DNA Nanotechnology: Methods and Protocols (Methods in Molecular Biology Vol 749)

Methods

in

Molecular Biology™

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



For further volumes: http://www.springer.com/series/7651

DNA Nanotechnology Methods and Protocols Edited by

Giampaolo Zuccheri and Bruno Samorì Department of Biochemistry, University of Bologna, Bologna, Italy

Editors Giampaolo Zuccheri, Ph.D. Department of Biochemistry University of Bologna Bologna, Italy [email protected]

Bruno Samorì Department of Biochemistry University of Bologna Bologna, Italy [email protected]

ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-61779-141-3 e-ISBN 978-1-61779-142-0 DOI 10.1007/978-1-61779-142-0 Springer New York Heidelberg London Dordrecht Library of Congress Control Number: 2011929163 © Springer Science+Business Media, LLC 2011 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. Printed on acid-free paper Humana Press is part of Springer Science+Business Media (www.springer.com)

Preface Giorgio Vasari, a painter, architect, and art historian during the Italian Renaissance, is credited with coining the expression “andare a bottega,” (“attending the studio”) referring to the internship that the apprentice would complete in the master’s studio in order to learn what could be uniquely transmitted in person and in that particular environment and that could then lead to making a unique artist of the apprentice. Nowadays, this same concept holds true in science, and despite the many opportunities for communication and “virtual presence”, the real physical permanence in a lab is still the best way for a scientist to learn a technique or a protocol, or a way of thinking. A book of protocols, such as this, humbly proposes itself as the second-best option. Not quite the same as being in person in a lab and witnessing the experts’ execution of a protocol, it still holds many more details and hints than the usually brief methods section found in research papers. This book of protocols for DNA nanotechnology was composed with this concept in mind: prolonging the tradition of Methods in Molecular Biology, it tries to simplify researchers’ lives when they are putting in practice protocols whose results they have learnt in scientific journals. DNA is playing a quite important and dual role in nanotechnology. First, its properties can nowadays be studied with unprecedented detail, thanks to the new instrumental nano(bio)technologies and new insight is being gathered on the biological behavior and function of DNA thanks to new instrumentation, smart experimental design, and protocols. Second, the DNA molecule can be decontextualized and “simply” used as a copolymer with designed interaction rules. The Watson–Crick pairing code can be harnessed towards implementing the most complicated and elegant molecular self-assembly reported to date. After Ned Seeman’s contribution, elegantly complicated branched structures can be braided and joined towards building nano-objects of practically any desired form. DNA nanotechnology is somewhat like watching professional tennis players: everything seems so simple, but then you set foot on the court and realize how difficult it is to hit a nice shot. When you see the structural perfection of a self-assembling DNA nanoobject, such as a DNA origami, you marvel at how smart DNA is as a molecule and wonder how many different constructs you could design and realize. Among the others, this book tries to show the procedures to follow in order to repeat some of the methods that lead to such constructs, or to the mastering of the characterization techniques used to study them. Many details and procedures are the fruit of the blending of artistry, science, and patience, which are often unseen in a journal paper, but that could be what makes the difference between a winning shot and hitting the net. Many research groups share their expertise with the readers in this book. For the sake of conciseness, we here mention the group leaders, while it is truly from the daily work of a complete team that the details of a protocol can be worked out. The chapters of this book can be roughly divided into two parts: some deal with the methods of preparing the nanostructures, from the rationale of the operations to the techniques for their handling; some other chapters deal more directly with advanced instrumental techniques that can manipulate and characterize molecules and nanostructures. As part of the first group, Roberto Corradini introduces the reader to the methods and choices for taming helix chirality, Alexander Kotlyar, Wolfgang Fritzsche, Naoki Sugimoto, and James Vesenka

v

vi

Preface

share their different methods in growing, characterizing, and modifying nanowires based on G tetraplexes; Hao Yan and Friedrich Simmel teach all the basics for implementing the self-assembly of branched DNA nanostructures, and then characterizing the assembly. Hanadi Sleiman tells about hybrid metal–DNA nanostructures with controlled geometry. Frank Bier shows the use of rolling circle amplification to make repetitive DNA nanostructures, while, moving closer to technological use of DNA, Arianna Filoramo instructs on how to metalize double-stranded DNA and Andrew Houlton reports on the protocol to grow DNA oligonucleotides on silicon. Also with an eye to the applicative side, Yamuna Krishnan instructs on how to insert and use DNA nanostructures inside living cells. On the instrument side, Ciro Cecconi and Mark Williams introduce the readers to methods for the use of optical tweezers, focusing mainly on the preparation of the ideal molecular construct and on the instrument and its handling, respectively. John van Noort and Sanford Leuba give us protocols on how to obtain sound data from single-molecule FRET and apply it to study the structure of chromatin. Claudio Rivetti teaches the reader how to extract quantitative data from AFM of DNA and its complexes, while Matteo Castronovo instructs on the subtleties of using the AFM as a nanolithography tool on self-assembled monolayers; Jussi Toppari dwelves on the very interesting use of dielectrophoresis as a method to manipulate and confine DNA, while Matteo Palma and Jennifer Cha explain methods for confining on surfaces DNA and those very same types of DNA nanostructures that other chapters tell the reader how to assemble. Aleksei Aksimientev shows the methods for modeling nanopores for implementing DNA translocation, a technique bound to find many applications in the near future. We hope this book will help ignite interest and spur activity in this young research field, expanding our family of enthusiastic followers and practitioners. There are certainly still many chapters to be written on this subject, simply because so much is happening in the labs at this very moment. There will certainly be room for the mainstreaming of protocols on the use of DNA analogues (starting with the marvelous RNA, of course), for the design and preparation of fully 3D architectures, for the development of routes towards functional DNA nanostructures, which will lead to applications. DNA nanostructures can be “re-inserted” in their original biological context, as microorganisms can be convinced to replicate nanostructures or even code them. And eventually, applications will require massive amounts of the nanostructures to be produced and to be manipulated automatically, possibly with a precision and output rate similar to that of the assembly of microelectronics circuitry nowadays. Our personal wish is that the next chapters will be written by some of our readers. Bologna, Italy Bologna, Italy

Giampaolo Zuccheri Bruno Samorì

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

v ix

  1 Synthesis and Characterization of Self-Assembled DNA Nanostructures . . . . . . . . Chenxiang Lin, Yonggang Ke, Rahul Chhabra, Jaswinder Sharma, Yan Liu, and Hao Yan   2 Protocols for Self-Assembly and Imaging of DNA Nanostructures . . . . . . . . . . . . Thomas L. Sobey and Friedrich C. Simmel   3 Self-Assembly of Metal-DNA Triangles and DNA Nanotubes with Synthetic Junctions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hua Yang, Pik Kwan Lo, Christopher K. McLaughlin, Graham D. Hamblin, Faisal A. Aldaye, and Hanadi F. Sleiman   4 DNA-Templated Pd Conductive Metallic Nanowires . . . . . . . . . . . . . . . . . . . . . . Khoa Nguyen, Stephane Campidelli, and Arianna Filoramo   5 A Method to Map Spatiotemporal pH Changes Inside Living Cells Using a pH-Triggered DNA Nanoswitch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Souvik Modi and Yamuna Krishnan   6 Control of Helical Handedness in DNA and PNA Nanostructures . . . . . . . . . . . . Roberto Corradini, Tullia Tedeschi, Stefano Sforza, Mark M. Green, and Rosangela Marchelli   7 G-Quartet, G-Quadruplex, and G-Wire Regulated by Chemical Stimuli . . . . . . . . Daisuke Miyoshi and Naoki Sugimoto   8 Preparation and Atomic Force Microscopy of Quadruplex DNA . . . . . . . . . . . . . James Vesenka   9 Synthesis of Long DNA-Based Nanowires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alexander Kotlyar 10 G-Wire Synthesis and Modification with Gold Nanoparticle . . . . . . . . . . . . . . . . . Christian Leiterer, Andrea Csaki, and Wolfgang Fritzsche 11 Preparation of DNA Nanostructures with Repetitive Binding Motifs by Rolling Circle Amplification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Edda Reiß, Ralph Hölzel, and Frank F. Bier 12 Controlled Confinement of DNA at the Nanoscale: Nanofabrication and Surface Bio-Functionalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Matteo Palma, Justin J. Abramson, Alon A. Gorodetsky, Colin Nuckolls, Michael P. Sheetz, Shalom J. Wind, and James Hone 13 Templated Assembly of DNA Origami Gold Nanoparticle Arrays on Lithographically Patterned Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Albert M. Hung and Jennifer N. Cha 14 DNA-Modified Single Crystal and Nanoporous Silicon . . . . . . . . . . . . . . . . . . . . Andrew Houlton, Bernard A. Connolly, Andrew R. Pike, and Benjamin R. Horrocks

1

vii

13

33

49

61 79

93 105 115 141

151

169

187 199

viii

Contents

15 The Atomic Force Microscopy as a Lithographic Tool: Nanografting of DNA Nanostructures for Biosensing Applications . . . . . . . . . . . . . . . . . . . . . . Matteo Castronovo and Denis Scaini 16 Trapping and Immobilization of DNA Molecules Between Nanoelectrodes . . . . . Anton Kuzyk, J. Jussi Toppari, and Päivi Törmä 17 DNA Contour Length Measurements as a Tool for the Structural Analysis of DNA and Nucleoprotein Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Claudio Rivetti 18 DNA Molecular Handles for Single-Molecule Protein-Folding Studies by Optical Tweezers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ciro Cecconi, Elizabeth A. Shank, Susan Marqusee, and Carlos Bustamante 19 Optimal Practices for Surface-Tethered Single Molecule Total Internal Reflection Fluorescence Resonance Energy Transfer Analysis . . . . . . . . . . . . . . . . Matt V. Fagerburg and Sanford H. Leuba 20 Engineering Mononucleosomes for Single-Pair FRET Experiments . . . . . . . . . . . Wiepke J.A. Koopmans, Ruth Buning, and John van Noort 21 Measuring DNA–Protein Binding Affinity on a Single Molecule Using Optical Tweezers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Micah J. McCauley and Mark C. Williams 22 Modeling Nanopores for Sequencing DNA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jeffrey R. Comer, David B. Wells, and Aleksei Aksimentiev

209 223

235

255

273 291

305 317

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359

Contributors Justin J. Abramson  •  Department of Mechanical Engineering, Columbia University, New York, NY, USA Aleksei Aksimentiev  •  Department of Physics, University of Illinois at Urbana-Champaign, Urbana, IL, USA Faisal A. Aldaye  •  Department of Systems Biology, Harvard Medical School, Boston, MA, USA Frank F. Bier  •  Department of Nanobiotechnology & Nanomedicine, Fraunhofer Institute for Biomedical Engineering, Branch Potsdam-Golm, Potsdam, Germany Ruth Buning  •  Leiden Institute of Physics, Leiden Universiteit, Leiden, The Netherlands Carlos Bustamante  •  Howard Hughes Medical Institute, Department of Physics, University of California, Berkeley, CA, USA Stephane Campidelli  •  CEA Saclay, Laboratoire d’Electronique Moléculaire, Gif-sur-Yvette Cedex, France Matteo Castronovo  •  Department of Biology, MONALISA Laboratory, College of Science and Technology, Temple University, PA, USA Ciro Cecconi  •  CNR-Istituto Nanoscienze S3, Department of Physics, University of Modena e Reggio Emilia, Modena, Italy Jennifer N. Cha  •  Department of Nanoengineering, UC San Diego, La Jolla, CA, USA Rahul Chhabra  •  University of Alberta, National Institute of Nanotechnology, Edmonton, AB, Canada Jeffrey R. Comer  •  Department of Physics, University of Illinois at Urbana-Champaign, Urbana, IL, USA Bernard A. Connolly  •  Chemical Nanoscience Laboratory, School of Chemistry, Newcastle University, Newcastle upon Tyne, UK Roberto Corradini  •  Dipartimento di Chimica Organica e Industriale, Univeristà di Parma, Parma, Italy Andrea Csaki  •  Institute of Photonic Technology (IPHT), Jena, Germany Matt V. Fagerburg  •  Departments of Cell Biology and Physiology and Bioengineering, University of Pittsburgh School of Medicine and Swanson School of Engineering, Petersen Institute of Nano Science and Engineering and University of Pittsburgh Cancer Institute, Pittsburgh, PA, USA Arianna Filoramo  •  CEA Saclay, Laboratoire d’Electronique Moléculaire, Gif-sur-Yvette Cedex, France Wolfgang Fritzsche  •  Institute of Photonic Technology (IPHT), Jena, Germany Alon A. Gorodetsky  •  Department of Chemistry, Columbia University, New York, NY, USA

ix

x

Contributors

Mark M. Green  •  Dipartimento di Chimica Organica e Industriale, Univeristã di Parma, Parma, Italy Graham D. Hamblin  •  Department of Chemistry, McGill University, Montreal, Canada Ralph Hölzel  •  Department of Nanobiotechnology & Nanomedicine, Fraunhofer Institute for Biomedical Engineering, Branch Potsdam-Golm, Potsdam, Germany James Hone  •  Department of Mechanical Engineering, Columbia University, New York, NY, USA Benjamin R. Horrocks  •  Chemical Nanoscience Laboratory, School of Chemistry, Newcastle University, Newcastle upon Tyne, UK Andrew Houlton  •  Chemical Nanoscience Laboratory, School of Chemistry, Newcastle University, Newcastle upon Tyne, UK Albert M. Hung  •  Department of Nanoengineering, UC San Diego, La Jolla, CA, USA Yonggang Ke  •  Dana-Farber Cancer Institute & Harvard Medical School, Boston, MA, USA Wiepke J.A. Koopmans  •  Leiden Institute of Physics, Leiden Universiteit, The Netherlands Alexander Kotlyar  •  Department of Biochemistry, The George S. Wise Faculty of Life Sciences, Tel Aviv University, Ramat Aviv, Israel Yamuna Krishnan  •  Biochemistry, Biophysics and Bioinformatics, National Centre for Biological Sciences, Bangalore, India Anton Kuzyk  •  Lehrstuhl für Bioelektronik, Physik-Department and ZNN/WSI, Technische Universität München, Garching, Germany Christian Leiterer  •  Institute of Photonic Technology (IPHT), Jena, Germany Sanford H. Leuba  •  Departments of Cell Biology and Physiology and Bioengineering, University of Pittsburgh School of Medicine and Swanson School of Engineering, Petersen Institute of NanoScience and Engineering, University of Pittsburgh Cancer Institute, Pittsburgh, PA, USA Chenxiang Lin  •  Dana-Farber Cancer Institute & Wyss Institute at Harvard University, Boston, MA, USA Yan Liu  •  Department of Chemistry and Biochemistry, The Biodesign Institute, Arizona State University, Tempe, AZ, USA Pik Kwan Lo  •  Department of Chemistry, McGill University, Montreal, Canada Rosangela Marchelli  •  Dipartimento di Chimica Organica e Industriale, Univeristà di Parma, Parma, Italy Susan Marqusee  •  Department of Molecular & Cell Biology, University of California, Berkeley, CA, USA Micah J. McCauley  •  Department of Physics, Northeastern University, Boston, MA, USA Christopher K. McLaughlin  •  Department of Chemistry, McGill University, Montreal, Canada Daisuke Miyoshi  •  Faculty of Frontiers of Innovative Research in Science and Technology (FIRST), and Frontier Institute for Biomolecular Engineering Research (FIBER), Konan University, Kobe, Japan

Contributors

Souvik Modi  •  Biochemistry, Biophysics and Bioinformatics, National Centre for Biological Sciences, Bangalore, India Khoa Nguyen  •  CEA Saclay, Laboratoire d’Electronique Moléculaire, Gif-sur-Yvette Cedex, France Colin Nuckolls  •  Department of Chemistry, Columbia University, New York, NY, USA Matteo Palma  •  Department of Mechanical Engineering & Applied Physics and Applied Mathematics, Columbia University, New York, NY, USA Andrew R. Pike  •  Chemical Nanoscience Laboratory, School of Chemistry, Newcastle University, Newcastle upon Tyne, UK Edda Reiß  •  Department of Nanobiotechnology & Nanomedicine, Fraunhofer Institute for Biomedical Engineering, Branch Potsdam-Golm, Potsdam, Germany Claudio Rivetti  •  Department of Biochemistry and Molecular Biology, University of Parma, Parma, Italy Denis Scaini  •  Sincrotrone Trieste, Basovizza, Trieste, Italy Stefano Sforza  •  Dipartimento di Chimica Organica e Industriale, Univeristà di Parma, Parma, Italy Elizabeth A. Shank  •  Harvard Medical School, Boston, MA, USA Jaswinder Sharma  •  Center for Integrated Nanotechnologies, Los Alamos National Laboratory, Los Alamos, NM, USA Michael P. Sheetz  •  Department of Biological Sciences, Columbia University, New York, NY, USA Friedrich C. Simmel  •  Physik Department, Technische Universität München, Munich, Germany Hanadi F. Sleiman  •  Department of Chemistry, McGill University, Montreal, Canada Thomas L. Sobey  •  Physik Department, Technische Universität München, Munich, Germany Naoki Sugimoto  •  Faculty of Frontiers of Innovative Research in Science and Technology (FIRST), and Frontier Institute for Biomolecular Engineering Research (FIBER), Konan University, Kobe, Japan Tullia Tedeschi  •  Dipartimento di Chimica Organica e Industriale, Università di Parma, Parma, Italy J. Jussi Toppari  •  Department of Physics, Nanoscience Center, University of Jyväskylä, Jyväskylä, Finland Päivi Törmä  •  Department of Applied Physics, School of science, Aalto University, Aalto, Finland John van Noort  •  Leiden Institute of Physics, Leiden Universiteit, Leiden, The Netherlands James Vesenka  •  Department of Chemistry and Physics, University of New England, Biddeford, ME, USA David B. Wells  •  Department of Physics, University of Illinois at Urbana-Champaign, Urbana, IL, USA

xi

xii

Contributors

Mark C. Williams  •  Department of Physics, Northeastern University, Boston, MA, USA Shalom J. Wind  •  Department of Applied Physics and Applied Mathematics, Columbia University, New York, NY, USA Hao Yan  •  Department of Chemistry and Biochemistry, The Biodesign Institute, Arizona State University, Tempe, AZ, USA Hua Yang  •  Department of Chemistry, University of British Columbia, Vancouver, Canada

Chapter 1 Synthesis and Characterization of Self-Assembled DNA Nanostructures Chenxiang Lin, Yonggang Ke, Rahul Chhabra, Jaswinder Sharma, Yan Liu, and Hao Yan Abstract The past decade witnessed the fast evolvement of structural DNA nanotechnology, which uses DNA as blueprint and building material to construct artificial nanostructures. Using branched DNA as the main building block (also known as a “tile”) and cohesive single-stranded DNA (ssDNA) ends to designate the pairing strategy for tile–tile recognition, one can rationally design and assemble complicated nanoarchitectures from specifically designed DNA oligonucleotides. Objects in both two- and three-dimensions with a large variety of geometries and topologies have been built from DNA with excellent yield; this development enables the construction of DNA-based nanodevices and DNA-template directed organization of other molecular species. The construction of such nanoscale objects constitutes the basis of DNA nanotechnology. This chapter describes the protocol for the preparation of ssDNA as starting material, the self-assembly of DNA nanostructures, and some of the most commonly used methods to characterize the self-assembled DNA nanostructures. Key words: DNA nanotechnology, Self-assembly, Electrophoresis, Atomic force microscopy

1. Introduction The notion that DNA is merely the gene encoder of living ­systems has been eclipsed by the successful development of DNA nanotechnology. DNA is an excellent nanoconstruction material because of its inherent merits: First, the rigorous Watson-Crick base-pairing makes the hybridization between DNA strands highly predictable. Second, the structure of the B-form DNA double helix is well-understood; its diameter and helical repeat have been determined to be ~2 and ~3.4 nm (i.e., ~10.5 bases), respectively, which facilitates the modeling of even the most complicated DNA nanostructures. Third, DNA possesses combined Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_1, © Springer Science+Business Media, LLC 2011

1

2

Lin et al.

structural stiffness and flexibility. The rigid DNA double helixes can be linked by relatively flexible single-stranded DNA (ssDNA) to build stable motifs with desired geometry. Fourth, modern organic chemistry and molecular biology have created a rich toolbox to readily synthesize, modify, and replicate DNA molecules. Finally, DNA is a biocompatible material, making it suitable for the construction of multicomponent nanostructures made from hetero-biomaterials. The field of structural DNA nanotechnology began with Nadrian Seeman’s vision of combining branched DNA molecules bearing complementary sticky-ends to construct two-dimensional (2D) arrays (1) and his experimental construction of a DNA object topologically equal to a cube (2). Today, DNA self-assembly has matured with such vigor that it is currently possible to build micro- or even millimeter-sized nanoarrays with desired tile geometry and periodicity as well as any discrete 2D or 3D nanostructures we could imagine (3–8). Modified by functional groups, those DNA nanostructures can serve as scaffolds to control the positioning of other molecular species (9–21), which opens opportunities to study intermolecular synergies, such as protein–protein interactions, as well as to build artificial multicomponent nanomachines (22–24). Generally speaking, the creation of a novel DNA motif usually requires the following steps: (1) Structural modeling: physical and/or graphic models are used to help the design of a new DNA motif; (2) Sequence design: in this step, specific sequences are assigned to all ssDNA molecules in the model; (3) Experimental synthesis of the DNA nanostructure; and (4) Characterization of the DNA nanostructure. The first two steps are crucial to program the outcome of self-assembly and assisted by computer software (25–30). In this chapter, we are going to describe the experimental protocols involved in steps 3 and 4.

2. Material All chemicals are purchased from Sigma-Aldrich (St. Louis, MO) unless otherwise noted. All buffer solutions are filtered and stored at room temperature unless otherwise noted. 2.1. Denaturing Polyacrylamide Gel Electrophoresis for the Purification of Synthetic SingleStranded DNA

1. Synthetic ssDNA (Integrated DNA Techonologies, Coralville, IA) with designated sequences. 2. TBE buffer (1×): 89 mM Tris–boric acid, pH 8.0, 2 mM ethylenediaminetetraacetic acid disodium salt (EDTA-Na2). 3. 20% urea-acrylamide Mix: 20% acrylamide (19:1 acrylamide:bis, Bio-Rad Laboratories, Hercules, CA), 8.3 M urea in 1× TBE buffer.

Synthesis and Characterization of Self-Assembled DNA Nanostructures

3

4. 0% Urea-acrylamide Mix: 8.3 M Urea in 1× TBE buffer. 5. Ammonium persulfate (APS): prepare 10% water solution and store at 4ºC. 6. N,N,N,N¢-tetramethyl-ethylenediamine (TEMED, Bio-Rad). 7. Bromophenol blue (BB) or xylene cyanole FF (XC) (2×): prepare 0.1% w/v solution of the dye in 90% formamide solution containing 10 mM NaOH and 1 mM Na2EDTA. 8. Ethidium bromide: prepare 300 mL 0.1 mg/mL solution in a glass tray for gel staining. 9. Elution buffer (1×): 500  mM ammonium acetate, 10  mM magnesium acetate, 2 mM EDTA-Na2. 10. 1-Butanol and 100% Ethanol. 11. Spin X centrifuge tube filters (Corning, Lowell, MA). 2.2. Self-Assembly of DNA Nanostructures

1. Polyacrylamide Gel Electrophoresis (PAGE) purified ssDNA.

2.3. Non-denaturing PAGE for the Characterization of Self-Assembled DNA Nanostructures

1. Self-assembled DNA nanostructures.

2. TAE-Mg buffer (10×): 0.4  M Tris–acetic acid, pH 8.0, 125 mM magnesium acetate, 20 mM EDTA-Na2. 2. 40% acrylamide (19:1 acrylamide:bis, Bio-Rad Laboratories, Hercules, CA) solution. 3. Non-denaturing loading buffer (10×): 0.2% w/v bromophenol blue and xylene cyanole FF in 1× TAE-Mg buffer containing 50% v/v glycerol. 4. DNA ladder with suitable size (Invitrogen, Carlsbad, CA). 5. TAE-Mg buffer (1×), TEMED, and 10% APS solution (vide supra). 6. Stains-All: prepare 0.01% w/v Stains-All in 45% v/v formamide solution.

2.4. Atomic Force Microscope Imaging of Self-Assembled DNA Arrays

1. Self-assembled DNA nanostructures. 2. TAE-Mg buffer (1×) (vide supra). 3. Mica discs (Ted Pella, Inc) and Atomic Force Microscope (AFM) cantilevers of choice with integrated probes (such as NP-S from Veeco, Inc for imaging in liquids).

3. Methods 3.1. Denaturing PAGE Purification of Synthetic ssDNA

With advanced solid state synthesis chemistry, DNA synthesizer can generate DNA strands with designated sequences up to 200base long. However, a significant yield drop is normally associated with the synthesis of longer DNA strands. For example,

4

Lin et al.

if the yield for the addition of one nucleoside is 99%, the yield for the synthesis of a 100-mer ssDNA is only ~37%. Therefore, it is very important to purify the synthetic DNA strands that are longer than 30 bases to maximize the self-assembly yield in the next step. Effective ways to purify ssDNA less than 200-base long include high performance liquid chromatography (HPLC) and PAGE. Here, we discuss the protocol for denaturing PAGE purification of synthetic DNA strands. 1. Set up the gel assembly following the manufacturer’s instruction (we use a Hofer SE 600 Ruby from GE Healthcare) (see Note 1). 2. Mix proper volume of 20% and 0% Urea-acrylamide stock solution to prepare the acrylamide solution with desired concentration. Each gel needs ~35 mL acrylamide solution. For example, to make an 8% polyacrylamide gel, take 14 mL of 20% Urea-acrylamide stock and mix with 21 mL of 0% Ureaacrylamide stock. Stir thoroughly to mix well. For each gel, add 262 mL of 10% APS solution and 14.7 mL of TEMED. Stir thoroughly to mix well. 3. Quickly cast the gel using 35 mL pipette and insert the comb. Make sure no air bubble is trapped in the gel. Leave the gel at room temperature for at least 30 min to allow it solidifies. 4. Prepare the DNA sample. Add DI water to each dry samples to make 0.5 OD260/mL DNA solution. Take 4 OD of each sample (8 mL) into newly labeled tubes (see Note 2) and the rest of the samples should be stored at −20ºC. Add 2× denaturing dye to each sample (BB, XC, or both) and add water to adjust the final volume to 20 mL. Heat the sample at 90ºC for 5 min to denature the DNA strands (see Note 3). 5. When the gel has polymerized, remove the combs and attach the upper buffer chamber (UBC) to the gel assembly. Add 1 × TBE buffer (running buffer) to the UBC and rinse the wells thoroughly with glass pipette. Drain the UBC and add fresh running buffer to cover all the wells. 6. Load the samples to each well. Load 10 mL/well into the gel wells (generally 2 OD per lane) using the gel loading tips. Be careful not to flush the sample out of the well (see Note 4). 7. Carefully put the UBC and gel assembly into the lower buffer chamber (LBC) with ~3.5 L 1× TBE buffer. Add buffers into both UBC and LBC to the marked MAX lines (see Note 5). 8. Turn on the circulating water and set the temperature to 35ºC. Secure the lid of the gel box and connect the electrodes to a DC power supply. Make sure the polarity is correct. Run gel at constant current ~30–40 mA per gel for around 2–3 h depending on the length of the interested DNA fragments.

5

Synthesis and Characterization of Self-Assembled DNA Nanostructures

Table 1 The tracking dye migration on polyacrylamide denaturing gels (Dyes migrate to the same point as DNA strand of the indicated size in a denaturing polyacrylamide gel) Polyacrylamide concentration

5%

6%

8%

10%

12%

Bromophenol blue (bp)

35

26

19

12

8

Xylene cyanole FF (bp)

130

106

76

55

26

The tracking dye in the loading buffer provides a rough marker of the migration of DNA fragments (Table 1). 9. Turn off the power supply and circulating water. Lift the gel from the gel assembly and carefully put it into a glass try containing ~300 mL ethidium bromide (see Note 6). Stain the gel for 5 min and destain it for 5 min in distilled water. 10. In a dark room, lift the gel gently and put it on the UV transilluminator. Turn on the UV at wavelength of 302  nm, use razor blade to cut the major band out (see Note 7). Turn off the UV lamp, chop the band into small pieces, and collect the small gel blocks into Spin X centrifuge tube filters. Add 500 mL of elution buffer into each filter; shake in cold room (4ºC) overnight before proceed to the next step (see Note 8). 11. Centrifuge the Spin X tube filters (4,600 × g for 6  min) to separate the elution buffer from gel blocks. Add 1  mL of 1-butanol to the collected elution buffer, vortex the tube for 1 min, and centrifuge it at 600 × g for 1 min. After the spin, discard the upper layer of 1-butanol with pipette into waste bottle under venting hood. The 1-butanol washing extracts ethidium bromide and tracking dyes from the DNA sample. 12. Add in 1 mL ethanol to the DNA sample and mix well. Leave the mixture in −20ºC freezer for 30 min. Spin at 16,200 × g for 30 min at 4ºC to precipitate DNA. Pour out the ethanol and wash the DNA pellet with 70% v/v ice cold ethanol if desired. Centrifuge the tube at 16,200 × g for 10  min after ethanol washing and pour out all liquid. 13. Use a vacuum concentrator (we use a Vacufuge from Eppendorf, Westbury, NY) to dry the purified DNA sample for 1 h at 30ºC. Add in 50 mL distilled H2O, vortex for 1 min to dissolve the DNA sample. Measure the absorbance of the DNA solution at 260 nm (OD260) using a UV-Vis spectrometer (we employ a Biophotometer from Eppendorf) and convert the measured OD260 value to molar concentration using

6

Lin et al.

the extinction coefficient (e) of the DNA strand provided by the oligonucleotide vendor. Adjust the concentration of all purified DNA strand solution to 30 mM (or any other value to the experimenter’s convenience) by adding distilled H2O. Store all DNA samples in −20ºC freezer. 3.2. Anneal DNA Strands to SelfAssemble DNA Nanostructures

The formation of hydrogen bonded DNA complex is a selfassembly process. The DNA strands are mixed at stoichiometric molar ratio in a near-neutral buffer containing divalent cations (usually Mg2+), heated to denature and then gradually cooled to allow the ssDNA molecules to find their correct partners and adopt the most energy-favorable conformation. 1. Add stoichiometric amount of DNA strands into one 1.5 mL tube (or any other suitable tube size). Add 10× TAE-Mg buffer and distilled H2O to adjust the final concentration of each DNA strand to be 1 mM or any other desired concentration. Mix well and close the tube tightly. 2. This mixture is then heated on a heat block to 95ºC for 5 min and cooled to the desired temperature by the following protocol: 20 min at 65ºC, 20 min at 50ºC, 20 min at 37ºC, and if desired, 20 min at room temperature. 3. To assemble large DNA constructs, such as 2D arrays, slow annealing is desirable. In this case, the mixture is placed on a floating rack, transferred to a 2 L water bath, which is preheated to about 90°C and placed inside a Styrofoam box, and allowed to cool slowly to the desired temperature over the period of 2 days. This slow annealing process can also be carried out on a thermal cycler (see Note 9).

3.3. Non-denaturing PAGE for the Characterization of Self-Assembled DNA Nanostructures

Non-denaturing PAGE is an effective assay to characterize the self-assembled DNA supermolecules. Well-formed DNA nanostructure should migrate as a distinct band after electrophoresis. Non-denaturing PAGE also provides information regarding the yield of self-assembly. A typical gel image showing the correct formation of four helix DNA tile (31) is shown in Fig. 1. 1. Set up the gel assembly following the manufacturer’s instruction as described in step 1, Subheading 3.1 (we use a Hoefer SE 600 Ruby, GE Healthcare). 2. Prepare non-denaturing acrylamide mixture from 40% acrylamide (acrylamide:bis 19:1) stock, 10× TAE-Mg buffer and distilled H2O. The final mixture should contain 1× TAE-Mg buffer. For example, to make an 8% non-denaturing gel, mix 7 mL of 40% acrylamide stock, 3.5 mL of 10× TAE-Mg buffer, and 24.5 mL H2O. Stir thoroughly to mix well. For each gel, add 262 mL of 10% APS solution and 14.7 mL of TEMED. Stir thoroughly to mix well.

Synthesis and Characterization of Self-Assembled DNA Nanostructures

7

Fig. 1. Nondenaturing gel (8% polyacrylamide) of the 4-helix complex stained with Stains-All. Equimolar mixtures of 1 mM of each strand were annealed, and the electrophoresis was run at room temperature. Lane  M is a 100  bp DNA ladder. Lanes 1–8 contain complexes with partial combination of the component strands. Strands included in the annealing are indicated with a schematic drawing above the lane. Lane 9 corresponds to the full complex with all of the component strands.

3. Quickly cast the gel using 35 mL pipette and insert the comb. Make sure no air bubble is trapped in the gel. Leave the gel at room temperature for at least 2 h to allow it solidify (see Note 10). 4. When the gel has polymerized, remove the combs and attach the UBC to the gel assembly. Add 1× TAE-Mg buffer (running buffer) to the UBC and rinse the wells thoroughly with glass pipette. Drain the UBC and add fresh running buffer to cover all the wells. 5. Add 10× non-denaturing loading buffer to the preannealed DNA samples (finally, the DNA should be in 1× loading buffer). Vortex to mix well. Immediately load the DNA samples to each well using the gel loading tips. (Be careful not to flush the sample out of the well.) Take note about the sequence of the samples loaded. A DNA ladder with proper size should be loaded into a separate lane as a reference. 6. Immerse the gel assembly (together with UBC) to the 1× TAE-Mg buffer in the LBC. Add buffers into both UBC and LBC to the marked MAX lines. It is important not to disturb the sample when adding buffer to UBC. Add buffer gently along the side of the chamber. 7. Turn on the circulating water and set the temperature to 20ºC. Secure the lid of the gel box and connect the electrodes to a DC power supply. Make sure the polarity is correct. Run gel at constant voltage ~200 V for 4–8 h depending on the size of the interested DNA complexes.

8

Lin et al.

8. Turn off the power supply and circulating water. Lift the gel from the gel assembly and carefully put it into a glass tray containing ~300 mL 0.01% Stains-All solution. (Wear gloves all time when working with Stains-All). Low percentage gels can be fragile thus should be treated with extreme care. Stain the gel for 2 h and then rinse the gel well using distilled H2O. 9. Place a slide of transparency on the white lamp of the transluminator, carefully put the gel on it, and turn on white light to destain the gel until the gel appears almost colorless. This takes approximately 10–20 min. Watch the color change to avoid overdestain. 10. Use Kimwipe to wick the water off as much as possible, and then cover the gel with another transparency. Make sure no air bubble is trapped between the gel and the transparency. Also avoid stripes and pattern caused by thin layer of water. Scan the gel on a desktop scanner and save the image. 3.4. AFM Imaging of Self-Assembled DNA Arrays

1. The protocol described here assumes the use of PicoPlus AFM (Agilent). To start the imaging session, turn on the computer, Pico Scan controller, and then the AC controller. Open software “Pico Scan.” 2. Choose tapping mode AFM (AC AFM) and insert proper AFM tip into the tip holder on the top of the scanner. For AAC (acoustic AC) mode, use the gold-coated silicon nitride tip (NP-S tip, Veeco) for imaging in liquid or the proper acoustic AC tip (Veeco) for imaging in air. For the NP-S tips, use the tip on the thinner and shorter cantilever for imaging. 3. Sample preparation: Assemble a piece of freshly cleaved mica as the bottom of the fluid cell on the sample stage. Spot a 2 mL of 1 mM NiCl2 solution on mica and leave it to adsorb on the surface for 2 min. Then, add a 2 mL of the sample to the spot and leave it to adsorb on the surface for another 2 min. Finally, add 400 mL 1× TAE-Mg buffer onto the mica in the fluid cell. The Ni2+ adsorbed on mica surface can help the DNA array stay on the surface during the scanning. Attach the sample stage to the magnetic posts on the AFM. 4. Place the scanner on the sample stage with the tip pointing down. Lock the scanner. Turn on laser switch and plug in the detector. Move the laser spot so that it is on the back of the cantilever tip. Adjust the position of the photodiode inside the detector to maximize the sum of signal. Also make sure that the reflected laser spot is at the center of the photodiode. 5. Tune the tip and choose drive frequency with maximum amplitude. Set the parameters for scanning. Proportional gain and integral gain are 0.5 both or larger (<1.2). Use the servo range of ~4,180  nm. Set 0.85 for the amplitude set point.

Synthesis and Characterization of Self-Assembled DNA Nanostructures

9

Fig. 2. AFM images of DNA arrays self-assembled from eight helix tiles. Image sizes are 4.5 × 4.5 mm2 on the left and 800 × 800 nm2 on the right. Each square on the zoom-in image represents an eight helix tile as shown in schematic at the bottom right corner.

Start approaching the tip to the surface and wait until the servo is active. The approaching process may take a few ­minutes if the tip was far from the surface initially. 6. When servo is active, start scanning. During the scanning, optimize the parameters (gains, amplitude set point, and servo range) to obtain optimal images. Scanning size depends on the size and morphology of the DNA assembly of interest. Zoom in on area of interest to observe the detail of the DNA nanostructures. Save the images on the computer. A typical AFM image of the DNA 2D arrays self-assembled from eight helix tiles (31) is shown in Fig. 2. 7. To end the imaging session, stop scanning and withdraw the tip from the surface. Take out the sample stage and dissemble the mica. Remove the tip from the scanner. Clean the sample stage, mica, and scanner head for future use.

4. Notes 1. It is crucial to clean all parts of the gel assembly, especially the glass plates and combs. The glass plates can be cleaned by rinsing extensively with distilled water followed by ethanol and acetone. Also check the edges of glass plates; they should be free of indentation. 2. Each lane of the gel can hold 2 OD of DNA; to purify larger amount, simply take more sample and lanes.

10

Lin et al.

3. Do not heat the sample longer than 10  min. Immediately place the heated tubes on ice to better denature DNA strands with complex secondary structures. 4. Take note about the sequence of the samples loaded. Leave one lane empty in between different samples, especially when their lengths (bp) are very close. The samples should not be loaded too slow to prevent diffusion, which may lead to band smearing. 5. It is important not to disturb the sample when adding buffer to UBC. Add buffer gently along the side of the chamber. 6. Wear nitrile gloves all time when working with ethidium bromide. Low percentage gels can be fragile thus should be treated with extreme care. 7. The major band may show darker than the surrounding because the UV absorbance of DNA is so high, especially when more than 2 OD of DNA is loaded in each lane. 8. To optimize the elution yield, freeze small blocks of gel at −20ºC for 10 min before adding elution buffer. This is especially helpful when purifying DNA strands longer than 100 bases. 9. The annealed structures should be handled gently (e.g., do not vortex) and stored at 4ºC. 10. Low percentage gels can take longer time to solidify. Always make sure that the gel solidifies completely before removing the comb.

Acknowledgments This work was supported by grants from the National Science Foundation (NSF), the Army Research Office (ARO), and the Technology and Research Initiative Fund from Arizona State University to Y.L. and by grants from NSF, ARO, Air Force Office of Scientific Research, Office of Naval Research, and the National Institute of Health to H.Y. References 1. Seeman, N. C. (1982) Nucleic acid junctions and lattices. J. Theor. Biol. 99, 237–247. 2. Chen J., and Seeman, N. C. (1991) The ­synthesis from DNA of a molecule with the connectivity of a cube. Nature 350, 631–633. 3. Seeman, N. C. (2003) DNA in a material world. Nature 421, 427–431.

4. Deng, Z. X., Lee, S. H., and Mao, C. D. (2005) DNA as nanoscale building blocks. J. Nanosci. Nanotechnol. 5, 1954–1963. 5. Turberfield, A. J. (2003) DNA as an engineering material. Phys. World 16, 43–46. 6. Lin, C., Liu, Y., Rinker, S., and Yan, H. (2006) DNA Tile based self-assembly: building

Synthesis and Characterization of Self-Assembled DNA Nanostructures

7. 8. 9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

c­ omplex nanoarchitectures. ChemphysChem 7, 1641–1647. Feldkamp, U., and Niemeyer, C. M. (2006) Rational fesign of DNA nanoarchitectures. Angew. Chem. Int. Ed. 45, 1856–1876. Aldaye, F. A., Palmer, A. L., and Sleiman, H. F. (2008) Assembling materials with DNA as the guide. Science 321, 1795–1799. Yan, H., Park, S. H., Ginkelstein, G., Reif, J. H., and LaBean, T. H. (2003) DNA templated self-assembly of protein arrays and highly conductive nanowires. Science 301, 1882–1884. Le, J. D., Pinto, Y., Seeman, N. C., MusierForsyth, K., Taton, T. A., and Kiehl, R. A. (2004) DNA-templated self-assembly of metallic nanocomponent arrays on a surface. Nano Lett. 4, 2343–2347. Zhang, J., Liu, Y., Ke, Y., and Yan, H. (2006) Periodic square-like gold nanoparticle arrays template by self-assembled 2D DNA nanogrids on a surface. Nano Lett. 6, 248–251. Sharma, J., Chhabra, R., Liu, Y., Ke, Y., and Yan, H. (2006) DNA-templated self-assembly of two-dimensional and periodical gold nanoparticle arrays. Angew. Chem. Int. Ed. 45, 730–735. Zheng, J., Constantinou, P. E., Micheel, C., Alivisatos, A. P., Kiehl, R. A., and Seeman N. C. (2006) Two-dimensional nanoparticle arrays show the organizational power of robust DNA motifs. Nano Lett. 6, 1502–1504. Sharma, J., Chhabra, R., Cheng, A., Brownell, J., Liu, Y., and Yan, H. (2009) Control of self-assembly of DNA tubules through integration of gold nanoparticles. Science 323, 112–116. Sharma, J., Ke, Y., Lin, C., Chhabra, R., Wang, Q., Nangreave, J., Liu, Y., and Yan, H. (2008) DNA-tile-directed self-assembly of quantum dots into two-dimensional nanopatterns. Angew. Chem. Int. Ed. 47, 5157–5159. Aldaye, F. A., and Sleiman, H. F. (2006) Sequential self-assembly of a DNA hexagon as a template for the organization of gold ­nanoparticles. Angew. Chem. Int. Ed. 45, 2204–2209. Liu, Y., Lin, C., Li, H., and Yan, H. (2005) Aptamer directed self-assembly of proteins on a DNA nanostructure. Angew. Chem. Int. Ed. 44, 4333–4338. Chhabra, R., Sharma, J., Ke, Y., Liu, Y., Rinker, S., Lindsay, S., and Yan, H. (2007) Spatially addressable multiprotein nanoarrays template by aptamer-tagged DNA nanoarchitectures. J.  Am. Chem. Soc. 129, 10304–10305.

11

19. Rinker, S., Ke, Y., Liu, Y., Chhabra, R., and Yan, H. (2008) Self-assembled DNA nanostructures for distance-dependent multivalent ligand-protein binding. Nat. Nanotechnol 3, 418–422. 20. Duckworth, B. P., Chen, Y., Wollack, J. W., Sham, Y., Mueller, J. D., Taton, T. A., and Distefano, M. D. (2007) A universal method for the preparation of covalent protein-DNA conjugates for use in creating protein nanostructures. Angew. Chem. Int. Ed. 46, 8819–8822. 21. Malo, J., Mitchell, J. C., Vénien-Bryan, C., Harris, J. R., Wille, H., Sherratt, D. J., and Turberfield, A. J. (2005) Engineering a 2D protein-DNA crystal. Angew. Chem. Int. Ed. 44, 3057–3061. 22. Liedl, T., Sobey, T. L., and Simmel, F. C. (2007) DNA based nano-devices. Nanotoday 2, 36–41. 23. Seeman N. C. (2005) From genes to machines: DNA nanomechanical devices. Trends. Biochem. Sci. 30, 119–125. 24. Bath, J., and Turberfield, A. J. (2007) DNA nanomachines. Nat. Nanotechnol. 2, 275–284. 25. Zuker, M. (2003) Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res. 31, 3406–3415. 26. Birac, J. J., Sherman, W. B., Kopatsh, J., Constantinou, P. E., and Seeman, N. C. (2006) GIDEON, A program for design in structural DNA nanotechnology. J. Mol. Graphics Model. 25, 470–480. 27. Williams, S., Lund, K., Lin, C., Wonka, P., Lindsay, S., and Yan, H. (2008) Tiamat: a three-dimensional editing tool for complex DNA structures. The 14th International Meeting on DNA Computing, Prague, Czech Republic. 28. Nanoengineer-1 is a molecular design program developed by Nanorex, Inc (Bloomfield Hills, MI). http://nanoengineer-1.com/ content/ 29. Seeman, N. C. (1990) De novo design of sequences for nucleic acid structure engineering. J. Biomol. Struct. Dynam. 8, 573–581. 30. Wei, B., Wang, Z., and Mi, Y. (2007) Uniquimer: software of de  novo DNA sequence generation for DNA self-assembly– an introduction and the related applications in DNA self-assembly. J. Comput. Theor. Nanosci. 4, 133–141. 31. Ke, Y., Liu, Y., Zhang, J., and Yan, H. (2006) A study of DNA tube formation mechanisms using 4-, 8-, and 12-helix DNA nanostructures. J. Am. Chem. Soc. 128, 4414–4421.

Chapter 2 Protocols for Self-Assembly and Imaging of DNA Nanostructures Thomas L. Sobey and Friedrich C. Simmel Abstract Programed molecular structures allow us to research and make use of physical, chemical, and biological effects at the nanoscale. They are an example of the “bottom-up” approach to nanotechnology, with structures forming through self-assembly. DNA is a particularly useful molecule for this purpose, and some of its advantages include parallel (as opposed to serial) assembly, naturally occurring “tools,” such as enzymes and proteins for making modifications and attachments, and structural dependence on base sequence. This allows us to develop one, two, and three dimensional structures that are interesting for their fundamental physical and chemical behavior, and for potential applications such as biosensors, medical diagnostics, molecular electronics, and efficient light-harvesting systems. We describe five techniques that allow one to assemble and image such structures: concentration measurement by ultraviolet absorption, titration gel electrophoresis, thermal annealing, fluorescence microscopy, and atomic force microscopy in fluids. Key words: DNA, Self-assembly, Atomic force microscopy, Fluorescence microscopy, Nanostructures

1. Introduction Many of the properties that make DNA useful for genetic information transfer also make it useful for self-assembly of nanostructures. Researchers from physics, chemistry, biology, and computer science use DNA self-assembly to examine the fundamental theories and optimal conditions of self-assembly (1), cooperative effects and emergence; thermodynamics and mechanics of polymers (2–4); and biochemical algorithm execution, logic modules and circuits, error correction techniques, and computational demonstrations (5, 6). DNA as a molecule has several advantages when compared with other molecules. It is simple enough to be relatively well

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_2, © Springer Science+Business Media, LLC 2011

13

14

Sobey and Simmel

understood (compared with proteins), complex enough to build technically advanced structures (compared with many natural and artificial polymers); it can be chemically synthesized (and nowadays ordered from companies over the internet); and it is stable, reliable, and predictable enough to be confidently handled by researchers with little in the way of chemistry background. A major challenge in the field to date has been coping with exact stoichiometry requirements needed for high numbers of and/or physically large perfectly assembled structures. Typically, several different short (~5–100 bases) strands bind to each other, and if they are not at the correct absolute and relative concentrations then significant defects occur in the final assembled structures. Three ideas have been introduced in recent years to overcome this problem. Structures have been assembled that require only one carefully designed sequence which takes advantage of sequence symmetry principles (7). Examples of this include ­single-sequence DNA nanotubes shown in Figs. 3 and 4. Two other techniques: error avoidance protocols and DNA “origami” that have been also introduced are left for another discussion (8–10). To face the challenges of stoichiometry, one measures the absolute or relative concentrations of DNA very precisely. This is done in one of the two ways. The absorption of DNA at a light wavelength of 260 nm is dependent on its concentration, base sequence length and structure. If the DNA has a known base sequence and length, and does not have any structure ­(secondary structure), then its concentration can be related to its absorption (11). If the DNA does have some structure, then the correct concentration ratio with its complementary strands can be chosen by mixing it at different ratios (“titrating”), and analyzing these using titration gel electrophoresis (12). Having determined the concentrations of the DNA strands, they can then be mixed in appropriate buffer conditions and slowly annealed over several days from ~90 to 20°C to assemble the desired structures. To visualize the structures several options may be used (see Note 1). Fluorescent molecules that bind to DNA may be added and the structures viewed with a fluorescence microscope, which is relatively quick and easy. Significantly more challenging is to use an atomic force microscope and fluid cell, visualizing the structures using a scanning probe. This provides much higher resolution. Several examples of structures that are relatively stable and/ or simple to assemble have been developed by Mao and colleagues. These include single sequence lattices and nanotubes and structures consisting of three sequences, including lattices and polyhedra (7, 13–15).

Protocols for Self-Assembly and Imaging of DNA Nanostructures

15

2. Materials The following lists of materials and equipment are suggested, along with recommended suppliers. There are often many other good suppliers for these, the following are suggestions only, in particular with regards to equipment. All water used should be 18 MW and of pH 7–8. 2.1. Absolute Concentration by Ultraviolet Absorption Measurements

1. UV spectrometer (see Notes 2 and 3) (we use a V-630Bio, Jasco, Japan). 2. Cuvettes: 2 (Hellma, Germany). 3. Ultrapure water: 18 MW, pH 7–8. 4. Pipettes: 100, 2.5 mL (Eppendorf, Germany). 5. Centrifuge tubes: 0.5 mL (Eppendorf, Germany). 6. Clean compressed air/nitrogen and/or lens cleaning tissue. 7. DNA (Integrated DNA Technologies, USA).

2.2. Relative Concentration by Polyacrylamide Gel Electrophoresis

1. Gel electrophoresis system (such as the PerfectBlue Dual Gel System, Peqlab Biotechnologie, Germany). 2. Electrophoresis Power supply (such as the EPS301, GE Healthcare, USA). 3. Circulating cooling water at 4°C (see Note 4). 4. Detergent (such as 1104-1, Alconox, USA). 5. Ethanol in squirt dispenser. 6. Acrylamide–bisacrylamide: (Rotiphorese Gel 40, Roth, Germany). Warning: Acryilamide is a neurotoxin and carcinogen and should be handled with care in a fume cupboard. 7. 10× TAE Buffer: 400  mM Tris–acetate, 10  mM EDTA, pH 8.3. 8. Ultrapure water: 18 MW, pH 7–8. 9. MgCl2: 1 M (Sigma-Aldrich, Germany). 10. TEMED (Tetramethylethylenediamine, Germany).

Sigma-Aldrich,

11. APS: (Ammonium persulfate, Sigma-Aldrich, Germany) prepare fresh solutions weekly at 10% w/v in water. 12. Glass beakers: 2, 150 mL (Duran Group, Germany). 13. Pipettes: 100, 2.5 mL (Eppendorf, Germany). 14. Vacuum chamber and pump (such as model 2478257, Duran Group, Germany, or model MVP 015-4, Pfeiffer Vacuum, USA). 15. Aspirating pipettes: 25 mL (BD Falcon, USA).

16

Sobey and Simmel

16. Pipette Filler (VWR International, USA). 17. Bulldog clips. 18. Centrifuge tubes: 0.5 mL (Eppendorf, Germany). 19. 20 mL Syringe. 20. Needle (G 14 0.60 × 30 mm). 21. Gel loading buffer: 4 g sucrose, 25 mg bromophenol blue, 25  mg xylene cyanol, 25  mg Orange G (Sigma-Aldrich, Germany), H2O to 10 mL. Store in small aliquots at 4°C. 22. DNA ladder (Low Molecular Weight, such as N3233, New England Biolabs, USA). 23. SYBR Gold (Invitrogen, USA). This is toxic and should be handled carefully according to the manufacturer’s instructions. 24. Stiff plastic/card sheet larger than the gel plates. 25. Aluminum foil. 26. Staining tray: opaque plastic box with lid slightly larger than the size of the gel. 27. Gel documentation system (Molecular Imager Gel Doc XR, Bio-Rad, USA). 28. DNA strands (Integrated DNA Technologies, USA). 2.3. Thermal Annealing of DNA Nanostructures

1. 10× TAE buffer: 400  mM Tris–acetate, 10  mM EDTA, pH 8.3. 2. MgCl2: 1 M (Sigma-Aldrich, Germany) (see Notes 5 and 6). 3. Water: 18 MW, pH 7–8. 4. Membrane filter: 0.02  mm (Anotop 25 Plus, Whatman, England). 5. Beaker: 2–4 L (Duran Group, Germany). 6. Styrofoam box to fit beaker (see Note 7). 7. Boiling water to fill beaker (see Note 8). 8. Screw-top microtubes: 0.5 mL (VWR International, USA). 9. Zip-lock bag. 10. Metal weights (nuts and bolts). 11. Glass thermometer 0–100°C. 12. Pipettes: 100, 2.5 mL (Eppendorf, Germany). 13. DNA strands (Integrated DNA Technologies, USA).

2.4. Fluorescence Microscopy of DNA Nanostructures

1. Fluorescence microscope (see Note 9) (Olympus IX71, Olympus, Japan). 2. 10× TAE buffer: 400  mM Tris–acetate, 10  mM EDTA, pH 8.3. 3. Water: 18 MW, pH 7–8.

Protocols for Self-Assembly and Imaging of DNA Nanostructures

17

4. YOYO-1 (Invitrogen, USA). Warning: This is toxic and should be handled carefully according to the manufacturer’s instructions. 5. Microscope slides or cover slips, thickness 0 (Menzel, Germany). 6. Fingernail varnish. 7. Wavelength filter (U-MWIB2, Olympus, Japan). 8. Light source (X-Cite Series 120, EXFO Photonic Solutions, Canada). 9. Ascorbic acid (see Note 10). 10. Pipettes: 100, 2.5 mL. 11. DNA product. 2.5. Atomic Force Microscopy in Fluid of DNA Nanostructures

1. Atomic force microscope (we use a Multimode V, Veeco Instruments, USA): operated in intermittent contact ­(tapping) mode. 2. Fluid cell (Veeco Probes, USA). 3. Mica (50, Ted Pella, USA). 4. Metal puck (Ted Pella, USA). 5. Cantilevers (model DNP-S10, Veeco Probes, USA). 6. 10× TAE buffer: 400  mM Tris–acetate, 10  mM EDTA, pH 8.3. 7. Ultrapure water: 18 MW, pH 7–8. 8. Membrane filter: 0.02  mm (Anotop 25 Plus, Whatman, England). 9. Optical microscope. 10. Tweezers (such as model 5599, Ted Pella, USA). 11. Pipettes: 100, 2.5 mL (Eppendorf, Germany). 12. DNA product.

3. Methods 3.1. Absolute Concentration by Ultraviolet Absorption Measurements

1. Single-stranded DNA can have significant secondary structure (where bases in the same strand bind to each other). This alters the extinction coefficient and leads to incorrect concentration determination. With current models and technology, there is no way around this (apart from using strands that are designed not to have secondary structure) and the best way to circumvent this is to use titration gel electrophoresis. However, titration gel electrophoresis requires much more time and effort, and thus is usually only conducted when it is found the

18

Sobey and Simmel

lattice is not forming as desired from the concentrations determined by UV absorption measurements. 2. Turn on spectrometer; allow lamp and system to stabilize for 1–2 h. 3. Appropriate DNA sequences can be dissolved in water to a concentration of 100 mM, this can be determined from the information sheet accompanying the sequences. These should be briefly heated to 60°C and well vortexed to ensure ­complete mixing. 4. Calculate a molar extinction coefficient for each DNA sequence using the nearest-neighbor model (16–18) – for example using Scitools on the internet from Integrated DNA Technologies (see Note 11). 5. Rinse the cuvette under flowing water, shake water out by hand hard, repeat several times. Dry the outer surface with compressed air/nitrogen and lens cleaning tissue. 6. Load the cuvettes with 100 mL water, set parameters (depending on the model of spectrometer, these exact options may not be possible, but there should be similar possibilities): (a) Wavelength scan: 350–220 nm (b) Scan rate: 400 nm/min (c) Bandpass: 1 nm (d) Response (integration) time: medium 7. Measure baseline, set baseline subtraction. 8. Add 2  mL of DNA to the measurement cuvette without removing it from the spectrometer, stir with pipette tip for 10–20 s. 9. Measure absorbance, ensure that the absorbance lies between 0.1 and 1 or add or dilute DNA until this is the case. Also ensure that absorbance between 320 and 350 nm is extremely close to zero or apply an offset if it is not, read-off absorbance at 260 nm (note this may not be the peak maximum), (see Notes 12 and 13) see Fig. 1 for an example. 10. Calculate the concentration of DNA using the Beer-Lambert law: Absorbance = path length ´ extinction coefficient



´ concentration

Concentration =



(1)

absorbance  (2) path length ´ extinction coefficient

For example, with an absorbance of 0.5 and an extinction coefficient of 100,000 L/mol·cm and a cuvette of width 1 cm:

19

Protocols for Self-Assembly and Imaging of DNA Nanostructures

Fig. 1. A representative ultraviolet absorbance curve of a single-stranded DNA sequence. Note that the maximum is not exactly at a wavelength of 260 nm (this is dependent on the sequence); however, the absorbance is measured at 260 nm because this is the value that the extinction coefficient is normally calculated at. Note also that a baseline is measured directly before adding DNA, thus there is no vertical offset necessary, and this can be seen by the 0 values at higher wavelengths (320–350 nm).

Concentration =

0.5 = 5 mM 1cm ´ 100, 000(L/mol cm)

(3) 

11. Calculate the amount of water needed to be added achieve a concentration of 1 mM (the DNA nanostructures are ­generally assembled at a DNA concentration of 1 mM or less). Continuing the example:



Concentration initial ´ volume initial = concentration final ´ volume final (4)

concentration initial ´ volume initial concentration final 

(5)

5 mM ´ 102 mL = 510 mL 1 mM 

(6)

Volumeneeded = 510 - 102 mL = 408 mL. 

(7)

Volume final =

Volumefinal =

Add this amount and mix with pipette tip. 12. Transfer solution to a centrifuge tube (loss of small amounts here is not critical, if well mixed, the concentration will not change). 13. Repeat steps from three onward for all DNA strands.

20

Sobey and Simmel

1. Clean electrophoresis plates thoroughly with detergent and rinse thoroughly with water, wipe with ethanol then wipe dry.

3.2. Relative Concentration by Polyacrylamide Gel Electrophoresis

2. Place plates with spacers together and set in the electrophoresis unit, with the gap for the comb upward and inward. 3. Squirt ethanol in between the plates until approximately 1/4 full and leave for several minutes to ensure there are no leaks. 4. Mix Acrylamide–bisacrylamide, TAE buffer, MgCl2 solution and water in the following ratio per 10  mL of resulting ­solution (see Note 14) (Table 1). 5. Place in vacuum chamber for 5 min to remove air from the solution (this speeds up polymerization). 6. Check electrophoresis plates to see that there are no leaks, pour out ethanol. If there are leaks, pull the plates apart, put them back together again and recheck. 7. Remove solution from vacuum chamber. Divide solution gently into two (for two gels) without mixing in unnecessary air. 8. Prepare pipettes and tips for the APS and TEMED solutions. 9. Under the fume-hood, working without pause, add APS ­solution at 50 mL per 10 mL of solution to the first flask, and swirl gently to mix. Add TEMED (closing TEMED lid immediately) at 10 mL per 10 mL of solution to the second flask, and swirl gently to mix. 10. Immediately pipette (slowly to avoid bubbles) solution between the first set of gel electrophoresis plates, making sure no air bubbles get trapped. Fill until the level reaches the bottom of the gap for the comb. 11. Insert the comb, ensure that it traps no air bubbles, if this is the case take it out and reinsert it. It is useful to have the top of the comb slightly (~1 mm) above the upper edge of the glass plates (see Note 15). Use bulldog clips to hold the comb securely into position (otherwise, the expanding polymerizing gel displaces it). 12. Repeat steps 9–11 for the second solution. 13. Optimally, wait 90  min for the gel to polymerize (shorter times and the gel does not have polymerized completely with even and static pore sizes, longer than a couple of hours and the gel swells and dries) (see Note 16).

Table 1 Pipetting instructions for 10 mL of a 20% TAE/Mg2+ polyacrylamide gel Polyacrylamide gel (%)

Acrylamide–bisacrylamide (37.5:1) (mL)

10 ¥ TAE (mL)

1 M MgCl2 (mL)

H2O (mL)

20

5.00

1

120

3.88

21

Protocols for Self-Assembly and Imaging of DNA Nanostructures

14. Fill buffer between the gels and into the reservoirs of the electrophoresis unit. 15. Remove one of the combs and immediately flush the wells completely with buffer using the syringe and needle to remove unpolymerized acrylamide (see Note 17). 16. Repeat for second comb. 17. Set out nine 0.5 mL centrifuge tubes in a holder. 18. Calculate – from the concentrations (in mM) determined by UV absorption – the volume of the first strand needed for 100 ng (shorter strands run relatively faster and spread relatively wider, thus may need to be relatively more concentrated when run with longer strands). Use the formula: Volume(mL) =

100 ´ 10 -9 g (8) molecular mass(g/mol) ´ concentration(mol/mL)

Add this volume of the first strand to each tube.



19. Calculate the quantity in moles of the first DNA strand that this volume holds using:

Quantity = concentration ´ volume 

(9)

20. Calculate the volume needed of the second strand for each these (suggested) factors of the first DNA strand: 0.25, 0.5, 0.75, 1, 1.25, 1.5, 1.75, 2.

Volume =

quantity ´ factor  concentration

(10)

21. Add each volume to one of the tubes (there is one tube left). 22. Calculate the amount of 1 M MgCl2 solution needed to be added to each tube to give a concentration of 12.5 mM. Add this amount to each respective tube. 23. Close the tubes, label them, and vortex briefly. 24. Using a Polymerase Chain Reaction (PCR) machine, heat tubes to 90°C (lid temperature 92°C) for 1  min and cool evenly in small steps to room temperature over 1  h. This should ensure hybridization of the strands. 25. Add 0.4  mL of low molecular weight DNA ladder to the remaining tube. Pipette 3 mL of loading buffer into each tube and vortex. 26. Prepare running buffer and salt (1 L or more depending on the size of the electrophoresis unit): 20 mL 50× TAE buffer, 12.5  mL 1  M MgCl2, and water to fill to 1  L. Refrigerate until at 4°C. The buffer stock should be the same as that used for the gel and at the same concentration (1X). 27. Flush wells again with buffer using the syringe and needle, immediately before loading wells with DNA.

22

Sobey and Simmel

28. Load the first well with prepared DNA ladder solution and the rest of the wells with the DNA strand solutions in order. 29. Connect the electrophoresis unit to the circulating cooling water. This ensures that the gels remain cool while running and do not thermally denature the hybridized DNA samples (see Note 4). 30. Connect the electrophoresis unit to the power supply and run at a constant voltage of 10 V per cm gel length until the yellow loading dye runs to the bottom of the gel (typically 1–3 h). 31. During this time, the staining solution can be prepared. An opaque plastic container with a flat bottom just larger than the gel is needed, and this is filled with buffer to a depth that would be the same as the thickness of the gel. SYBR Gold is added at a ratio of 1 mL per 10 mL, this is covered with an opaque lid or aluminum foil and allowed to mix on a rotator at a small angle to the horizontal at the lowest speed (<1 Hz). 32. Prepare a sheet of aluminum foil about three times the size of the gel flat on the table. 33. When the gel is finished, turn off the power supply, and remove the gel from the unit. Use a thin blade or plastic scraper to carefully remove the top gel plate. A few droplets of water between the plate and the gel can help. Spread the aluminum foil over the top of the gel, then place a flat stiff piece of plastic over the aluminum foil. Use this to support the gel “sandwich,” as it is flipped up so that the bottom gel plate is now on top. This plate is also removed, two opposite sides of the foil are trimmed to the gel width, the remaining two sides are used to lift and support the gel, and the whole lot is placed in the staining container. The lid is placed on and the gel is left to stain on the rotator for 30–60 min. 34. The rotator is then stopped, the staining solution is removed using the 25 mL pipette and disposed of as toxic waste. 35. The gel is lifted out using the aluminum foil support and onto the UV light box. The gel can then be slid off the ­aluminum foil onto the glass using a few drops of water as lubricant if necessary. 36. The aluminum foil is stored safely for next time or disposed of as toxic waste. 37. The gel is examined/photographed using the gel documentation system and an appropriate wavelength filter for SYBR Gold making sure the focal distance of the camera is set to reach the gel and not to the inner UV bulbs (see Fig. 2 for an example). 38. The correct ratio of DNA strands is chosen by comparing the bands to see the one-to-one binding ratio. This ratio can be used with the excess (not used in the gel analysis) DNA to self-assemble the desired structure.

Protocols for Self-Assembly and Imaging of DNA Nanostructures

23

Fig. 2. A native PAGE gel electrophoresis titration analysis of the concentration of two complementary DNA strands. Each strand is 8 bases long, the gel is 20% and was run at 10 V/cm for 3 h. Lane 1: low molecular weight DNA ladder (766–25 bp); Lanes 2–9: relative concentration increments from factors of 2 to 0.25 as listed in the protocol. The upper band represents the hybridized DNA, the lower band excess single-stranded DNA: Lane 6 has the correct ratio of the two strands in this case, as there is no excess singlestranded DNA.

3.3. Thermal Annealing of DNA Nanostructures

1. The DNA strands are mixed at the correct concentrations and thermally annealed in appropriate buffer conditions (see Note 18). 2. All stock buffer and salt solutions should be filtered with a 0.02 mm membrane filter before use, this is quite critical. 3. The final volume and concentration of required DNA product is decided. Typically, volumes between 20 and 2,000 mL are produced, at concentrations between 50 and 1,000 nM. As an example, 100 mL of 5-stranded lattice at 500 nM total concentration is chosen. 4. The concentration needed of each strand is calculated, this depends on the stoichiometry of the strands in the final structure. Concentration of strand =



stoichiometry of strand total stoichiometry ´ total concentration (11)

5. For a strand mixed at a ratio of 1 with a total of five strands, this is Concentration of strand =



1 ´ 500 nM = 100 nM 5 

(12)

6. The volume needed from each DNA strand solution is calculated using Eq.  4. For example, if each strand solution has been diluted to 1,000 nM, then for 100 mL final volume:

1, 000 nM ´ volume initial mL = 100 nM ´ 100 mL 

(13)

24

Sobey and Simmel



Volume initial =

100 nM ´ 100 mL = 10 mL  1, 000 nM

(14)

7. The calculated volume of each strand is pipetted into a screwtop microcentrifuge tube. 8. The volume of 50× TAE buffer needed for a 1× solution of 100 mL is calculated using the Eq. 4. For example:

50 ´ volumeinitial mL = 1 ´ 100 mL. 

(15)



Volume initial =

1 ´ 100 mL = 2 mL  50

(16)

This volume is pipetted into the screw-top microcentrifuge tube. 9. The volume of 1,000 mM (1 M) MgCl2 solution needed for a 12.5 mM solution of 100 mL is

1, 000 mM ´ volume initial mL = 12.5 mM ´ 100 mL.  (17) Volume initial =

12.5 mM ´ 100 mL mL = 1.25 mL  1, 000 mM

(18)

This is pipetted into the screw-top microcentrifuge tube. 10. Enough water is added (here 46.75 mL) to make up the total required volume (here 100 mL). 11. The lid is screwed tightly onto the tube, and it is briefly (10 s each) centrifuged, vortexed, then centrifuged again. 12. The tube is placed into the zip-lock plastic bag with enough weights (nuts and bolts) to make sure it sinks, the bag is rolled up and secured with a rubber band, and a few small holes are made in it to allow the air to escape. 13. Enough (tap) water is boiled to fill the large (2–4 L) beaker. The bag with weights and microcentrifuge tube is placed in the bottom of the beaker, along with the thermometer. The beaker is filled with water just below boiling point. 14. The beaker is placed inside the Styrofoam box which is closed, and this is placed in a safe place and left for 48 h or until the water has cooled to room temperature (~20°C). 15. The microcentrifuge tube is then taken out and dried, to ensure that no water droplets on the outside enter upon opening the lid of the microcentrifuge tube. 3.4. Fluorescence Microscopy of DNA Nanostructures

1. If the chosen DNA structure has dimensions on the order of several micrometers or greater (for example, a large twodimensional lattice), then it may be viewed with a fluorescence microscope if it is “dyed” using an intercalating

Protocols for Self-Assembly and Imaging of DNA Nanostructures

25

fluorescent molecule, such as YOYO-1 (Invitrogen). This binds between base pairs of double-stranded DNA, optimally at a ratio of 1 dye molecule per 5 base pairs (see Note 19). 2. Thus, if the total number of bases of the 5 strands is 100, then the number of base pairs is 50. The YOYO-1 stock solution is 1,000,000 nM (1 mM) and typically one dyes a DNA structure solution of 10 mL. The volume of YOYO-1 needed is 1, 000, 000 nM ´ volume



initial

mL =

50 base pairs 5 ´ 500 nM ´ 10 mL



(19)

(50 base pairs/ 5) ´ 500 nM ´ 10 mL mL 1, 000, 000 nM = 0.05mL (20)

Volumeinitial =

3. This volume is not realistic to pipette, so the YOYO-1 is diluted in 1× TAE buffer, for example (100× dilution): 0.5 mL YOYO-1 stock solution, 2 mL 50× TAE buffer, 97.5 mL water in a plastic (YOYO-1 binds to glass containers) microcentrifuge tube. 4. With 100× dilution, 5 mL of this is pipetted into a microcentrifuge tube. 5. Using a cut-off tip 10 mL of DNA structure solution is added (see Note 20). 6. Ascorbic acid is used to minimize photobleaching of the fluorescent molecules. It is prepared at 100 mM in a volume of, for example, 10 mL. With a molecular mass of 176.12 g/mol this is:

Mass = molecular mass ´ concentration ´ volume  (21)



Mass = 176.12(g/mol) ´ 100 ´ 10 -3 (mol/L) ´ 10 ´ 10 -3 L  = 176 mg

(22)

This mass is dissolved in 10 mL of water and stored in a light proof jar. New solutions should be made every week. 7. This is added to a final concentration of 10 mM, so with 5 mL of YOYO-1 solution and 10 mL of DNA solution, one adds approximately 1.5 mL. 8. The fluorescence microscope is prepared, the light source is switched on, the correct filter is loaded, and an appropriate objective (40× air) is chosen. 9. For a very quick look, 1 mL of dyed-DNA solutions can be pipetted using a cut-off tip onto a Number 0 cover slip and placed on the microscope. There are large amounts of background fluorescence, but normally the structures themselves can also be seen.

26

Sobey and Simmel

10. For a better image, this process is repeated but the droplet is covered with a second cover slip and the edges of the cover slip are sealed with fingernail varnish. There is much less ­fluorescence background using this technique. 3.5. Atomic Force Microscopy on Fluid of DNA Nanostructures

1. The DNA structures bind in solution to a mica surface given the correct conditions, and the topography of the structure can then be measured/“visualized” using an atomic force microscope. The precise details of this protocol vary greatly depending on the model of atomic force microscope used. 2. The microscope and control computer are switched on, and the software loaded. 3. It is generally much easier to set the correct engage height of the cantilever above the mica surface in air, as the surface of the mica is difficult to see with an optical microscope when submerged in buffer. 4. The cantilever holder is set to a distance far enough from the surface to ensure that upon loading the cantilever the tip of the cantilever does not contact the surface. 5. Using a small optical microscope and tweezers, the tip is loaded correctly in the fluid cell. 6. The mica is loaded into the microscope (initially without any sample). 7. The fluid cell/holder is loaded into the microscope. 8. With the aid of the optical microscope that comes with the atomic force microscope, the laser spot is aligned onto the very end and center of the cantilever. This is important! 9. If there is a reflecting mirror, its angle is adjusted so the laser shines close to the center of the photodiode window. 10. The photodiode position is adjusted so that the laser is reflected directly at its center. 11. The steps 8–11 are repeated to fine-tune the system to ensure that the detected signal is high (with the laser reflecting very close to the end of the tip) and the deflection signal (relating to the laser reflecting onto the center of the photodiode) is minimal. 12. The surface of the mica is brought into focus of the optical microscope. The surface can be difficult to observe (being semitransparent), it can help to move around looking for cracks on the surface. The cantilever is brought to a level just before it comes into perfect focus, indicating that it is very close to, but not in contact with the surface (see Note 21). As a guide, as the cantilever moves closer to the focal height, a double image of the cantilever is seen, and this merges into one at the focal height.

Protocols for Self-Assembly and Imaging of DNA Nanostructures

27

13. The mica is then removed (this probably entails removing the fluid cell/holder also) and a fresh surface is prepared using (opaque) masking tape (see Note 22). This is best done by pressing firmly a strip of tape flat onto the mica on a table, lifting the far edge of the mica up so that it stands perpendicular to the table on its bottom edge, and peeling the tape slowly and evenly downward. A thin, complete, shiny layer of mica should have adhered to the tape. Quality of results may depend on the orientation of the tape relative to the mica, there is an optimal direction found by experimenting. 14. 5 mL of DNA structure solution is carefully pipetted onto the center of the mica using a cut-off pipette. 15. 1× TAE 12.5 mM MgCl2 (filtered through a 0.02 mm ­membrane filter before use) buffer solution is added to the mica, and/or the surface of the fluid cell/holder and/or through a tube into the fluid cell, dependent on the system. Care should be taken that no air bubbles are trapped on the cantilever. 16. The mica is carefully reloaded into the microscope. 17. The buffer has a different refractive index so that the laser beam travels a slightly different path, steps 8–11 may need to be repeated with small changes to optimize the measured laser signal. 18. The cantilever is tuned (generally using a function in the ­software) to ~5% below its resonant vibration frequency (see  Notes 23 and 24). The amplitudes used are much smaller than those in air, and should be adjusted to be above the level at which the tip sticks to the surface when imaging, but not so large that the sample is damaged by the tip’s ­vibrations. This is best determined through trial and error. 19. The amplitude set point (ratio of the free amplitude of vibration to the amplitude while imaging) is generally set just below 1, for example 0.98 (or 98%). However, this can vary greatly dependent on the system. 20. The imaging parameters are then set. Initial scan sizes and speeds are set small (1  mM) and slow (0.5  Hz) to prevent damage to the tip as it first “contacts” the surface. 21. The most important two other parameters are the integral and proportional gains, these should be initially set extremely small (exact values are system dependent). 22. The number of measurements per scan line (pixels) can be set to 256. 23. The “engage surface” function of the microscope is actuated. Several errors may occur during this process (see Note 25). 24. When correctly engaged on the surface, the imaging parameters are optimized. It is generally helpful to first withdraw

28

Sobey and Simmel

the cantilever from the surface slightly (several hundred nanometers), retune the cantilever to the correct frequency and drive amplitude, before reengaging the surface. 25. There are generally at least two “views” in the software, an “image” view and an “oscilloscope” view of the trace and retrace of the current line scan profile. The amplitude set point, which is slowly increased until the tip just no longer contacts the surface, is the best seen in the oscilloscope mode when the trace and retrace scan profiles significantly depart vertically from each other. It is then decreased to just below the value when they come back vertically on top of each other for optimal imaging. 26. The integral gain is gradually increased so that the trace and retrace scan profiles correlate optimally with one another without excessive noise being introduced into the signals. 27. The proportional gain is then adjusted similarly. 28. The scan size and speed can then be increased and suitable DNA structures for imaging are found. 29. The desired scan size is set, and the scan speed is slowed to 1–3 Hz, and the number of measurements per line is increased to 512 or 1024. An image is then captured. 30. Care should be taken that no imaging artifacts like double tip images (coming from broken tips with two or more points) or material sticking to the tip occur, if so the cantilever should be changed and the whole process repeated. 31. When a new sample is required, the process can be simplified if care is taken. The tip is withdrawn approximately 100 mm from the surface. If the cantilever is not moved within its holder, a thin layer of mica is removed from the same mica sample, and the mica is returned afterward to the same position in the microscope, then the cantilever should be relatively close to the surface and should not need long for the engage procedure. 32. Examples of atomic force microscopy images taken using this method are shown in Figs. 3 and 4.

4. Notes 1. There are other options, such as Transmission Electron Microscopy, that are not discussed here. 2. There are now ultraviolet absorption spectrometer systems that are designed to quickly measure mL volumes in the mM range. These may not be accurate enough for the standards required here.

Protocols for Self-Assembly and Imaging of DNA Nanostructures

29

Fig. 3. An atomic force microscopy image of self-assembled DNA nanotubes that have clumped together. Excess DNA that did not form nanotubes can be seen as a background carpet. This height image was captured using “tapping mode” in buffer on mica. Scale bar 1 mm, height scale 15 nm.

Fig. 4. An atomic force microscopy image of a self-assembled DNA nanotube that has connected both its ends (by chance). Thinner-tangled nanotubes can also be seen. This height image was captured using “tapping mode” in buffer on mica. Scale bar 500 nm, height scale 30 nm.

3. For DNA concentration measurements, temperature control (via a programmable water bath or Peltier element) of the sample while measuring absorption is not usually necessary; however, this is useful for making DNA melting measurements, often used to assist in analyzing these structures. 4. The gels should be “run” at 4°C, and this can also be achieved by placing the system in a cool room or refrigerator.

30

Sobey and Simmel

5. Salts are critical for these structures – they provide electrostatic shielding that allows the negative DNA strands to bind together. In standard DNA hybridization, salts with monovalent ions like Sodium Chloride are used. For the structures discussed here, salts with divalent ions such as MgCl2 are used, and these allow the DNA to “fold” into the desired complex structures.  12.5  mM concentration is generally ­chosen, and this is high enough for binding and folding and allows the large DNA structures to bind to the mica surface in atomic force microscopy. Higher volumes may cause ­condensation of the DNA or unwanted significant binding of any excess single-stranded DNA to the mica surface. 6. Alternatively, a large thermos flask can be used instead of a beaker and polystyrene box. 7. Instead of annealing in hot water, a programmable PCR machine can be used with small temperature steps, ensuring that the lid is a few degrees warmer than the heating block. 8. This can be normal tap water. 9. A Total Internal Reflection Fluorescence (TIRF) microscope is advantageous to remove background fluorescent light from sources not in focus (at the surface), but imaging is certainly manageable without such a system. 10. This helps to prevent photobleaching of the fluorescent molecules. 11. The extinction coefficient is calculated using a nearest-­neighbor model. One can, for example, make use of the online calculator “Scitools” provided by Integrated DNA Technologies at h t t p : // w w w. i d t d n a . c o m / a n a l y z e r / A p p l i c a t i o n s / OligoAnalyzer/ 12. The absorption spectrum of DNA is sequence dependent, and thus the UV absorption peak of DNA may be found between approximately 260 ± 15 nm; however, the absorption should be measured at the wavelength that the extinction coefficient is calculated for, which is generally 260 nm. 13. Depending on the cuvette, it may be necessary to stir the sample with a pipette tip to remove an air bubbles and make a second measurement to ensure reproducibility. 14. This gel concentration is suitable for DNA strands up to 100 bases long, for longer strands smaller gel concentrations are needed. 15. The gels can be stored for several days in their glass plates if wrapped with tensioned rubber bands to keep the comb pressed securely into the wells (with the upper edge of the comb above the glass taking some of the tension) and kept in buffer. Metal clips oxidize in buffer and should not be used.

Protocols for Self-Assembly and Imaging of DNA Nanostructures

31

16. Gels may be stored for several days if wrapped securely with rubber bands – ensure that there is tension holding the comb correctly in place; otherwise, the wells fill with unpolymerized acrylamide – and stored in 1× TAE solution. 17. A small battery head lamp can help to make the gel wells more visible (Petzl Tikka). 18. We have also developed a technique that does not rely on thermal annealing but rather on the basis of dilution of DNA denaturing agents in the buffer (19). 19. One YOYO-1 molecule every five base pairs gives the best ratio of minimal structural deformation of the DNA helix to maximal fluorescence intensity, giving an optimal signal to noise ratio (20). 20. The DNA structures can be so large that the normal hole diameter of the pipette tip damages them as they pass through. 21. It is important to come into focus on the surface from a starting point far away from the surface; otherwise, one may focus on the reflection and not on the real surface. 22. One can see the thin peeled layers of mica with better ­contrast if the tape is opaque. 23. The feedback loop in the electronics of the microscope works optimally at values just below the resonance frequency of the cantilever. 24. In the 10  kHz range with small buffer volumes, there may be  a resonance in the buffer itself which can be heard as a high-pitched tone. This is normal. 25. Several errors often occur while engaging, if these occur, the  engage should be aborted. The amplitude may change significantly (more than 10%), thus the cantilever should be retuned with the correct amplitude. The deflection errors may increase significantly, particularly if the buffer was initially at a different temperature to the fluid cell/holder and/ or mica, thus the errors should be brought to a minimum. Once corrected, the engage can be restarted.

Acknowledgments The authors sincerely thank Helene Budjarek for her technical expertise and assistance, and Rob Fee and Ralf Jungmann for helpful discussions. The authors acknowledge financial support from the Center for Nanoscience (Ludwig-MaximilianUniversität, Germany), the International Doctorate Program NanoBioTechnology (Elite Network of Bavaria), and the Nanosystems Initiative Munich.

32

Sobey and Simmel

References 1. Pelesko, J. A. (2007) Self Assembly: The Science of Things That Put Themselves Together, Chapman & Hall/CRC. 2. Zheng, J. W., Lukeman, P. S., Sherman, W. B., Micheel, C., Alivisatos, A. P., Constantinou, P. E., and Seeman, N. C. (2008) Metallic Nanoparticles Used to Estimate the Structural Integrity of DNA Motifs, Biophys. J. 95, 3340–3348. 3. Green, S. J., Bath, J., and Turberfield, A. J. (2008) Coordinated Chemomechanical Cycles: A Mechanism for Autonomous Molecular Motion, Phys. Rev. Lett. 101. 4. Dirks, R. M., Bois, J. S., Schaeffer, J. M., Winfree, E., and Pierce, N. A. (2007) Thermodynamic Analysis of Interacting Nucleic Acid Strands, SIAM Review 49, 65–88. 5. Seelig, G., Soloveichik, D., Zhang, D. Y., and Winfree, E. (2006) Enzyme-Free Nucleic Acid Logic Circuits, Science 314, 1585–1588. 6. Zhang, D. Y., Turberfield, A. J., Yurke, B., and Winfree, E. (2007) Engineering EntropyDriven Reactions and Networks Catalyzed by DNA, Science 318, 1121–1125. 7. Liu, H. P., Chen, Y., He, Y., Ribbe, A. E., and Mao, C. D. (2006) Approaching the Limit: Can One DNA Oligonucleotide Assemble into Large Nanostructures?, Angew. Chem.Int. Edit. 45, 1942–1945. 8. Soloveichik, D., Cook, M., and Winfree, E. (2008) Combining Self-Healing and Proofreading in Self-Assembly, Natural Computing 7, 203–218. 9. Shih, W. M., Quispe, J. D., and Joyce, G. F. (2004) A 1.7-Kilobase Single-Stranded DNA That Folds into a Nanoscale Octahedron, Nature 427, 618–621. 10. Rothemund, P. W. K. (2006) Folding DNA to Create Nanoscale Shapes and Patterns, Nature 440, 297–302. 11. Tataurov, A. V., You, Y., and Owczarzy, R. (2008) Predicting Ultraviolet Spectrum of Single Stranded and Double Stranded

12.

13.

14.

15.

16.

17.

18.

19.

20.

Deoxyribonucleic Acids, Biophys. Chem. 133, 66–70. Lu, M., Guo, Q., Marky, L. A., Seeman, N. C., and Kallenbach, N. R. (1992) Thermodynamics of DNA Branching, J. Mol. Biol. 223, 781–789. Zhang, C., He, Y., Chen, Y., Ribbe, A. E., and Mao, C. D. (2007) Aligning One-Dimensional DNA Duplexes into Two-Dimensional Crystals, J. Am. Chem. Soc. 129, 14134-+. He, Y., Chen, Y., Liu, H. P., Ribbe, A. E., and  Mao, C. D. (2005) Self-Assembly of  Hexagonal DNA Two-Dimensional (2d) Arrays, J. Am. Chem. Soc. 127, 12202–12203. He, Y., Ye, T., Su, M., Zhang, C., Ribbe, A. E., Jiang, W., and Mao, C. D. (2008) Hierarchical Self-Assembly of DNA into Symmetric Supramolecular Polyhedra, Nature 452, 198–U141. Breslauer, K. J., Frank, R., Blocker, H., and Marky, L. A. (1986) Predicting DNA Duplex Stability from the Base Sequence, Proc. Natl. Acad. Sci. U. S. A. 83, 3746–3750. Sugimoto, N., Nakano, S., Yoneyama, M., and Honda, K. (1996) Improved Thermodynamic Parameters and Helix Initiation Factor to Predict Stability of DNA Duplexes, Nucleic Acids Research 24, 4501–4505. SantaLucia, J., Allawi, H. T., and Seneviratne, A. (1996) Improved Nearest-Neighbor Parameters for Predicting DNA Duplex Stability, Biochemistry 35, 3555–3562. Jungmann, R., Liedl, T., Sobey, T. L., Shih, W., and Simmel, F. C. (2008) Isothermal Assembly of DNA Origami Structures Using Denaturing Agents, J. Am. Chem. Soc. 130, 10062–10063. Doyle, P. S., Ladoux, B., and Viovy, J. L. (2000) Dynamics of a Tethered Polymer in  Shear Flow, Phys. Rev. Lett. 84, 4769–4772.

Chapter 3 Self-Assembly of Metal-DNA Triangles and DNA Nanotubes with Synthetic Junctions Hua Yang, Pik Kwan Lo, Christopher K. McLaughlin, Graham D. Hamblin, Faisal A. Aldaye, and Hanadi F. Sleiman Abstract The site-specific insertion of organic and inorganic molecules into DNA nanostructures can provide unique structural and functional capabilities. We have demonstrated the inclusion of two types of molecules. The first is a diphenylphenanthroline (dpp, 1) molecule that is site specifically inserted into DNA strands and which can be used as a template to create metal-coordinating pockets. These building blocks can then be used to assemble metal-DNA 2D and 3D structures, including metal-DNA triangles, described here. The second insertion is a triaryl molecule that provides geometric control in the preparation of 2D single-stranded DNA templates. These can be designed to further assemble into geometrically well-defined nanotubes. Here, we detail the steps involved in the construction of metal-DNA triangles and DNA nanotubes using these methods. Key words: DNA, Self-assembly, Nanostructure, Transition metal, Nanotube

1. Introduction The field of structural DNA nanotechnology has evolved a number of elegant strategies for organizing materials on the nanometer length scale (1–3). By taking advantage of the information-rich biomolecule DNA, simple self-assembly can now be used to ­prepare unique structures in multiple dimensions and template the arrangement of more functional components with remarkable precision and control. While DNA represents one of nature’s most predictable selfassembling systems, greater structural and functional diversity can be realized through strategic modification using small organic and

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_3, © Springer Science+Business Media, LLC 2011

33

34

Yang et al.

inorganic molecules. Such molecules can augment the non-covalent interactions of nucleic acids and provide chemical advantages in the form of enhanced stabilization, and photophysical, redox, magnetic, and catalytic properties that are not observed naturally for DNA. Conceptually, this brings the toolbox of supramolecular chemistry and the ability of this field to generate a diverse array of structures and functions using synthetic molecules, and combines this area with the programmability of DNA. We recently introduced the term “supramolecular DNA assembly” to describe the melding of these two fields, whereby the self-assembly properties of DNA are complemented by those of synthetic components that have been site specifically inserted within the nucleic acid components (2). This approach is in contrast with more conventional DNA assembly methods that rely entirely on unmodified DNA, and may allow for unique structural and functional diversity. In this chapter, we describe the necessary steps to build two  DNA nanostructures using such small synthetic moleculemodified DNA. In one case, a DNA triangle acts as a template for site-specific metal incorporation (4), while in the other, it becomes a building block for the assembly of geometrically well-defined nanotubes (5). Organic molecules are first made compatible with solid-phase DNA synthesis via conversion to phosphoramidite derivatives, and incorporated as vertices or ligands within oligonucleotides that are designed to form discrete polygons (4, 6). The assembly process yields a single-stranded template for either immediate functionalization with transition metals or as a module used as the starting point for nanotube preparation. Such methodology utilizes the properties of both DNA and synthetic molecules in tandem to prepare dynamic nanoscale two-dimensional (2D) and 3D products that are formed quantitatively. The first system combines the programmability of DNA with the redox, photoactivity, and magnetic properties of transition metal complexes. The second system allows the creation of DNA nanotubes of deliberately controlled geometries, for application as selective host structures, as templates for nanowire fabrication, and as drug delivery tools.

2. Materials 2.1. Solid-Phase DNA Synthesis and Purification

1. Standard reagents and phosphoramidites for automated solidphase DNA synthesis (see Note 1). 2. 3′-Phosphate functionalized controlled pore glass (CPG) (Chemgenes) with 1,000 angstrom pore size and loading densities of ca. 30 mmol/g.

Self-Assembly of Metal-DNA Triangles and DNA Nanotubes with Synthetic Junctions

35

3. Phosphoramidite derivatives of 1 and 2 are prepared as ­previously reported (see Note 2). 4. 5-Ethylthio-1H-tetrazole (ETT, Sigma–Aldrich). 5. Acetonitrile - low water (EMD). 6. Concentrated ammonium hydroxide (NH4OH) solution (28%, Fisher Scientific). 7. Sephadex G-25 (superfine DNA grade) (Amersham Bio­ sciences) in pre-packed columns. 2.2. Polyacrylamide Electrophoresis

1. TB (10×): 900  mM tris (hydroxymethyl)-aminomethane (Tris) and 900 mM boric acid, pH 8. Store at room tempe­ rature. 2. TB (1×): prepared by tenfold dilution of TB (10×). 3. TAMg (10×): 400  mM Tris, 76  mM MgCl2 × 6H2O, and 14 mM glacial acetic acid, pH 8 (adjusted with small amounts of glacial acetic acid). 4. TAMg (1×): prepared by tenfold dilution of TAMg (10×). 5. 40% Acrylamide/bis solution (Fisher Scientific) (WARNING: the unpolymerized solution is a neurotoxin and care should be taken to avoid exposure) and N,N,N ′,N ′tetramethylethylenediamine (TEMED, Sigma–Aldrich). 6. Denaturing polyacrylamide gel electrophoresis (PAGE) solution (24%, 100 mL): mix 42.04 g urea, 10 mL TB (10×), and 60 mL 40% acrylamide solution and add water to adjust the volume to 100 mL. 7. Native PAGE solution (8%, 100  mL): mix 10  mL TAMg (10×) and 20 mL 40% acrylamide solution and add water to adjust the volume to 100 mL. 8. Ammonium persulfate (APS). 9. Gel combs: preparative 1.5-mm-thick single lane comb (Hoefer) and 15-lane 0.75-mm-thick comb (Hoefer). 10. Denaturing PAGE loading solution: 8  M solution of urea in H2O. 11. Native PAGE loading solution: mixture of glycerin and H2O (7:3 v/v). 12. Dye mixture: 1 mL formamide, 10 mM Na2EDTA, pH 8.0, 0.1% (w/v) bromophenol blue, and 0.1% (w/v) xylene cyanol. 13. Stains-All solution: 10% (w/v) solution of Stains-All (Sigma– Aldrich) made in formamide (98%): H2O (1:1 v/v). 14. Autoclaved H2O.

36

Yang et al.

2.3. Metalation of DNA Triangle T1

2.4. MALDI-TOF

1. Reaction buffer: 10 mM NaH2PO4–Na2HPO4, pH 7.2. 2. Cu(I) solution: 0.5 mM Cu(CH3CN)4PF6 acetonitrile solution (or Cu(NO3)2 and TCEP·HCl (Tris[2-carboxyethyl] phosphine hydrochloride) 1:2 mixture in water, final CuI concentration 0.5 mM). 1. Solution 1: 40  mg 6-aza-thiothymine (ATT) in 250  mL HPLC-grade acetonitrile. 2. Solution 2: 1.26  mg spermine in 250  mL autoclaved water (25 mM). 3. Mix solutions 1 and 2. Centrifuge briefly since some ATT may not be dissolved. 4. Fucose solution: 8.2 mg fucose dissolved in 1 ml autoclaved water (50 mM).

2.5. Chemical Ligation

1. Cyanogen bromide (5  M in acetonitrile, Sigma–Aldrich) (Warning: this is a toxic liquid and care should be taken in the handling of this compound). 2. Reaction buffer: 250  mM morpholineethanesulfonic acid (MES), pH 7.6, and 20 mM MgCl × 6H2O (see Note 3). 3. Powdered dry ice (see Note 4). 4. Two percent (w/v) solution of lithium perchlorate (LiClO4) in acetone (spectral grade, Fisher). 5. Sterilized razor blades. 6. Microcon size-exclusion centrifugal filter devices (Millipore, YM 10).

2.6. Enzymatic Digestion of 2D Templates

1. Exonuclease VII (ExoVII, source: recombinant, Amersham Biosciences, 1  U converts 1  nmol nucleotide to acid-­ soluble nucleotide in 30 min at 37°C under standard assay conditions). 2. TAMg (1×) buffer.

2.7. Atomic Force Microscopy

1. SPI-1 grade highly orderd pyrolytic graphite (HOPG) (SPI Supplies). Cleave the mica with tape to remove top layer and create a clean surface for sample deposition. 2. Etched silicon cantilevers (OMCL-AC160TS, Olympus).

3. Methods A variety of geometrically well-defined single-stranded DNA ­templates can be prepared by the following methods. We focus on the preparation of triangular structures T1 and T2 that have

Self-Assembly of Metal-DNA Triangles and DNA Nanotubes with Synthetic Junctions

37

either metal-coordinating ligand 1 or triaryl vertex 2 insertions, respectively. It is of note that these methods can be readily modified to create other 2D shapes (such as squares, pentagons, and hexagons). T1 is composed of three DNA strands (T1a-c), each doubly modified with the dpp ligand 1. These are brought together by the complementary regions of each strand to yield a single-stranded triangle with three preorganized metal-binding pockets (as shown in Fig. 1a). Subsequent metalation can then be characterized by gel electrophoresis, circular dichroism (CD), and thermal denaturation experiments. T2 is a triangle made from a linear strand (T2L) that is cyclized via a complementary template (T2P) and then chemically ligated to the closed product. The single-stranded regions of this polygon allow for further assembly into DNA nanotubes. First, a triangular rung (T2R) is prepared by the addition of strands R1–3 and L1–3 as shown in Fig. 1b, with sticky ends oriented above and below the plane of the triangle. These rungs can be assembled into tubes with the addition of three complementary duplexes

Fig. 1. (a) Diphenyl phenanthroline (1, inset ) inserted into a single-stranded DNA sequ­ ence. Hybridization of designed sequences into a triangle that template the forma­ tion of three metal-coordination environments, which can site-selectively bind CuI. (b) Triaryl vertex (2) inserted into a single-stranded DNA template (T2L) followed by cyclization with cyanogen bromide (CNBr) to yield single-stranded DNA template T2. A triangular rung structure (T2R) is then assembled from T2 by addition of DNA strands that are designed to provide sticky ends with orientational control. The nanotube is then assembled by connecting the rungs with double-stranded linkers dLS1–3.

38

Yang et al.

(dLS1–3). Nanotube formation is characterized by electrophoresis and atomic force microscopy (AFM). 3.1. Solid-Phase DNA Synthesis and Purification

1. All strands (T1a–c, T2L, T2P R1–3, and L1–3) are initially designed to minimize secondary interactions (see Note 5). Strands with site-specific insertion of 1 and 2 are made on 1,000 Å CPG in 0.5 mmol quantities. DNA strands are synthesized on a solid support until the required 1 or 2 modification is to be inserted. Synthesis is then halted and the 5′ dimethoxytrityl (DMT) protecting group is removed. The phosphoramidite of 1 (5  mmol) is dissolved completely in 200 mL of acetonitrile in a glass vial under argon. Separately, in a second glass vial under argon, ETT (15 mg) is dissolved in 200 mL of acetonitrile. The phosphoramidite mixture is quickly added to the ETT solution via a 1-mL disposable syringe. This acetonitrile–phosphoramidite–ETT solution is briefly mixed and then drawn quickly back into the syringe and attached to the solid support (see Note 6). A second syringe is attached and manual coupling of 1 is performed for a duration of 5 min at room temperature. All liquid is then drawn back into one of the syringes and disposed of. The solid support is placed back on the synthesizer and standard washing, capping, and oxidation steps are performed before continuation of DNA synthesis (see Note 7). Successful insertion of 1 is monitored by the DMT cation response that occurs after addition of subsequent bases. The above procedure is replicated for site-specific insertion of the phosphoramidite of 2. Only in the case of T2L synthesis is the 3′-phosphorylated CPG used, as this is required for the subsequent chemical ligation steps. 2. After synthesis, oligonucleotides are deprotected and cleaved from the solid support by placing the CPG into to a screw-cap vial and adding 1 mL of concentrated NH4OH. The vials are placed in a heating bath (55°C) for 16  h and periodically shaken to assist in deprotection. 3. The NH4OH solutions are transferred to 1.5-mL centrifuge tubes and samples are dried to completeness under vacuum. The samples are then resuspended in 250  mL of H2O and vortexed gently to ensure dissolution; 250 mL of 8 M urea is then added. This sample is gently vortexed and kept at 4°C until purification.

