approximately six times higher than the turnover rate of electrons from the active site of GOx to its natural O2 electron acceptor (700 s−1). Therefore, electron transfer to SWCNT compete favorably with transfer to O2 and this makes such glucose sensor practically oxygen independent, which is very important in biosensing. In another work, Yu et al. (1) covalently linked enzymes onto the ends of vertically aligned SWCNTs forests (forests are meant in

K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_16, © Humana Press, a part of Springer Science + Business Media, LLC 2010

197

198

Pumera

this sense arrays of aligned nanoelectrodes), which were used as nanoelectrode arrays. Voltammetry of FeIII/FeII was observed for the iron heme enzymes myoglobin and horseradish peroxidase attached to carboxylated ends of the nanotube forests by amide linkages. It was suggested that the “trees” in the nanotube forest behaved electrically similar to a metal, conducting electrons from the external circuit to the redox sites of the enzymes. Carbon nanotube electrodes can incorporate enzymes in various ways. The most straightforward way is to incorporate enzyme in composite material where CNTs are mixed with some polymer binder. Wang et al. (3) combined CNT, Teflon binder, and enzyme and demonstrated that the electrocatalytic activity of CNT toward hydrogen peroxide and NADH permitted effective low-potential amperometric biosensing of glucose and ethanol in connection with the incorporation of glucose oxidase and alcohol dehydrogenase/NAD+ within the 3D CNT/Teflon matrix. It has been shown that accelerated electron transfer was tied with minimization of surface passivation. The advantages of CNT/Teflon composite devices were demonstrated with comparison to their graphite/Teflon counterparts, which clearly demonstrated higher sensitivity of CNT/Teflon biocomposite. CNT and glucose oxidase enzyme were incorporated in paste electrodes using oil as binder for glucose biosensing (4) and later on the variety of CNT/paste-incorporated enzymes was expanded towards lactate oxidase, polyphenol oxidase, and alcohol dehydrogenase/NAD+ (5). The vast majority of the articles in the area of carbon nanotube-based redox enzyme biosensors incorporate enzyme and carbon nanotubes in the organic or inorganic polymer binder without any chemical bond between enzyme and CNT and/or polymer matrix. Therefore, it is possible not to use polymer binder at all and it is possible to immobilize glucose oxidase on carbon nanotubes by noncovalent binding directly to create CNT-based enzymatic electrochemical glucose biosensor (6).

2. Materials 2.1. Construction of Carbon Nanotube and Graphite Electrodes

1. Carbon nanotubes with small diameter (about 5 nm), i.e., double-walled carbon nanotubes (DWCNT) (Sigma-Aldrich, Germany). 2. Graphite powder, particle size 50 mm (BDH, UK). 3. Nitric acid, diluted to 6 M concentration (Sigma-Aldrich). 4. Glassy carbon electrodes, diameter of 3 mm (Ecochemie, Netherlands).

Enzymatic Detection Based on Carbon Nanotubes

199

5. Abrasive paper and alumina paper (polishing strips) (Orion, Spain). 6. X-ray photoelectron spectrometer (XPS) for proof of the adsorption of glucose on carbon surface. 2.2. Chemicals, Analytes, and Electrolytes

1. Glucose oxidase enzyme (GOx) (Sigma-Aldrich). 2. Potassium dihydrogen phosphate and potassium hydrogen phosphate (Sigma-Aldrich). 3. Nitric acid (Sigma-Aldrich). 4. Hydrogen peroxide (30%) (Merck).

2.3. Electrochemical Biosensing

1. Platinum auxiliary electrode (BAS). 2. Double junction Ag/AgCl reference electrode (BAS). 3. Carbon materials (carbon nanotube and graphite) modified glassy carbon electrode as working electrode. 4. Autolab PGSTAT 20 (Eco Chemie, The Netherlands) connected to a personal computer for electrochemical measurements.

3. Methods 3.1. Fabrication of Noncovalently Modified Carbon Nanotube (Graphite) Glassy Carbon Electrodes

1. Purify the carbon nanotubes by stirring them in 2 M nitric acid at 25°C for 24 h. Wash the resulting mixture several times with distilled water until the pH of aqeuos solution reaches neutral pH. Dry carbon nanotubes. 2. Prepare dispersion of carbon nanotubes (0.5 mg/mL) and glucose oxidase (0.2 mg/mL) in distilled water. 3. Stir the solution of carbon nanotubes and glucose oxidase for 75 min. 4. Filter the solution via 0.2 mm membrane. 5. Check if the adsorption of glucose oxidase on carbon nanotube surface was successful using XPS (see Fig. 1). The presence of glucose oxidase will appear as signal for N 1s. High resolution XPS will provide confirmatory evidence of presence via strong signal of C–N bond (Fig. 2). High resolution XPS spectrum of DWCNT exhibits tailing peak at 284.5 eV, which is related to sp2 hybridization of carbon in double-wall carbon nanotube (Fig. 16.2a) (7, 8). The C 1s core-level spectrum of DWCNT/GOx shows more complex features (Fig. 16.2b) and it can be curve-fitted with four peak components, with binding energies at 284.5, 285.4, 286.1, and 287.6 eV assigned to the C–H, C–N, C–O and C=O species, respectively (9).

200

Pumera

Fig. 1. Wide scan X-ray photoelectron spectrum for DWCNT (black line) and noncovalently functionalized DWCNT/GOx (red line). Also shown detailed spectrum of N 1s in DWCNT/ GOx sample as inset. Reprinted with permission from (6)

Fig. 2. High-resolution X-ray photoelectron spectrum of C 1s core-level spectrum for DWCNT (a) and DWCNT/GOx (b). Reprinted with permission from (6)

All the above mentioned species are associated with amino acid residues existing in glucose oxidase protein. 6. Disperse the carbon nanotube/glucose oxidase at concentration 1 mg/mL and subject the suspension to ultrasonication for 1 min. 7. Drop the suspension on glassy carbon electrode. 8. Leave it to evaporate the solution to create random network of glucose oxidase modified carbon nanotubes on the surface of the glassy carbon electrode. 9. Repeat procedure 1–8 for graphite instead of carbon nanotubes to have a control material.

Enzymatic Detection Based on Carbon Nanotubes

3.2. Enzymatic Electrochemical Biosensing Using Carbon Nanotube/ Glucose Oxidase Electrodes

201

1. Glucose oxidase converts glucose to gluconic acid and uses oxygen as the cofactor, which is reduced to hydrogen peroxide as shown in Fig. 3. The hydrogen peroxide is the marker, which is actually detected on the electrode. 2. Prepare supporting electrolyte from phosphate salts. The resulting solution should have concentration of 50 mM and pH of 7.4. 3. Add hydrogen peroxide to the solution and test its electrochemical behavior on carbon nanotube (and graphite)modified electrode using cyclic voltammetry. The optimal concentration of hydrogen peroxide is about 5 mM. 4. Scan voltage from −200 mV to 1,100 mV with sweep rate of 50 mV s−1. (see Notes 1 and 2). 5. Determine electrochemical activation potential from cyclic voltammograms (for an example of cyclic voltammogram of hydrogen peroxide, see Fig. 4).

Fig. 3. Enzymatic oxidation of glucose on surface of glucose oxidase-modified carbon nanotubes. Reprinted with permission from (6)

Fig. 4. Cyclic voltammograms for 5 mM hydrogen peroxide using (a) graphite/GOx and (b) DWCNT/GOx electrodes. Conditions: Scan speed, 50 mV/s; supporting electrolyte, phosphate buffer (50 mM, pH 7.4). Reprinted with permission from (6)

202

Pumera

6. Figure 4 presents cyclic voltammograms of hydrogen peroxide (5 mM), which are performed using graphite/GOx (a) and DWCNT/GOx (b) electrodes. It is clearly observed that oxidation of hydrogen peroxide starts around +0.6 V in case of DWCNT/GOx-modified electrode while oxidation of H2O2 starts at potential about +0.8 V in case of graphite/ GOx-modified electrode. The cyclic voltammograms indicate that DWCNT-based electrode provides high surface compared to graphite and favorable low detection potential for detection of hydrogen peroxide, which is the coproduct of glucose oxidation by glucose oxidase enzyme. 7. Signal of DWCNT electrode towards oxidation of hydrogen peroxide is about 5–6 times higher than the one on graphite electrode. This observation is in line with surface-specific area measurements of DWCNT and graphite measured by BET method and which are 58.55 m2 g−1 for DWCNT and 11.07 m2 g−1 for graphite (about 5× larger surface area of DWCNT than of graphite). 8. The response graphite/GOx and DWCNT/GOx film electrodes toward glucose is scrutinized. Figure 5 presents hydrodynamic voltammograms for 10 mM glucose at the graphite/GOx (a) and DWCNT/GOx (b) film electrodes. (see Note 3). 9. No significant redox activity is observed for glucose at the graphite/GOx film electrode. This is because glucose oxidase enzyme is not adsorbed on graphite microparticles in large amount.

Fig. 5. Hydrodynamic voltammograms for 10 mM glucose working with (a) graphite/GOx and (b) DWCNT/GOx electrodes. Supporting electrolyte, phosphate buffer (50 mM, pH 7.4); stirring rate, ~550 rpm

Enzymatic Detection Based on Carbon Nanotubes

203

10. A significantly different situation prevails in the case of DWCNT/GOx film electrodes. Large oxidation currents, starting around +0.6 V, are observed for glucose at DWCNT/ GOx film electrode. The data of Fig. 5 indicate that the GOx redox enzyme was successfully adsorbed on surface of DWCNT. DWCNTs serve as a signal transducer. 11. The outstanding bioactivity of glucose oxidase adsorbed on DWCNT compared to standard graphite is in line with previous work employing enzyme/CNT composites. Asuri et al. (10) demonstrated that CNT–enzyme composites exhibited about 30-times higher overall enzymatic activity than control composites where the proteases were conjugated to nonnanostructure graphite. 12. Stability of the electrode is a fundamental issue. Investigate the stability of immobilized glucose oxidase at DWCNT. Figure 6 demonstrates an example of the amperometric signal of 2.5 mM glucose. No decrease should be observed over time periods in order of 10 min.

4. Notes 1. It is important to note that all electrochemical experiments used deionized water. 2. Solutions during acquiring cyclic voltammogram are static, not steered. 3. Amperometric sensing at fixed potential (i.e., observing stability of electrode) should be carried out in steered solution with magnetic stirrer.

Fig. 6. Response of DWCNT/GOx electrode to 2.5 mM glucose for stability test. Conditions: supporting electrolyte, phosphate buffer (50 mM, pH 7.4); stirring rate, ~550 rpm, detection potential of +0.85 V

204

Pumera

References 1. Yu X, Chattopadhyay D, Galeska I, Papadimitrakopoulos F, Rusling JF (2003) Peroxidase activity of enzymes bound to the ends of single-wall carbon nanotube forest electrodes. Electrochem Commun 5:408–411 2. Patolsky F, Weizmann Y, Willner I (2004) Long-range electrical contacting of redox enzymes by SWCNT connectors. Angew Chem Int Ed 43:2113–2117 3. Wang J, Musameh M (2003) Carbon nanotube/teflon composite electrochemical sensors and biosensors. Anal Chem 75:2075–2079 4. Rubianes MD, Rivas GA (2003) Carbon nanotubes paste electrode. Electrochem Commun 5:689–694 5. Rubianes MD, Rivas GA (2005) Enzymatic biosensors based on carbon nanotubes paste electrodes. Electroanal 17:73–78 6. Pumera M, Smid B (2007) Redox protein noncovalent functionalization of double-wall

7.

8.

9.

10.

carbon nanotubes: electrochemical binder-less glucose biosensor. J Nanosci Nanotechnol 7:3590–3595 Pirlot C, Willems I, Fonseca A, Nagy JB, Delhalle J (2002) Preparation and characterization of carbon nanotube/polyacrylonitrile composites. Adv Eng Mater 4:109–114 Vast L, Philippin G, Destrée A, Moreau N, Fonseca A, Nagy JB, Delhalle J, Mekhalif Z (2004) Chemical functionalization by a fluorinated trichlorosilane of multi-walled carbon nanotubes. Nanotechnology 15:781–785 Liu X, Neoh KG, Cen L, Kang ET (2004) Enzymatic activity of glucose oxidase covalently wired via viologen to electrically conductive polypyrrole films. Biosens Bioelectron 19:823–834 Asuri P, Karajanagi SS, Kane RS, Dordick JS (2007) Polymer-nanotube-enzyme composites as active antifouling films. Small 3:50–53

Chapter 17 Carbon Nanotube Biosensors Based on Electrochemical Detection Martin Pumera Abstract The advantages of carbon nanotubes, such as high surface area, favorable electronic properties, and electrocatalytic effect, attracted considerable attention very recently for the construction of electrochemical biosensors. We describe here the construction and application of carbon nanotube/epoxy rigid polymer composite electrochemical biosensor for the detection of important biomarkers, such as NADH and hydrogen peroxide. Key words: Carbon nanotubes, Biosensing, Electrochemistry, Composite

1. Introduction Distinctive properties of carbon nanotubes (CNT), such as a high surface area, ability to accumulate analyte, minimization of surface fouling, and electrocatalytic activity, are very attractive for electrochemical sensing (1, 2). Recent studies demonstrated that CNT exhibits strong electrocatalytic activity for a widerange of compounds, such as NADH (3), hydrogen peroxide (4), neurotransmitters (5), cytochrome c (6), amino acids (7), and DNA (8). Most of the CNT-based electrodes for electroanalytical applications are based on physical adsorption of CNT onto electrode surfaces, usually glassy carbon. However, it is important to note that CNT dispersed in mineral oil (8) or incorporated into Teflon (9) have been used. The method described in this chapter deals with the outstanding properties of CNT incorporated into an epoxy polymer, forming an epoxy composite hybrid material as a new electrode with

K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_17, © Humana Press, a part of Springer Science + Business Media, LLC 2010

205

206

Pumera

improved electrochemical sensing properties (10). This CNT–epoxy composite is simple, cheap, and results in interesting electrode material. The preparation and characterization of this multifunctional material is discussed, and electrochemical responses to important biomarkers such as NADH (b-nicotinamide adenine dinucleotide reduced form) and hydrogen peroxide are studied.

