Atomic Force Microscopy in Biomedical Research: Methods and Protocols (Methods in Molecular Biology Vol 736)

Methods

in

Molecular Biology™

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



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

wwwwwwwwwwwwww

Atomic Force Microscopy in Biomedical Research Methods and Protocols

Edited by

Pier Carlo Braga Department of Pharmacology, School of Medicine, University of Milan, Milan, Italy

Davide Ricci Robotics, Brain and Cognitive Sciences Department, Italian Institute of Technology, Genoa, Italy and Department of Biophysical Electronic Engineering, University of Genoa, Genoa, Italy

Editors Pier Carlo Braga Department of Pharmacology School of Medicine University of Milan Milan, Italy [email protected]

Davide Ricci Robotics, Brain and Cognitive Sciences Department, Italian Institute of Technology Genoa, Italy and Department of Biophysical Electronic Engineering University of Genoa Genoa, Italy [email protected]

ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-61779-104-8 e-ISBN 978-1-61779-105-5 DOI 10.1007/978-1-61779-105-5 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2011926794 © Springer Science+Business Media, LLC 2011 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or ­dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Printed on acid-free paper Humana Press is part of Springer Science+Business Media (www.springer.com)

Preface The invention and development of the optical microscope in the seventeenth century revealed the presence of a previously unseen and unimaginable world within and around us. Our lives would not be what they are today if optical microscopy had never existed or if it had not helped us to understand better what we are, how we function, and how we can improve our condition – first in the fields of biology and medicine, and then in many other fields. Another great step was made with the introduction of transmission and scanning electron microscopy in the 1930s, which was initially integrated with optical microscopy but subsequently developed its own identity and technology and opened up new horizons in human knowledge. Starting in 1986, further technological advances led to the development of atomic force microscopy (AFM), which is completely different from its predecessors: instead of being based on lenses, photons, and electrons, it directly explores the surface of the sample by means of a local scanning probe while the use of dedicated software allows the results to be visualized on a monitor. AFM has a number of special characteristics: very high magnification with very high resolution; minimal sample preparation (none of the dyes of optical microscopy, or the vacuum, critical point, or gold sputtering required by scanning electron microscopy); real three-dimensional topographical data that allow us to obtain different views of the samples from a single collected dataset; and the ability to work in a liquid in real time, thus making it possible to study the dynamic phenomena of living specimens in their biological environment and under near-physiological conditions. Over the years, an increasing number of researchers have started to use AFM and, in addition to a wide range of scientific articles, there are now also various books on the subject. In 2004, we edited a book published by Humana Press (Atomic Force Microscopy: Biomedical Methods and Applications) that described a series of practical AFM procedures in various applications with the aim of stimulating researchers to use the technique. We were therefore surprised when Humana Press proposed the publication of a second book on the subject so quickly after the first, and hesitated to accept the challenge. However, upon further reflection, we had to agree that the sheer breadth and originality of the new applications that have emerged since the first book was published more than justified this further review. The reason is quite simple: AFM is no longer simply just another form of microscopy, but has given rise to a completely new way of using microscopy that fulfils the dreams of all microscopists: being able to touch, move, and interact with the sample while it is being examined, thus making it possible to discover not only morphological, but also chemical and physical structural information. Optical microscopy made it possible to talk at the “micron” level (cells), and ­transmission and scanning electron microscopy introduced the idea of the “nano” level (sub-cellular), but still only in two dimensions; however, when speaking of AFM, it is not only usual to talk in three-dimensional “nano” terms, but it is also already possible to talk

v

vi

Preface

at the “pico” level (molecular). Together with continuous technical improvements, the reaching of this new dimensional range means that AFM can provide an opportunity to interact with individual molecules, observing them while we touch them and move them around in order to be able to discover their physical characteristics. All of this has also led to the development of a parallel “nano-technology” insofar as an AFM workstation has become a “nano-robot” that can dynamically interact with and manipulate samples on a “nano-scale”, and acquire information of sub-pico Newton “force spectroscopy” data on which to base the study of “nano-biology”. Functionalizing the AFM tip has made it possible to obtain “nano-biosensors” that can be used in the field of dynamic biomolecular processes in ways that could not even be imagined just a few years ago. Finally, combining AFM with other microscopic techniques, such as confocal or fluorescence microscopy is now being actively explored, and a number of interesting synergies have been discovered. This book brings together different types of applications in order to provide examples from different fields in the hope that this will stimulate researchers to apply their ingenuity in their own specialization and allow them to add significant originality to their studies. We gratefully acknowledge all of the contributions of our colleagues, each of whom donated their experience in order to cross-fertilize this new and fascinating technology. “GOD BLESS MICROSCOPY (ALL TYPES) …AND MICROSCOPISTS TOO ” because they show us what and how wonderful life is. Milan, Italy  Genoa, Italy 

Pier Carlo Braga Davide Ricci

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

v xi

Part I The Basics of Atomic Force Microscopy   1 How the Atomic Force Microscope Works? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bruno Torre, Davide Ricci, and Pier Carlo Braga   2 Measurement Methods in Atomic Force Microscopy . . . . . . . . . . . . . . . . . . . . . . Bruno Torre, Claudio Canale, Davide Ricci, and Pier Carlo Braga   3 Recognizing and Avoiding Artifacts in Atomic Force Microscopy Imaging . . . . . . Claudio Canale, Bruno Torre, Davide Ricci, and Pier Carlo Braga

3 19 31

Part ii Molecule Imaging   4 Imaging the Spatial Orientation of Subunits Within Membrane Receptors by Atomic Force Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stewart M. Carnally, J. Michael Edwardson, and Nelson P. Barrera   5 High Resolution Imaging of Immunoglobulin G Antibodies and Other Biomolecules Using Amplitude Modulation Atomic Force Microscopy in Air . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sergio Santos and Neil H. Thomson   6 Atomic Force Microscopy of Ex Vivo Amyloid Fibrils . . . . . . . . . . . . . . . . . . . . . Claudio Canale, Annalisa Relini, and Alessandra Gliozzi   7 Studying Collagen Self-Assembly by Time-Lapse High-Resolution Atomic Force Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clemens M. Franz and Daniel J. Muller   8 Atomic Force Microscopy Imaging of Human Metaphase Chromosomes in Liquid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Osamu Hoshi and Tatsuo Ushiki   9 Atomic Force Microscopy of Proteasome Assemblies . . . . . . . . . . . . . . . . . . . . . . Maria Gaczynska and Pawel A. Osmulski 10 Atomic Force Microscopy of Isolated Mitochondria . . . . . . . . . . . . . . . . . . . . . . . Bradley E. Layton and M. Brent Boyd 11 Imaging and Interrogating Native Membrane Proteins Using the Atomic Force Microscope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andreas Engel

vii

47

61 81

97

109 117 133

153

viii

Contents

Part III Nanoscale Surface Analysis and Cell Imaging 12 Atomic Force Microscopy Investigation of Viruses . . . . . . . . . . . . . . . . . . . . . . . . Alexander McPherson and Yurii G. Kuznetsov 13 Determination of the Kinetic On- and Off-Rate of Single Virus–Cell Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Christian Rankl, Linda Wildling, Isabel Neundlinger, Ferry Kienberger, Hermann Gruber, Dieter Blaas, and Peter Hinterdorfer 14 Atomic Force Microscopy as a Tool for the Study of the Ultrastructure of Trypanosomatid Parasites . . . . . . . . . . . . . . . . . . . . . . . . Wanderley de Souza, Gustavo M. Rocha, Kildare Miranda, Paulo M. Bisch, and Gilberto Weissmuller 15 Normal and Pathological Erythrocytes Studied by Atomic Force Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andreas Ebner, Hermann Schillers, and Peter Hinterdorfer 16 The Growth Cones of Living Neurons Probed by the Atomic Force Microscope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Davide Ricci, Massimo Grattarola, and Mariateresa Tedesco 17 Highlights on Ultrastructural Pathology of Human Sperm . . . . . . . . . . . . . . . . . . Narahari V. Joshi, Ibis Cruz, and Jesus A. Osuna 18 High-Speed Atomic Force Microscopy and Biomolecular Processes . . . . . . . . . . . Takayuki Uchihashi and Toshio Ando

171

197

211

223

243 259 285

Part IV Non-topographical Applications (Force-Spectroscopy) 19 Atomic Force Microscopy in Mechanobiology: Measuring Microelastic Heterogeneity of Living Cells . . . . . . . . . . . . . . . . . . . . . Evren U. Azeloglu and Kevin D. Costa 20 Force-Clamp Measurements of Receptor–Ligand Interactions . . . . . . . . . . . . . . . Félix Rico, Calvin Chu, and Vincent T. Moy 21 Measuring Cell Adhesion Forces: Theory and Principles . . . . . . . . . . . . . . . . . . . . Martin Benoit and Christine Selhuber-Unkel 22 Nanoscale Investigation on E. coli Adhesion to Modified Silicone Surfaces . . . . . . Ting Cao, Haiying Tang, Xuemei Liang, Anfeng Wang, Gregory W. Auner, Steven O. Salley, and K.Y. Simon Ng

303 331 355 379

Part V Investigating Drug Action 23 Imaging Bacterial Shape, Surface, and Appendages Before and After Treatment with Antibiotics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391 Pier Carlo Braga and Davide Ricci 24 Thymol-Induced Alterations in Candida albicans Imaged by Atomic Force Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401 Pier Carlo Braga and Davide Ricci

Contents

ix

25 Atomic Force Microscope-Enabled Studies of Integrin–Extracellular Matrix Interactions in Vascular Smooth Muscle and Endothelial Cells . . . . . . . . . 411 Zhe Sun and Gerald A. Meininger 26 Atomic Force Microscopy Studies on Circular DNA Structural Changes by Vincristine and Aspirin . . . . . . . . . . . . . . . . . . . . . . . . . . . 425 Zhongdang Xiao, Lili Cao, Dan Zhu, and Zuhong Lu

Part VI Atomic Force Microscopy as a Nanotool 27 Combined Atomic Force Microscopy and Fluorescence Microscopy . . . . . . . . . . . Miklós S.Z. Kellermayer 28 Chemical Modifications of Atomic Force Microscopy Tips . . . . . . . . . . . . . . . . . . Régis Barattin and Normand Voyer 29 Atomic Force Microscopy as Nanorobot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ning Xi, Carmen Kar Man Fung, Ruiguo Yang, King Wai Chiu Lai, Donna H. Wang, Kristina Seiffert-Sinha, Animesh A. Sinha, Guangyong Li, and Lianqing Liu Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

439 457 485

505

wwwwwwwwwwwwww

Contributors Toshio Ando  •  Department of Physics, Kanazawa University, Kanazawa, Japan; CREST, JST, Tokyo, Japan Gregory W. Auner  •  College of Engineering, Wayne State University, Detroit, MI, USA Evren U. Azeloglu  •  Department of Pharmacology and Systems Therapeutics, Mount Sinai School of Medicine, New York, NY, USA Régis Barattin  •  CEA-Grenoble, Grenoble, France Nelson P. Barrera  •  Department of Physiology, Pontificia Universidad Católica de Chile, Santiago, Chile Martin Benoit  •  Lehrstuhl für Angewandte Physik, LMU, Sektion Physik, München, Germany Paulo M. Bisch  •  Instituto de Biofísica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro, Ilha do Fundão, Rio de Janeiro, Brazil Dieter Blaas  •  Department of Medical Biochemistry, Max F. Perutz Laboratories, Vienna Biocenter, Medical University of Vienna, Vienna, Austria M. Brent Boyd  •  Department of Mechanical Engineering and Mechanics, Drexel University, Philadelphia, PA, USA Pier Carlo Braga  •  Department of Pharmacology, School of Medicine, University of Milan, Milan, Italy Claudio Canale  •  Nanophysics Unit, Italian Institute of Technology, Genoa, Italy Lili Cao  •  State Key Laboratory of Bioelectronics, School of Biological Science and Medical Engineering, Southeast University, Nanjing, China Ting Cao  •  College of Engineering, Wayne State University, Detroit, MI, USA Stewart M. Carnally  •  Department of Pharmacology, University of Cambridge, Cambridge, UK Calvin Chu  •  Miller School of Medicine, University of Miami, Miami, FL, USA Kevin D. Costa  •  Cardiovascular Research Center, Mount Sinai School of Medicine, New York, NY, USA Ibis Cruz  •  Department of Physiology, Laboratory of Andrology, University of Los Andes, Merida, Venezuela Wanderley de Souza  •  Instituto de Biofísica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro, Ilha do Fundão, Rio de Janeiro, Brazil; Diretoria de Programas, Instituto Nacional de Metrologia, Normalização e Qualidade Industrial – INMETRO, Rio Comprido, Rio de Janeiro, Brazil Andreas Ebner  •  Institute for Biophysics, University of Linz, Linz, Austria J. Michael Edwardson  •  Department of Pharmacology, University of Cambridge, Cambridge, UK

xi

xii

Contributors

Andreas Engel  •  Maurice E. Müller Institute for Structural Biology, Biozentrum, University of Basel, Basel, Switzerland; Department of Pharmacology, Case Western Reserve University, Cleveland, OH, USA Clemens M. Franz  •  DFG-Center for Functional Nanostructures, Karlsruhe Institute of Technology, Karlsruhe, Germany Carmen Kar Man Fung  •  Department of Electrical and Computer Engineering, Michigan State University, East Lansing, MI, USA Maria Gaczynska  •  Department of Molecular Medicine, Institute of Biotechnology, University of Texas Health Science Center at San Antonio, San Antonio, TX, USA Alessandra Gliozzi  •  Department of Physics, University of Genoa, Genoa, Italy Massimo Grattarola  •  Dipartimento di Ingegneria Biofisica ed Elettronica, University of Genoa, Genoa, Italy Hermann Gruber  •  Institute for Biophysics, University of Linz, Linz, Austria Peter Hinterdorfer  •  Institute for Biophysics, University of Linz, Linz, Austria Osamu Hoshi  •  Division of Microscopic Anatomy and Bio-Imaging, Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan Narahari V. Joshi  •  Department of Physiology, University of Los Andes, Merida, Venezuela Miklós S.Z. Kellermayer  •  Department of Biophysics and Radiation Biology, Semmelweis University, Budapest, Hungary Ferry Kienberger  •  Institute for Biophysics, University of Linz, Linz, Austria; Agilent Technologies Austria GmbH, Linz, Austria Yurii G. Kuznetsov  •  Department of Molecular Biology and Biochemistry, University of California, Irvine, CA, USA King Wai Chiu Lai  •  Department of Electrical and Computer Engineering, Michigan State University, East Lansing, MI, USA Bradley E. Layton  •  Applied Computing and Electronics, The University of Montana College of Technology, Missoula, MT, USA Guangyong Li  •  Department of Electrical and Computer Engineering, Michigan State University, East Lansing, MI, USA Xuemei Liang  •  College of Engineering, Wayne State University, Detroit, MI, USA Lianqing Liu  •  Department of Electrical and Computer Engineering, Michigan State University, East Lansing, MI, USA Zuhong Lu  •  State Key Laboratory of Bioelectronics, School of Biological Science and Medical Engineering, Southeast University, Nanjing, China Alexander McPherson  •  Department of Molecular Biology and Biochemistry, University of California, Irvine, CA, USA Gerald A. Meininger  •  Department of Medical Pharmacology and Physiology, Dalton Cardiovascular Research Center, University of Missouri-Columbia, Columbia, MO, USA

Contributors

xiii

Kildare Miranda  •  Instituto de Biofísica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro, Ilha do Fundão, Rio de Janeiro, Brazil; Diretoria de Programas, Instituto Nacional de Metrologia, Normalização e Qualidade Industrial – INMETRO, Rio Comprido, Rio de Janeiro, Brazil Vincent T. Moy  •  Miller School of Medicine, University of Miami, Miami, FL, USA Daniel J. Muller  •  Department of Biosystems Science and Engineering, ETH Zurich, Basel, Switzerland Isabel Neundlinger  •  Institute for Biophysics, University of Linz, Linz, Austria K.Y. Simon Ng  •  College of Engineering, Wayne State University, Detroit, MI, USA Pawel A. Osmulski  •  Department of Molecular Medicine, Institute of Biotechnology, University of Texas Health Science Center at San Antonio, San Antonio, TX, USA Jesus A. Osuna  •  Department of Physiology, Laboratory of Andrology, University of Los Andes, Merida, Venezuela Christian Rankl  •  Institute for Biophysics, University of Linz, Linz, Austria; Agilent Technologies Austria GmbH, Linz, Austria Annalisa Relini  •  Department of Physics, University of Genoa, Genoa, Italy Davide Ricci  •  Robotics, Brain and Cognitive Sciences Department, Italian Institute of Technology, Genoa, Italy; Department of Biophysical Electronic Engineering, University of Genoa, Genoa, Italy Félix Rico  •  Centre de Recherche, Institut Curie, UMR168-CNRS, Paris, France Gustavo M. Rocha  •  Instituto de Biofísica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro, Ilha do Fundão, Rio de Janeiro, Brazil; Diretoria de Programas, Instituto Nacional de Metrologia, Normalização e Qualidade Industrial – INMETRO, Rio Comprido, Rio de Janeiro, Brazil Steven O. Salley  •  College of Engineering, Wayne State University, Detroit, MI, USA Sergio Santos  •  School of Physics and Astronomy, University of Leeds, Leeds, UK Hermann Schillers  •  Institut fur Physiologie II, University Munster, Munster, Germany Kristina Seiffert-Sinha  •  Department of Electrical and Computer Engineering, Michigan State University, East Lansing, MI, USA Christine Selhuber-Unkel  •  Institute for Materials Science, University of Kiel, Kiel, Germany Animesh A. Sinha  •  Department of Electrical and Computer Engineering, Michigan State University, East Lansing, MI, USA Zhe Sun  •  Dalton Cardiovascular Research Center, University of Missouri-Columbia, Columbia, MO, USA Haiying Tang  •  College of Engineering, Wayne State University, Detroit, MI, USA

xiv

Contributors

Mariateresa Tedesco  •  Dipartimento di Ingegneria Biofisica ed Elettronica, University of Genoa, Genoa, Italy Neil H. Thomson  •  School of Physics and Astronomy, University of Leeds, Leeds, UK Bruno Torre  •  Italian Institute of Technology, Genoa, Italy Takayuki Uchihashi  •  Department of Physics, Kanazawa University, Kanazawa, Japan; CREST, JST, Tokyo, Japan Tatsuo Ushiki  •  Division of Microscopic Anatomy and Bio-Imaging, Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan Normand Voyer  •  Département de chimie, Université Laval, Quebec, QC, Canada Anfeng Wang  •  College of Engineering, Wayne State University, Detroit, MI, USA Donna H. Wang  •  Department of Electrical and Computer Engineering, Michigan State University, East Lansing, MI, USA Gilberto Weissmuller  •  Instituto de Biofísica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro, Ilha do Fundão, Rio de Janeiro, Brazil Linda Wildling  •  Institute for Biophysics, University of Linz, Linz, Austria Ning Xi  •  Department of Electrical and Computer Engineering, Michigan State University, East Lansing, MI, USA Zhongdang Xiao  •  State Key Laboratory of Bioelectronics, School of Biological Science and Medical Engineering, Southeast University, Nanjing, China Ruiguo Yang  •  Department of Electrical and Computer Engineering, Michigan State University, East Lansing, MI, USA Dan Zhu  •  State Key Laboratory of Bioelectronics, School of Biological Science and Medical Engineering, Southeast University, Nanjing, China

Part I The Basics of Atomic Force Microscopy

wwwwwwwwwwwwww

Chapter 1 How the Atomic Force Microscope Works? Bruno Torre, Davide Ricci, and Pier Carlo Braga Abstract This chapter aims at giving a quick but precise introduction of the atomic force microscope from the working principle point of view. It is intended to provide a useful starting point to those who first approach the instrument giving a general sketch of the working principles and technical implementations as well as last improvements. Subheading 1 is introductory: it gives an overview of what the instrument does and why it has been developed. Subheading 2 is focused on measurement ranges and on the comparison with scanning electron microscope (SEM) and transmission electron microscope (TEM) which have similar ranges and resolutions but different sample interactions and applications. Subheading 3 gives an overview of the working principles and the most diffused technical implementations on which most of the commercial microscopes rely, as we think it gives the useful base knowledge to understand possible applications, instrument capabilities, and results. In particular, technical improvements taking place over the past few years are highlighted. Despite of the simple and not very technical approach, it has a key importance in understanding concepts at the base of Chapter 3, which is, on the other side, useful for beginners and experienced users as well. Subheading 4 compares different instrument architectures and can, therefore, be useful for those who are going to choose an instrument having clear final applications. Latest solutions are once more highlighted. Subheading 5 gives an overview and some suggestions to start working, both in air and in liquid. Following the general philosophy of the book, it follows more an “how to do” concept than a general theoretical approach. Subheading 6 contains the future developments of the techniques. Key words: Introduction to AFM, AFM working principles, AFM basics

1. Introduction Microscopes have always been one of the essential instruments for research in the biomedical field. The capability of optical ­microscopes to magnify and resolve details well below 1 mm has soon reached its intrinsic physical limit due to the well-known “diffraction limit”: when radiation hits obstacles of size comparable with its wavelength (visible light: 380–750 nm), diffraction

Pier Carlo Braga and Davide Ricci (eds.), Atomic Force Microscopy in Biomedical Research: Methods and Protocols, Methods in Molecular Biology, vol. 736, DOI 10.1007/978-1-61779-105-5_1, © Springer Science+Business Media, LLC 2011

3

4

Torre, Ricci, and Braga

and interference became important, and smaller details cannot be distinguished. Historically, two solutions have been found to image samples with few nanometers resolution or better: the first one is to shorten radiation wavelength, using ultraviolet, X-ray, or electronbased microscopes, to push the diffraction limit from hundreds to few nanometers scale. Radiation-based microscopes (such as the light microscope and the electron microscope) have become trustworthy companions in the laboratory and have contributed greatly to our scientific knowledge. However, short-wavelength radiation can induce sample damaging because high-energy interaction can be involved; moreover, measurements often require special sample preparation or controlled (vacuum) conditions, that can often be incompatible with physiological environment or in vivo measurements. A second strategy relies on a completely different system: a very sharp tip is set in (weak) interaction with the sample and rastered on it while interaction is measured and controlled. In this way, the tip tracks surface morphology while its XYZ position is registered by the electronics to compose a 3D map of the sample surface. Since the interaction can be controlled and limited to very low values, this kind of imaging is usually nondestructive. Depending on the type of interaction measured, scanning probe microscopes (SPMs) take different names, such as scanning tunneling (STM; current between tip and samples), scanning near field optical (SNOM; optical coupling), atomic force (AFM; force between last part of the tip and sample) (1), etc. Historically, AFM has been invented after STM to allow measurements on insulating samples: very soon it was clear that it could archive nanometer resolution working in different environments – air, liquid, or vacuum – regardless of conductive or optical properties of the sample. Moreover, it measures (and controls) tip–sample interaction forces and, therefore, it allows to probe (nano) mechanical properties of the specimen applying pressure or pulling the sample. Hereafter, we refer mainly to AFM for its wide range of applicability in biological field. After using it for the first time, three things can be noticed: ●●

●●

●●

Despite of a rather high-sounding name, imaging with the AFM can be quite simple: no special sample preparation is required, and images with unexpected resolution can be obtained on the very first time; Images are real 3D ones: height is measured with even higher resolution, and subnanometer steps are commonly resolved; questions on surface corrugation and feature height can be easily answered. At first use, imaging appears quite slow: one image can take some minutes to acquire and video rate measurements are not possible with commonly used instruments: this is a

How the Atomic Force Microscope Works?

5

c­ ommon feature for all the mechanical scanning techniques. Nevertheless, a few minutes (without preparation needed) time interval is compatible with most experiments even for biological applications. After some experience, one learns that in some cases it is ­ ossible to push resolution to the “atomic” level (2–4) and p that  images do contain details not observable with any other instrument. A noteworthy feature is that imaging is only one of the experiments that can be performed with the instrument: the tip can be pushed on the sample, pulled out, used to make scratches, functionalized to bind to specific chemical groups, electrically connected to detect currents or potentials, and used to induce catalytic reactions or for lithography purpose. The number of experiments that can be redesigned on the nanometer scale seems to be limited just by applicant imagination: this capability gave rise to a new definition for AFM applications, “lab on tip.” These are, anyway, advanced techniques and are not described in this work: readers interested in the topic are referred to specialized literature.

2. Performance Range of AFM AFM images show significant information about surface features with unprecedented clarity. The AFM can perform nondestructive examinations on any sufficiently “rigid” surface either in air or in liquid, regardless if the specimen is insulating, conductive, transparent, or opaque. Modern instruments can be endowed with temperature control stages and closed chamber for environmental control; some of them are especially designed to be coupled with an optical microscope for simultaneous imaging through advanced optical techniques so that a huge variety of complementary information can be archived. The field of view can vary from the atomic and molecular scale up to sizes larger than 100 mm so that data can be coupled with other information obtained with lower resolution – and wider field of view – techniques. The AFM can also examine rough surfaces with (sub)nanometer resolution on the vertical range up to more than 10 mm; large samples can be fitted directly in the microscope without cutting. With stand-alone instruments, any area on flat or nearly flat specimens can be investigated. Compared with the SEM, AFM provides topographic contrast of surface features with quantitative height information. Moreover, as the sample need not be electrically conductive, no metallic coating of the sample is required. Hence, no dehydration of the sample is necessary as with SEM, and samples may be imaged in their hydrated state. This eliminates the shrinkage of biofilm associated with SEM imaging, yielding a nondestructive

6

Torre, Ricci, and Braga

technique. The resolution of AFM is higher than that of environmental SEM, where hydrated images can also be obtained and extracellular polymeric substances may not be imaged. Compared with transmission electron microscopes, where the electron beam gives a planar projection of the sample by flowing through it, AFM images give information on 3D properties of the surface: in this sense, these two techniques can be regarded as the most complementary ones, since one (TEM) provides contrast on inner structures of the sample, but it is intrinsically 2D, while the other one (AFM) gives real 3D images with similar resolution, but it can only access to the exposed surface. Finally, it can be commented that with respect to TEM, no expensive and destructive (cross-sectioning) sample preparation is needed. Moreover, image contrast is quantitative and can be expressed in nanometer units by default and this is a pretty unique characteristic, allowing direct comparison between different samples. In the following subheadings, we give a brief outline of how the AFM works followed by a description of the parts that can be added to the basic instrument. Our overview has no pretense of completeness but aims at simplicity. For a more thorough description of the physical principles involved in the operation of these instruments, we refer you to the specialized literature.

3. The Microscope In Fig. 1, a schematic diagram of the AFM working principle is shown (1, 5). In principle, AFM can remind one of those old style record players, but it incorporates a number of technical solutions that allow to detect atomic-scale corrugation: very sharp tips at the end of flexible cantilevers and a sensitive deflection sensing system capable of controlling with high accuracy the tip–sample relative position are used. A basic configuration is made up as follows: ●●

●●

●●

●●

A 3D positioning system, called scanner, to adjust tip–sample relative position: if the tip is attached to the scanner, the configuration is called scanning probe; otherwise (as marked with 1 in Fig. 1a), if the tip is fixed and the sample is moved it is called scanning sample. A sample holder where the specimen can be placed in a stable configuration (Fig. 1b). A sharp tip at the end of a flexible cantilever (marker 3 in Fig. 1a and b). A deflection detecting system: in Fig. 1a and c (marker 4), the widely used optical beam deflection (OBD) configuration is

How the Atomic Force Microscope Works?

7

Fig. 1. Schematic diagram of a scanned-sample AFM, based on five quadrant piezo scanner configuration (see below). In the case of scanned probe, it is the tip that is scanned instead of the sample. (a) The piezoelectric scanner (1) is the (nano)positioning element allowing movement: it works by applying opposite voltages to ±X and ±Y sectors to move the sample in X and Y directions, respectively; an additional Z sector moves the sample in the vertical direction. The sample (2) is positioned on the scanner; (3) cantilever; (4) optical beam deflection system (OBD) to detect tip displacement; (5) position-sensitive photodetector (PSD) and preamplifier; (6) electronics. (b) A magnification of tip and sample. (c) A detail of the OBD system.

shown; in this configuration, a tiny tip displacement is detected by a laser beam, amplifying the deflection of the cantilever holding the tip. Laser light, reflected from the rear of the cantilever, is centered on a (usually four vector) photodiode by means of mirrors placed on the optical path. This method allows good signal amplification and it is of rather simple use ; therefore, it is employed on almost all commercial instruments. ●●

●●

Some signal conditioning and preamplifying stage: in case of OBD system, signals from sector A, B, C, and D are used to calculate overall power SUM = A + B + C + D, normal deflection N = (A + B − C − D)/SUM, and lateral deflection as L = (A + C − B − D)/SUM. A digital control system to control tip–sample position on the basis of collected signals.

The following sections contain further details on single ­components and on the working principle.

8

Torre, Ricci, and Braga

3.1. The Scanner

High-resolution (nanometer) positioning can be performed using piezoelectric ceramic materials. These materials undergo a reversible deformation when an external (high) bias voltage is applied across two opposite faces of it: in a first approximation, such deformation can be considered to depend linearly on the applied voltage. A widely used scanner configuration relies on piezoelectric tubes made up by four or five sectors (Fig. 2). A differential bias (with respect to the inner part of the tube, grounded) applied to opposite electrodes induces a bending of the tube in one of the two main directions, while common mode voltage induces a contraction or elongation in the vertical direction. The same happens for the other two electrodes so that differential signals can be used for X and Y movement and common bias for Z movement and four electrodes are sufficient to provide a complete 3D positioning. Anyway, for technical reasons, it is preferable to decouple the Z movement from the XY one, by adding a fifth dedicated electrode (see Figs. 1a and 2) so that a common voltage to side electrodes is no more needed. As shown in Fig. 2, this configuration gives an undesired parabolic component to the motion, therefore this type of scanners are usually endowed with embedded positioning sensors that allow distortion compensation and linearization. This kind of distortion (commonly referred as bow) is mostly relevant for high scan ranges (above some micrometers) and becomes less important for smaller regions, that is to say in case of high resolution: for this reason, some instruments allow to operate also in open-loop mode (i.e., without sensor compensation) for high-resolution imaging, to further reduce electrical noise of readout circuitry. Modern microscopes use a slightly different configuration: single linear piezoelectric elements are embedded in a metallic frame machined by electroerosion to be easily deformable in

Fig. 2. Working principle of a five quadrant piezo tube: right image shows how deflection occurs upon differential ­biasing of two opposite sectors, here −X and +X; the Z sector is visible just below the sample holder on the upper part of the tube. Figure is not in scale and deflection is intentionally exaggerated to highlight the effect.

How the Atomic Force Microscope Works?

9

­ redefined directions (flexure system); for each axis, a ­deformation p occurs easily in the parallel direction to the piezoelectric strain so that the three directions are efficiently decoupled on the three axes. The frames usually incorporate low noise, often capacitive or inductive positioning sensors, and therefore are good candidates for metrology purposes. In some configurations, one of the axes (vertical one) is also mechanically and physically decoupled from the other two. A few words can be spent on the topic of positioning sensors to detect displacements on the nanometer scale. Neglecting for the moment interferometric solutions, that are often used for metrological standards but are not very easy to be integrated, three different accurate positioning sensors are commercially available: ●●

●●

●●

Strain gauge (resistive) sensors can be easily integrated even on piezoelectric tubes by simply gluing them: upon deformation, sensors change their electrical resistance that can be directly read by the electronics. Anyway, since resistors are intrinsically thermal noise generators, this kind of sensors is noisy and resolution is usually limited to a few nanometers. Capacitive sensors: basically they are made up by a capacitor with one plate coupled with the moving part and one fixed to a standing position; a change in the relative position of the two plates implies a change in the relative capacitance that is electrically detected. This type of sensors has very low noise and commonly allows sensitivities of the order of tens of nanometers or better. Integration of these sensors in a scanner is more difficult than for the strain-based ones, and parallelism between faces is often an issue, so they are more often found on flexure scanners than on piezo tubes motors. Inductive or eddy current sensors: recently, some commercially available microscopes have successfully employed inductive sensors for embedded position detection. These have reported resolutions of few tens of nanometer.

Readers interested on this topic can find further details in ref.  6. For our purpose, it is sufficient to keep in mind that in some lower cost microscope, where this compensation is not implemented, a postprocessing software correction can always be performed. 3.2. Tip and Cantilever

The tip, which is mounted at the end of a small cantilever, is the heart of the instrument because it is brought in closest contact with the sample and gives rise to images though its force interactions with the surface. When the first AFM was made, a very small diamond fragment was carefully glued to one end of a tiny piece of gold foil. Today, the tip–cantilever assembly typically is

10

Torre, Ricci, and Braga

Fig. 3. The essential parameters in a tip are the radius of curvature (R ) and the aspect ratio (ratio of H to W ).

f­abricated from silicon or silicon nitride and, using technology similar to that applied to integrated circuit fabrication, allows a good uniformity of characteristics and reproducibility of results (7, 8). The essential parameters to consider are the sharpness of the apex, measured by the radius of curvature (spherical approximation), and the aspect ratio of the whole tip (Fig. 3). Nowadays, a variety of cantilevers are commercially available: in addition to standard pyramid tips, usually 3  mm tall with approximately 30-nm apex radius, also tetragonal, high aspect ratio and conical tips can be found. The tip can end with diamond-like carbon spikes, carbon nanotubes, or whiskers for low curvature radius, and they can also be further machined by means of focused electron beam (FEB) or focused ion beam (FIB) to obtain even higher aspect ratios. Commercial tips commonly end with curvature radius smaller than 10 nm, but ultrasharp tips with R  70%).

68

Santos and Thomson

This is due to the effects of water layers partially or totally ­covering the molecules and screening the surface van der Waals forces that would otherwise be used as feedback to track the topography. Thus, in order to scan in the non-contact regime with small free amplitudes, it is advisable to image at lower relative humidity (e.g., RH  70%); potassium carbonate is soluble in water. In fact, similar studies carried out with the surface force apparatus show that some of the surface precipitates are water soluble and that the contamination layer can be removed by immersing the mica surface in water (25, 26). Unfortunately, as previously mentioned, at high relative humidity, the water layers have a screening effect on the intermolecular surface forces. This screening has a direct effect on the apparent molecular height, giving lower values in the Attractive regime at high RH, particularly if small free amplitudes (A  15–20 nm) are used. It is important to stress that this is only a tendency, since results might vary significantly with even small differences in tip radii, cantilever spring constant, drive frequency, free amplitude, Asp, and the thickness of the water layers. This is due to the nature of the screening effect of the water and its relationship to the tip–sample proximity and the tip radius. In the attractive force regime, and in particular, when there is no tip–sample intermittent contact, the dynamics of the cantilever are solely controlled by the longrange interactions (e.g., van der Waals interactions). When there is sufficient water around, the van der Waals forces due to the water molecules screen the van der Waals forces due to the surface. Thus, in order to “sense” the true solid surface, the tip must get as close as possible to the surface and ideally through the water layer. Since many parameters contribute to the final tip–sample separation in a non-linear fashion, the user must experimentally find the optimum separation in each situation. An example for dsDNA is shown in Fig.  2. Although dsDNA fragments have been used in this example,

70

Santos and Thomson

Fig. 2. Effects of relative humidity (RH) on the molecular width and height of dsDNA fragments on mica by AM AFM in the non-contact mode of operation. Height (z-piezo) images taken in the Low state. First, (a) and (b) together are an example for which the background impurities at low RH are of the order of 0.15–0.2 nm in apparent height. For the same molecules and cantilever-tip system, an increase in RH reduced the background impurities but the screening effect of the water molecules induces some loss in apparent height (b). A second sample and AFM probe have been used to produce the scans in (c) and (d) where the effects of the impurities are more dramatic at low RH (c). However, the same clearing effect of the background is observed at the high RH (d). Molecular average heights and widths are given on the bottomleft part for each scan in nm as H and W, respectively. In (a) and (c) the heights are given as two-part values and these correspond to the height of the molecules relative to the salt layers (lighter part of the background) and the true mica surface (darker part of the background). At high RH, these effects are minimised, so a single average value is given.

the effects are equally valid for antibody imaging or other single biomolecules. Figure  2 shows two different samples, one in (a and b), and another in (c and d). The four images have been obtained in the non-contact mode of operation (Low state) and in the attractive force regime with A = 3 nm and Asp/A = 0.9 driving at resonance. The differences in molecular and background contrast are due mainly to operating at different RH, namely, 30% in (a) and (c) and above 70% in (b) and (d). The apparent average height and width of the molecules are given on the bottom-left corner of each scan as H and W, respectively. The common outcome of these experiments as the humidity increases is twofold; first, the average apparent height of the DNA tends to decrease and second, the background becomes clearer as surface contamination is cleared and/or solubilises. In Fig. 2, two values of the height are given at low relative humidity, these are the

High Resolution Imaging of Immunoglobulin G Antibodies

71

heights of the molecules relative to the salt layer (lighter part of the ­background) and relative to the mica surface. The ­discrepancies in apparent height and width for a given RH can be attributed to slight differences in spring constant, tip radius, average tip–sample separation, the formation of molecular structures on the surface of the mica, and sample preparation and degeneration due to variations in RH. To sum up, the two examples in Fig. 2 below represent the two most common results when imaging biomolecules on a mica surface with small amplitudes (A  0.95. After engaging, the normalised Asp should be slightly reduced to a minimum of 0.9–0.85 in order to try and reach the repulsive regime. If the repulsive regime is not reached, then the tip should be disengaged and the free amplitude should be slightly increased (say 0.5 nm) and/or the drive frequency should be set slightly lower (Df~ −0.01%). The same procedure should then be followed until the repulsive regime is reached stably. The main aim is to reach the repulsive regime stably with a free amplitude as small as possible and with an Asp as high as possible. This is a direct consequence of the tip–sample forces ever increasing with free amplitude and decreasing Asp. Note that a region of bistability could be reached before stably reaching the Repulsive regime. As the set-point is decreased, constant switching between the two cantilever oscillation states can occur; an example of this is shown in Fig. 5. From top to bottom, the Attractive regime is first reached stably by setting the Asp high (Asp/A > 0.985). Then an area of bistability follows

High Resolution Imaging of Immunoglobulin G Antibodies

75

Fig. 6. (a) AFM height data showing sub-domain resolution on the Fab domains, within the tri-nodular structure. (b) Ribbon diagram of an IgG antibody structure rendered from X-ray coordinates from the protein data bank. The X-ray structure is rendered at ~2.3 times larger magnification than the AFM image. (Adapted from Fig. 5, with permission from Elsevier (13)).

as shown in the middle section of the image. Finally, further reducing the Asp to 0.956 allows stably reaching the Repulsive regime while increasing resolution. In particular, it is observed that the tri-nodular structure of the antibody IgG is clearly resolved. Care should be taken, however, since abruptly increasing the free amplitude or reducing the Asp might cause permanent damage to the sample or, in extreme cases, even the tip (12). In the best case scenario to date, IgG has been imaged with a resolution down to 25 kDa (13), that is, the two pairs of immunoglobulin folds that make up one fragment have been resolved (Fig. 6). 6. Using amplitude and phase distance (APD) curves to find optimal operational parameters to image in the Repulsive regime. Another method to try and reach the repulsive regime with a free amplitude as low as possible and an Asp as high as possible consists of systematically acquiring APD curves prior to scanning. In APD curves, the cantilever-sample separation (or z-piezo distance) is decreased and increased at a single point on the sample, while monitoring the amplitude reduction and the phase shift. First, a free amplitude and a drive frequency are set, then the z-distance is decreased through extension of the z-piezo and the amplitude goes down because of damping. It is not advisable to completely damp the amplitude while performing APD curves, since it might cause tip deformation (27). Usually, a trigger is set which limits the damping to a percentage of the free amplitude. Typically, the trigger is set so that only up to 70 or 80% of the cantilever free amplitude are damped. Once the amplitude reaches that minimum value, the z-distance is increased which is typically termed retraction. A method to establish the optimum free

76

Santos and Thomson

amplitude and set-point to image the antibodies is as follows; first, the drive frequency is set near resonance, typically at or above to avoid sudden jumps to the Repulsive regime. Then, a relatively low free amplitude is set, for example 5 nm. An APD curve is subsequently acquired by triggering the amplitude reduction to no more than 80%. If no switch to the repulsive regime is observed after several curves, the tip has to be disengaged and the free amplitude set slightly higher. Shifting the drive frequency slightly towards lower frequencies might also help reach the repulsive regime with a lower free amplitude. As soon as switching from the attractive to the repulsive regime is observed, the critical value of free amplitude (Ac) to reach the repulsive regime has been reached and scans can be performed with that value. It is advisable, however, to slightly increase the free amplitude, since scanning with a free amplitude of just Ac might cause constant switching (see the middle section of Fig. 5) or might require an Asp low enough to cause permanent damage to the molecules, for example Asp/A |20  mV|, VDACs are in a state that has a reduced conductance. Electron microscopy has shown that these pores form lattices whose unit cells comprise six VDACs when native mitochondrial outer membranes are treated by phospholipase A2 (33). NMR and x-ray analyses have produced several atomic structures of the single channel (34). However, only recently it has been possible to reveal these channels directly in the native outer membranes of mitochondria from potato (35) and from yeast (36). In some membrane domains, VDACs were found to be packed at high density like bacterial outer membrane porins (36), whereas in other domains, VDACs were loosely packed, exhibiting single pores and oligomeric clusters comprising two, three, four, and six channels (35) (Fig. 4). The strength of FM-AFM compared to that of contact mode AFM is demonstrated by the images of the VDAC in native potato mitochondrial membranes (35). Whereas in contact mode small oligomers were hard to observe and single VDAC channels were not found, monomers, dimers, trimers, tetramers, and hexamers

Fig. 4. AFM topographs of the voltage-dependent anion channel in outer mitochondrial membranes (OMMO). (a) Topography of OMM patches. Two different types of surface are evident: (1) the OMM surface and (2 ) the mica surface. (b) VDAC proteins arranged in hexagons are visualized by contact mode AFM. The selected VDAC hexagons marked by broken squares are displayed in the gallery at the bottom. Scale bar represents 75 nm and the frame size of the magnified particles in the gallery is 18 nm. (c) The surface structure and organization of single VDAC proteins in the native OMM are revealed by frequency modulation AFM. Various oligomeric states of the VDAC are displayed in the gallery at the bottom: monomers (marked 1 ), dimers (2 ), trimers (3 ), tetramers (4 ) and hexamers (6 ). The scale bar represents 75 nm and the frame size of the boxes in the gallery is 21 nm.

