Mechanobiology of Cell-Cell and Cell-Matrix Interactions

Mechanobiology of Cell‐Cell and Cell‐Matrix Interactions A. Wagoner Johnson l Brendan A.C. Harley Editors Mechano...

Author: A. Wagoner Johnson | Brendan A.C. Harley

200 downloads 1135 Views 7MB Size Report

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

Report copyright / DMCA form

DOWNLOAD PDF

Mechanobiology of Cell‐Cell and Cell‐Matrix Interactions

A. Wagoner Johnson

l

Brendan A.C. Harley

Editors

Mechanobiology of Cell‐Cell and Cell‐Matrix Interactions

Editors A. Wagoner Johnson Department of Mechanical & Industrial Engineering University of Illinois, Urbana‐Champaign 1206 W. Green St. Urbana, Illinois 61801 USA [email protected]

Brendan A.C. Harley Department of Chemical & Biomolecular Engineering University of Illinois, Urbana-Champaign 600 S. Mathews Ave. Urbana, Illinois 61801 USA [email protected]

ISBN 978-1-4419-8082-3 e-ISBN 978-1-4419-8083-0 DOI 10.1007/978-1-4419-8083-0 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2011921261 # Springer ScienceþBusiness 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 (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 Springer is part of Springer Science+Business Media (www.springer.com)

Preface

Cell behavior is modulated by a complex, spatio-temporally integrated set of biophysical mechanisms influenced by the biochemistry of extracellular and intracellular signaling, but also by the properties of the surrounding extracellular environment. A cell assimilates multiple cues from its microenvironment, including signals bound to the extracellular matrix and neighboring cells, mechanical stimuli, and soluble signals from both adjacent and distant cells. The cell then responds to these signals via multiple pathways, each involving multiple cascades of internal molecular interactions. While much as been learned over the past decade regarding the mechanobiology of how cells interact with their surrounding environment in cases of physiology and disease, many fundamental questions remain. New tools as well as experimental and modeling approaches now enable researchers to answer a host of complex questions regarding the biophysics of how cells sense and respond to each other as well as to the multitude of extrinsic signals present in their local microenvironment. Continued progress in the field requires continued, close interactions between researchers in engineering, biology, physics, and medicine, but promises significant advances in the fields of regenerative biology and tissue engineering. This peer-reviewed book is one result of the Society of Engineering Science (SES) 45th Annual Technical Meeting, held October 12–15, 2008 at the University of Illinois at Urbana-Champaign. The meeting brought together scientists, engineers and mathematicians from around the world with the common belief that solutions to critical modern problems transcend traditional disciplinary boundaries and require bringing together diverse, interdisciplinary groups of researchers to discuss advances in highly focused symposia. We had the privilege of organizing the Mechanobiology of Cell-Extracellular Matrix Interactions Symposium at the meeting. The symposium focused on interdisciplinary research involving both experimental and modeling approaches to understanding the mechanisms of how individual or populations of cells respond to distinct extracellular cues. There were 28 oral presentations spread across six sessions, with the keynote address given by Prof. Yu-Li Wang (Carnegie Mellon University, USA). We would like to thank the

v

vi

Preface

General Chair (H. Johnson, U. of Illinois at Urbana-Champaign, USA), Technical Program Chair (I. Jasiuk, U. of Illinois at Urbana-Champaign, USA), as well as the rest of the Organizing Committee for their support in helping to make the symposium a success. This book has been organized into four technical sections that roughly reflect the organization of technical sessions at the SES symposium. l l l l

Mechanisms of Cell Adhesion and Mechanotransduction Cooperative Cell Behavior and Mechanobiology Mechano-pathology of Disease Tools for Exploring Mechanobiology

We wish to acknowledge a number of key people who helped make this book a reality. We would like to thank Elaine Tham and Michael Luby at Springer for her initiation of the project and encouragement to publish this volume as well as his management of the process. This book would have not been possible without the technical assistance we received along the way, notably staff in the Department of Mechanical Science and Engineering including Ben Kaap, Susan Petry, Pam Vanetta, and Jennifer Carroll, our students Michael Poellmann and Emily Gonnerman, as well as our colleagues within and administrative support from the Depts. of Mechanical Science and Engineering, Chemical and Biomolecular Engineering, and the Institute for Genomic Biology at the University of Illinois. This book was only possible due to the substantial effort of all of our authors, and we are extremely grateful for their tireless work in preparing their own chapters and in many cases serving as reviewers of additional chapters. We were also assisted by a multitude of colleagues who agreed to serve as reviewers of each chapter. Finally, we are deeply grateful for the support and patience shown by our partners, Harley Johnson and Kathryn Clancy, throughout the process of putting this book together, as well as for the joy our children, Elise (7 years) and Will (3 years) Johnson and Joan Clancy-Harley (2 years), bring to our lives. Urbana, IL September 2010

Brendan A.C. Harley A. Wagoner Johnson

Contents

1

2

3

4

5

6

Responses of Cells to Adhesion-Mediated Signals: A Universal Mechanism. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andrew D. Rape, Wei-Hui Guo, and Yu-Li Wang Substrate Elasticity as a Probe to Measure Mechanosensing at Cell-Cell and Cell-Matrix Junctions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jessamine P. Winer, Anant Chopra, J. Yasha Kresh, and Paul A. Janmey

1

11

A Role for Integrin-ECM Bonds as Mechanotransducers that Modulate Adult Stem Cell Fate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nathaniel Huebsch and David J. Mooney

23

Cell-Generated Forces in Tissue Assembly, Function, and Disease. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . John Huynh, Joseph P. Califano, and Cynthia A. Reinhart-King

47

Cell-Cell Interactions and the Mechanics of Cells and Tissues Observed in Bioartificial Tissue Constructs . . . . . . . . . . . . . . Guy M. Genin, Teresa M. Abney, Tetsuro Wakatsuki, and Elliot L. Elson

75

Specific and Non-Specific Adhesion in Cancer Cells with Various Metastatic Potentials. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Xin Tang, Tony Cappa, Theresa Kuhlenschmidt, Mark Kuhlenschmidt, and Taher Saif

vii

viii

Contents

7

Systems Biology of Tumor Cell Migration in 3D: Protein Signaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 Jaya Srivastava and Muhammad H. Zaman

8

Development of Three-Dimensional Tumor Models for the Study of Anti-Cancer Drug Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 Wei Sun, Raj Rajagopalan, and Chwee Teck Lim

9

Mechanobiology of Epidermal Keratinocytes: Desmosomes, Hemidesmosomes, Keratin Intermediate Filaments, and Blistering Skin Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 John C. Selby

10

Quantifying Cell-Matrix Deformations in Three Dimensions . . . . . . . . 211 Christian Franck and Stacey A. Maskarinec

11

Tools for Studying Biomechanical Interactions in Cells . . . . . . . . . . . . . . 233 Rebecca E. Taylor, Vikram Mukundan, and Beth L. Pruitt

12

Biomaterials for Studies in Cellular Mechanotransduction. . . . . . . . . . . 267 Ross DeVolder and Hyunjoon Kong

13

Optical Sensing of Red Blood Cell Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . 279 YongKeun Park, Catherine A. Best, and Gabriel Popescu

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311

Contributors

Teresa M. Abney Department of Mechanical, Aerospace, and Structural Engineering, Washington University, St. Louis, MO, USA Catherine A. Best College of Medicine, University of Illinois at Urbana-Champaign, Urbana, IL, USA Joseph P. Califano Department of Biomedical Engineering, Cornell University, Ithaca, NY, USA Tony Cappa Department of Pathology, College of Veterinary Medicine, University of Illinois at Urbana-Champaign, Urbana, IL, USA Anant Chopra Department of Cardiothoracic Surgery, Drexel University College of Medicine, Philadelphia, PA, USA Ross DeVolder Department of Chemical and Biomolecular Engineering, University of Illinois at Urbana Champaign, Urbana, IL, USA Elliot L. Elson Department of Biochemistry and Molecular Biophysics, Washington University School of Medicine, St. Louis, MO, USA ix

x

Christian Franck School of Engineering, Brown University, Providence, RI, USA Guy M. Genin Department of Mechanical, Aerospace, and Structural Engineering, Washington University, St. Louis, MO, USA Wei-Hui Guo Department of Biomedical Engineering, Carnegie Mellon University, Pittsburgh, PA, USA Nathaniel Huebsch School of Engineering and Applied Sciences, Wyss Institute for Biologically Inspired Engineering, Harvard University, Boston, MA, USA John Huynh Department of Biomedical Engineering, Cornell University, Ithaca, NY, USA Paul A. Janmey Institute for Medicine and Engineering, University of Pennsylvania, Philadelphia, PA, USA Hyunjoon Kong Department of Chemical and Biomolecular Engineering, University of Illinois at Urbana Champaign, Urbana, IL, USA J. Yasha Kresh Department of Cardiothoracic Surgery, Drexel University College of Medicine, Philadelphia, PA, USA Mark Kuhlenschmidt Department of Pathology, College of Veterinary Medicine, University of Illinois at Urbana-Champaign, Urbana, IL, USA

Contributors

Contributors

Theresa Kuhlenschmidt Department of Pathology, College of Veterinary Medicine, University of Illinois at Urbana-Champaign, Urbana, IL, USA Chwee Teck Lim NUS School of Integrative Science and Engineering, National University of Singapore, Singapore; Department of Mechanical Engineering, Division of Bioengineering, National University of Singapore, Singapore; Mechanobiology Institute, National University of Singapore, Singapore Stacey A. Maskarinec Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, CA, USA David J. Mooney School of Engineering and Applied Sciences, Wyss Institute for Biologically Inspired Engineering, Harvard University, Boston, MA, USA Vikram Mukundan Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany YongKeun Park Department of Physics, Korea Advanced Institute of Science and Technology, Daejeon, Republic of Korea Gabriel Popescu Department of Electrical and Computer Engineering, Quantitative Light Imaging Laboratory, Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, USA Beth L. Pruitt Department of Mechanical Engineering and Cardiovascular Institute, Stanford University, Stanford, CA, USA

xi

xii

Raj Rajagopalan NUS School of Integrative Science and Engineering, and Department of Chemical and Biomolecular Engineering, National University of Singapore, Singapore Andrew D. Rape Department of Biomedical Engineering, Carnegie Mellon University, Pittsburgh, PA, USA Cynthia A. Reinhart-King Department of Biomedical Engineering, Cornell University, Ithaca, NY, USA Taher Saif Micro and Nanotechnology Laboratory, Department of Mechanical Science and Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, USA John C. Selby The College of Medicine, Department of Mechanical Science and Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, USA Jaya Srivastava Department of Biomedical Engineering, Boston University, Boston, MA, USA Wei Sun NUS School of Integrative Science and Engineering, National University of Singapore, Singapore Xin Tang Department of Mechanical Science and Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, USA Rebecca E. Taylor Department of Mechanical Engineering and Cardiovascular Institute, Stanford University, Stanford, CA, USA

Contributors

Contributors

Tetsuro Wakatsuki Department of Physiology, Medical College of Wisconsin, Milwaukee, WI, USA Yu-Li Wang Department of Biomedical Engineering, Carnegie Mellon University, Pittsburgh, PA, USA Jessamine P. Winer Institute for Medicine and Engineering, University of Pennsylvania, Philadelphia, PA, USA Muhammad H. Zaman Department of Biomedical Engineering, Boston University, Boston, MA, USA

xiii

Chapter 1

Responses of Cells to Adhesion-Mediated Signals: A Universal Mechanism Andrew D. Rape, Wei-Hui Guo, and Yu-Li Wang

This chapter is part of Section I: Mechanisms of Cell Adhesion and Mechanotransduction

Abstract Cells are exposed to a plethora of signals that typically coerce them to function properly, but aberrant signaling can lead to pathological conditions. In the treatment of diseases and the rational design of functioning tissues, it is vital to understand and be able to manipulate these inputs. In the past, much of the interest has been on chemical signaling but recently, there has been an explosion of research into a diverse array of mechanical signals. Mechanical signals have been shown to influence cellular growth, survival, migration, and differentiation. Despite its obvious importance, relatively little is known about the mechanism of mechanosensing. In this chapter, we describe what is currently known about potential mechanosensing molecules and then describe a model by which a wide array of mechanical signals can be interpreted by a common mechanism. By understanding this mechanism, one may be able to develop new therapeutic interventions for devastating diseases such as cancer and break through critical barriers facing the field of tissue engineering. We expect the knowledge gained from the study of basic biology to greatly impact the treatment of many patients in the clinical setting in the coming years. Keywords Mechanical signals Traction forces Durotaxis Focal adhesions Cancer

1.1

Introduction

Cellular behavior including growth, survival, migration, and differentiation is regulated by the complex interplay between cells and their environment. While much attention has been focused on chemical factors, it is becoming increasingly

Y.-L. Wang (*) Department of Biomedical Engineering, Carnegie Mellon University, 5000 Forbes Avenue, Pittsburgh, PA 15213, USA e-mail: [email protected]

A. Wagoner Johnson and Brendan A.C. Harley (eds.), Mechanobiology of Cell-Cell and Cell-Matrix Interactions, DOI 10.1007/978-1-4419-8083-0_1, # Springer Science+Business Media, LLC 2011

1

2

A.D. Rape et al.

evident that adhesion-mediated non-chemical signals such as mechanical forces and topography can play an equally important, complementary role. However, despite a large volume of phenomenological records, our understanding of the mechanisms of cellular responses to these signals remains fragmentary. While some signals such as mechanical forces may interact directly with intracellular components [1], many other signals are likely converted into intracellular chemical events near the plasma membrane where adhesion takes place. Adhesion-mediated non-chemical signals take a variety of forms. Applied mechanical forces are able to cause behavioral responses both directly through intracellular signaling and indirectly through changes in gene expression [2, 3]. Equally important extracellular signals include rigidity [4, 5], shape [6, 7], and topography [8]. For example, fibroblasts cultured on soft substrates undergo apoptosis while those on rigid substrates show enhanced growth [9]. In addition, the differentiation of mesenchymal stems cells in vitro appears to be dictated by substrate rigidity [10]. Interestingly, shape and geometry are able to elicit similar responses as mechanical forces. Adhesive cells show active growth when allowed to spread without constraints, and undergo apoptosis when inhibited from spreading [6]. Osteogenic cell differentiation is favored only within a range of spreading areas [7, 11], while adipogenic cells fail to differentiate when allowed to spread fully [12]. The fate of mesenchymal stem cells can similarly be directed via shape constraints [13]. Cell migration and cytoskeletal structures also respond to a similar set of adhesion-mediated signals. Contact guidance was discovered decades ago as the orientation of adhesive cells and their actin cytoskeleton along micrometer sized grooves [8]. In addition, motile adhesive cells were found to orient toward tensile forces [3, 14], while migrating fibroblasts turn preferentially toward stiff substrates, a phenomenon known as durotaxis [14]. Among different adhesion-mediated signals, mechanical forces are the best understood. While it is possible that each type of sensing uses separate mechanisms, various observations suggest that there may be a universal mechanism for sensing diverse forms of adhesion-mediated signals. We will first discuss the potential mechanism of force sensing, then propose a common force-dependent sensing mechanism that, with proper positive and negative feedback loops, may function universally for sensing a wide range of signals.

1.2

Mechanisms for Sensing Mechanical Forces

Mechanical forces may induce transmembrane signals by triggering the entry of calcium ions through stretch-activated channels [15–17], and/or by inducing structural changes at adhesion sites. For adherent cells, focal adhesions have been the focus of attention as they are both the direct link connecting the cell’s cytoskeleton to the extracellular matrix [18], and the site of concentration of important signal transduction enzymes such as the Src kinase and FAK [19]. Responses to

1 Responses of Cells to Adhesion-Mediated Signals

3

Fig. 1.1 A potential signal transduction mechanism based on force-induced conformational change. Pulling forces generated by the actin cytoskeleton cause the unfolding of associated proteins, which may expose binding sites for regulatory factors

mechanical forces include activation of the small GTPase, RhoA, coupled to an increase in the size of focal adhesions and enhancement of intracellular contractility and traction forces [20, 21]. How mechanical forces modulate the activities of signaling enzymes remains to be an area of active investigation. Integrins – the membrane spanning component of focal adhesions – were first hypothesized to be mechanosensors [22–24]. Integrin clustering, ECM-binding, and recruitment of focal adhesion proteins are known to be enhanced by mechanical forces [21, 25, 26]. Thus mechanosensing may involve the activation of integrins and the resulting concentration of focal adhesion kinase (FAK) and Src [27, 28]. A second possible mechanism involves inherent mechanosensitivities of these signaling molecules. Mechanical forces may directly induce conformational changes and expose autoregulated catalytic domains, shielded substrate domains, and/or cryptic binding sites of scaffold proteins (Fig. 1.1). Focal adhesion proteins, such as vinculin [29], may change their conformation in response to mechanical input. In addition, stretching of p130Cas induces a conformational change that enhances its phosphorylation by the Src kinase [30], which in turn activates the recruitment of binding partners including many small GTPases. Force-induced structural changes may occur not only at the intra-molecular but also inter-molecular level [31]. Focal adhesion-associated actin filaments show force-dependent assembly and retrograde flux. Differential association of focal adhesion components to actin filaments may then lead to differential transport and relative shear movements of these components (Fig. 1.2). Thus, interactions among focal adhesion proteins may be regulated according to their relative affinity for integrins/membrane components vs. the actin cytoskeleton. While an upstream event is required to regulate actin flux, this mechanism may serve to amplify the responses. In addition to direct responses to external forces, positive and negative feedback loops are likely to play an important role. A combination of local positive feedback and global negative feedback has long been recognized as a key component for the extreme sensitivity of chemotaxis [32–34]. A similar mechanism may function

4

A.D. Rape et al.

Fig. 1.2 A potential signal transduction mechanism based on force-induced protein shear movements. Pulling forces generated by the actin cytoskeleton cause lateral shear and relative movements of focal adhesion proteins, which may alter protein-protein interactions and affect enzymatic activities within focal adhesions

constitutively in mechanosensing. The so-called “inside-out” signaling may in fact represent the positive feedback mechanism. Inside-out signaling was first recognized as actin cytoskeleton and contractility-dependent enhancement of integrin-ECM interactions [35]. The process may include both transmembrane activation of integrins and force-induced conformational changes of fibronectin to expose cryptic binding sites [36, 37]. The latter then causes fibronectin molecules to change conformation and form a long, multi-molecular fibrillar structure [38], and enhances the mechanical input from the matrix. Together, these responses create a positive feedback loop that, upon the initial response to mechanical stimulations, increases the cytoskeletal contractility and further enhances mechanical stimulations and/or responses (Fig. 1.3).

1.3

A Universal Sensor for Diverse Adhesion-Mediated Signals

While passive force-sensing may be responsive only to external mechanical forces, a highly versatile sensing mechanism may be created by incorporating internal contractile forces as part of the mechanism. In addition to mediating inside-out signaling as discussed above, when transmitted to the extracellular matrix, such forces may function as probing forces for parameters such as substrate elasticity, cell shape, and size. This notion is further supported by the similarity of cellular responses to these diverse signals as discussed earlier. Substrate elasticity is characterized by the Young’s modulus, which under ideal conditions is a proportional constant between applied stress and the resulting strain (deformation). As a cell actively applies increasing forces on the substrate, both the strain and resistive counter forces increase. While the cell may use either the resistive force for a given deformation, or the deformation under a given

1 Responses of Cells to Adhesion-Mediated Signals

5

Fig. 1.3 Interplay between physical signals and chemical signaling pathways. Physical signals such as substrate stiffness, applied mechanical forces, and topographic cues activate intracellular signaling pathways and stimulate contractility (Outside-In signals). Increased contractility in turn amplifies focal adhesions and enhances signal sensitivity (Inside-Out signals). This positive feedback loop can then lead to large changes in cell motility, growth, and differentiation. From Discher et al. [4]

probing (and counter) force, for the detection of elasticity, one study supports the former mechanism by showing a relatively constant deformation of the substrate irrespective of its elasticity [39]. However this does not necessarily indicate that intracellular structures undergo a constant deformation. To the contrary, intracellular deformation is determined by the magnitude of counter forces against probing forces, and both the counter force and intracellular deformation must increase under a constant amount of substrate deformation as its rigidity increases (Fig. 1.4).

6

A.D. Rape et al.

Fig. 1.4 A universal model for the detection of physical and topographical signals. Focal adhesions are represented as parallelograms of different colors and skews. Cell migration and soft substrates share a common feature of minimizing structural changes at focal adhesions, due to the forward movement of the cell body or the backward movement of the substrate that reduces tension at focal adhesions. Cell immobilization, a stiff substrate, or a pulling force transmitted through flexible substrates causes an opposite effect and generates strong tension at focal adhesions. Thus a common mechanism may be able to sense a wide range of signals. Adapted from Guo and Wang [31]

The same mechanism may also be used for the detection of a cell’s own shape, spreading and migration. Translocation of the cell body cancels the deformation of a mechanically coupled substrate and cell body, thereby diminishing the signal. In addition, as an elastic self-spreading object, adhesive cells must generate increasing forces against the substrate to propel an increasing extent of spreading. Responses to the corresponding, increasing counter forces may then allow a cell to detect its extent of spreading. An adherent cell may further detect its own shape through a combination of local positive feedback to enhance activities in the extended region, where mechanical input is strong, and global negative feedback to suppress activities elsewhere. Importantly, these explanations are consistent with experimental observations. For example, it has been determined that focal adhesion size, cell spreading area, and traction force increase as a function of substrate stiffness [40]. The model also explains the phenomenon of durotaxis. Focal adhesions are reinforced in stiffer regions causing an increase in local forces, which in turn causes the cell to migrate away from soft regions. In addition, consistent with a force-based mechanism for the detection of spreading, cells confined either by micropatterning or decreasing ligand density exert markedly reduced forces on the substrate [41, 42]. However, although this model can succinctly explain the sensing of many forms of adhesionmediated signals, much needs to be done to gain a full understanding of the process particularly the initial mechanosensitive events and the mechanisms of feedback.

1 Responses of Cells to Adhesion-Mediated Signals

1.4

7

Implications in Biomedical Engineering and Disease Treatment

Despite the still limited mechanistic understanding of mechanotransduction, tissue engineers and pathologists have already realized the profound implications of cellular responses to various forms of adhesion-mediated signals in disease treatment and regenerative medicine. Mechanical sensing may add a critical dimension to the effective treatment of cancer, which generally fails due to uncontrolled cell growth and migration – both are regulated by chemicals as well as adhesion-mediated signals. Two aspects in particular contribute to the disease phenotype of cancer. First is the well known loss of anchorage dependence, defined as the need for most normal, nonhematopoietic cells to adhere firmly to a surface in order to survive [8, 43, 44]. Loss of anchorage dependence may allow cancer cells to survive following the penetration through the vasculature, and to float through the bloodstream before reaching distant colonization sites. Equally important may be the increase in stiffness in many tumors relative to their normal counterpart or the surrounding tissues [45–47]. Conversely, it was found that increased ECM stiffness acted to promote tumorigenesis in an integrin and cytoskeletal contractility dependent manner [48, 49]. The pathology may involve two potential aspects. First is the possible stimulation of extracellular matrix production/assembly by the surrounding fibroblasts, as evidenced by the involvement of “carcinoma-associated fibroblasts” in cancer progression [50]. Second is the stimulated durotaxis as a result of increased stiffness, which may cause tumor cells to migrate away from the home tissue, and blood vessels to grow into the tumor to provide nutrients. Clearly mechanotransduction pathways offer exciting new targets for cancer therapy. Insights from basic mechanobiology research will also facilitate engineering control of man-made tissues in regenerative medicine, a field with seemingly unlimited potential but challenged by limited success thus far. To guide stem cells toward desirable differentiation pathways and target sites, tissue engineers have begun to realize the critical need to control not only the chemical environment but also parameters such as stiffness and topography of the surrounding adhesive materials. For example, consistent with the tendency of adherent cells to disperse on stiff substrates and to form tissue-like aggregates on soft matrices [51], stem cells form organoids only under minimal mechanical input such as in hanging drops, and grow as spread monolayer on conventional tissue culture dishes [52]. Thus not only is it important to select scaffold materials of proper mechanical characteristics, but to develop new “smart” materials that allow the modulation of these parameters spatially and temporally. Complex tissues may be successfully engineered only through the regulation of mechanical and topographic environment at a matching complexity. In summary, knowledge in basic cellular mechanotransduction is finding rapid translation into medical applications. Conversely, lessons learned from clinical outcomes of cellular mechanical manipulations may complement basic research

8

A.D. Rape et al.

in understanding both normal mechanisms and pathological defects. Such interplay between basic “translatable” research and clinical research is likely to lead to significant breakthroughs and make the coming decade an exciting time for fruitful manipulations of cellular adhesive interactions.

References 1. Wang N, Tytell JD, Ingber DE (2009) Mechanotransduction at a distance: mechanically coupling the extracellular matrix with the nucleus. Nat. Rev. Mol. Cell Biol. 10:75–82. 2. Bershadsky AD, Balaban NQ, Geiger B (2003) Adhesion-dependent cell mechanosenstivity. Annu. Rev. Cell Dev. Biol. 19:677–695. 3. Wang JH-C, Thampatty BP (2006) An introductory review of cell mechanobiology. Biomech. Model. Mechanobiol. 5:1–16. 4. Discher DE, Janmey P, Wang YL (2005) Tissue cells feel and respond to the stiffness of their substrate. Science 310:1139–1143. 5. Mammoto A, Connor KM, Mammoto T, Yung CW, Huh D, Aderman CM, Mostoslavsky G, Smith LE, Ingber DE (2009) A mechanosensitive transcriptional mechanism that controls angiogenesis. Nature 457:1103–1108. 6. Chen CS, Mrksich M, Huang S, Whitesides GM, Ingber DE (1997) Geometric control of cell life and death. Science 276(5317):1425–1428. 7. Thomas CH, Collier JH, Sfeir CS, Healy KE (2002) Engineering gene expression and protein synthesis by modulating nuclear shape. Proc. Natl. Acad. Sci. U.S.A. 99:1972–1977. 8. Curtis A, Wilkinson C (1997) Topographical control of cells. Biomaterials 18:1573–1583. 9. Wang HB, Dembo M, Wang YL (2000) Substrate flexibility regulates growth and apoptosis of normal but not transformed cells. Am. J. Physiol. 279:C1345–C1350. 10. Engler AJ, Sen S, Sweeney HL, Discher DE (2006) Matrix elasticity directs stem cell lineage specification. Cell 126:677–689. 11. Carvalho RS, Schaffer JL, Gerstenfeld LC (1998) Osteoblasts induce osteopontin expression in response to attachment on fibronectin: demonstration of a common role for integrin receptors in the signal transduction processes of cell attachment and mechanical stimulation, J. Cell. Biochem. 70:376–390. 12. Spiegelman BM, Ginty GA (1983) Fibronectin modulation of cell shape and lipogenic gene expression in 3T3-adipocytes. Cell 35:657–666. 13. McBeath R, Pirone DM, Nelson CM, Bhadriraju K, Chen CS (2004) Cell shape, cytoskeletal tension, and RhoA regulate stem cell lineage commitment. Dev. Cell 6:483–495. 14. Lo CM, Wang HB, Dembo M, Wang, YL (2000) Cell movement is guided by the rigidity of the substrates. Biophys. J. 79:144–152. 15. Lee J, Ishihara A, Oxford G, Johnson B, Jacobson K (1999) Regulation of cell movement is mediated by stretch-activated calcium channels. Nature 400:382–386. 16. Sigurdson W, Ruknudin A, Sachs F (1992) Calcium imaging of mechanically induced fluxes in tissue cultured chick heart: role of stretch activated ion channels. Am. J. Physiol. Heart Circ. Physiol. 262:H1110–H1115. 17. Munevar S, Wang YL, Dembo M (2004) Regulation of mechanical interaction between fibroblasts and the substratum by stretch activated Ca2+ entry. J. Cell Sci. 117:85–92. 18. Geiger B, Spatz JP, Bershadsky AD (2009) Environmental sensing through focal adhesions. Nat. Rev. Mol. Cell. Biol. 10:21–33. 19. Carragher NO, Frame MC (2004) Focal adhesion and actin dynamics: a place where kinases and proteases meet to promote invasion. Trends Cell Biol. 14:24–249. 20. Chen CS, Tan J, Tien J (2004) Mechanotransduction at cell-matrix and cell-cell contacts. Annu. Rev. Biomed. Eng. 6:275–302.

1 Responses of Cells to Adhesion-Mediated Signals

9

21. Riveline D, Zamir E, Balaban NQ, Schwarz US, Ishizaki T, Narumiya S, Kam Z, Geiger B, Bershadsky AD (2001) Focal contacts as mechanosensors: externally applied local mechanical force induces growth of focal contacts by an mDia1-dependent and ROCK-independent mechanism. J. Cell Biol. 153:1175–1186. 22. Wang N, Butler JP, Ingber DE (1993) Mechanotransduction across the cell surface and through the cytoskeleton. Science 260:1124–1127. 23. Choquet D, Felsenfeld DP, Sheetz MP (1997) Extracellular matrix rigidity causes strengthening of integrin-cytoskeleton linkages. Cell 88:39–48. 24. Katsumi A et al (2004) Integrins in mechanotransduction. J. Biol. Chem. 279:12001–12004. 25. Garcia AJ, Huber F, Boettiger D (1998) Force required to break a5b1 integrin-fibronectin bonds in intact adherent cells is sensitive to integrin activation state. J. Biol. Chem. 273 (18):10988–10993. 26. Paszek MJ, Boettiger D, Weaver VM, Hammer DA (2009) Integrin clustering is driven by mechanical resistance from the glycocalyx and the substrate. PLoS Comput. Biol. 5:1–16. 27. Wang HB, Dembo M, Wang YL (2001) Focal adhesion kinase is involved in mechanosensing during fibroblast migration. Proc. Natl. Acad. Sci. U.S.A. 98:11295–11300. 28. Wang Y et al (2005) Visualizing the mechanical activation of Src. Nature 434:1040–1045. 29. Johnson RP, Craig SW (1995) F-actin binding site masked by the intramolecular association of vinculin head and tail domains. Nature 373:261–264. 30. Sawada Y et al (2006) Force sensing by mechanical extension of the Src family kinase substrate p130Cas. Cell 127:1015–1026. 31. Guo WH, Wang YL (2007) Retrograde fluxes of focal adhesion proteins in response to cell migration and mechanical signals. Mol. Biol. Cell 18:4519–4527. 32. Meinhardt H (1999) Orientation of chemotactic cells and growth cones: models and mechanisms. J. Cell Sci. 112:2867–2874. 33. Rappel WJ, Thomas PJ, Levine H, Loomis WF (2002) Establishing direction during chemotaxis in eukaryotic cells. Biophys. J. 83:1361–1367. 34. Postma M, Bosgraaf L, Loovers HM, Van Haastert PJM (2004) Chemotaxis: signalling modules join hands at front and tail. EMBO Rep. 5:35–40. 35. Schwartz MA, Schaller MD, Ginsberg MH (1995) Integrins: emerging paradigms of signal transduction. Annu. Rev. Cell. Dev. Biol. 11:549–599. 36. Zhong CL, Chrzanowska-Wodnicka M, Brown J, Shaub A, Belkin AM, Burridge K (1998) Rho-mediated contractility exposes a cryptic site in fibronectin and induces fibronectin matrix assembly. J. Cell Biol. 141:539–551. 37. Baneyx G, Baugh L, Vogel V (2002) Fibronectin extension and unfolding within cell matrix fibrils controlled by cytoskeletal tension. Proc. Natl. Acad. Sci. U.S.A. 99:5139–5143. 38. Wierzbicka-Patynowski I, Schwarzbauer JE (2003) The ins and outs of fibronectin matrix assembly. J. Cell Sci. 116:3269–3276. 39. Saez A, Buguin A, Silberzan P, Ladoux B (2005) Is the mechanical activity of epithelial cells controlled by deformations or forces? Biophys. J. 89:L52–L54 40. Pelham RJ Jr., Wang YL (1997) Cell locomotion and focal adhesions are regulated by substrate flexibility. Proc. Natl. Acad. Sci. U.S.A. 94:13661–13665. 41. Wang N, Ostuni E, Whitesides G, Ingber DE (2002) Micropatterning tractional forces in living cells. Cell Motil. Cytoskeleton 52:97–106. 42. Reinhart-King CA, Dembo M, Hammer DA (2003) Endothelial cell traction forces on RGDderivatized polyacrylamide substrata. Langmuir 19:1573–1579. 43. Danen EHJ, Yamada KM (2001) Fibronectin, integrins, and growth control. J. Cell Physiol. 189:1–13. 44. Schwartz MA (1997) Integrins, oncogenes, and anchorage independence. J. Cell Biol. 139:575–578. 45. Paszek MJ, Weaver VM (2004) The tension mounts: mechanics meets morphogenesis and malignancy. J. Mammary Gland Biol. Neoplasia 9:325–342.

10

A.D. Rape et al.

46. Levental I, Georges PC, Janmey PA (2007) Soft biological materials and their impact on cell function. Soft Matter 3:299–306. 47. Huang S, Ingber DE (2005) Cell tension, matrix mechanics, and cancer development. Cancer Cell 8:175–176. 48. Paszek MJ, Zahir N, Johnson KR, Lakins JN, Rozenberg GI, Gefen A, Reinhart-King CA, Margulies SS, Dembo M, Boettiger D, Hammer DA, Weaver VM (2005) Tensional homeostasis and the malignant phenotype. Cancer Cell 8:241–254. 49. Levental KR, Yu H, Kass L, Lakins JN, Egeblad M, Erler JT, Fong SF, Csiszar K, Giaccia A, Weninger W, Yamauchi M, Gasser DL, Weaver VM (2009) Matrix crosslinking forces tumor progression by enhancing integrin signaling. Cell 139:891–906. 50. Orimo A, Gupta PB, Sgroi DC, Arenzana-Seisdedos F, Delaunay T, Naeem R, Carey VJ, Richardson AL, Weinberg RA (2005) Stromal fibroblasts present in invasive human breast carcinomas promote tumor growth and angiogenesis through elevated SDF-1/CXCL12 secretion. Cell 121:335–348. 51. Guo WH, Frey MT, Burnham NA, Wang YL (2006) Substrate rigidity regulates the formation and maintenance of tissues. Biophys. J. 90:2213–2220. 52. Smith AG (2001) Embryo-derived stem cells: of mice and men. Annu. Rev. Cell Dev. Biol. 17:435–462.

Chapter 2

Substrate Elasticity as a Probe to Measure Mechanosensing at Cell-Cell and Cell-Matrix Junctions Jessamine P. Winer, Anant Chopra, J. Yasha Kresh, and Paul A. Janmey

This chapter is part of Section I: Mechanisms of Cell Adhesion and Mechanotransduction

Abstract In vivo, most cells are mechanically and chemically connected to other cells or to a variety of polymeric networks generically called the extracellular matrix (ECM). Adhesive contacts are formed by distinct classes of transmembrane protein complexes that have specific binding sites for extracellular targets on one side of the membrane and cytoplasmic domains that engage specific elements of the cytoskeleton and signal transduction systems. Engagement of cell-cell or cellmatrix contact both initiates and depends on mechanical signaling from inside and outside the cell, but also depends on the forces generated at the cell-cell or cell-ECM junction. This chapter will summarize some recent studies of mechanotransduction at cell adhesion sites and present examples of the interplay between cell-cell or cell-matrix contacts in fibroblasts, endothelial cells, cardiac myocytes, T lymphocytes and other cell types.

2.1

Introduction

With a few exceptions such as erythrocytes, nearly all cells in multicellular organisms are bound either to other cells or to extracellular matrices. Even single cell organisms such as yeast and bacteria form contacts between soft inner membranes and stiffer outer coats, or to the cells they invade. Such contacts are not only adhesive, they are also focal points for concentrating signaling proteins and lipids and help organize the architecture of the plasma membrane and the cytoskeleton. At the center of such contacts is one of several large classes of transmembrane proteins, bound at the outside to the matrix or to another cell and at the inside to a number of cytosolic proteins that bind either cytoskeletal elements or signal transduction intermediates (Fig. 2.1). Not all signaling that initiates at adhesion sites is triggered by ligation of the transmembrane receptors alone, and in recent

P.A. Janmey (*) Institute for Medicine and Engineering, University of Pennsylvania, Philadelphia, PA, USA e-mail: [email protected]

A. Wagoner Johnson and Brendan A.C. Harley (eds.), Mechanobiology of Cell-Cell and Cell-Matrix Interactions, DOI 10.1007/978-1-4419-8083-0_2, # Springer Science+Business Media, LLC 2011

11

12

J.P. Winer et al.

Fig. 2.1 Cell-cell and cell-ECM adhesion complexes. Various transmembrane proteins with extracellular domains specific for ligands expressed either at cell membranes or the extracellular matrix are linked to elastic elements with different mechanical properties depicted as springs with different stiffness

years it has become clear that forces generated at or applied to sites of cell-cell or cell-matrix contacts have important effects on cell structure and function on both the single cell [1, 2] and whole tissue level [3]. Transmitting force within the cell and transducing force to biochemical reactions that engage intracellular signaling pathways are important aspects of mechanobiology, and recognition of external force or of resistance to internally generated force has been termed mechanosensing. This chapter will emphasize one aspect of mechanosensing, the response of cells to the stiffness of the material to which they adhere. Mechanosensing through cell-cell and cell-matrix adhesion complexes are distinct processes that are mediated by distinct subsets of proteins and respond to different levels of force, displacement, or stiffness. A few example of mechanosensing will be discussed in an effort to suggest unifying themes that might reveal the mechanism of these processes.

2.2

Mechanosensing by Cell-Matrix Adhesions

The importance of the mechanical properties of the extracellular matrix (ECM) for cell morphology [4], motility [5] and differentiation [6] has been known for decades and recently re-emphasized as an essential control parameter in vivo distinct from and coordinated with biochemical signaling. Most cells adhere to extracellular matrices or to cell culture substrates using one or more of a large class of integrins, which are heterodimeric transmembrane complexes with variable affinity and specificity to ECM proteins such as collagen,

2 Measuring Mechanosensing at Cell-Cell and Cell-Matrix Junctions

13

fibronectin and laminin. Some integrins bind cell surface proteins such as ICAMs to mediate cell-cell contacts that differ in structure and function from those formed by cadherins. Integrins are heterodimers consisting of an alpha and a beta subunit and the ability of an integrin dimer to bind a particular protein depends on the composition of the dimer. For instance a1b2 dimers primarily bind collagen whereas avb3 dimers bind multiple proteins such as fibronectin containing the amino acid sequence RGD. In addition to having different extracellular binding partners, different integrins engage different cytoskeletal proteins and trigger different signaling pathways (Fig. 2.1). Many of the first studies demonstrating the effects of substrate stiffness on the phenotype of cells bound by integrins to deformable surfaces were done using cultured fibroblast or epithelial cell lines [7], but also early passage primary cells such as vascular smooth muscle cells [8] and mesenchymal stem cells [9]. The results of early studies are summarized in recent reviews [10–12]. As the mechanical responses of more cell types have been studied, one pattern that emerges is that changes in properties such as spread area, cytoskeletal structure, proliferation, differentiation [13–15], or cell stiffness [16, 17] do not necessarily change linearly or even monotonically with substrate stiffness. Rather, these responses saturate at different substrate stiffnesses or show maxima or minima at intermediate stiffnesses that approximate those of the tissue from which the cells were derived [18]. For example, neonatal rat heart cells exhibit optimal morphology and function when cultured on collagen-coated gels with intermediate stiffness (10–20 kPa) that approximates the stiffness of adult rat myocardium [19] and neonatal rat ventricular myocytes also produce a more nearly native phenotype when the substrate stiffness is within this intermediate range [20–22]. Cell motility is also reported to reach a maximum at intermediate substrate stiffness in a study of neutrophil migration on gels coated with fibronectin [23], and in several cases cell motility is directed to regions of increased substrate stiffness, as shown in studies using gels with stiffness gradients [24–26]. In many cases the trends observed with increasing stiffness in 2D systems are also observed in 3D systems of the same elastic modulus such as the differential growth of neurons and astrocytes grown in fibrin [27, 28] or the stiffening of endothelial cells grown in collagen gels [29], but in other cases cell responses to simple linear elastic gels like polyacrylamide that are coated with adhesion proteins are different from those of the same cell type bound to or within a 3D network made of the same protein [30]. Differences in morphology of cells bound in 2D or 3D are also strongly affected by formation of dorsal cell-ECM adhesions. A rapid, substrate stiffness-dependent transition from a well-spread, flat morphology to an elongated bipolar or stellate morphology closer to the structure in vivo occurs when a second ECM surface is placed on top of fibroblasts initially cultured in 2D [31, 32]. Stiffness responses can depend strongly on the nature of the adhesive ligand and therefore the type of integrin that engages the substrate. For example, fibroblasts show a much stronger stiffness-dependent morphology when spread on substrates coated with both collagen I and collagen V than compared to surfaces coated with collagen I alone [33]. Melanoma cells increase spread area with increasing substrate

14

J.P. Winer et al.