2. Materials 2.1. Construction of Carbon Nanotube/ Epoxy Composite Electrodes

1. Epoxy resin Epotek H77A and hardener Epotek H77B (Epoxy Technology, Billerica, MA, USA). 2. Multi-wall carbon nanotubes (CNT-200: length, 0.5–200 µm; O.D., 30–50 nm, wall thickness, 12–18 nm; and CNT-2: length, 0.5–2 µm; O.D., 20–30 nm; wall thickness, 1–2 nm; both produced by chemical vapor deposition method) with ~95% purity (Sigma-Aldrich). 3. Graphite powder (particle size 50 µm, BDH, U.K). 4. Nitric acid, diluted to 2 M concentration (Sigma-Aldrich). 5. Cylindrical PVC sleeve (6 mm i.d., 8 mm o.d., and 16 mm long). 6. Copper disk (6 mm o.d. and 0.5 mm thickness). 7. Abrasive paper and alumina paper (polishing strips) (Orion, Spain). 8. Spatula and standard laboratory glassware.

2.2. Analytes and Electrolytes

1. b-Nicotinamide adenine dinucleotide reduced form (NADH) (Sigma-Aldrich). 2. Potassium dihydrogen phosphate and potassium hydrogen phosphate (Sigma-Aldrich). 3. Nitric acid (Sigma-Aldrich). 4. Hydrogen peroxide (30%) (Merck, Germany).

2.3. Electrochemical Biosensing

1. Platinum auxiliary electrode (BAS). 2. Double-junction Ag/AgCl reference electrode (BAS). 3. Carbon nanotube/epoxy composite electrode as working electrode. 4. Autolab PGSTAT 20 (Eco Chemie, Netherlands) connected to a personal computer for electrochemical measurements. 5. Magnetic stirrer.

Carbon Nanotube Biosensors Based on Electrochemical Detection

207

3. Methods 3.1. Construction of Carbon Nanotube/ Epoxy Composite Electrodes

1. Purify the carbon nanotubes by stirring them in 2 M nitric acid at 25°C for 24 h. 2. Clean the copper disk by dipping it in HNO3/water solution (1:1) in order to remove copper oxide on the surface and rinse it well with bidistilled water. 3. Connect female of 2 mm of diameter and place a metallic thread and then solder this connection to the center of the copper disk. 4. Introduce this connection into the cylindrical PVC sleeve (6 mm i.d., 8 mm o.d., and 16 mm long). The metallic thread allows the connection to remain well-fixed at the end of the cylindrical PVC sleeve (Fig. 1). 5. Mix epoxy resin and hardener manually with spatula in the ratio 20:3 (w/w). 6. Carbon nanotubes/epoxy electrodes have been produced by loading the epoxy resin, before curing, with 20% w/w of carbon nanotubes. 7. When the resin and hardener are well-mixed, the carbon nanotubes should be added and mixed for 30 min. Place the resulting paste into a cylindrical PVC sleeve (6 mm i.d.). Electrical contact should be completed using copper disk and wire. 8. Cure conducting composite is to be cured at 40°C for 1 week. 9. Prior to use, the surface of the electrode should be polished with abrasive paper and then with alumina paper (see Note 1). 10. In the following text, the carbon nanotube/epoxy composite electrodes based on CNT-200 nanotubes will be marked as CNT-200-EC and the ones based on CNT-2 nanotubes will be marked as CNT-2-EC.

Fig. 1. Details on preparation of carbon nanotube/epoxy composite electrodes

208

Pumera

11. In the same way as described above (steps 1–9), prepare graphite/epoxy composite electrode (GEC) for comparative purposes. 3.2. Electrochemical Biosensing Using Carbon Nanotube/ Epoxy Composite Electrodes

1. Prepare supporting electrolyte from phosphate salts. The resulting solution should have a concentration of 50 mM and pH of 7.4 (see Note 2). 2. Add biomarker in the solution and test its electrochemical behavior using cyclic voltammetry. The optimal concentration of NADH and hydrogen peroxide is about 5 mM. 3. Scan voltage from −200 mV to 1,100 mV with sweep rate of 100 mV/s (see Note 3). 4. Determine electrochemical activation potential from cyclic voltammograms (for an example of cyclic voltammogram of NADH, see Fig. 2). 5. Figure 2 compares cyclic voltammograms for 1 mM NADH at carbon nanotube CNT-200-EC (a) and CNT-2-EC (b) electrodes with graphite epoxy composite electrode (c). A broad NADH oxidation peak is observed at +0.72 V (vs Ag/AgCl) for GEC electrode. For CNT-2-EC electrode, well-developed oxidation peak is observed at the very similar potential +0.74 V. Substantial shift to +0.45 V in the oxidation potential was observed at CNT-200-EC. It is interesting to note that the large potential shift to the lower potentials (similar to those observed by Musameh et al. (3)) is observed only at carbon nanotube CNT-200-EC electrode, while NADH oxidation potentials at CNT-2-EC and graphite powder epoxy electrodes are similar and they remain relatively high.

Fig. 2. Cyclic voltammograms for 1 mM NADH of the (a, b) carbon nanotube/epoxy composite electrodes and (c) graphite electrodes. Reprinted with permission from (10)

Carbon Nanotube Biosensors Based on Electrochemical Detection

209

6. In the same way add hydrogen peroxide to supporting electrolyte in the new electrochemical cell. Optimum concentration of hydrogen peroxide biomarker is 5 mM. 7. Scan voltage from −200 mV to 1,100 mV with sweep rate 100 mV/s. 8. Determine electrochemical activation potential from cyclic voltammograms (for an example of cyclic voltammogram of hydrogen peroxide, see Fig. 3). 9. Figure 3 displays cyclic voltammograms for 1 mM hydrogen peroxide at carbon nanotube CNT-200-EC (a), CNT-2-EC (b), and GEC (c) electrodes. The long-nanotube CNT200-EC electrode displays oxidation signal around +0.60 V (a), while the short-nanotube CNT-2-EC and graphite epoxy composite electrode (b and c, respectively) shows oxidation around +0.70 V. CNT-200-EC electrode also shows higher sensitivity towards oxidation of hydrogen peroxide (note the different scales). 10. Important characteristic of carbon nanotube-based electrochemical biosensor is the fact that it is prone to passivation. The oxidation of NADH at conventional solid electrodes is prone to the problems with passivation of the surface (3), the stability of response of the electrodes should be evaluated. 11. Figure 4 compares stability of oxidation of NADH at +0.55 V for CNT-200-EC (a), CNT-2-EC (b) and GEC (c) electrodes. 12. The results should indicate that carbon nanotube epoxy composite electrodes provide a good stability towards oxidation of NADH. See an example in Fig. 4.

Fig. 3. Cyclic voltammograms for 1 mM hydrogen peroxide of the (a, b) carbon nanotube/epoxy composite electrodes and (c) graphite electrodes. Reprinted with permission from (10)

210

Pumera

Fig. 4. Stability of 1 mM NADH response using the (a) CNT-200-EC, (b) CNT-2-EC, and (c) GEC electrodes. Operation potential, +0.55 V. Solution stirring ca. 400 rpm. 50 mM phosphate buffer, pH 7.4. Reprinted with permission from (10)

13. Figure 4 shows that carbon nanotube/epoxy composites hold about 96% (CNT-200-EC) and 93% (CNT-2-EC) of original response even after 45 min of immersion in the solution of NADH, while response of GEC electrode decreases by 33% in the same timescale. Similar passivation layer as on the surface of GEC electrode also shows glassy carbon electrode (3, 11). 14. The amperometric response of CNTEC and GEC electrodes towards NADH should be assessed from cyclic voltammograms. In our example in Fig. 2, it was assessed at +0.55 V (see Note 4). 15. Calibration curve should be constructed for both NADH and hydrogen peroxide by adding 100 µM increments of these analytes to electrochemical cell. Carbon nanotube-based electrodes will displays a well-defined concentration dependence over concentration range 0.0–1.0 mM (0.1 mM increments), with sensitivities in tens µA/mM with high correlation coefficients (about 0.995) for NADH. 16. For hydrogen peroxide, the concentration range 0.0–2.0 mM (0.2 mM increments) should provide linear calibration range with of tens µA/mM with correlation coefficients about 0.995. 17. The hydrogen peroxide response was not affected by regenerating (polishing) the surface. 18. Reproducibility of biosensor should be tested. Series of eight successive measurements, each recorded on a newly polished surface will yield to RSD = 5%.

Carbon Nanotube Biosensors Based on Electrochemical Detection

211

Fig. 5. SEM images for long-carbon nanotube-based epoxy electrode (a), short-carbon nanotube-based epoxy electrode (b) and conventional GEC electrode (c). All electrode surfaces have been polished in the same way as explained in the text. The same acceleration voltage (5 kV) and resolution are used. Reprinted with permission from (10)

4. Notes 1. It is important to characterize carbon nanotube/epoxy composites with scanning electrochemical microscopy (SEM) in order make sure that electrode was fabricated successfully. For an example, see Fig. 5. 2. It is important to note that all electrochemical experiments used bidistilled water. 3. As a standard, solutions during acquiring cyclic voltammogram are static, not stirred. 4. On other hand, amperometric sensing at fixed potential (i.e., observing stability of electrode) should be carried out in steered solution with magnetic stirrer (approximately 400 rpm). References 1. Wang J, Liu G, Jan MR (2004) Ultrasensitive electrical biosensing of proteins and DNA: carbon-nanotube derived amplification of the recognition and transduction events. J Am Chem Soc 126:3010 2. Patolsky F, Weizmann Y, Willner I (2004) Long-range electrical contacting of redox enzymes by SWCNT connectors. Angew Chem Int Ed 43:2113 3. Musameh M, Wang J, Merkoçi A, Lin Y (2002) Low-potential stable NADH detection

at carbon-nanotube-modified glassy carbon electrodes. Electrochem Commun 4:743 4. Hrapovic S, Liu Y, Male KB, Luong JHT (2004) Electrochemical biosensing platforms using platinum nanoparticles and carbon nanotubes. Anal Chem 76:1083 5. Luo H, Shi Z, Li N, Gu Z, Zhuang Q (2001) Investigation of the electrochemical and electrocatalytic behavior of single-wall carbon nanotube film on a glassy carbon electrode. Anal Chem 73:915

212

Pumera

6. Wang J, Li M, Shi Z, Li N (2002) Direct electrochemistry of cytochrome c at a glassy carbon electrode modified with single-wall carbon nanotubes. Anal Chem 74:1993 7. Wang JX, Li MX, Shi ZJ, Li NQ, Gu ZN (2004) Electrochemistry of DNA at single-wall carbon nanotubes. Electroanalysis 16:140 8. Pedano ML, Rivas GA (2004) Adsorption and electrooxidation of nucleic acids at carbon nanotubes paste electrodes. Electrochem Commun 6:10

9. Wang J, Musameh M (2003) Carbon nanotube/ teflon composite electrochemical sensors and biosensors. Anal Chem 75:2075 10. Pumera M, Merkoci A, Alegret S (2006) Carbon nanotube-epoxy composites for electroccihemical sensing. Sens Actuators B Chem 113:617 11. Lawrence NS, Deo RP, Wang J (2005) Comparison of the electrochemical reactivity of electrodes modified with carbon nanotubes from different sources. Electroanalysis 17:65

Chapter 18 Biosensors Based on Carbon Nanotube-Network Field-Effect Transistors Cristina C. Cid, Jordi Riu, Alicia Maroto, and F. Xavier Rius Abstract We describe in detail the different steps involved in the construction of a carbon nanotube field-effect transistor (CNTFET) based on a network of single-walled carbon nanotubes (SWCNTs), which can selectively detect human immunoglobulin G (HIgG). HIgG antibodies, which are strongly adsorbed onto the walls of the SWCNTs, are the basic elements of the recognition layer. The nonspecific binding of proteins or other interferences are avoided by covering the nonadsorbed areas of the SWCNTs with Tween 20. The CNTFET is a reagentless device that does not need labels to detect HIgG. Key words: Carbon nanotubes, Field-effect transistor, CNTFET, HIgG

1. Introduction Single-walled carbon nanotubes (SWCNTs) are a type of nanostructured materials that have been successfully used in the past years for sensing devices. One specific type of electrochemical SWCNT sensors is the so-called carbon nanotube field-effect transistors (CNTFETs). In a CNTFET, a unique semiconductor SWCNT or a network of SWCNTs is connected to two metal electrodes, source and drain, and by applying a source-to-drain voltage, an electronic current flows through the SWCNTs. This current is further modulated with the presence of a third electrode, the gate electrode (Fig. 1). The gate electrode is usually a doped substrate (e.g., Si) placed in indirect contact to the channel made using the SWCNTs through an insulating material (e.g., SiO2) of few hundreds of nanometres thick. When applying positive gate voltages, the conductance (or current) of the SWCNTs can be reduced by several orders of magnitude up to practically

K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_18, © Humana Press, a part of Springer Science + Business Media, LLC 2010

213

214

Cid et al.