162

Engel

were unambiguously identified in FM-AFM. This observation suggests that FM-AFM induces smaller lateral forces, thereby making the observation of single channels embedded in the bilayer possible.

4. Conclusion and Perspectives Since its invention more than two decades ago the AFM has become an important tool for structural biologists. It is the only instrument that allows surfaces of cells, supramolecular assemblies, and single molecules to be imaged in the native aqueous environment at nanometer-scale resolution. In addition, it makes manipulation of such structures at this scale possible. Recent developments of cantilevers, deflections sensors, imaging modes, and fast-scan systems demonstrate the potential possibilities to improve the AFM. Such progress will enhance the applicability of AFM to a wider range of biological questions, and will allow data to be acquired more efficiently than hitherto possible. AFM images of native membranes at submolecular resolution have provided a wealth of novel insights, and it is likely that this particular application of AFM will yield further important results. Measurements of forces between cells, within supramolecular aggregates, or forces dictating the fold of proteins can now be executed with great efficiency, allowing large datasets to be acquired, delivering quantitative information that were hitherto not accessible. As instrumentation development progresses, sample preparation methods have to be improved as well. There is room for improving the immobilization of native biological membranes to make them accessible to high-resolution imaging and manipulation with the AFM. The rapid progress over the past few years promises the AFM to deliver substantial new information about the structure, dynamics, and function of diverse native biological membranes.

5. Notes 5.1. Preparing the Sample Holders

Glue should be completely hardened and be uniformly distributed between mica and Teflon (or support), and be devoid of air bubbles. In AFMs that displace the supporting surface, air bubbles between mica and support can cause vibration or drift on the nanometer scale. Before mounting the sample support in the AFM, clean all contact surfaces using propanol and/or ethanol. Even small particles between the support and AFM can cause vibrations and drift.

Imaging and Interrogating Native Membrane Proteins Using the AFM

163

5.2. Damping of Vibrations

For high-resolution AFM imaging, an acoustic and vibration-isolated setup of the microscope is crucial. Isolate AFM from sources that may cause electronic and mechanic noise and thermal drift. Noise may be detected by vibration detectors or by scanning the mica surface in buffer solution at minimal forces of 50 pN or less. Sources of electronic noises may be found by switching off the devices individually. The AFM should be placed on an actively or passively damped table. For acoustic isolation, a glass bell may be used.

5.3. Choosing Suitable AFM Cantilevers

For achieving the best possible results, the cantilever properties must be adapted to the experiment. For high-resolution contact mode imaging AFM, cantilevers should be soft (0.1 N/m) and exhibit resonance frequencies in buffer solution, which allow tracking the surface features at the scanning speed applied (see (37)). For oscillating mode imaging, the cantilevers can be up to 30 times stiffer since the amplitude changes of the oscillating cantilever can be detected with sufficient accuracy to sense even very subtle force differences, which is required to prevent the deformation of membrane proteins (12). For specific applications, even stiffer cantilevers may be needed, but the sensitivity of the deflection detection will impose limits. In all cases, the cantilever stylus should have a nominal radius of less than 10 nm. SMFS requires soft cantilevers having a high resonance frequency. Otherwise, small forces may not be detected and the maximum sampling rate of the cantilever limits capturing fast unfolding events.

5.4. Sample Preparation

For AFM experiments, membranes are usually adsorbed to a chemically inert hydrophilic and flat solid support by properly adjusting pH and ionic strength. Such solid-supported membranes have allowed, and will still allow, important insights to be gained into the structure–function relationship of native membrane proteins. Solid supports that have proven suitable for highresolution imaging of membrane proteins include mica (28), HOPG (9, 38, 39), molybdenum disulfide (38), template-stripped gold (38, 40), and template-stripped platinum (38). Templatestripped metal surfaces appear to be particularly useful for combined topographical and electronic measurements (38). Detailed, step-by-step protocols for preparing biological membranes and for AFM imaging have been provided (7). Although high-resolution imaging is possible exclusively on solid-supported membranes, certain question may not be addressed by this preparation method. For example, membrane proteins in membranes directly attached to the support often exhibit impaired mobility (41–43), because the gap between membrane and support is only 0.5–2 nm. Moreover, adsorption forces may influence the conformation of membrane proteins, and it is known that lipids of a solid-supported lipid bilayer can show different structural features than the lipids of a vesicle or

164

Engel

a freestanding lipid bilayer (42). Various schemes have been proposed to circumvent this problem by using spacers that warrant a larger gap, or polymer cushions (44). Freestanding bacterial S-layers spanned over small wells have been imaged at high resolution in the AFM (45), representing an ideal situation. Nevertheless, there is room for further progress in sample preparation strategies to study structure and function of native membrane protein assemblies by AFM. 5.5. Imaging Membrane Channels at High Resolution Using the AFM

Tip, cantilever, deflection detector, piezo elements, fluid cell, and the electronic control system dictate the performance of an AFM. Since the early days, the tip was subject to attempts making it sharper, more reproducible, and more robust. However, even the best tip will inevitably change upon interaction with the sample. Therefore, it is up to the experience and skill of the operator to judge the performance of the tip to collect data when it is devoid of contaminants. Whereas the tip apex dominates the lateral resolution of an AFM, the vertical detection limit is ultimately given by thermal fluctuations of the cantilever, whose properties have been analyzed in great detail (for a recent summary see (37)). The AFM can be operated in several imaging modes that seek to minimize variations in the tip–sample interaction. In contact mode, the vertical deflection signal is the servo loop input for maintaining a constant deflection by a vertical displacement of sample or cantilever. Both the servo in- and output can produce AFM images. The deflection signal (i.e., the small deviation from a set cantilever deflection that the servo attempts to maintain) is mainly used to visualize topographical features in images with large height differences, such as overview images. Because it is a differential signal, edges are enhanced, allowing membrane patches to be easily identified. Lateral forces acting between the tip and the sample in contact mode often leads to a displacement of the sample and can limit the obtainable resolution. In oscillation type AFM, the cantilever is only intermittently in contact with the sample, thus reducing lateral forces. The servo loop in oscillation mode AFM uses the amplitude, phase, or frequency change for feedback. In FM-AFM a phase-locked loop measures the frequency shift, which is used as input for the servo loop operating the z-piezo. An advantage of FM-AFM is that the frequency shift gives quantitative information on the force acting between the tip and the sample (12, 46). When running the FM-AFM with small amplitudes (£1  nm), a most stable and quantitative operation yielding high-resolution images was achieved (12, 47). While these advances in instrumentation and novel imaging modes indeed make the acquisition of high-resolution images more reproducible, it is the tip–sample interactions that ultimately dictate the quality of topographical information obtained. Ionic strength and pH have a critical influence on the tip–sample interactions, and thus provide a handle to optimize them. Ideally, the

Imaging and Interrogating Native Membrane Proteins Using the AFM

165

Fig. 5. Selecting buffer conditions for high-resolution AFM imaging. (a) Small asperities of the AFM stylus contour the sample topography at high resolution. Interactions between AFM stylus and protein membrane can be divided into long- and short-range interaction forces. Long-range electrostatic repulsion forms a cushion that compensates forces applied to the tip. (b) Force–distance curves recorded reveal electrostatic repulsion most clearly from (1) to (2 ). Increasing ion concentration and valency screens electrostatic interactions, revealing the van der Waals attraction at high ionic strength (lowest curve). Balancing van der Waals attraction and electrostatic repulsion is used to promote adsorption of the sample onto the support, and to minimize the forces between asperities of the AFM stylus and the protein.

force applied to the tip (i.e., the set cantilever deflection) to obtain stable operation in the contact mode should be distributed over a significant surface area, while the tip apex barely touches the protein surface. In low ionic strength buffers, electrostatic forces can have a decay length of 100 nm or more, which is observed by the deflection of the cantilever long before the tip contacts the specimen surface. In contrast, van der Waals attraction has a range of about 1 nm and depends neither on pH nor on ionic strength. Therefore, the buffer can be used to adjust the electrostatic contribution to counteract the van der Waals attraction and to distribute the force applied to the tip, thereby preventing severe tip-induced sample deformations (25) (Fig. 5). The strategy for optimizing the recording buffer has been detailed recently (7).

Acknowledgments The author thanks Daniel J. Müller, Dimitrios Fotiadis, and Bart Hoogenboom for providing beautiful topographs and constructive discussions. This work was supported by the Maurice E. Müller Foundation of Switzerland and by the Swiss National Foundation. The used AFM facility was built with contributions from the Swiss University Conference and JPK-Instruments AG, Berlin, Germany.

166

Engel

References 1. Binnig, G., C. F. Quate, and C. Gerber. 1986. Atomic force microscopy. Phys. Rev. Lett. 56:930–933. 2. Drake, B., C. B. Prater, A. L. Weisenhorn, S. A. Gould, T. R. Albrecht, C. F. Quate, D. S. Cannell, H. G. Hansma, and P. K. Hansma. 1989. Imaging crystals, polymers and processes in water with the atomic force microscope. Science 243:1586–1589. 3. Hoh, J. H., R. Lal, S. A. John, J. P. Revel, and M. F. Arnsdorf. 1991. Atomic force microscopy and dissection of gap junctions. Science 253:1405–1408. 4. Buzhynskyy, N., J. F. Girmens, W. Faigle, and S. Scheuring. 2007. Human cataract lens membrane at subnanometer resolution. J Mol Biol 374:162–169. 5. Czajkowsky, D. M., E. M. Hotze, Z. Shao, and R. K. Tweten. 2004. Vertical collapse of a cytolysin prepore moves its transmembrane beta-hairpins to the membrane. Embo J 23:3206–3215. 6. Kienberger, F., H. Mueller, V. Pastushenko, and P. Hinterdorfer. 2004. Following single antibody binding to purple membranes in real time. EMBO Rep 5:579–583. 7. Müller, D. J., and A. Engel. 2007. Atomic force microscopy and spectroscopy of native membrane proteins. Nat Protoc 2:2191–2197. 8. Müller, D. J., G. M. Hand, A. Engel, and G. E. Sosinsky. 2002. Conformational changes in surface structures of isolated connexin 26 gap junctions. EMBO Journal 21:3598–3607. 9. Müller, D. J., and A. Engel. 1999. Voltage and pH-induced channel closure of porin OmpF visualized by atomic force microscopy. J Mol Biol 285:1347–1351. 10. Engel, A., and H. E. Gaub. 2008. Structure and mechanics of membrane proteins. Annu Rev Biochem 77:127–148. 11. Kedrov, A., H. Janovjak, K. T. Sapra, and D. J. Müller. 2007. Deciphering molecular interactions of native membrane proteins by single-molecule force spectroscopy. Annu Rev Biophys Biomol Struct 36:233–260. 12. Hoogenboom, B. W., H. J. Hug, Y. Pellmont, S. Martin, P. L. T. M. Frederix, D. Fotiadis, and A. Engel. 2006. Quantitative dynamicmode scanning force microscopy in liquid. Appl. Phys. Lett. 88:193109. 13. Nikaido, H., and M. Vaara. 1985. Molecular basis of bacterial outer membrane permeability. Microbiol. Rev. 49:1–32.

14. Schindler, H., and J. P. Rosenbusch. 1978. Matrix protein from Escherichia coli outer membranes forms voltage-controlled channels in lipid bilayers. Proc. Natl. Acad. Sci. USA 75:3751–3755. 15. Brunen, M., and H. Engelhardt. 1993. Asymmetry of orientation and voltage gating of the Acidovorax-delafieldii porin Omp34 in lipid bilayers. Eur. J. Biochem. 212:129–135. 16. Sen, K., J. Hellman, and H. Nikaido. 1988. Porin channels in intact cells of Escherichia coli are not affected by donnan potentials across the outer membrane. J. Biol. Chem. 263:1182–1187. 17. Todt, J. C., W. J. Rocque, and E. J. McGroarty. 1992. Effects of pH on bacterial porin function. Biochem. 31:10471–10478. 18. Delcour, A. H. 1997. Function and modulation of bacterial porins: insights from electrophysiology. FEMS Microbiol Lett 151:115–123. 19. Cowan, S. W., T. Schirmer, G. Rummel, M. Steiert, R. Ghosh, R. A. Pauptit, J. N. Jansonius, and J. P. Rosenbusch. 1992. Crystal structures explain functional properties of two E. coli porins. Nature 358:727–733. 20. Schirmer, T. 1998. General and specific porins from bacterial outer membranes. J. Stuct. Biol. 121:101–109. 21. Schulz, G. 1993. Bacterial porins: structure and function. Curr. Opin. Cell Biol. 5:701–707. 22. Phale, R. S., T. Schirmer, A. Prilipov, K.-L. Lou, A. Hardmeyer, and J. Rosenbusch. 1997. Voltage gating of Escherichia coli porin channels: role of the constriction loop. Proc. Natl. Acad. Sci. USA 94:6741–6745. 23. Schabert, F. A., C. Henn, and A. Engel. 1995. Native Escherichia coli OmpF porin surfaces probed by atomic force microscopy. Science 268:92–94. 24. Fotiadis, D., S. Scheuring, S. A. Müller, A. Engel, and D. J. Müller. 2002. Imaging and manipulation of biological structures with the AFM. Micron 33:385–397. 25. Müller, D. J., D. Fotiadis, S. Scheuring, S. A. Müller, and A. Engel. 1999. Electrostatically balanced subnanometer imaging of biological specimens by atomic force microscopy. Biophys J 76:1101–1111. 26. Müller, D. J., and A. Engel. 1997. The height of biomolecules measured with the atomic force microscope depends on electrostatic interactions. Biophys J 73:1633–1644.

Imaging and Interrogating Native Membrane Proteins Using the AFM 27. Israelachvili, J. 1991. Intermolecular & surface forces. Academic Press Limited, London. 28. Müller, D. J., M. Amrein, and A. Engel. 1997. Adsorption of biological molecules to a solid support for scanning probe microscopy. J Struct Biol 119:172–188. 2 9. Karshikoff, A., V. Spassov, S. W. Cowan, R. Ladenstein, and T. Schirmer. 1994. Electrostatic properties of two porin channels from Escherichia coli. J. Mol. Biol. 240:372–384. 30. Andersen, C., B. Schiffler, A. Charbit, and R. Benz. 2002. PH-induced collapse of the extracellular loops closes Escherichia coli maltoporin and allows the study of asymmetric sugar binding. J Biol Chem 277:41318–41325. 31. Yildiz, O., K. R. Vinothkumar, P. Goswami, and W. Kuhlbrandt. 2006. Structure of the monomeric outer-membrane porin OmpG in the open and closed conformation. Embo J 25:3702–3713. 32. Mari, S. A., S. Koster, C. A. Bippes, O. Yildiz, W. Kuhlbrandt, and D. J. Müller. 2010. pHinduced conformational change of the betabarrel-forming protein OmpG reconstituted into native E. coli lipids. J Mol Biol 396:610–616. 33. Mannella, C. A. 1986. Mitochondrial outer membrane channel (VDAC, porin) twodimensional crystals from Neurospora. Methods Enzymol 125:595–610. 34. Hiller, S., and G. Wagner. 2009. The role of solution NMR in the structure determinations of VDAC-1 and other membrane proteins. Curr Opin Struct Biol 19:396–401. 35. Hoogenboom, B. W., K. Suda, A. Engel, and D. Fotiadis. 2007. The supramolecular assemblies of voltage-dependent anion channels in the native membrane. J Mol Biol 370:246–255. 36. Goncalves, R. P., N. Buzhynskyy, V. Prima, J. N. Sturgis, and S. Scheuring. 2007. Supramolecular assembly of VDAC in native mitochondrial outer membranes. J Mol Biol 369:413–418. 37. Frederix, P. L., P. D. Bosshart, and A. Engel. 2009. Atomic force microscopy of biological membranes. Biophys J 96:329–338.

167

38. Frederix, P. L. T. M., P. D. Bosshart, T. Akiyama, M. Chami, M. R. Gullo, J. J. Blackstock, K. Dooleweerdt, N. F. de Rooij, U. Staufer, and A. Engel. 2008. Conductive supports for combined AFM-SECM on biological membranes. Nanotechnology 19. 39. Scheuring, S., D. J. Müller, P. Ringler, J. B. Heymann, and A. Engel. 1999. Imaging streptavidin 2D crystals on biotinylated lipid monolayers at high resolution with the atomic force microscope. J Microsc 193:28–35. 40. Cisneros, D. A., D. J. Müller, S. M. Daud, and J. H. Lakey. 2006. An approach to prepare membrane proteins for single-molecule imaging. Angew Chem Int Ed Engl 45:3252–3256. 41. Müller, D. J., A. Engel, U. Matthey, T. Meier, P. Dimroth, and K. Suda. 2003. Observing membrane protein diffusion at subnanometer resolution. J Mol Biol 327:925–930. 42. Tanaka, M., and E. Sackmann. 2005. Polymersupported membranes as models of the cell surface. Nature 437:656–663. 43. Wagner, M. L., and L. K. Tamm. 2000. Tethered polymer-supported planar lipid bilayers for reconstitution of integral membrane proteins: silane-polyethyleneglycol-lipid as a cushion and covalent linker. Biophys J 79:1400–1414. 44. Müller, D., and A. Engel. 2008. Strategies to prepare and characterize native membrane proteins and protein membranes by AFM. Curr Opin in Colloids & Interface Science 13:338–350. 4 5. Goncalves, R. P., G. Agnus, P. Sens, C. Houssin, B. Bartenlian, and S. Scheuring. 2006. Two-chamber AFM: probing membrane proteins separating two aqueous compartments. Nat Methods 3:1007–1012. 46. Uchihashi, T., M. J. Higgins, S. Yasuda, S. P. Jarvis, S. Akita, Y. Nakayama, and J. E. Sader. 2004. Quantitative force measurements in liquid using frequency modulation atomic force microscopy. Applied Physics Letters 85:3575–3577. 47. Fukuma, T., K. Kobayashi, K. Matsushige, and H. Yamada. 2005. True atomic resolution in liquid by frequency-modulation atomic force microscopy. Appl. Phys. Lett. 87:034101.

wwwwwwwwwwwwww

Part III Nanoscale Surface Analysis and Cell Imaging

wwwwwwwwwwwwww

Chapter 12 Atomic Force Microscopy Investigation of Viruses Alexander McPherson and Yurii G. Kuznetsov Abstract Atomic force microscopy (AFM) has proven to be a valuable approach to delineate the architectures and detailed structural features of a wide variety of viruses. These have ranged from small plant satellite viruses of only 17 nm to the giant mimivirus of 750 nm diameter, and they have included diverse morphologies such as those represented by HIV, icosahedral particles, vaccinia, and bacteriophages. Because it is a surface technique, it provides images and information that are distinct from those obtained by electron microscopy, and in some cases, at even higher resolution. By enzymatic and chemical dissection of virions, internal structures can be revealed, as well as DNA and RNA. The method is relatively rapid and can be carried out on both fixed and unfixed samples in either air or fluids, including culture media. It is nondestructive and even non-perturbing. It can be applied to individual isolated virus, as well as to infected cells. AFM is still in its early development and holds great promise for further investigation of biological systems at the nanometer scale. Key words: Imaging, Nanoscale, Structure, Infection, Nucleic acids, Icosahedra

1. Introduction A direct imaging technology that promises to have a significant impact on structural biology, and which is, in most ways, complementary to X-ray diffraction and electron microscopy, the classical approaches, is atomic force microscopy (AFM) (1–3). An immediate advantage of AFM is that it is based on relatively simple physical principles, unlike X-ray crystallography, and the instruments are mechanically and electronically rather straightforward, unlike electron microscopy. Unlike both of the other technologies, AFM is fairly inexpensive to institute and apply, even to biological specimens. The acuity and investigative size range of the AFM have proven to be quite remarkable and it is now permitting researchers new access to virus structure and the effects of viruses on organisms. Indeed, it has allowed us to Pier Carlo Braga and Davide Ricci (eds.), Atomic Force Microscopy in Biomedical Research: Methods and Protocols, Methods in Molecular Biology, vol. 736, DOI 10.1007/978-1-61779-105-5_12, © Springer Science+Business Media, LLC 2011

171

172

McPherson and Kuznetsov

visualize the surface features, internal structures, and the nucleic acid cores of many viruses. At the same time, it has proven to be an effective instrument for observing viruses emerging from animal cells, and the perturbations they produce to the cells.

2. Materials 1. Glutaraldehyde for sample fixation was made by diluting a commercial solution to a concentration of 5% w/v with distilled water, and kept in a light-free container. Glutaraldehyde should be freshly prepared at least every 2 weeks and maintained at 4°C. 2. Poly-l-lysine for coating cover slips was prepared by dissolving lyophilized polymer in distilled water to a final concentration of 1 mg/ml. This stock solution, maintained at 4°C, was further diluted to 0.1  mg/ml before application to substrates. 3. Magnesium and nickel chloride for treating substrates were made by dissolving reagent-grade salts in distilled water to concentrations of 50 mM. 4. Substrates for AFM were acid-washed glass or plastic, 1-cm diameter cover slips, which were extensively rinsed with distilled water. Alternately, the substrates were freshly cleaved mica. 5. Virus samples were diluted from their stocks into distilled water when possible, and into phosphate-buffered saline otherwise. Substrates with samples, unless air-dried directly, were rinsed with distilled water before air drying. 6. For scanning in fluids, the fluid cell was filled with either distilled water or phosphate-buffered saline. 7. Any other reagents, such as enzymes, detergents, or reducing agents, were prepared from the highest grade materials available and dissolved in distilled water or phosphate-buffered saline. These were maintained at −20°C in small aliquots and thawed for use as needed.

3. Methods 3.1. AFM Instrument Setup

AFM instruments can be operated in either contact mode, or what is referred to as tapping mode. In contact mode, a probe made of silicon or silicon nitride is placed in near contact with the surface of interest, say the capsid of a virus, and then translated in a systematic raster mode over the surface. The AFM probe is a

Atomic Force Microscopy Investigation of Viruses

173

sharp stylus similar in function to a minute phonograph needle. The tip ideally has a single point, with a very small radius of curvature. The probe is mounted at the end of a short cantilever, typically 100–250 mm in length, which has a low spring constant to minimize the force between the tip and the sample. Scanning is achieved by translating the sample beneath the probe, using a piezoelectric positioned x–y stage, along a continuous sequence of raster lines. As the probe tip passes over the surface, it interacts through “aggregate atomic forces,” which remain somewhat mysterious, with structural features on the surface. Encounters with these substructures cause the probe to be displaced vertically as the tip rides across. Exceedingly small displacements of the tip are amplified by deflection of a laser beam that is reflected from the upper surface of the cantilever, and these deflections are detected and tracked by a split photodiode. Photoelectric circuitry converts the deflections into height information. The resulting scan data, recorded as a digital topographical image, can then be presented in a number of visual formats. Sample perturbation and other problems arising from unfavorable probe–surface interactions have been obviated to a great extent by the development of “tapping” mode instruments (4). With tapping mode, the probe tip is not in continuous contact with the sample surface, but rapidly oscillates up and down as it is scanned over the surface, essentially “tapping” its way and gently sensing the heights of obstacles it encounters. In tapping mode, the vertical position of the sample is continually adjusted by a feedback mechanism to maintain the amplitude of the freely oscillating probe constant (see Note 1). The “tapping mode” approach has proven to be a significant boon in biological investigations as it has allowed the characterization of samples that would otherwise be too soft or too fragile to withstand contact mode examination. Operating with tapping mode in a liquid environment presents some complications due to fluid dynamics, but these are not severe. A constraint that sometimes presents obstacles during analysis in a liquid medium is that the specimen under study must be fixed to, or made to adhere firmly to the substrate surface of the fluid cell, which may be glass, cleaved mica, plastic, or any other hard material (see Note 2). One particular feature of AFM must be borne in mind whenever one is interpreting images. The one- or two-dimensional profile obtained of any object, or surface substructure, is the convolution of the tip shape with that of the feature being scanned. This is illustrated in Fig. 1. An image of an object scanned with a broad, dull tip is not the same as that acquired with a sharper tip. In particular, while the height of the object will be the same regardless of the tip shape (because the maximum vertical deflection of the cantilever tip would be the same), the lateral dimensions will not. A broader tip yields a broader object, and a sharper tip produces the more accurate size (see Note 3). Whereas height information is almost

174

McPherson and Kuznetsov

Fig. 1. Schematic illustration of the convolution of the shape of the AFM tip with the shape of the feature, or particle being scanned. The side of the cantilever tip contacts the object and begins to produce a deflection of the cantilever before the tip apex actually reaches the object. Similarly, the opposite side of the tip is still in contact with the object even after the apex itself has passed. Thus the total deflection implies a virtual lateral dimension for the object greater than its actual dimension. The difference between the virtual and actual dimensions is a function of the width of the cantilever tip. The sharper the tip, the more accurate the observed dimensions, and the greater the resolution attainable.

always trustworthy, lateral measurements are frequently suspect. The reliability of lateral measurements can, however, be increased if some standard having defined spatial features is first scanned and its known spacings or cell dimensions compared with those in the image. Such standards may be etched grids on silicon, or the surfaces of protein crystals (5). Height resolution for all samples is typically better than 1 nm (see Notes 4–6). Specimens, however, are not always best visualized under physiological conditions, particularly when high resolution is desired.

Atomic Force Microscopy Investigation of Viruses

175

Because cantilever tip pressure, even in “tapping mode,” may produce deformation, for example, of a cell membrane, in some cases fixation is the better option. This, as with light microscopy histological procedures, usually relies on glutaraldehyde, paraformaldehyde, or osmium tetroxide fixation, followed by dehydration and imaging in water–alcohol mixtures, or in air. These methods have been developed by microscopists for more than a century to preserve the natural morphology of a sample but still allow high-resolution imaging. While not as ideal as in situ observation, the cells are no longer alive or viruses infective, fine details of their structures can be visualized that would otherwise be obscured by membrane flexion. 3.2. Virus Imaging

The resolution of AFM, in the best of cases, is roughly that of current cryo-EM models (6). It is applied to individual particles and does not yield an average structure over an entire population as do many EM reconstructions. It does not require that a virus have symmetrical or uniform architecture, or even that all particles be the same in structure. Thus, it is equally applicable to small icosahedral viruses such as tomato bushy stunt virus, helical viruses such as tobacco mosaic virus, and completely irregular, complex viruses such as vaccinia or the retroviruses. There is no size restriction. It has been used to analyze small plant viruses such as turnip yellow mosaic virus (TYMV) (7) to massive icosahedral viruses such as PBCV-1, an algal virus (8), to mimivirus (9), the largest virus known. Viruses were first visualized by AFM in their crystalline form (7, 10, 11), as illustrated by Fig. 2, rather than as single isolated particles, in investigations of the growth of crystals of satellite tobacco mosaic virus (STMV) and TYMV (12–14). Because they were immobilized on the surfaces of crystals, conditions were suitable for direct imaging of even the small 17–30-nm diameter

Fig. 2. In (a), a low magnification AFM image shows the two-dimensional growth islands that characterize the surfaces of orthorhombic T = 3 Brome Mosaic Virus (BMV) crystals. In (b) is a high magnification AFM image of the surface of the same crystal. As with most virus crystals, vacancies frequently occur in clusters to produce large defects. In (c) is a low magnification AFM image of the reassembled T = 1 particles of BMV in a tetragonal lattice. The scan areas are (a) 2 mm × 2 mm, (b) 542 nm × 542 nm, (c) 272 nm × 272 nm.

176

McPherson and Kuznetsov

virions. Larger icosahedral plant viruses in crystalline form were studied subsequently (15–19). The first AFM studies of noncrystalline viruses were retroviruses on cell surfaces (20–22), again, principally because they were immobilized by their association with cell surfaces. Single particles of larger viruses, and helical viruses, were eventually visualized by AFM, and these included tobacco mosaic virus, cauliflower mosaic virus, Tipula iridescent virus (15, 23), herpes simplex virus (24), vaccinia virus (25, 26), and mimivirus (9). Although virus crystals were investigated using both contact and tapping mode, noncrystalline specimens were imaged exclusively with tapping mode, in both air and buffer. 3.3. Internal Structure Imaging

Because AFM images the surfaces of specimens, it might be thought that AFM would be of little use in visualizing the interior features of viruses or cells. This, however, is not the case. As has been shown in AFM investigations of a number of viruses, it is, in fact, an invaluable tool for deducing the interior architecture of virions, regardless of their external form or size. This is because it is possible to strip away layers of structure systematically by chemical, physical, and enzymatic means (23, 26, 27) and to accompany this process of dissection by AFM visualization. Using the same strategy as that used by conventional anatomists, it has been proven possible to disassemble viral specimens, see what is inside, and ascertain how the components are linked.

3.4. Rapid Shape Classification

A valuable qualitative result that emerges almost immediately from AFM images is what the virus looks like, what is its overall architecture, and how similar are particles to one another. Are they uniformly the same in appearance, or are there a variety of forms? Thus even a cursory investigation may quickly reveal certain general features that allow rapid classification. This is illustrated by the various structural classes of viruses shown in Fig. 3. The virions may be spherical, cylindrical, or filamentous. They may have symmetrically arranged capsomeres or other surface units, fibers, protruding vertices, prolate or icosahedral shapes, unusual morphologies, pleiomorphic character, etc. Tail assemblies may be observed directly, as on phages for example. AFM is, therefore, a useful tool for simply deducing the kind of virus one is dealing with, whether more than one kind of virus is present in a population, and the general level of contamination that may accompany the virus as a consequence – cellular material, degraded virions, and macromolecular impurities of all sorts.

3.5. Quantitative Dimensional Measurements

A fundamental parameter for virus particles is their diameter if they are spherical viruses, or their diameter and length if they are helical. AFM can provide measures of these in both the hydrated and dried states, which also gives an estimate of the degree of shrinkage they undergo as a result of dehydration. Because of the finite tip size, and tip-to-tip variation in radius of curvature, it is

Atomic Force Microscopy Investigation of Viruses

177

Fig. 3. A variety of viruses seen at relatively low magnifications by AFM. In (a) vaccinia virus, in (b) the retrotransposon Ty3, in (c) Tobacco Mosaic Virus, in (d) T4 bacteriophage, in (e) PBCV-1 algal virus, an irridovirus, and in (f) Moloney Mouse Leukemic Virus. The scan areas are (a) 850  nm × 850  nm, (b) 500  nm × 500  nm, (c) 200  nm × 200  nm, (d) 250 nm × 250 nm, (e) 185 nm × 185 nm, and (f) 200 nm × 200 nm.

risky to measure linear dimensions directly by AFM (see Note 3). It is, however, safe to measure the heights of objects above the substrate plane, and the distances between the points of maximum elevation (e.g., capsomere to capsomere) on particles, or centerto-center distances (e.g., particles in a crystal or in a cluster). As has been emphasized already, for spherical and cylindrically symmetric particles, measurements of particle heights above the substrate plane yield reasonably accurate values for their dia­ meters, and individual measurements are usually accompanied by rather modest error, generally of the order of 5% or less. By repeating measurements for a number of particles in the field, and using different scan directions, good statistics can be obtained, and histograms of size distributions compiled. Precision of a few angstroms is possible. Histograms of particle sizes, as illustrated in Fig. 4, are often informative (8, 22) (see Note 7). 3.6. Topography as a Function of Composition and Architecture

The surfaces of virus particles vary topographically as a function of their composition and architectures. Plant viruses, for example, generally exhibit protein capsids with few embellishments, and this is true of many animal viruses and bacteriophages as well. These capsids are generally based on icosahedral architectures,

178

McPherson and Kuznetsov

Fig. 4. The heights above substrate level were measured for about 200 isolated MuLV virions and plotted as a histogram. The spread of sizes is not due to error in the measurements, which is only a few nanometers at most, but represents the real variation in size of particles produced in infection. The very large and very small particles are aberrant virions.

Fig. 5. A phase-contrast AFM image of a herpes simplex virion absorbed onto mica. The large sheet of white material around the capsid is the membrane envelope of the virus that has been partially discarded. The capsid lattice of the virus is clearly evident. The scan area is 770 nm × 770 nm.

and clusters of coat protein subunits, or capsomeres, are symmetrically distributed (28, 29). Many animal viruses, on the contrary, though they may contain an icosahedral capsid in their interior, often have either a lipid membrane over their surface, as does the herpes virus in Fig. 5, a covering of protein clusters, or even a hair-like coating of fibers. These various surfaces are readily apparent by AFM, and can be identified and delineated with a high degree of precision with the aid of some histological procedures, such as osmium tetroxide fixation, or protease treatment.

Atomic Force Microscopy Investigation of Viruses

179

Fig. 6. At higher magnification, the surface features of many viruses emerge. Seen here in (a) is the surface lattice of PBCV-1, a giant algal virus belonging to the irridovirus ­family, in (b) in vitro reassembled particles of the Gag protein from Mason–Pfizer Monkey Virus, in (c) Ty3 retrotransposons, and in (d) the capsid of the giant Mimivirus. The scan areas are (a) 42 nm × 42 nm, (b) 63 nm × 63 nm, (c) 117 × 117 nm, and (d) 200 nm × 200 nm.

Icosahedral capsids, or bullet-shaped or elongated capsids based on that symmetry, can be characterized in terms of the structure of the fundamental capsomere, along with the icosahedral triangulation number, T (29). Some examples are shown in Fig. 6. This will vary from small integral numbers like T = 1 for satellite viruses, or the T = 1 reassembly particles of Brome Mosaic Virus (30) to T = 3 and higher for more conventional, small icosahedral viruses such as poliovirus or TYMV, to very large numbers for complex viruses such as the irridoviruses such as PBCV-1 (T = 169) and mimivirus (T one of nine possibilities lying between 972 and 1,200). In many cases, the exterior shell of a virus may not be icosahedral, but it might possess an inner capsid which is. For example, though membrane covered and of pleiomorphic external shape, herpes simplex virus possesses a nucleic acid containing capsid of icosahedral form T = 16. Mimivirus exhibits a complex outer surface coated with a forest of fibers, but it too contains an

180

McPherson and Kuznetsov

Fig. 7. On the left is a capsid of the Ty3 retrotransposon. If the pattern of five- and sixfold capsomeres are plotted on its surface, as on the right, then it can be deduced that the virion has T = 4 icosahedral symmetry as shown in the center drawing. Other Ty3 virions were shown to have T = 3 and T = 7l icosahedral architectures.

icosahedral core (9). The T number, then, provides much of the information one needs to describe an icosahedral capsid. The triangulation numbers of icosahedral viruses can frequently be deduced from AFM images, as for the retrotransposon Ty3 in Fig. 7. With very large icosahedral capsids, which include PBCV-1 and mimivirus, one determines the two indices h and k (29), which define T (T = h2 + hk + k2). This is done by following a row of hexagonal capsomeres from one pentagonal vertex to the next icosahedral edge, and by simply counting the number of capsomeres along one edge h and the other k (the h and k coordinates of the intersection point on the icosahedral edge) that one needs to traverse (31) (see Note 8). In the cases of Ty3 retrotransposon and of Mason–Pfizer monkey virus (MPMV) in Fig. 8, for example, particles of their truncated Gag proteins were reassembled in vitro and imaged at high magnification by AFM. From the images, individual protein subunits were visible, and this allowed the discrimination of two possible models for the capsomeres (5). A similar analysis was used in the case of the large algal virus PBCV-1 (8). Knowing the diameters of capsomeres is often of considerable importance, even when individual subunits cannot be resolved. In mimivirus, for example, capsomere diameter provided a crucial clue in delineating the capsid architecture and permitting subsequent detailed analysis and reconstruction by cryo-EM (9). Although capsids of native HIV have yet to be visualized by AFM, helical tubes of capsid protein reassembled in  vitro have (5). In these tubes, a hexagonal arrangement of coat proteins could be clearly seen, and this provided support for a capsid model based on modified icosahedral architecture (32). The tubes reassembled from HIV Gag protein should remind us that helical and rod-shaped structures having periodic substructure are also excellent specimens

Atomic Force Microscopy Investigation of Viruses

181

Fig.  8. Virus-like particles (VLP) reassembled in  vitro or in recombinant bacteria often appear to have icosahedral ­symmetry, but upon closer inspection, they do not, the pentameric vertices are replaced by random defects or overall disorder. This is true of the Ty3 Gag particles in (a) and the mutant Gag of MPMV seen in (b). VLPs often take on entirely different shapes than the native capsids, as seen in the tubular forms of reassembled HIV Gag in (c). The scan areas are (a) 250 nm × 250 nm, (b) 125 nm × 125 nm, and (c) 200 nm × 200 nm.

for AFM analysis. These can appear in investigations of intact viruses and even in studies of spherical viruses when their interiors are explored. 3.7. Virus Sample Preparation Techniques 3.7.1. Isolated Particles

With AFM, it is not essential that highly purified virus particles be used as samples (15), although that might be ideal. Because individual particles can be investigated whenever a good specimen is spatially distinct from the surrounding rubble of proteins, cellular debris, and biological detritus, it may still yield excellent images. A problem, however, is that biological debris often adhere to and foul the AFM tip, severely degrading the quality of images. Contaminated tips are one of the most frustrating and annoying accompaniments of biological AFM.