Fig. 2.2 Stiffness-dependent cell spreading and stiffening depends on the type of adhesion complex. Cellular stiffness measured by AFM (a) and adherent area (b) of A7 melanoma cells cultured for 24 h on polyacrylamide gels laminated with collagen I or fibronectin or a mixture of collagen I and fibronectin. Both proteins were added as saturating concentrations to the gels using methods described in [65]

stiffness whether the substrates are coated with collagen I or fibronectin, but only change their own stiffness to match that of the substrate when they adhere through collagen receptors [16]. Stiffening of cells bound by collagen receptors requires the function of the actin crosslinker filamin A, but spreading of cells bound by fibronectin receptors is evidently mediated by other cytoskeletal linkers [29, 34]. The specificity of different integrins and different integrin ligands in mechanotransduction from the ECM to the cell is illustrated in Fig. 2.2, where the stiffening and spreading of filamin A-expressing A7 human melanoma cells is compared on substrates with different stiffnesses that are coated with either fibronectin or collagen 1. Stiffening (Fig. 2.2a) but not spreading (Fig. 2.2b) of A7 melanoma cells depends very strongly on whether integrins specific for fibronectin or collagen are engaged. When A7 cells are plated on gels coated with saturating amounts of either Fn or collagen I, they spread to approximately the same extent, but the cells on collagen 1 are much stiffer than those adherent to Fn. When both Fn and collagen I are present, allowing both beta 1 and beta 3 integrins to bind, adherent area increases, but cell stiffness reaches an value intermediate between those on Fn or collagen I alone. Specificity in the mechanotransduction through different integrins has been observed in various other contexts. Using a spinning disk assay to apply shear stresses to cells, it has been shown that the residues flanking the RGD sequence influence how strongly the cells adhere to the substrate [35] and that adhesivity increases with cell spread area and time of adhesion [36]. When challenged by shear stress caused by fluid flow, osteoblasts adhere more strongly to fibronectin coated substrates than to vitronectin coated surfaces and least strongly to collagen coated substrates [37].

2 Measuring Mechanosensing at Cell-Cell and Cell-Matrix Junctions

15

Quantitative measurements of the forces exerted or resisted by single clusters of integrins bound to specific ligands have emerged from studies using atomic force microscopy. In one study, a fibronectin coated polystyrene bead was attached to a cantilever and brought in contact with a smooth muscle cells for 2, 5 or 8 min and then withdrawn. Three characteristic release forces of 40, 55 and 80 pN were measured, and adhesion could be blocked by adding antibodies blocking the function of a5b1 integrins prior to adhesion [38]. In another study CHO cells were allowed to spread on an AFM cantilever and the cell-coated tip was used to probe a collagen-I coated substrate. In this study the time allowed for adhesion was varied from 200 ms to 10 min. The 200 ms time point was used to measure a force of 50 pN for binding of a single integrin that is not reinforced by other adhesion proteins or the actin cytoskeleton. Allowing the cells to adhere for an extended time increased the release force to 500 pN [39]. Surprisingly this larger adhesion force was not affected by inhibition of actomyosin contraction and appears to result from integrin clustering and formation of large adhesion plaques.

2.3

Mechanosensing Through Cell-Cell Adhesions

Mechanical control of cell-cell contacts is mediated by homophilic binds between cadherins or the immunoglobulin superfamily glycoprotein N-CAM or by heterophilic contacts between I-CAM and integrins such as LFA1 [40]. The importance of cell mechanics for the function of cell-cell contact through cadherins was demonstrated by studies showing how differences in effective surface tensions, mediated in part by cadherin-cadherin binding energy, can drive cell sorting [41, 42]. The resistance of cell-cell contacts to disruption by force, especially for cadherins, has been fairly extensively studied (e.g. [43]). Cadherins are clearly implicated in mechanosensing, for example by transmitting signals from the plasma membrane to distant sites within the cell [44] and by reacting to applied force by increasing local cell adhesion size [45], but mechanical signaling through cell-cell contacts or the response of these contacts to differences in stiffness are relatively understudied compared to analogous studies of cell-ECM adhesions. At the neuromuscular junction in Drosophila, vesicle clustering at the presynaptic terminal depends on mechanical tension within the axons that leads to stress at the interface connecting the neuron to the muscle [46]. When this junction is mechanically broken, vesicle movements are randomized, but not prevented, and the axonal cytoskeleton is disrupted. Application of nN forces to the tip of the severed axon restores both cytoskeletal structure and anterograde bias to vesicle movements [46]. T lymphocyte activation has also been shown to be strongly dependent on force applied to the bond between the lymphocyte T-cell receptor complex (TCR) and the MHC-agonist peptide complex at the surface of an antigen-presenting cell [47, 48]. ICAM-LFA1 bonds and occupation of the TCR with an agonist peptide are each necessary but not sufficient for T-cell activation. The combination of both cell-cell contacts, along with force production by the lymphocyte on the

16

J.P. Winer et al.

antigen-presenting cell surface is hypothesized to be required for full activation of the T-cell response. The mechanical resistance provided by an antigen-presenting cell, or a fixed substrate, to the activated TCR might be essential to prevent inappropriate triggering of T-cells by soluble ligands. Quantitative measurements of the forces applied at sites of binding to cadherins and the dependence of cadherin signaling to the cytoskeleton on mechanical compliance are beginning to be elucidated using micropost and soft gel systems originally developed to examine cell-substrate effects. C2 myogenic cells cultured on polydimethylsiloxane (PDMS) microposts coated with an FC-N-cadherin construct that efficiently binds cellular N-cadherin developed stresses similar to those generated by the same cells bound to the posts through fibronectin receptors [49]. As the stiffness of N-cadherin-coated substrates increases, the size of the cell’s cadherin adhesion also increases as does the magnitude of force generation [50]. As a result, the morphology of the cell ranges from well spread on stiff substrates (~100 kPa) to poorly spread and rounded on soft substrates (~10 Pa) regardless of whether the substrates were PDMS pillars or polyacrylamide gels [50]. Similar effects are also seen when neonatal rat cardiomyocytes are cultured on N-cadherin coated gels [20], although in this case there is a pronounced optimal stiffness at which these cells spontaneously elongate and form myofibrils, as seen in Fig. 2.3. On soft N-cadherin-coated polyacrylamide gels (shear modulus G ~ 100 Pa) neonatal ventricular cardiomyocytes exhibit a rounded morphology,

Fig. 2.3 Single cardiomyocytes on N-Cadherin-coated substrates of varying stiffness display differential morphology

2 Measuring Mechanosensing at Cell-Cell and Cell-Matrix Junctions

17

with poor sarcomere organization. At physiological stiffness (G ~ 5, 10 kPa) cells exhibit striated F-actin and sarcomeric alpha actinin staining and a high aspect ratio. On stiff surfaces (G ~ 30 kPa, glass) cells exhibit prominent F-actin filaments devoid of striations and a polygonal shape. The optimal stiffness resembles that of the native tissue and is similar to that first reported for myotubes using gels coated with ECM proteins [18]. Although the qualitative effects of cell-cell and cell-ECM adhesions are similar using substrates of different stiffness, there are significant quantitative differences that likely reflect the different signaling through cadherins and integrins, and the different spatial distributions of these transmembrane proteins in the native tissue.

2.4

Crosstalk Between Cell-Matrix and Cell-Cell Mechanical Signaling

Signaling, adhesion, and stiffness [51] at cell-cell and cell-ECM junctions are accomplished by different mechanisms and sets of proteins, but the two systems have significant influence on each other [52]. For example, when endothelial cells are cultured on ECM protein-coated gels, their morphologies depend very strongly on substrate stiffness as long as the cells are subconfluent. However, when the cells also make cell-cell contacts, actin bundles resembling stress fibers can form in cells that touch each other but not on single cells on soft substrates, as shown in Fig. 2.4. When endothelial cells become confluent, their gross morphology is no longer apparently dependent on the stiffness of the substrate beneath them [53].

Fig. 2.4 Redistribution of actin cytoskeleton when cells on soft substrates make cell-cell contact. NIH 3T3 fibroblasts on soft gels (180 Pa) with F-actin stained by rhodamine phalloidin. The isolated fibroblast (left) appears to have no stress fibers. When the fibroblasts are able to make cell-cell contact (right), stress fibers form

18

J.P. Winer et al.

In some cases activation of one class of contacts enhances formation of the other, and in other cases signaling between cell-cell and cell-ECM contacts appears to be antagonistic. For example, engagement of integrins increases the strength of E-cadherin junctions in cultured fibroblasts and carcinoma cell lines by a mechanism that involves the Src kinase [54]. Cell-cell contacts between epithelial cells can also increase the force applied to cell-ECM contacts, as observed when clusters of MDCK cells cultured on fibronectin-coated microposts apply greater traction force than individual cells even though only one cell is pulling on an individual post at a time [55]. Similarly, mesodermal cells from the developing Xenopus laevis embryos apply stronger tractions (25 kPa) when in small clusters (4–7 cells) than individual cells (18 kPa) [56]. On the other hand, spreading of epithelial cells on artificial substrates appears to depend on a competition between cell-cell and cell-substrate adhesion [42]. Engagement of E-cadherin decreases lamellipodial protrusions at adjacent sites of integrin ligation in epithelial cells [57]. In this case, the difference between integrin and cadherin signaling for protrusion of the leading edge of cells would control the directionality and persistence of cell migration [57]. In confluent epithelial sheets, force magnitudes and directions are more complex than can be inferred from analysis of local forces applied by single cells. Forces are distributed throughout a moving sheet of cells and not just at the leading edge. Cells throughout the sheet exhibit both pushing and pulling forces [58]. Coordinated motions on length scales much larger than a single cell and maintained over long times are observed to depend strongly on the stiffness of the substrate, even when the overall cell morphologies are similar [59]. Cross-talk between integrins and cadherins is likely to be essential for coordinated movements of cell sheets such as endothelial and epithelial surfaces to enable the intact sheet to move across an ECM while maintaining constant intact cell-cell boundaries [60]. Stiffness cues might be particularly important for development of vasculature and other multicellular structures, as suggested by the finding that traction forces and the structure of multicellular arrays of endothelial cells depended strongly on substrate stiffness, with soft substrates leading to lower traction forces but more robust formation of branching networks of cells that could develop into discrete tubes rather than flat sheets [61, 62]. The stiffness of the substrate and presumably the ECM also affects the crosstalk between cell-cell and cell-ECM signaling. When embryonic fibroblasts were cultured on collagen coated polyacrylamide gels of two stiffnesses 2.7 kPa (soft) and 7.7 kPa (stiff), cells grew on soft gels initially as isolated cells that aggregated and grew as expanding clusters, whereas on stiff gels cells remained scattered, begin to spread and eventually form branching clusters [63]. Whether cells remain clustered or scatter on substrates also depends on the type of ECM ligand and therefore the class of integrins. For example, epithelial cells adhering to fibronectin or laminin maintain intercellular adhesions while in contrast, if these cells attach to collagen, the intercellular junctions dissipate and the cells disperse [64], suggesting that dominance between cell-cell and cell-ECM signaling depends on both chemical and mechanical signaling.

2 Measuring Mechanosensing at Cell-Cell and Cell-Matrix Junctions

2.5

19

Conclusions

Mechanosensing, as determined by the cell’s reaction to forces applied at its surface and to the resistance that its surroundings exert on the forces the cell generates internally, has recently regained interest as an essential component of tissue formation and function. Cells contact each other and their extracellular matrices through numerous transmembrane protein complexes that link the cell interior and engage signaling pathways in highly distinct ways. Not all adhesion sites are equally responsive to forces, and the mechanical stresses generated at any particular site depend strongly on the mechanical properties of the system, but also on the vast array of other signals the cell receives from both chemical and mechanical stimuli at other sites. The development of new soft materials and imaging methods has now revealed that both cell-ECM and cell-cell adhesion receptors are mechanically sensitive and that one system can strongly influence the mechanical response of the other. One challenge for the future is to identify not only the proteins and signals involved in mechanosensing but also the physical principles that enable cells to measure stiffness and force with the precision required to use these inputs to control their fate.

References 1. Tenney, RM, Discher, DE (2009) Stem cells, microenvironment mechanics, and growth factor activation. Curr Opin Cell Biol 21:630–5 2. Vogel, V, Sheetz, MP (2009) Cell fate regulation by coupling mechanical cycles to biochemical signaling pathways. Curr Opin Cell Biol 21:38–46 3. Mammoto, T, Ingber, DE (2010) Mechanical control of tissue and organ development. Development 137:1407–20 4. Keese, CR, Giaever, I (1991) Substrate mechanics and cell spreading. Exp Cell Res 195:528–32 5. Weiss, P, Garber, B (1952) Shape and movement of mesenchyme cells as functions of the physical structure of the medium: contributions to a quantitative morphology. Proc Natl Acad Sci U S A 38:264–80 6. Opas, M, Dziak, E (1994) bFGF-induced transdifferentiation of RPE to neuronal progenitors is regulated by the mechanical properties of the substratum. Dev Biol 161:440–54 7. Pelham, RJ, Jr., Wang, Y (1997) Cell locomotion and focal adhesions are regulated by substrate flexibility. Proc Natl Acad Sci U S A 94:13661–5 8. Klein, EA, Yin, L, Kothapalli, D, Castagnino, P, Byfield, FJ, Xu, T, Levental, I, Hawthorne, E, Janmey, PA, Assoian, RK (2009) Cell-cycle control by physiological matrix elasticity and in vivo tissue stiffening. Curr Biol 19:1511–8 9. Engler, AJ, Sen, S, Sweeney, HL, Discher, DE (2006) Matrix elasticity directs stem cell lineage specification. Cell 126:677–89 10. Georges, P, Janmey, P (2005) Cell type-specific response to growth on soft materials. J Appl Physiol 984:1547–53 11. Ghibaudo, M, Saez, A, Trichet, L, Xayaphoummine, A, Browaeys, J, Silberzan, P, Buguin, A, Ladoux, B (2008) Traction forces and rigidity sensing regulate cell functions. Soft Matter 4:1836–43

20

J.P. Winer et al.

12. Nemir, S, West, JL (2010) Synthetic materials in the study of cell response to substrate rigidity. Ann Biomed Eng 38:2–20 13. Evans, ND, Minelli, C, Gentleman, E, LaPointe, V, Patankar, SN, Kallivretaki, M, Chen, X, Roberts, CJ, Stevens, MM (2009) Substrate stiffness affects early differentiation events in embryonic stem cells. Eur Cell Mater 18:1–13; discussion 13–4 14. Leipzig, ND, Shoichet, MS (2009) The effect of substrate stiffness on adult neural stem cell behavior. Biomaterials 30:6867–78 15. Li, Z, Dranoff, JA, Chan, EP, Uemura, M, Sevigny, J, Wells, RG (2007) Transforming growth factor-beta and substrate stiffness regulate portal fibroblast activation in culture. Hepatology 46:1246–56 16. Byfield, FJ, Wen, Q, Levental, I, Nordstrom, K, Arratia, PE, Miller, RT, Janmey, PA (2009) Absence of filamin A prevents cells from responding to stiffness gradients on gels coated with collagen but not fibronectin. Biophys J 96:5095–102 17. Solon, J, Levental, I, Sengupta, K, Georges, PC, Janmey, PA (2007) Fibroblast adaptation and stiffness matching to soft elastic substrates. Biophys J 93:4453–61 18. Engler, AJ, Griffin, MA, Sen, S, Bonnemann, CG, Sweeney, HL, Discher, DE (2004) Myotubes differentiate optimally on substrates with tissue-like stiffness: pathological implications for soft or stiff microenvironments. J Cell Biol 166:877–87 19. Bhana, B, Iyer, RK, Chen, WL, Zhao, R, Sider, KL, Likhitpanichkul, M, Simmons, CA, Radisic, M (2010) Influence of substrate stiffness on the phenotype of heart cells. Biotechnol Bioeng 105:1148–60 20. Chopra, A, Tabdanov, E, Botta, GP, Kresh, JY, Janmey, P (2009) N-cadherin mediated mechanosensitivity and its role in myocyte cytoskeleton remodeling. Circulation 120:S775 21. Engler, AJ, Carag-Krieger, C, Johnson, CP, Raab, M, Tang, HY, Speicher, DW, Sanger, JW, Sanger, JM, Discher, DE (2008) Embryonic cardiomyocytes beat best on a matrix with heartlike elasticity: scar-like rigidity inhibits beating. J Cell Sci 121:3794–802 22. Jacot, JG, McCulloch, AD, Omens, JH (2008) Substrate stiffness affects the functional maturation of neonatal rat ventricular myocytes. Biophys J 95:3479–87 23. Stroka, KM, Aranda-Espinoza, H (2009) Neutrophils display biphasic relationship between migration and substrate stiffness. Cell Motil Cytoskeleton 66:328–41 24. Isenberg, BC, Dimilla, PA, Walker, M, Kim, S, Wong, JY (2009) Vascular smooth muscle cell durotaxis depends on substrate stiffness gradient strength. Biophys J 97:1313–22 25. Lo, CM, Wang, HB, Dembo, M, Wang, YL (2000) Cell movement is guided by the rigidity of the substrate. Biophys J 79:144–52 26. Nemir, S, Hayenga, HN, West, JL (2010) PEGDA hydrogels with patterned elasticity: novel tools for the study of cell response to substrate rigidity. Biotechnol Bioeng 105:636–44 27. Georges, PC, Miller, WJ, Meaney, DF, Sawyer, ES, Janmey, PA (2006) Matrices with compliance comparable to that of brain tissue select neuronal over glial growth in mixed cortical cultures. Biophys J 90:3012–8 28. Ju, YE, Janmey, PA, McCormick, ME, Sawyer, ES, Flanagan, LA (2007) Enhanced neurite growth from mammalian neurons in three-dimensional salmon fibrin gels. Biomaterials 28:2097–108 29. Byfield, FJ, Reen, RK, Shentu, TP, Levitan, I, Gooch, KJ (2009) Endothelial actin and cell stiffness is modulated by substrate stiffness in 2D and 3D. J Biomech 42:1114–9 30. Winer, JP, Oake, S, Janmey, PA (2009) Non-linear elasticity of extracellular matrices enables contractile cells to communicate local position and orientation. PLoS One 4:e6382 31. Beningo, KA, Dembo, M, Wang, YL (2004) Responses of fibroblasts to anchorage of dorsal extracellular matrix receptors. Proc Natl Acad Sci U S A 101:18024–9 32. Beningo, KA, Wang, YL (2007) Double-hydrogel substrate as a model system for threedimensional cell culture. Methods Mol Biol 370:203–12 33. Breuls, RG, Klumpers, DD, Everts, V, Smit, TH (2009) Collagen type v modulates fibroblast behavior dependent on substrate stiffness. Biochem Biophys Res Commun 380:425–9 34. Kasza, KE, Nakamura, F, Hu, S, Kollmannsberger, P, Bonakdar, N, Fabry, B, Stossel, TP, Wang, N, Weitz, DA (2009) Filamin A is essential for active cell stiffening but not passive stiffening under external force. Biophys J 96:4326–35

2 Measuring Mechanosensing at Cell-Cell and Cell-Matrix Junctions

21

35. Lee, MH, Adams, CS, Boettiger, D, Degrado, WF, Shapiro, IM, Composto, RJ, Ducheyne, P (2007) Adhesion of MC3T3-E1 cells to RGD peptides of different flanking residues: detachment strength and correlation with long-term cellular function. J Biomed Mater Res A 81:150–60 36. Gallant, ND, Capadona, JR, Frazier, AB, Collard, DM, Garcia, AJ (2002) Micropatterned surfaces to engineer focal adhesions for analysis of cell adhesion strengthening. Langmuir 18:5579 37. Takai, E, Landesberg, R, Katz, RW, Hung, CT, Guo, XE (2006) Substrate modulation of osteoblast adhesion strength, focal adhesion kinase activation, and responsiveness to mechanical stimuli. Mol Cell Biomech 3:1–12 38. Sun, Z, Martinez-Lemus, LA, Trache, A, Trzeciakowski, JP, Davis, GE, Pohl, U, Meininger, GA (2005) Mechanical properties of the interaction between fibronectin and a5b1-integrin on vascular smooth muscle cells studied using atomic force microscopy. Am J Physiol Heart Circ Physiol 289:H2526–35 39. Taubenberger, A, Cisneros, DA, Friedrichs, J, Puech, P-H, Muller, DJ, Franz, CM (2007) Revealing early steps of {alpha}2beta1 integrin-mediated adhesion to collagen type I by using single-cell force spectroscopy. Mol Biol Cell 18:1634–44 40. Staunton, DE, Marlin, SD, Stratowa, C, Dustin, ML, Springer, TA (1988) Primary structure of ICAM-1 demonstrates interaction between members of the immunoglobulin and integrin supergene families. Cell 52:925–33 41. Duguay, D, Foty, RA, Steinberg, MS (2003) Cadherin-mediated cell adhesion and tissue segregation: qualitative and quantitative determinants. Dev Biol 253:309–23 42. Ryan, PL, Foty, RA, Kohn, J, Steinberg, MS (2001) Tissue spreading on implantable substrates is a competitive outcome of cell-cell vs. cell-substratum adhesivity. Proc Natl Acad Sci U S A 98:4323 43. Pittet, P, Lee, K, Kulik, AJ, Meister, JJ, Hinz, B (2008) Fibrogenic fibroblasts increase intercellular adhesion strength by reinforcing individual OB-cadherin bonds. J Cell Sci 121:877–86 44. Wang, N, Tytell, JD, Ingber, DE (2009) Mechanotransduction at a distance: mechanically coupling the extracellular matrix with the nucleus. Nat Rev Mol Cell Biol 10:75–82 45. Liu, Z, Tan, JL, Cohen, DM, Yang, MT, Sniadecki, NJ, Ruiz, SA, Nelson, CM, Chen, CS (2010) Mechanical tugging force regulates the size of cell-cell junctions. Proc Natl Acad Sci U S A 107:9944–9 46. Siechen, S, Yang, S, Chiba, A, Saif, T (2009) Mechanical tension contributes to clustering of neurotransmitter vesicles at presynaptic terminals. Proc Natl Acad Sci U S A 106:12611–6 47. Kim, ST, Takeuchi, K, Sun, ZY, Touma, M, Castro, CE, Fahmy, A, Lang, MJ, Wagner, G, Reinherz, EL (2009) The alphabeta T cell receptor is an anisotropic mechanosensor. J Biol Chem 284:31028–37 48. Ma, Z, Finkel, TH (2010) T cell receptor triggering by force. Trends Immunol 31:1–6 49. Ganz, A, Lambert, M, Saez, A, Silberzan, P, Buguin, A, Mege, RM, Ladoux, B (2006) Traction forces exerted through N-cadherin contacts. Biol Cell 98:721–30 50. Ladoux, B, Anon, E, Lambert, M, Rabodzey, A, Hersen, P, Buguin, A, Silberzan, P, Mege, RM (2010) Strength dependence of cadherin-mediated adhesions. Biophys J 98:534–42 51. Potard, US, Butler, JP, Wang, N (1997) Cytoskeletal mechanics in confluent epithelial cells probed through integrins and E-cadherins. Am J Physiol 272:C1654–63 52. Janmey, PA, McCulloch, CA (2007) Cell mechanics: integrating cell responses to mechanical stimuli. Annu Rev Biomed Eng 9:1–34 53. Yeung, T, Georges, PC, Flanagan, LA, Marg, B, Ortiz, M, Funaki, M, Zahir, N, Ming, W, Weaver, V, Janmey, PA (2005) Effects of substrate stiffness on cell morphology, cytoskeletal structure, and adhesion. Cell Motil Cytoskeleton 60:24–34 54. Martinez-Rico, C, Pincet, F, Thiery, JP, Dufour, S (2010) Integrins stimulate Ecadherin-mediated intercellular adhesion by regulating Src-kinase activation and actomyosin contractility. J Cell Sci 123:712–22 55. du Roure, O, Saez, A, Buguin, A, Austin, RH, Chavrier, P, Silberzan, P, Ladoux, B (2005) Force mapping in epithelial cell migration. Proc Natl Acad Sci U S A 102:2390–5

22

J.P. Winer et al.

56. Dzamba, BJ, Jakab, KR, Marsden, M, Schwartz, MA, DeSimone, DW (2009) Cadherin adhesion, tissue tension, and noncanonical Wnt signaling regulate fibronectin matrix organization. Dev Cell 16:421–32 57. Borghi, N, Lowndes, M, Maruthamuthu, V, Gardel, ML, Nelson, WJ (2010) Regulation of cell motile behavior by crosstalk between cadherin- and integrin-mediated adhesions. Proc Natl Acad Sci U S A 107:13324–9 58. Trepat, X, Wasserman, MR, Angelini, TE, Millet, E, Weitz, DA, Butler, JP, Fredberg, JJ (2009) Physical forces during collective cell migration. Nat Phys 5:426–30 59. Angelini, TE, Hannezo, E, Trepat, X, Fredberg, JJ, Weitz, DA (2010) Cell migration driven by cooperative substrate deformation patterns. Phys Rev Lett 104:168104 60. Friedl, P, Gilmour, D (2009) Collective cell migration in morphogenesis, regeneration and cancer. Nat Rev Mol Cell Biol 10:445–57 61. Califano, JP, Reinhart-King, CA (2009) The effects of substrate elasticity on endothelial cell network formation and traction force generation. Conf Proc IEEE Eng Med Biol Soc 2009:3343–5 62. Califano, JP, Reinhart-King, CA (2010) Exogenous and endogenous force regulation of endothelial cell behavior. J Biomech 43:79–86 63. Guo, WH, Frey, MT, Burnham, NA, Wang, YL (2006) Substrate rigidity regulates the formation and maintenance of tissues. Biophys J 90:2213–20 64. Sander, EE, van Delft, S, ten Klooster, JP, Reid, T, van der Kammen, RA, Michiels, F, Collard, JG (1998) Matrix-dependent Tiam1/Rac signaling in epithelial cells promotes either cell-cell adhesion or cell migration and is regulated by phosphatidylinositol 3-kinase. J Cell Biol 143:1385–98 65. Kandow, CE, Georges, PC, Janmey, PA, Beningo, KA (2007) Polyacrylamide hydrogels for cell mechanics: steps toward optimization and alternative uses. Methods Cell Biol 83:29–46

Chapter 3

A Role for Integrin-ECM Bonds as Mechanotransducers that Modulate Adult Stem Cell Fate Nathaniel Huebsch and David J. Mooney

This chapter is part of Section I: Mechanisms of Cell Adhesion and Mechanotransduction

Abstract Mesenchymal stem cells (MSC), occasionally referred to as “adult stem cells,” are a multipotent cell population derived from bone marrow. MSC are an important cell population from therapeutic and fundamental science perspectives, and thus have been studied extensively. In particular, there has been substantial focus on using biomaterials to control the fate of these cells in the context of tissue regeneration. In this chapter, we review evidence for the role of substrate mechanical properties (and elastic modulus in particular) in regulating MSC fate in 2D and 3D cultures in vitro. Importantly, MSC fate appears to be markedly sensitive to the elasticity of the micro-environment in both cases, but mechanisms proposed for cellular mechanosensitivity that were based on 2D culture – in particular, a focus on morphological change as a means for sensing and responding to substrate mechanics – appear to be insufficient to explain MSC responses to substrate mechanics in 3D culture. Instead, we present recent evidence that molecular-scale changes in the cell-material interface, even absent gross morphology changes in cells, are consistent with cell fate changes in both 2D and 3D cultures. Remarkably, the mechanical interplay between cell traction forces and the material resisting this traction has both quantitative effects on occupancy of integrin adhesion receptors, as well as qualitative effects on which integrins are used for adhesion. The possibility that this is due to catch-bonds forming between integrins and materials is discussed, along with other explanations derived from the recent literature. Finally, an overview of the implications of these results for the fields of mechanotransduction and biomaterials engineering is presented. Keywords Integrin Synthetic Extracellular Matrix Mesenchymal Stem Cell F€ orster Resonance Energy Transfer (FRET) Cell therapy

D.J. Mooney (*) School of Engineering and Applied Sciences, Wyss Institute for Biologically Inspired Engineering, Harvard University, Boston, MA 02115, USA e-mail: [email protected]

A. Wagoner Johnson and Brendan A.C. Harley (eds.), Mechanobiology of Cell-Cell and Cell-Matrix Interactions, DOI 10.1007/978-1-4419-8083-0_3, # Springer Science+Business Media, LLC 2011

23

24

3.1

N. Huebsch and D.J. Mooney

Controlling Cell Fate with Synthetic Extracellular Matrix Analogs

During embryonic development, wound healing and regenerative processes, the Extracellular matrix (ECM) provides important chemical and physical cues to guide cell fate [1]. A prime example of the important role of the ECM in cell biology is the observation that many tissue cells undergo anoikis, a programme of controlled death, if they are not adherent to a substrate [2]. In the developing embryo, cell-ECM interactions are vital to many processes, including skeletogenesis [3]. Given the importance of cell-ECM interactions, biomaterials are increasingly designed to control them. These materials are increasingly used to model the micro-environment of cells in vitro, in many cases providing a context for assessing cell responses to soluble stimuli (for example, chemotherapeutics) that is more representative than standard tissue culture plastic [4–6]. A newly emerging application for these “synthetic extracellular matrices” is stem cell therapy. Currently, cell therapies hold great promise, but significant obstacles hamper their clinical translation. For example, many transplanted cells die, and even when extensive measures are taken to modify cells prior to transplantation, controlling their behavior within the patient remains a challenge [7]. Thus, ECMmimicking materials that interface effectively with cells may be useful for maintaining transplanted cell viability, and also to control cell fate in situ [8, 9]. Adult stem cells from bone marrow, often referred to as “mesenchymal stem cells” (MSC; Pittenger 1999), are a heterogeneous population of mesodermal progenitors which share the common characteristic of adhering to, and propagating on tissue culture plastic in vitro. As MSC can be isolated and propagated with relative ease, they are prime candidates for cell therapies, and have been applied clinically to treat injury in a variety of tissues, including bone and myocardium [10–12]. Given the therapeutic potential of MSC, studies have been performed to determine methods to control their fate using both soluble cues and ECM. Soluble cues which activate the mitogen activated protein kinase (MAPK) signaling pathway, for example, stimulate osteogenesis in human MSC in vitro, whereas inhibited MAPK promotes adipogenesis [13]. Specific ECM proteins (e.g. laminin) also affect MSC fate [14]. Interestingly, a common requirement for Extracellular signal-regulated protein kinase (ERK) signaling is involved in both cases. Recent work has also revealed that MSC respond to biophysical properties of the ECM, including its stiffness [15] and micro-architecture [4]. Altogether, these studies provide fundamental knowledge about adult stem cell biology, but also potentially help us to rationally design biomaterials to control the fate of those cells in situ. To improve design of these materials systems, and to identify specific molecular cues that could be used to fabricate them, natural ECM has been subjected to structure-function analysis over various length and time scales (Fig. 3.1). Over the millimeter-scale, ECM can influence the maintenance or formation of morphogen gradients by interacting with these factors. ECM can also influence cell patterning

3 Modulating Stem Cell Fate via Integrin Mechanotransduction

25

Fig. 3.1 Chemical features of extracellular matrix that influence cells over multiple spatial scales. On the nanometer scale, the ECM contains a variety of peptide epitopes (shaded rectangles, labelled with amino-acid sequences of the epitopes) that facilitate integrin-mediated adhesion and other receptor-linked functions. These epitopes are organized in a specific pattern on the nanometer scale within each protein molecule (left) and on the micrometer scale in fibrillar and other structures (center). The ECM may also regulate the diffusion of soluble proteins, mediating gradients of morphogens between cells on larger length scales (millimetres) (right); the shaded scale represents one such gradient, with the concentration (from high to low) of morphogen (e.g., vascular endothelial growth factor [VEGF]) proportional to intensity. Image adapted from Huebsch and Mooney [71]

on micron and larger scales by assembling into complex fibrillar structures. On the nanometer scale, ECM proteins interact directly with cellular transmembrane receptors. These bimolecular interactions, and the influence of ECM mechanical properties upon them, are the focus here. In this chapter, we first present an overview of previous work demonstrating that manipulating the chemical properties of synthetic ECM mimics can modulate fate in many cell types, including adult stem cells. We then focus on more recent work performed in 2D culture demonstrating that the mechanical rigidity of ECM has significant effects on cell fate. Following, we discuss the current understanding of how matrix mechanics are transduced by cells. Next, emerging data demonstrating that matrix mechanical properties also have significant effects on cell fate in 3D culture are presented. Contradictions between this 3D data and previously proposed mechanotransduction mechanisms proposed based on 2D cell culture are discussed. We then present evidence from the recent literature suggesting that the number and type of bonds formed between the ECM and integrin receptors depends on ECM mechanical properties, which provides a means for cells to respond to the rigidity of their microenvironment. Finally, implications of these findings are briefly reviewed.

26

3.2

N. Huebsch and D.J. Mooney

Controlling Cell Fate by Mimicking Extracellular Matrix Chemistry

Cells interact with natural and synthetic materials through a variety of mechanisms [16]. Perhaps the best-characterized are bimolecular interactions involving ECM proteins and the integrin family of transmembrane receptors [1, 17]. Integrins bind to adhesion epitopes including the Arg-Gly-Asp (RGD) sequence found in many natural ECM molecules (e.g. Fibronectin, FN; [18, 19]). A common strategy in designing synthetic ECM materials is to graft peptides like RGD onto synthetic or naturally derived polymers which normally do not interact with cells [20–24]. Early work with these materials demonstrated that quantitative changes in adhesion-dependent cell behaviors such as migration and morphology could be elicited by varying the density of peptide presented to cells [25, 26]. Presumably, adhesion ligand density affected cells by regulating the degree of integrin occupancy and subsequent signaling through these receptors. Studies with peptide-modified hydrogels also demonstrated that integrin-mediated adhesion to ECM regulated gene expression patterns of a variety of different cell types, including osteoblasts, in vitro [3, 27, 28]. For example, Gronthos et al. demonstrating that interactions between integrins and ECM proteins were essential in driving osteoprogenitors to the bone lineage [28]. In parallel, Alsberg et al. demonstrated that presenting RGD to a combined population of osteoblasts and chondrocytes substantially enhanced ectopic bone formation in vivo [29]. Other work confirmed an influence of adhesion ligand presentation by biomaterials on osteogenesis in vitro [24, 30, 31]. Hence, substrate-mediated changes to behaviors that are obviously adhesion dependent (e.g. migration) is often correlated to changes in other cell behaviors (e.g. osteogenesis).

3.3

A Role for Mechanics in Regulating Adhesion and Fate in 2D Culture

Historically, advances in molecular biology coupled with reductionist paradigms had the effect of de-emphasizing a role for mechanics in development [32, 33]. However, cell-matrix adhesion is one area of cell biology where there has been continual appreciation for the importance of mechanical forces. Mechanics provided both experimental tools – for example, centrifuge-based cell detachment assays [34] – as well as an intellectual foundation for considering how the various molecular components of cells might be integrated together in a physically reasonable manner. In one early study, Folkman and Moscona [35] showed a direct link between proliferation and spreading of cells on 2D substrates. Following this, Ingber et al. demonstrated that integrin-ligation by fibronectin (FN) activated both secondmessenger based soluble signaling similar to that which occurred via soluble growth factor stimulus, and physical interactions between integrins and cytoskeleton [36] .

3 Modulating Stem Cell Fate via Integrin Mechanotransduction

27

A distinct requirement for cell spreading to elicit ECM-dependent cell proliferation, regardless of the absolute density of FN presented to integrin receptors has also been demonstrated [37]. These results suggest that cell proliferation requires mechanical forces be transmitted from the ECM to the cytoskeleton through integrins, rather than integrins simply being bound to ECM. The mechanical continuity between integrins and the cytoskeleton has been tested using various engineering tools, including bead-twisting cytometry [38].

3.3.1

Using Biomaterials to Study Cell-Matrix Mechanics and Their Effects on Cell Fate

The studies described above and similar work led to a critical study performed by Pelham and Wang. In this work, hydrogel-forming polymers were combined with a surface-coating of ECM molecules (e.g. Collagen I; [39]). This provided a substrate in which mechanical rigidity could be varied independent of the density of adhesion ligand (Fig. 3.2a). Strikingly, the rigidity of these substrates – directly related to their ability to resist cell traction forces – regulated morphology, along with the structure of focal adhesions, and overall levels of tyrosine kinase phosphorylation throughout epithelial cells and fibroblasts [39]. Following this study, other investigators studied the effects of matrix stiffness in adhesion and migration of a variety of different cell types. For example, Peyton and Putnam demonstrated that cell migration speed followed a biphasic trend with respect to the rigidity of substrates (Fig. 3.2b) [40]. This is consistent with a model for migration proposed by DiMilla and Lauffenburger. In the DiMilla-Lauffenburger model, traction forces exerted through integrins can pull the cell body forward, while ECM-integrin adhesions at the trailing edge resist cell movement. At low densities of adhesion molecules, migration speed is low due to the poor adhesivity of the surface. At high densities of adhesion molecules, the cells cannot effectively detach from the surface, slowing migration. Migration speed is optimal at intermediate ligand densities, as cells can form sufficient numbers of adhesions to effectively pull against the material, but the density of adhesions is not so high as to impede the rear detachment required for movement [41, 42]. Assuming that substrate rigidity regulates the force required to form and detach adhesions, one would expect a shift in the optimum value of substrate rigidity that would maximize cell migration speed, which is consistent with the Peyton and Putnam’s results (Fig. 3.2b). A critical aspect of these types of studies is the specific method used to control the mechanical properties of the substrate, and how mechanical properties are characterized. Typically, the terms “rigidity,” “elasticity” and “stiffness” refer to the elastic modulus E of substrates. E is a quantitative metric of a material’s ability to resist elastic deformation. Strictly speaking, elastic deformation is a deformation during which all the energy imparted by mechanical deformation is stored within the material, and that removal of the load will lead to a complete return to the zero-strain initial state. Practically, E is measured by applying a small load

Fig. 3.2 Matrix elasticity and cellular mechanical forces regulate adult stem cell fate in 2D culture. (a) Schematic depicting a cell adherent to a hydrogel-based material crosslinked at specific levels to control elastic modulus E and modified to present specific quantities of adhesive moieties (e.g., RGD peptides). Features in the schematic are not meant to be drawn to scale. (b) Analysis of cell speed vs. substrate elastic modulus E for smooth muscle cells migrating on bisacrylamide hydrogel substrates presenting a high (dashed curve) or low (solid curve) density of fibronectin (image reproduced, with permission, from Peyton 2005). (c) Immunofluorescence micrographs depicting changes in the lineage markers b3-tubulin (neurogenesis), MyoD (myogenesis) and CBFa1 (osteogenesis), expression in human mesenchymal stem cells (hMSC) cultured on Collagen I coated bisacrylamide hydrogels with varying elasticity in the presence of standard growth media (image reproduced, with permission, from Engler et al. [15]). (d) Histochemical analysis of alkaline phosphatase activity (bone marker, blue) and neutral lipid deposition (fat marker, red) in hMSC plated onto poly(dimethylsiloxane) rubber substrates coated with fibronectin in which matrix elasticity was constant but cell morphology was controlled by defining where fibronectin was presented. Cells were exposed either to standard growth media (top) or media supplemented with several exogenous osteogenic and adipogenic soluble cues (Table 3.1; image reproduced, with permission, from McBeath et al. [47]). Scale bars: (c): 5 mm, (d): 50 mm. Refer to online version for necessary color discrimination for parts (c) and (d)

3 Modulating Stem Cell Fate via Integrin Mechanotransduction

29

(typically, deforming the material <10%) at a slow rate. Techniques used to measure E include millimeter-scale using compression or tensile testing, rheology. Micronscale resolution can be obtained with atomic force microscopy (AFM). For reference, tissue culture plastic has E in the range of 106 Pa, whereas non-mineralized mesodermal tissues range from ~2 103 Pa (fat) to 4 104 Pa (osteoid) [15, 43]. To control E, one can vary the density of crosslinks, N, within hydrogel substrates. Typically, E is proportional to N, and this is consistent with rubber elasticity theory [44]. When substrates are used for 2D studies, one can wash them extensively to extract unreacted crosslinker, allowing cytotoxic crosslinking molecules like bisacrylamide and cytotoxic reactions (e.g. carbodiimide) to be used for matrix formation. Furthermore, diffusional limitations posed by increasing N do not typically play a role in determining cell fate when all cells lie atop a substrate. Hence, one can decouple ECM mechanical properties from many other matrix aspects related to crosslink density that are likely to affect cell fate in 2D culture.

3.3.2

Matrix Mechanics Regulate Stem Cell Fate in 2D Culture

As with earlier work with substrates in which integrin-ECM binding was modulated by varying ligand density, studies of mechanically-controlled cell-matrix adhesion and spreading were paralleled by studies demonstrating an effect of ECM rigidity on cell fate. Kong et al. showed that differentiation of pre-osteoblasts occurred optimally on RGD-modified alginate hydrogel substrates with a relatively low elastic modulus (20 kPa). Conversely, stiffer materials (60–110 kPa) inhibited differentiation, but enhanced proliferation and non-viral gene delivery [45]. Using acrylamide gels, Engler et al. showed that substrate elasticity could also regulate myoblast differentiation [46], and, soon afterwards, also demonstrated that the commitment of human MSC populations could be modulated amongst three different lineages simply by tuning substrate elasticity (Fig. 3.2c; [15]) . Although the mechanisms for this differentiation are not entirely characterized, Engler et al. demonstrated a distinct requirement for acto-myosin mediated contractility for cells to transduce the elasticity of their matrix into fate decisions. If the acto-myosin ATPase, which mediates this contractility, was pharmacologically inhibited, they observed no elasticity-dependent cell differentiation. Hence, the ability of the ECM to resist traction forces that lead to increased cell spreading is correlated, in 2D, with osteogenic differentiation. Interestingly, in the above work, osteogenesis was most prominent on substrates with E values near 20–40 kPa, the same range found for non-mineralized osteoid tissues. This suggests that ECM mechanics may play a role in maintaining the specification of mesenchyme-derived cells, even in the presence of “noise” in terms of the soluble inputs (e.g. growth factors) associated with cell-fate decisions. A role for these soluble factors was identified in this work as well – even though human MSC commitment did not appear to require exogenous soluble factors per se, the quantitative level of commitment substantially increased in their presence

30

N. Huebsch and D.J. Mooney

Table 3.1 Examples of soluble media additives that influence MSC fate toward the osteogenic and adipogenic lineages Desired cell fate Chemical cocktail Role of each component Osteogenesis b-Glycerol phosphate Source of inorganic phosphate Ascorbic acid Cofactor required for hydroxylation of lys, pro residues in collagen, required for folding and secretion (matrix synthesis and mineralization) Dexamethasone Enhances osteogenesis within human cells but inhibits osteogenesis in rodent-derived cells [94] Serum Transforming growth factor, bone morphogenic protein and ECM molecules are all likely to be present in standard cell culture serum (FBS) and have well-characterized roles in osteogenesis Adipogenesis Dexamethasone Glucocorticoids are associated with suppressed osteoblast function in murine osteoprogenitors and their clinical use is correlated with osteopenia [95] Insulin General anabolic stimulus Indomethacin Inhibits Cox-mediated prostaglandin synthesis 3-IsobutylBlocks Adenosine receptors, activates cAMP 1-methylxanthine signaling involved with PPARg activation (IBMX)

[15]. In earlier work, functional markers of the bone and fat phenotypes were not observed to change unless authors controlled both intracellular tension and soluble factors (Fig. 3.2d; [47]) . Hence, a key aspect of mechanotransduction research, particularly in the stem cell arena, is a requirement for combinatorial presentation of both soluble and insoluble cues [43]. We provide a short overview of some relevant chemical factors below (Table 3.1). Importantly, the specific role played by several of these factors is not well characterized. It should also be noted that the majority of mechanotransduction studies are performed with serum (e.g. FBS) as a media component, and this is likely to add growth factors (e.g. bone morphogenic proteins) with roles in guiding stem cell fate. Further knowledge regarding the specific pathways on which these factors act may provide insight into how mechanotransduction pathways are integrated with signals downstream of soluble growth factors. It may be particularly revealing if changes in matrix elasticity elicit quantitative or qualitative changes in the amount or type of growth factor required to induce changes in stem cell fate. Experiments using defined media that provide greater control over chemical stimuli are likely to be an important facet of these analyses.

