Drug-DNA Interaction Protocols (Methods in Molecular Biology 613) (2nd Edition)

Methods

in

Molecular Biology™

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



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

Drug-DNA Interaction Protocols Second Edition

Edited by

Keith R. Fox School of Biological Sciences, University of Southampton, Southampton, UK

Editor Keith R. Fox School of Biological Sciences University of Southampton Southampton UK [email protected]

ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-60327-417-3 e-ISBN 978-1-60327-418-0 DOI 10.1007/978-1-60327-418-0 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2009940801 © Humana Press, a part of Springer Science+Business Media, LLC 1998, 2010 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or ­dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the date of going to press, ­neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Cover illustration: Background art is derived from Figure 6 in Chapter 9. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)

Preface DNA has been known to be the cellular target for many cytotoxic anticancer agents for several decades. The knowledge of its structure in atomic detail and the ease with which DNA fragments (both synthetic oligonucleotides and natural sequences) can be prepared and manipulated has aided the design of compounds that bind to it with improved selectivity. On the basis of this information, new generations of sequence reading compounds (including triplex forming oligonucleotides and minor groove binding ligands) have been prepared, which have the potential for targeting specific DNA sequences as anti-gene agents. Within the last 10 years, it has also become apparent that the familiar DNA duplex is not the only structure that can be targeted by DNA-binding ligands and there has been increased interest in triplex and quadruplex structures as drug targets, as well as proteinDNA complexes, such as those with nucleosomes or topoisomerases. Each of these advances has required the availability and development of an arsenal of techniques for probing the interactions in both qualitative and quantitative terms. This volume of Methods in Molecular Biology brings together several techniques that are currently useful for examining these interactions. Some of these are updates on ones that were included in the earlier volume (Methods in Molecular Biology 90), published 12 years ago, while others are new. Molecular science is a multidisciplinary enterprise, and while individuals and laboratories may become experts in a few techniques, a detailed description of DNA-ligand interactions requires a combination of approaches. This volume should therefore be useful for established workers who wish to broaden their experimental repertoire, as well as for those who are new to the field and need expert advice and guidance. The chapters have all been written by scientists who are experts in their own fields. They will obviously reflect their local preferences in experimental protocols, which can be modified to suit the requirements of the individual researcher. Each chapter begins with a short introduction, which outlines the background to the technique, the principles of its application, and the importance of the particular method. The most important part of each chapter is the methods section. These set out the experimental protocols in a stepby-step fashion and are accompanied by Notes sections which provide technical tips, based on experience, giving valuable information about potential problems and pitfalls and emphasizing the points at which special care is required. This volume should therefore be useful for post-graduates, post-doctoral workers, and established scientists, working in drug-DNA interactions. The chapters in this volume combine a wide range of approaches, from the cellular to the structural. The first nine chapters describe various biophysical techniques for quantifying drug-DNA interactions and for describing these in molecular and atomic detail, while the later chapters describe molecular and cellular approaches. Together these provide methods for assessing the strength and mode of binding, the sequence selectivity, and their effect on biological systems. Southampton, UK

Keith R. Fox

v

Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   1 Quantitative Analysis of Small Molecule–Nucleic Acid Interactions with a Biosensor Surface and Surface Plasmon Resonance Detection . . . . . . . . . . Yang Liu and W. David Wilson   2 Thermal Melting Studies of Ligand DNA Interactions . . . . . . . . . . . . . . . . . . . . . Aurore Guédin, Laurent Lacroix, and Jean-Louis Mergny   3 Circular and Linear Dichroism of Drug-DNA Systems . . . . . . . . . . . . . . . . . . . . . Alison Rodger   4 Drug Binding to DNA⋅RNA Hybrid Structures . . . . . . . . . . . . . . . . . . . . . . . . . . Richard T. Wheelhouse and Jonathan B. Chaires   5 Quantification of Binding Data Using Capillary Electrophoresis . . . . . . . . . . . . . . Fitsumbirhan Araya, Graham G. Skellern, and Roger D. Waigh   6 Determination of Equilibrium Association Constants of Ligand–DNA Complexes by Electrospray Mass Spectrometry . . . . . . . . . . . . . . . . . . . . . . . . . . Valérie Gabelica   7 Detection of Adriamycin-DNA Adducts by Accelerator Mass Spectrometry . . . . . Kate Coldwell, Suzanne M. Cutts, Ted J. Ognibene, Paul T. Henderson, and Don R. Phillips   8 Molecular Modelling Methods to Quantitate Drug-DNA Interactions . . . . . . . . . Hao Wang and Charles A. Laughton   9 Application of Anomalous Diffraction Methods to the Study of DNA and DNA-Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Derrick Watkins, Tinoush Moulaei, Seiji Komeda, and Loren Dean Williams 10 DNase I Footprinting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Antonia S. Cardew and Keith R. Fox 11 Methods to Characterize the Effect of DNA-Modifying Compounds on Nucleosomal DNA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vidya Subramanian, Robert M. Williams, Dale L. Boger, and Karolin Luger 12 REPSA: Combinatorial Approach for Identifying Preferred Drug–DNA Binding Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Michael W. Van Dyke 13 In vitro Transcription Assay for Resolution of Drug-DNA Interactions at Defined DNA Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Benny J. Evison, Don R. Phillips, and Suzanne M. Cutts 14 In vitro Footprinting of Promoter Regions Within Supercoiled Plasmid DNA . . . Daekyu Sun

vii

v ix 1 25 37 55 71

89 103

119

133

153

173

193

207 223

viii

Contents

15 Topoisomerase I-Mediated DNA Relaxation as a Tool to Study Intercalation of Small Molecules into Supercoiled DNA . . . . . . . . . . . . . . . . . . . . Paul Peixoto, Christian Bailly, and Marie-Hélène David-Cordonnier 16 A High-Throughput Assay for DNA Topoisomerases and Other Enzymes, Based on DNA Triplex Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Matthew R. Burrell, Nicolas P. Burton, and Anthony Maxwell 17 Measurement of DNA Interstrand Crosslinking in Individual Cells Using the Single Cell Gel Electrophoresis (Comet) Assay . . . . . . . . . . . . . . . . . . . Victoria J. Spanswick, Janet M. Hartley, and John A. Hartley 18 Measurement of DNA Interstrand Crosslinking in Naked DNA Using Gel-Based Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Konstantinos Kiakos, Janet M. Hartley, and John A. Hartley 19 An Evaluation Cascade for G-Quadruplex Telomere Targeting Agents in Human Cancer Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mekala Gunaratnam and Stephen Neidle

235

257

267

283

303

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315

Contributors Fitsumbirhan Araya  •  Institute of Pharmacy and Biomedical Sciences, University of Strathclyde, Glasgow, UK Christian Bailly  •  Jean-Pierre Aubert Research Center (JPARC), Institut de Recherches sur le Cancer de Lille, Lille, France IMPRT-IFR114, Lille, France Pierre Fabre Research Institute, Toulouse, France Dale L. Boger  •  Department of Chemistry, The Scripps Research Institute, La Jolla, CA, USA Matthew R. Burrell  •  Dept Biological Chemistry, John Innes Centre, Colney, Norwich, UK Nicolas P. Burton  •  Inspiralis Ltd., Norwich Bioincubator, Norwich Research Park, Norwich, UK Antonia S. Cardew  •  School of Biological Sciences, University of Southampton, Southampton, UK Jonathan B. Chaires  •  James Graham Brown Cancer Center, University of Louisville, Louisville, KY, USA Kate Coldwell  •  Department of Biochemistry, La Trobe University, Bundoora, VIC, Australia Suzanne M. Cutts  •  Department of Biochemistry, La Trobe University, Bundoora, VIC, Australia Marie-Helene David-Cordonnier  •  INSERM U-837, Jean-Pierre Aubert Research Center (JPARC), Institut de Recherches sur le Cancer de Lille, Lille, France IMPRT-IFR114, Lille, France Benny J. Evison  •  Department of Biochemistry, La Trobe University, Bundoora, VIC, Australia Keith R. Fox  •  School of Biological Sciences, University of Southampton, Southampton, UK Valérie Gabelica  •  Physical Chemistry and Mass Spectrometry Laboratory, Department of Chemistry, University of Liège, Belgium GIGA-Systems Biology and Chemical Biology, University of Liège, Belgium Aurore Guédin  •  Equipe Santé, Laboratoire ‘Régulation et Dynamique des Génomes’, Muséum National d’Histoire Naturelle USM 503, INSERM UR 565, CNRS UMR 5153, Paris, France Mekala Gunaratnam  •  The Cancer Research UK Biomolecular Structure Group, The School of Pharmacy, University of London, London, UK Janet M. Hartley  •  Cancer Research UK Drug-DNA Interactions Research Group, UCL Cancer Institute, University College London, London, UK John A. Hartley  •  Cancer Research UK Drug-DNA Interactions Research Group, UCL Cancer Institute, University College London, London, UK

ix

x

Contributors

Paul T. Henderson  •  Division of Hematology and Oncology, Department of Internal Medicine, UC Davis Cancer Center, University of California Davis Medical Center, Sacramento, CA, USA Konstantinos Kiakos  •  Cancer Research UK Drug-DNA Interactions Research Group, UCL Cancer Institute, Paul O’Gorman Building, University College London, London, UK Seiji Komeda  •  School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, GA, USA Laurent Lacroix  •  Equipe Santé, Laboratoire ‘Régulation et Dynamique des Génomes’, Muséum National d’Histoire Naturelle USM 503, INSERM UR 565, CNRS UMR 5153, Paris, France Charles A. Laughton  •  Centre for Biomolecular Sciences, School of Pharmacy, University of Nottingham, Nottingham, UK Yang Liu  •  Department of Chemistry, Georgia State University, Atlanta, GA, USA Karolin Luger  •  Department of Biochemistry and Molecular Biology, Colorado State University, Fort Collins, CO, USA Anthony Maxwell  •  Department of Biological Chemistry, John Innes Centre, Colney, Norwich, UK Jean-Louis Mergny  •  Equipe Santé, Laboratoire ‘Régulation et Dynamique des Génomes’, Muséum National d’Histoire Naturelle USM 503, INSERM UR 565, CNRS UMR 5153, Paris, France Tinoush Moulaei  •  School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, GA, USA Stephen Neidle  •  The Cancer Research UK Biomolecular Structure Group, The School of Pharmacy, University of London, London, UK Ted J. Ognibene  •  Chemistry, Materials, Earth and Life Sciences, Center for Accelerator Mass Spectrometry, Lawrence Livermore National Laboratory, 7000 East Avenue, L-452, Livermore, CA 94551, USA Paul Peixoto  •  INSERM U-837, Jean-Pierre Aubert Research Center (JPARC), Institut de Recherches sur le Cancer de Lille, Lille, France IMPRT-IFR114, Lille, France Don R. Phillips  •  Department of Biochemistry, La Trobe University, Bundoora, VIC, Australia Alison Rodger  •  Department of Chemistry, University of Warwick, Coventry, UK Graham G. Skellern  •  Institute of Pharmacy and Biomedical Sciences, University of Strathclyde, Glasgow, UK Victoria J. Spanswick  •  Cancer Research UK Drug-DNA Interactions Research Group, UCL Cancer Institute, Paul O’Gorman Building, University College London, London, UK Vidya Subramanian  •  Department of Biochemistry and Molecular Biology, Colorado State University, Fort Collins, CO, USA Daekyu Sun  •  Department of Pharmacology and Toxicology, College of Pharmacy, University of Arizona, BIO5 Institute, Tucson, AZ, USA Michael Van Dyke  •  Molecular & Cellular Oncology, M.D. Anderson Cancer Center, University of Texas, Houston, TX, USA

Contributors

Roger D. Waigh  •  Institute of Pharmacy and Biomedical Sciences, University of Strathclyde, Glasgow G4 0NR, UK Hao Wang  •  Department of Oral and Dental Science, University of Bristol, UK Derrick Watkins  •  School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, GA, USA Richard T. Wheelhouse  •  School of Pharmacy, University of Bradford, Bradford, West Yorkshire, UK Loren D. Williams  •  School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, GA, USA Robert M. Williams  •  Department of Chemistry, Colorado State University, Fort Collins, CO, USA The University of Colorado Cancer Center, Aurora, CO, USA W. David Wilson  •  Department of Chemistry, Georgia State University, Atlanta, GA, USA

xi

Chapter 1 Quantitative Analysis of Small Molecule–Nucleic Acid Interactions with a Biosensor Surface and Surface Plasmon Resonance Detection Yang Liu and W. David Wilson Abstract Surface plasmon resonance (SPR) technology with biosensor surfaces has become a widely-used tool for the study of nucleic acid interactions without any labeling requirements. The method provides simultaneous kinetic and equilibrium characterization of the interactions of biomolecules as well as small molecule-biopolymer binding. SPR monitors molecular interactions in real time and provides significant advantages over optical or calorimetic methods for systems with strong binding coupled to small spectroscopic signals and/or reaction heats. A detailed and practical guide for nucleic acid interaction analysis using SPR-biosensor methods is presented. Details of the SPR technology and basic fundamentals are described with recommendations on the preparation of the SPR instrument, sensor chips, and samples, as well as extensive information on experimental design, quantitative and qualitative data analysis and presentation. A specific example of the interaction of a minor-groove-binding agent with DNA is evaluated by both kinetic and steady-state SPR methods to illustrate the technique. Since the molecules that bind cooperatively to specific DNA sequences are attractive for many applications, a cooperative small molecule–DNA interaction is also presented. Key words: Biosensor, Surface plasmon resonance, Small molecule–nucleic acid interaction, Kinetics, Steady-state analysis, Cooperativity, Biacore, Minor groove

1. Introduction Biological systems function on a platform of complex and integrated biomolecular interactions (1–3). Signaling, transcription control of gene expression, and a host of other complex processes, frequently, are built on sets of sequential biomolecular interactions and subsequent reactions (4–8). These sequential processes that control and direct cellular functions can generally be understood

K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_1, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

1

2

Liu and Wilson

in terms of the interaction of a small number of molecules in each step. Appropriate assembly of the entire sequential array gives the beautifully intricate mechanism of cell function. The entire process can be understood and explained on the basis of thermodynamic theory that has been firmly established for over 100 years. The biomolecular associations involve macromolecular complexes as well as small molecule-macromolecule interactions that are essential for the control of cell processes. Most drugs, for example, are small molecules that must interact with a macromolecular receptor in order to affect the target cell function in a selective manner (5–12). The field of chemical biology is built around design and use of small molecules to control specific aspects of cell function through biopolymer interactions (13–15). In order to understand this intricate array of interactions and control systems, it is very informative to evaluate all of the specific sequential steps that can be isolated. This quantitative interaction information is essential to put available structural results into the appropriate context of cell function. To establish a basic quantitative characterization of the interactions, it is essential to determine a set of basic thermodynamic quantities (16–21). Two questions must be answered at the start of the investigations: (1) what do we want to know about the interactions and (2) how can we accurately and with a reasonable effort determine the desired information? In the optimum case, the binding affinity (the equilibrium constant, K, and Gibbs energy of binding, DG), stoichiometry (n, the number of compounds bound to the biopolymer), cooperative effects in binding, and binding kinetics (the rate constants, k, that define the dynamics of the interaction) should be determined. These fundamental parameters are the keys to a molecular understanding of the interactions and how they affect cellular functions. To determine these parameters, an accurate method of determining the concentration of each component that is not bound and the concentration of their complex is required. The information must be determined as a function of concentration at equilibrium for accurate K and n, and as a function of time for k. For each system then, the question becomes one of how to accurately determine the necessary concentrations as a function of time, reactant total concentrations, solution conditions, temperature, etc. The interacting systems can range from as little as two molecules to as many as required to form the final complex. As described above, a more complete understanding of the biomolecular complexes requires determination of additional thermodynamic quantities that characterize complex formation in detail. Because biological molecules have widely different molecular characteristics that can and generally do change on complex formation, it can be difficult to find methods to evaluate the full array of interactions under an appropriate variety of conditions.

Quantitative Analysis of Small Molecule–Nucleic Acid Interactions

3

This is the “reasonable effort” part of question (2) above. For many biologically important interactions, binding is very strong, and experiments must be conducted at very low concentrations, down to the nanomolar or lower levels. In a binding experiment, the compound concentration should vary from below to above the Kd in order to obtain accurate binding results, and this requires significant concentrations of both free and complexed molecules (16, 22–24). In many methods, for example, those that involve some separation of free and bound forms such as dialysis, the unbound concentration has to be determined, and if this is below around 100 nM, as expected for the large K values generally observed in biological systems and as generally required for effective drug molecules, measurement can be difficult or impossible. These concentrations fall below the detection limit for many systems and may require special methods, such as radiolabels or fluorescent probes, for added sensitivity in detection. The labels may significantly increase the sensitivity of detection; however, they may also perturb the interaction that is being investigated. An attractive alternative method, which is operational down to very low concentrations, is the use of biosensors with surface plasmon resonance (SPR) detection (17, 25–30). SPR responds to the refractive index or mass changes at the biospecific sensor surface on complex formation (27–32). Since the SPR signal responds directly to the amount of bound compound in real time, as versus indirect signals at equilibrium that are obtained for many physical measurements, it provides a very powerful method to study biomolecular interaction thermodynamics and kinetics. Use of the SPR signal and direct mass response to monitor biomolecular reactions also removes many difficulties with labeling or characterizing the diverse properties of biomolecules (25–31). To illustrate the power of the SPR method, small organic cation interactions with specific DNA sequences will be used as examples. Small molecule targeting of DNA has applications in therapeutics, from cancer treatment to antiparasitic applications, and in biotechnology, and development of such molecules for control of nucleic acid function is at the heart of chemical biology (5–10, 16, 25, 29, 32–47). The interaction of nucleic acids with small molecules has been a primary area of interest and research since before the discovery of the double helical structure of DNA. There are many successful anticancer agents that intercalate or alkylate DNA, and they can have quite varied structure and properties (34–37). Dicationic minor groove binding heterocycles which target eukaryotic organisms that cause parasitic diseases, such as malaria and sleeping sickness, have also been known for many years and are used in humans and animals (9, 10, 38–40). Therapeutic targeting of nucleic acids recently entered a new and very exciting area with the discovery that four-stranded, quadruplex

4

Liu and Wilson

DNA structures can occur in important cellular DNA regions from chromosomal telomeres to oncogene promoters (34, 35, 41–46). Several studies have clearly shown that interactions of a range of small molecules with quadruplexes can yield anticancer activity. Although this discussion focuses on DNA, essentially all of the methods can be applied to RNA interactions. Additional treatment of the biosensor surfaces may be required when using RNA, however, due to possible hydrolysis of RNA by a number of agents. Applications with small molecules can be very challenging, in terms of signal obtained on binding in a biosensor-SPR experiment. The larger the bound molecule, the larger the SPR signal and macromolecule binding can give large signal to noise ratios (27, 29, 31). We will show, however, that the biosensor-SPR method with a macromolecular sensor surface can be used under appropriate conditions to investigate small molecule binding with current state-of-the-art instruments. Because of the varied properties of the compounds, it is difficult and time-consuming to find other suitable methods that can quantitatively define their interactions with DNA (or RNA) under a variety of conditions. As described above, in biosensor-SPR instruments and experiments, it is the bound compound on the surface that is detected by the change in plasmon resonance angle, and this can be determined very accurately from a low to a high fraction of surface binding sites covered. The unbound or free solution compound concentration is not measured but is simply prepared by dilution and is in a constant concentration in the solution that flows over the sensor surface. Since the SPR signal responds to the refractive index or mass changes at the biospecific sensor surface on complex formation, no specific labeling of detection ability for the compounds is required (25–27, 31). For compounds that have molecular weights of approximately 300 or more, the SPR technology provides a very attractive method to study interactions with nucleic acids or proteins immobilized to form a biospecific target surface. 1.1. Fundamentals of Molecular Interactions by Biosensor-SPR Methods

For a single compound (C), binding to n equivalent sites on DNA (or other biopolymer) to give a complex (DNA·nCbound), the interaction process is described by kinetic and equilibrium equations as follows: DNA (site) + nCfree

ka kd

DNA.nCbound Ka =ka/kd = 1/ Kd  (1)

More complex models with nonequivalent sites are also well-known and are treated in a similar manner with more complex functions (27) (and described below under Subheading  3.2). Although a discussion of complex data fitting is beyond the

Quantitative Analysis of Small Molecule–Nucleic Acid Interactions

5

scope of this article, the topic is covered in most biophysical chemistry texts, and more complex models are included in commercial SPR instrument software packages as well as in software available from Myszka and coworkers (http://www. cores.utah.edu/interaction/software.html). In the most common case, it is of interest to study a variety of compounds for binding to a limited number of different DNAs. For this case, however, it is more convenient and efficient to immobilize, the DNAs on the sensor surface to monitor complex formation. Biacore T100 and 2000/3000 instruments have sensor chips with four channels such that three DNAs can be immobilized, with one flow cell left blank as a control for bulk refractive index subtraction. With a common sensor surface that has covalently attached streptavidin, a nucleic acid strand with biotin linked to either the 5¢ or 3¢ terminus can be captured to create the biospecific surface. The terminal attachment of biotin, through a flexible linker, leaves the nucleic acid binding sites open for complex formation. If streptavidin or the biotin linker creates problems with the interaction or nonspecific compound interactions, the nucleic acid can be synthesized with a terminal alkyl amine and the amine can be used to form a direct covalent amide bond with the surface through activation of carboxyl groups on the sensor surface. In Biacore instruments, which have been most widely used to date in the area of biosensor-SPR experiments, any molecule with free amino groups can be immobilized through amide bonds with activated carboxyl groups from sensor chip surface-linked carboxymethyl (CM) dextran. A range of other sensor chips surfaces and immobilization chemistries are also available and it is generally possible to find an appropriate surface for any biological interaction application (for Biacore instruments, see the web http://www.biacore.com/lifesciences/products/ sensor_chips/guide/index.html). The results from a biosensor-SPR experiment are typically presented as a set of sensorgrams, which plot a function that is directly related to the SPR angle versus time as shown in Fig. 1. The angle change is reported in Biacore technology as resonance units (RU) where a 1,000 RU response is equivalent to a change in surface concentration of about 1 ng/mm2 of protein or DNA (the relationship between RU and ng of material bound will vary with the refractive index of the bound molecule) (27, 29, 31). With a DNA sequence immobilized on the chip surface, a compound solution is injected and as the solution flows over the surface, compound binding to the DNA is monitored by a change in SPR angle (as RU). After a selected time, buffer flow is restarted and dissociation of the complex can be monitored for an additional selected time period (Fig.  1). Note that when the association phase collection time is long enough, a steady state plateau is reached such that the rate of binding of the small

6

Liu and Wilson

Fig. 1. Representative SPR sensorgrams for the interaction of DB293 with AATT and ATGA oligomer hairpin duplexes. The DB293 concentrations from bottom to top are 0–1 mM

molecule equals the rate of dissociation of the complex and no change of signal with time is observed. The response of interest is the difference between the response in cells with immobilized DNA minus the response of the blank flow cell without DNA. If the added molecule does not bind to a target/receptor at the surface, the SPR angle change in the sample and reference flow cells will be the same in a properly functioning instrument, and the signals, after subtraction, give a zero net RU response that is indicative of no binding. In the case in which binding does occur, an extra amount, relative to the blank surface, of the added molecule is bound at the sensor surface, and an additional SPR angle change is generated in the sample flow cell. Again, the amount of unbound compound in the flow solution is the same in the sample and reference flow cells, and is subtracted so that only the bound molecule generates a positive SPR signal. The concentration of the unbound molecule is constant and is fixed by the concentration in the flow solution. Both the association and dissociation phases of the sensorgram can be simultaneously fit to a desired binding model in several sensorgrams at different concentrations with a global fit routine (27–32). Global fitting allows the most accurate determination of the kinetics constants as well as calculation of the equilibrium constant, Ka, from the ratio of kinetic constants (Eq. 1). It is also possible to determine Ka independently of rate constants by fitting the steady-state response versus the concentration of the binding molecule in the flow solution over a range of concentrations. For the binding process in Eq. (1) the steady-state RU

Quantitative Analysis of Small Molecule–Nucleic Acid Interactions

7

at each compound concentration is determined and Ka can be obtained by fitting to the following equation:

RU = (RUmax • n • Ka • Cf)/(1 + Ka • Cf)

(2)

where RUmax is the maximum change in RU for binding to a single DNA site. It can be calculated, determined experimentally at high compound concentrations or used in the above equation as a fitting parameter such that Ka, n and RUmax are determined by fitting RU versus C f. Note that the common term “r”, the moles of compound bound per mole of DNA, is equal to RU/RU max. At high C f,where all binding sites are filled with compound, RU = RUmax • n. The refractive index change in SPR experiments generates essentially the same response for each bound molecule and can provide a direct determination of the stoichiometry, n, in Eqs. 1 and 2, provided that the amount of immobilized nucleic acid is known (29, 31). If the complex dissociates slowly, the surface can be regenerated before complete dissociation has occurred by a solution that causes rapid dissociation of the complex without irreversible damage to the immobilized DNA (29, 47). For example, a solution at low or high pH can unfold DNA and cause complex dissociation. It should be noted, however, that accurate fitting of the dissociation part of the curve requires collection of as much of the total dissociation signal as possible. Additional injections of buffer at pH near 7 allow the DNA to refold for the next binding experiment. If the immobilized DNA is composed of separate strands, the duplex must be reformed by hybridization. After the dissociation/regeneration phase is over and a stable baseline is reestablished, a second concentration sample can be injected to generate a second sensorgram. This process can be repeated with as many concentrations as needed to obtain a broad coverage of the fraction of compound bound to the nucleic acid site or sites (see Fig.  1). The kinetic and equilibrium constants that describe the reaction in Eq. 1.1 are obtained by global fitting the sensorgrams with equations from a kinetic model or by fitting steady-state RU versus concentration plots to an appropriate binding model. The models are the same for all types of binding experiments and are not unique to SPR methods. It should be emphasized that to obtain accurate kinetic information for a binding reaction, it is essential that the kinetics for transfer of the binding molecule to the surface immobilized nucleic acid (mass transfer) be faster than the binding reaction. Equilibrium information can be obtained, however, by fitting sensorgrams even when mass transfer limitations prevent accurate determination of kinetics (48). It should be emphasized that, when properly conducted, biosensor-SPR kinetic and equilibrium results are in excellent agreement with other methods (26, 27, 29, 30, 48–50).

8

Liu and Wilson

2. Materials 2.1. Required Materials for Instrument Cleaning

These materials are for the Biacore T100, 3000 and 2000 research instruments but similar materials are required for other instruments. 1. Maintenance chip with a glass flow cell surface. 2. 0.5% SDS (Biacore desorb solution 1). 3. 50 mM glycine pH 9.5 (Biacore desorb solution 2) (see Note 1). 4. 1% (v/v) acetic acid solution. 5. 0.2 M sodium bicarbonate solution. 6. 6 M guanidine HCl solution. 7. 10 mM HCl solution. (see Note 2)

2.2. Running Buffer for Immobilization of DNA

1. HBS-EP buffer: 10 mM HEPES pH 7.4, 150 mM NaCl, 3 mM EDTA, 0.005%, v/v polysorbate 20. (GE Healthcare Inc.) 2. HBS-N buffer: 10  mM HEPES pH 7.4, 150  mM NaCl. (GE Healthcare Inc.) 3. Filter and degas all solution quite thoroughly. 4. It should be emphasized that the internal flow system of the instrument has microcapillaries that can be damaged by particulate matter in any solution.

2.3. Sensor chip Preparation for DNA Immobilization: CM5 or CM4 Chip

1. A CM5 or CM4 sensor chip that has been at room temperature for at least 30  min (all sensor chips are available from GE Healthcare Inc.). 2. 100 mM N-hydroxysuccinimide (NHS) freshly prepared in water. 3. 400  mM N-ethyl-N¢-(dimethylaminopropyl) carbodiimide (EDC) freshly prepared in water. 4. 10 mM acetate buffer pH 4.5 (immobilization buffer). 5. 200–400 mg/ml streptavidin in immobilization buffer. 6. Amino-labeled nucleic acid solutions (~25 nM of single strand or hairpin dissolved in HBS-EP buffer). (5¢- end modified DNA obtained from Integrated DNA Technologies, Coralville, IA) 7. 1 M ethanolamine hydrochloride in water pH 8.5 (deactivation solution). 8. Dock the CM4 or CM5 chip, Prime with running buffer. Start a sensorgram in all flow cells with a flow rate of 5 µl/min. “Dock” and “Prime” are Biacore software commands that instruct the instruments to carry out specific operations. The commands and operations are listed in Table 1. 9. With NHS (100 mM) in one vial and EDC (400 mM) in other, use the “Dilute” command to make a 1:1 mixture of NHS/EDC.

Quantitative Analysis of Small Molecule–Nucleic Acid Interactions

9

Table 1 Biacore instrument commands Software commands

Function

Dock

Docks the sensor chip

Undock

Undocks the sensor chip

Prime

Flushes the flow system with running buffer

Dilute

Diluting samples with buffer or for preparing a defined mixture of two samples

Manual inject

Manually controlled injection

Desorb

Removes adsorbed samples from the autosampler and IFC using SDS and glycine

Sanitize

Cleans pumps, IFC and autosampler from micro-organisms using BIA disinfectant solution

10. Inject NHS/EDC for 10  min (50 µl) to activate the carboxymethyl surface to reactive esters. 11. Using “Manual Inject” with a flow rate of 5 µl/min, load the loop with ~100 µl of streptavidin in the appropriate buffer and inject streptavidin over all flow cells. Track the number of RUs immobilized, which is available in real time readout, and stop the injection after the desired level is reached (typically 2,500–3,000 RU for CM5 chip and 1,000–1,500 RU for CM4 chip). 12. Inject ethanolamine hydrochloride for 10  min (50 µl) to deactivate any remaining reactive esters. 13. Prime several times to ensure surface stability. 14. DNA is immobilized as described under Subheading 2.4. 15. To reduce the nonspecific binding to immobilized streptavidin by small molecules, covalent immobilization using aminolabeled DNA could also be used. In this way, DNA can be directly captured on the EDC/NHS activated carboxymethyl dextran surface of the sensor chip (from step 16 to step 19). 16. Start from step 8 to step 10 to activate the sensor chip surface. Then, start a new sensorgram with a flow rate of 2 µl/ min and select one desired flow cell on which to immobilize the nucleic acid. 17. Use Manual Inject, load the injection loop with ~100 µl of a 25 nM nucleic acid solution and inject over the low cell. Track the number of RUs immobilized and stop the injection after a desired level is reached (see Note 3).

10

Liu and Wilson

18. Inject ethanolamine hydrochloride for 10  min (50 µl) to deactivate any remaining reactive esters. 19. Prime several times to ensure surface stability. 2.4. Sensor chip Preparation for DNA Immobilization: SA Chip

1. A streptavidin-coated sensor chip (SA chip or prepared as outlined above) that has been at room temperature for at least 30 min. 2. HBS-EP buffer is used as running buffer. 3. Activation buffer: 1 M NaCl, 50 mM NaOH. 4. Biotin-labeled nucleic acid solutions (~25 nM of single-strand or hairpin dissolved in HBS-EP buffer). 5. Dock a streptavidin-coated chip and start a sensorgram with a 20 µl/min flow rate. 6. Inject activation buffer: 1 M NaCl, 50 mM NaOH for 1 min (20 µl) five to seven times to remove any unbound streptavidin from the sensor chip. 7. Allow buffer to flow at least 5 min before immobilizing the nucleic acids. 8. Start a new sensorgram with a flow rate of 2 µl/min and select one desired flow cell on which to immobilize the nucleic acid. Take care not to immobilize nucleic acid on the flow cell chosen as the control flow cell. Generally, flow cell 1 (fc1) is used as a control and is left blank for subtraction. It is often desirable to immobilize different nucleic acids on the remaining three flow cells (see Note 4). 9. Wait for the baseline to stabilize which usually takes a few minutes. Use Manual Inject, load the injection loop with ~100 µl of a 25 nM nucleic acid solution and inject over the low cell. Track the number of RUs immobilized and stop the injection after a desired level is reached (see Note 3). 10. At the end of the injection and after the baseline has stabilized, use the instrument crosshair to determine the RUs of nucleic acid immobilized and record this amount. The amount of nucleic acid immobilized is required to determine the theoretical moles of small molecule binding sites for the flow cell. 11. Repeat steps 4–6 for another flow cell (e.g., fc3 or fc4) (see Note 5).

2.5. Sensor chip Preparation for DNA Immobilization: HPA Chip

1. A HPA sensor chip that has been at room temperature for at least 30 min (see Note 6). 2. HBS-N buffer: 10  mM HEPES pH 7.4, 150  mM NaCl is used as running buffer (see Note 7). 3. 40  mM n-octyl glucoside (Sigma Chemical Co.) water solution.

Quantitative Analysis of Small Molecule–Nucleic Acid Interactions

11

4. 100 mM NaOH water solution. 5. Cholesterol conjugated nucleic acid solutions (~25  nM of hairpin dissolved in HBS-N buffer). (3¢- end conjugated DNA obtained from Integrated DNA Technologies, Coralville, IA or another source) 6. 0.1 mg/ml bovine serum albumin (BSA) prepared in HBS-N running buffer. 7. Before starting, the instrument must be kept scrupulously clean by running “Desorb” followed by “Sanitize”, then run on “Standby” or a low continuous flow rate overnight with distilled water. Switch to running buffer for the experiment. Make sure all solutions used are properly filtered and thoroughly degassed. 8. Liposomes for adsorption on HPA chips should be prepared in running buffer, using standard liposome preparation techniques (51). A liposome concentration of 0.5 mM, with respect to phospholipids, is usually sufficient. 9. Choose an appropriate experimental temperature based on the phase transition temperature (Tc) of liposome to be prepared. At temperatures below Tc, adsorption may be slow and the lipids may not form a monolayer on the surface (see Note 8). 10. Dock an HPA chip and start a sensorgram with a 20 µl/min flow rate using HBS-N running buffer. The chip is precleaned and conditioned by injecting twice with 40 mM n-octyl glucoside for 5 min every time. 11. Start a new sensorgram with a flow rate of 2 µl/min and select one desired flow cell on which to immobilize liposome preparation. Inject 60–180 µl liposome preparation (~0.5  mM), depending on lipid composition, liposome size and experimental temperature. Adsorption is seen as a steady increase in response, which flattens out as the surface coverage approaches completion. Typically, the maximum responses reached are in the region of 1,500–2,000 RU. 12. Inject 10 µl 100 mM NaOH water solution to remove loosely bound structures such as partially fused liposomes and multilayered structure. 13. Start a new sensorgram with a flow rate of 2 µl/min and select one desired flow cell (fc2–fc4) on which to immobilize cholesterol conjugated DNA hairpin. Use Manual Inject, load the injection loop with ~100 µl of a 25 nM nucleic acid solution and inject over the low cell. Track the number of RUs immobilized and stop the injection after a desired level is reached (see Note 3).

12

Liu and Wilson

14. Inject 10–50 µl 0.1 mg/ml BSA or another inert protein to block any exposed hydrophobic area on the sensor chip. 15. Prime several times to ensure surface stability. 2.6. Flow Solutions: Buffers and Samples (see Note 9)

1. HBS-EP buffer:10  mM HEPES pH 7.4, 150  mM NaCl, 3 mM EDTA, 0.005%, v/v polysorbate 20 (GE Healthcare Inc.). 2. MES10 buffer: 10 mM MES [2-(N-morpholino) ethanesulfonic acid] pH 6.25, 100 mM NaCl, 1 mM EDTA, 0.005%, v/v polysorbate 20. 3. CCL10 buffer: 10  mM CCL [cacodylic acid] pH 6.25, 100 mM NaCl, 1 mM EDTA, 0.005%, v/v polysorbate 20. 4. Tris10 buffer: 10  mM Tris-HCl pH 7.4, 100  mM NaCl, 1 mM EDTA, 0.005%, v/v polysorbate 20. 5. Regeneration solution: 10  mM Glycine-HCl (pH 2.5) (see Note 10)

3. Methods The Biacore software supplied with the instruments allows users to write a method or to use a software wizard to set up experiments. Several important factors, such as flow rate, association time and dissociation time, injection order and surface regeneration, must be considered in setting up experiments. A sample method used to collect small molecule binding results on nucleic acid surfaces is shown below. The structure of the compound (DB293) and the biotin-labeled DNA sequences (ATGA, AATT, ATAT) used in this example are shown in Fig. 2. 3.1. Data Collection and Processing

1. A Biacore T100 instrument (GE Healthcare Inc.) is used in this study. 2. Three biotin-labeled DNA hairpins are immobilized in different flow cells of a SA chip as described in Subheading  2.4. Approximately the same moles of each DNA oligomer a immobilized on the surface of these flow cells, so that the sensorgram saturation levels can be compared directly for stoichiometry differences. 3. Serial dilutions (concentration range is from 1 nM to 1 µM) of DB293 compound are prepared with the running buffer as the diluent to minimize the effect of bulk refractive index changes. (see Notes 11 and 12) 4. The flow rate is set to 25 µl/min (see Note 13).

Quantitative Analysis of Small Molecule–Nucleic Acid Interactions

13

Fig. 2. Structure of DB293 and sequences of 5¢-biotin-labeled DNA hairpins

5. A wait period of 5 min is used with running buffer flowing at the beginning of each concentration injection cycle to give a very stable baseline that is essential for accurate small molecule binding analysis. Several buffer samples are injected at the start of each experiment and these indicate whether the instrument is performing within specifications as well as serving as controls for data processing. (see Notes 14 and 15) 6. Inject 250 µl of each concentration of the compound solutions and set 600  s as the dissociation time (see Note 16). Inject samples from low to high concentration to eliminate the artifacts in the data from adsorption carry over on the instrument flow system (see Note 17 and 18). 7. At the end of the dissociation phase, inject two short pulses of regeneration solution (see Note 19), followed by a Mix command with excess volume of buffer and two 1-min injections of running buffer. 8. At the end of each cycle, 5 min waiting with running buffer flowing is also set to ensure that the chip surface is reequilibrated for binding (i.e., the dextran matrix is reequilibrated with running buffer) and the baseline has stabilized before the next sample injection. 9. After the data are collected, open the experimental sensorgrams in the BIAevaluation software for processing (see Note 20). First, zero the sensorgrams on the y-axis (RU) to allow the responses of each flow cell to be compared. Generally, the

14

Liu and Wilson

average of a stable time region of the sensorgram, prior to sample injection, should be selected and set to zero for each sensorgram. Then, zero on the x-axis (time) to align the beginnings of the injections with respect to each other. 10. Subtract the control flow cell (fc1) sensorgram from the reaction flow cell sensorgrams (i.e. fc2−fc1, fc3−fc1, and fc4−fc1). This removes any bulk shift contribution to the change in RUs. 11. Subtract a buffer injection, or better, an average of several buffer injections from the compound injections (different concentrations) on the same reaction flow cell (see Note 21). This is known as double subtraction and removes any flow cell specific baseline irregularities (27). At this point, the data should be of optimum quality and are ready for analysis as shown below. 3.2. Data Analysis

1. After the data are processed as described, kinetic and/or steady-state analysis is performed. Both kinetic and steadystate fitting can be done in Biacore software or in other available software packages (27, 29). As can be seen in Fig.  1, DB293 binding results reach a steady-state plateau in the injection period so that both kinetic and steady analysis can be used. In this case, the binding rate is not limited by mass transfer, and the association and dissociation rate constants can be determined. The average of the data over a selected time period in the steady-state region of each sensorgram can be obtained, converted to r (r = RU/RUmax), and plotted as a function of compound concentration in the flow solution (see Note 22). 2. Equilibrium constants can be obtained by fitting the results to the equivalent site model in Eq. 1.1. For two nonequivalent sites, the following equation can be used:



2

r = (K1 Cfree + 2 K1K2 C free)/(1 + K1Cfree + K1 K2Cfree2) (3) where K1 and K2 are the macroscopic thermodynamic binding constants (for a single site K2 = 0) and Cfree is the constant concentration of the compound in the flow solution (see Note 23). As described above, the binding stoichiometry can also be obtained directly from comparing the maximum response with the predicted response per compound. 3. Since in this example, equal moles of DNA hairpin duplexes were immobilized, the difference in maximum responses among the sets of sensorgrams is readily seen and directly reflects the difference in binding stoichiometry. The differences in kinetics constants, binding constants, stoichiometry

Quantitative Analysis of Small Molecule–Nucleic Acid Interactions

15

Fig. 3. Kinetic fitting to the AATT DNA data at low compound concentrations. The DB293 concentrations from bottom to top are 1, 4, 6, 10, 15 and 20 nM. The kinetic analysis is performed with mass transport kinetic 1:1 binding model. The smooth lines are the best fit lines using global fitting

and cooperativity for binding of DB293 to two different DNA hairpins, AATT and ATGA, can now be obtained as illustrated in Figs. 3 and 4. Under these experimental conditions, DB293 binds with a 1:1 ratio to the AATT site or the ATAT site (not shown), but with a 2:1 ratio to the ATGA site. 4. The sensorgram in Fig. 1 contains several distinct regions. In region (1), buffer flows over all surfaces, and a reference baseline is established. In region (2), DB293 is injected, and the kinetics of association can be determined. With time, a steady-state plateau region is established when binding and dissociation of DB293 are equal. In region (3), buffer flow is again started, and the DB293-DNA complex dissociates until the baseline is reached. If complete dissociation does not occur in a time period, a surface regeneration solution can be introduced. The RU on the surface is directly related to the DB293 bound. Based on the RU at saturation, we can determine that DB293 forms a 1:1 complex with the AATT DNA, as expected, and an unusual 2:1 complex with ATGA. 5. The steady state data for compound binding are fit with onesite (AATT) or two-site (ATGA) binding models (Fig. 4 and Table 2). The relative values of the macroscopic equilibrium constants, K1 and K2, reflect a highly cooperativity interac-

16

Liu and Wilson

Fig. 4. Comparison of the SPR binding affinity for AATT and ATGA. RU values from the steady-state region of SPR sensorgrams were converted to r (r = RU/RUmax) and are plotted against the unbound compound concentration (flow solution) for DB293 binding with AATT (squares) and ATGA (circles). The lines are the best fit values using appropriate binding models as described in the text

Table 2 Binding affinity comparison from steady-state fitting DNA binding site

K1 × 107 (1/M)

K2 × 107 (1/M)

CF (K2/K1) × 4

Binding mode

ATGA

0.41

0.98

9.6

Dimer, cooperative

AATT

4.87

—

—

Monomer

ATAT

0.62

—

—

Monomer

tion with ATGA. A cooperativity factor to assess the degree of cooperativity is defined as CF = (K2/K1) × 4. For interaction with no cooperativity, CF = 1, and CF > 1 for positive cooperativity and <1 for negative cooperativity (52). The positive cooperativity in binding of DB293 to ATGA can be easily seen from Table 2, suggesting that DB293 interacts as a cooperative dimer stacked with the ATGA sequence. 6. In cases where a steady-state plateau is reached, the ratio of the rate constants (ka/kd) should be compared to the steady-state

Quantitative Analysis of Small Molecule–Nucleic Acid Interactions

17

Table 3 Binding affinity comparison of kinetic and steady-state analysis for DB293 binding to the AATT DNA Exp.

ka (M−1 s−1)

kd (s−1)

KA (M−1)

kt [RU/(Ms)]

Chi2 (RU)2

ka × RUmax/kt

Kinetics

6.75 × 105

0.0155

4.35 × 107

1.56 × 1010

0.278

1.5 × 10−3

Steady-state

–

–

4.87 × 107

–

0.22

–

KA value. To illustrate a kinetic fit, the binding data for AATT with DB293 at low-concentration is globally fitted with a single site kinetic model with a mass transport term. The results are shown in Fig. 3 and Table 3. The binding constants obtained from kinetic and steady-state analyses are in excellent agreement, reflecting the high quality of data in the SPR experiments. In this case, the mass transfer effect is not significant and does not dominate the kinetics. The association rate constant ka is less than the transfer rate constant kt. 7. Kinetic analysis using global fitting of SPR data places great demand on obtaining high-quality data. Experimental design, analysis and optimization of kinetic studies have been described in detail elsewhere (27, 29) (see Note 24). In general, low surface densities of the immoiblized reactant and high analyte flow rate should be used to minimize the effects of mass transfer. Several criteria must be satisfied when considering whether a global kinetic fit is acceptable (48): (1) within experimental limits, RUmax is the same as the predicted value or from the steady-state results for one binding molecule; (2) the rate constants are within the range of small molecules; (3) the mass transport constant kt is in the 107range; (4) (ka × RUmax/kt) £ 5; (5) the half-life t1/2 from the dissociation phase of sensorgram is close to the calculated half-life using the fitted value (t1/2 = ln2/kd), suggesting the mass transport effect is minimized; (6) the residuals are within the instrumental noise, and there are no systematic deviations; (7) a low chi-squared value is obtained at convergence.

4. Notes 1. Maintenance chips are available from GE Healthcare Inc. For desorb running, 50% DMSO and 10% DMSO can be used instead of Biacore desorb solution 1 and Biacore desorb solution 2 to effectively remove sticky small molecules. For bio-

18

Liu and Wilson

logical samples, such as protein, Sanitize method is also used after Desorb to insure that no microbial growth is present in the liquid injection and flow system. 2. After running the regular Desorb, if the baseline is still not stable within ±1.0  RU/min (check with a new CM5 chip using running buffer), additional Super clean method may be used. First, run Desorb using 1% (v/v) acetic acid in place of Biacore desorb solution 1 and Biacore desorb solution 2, followed by one Prime to wash out the residual acetic acid. Then, run Desorb using 0.2 M sodium bicarbonate in place of SDS (solution 1) and glycine (solution 2), followed by one Prime to wash out the sodium bicarbonate residuals. Last, run Desorb using 6 M guanidine HCl for the SDS and 10 mM HCl for glycine. Prime the instrument a few times to thoroughly clean all residuals. 3. Typically, an immobilization amount of 300–450  RUs of hairpin nucleic acid (~20–30 bases in length) is immobilized for running steady-state experiments and a low amount of 100–150  RUs for kinetic experiments to minimize mass transfer effects. 4. A scrambled nonbinding sequence could be immobilized on flow cell 1 in some situations to provide a more similar control surface for subtraction. 5. To visualize and illustrate a difference in stoichiometry for different nucleic acid hairpins, equal moles of nucleic acid can be immobilized on different flow cells using the equation (RU = C × moles bound × MWnucleic acid), where C is a relatively constant conversion factor for nucleic acid, proteins, etc. A higher level of response units (RU) is required for a higher molecular weight hairpin. 6. Don’t open the package or dock the chip until the liposomes are prepared and ready for use to minimize adsorption of unwanted material from the air and running buffer on to the sensor chip. 7. Detergents should not be used. 8. Buffer degassing is very important, and is particularly important if experiments are run at temperatures above 25°C. 9. For the Biaore 2000 instrument, even 10 times lower surfactant concentration (0.0005%, v/v polysorbate 20) can be applied to the running buffer, and it works very well. While for Biacore 3000 and Biacore T100, higher concentration (0.005%, v/v polysorbate 20) is needed in the running buffer. Depending on the binding compound, a little higher binding affinity may be obtained when increasing surfactant concentration in the running buffer. Salt concentration can be adjusted based on the experimental requirement. The higher the salt concentration, the lower the binding affinity that will be obtained when cations bind to nucleic acids.

Quantitative Analysis of Small Molecule–Nucleic Acid Interactions

19

10. Good regeneration conditions should remove the analyte completely from the surface without removing or damaging the immobilized ligand. The generally used regeneration solutions are listed in Table 4. In general, milder conditions are initially used, but more stringent conditions are applied as need. Some other regeneration solutions for special sample are available from the Biacore website. In our studies, 10 mM Glycine-HCl (pH 2.5) is typically used as an efficient regeneration solution to remove small molecules from the DNA immobilized sensor chip surface. 11. The sample concentration to be used depends on the magnitude of the binding constant (KA). With a single binding site, for example, concentrations at least ten times above and below 1/KA should be used (i.e., a 100-fold difference between the lowest and highest concentrations). A larger concentration range above and below 1/KA will yield a more complete binding curve. For binding constants of 107– 109 M−1, as observed with many nucleic acid/small molecule complexes, small molecule concentrations from 0.01 nM to 10 µM in the flow solution allow accurate determination of binding constants. Injecting samples from low to high concentration is useful for eliminating artifacts in the data from adsorption or carry over. 12. If the KA is unknown, it is necessary to conduct a preliminary experiment with several concentrations spread over a broad range to obtain an estimate of KA. A more focused set of concentrations is then prepared to cover the specific binding range.

Table 4 Choice of regeneration solutions Type of bond

Acidic

Basic

Weak strength

pH > 2.5 Formic acid HCl 10 mM Glycine/ HCl

pH < 9 10 mM pH < 9 50% HEPES/NaOH ethylene glycol

Intermediate strength

pH 2–2.5 pH 9–10 NaOH Formic acid HCl 10 mM Glycine/ 10 mM Glycine/HCl NaOH H3PO4

Strong strength pH < 2 Formic acid HCl 10 mM Glycine/ HCl H3PO4

pH > 10 NaOH

Hydrophobic

Ionic 1 M NaCl

pH 9–10 50% ethylene glycol

2 M MgCl2

pH > 10 25–50% ethylene glycol

4 M MgCl2 6M ­guanidine chloride

20

Liu and Wilson

13. For the steady-state method, equilibrium, but not kinetics, constants can be obtained even when mass transfer effects dominate the observed kinetics. Thus, higher flow rates are not required in steady-state experiments, as long as a clear steady-state plateau is obtained for determining RU at the steady-state. Higher flow rates (>50 µl/min) are used for kinetic experiments to minimize mass transport effects. 14. The sample solution must be prepared in the same buffer used to establish the baseline − the running buffer. 15. If the small molecule requires the presence of a small amount of an organic solvent (e.g., <5% DMSO) to maintain solubility, the same amount of this organic solvent must be in the running buffer to minimize the refractive index difference. 16. A sufficient association phase with a plateau region is needed for steady-state analysis. For the most accurate fitting of the dissociation phase, it is best to allow sufficient time for the compound to dissociate at least 80% from the complex. 17. Many organic small molecules are easily adsorbed nonspecifically to the tubing of the injection micro-fluidics and are slowly released over the course of the experiment. 18. Possible problems at high sample concentrations: poor sensorgrams and non-specific binding may be obtained. 19. A short contact time, 30–60  s, is usually sufficient. Longer exposure to regeneration conditions involves greater risks for loss of binding activity on the surface, and often does not lead to improved regeneration. 20. Other software programs such as Scrubber2 and CLAMP are available for processing Biacore data. The results can also be exported and presented in graphing software such as KaleidaGraph for either PC or Macintosh computers. Although it is useful to experiment with different software packages, BIAevaluation (current version 4.1) is sufficient for most routine analyses of sensorgram data. For the Biacore T100 user, data processing can be performed automatically using the Biacore T100 evaluation software, which is much more convenient for new users. 21. These two data processing steps are referred to as “double referencing”. Typically, multiple buffer injections are performed and averaged before subtraction. In double referencing, plots are made for each flow cell separately overlaying the control flow cell− corrected sensorgrams from buffer and all sample injections. The buffer sensorgram is then subtracted from the sample sensorgrams. “Double referencing” removes the systematic drifts and shifts in baseline, and is helpful to minimize offset artifacts and also to correct the bulk shift that results from slight differences in injection buffer and running buffer (27).

Quantitative Analysis of Small Molecule–Nucleic Acid Interactions

21

22. In some instances, at low concentrations, in which the response does not reach the steady-state, the equilibrium responses can be obtained from kinetic fits of the sensorgrams utilizing the known RUmax from the higher concentration sensorgrams. This extrapolation method works well with sensorgrams wherein the observed response is at least 50% of the equilibrium RU. 23. Various fitting models are included in the Biacore evaluation software and users can also write other models to be used with the program. While more complex binding models should only be used if the materials are pure, the experiment has been conducted properly, and the more complex model improves the fit significantly above the experimental error. 24. In cases where a steady-state plateau is reached, the ratio of the rate constants (ka/kd) should be compared to the steadystate KA value. An agreement between the two methods suggests that the binding constant, KA, is correct but does not necessarily mean that the ka and kd values are correct due to possible mass transfer effects and possible correlation of constants (29, 47).

Acknowledgments We very much thank the NIH for funding the research that has made this review possible and the Georgia Research Alliance for funding of Biacore instruments. We also very much thank Professor David W. Boykin and his coworkers (Georgia State University, Atlanta, GA, USA) for supplying DB293 and many other DNA and RNA binding agents for SPR studies as well as for helpful discussions.

References 1. Charbonnier S, Gallego O, Gavin AC (2008) The social network of a cell: recent advances in interactome mapping. Biotechnol Annu Rev 14:1–28 2. Figeys D (2008) Mapping the human protein interactome. Cell Res 18:716–724 3. Reed JW, Bartel B (2008) Cell signaling and gene regulation. Curr Opin Plant Biol 11:471–473 4. Kornberg R (2007) The molecular basis of eukaryotic transcription. Angew Chem Int Ed Engl 46:6956–6965

5. Majmudar CY, Lum JK, Prasov L, Mapp AK (2005) Functional specificity of artificial transcriptional activators. Chem Biol 12:313–321 6. Berg T (2008) Inhibition of transcription factors with small organic molecules. Curr Opin Chem Biol 12:464–471 7. Xiao X, Yu P, Lim HS, Sikder D, Kodadek T (2007) Design and synthesis of a cell-permeable synthetic transcription factor mimic. J Comb Chem 9:592–600 8. Burnett R, Melander C, Puckett JW, Son LS, Wells RD, Dervan PB et  al (2006) DNA

22

Liu and Wilson

sequence-specific polyamides alleviate transcription inhibition associated with long GAA. TTC repeats in Friedreich’s ataxia. Proc Natl Acad Sci U S A 103:11497–11502 9. Wilson WD, Tanious FA, Mathis A, Tevis D, Hall JE, Boykin DW (2008) Antiparasitic compounds that target DNA. Biochimie 90:999–1014 10. Tidwell RR, Boykin DW (2003) Dicationic DNA minor-groove binders as antimicrobial agents. In: Demeunynck M, Bailly C, Wilson WD (eds) DNA and RNA binders: from small molecules to drugs. Wiley-VCH, pp 414–460 11. Cai Z, Greene MI, Berezov A (2008) Modulation of biomolecular interactions with complex-binding small molecules. Methods 36:39–46 12. Dominy BN (2008) Molecular recognition and binding free energy calculations in drug development. Curr Pharm Biotechnol 9:87–95 13. Seiler KP, George GA, Happ MP, Bodycombe NE, Carrinski HA, Norton S et  al (2008) ChemBank: a small-molecule screening and cheminformatics resource database. Nucleic Acids Res 36:351–359 14. Butcher RA, Schreiber SL (2005) Using genome-wide transcriptional profiling to elucidate small-molecule mechanism. Curr Opin Chem Biol 9:25–30 15. Mapp AK, Ansari AZ (2007) A TAD further: exogenous control of gene activation. ACS Chem Biol 2:62–75 16. Chaires JB (2008) Calorimetry and thermodynamics in drug design. Annu Rev Biophys 37:135–151 17. Papalia GA, Giannetti AM, Arora N, Myszka DG (2008) Thermodynamic characterization of pyrazole and azaindole derivatives binding to p38 mitogen-activated protein kinase using Biacore T100 technology and van’t Hoff analysis. Anal Biochem 383: 255–264 18. Perozzo R, Folkers G, Scapozza L (2004) Thermodynamics of protein–ligand interactions: history, presence, and future aspects. J Recept Signal Transduct Res 24:1–52 19. Velazquez Campoy A, Freire E (2005) ITC in the post-genomic era...? Priceless. Biophys Chem 115:115–124 20. Li L, Dantzer JJ, Nowacki J, O’Callaghan BJ, Meroueh SO (2008) PDBcal: a comprehensive dataset for receptor-ligand interactions with three-dimensional structures and binding thermodynamics from isothermal titration calorimetry. Chem Biol Drug Des 71: 529–532

21. Privalov PL, Dragan AI (2007) Microcalorimetry of biological macromolecules. Biophys Chem 126:16–24 22. Plotnikov VV, Brandts JM, Lin LN, Brandts JF (1997) A new ultrasensitive scanning calorimeter. Anal Biochem 250:237–244 23. Ciulli A, Williams G, Smith AG, Blundell TL, Abell C (2006) Probing hot spots at proteinligand binding sites: a fragment-based approach using biophysical methods. J Med Chem 49:4992–5000 24. Turnbull WB, Daranas AH (2003) On the value of c: can low affinity systems be studied by isothermal titration calorimetry? J Am Chem Soc 125:14859–14866 25. Wilson WD (2002) Analyzing biomolecular interactions. Science 295:2103–2105 26. Rich RL, Myszka DG (2007) Survey of the year 2006 commercial optical biosensor literature. J Mol Recognit 20:300–366 27. Myszka DG (2000) Kinetic, equilibrium, and thermodynamic analysis of macromolecular interactions with BIACORE. Methods Enzymol 323:325–340 28. Nagata K, Handa H. (2000) Real–time analysis of biomolecular interactions: applications of Biacore. Springer 29. Nguyen B, Tanious FA, Wilson WD (2007) Biosensor-surface plasmon resonance: quantitative analysis of small molecule-nucleic acid interactions. Methods 42:150–161 30. Jason-Moller L, Murphy M, Bruno J (2006) Overview of Biacore systems and their applications. Curr Protoc Protein Sci Chapter 19. doi:10.1002/0471140864 31. Davis TM, Wilson WD (2000) Determination of the refractive index increments of small molecules for correction of surface plasmon resonance data. Anal Biochem 284:348–353 32. Davis TM, Wilson WD (2001) Surface plasmon resonance biosensor analysis of RNA– small molecule interactions. Methods Enzymol 340:22–51 33. Koh JT, Zheng J (2007) The new biomimetic chemistry: artificial transcription factors. ACS Chem Biol 2:599–601 34. Rezler EM, Bearss DJ, Hurley LH (2003) Telomere inhibition and telomere disruption as processes for drug targeting. Annu Rev Pharmacol Toxicol 43:359–379 35. Neidle S, Parkinson GN (2008) Quadruplex DNA crystal structures and drug design. Biochimie 90:1184–1196 36. Hampshire AJ, Fox KR (2008) The effects of local DNA sequence on the interaction of ligands with their preferred binding sites. Biochimie 90:988–998

Quantitative Analysis of Small Molecule–Nucleic Acid Interactions 37. Chaires JB (2005) Competition dialysis: an assay to measure the structural selectivity of drug-nucleic acid interactions. Curr Med Chem Anticancer Agents 5:339–352 38. Nguyen B, Neidle S, Wilson WD (2009) A role for water molecules in DNA–ligand minor groove recognition. Acc Chem Res. 42:11–21 39. Wilson WD, Nguyen B, Tanious FA, Mathis A, Hall JE, Stephens CE et  al (2005) Dications that target the DNA minor groove: compound design and preparation, DNA interactions, cellular distribution and biological activity. Curr Med Chem Anticancer Agents 5:389–408 40. Werbovetz K (2006) Diamidines as antitrypanosomal, antileishmanial and antimalarial agents. Curr Opin Investig Drugs 7:147–157 41. De Cian A, Lacroix L, Douarre C, TemimeSmaali N, Trentesaux C, Riou JF et al (2008) Targeting telomeres and telomerase. Biochimie 90:131–155 42. White EW, Tanious F, Ismail MA, Reszka AP, Neidle S, Boykin DW et al (2007) Structurespecific recognition of quadruplex DNA by organic cations: influence of shape, substituents and charge. Biophys Chem 126:140–153 43. Parkinson GN, Cuenca F, Neidle S (2008) Topology conservation and loop flexibility in quadruplex-drug recognition: crystal structures of inter- and intramolecular telomeric DNA quadruplex–drug complexes. J Mol Biol 381:1145–1156 44. Huppert JL (2008) Four-stranded nucleic acids: structure, function and targeting of G-quadruplexes. Chem Soc Rev 37:1375–1384

23

45. Qin Y, Hurley LH (2008) Structures, folding patterns, and functions of intramolecular DNA G-quadruplexes found in eukaryotic promoter regions. Biochimie 90:1149–1171 46. Lane AN, Chaires JB, Gray RD, Trent JO (2008) Stability and kinetics of G-quadruplex structures. Nucleic Acids Res 36:5482–5515 47. Tanious FA, Nguyen B, Wilson WD (2008) Biosensor-surface plasmon resonance methods for quantitative analysis of biomolecular interactions. Methods Cell Biol 84:53–77 48. Karlsson R (1999) Affinity analysis of nonsteady-state data obtained under mass transport limited conditions using BIAcore technology. J Mol Recognit 12:285–292 49. Navratilova I, Dioszegi M, Myszka DG (2006) Analyzing ligand and small molecule binding activity of solubilized GPCRs using biosensor technology. Anal Biochem 355:132–139 50. Day YS, Baird CL, Rich RL, Myszka DG (2002) Direct comparison of binding equilibrium, thermodynamic, and rate constants determined by surface- and solution-based biophysical methods. Protein Sci 11:1017–1025 51. Cooper MA, Hansson A, Lofas S, Williams DH (2000) A vesicle capture sensor chip for kinetic analysis of interactions with membrane-bound receptors. Anal Biochem 277:196–205 52. Peixoto P, Liu Y, Depauw S, Hildebrand MP, Boykin DW, Bailly C et al (2008) Direct inhibition of the DNA-binding activity of POU transcription factors Pit-1 and Brn-3 by selective binding of a phenyl-furan-benzimidazole dication. Nucleic Acids Res 36:3341–3353

Chapter 2 Thermal Melting Studies of Ligand DNA Interactions Aurore Guédin, Laurent Lacroix, and Jean-Louis Mergny Abstract A simple thermal melting experiment may be used to demonstrate the stabilization of a given structure by a ligand (usually a small molecule, sometimes a peptide). Preparation of the sample is straightforward, and the experiment itself requires an inexpensive apparatus. Furthermore, reasonably low amounts of sample are required. A qualitative analysis of the data is simple: An increase in the melting temperature (Tm) indicates preferential binding to the folded form as compared to the unfolded form. However, it is perilous to derive an affinity constant from an increase in Tm as other factors play a role. Key words:  FRET, Telomeres, Telomerase inhibitor, G-quadruplex, G-Quartet, DNA ligands

1. Introduction One of the possible methods to demonstrate an interaction between a compound and a nucleic acid is to perform a melting experiment: In the presence of the molecule, the melting temperature of the DNA or RNA should increase. This simple approach does not require any spectroscopic condition for the ligand – no fluorescence or absorbance special properties are required and has been applied for decades, initially on polynucleotides (see for example (1)), then on unusual DNA structures such as triplexes (2, 3) or quadruplexes. Quadruplexes can be formed by certain guanine-rich sequences in the presence of monovalent cations and are stabilized by G-quartets (Fig.  1). The theory behind these melting experiments, which is not fundamentally different when larger molecules such as polypeptides are considered can be complicated, especially for polynucleotides (4–6). Hence, it is not straightforward to translate an increase in melting temperature into affinity constants. Although the two phenomena are K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_2, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

25

26

Guédin, Lacroix, and Mergny

Fig.  1. G-quadruplexes. Top: A G-quartet, with four coplanar guanines and a central cation (here potassium; in reality the ion is sandwiched between the planes of two consecutive quartets). Bottom: Two possible folds of an oligonucleotide having 4 blocks of 3 guanines. These two conformations involve 3 quartets, but the loops are arranged in a different manner, and the syn (dark blue)/anti (light grey) orientations of the guanine sugar/phosphate backbone are different.

correlated (the strongest ligands tend to lead to the largest increase in Tm), other factors play a role, such as the number of binding sites, cooperativity of binding, and affinity of the ligand for the single strands. In other words, the increase in the melting temperature of a quadruplex induced by the presence of the ligand leads to a semi quantitative measurement of the interaction between a ligand and a nucleic acid. An increase in the melting temperature of the nucleic acid may of course be followed by absorbance. Nucleic acids display a high absorbance around 260 nm, and it is known for decades that this signal increases when its secondary structure is denatured. For unusual structures, such as quadruplexes, one should select different wavelengths such as 240 or 295 nm (7). An alternative, although less used signal is provided by circular dichroism.

Thermal Melting Studies of Ligand DNA Interactions

27

Provided that an appropriate wavelength is chosen, one can study the melting of a DNA structure in the presence of ligands by recording ellipticity as a function of temperature (8). But it is also possible and often desirable to label the nucleic acid with fluorescent groups and obtain fluorescence vs. temperature melting profile. One popular approach is to attach a donor and a quencher at both ends of an intramolecular structure and follow its thermal denaturation by a concomitant increase in donor emission (9–14) (Fig. 2). There are several important advantages of a fluorescencebased assay: (1) The difference between the emission levels of the folded and unfolded forms can be very significant; (2) The synthesis or screening of large chemical libraries of potential ligands needs an inexpensive, reproducible and fast assay to test nucleic acid binding: A fluorescent melting assay may easily be converted into 96- or 384-well format, allowing for the weekly screening of thousands of compounds; (3) Lower concentrations are required; (4) many ligands have a significant absorbance in the region where DNA absorbs light, which may interfere with the absorbance or ellipticity signal resulting from complex dissociation. On the other hand, only a handful of molecules were found to interfere with a fluorescence-based melting assay; (5) finally, as the targeted nucleic acid is labeled, one may add massive amounts of “unlabeled” nonspecific nucleic acid competitors without interfering with the fluorescent signal.

1

1

0

0

Fig. 2. Example of a FRET melting experiment for a quadruplex-forming oligonucleotide. Top: Sequence of the oligonucleotide used. Bottom: Fluorescence emission vs. temperature plot for the sequence above determined in a 10 mM sodium cacodylate pH 7.2 buffer with 0.1 M NaCl (total monocation concentration: 110 mM). Fluorescence intensities at both wavelengths have been normalized to 1. Unfolding of the quadruplex leads to a decrease in the emission at 515 nm (donor, dark curve) and a concomitant increase at 588 nm (acceptor, grey curve).

28

Guédin, Lacroix, and Mergny

2. Materials 2.1. Nucleic Acids 2.1.1. Design and Synthesis

For absorbance, simple unlabeled oligonucleotides may be ordered. A synthesis on the 0.2 micromole scale is sufficient for most applications. Polynucleotides may also be tested. For fluorescence studies, modified oligonucleotides must be synthesized. Fortunately, these compounds are commercially available. For quadruplex stabilization assay, we designed a doublylabeled oligonucleotide that forms an intramolecular structure. Its sequence is 5¢d-GGGTTAGGGTTAGGGTTAGGG3¢. The FAMTAMRA and FAM-Dabcyl dual-labeled oligomers are called “F21T” and “F21D”, respectively. Other groups have chosen identical or different FRET pairs (15–19), such as for example fluorescein-Cy3, Cy3-Cy5, or pairs of chemically modified fluorescein, and rhodamine dyes (i.e., Texas Red or ROX; for a recent review on FRET pairs, see (20)) (see Notes 1 and 2). Other quadruplex forming sequences may be tested. We recently analyzed oligonucleotides mimicking the Plasmodium falciparum degenerate telomeric motif (GGGTTYA)n (21). Nontelomeric repeats or RNA quadruplexes (22) may also be studied; they are reviewed in (23). Figure 3 summarizes results obtained with four different

Fig. 3. Tm Stabilization by a quadruplex ligand. Top: Chemical formula of the G4 ligand (360A) (33, 34). Bottom: Stabilization obtained with the 360A ligand (22) for the human telomeric motif F21T (GGGTTA)3GGG; the Plasmodium telomeric repeat (GGGTTTA)3GGG (FPf1T), the human telomeric RNA (GGGUUA)3GGG (F21RT); and hairpin duplex with a HEG loop TATAGCTATAC18TATAGCT ATA (FdXT) (F FAM, T TAMRA). All measurements were performed in a 10 mM lithium cacodylate pH 7.2 buffer with 10 mM KCl and 90 mM LiCl (total monocation concentration: »110 mM).

Thermal Melting Studies of Ligand DNA Interactions

29

oligonucleotides (three quadruplex forming sequences and one hairpin duplex). We recommend checking the purity of these oligomers, using mass spectrometry, HPLC, or gel electrophoresis. 2.1.2. Determination of Concentration

Concentrations of all oligodeoxynucleotides were estimated using extinction coefficients provided by the manufacturer and calculated with a nearest neighbor model (24). The concentration should be checked even if the manufacturer provides an estimate of oligonucleotide amount.

2.2. Buffers

A proper ionic environment is crucial as this parameter will influence the stability of the nucleic acid and modulate the interaction between a cationic ligand and a negatively charged DNA or RNA. The ionic strength and nature of the salt will also modulate the Tm. The cation will be condensed around nucleic acids, which are negatively-charged polyelectrolytes. There is a near-linear relationship between Tm and the logarithm of salt concentration, reaching a plateau around 1 M salt. A “reasonable” buffer contains 0.1–0.15 M salt (NaCl, LiCl or KCl). We suggest choosing buffer conditions that lead to a melting temperature between 40 and 50°C (see Note 3). For guanine quadruplexes, this parameter can also be tuned with a partial substitution of Na+ or K+ with Li+ while keeping the total monocation concentration constant (14). For the human telomeric quadruplex, initial experiments were performed in a 0.1 M lithium chloride, 10 mM sodium cacodylate pH 7.2 buffer (25) (i.e. with 10 mM Na+ ions). Later, we changed the reference conditions and compared stabilization values of this quadruplex in two different buffers. Both contain 10 mM cacodylic acid buffered to pH 7.2 with LiOH and either (1) 100 mM sodium chloride or (2) 10 mM potassium chloride + 90  mM lithium chloride. When studying extremely stable quadruplexes, such as c-myc DNA or some RNA G4 sequences, one can choose buffer conditions with an even lower potassium concentration (for example, 1 mM) supplemented with LiCl in order to keep total cation concentration around 0.1 M (26). An intermediate Tm value is required to follow the stabilization obtained in the presence of very « strong » ligands.

2.2.1. Choice of the Ionic Strength

2.2.2. Storage and Preparation

2–10× buffer solutions may be prepared in advance, filtered and kept at 4°C for weeks.

2.3. Ligands

Ligands should be solubilized in the mM concentration range using an adequate solvent (typically H2O or DMSO). Stock solutions can be kept in the dark at −20°C for months. Fresh dilutions should be prepared before each melting experiment. High temperatures may lead to degradation of the ligand; in that case, one should reduce the duration of incubations at high temperatures.

2.3.1. Storage and Preparation

30

Guédin, Lacroix, and Mergny

2.3.2. Concentration Range

It is important to test the stabilization of the quadruplex at various ligand concentrations, as initially proposed by Neidle and coworkers (16). The final ligand concentration should be adjusted between two values: The lower corresponding to the oligonucleotide concentration as lower concentration may lead to complex melting profiles. On the other hand, the highest ligand concentration should not lead to a too-high stabilization (a Tm above 80°C is experimentally difficult to determine) and/or precipitation of the oligonucleotide (see Note 4). Comparison of the profiles obtained at different concentrations generates a DTm vs. concentration curve (16, 27).

3. Methods 3.1. Absorbance 3.1.1. Instrumentation

Different models and trademarks of high-quality spectrophotometers are available on the market. We use Uvikon 940 and Secomam XL spectrophotometers. Their performance and stability over time in the far UV region (and at high temperature) should be guaranteed. The user should be able to select a temperature gradient over a wide range, and be able to monitor the sample temperature (with an external probe immerged in a control cuvet). Both heating and cooling profiles should be recorded. A multisample cell holder will allow a reasonable throughput, even with experiments that last 15 h. Besides 260 nm (or 295 nm for quadruplexes), it is sometimes appropriate to record absorbance around the ligand absorbance maximum and at a control wavelength. It is therefore useful to have the capability of recording absorbance at several wavelengths.

3.1.2. Preparation of the Samples

Tm of the (DNA (or RNA) + ligand) mixture should always be compared with the Tm of the DNA alone. A multisample cell changer allows performing this control in parallel. Make sure that the total absorbance stays in the linear range of the spectrophotometer – the ligand may significantly absorb light at 260 nm (see Note 5).

3.1.3. Absorbance vs. Temperature measurements

Data acquisition may be programmed for most spectrophotometers. The exact protocol depends on the instrument. Melting experiments are typically performed at a concentration of 2–5 mM per strand. Most transitions are reversible, as shown by superimposable heating and cooling profiles at a fixed rate of 0.1–0.5°C/min. This indicates that the denaturation curves correspond to a true equilibrium process, but prevent us from accessing kinetic parameters (28, 29). One should keep in mind that profiles may be superimposable for the oligonucleotide alone but not the drug-DNA (RNA) complex.

Thermal Melting Studies of Ligand DNA Interactions

31

We recommend using the term T1/2 (instead of Tm) for experiments in which no demonstration of reversibility is provided. 3.2. Circular Dichroism

The choice for a circular dichroism spectrophotometer is more limited. Nevertheless, different models from Aviv Biosciences, and Jasco are available. We use a Jasco J-810 instrument equipped with a Peltier controlled cell changer.

3.3. Fluorescence (FRET)

Melting measurements may either be performed on: A traditional spectrofluorimeter (in our laboratory a Spex Fluoromax 3 instrument) with 600 µl of solution containing 0.2 µM of tagged oligonucleotide. An example of a FRET melting curve is provided in Fig. 2. A real time PCR apparatus (MX3000P, Stratagene; or Roche LightCycler or DNA engine Opticon, MJ Research), allowing the simultaneous recording of up to 96 independent samples as first proposed by S. Neidle and coworkers. The typical concentration of fluorescent oligonucleotide is 0.2 µM (strand concentration), but acceptable curves may be obtained with concentrations in the 0.2–0.5 µM range depending on the volume (from 20 to 100 µl), the gain and the type of detection of the quantitative PCR apparatus used. Each condition is tested at least in duplicate, in a volume of 25 µL for each sample (see Fig. 4 for examples of FRET melting curves recorded on a real time PCR machine) (see Note 6).

3.4. Tm Determination

The melting temperature (Tm) corresponds to the equilibrium temperature at which half of the sample is folded, and half is unfolded. A time-consuming but precise method to determine Tm has been described in detail (30). In cases where equilibrium completion is unknown, T1/2 should be used instead of Tm (see Note 7). A popular alternative to this baseline method is simply to determine the maximum of the first derivative of the absorbance or fluorescence signal (dA/dT or dF/dT). This is a simple and user-independent method, and many software programs provide an automatic “Tm” determination by this approach. However, one should keep in mind that this approach only gives an approximation of the true Tm. Nevertheless, this method is sufficient for most applications concerning nucleic acid ligands.

4. Notes 1. Some of the commonly used fluorescent labels have pH-sensitive emission properties. It is therefore recommended to check that the chosen pH does not quench the fluorescent properties of the dye.

32

Guédin, Lacroix, and Mergny

Fig. 4. Effect of competitors on T½ increase. Top: Melting profiles for the human telomeric motif F21T (GGGTTA)3GGG alone (circle), in the presence of 1 µM 360A ligand (full line) and in the presence of a duplex competitor (ds26, 3 µM, hatched/dotted line). All curves are presented in duplicate. Bottom: Melting profiles for the human telomeric motif F21T (GGGTTA)3GGG alone (circle), in the presence of 1 µM 360A ligand (full line) and in the presence of an RNA quadruplex competitor (RhTR41 (26), 3 µM, hatched/ dotted line). All curves are presented in duplicate. The duplex competitor does not lead to a significant decrease of the stabilization provided by the quadruplex ligand (top) while an excess of an RNA quadruplex nearly abolishes this effect: The curve reverts to the melting profile obtained in the absence of the ligand (arrow). All measurements were performed in a 10 mM lithium cacodylate pH 7.2 buffer with 10 mM KCl and 90 mM LiCl (total monocation concentration: »110 mM).

Thermal Melting Studies of Ligand DNA Interactions

33

2. The stabilization value depends on the nature of the fluorescent tags, the incubation buffer and the method chosen for Tm calculation, complicating a direct comparison of the results obtained by different laboratories. 3. The buffer plays an important and often underestimated role. We strongly recommend, cacodylate (pKa: 6.14 at 25°C) and acetate (pKa: 4.62) for near neutral and slightly acidic conditions, respectively. These buffers do not absorb light in the far UV-region and their pKa is not too temperature-dependent. This is a key issue as some nucleic acid structures are extremely pH-dependent (31). The pKa of other buffers such as MES, BES, TES, Tricine, HEPES, MOPS or TAPS is extremely temperature-dependent (over 2 pH units between 0 and 100 °C) (32). These are therefore not appropriate for Tm experiments. 4. The ligand may interfere with the fluorescent labels rather than with the nucleic acid. In that case, an increase in melting temperature reflects an interaction with the fluorescent dye, not with the target structure, generating false positives. DNA/RNA interactions should therefore be confirmed by other methods. 5. A problem associated with heating is due to the reduced gas solubility at high temperature, which leads to the formation of air bubbles in the sample. These air bubbles will seriously alter absorbance, CD, or fluorescence measurements if they are in the optical pathway. Two simple methods solve this difficulty: It is possible to preheat the sample at 90°C for 5–15 min or to degas the sample under vacuum at room temperature for an equivalent amount of time. The latter solution should be preferred if the sample is extremely heat-sensitive. One may also preheat the buffer component before adding the nucleic acid. 6. For FRET-based fluorescent systems, we found that following the emission of the “donor” (in our case, fluorescein or FAM) gives more reproducible results than the sensitized emission of the acceptor (TAMRA). 7. Manual baseline determination, which is normally required for true Tm determination (30), is actually impractical when facing the amount of data generated by a 96-plate reader. Only two feasible automatable alternatives are possible: (1) T1/2 determination and (2) first derivative analysis (14). Although both methods provide a less rigorous analysis than a full description of the melting process, they are much faster to implement, and values are obtained in a user-independent way. Although we tend to favour the first one (1), both methods are usually valid.

34

Guédin, Lacroix, and Mergny

Acknowledgments We thank all the past and present members of the “Laboratoire de Biophysique” in the Muséum National d’Histoire Naturelle. This work was supported by an E.U. FP6 “MolCancerMed” (LSHC-CT-2004-502943) grant. References 1. Stewart CR (1968) Broadening by acridine orange of the thermal transition of DNA. Biopolymers 6:1737–1743 2. Mergny JL, Duval-Valentin G, Nguyen CH, Perroualt L, Faucon B, Montenay-Garestier T, Bisagni E, Hélène C (1992) Triple helix specific ligands. Science 256:1691–1694 3. Escudé C, Nguyen CH, Mergny JL, Sun JS, Bisagni E, Garestier T, Hélène C (1995) Selective stabilization of DNA triple helices by benzopyridoindole derivatives. J Am Chem Soc 117:10212–10219 4. Crothers DM (1971) Statistical thermodynamics of nucleic acids melting transitions with coupled binding equilibria. Biopolymers 10: 2147–2160 5. McGhee JD (1976) Theoretical calculation of the helix-coil transition of DNA in the presence of large, cooperatively binding ligands. Biopolymers 15:1345–1375 6. McGhee JD, von Hippel PH (1974) Theore­ tical aspects of DNA Protein interactions: Cooperative and non cooperative binding of large ligands to a one dimensional homogeneous lattice. J Mol Biol 86:469–489 7. Mergny JL, Phan AT, Lacroix L (1998) Following G-quartet formation by UV-spectroscopy. FEBS Lett 435:74–78 8. De Cian A, Mergny JL (2007) Quadruplex ligands may act as molecular chaperones for tetramolecular quadruplex formation. Nucleic Acids Res 35:2483–2493 9. Mergny JL, Boutorine AS, Garestier T, Belloc F, Rougée M, Bulychev NV, Koshkin AA, Bourson J, Lebedev AV, Valeur B, Thuong NT, Hélène C (1994) Fluorescence energy transfer as a probe for nucleic acid structures and sequences. Nucleic Acids Res 22:920–928 10. Simonsson T, Sjoback R (1999) DNA tetraplex formation studied with fluorescence resonance energy transfer. J Biol Chem 274:17379–17383 11. Mergny JL (1999) Fluorescence energy transfer as a probe for tetraplex formation: The i-motif. Biochemistry 38:1573–1581

12. Mergny JL, Maurizot JC (2001) Fluorescence resonance energy transfer as a probe for G-quartet formation by a telomeric repeat. Chembiochem 2:124–132 13. Rachwal PA, Fox KR (2007) Quadruplex melting. Methods 43:291–301 14. De Cian A, Guittat L, Kaiser M, Sacca B, Amrane S, Bourdoncle A, Alberti P, TeuladeFichou MP, Lacroix L, Mergny JL (2007) Fluorescence-based melting assays for studying quadruplex ligands. Methods 42:183–195 15. Darby RAJ, Sollogoub M, McKeen C, Brown L, Risitano A, Brown N, Barton C, Brown T, Fox KR (2002) High throughput measurement of duplex, triplex and quadruplex melting curves using molecular beacons and a LightCycler. Nucleic Acids Res 30:e39 16. Schultes CM, Guyen W, Cuesta J, Neidle S (2004) Synthesis, biophysical and biological evaluation of 3, 6-bis-amidoacridines with extended 9-anilino substituents as potent G-quadruplex-binding telomerase inhibitors. Bioorg Med Chem Lett 14:4347–4351 17. Gomez D, Paterski R, Lemarteleur T, Shin-ya K, Mergny JL, Riou JF (2004) Interaction of telomestatin with the telomeric single-strand overhang. J Biol Chem 279:41487–41494 18. Juskowiak B, Galezowska E, Zawadzka A, Gluszynska A, Takenaka S (2006) Fluorescence anisotropy and FRET studies of G-quadruplex formation in presence of different cations. Spectrochim Acta A Mol Biomol Spectrosc 64:835–843 19. Moore MJB, Schultes CM, Cuesta J, Cuenca F, Gunaratnam M, Tanious FA, Wilson WD, Neidle S (2006) Trisubstituted acridines as G-quadruplex telomere targeting agents. Effects of extensions of the 3, 6-and 9-side chains on quadruplex binding, telomerase activity, and cell proliferation. J Med Chem 49:582–599 20. Sapsford KE, Berti L, Medintz IL (2006) Materials for fluorescence resonance energy transfer analysis: Beyond traditional donoracceptor combinations. Angew Chem Int Ed Engl 45:4562–4589

Thermal Melting Studies of Ligand DNA Interactions 21. De Cian A, Grellier P, Mouray E, Depoix D, Bertrand H, Monchaud D, Teulade-Fichou MP, Mergny JL, Alberti P (2008) Plasmodium telomeric sequences: Structure, stability and quadruplex targeting by small compounds. Chembiochem 9:2730–2739 22. De Cian A, Gros J, Guédin A, Haddi M, Lyonnais S, Guittat L, Roiu J-F, Trentesaux C, Sacca B, Lacroix L, Alberti P, Mergny LJ (2008) DNA and RNA quadruplex ligands. Nucleic acids Symp Ser 52:7–8 23. Juskowiak B, Takenaka S (2006) Fluorescence resonance energy transfer in the studies of guanine quadruplexes. Methods Mol Biol 335:311–341 24. Cantor CR, Warshaw MM, Shapiro H (1970) Oligonucleotide interactions. 3. Circular dichroism studies of the conformation of deoxyoligonucleotides. Biopolymers 9:1059–1077 25. Mergny JL, Lacroix L, Teulade-Fichou MP, Hounsou C, Guittat L, Hoarau M, Arimondo PB, Vigneron JP, Lehn JM, Riou JF, Garestier T, Hélène C (2001) Telomerase inhibitors based on quadruplex ligands selected by a fluorescent assay. Proc Natl Acad Sci USA 98:3062–3067 26. Gros J, Guédin A, Mergny JL, Lacroix L (2008) G-quadruplex formation interferes with P1 helix formation in the RNA component of telomerase hTERC. Chembiochem 9:2075–2078 27. Kaiser M, Sainlos M, Lehn JM, Bombard S, Teulade-Fichou MP (2006) Aminoglycosidequinacridine conjugates: Towards recognition

35

of the P6.1 element of telomerase RNA. Chembiochem 7:321–329 28. Rachwal PA, Findlow S, Werner JM, Brown T, Fox KR (2007) Intramolecular DNA quadruplexes with different arrangements of short and long loops. Nucleic Acids Res 35:4214–4222 29. Gros J, Rosu F, Amrane S, De Cian A, Gabelica V, Lacroix L, Mergny JL (2007) Guanines are a quartet’s best friend: Impact of base substitutions on the kinetics and stability of tetramolecular quadruplexes. Nucleic Acids Res 35:3064–3075 30. Mergny JL, Lacroix L (2003) Analysis of thermal melting curves. Oligonucleotides 13: 515–537 31. Mergny JL, Lacroix L, Han X, Leroy JL, Hélène C (1995) Intramolecular folding of pyrimidine oligodeoxynucleotides into an i-DNA motif. J Am Chem Soc 117:8887–8898 32. Fukada H, Takahashi K (1998) Enthalpy and heat capacity changes for the proton dissociation of various buffer components in 0.1  M potassium chloride. Proteins 33:159–166 33. Granotier C, Pennarun G, Riou L, Hoffschir F, De Cian A, Gomex D, Mandine E, Riou JF, Mergny JL, Mailliet P, Duttrilaux P, Boussin FD (2005) Preferential binding of a G-quadruplex ligand to human chromosome ends. Nucleic Acids Res 33:4182–9410 34. Pennarun G, Granotier C, Hoffschir F, Mandine E, Biard D, Gauthier LR, Boussin FD (2008) Role of ATM in the telomere response to the G-quadruplex ligand 360A. Nucleic Acids Res 36:1741–1754

Chapter 3 Circular and Linear Dichroism of Drug-DNA Systems Alison Rodger Abstract When a drug binds to DNA, its electronic structure is perturbed, and it perturbs the DNA’s electronic structure. The resulting change to the electronic spectroscopy can be used to probe the drug-DNA interaction. This chapter outlines how circular and linear dichroism spectroscopy can be used to provide information about drug-DNA systems. Circular dichroism spectroscopy involves measuring the difference in absorption of left and right circularly polarized light. It is uniquely sensitive to the helicity of the molecules being studied. Linear dichroism, as the name implies, involves measuring the difference in absorption of light linearly polarized parallel and perpendicular to an orientation axis. Linear dichroism provides information about the relative orientations of subunits of an interacting system. The material presented in this chapter is by no means comprehensive; the aim is to enable the user to collect reasonable quality data and to interpret it. Key words: Circular dichroism, Linear dichroism, Spectroscopy, DNA, Drug

1. Introduction When a drug binds to DNA, either a covalent bond is formed or the interaction is stabilized by various noncovalent interactions such as charge–charge interactions, hydrophobic interactions, dispersion interactions, dipole–dipole interactions, etc. All of these interactions affect the electronic structure of the molecules, and thus alter their electronic spectroscopy. This chapter outlines how circular and linear dichroism spectroscopy can be used to probe drug-DNA interactions. 1.1. Circular Dichroism Spectroscopy

Circular dichroism (CD) spectroscopy involves measuring the difference in absorption of left and right circularly polarized light (1). CD = Al - Ar

K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_3, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

37

Rodger 10 �/(mol�1cm�1dm3)

38

26% GC 42% GC 72% GC

5

100% GC 100% GC Z-DNA

0

–5

–10 200

220

240 260 280 Wavelength / nm

300

320

Fig. 1. CD spectrum of DNAs as a function of GC content. All sample in pH = 6.8, 1 mM sodium cacodylate, 5 mM NaCl. Clostridium perfringens (27%), calf thymus DNA (42%), Micrococcus Lysodeikticus (72%), poly[d(G-C)]2 (100%). Also shown is Z-form poly[d(GC)]2 induced by 50 µM [Co(NH3)6]3+

With circularly polarized light, the electric field of the photon follows a helix as it propagates; thus, CD spectroscopy is uniquely sensitive to any helical or chiral character in the electronic structure of molecules. In the case of DNA, over the wavelength range we can readily measure (down to say 190 nm) the spectrum that is almost completely determined by the transitions of the DNA bases, which are themselves achiral (1). However, because they are attached to a chiral sugar unit and stacked to make a helical staircase, transitions within a given base are perturbed by the chiral environment around them, and we get net CD spectra, such as those illustrated in Fig. 1. DNA CD spectra are more dependent on the polymorph present (so on the local structure) than on the base sequence, so we can use CD to identify which form of DNA is present as illustrated. When a drug binds to DNA, it gets an induced CD (ICD) spectrum because of the interaction with DNA. This may result from either a geometric change in the drug or a coupling between its electronic transitions, and those of the DNA. Similarly, the DNA gets an ICD contribution to its CD spectrum from its interaction with the drug. If, as is usually the case, the drug has transitions in the DNA region of the spectrum, what we observe is a combination of DNA CD, DNA ICD, drug CD (which is zero for an achiral drug and nonzero for a chiral drug), and drug ICD as illustrated in Fig. 2. To measure a UV-visible CD spectrum of a drug-DNA complex, one requires a solution containing both species of interest (and no others that will absorb a significant percentage of the photons); a quartz cuvette of a pathlength that ensures the absorbance maximum in the range of interest is no more than

Circular and Linear Dichroism of Drug-DNA Systems 200 40

250

300

350

400

39

450

30 CD / mdeg

20 10 0

100 mM GC + 2.5 mM +3 +5 + 6.5

–10 –20 –30 –40 200

250

300

350

+ + + +

400

10 13 20 25

450

Wavelength / nm

Fig. 2. CD titration series for a DNA-binding platinum compound showing the change in signal, when the platinum compound binds to poly[d(G-C)]2 DNA (100 µM base, 50 mM NaCl)

two, and a CD spectropolarimeter. Ideally, one is also equipped with an absorbance spectrum of each component and the mixture. Although many CD spectroscopists prefer cylindrical cuvettes, for drug-DNA work rectangular cuvettes have the advantage of fitting into standard absorbance spectrometers and peltier thermostatted cell holders. 1.2. Linear Dichroism Spectroscopy

Linear dichroism (LD) spectroscopy, as the name implies, involves measuring the difference in absorption of two linear polarizations of light. By convention, these are usually chosen to be parallel and perpendicular to a sample orientation direction, so we may write (1), LD = A// − A⊥ =

3 S (3cos 2 a − 1) 2

where S is the orientation parameter, and a is the angle between the transition moment polarization (direction of electron displacement during the transition) and the parallel direction. If the sample is isotropic (randomly oriented as in a CD cuvette, S = 0), then the LD is zero; thus, the main experimental challenge of LD spectroscopy is to orient the sample. For small molecule drugs, this is achieved by putting them into a stretched film. For DNA, the simplest method is to use a fairly long polymer and orient the DNA by imposing a shear force on it by flowing it past one or two stationary walls. A side benefit of DNA flow LD is that small drug molecules and globular proteins will not orient in a flow system, so they will have no LD signal unless they bind to the DNA. This is the major advantage of LD spectroscopy compared with almost any other characterization method for drug-DNA systems: drugs are invisible unless they are actually bound to the flow oriented DNA (2–5).