3.7.2. Viruses on the Surface of Host Cells

Viruses on the surfaces of host cells may be visualized as well as free particles, and sometimes with better results because they are better immobilized (8, 20, 22, 33). Moloney mouse leukemia virus (MuLV) emerging from an infected 3T3 cell are clearly delineated in Fig. 9. They may be seen entering cells upon infection, or budding from cells after replication and assembly. This often provides valuable insights into which cells in a population are producing virus, the distribution of virus particles on the surface of the cells (are there preferred sites for budding?), and some details of the budding process itself.

3.7.3. Mutant Viruses: Anomalous Features

Mutant viruses, naturally occurring or produced in the laboratory, can be imaged as well as native virions and virus-like particles (VLP) created in vitro from capsid proteins. In some cases, the phenotype of the mutant can be revealed by observing infected host cells for unique or anomalous features. This was done, as shown in Fig. 10, in a study of MuLV-infected 3T3 cells, where a

182

McPherson and Kuznetsov

Fig. 9. Viruses can also be observed while emerging from, or stall attached to the host cell plasma membrane. In (a) is a low magnification AFM image of a 3T3 cell in culture infected with MuLV. The virus is clearly seen as white spots over the surface of the cell. In (b) is a higher magnification image showing four MuLV particles budding from a host cell in culture. In (c) is a mass of HIV virus bursting from the surface of a human lymphocyte in culture, and in (d), a higher magnification image of four HIV budding from a lymphocyte surface. The scan areas are (a) 10  mm × 10  mm, (b) 2  mm × 2  mm, (c) 5 mm × 5 mm, and (d) 460 nm × 460 nm.

mutant lacked glycosylated Gag protein (20, 34). Prior evidence suggested that such mutants failed in some stage of viral budding. This was confirmed by AFM visualization of infected, virus-producing cells. As seen in Fig.  10, instead of normal, spherical virus emerging from the cell surface, bullet- and ­comet-shaped protrusions were found distributed all over the plasma membrane of the host cells. The comets were viruses that were apparently trying to escape, but were unable to pull away and terminate association with the host cell. From this, it was concluded that the failure of glycosylation produced a defect in late stages of the budding process. Other mutations in virus genomes may produce alterations in external features of virus particles that are readily observable by AFM.

Atomic Force Microscopy Investigation of Viruses

183

Fig. 10. Two mutant forms of Moloney Mouse Leukemic Virus (MuLV) are shown in (a) and (b). In (a), the mutant lacks the ability to make the envelope protein, and the lipid membrane of the virus is observed directly. Its wavy pattern is due to its fluid nature and its motion when subjected to the pressure of the AFM cantilever. In (b) is another mutant which does not properly glycosylate its Gag protein, resulting in an inability to bud from the host cell properly, resulting in the comet-shaped structures on the surface of the host 3T3 cell in culture. The scan areas are (a) 250 nm × 250 nm. In (b) are composites of serial, positionally incremented AFM images representing total scan areas of 2 mm × 8.44 mm and 2 mm × 4.72 mm.

MuLV particles that failed to make an envelope protein (gp120 protein), one of which is seen in Fig. 10, were examined in another study (22). While normal particles are characterized by a coating of protein tufts, about 100–150 in number, mutant particles were “bald” virions lacking any such protein clusters. Instead, only an outer lipid membrane was visible. Some viruses exhibit special external structures, or deviations from their general architectures. For example, MuLV particles generally have a single small bump or a brief protrusion somewhere on their otherwise uniformly crenulated surfaces. These are likely to be a “budding scar” resulting from breaking away from the host cell (22). Other MuLV particles, perhaps defectives, exhibited small sectors on their surfaces where protein was absent and a channel into the interior appeared (20). Other, more prominent features are the surface fibers on the surfaces of mimivirus and the lateral bodies of vaccinia (seen in Fig. 11) (25, 26). PBCV-1 Chlorella Virus, an irridovirus, exhibited a unique pentagonal assembly of proteins at every fivefold vertex of its icosahedral capsid (8), shown in Fig. 12. The assembly had a single protein in the center that could “push in” and “pull out” as demonstrated by the application of AFM tip pressure. Its exact function is speculative. Many bacteriophages have tail assemblies of one sort or

184

McPherson and Kuznetsov

Fig. 11. AFM can capture a variety of specialized structures produced by viruses. In (a) is a fragment of the anchor protein, at the center, which connects a corona of glycosylated surface fibers to the capsid. In (b) is the contractile DNA injection assembly, or tail, of bacteriophage T4 showing helical architecture. In (c) is the “stargate” apparatus found at a unique vertex of Mimivirus which, upon opening, allows escape of the DNA.

Fig.  12. Fivefold vertices on large viruses which have icosahedral capsids often have unique clusters of proteins or unusual structures. In (a) and (b) are fivefold vertices of the large algal virus PBCV-1. In (c) and (d) is seen the stargate apparatus of Mimivirus opening to allow expulsion of the DNA inside the capsid. The scan areas are (a) 96 nm × 96 nm, (b) 100 nm × 100 nm, (c) 500 nm × 500 nm, and (d) 800 nm × 800 nm.

Atomic Force Microscopy Investigation of Viruses

185

another for packaging and injecting their DNA. Mimivirus is, in a sense, similar to these phages and has an assembly, seen in Fig. 12c, d, of presumably similar function at a single, unique, fivefold vertex. This star shaped structure is undoubtedly analogous to the tail assemblies of phages and is, as is evident in the figures, a distinctive feature of the virions. It is a complex structure likely composed of many proteins, and AFM reveals much of that complexity. 3.8. Population Analysis

An important point that deserves particular emphasis is that all of the particles within a population of virus are not absolutely identical, and often there are very significant differences in the detailed features of individual virus particles. This is a point often obscured by the results of X-ray crystallography (35, 36) or cryo-EM reconstructions (37, 38). These techniques rely totally on an assumption of structural conformity and produce models that represent the average in time and space of the individuals that make up the population. AFM, on the contrary, allows the revelation of the eccentricities and unique features of the individuals, and these are instructive. They often define the extremes of what is possible among a large population of viruses having, presumably, the same genome and the same environment for replication and assembly. What we see with AFM is that anomalous and aberrant individuals are not only present, but are also common.

3.9. Internal Structure Determination

One might think that because AFM provides images of the surfaces of objects and does not peer into their interiors, as do X-ray diffraction and electron microscopy, they would be of little value in delineating the interior structure of viruses, the layers beneath the external surface. This is not true, however, as we can apply the same technique that has been used by anatomists for centuries – dissection. With the aid of chemical, enzymatic, and physical tools, we can systematically pare a complex entity, including a virus, down to its core, layer by layer (23). At each stage, AFM may then be used to visualize what remains and what has been removed as well (see Note 9). Among the most useful agents for chemical dissection have been detergents, usually 0.5–2% of some nonionic detergents such as NP40, and reducing agents such as DTT or DTE. The former causes protein structure to unravel gradually and detergents strip away the lipid membrane. The latter reduces disulfide bonds and liberates polypeptides otherwise bonded to one another. Disulfide bond reduction appears to be particularly important in large, complex viruses where such covalent linkages cross-link coat proteins and stabilize capsids (25, 39) (see Note 10). The most effective enzymatic tools have been proteases that degrade polypeptides. These are particularly useful because they have a range of activities and a spectrum of specificities. As a consequence, a whole variety of proteases have been employed, including trypsin, bromelin, proteinase K, subtilisin, and mixtures

186

McPherson and Kuznetsov

of pancreatic proteases. Viruses are usually exposed to the ­proteases for anywhere from 15  min to several hours, or even overnight, at concentrations of 0.5 mg/ml to as high as 5 mg/ml (see Note 11). Physical forces have also been used to disrupt viruses, and often fortuitous perturbations, resulting simply from preparation and handling, have proven to be structurally illuminating. Heat, for example, was used to open TYMV (7) to release its encapsidated RNA, and direct physical pressure was used on mimivirus sandwiched between two layers of atomically smooth cleaved mica, as well. There are also instances where “hammering” of individual particles with the AFM tip has been utilized, taking advantage of the fact that AFM can serve as a tool as well as an imaging device. In carrying out the dissection of a virus, or even in simply visualizing particles spread on a glass, plastic, or mica substrate, it is necessary to ensure that the virus particles adhere firmly to the substrate. Failure to do so allows the particles to move beneath the AFM tip, rendering imaging impossible. Occasionally, altering the charge on the substrate is sufficient (see Notes 12 and 13). Altering charge is, however, frequently insufficient for virions. To fix most viruses to the substrate, as well as a wide variety of other biological entities and materials, an effective procedure is to coat the substrate with poly-l-lysine before depositing the virus. Presumably, salt bridges between the e amino groups of the lysines and the glutamic and aspartic acid carboxyl groups on the particles lock them in place. After such substrate–particle attachment, the substrate can be rinsed with water several times without loss of sample (see Note 14). It is occasionally unnecessary to actually treat viruses with any chemical or biochemical agent to view the interior, as the physical stress of preparation and purification may result in damaged or partially degraded particles. These may expose interior structural features that are otherwise not apparent. Retroviruses, in particular, are physically fragile. Some MuLV, as shown in Fig. 13, when subjected to the shear forces of centrifugation, lose portions of the shell surrounding the capsid. This permits direct visualization of the virus core still embedded within the layers of envelope and matrix protein (22). HIV is another example where even the mildest procedures produce some damaged virions. Although the cores of HIV have not yet been visualized by AFM, likely due to their fragility, the remainder of the virus without the cores has been (22). Some examples can be seen in Fig.  13c, d. Such partially disrobed particles, both MuLV and HIV, provide specimens that can be subjected to quantitative examination and thereby yield the dimensions, the thicknesses of internal structural layers, and they give some clues as to their components as well. The best example of a complete dissection of a complex virus using AFM is that of vaccinia virus, a pox virus of about 300-nm

Atomic Force Microscopy Investigation of Viruses

187

Fig. 13. Virions are sometimes damaged in their preparation, especially fragile viruses such as retroviruses. These occasions often provide images of interior structure that would otherwise not be accessible to the AFM tip. In (a) and (b) are MuLV whose outer shell has been partially sheared away to reveal the nucleic acid containing capsid inside. In (c) and (d) are similarly sheared HIV particles. The interior capsids have been lost, leaving the empty outer shells behind. The scan areas are (a) 200 nm × 200 nm, (b) 200 nm × 200 nm, (c) 500 nm × 500 nm, and (d) 250 nm × 250 nm.

diameter that is delimited by a lipid membrane (25, 26). It ­contains a double-stranded DNA genome bounded by several protein shells. It also has two unusual protein assemblies of still unknown function, known as lateral bodies, associated with its inner core. Vaccinia was sequentially degraded with a 0.5% NP40 nonionic detergent combined with 0.05  M DTT, followed by exposure to this same mixture but containing either trypsin or proteinase K, or to the proteases alone. Six stages in this process are presented in Fig.  14. At the end, the innermost core was breached and the DNA was exposed. 3.10. Imaging of Nucleic Acids of Viruses

The nucleic acids of viruses, some of which are seen in Fig. 15, from a structural standpoint, are of considerable interest, and in particular, how they are condensed and packaged inside capsids and cores. Clearly, packaging is accomplished differently by ­specific families of viruses. It is unlikely, for example, that

188

McPherson and Kuznetsov

Fig. 14. A series of AFM images presenting the dissection of vaccinia virus using a variety of enzymatic and chemical reagents to remove successive layers of structure. In (a) is the intact virus in buffer before any treatment. In (b), the lipid membrane is pronounced as a corona surrounding a virion dried upon the substrate. The lateral body at the center is also more pronounced on the dried virions. In (c), the outer protein shell has been etched away to reveal the inner protein capsid which is perforated and still has the lateral bodies connected to it. In (d) is a mass of capsids having lost both their lateral bodies and their DNA. In (e) are higher magnification AFM images of the lateral bodies that decorate the capsids, and in (f) is a mass of vaccinia DNA released onto the AFM substrate. The scan areas are (a) 400 nm × 400 nm, (b) 350 nm × 350 nm, (c) 350 nm × 350 nm, (d) 2 mm × 2mm, (e) 500 nm × 500 nm, and (f) 2 mm × 2 mm.

bacteriophage and pox virus package their genomic doublestranded DNA the same way. The packing densities of the nucleic acid differ by more than tenfold (26). Also, is it unlikely that large, single-stranded RNA-containing viruses, such as retroviruses, package their genomes the same way as do T = 1 or T = 3 icosahedral viruses (40). Certainly, helical and filamentous viruses use entirely different mechanisms. AFM investigations have been conducted on RNA extracted by phenol from a series of small icosahedral viruses, and from tobacco mosaic virus, the classical rod-shaped, helical virus (41). The spherical viruses included poliovirus, STMV, TYMV, and brome mosaic virus. In this study, the gradual unraveling of the tertiary structure of the RNA, and ultimately the secondary structure as well, could be produced in stages simply by heating. A counter example was provided by the rod-shaped, helical tobacco mosaic virus RNA which appeared initially as a thread, a completely extended molecule lacking any secondary structure. With time, it began forming local secondary structural elements

Atomic Force Microscopy Investigation of Viruses

189

Fig. 15. Expulsions or emissions of nucleic acid by a variety of viruses. In (a), a shattered Ty3 retrotransposon disgorges a mass of nucleic acid. Although it is known that the genome of the retrotransposon has two single strands of RNA as its genome, the nucleic acid seen here has the characteristics of double-stranded DNA. In (b), the DNA core of the algal virus PBCV-1 throws out a splash of double-stranded DNA. In (c) is a mass of DNA released by vaccinia virus upon degradation with proteases. In (d), a shattered virion of STMV spreads its single-stranded RNA genome of 1,058 nucleotides around itself. In (e), virions of TYMV are losing their single-stranded RNA genomes after loss of a capsomere. In (f), damaged T4 bacteriophages release their DNA on the AFM substrate. The scan areas are (a) and (b) 1 mm × 1 mm, (c) 5 mm × 5 mm, (d) 200 nm × 200 nm, (e) 500 nm × 500 nm, and (f) 2 mm × 2 mm.

and eventually condensed into forms similar to those seen for the RNA from the icosahedral viruses (41). One conclusion of the study was that the single strands of RNA spontaneously condensed as linear arrangements of stem-loop substructures following synthesis, the condensed RNA bound coat protein to it, and the two cooperatively coalesced into the completed particle. In studies such as these, AFM proved itself as an able technique for directly visualizing nucleic acid structure, demonstrating its fluidity, and suggesting the mechanisms by which it is encapsidated. DNA and RNA appear quite different in AFM images, and this is evident in Fig.  16, which presents both kinds of nucleic acids. The former looks like strands and coils of stiff rope lacking any higher levels of structure, while the latter appears as complicated, linear sequences of self-involved secondary structure. Sometimes, however, the distinction is not entirely clear and further evidence may be needed to show whether a filament, strand, or complex is DNA or RNA.

190

McPherson and Kuznetsov

Fig. 16. Viral nucleic acid can be readily visualized by AFM after expulsion from the capsid. In (a) is a DNA plasmid used in recombinant DNA research to transform bacteria. In (b) is a tangle of double-stranded DNA from vaccinia virus. In (c) is the genomic single-stranded RNA of STMV, still condensed due to secondary structure, and in (d), RNA from poliovirus escaping from a condensed core obtained by phenol extraction of the virus. The scan areas are (a) 587 nm × 587 nm, (b) 500 nm × 500 nm, (c) 220 nm × 220 nm, and (d) 500 nm × 500 nm.

A method was devised for additional identification based on exposure of the nucleic acid to high concentrations of bovine RNase A (42). RNA, naturally, was hydrolyzed to small pieces by RNase A and left only fragments on the substrate which corresponded to protected stem-loops. DNA, on the contrary, became coated with the protein and the resulting strands exhibited diameters two to three times that of naked double-stranded DNA. Thus it is possible to practice a kind of crude histology with AFM. A second example of histological AFM is the immunolabeling of viruses with antibodies specific for certain proteins. Although individual IgG are not clearly identifiable by AFM when bound to a virion, IgG conjugated with gold particles generally are. In a sense, these are used in the same way as they are used in transmission electron microscopy immunolabeling, except that instead of visualizing points of high electron density, one images with AFM objects having the size and shape of the immuno-gold particles.

Atomic Force Microscopy Investigation of Viruses

191

Fig. 17. Virus can be labeled with gold particles conjugated with antibody against specific viral proteins. These can be visualized using AFM as seen here. In (a) and (b) are two different MuLV virions that have been exposed to antibody/gold particles directed against the envelope protein. Arrows indicate the gold particles. The scan areas are (a) 500 nm × 500 nm and (b) 600 nm × 600 nm.

Using gold–IgG conjugate particles against the envelope ­protein, as shown in Fig. 17, it was possible to show that protein tufts on the surfaces of MuLV were indeed envelope proteins (20, 22). The major problem with IgG–gold conjugates at this point is that their physical size limits the resolution of the method. Conjugated gold particles can bind only as close as their diameters allow. The answer to the question, what can AFM visualize that is of value to the structural biology of viruses, is that it can visualize virtually every part of a virus, and to resolutions that approach, and in some cases surpass, those of electron microscopy. At this time, lipid membranes have been identified, both RNA and DNA have been visualized, and large protein assemblies resolved. The capsids of icosahedral viruses, and the icosahedral capsids of nonicosahedral viruses have been seen at high resolution, in some cases sufficiently high to deduce the arrangement of coat protein units in the capsomeres, or to determine the triangulation number T. In addition, viruses have been recorded budding from infected cells and suffering the consequences of a variety of stresses. Mutant viruses have been examined and phenotypes described. Unusual structural features have appeared, and very importantly, the unexpectedly great amount of structural nonconformity within populations of virus particles has been well documented. It has, furthermore, been shown that the structures of viruses observed by AFM are entirely consistent with models derived by X-ray crystallography and cryo-EM (16). Although there are currently no examples, there is certainly no reason why structural information derived from X-ray crystallography and/or electron microscopy cannot be combined with AFM images, just as it has been for the latter two technologies.

192

McPherson and Kuznetsov

4. Notes 1. Tapping mode minimizes contact between the probe tip and the sample surface and greatly reduces lateral forces. An even more sensitive means of scanning in tapping mode is called phase modulation scanning. Here, phase changes are introduced into the tip oscillations due not only to height differences, but also to changes in the nature of the aggregate atomic interactions, caused in turn by variations in the physical or chemical properties of the sample surface. This approach has been shown to be useful for imaging very thin and delicate materials such as biological membranes (24). 2. To achieve this, it may be necessary to treat the substrate with various reagents, such as poly-l-lysine, to induce better adhesion of samples. If this condition is not met, then the specimen will move due to interaction with the probe, and no useful information will be gathered. 3. Because one does not, in general, know the tip shape one is working with at the time, the image cannot be easily deconvoluted to provide true lateral dimensions. 4. On large soft samples, such as living animal cells (33), lateral resolution may be more limited by the motion and deformation of the cell surface in response to tip pressure rather than tip structure. 5. Because visualization can be carried out in a fluid environment, specimens may suffer no dehydration as is generally the case with electron microscopy, and they usually require no fixing or staining. 6. Indeed, specimens can be observed over long periods, so long as they stay relatively unchanged and immobilized during a single frame interval. For the most part, even living cells seem oblivious to the presence of the probe tip (33). 7. If the distribution is a simple Gaussian, then it can be presumed that particles of only one general morphology, or icosahedra of only one triangulation number are present, but that their diameters vary to some degree about the mean, perhaps due to physiological state or degree of maturation. On the contrary, if a more complex distribution is observed, one having multiple peaks and shoulders, then particles of separate classes may be present. 8. While the T number describes the overall distribution of capsomeres on the surface of an icosahedral capsid, the more complete description of a virus structure would require the distribution of protein units in the individual capsomeres to be defined, and ultimately coordinates of the atoms

Atomic Force Microscopy Investigation of Viruses

193

comprising the virus coat proteins. The last can only be obtained by X-ray crystallography, but the distribution of subunits within capsomeres can sometimes be determined or deduced by AFM analysis. 9. This approach is particularly effective with large, complex viruses such as vaccinia virus (25, 26) or mimivirus (9). With these large assemblies, ordered and disordered protein shells, lipid membranes, and the nucleic acid within can be revealed and analyzed. By deconstruction, the architecture of particles is revealed, and, at the same time, the kinds of biochemical interactions that maintain each level of structure are delineated as well. 10. In some cases, nonionic detergents are insufficient to disrupt structure and more vigorous ionic detergents such as SDS must be used. There is difficulty with SDS, however. It tends to have an all-or-none effect, so that upon reaching a concentration sufficient to disrupt viruses, it completely degrades them uncontrollably. SDS can also produce artifacts due to drying on the substrate. 11. The proteases must be washed from the virions with buffer or water before imaging as they otherwise produce a dense, irregular background that makes imaging problematic, and they foul the cantilever tip. 12. Mica is negatively charged on its surface, but exposure to nickel or magnesium salt such as MgCl2 coats it with divalent ions and leaves it positively charged. 13. Some viruses or macromolecules, such as nucleic acids, may be firmly held by a positive surface if they are repelled by a negative surface, or vice versa. 14. The only serious disadvantage of coating with poly-l-lysine is that it produces a rather rough and irregular background. As a consequence, molecular objects, such as lipid membranes or nucleic acids, which rise only about a nanometer or two above the substrate plane, become difficult to identify and visualize. The method is excellent, however, for imaging cells and intact or partially degraded virions. References 1. Binning, G., and Quate, C.F. (1986). Atomic force microscope. Phys. Rev. Lett. 56: 930–933. 2. Bustamante, C., and Keller, D. (1995). Scanning force microscopy in biology. Phys. Today 48: 32–38. 3. Allen, S., Davies, M.C., Roberts, C.J., Tendler, S.J.B., and Williams, P.M. (1997). Atomic

force microscopy in analytical biotechnology. Trends in Biotech. 15: 101–105. 4. Hansma, P.K. (1994). Tapping mode atomic force microscopy in liquids. Appl. Phys. Lett. 64: 1738–1740. 5. Kuznetsov, Y.G., Ulbrich, P., Haubova, S., Ruml, T., and McPherson, A. (2007).

194

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

McPherson and Kuznetsov Atomic force microscopy investigation of Mason-Pfizer monkey virus and human immunodeficiency virus type 1 reassembled particles. Virology 360: 434–446. Baker, T.S., Olson, N.H., and Fuller, S.D. (1999). Adding the third dimension to virus life cycles: Three-dimensional reconstruction of icosahedral viruses from cryo-electron micrographs. Microbiology and Molecular Biology Reviews 63: 862–922. Kuznetsov, Y.G., and McPherson, A. (2006). Atomic force microscopy investigation of Turnip Yellow Mosaic Virus capsid disruption and RNA extrusion. Virology 352: 329–337. Kuznetsov, Y.G., Gurnon, J.R., Van Etten, J.L., and McPherson, A. (2005). Atomic force microscopy investigation of a chlorella virus, PBCV-1. J Struct Biol 149: 256–263. Xiao, C., Kuznetsov, Y.G., Sun, S., Hafenstein, S.L., Kostyuchenko, V.A., Chipman, P.R., Suzan-Monti, M., Raoult, D., McPherson, A., and Rossmann, M.G. (2009). Structural studies of the giant mimivirus. PLoS Biology 7: 958–966. Malkin, A.J., Land, T.A., Kuznetsov, Y.G., McPherson, A., and DeYoreo, J.J. (1995). Investigation of virus crystal growth mechanisms by in situ atomic force microscopy. Physical Review Letters 75: 2778–2781. Land, T.A., Malkin, A.J., Kuznetsov, Y.G., McPherson, A., and DeYoreo, J.J. (1996). Mechanisms of protein and virus crystal growth: An atomic force microscopy study of Canavalin and STMV crystallization. J. Cryst. Growth 166: 893–899. Malkin, A.J., Kuznetsov, Y.G., Lucas, R.W., and McPherson, A. (1999). Surface processes in the crystallization of turnip yellow mosaic virus visualized by atomic force microscopy. J Struct Biol 127: 35–43. Malkin, A.J., Kuznetsov, Y.G., and McPherson, A. (2001). Viral Capsomere Structure, Surface processes and growth kinetics in the crystallization of macromolecular crystals visualized by in situ atomic force microscopy. J. Cryst. Growth 232: 173–183. Malkin, A.J., Plomp, M., and McPherson, A. (2002). Application of atomic force microscopy to studies of surface processes in virus crystallization and structural biology. Acta crystallographica 58: 1617–1621. Kuznetsov, Y.G., Malkin, A.J., Lucas, R.W., Plomp, M., and McPherson, A. (2001). Imaging of viruses by atomic force microscopy. The Journal of general virology 82: 2025–2034. McPherson, A., Malkin, A.J., Kuznetsov, Y.G., and Plomp, M. (2001). Atomic force microscopy

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

applications in macromolecular crystallography. Acta Cryst. D57: 1053–1060. McPherson, A., Kuznetsov, Y.G., Malkin, A.J., and Plomp, M. (2004). Macromolecular crystal growth investigations using atomic force microscopy. Journal of synchrotron radiation 11: 21–23. Makino, D.L., Larson, S.B., and McPherson, A. (2005). Preliminary analysis of crystals of panicum mosaic virus (PMV) by X-ray diffraction and atomic force microscopy. Acta crystallographica 61: 173–179. Kuznetsov, Y.G., Daijogo, S., Zhou, J., Semler, B.L., and McPherson, A. (2005b). Atomic force microscopy analysis of icosahedral virus RNA. J Mol Biol 347: 41–52. Kuznetsov, Y.G., Datta, S., Kothari, N.H., Greenwood, A., Fan, H., and McPherson, A. (2002). Atomic force microscopy investigation of fibroblasts infected with wild-type and mutant murine leukemia virus (MuLV). Biophys J 83: 3665–3674. Kuznetsov, Y.G., Victoria, J.G., Robinson, W.E., Jr., and McPherson, A. (2003). Atomic force microscopy investigation of human immunodeficiency virus (HIV) and HIV-infected lymphocytes. J Virol 77: 11896–11909. Kuznetsov, Y.G., Low, A., Fan, H., and McPherson, A. (2004). Atomic force microscopy investigation of wild-type Moloney murine leukemia virus particles and virus particles lacking the envelope protein. Virology 323: 189–196. Malkin, A.J., Plomp, M., Leighton, T.J., McPherson, A., and Wheeler, K.E. (2006). Unraveling the Architecture and Structural Dynamics of Pathogens by High Resolution In vitro Atomic Force Microscopy. In Microscopy and Microanalysis, pp. 32–85. Cambridge Univ. Press, Cambridge. Plomp, M., Rice, M.K., Wagner, E.K., McPherson, A., and Malkin, A.J. (2002). Rapid visualization at high resolution of pathogens by atomic force microscopy: structural studies of herpes simplex virus-1. Am J Pathol 160: 1959–1966. Malkin, A.J., McPherson, A., and Gershon, P.D. (2003). Structure of intracellular mature vaccinia virus visualized by in situ AFM. J. Viol. 77: 6332–6340. Kuznetsov, Y., Gershon, P.D., and McPherson, A. (2008). Atomic Force Microscopy Investigation of Vaccinia Virus Structure. Journal of Virology 85: 7551–7566. Kaper, J.M. (1975). The Chemical Basis of Virus Structure, Dissociation and Reassembly. In The Frontiers of Biology Series.

Atomic Force Microscopy Investigation of Viruses

28. 29.

30.

31.

32.

33.

34.

(eds. A. Neuberger, and E.L. Tatum). NorthHolland, Amsterdam. Horne, R.W., and Wildy, P. (1961). Symmetry in virus architecture. Virology 15: 348–373. Caspar, D.L.D., and Klug, A. (1962). Physical principles in the construction of regular viruses. In Cold Spring Harbor Symposium on Quant. Biol., pp. 1–24. Lucas, R.W., Kuznetsov, Y.G., Larson, S.B., and McPherson, A. (2001). Crystallization of Brome Mosaic Virus (BMV) and T = 1 Brome Mosaic Virus Particles following a structural transition. Virology 286: 290–303 Reddy, V.S., Natarajan, P., Okerberg, B., Li, K., Damodaran, K.V., Morton, R.T., Brooks, C.L.I., and Johnson, J.E. (2001). Virus Particle Explorer (VIPER), a website for virus capsid structures and their computational analyses. J. Virol. 75: 11943–11947. Ganser, B.K., Li, S., Klishko, V.Y., Finch, J.T., and Sundquist, W.I. (1999). Assembly and analysis of conical models for the HIV-1 core. Science 283: 80–82. Kuznetsov, Y.G., Malkin, A.J., and McPherson, A. (1997). Atomic force microscopy studies of living cells: visualization of motility, division, aggregation, transformation, and apoptosis. J Struct Biol 120: 180–191. Low, A., Datta, S., Kuznetsov, Y., Jahid, S., Kothari, N., McPherson, A., and Fan, H. (2007). Mutation in the glycosylated gag protein of murine leukemia virus results in reduced in vivo infectivity and a novel defect

35. 36.

37.

38. 39.

40. 41.

42.

195

in viral budding or release. J Virol 81: 3685–3692. Rossmann, M.G. (2000). Fitting atomic models into electron-microscopy maps. Acta crystallographica 56: 1341–1349. Rossmann, M.G., Bernal, R., and Pletnev, S.V. (2001). Combining electron microscopic with x-ray crystallographic structures. J Struct Biol 136: 190–200. Adrian, M., Dubochet, J., Lepault, J., and McDowall, A.W. (1984). Cryoelectron microscopy of viruses. Nature (London) 308: 32–36. Carragher, B., Potter, C.S., and Sigworth, F.J. (2007). Software tools for macromolecular microscopy. J. Struct. Biol. 157: 1–288. Wikoff, W.R., Liljas, L., Duda, R.L., Tsuruta, H., Hendrix, R.W., and Johnson, J.E. (2000). Topologically linked protein rings in the bacteriophage HK97 capsid. Science 289: 2129–2133. McPherson, A. (2005). Micelle formation and crystallization as paradigms for virus assembly. Bioessays 27: 447–458. Kuznetsov, Y.G., Zhang, M., Menees, T.M., McPherson, A., and Sandmeyer, S. (2005c). Investigation by atomic force microscopy of the structure of Ty3 retrotransposon particles. J Virol 79: 8032–8045. Kuznetsov, Y.G., and McPherson, A. (2006b). Identification of DNA and RNA from retroviruses using ribonuclease A. Scanning 28: 278–281.

wwwwwwwwwwwwww

Chapter 13 Determination of the Kinetic On- and Off-Rate of Single Virus–Cell Interactions Christian Rankl, Linda Wildling, Isabel Neundlinger, Ferry Kienberger, Hermann Gruber, Dieter Blaas, and Peter Hinterdorfer Abstract Human rhinoviruses are the causative agents of the common cold. The serotypes belonging to the minor receptor group attach to members of the low-density lipoprotein receptor family and enter the host cell via receptor-mediated endocytosis. Receptor binding, the very first step in infection, was characterized by force spectroscopy measurements at the single molecule level. We demonstrate how kinetic on- and offrate constants can be derived from such experiments carried out with the atomic force microscope. Key words: Force spectroscopy, Molecular recognition, Viral receptor, Very low density lipoprotein receptor, Minor receptor group rhinovirus, PicoRNAvirus, Atomic force microscope

1. Introduction The first step of viral infection is the attachment of the virus to a cell receptor. This interaction initiates a cascade of processes resulting in the delivery of the viral genome into the host cell. The route can be receptor-mediated endocytosis (e.g., many naked and enveloped viruses (1)) or direct translocation of nucleoproteins via fusion of viral envelope and plasma membrane (2, 3). Once inside the cell, proteins encoded by the viral genome are translated and, ultimately, viral progeny is produced. Thus, understanding cell attachment of viruses is important. Here, we show how single molecule force spectroscopy is used to gain insight into this elementary process. The atomic force microscope was invented in the late 1980s (4). Its ability to measure in liquids and at room temperature made it a tool to investigate biological interactions at the single Pier Carlo Braga and Davide Ricci (eds.), Atomic Force Microscopy in Biomedical Research: Methods and Protocols, Methods in Molecular Biology, vol. 736, DOI 10.1007/978-1-61779-105-5_13, © Springer Science+Business Media, LLC 2011

197

198

Rankl et al.

molecule level, such as binding of receptor–ligand pairs (5), antibody–antigen interaction (6), and the interaction of complementary DNA strands (7). Recently, this technique was applied to investigate the interaction of single viruses with live cells (8). In order to perform force measurements of interacting components (molecules or molecular assemblies), one binding partner is immobilized on the cantilever tip and the other one on a sample surface. Approaching the tip until it touches the surface eventually leads to interaction between the molecules on the tip and on the surface. Subsequent retraction loads the bond with an increasing force due to cantilever bending. At a certain force, the bond ruptures and the bent cantilever jumps off the surface into its equilibrium position. This unbinding is caused by thermal fluctuations rather than by purely mechanical dissociation. If the thermal lifetime of the interaction is short compared to the time it takes to apply the bending force of the cantilever, no unbinding event will be observed. Faster loading results in measurable unbinding forces. Therefore, unbinding forces depend on the applied loading rate and on the details of the functional relation of bond lifetime and applied force. A detailed theoretical consideration yields the direct link between such single molecule pulling experiments and bulk experiments, where thermodynamic data are experimentally acquired. The single molecule approach gives access to the full spectrum of a property instead of an averaged value gained by bulk experiments. This chapter describes the protocol to measure kinetic onand off-rates of a human rhinovirus type 2 (HRV2) interacting with the human low-density lipoprotein receptor, overexpressed in mouse fibroblast cells. In addition, we show how the number of interacting receptors can be determined. We used a PicoPlus AFM (Agilent Technologies, Chandler, AZ, USA) with commercially available cantilevers (Microlevers, Veeco). The virus was bound to the tip using a homemade amine reactive cross-linker.

2. Materials 2.1. AFM

1. 5500 AFM (Agilent Technologies, Chandler, AZ, USA). 2. National Instruments data acquisition card (PCI-6121). 3. Cantilever: MSCT Microlever (Veeco, Santa Barbara, CA).

2.2. Linking Viruses to the Tip

1. HRV2 produced in suspension culture and purified by differential and sucrose density gradient centrifugation (9). 2. DMSO (Sigma, Taufkirchen, Germany): highest available purity. 3. Ethanolamine hydrochloride (Sigma, Taufkirchen, Germany): highest available purity.

Determination of the Kinetic On- and Off-Rate of Single Virus–Cell Interactions

199

4. Chloroform and ethanol (Fisher Scientific): analytical grade. 5. Triethanolamine (Sigma, Taufkirchen, Germany): highest available purity. 6. 4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid, N-(2hydroxyethyl)piperazine-N ¢-(2-ethanesulfonic acid) (HEPES) (Sigma, Taufkirchen, Germany): highest available purity. 7. NaCl (Sigma, Taufkirchen, Germany): highest available purity. 8. NaOH (Sigma, Taufkirchen, Germany): highest available purity. 9. NaCNBH3 (Sigma, Taufkirchen, Germany): highest available purity. 10. HBS buffer solution: 10 mM HEPES, 150 mM NaCl, pH 7.4. 11. HBS-Ca buffer solution: HBS with 2 mM CaCl2. 2.3. Mouse Fibroblast Cells

1. M4-LDLR cells (10), a SV40 large T-antigen-immortalized mouse fibroblast cell line deficient in endogenous LDLR and LRP (LDLR-related protein) overexpressing human lowdensity lipoprotein receptor (11, 12). 2. DMEM (Invitrogen, Cat. No. 21969-035). 3. l-Glutamin (Invitrogen, Cat. No. 25030-024). 4. Penicillin–streptomycin (Invitrogen, Cat. No. 151140-122). 5. Fetal bovine serum (Invitrogen, Cat. No. 10270-106). 6. Growth medium: DMEM with 1% l-glutamin, 1% penicillin– streptomycin, and 10% fetal bovine serum.

3. Methods 3.1. Immobilizing the Virus to the Tip

Attachment of ligand molecules to the measuring tip of an AFM converts it into a biospecific sensor by which cognate receptor molecules can be detected on a sample surface. Attachment of ligands to AFM tips via PEG chains is done in three steps: (1) amino groups are generated on the tip surface, (2) PEG chains are attached to the amino groups on the tip, and (3) a ligand molecule is coupled to the free-tangling end of PEG: 1. For amino functionalization of the cantilever chips, 3.3  g ethanolamine hydrochloride was dissolved in 6  mL DMSO by heating to 60°C in a glass beaker. 2. When all solid was dissolved, the beaker was removed from the heater and molecular sieve beads (3 Å) were added to the mixture and a small thin glass plate (prewashed in ethanol and dried with nitrogen gas) was placed on top of the beads, taking care to avoid inclusion of air bubbles. 3. The cantilevers were washed three times in chloroform and dried with nitrogen gas.

200

Rankl et al.

4. Cantilevers were immersed in the ethanolamine hydrochloride/ DMSO solution and incubated overnight. 5. The tips were washed with DMSO (3×), ethanol (3×), and dried with nitrogen gas. 6. For linking of aldehyde–PEG–NHS, 3.3 mg of this compound were dissolved in 0.5 mL chloroform and transferred into a small glass reaction chamber. 7. Triethylamine (30  ml) was added and the cantilevers were immersed into the solution at RT for 2 h. 8. After the reaction, the cantilevers were extensively washed in chloroform and dried with nitrogen gas. 9. For covalent binding of viral particles, a 5-mL aliquot of the virus stock solution (HRV2, ~1  mg/mL in HBS, freshly thawed from −25°C) was mixed with 45 ml of HBS. 10. The bottom of a disposable Petri dish was covered with parafilm and the dried cantilevers were moderately pressed onto the parafilm. The cantilevers were arranged in a circular manner so that the ends with the tips pointed to the center of the circle. Then, the 50-mL virus suspension was pipetted into the center of the circle such that all tips were covered with liquid. 11. Immediately, a 2-mL aliquot from a 1  M NaCNBH3 stock solution (freshly prepared by dissolving 32 mg solid NaCNBH3 in 500 mL 10 mM NaOH) was added and the cantilever was incubated for 60 min. 12. Free aldehyde functions on the tips were inactivated by addition of 5 mL 1 M ethanolamine hydrochloride (pre-adjusted to pH 9.6 with NaOH and stored in aliquots at −25°C). 13. After another 10 min reaction time at room temperature, the cantilevers were washed with HBS and stored in HBS-Ca at 4°C for less than 24 h. 3.2. Cell Preparation

1. For AFM measurements, cells were grown on a 25 mm cover slip. Best results were achieved when the cells were nearly confluent (~80%). 2. The cover slip was mounted in the AFM. It was taken care that the cells never dried. 3. Growth medium was exchanged against HBS-Ca buffer solution.

3.3. Calibration of the Optical Detection System of the AFM

Prior to interaction measurements, the optical lever detection (OLS) system was calibrated allowing to correlate photodiode voltage with the deflection of the cantilever. This calibration was done for every cantilever used. The result was a calibration factor

Determination of the Kinetic On- and Off-Rate of Single Virus–Cell Interactions

201

fols, which describes the relationship between physical cantilever deflection and measured photodiode voltages: 1. The AFM was set up according to the manual. 2. A cantilever containing virus immobilized to its tip was inserted into the AFM. It was taken care that the cantilever never dried, as this might denature the virus on the tip. 3. A freshly cleaned glass slide was mounted in the AFM. The liquid cell was filled with HBS buffer. 4. For sensitivity determination approx. 50 force distance cycles with 100 nm length and 1 s sweep time were acquired. In order to obtain most accurate results, the z-piezo was freshly calibrated and the force distance cycles were performed around 0 nm (i.e., start at 50 nm and end at −50 nm z-piezo position). 5. The slope of the contact part was determined in every force curve individually (see Note 1, Fig. 1a). 6. The calibration factor fols represents the negative average of the above determined inverse slopes. The obtained value depends on the cantilever and AFM system, typical values for fols are between 10 and 100 nm/V.