3.4

Examining the Role of Mechanics in 3D Culture

Emerging data from 3D cell culture studies involving MSC and other stem cell types, including embryonic stem cells, has revealed substantial differences in cell fate induced simply by altering matrix dimensionality. This may be due to a number

3 Modulating Stem Cell Fate via Integrin Mechanotransduction

31

of factors, ranging from changes in the ability of mitogens in the growth media or secreted by cells to diffuse within 3D matrices [48] to differences in the repertoire of adhesion receptors cells use to “see” the ECM [8, 49]. Remarkably, multiple investigators have shown indirectly that a5-integrins play a role in cell fate in 3D culture, even if the materials being used present ligands such as RGD but not more molecularly complex ligands (e.g. RGD combined with the synergy sequence PHSRN) purported to be required for binding this receptor in 2D culture [50]. In some cases, for example, a role for this integrin has been demonstrated using function blocking antibodies to eliminate a5-integrin mediated behaviors such as ostengenesis or self-renewal of embryonic stem cells [8, 51]. Thus, although previous 2D studies provided strong evidence that cell fate can be determined by the elastic modulus of the matrix, it is crucial to test whether or not similar effects can be observed in 3D cultures. Besides understanding which receptors are involved in 3D cell-ECM adhesion, basic questions involving how cells sense matrix elasticity need to be answered if material properties such as elasticity are to be exploited to manipulate stem cell fate. Although the subject of intensive investigations [43], the biophysical mechanisms that allow cells to sense matrix compliance remain unclear. The need to test for effects of E on stem cell fate in 3D is particularly important because in many applications involving transplantation, stem cells will encounter the material in a 3D context. Likewise, if synthetic ECM approaches are used to study biological phenomena in vitro, 3D cultures will typically be used to provide a more accurate model of the micro-environment found in vivo.

3.4.1

Biomaterials Approaches to Mechanosensing in 3D Culture

To directly address the effects of matrix elasticity on stem cell fate in 3D culture, studies have been performed using various biomaterials systems. Scaffolds with macro-scale pores (e.g. poly(lactide-co-glycolide), PLGA) systems, offer a means to culture cells in an environment that is 3D on the spatial scale of millimeters, facilitating studies of the effects of morphogen gradients on cells [52]. However, on the scale of cells (~20 mm), these materials essentially present a 2D surface. Thus, to study how cells react to 3D micro-environments, cells are typically encapsulated into hydrogel materials with mm-sized or smaller pores. One key difference between these types of substrates and materials used for 2D culture is that the method used to crosslink hydrogels must be compatible with cells. Hence, carbodiimide or other reactions that rely on cytotoxic reagents (e.g. 1-Ethyl3-(3-dimethylaminopropyl)carbodiimide, EDC) cannot be used to form these matrices. This rules out the use of bisacrylamide gel systems often used for 2D work. Instead, a variety of natural and synthetic polymers have been applied. Native ECM polymers are a natural choice for studying the role of mechanics on fate. These materials can typically be crosslinked in the presence of cells via enzymatic reactions or changes in pH. Using fibrin-based gels, for example, Ghajar et al.

32

N. Huebsch and D.J. Mooney

demonstrated a significant effect of matrix density on endothelial cell sprouting, an in vitro assay used to model angiogenesis [53]. One challenge involving the use of natural ECM systems to study mechanical responses is that many parameters of these systems that affect cells – including ligand presentation, susceptibility to matrix metalloproteinases (MMPs) and elastic modulus are coupled together. To overcome this, chemicals that specifically inhibit MMP activity or integrin binding have been applied to cells within these systems [53, 54]. Peptide modified synthetic and natural polymers capable of forming cellencapsulating hydrogels under physiological conditions provide a convenient means to study cell fate in 3D culture without complications involved with native ECM. Cytocompatible crosslinking reactions used to form these hydrogels include physical crosslinking of long-chain polymers (e.g. agarose) due to temperature changes, divalent cation bridging of negatively charged polymers (e.g. alginates), or light-mediated free radical polymerization of acrylated polymers (e.g. poly (ethylene glycol diacrylate), PEGDA). Examples of the relationships between N and E for these materials are given in Fig. 3.3a, b.

3.4.2

Matrix Elasticity Regulates Stem Cell Fate in 3D Culture

As matrix elasticity was shown to regulate cell fate in 2D culture, an important step was to determine whether this same parameter also influenced cell fate in 3D microenvironments. Several recent studies have indeed highlighted a role for E in determining cell fate, though the underlying mechanisms are not always the same as those identified in 2D work. For example, using cell-encapsulating alginate hydrogels without RGD-modification, Banerjee et al. demonstrated that E had significant effects on the fate of neural progenitors. These authors observed optimal proliferation and differentiation within the softest materials [55]. As no adhesive peptides were used in this study, it is likely that matrix crosslinking-mediated changes in morphogen transport affected cell fate in this work. Interestingly, however, this result is similar to results obtained in 2D with ligand-modified matrices with varying elasticity [15, 56], where transport limitations resulting from changes in substrate crosslinking are unlikely to play a substantial role in determining cell responses. Using ligand-modified materials, Ying et al. demonstrated a role for matrix elasticity in determining the fate of human MSC in 3D [57]. Interestingly, the range over which these authors observed osteogenesis was orders of magnitude less rigid (<1 kPa compared to ~30 kPa) than what was previously observed in 2D cultures [15]. This may be due to the shear-thinning nature of the materials used by Pek et al., or the different metrics used by these authors to assess MSC lineage specification (mRNA expression levels) contrasted with markers used for 2D studies (protein expression levels; [15]). One important issue in studies involving MSC is the inherent heterogeneity of this cell population. Because MSCs are derived based on their ability to adhere to tissue culture plastic in vitro, this cell population contains multipotent naı¨ve stem

a 60

b 120

Elastic Modulus (kPa)

Elastic Modulus (kPa)

3 Modulating Stem Cell Fate via Integrin Mechanotransduction

50 40 30 20 10

33

100 80 60 40 20 0

0 0 10 20 30 40 50 Crosslinker Concentration (mM)

0 0.5 1 1.5 polymer concentration (g/100mL)

Fig. 3.3 Matrix elasticity alters mesenchymal stem cell fate in 3D matrix culture. (a, b) Examples of data depicting elastic modulus as a function of the crosslinker type and concentration for a variety of different RGD-modified hydrogels suitable for cell encapsulation.1 (a) RGD-modified alginate at 10 mg/mL, where the polymers have high guluronic acid (filled diamond) or high mannuronic acid (filled square) content, and are crosslinked with calcium sulfate and varying concentrations. (b) RGD-modified standard agarose at concentrations from 2.5 to 150 mg/mL crosslinked by allowing a molten agarose solution to first cool to 45 C for cell encapsulation, then to 25 C for gelation. (c) Histochemical analysis of alkaline phosphatase activity (bone marker, Fast Blue Stain, blue) and neutral lipid deposition (fat marker, Oil Red O stain, red) in clonally derived mouse mesenchymal stem cells encapsulated into RGD-modified alginate hydrogels with a constant RGD density but varying elastic moduli. Scale bars: (c): 100 mm. Error bars are SD, n ¼ 3–4. Images reproduced from Huebsch et al. [8]. Refer to online version for necessary color discrimination in part (c)

cells as well as more committed cell types such as osteoblasts. Hence, it is possible that mechanotransduction pathways may primarily act on those more committed cells, and these cells, in turn, influence stem cells through paracrine signaling. To distinguish between this possibility and the possibility that stem cells themselves respond to matrix elasticity, one can use clonally derived cell populations. In one study, both primary and clonally derived MSC were shown to differentiate in response to the elasticity of RGD-modified 3D materials. Here, MSC were presented with similar peptide densities but varying rigidity within 3D matrices comprised of RGD-modified alginate, agarose or PEG-based polymers [8]. In all cases, MSC committed optimally to an osteogenic phenotype in matrices of 1

Cell encapsulation refers to the inclusion of cells into a hydrogel during the time of crosslinking. Because this procedure typically yields a spatially homogeneous material, it is assumed that encapsulation matches the 3-dimensional physiological micro-environment of tissue-derived cells, rather than artificially polarizing the cells (e.g. by plating them onto a 2D substrate).

34

N. Huebsch and D.J. Mooney

intermediate stiffness (11–30 kPa) and committed to an adipogenic phenotype in softer materials (Fig. 3.3c). Further, while population trends with human MSC were consistent with trends observed using the clonally derived murine MSC line, the human MSC exhibited more heterogeneous lineage commitment in response to ECM mechanical properties [8]. This may be due to the greater heterogeneity in the primary human MSC population – not only in terms of initial gene expression, but also in terms of responsiveness to matrix elasticity. This suggests that one important facet in studies of stem cell responsiveness to matrix elasticity is characterization of the baseline heterogeneity and differentiation potential of tissue-culture expanded naı¨ve stem cell populations. Importantly, studies with rigorously defined stem cell populations (e.g. embryonic stem cells) would determine whether or not the phenomenon of stem cell responsiveness to matrix elasticity is universal, or limited to mesoderm-derived multi-potent cells.

3.5

Mechanisms for 2D and 3D Mechanosensing

In 2D culture studies, there has been a strong focus on the link between matrix rigidity and the morphology of adherent cells (Fig. 3.4a, b; [15, 58]). In other work, morphology alone has been correlated to human MSC lineage commitment [47]. Using micro-contact printing to manipulate the extent of cell spreading, McBeath et al. demonstrated that the morphology of MSC strongly correlated with their lineage commitment: MSC that spread extensively committed preferentially to the osteoblast lineage, whereas cells in which spreading was limited tended to display markers of adipogenesis (Fig. 3.2d). These authors suggest that by controlling cell spreading, one controls the assembly of stress fibers and acto-myosin machinery required for cells to exert traction on their ECM. In turn, this traction, ultimately applied through the cytoskeleton, is responsible for allowing spread cells to commit to an osteogenic pathway. Support for this hypothesis is provided by the demonstration that spread cells commit to fat rather than bone phenotype when tractionmediated Rho-signaling is down regulated, and that cells that are not spread can be switched to an osteogenic phenotype with expression of constitutively active Rho [47]. Altogether, these findings suggests that changes in cell morphology may be required to induce stem cell fate changes in response to matrix elasticity, and that the Rho-GTPase is implicated in this process. Matrix elasticity also clearly has significant effects on stem cell fate in 3D culture, but there are a greater number of potential underlying mechanisms in this type of culture system. First: in 3D, because cells are exposed to all the components of the crosslinking reaction used to make hydrogels, changes in cell behavior may result from the small molecules used to control matrix elasticity. Alternatively, as crosslinks are added to the matrix, the diffusive transport of instructive morphogens may be substantially altered [53]. Appropriate control studies thus must be performed to study the possible contribution from each of these factors – for example, one may directly measure morphogen transport properties within materials by imaging diffusion of labeled macromolecules within

3 Modulating Stem Cell Fate via Integrin Mechanotransduction

35

Fig. 3.4 Morphology of stem cells vs. elastic modulus of substrates in 2D and 3D Culture. (a) Gross morphology of human mesenchymal stem cells (MSC), 4–24 h after plating onto Collagen I modified bisacrylamide hydrogels with varying elastic modulus. (b) Cytoskeletal morphology (F-actin, stained with phalloidin, red with blue nuclear counterstain) of human MSC on substrates of varying elastic modulus, taken 24 h after seeding cells. (c) Gross morphology of clonally derived mouse MSC encapsulated into RGD-modified alginate hydrogels of varying elastic modulus. (d) F-actin staining (red) of mouse MSC after encapsulation into RGDmodified alginate hydrogels of varying elastic modulus. Encapsulated cells were imaged 2 h after encapsulation, and no substantial changes were observed between 2 and 24 h. Scale bars: (a) 20 mm; (b) 5 mm; (c, d) 10 mm. Images reproduced, with permission, from Engler et al. [15] and Huebsch et al. [8]

or through the material, or using measurements of macromolecular release out of materials [8, 53, 59]. It may also be instructive to perform comparative studies of cellular responsiveness to matrix elasticity using entirely a range of materials in which the relationships between E and other factors vary. For example, the calcium

36

N. Huebsch and D.J. Mooney

ions used to crosslink cell-encapsulating alginate hydrogels may have direct effects on stem cells. However, in contrast to covalently crosslinked hydrogels such as PEGDA, ionically crosslinked alginate hydrogels exhibit a relatively weak relationship between network swelling (Q) and E [60]. Since Q is strongly related to macromolecular transport within hydrogels, using different chemical strategies to form hydrogels may allow effective decoupling of Q and E, allowing one to distinguish cell fate decisions that stem from mass transport changes vs. mechanotransduction pathways. This strategy has been important to the aforementioned study in which MSC fate depended strongly on the elasticity of 3D hydrogels [8]. Strikingly, 3D studies have revealed a much less substantial effect of E on cell morphology (Fig. 3.4c, d), even though cells clearly change their phenotype in response to matrix elasticity (Fig. 3.3c; [8]). In fact, whereas in 2D culture, cell spreading is typically proportional to E, in 3D culture, cell spreading is often inversely proportional to the same quantity [61]. This is likely caused by increased steric hindrance to cell extension as the cross-link density in the matrix is raised to increase E. These results contrast with results from 2D studies, which had suggested that gross cell morphology changes are required in order for cells to transduce matrix elasticity into a phenotypic response. Instead, these results suggest that while cell morphology is often a strong indicator of the degree of cell-matrix adhesion in 2D culture, changes in morphology per se are not an absolute requirement for mechanosensing. Instead, other aspects of cell-ECM interaction may be the means cells use to sense the rigidity of their environment, even in 3D culture where spreading is confined by ECM. Other possible mechanisms may involve intracellular changes in tensile forces exerted on the matrix, assembly of molecular components of focal adhesions, or changes in the bonds between integrins and ECM themselves. The possibility of cell-ECM bonds as mechanosensors is explored in the next section.

3.6

Integrin-ECM Bonds as 3D Mechanosensors

The finding that cells are able to respond to the elasticity of their matrix due predominantly to factors other than gross morphology of the cell or diffusion limitation of the matrix suggests that possibility of molecular (e.g. nanometer) scale mechanosensing mechanisms. A substantial body of work exists that is focused on mechanically-induced deformation of extracellular adhesion molecules and intracellular signaling proteins [17, 62–64]. These molecular scale changes may be independent of overall cell morphology and are likely to partially contribute to cell fate decisions in response to matrix elasticity. However, there has not been a similar level of focus on the molecular bonds between integrins and ECM, and whether these are changed by matrix elasticity. This discrepancy is especially interesting in light of findings that increases in matrix elasticity and ligand density often have a parallel effect on cell responses in 2D [46, 65]. In addition, theoretical

3 Modulating Stem Cell Fate via Integrin Mechanotransduction

37

and experimental work has demonstrated that mechanical forces can directly modulate the stability of receptor-ligand bonds involved in adhesion ([66–68]; Fig. 3.5a–c). In fact, integrins that ligate RGD, including the a5-subunit, have been demonstrated directly to act as “catch-bonds” [69]. As described by Dembo et al., a catch-bond mechanism suggests that a particular bond will have a relatively high dissociation constant (KD) when little force is applied to it, whereas at very high applied force, the bond will rupture. However, at an intermediate value of applied force, a catch-bond is activated to its high-affinity state (Fig. 3.5a). In the case of integrins and a few other adhesion receptors, a high-affinity state has been identified, and force-dependence of activation confirmed using single-molecules techniques (Fig. 3.5c; [67, 68]).

Fig. 3.5 Receptor-adhesion ligand bonds as mechanosensors. (a) Schematic of a catch-bond. Note that tensile forces applied through the bond are proportional to substrate elasticity, and that under applied force, the bond exhibits allosteric conformational change from a low affinity state to a high affinity state. At very high applied tensile forces, the bond ruptures. (b) Theoretical predictions suggesting that the force applied to a receptor-ligand bond might regulate its dissociation (image reproduced, with permission, from Bell [66]). The arrow denotes decreased baseline affinity of the receptor ligand bond or lower receptor density (KL); the y axis denotes the logarithm of the characteristic bond lifetime t*, and the x-axis denotes the scaled force f * applied to the bond. (c) Single-molecule data demonstrating that a5b1 integrins form catch bonds with fibronectin fragments. The lifetime of bonds is plotted against the force applied to the bond using atomic force microscopy (image reproduced, with permission, from Kong et al. [67]). (d) Response surface generated from FRET-based measurements of integrin-RGD bond number in 3D RGD-modified alginate hydrogels in which RGD density and elastic modulus were varied in parallel. Image reproduced from Huebsch et al. [8]

38

3.6.1

N. Huebsch and D.J. Mooney

Molecular Fluorescence Techniques to Analyze Cell-Material Interfaces

One challenge in addressing the hypothesis that changes in integrin-RGD bond formation underlie MSC fate changes induced by altering matrix elasticity is that traditional metrics of cell-matrix adhesion – for example spreading atop 2D substrates [70] – may not be accurate metrics of the nanometer-scale interface in 3D matrices that resist cell morphology changes. Thus, new methodologies have been developed to study integrin ligation and activation, independent of cell morphology. One powerful tool that has been utilized in this respect is F€orster resonance energy transfer (FRET; [8, 71–73]). FRET is a quantum-mechanical phenomenon in which dipole-dipole coupling between nearby fluorescent molecules allows the energy imparted by incident, high energy photons to one molecule, the “Donor,” to be transferred by a non-radiative means to the other molecule, the “Acceptor.” When FRET occurs, substantially less energy imparted by incident photons yields fluorescent emission (e.g. radiative decay) of the donor fluorophore, and this decrease, or the subsequent increase in acceptor emission, can be quantified [74, 71]. Using FRET between a membrane-embedded fluorescein dye and rhodamine labeled peptides, a biphasic relationship between integrin-ECM bond formation and matrix elasticity in MSCs has been demonstrated (Fig. 3.5c). This relationship is consistent with the hypothesis that integrins form catch-bonds with ECM and assuming that traction-forces cells apply to the matrix are proportional to the elastic modulus of the matrix resisting them. This latter assumption has been confirmed independently [75, 76]. The biphasic relationship between E and bond number was also maintained if different polymers and crosslinking molecules were used to present specific RGD density while varying E, suggesting that matrix elasticity (rather than the concentration or specific crosslinking molecule used to modulate E) specifically modulates integrin binding [8]. Using a similar FRET approach, Chigaev et al. were able to assess integrinactivation [77]. At low receptor occupancy, the FRET signal is likely to be predominantly caused by changes in ligand binding rather than changes in ligand conformation [72], so these investigators performed their FRET studies with saturating concentrations of receptor-binding peptides. With this approach, Chigaev et al. demonstrated that, even when certain cell types are suspended in solution, inside-out signaling mediated by Mn2+ ions was sufficient to activate integrins to a higher affinity state. A similar approach has been applied to test the hypothesis that integrin activation is coincident with focal adhesion assembly in 2D culture. Using fluorescent lifetime imaging microscopy (FLIM) to measure changes in donor lifetime, Humphries et al. demonstrate changes in integrin activation within focal adhesion sites via FRET [78]. To complement these experimental approaches, new models of the cell-ECM interface have been developed. Adding to the classic work by Bell and classic “catch-bond” work by Dembo and colleagues [66, 69], Hammer and colleagues have developed a simple theoretical model predicting that integrin clustering, vital to signaling processes, is driven by mechanical resistance both from the glycocalyx and the ECM [78].

3 Modulating Stem Cell Fate via Integrin Mechanotransduction

3.6.2

39

Biochemical Analysis of the Cell-ECM Interface

Besides changes in net integrin-ECM bond number, it is also possible that cells utilize a different repertoire of receptors to ligate the matrix when the stiffness or dimensionality of the ECM changes. A role for physical aspects of the ECM, and the resistance of ECM to cell-mediated deformation in particular, was identified by Geiger, Yamada and colleagues using glass substrates to which FN was either adsorbed or covalently bound. Although the chemical composition of substrate is the same in these cases, cells are capable of reorganizing adsorbed proteins in an acto-myosin dependent manner, whereas they are incapable of reorganizing FN that is covalently bound to glass. This physical change led to biochemical changes in the composition of cell-matrix contacts [80]. Those early studies by Geiger et al. paved the way for groundbreaking work by Yamada et al. in which the molecular contents of 2D vs. cell-derived 3D adhesions were assessed. Strikingly, these authors discovered that 3D adhesions were distinct morphologically from their 2D counterparts, and were enriched in a5-integrin. Importantly, these 3D adhesions more closely resembled adhesion structures found in vivo than did 2D cell-ECM adhesion plaques [49]. As the chemical composition of cell-ECM contacts plays a significant role in determining adhesion-mediated signaling, a variety of techniques have been developed to analyze them. In the work by Geiger, Yamada and colleagues, standard immunofluorescence approaches were used to analyze cell-ECM contacts, while immunoprecipitation was utilized to identify specific integrins bound to FN [80]. To provide additional, quantitative information about the molecular composition of the cell-matrix interface, additional biochemical techniques have also been developed. For example, Garcia et al. have applied a cell-impermeable crosslinker to covalently link integrins and the ECM-coated substrata with which they interact [81]. After solubolizing unbound proteins with sodium dodecyl sulfate (SDS), the investigators next analyzed the residual bound integrins using Western Blot or ELISA [81, 82]. A similar approach has been applied with RGD binding in 3D culture. By biotinylating RGD or other biomimetic peptides, one is able to retrieve these molecules and integrins bound to them using a neutravidin-coated substrate [8]. These new approaches have demonstrated that the specific integrins used to bind an adhesion molecule like RGD or FN depend strongly on the biophysical context in which these adhesive molecules are presented. For example, in 2D, FN only acts as a ligand for the a5-integrin when it is adsorbed to surfaces with specific chemistries [83]. In part, this may stem from these chemistries mediating a link between substrate and adsorbed FN that is more amenable to traction mediated reorganization [84]. In contrast, the RGD epitope from FN does not appear to act as an a5-integrin ligand in 2D even when presented from a compliant substrate; strikingly, however, this same molecule does ligate a5-integrins in 3D culture [8]. Quantitative analysis of a5-integrin binding to RGD in 3D culture using an ELISAbased technique further revealed a biphasic relationship between a5-integrin ligation to RGD and matrix elasticity, consistent with aforementioned FRET data (Fig. 3.5d).

40

3.7

N. Huebsch and D.J. Mooney

Indirect Effects of Traction Forces on the Cell-ECM Interface

Given a strong correlation observed between integrin-ECM binding and matrix rigidity, one may wonder whether catch-bond formation is the only mechanism underlying this effect. An additional possibility is that cells may mechanically reorganize the material, such that RGD is “gathered” and concentrated near integrins. Such a mechanism would be consistent with substantial evidence indicating that cells are capable of mechanically reorganizing natural ECM proteins such as FN, or RGD presented from the surface of compliant 2D substrates [75]. To directly address this possibility, FRET techniques have been developed [8, 75]. Intramolecular FRET has been utilized to assess changes in distance between labeled epitopes within one molecule (e.g. FN). FRET analysis of fibronectin revealed that as strain was applied to fibrils of this protein, these individual fibrils unfolded to reveal cryptic binding domains in the protein backbone, and this was correlated to strain-stiffening of the material [85]. Alternatively, intermolecular FRET has been used to demonstrate clustering of adhesion peptides normally separated by a large enough distance to prevent the FRET signal from occurring [75]. Traction mediated reorganization of compliant, 3D RGD-modified alginates was observed. Interestingly, the peak in this nanometer-scale reorganization was found to correlate with the observed peak in integrin ligation near 20 kPa [8]. This finding is consistent with previous studies, which had shown that cells cultured on very compliant substrates could assemble the cytoskeleton-associated adhesion complexes required to exert significant traction forces [86], whereas cells on very rigid substrates were incapable of deforming the material [75]. Another possible mechanism allowing matrix elasticity to alter the number of cell-ECM contacts, absent morphology changes, is that matrix mechanics indirectly alter the stability of intracellular bonds that bridge integrins to cytoskeleton by regulating the assembly of focal contacts. Theoretical and experimental studies indeed suggest this possibility [87–90]. For example, Galbraith et al. have shown that activated b1-integrins (the partner which forms heterodimers with a5-integrins) are transported to cell-ECM contacts along polymerizing actin fibers [88]. Mechanical strains applied to 2D substrates had a direct effect on the assembly of cytoskeletal polymers and thereby influenced membrane targeting of signaling proteins, including the Rho-GTPase [89, 91]. These results suggest that any changes in cell polarity or the ability of the ECM to resist traction, which would both alter intracellular force balances, would further influence integrin transport and activation states. Finally, modeling of cell-ECM mechanical coupling suggests optimal alignment of stress fibers when the elasticity of cells and ECM are similar [92], suggesting a positive-feedback mechanism that would maximize cells’ sensitivity to adhesive signaling when the matrix exhibits a certain range of elasticity. These varying mechanisms may work independently, or more likely, in concert with one another to regulate integrin-ECM bond formation. More broadly, there are several other pathways detailed in the literature through which ECM mechanics

3 Modulating Stem Cell Fate via Integrin Mechanotransduction

41

may regulate cell fate, and changes in the cell-ECM interface are likely to act in concert with these pathways as well. For example, the unfolding of signaling proteins within the focal complex due to cell traction forces has been demonstrated by Johnson et al. using a “Cys-shotgun” technique in which cysteine residues revealed by protein unfolding are tagged with a fluorophore, which allows these proteins to be identified by proteomic or other molecular analyses [62]. Cysshotgun analysis of cardiomyocytes has revealed several proteins which unfold in a manner dependent on matrix elasticity. This may be related to the functional difference in cardiomyocyte beating frequency that occurs when these cells are plated onto 2D substrates of different rigidity [93]. As the studies above imply, there many different proteins which interact in a systematic manner to, and respond in different ways, to matrix elasticity or to physical forces directly applied to cells. The fact that integrins, ECM and focal contacts are all in physical continuity with one another presents a key challenge to researches in this field: manipulations intended to selectively perturb one of these components (e.g., viral knockdown of ROCK-kinase, a key molecule in mediating cell traction forces) are likely to perturb others. This highlights the utility of techniques which are able to probe cellular mechanotransduction without mechanically perturbing the system, whether those techniques involve FRET, or noninvasive methodology [8, 62, 75, 85] (Baneyx 2005).

3.8

Implications for Cellular Reorganization of Synthetic Materials on the Nanoscale and of Integrin-ECM Bonds as Mechanosensors

Taken together, this body of work suggests integrin-adhesion ligand bonds may act as morphology-independent sensors of both matrix elasticity and dimensionality. Fundamentally, these data indicate that the physical context in which adhesive epitopes are presented to cells may partially determine how these epitopes are interpreted chemically. Differences in integrin ligation between standard tissue culture and compliant 3D micro-environments may partially explain discrepancies often observed between standard 2D tissue culture vs. in vivo, highlighting the importance of synthetic ECM approaches for both basic investigations and drug screening. From an engineering perspective, the finding that matrix dimensionality and stiffness regulate integrin bond number and stem cell fate suggests that when designing materials to manipulate cell fate, one must consider how mechanics will affect the cell-matrix interface. Moreover, if one wishes to modulate stem cells, it suggests that efforts should be focused to control cell-ECM coupling. Manipulating cell shape may be an indirect way to accomplish this rather than an absolute requirement for cells to transduce mechanical information from the substrate. More broadly, it is a natural tendency of cells to actively probe their micro-environment

42

N. Huebsch and D.J. Mooney

for adhesion ligands by deforming themselves and their matrix on multiple spatiotemporal scales. Rather than attempting to avoid cell-mediated changes in the structure of biomaterials or the interface they make with cell receptors, the mechanical and chemical processes like those describe here, if characterized sufficiently, may allow one to predict how a specific cell type will modify the material. This cell-based remodeling might one day be harnessed to process biomaterials in situ, much in the same manner that mechanical forces are applied during fabrication of traditional materials. This may present a generalized approach to create materials that are functionally complex in terms of their ability to elicit changes in transplanted cell fate, yet simple enough from a structural standpoint to be facile to manufacture.

References 1. Hynes RO. Integrins: bidirectional, allosteric signaling machines. Cell 2002; 110(6): 673–87. 2. Lodish H, Berk A, Kaiser CA, Krieger M, Scott MP, Bretscher A, Ploegh H, Matsudaira P. Molecular cell biology, 6th ed. New York: WH Freeman; 2007. 3. Damsky CH. Extracellular matrix-integrin interactions in osteoblast function and tissue remodeling. Bone 1999; 25(1): 95–6. 4. Lutolf MP, Gilbert PM, Blau HM. Designing materials to direct stem-cell fate. Nature 2009; 462(7272): 433–41. 5. Ghajar CM, Bissell MJ. Tumor engineering: the other face of tissue engineering. Tissue Eng. Part A 2010; 16(17): 2153–6. 6. Yamada KM, Cukierman E. Modeling tissue morphogenesis and cancer in 3D. Cell 2007; 130 (4): 601–10. 7. Passier R, van Laake LW, Mummery CL. Stem-cell based therapy and lessons from the heart. Nature 2008; 453(7193): 322–9. 8. Huebsch N, Arany PR, Mao AS, Shvartsman D, Ali OA, Bencherif SA, Rivera-Feliciano J, Mooney DJ. Harnessing traction-mediated manipulation of the cell-matrix interface to control stem cell fate. Nat. Mater. 2010; 9(6): 518–26. 9. Silva EA, Kim ES, Kong HJ, Mooney DJ. Material-based deployment enhances efficacy of endothelial progenitor cells. Proc. Natl. Acad. Sci. U.S.A. 2008; 105(38): 14347–52. 10. Bruder SP, Fink DJ, Caplan AI. Mesenchymal stem cells in bone development, bone repair and skeletal regeneration therapy. J. Cell. Biochem. 1994; 56: 283–94. 11. Caplan AI. Adult mesenchymal stem cells for tissue engineering versus regenerative medicine. J. Cell Physiol. 2007; 213: 341–7. 12. Wollert KC, Drexler H. Cell therapy for the treatment of coronary heart disease: a critical appraisal. Nat. Rev. Cardiol. 2010; 7: 204–15. 13. Jaiswal RK et al. Adult human mesenchymal stem cell differentiation to the osteogenic or adipogenic lineage is regulated by mitogen-activated protein kinase. J. Biol. Chem. 2000; 275: 9645–52. 14. Klees RF et al. Laminin-5 induces osteogenic gene expression in human mesenchymal stem cells through an ERK-dependent pathway. Mol. Biol. Cell 2005; 16(2): 881–90. 15. Engler AJ, Sen S, Sweeny HL, Discher DE. Matrix elasticity directs stem cell lineage specification. Cell 2006; 126: 677–89. 16. Leckband D, Israelachvili J. Intermolecular forces in biology. Quart. Rev. Biophys. 2001; 34(2): 105–267. 17. Geiger B, Bershadsky A. Exploring the neighborhood: adhesion-coupled mechanosensors. Cell 2002; 110(2): 139–42.

3 Modulating Stem Cell Fate via Integrin Mechanotransduction

43

18. Pierschbacher MD, Ruoslahti E. Cell attachment activity of fibronectin can be duplicated by small synthetic fragments of the molecule. Nature 1984; 309(5963): 30–3. 19. Ruoslahti E. RGD and other recognition sequences for integrins. Annu. Rev. Cell Dev. Biol. 1996; 12: 697–715. 20. Hern DL, Hubbell JA. Incorporation of adhesion peptides into nonadhesive hydrogels useful for tissue resurfacing. J. Biomed. Mater. Res. 1998; 32(2): 266–76. 21. Lin HB, Zhao ZC, Garciaecheverria C, Rich DH, Cooper SL. Synthesis of a novel polyurethane copolymer containing covalently attached RGD peptide. J Biomater. Sci. Polym. Ed. 1992; 3(3): 217–27. 22. Massia SP, Hubbell JA. Covalent surface immobilization of ARG-GLY-ASP-containing and TYR-ILE-GLY-SER-ARG-containing peptides to obtain well-defined cell-adhesive substrates. Anal. Biochem. 1990; 187(2): 292–301. 23. Rowley JA, Madlambayan G, Mooney DJ. Alginate hydrogels as synthetic extracellular matrix materials. Biomaterials 1999; 20(1): 45–53. 24. Stile RA, Healy KE. Thermo-responsive peptide-modified hydrogels for tissue regeneration. Biomacromolecules 2001; 2(1): 185–94. 25. Massia SP, Hubbell JA. An RGD spacing of 440nm is sufficient for integrin alpha V beta 3-mediated fibroblast spreading and 140nm for focal contact formation and stress fiber formation. J. Cell Sci. 1991; 114(5): 1089–100. 26. Xiao Y, Truskey GA. Effect of receptor-ligand affinity on the strength of endothelial cell adhesion. Biophys. J. 1996; 71(5): 2869–84. 27. Franceschi RT. The developmental control of osteoblast-specific gene expression: role of specific transcription factors and the extracellular matrix environment. Crit. Rev. Oral Biol. Med. 1999; 10(1): 40–57. 28. Gronthos S, Simmons PJ, Graves SE, Robey PG. Integrin-mediated interactions between human bone marrow stromal precursor cells and the extracellular matrix. Bone 2001; 28(2): 174–81. 29. Alsberg E, Anderson KW, Albeiruti A, Rowley JA, Mooney DJ. Engineering growing tissues. Proc. Natl. Acad. Sci. U.S.A. 2002; 99(19): 12025–30. 30. Burdick JA, Anseth KS. Photoencapsulation of osteoblasts in injectable RGD-modified PEG hydrogels for bone tissue engineering. Biomaterials 2002; 32(22): 4315–23. 31. Yang F, Williams CG, Wang DA, Lee H, Manson PN, Elisseeff J. The effect of incorporating RGD adhesive peptide in polyethylene glycol diacrylate hydrogel on osteogenesis of bone marrow stromal cells. Biomaterials 2005; 26(30): 5991–8. 32. Gilbert SF. Developmental biology, 5th ed. Sunderland: Sinauer Associates; 1997. 33. Thompson, DW. On growth and form. Cambridge: Cambridge University Press; 1917. 34. Curtis ASG, Lackie JM (eds). Measuring cell adhesion. West Sussex: Wiley; 1991. 35. Folkman J, Moscona A. Role of cell-shape in growth-control. Nature 1978; 273(5661): 345–9. 36. Schwartz MA, Lechene C, Ingber DE. Insoluble fibronectin activates the Na/H antiporter by clustering and immobilizing integrin a5b1 independent of cell shape. Proc. Natl. Acad. Sci. U.S.A. 1991; 88: 7849–53. 37. Ingber DE. Fibronectin controls capillary endothelial cell growth by modulating cell shape. Proc. Natl. Acad. Sci. USA 1990; 87: 3579–83. 38. Wang N, Butler JP, Ingber DE. Mechanotransduction across the cell surface and through the cytoskeleton. Science 1993; 260(5111): 1124–7. 39. Pelham RJ Jr., Wang YL. Cell locomotion and focal adhesions are regulated by substrate flexibility. Proc. Natl. Acad. Sci. U.S.A. 1997; 94(25): 13661–5. 40. Peyton SR, Putnam AJ. Extracellular matrix rigidity governs smooth muscle cell motility in a biphasic fashion. J. Cell Physiol. 2005; 204(1): 198–209. 41. DiMilla PA, Barbee K, Lauffenburger DA. Mathematical model for the effects of adhesion and mechanics on cell migration speed. Biophys. J. 1991; 60: 15–37. 42. Palecek SP, Loftus JC, Ginsberg MH, Lauffenburger DA, Horwitz AF. Integrin-ligand binding properties govern cell migration speed through cell-substratum adhesiveness. Nature 1997; 385: 537–40.

44

N. Huebsch and D.J. Mooney

43. Discher DE, Mooney DJ, Zandstra PW. Growth factors, matrices and forces combine to control stem cells. Science 2009; 5935: 1673–7. 44. Anseth KS, Bowman CN, Brannon-Peppas L. Mechanical properties of hydrogels and their experimental determination. Biomaterials 1996; 17(17): 1647–57. 45. Kong HJ, Polte TR, Alsberg E, Mooney DJ. FRET measurements of cell-traction forces and nano-scale clustering of adhesion ligands varied by substrate stiffness. Proc. Natl. Acad. Sci. USA 2005; 102(12): 4300-5. 46. Engler AJ et al. Substrate compliance versus ligand density in cell on gel responses. Biophys. J. 2004; 86: 617–28. 47. McBeath R, Pirone DM, Nelson CM, Bhadriraju K, Chen CS. Cell shape, cytoskeletal tension, and RhoA regulate stem cell lineage commitment. Dev. Cell 2004; 6: 483–95. 48. Fischbach C et al. Cancer cell angiogenic capability is regulated by 3-D culture and integrin engagement. Proc. Natl. Acad. Sci. U.S.A. 2009; 106(2): 399–404. 49. Cukierman E, Pankov R, Stevens DR, Yamada KM. Taking cell-matrix adhesions to the third dimension. Science 2001; 294(5547): 1708–12. 50. Aota S, Nomizu M, Yamada KM. The short amino acid sequence Pro-His-Ser-Arg-Asn in human fibronectin enhances cell-adhesive function. J. Biol. Chem. 1994; 269(40): 24756–61. 51. Lee ST et al. Engineering integrin signaling for promoting embryonic stem cell self-renewal in precisely defined niche. Biomaterials 2010; 31(6): 1219–26. 52. Fischbach C, Chen R, Matsumoto T, Schmelzle T, Brugge JS, Polverini PJ, Mooney DJ. Engineering tumors with 3D scaffolds. Nat. Methods 2007; 4(10): 855–60. 53. Ghajar CM et al. The effect of matrix density on the regulation of 3-D capillary morphogenesis. Biophys. J. 2008; 94(5): 1930–41. 54. Zaman MH et al. Migration of tumor cells in 3D matrices is governed by matrix stiffness along with cell-matrix adhesion and proteolysis. Proc. Natl. Acad. Sci. U.S.A. 2006; 103(29): 10889–94. 55. Banerjee A et al. The influence of hydrogel modulus on the proliferation and differentiation of encapsulated neural stem cells. Biomaterials 2009; 30: 4695–9. 56. Saha K et al. Substrate modulus directs neural stem cell behavior. Biophys. J. 2008; 95(9): 4426–38. 57. Pek YS, Wan ACA, Ying JY. The effect of matrix stiffness on mesenchymal stem cell differentiation in a 3D thioxtropic gel. Biomaterials 2010; 31: 385–91. 58. Jiang G, Huang AH, Cai Y, Tanase M, Sheetz MP. Rigidity sensing at the leading edge through aVb3 integrins and RPTPa. Biophys. J. 2006; 90: 1804–9. 59. Li RH, Altreuter DH, Gentile FT. Transport characterization of hydrogel matrices for cell encapsulation. Biotechnol. Bioeng. 1995; 50: 365–73. 60. Boontheekul T, Kong HJ, Mooney DJ. Controlling alginate gel degradation utilizing partial oxidation and biomodal molecular weight distribution. Biomaterials 2005; 26: 2455–65. 61. Peyton SR, Raub CB, Keschramrus VP, Putnam AJ. The use of poly(ethylene glycol) hydrogels to investigate the impact of ECM chemistry and mechanics on smooth muscle cells. Biomaterials 2006; 27: 4881–93. 62. Johnson CP, Tang HY, Carag C, Speicher DW, Discher DE. Forced unfolding of proteins within cells. Science 2007; 317(5838): 663–6. 63. Discher DE, Janmey P, Wang YL. Tissue cells feel and respond to the stiffness of their substrate. Science 2005; 310: 1139–43. 64. Baneyx G, Baugh L, Vogel V. FN extension and unfolding within cell matrix fibrils controlled by cytoskeletal tension. Proc. Natl. Acad. Sci. USA 2002; 99(8): 5139-43. 65. Chung EH et al. Biomimetic artificial ECMs stimulate bone regeneration. J. Biomed. Mater. Res. A 2006; 79(4): 815–26. 66. Bell GI. Models for the specific adhesion of cells to cells. Science 1978; 200(4342): 618–27. 67. Kong F, Garcia AJ, Mould AP, Humphries MJ, Zhu C. Demonstration of catch bonds between an integrin and its ligand. J. Cell Biol. 2009; 185(7): 1275–84.