3.2. Purification and Quantification of DNA

1. These instructions assume the use of a Hoefer SE-600 gel system, but are easily adaptable to other formats. Glass plates (20 × 15 cm) are washed extensively using detergent. Before gel preparation, the plates are rinsed with ethanol (95%) and assembled using 1.5-mm spacers and the casting unit. A preparative 1.5-mm-thick comb is used for preparative PAGE.

Self-Assembly of Metal-DNA Triangles and DNA Nanotubes with Synthetic Junctions

39

2. Prepare a 1.5-mm-thick, 15% gel by mixing 31  mL of the 24% PAGE/urea solution with 19 mL of H2O (see Note 8). Mix this solution and add 50 mL TEMED and 50 mg APS. Pour the gel, insert the comb, and wait for polymerization (ca. 10 min). After polymerization, invert the gel, remove the comb, and wash all gel lanes with H2O. 3. After allowing the gel to set for ca. 30  min, add running buffer TB (1×) to upper and lower chambers of the gel unit and clean lanes with a Pasteur pipette (see Note 9). Carefully add the 500 mL solution of DNA/urea using a micropipette with loading tip. Add 10 mL of the dye mixture to a small gel lane. 4. Complete assembly of the gel unit and connect to power ­supply. Run gel at 250 V (30 mA) for 30 min (see Note 10), then change voltage to 500 V and continue to run for 1 h. 5. After the gel has completed running, remove buffer from upper chamber of the gel unit. Carefully remove the glass plates from gel and transfer to a piece of plastic wrap. 6. Place wrapped gel on top of a 20 × 20 fluorescent TLC plate. Remove upper plastic so that the gel is exposed on one side, irradiate with short wavelength UV light (254 nm) using a hand-held UV lamp, and quickly excise DNA band from the gel (see Note 11) using a sterilized razor blade. Place the excised gel portion into a 15-mL sterilized tube (see Note 12). 7. Using a sterilized pipette tip, crush the excised PAGE band and add 13 mL of H2O. Place in 55°C bath for 12 h. Shake periodically to assist in DNA extraction. 8. Remove aqueous layer from the 15-mL tube and reduce ­volume to ca. 1 mL by evaporation on a Speed-Vac. 9. Wash the storage buffer from the pre-packed Sephadex ­columns with ca. 25  mL of H2O. Add DNA solution and allow the column to drain completely. Add autoclaved H2O and collect 1–1.5 mL fractions in 1.5-mL plastic tubes (4–5 fractions should be collected). 10. Using a UV/Vis spectrometer, assay collected DNA fractions and record absorbance at 260 nm to obtain each concentration (see Note 13). 3.3. Denaturing PAGE and MALDI-TOF Characterization of DNA

1. Prepare a 0.75-mm-thick, 12% analytical denaturing PAGE using a similar method as in Subheading 3.2, step 2, but with 10 mL 24% denaturing PAGE solution and 10 mL H2O. 2. Prepare 0.01  nmol of each sample in 10  mL water and add 5 mL urea loading solution. Load the samples to each lane of the gel. Load the dye mixture to monitor the running of gel.

40

Yang et al.

3. Run denaturing PAGE at 200 V for 3 h. Stain the gel using Stains-All solution. The purified DNA shows a single band on the gel. 4. Characterize the DNA by MALDI-TOF MS (KOMPACT MALDI III mass spectrometer) (7). For 0.2–1 nmol DNA, reduce volume to dryness and then add freshly prepared 1 mL ATT/spermine solution and 1  mL fucose solution as the matrix. 3.4. Assembly of DNA triangle and Metalation

1. To prepare T1, first lyophilize 0.09 nmol of the three strands T1a, T1b, and T1c to dryness, and dissolve each in 10 mL phosphate buffer in a 200-mL microtube. Then, mix T1a, T1b, and T1c and incubate at room temperature for 15 min to generate the triangle product. 2. Cu(I) coordination to the dpp-modified DNA structure: add 0.6  mL of freshly prepared 0.5  mM Cu(I) solution to T1 and incubate at room temperature for 15 min. At high DNA concentration, a red color can be observed. 3. Run denaturing PAGE using the same conditions as in Sub­ heading 3.3, step 1. The metalated product shows a slower mobility band on denaturing PAGE, while the non-metalated product shows the bands with the same mobility of the individual components (Fig. 2a). 4. Prepare 0.75-mm-thick native PAGE using a method similar to that in Subheading 3.3, step 1, but using 20 mL 8% native PAGE solution. 5. Add 2 mL glycerin loading solution to 10 mL of each sample. Load each sample into a different lane and the dye mixture to monitor the running of gel. 6. Run at 85 V for 14 h at 4°C and then stain with Stains-All. The native gel demonstrates the sequential assembly of T1 (Fig. 2b). 7. The circular dichroism (CD) experiments are carried out on a JASCO J-810 spectropolarimeter at 25°C. Prepare a 15 mM sample in 10 mM phosphate buffer in a cell with a path length of 0.1 cm. A scan rate of 100 nm/min from 200 to 500 nm is used, and three trials are averaged. The signals for metal complex chirality can be observed as positive bands in 300–400 nm region (Fig. 2c). 8. Thermal denaturation experiments are performed on a UV–Vis spectrometer (Varian Cary 300 biospectrophotometer). Prepare 5 nmol of sample in 1 mL 10 mM phosphate buffer. Monitor the absorbance at 260 nm, heating the sample from 5 to 90°C and then cooling down from 90 to 5°C at the rate of 0.2°C/min. Calculate the melting temperatures

Self-Assembly of Metal-DNA Triangles and DNA Nanotubes with Synthetic Junctions

41

Fig. 2.  Assembly of T1 and metal coordination. (a) Native PAGE analysis demonstrates the quantitative formation of T1 from its component strands T1a-c. (b) Denaturing PAGE analysis of T1. While the non-metalated T1 denatures to its components (lane 4 ), the CuI coordinated T1 remains intact, giving a band with slower gel mobility (lane 5 ). (c) Circular dichroism characterization of T1 and T1.Cu3I. The metal binding gives a positive peak at 300–400 nm. (d) Thermal denaturation study of T1 and T1.Cu3I. The denaturation temperature dramatically increases after metal coordination.

(Tm) from the maximum of the first derivative of absorbance over temperature. The CuI coordination increases the Tm from 49°C for T1 to 76°C for T1.CuI3 (Fig. 2d). 3.5. Chemical Ligation and Purification to Obtain T2

1. Dry 0.42 nmol of T2L and 0.42 nmol of the template strand T2P. Resuspend each DNA sample in 15 mL of MES buffer. Vortex the samples gently and then mix together in a 0.6-mL plastic tube. Cool the sample in ice for 10 min. Repeat this step for ten individual experiments. 2. Add 10 mL of CNBr solution to each experiment and incubate in ice for 15 min. 3. Add 345 mL of 2% LiClO4 solution (w/v) to each experiment to precipitate DNA and immediately cool in powdered dry ice for 15 min. Centrifuge the samples at about 14,000 g for 3 min.

42

Yang et al.

Decant the supernatant and lyophilize the samples to dryness with heat. 4. Run a 12% denaturing PAGE using the same methods described in Subheading 3.3, steps 1–3. Load each experiment on one lane and run the gel at 250 V (15 mA) for 30 min, ­followed by 500 V for 1 h. 5. After the gel has completed running, carefully remove it and using a sterilized razor blade, cut off a single gel lane in its entirety and place into the Stains-All solution. Store the rest of the gel in water to keep it hydrated. 6. After the single lane has been stained for ca. 2 h, place the gel on a clean surface and analyze where the cyclic product is (Fig. 3b). Remove remainder of the gel from water and place next to the stained gel lane, making sure to align the gel properly (see Note 14). Based on the stained lane, use a sterilized razor blade to excise bands on the unstained gel portion (9 lanes) that correspond to the cyclic product. 7. Separate the excised band into a 15-mL tube. Extract DNA using water and remove salts and urea using the same method described in Subheading 3.2, steps 7–9. 8. Using a small volume UV/Vis cell, directly assay recovered T2 using a UV/Vis spectrometer. Record absorbance value at 260 nm (see Note 13).

Fig. 3. (a) Chemical ligation of T2. Denaturing PAGE analysis reveals the generation of chemically ligated product T2 as a single band of slower electrophoretic mobility when the assemblies (lane 1) are chemically ligated using cyanogen ­bromide (lane 2 ). (b) ExoVII nuclease analysis of T2 on denaturing PAGE. When the crude ligated product (lane 1) is treated with ExoVII (lane 2 ), the closed structure T2 (upper band) resists enzymatic digestion, while the open structure degraded (lower band). (c) Assembly of rung T2R. The single-stranded and cyclic template T2 (lane 1) is sequentially titrated with the complementary strands L1–3 (lanes 2–4, respectively), and R1–3 (lanes 5–6, respectively) to generate quantitatively a fully assembled triangular rung T2R.

Self-Assembly of Metal-DNA Triangles and DNA Nanotubes with Synthetic Junctions

3.6. Characterization of T2 by Enzymatic Digestion

43

1. Characterize single-stranded DNA template T2 by digestion with ExoVII (see Note 15). To assay the products directly after chemical ligation using enzymatic digestion, repeat Subheading 3.5, steps 1 and 2. Use 0.44 nmol of T2L and 0.84 nmol of the template strand in a single experiment. 2. After chemical ligation is completed, directly centrifuge the product mixture using a centrifugal filter device (Millipore, YM 10) to a volume of ca. 10 mL (see Note 16). 3. Add 30 mL of TAMg (1×) buffer to the centrifuged product. Divide the sample into two 20-mL portions. The first portion is stored at room temperature. 4. Add 5 U of ExoVII directly to the second portion with gentle mixing. Incubate the mixture at 37°C for 22 min and then at 95°C for 5 min in a Flexgene Thermocycler (any thermocycler can be substituted in this step) (see Note 17). 5. Run a denaturing PAGE using the same conditions as those in Subheading 3.3, steps 1–3. Load the two portions into separate lanes. Run the gel at 250 V (15 mA) for 30 min, followed by 500 V for 1 h. 6. Remove the gel and place in Stains-All solution for ca. 2  h. The cyclic structure T2 remains undigested compared to other products in the mixture (see Fig. 3b).

3.7. Assembly of Fully Double-Stranded Triangular DNA Nanotubes

1. Add 0.50 mL TAMg (10´) to an Eppendorf microtube. 2. Add 7.0 pmol of T2 template in water. 3. Add 7.0 pmol of three complementary strands L1, L2, and L3 that contain sticky-end overhang cohesions in water. 4. Add 7.0 pmol of the three rigidifying strands R1, R2, and R3 in water. Hybridization of these three strands orients each of the sticky ends on L1–3 into one of two lateral directions. 5. Dilute to a final volume of 4.00 mL with water. The concentration of the rung assembly is 1.75 mM. 6. Incubate the assembly at 95°C for 10 min and then slowly cool to 5°C over a period of 3.5 h to assemble the triangle rung T2R. Monitor the sequential assembly of rung T2R by native PAGE (see Subheading 3.8.1). 7. Prepare three double-stranded linking strands dLS1–3 with appropriate sequences by mixing the complementary strands in a 1:1 ratio at a scale of 7.0 pmol in TAMg (1´). It is typically convenient to combine 0.10 mL TAMg (10´) with the appropriate amount of each DNA strand in water, and dilute the solution to a final volume of 1.00 mL with water. 8. Add the double-stranded linking strand dLS1 to the already assembled triangular rung T2R in a 1:1 ratio.

44

Yang et al.

9. Incubate the assembly at 56°C for 10 min and then cool to room temperature over a period of 1 h. 10. Add the other two double-stranded linking strands dLS2–3 to the mixture in a 1:1 ratio in order to generate a final assembly of 1.00 mM in terms of rung. 11. Incubate the assembly at 44°C for 10 min and then cool to room temperature over a period of 1 h. 3.8. Characterization of DNA Nanotubes 3.8.1. Polyacrylamide Gel Electrophoresis

1. Titrate single-stranded triangle T2 in TAMg (1´) sequentially with the complementary strands L1–L3 and R–3 in TAMg (1´) at 1:1 mole ratios. For staining purposes, 28 pmol total material per lane gives clean bands. E.g. 28 pmol T2 (lane 1), 14 pmol T2 and 14 pmol L1 (lane 2), etc. 2. Run an 8% native PAGE using the same method as described in Subheading 3.4, steps 4–6. Load the seven different assemblies into seven lanes on the native gel. Run for 16 h at 80 mV below 15 mA. Stain the gel in Stains-All solution for 2 h (Fig. 3c).

3.8.2. Atomic Force Microscopy

1. AFM sample preparation typically involves the deposition of 2 mL of the 1.00 mM self-assembled mixture in TAMg (1´) buffer onto freshly cleaved HOPG, followed by evaporation

Fig. 4. Phase images of atomic force microscopy (AFM) analysis of nanotubes reveal the formation of well-defined one-dimensional DNA assemblies that extend over several microns, whose corresponding cross-sectional height analysis shows them to be of the same height.

Self-Assembly of Metal-DNA Triangles and DNA Nanotubes with Synthetic Junctions

45

to achieve complete dryness (typically 30 minutes in ambient conditions, with a cover to minimize dust and disturbances). 2. Conduct imaging within 24 h to minimize time-dependant sample degradation. 3. Place the sample HOPG on the imaging stage of the AFM. 4. Acquire AFM images in air at room temperature. Perform the “tapping mode” at a scan rate of 1 Hz using etched silicon cantilevers with a resonance frequency of ~70 kHz, a spring constant of ~2 N/m, and a tip radius of <10 nm. Acquire all images with medium tip oscillation damping (20–30%) (Fig. 4).

4. Notes 1. DNA synthesis protocols are based on the use of either a Biosystems Expedite 8900 or a Bioautomation MerMade MM6 machine, but can easily be adapted to other instru­ ments. 2. Phosphoramidites of 1 are prepared according to the procedures outlined in ref. 4. In a similar fashion, the phosphoramidite derivative of 2 is synthesized according to ref. 6. 3. The chemical ligation protocol is pH sensitive and care should be taken to adjust the pH to as close to 7.6 as possible using NaOH (8). 4. Pelleted dry ice can also be used, but care should be taken to ensure efficient packing of the DNA/acetone mixtures to promote precipitation. 5. A variety of software programs are now available for modeling DNA nanostructures and minimizing unwanted secondary interactions. Among these are Tiamat (9), Uniquimer (10), and IDT (11). The sequences used here are as follows:

Metalated DNA triangle sequences Strand name

Sequence (5 ¢→3 ¢)

T1a

TCTAGGAGAC-1-ACATTAGGTA-1-CTTTCAACTT

T1b

AAGTTGAAAG-1-GTTTGCTGGG-1-GTGATGTCAT

T1c

ATGACATCAC-1-CCGCCGATTA-1-GTCTCCTAGA

46

Yang et al.

Triangular DNA nanotube sequences Strand name Sequence (5 ¢→3 ¢) T2L

TATTGGTTTG-2-TGACCAATAACACAAATCGG-2-TCAGTAATCTCTTGAAGGTA2-GGAAACGACA

T2P

CAAACCAATATGTCGTTTCC

L1

CTCAGCAGCGAAAAACCGCTTTACTACCTTCAAGAGATTACTGAGTCTTG GAGTCGGATTGAGC

L2

TCGGCAGACTCTACTTGGTGCAAACCAATATGTCGTTTCCAGCATAGGAC GGCGGCGTTAAATA

L3

CGGTGCATTTAGTCGTGTCGCCGATTTGTGTTATTGGTCACGCGAATCA TGCGTACTCGT

R1

ACTCCAAGACTTCGACACGACT

R2

ATGATTCGCGTTCACCAAGTAG

R3

GTCCTATGCTTTGTAAAGCGGT

dLS1a

TTTTCGCTGCTGAGTGAACCAGAAAGTGGTCCTGTATATTTAACGCCGCC

dLS1b

ACTTTCTGGTTCA

dLS2a

AGTCTGCCGAGCTACCAGGTGAATTTCAAACGCAATTACGACGAGTACGC

dLS3a

AAATGCACCGGCTACCAGGTGAATTTCAAACGCAATTACGGCTCAATCCG

dLS23b

GTTTGAAATTCACCTGGTAGC

6. The adapters required to perform this off-column coupling step will be dependent on the DNA synthesizer being used. For example, Mermade synthesizers require a Female luer x  female luer adapter (Cole-Palmer, WU-45501-22). The Biosystems Expedite 8900 does not require any adapters. 7. These steps can be performed manually, or protocols can be directly modified and automated within the synthesizer. 8. The purification conditions here are meant for strands of length between 30 and 50 bases, i.e., T1a–c, T2L, L1–3, and R1–3. For the template strand T2P, a 19% PAGE solution should be used to obtain best separation results. Refer to ref. 12 for recommended PAGE purification conditions with respect to oligonucleotide length. 9. Gel lanes should be extensively washed to remove any unpolymerized acrylamide that can cause unwanted mobility effects in PAGE or failure of the sample to penetrate the gel matrix (12). 10. The initial 30 min at 250 V helps to “stack” the DNA within the gel matrix, leading to a uniform and condensed band. This step helps toward reliable excision from gel without n − 1 products.

Self-Assembly of Metal-DNA Triangles and DNA Nanotubes with Synthetic Junctions

47

11. Care should be taken to avoid prolonged exposure of DNA (<15 s) to the short wavelength UV light as degradation can occur. 12. Effort should be taken to use only sterilized plasticware and, if possible, materials that are certified RNase–DNase free. 13. Molecule 2 has an extinction coefficient value of 3.5 × 104  L/mol cm at 260 nm. For accurate concentration values of T2L, this must be accounted for in the absorbance readings using Beer’s law (ATotal = A2 + ADNA). 14. It is important to align the stained lane with the remaining hydrated PAGE section to facilitate accurate excision of the product band. 15. ExoVII is a bidirectional (5′ and 3′ active) exonuclease that is selective for the digestion of single-stranded open form DNA and will not be effective at digesting the cyclic structure T2. 16. Precipitation using LiClO4/acetone was found to disrupt normal ExoVII function when assaying the product mixture directly after the chemical ligation step. 17. The 95°C period ensures that ExoVII is denatured. In cases where hybridized structures are being assayed, products should be loaded quickly onto gel and the heat denaturation phase should not be used. 18. DNA samples deposited on AFM mica should be dried ­overnight under vacuum in the summer time. References 1. Seeman, N. C. (2007) An overview of structural DNA nanotechnology, Mol. Biotech. 37, 246–257. 2. Aldaye, F. A., Palmer, A. L., and Sleiman, H. F. (2008) Assembling materials with DNA as the guide. Science, 321, 1795–1799. 3. Lin, C., Liu, Y., and Yan, H. (2009) Designer DNA nanoarchitectures. Biochemistry 48, 1663–1674. 4. Yang, H. and Sleiman, H. F. (2008) Templated synthesis of highly stable, electroactive and dynamic metal-DNA branched junctions, Angew. Chem. Int. Ed. 47, 2443–2446. 5. Aldaye, F. A., Lo, P. K., Karam, P. McLaughlin, C. K., Cosa, G. and Sleiman, H. F. (2009) Modular construction of DNA nanotubes of  tunable geometry and single- or doublestranded character, Nat. Nanotech. 4, 349–352. 6. Aldaye, F. A. and Sleiman, H. F. (2007) Dynamic DNA Templates for gold nanoparticle discrete structures: control of geometry, particle identity, write /erase and structural switching, J. Am. Chem. Soc. 129, 4130–4131.

7. Distler, A. M. and Allison, J. (2001) Improved MALDI-MS analysis of oligonucleotides through the use of fucose as a matrix additive, Anal. Chem. 73, 5000–5003. 8. Carriero, S. and Damha, M. J. (2003) Template-mediated synthesis of lariat RNA and DNA. J. Org. Chem. 68, 8328–8338. 9. Williams, S., Lund, K., Lin, C., Wonda, P., Lindsay, S., and Yan, H. (2009) Tiamat: a three-dimensional editing tool for complex DNA structures. Lecture Notes in Computer Science, 5347, 90–101. 10. Zhu, J., Wei, B. Yuan, Y., and Mi, Y. (2009) UNIQUIMER 3D, a software system for structural DNA nanotechnology design, ­analysis and evaluation. Nuc. Acid. Res. 37, 2164–2175. 11. IDT Inc. Oligo analyzer. http://www.idtdna. com/ANALYZER/Applications/OligoAnalyzer 12. Albright, L.M. and Slatko, B.E. (2003) Appendix 3B, denaturing polyacrylamide gel electrophoresis. Current Protocols in Nucleic Acid Chemistry. A.3B.1-A.3B.5. John Wiley and Sons, Inc

Chapter 4 DNA-Templated Pd Conductive Metallic Nanowires Khoa Nguyen, Stephane Campidelli, and Arianna Filoramo Abstract We here present a protocol for the metallization of DNA scaffolds by palladium. The method is based on the initial slow precipitation of palladium oxide onto DNA strands. A reduction step follows to create conductive metallic nanowires. The slow oxide precipitation approach enables the formation of thin and continuous coatings on the DNA strands with negligible parasitic metallization of the remaining substrate surface. Key words: DNA metallization, Metal nanowires, Templated self-assembly, Nanoelectronics, Molecular combing

1. Introduction In nanotechnology, self-assembly is a valuable strategy where the design, manufacture, and control on the scale of a few nanometers are governed by molecular and supramolecular affinities (like structural and chemical properties). In particular, the exceptional recognition properties of DNA make it an ideal candidate to encode instructions for such nanoscale assembly to create scaffolds (1, 2) and to incorporate other materials in the self-assembly process (3). In nanoelectronics, it is highly desirable to utilize DNA not only as a positioning scaffold but also for electrical interconnections. Since DNA transport properties are still under discussion (4), it is pragmatically envisioned to develop a method where electronic conduction is ensured by a metallic coating selectively deposited onto the DNA strand. During the past 10  years, numerous methods to metalize DNA scaffolds have been developed (5). Different metals have been used in the metallization process based on silver (6, 7)

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_4, © Springer Science+Business Media, LLC 2011

49

50

Nguyen, Campidelli, and Filoramo

or copper (8) ions, gold nanoparticles (9, 10), platinum (11–18), or palladium (19–24) complexes. Recently, Pd metallization appeared of particular interest (25) since it has been shown to be the best choice for electrodes connecting single-wall carbon nanotubes (SWNTs). SWNTs occupy a special place within molecular electronics (26) and a particular attention is given to methods that can lead to the self-assembling and electrical connection of these nanoobjects. In the DNAdirected vision, the metallization process should be the last step of the fabrication process to use the DNA recognition properties as long as possible. Therefore, the most proficient scenario consists of DNA already deposited in the definitive circuit arrangement onto the substrate, where the final metallization can take place. The ultimate goal of the metallization process is to obtain thin metallic nanowires, while minimizing parasitic metal cluster nucleation and growth on the surrounding surface. In addition, attention should be devoted to both structural and conductive homogeneity of the obtained nanowires; these two issues are crucial for the use of DNA scaffolding in a circuit. Neither a big resistance value nor a large desorption of nanostructures during the process are acceptable. Here, we present a protocol to metallize DNA with Pd. The DNA is first combed on a surface to facilitate the evaluation of the yield and uniformity of the metallization process. The surface is SiO2, the standard for micro and nanoelectronics, and it is passivated by a silanization process to minimize the parasitic metallization. The chosen silane is hexamethyldisilazane (HMDS). After the deposition of the DNA nanoobjects to be metallized (as for the case of combed DNA for which we here give instructions) a thin DNA-templated wire of palladium oxide is then grown with a strong preference over precipitation on the substrate ­surface. Reduction of the oxide to metal makes the obtained nanowires conductive. One of the main advantage of this method is that it allows to have uniform metallization and preserve samples against large desorption. This is not always the case for other published protocols where cycling of different solutions is needed and where fast reduction kinetics are employed. It is worth to note that these two aspects are generally difficult to control to achieve a good homogeneity of the nanowire geometry and properties. In Subheading  3.1, we describe the surface treatment of SiO2 substrates with hexamethyldisilazane (HMDS); this is performed in order to limit parasitic metallization of the substrate when DNA strands are metalized. Subheading 3.2 explains the surface characterization procedure for testing the above-mentioned surface treatment: contactangle measurement.

DNA-Templated Pd Conductive Metallic Nanowires

51

Young’s Equation g sv = g sl + g lv cosq g lv VAPOR LIQUID

g sl

θ

g sv

SOLID

Fig. 1. Definition of the contact angle. On a hydrophilic substrate q is small, while it gets larger (larger than a right angle) on a hydrophobic surface.

The contact angle (q) at the liquid–vapor interface is a widely used characterization method to evaluate the degree of wettability of the surfaces. It is defined as the angle between the tangent to a liquid droplet deposited on a planar solid surface and the solid surface itself. In a first approximation, the relation between the contact angle and the interfacial tensions of the three phases (vapor, liquid, and solid) can be described by the Young’s equation (see Fig. 1). In the case of a water droplet, the contact angle characterizes the hydrophobicity of the surface: in our case, it helps to monitor the presence of CH3 groups introduced by the silanization. Subheading  3.3 describes how to obtain substrates with adsorbed combed DNA, which are ideal for the verification and optimization of the metallization procedure. The combing principle and setups are well described in literature (27, 28). Subheading 3.4 describes the protocol for the mineralization and the metallization of the surface-adsorbed DNA molecules. The rationale of this protocol is the following (and it is described in Fig. 2): after depositing DNA on the substrate, a selective precipitation of the oxidized form of Pd on DNA is realized, toward the formation of a nanowire, according to the following reaction (29–31):

K 2PdCl 4 + H 2O → PdO ↓ + 2HCl + 2KCl. Here, the final diameter, continuity, and uniformity of the nanowire are realized. At this stage, however, the nanowire is nonconducting. The reduction of oxides with metallic behavior is then performed, only when the nanowire is completely formed.

52

Nguyen, Campidelli, and Filoramo

Fig. 2. Scheme of the metallization by precipitation/reduction approach. (a) The palladium oxides precipitate onto the DNA scaffold. A palladium oxide nanowire is obtained. (b) the palladium oxide nanowire is converted to a metallic palladium one by a reducing agent.

2. Materials 1. Deionized (DI) water, 18.2 MW resistivity (Millipore). 2. Si/SiO2 subtrates: As doped Si n type <100> orientation with 150 nm of SiO2 (Siltronix). 3. Hexamethyldisilazane (CH3)3 SiNHSi(CH3)3 HMDS (99.9% purity, Sigma-Aldrich). 4. Piranha solution: 1:3 by volume mixture of 30% hydrogen peroxide and concentrated sulfuric acid. (Caution: Piranha solution reacts violently with organic materials and should be handled with extreme care.) 5. Dipotassium tetrachloropalladium complex (K2PdCl4) (Sigma-Aldrich): the K2PdCl4 solution is prepared at 20 mM and let hydrolyze during 2 weeks at 4°C. The hydrolization degree is checked by measuring the pH of the solution in analogy of what is done for Pt complexes (18). See Fig. 3 for the pH trend over time. 6. Lambda phage DNA, methylated, from Escherichia coli (Sigma-Aldrich). 7. ULSI (ultra large-scale integration) grade acetone (SigmaAldrich). 8. MES–NaCl buffer: 10  mM MES buffer (Sigma), 50  mM NaCl, pH 5.5.

DNA-Templated Pd Conductive Metallic Nanowires

53

pH value

4,0

3,5

0

2

4

6

8

10

12

14

Time (days) Fig.  3. Evolution of pH vs. time of Pd complexes solution. It indicates the degree of hydrolysis.

9. Hydrogen gas, pure (compressed). 10. Lambda DNA solution: 7  ng/ml l-DNA (Sigma) in MES– NaCl buffer.

3. Methods 3.1. Silanization of SiO2 with Hexamethyldisilazane

1. A 2-in. silicon wafer (with a 1,500 Å surface layer of SiO2) is treated with piranha solution for 1  h, then carefully rinsed with deionized water. 2. A step of RIE (reactive ion etching) in a O2 plasma is then performed to increase the reactivity of the surface (i.e., the density of surface silanols). The parameters for the RIE are 130 V, 2 min, 5 mbar. 3. The wafer is then baked for 15 min at 170°C to achieve dehydration, thus preventing HMDS from reacting with water. 4. The substrate is put in a sealed glass chamber (500 ml) with a small open glass container with 100 ml of HMDS for 30 min (in air). 5. The wafer is carefully rinsed with deionized water then with acetone (see Note 1). The HMDS monolayer formed can then be characterized by two methods.

3.2. Surface ContactAngle Measurement

We measure the contact angle by using the simple experimental setup schematized in Fig. 4. 1. Place the solid substrate on a plane goniometer. 2. Deposit a 10 mL water droplet on the surface with the aid of a micropipette.

54

Nguyen, Campidelli, and Filoramo

Fig. 4. Scheme of contact-angle measurement setup.

Fig. 5. Example of recorded image for contact-angle measurement.

3. Shine white light on the droplet by using a lamp and record the image of the droplet on the surface by using a CCD camera. 4. Evaluate the contact angle from the picture (as shown graphically in Fig. 5). As an example, we report in Fig. 6 the measured contact angle vs. the HMDS silanization time (see Note 2). 5. AFM imaging can also be performed at this stage in order to test the flatness and cleanliness of the surface (see Note 3). 3.3. DNA Combing on HMDS-Coated Silicon

1. Dip coat the substrate in a l-phage DNA solution in MES–NaCl pH 5.5 by applying a meniscus receding speed of ~100 mm/s. We used a combing setup as schematized in Fig. 7. 2. Evaluate the results of combing by AFM imaging of the dry substrate. In Fig. 8, an AFM image of the combed DNA is shown.

DNA-Templated Pd Conductive Metallic Nanowires

55

Fig. 6. Pilot of the contact angle on HMDS–SiO2 as a function of the time of silylation.

Fig. 7. Scheme of combing technique by receding meniscus.

Fig. 8. l-DNA combed onto a HMDS/SiO2 substrate.

56

Nguyen, Campidelli, and Filoramo

Fig. 9. Micrographs of the sample surface before and after the metallization process. (a) AFM phase image of a 13 × 13 mm2 area after combing and prior to metallization. (b) SEM image of a 20 × 15  mm2 area of the same sample after metallization.

3.4. DNA Mineralization and Metallization

1. The sample with the combed DNA is immersed in the ­precipitation solution contained in a microtube. Then, it is incubated at 45°C for 42 h, in a thermostat (see Notes 4 and 5). 2. The sample is then removed from the microtube, carefully rinsed with deionized water and dried with nitrogen. 3. Reduce the nanowires by placing the sample in a static 1 bar hydrogen atmosphere overnight at 400°K (see Note 6). 4. XPS analysis can be used at this stage to verify the extent of reduction (see Note 7). 5. SEM or AFM imaging can give insight on the quality and homogeneity of the nanowires. In Fig. 9, we report, respectively, AFM (Fig. 9a) and SEM images (Fig. 9b) before and after the process. It is possible to evaluate that nearly no DNA desorption and breakage is induced by the process. Higher resolution micrographs can be analyzed in order to evaluate the nanowire thickness and homogeneity (25).

4. Notes 1. This rinsing step is intended to remove any physisorbed layer or hydrolyzed silane, from the reaction of HMDS with any residual water. 2. We note that the contact angle value saturates around 80° indicating a limitation in the number of the grafted trimethylsilyl (TMS) groups. This phenomenon has already been observed and it can be explained by a reached equilibrium between the grafting process and the partial destruction of TMS by the released ammonia following Belyakova et al. (32).

DNA-Templated Pd Conductive Metallic Nanowires

57

3. AFM tapping-mode phase imaging (such as that showed in Fig. 8) is useful in evidencing the presence of HMDS layers (especially easy to visualize on patterned substrates) as the height of the HMDS molecules covering the surface is below 3Å and a low contrast is commonly obtained in height or amplitude images. 4. During the incubation time, the PdO deposits onto DNA in a very slow and progressive way, allowing the control of morphology of the PdO nanowire from coarse discontinuous to continuous. In this way, it is possible to obtain extremely regular, thin, and continuous nanowires down to a diameter of 20–25  nm. At the end of this step of PdO deposition, the nanowires are completely formed and similar in terms of size and shape. Thanks to the surface passivation, the precipitation process is selective and extremely well templated by the DNA scaffold. Only limited parasitic precipitation on the substrate is observed for the typical duration of experiments yielding the desired nanowires. 5. The temperature and time of incubation are key parameters of the reaction to control the homogeneity of synthesized nanowires. Those reported are the results of parameter optimization in our conditions and adjustments might be necessary (see also ref. 25). 6. For the reduction step, two reducing agents were studied: dimethylamine borane (DMAB) and hydrogen. The DMAB reduction can be performed by 15 min immersion of the sample in a 20 mM DMAB solution. In this case, XPS measurement demonstrated that only first layer of Pd is reduced, but in our hands, hydrogen gas gave the best results (for reduction H2 no form of oxidized Pd can be detected). 7. It should be noted that XPS analysis is a surface technique and only the first few nanometers can be characterized. Therefore, any conclusion cannot be done for the more internal core of the nanowire. However, the metallic character of the nanowire was also confirmed by electrical measurements. The resistivity of single nanowires was determined by us from the I/V curves and was found to be 3 × 10−6 Wm which is one order of magnitude higher than that of bulk Pd (1.05 × 10−7 Wm) (see also ref. 25).

Acknowledgments This work was partially funded by the NUCAN-NMP Strep 013775 project and the French Ministry of Research through the ACI Bio-NT project.

58

Nguyen, Campidelli, and Filoramo

References 1. Seeman N.C. (2003) DNA in a material world. Nature 421, 427–431. 2. Seeman N. C. (1991) the use of branched DNA for nanoscale fabrication. Nanotechnology 2, 149–159. 3. Mirkin C. A. (2000) Programming the assembly of two- and three-dimensional architectures with DNA and nanoscale inorganic building blocks. Inorg. Chem. 39, 2258–2272. 4. Guo X. , Gorodetsky A. A., Hone J., Barton J. K., Nuckolls C. (2008) Conductivity of a single DNA duplex bridging a carbon nanotube gap Nature Nanotechnology 3, 163–167. Xu B. , Zhang P., Li X., Tao N. (2004) Direct Conductance Measurement of Single DNA Molecules in Aqueous Solution Nanoletters 4, 1105–1108 . Storm A. J., van Noort J., de Vries S., Dekker C. (2001) Insulating behavior for DNA molecules between nanoelectrodes at the 100 nm length scale App.Phys. Lett. 79, 3881–3883. 5. Richter J. (2003), Metallization of DNA Physica E, 16, 157. 6. Yan H., Park S.H., Finkelstein G., Reif J.H., LaBean T.H. (2003) DNA-Templated SelfAssembly of Protein Arrays and Highly Conductive Nanowires Science 301, 1882–1884. 7. Braun E., Eichen Y., Sivan U., Ben-Joseph G. (1998) DNA-templated assembly and electrode attachment of a conducting silver wire Nature 391, 775–778. 8. Monsoon C.F., Woolley A.T. (2003) DNATemplated Construction of Copper Nanowires Nanoletters 3, 359–363. 9. Maubach G., Born D., Csaki A., Fritzsche W. (2005) Parallel Fabrication of DNA-Aligned Metal Nanostructures in Microelectrode Gaps by a Self-Organization Process Small 1, 619–624. 10. Kumar A., Pattarkine M., Bhadbhade M., Mandale A.B., Ganesh K.N., Datar S.S, Dharmadhikari C.V., Sastry M. (2001) Linear Superclusters of Colloidal Gold Particles by Electrostatic Assembly on DNA Templates Advanced Materials 13, 341–344. 11. B. Lippert, in Cisplatin: chemistry and biochemistry of a leading anticancer drug; WILEY-VCH, Weinheim, Germany 1999. 12. Macquet J.P., Theophanides T. (1975) Spécificité de l’interaction DNA-platine dosage du platine, pH métrie Biopolymers 14, 781–799. 13. Ford E.W., Harnack O., Yasuda A., Wessels J. M. (2001) Platinated DNA as Precursors to

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24. 25.

Templated Chains of Metal Nanoparticles Advanced Materials 13, 1793–1797. Colombi Ciacchi L., Mertig M. , Seidel R., Pompe W., de Vita A. (2003) Nucleation of platinum clusters on biopolymers: a first principles study of the molecular mechanisms Nanotechnology 14, 840–848. Macquet J.P., Theophanides T. (1976) DNAplatinum interactions. Characterization of solid DNA-K2PtCl4 complexes Inorganic Chimica Acta 18, 189–194. Macquet J.P., Butour J.L. (1978) Circular dichroism study of DNA Platinum complexes – Differentiation between monofunctional cis—­ bisdentate and trans-bidentate platinum fixation on a series of DNA Eur. J. Biochem. 83, 375–387. Mertig M., Colombi Ciacchi L., Seidel R., Pompe W., De Vita A. (2002) DNA as a Selective Metallization Template Nanoletters 2, 841–844. Seidel R., Ciacchi L.C., Weigel M., Pompe W., Mertig M. (2004) Synthesis of Platinum Cluster Chains on DNA Templates: Conditions for a Template-Controlled Cluster Growth J. Phys. Chem. B 108, 10801–10811. Richter J., Seidel R., Kirsch R., Mertig M., Pompe W., Plaschke J., Schackert H.K. (2000) Nanoscale Palladium Metallization of DNA Advanced Materials 12, 507–510. Richter J., Mertig M., Pompe W., Monch I., Schackert H. K. (2001) Construction of highly conductive nanowires on a DNA template Appl. Phys. Lett. 78, 536–538. Deng Z., Mao C. (2003) DNA-Templated Fabrication of 1D Parallel and 2D Crossed Metallic Nanowire Arrays Nanoletters 3, 1545–1548. Richter J., Mertig M., Pompe W., Vinzelberg H. (2002) Low-temperature resistance of DNA-templated nanowires Appl. Phys. A 74, 725–728. Lund J., Dong J., Deng Z., Mao C., Parviz B. A. (2006) Electrical conduction in 7 nm wires constructed on l-DNA Nanotechnology 17, 2752–2757. Nguyen K., Streiff S., Lyonnais S., Goux-Capes L., Filoramo A., Goffman M., Bourgoin J-Ph. (2006) AIP Conference Proceedings 859, 39. Nguyen K., Monteverde M., Filoramo A., Goux-Capes L., Lyonnais S., Jegou P., Viel P., Goffman M. J.P. Bourgoin (2008) Synthesis of Thin and Highly Conductive DNA-Based Palladium Nanowire Advanced Materials 20, 1099–1104.

DNA-Templated Pd Conductive Metallic Nanowires 26. Avouris P., Chen Z., Perebeinos V. (2007) Carbon based electronics Nature Nanotechnology 2, 605–615. 27. Bensimon D., Simon A., Croquette V., Bensimon A. (1995) Stretching DNA with a Receding Meniscus: Experiments and Models Phys. Rev. Lett. 74, 4754–4757. F. Allemand, Thèse de Doctorat, Laboratoire de Physique Statistique (Paris VI, Paris VII) (1997). 28. Bensimon A., Simon A., Chiffaudel A., Croquette V., Heslot F., Bensimon D. (1994) Alignment and sensitive detection of DNA by a moving interface Science 265, 2096–2098.

59

29. M. Pourbaix, in Atlas d’Équilibres Électrochimiques, GAUTHIER VILLARD, France 1963. 30. Milic N.B., Bugarcic Z. (1984) Hydrolysis of the palladium(II) ion in a sodium chloride medium Transition Met. Chem. 9, 173–176. 31. Hérard C., Bowen P., Lemaître J., Dutta J. (1995) Chemical synthesis and characterization of nano-crystalline palladium oxide NanoStructured Materials 6, 313–316. 32. Belyakova L.A., Varvarin A.M. (1999) Surfaces properties of silica gels modified with hydrophobic groups Colloids and Surfaces 154, 285–294.

Chapter 5 A Method to Map Spatiotemporal pH Changes Inside Living Cells Using a pH-Triggered DNA Nanoswitch Souvik Modi and Yamuna Krishnan Abstract A few cellular compartments maintain acidic environments in their interiors that are crucial for their proper function. Alteration in steady state organelle pH is closely linked to several diseases. Although a few probes exist to measure pH of cell compartments, each has several associated limitations. We present a high-performance pH sensor, a DNA nanoswitch, a convenient method to map spatiotemporal pH changes in endocytic pathways. DNA has been used to make a variety of nanoswitches in vitro. However, the present DNA nanoswitch functions as a pH sensing device equally efficiently intracellularly as it does in vitro. This DNA nanoswitch functions as a FRET-based pH sensor in the pH regime of 5.5–7, with high dynamic range between pH 5.8 and 7. It is efficiently engulfed by Drosophila hemocytes through endocytosis and can be used to measure the acidity of the endocytic vesicles that it marks during their maturation till their lysosomal stage. Key words: Nanomachine, I-motif, FRET, pH sensor, Endocytosis

1. Introduction Eukaryotic cell function is tuned by an orchestra of compartments involved in the uptake and secretion of various biomolecules that are networked with each other via a series of controlled fusion events between compartmental membranes (1). One of the many determinants is believed to be the luminal acidity of these compartments (2) which is a balance between proton leakage, ion permeability of the compartment membranes, and various ATP-dependent proton pumps (3). The maintenance of acidity levels is a key feature in the biogenesis of the secretory granules, protein sorting, and transport along both biosynthetic

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_5, © Springer Science+Business Media, LLC 2011

61

62

Modi and Krishnan

and endocytic pathways. For example, the luminal pH of the secretory pathway plays a crucial role in the glycosylation and sorting of proteins and lipids to plasma membranes (4). It is also believed that several viruses and pathogens cleverly manipulate luminal acidity to mature and infect cells (5–7). Defective acidification also results in disease, given its impact on protein sorting. For example, genetic diseases, such as cystic fibrosis (CF) and Dent’s disease, have been known to arise mainly due to dysfunctional chloride channels and improper sodium conductance which perturb organelle acidification (8, 9). pH changes along different endocytic/secretory pathways have been measured using various probes such as small molecules, weak bases, and genetically engineered GFP variants such as pHluorin (4). Some of these methods utilize the permeability of compartments to weak bases due to the acidic nature of their lumen (10). Others use de novo esterase activity of enzymes to generate pH-responsive fluorophores inside various organelles (11). Green fluorescent proteins have proved to be a popular method to measure pH as it enables one to genetically encode a pH probe that marks a specific cellular compartment or pathway (12). However, one of the simplest and widely used methods to measure organelle pH is the targeted delivery of pH-sensitive probes conjugated to macromolecules that are internalized by fluid phase or via a specific receptor to various endocytic compartments (13, 14). However, each of these methods have accompanying limitations. For example, the popularly used fluorescent probe FITC photobleaches easily and its fluorescence progressively decreases with increasing acidity. Ratiometric fluorescent probes such as BCECF and Carboxy-SNARF-1, besides undergoing intracellular hydrolysis of their ester bonds, prove restricting due to high pKa and low quantum yield, respectively. The obvious advantages of a genetically encoded sensors, such as ratiometric GFP, are target-specific expression in a desired location and their non-perturbative nature. However, ratiometric GFPs suffer from the limitation that their fluorescence is quenched at acidic pH, thus making it hard to visualize below pH 6.0. Furthermore, given that the probing wavelength is fixed at the wavelength of GFP, this restricts its use to mark various pathways simultaneously in the background of several interesting GFP mutants. The DNA nanomachine that we describe as an intracellular pH sensor is called the I-switch. The advantages of the I-switch are that it is bright, photostable, and easy to use. Since it is a FRET-based pH sensor, the I-switch is not limited by fluorophore wavelengths as several FRET pairs can be used (15). It provides the highest dynamic range from pH 5.5 to 7, making it a versatile probe to assay a wide range of biological processes associated with pH change.

A Method to Map Spatiotemporal pH Changes Inside Living Cells

63

2. Materials 2.1. Oligonucleotide Samples Preparation

1. Prepare 200 mM stocks in Milli-Q (MQ) water (Millipore) of all oligonucleotides in Table 1 (oligonucleotides are obtained from MWG Biotech, Germany, or IBA GmbH, Germany, or other quality suppliers). These should be HPLC-purified and lyophilized. Store aliquoted oligonucleotides at −20°C. 2. Ethanol, absolute. 3. 3.0 M Potassium chloride solution: 2.23 g KCl dissolved in 10 ml MQ water. 4. Phosphate buffers: 100 mM (5×) KH2PO4: 1.36 g KH2PO4 dissolved in 100 ml MQ water. 100 mM (5×) K2HPO4: 1.74 g K2HPO4 dissolved in 100  ml MQ water. 100  mM H3PO4: 0.98 g H3PO4 adjusted to 100 ml. 5. A series of buffers with different pH must be prepared after diluting each potassium phosphate solution to 20 mM. Add 20 mM KH2PO4 and 20 mM K2HPO4 in varying ratio for pH ranges 5–7.3, whereas for pH 4–5, the pH of 20 mM KH2PO4 is adjusted by addition of 20  mM H3PO4. Buffers are prepared in increments of 0.2 pH units. 6. Acetate buffer: 100 mM potassium acetate: 0.981 g dissolved in 100 ml water and titrated to pH 4.5 with acetic acid. 7. Phosphate buffer saline (1×): For 1 L, dissolve 8 g of NaCl, 0.2 g of KCl, 1.44 g of Na2HPO4, and 0.24 g of KH2PO4; adjust pH to 7.4; and bring to 1 L with deionized water. 8. 50× Tris–acetate buffer: 42  g of Tris base, 57.1  ml glacial acetic acid, and 18.6 g of EDTA, adjusted to 1 L with deionized water (pH 8.3). 9. Heat block and water bath (Neo Labs).

Table 1 Oligonucleotide sequences and derivatizations for the I-switch Name

Sequence

O1

5¢-CCCCAACCCCAATACATTTTACGCCTGGTGCC-3¢

O2

5¢-CCGACCGCAGGATCCTATAAAACCCCAACCCC-3¢

O3

5¢-TTATAGGATCCTGCGGTCGGAGGCACCAGGCGTAAAATGTA-3¢

O-488

5¢-Alexa-488-CCCCAACCCCAATACATTTTACGCCTGGTGCC-3¢

O-TMR

5¢-Bodipy-TMR-CCCCAACCCCAATACATTTTACGCCTGGTGCC-3¢

O-647

5¢-CCGACCGCAGGATCCTATAAAACCCCAACCCC-Alexa-647-3¢

O3-Bio

5¢-Biotin-AATTATAGGATCCTGCGGTCGGAGGCACCAGGCGTAAAATGTA-3¢

64

Modi and Krishnan

2.2. Circular Dichroism

1. JASCO J-815 spectrophotometer (Jasco, Japan) equipped with a temperature controller, an in-built stop-flow device (Biologic), and a Mercury–Xe lamp as the light source. 2. Acetate buffer: see Subheading 2.1 above. 3. Phosphate buffer, pH 4.5: see Subheading 2.1 above. 4. Phosphate buffer, pH 7.3: see Subheading 2.1 above. 5. 100  mM phosphate buffer: 100  mM (potassium hydrogen phosphate/potassium dihydrogen phosphate) buffer at pH 7.3. 6. 3 M potassium chloride: see Subheading 2.1 above. 7. 1 cm quartz cuvette (Varian).

2.3. Steady State and Ratiometric Fluorescence Measurements

1. JASCO J-815 spectrophotometer (Jasco) equipped with a monochromator (FMO 427S/14) and temperature controller (CDS426S/15). 2. 5 mM Fluorescently labeled I-switch: The I-switch is prepared according to protocol 3.1 in phosphate buffer pH 5.5 supplemented with 100 mM KCl, employing fluorescently labeled oligonucleotides: O1 labeled with Alexa 488 and BodipyTMR (O1-488/TMR), and O2 labeled with Alexa 647 (O2-647) are used in this study (see Note 1). 3. 1 cm/0.5 cm dual Quartz cuvette (Hellma). 4. pH cycling is performed in Fluorolog-Spex spectrophoto­ meter (Horiba, Jobin-Yvon). 5. 1 N HCl solution: 10 ml of concentrated HCl adjudsted to 120 ml in Milli-Q ultrapure water. 6. 1 N KOH solution: 5.6 g of KOH dissolved in 100 ml Milli-Q ultrapure water. 7. 3 M KCl: see Subheading 2.1 above.

2.4. Time-Resolved Fluorescence Measurement

1. 5  mM I-switch sample containing donor label only (i.e., O-488/O2/O3) and both donor and acceptor labels (i.e., O-488/O647/O3) is diluted to 1 mM in 20 mM phosphate buffer of pH 5, 6, 6.5, and 7.3. 2. Multiphoton excitations are performed on a Zeiss LSM 510 Meta microscope (Carl Zeiss, http://www.zeiss.com) with 63´, 1.4 numerical aperture (NA) objective. 3. Femtosecond-pulsed Tsunami Titanium:Sapphire tunable pulsed laser (Spectra physics). 4. Hamamatsu R3809U multichannel plate photomultiplier tubes (PMTs; Hamamatsu Photonics, http://www.­ hamamatsu.com). 5. Becker & Hickl 830 card for Time-Correlated Single Photon Counting (TCSPC) (Becker and Hickl, http://www.beckerhickl).

A Method to Map Spatiotemporal pH Changes Inside Living Cells

65

6. Glass coverslips (gold seal, no: 1). 7. 10–16-nm gold nanoparticles (Ted Pella, Inc., Redding, CA) dried on a coverslip as a second harmonic generator. 2.5. Protein Conjugation, Agarose Gel Electrophoresis, and Size Exclusion Chromatography

1. PBS (1×): see Subheading 2.1 above. 2. 0.5 mg Streptavidin (Sigma) dissolved in 1 ml× PBS, pH 7.4. The solution can be stored at 4°C. 3. Biotinylated human holo transferrin (Sigma): 1 mg biotinylated human holo transferrin (Sigma) dissolved in 1 ml PBS and stored at 4°C. 4. Human holo transferrin: 1 mg human holo transferrin (Sigma) dissolved in 1 ml PBS and stored at 4°C. 5. Biocytin solution: 3  mg biocytin (Sigma) dissolved in 1  ml Milli-Q water and kept at 4°C. 6. Agarose powder (Sigma). 7. Ethidium bromide stock: 10 mg ethidium bromide in 1 ml Milli-Q water. 8. 50× TAE: see Subheading 2.1 above. 9. Gel loading dye (10×), 100 mg Orange G (Sigma) dissolved in 10 ml Milli-Q water, and 15 ml glycerol (Merck) in 15 ml Milli-Q water. Both solutions were mixed and adjusted to 50 ml using Milli-Q water. 10. Shimadzu HPLC system composed of a temperature controller, a photodiode array detector for absorbance, a fluorescence detector, an autosampler unit, and an auto-injection unit (Shimadzu, Japan). 11. SEC column BioSep-SEC-S3000 (Phenomenex), with 5 mm beads with dimensions of 300 × 4.6 mm.

2.6. Cell Culture and Labeling

1. Drosophila stocks. These are grown at 20°C in cornmeal agar bottles. After transferring parent flies to a new bottle, it is maintained at 20–25°C. After 8–9 days, third instar drosophila larvae are collected by picking healthy larvae walking near the sides and necks of the bottles using a drawing brush (as 1st and 2nd instar larvae do not leave the media). 2. Complete insect medium: Schneider’s insect medium supplemented with 10% nonheat-inactivated FBS (Gibco) and supplemented with 1 mg/mL of bovine pancreatic insulin (Sigma), 150 mg/mL of penicillin (Sigma), 250 mg/mL of streptomycin (Sigma), and 750 mg/mL of glutamine (Sigma). 3. 35 mm coverslip bottom dishes: Obtained by drilling 35 mm cell culture dishes (Nunc) and fixing coverslip (Gold seal no:1) by glue. 4. Medium 1 (1×), 150 mM NaCl, 5 mM KCl, 1 mM CaCl2, 1 mM MgCl2, 20 mM HEPES. NaCl 87 mg, KCl 3.7 mg,

66

Modi and Krishnan

CaCl2 1.10 mg, MgCl2 2.03 mg, and HEPES 47.6 mg were dissolved in 10  ml water and adjusted to pH 7.0. Filtersterilize using 0.22 mm membrane filter (Millipore). For labeling cells with I-switch and other endocytic probes, 1 mg/mL bovine serum albumin (BSA) and 2  mg/mL d-glucose are added to Medium 1 prior to use. 5. FITC-dextran solution (10×): 10 mg FITC-dextran (Sigma) dissolved in 1 ml PBS. 6. LysoTracker Red™ (1  mM solution in DMSO, Molecular probes, Invitrogen). 7. Paraformaldehyde (PFA, 20% (w/v) stock) is prepared by dissolving it in water, heating it to 60°C. It is then aliquoted, and quickly frozen to −20°C. Each vial is thawed prior to use by heating to 60°C. 8. Clamping buffers: 120 mM KCl, 5 mM NaCl, 1 mM CaCl2, 1  mM MgCl2, 20  mM HEPES, and 10  mM nigericin. KCl 89.472 mg, NaCl 2.92 mg, CaCl2 1.10 mg, MgCl2 2.03 mg, and HEPES 47.6 mg are dissolved in 10 ml and adjusted to pH 5–7. To this, 10 ml of 10 mM nigericin (Sigma) was added. 2.7. Fluorescence Microscopy and Image Analysis

1. Nikon inverted microscope (TE 300 Nikon, Japan). 2. Filters: 520/30 and 665 LP (long pass) (Chroma, USA). 3. Mercury arc illuminator (Nikon, Japan) and cooled CCD camera (Andor, USA) controlled by Metamorph software (Universal Imaging, PA). 4. Olympus Fluoview 1000 Confocal microscope (Olympus, Japan). 5. Argon ion laser (Spectra physics) for 488 nm excitation and He–Ne laser (Spectra physics) for 543 nm excitation with a set of dichroics, excitation, and emission filters suitable for each fluorophore provided by Olympus fluoview software. 6. Image analysis software: Image J (developed at NIH, freely available from website: http://rsbweb.nih.gov/ij/) and Metamorph software (Universal Imaging, PA).

2.8. Competition Study with mBSA

1. Maleic anhydride (Sigma). 2. BSA solution: 6 mg/ml solution in 0.1 M sodium carbonatebicarbonate buffer (pH 9).

3. Methods DNA nanomachines are structural scaffolds that change their states in response to an external trigger. We have created a DNA nanomachine called the I-switch that functions autonomously as

67

A Method to Map Spatiotemporal pH Changes Inside Living Cells

a pH sensor both in vitro and in vivo (15). We used a set of DNA strands complementary to each other that resulted in C-rich i-motif forming overhangs leaving a one-base gap. This is the linearly extended form adopted at pH 7.4 and is referred to as the open state (Fig.  1a). The one-base gap between the duplexed regions acts as a fulcrum between the duplex arms that help the scaffold to bend to in order to form an I-motif that results in the formation of the buckled, closed state at acidic pH (Fig. 1a). To characterize the two different states upon pH change, we first perform circular dichroism (Fig. 1b) and fluorescence resonance energy transfer (FRET) experiments using bulk and fluorescence lifetime measurements (Fig. 1d). The fluorescence properties of

a

pH 5 pH 7.3 Difference (pH 5 - pH 7.3)

b 8

OH−

Salt

pH 7.3 open

CD (mdeg)

Salt

H+

pH 5 closed

4

0

220

240

260

280

300

320

−4

Wavelength (nm) d Normalized counts

c

Normalized D/A

1.2

In vitro Intra cellular

0.8

0.4

1

pH 5.0 pH 6 .0 pH 6.5 pH 7.3

0.5 0 0 0

0.0

0 4.0

4.8

5.6

6.4

7.2

0

Time (ns)

2

pH Fig. 1. In vitro characterization of the I-switch. (a) Schematic of working principle of the I-switch in the “open” state (low FRET) at high pH and in the “closed” state (high FRET) at low pH. (b) Characterization of I-switch using CD spectroscopy. CD spectra of I-switch at pH 5 (Square) and pH 7.3 (Circle), and difference spectra (pH 5–7.3, triangle) are presented. (c) In vitro and intracellular calibration curve of I-switch. The normalized donor/acceptor (D/A) intensity (Alexa 488/Alexa 647) ratios, plotted as a function of pH, yield the calibration curve. Intracellular curve (Gray line) is overlayed on the in vitro pH profile (Black line). Error bar : Mean of two independent experiments ± standard error of the mean (SEM). (d) Time-resolved fluorescence spectra of I-switch assemblies at different pH. Lifetime decay of dual-labeled (Alexa 488/647) I-switch at indicated pH is measured using multiphoton excitation. 5 mM I-switch is diluted with phosphate buffer of appropriate pH containing 100 mM KCl before acquisition.

68

Modi and Krishnan

the I-switch are measured as a function of pH to characterize the pH sensitivity of the I-switch and its associated working regime. Changing the pH changes the ratio between the open and the closed states, which in turn changes the ratios of donor and acceptor intensities (D/A). A plot of D/A vs. pH results in a sigmoidal increase from 5.0 to 6.8, giving a standard in vitro pH calibration curve for the I-switch (Fig.  1c). Doubly labeled I-switch showed a ~5.5-fold increase in  vitro with an unprecedented dynamic range in the pH regime 6–7. As the trafficking route from early endosome to lysosome involves a characteristic pH change from 6.2 to 5, this leaves the I-switch well placed to sense the pH along this pathway. As proof of concept, we mark the anionic ligand-binding receptor (ALBR) pathway using I-switch and measure spatial and temporal pH across this pathway involving early endosome, late endosome, and finally lysosome (Fig.  2c). To make the I-switch suitable to measure pH in the

FITC-Dextran

I-switch

a

c

6.0

pH

5.8

Mean pH

5.6 5.4 5.2 5.0

d

b

4.8 4.4

D/A

4.0

Mean D/A 3.6 3.2

Lysotracker Red

I-switch

2.8

0

20

40

60 80 Time (min)

100 120

Fig. 2. Colocalization of the I-switch with endocytic vesicle marker. (a) FITC-dextran (left panel ). Cells are pulsed with a solution of FITC-dextran and Bodipy–TMR labeled I-switch (right panel ), chased, and imaged under a confocal microscope. (b). Colocalization of the I-switch (below right panel ) with lysosomal marker LysoTracker Red (below left panel ). (c). Spatiotemporal mapping of pH changes during endocytosis using the I-switch in living cells. After obtaining D/A values for a collection of endosomes, mean D/A value at indicated chase times was measured and converted to their respective pH values after comparing with intracellular calibration curve. (d) Real-time monitoring of the acidification during endocytosis. Cells are pulsed with I-switch, washed, and imaged over 2 h in wide-field microscope. Five distinct cells are imaged at each time point, and the mean D/A of two experiments ± SEM is plotted with time. Scale Bar : 5 mm. Insets show magnification of a represented vesicle.

A Method to Map Spatiotemporal pH Changes Inside Living Cells

69

Fig. 3. Quantification of I-switch internalization by Drosophila hemocytes. (a) Competition assay in the presence and absence of excess of competitor maleylated BSA (mBSA, 800 mg/ml) which saturates ALB receptors. I-switch internalization is quantified in terms of the total fluorescence intensity (at Alexa 647 channel) of cells pulsed with I-switch in the presence and absence of excess unlabeled mBSA. The contribution from autofluorescence is also shown. (b), Protein conjugation with the I-switch. Biotinylated I-switch (IB) is tagged to biotinylated transferrin (TfB) through streptavidin (SA), thus labeling the transferrin with the I-switch. Inset : 3% agarose–TAE gel to characterize the I-switch–SA conjugate (IB–SA) and the I-switch–SA–transferrin conjugate (IB–SA–TfB). (c). Size exclusion chromatography (SEC) to characterize complex stoichiometry. SEC chromatogram of (1) streptavidin (SA) and (2) I-switch (IB) shows single peaks, whereas 1:1 complex of SA and IB shows two peaks corresponding to IB–SA and IB, respectively (3). 50 mL of 1 mM I-switch or complex is injected and separated using an isocratic flow of PBS with a flow rate of 0.5 mL/min. Absorbance at 260 nm is followed over time. Chromatogram (4) showing free TfB, and when TfB is added to IB–SA in a 2:1 ratio, a new peak arises due to the formation of IB–SA–TfB (5). Transferrin has 475-nm absorbance peaks and it is followed to probe number of bound transferrins per IB–SA complex. Scale bar : 5 mm.

environment of a given protein of interest, we demonstrate a convenient method to conjugate it to proteins using biotin– streptavidin interactions (Fig. 3b). We demonstrate this on transferrin as a model protein to mark the receptor-mediated endocytic pathway of transferrin and also the recycling endosome. This method can be extended to conjugation to any biotinylated protein– ligand that may be internalized via a specific receptor in order to measure the pH environment around the ligand.

70

Modi and Krishnan

3.1. Oligonucleotide Sample Preparation

1. Prepare 100 mL solutions of the I-switch at 5 mM in a 500 mL Eppendorf tube by mixing 2.5 mL of each O1, O2, and O3 200  mM stocks, and adding 3.3  mL of potassium chloride solution. Bring the volume to 100  mL by adding 89  mL of 20 mM phosphate buffer of the desired pH and briefly vortex to mix all the solution. 2. Heat the tube at 90°C for 5 min in a heat block; cool to the room temperature over 3  h at a rate of 5°C/15  min. After attaining room temperature, equilibrate the sample tube at 4°C for 8–24 h. 3. Prior to use, the above tube of I-switch stock solution is subjected to a brief spin in a tabletop centrifuge, followed by mixing in a vortex mixer.

3.2. Circular Dichroism Spectroscopy

1. Dilute 110 mL of 5 mM I-switch stock solution to 340 mL by adding potassium phosphate buffer of (1) pH 4.5 and (2) pH 7.3. To this, add 11 mL of 3 M KCl and mix well by vortexing. The sample is equilibrated for 30 min prior to acquisition of spectra. 2. Of the solution, 350  mL is transferred into a 1  cm quartz cuvette, and CD spectra are recorded from 320 nm to 200 nm with 1 nm/s bandwidth and 0.25 s response time. An average of five successive scans is acquired. 3. For thermal melting experiments, samples are prepared similarly. For melting at pH 4.5, 5 mM I-switch samples are prepared in acetate buffer, whereas for melting at pH 7.3, phosphate buffer is used. 4. CD spectra at different pH are acquired by monitoring a specific wavelength at different temperatures or using a wavelength–temperature scan option available in the application software (this software also records CD spectra at each temperature, which enables one to monitor any desired wavelength). Wavelengths of 293 nm (for pH 4.5) and 278 nm (for pH 7.3) are monitored as a function of temperature from 20 to 80°C at a rate of 1°C/min.

3.3. Steady State and Ratiometric Fluorescence Measurements

1. Prepare fluorescence specimens at different pH values from 5  mM fluorescently labeled I-switch solutions by diluting samples to 80 nM in the desired phosphate buffer. 5 mL of stock is diluted with 295 mL of 20 mM phosphate buffer of pH ranging from 4 to 7.3. All samples are vortexed and equilibrated for 30 min at RT. 2. The samples are excited at 488 nm and emission is collected between 505 and 730  nm with a response time of 64  ms, bandwidth of 5 nm (for excitation) and 10 nm (for emission), and spectral scan speed of 1 nm/s (see Note 2).

A Method to Map Spatiotemporal pH Changes Inside Living Cells

71

3. FRET efficiencies are calculated using the formula E = 1−IDA/ID = 1/[1 + (R/Ro)6], where ID is the intensity of the donor in the absence of the acceptor, IDA is the intensity of the donor (D) in the presence of the acceptor (A), Ro is the Förster distance, and R is the interfluorophore distance. Ro for Alexa 488 and Alexa 647 is taken as 48–50 Å (16). 4. For in  vitro calibration curve, fluorescent intensities of the different pH buffers are recorded at 520 nm (D) and 669 nm (A), and then D is divided by A to generate an in vitro calibration curve (see Note 3), as seen in Fig. 1c. 3.4. Time-Resolved Fluorescence Measurement

1. Of the labeled I-switch, 10–20  mL is placed on a coverslip such that it forms a droplet, and the coverslip is mounted on microscope stage. 2. Fluorophores are excited by a two-photon excitation at 720  nm with a repetition rate of the pulsed laser fixed at 80.09  MHz (12  ns). Fluorescence decay is followed over 12 ns using the TCSPC card. 3. The instrument response function (IRF) is calculated similarly using gold nanoparticles dried on a coverslip. 4. The experimentally measured fluorescence decay is deconvoluted using IRF with the intensity decay function. Average lifetimes at different pH from both donor only and dually labeled samples obtained from the intensity decay data are fit to the appropriate equations by an iterative reconvolution procedure using a Levenberg–Marquardt minimization algorithm (17) developed in-house. A typical plot corresponding to in vitro characterization is shown in Fig. 1.

3.5. Protein Conjugation, Agarose Gel Electrophoresis, and Size Exclusion Chromatography

1. Prepare 5 mL of biotinylated I-switch (IB, i.e., 25 pmoles of O488/O647/O3-Bio) and mix with 5  mL of PBS in an Eppendorf tube. 2. To 5  mL of streptavidin, in another tube, the solution of I-switch in PBS is slowly added with constant agitation. After complete addition of I-switch, the resultant mixture is incubated at RT for 1  h. This mixture yields an IB–Streptavidin conjugate (IB–SA). 3. 5  mL of biotinylated human holo transferrin (50  pmole) is mixed with 10 mL of PBS and then slowly added dropwise to the tube containing IB–SA with constant agitation. After a brief vortexing, this mixture is then incubated for an additional 1  h. This yields a ternary conjugate of biotinylated transferrin linked to the biotinylated I-switch via streptavidin (IB–SA–TfB). See Fig. 3b for representative data. 4. 2 mL of biocytin (1,000 pmole) is then quickly added to the IB–SA–TfB mixture in order to quench unbound sites

72

Modi and Krishnan

on SA to minimize aggregation, and the sample is quickly transferred to 4°C or used immediately. 5. Prepare an agarose gel: 0.9 g of agarose is weighed in a 250 ml conical flask and mixed with 29.4  ml of Milli-Q water and 600  mL of 50× TAE buffer. This flask is then heated in a microwave oven for 1.5–2 min (to melt all the agarose and obtain a homogeneous solution). The contents are cooled to RT, 5 mL of EtBr stock is added, gently mixed, and poured in a gel tray containing a comb. 6. After the gel solidifies (~15 min), the tray containing gel is transferred into an electrophoresis unit (we use a Bio-Rad) containing 300 mL 1× TAE buffer and pre-run for 10 min at 50 V. 7. 10 mL of samples containing (1) IB, (2) IB–SA, and (3) IB–SA– TfB are each mixed with 5 mL of gel loading dye, loaded on separate wells in the gel, and run at 100 V for 2 h. 8. The gel is stopped after the loading dye reaches 75% of the gel length (~2 h) and imaged in a transilluminator/gel documentation system. 9. For characterization by size exclusion chromatography (SEC), IB, IB–SA, IB–SA–TfB, SA, and TfB are prepared as described earlier, and diluted to 1 mM concentration with PBS prior to injection. 10. HPLC system and SEC column are equilibrated in PBS for an hour before sample injection. 500  mL of 1  mM sample is loaded into a vial and kept inside the injection unit. All samples are separated using an isocratic flow (0.5  mL/min) of PBS (1×). 11. Each time ~50–100 mL of sample is injected into the column and absorbance at 260  nm (for DNA and proteins) and 475 nm (for only transferrin (TfB) and IB–SA–TfB) is followed for over 30 min. 12. Labeling cells using this IB–SA–TfB construct is described in Subheading 3.6 (see Fig. 3c and Note 4). 3.6. Cell Culture and Labeling

1. Third instar Drosophila larvae are surface sterilized using ­double-distilled water and bleach solution (50%), and transferred into a dissection dish containing complete insect medium. The larvae are then stretched by fixing them between two needles, one in the head and one in the tail. 2. Hemolymph is isolated by puncturing the larvae using dissection forceps to release the hemolymph into the complete insect medium. Three to five larvae provide enough cells to plate two dishes.

A Method to Map Spatiotemporal pH Changes Inside Living Cells

73

3. The medium containing hemocytes is then plated onto 35 mm coverslip dishes. These dishes are then transferred into a 20°C incubator and kept for 1.5 h before endocytic assay. 4. The cells are washed with Medium 1 prior to labeling. 5 mL of the I-switch sample is diluted to 300 mL in Medium 1 and then added to the coverslip containing hemocytes (this is referred to as “pulsing”). The cells are incubated for 5–10 min and then washed 3–4 times with Medium 1 and then chased for an additional 5–120 min. For a chase longer than 5 min, Medium 1 is replaced by complete insect medium and transferred to a 20°C incubator. For 5  min chases, the cells are washed and then imaged in Medium 1 (see Note 5). 5. For colocalization studies with the I-switch and FITCdextran, cells are co-pulsed with 300 mL of I-switch (O-TMR/ O2/O3) and FITC-dextran (1 mg/mL) for 15 min, washed five times, and chased for 15–30 min. The cells are then briefly fixed with 200  mL of 2.5% PFA for 20  min in Medium 1, quickly washed three times with Medium 1, and imaged (see Fig. 2a and Note 6). 6. For colocalization studies with LysoTracker Red™, 300 mL of 80 nM dual-labeled I-switch is co-pulsed with 1:1,000 diluted LysoTracker Red™ for 5  min, chased for 2  h in complete medium, washed three times with Medium 1 as described above, and imaged live (see Fig. 2b and Note 7). 7. For pH measurement experiments, after pulsing with duallabeled I-switch, cells are chased for (1) 5 min, (2) 60 min, and (3) 120 min. Each set of cells is then imaged live, acquiring three images for each set of cells: (1) by exciting donor (488 nm) (2) by exciting acceptor (630 nm), and (3) by exciting at 488 nm and acquiring at 647 nm (the FRET image). 8. The intracellular pH standard curve is obtained by adding 300 mL of 80 nM I-switch, incubating for 30 min (for development of a good enough signal), washing 3–4 times with Medium 1, and chasing for 5 min in the same medium. 9. The cells are then briefly fixed with 200 mL of 2.5% PFA for 1 min, quickly washed three times, and retained in Medium 1. 10. Add 1 mL of clamping buffer containing 10 mM of nigericin at the desired pH to the previously fixed cells, incubate for 15–30 min, and then image (see Note 8). 3.7. Fluorescence Microscopy and Image Analysis

1. All wide-field images are collected using a Nikon inverted microscope equipped with 60´, 1.4 NA objective. Cells are located and focused after acquiring a phase contrast image. Then fluorescence images of the cells are obtained by exciting Alexa 488 (3-s exposure time) using the 520 ± 30-nm emission

74

Modi and Krishnan

filter (in case of the donor image). The cells are then re-excited for 3  s at 488  nm and emission of the acceptor is imaged using the 665-nm long-pass emission filter (this is the FRET image). A third image is obtained by exciting the cells at 630 nm and imaged using 665-nm long-pass filter (acceptor image). 2. Confocal imaging is carried out on an Olympus Fluoview 1000 Confocal microscope using a similar strategy that employs an Argon ion laser for 488 nm excitation and He–Ne laser for 543 nm excitation. 3. Autofluorescence is measured on unlabeled cells. All the images (donor, acceptor, and FRET) are then background subtracted taking mean intensity over a large cell-free area. 4. Donor and acceptor images are colocalized using ImageJ, and endosomes showing colocalization are selected for further analysis. Each endosome in the donor and FRET image is selected by the ROI plugin in ImageJ, and total intensity and mean intensity in each endosome are measured and recorded in an MS Excel file. 5. A ratio of donor to acceptor intensities (D/A) is obtained from these values by dividing the intensity of a given endosome in the donor channel with the corresponding intensity in the FRET channel. Thus for a given time point, or a given pH-clamping experiment, the D/A of each endosome is collected and the mean D/A is determined for a collection of endosomes in a given condition. The mean D/A for 2–3 experiments is collected and plotted along with standard error of the mean (SEM) for a given experimental condition such as chase time, or pH clamp (see Fig. 2c for an example). 6. Real-time pH measurements are obtained by imaging cells in different areas on a single coverslip. After labeling cells for 5 min in Medium 1, the cells are washed three times with Medium 1 and then directly mounted on the microscope stage. 7. Imaging is performed as described earlier over a total of 2 h with time intervals of 5 min each. Imaging is done such that each time frame has five images covering five different places in the coverslip. 8. From those five fields of view, 40 endosomes at each time point are quantified and represented as earlier. An example of this experiment is shown in Fig. 2d. 3.8. Competition Study with mBSA

1. mBSA is prepared by labeling BSA with maleic anhydride: mix 3.3 ml of BSA solution. 20 mg of BSA (at 6 mg/ml) dissolved in 3.3  mL of 0.1  M sodium carbonate-bicarbonate buffer (pH 9) was mixed with 50 mg of maleic anhydride and stirred at room temperature. To this solution, solid sodium

A Method to Map Spatiotemporal pH Changes Inside Living Cells

75

carbonate was added to maintain the optimal pH at 9. After 30  min, the reaction mixture was dialyzed against 1× PBS overnight to remove unreacted mBSA. mBSA is concentrated and stored as 10 mg/mL stock in aliquots at −20°C. 2. Three different dishes of cells are prepared. One dish is incubated with mBSA and I-switch (referred to as “+mBSA”), one dish is incubated with I-switch alone (−mBSA), and a third dish is used to measure autofluorescence. 3. For the competition assay with mBSA, cells are incubated for 5  min with 300  mL of 0.8  mg/mL mBSA and chased for 5  min to allow all the receptors to internalize. Then these cells are again pulsed with 300 mL of 0.8 mg/mL mBSA containing 80 nM of dual-labeled I-switch (O-488/O-647/O3) for 5  min, chased for 5  min, and washed three times with Medium 1. 4. The cells are then fixed using 2.5% PFA in Medium 1 for 20 min, washed three times in Medium 1, and then imaged. 5. Imaging is performed in a wide-field microscope after exciting Alexa 647 at 630 nm with a 3-s exposure. 6. Images are background subtracted as described earlier and cell boundary is determined from the phase contrast image and stored as ROI in ImageJ. These ROIs are recalled after opening the fluorescence image and the total cell intensity is measured in all three dishes. After obtaining the mean intensity of all three dishes, it is normalized with respect to the mean intensity in the cells pulsed with I-switch alone (−mBSA) and is presented as the fraction of I-switch internalized (see Fig. 3a and Note 9).

4. Notes 1. It is critical to form samples with 20 mM phosphate buffer at pH 5.5 in order to reproducibly change the pH of the solution. 2. Note that at acidic pH, emission maxima of Alexa 488 is shifted by 2–4 nm, but this discrepancy is eliminated by simply plotting intensity at 520 nm as a function of pH. 3. The calibration curve is then normalized by dividing the D/A values at every pH by the D/A value at pH 7.3. This normalization accounts for errors due to differential intensity of excitation source across different days. 4. To label mammalian cells, it is necessary to deplete transferrin present in the serum. This may be achieved by 30-min incubation at 37°C in a serum-free media prior to labeling.

76

Modi and Krishnan

5. Hemocytes are not well adherent, so washing and labeling must be done very gently. It is also recommended not to expose these cells to air for longer than necessary. Washing and fixing should be performed as quickly as possible. 6. Fixing is done in Medium 1 lacking BSA and glucose. 7. For colocalization studies with LysoTracker Red™, it is necessary to image the cells live as fixing the cells abolishes the pH gradient across endosomes/lysosomes, leading to greatly diminished labeling by LysoTracker Red™. 8. When calibration curves with I-switch are prepared, cells are imaged in clamping buffers of the desired pH containing nigericin throughout the course of experiments. 9. Mean intensity of cells in competition experiment is normalized with respect to mean intensity of untreated cells (no mBSA). This kind of normalization accounts for day-to-day experimental variations resulting from differential labeling between experiments.

Acknowledgments We thank S. Mayor and M.G. Swetha for inputs on endocytosis assay, and the Nanoscience and Technology Initiative of the Department of Science and Technology, Govt of India (GoI). S.M. thanks the Council of Scientific and Industrial Research, GoI, for funding. Y.K. thanks the Department of Biotechnology, GoI, for the Innovative Young Biotechnologist Award and NCBS for a start-up grant. References 1. Mukherjee, S., Ghosh, R. N., Maxfield, F. R. (1997) Endocytosis. Physiological Reviews 77, 759–803. 2. McCoy, K.L. (1990) Contribution of endosomal acidification to antigen processing. Semin. Immunol 2, 239–246. 3. Weisz, O. A. (2003) Organelle Acidification and Disease. Traffic 4, 57–64. 4. Paroutis, P., Touret, N., Grinstein, S. (2004) The pH of the Secretory Pathway: Measurement, Determinants, and Regulation. Physiology 19, 207–215. 5. Palokangas, H., Metsikko, K., Vaananen, K. (1994) Active vacuolar H+-ATPase is required for both endocytic and exocytic processes during viral infection of BHK-21 cells. J. Biol. Chem. 269, 17577–17585.

6. Guinea, R., Carrasco, L. (1995) Requirement for vacuolar proton-ATPase activity during entry of influenza virus into cells. J. Virol 69, 2306–2312. 7. Yu, I. M., Zhang, W., Holdaway, H. A., Li, L., Kostyuchenko, V. A., Chipman, P. R., Kuhn, R. J., Rossmann, M. G., Chen, J. (2008) Structure of the immature dengue virus at low ph primes proteolytic maturation Science 319, 1834–1837. 8. Piwon, N., Gunther. W., Schwake, M., Bosl, M. R., Jentsch, T. J. (2000) ClC-5 Cl– channel disruption impairs endocytosis in a mouse model for Dent’s disease. Nature 408, 369–373. 9. Barasch, J., Kiss, B., Prince, A., Saiman, L., Gruenert, D., Al-Awqati, Q. (1991) Defective

A Method to Map Spatiotemporal pH Changes Inside Living Cells

10.

11.

12.

13.

acidification of intracellular organelles in cystic fibrosis. Nature 352, 70–73. Anderson, R. G., Pathak, R. K. (1985) Vesicles and cisternae in the trans Golgi apparatus of human fibroblasts are acidic compartments. Cell 40, 635–643. Wu, M. M., Llopis, J., Adams, S., McCaffery, J. M., Kulomaa, M. S., Machen, T. E., Moore, H. P., Tsien, R. Y. (2000) Organelle pH studies using targeted avidin and fluorescein-biotin. Chem. Biol. 7, 197–209. Miesenbock, G., De Angelis, D. A., Rothman, J. E. ( 1998) Visualizing secretion and synaptic transmission with pH-sensitive green fluorescent proteins Nature 394, 192–195. Ohkuma, S., Poole, B. (1978) Fluorescence probe measurement of the intralysosomal pH in living cells and the perturbation of pH by various agents. Proc. Natl. Acad. Sci. USA. 75, 3327–3331.

77

14. Sipe, D. M., Murphy, R. F. (1987) Highresolution kinetics of transferrin acidification in BALB/3 T3 cells exposed to pH 6 followed by temperature sensitive alkalinization during recycling. Proc. Natl. Acad. Sci. USA. 84, 7119–7123. 15. Modi, S., Swetha, M.G., Goswami, D., Gupta, G. D., Mayor, S., Krishnan, Y. (2009) A DNA nanomachine that maps spatial and temporal pH changes inside living cells. Nat Nanotech. 4, 325–330. 16. Majumdar, D. S., Smirnova, I., Kasho, V., Nir, E., Kong, X., Weiss, S., Kaback, H. R. (2007) Single-molecule FRET reveals sugarinduced conformational dynamics in LacY. Proc. Natl. Acad. Sci. USA. 104, 12640–12645. 17. Krishna, M. M., Srivastava, A., Periasamy, N. (2001) Rotational dynamics of surface probes in lipid vesicles. Biophys. Chem. 90, 123–133.

Chapter 6 Control of Helical Handedness in DNA and PNA Nanostructures Roberto Corradini, Tullia Tedeschi, Stefano Sforza, Mark M. Green, and Rosangela Marchelli Abstract Helical handedness and the twist and tilt parameters of the base pairs in duplex DNA can be affected by base sequence variation and change in environmental conditions as occurs in the transformation between right-handed B-DNA and left-handed Z-DNA. For duplexes of DNA with oligonucleotide analogs such as peptide nucleic acids (PNAs), less is known about the effects on structure such as the base pair twist and tilt parameters and handedness. However, in PNA:PNA duplexes, the absence of chiral information determining helical handedness allows the relationship between preferred helical handedness and structural design to be manipulated and, therefore, better understood. In this chapter, we report a protocol for switching between B- and Z-DNA:DNA duplexes, and the experimental procedures for obtaining right- or left-handed PNA:PNA duplexes. Key words: Helical handedness, Z-DNA, B-DNA, Chiral PNAs, PNA:PNA duplex

1. Introduction Helical handedness is a property that is essential in polymeric and supramolecular materials in which the effect of molecular chirality is “amplified” in macroscopic features (1–3). One interesting property of DNA as a nanomaterial is that a very large number of structural variants have been described, and the main DNA conformations (A, B, and Z, Fig. 1a) are sufficiently characterized to be used for designing nanostructures (4). The B-form of DNA is a right-handed helix with a pitch of 10–10.5 base pairs (bp), a rise between adjacent base pairs of 3.4 Å, and a twist angle of 36°; the full turn of the helix is 3.4  nm long, with a diameter of 1.84  nm, with the base pairs centered on the helical axis and

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_6, © Springer Science+Business Media, LLC 2011

79

80

Corradini et al.

a

Z-DNA

A-DNA

B-DNA

b

“Pro-Z” sequence

Mixed sequence

Mixed sequence

High Salt Additives (e.g Co(NH3)63+)

Negative supercoiling Antibodies/Proteins Z-DNA binding molecules B-Z junction

B-Z junction

B-DNA

Z-DNA

B-DNA

Unwinding of helix: 3.5 turns for d(CG)10 d(CG)10 Elongation of 0.3 Å/base pair Easier for long Z-tract and short B-tract

Fig. 1. (a) Different DNA conformations. (b) Schematic representation of a B–Z transition involving B–Z junctions.

­ erpendicular to it. The A form is also right handed, with a pitch p of 10.7  bp and a rise of 2.3 Å, and the full turn of the helix is 2.5 nm with a diameter of 2.6 nm. The change from B to A can be performed by changing the experimental conditions to low relative humidity and low salt concentration. Z-DNA occurs at high salt concentration with alternating purine–pyrimidine sequences, and is left handed, with alternating syn (for purine) and anti (for pyrimidine) base conformations and alternating sugar puckering (C3′ endo for purines and C2′ endo for pyrimidines). It has a helix pitch of 12  bp and a rise of 3.63–3.72 Å, with a twist of −60°. This helix has a path of 4.4–4.5 nm and a diameter of 1.6–1.9 nm. In DNA-based nanostructures, the structural parameters of DNA such as diameter, length, and pitch are very important for the design of 2D or 3D features (5). Indeed, the control of the B–Z transitions has been the key change in some simple “DNA nanomachines,” which can convert a chemical input into a mechanical motion (6), and also in the control of the topological handedness of DNA knots (7), a transition that can be followed by CD spectroscopy (Fig.  2). The overall handedness of DNA and DNA-like structures can be used as a tool for discriminating enantiomers in some applications (8), as a chiral microenvironment

81

Control of Helical Handedness in DNA and PNA Nanostructures

b

100 230 −100

250

270

290

310

330

De (M-1 x cm-1)

De (M-1 x cm-1)

300

350

λ(nm)

−300

−500

100 250

270

290

310

330

350

λ(nm)

−300

300

100

Poly d(me5CG)

1500

700

700

500 0.6

0

NaCl

MgCl2

250

200

g

Poly d(CG)

(∆ε–∆ε0) (M−1 cm−1)

2500

C half transition (mM)

C half transition (mM)

f

Poly d(CG)

2000

1000

270

290

310

330

350

λ(nm)

230 −100

250

270

290

310

330

350

λ(nm)

−300

−500

3000 2500

250

−300

d

−500

e

230 −100

−500

300

230 −100

100

De (M-1 x cm-1)

De (M-1 x cm-1)

c

300

Poly d(me5CG)

150

100

50

0

50 mM NaCl

100 mM NaCl

δ(∆ε–∆ε 0) (M-1cm-1 )

a

8 6 4

8 6 4 2 0 0

2

4

6

m

8

10

12

2 0 0

0.5

1

1.5

2

2.5

[Mg2+] (M) Fig. 2. Factors affecting DNA helicity. (a–d) Examples of detection of B–Z transition by circular dichroism. (a): d(GC)20 (bold lines) and d(GC)10 (thin lines) in 5 mM Tris–HCL buffer, pH 8.0, without additives (B-form) and (b) in 5 M NaCl (Z-form); duplex concentration is used circular dichroism calculation (additive effect of each base pair is, therefore, evident). (c) Mixed sequence containing two B–Z junctions (5′-GTAGATCACT(CG)10GTAGATCACT-3′ with complementary strand) in 1 mM cacodylate buffer, pH 6.8; bold curve: without additives; thin curve: in 5 M NaCl (B-form of terminal tracts and Z-form of CG repeats). (d) Difference between the 5 M spectrum and the contribution of the B component (half spectrum of the B-form) showing a typical Z-DNA contribution. (e, f ) Effect of methylation of cytosine on the salt concentration (e) and dependence of [Co(NH3)6]3+ on [NaCl] (f ) at midpoint of B–Z transition in alternate copolymers [according to data in Ref. (28). (g) Plot of circular dichroism at 260 nm subtracted of the initial contribution of the B-form data (De − De0) as a function of Mg2+ concentration and length of me5CG tract in a (me5CG)m(TA)16-m oligomer: filled circles: m = 12, empty circles: m = 10, filled triangles : m = 8, empty triangles : m = 6, filled squares: m = 4. Inset: plot of the net magnitude of the 260-nm band as a function of m. Reprinted from Ref. (39) with permission from Elsevier.

for asymmetric synthesis (9, 10), as a template for inducing asymmetry in achiral cyanine dyes (11) and in silver aggregates acting as nanowires (12), or for creating enantiomeric excess in racemic chiral carbon nanotubes (13). Conformational change from B- to Z-DNA results in a change in the length of the duplex, which is predictable for the Z-part, resulting in a rise per base pair that increases from 3.40 to

82

Corradini et al.

3.63–3.72 Å. However, this parameter is not well defined for the B–Z junctions (Fig. 1b), since the data in the literature vary significantly according to different conditions used (three base pairs in protein-stabilized Z-DNA crystal structure recently reported (14)). Synthetic “DNA-like” molecules are also able to form double-stranded structures, held together by Watson–Crick base pairing and base-stacking interactions, which in some cases also allow cross-hybridization with DNA (or RNA). The ability of DNA analogs to form right- or left-handed helices can be controlled by the appropriate choice of the stereochemistry of the monomeric units, a variable not available in nucleotides, and in turn the nature of the helicity of these analogs can affect their ability to bind to DNA (15). Among the most interesting structural analogs of DNA are peptide nucleic acids (PNAs, see Fig. 3), a structural motif introduced by Nielsen et  al. in the early 1990s (16). In the PNA structure, the sugar–phosphate backbone of DNA has been replaced by a polyamidic chain. A large number of experimental data are available for these compounds, as well as several structures obtained by X-ray crystallography or NMR studies, which allowed correlation of the stereochemistry of the PNA with its a

Base 2' O R'' 1' O 6 N 2 1 n* N 5 4 3 H R'

*

b

Acyclic PNA R' = R'' = H : standard achiral PNA R' = side chain, R'' = H : C2-substituted I R' = H, R'' = side chain : C5-substituted II R' = side chain, R' = side chain: C2,C5 disubstituted III

O

c

NH2

Boc

fmoc O N OH H R

N H

H2N

PNA or linker

NH3+

HN

Achiral PNA

R

Boc

N H

Monomer

N

R

H

NH2

fmoc O N R

H

OH

N H

Boc

PNA or linker

Coupling

PNA or linker

N H

a-Deprotection

Base O

N

NH3+

O

H

O

Helical induction by chiral PNA monomer(s) R

O

Epimerization-safe Coupling

Base O

NH2

Helical induction by C-terminal Amino acid

Submonomer

Boc

NH3+

Achiral PNA

H2N

d

O

NH3+

N H

O

H N R

H

N H

PNA or linker

Fig.  3. (a) Different types of chiral PNA derivatives with acyclic backbone (2-substituted, 5-substituted, and 2,5-­dis­ub­­stituted); (b) Helical induction in a PNA:PNA antiparallel duplex by a terminal amino acid. (c) Helical induction by a chiral PNA monomer. (d) Scheme for the synthesis of a 2-substituted chiral PNA monomer by a submonomer ­strategy (30, 31).

Control of Helical Handedness in DNA and PNA Nanostructures

83

affinity for DNA (17). PNA:DNA duplexes are also interesting as components of nanostructures, since PNA is able to distort the helical form of DNA, due to a far different helical pitch of 16 bases per turn, as observed in the crystal structure of one of these complexes (18). Seeman used PNA:DNA tracts as components of a two-dimensional array, finding consistently a helical pitch of 15.7 bases (19). Achiral PNA single strands form chiral PNA:PNA helical structures with equally probable right-handed and left-handed helices. Interestingly, by covalently linking a chiral amino acid at the C-terminus, a preference is gained for one or the other helical form (less if linking to the N-terminus) (20), with a mechanism of induction which is still under investigation (21), but with a propagation of chirality of the “sergeant and soldiers” type (Fig. 3a) (2). This preference is typical of PNA in solution, since it has not been observed in the solid state (22) similar to that reported for polymers (23). Helix handedness in PNA:PNA duplexes can be obtained by modification of PNA strands constructed of flexible, acyclic PNA monomers appended with side chains derived from amino acids (such as lysine or alanine), as depicted in Fig. 3a (24). It has been shown that the influence on helix handedness is stronger if the chiral monomers are placed in the middle of the PNA strands rather than at the teminus (25, 26). The effects of the configurations of chiral monomers with stereocenters at C2 or C5 (or both) (see Fig. 3) have been thoroughly investigated and rationalized. Therefore, we propose a second protocol to create helical PNA:PNA duplexes and control the correct helicity from CD spectra based on an established model (see Fig. 4a); a further confirmation is provided by the induced circular dichroism of cyanine dyes (in particular, 3,3′-diethylthiadicarbocyanine), as first described by Armitage and coworkers (see Fig. 4b, c) (27).

2. Materials 2.1. Switching the DNA Handedness Through the B–Z Transition

1. The synthetic double-stranded DNA poly d(GC) poly d(GC) and poly d(me5CG) (me5C = 5-methylcytosine) can be obtained from several companies (e.g., GE Healthcare, NJ, USA) or synthesized according to literature procedures (28). Oligomers containing a predefined stretch of alternating d(CG) residues can also be obtained from several firms at different degrees of purity. Either HPLC- or PAGE-purified grade oligomers should be used. We use HPLC-grade oligos (Thermo Fisher Scientific, Ulm, Germany). The DNA stock solutions are prepared by dissolving these polymers in double-distilled water, and stored ­frozen (−20°C) as such.

84

Corradini et al.

a

b

400

Right-Handed

300

De (M-1 x cm-1)

Left-Handed

100 0

220

270

320 l (nm)

-100

Right-Handed

De (M-1 x cm-1)

800

200

-200

1200

400 0 400

500

600

700 l (nm)

-400 -800

-300

Left-Handed

-400

-1200

c Left Handed LH

Cyanine Dye Aggregate Exciton Chirality

Right Handed RH

Fig. 4. (a) CD spectra of right-handed (full line, 2D) and left-handed (dotted line, 2L) chiral PNA (GTAGAT2-Lys CACT)–achiral PNA (complementary) duplex in phosphate buffer, pH 7.0 (c = 5  mM of each strand). (b) Induced CD spectra of 3,3′-­diethylthiadicarbocyanine dye in the presence of right-handed (full line) and left-handed (dotted line) chiral PNA (GTAGAT2-Lys CACT)–achiral PNA duplexes. Concentration used for De calculation is of the PNA duplex (hence of the aggregate), in order to allow direct comparison. (c) Scheme of generation of induced CD by aggregation on PNA:PNA duplexes of the achiral 3,3′-diethylthiacarbocyanine dye.

2. Cacodylate buffer: Dissolve the sodium salt of cacodylate (Sigma–Aldrich St. Louis, MO, USA) to a concentration of 1 mM. Add hydrochloric acid until the desired pH (6.5–7.0) is achieved (see Note 1). In addition, NaCl or other salts are used as described below. 3. A UV–vis spectrophotometer with quartz cuvettes and a temperature control device: we used a Perkin–Elmer (Norwalk, CT) k BIO 20 spectrophotometer, equipped with Peltier PTP 6 temperature programmer and a cell changer. 4. A spectropolarimeter with quartz cuvettes and a temperature control device: we used a Jasco J 715 spectropolarimeter (Tokyo, Japan) equipped with a Peltier PTC 348 temperature controller unit. Solvent and cell measurement must be subtracted from the sample spectrum. At least four accumulations are used for each spectrum, in order to reduce the signal-to-noise ratio.

Control of Helical Handedness in DNA and PNA Nanostructures

2.2. Inducing Handedness in PNA Segments

85

1. Achiral PNA oligomers are commercially available (from Panagene) and are synthesized with standard protocols for peptide synthesis (16b). We synthesize PNA on a ABI 433A (Applied Biosystems, Foster City, CA, USA) peptide synthesizer. Purification and analysis is performed on analytical HPLC-MS on a Waters 2695 separation module (Waters, Milford, MA) interfaced with a Micromass ZMD mass spectrometer (Micromass, Manchester, UK) equipped with an ESI interface and a single quadrupole analyzer. 2. Unmodified PNA monomers are commercially available from ASM (Burgwedel, Germany) [Boc(Z) and Fmoc(Bhoc) protected] or from Panagene (Daejon, Korea) [Fmoc(Bhoc) protected]; they are synthesized according to ­literature procedures (29). C2 or C5 PNA monomers are synthesized from d- or l-amino acids, according to our p ­ ublished procedures (30–32). 3. Chiral PNA monomers with stereogenic centers in position 2 can be affected by racemization during the solid-phase peptide synthesis (SPPS) procedure (33). In order to minimize this side reaction, a submonomeric protocol (Fig.  3c) has been set up on a solid phase support, which has allowed incorporation of the monomer into the PNA growing chains with very high optical purity (e.g., >94%) (30, 31). The submonomeric approach must be used to make sure that the C2-modified chiral PNA is synthesized with minimal racemization, which can be checked by chiral GC analysis (33). 4. For the insertion of chiral PNA monomers with stereogenic centers in position 5, which are not affected by racemization, and of achiral PNA monomers, normal SPPS procedures can be used (16b, 32). 5. Diethylthiadicarbocyanine solution: 500 mM 3,3′-diethylthiadicarbocyanine dye (Sigma–Aldrich St. Louis, MO, USA) in methanol. 6. Phosphate buffer for measurements: 10  mM phosphate ­buffer, pH 7.0.

3. Methods 3.1. Switching the DNA Handedness Through the B–Z Transition

The conditions able to induce a B–Z transition in a poly [d(CG)]2 and in a mixed sequence containing a d(CG)10 insert (see Note 2) in a B-DNA (Fig. 2a, d) domain are described (see Notes 3 and 4). 1. The concentration of DNA can be determined by spectrophotometry, according to the molar absorbance provided by the producer. For poly d(GC), the e257 of 8,400 mol−1 cm−1 must be used to calculate the concentration expressed as number of base residues independent of the molecular weight.

86

Corradini et al.