Fig. 1. (a) A force–distance cycle (dash dotted line) on mica was acquired so as to calibrate the optical lever detection system. The tip was aligned away from the surface and continuously approached toward it. As long as the tip does not contact the surface, no deflection is seen (right part of the graph). As long as the tip is in contact with the surface, further approaching results in an upward bending (left part of the graph). A linear fit (solid line) through the contact region yielded a calibration factor fols = 18.3 nm/V. (b) A force–distance cycle showing an unbinding event in the retrace; note the nonlinear stretching indicating specific binding. The rupture force F can be directly calculated from the height of the rupture event. The uncertainty of the rupture force s is determined by the standard deviations left (sl) and right (sr) of the rupture: s 2 = s 2l + s 2r . The slope at rupture keff is needed to determine the loading rate r = keffv.

202

Rankl et al.

3.4. Dynamic Force Spectroscopy Experiments

The interaction force is measured by performing force distance cycles; i.e., the tip is approached to the surface until it touches it and subsequently retracted. During retraction the linker stretches until the bond breaks, which results in a characteristic nonlinear force signal. The dissociation of the complex is mainly thermally driven and of stochastic nature. Therefore, many such force signals must be acquired to obtain statistically relevant distributions of the rupture forces. It is another consequence of the stochastic nature of this process that the rupture force distribution depends on the pulling speed. From this dependence, the kinetic off-rate is extracted: 1. A cell sample was mounted in the AFM and the growth medium was exchanged against HBS-Ca buffer solution. 2. A virus-carrying tip was mounted and placed over a cell making use of the optical access. During approach, it was taken care that the tip did not crash into the cell, so as to avoid unspecific adhesion of proteins to the AFM tip. Such a “dirty” tip can result in false force signals or in no force signals at all. 3. At least 1,000 force–distance cycles of typically 1,000  nm sweep length and 1  s sweep duration time were recorded. Again, the maximum force applied by the cantilever was kept as low as possible to prevent contamination of the tip. 4. Step 3 was repeated using different sweep duration times (0.1, 0.2, 0.5, 2, and 5 s). For very short duration times (0.1 and 0.2 s), at least 2,000 force curves were recorded so as to compensate for the low binding probabilities due to the short contact times. 5. The steps 3 and 4 were repeated using cantilevers of different spring constants (10, 20, and 30 pN/nm). 6. The specificity of the interaction was proven by blocking the interaction. Usually, one binding partner is added at high concentration, leading to a saturation of available binding sites. Here, we choose reversible inactivation. Activity of the LDL receptor depends on Ca2+ ions. In order to block the interaction with the virus, the buffer was exchanged against HBS buffer, containing 2 mM EGTA, which does not contain Ca2+. Again, 1,000 force curves with 1,000 nm sweep length and 1 s sweep duration were acquired. The binding probability was reduced to less than 3% (compared to 15% in HBS-Ca buffer). Next, the buffer was changed back to HBS-Ca, resulting in restoration of the binding probability to a value comparable before Ca2+ removal. 7. Step 6 was repeated for each cantilever used. Only data from cantilevers that showed the above-described reversible blocking behavior were used.

Determination of the Kinetic On- and Off-Rate of Single Virus–Cell Interactions

3.5. Determining the Spring Constant of the Cantilever

203

The cantilever acts as a small spring and, accordingly, the deflection can be converted into a force F acting on the cantilever, using Hooke’s law F = k × z. Here k is the cantilever spring constant and z the deflection of the cantilever. The spring constant can be measured in several ways, either using a calibrated reference (13), the added mass method (14), the thermal noise method (15), or the Sader method (16). Each of these methods has advantages and drawbacks, a good overview can be found in ref. (17). The thermal noise method is most widely used and commonly accepted. It is based on modeling the cantilever as a spring, making use of the equipartition theorem: 1. The washed and dried cantilever was mounted in the AFM with a freshly calibrated scanner (see Note 2). A cleaned mica sheet was used as sample. 2. The free (at least 100 mm away from the surface) cantilever movement was recorded using a data acquisition card. 3. The optical lever sensitivity was determined (see above). 4. Multiplying the recorded cantilever movement with the optical lever sensitivity converts the photodiode output into the deflection (nanometer) of the cantilever. 5. The power spectrum density of the data gained in step 4 was estimated using a fast Fourier transform (see Note 3). The resulting squared amplitudes of the Fourier transform are proportional to the power spectrum. However, there are several different conventions for the normalization of the power spectrum and many opportunities for making it wrong. In order to achieve the time-integral squared amplitude, the proper normalization factor is given by the sampling interval. A detailed discussion can be found in ref. 18. 6. This power spectrum estimate was fitted with a simple harmonic oscillator model:



A = Awhite +

4 A0w 0 2 2 2 æ ww 0 ö w - w0 +ç è Q ÷ø

(

)

2

,



(1)

where Awhite is a white noise floor, A0 is the zero frequency amplitude, w0 is the radial resonance frequency, and Q is the quality factor (Fig. 2).



7. The spring constant of the cantilever using the thermal noise method is finally given by: 4 kb T k =α , (2) A0ω 0 Q where kb is the Boltzmann constant and T the absolute temperature, a = 0.817 for rectangular cantilevers (19) and a = 0.764 for triangular cantilevers (20).

204

Rankl et al.

Fig. 2. The power spectral density (dots) of a freely oscillating cantilever used to calculate the spring constant is shown. It was calculated by applying a Fourier transform to the photo diode output shown in the inset and subsequent proper normalization. The harmonic oscillator fit (solid line) yielded a radial resonance frequency of ~92 krad/s, a quality factor of 25.6 and a zero frequency amplitude of 2.4 nm2/Hz. These parameters were used to estimate the spring constant of the cantilever (23.4 pN/nm).

3.6. Extraction of the Kinetic Off-Rate from Force Spectroscopy Experiments

A simple two-state binding model with a separating energy barrier can be used to describe dynamic force spectroscopy experiments (21, 22). It links dynamic force spectroscopy experiments to the kinetic off-rate constant and to the thermally averaged distance between the binding state and the barrier along the projection of the applied force: 1. A program (see Note 1) was written, which converts the force curves into proper units (pN) by multiplying the photo diode output with fols × k and allowed toggling through the proper converted force curves. Curves showing a nonlinear force signal, terminated by an abrupt return to the baseline were identified as an unbinding event. The program allowed extracting 2 2 2 the height F i and the uncertainty ( s i = s l,i + s r,i ) of a rupture event, and the slope at the time of rupture keff of proper unbinding events (Fig. 1b).



2. The probability density estimate of the rupture forces was estimated using: æ (F - Fi )2 ö 1 n 1 (3) pdf (F ) = å exp ç ÷. 2 n i =1 2ps i2 è 2s i ø 3. The most probable rupture force F * was determined from this pdf.

Determination of the Kinetic On- and Off-Rate of Single Virus–Cell Interactions

205

Fig. 3. The dependence of the applied loading rate on the distribution of rupture forces is shown. Theory predicts an increase of observed rupture forces with higher loading rate, which was indeed found.

4. The loading rate of a dataset was determined to be: r = keff

2 ´ scansize . sweeptime

5. Steps 1–4 were repeated for different pulling speeds resulting in a pdf and a loading rate estimate for each dataset (see Fig. 3). 6. The most probable unbinding force F * was plotted against the logarithm of the loading rate r (Fig. 4). 7. A straight line ( y = ax + b ) with slope a and y axis intersection b was fitted into the above plot. 8. Parameters were extracted according to -

3.7. Measuring the Kinetic On-Rate Constant



b

kT e a , x b = B , c.f. Notes 4 and 5 (21, 22). koff = a a The virus–receptor interaction was approximated with pseudo first-order kinetics. Estimation of the kinetic on-rate constant kon from single molecule unbinding force measurements requires the determination of the interaction time t and the effective concentration ceff: kon = (tc eff ) . -1

The interaction time t was determined from the binding probability at different encounter times. An effective concentration

206

Rankl et al.

Fig. 4. The most probable rupture force as a function of the loading rate is shown. As predicted (21, 22), the unbinding force depends linearly on the logarithm of the loading rate. The straight line shows the linear fit used to calculate the model parameters, koff = 0.6/s and x b = 0.4 nm.

was estimated from the number of binding partners nb (typically 1) within a free volume accessible by the virus tethered to the tip. The free volume was approximated by a sphere, whereby the radius reff of the sphere is the sum of the equilibrium length of the cross-linker (3  nm) plus the diameter of the virus (30  nm). Therefore, the kinetic on-rate constant can be calculated using the following formula (8, 23): kon

4preff3 N A = , 3nbt

(4)

where NA is the Avogadro constant = 6.022 × 1023/mol. 1. Virus tip and cell samples were mounted as described above. 2. The size and duration for force distance cycle were set to 1,000 nm sweep length and 1 s sweep time. 3. First, a dwell time of about 10 ms on the surface was used. As our AFM software does not directly support dwell times on the surface, we set a relative maximum deflection limit of 0.2  V. The dwell time was then determined from the force curve, i.e., the time needed to travel from the contact point to the maximum force limit and back again. 4. At least 1,000 force distance cycles were acquired. 5. Step 4 was repeated using different dwell times (10, 20, 50, 100, 200, 500, 1,000 ms, etc.), c.f. Note 6.

Determination of the Kinetic On- and Off-Rate of Single Virus–Cell Interactions

207

Fig. 5. Binding probability as a function of the contact time. A typical first-order reaction kinetics was found allowing for the determination of the kinetic on-rate constant kon.

6. The binding probability, i.e., the fraction of force distance cycles that showed an unbinding event was determined for every dwell time. 7. The unbinding probability as a function of the contact time was plotted. 8. This graph was fitted using p = A (1 - exp((t - t 0 ) / t )) , where t0 is the time lag, A the maximum observable binding probability, and t the interaction time (Fig. 5). 9. The suchlike gained t was used to determine the kinetic on-rate constant: 4preff 3N A . kon = 3nb t 3.8. Counting the Number of Bound Receptors

HRV2 has 12 receptor-binding sites (i.e., each receptor molecule can wind around each of the 12 vertices of the viral icosahedron simultaneously attaching via up to five symmetry-related sites; (24)). In order to determine how many receptors are bound to the virus, the distribution of rupture forces were further analyzed. Multiple receptor binding stabilizes the interaction and leads to higher rupture forces. If an individual receptor–virus bond breaks, the applied load is distributed over the remaining bonds and reduces the bond lifetime to nanosecond levels (25). Therefore, the remaining bonds rupture very shortly after the first one is broken and the finite bandwidth of the AFM causes the multiple bond breakages to be registered as a single rupture event. A direct consequence of this fact is that the governed rupture force distribution shows multiple force peaks. This distribution was fitted with æ F -m 2 ö 1 ( i ) ÷ , including N A exp ç a sum of Gaussians: p (F ) =

å

i =1

i

si 2p

çè

2si2

÷ø

208

Rankl et al.

Fig. 6. The rupture force distribution of virus interacting with a cell is shown at two different encounter times. (a) 330 ms and (b) 670 ms. Shorter encounter times show more single receptor–virus interaction breakings. With longer encounter times, the number of single receptor–virus binding decreases, concomitantly double receptor–virus bindings increase. N

åA

=1 the boundary condition i=1 i , where Ai is the fraction of rupture events corresponding to this peak, m i is the position of the peak, and si is the width of the peak. With increasing encounter time, the positions mi and widths of the peaks si were constant. In contrast, the fractions Ai changed and more multiple receptor bonds were found (Fig. 6). The force needed to unbind two receptors from a single virus is not necessarily twice the force needed to rupture a single receptor–virus bond. For uncorrelated binding, it is smaller than the double value. In addition, the most probable unbinding force for multiple receptors–virus interactions does not depend linearly on the logarithm of the loading rate anymore (8). Therefore, the kinetic off-rate constant must be determined from the loading rate behavior of the force peak of a single receptor–virus interaction.

4. Notes 1. MATLAB (The Mathworks, Natick, MA, USA) was used for all programs and calculations. 2. The thermal noise method to calibrate the spring constant depends as 1 / f ols 2 on the sensitivity fols; therefore, a proper calibration of the z-scanner is necessary to achieve highest accuracy prior to every spring constant measurements. A sensitivity-free method to determine the spring constant is the Sader method, which requires knowing the plain view geo­ metry of the cantilever. In our lab, a program was written that

Determination of the Kinetic On- and Off-Rate of Single Virus–Cell Interactions

209

is capable to determine the spring constant using the thermal noise and Sader method within the same measurement. 3. Another important detail to note is that the standard deviation of the Fourier transform is 100%, independent of the number of sample points. But the mean value can be estimated very accurately by averaging several measurements. For this, the recorded free cantilever movement was split into five to ten parts of equal length; a power density estimation of every part was performed. The final power density was calculated by averaging over all parts (18). 4. Accuracy estimates of koff and x b can be derived using bootstrapping (26). 5. In the last years, additional models describing dynamic force spectroscopy have been published (27–29), which could also be used to extract information about the binding energy landscape. 6. Prior to kinetic on-rate measurements, an expected t value should be estimated by solving Eq. 3 for t, using an estimate for kon. The dwell times on the surface should be adjusted to this estimate. References 1. Marsh, M. and A. Helenius (2006) Virus entry: Open sesame. Cell 124, 729–740. 2. Harrison, S.C. (2008) Viral membrane fusion. Nat Struct Mol Biol 15, 690–8. 3. Falanga, A., et  al. (2009) Membrane fusion and fission: enveloped viruses. Protein Pept Lett 16, 751–9. 4. Binnig, G., C.F. Quate, and C. Gerber (1986) Atomic Force Microscope. Phys. Rev. Lett. 56, 930–933. 5. Moy, V.T., E.L. Florin, and H.E. Gaub (1994) Intermolecular forces and energies between ligands and receptors. Science 266, 257–259. 6. Hinterdorfer, P., et  al. (1996) Detection and localization of individual antibody-antigen recognition events by atomic force microscopy. Proc. Natl. Acad. Sci. USA. 93, 3477–3481. 7. Lee, G.U., L.A. Chrisey, and R.J. Colton (1994) Direct measurement of the forces between complementary strands of DNA. Science 266, 771–773. 8. Rankl, C., et  al. (2008) Multiple receptors involved in human rhinovirus attachment to live cells. Proceedings of the National Academy of Sciences 105, 17778–17783. 9. Hewat, E.A., et al. (2000) The cellular receptor to human rhinovirus 2 binds around the

5-fold axis and not in the canyon: a structural view. EMBO J. 19, 6317–6325. 10. Herdy, B., et al. (2004) Identification of the human rhinovirus serotype 1A binding site on the murine low-density lipoprotein receptor by using human-mouse receptor chimeras. J Virol 78, 6766–74. 11. Ishibashi, S., et al. (1993) Hypercholesterolemia in low density lipoprotein receptor knockout mice and its reversal by adenovirus-mediated gene delivery. J Clin Invest 92, 883–893. 12. Herz, J., D.E. Clouthier, and R.E. Hammer (1992) LDL receptor-related protein internalizes and degrades uPA-PAI-1 complexes and is essential for embryo implantation. Cell 71, 411–421. 13. Gibson, C.T., G.S. Watson, and S. Myhra (1996) Determination of the spring constants of probes for force microscopy/spectroscopy. Nanotechnology 7, 259–262. 14. Cleveland, J.P., et al. (1993) A nondestructive method for determining the spring constant of cantilevers for scanning force microscopy. Review of Scientific Instruments 64, 403–405. 15. Hutter, J.L. and J. Bechhoefer (1993) Calibration of atomic-force microscope tips. Rev. Sci. Instrum. 64, 1868–1873.

210

Rankl et al.

16. Sader, J.E., et al. (2005) General scaling law for stiffness measurement of small bodies with applications to the atomic force microscope. J. Appl. Phys. 97, 124903–7. 17. Ohler, B. Practical Advice on the Determination of Cantilever Spring Constants. 2007; Available from: http://www.veeco.com/pdfs/appnotes/AN94%20Spring%20Constant%20 Final_304.pdf. 18. Press, W.H., (2007) Numerical Recipes: The Art of Scientific Computing. 3 ed. 2007: Cambridge University Press. 19. Butt, H.J. and M. Jaschke (1995) Calculation of thermal noise in atomic force microscopy. Nanotechnology 6, 1–7. 20. Stark, R.W., T. Drobek, and W.M. Heckl (2001) Thermomechanical noise of a free v-shaped cantilever for atomic-force microscopy. Ultramicroscopy 86, 207–215. 21. Evans, E. and K. Ritchie (1997) Dynamic strength of molecular adhesion bonds. Biophysical Journal 72, 1541–1555. 22. Izrailev, S., et al. (1997) Molecular dynamics study of unbinding of the avidin-biotin complex. Biophys. J. 72, 1568–1581.

23. Baumgartner, W., et  al. (2000) Affinity of Trans-interacting VE-cadherin Determined by Atomic Force Microscopy. Single Molecules 1, 119–122. 24. Verdaguer, N., et al. (2004) X-ray structure of a minor group human rhinovirus bound to a fragment of its cellular receptor protein. Nature Structural Molecular Biology 11, 429–434. 25. Sulchek, T., R.W. Friddle, and A. Noy (2006) Strength of Multiple Parallel Biological Bonds. Biophys J 90, 4686–4691. 26. Rankl, C., et al. (2007) Accuracy Estimation in Force Spectroscopy Experiments. Japanese Journal of Applied Physics 46, 5536. 27. Dudko, O.K., et al. (2003) Beyond the conventional description of dynamic force spectroscopy of adhesion bonds. Proc. Natl. Acad. Sci. USA. 100, 11378–11381. 28. Hummer, G. and A. Szabo (2003) Kinetics from nonequilibrium single-molecule pulling experiments. Biophys. J. 85, 5–15. 29. Raible, M., et al. (2006) Theoretical analysis of single-molecule force spectroscopy experiments: heterogeneity of chemical bonds. Biophys. J. 90, 3851–3864.

Chapter 14 Atomic Force Microscopy as a Tool for the Study of the Ultrastructure of Trypanosomatid Parasites Wanderley de Souza, Gustavo M. Rocha, Kildare Miranda, Paulo M. Bisch, and Gilberto Weissmuller Abstract Here, we describe the methodology currently used to analyze the structural organization of protozoa of the Trypanosomatidae family by atomic force microscopy. The results are compared with those obtained using light, scanning, and transmission electron microscopy. Key words: Parasitic protozoa, Trypanosomatids, Atomic force microscopy, Scanning electron microscopy

1. Introduction Prior to the 1980s, analysis of the structural organization of ­biological samples was carried out primarily by light, scanning, and transmission electron microscopy. In 1981, Gerd Binnig and Heinrich Rohrer, working at the IBM laboratory in Zurich, developed the scanning tunneling microscope, which was first microscope able to generate atomic-resolution three-dimensional images, and used it to analyze the surfaces of materials with high conductivity (1). A few years later, the same group developed the atomic force microscope (AFM). The AFM did not require a highly conductive surface, and thus could be applied to analyze the surfaces of a large number of materials, including biological materials (2). More importantly, especially from a biological point of view, AFM can provide nanometer-resolution images of living cells in gaseous and liquid environments. As a result, even cells in culture can be examined by AFM, which opens new avenues for  the application of this technique in the biological sciences. Pier Carlo Braga and Davide Ricci (eds.), Atomic Force Microscopy in Biomedical Research: Methods and Protocols, Methods in Molecular Biology, vol. 736, DOI 10.1007/978-1-61779-105-5_14, © Springer Science+Business Media, LLC 2011

211

212

de Souza et al.

Since its inception, AFM has been used extensively to analyze diverse biological samples, including parasitic protozoa (3–10). It is now well established that AFM has an enormous potential for imaging and analyzing cells. This potential has been increasingly realized in the last few years, especially with the release of commercially available equipments, some even dedicated to biological applications. AFM has also been applied to study the structural organization of trypanosomatid parasites, which was first attempted by Dvorak and collaborators almost 10 years ago (4). While this study drew the attention of the parasitology community to the potential of this technique, the published images did not add significant new information regarding the structure of these cells. However, a subsequent study that used a similar methodology to prepare the samples but with the addition of a pretreatment of the protozoa with detergent produced images wherein both the cell surface and the intracellular structures of epimastigote forms of Trypanosoma cruzi could be well recognized (10). In addition, new structures were described, including (a) a flagellar furrow separating the axoneme from the paraflagellar rod (PFR) and running from the point of emergence of the flagellum from the flagellar pocket to the flagellar tip, (b) a row of periodically organized structures localized in the flagellar furrow, and (c) the structural organization of the flagellar necklace, which appears as nine protrusions positioned about 400  nm from the flagellar base and distributed around the circumference of the flagellum. This chapter describes the methods we have used to analyze the structure of trypanosomatids by atomic force microscopy. The methodology used is explained in great detail so that it can be easily repeated and adapted for similar studies with other trypanosome species and other protozoa. There are many AFM scan modes. The intermittent contact mode produces topographical information. Phase imaging is a complementary imaging mode that, in addition to providing topographical information, allows for the analysis of adhesive and elasticity properties. Herein, we describe such approaches for acquiring images of parasitic protozoa of the Trypanosomatidae family.

2. Materials 2.1. Protozoan Culture Medium

1. Liver infusion tryptose (LIT) medium: contains 0.4% sodium chloride, 0.2% glucose, 0.65% anhydrous disodium hydrogen phosphate, 0.5% tryptose, 0.5% liver infusion broth, 0.04% potassium chloride, 0.05% hemin, and 0.02% folic acid. 2. LIT medium is supplemented with 10% fetal bovine serum.

Atomic Force Microscopy as a Tool for the Study of the Ultrastructure

213

3. Sodium phosphate buffer (to obtain 200 mL of 0.1 M sodium phosphate buffer pH 7.4, combine 19 mL of 0.2 M monobasic sodium phosphate, NaH2PO4, 81  mL of 0.2  M dibasic sodium phosphate, Na2HPO4, and 100  mL of distilled water). 4. Glass pipettes (5, 1 mL) and rubber bulb. 5. Calibrated pH meter. 6. Conventional centrifuge and a high-speed microcentrifuge. 7. Centrifuge tubes (conical, 15 mL) or microcentrifuge tubes. 8. Top loading balance for measuring buffer salts and resin components. 9. Hot plate with magnetic stirrer and stirring bar. 10. Magnetic stir bars. 11. Culture tubes or plastic flasks. 12. Dissecting needle. 2.2. Whole Intact Cell Analysis

1. 2.5% Glutaraldehyde (EM grade). 2. Sodium cacodylate buffer (to obtain 100 mL of 0.1 M sodium cacodylate buffer pH 7.4, combine 25 mL of 0.2 M sodium cacodylate, Na(CH3)AsO2 · 3H2O, 1.4  mL of 0.2  N hydrochloric acid, HCl, and 23.6 mL of distilled water). 3. Ethanol dehydration series (Merck – product number 100983). 4. Fume cupboard (Fume hood). 5. Critical point dryer (Baltec – CPD 030). 6. 0.01% Poly-l-lysine (70,000–150,000 MW; Sigma–Aldrich). 7. Ethanol dehydration series (Merck – product number 100983). 8. Fume cupboard (Fume hood). 9. Critical Point Dryer (Baltec – CPD 030).

2.3. Atomic Force Microscopy

1. PHEM buffer containing 60 mM pipes (1,4-piperazine diethylsulfonic acid), 20 mM Hepes (N-2-hydroxyethylpiperazine N-1-2-ethanesulfonic acid), 10  mM EGTA, and 2  mM MgCl2, pH 7.2. 2. 1% Nonidet NP-40. 3. V-shaped standard narrow cantilevers, model NP-S (Veeco Probes, Camarillo, CA, USA). 4. Tetrahedral-shaped cantilevers, model AC240TS (Olympus, Tokyo, Japan).

214

de Souza et al.

2.4. Field Emission Scanning Electron Microscopy

1. 2.5% Glutaraldehyde (EM grade). 2. Sodium cacodylate buffer (to obtain 100 mL of 0.1 M sodium cacodylate buffer pH 7.4, combine 25 mL of 0.2 M sodium cacodylate, Na(CH3)AsO2 · 3H2O, 1.4  mL of 0.2  N hydrochloric acid, HCl, and 23.6 mL of distilled water). 3. 0.01% Poly-l-lysine (70,000–150,000 MW; Sigma–Aldrich). 4. Osmium tetroxide (Electron Microscopy Sciences, stock aqueous solution). 5. Potassium ferrocyanide (to obtain 50 mL of 1.6% potassium ferrocyanide stock solution, combine 0.8 g potassium ferrocyanide, K4[Fe(CN)6] · 3H2O, 10  mM calcium chloride, CaCl2, and 50 mL of 0.2 M sodium cacodylate buffer). 6. Ethanol dehydration series (Merck – product number 100983). 7. Fume cupboard (Fume hood). 8. Critical Point Dryer (Baltec – CPD 030). 9. High-resolution ion beam coater (GATAN – model 681). 10. SEM specimen stubs. 11. Desiccated or dry compartment to store SEM specimen stubs.

2.5. Plasma Membrane Extraction

1. PHEM buffer containing 60 mM pipes (1,4-piperazine diethylsulfonic acid), 20 mM Hepes (N-2-hydroxyethylpiperazine N-1-2-ethanesulfonic acid), 10  mM EGTA, and 2  mM MgCl2, pH 7.2. 2. 1% Nonidet NP-40. 3. 2.5% Glutaraldehyde (EM Grade). 4. 0.01% Poly-l-lysine (70,000–150,000 MW; Sigma–Aldrich). 5. Ethanol dehydration series (Merck – product number 100983). 6. Fume cupboard (Fume hood). 7. Critical Point Dryer (Baltec – CPD 030).

3. Methods 3.1. Cultivation of Trypanosomatids

T. cruzi epimastigotes were cultivated in LIT medium supplemented with 10% fetal bovine serum and 1% hemin for 3–5 days at 28°C. Only parasites collected in the exponential phase of growth were used for experiments.

Atomic Force Microscopy as a Tool for the Study of the Ultrastructure

3.2. Preparation of Samples for Scanning Electron Microscopy

215

1. Fixation. Cells were collected by centrifugation (1600 × g for 10 min at 4°C) from 3- to 5-day-old cultures, washed with 0.1  M phosphate buffer, pH 7.2, and fixed for 60  min at room temperature in a solution containing 2.5% glutaraldehyde in 0.1 M phosphate or PHEM buffer, pH 7.2. For scanning electron microscopy following glutaraldehyde fixation, the cells were washed in buffer, postfixed for 1 h at 4°C in a solution containing 1% osmium tetroxide and 0.8% potassium ferrocyanide in 0.1 M sodium cacodylate buffer, pH 7.2. 2. Adhesion of protozoa. As trypanosomatids are single isolated cells, they were adhered for 10 min at room temperature into glass coverslips previously coated with 0.1% poly-l-lysine. Before use, the coverslips were cleaned with soap and water, rinsed in distilled water, and dried in a dust-free environment. For more details, see Note 1. 3. Plasma membrane extraction. For some experiments, the adhered cells were incubated with 1% Nonidet NP-40 diluted in PHEM buffer for 5–7 min in order to remove the plasma membrane. They were then fixed in glutaraldehyde as described above. For more details, see Note 2. 4. Dehydration. The cells were dehydrated using an ethanol series (30, 50, 70, 90, and 100%). The ethanol was of good quality. The cells were incubated for 10 min in each ethanol solution. 5. Critical point drying. To dry the samples and avoid surface tension, critical point drying (BALTEC – CPD030) was used. After dehydration, the cells still adhered to the glass coverslips were transferred to the equipment. The compartment containing the samples was closed and the temperature and pressure were changed to about 8°C and 50 atm. Subsequently, the ethanol was gradually replaced by liquid carbon dioxide. When the substitution was complete, the temperature and pressure of the compartment were elevated to over 31°C and 73 atm to pass to the critical point of the working fluid, thus avoiding the direct liquid–gas transition seen in ordinary drying. For more details, see Note 3. 6. Metal coating. After critical point drying, small pieces of the glass coverslips were adhered to special stubs that were then introduced into the chamber of the high-resolution ion beam coater (GATAN – model 681). The samples were coated for 5 min with a 2–3-nm thick gold or chromium layer using the appropriate voltages and currents indicated by the manufacturer (400 mA). 7. Scanning electron microscopy. Observation was carried out using secondary electron imaging in a Jeol JSM 6340F field emission scanning electron microscope operating at 5.0 kV,

216

de Souza et al.

Fig. 1. Light and electron micrographs of epimastigote forms of Trypanosoma cruzi. (a) Differential interference contrast micrograph of T. cruzi showing the whole parasite. It is possible to visualize the whole parasite using this technique. The arrow points to the anterior region where a free flagellum is seen; (b) Transmission electron micrograph of a thin section of T. cruzi. Subcellular structures, such as the nucleus (N), kinetoplast (K), Golgi complex (G), mitochondrion (asterisk), and the flagellum (arrow ), are seen. Microtubules of axonem are also observed (arrowheads ). (c) Field emission scanning electron micrograph of T. cruzi showing its elongated form and its smooth surface. Arrow indicates the flagellum. Bars – (a) 5 mm; (b) 1 mm; and (c) 2 mm.

12  A emission current, and a working distance of 8  mm. Figure  1 shows a general view of the epimastigote form of T. cruzi as seen by light and electron microscopy. These images show the general shape of the protozoan from the posterior to anterior end of the parasite. The flagellum of this trypanosomatid emerges from the flagellar pocket (FP) and remains tightly attached to the cell body along its length (Fig. 1b, c). Scanning electron microscopy shows the flagellar tip as a single and uniform cylindrical shaped structure (Fig. 2). Figure 3 shows a schematic view of the protozoan, based mainly on observations made both by scanning and transmission electron microscopy of thin sections in order to facilitate the interpretation of the images shown. For more details, see Note 4. 3.3. AFM Imaging Techniques

1. Contact mode analysis. The glass slide containing the sample was mounted onto the XY scanner of the AFM and a CCD camera was used to locate the parasites. For contact mode analysis, a soft cantilever was used. V-shaped standard narrow cantilevers, model NP-S (Veeco Probes, Camarillo, CA), were used. This model of cantilever includes four cantilevers differing in length and width. For the experiments, we used the softer cantilever with a nominal spring constant of 0.12 N/m and nominal frequency of 18 kHz. Cantilever elastic ­constants

Atomic Force Microscopy as a Tool for the Study of the Ultrastructure

217

Fig. 2. FESEM image of the end of the flagellum of an epimastigote form of T. cruzi. The flagellum appears as a cylindrical structure and its end is seen as a unique and uniform structure. Bar – 1 mm.

Fig. 3. Illustration showing some structures and organelles found in the epimastigote form of T. cruzi based on images obtained by transmission electron microscopy.

were obtained by the thermal noise method. Samples were scanned at a constant force with a low scan rate (0.6 Hz) to reduce noise and minimize sample damage. Force and integral gain was constantly monitored to use minimum force with good response of the feedback system to obtain better images. To get high-resolution images, samples were acquired with 512 points × 512 lines of resolution. In the contact mode, the cytoskeleton could be detected after plasma membrane extraction (Fig. 4). 2. Intermittent contact mode analysis. For the intermittent contact mode, tetrahedral-shaped cantilevers (AC240TS, Olympus, Tokyo, Japan – nominal spring constant 2  N/m and 70  kHz of nominal frequency) were used. These are

218

de Souza et al.

Fig. 4. Three-dimensional view of a height image of the flagellum (f) of a T. cruzi epimastigote previously extracted with detergent and scanned in contact mode. Microtubules (arrow ) of the flagellar axoneme can be observed in this image. The subpellicular microtubules (asterisk) and the basal body (bb) are also seen. Bar – 2 mm.

interesting because their tips are located on the very end of the cantilever, and they offer a very good lateral resolution. Cantilever elastic constants were obtained by the thermal noise method. Samples were scanned at a constant force with a low scan rate (0.5 Hz) to reduce noise and minimize sample damage. Force and integral gain were constantly monitored to use minimum force with good response of the feedback system to obtain better images. To get images with good resolution, samples were acquired with 512 points × 512 lines of resolution. Figure 5a shows the topographic signal obtained by scanning the flagellum region of a T. cruzi epimastigote using the intermittent contact mode. This image allows the observation of a furrow along the main axis of the flagellum. This structure has not been seen by any other kind of light or electron microscopy techniques. 3. Phase mode analysis. Phase mode occurs when the tip is in intermittent contact with sample. This mode determines the topographic, viscoelastic, and adhesive properties of the samples. In this mode, the phase shift of the oscillating cantilever is measured relative to the driving signal. Images were obtained at the same moment of intermittent contact mode. Figure 5b shows phase signal image of the same region detected by intermittent contact mode shown in Fig. 5a. The images obtained by phase mode have more contrast, allowing for better

Atomic Force Microscopy as a Tool for the Study of the Ultrastructure

219

Fig. 5. AFM images of the flagellum of T. cruzi previously extracted with detergent. The use of intermittent contact mode allowed a better view of the topography of the flagellum and the furrow that exists in its major axis (arrows) (a). Phase image (b) complements the information from the flagellum height signal (a). Moreover, phase mode is more evident when the phase signal is overlaid with the topographical image (c). Three-dimensional visualization reveals higher levels of detail in the analyzed structures.

v­ isualization of some flagellar structures, such as the PFR and the periodic organization of protrusions in the furrow. 4. Image analysis. AFM image processing (line-wise flattening only) was performed in IGOR-PRO (Wavemetrics, Portland, OR) using a MFP-3D template developed by Asylum Research. For better topographic visualization, topographic, phase, and an overlay of both images are displayed as three-dimensional views. Figure 5c shows an overlay of the phase image and the topographic image. For more details, see Note 5.

220

de Souza et al.

4. Notes 1. 106 cells/mL is a good cell density starting point to have the trypanosomatids well distributed in a 100 mm2 region. 2. The plasma membrane extraction could be done with many different detergents, or solvents, such as methanol. We found that 1% Nonidet NP-40 diluted in PHEM buffer worked well. However, it is important to note that this concentration, as well as the type of detergent and period of membrane extraction, can be changed. We suggest starting with 1% Nonidet NP-40 in PHEM buffer and then analyzing the refractive power of the cell by light microscopy. Transmission electron microscopy must also be used to confirm plasma membrane extraction. 3. To observe biological samples by SEM or in air conditions by AFM, cells must be critical point-dried because of the high vacuum in the SEM chamber, which could change the integrity of the specimen surface due to the boiling of reminiscent hydrated specimens. Air-dried specimens could cause deformations and the collapse of structures. Because of this, the use of a machine that substitutes the liquid embedding the samples for one with a lower surface tension could reduce damage to the sample. Using critical point drying, it is possible to pass from liquid to gas without any abrupt change in state. Critical point drying is the safest method for drying samples. Additionally, after plasma membrane extraction, critical point drying can be used to maintain the architecture of cells. 4. Field emission scanning electron microscopy allows for the acquisition of high-resolution images due to the field emission cathode in the electron gun of the scanning electron microscope, which produces narrower probing beams at both low and high electron energies, resulting in improved spatial resolution. 5. The image processing techniques used in all types of microscopy are intended to most accurately reproduce the information obtained. The simplest processing technique is the adjustment of the brightness and contrast of the image. Other techniques could include operations, such as image rotation, warping, color balancing, etc. These may change the results of the image and increase the fluorescence signal. For example, algorithmic programs like blur and sharpen can be used to reduce and enhance image details, respectively, by changing pixel values based on a subjective combination with surrounding pixels.

Atomic Force Microscopy as a Tool for the Study of the Ultrastructure

221

Acknowledgments The authors are grateful for support from Conselho Nacional de Desenvolvimento Cinetífico e Tecnológico (CNPq), Financiadora de Estudos e Projetos (FINEP), Fundação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), and Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ). References 1. Binnig, G. and Rohrer, H. (1982) Scanning tunneling microscopy. Helv. Phys. Acta. 55, 726–735. 2. Binnig, G., Quate, C. F. and Gerber, C. (1986) Atomic force microscopy. Phys. Rev. Lett. 56, 930–933. 3. Bustamante, C. and Dunlap, D. (1991) Application of scanning tunneling microscopy to structural biology. Semin. Cell Biol. 2, 179–185. 4. Dvorak, J. A., Kobayashi, S., Abe, K., Fujiwara, T., Takeuchi, T. and Nagao, E. (2000) The application of the atomic force microscope to studies of medically important protozoan parasites. J. Elec. Microsc. 49, 429–435. 5. Akaki, M., Nakano, Y., Nagayasu, E., Nagakura, K., Kawai, S. and Aikawa, M. (2001) Invasive forms of Toxoplasma gondii, Leishmania amazonensis and Trypanosoma cruzi have a positive charge at their contact site with host cells. Parasitol. Res. 87, 193–197.