3 Modulating Stem Cell Fate via Integrin Mechanotransduction

45

68. Marshall BT, Long M, Piper JW, Yago T, McEver RP, Zhu C. Direct observation of catch bonds involving cell-adhesion molecules. Nature 2003; 423(6936): 190–3. 69. Dembo M, Torney DC, Saxman K, Hammer D. The reaction-limited kinetics of membrane-tosurface adhesion and detachment. Proc. R. Soc. Lond. B Biol. Sci. 1988; 234(1274): 55–83. 70. Humphries MJ. Cell-substrate adhesion assays. In: Bonifacino JS, Dasso M, Harford JB, Lippincott-Schwartz J, Yamada K (eds) Current Protocols in Cell Biology; 9.1.1–9.1.11. New York: Wiley; 1998. 71. Huebsch N, Mooney DJ. Inspiration and application in the evolution of biomaterials. Nature 2009; 462(7272): 426–32. 72. Kong HJ, Boontheekul T, Mooney DJ. Quantifying the relation between ligand-recepto bond formation and cell phenotype. Proc. Natl. Acad. Sci. U.S.A. 2006; 103(49): 18534–9. 73. Lakowicz JT. Principles of fluorescence spectroscopy, 3rd ed. New York: Springer; 2006. 74. Lakowicz JR. Principles of Fluorescence Spectroscopy, 3rd. Ed. New York: Springer; 2006. 75. Kong HJ, Polte TR, Alsberg E, Mooney DJ. FRET measurements of cell-traction forces and nano-scale clustering of adhesion ligands varied by substrate stiffness. Proc. Natl. Acad. Sci. U.S.A. 2005; 102(12): 4300–5. 76. Lo CM, Wang HB, Dembo M, Wang YL. Cell movement is guided by the rigidity of the substrate. Biophys. J. 2000; 79: 144–52. 77. Chigaev A, Buranda T, Dwyer DC, Prossnitz ER, Sklar LA. FRET detection of cellular a4-integrin conformational activation. Biophys. J. 2003; 85: 3951–62. 78. Askari JA, Tyrian CJ, Webb SED, Martin-Fernandez ML, Ballestrem C, Humphries MJ. Focal adhesions are sites of integrin extension. J. Cell Biol. 2010; 188(6): 891–903. 79. Paszek MJ, Boettiger D, Weaver VM, Hammer DA. Integrin clustering is driven by mechanical resistance from the glycocalyx and the substrate. PLOS Comp. Biol. 2009; 5(12): e10000604. 80. Pankov R et al. Integrin dynamics and matrix assembly: tensin-dependent translocation of a5b1 integrins promotes early fibronectin fibrollogenesis. J. Cell Biol. 2000; 148: 1075–90. 81. Garcia AJ, Vega MD, Boettiger D. Modulation of cell proliferation and differentiation through substrate-dependent changes in FN conformation. Mol. Biol. Cell 1999; 10: 785–98. 82. Keselowsky BG, Garcia AJ. Quantitative methods for analysis of integrin binding and focal adhesion formation on biomaterial surfaces. Biomaterials 2005; 26: 413–8. 83. Keselowsky BG, Collard DM, Garcia AJ. Integrin binding specificity regulates biomaterial surface chemistry effects on cell differentiation. Proc. Natl. Acad. Sci. U.S.A. 2005; 102(17): 5953–7. 84. Lan MA, Gersbach CA, Michael KE, Keselowsky BG, Garcia AJ. Myoblast proliferation and differentiation on fibronectin-coated self assembled monolayers presenting different surface chemistries. Biomaterials 2005; 26(22): 4523–31. 85. Klotzsch E et al. Fibronectin forms the most extensible biological fibers displaying switchable force-exposed cryptic binding sites. Proc. Natl. Acad. Sci. U.S.A. 2009; 106(43): 18267–72. 86. Galbraith CG, Yamada KM, Sheetz MP. The relationship between force and focal complex development. J. Cell Biol. 2002; 159(4): 695–705. 87. Brown CM et al. Probing the integrin-actin linkage using high-resolution protein velocity mapping. J. Cell Sci. 2006; 119: 5204–14. 88. Galbraith CG, Yamada KM, Galbraith JA. Polymerizing actin fibers position integrins primed to probe for adhesion sites. Science 2007; 315(5814): 992–5. 89. Putnam AJ, Cunningham JJ, Dennis RG, Linderman JJ, Mooney DJ. Microtubule assembly is regulated by externally applied strain in cultured smooth muscle cells. J. Cell Sci. 1998; 111: 3379–87. 90. Ward MD, Dembo M, Hammer DA. Kinetics of cell detachment: peeling of discrete receptor clusters. Biophys. J. 1994; 67(6): 2522–34. 91. Putnam AJ, Cunningham JJ, Pillemer BBL, Mooney DJ. External mechanical strain regulates membrane targeting of Rho GTPases by controlling microtubule assembly. Am. J. Physiol. Cell Physiol. 2003; 284: C627–39. 92. Zemel A, Rehfeldt F, Brown AEX, Discher DE, Safran SA. Optimal matrix rigidity for stressfibre polarization in stem cells. Nat. Phys. 2010; 6: 468–73.

46

N. Huebsch and D.J. Mooney

93. Engler AJ et al. Embryonic cardiomyocytes beat best on a matrix with heart-like elasticity: scar-like rigidity inhibits beating. J. Cell Sci. 2008; 121: 3794–802. 94. Lian JB et al. Species-specific glucocorticoid and 1,25-dihydroxyvitamin D responsiveness in mouse MC3T3-E1 osteoblasts: dexamethasone inhibits osteoblast differentiation and vitamin D down-regulates osteocalcin gene expression. Endocrinology 1997; 138(5): 2117–27. 95. Gundle R, Stewart K, Screen J, Beresford JN. Isolation and culture of human bone-derived cells. In: Beresford JN, Owen ME (eds) Practical Animal Cell Biology Series: Marrow Stromal Cell Culture. New York: Cambridge University Press; 1998.

Chapter 4

Cell-Generated Forces in Tissue Assembly, Function, and Disease John Huynh, Joseph P. Califano, and Cynthia A. Reinhart-King

This chapter is part of Section II: Cooperative Cell Behavior and Mechanobiology

Abstract While the cell and tissue-level effects of exogenous, physiological forces like shear stress and pressure are well-documented, the effects of endogenous cellgenerated forces and the mechanics of the microenvironment have only recently gained significant attention. There is now mounting evidence that cells generate contractile forces that can elicit changes in the balance between cell-cell cohesion and cell-matrix adhesion within tissues. This balance is critical in governing tissue structure, formation and health. These cell-generated traction forces are altered by changes in the mechanics of the cellular microenvironment. Notably, changes in tissue stiffness accompany both the progression of many diseases including atherosclerosis, heart disease and cancer, and in normal physiological processes including development. Recent evidence suggests that the mechanics of the microenvironment may play a role in dictating cell function and tissue structure. Additionally, abnormal changes in tissue stiffness may promote disease progression. This chapter will discuss the role of cell-mediated forces and the mechanics of the microenvironment in the assembly and maintenance of cells into tissues. Recent advances in tools, techniques, and materials used to study cellular forces and the effects of matrix mechanics will be described. Additionally, the role of cellular traction forces and matrix mechanics in both normal and diseased states will be described, using examples primarily from the cardiovascular system to illustrate the relationship between mechanics and cell and tissue function.

4.1

Introduction

A balance between cell-cell and cell-matrix adhesion exists within tissues. This balance is critical in governing tissue formation, structure, health, and disease. Mechanical forces exerted by cells and the mechanical properties of the

C.A. Reinhart-King (*) Department of Biomedical Engineering, Cornell University, Ithaca, NY 14853, USA e-mail: [email protected]

A. Wagoner Johnson and Brendan A.C. Harley (eds.), Mechanobiology of Cell-Cell and Cell-Matrix Interactions, DOI 10.1007/978-1-4419-8083-0_4, # Springer Science+Business Media, LLC 2011

47

48

J. Huynh et al.

microenvironment play integral roles in dictating this balance. This chapter will focus on the roles of cell-generated forces in tissue development, tissue homeostasis, and tissue disorganization resulting in disease progression. Particularly, the effects of cell force on cell sorting, extracellular matrix remodeling, stem cell differentiation, and embryonic development will be discussed. Finally, an overview of how dysregulated cell forces result in disease will be presented, using examples primarily from the cardiovascular system, to illustrate the relationship between cell and matrix mechanics and tissue function.

4.2 4.2.1

Forces in Cell Sorting Differential Adhesion Hypothesis

During development and tissue morphogenesis, mixed cell types segregate and organize themselves into specific structures. Cell sorting was extensively studied in the 1960s and 1970s by the laboratory of Malcolm Steinberg who first showed that mixed cell type aggregates undergo segregation into several homogenous structures [1–4]. The authors speculated that this behavior is analogous to phase separation of immiscible liquids, like oil and water that do not form a homogeneous mixture. Instead, they separate to prefer adhesion with “like” elements. Time-lapse experiments by Steinberg and Garrod showed that co-cultured chick embryonic liver and limb bud mesoblast cells actively “sorted-out” such that the liver cells were grouped together in islands surrounded by limb bud cells [5]. The authors hypothesized that the more cohesive cell type, i.e. those that prefer cell-cell adhesions, remains in the middle while the less cohesive cell type, i.e. those that instead prefer cell-matrix adhesions, distributes on the outside. From these observations, Steinberg developed a thermodynamics-based hypothesis that described cell sorting as a phenomenon that minimizes the free energy of a system based on intercellular adhesivity and surface tension. This Differential Adhesion Hypothesis (DAH) set the conceptual framework for understanding cell sorting and tissue assembly by force-driven mechanisms. By the 2000s, numerous cell sorting and adhesion studies have experimentally supported Steinberg’s hypothesis. For example, differential adhesion is suggested to play a role in delineating adjacent embryonic mouse brain regions [6] and in the organization of endocrine cells in pancreatic islets [7]. In a particularly elegant study, Ryan et al. demonstrate that tissue spreading is based on a balance between cell-cell and cell-substrate adhesions [8]. In these experiments, embryonic mouse fibroblasts, normally non-cohesive, were transformed to express varying levels of cadherin, an important cell surface protein that regulates cell-cell interactions (discussed in more depth in Sect. 4.4 of this chapter), which promoted their cohesivity. In addition to controlling cell-cell adhesion, cell-substrate adhesivity was modulated by incorporating various amounts of polyethylene glycol onto the

4 Cell-Generated Forces in Tissue Assembly, Function, and Disease

49

substrate. After plating cell aggregates, the authors found that increasing cell-cell adhesion (through increased cadherin expression) decreases aggregate spreading rate whereas decreasing cell-substrate adhesion also decreases spreading rate. Although compelling, these measurements do not directly evaluate the thermodynamics hypothesis since surface tensions of cell aggregates were not quantified. To directly measure the surface energy across a cell aggregate, a parallel plate compression apparatus was designed which continuously monitors the force exerted by spherical cell aggregates and the changes in aggregate shape as the aggregate is compressed [9]. Foty and Steinberg were able to measure surface tensions of cell aggregates of varying cohesivities and show that aggregate surface tension increases linearly with the number of surface cadherins per cell [10], results that strongly support the DAH. While the DAH is the primary hypothesis appearing in the scientific literature to describe cell sorting, there is some evidence to suggest that the DAH is limited in its ability to predict all in vitro cell sorting outcomes. Advances in molecular biology and nanotechnology have enabled measurements of receptor binding strength which contribute to cell-cell adhesion. Utilizing atomic force microscopy (AFM), adhesion strengths and dissociation rates of chicken, canine, and Xenopus cadherin bonds have been quantified [11] to show that cell sorting outcomes do not correlate with cadherin bond adhesion strengths or dissociation rates. In another study, Krieg et al. quantified the cohesivities of cells from the endoderm, mesoderm, and ectoderm of embryonic zebrafish [12]. Using AFM, they measured the force required to pull apart two cells, where one cell is attached to a fixed substrate and the other on the cantilever. The cohesivity of the mesoderm was found to be highest, followed by the endoderm and then the ectoderm. Because the ectoderm has the lowest cohesivity, the DAH would predict that the ectoderm would distribute on the outside. However, when allowing pairs of cell populations to self-sort in vitro, the ectoderm cells were always found to aggregate in the middle. These same studies also quantified cell tension or contractility by using AFM to measure the force required to deform adherent cells, a method analogous to micropipette aspiration assays [12]. They show that ectoderm cells had the highest contractility, followed by the mesoderm and finally the endoderm. Because endoderm cells surround mesoderm cells and that the ectoderm always aggregated in the middle in in vitro sorting assays, the authors suggest that sorting correlates with actomyosin-dependent tensile forces rather than through differential cadherindependent adhesions. These findings support Harris’s Differential Surface Compaction Hypothesis (DSCH) which assumes that instead of cellular adhesions, cell-generated forces in cell-cell and cell-surface interactions guide cell sorting [13]. Findings from more recent computer simulations also acknowledge that the DSCH deserves additional attention [14]. Regardless of which hypothesis (either the DAH or the DSCH) holds true, physical force, either generated through the cytoskeleton or between cellular adhesions, is considered the underlying predictive mechanism by which cell sorting and tissue assembly occurs.

50

4.2.2

J. Huynh et al.

Extracellular Matrix Composition in Cell Sorting and Aggregation

Extracellular matrix (ECM) protein composition is known to influence the endogenous generation of cell forces important in cell sorting and aggregation. Here, we describe several in vitro studies that have laid the groundwork for understanding how ECM composition influences the balance between cell-cell and cell-matrix adhesions that drives cell sorting. Cell aggregation has been studied by altering ECM ligand concentration which modulates cell-substrate adhesivity. For example, increasing concentrations of Matrigel changes hepatocyte aggregate morphology from spheroids to monolayers [15]. Mixtures of primary rat hepatocytes and bovine aortic endothelial cells (ECs) preferentially sort with changes in collagen concentration. Low collagen concentrations promote cell aggregation where ECs surround a core of hepatocytes; intermediate collagen concentrations promote cell stacking where a layer of hepatocytes spread on top of the EC layer; high collagen concentrations allow both cell types to spread across the substrate surface [16]. The sorting and aggregation of cells also depends on matrix protein type. Rat dermal papilla cells plated on poly(ethylene-co-vinyl alcohol) form aggregate spheroids when the substrate is coated with fibronectin but not collagen [17]. It was shown that fibronectin promotes aggregation by enhancing cell attachment and motility, which is not as well-supported on collagen for this particular cell type. These data indicate that cell sorting and aggregate morphology is sensitive to both ECM protein concentration and type. The role of ECM composition in cell sorting and aggregation is important in organogenesis. During the development of some organs including the vasculature and kidneys, cell aggregates assemble preferentially in sprout or branching morphologies. The same factors that have been shown to affect cell sorting and aggregation also affect branching. Early experiments indicate tube formation is induced when substrate adhesivity is increased by increasing the density of fibronectin coatings [18]. Work in our own lab has determined that assembly of ECs into branching structures depends on a balance between cell-cell and cell-substrate interactions [19]. We show that spontaneous network formation can be induced by changing the density of matrix collagen derivatized to the substrate. In the kidney, ureteric buds branch when suspended in Matrigel diluted in low concentrations of hyaluronic acid, whereas high concentrations of hyaluronic acid inhibit branching [20]. Together, these data suggest that cell sorting and aggregation, which are important in development and organogenesis, are also mediated by matrix composition and density.

4.2.3

Substrate Mechanics in Cell Sorting and Aggregation

The role of matrix mechanics as a mediator of cell force generation has gained increasing attention in recent years. In general, rigid substrates promote increased

4 Cell-Generated Forces in Tissue Assembly, Function, and Disease

51

cell force generation through cytoskeletal and focal adhesion organization [21–25]. Interestingly, the mechanical properties of the ECM also mediate cell aggregation and tissue assembly. There is significant data indicating that substrate stiffness alone promotes self-assembly of ECs into networks or capillary-like structures; see methods and review for ECs [26, 27]. In our lab, we found that bovine aortic EC networks spontaneously assemble and organize on compliant (E 1 kPa) but not stiff (2.5 kPa E 10 kPa) substrates derivatized with the same concentration of collagen [19]. Other 2D studies also show that ECs exhibit a decrease in network formation when cultured on stiffened Matrigel [28, 29] or fibrin substrates [30–32] compared to softer substrates. Likewise, this same behavior occurs in 3D, where human umbilical vein ECs were shown to decrease network formation in collagen gels stiffened with increasing collagen concentration compared to low density gels [33]. Recent studies have begun to clarify some of the intracellular responses of ECs to substrate mechanics in regards to network formation and angiogenesis. For example, a decrease in ECM stiffness leads to increased matrix metalloproteinase mRNA expression and activity [34], indicating that the stiffness of the ECM may regulate degradation and remodeling of the ECM to permit cell assembly in vitro. Furthermore, angiogenic signaling during tube formation is affected by mechanosensitive transcriptional mechanisms. ECM stiffness influences p190RhoGAP, an upstream inhibitor of Rho, and downstream transcription factors regulating vascular endothelial growth factor (VEGF)-receptor expression and angiogenesis in vitro and in vivo [35]. These data suggest that optimal VEGF-receptor expression and vascular development require a specific ECM stiffness environment. Collectively these data indicate crosstalk between cell-substrate adhesions and growth factor responsiveness – compliant substrates promote endothelial network assembly, and tube formation may be modulated by substrate mechanics through signaling pathways active at the transcriptional level. Exogenous force cues supplied to cells through substrate stiffness is sufficient in guiding cellular aggregation and assembly in several other cell types in addition to ECs. When plated on compliant 2D polyacrylamide gels, 3T3 fibroblasts aggregate whereas those on stiffer substrates scatter [36]. Similarly, cells within neonatal rat heart explants migrate out of the tissue onto stiff substrates whereas those on compliant substrates remain adherent in the explant [36]. The ability of Chinese hamster ovary cell aggregates to self-assemble into metastable constructs depends on the adhesive and mechanical properties of the 3D NeuroGel matrix [37]. Additionally, ductal morphogenesis of T47D human breast carcinoma cells is inhibited by increasing collagen gel stiffness [38]. Taken together, these data indicate that tissue morphogenesis can be modulated by substrate stiffness in a number of cellular systems. The exact mechanism by which substrate mechanics guides tissue assembly remains unclear. It is thought that cell assembly during network formation or aggregation may result from an optimization of mechanical input [36]. On compliant substrates where mechanical resistivity is low, cells seek out cell-cell connections and form aggregates to increase mechanical input. In contrast, stiff mechanically-resistive substrates may provide optimal mechanical input creating

52

J. Huynh et al.

a preference for cell-substrate interactions and scattering [39]. Cellular aggregation on compliant substrates may result from weakened cell-substrate adhesions and contractile forces that drive cell-cell interactions [36]. Furthermore, modeling of EC network formation implicates cell-generated traction forces (discussed in more detail in Sect. 4.4) as mediators of matrix patterning and network formation [40]. In summary, cell force generation and tissue assembly are sensitive to the local mechanical environment, and clearly further consideration of the role of substrate mechanics in dictating tissue form and function is required to explore the full parameter space of mechanics across physiological systems.

4.3

Forces in Differentiation and Development

As described in previous sections, forces play a fundamental role in regulating cell processes in embryonic development and stem cell differentiation. Particularly, as morphogenic events proceed, tensile forces are necessary for the correct shaping of embryos [41]. Interestingly, modeling of Drosophila wing development indicates that non-uniform tissue growth leads to the accumulation of mechanical stress among cells growing near neighboring cells [42]. When compact tissue patches grow faster than their surroundings, the patch becomes compressed and strains the surrounding tissue. Such stresses may in turn feedback to regulate cell proliferation and apoptosis and ensure uniform tissue growth. As another example, when Xenopus laevis undergo gastrulation, the notochord, a flexible rod-shaped structure that defines the “backbone” of the embryo, stiffens in order to provide mechanical stabilization for the body [43]. Without sufficient stiffening, the notochord will be unable to properly elongate the embryo as it is bent or compressed. These data implicate force generation as a key regulator of proliferation and suggest that tissue growth is linked to force-dependent cellular contractility through intricate feedback mechanisms. Matrix rigidity also plays a critical role in development and is an emerging field in the study of differentiation and tissue formation. In a seminal study by Engler et al., substrate stiffness alone has been found to direct differentiation of mesenchymal stem cells (MSCs) [44]. When plated on substrates that have a rigidity similar to that of brain tissue, MSCs are directed into a neurogenic lineage, whereas on stiffer substrates that mimic the rigidities of muscle or collagenous bone, MSCs differentiate into myogenic or osteogenic lineages. Similarly, substrate elasticity has been found to modulate cell spreading, growth rate, gene expression, and osteogenic differentiation of pluripotent embryonic stem cells [45]. These studies have laid a foundation for what will surely be a fertile area of research for many years: that is, the role of mechanics in stem cell differentiation. In addition to changes in ECM stiffness, the stiffness of cells themselves also contributes to differentiation and embryogenesis. Notably, the nuclei of human embryonic stem cells stiffens sixfold as cells progress through differentiation [46], suggesting that cell stiffening plays a role in promoting differentiation.

4 Cell-Generated Forces in Tissue Assembly, Function, and Disease

53

Chowdhury et al. corroborated this finding by showing that mouse embryonic stem cells are ten times softer than their differentiated counterparts [47]. It has been hypothesized that by remaining soft, undifferentiated cells have the advantage of being more motile and retain the ability to migrate through solid tissues and small pores. As will be discussed in more detail later in the chapter, cell stiffness is directly influenced by the actomyosin-dependent contraction of cells [12]. Sordella et al. show that modulating the activity of Rho, an integral mediator of cell force generation, is sufficient to alter the differentiation program of adipocyte and myocyte precursors [48]. The authors also show that mouse embryo-derived fibroblasts lacking p190RhoGAP, an inhibitor of Rho, have excessive Rho activity and are defective for adipogenesis, opting instead to undergo myogenesis. Subsequent studies suggest that actomyosin-dependent tension, mediated by Rho and its downstream effector Rho-associated kinase (ROCK), is required for human MSCs to commit to osteogenesis [49]. McBeath et al. attribute lineage commitment to mechanical cues presented by changes in cell shape – they found that unspread, round MSCs underwent adipogenesis whereas flat, spread cells committed to an osteogenic lineage. In vivo studies also implicate the Rho-ROCK pathway as a central facilitator of differential growth in tissue assembly. In embryonic mouse lung morphogenesis, disrupting cell tension by inhibiting ROCK and other downstream effectors alters basement membrane thickness, inhibits new epithelial bud formation, and disrupts capillary blood vessels [50]. Conversely, increasing cell tension by activating Rho accelerates lung branching and increases capillary elongation. Taken together, these studies highlight the critical role cell-generated forces play in the correct differentiation of stem cells and embryonic development.

4.4

Forces in Cell-Substrate and Cell-Cell Interactions

As described above, cell sorting, differentiation, and tissue assembly are regulated by forces generated by cell-cell and cell-substrate adhesion [10, 12]. Cell-cell adhesion is mediated in part by cadherins, transmembrane, calcium-dependent, intercellular adhesion proteins [51], whereas cell-adhesion is largely attributed to integrins, heterodimeric, transmembrane glycoproteins that bind the cell to the ECM (Fig. 4.1) [52]. The integrin adhesome is comprised of a complex network of molecular players that is undergoing intense research [53, 54]. Significant crosstalk exists between cadherins, integrins, and growth factor receptors [55], and the action of cadherins and integrins as cell adhesion receptors are considered largely responsible for mechanotransduction during development and tissue morphogenesis [56]. Cadherins and integrins are physically linked intracellularly to actin, an important component of the cytoskeleton that mediates cellular contractility. Actin assembles into stress fibers that behave as tensed viscoelastic cables that contribute significantly to cell shape [57]. Integrins transmit force between the cytoskeleton and ECM. As a physical and chemical link between the intra- and extracellular domains, integrins

54

J. Huynh et al.

Fig. 4.1 Players mediating cell-cell and cell-substrate interactions. Cell-cell interactions are primarily mediated by cadherins, transmembrane proteins that dimerize in the intercellular space. Integrins also span the plasma membrane (PM) and bind intracellular actin to the extracellular matrix (ECM). Growth factor binding also activates intracellular mediators of contractility including Rho pathways. Note that cadherins and integrins are tied directly to actin stress fibers (SF) via a number of adapter proteins (AP)

play a key role in cellular mechanosensing [58–60] and traction force generation, critically regulating a variety of cell behaviors [61–64]. Endogenous traction forces are used for motility, tissue assembly, ECM remodeling, and mechanosensing. In recent years, many different methods have been developed to quantify cellular traction forces. Here, we briefly review these in vitro methods and highlight the experiments that have characterized the nature of traction forces.

4.4.1

Traction Force Quantification

Generally, cell-generated traction forces are exerted tangentially on a substrate and occur through specific structures, like focal adhesions, or through nonspecific frictional interactions [65]. Traction forces of adherent cells were first observed in early experiments as cell-generated wrinkles in silicone rubber substrates [66].

4 Cell-Generated Forces in Tissue Assembly, Function, and Disease

55

Fig. 4.2 (a) Landmark experiments with chick heart fibroblasts on deformable silicone substrates reveal the presence of cell-generated traction forces; bar is 100 mm. From Harris et al. [66]. Reprinted with permission from AAAS. (b, c) Micropost arrays: (b) Electron micrograph of a smooth muscle cell attached to a microfabricated post array. (c) Traction force vector map (white arrows) obtained by measuring post deflections spatially correlated with immunofluorescence localization of the focal adhesion protein vinculin; bars are 10 mm, arrow is 50 nN. From Tan et al. [74]. Copyright 2003 National Academy of Sciences, USA. (d–h) Traction force microscopy: (d) A cell is plated on a deformable gel substrate and imaged with phase contrast microscopy. Fluorescent beads embedded in the gel substrate are imaged immediately after the phase image (e, “Stressed”) and after cell removal with trypsin (f, “Null”). Arrowheads and arrows indicate bead locations before and after cell removal, respectively. (g) Merged fluorescent images depict the displacements of the marker beads in the substrate due to cell contractile forces in the boxed region of the phase image. (h) Marker displacements are used to calculate the magnitude (|T| per area) and direction (arrows) of tractions that a cell exerts on the substrate. Scale bar in (d–g) is 50 mm; beads in (d–g) are 500 nm in diameter. Reprinted from Journal of Biomechanics [26], with permission from Elsevier

These landmark experiments were the first to demonstrate the presence of contractile forces in cells (Fig. 4.2a). Interestingly, there was no evidence of pushing forces based on the location of the cell relative to the wrinkles. The most significant drawback to the wrinkling substratum method, despite its ease to implement, is that it is only semi-quantitative. It is difficult (and perhaps impossible) to derive

56

J. Huynh et al.

quantitative information about the magnitude of the forces based on the location or size of the wrinkles. This limitation was overcome by the incorporation of fluorescent beads into pre-stressed silicone substrates, which provided a more quantitative analysis of cellular traction forces [67] and set the stage for methods of measuring force through substrate displacement analysis. Regarded as an inverse problem where cell-generated stress is calculated from the measurement of underlying substrate displacement, one particular method for measuring traction forces from substrate displacements has gained prominence in recent years: traction force microscopy. Traction force microscopy (TFM) is a method developed by Dembo et al. to measure cell-generated traction stresses [68]. TFM is a tool that maps the magnitude, direction, and spatial orientation of traction forces exerted by an adherent cell on a deformable substrate. Beads embedded in the deformable substrate allow displacements to be tracked with an optical flow algorithm [69]. Forces are calculated using Bayesian statistics to find the most likely traction vector field that explains a given cell-generated strain field (Fig. 4.2d–h) [70]. TFM algorithms by Dembo et al. have been packaged in a licensed software library called LIBTRC. Similar techniques that also solve the inverse problem have been published by others in recent years [71–73]. A second prominent method for measuring traction, first described by Chen and coworkers [74], involves cell culture on microfabricated post array detectors (mPADs, Fig. 4.2b, c); see [75] for methods. Cell force and contractility causes the posts to deflect. These deflections are directly proportional to the traction forces exerted by the cell and are calculated at each post using elastic beam theory [74]. The modulus of mPADs can be altered by the geometry of the posts independent of the surface ECM coating that facilitates cell adhesion. The advantage of such a technique is that computationally, cellular forces are straightforward to calculate, as the calculation is based on simple beam-bending mechanics. Moreover, the force on each post is decoupled. A drawback of these measurements is that the cells are adherent to a noncontinuous surface, which deviates further from mimicking the in vivo environment. Alternate methods of force quantification have also been described, including the use of microfabricated cantilevers [76] and micropatterned silicone elastomeric substrates [77], but their appearance in the literature is limited.

4.4.2

Forces in Cell-Matrix Interactions

The results of experimental methods used to characterize traction forces have elucidated the complexity of cell-generated forces in vitro. Traction force generation is sensitive to cell size, shape, ECM density, substrate stiffness, and growth factors present in the environment. Early work showed that traction generation is dependent on both cell size and shape [78, 79]. Cell tractions on circular islands are greatest at sites of cell protrusion but exhibit no preferred direction of force. However, on square islands, high traction levels are associated with protrusions at island corners. Additional traction measurements of smooth muscle cells, human umbilical vein endothelial cells (ECs), bovine aortic ECs, mouse embryo

4 Cell-Generated Forces in Tissue Assembly, Function, and Disease

57

fibroblasts, and mammary epithelial cells indicate that while the magnitude of traction forces may vary significantly, there is a positive correlation between cellgenerated force and cell size [80, 81]. These data suggest that traction forces are sensitive to cell size and shape in a variety of cell types and underscore the importance of studying tractions of specific cell types. Traction forces are also sensitive to both substrate ECM coating and stiffness. Work in our own lab with polyacrylamide gels derivatized with increasing concentrations of ligand indicate that cell spreading and traction forces increase with ligand concentration and that force increases linearly with cell area [81, 82]. This coincides with the work described above indicating that traction forces are transmitted through cell-matrix contacts, where additional ligand on a surface increases the availability of cell-matrix adhesions to form. Cellular force also increases with substrate stiffness and concomitant increases in cell area [83, 84]. Since increasing substrate stiffness increases contractility and cell area, interplay between substrate stiffness, cell area, and traction force generation may exist. Using measurements of force, area, and stiffness coupled with linear regression modeling, our lab demonstrated that both substrate stiffness and cell area are significant predictors of traction force generation of single cells [85]. Moreover, we demonstrate that this trend holds for pairs of cells in contact. These data suggest that both substrate stiffness and cell area, whether modulated by stiffness directly or by ligand density, contribute significantly to traction force generation. Growth factors present in the extracellular environment also modulate cell contractility [86]. Notably, it has been suggested that cell sensitivity to growth factor, cytokines, or the type of ECM ligand present are governed by a mechanically resistive ECM that facilitates the production of cytoskeletal distortions [87]. Vascular endothelial growth factor (VEGF), a secreted glycoprotein that stimulates endothelial cell proliferation, migration [88], and angiogenesis [89] has been shown to stimulate stress fiber and focal adhesion formation [90] through Rho and ROCK activation [91]. Similarly, tissue growth factor-a (TGFa), a ligand for epidermal growth factor receptors, stimulates Rho and ROCK activity that lead to cytoskeletal reorganization [92]. While growth factors are typically only considered in the context of their effects on cell growth on proliferation, these data indicate that they may function in part through their ability to alter cell contractility profiles. It is likely that there is significant crosstalk between cell response to growth factors, substrate stiffness and cell-matrix adhesions, suggesting that these separate signals integrate intracellularly to produce a specific cell response. There is intense interest in understanding the specific mechanism by which cellmatrix adhesions transmit force from the cell to the substrate. Real-time measurements of cell force coupled with fluorescent imaging of focal adhesions indicate that focal adhesion area is linearly dependent on the local force exerted by a cell [77]. This relationship indicates that a constant stress is applied to focal adhesions in the cell despite variations in focal adhesion size and shape. Separate experiments with cells on mPADs confirm that focal adhesion size correlates with cell-generated stresses [74]. Notably, this correlation holds only for focal adhesions larger than 1 mm2 while small adhesions generate large forces that do not correlate with

58

J. Huynh et al.

adhesion size. This response may be due to the temporal stability of adhesion contacts. Small nascent adhesions (called focal complexes) at the leading edge are capable of exerting strong transient forces that drive migration [93]. Forces subsequently decrease as adhesions mature into large plaques that become centrally localized in the cell during migration. More recent work further illustrates the complex interaction of force generation and substrate adhesions. When 3T3 fibroblasts are seeded on magnetic microposts and deflected by an external magnetic field, changes in traction forces occur at peripheral adhesions that are not directly located at the force application site [94]. These data suggest that cell-generated traction forces are transmitted throughout the cytoskeleton. Less is known about the nature of this transmission, but research in this area is active.

4.4.3

Forces in Cell-Cell Interactions

While much of the work investigating cellular traction forces have focused on cell-matrix contractility, recent work has adapted traction force microscopy methods to the study of forces exerted at cell-cell contacts as well. Cadherins are intimately linked to intracellular signaling mechanisms that influence cellular contractility and force generation (Fig. 4.1). For example, vascular endothelial (VE)-cadherin mediated cell-cell contact activates a Rho signaling pathway, resulting in increased intracellular tension [95] and time-dependent adhesion strengthening [96]. Interestingly, this cadherin-mediated increase in contractility is actin dependent, similar to integrin-mediated contractility. Cadherins are capable of mediating cellular contractility, and there is growing interest in measuring these forces. To do this, recent work has coated deformable substrates with cadherin fragments, thereby simulating cell-cell contacts on a matrix used for traction measurements. Cadherins directly transmit traction forces [97, 98] that increase with increasing substrate rigidity and require actin assembly and myosin activity [99]. These data implicate cadherins as integral components of cellular mechanosensing.

4.5

Extracellular Matrix Remodeling by Cell-Generated Forces

Cells have the ability to remodel the ECM, altering its mechanical and chemical composition. ECM remodeling occurs in a variety of physiological and pathological states including embryonic morphogenesis, wound healing, tumor cell invasion, and metastasis. ECM remodeling involves both the deposition and degradation of matrix proteins. Importantly, it also involves the remodeling of the matrix through cellular traction forces, where cell contractile forces can assemble, collapse and/or realign matrix. Here we focus primarily on cell-mediated remodeling of ECM proteins collagen and fibronectin; however, many of these same principles apply to cell interactions with other ECM proteins.

4 Cell-Generated Forces in Tissue Assembly, Function, and Disease

4.5.1

59

Cell-Generated Forces in Collagen Remodeling

Collagen is the most abundant protein in the ECM and provides structural support for adherent cells. Collagen molecules, or tropocollagen, form a triple-helical structure and can spontaneous assemble into fibrils [100, 101]. Several forms of collagen can be found throughout the body, such as collagen I in skin, bone, and the vasculature and collagen IV in cartilage. Because collagen is easily polymerized into gels, researchers have exploited this property to create three-dimensional bioscaffolds to study the behavior of cells that are entrapped within. Mechanical properties of the gels, including stiffness and porosity, can be controlled by altering the density of collagen when polymerizing gels. It is worth noting that the entrapped cells themselves also contribute to the mechanical properties of the gel as a passive component of a composite material [102]. Additionally, entrapped cells can remodel collagen by remodeling the orientation of collagen fibers. They generate contractile forces which exceed the mechanical resistance of the matrix and lead to the inward compaction of the fibrillar network [103]. Obviously, collagen compaction results in changes in collagen gel density, matrix stiffness, and porosity, which in turn are mechanosensed by cells, oftentimes resulting in changes in cell phenotypes. Not surprisingly, cell morphology is governed by mechanical constraints of the collagen gel – cells entrapped within free-floating gels that do not resist compaction adopt a “balled-up” morphology, whereas in constrained gels attached to an adherent surface, cells are better spread and develop stress fibers [104]. Knowledge of the effects of cell-mediated collagen compaction is important for the design of tissue-engineered constructs [105, 106]. Particularly, the influence of endogenous force-driven collagen compaction on cell proliferation and scaffold mechanical properties is currently under intense investigation. In collagen gels that were not allowed to undergo compaction, cell viability and collagen synthesis were diminished and matrix metalloproteinease production increased, all of which contribute to a significant decrease in the mechanical strength of the matrix [107]. There is a clear need to predict local microstructures and mechanical responses of cell-compacted collagen gels due to physiological forces like mechanical stretch [108]. Such research is important in advancing our understanding of the relationship between ECM microarchitecture and macroscopic mechanical properties to better engineer tissue constructs that meet the rigorous mechanical demands within the body while preserving cellular function.

4.5.2

Cell-Generated Forces in Fibronectin Remodeling

Significant work has been done to elucidate the role of cellular force in fibronectin (Fn) remodeling and assembly. Fn is a dimeric glycoprotein found in most ECMs and basement membranes. Unlike collagen that self-polymerizes in physiological conditions, Fn is assembled into fibrils by cells [109] through cell-generated forces

60

J. Huynh et al.

that stretch and unfold dimer arms [110]. Rho-mediated contractility promotes Fn assembly by exposing cryptic self-assembly sites [111] in Fn type III repeats, domains containing a b-sandwich structure that can be unfolded to a mechanically stable state [112]. The orientation of Fn fibrils is subsequently guided directly by cellular traction forces [113]. Furthermore, experiments with transformed cells capable of assembling exogenously supplied Fn but incapable of producing Fn indicate that Fn polymerization is requisite for collagen gel contraction [114] and the maintenance of the fibrillar organization of ECM Fn [115]. The absence of continuous Fn assembly leads to a decrease in fibril formation and increased Fn degradation through caveolin-mediated endocytosis [116]. Additionally, the 3D organization of fibrillar Fn matrices stimulates integrin-mediated Fn assembly [117]. It is thought that cell adhesion to Fn elicits Rho-ROCK signaling, where ROCK modulation by adhesion requires cytoskeletal-tension mediated traction forces. Interestingly, cadherin adhesions also play a role in cellular contractility that direct Fn fibril formation [118]. These data demonstrate that cell-generated traction forces direct Fn fibrillogenesis and assembly and suggest that forces mediate Fn ECM maintenance important for tissue homeostasis.

4.5.3

Force-Mediated Matrix Remodeling During Tissue Assembly

Traction forces have been implicated in the remodeling of ECM during tissue assembly events. Previous work indicates that capillary development is regulated by mechanical interactions between the cell and flexible ECM networks that facilitate cell contraction and traction stress generation [18]. After weeks in culture, endothelial cells (ECs) in confluent monolayers spontaneously form networks after the accumulation of adhesive tendrils on top of the monolayer. EC generated tension results in elevation of ECM webs above the monolayer that support cell reorganization into capillary networks. Similarly, bovine aortic ECs, fibroblasts, smooth muscle cells, and TM3 murine leydig cells form cellular networks only after tiled networks of basement membrane matrix appear [119]. Further investigation indicates that bovine aortic ECs synthesize and organize type I collagen into extracellular cables that promote cell morphology changes where cells become spindle-shaped and form networks. This reorganization requires traction force and collagen I synthesis [120]. In 3D, human umbilical vein EC tubulogenesis in fibrin matrices requires Fn matrix fibrillogenesis to promote cytoskeletal organization and actomyosin-dependent tension. ECs require Fn assembly and myosin ATPase activity, which increases intracellular stiffness to match the surrounding matrix stiffness during neovessel formation [121]. In our own lab we have shown that EC network formation requires Fn assembly to stabilize cell-cell interactions during network assembly on compliant substrates [19]. In addition to promoting Fn assembly, the traction forces exerted by a cell on sufficiently compliant substrates may stiffen the surrounding substrate, which in turn

4 Cell-Generated Forces in Tissue Assembly, Function, and Disease

61

can be perceived by adjacent cells, causing cells to be drawn together in network formation [39]. This may help facilitate tension-based guidance pathways allowing cells to sense each other through the substrate and at a distance to organize into multicellular structures [122]. Taken together, these data suggest that ECM remodeling by endogenous force generation and ECM remodeling are critical for cell assembly and tissue formation events.

4.6 4.6.1

Forces in Cell Motility, Migration, and Proliferation Forces in Single Cell Motility

One of the core functions of traction generation is to drive cell motility [123], a process that is necessary during cell organization and tissue assembly. Differential motility among cell types is responsible for cell sorting and self-organization in tissues [124]. Local ECM chemistry and mechanics that affect cell-generated forces are likely responsible for the differential motility of cells during development and assembly. Early experiments indicate that changes in cell migration may be attributed to substrate ligand density, integrin expression level, and ligand binding affinity [125]. Maximal cell migration requires high levels of membrane activity and intermediate levels of cell-substrate adhesion [126]. Additionally, Maheshwari and co-workers show that ECM density affects growth factor stimulated migration, including migration speed and membrane protrusion activity. Specifically, when stimulated with epidermal growth factor, NR6 fibroblasts exhibit a biphasic response of cell speed and membrane protrusion activity that is maximal at intermediate concentrations of substrate fibronectin. Other studies show that in addition to substrate ligand density, substrate stiffness also mediates cell migration [127, 128]. The phenomenon of durotaxis, where cells migrate in response to a stiffness gradient, was identified a decade ago in fibroblasts [83]. Moreover, mechanical strains created by substrate manipulation near polarized cells were also found to guide motility. More recent work with bovine aortic smooth muscle cells indicates that durotaxis and cell orientation increase with increasing stiffness gradient magnitude [129]. Although more research is necessary to fully understand the role of substrate mechanics in cell migration, it is a clear parameter in dictating cell locomotion.