2. Dissolve DNA in a 80–100  mM concentration (in base) in cacodylate buffer (see Note 1), without the addition of salts. If no buffer is used, trace amounts of divalent cations must not be present (we used water with 18 MW × cm resistivity) (see Note 5). 3. Prepare a DNA solution without NaCl (or with 10  mM NaCl), and without any other additives (see Note 6). This solution is the reference B-DNA form (Fig. 2a, c). 4. Prepare a solution of poly d(CG) in the presence of 4–5 M NaCl (or 1 M MgCl2). Heat at 60°C for a few minutes and then slowly cool to room temperature. This will cause the formation of Z-DNA (Fig. 2b) (see Note 7). If a solution of poly d(me5CG) is used, 2  M NaCl is sufficient to obtain Z-DNA conversion (Fig. 2e). 5. Prepare a solution of mixed sequence containing a d(CG)10 insert in a B-DNA domain (as depicted in Fig.  1b) in the presence of 5  M NaCl, heat at 60°C for few minutes, and then slowly cool to room temperature. The resulting solution has the insert mostly in the Z-conformation, with the remaining part mostly in the B-conformation, with two B–Z junctions spanning from 3 to 10 bp each (Fig. 2c, d). 6. Record a UV spectrum of each solution between 200 and 350 nm at 25°C (or at the desired temperature). Slight differences are observed for the B- and Z-forms, with the latter having a lower absorption at 260 nm and presenting a shoulder in the 280–300-nm region. 7. Record a CD spectrum in the 200–350-nm range at 25°C. The B-form has a maximum at 280 nm, a crossover around 265–270 nm, and a minimum at 250 nm (plus other transitions more intense in the 200–230 range). The Z-form has a minimum in the 290–295-nm region, a crossover at 280 nm, and a maximum near 260–265 nm, plus other transitions more intense in the 200–230 range. In Fig. 2a, b, a typical B–Z transition is reported for d(CG)10 and d(CG)20, showing additivity of the base contribution. In Fig. 2 c, d, the transition obtained for a segment containing two B–Z junctions is reported. Assuming a typical spectrum of the B-DNA segments, the difference obtained by subtracting to the spectrum measured the contribution of the B-segments shows the typical profile and intensity of the Z-form of same length (see Fig. 2d). 8. B–Z transition. To a portion of the solutions in point 2, with 10  mM NaCl, add a solution of [Co(NH3)2]3+ to a final 20  mM concentration, under stirring (see Notes 8 and 9). This leads to a fast transition to the Z-form. The concentration of the [Co(NH3)2]3+ necessary to obtain a half transition

Control of Helical Handedness in DNA and PNA Nanostructures

87

is dependent on the concentration of NaCl (see Fig. 2f ), and increases as [Na+] increases (see Note 10). 9. Record changes in UV and CD spectra, in particular at 260 and 294 nm, and plot these as a function of salt concentration. A sigmoidal curve should be observed in each case, allowing for the detection of the mid-transition point, as illustrated by the graph reported in Fig. 2g. 3.2. Inducing Handedness in PNA Segments

1. Helix handedness prediction. Though most chiral singlestranded PNAs have low CD signals and, therefore, are not preorganized in helical structures, they show helical behavior when forming duplexes with the complementary achiral PNAs (Fig.  3c) (see Note 11). The CD results associated with helicity can be used to predict the preferred handedness of PNAs as a function of the stereochemistry of chiral monomers: (a) C2-substituted (synthesized from d-amino acid: 2D, Fig.  3a) or C5-substituted (synthesized from l-amino acid: 5L) or with both stereogenic centers (2D, 5L) simultaneously present: right-handedness preferred; (b) C2-substituted (from l-amino acid, 2L) or C5-substituted (from d-amino acid, 5D) or with both stereocenters (2L, 5D) simultaneously present: left-handedness preferred; (c) 2L and 5L stereocenters simultaneously present (2L, 5L): chiral conflict, but the induction exerted by the stereocenter in position 5 is stronger, thus right-handedness is preferred; (d) 2D and 5D stereocenters  (2D, 5D) simultaneously present: chiral conflict, but the  induction exerted by the stereocenter in position 5 is ­stronger, thus left-handedness is preferred. The preference for the ­helical sense caused by incorporation of a non-racemic chiral center at C2 and/or C5 will in turn affect the binding affinity of PNA for the nucleic acids (DNA or RNA) (see Note 12) (32). 2. Prepare a solution containing the chiral PNA (5 mM) and its complementary antiparallel achiral PNA strand (5  mM) in phosphate buffer for measurements (see Note 13). 3. Measure the circular dichroism spectrum between 220 and 350 nm. A prevalence of right-handed PNA helices is characterized by a negative Cotton effect at 275 nm and a higher (in absolute value) positive Cotton effect at 260 nm (full line in Fig. 4a). A prevalence of left-handed PNA helices is characterized by a positive Cotton effect at 275 nm and a higher (in absolute value) negative Cotton effect at 260 nm (dotted line in Fig. 4a). Though the exact percentages of right- and left-handed structures in PNA:PNA duplexes are not exactly known and are still under investigation, evidences indicate that at least for antiparallel PNA:PNA duplexes, two different PNA modifications can be compared as their ability to induce

88

Corradini et al.

a preferred handedness: if the length, the sequence, and the concentrations of the PNAs are the same, the intensity of the antiparallel CD spectra can be taken as a measure of the extent of the helical induction. 4. Prepare a solution containing the chiral PNA:PNA duplex (5 mM of each strand in phosphate buffer for measurements) in 10  mM phosphate buffer at pH 7. To this solution, add 3%  volume of diethylthiadicarbocyanine solution. The dye aggregates onto the PNA–PNA duplex and the handedness of the duplex is transferred to the dye aggregate (see Fig. 4c). 5. Measure the absorbance and circular dichroism spectrum between 400 and 700 nm. A shift in the absorption spectrum (from 648 to 534 nm) is observed if the PNA:PNA–dye interaction occurs. 6. A prevalence of right-handed PNA helices is characterized by a positive CD at about 540 nm and a negative one at about 520  nm (full line in Fig.  4b). A prevalence of left-handed PNA:PNA helices is characterized by opposite signs (dotted line in Fig. 4b) (see Note 14).

4. Notes 1. Other buffers that avoid the use of arsenic compounds are also suitable. We tested 5 mM Tris–HCl at pH 7.0, with similar results. 2. The occurrence of Z-DNA structures is strongly dependent on the sequence and base modification (34); the most favorable combinations for Z-DNA are repeated d(CG) sequences (35, 36). Unmodified d(CG)4, but not d(GC)4, undergoes the B–Z transition with a midpoint of transition at 3 M NaCl. For short sequences, the use of CG and not GC repeats is recommended. 3. The preference for the Z-forms varies as follows: d(CG) > d (TG)n > d(GGGC)n > d(TA)n (37). Substitution of cytosine with 5-methylcytosine (me5C) or other 5-substituted cytosines (bromo or iodo) has the effect of favoring the otherwise unstable Z-DNA structures, allowing this particular form to be obtained under conditions near normal physiological salt concentration. 4. In order to evaluate the possible effect of a molecule on the B–Z transition, select a condition under which B- and Z-forms are present in almost equal amounts. First, perform a study of the extent of the Z-form of the DNA as a function of salt concentration. Add the molecule under study to the solution

Control of Helical Handedness in DNA and PNA Nanostructures

89

in which B- and Z-forms are present in 1:1 ratio. Record changes in UV and CD spectra. If the 260-nm signal has increased and the 280-nm signal has decreased in UV, the extent of the Z-form has increased, and vice-versa for the B-form. The change in the percentage of B- and Z-forms can be better calculated by CD signals at 294 and 260 nm. 5. There are reports of Z-induction at very low salt concentration in the presence of traces of divalent cations. The effect of counterions has been thoroughly studied in the case of Ni2+induced B–Z transitions and the following order of Ni2+ concentration needed has been found: NaCl ≈ Me4NCl > LiCl >>  MgCl2 > no salt > NaBF4 ≈ NaNO3 ≈ NaClO4 (38). 6. Several additives can affect the B–Z transition depending on their chirality (15). Examples of Z-DNA inhibitors are actinomycin, adriamycin, and mitomycin; Z–B inducers are ethidium bromide, daunomycin, and netropsin. A more complete list is reported in Ref. (34). 7. The kinetics of formation of Z-DNA are slow. Therefore, incomplete conversion can lead to contrasting CD results. 8. Many cationic Z-DNA inducers at higher concentrations can cause DNA condensation (aggregation of several DNA segments), which competes with the formation of Z-DNA. Vigorous stirring is recommended. A UV spectrum from 350 to 200 nm should also be recorded in order to rule out the presence of a condensation process (in which case a drift of baseline is observed). 9. A solution of spermine added to a final 10 mM concentration can be used as an alternative to this method. Under stirring (see Note 8), it is heated at 60°C for 10 min and then slowly cooled to room temperature. This leads to conversion to the Z-form. 10. Other tools are available for detecting Z-DNA, such as gel electrophoresis, NMR, Z-DNA-specific antibodies (35), fluorescence energy transfer (FRET) (39), exciton chirality of appended porphyrins (40), and Raman spectroscopy (41). 11. The helical preference of a chiral PNA strand is most apparent when the PNA is in a well-defined helical conformation, i.e., when involved in a PNA–PNA double helix (with an achiral complementary PNA). Although in some cases the presence of a helical conformation might be hypothesized also in single-strand PNAs, caution is suggested when interpreting similarities in CD spectra or other experiments as “proofs” for a helical conformation of single-strand PNAs. 12. Usually, PNAs having a preference for left-handedness will nonetheless bind to DNA and RNA, by assuming the righthanded helical conformation dictated by the nucleic acids.

90

Corradini et al.

It  is exactly this forced unnatural conformation that is responsible for the decrease in stability in these PNA–DNA and PNA–RNA duplexes. 13. In order to avoid confusing effects, it is important that the complementary PNA is completely achiral, i.e., also amino acids linked at the C-terminus (as it is often the case with lysine) are to be avoided. 14. Aggregation of the dye onto the PNA–PNA duplexes depends on different factors, including steric hindrance and electrostatic repulsion or attraction, thus two different PNA modifications cannot be compared on their ability to induce a preferred handedness by considering the induced intensity of the dye in the CD spectra. References 1. (a) Hembury, G. A. , Borovkov, V.V. , and Inoue, Y. (2008) Chirality-Sensing Supramole­ cular Systems Chem Rev 108, 1–73. (b) Green, M. M. (2000) A model for how polymers amplify chirality. In Circular Dichroism (2nd Edition) Berova, N., Nakanishi, K., Woody, R.W. Eds Wiley-VCH, New York, 491–520. 2. (a) Green, M.M., Peterson, N. C., Sato, T., Teramoto, A., Cook, R., and Lifson, S. (1995) A helical polymer with a cooperative response to chiral information Science 268, 1860–6. (b) Green, M.M., Park, J.-W., Sato, T., Teramoto, A., Lifson, S., Selinger, R.L.B., and Selinger J.V. (1999). The macromolecular route to chiral amplification Angew Chem, Int Ed 38, 3139–54. 3. Tomar S., Green M. M., and Day L. A. (2007) DNA-Protein Interactions as the Source of Large Length Scale Chirality Evident in the Liquid Crystal Behavior of Filamentous Bacteriophages J Amer Chem Soc 129, 3367–75. 4. Hecht S.M. Ed. (1996) Bioorganic ChemistryNucleic Acids. Oxford University Press, Oxford-UK. 5. a) Seeman, N.C. (2007) An Overview of Structural DNA Nanotechnology Mol Biotechnol 37, 246–57. b) Brucale M., Zuccheri G., and Samorì B. (2006) Mastering the complexity of DNA nanostructures Trends in Biotech 24, 3427–34. 6. Mao, C.D., Sun, W.Q., Shen, Z.Y., and Seeman, N.C. (1999) A nanomechanical device based on the B-Z transition of DNA Nature 397, 144–6. 7. Du S. M., Stollar B. D., and Seeman, N.C. (1995) A Synthetic DNA Molecule in Three

8.

9. 10.

11.

12.

13.

14.

15.

Knotted Topologies J Am Chem Soc 117, 1194–1200. Michaud, M., Jourdan, E., Raavelet, C., Villet, A., Ravel, A., Grosset, C., and Peyrin, E. (2005) Immobilized DNA aptamers as target specific chiral stationary phases for resolution of nucleoside and amino acid derivative enantiomers Anal Chem 76, 1015–20. Roelfes, G., and Feringa, B.L. (2005) DNAbased asymmetric catalysis. Angew Chem Int Ed Engl 44, 3230–2. Li, X., and Liu, D. (2004) DNA-templated organic synthesis: nature’s strategy for controlling chemical reactivity applied to synthetic molecules Angew Chem Int Ed 43, 4848–70. Hannah, K.C., and Armitage, B.A. (2004) DNA-templated assembly of helical cyanine dye aggregates: a supramolecular chain polymerization Acc Chem. Res 37, 845–53. Shemer, G., Krichevski, O., Markovich, G., Molotsky, T., Lubitz, I., and Kotlyar, A.B. (2006) Chirality of Silver Nanoparticles Synthesized on DNA J Am Chem Soc 128, 11006–7. Dukovic, G., Balaz, M., Doak, P., Berova, N. D., Zheng, M., Mclean, R.S., and Brus, L.E. (2006) Racemic Single-Walled Carbon Nanotubes Exhibit Circular Dichroism When Wrapped with DNA J Am Chem Soc 128, 9004–5. Ha, S.C., Lowenhaupt, K., Rich A., Kim, Y.-G., and Kim, K.K. (2005) Crystal structure of a junction between B-DNA and Z-DNA reveals two extruded bases Nature 437, 1183–6. Corradini, R., Sforza, S., Tedeschi, T., and Marchelli R. (2007) Chirality as a Tool in

Control of Helical Handedness in DNA and PNA Nanostructures

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

Nucleic Acid Recognition: Principles and Relevance in Biotechnology and in Medicinal Chemistry Chirality 19, 269–94. a) Nielsen, P.E., Egholm, M., Berg, R.H., and Buchardt, O. (1991) Sequence-Selective Recognition of DNA by Strand Displacement with A Thymine-Substituted Polyamide Science 254, 1497–1500. b) Nielsen P.E. (Ed.) (2004) Peptide Nucleic Acids: Protocols and Applications (Second Edition) Horizon Bioscience, Norfolk (UK). Corradini, R., Sforza, S., Tedeschi, T., Totsingan, F., and Marchelli, R. (2007) Peptide nucleic acids with a structurally biased backbone: effect of conformational constraints and stereochemistry Curr Top Med Chem 7, 681–94. Menchise, V., De Simone, G., Tedeschi, T., Corradini, R., Sforza, S., Marchelli, R., Capasso, D., Saviano, and M., Pedone, C. (2003) Insights into peptide nucleic acid (PNA) structural features: the crystal structure of a D-lysine-based chiral PNA–DNA duplex Proc Natl Acad Sci USA 100, 12021–6. Lukeman, P.S., Mittal, A.C., and Seeman, N.C. (2004) Two dimensional PNA/DNA arrays: estimating the helicity of unusual nucleic acid polymers Chem Commun 1694–5. Wittung, P., Eriksson, M., Lyng, R., Nielsen, and P. E., Norden, B. (1995) Induced Chirality in PNA-PNA Duplexes J Am Chem Soc 117, 10167–73. Totsingan, F., Jain, V., Bracken, W. C., Faccini, A., Tedeschi, T., Marchelli, R., Corradini, R., Kallenbach, N.R., and Green, M.M. (2010) Conformational Heterogeneity in PNA:PNA Duplexes Macromolecules 43, 2692–2703. Rasmussen, H., Liljefors, T., Petersson, B., Nielsen, P. E., and Kastrup, J. S. (2004) The influence of a chiral amino acid on the helical handedness of PNA in solution and in crystals J Biomol Struct Dyn 21, 495–502. Pino, P., and Luisi, P.L. (1968) Optical activity and conformation in stereoregular vinyl polymers J Chimie Physique Physico-Chimie Bio. 65, 130–9. Puschl, A., Sforza S., Haaima, G., Dahl, O., and Nielsen, P.E. (1998) Peptide nucleic acids (PNAs) with a functional backbone Tetrahedron Lett 39, 4707–10. Sforza, S., Haaima, G., Marchelli, R., and Nielsen, P.E. (1999) Chiral peptide nucleic acids (PNAs). Helical handedness and DNA recognition Eur J Org Chem 197–204. Sforza, S., Corradini, R., Ghirardi, S., Dossena, A., and Marchelli, R. (2000) DNA Binding

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

91

of  a D-Lysine-Based Chiral PNA: Direction Control and Mismatch Recognition Eur J Org Chem 2905–13. Smith, J.O., Olson, D.A., and Armitage B.A. (1999) Molecular Recognition of PNA-Containing Hybrids: Spontaneous Assembly of Helical Cyanine Dye Aggregates on PNA Templates J Am Chem Soc 121, 2686–95. Behe, M., and Felsenfeld, G. (1981) Effects of methylation on a synthetic polynucleotide: The B-Z transition in poly(dG-m5dC) poly(dG-m5dC) Proc Natl Acad Sci USA 78, 1619–23. Uhlmann, E., Peyman, A., Breipohl, G., and Will D.W. (1998) PNA: Synthetic Polyamide Nucleic Acids with Unusual Binding Properties Angew Chem Int Ed 37, 2796–2823. Sforza, S., Tedeschi, T., Corradini, R., Ciavardelli, D., Dossena, A., and Marchelli, R. (2003) Fast, Solid-Phase Synthesis of Chiral Peptide Nucleic Acids with a High Optical Purity by a Submonomeric Strategy Eur J Org Chem 1056–63. Tedeschi, T., Sforza, S., Maffei, F., Corradini, R., and Marchelli R. (2008) A Fmoc-based submonomeric strategy for the solid phase synthesis of optically pure chiral PNAs Tetrahedron Lett 49, 4958–61. Sforza, S., Tedeschi, T., Corradini, R., and Marchelli, R. (2007) Induction of Helical Handedness and DNA Binding Properties of  Peptide Nucleic Acids (PNAs) with Two Stereogenic Centres Eur J Org Chem 5879–85. Corradini, R., Di Silvestro, G., Sforza, S., Palla, G., Dossena, A., Nielsen, P.E., and Marchelli, R. (1999) Direct Enantiomeric Separation of N-aminoethyl amino acids: Determination of the Optical Purity of Chiral Peptide Nucleic Acids (PNAs) by GC Tetrahedron Asymm 10, 2063–6. Fuertes, M.A., Cepeda, V., Alonso, C., and Pérez, J.M. (2006) Molecular Mechanisms for the B − Z Transition in the Example of Poly[d(G − C)·d(G − C)] Polymers. A Critical Review Chem Rev 106, 2045–64. Pohl, F.M., and Jovin, T.M. (1972) Saltinduced co-operative conformational change of a synthetic DNA: equilibrium and kinetic studies with poly (dG-dC) J Mol Biol 67, 375–96. Wang, A.H., Quigley, G.J., Kolpak, F.J., Crawford, J.L., van Boom, J.H., van der Marel, G., and Rich, A. (1979) The molecular Structure of the Left-Handed Z-DNA Double Helix at 1.0 Angstrom Atomic Resolution. Nature 282, 680–6.

92

Corradini et al.

37. Herbert, A., and Rich, A. (1999) Left-handed Z-DNA: structure and function Genetica 106, 37–47. 38. Spingler B. (2005) Anions or Cations: Who Is in Charge of Inhibiting the Nickel(II) Promoted B- to Z-DNA Transition? Inorg Chem 44, 831–3. 39. Jares-Erijman, E.A., and Jovin, T.M. (1996) Determination of DNA Helical Handedness by Fluorescence Resonance Energy Transfer J Mol Biol 257, 597–617. 40. a) Balaz, M., De Napoli, M., Holmes, A.E., Mammana, A., Nakanishi, K., Berova, and N.,

Purrello, R. (2005) A Cationic Zinc Porphyrin as a Chiroptical Probe for Z-DNA Angew Chem Int Ed 44, 4006-9. b) Balaz, M., Li, B.C., Steinkruger, J.D., Ellestad, G.A., Nakanishi, K., and Berova, N. (2006) Porphyrins conjugated to DNA as CD reporters of the salt-induced B to Z-DNA transition Org Biomol Chem 4, 1865–7. 41. Dai, Z.Y., Thomas, G.A., Evertsz, E., and Peticolas, W.L. (1989) The length of a junction between the B and Z conformations in DNA is three base pairs or less Biochemistry 28, 6991–6.

Chapter 7 G-Quartet, G-Quadruplex, and G-Wire Regulated by Chemical Stimuli Daisuke Miyoshi and Naoki Sugimoto Abstract Guanine-rich DNA, which is widely distributed in the human genome, can fold into a supramolecular structure called the G-wire. The G-wire possesses promising characteristics as a functional element for various applications in vitro and in vivo. Here, we describe the preparative procedures for the G-wire and signatures of G-wire formation. Procedures for the regulation of G-wire formation by chemical stimuli will be useful for in vivo and in vitro applications. Key words:  G-wire, G-quadruplex, G-quartet, Guanine, Molecular crowding, Metal ion, Polymorphism

1. Introduction There are 3.2 billion base pairs in the entire human genome, most of which may form the canonical B-form duplex via sequence specific Watson–Crick base pairs (1). Although the biological function of the genome is the storage of genetic information, sequence-specific formation of B-form duplexes is also useful for the construction of nanostructures (2, 3). In fact, many researchers have demonstrated that sequence-specific duplexes can be used to create one-dimensional, two-dimensional, and three-dimensional nanostructures (4–8). To facilitate the practical use of DNA ­nanostructures, functionalization of the nanostructures is an important topic in DNA nanotechnology. The Human Genome Project revealed that a very small fraction of the genome contains protein-encoding sequences (1). Our understanding of the functions of the remaining noncoding

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_7, © Springer Science+Business Media, LLC 2011

93

94

Miyoshi and Sugimoto

f­raction is rapidly increasing. It has also been revealed that ­repetitive DNA sequences widely distributed in the human genome have the potential to fold into noncanonical structures. These noncanonical DNA structures can undergo structural transitions depending on their sequences or surrounding conditions, which in turn can control gene expression (9–11). The polymorphic nature of noncanonical DNA structures is useful for designing functional DNA nanostructures. The most common and frequently observed repetitive DNA sequences throughout the genomes of most organisms are Guanine rich (G-rich) sequences and their complementary Cytosine-rich (C-rich) strands (12, 13). Especially, it is well known that G- and C-rich sequences are found at the ends of the chromosomes (telomeres). G- and C-rich telomeric sequences can fold into G-quadruplexes and i-motifs, respectively. G-quadruplexes can be formed by intermolecular or intramolecular association with four Hoogsteen-paired coplanar guanines called a G-quartet (Fig. 1a) (14). The G-quadruplex, which can be formed by intermolecular or intramolecular association of G-rich strands in antiparallel or ­parallel orientations (Fig.  1b), shows high structural polymorphism depending on its surrounding conditions. For example, Sugimoto and coworkers reported that only 1  mM of divalent cations largely  affects the thermodynamic parameters of the

a

b H

R

N

N

G

H

N H N

G

R N H

H

G

O

N H

N

N

N H

O H

N R

N

H

O N

H

M+

H

N

N

N

O

H

H

N

G N

N H H

N H

N R

Fig. 1. (a) Chemical structure of G-quartet. Center circle indicates metal ion. (b) Structure of antiparallel (top) and parallel (bottom) G-quadruplexes.

G-Quartet, G-Quadruplex, and G-Wire Regulated by Chemical Stimuli

95

­antiparallel G-quadruplex of d(G4T4G4) (15). G-quadruplexes are more polymorphic under molecular crowding conditions. Molecular crowding causes d(G4T4G4) and d((G4T4)3G4) to undergo transitions from antiparallel to parallel G-quadruplexes (Fig.  2a) (16). Moreover, it was demonstrated that a duplex formed of G- and C-rich DNAs under uncrowded conditions

Fig. 2. (a) Structural transition of d(G4T4G4) and d(G4T4)3G4 induced by molecular ­crowding. (b) Structure of G-rich and C-rich strands under dilute and molecular crowding conditions. (c) Schematic illustration for G-wire formation.

96

Miyoshi and Sugimoto

­ issociates upon molecular crowding and that the G- and C-rich d DNAs individually fold into quadruplexes (Fig.  2b) (17). Importantly, it was shown that the ­parallel G-quadruplex induced by molecular crowding was a high-order DNA nanostructure based on G-quartets called a G-wire (16). The G-wire consists of numerous strands of guanine-rich DNA strands aligned in parallel (Fig.  2c). Guanine monomers including guanosine and its synthetic derivatives, and guanine ­oligomers from several to thousands of consecutive guanines have been utilized for G-wire formations (17). Chen reported that Sr2+ can induce the formation of a G-wire by d((G4T2)3G4) (18). Using atomic force microscopy (AFM), Henderson and coworkers directly observed for the first time that d(G4T2G4) forms a G-wire in the presence of Mg2+ and spermidine (19). They also found that the G-wire is a one-dimensional nanowire with a uniform height and width, and few bends or kinks (20). Kotlyar et  al. developed a long G-wire using a polyG (~3,000 bases) strand that is prepared enzymatically with the Klenow fragment (21, 22). Macgregor and coworkers used d(NxGy) or d(GyNx), where y > 10 and x > 5, to prepare a DNA nanowire that they called a frayed wire (23, 24). Sugimoto and coworkers reported that Ca2+ induces G-wire formation of d(G4T4G4) (25). They further demonstrated that a single G to A substitution in the loops leads to a drastically different structure. Under molecular crowding conditions, which is one of the most drastic differences between chemical conditions of a ­living cell and test tube, d(T2(G3T2G)3G) folds into very long G-wires, whereas d((G3T2A)3G3) folds into an antiparallel G-quadruplex (26). Notably, studies on G-wires further showed that they have useful characteristics for various applications. For example, there is growing interest in G-rich sequences as functional ­elements in molecular electronics (27), since theoretical study suggests that the G-wire structure is promising for nanoscale biomolecular electronics (28). For such electronic applications, it is important that the G-wire tends to deposit on a surface with orientational distribution with three preferential directions (29). It is well known that the G-wire has promise for effecting separations and transportations of cations (30). The characteristics of G-wires are promising not only for material applications but also for biological applications. For example, it was demonstrated that cellular uptake and nuclear localization of the G-wire are more efficient than that of a single-stranded DNA, allowing us to utilize a G-wire for a drug delivery system (31). Moreover, G-wire assembly into a liquid crystalline phase (32) was applied for NMR study of detergent-solubilized membrane proteins (33).

G-Quartet, G-Quadruplex, and G-Wire Regulated by Chemical Stimuli

97

2. Materials 2.1. Preparation of DNA Samples

1. The phosphoramidite monomers, solid support (controlled pore glass bead), and all reagents and solvents for DNA ­oligonucleotide synthesis from Glen research (Sterling, VA). 2. Poly-Pak (Glen research) and Sep-Pak (Waters) C18 reversedphase cartridges for deprotection and crude preparation of oligonucleotides. 3. Reagents for purification by reverse-phase high performance liquid chromatography (HPLC) from Wako Pure Chemical Industries (Osaka, Japan). Reverse-phase column for HPLC from Tosoh (Tokyo, Japan). 4. G-wire buffer 1: 100 mM NaCl, 50 mM MES (pH 6.1), and 40 wt% polyethylene glycol (PEG 200 with average molecular weight 200) or PEG 2000. 5. G-wire buffer 2: 50 mM MES (pH 6.1), 100 mM NaCl, and 20 mM CaCl2.

2.2. Polyacrylamide Gel Electrophoresis

1. 10× Tris–borate–EDTA (TBE) buffer (1  L): 890  mM Tris, 890 mM boric acid, 20 mM Na2EDTA, and ddH2O to adjust the volume to 1 L. Stored at room temperature. 2. 20% Acrylamide solution (1 L): 190 g acrylamide, 10 g bisacrylamide, 100 mL 10× TBE buffer, and ddH2O to adjust the volume to 1 L. Stored at 4°C (see Note 1). 3. 10% Ammonium persulfate (APS) (1 mL): 100 mg APS and ddH2O to adjust the volume to 1 mL. Stored at 4°C. 4. Gel loading buffer dye for non-denaturing PAGE (5  mL): 40 wt% glycerol, 0.2 wt% Blue dextran, 0.5 mL 10× TBE buffer, and ddH2O to adjust the volume to 5 mL.

2.3. UV/Vis and Circular Dichroism Measurement

1. UV measurement: Appropriate buffer, quartz cuvettes with 1.0- and 0.1 cm path lengths, and temperature controller.

2.4. Atomic Force Microscope Measurement

1. Mica: K2O·Al2O3·SiO2 from Ted Pella Inc. (Redding, CA).

2. Circular dichroism (CD) measurement: Appropriate buffer, quartz cuvettes with 1.0- and 0.1 cm path lengths, and temperature controller.

2. AFM probe: RTESP for tapping mode in air and NP-S for tapping mode in liquid, from Veeco Instruments (Plainview, NY). 3. AFM specimen buffer: 100  mM NaCl, 50  mM MES (pH 6.1), and 100 mM CaCl2.

98

Miyoshi and Sugimoto

3. Methods 3.1. Preparation of DNA Samples

1. DNA is synthesized on solid supports using the standard ß-cyanoethyl phosphoramidite methods (34). 2. The synthesized DNA oligonucleotides containing the 5¢-end dimethoxytrityl (DMT) groups are removed from the solid support, and base blocking groups are removed by treatment with concentrated 25% ammonia at 55°C for 8 h. After drying in vacuum, the oligonucleotides are passed through a Poly-Pak cartridge with 2% trifluoroacetic acid to remove the 5¢-end DMT groups. After deblocking operations, the oligonucleotides are desalted through a C-18 Sep-Pak cartridge column. 3. The oligonucleotides are purified by HPLC on a TSK-gel Oligo DNA RP column (Tosoh) with a linear gradient of 0–50% MeOH/H2O containing triethylammonium acetate (pH 7.0) (see Note 2). The final purities of the oligonucleotides are confirmed to be >98% by HPLC. The purified oligonucleotides are desalted again with a C-18 Sep-Pak cartridge before use. 4. DNA synthesis is confirmed by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry (see Note 3). 5. Single-strand concentrations of the DNA oligonucleotides are determined by measuring the absorbance at 260 nm at a high temperature using a Shimadzu 1700 spectrophotometer (Shimadzu, Kyoto, Japan) connected to a thermoprogrammer. Single-strand extinction coefficients are calculated from mono­ nucleotide and dinucleotide data using the nearest-neighbor approximation (35). 6. To induce G-wire formation of d(G4T4G4) and d(G4T4)3G4, appropriate buffers are G-wire buffers 1 and 2 (see Note 4). 7. Before the measurement, the sample is heated to high temperature (typically around 80–95°C), gently cooled at a rate of 0.1–3.0°C/min, and incubated at the desired temperature for several hours (see Note 5).

3.2. UV Analysis

1. The UV absorbance is measured with a Shimadzu 1700 spectrophotometer (Shimadzu, Kyoto, Japan) equipped with a temperature controller. 2. Melting curves of G-wire and G-quadruplex are obtained by measuring the UV absorbance at 295 nm in appropriate buffers (see Note 6). 3. Confirm that the heating rate does not affect the shape of the melting curve. Generally, the heating rate for G-wire and

G-Quartet, G-Quadruplex, and G-Wire Regulated by Chemical Stimuli

99

Fig. 3. (a) Normalized UV melting curves for a typical G-quadruplex in a buffer of 100 mM KCl, 10 mM K2HPO4 (pH 7.0), and 1 mM K2EDTA containing 0, 10, 20, 30, or 40 wt% PEG 200 (from left to right ). (b) Differential UV spectrum of a typical G-quadruplex.

G-quadruplex is 0.1–0.5°C/min or possibly slower. DNA concentration is adjusted to an absorbance of around 1.0 at 260 nm. 4. The melting temperature (Tm) and thermodynamic parameters (enthalpy change, ∆H ˚; entropy change, T∆S ˚; and free energy change, ∆G ˚) for the G-wire and G-quadruplex formation are obtained from the UV melting curve (see Note 7). 5. The hypochromicity in the UV melting curve at 295 nm is observed for G-wire and G-quadruplex structures (Fig.  3a) (36). On the contrary, the duplex shows hyperchromicity at 295 nm. 6. The differential UV spectrum (UV spectra at a temperature higher than Tm − UV spectra at a temperature lower than Tm) shows a negative peak at around 295 nm in the case of G-wire and G-quadruplex structures (Fig. 3b) (37). 3.3. CD Analysis

1. CD experiments are carried out with a JASCO J-820 spectropolarimeter (JASCO, Hachioji, Japan). The temperature of the cell holder is regulated by a JASCO PTC-348 temperature controller (see Note 6). 2. The CD spectra are obtained by taking the average of at least three scans made from 200 to 350 nm. The buffer spectrum is subtracted from the DNA spectrum. 3. The CD spectra of the G-wire and parallel G-quadruplex have positive and negative peaks at around 260 and 240  nm, respectively (Fig. 4a). On the contrary, the CD spectra of the antiparallel G-quadruplex have positive and negative peaks at

100

Miyoshi and Sugimoto

Fig.  4. (a) CD spectra of 50  mM d(G4T4G4) in buffers containing 100  mM NaCl, 10  mM Na2HPO4 (pH 7.0), and 1  mM Na2EDTA without (bottom) or with 40 wt% co-solute at 4°C. Co-solutes used here are PEG 2000; PEG 200; tri(ethylene glycol), di(ethylene glycol), or ethylene glycol; and diEG (from top to second bottom at 260 nm). (b) Non-denaturing PAGE of d(G4T4G4) in the presence of 0, 10, 20, 30, or 40 wt% PEG 2000. Lane 1: 10-bp DNA marker, lane 2: single-stranded 24 mer DNA, lane 3: d(G4T4G4) with Mg2+ and spermidine, lanes 4–8: d(G4T4G4) with 0, 10, 20, 30, or 40 wt% PEG 2000.

around 295 and 260  nm. In addition, the mixed parallel G-quadruplex, in which three strands are in a parallel and one in an antiparallel direction, shows positive peaks at around 295 and 260 nm. 3.4. PAGE Analysis

1. The following instruction is for the Atto AE6220 gel system. This is easily adaptable to other systems. It is important to use clean glass plates. The glass plates are scrubbed with a rinsable detergent after use, rinsed extensively with distilled water, and dried. 2. Prepare a non-denaturing gel with 15 cm × 15 cm glass plates and 1.0  mm-thick spacer. For a non-denaturing acrylamide gel of 15  cm × 15  cm × 1.0  mm, 20  ml of gel solution is ­sufficient. For a 20% acrylamide gel, quickly mix 20  mL ­acrylamide solution, 180  mL APS, and 18  mL N,N,N ¢,N ¢tetramethylethylenediamine (TEMED); pour into the space between the two glass plates; insert the comb around 1 cm high; and leave it for 30–60 min (see Note 8). 3. After polymerization is complete, remove the comb and bottom spacers. Wash the gel plates with distilled water to remove any spilled acrylamide gel. 4. Fill the lower reservoir of the electrophoresis system with 1× TBE buffer. Set the gel plates to the gel system and fill the reservoir of the electrophoresis system with 1× TBE buffer until all of the sample wells are covered (see Note 9).

G-Quartet, G-Quadruplex, and G-Wire Regulated by Chemical Stimuli

101

5. Wash the wells with the filled 1× TBE buffer. Mix the loading buffer (5 mL) with 5 mL of sample and load 10 mL of each sample in a well. 6. Complete assembly of the gel system and connect to a power supply. The gel can be run for several ten minutes to several hours at 5 V/cm. 7. In order to stop running, the gel system is disconnected from the power supply and disassembled. The stacking gel with glass plates is decartelized and the gel is removed from the glass plate. 8. The separating gel is then immersed for 10 min in a dye solution such as GelStar nucleic acid gel stain (Cambrex, ME), SYBR gold (Invitrogen, Eugene, OR), or Stains-All (Sigma– Aldrich Co, St. Louis, MO) (see Note 10). 9. Pick up and immerse the gel in 1 × TBE buffer for 10 min to remove dyes binding to the gel. 10. In order to quantify the electrophoresis result, fluorescence intensity is analyzed with an image analyzer such as a Fujifilm FLA-5001 (Fujifilm, Tokyo, Japan). 11. Migration of the G-quadruplex is faster than that of a duplex of the same length. On the contrary, smearing and slower migration, including a ladder pattern, are observed for the G-wire (Fig. 4b). In addition, it is sometimes impossible for a larger G-wire to migrate through the gel (see Note 11). 3.5. AFM Measurement

1. A 25 mM DNA sample is dissolved in AFM specimen buffer and heated to 90°C for 10  min, gently cooled at a rate of 1.0°C/min, and incubated at 45°C for 3 h (see Note 12). 2. A 1 mL sample is deposited onto freshly cleaved mica three times, washed with 500 mL deionized water three times, and dried with a stream of N2 gas (see Note 13). 3. AFM images are obtained in the tapping mode using a Nanoscope III (Digital Instruments Inc., USA). The AFM image shows that the G-wire is a linear polymer, and the height and width of the G-wires are uniform with a few bends or kinks. Clear AFM images for G-wires were reported (38, 39).

4. Notes 1. Wear safety glasses, mask, and gloves since unpolymerized acrylamide is neurotoxic. 2. HPLC-grade water and methanol are recommended.

102

Miyoshi and Sugimoto

3. The matrix routinely used in MALDI-TOF mass spectrometry to analyze oligomer DNA is 3-hydroxypicolinic acid (3-HPA). 4. The pH value of buffers is adjusted with HCl and NaOH, KOH, or LiOH depending on the monovalent cation used. 5. Note that the highest temperature should be lower than the boiling temperature of the co-solute in the buffer, but higher than the melting temperature of the DNA structure. 6. When the measurement is carried out below room temperature, the cuvette-holding chamber is flushed with a constant stream of dry N2 gas to avoid condensation of water on the cuvette exterior. 7. A two-state transition between an unstructured random coil and structured states is required to evaluate the thermodynamic parameters. In the case of the G-wire, it is generally difficult to assume a two-state transition since the G-wire is not a homogeneous structure. 8. Remove air bubbles from the acrylamide gel solution before polymerization. Since the polymerization is an exothermic reaction, the completion of the reaction is detectable by ­temperature change. 9. Place the gel into the lower tank at an angle to avoid air ­bubbles forming at the bottom of the gel plates. 10. Wear safety glasses, mask, and gloves since these dyes are toxic. 11. Since the G-wire is thermally very stable, slower migration is often observed even in a denaturing gel. 12. Divalent metal ions such as Mg2+, Ca2+, or Ni2+ are required for immobilization of the DNA sample onto the mica surface, since freshly cleaved mica has a negative charge. Thus, TAE/Mg buffer (10× TAE–Mg (Tris–acetate–Mg2+) is 125  mM Mg–acetate, 400 mM Tris–HCl, and 10 mM Na2EDTA) can alternatively be used for AFM measurements of DNA nanostructure. 13. The preparation procedure for AFM measurements in liquid is different from that in air.

Acknowledgments This work was supported in part by Grants-in-Aid for Scientific Research, the Academic Frontier Project (2004–2009), and the Core Research project (2009–2014) of the Ministry of Education, Culture, Sport, Science and Technology (MEXT) of Japan; by the Long-range Research Initiative; and by the Hirao Taro Foundation of the Konan University Association for Academic Research.

G-Quartet, G-Quadruplex, and G-Wire Regulated by Chemical Stimuli

103

References 1. International Human Genome Sequencing Consortium (2004) Finishing the euchromatic sequence of the human genome. Nature 431, 931–945. 2. Seeman, N. C. (2003) DNA in a material world. Nature 421, 427–431. 3. Jaeger, L. and Chworos, A. (2006) The architectonics of programmable RNA and DNA nanostructures. Curr. Opin. Struct. Biol. 16, 531–543. 4. Sharma, J., Chhabra, R., Cheng, A., Brownell, J., Liu, Y., and Yan, H. (2009) Control of self-assembly of DNA tubules through ­integration of gold nanoparticles. Science 323, 112–116. 5. Liu, Y., Kuzuya, A., Sha, R., Guillaume, J., Wang, R., Canary, J. W., and Seeman, N. C. (2008) Coupling across a DNA helical turn yields a hybrid DNA/organic catenane doubly tailed with functional termini. J. Am. Chem. Soc. 130, 10882–10883. 6. He, Y., Ye, T., Su, M., Zhang, C., Ribbe, A. E., Jiang, W., Mao, C. (2008) Nature 452, 198–201. 7. Rothemund, P. W. (2006) Folding DNA to create nanoscale shapes and patterns. Nature 440, 297–302. 8. Goodman, R. P., Schaap, I. A. T., Tardin, C. F., Erben, C. M., Berry, R. M., Schmidt, C. F., and Turberfield, A. J. (2005) Rapid chiral assembly of rigid DNA building blocks for molecular nanofabrication. Science 310, 1661–1665. 9. Kumari, S., Bugaut, A., Huppert, J., and Balasubramanian, S. (2007) An RNA G-quadruplex in the 5’ UTR of the NRAS proto-oncogene modulates translation. Nat. Chem. Biol. 3, 218–221. 10. Rankin, S., Reszka, A. P., Huppert, J., Zloh, M., Parkinson, G. N., Todd, A. K., Ladame, S., Balasubramanian, S., and Neidle, S. (2005) Putative DNA quadruplex formation within the human c-kit oncogene. J. Am. Chem. Soc. 127, 10584–10589. 11. Qin. Y and Hurley, L. H. (2008) Structures, folding patterns, and functions of intramolecular DNA G-quadruplexes found in eukaryotic promoter regions. Biochimie, 90, 1149–1171. 12. Scaria, V., Hariharan, M., Arora, A., and Maiti, S. (2006) Quadfinder: server for identification and analysis of quadruplex-forming motifs in nucleotide sequences. Nucleic Acids Res. 34 (Web Server issue), W683–685.

13. Du, Z., Zhao, Y., and Li, N. (2008) Genomewide analysis reveals regulatory role of G4 DNA in gene transcription Genome Res. 18, 233–241. 14. Gellert, M., Lipsett, M. N., and Davies, D. R. (1962) Helix formation by guanylic acid. Proc. Natl. Acad. Sci. USA, 48, 2013–2018. 15. Miyoshi, D., Nakao, A., Toda, T., and Sugimoto, N. (2001) Effect of divalent cations on antiparallel G-quartet structure of d(G4T4G4). FEBS Lett. 496, 128–133. 16. Miyoshi, D., Nakao, A., and Sugimoto, N. (2002) Molecular crowding regulates the structural switch of the DNA G-quadruplex. Biochemistry 41, 15017–15024. 17. Miyoshi D, Matsumura S, Nakano S, Sugimoto N. (2004) Duplex dissociation of telomere DNAs induced by molecular crowding. J. Am. Chem. Soc. 126, 165–169. 18. Davis, J. T. (2004) G-quartets 40 years later: from 5’-GMP to molecular biology and supramolecular chemistry. Angew. Chem. Int. Ed. 43, 668–698. 19. Chen, F. M. (1992) Sr2+ facilitates intermolecular G-quadruplex formation of telomeric sequences. Biochemistry, 21, 3769–3776. 20. Marsh, T. C. and Henderson, E. (1994) G-wires: self-assembly of a telomeric ­oligonucleotide, d(GGGGTTGGGG), into large superstructures. Biochemistry 33, 10718–10724. 21. Marsh, T. C., Vesenka, J., and Henderson, E. (1995) new DNA nanostructure, the G-wire, imaged by scanning probe microscopy. Nucleic Acids Res. 23, 696–700. 22. Kotlyar, A., Borovok, N., Molotsky, T., Cohen, H., Shapir E., and Porath, D. (2005) Long Monomolecular G4-DNA Nanowires”. Adv. Mat. 17, 1901–1905. 23. Borovok, N., Molotsky, T., Ghabboun, J., Porath, D., and Kotlyar, A. (2008) Efficient procedure of preparation and properties of long uniform G4-DNA nanowires. Anal. Biochem. 374, 71–78. 24. Protozanova, E. and Macgregor, R. B. Jr. (1996) Frayed wires: a thermally stable form of DNA with two distinct structural domains. Biochemistry 35, 16638–16645. 25. Yanze, M. F., Lee, W. S., Poon, K., PiquetteMiller, M., and Macgregor, R. B. Jr. (2003) Cellular uptake and metabolism of DNA frayed wires. Biochemistry 42, 11427–11433. 26. Miyoshi, D., Nakao, A., and Sugimoto, N. (2003) Structural transition from antiparallel

104

27.

28.

29. 30.

31.

32. 33.

Miyoshi and Sugimoto to parallel G-quadruplex of d(G4T4G4) induced by Ca2+. Nucleic Acids Res. 31, 1156–1163. Miyoshi, D., Karimata, H., and Sugimoto, N. (2005) Drastic effect of a single base ­difference between human and tetrahymena telomere sequences on their structures under molecular crowding conditions. Angew. Chem. Int. Ed. 44, 3740–3744. Porath, D., Bezryadin, A., de Vries, S., and Dekker, C. (2000) Direct measurement of electrical transport through DNA molecules. Nature 403, 635–638. Calzolari, A., Felice, R. D., Molinari, E. (2002) G-quartet biomolecular nanowires. Appl. Phys. Lett. 80, 3331–3333. Vesenka, J., Bagg, D., Wolff, A., Reichert, A., Moeller, R., and Fritzsche, W. (2007) Autoorientation of G-wire DNA on mica. Colloids Surf B Biointerfaces 58, 256–263. Davis, J. T. and Spada, G. P. (2007) Supramolecular architectures generated by self-assembly of guanosine derivatives. Chem. Soc. Rev. 36, 296–313. Spada, G. P. and Gottarelli, G. (2004) Synlett 596–602. Lorieau. J., Yao, L., and Bax, A. (2008) Liquid crystalline phase of G-tetrad DNA for NMR

34. 35.

36. 37.

38.

39.

study of detergent-solubilized proteins. J. Am. Chem. Soc. 130, 7536–7537. Sanger. W. (1984) Principle of Nucleic Acids and Structure, Springer-Verlag, New York. Richards, E. G. (1975) Use of tables in calculation of absorption, optical rotatory dispersion and circular dichroism of polyribon ucleotides. In Fasman,G.D. (ed.), Handbook of Biochemistry and Molecular Biology, 3rd edn. CRC Press, Cleveland, OH, USA, Vol. 1, pp. 596–603. Mergny, J. L., Phan, A. T., and Lacroix L. (1998) Following G-quartet formation by UV-spectroscopy. FEBS Lett. 435, 74–78. Miyoshi, D., Nakamura, K. Karimata, H., Ohmichi, T., and Sugimoto, N. Hydration of Watson-Crick base pairs and dehydration of Hoogsteen base pairs inducing structural polymorphism under molecular crowding conditions. J. Am. Chem. Soc., 130, in press (2009). Mergny, J. L., Li, J., Lacroix, L., Amrane, S., and Chaires, J. B. (2005) Thermal difference spectra: a specific signature for nucleic acid structures. Nucleic Acids Res. 33, e138. Kunstelj, K., Federiconi, F., Spindler, L., and Drevensek-Olenik, I. (2007) Self-organization of guanosine 5’-monophosphate on mica. Colloids Surf B Biointerfaces 59, 120–127.

Chapter 8 Preparation and Atomic Force Microscopy of Quadruplex DNA James Vesenka Abstract The purpose of this chapter is to provide detailed instructions for the preparation and atomic force microscopy (AFM) imaging of linear chains of quadruplex DNA (a.k.a. “G-wire DNA”). Successful selfassembly of long chain quadruplex DNA requires pure concentrated guanine-rich oligonucleotide sequence (GROs) and monovalent cations in a growth buffer. AFM imaging of individual G-wire DNA strands requires many carefully monitored steps, including substrate preparation, G-wire concentration, adsorption onto substrate, rinsing, drying, appropriate selection/use of imaging probes, and dry atmosphere imaging conditions. Detailed step-wise instructions are provided. Key words: G-wire, Quadruplex DNA, Guanine-rich oligonucleotide, Self-assembly, Atomic force microscopy

1. Introduction G-DNA is a polymorphic family of four-stranded structures containing guanine tetrad motifs (1, 2) (see Fig. 1a). Guanine-rich oligonucleotides (GROs) that are self-complementary, as found in many telomeric (chromosome ends) repeat sequences (3, 4), form G-DNA in the presence of monovalent and/or divalent metal cations. The length and number of guanines and linker residues in GROs determine their diverse topologies. Quadruplex DNA has been constructed from mono, double, and quadruple strands of GROs, and are looped or linear, parallel or antiparallel (4), with a minimum of two base-stacked G-quartets (Fig. 1b). Naturally occurring hairpin structures, comprised of guanine quartets, are thought to play a role in telomerase activity that is essential for DNA replication (5). GROs can self-assemble in the

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_8, © Springer Science+Business Media, LLC 2011

105

106

Vesenka

Fig. 1. Quadruplex DNA is composed of G-quartets (a). The Tet1.5 monomer can form a dimer pair with a “closed,” “looped,” or “staggered” conformation as shown in (b). In either of the closed or looped conformations, no more growth of the G-wires can occur. In the staggered conformation, another dimer can attach to the G-wire ladder creating a succession of “sticky ends,” enabling multimers to assemble. The process is driven thermodynamically (14 ). Monomeric cation species, such as potassium or sodium, are known to stabilize the G-wires down the base-stacked core of the structure as seen in panel (b) (3, 4 ).

presence of monovalent cations to micrometer length linear chains, hence the term “G-wires.” The images shown in this work are from Tetrahymena thermophila with the oligonucleotide sequence of G4T2G4 (Tet1.5). Individual strands are easily imaged by atomic force microscopy (AFM) on the surface of mica because they are exceptionally stable to electrostatic collapse, unlike ­double-stranded DNA (see Fig.  2) (6). Long chain quadruplex DNA is of interest to nano- and biotechnological fields as ­templates for molecular wires (7) and as therapeutics (8). The procedure involved in successful growth of G-wire DNA involves starting with highly purified oligonucleotide sequences.

Preparation and Atomic Force Microscopy of Quadruplex DNA

b

K+ Mg2 G +

-

N

Nc -

-

G

M+ G G

107

-

-

Fig. 2. Electrostatic pinning (6 ) suggests that the greater internal attractive forces of G-wire DNA, comprised of guanine-quartet building blocks, four in a row, enable it to retain its solution-state and crystal-state structure when exposed to tether cations, here shown as magnesium. However, the stronger tethering force exerted on the unsupported phosphate backbone of duplex DNA, shown here as N–Nc (complementary) base pair, electrostatically pins the duplex DNA flat to the mica substrate.

These can either be purified from natural sources (9) or be ­synthetically constructed (10). The next important consideration is to include the appropriate monovalent cations in the buffer at sufficient oligonucleotide concentrations so that long G-wires are formed. Lastly, successful AFM imaging of G-wire DNA involves dilution (for individual strand observation), rapid adhesion onto smooth substrates, and thorough rinsing with deionized water to remove undesirable salt artifacts from the imaging process.

2. Materials With few exceptions (3), GROs self-assemble into G-wires in the presence of potassium or sodium. Quadruplex DNA formed in the presence of sodium tends to be interplanar or planar with the G-quartets, whereas those formed in the presence of potassium are almost exclusively interplanar (11, 12). Thus growth buffers must contain one of these two ions. Temperature seems to play little role in the growth or stability of many G-wires (13). The self-assembly process is concentration driven (14). Successfully grown G-wires require a growth concentration in the neighborhood of 100  mM, but must be diluted for AFM imaging of individual G-wires.

108

Vesenka

2.1. GRO Self-assembly

1. Growth buffer: 5 mM NaCl (or KCl), 10 mM MgCl2, 10 mM Tris–HCl, pH 7.5, and 1 mM spermidine (see Note 1). Though optimized for the Tet1.5 sequence, this growth buffer appears to work for many other naturally occurring GROs (J. Vesenka, unpublished data). 2. Obtain Tet1.5 from natural sources (9) or it can be synthetically made. For Tet1.5 (G4T2G4) sequence, 1OD = 1A260 of lyophilized G4T2G4 = 9.99  nmole. Dissolver this amount in 100 mL of growth buffer to provide a starting concentration of 100 mM, the minimum recommended for successful G-wire self-assembly (see Note 2).

2.2. Imaging Buffers

The presence of divalent cations, such as magnesium, improves the adsorption of duplex (15) and quadruplex (6) DNA onto the substrate of choice, mica. One consequence of the use of divalent cations to tether DNA onto mica is that it leads to collapse of duplex DNA, whereas the inherent stability of the quadruplex DNA helps to maintain its structure (Fig. 2). The growth buffer described previously will work satisfactorily as an imaging buffer. Salt artifacts can be further reduced in AFM images by substituting acetate (Ac) for chloride ions because of acetate’s greater volatility [e.g., 5 mM NaAc (or KAc), 10 mM MgAc2, and 10 mM Tris–Ac, pH 7.5].

3. Methods 3.1. GRO Self-assembly

No special temperature other than being in a liquid state is required for self-assembly of G-wires. After 24 h of incubation in the buffer described in Subheading  2.1, three samples were deposited and imaged on the surface of mica (procedure described shortly). At 100 mM concentration, the entire surface is coated with strands of DNA (Fig. 3a). At 10 mM concentration, the “DNA network” is

Fig. 3. Fresh, lyophilized Tet1.5 GRO incubated for 24 h in growth buffer. (a) 100 mM of incubated GRO adsorbed on mica for 10 min, rinsed, dried, and imaged. (b) 10 mM of incubated GRO adsorbed on mica for 10 min, rinsed, dried, and imaged. (c) 1 mM of incubated GRO adsorbed on mica for 10 min, rinsed, dried, and imaged. Note the transition from packed strands to a G-wire network to individual wires. Average length of individual wires in the image was 70 ± 30 nm and their average height was (equivalent diameter) 2.0 ± 0.1 nm.

Preparation and Atomic Force Microscopy of Quadruplex DNA

109

commonly observed (Fig. 3b). At 1 mM concentration, G-wires are observed with lengths in the 70 ± 30-nm range and 2 ± 0.1 nm in height (equivalent to diameter – Fig. 3c). If the concentration is decreased by another order of magnitude, typically no linear structures can be found, presumably because the G-wires have disassembled into constituent GROs. Low concentrations (1 mM) and longer adsorption times (at least 1  h) onto mica commonly lead to “auto-orientation” of quadruplex DNA with the hexagonal surface of mica (Fig.  4a) (16). Low concentrations and extremely long adsorption times (weeks) will lead to “rafting” of the G-wires (Fig. 4b, c). When growing G-wires at elevated temperature, e.g., 37°C, evaporation of buffer can present a problem. Evaporation can be remedied by injecting a layer of fresh mineral oil that will float over the

Fig. 4. A notable feature of the shorter segments of quadruplex DNA is their ability to “auto-orient” on the surface of mica (17 ). (a) After 1 h of incubation, the auto-orientation is easily observed. (b, c) Longer (weeks) incubation times lead to the self-assembly of G-wire “rafts,” still with noticeable auto-orientation. The rinse and blast drying do not appear to affect the orientation, presumably because of the negligible impact of surface tension on such small molecules.

110

Vesenka

GRO–buffer ­mixture (see Note 3). Long-term storage at elevated temperatures is not advisable as the samples appear to discolor, possibly due to impurities from the mineral oil interacting with the GROs–buffer. However, long-term storage at room temperature, 4°C, or freezing appears to have no impact on G-wire integrity. 3.2. AFM Sample Preparation and Imaging

1. Attach a mica disk (high-quality, 10 mm diameter disks were used in this work) to a ferrous disk (steel “puck”) using adhesive tabs (17). The metal disk can be secured to a magnet for the purposes of cleaving mica, rinsing, and drying. 2. Freshly cleave the mica substrate using transparent tape by firmly pressing the tape onto mica and pulling the tape off as if you were opening a hard cover book. Peeling mica off by rolling the tape will damage the mica. 3. Examine the fresh mica surface for mirror smoothness. This will aid in uniform spreading of the G-wire sample. Mica that appears “scratched” should be cleaved again or discarded. 4. Place 20  mL of sample onto the mica and check for uniform spreading of the solution over the disk. Uniform spreading is an indication of a clean substrate and will provide good G-wire adhesion. A sample that does not spread will result in poor AFM images. Repeat from step 2 if the sample does not spread. 5. Incubate on the mica for the desired length of time (from seconds to weeks). Do not let the sample dry, as dried buffer salts ruin AFM imaging (see Note 4). 6. Rinse with 1 ml of deionized water. This is most easily accomplished by direct deposition of the water onto the mica/puck attached to a secure magnet. Draining the sample is not necessary. 7. Immediately after rinsing, “blast dry” the sample with dry nitrogen at a gauge pressure of about 20  psi = 140  kPa. A nozzle connected to flexible Tygon™ tubing attached to a tank of dry nitrogen with an easy open valve works well. A 1  mL pipette tip with its end snipped off makes a good nozzle (see Note 5). 8. The sample should then be imaged immediately or stored in a dry environment until ready for imaging. Relative humidity above 25% will lead to migration of residual salts on the hygroscopic mica surface (18) and will ruin imaging. 9. To optimize dynamic atomic force microscopy resolution of the samples prepared above, four elements should be considered: A vibrationally, electronically, and thermally stable microscope; low humidity; sharp AFM probes; and slow scans with small RMS amplitudes. Recent advances in improving the stability of scanning probe microscopes have allowed for

Preparation and Atomic Force Microscopy of Quadruplex DNA

111

real-time corrections at the atomic scale (19). Until these processes are commercialized, high-resolution imaging requires active vibration isolation (20) and several hours of exercising the piezoelectric scanner. Active scanning at the desired scan range to work out piezoelectric hysteresis is microscope dependent. Some AFM systems allow this to be operator controlled (e.g., NDT), and other systems require “false engagement” (e.g., Digital Instruments/Veeco) to avoid tip contact (see Note 6). 10. Humidity and thermal equilibrium can be achieved by installing the sample and tip in the microscope surrounded by a plastic cover. Run a low-pressure stream (see Note 7) of dry helium through a hose into the cover and allow the microscope to equilibrate for several hours while electronic stabilization is being performed. The dry helium has an added effect of slightly improving the quality value of the resonance peak. 11. Sharp hydrophobic AFM probes are now available from numerous manufacturers. These include materials such as carbon nanotubes (21) and diamond-like carbon whiskers (22). Hydrophobicity of the probe reduces the likelihood of contamination from the G-wires and extends probe life. Follow the recommended manufacturer’s instruction on the use of these tips to optimize their imaging performance (23). In brief, the equipment should be well stabilized and the RMS oscillation of the cantilever should be around 1 nm to reduce ruinous impulse between the AFM tip and sample. Scan sizes should be small (about 250 nm or less) and scan speeds at this size slow (0.1 Hz). The scan speed can be proportionally increased with decrease in scan size. The importance of system stability becomes obvious as scanner drift can obscure the image. 12. After stabilization, adjust the scan size to “0” and adjust the set point while still scanning to the desired RMS value, and re-engage the microscope. When the microscope has automatically contacted the surface, the set point can be increased to back off the tip from the surface and the scan size restored to the desired value. When ready to image, decrease the set point manually to a value just below the automatic engagement value and capture the image.

4. Notes 1. The monovalent cation stabilizes the G-quartet. The magnesium ions and spermidine stabilize the G-wire phosphate backbone. Prepare buffer from fresh deionized water at 10× the

112

Vesenka

concentration described above and aliquot into sterile 1  ml Eppendorf tubes and freeze at −20°C for later use. 2. To estimate the quantity of single-stranded GROs needed in micrograms for different sequences, use the following formulas (24):

1 A 260 unit (= 1OD) of single-stranded DNA » 33 mg/mL



#(mg) ´ 3.0 (nmole/(mg)/N = #(nmole), where N is the number of bases in the oligonucleotide sequence. 3. When taking an aliquot from a concentrated GRO solution above room temperature with mineral oil on the top, the sample should first be allowed to cool to room temperature to avoid gas expansion (and incorrect volume measurements) in the pipette tip. The pipette tip can then be immersed through the oil layer with the plunger pre-depressed to a selected volume slightly greater than desired. After collecting the sample, wipe the outer edges of the pipette tip on wax film to remove excess oil. Deposit the aliquot on a fresh piece of wax film and use a fresh pipette tip to siphon off the desired volume of the aqueous droplet that remains, leaving behind any remaining mineral oil residue. This process can be used down to 1 mL effectively. 4. If the ambient humidity is low, evaporation will be easily noticed. At 10% relative humidity, a 20 mL sample will evaporate in less than 30  min. Long incubation times necessitate keeping the sample moist. This is most conveniently achieved by placing the freshly made sample on an elevated platform inside a Petri dish (e.g., 3 cm diameter) with 1 mL of deionized water on the bottom of the dish. Cover the dish and seal it with wax tape (e.g., Parafilm™) and secure in a safe place until you are ready to proceed to imaging. 5. A fast stream of nitrogen decreases the size of salt crystals, but can literally blow away a weakly anchored layer of mica. A clean white Styrofoam shipping box can help to monitor this procedure: examine the box after drying the sample to see if mica fragments are visible. If no DNA is found upon AFM imaging, the culprit may very well be that the sample layer was accidentally removed. In either case, the sample preparation must be repeated. 6. “False engagement” can be achieved on a Digital Instruments/ Veeco Multimode AFM by reducing the amplitude set point in Tappingmode™ to “0”. The software will cause the scanner to engage falsely and scan at the speed and size determined by the user. 7. A low stream of helium can be established by the “lip test.” Adjust the stream so that it just barely registers on whetted lips.

Preparation and Atomic Force Microscopy of Quadruplex DNA

113

References 1. Williamson, J.R., Raghuraman, M.K., and Cech, T.R. (1989) Monovalent cationinduced structure of telomeric DNA: the G-quartet model. Cell. 59, 871–880. 2. Williamson, J.R. (1993) G-Quartets in Biology: Reprise. Proc. Natl. Acad. Sci. USA. 90, 3124–3124. 3. Burge, S., Parkinson, G. N., Hazel, P., Todd, A. K., and Neidle, S. (2006) Quadruplex DNA: sequence, topology and structure Nucleic Acids Research, 34:19, 5402–5415. 4. Williamson, JR. (1994) G-quartet structures in telomeric DNA Annu Rev Biophys Biomol Struct. 23, 703–730. 5. Sen, D. and Gilbert W. (1992) Novel DNA superstructures formed by telomere-like oligomers. Biochem. 31, 65. 6. Muir, T., Morales, E., Root, J., Kumar, I., Garcia, B., Vellandi, C., Marsh, T., Henderson, E., and Vesenka, J. (1998) The morphology of duplex and quadruplex DNA on mica. J. Vac. Sci. Technol. A. 16, 1172–1177. 7. Weglarz, M., Fritzsche, W., Yerkes, S., Kmiec, E., and Vesenka, J. (2008) “Analysis of G5x Quadruplex DNA”, DNA-Based Nano-Scale Integration, AIP Conference Proceedings 1062, 123–128. 8. Borman, S., (2007) Ascent Of Quadruplexes: Nucleic acid structures become promising drug targets Chem. & Eng. News 85, 22, 12–17. 9. Marsh, T.C., Vesenka, J., and Henderson, E., (1995) A new DNA nanostructure, the G-wire, imaged by scanning probe microscopy, Nucleic Acids Res. 23:696–700. 10. Vesenka, J., Vellandi, C., Kumar, I., Marsh, T., and Henderson, E., (1998) The diameter of duplex and quadruplex DNA measured by Scanning Probe Microscopy. Scanning Microscopy 12:2, 329–342. 11. Howard, F. B., Frazier, J. and Miles, H. T. (1977) Stable and metastable forms of poly(G). Biopolymers, 16, 791–809. 12. Dingley, A. J., Peterson, R. D., Grzesiek, S. and Feigon, J. (2005) Characterization of the cation and temperature dependence of DNA quadruplex hydrogen bond properties using high-resolution NMR. J. Amer. Chem. Soc., 127, 14466–14472. Sci. USA, 102, 634–639.

13. Schwartz, T.R., Vasta, C.A. Bauer, T.L. Parekh-Olmedo, H., and Kmiec, E., (2008) G-Rich Oligonucleotides Alter Cell Cycle Progression and Induce Apoptosis Specifically in OE19 Esophageal Tumor Cells, Oligonucleotides 18, 51–63. 14. Marsh, T., and Vesenka, J. (2007) Properties of G-Wire DNA. Nano and Molecular Electronics Handbook, Sergy Lyshevski ed., CRC Press, New York, 13, 1–15. 15. Vesenka, J., Tang, C. L., Guthold, M., Keller, D., Delaine, E., and Bustamante, C. (1992) A substrate preparation for imaging biomolecules with the scanning force microscope. Ultramicroscopy, 42–44, 1243–1249. 16. Vesenka, J., Bagg, D., Wolff, A., Reichert, A., and Fritzsche, W., (2007) Auto-Orientation of G-wire DNA on Mica. Colloids and Surfaces B: Biointerfaces, 58, pp. 256–263. 17. E.g. (this is NOT a product endorsement) all supplies from http://www.tedpella.com/. 18. Vesenka, J., Manne, S., Yang, G., Bustamante, C., and Henderson, E., (1993) Humidity effects on atomic force microscopy of goldlabeled DNA on mica. Scanning Microscopy, 7, 781–788. 19. Perkins, T., King, G., Carter, A. and Churnside, A., (2008) Ultrastable atomic force microscopy: atomic-scale stability and registration in ambient conditions. AFMBiomed Conference, Monterey CA, October 15–18. 20. An animated summary can be found at http:// physics-animations.com/Physics/English/ spri_txt.htm. 21. Hall, A., Matthews, W. G., Superfine, R., Falvo, M. R., and Washburn, S., (2003) Simple and efficient method for carbon nanotube attachment to scanning probes and other substrates. Appl. Phys. Lett. 82, 2506. 22. Klinov, D., and Magonov, S., (2004) True molecular resolution in tapping-mode atomic force microscopy with high-resolution probes. Appl. Phys. Lett. 84, 2697. 23. E.g. (this is NOT a product endorsement) http://www.spmtips.com/howto/res/hr 24. E.g. (this is NOT a product endorsement) http://www.biosyn.com.

Chapter 9 Synthesis of Long DNA-Based Nanowires Alexander Kotlyar Abstract Here we describe novel procedures for production of DNA-based nanowires. This include synthesis and characterization of the one-to-one double-helical complex of poly(dG)–poly(dC), triple-helical poly(dG)– poly(dG)–poly(dC) and G4-DNA, which is a quadruple-helical form of DNA. All these types of DNAbased molecules were synthesized enzymatically using Klenow exo− fragment of DNA Polymerase I. All the above types of nanowires are characterized by a narrow-size distribution of molecules. The contour length of the molecules can be varied from tens to hundreds of nanometers. These structures possess improved conductive and mechanical properties with respect to a canonical random-sequenced DNA and can possibly be used as wire-like conducting or semiconducting nanostructures in the field of nanoelectronics. Key words: DNA nanowires, Enzymatic synthesis, Klenow exo−, Poly(dG)–poly(dC), G4-DNA, Triplex DNA

1. Introduction The DNA molecule is an attractive candidate to wire electrons over long molecular distances. Charge migration along DNA molecules has been the subject of scientific interest for many years. It is currently accepted by the scientific community that a native, random sequence DNA is not a good electrical molecular wire, due to its apparent poor intrinsic conductivity. It has been demonstrated that uniform DNA comprising repeating sequences improves conduction properties. Recent experimental demonstration of the conducting behavior in short poly(dG)–poly(dC) DNA oligomers (1, 2) and the results of theoretical calculations show that poly(dG)–poly(dC), a homopolymer consisting of a pair of poly(dC) and poly(dG) chains, exhibits better conductance than poly(dA)–poly(dT) homopolymer (3). This is mainly due to the fact that poly(dG)–poly(dC) provides better conditions

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_9, © Springer Science+Business Media, LLC 2011

115

116

Kotlyar

for p overlap compared to poly(dA)–poly(dT). In addition, ­guanine, which present in a high quantity in the former DNA is ­characterized by lowest ionization potential among DNA bases, thus promoting charge migration through the DNA (4). These facts emphasize the importance of guanine-rich DNA-based ­molecules as possible candidates for molecular electronics applications. The following guanine-rich DNA-based structures, whose syntheses is described here, may offer the desired electrical conduction properties (1) double-stranded poly(dG)–poly(dC), (2) triple-stranded poly(dG-dG)–poly(dC), and (3) four-stranded G4-DNA. A poly(dG)–poly(dC) is a double-stranded deoxyribopolynucleotide polymer, which consists of a pair of antiparallel poly(dC) and poly(dG) homopolymers. Commercial preparations of the DNA are available and were used by researchers in electrical conductivity studies (5, 6). We have demonstrated (7), however, that commercial preparations of poly(dG)–poly(dC) consists of long continuous poly(dC)-strand and relatively short poly(G) fragments, 500–1,500 bases long, associated with the C-strand, but not covalently connected to each other. The presence of the G-strand breaks along poly(dG)–poly(dC) must strongly reduce the ability of the polymer to conduct current and strongly limits the use of the ­molecules in nanoelectronics. The molecules prepared by our technique (7) lack the above disadvantages. The enzymatic synthesis, conducted as described below, yielded doublestranded poly(dG)–poly(dC) characterized by a well-defined length (up to 10 kb) and narrow-size distribution of molecules. The synthesized molecules composed of continuous dG- and dChomopolymers of equal length lacking strand nicks. In addition, the poly(dG)–poly(dC) may comprise a functional group attached to 5¢ ends of either one or both strands composing the DNA (7). The functional group may be a particular sequence of singlestranded DNA, fluorescent labels, thiol-groups, biotin moieties, and other groups. Thiols are known to interact specifically with gold (8, 9). Introduction of SH-groups at the ends of DNA was used to anchor the DNA fragments to flat gold surfaces (10). The ability to attach SH-groups to the 5¢ ends of the strands thus provides a tool for the selective binding of long poly(dG)–poly(dC) polymers to gold surfaces and gold nanoelectrodes. This property is especially useful for application of the polymer in nanoelectronics. Triple-stranded DNA structures have been a subject of research in the past 50 years ( for review, see Refs. 11–13). Most of these structures are composed of tens of triads. Long triplestranded DNA, poly(dG-dG)–poly(dC) have been reported by us only recently (14). A poly(dG-dG)–poly(dC) is an intramolecular triplex, composed of continuous dG- and dC-homopolymers. The length of the dG-homopolymer is twice the length of the

Synthesis of Long DNA-Based Nanowires

117

poly(dC), which enables the former strand to fold back to pair with the adjacent of poly(dG)–poly(dC) duplex. The length of the latter triplex structures can be varied from ten to hundreds of nanometers. We were able to obtain the triplex nanostructures characterized by very narrow-size distribution by applying the enzymatic method described below. As in the case of poly(dG)– poly(dC) molecules, various functional group including fluorescent labels, thiol-groups can be attached to 5¢ ends of either one or both strands composing the triplex (14). We have demonstrated that the triplex molecules are stiffer and more resistant to mechanical deformation compared to random sequence DNA and poly(dG)–poly(dC) (14). This property, together with the ability to functionalize the synthesized triplex molecules is essential for their application as elements in nanodevices. G4-DNA wire is a very stable molecule made of consecutive stacked arrangement of 4 guanine (G) bases (tetrads – a tetrad consists of 4 G bases). It is known for decades that G-rich DNA sequences containing runs of guanines (dG) can form G-quadruplex structures ( for review see Refs. 15–17). These structures, commonly named G4-DNA, are comprised of stacked tetrads; each of the tetrads arises from the planar association of four guanines by Hoogsteen hydrogen bonding. Most of the studies have been performed using short (16–32 bases) G-rich telomeric oligonucleotides (18, 19). Short G-rich oligonucleotides were shown to assemble spontaneously into long molecular wires in the presence of proper monovalent cations (20, 21). These wires are very polymorphic, and constructed of short oligomers, resulting in nonuniform polymers with gaps (noncovalently bonded backbone) between G-rich oligonucleotide fragments along the formed wires (20, 21). Guanine tetrads were proposed as building blocks of molecular nanodevices (16, 17, 22). However, the above wires, formed by many short DNA segments, and containing many nicks are probably not good candidates for application as molecular nanowires. For the utilization of G4-DNA in nanoelectronics, long, persistent, homogenous populations of molecules are required. Only recently, we have reported a method (described below) for synthesis of novel long (hundreds of ­nanometers) continuous G-based nanostructures, composed of ­hundreds of stacked tetrads (23). These nanostructures are characterized by a narrow length distribution and contain no gaps in their backbone. We have also demonstrated that these wires are characterized by higher stability, resistance to heat treatment and higher charge polarizability, as compared to double-stranded DNA (24). These properties make these structures very promising for nanoelectronic applications. Enzymatic synthesis of the long, di, tri, and tetra-stranded G-rich DNA-based nanostructures is described in detail below.

 

118

Kotlyar

2. Materials 2.1. Preparation of (dG)12–(dC)12 Template–Primer

1. 0.1 M NaOH solution: 4 g of NaOH. Add 1 L of fresh water purified by a Seralpur Pro 90 CN system (Merck Belgolabo, Overijse, Belgium), deionized water and filtered through 0.22  mm Millipore Express PLUS membrane filter, deionized/filtered water. 2. 0.1  M NaOH containing 10% acetonitrile: 4  g of NaOH. Add 0.9 L of deionized/filtered water. Add 100 mL of acetonitrile (Bio Lab, HPLC-S Gradient grade). 3. 1  M phosphate buffer (pH 7.5): 136  g of KH2PO4. Add 700 mL of deionized water; adjust the pH to 7.5 with 1 M KOH and add deionized water to 1  L. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 4. 20  mM Phosphate buffer (pH 7.5) containing 10% acetonitrile: to 880 mL of deionized water add 20 mL of 1 M Phosphate buffer (pH 7.5) and filter the solution through 0.22  mm Millipore Express PLUS membrane filter. Add 100 mL of acetonitrile (Bio Lab, HPLC-S Gradient grade). 5. 250  mM Phosphate buffer (pH 7.5) containing 10% acetonitrile: to 650 mL of deionized water add 250 mL of 1 M Phosphate buffer (pH 7.5) and filter the solution through 0.22  mm Millipore Express PLUS membrane filter. Add 100 mL of acetonitrile (Bio Lab, HPLC-S Gradient grade). 6. Alkaline 1 M NaCl solution containing acetonitrile: 58 g of NaCl (Merck). Add 0.9 L of 0.1 M NaOH and filter the solution through 0.22  mm Millipore Express PLUS membrane filter. Add 100 mL of acetonitrile (Bio Lab, HPLC-S Gradient grade). 7. Glacial acetic acid 100%. 8. 1 M Tris–acetate (pH 7.5): 121 g of Tris-base. Add 800 mL of deionized water; adjust the pH to 7.5 with glacial acetic acid and add deionized water to 1  L. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 9. 2 mM Tris–acetate (pH 7.5): to 998 mL of deionized/­filtered water add 2 mL of 1 M Tris–acetate (pH 7.5). 10. 20  mM Tris–acetate (pH 7.5): to 980  mL of deionized/­ filtered water add 20 mL of 1 M Tris–acetate (pH 7.5). 11. (dG)12 oligonucleotide from Alpha DNA (Montreal, Canada). 12. (dC)12 oligonucleotide from Alpha DNA (Montreal, Canada). 13. Dialysis tubing 10 mm (Sigma). Treat the tubing as follows: wash with running tap water for 3 h; treat with 0.3% (w/v) solution of sodium sulfide at 80°C for 1 min; wash with tap

Synthesis of Long DNA-Based Nanowires

119

water for 2–5 min at 60°C; treat with 0.2% (v/v) solution of sulfuric acid for 10 min at room temperature; wash with tap water for 10–15 min. Store the tubing in 25% ethanol at 4°C. Rinse the tubing with running deionized/filtered water before use. 14. Ion-exchange PolyWax LP column (4.6 × 200  mm, 5  mm, 1,000 Å) (Western Analytical Products). 15. Sephadex NAP-25 DNA-Grade column (15 × 50  mm) (GE Healthcare). 16. Ion-exchange HiTrap Q HP column (1 mL) (GE Healthcare). 17. Agilent 1100 HPLC system with a photodiode array detector. 18. Eppendorf table centrifuge (model 5424). 19. Laboratory Freeze Dryer Christ Alpha 1–4 (Osterode am Harz, Germany). 2.2. Preparation of Thiol-End-Labeled (dG)12–(dC)12 Template–Primer [SH-(dG)12–(dC)12-SH]

1. 0.1  M NaOH containing 10% acetonitrile: 4  g of NaOH. Add 0.9 L of deionized/filtered water. Add 100 mL of acetonitrile (Bio Lab, HPLC-S Gradient grade). 2. 5  M HCl solution: to 215  mL of deionized/filtered water add 285 mL of 32% HCL (Merck). 3. Alkaline 1 M NaCl solution containing acetonitrile: 58 g of NaCl. Add 0.9  L of 0.1  M NaOH and filter the solution through 0.22 mm Millipore Express PLUS membrane filter. Add 100  mL of acetonitrile (Bio Lab, HPLC-S Gradient grade). 4. 1 M Tris–acetate (pH 7.5): 121 g of Tris-base. Add 800 mL of deionized water; adjust the pH to 7.5 with glacial acetic acid and add deionized water to 1  L. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 5. 2 mM Tris–acetate (pH 7.5): to 998 mL of deionized/­filtered water add 2 mL of 1 M Tris–acetate (pH 7.5). 6. 20  mM Tris–acetate (pH 7.5): to 980  mL of deionized/­ filtered water add 20 mL of 1 M Tris–acetate (pH 7.5). 7. 0.4  M dl-dithiothreitol (DTT). 15.4  mg of DTT. Add 0.25 mL deionized/filtered water. Store at −18°C. 8. 5¢-(6-Mercapto-1-hexyl-phosphoric acid ester) of (dG)12 oligonucleotide, SH-(dG)12, from Alpha DNA (Montreal, Canada). 9. 5¢-(6-Mercapto-1-hexyl-phosphoric acid ester) of (dC)12 oligonucleotide, SH-(dC)12, from Alpha DNA (Montreal, Canada). 10. Dialysis tubing 10 mm (Sigma). Treat the tubing as follows: wash with running tap water for 3 h; treat with 0.3% (w/v)

120

Kotlyar

solution of sodium sulfide at 80°C for 1 min; wash with tap water for 2–5 min at 60°C; treat with 0.2% (v/v) solution of sulfuric acid for 10 min at room temperature; wash with tap water for 10–15 min. Store the tubing in 25% ethanol at 4°C. Rinse the tubing with running water and then with deionized/filtered water before use. 11. Ion-exchange PolyWax LP column (4.6 × 200  mm, 5  mm, 1,000 Å), Western Analytical Products. 12. Sephadex NAP-25 DNA-Grade (15 × 50 mm) (GE Healthcare).

prepacked

column

13. Ion-exchange HiTrap Q HP column (1 mL) (GE Healthcare). 14. Agilent 1100 HPLC system with a photodiode array detector. 15. Eppendorf table centrifuge (model 5424). 16. Laboratory Freeze Dryer Christ Alpha 1–4 (Osterode am Harz, Germany). 2.3. Synthesis of Poly(dG)–Poly(dC) 2.3.1. Enzymatic Synthesis of Poly(dG)–Poly(dC)

1. 5 M KOH solution: 140 g of KOH. Add 0.5 L of deionized water. 2. 1  M Phosphate buffer (pH 7.5): 136  g of KH2PO4. Add 700 mL of deionized water; adjust the pH to 7.5 with 5 M KOH and add deionized water to 1  L. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 3. 1 M MgCl2: 20.3 g of MgCl2·6H2O. Add 100 mL of deionized water. Filter the solution through 0.22  mm Millipore Express PLUS membrane filter. 4. 1 M EDTA: 29.2 g of Titriplex II (Merck). Add 100 mL of deionized water. Filter the solution through 0.22  mm Millipore Express PLUS membrane filter. 5. 0.4  M dl-dithiothreitol (DTT). 15.4  mg of DTT. Add 0.25 mL of deionized/filtered water. Store at −18°C. 6. 100 mM dCTP. Dissolve 23.2 mg of dCTP (Sigma) in 0.5 mL of deionized/filtered H2O. Store at −18°C. 7. 100 mM dGTP. Dissolve 25 mg of dGTP (Sigma) in 0.5 mL of deionized/filtered water. Store at −18°C. 8. 10  mM (dG)12–(dC)12 prepared as described below (see Subheading 3.1). 9. Klenow exo− (Klenow fragment of Escherichia coli DNA polymerase I, lacking the 3¢ → 5¢ exonuclease activity), 5  U/mL enzyme solution in glycerol from Fermentas (Lithuania). Store the solution at −18°C. 10. Dry bath incubator (MRC, Israel).

Synthesis of Long DNA-Based Nanowires 2.3.2. HPLC Purification of Synthesized Poly(dG)–Poly(dC)

121

1. 1  M acetic acid solution: to 470  mL of deionized/filtered water add 30 mL of 100% acetic acid (Merck). 2. 20  mM Tris–acetate (pH 8.0): 2.42  g of Tris-base (Fluka). Add 800 mL of deionized water; adjust the pH to 8.0 with 1  M acetic acid and add deionized water to 1  L. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 3. TSK-gel G-DNA-PW HPLC column (7.8 × 300  mm) (TosoHaas, Japan). 4. Agilent 1100 HPLC system with a photodiode array detector.

2.4. Synthesis of Thiol-End-Labeled Poly(dG)–Poly(dC), SH-Poly(dG)– Poly(dC)-SH 2.4.1. Enzymatic Synthesis of SH-Poly(dG)–Poly(dC)-SH

1. 5 M KOH solution: 140 g of KOH. Add 0.5 L of deionized water. 2. 1  M Phosphate buffer (pH 7.5): 136  g of KH2PO4. Add 700 mL of deionized water; adjust the pH to 7.5 with 5 M KOH and add deionized water to 1  L. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 3. 1 M MgCl2: 20.3 g of MgCl2·6H2O. Add 100 mL of deionized water. Filter the solution through 0.22  mm Millipore Express PLUS membrane filter. 4. 1 M EDTA: 29.2 g of Titriplex II (Merck). Add 100 mL of deionized water. Filter the solution through 0.22  mm Millipore Express PLUS membrane filter. 5. 0.4  M dl-dithiothreitol (DTT). 15.4  mg of DTT (Sigma). Add 0.25 mL deionized/filtered water. Store at −18°C. 6. 100 mM dCTP. Dissolve 23.2 mg of dCTP (Sigma) in 0.5 mL of deionized/filtered H2O. Store at −18°C. 7. 100 mM dGTP. Dissolve 25 mg of dGTP (Sigma) in 0.5 mL of deionized/filtered water. Store at −18°C. 8. 10  mM SH-(dG)12–(dC)12-SH prepared as described below (see Subheading 3.2). 9. Klenow exo− (Klenow fragment of E. coli DNA polymerase I, lacking the 3¢ → 5¢ exonuclease activity), 5  U/mL enzyme solution in glycerol from Fermentas (Lithuania). Store the solution at −18°C. 10. Dry bath incubator (MRC, Israel).

2.4.2. HPLC Purification of Synthesized SH-Poly(dG)– Poly(dC)-SH

1. Glacial acetic acid 100%. 2. 1 M Tris–acetate (pH 7.5): 121 g of Tris-base (Fluka). Add 800 mL of deionized water; adjust the pH to 7.5 with glacial acetic acid and add deionized water to 1 L. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 3. 2  mM Tris–acetate (pH 7.5): to 998  mL of deionized/ filtered water add 2 mL of 1 M Tris–acetate (pH 7.5).

122

Kotlyar

4. 20 mM Tris–acetate (pH 7.5): to 980 mL of deionized water add 20 mL of 1 M Tris–acetate (pH 7.5). 5. TSK-gel G-5000-PW HPLC column (7.8 × 300  mm), TosoHaas, Japan. 6. Agilent 1100 HPLC system with a photodiode array detector. 2.5. Synthesis of Poly(dG-dG)–Poly(dC) Triplex 2.5.1. Poly(dG-dG)– Poly(dC) Triplex Synthesis

1. 1  M KOH solution: 56  g of KOH. Add 1  L of deionized water. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 2. 1 M Phosphate buffer (pH 7.5): 136 g of KH2PO4 (Merck). Add 700 mL of deionized water; adjust the pH to 7.5 with 1 M KOH and add deionized water to 1 L. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 3. 1 M MgCl2: 20.3 g of MgCl2·6H2O. Add 100 mL of deionized water. Filter the solution through 0.22  mm Millipore Express PLUS membrane filter. 4. 1 M EDTA: 29.2 g of Titriplex II (Merck). Add 100 mL of deionized water. Filter the solution through 0.22  mm Millipore Express PLUS membrane filter. 5. 0.4  M dl-dithiothreitol (DTT). 15.4  mg of DTT. Add 0.25 mL deionized/filtered water. Store at −18°C. 6. 100 mM dGTP. Dissolve 25 mg of dGTP (Sigma) in 0.5 mL of deionized/filtered water. Store at −18°C. 7. 1 mM (in base pairs) 500–1,000 base pairs poly(dG)–poly(dC) prepared as described in Subheading 3.3. 8. Klenow exo− (Klenow fragment of E. coli DNA polymerase I, lacking the 3¢ → 5¢ exonuclease activity), 5  U/mL enzyme solution in glycerol from Fermentas (Lithuania). Store the solution at −18°C. 9. Dry bath incubator (MRC, Israel).

2.5.2. HPLC Purification of Synthesized Poly(dG-dG)–Poly(dC)

1. 1  M acetic acid solution: to 470  mL of deionized/filtered water add 30 mL of 100% glacial acetic acid (Merck). 2. 20  mM Tris–acetate (pH 7.5): 2.42  g of Tris-base. Add 800 mL of deionized water; adjust the pH to 7.5 with 1 M acetic acid and add deionized water to 1 L. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 3. TSK-gel G-DNA-PW HPLC column (7.8 × 300  mm), TosoHaas, Japan. 4. Agilent 1100 HPLC system with a photodiode array detector.

123

Synthesis of Long DNA-Based Nanowires

2.6. Synthesis of G4 (Quadruple)-DNA 2.6.1. Purification of (dC)20

1. 0.1 M NaOH solution: 4 g of NaOH. Add 1 L of deionized water. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 2. Glacial acetic acid 100% (Merck). 3. 1  M Phosphate buffer (pH 7.5): 136  g of KH2PO4. Add 700 mL of deionized water; adjust the pH to 7.5 with 1 M KOH and add deionized water to 1  L. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 4. 20  mM Phosphate buffer (pH 7.5) containing 10% acetonitrile; to 880 mL of deionized water add 20 mL of 1 M Phosphate buffer (pH 7.5) and filter the solution through 0.22  mm Millipore Express PLUS membrane filter. Add 100 mL of acetonitrile (Bio Lab, HPLC-S Gradient grade). 5. 0.5 M Phosphate buffer (pH 7.5) containing 10% acetonitrile; to 0.4 L of deionized water add 0.5 L of 1 M Phosphate buffer (pH 7.5) and filter the solution through 0.22 mm Millipore Express PLUS membrane filter. Add 100 mL of acetonitrile (Bio Lab, HPLC-S Gradient grade). 6. 1 M Tris–acetate (pH 7.5): 121 g of Tris-base (Fluka). Add 800 mL of deionized water; adjust the pH to 7.5 with glacial acetic acid and add deionized water to 1 L. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 7. Tris–acetate (pH 7.5): to 998 mL of deionized/filtered water add 2 mL of 1 M Tris–acetate (pH 7.5). 8. 10  mM (dG)12–(dC)12 Subheading 3.1).

prepared

as

described

(see

9. (dC)20 oligonucleotide from Alpha DNA (Montreal, Canada). 10. Dialysis tubing 10 mm (Sigma). Treat the tubing as follows: wash with running tap water for 3 h; treat with 0.3% (w/v) solution of sodium sulfide at 80°C for 1 min; wash with tap water for 2–5 min at 60°C; treat with 0.2% (v/v) solution of sulfuric acid for 10 min at room temperature; wash with tap water for 10–15 min. Store the tubing in 25% Ethanol at 4°C. Rinse the tubing with running deionized/filtered water before use. 11. Ion-exchange PolyWax LP column (4.6 × 200  mm, 5  mm, 1,000 Å), Western Analytical Products. 12. Sephadex NAP-25 DNA-Grade (15 × 50 mm), GE Healthcare.

prepacked

column

13. Ion-exchange HiTrap Q HP column (1 mL), GE Healthcare. 14. Agilent 1100 HPLC system with a photodiode array detector. 15. Eppendorf table centrifuge (model 5424). 16. Laboratory Freeze Dryer Christ Alpha 1–4 (Osterode am Harz, Germany).

124

Kotlyar

2.6.2. Poly(dG)–n(dC)20 Synthesis

1. 1  M KOH solution: 56  g of KOH. Add 1  L of deionized water. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 2. 1  M Phosphate buffer (pH 7.5): 136  g of KH2PO4. Add 700 mL of deionized water; adjust the pH to 7.5 with 1 M KOH and add deionized water to 1  L. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 3. 20 mM Phosphate buffer (pH 7.5): to 98 mL of deionized water add 2 mL of 1 M Phosphate buffer (pH 7.5). Filter the solution through 0.22  mm Millipore Express PLUS membrane filter. 4. 1 M MgCl2: 20.3 g of MgCl2·6H2O. Add 100 mL of deionized water. Filter the solution through 0.22  mm Millipore Express PLUS membrane filter. 5. 1 M EDTA: 29.2 g of Titriplex II (Merck). Add 100 mL of deionized water. Filter the solution through 0.22  mm Millipore Express PLUS membrane filter. 6. 0.4  M dl-dithiothreitol (DTT). 15.4  mg of DTT. Add 0.25 mL deionized/filtered water. Store at −18°C. 7. 100 mM dGTP. Dissolve 25 mg of dGTP (Sigma) in 0.5 mL of deionized/filtered H2O. Store at −18°C. 8. 10 mM (dG)12–(dC)12 template/primer prepared as described below (see Subheading 3.1). 9. HPLC-purified (dC)20 prepared as described below (see Subheading 3.6.1). 10. Klenow exo− (Klenow fragment of E. coli DNA polymerase I, lacking the 3¢ → 5¢ exonuclease activity), 5  U/mL enzyme solution in glycerol from Fermentas (Lithuania). Store the solution at −18°C. 11. Dry bath incubator (MRC, Israel).

2.6.3. HPLC Purification of Synthesized Poly(dG)–n(dC)20

1. 1  M acetic acid solution: to 470  mL of deionized/filtered water add 30 mL of 100% glacial acetic acid. 2. 20  mM Tris–acetate (pH 7.5): 2.42  g of Tris-base. Add 800 mL of deionized water; adjust the pH to 7.5 with 1 M acetic acid and add deionized water to 1 L. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 3. TSK-gel DNA-G-DNA PW HPLC column (7.8 × 300 mm) from TosoHaas, Japan. 4. Agilent 1100 HPLC system with a photodiode array detector.

Synthesis of Long DNA-Based Nanowires 2.6.4. Preparation of Poly(dG) Strands

125

1. 1 M NaOH solution: 40 g of NaOH. Add 1 L of deionized water. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 2. 0.1  M NaOH: to 900  mL of deionized/filtered water add 100 mL of 1 M NaOH. 3. TSK-gel DNA-G-DNA PW HPLC column (7.8 × 300 mm) from TosoHaas, Japan. 4. Agilent 1100 HPLC system with a photodiode array detector.

2.6.5. Preparation of G4-DNA

1. 1  M acetic acid solution: to 470  mL of deionized/filtered water add 30 mL of 100% glacial acetic acid. 2. 2  mM Tris–acetate (pH 8.0): 242  mg of Tris-base. Add 900 mL of deionized water; adjust the pH to 8.0 with 1 M acetic acid and add deionized water to 1 L. Filter the solution through 0.22 mm Millipore Express PLUS membrane filter. 3. Sephadex G-25 (NAP-5) prepacked DNA-Grade column (10 × 30 mm), GE Healthcare.

3. Methods 3.1. Preparation of (dG)12–(dC)12 Template–Primer 3.1.1. HPLC Purification of (dC)12

Complete purification of (dG)12 and (dC)12 oligonucleotides comprising a (dG)12–(dC)12 template–primer from shorter or longer oligonucleotides is required (see Notes 1 and 2). 1. Transfer ~1 mg of a dry oligonucleotide powder to 1.5 mL plastic tube. 2. Add 1 mL of deionized/filtered water. 3. Shake the sample and vortex vigorously for 2 min; incubate at room temperature for 30 min and vortex again. 4. Centrifuge the sample for 2  min at 5,000 × g at room temperature in order to get rid of insoluble compounds that might be present in the oligonucleotide preparation. 5. Transfer the entire supernatant to a new 1.5  mL plastic tube. 6. Connect an ion-exchange PolyWax LP column to the HPLC system. 7. Equilibrate the column with 50  mL of 20  mM Phosphate buffer (pH 7.5) containing 10% acetonitrile at a flow rate of 0.8 mL/min at room temperature. 8. Load 150 mL of the oligonucleotide sample at a flow rate of 0.8 mL/min. Do not overload the column; the large sample volume can significantly reduce the separation efficiency.

126

Kotlyar

Fig. 1. Purification of (dC)12 on a Poly WAX 300 Å column.

9. Elute the oligonucleotide in 10% acetonitrile with a linear Phosphate buffer gradient from 0.02 to 0.25 M for 30 min at a flow rate of 0.8 mL/min at room temperature. Monitor the elution by measuring absorbance at 260 nm. The elution profile is shown in Fig. 1. 10. Collect the fraction containing (dC)12 eluted between 33.5 and 35 min (as indicated by the arrows in Fig. 1). Total volume of the fraction should be ~1.0 mL. 11. Equilibrate the Sephadex G-25 DNA-Grade column with 30 mL of 2 mM Tris–acetate (pH 7.5) at room temperature. 12. Load 1 mL of the oligonucleotide sample obtained from the ion-exchange column (see step 10). Allow the sample to enter the column completely. Add 2.0 mL of Tris–acetate (pH 7.5). Allow the buffer to enter the column. 13. Place 2.0 mL plastic tube under the column; add 1.5 mL of 2 mM Tris–acetate (pH 7.5) and collect the eluant. 14. Transfer the sample into two 1.5 mL plastic tubes (1.0 mL per tube). 15. Freeze the sample in a dry ice/ethanol bath and lyophilize it to dryness. It takes approximately 15  h to completely lyophilize the sample. 16. Store the dry sample at −18°C. 3.1.2. HPLC Purification of (dG)12

1. Transfer ~1  mg of a dry oligonucleotide powder to plastic 1.5 mL plastic tube. 2. Add 1 mL of 0.1 M NaOH. 3. Shake the sample and vortex vigorously for 2 min.

Synthesis of Long DNA-Based Nanowires

127

Fig. 2. Purification of (dG)12 on a HiTrap Q HP column.

4. Centrifuge the sample for 2  min at 5,000 × g at room temperature in order to get rid of insoluble compounds that might be present in the oligonucleotide preparation. 5. Transfer the entire supernatant to a new 1.5  mL plastic tube. 6. Connect a HiTrap Q HP column to the HPLC system. 7. Equilibrate the column with 0.1  M NaOH containing 10% acetonitrile at a flow rate of 0.7  mL/min at room temperature. 8. Load 150 mL of the oligonucleotide sample at a flow rate of 0.7 mL/min. 9. Elute the oligonucleotide in 0.1  M NaOH containing 10% acetonitrile with a linear NaCl gradient from 0.5 to 1 M for 60 min at a flow rate of 0.7 mL/min at room temperature. Monitor the elution by measuring absorbance at 260  nm. The elution profile is shown in Fig. 2. 10. Collect the fraction eluted between 35 and 37 min (as indicated by the arrows in Fig. 2). Total volume of the fraction should be ~1.5 mL. 11. Equilibrate Sephadex G-25 DNA-Grade column with 30 mL of 2 mM Tris–acetate (pH 7.5) at room temperature. 12. Load 1.5  mL of the oligonucleotide sample obtained from the ion-exchange column (see step 10). Allow the sample to enter the column completely. Add 1.5  mL of 2  mM Tris– acetate (pH 7.5) buffer. Allow the buffer to enter the column. 13. Place 2.0  mL plastic tube under the column; add 2  mL of 2 mM Tris–acetate (pH 7.5) buffer and collect the eluant.