6. Dufrêne, Y. F. (2002) Atomic force microscopy, a powerful tool in microbiology. J. Bacteriol. 184, 5205–5213. 7. Dufrêne, Y. F. (2008) Towards nanomicrobiology using atomic force microscopy. Nat. Ver. Microbiol. 6, 674–680. 8. Wright, C. J. and Armstrong, I. (2006) The application of atomic force microscopy force measurements to the characterization of microbial surfaces. Surf. Interface Anal. 38, 1419–1428. 9. Garcia, C. R., Takeuschi, M., Yoshioka, K. and Miyamoto, H. (1997) Imaging Plasmodium falciparum-infected ghost and parasite by atomic force microscopy. J. Struct. Biol. 119, 92–98. 10. Rocha, G. M., Miranda, K., Weissmüller, G., Bisch, P. M., de Souza, W. (2008) Ultrastructure of Trypanosoma cruzi revisited by atomic force microscopy. Microsc. Res. Tech. 71, 133–139.

wwwwwwwwwwwwww

Chapter 15 Normal and Pathological Erythrocytes Studied by Atomic Force Microscopy Andreas Ebner, Hermann Schillers, and Peter Hinterdorfer Abstract Erythrocytes (red blood cells, RBCs) are the most common type of blood cells in vertebrates. Many diseases and dysfunctions directly affect their structure and function. Employing the atomic force microscope (AFM) physical, chemical, and biological/physiological properties of RBCs can be studied even under near-physiological conditions. In this chapter, we present the application of different AFM techniques to investigate and compare normal and pathological RBCs. We give a detailed description for nondestructive immobilization of whole intact RBCs and explain preparation techniques for isolated native RBC membranes. High-resolution imaging of morphological details and pathological differences are demonstrated with healthy and systemic lupus erythematosus (SLE) erythrocytes revealing substructural changes due to SLE. We also present the technique of simultaneous topography and recognition imaging, which was used to map the distribution of cystic fibrosis transmembrane conductance regulator sites on erythrocyte membranes in healthy and cystic fibrosis-positive RBCs. Key words: Erythrocytes, TREC, Membrane preparation, Molecular recognition, Recognition imaging, Cystic fibrosis, AFM tip chemistry

1. Introduction Erythrocytes (red blood cells, RBCs) are the most common type of blood cells in vertebrates. They are well known for their ability to transport oxygen. Many diseases and dysfunctions directly affect the structure and function of erythrocytes. Employing the atomic force microscope (AFM) physical, chemical, and ­biological/physiological properties of RBCs can be studied even under near ­physiological conditions. A number of AFM studies have focused on disease-related changes (for a review thereof, see ref. 1). In 1992, Zacheé already investigated changes in the shape of RBCs of patients after splenectomy (2). Investigations on RBC ­deformations Pier Carlo Braga and Davide Ricci (eds.), Atomic Force Microscopy in Biomedical Research: Methods and Protocols, Methods in Molecular Biology, vol. 736, DOI 10.1007/978-1-61779-105-5_15, © Springer Science+Business Media, LLC 2011

223

224

Ebner, Schillers, and Hinterdorfer

(3–5), nano-rheological properties (6), as well as viscoelastic s­ tudies (7) brought new insights into the behavior of RBCs. Hypercholesterolemia decreases the deformability of RBCs, which impairs their hemorheological behavior and ­promotes atherosclerosis (8). The hemolytic activity of the amphipathic peptides has been correlated to the phosphocholine-to-sphingomyelin ratio (PC/SM) (9). Hyperhomocysteinemia and lipid abnormalities are commonly found in patients with chronic renal failure; both are recognized as risk factors for atherosclerosis (10). Many diseases of the heart and circulatory system have been linked with insufficient deformability and increased aggregability of RBCs (11). Further­ more, in essential hypertension, the deformability of RBCs is reduced (12). Excessive amounts of NO leads to damage of erythrocyte plasticity. The loss of deformability is accompanied by increased fragility of erythrocyte membranes as measured by enhanced release of free hemoglobin (13). Changes in erythrocyte membrane parameters, especially in phospholipids, can produce serious metabolic disorders and influences the rheologic properties of erythrocytes in patients with Binswanger’s disease (14). The observed changes in erythrocyte membrane fluidity among Cyclosporin A-treated patients correlate with more frequent prevalence of hemolytic anemia among Cyclosporin A (15). Therefore, RBCs have been used as a prototypical cellular system to study drug-mediated plasma bilayer effects (16). Here, we describe the application of different AFM techniques to investigate and compare normal and pathological RBCs. Healthy and cystic fibrosis (CF)-positive RBCs, as well as systemic lupus erythematosus (SLE)-positive RBCs have been probed. Since AFM allows measurements under physiological conditions, it offers a well-suited technique to explore morphological details and follow functional changes in RBC. For imaging of morphological details and pathological differences, a whole intact RBC has to be immobilized in a nondestructive way (e.g., on a wheat germ agglutinin matrix) and should be scanned using a gentle imaging technique like tapping-mode AFM. In this chapter, such an experiment is demonstrated with healthy and SLE erythrocytes revealing substructural changes due to SLE. For the investigation of membrane proteins at the single protein level and for investigation of the inner side of the cellular membrane, further preparation techniques are needed. Hence, a second focus of this chapter is a preparation technique to produce isolated but native inside-out orientations of RBC membranes. RBCs are tightly attached to a support and, e.g., exposed to a fluid flow-imposed shear stress which opens the cells. By imaging the flat RBC membrane, more topographical details can be resolved, than when imaging the whole cell. A third main aspect of this chapter is the introduction of a recently developed AFM technique, which facilitates the simultaneous acquisition of topographical images and corresponding

Normal and Pathological Erythrocytes Studied by Atomic Force Microscopy

225

recognition maps. This technique is exemplified in a comparative study of RBC membranes from healthy individuals and cystic fibrosis patients. The cystic fibrosis transmembrane conductance regulator (CFTR) is a cAMP activatable membrane protein that acts not only as an ion channel, but also as a regulator of several membrane conductances (17). A mutation in the gene encoding for CFTR results in the severe disease of cystic fibrosis. The most common CF-associated mutation is the deletion of phenylalanine at residue 508, ∆F508 CFTR. CFTR bearing the ∆F508 mutation fails to progress through the normal biosynthetic pathway and fails to traffic to the plasma membrane. As a result, CFTR ∆F508 is mislocalized and is not present in the apical membrane of epithelia cells (18). Consequently, the apical membrane of CF cells is Cl− -impermeable resulting in an impaired electrolyte transport and fluid secretion by several epithelia, including the sweat duct, exocrine pancreas, and the pulmonary airways (19). However, it was shown that the membrane distribution of ∆F508 CFTR is tissue-specific and exhibits variation of expression from null to apparently normal amounts (20, 21). CFTR is not only found in epithelia, but also in human erythrocytes as shown by using Western blot techniques (22), in studies of Plasmodium falciparum-induced channel activation (23), by deformationinduced CFTR-dependent ATP-release (24) and functional studies (25, 26). With recognition imaging, it was possible to prove unequivocally that CF-positive erythrocytes have a significant lower number of CFTR proteins compared to healthy RBCs.

2. Materials 2.1. Tight Attachment of Whole RBCs

1. Mica (Muscovit) sheets (E. Groepl, Austria).

2.1.1. Aminofunctio­ nalization

3. Molecular sieve 0.4 nm (Merck, Germany).

2. Ethanolamine hydrochloride (Sigma–Aldrich, Austria). 4. Ethyleneglycol-bis(Succinimidyl.succinate) = EGS (Sigma–Aldrich, Austria). 5. Desiccator 5 L with O-ring.

2.1.2. Covalent Coupling of Capturing Biomolecules (e.g., Wheat Germ Agglutinin) 2.1.3. Red Blood Cell Preparation and Attachment

1. Phosphate buffered saline (PBS): 5 mM Na2HPO4, 150 mM NaCl, pH 7.4. 2. Capturig biomolecules (e.g. wheat germ agglutinin, WGA). 3. Small glass chamber, ~30 mm diameter. 1. Purified RBCs.

226

Ebner, Schillers, and Hinterdorfer

2.2. Isolated RBC Membranes 2.2.1. Membrane Preparation

1. PBS: 5 mM Na2HPO4, 150 mM NaCl, pH 7.4. 2. Ethylene glycol tetra-acetic (EGTA) (Sigma, Deisenhofen, Germany). 3. Very low salt buffer (vlsb): 0.3  mM Na2HPO4/NaH2PO4, 0.2 mM EGTA, pH 7.4. 4. Poly-l-lysine (PLL) (Sigma, Deisenhofen, Germany). 5. Paraformaldehyde (PFA) (Sigma, Deisenhofen, Germany). 6. Bovine serum Germany).

albumin

(BSA)

(Sigma,

Deisenhofen,

7. Glutaraldehyde (Sigma, Deisenhofen, Germany). 8. Masterflex PTFE diaphragm pump 7090-62 (Cole-Parmer Instruments, Chicago, USA). 9. Microcentrifuge 5417R (Eppendorf, Germany). 10. Shaker MTS 4 electronic (IKA-Works, Wilmington, USA). 2.2.2. Imaging of Isolated Membranes

1. Multimode AFM equipped with Nanoscope IIIa controller (Veeco, Santa Barbara, USA) or PicoPlus AFM equipped with MACMode controller (Agilent, Santa Clara, USA). 2. MSCT SiN3 cantilever k = 0.01 N/m (Veeco, Santa Barbara, USA).

2.2.3. Recognition Imaging of Isolated Membranes

1. Magnetically coated AFM cantilever, spring constant ~0.1 N/m. 2. Aminopropyl-triethoxysilane (APTES). 3. Triethylamine (TEA). 4. Heterobifunctional polyethylene(glycol) cross-linker. 5. Specific target molecule (e.g., antibody). 6. Dessicator (5 L) with an O-ring. 7. AFM setup equipped with a topography and recognition imaging (TREC) box (Agilent, Santa Clara, USA).

3. Methods The investigation of RBCs with the AFM can be performed on the outer or inner cell membrane. In the case of the outer membrane, it is possible to image whole intact cells, which should be gently fixed to increase the lateral resolution. This allows the exploration of morphological details and pathological differences of whole RBCs, as was observed in a study of Kamruzzahan et al. There, the morphology of erythrocytes from healthy humans with erythrocytes from SLE-positive patients was compared. To perform such measurements, the RBCs have to be mounted firmly on a flat solid

Normal and Pathological Erythrocytes Studied by Atomic Force Microscopy

227

support. For anchoring cells to a substrate, flexible linkers are appropriate tools to achieve immobilization avoiding direct physical contact to the sample surface (27). Thus, the cells are anchored at a small distance from the surface, which may facilitate unaffected cell physiology. Major tasks for the development of such binding protocols are the realization of defined and suitable coupling chemistry protocols and the adjustment of the surface density of the linkers. In Subheading 3.1 a tight and noninvasive immobilization technique to mount intact RBCs to a flat surface using a short flexible linker is presented. This anchoring protocol is a prerequisite to resolve structural details of the outer cellular membrane using a gentle AFM imaging technique like the tapping mode. To probe the inner side of a red blood cell, a completely different approach has to be used, whereby the RBCs need to be opened. This is described in detail in Subheading  3.2. Here, a technique called “shear opening” allows the generation of tightly attached erythrocyte membranes on a glass coverslip surface facing the inner side out. As a result, the membrane appears significantly stiffer and high-resolution AFM techniques like TREC are possible (28). Subheading 3.3 focuses on TREC, a recent development in dynamic force microscopy. By oscillating a functionalized tip close to its resonance frequency during the lateral scan across the surface, both the sample topography and the corresponding map of recognition sites can be simultaneously obtained. This highly sophisticated technique has proven to be very powerful in biophysical research (29–31) (e.g., in the investigation of RBC membranes (28)). 3.1. Tight Attachment of Whole RBCs

The coupling process of intact RBCs to a solid support can be performed in different ways. One possibility is the use of PLLcoated glass slides (1, 32). The attachment of RBCs is achieved by simple electrostatic interactions between the negatively charged cellular membrane and the positively charged PLL-coated glass surface. In this section, a tight but noninvasive and physiological way to anchor RBCs is shown (3), in which the cell is immobilized via small spacers. The protocol consists of several coupling steps. In the first step, the nonreactive flat surface (e.g., mica) has to be converted into a chemically addressable surface. The ethanolamine aminofunctionalization (33) turned out to be best suited for this purpose (Fig. 1(1)) (see Note 1). In the next step, a short homobifunctional cross-linker allows conversion of the outer amino groups into an amino-reactive surface. EGS has two reactive N-hydroxy-succinimide ester groups, highly reactive toward amino groups, whereas the ethyleneglycol in between acts as a spacer molecule. This space is short enough to avoid loop formation on the aminofunctionalized surface. The first EGS coupling step (to surface amino groups) is performed in a solvent

228

Ebner, Schillers, and Hinterdorfer

Fig. 1. Coupling scheme of erythrocytes via surface-bound lectin. (1) Aminofunction­ alization using ethanolamine hydrochloride. (2) EGS is used to invert the aminofunctionalized surface into a lysine reactive one. (3) Finally, the outer NHS-ester of EGS is used to couple WGA (a lectin with high affinity to membrane glycoproteins). Taken from ref. 3.

(Fig. 1(2)) while the second reactive group of EGS, which is used to covalently couple proteins via their lysine residue, is performed in aqueous buffer (Fig. 1(3)). WGA is used as a coupling protein for tight but noninvasive attachment of RBCs. WGA is a plant lectin that was found to bind carbohydrate moieties of the glycocalix of RBCs (34). In Fig. 3, the coupling scheme is shown. 3.1.1. Aminofunc­ tionalization of the Solid Support

1. Clean the surface by washing 3 × 5 min with ethanol (an ultrasonic bath is ideal). When using mica sheets, freshly cleave both sides directly before performing the following steps. 2. Dissolve 3.3 mg ethanolamine hydrochloride in 6 mL DMSO by gently heating to ~70°C. 3. Add 10% v/v molecular sieves (0.4 nm). 4. Cool down to room temperature. 5. Degas in a desiccator (at aspirator vacuum) for 30 min. 6. Place the freshly cleaved mica or the cleaned glass slides into the solution. 7. Incubate overnight.

Normal and Pathological Erythrocytes Studied by Atomic Force Microscopy

229

8. Wash 3× 5 min with DMSO and 3× 5 min with ethanol, dry in a gentle nitrogen or inert gas stream (see Note 2). 3.1.2. Coupling of WGA Via EGS

1. Dissolve 10  mg EGS in 10  mL chloroform in a small glass chamber. 2. Place the aminofunctionalized mica sheets or glass slides in the solution. 3. Add 0.5% (v/v) TEA and mix carefully. 4. React for 60 min. Afterward, wash extensively with chloroform and dry the slides in an inert gas or nitrogen stream. 5. Dissolve 0.5 mg/mL wheat germ agglutinin in PBS. 6. Incubate this solution on the EGS functionalized surface and react for 120  min. Afterwards remove unbound WGA by extensive rinsing with PBS. 7. Store in PBS at 4°C until further use.

3.1.3. Red Blood Cell Preparation and Attachment

1. RBCs should be freshly prepared from blood by a suitable isolation method (e.g., by discontinuous density gradient centrifugation). 2. Incubate the suspension of purified RBCs (less than 0.05% white blood cells) in PBS buffer on the WGA-coated surface for 20 min, and wash afterward extensively with PBS buffer. 3. Incubate the RBCs in 1% glutaraldehyde for 1 min and wash with PBS.

3.1.4. AFM Imaging

AFM imaging of RBCs at high lateral resolution is difficult because of the softness and the spherical shape of the cells. In very early studies, Häberle (35) presented an elegant solution to achieve a resolution of ~10 nm by using a special micropipette technique. However, the technique involved excessive technical effort and did not present a generally applicable approach. When using a commercial AFM, RBCs should be imaged with very soft AFM cantilevers to avoid any damage. Furthermore, intermittent contact imaging modes like acoustic or magnetic AC mode AFM are best suited. The AFM probe thus oscillates above the sample and touches the surface only at the end of the downward movement. Thereby, only a very low force is applied to the biological sample. In Fig. 2, AFM images of healthy (left A–D) and SLE-positive (right A–D) RBCs are shown, where the pathological SLE erythrocytes clearly exhibit deformations when compared to the healthy cells.

3.2. Isolated Membranes

The plasma membrane of a cell accommodates diverse membrane proteins, including integral membrane proteins such as receptors, ion channels, and transporters, as well as certain antigens that are peripherally associated with the membrane. The plasma membrane of eukaryotic cells is generally anything but flat. The curvature of

230

Ebner, Schillers, and Hinterdorfer

Fig. 2. Tapping-mode images of RBCs from healthy (left ) and SLE-positive (right ) patients at different scan sizes. Taken from ref. 3.

a cell is formed by lamellopodia, cell body, and cell nucleus. The membrane shows major structures like membrane ruffles (36), microvilli, and cilia, and submembranous structures like the cytoskeleton (37). A closer view reveals unevenness of the plasma membrane like humps and pits, representing endo- and exocytotic activity (38). A huge variety of proteins are heterogenically distributed within the membrane, and many membrane proteins are equipped with highly branched sugars forming the glycocalyx. The glycocalyx, a network of polysaccharides that protrudes up to 100 nm from cellular surfaces, limits the tip access to the membrane surface, thus reducing the resolution (39). These factors impede molecular resolution on whole cells, which is indispensable for dissolving the structure of membrane proteins with AFM. In imaging whole cells, the generated topographical information is limited to the outer cell surface, but several membrane proteins are located on the cytoplasmic side of the cell membrane. To gain more information about the structure of membrane proteins by AFM imaging, it is necessary to isolate cell membranes on a solid support in such a way that the intracellular face of the membrane is accessible for the AFM tip (“inside-out” orientation). 3.2.1. Plasma Membrane Preparation

The membrane of a RBC can be isolated according to a modified method of Swihart et al. (40). By this preparation, the RBC membrane is spread in patches on glass with an inside-out orientation, and gently shaken in a vlsb at 37°C for 20 min in order to remove hemoglobin and remnant cytoskeletal proteins. This approach

Normal and Pathological Erythrocytes Studied by Atomic Force Microscopy

231

offers several advantages: (a) the membrane is flat due to the fact that no curvature is imposed by underlying structures (e.g., cytoskeleton), (b) the membrane is stiff because it lies on a hard support instead of soft cytosol, and (c) the intracellular face of the membrane can be imaged by AFM at high resolution, since none of it stretches out over the surface. In general, the main principle in isolating cellular membranes is to “glue” the cell on a functionalized surface and subsequently remove the cell body, whereby membrane patches remain in an inside-out orientation (41, 42) on the support (Fig. 3). Preparation steps: 1. Wash glass coverslips with acetone and use a lens cleaning wipe. 2. Coat glass coverslips with 150 mL of 0.01% Poly-L-Lysine (PLL) in H2O for 30 min followed by two washing steps with H2O. 3. Collect ~5 mL of venous blood from donors with an EDTA blood draw container. Dilute four drops of blood with 1,000 mL isotonic PBS containing 0.2 mM EGTA. Spin down cells with 1,450 × g for 4  min and remove the supernatant. Repeat this washing procedure three times. Dilute packed RBC 1:600 with PBS. Transfer 150  mL of this suspension onto a PLL-coated glass coverslip allowing the RBC to adhere for 15 min. Then, remove the droplet. Attention: do not let the sample dry during all following steps!

Fig. 3. Schematic overview of shear opening procedure of red blood cells. Erythrocytes are exposed to fluid flow-imposed shear stress and, as a result, the cells are opened, exposing their cytoplasmic side of the membrane.

232

Ebner, Schillers, and Hinterdorfer

Fig. 4. Detailed scheme of shear opening procedure.

4. By applying shear stress to the immobilized RBC you get “inside-out” oriented RBC membranes. Use a stream of PBS buffer to remove the upper part of the cells by pressing PBS with a flow rate of 36 mL/min through a needle (21 G, distance and angle to the glass coverslip: 5  cm and 20°). The shear stress is applied for 12 s (Fig. 4). It is crucial to use the correct parameters. Changes in flow rate, time, or angle result in a deficient membrane preparation. 5. Wash the membranes, which are firmly attached to the glass surface, two times with a vlsb to clean membranes from hemoglobin and membrane residues. For this, add 150 mL vlsb to the glass coverslips and remove the droplet after 2 min. 6. To detach cytoskeletal proteins, add 150  mL of vlsb to the samples and shake the glass coverslips at 37°C for 20  min (1,000/m in a shaker: e.g., IKA-shaker MTS 4 electronic). 7. Wash the samples three times using 1,000 mL of PBS. This can be done in a 12-well-plate with one glass coverslip per well. 8. Fix the RBC membranes with 500 mL of 4% PFA in PBS for 40 min. Wash three times with 1,000 mL of PBS containing 3% of BSA. 9. To block against unspecific binding, incubate the samples in 1,000 mL of 3% BSA in PBS at room temperature (RT) for 60 min. 10. Wash the samples three times using 1,000 mL of PBS. 3.2.2. Imaging of Isolated Membranes

Soft cantilevers (e.g., with spring constants of 0.01 N/m) should be used for imaging of isolated RBC membranes to avoid damage to the membrane. Imaging can either be done in contact or in tapping mode. An optimal imaging speed is 1–20  mm/s. The above-described membrane preparation method results in a certain configuration of the sample, which is depicted in Fig. 5. Only region C has the required inside-out orientation of the RBC membrane. Thus, use an AFM equipped with a light microscope, so that the cantilever tip should be positioned over this region.

Normal and Pathological Erythrocytes Studied by Atomic Force Microscopy

233

Fig. 5. Scheme of the sample after preparation seen from the topview.

Fig. 6. AFM images demonstrating the membrane isolation procedure. (a) Erythrocytes attached to poly-l-lysine-coated glass, before exposure to shear stress and (b) after application of the jet stream, resulting in flat inside-out oriented membranes with circular shapes.

In Fig. 6, AFM images are presented showing intact RBCs on the left (Fig. 6a) and RBC membrane patches after application of shear stress on the right (Fig. 6b). The membrane patches attached on glass (Fig.  6b) exhibit flat surfaces with circular shapes and diameters of about 8 mm. Observed heights of the isolated plasma membranes, including the protruding membrane proteins, ranged between 10 and 15 nm. Overlapping membrane edges appeared frequently with heights of about 25  nm (yellow structures in Fig. 6b).

234

Ebner, Schillers, and Hinterdorfer

Fig. 7. High-resolution scan (5 mm × 5 mm) of the cytosolic face of a human erythrocyte. The high protein density together with the tip geometry (tip convolution) almost prevents imaging the lipid bilayer between the proteins. However, at some points the lipid bilayer is visible defining the zero height in this image (taken from ref. 44).

The fluid mosaic model describes a cell membrane as a twodimensional oriented solution of integral proteins in a viscous phospholipid bilayer. Current textbooks estimate a protein content of 50% for typical membranes (43). With this, not only ­proteins, but also a lipid bilayer should be visible in high-resolution­ AFM images of RBC membranes. Figure 7 shows a 5 mm × 5 mm scan of the cytosolic face of a human erythrocyte. A dense package of proteins with different shapes and sizes can be observed, yet the lipid bilayer is not visible. These observations are in good agreement to other AFM studies of human erythrocytes (45) where high-resolution images of the RBC membranes did not reveal ­lipids. The tip convolution contributes to this result, but not exclusively. Either globular shapes of huge intracellular polypeptide chains or a protein content much higher than expected could cause the observed appearance of the human red cell membrane without a visible lipid bilayer. 3.3. Recognition Imaging of Isolated Membranes

TREC allows mapping of topographical details of a specimen and simultaneous investigation of the distribution of receptors on the surface (see Note 3). A prerequisite for such measurements is the use of a sensing, ligand-functionalized AFM tip. The upgrade of a bare tip (usually silicon or silicon nitride) into a monomolecular sensor requires a number of coupling steps, which are listed in Subheading  3.3.1. Subsequently, optimal conditions for the

Normal and Pathological Erythrocytes Studied by Atomic Force Microscopy

235

TREC experiments are explained and an example for successful TREC imaging is presented with the sensing of CFTR in RBCs of healthy and CF diseased patients. 3.3.1. Tip Chemistry

The vapor phase deposition of APTES is the best suited method for aminofunctionalization of a given surface (see Note 1). By APTES coating, a nondense monomolecular silane layer with optimal physical and chemical behavior is introduced on the AFM tip. In addition, choosing the right parameters allow an accurate and easy adjustment of reactive sites per tip apex (33). A high number of distensible homo- and heterobifunctional cross-linkers have been developed. In the following, one of the most commonly used linkers, the NHS– PEG–aldehyde (46) linker, is exemplary presented to demonstrate ligand coupling to an AFM tip (see Note 4 and Fig. 8). 1. Wash the cantilevers 3× 5 min with chloroform and dry in a gentle nitrogen gas stream. Then, wash 3× 5 min with ethanol and dry again in a gentle nitrogen gas stream. 2. Flood a 5-L desiccator with argon. Open the desiccator and place a vial with 30 mL freshly distilled APTES and a vial with 10 mL TEA inside. 3. Place the cleaned cantilevers in the desiccator, rinse with argon for 60 s, and let it react for 120 min.

Fig. 8. Aminofunctionalized AFM tip gets upgraded into a monomolecular sensor. NHS– PEG–aldehyde linker is coupled via amide-bond formation. Subsequently, the aldehyde group is coupled to the amine residue of a lysine of a protein (e.g., of an antibody). Taken from ref. 28.

236

Ebner, Schillers, and Hinterdorfer

4. Remove the vials from the desiccator and rinse the desiccator with argon for 5–10  min. Store the APTES-coated tips for 2 days under argon before further use. 5. Dissolve the heterobifunctional cross-linker NHS–PEG– aldehyde in chloroform (3.3 mg/mL). Transfer the solution into a small reaction chamber, add the aminofunctionalized cantilevers, and start the reaction by adding 10 mL of TEA. 6. After 60 min, wash the tips with chloroform and dry them in a gentle stream of inert gas. 7. Place the aldehyde terminated tips on a clean surface (e.g., Parafilm™) and add 100 mL of the protein solution (typically 0.1–3 mg/mL protein in PBS buffer) on the cantilevers. 8. Add 2 mL 1 M NaCNBH3 and allow to react for 110 min. 9. Add 5 mL 1 M ethanolamine and allow to react for 10 min. 10. Wash the tips with PBS and store in PBS at 4°C. 3.3.2. Parameters for Recognition Imaging

When performing TREC experiments, ligand-functionalized tips have to be approached to the surface followed by the oscillation of the tip close to the sample surface. For successful measurements, the working amplitude has to be adjusted in a way that the upper part of the oscillation amplitude is damped (Fig. 7). Details for setting up the optimal imaging conditions can be found in Preiner et al. (47). Finally, the specificity of the interaction has to be proven. This can be done in two ways. By adding free ligands to the sample, the surface-embedded receptors form a receptor–ligand complex resulting in a hindrance of the tip–ligand surface–receptor ­complex  formation (“surface block”). Adding free receptors, on the other hand, passivates the AFM tip to complexation of free receptors with the tip-bound ligand (“tip block”) (see Note 5). For a more detailed explanation of TREC, other literature is recommended (29, 31, 48, 49). In the following, the necessary steps for successful TREC imaging are explained (see Fig. 9).

Fig.  9. TREC working principle: A ligand-functionalized AFM tip is oscillated over the sample surface. The lower part of the amplitude is used for driving the AFM feedback loop, resulting in the topography image, whereas the upper part is affected by molecular recognition, yielding a simultaneously acquired recognition image.

Normal and Pathological Erythrocytes Studied by Atomic Force Microscopy

237

1. Approach the functionalized cantilever to the sample surface (see Note 6). 2. Adjust the amplitude for TREC imaging. When using the typical PEG-18 cross-linker, start with a free amplitude of ~20 nm and adjust the set point amplitude to a value close to the free amplitude. Decrease the free amplitude until recognition signals are detected. If no recognition signals are detected, even when using various amplitudes, replace the cantilever. 3. Image with a scanning velocity less than 3 mm/s (in the fast scan axis). 4. Specificity proof experiment: use either a tip or a surface “block experiment” (see Note 5). All specifity proof experiments should performed at the same conditions as the unblocked recognition imaging experiments (see Fig. 10) (1) Tip block. After performing a successful TREC experiment, incubate the ligand-functionalized tip in a solution containing the corresponding receptor. Remove the excess of unbound receptor molecules by extensive washing with measuring buffer (2) Surface block. Alternatively, add a solution of free ligands to the immobilized RBC surface and incubate for 30–60 min. Wash the membrane with measuring buffer to remove unbound ligands. The following steps are optional to estimate the CFTR ­density using nanocrystalline fluorophores quantum dots: 5. To localize CFTR, incubate the isolated membranes at 4°C overnight with a mouse monoclonal antibody against a C-terminal epitope of CFTR (Mab25031, R&D Systems Inc., Minneapolis, MN) diluted to 8 mg/mL in PBS containing 3% BSA. 6. Wash the samples five times with 1,000 mL of 3% BSA in PBS for 5 min. 7. Incubate at RT for 1 h with anti-mouse antibody, labeled with the nanocrystalline fluorophores Quantum Dots (1102-1, Quantum Dot Corp., Hayward, CA) diluted to 10 nM in PBS with 3% BSA). 8. Wash the sample three times with 1,000 mL of PBS with 3% BSA for 5 min. 9. Wash the sample two times with 1,000 mL of PBS (without BSA) for 5 min. 10. Fix the membrane patches with 0.5% glutaraldehyde in PBS for 40 min.

238

Ebner, Schillers, and Hinterdorfer

Fig. 10. Topography and recognition images of isolated erythrocyte membranes. TREC imaging topography of a non-CF (a) and of a CF (d) erythrocyte membrane. Dark spots in the recognition images (b) and (e) represent the specific interaction sites between the modified tip (i.e., anti-CFTR antibody tip) and CFTR, corresponding to the same areas as shown in (a) and (d). The CF membrane (e) clearly reveals fewer recognition events compared to the non-CF membrane (b). Blocking the membrane of non-CF (c) and CF (f) erythrocytes with free anti-CFTR antibody results in the disappearance of the recognition signals (block efficiency >90%), confirming the specificity of recognition. Scale bar is 200 nm, z scale 80 nm. Taken from ref. 28.

11. Wash the samples with H2O three times to remove buffer salts 12. Dry the samples in air.

4. Notes 1. Aminofunctionalization: Two aminofunctionalization techniques are presented here. The vapor phase deposition of APTES is preferable for the tip coating. Although this coating requires accurate handling, it prevents possible damage to the magnetic coating of cantilevers used for TREC. In contrast, for the first step in the immobilization protocol of whole RBCs, the simpler ethanolamine procedure is sufficient but can also be substituted with the APTES vapor phase deposition.

Normal and Pathological Erythrocytes Studied by Atomic Force Microscopy

239

2. The stability of aminofunctionalized glass surfaces significantly depends on the storage conditions. When stored under argon, the critical oxidation of the amino groups can be avoided and the slides (or sheets) can be stored over weeks. 3. TREC is a sophisticated method allowing investigation of receptor distributions under near physiological conditions, which can only be briefly explained within this book chapter. For more details, please find literature focused on this (29, 31, 48, 49). 4. Tip chemistry. The shown aldehyde–PEG–NHS cross-linker is well suited to couple ligands like proteins via their lysines (using the primary amine group). In addition, an ever increasing number of other heterobifunctional cross-linker exists, using different coupling strategies. The most prominent ones are the site-directed coupling via a His6-tag (50) or the covalent coupling of a thiolated ligand via PDP–PEG–NHS (46). More possibilities and detailed information are given in ref. 51. 5. Specificity proof experiments. In general, the interaction between a tip-bound ligand and a surface-bound receptor can be hindered (blocked) in two ways. Either the tip-bound ligand gets inactivated due to binding of free receptor molecules forming a tip–ligand–receptor complex on the tip (=“tip block”), or the surface-bound receptor gets blocked by addition of free ligand molecules resulting in a surface–receptor– ligand complex. The terminus block in this context means the inactivation of ligands or receptor as specificity proof. 6. The initial approach with a functionalized tip is very critical, since hard contact at the onset can destroy the functional unit on the tip. To overcome this, the approaching velocity should be set very low and the feedback parameters high. References 1. Dulinska, I., Targosz, M., Strojny, W., Lekka, M., Czuba, P., Balwierz, W. & Szymonski, M. (2006). Stiffness of normal and pathological erythrocytes studied by means of atomic force microscopy. Journal Of Biochemical And Biophysical Methods 66, 1–11. 2. Zachee, P., Boogaerts, M. A., Hellemans, L. & Snauwaert, J. (1992). Adverse Role Of The Spleen In Hereditary Spherocytosis - Evidence By The Use Of The Atomic Force Microscope. British Journal Of Haematology 80, 264–265. 3. Kamruzzahan, A. S. M., Kienberger, F., Stroh, C. M., Berg, J., Huss, R., Ebner, A., Zhu, R., Rankl, C., Gruber, H. J. & Hinterdorfer et, a. (2004). Imaging morphological details and pathological differences of red blood cells

using tapping-mode AFM. Biological Chemistry 385, 955–960. 4. Wu, Y. Z., Hu, Y., Cai, J. Y., Ma, S. Y., Wang, X. P., Chen, Y. & Pan, Y. L. (2009). Timedependent surface adhesive force and morphology of RBC measured by AFM. Micron 40, 359–364. 5. Ho, M. S., Kuo, F. J., Lee, Y. S. & Cheng, C. M. (2007). Atomic force microscopic observation of surface-supported human erythrocytes. Applied Physics Letters 91. 6. Bremmell, K. E., Evans, A. & Prestidge, C. A. (2006). Deformation and nano-rheology of red blood cells: An AFM investigation. Colloids And Surfaces B-Biointerfaces 50, 43–48.

240

Ebner, Schillers, and Hinterdorfer

7. Strasser, S., Zink, A., Kada, G., Hinterdorfer, P., Peschel, O., Heckl, W. M., Nerlich, A. G. & Thalhammer, S. (2007). Age determination of blood spots in forensic medicine by force spectroscopy. Forensic Science International 170, 8–14. 8. Koter, M., Franiak, I., Strychalska, K., Broncel, M. & Chojnowska-Jezierska, J. (2004). Damage to the structure of erythrocyte plasma membranes in patients with type-2 hypercholesterolemia. International Journal of Biochemistry and Cell Biology 36, 205–215. 9. Belokoneva, O., Villegas, E., Corzo, G., Dai, L. & Nakajima, T. (2003). The hemolytic activity of six arachnid cationic peptides is affected by the phosphatidylcholine-tosphingomyelin ratio in lipid bilayers. BBABiomembranes 1617, 22–30. 10. de Gómez Dumm, N., Giammona, A. & Touceda, L. (2003). Variations in the lipid profile of patients with chronic renal failure treated with pyridoxine. Lipids in Health and Disease 2, 7. 11. Starzyk, D., Korbut, R. & Gryglewski, R. J. (1999). Effects of nitric oxide and prostacyclin on deformability and aggregability of red blood cells of rats ex vivo and in vitro. J Physiol Pharmacol 50, 629–37. 12. Sandhagen, B. (1999). Red cell fluidity in hypertension. Clin Hemorheol Microcirc 21, 179–81. 13. Starzyk, D., Korbut, R. & Gryglewski, R. (1997). The role of nitric oxide in regulation of deformability of red blood cells in acute phase of endotoxaemia in rats. Journal of physiology and pharmacology: an official journal of the Polish Physiological Society 48, 731. 14. Chen, C., Jia, H., Ma, H., Wang, D., Guo, S. & Qu, S. (1999). Rheologic determinant changes of erythrocytes in Binswanger’s disease. Zhonghua yi xue za zhi = Chinese medical journal; Free China ed 62, 76. 15. Wrobel, A., Kaminska, D. & Klinger, M. (2003). 16. Li, A., Seipelt, H., Müller, C., Shi, Y. & Artmann, G. (1999). Effects of salicylic acid derivatives on red blood cell membranes. Pharmacology & toxicology 85, 206–211. 17. Schwiebert, E., Benos, D., Egan, M., Stutts, M. & Guggino, W. (1999). CFTR is a conductance regulator as well as a chloride channel. Physiological reviews 79, 145–166. 18. Welsh, M., Denning, G., Ostedgaard, L. & Anderson, M. (1993). Dysfunction of CFTR bearing the delta F508 mutation. Journal of cell science. Supplement 17, 235. 19. Fuller, C. & Benos, D. (1992). Cftr! American Journal of Physiology- Cell Physiology 263, 267–286.

20. Dupuit, F., Kälin, N., Brezillon, S., Hinnrasky, J., Tümmler, B. & Puchelle, E. (1995). CFTR and differentiation markers expression in nonCF and delta F 508 homozygous CF nasal epithelium. Journal Of Clinical Investigation 96, 1601. 21. Kälin, N., Claaß, A., Sommer, M., Puchelle, E. & Tümmler, B. (1999). F508 CFTR protein expression in tissues from patients with cystic fibrosis. Journal Of Clinical Investigation 103, 1379–1389. 22. Sterling Jr., K., Shah, S., Kim, R., Johnston, N., Salikhova, A. & Abraham, E. (2004). Cystic fibrosis transmembrane conductance regulator in human and mouse red blood cell membranes and its interaction with ecto-apyrase. Journal Of Cellular Biochemistry 91. 23. Verloo, P., Kocken, C., Van der Wel, A., Tilly, B., Hogema, B., Sinaasappel, M., Thomas, A. & De Jonge, H. (2004). Plasmodium falciparum-activated chloride channels are defective in erythrocytes from cystic fibrosis patients. Journal Of Biological Chemistry 279, 10316. 24. Sprague, R., Ellsworth, M., Stephenson, A., Kleinhenz, M. & Lonigro, A. (1998). Deformation-induced ATP release from red blood cells requires CFTR activity. American Journal of Physiology- Heart and Circulatory Physiology 275, 1726–1732. 25. Stumpf, A., Almaca, J., Kunzelmann, K., Wenners-Epping, K., Huber, S., Haberle, J., Falk, S., Duebbers, A., Walte, M. & Oberleithner, H. (2006). IADS, a decomposition product of DIDS activates a cation conductance in Xenopus oocytes and human erythrocytes: new compound for the diagnosis of cystic fibrosis. Cell Physiol Biochem 18, 243–252. 26. Stumpf, A., Wenners-Epping, K., Wälte, M., Lange, T., Koch, H., Häberle, J., Dübbers, A., Falk, S., Kiesel, L. & Nikova, D. (2006). Physiological concept for a blood based CFTR test. Cellular Physiology And Biochemistry 17, 29–36. 27. Schilcher, K., Hinterdorfer, P., Gruber, H. J., Schindler, H. (1997). A non-invasive method for the tight anchoring of cells for scanning force microscopy. Cell Biology International 21, 769–778. 28. Ebner, A., Nikova, D., Lange, T., Haberle, J., Falk, S., Dubbers, A., Bruns, R., Hinterdorfer, P., Oberleithner, H. & Schillers, H. (2008). Determination of CFTR densities in erythrocyte plasma membranes using recognition imaging. Nanotechnology 19. 29. Ebner, A., Kienberger, F., Kada, G., Stroh, C. M., Geretschlager, M., Kamruzzahan, A. S. M., Wildling, L., Johnson, W. T., Ashcroft, B.,

Normal and Pathological Erythrocytes Studied by Atomic Force Microscopy

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

Nelson, J., Lindsay, S. M., Gruber, H. J. & Hinterdorfer, P. (2005). Localization of single avidin-biotin interactions using simultaneous topography and molecular recognition imaging. Chemphyschem 6, 897–900. Stroh, C., Wang, H., Bash, R., Ashcroft, B., Nelson, J., Gruber, H., Lohr, D., Lindsay, S. M. & Hinterdorfer, P. (2004). Single-molecule recognition imaging microscopy. Proceedings Of The National Academy Of Sciences Of The United States Of America 101, 12503–12507. Stroh, C. M., Ebner, A., Geretschlager, M., Freudenthaler, G., Kienberger, F., Kamruzzahan, A. S. M., Smith-Gill, S. J., Gruber, H. J. & Hinterdorfer, P. (2004). Simultaneous topography and recognition imaging using force microscopy. Biophysical Journal 87, 1981–1990. Nowakowski, R., Luckham, P. & Winlove, P. (2001). Imaging erythrocytes under physiological conditions by atomic force microscopy. Biochimica Et Biophysica Acta-Biomembranes 1514, 170–176. Ebner, A., Hinterdorfer, P. & Gruber, H. J. (2007). Comparison of different aminofunctionalization strategies for attachment of single antibodies to AFM cantilevers. Ultramicroscopy 107, 922–927. Salzer, U., Hinterdorfer, P., Hunger, U., Borken, C. & Prohaska, R. (2002). Ca(++)-dependent vesicle release from erythrocytes involves stomatin-specific lipid rafts, synexin (annexin VII), and sorcin. Blood 99, 2569–2577. Haberle, W., Horber, J. K. H. & Binnig, G. (1991). Force Microscopy On Living Cells. Journal Of Vacuum Science & Technology B 9, 1210–1213. Braet, F., Seynaeve, C., De Zanger, R. & Wisse, E. (1998). Imaging surface and submembranous structures with the atomic force microscope: a study on living cancer cells, fibroblasts and macrophages. Journal Of Microscopy 190, 328–338. Rotsch, C. & Radmacher, M. (2000). Druginduced changes of cytoskeletal structure and mechanics in fibroblasts: an atomic force microscopy study. Biophysical Journal 78, 520–535. Schneider, S., Sritharan, K., Geibel, J., Oberleithner, H. & Jena, B. (1997). Surface dynamics in living acinar cells imaged by atomic force microscopy: identification of plasma membrane structures involved in exocytosis, Vol. 94, pp. 316–321. National Acad Sciences. Le Grimellec, C., Lesniewska, E., Cachia, C., Schreiber, J., De Fornel, F. & Goudonnet, J. (1994). Imaging of the membrane surface of MDCK cells by atomic force microscopy. Biophysical Journal 67, 36–41.