4.6.2

Forces in Collective Cell Migration

While much of the work in elucidating the role of cell forces in migration has been performed with single cells, there is also a significant role for traction generation in collective cell migration. Collective cell migration, the phenomenon of migration of

62

J. Huynh et al.

cell groups [130], occurs during development, wound healing, and metastasis. Less is known about traction forces in cell aggregates, largely because few techniques exist to make these measurements. Early work has been done to investigate collective cell migration and traction force generation in Madin-Darby canine kidney cell (MDKC) epithelial sheets as a model system for the epithelial-mesenchymal transition [131]. Epithelial cell scatter of MDKCs increases with increasing concentrations of ECM protein and correlates with adhesion strength and actomyosindependent contractility. This contractility transmits tension to the cell periphery. Measurements of MDKC tractions on microfabricated polydimethylsiloxane (PDMS) pillars indicate that traction stresses are greatest at the monolayer edge and are greater than those measured in single cells [132]. Separate work indicates that large traction forces are also generated in the bulk of the cell sheet, many cell rows away from the leading edge [133]. These traction forces may act in a global tugof-war that pulls the cell sheet toward the leading edge. Additional work on PDMS substrates of anisotropic stiffness indicate MDKC assembly is anisotropic along the stiffest substrate direction and correlates with traction force and actin cytoskeletal orientation [134]. These data indicate that cells within cell sheets can act in concert to generate traction forces throughout the tissue and further implicate substrate stiffness as a regulator of tissue formation. As this is a nascent area, much work remains to be done to understand the role of traction generation in collective cell movements. As technologies are designed to investigate and quantify collective forces, this field will grow rapidly.

4.6.3

Forces in Cell Proliferation

During tissue formation, cells migrate and proliferate into new structures. Cell-generated forces are critically important during cell proliferation and division [135]. The same interplay described earlier of cell-substrate and cell-cell interactions also mediates proliferation of cells within cell aggregate models of tissues. Recent work suggests that differential patterns of proliferation correspond to local mechanical stresses and cytoskeletal generated tension [136]. Cell proliferation within collections of cells is decreased along the long edges of rectangular substrates compared to proliferation along the edges of square substrates, suggesting that proliferation is mediated by contractile tension from the bulk of the monolayer rather than cells at the edge. In similar experiments, fibroblast proliferation and differentiation on micropatterned PDMS islands correspond to high levels of mechanical stress at the perimeter regardless of island shape [137]. These data indicate that tension at the cell-matrix interface and along the edges of collections of cells can drive proliferation profiles. Forces generated through cell-cell contacts also regulate proliferation. A biphasic proliferative response is seen when cell proliferation is regulated by the amount of neighboring-cell contact – while a single-cell neighbor increases

4 Cell-Generated Forces in Tissue Assembly, Function, and Disease

63

proliferation, an increasing number of contacting cells inhibits proliferation. This relationship is potentially explained by the balance that exists between contractility, spreading, and proliferation. While increases in proliferation are dependent on Rho-mediated intracellular tension, cell-cell contacts may lower proliferation by decreasing cell-ECM contact [138]. Using ECs cultured on patterned PDMS substrates, Nelson and Chen show that VE-cadherin-mediated cell-cell contacts promote opposing proliferation signals [139]. VE-cadherin engagement inhibits cell spreading and arrests proliferation; however, when cell spreading is prevented by culturing cell pairs in bowtie-shaped wells, VE-cadherin stimulates proliferation through changes in Rho-ROCK and cytoskeletal tension. These data again point to the delicate balance between cell-matrix and cell-cell adhesion in regulating cell behavior.

4.7

Force Disruption in Disease

Other than the hematopoietic cells, all cell types require adhesion for normal function. As such, virtually all cell types are capable of generating forces and sensing their mechanical environment. When cell-generated forces are altered due to changes in the mechanical properties of their microenvironment, cell functions such as survival, death, proliferation, adhesion, and migration may become dysregulated. The perturbation of proper force balance or mechanotransduction signaling pathways is thought to result in diseases of the liver [140], heart [141], and vasculature [142], among others; see reviews [143, 144]. Importantly, changes in ECM elasticity and the disruption of force balance are also considered important factors in cancer progression [145]. As an example of cell-force in disease progression, the tumor environment itself is stiffer than its surrounding tissue due to elevated interstitial fluid pressure, increased matrix deposition, and cell proliferation [146]. Currently, there is an emphasis in cancer research on elucidating the mechanical mechanisms that drive tumor progression and metastasis. In a study by Paszek et al., increased Rhodependent cell contractility, mediated by matrix stiffness, was found to disrupt tissue organization and enhance the malignant transformation of mammary epithelial cells [84]. Of note, reducing cytoskeletal tension by inhibiting Rho decreased tumor colony size and proliferation and the expression of the malignant phenotype. These data were critical in establishing the role of matrix stiffness and cellgenerated forces in abnormal cell growth and tumorigenesis. Oftentimes, the mechanical compliance of tissues is altered in injury. After injury of the skin, for instance, platelets aggregate and form a temporary fibrin clot, and myofibroblasts lay down a provisional ECM comprised of collagen [147]. Although myofibroblasts quickly restore the mechanical integrity of damaged tissues, scar tissues have increased stiffness due to excessive collagen deposition, or fibrosis [148]. Unfortunately, fibrosis after injury can affect almost every organ in the body. For example, recent studies investigated the effects of matrix rigidity on cell

64

J. Huynh et al.

differentiation in the early stages of liver fibrosis. Li et al. show that myofibroblastic differentiation of liver portal fibroblasts depends on mechanical tension and that they exhibit increasingly myofibroblastic phenotypes with increasing matrix rigidity [149]. Subsequent in vivo studies confirmed that liver stiffness is influenced by matrix quantity and cross-linking [150]. This increase in stiffness was found to precede liver fibrosis and is thought to activate myofibroblast differentiation. In the cardiovascular system, myocardial infarction is generally followed by scarring of the heart tissue and fibrosis, which leads to increased stiffness and impaired cardiac output. To compensate for this decrease in cardiac output, cardiomyocytes increase in size in order to generate more force, which leads to increased ventricular wall thickness or dilation [151, 152]. Over time, elevated stress levels in the myocardium leads to the dysregulation of ECM remodeling resulting in myocyte apoptosis, necrosis, and cardiac failure [144, 153]. Interestingly, optimal beating of cardiomyocytes has been found to correspond to an optimal matrix elasticity [154]. Engler et al. cultured chick embryonic cardiomyocytes on substrates of varying compliances and observed significant differences in cardiac output and strength of beating. On stiffer matrices that mimic the modulus of fibrotic myocardium, large differences were found between cell strains and their underlying matrix strains, suggesting that cells were overstretching and overstraining themselves. Furthermore, cardiomyocytes on stiffer substrates rarely exhibited the typical striated myofibrils throughout the cytoplasm, and 48 h after plating, less than 10% of cells continued to beat, with cells only beating sporadically [154]. In comparison, cardiomyocytes on substrates softer than striated muscle were able to continuously beat but performed very little contractile work. These results show that for cells to function properly, they must be supported by a matrix of physiologic stiffness and may lose their native phenotype on matrices that are either too soft or too stiff, as occurs in many pathologies. Disruptions in cell-generated forces also have wide implications in disease progression of the vasculature, particularly in the development and advancement of atherosclerosis. Atherosclerosis is a multi-stage disease in which fatty deposits accumulate on the vessel wall and obstruct blood flow (Fig. 4.3). This results in the stiffening of the artery, which loses its elasticity due to excess ECM protein deposition by fibroblasts and smooth muscle cells [155]. The endothelial cell (EC) layer, normally a non-adhesive and non-thrombotic surface, is considered the first mediator in atherosclerosis progression. Dysregulation of the EC layer leads to increased permeability, immune cell attachment, and lipid deposition. Force balance, driven by the mechanical interaction between the endothelium and its basement membrane (Fig. 4.3a), is implicated as a critical factor in the maintenance of a functional endothelial barrier [156, 157]. It is hypothesized that Rho-dependent contractile forces drive ECs to separate, while cell-cell interactions provide tethering forces that maintain endothelial barrier integrity. In a recent study by Birukova et al., AFM was used to test this hypothesis by measuring the stiffness of human pulmonary artery ECs as they were stimulated with known barrier-disruptive or barrier-protective agonists [158]. The authors were able to show that barrier-disruptive agonists increase cell stiffness in the

4 Cell-Generated Forces in Tissue Assembly, Function, and Disease

65

Fig. 4.3 Disruption of force balance in the progression of atherosclerosis. (a) Increased vessel stiffness, particularly the rigidity of the internal elastic lamina, may disrupt endothelial cell force balance leading to increased cell contractility and barrier permeability. (b) Increased substrate and endothelial cell stiffness may recruit monocytes to adhere and transmigrate through the endothelial layer. (c) Increased ECM deposition and vessel stiffness causes vascular smooth muscle cells to proliferate and migrate into the intima, occluding the blood vessel

central region, while barrier-protective agonists increase stiffness at the cell periphery and decrease cell stiffness at the center. These results were confirmed by similar AFM studies on human umbilical vein ECs that show that cell stiffness and membrane tethering forces are directly related to the polymerization state of actin [159]. These results support the hypothesis that cell-generated forces maintain the endothelial barrier and directly link cell stiffness with specific permeability responses. As the permeability of the endothelium becomes dysregulated, lipoprotein deposition initiates the inflammation cascade and recruits immune cells to attach and transmigrate through the endothelial layer (Fig. 4.3b). The adhesion and migration of neutrophils are influenced in part by cell-generated forces and substrate rigidity [160]. Neutrophils are more well-spread and exert greater traction forces when allowed to adhere to stiffer substrates. Additionally, neutrophils migrate more slowly on stiffer substrates but also more persistently, resulting in longer distances being covered. Further studies show that a balance in cell-generated force corresponding to an optimal substrate stiffness maximizes neutrophil migration [161]. These in vitro data suggest that matrix stiffening, which occurs in atherosclerosis, may promote leukocyte recruitment and initiate the inflammatory cascade in vivo. Interestingly, in this same inflammatory cascade, heterogeneous cell-cell interactions alter cell mechanics. The adhesion of neutrophils to the EC layer increases the stiffness of adjacent ECs [162]. When taken together with the findings of Birukova et al. which suggest that cell-generated forces may regulate endothelial permeability, it is possible that neutrophil adhesion causes a local increase in

66

J. Huynh et al.

endothelial force generation and barrier permeability in preparation for their eventual transmigration. This possibility was explored in an elegant study done by Rabodzey et al. The authors measured the tangential forces exerted by neutrophils during transmigration through a confluent EC layer grown on a microfabricated pillar substrate [163]. In response to neutrophil transmigration, endothelial cell-cell junctions become disrupted and EC traction forces increase significantly. Moreover, traction forces induced by neutrophil transmigration increase with increasing substrate rigidity. Therefore, the relationship between matrix stiffness, EC stiffness, and EC force generation plays an integral role in regulating permeability and transendothelial migration. In the later stages of atherosclerosis, vascular smooth muscle cells (VSMCs) migrate from the media into the intima, largely contributing to the occlusion of the blood vessel (Fig. 4.3c). Coincident with this migration is an increase in collagen deposition and increased vessel stiffness. Recent in vitro studies indicate that substrate stiffness has a direct effect on the migration speeds of VSMCs. Data show that VSMC migration speed has a biphasic relationship with respect to substrate stiffness, suggesting that an optimal stiffness exists which maximizes cell migration speed and persistence [127]. On substrates fabricated to contain welldefined stiffness gradients [129], VSMCs orient themselves in the direction of the gradient and preferentially migrate from softer to stiffer regions. While these studies have advanced our understanding of the interplay between mechanics and cardiovascular disease, additional attention is required to fully understand the role of cell-generated forces in endothelial permeability, neutrophil transmigration, and VSMC migration in vivo.

4.8

Summary and Future Directions

As evidenced in this chapter, force balance between cell-cell and cell-matrix adhesion is critical for proper tissue organization, health, and disease. Cell-generated forces are dependent on the surrounding chemical and mechanical microenvironment. Important chemical factors include soluble growth factors and the protein composition of the underlying extracellular matrix, whereas substrate mechanical properties have been found to regulate a variety of cell phenotypes including cell proliferation, migration, and assembly. Interestingly, cells utilize traction forces to remodel their microenvironment, as occurs in collagen and fibronectin remodeling. This creates a mechanical feedback loop where cells are able to alter their chemical and mechanical microenvironment, which in turn affects their behavior in tissue formation and homeostasis. This regulatory circuit is critical for the correct differentiation and sorting of cells required in embryonic development, and disruption of force balance dysregulates the feedback loop, resulting in the progression of diseases such as atherosclerosis. Ever since the early works of Malcolm Steinberg, researchers have rigorously studied the nature and importance of cell-generated forces. The advent of

4 Cell-Generated Forces in Tissue Assembly, Function, and Disease

67

technologies such as traction force microscopy and microfabricated post array detectors allowed for the quantification of cell forces. However, these techniques require cells to be cultured on two-dimensional substrates that do not resemble the three-dimensional environments in which cells are found in vivo. To address this, future work should consider quantitative analysis of traction forces of cells suspended in three-dimensional matrices. Additionally, although current technology allows researchers to measure forces at the molecular level, such as between cadherin bonds, and at the cellular level, as described thoroughly in this chapter, additional work is required for the quantification of forces exerted by tissues or large cell aggregates. Novel methodologies in these areas will aid in characterizing traction forces found in vivo.

References 1. Steinberg MS (1962) Mechanism of tissue reconstruction by dissociated cells. II. Time-course of events. Science 137:762–763 2. Steinberg MS (1962) On the mechanism of tissue reconstruction by dissociated cells, Iii. Free energy relations and the reorganization of fused, heteronomic tissue fragments. Proc Natl Acad Sci U S A 48:1769–1776 3. Steinberg MS (1962) On the mechanism of tissue reconstruction by dissociated cells. I. Population kinetics, differential adhesiveness and the absence of directed migration. Proc Natl Acad Sci U S A 48:1577–1582 4. Steinberg MS (1970) Does differential adhesion govern self-assembly processes in histogenesis? Equilibrium configurations and the emergence of a hierarchy among populations of embryonic cells. J Exp Zool 173:395–433 5. Steinberg MS and Garrod DR (1975) Observations on the sorting-out of embryonic cells in monolayer culture. J Cell Sci 18:385–403 6. Inoue T, Tanaka T, Takeichi M, et al (2001) Role of cadherins in maintaining the compartment boundary between the cortex and striatum during development. Development 128:561–569 7. Jia D, Dajusta D, and Foty RA (2007) Tissue surface tensions guide in vitro self-assembly of rodent pancreatic islet cells. Dev Dyn 236:2039–2049 8. Ryan PL, Foty RA, Kohn J, et al (2001) Tissue spreading on implantable substrates is a competitive outcome of cell-cell vs. cell-substratum adhesivity. Proc Natl Acad Sci U S A 98:4323–4327 9. Foty RA, Forgacs G, Pfleger CM, et al (1994) Liquid properties of embryonic tissues: measurement of interfacial tensions. Phys Rev Lett 72:2298–2301 10. Foty RA and Steinberg MS (2005) The differential adhesion hypothesis: a direct evaluation. Dev Biol 278:255–263 11. Shi Q, Chien YH, and Leckband D (2008) Biophysical properties of cadherin bonds do not predict cell sorting. J Biol Chem 283:28454–28463 12. Krieg M, Arboleda-Estudillo Y, Puech PH, et al (2008) Tensile forces govern germ-layer organization in zebrafish. Nat Cell Biol 10:429–436 13. Harris AK (1976) Is cell sorting caused by differences in the work of intercellular adhesion? A critique of the Steinberg hypothesis. J Theor Biol 61:267–285 14. Brodland GW and Chen HH (2000) The mechanics of heterotypic cell aggregates: insights from computer simulations. J Biomech Eng 122:402–407

68

J. Huynh et al.

15. Powers MJ, Rodriguez RE, and Griffith LG (1997) Cell-substratum adhesion strength as a determinant of hepatocyte aggregate morphology. Biotechnol Bioeng 53:415–426 16. Powers MJ and Griffith LG (1998) Adhesion-guided in vitro morphogenesis in pure and mixed cell cultures. Microsc Res Tech 43:379–384 17. Young TH, Tu HR, Chan CC, et al (2009) The enhancement of dermal papilla cell aggregation by extracellular matrix proteins through effects on cell-substratum adhesivity and cell motility. Biomaterials 30:5031–5040 18. Ingber DE and Folkman J (1989) Mechanochemical switching between growth and differentiation during fibroblast growth factor-stimulated angiogenesis in vitro: role of extracellular matrix. J Cell Biol 109:317–330 19. Califano JP and Reinhart-King CA (2008) A balance of substrate mechanics and matrix chemistry regulates endothelial cell network assembly. Cell Mol Bioeng 1:122–132 20. Rosines E, Schmidt HJ, and Nigam SK (2007) The effect of hyaluronic acid size and concentration on branching morphogenesis and tubule differentiation in developing kidney culture systems: potential applications to engineering of renal tissues. Biomaterials 28:4806–4817 21. Genes NG, Rowley JA, Mooney DJ, et al (2004) Effect of substrate mechanics on chondrocyte adhesion to modified alginate surfaces. Arch Biochem Biophys 422:161–167 22. Engler A, Bacakova L, Newman C, et al (2004) Substrate compliance versus ligand density in cell on gel responses. Biophys J 86:617–628 23. Ghibaudo M, Saez A, Trichet L, et al (2008) Traction forces and rigidity sensing regulate cell functions. Soft Matter 4:1836–1843 24. Yeung T, Georges PC, Flanagan LA, et al (2005) Effects of substrate stiffness on cell morphology, cytoskeletal structure, and adhesion. Cell Motil Cytoskeleton 60:24–34 25. Pelham RJ, Jr. and Wang Y (1997) Cell locomotion and focal adhesions are regulated by substrate flexibility. Proc Natl Acad Sci U S A 94:13661–13665 26. Califano JP and Reinhart-King CA (2010) Exogenous and endogenous force regulation of endothelial cell behavior. J Biomech 43:79–86 27. Reinhart-King CA, Fujiwara K, and Berk BC (2008) Physiologic stress-mediated signaling in the endothelium. Meth Enzymol 443:25–44 28. Kuzuya M, Satake S, Miura H, et al (1996) Inhibition of endothelial cell differentiation on a glycosylated reconstituted basement membrane complex. Exp Cell Res 226:336–345 29. Deroanne CF, Lapiere CM, and Nusgens BV (2001) In vitro tubulogenesis of endothelial cells by relaxation of the coupling extracellular matrix-cytoskeleton. Cardiovas Res 49:647–658 30. Stephanou A, Meskaoui G, Vailhe B, et al (2007) The rigidity in fibrin gels as a contributing factor to the dynamics of in vitro vascular cord formation. Microvasc Res 73:182–190 31. Nehls V and Herrmann R (1996) The configuration of fibrin clots determines capillary morphogenesis and endothelial cell migration. Microvasc Res 51:347–364 32. Vailhe B, Ronot X, Tracqui P, et al (1997) In vitro angiogenesis is modulated by the mechanical properties of fibrin gels and is related to alpha(v)beta3 integrin localization. In Vitro Cell Dev Biol Anim 33:763–773 33. Sieminski AL, Hebbel RP, and Gooch KJ (2004) The relative magnitudes of endothelial force generation and matrix stiffness modulate capillary morphogenesis in vitro. Exp Cell Res 297:574–584 34. Krishnan L, Hoying JB, Nguyen H, et al (2007) Interaction of angiogenic microvessels with the extracellular matrix. Am J Physiol Heart Circ Physiol 293:H3650–H3658 35. Mammoto A, Connor KM, Mammoto T, et al (2009) A mechanosensitive transcriptional mechanism that controls angiogenesis. Nature 457:1103–1108 36. Guo WH, Frey MT, Burnham NA, et al (2006) Substrate rigidity regulates the formation and maintenance of tissues. Biophys J 90:2213–2220 37. Jakab K, Neagu A, Mironov V, et al (2004) Engineering biological structures of prescribed shape using self-assembling multicellular systems. Proc Natl Acad Sci U S A 101:2864–2869

4 Cell-Generated Forces in Tissue Assembly, Function, and Disease

69

38. Gehler S, Baldassarre M, Lad Y, et al (2009) Filamin A-beta1 integrin complex tunes epithelial cell response to matrix tension. Mol Biol Cell 20:3224–3238 39. Reinhart-King CA, Dembo M, and Hammer DA (2008) Cell-cell mechanical communication through compliant substrates. Biophys J 95:6044–6051 40. Namy P, Ohayon J, and Tracqui P (2004) Critical conditions for pattern formation and in vitro tubulogenesis driven by cellular traction fields. J Theor Biol 227:103–120 41. Wozniak MA and Chen CS (2009) Mechanotransduction in development: a growing role for contractility. Nat Rev Mol Cell Biol 10:34–43 42. Shraiman BI (2005) Mechanical feedback as a possible regulator of tissue growth. Proc Natl Acad Sci U S A 102:3318–3323 43. Adams DS, Keller R, and Koehl MA (1990) The mechanics of notochord elongation, straightening and stiffening in the embryo of Xenopus laevis. Development 110:115–130 44. Engler AJ, Sen S, Sweeney HL, et al (2006) Matrix elasticity directs stem cell lineage specification. Cell 126:677–689 45. Evans ND, Minelli C, Gentleman E, et al (2009) Substrate stiffness affects early differentiation events in embryonic stem cells. Eur Cell Mater 18:1–13; discussion 13–14 46. Pajerowski JD, Dahl KN, Zhong FL, et al (2007) Physical plasticity of the nucleus in stem cell differentiation. Proc Natl Acad Sci U S A 104:15619–15624 47. Chowdhury F, Na S, Li D, et al (2010) Material properties of the cell dictate stress-induced spreading and differentiation in embryonic stem cells. Nat Mater 9:82–88 48. Sordella R, Jiang W, Chen GC, et al (2003) Modulation of Rho GTPase signaling regulates a switch between adipogenesis and myogenesis. Cell 113:147–158 49. McBeath R, Pirone DM, Nelson CM, et al (2004) Cell shape, cytoskeletal tension, and RhoA regulate stem cell lineage commitment. Dev Cell 6:483–495 50. Moore KA, Polte T, Huang S, et al (2005) Control of basement membrane remodeling and epithelial branching morphogenesis in embryonic lung by Rho and cytoskeletal tension. Dev Dyn 232:268–281 51. Nishimura T and Takeichi M (2009) Remodeling of the adherens junctions during morphogenesis. Curr Top Dev Biol 89:33–54 52. Delon I and Brown NH (2007) Integrins and the actin cytoskeleton. Curr Opin Cell Biol 19:43–50 53. Zaidel-Bar R, Itzkovitz S, Ma’ayan A, et al (2007) Functional atlas of the integrin adhesome. Nat Cell Biol 9:858–867 54. Zaidel-Bar R and Geiger B (2010) The switchable integrin adhesome. J Cell Sci 123:1385–1388 55. Ogita H and Takai Y (2008) Cross-talk among integrin, cadherin, and growth factor receptor: roles of nectin and nectin-like molecule. Int Rev Cytol 265:1–54 56. Schwartz MA and DeSimone DW (2008) Cell adhesion receptors in mechanotransduction. Curr Opin Cell Biol 20:551–556 57. Kumar S, Maxwell IZ, Heisterkamp A, et al (2006) Viscoelastic retraction of single living stress fibers and its impact on cell shape, cytoskeletal organization, and extracellular matrix mechanics. Biophys J 90:3762–3773 58. Chicurel ME, Chen CS, and Ingber DE (1998) Cellular control lies in the balance of forces. Curr Opin Cell Biol 10:232–239 59. Maniotis AJ, Chen CS, and Ingber DE (1997) Demonstration of mechanical connections between integrins, cytoskeletal filaments, and nucleoplasm that stabilize nuclear structure. Proc Natl Acad Sci U S A 94:849–854 60. Katsumi A, Orr AW, Tzima E, et al (2004) Integrins in mechanotransduction. J Biol Chem 279:12001–12004 61. Chen CS (2008) Mechanotransduction – a field pulling together? J Cell Sci 121:3285–3292 62. Orr AW, Helmke BP, Blackman BR, et al (2006) Mechanisms of mechanotransduction. Dev Cell 10:11–20

70

J. Huynh et al.

63. Discher DE, Janmey P, and Wang YL (2005) Tissue cells feel and respond to the stiffness of their substrate. Science 310:1139–1143 64. Vogel V and Sheetz M (2006) Local force and geometry sensing regulate cell functions. Nat Rev Mol Cell Biol 7:265–275 65. Oliver T, Jacobson K, and Dembo M (1998) Design and use of substrata to measure traction forces exerted by cultured cells. Meth Enzymol 298:497–521 66. Harris AK, Wild P, and Stopak D (1980) Silicone rubber substrata: a new wrinkle in the study of cell locomotion. Science 208:177–179 67. Lee J, Leonard M, Oliver T, et al (1994) Traction forces generated by locomoting keratocytes. J Cell Biol 127:1957–1964 68. Dembo M, Oliver T, Ishihara A, et al (1996) Imaging the traction stresses exerted by locomoting cells with the elastic substratum method. Biophys J 70:2008–2022 69. Marganski WA, Dembo M, and Wang YL (2003) Measurements of cell-generated deformations on flexible substrata using correlation-based optical flow. Meth Enzymol 361:197–211 70. Dembo M and Wang YL (1999) Stresses at the cell-to-substrate interface during locomotion of fibroblasts. Biophys J 76:2307–2316 71. Sabass B, Gardel ML, Waterman CM, et al (2008) High resolution traction force microscopy based on experimental and computational advances. Biophys J 94:207–220 72. Del Alamo JC, Meili R, Alonso-Latorre B, et al (2007) Spatio-temporal analysis of eukaryotic cell motility by improved force cytometry. Proc Natl Acad Sci U S A 104:13343–13348 73. Butler JP, Tolic-Norrelykke IM, Fabry B, et al (2002) Traction fields, moments, and strain energy that cells exert on their surroundings. Am J Physiol Cell Physiol 282:C595–C605 74. Tan JL, Tien J, Pirone DM, et al (2003) Cells lying on a bed of microneedles: an approach to isolate mechanical force. Proc Natl Acad Sci U S A 100:1484–1489 75. Sniadecki NJ and Chen CS (2007) Microfabricated silicone elastomeric post arrays for measuring traction forces of adherent cells. Meth Cell Biol 83:313–328 76. Galbraith CG and Sheetz MP (1997) A micromachined device provides a new bend on fibroblast traction forces. Proc Natl Acad Sci U S A 94:9114–9118 77. Balaban NQ, Schwarz US, Riveline D, et al (2001) Force and focal adhesion assembly: a close relationship studied using elastic micropatterned substrates. Nat Cell Biol 3:466–472 78. Wang N, Ostuni E, Whitesides GM, et al (2002) Micropatterning tractional forces in living cells. Cell Motil Cytoskeleton 52:97–106 79. Parker KK, Brock AL, Brangwynne C, et al (2002) Directional control of lamellipodia extension by constraining cell shape and orienting cell tractional forces. FASEB J 16:1195–1204 80. Lemmon CA, Sniadecki NJ, Ruiz SA, et al (2005) Shear force at the cell-matrix interface: enhanced analysis for microfabricated post array detectors. Mech Chem Biosyst 2:1–16 81. Reinhart-King CA, Dembo M, and Hammer DA (2003) Endothelial cell traction forces on RGD-derivatized polyacrylamide substrata. Langmuir 19:1573–1579 82. Reinhart-King CA, Dembo M, and Hammer DA (2005) The dynamics and mechanics of endothelial cell spreading. Biophys J 89:676–689 83. Lo CM, Wang HB, Dembo M, et al (2000) Cell movement is guided by the rigidity of the substrate. Biophys J 79:144–152 84. Paszek MJ, Zahir N, Johnson KR, et al (2005) Tensional homeostasis and the malignant phenotype. Cancer Cell 8:241–254 85. Califano JP and Reinhart-King CA (2010) Substrate stiffness and cell area drive cellular traction stresses in single cells and cells in contact. Cell Mol Bioeng 3:68–75 86. Ren XD, Kiosses WB, and Schwartz MA (1999) Regulation of the small GTP-binding protein Rho by cell adhesion and the cytoskeleton. EMBO J 18:578–585 87. Ingber DE (2002) Mechanical signaling and the cellular response to extracellular matrix in angiogenesis and cardiovascular physiology. Circ Res 91:877–887 88. Leung DW, Cachianes G, Kuang WJ, et al (1989) Vascular endothelial growth factor is a secreted angiogenic mitogen. Science 246:1306–1309

4 Cell-Generated Forces in Tissue Assembly, Function, and Disease

71

89. Nicosia RF (1998) What is the role of vascular endothelial growth factor-related molecules in tumor angiogenesis? Am J Pathol 153:11–16 90. Huot J, Houle F, Rousseau S, et al (1998) SAPK2/p38-dependent F-actin reorganization regulates early membrane blebbing during stress-induced apoptosis. J Cell Biol 143:1361–1373 91. van Nieuw Amerongen GP, Koolwijk P, Versteilen A, et al (2003) Involvement of RhoA/Rho kinase signaling in VEGF-induced endothelial cell migration and angiogenesis in vitro. Arterioscler Thromb Vasc Biol 23:211–217 92. Appleton CT, Usmani SE, Mort JS, et al (2010) Rho/ROCK and MEK/ERK activation by transforming growth factor-alpha induces articular cartilage degradation. Lab Invest 90:20–30 93. Beningo KA, Dembo M, Kaverina I, et al (2001) Nascent focal adhesions are responsible for the generation of strong propulsive forces in migrating fibroblasts. J Cell Biol 153: 881–888 94. Sniadecki NJ, Anguelouch A, Yang MT, et al (2007) Magnetic microposts as an approach to apply forces to living cells. Proc Natl Acad Sci U S A 104:14553–14558 95. Nelson CM, Pirone DM, Tan JL, et al (2004) Vascular endothelial-cadherin regulates cytoskeletal tension, cell spreading, and focal adhesions by stimulating RhoA. Mol Biol Cell 15:2943–2953 96. Kris AS, Kamm RD, and Sieminski AL (2008) VASP involvement in force-mediated adherens junction strengthening. Biochem Biophys Res Commun 375:134–138 97. Ko KS, Arora PD, and McCulloch CA (2001) Cadherins mediate intercellular mechanical signaling in fibroblasts by activation of stretch-sensitive calcium-permeable channels. J Biol Chem 276:35967–35977 98. Ganz A, Lambert M, Saez A, et al (2006) Traction forces exerted through N-cadherin contacts. Biol Cell 98:721–730 99. Ladoux B, Anon E, Lambert M, et al (2010) Strength dependence of cadherin-mediated adhesions. Biophys J 98:534–542 100. Ramachandran GN (1988) Stereochemistry of collagen. Int J Pept Protein Res 31:1–16 101. Brodsky B and Baum J (2008) Structural biology: modelling collagen diseases. Nature 453:998–999 102. Wakatsuki T, Kolodney MS, Zahalak GI, et al (2000) Cell mechanics studied by a reconstituted model tissue. Biophys J 79:2353–2368 103. Tranquillo RT (1999) Self-organization of tissue-equivalents: the nature and role of contact guidance. Biochem Soc Symp 65:27–42 104. Roy P, Petroll WM, Cavanagh HD, et al (1997) An in vitro force measurement assay to study the early mechanical interaction between corneal fibroblasts and collagen matrix. Exp Cell Res 232:106–117 105. Jhun CS, Evans MC, Barocas VH, et al (2009) Planar biaxial mechanical behavior of bioartificial tissues possessing prescribed fiber alignment. J Biomech Eng 131:081006 106. Robinson PS, Johnson SL, Evans MC, et al (2008) Functional tissue-engineered valves from cell-remodeled fibrin with commissural alignment of cell-produced collagen. Tissue Eng Part A 14:83–95 107. Berry CC, Shelton JC, and Lee DA (2009) Cell-generated forces influence the viability, metabolism and mechanical properties of fibroblast-seeded collagen gel constructs. J Tissue Eng Regen Med 3:43–53 108. Sander EA, Stylianopoulos T, Tranquillo RT, et al (2009) Image-based multiscale modeling predicts tissue-level and network-level fiber reorganization in stretched cell-compacted collagen gels. Proc Natl Acad Sci U S A 106:17675–17680 109. Magnusson MK and Mosher DF (1998) Fibronectin: structure, assembly, and cardiovascular implications. Arterioscler Thromb Vasc Biol 18:1363–1370 110. Baneyx G, Baugh L, and Vogel V (2002) Fibronectin extension and unfolding within cell matrix fibrils controlled by cytoskeletal tension. Proc Natl Acad Sci U S A 99:5139–5143

72

J. Huynh et al.

111. Zhong C, Chrzanowska-Wodnicka M, Brown J, et al (1998) Rho-mediated contractility exposes a cryptic site in fibronectin and induces fibronectin matrix assembly. J Cell Biol 141:539–551 112. Gao M, Craig D, Lequin O, et al (2003) Structure and functional significance of mechanically unfolded fibronectin type III1 intermediates. Proc Natl Acad Sci U S A 100:14784–14789 113. Lemmon CA, Chen CS, and Romer LH (2009) Cell traction forces direct fibronectin matrix assembly. Biophys J 96:729–738 114. Hocking DC, Sottile J, and Langenbach KJ (2000) Stimulation of integrin-mediated cell contractility by fibronectin polymerization. J Biol Chem 275:10673–10682 115. Sottile J and Hocking DC (2002) Fibronectin polymerization regulates the composition and stability of extracellular matrix fibrils and cell-matrix adhesions. Mol Biol Cell 13:3546–3559 116. Sottile J and Chandler J (2005) Fibronectin matrix turnover occurs through a caveolin1-dependent process. Mol Biol Cell 16:757–768 117. Bhadriraju K, Yang M, Alom Ruiz S, et al (2007) Activation of ROCK by RhoA is regulated by cell adhesion, shape, and cytoskeletal tension. Exp Cell Res 313:3616–3623 118. Dzamba BJ, Jakab KR, Marsden M, et al (2009) Cadherin adhesion, tissue tension, and noncanonical Wnt signaling regulate fibronectin matrix organization. Dev Cell 16:421–432 119. Vernon RB, Angello JC, Iruela-Arispe ML, et al (1992) Reorganization of basement membrane matrices by cellular traction promotes the formation of cellular networks in vitro. Lab Invest 66:536–547 120. Vernon RB, Lara SL, Drake CJ, et al (1995) Organized type I collagen influences endothelial patterns during “spontaneous angiogenesis in vitro”: planar cultures as models of vascular development. In Vitro Cell Dev Biol Anim 31:120–131 121. Zhou X, Rowe RG, Hiraoka N, et al (2008) Fibronectin fibrillogenesis regulates threedimensional neovessel formation. Genes Dev 22:1231–1243 122. Davis GE and Senger DR (2005) Endothelial extracellular matrix: biosynthesis, remodeling, and functions during vascular morphogenesis and neovessel stabilization. Circ Res 97:1093–1107 123. Oliver T, Dembo M, and Jacobson K (1999) Separation of propulsive and adhesive traction stresses in locomoting keratocytes. J Cell Biol 145:589–604 124. Mori H, Gjorevski N, Inman JL, et al (2009) Self-organization of engineered epithelial tubules by differential cellular motility. Proc Natl Acad Sci U S A 106:14890–14895 125. Palecek SP, Loftus JC, Ginsberg MH, et al (1997) Integrin-ligand binding properties govern cell migration speed through cell-substratum adhesiveness. Nature 385:537–540 126. Maheshwari G, Wells A, Griffith LG, et al (1999) Biophysical integration of effects of epidermal growth factor and fibronectin on fibroblast migration. Biophys J 76:2814–2823 127. Peyton SR and Putnam AJ (2005) Extracellular matrix rigidity governs smooth muscle cell motility in a biphasic fashion. J Cell Physiol 204:198–209 128. Jannat RA, Dembo M, and Hammer DA (2010) Neutrophil adhesion and chemotaxis depend on substrate mechanics. J Phys Condens Matter 22:194117 129. Isenberg BC, Dimilla PA, Walker M, et al (2009) Vascular smooth muscle cell durotaxis depends on substrate stiffness gradient strength. Biophys J 97:1313–1322 130. Rorth P (2007) Collective guidance of collective cell migration. Trends Cell Biol 17:575–579 131. de Rooij J, Kerstens A, Danuser G, et al (2005) Integrin-dependent actomyosin contraction regulates epithelial cell scattering. J Cell Biol 171:153–164 132. du Roure O, Saez A, Buguin A, et al (2005) Force mapping in epithelial cell migration. Proc Natl Acad Sci U S A 102:2390–2395 133. Trepat X, Wasserman MR, Angelini TE, et al (2009) Physical forces during collective cell migration. Nat Phys 5:426–430

4 Cell-Generated Forces in Tissue Assembly, Function, and Disease

73

134. Saez A, Ghibaudo M, Buguin A, et al (2007) Rigidity-driven growth and migration of epithelial cells on microstructured anisotropic substrates. Proc Natl Acad Sci U S A 104: 8281–8286 135. Scholey JM, Brust-Mascher I, and Mogilner A (2003) Cell division. Nature 422:746–752 136. Nelson CM, Jean RP, Tan JL, et al (2005) Emergent patterns of growth controlled by multicellular form and mechanics. Proc Natl Acad Sci U S A 102:11594–11599 137. Li B, Li F, Puskar KM, et al (2009) Spatial patterning of cell proliferation and differentiation depends on mechanical stress magnitude. J Biomech 42:1622–1627 138. Gray DS, Liu WF, Shen CJ, et al (2008) Engineering amount of cell-cell contact demonstrates biphasic proliferative regulation through RhoA and the actin cytoskeleton. Exp Cell Res 314:2846–2854 139. Nelson CM and Chen CS (2003) VE-cadherin simultaneously stimulates and inhibits cell proliferation by altering cytoskeletal structure and tension. J Cell Sci 116:3571–3581 140. Wells RG (2008) The role of matrix stiffness in regulating cell behavior. Hepatology 47:1394–1400 141. Lammerding J, Kamm RD, and Lee RT (2004) Mechanotransduction in cardiac myocytes. Ann N Y Acad Sci 1015:53–70 142. Haga JH, Li YS, and Chien S (2007) Molecular basis of the effects of mechanical stretch on vascular smooth muscle cells. J Biomech 40:947–960 143. Ingber DE (2003) Mechanobiology and diseases of mechanotransduction. Ann Med 35:564–577 144. Jaalouk DE and Lammerding J (2009) Mechanotransduction gone awry. Nat Rev Mol Cell Biol 10:63–73 145. Butcher DT, Alliston T, and Weaver VM (2009) A tense situation: forcing tumour progression. Nat Rev Cancer 9:108–122 146. Huang S and Ingber DE (2005) Cell tension, matrix mechanics, and cancer development. Cancer Cell 8:175–176 147. Midwood KS, Williams LV, and Schwarzbauer JE (2004) Tissue repair and the dynamics of the extracellular matrix. Int J Biochem Cell Biol 36:1031–1037 148. Hinz B (2009) Tissue stiffness, latent TGF-beta1 activation, and mechanical signal transduction: implications for the pathogenesis and treatment of fibrosis. Curr Rheumatol Rep 11:120–126 149. Li Z, Dranoff JA, Chan EP, et al (2007) Transforming growth factor-beta and substrate stiffness regulate portal fibroblast activation in culture. Hepatology 46:1246–1256 150. Georges PC, Hui JJ, Gombos Z, et al (2007) Increased stiffness of the rat liver precedes matrix deposition: implications for fibrosis. Am J Physiol Gastrointest Liver Physiol 293: G1147–G1154 151. Grossman W, Jones D, and McLaurin LP (1975) Wall stress and patterns of hypertrophy in the human left ventricle. J Clin Invest 56:56–64 152. Sun Y and Weber KT (2000) Infarct scar: a dynamic tissue. Cardiovasc Res 46:250–256 153. Barry SP, Davidson SM, and Townsend PA (2008) Molecular regulation of cardiac hypertrophy. Int J Biochem Cell Biol 40:2023–2039 154. Engler AJ, Carag-Krieger C, Johnson CP, et al (2008) Embryonic cardiomyocytes beat best on a matrix with heart-like elasticity: scar-like rigidity inhibits beating. J Cell Sci 121:3794–3802 155. Zieman SJ, Melenovsky V, and Kass DA (2005) Mechanisms, pathophysiology, and therapy of arterial stiffness. Arterioscler Thromb Vasc Biol 25:932–943 156. Ingber DE (1997) Tensegrity: the architectural basis of cellular mechanotransduction. Annu Rev Physiol 59:575–599 157. Mehta D and Malik AB (2006) Signaling mechanisms regulating endothelial permeability. Physiol Rev 86:279–367

74

J. Huynh et al.

158. Birukova AA, Arce FT, Moldobaeva N, et al (2009) Endothelial permeability is controlled by spatially defined cytoskeletal mechanics: atomic force microscopy force mapping of pulmonary endothelial monolayer. Nanomedicine 5:30–41 159. Cuerrier CM, Gagner A, Lebel R, et al (2009) Effect of thrombin and bradykinin on endothelial cell mechanical properties monitored through membrane deformation. J Mol Recognit 22:389–396 160. Oakes PW, Patel DC, Morin NA, et al (2009) Neutrophil morphology and migration are affected by substrate elasticity. Blood 114:1387–1395 161. Stroka KM and Aranda-Espinoza H (2009) Neutrophils display biphasic relationship between migration and substrate stiffness. Cell Motil Cytoskeleton 66:328–341 162. Kang I, Wang Q, Eppell SJ, et al (2010) Effect of neutrophil adhesion on the mechanical properties of lung microvascular endothelial cells. Am J Respir Cell Mol Biol 43:591–598 163. Rabodzey A, Alcaide P, Luscinskas FW, et al (2008) Mechanical forces induced by the transendothelial migration of human neutrophils. Biophys J 95:1428–1438

Chapter 5

Cell-Cell Interactions and the Mechanics of Cells and Tissues Observed in Bioartificial Tissue Constructs Guy M. Genin, Teresa M. Abney, Tetsuro Wakatsuki, and Elliot L. Elson

This chapter is part of Section II: Cooperative Cell Behavior and Mechanobiology

Abstract Mechanical interactions among cells in living tissues play central roles in several physiologic and pathologic phenomena. A large body of work from the mechanical study of cells in bioartificial tissue constructs sheds light on these interactions. This chapter summarizes models used to derive mechanical responses of cells from tissue constructs and then presents examples of three classes of biophysical observations that can be understood through simplified analysis of mechanical interactions between cells. The first observation is the effective mechanical behavior of cells: contractile fibroblast cells appear to modulate both their effective stiffness and that of the extracellular matrix in which they are embedded in a way that is regulated by their volume fraction within a tissue construct; these cells and their extracellular matrix have matched effective relaxed moduli of approximately 10–20 kPa when the volume fraction of cells within the tissue construct is that associated with the percolation threshold. Additionally, cells appear to proliferate or die off to reach this volume fraction. The second observation is the ensemble average mechanical properties of subcellular protein structures, when contractile fibroblasts exist at a volume fraction above the percolation threshold, the dynamic behaviors of their cytoskeletons are evident from the mechanics of the tissue as a whole. The third observation is the pathological effects of cell-cell interactions: tissue constructs containing both cardiomyocytes and proliferative cardiac fibroblasts serve as a model of fibrosis of the heart.