40

Rodger

To measure a UV-visible LD spectrum of a drug in a film or drug-DNA complex in solution, one requires a method of measuring A // and A –| . In principle, this can be done with a normal absorbance spectrometer and a polarizer. Either the polarizer or the sample is then rotated to get the two spectra. This method is affected by variations in sample positioning (when it is rotated) or by light beam intensity (in the two polarizations if the polarizer is rotated), and also limited by the sensitivity of the spectrometer (usually no more than 3 decimal places in absorbance). A preferable approach, and the one described in this chapter, is to use a CD spectropolarimeter that has been converted for LD either by doubling the voltage across the photoelastic modulator or by inserting a quarter wave plate that changes the alternating left and right circularly polarized light into alternating parallel and perpendicular light.

2. Materials 2.1. Calibration

A neodymium or a holmium filter should be used to calibrate your instrument for wavelength accuracy according to the manufacturer’s instructions. Similarly, the manufacturer’s instructions for intensity calibration should be followed. This usually involves high chemical and enantiomeric purity camphor sulfonic acid or its ammonium salt. Another more stable standard with more peaks across the visible-UV range is currently being developed (6).

2.2. Samples

The samples for both CD and LD spectroscopy must not have too high an absorbance in the pathlength cell to be used over the wavelength range of interest. A maximum absorbance of two is a good working maximum, though one should be aware that CD spectropolarimeters are often used to lower wavelength than absorbance spectrometers, the lower limit of which is 200 nm unless they are purged with nitrogen gas. The molecules of interest (the DNA and drugs) must contribute a significant percentage of the absorbance. This can be checked by measuring an absorbance spectrum of the buffer/solvent with and without the DNA and drug.

3. Methods 3.1. Circular Dichroism

Measuring a CD spectrum is a routine procedure assuming one has access to a CD spectropolarimeter. If your sample is chiral (which DNA is) and gives a good UV-visible absorbance spectrum, then it is highly likely that you will get a good CD spectrum.

Circular and Linear Dichroism of Drug-DNA Systems

41

CD machines are much more expensive than standard absorbance instruments as the absorbance magnitude signal of a typical CD spectrum is a fraction of a percent of that observed in the normal absorbance spectrum, so the optics have to be significantly better. The radiation source in most UV-visible CD spectrometers is a high energy (150–450 W) xenon arc lamp, which provides light from ~900 nm to 170 nm, though the intensity is low at both extremes. Thus, there are no lamp change over issues. In some instruments, the lamp is water cooled. This can prove the most problematic part of the whole instrument since, if the local water supply is used for a flow through system, one is at the mercy of the water purity, hardness, etc. Another requirement of the high energy lamps used in CD machines is that the optics of the instrument usually need to be purged with nitrogen gas to avoid ozone being created and reacting with surfaces of mirrors (and the operator). Nitrogen purging is also required for the sample compartment, when running below 200 nm to avoid having O2 absorbing the incident radiation. In practice, this means a moderate nitrogen flow rate (3−5 cm3/min) at all times with an increase to 10 or more cm3/minute when collecting data below 200 nm. 3.1.1. The Sample

Most CD spectrometers are single beam instruments and can tolerate a maximum absorbance of about 2 before the number of photons passing through the sample become too small to be measured accurately. If at all possible, measure an absorbance spectrum (against air if it is a double beam absorbance spectrometer) of your sample before measuring a CD spectrum. The voltage on the photomultiplier tube can usually be plotted on the screen at the same time as the CD spectrum is measured to give an indication of absorbance (see Note 1). Avoid Cl− and Br− as counterions if possible (see Note 2).

3.1.2. The Cuvette

For UV-visible CD, high quality quartz cuvettes (see Note 3) that transmit the full wavelength range of UV-visible light are required. In the visible region, glass may be used, but it is generally advisable to use quartz even here. Plastic cuvettes typically have high intrinsic birefringence so should be avoided. In any case, the need to run a baseline of each cuvette used (see below) removes the usual attraction of disposable plastic cuvettes. The default cuvette pathlength for drug-DNA work is usually 1  cm. Most CD spectrometers have cuvette holders of better design than absorption spectrometers and will properly hold a range of cuvette sizes (and shapes). So, if the sample has an absorbance signal greater than 2, try using a shorter pathlength cell (see Note 5). If much of the absorbance is due to solvent or buffer, this is a particularly attractive option, especially if the analyte concentration can be increased.

42

Rodger

3.1.3. The Baseline and Zeroing

A machine baseline (i.e. CD spectrum of air) measured on a standard CD spectrometer will not be flat. In addition, the cuvette used for an experiment will also have its own CD spectrum (see Note 3). So it is essential to collect a baseline spectrum of the solvent/buffer under the same conditions as the sample spectrum using the same cuvette in the same orientation and same position with respect to the light beam. Subtract the baseline spectrum from the sample spectrum to produce the final CD plot. It is usually the case that with small CD signals, even when the baseline is subtracted, the CD is not exactly zero outside the absorption envelope. However, if the spectrum is flat outside the absorption envelope, it may be zeroed by adding or subtracting a constant (either within the spectrometer software or using your chosen data plotting software). See Note 6 for the effect of light scattering.

3.1.4. The Parameters

CD spectrometers usually (though not always) scan from longer wavelengths to shorter ones. Select a wavelength range so that there is at least 20  nm of zero absorbance beyond the normal absorption envelope(s) of interest to enable zeroing (see above). The choice of the number of accumulations, scan speed and time constant or response time are all linked. The signal to noise ratio in a CD spectrum increases with: √n, (n is the number accumulations that are averaged); √t (t is the time over which the machine averages each data point, some instruments use the label response time which is related to t); and√I (I is the intensity of the light beam). To optimize signal to noise effects, t should be selected to be as large as possible subject to τ × s ≤ b (b is the 2 bandwidth and s the scan speed). If t is too long for the chosen s and b, then maxima of peaks (both positive and negative) will be cut off and their wavelengths shifted (see Note 7). Scan speeds of 100 nm/min, t  = 1 s, b = 1 nm, and a data step of 0.5 nm seem to be a good starting point parameter set for most experiments, wherein the samples have the broad band shapes usually found for solution samples. If the bandwidth is fixed, the instrument will be programmed to control the slit width directly. If a number of spectra with the same parameters are to be collected, it is worth investing the time in determining whether scan speed, t or b can be increased without affecting the resulting spectrum. If a collected spectrum looks too noisy, there are some post collection smoothing options, but it is preferable to collect good data (see Note 8).

3.1.5. Units of CD Spectroscopy

The relationship between CD signal (in absorbance units) and sample concentration and pathlength is analogous to the Beer– Lambert law for absorbance, namely: DA = (De)C

Circular and Linear Dichroism of Drug-DNA Systems

43

where C is the sample concentration in mol dm−3, l is pathlength in cm, and De is in units of mol−1 dm3 cm−1. A general summary of possible CD units and their interrelationship follows. In absorbance units, we write CD = Al - Ar Alternatively, in molar units it is CD = el – er where er is the molar extinction coefficient for the absorption of right circularly polarized light. The units used for extinction coefficients are almost always mol–1 dm3 cm–1, which gives values that are ten times greater than those in SI units. The moles to which one refers may be either in terms of molecules or, for DNA, in terms of the bases or base pairs (you need to state which you are using). The ellipticity, q, is obtained from the ratio of the minor and major axes of the ellipse traced out by the electric field vector of the elliptically polarized light that emerges from a circularly dichroic sample onto which linearly polarized light was incident. For small ellipticities (which is what we measure for DNA samples) q / radians =

2.303C (el − er ) 4

Upon converting to millidegrees, as commonly presented by CD machines we have: q / millidegrees = 32, 980C  (e  − er ) = 32, 980 (A − Ar )

The related molar ellipticity in degrees is defined as Mq =

100q C

where M is molar mass in g, q is in degrees, C is in moles dm–3, and l is in cm. Thus, M q = 3, 298 (e  − er ) = 3, 298De A related unit that is often used is the mean residue ellipticity. In this case, as mentioned above, the concentration is given in terms of total numbers of residues without reference to their identity; hence it is an average over the types of residue present in the macromolecule. Most CD machines will perform the conversions between units on request.

44

Rodger 6 4

CD / mdeg

2 0 –2 100-1 80-1 60-1 40-1

–4 –6 300

400

500 600 Wavelength / nm

30-1 20-1 10-1 700

Fig. 3. ICD spectra (obtained by subtracting the free metal complex spectra from the ct-DNA plus metal complex CD spectra) spectra for constant complex concentration (15  mM) and varying ct-DNA concentrations for (P)-[Ru2(phen)4L]4+, where L is a bis(pyrdylimine) tetradentate ligand that spans the two ruthenium centres. DNA base to metal complex ratios are indicated on the figure. All spectra were run in 50 mM NaCl and 1 mM sodium cacodylate buffer with 18.2 MW water. The cell pathlength was 1 cm

3.1.6. Binding Constants

It is often very useful to measure a series of spectra where some variable such as the ionic strength, or mixing ratio (drug:DNA ratio), or temperature is changed. If the molar induced CD (ICD) intensity changes but the shape of the spectrum remains the same during such an experiment, then it can be deduced that the drug binding mode is unchanged though the amount of bound ligand may have changed. An example is given in Fig. 3. The signal strength in such experiments is often low. Thus, in such an experiment, if possible, it is best to keep the ligand concentration constant as, then, any changes in ICD above 300 nm are more easily apparent. If the ligand binds in a single binding mode, or in a number of sites whose relative proportions are independent of DNA:drug ratio, then we may write Lb = a ´ ICD = ar where Lb is the concentration of bound ligand, r is the ICD signal and a is a proportionality constant. If you can draw a plot that looks similar to the one in Fig. 4 with a reasonably straight part followed by a curve, and ideally a leveling off part, then one of a number of the analysis methods may be used to determine equilibrium binding constants and site sizes (see ref. 7) (see Note 9). It is frequently the case with CD of DNA-ligand complexes that one cannot get end-point data as a second mode becomes operative. See Note 9 for a further discussion of this.

45

60

20

ligand:DNA 0.008:1 0.015:1 0.022:1 0.029:1 0.037:1

0.045:1 0.056:1 0.067:1 0.083:1 0.091:1

0.11:1 0.12:1 0.14:1 0.15:1 0.16:1 0.17:1

–20

70 60 50 ICD

Induced CD / mdeg

Circular and Linear Dichroism of Drug-DNA Systems

40 30 20 10

–60 220

0

240

260 280 Wavelength / nm

300

0

50

100 150 Ligand:DNA ratio

200

Fig.  4. Poly[d(G-C)]2 DNA (160 µM base) with an intercalated anthracene chromophore that is coupled to a cationic spermine group. ICD signals are the total CD minus the free DNA CD. The binding curve is derived from 262 nm data.

3.1.7. Induced CD of Intercalating Ligands

When planar aromatic ligands bind to DNA, their absorbance bands are typically shifted to longer wavelength, reduced in intensity, and are broadened. The ICD signals reflect this and occur at the longer wavelength with the broader band shape. If one knows the polarization of the ligand transitions, one can identify the orientation of an intercalated ligand in the pocket by CD. As originally shown by Schipper and Nordén (8), long axis polarized transitions of intercalators oriented with their long axis “parallel” to the DNA bases will have a negative induced CD signal in a random sequence DNA. Ones whose long axis pokes out into the DNA grooves will have a positive induced CD signal for long axis polarized transitions. The converse signs are to be expected for short axis polarized transitions (1, 9). Note this analysis requires knowledge of the transition moment polarizations of the ligand.

3.2. Linear Dichroism

The general spectroscopy principles of LD spectroscopy and its instrumentation requirements are the same as for CD spectroscopy. The major difference relates to the need to orient the samples. In addition, one must either (1) change the data mode on the instrument from CD to LD, or (2) insert a quarter wave plate and convert the output from mdeg to DA by dividing by 32,980. For samples that are uniaxially oriented, as may be assumed to be the case for drug molecules in films and DNA in shear flow, the LD signal is LD =

3A S (3cos 2 a − 1) 2

where A is the isotropic absorbance of the same sample in the same pathlength, cos 2a is the average of the cosine square of the angle the transition moment makes with the orientation direction, and S is the orientation parameter which equals 1 for a perfectly oriented sample and 0 for an unoriented one.

46

Rodger 0.005

LD / absorbance units

0 –0.005

ct-DNA 500:1 200:1 160:1 120:1 100:1 80:1 40:1 20:1 10:1

– 0.01 –0.015 – 0.02 –0.025 – 0.03 200

300

400 500 600 Wavelength / nm

700

Fig.  5. The LD of a tetracationic di-iron triple helicate binding with calf thymus DNA showing the effect on the LD if the ligand bends the DNA. ct-DNA (500 µM, 20 mM NaCl, 1 mM sodium cacodylate buffer pH = 7); DNA:ligand ratios are shown on the figure.

We usually assume (hope) that the average of the cosine squared (〈cos2 a〉) is the same as the square of the cosine of the average of a (〈cos2 a〉). Determining S is required for a quantitative analysis of LD spectra and is often extremely difficult. For the bases in B-DNA, it is accepted (10) that 〈a〉 ~86°. If a drug binding to DNA absorbs at the same wavelength as the DNA, then one is usually left with assuming that S remains unchanged for a free DNA and a ligandbound DNA. As Fig. 5 shows, this is not always the case: the LD signal at 260  nm decreases much more rapidly than is possible due to any possible positive LD signal from the ligand. If S increases, this is indicative (though not conclusive) of an intercalating ligand that stiffens the DNA. If S decreases, the DNA is being bent or made more flexible. 3.2.1. Film LD

LD data are most useful if the transition polarizations (direction of electron displacement) are known. For DNA bases, we know they lie in the plane, and thus are approximately perpendicular to the DNA helix axis. For drug molecules, we sometimes have to determine their polarizations, and one method is by film LD. For a molecule to be aligned in a stretched film, it must either be an integral part of the film or be associated sufficiently strongly that when the film is stretched the molecule follows the film alignment axis. In practice, most molecules can be aligned in one of the two films: polyethylene (PE) for non-polar molecules and polyvinyl alcohol (PVA) for polar molecules.

Circular and Linear Dichroism of Drug-DNA Systems

47

3.2.1.1. Alignment in Polyethylene Films

Polyethylene (PE) is microcrystalline, and when it is mechanically stretched along the manufacturer’s stretch direction (usually apparent when you hold the plastic up to the light and definitely apparent when you try to stretch it the wrong way), a molecular orienting environment is produced. The keys to success with PE film LD are the choice of PE (see Note 10), identifying the manufacturer’s stretch direction, and the degree of stretching. By convention, the parallel direction of the polarized light is usually taken to be horizontal (see Note 11), so put the film with the stretch direction horizontal with the beam going through the relevant part of the film. It is advisable not to stretch too close to the breaking point of the polymer since the film has a tendency to become opaque and to suddenly rip. With a film stretcher, a factor of ~5 with PE is fairly straightforward. To measure a film LD spectrum with a PE film, it is most convenient to stretch the film first, then drop a few drops of the solvent in which you have dissolved the analyte onto the film, allow the solvent to dry, and measure a baseline spectrum over the wavelength range of interest. Then, introduce the analyte into the stretched film by adding drops of the analyte solution onto the surface of the film and allowing the solvent to evaporate. Endeavour to collect the sample spectrum with the same part of the film as the baseline. Finally, subtract the baseline spectrum from that of the analyte.

3.2.1.2. Polyvinyl Alcohol Films

Polyvinyl Alcohol (PVA) is usually the best film for polar molecules; the film is transparent in the UV (above 200  nm) and visible regions of the spectrum, though it has a strong absorption over large regions in the infrared (11). For small molecules, it has to be used in a dry form (less than a few percent of water). PVA films are more difficult to prepare than PE films, however, the quality of data is often better. To prepare a PVA film, one mixes well hydrolyzed commercial PVA powder in cold water (10% w/v) to make a slurry, which is then heated to near boiling point to form a viscous solution. Then, divide the solution into two parts: An aqueous solution of the analyte (typically ~5  mM solution in water, but the aim is to have the final film with an absorbance maximum between 0.1 and 1) is then added to one half of the PVA solution. The same volume of water is added to the other half of the PVA solution. The two solutions are then cast onto a glass plate and allowed to dry over a period of one to three days in a well-ventilated dust-free place. Finally, the two films are removed from the glass plate (using a scalpel or other sharp blade), loaded into a film stretcher, and stretched by the same factor (typically ~2) at an elevated temperature by holding the films in the hot air from a hair dryer as they are being stretched. It is advisable to stretch a “safe” amount first, and then measure the spectra before trying a larger stretch. The greater the stretch

48

Rodger

factor, the greater the LD signal magnitude, until of course the film breaks. One can usually assume that the drug molecule is uniaxially oriented within the film. To determine the component long axis and short axis spectra, one looks for the most positive point in the reduced LD spectrum (the ratio of LD to absorbance) and assumes the absorbance at that wavelength is purely long axis (z) polarized in which case its reduced LD, LDr LDr (positive max imum) =

LD(z) = 3S A

Alternatively, if the longest wavelength LD is negative, it will often be better to look for the largest magnitude negative LDr and assume it is purely y polarized, so LDr (negative max imum) =

LD(y ) 3 =− S 2 A

The component spectra in the z and y directions are then (1, 11) Az (l) = A (l) +

2 LD(l) 3S

and

Ay (l) = Ax (l) = A (l) −

LD(l) 3S

An example is given in Fig. 6. 3.2.2. Flow Orientation of DNA

The shear forces of a polymeric molecule, such as DNA dissolved in water flowing past a stationary surface at about 1 m/s is sufficient to give it enough orientation that A // and A –| are different if the DNA is long enough (see Note 12). If the walls past which the DNA flows are quartz, then we can measure LD spectra in the visible and UV regions. If the shear flow is provided by a linear flow through system, such as an HPLC pump this is very expensive on sample. Wada in 1964 (12, 13) solved this problem with the invention of a Couette flow cell, wherein the sample is endlessly flowed between two cylinders one of which rotates and one of which is stationary. A microvolume Couette flow cell that requires less than 50 µL of sample rather than the mLs of the previous Couette cells is now available (14, 15). The large volume cell described in reference (16) and a microvolume cell are illustrated in Fig. 7. To load a large volume cell of the kind illustrated in Fig. 7, use a micropipette (usually of 1 mL volume) and add sufficient sample to ensure the meniscus is above the windows through which the light passes. Turn on the rotation, and look down the light path to ensure there is indeed enough sample. It is easy to add additional material such as a DNA binding drug to such a cell to measure LD as a function of mixing ratio, pH, etc. (see Note 4). The microvolume option, as the name implies, requires less sample and is also available as a thermostatted unit (17). It, however,

Circular and Linear Dichroism of Drug-DNA Systems

49

0.8 0.7

Absorbance

0.6 0.5 0.4 0.3 0.2 0.1 0

300

400

500

600

700

0.15

LD

0.1 0.05 0 –0.05

300

400 500 600 Wavelength / nm

700

Absorbance

1.5 Ay (1) Az(1)

1

0.5

0 300

400 500 600 Wavelength / nm

700

Fig.  6. PVA film absorbance LD spectrum for a tetracationic di-iron triple helicate prepared as described in the text. Also shown are the component spectra determined as described in the text.

Lid Housing 100 mm

Quartz rod 2.4 mm Quartz Capillary OD 5 mm ID 2.9

Fig. 7. Large volume (16) and microvolume Couette flow (14, 15) cells.

500 m m path length

Rodger

has the disadvantages of being more difficult to load without trapping an air bubble in the light path; furthermore, adding material after the initial loading is also not as straightforward as with the larger cell. When collecting LD data, most of the issues mentioned above for CD must be considered. In particular, if the photomultiplier tube is not receiving many photons, it becomes unreliable. As with CD, the ideal absorbance is about 1 but a range from 0.1 to 2 is usually satisfactory. The approach of ensuring that the voltage on the photomultiplier tube is less than 600 V (see Note 1) is not always appropriate with LD. Almost by definition, the samples that one uses for LD are large enough to scatter light. This means that a significant percentage of the light reaching the photomultiplier tube from a light scattering sample may be scattered light rather than the unabsorbed photons transmitted directly through the sample, especially at lower wavelengths when xenon lamp instruments are struggling for light intensity. It is important to check whether the Beer–Lambert law is being followed by diluting the sample and checking the spectrum. False maxima are otherwise observed as illustrated in Fig. 8. For both CD and LD, it is essential that the light beam being detected passes only through the sample and not stray parts of the cuvette. The problem is particularly acute for the micro­ volume Couette flow cells, wherein it is essential that the light beam passes through the centre of the rod. The image of the light beam should be investigated with all new experimental arrangements or when the sample compartment configuration is changed (see Note 13). 0.14 0.12 0.1 LD / absorbance units

50

0.08 0.06 0.04 0.02 0 –0.02

200

220

240 260 280 Wavelength / nm

300

320

Fig. 8. Far UV LD spectra showing the apparent shift to shorter wavelength of the maximum signal as the concentration of F-actin (a fibrous polymer) is reduced. F-actin concentrations 93; 74; 62; 53 and 12 µM (the true spectrum, solid line).

51

Circular and Linear Dichroism of Drug-DNA Systems

a

b

0.002

0

–0.002

LD / absorbance units

LD / absorbance units

0

[Ethidium Bromide] / µM –0.004

0 5 10 15 20 25 30 35 40 45 50

–0.006 –0.008 –0.01 –0.012

0.01

200

250

300

350 400 450 Wavelength / nm

500

–0.01 –0.02

[DNA] 1000 µM [DNA] 1000 µM + [Hoechst] 50 µM

–0.03 –0.04 –0.05

550

600

–0.06

200

250

300

350 400 450 Wavelength / nm

500

550

600

Fig. 9. Linear dichroism of (a) ethidium bromide bound to calf thymus DNA (200 µM, 20 mM NaCl, pH = 7, ethidium concentrations shown on the figure); (b) Hoechst 33258 (50 µM) in calf thymus DNA (1,000 µM, 20 mM NaCl, pH = 7).

Figure  9 shows typical LD data for drug-DNA systems. Fig. 9a is the flow LD spectrum expected for an intercalator whose planar aromatic structure and alignment parallel to the DNA bases ensure that it has a negative LD spectrum of the same shape as its absorbance spectrum. If anything, the ligand’s reduced LD is larger than that of the average DNA base as it stiffens the DNA. Long axis polarized transitions of groove binders, by way of contrast, give a positive signal as shown in Fig. 9b. The iron triple helicate, the film LD of which was given above, bends the DNA significantly (as shown in Fig. 5), so the flow LD spectrum can also be used as a measure of DNA bending, but gives no idea of the local orientation of the molecule on the DNA. LD can be used to probe the indirect effects of ligand binding. For example, when a restriction enzyme cleaves DNA taking it from supercoiled to linear DNA, and then to shorter linear pieces of DNA (if the DNA has multiple restriction sites), we see first an increase in LD as the DNA lengthens then a decrease as it shortens (18).

4. Notes 1. The smaller the number of photons incident on the photomultiplier tube, the higher the voltage applied to it. Thus, the high tension (HT) voltage on a CD machine is a rough guide to the absorbance of a sample. Some instruments have reasonable HT to absorbance conversion routines. The HT voltage can usually be monitored while the CD spectrum is being collected. It is advisable to take this option even with routine samples. HT voltages above 600 V are indicative of a sample that is absorbing too highly. However, if the sample is prone to

52

Rodger

scatter light, even 600  V may not be safe. Changing the concentration or pathlength by e.g. a factor of two and seeing if the signal scales accordingly should be undertaken. 2. Salts such as Cl− and Br− absorb highly from ~215 nm, and therefore be kept at low concentrations. F− may be an alternative. 3. Either cylindrical or rectangular cuvettes may be used for CD. Cylindrical cells are usually deemed to have lower birefringence (baseline CD) than rectangular cuvettes; however, if UV and CD “matching” is requested when the cuvettes are purchased, rectangular cuvettes seem to be equally good. It should be noted that even “CD matched” cuvettes have slightly different intrinsic spectra. The CD spectrum of most cuvettes also varies with the part of the cuvette through which the light passes. Rectangular cells have a number of advantages over cylindrical cuvettes for 1  mm and longer pathlength experiments. They are cheaper, may be used in standard absorption spectrophotometers (so CD and normal absorption data may be collected on exactly the same sample), and may be used for serial titration experiments as~60% of a rectangular cell can be empty for the first spectrum and gradually filled (see Note 4). However, rectangular cuvettes require more sample. 4. Titration experiments where spectra are collected as a function of concentration, ionic strength, pH, etc. often involve adding solution to the cuvette. A simple way to avoid dilution effects is as follows. Consider a starting sample that has concentration x M of species X. Each time y µL of Y is added, also add y µL of a 2x M solution of X. The concentration of X remains constant at x M. An infinite number of variations on this theme are possible. 5. If pathlengths of 0.1  mm or less are required, it is probably best to use demountable cuvettes wherein the sample is dropped onto a quartz disk or plate that is etched to a predefined depth; and then, another quartz disk is carefully placed on top. In this case, sample recovery is very difficult so the smaller volume of cylindrical cells makes them more attractive than rectangular ones. Any cell holder for demountable cells must be located perpendicular to the light beam with the whole incident light beam passing through the cell windows, when data is being collected. The pathlength of demountable cells must be determined independently and by each user as the pressure used to assemble them affects the pathlength whose value is seldom as specified by the manufacturer. Potassium chromate in basic solution has an extinction coefficient at 372  nm of 4,830 mol–1/dm3/cm–1 and is ideal for determining pathlengths if an accurate balance is available to prepare the solutions. For many absorbance instruments, this means rectangular cells are to be preferred. 6. If the baseline is not flat outside the absorbance band, then this probably means that either there is a very weak absorbance

Circular and Linear Dichroism of Drug-DNA Systems

53

 band that has a large dissymmetry factor (and hence large CD signal compared with its normal absorbance intensity), or more probably, there is some light scattering of the sample. Sources of light scattering include slightly dirty cuvettes (inside or outside), undissolved sample, or condensation or aggregation of samples (a particular problem with DNA when highly charged cations are added). The scattering from fully solvated DNA itself is usually too small to be detected, however, if LD signals are small the scattering is apparent as a sloping signal outside the absorbance region. 7. A control scan using τ ' = τ (or s¢ = 2s) should be used to 2 check that spectra are not being distorted by the chosen parameters. The data interval determines how often a data point is collected. If your instrument works in a stopped scan mode (it stops at each point to collect data) then this parameter determines the scan speed. Some instruments let you deal with scan speed and data collection time more or less independently, in which case you need to ensure that you are not distorting the spectrum by scanning too quickly. The advantage of independent scan speed and time constant parameters is that with broad bands, the user can save time without collecting distorted spectra. However, care must be taken to ensure that this is indeed happening. 8. Sometimes, the sample concentration or the time available means the spectra collected are very noisy. There are many options for smoothing the data (some of them within the instrument software) in such cases, however, unless it is obvious “to the eye” what the result should be then avoid smoothing the data as you will probably either introduce or remove structure. To ensure the result of smoothing is reasonable, always overlay the noise-reduced result on the original data set and use your eye to decide the validity of the transformed spectrum. If it does not “look right”, reject the result. 9. The method originally developed by Crothers (19) for determining binding constants is widely used for DNAdrug systems, but it is strictly only valid for drugs which occupy one base pair as the binding site. An alternative method, which does not have this restraint is the induced spectroscopy method, which also does not require saturated binding data (1, 7). 10. Our current experience is that the best source of PE is the plastic bags supplied with micropipette tips. To stretch this PE, you need to have a mechanical stretcher into which the film can be fixed. In the absence of a film stretcher to stretch and hold the PE, light weight magazine wrappers made of PE can usually be stretched by hand and fixed into the light beam of the spectrometer using blue tack and one wall of the sample compartment.

54

Rodger

11. Most LD spectrometers take parallel as horizontal; however, this needs to be confirmed especially if a quarter wave plate is being used. 12. The DNA length lower limit for LD is ~300 basepairs (twice the persistence length), but this is hard work; 1,000 base pairs or more give much better data, and for drug-DNA, interactions are advisable. 13. Business cards are ideal for inserting in the light beam at ~ 550 nm (the green light most easily detected by our eyes) to see the beam width and height, though note that the beam width is dependent on the instrument slit width, which in turn may be designed to depend on the lamp energy, so may be larger in the UV region than at 550 nm. References 1. Rodger A, Nordén B (1997) Circular dichroism and linear dichroism. Oxford University Press, pp 150 2. Nordén B, Kubista M, Kuruscev T (1992) Linear dichroism spectroscopy of nucleicacids. Q Rev Biophys 25:51–170 3. Nordén B (1978) Applications of linear dichroism spectroscopy. Appl Spectros Rev 14:157–248 4. Dafforn TR, Rodger A (2004) Unravelling the configuration of protein fibres and membrane proteins. Curr Opin Struct Biol 14:541–546 5. Rodger A, Marrington R, Geeves MA, Hicks M, de Alwis L, Halsall DJ, Dafforn TR (2006) Looking at long molecules in solution: what happens when they are subjected to Couette flow? Phys Chem Chem Phys 8:3161–3171 6. Damianoglou A, Crust EJ, Hicks MJ, Howson SE, Knight AE, Ravi J, Scott P, Rodger A (2008) A new reference material for UV-visible circular dichroism spectroscopy. Chirality 20:1029–1038 7. Stootman FH, Fisher DM, Rodger A, AldrichWright JR (2006) Improved curve fitting procedures to determine equilibrium binding constants. Analyst 131:1145–1151 8. Schipper PE, Nordén B, Tjerneld F (1980) Determination of binding geometry of DNAadduct systems through induced circulardichroism. Chem Phys Lett 70:17–21 9. Schipper PE, Rodger A (1983) Symmetry rules for the determination of the intercalation geometry of host/guest systems using circular dichroism: a symmetry adapted coupled-oscillator model. J Am Chem Soc 105:4541–4550 10. Chou PJ, Johnson WC (1993) Base inclinations in natural and synthetic DNAs. J Am Chem Soc 115:1205–1214

11. Matsuoka Y, Nordén B (1982) Linear dichroism studies of nucleic acid bases in stretched poly(vinyl alcohol) film. Molecular orientation and electronic transition moment directions. J. Phys Chem 86:1378–1386 12. Wada A (1964) Chain regularity and flow dichroism of deoxyribonucleic acids in solution. Biopolymers 2:361–380 13. Wada A (1972) Dichroic spectra of bio­ polymers oriented by flow. Appl Spectros Rev 6:1–30 14. Marrington R, Dafforn TR, Halsall DJ, Rodger A (2004) Micro volume Couette flow sample orientation for absorbance and fluorescence linear dichroism. Biophys J 87: 2002–2012 15. Marrington R, Dafforn TR, Halsall DJ, Hicks MR, Rodger A (2005) Validation of new microvolume Couette flow linear dichroism cells. Analyst 130:1608–1616 16. Rodger A (1993) Linear dichroism. In: Riordan JF, Vallee BL (eds) Methods in enzymology, vol 226. Academic, San Diego, pp 232–258. 17. Marrington R, Seymour M, Rodger A (2006) A new method for fibrous protein analysis illustrated by application to tubulin microtubule polymerization and depolymerization. Chirality 18:680–690 18. Hicks MR, Rodger A, Thomas CM, Batt SM, Dafforn TR (2006) Restriction enzyme kinetics monitored by UV linear dichroism. Biochemistry 45:8912–8917 19. Schmechel DEV, Crothers DM (1971) Kinetic and hydrodynamic studies of the complex of proflavine with polyA.polyU. Biopolymers 10:465–480

Chapter 4 Drug Binding to DNA⋅RNA Hybrid Structures Richard T. Wheelhouse and Jonathan B. Chaires Abstract DNA·RNA hybrid duplexes are functionally important structures in gene expression that are underutilized as potential drug targets. Several tools are described here for the discovery and characterization of small molecules capable of the selective recognition of DNA·RNA hybrid structures. Competition dialysis and thermal denaturation of mixtures of polynucleotide structures can be used to identify small molecules that bind selectively to DNA·RNA hybrids. An assay that measures small molecule inhibition of RNase H can be used to measure a functional response to these ligands. Key words: DNA·RNA hybrid, RNase H, Dialysis, Thermal denaturation, Enzyme inhibition, UV spectrophotometry

1. Introduction The DNA⋅RNA hybrid helix was first discovered in 1960, only seven years after the discovery of the famed DNA double helix (1, 2). The hybrid structure was immediately recognized as a key component of genetic information transfer, even though mRNA and tRNA had yet to be discovered (2). After nearly half a century, DNA·RNA hybrid duplexes have yet to be exploited as targets in drug design, despite their unquestioned functional significance. This stands in contrast to other noncanonical nucleic acid structures (e.g., DNA triplexes and quadruplexes, Z-DNA, RNA secondary structures), all of which have been recognized as potential therapeutic targets. The reasons underlying the failure to target the DNA·RNA hybrid are most likely related to uncertainties and misunderstandings about the precise nature of the hybrid duplex structures, along with the lack of available lead ligands (whether of natural or synthetic origin) and, finally, the dearth of convenient assays to assess structural-selective binding to the hybrid. K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_4, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

55

56

Wheelhouse and Chaires

Structure- and sequence-selective ligands for DNA·RNA hybrids have a variety of potential pharmaceutical applications. For instance, the first-formed product of HIV reverse transcriptase is a DNA·RNA hybrid duplex. Generation of the viral DNA duplex and its integration into the host genome are dependent on the digestion of the RNA strand of this hybrid by RNase H (3). Loading DNA·RNA hybrid duplexes with small-molecule ligands inhibits RNase H (4, 5) and thus provides a new route to the control of HIV infection. The tumour-specific enzyme, telomerase, is a specialized reverse transcriptase, with protein structure domains closely-related to the HIV enzyme (6). A DNA·RNA hybrid duplex is generated in the course of the telomerase reaction, so telomerase is also amenable to such a small-molecule targeting strategy (7, 8). Indeed, the thermodynamic stability of the DNA-RNA hybrid appears to regulate telomerase activity (9). A long-term goal of nucleic acid-directed drug discovery is the control of gene expression. This may potentially be achieved through sequence-selective intervention in the RNA polymerase products of normal gene expression or binding to the Okazaki fragment RNA·DNA hybrid formed at the initiation of DNA replication (10). The consensus view of hybrid duplex structures in solution is that the overall helix conformation lies somewhere between the global B-form adopted by pure DNA and the A-form of pure RNA duplexes. The conformation of each strand is driven by the sugar puckers, which in turn are determined by the b-effect of the 3¢ and 2¢ substituents of the furanose rings. Thus in the DNA strand, “North” puckers predominate, driving that strand toward a B-conformation. In contrast, for the RNA strand, more heterogeneous sugar puckers are detected (11), resulting in an overall A-like structure. The duplex formed on combination of such diverse strands differs significantly from both pure DNA·DNA and RNA·RNA duplexes (12). The structures of different DNA·RNA hybrids can differ, depending on whether the DNA strand is purine or pyrimidine (13, 14). DNA·RNA hybrids of differing sequence may thus present different backbone conformations and groove geometries, distinctive features that may be selectively targeted by small molecules. Intuitive attempts to identify ligands for DNA·RNA hybrids using established DNA duplex groove binders or intercalators are, perhaps, doomed. The grooves of the hybrid are not recognized by the classic minor groove binders (for examples, see data in Fig. 1). Among intercalators, matching the planar surface shape and size of a ligand to the cross-sectional dimensions of triplex and quadruplex DNA proved a fruitful approach to the identification of structure-selective binding (15). However, the search for structure selective intercalators of DNA·RNA hybrids is constrained by the similarity of the Watson–Crick basepairs, and hence intercalation site dimensions, to those of the

Drug Binding to DNA·RNA Hybrid Structures

57

Fig. 1. Melting of mixtures data for “classic” duplex DNA minor groove binders (0.25 mM) that imply lack of binding to structures other than the DNA duplex

pure DNA or RNA duplexes. Nonetheless, some success has been achieved in identifying ethidium as possessing a marked binding preference for the poly(A)·poly(dT) hybrid duplex (16) and redesigning actinomycin to enhance hybrid selectivity (17, 18). The few hybrid binding ligands so far reported are dominated by ethidium and aminoglycoside structures (4, 8, 19, 20) and have recently been the subject of a comprehensive review (19). All of these investigations focus on measuring binding phenomena using spectrophotometric, fluorimetric and calorimetric techniques but offer little rationalization of the intermolecular interactions that drive the recognition event. Herein, we describe three assays used in our laboratories that are useful in exploring the structure and sequence-binding preferences of small molecules that target DNA⋅RNA hybrid duplexes. Competition equilibrium dialysis (21) delivers a detailed and thermodynamically-rigorous profile of the binding strengths and preferences of an investigational ligand across a broad range of up to 48 different nucleic acid structures and sequences (22–25). The assay, in its current form, typically includes both poly(dA)·poly(U) and poly(A)·poly(dT) hybrids alongside the counterpart pure DNA and pure RNA duplexes. The melting of mixtures assay (26) is an extension of the well-established DNA thermal denaturation assay (27, 28) that detects the effects of added ligand on thermal denaturation of four duplexes simultaneously. It is valuable for rapidly assessing binding preferences among DNA, RNA and hybrid duplexes; for the hybrid systems, it also provides evidence of selectivity between dA⋅U and A⋅dT sequences. Finally an RNase H assay is included that provides a pharmaceutically-relevant biological endpoint for studies of

58

Wheelhouse and Chaires

ligands for DNA·RNA hybrids. Various protocols for the detection and quantitation of telomerase inhibition are described in the literature (29–31). All three assays described here are UV spectrophotometric methods, directly detecting either ligand or nucleic acid without the use of radio-labels or fluorescent tags. This latter is particularly important since the presence of large aromatic groups, typical of fluorophores and quenchers, will itself alter the nucleic acid structures and thereby ligand-binding events. 1.1. Overview of Methods and Illustrative Results 1.1.1. Competition Dialysis

The competition dialysis assay is a simple extension of equilibrium dialysis binding assays (32, 33). In the competition dialysis assay, an array of nucleic acid structures is dialyzed against a common test ligand (“drug”) solution. Each structure in the array is at an identical concentration and is contained within its own dialysis unit, isolated from all other structures. All structures are in contact with the same ligand solution. At equilibrium, the free ligand concentration is identical for all structures in the array, but each structure will bind ligand according to its affinity for the ligand. The difference in the amount of ligand bound to each structure is proportional to the association constant of the ligand for that structure. A simple bar graph reveals, at a glance, the sequence or structural preferences of the ligand. Fig.  2 shows a representative result for the DNA intercalator ethidium. In this case, binding to an array of 13 nucleic acid structures reveals that ethidium binds with a slight, but significant preference to a DNA⋅RNA hybrid structure.

Fig. 2. Competition dialysis data for the interaction of ethidium with an array containing 13 nucleic acid structures. The data show preferential binding to the DNA·RNA hybrid poly(rA)·poly(dT) over all other structures and sequences

Drug Binding to DNA·RNA Hybrid Structures

59

Competition dialysis is particularly powerful when examining a library of small molecules. In such a case, competition dialysis difference plots can be used to highlight those ligands that bind with high affinity and selectivity to a particular structure or sequence of interest (34, 35). Fig. 3 shows an example in which ligands that bind to a DNA·RNA hybrid structure were sought. In this case, the binding of a library of 126 nucleic acid binding compounds to an array of 13 nucleic acid structure and sequences was studied. Fig. 3 shows three different plots that emphasize the difference binding to a DNA·RNA hybrid structure, to a standard DNA duplex, and to an RNA duplex. Four compounds stand out as having higher than average affinity for the DNA·RNA hybrid, while showing clear preferential binding to the hybrid over duplex DNA or RNA. These compounds are ellipticine (III), the topoisomerase poison TAS103 (I), thiazole orange (II), and ethidium (IV). These compounds are all DNA intercalators, and while they show modest preferential binding to the DNA·RNA hybrid, the intercalator motif may not provide optimal lead compounds

Fig. 3. Differential binding plots derived from competition dialysis studies of 126 compounds with 13 nucleic acids structures. (a) Difference in the amount bound to poly(rA)·poly(dT) relative to the average amount bound . Positive values indicate higher than average affinity. (b) Difference in the amount bound to poly(rA)·poly(dT) relative to duplex calf thymus DNA. Positive values indicate preferential binding to the hybrid form over this standard duplex DNA. (c) Difference in the amount bound to poly(rA)·poly(dT) relative to poly(rA)·poly(U). Positive values indicate preferential binding to the hybrid form over RNA. Four compounds are highlighted (I–IV) that bind with higher than average affinity to the hybrid structure while showing preferential binding to the hybrid form over DNA or RNA. The construction and use of differential binding plots is fully described in (34, 35).

60

Wheelhouse and Chaires

for more stringent targeting of hybrid structures and sequences. The primary point of Fig. 3 is to illustrate the power of compe­ tition dialysis for screening libraries to reveal binding preferences. Detailed, step-by-step protocols for the competition dialysis assay and for the analysis of competition dialysis data have been presented in several publications (21, 22, 25, 34, 35). These protocols will not be repeated here, and the interested investigator should consult the published protocols. Competition dialysis is an important element of the toolbox for the discovery of DNA×RNA hybrid binders, as this brief discussion indicates. 1.1.2. Melting of Mixtures

Thermal denaturation is a powerful tool for monitoring binding to nucleic acids (27, 28, 36, 37). Ligands that bind to structured nucleic acids stabilize the ordered structure and elevate the melting temperature for the denaturation of that structure. The magnitude of the change in melting temperature is a complex function of binding affinity and stoichometry (28). Qualitatively, an increase in the melting temperature for a nucleic acid structure provides unambiguous evidence for binding. A recent extension of the thermal denaturation assay resulted from the simple expediency of making a mixture of different nucleic acid structures that could be distinguished by their melting temperatures (26). Addition of low binding ratios of a test ligand altered the melting temperature of the nucleic acid structure to which it bound most avidly, providing a simple, rapid evaluation of its structural preference. A mixture containing DNA, RNA, and two DNA·RNA hybrid forms was described and tested (26), and a protocol for that preparation and use of that mixture will be described here. Examples of melting of mixtures data are presented for established DNA minor groove binders (Fig. 1) and a family of analogues of the biarylpyrimidine 1a (Fig.  4a.) These demonstrate the value of this assay in rapidly and simultaneously assessing thermal stabilization and selectivity among the four nucleic acid structures. Thus, although analogue 1d showed the greatest DTm for the poly(dA)·poly(U) duplex, it was the analogue 1b that showed the most selectivity between the structures, with minimal shifting of the DNA duplex. The data of Fig. 4 also show a direct correlation between the ligand-induced DTms of the poly(dA)·poly(U) duplex (Fig.  4a, inset) and the potency of RNase H inhibition (Fig. 4b). The rank order of DTms is the same as the rank order as inhibitors of RNase H digestion of that hybrid duplex. Thus, there is a strong correlation between strength of the ligand–hybrid interaction and RNase H inhibitor potency; this means that the relatively quick and inexpensive mixed melting assay may be used as a surrogate for the more expensive and time-consuming RNase H assay.

Drug Binding to DNA·RNA Hybrid Structures

61

Fig. 4. Nucleic acid melting and RNase H inhibition data for a family of analogues of biarylpyrimidine 1a; assays performed according to the protocols described in the text. (a) Melting mixtures assay, 4 × 10 mM bp nucleic acids, 0.25 mM ligand; inset shows an expansion of the poly(dA)·poly(U) melting curves. (b) inhibition of RNase H digestion of poly(dA)·poly(U) hybrid duplex; nucleic acid 10 mM bp; inhibitors, 3 mM

62

Wheelhouse and Chaires

1.1.3. Ribonuclease H Assay

This spectrophotometric assay is based on that previously described by Rasche et al (38). The release of nucleotides from the hybrid duplex substrate is followed as an increase in A260. The overall DA260 change detected for complete digestion of the RNA strand of a hybrid duplex approaches that observed following loss of the duplex secondary structure in the UV thermal melting experiment. This is a direct assay that monitors the disappearance of substrate: the signal is independent of the inhibitor under investigation. The scale on which the reaction is performed (enzyme and substrate concentrations) is determined by the sensitivity of the UV spectrophotometer (see Note 1) and there is no need for either radioactive labelling or the introduction of fluorescent tags. Solutions of the hybrids 10 mM bp were found to give a final DA260 » 0.06, appropriate for monitoring the reaction. Typical time courses for the digestion of poly(dA)·poly(rU) and poly(A)·poly(dT) are shown in Fig 5a. A burst phase is apparent at the start of the reaction which is typical of “ping-pong” kinetics involving a covalently-linked enzyme-substrate intermediate. This phase was usually flowed by a linear phase from which the vinital was estimated as the gradient of a best fit straight line; at high inhibitor concentrations, curvature of the line could complicate estimation of this gradient. As with any enzyme reaction, it was first necessary to determine the KM of the enzyme for the specific substrate under investigation. Figure  5b shows the dependence of vinit on [poly(dA)· poly(rU)] from which KM was determined (KM = 9.99±1.39 mM for poly(dA)·poly(rU)). Subsequent kinetic runs were performed with [S] = 4KM to ensure maximal velocity (39). Formally, the inhibitor strength should be measured as the Ki. However, this requires repetitions of the assay at ranges of [I] and [S] which may become expensive in both time and the quantity of enzyme consumed; for most purposes it is adequate to measure the IC50. Data obtained for a range of concentrations of the most potent inhibitor, compound 1d, and the curve-fit for the determination of the IC50 (6.53±0.03 mM) are shown in Fig. 5c. The RNase H assay can be used in a survey mode to screen series of compounds at a fixed concentration, in order to select the most potent inhibitors for detailed investigation. Figure 4b shows data for a range of analogues of biarylpyrimidine 1a inhibiting digestion of the poly(dA)·poly(U) hybrid. Significantly, the rank order of potency of inhibition thus detected, followed the magnitude of the Tm shifts detected in the melting of mixtures assay (Fig. 4a), even though this melting assay was performed at nonsaturating concentrations of ligand using a different buffer system.

Drug Binding to DNA·RNA Hybrid Structures

63

Fig. 5. Kinetics of the RNase H reaction. (a) A260 change during the course of reaction for the RNase H degradation of poly(dA)·poly(rU) (10 and 20 mM bp) and poly(rA)·poly(dT) (10 mM bp) hybrid duplexes showing the initial burst phase followed by the linear section of the curve used to estimate vmax; (b) Dependence of v on the concentration of poly(dA)·poly(rU) substrate from which KM was determined (KM = 9.99±1.39 mM, R = 0.996); (c) Effect of varying concentration of inhibitor 1d when [poly(dA)·poly(rU)] = 40 mM, from which IC50 = 6.53±0.03 mM (R = 0.997).

2. Materials 2.1. Competition Dialysis

Materials for the competition dialysis assay in a variety of different formats containing arrays of 13, 19 and 48 nucleic acid structures have been presented in several publications (21, 22, 25, 34, 35).

2.1.1. Dialysis Units

The first generation assay used 1.0 or 0.5 mL DispoDialyzer® units (Spectrum Medical Industries, Inc.; www.spectrumlabs.com) of the desired molecular weight cutoff, usually 3,500 MWCO.

64

Wheelhouse and Chaires

Subsequently, a less expensive and overall more suitable alternative was found, Pierce Slide-A-Lyzer® MINI Dialyzer units (Pierce Chemical Company; www.piercenet.com). Dialysis units may be reused several times after careful rinsing with buffer solution. 2.1.2. Nucleic Acids

The nucleic acid structures used in one version of the competition dialysis assay are listed in Table 1. All samples are dissolved in BPES buffer, consisting of 6  mM Na2HPO4, 2  mM NaH2PO4, 1 mM Na2EDTA, 0.185 M NaCl, pH 7.0. 1. Samples of DNA from Clostridium perfringens calf thymus and Micrococus lysodeikticus (Sigma Chemical Co., St. Louis, MO) were sonicated, phenol extracted, and purified as previously described (40). 2. Poly(dA), poly(dT), poly(U), poly(dC), poly(dA).poly(dT), poly(dAdT), and poly(dGdC) (Pharmacia Biotech, Inc., Piscataway, NJ). 3. Poly(rA) and poly(A).poly(U) (Sigma Chemical Co). 4. Deoxyoligonucleotides 5¢T2G20T2, 5¢G10T4G10 and 5¢AG3 (TTAG3)3 (Research Genetics, Huntsville, AL). Synthetic single-stranded and duplex polynucleotides were used without further purification. The poly(A).poly(dT) DNARNA hybrid was prepared by mixing poly rA and poly dT in a 1:1 molar ratio, heating to 90°C, and slowly cooling to room temperature. The DNA and RNA triplex forms prepared by mixing poly(dA).poly(dT) with poly(dT) (or poly(A). poly(U) with poly(U)) in a 1:1 mole ratio, heating to 90°C, and slowly cooling to room temperature. Quadruplex DNA and i-motif DNA were prepared by heating the oligonucleotides (5¢T2G20T2), (5¢G10T4G10), (5¢AG3(T2AG3)3) or poly(dC) to 90°C for 2  min, slowly cooling to room temperature, and then equilibrating for 48 h at 4°C before use. Left-handed Z DNA was prepared by bromination of poly(dGdC) as previously described (41).

2.1.3. Concentration Determinations

Concentrations of nucleic acid samples were determined by UV absorbance measurements using the extinction coefficients and absorption maxima listed in Table 1. Stock solutions of nucleic acids were prepared and maintained at 75 mM concentration, using the monomeric unit of each polynucleotide as the concentration standard. That means nucleotides (nt) for single stranded forms, base pairs (bp) for duplex forms, triplets for triplex forms and quartets for quadruplex forms. The extinction coefficients listed in Table 1, all refer to these concentration standards.

Drug Binding to DNA·RNA Hybrid Structures

65

Table 1 Nucleic acid conformation and samples used in competition dialysis experiments, together with data for the spectrophotometric quantitation of polynucleotides (26) Conformation

DNA/oligonucleotide

l (nm)

e (M−1cm−1)

Single strand purine

poly(dA)

257

8,600

poly(A)

258

9,800

poly(dT)

264

8,520

poly(U)

260

9,350

C. perfringens (31%GC)

260

12,476

calf thymus (42%GC)

260

12,824

M. lysodeikticus (72%GC)

260

13,846

poly (dA). poly (dT)

260

12,000

poly(dAdT)

262

13,200

poly(dGdC)

254

16,800

DNA–RNA hybrid

poly(A). poly(dT)

260

12,460

Duplex RNA

poly(A). poly(U)

260

14,280

Z DNAa

Br- poly (dGdC)

254

16,060

Triplex DNA

poly(dA). [poly (dT)]2

260

17,200

Triplex RNA

Poly(A). [poly(U)]2

260

17,840

Quadruplex DNA 1

(5¢T2G20T2)4

260

39,267

Quadruplex DNA 2

5¢AG3(TTAG3)3

260

73,000

Quadruplex DNA 3

(5¢G10T4G10)x

260

39,400

i-motif

[poly(dC)]4

274

27,200

Single-strand pyrimidine

Duplex DNA

Key: l: Wavelength; e: molar extinction coefficient

2.2. Melting of Mixtures

1. Polynucleotides poly(rA), poly(dA), poly(rU), poly(dT) (Sigma Aldrich, Poole, UK) were used without further treatment. 2. Assay buffer ¼ BPES: 1.5 mM Na2HPO4, 0.5 mM NaH2PO4, 0.25 mM, Na2EDTA, 46.25 mM NaCl, pH = 7.0.

2.3. Ribonuclease H Assay

1. RNase H was obtained from Promega UK Ltd. (Southampton, UK) in HEPES buffer–glycerol (1.5 U ml−1) and used directly, see Note 2. 2. Reaction buffer: 50 mM Tris pH 8.0 (Note 3), 50 mM NaCl, 10 mM MgCl2. All materials Sigma-Aldrich molecular biology grade. 3. Polynucleotides poly(dA), poly(dT), poly(U), poly(A) were used as supplied by Sigma.

66

Wheelhouse and Chaires

3. Methods 3.1. Competition Dialysis

Detailed, step-by-step protocols for the competition dialysis assay in different formats have been presented in several publications (21, 22, 25, 34, 35). 1. For each competition dialysis assay, place 400  ml of the dialysate solution containing 1 mM test ligand concentration into a beaker. 2. Pipette 180  uL (at 75 mM monomeric unit) of each of the DNA samples listed in Table 1 into a separate Slide-A-Lyzer® MINI dialysis unit with 7,000 molecular weight cutoff membrane. Place all 19 dialysis units in a MINI dialysis flotation device (Pierce Chemical Company) and then place the whole unit in the beaker containing the dialysate solution. 3. Cover the beaker with parafilm, wrap the beaker in foil, and equilibrate with continuous stirring for 24 h at room temperature (20–22°C). 4. At the end of the equilibration period, carefully collect DNA samples in the corner of the mini dialysis unit and transfer to microfuge tubes. Add 10% sodium dodecyl sulfate (SDS) to DNA samples to a final concentration of 1% (w/v) (SDS). 5. The total concentration of drug (Cf) with each mini dialysis unit is then determined spectrophotometrically using wavelengths and extinction appropriate for each ligand. The free ligand concentration (Cf) is determined spectrophotometrically using an aliquot of the dialysate solution, although its concentration usually does not vary appreciably from the initial 1 mM concentration. 6. The amount of bound drug is determined by difference, Cb = Ct–Cf. Plot as a bar graph using Orgin software (Version 5.1, Microcal, Inc., Northampton, MA) or any other available graphics package suitable for plotting and analysis.

3.2. Melting of Mixtures

1. Commercial polynucleotides were dissolved in a small volume of assay buffer to give stock solutions approximately 5 mg or 10 U per ml. The concentrations (in phosphate group) were then determined spectrophotometrically using the data of Table 1. 2. Preparation of duplexes: poly(dA)·poly(dT), poly(A)·poly(dT), poly(dA)·poly(U), poly(A)·poly(U). Working solutions (40 mM bp) of each duplex were prepared in 100 ml volumetric flasks by mixing the required volume (see Note 4) of each polynucleotide single strand solution and then making up to the line with buffer. Solutions were transferred to Sterilin screw-cap jars and annealed by heating at 100°C for 5 min in

Drug Binding to DNA·RNA Hybrid Structures

67

a water bath, followed by slow cooling overnight. All four working solutions were stable in the refrigerator over several months, Note 5. 3. Test compounds (approx 1 mg weighed accurately) were dissolved in water (0.5 ml) and diluted tenfold, if necessary, to give volumes convenient for handling. 4. In a typical assay (Fig 4a), solutions of each duplex (1 ml) were mixed in a polythene vial and the requisite amount of test compound (2–5  ml in water) added to give a final mixture 10 mM in each duplex, 0.25 mM in test compound. This solution was mixed gently and allowed to equilibrate for at least 12 h at 4°C before use. 5. Data were acquired at 260 nm using the “Thermal” module within the Cary WinUV software to provide the melting curve and first differential. In a typical run, equilibration for 30 min at 19°C was followed by a ramp from 20–80°C at 1°C min−1. 3.3. Ribonuclease H Assay

1. Concentrated stock solutions of poly(dA)·poly(rU) and poly(dT)·poly(rA) (100  mM bp) were prepared in reaction buffer. Formation of the duplexes was checked by thermal melting (Tm = 73.0°C, DA260 » 0.06, for poly(dA)·poly(rU) 10 mM bp), the solutions were stored at 0–4°C and were stable over several months at 4°C. 2. Test compounds dissolved in water (approx. 2 mg ml−1) were added to portions of diluted nucleic acid stock solution in glass or polythene vials to achieve the desired experimental concentrations, then equilibrated at 0–4°C for at least 12  h before use. 3. In a typical assay, the nucleic acid (or test compound–nucleic acid) solution was equilibrated at 37°C; 1.2 ml was transferred to a cuvette, 2 ml of enzyme preparation added directly and the reaction mixture mixed by gentle inversion 12 times. The sample was placed in the spectrophotometer block at 37°C and A260 monitored for 1–3  h following an initial 2  min delay. Cuvettes were cleaned thoroughly before reuse Note 6. 4. Data were transferred to software packages for analysis. Typical plots of dA 260 vs. time are shown in Fig.  5a. The initial burst phase was followed by a region of steady reaction with approximately linear gradient. Reaction rate, v, was determined as the gradient of this linear region: typically over a period of 20  min during the first 1.5  h of reaction, although in the presence of the more potent inhibitors, this extended up to 90  min. The linear region was identified graphically and the gradient calculated using a software-fitted a straight line (R>0.996). Concentration and velocity (see Note 7) data were transferred to the KaleidaGraph (42) package for calculation

68

Wheelhouse and Chaires

of KM, Vmax and IC50 using the integral curve-fitting routines of the program (see Fig. 5b, c).

4. Notes 1. The assays described herein were performed using a CARY400Bio spectrophotometer equipped with a multicell (6+6) block and Peltier heating control. A matched pair of Hellma masked quartz cuvettes (114-QS) was used. The test solution cuvette was placed in the front position and buffer in the back (or nothing where relative absorbance data were required); sample volumes were 1.2 ml. The temperature was monitored using a temperature probe placed in a third cuvette, containing buffer, placed in the position adjacent to the sample cuvette in the block. 2. Attempts at dilution of the enzyme stock into reaction buffer prior to use resulted in loss of activity. The enzyme preparation is sufficiently fluid at room temperature to be pipetted into the reaction cuvette. Thorough mixing by gentle inversion 12 times initiates the reaction. 3. Tris buffer was used as made up from the supplied solid; the pH was not checked or adjusted. 4. Precision digital pipettes were used for all test compound and nucleic acid measurements. 5. Before use, each of the four duplexes should be melted individually for quality assurance. The absorbance change observed for each duplex is about the same. 6. It is critical that the cuvettes are cleaned thoroughly: during the KM study, cuvettes were soaked for 3–6 h in 2% Hellmanex solution between reactions. 7. From the total absorbance change at “infinity” (about 6.5 h for poly(dA)·poly(U)), and assuming complete release of rU nucleotides from the hybrid, the absorbance data could be scaled to nmoles of rU released (1 Abs unit ≡ 287.08 nmoles rU) and hence rate data converted to pmol(rU) s−1.

Acknowledgments This work supported by award number R01GM077422 from the National Institute of General Medical Sciences. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIGMS or NIH.

Drug Binding to DNA·RNA Hybrid Structures

References 1. Rich A (1960) A hybrid helix containing both deoxyribose and ribose polynucleotides and its relation to the transfer of information between the nucleic acids. Proc Nat Acad Sci USA 46:1044–53 2. Rich A (2006) Discovery of the hybrid helix and the first DNA-RNA hybridization. J Biol Chem 281:7693–6 3. Williams DA, Lemke TL (2002) Foye’s principles of medicinal chemistry. Lippincott, Williams & Wilkins, Baltimore, MD, USA 4. Ren J, Qu X, Dattagupta N, Chaires JB (2001) Molecular recognition of a RNA:DNA hybrid structure. J Am Chem Soc. 123:6742–3 5. Barbieri CM, Li TK, Guo S et  al (2003) Aminoglycoside complexation with a DNA· RNA hybrid duplex: the thermodynamics of recognition and inhibition of RNA processing enzymes. J Am Chem Soc 125:6469–77 6. Gillis AJ, Schuller AP, Skordalakes E (2008) Structure of the Tribolium castaneum telome­ rase catalytic subunit TERT. Nature 455:633–7 7. Francis R, West C, Friedman SH (2001) Targeting telomerase via its key RNA/DNA heteroduplex. Bioorg Chem 29:107–17 8. Rangarajan S, Friedman SH (2007) Design, synthesis, and evaluation of phenanthridine derivatives targeting the telomerase RNA/ DNA heteroduplex. Bioorg Med Chem Lett 17:2267–73 9. Yu HQ, Zhang DH, Gu XB, Miyoshi D, Sugimoto N (2008) Regulation of telomerase activity by the thermodynamic stability of a DNA x RNA hybrid. Angew Chem Int Ed Engl 47:9034–8 10. Gmeiner WH, Cui W, Konerding DE et  al (1999) Shape-Selective Recognition of a Model Okazaki Fragment by GeometricallyConstrained Bis-Distamycins. J Biomol Struct Dyn 17:507–18 11. Saenger W (1984) Principles of Nucleic Acid Structure. Springer Verlag, New York 12. Noy A, Perez A, Marquez M, Luque FJ, Orozco M (2005) Structure, recognition properties, and flexibility of the DNA.RNA hybrid. J Am Chem Soc 127:4910–20 13. Gyi JI, Conn GL, Lane AN, Brown T (1996) Comparison of the thermodynamic stabilities and solution conformations of DNA.RNA hybrids containing purine-rich and pyrimidine-rich strands with DNA and RNA duplexes. Biochemistry 35:12538–48 14. Gyi JI, Lane AN, Conn GL, Brown T (1998) Solution structures of DNA.RNA hybrids with purine-rich and pyrimidine- rich strands:

15. 16.

17.

18.

19. 20.

21. 22.

23. 24.

25. 26.

27.

28.

69

comparison with the homologous DNA and RNA duplexes. Biochemistry 37:73–80 Jenkins TC (2000) Targeting multi-stranded DNA structures. Curr Med Chem 7:99–115 Alberti P, Ren J, Teulade-Fichou MP et  al (2001) Interaction of an acridine dimer with DNA quadruplex structures. J Biomol Struct Dyn 19:505–13 Shinomiya M, Chu WH, Carlson RG, Weaver RF, Takusagawa F (1995) Structural, Physical, and Biological Characteristics of RNA·DNA Binding-Agent N8-Actinomycin-D. Bio­ chemistry 34:8481–91 Takusagawa F, Takusagawa KT, Carlson RG, Weaver RF (1997) Selectivity of F8-actinomycin D for RNA·DNA Hybrids and its Anti-leukemia Activity. Bioorg Med Chem 5:1197–207 Shaw NN, Arya DP (2008) Recognition of the unique structure of DNA:RNA hybrids. Biochimie 90:1026–39 Shaw NN, Xi H, Arya DP (2008) Molecular Recognition of a DNA:RNA Hybrid: Subnanomolar Binding by a Neomycin-methidium Conjugate. Bioorg Med Chem Lett 18:4142–5 Ren J, Chaires JB (1999) Sequence and structural selectivity of nucleic acid binding ligands. Biochemistry 38:16067–75 Chaires JB (2002) A competition dialysis assay for the study of structure-selective ligand binding to nucleic acids. In: Beaucage SL, Bergstrom DE, Glick GD, Jones RA (eds) Current protocols in nucleic acid chemistry, vol 1. John Wiley & Sons, Inc, New York, pp 8.3.1–8.3.8 Ragazzon P, Chaires JB (2007) Use of competition dialysis in the discovery of G-quadruplex selective ligands. Methods 43:313–23 Ragazzon PA, Garbett NC, Chaires JB (2007) Competition dialysis: a method for the study of structural selective nucleic acid binding. Methods 42:173–82 Ren J, Chaires JB (2001) Rapid screening of structurally selective ligand binding to nucleic acids. Methods Enzymol 340:99–108 Shi X, Chaires JB (2006) Sequence- and structural-selective nucleic acid binding revealed by the melting of mixtures. Nucleic Acids Res 34:e14 Wilson WD, Tanious F, Fernades-Saiz M, Rigl CT (1997) Evaluation of drug-nucleic acid interactions by thermal melting curves. In: Fox KR (ed) Drug-DNA interaction protocols, vol 90. Humana, Totowa, NJ, pp 219–40 Shi X, Chaires JB (2006) Thermal denaturation of drug-DNA complexes: tools and

70

29.

30.

31.

32. 33.

34.

35.

Wheelhouse and Chaires tricks. In: Waring M (ed) Sequence-specific DNA Binding Agents. RSC Publishing, Cambridge, pp 130–51 Sun D, Hurley LH, Von Hoff DD (1998) Telomerase assay using biotinylated-primer extension and magnetic separation of the products. Biotechniques 25:1046–51 Kim NW, Wu F (1997) Advances in quantification and characterization of telomerase activity by the telomeric repeat amplification protocol (TRAP). Nucleic Acids Res 25:2595–7 Francis R, Friedman SH (2003) An interference-free fluorescent assay of telomerase for the high-throughput analysis of inhibitors. Anal Biochem 323:65–73 Craig LC, King TP (1962) Dialysis. Methods Biochem Anal 10:175–99 Muller W, Crothers DM (1975) Interactions of heteroaromatic compounds with nucleic acids. 1. The influence of heteroatoms and polarizability on the base specificity of intercalating ligands. Eur J Biochem 54:267–77 Chaires JB (2005) Competition dialysis: an assay to measure the structural selectivity of drug-nucleic acid interactions. Curr Med Chem Anti-Canc Agents 5:339–52 Chaires JB (2005) Structural selectivity of drug-nucleic acid interactions probed by

competition dialysis. In: Waring MJ, Chaires JB (eds) DNA binders and related subjects, vol 253. Springer-Verlag, Berlin, pp 33–54 36. Crothers DM (1968) Calculation of melting curves for DNA. Biopolymers 6:1391–404 37. McGhee JD (1976) Theoretical calculations of the helix-coil transition of DNA in the presence of large, cooperatively binding ligands. Biopolymers 15:1345–75 38. Raschke TM, Kho J, Marqusee S (1999) Confirmation of the hierarchical folding of RNase H: a protein engineering study. Nature Struct Biol 6:825–30 39. Tipton KF (1992) Principles of enzyme assay and kinetic studies. In: Danson MJ J (ed) Enzyme assays a practical approach. Oxford University Press, Oxford, UK 40. Chaires JB, Dattagupta N, Crothers DM (1982) Studies on interaction of anthracycline antibiotics and deoxyribonucleic acid: equilibrium binding studies on interaction of daunomycin with deoxyribonucleic acid. Biochemistry 21:3933–40 41. Moller A, Nordheim A, Kozlowski SA, Patel DJ, Rich A (1984) Bromination stabilizes poly(dG-dC) in the Z-DNA form under low-salt conditions. Biochemistry 23:54–62 42. KaleidaGraph: Synergy Software, 2004.

Chapter 5 Quantification of Binding Data Using Capillary Electrophoresis Fitsumbirhan Araya, Graham G. Skellern, and Roger D. Waigh Abstract The design of new DNA-targeted molecules, primarily for use in the therapy of diseases such as cancer, relies on the assessment of both affinity for DNA and selectivity of binding to choosen base pair sequences. Capillary electrophoresis, with a polymer added to the running buffer, is very well suited to the separation of oligonucleotides in the range 12–20 base pairs, with the separation based on length rather than base pair sequence. In this way, it is possible to conduct competition experiments using mixtures of up to four oligonucleotides and giving a direct measure of the relative affinity of high-affinity ligands, specifically those binding in the minor groove with slow on-off rates. The relative affinities can be securely quantified, even where the affinities are very high. Working from first principles, it is shown that the measurement of absolute affinities presents various problems, not least that the concentration of DNA and ligand used in the experiment will affect the magnitude of Kd, which is not constant. Key words: Ligand binding, Kd, Minor groove, Drug design, Quantification, Competition, Simulation, Capillary electrophoresis, Distamycin, Netropsin, Berenil, Hoechst 33258

1. Introduction If drugs that bind noncovalently to DNA are to be developed as successful medicines, it is highly desirable that they both bind with high affinity and that they are selective in the DNA sequence that they bind to. It is more convincing in a study of the binding of any two species to say that the dissociation constant has a specific numerical value, rather than stating that the binding is “weak,” “strong,” or “intermediate”. This has led to the application to capillary electrophoresis data of older techniques, such as the Scatchard plot, and more recently, computerized nonlinear curve fitting. Unfortunately, the use of these by nonspecialists often leads to entirely inappropriate conclusions, with implicit K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_5, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

71

72

Araya, Skellern, and Waigh

trust being invested in the numbers produced by the “black box”. The following is an attempt to put this kind of data treatment on a more secure footing, even if the conclusion is often that no precise numbers can be assigned. While the focus will be mainly on the use of capillary electrophoresis for binding studies with DNA and small ligands, many of the arguments apply more generally. The general nature of the binding studies that we have conducted over a number of years has been described in detail (1) and has concentrated on the use of short lengths of DNA, in the range 12–20 base pairs, particularly with compounds binding in the minor groove. The present paper will concentrate on data analyses rather than the practical experimental technique. A simple simulation technique is presented that may convince the researcher to be cautious in assigning numbers to binding equilibria. 1. Basic Theory

In a chemical synthesis, for example, in a production environment, it is possible to write a simple equation for a nonreversible reaction: A+B®C The rate of production of C is governed by the intrinsic reactivity of A and B as well as the concentration, the rate of mixing and the temperature. The intrinsic reactivity is in turn determined by the shapes of the molecules, the charge distribution and in some cases by the nature of the “frontier” orbitals. Together, these make up a reaction ‘frequency factor’ that governs the outcome of collisions between the molecules. If the concentrations of A and B are maintained by addition of more material, the reaction will proceed at a uniform rate, assuming that practical considerations such as temperature and mixing are constant. If the concentration of either A or B is halved, the reaction rate will be halved, since the frequency of collisions is halved. If the concentrations of both A and B are halved, the reaction rate will be reduced fourfold. Quantitative treatment of equilibria originated with the Law of Mass Action, put forward by Guldberg and Waage in 1864 (2). The equilibrium that they treated was of the form A+B → ← C+D In which the equilibrium constant K is given by K = [C] · [D]/[A] · [B] where the square brackets represent concentrations, or more properly activities. It may be noted that K is dimensionless and independent of concentration: the concentration term, such as “molar” is said to “cancel out”.

Quantification of Binding Data Using Capillary Electrophoresis

73

The value of K is determined by the values of two rate constants, “forward” and “reverse”; arbitrarily A+B is forward and C+D is reverse. Individually, the two rates are affected in the same way as a simple synthetic reaction, as above, but in this kind of equilibrium dilution slows the forward and reverse rates equally, so the only determinant of K is the “free energy” associated with each process. When one molecule binds to another to generate one product instead of two in a reversible reaction, in other words “complex formation”, the above does not apply. Where AB → ← A+B, the rate of the reverse (association) reaction is affected by dilution, which alters the probability of collisions between A and B. The rate of the forward (dissociation) reaction is not affected by the probability of collision. If we measure the position of equilibrium and then dilute the solution, we shall obtain a different value for Kd, the dissociation value at equilibrium, which is not a true constant. This concept is illustrated in Fig. 1. Before dilution (Fig. 1a), the equilibrium value, normally called the Kd value, is 10 concentration units (10 × 10/10). This solution is diluted to give b in Fig. 2b and this relaxes to c. Once equilibrium is reached, in c, the original box size now contains 8 concentration units of A and B, but only 2 units of AB. The “Kd” value is now 32 concentration units (8 × 8/2). The difference in Kd values is caused by the less frequent collisions between A and B that occur after dilution, while the decomposition of AB occurs with the same probability as before. 1.2. Molecular Events Affecting Binding

It may be helpful to consider the molecular processes that affect dissociation. At very low temperatures, the complex, once formed, will not break down. As the temperature is raised, buffeting from surrounding molecules, primarily water, will increase in both frequency and energy. These random collisions will occasionally occur with such force and in such combination that they produce major conformational changes, sufficient to separate the two components. In many cases, the components may recombine, since they are initially close together, but sometimes they part company. In the specific context of minor groove binders, which are mainly long thin molecules, capable of wrapping around the double helix, the buffeting from water molecules will often cause partial separation, but part of the molecule may remain in close contact so that there is a high probability that the ligand will return to a fully bound state. Random buffeting will only rarely separate a ligand completely, unless the fit is poor, and the result is a slow off-rate. Since binding requires a substantial re-alignment of the ligand with the helix, the on-rate may also be slow. Some special considerations apply to charged molecules; since opposite charges attract, there is a force-field that draws the two

74

Araya, Skellern, and Waigh

AB

1

B

B

A

AB

A

AB

B

AB

A AB

A

A

AB

B

B A AB B B A

A B

B

A

AB

A

B

dilute

2 AB

A AB

A

A

B

AB

AB

AB

A B

AB

B

A

AB

B

B

B

A AB

B A

A

B

A

A

B

B

equilibrate

3 A B

A

B A AB

A

B

A B

AB

B

A

A B

A

A

B

B

A B A

AB

B A

A

A

A

AB

B

B

B

A

B

Fig. 1. Diagrammatic representation of the effect of dilution on a reversible reaction in which a complex dissociates to give two species.

together and may hold them in proximity as an ion pair. Under certain conditions, this might result in the general considerations about the probability of collision being discounted, since the two species are closer to each other than would be expected from simple considerations of dilution. It is possible to imagine a situation in which dilution affects the rate of formation of ion pairs after mixing, but after equilibration, there is no effect of dilution on complex formation. If this were the case generally, the above comments about Kd not being constant would be invalid. In practice, the formation of simple ion pairs is very unlikely, since there are normally competing charged species. DNA will normally be

Quantification of Binding Data Using Capillary Electrophoresis

A

B

Absorbance at 260nm

2 1

5 3

75

C 9 6

10 11 7

4

8

12

Fig. 2. Electropherogram showing separation of three sets (A, B and C) of double stranded oligonucleotide mixtures: 1= (5¢-CGGCCGCGAAATGCGCCGGC/GCCGGCGCATTTCGCGGCCG-3¢) (20  bp), 2=(5¢-GCCGCGTTAAGCGCCGG/CCGGCGCTTAACGCGGC-3¢) (17bp), 3=(5¢-CCGCGAATTGCGCCG/CGGCGCAATTCGCGG-3¢) (15 bp), 4,7=(5¢-CGCGAATAGCGCC/GGCGCTATTCGCG-3¢) (13  bp), 5=(5¢-GCCGGCGCTATACGCGGCC/GGCCGCGTATAGCGCCGGC-3¢) (19  bp), 6,10=(5¢-GCCGCGATTAGCGCCG/ CGGCGCTAATCGCGGC-3¢) (16  bp), 8=(5¢-GCGTAAAGCGC/GCGCTTTACGC-3¢) (11  bp), 9=(5¢-GGCCGCGAAAAGCGCCGG/ CCGGCGCTTTTCGCGGCC-3¢) (18  bp), 11=CCGCGATATGCGCC/GGCGCATATCGCGG (14bp), 12=(5¢-CGCGATAAGCGC/ GCGCTTATCGCG-3¢) (12 bp), run in 3%HPC dissolved in 0.22 M borate buffer pH 7.5.

accompanied by metal ions such as sodium, magnesium or calcium, and the ligand will normally be introduced as a salt, such as a hydrochloride. Under normal circumstances, there will be a wide variety of ionic attractions and the chance that a ligand molecule will be close to a DNA molecule, but not fully bound, will be affected by dilution. It is conceivable that influences such as water structure (3) could mean that some ligands might spend more time near to DNA helices, without fully binding, than purely random arrangements would indicate, but this is not predictable and does not affect the general conclusion that Kd will be affected by dilution. It may be worth pointing out that the dissociation of protic acids, in other words, acids that liberate a proton when they dissociate, appears at first sight to be an example of the generalised dissociation AB → ← A+B. This is a special case (see Note 1). 1.3. Use of CE in Measuring Relative Binding Affinities

It is possible to carry out competition experiments using CE and these have several advantages (1). Typically, a mixture of oligonucleotides is treated with a ligand in increasing concentration and the resultant increase in complex formation for each oligo is plotted against the DNA/ligand ratio. In the present context, a major advantage is that the experiment is carried out at the same concentration for each oligo, so relative Kd values can be determined: these are true constants, without a concentration term,

76

Araya, Skellern, and Waigh

and should be fairly reproducible between different techniques. It is reasonable to assume that unpredictable effects in CE, such as sample stacking arising from differences in the sample buffer strength from that in the running buffer, will affect all the oligos in a mixture equally. For practical purposes, the relative Kd values are valuable. For example, in a program of drug discovery one of the most important parameters is the selectivity of a potential drug for one DNA sequence over other similar ones. Ultimately, when a potential DNA-binding drug is delivered to the body for treatment of disease, the concentration at the site of action is unlikely to be measurable. Neither the ability of the drug to reach the DNA will be known, nor will the ability to reach vulnerable binding sites on the DNA. The true, functional DNA concentration is also unlikely to be able to be determined. Overall, it seems that attempts to measure absolute binding affinities are of dubious value, particularly for high-affinity ligands, for both theoretical and practical reasons.

2. Materials 1. Minor groove binders such as distamycin, netropsin, berenil and Hoechst 33258, capillary electrophoresis grade disodium tetraborate and boric acid were all purchased from Sigma, dissolved in HPLC grade water and spectrophotometrically quantified using their appropriate extinction coefficient: distamycin (e303 = 35,000), netropsin (e296 21,500), berenil (e370 34,400) and Hoechst 33258 (e338 42,000). All ligands had purity of above 95% except berenil which was above 90%. 2. Hydroxypropylcellulose MW 80,000 was purchased from Sigma. The 3% hydroxypropylcellulose polymer solution was prepared by adding the powder to 0.22 M borate buffer (pH 7.5) and mechanically stirring for 24 h. The polymer solution was filtered using a 5 mm pore size syringe filter and degassed in an ultrasonic bath. 3. Single stranded oligonucleotides were purchased from Bioneer Corporation, Korea. They were dissolved in HPLC grade water and quantified by measuring their optical density at 260 nm using the extinction coefficient provided by the supplier. The oligonucleotide solutions were all stored at −20°C until the time of use. The duplexes were formed by dissolving equal amounts of both the complementary single strands in 0.02  M NaCl, annealed by heating the mixture at 65°C for 10 min followed by slow cooling, and kept at −20°C until the time of use (see Note 2).

Quantification of Binding Data Using Capillary Electrophoresis

77

3. Methods 3.1. Capillary Electrophoresis Separations

Experiments were conducted using a P/ACE MDQ capillary electrophoresis instrument from Beckman Coulter, with a photodiode array detector set at an appropriate wavelength. A plug of the sample mixture was pressure injected (15psi for 7 s) into the anodic end of the capillary (50 mm ID, 375 mm OD and 40 cm effective length, purchased from Polymicro Technologies, Phoenix, AZ, USA) and subjected to 25 kv electric field in the forward polarity. During all the runs, the temperature of the capillary inside the cartridge was thermostatically kept at 20°C. Each new capillary was treated with 1M NaOH at 60°C for 2 h to etch the fused silica, then cleaned by flushing it with water for 1 hour followed by 0.1M NaOH for 10  min, and finally filled with the 3%HPC polymer solution for 10 min. After each run, the capillary was flushed with water for 5 minutes followed by 0.1M NaOH for 3 min, then with water for 2 min, and filled with the polymer solution for 5  min and run for 10  min. The data obtained were analyzed using 32 Karat software. By this procedure, the capillary can be used for hundreds of runs before it requires replacement. Fig. 2 shows sections of electropherograms in which oligos containing (A/T)4 sites are separated from oligos containing similar sites. In principle, any desired base pair sequence can be separated from any other, at least in the range of lengths studied so far, by changing the total lengths of the oligos so that they differ by at least two base pairs. The separation is based on a partial entanglement mechanism, in which the longer oligos undergo more extensive interactions with the added polymer than the shorter ones. All analytes migrate towards the negative electrode, against the natural tendency of the negatively charged DNA: transient interactions with the neutral polymer cause the DNA to move more quickly towards the detector, at the negative end of the capillary, with the result that the longer ones elute first. In each electropherogram, the migration time for the first peak is of the order of 7 min. Figure. 3 shows UVspectra for DNA and equimolar complexes with four minor groove binders, showing the longwavelength peak that allows each complex to be detected in the presence of free DNA. Although the origin of this peak is not clear, it has been observed with every minor groove binder studied to date. Figure. 4 shows sections of a series of electropherograms in which netropsin is added in increasing amounts to a mixture of oligos containing different binding sites. The detection wavelength was set to record the complex rather than the free DNA, in order to avoid confusion between the two species; in many

Araya, Skellern, and Waigh

mAU

78

200

225

250

275

300

325

350

375

400

425

nm

Fig. 3. UV spectra of DNA and complexes with some well-studied minor groove binders: berenil (­ ·­), distamycin (—), Hoechst 33258 (····), netropsin (− ∙∙ −), free ODN (– –) . These are overlaid spectra, small differences in absorbance are probably attributable to experimental error.

5

6

Absorbance (315 nm)

4 3 2

1 Migration Time

Fig. 4. Sections of electropherograms at 315 nm showing increased peak size of the complex formed upon titration of four double stranded oligonucleotides in a mixture (5¢-CGCGATAAGCGC/GCGCTTATCGCG-3¢) (12 bp), (5¢-CCGCGATATGCGCC/ GGCGCATATCGCGG-3¢) (14 bp),(5¢-GCCGCGATTAGCGCCG/CGGCGCTAATCGCGGC-3¢) (16 bp),(5¢-GGCCGCGAAAAGCGCCGG/ CCGGCGCTTTTCGCGGCC-3¢) (18 bp) with increased amounts of netropsin. 1 = netropsin:DNA ratio (R) = 0, 2 = netropsin: DNA ratio (R) = 1, 3 = netropsin:DNA ratio (R) = 2, 4 = netropsin:DNA ratio (R) = 3, 5 = netropsin:DNA ratio (R) = 4, 6 = netropsin:DNA ratio (R) = 5, run with 3% HPC in 0.22 M borate buffer pH 7.5.

Quantification of Binding Data Using Capillary Electrophoresis

79

cases, the free and bound DNA migrate very close together. The peak height or area for each complex may be compared directly with the “MCE” (molarity of complex at equilibrium) in the simulations in Fig. 5, although in the simulation only one complex is considered.

peak height

0.02 0.015 MCE

0.01

FDE

0.005 0

1

2

3

4

5 6 ligand added

7

8

9

10

peak area

0.02 0.015 MCE

0.01

FDE

0.005 0

1

2

3

4

5 6 ligand added

7

8

9

10

0.025

peak area

0.02 0.015

MCE

0.01

FDE

0.005 0

1

2

3

4

5 6 ligand added

7

8

9

10

0.025

peak area

0.02 0.015

MCE

0.01

FDE

0.005 0

1

2

3

4

5 6 ligand added

7

8

9

10

Fig.  5. Simulated DNA-ligand CE titrations with relative Kd values 10−3 (at top), 10−4, 10−5 and 10−6. MCE Molarity of Complex at Equilibrium, FDE Free DNA at Equilibrium.