Fig. 1. Schematic configuration of a CNTFET device

Fig. 2. CNTFET transfer characteristics. (a) Threshold voltage, (b) transconductance, (c) maximum conductance, and (d) modulation

an insulating state because of the semiconducting p-type electrical behavior (1) of SWCNTs. If a constant bias voltage between source and drain electrodes is applied while sweeping the gate voltage, the resulting curve of conductance (or current) versus gate voltage is the so-called device characteristics. From this curve (Fig. 2), one can determine some typical parameters of the CNTFET, such as the maximum conductance value, the modulation (the signal compared to the noise level), the transconductance at zero gate voltage (indicating the kind of carrier mobility: electrons or holes), and the threshold voltage. One of the main advantages of SWCNTs is their high sensitivity to the presence of different molecules in their surrounding environment. These molecules can provide or withdraw charges to the electronic layer of the SWCNT, changing its conductance. Nevertheless, SWCNT sensitivity is also a drawback because

Biosensors Based on Carbon Nanotube-Network Field-Effect Transistors

215

molecules in the surrounding environment can interfere in the determination of the target analyte, modifying the selectivity of the sensing device. Therefore, it is necessary to protect the SWCNTs (by means of a specific functionalization process) to become selective only to the target analyte. CNTFET developed so far are able to selectively detect proteins (2–5), and among them, some immunoglobulins (6–8), DNA sequences (9) or bacteria (10, 11). SWCNTs are usually synthesized by using the chemical vapor deposition (CVD) method and integrated as the semiconductor channel of CNTFETs. The determination of human immunoglobulin G (HIgG) is very important in diagnosing illness, as several diseases are accompanied by changes in the concentration of this immunoglobulin. Moreover, HIgG has become a model for large charged molecules at physiological pH. The selective detection of HIgG can be therefore performed by functionalizing the SWCNTs with HIgG antibodies. HIgG antibodies are adsorbed over the surface of the SWCNTs, and the nonspecific binding of proteins and the effect of other interferences are avoided by covering the nonadsorbed areas of the SWCNTs with a suitable blocking molecule, like for instance Tween 20. In this way, we take advantage of the molecular recognition mechanism between an antibody and its antigen to selectively detect our target analyte (7). The functionalization process is represented in Fig. 3. The development of these CNTFET-based sensors can be divided into four main steps: (1) the synthesis of nanotubes, (2) the construction of the CNTFET, (3) the functionalization of SWCNTs, and (4) the process of recording the instrumental responses to characterize the CNTFET device. These CNTFETbased sensors can be used for the selective detection of many kinds of biomolecules by using a suitable molecular receptor selective to the target analyte. For instance, our research group has also developed CNTFET devices for the selective detection of Salmonella Infantis using anti-Salmonella antibodies (11) and xenoestrogens using a cellular receptor (12).

Fig. 3. Functionalization process of the SWCNTs for the selective detection of HIgG

216

Cid et al.

2. Materials 2.1. Synthesis of Single-Walled Carbon Nanotubes

1. Horizontal split tube furnace HST 12/600 (Carbolite , Hope Valley, UK). 2. Quartz tubular reactor (4 × 120 cm) (Afora, Barcelona, Spain). 3. Ultrasonic bath 100 W (Selecta, Barcelona, Spain). 4. Spin coater WS-400B-6NPP/LITE (Laurell Technologies Corporation, North Wales, PA, USA). 5. Mass controllers for methane and hydrogen including PC software (Bronkhorst, Ruurlo, The Netherlands). 6. Deionised and charcoal-treated water (18.2 MΩ cm specific resistance) obtained with Milli-Q PLUS reagent-grade water system (Millipore, Billerica, MA, USA). 7. Substrates of Si/SiO2. 500 nm of SiO2 grown thermally over n-type low resistivity Si. Dimensions: 1 × 1 cm (for screen printed CNTFETs, see Subheading 18.3.2) and 0.5 × 0.5 cm (for photolithographed CNTFETs, see Subheading 18.3.2) (D+T Microelectrónica, National Microelectronics Centre, Bellaterra, Spain).

2.2. CNTFET Fabrication

1. Conductive solution Electrodag® 1415M (Acheson Industries, Scheemdam, Netherlands). 2. Photolithography system (D+T Microelectrónica, National Microelectronics Centre, Bellaterra, Spain).

2.3. Functionalization Procedures

1. HIgG (heavy & light chain) antibody affinity purified produced in goat (Bethyl Laboratories Inc., Montgomery, TX, USA). Antibody concentration: 1 mg/mL. Physical state: liquid. Buffer: phosphate-buffered saline (PBS), pH 7.2. Preservative: 0.1% sodium azide (see Note 1). 2. Dulbecco PBS, 150 mM, pH: 7.1–7.5 (Sigma, Tres Cantos, Spain). 3. Deionised and charcoal-treated water (18.2 M W cm specific resistance) obtained with Milli-Q PLUS reagent-grade water system (Millipore, Billerica, MA, USA).

2.4. CNTFET Characterization

1. Precision semiconductor parameter analyzer Agilent 4157A (Agilent Technologies, Las Rozas, Spain). 2. MP 1008 manual probe station (Wentworth Laboratories, Sandy, UK) placed within a Faraday box (SIRM, Badalona, Spain) over an anti-vibration table (SIRM, Badalona, Spain).

Biosensors Based on Carbon Nanotube-Network Field-Effect Transistors

217

3. Methods 3.1. Synthesis of Single-Walled Carbon Nanotubes (4, 13, 14)

1. Prepare the catalyst solution. 100 mg/L of iron nitrate in isopropanol sonicated (ultrasonic bath, 100 W) about 10 min (see Note 2). 2. Clean the Si/SiO2 substrate by one cycle of consecutive immersions in the following solvents: acetone, isopropanol, and finally deionised water in the ultrasonic bath. Time of bath: approximately 5 min for each solvent. Dry under a gentle flow of nitrogen. 3. Spin-coat three times 20 ml of the catalyst solution over the surface of the substrate of Si/SiO2 when it spins at 100 g, and leave it spinning until the solvent completely evaporates (about 30 s). 4. Introduce the substrate into the quartz reactor (Figs. 4 and 5) (see Note 3). Set the furnace temperature at 900°C. An argon current of 1,000 sccm, acting as purging gas, flows through the reactor from room temperature up to 900°C. Once 900°C is reached, the argon flow is turned off and at the same time the flows of methane and hydrogen are opened. The methane and hydrogen flows are then kept constant (600 and 200 sccm, respectively) for about 20 min. After this time, the methane and hydrogen flows are turned off and the argon flow is turned on again while the system cools to room temperature (see Notes 4 and 5). 5. At this point, a network of SWCNTs is grown over the silica substrates.

Fig. 4. Image of the split tube furnace with the quartz reactor at 900°C used for the synthesis of SWCNTs using the chemical vapor deposition method

218

Cid et al.

Fig. 5. Schematic of the experimental setup for the chemical vapor deposition method

3.2. CNTFET Fabrication

To measure the electric current flowing through the SWCNTs network, the instrumental signal that we will relate to the concentration of the target analyte, electrical contacts are placed at both ends of the selected network. These two metal contacts, the source and drain electrodes, have been patterned using two different procedures. One involves a photolithographic process and the other was done using homemade screen printing. 1. As source and drain electrodes, double-layered chromium/ gold (2/30 nm) are used (Fig. 6). The gold layer (upper layer) is used as the electrode metal since it remains unaltered when it comes in contact with air due to its noble metal characteristics. The chromium layer (lower layer, placed directly over the Si/SiO2 substrate containing SWCNTs) is needed so the gold layer adheres properly to the Si/SiO2 substrate. The distance between pairs of Cr/Au electrodes (10 × 5 mm) is 2–3 mm. The metal pads (Fig. 6 left) for contacting the tips of the probe station are 200 × 200 mm. 2. For screen printing (Fig. 7), a conductive solution (commercially called Electrodag® 1415M, made of silver and 4-methylpentan-2-one) is placed manually over the Si/SiO2 substrate containing SWCNTs using a silicone mask. The mask allows the metallic electrodes to be printed at the desired sites. Subsequently, this substrate (Si/SiO2 with SWCNTs and metallic electrodes) is cured at 150°C for 10 min. With this process, the desired distance between source and drain electrodes can be achieved with lower precision than with optical lithography. The distance between the metallic electrodes ranged from 1 to 5 mm, according to what the device was

Biosensors Based on Carbon Nanotube-Network Field-Effect Transistors

219

Fig. 6. Electrodes deposited photolithographically. Picture of the whole substrate containing four pairs of metal electrodes (left) and a zoom with an environmental electron scanning microscope (right)

Fig. 7. Home screen-printed electrodes

Fig. 8. Tween 20 structure (w + x + y + z = 20)

to be used for. This is a fast and inexpensive method to roughly place electrodes in the laboratory. 3.3. Functionalization Procedures

The antibody is directly immobilised over the surface of the SWCNTs (8, 15–17). In this method, there are two functionalization steps. The first step is the adsorption of antibody molecules, which act as the sensing part of the CNTFET, over the SWCNTs (7). The second step is preventing the nonselective binding of undesired substances, such other proteins in the sample. This latter step is performed by immobilising a small molecule, Tween 20 (Fig. 8), over the remaining free surface of SWCNTs that has not been coated by the antibodies.

220

Cid et al.

1. Prepare a diluted solution (10 mg/L) from the commercial antibody (1,000 mg/L) in PBS solution (15 mM, pH = 7.4). 2. Immerse the CNTFET (built according Subheading 18.3.2) in the prepared antibody solution overnight to have the antibodies adsorbed over the SWCNTs. Subsequently, rinse with deionised water twice and dry with nitrogen. 3. Prepare a solution of 0.5% of Tween 20 in PBS. Immerse the CNTFET containing the HIgG antibodies adsorbed over the SWCNTs in the Tween 20 solution for 2 h. Subsequently, rinse with deionised water twice. 4. The devices are ready for detecting the HIgG (see Note 6). 3.4. CNTFET Characterization 3.4.1. Microscopy Characterization

Several powerful microscopic techniques are used to view onedimensional nanometric structures, such as SWCNTs. We used scanning electron microscopy (SEM), environmental scanning electron microscopy (ESEM), and atomic force microscopy (AFM) to characterize several parameters: 1. Density of the network of SWCNTs: for medium-high density networks of SWCNTs, SEM or ESEM is preferred. For lowdensity networks, AFM allows a more detailed characterization of the structure of SWCNTs (Fig. 9). 2. Diameter of individual SWCNTs: only AFM is suitable for measuring individual diameters. If the density is very high, it can be difficult to measure individual diameters. 3. Lengths: both SEM/ESEM and AFM are suitable, but AFM gives more accurate results.

3.4.2. Electrical Characterization

The characterization of the electrical properties of SWCNT networks is an essential step in assessing their behavior as the transducer part in FET-based sensors. It has been established

Fig. 9. Typical AFM image of a medium-low density network of SWCNTs

Biosensors Based on Carbon Nanotube-Network Field-Effect Transistors

221

that synthesis with CVD can produce both metallic and semiconducting SWCNTs. This creates nanotubes of both types in random spaghetti-like network. This network is grown on the SiO2 layer of the Si/SiO2 substrate. With this configuration, we use the silicon substrate as a back gate, i.e., the third electrode of the FET. The SiO2 acts as a dielectric layer, isolating nanotubes electrically from the silicon. The electrical properties of the CNTFET were measured using a semiconductor parameter analyzer at room temperature. Source and drain electrodes are contacted with thin tungsten tips of the probe station, and the tips are then connected to the semiconductor parameter analyzer. Using electrical characterization, we can check every step of CNTFET construction, from the synthesis of SWCNTs to the individual steps of the functionalization process. 1. Check the existence of an electrical channel that connects source and drain electrodes. Apply a source-to-drain sweep voltage (Vsd) (e.g., +0.25 to −0.25 V) (see Note 7) and record the electrical current value (I) at each Vsd while keeping constant at 0 V the gate voltage (Vg). With this test, we obtain: (a) the resistance (R) of the channel (R = Vsd/I) as the slope of the curve I versus Vsd (typical resistance values range from some kΩ to a few MΩ. These values depend, among other, on the density of SWCNTs and on the distance between electrodes (18)), (b) the optimum value of Vsd (see Note 8), and (c) the sourceto-gate current (“gate leakage”) at Vg = 0 V. If the values of the gate leakage (the electrical current value measured at the gate electrode while keeping Vg = 0 V) are above 1 nA, then the device cannot be used (see Note 9). If the three parameters we obtain with this test are adequate, there is a suitable electrical channel connecting source and drain electrodes. 2. Register the electrical current at a fixed Vsd (found in the previous paragraph) while sweeping the Vg (e.g., +5 to −5 V). This curve determines whether the nanotube network exhibits a semiconducting or metallic behavior response, and is referred to as the “device characteristics” (see Fig. 10) (see Note 10). An ideal CNTFET should have a modulation (Fig. 2) as high as possible (typical values are modulations of three to five orders of magnitude for low density networks and modulations of 1 order of magnitude for dense networks) and a threshold voltage (Fig. 2) at positive Vg values very close to zero (what means that the SWCNTs are in the off-state). If these two parameters are adequate, the CNTFET can be used for sensing purposes. The best value of Vg is usually selected according to the value that produces the highest source-todrain current, and at the same time maintains the capacity of modulate (i.e., with small variations of the gate voltage, a high change in the source–drain conductivity is observed).