128

Kotlyar

14. Transfer the solution into two 1.5 mL plastic tubes (1.0 mL per tube). 15. Freeze the sample in a dry ice/ethanol bath and lyophilize it to dryness. It takes approximately 15  h to completely lyophilize the sample. 16. Store the dry sample at −18°C. 3.1.3. Annealing of Purified (dC)12 and (dG)12

1. Dissolve HPLC-purified (dC)12 obtained as described in Subheading 3.1.1 in 200 mL of 0.1 M NaOH. 2. Withdraw 10 mL from the sample and add to a quartz cuvette filled with 1 mL of 20 mM Tris–acetate (pH 7.5). 3. Measure absorption of 100-times diluted sample at 260 nm. 4. Calculate the concentration of the oligonucleotide in the sample using an extinction coefficient of 90  mM−1  cm−1 at 260 nm. For example, the absorption of 0.9 corresponds to (dC)12 concentration in the stock solution of 1 mM. 5. Dissolve HPLC-purified (dG)12 obtained as described in Subheading 3.1.2 in 200 mL of 0.1 M NaOH. 6. Withdraw 10 mL from the sample and add to a quartz cuvette filled with 1 mL of 20 mM Tris–acetate (pH 7.5). 7. Measure absorption of 100-times diluted sample at 260 nm. 8. Calculate the concentration of the oligonucleotide in the sample using an extinction coefficient of 120 mM−1 cm−1 at 260 nm. For example, the absorption of 0.6 corresponds to (dG)12 concentrations in the stock solution of 0.5 mM. 9. Mix proper volumes of (dG)12 and (dC)12 samples (see above) to obtain a solution having equal final concentrations of both the above oligonucleotides. The final volume of the mixture should be 200–400  mL and the concentration of (dG)12– (dC)12 should be in the range of 10–30 mM. 10. Transfer the mixture to dialysis tubing and dialyze against 1 L of 20  mM Tris–acetate (pH 7.5) for 2  h at room temperature. 11. Withdraw 100  mL from the dialyzed sample and add to a quartz cuvette filled with 0.9  mL of 20  mM Tris–acetate (pH 7.5). 12. Measure absorption of tenfold diluted sample at 260 nm. 13. Calculate the concentration of the template–primer in the sample using an extinction coefficient of 177 mM−1 cm−1 at 260 nm. For example, the absorption of 0.177 corresponds to (dG)12–(dC)12 concentration in the stock solution of 10 mM.

Synthesis of Long DNA-Based Nanowires

129

14. Transfer the (dG)12–(dC)12 sample into several 0.5 mL plastic tubes (0.1 mL per tube). 15. Freeze the samples in a dry ice/ethanol bath and store at −18°C until ready to proceed with the DNA synthesis. 3.2. Preparation of SH-(dG)12–(dC)12-SH Template–Primer

Complete separation of SH-(dG)12 and SH-(dC)12 oligonucleotides, comprising a SH-(dG)12–(dC)12-SH template–primer, from shorter or longer oligonucleotides and oligonucleotides not containing thiol-groups is required.

3.2.1. HPLC Purification SH-(dC)12

1. Transfer ~1  mg of a dry oligonucleotide powder to plastic 1.5 mL plastic tube. 2. Add 1 mL 20 mM Tris–acetate (pH 7.5). 3. Shake the sample and vortex vigorously for 2 min; incubate at room temperature for 30 min and vortex again. 4. Centrifuge the sample for 2  min at 5,000 × g at room temperature in order to get rid of insoluble compounds that might be present in the oligonucleotide preparation. 5. Transfer the entire supernatant to a new 1.5  mL plastic tube. 6. Add 25 mL of 0.4 M DTT. 7. Incubate the sample for 40 min at room temperature. 8. Connect an ion-exchange PolyWax LP column to the HPLC system. 9. Equilibrate the column with 20  mL of 2  mM Tris–acetate (pH 7.5) at a flow rate of 0.8  mL/min at room temperature. 10. Load 150 mL of the oligonucleotide sample at a flow rate of 0.8 mL/min. Do not overload the column; the large sample volume can significantly reduce the separation efficiency. 11. Elute the oligonucleotide with a linear NaCl gradient from 0 to 0.5 M for 60 min at a flow rate of 0.8 mL/min at room temperature. Monitor the elution by measuring absorbance at 260 nm. 12. Collect the fraction containing SH-(dC)12. Total volume of the fraction should be ~1.5 mL. 13. Equilibrate Sephadex G-25 DNA-Grade column with 30 mL of 2 mM Tris–acetate (pH 7.5) at room temperature. 14. Load 1.5  mL of the oligonucleotide sample obtained from the ion-exchange column (see step 12). Allow the sample to enter the column completely. Add 1.5  mL of 2  mM Tris– acetate (pH 7.5). Allow the buffer to enter the column.

130

Kotlyar

15. Place 2  mL plastic tube under the column; add 2.0  mL of Tris–acetate (pH 7.5) buffer and collect the eluant. 16. Transfer the solution into two 1.5 mL plastic tubes (1.0 mL per tube). 17. Freeze the sample in a dry ice/ethanol bath and lyophilize it to dryness. It takes approximately 15  h to completely lyophilize the sample. 18. Store the dry sample at −18°C. 3.2.2. HPLC Purification of SH-(dG)12

1. Transfer ~1  mg of a dry oligonucleotide powder to plastic 1.5 mL plastic tube. 2. Add 1 mL of 0.1 M NaOH. 3. Shake the sample and vortex vigorously for 2 min. 4. Centrifuge the sample for 2  min at 5,000 × g at room temperature in order to get rid of insoluble compounds that might be present in the oligonucleotide preparation. 5. Transfer the entire supernatant to a new 1.5  mL plastic tube. 6. Add 25 mL of 0.4 M DTT. 7. Incubate the sample for 30 min at room temperature. 8. Connect an ion-exchange HiTrap Q HP column to the HPLC system. 9. Equilibrate the column with 0.1 M NaOH at a flow rate of 0.7 mL/min at room temperature. 10. Load 150 mL of the oligonucleotide sample at a flow rate of 0.7 mL/min. 11. Elute the oligonucleotide in 0.1 M NaOH with a linear NaCl gradient from 0.5 to 1  M for 200  min at a flow rate of 0.7  mL/min at room temperature. Monitor the elution by measuring absorbance at 260 nm. 12. Collect the fraction containing SH-(dG)12. Total volume of the fraction should be ~3 mL. 13. Equilibrate Sephadex G-25 DNA-Grade column with 30 mL of 2 mM Tris–acetate (pH 7.5) at room temperature. 14. Load 3 mL of the oligonucleotide sample obtained from the ion-exchange column (see step 12). Allow the sample to enter the column completely. 15. Place 5 mL plastic tube under the column; add 3 mL of 2 mM Tris–acetate (pH 7.5) buffer and collect the eluant. 16. Transfer the solution into three 1.5 mL plastic tubes (1 mL per tube).

Synthesis of Long DNA-Based Nanowires

131

17. Freeze the sample in a dry ice/ethanol bath and lyophilize the sample to dryness. It takes approximately 15 h to completely lyophilize the sample. 18. Store the dry sample at −18°C. 3.2.3. Annealing of Purified SH-(dC)12 and SH-(dG)12

1. Dissolve HPLC-purified SH-(dC)12 obtained as described in Subheading 3.2.1 in 200 mL of 0.1 M NaOH. 2. Withdraw 10 mL from the sample and add to a quartz cuvette filled with 1 mL of 20 mM Tris–acetate (pH 7.5). 3. Measure absorption of 100-times diluted sample at 260 nm. 4. Calculate the concentration of the oligonucleotide in the sample using an extinction coefficient of 90  mM−1  cm−1 at 260 nm. For example, the absorption of 0.9 corresponds to SH-(dC)12 concentration in the stock solution of 1 mM. 5. Dissolve HPLC-purified SH-(dG)12 obtained as described in Subheading 3.2.2 in 200 mL of 0.1 M NaOH. 6. Withdraw 10 mL from the sample and add to a quartz cuvette filled with 1 mL of 20 mM Tris–acetate (pH 7.5). 7. Measure absorption of 100-times diluted sample at 260 nm. 8. Calculate the concentration of the oligonucleotide in the sample using an extinction coefficient of 120 mM−1 cm−1 at 260 nm. For example, the absorption of 0.6 corresponds to SH-(dG)12 concentration in the stock solution of 0.5 mM. 9. Mix proper volumes of SH-(dG)12 and SH-(dC)12 samples (see above) to obtain a solution having equal final concentrations of  both the above oligonucleotides. The final volume of the ­mixture should be 200–400  mL and the concentration of SH-(dG)12–(dC)12-SH should be in the range of 10–30 mM. 10. Add 2.5 mL of 0.4 M DTT per each 100 mL the mixture and incubate at room temperature for 30 min. 11. Transfer the sample (200–400 mL) to dialysis tubing and dialyze against 100 mL of 20 mM Tris–acetate (pH 7.5) buffer containing 2 mM DTT for 1 h at room temperature. Change the dialysis buffer and continue dialysis for one more hour at room temperature. 12. Withdraw 50 mL from the sample and add to a quartz cuvette filled with 1 mL of 20 mM Tris–acetate (pH 7.5). 13. Measure absorption of 20 times diluted sample at 260 nm. 14. Calculate the concentration of the template–primer in the sample using an extinction coefficient of 177 mM−1 cm−1 at 260 nm. For example, the absorption of 0.177 corresponds to SH-(dG)12–(dC)12-SH concentration in the stock solution of 20 mM.

132

Kotlyar

15. Transfer the SH-(dG)12–(dC)12-SH sample into several 0.5 mL plastic tubes (0.1 mL per tube). 16. Freeze the samples in a dry ice/ethanol bath and store at −18°C until ready to proceed with the DNA synthesis. 3.3. Preparation of Poly(dG)–Poly(dC) 3.3.1. Enzymatic Synthesis of Poly(dG)–Poly(dC)

The method described here is different from classical PCRmethods of DNA synthesis. It is based on the unique property of DNA Polymerase (Klenow exo− fragment) to extend blunt-ended poly(dG)–poly(dC) molecules in the presence of dGTP and dCTP (see Note 3). 1. Prepare the mix for the DNA synthesis. Combine the following reagents in a 0.5 mL microcentrifuge plastic tube for each reaction: 85.25 mL of deionized/filtered water, 6 mL of 1 M Phosphate buffer (pH 7.5), 0.5 mL of 1 M MgCl2, 1.5 mL of 100 mM dCTP, 1.5 mL of 100 mM dGTP and 1.25 mL of 0.4 M DTT, for a total volume of 0.1 mL (you may scale up or down accordingly). Mix well by vortexing. 2. Add 2  mL of (dG)12–(dC)12 template/primer. Mix well by vortexing. 3. Add 2 mL of Klenow exo−, mix well by vortexing and incubate the reaction at 37°C in an air dry bath for 1  h. One-hour incubation leads to synthesis of approximately 2,000 base pairs poly(dG)–poly(dC) molecules (see Fig.  3). You may change the amount of bases in the DNA accordingly by extending or reducing the incubation time. 4. Add 2 mL of 1 M EDTA to terminate the reaction and vortex the sample.

Fig. 3. Time course of poly(dG)–poly(dC) synthesis reaction. Polymerase extension assay was performed as described in Subheading 3.3.1 in the presence of 0.2 mM (dG)12–(dC)12 and 20 mg/ml of Klenow exo−; the incubation was set at 37°C. Aliquots were withdrawn each 15 min for 2 h 15 min. (a) The reaction products were resolved on 1% agarose gel and stained with ethidium bromide. The marker bands of 1 kb DNA ladder (lane 1) are indicated to the left. Time-dependent products for 15, 30, 45, 60, 75, 90, 105, 120, and 135 min of the synthesis (lanes 2–10). (b) Dependence of the polymer length (in kb) estimated from (a) on the time of synthesis.

Synthesis of Long DNA-Based Nanowires 3.3.2. HPLC Purification of Synthesized Poly(dG)–Poly(dC)

133

In order to completely separate synthesized Poly(dG)–Poly(dC) from nucleotides, the template–primer, Klenow exo−, and other reaction components of the synthesis we recommend to use sizeexclusion HPLC. 1. Connect TSK-gel G-DNA-PW HPLC column to the HPLC system (see Note 4). 2. Equilibrate the column with 20 mM Tris–acetate (pH 8.0) at a flow rate of 0.5 mL/min at room temperature. 3. Load 100  mL of poly(dG)–poly(dC) sample obtained as described above (Subheading  3.3.1) at a flow rate of 0.5 mL/min. 4. Elute the DNA in 20 mM Tris–acetate (pH 8.0) at a flow rate of 0.5 mL/min at room temperature. Monitor the elution by measuring absorbance at 260 nm. 5. Collect the DNA fraction from the column. Total volume of the fraction should be ~1 mL. 6. Withdraw 100 mL from the DNA sample and add to a quartz cuvette filled with 1 mL of 20 mM Tris–acetate (pH 8.0). 7. Measure absorption of tenfold diluted sample at 260 nm. 8. Calculate the concentration of the DNA in the sample using an extinction coefficient of 14.8 mM−1 cm−1 at 260 nm for a GC pair. For example, the absorption of 0.148 corresponds to GC concentration in the stock solution of 0.1 mM. 9. Transfer the solution into several 0.5  mL plastic tubes (0.1 mL per tube) (see Note 5). 10. Freeze the DNA samples in a dry ice/ethanol bath and store at −18°C.

3.4. Preparation of SH-Poly(dG)– Poly(dC)-SH 3.4.1. Enzymatic Synthesis of SH-Poly(dG)–Poly(dC)-SH

The method enables to obtain homogeneous population of SH-poly(dG)–poly(dC)-SH characterized by high affinity to gold surfaces and electrodes. 1. Prepare the mix for the DNA synthesis. Combine the following reagents in a 0.5 mL microcentrifuge plastic tube for each reaction: 67 mL of deionized water, 6 mL of 1 M Phosphate buffer (pH 7.5), 0.5 mL of 1 M MgCl2, 1.5 mL of 100 mM dCTP, 1.5 mL of 100 mM dGTP and 1.5 mL of 0.4 M DTT, for a total volume of 0.1  mL (you may scale up or down accordingly). Mix well by vortexing. 2. Add 20  mL of SH-(dG)12–(dC)12-SH template/primer. Mix well by vortexing. 3. Add 2 mL of Klenow exo−, mix well by vortexing and incubate the reaction at 37°C in an air dry bath for 1  h. One-hour incubation leads to synthesis of approximately 500 base pairs

134

Kotlyar

of SH-poly(dG)–poly(dC)-SH molecules. You may change the amount of bases in the DNA accordingly by extending or reducing the incubation time (see Note 6). 4. Add 2 mL of 1 M EDTA to terminate the reaction and vortex the sample. 3.4.2. HPLC Purification of Synthesized SH-Poly(dG)–Poly(dC)-SH

1. Connect TSK-gel G-DNA-PW HPLC column to the HPLC system. 2. Equilibrate the column with 20 mM Tris–acetate (pH 7.5) at a flow rate of 0.5 mL/min at room temperature. 3. Load 100 mL of SH-poly(dG)–poly(dC)-SH sample obtained as described above (Subheading  3.4.1) at a flow rate of 0.5 mL/min. 4. Elute the DNA in 20 mM Tris–acetate (pH 7.5) at a flow rate of 0.5 mL/min at room temperature. Monitor the elution by measuring absorbance at 260 nm. 5. Collect the DNA fraction from the column. Total volume of the fraction should be ~1 mL. 6. Withdraw 100 mL from the DNA sample and add to a quartz cuvette filled with 0.9 mL of 20 mM Tris–acetate (pH 7.5). 7. Measure absorption of tenfold diluted sample at 260 nm. 8. Calculate the concentration of the DNA in the sample using an extinction coefficient of 14.8 mM−1 cm−1 at 260 nm for a GC pair. For example, the absorption of 0.148 corresponds to GC concentration in the sample of 0.1 mM. 9. Transfer the solution into several 0.5  mL plastic tubes (0.1 mL per tube) (see Note 7). 10. Freeze the DNA samples in a dry ice/ethanol bath and store at −18°C.

3.5. Preparation of Poly(dG-dG)–Poly(dC)

The method of poly(dG-dG)–poly(dC) synthesis described here is based on the extension of the G-strand of the poly(dG)– poly(dC) by the Klenow exo− fragment of DNA polymerase I, under conditions when only the G-strand is allowed to grow.

3.5.1. Enzymatic Synthesis of Poly(dG-dG)–Poly(dC)

1. Prepare the mix for the DNA synthesis. Combine the following reagents in a 0.5 mL microcentrifuge plastic tube for each reaction: 74.7 mL of deionized water, 6 mL of 1 M Phosphate buffer (pH 7.5), 0.33 mL of 1 M MgCl2, 0.5 mL of 100 mM dGTP, and 1.5  mL of 0.4  M DTT, for a total volume of 0.1 mL (you may scale up or down accordingly). Mix well by vortexing. 2. Add 15  mL of 1  mM (in base pairs) poly(dG)–poly(dC) ­solution. Mix well by vortexing.

Synthesis of Long DNA-Based Nanowires

135

3. Add 2 mL of Klenow exo−, mix well by vortexing and incubate the reaction at 37°C in an air dry bath for 4 h. 4. Add 1 mL of 1 M EDTA to terminate the reaction and vortex the sample. 3.5.2. HPLC Purification of Synthesized Poly (dG-dG)–Poly(dC)

1. Connect TSK-gel G-DNA-PW HPLC column to the HPLC system. 2. Equilibrate the column with 20 mM Tris–acetate (pH 7.5) at a flow rate of 0.5 mL/min at room temperature. 3. Load 100 mL of poly(dG-dG)–poly(dC) sample obtained as described above (Subheading  3.5.1) at a flow rate of 0.5 mL/min. 4. Elute the triplex DNA in 20 mM Tris–acetate (pH 7.5) at a flow rate of 0.5 mL/min at room temperature. Monitor the elution by measuring absorbance at 260 nm. 5. Collect the DNA fraction from the column. Total volume of the fraction should be ~1 mL. 6. Withdraw 100 mL from the DNA sample and add to a quartz cuvette filled with 0.9 mL of 20 mM Tris–acetate (pH 7.5) buffer. 7. Measure absorption of tenfold diluted sample at 260 nm. 8. Calculate the concentration of the DNA in the sample using an extinction coefficient of approximately 20  M−1  cm−1 at 260 nm for a GGC triad. For example, the absorption of 0.2 corresponds to GGC concentration in the sample of 0.1 mM. 9. Transfer the solution into several 0.5  mL plastic tubes (0.1 mL per tube). 10. Freeze the DNA samples in a dry ice/ethanol bath and store at −18°C.

3.6. Preparation of G4 (Quadruple)-DNA 3.6.1. HPLC Purification of (dC)20

1. Transfer ~1  mg of a dry oligonucleotide powder to plastic 1.5 mL plastic tube. 2. Add 1 mL of deionized/filtered water. 3. Shake the sample and vortex vigorously for 2 min; incubate at room temperature for 30 min and vortex again. 4. Centrifuge the sample for 2  min at 5,000 × g at room temperature in order to get rid of insoluble compounds that might be present in the oligonucleotide preparation. 5. Transfer the entire supernatant to a new 1.5 mL plastic tube. 6. Connect an ion-exchange PolyWax LP column to the HPLC system.

136

Kotlyar

7. Equilibrate the column with 20  mL of 20  mM Phosphate buffer (pH 7.5), 10% acetonitrile, at a flow rate of 0.8 mL/min at room temperature. 8. Load 150 mL of the oligonucleotide sample at a flow rate of 0.8 mL/min. Do not overload the column; the large sample volume can significantly reduce the separation efficiency. 9. Elute the oligonucleotide with a linear gradient from 0.02 to 0.5  M Phosphate buffer (pH 7.5), 10% acetonitrile, for 60 min at a flow rate of 0.8 mL/min at room temperature. Monitor the elution by measuring absorbance at 260 nm. 10. Collect the DNA fraction from containing (dC)20. Total volume of the fraction should be ~1 mL. 11. Equilibrate Sephadex NAP-25 DNA-Grade column with 30 mL of 2 mM Tris–acetate (pH 7.5) at room temperature. 12. Load 1 mL of the oligonucleotide sample obtained from the ion-exchange column (see step 10). Allow the sample to enter the column completely. Add 2 mL of 2 mM Tris–acetate (pH 7.5). Allow the buffer to enter the column. 13. Place 2 mL plastic tube under the column; add 2 mL of 2 mM Tris–acetate (pH 7.5) buffer and collect the eluant. 14. Transfer the solution into four 1.5 mL plastic tubes (0.5 mL per tube). 15. Freeze the sample in a dry ice/ethanol bath and lyophilize the sample to dryness. It takes approximately 15 h to completely lyophilize the sample. 16. Store the dry sample at −18°C. 3.6.2. Enzymatic Synthesis of Poly(dG)–n(dC)20

Klenow exo− fragment of DNA Polymerase is capable of producing long double-stranded poly(dG)–n(dC)20 molecules composed of a long continuous dG-strand and relatively short dC-­ oligonucleotides not covalently connected to each other in the presence of dGTP and (dC)20. 1. Prepare the mix for the DNA synthesis. Combine the ­following reagents in a 0.5  mL microcentrifuge plastic tube for each ­reaction: 78.5 mL of deionized water, 6 mL of 1 M Phosphate buffer (pH 7.5), 0.5 mL of 1 M MgCl2, 1.5 mL of 100 mM dGTP and 1.5 mL of 0.4 M DTT, for a total volume of 0.1 mL (you may scale up or down accordingly). Mix well by vortexing. 2. Dissolve HPLC-purified (dC)20 obtained as described in Subheading  3.6.1 in 200  mL of 20  mM Phosphate buffer (pH 7.5). 3. Withdraw 1 mL from the sample and add to a quartz cuvette filled with 1 mL of 20 mM Phosphate buffer (pH 7.5).

Synthesis of Long DNA-Based Nanowires

137

4. Measure absorption of 1,000-times diluted sample at 260 nm. 5. Calculate the concentration of the oligonucleotide in the sample using an extinction coefficient of 144 mM−1 cm−1 at 260 nm. For example, the absorption of 0.144 corresponds to (dC)20 concentration in the sample of 1 mM. 6. Add 10  mL of 1  mM (dC)20 to a final concentration of 100 mM. 7. Add 2  mL of (dG)12–(dC)12 template/primer. Mix well by pipetting. 8. Add 2 mL of Klenow exo−, mix well by vortexing and incubate the reaction at 37°C in an air dry bath for 1  h. Two-hour incubation leads to synthesis of poly(dG)–n(dC)20 molecules composed of approximately 2,000 base-long G-strand. You may change the amount of bases in the DNA accordingly by extending or reducing the incubation time. 9. Add 2 mL of 1 M EDTA to terminate the reaction and vortex the sample. 3.6.3. HPLC Purification of Synthesized Poly(dG)–n(dC)20

1. Connect TSK-gel DNA-G-DNA PW HPLC column to the HPLC system. 2. Equilibrate the column with 20 mM Tris–acetate (pH 7.5) at a flow rate of 0.5 mL/min at room temperature. 3. Load 100  mL of poly(dG)–n(dC)20 sample obtained as described above (Subheading  3.6.2) at a flow rate of 0.5 mL/min. 4. Elute the DNA in 20 mM Tris–acetate (pH 7.5) at a flow rate of 0.5 mL/min at room temperature. Monitor the elution by measuring absorbance at 260 nm. 5. Collect the DNA fraction from the column. Total volume the fraction should be ~1 mL.

3.6.4. Preparation of Poly(dG)

At pH higher than 12.5 the poly(dG)- and the (dC)20-fragments composing poly(dG)–n(dC)20 are separated from each other and are eluted separately from the HPLC column. 1. Connect TSK-gel G-DNA-PW HPLC column to the HPLC system. 2. Equilibrate the column with 0.1 M NaOH solution at a flow rate of 0.5 mL/min at room temperature. 3. Transfer 100  mL of poly(dG)–n(dC)20 sample obtained as described above (Subheading 3.6.3) to a 0.5 mL microcentrifuge plastic tube. 4. Add 15 mL of 1 M NaOH and incubate for 10 min at room temperature.

138

Kotlyar

5. Load the sample onto the column at a flow rate of 0.5 mL/min. 6. Elute the DNA in 0.1 M NaOH at a flow rate of 0.5 mL/min at room temperature. Monitor the elution by measuring absorbance at 260 nm. 7. Collect a poly(dG) fraction eluted between 14 and 16 min. Total volume of the fraction should be ~1 mL. 3.6.5. Preparation of G4-DNA

Folding of G-strands into G4-structures is taking place spontaneously upon pH reduction during chromatography of the alkaline strands solution on a Sephadex G-25 column. 1. Equilibrate a Sephadex G-25 column with 5  mL of 2  mM Tris–acetate (pH 8.0) at room temperature. 2. Load 0.5 mL of the alkaline G-strand sample obtained from the TSK-gel G-DNA-PW HPLC column (see Sub­ heading  3.6.4). Allow the sample to enter the column ­completely. Add 0.2  mL of 2  mM Tris–acetate (pH 8.0). Allow the buffer to enter the column. 3. Place 1.5 mL plastic tube under the column; add 0.7 mL of 2 mM Tris–acetate (pH 8.0) buffer and collect the eluant. 4. Measure absorption of the sample at 260 nm. 5. Calculate the concentration of the DNA (in tetrads) in the sample using an extinction coefficient of 36  mM−1  cm−1 at 260 nm. For example, the absorption of 0.36 corresponds to 10 mM G4-DNA. 6. The sample can be stored for 2–3 days at 4°C. Longer storage is not recommended. Do not freeze the sample.

4. Notes 1. Complete purification of (dG)12, (dC)12, SH-(dG)12, SH-(dC)12 comprising a (dG)12–(dC)12 and SH-(dG)12– (dC)12-SH ­template–primers from shorter or longer oligonucleotides that are usually present in minor quantities in commercial preparations is required. If primed by nonpurified template–primers, the synthesis yields DNA molecules with large length variability. 2. Steps 8–13 in Subheadings 3.1.1 and 3.1.2 can be repeated several times in order to obtain larger quantities of purified oligonucleotides for a large-scale synthesis of the poly(dG)–poly(dC).

Synthesis of Long DNA-Based Nanowires

139

3. The protocol of poly(dG)–poly(dC) synthesis (see Subheading  3.3) can be adapted for synthesis of poly(dA)– poly(dT), a double-stranded polymer composed of poly(dA)and poly (dT)-homopolymer strands of equal length. 4. If the length of the synthesized poly(dG)–poly(dC) is shorter than 1  Kbp, use TSK-gel G-5000-PW HPLC column (7.8 × 300  mm, TosoHaas, Japan) instead of TSK-gel G-DNA-PW HPLC column (7.8 × 300  mm, TosoHaas, Japan) for the HPLC purification of the DNA. 5. The G-containing structures are unstable at low pH and undergo acid hydrolysis. We thus recommend storage of the DNA samples at pH 8–8.5 at 4°C. 6. The rate of SH-poly(dG)–poly(dC)-SH synthesis is slower than that of non-thiolated poly(dG)–poly(dC) polymer. Relatively large quantities of the enzyme should therefore be added into the assay in order to obtain long SH-poly(dG)– poly(dC)-SH molecules. 7. Thiol-groups can undergo spontaneous oxidation to disulfides. We recommend to store SH-poly(dG)–poly(dC)-SH in the presence of 1  mM DTT to avoid disulfides formation. The molecules should be separated from DTT prior to deposition on gold surfaces. This can be done by passing the DNA sample through Sephadex G-25 DNA-Grade column equilibrated with 2 mM Tris–acetate (pH 7.5).

Acknowledgments This work was supported by the EC through the contracts IST2001-38951 (“DNA-Based Nanowires”) and FP6-029192 (“DNA-Based Nanodevices”). References 1. Porath, D., Bezryadin, A., de Vries, S. and Dekker, C. (2000) Direct measurement of electrical transport through DNA molecules. Nature, 403, 635–638. 2. Hwang, J. S. K., Kong, J., Ahn, D. G., Lee, S., Ahn, D. J. S. and Hwang, W. (2002) Electrical transport through 60 base pairs of poly(dG)-poly(dC) DNA molecules. Appl. Phys. Lett., 81, 1134–1136. 3. Hennig, D., Starikov ,E. B., Archilla, J. F. R. and Palmero, F. (2004) Charge transport in poly(dG)-poly(dC) and poly(dA) poly(dT) DNA polymers. J. Biol. Phys., 30, 227–238.

4. Yi, J. (2003) Conduction of DNA molecules: A charge-ladder model. Physic. Rev. B. 68, 193103. 5. Lee, H.-Y., Tanaka, H., Otsuka, Y., Yoo, K.-H., Lee, J.-O. and Kawai,T. (2002) Control of electrical conduction in DNA using oxygen hole doping. App. Phys. Lett., 80, 1670–1672. 6. Yoo, K.-H., Ha, D.H., Lee, J.-O., Park, J.W., Kim, J., Kim, J.J., Lee, H.-Y., Kawai, T. and Choi, H.-Y. (2001) Electrical conduction through Poly(dA)–Poly(dT) and Poly(dG)– Poly(dC)DNAmolecules. Phys. Rev. Lett., 87, 198102.

140

Kotlyar

7. Kotlyar, A. B., Borovok, N., Molotsky T., Fadeev L., and Gozin M. (2005) In Vitro synthesis of uniform Poly(dG)-Poly(dC) by Klenow exo– fragment of Polymerase I. Nucl. Acid Res. 33, 525–535. 8. Nuzzo, R. G. and Allara, D. L. (1983) Adsorption of bifunctional organic disulfides on gold surfaces. J. Am. Chem. Soc., 105, 4481–4483. 9. Sellers, H., Ulman, A., Shnidman, Y. and Eilerss, J.E. (1993) Structure and binding of alkanethiolates on gold and silver surfaces: implications for self-assembled monolayers. J. Am. Chem. Soc., 115, 9389–9401. 10. Hegner, M., Wagner, P. and Semenza, G. (1993) Immobilizing DNA on gold via thiol modification for atomic force microscopy imaging in buffer solutions. FEBS Lett., 336, 452–456. 11. Frank-Kamenetskii, M. D. and Mirkin, S. M. (1995) Triplex DNA structures. Annu. Rev. Biochem., 64, 65–95. 12. Sun, J. S., Garestier, T. and Helene, C. (1996) Oligonucleotide directed triple helix formation. Curr. Opin. Struct. Biol., 6, 327–333. 13. Radhakrishnan, I. and Patel, D. J. (1994) DNA triplexes: solution structures, hydration sites, energetics, interactions, and function. Biochemistry, 33, 11405–11416. 14. Kotlyar, A. B., Borovok, N., Molotsky, T, Klinov, D., Dwir, B. and Kapon E. Synthesis of novel poly(dG)-poly(dG)-poly(dC) triplex structure by Klenow exo- fragment of DNA polymerase I. 2005 Nucl. Acid Res. 33, 6515–6521. 15. Kerwin, S. M. (2000) G-Quadruplex DNA as a target for drug design. Curr. Pharmaceutic. Design, 6, 441–478.

16. Davis, J. T. (2004) G-quartets 40 years later: From 50-GMP to molecular biology and supramolecular chemistry, Angew. Chem. Intl. Ed. 43, 668–698. 17. Keniry M. A. (2001) Quadruplex Structures in Nucleic Acids. Biopolymers, 56, 123–146. 18. Parkinson, G. N., Lee, M. P. and Neidle, S. (2002) Crystal structure of parallel quadruplexes from human telomeric DNA. Nature, 417, 876–880. 19. Burge, S., Parkinson, G. N., Hazel, P., Todd, A. K. and Neidle, S. (2006) Quadruplex DNA: sequence, topology and structure. Nucleic Acids Res., 34, 5402–5415. 20. Sen, D. and Gilbert, W. (1992) Novel DNA superstructures formed by telomere-like oligomers. Biochemistry, 31, 65–70. 21. Marsh, T. C., Vesenka, J. and Henderson, E. (1995) A new DNA nanostructure the G-wire imaged by scanning probe microscopy. Nucleic Acids Res., 23, 696–700. 22. Kotlyar, A. B., Borovok, N., Molotsky, T., Cohen, H., Shapir, E. and Porath, D. (2005) Long monomolecular guanine-based nanowires, Adv. Mater. 17, 1901–1905. 23. Borovok, N, Molotsky, T, Ghabboun, J, Porath, D. and Kotlyar, A. (2008) Efficient procedure of preparation and properties of long uniform G4-DNA nanowires. Anal. Biochem. 374, 71–78. 24. Cohen, H., Sapir, T., Borovok, N., Molotsky, T., Di Felice, R., Kotlyar, A. B. and Porath, D. (2007) Polarizability of G4-DNA observed by electrostatic force microscopy measurements. Nano Letters, 7, 981–986.

Chapter 10 G-Wire Synthesis and Modification with Gold Nanoparticle Christian Leiterer, Andrea Csaki, and Wolfgang Fritzsche Abstract DNA molecules are well known for containing the genetic information of an individual. Furthermore, DNA is a biopolymer with the potential of building up nanoscale structures. These structures can be addressed sequence specifically and, therefore, they allow connecting and arranging with subnanometer accuracy. The extended work of the group of Nadrian Seeman (Nature 421:427–431, 2003) has shown that the self-assembly of DNA molecules offers great potential for the creation of bottom-up nanostructures for nanoelectronics, biosensors, and programmable molecular machines. Rothemund (Nature 440:297– 302, 2006) has shown that it is possible to generate a wide variety of 2D nanostructures by the assembly of synthetic desoxyoligonucleotides and M13mp18 DNA via Watson–Crick base pairing. Furthermore, DNA can form three- and four-stranded structures which offer even more possibilities for molecular construction. This chapter will deal with four-stranded DNA structures (G-wires) created from 10-bp deoxynucleotide units. Our focus will be especially on the synthesis, individualization, modification with gold nanoparticles, and characterization by high-resolution scanning force microscopy (AFM). Key words: G-wire, Quadruplex, Gold nanoparticle, AFM, Nanowires

1. Introduction G-wires assemble thanks to DNA–DNA hybridization of four guanine (G)-nucleotides (3, 4). In nature, such G-quartet motifs can be found at the telomeres of chromosomes and in ribozymes where they provide the DNA with extra structural stability against enzymatic digestion and chemical reactions. Artificial structures based on this motif can be prepared either by intermolecular (5) or intramolecular (6) G-quartet formation using oligonucleotides containing predominantly (or solely) guanines. Hydration layer scanning tunnel microscopy (HLSTM) (7) imaging has suggested that G-wires might function as semiconductors (8) due to base stacking of the G-quartets and the caged monovalent cations. This potential for conduction

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_10, © Springer Science+Business Media, LLC 2011

141

142

Leiterer, Csaki, and Fritzsche

and the formation of long stable structures makes G-wire interesting not only for molecular construction, but also for nanoelectronic applications (9, 10). The oligonucleotide with the sequence 5¢-GGGGTTGGGG-3¢ was used for the assembling of G-wires in this protocol. The four guanines on both ends can hybridize with three other oligonucleotides each, resulting in a polymerization reaction that leads to long (several tens up to hundreds of nanometers) G-rich quadruplex structures (5). The polymerization reaction is controllable by hybridization temperature, time, and buffer composition (11). Using biotinylated oligonucleotides in combination with streptavidin-conjugated gold nanoparticles or positively char­ ged gold, nanoparticles were specifically attached either to the ends or unspecifically to the backbone of the G-wire (12).

2. Materials 2.1. Synthesis and Immobilization of G-Wire

1. 10-nt oligonucleotide with the sequence 5¢-GGGGTTGGGG-3¢, (G4), from Jena Bioscience GmbH (Jena, Germany). 2. Synthesis buffer (P1): 50  mM NaCl, 50  mM  Tris–HCl, pH 7.4, and 10 mM MgCl2 (see Note 1). 3. Dilution buffer (P2): 10 mM Tris–HCl, pH 7.4, and 1 mM MgCl2. 4. Mica sheets for AFM, Hi-Grade quality (Plano Planet GmbH, Wetzlar, Germany). 5. Photolithographically structured (13–15) microelectrode chips on thermally oxidized silicon (1  mm). A 3–5-nm titanium adhesion layer and a 200-nm gold layer were deposited by sputtering, followed by a lift-off process. 6. Chip activation solution: 1:1 ratio mix of 30% H2O2 (puriss p.a.) and 25% NH4OH (puriss p.a.).

2.2. Isolation and Growth of G-Wire

1. G-buffer: 50 mM KCl, 10 mM MgCl2, and 50 mM Tris–HCl, pH 7.4.

2.3. Surface Passivation and Gold Nanoparticle Modification of G-Wires Using GENOgold

1. Mica sheets for AFM, Hi-Grade quality (Plano Planet GmbH, Wetzlar, Germany). 2. 1× PBS: 137  mM NaCl, 10  mM phosphate, and 2.7  mM KCl, pH 7.4. 3. PEG solution: 1 M PEG. 4. MgCl2 solution: 5 mM MgCl2. 5. Mg–acetate solution: 1 M Mg(CH3COO)2. 6. Genogold (BBI, UK): 6°× 10 11  particles/ml.

17–23  nm

in

diameter,

G-Wire Synthesis and Modification with Gold Nanoparticle

2.4. Modification of Biotinylated G-Wire with Streptavidin-Modified Gold Nanoparticle

143

1. Biotinylated oligonucleotide [G4T2G4-biotin, Jena Bioscience GmbH (Jena, Germany)]. 2. Streptavidin-conjugated 1.6°× 1014 particles/ml).

gold

nanoparticles

(5 

nm,

3. Methods A first objective was to synthesize long and separated G-wires, in order to immobilize them between two electrodes for conductivity measurements or further DNA construction experiments. Previous work had shown that in most cases, G-wire could only be assembled in tight network or very small single wires. The next step would be to synthesize long single G-wire for further investigations. To achieve this goal, G-wires were immobilized on different surfaces in a G-wire compatible buffer. Furthermore, G-wires were reassembled after synthesis in a repeated heating/ cooling cycle in order to get longer single G-wire structures. A second objective was the specific binding of gold nanoparticles to G-wires for further optical and electrical experiments. To this end, two strategies were used. One was direct binding of gold nanoparticles to the backbone of the G-wires using the affinity of positively charged gold nanoparticles to the negatively charged DNA backbone. Alternatively, streptavidin-conjugated gold nanoparticles were bound to G4T2G4-biotin-modified oligonucleotides which were previously integrated in the G-wire structures. 3.1. Synthesis and Immobilization of G-Wire

1. Prior to the synthesis of the G-wire, the lyophilized oligonucleotide was resuspended in ddH2O (250 mg/ml) and heated to 95°C for preventing unspecific binding between the oligonucleotides. 2. Afterward, the oligonucleotides were diluted in synthesis buffer (P1) (1:2) and incubated at 37°C for 4 days to assemble the G-wire structures. The dilution was covered with mineral oil to avoid evaporation and to keep a constant concentration during the whole assembly process. 3. After assembling, the G-wires were diluted in dilution buffer (P2) (1:20, final oligo concentration 6.25 ng/ml) and stored for up to 2 days at room temperature for immediate use, or at −20°C for storage. 4. For the immobilization of G-wires on mica sheets, the mica surface was activated by cleavage just before spreading (see Note 2).

144

Leiterer, Csaki, and Fritzsche

5. For immobilization on silicon (thermally oxidized), the surface was activated either by incubation in chip activation solution for 1  h at 70°C or by oxygen plasma etching at 50 W for 12 min. 6. On both surfaces (mica and oxidized silicon), the G-wire dilution was incubated for at least 4 min to achieve binding to the surface (see Note 3). Afterward, the mica sample was washed with 1 ml of ddH2O in order to remove remaining salt from the surface (see Note 4). 7. The G-wire can be visualized by AFM imaging in air, using a Nanoscope III with a Dimension 3100 AFM head. All images were taken in tapping mode™. In order to get more stable images by minimizing surface charges, the samples were stored at up to 1 week (see Note 5, Fig. 1). 3.2. Isolation and Growth of G-Wire

1. For the isolation and growth of the synthesized G-wire, the salt concentration was raised from synthesis buffer level to the G-buffer level for extra stabilization of the G-wire (see Note 6). 2. The G-wire suspension was diluted 1:2 and 1:4 (final oligo concentration 3 and 1.5 ng/ml, respectively) in order to get single wires. 3. For the growth of the G-wire, the suspension was heated and cooled (95°C/55°C) for seven cycles in order to get long G-wire structures (see Note 7). 4. For immobilization on mica, the mica surface is activated by cleavage just prior to spreading. 5. The G-wire dilution was incubated for at least 4  min to achieve the binding to the mica surface. 6. The samples were stored at room temperature for a few days in order to get dry and non-charged surface. Visualization of the G-wire was performed by AFM imaging in air, using a Nanoscope III with a Dimension 3100 AFM head. All images were taken in tapping mode™. In order to get more stable images by minimizing surface charges, the samples were stored up to 1 week (see Fig. 2).

3.3. Surface Passivation and Gold Nanoparticle Modification of G-Wires Using Genogold

1. To enable specific binding of gold nanoparticle to the G-wire, the surface must be passivated in order to prevent unspecific nanoparticle binding to the surface. 2. The mica surface was activated by cleavage just before use. 3. Then the surface was incubated in four different buffers: 1× PBS, PEG solution, MgCl2 solution, and Mg–acetate solution to saturate the negative charges on the surface. The surface was then washed with ddH2O and dried with gaseous N2.

G-Wire Synthesis and Modification with Gold Nanoparticle

145

Fig. 1. Immobilized G-wire on different surfaces and different conditions visualized by AFM imaging (all in P1/P2 mixture). (a) Negative control: AFM micrograph of the (G4T2G4) oligonucleotide spread on mica before G-wire synthesis; the oligonucleotide was stored at −20°C and 6.25 ng/ml. Only very short G-wire structures are evident. (b) AFM micrograph of the (G4T2G4) oligonucleotide spread on mica before G-wire synthesis; the oligonucleotide was stored at −20°C and 25 ng/ml. A carpet of G-wire formed on the mica surface. (c) Synthesized G-wires (made from 6.25 ng/ml oligonucleotide) immobilized on mica; these were forming a very tight arrangement of individual G-wire structures with few interconnections. (d) Same as (c) but immobilized on thermally oxidized silicon; such conditions lead to a significantly reduced number of immobilized G-wires compared to same conditions on mica.

4. The G-wire dilution was incubated on the mica surface for 4  min at room temperature and washed with ddH2O afterward. 5. Genogold was incubated for 15 min at room temperature on the surface and subsequently washed with ddH2O and dried with gaseous N2 (see Note 8).

146

Leiterer, Csaki, and Fritzsche

Fig. 2. Effect of heating cycle on the growth of the G-wires. G-wires were immobilized on mica visualized by AFM imaging. Changes in buffer concentration have been employed in order to get longer individual G-wire structures. (a) 4 days, 37°C synthesis at standard concentration (6.25 ng/ml) in P1/P2, single wires and network structures were formed. (b) Enlarging of G-wires synthesized for 4 days (cf. Fig. 2a) in P1/P2, using a heating/cooling cycle. Long network-like structures were formed probably due to re-hybridization of unstable loop endings. (c) 4 days, 37°C synthesized G-wires (6.25 ng/ml) in G-buffer were forming a tight network with hardly any visible end. Probably high salt concentration leads to stronger G-wire hybridization. (d) G-wires (3 ng/ml) were synthesized for 4 days at 37°C (same Fig. 2c) in G-buffer. Now the network-like structures visible in Fig. 2c are destroyed, and short single G-wires become visible. (e) G-wires (1.5 ng/ml) were synthesized for 4 days at 37°C in G-buffer, resulting in very short single G-wires at the surface.

6. Visualization of the nanoparticle-modified G-wire was realized by AFM imaging in air after a few days when all surface charges were neutralized (see Fig. 3). 3.4. Modification of Biotinylated G-Wire with Streptavidin-Modified Gold Nanoparticle

1. Biotinylated oligonucleotides (G4T2G4-biotin) were added to the previously synthesized G-wire at a concentration ranging between 10 and 100  mM (to achieve different degrees of modification) and stored at 37°C to insert a biotin modification in the G-wire. 2. For the modification of the biotinylated G-wire with streptavidin-conjugated gold nanoparticle, the dilution was incubated at room temperature for a few minutes. 3. To immobilize the nanoparticle-modified G-wire on mica, the mica surface was activated by cleavage.

G-Wire Synthesis and Modification with Gold Nanoparticle

147

Fig. 3. Binding of Genogold to G-wires immobilized on mica. Different buffers were used for the saturation of surface charges in order to passivate the surface against unspecific binding of gold nanoparticle. (a) 1× PBS-passivated surface with no G-wire visible. Gold nanoparticles were immobilized to the surface. (b) 1 M PEG-passivated surface, very short G-wires were immobilized. Gold nanoparticles were immobilized to the surface due to unspecific binding. (c) 5  mM MgCl2passivated surface. Short G-wire could be immobilized to the surface. Very few gold nanoparticles were bound unspecifically to the surface. Nearly no gold nanoparticle was bound to the G-wire. (d) 1 M Mg(CH3COO)2-passivated surface, G-wire network could be immobilized on the surface. Gold nanoparticles were incorporated in the G-wire network. Black arrows point to G-wires, white arrows point to gold nanoparticles.

4. The G-wire dilution was incubated for 4  min to provide binding to the mica surface. 5. The prepared samples were stored for drying and discharging up to 1 week. The samples were characterized by AFM imaging in air, using a Nanoscope III with a Dimension 3100 AFM head. All images were taken in tapping mode™. In order to get more stable images by minimizing surface charges, the samples were stored up to 1 week (see Fig. 4).

148

Leiterer, Csaki, and Fritzsche

Fig. 4. Streptavidin–Au-modified biotinylated G-wire on mica. G-wires were synthesized from a mixture of non-biotinylated (G4T2G4, 120 mM) and biotinylated oligonucleotides (G4T2G4-biotin) in different concentrations (10–100 mM). Afterward, streptavidin-conjugated gold nanoparticles were bound to the biotin modification on the G-wire. (a) 10 mM biotinylated oligonucleotides used for the synthesis of G-wire. The immobilized G-wire structures forming a tight network due to the small fraction (1/12) of biotinylated oligonucleotides which are causing a break in the polymerization reaction. A few streptavidin–Au nanoparticles were incorporated, (b) 20 mM biotinylated oligonucleotides, corresponding to 1/6 of the overall oligonucleotide concentration, forming a wide network of G-wires. More streptavidin–Au nanoparticles get incorporated due to the higher fraction of biotinylated oligonucleotides. (c) 50 mM biotinylated oligonucleotides (corresponding to 1/3) causing the beginning of dissociation of the network and the formation of big streptavidin–Au clusters in the G-wire network. (d) 100 mm biotinylated oligonucleotides, corresponding to a mixture of equal amounts of biotinylated and non-biotinylated oligonucleotides, do not form long G-wire structures. The biotin ending of the oligonucleotide probably hinders the polymerization reaction. Streptavidin–Au nanoparticles are evenly spread on the surface. Black arrows point to G-wires, white arrows point to gold nanoparticles.

4. Notes 1. All solutions were prepared in deionized water (ddH2O with a conductivity less than 20  mS/cm, no special filtering was necessary). 2. Cleavage was done by removing the upper mica layers from the mica sheets with a sticky tape. The result is a homogenous negatively charged and very planar surface.

G-Wire Synthesis and Modification with Gold Nanoparticle

149

3. The surface binding of G-wire is arranged by bivalent cations such as Mg2+. 4. Remaining salt from the buffer solution results in the growth of crystals on the surface, which hampers the AFM imaging; therefore, the crystals have to be removed. 5. Due to the previous surface activation, the surface remains charged even after immobilization. These remaining charges destabilize AFM imaging. It has been shown that the easiest way to get rid of the surface charges is storing the samples up to 1 week at room temperature and medium relative humidity (30% < x < 80%). 6. Higher concentration (up to 1 M max.) of small monovalent and bivalent cations, such as Na+, Mg2+, and K+, refers to higher G-wire stability. Therefore, the salt concentration was raised to prevent possibly destabilizing effects from low salt concentrations. 7. G-wire, in contrary to double-stranded DNA, do not denature at 95°C. The goal was to denature possibly unstable loops, which might prevent G-wire growth after a while during the synthesis process, on the ends of the assembled G-wires. 8. Genogold is a colloidal particle solution produced by BBI. The gold nanoparticles have an average size of about 20-nm diameter. Genogold has a high affinity to DNA due to its positively charged surface. It is mainly used for DNA staining in plotting techniques.

Acknowledgments We thank J. Vesenka for an introduction in G-wire and a longterm collaboration; M. Sossna, A. Ihring, and H. Porwohl for the chip preparation; and A. Wolff, C. Holste, and A. Sondermann for preparatory works. Support by EU (FP6-NMP STREP013775, NUCAN). References 1. Seeman, N. C. (2003) DNA in a material world Nature 421, 427–431. 2. Rothemund, P. W. (2006) Folding DNA to create nanoscale shapes and patterns Nature 440, 297–302. 3. Williamson, J. R., Raghuraman, M. K., Cech, T. R. (1989) Monovalent cation-induced structure of telomeric DNA: The G-quartet model Cell ; Vol/Issue: 59:5, Pages: 871–880. 4. Williamson, J. R. (1993) G-quartets in biology: reprise Proceedings of the National Academy

of Sciences of the United States of America 90, 3124-. 5. Marsh, T. C., Vesenka, J., Henderson, E. (1995) A new DNA nanostructure, the G-wire, imaged by scanning probe microscopy Nucleic Acids Res 23, 696–700. 6. Borovok, N., Molotsky, T., Ghabboun, J., Porath, D., Kotlyar, A. (2008) Efficient procedure of preparation and properties of long uniform G4-DNA nanowires Analytical Biochemistry 374, 71–78.

150

Leiterer, Csaki, and Fritzsche

7. Heim, M., Eschrich, R., Hillebrand, A., Knapp, H. F., Guckenberger, R., Cevc, G. (1996) Scanning tunneling microscopy based on the conductivity of surface adsorbed water. Charge transfer between tip and sample via electrochemistry in a water meniscus or via tunneling? J Vac Sci Technol B 14, 1498–1502. 8. Muir, T., Morales, E., Root, J., Kumar, I., Garcia, B., Vellandi, C., Jenigian, D., Marsh, T., Henderson, E., Vesenka, J. (1998) The morphology of duplex and quadruplex DNA on mica Papers from the 44th national symposium of the AVS 16, 1172–1177. 9. Rinaldi, R., Branca, E., Cingolani, R., Masiero, S., Spada, G. P., Gottarelli, G. (2001) Photodetectors fabricated from a self-assembly of a deoxyguanosine derivative Applied Physics Letters 78, 3541–3543. 10. Calzolari, A., Di Felice, R., Molinari, E., Garbesi, A. (2002) G-quartet biomolecular nanowires Applied Physics Letters 80, 3331–3333.

11. Sondermann, A., Holste, C., Moller, R., Fritzsche, W. Assembly Of G-Quartet Based DNA Superstructures (G-Wires). In: DNABASED MOLECULAR CONSTRUCTION: International Workshop on DNA-Based Molecular Construction; 2002; Jena (Germany): AIP; 2002. p. 93–98. 12. Holste, C., Sondermann, A., Moller, R., Fritzsche, W. Coupling G-Wires To Metal Nanoparticles. In: DNA-BASED MOLE­ CULAR ELECTRONICS: Inter­national Symposium on DNA-Based Molecular Elec­ tronics; 2004; Jena (Germany): AIP; 2004. p. 53–58. 13. Fuller, G. E. Handbook of semiconductor manufactoring technology. In: Dekker M, ed. New York; 2000:461. 14. Moreau, W. M. Semiconductor Lithography. New York: Plenium Press; 1988. 15. Bowden, M. J. Introduction to Microlithography. In: Thomson LF, Willson CG, Bowden MJ, eds. Washington: ACS; 1994.

Chapter 11 Preparation of DNA Nanostructures with Repetitive Binding Motifs by Rolling Circle Amplification Edda Reiß, Ralph Hölzel, and Frank F. Bier Abstract A long one-dimensional single-stranded DNA (ssDNA) molecule with a periodic sequence motif is an attractive building block for DNA nanotechnology because it allows the positioning of oligonucleotidelabeled particles or molecules with high spatial resolution via molecular self-assembly simply by hybridization reactions. In vitro enzymatic isothermal rolling circle amplification (RCA) produces such long concatemeric ssDNA molecules. These are complementary in sequence to their circular template. In this chapter, the preparation of stretched and surface-attached RCA products at the single molecule level is described. The methods presented comprise the enzymatic circularization of a ssDNA oligonucleotide, the covalent coupling of amino-modified primers to carboxylated fluorescence beads, the preparation of a hydrophobic glass substrate, the RCA in a flow-through system, the postsynthetic staining and stretching of the RCA products as well as the microscopic observation of individual ssDNA molecules. Key words: Rolling circle amplification, DNA nanostructure, Single-stranded DNA, Fluorescence microscopy, SYBR Green II, Phi29 DNA polymerase

1. Introduction In the mid-1990s it was discovered that certain polymerases are capable of using circular single-stranded DNA (ssDNA) molecules as a template for enzymatic in vitro DNA synthesis (1, 2). This enzymatic reaction, which has later been termed rolling circle amplification (RCA), produces long (>10 kb) ssDNA molecules with repetitive binding motifs which are complementary in sequence to their circular template molecule (2). The RCA reaction has been applied in genetic analysis and for signal amplification purposes, e.g., for mutation detection (3) or on microarray platforms (4). More recently, RCA has been applied as a tool in DNA nanotechnology (5–10), because the concatemeric Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_11, © Springer Science+Business Media, LLC 2011

151

152

Reiß, Hölzel, and Bier

ssDNA molecule which is generated by the RCA reaction is an attractive building block for the bottom-up construction of DNA nanostructures. The repetition of the binding motifs and especially the fact that they are of single-stranded nature allow the arrangement of nano-objects with high spatial resolution simply by molecular self-assembly via hybridization. Among others it has been shown that gold nanoparticles to which oligonucleotides have been coupled can be hybridized to their complementary binding sites on the RCA product (6, 7). In this chapter, a method is presented, in which RCA is carried out at individual surfaceanchored primers in a flow-through system (see Fig. 1 for an illustration). The protocols provided in this chapter describe the procedures that are necessary for the reproduction of the experiment in a detailed way. First the circularization of a 5′-end phosphorylated linear oligonucleotide (88 bases) using the enzyme CircLigase ssDNA ligase is described (see Subheading  3.1). Circular templates for RCA-applications can be prepared either by  chemical synthesis (11, 12) or more commonly by enzymatic ligation (12). In enzyme-based circularization protocols, usually short ssDNA oligonucleotides (splints) are employed in order to form double-helical splint complexes to juxtapose the 3′-hydroxyl and the 5′-phosphate ends of the linear precursor ssDNA molecules that are to be circularized (12). Thus the formation of a phosphodiester bond by T4 DNA ligase (i.e., circularization) of the ­linear precursor is enabled. In contrast to the ligation with T4 DNA ligase, no splint is needed, when the enzyme CircLigase ssDNA ligase is used for enzymatic circularization of the linear precursor oligonucleotide as described in this chapter. According to the manufacturer of the enzyme there are almost no linear or circular

Fig. 1.  Schematic drawing of enzymatic RCA starting at a bead-bound primer. (a) fluorescence bead; (b) primer; (c) synthesized RCA product; (d) Phi29 DNA polymerase; (e) circular template; (f) nBTCS-modified cover slip. Round arrow indicates synthesis direction.

Preparation of DNA Nanostructures with Repetitive Binding Motifs

153

concatemers produced under standard reaction conditions (13). Compared to the aforementioned circularization strategies the use of this special enzyme simplifies the circularization reaction significantly. Furthermore, a protocol for the covalent attachment of amino-modified DNA primers to 20  nm carboxylated fluorescence beads is provided under the Subheading  3.3. The annealing of the circular template to the primer-modified beads is described under the Subheading  3.4 and the adsorption of the beads with the preannealed template onto a hydrophobic cover slip under Subheading 3.6. After the cover slip has been mounted into the flow-through system, the RCA is carried out (see Subheading 3.7). In order to assure an efficient RCA, we employ the enzyme Phi29 DNA polymerase. This enzyme was originally isolated from the Bacillus subtilis bacteriophage Phi29. It combines the qualities of a high processivity with good strand ­displacement properties and does not need additional proteins when applied in in  vitro DNA replication reactions (14). The RCA reaction starting from the bead-immobilized templates produces long concatemeric ssDNA molecules that are stained with SYBR Green II and then stretched and attached to the surface by the passage of a water/air interface under microscopic observation (see Subheading 3.8). We could also show that the stretched and surface-attached RCA products prepared by our protocol are still accessible to hybridization reactions (15).

2. Materials Ultrapure water (conductivity 0.055 mS/cm; DNase/RNase free) is used for all buffer preparations and dilutions. Aseptic techniques are not needed for these protocols, but DNase-free polypropylene plasticware should be used in combination with clean handling of all materials throughout the protocols. 2.1. Preparation of Circular ssDNA Template

1. 5′-end phosphorylated oligonucleotide (termed cF5cF10c F9cF1): 5′-phosphate-CTT ATC GCT TTA TGA CCG GAC CCG TGT AGC CTT TGT ATT CGT CCC TTC ACG ATT GCC ACT TTC CAC GTA CTT CCT TAA ACG ACG CAG G-3′ (HPLC-purified grade; Thermo Electron, Ulm, Germany) (16). 2. CircLigase ssDNA ligase (100  U/mL; Epicentre Biotech­ nologies, Madison, WI, USA). Store at −20°C. Keep on ice when used. 3. CircLigase ssDNA ligase 10× reaction buffer (supplied with the enzyme): 500  mM MOPS (pH 7.5), 100  mM KCl, 50 mM MgCl2, 10 mM DTT.

154

Reiß, Hölzel, and Bier

4. 1 mM ATP, 50 mM MnCl2, sterile water (all supplied with the CircLigase enzyme) (see Note 1). 5. Escherichia coli exonuclease I (20  U/mL; Fermentas, St. Leon-Rot, Germany). 6. Qiagen Nucleotide Removal Kit (Qiagen, Hilden, Germany). 7. NanoDrop ND-1000 Spectrophotometer (Thermo Scientific, Wilmington, DE, USA). 2.2. Denaturing Polyacrylamide Gel Electrophoresis of Circularized DNA Template

1. 30% (w/v) Acrylamide/bis-acrylamide solution (mix ratio 19:1, electrophoresis grade; Sigma-Aldrich, Taufkirchen, Germany). Acrylamide is a neurotoxin. Work under a fume hood, wear suitable chemical-resistant gloves and safety glasses, do not inhale vapor, avoid contact with skin. Store at 2–8°C. 2. 10% (w/v) Ammonium persulfate (electrophoresis grade; Sigma-Aldrich) in water. Always prepare solution freshly. 3. N,N,N ′,N ′-tetramethylethylenediamine (TEMED; SigmaAldrich). Wear suitable chemical-resistant gloves and safety glasses when handling this chemical, do not inhale vapor. TEMED is sensitive to air and humidity and should be stored under inert gas. 4. Urea (analysis grade; Applichem, Darmstadt, Germany). 5. 10× TBEgel-buffer: 0.89 M Tris–borate, 0.025 M ethylenediaminetetraacetic acid (EDTA); pH 8.3 (for electrophoresis) or pH 8 (for SYBR Green II staining). Buffer should be 0.2 mm filtered (ME 24 membrane filter, mixed cellulose ester; Whatman, Dassel, Germany) into a clean glass bottle and can be stored at room temperature. Dilute stock solution with water to 1× TBEgel-buffer when needed. 6. Urea loading buffer: 8  M Urea, 20  mM EDTA, 5  mM Tris–HCl (pH 7.5), 0.5% (w/v) xylene cyanol FF, 0.5% (w/v) bromphenol blue (17). 7. 6× Non-denaturing loading buffer: 10  mM Tris–HCl (pH 7.6), 0.03% (w/v) bromophenol blue, 0.03% (w/v) xylene cyanol FF, 60% (v/v) glycerol, 60 mM EDTA. 8. Hyperladder Germany).

V

(DNA

ladder;

Bioline,

Luckenwalde,

9. 10,000× SYBR Green II stock solution in DMSO (Cambrex Bio Science, Rockland, ME, USA). Store at −20°C. Although SYBR Green is regarded to be less mutagenic than ethidium bromide, wear DMSO-resistant gloves and treat chemical as mutagenic. Always keep SYBR Green solutions protected from light.

Preparation of DNA Nanostructures with Repetitive Binding Motifs

2.3. Covalent Coupling of Amino-Modified Oligonucleotides to 20 nm Carboxylated Microspheres

155

All buffers used in this protocol are 0.2 mm filtered (Puradisc 30 Syringe Filter, cellulose acetate; Whatman) and stored aliquoted at −20°C. 1. FluoSpheres carboxylate-modified microspheres, 0.02  mm, crimson fluorescent (625/645), 2% solids (Invitrogen, Karlsruhe, Germany). 2. 5′-end amino-modified, 3′-end phosphothioate(*)-modified oligonucleotide (termed F5): 5′NH2C12-GGT CCG GTC ATA AAG CGA TA*A*G-3′ [HPLC-purified grade; resuspended to 100  mM in ultrapure water (pH adjusted to 7); Thermo Electron, Ulm, Germany] (16) (see Note 2). 3. N-(3-dimethylaminopropyl)-N ′-ethylcarbodiimide hydrochloride (EDC, commercial grade; Sigma-Aldrich). EDC is sensitive to humidity. Store in small portions and desiccated at −20°C. Allow to come to room temperature before usage. 4. N-hydroxysulfosuccinimide sodium salt (Sulfo-NHS, puriss.; Fluka/Sigma-Aldrich). Sulfo-NHS is sensitive to humidity. Store desiccated at room temperature. 5. MES-buffer: 100 mM 2-(N-morpholino)ethanesulfonic acid, pH 4.5. 6. 1× TE-buffer: 10 mM Tris–HCl, 1 mM EDTA; pH 8. 7. 10% (v/v) Tween 20 in H2O. 8. 0.02% (v/v) Tween 20 in H2O and 0.02% (v/v) Tween 20 in 1× TE-buffer. Prepare prior to use by diluting the 10% (v/v) Tween 20 stock solution 1:500 in H2O or in 1× TE-buffer. 9. 1 M Glycin solution in water. 10. Ultrafree-MC centrifugal filter units (Ultracel PL regenerated cellulose membrane, NMWL 30000; Millipore, Schwalbach/ Taunus, Germany).

2.4. Annealing of Circular ssDNA Template to OligonucleotideModified Microspheres 2.5. Preparation of n-Butyltrichlorosilane-Modified Cover Slips

3× Annealing buffer: 30 mM Tris–HCl, 150 mM NaCl, 3 mM EDTA; pH 8. Filter buffer 0.2  mm (Puradisc 30 Syringe Filter, cellulose acetate; Whatman) and store aliquoted at −20°C.

1. Microscope cover slips 15 × 15  mm, strength # 2 (MenzelGläser, Braunschweig, Germany). 2. Piranha cleaning solution: one part hydrogen peroxide (30%, p.a.; Carl Roth, Karlsruhe, Germany) and three parts sulfuric acid (96%, p.a.; Carl Roth). Always add the peroxide slowly to the sulfuric acid under mild stirring (PTFE-coated stir bar). A strongly exothermic reaction occurs and the solution heats up. Mixing piranha solution with organic compounds may cause an explosion. Handle with care. Wear safety glasses

156

Reiß, Hölzel, and Bier

and appropriate chemical-resistant gloves. Always handle piranha solution under a fume hood. 3. In-house made PTFE cover slips holders (Fraunhofer IBMT, St. Ingbert, Germany). 4. n-butyltrichlorosilane (nBTCS; 97%, ABCR, Karlsruhe, Germany). Keep under argon. 5. 500 mL GLS 80 wide neck glass bottle (Duran Group, Mainz, Germany). 2.6. Adsorption of OligonucleotideModified Microspheres with Preannealed Template to nBTCSModified Cover Slips

1. Phosphate-buffered saline (1× PBS): 10  mM Na2HPO4, 2  mM KH2PO4, 137  mM NaCl, 2.7  mM KCl; pH 7.4. Buffer should be 0.2 mm filtered (ME 24 membrane filter, mixed cellulose ester; Whatman) into a clean glass bottle and can be stored at room temperature. 2. 1× PBS + 0.05% (v/v) 20 and 0.1× PBS. 3. Humidity chamber: Plastic petri dish (diameter 14 cm) with a fitted piece of filter paper, wetted with 1 mL water.

2.7. RCA in a Flow-Through System 2.7.1. Flow System

1. Flow cell: In-house made of PVC, effective volume approximately 9  mL (Fraunhofer IBMT, Potsdam, Germany). Figure 2 shows a photo of the flow cell. 2. Flow cell sealing: O-ring (P574/NBR 55, size 8 × 1; C. Otto Gehrckens, Pinneberg, Germany). 3. Peristaltic pump: 520DU equipped with a 205CA pump head (Watson-Marlow, Rommerskirchen, Germany). 4. Tubing: Marprene pump hose (inner diameter: 0.25  mm; Watson-Marlow) (see Note 3) and PTFE capillary (inner diameter 0.25 mm; ERC, Riemerling, Germany). 5. Sodium hypochlorite (NaOCl) cleaning solution: Dilute NaOCl solution (available chlorine 10–13%, reagent grade; Sigma-Aldrich) 1:10 in water. Always prepare dilution freshly.

Fig. 2. Photo of the in-house made flow cell with o-ring. Pitch line: 1 mm.

Preparation of DNA Nanostructures with Repetitive Binding Motifs

157

6. Bovine serum albumin (BSA) blocking solution: Prepare a 5% (w/v) solution of protease-free BSA [Albumin Fraction V (pH 7.0), Blotting grade; Applichem] in water and filter through a 0.2  mm filter. Store in 1  mL aliquots at −20°C. Always prepare 1% BSA blocking solution freshly by mixing 350  mL water with 50  mL of 10× Phi29 DNA polymerase reaction buffer (see Subheading  2.7.2) and 100  mL of 5% BSA solution. 2.7.2. RCA Reaction

1. Phi29 DNA polymerase (10 U/mL; Fermentas). 2. 10× Phi29 DNA polymerase reaction buffer (supplied with the enzyme): 330  mM Tris–acetate (pH 7.9 at 37°C), 100  mM magnesium acetate, 660  mM potassium acetate, 1% Tween 20, 10 mM DTT. 3. 10 mg/mL BSA, purified (New England Biolabs, Frankfurt am Main, Germany). 4. 100  mM dNTP-Mix (each 25  mM dATP, dCTP, dGTP, dTTP; Bioline, Luckenwalde, Germany).

2.8. Staining of RCA-Product with SYBR Green II, Stretching, and Microscopic Detection

1. TBE-buffer: 45 mM Tris–borate, 1 mM EDTA; pH 8. Filter buffer 0.2 mm (Puradisc 30 Syringe Filter, cellulose acetate; Whatman) and store aliquoted at −20°C. Buffer is degassed to 10 mbar before use with a PC 2001 VARIO pumping unit controlled by a CVC 2000II pump controller (Vacuubrand, Wertheim, Germany). 2. 200× SYBR Green II predilution: Dilute SYBR Green II 10,000× stock solution (Cambrex Bio Science) 1:50 in DMSO. Store at −20°C (see Subheading 2.2, step 9 for safety precautions concerning SYBR Green handling). 3. SYBR Green II staining solution: Dilute 200× SYBR Green II predilution 1:200 in degassed TBE and mix carefully by pipetting up and down. Keep protected from light. 4. Ultrapure water, Degassed to 10 mbar before use. 5. Microscope: BX51 upright epifluorescence microscope (Olympus, Hamburg, Germany) equipped with a 100 W halogen lamp, a 100 W mercury arc lamp and an F-View II-camera, controlled by cell^M imaging software (Olympus). 6. Filter-sets: U-MWIBA3 (excitation filter 460–495  nm, dichroic mirror 505  nm, emission filter 510–550  nm) for  SYBR Green detection and U-MWIY2 (excitation filter 545–580 nm, dichromatic mirror 600 nm, and emission filter BA610IF) for bead-detection. 7. Objective: UPlanFLN (N.A. 1.3; Olympus).

100×

oil

immersion

objective

158

Reiß, Hölzel, and Bier

3. Methods 3.1. Preparation of Circular ssDNA Template

In this section, the circularization of ssDNA oligonucleotide ­ olecules with the enzyme CircLigase ssDNA ligase (Epicentre) m is described. This thermostable ATP-dependent ligase catalyzes the intramolecular ligation of ssDNA molecules that have a 5′-phosphate and a 3′-hydroxyl group without the need for a complementary oligonucleotide. Under standard reaction conditions, there are almost no linear or circular concatemers produced (13). 1. 20  mL circularization reaction: Pipette 12  mL of ultrapure water into a 0.2 mL reaction tube cooled on ice and add the following components: 2 mL 10× CircLigase reaction buffer, 2 mL of 5 mM oligonucleotide cF5cF10cF9cF1, 1 mL of 1 mM ATP, 1 mL of 50 mM MnCl2, and 2 mL of 100 U/mL CircLigase ssDNA ligase. For the negative control, replace the enzyme by water. For an acceptable yield prepare 12 parallel circularization reactions (prepare a 13× master mix for this purpose and divide it into 12 aliquots, 20 mL each). 2. Incubate the tubes on a thermal cycler using the following program: 2  h at 60°C, 10  min at 80°C, 4°C until further ­processed (see Note 4). 3. Digestion of remaining linear ssDNA template and adenylatedintermediate: Add 1  mL E. coli exonuclease I (20  U/mL) to each of the reaction tubes (do not add exonuclease I to the negative control as it will digest the linear ssDNA) and incubate on a thermal cycler using the following program: 90 min at 37°C, 15 min at 80°C, 4°C until further processed (see Note 5). 4. Removal of enzymes, dNTPs, and salts: Pool 12 positive ­samples and purify them using the Qiagen Nucleotide Removal Kit according to the manufacturer’s instructions. Elution from each column is done with 50 mL ultrapure water at pH 7.8. 5. Pool eluates from four columns (200  mL) and concentrate DNA in a vacuum concentrator to approximately 40 mL. 6. Determine the concentration of the circularized ssDNA using a NanoDrop ND-1000 Spectrophotometer by absorbance measurement at 260 nm (see Note 6).

3.2. Denaturing Polyacrylamide Gel Electrophoresis of Circularized DNA Template

These instructions refer to the use of a Mini-Protean 3 Cell vertical gel electrophoresis system (Bio-Rad, Hercules, CA, USA) with 0.75 mm spacer plates. If other systems are used, volumes might need to be adjusted.

Preparation of DNA Nanostructures with Repetitive Binding Motifs

159

1. In a 50 mL conical centrifugal tube mix 3 mL of 30% (w/v) acrylamide/bis-acrylamide solution with 0.6 mL 10× TBEgelbuffer, pH 8.3, and add 2.52 g urea. 2. Dissolve urea in an ultrasonic bath under mild warming (30–40°C). 3. Fill up volume to 6  mL with water and degas the resulting solution to 10 mbar (work under a fume hood) (see Note 7). 4. Meanwhile, prepare casting stand according to the manufacturer’s instructions. 5. Add 30  mL of a freshly prepared 10% (w/v) ammonium ­persulfate solution to the degassed acrylamide/bis-acrylamide solution from step 3 and mix by carefully inverting the conical tube (avoid the formation of air bubbles). 6. Add 3 mL TEMED, mix carefully (as in step 5) and immediately pipette the solution between the glass plates. Insert the comb and allow the gel to polymerize for at least 2 h. 7. When the gel has polymerized, carefully remove the comb and assemble the electrophoresis module. Fill 1× TBEgelbuffer, pH 8.3 into the inner and the lower buffer chamber of the assembly. Rinse wells carefully with 1× TBEgel-buffer, pH 8.3 using a microliter syringe. 8. Prerun gel for 30 min at 200 V. 9. Prepare samples: Add 6  mL urea loading buffer each to 6 mL of 0.5 mM circularized ssDNA (positive sample) and to 6 mL of 0.5 mM of the negative control and incubate for 5 min at 95°C on a thermal cycler, then immediately put samples on ice. 10. Also dilute size-marker Hyperladder V (Bioline GmbH, Luckenwalde, Germany) 1:8 into nondenaturing loading buffer. 11. Rinse wells again in order to remove urea that leaches into the sample wells. Load samples (each 12  mL) into separate wells of the gel and run gel at 200 V until the xylene cyanol FF-band reaches the bottom of the gel. 12. Disassemble the electrophoresis device and carefully transfer the gel to a plastic box with 1:10,000 diluted SYBR Green II-solution in 1× TBEgel-buffer, pH 8. 13. Stain gel for 30 min under mild shaking in the dark and read out fluorescence using a gel documentation system. An example for a typical electrophoresis run of the samples is given in Fig. 3. Compared to the linear form of the oligonucleotide (lane 1), the circularized molecules (lane 2) migrate retarded through the gel.

160

Reiß, Hölzel, and Bier

Fig. 3. Analysis of the circularization reaction on a 15% Polyacrylamide/7 M Urea gel. Lane 1: linear ssDNA form (negative control without CircLigase); lane 2: circularized ssDNA form; lane 3: DNA size marker Hyperladder V (Bioline).

3.3. Covalent Coupling of Amino-Modified Oligonucleotides to 20 nm Carboxylated Microspheres

In this protocol, the carboxyl groups at the surface of fluorescent polystyrene beads are activated with EDC. An amine-reactive O-acylisourea intermediate is formed and stabilized with SulfoNHS by converting it into an amine-reactive Sulfo-NHS ester. Upon reaction with the amino-modified oligonucleotide, a covalent bond is formed. The reaction is quenched with glycin and excess coupling reagent and oligonucleotide molecules are removed by several washing steps. Although the fluorescent microspheres are rather insensitive to photobleaching, it is advisable to carry out all steps in this protocol in the dark. 1. Vortex stock solution of 20 nm fluorescence beads (2% solids) and sonicate for 5 min in an ultrasonic bath. 2. Dry 10 mL of a 100 mM solution of amino-modified oligonucleotide F5 in a 0.5 mL reaction tube using a vacuum concentrator and then resuspend oligonucleotide in 26  mL of MES-buffer. 3. Add 10  mL of the bead suspension from step 1 to 390  mL MES-buffer and sonicate for 5 min in an ultrasonic bath.

Preparation of DNA Nanostructures with Repetitive Binding Motifs

161

4. Transfer diluted bead suspension to the filter cup of an Ultrafree-MC centrifugal filter unit and centrifuge at 3,000 × g until 20–40 mL remain (approximately 4 min). 5. Pipette bead suspension out of centrifugal filter unit and transfer it to a clean reaction tube. Take care to resuspend beads that have accumulated at the membrane of the filtration unit by carefully pipetting up and down without damaging the membrane. Bring volume to 40 mL with MES-buffer and sonicate for 5 min in an ultrasonic bath. 6. Add 4  mL of the bead suspension to the 26  mL of DNA ­solution from step 2. 7. Add 15  mL of a freshly prepared 100  mM Sulfo-NHS solution in MES-buffer and 15  mL of a freshly prepared 50  mM EDC solution in MES-buffer and vortex shortly (see Note 8). 8. Incubate for 2 h at 1,000 rpm at 25°C on a thermomixer. 9. Add 4 mL of a 1 M glycine solution (deactivation of unreacted carboxyl groups). 10. Incubate for 30 min at 1,000 rpm at 25°C on a thermomixer. 11. Add 336 mL 0.02% (v/v) Tween 20 in H2O, mix and transfer to an unused Ultrafree-MC centrifugal filter unit. 12. Centrifuge at 3,000 × g until 20–40 mL remain (approximately 4 min). 13. Discard filtrate and bring volume in the filter cup to 400 mL with 0.02% (v/v) Tween 20 in H2O. Resuspend beads well by  pipetting carefully up and down without touching the membrane of the filter cup. 14. Centrifuge at 3,000 × g until 20–40 mL remain (approximately 4 min). 15. Repeat steps 13 and 14 twice. 16. Discard filtrate and bring volume in the filter cup to 400 mL with 0.02% (v/v) Tween 20 in 1× TE-buffer. Resuspend beads well by pipetting up and down without touching the membrane of the filter cup. 17. Centrifuge at 3,000 × g until 20–30 mL remain. Transfer the suspension to a clean 0.5  mL reaction tube. Take care to resuspend beads that have accumulated at the membrane of the filter cup by carefully pipetting up and down without damaging the membrane. Bring volume to 30 mL with 0.02% (v/v) Tween 20 in 1× TE-buffer. 18. Store in the dark at 4°C. Oligonucleotide-modified beads are functional at least for 3 months.

162

Reiß, Hölzel, and Bier

3.4. Annealing of Circular ssDNA Template to OligonucleotideModified Microspheres

1. Dilute F5-modified beads (see Subheading 3.3) 1:10 in 0.02% (v/v) Tween 20 in 1× TE-buffer. 2. In a 0.2  mL reaction tube mix 4.25  mL H2O, 3  mL of 3× annealing buffer, 0.75 mL of 1 mM circularized oligonucleotide cF5cF10cF9cF1 (see Subheading 3.1) and 1 mL of the 1:10 diluted bead suspension. 3. Anneal on a thermal cycler by heating to 55°C for 5 min and then decreasing the temperature by −1°C/min to 20°C. Keep in the dark at 4°C until used. Preannealed template is functional for at least 2 weeks.

3.5. Preparation of nBTCS-Modified Cover Slips

This protocol describes a method to vapor-silanize cover slips with nBTCS. 1. Number cover slips in a corner with a diamond scriber and sort them into the PTFE-holder. 2. Wash cover slips for 2 min under running deionized water. 3. Sonicate cover slips for 5 min in a 1 L-beaker with ultrapure water in an ultrasonic bath set to 60°C. 4. Transfer precleaned cover slips in the PTFE-holder to a borosilicate glass beaker with freshly prepared piranha solution (see Note 9). 5. Sonicate cover slips in the piranha solution for 1  h in an ­ultrasonic bath set to 60°C. 6. Wash cover slips for 5 min under running deionized water. 7. Sonicate cover slips for 5 min in a 1 L-beaker with ultrapure water in an ultrasonic bath set to 60°C. 8. Centrifuge cover slips dry for 5 min at 1,000 × g. 9. Transfer cover slips to a clean PTFE-holder and put the ­latter into a clean 500  mL GLS 80 wide neck glass bottle (see Note 10). 10. Pipette 50  mL nBTCS in a small glass Petri dish in the 500  mL GLS 80 wide neck glass bottle and cap tightly immediately. 11. Incubate bottle 18–20 h at 90°C in a drying oven. 12. Open bottle under a fume hood and transfer PTFE-holder with cover slips to a clean beaker glass, cap the latter loosely with aluminum foil and incubate for 1 h at 110°C in a drying oven. 13. Store nBTCS-modified cover slips desiccated under vacuum. Cover slips can be used at least for 1 month (see Note 11).

Preparation of DNA Nanostructures with Repetitive Binding Motifs

3.6. Adsorption of OligonucleotideModified Microspheres with Preannealed Template to nBTCS-Modified Cover Slips

163

1. Dilute F5-modified beads to which circular template has been annealed (see Subheading 3.4) 1:200 in 1× PBS. 2. Pipette 100 mL of the dilution in the center of an nBTCSmodified cover slide (see Subheading 3.5). 3. Incubate for 1 h in a humidity chamber in the dark. 4. Wash for 1  min each in 1× PBS + 0.05% (v/v) Tween 20, 1× PBS, and 0.1× PBS. 5. Dry under a stream of nitrogen.

3.7. RCA in a Flow-Through System

A photo of the experimental setup (microscope, peristaltic pump, flow cell device) is given in Fig. 4. 1. Cleaning of flow system: Mount a clean 15 × 15  mm cover slip in the flow cell and assemble flow cell device. Clean flow system by pumping NaOCl cleaning solution for 30 min at 10  mL/min through it. Then pump water at 10  mL/min through flow system over night (16–20 h). 2. Blocking of flow system: After cleaning of the flow system, block it with BSA by pumping BSA blocking solution for  30  min at 10  mL/min through the flow system. Wash 10 min with water at 10 mL/min and then dry flow system by pumping air through the flow system with 10  mL/min. Disassemble flow cell device, remove cover slip, and dry flow cell under a stream of nitrogen. 3. Mount nBTCS-modified cover slip to which beads with the preannealed template have been adsorbed (see Subheading 3.6), upside down on the flow cell and assemble flow cell device.

Fig. 4. Photo of experimental setup. (a) microscope objective; (b) flow cell mounted into flow cell holder; (c) tubing; (d) peristaltic pump; (e) reagent supply.

164

Reiß, Hölzel, and Bier

4. Prepare Phi29 DNA polymerase reaction solution on ice by pipetting in a 1.5  mL reaction tube 158  mL of degassed water, 20  mL 10× Phi29 DNA polymerase reaction buffer, 4 mL of 10 mg/mL purified BSA, 8 mL of 100 mM dNTPMix, and 10 mL of 10 U/mL Phi29 DNA polymerase. Mix between each addition step carefully by pipetting up and down and finally spin down shortly. 5. Place the free end of the pump hose to the bottom of the reaction tube containing the Phi29 DNA polymerase reaction solution (inlet) and the end of the teflon tubing (outlet) slightly above the meniscus of the reaction solution. Seal reaction tube with a piece of Parafilm. 6. Pump Phi29 DNA polymerase reaction solution with 10 mL/min into the pump tubing until it reaches the inlet of the flow cell, then reduce speed to 3 mL/min and let the flow system fill completely (see Note 12). 7. Allow the Phi29 DNA polymerase reaction to take place over night (16–20 h). 3.8. Staining of RCA-Product with SYBR Green II, Stretching, and Microscopic Detection

1. After the RCA reaction has taken place, pump degassed water for 5 min at 3 mL/min into the flow system. 2. Pump SYBR Green II staining solution for 3 min at 3 mL/min into the system. 3. Pump 1× TBE-buffer for 30 min at 3 mL/min through the system. 4. Pump water for 30 min at 3 mL/min through the system. 5. Find correct focal plane by focusing onto the fluorescence beads using the U-MWIY2 filter cube and Hg-illumination (100× oil immersion objective) (see Note 13). 6. Observe SYBR Green II-stained RCA products microscopically using the U-MWIBA3 filter set and halogen illumination. 7. Stretching of SYBR Green II-stained molecules: Pump air at 3  mL/min for 8  min into the flow system, then put pump hose back into water. 8. When the air segment reaches the flow cell inlet, reduce pump speed to 0.4  mL/min (minimum speed available with the experimental setup used in this chapter). 9. Increase pump speed to 3  mL/min when the air segment reaches the outlet of the flow cell in order to allow refilling with water. 10. Observe SYBR Green II-stained ssDNA molecules microscopically using the U-MWIBA3 filter cube and halogen illumination. A typical fluorescence photomicrograph

Preparation of DNA Nanostructures with Repetitive Binding Motifs

165

Fig. 5. (a) Fluorescence photomicrograph showing an individual RCA product that has been stained with SYBR Green II and stretched out by the passage of a water/air-interface. Imaging conditions: 100× objective, halogen lamp, exposure time 500  ms, U-MWIBA3 filter cube. (b) Line plot showing the background-corrected intensity profile of the emitted ­fluorescence of the RCA product shown in (a).

showing a SYBR Green II-stained and stretched ssDNA molecule is given in Fig. 5a. Figure 5b shows the line plot of the emitted fluorescence of this molecule. The higher intensity at the starting point of the molecule is a result of the entangling of a part of the molecule at its bead-bound starting point.

4. Notes 1. According to the instruction sheet accompanying the enzyme (also available online (13)), the ligation of short ssDNA (as described in this chapter) is enhanced by the addition of 2.5 mM MnCl2, but should not be added when long ssDNA molecules (e.g., cDNA) are to be ligated. Furthermore, the yield of the circularization reaction can be strongly influenced by the sequence of the ssDNA. 2. Phosphothioate (PTO) modification: In PTO-modified ­oligonucleotides, one sulfur atom replaces an oxygen atom in the internucleotidic linkage. PTO-modification of the last two nucleotides at the 3′-end of the oligonucleotides protects them from degradation by the 3′–5′ exonuclease activity of Phi29 DNA polymerase (18). 3. Marprene pump hoses with 0.25  mm inner diameter were very stable concerning the calibration of the flow rate of the peristaltic pump and could be used for at least 500 h. 4. The enzyme manufacturer recommends an incubation time of 1 h at 60°C (13).

166

Reiß, Hölzel, and Bier

5. According to the instruction sheet of the enzyme 45  min incubation time for exonuclease I digestion is recommended. If the linear ssDNA sequence contains extensive secondary or hairpin structures, additional incubation with exonuclease III might be necessary (13). 6. For the oligonucleotide sequence used for circularization in this chapter, we obtained approximately 40 mL of 1 pmol/mL circularized product from 120  pmol linear educt (i.e., 33% yield). 7. Degassing removes oxygen, which is an inhibitor of the polymerization reaction, from the solution. 8. Work quickly as Sulfo-NHS and EDC are both not stable in aqueous solutions. Especially EDC will hydrolyze rapidly. 9. Only use tweezers made of PTFE or stainless steel for handling of the PTFE-holder in the piranha solution. 10. Only use clean equipment (500 mL GLS 80 bottle, cap, and small Petri dish) for the silanization step. Equipment is cleaned by soaking in a bath of saturated KOH in isopropyl alcohol over night, followed by soaking in 1% Hellmanex II-solution (Hellma Optik, Jena, Germany) for several hours. It is then washed thoroughly under running deionized water and dried in a drying oven at 90°C. The blue color of the cap of the bottle will turn purple due to the impact of HCl vapor that develops during the silanization reaction. This does not seem to have a major influence on the quality of the silanization of the cover slips. 11. If the silanization procedure has been carried out correctly, cover slips should have an equilibrium contact angle around 100°. Use only nBTCS-modified cover slips that are absolutely clear. If a white haze appears on the cover slips, they have not been absolutely clean or dry before vapor phase silanization. 12. If problems with air bubble-free filling of the flow cell occur, it can be put vertically during the filling process, with the outlet of the flow cell at the top. After complete filling, put it back onto the microscope stage in a horizontal position. 13. Finding the correct focal plane is facilitated greatly by focusing onto the o-ring seal of the flow cell under incident brightfield illumination with a lower magnification (e.g., 20× objective). Exactly adjusting the focal plane by focusing onto the fluorescence beads with the 100× oil immersion objective and the U-MWIY2 filter cube allows direct image acquisition of the SYBR Green-stained DNA molecules when switching to the U-MWIBA3 filter cube. This procedure minimizes photobleaching of the DNA molecules.

Preparation of DNA Nanostructures with Repetitive Binding Motifs

167

Acknowledgments The authors thank the group of C.M. Niemeyer (University of Dortmund) for communicating DNA oligonucleotide sequence information. This work was supported by the European Union’s sixth framework program, contract no. NMP4-CT-2004-013775, under the project name NUCAN (Nucleic Acid Based Nanostructures). References 1. Fire, A., and Xu, S. Q. (1995) Rolling replication of short DNA circles Proc. Natl. Acad. Sci. U. S. A. 92, 4641–5. 2. Liu, D., Daubendiek, S. L., Zillman, M. A., Ryan, K., and Kool, E. T. (1996) Rolling ­circle DNA synthesis: Small circular oligonucleotides as efficient templates for DNA ­polymerases J. Am. Chem. Soc. 118, 1587–94. 3. Lizardi, P. M., Huang, X., Zhu, Z., BrayWard, P., Thomas, D. C., and Ward, D. C. (1998) Mutation detection and singlemolecule counting using isothermal rollingcircle amplification Nat. Genet. 19, 225–32. 4. Schweitzer, B., Roberts, S., Grimwade, B., Shao, W., Wang, M., Fu, Q., Shu, Q., Laroche, I., Zhou, Z., Tchernev, V. T., Christiansen, J., Velleca, M., and Kingsmore, S., F. (2002) Muliplexed protein profiling on microarrays by rolling-circle amplification Nat. Biotechnol. 20, 359–65. 5. Beyer, S., Nickels, P., and Simmel, F. C. (2005) Periodic DNA nanotemplates synthesized by rolling circle amplification Nano Lett. 5, 719–22. 6. Deng, Z., Tian, Y., Lee, S. H., Ribbe, A. E., and Mao, C. (2005) DNA-encoded selfassembly of gold nanoparticles into onedimensional arrays Angew. Chem., Int. Ed. 44, 3582–5. 7. Zhao, W., Gao, Y., Kandadai, S. A., Brook, M. A., and Li, Y. (2006) DNA polymerization on gold nanoparticles through rolling circle amplification: Towards novel scaffolds for three-dimensional periodic nanoassemblies Angew. Chem., Int. Ed. 45, 2409–13. 8. Cheglakov, Z., Weizmann, Y., Braunschweig, A. B., Wilner, O. I., and Willner, I. (2008) Increasing the complexity of periodic protein nanostructures by the rolling-circle-amplified synthesis of aptamers Angew Chem., Int. Ed. 47, 126–30.

9. Zhao, W., Ali, M. M., Brook, M. A., and Li, Y. (2008) Rolling circle amplification: Applications in nanotechnology and ­biodetection with functional nucleic acids Angew. Chem., Int. Ed. 47, 6330–7. 10. Reiß, E., Hölzel, R., Nickisch-Rosenegk, M. v., and Bier, F. F. (2006) Rolling circle amplification for spatially directed synthesis of a solid phase anchored single-stranded DNA molecule, in DNA-Based Nanoscale Integration: International Symposium on DNA-Based Nanoscale Integration, Jena, Germany 18–20 May 2006 (Fritzsche, W., ed.) 2006, AIP Conference Proceedings 859, American Institute of Physics, Melville, NY, pp. 25–30. 11. Frieden, M., Pedroso, E., and Kool, E. T. (1999) Tightening the belt on polymerases: Evaluating the physical constraints on enzyme substrate size Angew. Chem., Int. Ed. 38, 3654–7. 12. Diegelman, A. M., and Kool, E. T. (2001) Chemical and enzymatic methods for preparing circular single-stranded DNAs Curr Protoc Nucleic Acid Chem Chapter 5, Unit 5.2. 13. Epicentre Biotechnologies. Protocol for CircLigase™ ssDNA Ligase (continued Lit. #222) [homepage on the Internet]. No date [cited 2009 Jan 15]. Available from: http:// www.epibio.com/litindex.asp. 14. Blanco, L., Bernad, A., Lázaro, J. M., Martín, G., Garmendia, C., and Salas, M. (1989) Highly efficient DNA synthesis by the phage Phi 29 DNA polymerase. Symmetrical mode of DNA replication J. Biol. Chem. 264, 8935–40. 15. Reiß, E., Hölzel, R., and Bier, F. F. (2009) Synthesis and stretching of rolling circle amplification products in a flow-through system Small 5, 2316–22. 16. Feldkamp, U., Schroeder, H., and Niemeyer, C. M. (2006) Design and evaluation of singlestranded DNA carrier molecules for DNAdirected assembly J. Biomol. Struct. Dyn. 23, 657–66.

168

Reiß, Hölzel, and Bier

17. Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A., and Struhl, K., (Eds.) (1999) Short Protocols in Molecular Biology 4th ed., John Wiley & Sons, USA.

18. Dean, F. B., Nelson, J. R., Giesler, T. L., and Lasken, R. S. (2001) Rapid amplification of plasmid and phage DNA using Phi29 DNA polymerase and multiply-primed rolling circle amplification Genome Res. 11, 1095–9.

Chapter 12 Controlled Confinement of DNA at the Nanoscale: Nanofabrication and Surface Bio-Functionalization Matteo Palma, Justin J. Abramson, Alon A. Gorodetsky, Colin Nuckolls, Michael P. Sheetz, Shalom J. Wind, and James Hone Abstract Nanopatterned arrays of biomolecules are a powerful tool to address fundamental issues in many areas of biology. DNA nanoarrays, in particular, are of interest in the study of DNA–protein interactions and for biodiagnostic investigations. In this context, achieving a highly specific nanoscale assembly of oligonucleotides at surfaces is critical. In this chapter, we describe a method to control the immobilization of DNA on nanopatterned surfaces; the nanofabrication and the bio-functionalization involved in the process will be discussed. Key words: DNA, Self-assembly, Nanoscale, Nanotechnology, Fluorescence microscopy

1. Introduction The ability to control the placement of biomolecules on surfaces with nanometer resolution is of great interest to the study of biological events at the single molecule level, and provides a platform for the development of biosensing devices with unparalleled sensitivity (1–9). In particular, the confinement of DNA on surfaces has gathered a great deal of interest, as it can be employed for bioanalytical (genomic) studies and can be used to drive the self-organization of biological (e.g., proteins) and inorganic (e.g., nanoparticles) moieties with nanometer resolution (10–17). Various chemical strategies have been employed to immobilize DNA on surfaces, ranging from electrostatic interactions, formation of (thiolated) self-assembled monolayers (SAMs), direct

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_12, © Springer Science+Business Media, LLC 2011

169

170

Palma et al.

covalent attachment, and through biotin–streptavidin interaction (18–24). Different fabrication methods have been used to control such immobilization in order to generate micro- and nanoscale DNA features: contact printing (25–27), AFM-based methods (28), and nanopipette deposition (29) are among the most notable examples of the different approaches pursued. In this chapter, we present a strategy merging top-down nanofabrication techniques with bottom-up self-assembly, to control the confinement of DNA molecules on substrates with nanometer resolution. We will first describe the fabrication procedure employed to nanopattern glass substrates with Au/Pd nanodots. This consists of a direct electron-beam (e-beam) writing step that creates nanometer-scale voids in a resist polymer spun on the glass substrate; a subsequent metal evaporation fills the voids with Au/Pd and produces an array of sub-50-nm metal dots in the desired geometry (defined by the e-beam writing). The Au/Pd nanodots so produced can be used as functional regions on the substrates for the controlled confinement of DNA. We will here describe a biotin–streptavidn-based functionalization methodology to immobilize oligonucleotide chains on the so-prepared nanopatterned surface (21, 22, 24, 30–32). The metal nanodots allow for the formation of SAMs of thiolated alkanes presenting biotin end groups, which can be used for the subsequent immobilization of streptavidin. Employing such a strategy, we immobilize double-stranded DNA (dsDNA) on every fabricated (and properly functionalized) nanodot. This functionalization procedure allows for the formation of non-sterically hindered DNA nanodomains at surfaces; a homogenous surface packing density can be envisioned, an advantage over thiolated DNA SAMs (20). We verify the validity of our approach by EPI-fluorescence microscopy imaging.

2. Materials 2.1. Metal on Glass Nanopattern Fabrication

In our laboratory, we have employed the following instrumentation: a CEE Brewer 100 resist spinner (Brewer Science), an electronbeam writing system (FEI, with Nabity NPGS stage writing control), and an electron beam evaporator (Semicore). 1. Glass coverslips, 22 mm × 22 mm, No. 1.5 (Corning). 2. Acetone, HPLC grade. 3. Ethanol, 200 proof. 4. Sulfuric acid (H2SO4), concentrated.