80

Araya, Skellern, and Waigh

3.2. Data Plotting

It is possible, starting from the simple dissociation equation AB → ← A+B, to write combined equations for competition experiments, conducted as in Fig. 4. Typical plots are shown in Figs. 6 and 7, after programming in Visual Basic (4). The significance of these plots is twofold. First, they show that distinctions can readily be made between Kd values that are reasonably close. Second, they show that competition experiments can provide comparative data even when the affinities are very high: compare this with the difficulties encountered where “absolute” affinities are modeled in Subheading 3.3. Experimental data can be plotted in an analogous manner, from experiments conducted as in Fig. 4, provided that the affinities are reasonably high for all the oligos in the mixture. An example is provided in Fig. 8. It should be noted that the curves have been normalized, to allow for the different peak heights or areas observed with oligos of different lengths.

Fig. 6. Theoretical binding curves of three oligodeoxynucleotides (ODNs) in a mixture titrated with increasing quantities of the ligand. The Kd values of ODN1, ODN2 and ODN3 are 10−5, 10−4, and 10−3 respectively and the curves are plotted using basic theoretical equations.

Fig. 7. Theoretical binding curves of three ODNs in a mixture titrated with increasing quantities of the ligand. The Kd values of the ODN1, ODN2 and ODN3 are 10−8, 10−6, and 10−4 respectively and the values are plotted using basic theoretical equations.

81

Quantification of Binding Data Using Capillary Electrophoresis

Normalized peak area

1

0.8 AAAA AATA

0.6

TATA AAAA

0.4

AATA TATA

0.2

0

0

0.5

1

1.5

2

2.5

3

Netropsin: oligomer ratio

Fig. 8. Curve fit plots of normalized peak area versus netropsin: ODN ratio (R) of complexes of netropsin and (—–) = 5¢-CGACCACGAAAAGCACCAGC/GCTGGTGCTTTTCGTGGTCG (AAAA), (– –) = 5¢-ACCACGTATAGCACCA/TGGTGCTATACGTGGT (TATA), (- - -) = 5¢-CACGAATAGCAC/GTGCTATTCGTG (AATA) with experimental points indicated by (), () and (▲) respectively. The relative Kd values obtained from curve fitting in this experiment were AAAA (1), AATA (3.3) and TATA (11.1). The highest affinity is for the oligo with the central AAAA sequence.

3.3. Modelling With a Spreadsheet

A spreadsheet, such as Microsoft Excel, can be used to explore the equilibrium situation in complex formation such as that described above. For the sake of simplicity, it is assumed that there is one binding site and that one molecule of ligand binds to one molecule of DNA. The spreadsheet can be set up as follows: A

B

C

D

E

F

G

H

I

J

TDM

MCE

FDE

DF

DFF

AFF

DR

AR

EP

Kd

Where TDM is the total DNA molarity, MCE is the molarity of complex at equilibrium, FDE is the free DNA at equilibrium, DF is the dilution factor, DFF is the dissociation frequency factor, AFF is the association frequency factor, DR is the dissociation rate, AR is the association rate, EP is the equilibrium position, and Kd is the equilibrium value. A–J are the cell names used by Excel. DFF and AFF are measures of the intrinsic probability of dissociation and association events: experimentally, they would be expressed as “per mole,” “per unit time,” but in this simulation the units are irrelevant: only the ratio of DFF to AFF matters. Entries are made as follows TDM

MCE

FDE

DF

DFF

AFF

DR

AR

EP

Kd

Num

=TDM-FDE

Num

Num

Num

Num

=MCE*DFF

=FDE*AFF

=1

result

82

Araya, Skellern, and Waigh

Assuming that the active line in Excel is line 2, the cell entries will look like this: A

B

C

D

E

F

G

H

I

J

N1

=A2-C2

N2

N3

N4

N5

=B2*E2

=C2*F2/D2

=G2/H2

=C2*C2/B2

Enter a value for the initial DNA molarity as N1; this is simply the concentration of DNA before adding any ligand. Enter a smaller value as N2, since the free DNA must be less than the total DNA. Enter a value of unity for N3 initially; this may be used later to compare the effects of dilution: it is important to realise that this is a simulation and that the value of the dilution factor has no absolute significance, it is solely a means of comparison between theoretical experiments conducted at different concentrations. Enter an arbitrary value as N3, for example, 1,000 and do the same for N4; it is instructive to change these values later, but initially we can set the dissociation and association frequencies to be the same. Start “Solver” and enter the target value $I$2 as unity (=1, the condition for equilibrium). Enter $C$2 as the value to be altered. The output looks like this, if the initial molarity of the DNA is set to 0.02, a normal working concentration for CE using a UV detector: A

B

C

D

E

F

G

H

I

J

0.02

0.01

0.01

1

1000

1000

10

10

1

0.01

Under these conditions, we find that there are equal concentrations of free and bound DNA, not surprising since we set the dissociation and association frequency factors to be equal. The value of Kd is 10−2 M. Let us now suppose that we have a fluorescent label on the DNA and we can use a suitable detector, which gives us two orders of magnitude greater sensitivity. We dilute the solution accordingly, so that the initial DNA concentration is 0.2mM. Consider the effect of concentration on the frequency of collision of X and Y, in this context, these are free DNA and free ligand. If we reduce the concentration of one of them by an order of magnitude and keep the other constant, the collision rate goes down tenfold. If we reduce both, the frequency is reduced 100-fold. In changing from UV to fluorescence detection, we have reduced the concentration of both DNA and ligand 100-fold and the dilution factor is 10,000. After running ‘Solver’ the simulation looks like this: A

B

C

D

E

F

G

H

I

J

0.0002

2E-08

0.0002

10000

1000

1000

2E-05

2E-05

1

1.9998

Quantification of Binding Data Using Capillary Electrophoresis

83

Thus Kd has a value of 2M and there is very extensive dissociation, much more than in the experiment using UV detection, but bear in mind that this statement applies only to the model and that the values of the dilution factor and the two frequency factors are arbitrary. It is only the relative degree of dissociation that is significant. In cyberspace, the first molarity chosen could have been different and both values for Kd would have been different. It is instructive to repeat the simulation using a wide variety of assumed conditions, as follows. Since the affinity of certain ligands for appropriate DNA sequences is known to be very high, we can set the association frequency factor to be very high, perhaps 1,000,000, and repeat the “UV detected” experiment: A

B

C

D

E

F

G

H

I

J

0.02

0.01998

2E-05

1

1000

1.0E6

19.98

19.98

0.999999

1.998E-08

It may be assumed that figures such as “1.998” are the result of rounding errors in binary arithmetic and the predicted value of Kd is 2 × 10−8, which seems intuitively reasonable. If the simulation is repeated, using fluorescence detection, with the same assumptions as before, the output looks like this: A

B

C

D

E

F

G

H

I

J

0.0002

1.82E-05

1.82E-4

1.0E4

1000

1.0E6

0.0181

0.0181

0.9999

1.81E-3

The values found in the four simulations are as follows: UV detection, DNA 0.02M, equal frequency factors, Kd = 10−2. Fluorescence detection, DNA 0.0002M, equal frequency factors, dilution factor ratio 104, Kd = 2.0. UV detection, DNA 0.02M, frequency factor ratio 103, Kd = 2 × 10−8. Fluorescence detection, DNA 0.0002M, dilution factor ratio 104, frequency factor ratio 103, Kd = 1.8 × 10−3. As a practical comment, it may be found that “Solver” cannot cope with the numbers produced when assumed values of Kd are numerically low, indicating high affinity. This is attributable to the way that “Solver” works, and it will be found that convergence to the correct values can be achieved manually, albeit very inconveniently. These arguments have assumed that the ligand and DNA bind in a 1:1 ratio. If two molecules bind side-by-side in the minor groove, which is more common than single occupancy, the calculation of the equilibrium value is more complicated. At its simplest, the effect of dilution where three molecular species come together will increase further the effect of dilution on Kd.

84

Araya, Skellern, and Waigh

In practice, many ligands that bind in pairs in the minor groove also associate in aqueous solution (5, 6). The stacking effect in free solution probably does not stop at two molecules in most cases, but this will be altered by dilution. If aggregation stops at two molecules and if the two molecules have high affinity for each other, the binding process can probably be treated numerically as if a single molecule were binding at half the added ligand concentration. In most cases, the situation will be more complicated and the correct way to analyze the data will be unknown. A further complication is that ligands such as distamycin can bind in either a 2:1 or a 1:1 ratio, depending on the specific base pair sequence in the binding site, which alters the width of the minor groove. Attempts to determine Kd are then full of uncertainty, but it is still possible to carry out competition experiments. (1) The main requirement is often to determine relative affinities and the question of stoichiometry is secondary: whether a ligand prefers to bind to site A as single molecules rather than site B as pairs, is often irrelevant from a practical point of view. Comparison of binding data between different techniques is fraught with difficulty (see Note 3). 3.4. Simulation of Titrations

In the practical context, a titration is conducted by measuring an appropriate amount of DNA; with UV detection this will be around 20mM in the final solution injected on the column. To the DNA solution are added aliquots of ligand solution, up to or slightly more than the molarity of the DNA, and ideally both the free DNA and the complex are measured as they pass the detector. The ligand-DNA complex normally has a long-wavelength absorption maximum for detection not possessed by the free DNA, otherwise the method relies on separation of the free and complexed DNA. (1) A spreadsheet can be set up to model the effect of addition of increasing amounts of ligand in a very similar manner to that used above. The graphic outputs for simulations with Kd values in the range 10−3–10−6 are presented in Fig. 5. The histograms are a direct representation of experimental data in which the height of a column represents the height (or area) of a peak. It should be clear that each histogram column has two DNA peaks, the first being free DNA and the second the complex. An example of an Excel table for one of these simulations is given as Table  1. As before, the target, whether using “Solver” or converging manually, is that EP = 1, so that the system is at equilibrium. The obvious conclusion from the simulations in Fig.  5 is that, using capillary electrophoresis, it would be difficult to distinguish with confidence between binding data corresponding to a Kd value of 10−4 from those obtained with Kd equal to 10−5 and

lig added

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

0.02

TDM

0.02

0.02

0.02

0.02

0.02

0.02

0.02

0.02

0.02

0.02

0.019558

0.017914

0.01596

0.013977

0.011985

0.00999

0.007993

0.005996

0.003998

0.001999

MCE

0.000442

8.59E-05

3.95E-05

2.32E-05

1.5E-05

9.98E-06

6.66E-06

4.28E-06

2.5E-06

1.11E-06

free lig eq

0.000442

0.002086

0.00404

0.006023

0.008015

0.01001

0.012007

0.014004

0.016002

0.018001

FDE

1

1

1

1

1

1

1

1

1

1

DFF

10E5

10E5

10E5

10E5

10E5

10E5

10E5

10E5

10E5

10E5

AFF

0.019558

0.017914

0.01596

0.013977

0.011985

0.00999

0.007993

0.005996

0.003998

0.001999

DR

0.019558

0.017914

0.01596

0.013977

0.011985

0.00999

0.007993

0.005996

0.003997

0.001999

AR

1.0

1.0

1.0

1.0

1.0

1.0

1.0

1.0

1.0

1.0

EP

1E-05

1E-05

1E-05

1E-05

1E-05

1E-05

1E-05

1E-05

1E-05

1E-05

Kd

Table 1 An Excel table for a simulated ligand-DNA titration giving a Kd value of 10−5. TDM Total DNA Molarity, lig added molarity of added ligand, free lig eq free ligand at equilibrium, FDE Free DNA at equilibrium, DFF Dissociation Frequency Factor, AFF Association Frequency Factor, DR Dissociation Rate, AR Association Rate, EP Equilibrium Position (= 1, the target in the iteration). Note that the values for the equilibrium position (EP) were obtained manually to five or six decimal places

Quantification of Binding Data Using Capillary Electrophoresis 85

86

Araya, Skellern, and Waigh

even harder to distinguish data with Kd equal to 10−5 from those with Kd equal to 10−6. It should be remembered that there is always experimental error and that this will normally be of the order of +/− 10% in peak area. Curve-fitting software would undoubtedly provide a numerical value in each case, when in truth the experiment cannot reliably provide an order-of-magnitude result. The higher the affinity, the more similar the curves appear. Molecules with potential therapeutic value will need to have very high affinity to compete for DNA against the proteins that require to be displaced from their biochemical function and it is clear that a simple titration of the kind simulated in Fig. 5 will be of little value. However, competition experiments do allow comparisons to be made for a ligand binding to different DNA sequences with very high affinity.

4. Notes 1. The dissociation of protic acids is normally characterised by the magnitude of pKa, which equates to the pH value at which 50% dissociation occurs. At first sight, this appears to be an example of the equilibrium AB → ← A+B and the pKa value should change with concentration, but in practice the solvent, which is almost always water, is able to supply the proton (the “B” part of the equilibrium). The assertion that the probability of the reverse reaction depends on concentration is no longer true, since “B” is not confined to the ions liberated by dissociation. In this special case, the acid dissociation constant is dependent only on the relative stabilities of AB and A, and is a true constant, provided that the solvent can act as an alternative source of protons. 2. The preferred method utilises a mixture of commercially available oligonucleotides, generally of 12–20 base pairs in length. This range of lengths was chosen to provide a high probability that the complementary single strands would anneal to form stable duplexes at the normal operating temperature of the equipment, around 20°C, while keeping the cost of synthesis down. In borate buffer, which was found empirically to give the best results, the use of an added polymer, either hydroxyethyl- or hydroxypropyl-cellulose, allows oligos to be reliably separated if they differ by at least two base pairs in length. It is occasionally necessary to use oligos that differ by three base pairs in length but rarely more. The method provides a sensitive means of distinguishing relative affinities, by allowing the oligos to compete for the ligand, but attempting to measure

Quantification of Binding Data Using Capillary Electrophoresis

87

absolute values is not normally worthwhile, for the reasons given above and below. 3. Taking the interaction of small molecules with DNA as representative of the general field, a number of techniques can be used to measure binding. In qualitative terms, they all appear to work well and include gel footprinting, chips on which DNA is immobilised, microcalorimetry, mass spectrometry and capillary electrophoresis (CE) as well as the older technique of UV spectroscopy. Given the necessity to include the concentrations of DNA and ligand, if we are to make a comparison of numerical values of Kd, there is a problem in that the concentrations used may well be different in each technique. Even in the more limited world of CE, the use of a fluorescence detector instead of UV may mean a reduction in working concentration of up to three orders of magnitude. Assuming that we know the initial concentration of DNA in a sample for CE, we then have the question of the true concentration where it matters, on the column. In some cases, the effect of the applied voltage and the flow through the column on concentration within the eluting band may be negligible, but there are also well documented instances of the analyte molecules bunching together, a phenomenon known as focusing, reducing the detected peak width to less than is theoretically possible for the volume loaded on the column (7–9). This effect is reasonably well understood and may arise from either pH differences or ionic strength differences between the sample loaded and the running buffer. More commonly, there may be some diffusion inside the column, which increases peak width. Either way, we cannot be totally certain of the true concentration. To add further confusion, it is certain that a positively charged ligand will migrate at a different rate from the DNA or the complex. This will happen progressively as the mixture being analysed travels down the capillary for any ligand that is not tightly bound to the DNA. For this reason, the state of the complex may be different when it passes the detector from the situation in the solution loaded onto the column. It has been pointed out earlier (1) that CE is not able to distinguish very small differences in peak height or area; the experimental error is considerable. To obtain an absolute measure of the equilibrium binding value, which as stated varies with concentration, we need to be making measurements in the concentration range where there is substantial dissociation. In theory, where affinity is high, this can be achieved by using a more sensitive detector and more dilute solutions, as explored above. In practice, the vast majority of DNA ligands are positively charged and if the experiment is conducted in such a way that there is measurable dissociation, the free ligand will stick

88

Araya, Skellern, and Waigh

to the capillary wall, which is negatively charged. The effects of this are unpredictable, but usually involve peak broadening and loss of resolution. It is theoretically possible to coat the inside of the capillary so that the negative charges on the silanol groups are masked, to avoid the ligand adhering to the wall, but in practice, the loss of the electroosmotic effect caused by such masking results in the loss of column efficiency (10, 11). References 1. Araya F, Huchet G, McGroarty I, Skellern GG, Waigh RD (2007) Capillary electrophoresis for studying drug-DNA interactions. Methods 42:141–149 2. h t t p : / / e n . w i k i p e d i a . o r g / w i k i / L a w _ of_mass_action 3. Plumridge TH, Waigh RD (2002) Water structure theory and some implications for drug design. J Pharm Pharmacol 54:1155–1179 4. Huchet, G., unpublished 5. Parkinson, J. A., unpublished data 6. Buurma NJ, Haq I (2008) Calorimetric and spectroscopic studies of Hoechst 33258: selfassociation and binding to non-cognate DNA. J Biol Chem 381:607–621 7. Burgi DS, Chien R-L (1996) Application and limits of sample stacking in capillary electrophoresis in Capillary Electrophoresis Guidebook:

8. 9.

10.

11.

Principles, Operation and Applications, Altria K.D. (ed). Methods Mol Biol 52:211–226 Shihabi ZK (2000) Stacking in capillary zone electrophoresis. J Chromatog A 902: 107–117 Britz-McKibbin P, Bebault GM, Chen DDY (2000) Velocity-difference induced focusing of nucleotides in capillary electrophoresis with a dynamic pH junction. Anal Chem 72:1729–1753 Barron AE, Sunada WM, Blanch HW (1995) The use of coated and uncoated capillaries for the electrophoretic separation of DNA in dilute polymer solutions. Electrophoresis 16:64–74 Horvath J, Dolnik V (2001) Polymer wall coatings for capillary electrophoresis. Electrophoresis 22:644–655

Chapter 6 Determination of Equilibrium Association Constants of Ligand–DNA Complexes by Electrospray Mass Spectrometry Valérie Gabelica Abstract Electrospray mass spectrometry can be used to detect ligand–DNA noncovalent complexes formed in solution. This chapter describes how to determine equilibrium association constants of the complexes. Particular attention is devoted to describing how to tune an electrospray mass spectrometer using a 12-mer oligodeoxynucleotides duplex in order to perform these experiments. This protocol can then be applied to any nucleic acid structure that can be ionized with electrospray mass spectrometry. Key words:  Mass Spectrometry, Electrospray, Nucleic acids, Oligonucleotides, Duplex, G-Quadruplex, Ligand, Equilibrium constants

1. Introduction The introduction of electrospray sources was a major breakthrough in the field of biomolecule mass spectrometry (1, 2). Electrospray allows producing intact ions directly from large biomolecules, which have been directly taken from a solution injected in to a capillary (3). The softness of electrospray (the absence of fragmentation, thanks to the limited energy given by thermal heating and collisional activation) is such that even noncovalent complexes can be detected intact. This is also the case for DNA and RNA noncovalent complexes (higher-order structures, nucleic acid complexes with small molecules or with proteins), as reviewed in references (4–8). In order to obtain soft source conditions that are suitable to maintain complexes intact, it is necessary to change some instrumental parameters compared to those typically used for covalent, less fragile analytes. Instead of giving the parameters that were optimized for our instruments, we will describe a tuning procedure K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_6, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

89

90

Gabelica

that can be applied to any electrospray mass spectrometer (Subheading 3.2). Peaks in the mass spectra are characterized by their m/z and their area. The masses of the peaks give immediate information on the stoichiometry of all complexes formed simultaneously in the solution, while in spectrophotometry methods, the stoichiometry is deduced by fitting of ligand binding curves. In addition, the relative peak areas in the mass spectra are proportional to the relative concentrations of the corresponding species in solution:

[X ] = R × A(X ) A(Y ) [Y ]

(1)

The easiest approximation is to consider that R = 1, i.e., the ratio between peak areas in the mass spectra is equal to the ratio between the concentrations of the corresponding species in solution. This is valid only if species X and Y under consideration have similar masses, shapes, hydrophobicity, and the ions produced have the same total charge. For nucleic acid–ligand complexes, it means that R = 1 is valid for small molecules binding to a nucleic acid target without changing dramatically the target’s conformation. The ideal case is that of minor groove binders of DNA duplexes (9). As there is currently no easy and widely applicable method to determine the value of R, all equations described here are based on the hypo­ thesis that R = 1. However, the protocol and equations can be applied to any nucleic acid structure that can be ionized with electrospray mass spectrometry, even if the validity of the R = 1 hypothesis is not known. In general, deviation from that ideal behavior manifests itself by the fact that the ratio between peak areas can depend on the charge state of the ions in the mass spectra. Therefore, in order to estimate the error due to the hypothesis, we describe a procedure involving the determination of the association constants for the different charge states.

2. Materials 2.1. Sample Preparation

1. 40 nmol dCGCGAATTCGCG, + 40 nmol of each synthetic oligonucleotide you want to include in your assay. We typically use OliGold® oligonucleotides from Eurogentec (www.eurogentec.com), with standard purification “process OliGold”. (see Note 1). Dissolve all lyophilized sequences to 400 µM in bidistilled water, make 100 µL aliquots and store at −20°C. 2. Ligand to be tested, as a powder or as a stock solution of concentration ³50 mM (see Note 2).

Determination of Equilibrium Association Constants of Ligand–DNA

91

3. Water, Na+ and K+ <1 mg/kg (see Note 3). We typically use deionized water that is further distilled twice in-house. 4. Methanol, Na+, and K+ <1 mg/kg (see Note 3), e.g., puriss. p.a., absolute, ACS reagent, ³99.8% (Fluka). 5. Ammonium acetate (NH4OAc) solution Bioultra, for molecular biology, 5M in H2O (Fluka) (see Note 3). 6. Microcon® YM-3 (cut-off = 3  kDa, yellow) Centrifugal Filter Units (Millipore). 7. Store all solvents and sample solutions in polypropylene tubes (see Note 4). 2.2. Electrospray Mass Spectrometer

1. In Subheading 3.2, we will describe a tuning procedure that can be applied to different mass spectrometers equipped with an electrospray source (see Note 5). In addition, we will provide indicative source tuning parameters for four instruments, from three different manufacturers. a. Finnigan LCQ classic (Thermo, San Jose, CA, USA) with standard ESI source block. b. Micromass Q-TOF II (now Waters, Manchester, UK) with standard Z-spray source. c. Waters Q-TOF Ultima Global (Waters, Manchester, UK) with dual MALDI/ESI Z-spray source. d. Bruker Apex-Qe 9.4T FTICR-MS (Bruker Daltonics, Bremen, Germany) with the Bruker Apollo 1 source. 2. A clean 250 µL syringe and a syringe pump capable of delivering flow rates of 1–10 µL/min.

3. Methods The methods will be described in the case of ligand-binding to double-stranded DNA. The general procedure is also applicable to other oligonucleotide structures like intramolecular folds, triplex DNA, or G-quadruplexes. For G-quadruplexes, the criteria used for tuning the instrumental parameters are different from the duplexes because G-quadruplexes embed a certain amount of specific cations. Specific variations of the protocol for G-quadruplexes are described in the notes section. 3.1. Sample Preparation

1. If not known precisely, determine the strand concentration by UV absorption spectrophotometry at 260 nm, using extinction coefficients calculated from http://biophysics.idtdna.com/. 2. Prepare 400 µL of 50 µM duplex [dCGCGAATTCGCG]2 in 150  mM NH4OAc by mixing 100 µL stock solution

92

Gabelica

(400 µM in single strand) with 300 µL NH4OAc 200  mM. Heat the solution in simmering water bath for 5  min, allow cooling to room temperature. Store at 4°C. 3. Prepare stock solutions for each synthetic oligonucleotide you want to include in your assay, final concentration is 50 mM in the desired structure (e.g., 50 µM [TG4T]4 requires 200 µM single strand) (see Note 6). 4. To prepare oligonucleotide solutions ready to inject in the mass spectrometer: mix (in that order) 5 µL oligonucleotide 50 µM with 35 µL aqueous ammonium acetate of the desired concentration, then add 10 µL methanol (see Note 7). Final concentration is 5 µM oligonucleotide and 20% methanol. 5. Oligonucleotide-ligand mixtures ready to inject in the mass spectrometer are 5 µM oligonucleotide, the desired concentration of ligand (typically 5 or 10 µM), and 20% methanol. If the ligand stock was in methanol, take that amount into account in the total 20%. 6. If mass spectra reveal it necessary to remove sodium or potassium from an oligonucleotide sample: (a) If not known, determine the extinction coefficient of the folded oligonucleotide structure at room temperature (this extinction coefficient is not the same as for unfolded single strand). For duplex DNA, use http://biophysics. idtdna.com/. (b) Place X µL of the solution to be desalted on a Microcon® YM-3 filter, centrifuge to near dryness. (c) Add 400 µL of 150 mM NH4OAc in the filter, and centrifuge to near dryness (repeat three times). (d) Add X µL of NH4OAc of the desired concentration to redilute the oligonucleotide and homogenize. (e) Place the filter upside down in a fresh Microcon vial, and centrifuge to get the oligonucleotide solution in the vial. (f) Oligonucleotide concentration has changed during that process. Determine the final concentration with UV spectrophotometry by using the extinction coefficient determined in (a). 3.2. Instrument Tuning

1. Switch the mass spectrometer to negative ion mode (see Note 8). 2. Calibrate the instrument in negative ion mode according to the manufacturer’s instructions. 3. Beware that optimal tuning parameters can vary from instrument to instrument, even from the same manufacturer. However, tuning parameters for a given machine remain constant over time if the instrument is kept clean, and can be

Determination of Equilibrium Association Constants of Ligand–DNA

93

reused later. If you already have tuning parameters for DNA complexes using your instrument, go to Subheading 3.3. 4. If it is the first time you tune your mass spectrometer for DNA complexes, start with the following parameters: (a) Finnigan LCQ classic (Thermo, San Jose, CA, USA) with standard ESI source block. Sheath gas flow rate (arb.) = 60, Aux gas flow rate (arb.) = 5, I Spray voltage (kV) = 3.7, Capillary temperature (°C) = 180, Capillary voltage = −10 V, Tube lens offset = 35. Multipole 1 offset = 4, Lens voltage (V) = 30, Multipole 2 offset (V) = 8, Multipole RF amplitude (Vp–p) = 600. (b) Micromass Q-TOF II (now Waters, Manchester, UK) with Z-spray source. Spray voltage = 2.2 kV, Cone voltage = 20 V, Source block temperature = 70°C, Desolvation temperature = 80°C, collision energy = 10 V, ion energy = 1.8. (c) Waters Q-TOF Ultima Global (Waters, Manchester, UK) with dual MALDI/ESI Z-spray source. First set the source pressure at 3.8  mbar (as measured on the pirani gauge). Spray voltage = 2.2  kV, Cone voltage = 100  V, RF Lens 1 voltage = 60  V, Source temperature = 70°C, Desolvation temperature = 80°C, collision energy = 10 V, ion energy = 1.8. (d) Bruker Apex-Qe 9.4T FTICR-MS (Bruker Daltonics, Bremen, Germany) with Apollo 1 source. Capillary = 3,400 V, End plate = 2,900  V, CapExit = −45  V, Skimmer1 = −30  V, Skimmer2 = −4 V, Offset = −2.5 V, Dry Gas Temp = 150°C, Col Cell Trap = 5 V. (e) For instruments not described here, start from any negative ion parameters, and be prepared to lower all voltages and temperatures significantly. 5. Inject a sample of duplex [dCGCGAATTCGCG]2 using a syringe pump at 4 µL/min. To find which of your instrumental parameters are critical for studying the complexes, change one parameter at a time and observe the ratio between the peak corresponding to the duplex5− (monoisotopic mass = 1456.8516, average mass = 1457.5421) and the peak corresponding to the single strand3− (monoisotopic mass = 1,213.8751, average mass = 1214.4587) (see Note 9). On the four instruments described here, the most important parameters are: (a) Finnigan LCQ classic (Thermo, San Jose, CA, USA) with standard ESI source block: Capillary temperature, Capillary voltage, and Tube lens offset. (b) Micromass Q-TOF II (now Waters, Manchester, UK) with Z-spray source: Cone voltage. Collision energy of 10  V is recommended by the manufacturer, but for very fragile complexes, you may have to lower this voltage down to 2 V.

94

Gabelica

(c) Waters Q-TOF Ultima Global (Waters, Manchester, UK) with dual MALDI/ESI Z-spray source: source pressure and RF Lens 1. Collision energy: see Q-TOF II. (d) Bruker Apex-Qe 9.4T FTICR-MS (Bruker Daltonics, Bremen, Germany) with Apollo 1 source: CapExit and Skimmer1. 6. In doing so, you have probably observed that softer conditions are obtained at the expense of sensitivity. To decide the final values of parameters, observe the relative intensity of the duplex5− peak and its first ammonium adduct (monoisotopic mass = 1460.2568, average mass = 1460.9481) (see Note 10). Choose instrumental parameters that give an intensity of first ammonium adduct peak between 20 and 30% of the intensity of the free duplex5−. Store these parameters. 3.3. Recording Electrospray Mass Spectra for the DNA-Ligand Complexes

1. Using the instrumental parameters determined previously, record a mass spectrum for a sample containing 5 µM of DNA, without ligand, injected at 4 µL/min. Try to assign all major peaks. You should observe your intact DNA at different charge states. You may also observe the separate constituents (single strands). The different m/z are given by m/z = (mDNA–z)/z, where mDNA is the mass of the neutral DNA in Daltons (see Note 11), and z is the charge state. 2. Fine tune the instrument for your particular oligonucleotide. For each intense charge state (signal-to-noise ratio >20), find source parameters at which the intensity of the first ammonium adduct peak (found at m/z = (mDNA + 17–z)/z) is about 20% of the adduct-free peak (see Note 12). Note down these parameters. If your oligonucleotide is a G-quadruplex, see Note 13. For troubleshooting, see Note 14. 3. Sodium adducts are found at m/z = (m DNA + n×22–z)/z. If the first sodium adduct peak is larger than 30% of the noadduct peak, desalt the oligonucleotide sample as described in step 6 of Subheading 3.1, then repeat steps 1 and 2. (see Note 15) 4. Inject a sample containing 5 µM DNA and 5 µM ligand, at a flow rate £ 4 µL/min (see Note 16). If the ligand binds to the nucleic acid, for each charge state, you see additional peaks at m/z = (mDNA + n×mligand–z)/z, where n is the number of ligands bound, and mligand is the mass of the “neutral” ligand (see Note 17). 5. Record separate mass spectra at different voltages: one optimal voltage for each abundant charge state. 6. Wash the electrospray source capillaries with 1 mL bidistilled water to remove ligand and DNA.

Determination of Equilibrium Association Constants of Ligand–DNA

95

7. Inject a sample containing 5 µM DNA and 10 µM ligand (same DNA and ligand as in step 4). Record separate mass spectra at different voltages: one optimal voltage for each abundant charge state. 8. Wash the electrospray source capillaries with 1 mL bi-distilled water to remove ligand and DNA. 9. Before running another DNA-ligand mixture, test with DNA alone that complexes are not detectable any longer (see Note 18). If you have a particular nucleic acid to which the ligand binds strongly, use that nucleic acid for this test. 10. Wash the electrospray source capillaries with 1 mL bi-distilled water to remove DNA. 11. To test another ligand with the same oligonucleotide, go back to step 4. To test another oligonucleotide, go back to step 1. 3.4. Calculation of Equilibrium Association Constant

1. For all mass spectra recorded for DNA-ligand mixtures (total concentration of DNA = CDNA = 5 µM, total concentration of ligand = CL), annotate all peaks corresponding to [1 DNA + n Ligands]z− (n = 0 to N). Determine the maximum number N of ligands bound. 2. Subtract the background from the mass spectra. Ensure that the peak shape is conserved. 3. For each charge state z, use the mass spectrum recorded at the optimum voltage for that charge state. Integrate (see Notes 19, 20) the peaks of each (1:n)z− species (n = 0 to N), and write the peak areas A(1:n) in a worksheet. 4. In the worksheet, for each charge state separately, calculate the concentration of free DNA (1:0) and each complex (1:n) as follows:



[1 : n ] = C DNA ×

A(1:n ) N

∑A

n =0

(1:n )



(2)

5. For each charge state separately, calculate the concentration of free ligand using the mass balance equation (see Note 21): N



[L ] = C L − ∑ n × [1 : n ]

(3)

n =1

6. Now you have the concentrations of all free species in solution, you can calculate, for each charge state and for each mixture, the stepwise equilibrium association (Ka,n) and dissociation (Kd,n) constants:

96

Gabelica



K a,n =

1 [1 : n ] = K d,n [L ][1 : (n − 1)]

(4)

7. For a given DNA target and a given ligand, examine the constants determined for each ligand concentration (at least 2). If the constants determined at different ligand concentrations do not agree with each other, repeat the mass spectrometry experiment, including sample preparation. If the discrepancy persists, see Note 21. 8. Constants determined from different charge states may differ. This is an intrinsic limitation of the method, due to the fact that relative intensities do not exactly reflect relative concentrations in solution. Reject outliers only if based on valid statistical tests. Calculate the mean value of the constant and the standard deviation. This standard deviation gives the uncertainty on the absolute value of the constant due to the electrospray mass spectrometry method.

4. Notes 1. The Oligold® purification process is a Reverse-Phase (RP) purification. The RP-cartridge devices used by Eurogentec are stable up to pH = 13, so the ammonium hydroxide solution after synthesis, diluted in water, is loaded directly on the packing. It is the level of desalting attained by this purification process that makes Oligold® oligonucleotides particularly suitable for mass spectrometry experiments. 2. What solvent can I use if my ligand is not water-soluble? Methanol. If not soluble in methanol either, prepare a concentrated solution in DMSO and then dilute with methanol. The final DMSO concentration in the solution injected in the electrospray source must be £0.5 %. 3. Sodium and potassium must be avoided, because during electrospray, the solvent droplets evaporate, and positive ions around the negatively charged DNA condense around the DNA. This leads to a distribution of species [DNA + n Na/K] that is heterogeneous in mass, and this translates in the mass spectra into a distribution of peaks. This alters the sensitivity of the method (total intensity is distributed over several peaks) and even the mass accuracy (if the peak without Na/K is not identified). A substitute salt is nevertheless needed in order to obtain the desired ionic strength. Ammonium acetate is the salt of choice because both ammonium and acetate are “volatile,” in the sense that an ammonium ion condensing

Determination of Equilibrium Association Constants of Ligand–DNA

97

around a phosphate will give a proton to the phosphate and leave as ammonia (NH3). Similarly, acetate anions can abstract protons and leave as neutral acetic acid. In electrospray mass spectro­metry, NH4OAc can be used at concentrations up to 150  mM, corresponding to physiological ionic strength. Finally, ammonium cations are also capable of stabilizing G-quadruplexes by insertion in between G-tetrads. 4. Glassware can be contaminated with NaCl and KCl that is not removed by autoclaves. Also, beware that some low quality polystyrene tubes may contain traces of polyethylene glycol that is solubilized by methanol, and gives undesirable peaks separated by 44 Da in the mass spectra. 5. Can nanospray be used instead of electrospray? Nanospray capillaries are thin glass capillaries terminated by a very thin needle. It is possible to spray DNA–ligand complexes using nanospray, but three difficulties can be encountered. (1) Glass capillaries can be contaminated by Na/K salts. If so, use quartz capillaries. (2) Thin capillaries clog rapidly when using 150 mM NH4OAc and results are not always reproducible quantitatively. With nanospray, it is recommended to use capillaries with reproducible tip opening (preopened tip) and to repeat the measurement at least three times to test the reproducibility of relative intensities. 6. Assays can be performed in any ammonium acetate concentration from 0 to 150 mM. Important: NH4OAc has no buffering capacities. It is an ampholyte (acid of one couple and base of another couple in equal quantities), hence the pH is the mean of the pKa’s (7.0). Use acetic acid to acidify your solution, or ammonium hydroxide basify your solution, if so desired. 7. By adding some methanol to the sample, the surface tension of the mixture is lower than in pure water. This makes droplet evaporation more favorable, resulting in a higher electrospray signal. You can test this on your own instrument by comparing the same sample with 0%, 10%, and 20% methanol. If the instrument is sensitive enough to work in 0% or 10% methanol, you can of course do so. However, although they were performed in 20% methanol, association constants determined by electrospray mass spectrometry were found in good agreement with those determined by other methods. It is reported that up to 50% methanol does not perturb DNA conformation (10), and that aliphatic alcohols even stabilize G-quadruplex DNA (11). In any case, always add methanol just before injection into the mass spectrometer (do not store DNA solutions in methanol). 8. Although it is possible to obtain mass spectra in positive ion mode, the relative intensities do not reflect the relative abundances in solution (12). Equilibrium association constants can be determined only in the negative ion mode.

98

Gabelica

9. In general, decreasing acceleration voltages and decreasing temperatures lead to softer conditions, i.e., less destruction of the complexes, and higher duplex/single strand ratio. Important: some parameters expressed in volts are not directly related to the voltages applied in the instrument (for example, in Thermo/Finnigan instruments, the tube lens offset voltage is an offset applied to a ramp of voltages that itself depends on the mass range chosen; the more positive the tube lens offset, the softer the conditions). Finally, increasing the pressures inside the instrument can have different effects depending on instrument design (for example, increasing the source pressure in Waters instruments leads to softer conditions, but increasing the pressure inside the collision cell leads to harder conditions). If you don’t know the instrument design in detail, tune all these parameters empirically. 10. Do not confound ammonium adduct (mass difference of +17/z) and sodium adduct (mass difference of +22/z) or potassium adduct (mass difference of +38/z). Ammonium adducts are removed (see Note 3) when source conditions get harsher, while sodium and potassium adducts are not. 11. The mass of the neutral DNA is calculated by assuming that all phosphates are neutralized by protons. 12. At this stage, change only one source voltage that allows you to go from very soft conditions (low voltage) to hard conditions (high voltage). On the instruments described here, these are: (a) on the LCQ, the capillary voltage; (b) on standard Z-spray sources, the cone voltage; (c) on the dual Z-spray source of the Q-TOF Ultima Global, the RF Lens 1 voltage; (d) on the Apollo Source the CapExit voltage. This optimum voltage can be the same for all charge states, but most often you will find the optimum voltage is lower for higher charge states, and increases as the charge state increases. If you see no ammonium adducts even at the lowest voltages allowed on your instrument, lower the source temperatures. 13. Tuning the instrument in the case of G-quadruplexes is different, because G-quadruplexes prepared in NH4OAc incor­ porate a specific number of ammonium cations in between their G-tetrads. In general, a quadruplex containing t consecutive G-tetrads incorporates (t-1) ammoniums, and the theoretical m/z of the intact quadruplex is given by: m/z = (mDNA + (t-1)×17–z)/z. In practice, a distribution of ammonium adducts is observed (from 0, if all inner ammoniums are removed, to > t-1 if nonspecific adducts on the phosphates are formed in addition to the specific adducts between the tetrads). For each charge state, tune the mass spectrometer to have the highest relative intensity for the (t-1) adduct. Note that this can be easy for some G-quadruplexes and very difficult for others (13).

Determination of Equilibrium Association Constants of Ligand–DNA

99

14. Troubleshoot if you have peaks you cannot assign. Check if these can correspond to a shorter sequence (for example one base is missing); in that case, have the DNA re-synthesized. Check if these can be due to backbone fragmentation of your oligonucleotide in the mass spectrometer (you can compute the most common fragments for any sequence with the Mongo Oligo Mass Calculator v2.06 freeware: http://library.med.utah.edu/masspec/mongo.htm); in that case, use softer source conditions. Check if these can correspond to unanticipated higher-order structures (several nucleic acids associating). Finally, check if this could correspond to ammonium, sodium, potassium adducts, adducts with your calibrant, etc. 15. If sodium adducts persist even when using clean solvents and desalting the best you can, you will have to take the intensity of all sodium adduct peaks into account in the calculation of the equilibrium association constants. It is more tedious and less precise than with a well-desalted sample, but the determination of the constants is still feasible. Ease of desalting is one of the reason we typically use DNA targets no longer than 30 bases: shorter oligonucleotides minimize the number of sodium binding sites. 16. Using low (5 µM total DNA) concentration and low (4 µL/ min) flow rate ensures that the production of ions is not limited by the surface concentration of charges available on the droplets compared to the concentration of analytes in the droplets (14). This ensures that all forms of DNA (free and complexed) have equal chances to be ionized. 17. Even if the ligand is not neutral in the experimental conditions, the mass of the “neutral” ligand is used in this formula because the total charge of the complex is already accounted for in “z”. To calculate the mass of the “neutral” ligand, for a given ligand structure, subtract i Daltons (i protons) if your ligand is a cation of charge i+, or add j Daltons (j protons) if your ligand is an anion of charge j-. In fact, in the charged complex, if a ligand is charged, protonation changes must occur on the phosphates in order for the total charge to remain the same. 18. Some ligands may stick to capillaries. If the ligand is more soluble in DMSO than in water, wash the capillaries as follows: first with 1  mL DMSO, then 500 µL methanol, then 500 µL water. Because of the possible remanence of ligands, when testing several ligand concentrations, it is recommended to conduct tests from lower concentration to higher concentration. When testing the binding of several ligands to a given target, design the test so as to alternate ligands of significantly different masses, in order to detect such remanence, and carry out washing if needed.

100

Gabelica

19. Peak areas are proportional to the ion current. However, if all peaks have exactly the same shape, which is usually the case when the signal-to-noise ratio is high, peak heights can be used instead of peak areas in the calculation of peak ratios. In principle, the whole ammonium/sodium adducts distribution of each (1:n)z− complex must be summed up in calculating the area of the (1:n)z− complex. In practice, if the shape of the adduct distribution is exactly the same for n = 0 to N, then the area of one representative peak can be used in the calculation of the peak ratios. Increasing the signal-to-noise ratio and smoothing the mass spectra help obtaining more similar peak shapes. For example, smoothing the isotope distribution is useful to avoid integrating each isotopic peak and summing them up afterwards. 20. Remember that a peak (1:n)z− can be quantified only if its signal-to-noise ratio is ³10. This sets the sensitivity limit of the method, and the order of magnitude of association constants that can be determined. With very large association constants, the signal of free DNA can be impossible to quantify. If so, run experiments with lower total concentrations. With very low association constants, the signal of the complexes can be impossible to quantify. If so, run experiments with higher ligand concentration. In all cases, as the signal-to-noise increases with the number of scans summed up, you can always improve your sensitivity by recording spectra for a longer time, if you have enough sample solution. 21. If the calculated [L] is negative, the association constant cannot be determined. If you find a negative free ligand concentration, this is often due to an error in ligand concentration, either in the stock solution (the concentration of the stock ligand solution cannot be determined precisely when weighing small quantities). Repeat the experiment including the preparation of a new stock solution of ligand. If this happens again, another possibility is that for the system under study the complexes respond in electrospray much better than the free nucleic acid, and that the intensities are not proportional to the concentrations (see Introduction).

Acknowledgments I would like to thank Frédéric Rosu, Nicolas Smargiasso, Joëlle Widart, Dominique Baiwir, and Edwin De Pauw for their comments on the manuscript. The Fonds de la Recherche ScientifiqueFNRS and the University of Liège are gratefully acknowledged for financial support. The author is a FNRS research associate.

Determination of Equilibrium Association Constants of Ligand–DNA

101

References 1. Whitehouse M, Dreyer RN, Yamashita M, Fenn JB (1985) Electrospray interface for liquid chromatographs and mass spectrometers. Anal Chem 57:675–679 2. Fenn JB, Mann M, Meng CK, Wong SF, Whitehouse CM (1989) Electrospray ioni­ zation for mass spectrometry. Science 246: 64–71 3. Fenn JB (2003) Electrospray wings for molecular elephants (Nobel lecture). Angew Chem Int Ed Engl 42:3871–3894 4. Hofstadler SA, Griffey RH (2001) Analysis of noncovalent complexes of DNA and RNA by mass spectrometry. Chem Rev 101:377–390 5. Beck J, Colgrave ML, Ralph SF, Sheil MM (2001) Electrospray ionization mass spectrometry of oligonucleotide complexes with drugs, metals, and proteins. Mass Spectrom Rev 20:61–87 6. Banoub JH, Newton RP, Esmans E, Ewing DF, Mackenzie G (2005) Recent developments in mass spectrometry for the characterization of nucleosides, nucleotides, oligonucleotides, and nucleic acids. Chem Rev 105:1869–1916 7. Hofstadler SA, Sannes-Lowery KA (2006) Applications of ESI-MS in drug discovery: interrogation of noncovalent complexes. Nature Rev Drug Discov 5:585–595

8. Rosu F, De Pauw E, Gabelica V (2008) Electrospray mass spectrometry to study drug-nucleic acid interactions. Biochimie 90:1074–1087 9. Gabelica V, Galic N, Rosu F, Houssier C, De Pauw E (2003) Influence of response factors on determining equilibrium association constants of non-covalent complexes by electrospray ionization mass spectrometry. J Mass Spectrom 38:491–501 10. Mel’nikov SM, Khan MO, Lindman B, Jönsson B (1999) Phase behavior of single DNA in mixed solvents. J Am Chem Soc 121:1130–1136 11. Smirnov IV, Shafer RH (2007) Electrostatics dominate quadruplex stability. Biopolymers 85:91–101 12. Rosu F, Pirotte S, De Pauw E, Gabelica V (2006) Positive and negative ion mode ESI-MS and MS/MS for studying drug–DNA complexes. Int J Mass Spectrom 253:156–171 13. Rosu F, Gabelica V, Houssier C, Colson P, De Pauw E (2002) Triplex and quadruplex DNA structures studied by electrospray mass spectrometry. Rapid Commun Mass Spectrom 16: 1729–1736 14. Kuprowski MC, Konermann L (2007) Signal response of coexisting protein conformers in electrospray mass spectrometry. Anal Chem 79:2499–2506

Chapter 7 Detection of Adriamycin-DNA Adducts by Accelerator Mass Spectrometry Kate Coldwell, Suzanne M. Cutts, Ted J. Ognibene, Paul T. Henderson, and Don R. Phillips Abstract There have been many attempts in the past to determine whether significant levels of Adriamycin-DNA adducts form in cells and contribute to the anticancer activity of this agent. Supraclincal drug levels have been required to study drug-DNA adducts because of the lack of sensitivity associated with many of the techniques employed, including liquid scintillation counting of radiolabeled drug. The use of accelerator mass spectrometry (AMS) has provided the first direct evidence of Adriamycin-DNA adduct formation in cells at clinically relevant Adriamycin concentrations. The exceedingly sensitive nature of AMS has enabled over three orders of magnitude increased sensitivity of Adriamycin-DNA adduct detection (compared to liquid scintillation counting) and has revealed adduct formation within an hour of drug treatment. The rigorous protocol required for this approach, together with many notes on the precautions and procedures required in order to ensure that absolute levels of Adriamycin-DNA adducts can be determined with good reproducibility, is outlined in this chapter. Key words: Adriamycin (doxorubicin), Adriamycin-DNA adducts, Accelerator mass spectrometry, [14C] Radioisotopes

1. Introduction Adriamycin (generic name doxorubicin) is an anticancer chemotherapeutic commonly used against a broad spectrum of cancers and known to act on nuclear processes (1). However, the mechanism of action of Adriamycin remains uncertain (2), and it is likely that it acts by many different mechanisms to kill tumour cells, which is consistent with its broad spectrum of activity. One proposed mechanism of cellular toxicity involves the formation of Adriamycin-DNA adducts that are mediated by intracellular formaldehyde (2). K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_7, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

103

104

Coldwell et al.

Adriamycin-DNA adducts are relatively unstable and have proved difficult to isolate and characterize, although this can now be readily explained with an understanding of their structural features (3–6). The limited stability of Adriamycin-DNA adducts is due to the characteristically unstable aminal linkage, which is both heat and alkali labile. Most Adriamycin-DNA adducts display similar instability with a half-life of 5–40 h at 37ºC and a preference for neutral pH (5). Conditions optimized for isolation of Adriamycin-DNA adducts involve neutral pH with minimal heating above 37ºC and avoidance of chaotrophic agents, which disrupt hydrogen bonding and result in more rapid decomposition of the DNA adducts (7, 8). Adriamycin-DNA adducts formed under physiological conditions have been measured in cells in culture and have been extensively characterized (2, 8, 9). Previously, the best approach for quantification of Adriamycin-DNA adducts utilized 14C radiolabeled Adriamycin (Fig. 1), wherein drug lesions were detected by liquid scintillation counting (LSC) of purified DNA (9). The benefits of decay counting by LSC are ease of use and the broad range of sensitivity of detection. However, LSC is relatively imprecise at its lower sensitivity limits, so is unable to accurately measure small quantities of 14C, necessitating the use of high specific activity radiolabeled compounds and/or the use of artificially high levels of the radiolabeled compounds. In order to enable detection of Adriamycin-DNA adducts by liquid scintillation counting, excessively high (and certainly pharmacologically irrelevant) Adriamycin concentrations were required, or artificial sources of formaldehyde were used to enhance Adriamycin-DNA adduct formation (9). These factors have called into question

Fig. 1. Structure of Adriamycin. The 14C radiolabel is at position 14.

Detection of Adriamycin-DNA Adducts by Accelerator Mass Spectrometry

105

whether adduct formation is a clinically relevant mechanism of action of the drug. Therefore, a more sensitive method was sought to investigate whether adducts are formed at clinically relevant Adriamycin concentrations. Observations of drug action at supraclinical Adriamycin concentrations (greater than 1 µM) are unlikely to be of significance since plasma concentrations of 10–250 nM are typically achieved in the clinic (10). Accelerator mass spectrometry (AMS) is a novel analytical technique that has been used for the detection of 14C in biological samples and is the gold standard for DNA adduct detection because of the greatly enhanced sensitivity relative to conventional means of detecting 14C-labeled biomolecules (11). Prior to development of AMS-based Adriamycin-DNA adduct measurement (12), the lowest Adriamycin-DNA adduct frequency typically measured in cells, with good reproducibility, was a few adducts per 104 base pairs by LSC following treatment of tumour cells in culture with 2 µM Adriamycin (9). Enhancement of Adriamycin-DNA adduct formation by the use of formaldehydereleasing prodrugs has now enabled measurement of adducts using even lower doses of Adriamycin (9). A drawback of AMS detection is that it does not differentiate between the 14C-labeled parent compound and other 14C sources, such as metabolites of the parent compound or contamination from external sources. Although the total radiolabel in a sample can be measured with considerable precision, the identity of the radiolabeled compound is not defined. Another potential limitation is that the extremely sensitive nature of AMS detection renders samples susceptible to contamination, particularly in laboratories which routinely use radioisotopes (13). For these reasons, a stringent design of sample preparation procedures and experimental controls are critically important. AMS has been used to measure [14C]Adriamycin uptake into tissues such as human tumour cells in culture and tissue samples from [14C]Adriamycin-treated tumour-bearing mice (14). For these pharmacokinetic studies of Adriamycin, AMS was over five orders of magnitude more sensitive than HLPC with fluorescence detection (14). However, this may have been partly due to the inability of AMS to differentiate between the parent compound, Adriamycin, and its metabolites, leading to an overestimate of total drug present in samples. Even at Adriamycin concentrations 105-fold lower than detectable by HPLC, clear differences between lower and higher doses were detectable. The limit of sensitivity of AMS detection described in tissues was femtomolar Adriamycin concentrations, which is >105-fold more sensitive than HPLC (14). AMS has not been used previously for the detection of Adriamycin-DNA adducts, nor for Adriamycin in any other subcellular fraction.

106

Coldwell et al.

2. Materials 2.1. Drug Preparation

1. [14-14C]Adriamycin (GE Healthcare). Specific activity 55 mCi/mmol, chemical purity 95.9% (Fig. 1). 2. Glass vials and conical glass inserts (Pico Pro). 3. 1 µL glass syringe (SGE, Texas, USA) to be used ONLY for dispensing the stock [14C]Adriamycin solution. 4. Sterile Milli-Q H2O (Type I water from a Millipore water purification system or equivalent).

2.2. Cell Culture and Drug Treatment

1. Growth medium: RPMI (JRH Biosciences) with 10% foetal calf serum. 2. Growth medium equilibrated in a ventilated vessel overnight at 37ºC in 5% CO2. 3. MCF-7 breast adenocarcinoma cells (15) and the MCF-7/ Dx Adriamycin-resistant subline (see Note 1). 4. Tissue culture 6-well plates. 5. Pipettes (non-AMS dedicated). 6. Disposable absorbent bench protector and disposable waste container for tips. 7. 10 mL centrifuge tubes. 8. Sterile 3 mL transfer pipettes. 9. Screw cap microfuge tubes. 10. Sterilized PBS. 11. Trypsin. 12. Hettich Universal 16 centrifuge for 10 mL tubes 13. Pump for use as a suction apparatus with sterile Pasteur pipettes.

2.3. DNA Extraction from Cells for AMS Sample Preparation

1. Disposable absorbent bench protector and disposable waste containers for tips and liquids. 2. 1.5 mL screw cap microfuge tubes (Interpath). 3. Sterile PBS. 4. QIAamp DNA Blood Mini kit reagents-buffer AL, columns and QIAGEN Protease. 5. Additional QIAGEN collection tubes. 6. RNase A prepared according to the manufacturer’s instructions (aliquoted and stored at −20ºC). 7. 70% ethanol. 8. Centrifuge for 1.5 mL tubes with a biohazard containment rotor lid, for use ONLY for AMS work.

Detection of Adriamycin-DNA Adducts by Accelerator Mass Spectrometry

107

9. Pipette to accommodate 20–200 µL (for use ONLY for AMS work). 10. 200 µl Filter tips. 11. Sterile transfer pipettes. 12. 50ºC heat source, preferably a heating block and not a water bath. 2.4. DNA Purification

1. Disposable absorbent bench protector and disposable waste containers for tips and liquids. 2. Centrifuge for 1.5 mL tubes with a biohazard containment rotor lid, to be used ONLY for AMS work. 3. Screw cap microfuge tubes (Interpath). 4. Pipette 20–200 µL, for use ONLY for AMS work. 5. Filter tips (20–200 µL) 6. QIAGEN DNA Kit reagents – buffer AW2 and collection tubes. 7. Sterile transfer pipettes. 8. Sterile Milli-Q H2O (Type I water from a Milli-Q water purification system or equivalent). 9. Tris-saturated ultrapure phenol is stored at 4ºC (see Notes 2 and 3). 10. Analytical grade chloroform (see Note 3). 11. DNA precipitation buffer: 9% (v/v) sodium acetate (dissolved first to 3 M, pH 5.2 in Milli-Q H2O) and 0.5% (v/v) 20 mg/ mL glycogen in DNase free analytical ethanol (Biolab BSPEL975). Stored at room temperature. Mix well before each use. 12. 70% (v/v) DNase-free analytical ethanol in Milli-Q H2O. 13. 1 mM NaCl in Milli-Q H2O, steam sterilized and stored at room temperature. 14. ND-1000 spectrophotometer (Nanodrop). 15. Liquid scintillation counter. 16. Ready Safe scintillation cocktail (Beckman Coulter, Fullerton, CA) and scintillation vials.

2.5. Preparation of Samples for Submission to AMS Resource at Lawrence Livermore National Laboratory (LLNL)

1. Disposable absorbent bench protector and disposable waste container for tips. 2. GF/A 21  mm filters (Whatman) placed in screw cap tubes using disposable forceps. 3. Screw cap microfuge tubes. 4. 20–200 µL pipette, for use ONLY for AMS work. 5. Filter tips (20–200 µL)

108

Coldwell et al.

6. 1 mM NaCl in Milli-Q H2O, steam sterilized and stored at room temperature. 7. Disposable forceps. 8. Plastic clip seal bags (approximately 20 cm square). 9. LLNL sample submission sheet – available from LLNL upon approval of project.

3. Methods The advantages of AMS for the detection of 14C-labeled biomolecules have been summarized previously (15). The procedure for measurement of [14C]Adriamycin-DNA adducts by AMS was adapted from a method for preparation of DNA extracts from Adriamycin-treated cells for LSC as described previously (9) plus a method for the preparation of DNA from Adriamycin-treated mouse tissues for AMS measurement (16), together with additional modifications as described herein. AMS samples are highly susceptible to 14C contamination, and hence rigorous procedures must be observed to minimize the possibility of contamination and maintain reproducible results (see Notes 4 and 5). Many points on preventing sample contamination during AMS samples have been previously documented (13, 16). Precautions were taken to minimize contamination of samples such as the use of disposable plastic ware, separate reagents, and equipment used solely for AMS sample preparation. Surface protection and gloves were changed frequently to minimize the spread of contamination. Separate equipment was dedicated solely for AMS sample preparations at the final stages of the DNA isolation procedure (when samples are most prone to contamination). Samples prepared for AMS measurement were only ever handled in a laminar flow hood to prevent 14C contamination from airborne sources (see Note 5). The method described below was developed during a phase of preliminary experimentation in collaboration with LLNL and generates reproducible results, typically ±10% (see Table 1). 3.1. Preparation of Drugs

1. Dilute [14C]Adriamycin to a stock concentration of 1  mM using non-AMS dedicated equipment (see Note 6) 2. Aliquot into glass vials and store at −20ºC. 3. In order to prepare a working stock of [14C]Adriamycin, pipette the required volume of Milli-Q H2O into a plastic tube using non-AMS dedicated pipettes and dispense 1 µL of 1  mM [14C]Adriamycin using a AMS-dedicated precision glass syringe.

Detection of Adriamycin-DNA Adducts by Accelerator Mass Spectrometry

109

Table 1 Reproducibility of AMS measurement of Adriamycin-DNA adducts (12). MCF-7 cells were treated with 0, 25, 100 and 500 nM Adriamycin for 4 h. The zero Adriamycin concentration samples were vehicle treated samples. Cellular DNA was isolated and prepared for AMS analysis as described. Values are expressed as Adriamycin-DNA adduct frequency per 107 bp of DNA Adducts/107 bp Adriamycin concentration (nM)

Level of Adriamycin-DNA adducts

Mean ± SDa (intra-expt)

Mean ± SEb (inter-expt)

0

None

0.04 ± <0.012

0.04 ± <0.002c

25

Low

5.32 ± 0.48

4.24 ± 0.96d

100

Intermediate

22.8 ± 3.0

19.0 ± 2.8e

500

High

109.2 ± 7.6

131 ± 22f

Results were obtained from three replicate samples prepared within one experiment, values shown are average ± standard deviation; intra-expt, intra-experimental b Results obtained from independent experiments, values presented are average ± standard error (reproduced from Fig. 2); inter-expt, inter-experimental c n = 3 d n = 3 e n = 5 f n = 2, error is range a

3.2. Drug Treatment of Cells

1. MCF-7 and MCF-7/Dx cells were maintained in growth medium by subculturing at subconfluence. 2. For experiments, seed 7.5 × 105 MCF-7 or MCF-7/Dx cells in 2  mL of growth medium in 35  mm wells, and allow to adhere overnight. 3. Refresh expired media overlaying cells with new growth media equilibrated overnight at 37ºC in 5% CO2. 4. Pipette the required volumes of [14C]Adriamycin. 5. At the end of treatment times, wash cells twice with PBS, and detach with 200 µL trypsin for 5 min at 37ºC. 6. Pellet cells in 10 mL PBS in 10 mL tubes by centrifugation at 440× g for 5 min in a non-AMS-dedicated centrifuge. 7. Aspirate the supernatant, and store cell pellets at −80ºC until required.

3.3. DNA Extraction from Cells for AMS Sample Preparation

1. Prepare QIAGEN kit reagents according to the manufacturer’s instructions. 2. Thaw and resuspend cell pellets in 150 µL PBS. Transfer to 1.5 mL screw capped microfuge tubes.

110

Coldwell et al.

3. Extract cellular DNA using the QIAamp Kit according to the manufacturer’s instructions, as outlined below, together with modifications as indicated at steps 4, 6 and 8 below (see Note 7). 4. Add RNase A to each sample to 2 mg/mL to eliminate RNA from DNA extracts. 5. Add 20 µL protease Q and 200 µL buffer AL to each sample and mix by pipetting 6. Conduct cell lysis at 50ºC for 30 min rather than 65ºC for 10 min (as recommended in kit manual) to minimize the loss of heat-labile adducts. 7. Carefully load the cell lysate onto columns using disposable transfer pipettes. Do not overfill the column. Centrifuge for 1 min at maximum speed. For large or frothy lysates, repeat this step with remaining sample if necessary. 8. Wash columns twice with 500 µL buffer AW2, as buffer AW1 has previously been shown to reduce adduct yield (9). 9. Spin columns in a fresh collection tube to remove any residual wash buffer. 10. Elute DNA from column into Milli-Q H2O rather than buffer AE. Add 200 µL Milli-Q H2O to the column and incubate for 5 min at room temperature. Elute by centrifugation for 1 min at maximum speed. 3.4. DNA Purification

1. Replace the bench surface protector and introduce equipment and reagents for the second phase of the procedure (see Note 8). 2. Transfer eluates from collection tubes into screw capped microfuge tubes. 3. Each DNA sample is then extracted by mixing with an equal volume (typically 200 µL) of phenol, centrifuged at 15.7× g for 5 min and the upper aqueous phase retained (see Notes 9 and 10). 4. Repeat phenol extraction (as per step 3 above). 5. Carry out one chloroform extraction (as per step 3 above). 6. Precipitate DNA with 2.1 volumes of DNA precipitation buffer, mix by inversion and incubate for 10 min at −80ºC. 7. Pellet DNA by centrifugation at 15.7× g for 10 min. 8. Aspirate supernatant and wash DNA pellets twice with 70% ethanol. 9. Air dry pellets thoroughly in the laminar flow hood (see Note 11).

Detection of Adriamycin-DNA Adducts by Accelerator Mass Spectrometry

3.5. Preparation of Samples for Submission to AMS Resource at Lawrence Livermore National Laboratory (LLNL)

111

1. Prepare filters by inserting one Whatman GF/A filter per tube using disposable forceps (see Note 12). 2. Resuspend DNA pellets in 200 µL of 1 mM NaCl. This is best achieved by solubilizing the DNA pellet for >10 min at room temperature, followed by at least 30 repetitions of pipetting (see Notes 13 and 14). 3. Transfer 100 µL of each sample to a separate tube to be used for DNA quantification and 14C estimation (designated LAB). The remainder (designated AMS) is then stored at −20ºC for sample submission to the AMS facility at LLNL. 4. Determine DNA concentration and quality (see Note 15) of LAB samples using a Nanodrop ND-1000 spectrophotometer. 5. Mix the remainder (preferably the majority) of LAB sample with 1  mL of scintillation fluid and subject to LSC (see Note 16). 6. Adjust DNA concentrations of AMS samples by diluting to the required concentration with 1 mM NaCl. DNA samples typically range from 0.5 to 2 µg DNA (see Note 17). 7. Prepare samples for submission to LLNL by drying the required volume of each sample onto Whatman GF/A filters. Retain the remainder of the AMS sample for further DNA measurements. 8. Prepare a blank filter and a vehicle control filter. 9. Dry filters overnight in a laminar flow hood by leaving tubes open overnight with all overhead white and UV lights switched off. If necessary, use aluminium foil to occlude light. 10. Using the remainder of each of the equalized AMS samples (obtained in step 7), measure the DNA content using a ND-100 spectrophotometer to determine the precise DNA concentration of the diluted sample that was submitted to LLNL (see Note 18). 11. Samples sent to LLNL for measurement by AMS require submission documentation including the quantity of DNA, composition of buffer and expected level of 14C for each sample (see Notes 19 and 20). 12. Once DNA samples are dried onto filters, label each sample with the project number, sample identifier, date and name of experimenter. Place samples in a sealed bag for shipping. 13. Staple the sample submission sheet to the bag, so as to indicate whether samples have been opened in transit prior to arrival at LLNL.

112

Coldwell et al.

14. Samples, submission sheet, and any other necessary shipping documentation are then sent to: Sample Submission, NIH Resource for Biomedical AMS, Bioscience Directorate, Lawrence Livermore National Laboratory, University of California, 7000 East Avenue, L-441, Livermore, CA 94550, USA. 15. Upon submission email your designated LLNL contact to alert them to receive your samples. LLNL contact(s) will then respond with notification, when the samples have arrived. Results are returned by email as Fraction Modern 14C per sample (F. Mod.), which is described further below. 3.6. Detection of Adriamycin DNA Adducts by AMS

1. The measurement of 14C in DNA samples is carried out at the Biomedical AMS resource at LLNL. 2. At the Biomedical AMS Center carrier carbon (Ccarrier) is added in the form of 1 µL of tributyrin (0.62  mg Ccarrier). Samples are prepared by combustion to CO2 followed by reduction to filamentous graphite using standard procedures, which have been described in detail elsewhere (17). Results were reported from LLNL as F. Mod. per sample (a ratio of total:14C in a sample). For each sample, F. Mod. per µg DNA was calculated and then corrected for the vehicle control. 3. The number of Adriamycin-DNA adducts were determined using the following calculations. In order to determine the quantity of Adriamycin bound to DNA, F. Mod. ratios were converted to an absolute quantity of 14C per sample (1 modern = 9.79 × 10−17 mol 14C/mg C). The quantity of DNA in each case added negligible carbon to the sample. For one F. Mod., the number of moles of 14C [n(14C)] is defined by Eq. (1) below. n(14C) = 1 modern 14C ×



9.79 × 14C × mg C

0.62 mg Ccarrier = 6.07 × mol 14C



(1)

4. For 1 F. Mod. the number of moles Adriamycin, n(Adr), was then calculated from the specific activity of [14C]Adriamycin (55 mCi/mmol; 0.8814 atoms 14C/molecule of Adriamycin) according to Eq (2) below.

n( 14 C) 6.07 × 10 −17 mol 14C = specific activity  0.8814 14C   molecule of Adr    = 6.89 × 10 −17 mol Adr

n(Adr) =

(2)

Detection of Adriamycin-DNA Adducts by Accelerator Mass Spectrometry

113

5. The quantity of DNA (µg) was converted to a molar quantity using MWbp of 653.72 g/mol and Adriamycin-DNA adducts per bp DNA were then calculated. One F. Mod. of 14C per µg DNA can then be converted to Adriamycin adducts per 107 bp DNA as defined by Eq (3) below, where f = n(14C) in F. Mod. and D = n(DNA) in µg. Adriamycin adducts per 107 bp DNAmol = f × 0.450 (3) D 3.7. Results

Adriamycin-DNA adduct detection by AMS exhibits specificity as evidenced by the linear dose response and the absence of detectable “adducts” in untreated controls (Fig. 2). MCF-7/Dx cells exhibited lower levels of adduct formation than the parent Adriamycin-sensitive cells line, consistent with drug efflux as a mechanism of resistance. The procedure presented here routinely produced samples of sufficient quality for AMS measurement with good reproducibility (12), both within experimental replicates and between separate experiments. The reproducibility of AMS measurement of radioisotopes itself is very high (18). A linear dose response resulted from 0 to 500 nM Adriamycin (Fig. 2), however, absolute Adriamycin-DNA adduct levels were 1.7-fold lower then results obtained by LSC (12). By AMS

Adducts /107 bp DNA

60

40

20

MCF-7 0

MCF-7/Dx 0

50

100

150

200

250

Adriamycin (nM) Fig.  2. Dose response of Adriamycin-DNA adduct formation in [14C]Adriamycin-treated cells (12). MCF-7 cells were treated with 0, 25, 50, 75, 100, 200, and 500 nM Adriamycin for 4 h. MCF-7/Dx cells were treated with 0, 50, 100, 200, and 500 nM Adriamycin for 4 h. The zero Adriamycin concentration samples were vehicle treated samples. Cellular DNA was isolated and prepared for AMS analysis as described. Values are expressed as Adriamycin-DNA adduct frequency per 107 base pairs (bp) DNA and were obtained from between two and five independent experiments. Error bars for 0–100 nM points are SEM (n = 3–5). Error bars for 200 and 500 nM points represent range (n = 2). The full data range (0–500 nM Adriamycin) is shown in the inset.

114

Coldwell et al.

measurement, 500  nM Adriamycin resulted in ~130 adducts/ 106 bp DNA whereas, by LCS, 2 µM Adriamycin resulted in ~1400 adducts/106 bp DNA. The lack of consistency is likely to be due to the limited accuracy and reliability of LSC detection at low levels of 14C. To address this variation, cells were treated with either 100  nM or 1 µM Adriamycin and samples prepared for AMS measurement by dilution of the samples to give approximately equivalent absolute 14C quantities. The tenfold difference in Adriamycin concentration resulted in an approximately proportional increase in Adriamycin-DNA adduct formation (12). The AMS and LSC procedures differ in that AMS samples are dried onto filters, and extra measures are taken to prevent contamination in the ultrasensitive AMS procedure although these factors are unlikely to result in the 1.7-fold higher adducts observed by LSC. Using AMS detection, the sensitivity of Adriamycin-DNA adduct detection was increased by three orders of magnitude over LSC (12). If the higher levels of adducts observed with LSC were due to sample contamination, untreated controls and Adriamycinalone treated controls would suffer from the contamination also, and they did not. Using AMS detection, the sensitivity of Adriamycin-DNA adduct detection was easily increased by three orders of magnitude over LSC (12) (see Table 2). MCF-7 cells were treated with 100  nM Adriamycin and Adriamycin-DNA adduct formation observed for up to 8  h (Fig.  3). Adducts are formed rapidly with the earliest observed adduct levels at 30 min after treatment. Adduct levels are increased in a linear fashion over 0–8  h, which indicates that adduct formation was not limited by exhaustion of Adriamycin or formaldehyde available in the cell in this timeframe. It is interesting to note that only 0.02% of the initial quantity of Adriamycin was detected as adducts at 8 h, indicating that further enhancement of Adriamycin-DNA adduct formation could be achieved by optimizing the timing of drug additions, since it has previously

Table 2 Comparison of Adriamycin-DNA adducts detected by LSC and AMS (12). Cells were treated for 4 h and Adriamycin-DNA adducts were measured by AMS and LSC as described Method of detection Adriamycin concentration Adducts/107 bp DNA AMS

100 nM

37.6

AMS

1 µM

442

LSC

2 µM

1,400 ± 540a

Error is SEM, n = 5.

a

Detection of Adriamycin-DNA Adducts by Accelerator Mass Spectrometry

115

Fig. 3. Time course of Adriamycin-DNA adduct formation (12). MCF-7 cells were treated with 100 nM Adriamycin for 0, 0.5, 1, 2, 4 and 8 h. Cellular DNA was isolated and prepared for AMS analysis as described. Values are expressed as mean Adriamycin-DNA adduct frequency per 107 base pairs (bp) DNA and were obtained from two replicate experiments. Error bars represent absolute error of two replicate experiments.

been shown that 2 h pretreatment with formaldehyde releasing prodrugs is optimal for enhancement of Adriamycin-DNA adduct formation (9). The linear time course shown in Fig. 3 suggests that additional adducts would form at longer treatment times. AMS has successfully been used here to measure AdriamycinDNA adducts from cells in culture. There is the potential to develop this procedure for other applications such as studies of Adriamycin treatment in mice and humans to establish the extent of formation of Adriamycin-DNA adducts in tumour and normal tissues. The same principle could also be applied to other DNA-binding chemotherapeutic drugs as a biomarker of adducts as a mechanism of action. For such applications, the AMS technique would need to be optimized to develop an appropriate experimental process, such as the need for radiolabeled compounds and suitable sample preparation. Highly sensitive detection of Adriamycin-DNA adducts has long been thought of the critical evidence required to support the hypothesis that adducts are a likely mechanism of action of Adriamycin (10). Here, a novel method is presented for highly sensitive and specific detection of adducts at clinically relevant Adriamycin concentrations using AMS methodology. AdriamycinDNA adducts formed at doses and in a timeframe which correlated with plasma concentrations of Adriamycin in patients. Further support for this hypothesis could be sought using an in vivo model and AMS detection to demonstrate whether Adriamycin-DNA adducts form in whole organisms at clinically relevant drug doses.

116

Coldwell et al.

4. Notes 1. MCF-7/Dx cells were generated by growth in escalating Adriamycin concentrations, which resulted in increased expression of the multidrug resistance pump, P-glycoprotein (19). Adriamycin-resistant cells were maintained under selective pressure of 300 nM Adriamycin as previously reported (20). 2. Phenol should be colourless. Pink colour of phenol indicates oxidization, and it should not be used as oxidized phenol binds covalently to DNA and may interfere in subsequent analysis. 3. Phenol and chloroform are hazardous substances. Appropriate personal protective equipment and ventilation should be used. 4. Avoid generating any 14C spills (i.e. proceed slowly and cautiously) as these can result in contaminated DNA later. Be cautious with every spill and situations, which may produce an aerosol. Clean spills with absorbent tissue and DECON then with 70% (v/v) ethanol or replace bench protector. 5. Although the isolation is carried out in a laminar flow hood (likely to be a cell culture facility), do not “sterilize” surfaces and equipment with alcohol. Instead, use the laminar flow to avoid introducing airborne radioisotope sources, and be conscious of what is carrying surface dirt into the hood. 6. Drug treatment of sample for AMS involved the use of high levels of 14C. It was therefore acceptable to use standard laboratory pipettes which had previously been used for conventional 14C work. 7. These modifications were previously established to maximize DNA yield with minimal adduct losses (9). 8. Store equipment and reagents for the DNA purification stage of the procedure separately from other equipment. 9. Laboratory vortexes are best avoided (as is any other equipment used for conventional laboratory work using radioisotopes) hence, vigorous mixing is best achieved by other means within the confines of the laminar flow hood (e.g. shaking and pipetting). 10. Collect tubes containing phenol-chloroform waste for appropriate solvent disposal. 11. Completely dry DNA pellets are apparent by a transparent, glassy appearance (not opaque and white). 12. Do not handle filters except with new, unused disposable forceps to avoid contamination at this late stage of the sample purification.

Detection of Adriamycin-DNA Adducts by Accelerator Mass Spectrometry

117

13. It is critically important that DNA is totally resuspended and thoroughly mixed prior to subsampling and determination of DNA bp concentration. 14. 1 mM NaCl is used to ensure DNA remains duplex throughout the procedure. 15. A260:A280 of 1.8–2.0 is expected for highly purified duplex DNA. 16. This preliminary LSC measurement of each sample was taken to establish that 14C levels were below the level of detection of LSC. Two to three readings of each sample were taken to be certain that there was no 14C above background levels. 17. Samples were prepared to result in up to 100 Frac. Mod. per sample, the appropriate level of 14C for AMS detection. Higher levels of 14C may be acceptable but 100 F. Mod. is the preferable maximum limit to be measured by AMS. Samples were usually prepared as equal volumes dispensed onto filters to yield approximately equal DNA quantities (although in some cases it was necessary to alter the quantity of DNA in each sample submitted for AMS measurement to ensure that the samples would fall within the acceptable range of detection of 0–100 F. Mod. per sample). 18. This retrospective quantification provided the most accurate measurement of DNA quantities that were sent for AMS detection without compromising sample quality. The quantities dispensed onto filters were intended to be approximately equal but were found to vary somewhat, and the precise determination of DNA quantities were used for subsequent calculations. 19. Estimated 14C levels were expressed as <1 DPM when no counts above background were measured in duplicate samples by LSC (or where indicated from previous experiments). 20. LLNL requirements for sample submission to the BioAMS resource are detailed by the resource staff at http://bioams. llnl.gov/sample.php.

Acknowledgments Work performed at the Research Resource for Biomedical AMS operated at LLNL under the auspices of the U.S. Department of Energy under contract DE-AC52-07NA27344. The Research Resource is supported by the National Institutes of Health, National Center for Research Resources, Biomedical Technology Program grant #P41 RR13461.

118

Coldwell et al.

References 1. Cummings J, McArdle CS (1986) Studies on the in vivo disposition of adriamycin in human tumours which exhibit different responses to the drug. Br J Cancer 53:835–8 2. Cutts SM, Nudelman A, Rephaeli A, Phillips DR (2005) The power and potential of doxorubicin-DNA adducts. IUBMB Life 57: 73–81 3. Wang AH, Gao YG, Liaw YC, Li YK (1991) Formaldehyde cross-links daunorubicin and DNA efficiently: HPLC and X-ray diffraction studies. Biochemistry 30:3812–3815 4. Zeman SM, Phillips DR, Crothers DM (1998) Characterization of covalent AdriamycinDNA adducts. Proc Natl Acad Sci U S A 95:11561–11565 5. van Rosmalen A, Cullinane C, Cutts SM, Phillips DR (1995) Stability of Adriamycininduced DNA adducts and interstrand crosslinks. Nucleic Acids Res 23:42–50 6. Cutts SM, Phillips DR (1995) Use of oligonucleotides to define the site of interstrand cross-links induced by Adriamycin. Nucleic Acids Res 23:2450–2456 7. Skladanowski A, Konopa J (1994) Interstrand DNA crosslinking induced by anthracyclines in tumour cells. Biochem Pharmacol 47:2269–2278 8. Cullinane C, Cutts SM, Panousis C, Phillips DR (2000) Interstrand cross-linking by adriamycin in nuclear and mitochondrial DNA of MCF-7 cells. Nucleic Acids Res 28:1019–1025 9. Cutts SM, Rephaeli A, Nudelman A, Hmelnitsky I, Phillips DR (2001) Molecular basis for the synergistic interaction of Adriamycin with the formaldehyde-releasing prodrug pivaloyloxymethyl butyrate (AN-9). Cancer Res 61:8194–8202 10. Gewirtz DA (1999) A critical evaluation of the mechanisms of action proposed for the antitumor effects of the anthracycline antibiotics adriamycin and daunorubicin. Biochem Pharmacol 57:727–741 11. Poirier MC, Santella RM, Weston A (2000) Carcinogen macromolecular adducts and their measurement. Carcinogenesis 21:353–359

12. Coldwell KE, Cutts SM, Ognibene TJ, Henderson PT, Phillips DR (2008) Detection of adriamycin-DNA adducts by accelerator mass spectrometry at clinically relevant Adriamycin concentrations. Nucleic Acids Res 36:e100 13. Buchholz BA, Freeman SPHT, Haack KW, Vogel JS (2000) Tips and traps in the 14C bioAMS preparation laboratory. Nucl Instrum Methods Phys Res B 172:404–408 14. DeGregorio MW, Dingley KH, Wurz GT, Ubick E, Turteltaub KW (2006) Accelerator mass spectrometry allows for cellular quantification of doxorubicin at femtomolar concentrations. Cancer Chemother Pharmacol 57: 335–342 1 5. Soule HD, Vazguez J, Long A, Albert S, Brennan M (1973) A human cell line from a pleural effusion derived from a breast carcinoma. J Natl Cancer Inst 51: 1409–1416 16. Dingley KH, Ubick EA, Vogel JS, Haack KW (2005) DNA isolation and sample preparation for quantification of adduct levels by accelerator mass spectrometry. Methods Mol Biol 291:21–27 17. Turteltaub KW, Vogel JS (2000) Bioanalytical applications of accelerator mass spectrometry for pharmaceutical research. Curr Pharm Des 6:991–1007 18. Ognibene TJ, Bench G, Brown TA, Peaslee GF, Vogel JS (2002) A new accelerator mass spectrometry system for C-14-quantification of biochemical samples. Int J Mass Spectrom 218:255–264 19. Calcabrini A, Meschini S, Stringaro A, Cianfriglia M, Arancia G, Molinari A (2000) Detection of P-glycoprotein in the nuclear envelope of multidrug resistant cells. Histochem J 32:599–606 20. Cutts SM, Nudelman A, Pillay V, Spencer DMS, Levovich I, Rephaeli A, Phillips DR (2005) Formaldehyde-releasing prodrugs in combination with adriamycin can overcome cellular drug resistance. Oncol Res 15:199–213

Chapter 8 Molecular Modelling Methods to Quantitate Drug-DNA Interactions Hao Wang and Charles A. Laughton Abstract We describe a molecular modelling method for calculating the binding affinity of ligands for DNA. Though theoretically applicable to any form of noncovalent interaction, we concentrate on the case of predicting the sequence selectivity of a minor-groove binding ligand. The method is based on performing molecular dynamics (MD) simulations on DNA sequences, with and without the ligand bound, and postprocessing the molecular dynamics trajectory data to obtain approximate free energies of binding. We discuss issues relating to the preparation of the structures for simulation, choices for the molecular dynamics simulation method itself, methods for evaluating the reliability and stability of the simulation data, and finally alternative approaches to postprocessing the data to extract approximate free energies of binding. Key words: Drug-DNA interactions, Thermodynamics, Free energy, Molecular dynamics, Molecular modelling, Poisson–Boltzmann, Generalized Born, Normal Mode Analysis, Configurational entropy

1. Introduction Over the last 15 years, there has been a drive to produce “smart” DNA targeting agents that interfere selectively with, for example, a gene of interest though their ability to read the genetic code and bind with high affinity only to a chosen DNA sequence. Such agents could be developed into far more selective and versatile pharmaceutical agents than the current generation of nonspecific DNA damaging agents that, despite their severe side effect profile, are still widely used in cancer chemotherapy. Although there have been some impressive developments in the field of “lexitropsins” (1–3), in general, we are still unable to design a pharmaceutically acceptable small molecule with the ability to target any DNA sequence we K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_8, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

119

120

Wang and Laughton

DG

Unbound = DNA + ligand

Bound

D G = Gbound – (GDNA + Gligand)

Fig.  1. The molecular systems whose simulation is required for the calculation of DNA-ligand binding energies.

choose. At least part of this is due to our limited understanding of the process of drug-DNA molecular recognition, so this has been a fertile field for theoretical and modelling studies. The overall process that a simulation, or set of simulations, attempts to mimic is illustrated in Fig. 1. In the process of the ligand binding to the DNA minor groove, the free energy change is

DG = Gbound – (GDNA + Gligand)

(1)

in which, Gbound refers to the absolute free energy of the complex (including its solvent environment), while GDNA and Gligand refer to the free energy of the DNA and ligand in the unbound state. The aim of simulation studies is to estimate each of these quantities, and so DG. A variety of different computational approaches of varying degrees of complexity are available to study this. Here, the method we are concerned with does this on the basis of undertaking, and then analyzing a series of molecular dynamics simulations. The approach was pioneered by Kollman, Case and coworkers (4, 5), and has since been used by many other groups, specifically to study nucleic acids and their interactions (see, e.g., (6–9)). Each free energy, G, may be calculated from the corresponding enthalpy (H), and entropy (S) terms:

G = H –TS

(2)

where T is the absolute temperature. The enthalpy term H may be estimated directly from a molecular dynamics simulation, as if this is well equilibrated and well enough sampled (both issues we will discuss later), then

H » <E >

(3)

where <E> is the average value of the molecular mechanics energy of the system, as calculated from analysis of every snapshot in the

Molecular Modelling Methods to Quantitate Drug-DNA Interactions

121

MD simulation. The calculation of the enthalpy change, DH, for drug-DNA binding, thus becomes DH » < E >bound –(< E >



DNA

+ < E >ligand)

(4)

Although in theory these individual averaged energies could be extracted directly from the appropriate simulation data, in practice, this is not ideal, as the individual terms are large and show large fluctuations. DH, is thus the (generally small) difference between large and fluctuating terms, and so difficult to estimate reliably. The reason for these individual energy terms showing large fluctuations is due to the dynamics of the large number of water molecules that are generally included in a molecular dynamics simulation to ensure that it mimics the biologically relevant aqueous environment. It is very difficult to run an MD simulation for long enough to ensure that the water environment is fully sampled, and so the error bars on <E> can be large. To overcome this problem, it is common to throw away the water molecules included in the simulation, and reanalyze the energy of each snapshot using a continuum, or implicit, water model instead. The Poisson–Boltzmann and Generalized Born implicit solvent models (5) are the two most commonly used combinations. These calculations may be further enhanced by including a solute– solvent interaction energy term that depends on the solvent accessible surface area, so the approaches are often described as the PB-SA or GB-SA methods. Each effect provides a correction to the energy of the solute because of the solvent that is an average over all possible solvent molecule configurations, so the ‘noise’ in the water energy terms is greatly reduced (Fig. 2).

B

A

E

E

snapshot

snapshot

Fig.  2. By stripping away the explicit water molecules included in the original MD simulation (a), and replacing them with an implicit water model (b), the ‘noise’ in the energy calculations is greatly reduced, so < E > can be estimated with greater precision.

122

Wang and Laughton

The entropy term, S, is more difficult to evaluate directly. For a typical biomolecular system, it has contributions from both the solvent (water) environment, and from the solute (the DNA and/or ligand, in this case):

S = Ssolute + Ssolvent

(5)

Fortunately, the continuum water models discussed above already take into account the solvent entropy term, so we are left with just Ssolute to determine from the MD simulation data. If we switch from considering individual thermodynamic terms to thinking about the changes that accompany the binding process, we can appreciate more clearly what needs to be evaluated (Eq. 8.6). DS = DStrans +DSrot + DSconfig+ DSvib)



(6)

Firstly, when the drug binds to the DNA, it loses its translational and rotational freedom. The entropy penalties associated with this (DStrans + DSrot) can be calculated from statistical mechanics theory, but we might want to note that if we are interested in the energy differences between a drug binding to one sequence and another, these terms will cancel out. Next, when the ligand and DNA interact, there are changes in the internal degrees of freedom in the molecules. Free rotations about single bonds may be impeded (DSconfig), and vibrational modes may be changed (DSvib). The calculation of Sconfig has frequently been seen as a difficult problem, since there can be a “grey area” between what might be regarded as a restricted rotation and what could be a low-frequency vibration. The simplest option is to ignore it, assuming that it is a small term in comparison with the other entropic terms, particularly the hydrophobic term and consider only Svib, which may be calculated from individual snapshots of the system using Normal Mode Analysis (NMA) methods (10–12) (Eq. 8.7), where vi are the vibrational frequencies, N the number of atoms in the molecular system, h the Planck’s constant, T the absolute temperature, and kB Boltzmann’s constant.



S vib

é ê 3N - 6 ê = å ê i =1 ê ê kBT ë

ù ú - hvi æ öú hvi kBT - ln ç 1 - e ÷ ú æ khvTi ö è øú B ç e - 1÷ ú è ø û

(7)

But, while it is true that Sconfig is frequently a small term, it is very often the case that we are interested in understanding the origins of differences in binding affinities, for example, between one ligand and another, or between one ligand and two alternative DNA sequences. In such cases, this entropic term may be much more significant. With this in mind, in our previous paper (13), the Schlitter method (14) was used to calculate this term from analysis of the eigenvalues, w, resulting from diagonalization of the

Molecular Modelling Methods to Quantitate Drug-DNA Interactions

123

Cartesian coordinate covariance matrix generated from the MD trajectory (Eq. 8.8). The method effectively calculates Sconfig + Svib, and there have been many reports of its successful application; however, in this particular case, we noted that the magnitude of the configurational entropy terms was unreasonably large.