241

40. Swihart, A., Mikrut, J., Ketterson, J. & Macdonald, R. (2001). Atomic force microscopy­ of the erythrocyte membrane ­skeleton. Journal Of Microscopy 204, 212–225. 41. Nikova, D., Lange, T., Oberleithner, H., Schillers, H., Ebner, A. & Hinterdorfer, P. (2006). Atomic force microscopy in nanomedicine. Applied scanning probe methods III. Springer, Berlin, 1–27. 42. Oberleithner, H., Schillers, H., Schneider, S. & Henderson, R. (2001). Nanoarchitecture of Plasma membrane visualized with atomic force microscopy. Ion channel localization methods and protocols methods in pharmacology and toxicology. Humana, Totowa, NJ, 405–424. 43. Cooper, G. & Hausman, R. (2000). The cell: a molecular approach, ASM Press Washington, DC. 44. Schillers, H. (2008). Imaging CFTR in its native environment. Pflugers Archiv-European Journal Of Physiology 456, 163–177. 45. Yamashina, S. & Katsumata, O. (2000). Structural analysis of red blood cell membrane with an atomic force microscope. Journal of Electron Microscopy 49, 445–451. 46. Ebner, A., Wildling, L., Kamruzzahan, A. S. M., Rankl, C., Wruss, J., Hahn, C. D., Holzl, M., Zhu, R., Kienberger, F., Blaas, D., Hinterdorfer, P. & Gruber, H. J. (2007). A  New, Simple Method for Linking of Antibodies to Atomic Force Microscopy Tips. Bioconjugate Chem. 18, 1176–1184. 47. Preiner, J., Ebner, A., Chtcheglova, L., Zhu, R. & Hinterdorfer, P. (2009). Simultaneous topography and recognition imaging: physical aspects and optimal imaging conditions. Nanotechnology 20. 48. Hinterdorfer, P. & Reich, Z. (2008). Molecular Recognition Force Microscopy: From Simple Bonds to Complex Energy Landscapes. Nanotribology and Nanomechanics: An Introduction, 279. 49. Preiner, J., Ebner, A., Chtcheglova, L., Zhu, R. & Hinterdorfer, P. (2009). Simultaneous topography and recognition imaging: physical aspects and optimal imaging conditions. Nanotechnology 20, 215103. 50. Kienberger, F., Kada, G., Gruber, H. J., Pastushenko, V., Riener, C., Trieb, M., Knaus, H.-G., Schindler, H. & Hinterdorfer, P. (2000). Recognition force spectroscopy studies of the NTA-His6 bond. Single Mol. 1, 59–65. 51. Ebner, A., Wildling, L., Zhu, R., Rankl, C., Haselgrubler, T., Hinterdorfer, P. & Gruber, H. J. (2008). Functionalization of probe tips and supports for single-molecule recognition force Microscopy. Topics in Current Chemistry 285, 29–76.

wwwwwwwwwwwwww

Chapter 16 The Growth Cones of Living Neurons Probed by the Atomic Force Microscope Davide Ricci, Massimo Grattarola, and Mariateresa Tedesco Abstract A detailed report of experimental findings concerning the use of atomic force microscopy to probe growth cones of chick embryo spinal cord neurons under vital conditions is given. The role played by indentation in the making of images and force-versus-distance curves is critically discussed. As a result, the thickness of growth cone regions is quantitatively estimated. By comparing the obtained images with descriptions given in the literature on the basis of other microscopy techniques, a central (C) region and a peripheral (P) region are identified, characterized by a different thickness and a different structural organization. Moreover, clusters of adhesion molecules are tentatively identified in regions where neuron arborizations were challenged by the atomic force microscope (AFM) tip. Key words: AFM, Indentation, Force maps, Living cells, Arborizations

1. Introduction There is a large body of recent literature describing the use of atomic force microscopy (1) for the study of living cells. These experimental findings clearly indicate that atomic force microscope (AFM) is a very valuable tool for the three-dimensional imaging of flat biological samples strongly adhering to a substrate, with a lateral resolution in between the resolutions of optical and electron microscopy. Moreover, a very relevant feature of AFM is the capability of analyzing local mechanical properties of living cells. The expression “flat biological samples” includes layers of cells, such as epithelia (2, 3), and single cells such as fibroblasts and glial cells (4, 5). The technique, in its present state, seems to be less appropriate for globular structures, such as neuron bodies (6), and for string-like structures, such as neuron arborizations (7, 8).

Pier Carlo Braga and Davide Ricci (eds.), Atomic Force Microscopy in Biomedical Research: Methods and Protocols, Methods in Molecular Biology, vol. 736, DOI 10.1007/978-1-61779-105-5_16, © Springer Science+Business Media, LLC 2011

243

244

Ricci, Grattarola, and Tedesco

On the contrary, neuronal growth cones are subcellular structures that seem to be very appropriate for AFM analysis: they are flat, highly specialized regions that very strongly adhere to the substrate. Moreover, the mechanical properties of these structures (i.e., the cytoskeleton local organization) are of great relevance for understanding the development of neural architectures and therefore, the potential of micromechanical information of the AFM is of particular value. On the basis of these premises, this chapter will be devoted to a detailed report of experimental findings concerning the use of AFM to probe growth cones of chick embryo spinal cord neurons under vital conditions.

2. Materials 2.1. Animals

1. Chick embryos 7–8 days old.

2.2. Chemicals and Reagents

1. Chick embryo extract (Gibco 16460-016). 2. Hanks’ balanced solution (HBSS, Gibco 24020-091). 3. Bovine serum albumin (BSA, Sigma A-7030). 4. Trypsin solution 0.25% (Gibco 25050-014). 5. Trypsin inhibitor (Sigma T-6522). 6. Deoxyribonuclease (DNAse), type I (Sigma D-5025). 7. DMEM-F12 (Gibco 31331-028). 8. Fetal bovine serum (heat inactivated) (Gibco 10108-157). 9. Horse serum (heat inactivated) (Gibco 26050-070). 10. Poly-d-lysine (Sigma P-7280) or poly-l-lysine (Sigma P-9155). 11. Stock supplement solution N-2 (Gibco 17502-048). 12. 5-Fluoro-2¢-deoxyuridine antimitotic agent (Sigma F-0503). 13. Phosphate-buffered saline (PBS, Gibco 14287-080). 14. Glutaraldehyde solution (Sigma G-6257).

2.3. Equipment

1. Atomic force microscope: Park Scientific Instrument Autoprobe CP (Thermomicroscopes, Sunnyvale, CA). 2. Silicon nitride pyramidal tips on cantilevers with 0.01  N/m nominal spring constant (Thermomicroscopes, Sunnyvale, CA). 3. Dissection microscope (Wild-Leitz). 4. Microdissection forceps (FTS Dumon #5 biologie). 5. Forceps (large, small) (FST). 6. Scissors (fine) (FST). 7. Disposable conical tubes (Falcon 2170, 2195 or equivalent).

The Growth Cones of Living Neurons Probed by the Atomic Force Microscope

245

8. Disposable cell culture dishes (100 mm ø Falcon, 35 mm ø Falcon). 9. Phase contrast microscope (Diavert-Leitz). 10. Thermo-controlled water bath (37.5°C). 11. Centrifuge. 12. Glass slides.

3. Methods The methods described below outline (1) the neuron cell culture and sample preparation, (2) the atomic force microscope setup for imaging, (3) the acquisition of force-versus-distance and indentation curves, (4) the results obtained and their interpretation, and (5) a comparison with other techniques. 3.1. Neural Cell Culture and Sample Preparation 3.1.1. Chick Embryo Spinal Cord Neuron Extraction

Spinal cord neurons were obtained through dissection of spinal cords from 8-day chick embryos and plated on treated coverslips. The dissected cords were minced in HBSS and enzymatically dissociated in 0.05% trypsin at 37°C for 25  min, then washed in CMF-HBSS containing 0.3% BSA, 0.005% DNAse (deoxyribonuclease, type I), and 0.025% trypsin inhibitor. After mechanical dissociation, the resulting single cells were suspended in MEMF12 (1:1) supplemented with 5% fetal bovine serum (FBS), 5% inactivated horse serum, and 5% chick embryo extract for plating on culture substrata.

3.1.2. Neural Cell Culture and Sample Preparation for AFM Investigations

To prepare the culture substrata, first glass slides were cut to 20 × 40 mm pieces and then cleaned and sterilized (see Note 1). They were then incubated overnight in a poly-d-lysine solution (5  mg in 50  ml distilled water), rinsed three times in distilled water, and then dried in a sterile hood (see Note 2). Plating was made on the glass slides kept in plastic Petri dishes, and the cultures were incubated at 37°C and 5% CO2 (see Note 3). Two days after plating, the medium was replaced with 96% MEM-F12, 3% horse serum, and 1% stock supplement solution N2. To free cultures from nonneural cells, an antimitotic agent (5-fluoro-2 deoxyuridine, 10−6  M) was added to the culture medium 72 h after plating.

3.1.3. Cell Fixation

For the purpose of comparing results obtained on living cells, fixated cells were also prepared. In this case, after keeping the cells in culture for 4 or 5 days, the medium was removed and cultures were briefly rinsed with PBS (Gibco BRL). The cells were fixed for 20–30 min using 0.8% glutaraldehyde in PBS. Finally, the slides were rinsed twice with PBS and dried.

246

Ricci, Grattarola, and Tedesco

3.2. Atomic Force Microscopy Setup for Imaging 3.2.1. AFM Setup

A Park Scientific Instruments Autoprobe CP (Sunnyvale, CA) atomic force microscope was used, equipped with a scanner tube allowing 100 mm (x, y) maximum scan size and 6 mm (z) excursion (see Note 4). All experiments were performed using cantilevers with 0.01 N/m nominal spring constant (see Note 5) and silicon nitride pyramidal tips (see Note 6). Special care was taken in order to avoid contact between liquids and scanner, which would cause permanent damage to the piezoelectric element and eventually to the high voltage electronics. For this purpose, the top half of the microscope containing the scanner was enclosed in a polyethylene film sheet. This allows the scanner to move freely and does not interfere with the magnetic coupling of the sample holder to the scanner. The cantilever chip was mounted on a chip holder that has a glass window behind the cantilever chip. In order to avoid air bubble formation, before mounting the chip holder into the microscope, the glass and cantilever chip were wetted with buffer solution using a syringe, and a droplet of water was trapped (kept in place by surface tension) between the chip and the glass window. The sample was then taken out of the Petri dish, taking care to keep a film of buffer solution onto the surface. To overcome the difficulties of gluing a wet glass slide to the sample holder metal disk and also to overcome the limitations of the x–y table that has only a 12 × 12 mm range, we used the following method. First, we fixed a whole glass slide with cyanoacrylate glue to the metal sample holder disk, which was then placed on the scanner as usual. Second, we placed Vaseline onto this glass slide and pressed the cell-covered glass slide firmly onto it. This allows to move the sample easily in search of a good area for imaging and also to change it quickly (see Note 7).

3.2.2. Tip to Sample Approach Procedure

The first step is to approach the tip to the sample as usual with the stepper motor until the drop hanging from the cantilever holder assembly meets the liquid covering the sample glass slide. A meniscus is then formed and from this moment, the surface of the sample can be seen through the on-axis optical microscope (see Note 8). The tip-to-sample approach was always performed on a glass area next to the cell to be imaged, and before scanning, the force set point was lowered to a small value (0.5 nN) in order to avoid cell damage.

3.2.3. AFM Settings for Imaging

Force-versus-distance curves prior to and after imaging were recorded routinely for cantilever deflection calibration purposes and for sample stiffness estimation. These curves were transformed into force-versus-indentation plots, using as reference a force-versus-distance curve obtained on glass during the same session. Images were taken by recording two acquisition channels

The Growth Cones of Living Neurons Probed by the Atomic Force Microscope

247

in the AFM simultaneously: the Z-piezo driving voltage and the error signal from the feedback loop. The first signal is proportional to the Z-piezo displacement necessary to maintain the cantilever deflection (force) at the set point during scanning, while the second one records deviations of the cantilever deflections (hence from the set force) from the set point value. In order to obtain imaging with higher spatial frequency resolution, we tuned the feedback loop parameters so that only the average cantilever deflection was kept near the set point value, allowing the system to generate a meaningful image from the error signal channel, which has a wider frequency band (9). Typical scanning speeds were between 13 and 41 mm/s (see Note 9). 3.3. Acquisition of Force-VersusDistance and Indentation Curves 3.3.1. Force-VersusDistance Curves

3.3.2. Force-VersusIndentation Curves

Force-versus-distance curves were obtained by using the standard PSI software, which records the cantilever deflection, while driving the piezo in the z-direction following a triangular wave. The software allowed us to set the wave frequency and to average the force-versus-distance curves obtained consecutively at the same point. The curves corresponding to a given image were stored in a digital file (1,024 points for each force curve) for further processing. The force scale for these curves was calibrated by using, as a reference substrate, the glass the cells adhered to. As the glass did not appreciably indent under the loads applied, from the slope of the linear portion (after tip contact) of the force-versusdistance curve, we derived the conversion factor from the error signal (in mV) to the cantilever deflection (in nm) and hence to the applied force (in nN), through the spring constant K of the cantilever (Force = K∙cantilever deflection, nominal K = 0.01 N/m). This conversion factor depended on the intensity of the laser beam reflected from the backside of the cantilever and on the area of the spot on the photodiode. Therefore, for each series of curves taken in the same session, we left the laser alignment unchanged, and began and finished the experiment by performing a calibration curve on the glass. When pushed against a soft sample, the tip of the AFM will indent the surface, and the shape of the indentation curve (i.e., the relationship between the load applied and the tip penetration) will give information on the stiffness of the sample. The force-versusindentation curves were calculated by using the approach portion of the force-versus-distance curves. The first step was to take a force-versus-distance curve on a naked glass portion of the sample as reference. From this curve, the coefficient of linear relationship between the Z-piezo displacement and cantilever deflection was derived. From each of the force-versus-distance curves taken on the cells, the calibration line was subtracted, thus obtaining the force-versus-indentation curve (see Note 10).

248

Ricci, Grattarola, and Tedesco

3.4. Results Obtained and Their Interpretation 3.4.1. Imaging

Figure 1 is a collage of various images (acquired in error mode) taken on the same growth cone of a spinal cord neuron, adhering to a treated slide just taken out of the incubator. Figure 1a shows a topview rendering of the growth cone. Filamentous cytoskeletal structures are evident in the thick region (arrows). Figure  1b shows the top region of the cone (partially missing in Fig. 1a). Small “dot-like” structures are visible (arrows). A further zoom of the top right corner of the cone is shown in Fig. 1c, showing a meshwork of cytoplasmic structures (arrows). Finally, Fig. 1d shows the image of the growth cone after about 10 min of continuous scanning. The background globular structure on the left (arrow), present in both images, can be used to align the two images. Most of the periphery of the growth cone has clearly retracted.

Fig. 1. Growth cone of a living spinal cord neuron adhering to a polylysine-coated glass slide. (a) Topview rendering of a error mode image. Filamentous cytoskeletal structures are evident in the thick region (arrows). (b) Scan of the top region of the cone (partially missing in a). Small “dot-like” structures can be seen in the thick domain. (c) Zoom of the top right corner of the cone. A meshwork of cytoplasmic structures appears (arrows). (d) Image of the growth cone after about 10 min of continuous scanning. Most of the periphery of the growth cone has retracted.

The Growth Cones of Living Neurons Probed by the Atomic Force Microscope

249

Fig. 2. 3D-shaded rendering of the z-piezo signal image acquired simultaneously with the image in Fig. 1a, with a pictorial representation of the possible “real” versus measured profile on the growth cone.

An increase in the relief of the filamentous structures projecting toward the neurite can be noticed. Figure 2 shows a 3D rendering of the growth cone, as derived from the z-piezo (topographic) image (not shown), taken simultaneously with the image in Fig. 1a. It should be noted that the cone thickness shown in the figure is affected by the indentation of the tip on the neuron. Nevertheless, “true” thickness can be estimated, as described in Subheading 3.4.2. In the 3D image, a thick and a flat region can be tentatively identified, separated by a continuous relief. Figure 3a, b show the growth cone of another neuron belonging to another slide analyzed immediately after taking it out of the incubator. A thick tubular zone is again evident toward the neurite. Careful inspection allows one to detect a surrounding lowcontrast region with flat protrusions (arrows). For comparison, Fig.  3c shows a similar growth cone after fixation. Similar to Fig. 3a, d shows a growth cone from another living neuron where one can identify a thick tubular region surrounded by spiky structures (arrows). Figure 4a shows a small whole neuron with several arborizations. Toward the apical end, most of them seem to be disrupted. Interestingly enough, a “trace” of the borders of the arborizations is evident (Fig.  4b, c). The trace is made of small (150-nm diameter) dot-like structures, which could be identified as clusters of adhesion molecules.

250

Ricci, Grattarola, and Tedesco

Fig. 3. Series of three images of different growth cones where the peripheral region has been detected by the AFM. (a) Growth cone of a living neuron analyzed immediately after taking it out of the incubator. A thick tubular zone is again evident. Careful inspection allows one to detect a surrounding low-contrast region with flat protrusions (arrowheads). Image obtained recording the z-piezo signal. (b) Image of a growth cone after fixation, shown for comparison purposes. Z-piezo signal image. (c) Growth cone from another living neuron where one can identify a thick tubular region, surrounded by spiky structures (arrows). Image obtained by recording the error signal.

3.4.2. Indentation, Topography, and Mechanical Properties

Figure 5 shows a series of representative force-versus-indentation curves acquired upon a growth cone of a living neuron. Curves 1–3 in Fig. 5 were taken moving at steps of 3 mm away from the growth cone edge toward the neurite and represent the typical behavior of a “soft” portion of a living cell (4, 7). The indentation at first shows a parabola-like trend followed by a quasi-vertical slope, at higher applied forces. The first part can be explained with the classical indentation theory of a solid punch into a half space, progressively deviating from such behavior as the glass substrate contribution becomes dominant (10). The quasi-vertical trend with increasing force indicates that the maximum compression of the cell material has been reached, and the indentation limit value attained will give an indication of the cell thickness (7). Let us now compare curves 3 and 4: curve 3 was taken onto an apparent depression of the surface, at a point like the one identified by A in Fig. 2, while curve 4 was recorded onto a stiff portion of the cell surface like the one identified by B. It is evident how in the

The Growth Cones of Living Neurons Probed by the Atomic Force Microscope

251

Fig. 4. (a) A small whole living neuron showing arborizations, image acquired using the error channel signal. The apical end of most of them seem to be disrupted. (b) Higher magnification error signal image of the apical end of an arborization. (c) Error signal and z-piezo signal simultaneous image of the same arborization apical end. The trace is made of small (approximately 150  nm in diameter) dot-like structures that could be identified as clusters of adhesion molecules.

case of curve 3 we are progressively indenting a thick portion of the cell until we reach the glass substrate, while in the second case, after a parabolic behavior for the first 100 nm of indentation, we reach a constant slope corresponding to an elastic spring constant of 0.0045 N/m. This means that after indenting the most external “soft” cell surface, the tip interacts with submembrane structures that exhibit an elastic response (4). At the force value of 0.5 nN that corresponds to the nominal set

252

Ricci, Grattarola, and Tedesco

Fig. 5. Series of representative force-versus-indentation curves acquired upon a growth cone of a living neuron. Curves 1–3 are taken moving away from the growth cone edge toward the neurite at steps of 3 mm. Each curve shows an increase in indentation with the applied force with a parabola-like behavior, until a quasi-vertical trend is reached. Curve 4 was taken next to the point where curve 3 was acquired, but on a protrusion. A totally different trend can be observed, where after an initial parabolic indentation, a linear dependence on the force increase is established.

point used during imaging in the case of curve 3, we have an indentation of 950 nm and for curve 4, 200 nm. We need to keep in mind these features and figures in order to understand the contrast mechanism of both the z-piezo signal and error signal images on such specimens. This means that the deep shallows next to high peaks found in the z-piezo image cannot purely be due to morphological features. In fact, on thick and “soft” locations, indentation can reach the micron range and on stiffer ones, it lowers down to the 100-nm range; the shallows and peaks in the “topographical” image must be essentially due to differences in stiffness of the sub-membrane growth cone structure encountered by the cantilever during scanning. A pictorial representation of the possible “real” versus measured profile on the growth cone is shown in the second half of Fig. 2. Similar effects are found in the error images, where the gray scale levels represent the deviations from the feedback force (cantilever deflection) set point. As the feedback loop has been tuned to give a high contrast in the error image, thus allowing temporary and relatively large deviations from the force set point, on the left hand side (scanning is from left to right) of an upward slope or an increase in stiffness, the pixels will be darker, while on the left hand side, it will be lighter. This can be clearly observed in Fig. 1a. Generally speaking, it is not possible to discriminate between a change in topography and a variation in stifness, unless one has independent knowledge of the properties of the surface. By use of force-versus-indentation curves, it is possible to discriminate the effects and estimate the thickness of the undeformed surface, at least at the point where the curve is taken. Extrapolation to similar areas can be made to have an estimate of numerical values.

The Growth Cones of Living Neurons Probed by the Atomic Force Microscope

253

A feature common to Figs. 1 and 3 is a thick tubular region extending toward the neurite. The thickness of this region as read on the z-piezo image is in the order of 1 mm, to which at least 200 nm must be added to take the indentation into account. 3.4.3. Identification of Growth Cone Regions 3.4.3.1. Lamellipodia

The filaments in evidence in Fig. 1a, d can be easily identified as microtubules. By comparing these with images generated by the other techniques and described in the literature, the thick region can be easily identified as the so-called C-domain. Note that submicrometer-size structures are visible in this region. In Figs. 1–3, this domain is surrounded by a flat area, with thickness in the order of a few hundred nanometers. This could be identified as a lamellipodia-rich P-domain. The flat protrusions shown in Fig. 3a, b can be identified as lamellipodia structures. Their thickness is in the order of 30–60 nm. Irregularities, distortion along the scanning direction, and “islands” of biological material underline the extent of the tip–sample interaction. For comparison, Fig. 3c shows a similar growth cone after fixation. Lamellipodia structures with a smooth profile are now evident. The thickness is now in the 100–200-nm range: the fixation process has affected the membrane stiffness so that only negligible indentation occurs during scanning.

3.4.3.2. Filopodia

The spiny protrusions surrounding the C-domain in Fig. 3d can be identified as filopodia structures. Interestingly enough, these protrusions appear to be made of globular subunits, often arranged in a discontinuous way. These subunits have a diameter of 120– 180 nm and a thickness ranging from 5 nm to 30 nm. They could be identified as clusters of proteins or patches of membrane adhering to the substrate, left after the tip–sample interaction. The distribution of proteins in the filopodia of growth cones is a subject of active research. Filopodia are known to be filled with bundles of actin filaments (11), and the presence of spots of tyrosinephosphorylated proteins has been recently demonstrated at the tips of growth cones by immunofluorescence techniques (12). The formation of focal contacts by the tip of filopodia with the substrate remains an open question, which could be addressed by further investigation of the described structures.

3.4.3.3. Arborizations

Finally, traces of discontinuous biological material are evident in the terminal regions of the arborizations of a whole neuron (Fig. 4). A fixed similar neuron is shown for comparison (Fig. 6). Here, the arborizations are smooth and continuous. The morphology of the biological details is better preserved, but no indications about the adhesion to the substrate can be inferred. On the contrary, we can conclude that the tip–living material interaction does somehow affect the morphology but, at the same time, it gives hints about the organization of the biological structure at the nanometer scale and about the way contact is made with the substrate.

254

Ricci, Grattarola, and Tedesco

Fig. 6. A small whole fixated neuron showing arborization, to compare with the living one imaged in Fig. 4.

3.5. Comparison with Other Techniques

Detailed AFM images of living flat cells, such as glia cells (5, 6), fibroblasts (4, 7) and epithelial cells (13) have been already analyzed in the literature. Low resolution images of whole neurons have also been produced (6). The other available techniques for studying growth cones are as follows: 1. Whole-mount electron microscopy that gives images with detailed information down to the nanometer scale (14), but on dead materials and without thickness quantification. 2. Fluorescence microscopy that allows one to identify cytoskeleton components by immunofluorescence staining. Timelapse analysis of stained (lipid probe Dioc6) living growth cones is also described (15). 3. Video-enhanced DIC imaging that is widely used for generating detailed images of unstained living growth cones. This technique has allowed the identification of two distinct domains: a central, relatively thick, organelle-rich region (C-domain) and a peripheral, thin region devoid of organelles (P-domain) (16, 17). AFM shares with the last technique the capability of imaging unstained samples. Moreover, compared to all the mentioned methods, it is the only one to have the potentiality of giving quantitative information on thickness. On the contrary, to put the last statement in the correct perspective, it should be underlined that any AFM-originated image of a soft sample is the result of a mechanical interaction between the tip and the sample. This implies

The Growth Cones of Living Neurons Probed by the Atomic Force Microscope

255

indentation, and force-versus-distance curves must be utilized to correct the data to obtain “real” thickness. At the same time, a challenge to the adhesion of the living cone to the substrate is exerted. Finally, it is worth mentioning that, in principle, information about viscosity could be also obtained, similar to that in the laser tweezers technique (18), by carefully comparing the forward and retraction portions of the force-versus-distance curves.

4. Notes 1. Cut glass slides are used as the AFM employed for these investigations is a “scanned sample” model that can accommodate only flat samples having a maximum width of 25 mm (width) and a length of about 50  mm. When using “scanned tip” instruments, it is possible to use the AFM directly onto the Petri dish. One advantage of using glass slides is the surface flatness with respect to Petri dishes. 2. Even if it is possible to grow cultures for some cell lines on bare substrates, for AFM investigation, a good adhesion with the substrate is essential, allowing the cells to withstand the lateral forces induced by the tip during scanning. 3. An advantage of using small glass slides is the possibility of preparing several samples in the Petri dish at one time that can be kept in the incubator until just before use. This allows to increase the throughput of one single primary cell line culture both in time and number of observable cells, as usually it is not possible to keep temperature and CO2 control during AFM measurements. Recently research groups have been developing systems that allow to control physiological environmental conditions during AFM imaging. 4. An important feature when using the AFM on living cells is the range available for the Z direction in the scanner, as variations in height of several microns can be found during scanning. Also X and Y range should be several tens of microns. 5. The use of small spring constants avoids damages to the cell surface or even detachment of the cell from the substrate. We have also tried using 0.003 N/m spring constant cantilevers, but the adhesion forces between the tip and sample did not allow achieving good imaging. 6. Standard silicon nitride tips are quite sufficient for good imaging in contact mode on living cells as the visco-elasticity of the cell is the main limiting factor in resolution: in fact, a sharper tip does not improve resolution but can produce damage. Moreover, silicon nitride seems to behave better than silicon or polysilicon with respect to tip-to-sample adhesion.

256

Ricci, Grattarola, and Tedesco

7. A fresh supply of the buffer solution used for the culture, possibly held at 37°C, should be kept at hand, in order to able to add it when required. 8. An on-axis microscope is essential for positioning the tip on the sample in the required position. Either a scanned-sample AFM equipped with a high magnification long focal length microscope, or a scanned tip AFM mounted onto an inverted microscope can be used. The latter obviously opens the possibility to use a wider range of optical microscopy techniques in conjunction with AFM. 9. One of the serious limitations of AFM is the low scanning speeds that have to be used on soft surfaces. A 10 mm × 10 mm image with 512 points per line is typically acquired in 8 min. 10. On the thinner portions of the cells, one observes at first the indentation process and then a constant relationship between load and Z-piezo travel is found. This means that all the cells have been compressed and the glass surface is “reached.” One can use this last linear portion of the force-versus-distance curve to derive the coefficient and obtain the corresponding force-versus-indentation curve. References 1. Binning, G., Quate C. F., and Gerber C. (1986) Atomic force microscope. Phys. Rev. Letters. 56, 930–933. 2. Schoenenberger C.A., and Hoh J. H. (1994) Slow cellular dynamics in MDCK and R5 cells monitored by time-lapse atomic force microscopy. Biophys. J. 67, 929–936. 3. Hoh, J. H., and Schoenenberger C.-A. (1994) Surface morphology and mechanical properties of MDCK monolayers by atomic force microscopy. J. of Cell Science. 107, 1105–1114. 4. Ricci, D., Tedesco M., and Grattarola M. (1997) Mechanical and morphological properties of living 3T6 cells probed via scanning force microscopy. Micros. Res. Tech. 36(3), 165–171. 5. Henderson, E., Haydon P. G., and Sakaguchi D. S. 1992. Actin filaments dynamics in living glial cells imaged by atomic force microscopy. Science. 257, 1944–1946. 6. Parpura, V., Haydon P., and Henderson E. (1993) Three - dimensional imaging of living neurons and glia with the atomic force microscope. J. of Cell Science. 104, 427–432. 7. Ricci, D., and Grattarola M. (1994) Scanning force microscopy on live cultured cells: Imaging and force-versus-distance investigations. J. of Microsc. 176, 254–261. 8. Butt, H.-J., Siedle P., Seifert K., Fendler K., Seeger T., Bamberg E., Weisenhorn A.L.,

Goldie K., and Engel A. (1993) Scan speed limit in atomic force microscopy. J. of Microsc. 169, 75–84. 9. Putman, C.A.J., van der Werf. K.O., de Grooth B.G., van Hulst N.F., Greve J., and Hansma P.K. (1992) A new imaging mode in Atomic Force Microscopy based on the error signal. Proc. SPIE. 1639, 198–204. 10. Sneddon, J.N. (1965) The relation between load and penetration in the axisymmetric Boussinesq problem for a punch of arbitrary profile. Int. J. Eng. Sci. 3, 47–57. 11. Lewis, A.K., and Bridgam P.C. (1992) Nerve growth cone lamellipodia contain two populations of actin filaments that differ in organization and polarity. J. Cell. Biol. 119, 1219–1243. 12. Da-Yu Wu and Golberg D.J. (1993) Regulated tyrosine phosphorylation at the tips of growth cone filopodia. The J. of Cell Biology. 123, 653–664. 13. Hoh, J.H., Sosinsky G.E., Revel J.-P., and Hansma P.K. (1993) Structure of the extracellular surface of the gap junction by atomic force microscopy. Biophys. J. 65, 149–163. 14. Bridgman, P.C., and Dailey M.E. (1989) The organization of myosin and actin in rapid frozen nerve growth cones. J. Cell. Biol. 108, 95–109.

The Growth Cones of Living Neurons Probed by the Atomic Force Microscope 15. Bridgman, P.C. (1991) Functional anatomy of the growth cone in relation to its role in locomotion and neurite assembly. In The nerve growth cone. Letourneau P.C., Kater S.B., and Macagno E.R., editors. Raven Press, New York. 39–53. 16. Gordon-Weeks, P.R., and Mansfield G.S. (1991) Assembly of microtubules in growth cones: the role of microtubule-associated proteins. In The nerve growth cone. Letourneau P.C., Kater S.B., and Macagno E.R., editors. Raven Press, New York. 55–64.

257

17. Goldberg, D.J., Burmeister D.W., and Rivas R.J. (1991) Video microscopic analysis of events in the growth cone underlying axon growth and the regulation of these events by substrate-bound proteins. In The nerve growth cone. Letourneau P.C., Kater S.B., and Macagno E.R., editors. Raven Press, New York. 79–95. 18. Dai, J., and Sheetz M.P. (1995) Mechanical properties of neuronal growth cone membranes studied by tether formation with laser optical tweezers. Biophys. J. 68, 988–996.

wwwwwwwwwwwwww

Chapter 17 Highlights on Ultrastructural Pathology of Human Sperm Narahari V. Joshi, Ibis Cruz, and Jesus A. Osuna Abstract Applications of atomic force microscopy to ultrastructural investigation of human spermatozoa are ­discussed, with particular emphasis to their most common pathological alterations, which are recognized to be associated with male infertility. Morphological alterations can be located in the head, neck piece, and/or in the flagellum. The consequences of these defects on infertility-related topics are examined in the light of aberrations caused in varicocele and in other spermatozoa morphological alterations like globozoospermia, oligoasthenospermia, and in semen from patients with HIV syndrome. Special attention is given to the temperature effects on sperm abnormalities. The application of the present approach to pharmacology, namely, the development of male contraceptive methods, is also referred. Key words: Atomic force microscope, Human sperm, Infertility, Pathological alterations

1. Introduction A conventional technique to investigate ultrastructure is ­transmission electron microscopy (TEM) (1) which permits to examine the materials, biological as well as nonbiological, with a resolution of the order of a few nanometers, even though some instruments have been recently developed with increased power of resolution pushing the limit further up to a fraction of nanometer. This widely used practice undergoes several serious limitations, namely, that TEM cannot be used for living biological systems. Moreover, in the sample preparation and the measurement, processes such as fixating, vacuum creation, staining, osmotic damage, and other deteriorating factors to the biological material are present. In a way, the measurement process itself seriously affects the pathological aspects in which we are mainly interested. Special methods, such as the use of betaine (2), is recommended to preserve spermatozoa for ultrastructural morphology, Pier Carlo Braga and Davide Ricci (eds.), Atomic Force Microscopy in Biomedical Research: Methods and Protocols, Methods in Molecular Biology, vol. 736, DOI 10.1007/978-1-61779-105-5_17, © Springer Science+Business Media, LLC 2011

259

260

Joshi, Cruz, and Osuna

however, its effectiveness is very limited. Applications of artifacts, which are indispensable, create additional defects and also damage or alter the surface completely or partially. These types of defects and irregularities are not only unavoidable but they change, up to a certain extent, the morphology and create doubts about the authenticity of the original data. In addition to these aspects, TEM is not able to provide topographical information which is really needed to examine the pathological aspects of living cells. Besides, TEM cannot provide three-dimensional images at ease and with precision, as measurements are not carried out by a scanning procedure and hence there is a lack of topographical information. These difficulties are overcome by a recently developed technique, namely, atomic force microscopy (AFM) (3, 4). AFM is a scanning probe microscope which consists of a microscale cantilever with a sharp tip (probe) at its end, and it is used to scan a specimen surface with high resolution, more than 1,000× the limit of an optical microscope. The spatial resolution of AFM depends upon the dimensions of the tip, and generally it has atomic resolution of the order of a few nanometer. Recent technology permits to fabricate the tip as small as 1 nm radius, and hence a slight improvement in spatial resolution is observed. As it is a scanning probe microscope, it provides information with precision in three dimensions. Obviously, it is a breakthrough in nanotechnology, and it has an extensive and sizable impact in biological and medical sciences. In biological systems, one encounters very flexible micromolecules and conventional AFM fails to yield useful images or topographical information. For this purpose, Cryo-AFM has been designed which permits to acquire images in liquid nitrogen vapor, and hence it is possible to control the temperature from 77 to 220 K. Cry-AFM is found to be very suitable for bio materials like immunoglobulins, examination of DNA, red blood cells (4), and other biological materials. Recently, temperature control unit is incorporated to AFM (5). Such achievements guarantee that new and useful information is available to examine the temperature dependence on defect formation in human sperm. This issue is addressed in the coming section. On the other hand, measurements at low temperature are not required for several biological entities, where ultrastructural images and topographical information at room temperature are essential for understanding pathological alterations in a living system at nanoscale level. With this view, an AFM is applied to Entamoeba histolytica (6, 7) in their natural environment and useful structural information has been reported. Andrology is not an exception and the applications of AFM to ultrastructural (in the range of nanometer) features and analysis of human sperm, both normal and pathological, have become a reality (8–12). In order to look at alterations caused in human sperm by a disorder of any kind, it is necessary to have complete and reliable

Highlights on Ultrastructural Pathology of Human Sperm

261

information of healthy human sperm in its physiological environment­, and a detailed topographical data have been reported earlier by Joshi et al. (8). This has been used to compare the alterations caused in the pathology of human sperm and to evaluate alterations caused by different deleterious conditions. Morphological alterations in human sperm are directly related with infertility, which is the inability of a couple to conceive after 1 year of regular unprotected intercourse. The prevalence of infertility has been estimated in about 15% in all couples of reproductive age (13). Many couples that come to an infertility clinic do not present apparently the causes related to infertility, and disturbances in fertility may remain occult for years, until a couple are interested in having a child. It is estimated that about 40–50% of the infertility cases is attributed to “male factor” (14–16). According to etiology, male infertility has been categorized in three groups: pretesticular causes, testicular causes (spermatogenesis factors), and post-testicular causes (17). Some aspects of infertility (treatable and untreatable) are certainly related with spermatozoon morphology. Normal testicular function is expressed in two different ways: sex hormone synthesis and spermatozoa production. Sperma­ togenesis is a very complex process in which different ana­ tomic and functional structures are involved: the central nervous system (the hypothalamo-pituitary axis) and, at the periphery, the testes (17). Spermatogenesis disruption is reflected in several pathological conditions. Primary testicular alterations are the most common causes of spermatogenesis absence or derangement, in which a variety of factors are involved: mainly genetic causes and chromosome disorders. Scrotal temperature is a crucial factor for testicular heat regulation, maintaining its temperature in the physiological range required for normal spermatogenesis. High testicular temperatures have been related to male infertility, probably altering sperm morphology (18). Some of these sperm morphological changes can be examined by TEM or by AFM, and could be an important source of information for the study of male infertility.

2. Experimental 2.1. Sperm Preparation

The initial evaluation of a man from an infertile couple should include a thorough clinical evaluation, with particular emphasis in his reproductive history, and at the same time semen analysis should be performed. It is a generally accepted criterion that semen analysis is the cornerstone of the laboratory evaluation of the infertile male; it provides information on semen volume as well as sperm concentration, motility, and morphology. Subjects from

262

Joshi, Cruz, and Osuna

our studies received standard written instructions for semen ­collection. Semen was obtained by masturbation, after 3–5 days of sexual abstinence, and the sample was delivered to the laboratory within 1 h of collection, avoiding exposure to extremes temperatures. Semen analysis was performed according to the manual published by the World Health Organization (WHO) (19). Samples were prepared for routine semen analysis, including estimation of sperm concentration, assessment of sperm morphology, hypoosmotic swelling test, and leukocytes quantification by the peroxidase technique. Spermatozoa separation was carried out by the “swim-up” technique. The conventional “culture medium” used in our experiments was Hanks 199 obtained from GIBCO laboratory, USA. It is observed that spermatozoa were maintained in healthy conditions (or without alteration) for 6–8  h, time enough to carry out AFM measurements. In some special cases, like sperm from HIV patients, a written permission was obtained and confidentiality about the identity was guaranteed. Selected sperms are placed on conventional microscopic slides. AFM images are recorded in normal atmosphere (not in vacuum) and saline environments without damaging the cell membrane. The present technique, without doubt, preserves the small cellular structure and also retains cytoplasmatic structure which is generally damaged in TEM measurement procedure. 2.2. Examination by Optical Microscopy

Sperms were examined by conventional optical microscopy (400×) for gross defects, such as bending, coiling, and relative length of flagellum. Very often, the form and the shape of the head can be viewed. Therefore, a preliminary study is generally carried out with an optical microscope. It is possible to detect some likely abnormalities by optical microscopy, even though the quantification can be carried out only by AFM.

2.3. AFM Operations and Precautions in the Measurement Techniques

AFM is a high-resolution instrument which permits to obtain images and examine ultrastructural details with precision in three dimensions in living and nonliving specimens. However, certain precautions are required to obtain good and reliable results: 1. Nano structure grown and/or grooved on semiconductor wafers like Si or GaAs is a good example of nonliving specimen where three-dimensional ultrastructure is found to be very useful. In this case, the sample is hard enough and the use of contact mode is very functional. However, in case of biological objects (living or nonliving) the use of contact mode is totally inadequate as the outer membrane is delicate and soft. Moreover, dragging the tip on the sample just damages the membrane and the sample. Therefore, the most important recommendation for non biological objects is measurements in contact mode are just not possible and its use should be avoided completely.