Controlled Confinement of DNA at the Nanoscale

171

5. Hydrogen peroxide (H2O2), 30% concentration. Store at 4°C. 6. 7× Cleaning solution (MDBio). 7. Polymethylmethacrylate (PMMA) photoresist: 495K and 25K molecular weight (MW), A3 concentration for 25K, A2 concentration for 495K (Microchem Corp.); store in the dark. 8. Aquasave conductive coating (Mitsubishi). 9. Isopropanol (IPA), HPLC grade, 99.8%. 10. Tweezers, round tip, stainless steel (SPI supplies). 11. Gold/palladium 60/40%, pellets (Plasmaterials). 12. Titanium (Plasmaterials). 2.2. Functionalization of Nanopatterned Surfaces

A Harrick PDC-32G plasma cleaner was used in our laboratory. 1. Toluene anhydrous, 99.8%. 2. Ethanol anhydrous, >99.5% (200 proof). 3. Acetic acid, glacial ACS reagent grade. 4. Ethanol, ACS reagent grade. 5. Acetone, ACS reagent grade. 6. Gibco™ Dulbecco’s phosphate buffer saline (PBS) 1× (no magnesium, no calcium, 2.7 mM potassium chloride, 0.14 M sodium chloride, 1.5 mM potassium phosphate, and 8 mM sodium phosphate, pH 7.4), Invitrogen; store at room temperature (RT). 7. ThermoScientific BupH™ PBS (TPBS) (0.1 M sodium phosphate and 0.15 M sodium chloride, pH 7.2); store at RT. 8. Deionized (DI) Millipore water (resistivity of 18 MWcm). 9. HS–(CH2)11–(C2H6O2)3–OH and HS–(CH2)11–(C2H6O2)3– biotin (ProChimia); store at −20°C. 10. Polyethylene glycol (PEG)-silane, 5,000 MW (mPEG5000) (LaysanBio); store at −20°C. 11. Glass syringes, metal needles for the anhydrous solvents (Popper). 12. Teflon mini-rack (Invitrogen). 13. 6-Well plates (Falcon). 14. Parafilm. 15. Streptavidin (Invitrogen); store at −20°C. 16. Chicken egg albumin (Sigma); store at 4°C. 17. Oligonucleotides (IDT): one 20-mer with a biotin functional group at the 5¢ position (5¢-/52-Bio/GTC ACT TCA GCT GAG ACG CA-3¢) and the complementary strand with a Cy3 fluorophore at the 5¢-end (5¢-/5Cy3/TGC GTC TCA GCT GAA GTG AC-3¢); store at 4°C wrapped in aluminum foil.

172

Palma et al.

2.3. Microscopy

In our laboratory, epifluorescence microscopy was performed on an inverted microscope, Olympus IX81 (Olympus) equipped with a Cascade II, 512 × 512 pixel CCD camera (Photometrics). 1. Cloning ring (Sigma/Aldrich). 2. High vacuum grease (Dow Corning).

3. Methods Throughout the steps of the method described here, a great deal of attention should be focused on the cleanliness of both the laboratory working environment and the materials and tools employed. Because of the nanometer scale of the features fabricated (and functionalized), we recommend carrying out the fabrication steps (Subheading 3.1) in an ultraclean environment, such as a clean room. Furthermore, in order to obtain a successful bio-functionalization (Subheading 3.2) of the fabricated nanopatterns, it is important to carry out the procedure rapidly with as short a time lapse as possible between steps. Moreover, all glassware and tweezers used must be dry, preferably stored in an oven at approximately 70°C, and cooled in air prior to use. It should be noted that in order to verify the validity of our approach (i.e., the controlled DNA functionalization of nanopatterned surfaces), we have used a biotinylated and Cy3-labeled dsDNA. We have hybridized in solution, prior to the attachment to the surface, a biotinylated oligonucleotide (20-mer chain) and its complementary oligonucleotide labeled with a Cy3 fluorophore. (In Subheading 4, we report the alternative procedure to immobilize biotinylated single-stranded DNA). 3.1. Metal on Glass Nanopattern Fabrication 3.1.1. Preparation and Cleaning

Cleaning is absolutely critical to the success of the procedure. Any contamination, even nanoscopic, not removed prior to patterning will result in defects in the surface passivation (see Note 1). 1. Prepare piranha solution (1/3 volume of H2O2 plus 2/3 of H2SO4). 2. Starting with a new glass coverslip, sonicate in ethanol for 2 min. 3. Blow dry with a stream of inert gas (Ar or N2) (see Note 2). 4. Dilute 7× detergent 1:4 with deionized (DI) water and bring to boiling temperature on a hot plate at 200°C (goes from cloudy to clear). 5. Immerse the slide for 2 min in boiling solution and rinse with DI water for 10 min. 6. Immerse in piranha solution and let soak for 5 min.

Controlled Confinement of DNA at the Nanoscale

173

7. Take the slide out of the piranha solution and rinse it for 10 min with DI water. 8. Rinse with ethanol. 9. Blow dry with a stream of inert gas (Ar or N2) (see Note 3). It is now possible to proceed to resist deposition. 3.1.2. Resist Deposition

Again, cleanliness is paramount here. All work should be performed in a clean room, class 10,000 or better. A bilayer of higher MW resist is spun on top of a lower molecular weight resist to aid with the subsequent metal deposition and liftoff (discussed later). The high MW top layer has a slightly different dose curve and will develop with a narrower opening, creating an overhang, which ensures proper liftoff (see Note 4). 1. Preheat a hot plate to 180°C and place cleaned samples on the hot plate for 1  min to force off any adsorbed water molecules. 2. Let cool for 10 s before placing the sample on a resist spinner chuck (see Note 5). 3. Spin lower MW resist (25K) first at 4,000 rpm for 45 s, use a ramp rate of 1,000 rpm/s (see Note 6). 4. Bake for 5 min at 180°C on a hot plate. 5. Let cool for 10 s before placing the sample on the resist spinner chuck. 6. Spin higher molecular weight resist (495K) as top layer at 4,000 rpm for 45 s, use a ramp rate of 1,000 rpm/s. 7. Bake for 5 min at 180°C. 8. Let cool for 10 s before placing on the resist spinner chuck. 9. Spin on the Aquasave conductive discharge layer at 1,000 rpm for 45 s at a ramp rate of 300 rpm/s. It is important to make sure to dab any droplets of Aquasave at the edges (particularly corners) so that the sample is completely dry before being placed in an electron-beam writer. It is now possible to proceed to the e-beam writing step.

3.1.3. Electron-Beam Pattern Writing

An electron-beam writing system is an SEM that controls beam shuttering and position to generate patterns from CAD files. Process testing will be necessary to determine optimal doses for generating the desired features (see Note 7). We use a pattern which is 50 mm by 50 mm, and consists of 1 mm register squares spaced every 10  mm, with sub-50-nm dots filling every 2  mm between them. This ensures that each individual dot is optically resolvable and discrete once functionalized and imaged with a fluorescence microscopy (see Note 8).

174

Palma et al.

1. Load samples in the e-beam writer, making sure that a good top contact is made with the mounting clips. 2. Pump down chamber. When pressure has reached or is below the specified value (in our case, 5 × 10−6 Torr), turn the beam on; use an acceleration voltage of 30 kV. 3. Check the beam currents of the spot sizes to be written within the Faraday cup. It is necessary to check beam currents each time in the Faraday cup, as currents drift over time. 4. Update the run files with the most recent beam currents to ensure that the doses are consistent. 5. Check gun tilt and stigmation, and adjust these until both are minimized (see Notes 9–11). 6. Move to the sample. Four-point focus (see Note 11) must be taken to correct for sample tilt. No sample is perfectly level, resulting in loss of focus over the writing area. Focus on the surface of the sample at four points around the region to be written, registering the points with the control software. A plane is then fit to these four points, and the focal depth interpolated and adjusted as the pattern is written (see Notes 10 and 11). 7. With all the steps taken to ensure proper focus and minimal stigmation over the area of the pattern, move to a position inside the four-point focus region. 8. Engage automated stage and beam control, and execute pattern writing in software. 9. After writing, it is necessary to remove the discharge layer, Aquasave. Rinse with DI water until all Aquasave is visibly gone. 10. Then blow dry with a stream of inert gas (Ar or N2). The samples are now ready for development. 3.1.4. PMMA Development

A cold ultrasonic development process is used to achieve nanometerscale features. The cold development sharpens resist contrast, resulting in the smallest possible features from the exposed regions. 1. Place rinsed and dry sample in a solution of 1:3 H2O:IPA at 4°C. Sonicate for 1 min in a water bath sonicator at 4°C (see Notes 12 and 13). 2. Quickly remove the sample and immerse in 100% IPA at room temperature to halt development (see Note 14). 3. Blow dry with a stream of inert gas (Ar or N2). 4. Examine pattern: large alignment marks should be visible; check further in an optical microscope at magnifications of 25–100×. The micron-sized features should be visible. It is now possible to proceed to the metal deposition step.

Controlled Confinement of DNA at the Nanoscale 3.1.5. Metal Deposition

175

In our laboratory, metal deposition is performed in a Semicore electron beam evaporator. A focused beam of electrons is used to heat a target material held in a crucible. As opposed to a thermal evaporator, the heating is highly localized to the surface and allows for precise control of the thickness (at the Ångström level). The sample is held above the target some distance away, allowing for a highly directional and uniform material flux onto the surface. 1. Place the developed samples on the sample holder in the evaporator. 2. Pump the system to less than the suggested threshold pressure and begin deposition procedure. 3. Evaporate a 1-nm adhesion layer of titanium. This is necessary for the adhesion of the Au/Pd to the glass. 4. Next, deposit 3 nm of Au/Pd on top of the titanium. 5. Follow proper power-down and venting procedures, and remove the sample. The metalized sample now consists of a layer of metal sitting atop the unexposed PMMA, with openings in the resist where the sample was exposed to electrons during the e-beam writing session (in these holes, the Ti adhesion layer and Au/Pd are deposited on the glass surface): see Fig. 1.

Fig. 1. Scanning electron microscope (JEOL JSM-5600 LV) image of nanopattern holes prior to liftoff of the PMMA resist layer. The pattern has been written in the electronbeam writer, developed, and the metal layers deposited; however, the unpatterned glass surface is still covered with the PMMA bilayer, with metal on top. The large 1 mm square registers are visible, with sub-50-nm holes in a square lattice with a 2 mm unit cell spacing in between.

Palma et al.

3.1.6. Liftoff

Liftoff is the process to remove the remaining, unexposed resist that is now covered with metal. Recall that a slight overhang was created due to the use of a bilayer. Therefore, metals deposited inside the features on the glass substrate are not connected to the bulk of the metal sitting on top of the resist. Solvent is used to dissolve the resist, removing the metal while leaving behind only the metal deposited in the holes (directly on the glass) as defined by beam writing. The metal will appear to float off of the surface, hence the term liftoff. 1. To remove the resist, place the metal-coated sample in spectroscopic grade acetone (see Note 15). 2. Seal the container with parafilm tightly to avoid evaporation, and let sit overnight to remove the resist and metal layer on top (see Note 16). 3. When the sample is visibly clean of the metal layer, transfer it to a fresh beaker of acetone. 4. Remove the sample from the acetone, and rinse first with acetone from a squirt bottle, and then with ethanol from a squirt bottle (see Note 17). 5. Finally, blow dry with a stream of inert gas (Ar or N2). The Au/Pd nanopatterns on glass are now ready for surface functionalization. Figure  2 displays an atomic force microscopy (AFM) image and profile of the sub-50-nm dots fabricated on a glass coverslip.

500 nm Line profile 2.5 2 1.5 1 0.5 0

0

47.548

200

0.103

0.764 nm

176

nm

400

600

Fig. 2. Atomic force microscopy image of the Au/Pd nanodots array on glass. Imaging was performed with a XE-100, advanced scanning probe microscope (PSIA). The image has been analyzed with the XEI software, version 1.6.

Controlled Confinement of DNA at the Nanoscale

177

Fig. 3. Scheme of the main steps of the bio-functionalization procedure.

3.2. Functionalization of Nanopatterned Surfaces

Figure  3 schematically displays the main steps of the bio-­ functionalization procedure. Starting with a nanopatterned surface, i.e., Au/Pd nanodots on a glass slide, carry out the following steps: 1. Prepare fresh piranha solution (see above). 2. Clean the nanopatterned surface slide by immersion for 3 min in 1 h 30 min-aged piranha solution (if piranha is used as freshly prepared, it will eat away the metal). The immersion can be done using a Teflon rack to hold the sample/s in the container. 3. Take the sample out of the piranha solution with clean tweezers and rinse in DI water for at least 5 min. 4. Rinse the sample with ethanol and blow dry with a gentle stream of inert gas (Ar or N2). 5. Place the dry sample in a plasma cleaner for 5 min at 18 W. 6. Prepare a 1 mM anhydrous ethanol solution of HS–(CH2)11– (C2H6O2)3–OH and HS–(CH2)11–(C2H6O2)3–biotin: mix at 3:1 ratio (see Note 18). 7. Pull the sample out of the plasma cleaner and immediately incubate (see Note 19) it in the freshly prepared EG/biotin– thiol solution (1.5 mL per 6-well plate is sufficient to incubate the glass slide of the dimension used here). Seal the 6-well plate container with parafilm, cover with Al foil, and incubate on a shaker overnight (12–18 h). In this way, a SAM exhibiting functional biotin end groups will be formed on the metal nanodots; such biotin groups can then be used to immobilize streptavidin on the dots.

178

Palma et al.

8. Prepare a fresh solution of 2  mg of mPEG 5000-silane in 25 mL of anhydrous toluene and add 30 mL of acetic acid (as a catalyst) (see Notes 20 and 21). 9. Rinse the sample in ethanol and blow dry with a gentle stream of inert gas (Ar or N2). 10. Place the dry sample in the PEG solution. Incubate in a glass container, sealed with parafilm and covered with Al foil for at least 24 h (up to 48 h). This step is fundamental to assure proper passivation of the glass surface surrounding the dots: the formation of a PEG layer prevents (or at least minimizes) any nonspecific protein and DNA adsorption from taking place. 11. When PEGylation is done, rinse the glass slide with acetone and then ethanol, and blow dry with Ar or N2. 12. Prepare a solution of PBS with 10 mg/mL of streptavidin and 1  mg/mL of albumin in PBS (see Note 22). Place the dry slide in 2 mL of such solution, seal the container with Parafilm, cover with Al foil, and incubate on the shaker at RT for 2 h (see Note 23). 13. Rinse the sample thoroughly with PBS (do not let the sample dry). Then, move the sample to a fresh well filled with PBS and incubate it for at least 30 min, while storing it foiled, on a shaker (see Note 24). 14. Meanwhile, prepare a solution of biotinylated DNA (see Note 25). For biotinylated dsDNA, prepare a solution of two complementary strands of 2 mM in TPBS at RT, one biotinylated at the 5¢ end and its complement labeled with a fluorophore (Cy3 in our case) at the 5¢ end. Heat the solution gradually (steps of approximately 2°C per min) to a temperature of 65°C; leave the solution at this temperature for 15 min. Then ramp the temperature up to 75°C and leave for 1 h 30 min. Then proceed in the reverse order, ramping down the temperature, let cool at 65°C for 15 min and then take gradually to RT. When the solution has reached RT, it can be employed as described in the next step. 15. Incubate the (wet) coverslip from step 13, in at least 1.5 mL of the 2  mM biotinylated dsDNA solution prepared as described above (see Note 25); cover the container with Al foil during the incubation and seal with Parafilm. The minimum incubation time recommended is 3 h (see Note 26). 16. Take the sample out of the previous solution and rinse it with PBS. Then move the sample to a fresh well and incubate it for 30 min in PBS, while storing it foiled, on a shaker. The sample is now ready: i.e., Au/Pd nanodot is properly and specifically functionalized with the oligonucleotide of interest.

Controlled Confinement of DNA at the Nanoscale

179

In the next section, we will describe how we characterized the functionalized surface, verifying by fluorescence microscopy the presence of Cy3-labeled dsDNA immobilized on sub-50-nm Au/Pd nanodots on glass. 3.3. Fluorescence Microscopy 3.3.1. Sample Preparation

Inverted microscopes are designed to accept coverslips mounted on standard glass slides. We have devised a simple mounting scheme with our coverslips for fluorescence microscopy (see Fig. 4). It is an open design, which allows for additional processes to be carried out in situ on the microscope. Microfluidics can and have been used: here, we present a setup offering most of the functionality without the complexity of microfluidics. 1. Remove the bio-functionalized sample from PBS rinse solution and place face up on a clean surface. 2. Apply vacuum grease to the edge of a cloning ring and place it on the coverslip, with the pattern centered as well as can be done by hand. The grease creates a watertight seal. 3. Fill the ring with buffer solution (PBS). At this point, check for leaks around the ring. If any are detected, a slight twist of the ring is often adequate to seal the leak; if not, additional grease applied to the region of the leak will solve the problem. 4. Suspend the slide across the cutout section of the aluminum holder of the microscope. We use an aluminum plate cut to fit

Fig. 4. Photograph (Nikon D80) of the fully functionalized nanopatterned slide prepared for fluorescence microscopy. The coverslip sits atop an aluminum carrier, machined to fit the microscope mounting hardware, and is suspended above a cutout section allowing clearance for the optics. It is held in place by a thin bead of grease along each edge. A cloning ring adhered with a thin bead of grease encircles the patterned region and contains the necessary buffer solution.

180

Palma et al.

the microscope mount, with a gap cut into it to allow for the objective. 5. Use vacuum grease to seal the coverslip to the aluminum holder along the edges. Proceed to the microscope for imaging. 3.3.2. Microscopy

An inverted fluorescence microscope capable of epifluorescence microscopy is used to image the samples. Oil immersion lenses with 60× and 100× magnification are best used for imaging of the nanopatterns. The camera we use is a photometrics Cascade II; it is cooled to −70°C and has on-chip amplification for low noise, high-sensitivity imaging. 1. Make sure the optics are aligned. 2. Find the pattern in differential image contrast (DIC) mode. 3. Switch from DIC to the Cy3 (excitation 550 nm/emission 568 nm) fluorescence channel. Use live imaging with <100-ms exposure to adjust the focus again as there is a slight difference in focal plane with the different optics. 4. Proceed to record desired data, static images, or time sequences (see Note 27). Figure 5 displays an epifluorescence microscopy image of the Cy3-labeled dsDNA on the sub-50-nm Au/Pd nanodot array.

Fig.  5. Epifluorescence image of the nanopatterned dots functionalized with a Cy3labeled dsDNA: excitation 550 nm, emission 568 nm. Every fifth sub-50-nm dot, a 1 mm register dot is clearly visible, and brighter in the fluorescence image. Exposure time: 300 ms. Image size 80 mm × 80 mm.

Controlled Confinement of DNA at the Nanoscale

181

4. Notes 1. Make sure all glassware used is clean, i.e., washed with detergent, rinsed in DI water, rinsed with ethanol, and blown dry with a stream of inert gas (Ar or N2). 2. When handling samples, especially washing and drying with tweezers, always make sure to rinse and blow dry toward the tweezers; this prevents the transfer of any contaminants from gloves and tweezers onto the samples. 3. Clean the coverslip immediately prior to the resist deposition; if it is allowed to sit in air, it will accumulate contaminants. 4. Time is of the essence; the quicker one works while spinning resist layers, the less likely the surface is to be contaminated. 5. Make sure the coverslip is cool before putting resist on, and never let resist sit for more than a few seconds before spinning. The solvent will evaporate slightly, especially at the edge, resulting in an uneven coating, which may or may not be obvious, but will affect the obtained features. 6. Observe the sample after each layer of resist is deposited. If there is streaking, this means the surface was not properly cleaned. Discard the sample and start again. Properly spun resist should uniformly coat the surface to the edges, looking smooth and glassy. 7. Larger features (micron sized) are written at spot size 5, while the nano-features are written with spot size 1. 8. The nanodots themselves are not visible in an optical microscope. The 1  mm registers are visible, and are intended to provide points of reference for the location of the nanodots, and to assist with focusing on the optical microscope. If you wish to characterize the nanodots, AFM is the best method. 9. It is important to correct for beam tilt and stigmation. If the beam is tilted, the projection of the feature will be oval rather than circle. This will affect feature shape and minimum achievable size, not to mention the fact that the focal plane of the beam will not be parallel with the plane of the sample. Stigmation results from the lens not being perfectly round, in fact no lens is. If you imagine the lens is an ellipse, and break the focus into orthogonal x–y coordinates aligned along the short and long axis of the lens, it becomes apparent that the lens will have different depth of focus for each axis. This shows when imaging has the ability to get one edge in perfect, crisp focus, while edges not parallel with the sharp edge will look out of focus, worsening as the angle of the two edges approaches 90° with respect to the properly focused edge.

182

Palma et al.

Adjusting the stigmation will correct for the aberration, resulting in better focus and thus smaller features. 10. A faceted, crystalline particle such as Al2O3 is a good target for stigmation correction. When all edges appear clear, regardless of orientation, stigmation is minimized. A commonly used technique is to sprinkle these particles at the corners of the slide and use them to define the four-point focus as well. However, the edge is actually a few hundred nanometer above the surface, so the focus is not entirely accurate. This method is best for features down to ~35 nm; below that, the “spot” method should be used (see below). 11. “Spot” method: This requires using a 15-nm film of aluminum as the discharge layer, instead of Aquasave. Focus as best you can on the surface. Use the spot function in the scan menu to expose a spot on the surface manually, and then subsequently optimize focus and stigmation on the spot until it is sharp. This must be done iteratively; the resulting spot comes out smaller and sharper each time, enabling further refinement of the stigmation and focus. When the spot size is no longer getting smaller, focus is optimized. Always adjust focus first, before stigmation is adjusted. This is the best method for performing four-point focus, as it verifies that the focus is optimized for exposure of the smallest possible area. Spotting four corners in a box, immediately around the area to be written, ensures the best focus for writing. The 15-nm film of aluminum should be removed in NaOH prior to development. 12. We recommend keeping the sonicator and water/IPA development solution in a refrigerator prior to and during development. If that is not possible, be sure that the bath and H2O/IPA are chilled to 4°C before proceeding. 13. Development solutions can be reused, but should be changed periodically (every ten uses) and stored at 4°C. 14. The development and quenching in room temperature IPA are time critical, so prepare everything in advance and move as quickly as possible. It is recommended to bring the IPA over to the cold sonication and after 1 min, transfer the sample directly into the IPA at room temperature. 15. It is important to prevent any re-deposition of the lifted off metal. 16. When placing the samples in the liftoff, keep them mostly vertical, but tilted slightly face down so the metal will fall away from the surface. Use one of the slotted Teflon racks as a holder. 17. Never allow the samples to go dry in the acetone or upon removal during the rinsing steps.

Controlled Confinement of DNA at the Nanoscale

183

18. Always use separate dedicated syringes and needles for the anhydrous solvents employed. Rinse the needles and syringes after use, and store them in an oven at 70°C: the syringe and the needle used for toluene should be rinsed in acetone and then in ethanol. Blow dry the needles before storage in oven. 19. Incubate the samples in the biotin solution directly after plasma cleaning: allowing the sample to sit too long in air has a detrimental effect over proper SAM formation. 20. When preparing the PEG-silane solution, make sure to work in a cold environment chamber (temperature below 5°C). 21. A functionalization procedure incubating the sample in the PEG-silane solution first and then thiolating the gold works, but it gives rise to a less reliable passivation, probably due to physisorption of biotin–alkane chains on the PEG layer, which can then play an active role as functional spots for the attachment of streptavidin on surfaces. 22. The purpose of albumin is to block nonspecific binding between streptavidin and any passivation defects on the glass surface. 23. It is recommended to incubate the sample in the streptavidin solution immediately after drying. 24. It is possible to check the intermediate step by using a fluorescent-labeled streptavidin (Invitrogen): suggested fluorophores include Alexa 488 and Cy5. It is also possible to use avidin or neutravidin, instead of streptavidin, although avidin is not recommended due to a high degree of nonspecific adsorption. 25. Either single-stranded (ssDNA) or dsDNA can be immobilized on the metal nanodots depending on the application: dsDNA nanodomains, for example, can be useful to study protein– DNA interactions, while bound biotinylated ssDNA is available for in situ hybridization to its complementary sequence, a process of interest for bioanalytical applications. Furthermore, immobilized ssDNA can serve as an anchoring point on the surface to drive the self-assembly of biological and inorganic nano-objects properly functionalized with the complementary strand: the information encoded in the double helix can be a powerful tool for the fine control of such organization, noteworthy at the nanoscale. For ssDNA, at step 15 of Subheading 3.2, incubate the sample in a ssDNA solution at a concentration of 2 mM in TPBS at RT for at least 3 h (up to 12–18 h, i.e., overnight). In order to proceed with a RT in situ hybridization on the nanodots (fabricated as described above and functionalized with ssDNA), incubate the sample in a 2 mM solution of the complementary strand in TPBS adding 3 mM of NaCl and 125 mM of MgCl2. The purpose of the high salt concentration is to screen any repulsive interactions

184

Palma et al.

taking place between the phosphate groups of the complementary DNA strands, as they can prevent an effective hybridization from taking place at RT. The incubation should be done in at least 1.5 mL of solution and left in a container sealed with Parafilm and covered with Al foil overnight (minimum incubation time: 12 h, up to 36 h): do not shake the container during the incubation. After incubation, rinse the sample in PBS and incubate in PBS for 1 h: the rinsing is very important to get rid of residual salt physisorbed on the substrate. 26. The incubation with biotinylated dsDNA can go up to 12–18 h (i.e., overnight). 27. Imaging is always a balancing act between exposure time and amplification to achieve detection threshold. Longer exposure times integrate more light and result in better signal to noise, but result in loss of time resolution and lead to faster photobleaching. Increasing the gain amplification allows for lower exposures but increases the effect of noise, which can swamp out the fluorescence signal. We typically work at exposures of 100–300 ms as this provides adequate signal to noise with most fluorophores we work with.

Acknowledgments We gratefully acknowledge support from the office of Naval Research under award number N00014-09-1-1117, National Institutes of Health through award number PN2EY016586 under the NIH Roadmap for Medical Research, and from the National Science Foundation under NSF award number EF-05-07086 and award number CHE-0936923. Additional support from the Nanoscale Science and Engineering Initiative of the National Science Foundation under NSF Award Number CHE-0641523 and from the New York State Office of Science, Technology, and Academic Research (NYSTAR) is also gratefully acknowledged. References 1. Whitesides, G. M. (2003) The ‘right’ size in nanobiotechnology, Nat Biotechnol 21, 1161–1165. 2. Torres, A. J., Wu, M., Holowka, D., and Baird, B. (2008) Nanobiotechnology and  cell biology: Micro- and nanofabricated surfaces to investigate receptormediated  signaling, Ann Rev Biophys 37, 265–288. 3. Rosi, N. L., and Mirkin, C. A. (2005) Nanostructures in biodiagnostics, Chem Rev 105, 1547–1562.

4. Langer, R., and Tirrell, D. A. (2004) Designing materials for biology and medicine, Nature 428, 487–492. 5. Wong, L. S., Khan, F., and Micklefield, J. (2009) Selective Covalent Protein Immobilization: Strategies and Applications, Chem Rev 109, 4025–4053. 6. Williams, B. A. R., Lund, K., Liu, Y., Yan, H., and Chaput, J. C. (2007) Self-assembled peptide nanoarrays: An approach to studying protein-protein interactions, Angew Chem Int Edit 46, 3051–3054.

Controlled Confinement of DNA at the Nanoscale 7. Winssinger, N., Pianowski, Z., and Debaene, F. (2007) Probing biology with small molecule microarrays (SMM), Top Curr Chem 278, 311–342. 8. (2004) Nanobiotechnology Wiley-VCH, Weinheim. 9. (2005) Nanofabrication Towards Biomedical Applications, Wiley-VCH, Weinheim. 10. Tan, P. K., Downey, T. J., Spitznagel, E. L., Xu, P., Fu, D., Dimitrov, D. S., Lempicki, R. A., Raaka, B. M., and Cam, M. C. (2003) Evaluation of gene expression measurements from commercial microarray platforms, Nucleic Acids Res 31, 5676–5684. 11. Becerril, H. A., and Woolley, A. T. (2009) DNA-templated nanofabrication, Chem Soc Rev 38, 329–337. 12. Niemeyer, C. M. (2001) Nanoparticles, proteins, and nucleic acids: Biotechnology meets materials science, Angew Chem Int Edit 40, 4128–4158. 13. Drummond, T. G., Hill, M. G., and Barton, J. K. (2003) Electrochemical DNA sensors, Nat Biotechnol 21, 1192–1199. 14. Rant, U., Arinaga, K., Scherer, S., Pringsheim, E., Fujita, S., Yokoyama, N., Tornow, M., and Abstreiter, G. (2007) Switchable DNA interfaces for the highly sensitive detection of label-free DNA targets, P Natl Acad Sci USA 104, 17364–17369. 15. Brucale, M., Zuccheri, G., and Samori, B. (2006) Mastering the complexity of DNA nanostructures, Trends Biotechnol 24, 235–243. 16. (2002) Methods in Molecular Biology Vol. 170, Humana Press, Totowa, NJ. 17. (2007) Nanobiotechnology II, Wiley-VCH, Weinheim. 18. Heise, C., and Bier, F. F. (2005) Immobilization of DNA on microarrays, Top Curr Chem 261, 1–25. 19. Luderer, F., and Walschus, U. (2005) Immobilization of oligonucleoticles for biochemical sensing by self-assembled monolayers: Thiol-organic bonding on gold and silanization on silica surfaces, Top Curr Chem 260, 37–56. 20. Murphy, J. N., Cheng, A. K. H., Yu, H. Z., and Bizzotto, D. (2009) On the Nature of DNA Self-Assembled Monolayers on Au: Measuring Surface Heterogeneity with Electrochemical in Situ Fluorescence Microscopy, J Am Chem Soc 131, 4042–4050. 21. Shumaker-Parry, J. S., Zareie, M. H., Aebersold, R., and Campbell, C. T. (2004) Microspotting streptavidin and doublestranded DNA Arrays on gold for highthroughput studies of protein-DNA

22.

23.

24.

25.

26.

27.

28.

29.

30.

31. 32.

185

interactions by surface plasmon resonance microscopy, Anal Chem 76, 918–929. Smith, C. L., Milea, J. S., and Nguyen, G. H. (2005) Immobilization of nucleic acids using biotin-strept(avidin) systems, Top Curr Chem 261, 63–90. Takahashi, S., Matsuno, H., Furusawa, H., and Okahata, Y. (2007) Kinetic analyses of divalent cation-dependent EcoRV digestions on a DNA-immobilized quartz crystal microbalance, Anal Biochem 361, 210–217. Ladd, J., Boozer, C., Yu, Q. M., Chen, S. F., Homola, J., and Jiang, S. (2004) DNA-directed protein immobilization on mixed self-assembled monolayers via a Streptavidin bridge, Langmuir 20, 8090–8095. Whitesides, G. M., Ostuni, E., Takayama, S., Jiang, X. Y., and Ingber, D. E. (2001) Soft lithography in biology and biochemistry, Annu Rev Biomed Eng 3, 335–373. Noh, H., Hung, A. M., Choi, C., Lee, J. H., Kim, J. Y., Jin, S., and Cha, J. N. (2009) 50 nm DNA Nanoarrays Generated from Uniform Oligonucleotide Films, Acs Nano 3, 2376–2382. Yu, A. A., Savas, T. A., Taylor, G. S., GuiseppeElie, A., Smith, H. I., and Stellacci, F. (2005) Supramolecular nanostamping: Using DNA as movable type, Nano Lett 5, 1061–1064. Demers, L. M., Ginger, D. S., Park, S. J., Li, Z., Chung, S. W., and Mirkin, C. A. (2002) Direct patterning of modified oligonucleotides on metals and insulators by dip-pen nanolithography, Science 296, 1836–1838. Rodolfa, K. T., Bruckbauer, A., Zhou, D. J., Korchev, Y. E., and Klenerman, D. (2005) Two-component graded deposition of biomolecules with a double-barreled ­nanopipette, Angew Chem Int Edit 44, 6854–6859. Cherniavskaya, O., Chen, C. J., Heller, E., Sun, E., Provezano, J., Kam, L., Hone, J., Sheetz, M. P., and Wind, S. J. (2005) Fabrication and surface chemistry of nanoscale bioarrays designed for the study of cytoskeletal protein binding interactions and their effect on cell motility, J Vac Sci Technol B 23, 2972–2978. (1990) Methods in Enzymology Vol. 184, Academic press. Nelson, K. E., Gamble, L., Jung, L. S., Boeckl, M. S., Naeemi, E., Golledge, S. L., Sasaki, T., Castner, D. G., Campbell, C. T., and Stayton, P. S. (2001) Surface characterization of mixed self-assembled monolayers designed for streptavidin immobilization, Langmuir 17, 2807–2816.

Chapter 13 Templated Assembly of DNA Origami Gold Nanoparticle Arrays on Lithographically Patterned Surfaces Albert M. Hung and Jennifer N. Cha Abstract Artificial DNA nanostructures such as DNA origami have garnered significant interest as templates for sub-20  nm lithography because their rational design allows for the incorporation of binding sites to assemble nanocomponents with 6 nm resolution. In addition, their overall size of 100 nm is easily accessible by top-down lithographic methods. Combining the strengths of top-down lithography and bottom-up self-assembly using DNA nanostructures may provide a commercially viable route to fabricating electronic and photonic devices with nanometer-scale features. We have demonstrated just such a comprehensive process in which 5 nm gold nanoparticles are first assembled in high yield on DNA origami. The constructs are then organized, rinsed, and dried on patterned silicon substrates, yielding large area arrays of both origami and nanoparticles. Key words: DNA origami, Gold nanoparticles, Template-directed self-assembly, Lithography

1. Introduction Nanomaterial-based devices hold great promise for many electronic and photonic applications, but cost-effective methods for reliably defining and addressing sub-20  nm components are lacking. Traditional “top-down” lithographic methods such as photolithography and electron beam (e-beam) lithography offer superior spatial registry and defect control, while “bottom-up” self-assembly can achieve higher resolution and chemical precision. Merging the two strategies could yield an effective fabrication process, but this is hindered by a gap between roughly 10 and 100 nm which both strategies have difficulty bridging (1, 2). Synthetic DNA nanostructures such as two-dimensional DNA “origami”(3) may be able to span this gap because they can be rationally designed to

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_13, © Springer Science+Business Media, LLC 2011

187

188

Hung and Cha

be hundreds of nanometers in size yet retain chemical functionality potentially down to the resolution of a single DNA helix (2 nm) (4, 5). We have demonstrated the assembly of DNA origami with controlled placement and orientation onto a lithographically patterned silicon substrate and used them as templates to direct the arrangement of 5 nm gold nanoparticles (6). To successfully employ DNA origami as a template for directing the assembly of nanoscale materials with sub-20 nm precision, three key processes are needed: (1) binding of the nanoscale materials (i.e., nanoparticles) to DNA templates with extremely high positional accuracy and yields, (2) stable and controlled deposition of the DNA templates onto patterned substrates, and (3) rinsing and drying of the sample without destroying or removing the DNA or nanoparticles. In addition, all methods must be compatible with one another. Binding of nanoparticles is achieved by conjugating them with single stranded DNA (ssDNA) and hybridizing them to complementary strands attached to the DNA origami. The surface of a thermal oxide layer on silicon is patterned with hydrophilic and hydrophobic domains by e-beam lithography. The origami–nanoparticle complexes selectively adsorb to the hydrophilic features and can be dried by rinsing in ethanol solutions.

2. Materials 2.1. Synthesis and Purification of DNA Origami

1. Single-stranded m13mp18 phage DNA (M13) at 10  nM concentration in water or buffer (see Note 1). Store at −20°C and allow to fully defrost before use. 2. Complete set of DNA staple strands, each at 100 mM concentration in water or buffer (Integrated DNA Technologies, Coralville, IA) (see Note 2). Store at −20°C and allow to fully defrost before use. 3. 10× TAE/125 mM MgCl2 buffer: 400 mM Tris acetate, 10 mM ethylenediamine tetraacetic acid (EDTA), 20  mM NaCl, and 125  mM MgCl2. This can be diluted one part to nine parts water or 125 mM MgCl2 in water to get 1× TAE/12.5 mM MgCl2 or 1× TAE/125 mM MgCl2 buffer, respectively. 4. 100  kDa MWCO centrifuge filters such as Microcon YM (Millipore, Billerica, MA) or Nanosep (Pall, Port Washington, NY).

2.2. Preparation of DNA-Coated Gold Nanoparticles

1. Citrate-stabilized 5 nm gold nanoparticle suspension (British Biocell International, Cardiff, UK). Store at 4°C. 2. Thiol-modified complementary DNA at a known concentration in water or buffer. Store at −20°C and allow to fully defrost before use.

Templated Assembly of DNA Origami Gold Nanoparticle Arrays

189

3. Bis ( p-sulfonatophenyl)phenylphosphine dihydrate dipotassium salt (BSPP) (Aldrich). 4. Sodium chloride (NaCl). 5. 30 kDa MWCO centrifuge filters. 2.3. Preparation of the Patterned Silicon Substrate

1. Silicon wafers with 100  nm thermal oxide coating, primed with hexamethyldisilazane (HMDS), and coated with a lithographically patterned photoresist layer (see Note 3). 2. Isopropyl alcohol (IPA). 3. N-methylpyrrolidone (NMP).

2.4. Assembly of DNA Origami and Gold Nanoparticles on Patterned Substrates

1. 50% Ethanol rinsing solution: 50% ethanol and 50% water (v/v). 2. 90% Ethanol rinsing solution: 90% ethanol and 10% water (v/v).

3. Methods DNA hybridization is an easy way to bind nanoparticles to DNA templates, and careful design, synthesis, and purification of both components are essential to achieving high yields (6). Complete coating of the nanoparticle surface with DNA is necessary as it helps the nanoparticles to remain stable in the high ionic strength buffers used with DNA origami. Purification is especially important as excess ssDNA will compete for binding and depress yields. When binding the nanoparticles to the origami in solution, the concentrations of both must be high enough to obtain good yields, but not too high as aggregation will occur. DNA adsorbs to hydrophilic silicon oxide at specific concentrations of multivalent cations, which are believed to coordinate to the negatively charged DNA and bridge it to the negatively charged silica surface (7). Placement of the origami can be controlled by functionalizing the surface with a hydrophobic blocking layer (HMDS) patterned with holes exposing “sticky” hydrophilic oxide (6, 8). The holes are created by patterning a layer of resist with e-beam lithography and burning away the exposed HMDS under oxygen plasma. Lastly, a simple rinsing procedure with ethanol can be used to remove the buffer and dry the complete assembly intact. While this final step is not necessary for template assembly, it is extremely useful in a practical sense for any subsequent characterization or device testing.

190

Hung and Cha

3.1. Synthesis and Purification of DNA Origami

1. Mix equal volumes of each 100 mM DNA staple solution to make a master staple mixture. This mixture can be stored at −20°C for several months. 2. For a single 50  mL aliquot of DNA origami solution, mix 5 mL M13, 5 mL 10× TAE/125 mM MgCl2 buffer, enough of the staple mixture to achieve a final concentration of 100  nM of each staple (roughly 10  mL; see Note 4), and enough water to bring the final volume up to 50 mL. Usually, 200–400 mL of the origami mixture is made at once and split into four to eight 50 mL aliquots. 3. In a standard thermal cycler, heat the mixture up to 95°C for 5 min and cool down to 20°C or below at the slowest rate allowed by the machine. Depending on the machine, this can take 1.5–4 h. 4. Load 100 mL of origami solution into the 100-kDa MWCO centrifuge filter (see Note 5). To wash the origami, add 400– 500 mL of 10× TAE/MgCl2 buffer to the solution, centrifuge until 25–50 mL is left, and discard the filtrate (see Note 6). Repeat this process two more times. 5. Recover the origami solution and add buffer to bring the volume up to 50 mL if necessary. This will be referred to as a filtered 2× origami solution. UV-vis absorption spectroscopy can be used to estimate the concentration of DNA origami in solution by measuring the absorbance at 260 nm and using an extinction coefficient of roughly 108 M−1 cm−1 (see Note 7). 6. DNA origami assembly and filtration can be confirmed by standard gel electrophoresis of the filtered and unfiltered solutions in a 1% (w/v) agarose gel. An example of a gel is shown in Fig. 1. DNA staples show up as a broad, fast band, while two slower bands indicate the presence of DNA origami and what is thought to be misfolded structures. The DNA origami is the faster of these two bands and should be the only one remaining after filtration. 7. Long-term storage of all origami solutions should be at −20°C. 4°C is adequate for short-term storage of 1–2 days. Unfiltered solutions are stable for many months, but filtered solutions may begin to degrade within a few weeks to a few months depending on the frequency of handling. The integrity of the DNA structures can be checked by gel electrophoresis or liquid-cell atomic force microscopy (AFM) of origami solution deposited on mica (see Note 8).

3.2. Preparation of DNA-Coated Gold Nanoparticles

1. Measure 20 mL of gold nanoparticle solution into a beaker or bottle and add 6 mg BSPP. Let it sit overnight at room temperature, covered and with mild stirring or agitation.

Templated Assembly of DNA Origami Gold Nanoparticle Arrays

191

Fig. 1. Gel electrophoresis of DNA origami and 5 nm gold nanoparticle solutions. (a) Electrophoresis of unfiltered (top) and filtered (bottom) solutions of triangular DNA origami showing the removal by centrifuge filtration of excess staples (fast, broad band) as well as what is suspected to be poorly folded M13 (slowest band). (b) BSPP-stabilized gold nanoparticles (bottom) slow down after conjugation with thiolated 30-mer thymine (T30) DNA strands (top).

2. Add NaCl slowly to the solution until the nanoparticles start to aggregate. The color of the solution should change from red to purple to bluish-brown (see Note 9). 3. Centrifuge at 10,500 rcf for 15 min and discard the supernatant. Resuspend the pelleted nanoparticles in 500 mL water. At this concentration, the solution should be a very dark red. 4. Filter out remaining salt by loading the nanoparticle solution into a 30-kDa MWCO centrifuge filter and spin down to <40 mL. Add 500 mL water and repeat. Resuspend the nanoparticles in 100 mL water. All nanoparticle suspensions must be stored at 4°C. Do not freeze. 5. Determine the nanoparticle concentration by UV-vis absorption, measuring the absorbance at 520  nm. The extinction coefficients are 1.205 × 107  M−1 cm−1 for 5  nm particles, 1.057 × 108 for 10 nm, and 4.303 × 108 for 15 nm. 6. Partition out a desired volume of nanoparticles to be functionalized with DNA. Add thiolated DNA at a molar ratio of 200:1 DNA strands to particles (see Note 10). Let the solution sit overnight at room temperature. 7. Filter the nanoparticle solution through 30 or 100 kDa centrifuge filters four times, each time adding water to bring the volume up to 500 mL and centrifuging down to 25 mL. For easier handling, water can be added to bring the final volume up to 50–200 mL if desired.

192

Hung and Cha

8. Confirm nanoparticle functionalization by gel electrophoresis using a 1% (w/v) agarose gel. A 50% (v/v) glycerol/water mixture can be used as a loading buffer (1.5  mL per 5  mL nanoparticle solution). The DNA-coated nanoparticles should move slower and as a single band. Broadening or multiple bands suggest incomplete coverage (9). 9. Measure the nanoparticle concentration by UV-vis absorption. Prepare a 40 nM solution of nanoparticles in 10× TAE/125 mM MgCl2. This can be made ahead of time and stored for a few months. The color of the solution should remain light red or pink, not purple (see Note 11). 3.3. Preparation of the Patterned Silicon Substrate

1. Immediately before use, remove the HMDS in the patterned areas with a short etch in oxygen plasma to expose a hydrophilic oxide surface (see Note 12). 2. Strip the remaining photoresist by sonicating the substrates in NMP at 55°C for 20  min (see Note 13). Briefly rinse the substrates in IPA and then water, and dry the substrates with a N2 stream.

3.4. Assembly of DNA Origami and Gold Nanoparticles on Patterned Substrates

1. Mix 5  mL of 40  nM DNA-coated gold nanoparticles and with 5 mL of 2× origami solution and anneal the solution at 37°C for 20  min. Add 10  mL of 10× TAE/125  mM MgCl2 to dilute the mixture (see Note 14). Solutions of origami–nanoparticle conjugates must be made fresh before use. 2. Place the substrate in a Petri dish or other container that is lined on the inside with a wet Kimwipe. Pipette 2 mL of the origami–nanoparticle solution onto the substrate to cover the patterned area, then loosely cap or cover the container (see Note 15). 3. Allow the DNA origami to adsorb to the substrate for 3 h at room temperature (see Note 16). 4. Prepare small beakers of three solutions: 10× TAE/125 mM MgCl2 buffer, 50% ethanol rinsing solution, and 90% ethanol rinsing solution. Take the substrate with droplet of origami solution still on it and dip it first in buffer for 2 s, then in 50% ethanol for 5 s, then immerse in 90% ethanol for 1 h. Do not agitate (see Note 17). 5. Remove the sample from ethanol and allow it to air dry. Image the samples by standard tapping mode AFM or other suitable techniques. An example of templated DNA origami and gold nanoparticles is shown in Fig. 2.

Templated Assembly of DNA Origami Gold Nanoparticle Arrays

193

Fig. 2. AFM tapping mode height image of templated arrays of triangular DNA origami and 5 nm gold nanoparticles. The triangles were modified with four adenine 30-mer (A30) strands at each corner, and the nanoparticles were coated with thymine 8-mer (T8) strands. The e-beam patterned substrate consisted of alternating rows of triangles (100 nm on each side) pointing left and pointing right (inset ). Scale bar is 500 nm.

4. Notes 1. Ultrapure water (18.2  MW  cm resistivity) is recommended for making all solutions and is what is meant by “water” in this protocol unless otherwise stated. Buffer for storing M13 and staple strands may be homemade TE buffer (10  mM Tris–HCl, 1 mM EDTA, pH 8.0) or whatever is purchased with the DNA. 2. Staples may be purchased prepurified at 100 mM concentration (“lab-ready”) in 96-well plates. The complete list of all roughly 200 staple sequences needed for any given origami shape and the corresponding diagram showing the position of each strand within the origami is too extensive to be given here but is available elsewhere (3). Modification of origami with DNA-based binding sites involves (1) choosing a location on the origami to add a binding site, (2) using the diagram and sequence list to identify the staple strand at that site, (3) appending the binding sequence to one end of the chosen staple sequence, and (4) replacing the selected staples with the modified ones upon making the master staple mixture. 3. Native oxide or other thicknesses of thermal oxide also work. DNA origami can also adsorb to quartz under similar conditions presented here, but it is more difficult to deposit a stable HMDS layer. This protocol assumes knowledge of or access to some

194

Hung and Cha

standard lithographic processes such as photolithography or e-beam lithography. Briefly, the oxide surface is cleaned and then functionalized with HMDS by either vapor or solution deposition. A photoresist layer is spin-coated on top and an electron beam is used to expose a pattern in the resist. The pattern is then developed in solvents that remove the resist in the exposed areas all the way down to the HMDS layer. Substrates should be kept in this state until use. 4. For the master staple mixture made as suggested, achieving 100  nM concentration of each staple in a 50  mL volume translates to adding a volume (in microliters) of the master mix equal the number of staples divided by 20 (e.g., for 208 staples, add 208/20 = 10.4 mL of the staple mixture). Smaller concentrations of staples may sometimes be used to make origami as well, in which case the calculation changes accordingly. 5. The effectiveness of centrifuge filtration in removing excess staples and M13 can depend on the brand of filter used and the buffer salt concentration. Generally, 100 kDa MWCO filters are sufficient, and the washing can be done with buffer concentrations down to 1× TAE/12.5 mM MgCl2 if desired (125 mM MgCl2 is required for DNA adsorption to oxide, however). 30 kDa MWCO Nanosep filters can also be used if origami loss is a concern, but then only buffers with 125 mM MgCl2 may be used. 6. All results presented use origami solutions filtered with 10× TAE/125 mM MgCl2, but 1× TAE/125 mM MgCl2 can also be used. 1× TAE may actually perform slightly better because monovalent salts tend to inhibit DNA adsorption to oxide surfaces (7). When washing, do not centrifuge down too close to dryness or else the origami will be lost to aggregation and adhesion to the filter. 7. Remember to use an aliquot of solution diluted such that the absorbance at the wavelength of interest does not exceed 1. A concentration of 1.0–1.5 nM DNA origami (2× solution) is recommended (see Note 14). DNA origami retention during filtering can vary depending on origami shape, with “stiffer” structures such as triangles giving higher yields. 8. For liquid-cell AFM on mica, drop 2–5 mL of origami solution on the mica and add 40–50  mL of 1× TAE/12.5  mM MgCl2 buffer. Allow 5–10  min for adsorption and image using standard protocols. DNA origami will adhere to mica in 12.5 mM MgCl2 but not 125 mM MgCl2. For silicon oxide surfaces, the case is exactly the opposite. 9. Add one to three small spatula-scoops of salt at a time and allow to dissolve before seeing if more is required.

Templated Assembly of DNA Origami Gold Nanoparticle Arrays

195

Larger nanoparticles will precipitate out faster than smaller ones. This entire protocol is one found to work well with 5  nm nanoparticles. Some parameters may need to be adjusted for larger nanoparticles. 10. This procedure aims to coat the entire surface of the nanoparticle with DNA. Larger nanoparticles may require more DNA per particle. Complete coverage is especially important for stabilizing the nanoparticles in the high ionic strength buffers used for patterning DNA origami. 11. 1× TAE/125 mM MgCl2 should also work. Purple color indicates aggregation and insufficient stabilization of the nanoparticles by DNA. Larger particle sizes and some DNA sequences may be more likely to cause precipitation in high salt solutions. If precipitation is a problem, one can add gradually increasing concentrations of MgCl2 to the nanoparticle solution during the incubation with thiolated DNA (10). This should cause the DNA strands to coil more tightly and allow for a denser coating of DNA on the nanoparticle surface. 12. It was found that UV-ozone was adequate for removing HMDS from large-feature patterns (300  nm lines) but not from finer e-beam patterned features. We employed a 7 s etch in a PX 250 (March Plasma Systems, Concord, CA). Etching times may vary between machines, and a directional etch may not be necessary for such short etch times. 13. Photoresist stripping conditions can vary depending on the resist, and milder conditions are preferred whenever possible because the HMDS layer can be hydrolyzed off. It is not uncommon for the static water contact angle to decrease from >60° freshly silanized to around 40° after stripping, yet there is still sufficient HMDS to inhibit DNA adsorption. 14. The concentrations of DNA origami and nanoparticles are free to be tuned for better results. Higher values of both result in better binding yields of nanoparticles to origami, but can also cause aggregation. Upon adsorption to patterned oxide surfaces, excess nanoparticles and DNA origami will also be deposited if the concentrations are too high. Low origami concentrations can leave too many empty areas of oxide. The protocol given anneals the DNA origami and nanoparticles together at relatively high concentrations to improve binding yields, then dilutes to mixture to allow for cleaner adsorption to patterned substrates. 15. A general technique for preventing solution evaporation is to moisten a single Kimwipe with water, roll it up into a cylinder, and use it to corral the area in the container into which the silicon substrates are placed. Do not allow the substrates to touch the Kimwipe. Loosely cap but do not seal the container

196

Hung and Cha

or else unwanted water condensation can occur. E-beam patterned areas are frequently small (1 mm x 1 mm or less), in which case 2 mL of origami solution is sufficient. 16. Longer adsorption times are not recommended as the HMDS will hydrolyze off. The sample may be imaged by liquid cell AFM after incubation but before drying. Use 125 mM MgCl2 buffer (1× or 10× TAE) for imaging. With careful handling, the sample can still be rinsed and dried with ethanol solutions as described. 17. The buffer should remove most of the excess nanoparticles, the 50% ethanol removes most of the buffer, and the 90% ethanol removes the remaining buffer and salt while fixing the DNA origami to the substrate (11). Water is necessary in the ethanol solutions to dissolve the trace salts, but prolonged immersion in anything less than 85% ethanol results in desorption of DNA.

Acknowledgments The authors thank Luisa Bozano for providing the lithographically patterned substrates and Christine M. Micheel for helping to develop the methods for binding nanoparticles to DNA origami. This work was financially supported by the Center on Polymer Interfaces and Macromolecular Assemblies (Award Number: NSF DMR 0213618), the Office of Naval Research (Award Number: N00014-09-01-0250), and UCSD startup funds. References 1. Seeman, N. C. (2003) DNA in a material world Nature 421, 427–431. 2. Whitesides, G. M., Grzybowski, B. (2002) Self-assembly at all scales Science 295, 2418–2421. 3. Rothemund, P. W. K. (2006) Folding DNA to create nanoscale shapes and patterns Nature 440, 297–302. 4. Aldaye, F. A., Palmer, A. L., Sleiman, H. F. (2008) Assembling materials with DNA as the guide Science 321, 1795–1799. 5. Lin, C. X., Liu, Y., Rinker, S., Yan, H. (2006) DNA tile based self-assembly: Building complex nanoarchitectures Chemphyschem 7, 1641–1647. 6. Hung, A. M., Micheel, C. M., Bozano, L. D., Osterbur, L. W., Wallraff, G. M., Cha, J. N. (2010) Large area spatially ordered arrays of gold nanoparticles directed by lithographically

confined DNA origami Nature Nano­technology, 5, 121–126. 7. Pastre, D., Hamon, L., Landousy, F., Sorel, I., David, M. O., Zozime, A., Le Cam, E., Pietrement, O. (2006) Anionic polyelectrolyte adsorption on mica mediated by multivalent cations: A solution to DNA imaging by atomic force microscopy under high ionic strengths Langmuir 22, 6651–6660. 8. Kershner, R. J., Bozano, L. D., Micheel, C. M., Hung, A. M., Fornof, A. R., Cha, J. N., Rettner, C. T., Bersani, M., Frommer, J., Rothemund, P. W. K., Wallraff, G. M. (2009) Placement and orientation of individual DNA shapes on lithographically patterned surfaces Nature Nanotechnology 4, 557–561. 9. Zanchet, D., Micheel, C. M., Parak, W. J., Gerion, D., Alivisatos, A. P. (2001)

Templated Assembly of DNA Origami Gold Nanoparticle Arrays Electrophoretic isolation of discrete Au nanocrystal/DNA conjugates Nano Letters 1, 32–35. 10. Demers, L. M., Mirkin, C. A., Mucic, R. C., Reynolds, R. A., Letsinger, R. L., Elghanian, R., Viswanadham, G. (2000) A fluorescencebased method for determining the surface coverage and hybridization efficiency of

197

thiol-capped oligonucleotides bound to gold thin films and nanoparticles Analytical Chemistry 72, 5535–5541. 11. Hansma, H. G., Bezanilla, M., Zenhausern, F., Adrian, M., Sinsheimer, R. L. (1993) Atomic Force Microscopy of DNA in Aqueous-Solutions Nucleic Acids Research 21, 505–512.

Chapter 14 DNA-Modified Single Crystal and Nanoporous Silicon Andrew Houlton, Bernard A. Connolly, Andrew R. Pike, and Benjamin R. Horrocks Abstract The functionalization of silicon as elemental crystalline wafer, nanoporous layers, or nanocrystalline particles with DNA oligonucleotides using automated solid phase synthesis is described. The procedures provide semiconductor surfaces covalently modified with oligomers suitable for capturing complementary oligonucleotide strands. Key words: DNA, Silicon, Oligonucleotide, Nanoporous, Monolayer

1. Introduction The functionalization of surfaces with DNA is an increasingly useful strategy for sensors and diagnostic devices, materials design and the bottom-up assembly of complex hierarchical systems (1, 2). Surface types to which DNA has been attached include insulators, commonly glass or the native oxide on silicon (3, 4), metals, parti­ cularly gold (5), and molecular-based materials such as synthetic polymers (6). Rather less well developed is the preparation of DNA-semiconductor surfaces though some binary inorganic Q-dots and silicon, in various forms, have been reported with oligonucleotide attachment (7–15). A key step in the preparation of DNA-modified silicon has been the development of the functional group chemistry of the hydrogen-terminated layer formed at silicon upon fluoride ­etching (16). Actually, in all cases to date irrespective of the method of attachment, either electrostatic (10) post-synthetic conjugation (11, 12) or, as here, on-chip synthesis (13, 14), a Si–C bonded

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_14, © Springer Science+Business Media, LLC 2011

199

200

Houlton et al.

Fig.  1. The steps involved in the on-chip synthesis of oligonucleotides at silicon surfaces. (1) Alkylation of hydrogen-­terminated silicon with 4,4¢-dimethoxytrityl-1-undecenol. (2) Solid-phase oligonucleotide synthesis resulting in 3¢-tethered strands. (3) Hybridization with complementary strand (reproduced from ref. 7 with permission from ACS).

monolayer, formed through reaction with the hydrogen-terminated surface, is used as a molecular interconnect (see Fig. 1). We use thermal hydrosilation to form the organic monolayer (13) for subsequent oligonucleotide synthesis as the alkene derivatives are widely available, the reaction is tolerant of functional groups and a metal-containing catalyst is not needed. The method is a straightforward route to forming well-ordered alkyl-based monolayers and the robust Si–C bond anchoring the monolayer is well suited for subsequent processing involved during oligonucleotide synthesis. Furthermore, patterning of the monolayers is also possible by selective etching, e.g., using photolithographic techniques, to mask areas of the surface. The method described here uses a simple protecting group approach to produce monolayers with a terminal –OH (Scheme); hydrogen-terminated ­silicon surfaces of oriented <111> single crystal or porous silicon are alkylated with 4,4¢-dimethoxytrityl-protected w-undecenol by refluxing in toluene solution. Automated solid-phase oligonucleotide synthesis using phosphoramidite reagents is then possible since the alkylated silicon surfaces presents, after de-blocking with acid, the necessary primary alcohol group for chain extension.

2. Materials 2.1. Preparation of w-dimethoxytri­ tylundecenol

1. 4,4¢-Dimethoxytrityl chloride (Sigma–Aldrich, Lancaster or Fluka) used as received but stored in a desiccator once opened to keep free from trace moisture. 2. w-Undecenol (Sigma–Aldrich, Lancaster or Fluka) used as received but stored in a desiccator once opened to keep free from trace moisture.

DNA-Modified Single Crystal and Nanoporous Silicon

201

3. Pyridine, distilled over calcium hydride under a nitrogen atmosphere directly before use. Alternatively, anhydrous ­pyridine in a Sure/Seal™ bottle (Sigma–Aldrich) can be used, and the pyridine is removed by syringe under a N2 purge. 4. Chloroform. 5. Magnesium sulfate. 6. Nanopure water, freshly collected deionized water from a Barnstead Nanopure Diamond™ system (18 MW cm). 2.2. Preparation of HydrogenTerminated Single Crystal Silicon Si–H

1. <111> silicon chips: <111> oriented silicon wafers, phosphorusdoped, n-type, 1–20  W cm resistivity (Compart Technology, Peterborough, UK) cut in 1  cm2 squares with the aid of a diamond pencil. 2. Organic cleaning solvents: trichloroethylene (TCE), 2-­propanol, and acetone of semiconductor grade or similar quality to ensure regular surfaces clear of contaminants for imaging by scanning probe microscopy (SPM). 3. Nanopure water, freshly collected deionized water from a Barnstead Nanopure Diamond™ system (18 MW cm). 4. Piranha solution, made in a sample vial by adding conc. H2SO4 to 30% H2O2 (1:4 v/v conc. H2SO4 and 30% H2O2). The solution initially becomes hot and is kept at a gentle boil by placing the vial directly onto a hot plate ca. 100°C. Warning: piranha solution reacts violently with organic ­materials and should be handled with extreme caution. 5. Aqueous NH4F: semiconductor grade 40% w/v aqueous NH4F is used in freshly degassed (1 h of bubbling N2 through the solution in a sealed PTFE cell) aliquots for each etching reaction (see Note 1).

2.3. Preparation of HydrogenTerminated Porous Silicon: Porous-Si–H

1. <100> silicon chips (approx. 1  cm2) cut with a diamond ­pencil from a silicon wafer (boron-doped, p-type, <100> oriented, 10 ± 5  W cm resistivity, Compart Technology, Peterborough, UK). 2. A potentiostat such as a Sycopel AEW2 (Boldon, Tyne & Wear, UK) with the supplied software. 3. Piranha solution (vide supra). 4. Nanopure water, freshly collected deionized water from a Barnstead Nanopure Diamond™ system (18 MW cm). 5. HF solution: 48% w/w aqueous HF (see Note 1). 6. Ethanol (semiconductor grade).

2.4. Alkylation of HydrogenTerminated Silicon

1. Toluene, dried over sodium wire and distilled under dry N2 prior to use. 2. Dichloromethane (DCM, semiconductor grade).

202

Houlton et al.

3. TCA solution: 3% Trichloroacetic acid (TCA) in DCM, ­prepared by dissolving 15 g of TCA in 500 mL of DCM and stored in a darkened glass bottle. Alternatively, the Expedite de-block mix purchased from Glen Research can be used (it can be collected directly from the DNA synthesizer using the prime reagent function and diverting the tubing into a small vial). 2.5. Oligonucleotide Synthesis on Alkylated Silicon Chips

1. Phosphoramidites: ULTRAMILD™ phosphoramidites and other synthesis reagents for automated DNA synthesis (Glen Research), with the exclusion of those listed below. 2. Expedite Cap A solution (PE Biosystems). 3. Phenoxyacetic anhydride THF solution (preferred to the ­acetic anhydride equivalent) used in conjunction with the ULTRAMILD™ bases. All reagents must be stored in a ­desiccator at 4°C until required. 4. Anhydrous methylamine (CH3NH2) supplied in a pressurized canister and stored at room temperature.

2.6. Hybridization

1. HEPES-based buffer: (60  mL nanopure water, 25  mmol HEPES (596 mg), 200 mmol NaCl (1.17 g), 1 mmol EDTA (29 mg), adjusted to pH 7.4 by NaOH) is prepared and is stored in 25 mL aliquots in a freezer at −4°C. 2. Complementary sequences, prepared on standard controlled pore glass (CPG) columns and purified by HPLC before hybridization to the surface bound strands. Alternatively, these can be purchased as purified and lyophilized oligonucleotides (Eurofins Genetic Services Ltd.). 3. Complementary sequences solution: 5–6  mL of complementary sequence (30 mM 12-mer or 24-mer) in 200–400 mL of HEPES-based buffer.

3. Methods 3.1. Preparation of w-Dimethoxytri­ tylundecenol

1. Hydroxyl protection of w-dimethoxytritylundecenol (dimethoxytrityl is abbreviated as DMT) involved stirring w-­undecenol (1.0 g, 5.9 mmol) with dimethoxytrityl chloride (2.2 g, 6.5 mmol) in anhydrous pyridine (10 mL) for 16 h. 2. The pyridine was removed in vacuo to leave an orange/brown gum (see Note 2). 3. The gum was redissolved in chloroform (20 mL) and washed (×3) with distilled water. 4. The organic layer was separated and dried over MgSO4.

DNA-Modified Single Crystal and Nanoporous Silicon

203

5. After filtration and removal of the chloroform in vacuo, the crude orange oil was purified by column chromatography on silica (petroleum ether:acetone, 15:2) to give a clear oil (1.23 g, 40%). 1H-NMR (CDCl3) d 1.42 (m, 14 H, (CH2)7) d 2.05 (m, 2 H, CH2–CH2–O) d 3.02 (t, 2 H, CH2–CH2–O) d 3.80 (s, 6H, OCH3) d 5.00 (m, 2 H, CH2=CH) d 5.82 (m, 1 H, CH2=CHCH2) d 6.82 (m, 4 H, aromatic) d 7.30 (m, 9 H, aromatic). MS: m/z (%): 472 (100), [M+]. 3.2. Preparation of HydrogenTerminated Single Crystal: Si<111>−H

1. <111> silicon chips were degreased with a series of washing steps: (1) boiling in TCE for 20 min, (2) soaking in 2-propanol, and (3) acetone for 5 min each and finally rinsing with and then storing in nanopure water. 2. An oxide layer was formed by immersing the chips in freshly prepared piranha solution for 1 h at ca. 100°C. 3. The oxide was removed and a hydrogen-terminated surface was formed by etching in aqueous NH4F for 3 min with the chip held in a vertical orientation. 4. Wafers were rinsed for 20  s in nanopure water and quickly blown dry under a stream of N2 to avoid surface oxidation.

3.3. Preparation of HydrogenTerminated Porous Silicon: Porous-Si–H

1. Porous silicon was formed by galvanostatic anodization of <100> silicon chips. Chips were degreased in acetone and then oxidized in freshly prepared piranha solution for at least 10 min at room temperature. 2. The oxide was removed and a hydride layer was formed by a subsequent immersion in aqueous HF (for 10 min). 3. Chips were rinsed for 20 s in nanopure water and blown dry with N2. 4. Ohmic contact to the back of the chip was made by scratching the surface and then coating with In/Ga eutectic. 5. The chip was placed in a PTFE cell with an electrolyte ­consisting of a 1:1 v/v solution of aqueous HF:ethanol and a current density of 12.7 mA/cm2 was applied until the charge passed reached 5.0 C/cm2 (ca. 6.5 min). 6. The chips were washed in nanopure water to remove ethanol and then immersed for a few seconds in aqueous HF to remove any residual oxide. 7. The chip was rinsed again with nanopure water and dried in a stream of dry N2. 8. Residual water in the pores was removed by heating to ca. 100°C for at least 3 h on a vacuum line. 9. Transmission FTIR spectra was collected. It indicated that samples so prepared were hydrogen-terminated with no oxide present and no hydrocarbon contamination and could be stored under nitrogen in a sealed Schlenk flask or under

204

Houlton et al.

v­ acuum for a few days. Samples stored for long times under vacuum occasionally showed traces of adventitious hydrocarbons (these can be removed easily by washing with hexane and/or DCM). Samples for preparation of DNA-modified surfaces were used immediately after preparation. 3.4. Alkylation of HydrogenTerminated Silicon: Preparation of Undecyl-ODMT Monolayers on Si<111>−H or Porous-Si–H

1. The reaction was performed using a double manifold ­vacuum/ N2 Schlenk line and flasks and under an atmosphere of dry N2 gas (see Note 3). A hydrogen-terminated silicon chip, as either oriented single crystal Si–H or porous Si–H ­prepared as described above was placed in a Schlenk flask purged with dry N2. Approx. 15 mL of a 0.02-M solution of 4,4¢-­dimeth oxytrityl-1-undecenol (prepared as in Subheading 3.1 above) in dry toluene was added to the flask and set to reflux overnight. 2. The surface was washed with excess toluene, followed by DCM and then dried under N2. The chip was stored under N2 prior to use. 3. The presence of the monolayer can be confirmed on porous silicon samples by normal transmission FTIR (aromatic C–H 3,074, 3,040, 3,000 cm−1, alkyl C–H 2,927, 2,854 cm−1, CC 1,641, 1,608 cm−1, sc(CH2) 1,465 cm−1). 4. Washing with TCA solution quantitatively removed all the bands associated with the DMT group and produced a broad band centered ca. 3,300 cm−1 due to –OH. 5. Modification of <111> silicon chips followed the same protocol (it is generally not possible to obtain FTIR spectrum in the normal transmission alignment for reasons of instrumental sensitivity). 6. After alkylation, the <111> silicon chips surface was washed with excess toluene and dried under N2 before storing under N2 prior to use.

3.5. Oligonucleotide Synthesis on Alkylated Silicon Chips

1. The alkylated silicon surfaces were loaded onto an Applied Biosystem Expedite DNA synthesizer (PE Applied Biosystems, Warrington, Cheshire, UK) in a column ­assembly modified to accept ca. 1 cm2 Si chips. This consisted of a cylindrical PTFE cell approximately 3.5  cm diameter and 2.3 cm in length. Holes (ca. 3 mm diameter bore) drilled in each face provided an inlet and outlet for the solution and were each terminated with a coarse glass frit (0.7 cm diameter). The cell was machined in two parts which screwed together to hold the silicon chip against a Viton O-ring seal with the face of the chip sandwiched between the porous frits so that the solution flowed over the

DNA-Modified Single Crystal and Nanoporous Silicon

205

surface of the wafer in a wall-tube configu­ration. The sequences (12-mer: 5-ATC-GTC-AGT-CCA-3; 24-mer 5-AGC-GGATAA-CAA-TTT-CAC-ACA-GGA-3) were synthesized using Ultramild base-phosphoramidites and the standard protocols for 0.2 mmol synthesis (see Note 4). 2. Deprotection of the oligonucleotides involved treatment with anhydrous, gaseous methylamine at room temperature for 20 min in a sealed Schlenk flask. 3. The samples were then washed with dry ethyl acetate followed by ether to remove traces of methylamine and the cleaved protecting groups. The samples were stored under an inert atmosphere of dry N2 (see Note 5). 3.6. Hybridization

1. The surface bound oligonucleotides were hybridized using 200–400  mL of complementary sequence solution HEPESbased buffer (see Note 6). 2. Samples were left immersed in solution for 1  h at room ­temperature to allow hybridization. After this, the wafers were exhaustively washed with fresh HEPES-based buffer, nanopure water, and then dried under a stream of dry N2. 3. Hybridization of the complement sequence to the surface bound oligonucleotides on <111> Si shows a degree of ordering in the STM images. Domains containing many duplex are evident and these domains are aligned with respect to one another into worm-like features (see Fig. 2).

Fig. 2. STM images of DNA-modified Si<111> surface with a 24-mer sequence (left ) and a 12-mer sequence (right ) after hybridization with complementary strands showing ordered worm-like features. Scales are 500 × 500  nm and 150 × 150 nm for left and right images, respectively (reproduced from ref. 7 with permission from ACS).

206

Houlton et al.

4. Notes 1. All work with NH4F and HF needs to be carried out in a ­designated ventilated area with suitable warning signs and clearly marked access to HF antidote gels. Particular care must be taken, gloves and protective clothing should be worn at all times and any trace contamination on hands should be washed immediately with plenty of water. A waste disposal beaker of aqueous CaF should be at hand to safely collect any nanopure water washings or spills. 2. To ensure complete removal of the pyridine repeated ­co-evaporation with toluene is required. More than trace amounts of pyridine in the product mixture can hinder the purification of 1,w-dimethoxytritylundecenol by column chromatography. 3. A standard double manifold (Schlenk) line was used providing vacuum/dry N2 feeds. Flasks were equipped with Young™ joints and taps so as to avoid the need for greasing standard glass equivalents. 4. The use of UltraMILD™ base phosphoramidites (Glen Research, VA) was preferred over normal reagents due to the modified deprotection step. This step for the former reagents employs brief treatment with anhydrous methylamine ­followed by rinsing with ethyl acetate and water. This compares with the standard treatment of 12 h in aqueous NH3 at 55°C. The former method yields much higher quality surfaces which are devoid of pitting as can be the case when using aqueous NH3. 5. If the quality of oligonucleotide synthesized at the silicon surface is to be confirmed the cleavable 3¢-phosphorylation reagent (Chemical phosphorylation agent, Glen Research, VA) should be inserted at the first position in the sequence. This is cleaved during deprotection and allows the quality of the oligonucleotides to be analyzed by standard gel electrophoresis methods. The cleaved oligonucleotides were removed from the silicon chip by deprotection as above followed by rinsing with water instead of dry organic solvents. The aqueous washings were pooled, washed with ethyl acetate, and then concentrated down. 6. In some cases ethanol was added to the hybridization buffer to improve the wetting of the hydrophobic porous silicon or the sample was heated to 60°C followed by cooling to room temperature to attempt to improve the yield.

DNA-Modified Single Crystal and Nanoporous Silicon

207

References 1. Mirkin, C. A. (2000) Programmed assembly of 2-, 3-D architectures with DNA and Inorganic building blocks. Inorg Chem. 39, 2258–72. 2. Nanobiotechnology Concepts, Applications and Perspectives. Weinheim: Wiley-VCH 2004. 3. Pirrung, M. C. (2002) How to make a DNA Chip. Angew. Chem., Int. Ed. 41, 1277–1287. 4. Lin, V. S.-Y., Motesharei, K., Dancil, K.-P. S., Sailor, M. J., and Ghadiri, M. R. (1997) A Porous Silicon-Based Optical Interferometric Biosensor Science 278, 840–843. 5. Kelley, S. O., Jackson, N. M., Hill, M. G., and Barton, J. K. (1999) Long-Range Electron Transfer Through DNA Monolayers. Angew. Chem., Int. Ed. 38, 941–945. 6. Watson, K. J., Park, S.-J., Im, J.-H., Nguyen, S. T., and Mirkin, C. A. (2001) DNA−Block Copolymer Conjugates. J. Am. Chem. Soc. 123, 5592–5593. 7. Patole, S. N., Pike, A. R., Connolly, B. A., Lie, L. H., Horrocks, B. R., and Houlton, A. (2003) STM study of DNA films synthesised on Si(111) surfaces. Langmuir 19, 5457–5463. 8. Lie, L. H., Patole, S. N., Pike, A. R., Ryder, L. C., Connolly, B. A., Ward, A. D., Tuite, E. M., Houlton, A., and Horrocks, B. R.. (2004) Immobilisation and synthesis of DNA on Si(111), nanocrystalline porous silicon and silicon nanoparticles. Faraday Disc. 125, 235–249. 9. Mitchell, G. P., Mirkin, C. A., and Letsinger, R. L. (1999) Programmed assembly of DNA

10.

11.

12.

13.

14.

15. 16.

functionalized quantum dots. J. Am. Chem. Soc. 121, 8122. Strother, T., Cai, W., Zhao, X., Hamers, R. J., and Smith, L. M. (2000) Synthesis and Characterization of DNA-modified Si(111) Surfaces J. Am. Chem. Soc. 122, 1205–1209. Strother, T., Hamers, R. J., and Smith, L. M. (2000) Covalent attachment of oligodeoxyribonucleotides to amine-modified Si (001) surfaces Nucl. Acid. Res. 28, 3535–3541. Lin, Z., Strother, T., Cai, W., Cao, X., Smith, L. M., and Hamers, R. J. (2002) DNA attachment and hybridisation at the silicon(100) surface. Langmuir 18, 788–796. Pike, A. R., Lie, L. H., Eagling, R. D., Ryder, L. C., Patole, S. N., Connolly, B. A., Horrocks, B. R., and Houlton, A. (2001) DNA on Silicon devices. On-chip synthesis, hybridization, and charge transfer. Angew Chem Int Ed. 41, 615–617. Pike, A. R., Ryder, L. C., Horrocks, B. R., Clegg, W., Connolly, B. A., and Houlton, A. (2005) Ferrocenyl-modified DNA: Synthesis, characterization and integration with semiconductor electrodes. Chemistry, Eur J. 11, 14–24. Buriak, J.M. Organometallic Chemistry on Silicon and Germanium surfaces. (2002) Chem. Rev. 102, 1271–1308. Bateman, J. E., Eagling, R. D., Worrall, D. R., Horrocks, B.R., and Houlton, A. (1998) Alkylation of porous silicon by direct reaction with alkenes and alkynes. Angew. Chem. Int. Ed. 37, 2683–2685.

Chapter 15 The Atomic Force Microscopy as a Lithographic Tool: Nanografting of DNA Nanostructures for Biosensing Applications Matteo Castronovo and Denis Scaini Abstract Current in  vitro techniques cannot accurately identify small differences in concentration in samples containing few molecules in single or few cells. Nanotechnology overcomes these limitations with the possibility of measuring protein amounts down to a hundred molecules and subnanomolar concentrations and in nanoliter to picoliter volumes. The nanoscale approach, therefore, permits measurements in samples consisting of single or few cells. Atomic force microscopy (AFM) nanografting can be utilized to prepare DNA nanopatches of different sizes (from few hundreds of nanometers to few microns in size) onto which DNA–antibody conjugates can be anchored through DNA-directed immobilization. AFM height measurements are used to assess the binding of the proteins as well as their subsequent interaction with other components, such as specific proteins from the serum. Recent results have contributed to demonstrate that nanografted patch arrays are well suited for application in biosensing and could enable the fabrication of multifeature protein nanoarrays. Key words: DNA, Building Blocks, Immobilization, Binding, Nanografting, Nanotechnology, AFM, Self-assembly, Monolayer, Supramolecular, Protein, Detection, Single cell, Analysis, Sensitive

1. Introduction The development of protein arrays with feature sizes at the micrometer length scale is currently of great interest for biomedical diagnostics and life sciences because these devices promise to evolve into a powerful technological platform for high-throughput analysis of biomolecular interactions with low requirements on the sample amount and hands-on processing time (1–4). An important area in the developments of protein and small molecule arrays

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_15, © Springer Science+Business Media, LLC 2011

209

210

Castronovo and Scaini

concerns their further miniaturization to the nanometer regime. Although various methods have been developed based on, e.g., dip-pen nanolithography (5–7), SNOM lithography (8), or microcontact printing (9), it is currently still very difficult to fabricate nanoarrays containing multiple features, such as a range of different proteins or small-molecule ligands. We here report a novel approach for the fabrication of protein nanoarrays, which takes advantage of atomic force microscopy (AFM)-based nanografting (see Fig. 1), previously developed and applied for hybridization studies of DNA

Fig. 1. (a) Illustration of nanografting technique. An AFM tip is exerted on an ethylene-glycol terminated self-assembled monolayer on a gold film in the presence of another thiol molecule. The height force applied by the tip allows an immediate exchange between thiolated molecules from the surface and thiolated molecules in solution. (b) An AFM image has shown that nanografting can be applied to the immobilization of DNA molecules. The process allows controlling the shape of nanografted patches with a resolution in order of the AFM tip size. (c) The AFM allows measuring the topographic height of DNA nanografted structure. A graph shows the side-by-side height profile of the DNA nanostructure in (b) along the black line.

The Atomic Force Microscopy as a Lithographic Tool

211

Fig. 2. Atomic force microscopy nanografting was utilized to prepare DNA nanopatches of different sizes (200 × 200 to 1,000 × 1,000 nm2) onto which DNA–protein conjugates can be anchored through DNA-directed immobilization. Height measurements were used to assess the binding of the proteins as well as their subsequent interaction with other components, such as antibodies. The results indicate that nanografted patch arrays are well suited for application in biosensing and could enable the fabrication of multifeature protein nanoarrays. Reproduced from ref. 24.

and other nucleic acids (10–14), and DNA-directed immobilization (DDI) of semisynthetic protein–DNA conjugates. The latter method, which takes advantage of the specific Watson–Crick hybridization of oligonucleotide-modified proteins to surfacebound complementary oligomers, has previously been developed (15) and largely applied for the generation of self-assembled protein arrays at the micrometer length scale (16–21). Although both AFM-based nanografting (22) and DDI (23) are by now considered established methods, it is only recently Scoles, Niemeyer and collaborators have shown that these two methods can indeed be combined and used to synergistically enable fabrication of protein nanoarrays with high control over lateral dimensions as well as options for reliable read-out, based on topographic AFM measurements (24). DDI is used to decorate nanometer-sized patches of single-stranded DNA (ssDNA) monolayers, previously prepared by AFM-based nanografting within alkylthiol self-assembled monolayers (SAMs) on ultraflat gold surfaces (10) with cDNA– protein conjugates (see Fig. 2).

2. Materials 2.1. Ultraflat Gold Substrate

1. Muscovite mica sheets (Goodfellow Cambridge Limited, Huntingdon, UK) freshly cleaved immediately before use. 2. High purity gold wire (99.999%, Goodfellow Cambridge Ltd., Huntingdon, UK).

212

Castronovo and Scaini

3. SU8-100 photoresist (MicroChem Co., Newton, MA). 4. (100) Silicon wafer, 500 (University Wafer, South Boston, MA) is cut in 5 × 5 mm2 slides using a diamond pen and successively washed in acetone (for HPLC, purity ³ 99.9%, Sigma– Aldrich) and ethanol (for HPLC, gradient grade, purity ³ 99.8%, Fluka), respectively and dried in a nitrogen flow. 5. Ethylene-glycol (EG) terminated alkanethiol (SH–(CH2)11– (O–CH2–CH2)3–OH, EG3 Thiol, ProChimia Surfaces Sp., Poland) is dissolved in 100  mM of absolute ethanol (for HPLC, gradient grade, purity  ³99.8%, Fluka). 6. Ethylene-glycol (EG) terminated alkanethiol (SH–(CH2)11– (O–CH2–CH2)6–OH, EG6 Thiol, ProChimia Surfaces Sp., Poland) is dissolved in 100  mM of absolute ethanol (for HPLC, gradient grade, purity ³ ³99.8%, Fluka). 2.2. Fabrication of DNA Nanostructures

1. Silicon-made AFM tips for the nanofabrication process (NSC19/noAl tip, nominal spring constants 0.63 Nm−1, tip radius < 10 nm, MikroMasch, Tallinn, Estonia) and subsequent DNA nanostructures characterization (CSC38/noAl tip, nominal spring constants 0.03  Nm−1, tip radius < 10  nm, MikroMasch, Tallinn, Estonia). 2. High purity and solvent resistant plastic polymer (TOPAS 6015, Topas Advanced Polymers GmbH, Frankfurt-Höchst, Germany) dissolved in toluene (CHROMASOLV Plus, for HPLC, ³99.9%, Sigma–Aldrich) in a 2:1 (v/v) ratio. 3. Nanofabrication buffer: 1  M sodium chloride, 10  mM tris(hydroxymethyl)aminomethane, 1  mM ethylenediaminetetraacetic acid (EDTA), pH 7 in ultrapure water (>18 MW cm). The buffer is filtered through a 0.22 mm pore size filter (GP Express PLUS Membrane, Millipore Co., Billerica, MA) just before use and stored at +4°C. 4. Characterization buffer: 1 M phosphate-buffered saline (PBS) (from tablets), 1  mM EDTA pH 7.2, made with ultrapure water (>18 MW cm). The buffer is filtered through a 0.22 mm pore size filter (GP Express PLUS Membrane, Millipore Co., Billerica, MA) just before use and stored at +4°C. 5. Thiolated 22 bases ssDNA D0 (5¢-TACAGTCAGAGTATGA GCCGAA-3¢), cD0 (SH-(CH2)6-5¢-TTCGGCTCATACT CTGACTGTA-3¢), cD1 (SH-(CH2)6-5¢CTTCACGATT GCCACTTTCCAC-3¢), cD2 (SH-(CH2)6-5¢-CTTATCGC TTTATGACCGGACC-3¢) (HPLC purified grade, Biomers.net GmbH, Ulm, Germany). Oligos are aliquoted in 15 mL batches at 100  mM concentration in TE buffer (10  mM Tris–HCl, 1 mM EDTA, pH 8.0) and stored at −20°C.

The Atomic Force Microscopy as a Lithographic Tool

213

6. Nanografting solution for EG3 SAM on gold: immediately before use thiolated ssDNA is diluted at a final concentration of 5 mM in an ethanol:nanofabrication buffer 1:1 (v/v). 7. Nanografting solution for EG6 SAM on gold: immediately before use thiolated ssDNA is diluted at a final concentration of 5 mM in nanofabrication buffer. 2.3. DNA–Protein Conjugation

1. Proteins: horseradish peroxidase (HRP) and glucose oxidase (GOx) (Sigma). Streptavidin (STV) was expressed as recombinant core streptavidin in E. coli using a published protocol and purification procedures; all the antibodies were obtained from Abcam (Cambridge, UK). 2. Thiolated 22 bases ssDNA D1 (SH-(CH2)6-5¢-GTGGAAA GTGGCAATCGTGAAG-3¢),D2(SH-(CH2)6-5¢-GGTCCGG TCATAAAGCGATAAG-3¢) (HPLC purified grade, Biomers. net GmbH, Ulm, Germany). 3. NAP5 and NAP10 size exclusion columns (Pharmacia). 4. Centricon 30 microcentrifuge ultrafiltration units (Millipore). 5. Centricon exchange buffer: 20 mM, pH 8.3. 6. MonoQ HR 5/5 column (Pharmacia). 7. Column buffer A: 20 mM Tris A pH 8.3. 8. Colum buffer B: 20 mM Tris, 1.5 M NaCl, pH 8.3.

3. Methods 3.1. Ultraflat Gold Substrate Fabrication

Ultraflat gold film substrates are prepared using a modified Ulman procedure. This method is a sort of mechanical template stripping of a gold film from its substrate (usually mica or silicon). This method offers the possibility to prepare a fresh and clean gold substrate to be used for molecules immobilization, just when it is needed. The sample preparation is rather quick and inexpensive (25). 1. 10 × 10  cm2 freshly cleaved mica sheets are mounted in an electron beam gold evaporator (EGE 450, Rial Vacuum S.r.l., Parma, Italy) about 50 cm above the gold source and maintained at a background pressure of 5 × 10−8 Torr and subsequently backed at 400°C for 24 h. 2. 200 Å of gold are deposited at an initial rate of 0.1 Å/s with substrate heated at room temperature. Afterward the evaporation rate is increased up to 1 Å/s. The evaporation is stopped when a final thickness of about 1,000 Å is obtained. The evaporation chamber is then allowed to cool down. Finally, the sample is then removed from the evaporation chamber

214

Castronovo and Scaini

and can be stored in a nitrogen chamber at room temperature without any further precaution. 3. The obtained gold-coated mica sample is cut into small samples of about 5 × 5 mm each. Cutting a mica sample can be done in a precise way when a sharp scalpel is used. The gold layer can be placed face down on the table over a protecting optical paper sheet. It is important that the gold film is not scratched during the process. Afterward the gold–mica samples are put over a glass slide and the gold surfaces are facing upward. 3.1.1. Preparation of Ultraflat Samples with a Very Reduced Thickness

The AFM sample holder often calls for a very thin sample. This submethod produces such substrates that can be about 0.1– 0.2 mm thick. 1. One drop of about 5–10 mL of epoxy SU8-100 glue is cast on each gold sample samples (SU8-100 is an optical resist and it is recommended to deal with it in a fume hood and, possibly, in a darkroom). Due to the fact that SU8-100 is a viscous preparation that only moderately wets a gold surface, it is recommended to allow the SU8 drops to equilibrate for few minutes after their delivery in order to spread evenly onto the gold samples. 2. To help the evaporation of the SU8 solvent from the drops a prebacking at 50°C for 6  h is recommended. The SU8 is cured by thermal annealing at 95°C for at least 3 h to complete the epoxy reticulation. Ultimately, the samples are slowly cooled down to room temperature.

3.1.2. Preparation of Ultraflat Samples with a Standard Thickness

1. One drop of about 5 mL of epoxy SU8-100 glue is cast on each gold sample. A 5 × 5 mm2 (or less) silicon slides are put in contact with the drops (make sure the silicon 100 surface contacts the SU8) and are thus gently pressed enough to flatten the SU8 glue (see Note 1). SU8 curing is the same one as described in step 2, paragraph 3.1.1 (10). 2. Using a sharp scalpel, either SU8 drops or silicon flakes are mechanically detached in air from the mica substrate. In turn, the gold film remains attached to the SU8 glue. The gold surface that was originally buried at the gold–mica interface is now exposed to the air for the first time. Such a gold surface has the advantage of reproducing the flatness of mica. In turn, they are ultraflat and have a roughness of about 3 Å over areas as large as hundreds of microns (10).

3.2. EG3/EG6 Passivation of the Gold Substrate

1. A 1.5 mL microcentrifuge tube is filled with a 100 mM ethanol solution of either EG3 or EG6 thiol. It is recommended always to use fresh EG3 or EG6 solutions. Storage is recommended for no longer than 1 month in clean glass bottles at −20°C. 2. Immediately after stripping, ultraflat gold sample are soaked into EG3 or EG6 solutions in a microcentrifuge tube.

The Atomic Force Microscopy as a Lithographic Tool

215

The substrate is kept in the thiol solution for 24 h at room temperature protecting it from the light. In this way, an ultraflat surface covered with an either EG3 or EG6 thiol SAM is obtained (10). Generally, EG terminated SAMs are known to resist the unspecific adsorption of biomolecules (26). 3. Sample must be extensively rinsed in ethanol before being used. 3.3. Nanografting of DNA Nanopatches

Nanografting is an emerging AFM-assisted nanolithograhy process (13). DNA nanografting for biosensing application will be described in the following. During nanografting AFM is normally carried out by a contact mode AFM (C-AFM) (see Note 2). 1. A freshly prepared EG3 or EG6 SAM-coated ultraflat gold film is setup on the stage of the AFM liquid cell. The sample is attached over metal or glass surfaces by using a toluenesoftened Topas 6015 polymer. The Topas glue is cured by keeping the sample for 30 min inside a nitrogen box at room temperature. 2. The liquid cell and/or sample are mounted into the AFM. An AFM cantilever is used to make markers on the gold surface by scratching it. To make these markers, it is necessary to apply a force in order of 300–500  nN. Once the tip is approached, markers are obtained by moving the sample by acting on the XY positioners. Markers are to be seen with a standard optical microscope as the one that is usually incorporated in the AFM system. Markers are essential to track the nanografting area on the surface when the AFM tip is changed. 3. A new, medium stiffness, AFM cantilever (NSC19) is mounted on the AFM. 4. The liquid cell is filled with the ssDNA solution. Use either nanografting solution for EG3 SAM on gold or nanografting solution for EG6 SAM on gold, depending on the SAM of choice (see Note 3) (10). 5. The AFM tip is approached on the surface far away from the markers. Nanografting is achieved by operating the AFM tip on the surface with forces ranging between 80 and 120 nN as it is shown in Fig. 1a. The scan rate during nanografting can be varied between 5,000 and 500 nm/s. During nanografting the EG3 (or EG6) SAM phase is locally disrupted and immediately exchanged with ssDNA molecules in solution. The DNA density inside the nanopatches critically depends on nanografting parameters, namely the DNA concentration, the nanografting solution, the density of scanning lines, and the loading force during nanografting (10) (see Note 4). 6. After nanografting, DNA patches can be imaged with the same AFM tip if the force is lowered at about 1 nN as it is shown in

216

Castronovo and Scaini

Fig. 1b. The relative height of each DNA patches with respect to the surrounding EG3 SAM is measured by scanning sideby-side the patches as it is depicted in Fig. 1c. 7. Before proceeding, it is useful to record an optical microscope image of the sample to remember the relative position between the cantilever (which stands right above the grafting area) and the markers. 8. After the nanofabrication process, the AFM tip is retracted from the surface. Liquid cell is drained and sample copiously washed respectively with ethanol and nanofabrication buffer (see Note 5). 3.4. DNA–Protein Conjugates Preparation

The D1-STV, D2-STV, D2-HRP, and D1-GOx conjugates were prepared using the bifunctional crosslinker sSMCC (sulfosuccinimidyl-4-(N-maleimido-methyl)-cyclohexan-1-carboxylte), following a previously described method (15, 17): 1. 100  mL of a 100  mM solution of the corresponding 5¢-thiol-modified oligonucleotide in TEN buffer was mixed with 60  mL DTT (1  M) and incubated overnight at 37°C to reduce any disulfide bonds formed upon storage of the oligonucleotide. 2. 200 mL of 200 mM protein solution (in phosphate buffer, pH 7.2) were incubated for 1 h at 37°C with sSMCC (2 mg in 60 mL DMF). Both DNA and protein solution were purified by two consecutive gel-filtration chromatography steps using NAP5 and NAP10 size exclusion columns. 3. The purified DNA and protein solutions, each of which had a volume of 1.5 mL, were then combined and incubated in the dark at room temperature for 3 h. The reaction mixture was concentrated to approximately 300  mL by ultrafiltration (using Centricon 30) and the buffer was exchanged to the Centricon exchange buffer during this step. 4. The conjugates were purified by anion exchange chromatography on a MonoQ HR 5/5 column using linear gradient between buffer A and buffer B, over 25  min (on a AKTA purifier, Amersham Bioscience). 5. The concentration was determined spectrophotometrically (15, 17).

3.5. Imaging of DNA Nanostructures and Proteins Immobilization Experiments

1. The medium stiffness AFM cantilever used for nanografting (MikroMasch NSC19) is substituted with a much softer one (MikroMasch CSC38). This cantilever allows imaging the surface with forces down to 500 pN or less. In this way, the DNA patches topography is safely measured. Measurements can be repeated for hours while the topographic height of DNA–protein patches remains unmodified. 2. The liquid cell is drained and filled with characterization buffer.

The Atomic Force Microscopy as a Lithographic Tool

217

3. The AFM tip is approached on surface and the cantilever is manually aligned to overlap with nanografting region with the help of the previous photograph. 4. A wide range AFM imaging allows the retrieval of the DNA patches. 5. Side-by-side scanning of DNA patches allows the measurement of the relative height between DNA and surrounding SAM. Depending on the AFM feedback performances the scan rate can be varied from 0.1 up to 50 mm/s or more without causing changes in the DNA topographic height. 6. DDI: To immobilize the protein, the sample was then incubated in a solution containing the DNA–protein conjugate (e.g., D1-STV, 2 mM in PBS buffer) for 2 h at room temperature (see Fig. 2 on the left and in the center) (24). 7. The DNA–protein conjugate solution is drained from the liquid cell and the sample is copiously washed with nanofabrication and characterization buffers. 8. The last two steps are repeated for each of the other DNA– protein conjugates (i.e., D2-STV, D2-HRP, and D1-GOx) (see Fig. 2 on the right). These steps are illustrated in Fig. 3, which demonstrate the applicability of our method to fabricate multiple-feature protein nanoarrays. In Fig. 3a, a prototypical device is made of three nanopatches with three different DNA oligonucleotides (cD0, cD1, and cD2). Oligomers cD1 and cD2 are complementary to conjugate D1-STV and D2-HRP respectively, while cD0 is a control sequence. Figure  3b was taken after the ssDNA patches A, B, and C were put in contact with a solution of D1-STV conjugate (1,000 nM PBS, pH 7.3). As shown in Fig. 3b, AFM measurements revealed an increase in height (4.5 ± 0.5 nm) only at patch B, which bears the complementary capture oligonucleotide. The same sample was then immersed in a solution containing the D2-HRP conjugate (500 nM PBS, pH 7.3). Again, AFM imaging revealed an increase in height of 3.8 ± 0.5 nm (Fig. 3c) only for patch C, thereby clearly indicating specific surface immobilization of the D2-HRP conjugate. The dimension of HRP molecules is known from X-ray crystallography (4.0 × 6.7 × 11.7 nm3) and the observed increase in height of patch C is in good agreement with previous studies on HRP immobilization. 9. Antibody binding: The patch array (like the one shown in Fig. 3) is eventually probed with a solution of anti-STV IgG (1,000  nM PBS, pH 7.3). As expected, only the height of patch B, which contained D1-STV conjugate increases (Fig. 3d). In contrast, no significant changes were observed for patches A and C, and also, the background heights remained entirely unchanged in the course of this multistep experiment.

218

Castronovo and Scaini

Fig. 3. AFM topographic images (8 × 2.5 mm2) and corresponding height profiles of a three-patch device, fabricated from DNA oligonucleotides cD0 (patch A, left ), cD1 (patch B, middle), and cD2 (patch C, right ): (a) pattern after nanografting of ssDNA; (b) pattern after hybridization of D1-STV conjugate (see the exclusive height increase in the center); (c) pattern obtained after further hybridization of D2-HRP conjugate (see the exclusive height increase on the right ); (d) pattern obtained after further binding of anti-STV IgG (see the exclusive height increase in the center). Reproduced from ref. 24.

3.6. Hybridization of Control DNA Nanostructures

1. The liquid cell is filled with the hybridization solution of D0 sequence (5 mM of complementary sequence in PBS buffer). 2. Sample is left in incubation with the complementary solution for 1 h at room temperature in a dark environment. 3. The hybridization solution is drained from the liquid cell and the sample is copiously rinsed with characterization buffer.

The Atomic Force Microscopy as a Lithographic Tool

219

4. Notes 1. Glass can be used instead of silicon as a substrate. Glass slides (BK7, Gerhard Menzel, Glasbearbeitungswerk GmbH & Co.) are cut into pieces of 5 × 5 mm2 and cleaned using acetone (CHROMASOLV Plus, for HPLC, ³99.9% (Sigma– Aldrich) and ethanol (for HPLC, gradient grade, purity ³ 99.8%, Fluka) and dried in a nitrogen flow. 2. It is recommended to use AFM systems that are endowed with closed-looped piezo-scanners and a closed liquid cell. It is also possible to operate with open-loop piezoscanners, in these cases, it is recommended to run the XY scanning at least overnight before starting the grafting experiments. In this way, most XY scanner drift will be avoided. In turn, the quality of the obtained DNA nanostructures will be comparable to the one that is normally achieved with AFMs that are endowed with closed-loop piezo-scanners. 3. When EG6 instead of EG3 thiol is used to form the matrix SAM, the grafting solution does not contain ethanol. A substantial advantage of this approach is the fact that it becomes possible to use an open AFM liquid cell. Usually this kind of apparatus has a significantly reduced commercial cost with respect to a closed liquid cell. Closed cells are essential to carry out nanografting into EG3 SAM matrix to avoid ethanol evaporation from the grafting solution. On the other hand, a possible disadvantage of former approach is that the maximum density within the obtained DNA nanopatches is lower than the one obtained with the latter. Moreover, the forces needed for EG6 shaving in water solution are generally higher than those used for EG3 by operating in the 1:1 ethanol/water mixture. 4. The “right” loading force for DNA nanografting has to be determined in preliminary experiments for each cantilever to be used. This is due to the slight differences of individual tip shapes and/or sizes among the cantilevers used. 5. After nanografting the sample may be stored in different ways: (1) it can be washed with ethanol and TE buffer and subsequently dried in a gentle flow of nitrogen. Eventually, it may be stored for many days in a nitrogen box at room temperature well protected by direct light. (2) For overnight storage we recommend not to dry the sample. In can be kept in TE buffer at 4°C.