222

Cid et al.

Fig. 10. Characteristic curve of a p-type semiconductor obtained from a typical CNTFET made with a dense network of SWCNTs. Source-to-drain current versus gate voltage

Fig. 11. Time dependence of the source–drain current, I (at Vg = −4 V and Vsd = 0.25 V), for increasing concentrations of HIgG. Arrows indicate the addition points of HIgG and the total concentration in the cell

3.5. HIgG Detection

1. Perform the measurements with the target solution containing the analyte, employing the best values of Vsd and Vg found in the previous section, and record the electrical current (ordinate axis) versus time (abscissa axis). Vsd and Vg values may vary slightly depending on the specific CNTFET used. However, the different CNTFET obtained using the same procedure should provide the response signal within the same logarithmic unit. 2. Liquid measurements are performed using a homemade silicone cell (see Note 11). The home made silicone cell has to

Biosensors Based on Carbon Nanotube-Network Field-Effect Transistors

223

be placed between the source and drain electrodes without touching them. 3. Although the change in the electrical current takes place within few minutes after the addition of each solution in the silicone cell, between each addition, the system has to be left undisturbed for about 5 min so that the electrical signal can stabilize. Figure 11 shows a typical experiment for detecting increasing concentrations of HIgG in solution. The electrical current starts to decrease within 60 s after each addition of HIgG (total concentration values from 1.25 up to 8.75 mg/L, i.e., from 8 up to 56 nM) and shows a stabilization of the signal about 10 min after each addition. The slight changes of the electrical current at 11.25 mg/L of HIgG were only due to the variability of the electrical current. Therefore, we concluded that almost all the anti-HIgG antibodies were already bound with HIgG at 8.75 mg/L HIgG. 4. Dry measurements are performed by immersing the CNTFET in the solution to be tested for 10 min. After being exposed to each solution, the devices are thoroughly rinsed with water and dried with nitrogen (see Note 12).

4. Notes 1. Storage temperature: 2–8°C. 2. The catalyst solution must be translucent and homogeneous without any kind of aggregate or precipitate. The solution can be kept in use for 1 month at room temperature if it maintains the mentioned properties. 3. The substrates have to be placed in the last third of the tubular oven. The gases should arrive at the substrate placement at maximum temperature. 4. Methane and hydrogen flow must be carefully controlled; this is why we used mass controllers. Since argon is the carrier gas, its flow can be controlled with a manometer. 5. Catalyst concentration and synthesis time (when methane and hydrogen gases flow through the reactor) influence the overall density of the network of SWCNTs. 100 mg/L of iron nitrate in isopropanol is the concentration normally used to obtain a medium-high density of SWCNTs. Lower concentrations, i.e., 50 and 25 mg/L, were used whenever a low density of SWCNTs was required. Additionally, the time of synthesis is directly related to the density of the nanotubes. If the density of the network is too high at 20 min of synthesis, it could be reduced down to 10 or 5 min.

224

Cid et al.

6. If the functionalized CNTFET devices are not used then, they must be kept dry and refrigerated (2–8°C). They can be stored at most for 24 h, but preferably the CNTFET devices should be used just after the functionalization process. 7. High Vsd sweeping values have to be avoided in order to prevent burning the SWCNTs. 8. In our devices, we considered the optimum Vsd value to be that which produces electrical current values of few microampers. This Vsd value is usually found between 0.1 and 0.25 V. 9. Sometimes a large gate leakage current is due to a high density of SWCNTs over the SiO2 layer. It can be reduced by isolating the network under study: a thin surface cut is made around the network of the SWCNTs containing the source and drain electrodes. 10. By observing the magnitude and polarity of the gate dependence, we can assess if the network grown by CVD behaves as a p-type semiconducting field effect transistor in air ( (4), (19)). The p-type behavior is due to the electron withdrawing effect of adsorbed O2 from the environment over the SWCNTs. In air, SWCNT conduction is made through positive carriers (holes), because O2 withdraws electrons from the surface of the SWCNT. When polarizing the system through the gate electrode, this hole conduction can be increased or decreased. If the hole conduction is completely decreased, which can happen at positive voltages in air, the SWCNTs are said to reach the off-state. 11. The homemade silicone cell is made by making a hole in a commercial silicone film. Typical dimensions of the cell are: inner diameter: 3 mm, outer diameter: 5 mm, with a height <5 mm. 12. The advantage of using dry measurements is that the ionic strength of the sample does not influence the measurements. References 1. Collins PG, Bradley K, Ishigami M, Zettl A (2000) Extreme oxygen sensitivity of electronic properties of carbon nanotubes. Science 287:1801–1804 2. Byon HR, Choi HC (2006) Network singlewalled carbon nanotube-field effect transistors with increased schottky contact area for highly sensitive biosensor applications. J Am Chem Soc 128:2188–2189 3. Shim M, Shi-Kam NW, Chen RJ, Li Y, Dai H (2002) Functionalization of carbon nanotubes for biocompatibility and biomolecular recognition. Nano Lett 2:285–288

4. Star A, Gabriel JP, Bradley K, Grüner G (2003) Electronic detection of specific protein binding using nanotube FET devices. Nano Lett 3:459–463 5. So HM, Won K, Kim YH, Kim BK, Ryu BH, Na PS, Kim H, Lee JO (2005) Single-walled carbon nanotube biosensors using aptamers as molecular recognition elements. J Am Chem Soc 127:11906–11907 6. Chen RJ, Bangsaruntip S, Drouvalakis KA, Kam NWS, Shim M, Li Y, Kim W, Utz PJ, Dai H (2003) Noncovalent functionalization of CNT for highly specific electronic

Biosensors Based on Carbon Nanotube-Network Field-Effect Transistors

7.

8.

9.

10.

11.

12.

biosensors. Proc Natl Acad Sci USA 100: 4984–4989 Cid CC, Riu J, Maroto A, Rius FX (2008) Carbon nanotube field effect transistors for the fast and selective detection of human immunoglobulin G. Analyst 133:1005–1008 Maehashi K, Katsura T, Kerman K, Takamura Y, Matsumoto K, Tamiya E (2007) Label-free protein biosensor based on aptamer-modified carbon nanotube field-effect transistors. Anal Chem 79:782–787 Star A, Tu E, Niemann J, Gabriel J-CP, Joiner CS, Valcke C (2006) Label-free detection of DNA hybridization using carbon nanotube network field-effect transistors. Proc Natl Acad Sci USA 103:921–926 So H-M, Park D-W, Jeon E-K, Kim Y-H, Kim BS, Lee C-K, Choi SY, Kim SC, Chang H, Lee J-O (2008) Detection and titer estimation of Escherichia coli using aptamer-functionalized single-walled carbon-nanotube field-effect transistors. Small 4:197–201 Villamizar R, Maroto A, Rius FX, Inza I, Figueras MJ (2008) Fast detection of Salmonella Infantis with carbon nanotube field effect transistors Biosens Bioelectron 24:279–283 Sánchez-Acevedo Z, Riu J, Rius FX (2009) Fast picomolar selective detection of bisphenol A in water using a carbon nanotube

225

field effect transistor functionalized with estrogen receptor-a Biosens Bioelectron 24: 2842–2846 13. Kong J, Soh HT, Cassell AM, Quate CF, Dai H (1998) Synthesis of individual single-walled carbon nanotubes on patterned silicon wafers. Nature 395:878–881 14. Dai H (2002) Carbon nanotubes: synthesis, integration, and properties. Acc Chem Res 35:1035–1044 15. Su X, Chew FT, Li SFY (1999) Self-assembled monolayer-based piezoelectric crystal immunosensor for the quantification of total human immunoglobulin E. Anal Biochem 273: 66–72 16. Veetil JV, Ye K (2007) Development of immunosensors using carbon nanotubes. Biotechnol Prog 23:517–531 17. Kojima A, Huon CK, Kamimura T, Maeda M, Matsumoto K (2005) Protein sensor using carbon nanotube field effect transistor. Jpn J Appl Phys 44:1596–1598 18. Hu L, Hecht DS, Grüner G (2004) Percolation in transparent and conducting carbon nanotube networks. Nano Lett 4: 2513–2517 19. Martel R, Schmidt T, Shea HR, Hertel T, Avouris P (1998) Single- and multi-wall carbon nanotube field-effect transistors. Appl Phys Lett 73:2447–2449

Chapter 19 Detection of Biomarkers with Carbon Nanotube-Based Immunosensors Samuel Sánchez, Esteve Fàbregas, and Martin Pumera Abstract A facile and capable method of preparation of sensitive carbon nanotube (CNT)/polysulfone/RIgG immunosensor is discussed in this chapter. The immunosensor is based on the modification of disposable screen-printed electrodes by phase inversion method. CNT/polysulfone membrane acts as the reservoir of immunomolecules as well as a transducer. This configuration offers large surface area, elevated porosity, and mechanical flexibility. The comparison with graphite/polysulfone/RIgG immunosensors shows a significantly improved sensitivity for those prepared with CNTs coupled with polysulfone (PSf). The immunosensing scheme is based on the competitive assay between free and labeled anti-RIgG for the available binding sites of immobilized rabbit IgG (RIgG). The RIgG is incorporated into the PSf immunosensor using a phase inversion method. Horseradish peroxidase enzyme is used as label and hydroquinone as electrochemical mediator. The limit of detection for competitive assay is 1.66 µg/mL. The sensitivity is six times higher for MWCNT-based than for graphite-based electrodes. Key words: Biosensors, Carbon nanotubes, Immunosensors, Biomarkers, Enzyme labeling

1. Introduction The extensive use of carbon nanotubes (CNTs) in immunosensing is due to their superior properties as immobilization platform and as transducer. Electrochemical immunosensors boast the specificity of immunochemical systems plus the advantages of electrochemical transducers (i.e., they are robust, sensitive, and cost-effective). The most crucial step in immunosensor design is the immobilization of the immunoreagent onto or into the electrode surface. Indeed, the quality of immobilization determines the sensitivity and reproducibility of the sensor. A few papers on electrochemical immunosensors that use CNTs have been pubK. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_19, © Humana Press, a part of Springer Science + Business Media, LLC 2010

227

228

Sánchez, Fàbregas, and Pumera

lished in recent years. Although the schemes reported in these publications enable surface immobilization or adsorption of antibodies as well as immunodetection based on electrochemical methods, none of them offer a cheap method for mass production of sensors for biomedical or clinical diagnosis. For an example, CNT has been used as both an electrode and as immobilization surface in an electrochemiluminescencebased biosensor (1). Such immunoassay was carried out in sandwich design by exposing CNT-poly(ethylene vinyl acetate) sheets to a sample containing a-fetoprotein (AFP) and anti-AFP antibodies conjugated with colloidal gold or Ru(bpy)32+. For potential mass production of immunosensing devices, the use of a mass production method for fabrication of electrodes is preferable. One of the most used methods is screen printing. The carbonnanotube thick-film or composite screen-printed immunosensor was constructed using polysulfone (PSf) as binder (2). This matrix retained the RIgG antibody at the surface of screenprinted electrode. The combination of CNT, PSf, and antibodies resulted into a novel composite material, consisting of an interconnected CNT-polymer network, and possessing mechanical flexibility, high toughness, and high porosity. The amperometric measurements demonstrated a six times higher sensitivity for CNT immunocomposite when compared with graphite immunocomposite. CNT/PSf biocomposite retains electrochemical behavior of CNT electrodes, the biocompatibility of PSf binder, and acts as integration matrix for all elements needed for the production of a complex biocomposite. Amperometric immunosensor based on the adsorption of antibodies onto perpendicularly oriented assemblies of single wall CNTs was developed (3). The forests were self-assembled from shortened SWCNTs by nitric acid onto Nafion/iron oxide coated pyrolytic graphite electrodes. Anti-biotin antibody was strongly adsorbed to the SWCNT forests. Improved fabrication of SWCNT forests utilizing aged nanotube dispersions provided higher nanotube density and conductivity (4). Authors concluded that the difference between mediated and unmediated assays is due to the fact that the average distance between horseradish peroxidase (HRP) labels and nanotube ends is too large for efficient direct electron exchange, which can be overcome by electron mediation. Electrochemical immunosensor for cholera toxin (CT) was developed based on poly(3,4-ethylenedioxythiophene)-coated CNTs (5). The sensing interface consists of monoclonal antibody against the B subunit of CT that is linked to poly(3,4-ethylenedioxythiophene) coated on Nafion-supported multiwalled CNT casted film on a glassy carbon electrode. The CT was detected by a “sandwich-type” assay on the electronic transducers, where the toxin is first bound to the anti-CT antibody and

Detection of Biomarkers with Carbon Nanotube-Based Immunosensors

229

then to the ganglioside-functionalized liposome loaded with potassium ferrocyanide for amplification. Here we describe the method for preparation of electrochemical biosensors based on PSf membrane encapsulating multiwall CNTs and immunoreagents layered on disposable screenprinted electrodes. This membrane is printed by serigraphy onto an electrode built on a polycarbonate sheet. Rabbit IgG is used as model antibody being easily labeled with enzymes. HRP is a very often used enzyme for immunological analysis as label, being a simple and cheap reagent available commercially. Direct and competitive immunoassays are carried out and the electrochemical response of HRP is followed by the addition of hydrogen peroxide to the solution.