æ e2 ö S » 0.5k å ln ç 1 + 2 ÷ a ø è i

(8)

where ai = Wi / kT , and the sum is over all nontrivial vibrations. At this stage, we have the methodology in hand to calculate all the thermodynamic terms involved in a (noncovalent) drugDNA recognition system. Some of the energy terms will come from the original molecular mechanics simulation forcefield, and the rest from the implicit solvation model. These ‘hybrid’ methods are consequently often referred to as the MM/GB-SA and MM/PB-SA methods. In Eq. (8.1), it would appear that this will require data from three separate simulations: One on the drug-DNA complex, plus one each on the isolated DNA and drug. However, in practice this is not always necessary. Because of the ‘post processing’ strategy and implicit solvation model used (Fig. 2), we can instead perform a single MD simulation on the drug-DNA complex, and then generate “pseudo trajectories” for the supposedly isolated DNA and drug by just extracting the relevant coordinates from this, and embedding them in their own implicit water environment (Fig. 3).

A

B

Fig.  3. The “single trajectory” approach. The original trajectory of the complex is stripped of solvent molecules and split to produce additional pseudo-trajectories of the DNA and ligand (a). Each is then analyzed within the context of a PB or GB-SA continuum solvent model (b).

124

Wang and Laughton

DD G

+

DNA1.lig

+

DNA2

DNA1

DNA2.lig

DD G = (GDNA1 +GDNA2.lig ) – (GDNA1.lig +GDNA2) Fig.  4. Simulating a competition dialysis experiment, enables relative binding free energies of a ligand to different sequences to be evaluated without performing a simulation on the unbound ligand.

There is, however, a serious approximation implicit in this approach, namely that the ligand and DNA have the same dynamics (i.e., sample the same configurations with the same probability) in the free state and in the complex. For a totally rigid system, this could be true, but while typical minor groove binding ligands are relatively rigid, DNA is not, and induced fit could be a major issue in the recognition process affecting both the enthalpic and entropic components of DG (15, 16). For this reason, it may be better to adopt a “twin trajectory” approach (13) in which a simulation of the free DNA is performed as well, and used for the energy calculations. What about a simulation of the free ligand? In many cases, it will only be relative binding affinities that are required, in which case we can avoid the need for a separate simulation of the unbound ligand by simulating competition dialysis experiments, as shown in Fig. 4.

2. Materials 2.1. Molecular Models

Initial models for the drugs, DNA sequences and their complexes may be obtained from a variety of sources. If crystal structures are available, this is an obvious start point. Crystal structures for the isolated ligands may be available from the Cambridge Structural Database (CSD), and/or for the DNA or drug-DNA complex from the Protein Data Bank (PDB). In the absence of crystallographic or NMR data, models for the ligands may be constructed using standard molecular modelling packages, while DNA sequences may be built either using tools included in many molecular modelling packages, or can be generated and downloaded from the web service at http://structure.usc.edu/ make-na/server.html.

Molecular Modelling Methods to Quantitate Drug-DNA Interactions

125

2.2. Molecular Dynamics Simulations

Many different MD simulation software packages are available. Most simulations on drug-DNA systems have been done using either AMBER (17), CHARMM (18), or NAMD (19) (implementing either an AMBER or CHARMM forcefield). Detailed discussion of MD simulation methods is outside the scope of this article.

2.3. Trajectory Validation

Confirming that an MD simulation has generated a well equilibrated and well sampled ensemble for analysis is not straightforward, but approaches based on Principal Component Analysis (20) are often very useful. Though many simulation software packages contain tools that can do this, a convenient “stand alone” package PCAZIP, is available from the Collaborative Computational Project for Biomolecular Simulation (CCPB) at: http://www.ccpb.ac.uk/software.

2.4. Thermodynamic Analysis

For the approach discussed here, the AMBER software suite provides some particularly useful tools, and these will be focussed on that which follows, but it should be pointed out that other packages should be able to perform more or less the same sort of analysis.

3. Methods 3.1. Molecular Models

Having obtained initial coordinates for the ligand(s) and DNA duplex(es) of interest by one of the approaches outlined in Subheading 2.1, it is necessary to prepare the full molecular systems for MD simulations. One of the first steps that may be required is to dock the ligand into the appropriate site on the DNA duplex of interest, if a crystal or NMR structure for this is not available. There are many approaches to this, but obviously it makes sense to make as much use as possible of existing information – e.g. structures of the same ligand bound to different DNA sequences, or of analogous ligands bound to the same DNA sequence. Interactive graphics should be used to check and refine the docked pose, as an unfavourable starting structure may seriously affect the speed and reliability of the following equilibration and relaxation steps. The next stage is usually to embed the solute in a volume of water (21), and ensure overall electrical neutrality by adding enough positive ions to counteract the negative charge of the DNA (see Notes 1 and 2). The exact methodology for doing this will depend on the chosen simulation package, but in general a system with periodic boundary conditions is established (see Note 3).

126

Wang and Laughton

3.2. Molecular Dynamics Simulations

The MD simulations pass through three key stages. Stage 1 is thermal equilibration. This involves assigning initial velocities to each atom of the system, and then running MD simulations during which energy is exchanged between atoms, and probably between the system and an external heat bath, until the temperature has stabilized at the required value (typically 300 K). This stage must be performed carefully, as initial problems in the system – e.g., bad contacts between the ligand and the DNA from a poor docking procedure, or between the solute and the initially placed water molecules, may cause unstable behavior. Our typical thermalization procedure is as follows: (a) Energy minimization of the structure only allows water to move (b) Energy minimization of the whole system (c) 10  ps of dynamics on the water only, at 100  K. Solute restrained to its initial position with a high force constant of 100 kcal/mol/Å2. (d) 10 ps of dynamics raising the temperature gradually to 300 K, solute still restrained. (e) A series of 10 ps simulations over which the restraints on the solute are reduced from 100 kcal/mol/Å2 to zero via 50, 10, 5, 2, and 1 kcal/mol/Å2 stages. Stage 2 is structural equilibration. Almost certainly, the initially built model of the system will not be in a “correct” conformation, and a stage of structural relaxation is required, during which the conformation of the system shifts more or less irreversibly towards a lower energy state, or rather envelope of low-energy conformations. This process typically takes much longer than the thermalization process described above. Once this is achieved, the MD moves on to stage 3: “Production” MD during which snapshots of the system are saved every picosecond or so to build up, over time, a representation of a Boltzmann-weighted ensemble of structures on which the thermodynamic analysis can be based.

3.3. Trajectory Validation

Determining when structural relaxation is complete in stage 2, and when a well sampled ensemble has been collected in stage 3, is greatly aided by the application of Principal Component Analysis methods. The process typically involves PCA of a trajectory, and examination of the time evolution of the projections of the major (maybe top five or so) eigenvectors. Drift in these projections is a good guide to relaxation, and a stable and smooth histogram of projection distributions suggests good sampling (Note 4). An example of the sort of data that can be obtained is shown in Fig. 5 (Hao and Laughton, unpublished data), where it can be seen that simulation (a) appears stable and well equilibrated, but simulations (b) and (c) show relaxation events in the projections of their first eigenvectors.

Molecular Modelling Methods to Quantitate Drug-DNA Interactions

127

Fig. 5. Example for principal component projection data. The graphs for the projections of the first three eigenvectors of a 5 ns simulation of Hoechst 33258 bound to one DNA sequence (a) indicate good sampling and equilibration. The graphs for the first component of the 5 ns simulations of Hoechst 33258 bound to two other sequences (b) and (c) clearly show relaxation events.

3.4. Thermodynamic Analysis

The essential steps in the form of thermodynamic analysis described in the introductory section are as follows. Firstly, snapshots from the equilibrated and well sampled section of the MD trajectory of the drug-DNA complex are extracted and stripped of their waters and counterions. If the “single trajectory” method is being used, then “pseudo trajectories” of the unbound DNA and unbound ligand are generated from these snapshots by extracting just the appropriate coordinates. The MM/GB-SA or MM/PB-SA energy of each snapshot is then calculated using an appropriate molecular simulation package. Ensemble averages are then determined, and since the continuum solvent models implicitly evaluate solvent-associated free energy terms, not pure enthalpies, we can equate <E> to G rather than H (Eq. 3) and through application of Eq.  8.1 the approximate free energy change on ligand binding is calculated. The AMBER simulation package includes a module (MM_PBSA) that more or less automates this whole procedure. As described above, there are sophistications that can be introduced to potentially improve the accuracy of the

128

Wang and Laughton

method. Firstly, a “twin trajectory” method may be used in which a separate simulation of the unbound DNA sequence is performed. This permits a correction to be made for any structural reorganization of the DNA that accompanies ligand binding. Secondly, the configurational entropy change may be estimated using either normal mode analysis or a PCA-based method such as the Schlitter equation (see Note 5).

4. Notes 1. Implicit solvent simulations. Since the approach described here involves “throwing away” the explicit solvent environment and recalculating energies using an implicit model, one could ask the question as to whether it would not be more straightforward to conduct the initial molecular dynamics simulations using an implicit solvent model as well. Although this can be done (22), in our experience this is probably not advisable. Firstly, though the Generalized Born solvation model can produce reasonable trajectories for drug-DNA systems, they seem less robust – there is quite a chance that at some stage the system will undergo a large, unrealistic and irreversible conformational change (maybe a loss of base pairing, or even a dissociation of the ligand from the DNA). This is probably in large part due to the absence of the frictional damping term that explicit water molecules provide. The second reason not to use an implicit solvent model for the MD simulation stage is that for a typical drug-DNA system, there is little if any gain in computational speed. Most MD packages have been extensively optimized for explicitly solvated simulations, and reliable implicitly solvated simulations typically require much longer nonbonded cutoffs (maybe 25 Å, vs. the 9–10 Å typically needed for an explicitly solvated simulation), with a corresponding increase in computational effort. 2. Ions. There has been considerable debate in the literature as to whether simulations of DNA and DNA-ligand systems are sensitive to questions of ionic strength. Though it may seem sensible to perform simulations of such charged systems under conditions that mimic physiological ionic strength, at present there is little definitive evidence that this produces better results than when simulations are performed using just enough positive counterions to compensate for the DNA negative charge. The second area that has received considerable attention is which counterion to use – e.g. sodium or potassium. This does seem to depend on the simulation package, and so forcefield, used, but only under conditions mimicking high ionic strength, when the ‘wrong’ combination of ion and

Molecular Modelling Methods to Quantitate Drug-DNA Interactions

129

solvent parameters in the simulation can lead to the formation of artifactual salt crystals (M. Orozco, personal communication) – so it is another reason to prefer minimal salt conditions. 3. Periodic boundary conditions. Since typical drug-DNA molecular systems are rather rod-like in shape, it may seem reasonable to use a rectangular periodic box for the simulations. There is a danger though that if over the course of the MD simulation the DNA system rotates in the box, it may reach the “edges” (Fig. 6a). Use of a cubic box will avoid this, but it is wasteful as water molecules in the corners will contribute little, but add to the computational expense (Fig. 6b). If the simulation package permits it, truncated octahedral boundary conditions immerse the DNA within an approximately spherical drop of water that avoids any artifactual behavior due to solute tumbling, but minimizes the number of water molecules used (Fig. 6c). 4. Assessing sampling from PC projections. While equilibration is fairly straightforward to detect by analysis of principal component (eigenvector) projections, assessing sampling is more difficult. It can always be argued that if the simulation was longer, other regions of conformational space would be sampled, and this is almost impossible to refute without actually extending the simulation. A realistic approach is to show that the simulations have reached a stage, where adding more snapshots does not significantly change the shapes of the histograms of the major projections. In a perfect, harmonic system, these histograms would be Gaussian (normal) distributions, and while this

A

B

C

Fig. 6. Use of conventional rectangular periodic solvent box may lead to problems if the DNA tumbles (a). Use of a cubic box avoids this, but at the expense of much more water to simulate (b). Truncated octahedral boundary conditions provide the best solution (c).

130

Wang and Laughton

should more or less be what is observed for the less important principal components, the major ones may show significantly non-normal distributions – if for example the simulation shuttles repeatedly between two local minima. However, the test of stability in the distributions remains valid. 5. Normal Mode versus Principal Component Analysis methods for configurational entropies. It should be borne in mind that while both NMA and PCA-based approaches have been used to estimate “configurational” entropy changes, they analyze the simulation data in rather different ways, and do not actually calculate the same term. This is perhaps most easily understood by considering the hypothetical energy surface shown in Fig. 7, and three configurations of the system on this, A–C. First, imagine that from a short simulation, we generate configurations A and B. Calculating the configurational entropy of this “ensemble” by the NMA method involves first a stringent energy minimization of each structure followed by the actual NMA procedure, which determines the curvature at the energy minimum. Clearly, in this case, structures A and B will relax to the same minimum, and so both will lead to the same estimate of the entropy. By the PCA method, the entropy of the two-configuration ensemble is estimated from the variance in the structures. Now, imagine that we extend the simulation, and as a result, gather structure C into our ensemble. NMA of this will involve first its relaxation into the alternative local minimum, then its entropy will be calculated, resulting in a value very similar to that obtained from consideration of A and B. By the PCA method, however, the total variance in the ensemble will now be much higher, and so will be the calculated entropy. Thus, NMA- and PCA-based methods may be expected to produce quite different predictions for “absolute” entropies; however, entropy differences, which are what really concern us, are not expected to be so sensitive to this issue.

Energy

A B

C

configuration Fig. 7. Difference between Normal Mode and Principal Component Analysis methods for configurational entropy estimation in the treatment of conformational variability. NMA will predict a very similar entropy for the ensembles (A + B) and (A + B + C), while PCA will predict that the entropy of (A + B) is considerably less than (A + B + C).

Molecular Modelling Methods to Quantitate Drug-DNA Interactions

131

References 1. Reddy BSP, Sharma SK, Lown JW (2001) Recent developments in sequence selective minor groove DNA effectors. Curr Med Chem 8:475–508 2. Goodsell DS (2001) Sequence recognition of DNA by lexitropsins. Curr Med Chem 8:509–516 3. Doss RM, Marques MA, Foister S, Chenoweth DM, Dervan PB (2006) Programmable oligomers for minor groove DNA recognition. J Am Chem Soc 128:9074–9079 4. Srinivasan J, Cheatham TE, Cieplak P, Kollman PA, Case DA (1998) Continuum solvent studies of the stability of DNA, RNA, and phosphoramidate – DNA helices. J Am Chem Soc 120:9401–9409 5. Kollman PA, Massova I, Reyes C, Kuhn B, Huo S, Chong L, Lee M, Lee T, Duan Y, Wang W, Donini O, Cieplak P, Srinivasan J, Case DA, Cheatham TE III (2000) Calculating structures and free energies of complex molecules: combining molecular mechanics and continuum models. Acc Chem Res 33:889–897 6. Cieplak P (2002) Application of the free energy calculations to study drug-enzyme and drug-DNA complexes. Mol Simul 28:173–186 7. Spackova N, Cheatham TE, Ryjacek F, Lankas F, van Meervelt L, Hobza P, Sponer J (2003) Molecular dynamics simulations and thermodynamics analysis of DNA-drug complexes. Minor groove binding between 4¢, 6-diamidino-2-phenylindole and DNA duplexes in solution. J Am Chem Soc 125:1759–1769 8. Anthony NG, Breerl D, Clarke J, Donoghue G, Drummond AJ, Ellis EM, Gemmell CG, Helesbeux JJ, Hunter IS, Khalaft AI, Mackay SP, Parkinson JA, Suckling CJ, Waigh RD (2007) Antimicrobial lexitropsins containing amide, amidine, and alkene linking groups. J Med Chem 50:6116–6125 9. Zhong H, Kirschner KN, Lee M, Bowen JP (2008) Binding free energy calculation for duocarmycin/DNA complex based on the QPLD-derived partial charge model. Bioorg Med Chem Lett 18:542–545 10. Hill TL (1986) An introduction to statistical thermodynamics. Dover, New York 11. Hagler AT, Stern PS, Sharon R, Becker JM, Naider F (1979) Computer simulation of the

12. 13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

conformational properties of oligopeptides. Comparison of theoretical methods and analysis of experimental results. J Am Chem Soc 101:6842–6852 McQuarrie DA (1976) Statistical mechanics. Harper and Row, New York Wang H, Laughton CA (2007) Molecular modelling methods for prediction of sequenceselectivity in DNA recognition. Methods 42:196–203 Schlitter J (1993) Estimation of absolute and relative entropies of macromolecules using the covariance matrix. Chem Phys Lett 215:617–621 Laughton C, Luisi B (1999) The mechanics of minor groove width variation in DNA, and its implications for the accommodation of ligands. J Mol Biol 288:953–963 Wellenzohn B, Flader W, Winger RH, Hallbrucker A, Mayer E, Liedl KR (2001) Significance of ligand tails for interaction with the minor groove of B-DNA. Biophys J 81:1588–1599 Case DA, Cheatham TE, Darden T, Gohlke H, Luo R, Merz KM, Onufriev A, Simmerling C, Wang B, Woods RJ (2005) The AMBER biomolecular simulation programs. J Comp Chem 26:1668–1688 Brooks BR, Bruccoleri RE, Olafson BD, States DJ, Swaminathan S, Karplus M (1983) CHARMM: A program for macromolecular energy, minimization, and dynamics calculations. J Comp Chem 4:187–217 Phillips JC, Braun R, Wang W, Gumbart J, Tajkhorshid E, Villa E, Chipot C, Skeel RD, Kalé L, Schulten K (2005) Scalable molecular dynamics with NAMD. J Comp Chem 26:1781–1802 Sherer EC, Harris SA, Soliva R, Orozco M, Laughton CA (1999) Molecular Dynamics Studies of DNA A-tract Structure and Flexibility. J Am Chem Soc 121:5981–5991 Berendsen HJ, Postma JPM, van Gunsteren WF, Hermans J (1969) Interaction models for water in relation to protein hydration. Nature 224:175–177 Sands ZA, Laughton CA (2004) Molecular dynamics simulations of DNA using the Generalized Born solvation model: quantitative comparisons with explicit solvation results. J Phys Chem B 108:10113–10119

Chapter 9 Application of Anomalous Diffraction Methods to the Study of DNA and DNA-Complexes Derrick Watkins, Tinoush Moulaei, Seiji Komeda, and Loren Dean Williams Abstract Anomalous scattering is commonly used to solve X-ray structures. As discussed here, anomalous scattering is also useful for characterizing complex systems with mixed and partial occupancies, where true electron density is represented by unresolvable ensemble averages. The solvent environment surrounding nucleic acids is an example of such a system, as are some DNA–ligand systems. The atomic number and wavelength dependencies of anomalous scattering allow one to filter out the electron densities of C, N, and O, and to cleanly visualize the electron densities of heavier atoms. Therefore, anomalous scattering can make beacons of selected atoms. In addition, anomalous scattering provides a model-independent method for determining atomic identities. Here, we describe applications of anomalous scattering to the structure determination of DNA–platinum complexes and in cation associations of free DNA, of DNA– anthracycline complexes, of chemically modified DNA, and of DNA–protein complexes. The utility of Rb+ and Tl+ as K+ substitutes is supported by similarities in Rb+ and Tl+ association with DNA. Key words: X-ray diffraction, Anomalous diffraction, Platinum–DNA complex, Anthracycline

1. Introduction to Anomalous Diffraction

“Anomalous scattering” is the unfortunate expression used to describe absorption-related effects on the scattering of X-rays by electrons. Anomalous scattering depends on X-ray energy, atomic number, and to lesser extent on chemical environment. Anomalous and normal X-ray scattering are both “elastic,” meaning that the input and scattered photons are of the same energy. Anomalous scattering can be conceptualized by imagining that some atoms absorb and immediately reemit photons. This absorption and reemission causes anomalously scattered photons to be retarded in phase, in comparison to normally scattered photons.

K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_9, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

133

134

Watkins et al. X-ray wavelength (Å) 15.0e 10.0e

1.65

CuK

1.37

1.18

1.03

0.91

0.82

0.75

f"

5.0e 0.0e 5.0e 10.0e

f'

Pt Tl Rb C

15.0e 20.0e

8000

10000

12000

14000

16000

X-ray energy (eV)

Fig. 1. Anomalous scattering; Plots of f¢ and f″ versus energy for Platinum, Thallium, Rubidium and Carbon.

The shifting of the phase of an anomalously scattered photon confers an imaginary component (f″) relative to normally scattered photons, causing it to rotate on the Argand plot. Any photon can be thought of as a vector with real and imaginary components and can be represented on an Argand diagram, with real and imaginary axes. The f″ scattering coefficient of an anomalously scattered photon is, by definition, nonzero (Fig. 1). A nonzero f″ breaks the equivalence between Bijvoet pairs, F+ and F−. A Patterson map calculated with (|F+| − |F−|)2 Fourier coefficients contains only peaks corresponding to vectors between pairs of anomalously scattering atoms. A correctly phased map with (|F+| − |F−|) Fourier coefficients contains only peaks corresponding to positions of anomalous scattering atoms. Anomalous scattering is used to solve X-ray structures (1, 2) but, as discussed here, is also useful for making beacons of selected atoms. The atomic number and wavelength dependencies of anomalous scattering allow one to filter out the electron density of C, N, and O, and to cleanly visualize the electron densities of heavier atoms. Anomalous scattering can be particularly useful for characterizing complex systems with mixed and partial occupancies, where true electron density is represented by unresolvable ensemble averages. In addition, anomalous scattering provides a modelindependent method for determining atomic identities. Here, we describe applications of anomalous scattering to the structure determination of DNA–platinum complexes and in cation associations of free DNA, of DNA–anthracycline complexes, of chemically modified DNA, and of DNA–protein complexes.

Application of Anomalous Diffraction Methods to the Study of DNA and DNA-Complexes

135

1.1. Solvent Characterization

A structural and thermodynamic understanding of DNA deformation and ligand binding, and RNA folding, requires accurate characterization of the solvent environment. In X-ray structures of polyanions, such as DNA and RNA, the characterization requires unambiguous determination of ordered solvent sites as water molecules, cations, or hybrids (with partial occupancy by both). Until around 1998, conventional X-ray methods had been used to support models in which nucleic acids in crystals are surrounded predominantly by neutral aqueous solvent, with fixed positions. These ordered water molecules were thought to share space with disordered counterions, which were crystallographically invisible, and unimportant for understanding structure and energetics. This view began to change when it was proposed that many of the solvent sites around DNA that were traditionally thought to be water molecules are, in fact, water–monovalent cation hybrids (3–6). Although this proposal was initially contested (7, 8), ­anomalous scattering experiments (9–12) have now demonstrated conclusively that water-only interpretations of solvent peaks surrounding nucleic acids are inaccurate, and that conventional crystallographic methods cannot generally distinguish between water molecules and water–cation hybrids, even at very high resolution. In our view, a next step should be accurate recharacterization and reinterpretation of the solvent environment surrounding DNA– ligand and DNA–protein complexes – in part by using the anomalous scattering methods described here.

1.2. Anomalous Scattering by Platinum

Sustained efforts by Lippard and others have resulted in the determination of three-dimensional structures of complexes of cis[PtX2(amine)2] compounds with DNA (13–17). These mononuclear Pt(II) compounds bind to and distort DNA as their primary mechanism of cytotoxicity (18, 19). The mode of DNA deformation is conserved throughout the cis-[PtX2(amine)2] family. Pt(II) gives a strong anomalous signal, which has been commonly used to help solve protein (20–22) and DNA (23) structures. Using the anomalous signal, one can readily identify and locate partially occupied Pt(II) atoms and estimate occupancies (detection limit is 5–10% occupancy for a well-ordered site). The anomalous signal (f″) of Pt(II) reaches a theoretical maximum at a wavelength of 1.07 Å (11.6 keV, Fig. 1), although the signal is significant and readily usable at 1.54 Å (CuKa). In practice, for a given experiment, the wavelength of the maximum in f″ should be experimentally determined by fluorescence.

1.3. Anomalous Scattering by Thallium

The anomalous signal of Tl+ obviates interpretation of subtle differences in coordination geometry and scattering power of K+ and Na+ versus H2O. The f″ for Tl+ reaches a theoretical maximum of 0.978 Å (12.7 keV, Fig. 1), although, as for Pt(II), the anomalous

136

Watkins et al.

signal of Tl+ is strong and usable at 1.54 Å. See Note 1 for information on thallium toxicity. 1.3.1. Proteins

Tl+ was initially investigated as a K+ mimic in biological systems by R.J.P. Williams (24, 25) and others (26–31). Tl+ has been used as a K+ substitute in the catalytic mechanisms of sodium–potassium pumps (32), fructose-1-6-bisphosphatase (33), and pyruvate kinase (34). Tl+ was used by Caspar and coworkers to determine counterion positions adjacent to insulin (35, 36), and by Gill and Eisenberg to characterize the positions of ammonium ions in the binding pocket of glutamine synthetase (37).

1.3.2. Nucleic Acids

We have used Tl+ as a probe for K+ in association with B-DNA (9, 11), with DNA–drug complexes (11), with DNA adducts (9) and with DNA–protein complexes (unpublished). Doudna and coworkers used Tl+ as a probe for K+ in the tetrahymena ribozyme P4-P6 domain (38). Draper and coworkers used Tl+ as a K+ probe in the structure of a fragment of the 23S rRNA (39). Correll used Tl+ as a K+ probe in the structure of the sarcin/ricin loop of the 23S rRNA (40). Tl+ has been shown to stabilize guanine quadruplexes in a manner analogous to K+ (41–43). See Note 2 for information on crystallization procedures using Tl+.

1.3.3. Thallium Specific Coordination Chemistry

Although Tl+ is a useful K+ mimic, some import differences in coordination chemistry distinguish Tl+ from group I cations. In crystals of inorganic salts, Tl+ frequently displays a low coordination number with one hemisphere of its coordination sphere unoccupied (44–46). In contrast, group I cations are spherically symmetric and tend to maximize their coordination number. Although Tl+ and group I cations both form Tl–O and Tl–N “bonds,” Tl+ also participates in Tl–p bonds, and long distance Tl–Tl “thallophilic” bonds. Thallium’s nonspherical outer shell and asymmetric coordination characteristics stem from an inert, partially hybridized 6s2 orbital, which is contracted due to strong interaction with the nucleus.

1.4. Anomalous Scattering by Rubidium

Although the anomalous signal of Rb+ is significantly weaker and less accessible than that of Tl+ (Fig. 1), it has been used with success, for example, by Egli (12, 47) and Steitz (48). In addition, the anomalous signal of Rb+ was used to identify K+ positions in a structure of the sodium–potassium pump (49). The theoretical absorption edge for Rb+, where f″ reaches a maximum, is at a wavelength of 0.815 Å (15.2 keV, Fig. 1), where the high energy of the X-rays can increase the rate of crystal decay. Rb+ and Tl+ associate with DNA in similar manners, supporting the utility of both as K+ mimics. See Note 2 for additional information on crystallization with Rb+.

Application of Anomalous Diffraction Methods to the Study of DNA and DNA-Complexes

137

2. Materials Synchrotron radiation is critical for most anomalous scattering experiments. The continuously tunable X-rays at modern synchrotrons allow selection of X-ray energies on, above and below the absorption edges of heavy atoms. However, for some heavy atoms, CuKa radiation can yield a strong signal.

3. Methods 3.1. Platinum–DNA Complexes

Work of Lippard (13–16, 50) and others has yielded deep insights into DNA binding and mechanism of Pt(II) complexes. The effort involved laborious purification protocols, recalcitrant crystal growth, and sometimes poor data quality. Only seven Pt(II)– DNA complexes, under a narrow range of DNA sequences and ligand-type, are contained in the structural databases. With the exception of a Z-DNA complex, all of the DNA–platinum structures in the database belong to the cis-[Pt(II)X2(amine)2] family.

3.1.1. Growth of DNA–Pt(II) Crystals

A “high-throughput” method of crystallization devised by Komeda (unpublished) has yielded numerous crystals of DNA–Pt(II) complexes both covalent and noncovalent. The method involves mixing double-stranded DNA fragments and Pt(II) compounds, and proceeding directly with crystallization, without purifying or attempting to minimize the number of species in solution. This method is fast, with low labor input, uses double-stranded DNA reactants, allows control of reaction extent, and is very efficient in yielding crystals. The method is much simpler and more direct than the traditional method of platination of DNA, followed by HPLC separation of products, and hybridizing purified-platinated DNA with a complementary strand. The high-throughput method commonly yields crystals with local disorder, with partial ligand occupancies and multiple reaction products/intermediates. The DNA is highly ordered, giving diffraction to high angle (high resolution).

3.1.2. Anomalous Scattering by Pt(II)

The anomalous signal of Pt(II) allows one to use X-ray crystallography, which generally gives best results for highly ordered systems, for characterizing and analyzing partially disordered systems. The strength of the Pt(II) anomalous signal allows for unambiguous assignment of various Pt(II) species, even at partial occupancy and with mixtures of intermediates/products. In part with the anomalous signal of Pt(II), we discovered a new mode of noncovalent DNA binding (51) with the potential to substantially increase the functionality and utility of small-molecule

138

Watkins et al.

Fig. 2. Backbone tracking by a trinuclear Pt(II) compound. Hydrogen bonds between TriplatinNC and DNA are indicated by dashed lines. (a) View perpendicular to the helical axis. (b) View along the helical axis. The phosphodiester backbone of the DNA is represented by a tube. The DNA phosphate groups are shown in ball and stick. TriplatinNC is represented by stick, except for the platinum and nitrogen atoms, which are ball and stick. (c) An anomalous map contoured at 5s (net). The locations of the Pt(II) atoms are indicated by the anomalous peaks.

DNA-binding agents. We used Komeda’s high-throughput method to obtain crystals of the trinuclear Pt(II) complex TriplatinNC (51, 52) in a complex with [d(CGCGAATTGCGC)]2 (Fig.  2). Diffraction data were collected on beamline X26C at the National Synchrotron Light Source, Brookhaven National Laboratory. 360° in phi of intensity data (oscillation angle of 1.0°) were collected at 175  K, at a wavelength of 1.05 Å. Data were recorded on an ADSC-Quantum 4 CCD and were merged and reduced with HKL2000 (53). The Bijvoet pairs were left unmerged in one dataset, used for calculation of |F+| − |F−| Fouriers, and combined in the other dataset, used for refinement of the final model.

Application of Anomalous Diffraction Methods to the Study of DNA and DNA-Complexes

139

After structure solution by molecular replacement, strong 2|Fo| − |Fc| and |Fo| − |Fc| Fourier peaks indicated positions of four Pt(II) atoms. The positions of the four Pt(II) ions were confirmed by |F+| − |F−| Fouriers (Fig. 2c), which were computed using phases by the most ordered subset of the Pt(II) atoms. See Note 3 for tips on synthesis and validation of |F+| − |F−| maps. Occupancies were optimized throughout the refinement by fitting statistics and by examination of +/− (|Fo| − |Fc|) Fourier maps. 3.1.3. A New DNA Binding Mode

The TriplatinNC complex reveals a new DNA binding mode that we call the Phosphate Clamp. In a phosphate clamp, two amino protons, from adjacent amino groups of Pt(II) tetra-am(m)ine form hydrogen bonds with a nonbridging phosphate oxygen of DNA. Phosphate Clamps are highly selective for phosphate oxygens over other functional groups of DNA. Appropriately linked Phosphate Clamps associate with DNA by a binding mode that we call Backbone Tracking (Fig. 2a, b). In this binding mode, the elongated ligand follows the path of the phosphodiester backbone.

3.2. Thallium in B-DNA Crystals

The anomalous signal of Tl+ was used to demonstrate that monovalent cations can localize around B-DNA (Fig. 3) at sequencespecific sites (11). Because some members of the nucleic acid structure community were convinced that monovalent cations could not dehydrate and associate at defined positions in direct contact with DNA bases, a series of rigorous and independent crystallographic methods were used to determine the Tl+ positions. Tl+ positions in [d(CGCGAATTCGCG)]2 crystals were determined by anomalous methods, isomorphous difference methods and anomalous/isomorphous Patterson methods.

3.2.1. |F+| − |F−| Difference Fouriers

For calculation of anomalous difference |F+| − |F−| Fouriers, a data set was collected from a single crystal, over 360° in phi at 90 K using 1.54 Å Cu Ka radiation from an in-house Rigaku/MSC rotating anode generator with Osmic blue confocal mirrors and an R-AXIS4++ image plate detector. 131,329 reflections to 1.55 Å resolution were reduced to a unique dataset containing 10,320 reflections, preserving Bijvoet pairs, using the dtprocess software in the program CrystalClear 1.3.0 (Rmerge = 0.048, 99.4% complete, DF/F = 0.060). Phase bias can be introduced into |F+| − |F−| maps in the same way that bias can be introduced into 2Fo − Fc maps. To avoid bias, |F+| − |F−| Fouriers were calculated using phases from a [d(CGCGAATTGCGC)]2/K+ structure (below). The map (Fig. 3) was calculated using data to 2.0 Å resolution. The resulting map showed peaks for Tl+ ions in the solvent regions and weaker peaks from the anomalous scattering of phosphorous atoms of the DNA (see Note 3).

140

Watkins et al.

Fig. 3. Tl+ ions are observed to localize in the grooves of [d(CGCGAATTCGCG)]2 by three independent crystallographic methods. Tl+ positions are indicated by crossed circles. Tl+ positions determined by isomorphous |FTl| − |FK| Fouriers are indicated by the dashed net. Tl+ positions determined by anomalous |F+| − |F−| Fouriers are indicated by the solid net. The DNA is represented in stick. The figures were created with Swiss-PDB Viewer (79) and rendered with Povray 3.1.

3.2.2. Isomorphous |FTl| − |FK| Difference Fouriers

To calculate isomorphous |FTl| − |FK| difference Fouriers (where |FTl| represents structure factor amplitudes from a crystal grown in the presence of Tl+, while |FK| represents structure factor amplitudes from a crystal grown in the presence of K+), high ­resolution [d(CGCGAATTCGCG)]2/K+ (3) and [d(CGCGAATTCGCG)]2/ Tl+ (11) data sets were collected at Brookhaven National Laboratory on beamline X26C with an ADSC Quantum 4 CCD detector using X-rays of 1.1 Å wavelength. Crystals were grown from solutions without competing monovalent cations, and were maintained at 100  K during data collection. For the DNA–Tl+ data, a total of 290,894 reflections were reduced to 21,760 unique reflections (resolution = 1.2 Å). For the DNA-K+ data, a total of 489,992 reflections were reduced to 19,244 unique

Application of Anomalous Diffraction Methods to the Study of DNA and DNA-Complexes

141

reflections (resolution = 1.2 Å). Data were indexed and integrated using MOSFLM 6.0 (54), reduced with SCALA (55). The two datasets were anisotropically scaled by shell, from 21 to 1.6 Å using the Xmerge routine of XtalView (56). Isomorphous difference Fouriers were calculated using the scaled |FTl| − |FK| as coefficients and phases from [d(CGCG AATTCGCG)2]/K+ model. The assumption of isomorphism of the DNA in the [d(CGCGAATTCGCG)2]/Tl+ and [d(CGCG AATTCGCG)2]/K+ crystals was based on (1) RMS deviations between the final CGCGAATTCGCG/Tl+ and CGCGA ATTCGCG/K+ models (all DNA atoms) of 0.32 Å, (2) differences in unit cell parameters of Da = 1.3%, Db = 0.2%, Dc = 0.7%, and (3) the observed Riso (S|FTl − FK|/|FK|) for common reflections between 21 and 1.6 Å resolution is 0.29 (average over all shells), which is similar to the predicted Riso calculated with the method of Crick and Magdoff (57). 3.2.3. (|F+| − |F−|)2 and (|FTl| − |FK|)2 Difference Patterson Fouriers

Difference Pattersons were calculated using (|F+| − |F−|)2 or (|FTl| − |FK|)2 as coefficients. The two methods gave similar results. A search of (|F+| − |F−|)2 Patterson space using the program CNS (58) with data to 2 Å resolution yielded three Tl+ sites, which, after enantiameric correction, corresponded to the three most highly occupied Tl+ sites in the refined model. An |F+| − |F−| Fourier was calculated using phases from the three sites identified in the (|F+| − |F−|)2 Patterson search. Three additional Tl+ sites corresponding to those in the refined model were clearly indicated in |F+| − |F−| Fourier. By Patterson methods, a total of six Tl+ sites were identified from the |F+| − |F−| data, without additional phase information.

3.2.4. Internal Consistency by Many Crystallographic Methods

Tl+ ions were identified by 2|Fo| − |Fc| and |Fo| − |Fc| Fouriers, by |F+| − |F−| and |FTl| − |FK| Fouriers, and by (|F+| − |F−|)2 and (|FTl| − |FK|)2 Patterson Fouriers. The methods give consistent results. Eleven of the thirteen Tl+ sites determined during the refinement are confirmed by both the |F+| − |F−| and |FTl| − |FK| Fouriers. One Tl+ site in the minor groove at the ApT step is observed in the |F+| − |F−| Fourier but not the |FTl| − |FK| Fourier. One site is based only in the |Fo| − |Fc| Fourier and is not observed in either the |F+| − |F−| or the |FTl| − |FK| Fourier, and so is considered tentative.

3.2.5. Cations in the Grooves

The most highly occupied Tl+ sites, located within the G-tract major groove, have estimated occupancies ranging from 20% to 35%. The occupancies of the minor groove sites are estimated to be around 10%. The Tl+ positions are, in general, not in direct proximity to phosphate groups. The A-tract major groove appears devoid of localized cations. The locations of the Tl+ ions are dictated by coordination geometry, electronegative potential,

142

Watkins et al.

avoidance of electropositive amino groups, and in several cases, cation–p interactions. 3.3. Thallium in DNA–Anthracycline Crystals

Anthracyclines bind preferentially at pyrimidine–purine steps (59) and interfere with the action of topoisomerase II (60). In X-ray structures of several DNA–anthracycline complexes, sodium ions mediate interactions between the oxygen atoms of the inter­ calated chromophore and the N7 atom of an adjacent guanosine (61, 62). These monovalent cations appear to influence stability, conformation and sequence-specificity. However, some X-ray structures of DNA–anthracycline complexes lack localized cations (63, 64). We have used Tl+ substitution and anomalous methods to resolve the discrepancies, and more generally, to determine roles of localized cations and electrostatic forces in structure, thermodynamics, and sequence-specificity of DNA–ligand ­complexes (10).

3.3.1. DNA–Anthracycline Binding Energetics

The DNA-binding energetics of daunomycin were experimentally decomposed by Chaires and coworkers into functional-group specific contributions (65). The daunosamine contributes a large and favorable binding free energy (2  kcal/mol). The 3¢-amino group contributes 0.7 kcal/mol above the polyelectrolyte contribution of the positive charge. The hydroxyl groups at the 9 and 14 positions contribute approximately 1 kcal/mol. Chaires also identified water as an important thermodynamic participant, observing uptake of water during binding (66). The thermodynamic results are consistent with three-dimensional structure determinations of a series of DNA–anthracycline complexes (10, 61–63, 67–71). The three-dimensional structures demonstrate that the daunosamine is in intimate contact with the DNA; hydrogen bonds and van der Waals contacts secure the daunosamine to the floor of the minor groove. The 9-hydroxy group forms multiple hydrogen bonds with a flanking guanine. A large number of solvent molecules are conserved in crystal structures of DNA/ anthracycline complexes.

3.3.2. Locating Tl+ Ions

A 1.34 Å resolution structure of a [d(CGATCG)]2-adriamycin2 complex was obtained from crystals grown in the presence of Tl+ (Fig. 4). For this work, data were collected with 1.54 Å CuKa radiation on the in-house rotating anode generator and detector described for Pt(II), above. Data were collected over 360° in phi at 92 K. 31,250 reflections to 1.26 Å resolution were reduced to a unique dataset containing 5,559 reflections, preserving Bijvoet pairs using the dtprocess software in the program CrystalClear 1.3.0 (Rmerge = 4.2%, 90% complete, DF/F = 0.083). Data used for refinement included 5,034 reflections from 35 to 1.34 Å. The structure was solved by molecular replacement. A subset of water molecules were converted to Tl+ ions during the refinement.

Application of Anomalous Diffraction Methods to the Study of DNA and DNA-Complexes

143

Fig. 4. Intercalated adriamycin and an adjacent guanosine associates with Tl+ ions. (a) A Residual |Fo| − |Fc| Fourier difference electron density map. This 5 s map was calculated with phases that included contributions from the water molecule (sphere) within the difference density. That water molecule was subsequently converted to a Tl+ ion. (b) An anomalous density map. The map is contoured at 3.5 s. Tl+ ions are depicted as small spheres. These figures were generated with SwissPDB Viewer and rendered with Povray 3.1. DNA and adriamycin atoms are shown in stick representation.

The conversions were based on |F+| − |F−| and |Fo| − |Fc| Fourier maps. Although these two methods do not bias each other, they give consistent results. |F+| − |F−| maps were calculated using data to 1.4 Å and phases from [d(CGATCG)]2-adriamycin2 alone (minus water and ions). The most intense peak (30 s) in the |F+| − |F−| map is 0.05 Å from the most highly occupied Tl+ site independently determined from |Fo| − |Fc| map. A shoulder (6 s) on the 30 s peak is nearly superimposed on another Tl+ site determined from |Fo| − |Fc| maps. An additional |F+| − |F−| peak (4.3 s) is 0.18 Å from the third Tl+ site from |Fo| − |Fc| maps. Other peaks observed the |F+| − |F−| map were not assigned as Tl+ ions because of a lack of evidence for Tl+ at these sites in |Fo| − |Fc| maps. Each of the three Tl+ sites is partially occupied in the final model.

144

Watkins et al.

Tl+ occupancies were estimated during the refinement by manually minimizing the magnitudes of the |Fo| − |Fc| and |Fc| − |Fo| peaks. 3.3.3. General Principles of DNA–Ligand–Cation Interactions

Consistent with previous observations with free DNA (5, 11), localized cations in the [d(CGATCG)]2–adriamycin2–Tl+ complex do not interact directly with phosphate groups or exocyclic amino groups of DNA bases. The results indicate cations associated with DNA can interact simultaneously, and with apparent cooperativity, with both ligands and DNA. The results of a survey of adriamycin-DNA structures, coupled with the Tl+ substitution experiments, are summarized in Fig. 5.

3.4. Thallium in Covalently Modified B-DNA Crystals

A 1.6 Å resolution X-ray structure of [d(CGCGAATTGCGC)]2 was obtained that has been covalently modified by the tethering of four cationic charges (Fig.  6). This modified dodecamer is composed of [d(CGCGAAXXCGCG)]2, where X is a thymine residue linked at the 5-position to an n-propyl-amine. To obtain diffraction quality crystals of this modified dodecamer, it was

3.4.1. Tethering Cations to DNA

(C,T) O2 OH

O 12

1

O

11

10

C 13 14 CH2 OH

9

2

D

C

B

A

3

OH 8

5

4

MeO Tl+(2)

7

6

O

OH Tl+(1)

O 5'O

Me 3'

G N7 G O6

NH3 OH 4'

(T) O2 (A) N3 Fig. 5. Consensus water and ion sites around DNA–adriamycin complexes. The circles and pie-shaped elements represent water molecules. A shaded full circle indicates that a water molecule interacts with the associated oxygen or nitrogen in 100% of DNA–adriamycin structures survey in the database. The pie-shaped elements indicate quantitatively the fraction of surveyed structures in which a water molecule interacts with the oxygen or nitrogen atom. The circles that contain plus signs indicate Tl+ sites. The nonshaded cation is nearly identical to the location of sodium ions in previous structures.

Application of Anomalous Diffraction Methods to the Study of DNA and DNA-Complexes

145

Fig. 6. Anomalous density contoured at 2.5 s surrounding [d(CGCGAATTCGCG)]2 that is modified by tethered cations. The crystals were soaked with Tl+ ions. The density surrounding the phosphorous atoms is drawn in green net, while that surrounding the Tl+ ions is blue net. The map was calculated using only the positions and occupancies of the Tl+ ions. The observed peaks around the phosphorous atoms serve as a positive control.

146

Watkins et al.

necessary to incorporate 20% v/v ANAPOE®-C12E10 (Anatrace), a nonionic detergent, into the crystallization droplet. The structure was determined from crystals soaked with Tl+, because cocrystallization experiments were unsuccessful. Three of the tethered cations are directed radially out from the DNA. The radially directed tethered cations do not appear to induce structural changes or to displace counterions. One of the tethered cations is directed in the 3¢ direction, toward a phosphate group near one end of the duplex. This tethered cation appears to interact electrostatically with the DNA, displacing counterions from the major groove. 3.4.2. Locating Tl+ Ions

The initial positions of Tl+ ions were determined from an |F+| − |F−| Fourier phased using either all atoms of the refined model, or the backbone phosphorous atoms of the refined model. Maps calculated using only the phosphorous atoms for phasing were much cleaner and easier to interpret. Other peaks in the anomalous map, which corresponded to the position of water molecules in the initial refinement were selected as candidate sites for Tl+ ions. The appropriate water molecules were converted to Tl+ ions, and their occupancies were manually refined by monitoring the refinement statistics and +/− (|F+| − |F−|) Fouriers. This process was repeated with the initial Tl+ ions used in phasing the anomalous map. Additional Tl+ sites were identified and refined.

3.5. Thallium and Rubidium in DNA– Protein Crystals

The P22 c2 repressor protein is a member of the lambdoid family of repressors (72). This family of repressors helps determine if a phage will enter a lytic or lysogenic phase of development (73). As with other members of the lambdoid family, P22 c2 repressor binds as a dimer to one of six DNA operators that are separated into two regions of a phage plasmid, designated OR and OL. Each region contains three different operator sites, OR1, OR2, OR3, for the OR region and OL1, OL2, OL3 for the OL region. The ability of the repressor to discriminate between the three operators of each region is the key to its ability to regulate phage development (74). The discrimination of the different operators by the P22 c2 repressor is partially dependent on the identification of the bases at the center of the operator that are not contacted by the repressor. These noncontacted bases influence affinity, a mode of recognition termed “indirect readout.” The central bases cause sensitivity of binding affinity on salt concentration. Taken together, the results suggest the involvement of monovalent cations in binding. Crystals of P22 c2 repressor N-terminal domain (P22R NTD), with DNA operators grown in the presence of Na+, Tl+ or Rb+, allow the mapping of monovalent cations in these complexes.

3.5.1. The P22 c2 Repressor Protein

3.5.2. P22R NTD-DNA/Na+, Rb+, Tl+

For crystals of P22R NTD-DNA grown in the presence of Tl+, anomalous difference maps contain clear peaks indicating sites of localized Tl+ ions (Fig. 7). These sites overlap solvent peaks of the

Application of Anomalous Diffraction Methods to the Study of DNA and DNA-Complexes

147

Fig. 7. Anomalous peaks calculated from a |F+| − |F−| Fourier map of P22NTD-DNA/Tl+, contoured at 2.5 s. Density (mesh) is observed corresponding to a Tl+ (sphere) in the major groove of DNA (stick). Phases for the anomalous map were calculated from P22NTD-DNA without solvent or Tl+ ions.

2|Fo| − |Fc| Fourier maps, indicating, as expected, that some water molecules in conventional structures (lacking anomalous scatters) are, in reality, partially occupied by cations. Data for the P22R NTD-DNA/Tl+ derivative were collected at wavelength of 0.97784 Å on SER-CAT beamline 22BM at the Advanced Photon Source, Argonne National Laboratory with a MAR225 CCD detector. Reflections were merged and reduced with HKL2000 (53). Initial phases were determined by molecular replacement. Models were refined with program Refmac5 (75, 76). |F+| − |F−| Fourier maps were calculated using CCP4 (55). Phases for the anomalous maps were calculated with solvent molecules removed, using the program Sfall (77). The locations of Tl+ ions determined by Fo − Fc were consistent with those determined by |F+| − |F−| and isomorphous |FTl| − |F|Na| Fouriers.

148

Watkins et al.

The utility of Rb+ and Tl+ as K+ substitutes is supported by similarities in Rb+ and Tl+ binding sites. Crystals of P22R NTDDNA grown in the presence of Rb+ yield anomalous difference peaks that are nearly superimposable on the Tl+ anomalous peaks. However, the weaker anomalous signal of Rb+ compared to Tl+ is clearly indicated by the relative intensities of the peaks. 3.5.3. Principles of DNA–Protein–Cation Interactions

Traditional (nonanomalous) crystallographic methods appear inadequate for accurately characterizing the solvent environment surrounding DNA/protein complexes. The use of anomalous scatters now allows one to identify cations sites in a model independent process. Here, the results indicate that contacts of the P22 c2 repressor with DNA are dependent on the conformation of DNA, and that the conformation of DNA is influenced by the localization of monovalent cations.

4. Notes 1. Thallium is extremely toxic in part due to its close mimicry of K+. Tl+ readily enters cells via potassium uptake pathways. However, the affinity of Tl+ for sulfur ligands, due the accessibility of the empty d-orbitals of Tl+, disrupts a host of cellular processes. Tl+ must be handled with appropriate caution and disposed of properly. 2. Rb+ has the advantage of being soluble as a chloride salt, whereas TlCl is insoluble. RbCl can often replace KCl or NaCl with little or no requirement for reoptimization of crystallization conditions. The anomalous experiments described here require clean conditions, with just one type of monovalent cation in the crystallization solution. The presence of either Na+ or K+ in the crystallization solution results in reduction or loss of the anomalous signal of Rb+ or Tl+. However, due to the high toxicity of Tl+ (see Note 1), it is recommended that the replacement of Na+ or K+ with Tl+ in the protein buffer be carried out via buffer exchange after all other steps of protein purification are complete. Crystallization in the presence of Tl+ generally requires reoptimization of conditions. This requirement is related to the insolubility of TlCl. Therefore, all Cl− must be omitted from all buffers and crystallization solutions. For the P22R NTD-DNA/Tl+ complex, Tl+ acetate was used to replace NaCl in both the protein buffer and in the crystallization solutions. pH of buffers was adjusted with acetic acid as opposed HCl. Salts such MgCl2 or LiCl were replaced with acetate derivatives. It should be noted that most commercial screens contain at least some Cl−, in that the pH of most of the buffers is adjusted with HCl. It is recommended

Application of Anomalous Diffraction Methods to the Study of DNA and DNA-Complexes

149

here that if initial crystallization conditions are not known, that they be determined with NaCl, KCl, or commercial screens, followed by buffer exchange and crystallization reoptimization with Tl+. 3. The correctness of |F+| − |F−| map synthesis can be indicated by the observation of peaks on the phosphorus atoms of DNA, see reference (78). The weak anomalous signal of phosphorus serves as an internal control, indicating the experiment has been done correctly. For anomalous map calculation, pairs of reflections with |F|/Sigma(F) < 1.0 for either Bijvoet pair are excluded. References 1. Ealick SE (2000) Advances in multiple wavelength anomalous diffraction crystallography. Curr Opin Struct Biol 4:495–499 2. Hendrickson WA (1991) Determination of macromolecular structures from anomalous diffraction of synchrotron radiation. Science 254:51–58 3. Sines CC, McFail-Isom L, Howerton SB, VanDerveer D, Williams LD (2000) Cations mediate B-DNA conformational heterogeneity. J Am Chem Soc 122:11048–11056 4. McFail-Isom L, Sines CC, Williams LD (1999) DNA structure: cations in charge? Curr Opin Struct Biol 9:298–304 5. Shui X, Sines C, McFail-Isom L, VanDerveer D, Williams LD (1998) Structure of the potassium form of CGCGAATTCGCG: DNA deformation by electrostatic collapse around inorganic cations. Biochemistry 37:16877–16887 6. Shui X, McFail-Isom L, Hu GG, Williams LD (1998) The B-DNA dodecamer at high resolution reveals a spine of water on sodium. Biochemistry 37:8341–8355 7. Chiu TK, Kaczor-Grzeskowiak M, Dickerson RE (1999) Absence of minor groove monovalent cations in the crosslinked dodecamer CGCGAATTCGCG. J Mol Biol 292:589–608 8. Chiu TK, Dickerson RE (2000) 1 Å Crystal structures of B-DNA reveal sequence-specific binding and groove-specific bending of DNA by magnesium and calcium. J Mol Biol 301:915–945 9. Moulaei T, Maehigashi T, Lountos GT, Komeda S, Watkins D, Stone MP, Marky LA, Li JS, Gold B, Williams LD (2005) Structure of B-DNA with cations tethered in the major groove. Biochemistry 44:7458–7468 10. Howerton SB, Nagpal A, Williams LD (2003) Surprising roles of electrostatic interactions in DNA-ligand complexes. Biopolymers 69:87–99

11. Howerton SB, Sines CC, VanDerveer D, Williams LD (2001) Locating monovalent cations in the grooves of B-DNA. Biochemistry 40:10023–10031 12. Tereshko V, Wilds CJ, Minasov G, Prakash TP, Maier MA, Howard A, Wawrzak Z, Manoharan M, Egli M (2001) Detection of alkali metal ions in DNA crystals using state-of-the-art X-ray diffraction experiments. Nucleic Acids Res 29:1208–1215 13. Takahara PM, Rosenzweig AC, Frederick CA, Lippard SJ (1995) Crystal structure of double-stranded DNA containing the major adduct of the anticancer drug cisplatin. Nature 377:649–652 14. Ohndorf UM, Rould MA, He Q, Pabo CO, Lippard SJ (1999) Basis for recognition of cisplatin-modified DNA by high-mobility-group proteins. Nature 399:708–712 15. Spingler B, Whittington DA, Lippard SJ (2001) 2.4 Å Crystal structure of an Oxaliplatin 1, 2-d(GpG) intrastrand cross-link in a DNA dodecamer duplex. Inorg Chem 40:5596–5602 16. Silverman AP, Bu W, Cohen SM, Lippard SJ (2002) 2.4-Å Crystal structure of the asymmetric platinum complex [Pt(ammine)(cyclohexylamine)]2+ bound to a dodecamer DNA duplex. J Biol Chem 277:49743–49749 17. Wang D, Lippard SJ (2005) Cellular processing of platinum anticancer drugs. Nat Rev Drug Discov 4:307–320 18. Jamieson ER, Lippard SJ (1999) Structure, recognition, and processing of cisplatin-DNA adducts. Chem Rev 99:2467–2498 19. Barnes KR, Lippard SJ (2004) Cisplatin and related anticancer drugs: recent advances and insights. Met Ions Biol Syst 42:143–177 20. Veerapandian B, Gilliland GL, Raag R, Svensson AL, Masui Y, Hirai Y, Poulos TL (1992) Functional implications of interleukin-1-beta

150

21.

22.

23.

24. 25.

26. 27.

28.

29. 30. 31.

32.

33.

34.

Watkins et al. based on the 3-dimensional structure. Proteins 12:10–23 Xia ZX, Dai WW, Xiong JP, Hao ZP, Davidson VL, White S, Mathews FS (1992) The threedimensional structures of methanol dehydrogenase from two methylotrophic bacteria at 2.6-A resolution. J Biol Chem 267:22289–22297 Quillin ML, Wingfield PT, Matthews BW (2006) Determination of solvent content in cavities in Il-1beta using experimentally phased electron density. Proc Natl Acad Sci U S A 103:19749–19753 Lipscomb LA, Zhou FX, Presnell SR, Woo RJ, Peek ME, Plaskon RR, Williams LD (1996) Structure of a DNA-porphyrin complex. Biochemistry 35:2818–2823 Williams RJP (1971) Biochemistry of Group Ia and IIa cations. Adv Chem Ser 100:155–173 Manners JP, Morallee KG, Williams RJP (1970) Thallium(I) as a potassium probe in biological systems. J Chem Soc Chem Commun 965–966 Kayne FJ (1971) Thallium (I) activation of pyruvate kinase. Arch Biochem Biophys 143:232 Post RL, Kume S, Tobin T, Orcutt B, Sen AK (1969) Flexibility of an active center in sodium-plus-potassium adenosine triphosphatase. J Gen Physiol 54:306–326 Inturrissi CE (1969) Thallium-induced dephosphorylation of a phosphorylated intermediate of (sodium + thallium-activated) ATPase. Biochim Biophys Acta 178:630 Inturrissi CE (1969) Thallium activation of K+-activated phosphatases from beef brain. Biochim Biophys Acta 173:567 Britten JS, Blank M (1968) Thallium activation of (Na+-K+)-activated ATPase of rabbit kidney. Biochim Biophys Acta 159:160–166 Reuben J, Kayne FJ (1971) Thallium-205 nuclear magnetic resonance study of pyruvate kinase and its substrates – evidence for a substrate-induced conformational change. J Biol Chem 246:6227–6234 Pedersen PA, Nielsen JM, Rasmussen JH, Jorgensen PL (1998) Contribution to Tl+, K+, and Na+ Binding of Asn776, Ser775, Thr774, Thr772, and Tyr771 in cytoplasmic part of fifth transmembrane segment in alpha-subunit of renal Na, K-ATPase. Biochemistry 37:17818–17827 Villeret V, Huang S, Fromm HJ, Lipscomb WN (1995) Crystallographic evidence for the action of potassium, thallium, and lithium ions on fructose-1, 6-bisphosphatase. Proc Natl Acad Sci U S A 92:8916–8920 Loria JP, Nowak T (1998) Conformational changes in yeast pyruvate kinase studied by 205 Tl+ NMR. Biochemistry 37:6967–6974

35. Badger J, Li Y, Caspar DL (1994) Thallium counterion distribution in cubic insulin crystals determined from anomalous X-Ray diffraction data. Proc Natl Acad Sci U S A 91:1224–1228 36. Badger J, Kapulsky A, Gursky O, Bhyravbhatla B, Caspar DL (1994) Structure and selectivity of a monovalent cation binding site in cubic insulin crystals. Biophys J 66:286–292 37. Gill HS, Eisenberg D (2001) The crystal structure of phosphinothricin in the active site of glutamine synthetase illuminates the mechanism of enzymatic inhibition. Biochemistry 40:1903–1912 38. Basu S, Rambo RP, Strauss-Soukup J, Cate JH, Ferre-D’Amare AR, Strobel SA, Doudna JA (1998) A specific monovalent metal ion integral to the AA platform of the RNA tetraloop receptor. Nat Struct Biol 5:986–992 39. Conn GL, Gittis AG, Lattman EE, Misra VK, Draper DE (2002) A compact RNA tertiary structure contains a buried backbone-K+ complex. J Mol Biol 318:963–973 40. Correll CC, Swinger K (2003) Common and distinctive features of GNRA tetraloops based on a Guaa tetraloop structure at 1.4 Å resolution. RNA 9:355–363 41. Basu S, Szewczak AA, Cocco M, Strobel SA (2000) Direct detection of monovalent metal ion binding to a DNA G-quartet by 205Tl NMR. J Am Chem Soc 122:3240–3241 42. Gill ML, Strobel SA, Loria JP (2006) Crystallization and characterization of the thallium form of the oxytricha nova G-quadruplex. Nucleic Acids Res 34:4506–4514 43. Caceres C, Wright G, Gouyette C, Parkinson G, Subirana JA (2004) A thymine tetrad in d(TGGGGT) quadruplexes stabilized with Tl+/ Na+ ions. Nucleic Acids Res 32:1097–1102 44. Janiak C (1997) (Organo)Thallium (I) and (II) chemistry: syntheses, structures, properties and applications of subvalent thallium complexes with alkyl, cyclopentadienyl, arene or hydrotris(pyrazolyl)borate ligands. Coord Chem Rev 163:107–216 45. Kristiansson O (2002) Structures of complexes of Thallium(I) and functionalized ­benzoate ligands with pronounced stereoactivity of the lone pair of electrons and metal-phenyl pibonding. Eur J Inorg Chem 2002: 2355–2361 46. Wiesbrock F, Schmidbaur H (2003) Complexity of coordinative bonding in Thallium(I) anthranilates and salicylates. J Am Chem Soc 125:3622–3630 47. Tereshko V, Minasov G, Egli M (1999) A “Hydrat-Ion” spine in a B-DNA minor groove. J Am Chem Soc 121:3590–3595 48. Klein DJ, Moore PB, Steitz TA (2004) The contribution of metal ions to the structural

Application of Anomalous Diffraction Methods to the Study of DNA and DNA-Complexes stability of the large ribosomal subunit. RNA 10:1366–1379 49. Morth JP, Pedersen BP, Toustrup-Jensen MS, Sorensen TL, Petersen J, Andersen JP, Vilsen B, Nissen P (2007) Crystal structure of the sodium-potassium pump. Nature 450: 1043–1049 50. Bond PJ, Langridge R, Jennette KW, Lippard SJ (1975) X-ray fiber diffraction evidence for neighbor exclusion binding of a platinum metallointercalation reagent to DNA. Proc Natl Acad Sci U S A 72:4825–4829 51. Komeda S, Moulaei T, Woods KK, Chikuma M, Farrell NP, Williams LD (2006) A third mode of DNA binding: phosphate clamps by a polynuclear platinum complex. J Am Chem Soc 128:16092–16103 52. Farrell N, Kloster MGB. High affinity DNA binding compounds as adjutants in antisense technology. US Patent 6,310,047 53. Otwinowski Z, Minor W (1997) Processing of X-ray diffraction data collected in oscillation mode. In: Carter CW Jr, Sweet RM, (eds) Methods in Enzymology, Macromolecular Crystallography, vol 276, Part A. Academic Press, New York, pp 307–326 54. Powell HR (1999) The Rossmann Fourier autoindexing algorithm in Mosflm. Acta Crystallogr D Biol Crystallogr 55:1690–1695 55. Collaborative Computational Project, Number 4 (1994) The Ccp4 Suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 50:760–763 56. McRee DE (1992) Xtalview: a visual protein crystallographic software system for X11/ Xview. J Mol Graph 10:44–46 57. Crick FHC, Magdoff BS (1956) The theory of isomorphous replacement for protein crystals. Acta Crystallogr A 9:901–908 58. Brunger AT, Adams PD, Clore GM, DeLano WL, Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J, Nilges M, Pannu NS, Read RJ, Rice LM, Simonson T, Warren GL (1998) Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr D Biol Crystallogr 54:905–921 59. Chaires JB, Fox KR, Herrara JE, Britt M, Waring MJ (1987) Site and sequence specificity of the daunomycin-DNA interaction. Biochemistry 26:8227–8236 60. Capranico G, Kohn KW, Pommier Y (1990) Local sequence requirements for DNA cleavage by mammalian topoisomerase II in the presence of doxorubicin. Nucleic Acids Res 18:6611–6619 61. Wang AH, Ughetto G, Quigley GJ, Rich A (1987) Interactions between an anthracycline

62.

63.

64.

65.

66. 67.

68.

69.

70.

71.

72.

151

antibiotic and DNA: molecular structure of daunomycin complexed to d(CpGpTpApCpG) at 1.2 Å resolution. Biochemistry 26:1152–1163 Frederick CA, Williams LD, Ughetto G, van der Marel GA, van Boom JH, Rich A, Wang AH-J (1990) Structural comparison of anticancer drug-DNA complexes: adriamycin and daunomycin. Biochemistry 29:2538–2549 Williams LD, Egli M, Ughetto G, van der Marel GA, van Boom JH, Quigley GJ, Wang AH-J, Rich A, Frederick CA (1990) Structure of 11-Deoxydaunomycin bound to DNA containing a phosphorothioate. J Mol Biol 215:313–320 Lipscomb LA, Peek ME, Zhou FX, Bertrand JA, VanDerveer D, Williams LD (1994) Water ring structure at DNA interfaces: hydration and dynamics of DNA-Anthracycline complexes. Biochemistry 33:3649–3659 Chaires JB, Satyanarayana S, Suh D, Fokt I, Przewloka T, Priebe W (1996) Parsing the free energy of anthracycline antibiotic binding to DNA. Biochemistry 35:2047–2053 Qu X, Chaires JB (2001) Hydration changes for DNA intercalation reactions. J Am Chem Soc 123:1–7 Quigley GJ, Wang AH-J, Ughetto G, van der Marel GA, van Boom JH, Rich A (1980) Molecular structure of an anticancer drugDNA complex: daunomycin plus d(CpGpTpApCpG). Proc Natl Acad Sci U S A 77:7204–7208 Williams LD, Frederick CA, Ughetto G, Rich A (1990) Ternary interactions of spermine with DNA:4¢-epiadriamycin and other DNA:anthracycline complexes. Nucleic Acids Res 18:5533–5541 Hu GG, Shui X, Leng F, Priebe W, Chaires JB, Williams LD (1997) Structure of a DNAbisdaunomycin complex. Biochemistry 36:5940–5946 Langlois d’Estaintot B, Gallois B, Brown T, Hunter WN (1992) The Molecular structure of a 4¢-epiadriamycin complex with d(TGATCA) at 1.7 Å resolution: comparison with the structure of 4¢-epiadriamycin d(TGTACA) and d(CGATCG) complexes. Nucleic Acids Res 20:3561–3566 Moore MH, Hunter WN, Langlois d’Estaintot B, Kennard O (1989) DNA-drug interactions: the crystal structure of d(CGATCG) complexed with daunomycin. J Mol Biol 206:693–705 Schumann W, Lindenblatt E, Bade EG (1976) Bacteriophage-specific DNA-binding proteins in P22-lysogenic and in P22-infected Salmonella typhimurium. J Virol 20:334–338

152

Watkins et al.

73. Poteete AR, Ptashne M (1982) Control of transcription by the bacteriophage P22 repressor. J Mol Biol 157:21–48 74. Maniatis T, Ptashne M, Backman K, Kield D, Flashman S, Jeffrey A, Maurer R (1975) Recognition sequences of repressor and polymerase in the operators of bacteriophage lambda. Cell 5:109–113 75. Murshudov GN, Vagin AA, Dodson EJ (1997) Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystallogr 53:240–255 76. Vagin AA, Steiner RA, Lebedev AA, Potterton L, McNicholas S, Long F, Murshudov GN (2004) Refmac5 dictionary: organization of prior chemical knowledge and guidelines for

its use. Acta Crystallogr D Biol Crystallogr 60:2184–2195 77. Agarwal RC, Isaacs NW (1978) Fast leastsquares method of structure refinement using the Fast-Fourier Algorithm – refinement of insulin at 1.5a resolution. Acta Crystallogr A Found Crystallogr 34:S46 78. Dauter Z, Adamiak DA (2001) Anomalous signal of phosphorus used for phasing DNA oligomer: importance of data redundancy. Acta Crystallogr D Biol Crystallogr 57:990–995 79. Guex N, Peitsch MC (1997) Swiss-Model and the Swiss-Pdbviewer: an environment for comparative protein modeling. Electrophoresis 18:2714–2723

Chapter 10 DNase I Footprinting Antonia S. Cardew and Keith R. Fox Abstract Footprinting is a method for determining the sequence selectivity of DNA-binding compounds in which ligands protect DNA from cleavage at their binding sites. Footprinting templates are typically 50–200 base pairs long, and DNase I is the most commonly used nuclease for these experiments. This chapter describes the preparation and labelling of suitable DNA footprinting substrates, the footprinting experiment itself, and the way in which these data can be used to estimate the dissociation constant of the interaction. Key words:  Footprinting, Sequence recognition, DNase I

1. Introduction Footprinting was first used in 1978 for studying the interaction between proteins and DNA (1). It has since been developed as a versatile method for identifying the sequence-specific interaction of many drugs with DNA. DNase I is the most commonly used cleavage agent for these experiments. It is relatively cheap and easy to use, and DNase I footprinting has been employed for identifying the DNA binding sites of a large number of ligands. 1.1. Footprinting

Footprinting is a protection assay designed to elucidate the sequence-specific binding sites for a ligand on double-stranded DNA. The ligand of interest protects the target DNA from digestion by a suitable cleavage agent. The DNA substrate for the reaction is usually between 50 and 200 base pairs long and is radiolabelled at one end of one strand. This substrate is cleaved in the presence and absence of a DNA-binding agent, and the regions to which the ligand is bound are protected from digestion, creating

K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_10, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

153

154

Cardew and Fox

Decreasing ligand concentration

5’

3’

Fig. 1. Schematic representation of a footprinting experiment. On the left is a representation of a duplex DNA template to which a ligand (filled box) is bound. The asterisk indicates the position of the radiolabel. On the right is a schematic representation of a footprinting gel. The first lane shows the control cleavage pattern in the absence of added ligand, while the other lanes indicate the pattern in the presence of decreasing concentrations of the ligand.

a gap or ‘footprint’ in the ladder of cleavage products when these are resolved on denaturing polyacrylamide gels (see Fig.  1). The reaction requires that each DNA molecule is only cleaved once (single hit kinetics) (see Note 1). Footprinting has been an invaluable tool for elucidating the sequence selectivity of many small molecules (2) including actinomycin (3, 4), mithramycin (5, 6), quinoxaline antibiotics (7, 8), nogalamycin (9), AT-selective minor groove binding ligands (3, 4, 10) and hairpin polyamides (11–13). It has also been widely used to examine the binding of triplex forming oligonucleotides (TFOs) to DNA (14–16). Although this chapter focuses on the use of DNase I, a number of other enzymatic and chemical cleavage agents have also been used for footprinting studies including micrococcal nuclease (17), methidiumpropyl-EDTA.Fe(II) (MPE) (18, 19), DNase II (7), copper phenanthroline (20), uranyl photoclevage (21, 22), and hydroxyl radicals (23–25). Each of these has distinct advantages

DNase I Footprinting

155

and disadvantages and has been used to examine different aspects of ligand interactions with DNA. An ideal cleavage agent should be sequence neutral, so as to produce an even distribution of cleavage products. Although DNase I is the most commonly used, due to its low cost and ease of use, it generates an uneven ladder of cleavage products, as the efficiency of the enzyme is affected by the global and local DNA structure (26, 27). In contrast, hydroxyl radicals, which are less versatile footprinting probes, generate a much more even ladder of cleavage products (25), though these cannot be used for footprinting ligands that only bind in the major groove (such as triplex-forming oligonucleotides). 1.2. DNase I

DNase I is a monomeric glycoprotein with a molecular weight of about 30,400. It is a double strand-specific endonuclease, though it will cut single-stranded DNA, albeit at a much lower rate. It requires the presence of divalent cations and makes single strand nicks in the DNA backbone, cleaving the O3¢-P bond. Optimal cleavage is obtained using calcium or magnesium, while manganese ions enhance the activity (see Note 2). Several crystal structures of DNase I have been determined in both the presence and absence of its DNA substrate (28–31). These structures show that the enzyme binds by inserting an exposed peptide loop into the DNA minor groove (see Note 3). This loop interacts with the walls of the groove as well as the phosphate backbone. Although DNase I cuts all sequences, it generates an uneven cleavage pattern due to local sequencedependent variations in DNA structure (26). Runs of As or Ts are poor substrates as the minor groove is too narrow for the insertion of the exposed loop (26). The crystal structures also reveal the enzyme causes the DNA to bend by about 21° towards the major groove, away from the enzyme. The bending is thought to be an important part of the cleavage mechanism. GC rich regions are therefore poor substrates as a result of their more rigid structure that is harder to bend. In general, YpR steps (Y = C or T; R = A or G) are cut less well than RpY; regions of (AT)n therefore generate an alternating pattern of light and dark bands with ApT cut much better than TpA. The crystal structures also suggest that the tyrosine at position 76 interacts with the DNA base 2 positions on the 5¢-side of the cleavage site. Arginine-41 also binds to the base at position -3, and this interaction is sterically hindered by a GC base pair at this position. From studying the crystal structures and cleavage pattern of DNase I, it has been suggested that the best cleavage site for the enzyme is WYW/WVN (W = A or T; Y = C or T; V = A, G, C) (32). DNase I binds to about 10 base pairs of the DNA duplex; one complete turn of the helix. As a result, the enzyme overestimates the size of drug binding sites. Since DNase I binds across the

156

Cardew and Fox

width of the minor groove, which runs at an angle relative to the helix axis, the enzyme can approach closer to the 5¢-side of the ligand binding site than the 3¢-side. As a result, footprints are often staggered by 2–3 base pairs across the two strands, relative to the actual binding site. This problem can be overcome by comparing the location of the footprints on both strands of the template individually.

2. Materials 2.1. DNase I

The DNase I used for footprinting experiments does not need to be especially pure; there is usually no need to use HPLC-pure or RNase-free enzyme. We currently use the type IV enzyme, from bovine pancreas, purchased from Sigma-Aldrich. The enzyme should be dissolved in 0.15 M NaCl containing 1 mM MgCl2 at a concentration of 7,200 Kunitz U/ml (see Note 4) and should be stored at −20°C. It is stable to frequent freezing and thawing. DNase I should be diluted to the working concentrations from this stock immediately before use, and the diluted enzyme should be discarded at the end of each experiment.

2.2. Choice of DNA Fragment

DNA footprinting substrates are typically 50–200 base pairs long; the length is only limited by the resolution of polyacrylamide gels. Although the DNA substrates employed vary from one laboratory to another, the tyrT fragment has been used more widely than many others (3, 7, 33). This is a 160 base pair fragment derived from the promoter of the tyrosine tRNA gene; its sequence is shown in Fig. 2 and it is freely available from the author’s laboratory on request (see Note 5). Other substrates can be excised from plasmid vectors, using appropriately chosen restriction enzymes. A problem when using natural DNA sequences is that binding sites for ligands that recognise a large number of bases may not be represented. The affinity of ligands for their binding sites may also be influenced by the flanking sequences, which affect the local DNA structure. Natural DNA fragments with a random distribution of available sequences may not be best suited for testing all possible binding sites of a ligand with unknown properties.

2.2.1. Natural DNA Restriction Fragments

2.2.2. Synthetic Oligonucleotides

If some properties of the ligand binding site are already known, then it may be easiest to design a synthetic oligonucleotide to use as a template in footprinting studies. For example, a ligand with unknown binding properties might first be tested on a natural fragment to obtain a general idea of its sequence preference, followed by the use of a synthetic template containing the putative binding site(s) (37, 38).

DNase I Footprinting

157

Ty rT(43-59) 3’-AAGGCCAATGGAAATTAGGCAATGCCTACTTTTAATGCGTTGGTCAAGAAAAAAGAGAAGGATTGTGAAATGTCGCCGCGCAGTAAA 10 20 30 40 50 60 70 80 CTATACTACGCGGGGCGAAGGGCTCTTCCCTCGTCCGGTCATTTTTCGTAATGGGGCACCACCCCCAAGGG-5’ 90 100 110 120 130 140 150 MS2 5’-GGATCCGCATTCGAGGCTGAGATGACAAACCAGACCCCACCGGACGTACTTTACATAACTCTTCACGCCCTAATTGCTATACCAGGATAGAAC GGGAGCTTAACCTTGATCGCGCTACGACTAGTGCAGTTGGAAATCGGCCATGTGTATTGCCGCATATGGATCC HexA 5’-GGATCCCGGGATATCGATATATGGCGCCAAATTTAGCTATAGATCTAGAATTCCGGACCGCGGTTTAAACGTTAACCGGTACCTAGGCCTGC AGCTGCGCATGCTAGCGCTTAAGTACTAGTGCACGTGGCCATGGATCC-3’ HexB 5’-GGATCCGGCCGATCGCGAGCTCGAGGGCCCTAATTAGCCGGCAATTGCAAGCTTATAAGCGCGCTACGTATACGCGTACGCGCGTATATACA TATGTACATGTCGACGTCATGATCAATATTCGAATTAATGCATGGATCC-3’

Fig. 2. Sequences of some commonly used footprinting fragments. TyrT(43–59) is a derivative of the tyrT fragment that contains a 17 base pair oligopurine tract between positions 43 and 59. The changes from the original tyrT sequence are indicated in bold, including the AvaI site generated at position 110. The sequence shown corresponds to the strand that is seen when labelling the EcoRI end of the restriction fragment. MS2 is an insert that contains every possible tetranucleotide sequence. The figure shows the strand that is seen when the fragment (which is cloned into the BamHI site of pUC18) is labelled at the 3¢-end of the HindIII site. MS1 corresponds to the same sequence cloned in the opposite orientation, visualising the complementary strand. HexA and HexB are sequences that between them contain all 64 symmetrical hexanucleotide sequences.

A study with designed synthetic fragments will also be able to examine the effects of the flanking sequences. It is often worth examining sequences that are closely related to the predicted site(s), to ensure that alternative binding sites have not been excluded. 2.2.3. Synthetic Restriction Fragments

A more common procedure is to clone a chosen designed sequence into a plasmid vector and isolate this within a restriction fragment for use as a footprinting substrate. Several useful target DNA sequences have been created in this way to aid footprinting of ligands with unknown binding properties, particularly small drug molecules with binding sites of 3–6 bases. For example, fragments for use with AT-selective minor groove binders contain a selection of (A/T)4 sites separated by GC-rich regions (10). These synthetic fragments are often cloned into the polylinker site of pUC vectors (6, 10, 39, 40) (see Note 6). More recently, a set of templates have been created which are particularly useful for ligands with small recognition sites. First, MS1 and MS2 are fragments that contain every possible tetranucleotide combination (41) (Fig.  2). These two fragments contain the same sequence which is cloned in opposite orientations to facilitate visualisation of binding sites that are located at opposite ends of the fragments. Similarly, the HexA and HexB fragments (Fig. 2) contain all the 64 symmetrical hexanucleotide sequences (these are each available with the insert cloned in either direction). These are especially

158

Cardew and Fox

useful for (pseudo)-symmetrical ligands, which are likely to bind to symmetrical DNA sequences (42). These too are available from the author’s laboratory. 2.3. Buffers 2.3.1. Plasmid Preparation

1. Commercial kits are recommended for plasmid purification. 5 ml of an overnight bacterial culture usually provide sufficient plasmid for one batch of labelling. 2. Sterile 2YT medium: 16 g tryptone, 10 g yeast extract and 5 g NaCl/l.

2.3.2. General Buffers

1. TE1: 10 mM Tris–HCl, pH 7.5, containing10 mM EDTA is used for eluting DNA from gel slices. 2. TE2: 10 mM Tris–HCl pH 7.5, containing 0.1 mM EDTA is used to resuspend labelled DNA after gel purification, ready for footprinting experiments. 3. 10 mM Tris–HCl, pH 7.5 containing 10 mM NaCl is used for preparing drug solutions. Other buffers can be used as appropriate (e.g., 50 mM sodium acetate pH 5.0 for work with triplex-forming oligonucleotides). 4. DNase I solution: 20  mM NaCl, 2  mM MgCl2, 2  mM MnCl2. 5. Loading dye for non-denaturing gels: 20% Ficoll, 10  mM EDTA, 0.1% bromophenol blue. 6. DNase I stop solution: Formamide containing 10 mM EDTA, 0.1% bromophenol blue and 1 mM NaOH.

2.3.3. Electrophoresis

1. TBE electrophoresis buffer (10× stock). Make up as a 10× stock: 108  g Tris, 55  g boric acid and 9.4  g EDTA in 1  l water. 2. Acrylamide solutions. Polyacrylamide sequencing gels are made from acrylamide:bisacrylamide in a 19:1 ratio. (Sequagel for denaturing gels or Accugel for non-denaturing gels: National Diagnostics). 3. 8 M urea. 4. TEMED. 5. 20% ammonium per sulphate.

2.3.4. Radioactivity

a-32[dNTP] or g-32P[ATP] are purchased at a concentration of 3,000 Ci/mmol (Perkin Elmer) (see Note 7).

2.3.5. Preparing and Labelling Restriction Fragments

1. AMV reverse transcriptase (Sigma or Promega). 2. Restriction enzymes. 3. Digestion buffer (Promega buffer B).

DNase I Footprinting

159

3. Methods 3.1. Plasmid Preparation

Several plasmid preparation kits are commercially available and give good yields of plasmids of appropriate quality. Mini-prep kits (suitable for plasmid extraction from 3 to 5 ml culture) are appropriate and we routinely use those from Promega or Qiagen (see Note 8).

3.2. Radiolabelling the DNA

Restriction fragments can be labelled at the 3¢-end by filling in appropriate sticky ends with a-32P[dNTP] using a polymerase, such as Klenow fragment or reverse transcriptase. Synthetic oligonucleotides are usually labelled at the 5¢-end using polynucleotide kinase and g-32P[ATP]. Generally, 3¢ labelling is preferred for DNase I footprinting studies as the DNase I digestion patterns are easier to interpret. DNase I cleaves at the O3¢-P bond, producing products with a 3¢-hydroxyl and 5¢-phosphate groups. GA tracks, used as markers lanes (Subheading  3.3), produce fragments with phosphate groups at both 3¢- and 5¢-ends. When using 3¢ labelled DNA fragments, the radiolabelled DNase I cleavage products and the GA marker lanes will both possess 5¢-phosphates and will comigrate on polyacrylamide gels. However, if 5¢ labelled DNA is used, then the radiolabelled DNase I cleavage products will terminate with a 3¢-hydroxyl compared with a 3¢-phosphate for the marker lanes; the marker lanes, therefore, migrate faster than the corresponding cleavage products, an effect which is especially pronounced for short fragments.

3.2.1. 3¢-End Labelling with Reverse Transcriptase

The fragment of choice is excised from the plasmid using a pair of restriction enzymes, at least one of which produces sticky ends with a 5¢-overhang, which is suitable for filling with a polymerase using a-32P[dNTP]. If both restriction enzymes produce similar sticky ends, then the second restriction digest must be performed after the labelling reaction, so as to avoid labelling both ends of the fragment. Typical pairs of restriction enzymes are EcoRI and AvaI (or SmaI) for the tyrT fragment; HindIII and SacI or EcoRI and PstI for pUC-derived vectors. It is convenient to use a buffer that is compatible with both restriction enzymes as well as reverse transcriptase, so as to avoid the need for changes of buffer during the process. We routinely use Promega Buffer B for these combinations of restriction enzymes. Klenow fragment of DNA polymerase I has frequently been used for end-filling, but we routinely obtain better yields with AMV reverse transcriptase (even though this is actually an RNA-dependent DNA polymerase) (see Note 9). This is because the exonuclease activity of the Klenow fragment can also remove the radiolabel (43).

160

Cardew and Fox

Restriction Digestion and 3¢-End Labelling

The following describes the procedure for a HindIII-SacI fragment (see Notes 10 and 11): 1. Mix 45 µl of plasmid DNA (the entire preparation from a 5 ml culture, dissolved in water) with 5 µl 10 × restriction enzyme buffer B, 1 µl HindIII and 1 µl SacI. Incubate at 37°C for at least 1 h. 2. Add 1 µl a-32P[dATP] (3,000  Ci/mmol, Perkin Elmer) and 0.5 µl AMV reverse transcriptase. Incubate at 37°C for a least 1 h.

Purification Radiolabelled DNA Fragments

1. Add 10 µl of loading dye to the labelling reaction and load the sample onto a 6–8% non-denaturing polyacrylamide gel. We use gels that are 0.3 mm thick and 40 cm long. The gels should be run at a temperature that does not denature the DNA. We typically apply 800 V (about 20 W). 2. Allow the loading dye to reach the bottom of the gel (typically 1.5–2 h). This runs any unincorporated label off the gel. 3. Separate the glass plates, cover the gel with Saran wrap and expose to X-ray film for 3–5 min. 4. Use the X-ray film to identify the location of the radiolabelled bands and excise these from the gel with a sharp razor blade. 5. Elute the radiolabelled DNA from the gel slice. A simple, effective and cheap extraction can be achieved as follows: Insert a small plug of glass wool at the bottom of a P1000 pipette tip and carefully seal the end of the tip with parafilm. Place the gel slice on top of the glass wool and cover this with 300–400 µl of TE1 buffer. Cover the top of the tip with parafilm, place in an open Eppendorf tube (in case it leaks) and gently agitate for at least 3 h. 6. Remove the parafilm from the top and bottom of the tip and expel the buffer, now containing the radioactive DNA, into an Eppendorf tube using a P1000 pipette. Residual buffer can be expelled by briefly centrifuging the tip, placed in an open Eppendorf tube. (see Note 12). 7. A small amount of polyacrylamide (and other insoluble material) is sometimes present in the eluate. Remove this by centrifuging the sample for 1 min and transferring the radioactive supernatant to a new tube. 8. Precipitate the DNA by adding at least three volumes of ethanol. Mix by inversion and leave on dry ice for at least 10 min. Spin the sample at 14,000 g for 10 min to collect the DNA. (see Note 13). 9. Carefully remove the supernatant in 200 µl aliquots. Check each aliquot against a hand-held Geiger counter to confirm that it does not contain any radioactivity.

DNase I Footprinting

161

10. Wash the precipitate with 200 µl of 70% ethanol and spin again for 1 min. Remove the supernatant and dry the samples under vacuum (see Note 14). 11. Dissolve the radiolabelled DNA in an appropriate buffer such as TE2 at a concentration of approximately 10  cps/µl, as determined using a hand-held Geiger counter (see Note 15). For footprinting experiments, it is generally not necessary to know the exact concentration of the DNA template, so long as it is lower than the ligand concentration. 3.2.2 5¢-End Labelling with Polynucleotide Kinase

DNA fragments can also be labelled at the 5¢-end of one strand using polynucleotide kinase and g-32P[ATP], provided that it is not already phosphorylated. This is, therefore, especially useful for labelling one strand of a synthetic DNA fragment before annealing with the unlabelled complement. 1. Mix 13 µl of water with 2 µl of 5 µM oligonucleotide. 2. Add 2 µl of g-32P labelled ATP, 2 µl of polynucleotide kinase buffer and 1 µl of polynucleotide kinase. 3. Incubate for 1 h at 37°C. 4. Add 5 µl of DNase I stop solution and heat the sample at 100°C for 3  min to ensure the DNA is completely denatured. 5. Load the sample onto a denaturing polyacrylamide gel (8–14%, depending on the length of the fragment). The gel should be run hot (1,500 V (40 W) for a 40 cm gel, 0.3 mM thick). (see Note 16). 6. When the loading dye has reached the bottom of the gel, the labelled DNA is isolated and extracted in the same way as described in “Purification Radiolabelled DNA Fragments”.

3.3. Maxam–Gilbert Marker Lanes

The identity of bands in the DNase I cleavage patterns is determined by comparison with appropriate sequence marker lanes. Maxam–Gilbert sequencing methods, for G or G+A have typically been used to create marker lanes for footprinting gels. However, this procedure is time consuming and involves several steps. We therefore use a much simpler empirical method for generating purine (G+A) marker lanes, which can be freshly prepared for each footprinting experiment. 1. Mix 1.5 µl of radiolabelled DNA with 20 µl of distilled water and 5 µl of DNase I stop solution in an Eppendorf tube. 2. Heat at 100°C for 30 min with the cap open then crash cool on ice. Load the entire sample onto the gel alongside the samples that have been digested with DNase I.