G.M. Genin (*) Department of Mechanical, Aerospace, and Structural Engineering, Washington University, St. Louis, MO 63130, USA e-mail: [email protected]

A. Wagoner Johnson and Brendan A.C. Harley (eds.), Mechanobiology of Cell-Cell and Cell-Matrix Interactions, DOI 10.1007/978-1-4419-8083-0_5, # Springer Science+Business Media, LLC 2011

75

76

5.1

G.M. Genin et al.

Tissue Constructs as Analytical Systems for Studying Cell and Tissue Structure and Function

Interactions and interconnections between cells have important consequences for the functional characteristics of a tissue. For example, gap junctions coordinate the contractions of cardiomyocytes, while cadherin-based anchoring junctions including desmosomes and adherens junctions are of central importance for fibroblast networks. Formation of multiple, possibly interconnected, networks affects the mechanical and electrical environment of cells within tissues. This chapter describes ways that the mechanics of such networks can be quantified and presents examples of the physiologic importance of understanding mechanical cell-cell interactions. The focus here is in exploiting these interactions in tissue constructs. We prefer these to natural tissues for study of cell mechanics because, while cells are often a minor contributor to overall mechanics of natural structural tissues such as tendons, tissue constructs can be designed to emphasize cellular contributions. We prefer three-dimensional (3D) tissue constructs over traditional 2D cultures for two reasons. First, cells in 3D constructs behave more like cells in natural tissues than do cells in monolayer cultures. Increasing evidence indicates that cells in 2D cultures differ substantially in structure and function from cells in a natural tissue [1]. Cellular signal transduction, migration, growth, development, and mechanical interaction with ECM are all markedly different in 2D vs. 3D cultures [2–5]. Since the mechanical environment of cells dictates many features of development, the absence of out-of-plane stiffness and the rigidity of the substratum may be important factors in these differences [6, 7]. Second, quantifying cell mechanics in monolayer culture poses a fundamental challenge because accessing the natural structural network of a cell is difficult when the cell cytoskeleton remodels to adhere to a planar substratum, which might explain in part why reported values of cell stiffness are much higher in 3D than in 2D [8, 9]. A wealth of techniques exist for quantifying cell tractions in 2D culture, but interpretation of data is confounded by the way that cell architecture changes when cells are plated on a 2D substrate [10, 11]. Three-dimensional methods are available as well [12] (also cf. Chap. 1 of this text), but these continue to face a difficult trade-off: ECM materials that are straightforward to characterize mechanically often bear little resemblance to the fibrous ECM with which cells interact in vivo. While methods exist for tracking strain fields and ECM fibers with great detail [13], and of relating these to local predictions of the mechanical environment around individual living cells [14–16], our focus here is the quantification of mechanical properties of ECM surrounding populations of interacting cells, and of the mechanical properties of the cells themselves. Bio-artificial tissue constructs mimic the 3D environment of a natural tissue (e.g. [7, 17, 18]) and capture the strong interactions between cells and ECM components. While studying cell mechanics in cell-cell interactions using these constructs overcomes many of the aforementioned problems, it introduces a host of new ones. After a brief description of the tissue constructs of interest, this chapter presents an overview of models designed to overcome the challenge of interpreting

5 Cell-Cell Interactions and the Mechanics of Cells and Tissues

77

cellular mechanics from the overall mechanics of a tissue construct, then presents three examples of mechanical cell-cell interactions that can be observed using these constructs.

5.1.1

Quantitative 3D Culture Systems

The system analyzed in this chapter involves ring-shaped collagen gels with prescribed populations of interacting cells (Fig. 5.1). The synthesis of two systems of interest are summarized here: fibroblast populated matrices (FPMs) and engineered heart tissues (EHTs). FPMs are synthesized by combining prescribed numbers of contractile fibroblasts, extracted from rat or 10-day-old chick embryos, with neutralized type I rat tail collagen and Dulbecco’s modified Eagle’s medium (DMEM) at prescribed concentrations [19]. Typically 1 ml of this cell suspension is poured into an annular Teflon mold (9.5 mm inner and 14.9 mm outer diameter), and, after 3 days of culture in DMEM enriched with 10% fetal calf serum (FCS), ring-shaped tissue construct specimens compressed to a measured fraction of their original cross-sectional area [19, 20] are obtained. These FPMs have sufficient mechanical integrity to be stretched on a custom built testing apparatus (Fig. 5.1). FPMs and analogous 3D culture systems have a rich history as qualitative tools for studying cell-cell and cell-ECM interactions, with much early work focused on the rate and extent of the contraction of polymerized collagen gel discs by

Control and data acquisition computer

Micrometer drive

Stepper motor

Force transducer

Tissue construct Organ bath

Fig. 5.1 Tissue constructs made of collagen and defined populations of cells serve as testbeds for quantifying the mechanics of cells and cell-cell interactions

78

G.M. Genin et al.

fibroblasts [17, 21, 22]. Methods to measure isometric contractile forces exerted by tissues have existed for some time [18, 23–27], as have methods for measuring culture-dependent stiffening and breaking strength [28] and creep response [29]. However, each of these is challenging to analyze quantitatively in terms of cellular mechanical properties, motivating the above approach. Central to the strategy for dissecting ECM and cellular contributions to the measured mechanical properties of a tissue construct is the ability to selectively activate, deactivate, or remove cells and cellular components through the application of drugs, detergents, and biochemical inhibitors. A typical assay involves two sets of experiments: a relaxation test performed on a preconditioned tissue construct, and a subsequent relaxation test performed after treatment of the tissue constructs by deoxycholate to dissolve away nearly all cytoskeletal components in the cells, leaving behind an ECM containing fluid-filled voids (“deoxycholatetreated construct”). Analysis of such experiments is a focus of this chapter. Cardiomyocyte constructs have been formed by embedding cells in matrices composed either of natural ECM constituents, such as collagen, or in a synthetic polymer meshwork [30–35]. Only the former approach is useful for delineating cell contribution from overall response because of the stiffness mismatch between cells and synthetic polymer matrices. Here, chick embryo cardiomyocytes and fibroblasts are both embedded in a collagen ECM, the former forming gap and adherens junctions to create a mechanically and electrically connected network that twitches synchronously and rhythmically. While EHTs are not yet suitable as grafts for clinical application, the cells are suitable models for heart cells [36], and the EHTs themselves retain many structural and functional properties of differentiated myocardium [37]. EHTs show “intensively interconnected, longitudinally oriented cardiac muscle bundles” that resemble adult myocardium, display a high twitch/baseline force ratio (>1), and exhibit a strong inotropic response to b-adrenergic stimulation [34]. Rat and chick EHTs are largely similar but differ in twitch rate, sensitivity to external calcium ion concentration and dependence of twitch amplitude on frequency. Oscillatory stretch of either induces a hypertrophic response, increasing myocyte size and longitudinal orientation, the density and length of myofilaments, and the RNA/DNA and protein/cell ratios. Furthermore, stretching increases contractile force by two- to four-fold [38] and also increases the twitch force, analogous to the Frank-Starling law of the heart [7]. We will describe below how these constructs can model cell-cell interactions in models of fibrosis.

5.2

Quantifying Cell Mechanics: Mechanical Models of Tissue Constructs

To interpret averaged cell mechanics from the responses of tissue constructs, integrated experiment and analysis is required. This section describes models designed specifically to interpret the mechanics of FPM rings (Fig. 5.1). From

5 Cell-Cell Interactions and the Mechanics of Cells and Tissues

79

the experimental standpoint, measurements involve recording the mechanical response of a preconditioned FPM to prescribed stretches, and then recording the response of the same FPM following treatment with a drug, detergent, or inhibitor. If the volume fraction of cells and their approximate dimensions and orientation distribution are known, and if a detergent such as deoxycholate is used to wash away the cells from the FPM prior to the second test, the effective mechanical properties of the ECM and cells can be backed out from the results of these tests (e.g. Fig. 5.7). The analytical methods appropriate for this approach incorporate and extend three classes of models from the mechanics of composite materials and paper. The first is homogenization methods, based largely upon the Eshelby solution [39, 40], including self-consistent (mean field) approaches [41, 42], in which elastic fields within each constituent are replaced by average values. These approaches and their extensions [43–49] relate cell shape, cell and ECM mechanical properties, and, to some degree, the statistics of the cell distribution to the overall mechanics of a FPM. A second class of methods is statistical approaches like the Zahalak model [50] and its extensions [8, 9, 51]. These relate the active and passive mechanical responses of cells and ECM to the overall mechanical responses of tissue constructs through consideration of the statistics of cell morphology and distribution. Related methods trace back to Flory and have enjoyed broad usage for incorporating collagen fiber geometry into macroscale constitutive models of soft tissues [52–57]. The third class is involves unit cell or periodic microfield models based upon numerical simulation of repeating microstructures [58, 59]. These have been used widely to study the mechanical environment of chondrocytes [60, 61]. The methods described below involve a suite of models that incorporate all three approaches.

5.2.1

The Zahalak Model and its Modifications

The first class of models to be discussed is the Zahalak model [50] and its extensions, which describe the active and passive mechanics of a tissue construct containing a single type of cell. We then describe briefly the extensions of Marquez et al. to this modeling framework and proceed to describe several classes of cell-cell interactions that are evident from application of this model to FPMs. Zahalak considered the mechanical response of a mechanically loaded tissue construct over timescales that are rapid compared to those associated with tissue remodeling (see, e.g. [12, 62–66]) and cellular remodeling [67, 68]. Cells were idealized as thin contractile rods, as is appropriate for spindle-shaped cells in tissue constructs [19]. The model is based upon a statistical averaging of the passive and active contributions of the cells and ECM to overall tissue construct mechanics, and, in its earliest form, assumed these contributions to be

80

G.M. Genin et al.

in parallel, such that the stress at any point is the sum of matrix (m) and cell (c) contributions: sij ¼ ð1 fc Þsij ðmÞ þ fc sij ðcÞ ;

(5.1)

where sij(m) and sij(c) are second Piola Kirchhoff stresses weighted by cell and ECM volume fractions fc and fm ¼ 1 fc. Equation (5.1) is an obvious oversimplification, but suffices to convey the essence of the model and will be corrected easily below. Zahalak also assumed that deformations are sufficiently characterized by the small-strain tensor: eij ¼ 1/2(ui,j + uj,i), where ui is the macroscopic continuum displacement of the tissue and commas indicate partial differentiation. This could also be corrected, but doing so poses an unnecessary complication for the small strains (<10%) of interest. For the cells, Zahalak applied the Hill model [69] for skeletal muscle (lower grouping of mechanical elements in the 1D idealization shown in Fig. 5.2), including three elements: (a) a parallel elastic component accounting for static passive mechanical properties, (b) a series elastic element to account for increased dynamic stiffness, and (c) a damped contractile element which embodies the active force generating and quasi-viscous force-velocity response of the cell. Zahalak adopted a linearized version of the Hill model, which yields the following equation governing the mechanical behavior of a cell: dF F Fo 1 dl lo kp l þ ¼ þ þ lo ðks þ kp Þ 1 ; tc lo dt tc tc lo dt

(5.2)

where F is the total force a cell exerts on the ECM; l and lo are the current and initial cell length, respectively; ks and kp are the series and parallel elastic stiffness values of the cell, respectively; and tc ¼ c/ks is the cell time constant in which c is the cell’s effective viscosity. The contractile force, F0, is generated by the cell and can increase in response to activators such as fetal bovine serum (FBS) or thrombin. To generalize to three dimensions, Zahalak treated cells as slender contractile rods producing a combination of active and passive forces along their lengths, with each cell directed along a unit vector ni. These unit vectors followed a prescribed or measured orientation distribution p(ni), such that the fraction of the total cell Rpopulation with orientations within the solid angle dO is given by p(ni)dO, and p(ni) dO(ni) ¼ 1. Homogenizing over the orientation distribution [70]:

Fig. 5.2 A one-dimensional embodiment of the Zahalak model

5 Cell-Cell Interactions and the Mechanics of Cells and Tissues ðcÞ sij

81

Z

¼ Nl0 Fni nj ¼ Nl0

O

Fðnp Þni nj pðnp ÞdOðnp Þ;

(5.3)

where NRis the number of cells per unit volume; lo is the initial cell length; and <*> ¼ (*)p(np)dO(np) is the orientation averaging operator. Note that the term Nlo arises from the assumption that cells and ECM deform in parallel mechanically: the cross-sectional area of a fictional tissue construct in which all cells were aligned end-to-end is 1/Nlo. Equation (5.2) then takes the form: ðcÞ

ðcÞ

@spq spq þ ¼ @t tc

3 @ o s0 Apq þ k þ Bpqij eij ; tc @t tc

(5.4)

where repeated indices imply summation, and: 1 s0 ¼ Nl0 F0 ; 3

k ¼ Nl20 ðks þ kp Þ;

o ¼ Nl20 kp :

(5.5)

The two “cell anisotropy tensors” translate the anisotropy of the microscopic cell distribution into anisotropic mechanical behavior of the macroscopic tissue: Aij ¼ ni nj ¼ Bijpq ¼ ni nj np nq ¼

Z Z

ni nj pðnk ÞdOðnk Þ; (5.6) ni nj np nq pðnk ÞdOðnk Þ

Zahalak assumed the contractile force F0 is independent of cell orientation; s0 represents the intensity of contractile stress. The parameter o represents the macroscopic manifestation of cellular parallel elasticity (“slow” elasticity), and k the combination of parallel and series elasticity (“fast” elasticity) (Fig. 5.2). For the case of a Fung QLV solid element for the matrix, combining the above yields the following integral form of the constitutive law: ðmÞ

sij ¼ ð1 fc Þsij Zt þ fc 1

t t0 3 @ o exp s0 Aij þ k 0 þ ðBijpq epq Þ dt0 ; tc tc @t tc

(5.7)

where:

ðmÞ sij ðtÞ

Zt ¼ pdij þ 2 1

ðeÞ

mðt t0 Þ

@sij ðeij ðt0 ÞÞ 0 dt ; @t0

(5.8)

82

G.M. Genin et al.

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 0 i ðeÞ e ðt Þ in which sij ðeij ðt 0 ÞÞ ¼ eij ðt 0 Þ eeffe0ðt0 Þ exp effe0 1 , eeff ¼ 23eij eij , and eo is a fitting parameter. Applying this model to interpret relaxation experiments, Zahalak et al. [50] estimated the active component of cell stress as so 0.4 kPa, which is on the same order as more recent estimates using traction force microscopy. The adaptation of the Zahalak model to situations in which cells and ECM do not stretch in parallel involves correcting the average strain fields experienced by cells and ECM. This correction captures the mechanical interaction between neighboring cells. In a case in which cells are sparse and are much stiffer than the ECM, cells experience a strain that differs from that of the ECM, and (5.1) applies only in the volume averaged sense. The relationship between the macroscopic strain applied to a tissue (resolved in a cell’s axial direction), and the axial strain experienced by the cell is approximately constant for all cell orientations, and is surprisingly insensitive to cell anisotropy [71]; we call this ratio the “strain factor,” S. The differential equation (5.4) governing cell force is then:

h

ðcÞ @spq 1 3 @ o Bpqij Se1 þ sðcÞ ¼ s A þ k þ o pq pq ij ; @t tc tc @t tc

(5.9)

and the governing equation for the tissue construct (5.7) becomes: 2 sij ¼ð1 fc Þ4pdij þ Zt þ fc 1

Zt 1

! 3 ðeÞ @sij ðeij ðt0 ÞÞ 0 2mðt t ÞMðt Þ dt5 @t0 0

0

t t0 3 @ o Bijpq Sðt0 Þepq dt0 exp so Aij þ k 0 þ tc @t tc tc

(5.10)

ðmÞ

where MðtÞ ¼ eij ðtÞe1 ij ðtÞ ¼ 1 þ ð1 SðtÞÞfc =ð1 fc Þ corrects the average strain field in the extracellular matrix, fc ¼ NloAc is the volume fraction of cells, and Ac is the average cell cross-sectional area. From (5.10), calculation of the instantaneous and relaxed effective moduli of cells and ECM is straightforward: the problem reduces to linear elasticity, with the caveat that the moduli must be understood as tangent moduli at a particular strain level [72]. For this case, and limiting to the case of isotropic cells and ECM, the strain factor for a thin tissue construct (thickness on the order of cell length) can be written in closed form [8] using a self-consistent-type approach [39, 41, 42, 73]:

2 1 þ 3ðt=lo Þ þ 2 t=lo 1 1 S¼ 1 þ ðt=lo Þ þ 2ðt=lo ÞðEc =Em Þ 1 þ 2ðt=lo ÞðEc =Em Þ 1 þ KY

(5.11)

where Ec is the elastic modulus of the cells, Em is the corresponding average tissue modulus, t is the (elliptical) cell width, lo is the (elliptical) cell length, Y ¼ tEc/loEm,

5 Cell-Cell Interactions and the Mechanics of Cells and Tissues

83

and K is a constant. In general the ratio Ec/Em can be expected to be different for the instantaneous and fully relaxed responses of the cell and ECM, and S would then vary as a function of time. Determining K requires an additional set of simulations involving study of idealized, periodic microstructures of a tissue construct. Such simulations show that K 2.2 for “rectangular,” planar cells; since this is very close to the value of 2 for very thin elliptical cells, we can conclude that the solution is not highly sensitive to the details of cell shape. A 3D generalization of this is also available [8, 9].

5.2.2

Cells Interact to Affect the Mechanical Environments of Their Neighbors at High Cell Concentrations

As cells encroach upon their neighbors, several types of collective action are observed in tissue constructs. The first is active: cells make contact through and link through adherens and gap junctions, leading in the case of tissue constructs containing cardiomyocytes to coordinated contractions in the tissue construct. The second is a passive effect due to mechanical interactions between cells [74]. We summarize these effects in this section by considering behavior of idealized, passive, cylindrical “cells” near the percolation threshold [75]. Simulations of such ellipsoids in a uniform isotropic matrix show a sharp rise in the strain factor at a concentration that depends strongly on the cell geometry but only weakly on the cell mechanical properties (data, Fig. 5.3). We describe here 0

10

Y *= 0.5 −1

Strain factor, S

10

Y *= 3

−2

10

Y *=30

−3

10

Y *= 300 Model

−4

10

Y *= 3000

Simulation

−5

10

−2

10

−1

10

10

0

10

1

10

2

Dimensionless cell concentration, C

Fig. 5.3 Stiff cells stretch much less than the surrounding ECM. Beyond the percolation threshold (here C 4), cells and ECM begin to stretch in registry (S 1)

84

G.M. Genin et al.

models for cells in tissue constructs that are sufficiently thin that they can be idealized with planar models. Since results presented later indicate that the effective cylinder that would represent a fibroblast cell has under most testing conditions a modulus greater than that of the ECM, we further specialize here to that case. Models exist for cells more compliant than the ECM and for thicker tissue constructs [8, 9]. At low cell concentrations, cells behave like isolated particles. In the case of very stiff particles, the axial strain experienced by the cells is small compared to the axially resolved strain experienced by the tissue construct. At the percolation threshold, the strain factor rises close to a value of 1, and the cells and ECM deform nearly in register. The model of (5.11) predicts this rise with two modifications [8, 9]. First, the modulus ratio must be replaced by a “self-consistent” estimate [41, 42] that accounts for the stiffening of the environment directly surrounding cells by the presence of neighboring cells. For cells in a thin tissue construct following a uniform, planar orientation distribution, Marquez et al. derived [8]: 3 Eeff m ¼ Em þ 8 SðEc Em ÞCt=lo ;

(5.12)

where for a thin tissue construct the dimensionless cell concentration is C ¼ Ac Nl2o =t ¼ fc lo =t. Note that for a thicker tissue construct with thickness much greater than the dimensions of cells, the definition C3D ¼ Nlo3 is the appropriate nondimensionalization, and a different expression is required for the effective modulus [9]. Second, the effective length of cells must be scaled to account for physical linkages in the case of overlap. In dense populations of straight, overlapping cells, the number of nearby cells that a cell can expect to intersect is C/p [8, 76]. Adjusting the effective cell length accordingly: leff ¼ lo ð1 þ C=2pÞ:

(5.13)

Expressions exist for cells in thicker tissue constructs with a variety of orientation distributions; for a thicker

construct with a 3D uniform orientation distribution, the expression is l ¼ l 1 þ 2dC pl = eff o 3D o [9]. Defining an effective modulus ratio and using this in place of Y in the scaling law (5.11), a very good Y ¼ tEc leff Eeff m estimate of the strain factor is obtained (Fig. 5.3, K ¼ 2.2); the model predicts the results of finite element analyses on representative periodic random cell distributions. Of particular interest in Fig. 5.3 is the sudden rise in the strain factor near C ¼ 4 (note that in a thicker tissue construct with a uniform 3D fiber orientation distribution, this corresponds to C3D 11). This is the concentration at which statistics of fiber networks predicts that a contiguous network of cells will form: in both the 2D and 3D cases, this percolation threshold occurs just above the value of C at which cells can be expected to intersect with a single neighbor. Zahalak’s assumption of S ¼ 1 is reasonable for very high cell concentrations; however, it is a poor approximation for lower cell concentrations, especially in tissues with a large stiffness mismatch between cells and matrix (high Y*).

5 Cell-Cell Interactions and the Mechanics of Cells and Tissues

85

Estimating the cell modulus is possible from the definition of Y* in (5.7) if the matrix properties and cell dimensions are characterized. The following equation relates the measurable quantity o (Fig. 5.2) to SY* [50]:

2 SY ¼ oN 2=3 E m lo : (5.14) The relationship between Y* and SY* is found numerically from (5.11) as a function of dimensionless cell concentration C, and, given the modulus of the matrix and a description of the cell morphology and arrangement, the effective cell modulus Ec can be calculated from the definition of Y*.

5.2.2.1

Homogenization Methods

An additional strategy for dissecting the mechanical responses of cells and ECM involves direct application of classical homogenization methods. Two mechanics challenges must then be addressed: (1) subtracting the effects of fluid-filled voids of negligible stiffness from the overall mechanical response of deoxycholate-treated construct and (2) determining the cell moduli from those of the intact tissue construct. Marquez et al. [51] performed Monte Carlo simulations on idealizations of tissue constructs with isotropic, nearly incompressible constituents (Poisson’s ratio n ¼ 0.49) to establish both of these sets of relations for thicker tissue constructs containing spindle-shaped cells with a nominally uniform orientation distribution. The relationship between the observed elastic modulus of deoxycholate-treated constructs, Ep, varied with respect to void (cell) volume fraction, fc, in a way that followed the model of Pabst and Gregorov [77] for porous ceramics: Ep ð fc Þ=Em ¼ ð1 fc Þð1 fc =fo Þ;

(5.15)

where fo ¼ 0.872 can be interpreted as the pore volume fraction beyond which the ECM is no longer continuous; the fitting (dashed line in Fig. 5.4) is excellent, with R2 ¼ 0.998. For the second relationship, Marquez et al. found that a generalization of the model of Pabst et al. was accurate for all tissue constructs in which cells are more compliant than ECM: E t ð fc Þ Ec 1 þ aEc =Em Ep ð fc Þ ¼1þ 1 1 ; (5.16) Em Em 1 þ bEc =Em Em where Ep(fc)/Em is given by (5.15), and a ¼ 11.57 and b ¼ 15.00 are dimensionless scaling parameters obtained by least squares fitting (Fig. 5.4). For cases of cells stiffer than the ECM, homogenization bounds are very broad, and no adequate approximation exists. However, a line midway between the three point bounds for overlapping spherical cells [48] provides an adequate approximation (see [51] for these formulae); the worst case fit with Monte Carlo data was for cells much stiffer than the ECM (R2 ¼ 0.85 for Ec/Em ¼ 1,000).

86

G.M. Genin et al. 1.0

Modulus ratio, Et / Em

Ec / Em = 0.50 0.8

0.6

Ec / Em = 0 (voids) 0.4

Ec / Em = 0.05 Ec / Em = 0.10

0.2

Ec / Em = 0.25

Eq. 16 Simulation

0.0 0.0

0.1

0.2

0.3

0.4

0.5

Cell volume fraction, fc

Fig. 5.4 The effective modulus Em of the ECM in a tissue construct containing ECM and voids can be estimated from the measured modulus Et of the tissue construct. Knowing Em and the modulus of a tissue construct containing cells, the effective modulus of the cells can be estimated using (5.16) (cases in which Ec Em) or analogous but lengthier relations (Ec > Em) [51]

5.3

Examples of Cell-Cell Interactions in Tissue Constructs

We present here three examples of cell-cell interaction in tissue constructs. The first shows ways that contractile fibroblast cells modify their own mechanics and that of their ECM as a function of their proximity to one another, and highlights that the percolation threshold appears to be a state, actively sought by myofibroblasts, at which myofibroblasts are able to match the effective modulus of the ECM to their own modulus. The second exploits the observation that at cell concentrations above the percolation threshold, the mechanical contributions of cells are in parallel with those of the ECM, allowing protein-level observations to be made about how cells transmit force to their neighbors and to the ECM. The third is an ongoing area of research to study how different types of cells interact in tissue constructs that serve as models of cardiac fibrosis.

5.3.1

Mechanics of Fibroblast Cells

Contractile fibroblast cells remodel the tissue construct in which they are embedded and proliferate or die off depending on culture conditions [78]. In the process, they can change the stiffness of the ECM and affect their volume fraction. The interdependence of cell concentration, ECM remodeling, and mechanics is important to the study of wound healing [79]. Using the methods described above, Marquez et al. [51] asked

5 Cell-Cell Interactions and the Mechanics of Cells and Tissues

87

Final cell concentration, N (millions of cells/ml)

the questions, “given their choice, what mechanical environment will contractile fibroblast cells create for themselves?” and “how will this affect their proliferation and mechanics?” The results summarized here suggest that cell-cell interactions are important determinants of both of these questions. The experiments of Marquez et al. [78] involve measurement of the mechanical relaxation of a tissue construct following rapid increments of stretching. The experiments are as described above: tissue constructs containing activated cells were compared to tissue constructs containing the voids remaining after cells were eliminated with deoxycholate. They are coupled with measurements of the volume fraction of cells both at the beginning of tissue culture and at the end of the 3-day interval after which mechanical testing occurred. These experiments provide many clues about the ways that cells interact mechanically in a tissue model. We will emphasize four aspects of the results here. The first is the way that cells remodel the tissue constructs (Fig. 5.5). As others have observed (e.g. [81]), remodeling is dependent upon culture conditions. The degree to which cells reduce the volume of a tissue construct during remodeling increases with increasing number of cells at the start of the 3 day culture period for the tissue construct. However, cells in constructs that contain initially the lowest concentrations of cells divide and multiply very quickly, while those in the FPMs that contain initially the highest concentrations of cells decrease in number.

10

3

number of cells increases during culture 10

number of cells decreases during culture

2

10

1

percolation threshold 10

0

−1

10

0

1

2

3

4

Initial cell concentration, N (millions of cells/ml)

Fig. 5.5 The cells in a tissue construct remodel the matrix and either proliferate or die off over the 3 day course of incubation. The cell concentration increases from the initial cell concentration in all cases. In the darker region, cell proliferation contributes to the increase in cell concentration during incubation of the tissue constructs. In the lighter region, compaction of tissue constructs leads to an increase in cell concentration during incubation, but the total number of cells within the tissue constructs decreases. In both cases, the number of cells decreases or increases during incubation so that the final cell concentration approaches the percolation threshold

88

G.M. Genin et al.

Between these is a concentration of cells at which the number of cells does not change over the 3 day course of tissue culture: tissue constructs that begin with 1 million cells in 1 ml of collagen solution end with this same number of cells. The final tissue construct shrinks over 3 days of incubation so that the cell concentration, N, is approximately 11 million cells per ml of tissue construct. Since the mean cell length in this system is 100 15 mm (mean SD) and cell diameter is 10 1 mm, the dimensionless cell concentration C3D is approximately 11 in this case; this is the percolation threshold. Taken as a whole, these results suggest that the cells in these tissue constructs may die off or proliferate to reach the percolation threshold. The second feature of interest is the force relaxation. Specimens were preconditioned with a stretch of 10% over 150 s, returned to their reference length at this same rate, then allowed to relax for 3,600 s. The specimens were then stretched rapidly (over 10–15 ms) to prescribed nominal strains. Isometric force was monitored over a 3,600 s relaxation period. FPMs were stretched to nominal strains of e ¼ {0.02, 0.08, 0.14, and 0.2}; the specimens were not unloaded between stretches. The peak isometric force in tissue constructs both with cells and with cells removed rose to a maximum then decayed over time. In both cases, the force relaxation was logarithmic with respect to time, as would be expected for a material that expresses a continuous spectrum of relaxation times [78]: Fðt; e; fc Þ ðao ð fc Þ þ a1 ð fc ÞeÞ þ b1 ð fc Þe lnðt=to Þ;

(5.17)

where a1, a2, and b1 are functions of the final cell volume fraction fc only, and to is an arbitrary normalization constant. The contribution of active cellular contraction, represented by ao, is largely independent of linearized strain e over the range tested, and increased exponentially with increasing final cell concentration. For a tissue construct with C3D ¼ 11.8 (i.e., near the percolation threshold), a value of ao ¼ 2.2 mN was extracted, corresponding to an active stress of approximately so ¼ 410 Pa in the construct, independent of mechanical strain. An additional aspect of note is that cell-cell interactions play a very strong role: Marquez et al. [78] showed that the ability of cells to transmit active, isometric stress to the boundaries of a tissue construct dropped sharply at cell concentrations below the percolation threshold. The third feature is the variation of the passive, incremental tangent moduli of cells and ECM with respect to cell concentration (Fig. 5.6). The stiffness of the ECM increased monotonically with increasing cell concentration [51]. Note that the stiffness was calculated from the second two terms in (5.17), an approximation that is most valid when ao is small. The collective action of cells to achieve this involved more than simply effects of cells squeezing water from the ECM (syneresis): the modulus of the ECM increased well beyond what could be accounted for simply through compression of the collagen network, suggesting that other factors such as cross-linking of collagen and collagen production by cells might play a role. The collective remodeling action of cells manifests itself in ECM stiffness that increases with increasing cell concentration. The effective modulus of cells, however, decreases with increasing cell concentration. At the lowest cell concentrations

Fig. 5.6 The modulus of ECM in a tissue construct increases with increasing cell concentration, and the modulus of the cells decreases. The crossover occurs at a cell concentration that corresponds to the percolation threshold

Long-term elastic modulus (Pa)

5 Cell-Cell Interactions and the Mechanics of Cells and Tissues

89

106

104

ECM data Cell data ECM model Cell model

102 10−1 100 101 102 103 Cell concentration, N (millions of cells per ml)

considered, the model of a cell as a contractile rod in a homogeneous ECM is likely invalid, as we and others have observed pockets of enhanced remodeling around cells in sparsely populated tissue constructs; indeed, for the very lowest concentrations in Fig. 5.6, a contractile rod that would mimic the cellular contribution to tissue construct mechanics would require infinite stiffness. As the cell concentration increases, the reliability of this model increases because the ECM becomes more spatially homogenous though the collective action of cells. At the highest cell concentrations considered, the model shows cells are more compliant than the ECM. The final feature of these data to note is the crossover point at which cell and ECM moduli are equal. At this point, the cells match the modulus of the ECM to their own modulus, which has implications in terms of reducing stress and strain concentrations in response to mechanical loading. The cell concentration at which this occurs corresponds to the percolation threshold, suggesting that this matching of modulus is a result of collective action by cells. The magnitude of modulus at which this occurs is also of interest: for the long-term modulus (tangent modulus 3,600 s after stretch) of the tissue construct is just over 10 kPa, which is the substrate modulus observed by Engler et al. [80] to steer mesenchymal stem cells to differentiate into a myofibroblastic cells. The crossover modulus shortly after stretch is approximately four times greater [51]. These data together hint that contractile fibroblasts in tissue constructs might modify themselves and their ECM to approach the state in which they exist in a network that is just at the percolation threshold, and in which cell and ECM modulus match. Such a picture of cell proliferation relating to mechanical environment is consistent with the work of Zhu et al. [81], who observed differences in the fate of cells within tissue constructs as a function of mechanical constraints during incubation. Myofibroblasts in tissue constructs proliferate or die off to approach this critical volume fraction. This is not unexpected in the context of wound healing, in which formation of a contiguous myofibroblast network is desirable. The fact that the cellular and ECM moduli match at this cell concentration has mechanical

90

G.M. Genin et al.

implications as well: stress concentrations at cell/ECM attachments should be smallest when the cell and ECM moduli are matched. However, in the experiments summarized here, as in those of Zhu et al., the mechanics are certainly not the only aspect of the experiment that could drive cell concentration dependent effects on mechanics and proliferation. Inter-related features such as nutrient transport, steric effects of ECM, and the concentration of ECM are also factors. Models for flow through fibrous networks are increasingly mature (e.g. [82]), and these offer hope for dissecting these effects through coupled modeling and experimentation.

5.3.2

Cytoskeletal Mechanics Observed Through Testing of Tissue Constructs

The previous example focused on analysis of force relaxation data taken either immediately after application of a step stretch to a tissue construct, or after the tissue construct has had 3,600 s to relax. From the mechanical response of the tissue construct in the intervening time interval and associated imaging of the actin cytoskeleton, additional information can be obtained about the way that individual cells bear and react to mechanical stretches. The data discussed here are those of Nekouzdeh et al. [83], who tested specimens like those described above, with final cell concentrations just above the percolation threshold (C3D 12). This allowed a straightforward delineation of the contributions of cells to the measured mechanical response of a tissue construct, because the contributions of cells act in parallel to those of the ECM in these cases (the strain factor S 1). In addition to obtaining the mechanical relaxation data described above, specimens were fixed and stained with rhodamine phalloidin to image the actin cytoskeleton at prescribed times after application of a rapid mechanical stretch of 20% to a preconditioned tissue construct. When an isometric force relaxation curve for the ECM in a tissue construct (here measured by treating cells with Triton X100 rather than deoxycholate) is subtracted from that of the overall tissue construct, the result is a force relaxation curve that scales with the mechanical response of the cells (Fig. 5.7). The cytoskeletal basis for these changes in the mechanical response of the cell can be understood by quantitative comparison to the preponderance of stress fibers in images of the tissue constructs (Fig. 5.8). The technique used (Fourier-space automated band-pass length estimation, or “FABLE” [83]) involves integrating over a band-pass-filtered power spectral density of an image to estimate the total length of features of a particular range of widths; images are pre-processed using median filtering to reduce artifacts. Results indicate that the total contribution of cells to tissue construct mechanics scales linearly with the total length of stress fibers observed in images of phalloidin-stained tissue constructs: the curve superimposed on data in Fig. 5.8 is renormalized data from Fig. 5.7. Several aspects of the mechanics of myofibroblasts are evident from these data. The first is that stress fibers are a candidate as the dominant determinants of the

5 Cell-Cell Interactions and the Mechanics of Cells and Tissues

91

Deoxycholate-treated construct

Force (mg)

Force (mg)

Tissue construct

Time (s)

Time (s)

Cellular contribution

B

F E

A

D C

Time (s)

Fig. 5.7 Mechanical responses of a FPM and its constituents to a rapidly applied stretch. The subsequent isometric mechanical response of the tissue construct as a whole shows a characteristic rise followed by a drop that is too fast to be explained by viscoelastic relaxation. This is followed by a rise in force due to active cellular contraction. The cellular contribution to this mechanical response can be deduced by subtracting the mechanical response of the ECM (“deoxycholatetreated construct”). The cellular contribution shows a pre-stretch baseline force (A), a rise following stretch (B), a rapid drop shown below to result from cytoskeletal depolymerization (C), a rapid active response (D), a plateau (E), and a gradual active response (F). The final isometric force exceeds the pre-stretch value. This behavior is repeated upon subsequent loading cycles

force that cells exert on their environment. The linear scaling between the total length of stress fibers in images of stained tissue constructs and the force exerted by cells suggests that the contributions of other sub-cellular protein structures is very small. The contributions of protein structures such as microtubules either scales linearly with that of the actin cytoskeleton, or is far smaller in 3D tissue constructs than in 2D culture systems [84]. Immediately following the stretch, a rapid drop in isometric force occurs as approximately half of the stress fibers disappear. This drop is faster than can be accounted for through the viscoelasticity of cells or ECM, as the time constants associated with it are too fast [83, 85–87]. The isometric force rises from this drop in two steps, a rapid active response that occurs over the course of tens of seconds, and a gradual active response that occurs over tens of minutes. In both cases, the cell force scales linearly with the total length of stress fibers. The end point is a stress level that exceeds the pre-stretch value, but is lower than the immediate poststretch value; the actin stress fiber network at the end of the gradual active response is indistinguishable from that which exists pre-stretch. This suggests that some but

G.M. Genin et al.

Polymerized actin fraction

92

1.0 0.8 0.6 0.4 Normalized cell contribution

0.2 0

500 Time (s)

1000

Fig. 5.8 Breakdown and remodeling of the actin cytoskeleton in response to rapid external stressing quantified by the density of F-actin at four time-points: (1) After preconditioning but before the application of a rapid stretch; (2) at the end of the rapid reduction of force following a step stretch; (3) at the end of the rapid active response; and (4) at the end of the gradual active response. The trend-line is the cellular response curve from Fig. 5.7 scaled to match the density at the end of the gradual active response. Error bars represent the range of data. The density of actin filaments scaled with the active cell force

not all of the passive resistance of cells due to stretch is eliminated by active cellular remodeling. The active responses are repeatable, suggesting that cells return to their original, pre-stretching state after release of the mechanical stretch. Why do cells exhibit these responses? The Fredberg and Navajas groups observed that, following mechanical stretch in 2D culture, single airway smooth muscle cells exhibit a behavior analogous to the rapid drop in force observed by Nekouzadeh et al. that they term a “fluidization” [88, 89]. Fluidization, Fredberg argues, would play the specific role of relieving the mechanical constriction of the airway; the absence or restriction of this post-stretch fluidization might be a cause of pathologies such as asthma. Analogous behavior occurs in myofibroblasts and smooth muscle cells. However, open questions persist regarding the mechanisms by which the breakdown and recovery occur, by which the associated mechanical forces are transmitted from cell to cell, and about the nature of these phenomena in living tissues.

5.3.3

Cell-Cell Interactions in Cardiac Fibrosis

Fibrotic cardiomyopathy can result from hypertension or coronary infarction [90]. Although cardiomyocytes occupy most of the tissue volume of the myocardium,

5 Cell-Cell Interactions and the Mechanics of Cells and Tissues

93

they comprise only one third of the cell number. The balance includes fibroblasts, vascular smooth muscle cells, and endothelial cells [91, 92]. Fibroblasts are the most numerous, outnumbering cardiomyocytes by up to 2:1, and play an important role in the maintenance of the physiologic structure of the heart [93–96]. Fibroblasts are small and mechanically quiescent in the normal myocardium, but convert to the larger, contractile myofibroblast phenotype under the influence of mechanical stress and agonists such as TGF-b [97, 98]. The conversion of fibroblasts to myofibroblasts is central to the development of cardiac fibrosis in response to hypertension [93]; in ventricular hypertrophy, hypertension, or infarction, the number of fibroblasts in the heart has been shown to increase [93]. A scar tissue formed after myocardial infarction also is stiffened by both reparative fibrosis and actively contracting cardiac fibroblasts. These myofibroblasts influence myocardial function by increased collagen secretion and contractility causing a stiffening of the heart muscle that can lead to diastolic dysfunction and heart failure [99] and also by possibly interfering with the electrical connectivity of the cardiomyocytes [93]. In this final example we describe how tissue constructs serve as model systems in which to study how fibroblasts and cardiomyocytes interact to control contractile force and tissue stiffness. While the EHTs described earlier differ substantially from myocardium, they provide a simplified system in which to dissect phenomena associated with the cellcell interactions. We begin with a description of their mechanical function and justification of their use as models of pathologically remodeled heart tissue. The mechanical response of EHTs, when observed by uniaxial or isometric stretching (loading) followed by relaxation (unloading) at a constant rate displays two distinct components: (1) a steady, contractile baseline force, analogous to diastole, and (2) a periodic twitch force superimposed on the baseline, analogous to systole. Both baseline and twitch forces increase nonlinearly when an EHT is stretched (Fig. 5.9a). The developed tension (the difference between end systole and end diastole) ranges from ~0.5 to ~1 mN/mm2 when the EHT is stretched up to 20% strain [19, 34]. This is significantly smaller than the developed tension observed in human tissue samples (~15–22 mN/mm2) [100]. The baseline tension observed here was several-fold higher than the twitch tension but is in the range observed in human myocardium (~4 to ~10 mN/mm2) [100]. Mechanical effects certainly play a role in these phenomena, since mechanical conditioning has been observed to reduce the baseline tension and improve cardiac mechanical function and structure [34]. While the twitch forces are smaller in EHTs than in normal myocardium, many important qualitative features are retained. EHTs conform to the Frank-Starling law, an important regulatory mechanism characteristic of heart muscle by which cardiac output is adjusted to demand. The pressure developed by contraction of the heart increases with increasing ventricular filling volume or applied pressure. The dependence of developed tension on strain in skeletal muscle is thought to result from stretch-dependent variation of the degree of overlap of thick (myosin) and thin (actin) filaments (number of functioning myosin cross bridges). The much steeper dependence of developed tension on change of tissue length observed in cardiac muscles [101] and in EHTs is ascribed to (1) a strain- or stress-dependent increase

94

G.M. Genin et al.