2. Materials 2.1. Immunoassay

1. IgG from rabbit (RIgG) (Sigma–Aldrich). 2. Anti-Rabbit IgG peroxidase (HRP) conjugated (GaRIgGHRP) (Sigma–Aldrich). 3. PBS buffer: 0.1 M KCl, 0.1 M phosphate sodium, pH 7.0. 4. Blocking buffer: TRIS pH = 7.5, 2% w/v BSA, 0.1% w/v Tween 20, 5 mM EDTA. 5. Washing buffer: 10 mM potassium phosphate pH 6.5, 0.5 M NaCl, 0.05% w/v Tween 20, 0.1% w/v BSA, 1 mM EDTA. 6. TRIS buffer: 0.1 M Tris–HCl, 0.15 M NaCl solution, pH 7.5. 7. H2O2 solution should be prepared just before the experiments by diluting 30% stock solution of H2O2. 8. Hydroquinone solutions (1.8 mM) should be prepared just before the experiments and deoxygenized purging nitrogen into the solution in order to achieve reproducibility.

2.2. CNT-Based Screen-Printed Electrodes

1. PSf (BASF Ultrasons S 3010 natur, Frankfurt, Germany). 2. Multiwalled CNTs (MWCNT) (Sigma–Aldrich). 3. MWCNTs are purified by stirring the MWCNTs in 2 M nitric acid for 24 h and drying at 80°C in a furnace. 4. Graphite powder (particle size 2–10 mm) (BDH, UK). 5. Acheson carbon ink Electrodag 400B (Acheson colloids Co., The Netherlands). 6. Conductive silver ink Electrodag 6037 SS (Acheson colloids Co., The Netherlands).

230

Sánchez, Fàbregas, and Pumera

7. Insulating ink Minico M 7000 (Acheson colloids Co., The Netherlands). 8. Polycarbonate substrates (Servei Estació, Barcelona, Spain). 9. Dek248 semi-automatic system (Asflex S.A. Int., Spain). 10. Scanning electron microscope (SEM) equipped with energy dispersive X-ray spectrometer (EDX) (Hitachi S-4800, Tokyo, Japan). 11. Atomic force microscope (Molecular Imaging, Phoenix, USA). 2.3. Electrochemical Detection

1. Bioanalytical system (BAS) LC-4C amperometric controller connected to a BAS X-Y recorder. 2. Potentiostat: AUTOLAB PGSTAT30 Electrochemical Analyzer (Eco Chemie BV, The Netherlands). 3. Ag/AgCl reference electrode (900200, Orion Thermo, West Palm Beach, FL, USA) filled with 0.1 M KCl as external reference solution. 4. Eppendorf thermomixer model 5436 (Kisker-biotech, Barcelona, Spain).

3. Methods The experiments indicate that MWCNT can be used, better than graphite, to prepare attractive soft immunocomposites for amperometric immunosensing. MWCNT was selected as the best material in connection with PSf for this biocomposite membrane due to its higher sensibility, easy to dissolve homogenously into the PSf/DMF solution, suitable surface roughness and mechanical and physical properties. The incorporation of immunoreagents into the composite aids to reduce the roughness of studied materials. RIgG was incorporated into PSf membrane combined with MWCNT and incubated with anti-RIgG-HRP following the reaction with H2O2 using hydroquinone as mediator. Immunoreagents are easily immobilized within immunosensor by inversion phase technique with negligible waste. In direct immunoassay, the response is six times higher for MWCNT than for graphite. The linear range was determined from 2 to 5 µg/ mL in competitive assay and the detection limit was determined to be 1.66 mg/mL. Such new immunosensor combines the conducting properties of MWCNT with the biocompatibility and flexibility of PSf polymer and it provides the advantages of mass production of immunosensor via screen printing technology.

Detection of Biomarkers with Carbon Nanotube-Based Immunosensors

3.1. CNT-Based Screen-Printed Electrodes

231

1. The carbon/PSf/RIgG screen-printed amperometric immunosensor consists of a single working screen-printed electrode deposited onto polycarbonate substrate. The scheme of fabrication of the screen printed immunosensor is shown in Fig. 1. 2. Serigraphy is employed to screen print the carbon/PS/RIgG immunocomposite onto the strips of 12 working electrodes. 3. The working electrodes are constructed by screen printing, using soft polymer-type squeegees. The applied pressure for the printing process is set to 7 kg/cm2. It is important to control the pressure to print the ink properly without breaking the mask (screen). If the pressure is too low, the ink does not cover the whole surface, and the electrode are not reproducible. 4. A double-sweep process is programmed at a speed of 20 mm/s. 5. Silver conducting ink, carbon, and insulating ink are printed and cured consecutively in the furnace at 60°C overnight. Figure 1 shows the structure of carbon/PSf/RIgG screenprinted electrode. 6. The PSf composite membranes are prepared by mixing 150 µl of 7.5wt%. PSf–DMF solution with graphite or MWCNT. Either MWCNT or graphite suspension are mixed with the PSf–DMF suspension for 10 min under continuous stirring (see Note 1). 7. The carbon loading ratio (carbon/PSf) is optimal for graphite of 17.6wt% and for MWCNT of 6.5wt%. If the carbon content is too low, the conductivity of electrode is very low.

Fig. 1. Schematic of screen-printed immunosensor. (a) The MWCNT/PSf screen-printed electrochemical detector, top view; (b) schematic drawing of showing structure of MWCNT/PSf/RIgG composite; and (c) cross section of the detection area of MWCNT/PSf/RIgG screen-printed immunosensors, including schematic drawing of the immunoreaction after incubation with anti-RIgG-HRP antibody. (a) Polycarbonate substrate, (b) insulator layer, and (c) MWCNT/PSf conducting composite. Reprinted from ref. (2) with permission

232

Sánchez, Fàbregas, and Pumera

If the carbon content is higher, electrodes are too dry and physically unstable. The graphite/PSf membrane should show resistance about 100 mW/sqr, graphite/PSf/RIgG of about 350 mW/sqr, MWCNT/PSf of 130 mW/sqr, and MWCNT/PSf/RIgG of 450 mW/sqr. 8. Immediately after the printing, the PSf solution precipitates (see Note 2). This is achieved by immersing the electrode into a RIgG solution where the nonsolvent (H2O) displaces the solvent (DMF) and makes PSf to precipitate (phase inversion method). The strips are cut and the electrodes are separated. 9. The composite obtained is rinsed with washing buffer for 2 min under stirring in order to eliminate the RIgG, which was not immobilized into the porous membrane. 10. In order to evaluate the unspecific adsorption onto the PSf membrane, control sensors should be prepared as described above but omitting the addition of RIgG in the inversion phase step. 11. It is important to characterize the prepared MWCNT-based electrode. Figure 2 compares surfaces of MWCNT/PSf and graphite/PSf electrodes studied using atomic force microscopy. It is important to note that homogeneous and not extremely high roughness is desirable. The homogeneous dispersion of either MWCNT or graphite is achieved by stirring the suspension for 10 min. 12. Another important characterization method for studying surface is scanning electron microscopy (SEM). Figure 3 demonstrates characteristic surfaces of carbon and MWCNT electrochemical immunosensors before and after incubation with antibody.

Fig. 2. Atomic force microscopy of composites in 3D images MWCNT/PSf (left ) and graphite/PSf (right ). Insets show the 2D images of corresponding composite. Reprinted from ref. (2) with permission

Detection of Biomarkers with Carbon Nanotube-Based Immunosensors

233

Fig. 3. SEM images of carbon/PSf screen-printed immunosensors. (a) MWCNT/PSf/RIgG immunosensor; (b) MWCNT/PSf/RIgG immunosensor incubated with anti-RIgG-HRP; (c) graphite/PSf/RIgG immunosensor; and (d) graphite/PSf/RIgG immunosensor incubated with anti-RIgG-HRP. Reprinted from ref. (2) with permission

3.2. Immunoassay 3.2.1. Direct Immunoassay

1. The primary antibody is incorporated into the PSf membrane as described in Subheading 2.1. 2. The immunoassay is performed as follows: the electrode undergoes 5 min of pre-treatment with the blocking buffer dipping the electrode into a 500 µl of blocking buffer solution and stirring continuously at a 500 rpm at controlled temperature of 25°C. 3. Incubation step is carried out by immersing the sensor into a solution of labeled anti-RIgG-HRP diluted in a 2% BSA solution (blocking buffer) at 37°C under stirring for 30 min.

234

Sánchez, Fàbregas, and Pumera

Fig. 4. (a) Calibration plot for (a) MWCNT200/PSf/RIgG and (b) graphite/PSf/RIgG screen-printed immunosensors modified with 5 g/mL of RIgG and incubated with different concentrations of anti-RIgG-HRP. Standard deviation is represented in figure for three different electrodes. (b) Calibration plot for competitive assay for MWCNT200/PSf/ RIgG screen-printed immunosensor modified with 5 g/mL of RIgG and incubated with 1.37 g/mL of anti-RIgG-HRP. Standard deviation is represented in figure for three different electrodes. Unspecific adsorption of responses is subtracted. Reprinted from ref. (2) with permission

4. The biosensors are washed and rinsed with washing buffer for 5 min. 5. The incubated biosensor undergoes electrochemical detection process described in the Subheading 3.3 (see Fig. 4a). 3.2.2. Competitive Direct Assay

1. Same procedure as described in Subheading 3.2.1 is followed but in competitive direct assay the incubation solution contains different concentrations of anti-RIgG solution varying

Detection of Biomarkers with Carbon Nanotube-Based Immunosensors

235

from 0 to 15 µg/mL mixed with 2.4 µg/mL of the solution of anti-RIgG-HRP conjugate and enough 2% (w/v) BSA. 2. Immunosensors are incubated at 37°C for 30 min with continuous stirring in the above-mentioned solution. 3. After that, the immunosensor is rinsed with the washing buffer for 5 min (see Fig. 4b). 3.3. Electrochemical Detection

1. Measurements are carried out in PBS buffer supporting electrolyte under stirring conditions. A potential of −200 mV vs. S.C.E. electrode is applied to the working electrode. Electrodes are placed close together to minimize IR drop. A stirring bar (7 × 2 mm) and a magnetic stirrer provide the convective transport during amperometric measurements (400 rpm) (see Notes 3 and 4). 2. Hydrogen peroxide is added to the bulk solution for the characterization of RIgG-biosensors using hydroquinone as mediator. 3. Compare MWCNT and graphite carbon materials. 4. All measurements are performed at room temperature (25°C). 5. The attractive behavior of new MWCNT/PSf/RIgG immunocomposite screen-printed immunosensor is demonstrated in connection with HRP-labeled antibody. Figure 5 compares calibration plots for hydrogen peroxide obtained at carbon/PSf/RIgG immunosensors with MWCNT/PSf/ RIgG immunosensor. The signal response is the highest for

Fig. 5. Calibration curves for 4.6 mM hydrogen peroxide using (a) MWCNT/PSf/RIgG immunosensor, (b) graphite/PSf/RIgG, modified with 2 µg/mL of RIgG and incubate with 5 µg/mL of anti-RIgG-HRP. Other conditions as in Fig. 2. Incubation time: 30 min. Reprinted from ref. (2) with permission

236

Sánchez, Fàbregas, and Pumera

MWCNT/PSf/RIgG composite (Fig. 5a). These results show that the MWCNT maintain their high conducting properties although they are immersed in a PSf matrix modified with RIgG antibody (see Note 5). 6. The unspecific adsorption of anti-RIgG-HRP is lower for MWCNT/PS composites when compared to graphite/PSf composite. Since there is no specific antibody (RIgG) on the active surface of composite electrode, the signal obtained in these experiments is due to the direct adsorption of antiRIgG-HRP on the carbon/PSf composite itself. 7. Different concentrations of RIgG antibody are incorporated into the membrane (inversion phase solution of 1, 2, 3, 4, and 5 mg/mL). For this experiment, use different antiRIgG-HRP concentrations (0.5–2 µg/mL). 8. For these experiments, a concentration of 5 mg/mL of RIgG is selected due to the highest signal obtained versus unspecific adsorption. Above-described experiments demonstrates that MWCNT/PS is more sensitive membrane for the preparation of screen-printed immunosensors in the means of the most favorable signal and low unspecific adsorption (see Notes 6 and 7). 9. Figure 6 shows the dynamic response of the MWCNT/PS/ RIgG and graphite/PS/RIgG sensors at working potential of −200 mV for successive injections of H2O2. Figure 6 clearly demonstrates the fast and high sensitivity of the MWCNT sensor to H2O2 in contrast to the low sensitivity of the graphite sensor. The time required to reach 95% of the maximum steady-state current is less than 60 s.