162

Cardew and Fox

3.4. DNase I Footprinting

The basic protocol for DNase I footprinting experiments is relatively quick and simple; a footprinting experiment can be performed in 1 day, once the radiolabelled template has been created. It is also easily adaptable to any desired conditions. 1. Mix 1.5 µl of radiolabelled DNA template (dissolved in TE2) with 1.5 µl of the ligand in an appropriate binding buffer at 2× the desired concentration (see Note 17). Several different binding reactions are usually set up in parallel containing serial dilutions of the ligand (see Note 18). Include at least one control reaction incubating the DNA with buffer alone (no ligand). 2. Leave to equilibrate for an appropriate amount of time at the required temperature; this may be 30  min or less for many small molecules, but may require overnight incubation to achieve equilibration with triplex-forming oligonucleotides. 3. Prepare a fresh solution of DNase I, diluted from the stock in DNase I buffer. The concentration is determined empirically by trial experiments, and depends on the reaction conditions, but is typically about 0.02 Units/ml (see Note 19). A general scheme for dilution is to add 2 µl of stock DNase I to 1 ml of DNase I buffer, mix gently and then add 2 µl of this dilution to a further 1 ml of DNase I buffer. This second dilution is then used in the footprinting experiments. 4. Begin the DNase I digestion by adding 2 µl of diluted DNase I to each sample. Leave to digest for 1 min before stopping the reaction by adding 4 µl of DNase I stop solution (see Notes 20 and 21).

3.5. Electrophoresis and Autoradiography

1. Boil the samples for 3 min and then crash cool on ice, immediately before loading onto denaturing polyacrylamide gel. The acrylamide concentration will need to be varied depending on the length of the sequence, but is typically between 7 and 12% (8% for the tyrT, MS1, MS2, HexA and HexB fragments). These gels should be run hot (1500  V (40  W) for 0.3 mm thick, 40 cm long gels) (see Note 22). Gels are prepared and run in 1× TBE buffer and are typically preheated by running for 30 min before loading the samples. 2. Run the gel until the bromophenol blue reaches the bottom of the gel (1.5–2 h). 3. Separate the glass plates and immerse the gel in 2  l of 10% (v/v) acetic acid for at least 10 min (see Notes 23 and 24). This fixes the DNA in the gel and removes a large amount of the urea before drying. High percentage gels (14% and above) will expand in the acetic acid and may crack when dried. This can be reduced by soaking the gel in 10% methanol until the gel has shrunk back to its original size.

DNase I Footprinting

163

4. Remove the gel from the acid and drain off most of the liquid. Lay a cut sheet of Whatman 3MM paper onto the wet gel and then peel this back, transferring the gel onto the filter paper. Cover this with Saran wrap and dry at 80°C for 1 h on a gel dryer (see Note 25). 5. After drying, the gel is ready to be imaged. This was originally achieved by exposing to X-ray film. However, this is now more commonly achieved with phosphorimaging. We use a Molecular Dynamic STORM phosphorimager. Dried gels are typically exposed to the phosphor screen overnight and then scanned at a pixel size of 100 µm. 3.6. Analysis 3.6.1. Visual Inspection

3.6.2. Differential Cleavage Plots

For most footprints, the binding sites can be assessed by visual inspection and simply aligned with the DNA sequence. If the DNA fragment contains many binding sites for the ligand, then some of these may be too close together, causing the footprints to overlap. This generates a large region of protection, which is difficult to analyse. If it is possible to perform separate experiments with the DNA labelled on both strands, then the exact binding site may be easier to identify as it will be in the region of overlap between them. If the ligand has low specificity, or recognises regions of DNA by virtue of their structure, rather than their sequence, then the footprinting patterns can be difficult to interpret. In these instances, it is worthwhile trying a number of possibilities to see which one best fits with the data (is the binding site a specific base sequence, runs of purines/pyrimidines, specific arrangement of purines and pyrimidines?). The interpretation should be internally consistent; that is, all sequences of the type should be within a footprint, and no footprints should be present in other sequences. If there are still uncertainties over the sequence selectivity of the ligand, then these are best resolved by preparing a synthetic footprinting fragment that contains the putative binding site (and variants of it) and performing further footprinting experiments with this. A laborious method for identifying the preferred binding sites, for which a reader does not have to view the original gels, is to present the data in the form of a differential cleavage plot (7, 44). For this, the relative intensity of each band in the cleavage pattern is determined by densitometry, and the intensity of bands in the ligand-treated lanes is compared with that in the control. The data are normalised according to either the total radioactivity in each lane or the intensity of a band (or bands) which is not affected by the ligand (see Note 26). These differential cleavage values are then plotted (often on a logarithmic scale) against the sequence. Values less than one indicate protections from cleavage, while values greater than one indicate enhanced cleavage (see Note 27).

164

Cardew and Fox

3.6.3. Enhanced Cleavage

It is not uncommon to observe enhanced DNase I cleavage in regions that surround ligand-binding sites. This can be the result of two different effects. 1. If a large proportion of the DNA is occluded by ligand binding, then the ratio of free DNA to enzyme will increase. This will necessarily result in increased cleavage of all the DNA regions that are not bound by the ligand. 2. The ligand may produce changes in the local DNA structure which render it more susceptible to DNase I cleavage. This is often seen around the binding sites for intercalators, especially if these are in regions that are normally poor substrates for the enzyme (such as An.Tn tracts). Single bonds of enhanced cleavage are often seen at the 3¢-end of the footprints produced by triplex-forming oligonucleotides.

3.7. Quantitative Footprinting

A number of methods have been outlined for rigorously analysing concentration-dependent footprinting data, in order to obtain quantitative estimates of binding constants (45, 46). In general, the data do not warrant such a detailed approach and good estimates of dissociation constants can usually be obtained by applying a simple binding equation to densitometer scans of the footprinting data. This analysis calculates the ligand concentration that reduces the intensity of bands in the footprint by 50% (termed the C50 value). This is achieved by comparing the intensity of bands within the footprint (or the entire footprint see Note 28), derived from a densitometer or ImageQuant analysis of the phosphorimage, at different ligand concentrations. These are then normalised by either dividing by the intensity of a band which is not affected by the ligand or by the total intensity of all bands in each lane (see Note 29). These normalised intensities at different ligand concentrations are combined to produce a ‘‘footprinting plot’’ of intensity against concentration which is fitted by the simple binding equation I/I0 = C50/(L + C50), where I and I0 are the relative band intensities in the presence and absence of the ligand, respectively, and L is the ligand concentration. This analysis assumes that the concentration of free ligand is unchanged by binding to the very low concentration of DNA, and is therefore only valid if the DNA concentration is lower than the dissociation constant of the interaction (see Note 30). C50, the ligand concentration at which the intensity of the bands in the footprint is reduced by 50%, then approximates to the dissociation constant of the ligand from the DNA (Kd). This analysis assumes that there is no DNase I cleavage when the binding site is fully occupied by the ligand (i.e., that I/I0 tends to zero at high ligand concentrations). However, we find that there can be some residual band intensity which is not reduced at high ligand concentrations. This could be because

DNase I Footprinting

165

DNase I is still able to cut the ligand–DNA complex (though with much reduced efficiency). This is unlikely for most minor groove binding ligands, though it may be possible for compounds that indirectly block DNase I activity by binding in the major groove (such as triplex forming oligonucleotides). The residual cleavage may also be caused by impurities in the DNA preparation, which yield bands that are not a direct result of DNase I cleavage (see Note 31). It is possible to allow for this residual cleavage intensity, by fitting the data to the equation I/I0 = C50/ (L + C50) + R, where R is the residual cleavage (which must be positive). Any regions of enhanced DNase I cleavage, which are observed adjacent to footprinting sites, can also be used to obtain a quantitative estimate of the ligand’s affinity. 3.8. An Example

Figure 3 shows a typical DNase I footprint obtained for the interaction of a 12-mer triplex-forming oligonucleotide with a target site in the tyrT(43–59) DNA fragment. This fragment contains a 17 base pair oligopurine tract between positions 43 and 59. The control shows a typical uneven cleavage pattern with two regions of especially poor cleavage. These are located toward the bottom of the gel in the oligodT tract around position 30, and in the oligopurine triplex target site itself, though one ApG step within this region shows good cleavage. The triplex-forming oligonucleotide generates a clear region of protection at the intended target site, which extends above (5¢) this site by about 2 bases. The footprint, which is most clearly seen by the reduced cleavage of the one strong band, is concentration-dependent and the cleavage pattern at the lowest concentrations resembles that in the control. A single band of enhanced DNase I cleavage is evident at the 3¢-duplex–triplex boundary (at the bottom of the footprint). This localised enhancement is not explained by changes in the ratio of free to bound DNA and appears to correspond to a change in the local DNA structure rendering it more susceptible to DNase I cleavage. This enhancement is also concentration-dependent. It is worth remembering that the DNA concentration in these experiments is very low (typically lower than 10 nM). The disappearance of the footprint at lower oligonucleotide concentrations is therefore not because of the stoichiometry (even at the lowest concentrations, there will be an excess of ligand to DNA substrate). The reduced footprint is a consequence of the binding reaction in which the concentration of the triplex-forming oligonucleotide has fallen well below the dissociation constant for this interaction. The target site will be 50% occupied at a concentration equivalent to the dissociation constant of the ligand. Quantitative analysis of these data can therefore be used to estimate the dissociation constant of the interaction. Footprinting plots, derived from either the reduced intensity of the footprint

166

Cardew and Fox

Fig.  3. DNase I footprinting patterns for the tyrT(43–59) fragment in the presence of varying concentrations of the 12-mer triplex-forming oligonucleotide 5¢-TMTMTBTBTBMT (M = 3-methyl-2-aminopyridine; B = bis-amino-dU). The experiment was performed in 50 mM sodium acetate pH 5.0. The oligonucleotide concentration (µM) is shown at the top of each lane. Con indicates control cleavage in the absence of added ligand. GA is a marker lane that is specific for purines. The filled box indicates the location of the 17 base oligopurine tract; the open box shows the extent of the DNase I footprint, while the bracket shows the position of the intended target site. The arrow shows the position of enhanced DNase I cleavage at the 3¢-end of the target site.

or the enhanced cleavage at the triplex–duplex boundary are illustrated in Fig. 4.

4. Notes 1. If the enzyme concentration is too great, or the DNA is digested for too long, then the short fragments (towards the bottom of the gel) will be overrepresented, as they arise from multiple cleavage events. The intensity of longer products (towards the top of the gel) is then greatly reduced or they are absent altogether. If this happens, then the experiment should be repeated with a lower concentration of the enzyme. Overdigested cleavage patterns can be used to obtain a crude estimate of the binding sites, but should not be published.

DNase I Footprinting

167

Fig. 4. Footprinting (C50) plots derived from the data shown in Fig. 10.3. Open circles correspond to data obtained from the intensity of all the bands in the footprint, while filled circles correspond to the intensity of the strong band located towards the top of the footprint. Each of these was normalised with respect to the intensity of all the cleavage products in the lane. The data are further normalised so that the value for the free DNA is unity. Triangles show data derived from the intensity of the enhancement, located at the bottom (3¢-end) of the footprint. This was normalised relative to the intensity of all the bands in each lane and was adjusted so that the intensity was 1 in the presence of 3 µM triplex-forming oligonucleotide. Simple binding curves were fitted to these data and yield C50 values of 0.35 ± 0.10; 0.43 ± 0.07 and 0.66 ± 0.12 µM, respectively.

The uncut DNA at the top of the gel should normally be retained in figures for publication, though this is often removed and only the region of interest is presented. 2. Since divalent cations are essential for DNase I, the reaction can be easily terminated by adding EDTA. Some publications suggest that calcium is essential, while we find that magnesium alone is sufficient. Manganese can be used to speed the rate of reaction. A few publications suggest that manganese alters the mechanism of cleavage, though we have observed no effects on the cleavage pattern. The reaction is inhibited by some transition metals such as zinc. 3. DNase I should therefore only produce footprints for ligands that occlude the minor groove. However, it is also widely used for footprinting triplex-forming oligonucleotides, which bind in the major groove. In this case, it is assumed that the cleavage is inhibited as a result of changes in DNA structure and/ or flexibility, rather than by steric occlusion. 4. The activity of DNase I is usually expressed in Kunitz units, where one unit is the amount of DNase I that causes an increase of 0.001 absorbance units per min when added to 1 mg/ml salmon sperm DNA when assayed in a 0.1 M sodium acetate, pH 5.0.

168

Cardew and Fox

5. Several versions of the original tyrT DNA fragment (which was cloned into plasmid pKM (26)) have now been prepared, cloned into pUC vectors (34, 35). Point mutations have been introduced into some of these for studies on triple–helix formation (36). The most commonly used form (shown in Fig. 2) also contains a point mutation around position 110 which has introduced an additional AvaI site. Cutting with EcoRI and AvaI yields a shorter fragment of 110 base pairs. The full length 160 base pair fragment can be isolated by cutting with EcoRI and SmaI. 6. If blue-white selection (with ß-galactosidase and X-gal) is used to select for successful clones, then ensure that the inserted oligonucleotide is either not a multiple of three nucleotides or that it contains an in-frame stop codon. 7. Until recently, the major supplier of radioactive nucleotides in the UK was Amersham (GE Healthcare). They no longer supply radioactivity. 8. Some laboratories prepare high purity plasmids by caesium chloride density gradient centrifugation. However, since the radiolabelled fragments are subsequently isolated by gel purification, a high level of purity is not required at this stage, though the plasmid DNA must be free of nucleases and any material that would inhibit restriction enzymes or polymerases. 9. We find that Mlu reverse transcriptase does not work. 10. If both enzymes produce sticky ends that can be filled in with dATP then the second digest must be done after incorporating the radiolabel. For example, cut with HindIII, label with reverse transcriptase and a-32P[dATP], heat at 65°C for 5  min to inactivate the reverse transcriptase, cool to 37°C and add EcoRI. 11. The restriction fragments must not contain internal copies of these restriction sites. For instance, HexA contains internal EcoRI and PstI sites and so the fragment is usually prepared by cutting with HindIII and SacI, whereas HexB contains HindIII and SacI sites and so the footprinting fragment is released with EcoRI and PstI. 12. Check to confirm that most of the radioactivity has eluted from the gel slice with a handheld Geiger counter. For fragments of 250 base pairs or shorter, we routinely elute at least 95% of the radioactivity in this way, though the efficiency decreases for longer fragments. If too much radioactivity is left in the slice, elute with more buffer for longer, making sure that the slice is completely submerged. 13. The DNA pellet will be much too small to see by eye (though a small amount of precipitated salt may be visible). It is therefore sensible to spin all the Eppendorf tubes in the same

DNase I Footprinting

169

orientation in the centrifuge; i.e., with the hinge up, so as to be sure of the position of the radioactive pellet. 14. Ensure that all the residual ethanol has been removed, but do not dry the samples for too long, or they will be difficult to redissolve. 15. The DNA should dissolve easily with only mild agitation. It should not require prolonged agitation. 16. In principle, synthetic oligonucleotides can be separated from the excess radiolabel by gel filtration, using, e.g., Nap columns. However, oligonucleotides are chemically synthesised in the 3¢-5¢ direction, and may have ragged 5¢-ends (i.e., with a small amount of n-1 products). These will also be labelled and, if they are not removed from the gel, will give shadow bands in footprinting experiments. We therefore recommend that radiolabelled synthetic oligonucleotides for footprinting should be isolated by gel electrophoresis. 17. If there is a large difference between the pH required for the ligand and that of the TE2 buffer (pH 7.5), then the volume of the added ligand should be increased. For example, with triplex-forming oligonucleotides, which require conditions of low pH, we add 3 µl of the TFO (at 1.5 times the final concentration) dissolved in 50 mM sodium acetate pH 5.0. 18. It is usually best to use a range of ligand concentrations. For initial experiments, these might be separated by a factor of 10 (e.g., 100 µM, 10 µM, 1 µM, 0.1 µM, 0.01 µM). For quantitative experiments, a much narrower range of concentration is used which are varied around the effective concentration (say 10 µM, 3 µM, 1 µM, 0.3 µM, 0.1 µM). 19. DNase I has a broad pH profile and can be used between pH 5.0 and 9. It is inhibited by high ionic strengths (above 100 mM NaCl), but the concentration of the enzyme can be adjusted to achieve cleavage in 1 M NaCl. 20. If the DNA is over- or under-digested, then it is usually better to adjust the DNase I concentration, rather than altering the digestion time. 21. For these experiments, it is important that all the samples are digested for the same length of time with fresh DNase I. With practise, it is possible to perform 15 (or more) digestions in parallel, adding the DNase I in turn to each one during the 1  min digestion, followed by stopping the reaction in the same order. 22. The temperature of some electrophoresis tanks can be controlled (use about 55–60°C). On older apparatus, the gels tend to “smile” as they become hotter in the middle than at the edges. This effect can be overcome by clamping a metal plate over the glass gel to ensure even heat distribution.

170

Cardew and Fox

23. The gel can be left in the acid for a long time (overnight if necessary). 24. Two litres of 10% (v/v) acetic acid can be used for fixing three gels before it should be discarded. 25. Do not break the seal on the gel drier too soon, or the gel will crack into small pieces and will be unusable. 26. This normalisation corrects for differences in loading, but does not correct for differences in the extent of digestion between different lanes if the conditions do not fulfil singlehit kinetics. 27. A further complication with this approach is that it is not possible to determine accurate values for bands that are cut very poorly in the control. 28. This can provide a more accurate estimate of the C50 but assumes that the footprint only contains a single ligand binding site. 29. The intensity should refer to the volume of each band in the footprint, corrected for any background. However, this can often be difficult to determine as there may be overlap between the individual bands, especially if these are located towards the top of the gel. Many densitometric analyses simply drop vertical lines from the minima in the trace, rather than properly fitting the peaks with Gaussian curves. It may therefore be appropriate to compare the heights of the bands rather than their areas. This problem is less pronounced when comparing the sum of the intensities of all the bands in a footprint. 30. The concentration of radiolabelled DNA is usually less than 10 nM and this condition is therefore fulfilled for ligands with a Kd of 100 nM or higher. However, this can be a problem for compounds which have very high affinities (such as some hairpin polyamides, or triplex-forming oligonucleotides). In these instances, the labelled DNA must be diluted and the footprinting reaction is performed in much larger volumes (up to 1 ml). 31. Take especial care not to contaminate the stock of radiolabelled DNA with traces of DNase I. It may be worth including a lane containing uncut DNA on the gel to check for any contaminants.

Acknowledgements Work in the author’s laboratory is currently funded by the BBSRC.

DNase I Footprinting

171

References 1. Galas DJ, Schmitz A (1978) DNAase footprinting – simple method for detection of protein–DNA binding specificity. Nucleic Acids Res 5:3157–3170 2. Hampshire AJ, Fox KR (2008) The effects of local DNA sequence on the interaction of ligands with their preferred binding sites. Biochimie 90:988–998 3. Fox KR, Waring MJ (1984) DNA structural variations produced by actinomycin and distamycin as revealed by DNAase-I footprinting. Nucleic Acids Res 12:9271–9285 4. Van Dyke MW, Hertzberg RP, Dervan PB (1982) Map of distamycin, netropsin, and actinomycin binding-sites on heterogeneous DNA–DNA cleavage-inhibition patterns with methidiumpropyl-EDTA.Fe(II). Proc Natl Acad Sci USA 79:5470–5474 5. Van Dyke MW, Dervan PB (1983) Chromomycin, mithramycin, and olivomycin binding-sites on heterogeneous deoxyribonucleic-acid–Footprintingwith(MethidiumpropylEDTA)Iron(II). Biochemistry 22:2373–2377 6. Carpenter ML, Marks JN, Fox KR (1993) DNA-sequence binding preference of the GC-selective ligand mithramycin – Deoxyribonuclease-I/deoxyribonuclease-II and hydroxy-radical footprinting at CCCG, CCGC, CGGC, GCCC and GGGG flanked by (AT)n and An⋅Tn. Eur J Biochem 215:561–566 7. Low CML, Drew HR, Waring MJ (1984) Sequence-specific binding of echinomycin to DNA – evidence for conformational-changes affecting flanking sequences. Nucleic Acids Res 12:4865–4879 8. Van Dyke MW, Dervan PB (1984) Echinomycin binding-sites on DNA. Science 225:1122–1127 9. Fox KR, Waring MJ (1986) Nucleotidesequence binding preferences of nogalamycin investigated by DNAse-I footprinting. Biochemistry 25:4349–4356 10. Abu-Daya A, Brown PM, Fox KR (1995) DNA sequence preferences of several AT-selective minor groove binding ligands. Nucleic Acids Res 23:3385–3392 11. Trauger JW, Dervan PB (2001) Footprinting methods for analysis of pyrrole-imidazole polyamide/DNA complexes. Methods Enzymol 340:450–466 12. Mrksich M, Parks ME, Dervan PB (1994) Hairpin peptide motif. A new class of oligopeptides for sequence-specific recognition in the minor groove of double-helical DNA. J Am Chem Soc 116:7983–7988

13. Trauger JW, Baird EE, Mrksich M, Dervan PB (1996) Extension of sequence-specific recognition in the minor groove of DNA by pyrrole-imidazole polyamides to 9–13 base pairs. J Am Chem Soc 118:6160–6166 14. Chandler SP, Fox KR (1996) Specificity of antiparallel DNA triple helix formation. Biochemistry 35:15038–15048 15. Gowers DM, Bijapur J, Brown T, Fox KR (1999) DNA triple helix formation at target sites containing several pyrimidine interruptions: Stabilization by protonated cytosine or 5-(1-propargylamino)dU. Biochemistry 38: 13747–13758 16. Rusling DA, Powers VEC, Ranasinghe RT, Wang Y, Osborne SD, Brown T, Fox KR (2005) Four base recognition by triplexforming oligonucleotides at physiological pH. Nucleic Acids Res 33:3025–3032 17. Fox KR, Waring MJ (1987) The use of micrococcal nuclease as a probe for drug-binding sites on DNA. Biochem Biophys Acta 909:145–155 18. Van Dyke MW, Dervan PB (1982) Footprinting with MPE.Fe(II) – Complementary-strand analyses of distamycin-binding and actinomycin-binding sites on heterogeneous DNA. Cold Spring Harbor Symp Quant Biol 47: 347–353 19. Van Dyke MW, Dervan PB (1983) Methidiumpropyl-EDTA.Fe(II) and DNase-I footprinting report different small moleculebinding site sizes on DNA. Nucleic Acids Res:11:5555–5567 20. Spassky A, Sigman DS (1985) Nuclease activity of 1, 10-phenanthroline copper-ion – conformational-analysis and footprinting of the lac operon. Biochemistry 24:8050–8056 21. Nielsen PE, Hiort C, Holst Sönnichsen S, Buchardt O, Dahl O, Nordén B (1992) DNA binding and photocleavage by uranyl(VI) (UO 22+) salts. J Am Chem Soc 114: 4967–4975 22. Nielsen PE (1992) Uranyl photofootprinting of triple helical DNA. Nucleic Acids Res 20:2735–2739 23. Churchill MEA, Hayes JJ, Tullius TD (1990) Detection of drug binding to DNA by hydroxyl radical footprinting. Relationship of distamycin binding sites to DNA structure and positioned nucleosomes on 5S RNA genes of Xenopus. Biochemistry 29: 6043–6050 24. Cons BMG, Fox KR (1989) High-resolution hydroxyl radical footprinting of the binding of

172

25. 26. 27. 28. 29.

30. 31.

32. 33.

34.

35.

36.

Cardew and Fox mithramycin and related antibiotics to DNA. Nucleic Acids Res 17:5447–5459 Jain SS, Tullius TD (2008) Footprinting protein–DNA complexes using the hydroxyl radical. Nat Protoc 3:1092–1100 Drew HR, Travers AA (1984) DNA structural variations in the Escherichia-coli tyrT-promoter. Cell 37:491–502 Drew HR (1984) Structural specificities of 5 commonly used DNA nucleases. J Mol Biol 176:535–557 Lahm A, Suck D (1991) DNase I-induced DNA conformation. 2Å structure of a DNase I-octamer complex. J Mol Biol 222:645–667 Suck D, Oefner C (1986) Structure of DNase-I at 2.0Å resolution suggests a mechanism for binding to and cutting DBA. Nature 321:620–625 Suck D, Lahm A, Oefner C (1988) Structure refined to 2Å of a nicked DNA octanucleotide complex with DNase-I. Nature 332:464–468 Weston SA, Lahm A, Suck D (1992) X-ray structure of the DNase I-d(GGTATACC)2 complex at 2.3Å resolution. J Mol Biol 226:1237–1256 Herrera JE, Chaires JB (1994) Characterization of preferred deoxyribonuclease-I cleavage sites. J Mol Biol 236:405–411 Fox KR, Howarth NR (1985) Investigations into the sequence-selective binding of mithramycin and related ligands to DNA. Nucleic Acids Res 13:8695–8714 Brown PM, Madden CA, Fox KR (1998) Triple-helix formation at different positions on nucleosomal DNA. Biochemistry 37: 16139–16151 Brown PM, Fox KR (1999) DNA triple-helix formation on nucleosome core particles – Effect of length of the oligopurine tract. Eur J Biochem 261:301–310 Osborne SD, Powers VEC, Rusling DA, Lack O, Fox KR, Brown T (2004) Selectivity and affinity of triplex-forming oligonucleotides containing 2¢-aminoethoxy-5-(3-aminoprop1-ynyl)uridine for recognizing AT base pairs in duplex DNA. Nucleic Acids Res 32:4439–4447

37. Bishop KD, Borer PN, Huang Y-Q, Lane MJ (1991) Actinomycin D induced DNase I hypersensitivity and asymmetric structure transmission in a DNA hexadecamer. Nucleic Acids Res 19:871–875 38. Huang YQ, Rehfuss RP, Laplante SR, Boudreau E, Borer PN, Lane MJ (1988) Actinomycin-D induced DNAase-I cleavage enhancement caused by sequence specific propagation of an altered DNA-structure. Nucleic Acids Res 16:11125–11139 39. Abu-Daya A, Fox KR (1997) Interaction of minor groove binding ligands with long AT tracts. Nucleic Acids Res 25:4962–4969 40. Waterloh K, Fox KR (1991) The effects of actinomycin on the structure of dAn.dTn and (dA-dT)n regions surrounding its GC binding site. A footprinting study. J Biol Chem 266:6381–6388 41. Lavesa M, Fox KR (2001) Preferred binding sites for [N-MeCys(3), N-MeCys(7)] TANDEM determined using a universal footprinting substrate. Anal Biochem 293: 246–250 42. Hampshire AJ, Fox KR (2008) Preferred binding sites for the bifunctional intercalator TANDEM determined using DNA fragments that contain every symmetrical hexanucleotide sequence. Anal Biochem 374:298–303 43. Oyama F, Kikuchi R, Omori A, Uchida T (1988) Avian-Myeloblastosis virus reversetranscriptase is easier to use than the klenow fragment of DNA-polymerase-I for labeling the 3¢-end of a DNA fragment. Anal Biochem 172:444–450 44. Fox KR, Waring MJ (2001) High-resolution footprinting studies of drug–DNA complexes using chemical and enzymatic probes. Methods Enzymol 340:412–430 45. Goodisman J, Rehfuss R, Ward B, Dabrowiak JC (1992) Site-specific binding constants for actinomycin D on DNA determined from footprinting studies. Biochemistry 31: 1046–1058 46. Goodisman J, Dabrowiak JC (1992) Structural changes and enhancements in DNase I footprinting experiments. Biochemistry 31: 1058–1064

Chapter 11 Methods to Characterize the Effect of DNA-Modifying Compounds on Nucleosomal DNA Vidya Subramanian, Robert M. Williams, Dale L. Boger, and Karolin Luger Abstract Eukaryotic DNA forms a complex with an equal mass of proteins to form chromatin. To fully understand the action of DNA-reactive antitumor antibiotics in the cell, their effect must be studied in a chromatin context. In particular, it is of interest to investigate how the distortion of DNA, in the context of a nucleosome, affects the action of drugs with either monoalkylation or crosslinking activity, and how modified DNA is assembled into chromatin. Here, we present experimental approaches that allow one to compare the effect of such drugs on free DNA and nucleosomes. We find significant differences that likely arise from the different geometry of nucleosomal DNA compared to free DNA and also find that drug-mediated DNA crosslinking affects nucleosome assembly. Key words:  Nucleosome, Chromatin, Alkylation, Crosslinking, Mitosene, Duocarmycin

1. Introduction Many antitumor antibiotics function through the covalent modification of DNA, particularly affecting rapidly dividing cancer cells. Several agents modify DNA via interstrand crosslinks or by intercalation between nucleotide bases, while others form covalent monoadducts that do not result in interstrand crosslinks (ISLs) (1). In many cases, the mechanism of action of these drugs is tested in various in vitro assays using free DNA as substrate. This can be misleading since the majority of eukaryotic DNA is packaged into nucleosomes, resulting in profound effects on DNA conformation and accessibility. Here, we present assays that allow a comparison of how free and nucleosomal DNA interacts with DNA damaging agents, and K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_11, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

173

174

Subramanian et al.

how DNA damaging agents may affect chromatin assembly. In our example, we study representatives of two classes of DNA-modifying agents: (1) Mitosene-based drugs – (FK317 (4) and NVOC-derivative (5), Fig. 1A): these are bifunctional DNA alkylating agents capable of DNA crosslinks and monoadducts (2); and (2) duocarmycins (DSAs) (Fig.  1C), a class of sequence-selective DNA alkylating agents (3). 1.1. Mitosene-Based Bioreductive Alkylating Drugs

Bioreductive alkylating agents include families of bifunctional compounds such as FR900482 and FK317 that are capable of monoalkylating DNA, eventually resulting in interstrand DNA crosslinks under highly reducing conditions. A representative of clinically significant drug is Mitomycin C (MMC 1, Fig. 1A) (4) which has been used in the clinic for over 30 years as an antitumor compound (5–7). FR900482 (Fig. 1A; (2)), a natural metabolite obtained from Streptomyces sandaensis No. 6897, is structurally and functionally related to MMC (8). Most recently, a semisynthetic derivative of FR900482-FK317 ((4); Fig.1A) has exhibited promising antitumor activity and has advanced from Phase I to Phase II human clinical trials. The activation of these compounds is mainly via a reduction pathway resulting in a bis-electrophilic mitosene moiety which results in the formation of ISLs as well as monoadducts specifically at 5¢CpG3¢ steps (Fig. 1B) (5, 9, 10). Recently, a synthetic mitosene-progenitor (5) was developed from FR900482 referred to in this study as the compound 5. This derivative is activated photochemically rather than via the reductive pathway typical of bioreductive agents. This derivative was designed to overcome the rate-limiting step, namely the formation of the reactive bis-electrophilic mitosene moiety (Fig. 1B). The functional and sequence-specific properties of this derivative, however, remain similar to that of its parent compounds (2). In this study, we use FK317, which is activated under highly reducing conditions and compound 5, which is photoactivated. This allows us to determine the method of activation that is most effective for DNA crosslinking.

1.2. Duocarmycin – Sequence-Selective Alkylating Agents

DSA belongs to this family of naturally occurring antitumor antibiotics (Fig. 1C) (11). The sequence selectivity is achieved through the noncovalent binding specificity of the agent. DNA-binding induces a conformational change in the agent that makes it amenable to nucleophilic attack (3, 12). This conformational change is dependent on the shape of the minor groove and the length of an AT-rich sequence (12). The result of the nucleophilic attack is an efficient DNA alkylation reaction via the N3 adenine and the cyclopropane moiety (Fig.  1D). Recent studies have shown that the natural enantiomer DSA is capable of alkylation of B form as well as nucleosomal DNA (13). The same study also performed modeling analysis which revealed that DSA remains unaltered in its ability

175

Methods to Characterize the Effect of DNA-Modifying Compounds on Nucleosomal DNA

a O

OCONH2

H2N

OMe N

Me

NH

R1

O

1, mitomycin C (MMC)

OH

OCONH2

OR2

OR N

H NCO2Me

NR3

O

O

MOMO

4

MeO2C

2, FR900482 R1=CHO, R2=R3=R4=H 3, FR66979 R1=CH2OH, R2=R3=R4=H 4, FK317 R1=CHO, R2=Me, R3=R4=Ac

H

N NVOC

O

5, NVOC =

NO2 O OMe OMe

5' GpG

5' CpG

OH

c

b

a

b

5' GpC

~3.4 A

R1

NH

N

R2 R3

mitosene core

4.6 A

4.2 A

3.1 A

c

MeO2C Me HN

MeO2C HN N

O

OMe

O

N H

OMe OMe

O

O

OMe

N O

N H

OMe OMe

(+)–duocarmycin A

(+)–duocarmycin SA d

O dR

N

N

H

O

H+ H N

OMe

N

N N

OMe

N dR

OMe NH

N MeO2C

O N H

OH

Fig. 1. Bioreductive alkylating agents and sequence-selective alkylating agents. (A) Structures of Mitomycin C, FR900482 and cogeners. Compound is the photoactivated mitosene progenitor. (B) Geometrical basis for the mitosene-based crosslinking specificity. (C) Molecular structures of (+)-duocarmycin SA (DSA) and (+)-duocarmycin A. (D) Schematic representation of the alkylation of the N3 adenine (A19) by DSA.

176

Subramanian et al.

to carry out sequence-specific monoalkylation of nucleosomal DNA. Since DSA is an obligate mono-alkylating agent (13), it allows us to demonstrate the ability of our techniques to resolve interstrand crosslinks from monoalkylation patterns. 1.3. Nucleosomal Versus Free DNA

Most of the structural, functional and chemical insights into these types of compounds are based on their effect on protein-free B form DNA. Approximately 80% of eukaryotic DNA exists in the form of chromatin, in which DNA is wrapped tightly around a protein core of eight histones to form nucleosomes (14). From the high-resolution X-ray structure of the nucleosome, it is obvious that the conformation of nucleosomal DNA deviates significantly from that of free DNA (15). It is highly bent and distorted, with dramatic implication on groove widths (14) and, in many instances, on base pairing interactions and local DNA geometry. Many parameters, although in most cases still within the parameters of B-form DNA, exhibit characteristic fluctuations (15) that likely affect the interaction of nucleosomal DNA with DNA damaging compounds, with potential effects on the outcome of the interaction. DNA crosslinking agents such as psoralen are restricted to the nucleosome-free linker region, with only limited activity in the nucleosome core DNA (16–18). Cisplatin, another DNA intercalating/crosslinking agent with clinical applications, also favors internucleosomal linker DNA over nucleosomal DNA (19, 20). A previous study demonstrated that an increase in the acetylation state of chromatin resulted in a corresponding increase in the amount of cisplatin-nucleosomal DNA adducts in a concentration-dependent manner (21). For other DNA crosslinking agents, such as nitrogen mustard and melphalan, both interstrand crosslinks and monoalkylation were similar in free and nucleosomal DNA. The same study revealed that the efficiency of mitomycin C (MMC)-mediated crosslinks was reduced at the nucleosomal dyad and more pronounced near the ends (22). MMC-mediated monoalkylation was not monitored.

1.4. Overview and Versatility of the Approaches Described Here

Here, we describe techniques that allow visualization of the bifunctional activity of DNA modifying agents – monoalkylation and interstrand crosslinking – on nucleosomal DNA and free DNA. We have adapted previously published methods to investigate the effect of DNA modification on in vitro chromatin assembly, as deficiencies in this essential activity may also contribute to apoptosis of rapidly dividing cancer cells. Quantitative Agarose Gel Electrophoresis (QAGE)/Multigel analysis is used to quantify the number of nucleosomes assembled, to test whether any of the compounds affects chromatin assembly (23–25). This method was previously used to determine the difference in surface charge density of linear versus assembled DNA to measure the actual

Methods to Characterize the Effect of DNA-Modifying Compounds on Nucleosomal DNA

177

number of nucleosomes on a DNA template (in the presence of a known control – T3 phage) (23–25). By carrying out the assembly on untreated DNA and drug-treated DNA (the latter after thorough characterization as described below), one can discern differences in the number of nucleosomes assembled and relate these to the type of DNA damage introduced.

2. Materials 2.1. Drug Stock Solutions

1. NVOC-derivative (compound 5) was synthesized as mentioned in (26). A stock of 10 mg/ml was made with the compound in 100% DMSO and stored in a light safe box at −80°C as single use aliquots. Appropriate dilutions from this stock were made fresh in 100% DMSO just before use. 2. FK317 was a generous gift from Fujisawa Pharmaceutical Co., Japan (27). A stock of 10 mg/ml was made in 90% DMSO and stored in a light safe box at −80°C as single use aliquots. Appropriate dilutions from this stock were made fresh in 90% DMSO just before use. 3. DSA was a generous gift from Dr. Dale L. Boger (The Scripps Research Institute). A 5  mg/ml stock of DSA was made in 100% DMSO and stored in a light safe box at −80°C as single use aliquots. Appropriate dilutions are made fresh in 100% DMSO just before use.

2.2. Nucleosome Preparation

Purification of DNA – a palindromic 146 bp DNA fragment derived from human a-satellites (28) and a 146  bp DNA fragment derived from the 5 S DNA gene of Xenopus borealis (29) was conducted as described (28). Recombinant histones were purified, refolded to octamers and subsequently used to reconstitute nucleosomes as described in (28). The nucleosomes were then incubated for 1 h at 37°C and analyzed for their homogeneity and integrity on 5% native PAGE gels using 0.2× Tris–Borate EDTA (TBE) as the running buffer (28). If appropriate, free DNA was removed by preparative gel electrophoresis (21). The use of recombinant histones and purified DNA with strong positioning sequences ensures a single stable nucleosome species, in terms of its positioning on the DNA after heat treatment. This is important for subsequent interpretation of crosslinking and monoalkylating reactions.

2.3. DNA Crosslinking Reactions

1. Ultraviolet lamp (Rayonet) containing 350 nm bulbs.

2.3.1. Crosslinking Reaction with B Form DNA and Nucleosomal DNA

3. Proteinase K: A stock solution of 2.5  mg/ml made in Tris– EDTA (TE) – 10 mM Tris–HCl pH 7.5, 0.1 mM EDTA.

2. 100 mM DTT stock freshly made before use.

4. 4 M NaCl.

178

Subramanian et al.

5. 20 mg/ml glycogen. 6. Phenol chloroform isoamylalcohol (25:24:1 v/v/v). 7. 3 M sodium acetate pH 5.2 8. 100% ethanol 9. To ensure nucleosomes are intact subsequent to incubation with the various DNA-modifying agents, the nucleosomes are analyzed on a 5% native gel as mentioned in (28). 2.3.2. Visualization of DNA Crosslinking Reactions

1. 10× alkaline agarose gel electrophoresis buffer: 500  mM NaOH, 10 mM EDTA. 2. 6× alkaline gel loading buffer: 300 mM NaOH, 6 mM EDTA, 18% sucrose, 0.15% bromophenol blue and 0.25% xylene cyanol. 3. Neutralization buffer: 1 M Tris–HCl pH 7.6, 1.5 M NaCl. 4. SyBr Gold (Invitrogen, Molecular Probes). A 1:10,000 dilution of SyBr gold was made in 500 ml of Tris–Acetate EDTA (TAE) buffer. The stain was made in a plastic container and stored in the dark (see Note1).

2.4. Analysis of Monoalkylation

1. [g32-P] ATP: Stock concentration of 3,000–7,000 Ci/mmole. 2. T4 polynucleotide kinase and reaction buffer.

2.4.1. DNA Radiolabeling Reaction 2.4.2. Eight Percent Denaturing PolyAcrylamide Gel Electrophoresis Analysis

1. SequaGel (National Diagnostics) Denaturing DNA solutions – 19:1 Acrylamide/bisacrylamide solution. 2. Ammonium persulfate: Prepare 10% solution in water and store at 4°C. 3. N,N,N,N¢ Tetramethyl-ethylenediamine (TEMED).

2.4.3. Formamide Dye

This recipe contains 0.005 g xylene cyanol, 0.005 g bromophenol blue, 400  ml of 0.5  M EDTA pH 8.0, 100  ml of nanopure water and 9.5 ml of ultrapure formamide. This solution is stored at −20°C. The same dye is also used for loading transcript generated from in vitro transcription reactions.

2.5. In vitro Chromatin Assembly

1. Native Drosophila core histones, recombinant ACF1 and drosophila Nucleosome Assembly Protein 1 (dNap1) were purified as described (30, 31). Dilutions for ACF1 are made in 1× assembly buffer and for dNap1 and core histones in Buffer R (recipe provided below).

2.5.1. Purification of Various Chromatin Assembly Components

2. Plasmid template: p4TxRE/G-less plasmid was prepared using Qiagen Mega plasmid preparation and stored at a concentration of 500 mg/ml at −20°C in triple-use aliquots (see Note 2) (32).

Methods to Characterize the Effect of DNA-Modifying Compounds on Nucleosomal DNA

179

3. Buffer R: 10  mM HEPES pH 7.5, 10  mM KCl, 0.5  mM EDTA, 10% v/v glycerol, 10 mM – glycerophosphate, 1 mM DTT and 0.2 mM phenymethylsulfonyl fluoride (PMSF). 4. 2× HEG buffer (Preincubation buffer for Nap1 and core histones): 50 mM HEPES pH 7.6, 0.1 mM EDTA pH 8.0, 10% glycerol. 5. 10× Assembly buffer: 100 mM HEPES pH 7.6, 500 mM KCl, 50 mM MgCl2, 50% glycerol, 10% PEG, 30 mM ATP, 0.1% NP40. 6. ATP regeneration system: Stocks of 3  M phosphocreatinine (PC) and 100  mg/ml creatinine phosphokinase (CPK) were prepared in the recommended buffers. Dilutions of the various components of the ATP regeneration system were made in 1× assembly buffer. Diluted samples of CPK cannot be reused. 2.5.2. Qualitative Analysis of Chromatin Assembly: Micrococcal Nuclease Digestion

1. Micrococcal nuclease (MNase) stocks were made in assembly buffer at a concentration of 500 U/ml and stored at −80°C in 50 ml aliquots. 2. Assembly/Transcription Stop Buffer: 20 mM EDTA pH 8.0, 0.2 M NaCl, 1% (w/v) SDS and 0.25 mg/ml glycogen (Sigma Aldrich) (see Note 3). 3. 1 M Calcium chloride [CaCl2] freshly made. 4. Proteinase K stocks were made as above. 5. 4 M ammonium acetate for precipitation of DNA after MNase digestion. 6. TBE running buffer: 90  mM Tris–borate and 2  mM EDTA pH 8.0. 7. 1.2% agarose gel is made in 1× TBE. 8. The gel size for this assay is 15 × 20 cm and 0.5 cm in thickness (150 ml volume of agarose solution). The electrophoresis unit used was the Subcell GT Systems from Biorad. 9. Agarose staining solution was made in 1× TBE containing 1:10,000 dilution of SyBr Gold (Invitrogen, Molecular Probes).

2.5.3. Qualitative Analysis of Chromatin Assembly – Qualitative Agarose Gel Electrophoresis

1. Multigel set up containing the gel bed, slot former, appropriate 18-lane comb, grooves, plates and buffer tank (25) (Fig. 4B). 2. Agarose for Multigel analysis (Molecular biology grade, low Electro Endo Osmosis-EEO, Research Organics). 3. Kimble borosilicate tubes for preparation of the agarose stocks and the appropriate dilutions of the stock.

180

Subramanian et al.

4. Water bath set at 60–70°C (Blue M constant temperature shaker bath). 5. Cole-Parmer oscillating pump used with a steel-cased variable voltage controller run at about 40–50% maximum output voltage, 120 V. 6. TAE electrophoresis buffer: 40  mM Tris–acetate, 1  mM EDTA. 7. Intact T3 phage (generous gift from Dr. Philip Serwer, University of Texas-San Antonio).

3. Methods Nucleosomes can be stored for over a month on ice or at 4°C. Plasmid and DNA fragments used for nucleosome preparation can be kept frozen at −20°C. Assembly components – ACF, dNap1 and CPK are all stored in 50% glycerol at −80°C. The plasmid DNA template that is used for in  vitro chromatin assembly is stored in TE at −20°C in suitable aliquots (31). All other components (i.e., gels, buffers) should be made freshly where indicated. 3.1. Compound 5-Mediated Crosslinking of Free and Nucleosomal DNA Substrates 3.1.1. Crosslinking Reaction

1. The final volume of each reaction is 20 ml, containing 6.25 mM free DNA (5  S and alpha satellite) or 12.5  mM nucleosome core particles (NCP). The reaction must be carried out specifically in clear Eppendorf tubes, in order to enable effective UV photoactivation. 2. Compound 5 is added to the reaction in increasing molar ratios. Dilutions of compound 5 are made in 100% DMSO. Appropriate controls such as single-stranded (ss) and doublestranded (ds) 146 bp DNA are used to ensure that the alkaline agarose gel is indeed capable of denaturing dsDNA. DNA (both unassembled and assembled) treated with DMSO alone is also used as a control to determine the presence of any background due to the presence of the organic solvent. 3. The reactions are irradiated with UV lamps (Rayonet lamps containing a wavelength of 350  nm) for 1  h in the dark at room temperature. In order to ensure that the compound 5 treated-nucleosomes are still intact after treatment, the samples are run on a 5% native gel (28). Dissociation of nucleosomes and release of free DNA can skew data interpretation with respect to the ability of the drug to modify nucleosomal DNA versus free DNA. For this reason, nucleosome samples that are contaminated with free DNA (as visualized on 5% native gels) should not be used in this assay.

Methods to Characterize the Effect of DNA-Modifying Compounds on Nucleosomal DNA

181

4. The UV-activated, drug-treated free DNA samples are ethanol precipitated (added 1/10th the volume of 3 M sodium acetate pH 5.2 and 2.5 times the volume of 100% cold ethanol) at −20°C overnight. This ensures removal of excess drug and allows for recovery of the treated substrates. 3.1.2. Recovery of Drug-Treated Nucleosomal DNA and Analysis of Drug-Treated DNA Substrates

1. For nucleosomal DNA, removal of excess drug is carried out by addition of 5.2  ml of 4  M NaCl, 4  ml of 20  mg/ml of ­glycogen and 60 ml of 100% ethanol to 20 ml of the crosslinking reaction mixture. Incubate at −20°C overnight. DNA is ­recovered by spinning the samples at 16,363 × g at 4°C for 30 min. 2. The recovered nucleosomal DNA is subjected to Proteinase K treatment (for histone protein removal) in 100 ml of T.E and 0.01 mg/ml Proteinase K. This mixture is incubated at 37°C for 1 h. Proteins are separated from the DNA through phenol chloroform extraction (1:1 v/v of phenol:chloroform/isoamylalcohol to sample volume) and ethanol precipitation (as above). 3. All samples are analyzed on a 1.4% alkaline agarose gel (33). The gels are run in 1× alkaline agarose electrophoretic buffer at 75 V for 50–55 min and stained with SyBr Gold solution. The band intensities are quantified using the Image Quant Analysis (version 5.1) software, or similar analysis software. 4. The efficiency of crosslinking is calculated as the ratio of the intensity of crosslinked band to the total intensity (sum of the intensities of the crosslinked and uncrosslinked bands). A representation of the percentage of crosslinking as a function of the fold excess of the drug used for both free DNA and nucleosomes is particularly useful for comparison of crosslinking efficiency between different DNA substrates and different drug derivatives (34).

3.2. FK317-Mediated DNA Crosslinking Reaction with B-Form and Nucleosomal DNA

FK317 differs from compound 5 in its activation mechanism. The appropriate volume of FK317 needed for each titration is first incubated with 5  mM DTT (in a small final volume) at room temperature for 1 h. This is the activation step for FK317. The substrates for the crosslinking reactions are then added to the activated reaction in concentrations, as mentioned above. All subsequent steps for analyses are as described in the Subheading 3.1. As seen in Fig. 2E (quantification of Fig. 2A–D), there is not only a difference in FK317-mediated crosslinking efficiency between the free 5  S and a-satellite DNA substrates, but also between free and nucleosomal DNA.

182

Subramanian et al.

a

Increasing molar excess of FK317

b

Increasing molar excess of FK317 FK317

FK317

1

2

3

4

5

6

1

2

3

4

5

6

7

α DNA

5S DNA

c

1

2

3

4

5

6

7

8

d

1

2

3

4

5

6

7

8

α nDNA

e

f

Percentage Cross-linking

5SnDNA

14 12 5S DNA : FK317

10 8

5S NucDNA : FK317

6

Alpha NucDNA : FK317

4

Alpha DNA : FK317

2 0

10 20 30 40 50 60 70 80 Molar Excess of FK317 (Fold excess)

Increasing molar excess of DSA

DSA 1

2

3

4

5

6

7

8

α nDNA

Fig. 2. Alkaline agarose gel electrophoresis-DNA crosslinking assay. (A) and (B) represent denaturing alkaline agarose gels of 5 S and a-satellite DNA treated with increasing ratios of FK317. Lane 1 in (A) and (B) represents untreated DNA, while lanes 2–7 represents a 1:1, 1:5, 1:10, 1:20, 1:40, 1:60 and 1:80 ratios of DNA to FK317. (C) and (D) represent denaturing alkaline agarose gels of 5 S and a-satellite nucleosomal DNA (nDNA) treated with increasing ratios of FK317. Lane 1 in (C) and (D) represents untreated DNA, while lanes 2–8 represents a 1:1, 1:5, 1:10, 1:20, 1:40, 1:60 and 1:80 ratios of nucleosomal DNA to FK317. The alkaline agarose gels were stained using Sybr Gold and visualized and the intensity of the crosslinked and uncrosslinked bands were determined using the ImageQuant v 5.1 software. The migration of the various species is depicted along the sides of the gels. (E) Quantitative representation of the percentage efficiency of FK317 mediated crosslinking on 5 S and a-satellite B form and nucleosomal DNA as a function of fold excess of FK317. The efficiency of crosslinking was determined as described above and results were plotted with drug-mediated percentage crosslinking as a function of fold/molar excess of FK317 over nucleosomal or free DNA (molar excess). The error bars are representative of deviations of three independent crosslinking experiments from the average. (F) Represents the effect of increasing amount of DSA (obligate monoalkylating agent) on a-nucleosomal DNA. Lane 1 represents untreated a DNA, lane 2 represents untreated nucleosomal a DNA while lanes 3–8 represents a 1:1, 1:10, 1:20, 1:40, 1:60 and 1:80 ratios of NCP to DSA.

Methods to Characterize the Effect of DNA-Modifying Compounds on Nucleosomal DNA

183

3.3. DSA-Mediated Alkylation of Nucleosomal DNA

The alkylating reactions are carried out in a final volume of 20 ml with 12.5  mM alpha-satellite DNA-containing nucleosomes in 100 mM NaCl, 10 mM Tris–HCl pH 7.5 and 1 mM EDTA pH 8.0. DSA is appropriately diluted in 100% DMSO and added in sequential molar excess to the nucleosomes. The reactions are incubated at room temperature for 1 h or for 24 h at 4°C. The excess drug is removed, as described above, and the samples thus recovered were analyzed the same as above. Analysis by alkaline agarose gel electrophoresis is effective in determining DNA crosslinking efficiency in a semiquantitative manner. For example, it is clear from Fig. 2A and B and Fig. 2C and D that nucleosomal DNA is more resistant to crosslinks that B-form DNA. Figure 2E further confirms that 5 S DNA is a better substrate than a-satellite DNA due to a higher number of target sites. Alkaline agarose electrophoresis is limited, in that it does not have the resolution to distinguish between DNA monoadducts on single stranded DNA. This is apparent for DSA (a monoalkylating agent), where no bands corresponding to crosslinked DNA or other DNA adducts is detected (Fig. 2F). To discern the presence of monoadducts on single stranded DNA, 8% denaturing PAGE can be utilized.

3.4. Analysis of Drug-Mediated Monoalkylation

1. Free and nucleosomal DNA are treated and recovered as described above. The treated and untreated (control) DNA substrates are radiolabeled in a final reaction volume of 50 ml containing 50 pmol of DNA, 50 pmol of [g32-P] ATP (stock concentration of 3,000–7,000 Ci/mmol) and 10 units of T4 polynucleotide kinase (New England Biolabs). 2. The excess enzyme and radiolabel is removed by phenol chloroform extraction and ethanol precipitation, as mentioned above (1.b.i-ii) (see Note 4 for alternative method to remove excess radiolabel). 3. The samples are loaded onto an 8% denaturing PAGE. 30 ml of the PAGE gel is prepared by mixing 9.6 ml of the SequaGel gel concentrate, 3  ml of the SequaGel 10× Denaturing PolyAcrylamide Gel Electrophoresis (DPAGE) buffer, 17.5 ml of the SequaGel diluent, 12 ml of TEMED and 240 ml of 10% APS, as per the manufacturer’s instructions. The dimensions for the gels used in the monoalkylation analyses were 20 × 20 cm with 0.4  mm thick spacers. The gel is prerun at 30  W for 30 min. This ensures that the gel is heated up to a temperature of ~50°C which is essential for efficient denaturation of doublestranded DNA. The samples are boiled in formamide buffer at 95°C for 10 min. The wells are rinsed with running buffer and the boiled samples are loaded. The gels are run at 30 W for 40  min in 1× TBE. The gels are dried and exposed to a Phosphoimager Exposure cassette (Molecular Dynamics) for

184

Subramanian et al.

an appropriate amount of time (between 2 and 15  h). Radioactivity is visualized using a Storm® (Molecular Dynamics) and analyzed with the Image Quant (version 5) software. This technique enables resolution of the various drug-mediated monoadducts on single-stranded DNA as well as on crosslinked DNA products. Together, the two electrophoresis approaches are instructive on the outcome of drug treatment of nucleosomal and free DNA. For example, Fig.  3A demonstrates that while a 1

α DNA 2 3 4

α nDNA 5 6 7 5‘

3‘ N

3‘

5‘

5‘

3‘

Positional Isomers

N

3‘

5‘

5‘ N

Monoadduct 3‘

b

1

2

α nDNA 3

4

Alkylated products

ssDNA Fig. 3. DPAGE analysis for visualization of drug-mediated monoalkylation. (A) DPAGE analysis on drug-treated nucleosomal and free a-satellite DNA. Lane 1: Free a-DNA, lane 2: Crosslinked free a-DNA control (The crosslinked DNA was purified from free single stranded DNA in an alkaline gel and radiolabeled). The two different orientation isomers are illustrated at the side of the gel. Lane 3 and 4: Free a-DNA treated with a 20 and 40 molar excess respectively of NVOC to DNA. Lane 5: Nucleosomal a-DNA treated with Proteinase K and hence free of protein content, lane 6 and 7: Nucleosomal DNA treated with 20 and 40 molar excess of NVOC, compound 5 respectively. From (34), with permission. (B) DPAGE analysis of a-nucleosomal DNA with DSA. Lane 1: a-nucleosomal DNA (anDNA) treated with Proteinase K, lanes 2–4: a-nucleosomal DNA treated with 10, 20 and 40 molar excess of DSA respectively.

Methods to Characterize the Effect of DNA-Modifying Compounds on Nucleosomal DNA

185

a-satellite DNA is capable of crosslinks, a-satellite nucleosomal DNA (anDNA) shows a reduced ability to be crosslinked but a higher propensity to get monoalkylated. DSA, as expected, causes no crosslinks (Fig.  2F), but a well-defined pattern of monoalkylated bands on anDNA (Fig. 3B). 3.5. Qualitative and Quantitative Analysis of In Vitro Chromatin Assembly

The effect of DNA modifying agents on chromatin assembly has not been well studied. The assembly of nucleosomes on DNA forms a very important checkpoint during cell cycle progression (from the S to the M phase). The decrease in nucleosome assembly mediated by other factors (35) has been shown to result in cells arrested at the S phase, which subsequently results in apoptosis (36). In our study, we find a 22% decrease in the nucleosome occupancy on crosslinked DNA mediated by compound 5, indicating that the crosslinked DNA is incapable of the flexibility required for histone deposition and subsequent nucleosome formation (34). In this study, we use an in vitro chromatin assembly reaction and quantitatively analyze the outcome of chromatin assembly on crosslinked DNA. The method can be used to study the effect of any DNA modifying agent on nucleosome assembly.

3.5.1. In vitro Chromatin Assembly

1. The various assembly components were made as in (30). The storage of the various assembly components is described in Subheading 3.1. 2. The assembly begins with preincubation of core histones with dNap1 on ice for 30 min in a final volume of 30 µl containing 1× HEG buffer. The reaction should contain at least 1.5 µg of purified Drosophila native core histones (0.11 nmol of each histone) and 12.012 µg of dNAP1 (see Note 5). The same assay can also be performed with recombinant histone octamer (31). 3. The assembly reaction is carried out in a final volume of 98 ml. The reaction includes 0.36 mg of ACF (1.285 pmol ACF peptide) in 1× assembly buffer, an ATP-regenerating system (30  mM phosphocreatinine and 1  mg/ml creatinine kinase) and 140  ng of BSA. The reaction volume also includes the preincubated dNap1-dimer complex (as above), and 2.1 mg of the p4TxRE plasmid DNA or any other plasmid DNA (untreated and treated) (31). The final assembly reaction is incubated at 27°C for a minimum of 4 h and a maximum of 12 h.

3.5.2. Qualitative Analysis of Chromatin Assembly – Micrococcal Nuclease Digestion

MNase digestion demonstrates whether the assembled DNA template contains evenly spaced nucleosomes along the plasmid template. This method is based on the principle that nucleosomal DNA is protected from DNA digestion (MNase) by virtue of the

186

Subramanian et al.

associated histones. If the nucleosomes are indeed evenly spaced, the digestion pattern of such a DNA template would resemble a definite ladder with each band separated from the next by a precise linker DNA length (Fig. 4A). This assay is therefore essential to determine the homogeneity of the chromatin assembly reaction. MNase cuts DNA between nucleosomes. Relatively precise nucleosome positioning results in a ladder with a certain number of “rungs” during the earlier digestion time points that depend on the number of nucleosomes occupied. In case of a partially occupied plasmid, the number of nucleosome rungs would be different at these earlier time points and there will be a difference in pattern when compared to the full occupancy plasmid. Since DNA crosslinking is resistant to MNase digestion, the inefficiency of MNase activity can lead to misinterpretation of the digestion patterns (37). Hence, this assay is used purely to check for the quality of the chromatin assembly on uncrosslinked DNA template and cannot be used for assemblies on crosslinked DNA templates. It is therefore necessary to back this assay with a more quantitative analysis that remains unbiased by the effects of DNA crosslinking. 1. To 70 ml of 10× assembly buffer, 593.8 ml of water and 3.62 ml of 1 M CaCl2 is added (1× AB + CaCl2). 135 ml of this solution is added to the 98 ml chromatin assembly reaction mixture (as prepared above). The mixture is incubated in a 37°C water bath. 2. The MNase stock (500 U/ml) is diluted to 1:25 v/v using the 1× AB + CaCl2 buffer and is incubated at 37°C. 7  ml of the diluted enzyme is added to assembly reaction above. 3. 60 ml of the assembly reactions with MNase is removed at 1, 2, 4 and 8 min and added to tubes containing 96 ml of transcription stop buffer and 9.6 ml of Proteinase K. 4. These reactions are then incubated at 37oC for at least 1  h. Following Proteinase K reaction, the mixture is subjected to phenol chloroform extraction and ethanol precipitation. 5. Once the ethanol is completely evaporated, the precipitate is resuspended in 10–15 ml of water. The resuspended sample is

Fig. 4. (continued ) with the various percentage of agarose indicated below their respective lanes. (vii) Represents the final set up which includes a pump that enables constant circulation of the running buffer thereby preventing the accumulation of ions during overnight electrophoresis. (c) A representative QAGE/Multigel analysis performed on NVOCtreated p4TxRE/chromatin assembled on crosslinked p4TxRE respectively. The numbers above each lane represents the agarose concentrations. (d) The QAGE results are represented quantitatively as Ferguson Plots. Intact T3 phage was used as a migration control to determine the unknown m0 values for unassembled and assembled DNA. These m0 values where then corrected for the electro osmotic nature of the buffer to finally determine the gel-free mobility (m0¢) (24). The values thus obtained from these plots were used to calculate the number of nucleosomes assembled. These values and the final nucleosome counts are listed in Table 1. A, C, D from (34), with permission.

Methods to Characterize the Effect of DNA-Modifying Compounds on Nucleosomal DNA

a

b i

p4TxRE p4TxRE+hu p4TxRE:40x NVOC

187

ii

1 2 3 4 5 6 7 8 9 101112131415

18-lane slot former, comb and electrophoresis unit iii

Chromatin Fragments

c

Framing gel setup iv

1.5% framing gel casted Framing gel with 18 lanes with slot former and comb 0.7 0.6 0.5 0.4 0.3 0.2 1.0 0.9 0.8 0.8 0.7 0.6 0.2 0.3 0.4 0.5 0.6 0.7

vi

v Intact T3 phage

Intact T3 phage

1

2 3

4

5

6

NVOC-treated p4TxTRE

9 10 11 12 13 14 15 16 17 18

Reset 18-lane comb for Specific percentage of casting single lanes agarose cast in each lane as indicated below

Assembly on NVOC-treated p4TxRE

10

vii

cm 2 / V/sec

d

7 8

Assembled p4TxRE –m

T3 + Assembled p4TxRE T3 + p4TxRE p4TxRE Assembled p4TxRE + NVOC T3 +Assembled p4TxRE + NVOC T3 + p4TxRE + NVOC p4TxRE + NVOC 1 0

0.2

0.4

0.6

0.8

1.0

1.2

Final electrophoresis set up includes circulating pumps for the running buffer

Percentage Agarose (%)

Fig. 4. Analysis of the in vitro chromatin assembly. (a) Micrococcal nuclease (MNase) analysis. The various chromatin templates are indicated above the lanes. This panel represents the MNase digestion on chromatin templates. Lanes 2–5, 7–10, 12–15 shows digestion times of 1, 2, 4 and 8 min. Represented along the gel are the various chromatin fragments/ intermediates that are generated at various time points during MNase digestion. (b) Quantitative Agarose Gel electrophoresis/Multigel analysis. Pictures representing the various steps involved in the preparation and running of a Multigel (Quantitative Agarose Gel Electrophoresis). (i) Shows the various components of the Multigel eletrophoresis setup (18lane slot former and comb specifically) while (ii) and (iii) shows the arrangement of the slot former and comb for casting of the 1.5% framing gel. (iv) and (v) show the formation of the 18-lane slots in the framing gel as well as the resetting of the 18-lane comb in the frame for the casting multi-percentage agarose. (vi) Shows a casted 18-lane multi-agarose gel

188

Subramanian et al.

then loaded with appropriate amount of DNA-loading dye onto a 1.2% TBE agarose gel. The gel is then run at 120 V for 3–4 h. The gel is stained with SyBr Gold solution and visualized using a Storm® gel and immunoblot imaging system (Molecular Dynamics) and analyzed using the Image Quant software (version 5.1). 3.5.3. Quantitative Analysis of Chromatin Assembly by Quantitative Agarose Gel Electrophoresis

The number of nucleosomes assembled on a plasmid DNA template can be determined by QAGE, which entails the measurement of the average surface charge density of assembled and unassembled DNA (23–25). 1. The apparatus and method has been described (23–25); (also see Fig. 4B). Briefly, a frame of 1.5% agarose contains narrow slots with gels of varying defined agarose concentration (0.2– 1% agarose; running gel) (23–25) (Fig. 4B v and vi). 2. The running and framing gels are cast in 1× TAE (40  mM Tris–acetate/1 mM EDTA pH 7.8). The agarose used in this experiment is low EEO from Research Organics (Fig. 4B). 3. Each lane is loaded with 0.8–2.1 mg of the chromatin assembly reaction along with 0.5–1.0 mg of intact T3 phage and a final concentration of 10% glycerol in each sample. 4. The running buffer is circulated throughout the experiment (using a Cole-Parmer oscillating pump). The temperature is maintained between 24 and 25°C. The gels are run at 48 V for 6 h (Fig. 4B.vii). 5. After electrophoresis, the gel is carefully lifted out and placed in the ethidium bromide staining solution for 1 h. 6. The gels are then destained in nanopure water for better contrast. 7. The stained gels photographed in the presence of a graduated scale and the migration (m) measured using the ruler function available in Scion Image (NIH software-O’Neill et.al., 1989). The standardization is done by first determining the number of pixels that correspond to a certain defined distance (in centimeters) on the graduated scale in the photographed image. This allows for further quantitation of distance/migration of the various assembled products. This calibration has to be done for every new image obtained. 8. In the example shown here, an 18-lane running gel frame was used, where the first 9 lanes (with varying agarose concentrations) were loaded with plasmid DNA (drug treated). The second 9-lane section of the gel contains the assembled DNA template (chromatin assembly on drug treated plasmid DNA) at various agarose concentrations. This ensures that the two samples – unassembled and assembled crosslinked plasmid

Methods to Characterize the Effect of DNA-Modifying Compounds on Nucleosomal DNA

189

DNA templates – are subjected to the same experimental conditions (Fig. 4C). 9. The migration of the various species (m) are calculated by measuring the distance (in centimeters) from the middle of the well to the center of the band of interest and plotted as a function of agarose concentration. The scatter plot is first fit to an exponential equation. This semilogarithmic plot is referred to as the Ferguson plot (24) and enables calculation of the true average surface charge density m0(Fig. 4C). The extrapolation of the migration distance for a species at 0% agarose gives the m¢0 gel free mobility value. This value contains migrations effects due to the electro-osmotic nature of the buffer. To obtain the true m0, the m¢0’ the value is corrected for electroosmotic effects of the buffer the EEO value for the buffer has to be subtracted from m¢0. The following equation is used: m0 = [R ´ (m0 of T3 phage + EEO)] –EEO R is m¢0 calculated from the Ferguson plot (semi logarithmic plot) for DNA/nucleosome assembly divided by the m¢0 for the T3 phage. The latter is a known value and should be around – 0.66 × 10−4 cm2/V/s (in 1× TAE) (23–25). The EEO of the buffer (in this case 1× TAE) is calculated to be around 0.1 × 10−4 cm2/V/s. The usual m0 values for DNA in 1× TAE buffer is 2.4 × 10−4 cm2/V/s and for nucleosomal arrays (23–25). References (23–25) indicate that these values were calculated using a 208-12 DNA template reconstituted with native chicken erythrocyte histone octamers. This nucleosome array therefore had 12 tandem nucleosomes assembled on 208-bp repeats of Lytechinus 5 S rDNA (38) −1.9 × 10−4 cm2/V/s. The values for m0 for DNA (both treated and untreated) are similar since the difference between the two substrates is the presence of covalent crosslinks, which does not contribute to a substantial change in surface charge. This method therefore is transparent toward the presence of crosslinked DNA. The amount of nucleosomes assembled can be calculated first by determining the difference in charge between DNA and assembled DNA using the following equation: m0 of DNA − m0 of nucleomal array × (No. of kb in DNA × 2) m0 of DNA The factor 2 is due to the charge contributed by the two phosphate groups present per base pair of DNA. This difference in charge is then divided by the charge of a single octamer in the nucleosomal context. The charge for a single octamer in a nucleosome was calculated in (23, 24) to be ~91 positive charges on the nucleosome surface. This allows one to

190

Subramanian et al.

determine the amount or number of nucleosomes assembled on a DNA template. In Fig. 4D, the in vitro chromatin assembly performed and similar calculations were done on drug-treated unassembled as well as assembled plasmid DNA. Table I shows untreated plasmid DNA has 18.5 ± 2 nucleosomes associated with plasmid DNA. This number has also been corroborated by other methods such as 2D topology gel analyses as well as atomic force microscopy data (data not shown). On the other hand, compound 5-mediated crosslinked plasmid DNA shows 14.5 ± 1 nucleosomes assembled on the template.

4. Notes 1. The SyBr Gold solution is always made in a plastic container. Glass containers should not be used for this purpose, since the SyBr Gold sticks to glass, thus resulting in low DNA staining efficiency. 2. The p4TxRE plasmid is prepared by transformation into special E.coli cells called TG2 cells. Other E.coli strains like HB101 and DH5a do not provide a very high copy number of this plasmid, largely because this plasmid has viral components. 3. The transcription stop buffer is made by autoclaving the stocks of some of the components individually. 0.5 M EDTA pH 8.0, 10% SDS and 4 M NaCl are autoclaved separately. A stock of 5 mg/ml of glycogen is made in filter sterilized TE buffer. The final mixture is again filter sterilized and stored in a sterile Falcon® tube. 4. In order to avoid mixed radioactive waste (phenol:chloroform :isoamylalcohol and free radiolabel), the excess radiolabel can be cleaned away using G25 resin. 8–10  ml of swollen and autoclaved G25 resin is packed in a 10 ml syringe. The resin is washed with water first with 2 column volumes (CV) followed by 2 CVs of TE. The 5¢ DNA labeling reaction is then loaded on top of the G25 column along with 2 ml of TE. The column is placed in a 50 ml Falcon® tube and along with an appropriate balance spun at 1,000 rpm for 1 min. The flow through contains the labeled DNA. Subsequent washes will release the free radiolabel from the column. The column traps the free radiolabel and can be disposed as solid radioactive waste. The G25 column has to be poured fresh before use. 5. The ratio of Drosophila core histones to dNap1 in this study was found to be ideal for the above chromatin assembly reaction, however differences in protein quantitation can skew the

Methods to Characterize the Effect of DNA-Modifying Compounds on Nucleosomal DNA

191

ratio at which ideal Nap1-histone complexes form, which can in turn affect the chromatin assembly process. It is therefore recommended that multiple NAP1-histone ratios be set up for ideal chromatin assembly. The ideal assembly conditions can be discerned from the MNase assay. Figure  4A represents microccal digests of a well-spaced assembly. Any deviation from this pattern is indicative of a less regular chromatin assembly reaction, in terms of nucleosome spacing on the DNA template.

Acknowledgments We thank Fujisawa Pharmaceutical (Astellas) for the generous gift of FK317. We are grateful to Drs. Jeffrey Hansen and Xu Lu for their help with multigels, and to Dr. Paul Laybourn and Kasey Konesky for their help and for providing reagents for chromatin assembly. Supported by NIH GM061909 to KL, by NIH CA51875 to RMW, and by CA041986 to DLB. References 1. Zwelling LA, Anderson T, Kohn KW (1979) DNA-protein and DNA interstrand crosslinking by cis-platinum(II) and transplatinum(II) Diamminedichloride in L1210 mouse leukemia-cells and relation to cytotoxicity. Cancer Res 39:365–369 2. Judd T, Williams RM (2002) Synthesis and DNA cross-linking of a phototriggered FR900482 mitosene progenitor. Org Lett 4:3711–3714 3. Boger DL, Garbaccio RM (1997) Catalysis of the CC-1065 and duocarmycin DNA alkylation reaction: DNA binding induced conformational change in the agent results in activation. Bioorg Med Chem 5:263–276 4. Iyer VN, Szybalski W (1963) A molecular mechanism of mitomycin action: linking of complementary DNA strands. Proc Nat Acad Sci U S A 50:355–362 5. Tomasz M, Lipman R, Chowdary D, Pawlak J, Verdine GL, Nakanishi K (1987) Isolation and structure of a covalent cross-link adduct between mitomycin-C and DNA. Science 235:1204–1208 6. Tomasz M (1995) Mitomycin-C-small, fast and deadly (but very selective). Chem Biol 2:575–579 7. Pratt WB, Ruddon RW, Ensminger WD, Maybaum J (eds) (1994) The anti-cancer

8.

9.

10.

11.

12. 13.

drugs, 2nd edn. Oxford University Press, New York Uchida I, Takasae S, Kayakiri H, Kiyoto S, Hashimoto M (1987) Structure of FR900482, a novel antitumor antibiotic from a streptomyces. J Am Chem Soc 109:4108–4109 Williams RM, Rajski SR, Rollins SB (1997) FR900482, a close cousin of mitomycin C that exploits mitosene-based DNA cross-linking. Chem Biol 4:127–137 Millard JT, Weidner MF, Raucher S, Hopkins PB (1990) Determination of the DNA crosslinking sequence specificity of reductively activated mitomycin-C at single-nucleotide resolution – deoxyguanosine residues at CpG are cross-linked preferentially. J Am Chem Soc 112:3637–3641 Boger DL, Searcey M, Tse WC, Jin Q (2000) Bifunctional alkylating agents derived from duocarmycin SA: Potent antitumor activity with altered sequence selectivity. Bioorg Med Chem Lett 10:495–498 Wolkenberg SE, Boger DL (2002) Mechanisms of in situ activation for DNA-targeting antitumor agents. Chem Rev 102:2477–2495 Trzupek JD, Gottesfeld JM, Boger DL (2006) Alkylation of duplex DNA in nucleosome core particles by duocarmycin SA and yatakemycin. Nat Chem Biol 2:79–82

192

Subramanian et al.

14. Luger K, Mader AW, Richmond RK, Sargent DF, Richmond TJ (1997) Crystal structure of the nucleosome core particle at 2.8 angstrom resolution. Nature 389:251–260 15. Richmond TJ, Davey CA (2003) Structure of DNA in the nucleosome core. Nature 423:145–150 16. Cech T, Pardue ML (1977) Cross-linking of DNA with trimethylpsoralen is a probe for chromatin structure. Cell 11:631–640 17. Wieshahn GP, Hyde JE, Hearst JE (1977) Photoaddition of trimethylpsoralen to drosophila-melanogaster nuclei – probe for chromatin substructure. Biochemistry 16:925–932 18. Hanson CV, Shen CK, Hearst JE (1976) Cross-linking of DNA in situ as a probe for chromatin structure. Science 193:62–64 19. Hayes J, Scovell WM (1991) Cisdiamminedichloroplatinum (II) modified chromatin and nucleosomal core particle. Biochim Biophys Acta 1089:377–385 20. Foka M, Paoletti J (1986) Interaction of cisdiamminedichloroplatinum(II) to chromatin – specificity of the drug distribution. Biochem Pharm 35:3283–3291 21. Bubley GJ, Xu J, Kupiec N, Sanders D, Foss F, O’Brien M, Emi Y, Teicher BA, Patierno SR (1996) Effect of DNA conformation on cisplatin adduct formation. Biochem Pharm 51:717–721 22. Millard JT, Spencer RJ, Hopkins PB (1998) Effect of nucleosome structure on DNA interstrand cross-linking reactions. Biochemistry 37:5211–5219 23. Fletcher TM, Serwer P, Hansen JC (1994) Quantitative-analysis of macromolecular conformational-changes using agarose-gel electrophoresis – application to chromatin folding. Biochemistry 33:10859–10863 24. Fletcher TM, Krishnan U, Serwer P, Hansen JC (1994) Quantitative agarose-gel electrophoresis of chromatin – nucleosome-dependent changes in charge, shape, and deformability at low ionic-strength. Biochemistry 33:2226–2233 25. Sewer P (1986) Use of gel electrophoresis to characterize multimolecular aggregates. Methods Enzymol 130:116–132 26. Judd TC, Williams RM (2002) Synthesis and DNA cross-linking of a phototriggered FR900482 mitosene progenitor. Org Lett 4:3711–3714

27. Naoe Y, Inami M, Matasumoto S, Nishigaki F, Tsujimoto S, Kawamura I, Miyayasu K, Manda T, Shimomura K (1998) FK317: A novel substituted dihydrobenzoxazine with potent antitumor activity which does not induce vascular leak syndrome. Cancer Chemother Pharmacol 42:31–36 28. Dyer PN, Edatathumangalam RS, White CL, Bao Y, Chakravarthy S, Muthuraja UM, Luger K (2004) Reconstitution of nucleosome core particles from recombinant histones and DNA. Methods Enzymol 375:23–44 29. Simpson RT, Stafford DW (1983) Structural features of a phased nucleosome core particle. Proc Natl Acad Sci U S A 80:51–55 30. Fyodorov DV, Kadonga JT (2003) Chromatin assembly in  vitro with purified recombinant ACF and NAP-1. Methods Enzymol 371:499–515 31. Konesky KL, Laybourn PJ (2007) Biochemical analyses of transcriptional regulatory mechanisms in a chromatin context. Methods 41:259–270 32. Matthews MA, Markowitz RB, Dynan WS (1992) In vitro activation of transcription by the human T-cell leukemia-virus type-I tax protein. Mol Cell Biol 12:1986–1996 33. Sambrook J, Russell DW (2001) Molecular cloning – a laboratory manual, 3rd edn. Cold Spring Habour Laboratory Press, New York 34. Subramanian V, Ducept P, Williams RM, Luger K (2007) Effects of photochemically activated alkylating agents of the FR900482 family on chromatin. Chem Biol 14:553–563 35. Kaufman PD, Kobayashi R, Kessler N, Stillman B (1995) The p150 and p60 subunits of chromatin assembly factor-I – a molecular link between newly synthesized histones and DNA-replication. Cell 81:1105–1114 36. Nabatiyan A, Krude T (2004) Silencing of chromatin assembly factor 1 in human cells leads to cell death and loss of chromatin assembly during DNA synthesis. Mol Cell Biol 24:2853–2862 37. Pan SS, Iracki T, Bachur NR (1986) DNA alkylation by enzyme-activated mitomycin-C. Mol Pharm 29:622–628 38. Simpson RT, Thoma F, Brubaker JM (1985) Chromatin reconstituted from tandemly repeated cloned DNA fragments and core histones – a model system for study of higherorder structure. Cell 42:799–808

Chapter 12 REPSA: Combinatorial Approach for Identifying Preferred Drug–DNA Binding Sequences Michael W. Van Dyke Abstract Many DNA-binding small molecules, typically those with a molecular mass less than 1,000 g/mol, recognize duplex DNA with some degree of sequence specificity. These include drugs used to treat several human diseases, including viral and bacterial infections, malaria, and cancer. Determining the binding specificity of DNA-binding molecules can be important for their development, especially if they are being designed to target specific DNA sequences. A limited amount of information can be obtained through the study of small molecule binding to defined naturally occurring or synthetic DNA sequences; however, a full picture of a small molecule’s binding specificity can only be obtained through combinatorial means, whereby vast libraries of sequences are screened. Several combinatorial methods have been developed for the study of ligand–DNA interactions, but only one method, Restriction Endonuclease Protection Selection and Amplification (REPSA), is generally applicable to the study of native small molecule–DNA complexes under physiologic conditions. REPSA may be used with both covalent and noncovalent small molecule– DNA complexes and with mixtures of small molecules with relatively unknown identities and properties. Thus, REPSA is a powerful, versatile, general method for the combinatorial determination of small molecule–DNA binding specificity and a functional means for drug discovery and characterization. Key  words:  Antibiotic, Anticancer agent, Combinatorial selection, Double-stranded DNA, Polymerase chain reaction, Sequence specificity, Type IIS restriction endonuclease

1. Introduction Combinatorial methods are powerful means of identifying preferred ligand-binding sites on nucleic acids. These methods have been referred to as in vitro genetics, cyclic amplification and selection of targets, directed molecular evolution, and SELEX (1–4). Typically, these methods involve large pools of oligonucleotides containing randomized sequences, purification of ligand–oligonucleotide complexes, and amplification of nucleic acids from these complexes. K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_12, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

193

194

Van Dyke

Repeating this process through multiple cycles yields a population of nucleic acids that are enriched in sequences exhibiting higher affinities for the ligand being investigated. A flow diagram describing these combinatorial methods is shown in Fig. 1a. Such methods have primarily been used to identify high-affinity binding sites of proteins on both DNA and RNA; however, they have also been successfully applied to multiplex-forming nucleic acids and some small molecule (<1,000 Da) ligands. One limitation of these combinatorial methods is that they require the physical separation of ligand–nucleic acid complexes from unbound nucleic acids. Typically, this is achieved through the altered physical properties of ligand-bound nucleic acids (e.g., reduced electrophoretic mobility) or through ligand-specific affinity methods (e.g., immunoprecipitation). For most small molecule ligands, however, these selection methods can be problematic. Because of their small mass relative to the oligonucleotides used for combinatorial selection, small molecules do not greatly change the density, electrophoretic mobility, hydrodynamic mobility, or lipophilicity of their nucleic acid complexes. Likewise, these molecules are not amenable to most affinity methods that directly recognize bound ligands unless they are first modified for this purpose, for example, through biotinylation or the introduction of chemically reactive moieties (5). Modification of small molecules can affect their intrinsic nucleic acid-binding properties, thus complicating these approaches. We have developed a novel combinatorial method, Restriction Endonuclease Protection, Selection, and Amplification (REPSA), that permits the identification of preferred ligand-binding sites on duplex DNA. Unlike most combinatorial methods, REPSA relies on the ability of bound ligands to inhibit an enzymatic cleavage process that renders unbound DNA otherwise unsuitable for amplification (Fig. 1b). To accomplish this, REPSA uses type IIS restriction endonucleases (IISREs), which cleave duplex DNA at a fixed distance from their binding site without regard to sequence specificity (6). REPSA also uses selection templates that direct IISRE cleavage to a region of randomized sequence (Fig. 2). We used REPSA to identify the preferred binding sites of a purine motif triple-helical DNA-forming deoxyribonucleotide 5¢-TG3TG4TG4TG3T-3¢ (7) and several DNA-binding proteins, including the restriction endonuclease Bsg I (7), an uncharacterized Bacillus sphaericus DNA-binding protein (7), and the TATAbinding subunit of the human general transcription factor TFIID (8). We were also able to successfully identify the preferred DNA binding sites of several small molecules, including the natural product antibiotics distamycin A and actinomycin D (9, 10), the synthetic hairpin polyamides ImPyPyPy-g-PyPyPyPy-b-Dp and ImPyPyPy-g-ImPyPyPy-b-Dp (11), and the covalent DNAbinding polyamide tallimustine (12). Taken together, our findings

REPSA: Combinatorial Approach for Identifying Preferred Drug–DNA Binding Sequences

195

Fig. 1. Flow diagrams of combinatorial methods used to select for duplex DNAs containing preferred ligand-binding sites. (Left ) Cyclic Amplification and Selection of Targets (CASTing). The steps in each CASTing cycle include (1) complex formation, (2) physical separation of complexes from unbound DNA and ligand, and (3) PCR amplification of selected DNAs. (Right ) Restriction Endonuclease Protection Selection and Amplification (REPSA). The steps in each REPSA cycle include (1) complex formation, (2) unbound template inactivation after DNA cleavage by IISREs, and (3) PCR amplification of selected, uncleaved DNAs.

196

Van Dyke

Fig. 2. Schematic representation of REPSA selection template ST2. Restriction endonuclease binding sites are indicated by brackets (in the previous section), and their cleavage sites are indicated by arrows. Bp Bpm I cleavage site, Bs Bsg I cleavage site, F Fok I cleavage site, H Hph I cleavage site.

demonstrate that REPSA can be successfully used to determine the consensus-binding sequences for a variety of DNA-binding ligands, including other nucleic acids, proteins, and small molecules. Notably, REPSA is unique among combinatorial methods in that it is suitable for the analysis of unmodified small molecules, such as various anticancer agents and oligomeric small molecules designed to target specific DNA sequences (13, 14).

2. Materials 2.1. Oligonucleotides

1. Template 63R14 (5¢-GTCCAAGCTTCTGGAGGGATGG TAAN14ATTCACCTCTGCACG AATTCCTAG-3¢, where N = an equimolar distribution of A, C, G, and T), PAGE-purified (Sigma-Genosys, The Woodlands, TX). Prepare as a 100 ng/mL stock and 5 ng/mL working solution in Tris–EDTA (TE) buffer. Store at −20°C. 2. Amplimers 63AL (5¢-CTAGGAATTCGTGCAGAGGTGAAT-3¢) and 63AR (5¢-GTCCAAGCTTCTGGAGGGATGGTAA-3¢), desalted (Sigma-Genosys, The Woodlands, TX). Prepare as a 100 ng/mL working solution in TE buffer. Store at −20°C. 3. TE buffer: 10 mM Tris–HCl (pH 7.9) and 1 mM ethylenediaminetetraacetic acid. Store at 4°C. 4. M13 Forward (-20) and M13 Reverse primers (Invitrogen, Carlsbad, CA). Prepare as a 100 ng/mL working solution in TE buffer. Store at −20°C.

2.2. Polymerase Chain Reactions

1. Taq DNA polymerase (Sigma-Aldrich, St. Louis, MO). 2. Taq DNA polymerase reaction buffer (10×): 100  mM Tris– HCl (pH 8.3), 500  mM KCl, 15  mM MgCl2, and 0.01% (w/v) gelatin.

REPSA: Combinatorial Approach for Identifying Preferred Drug–DNA Binding Sequences

197

3. Deoxyribonucelotide mixture (50×): 10  mM each dATP, dCTP, dGTP, and TTP. 4. MinElute® PCR Purification Kit (Qiagen, Valencia, CA). 2.3. Type IIS Restriction Endonuclease Cleavage Reactions

2.4. Polyacrylamide Gel Electrophoresis

1. Ligand-binding buffer (10×): 100  mM Tris–HCl (pH 7.9), 500 mM KCl, 100 mM MgCl2, 1% (w/v) Nonidet P-40, and 10 mM dithiothreitol. Store in aliquots at −20°C. 2. Type IIS restriction endonucleases Fok I and Bpm I (New England BioLabs, Beverly, MA). 1. 15% Ready Gel TBE Precast Gel, 15-well format (Bio-Rad, Hercules, CA). 2. TBE buffer: 89 mM Tris-base, 89 mM boric acid, and 2 mM ethylenediaminetetraacetic acid.

2.5. Subcloning and Sequencing

1. Zero Blunt® TOPO PCR Cloning Kit (Invitrogen, Carlsbad, CA). 2. Kanamycin sulfate solution (100×), equivalent to 10 mg/mL kanamycin (free base), filter sterilized (Sigma-Aldrich, St. Louis, MO). 3. LB medium: 10  g/L bacto-tryptone, 5  g/L bacto-yeast extract, and 10 g/L NaCl. Autoclave to sterilize. Add 100× kanamycin sulfate solution to 1× final concentration once medium has cooled to <40°C. 4. LB-agar medium: 10 g/L bacto-tryptone, 5 g/L bacto-yeast extract, 10 g/L NaCl, and 15 g/L bacto-agar. Autoclave to sterilize. Add 100× kanamycin sulfate solution to 1× final concentration once medium has cooled to <40°C. 5. Glycerol (Sigma-Aldrich). Autoclave to sterilize. 6. QuickLyse Miniprep Kit (Qiagen, Valencia, CA).