220

Castronovo and Scaini

Acknowledgments The authors are grateful to Dr. Loredana Casalis (ELETTRA Synchrotron Laboratory, Trieste, Italy), Prof. Giacinto Scoles (ICSUNIDO and SISSA, Trieste, Italy and Temple University, Philadelphia, PA), the SISSA – Scuola Internazionale Superiore di Studi Avanzati (Trieste, Italy), the Italian Institute of Technology (Trieste Unit @ SISSA, Trieste, Italy), the CBM S.c.r.l. – Cluster in Molecular Biomedicine (Trieste, Italy), Dr. Jian Liang, (Columbia University, NY), and Prof. Gang-yu Liu (UC Davis, CA). References 1. Jonkheijm, P., Weinrich, D., Schroder, H., Niemeyer, C. M., Waldmann, H. (2008) Chemical Strategies for Generating Protein Biochips Angewandte Chemie-International Edition 47, 9618–9647. 2. Sobek, J., Bartscherer, K., Jacob, A., Hoheisel, J. D., Angenendt, P. (2006) Microarray technology as a universal tool for high-throughput analysis of biological systems Combinatorial Chemistry & High Throughput Screening 9, 365–380. 3. Phizicky, E., Bastiaens, P. I. H., Zhu, H., Snyder, M., Fields, S. (2003) Protein analysis on a proteomic scale Nature 422, 208–215. 4. Tomizaki, K.-Y., Usui, K., Mihara, H. (2005) Protein-Detecting Microarrays: Current Accomplishments and Requirements ChemBioChem 6, 782–799. 5. Demers, L. M., Ginger, D. S., Park, S. J., Li, Z., Chung, S. W., Mirkin, C. A. (2002) Direct patterning of modified oligonucleotides on metals and insulators by dip-pen nanolithography Science 296, 1836–1838. 6. Lim, J.-H., Ginger, D. S., Lee, K.-B., Heo, J., Nam, J.-M., Mirkin, C. A. (2003) DirectWrite Dip-Pen Nanolithography of Proteins on Modified Silicon Oxide Surfaces13 Angewandte Chemie International Edition 42, 2309–2312. 7. Salaita, K., Wang, Y. H., Mirkin, C. A. (2007) Applications of dip-pen nanolithography Nature Nanotechnology 2, 145–155. 8. Leggett, G. J. (2005) Biological nanostructures: platforms for analytical chemistry at the sub-zeptomolar level Analyst 130, 259–264. 9. Coyer, S. R., Garcia, A. J., Delamarche, E. (2007) Facile preparation of complex protein architectures with sub-100-nm resolution on surfaces Angewandte Chemie-International Edition 46, 6837–6840.

10. Mirmomtaz, E., Castronovo, M., Grunwald, C., Bano, F., Scaini, D., Ensafi, A. A., Scoles, G., Casalis, L. (2008) Quantitative Study of the Effect of Coverage on the Hybridization Efficiency of Surface-Bound DNA Nano­ structures Nano Letters 8, 4134–4139. 11. Liu, M. Z., Liu, G. Y. (2005) Hybridization with nanostructures of single-stranded DNA Langmuir 21, 1972–1978. 12. Xu, S., Miller, S., Laibinis, P. E., Liu, G. Y. (1999) Fabrication of nanometer scale patterns within self-assembled monolayers by nanografting Langmuir 15, 7244–7251. 13. Liu, M. Z., Amro, N. A., Chow, C. S., Liu, G. Y. (2002) Production of nanostructures of DNA on surfaces Nano Letters 2, 863–867. 14. Castronovo, M., Radovic, S., Grunwald, C., Casalis, L., Morgante, M., Scoles, G. (2008) Control of Steric Hindrance on Restriction Enzyme Reactions with Surface-Bound DNA Nanostructures Nano Letters 8, 4140–4145. 15. Niemeyer, C. M., Sano, T., Smith, C. L., Cantor, C. R. (1994) Oligonucleotidedirected self-assembly of proteins: semisynthetic DNA – streptavidin hybrid molecules as connectors for the generation of macroscopic arrays and the construction of supramolecular bioconjugates Nucl Acids Res 22, 5530–5539. 16. Becker, C. F. W., Wacker, R., Bouschen, W., Seidel, R., Kolaric, B., Lang, P., Schroeder, H., Müller, O., Niemeyer, C. M., Spengler, B., Goody, R. S., Engelhard, M. (2005) Direct Readout of Protein-Protein Interactions by Mass Spectrometry from Protein-DNA Microarrays13 Angewandte Chemie International Edition 44, 7635–7639. 17. Fruk, L., Müller, J., Weber, G., Narvaez, A., Dominguez, E., Niemeyer, C. M. (2007) DNA-Directed immobilization of horseradish

The Atomic Force Microscopy as a Lithographic Tool

18.

19.

20.

21.

peroxidase-DNA conjugates on microelectrode arrays: Towards electrochemical screening of enzyme libraries Chemistry-a European Journal 13, 5223–5231. Schroeder, H., Adler, M., Gerigk, K., MüllerChorus, B., Götz, F., Niemeyer, C. M. (2009) User Configurable Microfluidic Device for Multiplexed Immunoassays Based on DNADirected Assembly Analytical Chemistry 81, 1275–1279. Schroeder, H., Ellinger, B., Becker, C. F. W., Waldmann, H., Niemeyer, C. M. (2007) Generation of live-cell microarrays by means of DNA-directed immobilization of specific cell-surface ligands Angewandte ChemieInternational Edition 46, 4180–4183. Wacker, R., Niemeyer, C. M. (2004) DDImFIA – A Readily Configurable MicroarrayFluorescence Immunoassay Based on DNA-Directed Immobilization of Proteins ChemBioChem 5, 453–459. Wacker, R., Schröder, H., Niemeyer, C. M. (2004) Performance of antibody microarrays fabricated by either DNA-directed immobilization, direct spotting, or streptavidin-biotin

22.

23. 24.

25.

26.

221

attachment: a comparative study Analytical Biochemistry 330, 281–287. Liu, M., Amro, N. A., Liu, G. Y. (2008) Nanografting for surface physical chemistry Annual Review of Physical Chemistry 59, 367–386. Niemeyer, C. M. (2007) Functional devices from DNA and proteins Nano Today 2, 42–52. Bano, F., Fruk, L., Sanavio, B., Glettenberg, M., Casalis, L., Niemeyer, C. M., Scoles, G. (2009) Toward Multiprotein Nanoarrays Using Nanografting and DNA Directed Immobilization of Proteins Nano Letters 9, 2614–2618. Gupta, P., Loos, K., Korniakov, A., Spagnoli, C., Cowman, M., Ulman, A. (2004) Facile route to ultraflat SAM-protected gold surfaces by “amphiphile splitting” Angewandte Chemie-International Edition 43, 520–523. Generally EG terminated SAMs resist the unspecific adsorption of biomolecules. (a) Kane, R. S.; Deschatelets, P.; Whitesides, G. M. Langmuir 2003, 19, 2388. (b) Harder, P.; Grunze, M.; Dahint, R.; Whitesides, G. M.; Laibinis, P. E. J. Phys. Chem. B 1998, 102, 426.

Chapter 16 Trapping and Immobilization of DNA Molecules Between Nanoelectrodes Anton Kuzyk, J. Jussi Toppari, and Päivi Törmä Abstract DNA is one of the most promising molecules for nanoscale bottom-up fabrication. For both scientific studies and fabrication of devices, it is desirable to be able to manipulate DNA molecules, or self-­assembled DNA constructions, at the single unit level. Efficient methods are needed for precisely attaching the single unit to the external measurement setup or the device structure. So far, this has often been too cumbersome to achieve, and consequently most of the scientific studies are based on a statistical analysis or measurements done for a sample containing numerous molecules in liquid or in a dry state. Here, we explain a method for trapping and attaching nanoscale double-stranded DNA (dsDNA) molecules between nanoelectrodes. The method is based on dielectrophoresis and gives a high yield of trapping only single or a few molecules, which enables, for example, transport measurements at the single ­molecule level. The method has been used to trap different dsDNA fragments, sizes varying from 27 to 8,416 bp, and also DNA origami constructions. We also explain how confocal microscopy can be used to determine and optimize the trapping parameters. Key words: Dielectrophoresis, Trapping, Immobilization

1. Introduction Due to its exceptional self-assembly properties, DNA could become a key player in bottom-up fabrication of nanoscale systems (1–3) and (functional) devices for nanoelectronics and optics (4, 5). Whether fabricating devices constructed of single molecules, or of larger self-assembled structures constituting of many different kinds of nanocomponents, the final step still includes embedding of the device into the rest of the circuit; thus, precisely controlled positioning of DNA molecules or DNA constructs is required.

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_16, © Springer Science+Business Media, LLC 2011

223

224

Kuzyk, Toppari, and Törmä

Dielectrophoresis (DEP) is a manipulation method based on the movement of a polarizable particle in a nonuniform electric field (6). In the case of positive (negative) DEP, polarized objects are moving toward the electric field maximum (minimum). Negative DEP is a consequence of the surrounding medium polarizing more than the objects themselves. In the case of an AC electric field, the dielectric force is determined by FDEP = ½α—(E 2), where α is the in-phase component of the polarizability of the object and E is the root mean square value of the electric field (assuming sinusoidal time dependence). Dielectrophoresis can be used as well with DC fields. However, utilization of AC-DEP helps to eliminate certain undesired effects, such as electrophoresis (in the case of charged particle), electrolysis, etc. For effective trapping, the dielectrophoretic force should overcome the force exerted on the particle by the random thermal motion (Brownian force), whose maximum value is given roughly by FT » kBT/V  1/3, where V is the volume of the particle and T is the temperature. In the micrometer scale, DEP has been widely used as a trapping technique for a variety of objects, such as cells, viruses, long DNA molecules, and nanotubes (7, 8). It has also been demonstrated that DEP can be used for manipulation of objects in the nanoscale, such as proteins (9, 10), short DNA molecules (10–12), and metal (13) and semiconducting (14) nanoparticles. Dielectrophoretic large-scale assembly of carbon nanotubes was also demonstrated (15). Due to its highly polarizable counter ion cloud in an aquatic buffer, DNA is well suited to DEP. DEP has proven to be an extremely powerful experimental technique for manipulating and trapping DNA molecules. We have shown that DEP can be used for trapping DNA molecules as small as 27 bp (11); moreover, it can be used for reproducible assembly of single DNA (16) or single DNA origami structures (17, 18) between nanoelectrodes with reasonably high yields.

2. Materials 2.1. Fabrication of Electrodes

For the fabrication of the metallic nanoelectrodes needed for DEP trapping, standard electron-beam lithography facilities are sufficient. Moreover, any other lithography technique that reaches the same resolution (linewidth 20–50  nm and lineheight ~20  nm) can be used. To extend the method to even smaller scales, for example, carbon nanotubes (CNT) can be utilized as an electrode (12). Below is the list of the materials used in our experiments (11, 16–18): 1. Boron-doped (100) silicon wafer, with 200-nm thermally grown SiO2 at the top. 2. Polymethylmethacrylate (A2 PMMA, Microchem).

Trapping and Immobilization of DNA Molecules Between Nanoelectrodes

225

3. Raith eLine electron beam lithography system or scanning electron microscope (SEM) LEO 1430+ equipped with Raith Elphy Plus lithography software. 4. Isopropyl alcohol (IPA). 5. (1:3 v) solution of methyl isobutyl ketone (MIBK) and IPA. 6. Oxford Plasmalab 80 Plus RIE equipped with O2. 7. UHV e-beam evaporator equipped with Ti and Au. 8. Acetone. 2.2. Fabrication of DNA Fragments

While any biological or synthetic technique is suitable for fabrication of the DNA to be trapped, we list here the materials used in our experiments to fabricate double-stranded DNA (dsDNA) fragments of varying lengths (27–8,461 bp) (11, 12, 16). 1. Primers (primers 1–5, 8, and 9; TAG Copenhagen A/S, Denmark; primers 6 and 7; Synthegen, Houston, TX) (see Table 1). 2. Taq DNA polymerase (Fermentas). 3. 1% Agarose gel and GFX™ PCR purification kit (Amersham Biosciences). 4. Transformed Escherichia coli JM109 cell line (Stratagene). 5. Plasmid purification kit (Macherey-Nagel, Düren, Germany). 6. pBVboostFG and pFastBac1 plasmids (Invitrogen). 7. ApaI, BglI, HindIII, and SpeI restriction enzymes (Promega).

Table 1 The oligonucleotides used in fabrication of dsDNA fragments Name

Base sequence

Primer 1

5¢-GGT GAA TTC GCC GGC ACC TAC ATC ACA-3¢

Primer 2

5¢-TGT GAT GTA GGT GCC GGC GAA TTC ACC-3¢

Primer 3

5¢-CCC GAT GGT CAT GTT GGC GCC CAG ATC GTT GGT-3¢

Primer 4

5¢-CTG CTA GAT CTA TGG TGC ACG CAA CCT CCC C-3¢

Primer 5

5¢-GAG TGA AGA TGA TGA TGC CGA CC-3¢

Primer 6

5¢-HS-(CH2)6-GCC AGA AAG TGC TCG CTG AC-3¢

Primer 7

5¢-HS-(CH2)6-TTC TCG ACA AGC TTT GCG GG-3¢

Primer 8

5¢-DTPA-GCC AGA AAG TGC TCG CTG ACT G-3¢

Primer 9

5¢-DTPA-CTT CTC GAC AAG CTT TGC GGG-3¢

Primers 1–5, 8, and 9 were purchased from TAG Copenhagen A/S, Denmark. Primers 6 and 7 were purchased from Synthegen, Houston, TX

226

Kuzyk, Toppari, and Törmä

2.3. Preparation of the DNA Solution

1. Buffer 1 (1×): 3 mM HEPES and 1 mM NaOH (pH ~ 6.9, conductivity s ~ 20 mS/cm) (see Note 1). 2. Buffer 2 (1×): 6.5 mM HEPES and 2 mM NaOH (pH ~ 6.9, s ~ 90 mS/cm) (see Note 1). 3. dsDNA-specific fluorescent label PicoGreen (Invitrogen). 4. Reduction agent: NaBH4 (sodium borohydride) (see Notes 2 and 3).

2.4. Dielectrophoretic Trapping

1. AC signal generator (Agilent 33120A arbitrary waveform generator). 2. Distilled water. 3. Humidity-tight container with electric feedthroughs.

2.5. Imaging

Basically any standard AFM and confocal imaging facilities are suitable. Below is the list of the equipment used in our experiments (11, 12, 16–18): 1. AFM (Veeco, Dimension 3100). 2. Probes: Veeco MPP-11100, resonance frequency ~300 kHz, spring constant ~40 N/m. 3. Probes: Nanosensors PPP-NCH, resonance frequency ~330 kHz, spring constant ~42 N/m. 4. Confocal microscope (Zeiss Axiovert LSM 510, Zeiss “Fluar” 40×/1.3 oil objective).

3. Methods Trapping of single DNA molecules is a procedure involving a large number of parameters, such as the geometry of the electrodes and the type of molecules to be trapped, which can be chosen as desired. The starting point is the fabrication of the electrodes, which can be done in several ways. The most critical parameter in the electrodes, in addition to their good conductivity, is the sharpness of the trapping structure, that is, the radius of the curvature at the end of a lithographically fabricated line. This, together with the maximum applicable voltage, determines the maximum gradient of the electric field and thus the maximum of the DEP force, FDEP µE —E. Electric field within more complex electrode geometries can be simulated, for example, by finite element method (FEM) (11, 12). The choice of the fabrication technique for the electrodes is not important if the required scale and resolution of the structures are achieved (linewidth

Trapping and Immobilization of DNA Molecules Between Nanoelectrodes

227

20–50  nm). Standard electron beam lithography method is presented here in detail as an example. The molecules to be trapped have to be obtained or fabricated as well. For this purpose, any biological or synthetic technique for DNA fabrication is suitable. Here, we briefly explain the fabrication of the dsDNA fragments of varying lengths (27–8,461 bp) (11). Three different methods were used: (1) annealing of the synthetic oligonucleotides, (2) polymerase chain reaction (PCR), and (3) restriction enzyme digestion of the plasmids multiplied in bacteria. More detailed recipes for DNA fabrication can be found in other sources (11). In addition, in order to immobilize the DNA molecules, they have to be equipped with linkers at both ends (see Note 4). This was done by using 5¢-modified oligonucleotides as primers in PCR as described below. The oligonucleotides were modified with either hexanethiol or dithiol phosphoramidite (DTPA) linkers. Table 1 lists all the primers used. Since DNA is negatively charged in a water solution, the trapping has to be done with an AC electric field to average out the electrophoretic effects. This means that the frequency has to be high enough so that the DNA molecules do not have time to follow the electrophoretic force and be attracted to the positive electrode. The trapping itself is recommended to be done first with fluorescently labeled molecules and in situ under continuous confocal fluorescence microscope imaging. This allows a coarse estimation of the correct parameters for the system under study (see Fig. 1). After the coarse tuning of the parameters, the trapping can be done with molecules without any labeling (if that is what is desired) and the result can be examined with the help of AFM imaging (see Fig. 2). By repeating this, the parameters can be fine tuned. 3.1. Fabrication of the Electrodes

1. Slightly boron-doped (100) silicon wafer, with 200-nm thermally grown SiO2 at the top as a passivation layer, was used as a substrate. 2. For a mask, polymethylmethacrylate was spin coated at 2,000– 2,500 rpm and baked for 5 min on a hot plate (160°C). 3. Patterning was done either by SEM LEO 1430+ equipped with Raith Elphy Plus lithography software, or Raith eLine ultra-high-resolution electron beam lithography system. 4. After patterning, the resist was developed by immersing the sample in a mixed (1:3) solution of MIBK and IPA for about 30 s at room temperature (22°C). 5. The sample was then rinsed with IPA. 6. Undeveloped residues from mask openings were removed using a short oxygen flash in a reactive ion etcher (Oxford Plasmalab 80 plus RIE).

228

Kuzyk, Toppari, and Törmä

Fig. 1. (a) A schematic view of the experimental setup used in the DNA DEP experiments performed in situ under the confocal microscope. The big image shows the nanoelectrode structure and the liquid chamber containing the DNA ­solution on top of it, in addition to the electrical feeds. A closer view of the DEP trap is presented in the upper left inset. An example of a DEP movie obtained from the captured confocal microscope images is shown in the upper middle inset. Grey /black is the reflected image and bright white is the fluorescence from the DNA (schematic 3D images by Tommi Hakala). Still image from a DEP movie obtained by using (b) 5 MHz frequency and 2 × 107 V/m field strength, and (c) 1 MHz frequency and 1.4 × 107 V/m field strength (16). It can be seen that the smaller frequency traps more efficiently, but the electrophoretic forces have more time to spread the molecules, resulting in a wider trapping spot.

Fig. 2. AFM images of one, two, and three thiol-modified 414-bp long dsDNA molecules trapped between the nanoelectrodes using DEP. Note that the electrodes are clipped in z-dimension.

7. Metal evaporation was done in an UHV (ultra-high vacuum) chamber, pressure being of the order of 10−8 mbar during the evaporation. 8. The thickness of the evaporated gold layer was 15–20  nm, under which 1–2  nm of titanium was used as an adhesion layer. 9. After evaporation, the PMMA mask was removed by lift-off in hot acetone.

Trapping and Immobilization of DNA Molecules Between Nanoelectrodes

229

10. PMMA residues were removed from the substrate using oxygen plasma (Oxford Plasmalab 80 plus RIE, parameters: 100 sccm O2 flow, 50 W RF power, and 1 min time). 3.2. Fabrication of DNA Fragments

1. Fragments of 27 bp were generated by mixing equal amounts of the complementary oligonucleotides Primer 1 and Primer 2 in buffer 2 (see Table 1). 2. DNA fragments of 145 and 444 bp were produced by PCR using TAQ polymerase with the oligonucleotides Primer 3 and Primer 4 or Primer 4 and Primer 5, respectively (see Table 1), as primers. Chicken avidin complementary DNA in pFastBac1 plasmid was used as a template in the PCR. 3. The plasmids (pBVboostFG and pFastBac1) were produced by cultivating transformed E. coli JM109 cell line at 37°C in the suspension and isolating the plasmids from the overnight cultures by using the plasmid purification kit. 4. DNA fragments of 1,065 bp were generated by digesting the pBVboostFG plasmid using BglI and SpeI restriction enzymes. 5. Fragments of 5,141  bp were produced by linearizing the pFastBac1 plasmid using HindIII enzyme. 6. Fragments of 8,461 bp were generated by linearizing pBVboostFG using ApaI enzyme. 7. The fragments were purified by 1% agarose gel electrophoresis and isolated with GFX™ PCR DNA and Gel Band purification kit. 8. The concentrations of final products were measured spectrophotometrically. 9. Oligonucleotide Primer 6 was used as a forward primer and Primer 7 as a reverse primer in a PCR process for hexanethiolmodified 414-bp DNA. The PCR was done as above (see Note 4). 10. Oligonucleotide Primer 8 was used as a forward primer and Primer 9 as a reverse primer in a PCR process for DTPAmodified 415-bp DNA. The PCR was done as above (see Note 4).

3.3. DNA Solution for the Experiments

1. For the fluorescence detection of dielectrophoretic trapping under confocal microscope, the DNA fragments (27– 8,416  bp) were labeled with fluorescent label PicoGreen, which is highly selective for dsDNA and gives linear fluorescence signal in the range of 0.2–200 nt of DNA. 2. PicoGreen stock solution was first diluted 1:100 into buffer 1 and then mixed 1:1 with the DNA solution to obtain the final solution, i.e., 1:200 diluted PicoGreen. 3. The final concentrations of the fragments were chosen so that the concentration of the nucleotides (bases) remains the same

230

Kuzyk, Toppari, and Törmä

in all cases (17 mM nucleotides). The PicoGreen dye attaches approximately uniformly along the double helix, which ideally results in the same amount of fluorescence in the solutions of DNA molecules of different length if the concentration of the nucleotides is the same. 4. In the case of immobilization of individual 414/415-bp long thiol-modified dsDNA molecules, the DNA concentration was 1  nM in buffer 2. (The trapping experiments aiming immobilizations of single molecules were not done under the confocal microscope imaging; therefore, fluorescent labels were not used in those experiments.) 5. The ready DNA solution (buffer 1 or buffer 2) was stored in single-use aliquots at 4°C (see Note 5). 3.3.1. Dielectrophoretic Trapping Under Confocal Microscope

1. Dielectrophoresis studies under the confocal microscope (Zeiss Axiovert LSM 510, Zeiss “Fluar” 40×/1.3 oil objective) were done by (fluorescent) imaging a square area (10 × 10 mm2) around the DEP trap (i.e., the area around the gap between the nanoelectrodes on the chip) simultaneously while applying an AC signal (Agilent 33120A waveform generator) to the electrodes used for trapping. Prior to the DEP experiment, the sample (Si-chip containing the nanoelectrodes) was cleaned with a short flash of oxygen plasma (Oxford Plasmalab 80 plus RIE, parameters: 100 sccm O2 flow, 50 W RF power, and 1 min time) in order to remove possible organic contamination from the surface of the sample and the electrodes. 2. To prepare a sample suited for the DEP experiment under the confocal microscope imaging, the sample was attached to a regular microscope slide using scotch tape and the electrical connections were realized by ultrasonic bonding with 25-mmthick Al wire. 3. The DNA solution is placed into a liquid chamber between the nanoelectrode sample and the glass coverslip. The side walls of the chamber were formed by the same scotch tape which also kept the sample on top of the microscope slide (see Note 6). 4. The experimental setup is represented in Fig. 1a. 5. Argon laser (488 nm) with power of 0.45 mW (see Note 7) was used for imaging. Fluorescence data were collected simultaneously using two channels: (1) fluorescence channel (containing 505-nm high pass), which corresponds to the amount of DNA, and (2) reflection channel (containing 475–525-nm band pass), which shows the location of the gap between the electrodes since the laser reflects differently from the substrate and the gold electrodes (see Fig. 1b, c). 6. To optimize the measurement for obtaining accurate information, the vertical resolution of the confocal microscope

Trapping and Immobilization of DNA Molecules Between Nanoelectrodes

231

was minimized to about the height of the electrodes (in our case, ~20  nm), so that the fluorescent background due to freely moving DNA above it is not collected. In addition, the detector sensitivity was maximized by fine-tuning the detector gain and the amplification offset according to the fluorescence background of each sample. 7. The DEP movies were obtained by capturing two 128 × 128 pixel frames per second (image refresh rate 2 Hz). 8. It was found that for successful trapping, electrical fields with a frequency between 100 kHz and 15 MHz should be used with more than 3 Vp-p peak-to-peak amplitude (field strength ~107 V/m for the 100-nm gap between electrodes). Too high voltages should be avoided since they can cause several unwanted effects (see Note 8). 9. Results of successful trappings are shown in Fig.  1b, c (see Note 9). 3.3.2. Dielectrophoretic Trapping of Individual Molecules

1. Samples were cleaned with a short flash of oxygen plasma prior to DEP experiments as above. 2. The samples were connected to a homemade probe station equipped with two electrical feeds designed for fast connectivity. 3. For the attachment of individual thiol-modified 414/415-bp long dsDNA molecules, it was necessary to prevent the molecules becoming attached to each other via the sulfur–sulfur bonds between separate DNA molecules. Thus, a reducing agent (NaBH4) was added to the ready DNA solution at 2 mM concentration about 1 h before the DEP experiment, to break the sulfur–sulfur bonds (see Notes 2 and 3). 4. The trapping was done by incubating a few microliter drop of 1  nM thiol-modified DNA solution onto the substrate containing the electrodes and keeping the sample in a moist chamber to prevent the drop from drying while applying AC voltage (Agilent 33120A waveform generator) to the electrodes. 5. The voltage was applied for ~20 min. 6. After DEP trapping, the drop was gently rinsed with buffer 2 and water, and nitrogen dried (see Note 10). 7. Coarse optimization of the parameters was done beforehand by DEP experiments under the confocal microscope (see Subheading 3.4.1). 8. It was found that for a successful trapping of individual molecules, electrical fields with a frequency of ~1 MHz should be used with more than 3  Vp–p peak-to-peak amplitude (field strength ~107 V/m for the 100-nm gap between electrodes) (see Note 8).

232

Kuzyk, Toppari, and Törmä

9. Examples of individual DNA molecules trapped between electrodes are given in Fig.  2. The yield of trapping was as high as 50%. The procedure can also be used for trapping of DNA self-assembled objects, e.g., DNA origami structures (see Note 11). 3.4. AFM Imaging

1. The AFM (Veeco Dimension 3100) was operated in tapping mode. 2. For imaging in air, the following probes were used: (a) Veeco MPP-11100, resonance frequency ~300  kHz, spring constant ~40 N/m (b) Nanosensors PPP-NCH, resonance frequency ~330 kHz, spring constant ~42 N/m

4. Notes 1. Buffers should have low enough conductivity (well below 1 mS/ cm) for DEP trapping to work. HEPES is a good choice for the buffer, since it is weakly acid and has a permanent polarization; thus the positive ends act as counter ions for the dsDNA, yet, without increasing the ionic conductivity of the buffer. 2. Other reducing agents such as dithiothreitol (DTT) or tris (2 carboxyethyl) phosphine (TCEP)–HCl can also be used. 3. The use of a reducing agent is essential when aiming at the immobilization of individual molecules (16); without the reducing agent, mostly multimers are trapped in or near the gap. 4. Without any linkers, DNA molecules diffuse very fast from the DEP trapping region after the DEP voltage is switched off; that is, unspecific physisorption is not strong enough to keep the objects in the trap (11). 5. Both buffers 1 and 2 were used and no significant difference was noticed. 6. In this configuration, the ends of the liquid chamber are open, which increases the speed of evaporation of the DNA solution. Effective operation time after inserting the DNA solution was between 20 and 40  min. The end can be closed, e.g., by vacuum crease, which not only increases the operation time to hours, but also increases the difficulties and risks during the preparations. 7. Before the actual DEP studies, bleaching tests were performed to confirm that the laser excitation does not cause significant bleaching of the fluorescent dye.

Trapping and Immobilization of DNA Molecules Between Nanoelectrodes

233

8. Too high voltage can cause Joule heating, which further induces convection flows that can drag the DNA away from the trapping area. The heating can also induce extensive evaporation and thus quick drying of the sample. The higher the voltage, or the more conductive the solution, the more probable overheating is. The flows due to heating are usually of the order of mm/s, and forces due to the flows are smaller than the DEP force. However, a phenomenon known as AC-electro-osmosis also happens within the double-layer capacitances formed at the top of the electrodes. The flows due to it can be significant, and the flow is always directed away from the trapping area. Care should be taken not to operate within the range of the parameters where osmosis happens. A good description of various hydrodynamic effects in combination with DEP can be found in ref. 19. 9. It can be seen that smaller frequencies trap more efficiently, i.e., smaller voltages can be used, but the electrophoretic forces have more time to spread the molecules, resulting in a wider trapping spot. 10. Gentle washing with water is an essential step; without it, the surface of a sample is usually covered with salt residues, which makes it extremely difficult to resolve DNA molecules during AFM imaging. On the contrary, extensive washing with water usually results in removal of DNA molecules. 11. One of the difficulties in such experiments is due to the fact that self-assembly is usually done in buffers of relatively high ionic strength (conductivity) to reduce the repulsion between negatively charged DNA backbones. On the contrary, DEP trapping should be performed in a buffer with a relatively low conductivity. Thus, before DEP, buffers should somehow be changed. It is possible to overcome this problem by ligation, which is known to increase mechanical and thermal stability of DNA structures and thus allow changing of buffers. More detailed information about DEP trapping of DNA origami structures can be found in refs. 17, 18.

Acknowledgments This work was supported by the Academy of Finland (Projects No. 117937, No. 118160, No. 115020, No. 213362) and conducted as part of a EURYI scheme program (see http://www.esf. org/euryi). A. K. thanks the National Graduate School in Nanoscience.

234

Kuzyk, Toppari, and Törmä

References 1. LaBean, T.H., and Li, H. (2007) Constructing novel materials with DNA. Nano Today 2, 26–35. 2. Seeman, N. C. (2003) DNA in a material world. Nature 421, 427–431. 3. Aldaye, F. A., Palmer, A. L., and Sleiman, H. F. (2008) Assembling materials with DNA as the guide. Science 321, 1795–9. 4. Bidault, S., Garcia de Abajo, F. J., and Polman, A. (2008) Plasmon-based nanolenses sssembled on a well-defined DNA template. J. Am. Chem. Soc. 130, 2750–1. 5. Keren, K., Berman, R. S., Buchstab, E., Sivan, U., Braun, E. (2003) DNA-templated carbon nanotube field-effect transistor. Science 302, 1380–2. 6. Pohl, H. (1978) Dielectrophoresis the behavior of neutral matter in nonuniform electric fields. (Cambridge University Press., Cambridge.). 7. Burke PJ (2004) Nanodielectrophoresis: Electronic Nanotweezers. Encyclopedia of Nanoscience and Nanotechnology 6, 623–641. 8. Hughes, M. (2000) AC electrokinetics: applications for nanotechnology. Nanotechnology 11, 124–32. 9. Hölzel, R., Calander, N., Chiragwandi, Z., Willander, M. and Bier, F. F. (2005) Trapping single molecules by dielectrophoresis. Phys. Rev. Lett. 95, 128102. 10. Clarke, R.W., Piper, J.D., Ying, L., and Klenerman, D. (2007) Surface conductivity of biological macromolecules measured by nanopipette dielectrophoresis. Phys. Rev. Lett. 98, 198102. 11. Tuukkanen, S., Kuzyk, A., Toppari, J. J., Häkkinen, H., Hytönen, V. P., Niskanen, E., Rinkiö, M., and Törmä, P. (2007) Trapping of 27 bp-8 kbp DNA and immobilization of

12.

13.

14.

15.

16.

17. 18.

19.

thiol-modified DNA using dielectrophoresis. Nanotechnology 18, 295204. Tuukkanen, S., Toppari, J. J., Kuzyk, A., Hirviniemi, L., Hytönen, V. P., Ihalainen, T. and Törmä, P. (2006) Carbon nanotubes as electrodes for dielectrophoresis of DNA. Nano Lett. 6, 1339–43. Barsotti, R. J, Vahey, M. D., Wartena, R., Chiang, Y. M., Voldman, J., and Stellacci, F. (2007) Assembly of metal nanoparticles into nanogaps. Small 3, 488–99. Hakala, T.K., Linko, V., Eskelinen, A-P., Toppari, J. J., Kuzyk, A., Törmä. P. (2009) Field induced nanolithography for high-throughput pattern transfer. Small 5, 2683. Vijayaraghavan, A., Blatt, S., Weissenberger, D., Oron-Carl, M., Hennrich, F., Gerthsen, D., Hahn, H., and Krupke, R (2007) Ultra-largescale directed assembly of single-walled carbon nanotube devices. Nano Lett. 7, 1556–60. Tuukkanen, S., Kuzyk, A., Toppari, J. J., Hytönen, V. P., Ihalainen, T., and Törmä, P. (2005) Dielectrophoresis of nanoscale doublestranded DNA and humidity effects on its elec­ trical conductivity. Appl. Phys. Lett. 87, 183102 . Kuzyk, A., Yurke, B., Toppari, J.J., Linko, V., and Törmä P (2008) Dielectrophoretic trapping of DNA origami. Small 4, 447–50. Linko, V., Paasonen, S. T., Kuzyk, A., Törmä, P., and Toppari, J. J. (2009) Characterisation of the conductance mechanisms of the DNA origami by AC impedance spectroscopy. Small 5, 2382. Castellanos, A., Ramos, A., González, A., Green, N. G., and Morgan, H. (2003) Electrohydrodynamics and dielectrophoresis in microsystems: scaling laws. J. Phys. D: Appl. Phys. 36, 2584–97.

Chapter 17 DNA Contour Length Measurements as a Tool for the Structural Analysis of DNA and Nucleoprotein Complexes Claudio Rivetti Abstract The atomic force microscope (AFM) is a widely used tool to image DNA and nucleoprotein complexes at the molecular level. This is because the AFM is relatively easy to operate, has the capability to image biomolecules under aqueous solutions, and, most importantly, can image mesoscopic macromolecular structures that are too complex to be studied by X-ray or NMR and too small to be visualized with the optical microscope. Although there are many AFM studies about the structure and the physical properties of DNA, only in few cases a rigorous method has been applied to analyze AFM images. This chapter describes procedures to prepare DNA and nucleoprotein complexes for AFM imaging and methods used to carry out simple image measurements to obtain structural data. In particular, methods to measure DNA contour length and the volume of free or DNA-bound proteins are presented and discussed. Key words: Atomic force microscopy, Image analysis, Contour length, Volume, DNA, Nucleoprotein complex

1. Introduction Imaging DNA and nucleoprotein complexes by atomic force microscopy (AFM) can provide valuable information about the physical properties and the structure of these molecules (1–4). Although the resolution of the AFM does not allow direct visualization of the atomic details of biomolecules, analysis of mesoscopic molecular features can be used to infer microscopic structural properties. For example, the measure of the distance between two consecutive base pairs of the DNA helix cannot be performed directly, because AFM cannot resolve single bases

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_17, © Springer Science+Business Media, LLC 2011

235

236

Rivetti

within a DNA fragment even with very sharp tips (radius of curvature of ~1  nm). However, the contour length of a DNA molecule is a function of the base pair distance, and thus, based on the periodic structure of DNA, it is possible to determine an average rise per base pair from the measured DNA contour length and the number of base pairs of the DNA fragment. With this approach, it has been shown that DNA deposited onto mica has an average rise per base pair of 3.2 Å, a value in between the canonical rise per base pair of B-form DNA (3.4 Å) and that of A-form DNA (2.6 Å) determined by X-ray crystallography (5). The contour length and the shape of DNA fragments imaged by AFM can also be used to measure physical properties such as the persistence length of DNA. The persistence length is related to the energy required to bend DNA: it is a measure of DNA flexibility and can be defined as the decay length through which the memory of the initial orientation of the molecule persists. DNA fragments imaged by AFM and analyzed using the worm-like chain (WLC) model gave a persistence length of 53  nm (2). Likewise, an extended version of the WLC model has been successfully applied to bent DNA molecules (3), showing that an intrinsic or protein-induced DNA bend angle can be determined from the position of the bend, the mean square end-to-end distance, and the contour length of the DNA fragment. Many cell processes are regulated by DNA-binding proteins that bind to specific DNA sequences. The AFM analysis of nucleoprotein complexes can provide useful information regarding their DNA-binding properties. For instance, by accurately measuring the distance of the DNA bound protein from the DNA ends, it is possible to locate DNA-binding sites within large DNA fragments with a resolution up to 10  bp (6, 7). This approach has been applied to study the interactions of the repressor TraA with key DNA-binding sites located within the regulatory gene cluster involved in the sex pheromone-stimulated mating response of Enterococcus faecalis (7). The study has shown that the pheromone peptide cPD1 modifies the DNA specificity of TraA in such a way that TraA moves from the transcriptional repression site to a nearby DNA site with consequent transcription activation. AFM images of these nucleoprotein complexes have also revealed binding of TraA to DNA sites located within the ipd terminator region, which may play an important role in controlling the expression of downstream genes via a transcriptional read-through mechanism. DNA contour length measurements can also give a measure of the extent of protein–DNA interaction (DNA wrapping) of nucleoprotein complexes (8–13). These measurements are based on the observation that the DNA wrapped around a protein shortens the overall DNA contour length of the complexes by an amount that is proportional to the degree of wrapping. An example is given by Escherichia coli RNA polymerase (RNAP) which

DNA Contour Length Measurements

237

binds to the promoter, forming a transcriptional competent open complex (RPo). DNA contour length analysis of these RPo has shown that binding of the RNAP to the promoter determines a DNA foreshortening of 30 nm, a length that is compatible with one turn of DNA wrapping around the RNAP (8). Further extending this analysis to wild-type and mutated promoters, it has been shown that DNA wrapping depends on the sequence of the promoter, and mutations aimed to eliminate AT-rich sequences from the upstream region of the promoter had, as effect, partial or total loss of DNA wrapping (12, 13). Accurate measurement of the DNA contour length from AFM images is thus a valuable tool for the structural analysis of DNA and protein–DNA complexes and provides data that are difficult to obtain with other techniques. This chapter describes methods for preparing and imaging DNA with the AFM microscope and image processing procedures used to estimate the DNA contour length and the volume of nucleoprotein complexes accurately.

2. Materials 1. TAE buffer: 40  mM Tris–acetate, pH 8.0, and 2  mM EDTA. 2. TE buffer: 10 mM Tris–HCl, pH 7.5, and 1 mM EDTA. 3. Agarose powder for DNA electrophoresis. Gel purification is usually performed in 1% agarose in TAE buffer. 4. Elutrap device (Schleicher & Schuell – http:// www.whatman. com). Assemble the Elutrap following the instructions given in the user guide. Use TAE as running buffer. If the gel slice is too large, cut it into small pieces. 5. QIAquick gel extraction kit (Qiagen – http:// www.qiagen. com). 6. Glutaraldehyde solution 50% in H2O (Sigma–Aldrich – http:// www.sigmaaldrich.com). 7. Mica substrate. With scissors, cut a square piece (~1 × 1 cm) of ruby mica (Ted Pella, Redding CA – http:// www.tedpella. com) and glue it on a metal disk with superglue adhesive. Cleave the mica with scotch tape immediately before sample deposition. A mica substrate can be cleaved a dozen times after which it can be unglued by soaking in acetone. 8. AFM cantilevers with integrated probes. Regular tapping mode probes are usually sharp enough for DNA and nucleoprotein imaging. We use NSC15/AlBS from MikroMasch (http:// www.spmtips.com) with a resonant frequency around 320 kHz and a force constant of about 40 N/m.

238

Rivetti

9. Deposition buffer: 4  mM HEPES, pH 7.4, 10  mM NaCl, and 2 mM MgCl2. 10. Milli-Q water (Millipore – http:// www.millipore.com). 11. Software: Image processing and data analysis are performed with scripts written in Matlab and available on the Internet or from the author on request. Image files in Nanoscope III format are read directly into Matlab, and the 16-bit integer intensities of the image are converted into nanometer scale using conversion factors written in the file header. The “Image processing toolbox” of Matlab is also recommended because it contains a number of useful built-in routines for image transformation, display, and analysis.

3. Methods 3.1. DNA Preparation

There are several methods that can be used to prepare DNA suitable for AFM imaging: isolation from cells, restriction digestion, or PCR. Regardless of the preparation procedure used, the DNA must be clean and possibly stored in a low salt buffer ­solution containing EDTA. It is often convenient to gel-purify restriction digestion or PCR products in order to obtain DNA fragments homogeneous in size and to eliminate the excess of oligonucleotides from the PCR. A good and reliable method to recover DNA from the gel slice is electroelution by means of an Elutrap device, followed by phenol–chloroform extraction and ethanol precipitation. Alternatively, commercially available DNA purification kits can be used. The recovered DNA is resuspended in TE buffer and stored at 4°C. Although there is no limitation on the size of the DNA fragments that can be imaged with the AFM, practically, DNA shorter than 100 bp are difficult to image because they have the tendency to contract, whereas DNA longer than 10 kbp cross themselves very often, making complex shapes that are difficult to interpret. 1. Assemble and carry out the restriction digestion or the PCR. 2. Run the product in an agarose gel together with a size marker. 3. Stain the gel with ethidium bromide or other DNA staining solution. 4. With a sharp blade, cut the gel slice containing the DNA band. 5. If a DNA purification kit is used, follow the protocol reported in the product guide and skip the following steps. 6. Load the gel slice in the Elutrap device and run the electroelution.

DNA Contour Length Measurements

239

7. Recover the solution containing the DNA (see Note 1). 8. Extract the recovered solution with phenol twice, phenol/ chloroform once, and chloroform once. 9. Precipitate the DNA with 95% ethanol and wash the pellet with 70% ethanol. 10. Dry and resuspend the DNA pellet in TE buffer. 11. Store at 4°C. 3.2. Preparation of Nucleoprotein Complexes

The preparation procedure of protein–DNA complexes suitable for AFM cannot be standardized because different protein–DNA complexes may require different buffer and different incubation conditions in order to be formed. However, there are several aspects of the preparation that can be applied to most cases. It is generally convenient to prepare the complexes under optimal buffer and temperature conditions and at concentrations in the range of 10–100 nM DNA. The solution can thus be diluted in deposition buffer to a final concentration of 1–2  nM DNA. Usually, this step also dilutes the salts of the reaction to a concentration that does not interfere with the adhesion of the complexes to the mica surface. In the case of proteins that have a single binding site within the DNA fragment used, avoid using a large excess of protein because it can saturate the mica surface either in a way that the complexes can no longer bind to the surface or in a way that makes the picture too crowded to be meaningful. In some cases, particularly when dealing with oligomeric proteins, it may happen that the protein–DNA complexes dissociate during deposition. To increase complex stability, the nucleoprotein complexes can be cross-linked with glutaraldehyde (0.11% final concentration). The complexes can then be diluted in deposition buffer and deposited onto the freshly cleaved mica substrate. 1. In 20 ml of an appropriate reaction buffer, add DNA to a final concentration of 20 nM. 2. Add the DNA-binding protein in order to have a protein/ DNA-binding-site molar ratio between 1 and 5 (the amount of protein can be adjusted after a first inspection of the images). 3. Incubate the reaction at an appropriate temperature (in most cases, 20 or 37°C) and for an amount of time that can vary from case to case. 4. If necessary, cross-link the complexes by adding 0.2 ml of 11% glutaraldehyde and incubating the reaction at 4°C or room temperature for few minutes. Quench the cross-linking reaction by adding 1 ml of 1 M Tris–HCl, pH 8.0 (see Note 2). 5. Add 2  ml of the reaction to 18  ml of deposition buffer and immediately deposit 20 ml of this solution onto freshly cleaved mica as described below (see Note 3).

240

Rivetti

3.3. Deposition of DNA or Nucleoprotein Complexes onto Mica

A simple and reliable way to deposit DNA for AFM imaging is to use freshly cleaved mica as a substrate and Mg++ cations as a means to stick the negatively charged DNA onto the negatively charged mica surface (2, 14). Other methods for DNA deposition include the use of APTES-coated mica (15), polylysine-coated mica (16), or l-ornithine-coated mica (17). Deposition of DNA onto mica (or onto any other surface) implies that the molecules undergo a change in dimensionality from 3D (in solution) to 2D (on the surface). A polymer-like DNA can undergo the 3D → 2D transition in two different ways: (a) the molecules may freely move and equilibrate on the surface as in a 2D solution before they are captured in their final conformation; and (b) DNA may stick to the surface so strongly that discrete parts of the polymer are unable to move from the initial binding site and the DNA is trapped in a conformation that resembles that of the 3D → 2D projection of the molecule. Any intermediate state between these two extremes is also possible. It has been shown that meaningful information about the DNA structure and flexibility can be extracted either from DNA molecules at equilibrium in 2D or from DNA molecules trapped on the surface. However, in our experience, trapped DNA molecules are difficult to trace and measure because of the many kinks and crossovers; in addition, the conformation of trapped DNA molecules does not exactly correspond to their 3D → 2D projection because during the deposition process, the DNA contour length must be conserved, forcing the molecules to deviate from a pure 2D projection. For these reasons, we ­recommend deposition conditions that favor DNA equilibration. 1. Prepare a freshly cleaved mica disk by peeling off the top layer with scotch tape. 2. Dilute DNA or protein–DNA complexes in deposition buffer to a concentration of about 2  nM (this concentration is referred to the DNA molecules). 3. Deposit 20 ml of the diluted solution onto the freshly cleaved mica. 4. Incubate the sample for about 2 min at room temperature. 5. Rinse the surface with Milli-Q water using a squeeze bottle. 6. Blot the excess of water and dry the surface with a weak flux of nitrogen (see Note 4). 7. Load the sample on the AFM and start imaging. If the surface appears too crowded or too scantly covered with molecules, adjust the deposition time at point 4 accordingly.

3.4. DNA Contour Length Measurements

Determination of the DNA contour length from AFM images, in which molecules are described by a subset of pixels in a 2D grid, is not a trivial task. This is because the DNA contour in the digital image is only an approximation of the exact contour of the DNA

DNA Contour Length Measurements

241

molecule; thus, the contour length can only be estimated rather than exactly determined. The accuracy of such estimate depends on both image resolution and the method employed to compute the contour length from the string of pixels representing the molecule. While the resolution is an inherent property of the microscope or of the microscope–sample system (18), two operational steps of image processing can significantly affect the reliability of DNA contour length determination from AFM images. These are the correct identification of the subset of pixels that best describe the backbone of the DNA molecule in the image (segmentation) and the use of an algorithm capable of yielding the most accurate estimate of the DNA contour length from the image contour. 3.4.1. Digitization of the DNA Backbone

Image segmentation methods are used to identify the contour of single DNA molecules within an AFM image. These methods can be completely automated or can involve user action during the digitization process (5, 19, 20). Fully automated procedures are very efficient and can detect thousands of DNA molecules in few minutes. However, their application to AFM images of DNA is complicated by the heterogeneity of the sample (particularly in the case of protein-bound DNA molecules), image noise, and the fact that long DNA fragments have often a complex shape. Manual procedures in which DNA molecules are digitized with the aid of a user intervention (generally through mouse clicks) are more time consuming, but are usually preferred because they provide a more homogeneous data set. There are a number of software that can be utilized to measure filaments imaged by AFM; however, many of them use different algorithms to compute the contour length and, in most cases, the algorithm employed is not ideal to measure DNA. It is, therefore, convenient to use ad hoc computer programs employing algorithms that have been optimized for DNA contour length measurements. Matlab is a very powerful numerical computing environment and programming language that can be readily used for processing AFM images with little coding.

3.4.2. Computation of the DNA Contour Length

There are several algorithms that can be used to compute the DNA contour length from a given chain code; they mainly differ in accuracy and ease of implementation. Several of these algorithms are described in details in ref. 5, and their applicability to the determination of the DNA contour length has been tested with real and simulated DNA molecules. Here, the use of a recently proposed “DNA estimator” (21) that is both accurate for DNA contour length measurements and easy to implement is described. The DNA estimator is based on the (ne,no) characterization which considers the string of pixels of a segment as composed by a number of even (ne) and odd (no) chain code elements. Even chain code elements correspond to horizontally or vertically

242

Rivetti

connected pixels, whereas odd chain code elements correspond to diagonally connected pixels. At variance with straight lines, it has been shown that a DNA length estimator depends on both the DNA flexibility (i.e., DNA persistence length) and the image resolution; thus, the coefficients a and b of the (ne, no) characterization general formula ( L (ne , no ) = ane + bno) are different under different imaging conditions (see Table 1 in ref. 21). DNA imaged under standard buffer conditions has a persistence length of ~50 nm, and in the case of AFM images collected with a scan size of 2,000  nm and an image size of 512 × 512 pixels (3.91  nm/ pixel), the resulting formula of the DNA contour length estimator is as follows (21):

L DNA = (0.964n e + 1.363n o )S , Where LDNA is the DNA contour length, ne and no are the number of even and odd chain code elements, and S is the scale factor expressed in nanometer per pixel. The scale factor is obtained from the ratio between the scan size and the number of pixels per scan line. The image processing procedures required to measure the DNA contour length can be described with the following steps: Automatic DNA tracing: 1. By using Matlab, load the AFM image matrix (see Fig.  1a) (see Note 5). 2. Compute the threshold level to convert the grayscale image to a binary image by using the graythresh function (see Fig. 1d). 3. Convert the grayscale image into a binary image using the im2bw function (see Fig. 1b). 4. Perform morphological operations on the binary image in order to obtain a continuous eight-connected object in which the pixels are contacted by no more than two other pixels (see Fig. 1c). 5. Invoke the bwboundaries function to find the boundaries of the objects in the binary image and to retrieve their pixel coordinates. 6. By looping through the pixels of an object, the number of even and odd chain code elements is counted based on the scheme shown in Fig. 1g. 7. DNA contour length is determined with the formula shown above. Manual DNA tracing: 1. By using Matlab, load the AFM image matrix and display it on the computer screen (see Fig. 1d and Note 5).

243

DNA Contour Length Measurements

g

Eight-connected chaincode 2 3

1

4

0

7

5 6 Chaincode string: 100111210077756

ne = 6

no = 9

Fig. 1. Automatic and manual image processing operations for DNA contour length measurements. (a) A view of a DNA molecule imaged by AFM. (b) Binary transformation of the image in (a). (c) Eight-connected contour of the DNA molecule shown in (a). (d) DNA molecule imaged by AFM and manually traced with the mouse. (e) Binary image created using the coordinates of the manually traced path. (f) Eight-connected contour of the DNA molecule shown in (a). (g) Scheme of an eight-connected object and the corresponding chain code.

2. Invoke the improfile function to draw the DNA path by ­clicking with the mouse along the DNA from one end to the other (see Fig. 1d and Note 6). 3. Matlab retrieves the image profile and returns the coordinates of the pixels crossed by the line path. These coordinates are

244

Rivetti

used to create a binary image in which the pixels of the path are set to one (see Fig. 1e and Note 7). 4. Morphological operations are performed on the binary image in order to obtain a continuous eight-connected object in which the pixels are contacted by no more than two other pixels (see Fig. 1f ). 5. By looping through the pixels of the object, the number of even and odd chain code elements is counted based on the scheme shown in Fig. 1g. 6. DNA contour length is determined with the formula shown above. Two Matlab scripts that can be used to measure the contour length of DNA imaged by AFM are reported in Appendix. The first script utilizes image segmentation procedures built-in in the Matlab environment to find the pixels of the DNA backbone without user intervention, whereas the second script requires a manual tracing of the DNA in order to identify the sequence of pixels corresponding to the DNA backbone. Both scripts then use the DNA estimator shown above to compute the DNA contour length from the number of even and odd chain code elements. 3.5. Volume Measurements for Protein Stoichiometry Analysis

AFM is a topographic imaging technique and when operating in “height mode” the pixels values correspond to the height of the sample at each x,y position of the scanned area (see Fig. 2a, b). In  the case of DNA or nucleoprotein complexes, the height is referred to a base plane represented by the flat substrate onto which the molecules are adsorbed. Therefore, from these AFM images, it is possible to measure objects in 3D and to measure not only their surface area but also their volume. It must be pointed out that neither the height nor the surface area of a particle imaged by AFM are accurate: usually, particles are compressed by the scanning tip particularly when operating in tapping mode, and their lateral dimensions are enlarged by the tip-broadening effect. However, when images are obtained under very similar conditions and volume data are used in a comparative manner, it is possible to obtain information about the molecular weight and the stoichiometry of protein assemblies. The volume of globular features in AFM images is often calculated with a geometrical approach, assuming that the particle has the shape of an oblate spheroid and using dimensions (height and width) obtained from cross-section profiles (22, 23). Although in principle correct, this procedure is approximated and the error adds to the uncertainties introduced by the tip effect. A better approach for volume determination is to use the entire 3D information of the image. Namely, in the AFM image, each particle

Fig. 2. Volume determination of a particle imaged by AFM. (a) Top view of the particle. Contrast is obtained by mapping the pixel intensities (in nanometers) into a grayscale colormap. (b) Tilted view of the image in (a) showing the threedimensionality of the particle. (c) Two-dimensional arrangement of the pixel grid of the image. (d) Lateral view of the bar representation of the image. Each bar represents a single pixel. Note that the bars corresponding to the particle emerge from the background and that the average background height is about 4 nm. This offset must be taken into account when computing the volume. (e) The black pixels define the boundary of the particle (ROI) which can be found with image thresholding procedures. (f) Three-dimensional representation of the particle boundary. The average value of these pixels can be used to determine the background offset. (g) Binary representation of the image. The volume of the particle is calculated by multiplying the area of the particle (black pixels) by the average height, relative to the background, of the same set of pixels.

246

Rivetti

is represented by a group of pixels within a region of interest (ROI) defining its contour (see Fig. 2c, f). Each pixel has a welldefined size: the lateral dimension is given by the ratio of the scan size to the number of pixels in each scan line and the height corresponds to the pixel intensity expressed in nanometers. The volume of the particle can thus be calculated from the sum of the volumes of all the pixels within the particle ROI minus the volume of the base plane (background) with the same area. In other words, the volume of a particle is calculated by multiplying the area of the particle (see Fig. 2g) by the average pixel height, relative to the baseline height, of all the pixels within the ROI (24). A simple way to determine the baseline height is to average the intensities of the ROI [i.e., the average value of the outer contour of the particle (see Fig. 2e, f)]. This procedure of volume determination is of general application: it does not require assumptions about the particle shape and can be readily applied to any particle imaged by AFM. 1. By using Matlab, load the AFM image matrix (see Note 8). 2. Compute the threshold level to convert the grayscale image to a binary image by using the graythresh function. 3. Convert the grayscale image into a binary image using the im2bw function. 4. Perform morphological operations on the binary image in order to enlarge the particle for better background determination. 5. Invoke the bwboundaries function to find the boundary of the particle in the binary image and to retrieve its pixel coordinates. 6. If necessary, convert the image values in nanometers using an appropriate scale factor. 7. By looping through the boundary pixels of the particle, compute the average background of the particle surrounding. 8. Compute the object volume by summing up the volume, relative to the background, of pixels within the particle. A Matlab script that can be used to measure the volume of particles in AFM images is reported in Appendix. Estimate of the protein molecular weight from volume measurements is complicated by the reasons exposed above. However, it is possible to determine the molecular weight of an unknown protein or protein–DNA complex by using a calibration curve obtained by measuring the volume of proteins with known molecular weights (22, 25). Because the volume measured by AFM is strongly affected by the tip sharpness, it is recommended to acquire the images with the same tip and under similar imaging conditions.

DNA Contour Length Measurements

247

4. Notes 1. Before pipetting out the solution from the Elutrap trapping well, invert the power supply current for 10  s in order to detach the DNA from the BT1 membrane. Recover the solution and rinse with 200 ml of buffer. 2. In order to cross-link the nucleoprotein complexes with glutaraldehyde, it is necessary to avoid amine-containing buffers, such as Tris, during the preparation of the complexes. 3. We have found that sometimes the monovalent salt concentration of the solution containing the protein–DNA complexes is too high to allow sticking of the complexes to the mica. In these cases, the deposition buffer must be supplemented with extra MgCl2 up to a final concentration that is at least 1/10 of the monovalent salt concentration. Usually, rising the MgCl2 concentration to 4 or 5  mM is sufficient to obtain good DNA deposition. 4. Perform the washing and drying steps by holding the mica disk with tweezers far from possible sources of dust. 5. Depending on the type of microscope used, AFM images can be saved in different formats. However, for DNA contour length measurements, images can be exported as tiff or jpg files without loss of information. Alternatively, Matlab can also be programmed to read the raw image data directly from the microscope file if the file format is known. 6. In the case of nucleoprotein complexes, the DNA that is in contact with the protein is not visible because of the tipbroadening effect; therefore, the DNA trace is made such to pass through the center of the protein. Alternatively, only the visible DNA can be traced going from the ends up to the protein edges. 7. To increase the accuracy of the measurements, the DNA path can be adjusted in order to match the pixels with higher intensity within neighbor pixels. 8. Volume determination of particles requires that the image values are expressed in nanometers. Thus, particular attention must be given to image export procedures in order to maintain the height information. Saving the images as Gif or Jpeg converts heights into RGB colors, making the image useless for volume measurement. It is more convenient to export the image as a grayscale tiff file or as a raw data ASCII file. However, in both cases, the image values must be transformed in nanometers using appropriated conversion factors.

248

Rivetti

Acknowledgments This work was supported by a grant from the Fondazione Cariparma and by the University of Parma.

5. Appendix

%MATLAB SCRIPT TO MEASURE THE DNA CONTOUR LENGTH USING %A FULLY AUTOMATED PROCEDURE %Requires: Matlab 7.0(R14) and the Image Processing Toolbox

%Set the pixel dimension (the ratio between the scansize %(in nanometers) and the image size (in pixels) PixelSize = 3.9063;

%Read the grayscale image of a DNA molecule from a tiff file I = imread('dna.tif');

%Thresholding: compute the image threshold using Otsu's method threshold = graythresh(I);

%Convert the grayscale image to a black and white (binary) image BW = im2bw(I,threshold);

%Perform morphological operations on binary image to obtain %an 8-connected object BW = bwmorph(BW, 'thin', Inf);

%Trace the DNA boundaries B = bwboundaries(BW,8,'noholes');

DNA Contour Length Measurements

%Eliminate duplicated rows from boundary boundary = unique(B{1}, 'rows');

%Initialize variables for counting even and odd chaincode elements Neven = 0; Nodd = 0;

%Loop through the boundary pixels to count even and odd %chaincode elements for i=1:size(boundary,1) row = boundary(i,1); col = boundary(i,2); block = BW(row-1:row+1,col-1:col+1); Neven = Neven + block(2) + block(4) + block(6) + block(8); Nodd

= Nodd + block(1) + block(3) + block(7) + block(9);

BW(row,col) = 0; end

%Compute the DNA contour length

using the DNA estimator

%formula and the scale factor DNA_LENGTH = (0.964*Neven + 1.363*Nodd) * PixelSize;

249

250

Rivetti

%MATLAB SCRIPT TO MEASURE THE DNA CONTOUR LENGTH USING %A MANUAL PROCEDURE %Requires: Matlab 7.0(R14) and the Image Processing Toolbox

%Set the pixel dimension (the ratio between the scansize %(in nanometers) and the image size (in pixels) PixelSize = 3.9063;

%Read the grayscale image of a DNA molecule from a tiff file I = imread('dna.tif');

%Initialize an empty binary image BW = zeros(size(I));

%Display the image on the screen imshow(I);

%With the mouse click along the DNA backbone from one end %to the other. Hit 'Return' when done [x,y,p] = improfile;

%Round the profile coordinates to the nearest integer x=round(x); y=round(y);

%In the binary image set the DNA profile pixels to one for i=1:length(x) BW(y(i),x(i))=1; end

DNA Contour Length Measurements

%Perform morphological operations on binary image to fill %holes and obtain an 8-connected object BW = bwmorph(BW, 'bridge'); BW = bwmorph(BW, 'thin', Inf);

%Trace the DNA boundaries B = bwboundaries(BW, 8, 'noholes');

%Eliminate duplicated rows from the boundary boundary = unique(B{1}, 'rows');

%Initialize variables for counting even and odd chaincode elements Neven = 0; Nodd = 0;

%Loop through the boundary pixels to count even and odd %chaincode elements for i=1:size(boundary,1) row = boundary(i,1); col = boundary(i,2); block = BW(row-1:row+1,col-1:col+1); Neven = Neven + block(2) + block(4) + block(6) + block(8); Nodd

= Nodd + block(1) + block(3) + block(7) + block(9);

BW(row,col) = 0; end

%Compute the DNA contour length

using the DNA estimator

%formula and the scale factor DNA_LENGTH = (0.964*Neven + 1.363*Nodd) * PixelSize;

251

252

Rivetti

%MATLAB SCRIPT TO MEASURE THE VOLUME OF OBJECTS IN AFM IMAGES %Requires: Matlab 7.0(R14) and the Image Processing Toolbox

%Set the pixel dimension (the ratio between the scansize %(in nanometers) and the image size (in pixels) PixelSize = 3.9063;

%Set the conversion factor used to transform the pixel %values into nanometers HeightScaleFactor = 6.2564;

%Read the grayscale image of a protein molecule from a tiff file I = imread('protein.tif');

%Thresholding: compute the image threshold using Otsu's method threshold = graythresh(I);

%Convert the grayscale image to a black and white (binary) image BW = im2bw(I,threshold);

%Enlarge the object boundaries BW = bwmorph(BW,'dilate');

%Trace the protein boundaries [B,L] = bwboundaries(BW, 8, 'noholes');

%Define the coordinates of the boundary x = B{1}(:,1); y = B{1}(:,2);

DNA Contour Length Measurements

253

%Convert the image grayscale values into heights in nanometers H = double(I) * HeightScaleFactor;

%Compute the image background by averaging the contour %pixel heights Bk=[]; for i=1:length(x) Bk=[Bk;H(x(i),y(i))]; end Background=mean(Bk);

%Find the pixels corresponding of the binary object Obj = find(L==1);

%Initialize the volume variable Volume = 0;

%Loop through the pixel of the object to compute %the volume in nm3 for (i=1:length(Obj)), Volume = Volume + (H(Obj(i))-Background) * (PixelSize^2); end

References 1. Bustamante, C, Rivetti, C (1996). Visualizing protein-nucleic acid interactions on a large scale with the scanning force microscope. Ann. Rev. Biophys. Biomol. Struc. 25, 395–429. 2. Rivetti, C, Guthold, M, Bustamante, C (1996). Scanning Force Microscopy of DNA deposited on mica: Equilibration versus kinetic trapping studied by polymer chain analysis. J. Mol. Biol. 264, 919–932. 3. Rivetti, C, Walker, C, Bustamante, C (1998). Polymer chain statistics and conformational analysis of DNA molecules with bends or sections of different flexibility. J. Mol. Biol. 280, 41–59. 4. Zuccheri, G, Scipioni, A, Cavaliere, V, Gargiulo, G, De Santis, P, Samori, B (2001).

Mapping the intrinsic curvature and flexibility along the DNA chain. Proc. Natl. Acad. Sci. USA. 98, 3074–3079. 5. Rivetti, C, Codeluppi, S (2001). Accurate length determination of DNA molecules visualized by atomic force microscopy: evidence for a partial B- to A-form transition on mica. Ultramicroscopy 87, 55–66. 6. Moreno-Herrero, F, Herrero, P, Colchero, J, Baro, AM, Moreno, F (2001). Imaging and mapping protein-binding sites on DNA regulatory regions with atomic force microscopy. Biochem. Biophys. Res. Commun. 280, 151–157. 7. Folli, C, Mangiarotti, L, Folloni, S, Alfieri, B, Gobbo, M, Berni, R et al. (2008). Specificity

254

8.

9.

10.

11.

12.

13.

14.

15.

Rivetti of the TraA-DNA interaction in the regulation of the pPD1-encoded sex pheromone response in Enterococcus faecalis. J. Mol. Biol. 380, 932–945. Rivetti, C, Guthold, M, Bustamante, C (1999). Wrapping of DNA around the E. coli RNA polymerase open promoter complex. EMBO J. 18, 4464–4475. Verhoeven, EE, Wyman, C, Moolenaar, GF, Hoeijmakers, JH, Goosen, N (2001). Archi­ tecture of nucleotide excision repair complexes: DNA is wrapped by UvrB before and after damage recognition. EMBO J. 20, 601–611. Rivetti, C, Codeluppi, S, Dieci, G, Bustamante, C (2003). Visualizing RNA extrusion and DNA wrapping in transcription elongation complexes of bacterial and eukaryotic RNA polymerases. J. Mol. Biol. 326, 1413–1426. Heddle, JG, Mitelheiser, S, Maxwell, A, Thomson, NH (2004). Nucleotide binding to DNA gyrase causes loss of DNA wrap. J. Mol. Biol. 337, 597–610. Cellai, S, Mangiarotti, L, Vannini, N, Naryshkin, N, Kortkhonjia, E, Ebright, RH et al. (2007). Upstream promoter sequences and alphaCTD mediate stable DNA wrapping within the RNA polymerase-promoter open complex. EMBO Rep. 8, 271–278. Mangiarotti, L, Cellai, S, Ross, W, Bustamante, C, Rivetti, C (2009). Sequence-dependent upstream DNA-RNA polymerase interactions in the open complex with lambdaPR and lambdaPRM ­promoters and implications for the mechanism of promoter interference. J. Mol. Biol. 385, 748–760. Sushko, ML, Shluger, AL, Rivetti, C (2006). Simple model for DNA adsorption onto a mica surface in 1:1 and 2:1 electrolyte solutions. Langmuir 22, 7678–7688. Lyubchenko, YL, Shlyakhtenko, LS, Harrington, RE, Oden, PI, Lindsay, SM (1993). Atomic force microscopy of long DNA: Imaging in air and under water. Proc. Natl. Acad. Sci. USA. 90, 2137–2140.

16. Bussiek, M, Mucke, N, Langowski, J (2003). Polylysine-coated mica can be used to observe systematic changes in the supercoiled DNA conformation by scanning force microscopy in solution. Nucleic Acids Res. 31, e137. 17. Podesta, A, Imperadori, L, Colnaghi, W, Finzi, L, Milani, P, Dunlap, D (2004). Atomic force microscopy study of DNA deposited on poly L-ornithine-coated mica. J. Microsc. 215, 236–240. 18. Bustamante, C, Rivetti, C, Keller, DJ (1997). Scanning Force Microscopy under aqueous solution. Current Opinion in Structural Biology 7, 709–716. 19. Ficarra, E, Benini, L, Macii, E, Zuccheri, G (2005). Automated DNA fragments recognition and sizing through AFM image processing. IEEE Trans. Inf. Technol. Biomed. 9, 508–517. 20. Marek, J, Demjenova, E, Tomori, Z, Janacek, J, Zolotova, I, Valle, F et al. (2005). Interactive measurement and characterization of DNA molecules by analysis of AFM images. Cytometry A. 63, 87–93. 21. Rivetti, C (2009). A simple and optimized length estimator for digitized DNA contours. Cytometry A. 75A, 854–861. 22. Tang, M, Cecconi, C, Bustamante, C, Rio, DC (2007). Analysis of P element transposase protein-DNA interactions during the early stages of transposition. J. Biol. Chem. 282, 29002–29012. 23. Minh, PN, Devroede, N, Massant, J, Maes, D, Charlier, D (2009). Insights into the architecture and stoichiometry of Escherichia coli PepA*DNA complexes involved in transcriptional control and site-specific DNA recombination by atomic force microscopy. Nucleic Acids Res. 37, 1463–1476. 24. Russ, J. C. (1995) The Image Processing Handbook. 2nd Ed. CRC, London. 25. Wyman, C, Grotkopp, E, Bustamante, C, Nelson, HCM (1995). Determination of heat-shock transcription factor 2 stoichiometry at looped DNA complexes using scanning force microscopy. EMBO J. 14, 117–123.

Chapter 18 DNA Molecular Handles for Single-Molecule Protein-Folding Studies by Optical Tweezers Ciro Cecconi, Elizabeth A. Shank, Susan Marqusee, and Carlos Bustamante Abstract In this chapter, we describe a method that extends the use of optical tweezers to the study of the folding mechanism of single protein molecules. This method entails the use of DNA molecules as molecular handles to manipulate individual proteins between two polystyrene beads. The DNA molecules function as spacers between the protein and the beads, and keep the interactions between the tethering surfaces to a minimum. The handles can have different lengths, be attached to any pair of exposed cysteine residues, and be used to manipulate both monomeric and polymeric proteins. By changing the position of the cysteine residues on the protein surface, it is possible to apply the force to different portions of the protein and along different molecular axes. Circular dichroism and enzymatic activity studies have revealed that for many proteins, the handles do not significantly affect the folding behavior and the structure of the tethered protein. This method makes it possible to study protein folding in the physiologically relevant low-force regime of optical tweezers and enables us to monitor processes – such as refolding events and fluctuations between different molecular conformations – that could not be detected in previous force spectroscopy experiments. Key words: Laser tweezers, DNA handles, Protein–DNA chimeras, Single-molecule mechanical manipulation, Protein folding

1. Introduction During the last decade, single-molecule force spectroscopy has emerged as a new and powerful technique to study protein folding. By manipulating one molecule at a time with either an atomic force microscope (AFM) or optical tweezers, scientists have been able to investigate aspects the folding mechanism of proteins that were previously inaccessible to experimental investigation, such as distances from the folded and unfolded states to their corresponding transitions states or anisotropy of a protein’s Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_18, © Springer Science+Business Media, LLC 2011

255

256

Cecconi et al.

energy landscape (1–4). The vast majority of these studies have been carried out by stretching engineered polymeric proteins between a gold substrate and an AFM silicon nitride tip (5). Using this method, scientists have applied AFM to the study of the unfolding processes of a large variety of proteins (6, 7). Optical tweezers instead have had much more limited use. The microscopic beads used in optical tweezers experiments are in fact not suited to manipulate molecules whose structures span only a few nanometers, such as most globular proteins and many of the polymeric proteins used in AFM studies; at such short distances, the surfaces of the tethering beads interact and severely compromise the measurements. For these reasons, optical tweezers have long been restricted to the characterization of the mechanical properties of proteins such as titin, whose native structures are micrometers long (8). In this chapter, we present a method that extends the use of optical tweezers to study the unfolding and refolding trajectories of individual globular proteins (9). This method relies on the use of molecular handles, ~500 bp DNA molecules, to connect the protein to polystyrene beads and minimize the interactions between the tethering surfaces of the bead (see Fig.  1). One end of each DNA molecule is covalently

Laser

digoxigenin

DNA

S S

protein S biotin

S

Polystyrene bead

pipette

Fig.  1. Experimental setup (not to scale). A single protein is manipulated between two micrometer-sized polystyrene beads by means of DNA molecular handles. One handle binds to a bead (held in the optical trap) through a digoxigenin– antibody interaction; the other handle binds to a bead (held by suction to the end of a pipette) through a streptavidine– biotin interaction. Each handle is covalently bound to the protein through a disulfide bond (see Fig. 2). During the experiment, the protein tethered between the beads is stretched and relaxed by moving the pipette relative to the optical trap by means of a piezoelectric actuator. (Adapted from Fig. 3 of ref. 13 with kind permission from IOS Press).

DNA Molecular Handles for Single-Molecule Protein-Folding Studies by Optical Tweezers

257

attached to a cysteine residue of the protein through a disulfide bond, while the other end is bound to a bead through either streptavidin–biotin or digoxigenin–antibody interactions. During the experiment, the protein is stretched and relaxed by moving the tethered beads relative to each other by means of piezoelectric actuators. We chose DNA molecules as our molecular handles because they are easy to synthesize in the laboratory by polymerase chain reaction (PCR) and because their mechanical ­properties have been extensively characterized by optical tweezers (10, 11), making it possible to distinguish their contributions easily in the recorded traces. In this chapter, we detail the protocols to generate DNA– protein chimeras and manipulate them with optical tweezers. We primarily consider DNA molecules made of ~500 bp because they have been the handles of choice for most of our experiments; handles of different lengths, however, can be used (9). The method described in this chapter represents a unique approach to single-molecule manipulation studies and presents a number of unique capabilities. (1) Individual globular proteins are directly manipulated with no need for the engineering of polymeric proteins, such as AFM experiments require. (2) The DNA handles are attached to cysteine residues that can be located anywhere on the surface of the protein, allowing a multitude of different pulling geometries to be used to study the anisotropy of the molecule’s energy landscape. (3) The attachment of the protein– DNA chimeras to the beads is achieved through specific bonds (biotin–streptavidin or digoxigenin–antibody), ensuring reproducibility between different experiments and easy interpretation of the data. (4) Unwanted multiple attachments between the tethering surfaces can be easily recognized. The response of DNA to force is known to be characterized by an overstretching transition at approximately 65 pN (11). If two or more molecules are caught between the two beads and pulled in parallel, the DNA molecules share the load and overstretch at higher forces. (5) The low spring constant of the optical trap and the force resolution of the instrument (RMS force noise of ~0.5 pN) allow the refolding process – as well as fluctuations between different molecular conformations – to be monitored directly. This method has been used to manipulate RNase H (12, 13) and T4 lysozyme (14) and should be applicable to any protein in which two exposed cysteine residues can be engineered. The DNA handles do not seem to affect the stability and folding mechanism of the tethered proteins, as shown by circular dichroism studies (9). Moreover, RNase H bound to two 558 bp molecular handles retains its enzymatic activity, proving that the overall three-dimensional structure of the tethered protein is conserved (12). This method presents some drawbacks as the components required to generate the DNA–protein chimeras are quite ­expensive and the range of forces that can be applied is limited to

258

Cecconi et al.

between 0 and 64 pN; at higher forces, the DNA overstretches and the digoxigenin/antibodies interaction becomes very labile, making interpretation of the data difficult. It should be noted, however, that the probability of observing a protein unfold depends not only on the applied force but also on the loading rate (r) used to pull on the molecule; r (pN/nm) equals pulling speed (nm/s) times spring constant (pN/nm). Owing to the low spring constant of an optical trap, the loading rates used in optical tweezer experiments are quite low and thus even proteins that are mechanically quite resistant can be observed to unfold below 64 pN of force.

2. Materials 2.1. Generation of Protein Variants

1. QuikChange site-directed mutagenesis kit (Stratagene). 2. High copy plasmid under a T7 promoter, such as pAED4 (from the Paul Matsudaira lab, The Whitehead Institute, Cambridge, MA, USA) or pET27 (Novagen). 3. Expression strain of Escherichia coli cells (BL21(DE3)plysS) (Promega). 4. Dithiothreitol (DTT) or Tris(2-carboxyethyl)phosphine (TCEP). 5. Column equilibration buffer: 0.1 M sodium phosphate buffer, pH 5.5. 6. Column equilibration buffer: 0.1 M sodium phosphate buffer, pH 7.0. 7. Disposable 5 mL polypropylene columns (Pierce). 8. Sodium azide (NaN3). 9. Sephadex G-25 coarse resin (Sigma). 10. DTPD solution: a 10  mM 2,2¢-dithiodipyridine solution is prepared by first dissolving 22.22 mg of DTDP in 1.5 mL of acetonitrile and then adding pH 5.5 column equilibration buffer to a final volume of 10 mL. Aliquots of 22.22 mg of DTDP dissolved in 1.5  mL of acetonitrile can be prepared and stored at −80 °C for months. 11. pH 5.5/G-25 columns: three disposable 5 mL polypropylene columns are cast with Sephadex G-25 coarse resin previously swelled overnight (O/N) in 0.2% NaN3 at room temperature (RT). The gel is allowed to settle in the column for at least 30 min at RT and it is then equilibrated with five column volumes of pH 5.5 column equilibration buffer, flown through via gravity. Immediately before being used for buffer exchange (i.e., loaded with protein), these columns must be spun “dry” at 1,000 × g for 1.5 min to remove excess buffer (see Note 1).

DNA Molecular Handles for Single-Molecule Protein-Folding Studies by Optical Tweezers

259

12. pH 7.0/G-25 columns: three disposable 5 mL polypropylene columns are cast with Sephadex G-25 coarse resin previously swelled overnight (O/N) in 0.2% NaN3 at RT. The gel is allowed to settle in the column for at least 30 min at RT and it is then equilibrated with five column volumes of pH 7.0 column equilibration buffer, flown through via gravity. Immediately before being used for buffer exchange (i.e., loaded with protein), these columns must be spun “dry” at 1,000 × g for 1.5 min to remove excess buffer. 13. Ready Gel® precast polyacrylamide gels 4–20% (Bio-Rad). 2.2. Generation of DNA Molecular Handles

1. PTC-200 Peltier thermal cycler (MJ Research). 2. Taq DNA polymerase (Qiagen). 3. pGEMEX1 plasmid DNA (Promega). 4. DNA primers (Integrated DNA Technology). 5. HiSpeed Plasmid Maxi kit (Qiagen). 6. Maxi kit DTT column elution buffer: 15  mM NaPO4 and 3 mM DTT, pH 7.0. 7. Handle buffer: 15 mM NaPO4 and 3 mM DTT, pH 7.0. 8. 30 MWCO Microcon ultrafiltration cartridges (Millipore). 9. Micro Bio-Spin columns with Bio-Gel P-6 in Tris buffer (Bio-Rad). 10. Spin column buffer: 0.1 M NaPO4 and 1 mM EDTA, pH 8.0.

2.3. DNA–Protein Coupling

1. PAGE reagents: 30% acrylamide/bisacrylamide 29:1, Tris(hydroxymethyl)aminomethane hydrochloride, sodium dodecyl sulfate, ammonium persulfate, and N,N,N ¢,N ¢tetramethylethylenediamine. 2. Ready Gel® precast polyacrylamide gels 4–20% (Bio-Rad). 3. SYBR Green I nucleic acid gel stain (Molecular Probes). SYPRO Red protein gel stains (Molecular Probes). Typhoon 8600 (Molecular Dynamic). DNA silver stain kit (GE Healthcare). 4. AFM deposition buffer: 10 mM HEPES, 10 mM NaCl, and 2 mM MgCl2 (pH 7.5). 5. A Veeco Nanoscope III AFM was used by us. 6. Pointprobes, type NCH-100 AFM probes (Nanosensors). 7. Ruby mica sheets (Asheville–Schoonmaker).

2.4. Preparation of Anti-digoxigenin Antibody-Coated Beads

1. PBS (7.0): 0.14 M NaCl, 2.7 mM KCl, 61 mM K2HPO4, and 39 mM KH2PO4, pH adjusted to 7.0 with HCl. 2. PBS (7.4): 0.14 M NaCl, 2.7 mM KCl, 80.2 mM K2HPO4, and 20 mM KH2PO4, pH adjusted to 7.4 with HCl.

260

Cecconi et al.

3. Cross-linking buffer: 100  mM Na2HPO4, pH 8.5, and 100  mM NaCl (or other non-amine containing buffer pH 7–9). 4. Protein G-coated polystyrene particles, 3.18  mm diameter, 0.5% w/v, 5 mL (Gentaur Molecular Products). 5. Anti-Dig solution: 1  mg/mL anti-digoxigenin (anti-Dig) antibody solution prepared by dissolving 200  mg of sheep polyclonal anti-Dig antibody (Roche) in 200 mL of PBS (pH 7.4). 6. DMP solution: A 10  mM dimethyl pimelimidate (DMP) (Pierce) solution is prepared by dissolving 50 mg of DMP in 1 mL of cross-linking buffer. DMP must be used immediately after its preparation because of its instability. 7. Tris base solution: 2 M Tris. 2.5. Tethering of Protein–DNA Chimeras to Polystyrene Beads

1. Binding buffer:10  mM Tris, 250  mM NaCl, and 10  mM MgCl2, pH 7.0. 2. Streptavidin-coated, 2.10-mm beads (Spherotech).

3. Methods Each molecular handle is attached to a protein via the formation of a disulfide bond between a thiol group present at the end of the DNA molecule and a thiol group of a cysteine residue in the ­protein (Fig. 1). The thiol–thiol reaction is mediated by DTDP (15–17) (Fig. 2). Disulfide bonds can form spontaneously, especially at high pH and temperatures (18, 19). However, DTDP speeds up the reaction and allows the time course of the DNA–protein coupling to be monitored spectrophotometrically at 343 nm via the release of the leaving group pyridine-2-thione (9). DTDP can be used to activate thiol groups of either DNA or proteins; in our experiments, however, we activate proteins because in this way the success of the reaction can be assessed by mass spectroscopy and because the activated molecule can then be used to generate polymeric proteins if desired. In the following sections, we describe in detail protocols to generate DNA–protein chimeras and to use them in single-molecule optical tweezer experiments. 3.1. Generation and Activation of CysteineBearing Protein Mutants

The following is the method to activate monomeric proteins for coupling with DNA, and the variant for coupling protein oligomers. Some proteins spontaneously react to form polymeric ­proteins without DTDP if stored in the absence of a reducing agent (18). However, in our experience, the speed of the reaction is greatly accelerated and more likely to continue to completion when DTDP is used.

DNA Molecular Handles for Single-Molecule Protein-Folding Studies by Optical Tweezers

261

Pyridine-2-thione 2

a

+

SH

HS

Cysteine-modified protein

2

N

S S

S

N N

N

2,2'-Dithiodipyridine disulfide

S S

S

S

N

Thiol-pyridine-activated protein

b DNA

2

SH

+ N

S S

2 S

N

S S

S S

N

S S

Protein-DNA chimera

Fig. 2. Schematic of the reactions used to attach DNA molecules covalently to proteins. (a) A cysteine-bearing protein is first activated with DTDP. (b) After cleaning up the reaction product, it is then allowed to react with DNA molecules bearing a thiol group at one end. The kinetics of both the protein thiol–pyridine activation and the protein–DNA coupling can be followed spectrophotometrically at 343 nm via the release of the leaving group pyridine-2-thione. (Adapted with kind permission of Springer Science + Business Media from ref. 9).

3.1.1. Generation of Monomeric Proteins

1. A protein variant bearing only two cysteine residues exposed on the surface of the protein (see Note 2) is generated through site-directed mutagenesis. Other cysteines naturally occurring in the protein are substituted with structurally neutral residues, such as alanine. 2. The cysteine-bearing variant is purified according to the protocol used for its cysteine-free variant with the only difference being that 1 mM DTT is added to all purification solutions (see Note 3). Depending on the inherent stability of the protein, the purified protein can be kept soluble at 4 °C for shortterm storage, or at −80°C after either lyophilization or liquid nitrogen freezing for long-term storage. 3. Before DTDP activation, the purified cysteine-bearing protein in buffer at pH 7.0 is allowed to react with 10–30 molar excess of DTT or 10 mM final TCEP (see Note 3) for ~1 h at RT to ensure full reduction of the thiol groups. 4. A volume of 150–200 mL (300 mL max) of reduced protein is carefully loaded onto one pH 5.5/G-25 column placed in a 15 mL conical tube and spun down at 1,000 × g for 1.5 min; during loading, attention should be paid to avoid disrupting the resin bed. To prevent the formation of protein oligomers after removal of the reducing agent, 150–200 mL of 10 mM DTDP is placed at the bottom of the 15 mL Falcon tube prior to centrifugation.

262

Cecconi et al.

5. The initial reaction of DTDP with a new protein variant should be monitored spectroscopically at 343 nm for production of the leaving group pyridine-2-thione (see Note 4 and Fig. 2). The activation of the cysteine residues is usually complete in a few minutes as seen by the time course of the release of the leaving group pyridine-2-thione during the reaction (9), although less accessible cysteines might take longer. Even rapid reactions that appeared to be complete in a matter of minutes were typically permitted to continue O/N at RT to ensure complete activation of the protein by DTDP. Some research groups, however, may elect to accelerate sample preparation for very reactive protein variants by allowing the reaction to proceed only for 2–4 h. 6. The thiol–pyridine-activated protein is isolated from the excess of unreacted DTDP by sequentially spinning the reaction solution through two pH 5.5/G-25 columns at 1,000 × g for 1.5  min. The protein is now activated and ready to be attached to DNA handles. 7. The activated protein can be stored at 4 °C for several days (depending on the inherent stability of the protein). 8. The success of the thiol–pyridine activation can be assessed through mass spectroscopy (9). This is particularly recommended for the initial reactions with new protein substrates to ensure complete activation by DTDP. 3.1.2. Synthesis of Polymeric Proteins

1. Reducing agents were removed from nonactivated protein by sequential spins through two pH 7.0/G-25 columns at 1,000 × g for 1.5 min (see step 3 above and Note 3). Because polymers are desired in this case, the nonactivated protein does not need to be reduced with DTT or TCEP as was done prior to DTDP activation. 2. Thiol–pyridine-activated protein (from step 8 above) in pH 5.5 column equilibration buffer is mixed with nonactivated protein at a molar ratio of 1:2 (see Fig. 3a). Typically, the final concentration of the activated protein is ~25 mM. The reaction is allowed to proceed O/N at RT. 3. The following day, more thiol–pyridine-activated protein (typically ¼ of the amount used in the first reaction) is added to the polymerization solution and allowed to again react O/N at RT. This second step aims to cap the ends of the polymers with activated monomers, which can then react with molecular DNA handles bearing thiol groups. 4. The degree of polymerization reached can be assessed using a SDS–PAGE gradient gel (4–20%) (see Fig. 3b, c).

263

DNA Molecular Handles for Single-Molecule Protein-Folding Studies by Optical Tweezers

a

S N

S S

S S

N

+

HS

SH

Thiol-pyridine-activated Cysteine-modified protein protein

b

N N

S S

-S--S-

-S--S-

n

S S

N

Poly-protein

c

a

b

Fig. 3. Synthesis of polymeric proteins. (a) Thiol–pyridine-activated proteins are mixed with thiol-modified proteins to generate tandem repeats of the molecule (adapted from ref. 9, with kind permission of Springer Science+Business Media). (b) SDS–PAGE analysis of an RNase H*Q4C/V155C polymerization reaction: unreacted protein (lane 1) and the product of the reaction (lane 2 ). As expected, the product of the reaction is a population of polymers of different lengths. The gel was stained with SYPRO Red, analyzed with a Typhoon 8600, and is shown in gray scale. (c) Intensity profile of the gel along the dark line in lane 2. Each peak of the profile corresponds to a different polymerization product, starting with the monomer from the left. The longest detectable polymers for this reaction were made of 24 monomers.

3.2. Preparing DNA Molecular Handles

1. Handles of 558 bp were generated by PCR. One handle was synthesized using the primers 5¢-thiol-GCT-ACC-GTA-ATTGAG-ACC-AC and 5¢-biotin-CAA-AAA-ACC-CCT-CAAGAC-CC. The other handle was synthesized using the same 5¢-thiol primer together with 5¢-digoxigenin-CAA-AAAACC-CCT-CAA-GAC-CC. PCR was performed using Taq polymerase and pGEMEX1 as a template. DTT at a final concentration of 20 mM was added to the PCR to keep the thiol groups of the primers reduced. Large amounts of handles (400–500  mg) were typically synthesized using as much as 9  mL of PCR reactions (PCR conditions are optimized for production yield and purity). The DNA handles were then purified using HiSpeed plasmid maxi kit columns and eluted with maxi kit DTT column elution buffer. The purified handles can be stored at −20 °C for months.

264

Cecconi et al.

2. The two types of handles are mixed in equal amounts to obtain ~1 mL of ~200 mg/mL digoxigenin/biotin (dig/bio) handles in handle buffer. 3. The dig/bio handle solution is then concentrated down to 50–60 mL with a 30-kDa MWCO Microcon centrifuge tube. 4. Reducing agents are removed from the handles by sequentially spinning them through three Micro Bio-Spin P6 columns equilibrated with the spin column buffer. The final handle concentration is usually 20–25 mM. 3.3. DNA–Protein Coupling 3.3.1. Monomeric Proteins

1. The dig/bio handles are mixed with thiol–pyridine-activated protein in pH 5.5 column elution buffer in a DNA handles: protein molar ratio of 4:1; typically, ~20 mM of DNA handles are reacted with ~5 mM of activated protein. The reaction is allowed to proceed O/N at RT (see Note 5). 2. The kinetics of the attachment of DNA molecules to proteins is slow, usually taking between 24 and 48 h to reach completion (9). The extent of the DNA–protein coupling can be assessed either by a 4% SDS–PAGE gel, prepared according to ref. 20, or by AFM (see Fig. 4). 3. To visualize the product of the protein–DNA reaction by AFM, molecular constructs were diluted to a final concentration of 2 nM in AFM deposition buffer. A volume of 20 mL of solution was deposited onto freshly cleaved mica and allowed to adsorb onto the surface for 1 min. The mica ­surface

Fig. 4. Characterization of the protein–DNA coupling reaction. (a) AFM image of the product of the reaction between 5­ 58 bp DNA molecules and RNase H*Q4C/V155C. Two protein–DNA chimeras made of two handles and one RNase H molecule are clearly visible. (b) 4% SDS–PAGE of the attachment of DNA handles to T4 lysozyme*T21C/K124C: DNA handle alone (lane 1), proteins bound to one handle (lane 2 ), handle dimers formed through the formation of disulfide bonds between the DNA thiol groups (lane 3 ), and protein bound to two handles (lane 4 ). The gel was stained with SYBR Green II, analyzed with a Typhoon 8600, and is shown in gray scale. (Adapted from ref. 9, with kind permission of Springer Science + Business Media).

DNA Molecular Handles for Single-Molecule Protein-Folding Studies by Optical Tweezers

265

Fig.  5. Isolation of long polymeric proteins through gel filtration. A 4–20% gradient SDS–PAGE gel of polymers of RNase H*Q4C/V155C (lane 1) and of different gel filtration fractions, starting from the latest (lane 2 ) and finishing with the slowest (lane 6 ). Gel filtration was performed using a size-exclusion column TosoHaas G2000 SWXL, equilibrated with 300  mM NaCl and 25  mM NaPO4, pH 7.4, and run at 0.75  mL/min. The gel was stained with SYPRO Red and analyzed with a Typhoon 8600, and is shown in gray scale. Each fraction was collected manually and they were each approximately 1 mL in volume.

was then gently washed with doubly distilled water and dried with a stream of nitrogen. The sample was imaged in air in tapping mode. 3.3.2. Polymeric Proteins

1. After the polymerization reaction, the longer polymers can be isolated from smaller protein oligomers by gel filtration (see Fig. 5). This step, although not strictly necessary, ensures that only long polymers are used for single molecules studies, which provides more data per single pull of the polymeric sample. 2. The polymeric proteins are then coupled to DNA handles following the same protocol used with monomers. For this reaction, the concentration of polymeric protein ends can be estimated from the overall absorbance of the protein sample and from the intensity of the different bands in polyacrilamide gels. Depending on the relative sizes of the DNA handles and proteins being used, the success of the reaction can be assessed either by SDS–PAGE (9) or simply by manipulating the molecules using optical tweezers (see Subheading 3.4).

3.4. Preparation of Anti-digoxigenin Antibody-Coated Beads

1. Protein G-coated polystyrene beads are spun down on a bench centrifuge at 1,000 × g for 5 min and then resuspended in 1 mL of cross-linking buffer.

266

Cecconi et al.

2. A volume of 60 mL of anti-Dig solution and 30 mL of DMP solution are added to the resuspended protein G-coated beads and the reaction is tumbled at RT for 60 min. The beads are then spun down at 1,000 × g for 5 min, resuspended in 1 mL of Tris base solution, and vortexed for 2 h at RT at minimal speed to quench the reaction. Notice that the Tris base quenching is a recommended but not necessary step. The beads are then diluted threefold in PBS (7.0), spun down at 1,000 × g for 15  min, and resuspended in PBS (7.0) three times to isolate the beads from the unreacted antibodies and DMP molecules. The beads are now ready to be used in laser tweezer experiments and can be kept at 4 °C for months. 3.5. Tethering of Protein–DNA Chimeras to Polystyrene Beads

1. Protein–DNA chimeras are allowed to react with anti-Dig beads in binding buffer for about 15 min at RT. The volume of a typical reaction was 10  mL total, with 8  mL of buffer, 1  mL of beads, and 1  mL of diluted DNA–protein chimera (3–20 pM final protein concentration in the 10 mL reaction). At the end of the reaction, the DNA–protein chimeras will be attached to the beads via the dig-labeled handle, while the other bio-labeled handle remains free and available for binding to streptavidin (see Fig. 6). 2. A streptavidin bead was suctioned onto the micropipette tip in the optical tweezer chamber. 3. The anti-Dig beads – now bearing protein–DNA chimeras – are further diluted into binding buffer (~1  mL) and then flowed into the optical tweezer fluid chamber until one bead is caught. The micropipette streptavidin bead is brought into close proximity to the anti-Dig chimera bead to allow the DNA–biotin moiety to react with streptavidin. Once the biotin/streptavidin bond is formed, the protein–DNA chimera is tethered between the two differently derivatized beads.

3.6. Manipulation of Protein–DNA Chimeras with Optical Tweezers 3.6.1. Monomeric Proteins

Once tethered between two beads, a protein–DNA chimera can be stretched and relaxed multiple times by moving the pipette relative to the optical trap. In our setup, the applied force is determined by measuring the change in light momentum of the beams leaving the optical trap, while the extension of the molecule is determined by means of a “light lever system” (21). Under tension, the unfolding of a protein is associated with a large change in its extension, as the molecule goes from a compact native state to an elongated unfolded state; on the contrary, the refolding process causes a sharp compaction of the protein. These sudden changes in molecular end-to-end distance give rise to sharp transitions in the force vs. extension traces. As an example, Fig. 7a shows a force–extension curve obtained by manipulating an RNase H molecule. By analyzing the size of

DNA Molecular Handles for Single-Molecule Protein-Folding Studies by Optical Tweezers

a

267

b

10 -15 min, RT

Fig. 6. Schematic of the experimental procedure used to tether single proteins in ­optical tweezer experiments. (a) Polystyrene beads covered with antibodies against digoxigenin are first allowed to react with the DNA–protein chimeras for 15 min at RT. (b) The DNA– protein chimera beads are then flown into the fluid chamber of the optical tweezer instrument and caught in the optical trap. A second bead covered with streptavidin was previously attached to a micropipette by suction. This streptavidin bead is then brought close to the optical trap bead to facilitate its binding to the DNA biotin moiety on the free end of the DNA–protein chimera.

the transitions in the recorded traces, it is possible to calculate the number of amino acids involved in the unfolding or refolding processes. Due to the low spring constant of the optical trap (usually 0.1  pN/nm or less), fluctuations between different molecular conformations can be monitored in real time. Figure 7b shows an extension vs. time trace obtained by keeping an RNase H molecule at a constant force using the force-feedback mode of the instrument; under these experimental conditions, the protein is observed to fluctuate between its intermediate and unfolded states.

268

b

25

Force (pN)

20

N

15 10 5 0

I N

U

50 nm

I

2955 2950

U

Extension (nm)

a

Cecconi et al.

2945 2940 2935 2930 634

Extension (nm)

636

638

640

642

644

646

648

Time (sec)

Fig. 7. Mechanical manipulation of single protein by optical tweezers. (a) Force–extension curve obtained by stretching and relaxing an RNase H*Q4C/V155C molecule. The protein unfolds at about 17 pN (native state N → U unfolded state) and refolds into an intermediate state (I ) at about 5 pN (U → I ). The intermediate state refolds into the native structure at lower forces ( I → N). (b) Extension vs. time trace showing an RNase H*Q4C/V155C protein hopping between its unfolded and intermediate states when holding the protein at a constant force of 5.5 pN. The extension values are the distances between the centers of the tethering beads.

35

Force (pN)

30 25 20 15 10 50 nm

5

Extension (nm) Fig. 8. Mechanical manipulation of a polymeric protein. Force–extension cycles obtained by stretching and relaxing a polymer of RNase H*Q4C/V155C multiple times. (Reprinted with kind permission of Springer Science + Business Media from ref. 9).