Fig. 6. Current versus time recordings for the successive additions of 1 mM hydrogen peroxide at (a) polysulfone (PSf)-CNT-RIgG and (b) PSf-graphite-RIgG immunosensor. Applied potential, −0.2 V. Reprinted from ref. (2) with permission

Detection of Biomarkers with Carbon Nanotube-Based Immunosensors

237

4. Notes 1. Use ultrasound bath to assure a homogeneous mixture of PSf and CNT in DMF solution. After that, use stirrer when the suspension is not being printed. 2. It is important to coagulate the composite into aqueous solution immediately after the printing. Otherwise, PSf is sensible to the room humidity, and therefore the reproducibility of the composite cannot be controlled. 3. It is important to note that all electrochemical experiments used bi-distilled water. 4. As standard, solutions during acquiring cyclic voltammogram are static, not steered. 5. On the other hand, amperometric sensing at fixed potential (i.e., observing stability of electrode) should be carried out in steered solution with magnetic stirrer (approximately 400 rpm). 6. Immunoreagents can be stored for a few weeks at 4°C in aqueous solution. 7. Control sensors should be done when preparing immunosensors using exactly the same conditions in order to control the unspecific adsorption. References 1. Wohlstadter JN, Wilbur JL, Sigal GB, Biebuyck HA, Billadeau MA, Dong LW, Fischer AB, Gudibande SR, Jamieson SH, Renten JH, Leginus J, Leland JK, Massey RJ, Wohlstadter SJ (2003) Carbon nanotubebased biosensor. Adv Mater 15:1184 2. Sánchez S, Pumera M, Fàbregas E (2007) Carbon nanotube/polysulfone screen-printed electrochemical immunosensor. Biosens Bioelectron 23:332 3. O’Connor M, Kim SN, Killard AJ, Forster RJ, Smyth MR, Papadimitrakopoulos F, Rusling JF

(2004) Mediated amperometric immunosensing using single walled carbon nanotube forests. Analyst 129:1176 4. Yu X, Kim SN, Papadimitrakopoulos F, Rusling JF (2005) Protein immunosensor using singlewall carbon nanotube forests with electrochemical detection of enzyme labels. Mol Biosyst 1:70 5. Viswanathan S, Wu L-C, Huang M-R, Ho J-A (2006) Electrochemical immunosensor for cholera toxin using liposomes and poly(3, 4-ethylenedioxythiophene)-coated carbon nanotubes. Anal Chem 78:1115

Chapter 20 Carbon Nanotube Biosensors with Aptamers as Molecular Recognition Elements Hye-Mi So, Dong-Won Park, Hyunju Chang, and Jeong-O Lee Abstract In this chapter, we discuss in detail the fabrication of carbon nanotube biosensors that use a single-walled carbon nanotube field-effect transistor (SWNT-FET) as a transducer, and aptamers as molecular recognition elements. We use a patterned growth technique to grow SWNTs on Si/SiO2 substrates, and standard microfabrication procedures are then employed to fabricate sensing devices. Key words: Single-walled carbon nanotubes, Field-effect transistor, Aptamer, Biosensor

1. Introduction The detection mechanism of sensors employing single-walled carbon nanotube field-effect transistor (SWNT-FET) transducers can be termed an “electrostatic gating effect” or a “chemically gated field effect transistor” (CHEM-FET) (1). This terminology is appropriate because FETs can detect extra charges from adsorbed molecules, and such molecules act as a molecular (or chemical) gate. The use of FETs for biosensing offers highly desirable miniaturization, and it has been thought that ultimate sensitivity might be achieved using nanoscale FETs. This is especially true when SWNTs are employed, as all atoms of an SWNT are on the surface. Minor disturbances can thus yield large changes in conductance. Since the first demonstration of gas sensing with an SWNT-FET in the year 2000 (2), sensor applications of SWNTs have appeared very promising. Although the sensing mechanism of SWNT-FETs is unclear, either being a field effect or resulting from the Schottky barrier (3), various biosensors based on SWNT-FETs have been successfully constructed to date (4–6). In this chapter, we K. Balasubramanian and M. Burghard (eds.), Carbon Nanotubes: Methods and Protocols, Methods in Molecular Biology, vol. 625, DOI 10.1007/978-1-60761-579-8_20, © Humana Press, a part of Springer Science + Business Media, LLC 2010

239

240

So et al.

concentrate on SWNT-FET biosensors employing aptamers as molecular recognition elements. The most important performance factor for sensors is selectivity; this is the ability to discriminate different substances, avoiding false signals originating from interference. Therefore, choosing recognition elements that afford superior selectivity to sensing transducers is very important. Antibody-antigen pairing is the most widely used recognition element-target pairing because of the high specificity and selectivity of this process. Antibodies are proteins found in vertebrate body fluids; they are generated by the immune system to identify and neutralize foreign substances. Most antibodies consist of two heavy chains and two light chains, and the target antigen binding sites (the variable regions) are at the tips of the Y-shaped antibody. As the height of the antibody is about 10 nm, binding of antigen and antibody can take place no closer than 10 nm from the sensor surface, even when no linkers are used for antibody immobilization. This can be a serious disadvantage in sensors employing surface interaction, including surface plasmon resonance and FET sensors. It is almost impossible to detect the binding of antibody to antigen using FET sensors in an analysis of bodily fluids because such binding will occur at least 10 nm away from the FET surface (assuming that no linkers are used for immobilization of antibodies), a distance that is well outside the Debye screening length (~1 nm for bodily fluids). In the bulk regime, outside of the Debye screening length, all extra charges are screened by surrounding ions (7). Therefore, a smaller and high-performance recognition element is essential for successful implementation of FET-based biosensors. Aptamers are ideal candidates since they are much smaller (~2 nm) than antibodies (8). Aptamers are functional nucleic acids that can bind a large variety of target molecules, such as proteins, small molecules, and microorganisms. Aptamers are single-stranded DNA or RNA oligomers, and, unlike antibodies that are obtained from animal cell culture, aptamers are obtained through an in vitro process termed SELEX (systematic evolution of ligands by exponential enrichment). First identified by Marciniak et al. in 1990 (9), aptamers have been actively studied by many researchers who seek new drug candidates, explore novel separation methods, or require recognition elements for sensors. As recognition elements, aptamers have advantages other than just small size. Being nucleic acids, they are more stable than protein antibodies, and large-scale production is possible once the required sequences are known. Aptamers may be immobilized and functionalized in a straightforward manner and show selectivity superior to that of antibodies. Also, for FET-based sensors, aptamers are ideal recognition elements because of their small size. Moreover, because the aptamer backbone is negatively charged, it is possible that target binding to aptamers could yield larger changes in electrical conductance

Carbon Nanotube Biosensors with Aptamers as Molecular Recognition Elements

241

than seen upon antibody-antigen binding. Recently, Maehashi et al. showed that SWNT biosensors with aptamers as recognition elements offered superior performance when compared to sensors with antibodies as recognition elements (10). The chapter is divided into three parts: (1) the fabrication of SWNT-FETs, (2) the immobilization of aptamers, and (3) the response and regeneration of aptamer-SWNT-FET sensors. In the first part, experimental procedures for SWNT-FET fabrication will be discussed. There are various ways to form SWNT-FETs, but we deal only with SWNT-FETs produced using patterned growth techniques. In the second section, the immobilization of aptamers on a nanotube surface will be described in detail. To retain the excellent electronic properties of SWNTs, we use non-covalent binding linkers for immobilization. Two different linkers, CDITween 20 and 1-pyrenebutyric acid N-hydroxysuccinimide ester, were employed. To use these linkers, the 3¢ or 5¢ end of an aptamer must be functionalized with an NH2 group. The synthesis, characterization, and evaluation of linkers will be discussed in this section. Finally, we will describe how sensor response can be measured. Sensitivity assessment, as well as a regeneration procedure for aptamer-SWNT-FET sensors, will be described.

2. Materials 2.1. Solutions and Reagents

1. Si wafers with 500 nm SiO2 (Nova Electronics Materials Ltd. Carrollton, TX). 2. Poly(methyl methacrylate); PMMA (Microchem, Newton, MA). 3. PMGI resist (Microchem). 4. GXR 601 resist (AZ electronic materials Korea, Seoul, Korea). 5. SU8-2002 resist (Microchem). 6. SU-8 developer (Microchem). 7. AZ 300 MIF developer (AZ electronic materials). 8. Iron (III) nitrate nonahydrate (Aldrich). 9. Aluminum oxide nanopowder (Aldrich or Degussa). 10. Bis(acetylacetonato)dioxomolybdenum (Aldrich) . 11. Evaporation sources Ti, Au, and SiOx all of 99.999% purity (Kurt Lesker, Clairton, PA). 12. High purity Ar (99.999%), CH4 (99.9%), and H2 (99.9%). 13. Methyl(isobutyl) ketone; MIBK (Aldrich). 14. 1,1¢-carbonyldiimidazole, reagent grade (Aldrich).

242

So et al.

15. TWEEN® 20, viscous liquid (Aldrich). 16. Tetrahydrofuran, anhydrous (Aldrich). 17. Dibutyltin dilaurate (Aldrich). 18. n-Hexane, anhydrous (Aldrich). 19. 1-pyrenebutyric acid N-hydroxysuccinimide ester (Aldrich). 20. N,N-dimethylformamide (Sigma, St. Louis, MO). 21. Ethanolamine (Aldrich). 22. Thrombin solution (Sigma). 23. Elastase (Sigma). 24. 6 M Guanidine hydrochloride solution (Qiagen). 25. Thrombin aptamer; 5¢-GGTTGGTGTGGTTGG-3¢; the 3¢-end of the aptamer is modified with an NH2 group (Bioneer Inc., Daejeon, Korea). 2.2. Equipment

1. Furnace for chemical vapor deposition (CVD)-growth (Lindberg Minimite). 2. Spin coater (Midas Systems, Daejeon, Korea). 3. Mask aligner (deep-UV and near-UV) (Midas). 4. Thermal evaporation system (Edward Vacuum). 5. Drying oven. 6. Ultrasonic bath (Branson, Danbury, CT). 7. 300 MHz NMR. 8. Condensing unit. 9. Probe station (MS Tech., South Korea). 10. DAQ 6025 data acquisition board with BNC 6025 (National Instruments, Austin, TX). 11. “No-leak” Ag/AgCl reference electrode (Cypress Systems, Chelmsford, MA). 12. LabView software version 7.0 (National Instruments).

3. Methods 3.1. Device Fabrication

1. To make universal markers, coat PMGI resist onto a 4-in. Si/ SiO2 wafer at ~5.5 g for 10 s and ~180 g for 30 s. Bake the PMGI-coated wafer on a 145°C hot plate for 5 min. Next, spin on GXR 601 resist at ~5.5 g for 10 s and ~180 g for 30 s. Then, bake the wafer on a 90°C hot plate for 3 min. Patterns are made by exposing the sample to a near-UV mask aligner (power ~22 mW/cm2) for 2.5 s and development in MIF developer for approximately 1 min. Use a bilayer of photore-

Carbon Nanotube Biosensors with Aptamers as Molecular Recognition Elements

243

sist to define undercuts that ensure easy lift-off and flat electrode edges. 2. For universal markers, evaporate 5 nm Cr as an adhesion layer (see Note 1) and then deposit ~20 nm Au. Lift off in 80°C Remover PG. Prepare two tanks of Remover PG, and conduct the lift-off process for 30 min in each tank (to give a total of 1 h in 80°C Remover PG). 3. For CVD growth of nanotubes, catalyst patterns should be made with respect to pre-fabricated markers. We used a solutioncatalyst in an organic solvent (methanol), so photoresist cannot be employed. To make catalyst patterns, spin PMMA resist over the 4-in. wafer containing pre-fabricated markers at ~180 g for 30 s. Patterns for catalyst are made employing a deep-UV mask aligner (220 nm, 10 mW/cm2) by exposing the samples to quartz-mask for 1 h (see Note 2). Patterns can be developed in methyl (isobutyl) ketone (MIBK) for 7–10 min. 4. Prepare the catalyst as follows. Mix 20 mg of Fe(NO3)3⋅9H2O, 15 mg of Alumina nanopowder, and 5 mg of MoO2 (acac)2 in 15 mL methanol, and stir the mixture with a magnetic stirrer for 24 h. Sonicate the mixed catalyst solution for 1 h just before use. 5. Spin-coat or pour the catalyst onto the patterned substrate. We usually pour the catalyst solution onto the substrate, wait for 30 s, and then blow dry with a dry N2 gun. Bake the sample at 150°C in a dry oven for 3–5 min to remove any remaining solvent. Lift off the sample in boiling acetone. In order to avoid nanotube formation in regions where there are no catalyst patterns, wash the sample several times with clean acetone. 6. Transfer the sample to the growth furnace. Heat the furnace to 900°C in an Ar atmosphere, and then introduce 1,000 sccm CH4 and 400 sccm H2 into the chamber. Allow the growth process to take place for ~10 min, then shut off the CH4 and H2 streams, and cool to room temperature in an Ar atmosphere (see Note 3). 7. To pattern electrodes, bilayer resists of PMGI with GXR 601 are used. The process is the same as described for marker fabrication. As electrode material, a 3 nm Ti adhesion layer and a 15 nm Au layer are deposited using the thermal evaporation technique. Again, lift-off is performed in 80°C Remover PG for 30 min, and substrates are then transferred to clean 80°C Remover PG for further lift-off. 8. For biosensing applications, electrode insulation is necessary. We use either SU8-2002 negative photoresist or we evaporate SiO2 onto the electrodes. For SU8-2002 insulation, SU8-2002 resist diluted 1:1 with thinner (SU8-2000.5 resist can also be used) is spun over the prefabricated SWNT-FET