3. Methods Identifying preferred small molecule-binding sites on duplex DNA by REPSA involves several steps, including the following: (1) preparing the selection template, (2) performing REPSA selections, (3) assaying for the enrichment of ligand-binding sequences, (4) sequence isolation and characterization, and (5) data analysis. Each step is detailed in the following sections. 3.1. Preparing the Selection Template ST2

1. Assemble the selection template ST2 (see Note 1) synthesis reaction on ice in this order: 40 mL of water, 5 mL of 10× Taq DNA polymerase reaction buffer, 1  mL of template 63R14

198

Van Dyke

(5  ng/mL), 1  mL each of amplimers 63AL and 63AR (100 ng/mL), and 1 mL of the 50× deoxyribonucelotide mixture. Mix gently. Add 1 mL (5 units) of Taq DNA polymerase and mix gently. 2. Perform ST2 synthesis. Place the samples in a thermocycler. Use the following amplification protocol: 94°C for 1  min, 55°C for 1 min, and 72°C for 2 min. Repeat for a total of six cycles (see Note 2) and then stop synthesis by incubating the samples at 4°C. 3. Purify ST2 from the reaction using a MinElute® PCR Purification Kit, following the manufacturer’s instructions. 4. Quantitate ST2 by ultraviolet (UV) spectrophotometry. Use the standard extinction coefficient for duplex DNA (1 OD260 = 50 ng DNA/mL). 5. Prepare a working solution (2 ng/mL) of ST2 in TE buffer. Store in aliquots at −20°C. 3.2. Performing REPSA

1. Assemble the ligand-binding reaction on ice in this order: 7  mL of water, 1  mL of 10× ligand-binding buffer, 1  mL of selection template ST2 (2 ng/mL), and 1 mL of 10 mM ligand (see Note 3). 2. Incubate the ligand-binding reaction for 30 min at 37°C (see Note 4). 3. To initiate IISRE cleavage, add 0.25 unit of Fok I (0.5 mL) and incubate for 5 min at 37°C (see Note 5). 4. Stop the IISRE cleavage reaction by placing the sample on ice (see Note 6). 5. Assemble the PCR. To the forementioned IISRE cleavage reaction mixture, add 31 mL of water, 4 mL of 10× Taq DNA polymerase reaction buffer, 1 mL each of amplimers 63AL and 63AR (100  ng/mL), and 1  mL of 50× deoxyribonucelotide mixture. Mix gently. Add 1 mL (5 units) of Taq DNA polymerase and again mix gently (see Note 7). 6. Perform the PCR. Place the samples in a thermocycler. Use the following amplification protocol: 94°C for 1  min, 55°C for 1 min, and 72°C for 2 min. Repeat for a total of six cycles and then stop synthesis by incubating the samples at 4°C. 7. Purify REPSA-selected ST2 from the PCR using a MinElute® PCR Purification Kit, following the manufacturer’s instructions. Purified DNA is eluted in a minimal volume (10 mL). 8. Use 2 mL of purified DNA to quantitate the PCR product by UV spectrophotometry, using the standard extinction coefficient for duplex DNA (1 OD260 = 50 ng DNA/mL).

REPSA: Combinatorial Approach for Identifying Preferred Drug–DNA Binding Sequences

199

9. Use 2 ng of REPSA-selected ST2 to seed a subsequent round of REPSA. Retain the remainder for PAGE analysis or subcloning. Store at −20°C. Repeat steps 1–9 for a total of 9 REPSA cycles (see Note 8). 10. For the final round of REPSA, repeat steps 1–8, except use 0.5 unit of Bpm I (0.5 mL) as the IISRE in step 3. 3.3. PAGE Analysis of REPSA Products

1. Assemble 15% Ready Gel TBE Precast Gel into an appropriate gel electrophoresis apparatus (e.g., Mini PROTEAN Tetra system, Bio-Rad, Hercules, CA). Use 1× TBE as the running buffer. 2. Add 1  mL of 5× loading dyes, which are supplied with the MinElute® PCR Purification Kit, to 5  mL of purified DNA and load onto the assembled gel. Note: a single lane containing 20 ng of unselected ST2 serves as a concentration and molecular weight marker. 3. Perform electrophoresis at 70 V until the bromophenol blue (dark blue) tracking dye is within 1.5 cm of the gel bottom. 4. Disassemble the apparatus and incubate the gel by rocking it in a tray containing 50 mL of 1 mg/mL ethidium bromide solution for 15 min at ambient temperature. Remove ethidium bromide solution and destain with 50 mL of water for 15 min at ambient temperature. Caution: ethidium bromide is a mutagen and suspected carcinogen. Handle with care and dispose of as chemical waste. 5. Visualize DNA using a UV transilluminator (305  nm) and record it photographically or with a charged-coupled device (CCD) camera.

3.4. Subcloning REPSA-Selected DNA

1. Using a Zero Blunt® TOPO PCR Cloning Kit (Invitrogen, Carlsbad, CA), add 2  mL of purified DNA from the final REPSA round to 1 mL of supplied salt solution, 2 mL of water, and 1 mL of pCR®II-Blunt-TOPO®. Mix gently and incubate it for 5  min at ambient temperature to affect ligation. The ligated plasmid may be stored overnight at −20°C. 2. Add 2 mL of the above ligation reaction mixture to the supplied One Shot® Mach1™-T1R chemically competent E. coli cells. Mix gently. Incubate for 15  min on ice. Heat shock cells for 30 s at 42°C and immediately quench them on ice. Add 250  mL of ambient temperature, kit-supplied SOC medium and incubate the mixture for 1 h at 37°C to allow bacterial expression of plasmid-encoded kanR gene. 3. Spread 40 mL of the above bacterial culture onto a prewarmed (37°C) LB-agar plate containing 50 mg/mL kanamycin. Incubate the plate overnight at 37°C to allow the colonies to appear.

200

Van Dyke

3.5. Functional Analysis of REPSASelected DNA

1. Isolate individual clones and prepare the probes for analysis by restriction endonuclease protection assay (REPA) (15). Touch the selected bacterial colony with a sterile toothpick and use it to inoculate a gridded LB-agar patch plate containing 50 mg/ mL kanamycin. Then, use the same toothpick to inoculate an ice-cold PCR reaction mixture containing 42  mL of water, 4  mL of 10× Taq DNA polymerase reaction buffer, 1  mL of each M13 Forward (-20) and M13 Reverse primers (100 ng/mL), 1 mL of 50× deoxyribonucelotide mixture, and 1 mL (5 units) of Taq DNA polymerase. Mix gently. 2. Make REPA probe DNA. Transfer the PCR reaction mixture to a thermocycler. Use the following amplification protocol: 94°C for 1 min, 55°C for 1 min, and 72°C for 2 min. Repeat for a total of 25 cycles and then stop synthesis by incubating the samples at 4°C. Purify the PCR products using a MinElute® PCR Purification Kit, following the manufacturer’s instructions. Purified probe DNA is eluted in a minimal volume (10 mL). Use 2 mL of purified DNA to quantitate by UV spectrophotometry using the standard extinction coefficient for duplex DNA (1 OD260 = 50 ng DNA/mL). Dilute probe DNA to 25 ng/mL with TE buffer. Store at 4°C. 3. Generate the clone library. Incubate the patch plate overnight at 37°C. Once colonies have appeared, they can be stored long-term at 4°C. 4. Perform REPA. Assemble cleavage reactions on ice in this order: 7 mL of water, 1 mL of 10× ligand-binding buffer, 1 mL of probe DNA (25 ng/mL), and either 1 mL of 10 mM ligand for the test reaction or 1 mL of water for the control reaction. Incubate the mixture for 30  min at 37°C to affect binding. Then, add 0.25 unit of Fok I (0.5 mL) and incubate for 5 min at 37°C to permit cleavage. Stop the cleavage reaction by placing the reaction mixtures on ice. Add 2 mL of 5× loading dyes to the cleavage reaction and load onto an assembled 15% Ready Gel TBE Precast Gel. Perform electrophoresis, staining, and recording as described earlier (Subheading  3.3). Note: the use of a CCD camera and image analysis software will allow quantitation of cleavage reactions, which will be useful when categorizing clones as containing low- or highaffinity ligand-binding sites. An example of typical results is shown in Fig. 3.

3.6. Sequencing REPSA-Selected DNA

1. Using a single colony from the clone library, inoculate 4 mL of LB medium containing 50 mg/mL kanamycin in a 15-mL snap-cap polypropylene culture tube and incubate overnight at 37°C with vigorous shaking.

REPSA: Combinatorial Approach for Identifying Preferred Drug–DNA Binding Sequences

201

Fig. 3. Assaying selection template clones by REPA. Radiolabeled PCR probes from individual clones were digested with the IISRE Fok I after incubation without (−/+) or with (+/+) 50 mM tallimustine. (−/−), undigested control. Shown is an autoradiogram of PAGE-resolved DNA fragments from uncleaved (U) and cleaved (C) probes. An asterisk (*) indicates a PCR artifact. Reprinted in part with permission from (12). Copyright 2005 American Chemical Society.

2. Place 1.5 mL of bacterial culture in a supplied 2 mL QuickLyse lysis tube and pellet bacteria by centrifugation at 18,000 × g for 1 min at ambient temperature in a microfuge. Purify plasmid DNA from bacteria following the protocol provided in the QuickLyse Miniprep Kit (Qiagen, Valencia, CA). The yield should be approximately 5 mg of plasmid DNA at a concentration of 100 ng/mL. Store at 4°C. 3. Submit 10  mL (1  mg) of plasmid solution for automated sequencing. Either the M13 Forward (-20) or M13 Reverse primer may be used. 4. Selected bacteria can be stored long-term by mixing 0.85 mL of bacterial culture with 0.15 mL of sterile glycerol and freezing at −80°C. 3.7. Analysis of REPSA-Selected Sequences

1. The sequences present in the randomized cassettes of the REPSA-selected clones can be analyzed using motif discovery and search software such as MEME (16). A web-based submission form can be found at http://meme.sdsc.edu/ meme4/cgi-bin/meme.cgi. Sequences should be provided in FASTA format. Parameters initially selected include (a) zero or one motif per sequence, (b) a minimum motif width of 4 and maximum of 14, and (c) a maximum of 5 motifs to find. Longer consensus sequences can then be identified by sequentially increasing the selected minimum motif width. An example of a typical MEME analysis is shown in Fig. 4.

202

Van Dyke

Fig. 4. Identifying consensus ligand-binding sequences from REPSA-selected DNA. Shown is a MEME alignment of randomized cassette sequences from clones obtained after seven rounds of REPSA that exhibited substantial tallimustine-dependent cleavage protection by REPA. Sequences from either the (+) or (−) strand are indicated. The consensus sequence is boxed.

4. Conclusions REPSA is a versatile combinatorial method for identifying preferred ligand-binding sequences on duplex DNA. It can be used for almost any type of DNA-binding ligand that is capable of interfering with enzymatic DNA cleavage, including unmodified small molecules like anticancer drugs. REPSA works well under mesophilic physiologic conditions, those commonly required by most IISREs. However, given the existence of extremophile organisms containing IISREs (e.g., Bacillus stearothermophilus F and BsmF I), it should be possible to expand REPSA studies to many different environments. REPSA requires essentially no foreknowledge of the ligands under investigation and can provide information on preferred binding sites even when multiple, uncharacterized ligands are present in the selection (see ref. 7). REPSA can even be used to identify the preferred binding sites of ligands that covalently modify DNA (7, 12), as long as one DNA strand of the selection template can be rendered capable of amplification by PCR. REPSA is a technically simple method, requiring few steps and sample manipulations. Thus, unlike conventional combinatorial methods, REPSA should be amenable to automation and high-throughput analysis. We believe that REPSA is a powerful method for studying molecules that bind to DNA and that it will ultimately become a standard tool in structural biology and medicinal chemistry laboratories.

REPSA: Combinatorial Approach for Identifying Preferred Drug–DNA Binding Sequences

203

5. Notes 1. ST2 is one selection template that can be used to identify preferred drug–DNA binding sites by REPSA. Others include ST1 (7) and ST3 (9). These selection templates differ in the length of the randomized cassette and the IISREs that can be used with them. In fact, most DNA that has (1) a region of random sequence, (2) defined flanks suitable for PCR amplification, and (3) binding sites for one or more IISREs, positioned such that they direct cleavage within the randomized region, can serve as a selection template for REPSA. Note that the length of the randomized cassette and the choice of IISRE used should be considered. For example, it is difficult to get a good representation of all possible sequences in a reasonable quantity (2  ng) of selection template when the randomized region is longer than 14 base pairs. In addition, in our hands, the following IISREs cleaved DNA with relative efficiency: Fok I > Bpm I > Bsg I, Eco57 I, and Hph I. Therefore, for small molecule ligands that recognize three to eight base pairs, we have found the selection template ST2 to be sufficient for their study by REPSA. 2. The number of PCR cycles is intentionally kept low (six cycles) so as not to exhaust the supply of amplimers required for the amount of input selection template used (2  ng). Should amplimer exhaustion occur, subsequent rounds of melting and annealing in latter cycles of the PCR will result in selection templates that are properly annealed on their flanking regions but not within the randomized cassette. These templates will not be cleaved by IISREs in subsequent REPSA rounds and will be amplified, thereby limiting the selection efficiency of REPSA. 3. For many small molecule ligands, a 1 mM final concentration is a good starting point for investigating their binding specificity by REPSA. However, if the dissociation constant is already known for a particular ligand:DNA complex, this information can be used to determine the ligand concentration to be investigated. Alternatively, a series of REPSA experiments may be performed through a range of ligand concentrations to identify the highest-affinity binding sites. 4. The duration and temperature of the binding reaction should be optimal for achieving equilibrium. Time is an easier parameter to adjust, given the temperature requirements of the IISRE being used in the REPSA selection. Alternatively, experiments may be performed with reduced incubation times, thereby permitting identification of kinetically preferred ligand-binding sites.

204

Van Dyke

5. The type of IISRE chosen, the amount of units used, and the duration of the cleavage reaction are based on conditions that provide a maximum difference in cleavage efficiency when a ligand is bound to a high-affinity site. Ideally, conditions should result in >90% cleavage of selection templates when a ligand is not present. The time of the cleavage reaction should be kept to a minimum, to maximize the detection of templates with high-affinity sites when investigating ligands with rapid binding kinetics. Conversely, ligands with relatively slow binding kinetics may benefit from extended cleavage reactions. 6. When investigating ligands that form covalent DNA adducts or intrinsically interfere with PCR amplification (e.g., tallimustine), it may be beneficial to purify the selection templates at this step. Unbound small molecule ligands can often be removed by extraction with an organic solvent (e.g., n-butanol or phenol). Further purification, if necessary, can be achieved using a MinElute® PCR Purification Kit, as described above. 7. Alternatively, a small fraction (equivalent to 10,000 cpm) of an amplimer can be 5¢ 32P end-labeled. This will allow the analysis of selection templates by REPA (Subheading  3.5) using far less material and more sensitive detection methods (e.g., autoradiography or phosphoimaging). 8. While it is possible to proceed through REPSA selections blindly, the number of REPSA rounds necessary to select for preferred ligand-binding sequences can be determined empirically by analyzing the amplified selection templates by REPA after each round, as described earlier(Subheading  3.5). The appearance of an emergent, ligand-dependent, cleavage-resistant selection template population demonstrates that REPSA has enriched the population for high-affinity ligand-binding sites. Note: it is not advisable to continue REPSA selections ad infinitum, given the propensity of selections with the IISRE Fok I to generate truncated selection templates that lack most randomized sequences. Such truncated species have a tendency to persist in subsequent REPSA selections.

Acknowledgments We thank Paul Hardenbol, who initially developed REPSA, and all the others in our laboratory who have used this method in their research: Jo C. Wang, Jing Shen, Xiaohong Yang, Y.N. Vashisht Gopal, and Gulshan Sunavala. We would also like to especially thank Nam Tonthat for his contemporary application of REPSA and auxiliary methods, which has made this manuscript possible. Our REPSA research was supported by grants from the American Cancer Society and the Robert A. Welch Foundation (G-1199).

REPSA: Combinatorial Approach for Identifying Preferred Drug–DNA Binding Sequences

205

References 1. Szostack JW (1992) In vitro genetics. Trends Biochem Sci 17:89–93 2. Wright WE, Funk WD (1993) CASTing for multicomponent DNA-binding complexes. Trends Biochem Sci 18:77–80 3. Kenan DJ, Tsai DE, Keene JD (1994) Exploring molecular diversity with combinatorial shape libraries. Trends Biochem Sci 19:57–64 4. Gold L, Polisky B, Uhlenbeck O, Yarus M (1995) Diversity of oligonucleotide functions. Annu Rev Biochem 64:763–797 5. Wilson DS, Szostack JW (1999) In vitro selection of functional nucleic acids. Annu Rev Biochem 68:611–647 6. Szybalski W, Kim SC, Hasan N, Podnajska AJ (1991) Class-IIS restriction enzymes – a review. Gene 100:13–26 7. Hardenbol P, Van Dyke MW (1996) Sequence specificity of triplex DNA formation: analysis by a combinatorial approach, restriction endonuclease protection selection and amplification. Proc Natl Acad Sci U S A 93:2811–2816 8. Hardenbol P, Wang JC, Van Dyke MW (1997) Identification of preferred hTBP DNA binding sites by the combinatorial method REPSA. Nucleic Acids Res 25:3339–3344 9. Hardenbol P, Wang JC, Van Dyke MW (1997) Identification of preferred distamycin-DNA binding sites by the combinatorial method REPSA. Bioconjug Chem 8:617–620

10. Shen J, Wang JC, Van Dyke MW (2001) Identification of preferred actinomycinDNA binding sites by the combinatorial method REPSA. Bioorg Med Chem 9:2285–2293 11. Vashisht Gopal YN, Van Dyke MW (2003) Combinatorial determination of sequence specificity for nanomolar DNA-binding hairpin polyamides. Biochemistry 42: 6891–6903 12. Sunavala-Dossabhoy G, Van Dyke MW (2005) Combinatorial identification of a novel consensus sequence for the covalent DNAbinding polyamide tallimustine. Biochemistry 44:2510–2522 13. Neidle S, Waring MJ (1993) Molecular aspects of anticancer drug–DNA interactions. Macmillan, London 14. Dervan PB (2001) Molecular recognition of DNA by small molecules. Bioorg Med Chem 9:2215–2235 15. Ward B (1996) Type IIS restriction enzyme footprinting I. Measurement of a triple helix dissociation constant with Eco57I at 25°C. Nucleic Acids Res 24:2435–2440 16. Bailey TL, Elkan C (1994) Fitting a mixture model by expectation maximization to discover motifs in biopolymers. In: Proceedings of the second international conference on intelligent systems for molecular biology. AAAI Press, Menlo Park, California, pp 28–36

Chapter 13 In vitro Transcription Assay for Resolution of Drug-DNA Interactions at Defined DNA Sequences Benny J. Evison, Don R. Phillips, and Suzanne M. Cutts Abstract A major class of anticancer agents in current clinical use exerts its anticancer effects by binding covalently or non-covalently to DNA. A detailed understanding of the nature of these drug-DNA complexes would be expected to lead to better uses of these drugs, and also assist with the design of improved drug derivatives. Here, we present a transcriptional footprinting assay that can be implemented to define the DNA sequence specificity and kinetics associated with drug-DNA complexes. The basic steps involve the formation of drug-DNA complexes, the formation of synchronised initiated transcripts, and finally transcriptional elongation to reveal drug blockage sites that impede the progression of RNA polymerase. We have used the “in vitro transcription assay” to investigate many covalent drug-DNA interactions; most notably those obtained using anthracycline anticancer agents such as doxorubicin and anthracenedionebased anticancer agents, including mitoxantrone and pixantrone. Key words: Drug, DNA, Transcription, Sequence specificity

1. Introduction Many anticancer agents in current clinical use interact with DNA covalently or non-covalently. A detailed understanding of these drugDNA interactions would be expected to provide the basis for the design of new generations of more active drug derivatives. Funda­ mental understanding of the DNA interaction includes knowledge of the kinetics and affinity of the interaction and the DNA sequence involved. Molecular biology-based approaches used to gauge the nature of these properties include DNA footprinting (utilising DNase I or hydroxyl radicals) and inhibition of the processivity of DNAdependent enzymes such as exonucleases and polymerases (1–3). Here, we present a transcription-based method (an “in vitro transcription assay”) that was developed in our laboratory and K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_13, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

207

208

Evison, Phillips, and Cutts

relies on the tendency of drug-occupied DNA sites to pose a physical blockage to RNA polymerase. This approach provides quantitative nucleotide resolution of the DNA specificity of drugoccupied sites as well as kinetic information associated with the interactions. Detailed theoretical information concerning this method has already been presented elsewhere (4, 5). We have used our version of the in vitro transcription assay to characterize the non-covalent DNA interactions of drugs, including actinomycin D, echinomycin, mitoxantrone, and doxorubicin (6–8) as well as probing long-lived or covalent DNA interactions produced by drugs such as doxorubicin, nitrogen mustards, mitoxantrone, and pixantrone (9–12). The experimental steps in this procedure are outlined in Subheading 3.2 later in the chapter. The DNA template we use contains the lac UV5 promoter since this promoter allows for robust production of synchronized transcripts using E. coli RNA polymerase. A variety of other promoter systems could alternatively be employed (4). The three main parts of the transcriptional footprinting procedure presented in this chapter entail the following key steps: 1. Reaction of DNA template containing the lac UV5 promoter with the drug of interest. Producing stable drug-DNA interactions suitable for transcriptional analysis requires conditions specific to the drug of choice. The drug-DNA complex must be relatively stable in order to survive the cleanup procedure presented below. Alternatively, the cleanup procedure can be omitted if the reactants will not be detrimental to subsequent transcription steps. Another version of the transcription assay entails initiation of transcription prior to drug exposure. This variation is often used to probe non-covalent drug-DNA interactions, where the interaction will not be inhibitory to the initiated transcription complex and this procedure has been presented in detail elsewhere (4, 5). 2. Initiation phase of transcription to form synchronized transcripts. Initiation of transcription is achieved quickly and with high fidelity using E. coli RNA polymerase and the lac UV5 promoter. Inclusion of a GpA dinucleotide ensures that the first nucleotide of the newly synthesized transcript is always at the -1 position, and therefore initiated transcripts are synchronized from a common starting point. Transcripts are stalled at 10 nucleotides in length due to the omission of CTP in the initiation mixture. Inclusion of three radiolabeled UTP (or ATP if preferred) nucleotides within the stalled initiated transcript allows for unambiguous and quantitative detection of equally labeled transcripts after electrophoresis. The synchronized initiation complex is represented diagrammatically in Fig. 1.

In vitro Transcription Assay for Resolution of Drug-DNA

209

Fig. 1. Synchronized transcription initiation complex. Initiation of the lac UV5 promoter with E. coli RNA polymerase GpA, ATP, GTP and [a −32P] UTP (in the absence of CTP) results in a stable transcription complex containing a nascent RNA ten nucleotides in length (indicated by an arrowhead). The nascent RNA begins at the −1 position with G of GpA in the initiation mixture. Radiolabelled [a–32P] UTP is incorporated into the nascent RNA in three locations (shown by asterisks).

3. Elongation phase of transcription to yield blocked transcripts. RNA polymerase progression is highly sensitive to the presence of drugs (and other physical perturbations) on the DNA template and becomes blocked at that site on the DNA. The length of transcripts therefore reveals the original location of drug-DNA complexes. Elongation of the initiated transcription complex is accomplished in the presence of high concentrations of unlabeled nucleotides. Since nucleotides are in vast excess to those used for initiation, further incorporation of radiolabeled UTP (or ATP) is not an issue, and therefore does not impede subsequent quantitative analysis.

2. Materials 2.1. Preparation of 512 bp DNA Fragment Containing the lac UV5 Promoter

1. Glycerol stock of plasmid containing the lac UV5 promoter (selection of promoter, see Note 1). We routinely use the plasmid pCC1 (13). 2. Qiagen maxi plasmid purification kit (QIAGEN, Hilden, Germany). 3. Restriction endonucleases PvuII and HindIII. 4. Elutrap electroelution apparatus and membranes (Whatman, Kent, UK).

210

Evison, Phillips, and Cutts

5. Agarose, molecular biology certified (Bio-Rad, Hercules, California). 6. Ultrapure buffer-saturated phenol (Invitrogen, Carlsbad, California). 7. Chloroform for liquid chromatography (Merck, Darmstadt, Germany). 8. Sodium acetate, anhydrous (Merck, Darmstadt, Germany). 9. Ethanol (Biolab, Clayton, Australia). 10. 1× TBE buffer (Tris-borate-EDTA) buffer: 90  mM Tris, 90  mM boric acid, 2  mM EDTA, pH 8.3. Store at room temperature as a 10× stock (see Note 2). 11. 1× TE buffer (Tris-EDTA) buffer: 10 mM Tris-HCl, 1 mM EDTA, pH 8.0. Store in aliquots at −20°C. 12. Transilluminator (Model TVC-312A, 312  nm, Spectroline, Spectronics, New York). 2.2. In vitro Transcription Assay 2.2.1. Individual Reagents

1. Dinucleotide GpA (guanylyl (3¢-5¢) adenosine) (Ribomed Technologies, Inc., Carlsbad, California). Make up a 2 mM stock solution in Milli-Q water and store at −20°C. 2. Ribonucleotides consisting of ATP, CTP, GTP, and UTP (RNase free). Supplied individually in purified water as 100 mM (pH 7.5) solutions (GE Healthcare, Buckinghamshire, UK). 3. Methoxy nucleotides for sequencing (eg. 3¢-methoxy CTP and 3¢-methoxy ATP) (see Note 3). 4. E. coli RNA polymerase, 1U/mL (USB Corp, Cleveland, Ohio). 5. [a-32P] UTP, 3,000 Ci/mmol (Perkin Elmer, Waltham, MA, USA). 6. Dithiothreitol (BioVectra, Charlottetown, Canada). Make up a 200 mM solution in Milli-Q water (see Note 4). 7. BSA, RNase/DNase free, approximately 2.5  mg/mL (GE Healthcare, Buckinghamshire, UK). 8. RNAguard™ (human placental), approximately 30 U/mL (GE Healthcare, Buckinghamshire, UK). 9. Heparin, ammonium salt (porcine intestinal mucosal) (Sigma, St Louis, MO). Make a 2  mg/mL stock solution in 1 × Tc (10 × Tc recipe is given below in subheading 2.2.2) and store at −20°C.

2.2.2. Reagent Mixes

1. 10× Transcription buffer (Tc): 400 mM Tris-HCl (pH 8.0), 1 M KCl, 30 mM MgCl2, and 1 mM EDTA. Autoclave, then store in aliquots at −20°C (see Note 5). 2. 6× Initiation mix (IM): 1.2  mM GpA, 30 mM GTP, 30 mM ATP, and 2 mCi/mL [a-32P] UTP in 1× Tc.

In vitro Transcription Assay for Resolution of Drug-DNA

211

3. 3× Elongation nucleotide mix (EM): 6  mM each of CTP, ATP, GTP, and UTP in transcription buffer containing 1.2 M KCl. 4. Sequencing solutions [eg. 10% Methoxy CTP/90% CTP (MeC mix)]. Make a solution in transcription buffer containing 30 mM 3¢ methoxy CTP, 0.27  mM CTP, 6  mM ATP, 6 mM GTP, and 6 mM UTP and 1.2 M KCl (see Note 6). Store in small aliquots at −20°C. 5. Termination/loading buffer: 9 M urea, 10% sucrose, 40 mM EDTA, 0.1% xylene cyanol, 0.1% bromophenol blue in 2× TBE buffer, pH 7.5.

3. Methods To facilitate optimal transcription, a high purity DNA template is essential. We have found that the DNA purification method presented below results in a high yield of good quality DNA template. The in vitro transcription method is presented in six steps (represented diagrammatically in Fig. 2, and a sample protocol is outlined in Fig.  3). The initial four steps (Steps 1–4 of Subheading 3.2) are sufficiently detailed to perform transcription in the absence of drug. It is recommended that the assay be tested in this manner to verify that initiated transcripts are no longer than 10 nucleotides in length, and that elongated transcripts produced in the absence of drug are predominantly full length. Sequencing lanes must also be well resolved in the transcript range to be analysed for subsequent drug-DNA interactions. When these criteria are satisfied, drug-DNA complexes can be produced (Subheading  3.2.5), and larger scale transcription assays can be employed to probe drug-DNA interactions (Subheading 3.2.6). Resolution and quantitation of truncated transcripts produced by drug blockages is the final step (Subheading 3.2.7). The transcription assay can then be used to yield further information, such as drug dissociation rates (Subheading 3.2.8). The examples presented relate to the characteristics of DNA adducts produced by formaldehyde-activated pixantrone (Figs. 4–6). 3.1. Preparation of 512 bp DNA Fragment Containing the lac UV5 Promoter

1. Set up a restriction digest of 20 mg pCC1 using PvuII and HindIII (see Note 7). Subject a small aliquot of this mix to electrophoresis using a 1% agarose gel to assess if liberation of the 512 bp fragment has been completed. 2. To separate the two DNA fragments, subject the restriction digest to electrophoresis using a 1.5% mini submarine agarose gel in 1× TBE buffer (lacking ethidium bromide) for 2 h at 10 V/cm (see Note 8).

212

Evison, Phillips, and Cutts

Fig. 2. Diagrammatic representation of the transcription assay. The major steps are formation of drug-DNA adducts within the 512 bp DNA fragment, binding of RNA polymerase selectively to the lac UV5 promoter (in the presence of heparin), formation of a synchronized initiated transcription complex, and elongation of the transcription complex to yield druginduced blocked transcripts of varied lengths which terminate at defined locations of the DNA template.

3. Cut a thin lengthwise slice from each side of the gel and stain with 0.5 mg/mL ethidium bromide. 4. Place the gel on a transilluminator and align the gel slice with the main gel. Visualise the location of the 512 bp fragment in the gel slice and excise the 512 bp fragment from the main agarose gel. 5. Place the agarose gel slice in an Elutrap chamber (Whatman) and electroelute the DNA fragment (see Notes 9 and 10).

In vitro Transcription Assay for Resolution of Drug-DNA

213

Fig. 3. Typical in vitro transcription protocol for drug-DNA concentration dependence studies. The 512 bp DNA template is reacted with drug under suitable conditions and then subjected to cleanup. The transcription mix (TM) is prepared and other samples required are initiation (I), sequencing (e.g., A and C), control (E) and drug-treated (S1-S4). The circled letters represent the end point of transcription for each of these samples where termination/loading buffer is added and samples subjected to electrophoresis.

6. Purify the DNA further by extracting with an equal volume of phenol followed by extraction with an equal volume of chloroform and then subject to ethanol precipitation. 7. Redissolve the DNA in TE buffer to a concentration of approximately 100 ng/mL (200 nM). 3.2. In vitro Transcription Assay 3.2.1. Initiation of Transcription

1. To a sterile Eppendorf tube, add lac UV5 DNA fragment (approximately 25 nM final concentration), 10× Tc (1× final concentration), DTT (10  mM final concentration), BSA (125 µg/mL final concentration), RNase Inhibitor (1 U/mL final concentration), and Milli-Q water in a total volume of 20 mL. 2. Add 1 mL E. coli RNA polymerase, mix gently, and incubate for 15 min at 37°C. 3. Add 5 mL of heparin and incubate for 5 min at 37°C.

214

Fig. 4. Comparison of blocked transcripts induced by the anticancer drugs mitoxantrone, pixantrone and doxorubicin. Multiple reactions were run in parallel, each containing the 512 bp DNA fragment (25 mMbp) and either of the following: 0–20 mM mitoxantrone and 5 mM formaldehyde; 0–5 mM pixantrone and 2 mM formaldehyde; 0–1 mM doxorubicin and 1 mM formaldehyde. Each reaction was incubated for 4 h, ethanol precipitated, resuspended and subjected to transcription. Transcription was initiated from the lac UV5 promoter of each sample, allowed to elongate for 5 min and subsequently terminated. Lane “I” represents the initiated complex prior to elongation, “E” represents elongated complex in the absence of drug and lanes “A” and “C” are sequencing lanes. The lengths of some of the transcripts are shown on the left-hand side of the autoradiogram.

215

Fig. 5. The sequence selectivity of (a) mitoxantrone, (b) pixantrone, and (c) doxorubicin. The % transcriptional blockage of each transcript in either the 20 mM mitoxantrone, 5 mM pixantrone or 0.25 mM doxorubicin lane (from Fig.  4) was quantitated and is expressed as a function of the sequence of a portion of the 512 bp DNA fragment. The nucleotide sequence of the non-template DNA strand is shown. To simplify the histogram transcriptional blockages less than an arbitrary value of 1% have been omitted (reprinted from Evison et  al. (12), Copyright© 2008, by permission of the ­publisher American Society for Pharmacology and Experimental Therapeutics).

216

Evison, Phillips, and Cutts

Fig. 6. Elongation of the transcription complex past pixantrone-induced blockage sites. (a) The 512 bp fragment was initially reacted with 5 mM pixantrone and 2 mM formaldehyde for 4 h. Following ethanol precipitation, drug-reacted DNA was resuspended and transcription initiated from the lac UV5 promoter. Elongation of the initiated complex was then allowed to proceed at 37°C for time periods ranging from 5 to 240 min. Lanes E1 and E2 are control lanes representing the initiated transcription complex that had been allowed to elongate in the absence of drug for 5 and 240 min, respectively. (b) Several drug-induced blockage sites, including 52-mer (open square), 59-mer (solid square), 80-mer (solid circle), 108-mer (open circle) and 120-mer (solid diamond) were quantitated and subjected to first-order kinetic analysis (reprinted from Evison et al. (12), Copyright© 2008, by permission of the publisher American Society for Pharmacology and Experimental Therapeutics). The half-life of each pixantrone-induced blockage is summarised in Table 1.

4. Add 5 mL of initiation mix (IM) and return to incubation at 37°C. 5. After 5 min at 37°C this mixture results in the formation of the initiated complex. Take a 5 mL aliquot and add to 5 mL termination/loading buffer on ice. 3.2.2. Elongation of Initiated Transcripts

1. Place a 10 mL aliquot of the initiated complex into an Eppendorf tube containing 5 mL of elongation mix (EM) and return to incubation at 37°C. 2. After 1  min at 37°C, take a 5 mL aliquot of the elongated complex and add to 5 mL termination/loading buffer on ice.

In vitro Transcription Assay for Resolution of Drug-DNA

217

3. After 5  min at 37°C, take a 5 mL aliquot of the elongated complex, and add to 5 mL termination/loading buffer on ice. 4. After 15  min at 37°C, take a 5 mL aliquot of the elongated complex, and add to 5 mL termination/loading buffer on ice (see Note 11). 3.2.3. Sequencing of Initiated Transcripts

1. Place two 5 mL aliquots of the initiated transcript into separate Eppendorf tubes. 2. Add 2.5 mL of MeC mix to one and 2.5 µL MeA mix to the other. 3. Incubate at 37°C for 5  min, then add an equal volume of termination/loading buffer to both samples, and place on ice.

3.2.4. Denaturing Gel Electrophoresis

1. Prepare a 12% acrylamide denaturing sequencing gel (19:1 acrylamide:bisacrylamide, containing 7 M urea) in TBE buffer. 2. Subject gel to pre-electrophoresis for 30 min to 1 h to heat gel to approximately 60°C (typically 2000 V, approximately 100 W). 3. Denature all samples (generated in steps 1–3 of Subheading 3.2 above) at 90°C for 5 min, then quench immediately on ice. 4. Rinse the wells of the gel with 1× TBE buffer then load 4–6 mL of each sample into each well. 5. Subject gel to electrophoresis until the bromophenol blue migrates approximately 75% of the length of the gel (1–2 h). Fix gel in 10% glacial acetic acid/10% methanol, rinse in distilled water, and then transfer onto Whatman 3MM paper. 6. Vacuum dry the gel.

3.2.5. Formation of Drug-DNA Adducts

1. Incubate the lac UV5 DNA fragment (approximately 400 ng/ sample) with the drug of interest using the appropriate drug activation conditions (see Notes 12 and 13). A good parameter to investigate initially is a range of drug concentrations. For an example of drug reaction conditions, refer to Fig. 4. Include a mock reaction that can be used for initiation, control, and sequencing lanes in subsequent transcription reactions. 2. Make up the volume of each sample to 200 mL using TE. 3. Add 200 mL of Tris-saturated phenol, vortex and centrifuge at full speed in a benchtop microcentrifuge for 5 min. 4. Transfer the top aqueous phase to a fresh Eppendorf tube. 5. Repeat steps 3 and 4 (see Note 14). 6. Add 200 mL of chloroform, vortex and centrifuge at full speed in a benchtop microcentrifuge for 5 min. 7. Ethanol precipitate the drug-reacted DNA using standard procedures. 8. Redissolve the DNA samples in TE buffer to a concentration of approximately 50 ng/mL (100 nM).

218

Evison, Phillips, and Cutts

3.2.6. Transcription of Drug-Reacted DNA

Set up transcription of mock reacted DNA as outlined in Sub-headings 3.2.1–3.2.3 in parallel with transcription of drugreacted template. However, this time make up a transcription mix that can be added to each individual DNA sample. Refer to Fig. 3 for a sample protocol. 1. To a sterile Eppendorf tube, add DTT (10 mM final concentration), 10× Tc (1× final concentration), BSA (125 mg/mL final concentration), RNase Inhibitor (1 U/mL final concentration), RNA polymerase (100  nM final concentration) and Milli-Q water in a total volume of 35 mL. This mixture comprises the transcription mix (TM) and is sufficient for all transcription control samples and approximately 4 drugreacted samples. 2. For control transcription reactions, add 4  mL mock treated DNA (control sample as outlined in Subheading 3.2.5) to a separate sterile Eppendorf tube. For each drug reacted DNA sample, add 1.5 mL per fresh Eppendorf tube. 3. Add 12 mL of TM to the control DNA sample and 4.5 mL TM to each drug-reacted DNA sample. 4. Incubate samples for 15 min at 37°C. 5. Add heparin to each tube; 4  mL to the control sample and 1.5 mL to each drug reacted DNA sample. 6. Incubate for 5 min at 37°C. 7. Add IM to each tube; 4 mL to the control sample and 1.5 mL to each drug reacted DNA sample. 8. Incubate for 5 min at 37°C. 9. Take a 5 mL aliquot of the control mixture and add to 5 mL termination/loading buffer on ice. 10. Take 2 × 5 mL aliquots of the control mixture and add to sequencing mixes; one to 2.5 mL MeC mix and one to 2.5 mL MeA mix. 11. Take a 5 mL aliquot of the control mixture and add to 2.5 mL EM. 12. Add 4.5 mL EM to each tube of drug-reacted DNA sample. 13. Incubate tubes from points 10–12 above at 37°C for 1, 5 or 15  min (as determined to be appropriate from Sub­ heading 3.2.2). 14. Add an equal volume of loading/termination buffer to each sample and place samples on ice. 15. Resolve truncated transcripts by electrophoresis as described in Subheading 3.2.4.

In vitro Transcription Assay for Resolution of Drug-DNA 3.2.7. Phosphor Imaging and Analysis

219

1. Place the dried gel in contact with a phosphor plate for 1 h (or up to overnight if necessary). 2. Scan the phosphor plate with a phosphorimaging system. 3. Analyse and quantitate the scanned image using an appropriate software program such as Image Quant. Normalise each transcript with respect to the total intensity in each lane to yield the percentage of each transcript in each reaction mixture. The sequencing lanes are used to determine the precise length of each transcript. The percentage of RNA blocked at each site is an indication of the relative occupancy at each site.

3.2.8. Drug Dissociation Kinetics

The rate of decay of drug-DNA interactions can be determined by monitoring their persistence as a function of time in the elongation phase of transcription. To optimise conditions for this purpose, a drug concentration that results in approximately 90% of full length transcripts (i.e., only approximately 10% of the total binding sites are occupied) must be chosen and a range of elongation times need to be chosen. After quantitation of blockage sites, In [RNA] vs. elongation time can be plotted. The results will be the most accurate for the initial blockage sites encountered since they are less affected by read-through of RNA polymerase from upstream (earlier) blockage sites. As an example, the stability of pixantrone-DNA adducts at a range of blockage sites has been determined and is shown in Fig. 6, and the half-lives at each site are summarised in Table 1.

Table 1 The stability of pixantrone-DNA adducts at discrete binding sites. The underlined nucleotide (non-template sequence) represents the site of each transcriptional blockage that was selected for quantitation in Fig. 6b (reprinted from Evison et al. (12), Copyright© 2008, by permission of the publisher American Society for Pharmacology and Experimental Therapeutics) Site

Sequence

Half-life (min)

52

ACGG

30

59

TCAC

40

80

ACAG

>180

108

ACGC

>180

120

TCGT

134

220

Evison, Phillips, and Cutts

4. Notes 1. The promoter selected for this assay must fulfil a number of suitable criteria. The promoter must accommodate a high fidelity of transcription from its start site and permit the initiated transcription complex to possess a half-life of at least several hours. Ideally, no additional transcription activating elements such as CAP and cAMP should be required. These requirements have been fully summarized previously along with a list of promoters that satisfy these criteria (4, 5). 2. It is important to use high purity sterile water to prepare all solutions for this procedure since the presence of trace amounts of metal ions, bacteria or nucleases can completely destroy transcription complexes. 3. Two methoxy nucleotides are generally sufficient for sequencing purposes. These should be chosen based on the expected sequence specificity of the agent under investigation. Alternatively, dideoxynucleotides can be utilised. 4. DTT has a limited half-life. Store in frozen aliquots and use a fresh aliquot for each experiment. 5. The exact MgCl2 concentration is critical to ensure that transcription proceeds efficiently and that natural pausing by the RNA polymerase is minimised. 6. The sequencing mixes are made up as 3× mixes and can be made for any of the four ribonucleotides. Dideoxynucleotides can be used as an alternative to methoxy nucleotides for sequencing reactions. However, a higher concentration is required to ensure adequate incorporation of the dideoxynucleotide. 7. Alternatively, use appropriate restriction enzymes or PCR primers to isolate the lac UV5 promoter from other sources as appropriate. 8. If ethidium bromide is included in the agarose gel, then single-strand nicks may be induced in the DNA, and this will result in a high background of truncated transcripts during the elongation phase of the transcription assay. 9. The Elutrap electroelution procedure has a high efficiency of recovery of DNA from agarose (typically greater than 95%). 10. Pierce the agarose slice with a pipette tip to inject a small amount of loading dye containing bromophenol blue before electroelution. Location of this dye in the electroelution trap after electroelution will help to assess that the process is complete. 11. Comparing the amount of full length transcript produced after three different time periods will allow selection of the

In vitro Transcription Assay for Resolution of Drug-DNA

221

optimal conditions for further experiments, and the repetition involved also allows experimenters to become competent with the technique before progressing to more complex experiments involving the transcription of multiple drugreacted DNA samples. 12. This procedure can be simplified by incubating the initiated transcript directly with drug as previously described (4, 5). This is mandatory if non-covalent drug-DNA interactions are being analysed. Some covalent drug-DNA interactions can also be assessed using this technique, but only if the initiated transcription complex is not damaged by the drug of choice. 13. Subsaturating levels of drug are ideal as this ensures that most drug binding sites are unoccupied, and subsequently a range of truncated transcript lengths that define different blockage sites can be obtained. If drug levels were saturating, transcription would terminate at the first blockage site encountered, thus revealing limited information. 14. The cleanup procedure chosen needs to be relevant to the drug of choice. The procedure is required to remove noncovalently bound drug that may interfere with transcription and other agents that are detrimental to subsequent formation of the transcription complex.

Acknowledgments We thank the Australian Research Council (ARC), National Health and Medical Research Council (NHMRC), and The CASS Foundation for funding our research.

References 1. Hampshire AJ, Rusling DA, Broughton-Head VJ, Fox KR (2007) Footprinting: a method for determining the sequence selectivity, affinity and kinetics of DNA-binding ligands. Methods 42:128–140 2. Murray VA (1999) A survey of the sequencespecific interaction of damaging agents with DNA: emphasis on antitumor agents. Prog Nucleic Acid Res Mol Biol 63:367–415 3. Portugal J (1989) Footprinting analysis of sequence-specific DNA-drug interactions. Chem Biol Interact 71:311–324 4. Phillips DR, Cutts SM, Cullinane CM, Crothers DM (2001) High-resolution transcription assay for probing drug-DNA interactions at

individual drug sites. Methods Enzymol 340: 466–485 5. Phillips DR, Cullinane CM, Crothers DM (1998) An in  vitro transcription assay for probing drug-DNA interactions at individual drug sites. Mol Biotechnol 10:63–75 6. White RJ, Phillips DR (1988) Transcriptional analysis of multisite drug-DNA dissociation kinetics: delayed termination of transcription by actinomycin D. Biochemistry 27: 9122–9132 7. Trist H, Phillips DR (1989) In vitro transcription analysis of the role of flanking sequence on the DNA sequence specificity of adriamycin. Nucleic Acids Res 17:3673–3688

222

Evison, Phillips, and Cutts

8. Panousis C, Phillips DR (1994) DNA sequence specificity of mitoxantrone. Nucleic Acids Res 22:1342–1345 9. Cullinane C, Phillips DR (1990) Induction of stable transcriptional blockage sites by adriamycin: GpC specificity of apparent adriamycin-DNA adducts and dependence on iron(III) ions. Biochemistr y 29: 5638–5646 10. Gray PJ, Cullinane C, Phillips DR (1991) In vitro transcription analysis of DNA alkylation by nitrogen mustard. Biochemistry 30: 8036–8040

11. Parker BS, Cutts SM, Phillips DR (2001) Cytosine methylation enhances mitoxantroneDNA adduct formation at CpG dinucleotides. J Biol Chem 276:15953–15960 12. Evison BJ, Chiu F, Pezzoni G, Phillips DR, Cutts SM (2008) Formaldehyde-activated Pixantrone is a monofunctional DNA alkylator that binds selectively to CpG and CpA doublets. Mol Pharmacol 74:184–194 13. Cullinane C, Phillips DR (1993) Thermal stability of DNA adducts induced by cyanomorpholinoadriamycin in vitro. Nucleic Acids Res 21:1857–1862

Chapter 14 In vitro Footprinting of Promoter Regions Within Supercoiled Plasmid DNA Daekyu Sun Abstract Polypurine/polypyrimidine (pPu/pPy) tracts, which exist in the promoter regions of many growth-related genes, have been proposed to be very dynamic in their conformation. In this chapter, we describe a detailed protocol for DNase I and S1 nuclease footprinting experiments with supercoiled plasmid DNA containing the promoter regions to probe whether there are conformational transitions to B-type DNA, melted DNA, and G-quadruplex structures within this tract. This is demonstrated with the proximal promoter region of the human vascular endothelial growth factor (VEGF) gene, which also contains multiple binding sites for Sp1 and Egr-1 transcription factors. Key words: Plasmid footprinting, DNA Secondary structure, G-quadruplex, VEGF

1. Introduction Polypurine/polypyrimidine tracts are known to exist at multiple sites in mammalian genomes, particularly in the proximal promoter regions of growth-related genes (1–7), including VEGF (Fig.  1). These cis-regulatory elements contain multiple Sp1binding sites, and several studies have independently reported the presence of DNase I- or S1 nuclease-hypersensitive sites within the regions of DNA harboring this tract in both chromatin and negatively supercoiled plasmid DNA, suggesting that this tract is structurally dynamic and easily converted into alternative conformations different from the typical B-DNA structure (8–11). In general, the structural transition of B-DNA to alternative secondary structures is preceded by the local melting or unwinding of duplex DNA, which is facilitated by a negative supercoiling stress naturally generated behind the translocating RNA polymerase complex during the transcription of the genes (12–14). Thus, the K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_14, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

223

224

Sun

Fig. 1. Schematic diagram showing the location of the pPu/pPy tract in the proximal promoter region of the VEGF gene.

structural transition of B-DNA to alternative non-B-conformations is believed to temporarily relieve a negative supercoiling stress generated under normal physiological conditions (15, 16). The proximal region of the VEGF promoter contains such a pPu/ pPy tract (Fig.  1), and this chapter uses this as an example for examining the structural dynamics, determining whether it can assume a number of different topological forms. 1.1. VEGF

Most primary solid tumors go through a dormant state in which the maximum attainable size is about 1–2 mm in diameter when the tumor cells use only preexisting host blood vessels (17, 18). However, the growth of new blood vessels from preexisting vessels by a process called angiogenesis allows the tumor cells to progressively expand and disseminate to distant organs (17, 18). Therefore, angiogenesis represents an essential step for tumor growth and metastasis by providing not only oxygen and nutrients to proliferating tumor cells, but also escape routes for metastatic tumor cells (17, 18). The switch to an angiogenic phenotype is mediated by a number of key regulators, such as fibroblast growth factors (FGFs), vascular endothelial growth factors (VEGFs) and angiopoietins, semaphorin, ephrin, Notch/Delta, and the roundabout/slit families of proteins (17, 18). Among them, VEGF (or VEGF-A) has been considered to be the key mediator of tumor angiogenesis by stimulating proliferation, migration, survival, and permeability of endothelial cells (19–21). VEGF expression is mainly regulated at the transcriptional level, and its expression is induced by a variety of factors, including hypoxia, pH, activated oncogenes, inactivated tumor suppressor genes, and growth factors (22–29). VEGF is frequently overexpressed in many types of cancer and the stable expression of VEGF appears to arise from increased VEGF promoter activity (20–30). The VEGF promoter region contains binding sites for several putative transcription factors, such as HIF-1, AP-1, AP-2, Egr-1, Sp1, and many others, suggesting that they may be involved in VEGF transcriptional regulation (3, 22–29).

In vitro Footprinting of Promoter Regions Within Supercoiled Plasmid DNA

225

Functional analysis of the human VEGF promoter using the fulllength VEGF promoter reporter revealed that the proximal 36-bp region (−85 to −50 relative to transcription initiation site) is essential for basal or inducible VEGF promoter activity in several human cancer cells (3). 1.2. Probes for Unusual DNA Structures

The presence of the pPu/pPy tract within the proximal region of the VEGF promoter (Fig. 1) led us to speculate that this region might be structurally dynamic and could potentially assume a number of different topological forms. Since our previous studies demonstrated that oligonucleotides representing the coding strands of this tract could adopt G-quadruplex structures (30), we further investigated the structural dynamics and the overall forms of the pPu/pPy tract within the promoter of the VEGF gene. We employed in vitro footprinting analysis utilizing a negatively supercoiled plasmid DNA containing this region in order to mimic the in vivo situation, where negative supercoils prevail due to the active DNA transaction (12–16, 30). The footprinting agents used in this study include DNase I, and S1 nuclease, since the reactivity of these probes is very sensitive to the conformation of DNA molecules (31, 32). These reagents have been utilized in many previous studies to probe structural transitions from B-DNA to non-B-type DNA structures, such as melted DNA, hairpin structures, G-quadruplex structures, and others (9–11). While DNase I preferentially cleaves locally unwound or normal duplex regions over single-stranded regions, S1 nuclease preferentially cleaves single-stranded regions of DNA over duplex DNA (9–11). However, both enzymes show the lowest cleavage activity toward highly organized secondary structures, such as hairpins or G-quadruplex structures (9–11, 31, 32). For these reasons, the combined use of both nucleases in in vitro footprinting experiments reveal pertinent information about unusual structural features of defined elements within the global region of DNA duplex molecules (31, 32).

2. Materials 2.1. Labeling 5¢-Termini of Nucleic Acids With [ 32P]

1. T4 polynucleotide kinase (Fermentas). 2. Kinase buffer (10×): 500  mM Tris-HCl (pH 7.6), 100  mM MgCl2, 50 mM DTT, 1 mM spermidine, and 1 mM EDTA. 3. Adenosine 5´-gamma 32P triphosphate (g-32P ATP), triethylammonium salt (6,000  Ci/mmole, 10  mCi/mL, GE, Healthcare). 4. Micro Bio-Spin™ 30 Columns (Bio-Rad).

226

Sun

2.2. Digestion of Plasmid DNA With Nucleases

1. DNase I (Promega). 2. S1 nuclease (Promega). 3. A supercoiled plasmid such as pGL3-V789 (see Note 1). 4. KCl buffer (10×): 1M KCl, 100 mM Tris-HCl (pH 7.4 at 25°C) (see Note 2). 5. Control buffer (10×): 100 mM Tris-HCl (pH 7.4 at 25°C).

2.3. Radioactive Cycle Sequencing and Linear Amplification

1. Thermo Sequenase DNA Polymerase (USB); 4  units/ml, 0.0006 units/ml Thermoplasma acidophilum inorganic pyrophosphatase; 50  mM Tris-HCl, pH 8.0, 0.1  mM EDTA, 1 mM dithiothreitol (DTT), 0.5% Tween®−20, 0.5% Igepal™ CA-630, 50% glycerol. 2. Reaction Buffer (concentrate): 260 mM Tris-HCl, pH 9.5, 65 mM MgCl2. 3. Telomestatin (see Note 3). 4. Gene-specific primer: 0.5  pmol/ml; (such as VEGF) 5¢CCCAGCGCCACGACCTCCGAGCTACC -3¢ (see Notes 4–5). 5. dNTP (10 mM) solution: 10 mM each dATP, dGTP, dCTP, dTTP (see Note 6). 6. ddG Termination Mix: 150 mM each dATP, dCTP, 7-deazadGTP, dTTP; 1.5 mM ddGTP. 7. ddA Termination Mix: 150 mM each dATP, dCTP, 7-deazadGTP, dTTP; 1.5 mM ddATP. 8. ddT Termination Mix: 150 mM each dATP, dCTP, 7-deazadGTP, dTTP; 1.5 mM ddTTP. 9. ddC Termination Mix: 150 mM each dATP, dCTP, 7-deazadGTP, dTTP; 1.5 mM ddCTP. 10. Stop Solution: 95% formamide, 20 mM EDTA, 0.05% bromophenol blue, 0.05% xylene cyanol.

2.4. Denaturing PAGE

1. TBE electrophoresis buffer (10×): 0.89M Tris, 0.89M boric acid, 20 mM EDTA, pH 8.0. Store at room temperature. 2. Sixteen percent acrylamide/bisacrylamide (29:1 with 3.3% C) with 8M urea and N,N,N¢,N¢- TEMED, Bio-Rad, Hercules, CA. 3. Ammonium persulfate: prepare 10% solution in water. Store at 4°C up to 1 month.

In vitro Footprinting of Promoter Regions Within Supercoiled Plasmid DNA

227

3. Methods In order to test if the pPu/pPy tract of a promoter region is structurally dynamic and can easily adopt a non-B-DNA conformation under physiological conditions, in vitro footprinting with DNase I and S1 nuclease is performed with a supercoiled form of a plasmid (such as pGL3-V789, which contains the VEGF promoter region from –727 to +50). This plasmid is incubated in the absence of any salt, or in the presence of 100 mM KCl to facilitate the evolution of the secondary structures from the pPu/pPy region. To test the binding of the G-quadruplex interactive agent to the secondary structures, the plasmid DNA is incubated with and without 1 mM telomestatin (33) for 1 h at 37°C, and then treated with DNase I or S1 nuclease for 2 min. To map S1 and DNase I cleavage sites, linear amplification by PCR was performed with 32P-labeled-gene-specific primers to amplify the top strand of both nucleases treated plasmid DNA. An overall strategy to perform in vitro footprinting of the wild-type VEGF promoter contained in a supercoiled plasmid in the presence of K+ and G-quadruplex-interactive compounds is shown schematically in Fig. 2.

Fig.  2. Flowchart of DNase I and S1 nuclease footprinting experiments of the VEGF promoter region in a supercoiled plasmid.

228

Sun

3.1. Isolation of Supercoiled Plasmids

The supercoiled plasmids are isolated from transformed E. coli strain DH5a using the QIAGEN Plasmid Maxi Kit. This method is based on modified SDS-alkaline lysis of bacterial cells in combination with selective binding of the DNA to silica beads in the presence of certain salts (see Note 7).

3.2. Treatment Plasmid DNA With Nuclease

1. In an empty tube, supercoiled plasmid (2  mg) and 2.5  ml of 10× KCl or control buffer are mixed and brought to 25  ml with the addition of DDW (see Note 8). 2. The incubation proceeds at 37°C for over 12 h or overnight to allow the secondary structures to evolve from the pPu/pPy region. 3. For testing the drug binding to the secondary structures, add 1 ml of diluted drug solution in KCl or control buffer (e.g., 25 mM telomestatin in KCl or control buffer) to 25 mL DNA solution and mix them by vortexing and centrifuge briefly. 4. Incubate the reaction mixture at 37°C for 1 h. 5. Add 2 mL of diluted DNase I (0.2 U) or 200 U of S1 nuclease to the tube, mix gently by pipetting up and down several times, cap the tubes and centrifuge briefly. After 1 min digestion, add 100 mL of 0.3M sodium acetate to the reactions followed by DNA precipitation with two volumes 100% ethanol and placement at −20°C overnight. 6. Spin in microfuge for 30 min and allow pellet to air dry. 7. Resuspend the dried pellet completely in 25 mL of TE buffer.

3.3. Labeling 5¢-Termini GeneSpecific Primers With [32P]

1. Prepare a reaction mixture (25  mL), containing oligonucleotide (4 mM), 3 mL g-32P ATP (6,000 Ci/mmole, 10 mCi/ mL), T4 polynucleotide kinase (10 U), 2.5 mL 10× kinase buffer, and water. 2. Incubate the reaction mixture at 37°C for 1 h in water bath for labeling 5¢-termini of oligonucleotides with g-[32P]-ATP. 3. After completion of the reaction, use Micro Bio-Spin™ 30 Columns (Bio-Rad) to remove unincorporated radioactive g-32P ATP (6,000  Ci/mmole, 10  mCi/mL) from labeled DNA. The instructions for use of Bio-Spin™ 30 Columns are based on recommendations from the manufacturer. In brief, the reaction mixture (25 mL) is loaded at the top of the column after centrifuging the column at 1,000× g for 4 min in a swinging bucket and removing the packing buffer. The column is then centrifuged for 4 min at 1,000× g to collect the purified 5¢-end-labeled oligonucleotide in water (see Note 9).

3.4. Radiolabeled Primer Cycle Sequencing

1. Label four tubes representing G, A, T and C. 2. Place 4 ml of the ddGTP termination mix in the tube labeled G. Similarly, fill the A, T and C tubes with 4 ml of the ddATP, ddTTP and ddCTP termination mixes respectively.

In vitro Footprinting of Promoter Regions Within Supercoiled Plasmid DNA

229

3. In a separate microcentrifuge tube, combine the following: 1  mL of plasmid DNA (20  pmole), 0.5  mL of concentrated reaction buffer, 1.0 mL of labeled primer, 1.0 mL of distilled water and 0.5 mL of Thermo Sequenase DNA Polymerase (see Note 10). 4. Transfer it to the PCR tube (from step 1), mix gently by pipetting up and down several times, cap the tubes and place them in the thermal cycler (see Note 11). 5. Carry out PCR using cycling conditions consisting of an initial 10-min denaturation step at 94°C, 1 min at 60°C, and 1 min at 72°C, for a total of 42 cycles (see Note 12). 6. Add 4 ml of stop solution to each of the termination reactions, mix thoroughly and centrifuge briefly. 3.5. Linear Amplification of the Plasmid DNA Digested With DNase I or S1 Nuclease Using 32 P-Labeled Primers

1. Place 4 ml of the plasmid DNA digested with DNase I or S1 nuclease from Subheading 3.2 in each PCR tube. 2. In a separate microcentrifuge tube, combine the following: 0.5  mL of 10  mM dNTP solution, 0.5  mL of concentrated reaction buffer, 0.5 mL of labeled primer, 2.0 mL of distilled water and 0.5 mL of Thermo Sequenase DNA Polymerase (see Note 9). 3. Transfer it to the PCR tube (from step 1), mix gently by pipetting up and down several times, cap the tubes and place them in the thermal cycler (see Note 13). 4. Carry out PCR using cycling conditions consisting of an initial 10-min denaturation step at 94°C, 1 min at 60°C, and 1 min at 72°C, for a total of 42 cycles (see Note 14).

3.6. Separation of PCR Products on Denaturing PAGE

1. Set up a denaturing 10% polyacrylamide gel of 30 cm × 30 cm × 0.4 mm. 2. Prepare 60  mL of gel solution by mixing 6  mL TBE buffer (10×), 15  mL of 40% acrylamide/bisacrylamide (29:1), and 30 g urea and adding water to 60 mL. After adding 100 mL ammonium persulfate solution and 20 mL TEMED, pour the gel and insert the comb. 3. Once the gel is polymerized, carefully remove the comb, and wash the well with TBE buffer (1×) using a pasture pipette. 4. Attach the gel plates to the electrophoresis apparatus, and fill both reservoirs of the electrophoresis tank with 1× TBE. Prerun and warm the gel for at least 30 minutes at 1,400 V (constant voltage) using a DC power supply. 5. Heat the samples and sequencing ladders at 95°C for 3 min, and chill the sample on ice before loading. Run the gel at about 1,400 V.

230

Sun

6. After the desired resolution is obtained, detach the gel plates from the electrophoresis apparatus, and carefully separate both plates, leaving the gel attached to one plate. 7. Place a piece of thin chromatography paper (DE81) on top of the gel, and slowly pull back on the paper to transfer gels to the paper. 8. Place a piece of Whatman paper (3MM) underneath, and cover the wet gel with plastic wrap on top. 9. Put the gel sandwich in a dryer between a plastic fiber mat and clear plastic sheet, and dry the gel at 80°C for at least 1 h with a vacuum. 10. Place the dried gel in an X-ray film cassette. Obtain an autoradiogram by exposing the X-ray film to the dried gel. Alternatively, the image can be obtained by exposing the dried gel to the phosphor screen for an appropriate time and scanning the phosphor screen. Fig.  3 is an example of an

Fig. 3. In vitro footprinting of the VEGF promoter region with DNase I, S1 nuclease or DMS. Autoradiograms showing S1 nuclease and DNase I cleavage sites on the G-strand of a supercoiled pGL3-V789 plasmid. This plasmid was incubated in control (lane 1), or in KCl buffer without (lane 2), and with 1 mM telomestatin (lane 3) at 37°C for 1 h before digesting with nucleases. Arrows indicate the hypersensitive cleavage sites to nucleases. The primer extension reaction revealed a long protected region at approximately −53 to −123 bp, including the G-rich sequences, when a supercoiled pGL3V789 plasmid was incubated with 100 mM KCl and digested with DNase I (compare lanes 1 and 2). This indicates a possible transition from B-DNA to a G-quadruplex structure in the VEGF promoter region, which is consequently resistant to DNase I digestion. Significantly, a striking hypersensitivity was found in the presence of KCl and telomestatin at a cytosine located at the 3¢-side of the G-quadruplex-forming region (underlined sequence), which is the junction site separating the putative G-quadruplex from the adjacent normal B-DNA (see arrow “A” in lane 3). The reactivity of S1 nuclease at the VEGF proximal promoter region was also moderately reduced in the presence of telomestatin and KCl, and the hypersensitivity site observed with S1 nuclease corresponds to one of those obtained with DNase I in the presence of telomestatin and KCl (lane 3) (Figure modified from ref (30)).

In vitro Footprinting of Promoter Regions Within Supercoiled Plasmid DNA

231

autoradiogram of a 10% polyacrylamide sequencing gel, showing the results of S1 and DNase I footprinting experiments carried out with a supercoiled pGL3-V789 plasmid.

4. Notes 1. Plasmid pGL3-V789 was originally constructed by Dr. Keping Xie by subcloning a 789  bp fragment containing 5¢ VEGF promoter sequences from −729 to +50 relative to the transcription initiation site into the KpnI and NheI sites of pGL3-basic (Promega, Madison, WI), which contains firefly luciferase coding sequences (3). 2. KCl Buffer provides optimum conditions for the formation of G-quadruplex structures from the single stranded DNA. 3. Telomestatin was kindly provided by Dr. Kazuo Shin-ya. 4. It is also a good idea to check the sequence of the primer for possible self-annealing (dimer formation could result) and for potential “hairpin” formation, especially those involving the 3¢-end of the primer. 5. Finally, check for possible sites of false priming in the vector or other known sequence if possible, again stressing matches which include the 3¢-end of the primer. 6. All enclosed reagents should be stored frozen at −20°C and keep all reagents on ice once removed from storage for use. 7. The protocol is suitable for obtaining pure plasmid DNA up to 100  mg from 30~100  ml bacterial culture grown in LB medium. The culture volume should be reduced to half or less when bacteria grown in rich medium are used. 8. It is best to prepare one large reaction mix and then aliquot 25 ml into each sample tube. 9. No further purification is required for most sequencing. 10. Use 1.0 pmol of fresh 32P-labeled primers, but two- to fivefold more primer can be used for shorter exposure times. 11. It is best to prepare one large reaction mix and then aliquot 4 ml into each sample tube. 12. The specific cycling parameters used will depend on the primer length and sequence and the amount and purity of the template DNA. 13. It is best to prepare one large reaction mix and then aliquot 4 ml into each sample tube. 14. If your gels seem to require longer exposures, and more template is not available, increase the number of PCR cycles.

232

Sun

Acknowledgments This research was supported by grants from the National Institutes of Health (CA109069). We are grateful to Drs. Allison Hays and Keith Fox for proofreading and editing the final version of the manuscript and figures. We also thank Drs Keping Xie and Kazuo Shin-ya for providing pGL3-V789 and telomestatin, respectively, for this study. References 1. McCarthy JG, Heywood SM (1987) A long polypyrimidine/polypurine tract induces an altered DNA conformation on the 3¢ coding region of the adjacent myosin heavy chain gene. Nucleic Acids Res 15:8069–8085 2. Michelotti GA, Michelotti EF, Pullner A, Duncan RC, Eick D, Levens D (1996) Multiple single-stranded cis elements are associated with activated chromatin of the human c-myc gene in  vivo. Mol Cell Biol 16:2656–2669 3. Shi Q, Le X, Abbruzzese JL, Peng Z, Qian CN, Tang H et  al (2001) Constitutive Sp1 activity is essential for differential constitutive expression of vascular endothelial growth factor in human pancreatic adenocarcinoma. Cancer Res 61:4143–4154 4. Rustighi A, Tessari MA, Vascotto F, Sgarra R, Giancotti V, Manfioletti G (2002) A polypyrimidine/polypurine tract within the Hmga2 minimal promoter: a common feature of many growth-related genes. Biochemistry 41:1229–1240 5. Cogoi S, Xodo LE (2006) G-quadruplex formation within the promoter of the KRAS protooncogene and its effect on transcription. Nucleic Acids Res 34:2536–2549 6. De Armond R, Wood S, Sun D, Hurley LH, Ebbinghaus SW (2005) Evidence for the presence of a guanine quadruplex forming region within a polypurine tract of the hypoxia inducible factor 1a promoter. Biochemistry 44:16341–16350 7. Guo K, Pourpak A, Beetz-Rogers K, Gokhale V, Sun D, Hurley LH (2007) Formation of pseudosymmetrical G-quadruplex and i-motif structures in the proximal promoter region of the RET oncogene. J Am Chem Soc 129:10220–10228 8. Pullner A, Mautner J, Albert T, Eick D (1996) Nucleosomal structure of active and inactive c-myc genes. J Biol Chem 271: 31452–31457

9. Wang Z, Lin XH, Qiu QQ, Deuel TF (1992) Modulation of transcription of the plateletderived growth factor A-chain gene by a promoter region sensitive to S1 nuclease. J Biol Chem 267:17022–17031 10. Siebenlist U, Henninghausen L, Battey J, Leder P (1984) Chromatin structure and protein binding in the putative regulatory region of the c-myc gene in Burkitt lymphoma. Cell 37:381–391 11. Evans T, Efstratiadis A (1986) Sequencedependent S1 nuclease hypersensitivity of a heteronomous DNA duplex. J Biol Chem 261:14771–14780 12. Benham CJ (1985) Theoretical analysis of conformational equilibria in superhelical DNA. Ann Rev Biophys Biophysical Chem 14:23–45 13. Liu LF, Wang JC (1987) Supercoiling of the DNA template during transcription. Proc Natl Acad Sci USA 84:7024–7027 14. Williams DL, Kowalski D (1993) Easily unwound DNA sequences and hairpin structures in the Epstein-Barr virus origin of plasmid replication. J Virol 67:2707–2715 15. Kouzine F, Levens D (2007) Supercoil-driven DNA structures regulate genetic transactions. Front Biosci 12:4409–4423 16. Kouzine F, Sanford S, Elisha-Feil Z, Levens D (2008) The functional response of upstream DNA to dynamic supercoiling in  vivo. Nat Struct Mol Biol 15:146–154 17. Folkman J (2002) Role of angiogenesis in tumor growth and metastasis. Semin Oncol 29:15–18 18. Sullivan DC, Bicknell R (2003) New molecular pathways in angiogenesis. Br J Cancer 89:228–231 19. Martiny-Baron G, Marme D (1995) VEGFmediated tumour angiogenesis: a new target for cancer therapy. Curr Opin Biotechnol 6:675–680

In vitro Footprinting of Promoter Regions Within Supercoiled Plasmid DNA 20. Goodsell DS (2003) The molecular perspective: VEGF and angiogenesis. Stem Cells 21:118–119 21. Jain RK (2002) Tumor angiogenesis and accessibility: role of vascular endothelial growth factor. Semin Oncol 29:3–9 22. Gunningham SP, Currie MJ, Han C, Turner K, Scott PA, Robinson BA et  al (2001) Vascular endothelial growth factor-B and vascular endothelial growth factor-C expression in renal cell carcinomas: regulation by the von Hippel-Lindau gene and hypoxia. Cancer Res 61:3206–3211 23. Schafer G, Cramer T, Suske G, Kemmner W, Wiedenmann B, Hocker M (2003) Oxidative stress regulates vascular endothelial growth factor-A gene transcription through Sp1- and Sp3-dependent activation of two proximal GC-rich promoter elements. J Biol Chem 278:8190–8198 24. Maeno T, Tanaka T, Sando Y, Suga T, Maeno Y, Nakagawa J et al (2002) Stimulation of vascular endothelial growth factor gene transcription by all trans retinoic acid through Sp1 and Sp3 sites in human bronchioloalveolar carcinoma cells. Am J Respir Cell Mol Biol 26:246–253 25. Chen H, Ye D, Xie X, Chen B, Lu W (2004) VEGF, VEGFRs expressions and activated STATs in ovarian epithelial carcinoma. Gynecol Oncol 94:630–635 26. Pal S, Datta K, Khosravi-Far R, Mukhopadhyay D (2001) Role of protein kinase Czeta in Rasmediated transcriptional activation of vascular permeability factor/vascular endothelial growth factor expression. J Biol Chem 276: 2395–2403

233

27. Tanaka T, Kanai H, Sekiguchi K, Aihara Y, Yokoyama T, Arai M et al (2000) Induction of VEGF gene transcription by IL-1 beta is mediated through stress-activated MAP kinases and Sp1 sites in cardiac myocytes. J Mol Cell Cardiol 32:1955–1967 28. Finkenzeller G, Sparacio A, Technau A, Marme D, Siemeister G (1997) Sp1 recognition sites in the proximal promoter of the human vascular endothelial growth factor gene are essential for platelet-derived growth factor-induced gene expression. Oncogene 15:669–676 29. Forsythe JA, Jiang BH, Iyer NV, Agani F, Leung SW, Koos RD et al (1996) Activation of vascular endothelial growth factor gene transcription by hypoxia-inducible factor 1. Mol Cell Biol 16:4604–4613 30. Sun D, Guo K, Rusche JJ, Hurley LH (2005) Facilitation of a structural transition in the polypurine/polypyrimidine tract within the proximal promoter region of the human VEGF gene by the presence of potassium and G-quadruplex-interactive agents. Nucleic Acids Res 33:6070–6080 31. Saluz HP, Jost JP (1993) Approaches to characterize protein-DNA interactions in  vivo. Crit Rev Eukaryot Gene Expr 3:1–29 32. Dabrowiak JC, Goodisman J, Ward B (1997) Quantitative DNA footprinting. Methods Mol Biol 90:23–42 33. Kim MY, Vankayalapati H, Shin-Ya K, Wierzba K, Hurley LH (2002) Telomestatin, a potent telomerase inhibitor that interacts quite specifically with the human telomeric intramolecular G-quadruplex. J Am Chem Soc 124: 2098–2099

Chapter 15 Topoisomerase I-Mediated DNA Relaxation as a Tool to Study Intercalation of Small Molecules into Supercoiled DNA Paul Peixoto, Christian Bailly, and Marie-Hélène David-Cordonnier Abstract Several biochemical and biophysical methods are available to study the intercalation of a small molecule between two consecutive base pairs of DNA. Among them, the topoisomerase I-mediated DNA relaxation assay has proved highly efficient, relatively easy to handle and very informative to investigate drug binding to DNA. The test relies on the use of a supercoiled plasmid to mimic the topological constraints of genomic DNA. The three main components of the assay – the topoisomerase I enzyme, DNA helix and intercalating small molecules – are presented here in a structural context. The principle of the assay is described in detail, along with a typical experimental protocol. Key words:  DNA topoisomerase, DNA relaxation, Intercalation process, Drug/DNA binding, Mechanism of action

1. Introduction The method presented here has been routinely used over many years to characterise the interaction of small molecules with DNA, in particular anticancer agents. The relaxation assay using topoisomerase I is one of the most robust approaches to evidence intercalation of small molecules into DNA. It is based on a process naturally occurring in every living cell. Here, we describe the typical experimental protocol and, to begin with, the players and the principle of the assay. 1.1. DNA Topoisomerase I

DNA is a long-established anticancer target. Many vital cellular processes, such as DNA replication and repair, transcription, chromosome aggregation and segregation, are highly active in proliferative

K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_15, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

235

236

Peixoto, Bailly, and David-Cordonnier

tumour cells. These processes induce extensive constraints and tangles in the DNA helix that are resolved by specific proteins capable of manipulating DNA structures: the DNA topoisomerases (1–8). Topoisomerase forms a protein/DNA complex in a 1:1 (topoisomerase I) or 2:1 (topoisomerase II) stoichiometry, requiring ATP and the presence (topoisomerase II) or absence (topoisomerase I) of divalent cations (Mg2+), cutting one (topoisomerase I) or both strands (topoisomerase II) of the DNA helix. The enzymatic cycle for topoisomerase I is presented in Fig. 1A and B. Structurally, topoisomerase I is subdivided into four domains, namely I, II, III and IV (only domains I to III are presented in Fig. 1C) and a linker helical domain (L) (9–12). After DNA recognition (step a in Fig. 1A and B), a transesterification reaction links the reactive tyrosine residue of the protein (Y723, localised in domain III for human topoisomerase IB, see Fig. 1B) to the 3¢-phosphate end of a DNA fragment (steps a–b). This reaction leads to the formation of a covalent topoisomerase–DNA complex (step b), referred to as the “cleavable complex” that slackens off local DNA constraints at the free strand. Subsequently, a reversible nicking reaction (stage c) occurs from a second nucleophilic attack between the free 5¢-OH group (for topoisomerase I) of the opened strand and the 3¢-phosphotyrosyl bond, to give back a native and relaxed DNA helix (step d) and a functional free topoisomerase I enzyme, ready to initiate a new enzymatic cycle. Structurally, topoisomerase IB surrounds DNA like a hand with its fingers tightly clamped around the DNA helix (Fig. 15.1C) and the arm being the linker domain L. Topoisomerase IB cuts the DNA on one strand, links to the 3¢-end, allowing the revolution of the 5¢ part of the open-strand over the non-cut strand, acting as a swivel. Then, 3¢- and 5¢-ends are religated to give back the full-length double-stand DNA helix, without the loss of genetic information. From X-ray data, the “fingers” surround the DNA downstream from the scissile strand very closely and are maintained closed by a disulphide bond so that to block full rotation of the DNA strand. In this way, no extensive opening of the protein clamp can occur, suggesting that only small shifts in the orientation of the cap and linker domains accommodate with full rotation of the open strand (13). Consequently, unwinding of the DNA by topoisomerase IB does not lead to a total release of all

Fig. 1. (continued) to the protein, DNA interaction step; b, cleavage of the DNA from a transesterification process leading to the formation of the cleavable complex; c, religation of the DNA resulting from the second transesterification process and d, dissociation of the topoisomerase I, DNA complex leading to the release of the functional enzyme and relaxed DNA (compare constraint and relaxed DNA helix in a and c steps on the bottom panel). The structure of topoisomerase I corresponds to that obtained from the 70-kDa protein lacking part of the C-terminal domain. (C) Topoisomerase binding to DNA matching the position of a hand with fingers holding tight the helix. The position of the gap between the two parts of the surrounding hand is presented as an arrow. Structural subdomains I, II, III and linker domain L are localised.

Topoisomerase I-Mediated DNA Relaxation as a Tool to Study Intercalation

237

Fig. 1. Topoisomerase I enzymatic reaction cycle and global structure. The reaction cycle is presented as a typical cycle (A) or as the chemical transesterification process (B). Four different reaction steps are identified: a, recognition corresponding

238

Peixoto, Bailly, and David-Cordonnier

supercoils in one step, but is a multi-step phenomenon with the “hand” holding tight and then releasing the DNA at each step. The resulting number of released supercoils is, however, not linear, but follows an exponential distribution as a consequence of the torque sensitivity of topoisomerase IB: the more twisted the DNA, the higher number of released supercoils (14). This phenomenon is the basis for topoisomerase-induced DNA relaxation studies to identify the intercalative properties of a DNA-binding compound, based on the helicity of the DNA and the constraints produced by the DNA intercalation of a small molecule between two adjacent base pairs. 1.2. DNA: The Properties of a Helix

1.3. Intercalation of Small Compounds into the DNA: Structural Consequences

In cells, the DNA helix is mainly organised as a right-handed helix (Fig. 2a). This “right-hand screw” has however a complex geometry at the origin of the internal curvature and flexibility of the DNA (15–18). Rotation over the three geometrical axes can occur (Fig.  15.2b), leading to multiple structural movements such as twist (W), tilt (t), roll (r) and slide (s) effects of the base steps (see Fig.  2c). These parameters directly depend on the nature of the bases following each other along the helix, but also indirectly on the presence and nature of salts or proteins bound to the DNA. Indeed, the degree of propeller twist is higher in A·T than in G·C base pairs and follows from the flexibility of hydrogen bonds between base pairs to allow better overlapping of the bases from the van der Waals stacking. Overall, each type of base steps present a different set of W, r and s values (16, 19–24) that are also affected by the nature of both 5’- and 3’-bases, widening the variability in structure conformations of the DNA helix. W is ~30° for ApT step, contrasting with W of ~40° for TpA step, exemplifying the zigzag conformation of alternative Y/R and R/Y base pairs DNA (25); runs of adenines present typical propeller twist with only poor W, r and s variations leading to stiff DNA (26, 27) with a full helix turn obtained with only 10.0 bp instead of 10.6 for mixed-bases DNA. GG, CG and GC steps present the highest s and lowest W values, and are therefore more flexible than AA and AT steps belonging to the rigid subgroup with lowest s and highest W effects (17, 28–32). By contrast, W and s values for TA step are closer to that for GG, CG, GC than for AT or AA steps. Topoisomerase I exquisitely manipulates the various W/t/ r/s parameters in DNA helix. Many small molecules, particularly (but not exclusively) those containing a flat aromatic chromophore, can insert between two consecutive base pairs as a DNA intercalation process that produces specific DNA constraints. As a helical stair representation of the DNA, with base pairs being the steps, the space between two

Topoisomerase I-Mediated DNA Relaxation as a Tool to Study Intercalation

239

Fig. 2. DNA helix and helical turns. (a) Right-handed DNA as a typical B-form with the medium 36° of twisting between adjacent base pairs. The helicity of the DNA is presented for one DNA strand as an arrow in front of (black arrow) or behind (grey arrow) the successive base pairs. (b) Illustration of the tri-dimensional coordinates (x, y and z) for base pairs movements. The base pairs are presented as a rectangle. Each side of the base pairs corresponds to a different colour on the frame. (c) Examples of the orientation changes in DNA base pairs by slipping (slide) along the y axis or rotation along the x (tilt), y (roll) or z (twist) axes. (d) Examples of slide (s) and roll (r) effects. This figure (adapted from Calladine and Drew (15)) evidences the global torsion of the DNA structure visualised by the DNA axis (central grey lane) and DNA strand distortions (black, grey arrows) and concomitant reduction of the length of the DNA fragment when both slide and roll effects are applied (right panel).

successive base pairs is around 0.34  nm (3.4 Å). Subsequent incorporation of a planar molecule in this space, sandwiched between the two base pair raises the height of the step to 0.68 nm, at least in theory (Fig. 3).

240

Peixoto, Bailly, and David-Cordonnier

Fig. 3. DNA relaxation upon mono- or bis-intercalation. Normal DNA helix (a), mono-intercalation (b) or bis-intercalation of the DNA (c). The intercalating chromophores (Interc) are presented in blue. Their respective positioning between adjacent base pairs (−1 and +1) induces a different twisting angle (x° or y°) that is specific of the nature of the intercalating, bis-intercalating compound. The intercalation, bis-intercalation processes lead to an increase in the length of the DNA visualised by a brace (D length) and an unwinding of the DNA helix as visualised by the arrow on the top of each panel.

Different examples of DNA intercalating drugs are given in Fig.  4. The planar chromophores could have different orientations relative to the base pair axis from perpendicular (33–37) to parallel (38–41) (Fig. 4). The intercalation process occurs preferentially between alternating base pairs (preferentially CpG/CpG), but can also occur between identical base pairs as reported with cryptolepine easily intercalated between two successive C·G base pairs (CpC/GpG site) (42). DNA intercalation produces an unwinding of the helix that can be measured and varies depending on the structure of the intercalating agent (Fig.  3): the mono-intercalators ethidium bromide and actinomycin D induce an unwinding angle of 26°, close to that measured with methylene blue (24°C) (43), whereas the anthracycline derivative daunorubicin or imide derivatives of 3-nitro-1,8-naphthalic acid unwind the DNA less profoundly with a smaller angle of 11° (33, 44). Bis-intercalators bearing two planar chromophores can double (or nearly) the unwinding value (Fig. 3), for example, that for the bis-intercalator echinomycin is 1.82 ± 0.30 times that of ethidium bromide (45, 46); the synthetic bis-quinoxaline drug TANDEM extensively unwinds DNA with an unwinding angle of 45° (47, 48) and the bis-naphthalimide drug elinafide (LU79553) by 37° (49, 50). Typically, the intercalation process differs from the groove binding mode by a weaker binding constant (ranging from 10−4

241

Topoisomerase I-Mediated DNA Relaxation as a Tool to Study Intercalation

–1 CH3 +N

Interc

NH2

N H

+1

N

Cryptolepine perpendicular O

parallel

N

HN

OCH3

COCH2R OH OH

OCH3 O

Cl

N

O

–

Quinacrine

Br

CH3

OH NH3 H3C

Daunomycin R = H Adriamycin R = OH

+

CH3

NH2

Ethidium bromide

CH3

Methylene blue

O

N

N O

2-methyl-9-hydroxyellipticinium

N

S

CH3

N+

N H

+

N

N

H2N

N

+

HO

9-amino-DACA (acridine-carboxamide)

CH3

OH

+ N

N H

O

O

O

NH N

NH O

O

N

O

N

NH2

O

O

N O

N

O H N

CH3

O

HN

O O

N

N H

CH3

Ellipticine

Actinomycin D Fig. 4. Orientation of the intercalating chromophore in DNA helix. The intercalating compounds could be positioned parallel (along the y-axis) or perpendicular (along the x-axis) relatively to the adjacent base pairs. In perpendicular orientation, parts of the intercalating molecule could protrude in either the minor or the major groove of the DNA helix.

to 10−6 M−1 and 10−5 to 10−9 M−1, respectively) and high rates for association/dissociation constants (for example, anthraquinone derivatives stay for less than a millisecond in its intercalative site (51)). But of course, there are always exceptions to the rule; highaffinity intercalators and low-affinity groove binders do exist.

2. Materials 2.1. Plasmid Purification on CsCl Gradient

Circular plasmid DNA is a very useful template to study the activity of topoisomerase I, drugs inhibiting topoisomerase I or compounds modifying the DNA topology such as DNA intercalators. The global folding of closed circular duplex DNA depends not only on the slide/roll/twist of the base pairs and on bending/

242

Peixoto, Bailly, and David-Cordonnier Sc Topo I Interc

electrophoretic migration

well 0 –1 –2 –3 –4 –5 –6 –7 –8 negative Right-handed supercoils

– –

0 +1 +2 +3 +4 +5 +6 +7 +8

+ +

positive Left-handed supercoils

Fig. 5. Schematic representation of the DNA relaxation process of supercoiled plasmid DNA. Migration on an agarose gel in the absence of ethidium bromide reveals different topoisomers (illustrated for eight different possible topoisomers) generated by the cleavage, religation cycle of topoisomerase I. The supercoiled plasmid DNA (Sc), presented as a control of migration present underwound helical turns (negative or right-handed supercoils). Supercoiled plasmid incubated with increasing concentration on intercalating agent (“Interc”) and then treated with topoisomerase I (lanes “Topo I”) reveals the transition from negative topoisomers (numbered and illustrated on the left of the panel) to fully relaxed DNA (“0”) and then to positively (left-handed) supercoiled DNA (numbered and illustrated on the right of the panel).

twisting of the helix but also on the number of supercoils, also referred as the linking number (Lk, where Lk = T + W, T being the twist and W the writhe of the DNA double helix) and corresponding to the number of times the two strands of the helix are wrapped around one another before resealing. It is important to bear in mind that the supercoiled DNA can be underwound (negative supercoils) or overwound (positive supercoils). The three-dimensional representations of supercoiled DNA at minimal energy define different organisations with, for example, linear inter-wound (as presented in Fig. 5, see “-1” form), asymmetric three lobe branches or trefoil structure (17, 52–56). In bacteria, the plasmids are negatively supercoiled (57) with up to one superhelical turn per 200–250 bp (58). Various double-strand supercoiled plasmids could be used as substrates. However, the greater the number of base pairs, the greater the number of topoisomers and, as a consequence, plasmids containing between 2,500 and 3,500 bp present a sufficient number of topoisomers and should be preferred to vectors of 4,000 or more base pairs, which are usually more difficult to manipulate (requiring lower resolution on gels).