3.6.2. Polymeric Proteins

When a polymeric protein is stretched and relaxed, the subsequent unfolding and refolding of the individual domains give rise to characteristic saw-tooth patterns in the force–extension traces (Fig. 8). Each tooth corresponds to the unfolding or refolding of an individual monomer within the polymeric protein–DNA chimera. The number of teeth in the force–extension traces will vary because of the heterogeneity in length of the polymeric proteins generated through the method described above; typically, however, between four and ten teeth were observed in our curves.

DNA Molecular Handles for Single-Molecule Protein-Folding Studies by Optical Tweezers

269

Pulling on polymers presents the advantage of providing information on the mechanical properties of several domains at the same time, and thus it is easier to collect a statistically relevant set of data. The interpretation of the results, however, is usually more difficult than with monomeric proteins.

4. Notes 1. We use pH 5.5 for protein activation by DTDP to keep the reaction conditions stringent for activation of cysteine residues. However, at higher pH, the rate of this reaction and the rate of protein polymerization increases, and this faster kinetics could be useful with protein variants that react more slowly than those we investigated. Similar buffer exchange columns at pH 7.0 will be used to prepare other reaction components later in the protocols; they will be referred to as “pH 7.0/G-25 columns.” 2. It is technically possible to construct protein variants for optical tweezer experiments in which the structure of the protein is unknown or the engineered cysteines are not on the exposed protein surface. For instance, denaturants can be used to expose buried cysteines transiently and permit DNA handle attachment, but the disrupted structures of these DNA– protein chimeras will prevent any meaningful data from being obtained. 3. DTT is preferred over b-mercaptoethanol during protein purification because it is a more efficient reducing agent. TCEP, a very potent reducing agent, can also be used. We recommend TCEP for storing concentrated, pure proteins and for ensuring complete reduction of cysteines prior to activation with DTDP, but it is too expensive to justify its use in protein purification buffers. 4. The activation rate of a particular protein variant with DTDP varies depending on the accessibility of the cysteine, necessitating spectroscopic monitoring during the initial reaction. We also recommend monitoring the ratio of the 343 and 405 nm absorbance values rather 343 nm alone to eliminate any potential shifts in the baseline due to either instrument drift or protein precipitation. Some proteins did exhibit minor precipitation that caused an anomalous baseline shift; this precipitation appeared to be variant specific and potentially related to the final concentration of acetonitrile (used to dissolve DTDP) in the reaction. Proteins that are more sensitive to the presence of organic solvents could be accommodated by changing the volumetric ratios of the DTDP and the ­protein

270

Cecconi et al.

stock solutions, while preserving the molar ratios, to decrease the final concentration of acetonitrile in the reaction. 5. Only 50% of the protein–DNA chimeras synthesized through this method bear the correct combination of handles: that is, one handle labeled with biotin and the other with digoxigenin. The rest of the chimeras carry identically labeled handles. However, because such chimeras cannot be tethered between two differently derivatized polystyrene beads, these molecules are not functional in optical tweezer experiments. The use of these mixed populations of DNA–protein chimeras – rather than a pure population of sample containing one digoxigeninand one biotin-labeled DNA handle – also does not appear to impact the optical tweezer experiments negatively: identical data were obtained from mixed sample populations and those engineered to contain a pure population (9). References 1. Junker, J. P., Ziegler, F., Rief, M. (2009) Ligand-Dependent Equilibrium Fluctuations of Single Calmodulin Molecules Science 323, 633–637. 2. Oberhauser, A. F., Carrion-Vazquez, M. (2008) Mechanical biochemistry of proteins one molecule at a time Journal of Biological Chemistry 283, 6617–6621. 3. Borgia, A., Williams, P. M., Clarke, J. (2008) Single-molecule studies of protein folding Annual Review of Biochemistry 77, 101–125. 4. Garcia-Manyes, S., Brujic, J., Badilla, C. L., Fernandez, J. M. (2007) Force-clamp spectroscopy of single-protein monomers reveals the individual unfolding and folding pathways of I27 and ubiquitin Biophysical Journal 93, 2436–2446. 5. Fowler, S. B., Best, R. B., Toca Herrera, J. L., Rutherford, T. J., Steward, A., Paci, E., Karplus, M., Clarke, J. (2002) Mechanical Unfolding of a Titin Ig Domain: Structure of Unfolding Intermediate Revealed by Combining AFM, Molecular Dynamics Simulations, NMR and Protein Engineering J Mol Biol 322, 841–849. 6. Forman, J. R., Clarke, J. (2007) Mechanical unfolding of proteins: insights into biology, structure and folding Current Opinion in Structural Biology 17, 58–66. 7. Bustamante, C., Chemla, Y. R., Forde, N. R., Izhaky, D. (2004) Mechanical processes in biochemistry Annu Rev Biochem 73, 705–748. 8. Kellermayer, M. S., Smith, S. B., Granzier, H.  L., Bustamante, C. (1997) Foldingunfolding transitions in single titin molecules ­characterized with laser tweezers Science 276, 1112–1116.

9. Cecconi, C., Shank, E. A., Dahlquist, F. W., Marqusee, S., Bustamante, C. (2008) ProteinDNA chimeras for single molecule mechanical folding studies with the optical tweezers European Biophysics Journal with Biophysics Letters 37, 729–738. 10. Wang, M. D., Yin, H., Landick, R., Gelles, J., Block, S. M. (1997) Stretching DNA with optical tweezers Biophysical Journal 72, 1335–1346. 11. Smith, S. B., Cui, Y., Bustamante, C. (1996) Overstretching B-DNA: the elastic response of individual double-stranded and single-stranded DNA molecules Science 271, 795–799. 12. Cecconi, C., Shank, E. A., Bustamante, C., Marqusee, S. (2005) Direct observation of the three-state folding of a single protein molecule Science 309, 2057–2060. 13. Cecconi, C., Shank, E. A., Marqusee, S., Bustamante, C. Studying protein folding with laser tweezers. In: Broglia RA, Serrano L, Tiana G, eds. Proceedings of the International School Enrico Fermi - Course CLXV: IOS Press; 2006:145–160. 14. Shank, E. A., Cecconi, C., Dill, J. W., Marqusee, S., Bustamante, C. (2010) The folding cooperativity of a protein is controlled by its chain topology Nature 465, 637–640. 15. Riener, C. K., Kada, G., Gruber, H. J. (2002) Quick measurement of protein sulfhydryls with Ellman’s reagent and with 4,4¢-­dithiodipyridine Anal Bioanal Chem 373, 266–276. 16. Pedersen, A. O., Jacobsen, J. (1980) Reactivity of the thiol group in human and bovine albumin at pH 3–9, as measured by exchange with 2,2’-dithiodipyridine Eur J Biochem 106, 291–295.

DNA Molecular Handles for Single-Molecule Protein-Folding Studies by Optical Tweezers 17. Grassetti, D. R., Murray, J. F., Jr. (1967) Determination of sulfhydryl groups with 2,2’or 4,4’-dithiodipyridine Arch Biochem Biophys 119, 41–49. 18. Dietz, H., Berkemeier, F., Bertz, M., Rief, M. (2006) Anisotropic deformation response of single protein molecules Proceedings of the National Academy of Sciences of the United States of America 103, 12724–12728. 19. Dietz, H., Bertz, M., Schlierf, M., Berkemeier, F., Bornschlogl, T., Junker, J. P., Rief, M.

271

(2006) Cysteine engineering of polyproteins for single-molecule force spectroscopy Nature Protocols 1, 80–84. 20. Sambrook, J., Fritsch, E. F., Maniatis, T. Molecular Cloning: A Laboratory Manual. Second Edition ed: Cold Spring Harbor Laboratory Press; 1989. 21. Smith, S. B., Cui, Y., Bustamante, C. (2003) Optical-trap force transducer that operates by direct measurement of light momentum Methods Enzymol 361, 134–162.

Chapter 19 Optimal Practices for Surface-Tethered Single Molecule Total Internal Reflection Fluorescence Resonance Energy Transfer Analysis Matt V. Fagerburg and Sanford H. Leuba Abstract Single molecule fluorescence microscopy can be used to follow the mechanics of molecular biology processes in real time. However, many factors, from flow cell preparation to improper data analysis can negatively impact single molecule fluorescence resonance energy transfer (smFRET) experiments. Here, we describe some best practices for ensuring that smFRET data are of the highest quality. In addition to instrumentation, we describe sample preparation and data analysis. Key words: Single pair fluorescence resonance energy transfer, Total internal reflection fluorescence microscope, Single molecule fluorescence, Alternating laser excitation

1. Introduction There is great interest in using single molecule approaches in biology (1). Single molecule fluorescence techniques offer the promise of interrogating biological processes at the molecular level, with nanometer resolution. However, largely due to the sensitivity of the technique, great care must be exercised in such studies in order to minimize the occurrence of experimental artifacts. The construction of an appropriate microscope for fluorescence resonance energy transfer (FRET) data collection has been described in great detail in several sources (2–7); this work focuses on some of the lesser discussed practical details of FRET data collection that will ensure the best possible data quality. Poor surface preparation, sample photobleaching, and optical field alignment all present obstacles to sound data interpretation. Here, we

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_19, © Springer Science+Business Media, LLC 2011

273

274

Fagerburg and Leuba

describe some of the techniques that our laboratory has adopted to maintain the integrity of our surface-tethered samples, and maximize the quality of collected data. Although our laboratory has worked almost exclusively with a prism-based total internal reflection fluorescence (TIRF) microscope (8), the methods presented here apply equally well to objective-based TIRF systems.

2. Materials 2.1. Flow Chamber Construction and Preparation

1. Glass slides (Fisher Scientific) or fused silica slides (G. Finkenbeiner) (see Note 1). 2. Glass coverslips. 3. 1 mm diamond-core drill bits (McMaster Carr) (see Note 2). 4. Small butane torch (McMaster Carr) or Bunsen burner. 5. Parafilm (Fisher Scientific). 6. Heat block, heated to ~100°C. 7. Silicone tubing (0.8 mm ID, 3 mm OD; BioRad).

2.2. Slide Cleaning

1. Ultrasonic bath (Branson model 2510 or similar). 2. Liqui-nox detergent (by Alconox, or similar). 3. Acetone. 4. Ethanol. 5. KOH stock solution: 10 M KOH, to be diluted to 1 M for cleaning. 6. Slide staining chamber (Fisher 08-815 or equivalent).

2.3. Flow Cell Surface Preparation

1. T50 buffer: 10 mM Tris–HCl, pH 7.5, 50 mM NaCl. Store at room temperature.

2.3.1. bBSA Slide Treatment

2. bBSA solution: 1 mg/ml biotin-labeled BSA (Sigma-Aldrich) in T50 buffer. 3. NeutrAvidin solution: 0.2 mg/ml NeutrAvidin (Pierce) suspended in T50 buffer.

2.3.2. PEG Slide Treatment

1. Flame-treated slides and coverslips. 2. Silanizing solution: 5% acetic acid, 1% (3-Aminopropyl)triethoxysilane (APTES, Sigma-Aldrich) (see Note 3), in methanol. 3. Carbonate buffer (see Note 4): 2 parts of 100 mM Na2CO3, 9 parts of 100  mM NaHCO3 (check with pH indicator/ meter that this buffer is pH 9) 4. Methoxy poly(ethylene glycol) succinimidyl valerate (mPEGSVA), MW 5,000 (Laysan Bio, Inc.)

Optimal Practices for Surface-Tethered Single Molecule

275

5. Biotin-poly(ethylene glycol) succinimidyl valerate (biotinPEG-SVA), MW 5,000 (Laysan Bio, Inc.) 6. PEG-treatment solution: 11% mPEG; 0.4% Bio-PEG (w/v) biotin-PEG-SVA in carbonate buffer (see Note 5). 7. Nitrogen gas. 8. Humidity chamber (see Note 6). 2.4. Considerations for Data Collection and Analysis

1. T50 buffer: 10 mM Tris–HCl, pH 7.5, 50 mM NaCl. Store at room temperature. 2. d-glucose solution: 20% (w/v) d-glucose (Sigma) in H2O.

2.4.1. Preparation and Use of Imaging Buffer

3. beta-Mercaptoethanol (BME).

2.4.2. Overview of Data Analysis

1. TransFluoSpheres: NeutrAvidin-coated fluorescent microspheres (Molecular Probes).

4. 100× gloxy solution: dissolve 10  mg of glucose oxidase (Sigma) in 80 ml of T50 buffer, then add 20 ml of 20 mg/ml catalase solution (Roche) in T50 buffer (see Note 7). Pipette up and down to mix completely. Centrifuge at low speed for 1 min. Pour off the supernatant and store at 4°C for up to 3 months.

2. smTIRF microscope with dual-channel fluorescence emission optics.

3. Methods 3.1. Flow Chamber Construction and Preparation

The details of design of a suitable flow chamber for carrying out FRET experiments ultimately depend on the specific research interests of a laboratory and the FRET system being used (e.g., geometry of the microscope stage, availability of an automated flow system, etc.). It is important to consider the type of slide material to be used in the construction of the flow chamber. FRET experiments on a TIRF microscope are commonly carried out in one of two ways. In prism-based TIRF-FRET, a prism is optically coupled to the flow cell and is used to introduce the excitation laser beam. In contrast, objective-based TIRF-FRET uses the microscope objective to introduce the laser excitation into the sample. Each technique has certain limitations and implications for the practical details of flow cell construction. Objective-based TIRF-FRET is, perhaps, the easier of the two techniques to use on a regular basis. Once the system optics have been configured, the excitation area is well-defined and remains the same from experiment to experiment. Since the prism is not used, there are no components on top of the flow cell to impede access, and, thus, no constraints on flow cell hole placement.

276

Fagerburg and Leuba

An additional attraction of objective-based TIRF-FRET is that there is no need of special materials in the construction of the flow cell. Standard glass slides and coverslips are suitable for the construction of flow cells used in an objective-based system. In contrast, a prism-based system requires that a small prism be physically mounted on the surface of the flow cell. This kind of setup requires that holes for the flow chamber be placed to accommodate the prism footprint as well as the footprint of any interface hardware such as that used in an automated flow system or temperature regulation. 1. Drill holes in the slides to define the entry and exit ports of the flow cell (see Note 8). Clean the slides as described in Subheading  3.2 (or begin with slides that have been previously cleaned and stored under distilled H2O). Blot the slides on one edge to remove excess H2O. Burn both surfaces of the slides with a clean flame to eliminate debris. CAUTION! Standard glass slides (used in objective-based TIRF systems) cannot sustain the stresses created by this heating step and should only be briefly exposed to the flame. Fused silica is much more resilient, and should be held in the flame for approximately 30 s on each side. Coverslips should be passed briefly through the flame as well (they will warp and/or shatter if heated too long). After the slides and coverslips have thoroughly cooled, they are ready for either subsequent polyethylene glycol (PEG) treatment (see Subheading  3.3), or the assembly of the flow cell can be completed as described in the following remaining steps: 2. Cut a channel (see Note 8) out of the Parafilm, and stack this between a slide and a coverslip (Fig. 1). Using the coverslip as a guide, cut away excess Parafilm. 3. Fuse this sandwich assembly by placing it (coverslip-down) atop a heat block heated to 100°C for ~10  s or until the Parafilm layer has become transparent. Remove the flow cell from the heat block, and press the coverslip into the warmed wax using a blunt instrument such as a capped pen. Be careful not to run the blunt instrument over the part of the flow cell that will be imaged, as this can cause scratches and smudges that will negatively impact data collection. Repeat this heating step once more. 4. Solutions can be introduced into such a flow chamber by first aspirating the solution into a micropipette tip (of the 10–200 ml variety). Next, a small (~5 mm) length of sterilized silicon tubing is placed over the end of the pipette tip, and this is used to make an airtight seal over one of the holes in the flowcell. The solution is introduced by slowly pressing

Optimal Practices for Surface-Tethered Single Molecule

277

a

b

Fig. 1. Flow cell construction. (a) The flow cell is constructed from a predrilled glass slide (fused silica is necessary for prism-based TIRF), pre-cut Parafilm channel and a glass coverslip. The assembly is fused together by heating until the Parafilm melts slightly. (b) For prism-based TIRF, multiple channels should be planned so that prism placement is not impeded. Orienting channels along the long edge of the slide ensures that a large region of the channel can be excited/observed without blocking the solution injection ports with the prism.

down the pipette plunger, while a Kimwipe absorbs displaced fluid at the other flow cell port (see Note 9). 3.2. Slide Cleaning

The following procedure is appropriate for cleaning both glass and fused silica slides before assembling them into flow cells. When reusing fused silica slides from previously constructed flow cells, first, microwave the used flow cells in water for several min, and then use a razor blade to remove coverslips and spacer material, being careful not to scratch the slide. 1. Load slides into slide staining chamber (see Note 10), and fill each with a 20% Liqui-nox solution. Sonicate for 60 min. 2. Pour off detergent solution, and rinse slides under a flow of distilled H2O until suds no longer form. Fill the chambers with distilled H2O and sonicate for 5 min. 3. Pour off H2O and fill chambers with acetone. Sonicate for 15 min. 4. Pour off acetone, rinse slides under distilled H2O, then fill chambers with distilled H2O and sonicate for 5 min.

278

Fagerburg and Leuba

5. Pour off H2O and fill chambers with 1 M KOH. Sonicate for 15 min. Do not let slides sit in KOH for prolonged periods because it will slowly etch them. 6. Pour off and save KOH solution. Rinse slides under distilled H2O, then fill chambers with ethanol and sonicate for 15 min. 7. Rinse slides with distilled H2O, and repeat step 5 using the KOH solution saved from step 6. 8. Pour off KOH solution (do not save this time). Rinse slides with distilled H2O, fill chambers with distilled H2O and sonicate for 15 min. At the end of this cycle, replace the distilled H2O in the slide staining chambers with fresh H2O. Store slides in H2O until they are to be used. 3.3. Flow Cell Surface Preparation

Proper flow cell preparation is important because reagents can interact with flow cell surfaces in ways that can compromise the sample, lead to poor signal collection or both. In choosing a proper surface preparation technique, one must weigh the amount of time, materials and expertise that it requires against the need and appropriateness for the experiment. All surface-bound FRET experiments require a method for tethering the sample to one surface of the flow cell. Although this can be achieved by covalently linking the sample to the flow cell surface (e.g., using properly functionalized silanes), it is typically easier to use ligand-binding techniques instead such as avidinbiotin or Ni-HexaHistidine. A more sophisticated surface-tethering technique employs nanoscale lipid vesicles, whose formation and stoichiometry are carefully controlled so that each individual vesicle contains a single labeled molecule. This technique requires some delicate manipulations, and because the vesicles are selectively porous to different reagents, it may pose difficulties for changing buffer conditions within them. Our laboratory typically uses either a functionalized PEG preparation in conjunction with amino-silanes (9), or the less sophisticated technique of using biotinylated BSA. This latter method is less labor-intensive, but is inferior for experiments involving proteins. However, it yields excellent results for systems composed of labeled DNA, and it is fine for microscope field alignment (see Subheading 3.4) or pilot studies to determine the appropriate substrate dilutions for optimal signal density.

3.3.1. bBSA Slide Treatment

This method works well if the sample consists only of DNA molecules biotinylated at one end. Proteins will likely stick to the bBSA-treated surfaces; thus, for protein work, it is recommended to use the PEG treatment described in Subheading 3.3.2. 1. Starting with a flow cell prepared according to the description above, inject enough bBSA solution to fill the chamber (typically ~50 ml). Let sit for 5 min.

Optimal Practices for Surface-Tethered Single Molecule

279

2. Flush out unbound bBSA with two sequential injections of 100 ml each of T50 buffer. 3. Inject 100 ml of NeutrAvidin solution. Let sit for 5 min. 4. Wash out unbound NeutrAvidin with two sequential injections of 100 ml each of T50 buffer. The flow cell is now ready for use. 3.3.2. PEG Slide Treatment

Unlike the bBSA procedure described in Subheading 3.3.1, this protocol does not begin with an assembled flow cell; instead, start with cleaned slides and coverslips, flame-treated as described in Subheading 3.1, step 1. 1. Place the slides and coverslips into a slide staining dish; cover them with silanizing solution (solution must fully cover the slides/coverslips). Let sit at room temperature for 10  min, sonicate for 1 min and let sit for another 10 min. 2. Rinse the slides with methanol, and dry them under a stream of nitrogen gas. Lay out the slides in a humidity chamber. 3. Deposit ~70 ml of the PEG-treatment solution onto a slide, and carefully cover with a coverslip. Try to ensure that the solution spreads out evenly between the slide and coverslip and does not initiate any air-bubble formation. If air bubbles do form, try to coax them out by maneuvering the coverslip. Once all slide/coverslips have been treated, close the humidity chamber and incubate overnight at room temperature. 4. Disassemble the slide/coverslip sandwiches and rinse them thoroughly with filtered, deionized H2O then dry thoroughly under a gentle stream of nitrogen. Handle the slides/coverslips by the edges and avoid touching their surfaces. 5. Proceed with construction of a flow cell as described in Subheading 3.1; DO NOT burn the slides/coverslips again with a flame (see Note 11).

3.4. Considerations for Data Collection and Analysis 3.4.1. Preparation and Use of Imaging Buffer

The imaging performance of all single molecule dyes are impacted to differing degrees by the photophysical processes of photobleaching and blinking due to metastable state transitions (10,  11). If not addressed, these processes severely limit the amount of time that meaningful data can be collected to a few seconds; furthermore, ubiquitous blinking artifacts will confound any attempt at analysis of dynamic events. Thus, it is advantageous to employ an imaging buffer that minimizes these effects during acquisition of experimental data. The imaging buffer should reduce the rate of chemical reactions that lead to inactivate dye molecules and also reduce the lifetime and/or entry rate of transitions to metastable “dark” states. The system that is most popular in the literature consists of a combination of glucose oxidase (GOD), catalase, and glucose as well as 1% beta-mercaptoethanol

280

Fagerburg and Leuba

(BME). The enzymes, using glucose as a substrate, remove reactive oxygen species via: b-d-Glucose + O2 + H2O → d-Glucono-1,5-Lactone + H2O2 (GOD)(12)



2 H2O2 → 2 H2O + O2 (catalase)

(13)

and BME has been shown to be an effective triplet-state quencher. Typically, our laboratory will prepare a flow cell as described in the previous section, incubate it with the sample construct, rinse with buffer, and then introduce scavenger (prepared in a 1× buffer appropriate to the experiment) just before imaging is performed. The reaction described above will slowly lower the pH of the buffer over time; thus, samples should be imaged in a timely fashion. We follow the recipe of the Ha laboratory (14): 1. The final imaging buffer should be based on the appropriate experimental buffer and contain: 1× gloxy 0.8% (w/v) d-glucose 1% (v/v) BME (see discussion below) For example, we usually prepare and inject imaging buffer just before imaging a sample. We typically use 100 ml, and so we might mix: 10 ml

10× T50 buffer

4 ml

20% (w/v) d-glucose

1 ml

100× Gloxy

1 ml

BME

84 ml

H2O

(see Note 12) 3.4.2. Overview of Data Analysis

The specific methods used for the analysis of FRET data depend on the details of the experiment being performed. Here we focus on the first, and most universal step in the data analysis, which consists of extracting individual dye fluorescence intensity trajectories from a series of images. The image analysis programs (available upon request) we use were a kind gift from the laboratory of Dr. T. Ha and were written using Image Development Laboratory (IDL), a software package popularly used by astronomers and members of the remote sensing community. Indeed, single-molecule FRET data resembles astronomical data in appearance, and, thus, many of the algorithms and techniques employed by the astronomical community are of utility. Typical FRET data will consist of a field of fluorescence intensity signals, appearing as diffraction-limited spots. Each half of the image field corresponds to the observed sample as imaged through two different fluorescence emission filters. Thus, one half of the

Optimal Practices for Surface-Tethered Single Molecule

281

image contains signals at the donor-dye emission wavelength, and the other half contains signals at the acceptor-dye emission wavelength (see Fig.  2a). Ideally, starting with a signal peak in one channel, one could find the “companion peak” (i.e., intensity of

Fig. 2. Analysis of simulated FRET data. (a) A single image frame of simulated FRET data; left side represents donorchannel signals, and the right side represents acceptor channel signals. (b) Mapping left channel of data in (a) to the right channel, using an appropriately defined mapping file. (c) Mapping left channel of data in (a) to the right channel, using an inappropriately defined mapping file. If the mapping file is old, and/or if the microscope has been disturbed or new filters or other optical components have been added, then mapping of the two channels will result in incorrect alignment of the two channels for automated analysis. The remedy is to make a new mapping file using fluorescent beads. (d) A binary mask of active acceptors generated from ALEX data. This mask is used to select for signal peaks that represent molecule pairs containing an active acceptor dye, as described in the text. (e) Sample FRET histograms produced from the data in (a). The upper histogram is made by finding all of the signal peaks in the data. It thus includes a large “zero-peak” representing molecules that consist only of a donor dye molecule. The lower histogram is generated using the ALEX-mask in (d), which eliminates these donor-only signals from the analysis. Removal of the zero-peak is particularly useful in experiments where low FRET states are expected.

282

Fagerburg and Leuba

the same molecule, but in the other fluorescence emission channel) by simply moving horizontally a set number of pixels. However, inherent optical distortions introduced by the different optical paths for each channel inevitably lead to a complication of this ideal relationship. In reality, both channels are typically rotated and offset with respect to each other, and may also exhibit barrel distortion or other spatial aberrations (for convenience, one channel can be considered to be “correct,” and the other can be considered to consist of distorted peak positions). These distortions must be corrected for in order for automated peak identification to be successful. One powerful solution to this problem is to use data from fluorescent microspheres to register the two image field halves via a mapping file (see Note 13). These spheres can be imaged in both the donor and acceptor channels simultaneously. Images of such fluorescent beads reveal how each imaging channel distorts the position of the bead location. It is then mathematically possible (though decidedly nontrivial!) to map one channel onto the other, and, thus, determine where a peak’s companion is in the other channel. IDL provides a pair of powerful functions, POLYWARP and POLY_2D, to perform this mapping. POLYWARP accepts as arguments an initial set of (x, y) coordinates, and a corresponding set of distorted coordinates. It then uses a least squares estimation to determine the coefficients of a polynomial transformation that best maps the initial coordinates onto the distorted ones. POLY_2D is the inverse operation of POLYWARP; given an initial set of points (e.g., donor peak locations) and a set of coefficients, it determines a new, transformed set of points (e.g., acceptor peak locations). In practice, our laboratory will take images of Transfluospheres, analyze these images with an IDL program that uses POLYWARP to construct a mapping file (consisting of the described polynomial coefficients) and then use this mapping file with subsequently acquired FRET data. A different IDL analysis program uses this mapping file and POLY_2D to map the donor channel onto the acceptor channel. Data from the two channels is then superimposed, and from this composite field, we identify our FRET signal peaks (see Fig. 2b, c) by simple thresholding. To minimize and correct for background produced by CCD thermal noise, we run our Andor iXon camera as cold as possible (−90°C with antifreeze cooling) and employ the following algorithm to determine the local background around a signal. First, the intensity of the candidate signal is integrated over a 5 × 5 pixel area centered on the signal peak (see Fig. 3). Next, the intensity over a 7 × 7 pixel area centered on the signal peak is integrated. The 5 × 5 integrated intensity is subtracted from the 7 × 7 integrated intensity, and the result is divided by 24 to give the average pixel intensity due to the local background. Finally, this value is

Optimal Practices for Surface-Tethered Single Molecule

283

Fig. 3. Background subtraction. A representative fluorescence signal is shown greatly magnified along with its surrounding pixels (image is inverted for clarity). Local background is calculated from the pixels surrounding the signal peak (in the region marked with slashes) as described.

multiplied by 25 and subtracted from the 5 × 5 integrated intensity to yield the background-corrected, integrated intensity of the signal peak. Ideally, this procedure should be carried out for each peak in each frame of a data series. 3.5. Guidelines for Alternating Laser Excitation

One method for reducing the large zero-peak in FRET data employs an additional laser to periodically and directly excite acceptor dye molecules. Such a scheme is referred to as alternating laser excitation (ALEX; (15)). Data taken using an ALEX setup contains information about which acceptor dye molecules are still viable, and this information can be used to determine whether a FRET signal exhibits legitimate low-FRET, or if one of the dye partners has photobleached. By eliminating these latter signals, the final distribution of FRET signals will reflect only signals from experimentally relevant (i.e., still active) dye pairs. Figure 2e demonstrates the effect that an ALEX system can have on the measured static distribution of signals. Perhaps of greater import, however, is the ability of such a system to distinguish biologically interesting events from photophysical acceptor-dye blinking in dynamic data. Implementing an ALEX system is fairly straightforward, but  several factors need to be considered in its construction: (1) appropriate optics (filters, shutters, and alignment), (2) method for periodic laser alternation, synchronized with camera, and (3) analysis programs that use the ALEX information to discriminate bona fide signals from signals in which the acceptor has been compromised. Addressing the optical challenges of an ALEX system entails choosing appropriate lasers and emission filters for the dyes

284

Fagerburg and Leuba

being  used. Our laboratory performs the majority of our work using the popular Cy3/Cy5 cyanine dye pair (16) and we have a corresponding pair of donor and acceptor-excitation lasers operating at 532  nm and 647  nm, respectively. In our optical path downstream from a 610 nm dichroic mirror, a 580/40 bandpass filter defines the bandwidth for donor spectral data, and a 660 nm long pass filter blocks the acceptor laser light while admitting acceptor signal. In order to successfully realize an ALEX system, the beam paths of both excitation lasers must be made to coincide prior to their excitation of the sample. The most straightforward way to accomplish this is to use an appropriate dichroic mirror to ­combine the beam paths. We use a Chroma Z532BCM to reflect (at 90°) the 532 nm laser while allowing the 647 nm light to pass through (Fig.  4). Beam dumps should be used to terminate and absorb

Fig.  4. ALEX system. A simplified schematic of our prism-based ALEX-TIRF system. Details of the basic FRET optical system can be found in the text and also in ref. (4). INSET: (1) control pulse train from the CCD’s “fire” port; (2) and (3) are pulse trains (created from (1) by digital logic circuitry) that control donor and acceptor-excitation lasers.

Optimal Practices for Surface-Tethered Single Molecule

285

any reflected laser light. A simple beam dump can be made by taping a stack of single-sided razor blades together. If the underlying TIRF-FRET system is of the prism-based variety, co-alignment of the two laser beams requires some consideration. In our experience, chromatic dispersion by the prism will lead to misalignment of the excitation regions if the two lasers are simply coaxially aligned. Thus, it is important to mount at least one of the lasers on a three-axis positioning stage. The excitation optics are then adjusted for one of the lasers (i.e., the one without positioning stages) and locked in place. The position of the second excitation region (observed through the microscope) is then adjusted using the positioning stages for the other laser. We have found that once the two-excitation regions have been superimposed, the system requires no additional adjustment over many months. ALEX data is taken by exciting the sample with the donorexcitation beam, while periodically interrupting this beam and illuminating with the acceptor-excitation beam. Thus, a practical method is required to achieve this periodic switching between the two lasers. Additionally, these excitation periods need to be synchronized with the period of time when the camera is taking an exposure. With a properly setup system, the resulting data will consist of a sequence of images, where every nth image is illuminated with the acceptor-excitation laser. Thus, it is crucial to be able to modulate the output of the two lasers using an electrical signal, and also to have some way of synchronizing this signal with the operation of the camera. Newer diode lasers usually have a port that accepts a TTL-level signal that allows for switching the laser output on and off at rates of a few kilohertz or faster. Lasers without such a control port can still be modulated by introducing fast, electronically controlled shutters such as the Uniblitz VMMD1 or similar. Most (if not all) scientific-grade CCD cameras can be configured to take an exposure when triggered by an external electrical signal. Additionally, cameras such as the Andor iXon will output a TTL-level pulse when an exposure is being made. In the former scenario, an external function generator (realized either in hardware, or produce using software such as LabView and a computer) is used to generate the pulse sequence that drives both the lasers and the camera. In the latter scenario, the camera “fire” signal can be conveniently used to drive the lasers. In either case, a method is needed to produce the correct train of pulses in order to realize direct acceptor excitation for every nth frame. In order for the camera to take a regular series of exposures, it should be driven with a simple square wave, whose period determines the rate at which data is logged, and whose duty cycle determines the length of each individual exposure. The pulse train that drives the donor-excitation laser should coincide with this camera-signal except that every nth pulse should be dropped.

286

Fagerburg and Leuba + 5V 24

17 1 16

+ 5V

15 14

16

74154

4

7

20

11

6

21

2

22

3

23

5

CLOCK_IN

74193

13 11 10 9 8 7 6 5

18

+ 5V 14 1

2

14

7

4 3

19

2 1

8

7404

N 15 14 13 12 11 10 9 8 7 6 5 CLOCK_IN 4 3 2 1

To Donor Laser

To Acceptor Laser

12

(to N)

Fig. 5. Pulse control circuit schematic. This circuit generates the pulse trains that control the donor and acceptor lasers (i.e., (2) and (3) in inset of Fig. 4). The inputs labeled “CLOCK_IN” are driven by the pulse train that drives the CCD camera (or equivalently, the pulse train output by the CCD camera).

Conversely, the signal driving the acceptor-excitation laser should be high only during these nth pulses. These signals are depicted in the inset in Fig. 4. Producing these signals is trivial if using a software-driven system, and not much more difficult if the system is realized in hardware. We tap the iXon CCD “fire” signal (corresponding to the pulse train labeled 1 in the inset in Fig. 4) and use it to drive a count-to-N-and-reset circuit built from standard 7400-series TTL logic chips. The signal on any one of this circuit’s outputs (i.e., from 0 to N) corresponds to the pulse train labeled 3; if this is combined with the original pulse train 1 using an XOR operation, the result is the pulse train labeled 2. This circuit is simple in its implementation (see Fig. 5), and allows us to set N from 0 (no ALEX) to 16. A more sophisticated circuit using a microprocessor (such as a PIC) can allow for a much greater range of values of N. Once robust ALEX data has been acquired, there are many ways to capitalize on the additional information provided by direct acceptor-excitation. Our laboratory has adopted a simple but effective method to select for active donor-acceptor pairs within the sea of fluorescence signals. For a brief explanation, we shall consider a series of N images in which the first one in the series contains the signals from directly excited acceptor dyes. As described above, we identify signal peaks in our data frames by combining both the donor-side and acceptor-side subimages into a composite field (while simultaneously correcting for distortion due to the optical system). When working with ALEX data, we

Optimal Practices for Surface-Tethered Single Molecule

287

use the acceptor-side subimage of the first frame of the series to construct a binary mask (Fig. 2d) that we overlay on subsequent data frame composite fields, ensuring that signal peaks in the composite field will be found only if they contain an active acceptor dye (see Fig. 2). To make this mask, we first threshold the first-frame acceptorside subimage, resulting in a binary-valued image of active acceptor dyes. This image essentially represents diffraction-limited spots; we further increase the diameter of these spots by application of the dilation transformation (17). The resulting image is the mask that we use on the remaining N − 1 data frames. To apply the mask to a given composite field, our analysis program performs a pixelby-pixel multiplication; signals that persist in the remaining processed composite field represent those that contain an active acceptor dye. Occasionally, a donor-only signal will fall close enough to a bona fide dye-pair signal that it is retained after masking; however, with conservative use of the dilation operation when constructing the mask, these spurious signals are quite rare. When using the background calculation described in the previous section on data analysis, it is important to dilate the mask signals enough to allow for proper background subtraction. Finally, extension of this masking technique to longer data runs is straightforward; the data series is simply broken into l × N sets (where l = (total_#_of_ frames)/N ) and analysis proceeds as described above.

4. Notes 1. A prism-based excitation system directs the excitation beam through the comparatively-thick glass slide of the flow chamber, and it is of paramount importance to use fused silica slides in a prism-based TIRF system. This is because even the highest quality standard glass slides are riddled with microscopic optical defects that will scatter the incident excitation beam, producing an unacceptable level of background and false signal artifacts for single molecule work. Fused silica slides are more expensive than standard glass, but can be cleaned and reused for multiple experiments. 2. Be sure to use diamond-core type bits, and not spiral-type; it is very easy to shatter the slides by drilling with the latter. 3. APTES should be stored in a desiccator, under vacuum. 4. The components of this buffer should be freshly prepared. 5. We typically prepare 70  ml of this PEG-treatment for each flow cell we construct. Instead, we now make a solution of 0.5 mg/ml Bio-PEG and store in −80°C. Care should be taken

288

Fagerburg and Leuba

in suspending the PEG in carbonate buffer in order to prevent introducing air ­bubbles (do not vortex). It is best to centrifuge the solution for at least 1 min to remove bubbles. 6. We use an empty micropipette tip box as a humidity chamber. Fill the bottom with ~2 cm of H2O. The tip rack inside supports the slides, and the chamber can be closed and easily transported. 7. Catalase will precipitate out of solution; we resuspend it by inverting the bottle a few times. It is normal for this resuspended catalase solution to appear cloudy. 8. It is advantageous to design your flow cell to contain several channels; multiple channels allow for several experiments to be performed using the same flow cell (while allowing for varying sample solutions or conditions), and reduce the number of slides that must be prepared for a given experiment. We use a flow cell design that utilizes three channels, each 2.5 cm long, for a total per-channel volume of ~30 ml (Fig. 1). It is possible to fit even more channels (albeit smaller ones) on one slide, but we have found that (at least for the prism-based setup that we use) the advantage of such a setup is offset by the limited amount of resulting observable area. 9. It is important to strive to avoid introducing air bubbles into the chamber during this procedure. In the event that one is introduced, often it can be “pushed out” by injecting a small amount of buffer into the opposite port. 10. One chamber will hold seven slides if they are loaded “on the diagonal.” 11. Unused flow cells can be stored in foil-wrapped 50 ml conical tubes, under vacuum at room temperature and used within 2 weeks. 12. A modification to the described imaging buffer that is gaining popularity in the literature replaces the BME (which can inhibit the activity of some proteins) with Trolox (18). Our experience with this vitamin-E derivative suggests that it is indeed a superior triplet-state quencher, producing noticeably less blinking than BME. Briefly, the task is to make a saturated aqueous solution of Trolox, which should result in ~2  mM solution. The actual concentration is quantified by UV absorption (extinction coefficient = 2,350 ± 100 M−1 cm−1 @290 nm (14)), and the solution should be stored at 4°C for no more 1 month. This solution becomes the basis for making the imaging buffer described above (replacing the H2O in the example). As long as the final concentration of Trolox is greater than 1 mM in the final imaging buffer, the benefits of the Trolox should be manifest. Finally, it is worth noting that

Optimal Practices for Surface-Tethered Single Molecule

289

there are likely many other antioxidants yet to be explored that can effectively suppress blinking. 13. We favor the 40 nm-diameter TransFluoSpheres (Molecular Probes) with excitation/emission at 488/605 nm.

Acknowledgments This work was supported by RO1 GM077872 (S.H.L.) from the National Institutes of Health. References 1. Leuba, S. H. and Zlatanova, J. (Eds.) (2001) Biology at the single-molecule level, Pergamon, Amsterdam. 2. Ha, T. (2001) Methods 25, 78–86. 3. Zheng, H., Goldner, L. S., and Leuba, S. H. (2007) Methods 41, 342–352. 4. Zheng, H. C., Tomschik, M., Zlatanova, J., and Leuba, S. H. (2005) in “Protein-Protein Interactions, A Molecular Cloning Manual, 2nd Ed.” (Golemis, E., and Adams, P., Eds.), pp. 429–444, Cold Spring Harbor Laboratory Press, Cold Spring Harbor. 5. Gell, C., Brockwell, D., and Smith, A. (2006) Handbook of Single Molecule Fluorescence Spectroscopy, Oxford University Press, Oxford, UK. 6. Roy, R., Hohng, S., and Ha, T. (2008) Nat Methods 5, 507–516. 7. Selvin, P. R. and Ha, T. (Eds.) (2008) Single molecule techniques: a laboratory manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 8. Axelrod, D., Burghardt, T. P., and Thompson, N. L. (1984) Annu Rev Biophys Bioeng 13, 247–268. 9. Hermanson, G., T. (1996) Bioconjugate techniques, Academic Press, San Diego.

10. Eggeling, C., Widengren, J., Brand, L., Schaffer, J., Felekyan, S., and Seidel, C. A. (2006) J Phys Chem A 110, 2979–2995. 11. Lakowicz, J. R. (2006) Principles of fluorescence spectroscopy. 3ed., Springer, New York. 12. Keilin, D., and Hartree, E. F. (1952) Biochem J 50, 331–341. 13. Reid, T. J., 3  rd, Murthy, M. R., Sicignano, A., Tanaka, N., Musick, W. D., and Rossmann, M. G. (1981) Proc Natl Acad Sci USA 78, 4767–4771. 14. Joo, C. and Ha, T. (2008) in “Single-molecule techniques” (Selvin, P. R. and Ha, T., Eds.), pp. 3–36, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 15. Kapanidis, A. N., Lee, N. K., Laurence, T. A., Doose, S., Margeat, E., and Weiss, S. (2004) Proc Natl Acad Sci USA 101, 8936–8941. 16. Mujumdar, R. B., Ernst, L. A., Mujumdar, S. R., Lewis, C. J., and Waggoner, A. S. (1993) Bioconjug Chem 4, 105–111. 17. Burger, W. and Burge, M. J. (2008) Digital image processing: an algorithmic introduction using Java, Springer, New York. 18. Rasnik, I., McKinney, S. A., and Ha, T. (2006) Nat Methods 3, 891–893.

Chapter 20 Engineering Mononucleosomes for Single-Pair FRET Experiments Wiepke J.A. Koopmans, Ruth Buning, and John van Noort Abstract In DNA nanotechnology, DNA is used as a structural material, rather than as an information carrier. The structural organization of the DNA itself determines accessibility to its underlying information content in vivo. Nucleosomes form the basic level of DNA compaction in eukaryotic nuclei. Nucleosomes sterically hinder enzymes that must bind the nucleosomal DNA, and hence play an important role in gene regulation. In order to understand how accessibility to nucleosomal DNA is regulated, it is necessary to resolve the molecular mechanisms underlying conformational changes in the nucleosome. Exploiting bottom-up control, we designed and constructed nucleosomes with fluorescent labels at strategically chosen locations to study nucleosome structure and dynamics in molecular detail with single-pair Fluorescence Resonance Energy Transfer (spFRET) microscopy. Using widefield total internal reflection fluorescence (TIRF) microscopy on immobilized molecules, we observed and quantified DNA breathing dynamics on individual nucleosomes. Alternatively, fluorescence microscopy on freely diffusing molecules in a confocal detection volume allows a fast characterization of nucleosome conformational distributions. Key words: Single-molecule fluorescence, Nucleosome, Reconstitution, Fluorescence resonance energy transfer

1. Introduction DNA nanotechnology uses DNA molecules as smart building blocks for construction at the nanoscale (1). It exploits the unique molecular recognition properties of DNA to direct the self-assembly of new nanostructures and devices with bottom-up control. A major goal of DNA nanotechnology is to use DNA as a scaffold to structure other molecules (e.g., proteins or electronic components), relying on its chemical stability, rigidity, and predictable structure (2). To date, DNA has been successfully used to create a variety of complex 2D and 3D architectures, ranging from cubes Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_20, © Springer Science+Business Media, LLC 2011

291

292

Koopmans, Buning, and van Noort

(3) and tetrahedrons (4) to elaborate world-maps (5). For this purpose, DNA is used as a structural material, rather than as an information carrier. At a higher level, it is the structural organization of the DNA itself that determines accessibility to its underlying information content in  vivo. The basic unit of DNA organization in eukaryotic nuclei is the nucleosome. A nucleosome core particle consists of 50  nm of DNA wrapped in nearly two turns around a histoneoctamer core (6). Arrays of nucleosomes can fold into fibers, which in turn can condense into higher-order structures. Since nucleosomes sterically hinder enzymes that bind the nucleosomal DNA, they play an important role in gene regulation. To understand the mechanisms underlying gene regulation, it is essential to resolve the structural and dynamic properties of nucleosomes in detail (7). We use a bottom-up approach similar to DNA nanotechnology to study DNA organization in the nucleosome using single-pair Fluorescence Resonance Energy Transfer (spFRET) microscopy (8, 9). We designed and chose individual DNA and histone components exactly such that DNA folding can be studied at any desired location in the nucleosome. Nucleosomes are assembled on a DNA template through a salt dialysis reconstitution with purified core histones (10). The fluorescently labeled DNA template contains a strong nucleosome positioning element (11), so that nucleosomes are exactly positioned at a specific location on the DNA. This level of control ensures that the fluorescent labels in the DNA are incorporated at the desired location in the nucleosome, resulting in efficient FRET. Further modifications of the DNA allow specific immobilization of the nucleosomes to a surface. The FRET efficiency of individual nucleosomes can be monitored in detail with single-molecule fluorescence microscopy. For these experiments, nucleosomes are diluted to single-molecule concentrations in optimized buffer conditions. Widefield total internal reflection fluorescence (TIRF) microscopy on immobilized molecules allows the monitoring of individual nucleosomes for tens of seconds to minutes, revealing their dynamic behavior in time (8, 9). Alternatively, fluorescence microscopy on freely diffusing molecules in a femtoliter confocal detection volume allows a fast characterization of nucleosome conformational distributions (12, 13).

2. Materials 2.1. DNA Preparation and Purification

1. Template DNA containing the 601 nucleosome positioning sequence. 2. 1× TE buffer: 10 mM Tris–HCl, 1 mM EDTA, pH 8. This can be prepared and stored at 50×.

Engineering Mononucleosomes for Single-Pair FRET Experiments

293

3. HPLC-purified, fluorescently labeled forward and reverse primers (IBA GmbH) dissolved at 50 mM in 1× TE. Store in aliquots at −20°C. 4. FastStart PCR kit (Roche) with thermostable hot start polymerase (5  U/ml), nucleotide mix (10  mM of each dNTP), and 10× reaction buffer. Store at −20°C. 5. Thin-walled PCR tubes (Eppendorf). 6. QIAquick PCR purification kit, or QIAquick Gel Extraction kit (Qiagen). 2.2. Mononucleosome Reconstitution

1. 5 M NaCl solution. 2. 5–20 mM recombinant histone octamers in 1× TE, 2 M NaCl, and 5 mM 2-mercaptoethanol; 2 mM mixed sequence competitor DNA (~147 bp) in 1× TE, 2 M NaCl, or 3. 3–10  mM micrococcal nuclease-digested nucleosome core particles. 4. Dialysis tubes: Slide-A-Lyzer MINI dialysis units (10  K MWCO, Pierce).

2.3. Polyacrylamide Gel Electrophoresis

1. Running buffer: 5× TB (450 mM Tris, 450 mM Boric Acid). Store at room temperature. 2. 40% acrylamide:bisacrylamide solution (29:1, Bio-Rad) (this is a neurotoxin when unpolymerized, so avoid exposure). Store at 4°C. 3. N,N,N,N ¢-Tetramethyl-ethylenediamine (TEMED, Bio-Rad). 4. Ammonium persulfate solution (APS): 10% solution in water (see Note 1), stored in aliquots at −20°C. 5. Loading buffer (6×): 10 mM Tris–HCl pH 8, 60 mM EDTA, 60% glycerol.

2.4. Single-Molecule FRET Measurements

1. Microscope cover slips (24 × 60 mm # 1.5, Menzel).

2.4.1. Cover Slide Preparation

3. 96% AR grade ethanol (Biosolve).

2. RBS-50 detergent (Fluka). 4. 10× poly-d-lysine solution: poly-d-lysine (Sigma) dissolved at 0.1 mg/ml in water, and stored at 4°C. 5. NCO star PEGs (kind gift of Dr. Groll, RWTH Aachen, see Note 2). 6. Tetrahydrofuran (THF) (Sigma). 7. Biocytin solution: biocytin (Sigma) dissolved at 1 mg/ml in water, and stored at −20°C.

2.4.2. Single-Molecule Imaging

1. T50 buffer (see Note 3): 10 mM Tris–HCl (pH 8), 50 mM NaCl. 2. Oxygen scavenger system: Catalase (Fluka) stored at 4°C, glucose oxidase (Sigma) stored at −20°C.

294

Koopmans, Buning, and van Noort

3. b-d-glucose (Sigma). 4. Triplet quencher (trolox): tetramethylchroman-2-carboxylic acid (Sigma). 5. Sodium hydroxide solution (120 mM NaOH). 6. 100× bovine serum albumin (BSA) solution: 10  mg/ml in T50. 7. 10 mM Tris–HCl (pH 8). 8. Neutravidin solution: neutravidin (Pierce), dissolved at 5 mg/ml in water and 10% glycerol, stored in aliquots at −80°C. 9. CoverWell perfusion chamber gaskets (Invitrogen).

3. Methods To obtain exact positioning of a FRET pair at a specific location in the nucleosome, the use of a nucleosome positioning DNA sequence is necessary. We use the 601 sequence, which has a ­single dominant position for nucleosome formation (11). Strategic locations for labeling the DNA can be deduced from a highresolution crystal structure (14). Donor and acceptor fluorophores are then incorporated in the template DNA through a PCR with fluorescently labeled primers. Long, ~80 base pair (bp), primers are needed to label the DNA at internal positions in the nucleosome. Mononucleosomes are then assembled on the DNA with a salt dialysis reconstitution (10). It is important to mix DNA and histone proteins in the right stoichiometry: a too low octamer-toDNA ratio results in a subsaturated reconstitution, a too high ratio results in the formation of DNA-histone aggregates. The optimal stoichiometry is found by titrating the DNA with increasing amounts of histone octamers. To prevent formation of aggregates, mixed sequence competitor DNA can be included in the reaction. Nucleosomes will preferentially form on the labeled DNA containing the nucleosome positioning element. Excess histones will bind to the competitor DNA. Alternatively, an exchange reconstitution is used. Histone octamers are then supplied in the right stoichiometry in the form of micrococcal nuclease-digested nucleosome core particles (NCPs). In a reconstitution with a five to tenfold excess of NCPs over the fluorescently labeled DNA, nucleosomes will first form on the nucleosome positioning DNA. The reconstitution yield in both cases is checked with polyacrylamide gel electrophoresis (PAGE). The FRET efficiency is obtained from a bulk fluorescence emission spectrum.

Engineering Mononucleosomes for Single-Pair FRET Experiments

295

In order to image single nucleosomes, the concentration should be sufficiently low (10–100  pM). To prevent dilutiondriven dissociation (15), nucleosomes are diluted in a buffer containing BSA and 10–100  nM unlabeled nucleosomes. Imaging takes place in a buffer containing an oxygen scavenger system, to prevent photobleaching and blinking (see Note 4). Widefield TIRF microscopy is performed on nucleosomes that are immobilized to the microscope cover slide through biotin-neutravidin linkage. In this way, individual nucleosomes can be monitored for tens of seconds to minutes. The cover slide has to be treated with a special starPEG coating to prevent nonspecific adsorption of the sticky histone proteins to the glass (9). Alternatively, short bursts of fluorescence from freely diffusing molecules can be used to determine the conformational distribution. With this approach, immobilization to a coated cover slide is not necessary. 3.1. Choice of Label Positions and Primer Design

1. A good indication for where each base is located in the nucleosome core particle, can be derived from high-resolution nucleosome crystal structures (14). We mapped out the baseto-base distance for each possible combination of bases on opposite DNA strands, as shown in Fig. 1a. Bases separated a full nucleosomal turn (~80  bp) are spaced less than 2  nm apart, so that efficient FRET can occur when these locations are labeled with a FRET pair. Note that the average distance between the fluorophores is slightly different than depicted in Fig.  1a, since they are attached to the bases with a short carbon linker.

Fig. 1. Choice of labeling positions. (a) Map of the distance Rij between DNA base i and base j on the complementary strand for crystal structure 1kx5 (14). The bases come within 2 nm proximity when they are separated by a full nucleosomal turn (~80 bp). When facing outward from the nucleosome, these are ideal locations for placing a FRET pair (e.g., locations 1–3). The diagonal represents paired bases. (b) Strategic positions for placing a FRET pair indicating in the nucleosome crystal structure: at the nucleosome exit (1), or reporting on internal sites in the nucleosome (2, 3).

296

Koopmans, Buning, and van Noort

2. The forward strand is labeled with a donor fluorophore, the reverse strand with an acceptor. We chose Cy3B as donor and ATTO647N as acceptor, because of the photostability, high extinction coefficients, and high quantum yields of these dyes. The Förster radius is approximately 5.5 nm for this pair. 3. Fluorescent labeling is done with modified bases. Thymine bases are optimal targets for this, since they can easily be replaced with a modified dUTP. Cytosine bases are an alternative; they can be replaced with a modified dCTP. The dyes are attached to the base with a carbon linker, so that the fluorophore can rotate freely. Furthermore, it is important to only label bases that face outward from the nucleosome to prevent interactions of the fluorophores with the histone protein core. 4. The fluorescent labels are inserted in the DNA with labeled single-stranded DNA primers through a PCR reaction. A biotin modification can be applied at one of the 5¢-ends to allow for immobilization of the DNA to a neutravidin-coated surface. To label the DNA deep inside the nucleosome, long (>80 bp) primers are needed. It is important to place the label not too close (<5 bp) to the primer end, to prevent stalling of the polymerase reaction. We successfully used the following primers on the 601 nucleosome positioning sequence: forward primer: 5¢-TTGG CTGGAGAATC CCGGTGCCGA GGC­ CGCTCAA TTGGTCGTAG ACAGCTCTAG CACC­ GCTTAA ACGCACGTAC GCGCTG-3¢, reverse primer 5¢-biotin-TTGGACAGGA TGTATATATC TGACACGTGC CTGGAGACTA GGGAGTAATC CCCTTGGCGG TTA­ AAACGCG GGGGACAGC-3¢. Label positions are underlined. This set of primers places one label at the nucleosome exit, and the other internally at the nucleosome dyad axis, to monitor DNA unwrapping starting from one nucleosome end. If different label positions are chosen, the same primers can be used to monitor a variety of positions inside the nucleosome. 3.2. DNA Preparation and Purification

1. In a 0.5 ml PCR tube, mix 5 ml PCR buffer containing MgCl2, 1 ml solution of dNTPs, 1.5 ml fluorescently labeled forward primer, 1.5 ml fluorescently labeled reverse primer, 0.5 ml faststart Taq polymerase, 200 ng template DNA containing the 601 sequence, and water in a total volume of 50 ml. 2. Place the tube in a thermal cycler equipped with a heated lid, such as the Techgene (Techne, Cambridge, UK). 3. Amplify the DNA with 35 cycles of the following two-step PCR cycle: 30 s denaturation at 95°C, 1 min annealing and polymerization at 72°C. The first cycle is preceded by a 5 min initial denaturation step at 95°C. The last cycle is followed by

Engineering Mononucleosomes for Single-Pair FRET Experiments

297

a final 10 min polymerization step at 72°C. We use a two-step cycle because of the long, ~80 bp, primers involved. 4. Purify the DNA with a Qiagen PCR purification kit according to the instructions supplied in the manual. We observed that it is difficult to remove all free primer DNA in this way. Alternatively, for a higher degree of purity, analyze the PCR reaction on a 1% agarose gel, excise the desired DNA band, and purify with a Qiagen Gel Extraction kit according to the instructions supplied in the manual. 3.3. Mononucleosome Reconstitution 3.3.1. Titration with Stoichiometric DNA-toOctamer Ratios

1. Titrations are carried out by varying the DNA-to-octamer ratio, typically from 1:0.5, 1:1, 1:1.5, 1:2, to 1:2.5, or as high as necessary. We use a minimum reaction volume of 30 ml for the dialysis tubes. The initial salt concentration should be 2 M NaCl, adjusted with 5 M NaCl solution. 2. In a 1.5 ml centrifuge tube, mix 2 mg of labeled DNA, and 2  mg of unlabeled competitor DNA in 30  ml of 2  M NaCl buffered with 1× TE (pH 8). 3. Add the appropriate amount of histone octamers, and mix. 4. Incubate 30 min on ice. 5. Transfer the sample to a presoaked dialysis tube. Cap the tube and place the tube in floating device in a beaker containing the buffer for the first dialysis step (1× TE with 0.85 M NaCl). 6. Dialyze at 4°C against 500  ml of 1× TE containing 0.85, 0.65, 0.5, and finally 0 M NaCl for at least 60 min per step. Continuously stir the buffer with a magnetic stirrer at a low speed setting. 7. Recover the contents from the dialysis tube, and analyze with bulk fluorescence spectroscopy and 5% PAGE as described below. An example result is shown in Fig. 2a. The reconstituted material can be stored at 4°C for several weeks (see Note 5). The titration point with the highest reconstitution yield (typically 80%) is subsequently used for single-molecule experiments.

3.3.2. Nucleosome Exchange Reconstitution

1. Exchange reconstitutions are carried out by transferring nucleosome cores from micrococcal nuclease-digested nucleo­ some core particles to the fluorescently labeled DNA containing the 601 sequence. We use a minimum reaction volume of 30 ml for the dialysis tubes. It is important to completely dissociate the nucleosome core particles. Therefore, the initial salt concentration should be 2 M NaCl, adjusted with a 5 M NaCl solution. 2. In a 1.5 ml centrifuge tube, mix 2 mg of labeled DNA, and a fivefold excess of nucleosome core particles in 30 ml of 2 M NaCl buffered with 1× TE (pH 8).

298

Koopmans, Buning, and van Noort

a

1

2

3

4

5

b

1

2

N

I (a.u.)

c

D

2.0

Nucleosomes Free DNA

1.5 1.0 0.5 0.0 500

600

700

λ (nm) Fig. 2. Nucleosomes analyzed on 5% PAGE (acceptor fluorescence with excitation at 633 nm) and with bulk fluorescence spectroscopy. (a) Octamer titration reconstitution. Lane 1: free DNA, Lanes 2–5: DNA:octamer ratios ranging from 1:2 to 1:3.5. All lanes show a band of nucleosomes (N) which migrate slower through the gel than the DNA (D). Lanes 4 and 5 show fluorescence in the slots, originating from DNA-histone aggregates due to an excess of histone octamer proteins. (b) Exchange reconstitution. Lane 1: free DNA. Lane 2: nucleosome reconstitution. The reconstitution yield in this case is >95%. (c) Fluorescence emission spectra of free DNA and reconstituted nucleosomes. Excitation was at 515 nm, the emission was recorded from 535 to 735 nm. Reconstituted nucleosomes show efficient (>60%) FRET, indicated by the decrease in donor fluorescence around 560 nm, and the increase in acceptor fluorescence at 670 nm.

3. Incubate 30 min on ice. 4. Transfer the sample to a presoaked dialysis tube. Cap the tube and place the tube in floating device in a beaker containing the buffer for the first dialysis step. 5. Dialyze at 4°C against 500  ml of 1× TE containing 0.85, 0.65, 0.5, and finally 0 M NaCl (for at least 60 min per step). Continuously stir the buffer with a magnetic stirrer at a low speed setting. 6. Recover the contents from the dialysis tube, and analyze with bulk fluorescence spectroscopy and 5% PAGE as described below. The reconstituted material can be stored at 4°C for several weeks (see Note 5). The reconstitution yield is typically >90%. An example result is shown in Fig. 2b. 3.4. Polyacrylamide Gel Electrophoresis

1. We use a Amersham Bioscience Hoefer SE 400 vertical gel slab unit (14  cm wide, 14  cm high), with a custom made pump unit for buffer recirculation. Buffer recirculation is necessary to prevent depletion of the low ionic strength running buffer (0.2× TB). Low ionic strengths are needed to prevent dissociation of nucleosomes. 2. The glass plates should be scrubbed clean with a detergent and rinsed extensively with distilled water. Rinse with 70% ethanol and air-dry. 3. Prepare a 1.5  mm thick 5% polyacrylamide gel (29:1 acrylamide:bisacrylamide, 0.2× TB) by mixing 5 ml acrylamide,

Engineering Mononucleosomes for Single-Pair FRET Experiments

299

1.6 ml 5× TB, 33 ml water. Degas for 10 min and mix with 200  ml APS and 80  ml TEMED to initiate polymerization. Pour the gel, insert the comb, and allow to polymerize for 30 min. 4. Remove the comb and rinse the wells with 0.2× TB buffer. Add 0.2× TB running buffer to the upper and lower chambers of the gel unit. 5. Prerun the gel for at least 60 min at 4°C at 19 V/cm, while continuously recirculating the buffer. 6. Load 0.2–1 pmol of reconstituted nucleosome core particles in 6  ml 1× loading buffer. The use of dyes such as xylene cyanol and bromophenol blue is not recommended, since their autofluorescence may interfere with the fluorescence from the nucleosomes. To track the migration of DNA in the gel, load the dye in a separate lane of the gel. 7. Run the gel for 75 min at 4°C at 19 V/cm, while continuously recirculating the buffer. 8. Image the fluorescence with a gel imager such as the Typhoon 9400 (GE). To assess the reconstitution yield, excite the acceptor fluorophore (as shown in Fig. 2a, b) since this is a direct reporter of the amount of DNA in each band. To obtain an indication of the FRET efficiency, the donor and acceptor fluorescence should be recorded separately while exciting the donor fluorophore. If the reconstitution yield is known, the FRET efficiency can be further quantified with a bulk fluorescence emission spectrum, as shown in Fig. 2c. We use the ratioA method as described in detail by Clegg (16). 3.5. Single-Molecule FRET Measurements

1. Place the microscope slides in a rack. Use clean tweezers.

3.5.1. Microscope Slide Cleaning

3. Sonicate 60 min in ethanol. Rinse with water.

2. Sonicate 15 min in 1% RBS-50 at 90°C. Rinse with water. 4. Flame dry slides with a Bunsen burner to remove any remaining traces of organic impurities. Hold the slides with a reverse action tweezers, and gently move the slide through the flame. 5. Clean the slides with an UVO ozone cleaning device (Jelight) for at least 60 min to obtain hydrophilic slides without any fluorescent impurities.

3.5.2. Passivation and Functionalization

1. Cover the clean slides with 100 ml of 1× poly-d-lysine solution, and incubate for 1 min. This step is needed for aminofunctionalization of the slides. Rinse with water and blow dry in a nitrogen stream. 2. Dissolve six-arm NCO PEG stars (MW 12 kDa) in THF at a concentration of 20  mg/ml, and dilute in water to a con­ centration of 2  mg/ml. This will initiate a crosslinking

300

Koopmans, Buning, and van Noort

reaction between the PEGs. Add biocytin (Sigma) to a final concentration of 1 mg/ml to obtain sparse biotinylation. 3. Five minutes after mixing, sterilize the solution by filtration through a 0.22 mm syringe filter. 4. Layer the solution onto the amino-functionalized cover slide. Spincoat the fully covered slide for 45 s at 2,500 rpm. 5. Incubate the slides at room temperature overnight to complete the crosslinking reaction. Slides can be stored in the dark for up to 1 week. 3.5.3. Oxygen Scavenger System

A 1× oxygen scavenger system consists of 2.2 mg/ml (2,170 U/ml) catalase, 0.92  mg/ml (165  U/ml) glucose oxidase, 0.4–4% glucose, and 2 mM trolox. 1. Mix 11  ml catalase with 1  ml T50, filtrate with a 0.22  mm syringe filter and centrifuge for 15 min (at 4°C, 16,000 RCF) in a table-top centrifuge, such as Eppendorf 5415R. Take 200 ml of the supernatant. 2. Dissolve 92 mg glucose oxidase in 1 ml T50, filtrate with a 0.22  mm syringe filter and centrifuge for 15  min (at 4°C, 16,000 RCF) in a table-top centrifuge. Take 200  ml of the supernatant. 3. Mix the catalase and glucose oxidase to obtain a 50× stock solution, and keep on ice. This is more than sufficient for a day of measurements. Further storage is also possible (see Note 6). 4. Prepare a 40% w/v solution of b-d-glucose and sterile filtrate with a 0.22 mm syringe filter. Store at 4°C. 5. Dissolve 250 mg trolox in 1 ml 120 mM NaOH (by vortexing and sonicating) to obtain a 100 mM (50×) trolox stock solution. Sterile filtrate with a 0.02 mm syringe filter. Store in aliquots at −80°C.

3.5.4. Single-Pair FRET Microscopy with Widefield Microscopy

1. Assemble a flow channel by placing a CoverWell perfusion chamber gasket on a functionalized slide. 2. Fill the flow chamber with T50 buffer and hydrate the slide for 5 min. 3. Inject a neutravidin solution (0.01–0.1 mg/ml) and incubate for 5  min. Wash excess neutravidin away with two to three flow chamber volumes of T50. 4. Dilute the labeled, biotinylated nucleosomes to 10–100 pM in a buffer containing 0.1  mg/ml BSA, 10–100  nM unlabeled nucleosomes, and 1× oxygen scavenger (see Note 7 and Subheading  3.5.3). Inject the sample in the flow chamber and seal by covering the holes with a glass slide. Immobilization is virtually instantaneous. Record the fluorescence with a widefield microscope. Example data is shown in Fig. 3a–c.

Engineering Mononucleosomes for Single-Pair FRET Experiments

1.0 0.5 0.0

I (kHz)

0

2

4

6

8

10

0

2

4

6

8

10

t (s)

150 100 50 0 50 100 150

Donor Acceptor

0.0

0.2

0.4

0.6

t (s)

0.8

1.0

Frequency

0

100

50

0 0.0 0.2 0.4 0.6 0.8 1.0

E

e

250

Frequency

1000

b

d

c

2000

E

I (a.u.)

a

301

200 150 100 50 0 0.0 0.2 0.4 0.6 0.8 1.0

E

Fig. 3. Single-pair FRET microscopy on nucleosomes. (a) Fluorescence intensity time trace of donor and acceptor fluorescence of a single immobilized nucleosome. (b) Resulting FRET efficiency E as a function of time. Fluctuations in E are caused by breathing dynamics of the nucleosome. (c) FRET efficiency histogram of the time trace in (b). (d) Fluorescence intensity time trace of fluorescence originating from freely diffusing molecules. Bursts of fluorescence arise from the passage of a single molecule through the excitation volume. (e) Histogram of FRET efficiencies calculated from bursts as in (d). A high and a low FRET population can clearly be distinguished. These reflect different conformations of the nucleosome.

3.5.5. Single-Pair FRET Confocal Microscopy

1. Alternatively, image freely diffusing molecules using confocal microscopy. Immobilization to a functionalized cover slide is not needed in this case, so the steps described in Subheading 3.5.2 can be omitted. The observation time is limited to a few milliseconds per molecule. 2. Mount a clean cover slide on the microscope, and place a droplet (~50 ml) of 10–100 pM labeled nucleosomes in a buffer containing 0.1  mg/ml BSA, and 10–100  nM unlabeled nucleosomes. Focus 25 mm above the glass surface, and record the fluorescence. Example data is shown in Fig. 3d, e.

4. Notes 1. Unless stated otherwise, all solutions are prepared with 18.2  MW  cm MilliQ water. This standard is referred to as “water” in the text. 2. StarPEGs are not commercially available. With commercially available linear PEGs, we observed nonspecific adsorption

302

Koopmans, Buning, and van Noort

and dissociation of nucleosomes (9). It has been reported that a repeated PEGylation step may increase the specificity and reduce nonspecific adsorption of slides coated with linear PEGs, however (17). 3. Care must be taken to avoid autofluorescent contamination of buffers, samples and microscope slides. Work at a clean lab bench, using clean glassware, and always wear gloves. Buffers for single molecule imaging are made sterile by filtering through a 0.02 mm syringe filter (Whatman Anotop 25). This removes virtually all autofluorescent impurities. 4. We recommend using Alternating LASER Excitation (ALEX) (18) to sort molecules based on their label stoichiometry, and to discriminate residual bleaching and blinking events from real conformational transitions. 5. Store the nucleosomes at the highest concentration possible (preferably >100 nM) to prevent dilution-driven dissociation. The use of low protein-binding tubes is recommended to minimize sticking to the tube. 6. If desired, 50 ml aliquots of 50× glucose oxidase and catalase can be stored at −80°C for several weeks. Although it is recommended not to freeze the catalase, we did not observe a major loss in oxygen scavenging activity from these aliquots. 7. Buffers for single molecule imaging are degassed prior to use, to remove the oxygen already in solution. When using oxygen scavenger the sample can be imaged for 30 min; after that the pH will drop as a result of the production of gluconic acid as a byproduct of the oxygen scavenging reaction.

Acknowledgments We thank Andrew Routh (MRC Cambridge) for samples of micrococcal nuclease-digested nucleosome core particles and useful discussion, Alexander Brehm (University of Marburg) for histone octamer preparations, and Jürgen Groll (RWTH Aachen) for providing samples of the NCO-star PEG material and support with the coating procedure. This work is part of the research programme of the “Stichting voor Fundamenteel Onderzoek der materie (FOM),” which is financially supported by the “Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO).”

Engineering Mononucleosomes for Single-Pair FRET Experiments

303

References 1. Seeman, N. and Lukeman, P. (2005) Nucleic acid nanostructures: bottom-up control of geometry on the nanoscale. Reports On Progress In Physics 68, 237–270. 2. Seeman, N. (1998) DNA nanotechnology: Novel DNA constructions. Annual Review of Biophysics and Biomolecular Structure 27, 225–248. 3. Chen, J. and Seeman, N. (1991) Synthesis from DNA of a molecule with the connectivity of a cube. Nature 350, 631–633. 4. Goodman, R., Schaap, I., Tardin, C., Erben, C., Berry, R., Schmidt, C., and Turberfield, A. (2005) Rapid chiral assembly of rigid DNA building blocks for molecular nanofabrication. Science 310, 1661–1665. 5. Rothemund, P. (2006) Folding DNA to create nanoscale shapes and patterns. Nature 440, 297–302. 6. Luger, K., Mader, A., Richmond, R., Sargent, D., and Richmond, T. (1997) Crystal structure of the nucleosome core particle at 2.8 Å resolution. Nature 389, 251–260. 7. Luger, K. (2006) Dynamic nucleosomes. Chromosome Research 14, 5–16. 8. Koopmans, W. J. A., Brehm, A., Logie, C., Schmidt, T., and van Noort, J. (2007) Singlepair FRET microscopy reveals mononucleosome dynamics. J.Fluoresc. 17, 785–795. 9. Koopmans, W. J. A., Schmidt, T., and van Noort, J. (2008) Nucleosome Immobilization Strategies for Single-Pair FRET Microscopy. ChemPhysChem 9, 2002–2009. 10. Dyer, P., Edayathumangalam, R., White, C., Bao, Y., Chakravarthy, S., Muthurajan, U., and Luger, K. (2004) Reconstitution of nucleosome core particles from recombinant

11.

12.

13.

14.

15.

16. 17. 18.

histones and DNA. Chromatin and Chromatin Remodeling Enzymes, Pt A 375, 23–44. Lowary, P. and Widom, J. (1998) New DNA sequence rules for high affinity binding to histone octamer and sequence-directed nucleo­ some positioning. J.Mol.Biol. 276, 19–42. Gansen, A., Hauger, F., Toth, K., and Langowski, J. (2007) Single-pair fluorescence resonance energy transfer of nucleosomes in free diffusion: Optimizing stability and resolution of subpopulations. Anal.Biochem. 368, 193–204. Kelbauskas, L., Chan, N., Bash, R., Yodh, J., Woodbury, N., and Lohr, D. (2007) Sequence-dependent nucleosome structure and stability variations detected by Förster resonance energy transfer. Biochemistry 46, 2239–2248. Davey, C., Sargent, D., Luger, K., Maeder, A., and Richmond, T. (2002) Solvent mediated interactions in the structure of the nucleosome core particle at 1.9 Å resolution. J.Mol.Biol. 319, 1097–1113. Claudet, C., Angelov, D., Bouvet, P., Dimitrov, S., and Bednar, J. (2005) Histone octamer instability under single molecule experiment conditions. J.Biol.Chem. 280, 19958–19965. Clegg, R. (1992) Fluorescence resonance energy-transfer and nucleic-acids. Methods Enzymol. 211, 353–388. Roy, R., Hohng, S., and Ha, T. (2008) A practical guide to single-molecule FRET. Nature Methods 5, 507–516. Lee, N., Kapanidis, A., Wang, Y., Michalet, X., Mukhopadhyay, J., Ebright, R., and Weiss, S. (2005) Accurate FRET measurements within single diffusing biomolecules using alternatinglaser excitation. Biophys.J. 88, 2939–2953.

Chapter 21 Measuring DNA–Protein Binding Affinity on a Single Molecule Using Optical Tweezers Micah J. McCauley and Mark C. Williams Abstract DNA–protein interactions may be observed on single molecules with a variety of techniques. However, quantifying the binding affinity is difficult and often requires many (~100) individual events to characterize the interaction. We use a single l DNA molecule that provides a lattice of binding sites for proteins. Extending and relaxing the tethered molecule reversibly melts DNA, allowing it to be converted between double-stranded (ds) and single-stranded (ss) forms. By monitoring changes in the properties of the DNA as a function of added protein concentration and fitting to binding models, the DNA–protein interaction may be characterized and quantified. As an example, the high mobility group protein HMGB1(box A + B) is observed to stabilize dsDNA. Measuring the strength of this effect allows us to determine the equilibrium association constant for HMGB1(box A + B) binding to dsDNA. Key words: Single molecule, Optical tweezers, Force spectroscopy, DNA binding, DNA melting

1. Introduction Optical tweezers experiments represent a subset of single molecule experiments which study individual processes in detail (1–4). The optical tweezers utilized in the following experiments are composed of a pair of counter-propagating beams brought to a narrow, overlapping focus approximately a micron in diameter. These beams will hold a spherical bead of several microns in diameter, up to forces of several hundred picoNewtons (pN). In principle, the experiments described here can be done with most optical tweezers instruments, although a dual beam instrument is needed for higher force measurements. Streptavidin – biotin linkages allow nucleic acids to be attached to this bead. The following experiments will be performed using a single DNA molecule from

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_21, © Springer Science+Business Media, LLC 2011

305

306

McCauley and Williams

bacteriophage l, 48,500  base pairs in length, which provides multiple sites for protein binding. The opposite end of the molecule is attached to another bead, which is held on a micropipette by suction, as shown in Fig. 1a, b. Held between the two beads, the DNA molecule may be subjected to tension by moving the micropipette relative to the bead in the optical trap. This tension displaces the bead from the center of the optical trap. The displaced sphere, in turn, slightly deflects the trapping lasers. The laser deflection and the applied tension are directly proportional. Thus, after initial calibration by applying a known force, subsequent measurement of the displacement of the trapping beams allows the applied force to be determined for a given extension. Force-extension data for dsDNA show elastic responses, where the force increases monotonically as the molecule is extended (see Fig. 1c). DNA is well characterized by a polymer model of elasticity at these forces (5). At ~60 pN, however, the molecule abruptly lengthens, nearly doubling in length (6–8). During this force-induced DNA melting transition, base stacking and pairing are disrupted and dsDNA is converted into ssDNA. The extension across the plateau corresponds to the fraction melted (9–12). If the extension is reversed and the molecule is relaxed

c 120 250 mM Na+

100

100 mM Na+

b

Force (pN)

25 mM Na+

68 66

80

64

60

62 60

40

Force (pN)

a

58 20 0 0.28

56 54 0.32 0.36 0.40 0.44 0.48 Extension (nm per base pair)

Fig. 1. Optical tweezers experiments isolate and melt dsDNA. (a) Phage l DNA is tethered between two polystyrene beads (spheres) via a streptavidin/biotin linkage as detailed in the text. While one bead is fixed onto a micropipette the other is held in the optical trap (indicated by the white dot ). The isolated DNA molecule may be extended and relaxed within the flow cell under specific solution conditions, including varying protein concentrations. HMGB1(box A + B) is represented by the L-shaped symbols. (b) An image of the trap taken with a CCD camera shows that a stretched DNA molecule (not visible) pulls the bead from the optical trap, deflecting the trapping laser (white dot ). (c) Force-extension (solid symbols) and -relaxation data (open symbols) for DNA in 25 mM (triangles), 100 mM (circles ), and 250 mM Na+ (squares ). The scales of the graph are split to highlight the changes in the force; data to the right of the split are on the expanded force scale. As the extension is increased, the force rises until base stacking and pairing are disrupted and dsDNA is melted, and converted into ssDNA, which has a longer contour length. Lower counterion concentration destabilizes dsDNA by reducing the screening between charged backbones. The end of the melting transition and the strand separation region are omitted.

Measuring DNA–Protein Binding Affinity on a Single Molecule

307

before the strands separate, then the base pairing and stacking may reform, and the reaction is reversible. When this experiment is repeated in the presence of DNA-binding ligands or under changing solution conditions, the observed force-extension curve changes in response to the interaction of the ligand with dsDNA or ssDNA. For example, binding ligands that interact preferentially with dsDNA will increase the melting transition force, while ligands that interact with ssDNA will lower the melting transition force. Thus, as a single DNA molecule is stretched and relaxed the effects of varying solution conditions and binding ligands may be probed (13–15). Below we describe the procedure for labeling DNA, capturing a single DNA molecule between two polystyrene beads using a fluid flow cell in an optical tweezers instrument, and obtaining a DNA force-extension curve. We then describe in detail the specific example of an experiment to measure the DNA binding affinity of high mobility group protein HMGB1(box A + B), which increases the DNA melting force. The change in DNA melting force is used as a measure of protein binding to dsDNA, and the observed fractional binding is fit to a lattice binding model to obtain the equilibrium protein–DNA binding affinity for this protein. An analogous procedure can be applied to any protein or ligand that binds nonsequence specifically to DNA and alters the DNA force-extension curve.

2. Materials 2.1. Biotinylated Phage Lambda DNA

1. Phage l DNA (48,500 base pairs), 300 mg/ml (Roche, Basel, Switzerland) (see Note 1). 2. Biotin-14-dCTP and Biotin-14-dATP (0.4 mM) (Invitrogen, Carlsbad, CA). 3. dGTP and dTTP (100 mM) (Promega, Madison, WI). 4. Polymerase I Klenow fragment, exonuclease− and 10× buffer (Fermentas, Glen Burnie, MD) (see Note 2). 5. T4 DNA ligase and 5× rapid ligase buffer (Fermentas, Glen Burnie, MD) (see Note 3). 6. Buffer – 10  mM HEPES, pH 7.5, 100  mM Na+ (5  mM NaOH used to set the pH and 95 mM NaCl). 7. Buffer – 10 mM Tris–HCl, pH 8.0, 1 mM EDTA. 8. 24:1 Chloroform:Isoamyl alcohol, stored at 4°C. 9. Equilibrated phenol, pH 8.0, stored at 4°C. 10. 3.0 M sodium acetate, pH 4.5, stored at 4°C. 11. 100% ethanol, stored at −20°C.

308

McCauley and Williams

2.2. Flow Cell

1. Cover slips, #1, 22 × 30 mm (Fisher, Atlanta, GA). 2. Spacer, 22 × 30 × 3 mm, Plexiglas (custom made). 3. Tubing, various sizes, polyethylene (Cole-Parmer, Vernon Hills, IL) (Warner Instruments, Hamden CT). 4. 15 mL tubes, polypropylene (Fisher, Atlanta, GA). 5. Glass micropipette, 1.0 mm diameter tip (WPI, Sarasota, FL). 6. Polystyrene beads (5 mm diameter), streptavidin coated, 1% solids w/v (Bangs Labs, Fishers, IN).

2.3. Dual Beam Optical Tweezers

1. Pair of 200  mW, CW diode IR lasers, 830  nm wavelength (JDSU, Carlsbad, CA) (see Note 4). 2. Pair of high numerical aperture microscope objectives (Nikon USA, Melville, NY). 3. Lateral effect diode detectors (CVI Melles Griot, Carlsbad, CA). 4. Piezoelectric stage and electronic controller (Npoint, Madison, WI).

3. Methods 3.1. Labeling DNA

1. Heat 20 mL of phage l DNA to 65°C for 10 min to remove circular and linear DNA aggregates. 2. To the DNA add 12.5 mL of each of the Biotin-labeled bases and 0.5 mL of each of the unlabeled bases. Then add 1.0 mL of polymerase and 5.0 ml of 10× buffer for a total volume of 52 mL. Incubate at 37°C for 1 h. 3. Biotinylated DNA may be ligated by the addition of 1 mL ligase and 12 mL of 5× buffer. As mentioned elsewhere, this step is optional (see Note 3). If this step is performed, the polymerase should first be inactivated by heating to 65°C for 15 min. 4. Add 0.5× sample volume of equilibrated phenol and 0.5× volume of the chloroform:isoamyl alcohol mixture, vortex and centrifuge at 11,000 × g for 3 min. Carefully remove the top, aqueous layer (which should be about one-half of the sample volume) and transfer to a fresh tube (see Note 5). 5. Add 1.0× of the chloroform:isoamyl alcohol mixture and centrifuge at 11,000 × g for 3  min. Remove the top layer (again about one-half of the sample volume), and transfer to a fresh tube. 6. Add 0.1× sample volume of sodium acetate and 2.5× of ethanol. Labeled DNA should now be visible as a diffuse precipitate

Measuring DNA–Protein Binding Affinity on a Single Molecule

309

that floats in the liquid under gentle shaking. Store at −20°C for ~1 h (see Note 6). Spin at 11,000 × g for 15 min at 4°C. Aspirate the supernatant and allow the pellet to dry (see Note 7). 7. For samples to be used immediately, resuspend the pellet in 250 ml of HEPES buffer. Labeled DNA will be suitable for stretching experiments for up to a year. If the samples are to be frozen for long-term storage, resuspend in 250 mL of TE buffer and aliquot before freezing. Avoid freezing the labeled DNA more than once (see Note 1). 3.2. Flow Cell Construction

1. Various designs have been utilized for different experiments, as different criteria, including the objective working distance and trapping geometry, demand unique cell constructions. All designs incorporate a pair of clear optical windows, a spacer, and tubing to allow the sample flow. Our cells consist of windows glued to the spacer, with the micropipette glued directly into the cell’s center (see Note 8). 2. The cells are finished by gluing the sample tubing, one for each different type of sample to prevent mixing, and attaching the tubing to fixed 15 ml tubes for use as sample reservoirs (see Note 9). 3. The completed cell is mounted onto a piezo stage and between the objectives, separated by their working distance, leaving enough room for a water meniscus (most high numerical aperture objectives are water or oil immersion). Introducing buffer flow into the cell should complete the optical path (a collimated laser incident upon an objective should emerge from the trap fully collimated).

3.3. Trap Alignment and Characterization

1. The key components of our optical trap are the trapping lasers, microscope objectives, flow cell, detectors, and piezo stage. The following protocol will focus upon the essential alignments that should be checked daily and will provide the necessary controls for data collection. Further details of optical tweezer design and construction are discussed within several reviews (and the references therein) (3, 4, 16, 17). 2. The output of the trapping lasers must be checked for mode quality. The beam cross-sections should reveal a clear Gaussian intensity profile with optimum power to guarantee a uniform trap (see Note 4). A convenient way to monitor beam quality is to use a confocally arranged CCD camera for each beam. Such cameras will also provide an image of the focal plane, including the micropipette tip and the beads. 3. The counter-propagating beams should overlap along the full length of the trapping axis (see Note 10). This can be checked

310

McCauley and Williams

with a thin piece of paper or tissue on the input side of each objective. Within the trap, the overlap may be checked using the camera image (described above), and may be fine tuned with a coarse piezo or fine micrometer stage mounted to one of the objectives. 4. Beam displacements on the detectors may now be zeroed (see Note 11). 5. A solution of coated beads (diluted roughly 1,000:1 from the stock solution) should be introduced into the cell (see Note 12). When a single bead appears in the trap, it may be drawn upon the micropipette tip. 6. The held bead may be moved back and forth across the trap (it should remain upon the tip), using the piezo for precise movement. The displacement and power of each trapping laser should be recorded at each stage position (see Note 13). This procedure determines the distance the bead is displaced from the trap center for a given deflection of the laser. Known as the stiffness correction, it is applied as a correction to the length moved by the piezo stage during DNA extension measurements (see Note 14). 3.4. Isolating a Single Molecule and Collecting ForceExtension Data

1. While retaining the bead held to the micropipette, another individual bead should be caught in the trap. The cell should be rinsed of all other beads, using the experimental buffer. 2. A solution of labeled DNA, diluted ~4,000:1 from the stock solution into the experimental buffer may now be flown into the cell (see Note 12). The bead fixed to the micropipette should be placed behind the bead in the optical trap (downstream), with a separation of about one contour length (16.5 mm, in this instance). As a DNA molecule flows through the cell, one end will attach to each bead. Though the molecule cannot be seen directly, movement of the trapped bead due to movement of the micropipette indicates a successful catch. Figure 1a shows the trapping geometry and Fig. 1b is a CCD image of the trap. The cell should again be rinsed fully with the experimental buffer. 3. Recording the beam deflections for various stage extensions gives the data shown in Fig.  1c. Several cycles of extension and relaxation should be collected and compared to deduce instrument stability, particularly for the baseline force (see Note 15). Irregularities between different cycles may suggest nicked DNA, a weak biotin–streptavidin linkage, or multiple DNA tethered between the beads. These molecules and attached beads should be discarded. 4. The trap may be calibrated, using the midpoint of the melting plateau (see Note 16). The midpoint extension may be found

Measuring DNA–Protein Binding Affinity on a Single Molecule

311

by averaging the contour lengths of dsDNA and ssDNA. The  force may be averaged over a linear range around this extension. This midpoint force is well known for various ­solution conditions, and is shown in Fig. 1c (18). 3.5. Characterizing the Concentration Dependence

1. Force-extension data for each detector are added, calibrated, and corrected for beam power and the stiffness of the trap, as described above. The calibration and stiffness corrections are held constant over the remaining extension cycles. Varying concentrations of protein may now be flown into the cell (see Note 17). 2. Averaging over several stretching curves will verify the stability of the trap and the equilibrium nature of protein binding. Individual stretching and relaxation cycles for various concentrations of HMGB2(box A + B) are shown in Fig.  2a. Distortions in the data at low forces are due to proteininduced aggregation and protein-stabilized looping. 3. The averaged melting force, plotted as a function of protein concentration in Fig. 2b, may be fit to a binding model. The model of McGhee and von Hippel includes the effects of site exclusion, and states that the occupancy of a given binding site (g) should increase with the protein concentration (c);

(1 - g )

n



g = Knc

a 120

b 76

100

68

60 56

40 100 nM HMGB1(box A+B) 20

2 nM HMGB1(box A+B) 0 nM HMGB1(box A+B)

0 0.28

0.52

(1)

74

68 66

52 48 44

0.32 0.36 0.40 0.44 0.48 Extension (nm per base pair)

Force (pN)

60

.

70

Force (pN)

Force (pN)

64

n -1

72

72

80

(1 - g + g / n )

64 62 0.1

1 10 100 Protein Concentration (nM)

Fig. 2. HMGB1(box A + B) stabilizes dsDNA. (a) Extension (solid symbols) and relaxation data (open symbols) for DNA in the absence (open symbols), and presence of 2 mM (triangles) and 100 mM (squares) of HMGB1(box A + B). Doublestranded DNA is stabilized with the addition of protein. (b) The increase in the midpoint of the melting force is averaged over several cycles and plotted vs. protein concentration. The fit to a binding isotherm determines the equilibrium association binding constant (solid line), as discussed in the text.

312

McCauley and Williams

The fit to this data determines the equilibrium association binding constant (K), while the binding site size (n) is estimated from previous crystallography and NMR work (see Note 18) (19–21). 4. The occupancy is deduced from the averaged melting force ( Fm ), scaled against the melting force for DNA in the absence of protein ( Fmo ) and in saturating concentrations of protein ( Fms ) (see Note 19);

g=

Fm - Fmo . Fms - Fmo

(2)

5. Least squares fits to these equations will determine the equilibrium association binding constant for this binding mode. Figure 2b shows the result of a fit for HMGB2(box A + B), which gives Kn = 7.2 ± 1.7 × 108  M−1. This protein binds to and stabilizes dsDNA with a high affinity (20, 22).

4. Notes 1. Once thawed, phage l DNA should be stored in the 4°C refrigerator. The backbone of long DNA may be sheared as ice forms in solution, introducing nicks. Though the probability of shearing is low, repeated freezing and thawing must be avoided. Once thawed, DNA will have a shelf life of 6 months to 1 year. 2. The lack of exonuclease activity will prevent removal of the newly added biotinylated bases. 3. DNA ligase is used to repair any nicks introduced into the backbone by normal pipetting (long DNA is easily sheared by small micropipette tips). However, any subsequent pipetting could reintroduce breaks in the DNA. Therefore, this step is optional, as careful handling should be the main technique used to minimize nick formation. If the DNA appears to break or show excessive hysteresis during stretching experiments, the samples may be ligated and purified, though it is preferable to simply prepare a new sample. 4. Only the critical components of a dual beam, single trap are shown here. Further details, including additional components, instrument construction and variations upon the laser tweezers design are discussed in several reviews (2–4, 9, 16, 17). 5. These steps are a standard phenol extraction protocol. Alternatively, shorter lengths of DNA may be successfully separated from the reactants with various filter kits. However,

Measuring DNA–Protein Binding Affinity on a Single Molecule

313

the highest and purest yields for long (>15  kbp), labeled DNA are obtained through phenol extraction. 6. If the precipitate appears especially prominent and the centrifuge may be temperature controlled, this storage step may be omitted. 7. An optional rinsing step may be performed here to further rinse unincorporated nucleotides from the sample. A volume of ~100 mL (roughly the same as the amount of ethanol added in the previous step), of 70% ethanol and 30% water should be added to the pellet. Tilt the tube gently a few times, then spin the sample at 11,000 × g for 15 min at 4°C. Aspirate the supernatant and allow the pellet to dry. 8. A microscope is helpful for precise placement of the tip. 9. The tubing assembly lasts for a few weeks of experiment use. High concentrations of protein may clog the tip (preventing bead attachment), though excess DNA is usually the contaminant. The cells should be dried and may even be refrigerated between experiments to discourage bacterial growth. 10. The objectives are assumed to be underfilled (the beam diameter is less than the input aperture of the optics). 11. Polarization beam optics are used to steer each beam onto a unique detector (16). 12. The pressure used to deliver the flow may be of the order of 1.0 kPa and should not exceed 10 kPa, though this depends somewhat upon the design of the spacer. 13. The position piezo control and data collection are performed with custom software written in LabWindows CVI. 14. The correction is generally taken to be linear, though for high force work near the limit of the trap a quadratic term may be added. Furthermore, a power correction must be applied at these forces, to correct for power lost due to clipping of the beams. 15. Thermal fluctuations in the cell and the objectives will cause opposing deflections in each beam. Averaging the displacements of the two beams will cause these fluctuations to cancel, a key advantage of the dual beam geometry. Fluctuations due to the drifting of a single beam, however, will not cancel and will become most evident as a change in the zero force baseline. 16. A viscous drag force measurement may also may be used for an absolute calibration of the trap. 17. The total volume flown into the cell should be at least a few times the internal volume of the flow cell. If the flow into the chamber is fully laminar, then a volume comparable to the

314

McCauley and Williams

internal volume of the cell may be used. Greater volumes are required for protein diffusion to reach equilibrium. 18. This parameter may be determined directly from the fits for low values of the binding site size (n £ 3). However, values of c2 for fits including large values will be only weakly dependent upon the exact value of n, leading to large uncertainties. 19. Care must be taken to determine the maximum melting force. Adding enough protein will eventually form a filament for this protein that will show a very stiff elastic response. The protein concentration should be kept low enough to prevent this.

Acknowledgments This work was supported by NIH (GM75965) and NSF (MCB02381890). L. James Maher and Jeff Zimmerman are thanked for purified samples of HMGB2(box A + B). Additionally, the authors would like to thank Karin Musier-Forsyth and Penny J. Beuning for technical assistance and advice with the DNA labeling protocol. References 1. Ashkin, A., Schutze, K., Dziedzic, J. M., Euteneuer, U., and Schliwa, M. (1990) Force generation of organelle transport measured in vivo by an infrared laser trap, Nature 348, 346–348. 2. Bustamante, C., Bryant, Z., and Smith, S. B. (2003) Ten years of tension: single-molecule DNA mechanics, Nature 421, 423–427. 3. McCauley, M. J., and Williams, M. C. (2009) Optical tweezers experiments resolve distinct modes of DNA-protein binding, Biopolymers 91, 265–282. 4. Neuman, K. C., and Block, S. M. (2004) Optical trapping, Rev Sci Instrum 75, 2787–2809. 5. Marko, J. F., and Siggia, E. D. (1995) Stretching DNA., Macromolecules 28, 8759–8770. 6. Cluzel, P., Lebrun, A., Heller, C., Lavery, R., Viovy, J.-L., Chatenay, D., and Caron, F. (1996) DNA: An extensible molecule, Science 271, 792–794. 7. Smith, S. B., Cui, Y., and Bustamante, C. (1996) Overstretching B-DNA: the elastic response of individual double-stranded and single-stranded DNA molecules, Science 271, 795–799. 8. Williams, M. C., Rouzina, I., and Bloomfield, V. A. (2002) Thermodynamics of DNA

interactions from single molecule stretching experiments, Acc Chem Res 35, 159–166. 9. van Mameren, J., Gross, P., Farge, G., Hooijman, P., Modesti, M., Falkenberg, M., Wuite, G., and Peterman, E. (2009) Unraveling the structure of DNA during overstretching using multicolor, single-molecule fluorescence imaging, Proc Natl Acad Sci USA, 10.1073/PNAS.0904322106. 10. Shokri, L., McCauley, M. J., Rouzina, I., and Williams, M. C. (2008) DNA overstretching in the presence of glyoxal: Structural evidence of force-induced melting, Biophys J 95, 1248–1255. 11. Rouzina, I., and Bloomfield, V. A. (2001) Force-induced melting of the DNA double helix. 2. Effect of solution conditions, Biophys J 80, 894–900. 12. Rouzina, I., and Bloomfield, V. A. (2001) Force-induced melting of the DNA double helix 1. Thermodynamic analysis, Biophys J 80, 882–893. 13. Cruceanu, M., Urbaneja, M. A., Hixson, C. V., Johnson, D. G., Datta, S. A., Fivash, M. J., Stephen, A. G., Fisher, R. J., Gorelick, R. J., Casas-Finet, J. R., Rein, A., Rouzina, I., and

Measuring DNA–Protein Binding Affinity on a Single Molecule

14.

15.

16.

17. 18.

Williams, M. C. (2006) Nucleic acid binding and chaperone properties of HIV-1 Gag and nucleocapsid proteins, Nucleic Acids Res 34, 593–605. Pant, K., Karpel, R. L., Rouzina, I., and Williams, M. C. (2005) Salt dependent binding of T4 gene 32 protein to single- and double-stranded DNA: Single molecule force spectroscopy measurements, J Mol Bio 349, 317–330. Vladescu, I., McCauley, M., Nunez, M. E., Rouzina, I., and Williams, M. C. (2007) Quantifying force-dependent and zero-force DNA intercalation by single-molecule stretching, Nature Methods 4, 517–522. Smith, S. B., Cui, Y., and Bustamante, C. (2003) Optical-trap force transducer that operates by direct measurement of light momentum, Methods Enzymol 361, 134–162. Wang, M. D., Yin, H., Landick, R., Gelles, J., and Block, S. M. (1997) Stretching DNA With Optical Tweezers, Biophys J 72, 1335–1346. Wenner, J. R., Williams, M. C., Rouzina, I., and Bloomfield, V. A. (2002) Salt Dependence

19.

20.

21.

22.

315

of the Elasticity and Overstretching Transition of Single DNA Molecules, Biophys J 82, 3160–3169. McCauley, M. J., Shokri, L., Sefcikova, J., Venclovas, C., Beuning, P. J., and Williams, M. C. (2008) Distinct Double- and SingleStranded DNA Binding of E. coli Replicative DNA Polymerase III alpha Subunit, ACS Chem Biol 3, 577–587. McCauley, M. J., Zimmerman, J., Maher, L. J., 3rd, and Williams, M. C. (2007) HMGB binding to DNA: single and double box motifs, J Mol Biol 374, 993–1004. McGhee, J. D., and von Hippel, P. H. (1974) Theoretical aspects of DNA-protein interactions: Cooperative and non-cooperative binding of large ligands to a one-dimensional homogeneous lattice. Journal of Molecular Biology 86, 469–489. McCauley, M., Hardwidge, P. R., Maher, L. J., 3rd, and Williams, M. C. (2005) Dual binding modes for an HMG domain from human HMGB2 on DNA, Biophys J 89, 353–364.