244

So et al.

at ~180 g for 30 s. We obtain an SU8 layer ~800 nm thick under these conditions. Pre-baking is performed on a 95°C hot plate followed by ~3 s exposure to a near-UV mask aligner (~22 mW/cm2). Bake the sample on a 95°C hot plate for 1–2 min, and develop with SU-8 developer (~1 min). The SU8 insulation layer can be further cross-linked for better performance. We usually bake the SU8-patterned samples on a hot plate with increasing temperature up to 200°C (1°C/min), and then cool slowly. Samples treated this way are very robust, showing excellent insulation characteristics and stability even in organic solvents such as acetone or dimethylformamide (DMF). Alternatively, thermally evaporated SiO2 can be used for insulation. With this technique, the pattern for the insulation layer should be defined on a PMGI/GXR 601 bilayer resist, and, after evaporation of SiO2, lift-off is achieved in hot remover solution (see Note 4). 9. Characterization of samples follows. We usually measure the I − Vg characteristics and also acquire AFM images. From the I − Vg measurements, samples that show complete depletion of conductance at positive gate voltages, single semiconducting behavior, are chosen for biosensing experiments. 10. Figure 1 is a schematic diagram of the fabrication process and includes a photograph of the SWNT-FET arrays. 11. Figure 2 shows AFM images of SU8-insulated and SiOxinsulated SWNT-FETs. Note the roughness of the surfaces.

Fig. 1. Schematic diagram showing the fabrication process of SWNT-FETs. (a) Universal marker fabrication; (b) Pattern catalyst fabrication using deep-UV; (c) CVD growth of the SWNT; (d) Electrode fabrication using photolithography and thermal evaporation; (e) Electrode insulation; and (f) Photograph of an SWNT-FET array (108 transistors on a 2 × 2 cm2 chip)

Carbon Nanotube Biosensors with Aptamers as Molecular Recognition Elements

245

Fig. 2. Atomic force microscopy (AFM) images of SWNT-FETs. (a) A sample with SU82002 insulation; (b) A sample with SiOx insulation. Scale bar is 1 mm

3.2. Immobilization 3.2.1. Synthesis of Linking Molecule, CDI-Tween 20

1. Dissolve 1,1¢-carbonyldiimidazole (CDI; 486 mg, 3 mmol) in tetrahydrofuran (THF, 10 mL). 2. Dissolve Tween 20 (1.03 mL, 1 mmol) in THF (3 mL). 3. Mix the CDI and Tween 20 solutions in a round-bottomed flask. 4. Add five drops of dibutyltin dilaurate (DBTDL) to the reaction mixture. 5. Boil the reaction mixture at 70°C and stir for 8 h under reflux. 6. Cool the reaction mixture to room temperature and wash five times with THF:n-hexane (1:9; v/v). 7. After removal of solvents, obtain the desired product as a clear yellow viscous liquid. 8. The quality of CDI-Tween 20 should be confirmed by NMR. Figure 3 shows NMR spectra measured using 300 MHz NMR. Compare the 1H absorption peak from position a, b, and c.

3.2.2. Immobilization of Thrombin Aptamer Using CDI-Tween 20

1. Immerse the sample in 3 mM CDI-Tween 20 in DI water for 1 h with stirring at room temperature. 2. Rinse with DI water and blow dry using a dry N2 gun.

246

So et al.

Fig. 3. NMR spectra measured from synthesized CDI-Tween 20. CDCl3 was used as a solvent

3. Drop 10 mL of aptamer solution onto the sample and hold for 12 h at 4°C. 3.2.3. Immobilization of Thrombin Aptamer Using 1-Pyrenebutyric Acid N-Hydroxysuccinimide Ester (Pyrene)

1. Prepare 6 mM pyrene in DMF. 2. Immerse the sample in the solution for 1 h with stirring at room temperature. 3. Rinse with pure DMF and blow dry with a dry N2 gun. 4. Immerse the sample in DI water for 1 h with stirring at room temperature (see Note 5). 5. Rinse with pure DI water and blow dry with a dry N2 gun. 6. Drop 10 mL of aptamer solution onto the sample and hold for 12 h at 4°C. 7. Immerse the device in DI water for 5 min and dry with an N2 gun.

3.2.4. Blocking

1. To block unbound linkers, immerse the sample in 0.1 M aqueous ethanolamine solution (~pH 9, with HCl) for 2 h with stirring. 2. Wash with pure DI water.

3.3. Sensor Response and Regeneration

1. Bring probes into metal contact on the FET. Figure 4 shows the measurement setup using a probe station and LabView controlled electronics (see Note 6).

Carbon Nanotube Biosensors with Aptamers as Molecular Recognition Elements

247

Fig. 4. Real-time measurement setup in the probe station

2. Drop 5 mL of DI water (buffers can be used also) onto the thrombin aptamer-modified SWNT-FET. 3. Place the Ag/AgCl reference electrode in the center of the droplet, and measure the conductance of the device in real time. Typical source-drain and liquid gate biases applied to the Ag/AgCl reference electrode are 100 mV and −300 mV, respectively. 4. Wait until the electrical signal from the SWNT-FET has stabilized, and then add 3–5 mL of thrombin solution of known concentration to the thrombin-aptamer functionalized SWNT-FETs while monitoring conductance from the device. Figure 5(a) shows real-time measurement of thrombin binding to aptamer-functionalized SWNT-FETs. As shown in the figure, the binding of thrombin appears as a decrease in conductance, presumably because bound thrombin screens negative charges from thrombin aptamers. Figure 5(b) shows the change of electrical conductance as a function of thrombin concentration (see Note 7). 5. Step 4 is the real-time measurement. I – Vg measurement using back-gate can be done after rinsing samples with clean buffer or DI water several times, and blow dry with nitrogen. 6. To test selectivity for the thrombin aptamer, repeat steps 2–5 with elastase, or another protein of similar molecular weight and pI value to thrombin.

248

So et al.

Fig. 5. (a) Real-time measurement of electrical conductance showing the binding of thrombin to a thrombin aptamerfunctionalized SWNT-FET. Insets, AFM images of a SWNT before and after the thrombin binding. Scale bar, 100 nm. (b) Sensitivity of the thrombin aptamer-functionalized SWNT-FET. (Reprinted with permission from ref. (5), copyright American Chemical Society)

7. To measure sensitivity, prepare solutions of different thrombin concentrations, and measure conductance changes with each. With aptamer sensors, bound targets can be removed from the aptamers, so the same sensor can be used repeatedly for sensitivity measurements. 8. Recycling of target-bound SWNT-FET aptamer sensor can be achieved by immersing samples in 6 M guanidine hydrochloride for 20 min. 9. Rinse with DI water and dry with nitrogen gas.

4. Notes 1. Chromium is recommended for marker layers because the metal provides a more robust adhesion layer than does Ti. Markers made with a Ti adhesion layer do not readily survive the piranha cleaning process. 2. Masks for near-UV aligners can be made with soda-lime glass, but for deep-UV applications, masks should be made with quartz as soda-lime glass is not transparent in the deep-UV region. 3. Growth conditions can be further optimized. Also, the minimum pattern size required for tube growth is about 200 nm. The larger the pattern size, the greater the probability that multiple carbon nanotubes will be obtained from a single catalyst pattern.

Carbon Nanotube Biosensors with Aptamers as Molecular Recognition Elements

249

4. Depending on experimental conditions, surfaces of samples with SU8 insulation layers can be very dirty, as shown in the AFM image of Fig. 20.2. If resist residues persist, further functionalization with aptamer is not possible, and care should be taken to ensure a clean nanotube surface. 5. This step is important to guarantee clean substrates. If this process is omitted, we see binding of aptamers or proteins not only onto SWNTs but onto substrates also. 6. It is useful to make a PDMS or plexiglass liquid cell that can hold both an adequate amount of liquid and the reference electrode. 7. It is important to note that the very same device was used for all measurements; this was possible because aptamers were employed as molecular recognition elements. Usually, it is not trivial to separate bound target molecules from antibodies, although low-pH buffer solutions can be used for that purpose. Aptamers lose their three-dimensional configuration under conditions of very high salt concentrations or on heating. Therefore, targets bound to aptamers can be easily removed by simply heating, or by washing in high-salt solutions. We found that the device recovered the original conductance when samples were washed with 6 M guanidine hydrochloride. References 1. Patolsky F, Zheng G, Lieber CM (2006) Nanowire sensors for medicine and life sciences. Nanomedicine 1:51–65 2. Kong J, Franklin NR, Zhou C, Chaplin MG, Peng S, Cho K, Dai H (2000) Nanotube molecular wires as chemical sensors. Science 287:622–625 3. Chen RJ, Choi HC, Bangsaruntip S, Yenilmez E, Tang X, Wang Q, Chang Y-L, Dai H (2004) An investigations of the mechanisms of electronic sensing of protein adsorption on carbon nanotube devices. J Am Chem Soc 126:1563–1568 4. Besteman K, Lee J-O, Wiertz FGM, Heering HA, Dekker C (2003) Enzyme-coated carbon nanotubes as single molecule biosensors. Nano Lett 3:727–730 5. So H-M, Won K, Kim YH, Kim B-K, Ryu BH, Na PS, Kim H, Lee J-O (2005) Carbon nanotube biosensors with aptamers as molecular recognition elements. J Am Chem Soc 127:11906–11907 6. So H-M, Park D-W, Jeon E-K, Kim Y-H, Kim BS, Lee C-K, Choi SY, Kim SC, Chang H, Lee

7.

8. 9.

10.

J-O (2008) Detection and titer estimation of Escherichia coli using aptamer-functionalized single-walled carbon nanotube field effect transistors. Small 4:197–201 Lee J-O, So H-M, Jeon E-K, Chang H, Won K, Kim YH (2008) Aptamers as molecular recognition elements for electrical nanobiosensors. Anal Bioanal Chem 390: 1023–1032 Lee JF, Stovall GM, Ellington AD (2006) Aptamer therapeutics advance. Curr Opin Chem Biol 10:282–289 Marciniak RA, Garcia-Blanco MA, Sharp PA (1990) Identification and characterization of a HeLa nuclear protein that specifically binds to the trans-activation-response (TAR) element of human immunodeficiency virus. Proc Natl Acad Sci USA 87:3 624–3628 Maehashi K, Katsura T, Kerman K, Takamura Y, Matsumoto K (2007) Label free protein biosensor based on aptamer-modified carbon nanotube field-effect transistors. Anal Chem 79:782–787

Index A

C

Absorbance............................................. 6, 7, 14, 45, 47, 48, 69, 71, 72, 106, 141, 189 Actuation.......................................................................... 28 Adipocyte......................................................................... 46 Adsorption................................. 3, 4, 6, 7, 132, 145, 154, 155, 157, 164–165, 199, 205, 219, 228, 232, 234, 236, 237 Amide ............................................... 12–14, 23, 24, 28, 198 Amperometric................................................198, 203, 210, 211, 228, 230, 231, 235, 237 Amphiphilic..................................................................... 28 Analyte.............................199, 205, 206, 210, 215, 218, 222 Antibody................................................ 10, 43, 56, 62, 135, 181, 182, 191, 192, 215, 216, 219, 220, 228, 229, 231–233, 235, 236, 240, 241 Apoptosis............................68, 73, 74, 78, 79, 130–132, 165 Aptamer................................................................. 239–249 Atomic force microscopy (AFM)................. 4, 7, 10, 14, 15, 20, 24–25, 29, 31–35, 220, 230, 232, 245, 248, 249 Attachment............................................. 3–7, 15, 25, 86, 90 Attenuated total reflection (ATR)........................ 13–14, 16