Topoisomerase I-Mediated DNA Relaxation as a Tool to Study Intercalation

243

The classical pUC18 or pUC19 plasmids (2,686 bp; Fermentas GMBH, St Leon-Rot, Germany), pBluescript-II-SK/KS (2,961 bp; Stratagene), (pSP64/65, 2,999 bp; Promega), pGEM-T (3,000 bp; Promega) or pBS(−) (3,204 bp; Stratagene) are typical examples of routinely used templates for topoisomerase I-induced DNA relaxation assays. In order to get an enrichment of the supercoiled form of the plasmid DNA over the nicked or linear forms, the method of purification is important. 2.1.1. Amplification of the Plasmid in Bacteria

– LB (Luria Bertoni) medium: peptone 10 g/L; yeast extract 5 g/L, NaCl 86 mM; pH 7.5; sterilised. – Antibiotic: ampicillin 0.5 mg/mL final solution, for the plasmid vectors presented above. – A shaker with the temperature fixed at 37°C for 500 mL culture bacteria vials. – A centrifuge for collecting the bacteria from medium with 200 mL centrifuge vials and an appropriate rotor.

2.1.2. Lysis of Bacteria

– Buffer 1: 25 % saccharose in 50 mM Tris–HCl, pH 8.0. – Lysozyme : stock at 10 mg/mL. – Buffer 2: 50 mM Tris–HCl, pH 8.0, EDTA 20 mM, Triton X-100 0.1%.

2.1.3. Elimination of Bacterial Remnants from Lysis.

– Swinging rotor such as the SW-41 Ti rotor from Beckman Coulter.

2.1.4. Separation of Plasmid DNA from Bacterial Genomic DNA

– Ethidium bromide stock solution: 10 mg/mL.

2.1.5. Recovery of the Plasmid

– G45 needle.

– 9 g of CsCl (precisely weighted). – Ultracentrifugation system and fixed angle rotor (such as the TI80 rotor from Kontron).

– 2 mL syringe. – UV transilluminator (366 nm).

2.1.6. Removal of Ethidium Bromide

– Tris/EDTA dialysis buffer: 10 mM Tris–HCl, 1 mM EDTA, pH 8.0. – Buffer BPE: 6 mM Na2HPO4, 2 mM NaH2PO4, 1 mM EDTA, pH 7.4. – UV spectrophotometer coupled to a thermostated element (for example an UVIKON 943 spectrophotometer coupled to the Neslab RTE111 temperature controller system). – 1-cm path length quartz cuvettes.

244

Peixoto, Bailly, and David-Cordonnier

2.1.7. Alternative Plasmid Purification System

– Purification columns: NucleoBond PC (Machery-Nagel Gmbh, Germany) or HiSpeed Plasmid Kits (Qiagen SA, France), for examples.

2.2. Agarose Gel

– Agarose: Ultra-pure agarose powder (Qbiogene). – 1X TBE buffer: 89-mM Tris–borate pH 8.3, 1 mM EDTA. – Electrophoresis apparatus with small (11 × 14  cm) or large (20 × 25 cm) electrophoresis sets. – Microwave. – Heat-resistant adhesive tape.

2.3. Topoisomerase I-Induced Relaxation of Supercoiled Plasmid in the Presence of Increasing Concentrations of the Test Compound 2.3.1. Drug/DNA Interaction

– Topoisomerase I reaction buffer: 10 mM Tris–HCl, pH 7.9, 150 mM NaCl, 0.1% BSA, 0.1 mM spermidine.

2.3.2. Topoisomerase I-Induced Relaxation

– Commercially available Topoisomerase I enzyme (for example from Topogen, Colombus, OH).

2.3.3. Degradation of the Protein

– SDS stock solution: SDS powder (Sigma) 10% in water.

2.3.4. Separation of the Forms of DNA on Agarose Gel

– Electrophoresis dye mixture: saccharose 40%, bromophenol blue 0.25%, EDTA 1 mM ; pH 8.0.

2.3.5. Staining of the Gel Using Ethidium Bromide

– Ethidium bromide (Sigma): 10 mg/mL.

2.3.6. Visualisation of the DNA Forms

– UV transilluminator coupled with a camera.

– Proteinase K stock solution: 5 mg/mL of proteinase K powder (from Sigma).

– 1 Kb molecular weight DNA ladder.

– A bath tray reserved for ethidium bromide staining.

3. Methods The following part describes the different steps of a so-called “DNA relaxation assay” aimed at characterising the effect of a small molecule on DNA structure, using topoisomerase I as a conformation-

Topoisomerase I-Mediated DNA Relaxation as a Tool to Study Intercalation

245

sensitive sensor. Typically, supercoiled plasmid DNA at a fixed concentration is incubated with graded concentrations of a test compound prior to being relaxed using a fixed amount of topoisomerase I. The various forms of DNA resulting from the relaxation reaction are then resolved on a native agarose gel (without ethidium bromide) and the DNA bands are visualised after electrophoresis by staining the gel in a solution of ethidium bromide. An illustration of the expected forms (and bands) of the plasmid DNA after gel electrophoresis is presented in Fig. 5. The different parts of the protocol are exemplified below. 3.1. Plasmid Purification on CsCl Gradients

The aim of this method is to purify a high quantity with the highest quality of supercoiled plasmid DNA from transformed bacterial cultures, mainly Escherichia coli. Usually, this method gives 90–98% of supercoiled DNA (usually referred as form I), the remaining 2–10% correspond to the plasmid nicked on one of the two strands of the DNA helix (open circular DNA, usually referred as form II) or on both two strands (linear DNA, also referred as form III).

3.1.1. Amplification of the Plasmid in Bacteria

One transformed bacterial clone is cultured at 37°C under agitation in 10  mL of LB medium (supplemented with appropriate antibiotic for 8 h); this pre-culture is then added to a larger flask containing 200  mL of LB medium plus antibiotic for selection (specific to the plasmid to be amplified). The bacteria are grown at 37°C in the shaker for 18–24 h and then collected by centrifugation at 4,000 g for 15 min at 4°C.

3.1.2. Lysis of Bacteria

The pellet is then resuspended in 2.5 mL of buffer 1 and the bacterial membrane lysed by the addition of 1 mL of lysozyme stock solution. This is incubated for 10 min on ice until the suspension becomes slightly viscous, followed by the addition of 4.5 mL of buffer 2. The solution is mixed strongly to get a very viscous solution and incubated for a further 20 min.

3.1.3. Elimination of Bacterial Remnants from Lysis

A centrifugation at 10,000  g for 30 min at 4°C in a swinging rotor (such as the SW-41 Ti rotor from Beckman Coulter) is performed to remove the lysed bacterial walls.

3.1.4. Separation of Plasmid DNA from Bacterial Genomic DNA

The cleared supernatant (9 mL) is transferred to a fresh tube containing 9 g of CsCl (precisely weighted). After dissolution, 45 µl of ethidium bromide (10 mg/mL) are added. Ethidium bromide has long been used as a potent DNA binder (59) and its intrinsic fluorescence is strongly enhanced when bound to DNA (60). This is a cheap and routinely used laboratory drug, but nevertheless, it is important to remember that this carcinogenic compound must be handled with a great care (and gloves). The tubes are filled with buffer 2 and sealed with a specific pair of pliers.

246

Peixoto, Bailly, and David-Cordonnier

An isopycnic CsCl gradient is then obtained by centrifugation at 300,000 g for 20 h at 16°C in a fixed angle rotor. The deceleration rate should be very slow (usually no brake at all) to avoid perturbation of the CsCl gradient. From this isopycnic centrifugation, the denatured proteins stay at the top of the tube and two bands of DNA are separated in the middle of the tube: the plasmid DNA and the genomic DNA. The circular genomic DNA varies in size from 160  Kb for Carsonella ruddii (61) to more than 12 Mb for the myxobacterium Sorangium cellulosum (62) and is around 4.6 Mb for the Escherichia coli strain typically used to amplify plasmid DNA. This genomic DNA, therefore, binds a larger amount of ethidium bromide than the short plasmid DNA, which appears denser. In a similar manner, the supercoiled DNA incorporates less ethidium bromide than the linearised plasmid DNA and migrates in a denser (lower) portion of the tube and will be separated from the linear DNA (broken plasmid) and open circular DNA which generally migrates between the supercoiled plasmid band and the genomic DNA band or could rather co-migrates with the genomic DNA band. The RNA fraction, which is highly dense due to important folding of the molecules, migrates towards the bottom of the tube (Fig. 6). In order to isolate the supercoiled DNA, a needle is inserted at the top of the tub (to allow the outflow of the DNA without internal pressure disturbance during the sucking up) prior to the insertion of the collecting syringe equipped with a needle that is picked just under the supercoiled DNA band. The syringe is filled slowly with the supercoiled plasmid DNA fraction, avoiding mixing and recovery of the bacterial DNA just above (Fig. 6). Usually, the quantity of DNA is enough to see DNA bands under visible light as a pink band in the tube, but in cases of low levels of DNA production, the DNA sample can be observed and recovered proteins linear + open circular plasmid DNA CsCl density

3.1.5. Recovery of the Plasmid

1.60

supercoiled plasmid DNA

1.65 1.70 1.75 1.80

RNA

Fig. 6. Schematic representation of the supercoiled plasmid DNA purification on isopycnic CsCl gradients. The centrifuged tube was firstly pierced on the top to avoid pressure variations during the recovery of the DNA. A syringe was then used to collect the plasmid DNA fraction. The internal density obtained from isopycnic CsCl gradient is indicated on the left of the tube.

Topoisomerase I-Mediated DNA Relaxation as a Tool to Study Intercalation

247

under UV light (366 nm). In this latter case, the UV illumination needs to be as short as possible and DNA collection should be done rapidly to avoid UV-induced DNA damage. In general, less than 1 mL of concentrated DNA sample is collected and it is better to keep this precious concentrated DNA solution out of the day light. 3.1.6. Removal of Ethidium Bromide

Ethidium bromide is removed from the DNA sample by three washings with three volumes of isobutanol. The extracted DNA is dialysed twice against 1 L of Tris/EDTA buffer. The DNA concentration is quantified in a quartz cuvette by determining the absorbance at 260 nm (typically, 10 µl of stock solution in 1-mL BPE buffer). An extinction coefficient of 6,600 M−1 cm−1 is typically used to estimate the DNA concentration (in bases), with naturally occurring DNA with a balanced AT/GC content. If it is too dilute, the DNA sample precipitated with ethanol and redissolved in an appropriate amount of TE buffer. The supercoiled DNA is stored frozen in small aliquot at −20°C.

3.1.7. Alternative Plasmid Purification System

An alternative procedure, more flexible but generally less efficient, employs purification columns, such as NucleoBond PC (MacheryNagel Gmbh, Germany) or HiSpeed Plasmid Kits (Qiagen SA, France), used according to the manufacturer protocols. However, these approaches generally lead to a higher amount of open circular DNA compared to our favoured (but manual) CsCl procedure.

3.2. Agarose Gel

Gels are prepared by dissolving 1% of agarose powder in 1X TBE buffer (see Note 1). Depending on the number of wells needed for the experiment, small or large electrophoresis sets are used. These should be filled with an appropriate volume of agarose mixture, usually 120 mL or 400 mL for the small and large gels, respectively. The agarose is melted for approximately 3  min at 700 W in a microwave (see Note 2). The electrophoresis gel is then poured on the base set, previously sealed at each end using adhesive tape and equipped with a comb of 5  mm length and 1  mm thick (depending on the volume of the samples to be loaded) to generate appropriates wells (see Note 3). After total stiffening of the gel at room temperature (see Note 4), removing of the comb and adhesive tape, the agarose gel is maintained wet by dipping it in the electrophoresis apparatus filled with 1X TBE buffer. Since drying of the wells will interfere with proper migration of the DNA samples, the gels should be completely covered with the 1X TBE buffer.

248

Peixoto, Bailly, and David-Cordonnier

3.3. Topoisomerase I-Induced Relaxation of Supercoiled Plasmid in the Presence of Increasing Concentrations of the Test Compound 3.3.1. Drug/DNA Interaction

Graded concentrations (usually up to 50 µM) of the compounds to be evaluated are incubated with supercoiled plasmid DNA (130 ng) in 20 µl of Topo I reaction buffer for 15 min at room temperature to ensure binding equilibrium.

3.3.2. Topoisomerase I-Induced Relaxation

Human recombinant topoisomerase I (5U) is added to the mixture and incubated at 37°C for 45 min to ensure relaxation of the DNA (see Note 5).

3.3.3. Degradation of the Protein

The reactions are then stopped by adding SDS and proteinase K to a final concentration of 0.25% and 250 µg/mL, respectively. The samples are incubated at 50°C for 30 min to totally remove the protein from DNA. At this stage, there is no need to extract DNA with solvents; the degraded protein fragments will not hinder DNA migration during the electrophoresis.

3.3.4. Separation of the Forms of DNA on Agarose Gel

DNA samples are completed by adding 3 µl of the electrophoresis dye mixture and loaded on to the 1% agarose gel, which is covered with 1X TBE buffer. The different forms of DNA are then separated by electrophoresis at room temperature for 2 h and 45 min at 120 V for small gels or 150 V for large ones. Various controls need to be loaded at the same time: supercoiled and linearised DNA plasmids (both of them in the same amount and in the same buffer as the samples of interest) and, if possible, a sample of molecular weight ladders (usually a commercially available 1-kb ladder).

3.3.5. Staining of the Gel Using Ethidium Bromide

Gels are run without ethidium bromide in order to obtain the various separated topoisomers. To visualise the DNA forms, the gel is stained post-electrophoresis by soaking in a bath containing 25-µM ethidium bromide for 30  min at room temperature. By contrast, gels used for DNA cleavage assays usually contain ethidium bromide, which is added when preparing the agarose gel prior and so do not need to be stained post-migration. The gels are finally washed for at least 4 h to overnight in 300 mL of purified water (see Note 6).

3.3.6. Visualisation of the DNA Forms

The gels are photographed under UV light (366 nm) using a UV transilluminator coupled with a camera and the data stored for further analysis (see Note 7). It should be remembered that the greater the number of supercoils, the faster the migration on agarose gel. Visualisation of DNA in agarose gels is based on the increase of

Topoisomerase I-Mediated DNA Relaxation as a Tool to Study Intercalation

249

fluorescence emitted by DNA-bound ethidium bromide molecules. This was initially demonstrated by LePecq and Paoletti (60) who evidenced a 21-fold activation of the fluorescence of ethidium bromide when bound to DNA. It is important to notice that ethidium bromide does not stain the different forms of plasmid DNA to the same extent. Indeed, the open circular plasmid can incorporate more ethidium bromide molecules than closed circular DNA. This characteristic is used to separate the supercoiled plasmid DNA from linear DNA by isopycnic centrifugation, as described in Subheading 1.1. This difference was quantified as a 1.8 ratio for the pBR322 vector (63, 64) and explains why the level of fluorescence in total DNA from each particular lane changes during the topoisomerase I-induced DNA relaxation process, evidenced by agarose gel migration and ethidium bromide staining. The different staining capacity of linear vs. supercoiled DNA should also be considered if the proportions of each DNA band need to be quantified. 3.4. An Example of DNA Relaxation Induced by Intercalating Agents

In the present example, ethidium bromide is used as a reference compound to evidence the topoisomerase I-induced DNA relaxation (Fig.  7a). Compounds 1, 2 and 3 are used as test compounds (Fig. 7b). The pUC19 plasmid (2,686 bp) was used and gave seven clearly identified topoisomers (marked by arrow heads in Fig. 7). In this case, as mentioned above, the gels were stained with ethidium bromide after electrophoresis and were washed extensively to give uniform staining, independent of the initial concentration of ethidium bromide used in the staining solution. The drug-free DNA sample is negatively supercoiled and migrates through the gel as a single band. A tiny proportion of nicked “open circular” DNA can be detected (lane “DNA”). Its profile changes markedly in the presence of increasing concentrations of ethidium bromide. Two phases can be distinguished. First at low concentrations, the intercalation of ethidium bromide between DNA base pairs induces a relaxation of DNA and the bulky, fully relaxed form migrates slowly through the gel. The top band in lane, containing 0.25-mM ethidium bromide corresponds to the formation of fully relaxed DNA molecules (which co-migrate with open circular DNA molecules). As the ethidium bromide concentration is further increased, the DNA molecules wind in the opposite way so as to produce more compact, positive supercoils (e.g. in lane corresponding to 0.75 mM ethidium bromide), which then migrate faster than the negatively supercoiled DNA topoisomers. When the DNA is fully positively supercoiled (e.g. in the lane corresponding to 2.5 mM ethidium bromide), it migrates as a single band with an electrophoretic mobility close (but nevertheless slightly reduced compared) to that of the native negatively supercoiled plasmid (lane “DNA”). At high ethidium bromide concentrations (>2.5  mM), a gel-shift is observed; the

250

Peixoto, Bailly, and David-Cordonnier

Topoisomerase I

a

0.01 0.025 0.05 0.1 0.25 0.5 0.75 1 2.5 5 7.5 10 0 DNA

Ethidium bromide (µM)

relaxed topoisomers

bound

supercoiled helical superturns

– – – –

0

+ + + + + + +

–+

Topoisomerase I

b

Cpd-2

Cpd-3 (µM)

DNA 0 5 10 20 30 40 50 5 10 20 30 40 50 1 2 5 10 20 50 DNA Topo I

Cpd-1

relaxed topoisomers bound

supercoiled helical superturns

–– – 0 + + + + –0 + + + + – –0 + + + – –

Fig. 7. Topoisomerase I-induced DNA relaxation in the presence of ethidium bromide. Negative supercoiled pUC19 vector was incubated with ethidium bromide at the concentrations indicated on the top of the lanes prior to be treated with topoisomerase I. Control DNA (lane “DNA”) was treated in the same manner but in the absence of topoisomerase I and ethidium bromide. The supercoiled (full arrow), relaxed (open arrow) and various topoisomers forms (spears of arrow) of DNA are localised on the left. Supercoiled DNA with bound ethidium bromide inducing a gel shift is localised using a bracket on the right of the panel. The positive or negative helical super-turns are represented as plus or minus symbols.

DNA is saturated and non-specific DNA binding occurs. The upand-down profile seen here with ethidium bromide (negative supercoiled → relaxed → positive supercoils) is typical of an intercalating agent. It is certainly one of the most robust approaches (together with viscometry) to characterise a DNA intercalating agent. Similar profiles can be seen with the test compounds 1–3 in Fig.  7. In all three cases, the negatively supercoiled plasmid substrate is first relaxed and then intercalation leads to the formation of positively supercoiled DNA. The concentration of the compound required to produce the fully relaxed DNA species

Topoisomerase I-Mediated DNA Relaxation as a Tool to Study Intercalation

251

varies from one compound to the other, with the following ranking: cpd-3 > cpd-1 > cpd-2 and this reflects their respective binding affinities. The three molecules are less potent DNA intercalators than ethidium bromide, but there is no doubt that they are typical intercalators, producing the characteristic up-and-down profile. With cpd-3 at high concentrations (20–50  mM), the positively supercoiled DNA is saturated and a gel-shift is observed. This type of behaviour is frequently seen with non-specific “sticky” DNA-binding drugs. Non-specific binding can be prevented, or at least reduced, using one of the two following procedures. First, by adding SDS (1%) in the samples as well as in the gel at the preparation stage. However, this method is not recommended, because it requires extensive washings of the gel after electrophoresis to be able to stain it properly with ethidium bromide. Since SDS is difficult to totally eliminate, some traces remain and limit the post-electrophoresis staining. Second, by extracting DNA prior to electrophoresis. The elimination of ethidium bromide in the samples of DNA is in this latter case performed before loading on the gel. For this purpose, a phenol/chloroform extraction is performed (vol./vol.), the samples are centrifuged 5 min at 13,000 g. The aqueous phase is recovered, transferred to a fresh tube and subjected to a chloroform extraction (vol./vol.). After similar centrifugation, the DNA fraction is precipitated with 2.5 vol. of cold pure ethanol followed by a last centrifugation 20 min at 13,000 g. The DNA pellet is dried and subsequently dissolved in an appropriate amount of 1X TBE buffer prior to the addition of the loading buffer and migration on the agarose gel, as usual. Because of the multi-step process, much attention needs to be given at each step to ensure the homogeneity in the quantity of recovered DNA in each point of the experiment. This second method works well, but it is time-consuming and requires practise. In most cases, the standard procedure, without extraction, can be recommended. 3.5. Conclusions

The DNA relaxation assay is a robust method to characterise the intercalation of small molecules into DNA and more generally to investigate drug–DNA interactions. It demands a good knowledge of the principle, based on the topoisomerase I-guided manipulation of DNA conformation. It also requires technical skills and proficiency, but overall, the assay is relatively simple to execute. In our highly complex, technology and machine-driven scientific society, this gel-based assay is a successful old-fashioned manual tool. It is a robust and highly reproducible method, which has been routinely used in our laboratories, by many people including many students.

252

Peixoto, Bailly, and David-Cordonnier

4. Notes 1. Ready-to-use solutions of 0.5–2% agarose in 1× TBE buffer can be purchased. 2. The mixture must be allowed to boil, but the heating must be stopped in time to avoid overflowing of a foaming hot gel. Complete dissolution of the agarose powder needs to be checked carefully to ensure a good quality gel (free of bubbles or “fishes”), essential to obtain good migration profiles of DNAs. The dissolved agarose mixture is very hot, and should be allowed to cool for 3–5 min (to ~50°C) prior to pouring the gel. 3. Bubble formation needs to be avoided at this step or removed prior to further cooling and stiffening of the agarose gel. 4. The gel should be prepared at least 1 h before use. 5. Commercially available topoisomerase I enzymes are usually satisfactory (which is not the case for topoisomerase II). Alternatively, topoisomerase I can be extracted and purified from eukaryotic cells or bacteria, according to published procedures (65–73). 6. However, due to the mutagenic effect of ethidium bromide, all liquid and solid wastes have to be eliminated using specific elimination procedures. Alternatively, gels can be stained using other dyes, such as chloroquine (Sigma), SYBR Gold (Invitrogen), SYBR Green (Molecular Probes), GoldView (SBS Genetech), GeneFinder (Bio-v Company) or GelRed (FluoProbes) (74– 80). However, these chemicals are more expensive than ethidium bromide, certainly very efficient as DNA stains, but would also be considered as potential mutagenic agents. That is why ethidium bromide is our method of choice. 7. Under UV light, the gel needs to be photographed rapidly to avoid fluorescence fading of the ethidium bromide/DNA complex (81). The variation of fluorescence during irradiation of ethidium bromide-stained DNA depends on the emission/ excitation wavelength and time of exposition to the light. The photodestruction of the ethidium fluorophore is sufficiently slow to allow a proper analysis.

Acknowledgements M.-H.D.-C. thanks the Institut pour la Recherche sur le Cancer (IRCL), Association pour la Recherche contre le Cancer (ARC) and the Ligue Nationale contre le Cancer (Comité du Nord) for grants. P.P. thanks the Institut pour la Recherche sur le Cancer

Topoisomerase I-Mediated DNA Relaxation as a Tool to Study Intercalation

253

sur le Cancer de Lille (IRCL), the Conseil Régional Nord-Pasde-Calais and the ARC for a PhD fellowship. The authors are grateful to Sabine Depauw for her technical expertise. References 1. Wang JC (1996) DNA topoisomerase. Annu Rev Biochem 65:635–692 2. Wang JC (2002) Cellular roles of DNA topoisomerases: a molecular perspective. Nat Rev Mol Cell Biol 3:430–440 3. Champoux JJ (2001) DNA topoisomerases: structure, function, and mechanism. Annu Rev Biochem 70:369–413 4. Corbett KD, Berger JM (2004) Structure, molecular mechanisms, and evolutionary relationships in DNA topoisomerases. Annu Rev Biophys Biomol Struct 33:95–118 5. Leppard JB, Champoux JJ (2005) Human DNA topoisomerase I: relaxation, roles, and damage control. Chromosoma 114:75–85 6. Giles GI, Sharma RP (2005) Topoisomerase enzymes as therapeutic targets for cancer chemotherapy. Med Chem 1:383–394 7. Forterre P, Gribaldo S, Gadelle D, Serre MC (2007) Origin and evolution of DNA topoisomerases. Biochimie 89:427–446 8. McClendon AK, Osheroff N (2007) DNA topoisomerase II, genotoxicity, and cancer. Mutat Res 623:83–97 9. Champoux JJ (2002) A first view of the structure of a type IA topoisomerase with bound DNA. Trends Pharmacol Sci 23: 199–201 10. Staker BL, Hjerrild K, Feese MD, Behnke CA, Burgin AB Jr, Stewart L (2002) The mechanism of topoisomerase I poisoning by a camptothecin analog. Proc Natl Acad Sci U S A 99:15387–15392 11. Staker BL, Feese MD, Cushman M, Pommier Y, Zembower D, Stewart L, Burgin AB (2005) Structures of three classes of anticancer agents bound to the human topoisomerase I-DNA covalent complex. J Med Chem 48: 2336–2345 12. Xiong B, Burk DL, Shen J, Luo X, Liu H, Shen J, Berghuis AM (2008) The type IA topoisomerase catalytic cycle: A normal mode analysis and molecular dynamics simulation. Proteins 71:1984–1994 13. Carey JF, Schultz SJ, Sisson L, Fazzio TG, Champoux JJ (2003) DNA relaxation by human topoisomerase I occurs in the closed clamp conformation of the protein. Proc Natl Acad Sci U S A 100:5640–5645

14. Koster DA, Croquette V, Dekker C, Shuman S, Dekker NH (2005) Friction and torque govern the relaxation of DNA supercoils by eukaryotic topoisomerase IB. Nature 434: 671–674 15. Calladine CR, Drew HR (1997) The molecule and how it works. In: Understanding DNA, 2nd edn, p 56 16. Calladine CR, Drew HR, Luisi BF, Travers AA (2004) Different kinds of Double Helix. In: Understanding DNA, 3rd edn, pp 39–63 17. Olson WK, Babcock MS, Gorin A, Liu G, Marky NL, Martino JA, Pedersen SC, Srinivasan AR, Tobias I, Westcott TP, Zhang P (1995) Flexing and folding double helical DNA. Biophys Chem 55:7–29 18. Olson WK, Marky NL, Jernigan RL, Zhurkin VB (1993) Influence of fluctuations on DNA curvature. A comparison of flexible and static wedge models of intrinsically bent DNA. J Mol Biol 232:530–554 19. Dickerson RE, Drew HR (1981) Structure of a B-DNA dodecamer: II. Influence of base sequence on helix structure. J Mol Biol 149:761–786 20. Dickerson RE, Klug A (1983) Base sequence and helix structure variation in B and A DNA. J Mol Biol 166:419–441 21. Calladine CR, Drew HR, McCall MJ (1988) The intrinsic curvature of DNA in solution. J Mol Biol 201:127–137 22. Hunter CA, Lu X-J (1997) DNA base-stacking interactions: a comparison of theoretical calculations with oligonucleotide X-ray crystal structures. J Mol Biol 265:603–619 23. Pedone F, Mazzei F, Matzeu M, Barone F (2001) Torsional constant of 27-mer DNA oligomers of different sequences. Biophys Chem 94:175–184 24. Anselmi C, De Santis P, Paparcone R, Savino M, Scipioni A (2002) From the sequence to the superstructural properties of DNAs. Biophys Chem 95:23–47 25. Klug A, Jack A, Viswamitra MA, Kennard O, Shakked Z, Steitz TA (1979) A hypothesis on a specific sequence-dependent conformation of DNA and its relation to the binding of the lac repressor protein. J Mol Biol 131: 669–680

254

Peixoto, Bailly, and David-Cordonnier

26. Drew HR, Dickerson RE (1981) Structure of a B-DNA dodecamer. III. Geometry of hydration. J Mol Biol 151:535–556 27. Nelson HC, Finch JT, Luisi BF, Klug A (1978) The structure of an oligo(dA). oligo(dT) tract and its biological implications. Nature 330:221–226 28. El Hassan MA, Calladine CR (1996) PropellerTwisting of Base-pairs and the Conformational Mobility of Dinucleotide Steps in DNA. J Mol Biol 259:95–103 29. El Hassan MA, Calladine CR (1996) Structural mechanics of bent DNA. Endeavour 20:61–67 30. Suzuki M, Amano N, Kakinuma J, Tateno M (1997) Use of a 3D structure data base for understanding sequence-dependent conformational aspects of DNA. J Mol Biol 274:421–435 31. Zuccheri G, Scipioni A, Cavaliere V, Gargiulo G, De Santis P, Samori B (2001) Mapping the intrinsic curvature and the flexibility along the DNA chain. Proc Natl Acad Sci U S A 98:3074–3079 32. Bharanidharan D, Gautham N (2006) Principal component analysis of DNA oligonucleotide structural data. Biochem Biophys Res Commun 340:1229–1237 33. Frederick CA, Williams LD, Ughetto G, Van der Marel GA, Van Boom JH, Rich A, Wang AHJ (1990) Structural comparison of anticancer drug-DNA complexes: adriamycin and daunomycin. Biochemistry 29: 2538–2549 34. Liao LB, Zhou HY, Xiao XM (2005) Spectroscopic and viscosity study of doxorubicin interaction with DNA. J Mol Struct 749:108–113 35. Lerman LS (1961) Structural considerations in the interaction of DNA and acridines. J Mol Biol 3:18–30 36. Adams A (2002) Crystal structures of acridines complexed with nucleic acids. Curr Med Chem 18:1667–1675 37. Monnot M, Mauffret O, Lescot E, Fermandjian S (1992) Probing intercalation and conformational effects of the anticancer drug 2-methyl-9-hydroxyellipticinium acetate in DNA fragments with circular dichroism. Eur J Biochem 204:1035–1039 38. Neidle S, Thurston DE (1994) New targets for cancer chemotherapy. In: Kerr DJ, Workman P (eds) CRC Press, Boca Raton, FL, pp 159–75 39. Lerman LS (1963) The structure of the DNAacridine complex. Proc Natl Acad Sci U S A 49:94–104

40. Adams A, Guss JM, Collyer CA, Denny WA, Wakelin LPG (1999) Crystal structure of the topoisomerase II poison 9-amino-[N-(2dimethylamino)ethyl]acridine-4-carboxamide bound to the DNA hexanucleotide d(CGTACG)2. Biochemistry 38:9221–9233 41. Rohs R, Sklenar H, Lavery R, Roder B (2000) Methylene blue binding to DNA with alternating GC base sequence: a modeling study. J Am Chem Soc 122:2860–2866 42. Lisgarten JN, Coll M, Portugal J, Wright CW, Aymami J (2002) The antimalarial and cytotoxic drug cryptolepine intercalates into DNA at cytosine-cytosine sites. Nat Struct Biol 9:57–60 43. OhUigin C, McConnell DJ, Kelly JM, van der Putten WJM (1987) Methylene blue photosensitised strand cleavage of DNA: effects of dye binding and oxygen. Nucleic Acids Res 15:7411–7427 44. Waring MJ, González A, Jiménez A, Vázquez D (1979) Intercalative binding to DNA of antitumour drugs derived from 3-nitro-1, 8-naphthalic acid. Nucleic Acids Res 7:217–230 45. Waring MJ, Wakelin LP (1974) Echinomycin: a bifunctional intercalating antibiotic. Nature 252:653–657 46. Wakelin SP, Waring MJ (1976) The binding of echinomycin to deoxyribonucleic acid. Biochem J 157:721–740 47. Lee JS, Waring MJ (1978) Interaction between synthetic analogues of quinoxaline antibiotics and nucleic acids. Changes in mechanism and specificity related to structural alterations. Biochem J 173:129–144 48. Viswamitra MA, Kennard O, Cruse WB, Egert E, Sheldrick GM, Jones PG, Waring MJ, Wakelin LP, Olsen RK (1981) Structure of TANDEM and its implication for bifunctional intercalation into DNA. Nature 289:817–819 49. Bailly C, Braña M, Waring MJ (1996) Sequence-selective intercalation of antitumour bis-naphthalimides into DNA. Evidence for an approach via the major groove. Eur J Biochem 240:195–208 50. Gallego J, Reid BR (1999) Solution structure and dynamics of a complex between DNA and the antitumor bisnaphthalimide LU-79553: intercalated ring flipping on the millisecond time scale. Biochemistry 38:15104–15115 51. Armitage BA, Yu C, Devadoss C, Schuster GB (1994) Cationic anthraquinone derivatives as catalytic DNA photonucleases: mechanisms for DNA damage and quinone recycling. J Am Chem Soc 116:9847–9859

Topoisomerase I-Mediated DNA Relaxation as a Tool to Study Intercalation 52. Rybenkov VV, Cozzarelli NR, Vologodskii AV (1993) Probability of DNA knotting and the effective diameter of the DNA double helix. Proc Natl Acad Sci U S A 90: 5307–5311 53. Katritch V, Bednar J, Michoud D, Scharein RG, Dubochet J, Stasiak A (1996) Geometry and physics of knots. Nature 384:142–145 54. Podtelezhnikov AA, Cozzarelli NR, Vologodskii AV (1999) Equilibrium distributions of topological states in circular DNA: interplay of supercoiling and knotting. Proc Natl Acad Sci U S A 96:12974–12979 55. Metzler R, Hanke A (2006) Knots, bubbles, untying, and breathing: probing the topology of DNA and other biomolecules, handbook of theoretical and computational nanotechnology. American Scientific Publishers, Stevenson Ranch, CA 56. Burnier Y, Dorier J, Stasiak A (2008) DNA supercoiling inhibits DNA knotting. Nucleic Acids Res 36:4956–4963 57. Drlica K (1992) Control of bacterial DNA supercoiling. Mol Microbiol 6:425–433 58. Shishido K, Komiyama N, Ikawa S (1987) Increased production of a knotted form of plasmid pBR322 DNA in Escherichia coli DNA topoisomerase mutants. J Mol Biol 195:215–218 59. Waring MJ (1964) Complex formation with DNA and inhibition of Escherichia coli RNA polymerase by ethidium bromide. Biochim Biophys Acta 87:358–361 60. LePecq JB, Paoletti C (1967) A fluorescent complex between ethidium bromide and nucleic acids. Physical–chemical characterization. J Mol Biol 27:87–106 61. Nakabachi A, Yamashita A, Toh H, Ishikawa H, Dunbar H, Moran N, Hattori M (2006) The 160-kilobase genome of the bacterial endosymbiont Carsonella. Science 314:267 62. Pradella S, Hans A, Spröer C, Reichenbach H, Gerth K, Beyer S (2002) Characterisation, genome size and genetic manipulation of the myxobacterium Sorangium cellulosum So ce56. Arch Microbiol 178:484–492 63. Hertzberg RP, Caranfa MJ, Hecht SM (1989) On the mechanism of topoisomerase I inhibition by camptothecin: evidence for binding to an enzyme-DNA complex. Biochemistry 28:4629–4638 64. Boege F, Straub T, Kehr A, Boesenberg C, Christiansen K, Andersen A, Jakob F, Köhrler J (1996) Selected novel flavones inhibit the DNA binding or the DNA religation step of eukaryotic topoisomerase I. J Biol Chem 271:2262–2270

255

65. Attardi DG, De Paolis A, Tocchini-Valentini GP (1981) Purification and characterization of Xenopus laevis type I topoisomerase. J Biol Chem 256:3654–3661 66. Ishii K, Hasegawa T, Fujisawa K, Andoh T (1983) Rapid purification and characterization of DNA topoisomerase I from cultured mouse mammary carcinoma FM3A cells. J Biol Chem 258:12728–12732 67. Tanizawa A, Pommier Y (1992) Topoisomerase I alteration in a camptothecin-resistant cell line derived from Chinese hamster DC3F cells in culture. Cancer Res 52:1848–1854 68. Jensen AD, Svejstrup JQ (1996) Purification and characterization of human topoisomerase I mutants. Eur J Biochem 236:389–394 69. Rossi F, Labourier E, Forné T, Divita G, Derancourt J, Riou JF, Antoine E, Cathala G, Brunel C, Tazi J (1996) Specific phosphorylation of SR proteins by mammalian DNA topoisomerase I. Nature 381:80–82 70. Zhu CX, Tse-Dinh YC (1999) Overexpression and purification of bacterial DNA topoisomerase I. Methods Mol Biol 94:145–151 71. Stewart L, Champoux JJ (1999) Purification of baculovirus-expressed human DNA topoisomerase I. Methods Mol Biol 94:223–234 72. Bronstein IB, Wynne-Jones A, Sukhanova A, Fleury F, Ianoul A, Holden JA, Alix AJ, Dodson GG, Jardillier JC, Nabiev I, Wilkinson AJ (1999) Expression, purification and DNAcleavage activity of recombinant 68-kDa human topoisomerase I-target for antitumor drugs. Anticancer Res 19:317–327 73. Takahashi T, Matsuhara S, Abe M, Komeda Y (2002) Disruption of a DNA topoisomerase I gene affects morphogenesis in Arabidopsis. Plant Cell 14:2085–2093 74. Jaxel C, Kohn KW, Wani MC, Wall ME, Pommier Y (1989) Structure-activity study of the actions of camptothecin derivatives on mammalian topoisomerase I: evidence for a specific receptor site and a relation to antitumor activity. Cancer Res 49:1465–1469 75. Kiltie AE, Ryan AJ (1997) SYBR Green I staining of pulsed field agarose gels is a sensitive and inexpensive way of quantitating DNA double-strand breaks in mammalian cells. Nucleic Acids Res 25:2945–2946 76. Miller SE, Taillon-Miller P, Kwok PY (1999) Cost-effective staining of DNA with SYBR green in preparative agarose gel electrophoresis. Biotechniques 27:34–36 77. Huang Q, Fu WL (2005) Comparative analysis of the DNA staining efficiencies of different fluorescent dyes in preparative agarose gel electrophoresis. Clin Chem Lab Med 43:841–842

256

Peixoto, Bailly, and David-Cordonnier

78. Maxwell A, Burton NP, O’Hagan N (2006) High-throughput assays for DNA gyrase and other topoisomerases. Nucleic Acids Res 34:e104 79. Li D, Li G, Guo W, Li P, Wang E, Wang J (2008) Glutathione-mediated release of functional plasmid DNA from positively charged quantum dots. Biomaterials 29:2776–2782 80. Sun YX, Zeng X, Meng QF, Zhang XZ, Cheng SX, Zhuo RX (2008) The influence

of RGD addition on the gene transfer characteristics of disulfide-containing polyethyleneimine, DNA complexes. Biomaterials 29:4356–4365 81. Wahl P, Paoletti J, Le Pecq JB (1970) Decay of fluorescence emission anisotropy of the ethidium bromide-DNA complex. Evidence for an internal motion in DNA. Proc Natl Acad Sci U S A 65:417–421

Chapter 16 A High-Throughput Assay for DNA Topoisomerases and Other Enzymes, Based on DNA Triplex Formation Matthew R. Burrell, Nicolas P. Burton, and Anthony Maxwell Abstract We have developed a rapid, high-throughput assay for measuring the catalytic activity (DNA supercoiling or relaxation) of topoisomerase enzymes that is also capable of monitoring the activity of other enzymes that alter the topology of DNA. The assay utilises intermolecular triplex formation to resolve supercoiled and relaxed forms of DNA, the principle being the greater efficiency of a negatively supercoiled plasmid to form an intermolecular triplex with an immobilised oligonucleotide than the relaxed form. The assay provides a number of advantages over the standard gel-based methods, including greater speed of analysis, reduced sample handling, better quantitation and improved reliability and accuracy of output data. The assay is performed in microtitre plates and can be adapted to high-throughput screening of libraries of potential inhibitors of topoisomerases including bacterial DNA gyrase. Key words: Topoisomerase, DNA gyrase, Triplex formation, Supercoiling, Relaxation, High-throughput screening

1. Introduction DNA topoisomerases are essential, ubiquitous enzymes that control the topological state of DNA in cells (1). As such, the enzymes are important, established targets of anti-bacterial and anti-tumour drugs and are potential herbicide and anti-viral targets. All topoisomerases can relax supercoiled DNA, and DNA gyrase, essential in bacteria, can also introduce negative supercoils into DNA. Topoisomerases are also capable of catenation and decatenation, and knotting and unknotting. The basic reaction catalysed by topoisomerases, the interconversion of relaxed and supercoiled DNA, is not readily monitored. The standard assay resolves the reaction products on the basis of their different mobilities on an agarose gel, and while it is K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_16, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

257

258

Burrell, Burton, and Maxwell

information-rich, it suffers from the drawbacks of being slow and, due to the electrophoresis step, requires a lot of sample handling. Therefore, the screening of combinatorial chemical libraries for novel topoisomerase inhibitors is greatly hindered by the lack of a rapid high-throughput assay. To address this issue, we have developed a rapid microtitre plate assay, based on DNA triplex formation, that is capable of monitoring topoisomerase activity with large number of samples (2). The underlying principle of the assay is the greater efficiency of triplex formation in negatively supercoiled DNA compared with the relaxed form (3, 4). Previously, immobilised biotinylated triplex-forming oligonucleotides have been shown to be able to capture supercoiled plasmid DNA (5). Using this principle, supercoiling or relaxation of DNA is readily monitored through fluorescence staining of plasmid DNA trapped to a microtitre plate through an intermolecular triplex (2).

2. Materials Unless otherwise stated, all materials are purchased from Sigma and are of the highest grade available (see Note 1). Ultrapure water with a resistivity of ~18 MW cm should be used throughout. 2.1. Buffers

1. Wash buffer: 20  mM Tris–HCl (pH 7.6), 137  mM NaCl, 0.01% (w/v) bovine serum albumin (acetylated), 0.05% (v/v) Tween-20. Stored at 4°C. 2. Triplex Formation (TF) buffer: 50  mM sodium acetate (pH 5.0), 50  mM NaCl, 50  mM MgCl2. Stored at room temperature. 3. T10 buffer: 10 mM Tris–HCl (pH 8.0), 1 mM EDTA. Stored at room temperature. 4. DNA Gyrase Supercoiling buffer: 35 mM Tris–HCl (pH 7.5), 24 mM KCl, 4 mM MgCl2, 2 mM DTT, 1.8 mM spermidine, 1 mM ATP, 6.5% (w/v) glycerol, 0.1 mg/mL bovine serum albumin. The buffer is stored as a 5× concentrate at −20°C. 5. DNA Gyrase Dilution buffer: 50  mM Tris–HCl (pH 7.5), 100 mM KCl, 2 mM DTT, 1 mM EDTA, 10% (w/v) glycerol. Stored at −20°C. 6. Topoisomerase I Relaxation buffer: 20  mM Tris–HCl (pH 7.5), 50 mM NaCl, 0.1 mM EDTA. The buffer is stored as a 10× concentrate at −20°C. 7. Topoisomerase I Dilution buffer: 10 mM Tris–HCl (pH 7.5), 1 mM DTT, 1 mM EDTA, 50% (v/v) glycerol, 100 µg/mL bovine serum albumin.

A High-Throughput Assay for DNA Topoisomerases and Other Enzymes

259

Table 1 Oligonucleotides used in the high-throughput assay

2.2. DNA

Name

Sequence (5¢–3¢)

5¢ Modification

TFO1

TCTCTCTCTCTCTCTC

Biotin

TFO1W

TCGGAG AGAGAGAGAGAGAG

TFO1C

CCGATCTCTCTCTCTCTCTC

1. Plasmid pNO1 is constructed from plasmid pBR322* (2) (a high-copy number version of pBR322, based on the work of Boros et al. (6)) by ligating the annealed oligos TFO1W and TFO1C (Table 1) into the AvaI site of the plasmid. Supercoiled pNO1 is prepared by transforming it into Escherichia coli competent cells (e.g. Top10, Invitrogen), growing the cells overnight in Luria–Bertani (LB) medium containing 100 µg/mL ampicillin and purifying the DNA using Qiagen mini- or midiprep kits according to the manufacturer’s instructions. To prepare DNA with a specific linking difference (superhelix density, s) of ~0.06, the plasmid can then be further treated with Topoisomerase I (200 µg plasmid with 200 units Topoisomerase I) in Topoisomerase I Relaxation buffer in the presence of ethidium bromide (30 µg/mL) for 90 min at 37°C. The DNA is extracted with butan-1-ol to remove the ethidium bromide followed by phenol/chloroform extractions and ethanol precipitation. The plasmid is resuspended in T10 buffer to a concentration of 1 mg/mL. 2. Relaxed pNO1 is prepared by incubating the supercoiled form with topoisomerase I (~40–50 µg plasmid with 200 units topoisomerase I in Topoisomerase I Relaxation buffer) for 1 h at 37°C. The DNA is extracted with two phenol/chloroform extractions and purified by ethanol precipitation. 3. TFO1 oligo (Table 1) with a 5¢ biotin tag (e.g. Sigma-Genosys, Bioneer) is resuspended to 100 µM in T10 buffer and stored at −20°C.

2.3. Enzymes

1. E. coli DNA gyrase subunits GyrA and GyrB are expressed in E. coli and purified according to the literature methods (7). The subunits are stored separately in DNA Gyrase Dilution buffer at −80°C (see Note 2). The complete enzyme is reconstituted at 4°C by mixing equal concentrations of GyrA and GyrB prior to use. 2. Human topoisomerase I (see Note 1) may be prepared by overexpressing in baculovirus-infected insect cells (Spodoptera frugiperda) and purified as described by Stewart et al. (8). It is stored at −80°C in Topoisomerase I Dilution buffer.

260

Burrell, Burton, and Maxwell

3. All restriction enzymes are purchased from New England Biolabs and stored according to the manufacturer’s instructions. 2.4. Equipment and DNA Staining

1. Black streptavidin-coated ReactiBind 96-well plates (Greiner) are used for the assay. The wells are rehydrated with three 200 µL volumes of Wash buffer before use. Plates should be stored, covered, at 4°C. 2. DNA is stained with SYBR Gold (Invitrogen), which is stored as a 10,000× concentrate at −20°C before use. SYBR Gold is diluted 10,000-fold in T10 buffer. 3. Fluorescence measurements are made using a SPECTRAmax Gemini fluorimeter and Softmax Pro Software. Alternative plate readers with fluorescence measurement capability can be used.

3. Methods 3.1. DNA Gyrase Supercoiling Assay

1. Wash microtitre plate wells with three 200 µL volumes of Wash buffer (see Note 3). 2. Load 100 µL 500  nM biotinylated TFO1 oligo (diluted in Wash buffer) into wells and allow immobilisation to proceed for 2 min at room temperature. 3. Remove oligo solution and wash carefully with three 200 µL volumes of Wash buffer (see Note 3). 4. The DNA gyrase supercoiling reaction is performed in the wells in a 30 µL volume containing the following: 1–6 µL reconstituted DNA gyrase (1–2 units (see Note 4); the total volume of gyrase is made up to 6 µL with DNA Gyrase Dilution buffer), 1 µL 1 µg/µL relaxed pNO1, 6 µL 5× DNA Gyrase Reaction buffer and H2O to 30 µL (see Note 5). Incubate the reaction at 37°C for 30  min (this can be carried out in the plate reader if temperature control is available). 5. The reaction is stopped with the addition of 200 µL TF buffer, which lowers the pH, and the plate is incubated at room temperature for 30  min to allow triplex formation. Supercoiled DNA becomes trapped on the plate while relaxed DNA remains in solution. 6. Remove unbound relaxed and linear plasmid by washing the wells thoroughly with three 200 µL volumes of TF buffer. 7. Drain the wells and stain DNA with 200 µL 1× SYBR Gold (diluted in T10 buffer). The plate is incubated for a further 20 min at 37°C. After incubation, mix the contents of the well.

A High-Throughput Assay for DNA Topoisomerases and Other Enzymes

b

261

80,000

Fluorescence

a Gyrase U 0 0.063 0.125 0.25 0.5

1

2.5

5

60,000 40,000 20,000 0

c

0

2

4

6

[gyrase] U

250

Fluorescence

200 150 100 50 0

0

1

2

3 4 [gyrase] U

5

6

Fig. 1. Monitoring the DNA gyrase supercoiling reaction using the high-throughput and gel-based assays. Assays were performed with relaxed pNO1 using the indicated amounts of enzyme in units. Samples were analysed by gel electrophoresis ((a) and (b)) and by SYBR Gold fluorescence using the microtitre plate assay (c). (Reproduced from Maxwell et al. (2), with permission from Oxford University Press.)

8. Read fluorescence using excitation at 495 nm and emission at 537 nm. Wells are washed with three 200 µL volumes of Wash buffer and the plate is stored at 4°C for future use (see Notes 6 and 7). Supercoiling activity is calculated by reference to control reactions containing 1 µg supercoiled pNO1 without enzyme (see Fig. 1 for example data). 3.2. Topoisomerase I Relaxation Assay

1. All steps of the relaxation assay are as described above for the supercoiling assay, except for the reaction step, which is performed as follows: the reaction is performed in a 30 µL volume containing 1–2 units topoisomerase I (the enzyme is made up to 6 µL with Topoisomerase I Dilution buffer), 1 µL 1 µg/µL supercoiled pNO1, 3 µL 10× Topoisomerase I Reaction buffer and H2O to 30 µL. The reaction is incubated at 37°C for 30 min.

3.3. Other Enzymes

In principle, the high-throughput assay can be used to monitor the activity of any enzyme that causes a change in the linking number of the plasmid substrate. For example, the method can be used for rapidly monitoring restriction endonuclease activity.

262

Burrell, Burton, and Maxwell

1. To assay the activity of a restriction enzyme that cuts the pNO1 plasmid, such as AatII, all steps are performed as described for the supercoiling assay, except the reaction step. The reaction is performed in the plate in a 30 µL volume containing 1 µg supercoiled pNO1, 1 unit of AatII and the reaction buffer supplied by the manufacturer and is incubated at 37°C for 60  min. A negative control using AvaI, which does not cut the plasmid, may also be performed. Restriction enzyme activity results in a loss of fluorescence compared to untreated plasmid due to the formation of linear DNA, which is unable to form the triplex. 3.4. High-Throughput Screening of Gyrase Inhibitors and Determination of IC50 Values

The high-throughput assay can be used for screening chemical libraries for potential topoisomerase inhibitors. For example, the assay has been used to characterise inhibition of DNA gyrase by a series of modified aminocoumarin antibiotics that contain alterations to the prenylated hydroxbenzoate ring (9). Because several inhibitor concentrations can be quickly tested on a single plate and the resulting output is a series of reliable quantitative measurements of enzyme activity, the assay lends itself to the rapid and accurate determination of IC50 values for compounds of interest. A general protocol for screening inhibitors is given below. 1. DNA gyrase supercoiling assays can be performed as described above using an enzyme concentration that gives ~50% supercoiling (i.e. ~1 unit; see Note 8). All reaction components except inhibitors are added in a ‘master-mix’, which is stored on ice until use to prevent the supercoiling reaction from occurring. The master mix is loaded onto the microtitre plate wells containing immobilised TF oligo, inhibitors are added and mixed well and the plate is incubated at 37°C for 30 min. Novobiocin (or a similarly characterised inhibitor) should be included as a reference. The solvent that test compounds are dissolved in should be included in control reactions (see Note 9). 2. To determine IC50 values for the inhibition of DNA gyrase by selected compounds, supercoiling assays are performed with a wide range of inhibitor concentrations (see Note 10). Typically, eight inhibitor concentrations, spanning three orders of magnitude, would be used (see Fig. 2 for example data). Controls with a known DNA gyrase inhibitor should also be performed. In addition, the highest concentration of the test compound used for the IC50 determination should be added to a reaction without DNA gyrase to ensure that the inhibitor does not increase or decrease the background fluorescence. 3. IC50 values are computed by plotting the inhibition data and using the exponential decay equation to fit them using a curvefitting program such as SigmaPlot.

A High-Throughput Assay for DNA Topoisomerases and Other Enzymes

263

Fig. 2. Characterisation of the activity of DNA gyrase inhibitors using the high-throughput assay. Supercoiling activity was determined at a range of concentrations of either novobiocin (a) or a novobiocin analogue (9) (b), and normalised to control data obtained in the absence of drug. An exponential decay equation was used to fit the data and determine IC50 values. Error bars show the spread of the data for three separate measurements

4 Notes 1. The triplex-based microplate assay described is protected by patent application WO06/051303. Commercial performance of the assay requires a licence, available from Plant Bioscience Ltd. (Norwich, UK; http://www.pbltechnology.com/).

264

Burrell, Burton, and Maxwell

The assay is available as a kit from Inspiralis Ltd. (Norwich, UK; http://www.inspiralis.com/) who also supply pNO1 and a range of topoisomerase enzymes including E. coli DNA gyrase and topoisomerase IV, and human topoisomerases I and II. 2. DNA gyrase and other topoisomerases can lose activity upon repeated freeze-thaw cycles and therefore should be stored at −80°C in the minimum practical aliquot size. 3. It is essential that wells are washed and buffer removed completely, where indicated in the assay procedure. Unbound TFO1 oligo remaining in the wells will interfere with the binding of pNO1 to the immobilised oligo causing invalid results and residual buffers may also affect subsequent steps. 4. One unit of supercoiling activity is defined as the amount of enzyme required to fully supercoil 0.5 µg relaxed pNO1 at 37°C in 30 min. The extent of supercoiling is determined by the inclusion of a control containing 1 µg supercoiled pNO1 without enzyme. Conversely, a unit of relaxation activity is defined as the amount of enzyme required to relax 0.5 µg supercoiled pNO1 at 37°C in 30 min. Units of enzyme activity are quantified by performing the reaction over a range of enzyme concentrations. For example, to quantify units of DNA gyrase supercoiling activity using the triplex assay, supercoiling would be assayed with gyrase between 2 and 20 nM, and the amount of enzyme corresponding to one unit can then be extrapolated from a linear regression of these data. 5. The assay can be carried out with less DNA (e.g. 0.75 µg) and correspondingly less enzyme. 6. Microtitre plates can be re-used for at least three subsequent assays without significant loss in signal quality (2). To re-use a plate, the wells should be washed with Wash buffer thoroughly after SYBR Gold staining and fluorescence measurement. The TFO1 oligo remains bound to the plate so does not need to be re-immobilised prior to subsequent reactions. 7. When troubleshooting problems with the assay (e.g. lack of signal), it is useful to perform the reactions with twice the normal volume and analyse the products of half the reaction mixture on a 1% (w/v) agarose gel. This will show whether the problem is related to the high-throughput assay system or the enzyme reaction step. For agarose gel analysis, 30 µL of the mixture is removed and the reaction terminated with the addition of an equal volume of STEB (40% sucrose, 100 mM Tris– HCl (pH 7.5), 100 mM EDTA, 0.5 mg/mL bromophenol blue),

A High-Throughput Assay for DNA Topoisomerases and Other Enzymes

265

 and the DNA extracted with chloroform:isoamyl alcohol (24:1). A sample (15 µL) of the upper blue phase is loaded onto a 1% (w/v) agarose gel in 1× TAE (40 mM Tris-acetate, 2 mM EDTA) and run at ~7.5 V/cm for ~2 h. The gel is stained in 1 µg/mL ethidium bromide for ~10 min, destained in 1× TAE for ~10  min and bands are visualised on a gel documentation system (Syngene). Supercoiling or relaxation activity can be determined by reference to supercoiled and relaxed pNO1 control reactions. 8. The advantage of using an enzyme concentration that gives less than 100% supercoiling (e.g. 50%) is that it makes the assay more sensitive to small changes in supercoiling activity. For large-scale drug screening, using an amount of enzyme that gives full supercoiling is likely to be acceptable in a primary screen. However, the actual enzyme concentration should be borne in mind when carrying out screening as this will give a lower limit in sensitivity to the assay in terms of the IC50s of test compounds (9). 9. Topoisomerase inhibitors, such as aminocoumarin antibiotics, are often poorly soluble in water and are typically dissolved in dimethyl sulphoxide (DMSO). Appropriate stock concentrations of inhibitors should be used such that the final DMSO concentration in the assay does not exceed 3–5% (v/v), but this should be determined empirically for each enzyme. When a range of inhibitor concentrations is used, appropriate dilutions of the stock inhibitor are made so that the DMSO concentration is the same in each reaction. 10. IC50 values obtained with the high-throughput assay are generally lower than those obtained with the standard gel-based assay. This discrepancy is due to the fact that the gel assay is unable to resolve supercoiled topoisomers with s greater than~−0.3, and as a result, significant inhibition occurs before a change is observed on the gel and IC50 values are therefore overestimated (2). Despite the difference in absolute IC50 values, the relative values obtained with either method will be similar.

Acknowledgements This work was supported by BBSRC (UK) and Plant Biosciences Ltd; we thank Martin Stocks and James Taylor for helpful comments.

266

Burrell, Burton, and Maxwell

References 1. Bates AD, Maxwell A (2005) DNA topology. Oxford University Press, Oxford 2. Maxwell A, Burton NP, O’Hagan N (2006) High-throughput assays for DNA gyrase and other topoisomerases. Nucleic Acids Res 34:e104 3. Hanvey JC, Shimizu M, Wells RD (1988) Intramolecular DNA triplexes in supercoiled plasmids. Proc Natl Acad Sci USA 85: 6292–6296 4. Sakamoto N, Akasaka K, Yamamoto T, Shimada H (1996) A triplex DNA structure of the polypyrimidine:polypurine stretch in the 5¢ flanking region of the sea urchin arylsulfatase gene. Zoolog Sci 13:105–109 5. Kawabata Y, Ooya T, Lee WK, Yui N (2002) Self-assembled plasmid DNA network prepared through both triple-helix formation and

6. 7.

8. 9.

streptavidin–biotin interaction. Macromol Biosci 2:195–198 Boros I, Pósfai G, Venetianer P (1984) Highcopy-number derivatives of the plasmid cloning vector pBR322. Gene 30:257–260 Maxwell A, Howells AJ (1999) Overexpression and purification of bacterial DNA gyrase. In: Bjornsti M-A, Osheroff N (eds) DNA topoisomerase protocols I. DNA topology and enzymes. Humana, Totowa, New Jersey, pp 135–144 Stewart L, Ireton GC, Champoux JJ (1996) The domain organization of human topoisomerase I. J Biol Chem 271:7602–7608 Anderle C, Stieger M, Burrell M, Reinelt S, Maxwell A, Page M, Heide L (2008) Biological activities of novel gyrase inhibitors of the aminocoumarin class. Antimicrob Agents Chemother 52:1982–1990

Chapter 17 Measurement of DNA Interstrand Crosslinking in Individual Cells Using the Single Cell Gel Electrophoresis (Comet) Assay Victoria J. Spanswick, Janet M. Hartley, and John A. Hartley Abstract The Single Cell Gel Electrophoresis (Comet) assay, originally developed to allow visualisation of DNA strand break damage in individual cells, has been adapted to measure DNA interstrand cross-links. DNA interstrand cross-links are formed in cells by a number of commonly used cancer chemotherapy agents and are considered to be the critical lesion formed by such agents. This technique allows the analysis of DNA interstrand cross-link formation and repair at a single cell level, requires few cells, allows the determination of heterogeneity of response within a cell population and is sensitive enough to measure DNA interstrand cross-links at pharmacologically relevant doses. The method can be applied to any in vitro or in vivo application where a single cell suspension can be obtained. The method has also become invaluable in studies using human tissue and can be used as a method for pharmacodynamic analysis in early clinical trials. Key words: Single gel electrophoresis (Comet) assay, Comet assay, DNA interstrand, Cross-links, DNA repair, Cross-linking drugs, Nitrogen mustards, Platinum drugs, Pharmacodynamic analysis

1. Introduction Drugs capable of producing interstrand cross-links in cellular DNA have been widely used in cancer chemotherapy for many years for the treatment of both solid tumours and haematological malignancies. These include members of the nitrogen mustard class (e.g. mechloroethamine, chlorambucil, melphalan, cyclophosphamide, ifosfamide), chloroethyl-nitrosoureas (e.g. BCNU), dimethanesulphonates (e.g. busulphan), some natural products (e.g. mitomycin C) and platinum drug (e.g. cisplatin, carboplatin and oxaliplatin). In addition, novel interstrand cross-linking agents continue to be developed (e.g. SJG-136 (1)), and conversion of K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_17, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

267

268

Spanswick, Hartley, and Hartley

prodrugs into potent DNA cross-linking agents is being tested clinically in a variety of applications including direct (2), antibodydirected (3), gene-directed (4), and virus-directed prodrug therapies (5). The covalent linkage of the two complementary strands of DNA produced by an interstrand cross-link prevents separation of the strands during critical cellular processes, such as replication and transcription, and it is therefore highly cytotoxic DNA damage. In addition, the involvement of both strands of DNA poses particular problems for the cellular DNA repair machinery and repair of interstrand cross-links is complex (6). The inability of tumour cells to repair these specific lesions contributes to the inherent sensitivity of some cancers (7), and increased repair can contribute clinically to acquired resistance to cross-linking drugs (7, 8). The inability of the two strands of DNA to separate under denaturing conditions resulting from interstrand cross-linking has been exploited in methods to detect cross-linking in naked DNA (see Chapter 18, this volume) and in intact cells. Indeed, the ability to measure DNA interstrand cross-linking in cells at pharmacologically relevant doses has been possible for several decades using the technique of alkaline elution developed by Kohn and co-workers (9). This method, however, requires a relatively large number of cells and the prior radiolabelling of cellular DNA. As a result, it cannot easily be adapted for studies in vivo. More recently, the single cell gel electrophoresis (Comet) assay has been adapted to measure interstrand cross-links (10). Originally developed to measure strand breaks (11), the comet assay allows visualization of DNA damage in individual cells. In order to detect DNA interstrand cross-links, cells are irradiated immediately prior to analysis to deliver a fixed level of random strand breaks to the genome. Cells are then embedded in agarose on a microscope slide and lysed to remove cellular proteins. The DNA is then denatured under alkaline conditions and subjected to electrophoresis. During electrophoresis, any relaxed or broken DNA fragments migrate further than the supercoiled, undamaged DNA. After appropriate staining, the DNA resembles a “comet” with a brightly stained head, and a tail whose length and intensity is determined by the level of strand breakage produced within that cell. The presence of DNA interstrand cross-links retards the migration of the irradiated DNA during electrophoresis compared to irradiated, non-cross-linked controls. The extent of retardation is proportional to the level of interstrand cross-linking. Removal (“unhooking”) of the interstrand cross-link from the DNA can also be assessed with time following a drug-free incubation period. The technique is highly reproducible, more sensitive than alkaline elution, requires fewer cells, and has the important

Measurement of DNA Interstrand Crosslinking in Individual Cells

269

advantage that analysis can be made at a single cell level. It is therefore possible to determine heterogeneity of interstrand cross-link formation and repair in a cell population. The method is applicable to any application where a single cell suspension can be obtained. This includes cells cultured in vitro, and cells isolated from in vivo sources. Recently, the method has become invalu­ able in mechanistic studies using human tissue (e.g. lymphocytes, haematological and solid tumour cells (7, 8, 12)), and is increasingly being used as a method for pharmacodynamic analysis in early clinical trials (2, 3, 13–15).

2. Materials 2.1. Cell Preparation and Drug Treatment

1. Appropriate tissue culture medium.

2.1.1. Suspension Cell Lines

3. FCS (Autogen Bioclear).

2.1.2. Adherent Cell Lines

1. Appropriate tissue culture medium.

2. L-glutamine (Autogen Bioclear, Calne, UK).

2. L-glutamine (Autogen Bioclear). 3. FCS (Autogen Bioclear). 4. Trypsin/ethylenediaminetetraacetic acid (EDTA) (1×) (Autogen Bioclear). 2.1.3. Human Lymphocytes

1. Vacutainer® CPT™ tubes (Becton Dickinson, Oxford, UK) (see Note 1). 2. RPMI 1640 tissue culture medium (Autogen Bioclear). 3. L-glutamine (Autogen Bioclear). 4. Foetal calf serum (FCS) (Autogen Bioclear).

2.1.4. Solid Tumour Tissue/ Aspirates

1. RPMI 1640 tissue culture media (Autogen Bioclear). 2. L-glutamine (Autogen Bioclear). 3. FCS (Autogen Bioclear).

2.1.5. Ascites

1. Dulbecco’s modification of Eagle’s medium (DMEM) (Autogen Bioclear). 2. L-glutamine (Autogen Bioclear). 3. FCS (Autogen Bioclear).

2.2. Single Cell Gel Electrophoresis (Comet) Assay

1. Single-frosted glass microscope slides and glass 24 × 40  mm coverslips. 2. Agarose, Type 1-A (Sigma, Poole, UK). 3. Agarose, Type VII: low gelling temperature (LGT) (Sigma).

270

Spanswick, Hartley, and Hartley

4. Lysis buffer: 100 mM disodium EDTA, 2.5 M NaCl, 10 mM Tris-HCl, pH to 10.5–11.0 with sodium hydroxide pellets. 1% triton X-100 to be added immediately before use. Store at 4°C. 5. Alkali buffer: 50  mM NaOH, 1  mM disodium EDTA, pH 12.5. Caution: Corrosive. Store at 4°C. 6. Neutralisation buffer: 0.5 M Tris-HCl, pH 7.5. Store at 4°C. 7. Phosphate buffered saline (PBS), pH 7.4. Store at 4°C. 8. Flat bed electrophoresis. This should be of sufficient size to hold a large number of slides e.g., 30 × 25 cm gel tank from Flowgen Bioscience, Nottingham, U.K. which holds up to 45 slides. 2.3. Staining and Visualisation

1. Propidium iodide (Sigma), 2.5 µg/mL. Make up fresh before use. Caution: Toxic and light sensitive. 2. Glass coverslips, 24 × 40 mm. 3. Double distilled water. 4. Epi-fluorescence microscope equipped with high pressure mercury light source using a 580 nm dichroic mirror, 535 nm excitation filter and 645  nm emission filter for propidium iodide staining (e.g. Olympus BX51 inverted microscope with Olympus U-RFL-T mercury lamp and Sony XCD-X710 digital camera). 5. Images are visualised, captured and analysed using a suitable image analysis system. Our laboratory uses Komet 5.5 analysis software from Andor Technology, formerly Kinetic Imaging (Belfast, UK) (see Note 2).

3. Methods 3.1. Cell Preparation and Drug Treatment 3.1.1. Suspension Cell Lines

1. Exponentially growing cells should be used at a density of 2.5–3.0 × 104 cells/mL in an appropriate medium containing 2 mM glutamine and 10% FCS. 2. A minimum of 2  mL cells are treated with the cross-linking agent and incubated for the appropriate time at 37°C in a humidified atmosphere with 5% carbon dioxide (see Note 3). 3. Pellet cells by centrifugation at 200× g for 5  min at room temperature. 4. Remove supernatant and resuspend cells in 2  mL of fresh drug-free medium containing 2 mM glutamine and 10% FCS maintained at 37°C. 5. Incubate cells using the above conditions for the required post-treatment time (see Note 4).

Measurement of DNA Interstrand Crosslinking in Individual Cells

271

6. Cells are now ready to be processed as described in Subheading 3.2, step 2. 7. Alternatively, cells can be frozen and stored at −80ºC following treatment with the cross-linking agent. This allows samples to be taken at different time points e.g. for repair studies. The samples can then be processed in a single assay reducing the possibility of inter-assay variation. Following treatment, centrifuge the cells at 200× g for 5  min at 4ºC. Discard the supernatant and resuspend the pellet in 2  mL freezing mixture (FCS containing 10% dimethylsulphoxide) (Sigma). Aliquot into 2 × 1  mL freezing vials and freeze at −80ºC (see Notes 5 and 6). 3.1.2. Adherent Cell Lines

1. Exponentially growing cells are treated with the cross-linking agent and incubated for the appropriate time at 37°C in a humidified atmosphere with 5% carbon dioxide (see Note 3). 2. After the appropriate incubation, carefully remove the media and replace with fresh drug-free medium containing 2  mM glutamine and 10% FCS maintained at 37°C. 3. Incubate cells for the required post-treatment time (see Note 4). 4. Remove media and trypsinize cells with tryspin/EDTA solution until all cells have rounded up and detached (see Note 7). 5. Neutralise trypsinisation by the addition of fresh media containing 2 mM glutamine and 10% FCS. 6. Transfer cells to a universal tube, wash twice with media containing 2 mM glutamine and 10% FCS maintained at 4°C by centrifuging at 200× g for 5 min at 4ºC. 7. Cells are now ready to be processed as described in Subheading 3.2, step 2. 8. Alternatively, cells can be frozen and stored at −80ºC as described in Subheading 3.1.1, step 7, and the Single Cell Gel Electrophoresis (Comet) assay performed at a later date.

3.1.3. Human Lymphocytes

1. Collect whole blood using Vacutainer® CPT™ system. This allows sterile blood collection and cell separation using a single centrifugation step (see Note 1). 2. Centrifuge at 1,500× g for 20 min at room temperature (see Note 8). 3. Remove layer of lymphocytes at the interface and wash twice with RPMI 1640 media containing 2  mM L-glutamine and 10% FCS maintained at 37°C. 4. For ex vivo experiments, i.e., where isolated lymphocytes are treated with drug, count the lymphocytes using a haemocytometer and dilute to 2.5–3.0 × 104 cells/mL in RPMI 1640 media containing 2  mM L-glutamine and 10% FCS.

272

Spanswick, Hartley, and Hartley

 Treat minimum of 2 mL of lymphocytes with the relevant concentration of cross-linking agent and incubate for the appropriate time at 37ºC in a humidified atmosphere with 5% carbon dioxide (see Note 3). 5. Remove the cross-linking agent by centrifuging the lymphocytes at 200× g for 5 min at room temperature. 6. Remove supernatant and resuspend the lymphocytes in fresh drug-free RPMI 1640 containing 2  mM L-glutamine and 10% FCS maintained at 37ºC. 7. Incubate lymphocytes at 37°C for the required post-treatment time (see Note 4). 8. Lymphocytes are now ready to be processed using the Single Cell Gel Electrophoresis (Comet) assay as described in Subheading 3.2, step 2. 9. For in  vivo investigations, i.e., where DNA interstrand cross-links are to be detected in patients treated with DNA cross-linking agents, isolate lymphocytes as quickly as possible after sampling (see Note 8). Wash lymphocytes twice with RPMI 1640 containing 2 mM L-glutamine and 10% FCS maintained at 4ºC and dilute to a final concentration of 2.5 × 104/mL. Continue assay from Subheading 3.2, step 4. 10. Alternatively, for both ex vivo and in vivo experiments, lymphocytes can be frozen and stored at −80ºC as described in Subheading 3.1.1, step 7. (see Note 5). 3.1.4. Solid Tumour Tissue/ Aspirates

1. Solid tumour/aspirate samples should be placed in 10  mL RPMI 1640 medium containing 2  mM L-glutamine and 10% FCS maintained at 4°C immediately after collection (see Note 8). 2. Place the tumour material in a 10  cm petri dish and cover with 1–2 mL of cold RPM1 1640 medium. This should be carried out in a class II biological safety cabinet. 3. Using two sterile scalpels, finely chop the tumour tissue using a cross-cutting action until a single cell suspension is formed. The petri dish must be kept on ice throughout the procedure (see Note 9). 4. Transfer cell suspension to a 15 mL falcon tube and make up the volume to 5 mL with cold RPMI 1640 medium containing 2 mM L-glutamine and 10% FCS and centrifuge at 200× g for 5 min at 4ºC. 5. Drug treatments (ex vivo), in vivo investigations and freezing procedure can be performed as described in Subheading 3.1.3, steps 4 and 10.

Measurement of DNA Interstrand Crosslinking in Individual Cells 3.1.5. Ascites

273

1. Ascitic fluid containing tumour cells and normal mesothelial cells, for example, from ovarian cancer patients, should be collected under sterile conditions. 2. Aliquot the fluid into 50 mL conical tubes and centrifuge at 200× g for 5 min at 4ºC. 3. Discard the supernatant and resuspend in 40  mL DMEM medium containing 2  mM L-glutamine and 10% FCS and seed into 150 cm2 tissue culture flasks. 4. Incubate cells for 1 hour at 37ºC with 5 % carbon dioxide in a humidified atmosphere. 5. After 1 h, remove the entire volume of tissue culture medium containing unattached cells from the flask and transfer to new flask and incubate at 37ºC with 5% carbon dioxide in a humidified atmosphere. 6. Tumour cells require a significant length of time to adhere to plastic, whereas normal mesothelial cells generally attach within the first hour of incubation. 7. Once a significant tumour cell population has been achieved, cells are maintained at 37ºC with 5% carbon dioxide in a humidified atmosphere. 8. Tumour cells are treated with the cross-linking agent in duplicate as described in Subheading  3.1.3, step 4, and incubated for the appropriate time at 37°C in a humidified atmosphere with 5% carbon dioxide (see Note 3). 9. After the appropriate incubation, carefully remove the media and replace with fresh drug-free medium containing 2 mM glutamine and 10% FCS maintained at 37°C. 10. Incubate cells using the above conditions for the required post-treatment time (see Note 4). 11. Remove media and trypsinize cells with tryspin/EDTA solution until all cells have rounded up and detached (see Note 7). 12. Neutralise trypsinisation by the addition of fresh media containing 2 mM glutamine and 10% FCS. 13. Transfer cells to a universal tube, wash twice with media containing 2 mM glutamine and 10% FCS maintained at 4°C by centrifuging at 200× g for 5 min at 4ºC. 14. Cells are now ready to be processed as described in Subheading 3.2, step 2. 15. Alternatively, cells can be frozen and stored at −80ºC as described in Subheading 3.1.1, step 7, and the Single Cell Gel Electrophoresis (Comet) assay performed at a later date.

274

Spanswick, Hartley, and Hartley

3.2. Single Cell Gel Electrophoresis (Comet) Assay

Important: All stages of this assay should be carried out on ice under subdued lighting, solutions maintained at 4°C, and incubations performed in the dark where indicated (see Notes 10 and 11). 1. Precoat microscope slides with 1% type 1-A agarose in water by pipetting 1 mL of molten agarose onto the centre of the slide and place a coverslip on top. Allow to set and remove the coverslip. Slides are then allowed to dry overnight at room temperature. The slides must be dry before use (see Note 12). 2. After required drug exposure/repair time, pellet cells by centrifuging at 200× g for 5  min at 4°C. Remove supernatant and resuspend cells to a final concentration of 2.5 × 104 cells/ mL in the appropriate tissue culture media maintained at 4°C ensuring that a single cell suspension has been achieved. 3. Alternatively, if samples are frozen following isolation and treatment, thaw and remove freezing mixture by centrifugation. Resuspend samples in the appropriate tissue culture medium at 4°C and centrifuge at 200× g for 5 min at 4°C. Discard the supernatant, resuspend sample in the appropriate tissue culture medium at 4°C to a final concentration of 2.5 × 104 cells/mL as a single cell suspension. 4. Irradiate samples on ice with the appropriate X-ray dose, except for the untreated unirradiated control (see Notes 13–15). 5. Take 0.5 mL of resuspended cells and put in a 24 well plate on ice. Add 1 mL of molten 1% LGT agarose in water cooled to 40°C, mix, pipette 1 mL onto the centre of the slide on ice and place a coverslip on top (see Note 16). Once set, remove coverslip and place in a tray on ice. Duplicate slides should be prepared (see Note 17). 6. Add ice cold lysis buffer containing 1% triton X-100 ensuring that all slides are sufficiently covered. 7. Incubate on ice for 1 h in the dark. 8. Carefully remove lysis buffer ensuring that the gels are intact and remain on the slides (see Note 18). 9. Add ice cold double distilled water to completely cover the slides. Incubate on ice for 15 min in the dark. This should then be repeated three times. 10. Remove slides from tray and transfer carefully to an electrophoresis tank (see Note 19). 11. Cover slides with ice cold alkali buffer and incubate for 45 min in the dark (see Note 20). 12. Electrophorese for 25  min at 18  V (0.6  V/cm), 250  mA. This must be carried out in the dark (see Note 21).

Measurement of DNA Interstrand Crosslinking in Individual Cells

275

13. Carefully remove slides from the buffer and place on a horizontal slide rack. 14. Flood each slide twice with 1 mL neutralisation buffer and incubate for 10 min. 15. Rinse slides twice with 1 mL PBS and incubate for 10 min. 16. Remove all excess liquid from slides and allow to dry overnight at room temperature. 3.3. Staining and Visualisation

1. Re-hydrate slides in double distilled water for 30 min. 2. Flood each slide twice with 1  mL 2.5 µg/mL propidium iodide solution and incubate for at least 30 min at room temperature in the dark (see Note 22). 3. Rinse slides twice with double distilled water for 10 min and once for 30 min. 4. Allow slides to dry at 40ºC in the dark (see Note 23). 5. Once dry, place a few drops of distilled water onto the slide and cover with a coverslip (see Note 24). 6. Examine individual cells and comets at 20× magnification analysing a minimum of 25 images per duplicate slide (i.e., minimum 50 in total) (see Note 25). 7. Remove coverslip and store slides in a light-proof box at room temperature (see Note 26) 8. The tail moment is used as a measure of DNA damage and is defined as the product of the percentage DNA in the comet tail, and the distance between the means of the head and tail distributions, based on the definition by Olive et  al. (16). The percentage decrease in tail moment is therefore calculated using the following formula:



é æ (TMdi - TMcu )ö ù % decrease in tail moment = ê1 - ç ÷ ú ´ 100 ëê è (TMci - TMcu )ø ûú where TMdi = tail moment of drug treated irradiated sample. TMcu = tail moment of untreated unirradiated control. TMci = tail moment of untreated irradiated control 9. In samples treated under conditions that can also produce strand breaks, e.g., combination of a cross-link agent with an agent known to produce single strand breaks, e.g., gemcitabine. Cross-linking is expressed as percentage decrease in tail moment compared to irradiated controls calculated by the formula below. This formula is used to compensate for the additional single strand breaks induced by agents such as gemcitabine in addition to those produced by the irradiation step.

276

Spanswick, Hartley, and Hartley

% decrease in tail moment é æ öù (TMdi - TMcu ) = ê1 - ç ÷ ú ´ 100 êë è (TMci - TMcu ) + (TMdu - TMcu )ø úû



where TMdi = tail moment of drug treated irradiated sample. TMcu = tail moment of untreated unirradiated control. TMci = tail moment of untreated irradiated control. TMdu = tail moment of drug treated unirradiated sample.

The percentage decrease in tail moment is proportional to the level of DNA cross-linking. 3.4. Examples

Figure 1 illustrates typical comet images obtained following the in vitro treatment of NIH187 human small cell lung cancer cells with melphalan. In the control untreated, unirradiated cells (a), no DNA damage is detected and the high molecular weight supercoiled DNA remains intact. Following irradiation of cells with 15 Gy X-ray to introduce a fixed number of random DNA strand breaks, the resulting shorter fragments of DNA migrate from the bulk of the DNA during electrophoresis to produce typical comet images (b). In irradiated samples following treatment with 25 mM melphalan, (c), comet tails are visible but with reduced length and intensity when compared to irradiated controls (b). Increasing the dose to 100 mM melphalan (d) results in a further reduction in comet tail length and intensity due to the retention of DNA by the melphalan-induced DNA interstrand cross-links in the head of the comet. When expressed as percentage decrease in tail moment compared with untreated irradiated controls, a dose-dependent increase in DNA interstrand crosslink formation can be observed in NIH187 cells following treatment with increasing doses of melphalan (Fig. 2). The formation and repair (“unhooking”) of cisplatin-induced DNA interstrand cross-links in A549 human non-small cell lung cancer cells is shown in Fig. 3. Peak of DNA interstrand cross-link formation is observed 9 h following a 1 h treatment with cisplatin. This is followed by significant repair of cisplatin-induced DNA interstrand cross-links at 24 and 48 h post-treatment.

4. Notes 1. The traditional method for isolating lymphocytes using FicollHypaque can also be used. However, in comparison to the Vacutainer® CPT™ system, the processing time is significantly increased and the whole blood must be layered on to the Ficoll-Hypaque before centrifugation. 2. Other Comet analysis programs are available incorporating all major measurement parameters, such as percentage head/tail

Measurement of DNA Interstrand Crosslinking in Individual Cells

277

Fig. 1. Typical comet images from NIH187 human small cell lung cancer cells following in vitro treatment with melphalan. (a) Untreated unirradiated control cells, (b) Untreated irradiated (15 Gy) control cells, (c) Irradiated melphalan treated (25 mM) cells and (d) Irradiated melphalan treated (100 mM) cells. Cells were treated with melphalan for 1 h followed by a 16 h post-incubation in the absence of drug.

DNA, tail length and Olive tail moment, for example Comet Assay IV, Perceptive Instruments, Haverhill, UK and LAI Comet Analysis System (LACAS), Loats Associates Incorporated, Westminster, Maryland, USA. Automatic analysis programs

Spanswick, Hartley, and Hartley 100

75 % decrease in tail moment

278

50

25

0

0

25

50

75

100

Melphalan (uM)

Fig. 2. In vitro formation of melphalan-induced DNA interstrand cross-links in NIH187 human small cell lung cancer cells. Cells were treated with melphalan for 1 h followed by a 16 h post-incubation in the absence of drug. Results are expressed as percentage decrease in tail moment for 50 cells analysed (mean ± standard error).

100

75 % decrease in 50 tail moment 25

0 0

12

24

36

48

Time (hours post-treatment)

Fig. 3. Formation and repair of cisplatin-induced DNA interstrand cross-links in A549 human non-small cell lung cancer cells. Cells were treated with 150 mM cisplatin for 1 h, then incubated in drug-free medium and samples taken at various time points. Results are expressed as percentage decrease in tail moment for 50 cells analysed (mean ± standard error).

are available allowing unattended automatic comet acquisition and measurement of comet parameters, for example Metafer CometScan, Metasystems, Altlussheim, Germany and AutoComet III, TriTek Corporation, Sumerduck, Virginia, USA. 3. If the chosen drug is to be reconstituted in solvents such as dimethylsulphoxide, the final concentration of solvent added to the cells should be no greater than 0.1%. This is to avoid any additional DNA damage and cell death.

Measurement of DNA Interstrand Crosslinking in Individual Cells

279

4. The length of the post-treatment time is dependent on the type of cross-linking agent used and the peak of interstrand cross-link formation. A time-course experiment should be performed to ascertain this. For example, following a 1 h treatment the peak of interstrand cross-linking for chlorambucil is reached following a 3  h post-incubation, while melphalan requires a post-incubation of 16 h. For repair experiments, the post-treatment time can be further extended. 5. Samples should be resuspended fully in the freezing mixture to prevent DNA damage and cell death. This should be carried out by gently resuspending the sample using a Pasteur pipette rather than vortexing. 6. Our laboratory has validated the effects of long term storage at −80ºC and samples can be stored up to 12 months without any detrimental effects to the integrity of the cells and DNA. 7. Trypsinise cells at 37ºC as quickly as possible to avoid any additional DNA damage and to prevent crosslink repair. Alternatively a non-enzymatic preparation can be used such as cell disassociation solution (Sigma). It is imperative that a single cell suspension is achieved. If several cells migrate together through the gel, an overestimated comet tail moment will result. 8. Lymphocytes and tumour cells should be separated within 30 min of collection to prevent any DNA damage and cell death induced by long term storage and to reduce repair of DNA interstrand cross-links when performing in  vivo investigations. 9. Other methods can be used such as cell disaggregation enzymes. The disadvantage with these is that samples require incubation at 37ºC for up to 1 h. Significant repair of DNA damage is therefore likely to occur in samples from treated patients or animals. 10. It is imperative that the comet assay should be performed on ice and in subdued lighting and incubations carried out in the dark where indicated throughout the text. This is to prevent any DNA repair. All solutions should be ice cold and maintained at 4°C. 11. Reagent kits for the Single Cell Gel Electrophoresis (Comet) assay are available (CometAssay™, Trevigen Incorporated, Gaithersburg, Maryland, USA and CometAssay™, R & D Systems, Abingdon, UK). These kits contain precoated Comet slides, LGT agarose, lysis solution, EDTA and SYBR® Green I nucleic acid gel stain and provide enough reagents for 25 slides. The assay protocols for these kits have not, however, been optimized for individual requirements.

280

Spanswick, Hartley, and Hartley

For example, lysis, alkali denaturation and electrophoresis incubation times are not standardized and require significant optimization by the user to achieve consistent results. Also such kits are significantly more expensive when compared to purchasing and preparing reagents individually. 12. Precoated slides should be prepared and dried in advance. Slides can be stored dry at room temperature for up to 6 months in an airtight container. 13. Our laboratory uses an AGO HS MP-1 X-ray machine with a Varian ND1-321 tube to produce X rays at a dose rate of 2.5 Gy/min. 14. Each experiment should include an untreated unirradiated control. In addition, an untreated irradiated control should also be included with every group of irradiated samples to allow for variation. It is also important to determine in the first instance if the drug under test will produce any detectable single strand breaks in addition to cross-links. This may be achieved by performing the comet assay on drug treated cells but excluding the irradiation step. This also applies when samples are treated in combination with an agent known to cause single strand breaks. 15. A standard curve for irradiation dose in non-drug treated cells should be performed to establish the optimum radiation dose for a given cell type. Ideally, the dose should give a head to tail DNA ratio of approximately 1:1. Our laboratory finds 15 Gy X-rays optimum for most cell lines, lymphocytes and tumour samples. 16. When sample sizes are small, for example lymphocytes, ascitic and tumour/aspirate samples from clinical investigations, the sample and gel size may be reduced. Take 100 mL cell suspension at a final concentration 2.5 × 104/mL and add to a 96 well plate. Add 200mL molten 1% LGT in water cooled to 40ºC, mix and pipette 300 mL onto the centre of the slide. Place 13  mm diameter circular coverslip on top. Once set, remove coverslip and place slide in a tray on ice. Continue assay from Subheading 3.2, step 4. 17. Molten 1% LGT agarose should be maintained at 40°C to aid uniform gel preparation. The thickness of the gel must be consistent between slides to ensure uniform DNA migration and reduce assay variability. All gels should have the dimensions of the coverslip. The gel should not flood the entire slide or the frosted section and should not contain air bubbles. 18. A number of protocols have stated that following lysis the slides can be kept overnight or even days in this solution prior to alkali DNA unwinding and electrophoresis.

Measurement of DNA Interstrand Crosslinking in Individual Cells

281

We do not advise that this should be carried out as the gels tend to break up and slip off the slide. It is therefore recommended that Subheading 3.2, steps 2–16 should be carried out in a single day. 19. Slides should be placed in a flat bed electrophoresis tank lengthways with the frosted end towards the anode. It is essential that the tank is level and all slides face the same direction to ensure low variability between slides. 20. The volume of alkali buffer added to the electrophoresis tank should be consistent from one experiment to the next. It is advisable to measure the volume of buffer required ensuring that all slides are covered by at least 5 mm buffer. 21. These electrophoresis parameters are optimal for our equipment. The current can be adjusted to suit individual requirements. 22. The most commonly used fluorescent stains are propidium iodide and ethidium bromide. The highly sensitive fluorochrome SYBR® Green I nucleic acid gel stain has also been used successfully. It has the advantage of being far more sensitive that propidium iodide and produces no background fluorescence. However, it fades much more rapidly under intense UV light. Comet images can also be visualised using silver staining (CometAssay™ Silver Staining Kit, Trevigen Incorporated, Gaithersburg, Maryland, USA and CometAssay™ Silver kit, R & D Systems Europe Limited, Abingdon, UK). This can be carried out after comets have been analysed using other staining methods and allows visualisation by standard light microscopy and provides a permanent staining for sample archiving. 23. Visualising the slides dry produces optimum results as all the cells are in the same plane giving clear cellular definition (17). This is favoured instead of the traditional wet slide method, which can cause difficulties in focusing and quantitation. 24. Slides should be analysed as quickly as possible. If the coverslip is left on for a considerable length of time, it will become permanently stuck. 25. Each slide should ideally be scored blind to avoid any bias, taking care to ensure that comets are measured from the entire gel area and no part of the slide is analysed more than once. 26. Slides may be stored for at least 5 years and re-analysed at any time as described in Subheading 3.3, steps 5–9. Slides may also be re-stained as described in Subheading 3.3, steps 1–4 if the staining has faded during storage.