Highlights on Ultrastructural Pathology of Human Sperm

263

2. The scanning rate should be adjusted to match the software and hardware of a given equipment. It is always recommended to try a couple of scanning rates before the final imaging. In most cases, 1 line/s was found to be an adequate rate and the images reveal all the details. The speed higher than 3 lines/s is not recommended, unless the software is specially designed for this purpose. 3. In biological samples, noncontact mode, commonly known as tapping mode, should be essentially employed. In this mode, three parameters for measurements procedure should be considered. The first is the stiffness of the cantilever. If the cantilever is not stiff enough, it will stick to the water molecules or liquid environment. The recommended value is about 1 nN/m. A value lower than this might be unsuitable for reliable measurements if the environment contains lipids or sticky solutions as it is in the present case. 4. The magnitude of the applied force is a critical parameter and it approximately varies from 0.2 to 0.8  nN/m depending upon the type of membrane and its elasticity. For example, for E. histolytica (6, 7) the optimum force is 0.8 nN/m, meanwhile for a healthy spermatozoon it is 0.2  nN/m. A slight variation is expected from one sperm to another and it is generally determined by experiment. It is worth to mention that when the sperm is deteriorated (e.g., incubated at 40°C), a lower force is recommended. 5. The resonating frequency of the vibrating tip is also a key parameter for high-resolution imaging, particularly when the sample is soft and the structure below the membrane is not compact. This is the present situation, particularly near the neck pieces. The vibrating frequency used for the spermatozoon lies between 170 and 190 kHz. For soft parts, sometimes, a separate scanning is required with lower frequency. High-resolution good images can be obtained with the help of a delicate balance between the force and the vibrating frequency (8). 2.4. Data Collection System and Imaging

Three-dimensional images were obtained by image processing software which was provided with the equipment. These images offer information about alterations caused in spermatozoon’s morphology of a specific disorder, and could help to correlate the cause for infertility; however, they do not permit to quantify the alterations, and the only way is to make them available with the help of longitudinal and transverse profiles of their structures (e.g., the head, the neck, or the tail). Therefore, we have supplied these profiles for healthy and pathological sperms. A typical spermatozoon is a stripped-down cell, which has a strong flagellum to propel it through the seminal fluid (see Fig. 1). The mature spermatozoon is devoid of cytoplasmic organelles

264

Joshi, Cruz, and Osuna

Fig. 1. Image obtained by AFM of a healthy spermatozoon. The height scale shown on the right hand side provides only a rough idea and hence the topography is a must. Reproduced from Archives of Andrology, Joshi et al. (8).

such as ribosomes, endoplasma reticulum, and Golgi apparatus. Meanwhile, it contains numerous mitochondria strategically located, particularly where they are needed to efficiently propel the flagellum (20). Spermatozoa consist of two functional distinct regions: the head, and the tail or flagellum. The neck region is considered as the beginning of the tail. However, morphologically the neck has distinct features, and therefore in the present work, we have classified it into three different sections, namely, the head, neck piece, and the flagellum. Each part has specific functions and pathological alterations or abnormalities, and each of them has a definite modified response. For example, if the tail has some defect, its motility will have an irregular response. We, therefore, believe that it is more useful if pathological irregularities are presented by subsection head, neck piece, and tail region.

3. Sperm Anatomy 3.1. Head

The head of the spermatozoon is occupied mostly by the nucleus and the acrosome, with lesser amounts of cytoskeletal structures and cytoplasm. The nucleus and the acrosome are usually symmetrical structures. Human spermatozoa are uniform as far as the size and the shape properties are concerned. Variations in them may be less than 5%. Therefore, it is just appropriate to investigate morphological details of healthy sperms and examine the alterations caused in them. Deviation, if it is observed, is due to external factors, which may change their morphology, as in some diseases like varicocele. The temperature at which the sperms are formed and maintained is also important. The spermatozoa nuclei are haploid, containing only one member of each chromosome pair. The chromatin, which is the male DNA contribution in the fertilization process, becomes highly condensed at the latter stages

Highlights on Ultrastructural Pathology of Human Sperm

265

of the spermatogenesis; at these stages, the histones, high in lysine content compared to arginine, are replaced by protamines, highly basic proteins rich in arginine and cysteine. The m-RNA encoding for mouse protamines are synthesized in spermatids, indicating that protamines are products of the haploid genome. The highly condensed protamine–DNA complex is stabilized by disulfide bonds between the protamines (21, 22). Therefore, the major nuclear proteins associated with sperm DNA are protamines. An increase in the height of the sperm indicates that chromatin is not in a compact form, and protamines and arginine are not in the required balance proportion. Defects in the region of head or neck pieces are very common, and they lack of specificity for any pathological condition of the testes. They are usually found in patients with severe low sperm counts (oligozoospermia), but also in patients with normal sperm count. The causes for oligozoospermia remain elusive, and they are the expression of damage of germinal epithelium. Defects of the head and neck pieces are included in the teratozoospermia category (23). Systemic illness, chronic fever, frequent exposure to high temperatures, cytotoxic chemotherapy and radiation therapy, drugs and toxins, and a large list of other factors may be related to aberrations in the spermatozoa head and neck pieces. For an evaluation of any kind of anomalous morphological structure, it is necessary to know the standard features. We are, therefore, presenting here an image with transverse and longitudinal topographical details of healthy spermatozoon in its natural environment. The pathological features are compared with these details. Figure  1 shows the image of a healthy spermatozoon. Figure 2a–c are head image, the longitudinal, and the transversal topography, respectively. A table of the magnitudes of lengths and heights appear at the bottom of each figure (Fig.  2b, c). These values are given with a precision of 0.1 nm. This is a measurement accuracy of the system used. However, as mentioned earlier, these values might vary by about 5% from person to person, and sperm to sperm. This aspect should be considered while comparing normal values with the pathological ones. The height measurements, sometimes, are provided with the help of a histogram, however, they are not suitable to quantify and hence their use is avoided in the present work. Figure 2b reveals that the longitudinal topography discloses some structural information like acrosome cap and a well-marked boundary between head and neck which could be an important feature. The length of the acrosome cap is about 1,880  nm and approximately the same value is reported by Kumar et al. (24, 25). The transverse profile, in this case, does not throw much light on the structural details and generally its form is symmetric if measured perpendicular to the length. For a pathological study, the longitudinal profile is

266

Joshi, Cruz, and Osuna

Fig.  2. (a) shows the image of the head of the spermatozoon. (b) and (c) show longitudinal and transverse profiles. Reproduced from Archives of Andrology, Joshi et al. (8).

more significant for the analysis of the head, and we take it into account for further discussion. A morphological examination of the head necessarily should focus the attention on the acrosomal region. It is a large secretory granule or vesicle that closely surrounds and overlies the anterior end of the nucleus. The acrosome originates from Golgi complex in the spermatids and contains hydrolytic enzymes necessary for the sperm to fuse and penetrate the egg’s outer coat to achieve fertilization. When a sperm comes in contact with an egg, the contents of the acrosome vesicle are released by exocytosis in the so-called acrosomal reaction (21, 23). This is a critical moment in the fertilization process. Specific proteins are also released which are essential to a normal sperm–ovum interaction. This membranous structure sits as a cap over the nucleus in the anterior region of the sperm head. The morphology of this cover cap and the alterations caused on it can be viewed with AMF (see Fig. 2b), and this is one of our concerns in the present investigation as it is directly related with infertility in male.

Highlights on Ultrastructural Pathology of Human Sperm

267

The other important component of the head of the sperm is chromatin which can also be viewed with the help of topographical configurations, as shown in Fig. 2b, c. These figures permit to calculate the volume of the desired area. The height of the head of a healthy sperm is about 750 nm, and Fig. 2b reveals a clear distinction between the acrosome and the chromatin regions. By this or by using an equivalent method, it is possible to measure the volume of the desired area. Sperm chromatin contains DNA material and it is located at the central part of the head of the sperm. Considering its importance, Lee et al. (26) have investigated chromatin volumes in human sperm nuclei for seven of nine classes of head shape abnormalities. On the basis of experimental data, it is concluded that the nuclear volume is essentially identical even though the shape and the projected area have substantially different forms. This study indicates that different head forms in fertile men are not originated from sperm chromatin or DNA content of the sperm nucleus, and the way in which chromatin is organized. Such information is valuable to compare the results obtained by TEM and with other observational tools. 3.1.1. Varicocele

Varicocele is the most commonly identified and correctable cause of male infertility. It is about three times more common in infertile men than in men of proven fertility. Varicocele is found predominantly on the left scrotal sac. The seminal profile seen in infertile men with varicocele was first described by MacLeod in 1965 (27). Oligozoospermia of varying degrees was observed, but the striking findings were marked impairment of sperm motility as well as a noticeable increased of immature and tapered forms in the seminal fluid (stress pattern). Nevertheless, the semen quality in men with varicocele varies from normal to azoospermia. A study from the WHO showed on 9.043 men that the incidence of varicocele was 25% on men with abnormal semen, and 11.7% on men with normal semen (23). Zucchi et al. (28) found that 18 out of 43 patients had oligoasthenospermia, 20 had asthenospermia, and 5 had normal values. Several theories have been postulated for the effect of varicocele on the testicular function, including vascular stasis, back pressure, reflux of renal or adrenal substances into the pampiniform plexus, and interference with the heat exchange mechanism of the pampiniform plexus; nevertheless, the precise mechanisms have not been elucidated (29). Recently, Mieusset et al. (30) studying a group of 13 fertile men have reached to the conclusion that clothing and seated with legs crossed position has a persisting thermogenic effect, stressing the fact that the scrotum regulatory system plays a fundamental role to keep the testes at a temperature lower than the rest of the body. The temperature effect seems to be very critical, and for this purpose a separate investigation was carried out to examine the effect of temperature on sperm morphology and it is discussed separately.

268

Joshi, Cruz, and Osuna

The above discussion confirms that varicocele is one of the predominant causes for male infertility and it has several implications, since it causes different types of abnormalities not only in head, but also in the neck piece and flagellum region. Therefore, we have given importance to the morphological investigation of sperm of a patient suffering from varicocele. By using AFM technique, we have demonstrated earlier that head abnormalities were predominant in sperm obtained from varicocele patients (9); probably, the difference is not remarkable between varicocele grade I and grade III. (11). In semen from patients with varicocele, the morphological and topographical variations in the head form of the spermatozoon can be appreciated with the help of Fig. 3. In normal sperm in the longitudinal topography, the acrosome cap is clearly visible (see Fig. 2b), and it is about 1,800 nm length while it is altered for sperm obtained from varicocele patients grade I which is about 1,450  nm (not well marked), and it is nearly absent for the sperms obtained from varicocele grade III. This is a clear indication that some spermatozoa of varicocele grade III have practically no possibility in the fertilizing process. The variation in height is also noticeable. In a normal cell it is about 750 nm; meanwhile, in varicocele patients (both grade I and III) the height is above 1,050 nm. In case of varicocele grade I, the increase in height is abrupt; meanwhile, in varicocele III it is rather continuous. This also indicates that the damage is not localized but spread over the entire region. An increase in height indicates that chromatin (or DNA) material is not sufficiently condensed or compact, and it may affect the fertilizing process. There is no significant variation in length and it is

Fig. 3. Images of the head along with the longitudinal and the transverse profiles of the sperm of a patient with varicocele I and varicocele III. Reproduced from Archives of Andrology, Joshi et al. (11).

Highlights on Ultrastructural Pathology of Human Sperm

269

of the order of 5,000  nm for varicocele grade I. In case of ­varicocele grade III, the length is difficult to determine because the depression corresponding to the neck is not detected, and hence it is not possible to pinpoint the exact location where the neck region starts. In this case, much information is obtained from the longitudinal topography. 3.1.2. Oligoasthenoter­ato­ zoospermia

Oligoasthenoteratozoospermia (OAT) signifies the alterations of three variables of the seminal fluid; characterized by abnormal low sperm concentration, low motility, and also abnormal sperm morphology. Different factors may be responsible for the OAT condition, including hypothalamic pituitary dysfunction, causing low luteinizing hormone (LH), follicle-stimulating hormone (FSH) syntheses, and inadequate stimulation of spermatogenesis; in cases of Leydig cell and germinal cell dysfunction, as may be the case of varicocele; and in the focal Sertoly-Cell-Only (SCO) syndrome (31). Because of the above reasons, understanding the morphological defects in sperms of OAT patients is very important. The morphological details of the sperm obtained from OAT patients were examined carefully as alterations are expected for the reasons mentioned above (see Fig. 4). The general trend is similar to that obtained from varicocele patients. The height of the head is increased significantly up to approximately 1,425 nm instead of about 750 nm. Meanwhile the length is not altered, it has the value approximately 5,000–5,200 nm (in this case also the beginning of the neck is not located and hence the value is not precise). Therefore, the increase in height appears more noticeable. In globozoospermic sperm, which is a severe form of teratozoospermia, probably genetic nature is characterized by round nuclei and the absence of acrosome, as a consequence the Golgi apparatus does not evolve into the acrosome (32). In globozoospermic spermatozoon, the situation is a little different (9) as compared with OAT. The height of the sperm lies between two extreme values, and it is about 1,300 nm, but the common factor is that the layer of acrosome is absent in both types. In this case also, the neck structure is not visible, which is expected to be located at 5,200 nm, but the most curious aspects is that the variation in the surface is smooth and continuous up to approximately 7,000 nm. This means that in the region of the head, neck piece, and the flagellum are merged without showing any structural variation, indicating that there are serious alterations in the sperm of patients with globozoospermia.

3.1.3. HIV Syndrome

Defects in the region of the head or the neck pieces are very common, and they lack specificity for any pathological condition of the testes. They are usually found not only in patients with severe oligozoospermia, but also in patients with normal sperm count.

270

Joshi, Cruz, and Osuna

Fig. 4. The image and the topography of the sperm of a patient having OAT. Reproduced from Fertility and Sterility, Joshi et al. (9).

Systemic illness, fever, and other factors previously mentioned may be related to changes in the spermatozoa head and neck pieces. AFM is an adequate research tool which may contribute to a precise knowledge of teratozoospermia and its implications in male infertility. Defects in the region of head or neck pieces are very common in sperms from other conditions, for example patients affected by HIV and diseases characterized by high and prolonged fever. In case of HIV patients, the situation is more interesting and complex as some of the patients are going through antiretroviral therapy (HAART) treatment for survival, and therefore AFM investigation had been carried out in both cases, patients with and without HAART treatment, and the results are very valuable (see Figs. 5 and 6).

Highlights on Ultrastructural Pathology of Human Sperm

271

Fig. 5. The image and the topography of the head of the sperm of a patient obtained from HIV patient. Reproduced from Archives of Andrology, Barboza et al. (12).

In the case of HIV patients, the height of the head of the sperm is slightly reduced (approximately 600 nm) but the length of the head is increased. The acrosome layer is not well marked or it might be very thin. After going through HAART treatment, these values try to normalize. The height of the spermatozoon approaches to the normal value. However, in this case the evaluation and conclusions are more difficult because the situation varies from patient to patient as they are undergoing different combination of drugs with different doses. The observation reveals only a tendency. Here, it is worth to mention that AFM does show the presence of HI-virus on the head and neck regions (12). This is the direct evidence of the active role of the spermatozoon in HI-virus transportation process. 3.1.4. Temperature Effects

From the aspects discussed in varicocele patients, it is clear that the thermogenic effect is very significant, not only from the point of view of sperm concentration, but also for the temperature at which the sperm is formed and maintained. High temperatures affect motility as well as sperm morphology. Healthiness of the

272

Joshi, Cruz, and Osuna

Fig. 6. The image and the topography of the head of the sperm of a HIV patient who is going through HAART. Reproduced from Archives of Andrology, Barboza et al. (12).

sperm is very sensitive to temperature. Moreover, from the ­clinical practice, it is well known that patients suffering from fever for a long time (chronic high temperature for days or weeks) show problems related with infertility, even though they are reversible. It is also known that the effect of prolonged sauna exposure on the testes is negative (33). In this case, the environment of the testes is inadequate, and hence spermatozoa are expected to have modification and malfunctioning (34). With the above consideration, a morphological investigation has been carried out by incubating sperms at 36.5, 38, and 40°C for 24 h (unpublished work), and the results are shown in Figs. 7 and 8. Obviously, the sperm morphology at 36.5°C is normal and coincides with the image and topography as given in Figs. 1 and 2. The difference is visible for the sperm incubated at 38°C, however, it is not drastic. The acrosome layer is visible but its form is slightly

Highlights on Ultrastructural Pathology of Human Sperm

273

Fig. 7. The image and the longitudanl and transeverse the topography of the head and the neck of a sperm incubated at 38°C.

distorted (Fig. 7). The height of the head does not really change but the structure near the neck pieces acquires an imperfect form. This means that DNA is still in a compact form. However, the sperm incubated at 40°C shows dramatic alterations (Fig. 8). The height of the head is increased from the expected value 750– 1,150 nm. The surface is rough and the head has a bulky form. In this case, DNA is not in a compact form, and the bulkiness of the head may cause impediments in the swimming process also. These features, namely, irregularities in the head form, increase in the height, and the absence of the acrosome structure, have been used positively by pharmacologists to control and examine the quality of a given drug toward the male contraceptive (24, 25). By injecting a polyelectrolytic compound (a special drug) in a

274

Joshi, Cruz, and Osuna

Fig. 8. The image and the topography of the head and neck of a sperm incubated at 40°C.

healthy person, the defect in the sperms and the ­alterations caused due to the drug are studied with the help of AFM, and its effectiveness toward a male contraceptive is obtained. The faulty head form revealed by the topography together with the absence of the acrosome layer is the fundamental feature of the pathology of the head of the human sperm. The abrupt and asymmetrical increase in height, which is also a common feature in pathological sperm, has also a remarkable negative effect on the floating and the swimming capacity of the sperm due to the principle of buoyancy, and this becomes as an additional ­impediment in the fertility process.

Highlights on Ultrastructural Pathology of Human Sperm

275

In this study, we have given special importance to the a­ bnormalities in the head region with reference to the sperm chromatin form; if it is intact, compact or not. These features can be viewed with the help of AFM. This is a key aspect in the artificial reproduction process and is referred again. New methods to evaluate the sperm chromatin and particularly sperm DNA fragmentation have been developed in recent years (35). The combination of one of these methods with AFM could be of great value to increase the knowledge about spermatozoa functional capability for fertilization. 3.2. Neck Structure

The neck of the spermatozoon is very often called “the connecting piece” as it is a bridge between the head and the upper part of the flagellum. The main structural components of the connecting piece are the basal plate or capitulum and the segmented columns. Interior in the connecting piece, and distally to the basal plate, a transversely oriented proximal centriole lies between the longitudinal oriented distal centriole and a depression in the capitulum. The connecting piece joins the middle piece distally at the beginning of the mitochondrial sheath (20). The tapping mode of AFM allows imaging of light and dense materials equally well. It is known that the mammalian spermatozoa have significant importance in mitochondrial activity as it deals with ATP, and hence the energy for the motility and its healthiness. Abnormalities in the neck piece are a direct indication that the motility is affected in one or in other form. Neck pieces along with the flagellum are related with the motility, one of the key qualities, of the sperm in the fertility process. Pathological modifications are not only limited to the head structure, but they are also extended to the neck pieces and the flagella regions. Because of the presence of the “neck piece” and the centriole structure, a depression or a discontinuity is created around the neck and this can be visualized with the help of longitudinal and transverse topography (Figs. 9 and 10). In this case, transverse topography reveals significant information like the width and the depth of the depression between the neck pieces. It is observed that the depth is of the order of 30 nm and the width lies between 670 and 690 nm. In case of varicocele, these details are more easily noticeable in the transverse topography than in the longitudinal. Therefore, only the transverse profile is provided. In normal sperm, the depression is clearly observed (Fig. 9) meanwhile in varicocele I, a slight tendency for the depression is observed and in varicocele III, the depression corresponding to the neck structure is totally absent and instead of it a peak is observed (see Fig.  10). This indicates that the deterioration and/damage of the neck pieces are progressive from varicocele grade I to varicocele grade III. The absence of the central structure and the disorganization

276

Joshi, Cruz, and Osuna

Fig. 9. The image of the neck structure along with the longitudinal and transverse topographies, of a healthy sperm. Reproduced from Archives of Andrology, Joshi et al. (8).

around the neck are two noteworthy features. The image of the neck structure also confirms similar conclusions. These defects are not only observed in varicocele, but are also observed in globozoospermia and hence these particular features cannot be used for identification purpose, even though they are useful when varicocele is the outstanding clinical feature accompanied with other clinical findings. In the case of OAT patients, the membrane near the “neck pieces” region is completely damaged (Fig. 11) and an unidentified granular structure is detected. This creates obstructions both in the motility and the energy-related features giving rise to slow, sluggish, and incoherent movements, affecting the sperm fertilizing capability, and therefore this sperm is not properly functional. The effects on the neck pieces are more drastic in case of HIV patients. Figure  6 shows that the neck structure is completely damaged and it is not recovered even after HAART. On the contrary, the region is distorted and bulky patches, probably the agglomeration of protein, are extended. A close look shows that the height of the neck is slightly reduced, but the more remarkable feature is that the depression corresponding to the neck structure is absent.

Highlights on Ultrastructural Pathology of Human Sperm

277

Fig. 10. The transverse profile of the neck of the sperm of a patient with varicocele grade I and grade III. Reproduced from Archives of Andrology, Joshi et al. (11).

Fig. 11. Image of the neck piece of a sperm suffering OAT. Reproduced from Fertility and Sterility, Joshi et al. (9).

The effect of the temperature on the neck pieces is still ­ ramatic and can be seen by Figs. 7 and 8 with the corresponding d longitudinal and transverse profiles. Images of the sperms incubated at 38 and 40°C show very strong damage, an uneven surface and irregular patches. At 38°C, the longitudinal profile reveals the region of the neck pieces, which is defective and such

278

Joshi, Cruz, and Osuna

structure disappears totally for the sperm incubated at 40°C. Moreover, there is a steep fall in the height without indication of the presence of any structure. In short, there is a complete damage of the centriole structure. 3.3. Flagellum

The structural components of the flagellum are the axoneme, the mitochondrial sheath, the outer dense fibers, and the fibrous sheath. The axoneme is composed of a “9 + 2” complex of microtubules which extends up to the end of the flagellum. The main structural components of the axoneme are a circle of nine microtubule doublets, interlinked by nexin and connected to the central sheath of the central pair of microtubules by radial spokes. Protruding from each doublet, there is a row of outer and inner dynein arms, which are the motors for the flagellum to move further in a special fashion (20). These structural details are covered under the surface and they are not possible to be detected by AFM unless the membrane is damaged. In stained human spermatozoon, the tail should be straight, uniform, and thinner than the mid piece, uncoiled and approximately 45  mm long. (WHO Manual). In Fig.  12, a complete

Fig. 12. The image and topographical data of the flagellum of a healthy sperm. Reproduced from Archives of Andrology, Joshi et al. (8).

Highlights on Ultrastructural Pathology of Human Sperm

279

length of the flagellum is not shown. Spermatozoa with abnormal flagella are relatively common in human semen. Fertile men have been reported to have 18% or more sperm with abnormal tails (36). In cases of total immotile spermatozoon (severe astenozoospermia), ovum fertilization can be achieved by the new assisted reproductive technologies like ICSI (36, 37). Defects in sperm motility in infertile men might be originated from axonemal defects, which could be related to the absence of dynein arms; such investigation has been carried out by TEM (37). In healthy sperm, the central channel or nanogroove is a wellorganized part from the end of the neck pieces to the end of the tail. This nanogroove is supposed to run throughout its length in a uniform manner (see Fig. 12) without obstruction or conglomeration of protein patches. The presence and details of the channel are appreciated more in the transverse section than in the longitudinal one. It is obvious that in varicocele grade I, the central canals deep is getting covered with proteins and the plateau is recorded. This process is continued even in varicocele III where instead of a dip, a peak is observed. This confirms that the damage is progressive from varicocele I to varicocele III (see Fig. 13). Considering abnormality in the head and also the flagellum (means motility), the sperm of the patient of varicocele III has serious limitations in the fertilizing process. The pathological alterations are more severe in HIV patients (12). The defective nature of the flagellum is all over the length. Not only the channel is absent but the tail is twisted and there are random patches and irregularity over the entire surface of the tail. These types of aberrations are typical in sperms of HIV patients (12).

Fig. 13. The image and topographical data of the flagellum of sperms from varicocele grade I and grade III. Reproduced from Archives of Andrology, Joshi et al. (11).

280

Joshi, Cruz, and Osuna

The flagellum of the sperm incubated at temperatures 38 and 40°C also show considerable alterations which affect their motility. In this particular case, both the transverse and the longitudinal profiles are useful. The last one provides data about the variation in thickness of the sperm. In normal sperm, the thickness varies in a smooth and undulatory manner meanwhile in the case of the sperm incubated at 38°C, the variation in the thickness is not smooth instead it is uneven. However, the overall structure remains the same with little alterations. The situation becomes worst in the case of the sperm kept warm at 40°C. Where the thickness of the tail varies dramatically and randomly, at certain points it reaches up to 70 nm and suddenly drops to 20 nm. This might be due to the conglomeration of some proteins which are unevenly deposited on the flagellum (see Fig. 14). The sperm incubated at 38°C shows the tendency of the ­presence of the central canal. This is observed by the transverse

Fig. 14. The image and longitudinal topographical data of the flagellum of a sperm incubated at 40°C.

Highlights on Ultrastructural Pathology of Human Sperm

281

profile and the images themselves. In healthy sperm, the groove is well marked from the neck to the end of the tail. This is not so for the incubated sperm. When it is incubated at 40°C, the central groove disappears completely and a plateau is observed (Fig. 14). The width of the present plateau is considerable (approximately 540 nm). All these changes cause impediments in the floating and swimming motion, and hence they fail the “swim-up test.” As mentioned earlier, this view, namely, the pathological alteration for the evaluation of the sperm quality has been applied to male contraceptives and the tail has been examined after the treatment. Certainly, the tail has notable alterations not only topographical, but also in its form. The coiling of the tail is clearly visible: obviously, the sperm fertilization capability is out of question (25). It is known that at a high temperature of the scrotal (18), the sperm production is reduced. The present investigation shows that at high temperature, the quality of the sperm is also considerably lowered and abnormalities, which affect functioning, are spread over the entire part of the sperm.

4. Supplementary Information AFM is the only technique which provides information at subnanostructure of the living organisms in their natural environment. It is, therefore, expected that this technique could reveal some facts or specific data which were not revealed before. This has been precisely observed in the case of HIV patients. The role of the sperms in HIV is an interesting aspect in the last decade. It was thought that HI-virus cannot enter in the sperm. However, AFM clearly reveals that the images where virus is on the head of the surface (12) and they are merging or penetrating. This is because HIV surface glycoprotein, gp 120, binds specifically to the galactosyl–acylglycerol (GALAAG) and its sulfate form. The seminolipid (SGALAAG) acts as an HIV receptor. Such valuable information was expected but never detected experimentally. This is the first time that the presence of the virus and its merging process is clearly perceived with the help of AFM technique. Thus, it seems that semen is a safe haven for HI-virus. This awareness is very significant for pharmacological research in HIV treatment.

5. Applications and Prospects The outcome of the ultrastructural investigation of human sperm has direct application in identifying certain infertility-related diseases and also in the developing of certain drugs for contraceptive purposes.

282

Joshi, Cruz, and Osuna

The morphology of the sperm is a significant prognostic f­actor for fertility-related issues, and potential targets are in the development of male contraceptive drugs. The investigation shows that there are certain drugs which produce charge ­imbalance on the sperm membrane, and consequently the sperm surface is destabilized and the acrosomal contents are dispersed (24, 25). Obviously, the morphology and the topology are altered and the effect of the drugs can be visualized and evaluated with the help of AFM. Such work has been recently carried out by Kumar et al. (24, 25), and it is expected that male contraceptives are available in a short time. It is necessary to examine the causes of infertility in men who are trying to surpass an infertility condition by assisted reproductive techniques for artificial reproduction purpose. The patients having globozoospermia are the most suitable candidates for artificial reproduction as the only defect is in the acrosome cap, which can be viewed by AFM. If the height of the head is normal and the acrosome is missing then independent of the morphology of the tail, the sperm is just fit for artificial reproduction. This suggests that AFM technique should be in the hands of the andrologists for routine procedures. Investigation on the temperature dependence of the alterations in the morphology of the sperm is now possible to be detected with very recent developments in AFM technology (5). It has been reported that atomic force microscopic measurements can be carried out with the precision of 0.025°C with long-term reproducibility. This clearly suggests that now it is time to examine pathological alterations in human sperm very precisely, and to examine what type of alterations and how they are taking place in human sperm to evaluate the impact of fever and related diseases on the fertility of the patients. Such information is truly needed for research in genetics-related fields and for reproductive medicine. It is our great concern to get more information about the environmental effects on the healthiness of the sperm. Morphological alterations also help to assess the chromatin structure as well as intact and reacted acrosome in human sperm. These features are decisive in fertility-related diseases. In the last couple of years, substantial improvements have been made in the technology of AFM both in speed and resolution and now it is emerging as a powerful tool in the hands of scientists who are working in reproductive medicine. The pathology of human sperm has impact not only in pharmacology, but also in biotechnology, microbiology, histology, and genetics. These features will be exploited in the near future. Thanks to AFM.

Highlights on Ultrastructural Pathology of Human Sperm

283

Acknowledgments We are thankful to CONICIT (Venezuela) for financial support. We also thank to Honorio Medina and J.M. Barboza for their valuable help and collaborations in this project. References 1. Egerton R.F. (ed.) (2008) Principles of electron microscopy: An introduction to TEM, SEM& AEM, Springer. 2. Swam M.A.(1997) Improved preservation of ultrastructural morphology in human spermatozoa using betaine in the primary fixation International journal of andrology 20, 45–54 3. Cohen S.H. and Lightbody M.L. (1999) Eds Atomic Force Microscopy, Kluwer academic, New York. 4. Han W., Mou J., Sheng J., Yang J. and Shao Z. Cryo (1995) Atomic Force Microscopy: a new approach for biological imaging at high resolution. Biochemistry 34 (26), 8215–20. 5. Agilent Technologies. www.agilent.com/find. 5600LS AFM, data sheet. 6. Joshi N.V., Medina H, Urdanedta H, Barboza J (2000) Atomic force microscopic study of ultrastructure of entamoeba histolytica in Scanning and force microscopies for biomedical applications II. Edt. Nie S, Tamiya E. and Yeng E. published in SPIE, 3922, 222–227. 7. Joshi N.V., Medina H., Urdaneta H. and Berrueta L. (1999) In vivo nanoimaging and ultrastructure of entamoeba histolica by using atomic force microscopy Experimental Parasitology 93, 95–100. 8. Joshi N.V., Medina. H, Colasante C. and Osuna A (2000) Ultrastructural investigation of human sperm using atomic force microscopy, Archives of Andrology 44, 51–57. 9. Joshi N., Medina H., Cruz I. and Osuna J. (2001) Determination of the ultrastructural pathology of human sperm by atomic force microscopy Fertility and Sterility, 75, 961–965. 10. Allen M.J, Bradbury E.M and Balhorn R. (1996) The chromatin structure of well spread demembranated human sperm nuclei revealed by atomic force microscopy. Scanning Microscope, 10, 989–94. 11. Joshi N.V., Medina H. and Osuna J.A. (2001) Ultrastructural pathology of varicocele spermatozoa by using atomic force microscopy Archives of Andrology 47, 143–152.

12. Barboza J.M., Medina H., Doria M. Rivero L. Hernandez L. and Joshi N. V. (2004) Use of atomic force microscopy to reveal sperm ultrastructure in HIV-patients on highly active antiretroviral therapy. 50, 121–129. 13. Irvine D.S. (1998). Epidemiology and etiology of male infertility. Hum. Reprod. 13 (Suppl 1):33–44. 14. Bhasin S, de Krester D.M and Baker H. W. (1994) Pathophysiology and natural history of male infertility. J Clin Endocrinol Metab. 79:1525–9. 15. World Health Organization. (1991) Infertility: a tabulation of available data on prevalence of primary and secondary infertility. Geneva., WHO Programme on Maternal and Child Health and Family Planning, Division of Family Health. 16. Gordon Baker H.W. Clinical Management of Male Infertility (2006). In De Groot L.J, Jameson J.L (eds) Endocrinology. Fifth ed. Vol.3 Elsevier Saunders, Philadelphia, PA, USA. 3199–3225. 17. Kerr J.B, de Krester D. Functional Morphology of the Testis.In: De Groot L.J, Jameson J.L. (2006) (eds) Endocrinology. Fifth ed. Vol.3 Elsevier Saunders, Philadelphia, PA, USA. 3089–3138. 18. Mieusset R, Bujan L. (1995) Testicular heating and its possible contributions to male infertility: a review. Int. J. Androl. 18, 169–184. 19. WHO laboratory manual for the examination of human semen and sperm-cervical mucus interaction (1999). Fourth edition. World Health Organization. Cambridge University Press. 20. Eddy E.M. (1998) The spermatozoon in “The physiology of reproduction”. Eds. Knobil E. and Neil J., Raven press, New York, 27–88. 21. Davis J.R. and Langfordt G.A. Testicular proteins. In: Johson A.D, Gomes W.R, and Vandemark N.L. (1970) The Testis. Vol.II. Academic Press New York and London. 259–306.

284

Joshi, Cruz, and Osuna

22. Hecht N.B, Bower P.A, Waters S.H, Yelick PC, Distel R.J. (1986) Evidence for haploid expression of mouse testicular genes. Exp. Cell. Res. 164, 183–90. 23. WHO: The influence of varicocele on parameters of fertility in a large group of men presenting to infertility clinics. World Health Organization. Fertil Steril. (1992) 57: 1289–93. 24. Kumar S. Chaudhary K. Sen P. and Guha S.K. (2006) Topological alterations in human spermatozoa associated with polyelectrolytic effect of RISUG Micron 37, 526–32. 25. Kumar S. Chaudhary K. Sen P. and Guha S.K. (2005) Atomic Force Microscopy: a powerful tool for high-resolution imaging of spermatozoa J. nanobiology. 3, 9–16. 26. Lee I.D.. Allen M-J- and Balhorn (1997) Atomic force microscope analysis of chromatin volumes in human sperm with head shape abnormalities. 56, 42–49. 27. MacLeod J. (1965). Seminal cytology in the presence of varicocele. Fertil Steril. 16, 735–57. 28. Zucchi A, Mearini L, Mearini E, Fioretti F, Bini V, Porena M. (2006) Varicocele and Fertility: Relatioship Between Testicular Volume and Seminal Parameters Before and After Treatment. J Androl. 27, 548–51. 29. Sandlow J. (2004) Pathogenesis and treatment of varicocele. Br Med. 328, 967–68. 30. Mieusset R, Bengoudifa B, and Bujan L.(2007) Effect of Posture and Clothing on Scrotal Temperature in Fertile men. J. Androl. 28:170–75.

31. Bergmann M, Behre HM, Nieschlag E. (1994) Serum FSH and testicular morphology in male infertility. Clin. Endocrinol. 40, 133–36. 32. Holstein A.F, Schirren C.G, Schirren C. Human spermatids and spermatozoa lacking acrosome. J Reprod. Fertil. (1973), 35, 489–91. 33. Brown-Woodman P.D., Post E.J., Gass C. and White I.G. (1984). The effect of a single sauna exposure on spermatozoa Arch. Andrology 12, 9–15. 34. Wang C. McDonald V., Leung A. Superlano L.; Berman N., Hull L. and Swerdloff R.S. (1997) Effect of increased scrotal temperature on sperm production in normal men. Fertility and sterility 68, 334–339. 35. Chohan K.R, Griffin J.T, Lafrombroise M, De Jonge C.J, Carrel D.T. (2006) Comparison of Chromatin Assays for DNA Fragmentation Evaluation in Human Sperm. J. Androl. 27, 53–59. 36. Ord T, Patrizio P, Marello E, Balmaceda J.P, ASCH RH. Mini-Percoll ( 1990): A new method of semen preparation for IVF in severe male factor infertility. Hum Reprod.; 5 :987–989. 37. Palermo G, Joris H, Derde M.P, Camus M, Devroey P, Van Steirghem A.C. ( 1993) Sperm characteristics and outcome of human assisted fertilization by sub-zonalinsemination and intracytoplasmatic sperm injection. Fertil.& Steril. 59, 826-835.

Chapter 18 High-Speed Atomic Force Microscopy and Biomolecular Processes Takayuki Uchihashi and Toshio Ando Abstract Atomic force microscope (AFM) is unique in its capability to capture high-resolution images of biological samples in liquids. This capability will become more versatile to biological sciences if AFM additionally acquires an ability of high-speed imaging, because “direct and real-time visualization” is a straightforward and powerful means to understand biomolecular processes. However, the imaging speed of conventional AFM is too slow to capture moving protein molecules at high resolution. In order to fill this large gap, various efforts have been carried out in the past decade. In this chapter, the past efforts for increasing the scan rate and reduction of tip–sample interaction force of AFM and demonstration of direct visualization of biomolecular processes are described. Key words: AFM, Protein, Bio-imaging, Dynamics, High-speed AFM

1. Introduction Atomic force microscopy (AFM) allows us to visualize individual protein molecules in aqueous solutions directly at submolecular resolution (1). However, its application range in biological research is limited because of its low imaging rate. With conventional AFM instruments, we cannot study dynamic molecular processes carried out by proteins which occur in the subsecond time scale. Over the past decade, significant efforts have been carried out to increase the scan speed of AFM and to achieve low invasiveness under high-speed imaging conditions (for recent reviews, see refs. 2–4). The most advanced high-speed AFM can now capture successive images at 30–60  ms/frame under the condition of a scan range of ~250 nm and ~100 scan lines (5, 6).

Pier Carlo Braga and Davide Ricci (eds.), Atomic Force Microscopy in Biomedical Research: Methods and Protocols, Methods in Molecular Biology, vol. 736, DOI 10.1007/978-1-61779-105-5_18, © Springer Science+Business Media, LLC 2011

285

286

Uchihashi and Ando

The tip–sample interaction force is greatly reduced without sacrificing the imaging rate significantly. As a result, dynamic biomolecular processes including molecular interactions, conformational changes, and diffusion of proteins can be captured on video (7–9). This chapter focuses on key techniques that realize fast and low-invasive imaging and gives examples of fast imaging of biomolecular processes.

2. Materials 1. AFM cantilever (AC-10, Olympus, Tokyo, Japan). 2. Mica (natural muscovite or synthetic fluorophlogopite). 3. Epoxy adhesive. 4. Highly oriented Pyrolytic Graphite (HOPG). 5. Phosphatidyl choline (PC, Avanti Polar Lipids, Inc, Alabama, USA). 6. Phosphatidyl ethanolamine (PE, Avanti Polar Lipids, Inc, Alabama, USA). 7. His-tag. 8. Dioleoyl-phosphatidyl-choline (DOPC, Avanti Polar Lipids, Inc, Alabama, USA). 9. Streptavidin (Sigma-Aldrich). 10. Phenol (Wako, Osaka, Japan). 11. Plasma etcher (South Bay Technology, CA, USA). 12. GroEL and GroES (Sigma-Aldrich). 13. Caged adenosine triphosphate (ATP, DOJINDO, Kumamoto, Japan). 14. The imaging buffer I: 50  mM HEPES–NaOH, pH 7.4, 50 mM KCl, 10 mM MgCl2, 400 mM caged ATP and 70nM GroES. 15. UV laser light. 16. Cytoplasmic dynein (gifted by Prof. Y. Toyoshima, Tokyo Univ.). 17. Microtubule (purified from chick brains). 18. Imaging buffer II: 80 mM PIPES–NaOH, pH 6.8, 25 mM potassium acetate, 0.1  mM ATP, 10% dimethyl sulfoxide (DMSO), and 2  mM taxol. Taxol was used to stabilize microtubules.