Force (mN)

18

4

16

2

449 450 451 1

10

2

Force (mN)

Force (mN)

Force (mN)

20

3

3 1 898 899 900 Time (s)

0 0

300

600

900

Developed tension (% maximum)

Time (s)

100

50

Skeletal muscle Cardiac muscle EHT

0 70

80 90 Muscle length (% length)

100

Fig. 5.9 Twitch and baseline force increase nonlinearly with the degree of stretch of an EHT (top). The Frank-Starling law in can be observed in EHTs as well as heart muscle (bottom)

of release of Ca2+ from sarcoplasmic reticulum and (2) a change in Ca2+ sensitivity of the contractile apparatus [102]. The molecular mechanisms underlying the Ca2+ sensitivity to stretch could be related to the length-dependent change in lateral spacing between thin and thick filaments [103]. EHTs display a corresponding steep slope in the variation of twitch force with baseline force (Fig. 5.9b) [7]. Following the observation that EHTs exhibit a response fundamental to heart tissue and dependent upon the combined function of both cardiomyocytes and

5 Cell-Cell Interactions and the Mechanics of Cells and Tissues

95

supporting cells, these tissue constructs have begun to be applied to study the problem of cardiac fibrosis. The data presented here involve EHTs enriched in fibroblasts to test the fibroblast contributions over time. EHTs were produced in which 25% of the cardiomyocytes were replaced with cardiac fibroblasts. The fibroblast-enriched EHTs began to twitch spontaneously and coherently much earlier (3–4 days after tissue formation) than the typical EHT without added fibroblasts (6–7 days). The twitch force developed by the EHT containing 25% fibroblasts was significantly higher than that of the fibroblast-depleted construct on the fourth day after the formation of the construct (Fig. 5.10a), whereas the baseline force was not much different. This is a reflection of the supportive contribution of fibroblasts to EHT development. On the sixth day of culture, however, the fibroblast-enriched EHT stopped twitching, yet the EHT without added fibroblasts continued to twitch and with a stronger force than that recorded on fourth day (Fig. 5.10b). The baseline force of the EHT with 25% fibroblasts was significantly higher than that of the typical construct, indicating elevated tonic tension of the EHT (Fig. 5.10b, d).

Twitch force (mN)

After 96 hrs of culture

After 144 hrs of culture

2.5 1.0

2.0 0.5 0

1

2

3

0

Time (s)

2

3

Time (s) 15

15 Baseline force (mN)

1

25% of cardiomyocytes substituted by myofibroblasts

10

10

control EHT

5

0

5

0

10 Strain (%)

20

0

0

10 Strain (%)

20

Fig. 5.10 Baseline (lower row) and twitch force (upper row) of EHTs stretched then unstretched 20%. Tissue constructs synthesized using the usual protocol (black curves) were compared to tissue constructs in which 25% of myofibroblasts were replaced with myofibroblasts prior to incubation (gray curves). After 96 h of culture (left column), an EHT initially containing extra myofibroblasts exhibited a higher twitch and baseline force. However, after 144 h, these tissue constructs ceased to beat (right column) while the control EHTs improved in function

96

G.M. Genin et al.

What leads to these effects? Cells in the heart interact through both paracrine and autocrine pathways and by direct contact with the formation of gap and adherens junctions and desmosomes. In adherens junctions cadherins on one cell bind to cadherins on another cell in contact and link intracellularly to the actin cytoskeleton via catenins. Desmosomes link to intermediate filaments. Adherens junctions and desmosomes mechanically connect cardiomyocytes and so distribute contractile force within the myocardium. Gap junctions provide intercellular channels for ionic communication that allows the rapid and coordinated spread of excitation throughout the heart. We dissect here the electrical and mechanical cell-cell phenomena that might underlie the above observations. From the mechanical perspective, a possible explanation is that the eventual domination of the construct by the proliferative myofibroblasts stiffens the tissue constructs in a way that retains the ability to produce a steady baseline force, perhaps exerted mainly by myofibroblasts, while losing the ability to generate myocyte-dependent twitch force. The stiffening of fibrotic myocardium can result from secretion of excessive ECM from the myofibroblasts [96], and likely from increased ECM remodeling and increased myofibroblast contractility as well. The stiffening of the extracellular environment by myofibroblasts and associated rise in baseline force may overwhelm the actomyosin contractile mechanism in the cardiac myofibroblasts, constraining it to a low number of cross-bridge connections by limiting the motion of the contractile apparatus. The reduction in cardiomyocyte volume fraction as myofibroblasts proliferate also plays a role. As is obvious from (5.1), the tissue-level mechanical contribution of a species of cells can be overwhelmed by a reduction in its volume fraction (e.g. [49]). A final mechanical factor to note is the cellular orientation distribution. The degree of alignment of cardiomyocytes exceeds that of myofibroblasts, and this should help emphasize the mechanical contribution of cardiomyocytes over that of myofibroblasts. While the degree to which this affects the mechanics of a fibrotic tissue or EHT is unknown, the models discussed above provide estimates. The relative reduction in the myofibroblast contribution due to the orientation distribution is 5/8 for a thin tissue construct with a planar uniform distribution of myofibroblasts (cf. (5.12)) or 4/5 for a thicker tissue construct [9]. In addition to these mechanical effects, myofibroblasts can impair both heart and EHT contractile function by distorting excitatory conduction. Under normal conditions the numerous fibroblasts in the heart maintain the ECM that provides the underlying structure for a continuous network of cardiomyocytes in electrical contact via gap junctions. The spread of electrical excitation in this network is organized to stimulate an orderly contraction first of the atria and then the ventricles to promote optimal pumping efficiency of the heart. Evidently, the presence of the fibroblasts in normal heart muscle does not perturb this orderly impulse conduction. Myofibroblasts can distort the propagation of the excitatory wave both by disrupting the normal interactions (gap junction formation) among the cardiomyocytes and by forming gap junctions and therefore electrical contact with the cardiomyocytes. However, delineating the influence of mechanical stretch vs. mechano-electrical coupling is difficult.

5 Cell-Cell Interactions and the Mechanics of Cells and Tissues

97

Myofibroblasts can be coupled electrotonically to cardiomyocytes in vitro via gap junctions mediated by the connexins Cx43 and Cx45 [94, 104]. In these experiments strands of cardiomyocytes were coated with cardiac fibroblasts that had converted to the myofibroblast phenotype. Both the conduction velocity and the maximum upstroke velocity, increased with increasing myofibroblast: cardiomyocyte ratio for low values of this ratio. At higher values of the myofibroblast: cardiomyocyte ratio conduction velocity and maximum upstroke velocity decreased with increasing ratios. This biphasic dependence is in accord with a gradual depolarization of the cardiomyocytes by the myofibroblasts. These experiments suggest that direct electrotonic coupling of myofibroblasts to cardiomyocytes could contribute to arrhythmogenesis. This coupling suggests that myofibroblasts might not only provide a barrier to the electrical interaction of cardiomyocytes but might also provide a conductive link between them. This was demonstrated in vitro by connecting two strands of neonatal rat ventricular cardiomyocytes by a band of cardiac myofibroblasts [105]. Impulse propagation initiated by an extracellular electrode was measured using a potential-sensitive fluorescent dye. The observed excitation of the strand of cardiomyocytes distal to the electrode showed that the impulse could be propagated through the band of myofibroblasts. The passage of the excitation across a 134 mm wide band of myofibroblasts, however, incurred a delay of about 30 ms and the delays increased with the width of the band until, for bands wider than 300 mm, the signal could no longer cross to the distal cardiomyocytes. These experiments demonstrate that myofibroblasts both distributed throughout the heart muscle and in border zones of healing infarcts can play a complex role in impulse propagation, imposing steric blockage, providing alternative but slower conduction pathways and predisposing the tissue to arrhythmia [96]. Finally, a model system similar to that described above has demonstrated that contact with myofibroblasts can cause spontaneous activation of cardiomyocytes that could be analogous to ectopic activity in the heart [106]. The combination of the mechanical and electrical functional degradation could be cooperatively responsible for arrhythmias and conduction problems in the heart, as could coupling between the two. Tissue constructs provide a promising platform for understanding these interactions.

5.4

Concluding Remarks

The above examples illustrate the application of tissue constructs as simplified platforms for quantifying cell-cell interactions, cell-ECM interactions, and subcellular mechanics. The approach involves tightly coupled experimentation and modeling, with selective elimination of tissue construct components through biochemical treatments serving as a bridge. The mathematical models used to interpret the experiments presented in this chapter are founded on straightforward homogenization theory for the mechanics of multiphase materials. These models are

98

G.M. Genin et al.

sufficient to quantify the mechanical roles of specific sub-cellular structures in a tissue construct, and to quantify whole cell mechanical responses. We conclude with thoughts on the limitations of these approaches and of associated needs for future work. At the whole-cell level, the models presented were adequate for tissue constructs that had a nominally homogeneous ECM, but this is not the case in tissue constructs that are not highly remodeled by cells or that are populated sparsely by cells. This was evident in the observation that even an infinitely stiff rod with the nominal dimensions of a cell was inadequate for representing the mechanical contributions of a cell. A better understanding of the ways that cells interact with ECM and other cells in sparsely populated and lightly remodeled tissue constructs is important to understanding the early stages of the wound healing process. At the level of sub-cellular mechanics, the approaches presented are limited to cellular populations near the percolation threshold because of challenges associated with predicting the viscoelastic relaxation that accompanies cytoskeletal remodeling. Dissecting the effects of viscoelastic relaxation and separating these from the mechanics of cells with dynamic actin cytoskeletons is an open challenge for cells distributed sparsely in an ECM, as is identifying the mechanisms by which the signals to depolymerize are transmitted from one cell to another. In the third example presented, multi-scale models are required that bridge cellular level mechanical and electrical cell-cell interactions with tissue level phenomena. A central question that remains open is how myofibroblasts and cardiomyocytes couple electronically in pathologic myocardium, and how these couplings affect conduction and contraction. Tissue constructs are a promising experimental test-bed for the characterization of thee phenomena. Acknowledgments This work was supported in part by the National Institutes of Health (HL079165 and AR047591), by the Center for Materials Innovation at Washington University in St. Louis, by the National Science Foundation (0538541), and by a National Science Foundation graduate fellowship to TMA.

References 1. Abbott, A., 2003, “Cell Culture: Biology’s New Dimension,” Nature, 424, pp. 870–2. 2. Cukierman, E., Pankov, R., Stevens, D. R., and Yamada, K. M., 2001, “Taking Cell-Matrix Adhesions to the Third Dimension,” Science, 294, pp. 1708–12. 3. Cukierman, E., Pankov, R., and Yamada, K. M., 2002, “Cell Interactions with ThreeDimensional Matrices,” Curr Opin Cell Biol, 14, pp. 633–9. 4. Even-Ram, S., and Yamada, K. M., 2005, “Cell Migration in 3d Matrix,” Curr Opin Cell Biol, 17, pp. 524–32. 5. Nelson, C. M., and Bissell, M. J., 2005, “Modeling Dynamic Reciprocity: Engineering Three-Dimensional Culture Models of Breast Architecture, Function, and Neoplastic Transformation,” Semin Cancer Biol, 15, pp. 342–52.

5 Cell-Cell Interactions and the Mechanics of Cells and Tissues

99

6. Stegemann, J. P., Hong, H., and Nerem, R. M., 2005, “Mechanical, Biochemical, and Extracellular Matrix Effects on Vascular Smooth Muscle Cell Phenotype,” J Appl Physiol, 98, pp. 2321–7. 7. Asnes, C. F., Marquez, J. P., Elson, E. L., and Wakatsuki, T., 2006, “Reconstitution of the Frank-Starling Mechanism in Engineered Heart Tissues,” Biophys J, 91, pp. 1800–10. 8. Marquez, J. P., Genin, G. M., Zahalak, G. I., and Elson, E. L., 2005, “Thin Bio-Artificial Tissues in Plane Stress: The Relationship Between Cell and Tissue Strain, and an Improved Constitutive Model,” Biophys J, 88, pp. 765–77. 9. Marquez, J. P., Genin, G. M., Zahalak, G. I., and Elson, E. L., 2005, “The Relationship Between Cell and Tissue Strain in Three-Dimensional Bio-Artificial Tissues,” Biophys J, 88, pp. 778–89. 10. Munevar, S., Wang, Y., and Dembo, M., 2001, “Traction Force Microscopy of Migrating Normal and H-Ras Transformed 3t3 Fibroblasts,” Biophys J, 80, pp. 1744–57. 11. Franck, C., Hong, S., Maskarinec, S. A., Tirrell, D. A., and Ravichandran, G., 2007, “ThreeDimensional Full-Field Measurements of Large Deformations in Soft Materials Using Confocal Microscopy and Digital Volume Correlation,” Exp Mech, 47, pp. 427–38. 12. Barocas, V. H., Moon, A. G., and Tranquillo, R. T., 1995, “The Fibroblast-Populated Collagen Microsphere Assay of Cell Traction Force – Part 2: Measurement of the Cell Traction Parameter,” J Biomech Eng, 117, pp. 161–70. 13. Roeder, B. A., Kokini, K., Robinson, J. P., and Voytik-Harbin, S. L., 2004, “Local, ThreeDimensional Strain Measurements Within Largely Deformed Extracellular Matrix Constructs,” J Biomech Eng, 126, pp. 699–708. 14. Evans, M. C., and Barocas, V. H., 2009, “The Modulus of Fibroblast-Populated Collagen Gels is Not Determined by Final Collagen and Cell Concentration: Experiments and an Inclusion-Based Model,” J Biomech Eng, 131, p. 101014. 15. Raghupathy, R., and Barocas, V. H., 2009, “A Closed-Form Structural Model of Planar Fibrous Tissue Mechanics,” J Biomech, 42, pp. 1424–8. 16. Sander, E. A., Stylianopoulos, T., Tranquillo, R. T., and Barocas, V. H., 2009, “Image-Based Multiscale Modeling Predicts Tissue-Level and Network-Level Fiber Reorganization in Stretched Cell-Compacted Collagen Gels,” Proc Natl Acad Sci U S A, 106, pp. 17675–80. 17. Bell, E., Ivarsson, B., and Merrill, C., 1979, “Production of a Tissue-Like Structure by Contraction of Collagen Lattices by Human Fibroblasts of Different Proliferative Potential in Vitro,” Proc Natl Acad Sci U S A, 76, pp. 1274–8. 18. Kolodney, M. S., and Wysolmerski, R. B., 1992, “Isometric Contraction by Fibroblasts and Endothelial Cells in Tissue Culture: A Quantitative Study,” J Cell Biol, 117, pp. 73–82. 19. Wakatsuki, T., Kolodney, M. S., Zahalak, G. I., and Elson, E. L., 2000, “Cell Mechanics Studied by a Reconstituted Model Tissue,” Biophys J, 79, pp. 2353–68. 20. Isenberg, B. C., and Tranquillo, R. T., 2003, “Long-Term Cyclic Distention Enhances the Mechanical Properties of Collagen-Based Media-Equivalents,” Ann Biomed Eng, 31, pp. 937–49. 21. Grinnell, F., 1994, “Fibroblasts, Myofibroblasts, and Wound Contraction,” J Cell Biol, 124, pp. 401–4. 22. Guidry, C., 1993, “Fibroblast Contraction of Collagen Gels Requires Activation of Protein Kinase C,” J Cell Physiol, 155, pp. 358–67. 23. Delvoye, P., Wilquet, P., Leveque, J.-L., Nusgens, B. V., and Lapiere, C. M., 1991, “Measurement of Mechanical Forces Generated by Skin Fibroblasts Embedded in a Three-Dimensional Collagen Gel,” J Invest Dermatol, 97, pp. 898–902. 24. Kolodney, M. S., and Elson, E. L., 1993, “Correlation of Myosin Light Chain Phosphorylation with Isometric Contraction of Fibroblasts,” J Biol Chem, 268, pp. 23850–5. 25. Eastwood, M., Mcgrouther, D. A., and Brown, R. A., 1994, “A Culture Force Monitor for Measurement of Contraction Forces Generated in Human Dermal Fibroblast Cultures: Evidence for Cell-Matrix Mechanical Signalling,” Biochim Biophys Acta, 1201, pp. 186–92.

100

G.M. Genin et al.

26. Eastwood, M., Porter, R., Khan, U., Mcgrouther, G., and Brown, R., 1996, “Quantitative Analysis of Collagen Gel Contractile Forces Generated by Dermal Fibroblasts and the Relationship to Cell Morphology,” J Cell Physiol, 166, pp. 33–42. 27. Brown, R. A., Prajapati, R., Mcgrouther, D. A., Yannas, I. V., and Eastwood, M., 1998, “Tensional Homeostasis in Dermal Fibroblasts: Mechanical Responses to Mechanical Loading in Three-Dimensional Substrates,” J Cell Physiol, 175, pp. 323–32. 28. Huang, D., Chang, T. R., Aggarwal, A., Lee, R. C., and Ehrlich, H. P., 1993, “Mechanisms and Dynamics of Mechanical Strengthening in Ligament-Equivalent Fibroblast-Populated Collagen Matrices,” Annals of Biomedical Engineering, 21, pp. 289–305. 29. Tranquillo, R. T., Girton, T. S., Bromberek, B. A., Triebes, T. G., and Mooradian, D. L., 1996, “Magnetically Orientated Tissue-Equivalent Tubes: Application to a Circumferentially Orientated Media-Equivalent,” Biomaterials, 17, pp. 349–57. 30. Eschenhagen, T., Fink, C., Remmers, U., Scholz, H., Wattchow, J., Weil, J., Zimmermann, W., Dohmen, H. H., Schafer, H., Bishopric, N., Wakatsuki, T., and Elson, E. L., 1997, “Three-Dimensional Reconstitution of Embryonic Cardiomyocytes in a Collagen Matrix: A New Heart Muscle Model System [in Process Citation],” Faseb J, 11, pp. 683–94. 31. Eschenhagen, T., and Zimmermann, W. H., 2005, “Engineering Myocardial Tissue,” Circ Res, 97, pp. 1220–31. 32. Zimmermann, W. H., and Eschenhagen, T., 2003, “Cardiac Tissue Engineering for Replacement Therapy,” Heart Fail Rev, 8, pp. 259–69. 33. Zimmermann, W. H., Fink, C., Kralisch, D., Remmers, U., Weil, J., and Eschenhagen, T., 2000, “Three-Dimensional Engineered Heart Tissue from Neonatal Rat Cardiac Myocytes,” Biotechnol Bioeng, 68, pp. 106–14. 34. Zimmermann, W. H., Schneiderbanger, K., Schubert, P., Didie, M., Munzel, F., Heubach, J. F., Kostin, S., Neuhuber, W. L., and Eschenhagen, T., 2002, “Tissue Engineering of a Differentiated Cardiac Muscle Construct,” Circ Res, 90, pp. 223–30. 35. Radisic, M., Park, H., Shing, H., Consi, T., Schoen, F. J., Langer, R., Freed, L. E., and Vunjak-Novakovic, G., 2004, “Functional Assembly of Engineered Myocardium by Electrical Stimulation of Cardiac Myocytes Cultured on Scaffolds,” Proc Natl Acad Sci U S A, 101, pp. 18129–34. 36. Akins, R. E., 2002, “Can Tissue Engineering Mend Broken Hearts?,” Circ Res, 90, pp. 120–2. 37. Zimmermann, W. H., Didie, M., Wasmeier, G. H., Nixdorff, U., Hess, A., Melnychenko, I., Boy, O., Neuhuber, W. L., Weyand, M., and Eschenhagen, T., 2002, “Cardiac Grafting of Engineered Heart Tissue in Syngenic Rats,” Circulation, 106, pp. I151–7. 38. Fink, C., Ergun, S., Kralisch, D., Remmers, U., Weil, J., and Eschenhagen, T., 2000, “Chronic Stretch of Engineered Heart Tissue Induces Hypertrophy and Functional Improvement,” Faseb J, 14, pp. 669–79. 39. Eshelby, J. D., 1957, “The Determination of the Elastic Field Outside an Ellipsoidal Inclusion and Related Problems,” Proc R Soc A, 241, pp. 376–96. 40. Eshelby, J. D., 1959, “The Elastic Field Outside an Ellipsoidal Inclusion,” Proc R Soc Lond Ser A, 252, pp. 561–9. 41. Budiansky, B., 1965, “On the Elastic Moduli of Some Heterogeneous Materials,” J Mech Phys Solids, 13, pp. 223–7. 42. Hill, R., 1965, “A Self-Consistent Mechanics of Composite Materials,” J Mech Phys Solids, 13, pp. 213–22. 43. Mori, T., and Tanaka, K., 1973, “Average Stress in Matrix and Average Elastic Energy of Materials with Misfitting Inclusions,” Acta Metall, 21, pp. 571–4. 44. Chen, C. H., Cheng, C. H., 1996, “Effective Elastic Moduli of Misoriented Short-Fiber Composites.,” Int J Solids Struct, 33, pp. 2519–39. 45. Chou, T. W., 1992, Microstructural Design of Fiber Composites, Cambridge University Press, Cambridge.

5 Cell-Cell Interactions and the Mechanics of Cells and Tissues

101

46. Budiansky, B., and Cui, Y. L., 1995, “Toughening of Ceramics by Short Aligned Fibers.,” Mech Mater, 21, pp. 139–46. 47. Tucker, C. L., and Liang, E., 1999, “Stiffness Prediction for Unidirectional Short-Fiber Composites: Review and Evaluation,” Comp Sci Tech, 59, pp. 655–71. 48. Milton, G., 2002, The Theory of Composites, Cambridge University Press, Cambridge. 49. Genin, G. M., and Birman, V., 2009, “Micromechanics and Structural Response of Functionally Graded, Particulate-Matrix, Fiber-Reinforced Composites,” Int J Solids Struct, 46, pp. 2136–50. 50. Zahalak, G. I., Wagenseil, J. E., Wakatsuki, T., and Elson, E. L., 2000, “A Cell-Based Constitutive Relation for Bio-Artificial Tissues,” Biophys J, 79, pp. 2369–81. 51. Marquez, J. P., Elson, E. L., and Genin, G. M., 2010, “Whole Cell Mechanics of Contractile Fibroblasts: Relations between Effective Cellular and Extracellular Matrix Moduli,” Philos Trans A Math Phys Eng Sci, 368, pp. 635–54. 52. Lanir, Y., 1983, “Constitutive Equations for Fibrous Connective Tissues,” J Biomech, 16, pp. 1–12. 53. Lanir, Y., Salant, E. L., and Foux, A., 1988, “Physico-Chemical and Microstructural Changes in Collagen Fiber Bundles Following Stretch in-Vitro,” Biorheology, 25, pp. 591–603. 54. Sacks, M. S., 2003, “Incorporation of Experimentally-Derived Fiber Orientation into a Structural Constitutive Model for Planar Collagenous Tissues,” J Biomech Eng, 125, pp. 280–7. 55. Horowitz, A., Lanir, Y., Yin, F. C., Perl, M., Sheinman, I., and Strumpf, R. K., 1988, “Structural Three-Dimensional Constitutive Law for the Passive Myocardium,” J Biomech Eng, 110, pp. 200–7. 56. Hurschler, C., Loitz-Ramage, B., and Vanderby, R., Jr., 1997, “A Structurally Based StressStretch Relationship for Tendon and Ligament,” J Biomech Eng, 119, pp. 392–9. 57. Wren, T. A., and Carter, D. R., 1998, “A Microstructural Model for the Tensile Constitutive and Failure Behavior of Soft Skeletal Connective Tissues,” J Biomech Eng, 120, pp. 55–61. 58. Bahei-El-Din, Y. A., 1996, “Finite Element Analysis of Viscoplastic Composite Materials and Structures,” Mech Adv Mater Struct, 3, pp. 1–28. 59. Bao, G., Hutchinson, J. W., and Mcmeeking, R. M., 1991, “Particle Reinforcement of Ductile Matrices Against Plastic Flow and Creep,” Acta Metall Mater, 39, pp. 1871–82. 60. Wu, J. Z., Herzog, W., and Epstein, M., 1999, “Modelling of Location- and Time-Dependent Deformation of Chondrocytes During Cartilage Loading,” J Biomech, 32, pp. 563–72. 61. Wu, J. Z., and Herzog, W., 2000, “Finite Element Simulation of Location- and TimeDependent Mechanical Behavior of Chondrocytes in Unconfined Compression Tests,” Ann Biomed Eng, 28, pp. 318–30. 62. Barocas, V. H., and Tranquillo, R. T., 1997, “A Finite Element Solution for the Anisotropic Biphasic Theory of Tissue-Equivalent Mechanics: The Effect of Contact Guidance on Isometric Cell Traction Measurement,” J Biomech Eng, 119, pp. 261–8. 63. Moon, A. G., and Tranquillo, R. T., 1993, “The Fibroblast-Populated Collagen Microsphere Assay of Cell Traction Force: Part 1. Continuum Model,” AIChE J, 39, pp. 163–77. 64. Tranquillo, R. T., 1999, “Self-Organization of Tissue-Equivalents: The Nature and Role of Contact Guidance,” Biochem Soc Symp, 65, pp. 27–42. 65. Tranquillo, R. T., and Murray, J. D., 1993, “Mechanistic Model of Wound Contraction,” J Surg Res, 55, pp. 233–47. 66. Tranquillo, R. T., and Murray, J. D., 1992, “Continuum Model of Fibroblast-Driven Wound Contraction: Inflammation-Mediation,” J Theor Biol, 158, pp. 135–72. 67. Zemel, A., and Safran, S. A., 2007, “Active Self-Polarization of Contractile Cells in Asymmetrically Shaped Domains,” Phys Rev E Stat Nonlin Soft Matter Phys, 76, p. 021905. 68. Zemel, A., Bischofs, I. B., and Safran, S. A., 2006, “Active Elasticity of Gels with Contractile Cells,” Phys Rev Lett, 97, p. 128103.

102

G.M. Genin et al.

69. Hill, A., 1938, “The Heat of Shortening and the Dynamic Constants of Muscle,” Proc R Soc Lond B, 126, pp. 136–95. 70. Bird, R. B., Curtiss, C. F., Armstrong, R. C., and Hassager, O., 1987, Dynamics of Polymeric Liquids, Volume Ii: Kinetic Theory, Wiley, New York. 71. Marquez, J. P., Genin, G. M., and Elson, E. L., 2006, “On the Application of Strain Factors for Approximation of the Contribution of Anisotropic Cells to the Mechanics of a Tissue Construct,” J Biomech, 39, pp. 2145–51. 72. Prager, W., 1969, “On the Formulation of Constitutive Equations for Living Soft Tissues,” Appl Math, 27, pp. 128–32. 73. Mura, T., 1982, Micromechanics of Defects in Solids, Martinus Nijhoff, The Hague. 74. Bug, A. L. R., Safran, S. A., Grest, G. S., and Webman, I., 1985, “Do Interactions Raise or Lower a Percolation Threshold?,” Phys Rev Lett, 55, pp. 1896–9. 75. Bug, A. L. R., Safran, S. A., and Webman, I., 1985, “Continuum Percolation of Rods,” Phys Rev Lett, 54, pp. 1412–5. 76. Kallmes, O., and Corte, H., 1960, “The Structure of Paper: The Statistical Geometry of an Ideal Two-Dimensional Fiber Network.,” Tappi, 43, pp. 737–52. 77. Pabst, W., and Gregorov·, E., 2004, “Effective Elastic Properties of Alumina-Zirconia Composite Ceramics – Part 2. Micromechanical Modeling,” Ceram Silikaty, 48, pp. 14–23. 78. Marquez, J. P., Genin, G. M., Pryse, K. M., and Elson, E. L., 2006, “Cellular and Matrix Contributions to Tissue Construct Stiffness Increase with Cellular Concentration,” Ann Biomed Eng, 34, pp. 1475–82. 79. Enever, P. A., Shreiber, D. I., and Tranquillo, R. T., 2002, “A Novel Implantable Collagen Gel Assay for Fibroblast Traction and Proliferation During Wound Healing,” J Surg Res, 105, pp. 160–72. 80. Engler, A., Bacakova, L., Newman, C., Hategan, A., Griffin, M., and Discher, D., 2004, “Substrate Compliance Versus Ligand Density in Cell on Gel Responses,” Biophys J, 86, pp. 617–28. 81. Zhu, Y. K., Umino, T., Liu, X. D., Wang, H. J., Romberger, D. J., Spurzem, J. R., and Rennard, S. I., 2001, “Contraction of Fibroblast-Containing Collagen Gels: Initial Collagen Concentration Regulates the Degree of Contraction and Cell Survival,” In Vitro Cell Dev Biol Anim, 37, pp. 10–6. 82. Stylianopoulos, T., Yeckel, A., Derby, J. J., Luo, X. J., Shephard, M. S., Sander, E. A., and Barocas, V. H., 2008, “Permeability Calculations in Three-Dimensional Isotropic and Oriented Fiber Networks,” Phys Fluids, 20, pp. 123601. 83. Nekouzadeh, A., Pryse, K. M., Elson, E. L., and Genin, G. M., 2008, “Stretch-Activated Force Shedding, Force Recovery, and Cytoskeletal Remodeling in Contractile Fibroblasts,” J Biomech, 41, pp. 2964–71. 84. Wang, N., Naruse, K., Stamenovic, D., Fredberg, J. J., Mijailovich, S. M., Tolic-Norrelykke, I. M., Polte, T., Mannix, R., and Ingber, D. E., 2001, “Mechanical Behavior in Living Cells Consistent with the Tensegrity Model,” Proc Natl Acad Sci U S A, 98, pp. 7765–70. 85. Nekouzadeh, A., Genin, G. M., Bayly, P. V., and Elson, E. L., 2005, “Wave Motion in Relaxation-Testing of Nonlinear Elastic Media,” Proc R Soc A Math Phys Eng Sci, 461, p. 1599. 86. Nekouzadeh, A., Pryse, K. M., Elson, E. L., and Genin, G. M., 2007, “A Simplified Approach to Quasi-Linear Viscoelastic Modeling,” J Biomech, 40, pp. 3070–8. 87. Pryse, K. M., Nekouzadeh, A., Genin, G. M., Elson, E. L., and Zahalak, G. I., 2003, “Incremental Mechanics of Collagen Gels: New Experiments and a New Viscoelastic Model,” Ann Biomed Eng, 31, pp. 1287–96. 88. Krishnan, R., Park, C. Y., Lin, Y. C., Mead, J., Jaspers, R. T., Trepat, X., Lenormand, G., Tambe, D., Smolensky, A. V., Knoll, A. H., Butler, J. P., and Fredberg, J. J., 2009, “Reinforcement Versus Fluidization in Cytoskeletal Mechanoresponsiveness,” PLoS One, 4, p. e5486.

5 Cell-Cell Interactions and the Mechanics of Cells and Tissues

103

89. Gavara, N., Roca-Cusachs, P., Sunyer, R., Farre, R., and Navajas, D., 2008, “Mapping Cell-Matrix Stresses During Stretch Reveals Inelastic Reorganization of the Cytoskeleton,” Biophys J, 95, pp. 464–71. 90. Wakatsuki, T., Schlessinger, J., and Elson, E. L., 2004, “The Biochemical Response of the Heart to Hypertension and Exercise,” Trends Biochem Sci, 29, pp. 609–17. 91. Weber, K. T., 2000, “Fibrosis and Hypertensive Heart Disease,” Curr Opin Cardiol, 15, pp. 264–72. 92. Manabe, I., Shindo, T., and Nagai, R., 2002, “Gene Expression in Fibroblasts and Fibrosis: Involvement in Cardiac Hypertrophy,” Circ Res, 91, pp. 1103–13. 93. Camelliti, P., Borg, T. K., and Kohl, P., 2005, “Structural and Functional Characterisation of Cardiac Fibroblasts,” Cardiovasc Res, 65, pp. 40–51. 94. Miragoli, M., Gaudesius, G., and Rohr, S., 2006, “Electrotonic Modulation of Cardiac Impulse Conduction by Myofibroblasts,” Circ Res, 98, pp. 801–10. 95. Maccannell, K. A., Bazzazi, H., Chilton, L., Shibukawa, Y., Clark, R. B., and Giles, W. R., 2007, “A Mathematical Model of Electrotonic Interactions Between Ventricular Myocytes and Fibroblasts,” Biophys J, 92, pp. 4121–32. 96. Rohr, S., 2009, “Myofibroblasts in Diseased Hearts: New Players in Cardiac Arrhythmias?,” Heart Rhythm, 6, pp. 848–56. 97. Tomasek, J. J., Gabbiani, G., Hinz, B., Chaponnier, C., and Brown, R. A., 2002, “Myofibroblasts and Mechano-Regulation of Connective Tissue Remodelling,” Nat Rev Mol Cell Biol, 3, pp. 349–63. 98. Rosenkranz, S., 2004, “Tgf-Beta1 and Angiotensin Networking in Cardiac Remodeling,” Cardiovasc Res, 63, pp. 423–32. 99. Kass, D. A., Bronzwaer, J. G., and Paulus, W. J., 2004, “What Mechanisms Underlie Diastolic Dysfunction in Heart Failure?,” Circ Res, 94, pp. 1533–42. 100. Holubarsch, C., Ruf, T., Goldstein, D. J., Ashton, R. C., Nickl, W., Pieske, B., Pioch, K., Ludemann, J., Wiesner, S., Hasenfuss, G., Posival, H., Just, H., and Burkhoff, D., 1996, “Existence of the Frank-Starling Mechanism in the Failing Human Heart. Investigations on the Organ, Tissue, and Sarcomere Levels,” Circulation, 94, pp. 683–9. 101. Allen, D. G., and Kentish, J. C., 1985, “The Cellular Basis of the Length-Tension Relation in Cardiac Muscle,” J Mol Cell Cardiol, 17, pp. 821–40. 102. Fuchs, F., and Smith, S. H., 2001, “Calcium, Cross-Bridges, and the Frank-Starling Relationship,” News Physiol Sci, 16, pp. 5–10. 103. Moss, R. L., and Fitzsimons, D. P., 2002, “Frank-Starling Relationship: Long on Importance, Short on Mechanism,” Circ Res, 90, pp. 11–3. 104. Kohl, P., Camelliti, P., Burton, F. L., and Smith, G. L., 2005, “Electrical Coupling of Fibroblasts and Myocytes: Relevance for Cardiac Propagation,” J Electrocardiol, 38, pp. 45–50. 105. Gaudesius, G., Miragoli, M., Thomas, S. P., and Rohr, S., 2003, “Coupling of Cardiac Electrical Activity over Extended Distances by Fibroblasts of Cardiac Origin,” Circ Res, 93, pp. 421–8. 106. Miragoli, M., Salvarani, N., and Rohr, S., 2007, “Myofibroblasts Induce Ectopic Activity in Cardiac Tissue,” Circ Res, 101, pp. 755–8.

Chapter 6

Specific and Non-Specific Adhesion in Cancer Cells with Various Metastatic Potentials Xin Tang, Tony Cappa, Theresa Kuhlenschmidt, Mark Kuhlenschmidt, and Taher Saif

This chapter is part of Section III: Mechano-pathology of Disease

Abstract The majority of cancer deaths are caused by metastasis, not by the parent tumor. During metastasis, cancer cells detach from their neighbors in the tumor, enter the circulation system, and invade other organs. Thus metastasis initiates with the decrease of adhesion between cancer cells and regaining of adhesion to invade new organs during circulation. In this chapter, we first introduce the altered tumor cells’ behavior during malignant transition with the focus on alterations of cell adhesion. Second, we discuss the cell-cell vs. cell-matrix interaction during cancer metastasis in the context of hierarchical changes in cell-properties during metastasis. Finally, we review the current state-of-the-art devices of single-cell adhesion measurement and present our results on non-specific adhesion measurement between cancer cells and a glass micro probe, using a novel micro force sensor. Adhesion measurements are compared for malignant, metastatic cells (HCT-8), malignant, non-metastatic adenocarcinoma cells (Caco-2), and normal monkey kidney MA104 cells. The non-specific adhesion force for HCT-8 is about twice that compared to Caco-2, and insignificant for MA104. We then measured the specific homotypic cell-cell adhesion rates using a Coulter Counter for PC3M (highly metastatic human prostate cancer), PC3 (less metastatic than PC3M), HCT8 and MA104 cell lines. We found, specific cell adhesion rates decrease with increasing metastatic potential. These results suggest, while specific adhesion of malignant cancer cells are often reduced to facilitate their detachment from the parent tumor, the high non-specific adhesion force on them can strategically maximize their adherence and invasion to diverse tissues during metastasis. Keywords Cancer metastasis Non-specific cancer adhesion Homotypic cell-cell adhesion rate Bio-MEMS force sensor Coulter-counter adhesion assay T. Saif (*) Department of Mechanical Science and Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, USA and Micro and Nanotechnology Laboratory, University of Illinois at Urbana-Champaign, Urbana, IL, USA e-mail: [email protected] A. Wagoner Johnson and Brendan A.C. Harley (eds.), Mechanobiology of Cell-Cell and Cell-Matrix Interactions, DOI 10.1007/978-1-4419-8083-0_6, # Springer Science+Business Media, LLC 2011

105

106

6.1

X. Tang et al.

Introduction to Metastasis

Despite the significant improvement in both the early diagnosis and treatment of cancer patients, metastasis is still the major cause of mortality. Ninety percent of cancer deaths are caused by metastasis of malignant cells, not by the primary tumor itself [1–3]. During metastasis, malignant cancer cells turn off cell adhesion activity, de-adhere from their neighbors or the extracellular matrix (ECM), enter the lymphatic system or the blood stream as cell suspensions, invade new host tissues and organs, regain adhesive activity and form new tumor colonies [1, 4–7]. Thus, adhesion plays a central role in management of successful metastasis. In spite of the central role that adhesion plays during metastasis, measurement of adhesion at the single cell scale for cancer cells is not widely studied, primary due to limitation of instrumentation. In this chapter, we first introduce the altered tumor cells behaviors during malignant transition with the focus on the alterations in their adhesive phenotype. Second, we discuss the cell-cell vs. cell-matrix interaction during cancer metastasis in the context of the changes in hierarchical cellproperties. Finally, we review the current state-of-the-art devices of single-cell adhesion measurement and, in particular, present a novel micro-scale Bio-MEMS instrumentation for the sensitive measurement of adhesive forces, at a single cell scale, to define the relationship between adhesive force and metastatic potential. Such quantitative data on cancer cell adhesion at various stages of metastasis may be used to ultimately correlate adhesive strength with cell adhesion molecule (CAM) gene expression. Quantitative measurements of cancer cell adhesion offers the potential of discovering molecular mechanisms that cancer cells employ to regulate adhesion during metastasis. It thus might be possible to rapidly screen anti-metastatic drugs targeted at regulation of cell adhesion.

6.2

6.2.1

Effects from Both Neighbors and Microenvironment During Metastasis Altered Tumor Cell Behavior During Metastasis

Adherence of cells to neighboring cells or extacellular matrix (ECM) allows a variety of signaling events that regulate normal cell functionality. A fundamental characteristic of cancer cells is altered adhesion properties, both to other cells and to the ECM. In non-cancerous cells, adhesion contributes to the regulation of cellular movement, division, organization into well-coordinated clusters to form tissues and organs, and survival. Altered properties of adhesion in cancer cells can not only promote metastasis and cell division, but can also provide inappropriate survival signals that can impact the ability to treat cancers selectively. Normal cells, when detached or mis-oriented from their normal tissue architecture, initiate anoikis, a form of apoptosis or programmed cell death [5]. Metastatic cancer cells, however, behave

6 Specific and Non-Specific Adhesion in Cancer Cells

107

quite differently when detached from the primary tumor. When they de-adhere from neighbor cells or the extracellular matrix (ECM) and enter the lymphatic system or the blood stream, they can survive for several days (3–5 days) by inhibiting or delaying anoikis [5]. During this time the cells find a new host tissue(s), which presents the appropriate microenvironment, and regain adhesive activity, local cellular motility and invade the new host tissue (metastases). The causes of low adhesivity of cancer cells, which de-adhere from primary tumor and prior to their arrival to a secondary site, fall into mainly three categories: (1) In concert with metastasis, most malignant cells down-regulate their surface adhesion molecules, such as E-Cadherin [1, 2, 4, 8, 9], or the key components mediating the adhesion molecules, such as the cytoplasmic protein a-catenin [10–12]; (2) Calcium plays a critical role in regulating cell-cell adhesion and thus, reducing Calcium is a common approach for tissue disassembly in cell culture experiments [13]. Coman [14, 15] showed that the cancer cells contain only about half the [Ca2+] than do the normal cells from the same organ, suggesting Ca2+ deficiency in cancer cell may be associated with weakening of their mutual adhesivity [14, 15]; (3) During metastasis, cancer cells express surface proteases to degrade the ECM and normal cell adhesion in tissues in order to create free space for invasion. For example, using time-resolved multimodal microscopy, Wolf et al. [16] showed invasive breast cancer cells degrade and remodel extracellular matrix (ECM) by secreting MT1 matrix metalloproteinase [16, 17]. Such excessive release of proteolytic enzymes in turn can disrupt the adhesion molecules on the cancer cells themselves [18–20]. Cancer cells also express altered mechanical phenotypes throughout the metastasis process. For example, cancer cells show increased cell-microenvironment traction forces associated with increased Rho and ROCK activity [21–23], altered force-bearing cytoskeleton architectures [24–26], and increased invasion migration speed [27–30].