Chapter 22 Modeling Nanopores for Sequencing DNA Jeffrey R. Comer, David B. Wells, and Aleksei Aksimentiev Abstract Using nanopores to sequence DNA rapidly and at a low cost has the potential to radically transform the field of genomic research. However, despite all the exciting developments in the field, sequencing DNA using a nanopore has yet to be demonstrated. Among the many problems that hinder development of the nanopore sequencing methods is the inability of current experimental techniques to visualize DNA conformations in a nanopore and directly relate the microscopic state of the system to the measured signal. We have recently shown that such tasks could be accomplished through computation. This chapter provides step-by-step instructions of how to build atomic scale models of biological and solid-state nanopore systems, use the molecular dynamics method to simulate the electric field-driven transport of ions and DNA through the nanopores, and analyze the results of such computational experiments. Key words: Molecular dynamics, Transmembrane transport, Nucleic acids, Membrane proteins, Bionanotechnology, Computer simulations

1. Introduction The successful completion of the human genome project created the opportunity for even more ambitious endeavors in the field of genomic research. For example, the personal genome project aims to establish the relationship between the variations in the DNA sequence among individuals and their health conditions and response to drugs and treatments. The goal of the cancer genome atlas project is to determine which DNA mutations lead to cancer in different human organs and tissues. To make wholegenome sequencing a routine procedure, the time and cost of sequencing must be further reduced by two orders of magnitude or more to less than a day and $1,000, respectively. Among many new approaches to sequencing DNA that are being explored, the so-called nanopore methods promise the

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0_22, © Springer Science+Business Media, LLC 2011

317

318

Comer, Wells, and Aksimentiev

most radical reduction in sequencing time and cost (1). In a ­typical nanopore measurement, negatively charged DNA is driven through a nanometer-sized pore in a nanometer-thick membrane by applying a voltage difference. As DNA passes through the nanopore, the sequence of nucleotides is detected by a readout mechanism. This enables detection of the nucleotide sequence directly from the DNA strand, requiring small amounts of reagents, simple sample preparation procedures and no limit on the length of the DNA fragment that can be read in one measurement. At present, several types of observable signals are being explored as a readout mechanism for nanopore sequencing, including transverse tunnelling current (2), capacitance readout (3), and fluorescence (4). The originally proposed and most explored readout method (5) relies on an ionic blockade current, uniquely determined by the identity of the DNA nucleotide occupying the narrowest constriction in the pore. To sequence DNA using a nanopore, the sequence-specific signal must be deciphered from the background of the conformational noise. Hence, to elucidate the molecular origin of the ionic current blockades, the conformation of DNA in a nanopore must to be characterized with atomic precision. In the absence of an experimental approach capable of accomplishing this task, molecular dynamics simulations have emerged as a kind of a computational microscope that can not only provide the atomic scale images of DNA conformations in a nanopore but also accurately predict the ionic current blockades (6, 7) and characterize the forces involved (8, 9). This chapter provides step-by-step instructions for using the molecular dynamics method to design nanopore systems for sequencing DNA. The chapter is organized as follows. In Subheading  2, we describe the software required, and the initial atomic structures that will be assembled into atomic-scale models of the nanopore systems. In Subheading  3, we first describe how to build and ­simulate systems containing DNA (see Subheading  3.1), the ­biological pore a-hemolysin (see Subheading  3.2), crystalline ­silicon nitride (see Subheading  3.3) and amorphous silicon ­dioxide (see Subheading  3.4) nanopores. Following that, we describe procedures for adding DNA to a-hemolysin (see Subheading  3.5) and synthetic nanopore (see Subheading  3.6) systems. Next we describe methods to simulate the a-hemolysin and a-hemolysin-DNA systems under a transmembrane bias (see Subheading  3.7), use grid-steered molecular dynamics (10) to study DNA transport through a-hemolysin (see Subheading 3.8), and to study synthetic nanopore systems under a transmembrane bias (see Subheading  3.9). The last Subheading  3.10 describes the most common analysis tasks that we use to characterize the outcomes of our computational experiments.

Modeling Nanopores for Sequencing DNA

319

2. Materials 2.1. Software and Scripts

1. VMD.  Download VMD from http:// www.ks.uiuc.edu/ Research/vmd. VMD is an extremely useful visualization and analysis package, and will be used throughout this chapter. VMD supports all major computer platforms. It will run on virtually any system, but benefits greatly from the latest graphics­ hardware and high memory capacity. For more information on VMD, please refer to the VMD User Guide (11) and VMD Tutorial (12). VMD is run by typing vmd on the terminal for Linux, with its icon in Windows, or either method in Mac OS X. The code presented here was tested with VMD version 1.9. Not all steps may work correctly with other versions. 2. NAMD.  Download NAMD (13) from http:// www.ks.uiuc. edu/Research/namd. NAMD is a state-of-the-art, highly scalable molecular dynamics code. NAMD binaries are available for Linux/ UNIX, Mac OS X, and Windows. NAMD will run on a desktop or laptop, but for systems of more than a few thousand atoms it is highly recommended that simulations be run on parallel clusters or supercomputers. In all simulations – except those for annealing SiO2 (Subheading 3.4, steps 8 and 9) – a multiple time-stepping method is used in which bonded forces are calculated on intervals of 1 fs, Lennard-Jones and short-range electrostatic forces are calculated on intervals of 2 fs, and long-range electrostatic forces are calculated on intervals of 4 fs. Note that the use of rigid water molecules and rigid bonds to hydrogen (as accomplished by specifying “rigidBonds all” in the NAMD configuration file) and a 2 fs timestep may be recommendable for simulations using the CHARMM force field, although they are not used in this work. In this case 2 fs, 2 fs, and 6 fs may be used for the above multiple time-stepping intervals. In all simulations, electrostatic forces are computed by the particle mesh Ewald algorithm using a grid spacing <1.5 Å. For more information on NAMD, please refer to the NAMD User Guide (14) and NAMD Tutorial (15, 16). The code presented here was tested with NAMD version 2.7. Not all steps may work correctly with other versions. 3. Associated files.  Scripts and other support files are availble in an archive on the publisher Web site and http://bionano.physics. illinois.edu/Tutorials/nanopore-protocols.tar.gz. Extract the archive in a working directory on your filesystem. The files are contained in the nanopore–protocols directory. All file locations referred to in this work are relative to this directory. Figure 1 displays the organization of the directories which contain the files necessary for this work. Each of these directories includes a subdirectory called output, which contains example output from performing the procedures in each section. If you encounter problems with any of the procedures, check the output directory for possible solutions.

320

Comer, Wells, and Aksimentiev grid

nanopore-protocols

building-dna

(§3.1)

building-ahl

(§3.2)

building-sin

(§3.3)

building-sio

(§3.4)

building-ahl+dna

(§3.5)

building-sin+dna

(§3.6)

running-ahl

(§3.7)

running-ahl+dna

(§3.8)

running-sin

(§3.9)

analysis

(§3.10)

Fig.  1. Subdirectories within nanopore-protocols containing the files ­necessary for each section. The section number is shown to the right of the corresponding working directory.

4. CHARMM topology and parameter files.  Download the CHARMM topology and parameter files from http://mackerell. umaryland.edu/CHARMM_ff_params.html. Select the c32b1 version. Extract the tarball to your nanopore–protocols directory. The CHARMM topology files top_all27_na.rtf and top_all27_prot_lipid.rtf as well as the CHARMM parameter files par_all27_na.prm and par_all27_prot_ lipid.prm are used in this work. We must edit the topology files to remove the H–H bond in water molecules. In both top_all27_na.rtf and top_ all27_prot_lipid.rtf, change the line BOND OH2 H1 OH2 H2 IS NEEDED FOR SHAKE

H1

H2

!

THE

LAST

BOND

\

to BOND OH2 H1 OH2 H2

This change should be made in any topology file you use which contains water when simulating with NAMD. 5. SOLVATE.  Download the source code for SOLVATE from h t t p : // w w w. m p i b p c . m p g . d e / h o m e / g r u b m u e l l e r / downloads/solvate/index.html. Download the archive files to the building-ahl subdirectory. Extract the files from the archive – the source files should be in the directory solvate_1.0. Change to the directory solvate_1.0 and enter the following in the Linux terminal: cc –ansi –O –o solvate solvate.c −lm cp solvate ..

The compiled SOLVATE program should be ready for use in Subheading 3.2. On Windows OS, follow the usual instructions for compiling a C code.

Modeling Nanopores for Sequencing DNA

321

6. Grid manipulation programs.  Included in the support files is C++ source code for manipulating the potential grids used in some of the steps. Here we give an example of how to compile them using the Gnu Compiler Collection. In the Linux terminal, change to the subdirectory grid and enter the following commands: g++ –O2 –Wall –o gridExternalField gridExternalField.C g++ –O2 –Wall –o thirdForce thirdForce.C g++ –O2 –Wall –o gridSourcePore gridSourcePore.C g++ –O2 –Wall –o gridShave gridShave.C

On the Windows OS, follow the usual instructions for ­compiling a C++ code. 2.2. D  NA

1. Create a canonical B-DNA structure. An exemplary B-DNA structure (dsdna_raw.pdb) is included in the buildingdna subdirectory. However, such structures can be generated using 3D-DART (17). Use your Web browser to navigate to http://haddock.chem.uu.nl/services/3DDART/. Enter A for the sequence and 40 for the number of repeats. Also check the box “Convert nucleic acid 1 letter to 3 letter notation” under “Step 3: PDB formatting options.” Click “Submit,” then download and unzip the resulting file. 2. Move and rename PDB file. In the unzipped directory, the DNA file will be named dna1_fixed.pdb in the jobnr8-­ PDBeditor directory. An image of the structure is shown in Fig. 2a. You can copy this file over building-dna/dsdna_ raw.pdb.

2.3. S  ilicon Nitride



The unit cell used to create the Si3N4 structures in subsequent sections is included with VMD. The CHARMM format parameter file silicon_nitride.par defines the interaction energy between the atoms of the membrane and all atoms of the system. The bond energy between the membrane’s silicon and nitrogen atoms is given by V Si-N = K (r - b)2 , where r = rSi - rN , K = 5.0  kcal/(mol Å2), and b = 1.777 Å (18). The nonbonded interactions of the membrane’s atoms consist of a Coulomb portion and a Lennard-Jones portion: V NB

1 qi q j = + eij 4pe0 rij

éæ R êç ij êçè rij ë

12

ö æ Rij ÷ - 2ç ÷ ç rij ø è

ö ÷ ÷ ø

6

ù ú ú û

(1)

with eij = ei × e j and Rij = Rimin 2 + R jmin 2 . The atom-specific parameters for this function are given in Table 1. (See Note 1). To maintain the structure of the Si3N4 membrane, harmonic restraints are applied to all of its atoms. These restraints, along with

322

Comer, Wells, and Aksimentiev

the bond parameters, are chosen to give the Si3N4 a relative permittivity of 7.5 (19). The internal (away from the surface) atoms of the membrane are restrained to their X-ray coordinates r0 with a force given by Frestraint (r) = - 2k(r - r0 ) , where K = 1.0 kcal/(mol Å2). The surface atoms are restrained with K  = 10.0  kcal/(mol Å2) to prevent large distortions of the Si3N4 surface.

Fig. 2. (a) Canonical B-DNA. Atoms are shown as van der Waals spheres. Hydrogen is not shown. (b) Single-stranded poly(dA)40 DNA. (c) DNA after using the phantom pore method to ensure that the strand will fit inside a-hemolysin when the systems are combined later.

Table 1 Parameters for the energy function Eq. 1 used in ­simulations of Si3N4 structures (19, 20). Atom

q (e )

e (kcal/mol)

Rmin /2 (Å)

Si

0.767900

0.31

2.1350

N

−0.575925

0.19

1.9975

Modeling Nanopores for Sequencing DNA

323

Furthermore, in our simulations of DNA–Si3N4 systems, we apply an additional DNA-specific force to reduce adhesion of the DNA to the pore walls (18). Each atom i of the DNA feels a repulsive force due to each Si3N4 atom j given by: surf ij



F

ì F0 eij ï = íF0 (1 - (rij - R) s)eij ï0 î

if rij < R if R < rij < R + s otherwise

(2)

Here we use R = 2 Å, s = 2 Å, and F0 = 2 kcal/(mol Å). This force is implemented using grid-steered molecular dynamics (10). In Subheading 3.6, we create a grid of the potential energy due to this force term. Note that in that example we produce grids (dx files) with F0 = 1, and a scaling factor of 2 is applied in the NAMD configuration files to obtain F0 = 2 kcal/(mol Å). 2.4. Silicon Dioxide

The coordinates of a unit cell of crystalline SiO2 are included with VMD. To construct an amorphous SiO2 structure, the crystalline structure is annealed (21, 22) using the BKS potential (23) with the short range modification of Vollmayr et al. (24). The properties of the amorphous structures produced by this potential (bond angle distributions, coordination numbers, etc.) are in good agreement with experiment (24). The BKS potential includes a Coulomb electrostatic term and the Buckingham potential describing the van der Waals and exclusion interactions: V BKS =



Cij 1 Q iQ j + Aij exp -Bij rij - 6 4pe 0 rij rij

(

)

(3)

The parameters of this potential are displayed in Table  2. The Buckingham terms of the potential are also described in the parameter file SiOtab.par and the tabulated potential file bkstab.dat. In our approach, the BKS potential is used only to obtain amorphous SiO2 structures. Other potential functions are used to simulate interactions of such SiO2 structures with water, ions and biomolecules (21, 25). In the latter case, the atoms of SiO2 are

Table 2 Parameters for the energy function Eq. 3 used for annealing SiO2 structures (23, 24) Atom pair

A (kcal/mol)

B (Å−1)

C (Å−6 kcal/mol)

Q (e)

Si Si

0.000 × 103

0.00000

0.0000

qSi = 2.4

OO

3.026 × 10

2.76000

175.0000

qO= −1.2

Si O

415.176 × 10

4.87318

133.5381

3 3

324

Comer, Wells, and Aksimentiev

Table 3 Parameters for the energy function Eq. 1 used for simulations of SiO2 structures along with water, ions, and biomolecules (21) Atom

q (e)

e (kcal/mol)

Rmin /2 (Å)

Si

1.00

0.30

2.1475

O

– 0.50

0.15

1.7500

restrained to their coordinates obtained at the end of the annealing procedure. The restraint forces are defined as Frestraint (r) = - 2k(r - r0 ) with the force constant K = 20.0  kcal/(mol Å2). Combined with the Si–O bond energy term V Si -O = K (r - b)2, where K = 1.0 kcal/ (mol Å2) and b = 1.6 Å, the restraints give the bulk amorphous SiO2 a dielectric constant of ~5, which can also depend on the density of the amorphous material. Table 3 lists the Lennard-Jones parameters and the atomic charges of SiO2 used in our simulations of the SiO2 nanopore systems. These parameters are also described in the CHARMM format parameter file silica.par. 2.5. a-Hemolysin

1. Download structure from the Protein Data Bank.  Using your Web browser, navigate to http://www.pdb.org and search for PDB code 7AHL. Click the “Download Files” link on the right-hand side and choose “PDB File (Text)”. The file will be saved as 7AHL.pdb. This file will contain the X-ray structure of heptameric a-hemolysin and the associated crystallographic water. You may wish to examine the site further, as it provides a plethora of information about this and other proteins and other structures in the repository.

3. Methods Here we describe our protocols for preparing models of nanopore systems and simulating these models using Molecular Dynamics (MD). If you wish to build and simulate a system containing DNA, go to Subheading 3.1. The three sections that follow cover model construction and simulation for three types of nanopores commonly used in experiments. The details of how to build and simulate an a-hemolysin pore, a silicon nitride pore, and amorphous silica pore can be found in Subheadings 3.2–3.4, respectively. In Subheading 3.5, we detail methods for combining the a-hemolysin pore built in Subheading 3.2 with the DNA built in Subheading 3.1. In Subheading 3.6, we do likewise for a synthetic pore.

Modeling Nanopores for Sequencing DNA

325

External electric fields are applied in many experiments to cause the translocation of the DNA through nanopores. To simulate the a-hemolysin systems with and without DNA under an external electric field, go to Subheading 3.7. In Subheading 3.8, we describe the implementation of grid-steered molecular ­dynamics to simulate transport of DNA through the a-hemolysin ­system. The simulations of a synthetic nanopore system under an external electric field are described in Subheading 3.9. Finally, in Subheading 3.10, we detail methods for quantitative analysis of the simulations described in the other sections, such as calculations of the ionic current and the DNA transport rate. 3.1. Building DNA

Our first task is to build a DNA structure from the PDB file that we have already downloaded. For MD simulations, every atomic structure requires two files, a PDB file and a PSF file. The PDB file contains atomic coordinates, while the PSF file contains information about the bonds, angles, dihedrals, and improper dihedrals that describe the bond structure of the molecules, as well as the types, masses, and charges of the atoms of the molecule, which are used to compute the interactions between the atoms. Here we construct both single-stranded DNA (ssDNA) and doublestranded DNA (molecules), which we will later place inside alphahemolysin and a synthetic pore, respectively. The scripts used in this section are located in the building-dna subdirectory. 1. Open VMD.  First open VMD, either by double-clicking its icon in Mac OS X or Windows, or by entering vmd on the terminal of Mac OS X or Linux. 2. Open the TkConsole.  We will enter commands to VMD through its Tcl-based console, called the TkConsole, or TkCon for short. To open the TkCon, select the “Extensions → Tk Console” menu item from the VMD Main window. 3. Change directories.  If you are not currently in the buildingdna directory, change to it by entering the following in the TkCon: cd your_working_directory/nanopore-protocols/building-dna

4. Separate chains.  We will use the VMD plugin psfgen to generate the PDB. This program requires each chain to be in its own PDB file. To do this, run the script separate.tcl by typing source separate.tcl

in the TkCon. The content of the script is reproduced here: mol load pdb dsdna_raw.pdb set all [atomselect top all] $all moveby [vecinvert [measure center $all weight \ mass]]

326

Comer, Wells, and Aksimentiev $all moveby "0 0 20" foreach chain {A B} { [atomselect top dsdna_$chain.pdb

"chain

$chain"]

writepdb

\

}

The script first loads the PDB downloaded previously, creates atom selections, moves the atoms so that their center of mass is at position (0,0,20), then creates a selection for each chain separately and writes a PDB of each. 5. Generate PSF file.  Next we use psfgen to make the PSF file. Run the script make-psf.tcl in the VMD TkCon. The script uses various commands for generating the PSF structure. The segment command creates a segment for each chain, the pdb command reads which atom resids and names are in the segment, coordpdb reads atomic coordinates, and guesscoord places atoms missing from the PDB, notably hydrogen atoms. (See Notes 2 and 3). 6. Make single-stranded DNA.  To generate ssDNA, we use psfgen to remove one strand from our system. Run the script make–ssdna.tcl. Warnings of poorly guessed coordinates may be safely ignored. The delatom command as used deletes the entire DNAB segment. We are left with a strand of poly(dA)40 as desired. You should have a system similar to that shown in Fig. 2b. 7. Solvate ssDNA.  Use the Solvate VMD plugin to place water around the DNA we just made. Run the script solvate– ssdna.tcl. The line solvate ssdna.psf ssdna.pdb –minmax {{–30 –30 –70}\ {30 30 110}} –o ssdna+solv

instructs Solvate to create a right rectangular box of TIP3 water from coordinate (−30,−30,−70) to (30,30,110). Water molecules overlapping the DNA are removed, and the complete system is saved as ssdna.psf and ssdna.pdb. 8. Add ions.  We now use the Autoionize VMD plugin to add ions to our system to achieve a NaCl solution having a molality of 1.0  mol/kg (see Note 4). Run the script ionize– ssdna.tcl. The contents of the script are shown below. Each nucleotide of the DNA molecule has a charge of −e, except the the two terminal nucleotides, which together sum to −e. Thus, we add an additional 39 Na+ ions to neutralize the system. if you wish to use KCl instead of NaCl, see Subheading 3.3 step 11 for instructions on how to convert the sodium ions to potassium ions. mol delete all mol load psf ssdna+solv.psf pdb ssdna+solv.pdb

Modeling Nanopores for Sequencing DNA

327

set conc 1.0 set water [atomselect top "name OH2"] set num [expr {int(floor($conc*[$water num]/(55.523 + \ 2.0*$conc) + 0.5))}] set nna [expr {$num+39}] package require autoionize autoionize –psf ssdna+solv.psf –pdb ssdna+solv.pdb \ [list "SOD $nna" "CLA $num"] –o ssdna+ions

9. Minimize.  The first simulation step using NAMD is to minimize the system. Minimization takes the system to the nearest local energy minimum (hence the name), resolving steric clashes and other high-energy configurations that would lead to large forces and unstable dynamics. Run NAMD by typing the following command on the terminal: namd2 min.namd > min.log &

This will take some time to run. On Linux and Mac OS X, you may monitor the progress of your job using the command less min.log

and typing shift-F. You may exit by typing control-C followed by Q. Minimization is a relatively fast process and should take several minutes on a PC. 10. Heat and equilibrate.  Next bring the system up to temperature at constant volume using the Langevin thermostat by running NAMD with the heat.namd config file, followed by equilibration at constant pressure using the eq.namd config file. The first simulation raises the temperature of the DNA system to 310 K. The second maintains that temperature, and further achieves a pressure of 1 atm by changing the volume of the periodic simulation cell. Equilibration allows the system to relax and fluctuate around an equilibrium state under given external conditions. The heating procedure should again finish in a few minutes on a PC, while the equilibration is run for 1 ns of simulation time and will take a couple of hours. Refer to the NAMD manual for further information about the details of the equilibration procedure 3.2. Building and Equilibrating a-Hemolysin

In this section, we build a system containing a-hemolysin, a lipid bilayer membrane, water, and ions. We then minimize and equilibrate the system. Scripts for this section are located in the buildingahl subdirectory. 1. Load PDB into VMD. Load the PDB of a-hemolysin by typing the following command in the TkCon: mol load pdb 7AHL.pdb

328

Comer, Wells, and Aksimentiev

2. Separate individual chains.  psfgen requires each chain of the downloaded PDB file to be split into its own separate PDB file. Run the script separate.tcl. 3. Make PSF.  We use VMD’s psfgen tool to make the PSF file for the system. During this step, we set the protonation states for histidines, add hydrogens, and produce structure files. Run the script make–psf.tcl. 4. Rotate and reposition a-hemolysin.  We want to align a-hemolysin with the z-axis, and want the center of the beta barrel to be in the center of the membrane, i.e. z = 0. This is most easily accomplished now, before combining systems. Run the script move.tcl. As seen below, the center of the membrane will also be placed at z = 0. 5. Solvate the protein.  We will now place a layer of water around a-hemolysin. For this purpose, we use the SOLVATE program (26). Unlike the Solvate plugin for VMD, used later, SOLVATE individually places water molecules around a protein based on the protein surface geometry. Type the following on the Linux command line: ./solvate -t 3 -n 8 -w protein solvate_raw This produces a 3Å-thick shell of water around a-hemolysin. The -n 8 option instructs the program to use 8 gaussians to approximate the protein surface. The command creates the file solvate_raw.pdb, containing only water thanks to the -w option. Now we must make a PSF file for the water. Using the VMD TkCon, run the script solvate.tcl. This scripts separates each segment into a separate PDB file, then create a PSF file and new PDB file. Example results of running SOLVATE are shown with a-hemolysin in Fig. 3a.

Fig. 3. (a) a-hemolysin solvated using the SOLVATE program (26). (b) a-hemolysin combined with lipid bilayer membrane. (c) a-hemolysin and lipid after solvating using the Solvate VMD plugin. (d) Final a-hemolysin system, with ions added using the Autoionize VMD plugin.

Modeling Nanopores for Sequencing DNA

329

6. Build the lipid membrane.  To build the lipid bilayer membrane in which a-hemolysin will sit, we use the Membrane plugin for VMD (27). This plugin uses pre-equilibrated patches of either POPC or POPE lipid bilayers, tiling and trimming them to achieve the desired size. To use the plugin, run the script make-membrane.tcl. The resulting files, membrane.pdb and membrane.psf, describe the structure of the membrane. They also contain a thin layer of pre-equilibrated water. These will now be loaded into VMD and become the top molecule. For convenience, the script translates the ­coordinates such that the center of mass of the bilayer is at the origin. 7. Combine protein, membrane, and water.  Run the script combine.tcl. See Note 5. 8. Remove overlapping lipids and associated water.  a-hemolysin is now combined with the membrane, but there are also many undesirable overlaps between the atoms. We therefore next remove lipid and water molecules that were placed too close to the protein, as well as water from within the lipid membrane. Run the script fix-solv.tcl. The script will remove any lipid molecules located within 2 Å of the a-hemolysin stem, water placed by the Membrane plugin and located within 2 Å of a-hemolysin or in its stem, and any water placed by SOLVATE that lies outside the stem and within the lipid bilayer. Your system should look similar to that shown in Fig. 3b. 9. Solvate the system.  We now use the Solvate VMD plugin to place our system in a water box. Run the script vmdsolvate.tcl. The script not only generates the water box we need, but also removes some of the water added by Solvate that extends outside of the system. Your system should look similar to that shown in Fig. 3c. 10. Add ions.  Finally, we use the Autoionize VMD plugin to add ions to our system to achieve a NaCl solution having a ­molality of 1.0 mol/kg as we did for DNA earlier. See Notes 4 and 6. Here we reduce the number of K+ ions by 14 to neutralize the charge of the a-hemolysin. Execute ionize.tcl in the VMD TkCon. See Subheading 3.3 step 11 for instructions on how to convert the sodium ions to potassium ions. The resulting system is shown in Fig. 3d. 11. Make the constraints file.  During minimization and equilibration, we will apply harmonic restraints to Ca atoms of a-­ hemolysin. We tell NAMD which atoms are subject to the restraints with the help of a PDB file in which the restrained atoms have a beta value of 1.0, while unrestrained atoms have a beta value of zero. Run the script make–constraints.tcl. 12. Prepare the minimization input file.  Minimization is the first step after building a system. Minimization takes the system to

330

Comer, Wells, and Aksimentiev

the nearest local energy minimum in configuration space, alleviating conditions such as steric clashes. Examine the file min.namd. 13. Minimize the system.  Minimization for this system is done in two stages. During the first stage, heavy protein atoms are held fixed to allow water, lipid, and protein hydrogen atoms to relax. In the second stage, protein alpha carbons are harmonically restrained. Run the minimization as before, first using the min.namd config file, followed by the min2.namd   file. 14. Heat the system.  Next we will heat the system using the temperature reassign feature of NAMD. Examine the file heat. namd. Run this on the Linux command line using: namd2 heat.namd > heat.log & 15. Equilibrate the system.  Now we will equilibrate the system. This allows the ions, the sidechains of a-hemolysin, and the overall system size, among other things, to relax. Examine the file eq.namd. Notice the lines regarding the Langevin piston. This simulation is run in the NPT ensemble, meaning that the periodic cell size is now variable, and is changed by NAMD to achieve the target pressure of 1  atm. The useFlexibleCell and useConstantRatio keywords tell NAMD to change the z dimension independently of the x and y dimensions, and to always keep the ratio x/y the same (in our case equal to 1). Next run the simulation. This may be accomplished on the Linux command line in the same fashion as in the previous steps; however, it is highly recommended that the simulation be run on a parallel machine, as running this system for 1  ns on a single CPU would take on the order of 1,000 h. 16. Determine the average system size.  Before beginning production simulations of the ionic current, we determine the average steady-state size of the periodic cell during equilibration. This is necessary because we will be simulating in the NVT ensemble, which is always advisable when applying external forces such as an electric field to your system. Run the script average-size.tcl. This script calculates the average system size in the x dimension (the y dimension is the same since we used the useConstantRatio yes keyword in the NAMD config file) and the z dimension, and prints them to the screen. 17. Minimize the system using the new system dimensions.  Examine the file posteq-min.namd. Replace <xymean> and with the values computed in the previous step, and run the simulation.

Modeling Nanopores for Sequencing DNA

331

Fig.  4. Transmission electron micrograph of a nanopore used in DNA translocation experiments having a minimum diameter of approximately 2.0 nm. Taken from Comer et al. (18) with permission from Elsevier.

3.3. Building and Equilibrating a Silicon Nitride Pore

Suppose that experimentalist colleagues ask you to model their DNA translocation experiments conducted using silicon nitride nanopores. They write out the details of their experiments, including the electrolyte concentrations, etc., and hand you the transmission electron micrograph shown in Fig. 4 as an example of the nanopores fabricated in their lab. The instructions below describe how you might proceed in performing MD simulations that model their experiments. In this section, we construct a pore from crystalline Si3N4, add water and ions, and perform simulations of the complete ­system. Scripts for this section are located in the building-sin subdirectory. 1. Define geometry of the Si3N4 membrane.  We are told that the membrane in which the experimentalist’s pore is housed has a thickness of ~10  nm. To produce an appropriate model, create a hexagonal prism having a length along the z axis of 36 Si3N4 unit cells (10.4472 nm), which approximately corresponds to the correct thickness for the membrane. Make the hexagonal cross-section of the structure in the xy plane to have an inscribed diameter of 12 Si3N4 unit cells (9.114 nm). The Inorganic Builder plugin for VMD can be used to conveniently generate such structures (see Notes 7 and 8). To define the geometry for use with Inorganic Builder, enter the following in the VMD TkCon: package require inorganicbuilder inorganicBuilder::initMaterials set box [inorganicBuilder::newMaterialHexagonalBox \ Si3N4 {0 0 0} 12 36]

332

Comer, Wells, and Aksimentiev

Fig. 5. Building a pore made of Si3N4. (a) Defining the geometry. (b) Replicating the unit cell. (c) Sculpting the pore.

We also have the option to display the geometry graphically in VMD (as illustrated in Fig. 5a) by entering the commands below in the TkCon: set m [mol new] ::inorganicBuilder::drawHexBox $box $m display resetview

2. Replicate the b-Si3N4 unit cell.  Inorganic Builder will replicate the b-Si3N4 unit cell given the geometry defined in step 1 (see Note 8). package require psfgen inorganicBuilder::buildBox $box sin

The structure is written to the files sin.psf and sin.pdb. The resulting structure should resemble that shown in Fig. 5b. 3. Record the periodic cell vectors for the entire structure.  To write the periodic basis vectors to a file called cell_basis.txt, enter the following in the TkCon: set out [open cell_basis.txt w] foreach v [inorganicBuilder::getCellBasisVectors $box] \ {puts $out $v} close $out

4. Add bonds.  Add bonds between all pairs of silicon and nitrogen atoms with distances between them <1.9 Å by entering the following in the TkCon: inorganicBuilder::buildSpecificBonds $box {{SI N 1.9}} \ {true true false} top

The parameter {true true false} specifies that bonds are added across the periodic boundaries in the xy plane, but not along the z axis. The hexagonal faces will be free surfaces. This step may take a few minutes to complete (see Note 9).

Modeling Nanopores for Sequencing DNA

333

5. Write the structure files.  Enter the following in the TkCon: set all [atomselect top all] $all writepsf sin_bonded.psf $all writepdb sin_bonded.pdb

6. Sculpt a double-cone pore.  Remove atoms that satisfy the criterion

x 2 + y 2 < d0 / 2 + z tan(g )

(4)

where (x  y  z) is the position of the atom’s center, d0 is the minimum diameter of the pore, and g is the angle that the pore walls make with the z axis. Here we choose d0 = 2.4 nm and g = 10° , a geometry suggested by electron microscopy of real pores (28, 29). Here in our example, we wish to create a pore to mimic one shown in Fig. 4, an electron transmission micrograph with an apparent minimum radius of 2.0  nm. The micrograph shows roughly the extent of the electron clouds of the atoms in the pore; thus, the apparent pore diameter in the image is with respect to the surfaces of the electron clouds. However, when we form the pore, we delete atoms based on their centers. Thus, setting d0 = 2.0 nm in Eq. 4 would yield a pore that is too small. As a heuristic for determining the value of d0, we consider that the r -12 portion of the Lennard-Jones potential is supposed to represent repulsion due to overlapping electron clouds of the atoms involved. We therefore assume that the interaction potential between the electrons produced by the microscope and an atom of the nanopore is shaped something like the r -12 portion of the Lennard-Jones potential between a particle of zero radius and that atom. The r -12 portion of the Lennard-Jones potential becomes very steep near the radius at which the potential crosses zero; thus, we take R apparent = R min 2 ´ 21 6 , which gives Rapparent ~0.2  nm for both silicon and nitrogen atoms. Adding this radius to atoms on both sides of the pore leads us to choose d0 = 2.4 nm. To remove the atoms using VMD and psfgen, execute cutPore.tcl. An illustration of the resulting pore is shown in Fig. 5c. 7. Change the types of the nitrogen atoms.  Inorganic Builder gives the nitrogen atoms of the membrane the type “N” in the PSF file. However, other atoms in the CHARMM force field use the type “N”. For this reason, we need to change the atom types of the nitrogen from “N” to “NSI”. The contents of the script changeTypesNitrogen.tcl are shown below: set nit [atomselect top "type N"] $nit set type NSI set all [atomselect top all]

334

Comer, Wells, and Aksimentiev $all writepsf sin_pore_types.psf $all writepdb sin_pore_types.pdb

Enter source changeTypesNitrogen.tcl in the VMD  TkCon to execute this script. 8. Set the charges.  The charges used in this Si3N4 model are qSi = 0.7679  e and qSi = 0.575925  e, taken from Wendell and Goddard (20). However, when atoms are removed to cut the pore, the ratio of silicon to nitrogen atoms is not maintained. The script setCharges.tcl sets the charges, shifting those on some atoms by a small amount to obtain a neutral system. Execute this script now by entering source setCharges. tcl in the VMD TkCon (see Notes 10 and 11). 9. Solvate.  We add water above and below the membrane to obtain a total water thickness 1.5 times the thickness of the membrane. We do this so that the two sides of the membrane are effectively electrically isolated from each another while also keeping the number of atoms as low as possible. The water can be added to the system using the VMD’s Solvate plugin. Enter the following in the VMD TkCon: package require solvate solvate sin_pore_charges.psf sin_pore_charges.pdb \ –z 75 +z 75 -o sin_sol

The water is added in cuboid as shown in Fig. 6b. 10. Cut the water to the system dimensions.  We need to remove some of water in the cuboid to conform to the boundary conditions of the system. Enter source cutWaterHex. tcl in the VMD TkCon to cut the system to a hexagon as illustrated in Fig. 6c.

Fig. 6. Solvating the Si3N4 pore. (a) The Si3N4 pore before solvation. (b) Adding a cuboid of water. (c) Cutting the water to the system dimensions.

Modeling Nanopores for Sequencing DNA

335

11. Add ions.  Create a solution having a KCl concentration of 1.0 mol/kg. For this purpose, we use VMD’s Autoionize plugin. The contents of the script addIons.tcl are shown below. We explicitly give Autoionize the number of ions to ensure that we have the correct molality in mol/kg (see Note 4). resetpsf mol load psf sin_hex.psf pdb sin_hex.pdb set conc 1.0 set water [atomselect top "name OH2"] set num [expr {int(floor($conc*[$water num]/(55.523 + \  2.0*$conc) + 0.5))}] package require autoionize autoionize –psf sin_hex.psf –pdb sin_hex.pdb \  -nions [list "POT $num" "CLA $num"] –o sin_ions

Enter source addIons.tcl in the VMD TkCon. 12. Check the concentrations.  The script getConc.tcl displays the concentrations of ions and total charge of the system. Enter source getConc.tcl in the VMD Tkcon to check that the concentrations are nearly 1.0 mol/kg and that the total charge is nearly zero. 13. Define harmonic restraints.  To maintain the solid structure of the membrane, we apply harmonic restraints to all atoms of the membrane. Surface atoms and internal atoms are restrained with energy constants of 10.0 and 1.0 kcal/(mol Å2), respectively. The contents of the script markRestraints.tcl are shown below: mol load psf sin_ions.psf pdb sin_ions.pdb set all [atomselect top all] $all set beta 0.0 set sel [atomselect top "resname SIN"] $sel set beta 1.0 set surf [atomselect top "resname SIN and \  ((name \"SI.*\" and numbonds<=3) or (name \"N.*\" and \ numbonds<=2))"] $surf set beta 10.0 $all writepdb sin_restrain.pdb

Enter source markRestraints.tcl in the VMD TkCon to execute this script. 14. Thermostat the membrane.  During the simulations, a Langevin thermostat will apply to only the atoms of the Si3N4 membrane, using a damping constant of 1.0 ps−1 We can produce a PDB files to implement the restraints and temperature control in NAMD by executing the commands below in the VMD TkCon. These commands must be entered immediately following those in the last step. $all set beta 0.0 $sel set beta 1.0 $all writepdb sin_langevin.pdb

336

Comer, Wells, and Aksimentiev

15. Minimize.  Before beginning the minimization we need to copy the basis vectors for the periodic cell into the NAMD configuration file. The parameters cellBasisVector1 and cellBasisVector2 should be set to the first and second lines, respectively, of cell_basis.txt. We need to add 150 to the value in the third row and third column of cell_ basis.txt to account for the space occupied by water (150 Å along the z axis). Run NAMD by entering the following in the Linux shell: namd2 sin_min.namd > sin_min.log &

16. Equilibrate.  We first simulate the system for 600  ps with pressure control applied. In this simulation, the system size will fluctuate. Enter the following in the Linux shell: namd2 sin_eq.namd > sin_eq.log &

3.4. Building and Annealing a Silica Pore

In this section, we anneal a SiO2 structure to create an annealed amorphous SiO2 pore. We then add water and ions and equilibrate the system. Scripts for this section are located in the building–sio subdirectory. 1. Set the system geometry.  Here we create a system with similar dimensions to those used for the Si3N4 pore. Enter the following in the Linux terminal: cp ../building-sin/cell_basis.txt grid_basis.txt

Edit the file grid_basis.txt in a text editor and add 15 to the value in the third row and third column. This will add 15 Å of vacuum between the surfaces of the membrane. 2. Define the region from which SiO2 atoms will be expelled.  Run the program gridSourcePore by entering the following in the Linux shell. ../grid/gridSourcePore grid_basis.txt 2 104.472 24 10 \ pore_points.txt

The first parameter determines the dimensions of the system. For this we have given the program grid_basis. txt, which has the system basis vectors, see Subheading 3.3. The next parameter sets the approximate spacing of the grid points, which here is 2 Å. The next three parameters describe the pore geometry. The thickness of the membrane is 104.472 Å, its minimum diameter is 20.0 Å, and the angle between the pore walls and the pore axis is 10°. The points in the region from which we want to expel the SiO2 atoms are contained in pore_points.txt. Note that gridSourcePore displays the volume of the different regions of the system. Record the value of the “Remaining volume”, which we will need below to calculate the density of SiO2.

Modeling Nanopores for Sequencing DNA

337

3. Create a grid to sculpt the pore.  The program thirdForce can be used to create a grid that forces the atoms of the membrane into the desired shape. ../grid/thirdForce pore_points.txt 1 3 3 anneal_grid.dx

grid_basis.txt

\

The first parameter contains points we created in the last step. The next parameter gives the dimensions of the system. The third parameter is the approximate resolution of the grid, which here is 1 Å. The next two parameters specify the values of R and s from Eq. 2. The last parameter is the name of the resultant grid. 4. Create a crystal of SiO2 with the appropriate number of atoms. Each unit cell of the silica has a mass of mu =240.337 Da. The number of unit cells required is therefore n u = rV mu , where r is the desired density of SiO2 and V is the “Remaining volume” calculated above. Here we choose a density value of 1.8 Da/Å3 » 3.0 g/cm3. You can use a density value of your choice. In the script buildSystem.tcl, we set nu using this formula and then employ Inorganic Builder to obtain a PDB file with the correct number of silicon and oxygen atoms. Enter source buildSystem.tcl in the VMD TkCon immediately after performing the step above. The contents of this script are listed below: set targetDensity 1.8 set remainingVolume 664947.9068 set nu [expr {$targetDensity*$remainingVolume/240.337}] set n [expr {int(ceil(sqrt($nu)))}] inorganicBuilder::initMaterials set box [inorganicBuilder::newMaterialBox SiO2 {0 0 0} \ [list $n $n 1]]

5. Randomly place the atoms within the membrane.  The script distributeAtoms.tcl takes the atoms from the crystal we just created and distributes them within the membrane. Enter source distributeAtoms.tcl in the VMD TkCon. The system should look like that illustrated in Fig. 7a. 6. Set the charges and types for annealing.  Because the type “O” already exists in the CHARMM parameter set, we change the type of the oxygen atoms to “OSI”. Also, we need to set the charges to those of the BKS force field (24), which are shown in Table 2. The contents of the script setChargesAnneal. tcl are shown below: mol load psf sio.psf pdb sio_ready.pdb set all [atomselect top all] set sil [atomselect top "type SI"]

338

Comer, Wells, and Aksimentiev

Fig.  7. Annealing a pore made of SiO2. (a) Random placement of atoms within the membrane. (b) After energy minimi­zation. (c) Annealing at 7,000 K. (d) Annealing at 300 K.

$sil set charge 2.4 set oxy [atomselect top "type O"] $oxy set charge –1.2 $oxy set type OSI $all writepsf sio_ready.psf

Enter source setChargesAnneal.tcl in the VMD TkCon to execute this script. 7. Mark the atoms for gridforce.  Immediately after entering the commands above, enter the following in the VMD TkCon: $all set beta 1.0 $all set occupancy 1.0 $all writepdb sio_all.pdb

8. Minimize.  We now will simulate the SiO2 using the BKS force field. Run NAMD by entering the following in the Linux shell:

Modeling Nanopores for Sequencing DNA

339

namd2 sio_anneal_min.namd > sio_anneal_min.log &

If you have encounter errors running this simulation, see Note 17. After minimiztion, the system should look like that illustrated in Fig. 7b. 9. Anneal.  Next we increase the temperature of the system to a high value and then slowly cool it to obtain a relaxed amorphous structure (see Note 12). The annealing schedule is: 20 ps at 7,000 K, 20 ps at 5,000 K, 50 ps at 2,000 K, and 50 ps at 300 K. Run NAMD by entering the following in the Linux shell: namd2 sio_anneal.namd > sio_anneal.log &

At the end of the 7,000  K portion of the simulaion, the ­system should look like that illustrated in Fig. 7c, while at the end of the annealing it should look similar to Fig.  7d (see Note 18). 10. Set the charges for production simulations.  Now that the annealing is finished, we will use a force field for silica compatible with CHARMM parameters for water, ions, and biomolecules (21). The contents of the script setChargesProduction.tcl, which sets the charges to those in Table 3, are shown below: mol load psf sio_ready.psf mol addfile sio_anneal.restart.coor set sil [atomselect top "type SI"] $sil set charge 1.0 set oxy [atomselect top "type OSI"] $oxy set charge –0.5

Enter source setChargesProduction.tcl in the VMD TkCon to execute this script. 11. Add bonds.  We will add bonds between all pairs of silicon and oxygen atoms with distances between them < 2.2 Å using Inorganic Builder. Since we used the same system size as in Subheading 3.3, we can initialize the system size given to Inorganic Builder in the same way as in Subheading  3.3. The contents of the script addBonds. tcl are shown below: set box [inorganicBuilder::newMaterialHexagonalBox Si3N4 \ {0 0 0} 12 36] inorganicBuilder::buildSpecificBonds $box {{SI OSI 2.2}} \ {true true false} top set all [atomselect top all] $all writepsf sio_annealed.psf $all writepdb sio_annealed.pdb

340

Comer, Wells, and Aksimentiev

Enter source addBonds.tcl in the VMD TkCon to execute this script, which may take some minutes to run. 12. Solvate.  Enter the following lines in the VMD TkCon. Here we do the same as in Subheading 3.3 step 9. package require solvate solvate sio_annealed.psf sio_annealed.pdb -z 75 +z 75 -o\ sio_sol

13. Cut the water to the system dimensions and add ions.  Perform steps 10 and 11 of Subheading 3.3, replacing the substring sin with sio in all the scripts and script files. 14. Define harmonic restraints.  Here we use harmonic restraints of 20.0  kcal/(mol Å2). The contents of the script markRestraintsSio.tcl are shown below: mol load psf sio_ions.psf pdb sio_ions.pdb set all [atomselect top all] $all set beta 0.0 set sel [atomselect top "resname SIO2"] $sel set beta 20.0 $all writepdb sio_restrain.pdb

Enter source markRestraintsSio.tcl in the VMD TkCon to execute this script. 15. Thermostat the membrane.  Immediately after executing the script above, type the following in the VMD TkCon: $sel set beta 1.0 $all writepdb sio_langevin.pdb

16. Minimize and equilibrate.  Run NAMD by entering the ­following in the Linux shell: namd2 sio_min.namd > sio_min.log & namd2 sio_eq.namd > sio_eq.log &

3.5. Building the a-Hemolysin-DNA System

In this section, we combine DNA and a-hemolysin. This enables the study of the ionic current modulation produced by different sequences of DNA, as well as the transport of DNA itself through a-hemolysin. Scripts for this section are located in the building– ahl+dna subdirectory. 1. Make target file for the phantom pore method.  We must first ensure that the ssDNA we equilibrated earlier will fit in the a-hemolysin pore. To do this, we use the so-called phantom pore method (19) to obtain a conformation of the ssDNA that fits into the a-hemolysin without steric clashes. The method uses the TclBC feature of NAMD to push DNA that will be in the a-hemolysin constriction into a 5 Å-radius cylinder, and keeps the rest of the DNA within a cylinder of 15 Å radius. Before

Modeling Nanopores for Sequencing DNA

341

running the simulation, we make a PDB file designating the atoms to be forced using the beta column. Do this by running the script make–target–ssdna.tcl in the VMD TkCon. 2. Phantom pore.  Run the phantom pore simulation using the config file phantom.namd. An example of the result is shown in Fig. 2c (see Notes 13 and 14). 3. Remove water and ions.  We remove water and ions to leave just the single-stranded DNA, which will later be combined with a-hemolysin. Run the script extract-ssdna.tcl in the VMD TkCon. 4. Combine a-hemolysin and DNA.  Using the a-hemolysin and DNA systems we already have, we first combine the two systems. Run the script combine.tcl in the VMD TkCon. Notice that we use the output of the equilibration simulation as our initial conformation of a-hemolysin. This way both a-hemolysin and the DNA are already equilibrated separately, which will reduce the equilibration time for the combined system. 5. Remove clashes from the a-hemolysin-DNA system.  We must now remove any atoms that overlap DNA in the system we just created. Run the script fix.tcl in the VMD TkCon. 6. Neutralize the a-hemolysin-DNA system.  DNA is charged, and we may have deleted ions in the last step, so we must next reneutralize the system using the Autoionize VMD plugin. Run the script neutralize.tcl in the VMD TkCon. See Note 6. You should now have a system similar to the one shown in Fig. 8. 7. Define restraints.  To prepare for minimization, heating, and equilibration, run the script make-target-ahl+dna.tcl in the VMD TkCon. 8. Minimize and equilibrate the a-hemolysin-DNA system. Minimize, heat, and equilibrate the a-hemolysin-DNA system just as you did for the a-hemolysin system, using the ahl+dna.psf and ahl+dna.pdb files you just created. Only the structure, coordinates, and output keywords need modification. 9. Determine the average system size and minimize again.  Using the same average–size.tcl script used for a-hemolysin alone, calculate the average system size, then minimize the system in the same fashion as before. 3.6. Building the Synthetic Pore-DNA System

In this section, we add the DNA created in Subheading 3.1 to the Si3N4 nanopore we constructed in Subheading  3.3. A similar approach could be used with the SiO2 pore. After adding the DNA, we generate a grid to implement the DNA-specific ­interaction. As in Subheading 3.3, water and ions are added, and the system is equilibrated. The files required for this section are in the building–sin+dna subdirectory.

342

Comer, Wells, and Aksimentiev

Fig. 8. The a-hemolysin-DNA system. Water and ions are not shown.

1. Combine the nanopore and DNA. The following commands use psfgen to combine the Si3N4 nanopore produced in Subheading 3.3 and the DNA produced in Subheading 3.1. To run these commands, enter source combine.tcl in the VMD TkCon. package require psfgen resetpsf readpsf ../building–dna/dsdna.psf coordpdb ../building–dna/dsdna.pdb readpsf ../building–sin/sin_pore_charges.psf coordpdb ../building–sin/sin_pore_charges.pdb writepsf sin+dna.psf writepdb sin+dna.pdb

2. Adjust the DNA position.  In the structure we just created, the DNA is already threaded through the pore and clashes with atoms of the pore. Because the pore is only 20 Å in diameter, the DNA cannot be threaded through the pore in

Modeling Nanopores for Sequencing DNA

343

its canonical form. In subsequent simulations, we will apply an electric field to the system so that the DNA will be forced through the constriction, distorting from its canonical conformation. Now, we place the DNA just above the constriction of the pore, so that we can observe the onset of DNA translocation through the constriction when a field is applied. The contents of the script adjustPos.tcl are shown below: mol load psf sin+dna.psf pdb sin+dna.pdb set all [atomselect top all] set sel [atomselect top "segname DNAA DNAB"] $sel moveby {0 4 65} $all writepdb sin+dna_placed.pdb

To run these commands, enter source adjustPos.tcl in the VMD TkCon. Now all atoms of the DNA should be more than 4.5 Å away from the atoms of the Si3N4. 3. Solvate.  Enter the following commands in the VMD TkCon to make a system of similar size to that in Subheading 3.3: package require solvate solvate sin+dna.psf sin+dna_placed.pdb \ –minmax {{–55 –55 –97} {55 55 157}} –o sin+dna_sol

4. Cut the water to the periodic boundaries.  Enter source cutWaterHex.tcl in the VMD TkCon. 5. Add ions.  Each DNA nucleotide has a charge of −e. To neutralize the system, we need to add a K+ ion for every nucleotide, beyond those required to obtain a 1  M solution. Source addIons.tcl in the TkCon to add the appropriate numbers of ions. Also source getConc.tcl to check that the ions were added correctly. Note that the potassium concentration is somewhat larger than 1 mol/kg. 6. Define harmonic restraints.  The script defineRestraints. tcl contains the following commands. Enter source defineRestraints.tcl in the VMD TkCon to execute them. mol load psf sin+dna_ions.psf pdb sin+dna_ions.pdb set all [atomselect top all] $all set beta 0.0 set sel [atomselect top "resname SIN"] $sel set beta 1.0 set surf [atomselect top "resname SIN and\ ((name \"SI.*\" and numbonds<=3) or (name \"N.*\" and \ numbonds<=2))"] $surf set beta 10.0 $all writepdb sin+dna_restrain.pdb

344

Comer, Wells, and Aksimentiev

7. Thermostat the membrane.  Enter the following commands immediately after those in the last step. $all set beta 0.0 $sel set beta 1.0 $all writepdb sin+dna_langevin.pdb

8. Write the positions of the membrane atoms to a file.  In order to make a grid to implement the DNA-specific force that acts according to Eq. 2, we need to write the positions of the membrane’s atoms to a file. Immediately after entering the commands in the last step, enter source writePos. tcl in the VMD TkCon. The contents of writePos.tcl are shown below: set sel [atomselect top "resname SIN"] foreach quiet {0} { set pos [$sel get {x y z}] } set out [open sin_positions.txt w] foreach r $pos { puts $out $r } close $out

9. Mark the DNA for the DNA-specific force.  Here, we make a PDB file in which the beta column marks the atoms to which the DNA-specific force is applied, whereas the occupancy column specifies the coupling of the atoms to that force. Immediately after entering the commands in the last step, enter source markDna.tcl in the VMD TkCon. The contents of markDna.tcl are shown below: set all [atomselect top all] set sel [atomselect top "segname DNAA DNAB"] $sel set beta 0.0 $sel set beta 1.0 $sel set occupancy 1.0 $all writepdb specific.pdb

10. Generate the grid defining the DNA-specific force.  The DNAspecific force is used to reduce the interaction between the DNA and the pore surface and thus prevent irreversible binding of DNA to the pore walls (6, 18). Note that we do not use this DNA-specific interaction for SiO2 pores. Copy ../building-sin/cell_basis.txt into the current directory. Edit cell_basis.txt and add 20 to the value in the third row and third column. Save the file as grid_ basis.txt. This basis will allow the grid for the DNAspecific interaction to extend 15 Å above and below the upper and lower surfaces of the ­membrane. Run the following command in the Linux shell: ../grid/thirdForce sin_positions.txt grid_basis.txt 1 2 2 \ specific2–2.dx

Modeling Nanopores for Sequencing DNA

345

The first parameter sin_positions.txt contains the equilibrium positions of the membrane’s atoms. The system size is defined by the basis vectors in grid_basis.txt. The third parameter is the approximate resolution of the grid, which here is 1 Å. The next two parameters specify the values of R and s in Eq. 2. 11. Minimize.  Run NAMD by entering the following in the Linux shell: namd2 sin+dna_min.namd > sin+dna_min.log & If you have encounter errors running this simulation, see Note 17. 12. Equilibrate.  Enter the following in the Linux shell: namd2 sin+dna_eq.namd > sin+dna_eq.log & 3.7. Simulating a-Hemolysin Under an External Electric Field

We are now ready to study some transport properties of a-hemolysin. We will first simulate the system in an external electric field corresponding to a 1.2 V transmembrane bias. Such a simulation allows the open-pore ionic conductance and the distribution of the electrostatic potential to be computed. The latter will also be used for G-SMD simulations of DNA transport. Files for this section are located in running–ahl. 1. Make SMD file.  We will be running the simulation in the external electric field without harmonic restraints. However, we would nevertheless like to keep a-hemolysin and the membrane from moving, because it makes calculation of the average electrostatic field easier. To accomplish this, we will use the SMD feature of NAMD. SMD is most often used to accelerate the dynamics of a simulation, but it can also be used to restrain the center of mass of a group of atoms. Run the script make–targets.tcl (see Note 15). 2. Compute the electric field.  We must first compute the electric field corresponding to the desired bias in a system of our size. We compute this as E z = V Lz where V = 1.2 V and Lz is the value zmean computed in Subheading 3.2 step 17 (see Note 20). Furthermore, NAMD requires the input electric field in units of kcal/(mol Å e). The conversion factor is 1 V/Å = 23.0451 kcal/ (mol Å e). Hence, if Lz = 180Å, then Ez = 0.1536 kcal/(mol Å e). 3. Simulate a-hemolysin in an external electric field.  Examine and edit the file electric.namd. Replace <efield> with the value calculated in the previous step, and run NAMD. 4. Calculate the average electrostatic potential in the a-hemolysin system.  We will now use the PMEPot VMD plugin to ­compute the average distribution of the electrostatic potential in the system using the trajectory generated in the previous step (electric.dcd). After adding the external electric field,

346

Comer, Wells, and Aksimentiev

we will then have a map that we can use to accelerate the transport of ssDNA through a-hemolysin. Type the following into the TkCon: mol delete all mol load psf ../building-ahl/ahl.psf mol addfile electric.dcd first 100 waitfor all package require pmepot pmepot –frames all –grid {96 96 128} –xscfile electric.xsc \ –dxfile electric_raw.dx

This command computes the electrostatic potential for each frame, and writes the average over the trajectory to the file electric_raw.dx. Notice that we did not load the first 100 frames, which gave the ions time to approach their steady state distribution after being added randomly in Subheading 3.5 step 6. (See Note 6). 5. Add external electric field to the potential map.  The potential map calculated by PMEPot is the reaction field of the a-hemolysin system, and does not include the potential of the external field we applied. Therefore, we must add the external field back into the map to get the complete electrostatic potential. To accomplish this, type the following command in your terminal: ../grid/gridExternalField electric_raw.dx 0.02585 1.2 \ electric.dx

The first argument is the input file. The second argument scales all grid values by that value, which is the proper factor to ­convert from kB T e with T = 300  K (the units used by the PMEPot plugin) to volts. The third argument indicates the requested voltage drop to be added to the grid. Finally, the fourth argument is the output file (see Note 15 and Fig. 9). 6. Simulate a-hemolysin and DNA in an external electric field. Perform procedures analogous to steps 1–3 in the directory running-ahl+dna, using the ahl+dna.psf and ahl+dna.pdb files. Results of this simulation will allow the reduction in ionic current caused by the DNA to be computed. Current calculations will be discussed in Subheading 3.10. 3.8. Simulating a-Hemolysin Using Grid-Steered Molecular Dynamics

The grid-steered molecular dynamics (G-SMD) feature of NAMD allows a potential energy defined on a grid to be applied to select atoms. The G-SMD allows us to simulate permeation of DNA through a-hemolysin much faster than using conventional MD. Files for this section are located in the running-ahl+dna directory. 1. Create G-SMD file.  Using the equilibrated a-hemolysinDNA system just produced, we will next simulate the accelerated translocation of DNA through a-hemolysin using the G-SMD. It requires a PDB telling NAMD which atoms to

Modeling Nanopores for Sequencing DNA

347

Fig. 9. Example average electrostatic potential of a-hemolysin in an external electric field. White indicates low voltage while black indicates high voltage.

force, which we make now. Run the script make-targets. tcl in the VMD TkCon. 2. Enter electric field value.  Replace <efield> in grid.namd with the value calculated in Subheading 3.7 step 2. 3. Simulate a-hemolysin and DNA.  Examine the file grid. namd. For the electric field, use the value calculated for simulating a-hemolysin in an external field above. G-SMD is controlled with the following keywords: gridforce on gridforcefile ahl+dna_DNA.pdb gridforcecol B gridforcepotfile ../running–ahl/electric.dx gridforcevolts yes gridforcescale 0 0 10 gridforcecont1 yes gridforcecont2 yes gridforcecont3 yes gridforcevoff 0 0 –1.2

This simulation uses the electrostatic potential detemined previously (in Subheading 3.7 step 4), and applies the force derived from it to DNA only. Thus, DNA transport is accelerated, but with a realistic potential, thereby resulting in a realistic permeation event (10). The gridforcescale keyword sets

348

Comer, Wells, and Aksimentiev

Fig. 10. Translocation of ssDNA through a-hemolysin using G-SMD to accelerate the process.

factors by which to scale the x, y, and z components of the force calculated from the potential file. Thus, above we are only applying force in the z direction. We apply an electric field during this simulation as well, in order to achieve a realistic ion distribution. Therefore, the effective force on DNA in the z direction is 11 times normal, radically accelerating its translocation. An example trajectory is depicted in Fig.  10 (see Note 16). 3.9. Simulating a Synthetic Nanopore Under an External Electric Field

Electrically driven transport of DNA through a nanopore is an essential part of many nanopore sequencing experiments (4, 30–38). In this section, we describe how simulations of such experiments can be performed. Often the change in the current of ions through the pore caused by the presence of DNA is used to detect trans­ location events. Therefore, in this section, we perform simulations of the pore with and without DNA so that the difference in current between the two states can be estimated. Furthermore, we also simulate the translocation of DNA through the pore. Files for this section are located in the running–sin directory. 1. Scale the system size.  In subsequent simulations, we will be applying an external electric field. Such external forces can cause spurious behavior of the pressure control. Thus, we perform the production simulations at constant volume, with no pressure control applied. Here we compute the mean size of the system from the pressure controlled simulation to use in the constant volume simulation. The script scaleToMeanNptSize.tcl determines the mean system size from the xst file (ignoring data before 50  ps, after which time the system should have reached a steady state). It then creates an xsc file that can be used in NAMD to simulate the system at this average size. It also scales the final frame of the equilbration trajectory to the mean system size and writes a pdb which can be used as initial coordinates for the production

Modeling Nanopores for Sequencing DNA

349

simulations. The scaled_sin+dna_ions.cz.dat contains the system size versus timestep data. It can be plotted easily using any plotting software. Execute the script scaleToMeanNptSize.tcl in the VMD TkCon. 2. Simulate the open pore under a transmembrane voltage bias. The electric field applied by NAMD is chosen by E z = V Lz where V is the voltage drop across the membrane and Lz is the length of the system along the z axis. The lines located at the end of sin_20V.namd extract the system length from the xsc file created by scaleToMeanNptSize.tcl compute the electric field using the appropriate conversion factor (from V/Å to kcal/(mol Å e)) (see Notes 19 and 20). set inStream [open $xsc r] set lengthZ [lindex [lindex [split [read $inStream] \ \n] 2] 9] close $inStream eFieldOn yes eField 0.0 0.0 [expr 23.06054917 * $voltage / $lengthZ]

Run NAMD by entering the following in the Linux shell: namd2 sin_20V.namd > sin_20V.log &

3. Scale the system size.  Open the tcl script scaleTo MeanNptSize.tcl in a text editor. Change the line set sys sin to set sys sin+dna. Execute the script. 4. Simulate the pore with DNA under an applied voltage.  Run NAMD by entering the following in the Linux shell: namd2 sin+dna_20V.namd > sin+dna_20V.log &

You should observe complete translocation of the DNA within 1 ns as shown in Fig. 11a (see Note 19). 3.10. Analysis of Nanopore Simulations

Below we describe how one can compute a number quantities associated with transport of different species through the nanopore. First, we analyze the DNA motion by obtaining a trace of the number of DNA molecules that have passed through the pore. This analysis is done using our simulation of DNA translocation through the Si3N4 pore, but that could easily be done for a-hemolysin as well. Next, we compute the current of ions through the pore. Measurements of ionic current have long been used to detect the passage of molecules through nanopores (39), and futhermore have been used to assay the DNA and obtain sequence information (3, 4, 30–35, 38). Files for this section are located in the analysis directory. 1. Remove the water and the membrane from the trajectory.  For some analyses we are interested in the behavior of only some parts of the system. By removing the other parts of the system from the trajectory, we can significantly reduce the time

350

Comer, Wells, and Aksimentiev

required for loading data and performing calculations. To determine the translocation rate of the DNA or the ionic current through the pore, we need the positions of only the DNA and ions. The script removeWaterDcd.tcl contains a procedure called removeWater that can remove the water and atoms of the membrane from the trajectories created in Subheading 3.9. Enter the lines below in the VMD TkCon: source removeWaterDcd.tcl removeWater ../building-sin/sin_ions ../running-sin/\ sin_20V.dcd removeWater ../building-sin+dna/sin+dna_ions ../running-sin/\ sin+dna_20V.dcd

The resulting DCD files are written to nw_sin_20V.dcd and nw_sin+dna_20V.dcd in the current directory. The associated structure files, nw_sin_ions.psf and nw_ sin+dna_ions.psf, are also written. 2. Compute the number of nucleotides permeated.  The script permeationTraj.tcl contains a procedure to compute the number of nucleotides that have passed through the plane z = 0. The results allow us to determine the translocation rate and compare it with experimental results. The procedure compute in permeationTraj.tcl takes four arguments. The first specifies the simulation name, which is part of the name of the output files. The second argument is the prefix of the PSF and PDB files used for the simulation. The third argument is a list of DCD trajectory files. The final argument is the duration in femtoseconds between DCD frames (the value of the dcdFreq parameter in NAMD for a 1 fs timestep). In the VMD TkCon, enter the following: source permeationTraj.tcl compute sin+dna_20V nw_sin+dna_ions\     {nw_sin+dna_20V.dcd} 5000

The results are recorded in a the two-column file nucleotides_sin+dna_20V.dat. The first column contains the time in nanoseconds and the second contains the number of permeated nucleotides. A plot of the resulting data (nucleotides_sin+dna_20V.dat) should be similar to Fig. 11b. From the slope of the curve we can determine the translocation rate over portions of the trajectory. 3. Compute the ionic current.  In the steady state, the mean ionic current through any plane z = z0 must be the same for all z0 , If we compute the current over all ions in the system, the uncertainty of our current estimate will grow with the system size. To minimize the uncertainty in the current estimate, we compute the current only over the region within the pore. Because the membrane has a

Modeling Nanopores for Sequencing DNA 0 ns

0.12 ns

0.24 ns

0.48 ns

b

0.64 ns

nucleotides permeated

a

351

80 60 40 20 0 0

0.24 0.48 time (ns)

0.72

Fig. 11. (a) Snapshots of the translocation of dsDNA through an Si3N4 nanopore from the simulation in Subheading 3.9 step 4. An external voltage of 20 V is applied. (b) Number of nucleotides passing through the pore as a function of time in the simulation above. The data shown in the graph is calculated in Subheading 3.10 step 2.

thickness of ~100 Å, we compute the current over −ℓ/2 £ z £ ℓ/2, where ℓ = 90 Å. In the script below, we compute the ionic current by I (t + Dt / 2) =



1 N å qi [zi (t + Dt ) - zi (t )] Dt  i

(5)

where

ì zi (t ), zi (t ) £  / 2 ï zi (t ) = í - / 2, zi (t ) <  / 2 ï  / 2, z (t ) >  / 2 i î

(6)

and zi and qi are respectively the z coordinate and charge of ion i, N is the total number of ions, and Dt = 5 ps was the time between trajectory frames. By entering the following in the VMD TkCon, we can compute first the ionic current for the pore in the absence of DNA and then the ionic current while the DNA occupies the pore. Here, the compute procedure takes arguments similar to the last step, except that there are two additional arguments that define the z coordinates between which the current is computed (−ℓ/2 and ℓ/2). The output for each call to compute consists of three files, which respectively contain the current due to K+, the current due to Cl– and the total current, each as a function of time. The time and the current are recorded in nanoseconds and nanoamperes, respectively (see Fig. 12). source currentTraj.tcl compute sin_20V nw_sin_ions {nw_sin_20V.dcd} 5000 –45 45 compute sin+dna_20V nw_sin+dna_ions {nw_sin+dna_20V.dcd}\ 5000 –45 45

Comer, Wells, and Aksimentiev 100 open with DNA

current (nA)

352

50

0

0

0.4

0.8 time (ns)

1.2

Fig.  12. Total ionic current through the Si3N4 nanopore with and without DNA. The current traces for the open pore (in the absence of DNA) and pore containing DNA are shown by circles and triangles, respectively, which represent block averages over 0.1 ns intervals. The current and averages are computed in Subheading 3.10 steps 3 and 4. In all cases, the transmembrane bias voltage is 20 V. For the open pore, we see a sharp transient in the current at the beginning of the trajectory, which drops to a steady state value of 32.1±0.7 nA. The trace in the presence of DNA can be compared to the images and permeation plot in Fig. 11. After a small transient the current is below the open pore value for t < 0.3 ns. As the DNA exits the pore, there is a sharp rise in the current, which decays back to the open pore value.

4. Compute block averages.  The instanteous ionic current has large fluctuations due to thermal and shot noise. To obtain reliable current values that can be compared with experiment, we must average the instantaneous current over long trajectories. Furthermore, it can take several nanoseconds after the application of the field for the ionic current to reach a steady state. First, we compute current averages over 0.1 ns blocks. From these averages, we can determine at which time the current appears to reach a steady state and compute a mean current for all data gathered beyond that time. In the Linux terminal, enter the following command: tclsh blockAvgSe.tcl 20 curr_*.dat

From the total current data files curr_sin_20V.dat and curr_sin+dna_20V.dat the script yields curr_ sin_20V.dat.20.block and curr_sin+dna_20V. dat.20.block. These files contain three values for each 0.1 ns – the mean time, the mean current, and the standard error over the block. The standard error is computed under the assumption that each sampling of the current in our original data is independent. This standard error is a reliable measure of the uncertainty of a current value for many purposes; however, for some systems significant correlations between samples taken at 5 ps intervals can exist. 5. Compute the mean currents.  Plot the block averaged currents and estimate beyond what time the current fluctuates about a mean steady state value. To compute the mean open pore

Modeling Nanopores for Sequencing DNA

353

current, execute the command below in the Linux terminal, replacing <steadyTime> with the time beyond which you estimate that the current has reached a steady state. tclsh meanValueCut.tcl <steadyTime >curr_sin_20V.dat

6. Compute the electrostatic potential.  The script pmeTraj.tcl computes the electrostatic potential of the system using VMD’s pmepot command. The first four arguments given to compute are the same as those above. The last argument allows one to skip some frames of the trajectory. For example, the value 4 means that the electrostatic potential will be computed for every fourth frame of the trajectory – in this case for frames separated by 20 ps. Note that we cannot use the DCD files from which we have removed the water and membrane because the water and membrane make large contributions to the electrostatic field. Enter the following in the VMD TkCon: source pmeTraj.tcl cp ../building–sin/sin_eq.restart.xsc ../building-sin/sin_ions.xsc compute sin_20V ../building–sin/sin_ions \ {../running–sin/sin_20V.dcd} 5000 4

7. Add the external electric field to the potential map.  As in Subheading 3.7 step 5, we need to add the external electric field that was applied during the simulation to obtain the total electrostatic potential. ../grid/gridExternalField pot_sin_20V.dx 0.02585 20 \ pot_sin_20V_ext.dx

The resulting grid sin_20V_ext.dx can now be viewed in VMD.

4. Notes 1. In Subheading 2.3, the harmonic restraint values are given as Frestraint (r) = - 2k(r - r0 ) . The energies are therefore given 2 by V restraint = k r - r0 . For the bond constants given in Subheading 2.3 and implemented in silicon_nitride. par, the energy appears as Vbond = K (r - b)2 . Other features of NAMD 2.7 (notably the SMD feature) use a different convention for which a factor of 1/2 appears in the energy formulas. 2. In Subheading  3.1 step 4, we apply the patches DEO1 for pyrimidines and DEO2 for purines while generating the DNA structure with psfgen. Applying these patches is absolutely essential for creating DNA structures. Without them, the resulting structure is RNA, which behaves very differently.

354

Comer, Wells, and Aksimentiev

3. In Subheading  3.1 step 4, we make the segment name relatively long and descriptive so that it will not clash with any segment names later when we merge the DNA with other systems. 4. For convenience, we express ion concentrations in molality, i.e. moles of solute divided by the mass in kilograms of solvent. Molality is well defined for our computer models because we control the number of water molecules and the number of ions. Molarity, on the other hand, might be ambiguous because the density of the solution depends on the temperature, the ion concentration, and even the water model used. One should always check carefully that the ion concentrations in the system accurately represent those in experiments to which one is comparing. The finite size of simulated systems can cause many spurious effects in the concentration. For example, in small systems with low ion concentrations, binding of ions to the Si3N4 surface can considerably deplete the number of free ions in the system. The ion concentration even far from the membrane can be much less than the bulk concentration computed from the raw number of ions and water molecules. 5. When combining structures (as in Subheading 3.2 step 3) be sure that no parts of the systems to be combined have identical segment names. 6. Caution! It can take some time for ions added at random locations to approach their equilibrium distribution. In Subheading 3.2 step 10, we add ions in the solvent around a-hemolysin. The pore of the a-hemolysin is both narrow and charged; therefore, it can be particularly difficult for the ion distribution to equilibrate. Again in Subheading 3.5 step 6, we add ions randomly to neutralize the charge of the DNA inside the a-hemolysin. Here, the time required for the ions to approach their equilibrium (or steady-state, when an electric field is applied) distributions can also be quite long. 7. The basis vectors of the b-Si3N4 unit cell are e1 = (a 0 0) e 2 = (a 2 a 3 2 0) and e3 = (0 0 c) where a = 0.7595 nm and c = 0.2902  nm (40). Thus, the structure in Subheading  3.3 step 1 has the dimensions I1 = n1e1 I 2 = n 2e 2 I3 = n3e3 where n x , n y , and nz are the number of replications along each crystal axis. Choose n x , n y , and nz to produce your desired geometry. n x  = n y is required to transform from a parallelepiped to a hexagonal prism. 8. A hexagonal prism of material shown in Fig. 5b is equivalent to the parallelepiped produced by replications of the unit cell. Transforming the structure to a hexagonal prism is not entirely necessary, but can facilitate creation and visualization of the pore. To perform the transformation, map atoms to the periodic image of the structure that puts them nearest

Modeling Nanopores for Sequencing DNA

355

to the centroid of the original structure. The periodic images of the entire structure are defined by displacements d = i1n1e1 + i2n2e 2 + i3n3e3 where i1 , i2 , i3 are any integers. Inorganic Builder performs the transformation automatically when inorganicBuilder::newMaterialHexagonal Box is used. 9. When viewing the Si3N4 pore in current versions of VMD using any representation that displays bonds, some of the bonds will appear to crisscross the pore. This occurs because the system has bonds to its periodic images. 10. In Subheading 3.3 step 8, we shift the charges of the nitrogen atoms by q N = - N Siq Si N N , where N i and qi are the number and charge of each species, respectively. This formula typically leads to a relative change in the nitrogen charge <2%; therefore, should have a negligible effect on short range interactions between atoms of the system. Even after the application of the above formula, the total charge of the structure can still be on the order of the elementary charge because of the woefully limited precision of text-based structure files. A second stage of neutralization is required to completely neutralize the structure. We compute the total charge of the pore, and subtract this value from the charge of a single atom. We choose an atom that is not on the surface of the structure to minimize the effect of the charge adjustment. 11. If you wish to give a nonzero charge to your Si3N4 nanopore, you can modify the neutralization procedure in Subheading 3.3 step 8. Note, that it is often desirable for the charge of the simulated system to be zero. Thus, you should make the total charge of the Si3N4 as close as possible to an integer, so that the charge of the entire system can be neutralized by adding counterions to the solution. 12. When performing annealing using the BKS force field, as in Subheading  3.4, use of multiple time-stepping can lead to unstable dynamics. Therefore, the NAMD configuration files for these simulations contain the following lines: timestep

1

nonBondedFreq

1

fullElectFrequency

1

13. In Subheading 3.5 step 2, we constrain the DNA with a phantom pore to allow it to be placed inside the a-hemolysin. One can also use the shape of a-hemolysin itself as the basis for the phantom pore shape, rather than a simple cylinder (41). 14. Another option for placing DNA in a-hemolysin is to place DNA near the a-hemolysin vestibule and use the G-SMD method described in Subheading 3.8 to pull the DNA through the pore.

356

Comer, Wells, and Aksimentiev

15. In Subheading 3.7 step 1, we define restraints for the center of mass of the a-hemolysin. If we do not restrain the a-hemolysin using SMD, we would need to first align the trajectory frames so that the a-hemolysin center of mass did not move before computing the distribution of the electrostatic potential as in Subheading 3.7 step 4. 16. In Subheading 3.8 step 3, we use G-SMD to simulate translocation of ssDNA through a-hemolysin. Smaller values of the z acceleration factor result in less distortion of the DNA and may be desirable (10). 17. While running NAMD with gridforce using a nonorthogonal basis, you may encounter the error “Gridforce too long for periodic cell”. This problem seems to be due to roundoff error in the NAMD code. The best way to avoid this problem is to use the newest version of NAMD, in the problem no longer exists. A workaround is to run the program gridShave on the grid. For example, you can run the following in the Linux terminal: ../grid/gridShave 20 20 4 anneal_grid.dx anneal_grid.dx

This loads anneal_grid.dx, removes grid points along each direction, and writes the result back to anneal_ grid.dx. 18. The pore created in Subheading 3.4 step 9 has a somewhat greater diameter than the Si3N4 pore created in Subheading 3.3. The size of the pore is determined in Subheading 3.4 step 2. By changing the diameter in the command below from 24 to a smaller value a smaller pore can be obtained. ../grid/gridSourcePore grid_basis.txt 2 104.472 24 10 \ pore_points.txt

19. In Subheading 3.9 steps 2 and 4, we apply 20 V to drive ions and DNA through the pore. This value of the transmembrane bias is several times larger than the largest values used in experiments. You should use more realistic values of the transmembrane bias in your production simulations, although longer simulations will be required to obtain reliable current estimates and to observe DNA translocation. 20. The electrolyte used here is a sufficiently good conductor that nearly all the voltage applied to the system drops across the membrane. Hence, the electric field distribution within a given pore is defined only by the voltage applied transverse to the membrane and not by the distance between the electrodes applying the voltage. Experiments applying the same voltages but having different system dimensions can therefore be compared – as can simulations with system dimensions much smaller than those in experiments. In our simulations we define the voltage across the entire system by applying an electric field E z = V Lz where Lz is the length of the simulated

Modeling Nanopores for Sequencing DNA

357

system. However, for a system having an insufficient number of charge carriers, a portion of the voltage applied to the entire system would drop across the water above and below the membrane due to its finite dielectric constant. References 1. Branton, D., Deamer, D., Marziali, A., Bayley, H., Benner, S., Butler, T., Di Ventra, M., Garaj, S., Hibbs, A., Huang, X., et  al. (2008) The potential and challenges of nanopore sequencing Nature Biotech 26, 1146–1153. 2. Lagerqvist, J., Zwolak, M., and Ventra, M. D. (2006) Fast DNA Sequencing via Transverse Electronic Transport Nano Lett 6, 779–782. 3. Sigalov, G., Comer, J., Timp, G., and Aksimentiev, A. (2008) Detection of DNA sequence using an alternating electric field in a nanopore capacitor Nano Lett 8, 56–63. 4. Soni, G. V. and Meller, A. (2007) Progress toward ultrafast DNA sequencing using solidstate nanopores Clinical Chemistry 53, 1996–2001. 5. Bransburg-Zabary, S., Nachliel, E., and Gutman, M. (2002) A Fast in Silico Simulation of Ion Flux through the Large-Pore Channel Proteins Biophys J 83, 3001–3011. 6. Aksimentiev, A., Heng, J. B., Timp, G., and Schulten, K. (2004) Microscopic kinetics of DNA translocation through synthetic nanopores Biophys J 87, 2086–2097. 7. Aksimentiev, A. and Schulten, K. (2005) Imaging alpha-hemolysin with molecular dynamics: Ionic conductance, osmotic permeability and the electrostatic potential map Biophys J 88, 3745–3761. 8. Luan, B. and Aksimentiev, A. (2008) Electroosmotic screening of the dna charge in a nanopore Phys Rev E 78, 021912. 9. Dorvel, B., Sigalov, G., Zhao, Q., Comer, J., Dimitrov, V., Mirsaidov, U., Aksimentiev, A., and Timp, G. (2009) Analyzing the forces binding a restriction endonuclease to DNA using a synthetic nanopore Nucl Acids Res 37, 4170–4179. 10. Wells, D. B., Abramkina, V., and Aksimentiev, A. (2007) Exploring transmembrane transport through a-hemolysin with grid-steered molecular dynamics J Chem Phys 127, 125101. 11. VMD User Guide. URL: http://www.ks.uiuc. edu/Research/vmd/current/ug/ug.html. 12. VMD Tutorial. URL: http://www.ks.uiuc. edu/Training/Tutorials/vmd/tutorialhtml/index.html. 13. Phillips, J. C., Braun, R., Wang, W., Gumbart, J., Tajkhorshid, E., Villa, E., Chipot,

14. 15. 16. 17. 18.

19.

20.

21.

22.

23.

24.

25.

C., Skeel, R. D., Kale, L., and Schulten, K. (2005) Scalable molecular dynamics with NAMD J Comp Chem 26, 1781–1802. NAMD User Guide. URL: http://www. ks.uiuc.edu/Research/namd/current/ug/. NAMD Tutorial. URL: http://www.ks.uiuc. edu/Training/Tutorials/namd/namd­tutorial-unix-html/index.html. NAMD Tutorial. URL: http://www.ks.uiuc. edu/Training/Tutorials/namd/namdtutorial-­win-html/index.html. van Dijk, M. and Bonvin, A. M. J. J. (2009) 3D-DART: a DNA structure modelling server Nucl Acids Res 37, W235–W239. Comer, J., Dimitrov, V., Zhao, Q., Timp, G., and Aksimentiev, A. (2009) Microscopic mechanics of hairpin DNA translocation through synthetic nanopores Biophys  J 96, 593–608. Heng, J. B., Aksimentiev, A., Ho, C., Marks, P., Grinkova, Y. V., Sligar, S., Schulten, K., and Timp, G. (2006) The electromechanics of DNA in a synthetic nanopore Biophys  J 90, 1098–1106. Wendel, J. A. and Goddard, III, W. A. (1992) The Hessian biased force-field for silicon nitride ceramics: Predictions of the thermodynamic and mechanical properties for a- and b-Si3N4 J Chem Phys 97, 5048–5062. Cruz-Chu, E. R., Aksimentiev, A., and Schulten, K. (2006) Water-silica force field for simulating nanodevices J  Phys Chem  B 110, 21497–21508. Aksimentiev, A., Brunner, R., Cruz-Chu, E. R., Comer, J., and Schulten, K. (2009) Modeling transport through synthetic nanopores IEEE Nanotechnology Magazine 3, 20–28. van Beest, B. W. H., Kramer, G. J., and van Santen, R. A. (1990) Force fields for silicas and aluminophosphates based on ab initio calculations Phys Rev Lett 64, 1955–1958. Vollmayr, K., Kob, W., and Binder, K. (1996) Cooling-rate effects in amorphous silica: A computer-simulation study Phys Rev B 54, 15808–15827. Cruz-Chu, E. R., Aksimentiev, A., and Schulten, K. (2009) Ionic current rectification through silica nanopores J Phys Chem C 113, 1850–1862.

358

Comer, Wells, and Aksimentiev

26. Grubmüller, H., Heymann, B., and Tavan, P. (1996) Ligand binding: Molecular mechanics calculation of the streptavidin-biotin rupture force Science 271, 997–999. 27. Aksimentiev, A., Balabin, I. A., Fillingame, R. H., and Schulten, K. (2004) Insights into the molecular mechanism of rotation in the Fo sector of ATP synthase Biophys  J 86, 1332–1344. 28. Ho, C., Qiao, R., Chatterjee, A., Timp, R. J., Aluru, N. R., and Timp, G. (2005) Electrolytic transport through a synthetic nanometerdiameter pore Proc Natl Acad Sci USA 102, 10445–14450. 29. Kim, M., McNally, B., Murata, K., and Meller, A. (2007) Characteristics of solid-state nanometre pores fabricated using a transmission electron microscope Nanotechnology 18, 205302. 30. Akeson, M., Branton, D., Kasianowicz, J. J., Brandin, E., and Deamer, D. W. (1999) Microsecond time-scale discrimination among polycytidylic acid, polyadenylic acid, and polyuridylic acid as homopolymers or as segments within singe RNA molecules Biophys  J 77, 3227–3233. 31. Meller, A., Nivon, L., Brandin, E., Golovchenko, J., and Branton, D. (2000) Rapid nanopore discrimination between single polynucleotide molecules Proc Natl Acad Sci USA 97, 1079–1084. 32. Vercoutere, W., Winters-Hilt, S., Olsen, H., Deamer, D., Haussler, D., and Akeson, M. (2001) Rapid discrimination among individual DNA hairpin molecules at single-nucleotide resolution using an ion channel Nature Biotech 19, 248–252. 33. Vercoutere, W. A., Winters-Hilt, S., DeGuzman, V. S., Deamer, D., Ridino, S. E., Rodgers, J. T., Olsen, H. E., Marziali, A., and Akeson, M. (2003) Discrimination among individual Watson-Crick base pairs at the ter-

mini of single DNA hairpin molecules Nucl Acids Res 31, 1311–1318. 34. Nakane, J., Wiggin, M., and Marziali, A. (2004) A nanosensor for transmembrane capture and identification of single nucleic acid molecules Biophys J 87, 615–621. 35. Ashkenasy, N., Sánchez-Quesada, J., Bayley, H., and Ghadiri, M. R. (2005) Recognizing a single base in an individual DNA strand: A step toward DNA sequencing in nanopores Angew Chem Int Ed Engl 44, 1401–1404. 36. Gracheva, M. E., Xiong, A., Leburton, J.-P., Aksimentiev, A., Schulten, K., and Timp, G. (2006) Simulation of the electric response of DNA translocation through a semiconductor nanopore-capacitor Nanotechnology 17, 622–633. 37. Gracheva, M. E., Aksimentiev, A., and Leburton, J.-P. (2006) Electrical signatures of single-stranded DNA with single base mutations in a nanopore capacitor Nanotechnology 17, 3160–3165. 38. Cockroft, S., Chu, J., Amorin, M., and Ghadiri, M. (2008) A Single-Molecule Nanopore Device Detects DNA Polymerase Activity with Single-Nucleotide Resolution J Am Chem Soc 130, 818–820. 39. Kasianowicz, J. J., Brandin, E., Branton, D., and Deamer, D. W. (1996) Characterization of individual polynucleotide molecules using a membrane channel Proc Natl Acad Sci USA 93, 13770–13773. 40. Grün, R. (1979) The crystal structure of b-Si3N4; structural and stability considerations between a- and b-Si3N4 Acta  Cryst B35, 800–804. 41. Mathé, J., Aksimentiev, A., Nelson, D. R., Schulten, K., and Meller, A. (2005) Orientation discrimination of single stranded DNA inside the a-hemolysin membrane channel Proc Natl Acad Sci USA 102, 12377–12382.

Index A Agarose gel electrophoresis....................65, 71–72, 192, 229 Antibody coated microbeads...................259–260, 265–266 Atomic force microscope (AFM) of DNA................................ 3, 17, 26–30, 105–112, 190, 209–219, 235, 255 imaging.................................3, 28, 29, 44, 106, 176, 235 volume measurements............................................... 112

B B-DNA modeling........................................................... 321, 322 structure.................................................................... 321 bBSA Slide Treatment................................................... 274 Bottom-up assembly................................170, 187, 199, 291 BSA surface treatment....................................274, 278–279 Butyl-trichlorosilane modified glass....................... 155–156 B-Z transition.................................................80, 81, 83–89

C Cacodylate buffer..................................................81, 84, 86 CHARMM............................. 319–321, 324, 333, 337, 339 Chemical ligation of DNA..............................36, 38, 41–42 CircLigase...............................................152–154, 158, 160 Circular dichroism spectroscopy......................64, 70, 87, 88 Circular single-stranded DNA (ssDNA)....................... 2–6, 151–155, 158–160, 162, 164–166, 183, 188, 189, 211–213, 215, 217, 218, 306, 307, 311, 325, 326, 340, 346, 348, 356 Confocal microscope observation................................... 301 Contour length estimator........................237, 241, 242, 244 Coupling of DNA to microspheres.................155, 160–161

D Deposition of DNA on mica.................................. 239, 240 Dielectrophoresis (DEP).........................224, 226, 228–233 Dielectrophoretic (DEP) trapping...........224, 226, 229–233 Diphenylphenanthroline.................................................. 37 Dithiopyridine........................................................ 258, 261 DNA annealing................6, 16, 23–24, 31, 155, 162, 227, 296 binding on metal nanostructures.............................. 143

combing.....................................................50, 51, 54–56 concentration determination.....................14, 17, 18, 85, 98, 158, 216 conformations....................................... 6, 80, 81, 86, 89, 240, 257, 318, 340, 341 contour length...................................235–247, 306, 311 enzymatic labelling..........................................66, 74, 75 fluorescent labeling............ 226, 229, 292, 294, 296, 297 helix handedness................................................... 79–90 immobilization at surfaces......... 102, 179, 183, 292, 296 labelling by PCR...................................................... 296 metallization................................................... 49–51, 56 mineralization....................................................... 51, 56 nanografting..................................................... 209–219 nanopatches.......................................211, 215–217, 219 nanoswitch............................................................ 61–76 nanowires.......................................................... 115–139 origami preparation.......................................... 188, 190 overstretching................................................... 257, 258 persistence length............................................. 236, 242 polymerase I.............................. 121, 122, 124, 134, 307 strand annealing..................................................... 6, 14 strand titration............................................................ 23 triangles......................... 33–47, 193, 194, 306, 311, 352 DNA contour length, measurement of................... 235–247 DNA-directed immobilization....................................... 211 DNA-nanomachines.........................................2, 62, 66, 80 DNA nanostructure.................. 1–10, 13–31, 34, 45, 79–90, 93, 94, 96, 102, 117, 151–166, 187, 209–219, 291 design......................................................79, 80, 94, 318 preparation............................... 8, 34, 151–166, 188–192 DNA-protein conjugates.........................211, 213, 216, 217 DNA-protein coupling............................259, 260, 264–265 DNA-wrapping...............................................236, 237, 292 Double-helix structure........................................................ 1 Drosophila cell culture................................................ 65, 72 Drosophila larvae.............................................................. 72

E Electric field computation.............................................. 345 Electroelution of DNA................................................... 238 Electron beam lihography.......................187, 224, 225, 227 Electron beam writing.....................................170, 173–175

Giampaolo Zuccheri and Bruno Samorì (eds.), DNA Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 749, DOI 10.1007/978-1-61779-142-0, © Springer Science+Business Media, LLC 2011

359

DNA Nanotechnology 360  Index

  

Electrophoresis tracking dyes mobilities............................. 5 Exonuclease VII......................................................... 36, 47

F Flow cell realization........................................................ 275 Fluorescence microscopy..........................16–17, 24–26, 30, 66, 73–74, 179–180, 292 of DNA.............................................16–17, 24–26, 292 imaging buffer for..................................................... 279 Fluorescence resonance energy transfer (FRET)....... 62, 67, 71, 73, 74, 89, 273–289, 291–302 Fluorophore choice for DNA labelling........................... 295 Force-induced DNA melting..................306, 307, 311, 312 Force loading-rate.......................................................... 258

G G4-DNA................................. 116, 117, 123–125, 135–138 GENOgold.............................................142, 144–147, 149 Glass slide cleaning and preparation...................... 179, 277 Gold-nanoparticle modified G-wire...................... 141–149 Gold nanoparticles.........50, 65, 71, 141–149, 152, 187–196 concentration............................................................ 188 oligonucleotide coated...................................... 143, 152 oligonucleotide-coated Gold Nanoparticles, preparation of..................................................... 190 Gold/palladium nanodots................................170, 176–180 Gold surface passivation..................................142, 144–146 G-protein coated microparticles............................. 260, 265 G-quadruplex................................... 93–103, 106, 107, 117, 123–125, 135–138, 142 G-rich oligonucleotides...........................105, 108–110, 117 G-rich oligonucleotides, assembly of...............105, 107–110 G-rich sequences.................................................94, 96, 117 G-wire......................................... 93–102, 106–111, 141–149 alignment.................................................................... 96 applications of............................................................. 96 growth.............................................. 106, 107, 142, 144, 146, 149 immobilization on mica.................................... 143–147 immobilization on silicon..........................142, 144, 145 preparation.................................................110, 141–149

H Helical induction...................................................82, 83, 88 a-Hemolysin.......................... 318, 322, 324, 325, 327–331, 340–342, 345–349, 354–356 a-Hemolysin modeling..................................318, 322, 324, 325, 327–331, 340–342, 345–349, 354–356 Hexamethyldisilazane surface passivation........................ 50 Hexamethyldisilizane (HMDS)...........................50, 52–57, 189, 192–196 Higher level structural organization of DNA................. 292 HPLC purification of DNA......................... 63, 97, 98, 121, 122, 124, 125, 133–135, 137–139

I Intracellular pH measurement.......................................... 73 I-switch......................................... 62–64, 66–71, 73, 75, 76

K Klenow fragment........96, 120–122, 124, 132, 134, 136, 307

L Lambda phage DNA................................................ 52, 307 Lift-off lithography........................................................ 142 Lipid membrane modeling............................................. 329

M MALDI-TOF. See Matrix-assisted laser desorption/ ionization time-of-flight Matrix-assisted laser desorption/ionization time-of-flight...............................36, 39–40, 98, 102 Mechanical manipulation of single molecules................ 268 Metal deposition........................................49, 173–176, 182 Metallic nanowires..................................................... 49–57 Metal nanostructures................... 49–57, 170, 177, 183, 224 Metal surface functionalization.......................170, 172, 177 Mica cleavage..............................................143, 144, 146, 148 surface passivation............................................ 144–147 Microscope data analysis.............................................................. 275 flow cell construction........................................ 275, 278 slide cleaning............................................................ 299 Molecular dynamics (MD).............................259, 318, 319, 323–325, 331, 346–348 Molecular dynamics (MD), grid-steered................318, 323, 325, 346–348 Molecular handles.................................................. 255–270 Mononucleosomes.................................................. 291–302

N NAMD software.....................319, 320, 327, 329, 330, 335, 336, 338–340, 345, 346, 348–350, 353, 355, 356 Nanoarray fabrication..................................2, 210, 211, 217 Nanoarrays..................................................2, 210, 211, 217 Nanofabrication....................... 169–184, 212, 213, 216, 217 Nanopore................................................................ 317–357 DNA sequencing.............................................. 317–357 translocation data, analysis................................ 349, 350 Nucleosome exchange reconstitution..................... 297–298 Nucleosome core particle (NCP)............292–295, 297, 299 Nucleosome reconstitution......................292–294, 297–298 Nucleosomes.......................................................... 292–302

O Oligonucleotide purification by electrophoresis............. 2–3 On-chip oligonucleotide synthesis................................. 200

DNA Nanotechnology 361 Index     

Optical tweezers......................................255–270, 305–314 Optical tweezers, alignment and operation............ 309–310

P Palladium.................................................................. 50, 171 Palladium oxide................................................................ 50 Palladium oxide reduction................................................ 50 Phi29 DNA polymerase.................. 152, 153, 157, 164, 165 Phosphothioate modified oligonucleotides............. 155, 165 pH sensor................................................................... 62, 67 Piranha solution................................. 52, 53, 155, 156, 162, 166, 172, 173, 177, 201, 203 PNA...............................................................................79–90 chirality............................................................83, 84, 89 helix handedness, prediction of................................... 87 PNA:PNA duplexes.................................. 83, 84, 87, 88, 90 Poly(dC)...................115–117, 120–122, 132–135, 138, 139 Poly(dG).......................... 115–117, 120–122, 124, 132–139 Polyacrylamide gel electrophoresis.....................2–3, 15–16, 20–23, 35, 44, 97, 154, 158–160, 259, 264, 293, 298–299 Polyacrylamide gel electrophoresis, denaturing............. 2–3, 35, 154, 158–160 Polymethylmethacrylate mask........................................ 227 Protein-DNA interaction....................................... 183, 236 Protein unfolding....................................256, 258, 266, 268 Purification of DNA........................ 2–6, 10, 34–35, 38–40, 139, 188–190, 216, 238, 263, 292–293, 296–297

Q Quadruple helix.............................................................. 6, 7

R Relative DNA concentration, determination of............... 14 Rolling circle amplification.................................... 151–166

S Self-assembled monolayers (SAMs).......................169, 170, 177, 183, 210, 211, 213, 215–217, 219 Silanization of silicon oxide........................................ 50, 53 Silica nanopore modeling............................................... 203 Silicon......................................36, 45, 53, 54, 144, 145, 188, 189, 192, 195, 199–206, 212–214, 219, 224, 227, 276, 321, 332–334, 337, 339 alkylation of...................................................... 200, 204 hydrogen terminated..................................201–202, 204 porous........................................................201, 203–204 Silicon dioxide modeling................................................ 318 Silicon nitride modeling....................................................318, 324, 331 nanopore................................................................... 331 Silicon oxide........................................................... 189, 194

Single molecule fluorescence...........................273–289, 292 alternating laser excitation method............275, 283–287 data analysis.......................................275, 280–283, 287 Single-molecule force spectroscopy................................ 255 Single-molecule manipulation........................................ 257 Single-pair fluorescence resonance energy transfer (spFRET) microscopy........................................ 292 Single-stranded DNA modeling.............322, 325, 326, 341 Size exclusion chromatography (SEC)............65, 69, 71–72 Solid phase DNA synthesis.................................. 34–35, 38 SOLVATE software................................320, 326, 328, 329 Stigmation adjustment........................................... 174, 182 Streptavidin-coated beads............................................... 260 Structural DNA nanotechnology................................. 2, 33 Surface bound rolling circle amplification.............. 152, 153 Surface contact-angle, measurement................................ 54 Surface hybridization of oligonucleotides....................... 205 SYBR Green II staining.......... 153, 154, 157, 164–165, 264 Synthesis of polymeric proteins.......................260, 262–263 Synthetic nanopores................................318, 325, 348–349

T Thermal annealing of DNA nanostructures....16, 23–24, 31 Thiol-modified DNA..............................188, 228, 230, 231 Thiol modified oligonucleotides..................................... 216 Time-resolved fluorescence measurements........... 64–65, 71 TIRF microscopy. See Total internal reflection fluorescence (TIRF) microscopy Top-down and bottom-up merging................................ 170 Total internal reflection fluorescence (TIRF) microscopy............................. 30, 273–289, 292, 295 Triple helix preparation.................................................. 116 Triple-stranded structures.............................................. 116 Trolox.................................................................288, 294, 300

U Ultraflat gold substrates.......................................... 211–214 UV analysis of G wire................................................. 98–99 UV spectroscopy................................... 5, 15, 39, 40, 42, 84, 86, 87, 89, 99, 190

V VMD software............................... 319, 321, 323, 325–329, 331–335, 337–345, 347, 349–351, 353, 355

Y YOYO–1...............................................................17, 25, 31

Z Z-DNA induction.................................................................... 89 structure...................................................................... 88