High-Speed Atomic Force Microscopy and Biomolecular Processes

287

3. Methods 3.1. Theoretical Background for Imaging Speed

Our high-speed AFM is based on the tapping mode (10, 11) in which the AFM tip is vertically oscillated and periodically brought into contact to a sample surface during scanning. The tip oscillation reduces the lateral force between tip and sample, and thus minimizes damage and/or deformation of biological molecules. The vertical tip force acting on the sample is controlled by a Proportional-Integral-Derivative (PID) feedback control so that the oscillation amplitude of the cantilever is kept constant. Precise and fast feedback control is the mandatory requirement for fast and low-invasive imaging (see Note 1). Here we consider factors that limit the scanning speed of AFM. Supposing that an image is taken in a time T for a scan range W × W with scan lines N, the scan velocity Vs in the x-direction is then given by Vs = 2WN/T. For W = 240  nm, N = 100, and T = 30 ms, Vs becomes 1.6 mm/s. Assuming that the sample surface has a sinusoidal shape with a periodicity l in the x-direction, the sample stage is moved in the z-direction with a frequency of f = Vs/l for the tip–sample distance to be maintained constant. When l = 10 nm and Vs = 1.6 mm/s, f becomes 160  kHz. The feedback bandwidth fB should be equal to f or higher and thus can be expressed as fB ³ 2WN/lT. Because of the chasing-after nature of feedback control, the sample topo­ graphy is always traced with a phase delay. The phase delay q is given by ~2 × 2pfDt, where Dt is the open-loop time delay (the sum of time delays of devices contained in the feedback loop). The main delays in tapping-mode AFM are the reading time of the cantilever’s oscillation amplitude, the cantilever’s response time, the z-scanner’s response time, the integral time of error signals in the feedback controller, and the parachuting time (tp). “Parachuting” means that the cantilever tip completely detaches from the sample surface at a steep down-hill region of the sample and thereafter takes time until it lands on the surface again. The feedback bandwidth is usually defined by the feedback frequency that results in a phase delay of p/4. With this definition, we obtain fB = 1/(16Dt). Tip parachuting significantly deteriorates obtained images under high-speed scanning and hence its avoidance or minimization is the most important subject in developing high-speed AFM. The parachuting time tp is a function of various parameters such as the sample height h0, the ratio r of the cantilever amplitude set point As to the free oscillation amplitude A0 (see Note 2), the phase delay q, and the cantilever’s resonant frequency fc (for details, see ref. 2).

288

Uchihashi and Ando

3.2. Devices of High-Speed AFM

The basic structure of our high-speed AFM instrument is similar to that of conventional AFM setups. However, various devices are optimized for high-speed scanning (2, 12). The key devices are given below.

3.2.1. Cantilevers

Cantilevers for fast and low-invasive imaging should have a highresonant frequency and a small spring constant. As a result, cantilevers should have small dimensions (see Fig. 1 and Note 3). Most advanced cantilevers for high-speed AFM are made of silicon nitride (13). They are ~6  mm long, ~2  mm wide, and ~90  nm thick, and are coated with a layer of ~20-nm-thick gold, which results in a resonant frequency fc of ~3.5 MHz in air and ~1.2 MHz in water, spring constant kc of ~0.2 N/m, and quality factor Q c of ~2.5 in water. Therefore, their response time tc(=Q c/pfc) is 0.66 ms in water. This type of cantilever is not yet commercially available, but cantilevers with a resonant frequency of 600  kHz in water and a spring constant of ~0.1 N/m are commercially available.

3.2.2. High-Speed Scanner

Several conditions are required to establish a high-speed scanner: (a) high resonant frequencies, (b) a small number of resonant peaks in a narrow frequency range, (c) sufficient maximum displacements, (d) small crosstalk between the three-dimensional (3D) axes, and (e) low quality factors. We employ flexure stages made of blade springs for the x- and y-scanners (see Fig. 2 and Note 4). The flexure stages are made by monolithic processing to minimize the number of resonant peaks (12). The y-scanner displaces the x-scanner, and the x-scanner displaces the z-scanner over which a sample stage is placed. The maximum displacements of the x- and y-scanners at 100 V are 1 and 3 mm, respectively. The x-piezoactuator is held at both ends with flexures, so that its center of mass is hardly displaced and, consequently, no large mechanical excitation is produced. The x-scanner has resonant

Fig. 1. Electron micrograph of a small cantilever developed by Olympus. Scale bar, 1 mm.

High-Speed Atomic Force Microscopy and Biomolecular Processes

289

Fig. 2. Sketch of the high-speed scanner currently used for imaging studies. A sample stage is attached on the top of the upper z-piezoelectric actuator (the lower z-piezoelectric actuator used for counterbalancing is hidden). The dimensions (W × L × H  ) of the z-actuators are 3 × 3 × 2 mm3.

peaks at 45 and 65 kHz, and higher frequencies; however, these peaks are not large. The z-piezoactuator (maximum displacement, 2 mm at 100 V; self-resonant frequency, 360 kHz) is held only at the four side rims parallel to the displacement direction. The z-piezoactuator can be displaced almost freely in both counter directions, and consequently, impulsive forces are barely exerted on the holder. This holding method has an additional advantage in that the resonant frequency is not lowered by holding, although the maximum displacement decreases by half. The x-scanner is actively damped either by the previously developed Q-control technique (14) or by feedforward control using inverse compensation (15, 16). The z-scanner is also actively damped either by the Q-control technique or by inverse compensation. An electronic circuit that automatically produces an inverse transfer function for a given transfer function was developed (6, 17). By using this compensation, the z-scanner bandwidth fs is extended to ~500 kHz, and the quality factor Q  s is reduced to ~0.5. Therefore, its response time ts (=Q  s/pfs) is ~0.32 ms. 3.2.3. Feedback Controller

The tip–sample interaction force has to be minimized for lowinvasive imaging. This is practically carried out by using an amplitude set point r close to 1. However, under this condition, the oscillating cantilever tip easily detaches from the sample surface at steep down-hill regions of the sample. Once detached completely, like “parachuting,” it takes time for the tip to touch the surface again. Parachuting results in a loss of sample topography information and a low feedback bandwidth. During parachuting, the error signal is saturated at (2A0 − As) = 2A0(1 − r), where 2A0 is the free oscillation peak-to-peak amplitude. A shallower set point results in a smaller saturated error signal and hence prolongs the parachuting time tp. The feedback gain cannot be increased to shorten tp, as a larger gain induces overshoot at up-hill regions of the sample, resulting in the instability of the feedback operation.

290

Uchihashi and Ando

Fig. 3. Schematic showing principle of dynamic PID control. Solid line : an amplitudedistance curve; gray line : an error signal used in conventional PID control; and broken line : an error signal used in dynamic PID control.

To solve this problem, a novel PID controller named “dynamic PID controller” was developed. It can automatically change the feedback gain depending on the oscillation amplitude (18). Namely, the feedback gain is increased when the error signal exceeds a threshold level (see Fig. 3). By using this technique, the parachuting time tp is significantly reduced, or parachuting is avoided completely, resulting in a significant increase in the feedback bandwidth and a decrease in the tapping force. 3.2.4. Other Devices

To drive the z-scanner at a high frequency, a piezodriver with a high slew rate and low electronic noise is required. The piezodriver is custom-made and characterized with a maximum output voltage of 50 V, rms noise 95% purity ethanol. 6. 3-aminopropyltriethoxysilane (3-APTES) (see Note 1).

2.1.3. Functionalization of Gold-Coated AFM Tips

1. 11-Mercaptoundecanoic acid. Store in acid cabinet. 2. 1-ethyl-3-(3-dimethlaminopropyl)carbodiimide (EDC). 3. N-hydroxysuccinimide (NHS). 4. HPLC grade ethanol.

2.1.4. Mica Substrate

1. 9.5 mm diameter V1-Mica disk or 15 mm2 or 25 mm2 square. 2. Die and punch set: for cutting round mica disks from a larger sample. This is necessary only if your mica disks/squares are too large to fit onto a glass slide. 3. Optically clear epoxy. 4. Scotch tape. 5. Glass slide/hard substrate. 6. Biotinylated BSA.

2.2. Atomic Force Microscope

An AFM consists of four major components: a cantilever, ­low-noise photodiode, laser source, and a piezoelectric element pictured in Fig. 1. A thermoelectric module or a similar temperature regulation device can also be used to regulate the temperature of the sample. In the configuration pictured in Fig. 1, the alignment of the laser with a beam splitter allows for a free optical path so that

334

Rico, Chu, and Moy beamsplitter piezo

light source

photdiode

dish

objective

cantilever

thermoelectric module

Fig.  1. Atomic force microscope. Schematic diagram of an atomic force microscope coupled to an inverted optical microscope with the relevant components: laser source, photodiode, piezoelectric element, and cantilever. The sample can be placed on a thermoelectric module for temperature control.

the AFM can be used on top of an inverted optical ­microscope, mechanically and acoustically isolated. Alternatively, the AFM can also be used as a stand-alone system on top of a mechanically ­isolated table. 2.2.1. Optical Detection Method

The cantilever is described in detail in the next section (see Note 3). Essentially, the cantilever behaves like a spring with spring constant k, which follows the Hooke’s law F = kd for small deflections, where F is force and d is the cantilever deflection. The deflection is measured by monitoring the position of the reflected laser with a segmented photodiode. Traditionally, laser diodes have been used to measure cantilever deflection, however, optical interference due to the high spatial coherence of lasers results in a sinusoidal-like wave in the deflection signal of force–distance curves (23, 24). The amplitude of interference is sometimes two to three times greater than single molecule rupture forces, thus compromising single molecule detection. To reduce optical interference, superluminescent diodes (SLDs) with a low coherence can be used in place of laser diodes. The light beam is focused on a spot of ~10 mm diameter using a microfocusing lens. The spot is positioned on the end of the cantilever and the reflected light is then collected by the segmented photodiode (Fig. 1). The difference in the photocurrents between the vertical quadrants is then transformed into the deflection of the cantilever. Thus, calibration of the photodiode sensitivity (sometimes referred as InvOLS, inverse optical lever sensitivity (25)) is required when carrying out force spectroscopy.

Force-Clamp Measurements of Receptor–Ligand Interactions

335

2.2.2. Piezoelectric Element

The piezoelectric element pictured in Fig. 1 moves the ­cantilever in the z-direction allowing the functionalized cantilever to make contact with the stationary protein-coated substrate. Another common configuration is to use the piezoelectric ­element to move the sample stage while keeping the cantilever stationary. The position of the piezoelectric element is monitored using either strain gauge (resistive) or capacitive elements. In the strain gauge method, as the piezoelectric element expands and contracts the resistance changes and the position can be monitored. The most common capacitive sensors are linear variable differential transformers (LVDTs), which measure the capacitance change between a fixed electrode and another electrode that moves along with the piezoelectric element. Since capacitive elements move with the moving piezoelectric element, capacitive elements ­generally provide more accurate position monitoring than strain gauge methods (Physik Instrumente has an excellent online piezo tutorial at http://www.physikinstrumente.com).

2.2.3. Piezoelectric Feedback Control

Piezoelectric elements are inherently nonlinear and present hysteresis and creep. This reduces the repeatability of the measurements and the accuracy in repositioning the AFM tip relative to the sample. Creep results in further displacement of the piezoelectric element, when the applied voltage signal is constant. Hysteresis is a complex process where the current displacement behavior is dependent on the previous history. A proportional-integrative-derivative (PID) controller can be used to correct for these nonlinear effects by computing a feedback signal from the piezoelectric position sensors mentioned in the last section to correct for these effects (26). AFM operation without nonlinear piezoelectric correction is referred to as open-loop while closed-loop refers to feedback correction. The advantages of closed-loop operation are a high reproducibility in the positioning of the tip and a dramatic reduction of nonlinearity, creep, and hysteresis. In force spectroscopy measurements, these corrections have an important effect in the applied velocity and the calibration of the photodiode sensitivity. The disadvantage of closed-loop operation is that the deflection signal is noisier as both electronic noise and mechanical/frictional noise, arising from feedback correction, are present. On the other hand, open-loop operation reduces these noise sources; hence, smaller forces can be resolved, but at the tradeoff of nonlinear cantilever displacement. The advantages of closed-loop operation are very useful in AFM imaging and lateral positioning, and most of the new commercial AFM systems include closed-loop operation.

2.3. Optimal Cantilevers for Single Molecule Rupture Force Measurements

AFM cantilevers are force transducers that can measure and apply pico to nano-Newton forces to the sample of interest. They come in many different sizes, shapes, and can be coated with different reflective metal coatings (Au, Al…). Each cantilever is optimized

336

Rico, Chu, and Moy

for a different function; hence, it is crucial for force spectroscopy measurements to select the cantilever optimal for the specific system. The tip at the end of each cantilever also comes in different sizes and shapes and may influence force spectroscopy measurements, as it is the actual part of the AFM to come into contact with the sample. A partial selection of cantilevers used for SMFS is provided in Table 1. 2.3.1. Composition and Dimensions of Cantilevers/Tips

Cantilevers commonly used for SMFS are mostly made of silicon nitride (Si3N4), and are mainly V-shaped or rectangular (see Fig. 2). These cantilevers vary in length from 60 mm for the small Biolevers (Olympus, Japan) to 320  mm for the Microlevers (Veeco, Santa Barbara, CA). At the free end of the cantilever, there is a tip that can be pyramidal, half pyramidal, or conical. Since the tip is the part of the AFM probe that comes into ­contact with the substrate, the tip shape affects the area of contact, which ultimately influences the probability of bond formation between the cantilever tip and the substrate. Generally, unsharpened cantilever tips have been mainly used for SMFS as they provide a smoother and slightly larger area of contact between the tip and the substrate while also reducing the local strains on biological systems, such as cells (2, 27–29). However, sharpened cantilevers have also been used effectively for SMFS (30). A possible drawback of sharpened tips is that they wear relatively faster than unsharpened ones after successive force scans, thus affecting the protein coating as well.

2.3.2. Reflective Coating of Cantilevers/Tips

The top or reflective surface of the cantilever is commonly coated with a thin gold layer to enhance the maximum amount of reflected light (2, 29, 30). Aluminum is also used as a reflective coating. Sometimes, both the reflective and tip surfaces of the cantilever can be coated with gold, in which case thiol chemistry is required for tip functionalization (this is the case of Olympus Biolevers, see Subheading 3.1) (27, 28). Lastly, it is also possible to obtain cantilevers without a reflective coating. Removal of the reflective coating allows for a clear optical light path through the cantilever and minimizes thermal drifts. The disadvantage of using cantilevers without a reflective coating is that the reflected laser signal is much less intense, leading to possibly lower sensitivity.

2.3.3. Ideal Range of Spring Constants

The cantilever spring constant determines the amount of deflection or displacement in the cantilever in response to an applied force. When a molecule tethered to the cantilever tip interacts with an apposing molecule attached to the substrate, the amount of cantilever deflection (d) upon retraction of the functionalized cantilever from the surface depends on the stiffness of the cantilever (kc), the stiffness of the linker and the protein (kl), and lastly the molecular displacement (x) upon separation of the cantilever from the substrate which are related by (31)

Si3N4

Olympus

Veeco

Veeco

Olympus

Biolever BL150VB-C1 (A/B)

NP series (B/D)

MSCT (B/C/D) sharpened MLCT

OMCLTR400PSA

Si3N4

Si3N4

Si3N4

Si3N4

Composition

Veeco

Manufacturer

MLCT-AUHW (B/C/D)

Catalog number

V-shaped

Rect./ V-shaped

V-shaped

Rectangular

Rect./ V-shaped

Shape

200

200/320/ 220

196

60/100

200/320/ 220

Length (mm)

Table 1 Cantilevers for single molecule force spectroscopy measurements

13.4/0.4

22/0.6

(33/18)/ 0.6

30/0.18

22/0.6

Width/ thickness (mm)

4-sided pyramid/ 2.9/20

4-sided pyramid/ 2.5-3.5/20

4-sided pyramid/ 2.5-3.5/20

Semi 4-sided pyramid/ 7/30

4-sided pyramid/ 2.5-3.5/50

Tip shape/ height (mm)/ radius(nm)

0.02

0.02/0.01/ 0.03

0.06/0.03

0.03/0.006

0.02/0.01/ 0.03

Nominal spring constant (N/m)

11

15/7/15

14/12

37/13

15/7/15

Resonant frequency in air (kHz)

(44)

(30)

(29)

(27, 28)

(2, 3)

References

338

Rico, Chu, and Moy

a

b

V-shaped cantilever

Bottom view

length

thickness

width

Rectangular cantilever Bottom view Reflective/top side Tip/bottom side

Side view

Fig. 2. AFM cantilevers. (a) Chip of AFM cantilevers with a regular size paperclip. Inset, magnification of one end of the chip showing four V-shaped cantilevers and one rectangular cantilever (MLCT-AUHW, Veeco). (b) Schematics showing the bottom and lateral views of V-shaped and rectangular cantilevers.



d=

kl x kl + kc

(1)

From Eq. 1, if kc » kl, the deflection of the cantilever is rather small and may be smaller than the resolution of the system. On the other hand, if kc « kl, it may require a larger amount of cantilever deflection (d) to stretch the molecule, which means a larger displacement of the piezoelectric element. Hence, it is important to keep these factors in mind when choosing an optimal cantilever for SMFS. In addition, it has been suggested that the spring constant of the cantilever may affect the energy landscape of the interaction (32–34). Cantilevers with spring constants ranging from 0.006 to 0.06 N/m have been successfully used previously (2, 27–30). The most commonly used cantilevers for SMFS generally have a nominal spring constant of 0.01 N/m (2, 30). 2.3.4. Noise Limit of Cantilevers



Assuming that the system is mechanically and electrically well ­isolated, the noise in the AFM detection system is mainly limited by the thermally driven motion of the cantilever. If the thermal noise in the cantilever is on the same scale or larger than the measured force of interest, then these cannot be well resolved. The thermal noise affecting the cantilever deflection is approximated by ∆Frms =

4g kBTf s

(2)

where kB is Boltzman’s constant, T is the absolute temperature, fs is the cutoff frequency, which should be well above the frequencies of interest over which SMFS measurements are acquired (31), and g is the viscous drag coefficient of the cantilever, which strongly depends on the separation between the cantilever and

Force-Clamp Measurements of Receptor–Ligand Interactions

339

the substrate (35, 36). Since the viscous drag coefficient is ­dependent on the size of the cantilever, it is thus possible to reduce thermal noise by reducing cantilever size (37). One of the smallest cantilevers commercially available to date is the Olympus Biolever (BLRC150V, Table 1). In addition, a higher tip increases the distance between the cantilever and the substrate, reducing the effective viscous drag. Low-pass filtering is also an effective way to reduce the noise in the deflection signal. In the case of the small Biolevers, a theoretical thermal noise limited detection of 5 pN is reasonably expected (38).

3. Methods 3.1. AFM Tip Functionalization

3.1.1. Protein Functionalization of Gold-Coated Tips via Thiol Self-Assembled Monolayers

The most common method of adsorbing proteins to AFM tips is either by coupling via amino (2, 30, 39–46) or carboxy termini (27, 28) (Notes 13 and 15). In addition, it is also possible to link proteins to AFM tips through specific tags if the protein of interest contains a tag and the tip is coated to specifically recognize the tag (2, 42, 46, 47) (see Note 14). Certain cantilevers manufactured with goldcoated tips (i.e., Olympus BL-R150VB and OMCL-TR400PB-1) can be functionalized with thiol-­reactive groups (-SH) to immobilize cross-linkers on the gold layer. Even more, the protein of interest can be linked directly to the gold-coated AFM tips if the protein of interest is labeled with a reactive thiol group (48). A more detailed description of common methods to coat AFM cantilevers with proteins is given in the chapter by Voyer and coworkers. 1. Wash tweezers in Alconox solution and rinse extensively (see Notes 6 and 7) 2. Shake cantilevers in a cleaned glass dish containing nanopure water for 1 min 3. Irradiate cantilevers with UV light for 5 min 4. Wash cantilevers with ethanol by completely submerging each cantilever once in each of five wells of a 24-well tissue culture plate containing 1 mL of ethanol 5. Completely submerge cantilevers in glass dish containing Nanothinks™ ACID11 for 24 h at room temperature. Nanothinks™ ACID11 contains a thiol group, a relatively short linker and a free carboxy terminus. A self-assembled monolayer then forms on the gold surface of the AFM tip, in which thiol groups interact with the gold surface while carboxyl groups remain free for cross-linking proteins to the tip (47, 49). 6. Rinse cantilevers five times with ethanol 7. Prepare activation buffer: 20 mg each of EDC and NHS in 1 mL of PBS pH 6.5–7.2. Equilibrate both EDC and NHS to

340

Rico, Chu, and Moy

room temperature before use. Mix thoroughly and then ­vortex for 1–2 min (see Note 8) 8. Completely submerge cantilevers in activation buffer EDC is a bifunctional cross-linker that reacts with carboxyl groups of the formed monolayer. Once reacted with a carboxyl group, EDC reacts with amine groups in the protein of interest. NHS prevents the rapid hydrolyzation of the EDC intermediate formed with ACID11 (see Pierce product info on NHS and EDC for more details). 9. Rinse cantilevers five times with PBS (see Note 9) 10. Dry the cantilever by gently touching it to the bottom surface of the well of a clean tissue culture plate (see Note 10) 11. Place a small square piece of parafilm (50 × 50  mm2) in the center of a clean Petri dish 12. Place the dry cantilever chips on top of the parafilm forming a circle with the cantilevers facing each other (see Fig.  3). Usually, around four cantilevers per protein drop 13. Gently press the center of each cantilever chip against the parafilm to prevent the cantilever from moving 14. Gently pipette 30–50 mL of the protein solution diluted to the desired protein concentration (see Note 11) 15. Check to see that all of the cantilevers are completely submerged in the solution, and that there are no air bubbles around the cantilevers 16. Transfer the entire tissue culture dish and protein-coated cantilevers into a humidified chamber (see Note 12)

a

b

Fig. 3. Handling of AFM cantilevers (see Note 4). (a) Stainless steel tweezers to manipulate c­ antilevers. (b) Four cantilever chips immersed in a drop of protein solution on top of parafilm forming a circle with the cantilevers facing each other (step 14 of Subheading 3.1.1).

Force-Clamp Measurements of Receptor–Ligand Interactions

341

17. Allow the proteins to coat the cantilevers either overnight at 4°C or for 3 h at room temperature 18. Rinse the excess protein from the protein-coated cantilevers five times in PBS or the desired experimental buffer 19. Store protein-coated cantilevers in PBS at 4°C for up to 5 days 20. Prior to using the cantilever in an experiment, block nonspecific binding sites by incubating for 30–60 min in 1% (w/v) BSA at room temperature 21. Rinse cantilevers five times in experimental buffer In our particular case, we used 50 mg/mL streptavidin as the protein solution. 3.1.2. Removing Thiols from Gold-Coated Cantilevers

Since the gold-coated cantilevers can be relatively expensive, and also coating cantilevers with gold is expensive and time consuming, it might be desired to reuse the cantilevers or any gold-coated surface. The protocol below shows how to remove thiols from gold-coated cantilevers by photooxidation (40). 1. Irradiate cantilever with UV light for 10  min (wavelength 254 nm @ 8 mW/cm2 or equivalently 60 min @ 1.25 mW/cm2) 2. Wash five times with nanopure water 3. Wash five times with ethanol 4. Store cantilever in air or argon

3.2. Substrate Coating

Hydrophilic surfaces are the most commonly used substrates to immobilize proteins for force spectroscopy. Hydrophilic surfaces include tissue culture dishes, glass coverslips/slides, beads, and mica. These are generally readily available and allow for relatively easy protein adsorption. Hydrophobic surfaces can also be used. The drawback of immobilizing proteins on hydrophilic surfaces is the increase in nonspecific binding compared to hydrophobic surfaces.

3.2.1. Mica

Mica is a mineral that exists as densely compact thin sheets. Commercially, it comes in different grades V1–V10, varying also in shape and pricing. Some of the nice properties of V1 mica are that each sheet is atomically flat, after cleavage each sheet is tremendously clean and optically clear, and many sheets can be cleaved from one sample. In most buffers, mica is negatively charged, making it relatively easy to functionalize mica surfaces. One of the most common methods to immobilize proteins to mica surfaces is through ionic interactions with a salt solution, namely NaCl (44). Mica can also be silanized to covalently crosslink adhesion molecules to mica (50, 51). The preparation of

342

Rico, Chu, and Moy

mica substrates is explained below in a two part protocol ­consisting of, firstly, fabrication and, secondly, protein coating of mica substrates (Notes 17 and 18 on alternative substrates). 3.2.1.1. Fabricating Mica Substrates

1. If your mica substrate fits onto a glass slide then proceed to step 2, otherwise use a dye and punch set and cut out a mica disk smaller than a glass slide. Avoid using scissors to cut mica. Some manufacturers sell precut mica disks/squares close to the desired size 2. Wash the glass slide or desired support with acetone 3. Glue the mica disk onto the dry clean glass slide with optically clear epoxy (EPOTEK 377). Follow manufacturer’s instructions for handling epoxy and curing time.

3.2.1.2. Functionalizing Mica Using a Salt Solution

1. Cleave the mica by taking a piece of scotch tape and pressing it against the mica disk. In one smooth, complete motion, rip the tape off (see Note 16) 2. Dilute the protein of interest in NaCl to a final molar concentration of 1 mM NaCl 3. Coat the mica disk with protein in 1 mM NaCl for 10 min 4. Rinse the protein-coated mica surface with 1 mM NaCl five times In our case, we use passive adsorption of biotinylated bovine serum albumin (bBSA) by incubating 1% bBSA overnight at 4°C.

3.3. AFM Force-Clamp Measurements

Force-clamp measurements require a fast enough feedback circuit to keep the deflection (force) constant, together with low-noise conditions to detect minimal forces. Low-noise conditions can be effectively achieved by vibration and acoustic isolation of the AFM system and by the reducing the size of the cantilever and the ­bandwidth of data acquisition. Most commercial AFM systems use  digital feedback to keep the force constant during contact mode imaging. However, its application to measurements in which the  force is held constant during pulling is not such a common feature of commercial software. We used a commercial AFM ­system MFP-3D-BIO (Asylum Research, Santa Barbara, CA), ­coupled to the stage of an inverted optical microscope (Nikon, Japan), which already provided force-clamp control procedures. The cantilevers used, biolevers, had a nominal spring constant of 6 pN/nm. The InvOLS was determined by calculating the slope of deflection (in V) versus distance (nm) from five consecutive force curves on the hard substrate, resulting in ~40  nm/V (this value depends on the specific electronics, on the geometrical configu­ ration of the AFM system, and on the specific cantilever used, Fig. 4). The spring constant of the cantilever was calibrated in ­liquid prior the measurements using a method provided by the AFM

Force-Clamp Measurements of Receptor–Ligand Interactions

343

0.8

Deflection (V)

0.6

slope=OLS

0.4

0.2

0

−20

−40

−60

−80

−100

Displacement (nm) Fig. 4. Sensitivity calibration. Representative example of a calibration curve obtained on a mica surface coated with BSA. The slope of the linear part of the curve was used to determine the optical lever sensitivity, OLS = 1/InvOLS.

manufacturer­ based on the thermal fluctuations method (52, 53), which lead to spring constants ranging from 5 to 7 pN/nm. The detailed force-clamp measurement procedure is described below. 1. Mount the cantilever on the cantilever holder (see Note 19) 2. Immediately immerse the cantilever in the measurement buffer to minimize any contact with air 3. Allow the system to equilibrate (15–30 min) 4. Be sure that the tip is far away from the surface (at least 20 mm) and acquire the thermal fluctuations of the cantilever to calibrate the spring constant using the thermal fluctuations method (52, 54) 5. Approach the tip to the substrate and make contact 6. Acquire five force-distance curves with a deflection ~1 V (this value depends on the particular AFM system) to calibrate the photodiode signal and determine the InvOLS by the inverse of the resulting slope (Fig. 4) 7. Position the tip on the protein-coated surface and make slight contact 8. Acquire force-clamp curves consisting in the following steps (Fig. 5): (a) Approach the tip to apply a force of few tens of pN (region I)

344

Rico, Chu, and Moy

(b) Maintain contact at this compression force for a certain time so as to assure that only 10–30% of the curves lead to bond formation (see Note 20) (region II) (c) Retract the cantilever at ~2 mm/s until it reaches the set clamping force (e.g., 10 pN) (region III) (d) If a bond is formed, hold the cantilever at this force until the bond breaks (region IV). If no bond is formed, the cantilever will reach its initial position (e) Retract the cantilever until the initial, resting position (regions V and VI) (f) Repeat the cycle (see Notes 20–22) 3.4. Control Measurements

Control measurements are crucial in adhesion measurements. In general, hundreds of force curves are acquired under determined conditions (compression force and contact time) and the adhesion frequency is calculated. Control measurements consist of acquiring force curves under the exact experimental conditions only modifying­ the ligand- or receptor-adhesive conditions and determining the resulting adhesion frequency. There are different methods to prove the specificity of the interaction that depend on each experimental system, such as blocking the receptor or ligand with antibodies, chelating ions required for the interaction, using cells not expressing the protein of interest, etc. An alternative method is using AFM tips (or substrates) in which the last step of protein coating was omitted. In our particular case, we blocked the biotinylated surface using free streptavidin as the most straightforward method, finding a dramatic reduction of the adhesion frequency. The relative adhesion frequency averaged over three positions of the sample surface dropped from 100% to 2 ± 1% after adding 100 mL of 100 mg/ml free streptavidin to the measurement buffer to block the biotin-coated substrate (final concentration ~50 mg/mL).

3.5. Data Processing and Analysis

Only force curves presenting a single adhesion event are considered. A representative example of a force-clamp curve is shown in Fig.  5, which represents the detected force and the piezo ­displacement as a function of time. Only the force-clamping region (region IV) is used to determine the two relevant parameters. 1. The actual clamping force, which depends on the resting force level, determined as the average level of force during clamping relative to the resting force level at zero velocity (see Note 23) 2. The bond lifetime measured from the beginning of the clamping regime until the force drops. The resulting list of forces and lifetimes is then pooled according to levels of force (in our case 10, 20, and 25 pN). The mean

Force-Clamp Measurements of Receptor–Ligand Interactions

345

(IV)

a 500 contact (II)

Displacement (nm)

400

(III)

300 200

approach (I)

100 0

Force (pN)

40

(V) (VI)

20

0

-20 0.2

b

Time (s)

0.6

0.8

60

40

Force (pN)

0.4

retraction clamping (III) (IV)

20

resting (VI)

repositioning (V)

0 clamping force

-20 lifetime

-40 0.80

0.85

0.90 Time (s)

0.95

1.00

Fig. 5. Force-clamp measurements. (a) Distance (top) and force (bottom) curves showing an example of a force-clamp measurement. The relevant steps of the curves are shown: (I) approaching regime in which force increased from zero (noncontact) to ~50 pN (trigger force); (II) in the contact regime, the force was maintained at the trigger value during a dwell time of ~0.5 s; (III) during retraction, the cantilever was withdrawn until the clamping force was reached (~20 pN, negative); (IV) during the clamping process, the force was maintained constant until the bond broke; (V) the cantilever was withdrawn 500 nm during repositioning; and (VI) the cantilever remained immobile during the resting regime. (b) Magnification of the four last steps of the force trace showing how to determine the lifetime of the bond and the actual clamping force.

346

Rico, Chu, and Moy

force of each group is thus calculated together with the standard deviation. For each set of clamping force, we generate the cumulative probability distribution of lifetimes by plotting the number of events with a lifetime of t or more against t (Fig. 6a), which can be modeled as an exponential decay (Aexp[-t/t (F) ]). The intrinsic lifetime of the interaction at the applied force, t(F), is then estimated by fitting the exponential to the data. More accurate methods can be applied to determine intrinsic dissociation lifetimes (55, 56). The lifetime obtained for each set is then represented against the average clamping force (Fig. 6b). The Bell model is the more generalized approach to interpret the effect of force on the lifetime of a bond (1, 5) and can be summarized by the following expression (see Notes 24 and 25)

3.6. Interpretation of the Results

t (F ) = t 0e



x b F / kBT



where xb is the effective range of the interaction, kB is the Boltzmann constant, and T, the absolute temperature. As observed, the lifetime of the bond decreases exponentially with the applied force, as is reflected from our experimental results (Fig. 6b). By fitting the above equation to our lifetime versus force data, we obtain the parameters of the bond, its intrinsic lifetime at no force (t0), and the width of the potential energy. In our case, we found a width of ~1 nm and a lifetime of ~9 s, which is in good agreement with previous results, given the low number of data points a

b 100

1

8 6

8

10 pN 20 pN 25 pN

6 4

2

Lifetime (s)

Number of events

4

10

8 6 4

2

0.1

8 6 4

2

2

1

8 6

0.01 0.0

0.5

1.0 1.5 Lifetime,t (s)

2.0

2.5

5

10

15 20 Force (pN)

25

30

Fig. 6. Dissociation kinetics. (a) Lifetime distributions representing the number of events with a lifetime of t or more against t at three different levels of clamping force (10, 20 and 25 pN, circles, crosses and squares, respectively). Solid lines are best fits of an exponential decay to determine the mean lifetime at each force. (b) Fitted lifetimes against actual forces. Error bars represent one standard deviation. The Bell model (solid line) was fitted to the curve obtaining a potential width xb = 1.0 ± 0.1 nm and an intrinsic lifetime at zero force t0 = 9 ± 1 s.

Force-Clamp Measurements of Receptor–Ligand Interactions

347

(11, 22). Given the virtually zero loading rate and the relatively low forces applied, it is possible that we were probing the outermost barrier of the streptavidin–biotin interaction, which has been described before at 1.3 nm (57). From previous studies of forced dissociation of the streptavidin–biotin complex, it is expected that an inner barrier would dominate the dissociation pathway at higher forces. This effect would be characterized by a weaker dependence of lifetime on the applied force (11, 21, 22).

4. Notes 1. APTES is a common cross-linker used to derivatize siliconbased substrates (i.e., glass or many types of cantilevers) and may also be used for quartz surfaces. The free amino group allows the cross-linking to proteins with a free carboxy terminus. This solution is very hygroscopic and can hydrolyze very quickly, so it is best to purchase in small volumes. 2. Handling AFM cantilevers is a difficult task when it is done for the first time. Handling broken cantilevers beforehand helps to acquire some practice without damaging new cantilevers. 3. Generally, cantilevers are mounted in an angle of 8–10° from the axis parallel to the substrate surface. 4. Cantilevers are flexible but fragile, and it is common to lose some of them during the various steps of the coating protocol. Thus, it is recommended to functionalize more than one cantilever, i.e., more than one chip, for each experiment (5–10 chips). Coated cantilever can be stored in PBS at 4°C for about 1 week. 5. The AFM cantilever holder can be made out of plexiglass or quartz. Quartz is a nice alternative to plexiglass as it is relatively more inert than plexiglass and is resistant to scratching. The disadvantage of quartz holders is the difficulty in machining. 6. Before starting the functionalization protocol, check to see whether all of the cantilevers on the chip are intact and undamaged to prevent wasting valuable protein and also your own time. 7. Long cantilevers (~320 mm) have the tendency to bend back onto themselves due to the surface tension of the fluid giving the appearance of a missing cantilever. To ensure that the cantilever is intact, gently tap the side of the vessel or use a pair of forceps and pick up the cantilever chip and tap it against the bottom of the vessel.

348

Rico, Chu, and Moy

8. Prepare fresh before using, and must be used immediately after preparing as the amine group on EDC and NHS can hydrolyze rapidly. Store both NHS and EDC in desiccant sieves at 4°C. 9. Notice that Tris–HCl buffer contains free amine groups. Avoid the use of this buffer for protein storage and during the functionalization process, as it reduces the coupling efficiency. PBS or HEPES buffers are recommended. 10. Residual fluid remaining on the cantilever dilutes the final protein concentration, but if the cantilever is too dry then it may be difficult to coat the cantilevers with protein solution. Just leave a small trace amount of fluid remaining on the chip, probably something less than 2 mL. 11. Depending on the system, the protein solution concentration for SMFS can range from 5 to 100  mg/mL, so generally a good starting protein concentration is 50 mg/mL. Usually a 30–50 mL is enough to coat four cantilever chips at once. 12. A humidified chamber can be made easily from a plastic box, e.g., a small sandwich lunchbox or an empty pipette tip box. Fill the bottom 10  mm of the box with water and place a stable platform protruding above the water so the Petri dish can sit inside the box without touching the water. 13. Another common method for coupling proteins to silicone nitride tips is through heterobifunctional cross-linker attached to the silanized silicone nitride surface (7, 27). 14. Most of the used methods to functionalize the tips, including those explained here, link the protein via any amine group, with no defined orientation. Alternatively, precise orientation of the protein can be achieved by making use of a specific tag present in it, such as His- or Avi-tags. Coating the tip with a receptor recognizing such tags or with an antibody binding to a known epitope of the protein would allow us to control the protein orientation. This method has been successfully used on different systems: His-tag (46, 47) and biotin (2, 42). The two major advantages of attaching proteins to AFM tips with site-specific tags are specificity and controlled protein orientation. However, it is important to first characterize the binding strength of such linking bonds because it may not be strong enough and may break before rupture of the complex under study (46). 15. Attachment of proteins to cantilevers with long linkers diminishes nonspecific adhesion of the tip with the functionalized surface. The long linker adds an elastic component to the single molecule rupture force profile. It also shifts the initial loading regime .The actual elastic response to force, usually described as a worm-like chain (58, 59), and the shift of the start of the loading regime depend on the length of the linker

Force-Clamp Measurements of Receptor–Ligand Interactions

349

and have been used as a signature of specific interaction (60). Two types of commonly used long linkers include polyethylene glycol (PEG) linkers (27, 28, 41, 60) and carboxymethyl amylose chains (49). Both of these moieties can be modified to obtain tethers of different length. 16. A nice cleavage should yield a very smooth surface devoid of any smaller pieces of mica. 17. Other convenient substrates on which proteins can be passively adsorbed are polystyrene dishes. As mentioned before, control measurements are required to confirm the specificity of the interaction. In general, most nonspecific binding in hydrophilic surfaces can be blocked with 1% BSA, but a small amount of nonspecific binding can still occur. The major advantage of hydrophobic surfaces is the reduction in nonspecific binding compared to hydrophilic surfaces. Ideal hydrophobic surface are Petri dishes manufactured to culture bacteria. Unlike tissue culture dishes, Petri dishes have not been chemically modified and are thus hydrophobic. One drawback of conducting adhesion measurements on hydrophobic surfaces is that for an equivalent protein concentration, less protein adsorbs to the hydrophobic than to the hydrophilic surface, hence, more concentrated protein solutions are generally required to functionalize hydrophobic surfaces. However, this is a small price to pay because there is either very little or, in some cases, absolutely no nonspecific binding to hydrophobic surfaces blocked with 1% w/v Pluronic (BASF), instead of BSA. 18. An alternative way of immobilizing proteins on a substrate is the use of commercially available beads. Beads come in many different materials and can vary greatly in size. One of the major advantages of beads is that many of them are commercially available with a wide range of conjugated proteins designed for protein purification, protein labeling, and other techniques requiring molecular protein specificity. Hence, if the adhesion molecule of interest is labeled with a tag specific (e.g., biotin, histidine, NTA, etc.) to the protein conjugated to the bead, then the molecule of interest can specifically adhere to the bead. Depending on the material composition, beads can also vary in stiffness, which may affect the contact area with the tip. In addition, the material composition may also affect the surface roughness of the beads. In order to work with beads in force spectroscopy, they must be immobilized to the substrate at least to the point where force curves do not affect bead position. Bead immobilization can be accomplished by using the same principle used for detection, i.e., using their own specific tag (61).

350

Rico, Chu, and Moy

19. Avoid exposure of all functionalized cantilevers to air, as ­proteins may be damaged. Thus, mount the cantilever on the AFM holder and immediately immerse it on the experimental buffer. 20. The adhesion frequency (the frequency of binding events) has to be kept