282

Spanswick, Hartley, and Hartley

References 1. Hartley JA, Spanswick VJ, Brooks N, Clingen PH, McHugh PJ, Hochhauser D, Pedley RB, Kelland LR, Alley MC, Schultz R, Hollingshead MG, Sausville EA, Gregson SJ, Howard PW, Thurston DE (2004) SJG-136 (NSC 694501) A novel rationally designed DNA minor groove interstrand cross-linking agent with potent and broad spectrum antitumour activity. Part 1: Cellular pharmacology, in vitro and initial in vivo antitumour activity. Cancer Res 64:6693–6699 2. Sarker D, Anderson D, Spanswick VJ, Davies S, Agarwal R, Aitken G, Kerr D, Hartley JA, Judson I, Middleton MR (2008) Preliminary results of a Cancer Research UK phase I trial combining the dinitrobenzamide prodrug CB1954 (tretazicar) and the NQO2 substrate EP-0152R (caricotamide) intraveneously (IV) every 3  weeks. J Clin Oncol 26(May 20 suppl):2505 3. Mayer A, Francis RJ, Sharma SK, Tolner B, Springer CJ, Martin J, Boxer GM, Bell J, Green AJ, Hartley JA, Cruickshank C, Wren J, Chester KA, Begent RH (2006) A phase I study of single administration of antibodydirected enzyme prodrug therapy with the recombinant anti-carcinoembryonic antigen antibody-enzyme fusion protein MFECP1 and a bis-iodo phenol mustard prodrug. Clin Cancer Res 12:6509–6516 4. Hedley D, Ogilvie L, Springer C (2007) Carboxypeptidase-G2-based gene-directed enzyme-prodrug therapy: a new weapon in the GDEPT armoury. Nat Rev Cancer 7:870–879 5. Palmer DH, Mautner V, Mirza D, Oliff S, Gerritsen W, van der Sijp JR, Hubscher S, Reynolds G, Bonney S, Rajaratnam R, Hull D, Horne M, Ellis J, Mountain A, Hill S, Harris PA, Searle PF, Young LS, James ND, Kerr DJ (2004) Virus-directed enzyme prodrug therapy: intratumoural administration of a replication-deficient adenovirus encoding nitroreductase to patients with resectable liver cancer. J Clin Oncol 22:1535–1537 6. McHugh PJ, Spanswick VJ, Hartley JA (2001) Repair of DNA interstrand cross-links: molecular mechanisms and clinical relevance. Lancet Oncol 2:483–490 7. Spanswick VJ, Craddock C, Sekhar M, Mahendra P, Shankaranarayana P, Hughes RG, Hochhauser D, Hartley JA (2002) Repair of DNA interstrand cross-links as a mechanism of clinical resistance to melphalan in multiple myeloma. Blood 100:224–229 8. Wynne P, Newton C, Ledermann JA, Olaitan A, Mould TA, Hartley JA (2007) Enhanced

9.

10.

11.

12.

13.

14.

15.

16.

17.

repair of DNA interstrand cross-linking in ovarian cancer cells from patients following treatment with platinum-based chemotherapy. Br J Cancer 97:927–933 Kohn KW, Ewig RAG, Erickson LC, Zwelling LA (1981) Measurement of strand breaks and cross-links by alkaline elution. In: Friedberg EC, Hanawalt PC (eds) DNA repair. Dekker, New York, pp 379–408 Hartley JM, Spanswick VJ, Gander M, Giacomini G, Whelan J, Souhami RL, Hartley JA (1999) Measurement of DNA cross-linking in patients on ifosfamide therapy using the single cell gel electrophoresis (Comet) assay. Clin Cancer Res 5:507–512 Ostling O, Johanson KL (1984) Microelectrophoretic study of radiationinduced DNA damages in individual mammalian cells. Biochem Biophys Res Commun 123:291–298 Webley SD, Francis RJ, Pedley RB, Sharma SK, Begent RH, Hartley JA, Hochhauser D (2001) Measurement of the critical DNA lesions produced by antibody-directed enzyme prodrug therapy (ADEPT) in  vitro, in  vivo and in clinical material. Br J Cancer 84:1671–1676 Corrie PG, Shaw J, Spanswick VJ, Sehmbi R, Jonson A, Mayer A, Bulusu R, Hartley JA, Cree I (2005) Phase I trial combining gemcitabine and treosulfan in advanced cutaneous and uveal melanoma patients. Br J Cancer 92:1997–2003 Lederman J, Gabra H, Jayson GC, Spanswick VJ, Rustin GJ, Jital M, James LE, Hartley JA (2007) Combination chemotherapy with carboplatin and gemcitabine in patients in platinum-resistant ovarian cancer chemotherapy – a phase II study demonstrating inhibition of DNA cross link repair by gemcitabine. Eur J Cancer Suppl 5:320 Puzanov I, Lee W, Berlin JD, Calcutt MW, Hachey DL, Vermeulen WL, Spanswick VJ, Hartley JA, Chen A, Rothenburg ML (2008) Final results of phase I and pharmacokinetic trial of SJG136 administered on a daily x3 schedule. J Clin Oncol 26, May 20 suppl., abstract 2504 Olive PL, Banath JP, Durand RE (1990) Heterogeneity in radiation-induced DNA damage and repair in tumour and normal cells measured using the “comet” assay. Radiat Res 122:86–94 Klaude M, Erikkson S, Nygren J, Ahnstrom G (1996) The comet assay: mechanisms and technical considerations. Mutat Res 363:89–96

Chapter 18 Measurement of DNA Interstrand Crosslinking in Naked DNA Using Gel-Based Methods Konstantinos Kiakos, Janet M. Hartley, and John A. Hartley Abstract Bifunctional DNA damaging agents continue to be the mainstay in various chemotherapeutic regimens used in the clinic. DNA interstrand crosslinks are considered to be the critical cytotoxic lesions for the biological activity of such agents. Gel-based electrophoretic assays can efficiently separate denatured singlestranded DNA from double-stranded, covalently-linked DNA resulting from the presence of an interstrand crosslink. The methods described here offer a simple way for the assessment of crosslinking efficiencies of bifunctional agents in both long fragments of DNA (e.g. 1–5 kb) and short oligonucleotide DNA duplexes. As the repair of interstrand crosslinks is a key determinant of cellular and clinical chemosensitivity, these methods can be useful for the characterization and isolation of site-directed adducted substrates for use in subsequent biochemical analysis of cellular recognition and DNA repair processes. Key words:  Gel electrophoresis, Agarose gel, Polyacrylamide gel, Crosslinking drugs, DNA interstrand crosslinks, Plasmid DNA, Oligonucleotide duplexes, DNA repair

1. Introduction DNA damaging agents still remain at the core of several single agent or combination chemotherapeutic regimens. Despite a shift in cancer drug discovery towards the development of targeted agents for cancer therapeutics, conventional cytotoxic drugs are in many instances the only therapeutic option still available. Some of the most potent chemotherapy agents used in the clinic are bifunctional, possessing two reactive moieties capable of forming crosslinks. These agents, in general, tend to be considerably more cytotoxic than their monofunctional counterparts (1). DNA interstrand crosslinks, in particular, are believed to be the most critical lesions for exerting the observed biological activity of these drugs and the most challenging for the cellular DNA repair mechanisms. K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_18, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

283

284

Kiakos, Hartley, and Hartley

Electrophoretic methodologies have been developed for the assessment of the ability of bifunctional agents to induce this type of lesions (2). The principle of these methods is based on the fact that the presence of a covalent interstrand crosslink can prevent complete denaturation of the two DNA strands. Electrophoresis can therefore separate permanently denatured single-stranded DNA from double-stranded, covalently-linked DNA. A simple agarose-based method described here allows the detection of interstrand crosslink formation in long fragments of linear DNA (e.g. linear plasmid DNA). The denaturation conditions of this method can be adjusted so that crosslinks of different nature can be detected (Figs. 1 and 2). The crosslinking potential of agents determined in plasmid DNA is relevant to their biological activity as it often correlates with in vitro cytotoxic potency and crosslinking efficiencies in cells (3–5). In addition to conventional clinical agents, novel bifunctional alkylating agents continue to emerge and to be evaluated in clinical trials. Some of these agents are rationally designed to target predetermined sequences. Polyacrylamide gel-based methods presented here employ synthetic oligonucleotides for the study of the interaction of these drugs with duplexes containing specific potential target binding sites (Fig.  3). 5¢-end-labeled duplexes can confirm sequence specificity, probe crosslinking reactivities within different sequence contexts, and offer an insight into the effect of base alterations flanking the site of interaction and spatial requirements such as reactive base pair span and the linker length of the agent (Fig. 4) (6, 7). The ability to repair interstrand crosslinks is a critical determinant of cellular sensitivity to these agents. The polyacrylamide gel-based methods employing non-labeled duplexes can be used for the characterization and isolation of DNA duplexes with a defined interstrand crosslink placed at a unique site (Fig. 5). These can serve, alone or incorporated into closed-circular DNA, as substrates for use in biochemical assays, which can monitor the recognition and processing of different interstrand crosslinks by cellular repair mechanisms.

2. Materials 2.1. Agarose Gel-Based Method for the Detection of DNA Interstrand Crosslinking in Long Fragments of Linear DNA

1. DNA. As an example, pUC19 plasmid DNA (2686 base pairs) is used. 2. Restriction enzymes and their appropriate buffers. BamHI (10 U/mL) (New England Biolabs (UK)) was used to digest pUC19 (see Note 1).

Measurement of DNA Interstrand Crosslinking in Naked DNA Using Gel-Based Methods

100

% Double-stranded (Crosslinked) DNA

c

285

80 60 40 20 0 0.1

1

10

100

Cisplatin concentration (mM)

Fig. 1. Autoradiographs of representative agarose gels showing DNA interstrand crosslinking by cisplatin in linear pUC19 plasmid DNA. Drug treatments were for 2 h at 37ºC. DS is the double-stranded undenatured control and SS is the singlestranded heat denatured control. All other samples were subjected to (a). heat or (b). alkali denaturation. (c). Doublestranded and single-stranded DNA bands in each lane were quantified by densitometry to obtain a dose response curve plotted as % double-stranded (crosslinked) DNA against the cisplatin concentration. With cisplatin, the result from gels in (a) and (b) is identical for both heat (solid line) and alkali (dotted line) denaturation. The increased mobility of doublestranded DNA at high cisplatin concentration is due to the formation of intrastrand crosslinks.

3. Bacterial alkaline phosphatase (BAP): 150  U/mL in 10  mM Tris-HCl, pH 8.0, 120  mM NaCl, 50% (v/v) glycerol (Invitrogen, Paisley, UK). 4. Dephosphorylation buffer (10×): 100 mM Tris-HCl, pH 8.0 (Invitrogen).

286

Kiakos, Hartley, and Hartley

100

% Double-stranded (Crosslinked) DNA

c

80 60 40 20 0 0.0001

0.001

0.01

0.1

1

10

100

SJG-136 concentration (mM)

Fig. 2. Autoradiographs of representative agarose gels showing DNA interstrand crosslinking by the pyrrolobenzodiazepine dimer SJG-136 in linear pUC19 plasmid DNA. Drug treatments were for 2 h at 37ºC. DS is the double-stranded undenatured control and SS is the singe-stranded heat denatured control. All other samples were subjected to (a) heat or (b) alkali denaturation. (c) Double-stranded and single-stranded DNA bands in each lane were quantified by laser densitometry to obtain a dose response curve plotted as % double-stranded (crosslinked) DNA against the SJG-136 concentration. With SJG-136, the result from gels A and B is different for heat (solid line) and alkali (dotted) denaturation due to some heat lability of the crosslinks.

5. Phenol:chloroform:isoamyl alcohol (25:24:1) (v/v/v) saturated with 10  mM Tris, pH 8.0, 1  mM EDTA (SigmaAldrich). 6. Chloroform (Sigma-Aldrich).

Measurement of DNA Interstrand Crosslinking in Naked DNA Using Gel-Based Methods

287

a 5’-GTACTATTTATAATAGATCTAAATATTA-3’ 3’-ATAAATATTATCTAGATTTATAATGATC-5’ b 5’-GTACTATTTATAATTGCATAATATTATTA-3’ 3’-ATAAATATTAACGTATTATAATAATGATC-5’ Fig. 3. Oligomer duplexes incorporating a single, centrally located interstrand crosslink site for (a) SJG-136 at 5¢-GATC-3¢ and (b) Cisplatin at 5¢-GC-3¢ were designed. The crosslinked duplexes isolated from gels (a) and (b) in Fig. 5 were successfully ligated in plasmid DNA.

Fig.  4. Autoradiograph showing interstrand DNA crosslinking in 5  mg of the 5¢-end radiolabeled oligonucleotide duplex A (Fig. 3) following incubation at the denoted concentrations with SJG-136, at 37ºC overnight. The crosslinked DNA duplex has a lower electrophoretic mobility than the denatured, single-stranded DNA. The latter is electrophoresed as control beside the drug treated samples to ensure that non crosslinked DNA is fully denatured, hence any double-stranded products are the result of the presence of drug-induced interstrand crosslinks.

7. Sodium acetate buffer: 3 M, pH 7.2. 8. Industrial Methylated Spirit (IMS) (Surgipath Europe Limited, Bretton, UK). 9. 95% and 70% Ethanol. 10. Lyophilizer or vacuum dryer.

288

Kiakos, Hartley, and Hartley

Fig. 5. (a) UV shadowing of 50 mg of oligonucleotide duplex A (Fig. 3) treated with SJG-136 at the denoted concentrations at 37ºC, overnight. 0 is the untreated control. DS corresponds to the crosslinked species and SS to the denatured, single-stranded oligonucleotides. (b) UV shadowing of 250 mg of oligonucleotide duplex B (Fig. 3) treated with cisplatin at the concentrations shown at 22ºC, overnight.

Measurement of DNA Interstrand Crosslinking in Naked DNA Using Gel-Based Methods

289

11. T4 polynucleotide kinase (PNK): 10  U/mL in 50  mM Tris-HCl, pH 7.6, 25 mM KCl, 5 mM DTT, 0.1 mM ATP, 0.2 mg/mL BSA, 50% glycerol (v/v) (Invitrogen). 12. Forward reaction buffer (5×): 350  mM Tris-HCl, pH 7.6, 50  mM MgCl2, 500  mM KCl, 5  mM 2-mercaptoethanol, 350 mM ADP (Invitrogen). 13. Adenosine-5¢-triphosphate [g-32P], 9.25  MBq, 6000  Ci/ mmol (Perkin Elmer, Boston, MA, USA). 14. Ammonium acetate buffer: 7.5 M. 15. TEOA buffer: 25  mM triethanolamine, 1  mM Na2EDTA, pH 7.2. 16. Alkylation stop solution: 0.6  M sodium acetate, 20  mM Na2EDTA, 100 mg/mL tRNA (see Note 2). 17. Non-denaturing loading buffer: 6% sucrose and 0.04% bromophenol blue in distilled and deionized water. 18. Strand separation buffer: 30% dimethyl sulphoxide, 1  mM EDTA, 0.04% bromophenol blue and 0.04% xylene cyanol. 19. Alkali denaturation buffer (6% sucrose, 0.4% sodium hydroxide, 0.04% bromophenol blue). 20. Agarose (Invitrogen). 21. TAE agarose gel and running buffer: 40 mM Tris, 20 mM acetic assay, 2 mM EDTA, pH 8.1. 22. Horizontal gel electrophoresis unit e.g. AGT4 submarine gel tank (VWR International Limited, Lutterworth, UK). Typically, for plasmid DNA a 20 cm × 20 cm × 0.5 cm gel was used. 23. Power supply e.g. VWR Power supply 300 V 500 MA. 24. 3 MM filter paper (Whatman, Maidstone, UK). 25. DE 81 filter paper (Whatman). 26. Saran wrap. 27. Standard vacuum heated gel drying unit. 28. Autoradiography cassettes. 29. X-ray film and developing facilities or imaging densitometer. 2.2. Polyacrylamide Gel-Based Method for the Detection and Analysis of DNA Interstrand Crosslinking in DNA Oligonucleotide Duplexes

1. HPSF (High purity salt free) purified oligonucleotides approximately 20–30 bases (0.05-mmol scale) (see Note 11). 2. TEOA buffer: 25  mM triethanolamine, 1  mM Na2EDTA, pH 7.2. 3. Sodium acetate: 3 M, pH 5.2. 4. 95% ethanol. 5. T4 polynucleotide kinase (PNK): 10 U/mL in 50 mM TrisHCl, pH 7.6, 25  mM KCl, 5  mM DTT, 0.1  mM ATP, 0.2 mg/mL BSA, 50% glycerol (v/v) (Invitrogen).

290

Kiakos, Hartley, and Hartley

6. Forward reaction buffer (5×): 350  mM Tris-HCl, pH 7.6, 50  mM MgCl2, 500  mM KCl, 5  mM 2-mercaptoethanol, 350 mM ADP (Invitrogen). 7. Adenosine 5¢-triphosphate [g-32P], 9.25  MBq, 6000  Ci/ mmol (Perkin Elmer). 8. Bio-spin disposable chromatography columns (Biorad Laboratories, UK). 9. Biogel-P6 spin column solution: 8% P6-biogel (w/v) (Biorad), 0.02% sodium azide (w/v) in distilled and deionized water. 10. TE buffer: 10 mM Tris-HCl, pH 7.6, and 1 mM EDTA. 11. Glycogen (20 mg/mL) (Roche diagnostics). 12. Ultrapure Sequagel™ Sequencing system comprising concentrate (19:1 acrylamide: bisacrylamide, 8.3  M urea), diluent (8.3 M Urea) and buffer (10× TBE, 8.3 M urea pH 8) solutions (National Diagnostics, Hull, UK) 13. Vertical slab gel electrophoresis unit e.g. Sequi-Gen® GT Sequencing Cell with glass plates 21 × 50 cm, 0.4 mm apart (BioRad). 14. 150 mL syringe and tubing. 15. High voltage power supply e.g. Power pac 3000 (Biorad). 16. Silanization Solution II (Fluka Chemicals). 17. Ammonium persulphate (Sigma). 18. Tetramethylenediamine (TEMED) (Sigma). 19. TBE buffer: 89  mM Tris base, 89  mM Boric acid, 2  mM EDTA (disodium salt), pH 8.3. 20. Formamide loading buffer (96% (v/v) formamide (deionized), 20 mM EDTA, 0.03% (w/v) bromophenol blue and 0.03% (w/v) xylene cyanol). 21. 3MM filter paper (Whatman, Maidstone, UK). 22. DE 81 filter paper (Whatman). 23. Saran wrap. 24. Standard vacuum heated gel drying unit. 25. Autoradiography cassettes. 26. X-ray film and developing facilities or imaging densitometer. 2.3. Isolation of Double Stranded Oligonucleotide Containing a Single Crosslink

1. Complementary HPSF purified oligonucleotides approximately 20–30 bases. 2. TEOA buffer: 25  mM triethanolamine, 1  mM Na2EDTA, pH 7.2. 3. 95% ethanol.

Measurement of DNA Interstrand Crosslinking in Naked DNA Using Gel-Based Methods

291

4. Ultrapure Sequagel™ Sequencing system system (National Diagnostics, UK) 5. Vertical slab gel electrophoresis unit e.g. Sequi-Gen® GT Sequencing Cell with glass plates 21 × 50 cm, 0.4 mm apart (BioRad). 6. 150 mL syringe and tubing. 7. High voltage power supply e.g. Power pac 3000 (Biorad). 8. Silanization Solution II (Fluka Chemicals). 9. Ammonium persulphate (Sigma). 10. Tetramethylenediamine (TEMED) (Sigma). 11. TBE buffer: 89  mM Tris base, 89  mM Boric acid, 2  mM EDTA (disodium salt), pH 8.3. 12. Formamide loading buffer (96% (v/v) formamide (deionized), 20  mM EDTA, 0.03% (w/v) bromophenol blue and 0.03% (w/v) xylene cyanol) with and without the marker dyes. 13. Saran wrap. 14. Gel elution buffer: Ammonium acetate 0.5 M, Magnesium acetate 10 mM, EDTA 1 mM pH 8.0, SDS 1% 15. Aluminium backed silica gel 60 F254 TLC plate  20 × 20  cm (Machery Nagel, Duren, Germany) 16. Handheld shortwave UV lamp (254 nm).

3. Methods 3.1. Agarose GelBased Method for the Detection of DNA Interstrand Crosslinking in Long Fragments of Linear DNA 3.1.1. Linearization of Plasmid DNA

1. Linearize 20 mg of the closed circular pUC19 plasmid DNA with the restriction enzyme BamHI (10 U) in its appropriate buffer, incubating at 37 ºC for 45 min. 2. Terminate the restriction reaction by adding 0.1 vol 3  M sodium acetate, and precipitate the DNA with 3 vol of 95% ethanol. 3. Vortex and freeze the samples in a dry ice/IMS bath for 15 min. 4. Centrifuge at top speed (16,000 g) in a cooling centrifuge for 15 min. 5. Remove the supernatant and wash the pelleted DNA with 2 × 200 mL of 70% (room temperature) ethanol. 6. Dry under vacuum. 7. Resuspend the dry DNA pellet in an appropriate volume of distilled and deionized water to provide a concentration of 0.5 mg/mL.

292

Kiakos, Hartley, and Hartley

3.1.2. Dephosphorylation of Linearized DNA

1. Dephosphorylate the linearized DNA with bacterial alkaline phosphatase (3 U/mg) in the dephosphorylation buffer provided and in a final volume of 100  mL of distilled water, at 65ºC for 1 h. 2. Allow the mixture to cool at room temperatute. 3. Extract the dephosphorylated DNA with phenol:chloroform:isoamyl alcohol (25:24:1).

2

vol

of

4. Vortex, spin at top speed (16,000 g) for 5 min, and transfer the upper aqueous layer to another Eppendorf tube. 5. Vortex with an equal volume of chloroform, spin, and collect again the aqueous phase in a fresh tube. 6. Back extract the organic layers with 100 mL of distilled water, and combine the aqueous layers of the Eppendorfs in steps 4 and 5. 7. Precipitate with sodium acetate and 95% ethanol, centrifuge and vacuum to dryness as described in steps 2–6 in Subheading 3.1.1. 8. Resuspend the DNA in 40 mL of distilled and deionized water providing a stock of linearized and dephosphorylated DNA at a concentration of 5 mg/10mL and store at −20ºC. 3.1.3. 5¢-End Labeling of Plasmid DNA

1. 5¢-end label 5 mg of the linearized and dephosphorylated DNA (10 mL of the stock of step 8 in Subheading 3.1.2) with [g-32P] ATP (10  mCi), T4 PNK (2  U/mg) and 1× forward reaction buffer in a final volume of 20 mL, with the addition of distilled and deionized water. 2. Incubate the kinase reaction mixture at 37ºC for 45 min. 3. Add an equal volume of 7.5 M ammonium acetate, and precipitate the DNA with 3 vol 95% ethanol (see Note 3). 4. Freeze, centrifuge, and dry as in steps 3, 4, and 6 in Subheading 3.1.1. 5. Resuspend the dry DNA pellet in 50 mL of distilled and deionized water. 6. Precipitate the DNA a second time with 0.1 vol of 3 M sodium acetate and 3 vol 95% ethanol, freeze, centrifuge, and vacuum or lyophilize to dryness. 7. Resuspend the DNA in 40 mL of distilled and deionized water yielding a concentration of 125 ng/mL.

3.1.4. Drug-DNA Incubations

1. Dilute 8  mL of the 5¢-end labeled DNA (125  ng/mL) in 100 mL of TEOA buffer, allowing for 10 experimental points, each containing 10 ng of DNA per 10 mL of buffer.

Measurement of DNA Interstrand Crosslinking in Naked DNA Using Gel-Based Methods

293

2. Dilute a stock solution of drug dissolved in the appropriate solvent, in TEOA buffer. 3. For drug treated samples, add the appropriate drug dilution volume to the 10 mL of DNA. Make all samples up to a final volume of 50  mL with TEOA buffer, per experimental point. 4. Incubate the DNA with a range of drug concentrations at 37ºC for the appropriate time (see Note 4). 5. Terminate the drug-DNA reactions by addition of an equal volume of alkylation stop solution, followed by 3 vol of 95 % ethanol. 6. Follow steps 3–6 (see Subheading 3.1.1). 7. Resuspend the undenatured double-stranded control samples in 10 mL of non-denaturing loading buffer. Dissolve the drugtreated samples and the single-stranded controls to be subjected to appropriate denaturation conditions, in an equal volume of either strand separation buffer or alkali denaturation buffer (for heat or alkali denaturation, respectively) (see Note 5). 8. Heat denature the relevant samples by incubating at 90ºC for 3 min. Chill immediately in an ice water bath to prevent reannealing. Samples for alkali denaturation should be vortexed vigorously for 3 min prior to gel loading. Control non-denatured samples are loaded directly on the agarose gel. 3.1.5. Agarose Gel Electrophoresis

1. Prepare a 0.8% neutral agarose gel in TAE buffer. For plasmid DNA, a 20 cm × 25 cm × 0.5 cm gel was prepared and horizontally submerged in the same running buffer. 2. Load the samples and electrophorese at 40  V overnight. Alternatively, electrophoresis can be carried out for less time at a higher voltage.

3.1.6. Autoradiography

1. Transfer gels onto one layer of Whatman 3 MM and one layer of DE 81 filter paper. Cover with Saran wrap. 2. Dry at 80ºC for 2 h on a gel drier. 3. Expose X-ray film to the dried gel to visualize the DNA fragments. Exposure times vary, depending on the amount of radioactivity. Overnight exposure of the gel at room temperature is usually sufficient to provide sharp images of satisfactory intensity (see Note 6). Typical examples are shown in Figs. 1a and 2a. 4. Quantify film bands using an imaging densitometer (see Notes 7–10).

294

Kiakos, Hartley, and Hartley

3.2. Polyacrylamide Gel-Based Method for the Detection and Analysis of DNA Interstrand Crosslinking in DNA Oligonucleotide Duplexes 3.2.1. Oligonucleotide Annealing 3.2.2. Oligonucleotide Labeling

1. Reconstitute each of the two oligonucleotides required to make the target duplex in distilled and deionized H2O to a concentration of, e.g., 1 µg/µL (see Notes 12 and 13). 2. Anneal the oligonucleotides by mixing equal volumes of each (e.g. 2.5 mg) in a total volume of 20 mL in TEOA buffer, and heat at 90ºC for 2 min in a heating block. Transfer the samples to a preheated waterbath set at 90ºC and switch it off, allowing them to slowly cool and fully anneal (see Note 14 and 15).

1. 5¢-end radiolabel e.g. 5  mg of annealed duplex with [g-32P] ATP (10  mCi), T4 PNK (2  U/mg) and 1× forward reaction buffer in a final volume of 20 mL, with the addition of distilled and deionized water (see Notes 16). 2. Incubate the kinase reaction mixture at 37ºC for 45 min. 3. Make up a total volume of 100 mL with the addition of TE buffer. 4. Load the kinase reaction mixture to a biogel spin column (see Note 17). 5. Spin for 5 min at 1,200 g to elute the end-labeled DNA. 6. Collect the radioactive probe.

3.2.3. Drug Treatment

1. Drug treat 10 ml of the eluted probe (~0.5 mg) for the appropriate time in a final volume of 50  mL of TEOA buffer (see Note 18). 2. Add 0.1 vol 3  M sodium acetate, and precipitate the DNA with 3 vol of 95% ethanol (see Note 19). 3. Vortex and freeze the samples in a dry ice/IMS bath for 15 min. 4. Centrifuge at top speed (16,000 g) in a cooling centrifuge for 15 min. 5. Remove the supernatant and dry under vacuum (see Note 20). 6. Lyophilize to dryness.

3.2.4. Denaturing Polyacrylamide Gel Electrophoresis

1. Prepare a 20% denaturing polyacrylamide gel. Polymerize the Ultrapure Sequagel™ Sequencing system solutions mixture used in this study, by adding 0.32 mL of freshly prepared 25% ammonium persulphate and 40  mL of TEMED and mixing immediately prior to pouring the gel. 2. Cast the gel using a syringe (applicable for the Sequi-Gen® GT Sequencing Cell) (see Note 21). 3. Insert a 16-well comb at the top of the gel and allow to polymerize.

Measurement of DNA Interstrand Crosslinking in Naked DNA Using Gel-Based Methods

295

4. Pre-run the gel in TBE buffer for 30 min, at 1,200 V. 5. Reconsitute the dry samples in 7  mL of formamide loading buffer and load directly onto the gel (see Notes 22 and 23). 6. Electrophorese at 1,200  V for about 5  h (or until the bromophenol blue marker has migrated 20–25 cm) to allow sufficient separation of the single-stranded and double-stranded DNA species (see Note 24). 7. Transfer the gel onto Whatman 3 MM paper, supported by a layer of DE 81 filter paper and cover with Saran wrap. 8. Dry on gel dryer (see Note 25). 3.2.5. Autoradiography and Densitometry

Expose X-ray film to the dry gel overnight at room temperature. The crosslinked DNA duplex has reduced mobility compared to the corresponding labeled denatured single strands (e.g. Fig. 4) and can be quantified using densitometry. Alternatively, the gel can be placed in contact with a phosphorimage plate for 2 h and scanned with a phosphorimager.

3.3. Isolation of Double Stranded Oligonucleotide Containing a Single Crosslink

1. Reconstitute each of the complementary oligonucleotides in H2O to yield a concentration of 10 µg/µL (see Notes 11–13 and 26).

3.3.1. Oligonucleotide Drug Treatment

2. Anneal the oligonucleotides as in step 2 in Subheading 3.2.1 (see Notes 14 and 15). 3. Drug treat the DNA duplexes (e.g. 5–250 mg) in a total volume of 100 mL TEOA at 37ºC for appropriate time (see Notes 18 and 27). 4. Precipitate the DNA with addition of 0.1 vol of 3 M sodium acetate and 3 vol of 95% ethanol. 5. Freeze in a dry ice/IMS bath for 20 min. 6. Centrifuge at 16,000 g for 15 min at 4ºC. 7. Remove the supernatant and dry (see Note 28). 8. At this point the samples can be stored at −80ºC to be run on the denaturing gel at a later date.

3.3.2. Denaturing Gel Separation

1. Cast and pre-run a 20% polyacrylamide denaturing gel (see Subheading 3.2.4). 2. In the first well, load 10 mL of formamide loading buffer containing the marker dyes to monitor migration during electrophoresis. 3. Next, load a control untreated sample to act as a marker of the single-stranded DNA species and a control drug-treated sample to act as a marker of the double-stranded DNA species. Allow a space of 1–2 wells between the two controls. These samples should be dissolved in dyeless formamide buffer (see Note 29)

296

Kiakos, Hartley, and Hartley

4. Load the drug treated samples for elution in the remaining wells, allowing enough space between these and the two controls. These samples should also be dissolved in dyeless formamide loading solution. 5. Run the gel at 1200 V until the bromophenol blue marker dye has migrated 20–25 cm from the well (approximately 5 h) (see Note 24). 6. Switch off the power, disconnect the electrodes, and remove the gel setup. Lay the setup on the bench, and gently remove the upper plate leaving the gel intact on the lower plate (see Note 21). 7. Cover the gel with Saran wrap, and gently peel it off the glass plate. It is best to start at a corner, and slowly peel back across and then down the gel. 8. Lay the gel with the Saran wrap side down, and cover the exposed side with Saran wrap. 3.3.3. Visualizing and Isolating the Crosslinked Products

1. Place an aluminium TLC plate with fluorescent indicator under the wrapped gel at the area where the DNA samples should be located, as identified by comparison to formamide marker in the first well (see Note 30). 2. Cover the samples for elution to prevent DNA damage. Shine the UV light on the rest of the gel to visualize the control samples. 3. DNA is identified as a shadow cast on the TLC plate under a shortwave UV lamp. The untreated control should present as one band corresponding to the denatured single-stranded DNA. The drug treated control is expected to present as two bands (if not fully crosslinked). The one migrating slower should correspond to the double-stranded, crosslinked species. 4. Mark the position of the crosslinked product of the treated control on the gel. Turn off the UV light before uncovering the rest of the gel. 5. Using the marked area of the control as a guide, identify the crosslinked species of the treated samples. 6. Cut out the areas of gel containing the crosslinked product with a scalpel, remove the Saran wrap, and place the gel slice in an Eppendorf tube. 7. Crush the piece of the gel to powder using a pipette tip, and add 1 mL of gel elution buffer. Mix well (see Note 31). 8. Incubate at 37ºC overnight. 9. Centrifuge at 16,000 g for 10 min and remove the supernatant to a fresh Eppendorf tube.

Measurement of DNA Interstrand Crosslinking in Naked DNA Using Gel-Based Methods

297

10. Precipitate the DNA by adding 0.1 vol of 3  M sodium acetate, 3 vol 95% ethanol and placing in a dry ice/ethanol freezing in a dry ice/IMS bath for 20 min. 11. Centrifuge at 16,000 g for 10 min and lyophilize to dryness (see Note 28). 12. Store dry at −20ºC until use (see Note 32).

4. Notes 4.1. Agarose GelBased Method

1. The restriction endonucleases considered should have a unique restriction site on the plasmid DNA to be linearized. Such enzymes can cleave the substrate DNA with varying efficiencies. 1 U/mg of BamHI efficiently cleaves supercoiled pUC19 under standard reaction conditions (as opposed to 5  U/mg required by HindIII or BspMI which even at 100  U/mg can achieve less than 10% cleavage). Complete digestion can be confirmed by electrophoresis of a small amount (1 mg) of the linearized plasmid and an equal amount of undigested closed-circular control on a 1% ethidium stained agarose minigel in TAE buffer. DNA is dissolved in nondenaturing loading buffer, loaded onto the gel and electrophoresed for 1 h at 75 V. Bands are visualized under UV. 2. Inclusion of tRNA in the stop solution is important as it facilitates DNA precipitation. 3. As polynucleotide kinase is inhibited by ammonium ions, precipitation steps prior to radiolabeling must only be carried out with sodium acetate. 4. The rate of crosslink formation varies markedly between different agents. For example, crosslinking agents such as the clinically used Busulphan produce low levels of crosslinks that form slowly, and are still increasing at 24 h. In contrast, crosslinks produced by mechlorethamine form within 1 h of exposure. Incubation times must be therefore adjusted according to the reactivity of the agent under investigation, as established by time course experiments. 5. The choice of appropriate denaturation conditions depends on the crosslinking agent under investigation. Although heat denaturation is most readily used, particularly for agents which crosslink via purine-N7 positions, some agents, e.g. pyrrolobenzodiazepine dimers, which crosslink through guanine-N2 positions induce interstrand crosslinks, which show some lability to heat treatments. Use of heat denaturation can therefore result in inaccurate estimate of the extent

298

Kiakos, Hartley, and Hartley

and efficiency of crosslinking (see Fig. 2). Other agents e.g. CC-1065-type bifunctional analogues such as bizelesin, which produce crosslinks through adenine-N3 positions, upon thermal denaturation produce DNA single-strand breaks due to the release of the covalently modified purines, in which case the sole presence of single-stranded DNA species can be misinterpreted as inability of those agents to crosslink DNA. Crosslinks of heat-labile nature should therefore be assessed with this assay by alkali denaturation. Finally, in rare cases, drugs may induce alkali-labile sites, necessitating the use of heat denaturation. 6. Autoradiographs can alternatively be obtained after shorter exposure of film with an intensifying screen, for 2–4  h at −70ºC, at the expense of the sharpness of the image. 7. Percentage crosslinking in each lane is calculated by measuring the total amount of DNA (summed density for the double-stranded and single-stranded bands) relative to the amount of crosslinked DNA (density of double-stranded band alone). Such quantitation provides a measure of crosslinking in a given DNA sample. The determined percentage level of crosslinked DNA can be then plotted against a range of drug concentrations providing a dose-response curve (see Figs. 1b and 2b). From this the XL50 value for a specific agent (the concentration of agent required for 50% crosslinking of plasmid DNA) can be extrapolated and the relative crosslinking efficiencies of various agents, under the conditions employed, compared. 8. Alternatively, the percentage level of crosslinked DNA can be plotted against time, providing a measure of the rate of formation of DNA interstrand crosslinks by an agent. In this case, drug treatments are continuous at 37ºC, and aliquots from a single reaction mixture are removed (and stopped) at relevant time intervals. Further processing of these samples proceeds as described in Subheading 3.1.4. When performing this experiment, concentrations which induce <100% of the plasmid DNA molecules to be crosslinked are selected since the rate to achieve maximum crosslinking can only be determined under these conditions. 9. This assay can also be used to specifically monitor the “second-arm” reaction as described in (8). Briefly, following an initial drug exposure, unbound drug is removed and DNA is then incubated in drug-free buffer. This allows crosslink formation during the postincubation period to be followed, resulting from the second-arm reaction of monoadducts going on to form crosslinks. Plots of the percentage of crosslinked DNA against time of the total and the second-arm

Measurement of DNA Interstrand Crosslinking in Naked DNA Using Gel-Based Methods

299

reaction provide a direct measure of the relative rates of the two arms of the crosslink reaction for a given agent. 10. Phosphorimaging can alternatively be used for the visualization and quantitation of crosslinked DNA. 4.2. Polyacrylamide Gel-Based Methods

11. Oligonucleotides can be complementary, self complementary, partially complementary, incorporate single or multiple sites of reactivity to the drug tested, bear modifications such as biotinylated ends, etc. Drug target sequences can also be manipulated with the inclusion of inosine (when for example probing the role of the N2 position of guanine in the crosslink formation) or introduction of mismatched bases. It is advisable that the oligonucleotide duplexes contain only one occurence of the potential interstrand crosslink sequence for the agent under investigation. This ensures that a single interstrand crosslink at only one site will be formed during drug treatment. Other potential drug binding sites (e.g. sequences of high reactivity favourable for the occurrence of monoadducts or intrastrand crosslinks) should be avoided in order to facilitate later biochemical analysis. The results described here employ complementary oligonucleotides centrally incorporating the consensus binding sequence (5¢-Pu-GATC-Py-3¢) of the potent crosslinking pyrrolobenzodiazepine dimer, SJG-136 (Figs. 3–5a) or that of cisplatin (5¢-GC-3¢) (Figs. 3 and 5b). 12. The HPSF purchased oligonucleotides do not require further purification. 13. Oligonucleotide stocks and their aliquots are stored at −20ºC. 14. The particular method described here has consistently produced fully annealed duplex products. The annealing process can alternatively be performed with a programmable thermal cycler. A simple protocol may comprise of an initial denaturation step at 90ºC for 3 min, a cycle at 70ºC for 40 s followed by repeated cycles of incremental decrease of temperature (e.g. 0.5ºC/cycle). Confirmation that oligonucleotides are in the duplex form can be checked using non-denaturing polyacrylamide gels. 15. It is not essential to heat the oligonucleotides to 90ºC. They can be annealed by heating a few degrees above their melting temperature for 5 min and then allowing to cool slowly. 16. Alternatively, singly end-labeled DNA duplexes can be prepared by radiolabeling one strand before combining and annealing it to its complementary strand to form the duplex. This approach can be useful if the exact base pair position of the crosslink on each DNA strand is to be analyzed.

300

Kiakos, Hartley, and Hartley

Briefly, the crosslinked product is excised from the denaturing polyacrylamide gel, eluted from the gel slices and subjected to a cleavage treatment which quantitatively converts the site of linkage to a strand break e.g. hot piperidine for guanine-N7 or heat for adenine-N3. Subsequent denaturing polyacrylamide gel electrophoresis can then assign the exact base position of the cleavage products and thereby the sequence of the exact sites of crosslink formation (2, 6). 17. The column is prepared immediately prior to use by loading 1 mL of Biospin-6 solution into an empty column, which is then spun at 1,200  g to pack and remove residual storage buffer. The packed column is then washed through twice with 200 mL of distilled H2O or TE buffer. 18. Drug treatment times and temperatures vary, and this will have to be optimized for each crosslinking agent. Oligonucleotide duplexes generally require higher doses and longer time of treatment to achieve similar levels of crosslinking as those observed in linear plasmid DNA. A preliminary experiment with overnight exposure is recommended. Drug reactivity and adduct stability should also be checked. 19. Addition of traces (1  mL) of glycogen (20  mg/mL) can enhance the precipitation process. 20. At this stage, it is advisable not to wash the pelleted DNA with 70% ethanol as it is possible that some DNA might go into solution. 21. Before using the gel apparatus, thoroughly clean the glass plates with ethanol. Siliconize the front plate by spreading approximately 2 mL of silanization solution evenly over the plate with a paper towel to ensure easy removal of the sequencing gel from the glass plates. This procedure must be carried out in a fume hood as the silanization solution is highly volatile and toxic. 22. Carefully remove the comb from the top of the gel. Before loading the samples, reshape the wells using a pipette tip if they have been damaged or deformed. Ensure that wells are well isolated from each other to avoid samples co-eluting. 23. Although crosslinks, which are not heat labile, can withstand the 3  min heat denaturation step described in the agarose gel-based method, samples can be loaded directly onto the polyacrylamide denaturing gel. The presence of formamide in the loading solution and urea in the sequagel buffers ensures full denaturation of duplex DNA under the conditions described. 24. The gel is run at low voltage so that the temperature does not rise above 25ºC avoiding potential breakdown of

Measurement of DNA Interstrand Crosslinking in Naked DNA Using Gel-Based Methods

301

thermally labile crosslink products and degradation of the oligonucleotides. 25. Polyacrylamide gels of such high percentage can often fragment upon drying. It is therefore essential that the temperature is kept below 60ºC and the dryer is connected to a good vacuum. If problems persist, the wet gel can be covered in Saran wrap and directly exposed to a film. 26. In some cases, e.g., mechlorethamine, in which crosslinked products are less stable, it is advisable to reconstitute the oligonucleotide, which will form the target duplex in 40 mM sodium cacodylate buffer. 27. A range of amounts of oligonucleotide duplex should be investigated. The crosslinking efficiency of some agents is low and the amount of crosslinked product may be too small to be visualized by UV shadowing. The limit of detection is approximately 2 µg per band. 28. Precipitated DNA samples must be thoroughly dried. 29. The control samples, as well as the samples to be eluted, should be dissolved in dyeless formamide loading solution. This facilitates the identification of bands and isolation of the crosslinked products, as the dyes interfere by presenting, like the DNA, as shadows on the TLC plate when visualized under UV light. 30. The TLC plate should be thin, so it can easily be placed under the gel for visualization. 31. The simple “crush-soak” method described here has been found to be very effective for recovery of DNA of ligation quality. Alternatively, commercially available kits for the isolation and purification of DNA from polyacrylamide gels can be used. 32. Purified crosslinked oligonucleotides can be reconstituted in distilled water. Before using them in subsequent biochemical procedures (e.g. phosphorylation and ligation into plasmids), the DNA concentration can be determined spectrophotometrically. References 1. Clingen PH, De Silva IU, McHugh PJ, Ghadessy FJ, Tilby MJ, Thurston DE, Hartley JA (2005) The XPF-ERCC1 endonuclease and homologous recombination contribute to the repair of minor groove DNA interstrand crosslinks in mammalian cells produced by the pyrrolo[2, 1-c][1.4]benzodiazepine dimer SJG-136. Nucleic Acids Res 33:3283–3291

2. Hartley JA, Souhami RL, Berardini MD (1993) Electrophoretic and chromatographic separation methods used to reveal interstrand crosslinking of nucleic acids. J Chromatogr 618:277–288 3. Sunters A, Springer CJ, Bagshawe KD, Souhami RL, Hartley JA (1992) The cytotoxicity, DNA crosslinking ability and DNA sequence selectivity of the aniline mustards

302

Kiakos, Hartley, and Hartley

melphalan, chlorambucil, and 4-[bis(2chloroethyl)amino] benzoic acid. Biochem Pharmacol 44:59–64 4. Bose DS, Thompson AS, Smellie M, Berardini MD, Hartley JA, Jenkins TC, Neidle S, Thurston DE (1992) Effect of linker length on DNA-binding affinity, cross-linking efficiency and cytotoxicity of C8-linked pyrrolobenzodiazepine dimers. J Chem Soc Chem Commun 20:1518–1520 5. Gregson SJ, Howard PW, Gullick DR, Hamaguchi A, Corcoran KE, Brooks NA, Hartley JA, Jenkins TC, Patel S, Guille MJ, Thurston DE (2004) Linker length modulated DNA cross-linking reactivity and cytotoxic potency of C8/C8¢ ether-linked C2-exo-Unsaturated Pyrrolo[2, 1-c][1, 4] benzodiazepine(PBD) dimers. J Med Chem 47:1161–1174

6. Berardini MD, Souhami RL, Lee CS, Gibson NW, Butler J, Hartley JA (1993) Two structurally related diaziridinylbenzoquinones preferentially cross-link DNA at different sites upon reduction with DT-diaphorase. Biochemistry 32:3306–3312 7. Smellie M, Bose DS, Thompson AS, Jenkins TC, Hartley JA, Thurston DE (2003) Sequence-selective recognition of duplex DNA through covalent interstrand crosslinking: kinetic and molecular modeling studies with pyrrolobenzodiazepine dimers. Biochemistry 42:8232–8239 8. Hartley JA, Berardini MD, Souhami RL (1991) An agarose gel method for the determination of DNA interstrand crosslinking applicable to the measurement of the rate of total and “second-arm” crosslink reactions. Anal Biochem 193:131–134

Chapter 19 An Evaluation Cascade for G-Quadruplex Telomere Targeting Agents in Human Cancer Cells Mekala Gunaratnam and Stephen Neidle Abstract The targeting of telomerase and telomere maintenance in human cancer cells can be achieved by small molecules that induce the 3¢single-stranded ends of telomeric DNA to fold up into four-stranded quadruplex structures that inhibit the action of the telomerase enzyme complex. In this chapter, we describe a series of biochemical, biophysical, and cellular assays that are used to evaluate the activity of new compounds, and so assess whether they are suitable for examination in xenograft models of human cancer. These assays evaluate quadruplex stabilisation properties, short- and long-term cell viability, telomerase enzymatic activity, cellular senescence, and telomere length changes. Key words: Telomere, Telomerase, Quadruplex, Senescence, Small molecule drugs

1. Introduction The role of eukaryotic telomeres is to protect chromosomal ends from unwanted recombination and degradation. They comprise an array of specialized telomeric proteins bound to telomeric DNA (1, 2), which itself consists of tandem repeats of simple guaninerich sequences, with a total length that ranges between ca 2 and 12  kb. In humans, this repeat is the sequence TTAGGG, and telomeric DNA is in the normal duplex form except for the final 100–200 3’-terminal residues, which are single-stranded. Telomeric DNA in normal somatic cells progressively shortens at each round of cell division as a consequence of the inability of DNA polymerase to fully replicate the ends. The consequence of this is that telomeres progressively shorten, until the Hayflick limit is reached, when cells stop replicating and enter a senescence phase. This normally precedes entry to apoptosis. Cancer cells maintain telomere length integrity, which plays a key role in cellular transformation and K.R. Fox (ed.), Drug-DNA Interaction Protocols, Methods in Molecular Biology, vol. 613, DOI 10.1007/978-1-60327-418-0_19, © Humana Press, a part of Springer Science + Business Media, LLC 1998, 2010

303

304

Gunaratnam and Neidle

immortalization (3). In the majority of cancer cell types, (>80%), this maintenance is the result of the expression of the telomerase specialized reverse transcriptase enzyme. Telomerase catalyses the synthesis of TTAGGG repeats during replication with the effect that telomeric DNA length remains constant in cancer cells. The enzyme has been validated as a therapeutic target (4) by a variety of approaches (notably using antisense, siRNA, dominant negative mutants or catalytic inhibitors), which have shown that inhibi­ tion of telomerase results in telomere shortening, and eventual senescence and selective cancer cell death. Telomerase function and its significance as an anti-cancer target have been extensively reviewed; especially recommended reviews are (5, 6). The concept that small molecules binding to the singlestranded overhang of human telomeric DNA can inhibit the activity of the telomerase enzyme was first demonstrated in vitro using a small library of amidoanthraquinone derivatives (7) with an amidoanthraquinone molecule stabilising the quadruplex. A large number of small molecules have subsequently been investigated as telomeric quadruplex ligands (8, 9). Cell-based studies have shown that they inhibit telomerase, resulting in telomere shortening and senescence (7, 8). An initially unexpected finding was that they produce short-term growth arrest, as a consequence of their ability to competitively displace bound telomeric proteins, especially hPOT1 and hTERT at the single-strand overhang (9–14). This displacement effectively exposes telomeric DNA, invoking a DNA damage response and consequent apoptosis. We describe here a cascade of assays currently in place in our laboratory (14–17), which enables a library of compounds, synthetic or natural products, to be systemativcally examined for evidence that they are active as quadruplex-binding telomere targeting agents. These assays comprise: ●●

●●

●●

●●

●●

Acute and long-term cytotoxicity: Effects on a panel of cancer cells and on normal cells, both to determine concentrations required to inhibit 50% of cell growth (IC50 values) and to determine whether at sub-IC50 concentrations cell growth can be inhibited following incubation, which is characteristic of telomerase inhibition and telomere uncapping. Quadruplex affinity: Selectivity for quadruplex vs. duplex DNA using a high-throughput assay that exploits the Fluorescence Resonance Energy Transfer (FRET) effect with quadruplex DNA (18). Inhibition of telomerase catalytic activity. The LIG-TRAP protocol is described (16), which ensures that the second PCR part of this assay is uncontaminated by ligand. Telomere length determination using Southern blotting – see Fig.  1 for an illustration of telomere shortening following long-term exposure. Induction of senescence using a b-galactosidase assay

An Evaluation Cascade for G-Quadruplex Telomere Targeting Agents

305

Fig.  1. Telomere length analysis of A2780 cells treated with the experimental drug BRACO19 for up to 5 weeks.

The reader may refer to the relevant literature for information on other important assays, especially for displacement of telomerecapping proteins hPOT1 and hTERT (11–13, 19), and for in vivo studies of anti-tumour activity (20, 21).

2. Materials 2.1. Cell Lines and Protein Extraction

The human ovarian cancer cell line A2780 was purchased from American Type Cell Culture and maintained in Dulbecco’s Modified Eagles Media containing 10% foetal bovine serum (Invitrogen, UK), 0.5 mg/ml hydrocortisone (Acros Chemicals, Loughborough, UK), 2  mM l-glutamine (Invitrogen, Netherlands), and nonessential amino acids 1× (Invitrogen, Netherlands) at 37°C, 5% CO2.

306

Gunaratnam and Neidle

Protein was extracted from exponentially growing cells and was used as the enzyme source. Total cell protein extraction and quantification was carried out using the Bradford assay. Cells were lysed in lysis buffer [50 mM HEPES (pH 7.4), 250 mM NaCl, 0.1% NP40/IGEPAL and protease cocktail inhibitors] on ice for 30 min (see Note 1). Samples were then centrifuged at 14,000 rpm for 15 min at 4°C and lysates were stored at −80°C. Bradford assay was performed by setting up a standard curve with BSA (concentration range of BSA from 1 to 30 µg/ml). Samples were diluted (1:200) in water and absorbance of both standards and samples in duplicates were read at 596 nm. 2.2. Ligands

Ligands in general should have their purity analytically confirmed. 10 mM stock solutions of the free bases are made in 100% DMSO followed by a further dilution to 1 mM in distilled water, adding 1% HCl. These 1 mM stock solutions are always freshly prepared before use in the assays.

2.3. Buffers and Reagents

1. TE buffer: 10  mM Tris–HCl pH 8.0 containing 1  mM EDTA 2. FRET buffer: 60 mM KCl, Kcacodylate, pH 7.4 3. TRAP buffer: 20 mM Tris–HCl, pH 8.3, containing 68 mM KCl, 1.5 mM MgCl2, 1 mM EGTA, 0.05% v/v Tween-20. 4. Lysis buffer: 10  mM Tris–HCl, pH 7.5, containing 1  mM MgCl2, 1  mM EGTA, 0.5% CHAPS, 10% glycerol, 5  mM b-mercaptoethanol, 0.1 mM AEBSF. 5. EB Buffer: 10 mM Tris–HCl, pH 8.5. 6. QIA quick nucleotide purification kit (Qiagen).

3. Methods 3.1. The FRET Assay

FRET oligonucleotides (Eurogentec Ltd., U.K.) have the sequences: F21T, 5¢FAM-d[G3(T2AG3)3]-TAMRA3¢; duplex, 5¢FAM-d [(TA)2GC(TA)2T6(TA)2GC(TA)2]-TAMRA3¢, where FAM is 6-carboxyfluoresein and TAMRA 6-carboxytetramethylrhodamine. 1. The oligonucleotide is suspended in FRET buffer at a concentration of 400  nM and heated to 85°C for 10  min prior to cooling to room temperature. 2. It is distributed (50 ml) across a 96 well RT-PCR plate (BioRad) to which ligand is added (50  ml; stored as a 20  mM DMSO stock, −20°C; diluted to 1  mM in HPLC grade DMSO) to afford the required concentration. 3. FRET buffer is used as a negative control.

An Evaluation Cascade for G-Quadruplex Telomere Targeting Agents

307

4. DNA melting is assessed with a MJ Research Opticon DNA Engine Continuous Fluorescence Detector exciting at 450– 495 nm. This is an RT-PCR machine, and a number of others are available that are suitable for this purpose. 5. Fluorescence values are recorded at 515–545  nm at 0.5°C intervals as the plate is heated from 30°C to 100°C. 6. The data are analyzed with the Origin 7.0 software package (Origin Lab Corp., Northampton, MA). 7. Melting curves can be fitted to sigmoid curves prior to analysis. 8. The change in melting temperature at 1 mM ligand concentration (DTm1mM) is calculated from four experiments by subtraction of the averaged negative control from the averaged 1  mM ligand melting temperature ± the maximum standard deviation (sd). 9. Competition Assay: This is performed as described above for the G-quadruplex experiment but with addition of varying concentrations of calf thymus DNA (Sigma-Aldrich, UK). 10. Concentrations expressed as G-tetrad:base pair ratios. The experiments were run at the ratios 1:1, 1:10, 1:100 and 1:300. 11. The percentage retained stabilization is calculated from three experiments, and normalized to the DTm1mM for that ligand with no CT-DNA competitor (100%) ± normalized sd. 3.2.The LIG-TRAP Telomerase Assay (see Fig. 2)

The TRAP assay has been modified from a two-step to the threestep TRAP-LIG procedure: step 1, initial primer elongation by telomerase and addition of ligand; step 2, subsequent removal of the ligand; step 3, PCR amplification of the products of telo­ merase elongation. 1. Step 1 Prepare a master mix containing the TS forward primer (0.1 mg; 5’-AAT CCG TCG AGC AGA GTT-3’), TRAP buffer, bovine serum albumin (0.05 mg), dNTPs (125 mM each), and protein extract (1000 ng/sample) diluted in lysis buffer. 2. The PCR master mix is added to tubes containing freshly prepared ligand at various concentrations and to a negative control containing no ligand. 3. The initial elongation step is carried out for 10 min at 30°C, followed by 94°C for 5 min and a final maintenance of the mixture at 20°C. 4. Step 2 To purify the elongated product and to remove the bound ligands, the QIA quick nucleotide purification kit (Qiagen) is used according to the manufacturer’s instructions. This kit is especially designed for the purification of both double and single-stranded oligonucleotides from 17 bases in length. It employs a high-salt buffer to bind the negatively

308

Gunaratnam and Neidle

Fig. 2. Inhibition of telomerase enzyme activity determined by the LIG-TRAP telomerase assay. Fig 2A (a) shows inhibition of telomerase activity in A2780 human ovarian protein extract by the known telomerase inhibitor AZT-triphosphate. Almost complete inhibition of telomerase is evident at a 50 µM concentration of compound. Fig 2B (b) shows LIG-TRAP gel of the inactive counterpart, the parent molecule AZT itself, where no inhibition of telomerase is seen up to 50 µM. Negative and positive control lanes are at left- and right-hand sides of each gel, respectively.

charged oligonucleotides to the positively charged spin-tube membrane through centrifugation, so that all other components, including positively charged and neutral ligand molecules would be eluted. PCR-grade water is then used (rather than the manufacturers’ recommendation of an ethanolbased buffer) to wash any impurities away before elution of the DNA using a low-salt concentration solution (see Note 2). 5. The purified samples are freeze-dried and then re-dissolved in PCR-grade water at room temperature prior to the second amplification step (see Notes 3 and 4). 6. Step 3 The purified extended sample is then subject to PCR amplification 7. A second PCR master mix is prepared consisting of ACX reverse primer (1 mM; 5¢-GCGCGG[CTTACC]3CTAACC-3¢), TS forward primer (0.1 mg; 5¢-AAT CCG TCGAGCAGAGTT-3¢), TRAP buffer, BSA (5 µg), 0.5 mM dNTPs, and 2U of TAQ polymerase (RedHot, ABgene, Surrey, UK).

An Evaluation Cascade for G-Quadruplex Telomere Targeting Agents

309

8. An aliquot of 10 µl of the master mix is added to the purified telomerase extended samples and amplified for 35 cycles of 94°C for 30 s, at 61°C for 1 min and 72°C for 1 min. 9. Samples are separated on a 12% PAGE and visualised with SYBR green (Aldrich) staining (see Note 5). 10. Gels are quantified using a gel scanner and gene tool software (Sygene, Cambridge, UK). 11. Drug samples are normalised against positive control containing protein only. 12. All samples are corrected for background by subtracting the fluorescence reading of negative controls. Data for all ligands are collected at a range of concentrations in order to obtain dose–response curves from which EC50 values (the concentration required for 50% enzyme inhibition, corresponding to a 50% decrease in total integrated ladder intensity) can be obtained by inspection. Graphs may be fitted to dose–response curves using, for example, the Origin 6.0 software package. 3.3. Sulforhodamine B short-term cytotoxicity assay

1. Cells are seeded (4,000 cells/wells) into the wells of 96-well plates in DMEM and incubated overnight to allow the cells to attach (see Notes 6 and 7). 2. Subsequently, cells are exposed to freshly made solutions of ligand at increasing concentrations and incubated for a further 96 h. 3. Following this, the cells are fixed with ice-cold trichloroacetic acid (TCA) (10%, w/v) for 30 min (see Note 8) and stained with 0.4% Sulforhodamine B (SRB) dissolved in 1% acetic acid for 15  min. All incubations are carried out at room temperature, except for the TCA incubation, which is at 4°C. 4. The IC50 value, the concentration required to inhibit cell growth by 50%, can be determined from the mean absorbance at 540 nm for each ligand concentration expressed as a percentage of the control untreated well absorbance.

3.4. Sub-Cytotoxic Long-Term Growth Inhibition Assay (see Fig. 3)

1. 1 × 105 cells are seeded in 75  cm3 tissue culture flasks and exposed to appropriate concentrations of ligands as single agents (or in combination) (see Note 9). 2. The concentrations are chosen according to individual IC50 values as determined in the SRB assay. 3. Cells are grown in a final volume of 10 ml DMEM and incubated as described previously. 4. Cells are exposed to ligands twice a week by replacing them with fresh media containing drug on day 3. 5. On day 7, media is removed and cells are washed with PBS once and trypsinised using 3 ml of trypsin.

310

Gunaratnam and Neidle

Accumulative Population Doublings

30

VC 2µM BRACO19 3µM BRACO19

25

20

15

10

5

0 0

1

2

3

4

Time (weeks) Fig. 3. Accumulated population doublings of the cis-platinum resistant human ovarian cancer cell line (A2780cis) exposed to sub-cytotoxic concentrations of the acridine compound BRACO19 over a period of 4 weeks.

6. Cells are then pelleted and re-suspended in 10 ml of DMEM and viability is determined with a haemocytometer. 7. Cells are counted with a Neubauer hemocytometer (Assistant, Germany) and the number of cellular population doublings are assessed by the equation n = (log Pn − log P0)/log2, where Pn is the number of cells collected and P0 the initial seeding density. 8. From this 1 × 105 cells are reseeded and the experiment is continued for a total of (for example) four weeks. 3.5. Staining for senescenceassociated b-galactosidase activity

1. MCF7 Cells (1 × 105, 10 ml media, ATCC-LGC Promochem) are exposed to two independent sub-cytotoxic concentrations of the required ligand over a 1 week period with a biweekly treatment. 2. A media-negative control is also screened. 3. Cells are stained for senescence using the b-galactosidase staining kit (Cell Signalling Technology) according to the manufacturer’s instructions (see Note 10). Cells are seeded (1 × 105, 2  ml) in a 6-well plate (Fisher-Scientific) with the required ligand concentration and incubated overnight. 4. The medium is removed, the well washed with PBS (2  ml) prior to fixing (1× fixative solution, 10 min).

An Evaluation Cascade for G-Quadruplex Telomere Targeting Agents

311

5. The fixative is removed and the well washed with PBS ( × 2 x 2 ml) prior to the addition of the staining solution (1 ml) and the plates incubated overnight. 6. Three independent fields of cells are visualized (200× magnification) from both repeats with the mean percentage of blue senescent cells reported ± sd. 3.6. Telomere length determination

1. DNA is extracted from cell pellets using the QIAGEN Blood and cell culture DNA Mini Kit following manufacture’s instructions. 2. The DNA concentration is determined by measurement of the absorbance at 260 nm using a GeneQuant spectrophotometer. Purity of each sample is determined as a ratio of absorbance at 260 nm/280 nm. 3. 20 mmol of the C-rich oligonucleotide d(CCTAACCCTAACCCTAACCC) is incubated with 80 µCi of g-32P -ATP, 1 ml of T4PNK enzyme (BioLabs), 3 ml of T4 Phosphonucleotide kinase buffer (BioLabs) and 16 µl of TE buffer for 1  h at 37°C. 4. Labelled oligonucleotide is purified using the QIAquick Nucleotide Removal Kit following manufacturer’s instructions. 5. The oligonucleotide is resuspended in 100 µl of EB buffer. 6. 2 µg of genomic DNA of each sample is digested with 1.5 µl of Hinf I and 1.5 µl Rsal restriction enzymes (Roche, Germany) and labeled with 3.5 ml of 32P-C-rich probe, in 3 µl of NEB buffer 2 (BioLabs), and up to 30 µl of TE buffer (pH 8.0) at 37°C overnight. 7. 5 ml of loading buffer is added to each tube to stop the reaction and samples are electrophoresed in 0.7% agarose gels with 0.5 mg/ml of ethidium bromide for 2.5 h at 115 V in 1× TBE buffer along with 35S DNA Marker (Amersham Biosciences, UK). 8. Subsequently, the gel is dried, firstly for two hours in filter paper and paper towel sandwich, followed by in a gel dryer for 20  min. The gel is then exposed to an X-ray film overnight (Molecular Dynamics) and telomeric DNA smears are visualized using a phosphorimager (Molecular Dynamics). See Fig. 19.1 for an example.

4. Notes 1. Protease cocktail inhibitors should be added to the lysis buffer freshly prior to the lysis of the pellets to increase yield.

312

Gunaratnam and Neidle

2. The TRAP assay is sensitive to even trace amounts of ethanol, which is present as residual in the PCR reaction mixture from the spin tube purification step. Even when 1% ethanol is present, it is found to inhibit the subsequent PCR step severely. To overcome this problem, it is first necessary to freeze dry the eluted samples in order to remove any traces of ethanol from the purification step. 3. TRAP Assay – extended products from the first PCR reaction should be dried thoroughly, but care should be taken to avoid over drying of samples, as this has shown to damage the samples and resulting in poor PCR extension. 4. Resuscitate freeze-dried extended products at room temperature, allow around 20  min for complete resuscitation, and add second PCR master mix directly to the same PCR tube. Avoid transferring resuscitated samples between tubes to prevent carry over sample loss. 5. Prior to running samples pre-run the gel for 5  min to equilibrate. 6. Cell culturing – Seeding density should be kept at 1:20 for subculturing A2780cis cells, as the doubling time is rapid. 7. Sulforhodamine B assay – optimum seeding density should be verified for each cell line in use as this varies considerably between different cancer lines. 8. TCA should be prepared freshly on the day of use and add enough TCA to fill about 2/3 of the well. 9. Long-term growth inhibition study – seeding density should be adjusted for each cell line in use 10. For staining for senescence-associated b-galactosidase staining the fixative and X-gal should be prepared freshly every time. References 1. de Lange T (2005) Shelterin: the protein complex that shapes and safeguards human telomeres. Genes Dev 19:2100–2110 2. Autexier C, Lue NF (2006) The structure and function of telomerase reverse transcriptase. Annu Rev Biochem 75:493–517 3. Kim NW, Piatyszek MA, Prowse KR, Harley CB, West MD, Ho PLC, Coviello GM, Wright WE, Weinrich R, Shay JW (1994) Specific association of human telomerase activity with immortal cells and cancer. Science 266: 2011–2015 4. Hahn WC, Stewart SA, Brooks MW, York SG, Eaton E, Kurachi A, Beijersbergen RL, Knoll JH, Meyerson M, Weinberg RA (1999)

Inhibition of telomerase limits the growth of human cancer cells. Nat Med 5:1164–1170 5. Oganesian L, Bryan TM (2007) Physiological relevance of telomeric G-quadruplex formation: a potential drug target. Bioessays 29:155–165 6. De Cian A, Lacroix L, Douarre C, TemimeSmaali N, Trentesaux C, Riou J-F, Mergny J-L (2008) Targeting telomeres and telomerase. Biochimie 90:131–155 7. Sun D, Thompson B, Cathers BE, Salazar M, Kerwin SM, Trent JO, Jenkins TC, Neidle S, Hurley LH (1997) Inhibition of human telomerase by a G-quadruplex-interactive compound. J Med Chem 40:2113–2116

An Evaluation Cascade for G-Quadruplex Telomere Targeting Agents 8. Monchaud D, Teulade-Fichou MP (2008) A hitchhiker’s guide to G-quadruplex ligands. Org Biomol Chem 6:627–636 9. Tan JH, Gu LQ, Wu JY (2008) Design of selective G-quadruplex ligands as potential anticancer agents. Mini Rev Med Chem 8:1163–1178 10. Mergny J-L, Lacroix L, Teulade-Fichou MP, Hounsou C, Guittat L, Hoarau M, Arimondo PB, Vigneron J-P, Lehn J-M, Riou J-F, Garestier T, Hélène C (2001) Telomerase inhibitors based on quadruplex ligands selected by a fluorescence assay. Proc Natl Acad Sci U S A 98:3062–3067 11. Gomez D, Wenner T, Brassart B, Douarre C, O’Donohue M-F, El Khoury V, Shin-ya K, Morjani H, Trentesaux C, Riou J-F (2006) Telomestatin-induced telomere uncapping is modulated by POT1 through G-overhang extension in HT1080 human tumor cells. J Biol Chem 281:38721–38729 12. Gomez D, O’Donohue M-F, Wenner T, Douarre C, Macadré J, Koebel P, GiraudPanis M-J, Kaplan H, Kolkes A, Shin-ya K, Riou J-F (2006) The G-quadruplex ligand telomestatin inhibits POT1 binding to telomeric sequences in  vitro and induces GFPPOT1 dissociation from telomeres in human cells. Cancer Res 66:6908–6912 13. Leonetti C, Amodei S, D’Angelo C, Rizzo A, Benassi B, Antonelli A, Elli R, Stevens MFG, D’Incalci M, Zupi G, Biroccio A (2004) Biological activity of the G-quadruplex ligand RHPS4 (3, 11-difluoro-6, 8, 13-trimethyl8H-quino[4, 3, 2-kl]acridinium methosulfate) is associated with telomere capping alteration. Mol Pharm 66:1138–1146 14. Gunaratnam M, Greciano O, Martins C, Reszka AP, Schultes CM, Morjani H, Riou J-F, Neidle S (2007) Mechanism of acridinebased telomerase inhibition and telomere shortening. Biochem Pharmacol 74:679–689 15. Schultes CM, Guyen B, Cuesta J, Neidle S (2004) Synthesis, biophysical and biological

16.

17.

18.

19.

20.

21.

313

evaluation of 3, 6-bis-amidoacridines with extended 9-anilino substituents as potent G-quadruplex-binding telomerase inhibitors. Bioorg Med Chem Lett 14:4347–4351 Reed JE, Gunaratnam M, Beltran M, Reszka AP, Vilar R, Neidle S (2008) TRAP-LIG, a modified TRAP assay to quantitate telomerase inhibition by small molecules. Anal Biochem 380:99–105 Moore MJ, Schultes CM, Cuesta J, Cuenca F, Gunaratnam M, Tanious FA, Wilson WD, Neidle S (2006) Trisubstituted acridines as G-quadruplex telomere targeting agents. Effects of extensions of the 3, 6- and 9-side chains on quadruplex binding, telomerase activity, and cell proliferation. J Med Chem 49:582–599 De Cian A, Guittat L, Shin-Ya K, Riou J-F, Mergny J-L (2005) Affinity and selectivity of G4 ligands measured by FRET. Nucleic Acids Symp Ser (Oxf) 235–236 Brassart B, Gomez D, De Cian A, Paterski R, Montagnac A, Qui KH, Temime-Smaali N, Trentesaux C, Mergny J-L, Gueritte F, Riou J-F (2007) A new steroid derivative stabilizes G-quadruplexes and induces telomere uncapping in human tumor cells. Mol Pharmacol 72:631–640 Burger AM, Dai F, Schultes CM, Reszka AP, Moore MJ, Double JA, Neidle S (2005) The G-quadruplex-interactive molecule BRACO19 inhibits tumor growth, consistent with telomere targeting and interference with telomerase function. Cancer Res 65:1489–1496 Leonetti C, Scarsella M, Riggio G, Rizzo A, Salvati E, D’Incalci M, Staszewsky L, Frapolli R, Stevens MF, Stoppacciaro A, Mottolese M, Antoniani B, Gilson E, Zupi G, Biroccio A (2008) G-quadruplex ligand RHPS4 potentiates the antitumor activity of camptothecins in preclinical models of solid tumors. Clin Cancer Res 14:7284–7291

Index A Absorbance......................................... 25–28, 30, 33, 38–42, 45, 47–53, 64, 68, 78, 167, 247, 306, 309, 311 Acceleration voltage......................................................... 98 Accelerator mass spectrometry............................... 103–117 Actinomycin..............................................57, 194, 240, 241 Adherent cell lines.................................................. 269, 271 A-DNA.......................................................................... 184 Adriamycin...............103–106, 108–109, 112–116, 143, 144 Adriamycin-DNA adducts..................................... 103–117 Agarose gel electrophoresis............. 176, 178–180, 182, 183, 187–190, 293 Alkylation........................................ 174, 175, 183, 289, 293 AMBER................................................................. 125, 127 Amidoanthraquinone..................................................... 304 Aminoglycoside................................................................ 57 Analyte.....................................17, 19, 41, 47, 77, 87, 89, 99 Angiogenesis.................................................................. 224 Anomalous diffraction............................................ 133–149 Anomalous map......................................138, 146, 147, 149 Anomalous scattering............................................. 133–139 Anthracycline......................................................... 142, 240 Anthraquinone....................................................... 241, 304 Antitumor antibiotic...................................................... 173 Apoptosis.........................................................176, 303, 304 Argand plot.................................................................... 134 Ascites.................................................................... 269, 273 Association constant........................................... 58, 89–100 Association frequency factor (AFF)..................... 81–83, 85 Autoradiography...............162–163, 204, 289, 290, 293, 295

B Backbone conformation.................................................... 56 Backbone tracking.................................................. 138, 139 Baseline determination..................................................... 33 Beer–Lambert law...................................................... 42, 50 Berenil........................................................................ 76, 78 Biacore................................................. 5, 8, 9, 12, 14, 17–21 Bijvoet pair...................................... 134, 138, 139, 142, 149 Binding constant..................14, 17, 19, 21, 44, 53, 164, 240 Binding curve.......................................... 19, 45, 80, 90, 167 Biosensor...................................................................... 1–21 Biotin......................................................... 5, 10, 12, 13, 259 Biotinylated TFO................................................... 258, 260 Boundary conditions.............................................. 125, 129

BRACO19............................................................. 305, 310 Busulphan............................................................... 267, 297

C Calorimetry...................................................................... 57 Cambridge Structural Database..................................... 124 Capillary electrophoresis............................................ 71–88 Carboplatin.................................................................... 267 Cell culture..............................................106, 116, 305, 311 CHARMM.................................................................... 125 Chlorambucil.......................................................... 267, 279 Chloroethyl-nitrosourea................................................. 267 Chloroquine................................................................... 252 Chromatin....................... 174, 176, 178–180, 185–191, 223 Circular dichroism...........................................26, 31, 37–45 Cisplatin...........................176, 267, 278, 285, 287, 288, 299 C-myc............................................................................... 29 Combinatorial selection.................................................. 194 Comet assay............................................................ 267–281 Competition assay.......................................................... 307 Competition equilibrium dialysis..................................... 57 Configurational entropy..................................123, 128, 130 Cooperativity........................................................ 14–16, 26 Copper phenanthroline.................................................. 154 Coumarin antibiotics.............................................. 262, 265 Crosslinking...........................................174–178, 180–183, 186, 267–281, 283–301 Cryptolepine........................................................... 240, 241 Crystal structure..............................................124, 142, 155 Cy3................................................................................... 28 Cy5................................................................................... 28 Cyclic amplification and selection of targets (CASTing)......................................................... 195 Cyclophosphamide......................................................... 267 Cytotoxicity assay........................................................... 309

D Dabcyl.............................................................................. 28 DB293.................................................................... 6, 12–17 Densitometry...................................................163, 285, 295 Dephosphorylation................................................. 285, 292 Differential cleavage plot................................................ 163 Dimethylsulphate (DMS).............................................. 230 Directed molecular evolution......................................... 193 Dissociation constant.......................... 71, 86, 164, 165, 203

315

Drug-DNA Interaction Protocols 316  Index

  

Dissociation kinetics....................................................... 219 Distamycin....................................................76, 78, 84, 194 DNA concentration determination................................ 64, 76, 111, 164, 165, 247, 301, 311 DNA gyrase...................................................... 257–264 DNA-platinum complex.......................................... 134 DNA triplex............................................... 55, 257–265 extraction...................................................106, 109–110 hairpins................................................................. 11–14 purification................................ 107, 110, 116, 211, 246 relaxation................... 125–127, 235–252, 258, 264, 265 repair..................................................235, 268, 279, 283 replication............................................56, 235, 268, 304 secondary structure............................223, 225, 227, 228 supercoiling................................................258, 260–265 DNA·RNA hybrids.................................................... 55–68 DNase I........................... 153–170, 207, 223, 225, 227–231 DNase II......................................................................... 154 Doxorubicin.............................................103, 208, 214, 215 Duocarmycin.......................................................... 174–176 Dynamics.................................... 2, 120, 121, 124, 224, 225

E Echinomycin.......................................................... 208, 240 Egr-1 transcription factor............................................... 224 Electropherogram....................................................... 75, 77 Electrophoresis...............................................158, 162–163, 169, 179, 184, 186–188, 199, 200, 208, 211, 213, 217, 218, 229, 230, 244, 245, 247–249, 251, 258, 268, 270, 276, 280, 281, 284, 293, 295, 297 Electrospray mass spectrometry................................ 89–100 Electrostatic forces.......................................................... 142 Elinafide......................................................................... 240 Ellipticine................................................................. 59, 241 Ellipticity.................................................................... 27, 43 Elongation.............................................. 209, 211, 212, 214, 216–217, 219, 220, 307 3’-End labelling...................................................... 159–161 5’-End labelling...............................................161, 284, 292 Energy-minimisation............................................. 126, 130 Enzyme inhibition.......................................................... 309 Equilibrium constant.......................................2, 6, 7, 14, 72 Ethidium.............................................. 51, 57–59, 188, 199, 211, 212, 220, 240–252, 259, 265, 281, 297, 311 Extinction coefficient......................... 29, 43, 64, 76, 91, 92, 198, 200, 247

F FAM....................................................................28, 33, 306 Film LD........................................................................... 46 FK317......................................................174, 177, 181, 182 Fluorescein................................................................. 28, 33 Fluorescence detection........................................................82, 83, 105

donor.................................................................... 27, 33 quencher............................................................... 27, 58 Fluorescence resonance energy transfer (FRET)........................ 27, 28, 31, 33, 304, 306–307 Footprinting.............................. 87, 153–170, 208, 223–231 Force-field.......................................................123, 125, 128 Formaldehyde...................103–105, 114, 115, 211, 214, 216 Fourier coefficients......................................................... 134 Free energy................................................73, 120, 127, 142

G b-Galactosidase activity.......................................... 310–311 Gaussian distribution..................................................... 129 Gel electrophoresis...........................................29, 169, 177, 199, 217, 245, 247, 289–291 Gel purification...................................................... 158, 168 Gene expression............................................................ 1, 56 Generalized Born................................................... 121, 128 Gibbs fee energy................................................................. 2 Global fitting.................................................................. 241 G-quadruplex........................................... 26, 91, 94, 97, 98, 225, 227, 230, 231, 303–312 G-quartet................................................................... 25, 26 Growth inhibition assay......................................... 309–310

H Hairpin polyamide..........................................154, 170, 194 Hayflick limit................................................................. 303 High-throughput screening.................................... 262–263 Histones...........................176–178, 181, 185, 186, 189, 190 Hoechst 33258..............................................51, 76, 78, 127 HPLC........................................... 29, 48, 76, 105, 137, 306 HTert...................................................................... 304, 305 Hybridization..................................................................... 7 Hydrophobic interactions................................................. 37 Hydroxyl radical..............................................154, 155, 207

I IC50.................................62, 63, 68, 262–263, 265, 304, 309 Ifosfamide....................................................................... 267 I-motif.............................................................................. 64 Implicit solvent simulations............................................ 128 Induced CD (ICD)...............................................38, 44, 45 Initiation of transcription................................208, 213, 216 Intercalation..............................................56, 173, 235–252 Intercalator.........................45, 51, 56, 58, 59, 164, 241, 251 Intermolecular triplex..................................................... 258 Interstrand crosslink................ 173, 176, 267–281, 283–301 In vitro chromatin assembly............176, 178–180, 185–190 In vitro transcription.......................................178, 207–221 Ionic strength.................................................29, 44, 52, 87, 96, 97, 128, 169 Isomorphous replacement...............................139–141, 147 Isotropic absorbance......................................................... 45

Drug-DNA Interaction Protocols 317 Index     



K

P

Kinetics....................................................................2, 3, 6, 7 Klenow fragment............................................................ 159 KM..........................................................................62, 63, 68

Patterson map................................................................. 134 Phosphorimaging............................................163, 219, 299 Ping-pong kinetics........................................................... 62 Pixantrone....................................... 208, 211, 214–216, 219 Plasmid footprinting.......................156–160, 168, 223–231 Plasmid preparation.........................................158, 159, 178 Poisson-Boltzmann........................................................ 121 Poly(U)............................................................................. 64 Polyelectrolyte.......................................................... 29, 142 Polymerase chain reaction (PCR).....................31, 195–204, 220, 227, 229, 231, 304, 306–308, 312 Polynucleotide kinase (PNK).................................159, 161, 178, 183, 225, 228, 289, 292, 294, 297, 311 Poly(A)·poly(dT).............................................57, 62, 64, 66 Poly(A)·poly(U).......................................................... 64, 66 Poly(dA)·poly(dT)...................................................... 64, 66 Positive supercoils............................................242, 249, 250 Pot1........................................................................ 304, 305 Principal component analysis..........................125, 126, 130 Promoters................................................... 4, 156, 208, 209, 220, 223–231 Protein data bank (PDB)................................................ 124 Psoralen.......................................................................... 176 pUC vectors............................................................ 157, 168

L Lac uv5 promoter....................................208–214, 216, 220 Law of mass action........................................................... 72 Lexitropsins.................................................................... 119 LIG-TRAP telomerase assay.................................. 307–309 Linear dichroism (LD)............................................... 37–54 Liquid scintillation counting..................................104, 105, 108, 111, 113, 114, 117 Luciferase....................................................................... 231 Lymphocytes........................... 269, 271–272, 276, 279, 280

M Mass spectrometry...........................................29, 87, 89, 96 Mass transfer effect.........................................17, 18, 20, 21 Maxam–Gilbert.............................................................. 161 Mechloroethamine......................................................... 267 Melphalan...............................................176, 267, 276–279 Methidiumpropyl-EDA.Fe(II)....................................... 154 Micrococcal nuclease...............................154, 179, 185–188 Minor-groove-binding ligand...................................................124, 154, 165 Mithramycin................................................................... 154 Mitomycin (MMC).........................................174–176, 267 Mitosene................................................................. 174, 175 Mitoxantrone...................................................208, 214, 215 Molecular dynamics simulation..............................120, 121, 125, 126, 128 Molecular modelling.............................................. 119–130 Monoalkylation.......................................176, 178, 183–185 Multidrug resistance....................................................... 116

N Negative ion mode..................................................... 92, 97 Negative supercoils.......................... 223–225, 242, 250, 257 Netropsin.............................................................. 76–78, 81 Nitrogen mustards...........................................176, 208, 267 NMR...................................................................... 124, 125 Nogalamycin................................................................... 154 Non-covalent interaction.................................................. 37 Normal mode analysis (NMA)........................122, 128, 130 Nucleosomal DNA................................................. 173–191 Nucleosome....................................................173, 176–178, 180, 181, 183, 185, 186, 188–191

O Oligonucleotide annealing.............................................. 294 Oncogene................................................................... 4, 224 Oxaliplatin...................................................................... 267

Q Quadruplex................... 3, 4, 25–30, 32, 55, 56, 64, 136, 304 Quantitative footprinting....................................... 164–165 Quinoxaline antibiotic.................................................... 154

R Radiolabel................................................... 3, 104, 105, 154, 159, 168, 169, 183, 190, 294 Random collisions............................................................ 73 Rate constant..................................... 2, 6, 14, 16, 17, 21, 73 Real time PCR................................................................. 31 Refractive index...............................................3–5, 7, 12, 20 Repressor protein............................................................ 146 Restriction endonuclease protection assay (REPA)................................................200–202, 204 Restriction Endonuclease Protection Selection and Amplification (REPSA)...................... 193–204 Reverse transcriptase......................... 56, 158–160, 168, 304 Rhodamine....................................................................... 28 RNA polymerase..............................................56, 208–210, 212, 213, 218–220, 223 RNA secondary structures................................................ 55 RNase H....................................................56, 57, 60–63, 65 Rubidium........................................................134, 136, 146

S S1 nuclease......................................................223, 225–230 Scatchard plot................................................................... 71

Drug-DNA Interaction Protocols 318  Index

  

SELEX........................................................................... 193 Senescence.......................................................303, 304, 310 Sensorchip.....................................................5, 8–12, 18, 19 Sensorgram..................................................5–15, 17, 20, 21 Sequence recognition...................................................... 157 Sequence specificity.................................142, 194, 220, 284 Sequencing............................................. 197, 201, 210, 211, 213, 217, 226, 228, 231, 290, 291, 294 Single cell gel electrophoresis (Comet) assay............................................. 267–281 Single hit kinetics................................................... 154, 170 Sodium adducts.................................................. 94, 98–100 Solid tumour tissue................................................. 269, 272 Southern blotting........................................................... 304 Sp1 transcription factor.......................................... 223, 224 Spectropolarimeter..................................................... 39, 40 Spectroscopy..........................................................37, 45, 53 SPR angle....................................................................... 5, 6 Spreadsheet................................................................ 81, 84 Stacking...............................................................76, 84, 238 Steady state..............................................5–7, 14–18, 20, 21 Stoichiometry.....................................................2, 7, 12, 14, 18, 84, 90, 165, 236 Streptavidin.......................................................5, 8–10, 260 Sugar pucker..................................................................... 56 Sulforhodamine B (SRB)....................................... 309, 312 Supercoiled DNA.................................................. 235–252, 257, 258, 260, 276 Surface plasmon resonance (SPR)................................ 1–21 Suspension cell line................................................ 269, 270 SYBR gold.....................................................178, 179, 181, 182, 188, 190, 252, 260, 261, 264 Synchrotron.................................................................... 137

T T1/2.............................................................................. 31–33 TAMRA..............................................................28, 33, 306 Telomerase........................................... 56, 58, 304, 307–309 Telomere............................................................. 4, 303–312 Telomere length determination.............................. 304, 311 Telomeric repeat amplification protocol (TRAP) assay..............................304, 306–309, 312

Telomestatin............................................226–228, 230, 231 TFIID............................................................................ 194 Thallium...........................134–136, 139, 142, 144, 146, 148 Thermal denaturation....................................27, 57, 60, 298 Thermal melting..............................................25–33, 62, 67 Thermodynamics........................................................ 3, 142 Thermodynamic stability.................................................. 56 Thiazole orange................................................................ 59 Tm...........................................6, 28–31, 33, 60, 62, 299, 307 Topoisomerase................................... 59, 236, 238, 257–265 Topoisomerase I.............................. 235–252, 258, 259, 261 Topoisomerase II.............................................142, 236, 252 Trajectory validation............................................... 125–127 Transcriptional footprinting........................................... 208 Transition moment polarization................................. 39, 45 Triplex.......................................................... 25, 56, 64, 165, 166, 258, 260, 262, 264 Triplex forming oligonucleotide (TFO).................154, 155, 158, 162, 164–167, 169, 170, 258 Type IIS restriction endonuclease.......................... 194, 197

U Uranyl photoclevage....................................................... 154 UV spectrophotometry......................................92, 198, 200

V Vascular endothelial growth factor (VEGF).......................................223–227, 230, 231 Vibrational frequencies................................................... 122 Vibrational modes.......................................................... 122

W Water molecules....................................... 73, 121, 126, 128, 129, 135, 142–144, 146, 147

X X-ray scattering.............................................................. 133 X-ray structure........................................................ 144, 176

Z Z-DNA.................................................................... 55, 137