Capillary electrophoresis (CE)............................... 153–167 Carbon nanotubes (CNTs) aggregate......................................... 6, 27, 28, 31, 34, 68, 81, 116, 146, 148, 176 arrays......................................... 172, 173, 176, 198, 244 bundles......................................................27, 28, 31, 34 cell trafficking................................................... 135–149 cellular uptake............................................123–133, 154 composite...................198, 205–211, 228, 231, 232, 237 cytotoxicity...............................................57, 68, 72, 86, 87, 91, 92, 95–106, 123–133, 179 dispersion/dispersing..................................6, 21, 27–36, 78, 126, 148, 199, 228, 232 double-walled (DWCNT)............................... 198–203 drug delivery.................................. 56, 85, 116, 136, 138 field-effect transistor (CNT-FET)............213–224, 239 fluorescently labelled............................60, 154–161, 166 functionalizing/functionalization......................... 14, 15, 20, 21, 25, 68, 95, 136, 146, 154–155, 215, 216, 219–221, 224, 249 genotoxicity...............................................110, 117, 136 internalization..................................................50, 51, 56 library................................................................. 95–106 metallic................................................30, 207, 218, 221 metal residues/metal impurities.......................... 74, 115 microelectrode array (MEA)............................ 173, 176 multi-walled (MWNT)..............................9–16, 19–26, 47, 60, 68, 69, 72, 73, 75, 82, 96–106, 154, 156, 174, 229–233, 235, 236 network......................................................213–224, 228 oxidized.................. 10, 12, 21, 23, 44, 68, 101, 154, 156 purity/purified/purification......................68, 73, 74, 206 semiconducting...............30, 31, 138, 214, 221, 224, 244 sidewall................................................ 15, 25, 28, 29, 31 single-walled (SWNT)..................... 3–7, 15, 19, 27–36, 42, 45, 47, 50, 51, 72, 86–88, 90–93, 136–140, 142–145, 148, 149, 154–157, 161–166, 197, 213–224, 228, 239–241, 243–245, 247–249 surface.........................................................7, 10, 27, 28, 51, 68, 69, 78, 95, 96, 99, 101, 102, 123–126, 132, 136, 139, 144, 145, 148, 171, 172, 197–199, 201–203, 205, 207–211, 215, 217, 219, 224, 227, 228, 230–232, 239–241, 244, 249 toxicity............ 41, 55–64, 68, 85, 95, 105, 116, 129, 136

B Bile salt............................................................................... 4 Bioavailability................................................................... 56 Biocompatibility.................... 67–83, 95, 104, 105, 228, 230 Biomarker........................................ 206, 208, 209, 227–237 Biosensor calibration curve/ calibration plot..................... 141, 161, 210, 234, 235 electrochemical..................................205–211, 229, 234 enzymatic......................................................... 197–203 field-effect transistor (FET)..............213–224, 239, 240 immunosensor.................................................. 227–237 reproducibility.......................................................... 210 selectivity...............................................9, 215, 240, 247 sensitivity.......................................... 198, 209, 214, 227, 228, 236, 239, 241, 248 Blocking................. 43, 49, 55, 181, 191, 215, 229, 233, 246 Blood-brain barrier (BBB)......................................... 55, 56 Bone......................................................................42, 44, 50 Bovine serum albumin (BSA).............................43, 97, 102, 181, 189, 229, 233, 235 Brain.................................................... 55–64, 68, 77, 81, 82 Bundling........................................................................... 32

251

Carbon Nanotubes 252 Index

Carboxylation...................10–12, 20, 21, 86, 87, 91, 92, 198 Cell adhesion.....................................................171, 179–193 anchoring.................................................................. 172 association.........................................124, 125, 130–132 autofluorescence........................................................ 136 culture.............................................................42, 58, 69, 71–73, 87–90, 92, 97, 103, 111, 112, 126–127, 139, 142, 143, 155, 180, 182, 183, 185, 188, 193, 240 death.................................. 69, 73–74, 80, 110, 130, 132 incubation...............................................72, 74–77, 103, 104, 126, 127, 129, 139, 140, 142–144, 160, 183 lysis................................ 43, 49, 155, 181, 187–189, 193 membrane............................................... 51, 75, 76, 130, 132, 136, 138, 140–143, 145, 146, 155, 182 proliferation........................... 68, 69, 71, 72, 87, 97, 103 spreading...............................................86, 90, 179–193 trafficking......................................................... 135–149 viability............................................... 45, 71–73, 86, 98, 103, 104, 111, 113, 127, 133, 180, 182–183, 192 Cell lines breast cancer MCF7..................................137, 139, 140 colon cancer HT-29.......................................... 137, 142 HepG2..................................................................... 180 lung epithelial A549......................................... 125, 126 Cellular dynamics....................................................85, 86, 88–91 signaling..................................................................... 96 Central nervous system (CNS)................................... 55–57 Centrifugation.............................................. 4–6, 11, 21, 35, 68, 100, 101, 105, 112, 144, 156, 162, 166 Chemical vapor deposition (CVD)........................... 10, 20, 42, 97, 136, 172–176, 206, 215, 217, 218, 221, 224, 242–244 Chemiluminescence................................................. 49, 228 Chemotherapies............................................................... 55 Circular dichroism (CD).......................................33, 35, 36 CNT-polymer.................................................198, 205, 228 Colorimetry.....................................................104, 124, 125 Combinatorial library............................................... 95–106 Comet assay............................................................ 110–117 Confocal.........................................................58, 60, 61, 78, 81, 124, 132, 137, 140, 142–145, 172 Conformation........................................................28, 29, 35 Conjugation.................................................................. 9–16 Covalent.....................3–7, 9–16, 19–27, 136, 138, 154, 197 Cyclic voltammetry................................................ 201, 208 Cytokines................................................................... 56, 64 Cytotoxicity.............................45, 57, 68, 72, 77–78, 86, 87, 91, 92, 95–106, 110, 116, 123–133, 141, 166, 179

D Debye screening length.................................................. 240 Deoxyribonucleic acid (DNA)

breaks.........................................................110, 116, 117 damage....................................................... 68, 109–118 migration................................... 110, 114–115, 117, 118 modulator................................................................... 95 repair................................................................. 110, 117 transfection................................................................. 57 Dialysis............................................................4–7, 144, 167 Differentiation.................................................... 41–52, 103 DiIC18, . ..........................................................137, 140–143 Dispersion.................................................... 5, 6, 21, 27–36, 68, 78, 126, 148, 199, 228, 232

E Electrocatalytic activity........................................... 198, 205 Electrolyte................................. 86, 201–203, 208, 209, 235 Electron mediation......................................................... 228 Electron microscopy scanning electron microscopy (SEM).................. 57–58, 79, 80, 144, 180, 184–186, 211, 220, 230, 232, 233 transmission electron microscopy (TEM)................................................10–12, 50, 51, 57–58, 60, 98, 101, 124, 132, 156, 180, 183–185 Electron transfer..............................................138, 197, 198 Electropherogram............................ 157–159, 161, 163, 164 Electrophoresis........................ 109–118, 153–167, 187–190 Elemental analysis.................................................. 102, 106 Enzyme enzyme activity............................................................. 4 glucose oxidase.............................................. 4, 197–203 horseradish peroxidase (HRP)...............4, 198, 228, 229 Epoxy...........................................................14, 24, 205–211 Expression.......................................... 46, 49, 50, 57, 61, 64, 113, 185, 189

F Flow cytometry..................... 59, 62, 64, 123–133, 155, 166 Fluorescein..................................................................... 155 Fluorescence background............................................................... 115 energy transfer.................................................. 136–140 microscopy.................................................110, 143, 172 Force field....................................................................... 147 Fullerene C70...................................................137, 140, 141 Functionalization..................................................14, 15, 20, 21, 25, 27, 68, 95, 136, 146, 215, 216, 219–221, 224, 249

G G band..............................................................30, 140, 149 Gene delivery....................................................28, 149, 179 Glassy carbon electrode...........................198–200, 210, 228 Glia cells............................................................77, 172, 173 Gliomas............................................................................ 56 Glucose oxidase.................................................. 4, 197–203

Carbon Nanotubes 253 Index

H

N

a-Helix....................................................................... 29, 35 Horse radish peroxidase (HRP)...........................4, 5, 7, 43, 182, 228–231, 233–236 Human immunoglobulin G (HIgG)................................ 215, 216, 220, 222–223 Human lymphocytes...................................................... 115 Hybridization........................................................... 20, 199 Hydrophilicity.................................................................. 28 Hydrophobicity.......................................................... 28, 98

Nanoparticles............................................. 56, 68, 101, 106, 124, 125, 136–137, 141 Near infra-red (NIR).......................................4, 5, 136, 138 Necrosis.................................................................. 130–132 Neuroblastoma.......................................... 68, 69, 72, 74, 75 Neuro-chip..................................................................... 172 Neurons..................................... 68, 70–71, 77–81, 171–173 b-Nicotinamide adenine dinucleotide (NADH), reduced form........................ 198, 205, 206, 208–210 Nitrous oxide (NO).............................96–98, 103–104, 106 Non-covalent.....................................................3–7, 27, 241 Non-specific binding.............................................. 144, 215 Nuclear magnetic resonance (NMR).................101, 102, 105, 106, 242, 245, 246

I Immobilize/immobilization....................................198, 203, 227, 228, 230, 232, 240, 241, 245–246 Immunoblotting......................................180–182, 188–192 Immunosensor........................................................ 227–237 Impedance.................................................................. 86, 87 Inflammation.................................................................. 106 p-Interactions................................................................... 29 Interband transitions........................................................ 30 Intracellular carriers................................................ 153–167 Intracerebral..........................................................78, 81, 82 Intracranial................................................................. 57, 63 Isoelectric point................................................................ 29

L Labeling...............................................................55, 58, 60, 137, 143–144 Laser induced fluorescence (LIF)........................... 153–167 Leukemia...................................................97, 153, 154, 157 Library...................................................................... 95–106 Ligands..................................................................... 56, 240 Light scattering.......................................106, 125, 130, 132 Linker ............................................... 29, 156, 240, 241, 246 Lipid bilayers.......................................................... 145–148 Lysophospholipid............................................136, 138, 140

M Macrophage..........................................................56, 63, 64, 98, 103, 104, 106 Micellar electrokinetic chromatography (MEKC)..................................................... 161–163 Microglia.............................................................. 56, 61–63 Molecular dynamics................................................ 145–147 Molecular models atomistic................................................................... 148 coarse grained........................................................... 146 multiscale.................................................................. 148 Monitoring......................................... 85–93, 101, 102, 136, 141, 147, 247 MTT assay...................................... 45, 71, 72, 86, 125, 192 Multidrug resistance (MDR)................................. 153–167

O Oligonucleotides.............................................................. 20 Osteoblasts................................................................. 41–52 Oxidation.................................... 68, 82, 201–203, 208, 209 Oxidative stress.........................................68, 70, 74–77, 86

P P-glycoprotein.................................................153, 159, 161 pH........................................................ 5, 10, 20–22, 27–36, 43, 44, 48, 57, 60, 70, 98, 102, 110–112, 142, 145, 155, 181, 199, 201–203, 208, 210, 215, 216, 220, 229, 246, 249 Phosphatase.......................................................... 42, 45–46 Photolithography.............................................173, 216, 244 Photoresist.............................................................. 174, 243 Photothermal................................................................... 28 Plasmid............................................................................. 57 Pluronic....................................... 57–59, 68, 72, 78, 79, 126 Polyethylene glycol (PEG)......................................... 16, 25 Poly-L-lysine..................................... 20–23, 28, 29, 32, 173 Polymerase chain reaction (PCR)............................ 22, 180–181, 185, 187–188 Polysulfone............................................................. 228, 236 Pore size.............................................. 7, 10, 11, 20, 21, 102 Primary neurons........................................69–71, 77–79, 81 Proliferation......................41–52, 68, 69, 71, 72, 87, 97, 103 Protein.........................................4–7, 12–15, 25, 42, 43, 46, 48–50, 52, 88, 96–98, 102–106, 124, 131, 136, 146, 188–190, 193, 200, 240, 247

Q Quantum dots................................................................ 125 Quencher/quenching................. 32, 103, 138, 139, 154, 162

R Radial breathing mode (RBM)......................................... 30 Radioisotope labeling............................................. 143–144

Carbon Nanotubes 254 Index

Raman spectra................................................................. 33, 140 spectrometer..................................................... 137, 149 Rapid screening...................................................... 124, 125 Real-time assay........................................................... 85–93 Receptor..............................................................56, 87, 215 Redox activity................................................................. 202 Reverse transcriptase.........................................57, 181, 187 Rhodamine.............................. 136, 138–140, 148, 154, 155 Ribonucleic acid (RNA)............................................. 19–26

S Scaffold............................................................................. 85 Screen-printed electrode..................................219, 228–233 Self-assembly.......................................................... 172, 228 Serigraphy.............................................................. 229, 231 Single cell gel electrophoresis................................. 109–118 Sodium cholate............................................................... 4–7 Specific area.................................................................... 202 Staining..........................................................46–48, 50, 52, 69, 73–76, 78, 81, 112, 114, 124, 126, 130–133, 180, 184 Stern-Volmer equation........................................... 103, 106 SU-8....................................................................... 241, 244 Surfactant.................................................. 4, 28, 68, 78, 154

Tissue engineering................................. 19, 41, 85, 135, 179 Topography.................................................................... 171 Toxicity...........................................................41, 55–64, 68, 85, 95, 105, 116, 126, 129, 136 Transconductance........................................................... 214 Transmission electron microscopy (TEM).......................................... 10, 57–58, 60, 96, 124, 132, 156, 180, 183–184 Trypan blue exclusion assay............................................................ 124, 193 Tumor........................................................55–57, 62–64, 85

U Ultracentrifugation................................................44, 60, 63 Ultrasonication...................................... 16, 25, 92, 193, 200 Uptake.................................................. 55–64, 85, 123–133, 136–142, 154–156, 159, 162, 164

V Vectors........................................................................ 57, 70

W Western blot....................................................43–44, 48–50

T

X

Threshold voltage................................................... 214, 221 Thrombin........................................................242, 245–248

X-ray photoelectron spectroscopy (XPS)..................................... 10–13, 20, 22–24, 199