6.2.2

Remodeling the Tumor Microenvironments

Stromal-epithelial interaction plays an important role in regulating normal cell behaviors in normal tissue. However, deregulation of these interactions during tissue disease or inflammation can promote tumorigenesis and cancer invasion [7, 23, 31–33]. For example, Collagen I is one of ECM components present vastly in the stromal extracellular matrix and it interacts with the cells adhesion molecules, integrin a2b1. It has been reported that patients with collagen-dense breast tissue experience a fourfold higher risk of breast carcinoma and metastasis compared to those with normal collagen-levels [34, 35]. Using nonlinear microscopy techniques such as multiphoton laser-scanning microscopy and second harmonic generation, Provenzano et al. [36] revealed the alterations in 3D local collagen density in the neoplastic mammary stromas, and discovered that the breast cancer cell invasion in mammary gland is facilitated by the radially aligned collagen fibers [36].

108

X. Tang et al.

Thus, during cancer epithelial-mesenchymal transition (EMT), the local microenvironment of primary tumor is remolded and modified to assist the cell invasion and transendothelial migration. Both malignant cancer cells and the recruited stroma cells, such as macrophages, carry out these remodeling and degradation of microenvironment [37, 38]. In addition to remodeling tissue microenvironment around the primary tumor, cancer cells can also alter the microenvironment of distant secondary organs to prepare for metastasis. It has been clinically recognized that certain types of tumors preferentially metastasize to particular secondary sites. For example, colon cancers preferentially metastasize to liver, while breast and prostate cancers often metastasize to bone [1, 2, 9]. Since this directed distribution of metastases cannot be solely explained by the patterns of blood flow which carry the de-adhered tumor cells throughout the body, a “seed and soil” theory was proposed by Paget in 1889, stating that cancer cells (seeds) display survival dependence on the microenvironment of secondary organs (soil) [39, 40]. A recent study by Kaplan [41] shed new light on unrevealing this phenomenon [41]. They found that bone marrow-derived haematopoietic progenitor cells (BMDC), which express vascular endothelial growth factor 1 (VEGFR1), colonize tumor-specific pre-metastatic sites and form cellular clusters before the arrival of tumor cells. These VEGFR1-expressed progenitors generate a conducive microenvironment to promote tumor adherence and growth at the selected secondary sites, resulting in organ-specific tumor spread.

6.2.3

Cell-Cell vs. Cell Matrix Interaction During Metastasis

Considerable evidence now indicates cancer cells respond differently in their gene expression, adhesive activity, intracellular mechanical rheology, growth, and susceptibility to therapy depending on the microenvironment they encounter during metastasis [1, 3, 22, 42, 43]. For example, Paszek (2005) have shown mammary epithelial cells (MECs) display both structural and transcriptional hallmarks of tumor transformation, invasion and metastasis when cultured in ECMs of the mechanical stiffness resembling in vivo tumor stroma, but not in the 5–20-times softer ECMs [44]. Baker (2009) found the effective intracellular compliance and stiffness of prostate cancer cells increased an order-of-magnitude as the matrix environment switched from 2D to 3D, and was attributed to the increased cellmatrix integrin adhesion and engagements in the 3D matrix architecture [43]. Moreover, Dalton and Jove (1999) has shown soluble cytokines, such as IL-6 and cell adhesion molecules (CAMs) contribute to the microenvironment-mediated responses of tumor cells to chemotherapy [45]. Sehgal et al. (2006) has shown that during the initial phase of metastasis, the cells over-express urokinase receptor (uPAR), a molecule that has various roles on cell functionality including degradation of ECM and cell adhesion activity, inhibition of anoikis, motility, and increased adhesion during tumor invasion [46]. Inhibition of uPAR has been found to reduce colonization of extra-prostatic sites in animal models. Thus, disruption of adhesion

6 Specific and Non-Specific Adhesion in Cancer Cells

109

mechanisms of cancer cells to its microenvironment during metastasis may represent a new approach for the treatment of cancer. Therefore, understanding cellular adhesion mechanisms and how they are different in metastatic cancer cells is crucial for understanding metastasis and discovery of possible new approaches and targets for cancer therapy. There is increasing experimental and computational evidence suggesting extracellular and intracellular mechanical forces have a profound influence on a wide range of cell behavior including adhesion and signal transduction, growth, differentiation, apoptosis, and gene expression [47–56]. Cells transmit these forces through localized adhesion sites at which they are adhered to other cells or extracellular matrix (ECM). These adhesion sites are formed by transmembrane proteins, called integrins, to anchor the cell to a matrix, or adhesion molecules such as cadherins and selectins, to anchor to other cells. Both the integrins and adhesion molecules are attached to the tensile members of the cytoskeleton, the actin filaments, through focal adhesion complex – a highly organized cluster of molecules [57, 58]. The integrins and cell-cell adhesion molecules, together with the cytoskeleton, form a hard-wired network for signal transmission [53]. It has been shown by Assoian and Klein [59] that several signaling cascades, such as FAK, Rho GTPases and ERK, transmitting through this hard-wired network, can control cancer cell growth and essential cell-cycle progression via regulating cyclin D1 gene expression [59–61]. Moreover, Paszek and co-workers suggested that malignant transformation of the breast might be functionally related to abnormal integrin organizations resulting from perturbed tensional-homeostasis, increased mammary gland tension and elevated ECM stiffness [56, 62]. Furthermore, a recent study by Levental (2009) showed that induction of collagen crosslinking to stiffen ECMs can promote cancer cells’ focal adhesions, enhance their PI3L kinase activity and induce the invasion of an oncogene-initiated epithelium [63]. Cell-cell interaction is also very critical throughout metastasis. As cancer cells invade adjacent tissues, they often form ECM-degrading membrane adhesive protrusions on cell front called invadopodia. Yamaguchi and Condeelis (2007) showed that the tumor-associated macrophages can interact with carcinoma cells through EGF/CSF-1 paracrine loop [64–66]. Upon activation by cancer cells, macrophages generate another type of membrane adhesive protrusion called podosomes [67], which share the same molecular compositions and ECM-degrading function with invadopodia to promote cancer invasion and metastasis. Clinical data indicate the combined presence of tumor-associated macrophages and cancer cells in tumors correlates with poor prognosis in human cancers [68, 69].

6.2.4

Techniques Used to Study Cell Adhesion

In spite of the central role of adhesion in metastasis, adhesive interactions have not been quantified, especially as they are influenced by different microenvironments in which the cells find themselves during the metastatic process. This is primarily

110

X. Tang et al.

due to the limitation of instrumentation that can probe and measure adhesion at a cellular scale. Such a cellular scale measure of adhesion will not only test the hypothesis that alterations in intercellular adhesion are responsible for metastatic behavior, but will also add a new mechanistic dimension in the understanding of the metastatic process and potential prevention strategy. For example, if the adhesive forces responsible for cell-cell and cell-matrix adhesion could be reproducibly measured in an in vivo-relevant, in vitro analysis system, then straightforward experiments could be designed to isolate, characterize and compare the relevant cell surface adhesion receptors between normal and highly metastatic cancer cells, as well as the signaling mechanisms which regulate expression of these adhesion molecules within the various microenvironments occupied by metastasizing cancer cells. Furthermore, such an assay system would allow rapid screening of putative anti-metastatic drugs that have the potential to block adhesivity of metastatic cancer cells to their respective target organ microenvironment and thereby block metastatic growth. There are several outstanding state-of-the-art devices developed to study cellularand molecule-level cell adhesion, and they have shed new light on several longlasting biological questions related to cell adhesion. In particular, by combining AFM with side-view fluorescent microscopy, Chaudhuri et al. (2009) directly observed cell shape evolution, cytoskeleton reorganization and stress-fiber formation while simultaneously measuring the contractile force applied by single cells [70, 71]. This instrument may help to reveal the molecular mechanisms underlying the cell response upon mechanical stimulus. Additionally, this instrument was used to measure the long-term adhesion interaction between leukocytes and endothelial cells to mimic in vivo leukocytes circulating in the blood vessels. In D. Wirtz’s group [10, 11, 72], the single-molecule force spectroscopy (MFP) was used to measure adhesion characteristics between single E-cadherin or N-cadherin homophilic bonds. Aided by the high-precision and versatility of MFP, they found that, although bond lifetime for those two adhesion molecules is similar, their adhesion force, reactive compliance and single-barrier energy potential differ markedly [72]. Using the same methodology, Bajpai et al. (2008, 2009) also discovered the disruption of a-catenin could dramatically reduce the lifetime, tensile strength and mechanical stability of single cadherin bonds between human breast cancer cells, suggesting that down-regulation of cell surface E-cadherin may not be entirely responsible for the loss of intercellular adhesion during cancer progression [10, 11]. It is known that during metastasis, cancer cells down-regulate their surface adhesion molecules, such as E-Cadherin, to enhance the escape from the primary tumor [4, 8, 12, 73]. However, it is not clear how can these E-Cadherin-deficient cancer cells manage to adhere to and invade new target organs during metastasis. In this chapter, we hypothesize that cancer cells manage their invasive behavior by regulating two types of surface adhesion, the specific and non-specific surface adhesivities. A novel and versatile microelectromechanical systems (bio-MEMS) force sensor is developed to quantify the strength of non-specific adhesion between a series of living cancer cells and the Silicon probe [74–76]. Unlike the AFM, our bio-MEMS force sensors apply the mechanical stimulus horizontally and, thus

6 Specific and Non-Specific Adhesion in Cancer Cells

111

optical measurement can be conveniently performed in an inverted phase-contrast microscope. The classic Coulter counter is also used to measure the homotypic specific cell-cell adhesion rates between cells of the same type [77]. Here the non-specific adhesion activity may arise from van der Waals attraction, or electrostatic attraction and repulsion depending on chemophysiological composition of cell surface proteins [78–80]. Specific adhesion refers to ligand-receptor binding among the cells of the same type. It has been reported that cells of one type tend to preferentially adhere to each other, namely, cells often show higher specific homotypic rather than heterotypic adhesion [80–82]. However, during metastasis, malignant cells in travel encounter various secondary tissues consisting of cells different from their own type. Thus, it is hypothesized that non-specific adhesion plays an essential role in the invasion process but this adhesive activity is poorly understood. Our results suggest that, cancer cells exhibit down-regulated specific adhesion activity but possess higher non-specific adhesion strength compared to that of the normal cells. This high non-specific adhesion strength might enhance the ability of cancer cells to adhere to the target organs during metastasis followed by successful invasion.

6.3

6.3.1

Bio-MEMS Force Sensor and Non-Specific Adhesion Study Bio-MEMS Force Sensor

The sensor consists of a pair of structural beams (Fig. 6.1a), anchored at both ends. They are connected at the middle by a rigid backbone, which also serves as a probe to contact a cell. When a force is applied to the probe, the beams deform together. Their deformation, x, is measured with respect to a reference probe. The force is given by F ¼ K x, where K is the spring constant of the double beam sensor. The spring constant is calibrated by an Atomic Force Microscope. The displacement x is measured optically. The sensor is fabricated from single-crystal Silicon using the SCREAM process. The details of the fabrication process are reported in [74, 75, 83]. Typical length of the sensor beams is 1–2 mm. They are 1–3 mm wide and about 10 mm deep. Thus the spring constant of the sensor is in the range of 1 nN/mm to 1 mN/mm [76]. The sensor used in this study is 14 nN/mm. The minimum displacement that can be measured using correlation methods is about 0.1 mm. Thus, the force resolution is about 1.4 nN. The probe of the sensor employed in this study is a flat vertical plate with dimension 5 10 mm. The sensor is held and manipulated by an x–y–z piezo actuator (Fig. 6.1b), which brings the probe in contact with the cells. The sensor is slightly tilted (about 5 ) with the horizontal so that the probe is at the lowest point. As the cells are proliferating on polystyrene substrate, they rarely remain single, but often form cell colonies that grow in size until the cells form a monolayer. The adhesion experiments have to be performed before this monolayer is formed.

112

X. Tang et al.

Fig. 6.1 (a) The schematic of bio-MEMS force sensor to measure the non-specific cell surface adhesion (top view). The probe, a vertical glass plate 5 10 mm, interfaces the cell or cell colony for 2 min and forms a non-specific adhesion. The probe is then retracted from the cell. If there is an adhesion between cell and the probe, then the sensor springs deform as shown by d. The cell force is given by K d, where K is the spring constant of he sensor beams, calibrated by Atomic Force Microscope. The maximum force prior to detachment is used as the non-specific adhesion of the cell/cell colony. The inset shows phase-contrast image of a HCT-8 cell colony being pulled by retracting bio-MEMS force sensor while the non-specific adhesion is measured. Scale bar: 50 mm. (b) The schematic of the experimental setup for adhesion measurement (side view). The sensor and the probe are manipulated by the piezo-actuators. The digital video recording is performed to monitor experiments and for data analysis and interpretation

In the current study, all experiments are consistently performed as the cell colonies diameter is around 100 mm, 2 days after plating (Fig. 6.1a, inset). The probe indents the colony along the plane of the culture substrate by about 5–8 mm, which is less than 10% of cell colony diameter, (about 100 mm), thus

6 Specific and Non-Specific Adhesion in Cancer Cells

113

forming an area of contact. After holding the probe for 2 min, the probe is retracted. Due to the short time of probe-cell contact, focal adhesion complex is unlikely to establish, and the adhesion between the cell and the probe, if any, is non-specific in nature. We have empirically found that until about 2 h of contact, the adhesion between non-functionalized micro-probe and cells is time-independent. After 2 h of contact, nascent focal adhesion sites might be formed between the probe and the cell. The characteristics of focal adhesion rupture strengths are reported previously by Yang [76]. After 2 min of contact, if there is adhesion between the probe and the cell (which is part of the colony), then the cell applies a resisting force during retraction, which is measured by the sensor. With increasing retraction, the force-retraction relation is obtained (Fig. 6.2a–c). In this study we use the maximum interaction force as a measure of adhesion, since the area of contact cannot be

Fig. 6.2 (a) The force sensor probe is approaching the cell sample. Note the reference and measurement pins are well aligned. (b) After cell-probe contact of 2 min, the probe starts to pull back. The non-specific adhesion between the probe and cell allows the force to increase (reflected by + d1). (c) The largest pulling force is attained before the cell detaches from the probe, as reflected by + d2. After detachment, the probe realigns back with the reference pin. The phase-contrast picture series for a Caco-2 adhesion test are shown beside the schematics. Scale bar: 50 mm

114

X. Tang et al.

ascertained with certainty – a common problem in cell adhesion measurements. In all the experiments reported here, the culture conditions as well as the probe geometry and the retraction speed are kept the same.

6.3.2

Malignant Cancer Cells Possess High Non-Specific Adhesion

In order to evaluate the relation between non-specific adhesivity and cell malignancy, we strategically choose three cell lines and culture them on 2D polystyrene surface of the tissue culture Petri dishes: malignant, metastatic (HCT-8), malignant, nonmetastatic (Caco-2) human colorectal adenocarcinoma cells [84, 85], and normal monkey kidney MA104 cells. During culture, the cells form clusters or colonies. We measure the adhesion between the colony and the probe. Figure 6.3 shows the adhesion strengths. The aggressive HCT-8 cells show high surface non-specific adhesivity (49.58 11.93 nN). The less aggressive Caco-2 cells have strengths of 27.50 3.49 nN. Finally, the normal MA104 cells show no surface adhesion. Figure 6.4 shows the representative adhesion force vs. displacement curve of HCT-8 cancer cells. The curve shows a 2-slope force behavior, which is followed by a sudden adhesion failure (rupture). It is qualitatively different from the earlier cell adhesion measurements where the adhesion between a monkey kidney

Fig. 6.3 Non-specific adhesion between a glass probe (a micro plate) and a cell colony of cancer cells with different metastatic potentials, or normal cells. Adhesion strength is defined by the maximum force to detach the probe from the cell colony. Human colon carcinoma cells (HCT-8), human colon adenocarcinoma cells (Caco-2) and normal kidney epithelial cells (MA104) are tested. HCT-8 are more aggressive than Caco-2 (n ¼ 15 for each column). Higher the metastatic potential, higher is the non-specific adhesion. Normal cells show no non-specific adhesion with the glass micro plate

6 Specific and Non-Specific Adhesion in Cancer Cells

115

Fig. 6.4 (a) The force displacement curve of adhesion measurement for HCT-8 cells. The largest adhesion strength is measured prior to the sudden detachment between probe and cell membrane. (b) The time-lapse phase-contrast images of adhesion measurement process corresponding to Fig. 6.4a. Cell force is measured from the misalignment between the reference marker and the probe

fibroblast and a fibronectin coated probe [76] was formed. With increasing cell stretch (probe retraction), the cell force increases, reaches a maximum, and then gradually decreases until detachment. The force maxima and decay may result from progressive failures of the focal adhesions that might have formed between the cell and the fibronectin coated probe. In current situation (Fig. 6.4a), the force response has no maxima, but it has two linear regimes. Initially, the cell force with stretch shows a high slope, about 4.2 nN/mm, followed by a slope of 1.1 nN/mm. The cell then detaches from the probe abruptly. Figure 6.4b shows the sensor and the cell colony at different stages of cell force response.

6.3.3

Homotypic Cell-Cell Adhesion Rates Decrease in Malignant Cancer Cells

Specific homotypic cell-cell adhesion rates for HCT-8 cells and normal Ma104 cells were compared using a Coulter counter assay as previously described [77]. To determine the change of the homotypic cell-cell adhesion of cancer cells before and after the metastasis, we also tested another two cell lines, PC3 and PC3M. The PC3 cells are malignant prostate cancer cells harvested from the primary tumor before metastasis, while PC3M cells are the metastasized PC3 cells harvested from the lymph nodes after PC3 cells were injected into mice models. We choose these two cell lines with known metastatic potential to investigate the correlation between specific cell adhesion strength and cell metastasis.

116

X. Tang et al.

In all cases, the cells were harvested from the polystyrene substrates, individualized by trypsin treatment, and allowed to recover from trypsinization under the same incubation conditions prior to measuring intercellular adhesion rates as described in Materials and Methods. The Coulter counter measures the rate and extent of cell adhesion by quantifying the reduction in the number of single cells in suspension as cell aggregates form with time. Interestingly, the highly malignant PC3M cells displayed a markedly lower extent and rate of cell-cell adhesion as compared to the PC3 cells, HCT-8 cells and normal MA-104 cells (Fig. 6.5). After 120 min of incubation, 92.2 2.3% of the PC3M cells remained as single cells, in contrast to 73.9 4.8% of original PC3 cells, 52.8 1.0% of HCT-8 cells and 9.8 2.1% of normal Ma104 cells (Fig. 6.5). This rather remarkable result is also consistent with the current understanding that cancer cells reduce surface cell-cell adhesion molecules, such as E-Cadherin, to enhance the escape from primary tumor [4, 8, 12]. Taken together, these results strongly support the hypothesis that cancer cells with higher metastatic potential have lower specific cell-cell adhesion rates, but higher non-specific adhesion. This suggests that highly metastatic cancer cells may employ non-specific adhesion to adhere to target tissues in order to invade and form new colonies.

Fig. 6.5 Cell-cell adhesion measured by Coulter Counter. It measures the percentage of single cells in cell suspension as a function of time. Cells adhere to one another quickly if cell-cell adhesion is high. Thus, higher the percentage of single cells at a given time, lower is the cell-cell adhesion. Data show homotypic cell-cell adhesion for PC3M, PC3, HCT-8 (cancer cells arranged in order of decreasing metastatic potential), and MA-104 cells (normal cells) at different time points. The cells with higher malignancy express lower surface cell-cell specific adhesion strength

6 Specific and Non-Specific Adhesion in Cancer Cells

6.3.4

117

Discussion and Conclusions

Metastasis is responsible for the majority of cancer deaths, and cell adhesion plays one of the essential roles in assisting the ability of cancer cells to successfully metastasize [1, 2, 4, 73]. Therefore, understanding the nature of cancer cells’ adhesion during their malignant evolution is essential for preventing metastasis. Unfortunately, it is rather challenging to measure the surface adhesion properties of cancer cells at a single cell scale. In this study, we designed and developed a novel micro-electrical-mechanical-system (MEMS) force sensor to measure the single cell surface adhesion. We found cancer cells show strong non-specific adhesion with the SiO2 probe. The non-specific adhesion strength for highly metastatic cancer cells is higher than the less metastatic cancer cells. For normal, noncancerous cells, the non-specific adhesive activity is undetectable. We also performed the Coulter counter assay to study the specific cell-cell adhesion for cancer cells of different metastatic potential. We found, cells with higher metastatic potential show lower rates of specific cell-cell adhesion. These results are consistent with earlier reports that suggested the more metastatic cancer cells down-regulate their surface adhesion molecule (CAMs), such as E-cadherin, to facilitate the escape from neighbor tumor cells and cell migration [4, 8, 12]. It is known that post-metastatic, circulating cancer cells need to regain their adhesivity in order to invade new tissues. This recovery of adhesivity; however, is unlikely to be specific so as to maximize the adherence of the cells to diverse tissues. Our results suggest that non-specific adhesion activity might be up-regulated purposely to assist invasion and growth of new metastases. In summary, we found that the cancer cells have two types of surface adhesion, the specific adhesion and non-specific adhesion. Cells with higher metastatic potential show higher non-specific surface adhesion. During metastasis, the E-cadherindeficient cancer cells may take advantage of non-specific adhesion to stick with target organs. The normal cells show no non-specific surface adhesion.

6.4 6.4.1

Materials and Methods Cell Culture and Experiments Setup

Three types of living cells were selected in the non-specific adhesivity study. They are metastatic human colon adenocarcinoma HCT-8 cells, non-metastatic human colon adenocarcinoma Caco-2 cells [84, 85], and non-malignant monkey kidney epithelial MA-104 cells. Four types of living cells were selected in the specific cellcell adhesivity study. They are prostate cancer cells PC3, the post-metastatic cancer cells PC3M, the metastatic human colon carcinoma HCT-8 cells and non-metastatic monkey kidney epithelial MA-104 cells. All cells tested in the present study were purchased from American Type Culture Collection (ATCC). They were cultured in

118

X. Tang et al.

normal 35-mm Petri dishes with culture medium consisting of 90% Dulbecco’s Modified Eagle Medium (DMEM), 10% Fetal Bovine Serum (FBS) and 1% Penicillin-Streptomycin (PenStrip) in a cell-culture incubator with 37 C temperature, 5% CO2, and sufficient humidity. The humidity in incubator is created by full interior-tank of clean H2O (with 5 mL Peni-Strep antibiotic solution). The 5% CO2 is transmitted into the incubator by blending external 100% CO2 with air, monitored by a unique InfraRad (IR) CO2 sensor. During the experiment 37 C heat plate was used to maintain the psychological temperature of cells in Petri dishes. The experiments setup is schematically shown in Fig. 6.1.

6.4.2

The Coulter Counter Cell-Cell Adhesion Measurement

The cells were harvested and individualized by trypsin/EDTA treatment and were restored in complete culture medium containing serum to neutralize residual trypsin. The cell suspensions were placed in 17 100 mm capped polypropylene tubes (Falcon No.: 352059) and were rotated end over end at 7–8 revolutions per minute in a Labquake shaker (Barnstead/Thermolyne Model No.: 41510) for 1 h at 37 C to allow recovery of any surface cell adhesion molecules (CAMs) or other proteins. The recovery of CAMs is determined by the measurement of remaining single and non-aggregated cell percentage in Coulter counter. The 1 h preincubation is sufficient, since the cells in suspension regain 40% of adhesion within 1 h and 80% within 3 h, and the rest of the adhesion takes about 18 h [86]. Therefore, the rate of adhesion is best measured within the first 3 h of suspension. Portions of the pre-incubated cells (0.3 mL, approximately 5 105 cells) were placed in flat bottom three dram shell vials (Fisher catalog No.: 0333926D) and rotated in a gyratory water bath shaker (G-76, New Brunswick) at 12 rpm at 37 C for 5, 10, 20, 40, 60, 80, 100 and 120 min. At the end of each time period, cells were diluted with 8 mL 0.9% saline and placed on ice to stop further cell aggregation. The number of single cells present at each time point was measured in the Coulter counter as described elsewhere [77]. Acknowledgements The work was supported by NSF ECCS 07-25831.

References 1. Chambers AF, Groom AC, and MacDonald IC (2002) Dissemination and growth of cancer cells in metastatic sites. Nature Reviews 2:563–572 2. Weinberg RA (2006) The Biology of Cancer. Garland Science, New York 3. Sahai E (2007) Illuminating the metastatic process. Nature Reviews 7:737–749 4. Bissell MJ and Radisky D (2001) Putting tumours in context. Nature Reviews Cancer 1:46–54 5. Liotta LA and Kohn E (2004) Anoikis: cancer and the homeless cell. Nature 430:973–974

6 Specific and Non-Specific Adhesion in Cancer Cells

119

6. Oppenheimer SB (2006) Cellular basis of cancer metastasis: a review of fundamentals and new advances. Acta Histochemica 108:327–334 7. Sahai E (2005) Mechanisms of cancer cell invasion. Current Opinion in Genetics & Development 15:87–96 8. Birchmeier W and Behrens J (1994) Cadherin expression in carcinomas: role in the formation of cell junctions and the prevention of invasiveness. Biochimica et Biophysica Acta 1198:11–26 9. Lodish HA et al (2007) Molecular Cell Biology. W. H. Freeman, New York 10. Bajpai S et al (2009) Loss of alpha-catenin decreases the strength of single e-cadherin bonds between human cancer cells. The Journal of Biological Chemistry 284:18252–18259 11. Bajpai S et al (2008) Alpha-catenin mediates initial E-cadherin-dependent cell– cell recognition and subsequent bond strengthening. Proceedings of the National Academy of Science 105:18331–18336 12. Vermeulen SJ et al (1999) The alpha-E-catenin gene (CTNNA1) acts as an invasion-suppressor gene in human colon cancer cells. Oncogene 18:905–915 13. Patel SD et al (2003) Cadherin mediated cell-cell adhesion: sticking together as a family. Current Opinion in Structural Biology 13:690–698 14. Coman DR (1944) Decreased mutual adhesiveness, a property of cells from squamous cell carcinomas. Cancer Research 4:625–629 15. Coman DR (1953) Mechanisms responsible for the origin and distribution of blood-borne tumor metastases: a review. Cancer Research 13:397–404 16. Wolf K et al (2007) Multi-step pericellular proteolysis controls the transition from individual to collective cancer cell invasion. Nature Cell Biology 9:893–904 17. Friedl P and Gilmour D (2009) Collective cell migration in morphogenesis, regeneration and cancer. Nature Reviews 10:445–457 18. Wilson S et al (2005) The membrane-anchored serine protease, TMPRSS2, activates PAR-2 in prostate cancer cells. Biochemistry Journal 388:967–972 19. Mazzocca A et al (2005) A secreted form of ADAM9 promotes carcinoma invasion through tumor-stromal interactions. Cancer Research 65:4728–4738 20. Oppenheimer SB (2004) Cancer, a Biological and Clinical Introduction. Pearson Education, Upper Saddle River 21. Wang HB et al (2000) Substrate flexibility regulates growth and apoptosis of normal but not transformed cells. American Journal of Physiology – Cell Physiology 279:1345–1350 22. Kumar S and Weaver VM (2009) Mechanics, malignancy, and metastasis: the force journey of a tumor cell. Cancer Metastasis Review 28:113–127 23. Mierke CT et al (2008) Contractile forces in tumor cell migration. European Journal of Cell Biology 87:669–676 24. Kokkinos MI (2007) Vimentin and epithelial-mesenchymal transition in human breast cancer – observations in vitro and in vivo. Cells Tissues Organs 185:191–203 25. Pagan R (1996) Vimentin filaments follow the preexisting cytokeratin network during epithelial-mesenchymal transition of cultured neonatal rat hepatocytes. Experimental Cell Research 222:333–344 26. Willipinski-Stapelfeldt B (2005) Changes in cytoskeletal protein composition indicative of an epithelial-mesenchymal transition in human micrometastatic and primary breast carcinoma cells. Clinical Cancer Research 11:8006–8014 27. Gao CF et al (2005) Proliferation and invasion: plasticity in tumor cells. Proceedings of the National Academy of Science 102:10528–10533 28. Jiang WG et al (2006) Expression of autocrine motility factor (AMF) and its receptor, AMFR, in human breast cancer. Journal of Histochemistry & Cytochemistry 54:231–241 29. Xue C (2006) Epidermal growth factor receptor overexpression results in increased tumor cell motility in vivo coordinately with enhanced intravasation and metastasis. Cancer Research 66:192–197 30. Wang W et al (2005). Tumor cells caught in the act of invading: their strategy for enhanced cell motility. Trends in Cell Biology 15:138–145

120

X. Tang et al.

31. Hooper S (2006) Tumor cell migration in three dimensions. Methods in Enzymology 406:625–643 32. Pinner S and Sahai E (2008) PDK1 regulates cancer cell motility by antagonising inhibition of ROCK1 by RhoE. Nature Cell Biology 10:127–137 33. Mierke CT (2008) Breakdown of the endothelial barrier function in tumor cell transmigration. Biophysical Journal 94:2832–2846 34. Boyd N (1998) Mammographic densities and breast cancer risk. Cancer Epidemiol Biomarkers Preview 7:1133–1144 35. Boyd N (2001) Mammographic densities as a marker of human breast cancer risk and their use in chemoprevention. Current Oncology Reports 3:314–321 36. Provenzano PP et al (2006) Collagen reorganization at the tumor-stromal interface facilitates local invasion. BMC Medicine 4:1–16 37. Condeelis J and Pollard JW (2006) Macrophages: obligate partners for tumor cell migration, invasion, and metastasis. Cell 124:263–266 38. Pollard JW (2004) Tumour-educated macrophages promote tumour progression and metastasis. Nature Reviews 4:71–78 39. Paget S (1989) The distribution of secondary growths in cancer of the breast. Cancer Metastasis Review 8:98–101 40. Poste G and Paruch L (1989) Stephen Paget (1855–1926): a retrospective. Cancer Metastasis Review 8:93–97 41. Kaplan RN (2005) VEGFR1-positive haematopoietic bone marrow progenitors initiate the pre-metastatic niche. Nature 438:820–827 42. Psaila B and Lyden D (2009) The metastatic niche: adapting the foreign soil. Nature Reviews 9:285–293 43. Baker EL (2009) Extracellular matrix stiffness and architecture govern intracellular rheology in cancer. Biophysical Journal 97:1013–1021 44. Paszek MJ (2005) Tensional homeostasis and the malignant phenotype. Cancer Cell 8:241–254 45. Dalton W and Jove R (1999) Drug resistance in multiple myeloma: approaches to circumvention. Seminars in Oncology 26:23–27 46. Sehgal I et al (2006) Prostate cancer cells show elevated urokinase receptor in a mouse model of metastasis. Cancer Cell International 6:1–9 47. Ingber DE (1998) In search of cellular control: signal transduction in context. Journal of Cell Biochemistry 30:232–237 48. Chicurel ME (1998) Cellular control lies in the balance of forces. Current Opinion in Cell Biology 10:232–239 49. Essen D (1997) A tension-based theory of morphogenesis and compact wiring in the central nervous system. Nature 385:313–318 50. Sukharev SI (1993) Two types of mechanosensitive channels in the Escherichia coli cell envelope: solubilization and functional reconstitution. Biophysical Journal 65:177–183 51. Vogel V and Sheetz M (2006) Local force and geometry sensing regulate cell functions. Nature Reviews 7:265–275 52. Engler AJ et al (2006) Matrix elasticity directs stem cell lineage specification. Cell 126:677–689 53. Chen CS (1997) Geometric control of cell life and death. Science 276:1425–1428 54. Janmey PA and McCulloch CA (2007) Cell mechanics: integrating cell responses to mechanical stimuli. Annual Review of Biomedical Engineering 9:1–34 55. Pickett-Heaps JD et al (1997) Traction fibre: toward a “tensegral” model of the spindle. Cell Motility and Cytoskeleton 37:1–6 56. Paszek MJ (2009) Integrin clustering is driven by mechanical resistance from the glycocalyx and the substrate. PLOS Computational Biology 5:1–16 57. Palecek SP (1997) Integrin-ligand binding properties govern cell migration speed through cellsubstratum adhesiveness. Nature 385:537–540 58. Keely P (1998) Integrins and GTPases in tumour cell growth, motility and invasion. Trends in Cell Biology 8:101–106

6 Specific and Non-Specific Adhesion in Cancer Cells

121

59. Assoian RK and Klein EA (2008) Growth control by intracellular tension and extracellular stiffness. Trends of Cell Biology 18:347–352 60. Walker JL (2005) Regulation of growth factor signaling and cell cycle progression by cell adhesion and adhesion-dependent changes in cellular tension. Cytokine & Growth Factor Reviews 16:395–405 61. Klein EA (2007) Cell adhesion, cellular tension, and cell cycle control. In D.A Cheresh and S.P. Colowick (eds.), Methods in Enzymology. Elsevier 62. Paszek MJ and Weaver VM (2004) The tension mounts: mechanics meets morphogenesis and malignancy. Journal of Mammary Gland Biology and Neoplasia 9:325–342 63. Levental KR (2009) Matrix crosslinking forces tumor progression by enhancing integrin signaling. Cell 139:891–906 64. Yamaguchi H (2005) Molecular mechanisms of invadopodium formation: the role of the N-WASPArp2/3 complex pathway and cofilin. Journal of Cell Biochemistry 168:441–452 65. Yamaguchi H (2006) Invadopodia and podosomes in tumor invasion. European Journal of Cell Biology 85:213–218 66. Yamaguchi H and Condeelis J (2007) Regulation of the actin cytoskeleton in cancer cell migration and invasion. Biochimica et Biophysica Acta 1773:642–652 67. Collin O et al (2008) Self-organized podosomes are dynamic mechanosensors. Current Biology 18:1288–1294 68. Goede V et al (1999) Induction of inflammatory angiogenesis by monocyte chemoattractant protein-1. International Journal of Cancer 82:765–770 69. Leek RD et al (1996) Association of macrophage infiltration with angiogenesis and prognosis in invasive breast carcinoma. Cancer Research 56:4625–4629 70. Chaudhuri O et al (2009) Combined atomic force microscopy and side-view optical imaging for mechanical studies of cells. Nature Methods 6:383–388 71. Choy JL et al (2007) Differential force microscope for long time-scale biophysical measurements. Review of Scientific Instruments 78:0437111–0437116 72. Panorchan P (2005) Single-molecule analysis of cadherin-mediated cell-cell adhesion. Journal of Cell Science 119:66–74 73. Bacac M and Stamenkovic I (2008) Metastatic cancer cell. The Annual Review of Pathology: Mechanisms of Disease 3:221–247 74. Yang S and Saif MT (2007) Force response and actin remodeling (agglomeration) in fibroblasts due to lateral indentation. Acta Biomaterialia 3:77–87 75. Yang S and Saif MT (2005) Micromachined force sensors for the study of cell mechanics. Review of Scientific Instruments 76: 44301 76. Yang S (2007) Micromachined sensors and their applications in the study of mechanical response of single living cells. In Mechanical Science and Engineering. University of Illinois at Urbana-Champaign, Urbana, p. 106 77. Kuhlenschmidt MS et al (1982) Studies on the intercellular adhesio of rat and chicken hepatocytes, conditions affecting cell-cell specificity. The Journal of Biological Chemistry 257:3157–3164 78. Lee MH (2005) The effect of non-specific interactions on cellular adhesion using model surfaces. Biomaterial 26:1721–1730 79. Xia Z (1993) Kinetics of specific and nonspecific adhesion of red blood cells on glass. Biophysical Journal 65:1073–1083 80. Busscher HJ (2008) Specific molecular recognition and nonspecific contributions to bacterial interaction forces. Applied and Environmental Microbiology 74:2559–2564 81. Curtis AG (1970) On the occurrence of specific adhesion between cells. Journal of Embryological Experimental Morphogenesis 23:253–272 82. Spits H (1986) Alloantigen Recognition is preceded by nonspecific adhesion of cytotoxic T cells and target cells. Science 18:403–405 83. Siechen S (2009) Mechanical tension contributes to clustering of neurotransmitter vesicles at presynaptic terminals. Proceedings of the National Academy of Science 106:12611–12616

122

X. Tang et al.

84. Hoorde VL (2000) Induction of invasion in vivo of alpha-catenin-positive HCT-8 human colon-cancer cells. International Journal of Cancer 88:751–758 85. Hamada K (2008) Liver metastasis models of colon cancer for evaluation of drug efficacy using NOD/Shi-scid IL2Rgammanull (NOG) mice. International Journal of Oncology 32:153–159 86. Bosco D (2007) Differential expression of E-cadherin at the surface of rat b-cells as a marker of functional heterogeneity. Journal of Endocrinology 194:21–29

Chapter 7

Systems Biology of Tumor Cell Migration in 3D: Protein Signaling Jaya Srivastava and Muhammad H. Zaman

This chapter is part of Section III: Mechano-pathology of Disease

Abstract Without efficient communication, a cell’s ability to survive in a system would be improbable. Moreover, the efficiency with which cancer cells metastasize is attributed to the proficiency with which tumorigenic cells integrate signals from both their surrounding environment and intracellular processes. The mechanosensing ability of a cell is such a process that impacts cellular differentiation, migration, proliferation, and apoptosis. The migratory ability of a cell requires rearrangement of cellular actin scaffolds for the formation of protrusive structures, such as pseudopodia, and secretion of proteolytic enzymes that degrade the extracellular environment. Adhesive structures, such as cadherins and integrins, are tightly regulated processes to efficiently mediate cell migration and apoptosis. In order to properly investigate the behavior of molecular level interactions, scientific research has begun to probe into such dynamic processes at the systems level. This chapter explores the signaling cascades associated with cancer cell migration and those known to be influenced by the mechanical variations of the extracellular matrix. Keywords Metastasis Cell migration Extracellular matrix (ECM) Signaling Adhesion Matrix metalloproteinases (MMPs) 3D

The reductionist approach has successfully identified most of the components and many interactions, but, unfortunately, offers no convincing concepts and methods to comprehend how system properties emerge [1].

M.H. Zaman (*) Department of Biomedical Engineering, Boston University, Boston, MA 02215, USA e-mail: [email protected]

A. Wagoner Johnson and Brendan A.C. Harley (eds.), Mechanobiology of Cell-Cell and Cell-Matrix Interactions, DOI 10.1007/978-1-4419-8083-0_7, # Springer Science+Business Media, LLC 2011

123

124

7.1

J. Srivastava and M.H. Zaman

Introduction

Within a burgeoning system, whether it be a bustling city population or the reproductive cycle of an organism, there is a multitude of ongoing processes that continually ensure functionality and progression of the particular system. In the realm of biological sciences, the approach of systems biology seeks to integrate the myriad of signaling processes that occur within an organism to better understand the variables that impact its function. Signals are processed by cells and manifested into a variety of physical processes, including the cell division and migration. While investigating interactions at a molecular level is integral to understanding signaling events that mediate cell migration, current scientific research has turned to clarifying such dynamic processes at the systems level. In short, the systems biology approach investigates a biological system by perturbing individual components and observing the resultant effects on the entire system. These perturbations can be conducted at genetic, chemical, biological, or mechanical levels, and responses to such perturbations may be monitored on multiple strata that can range from observing fluctuations in protein and gene signaling pathways to population responses [2]. The systems biology approach to investigating component behavior within a system must consider a number of technical challenges. These difficulties include: (1) the identification and quantification of system-wide components on a molecular level (i.e., mRNA and metabolites); (2) the experimental identification of component interactions; (3) the computational interpretation of a component’s structure, conformation, and quantity within a specific interaction; and (4) integration of experimental data [1]. The systems biology approach occurs in a number of well-ordered steps, each of which considers the individual components of a system within the experimental environment. Briefly, a hypothesis is proposed to describe an observed biological phenomenon. An analytical model, based on existing scientific and mathematical theories, is then developed to predict the observed system behavior. Subsequent biological studies are performed to examine this behavior, and the empirical data obtained from these studies is then compared with results predicted by the analytical model. Based on gaps between the model predictions and empirical data, the model is updated, and additional biological studies may be performed to further examine or validate the model. Once the analytical model is finalized, it may be utilized to propose a new theory that describes the examined system behavior (Fig. 7.1). Given the benefits of a holistic approach of systems biology, how does this approach practically contribute to the future of science? Perhaps the most well known integration of systems biology is exemplified by the Human Genome Project. The information gathered from this project provides a catalog of the genomic sequences found in the Human Genome – but what can be done with such information? One such use is that investigating the impact of environmental stimuli on gene regulation correlates the impact of such interactions on the surrounding ecological system – the same type of studies carried out in this example encompass the basis of systems biology [2].

7 Systems Biology of 3D Tumor Cell Migration

125

Fig. 7.1 The systems biology approach. Schematic of a probable working model used in systems biology

By investigating biological events on a systems level, the pharmaceutical and clinical fields can utilize such results to better tune drug discovery and research approaches. For example, by applying a systems-level approach to cellular regulation cascades, therapeutics can be more aptly designed for efficiency and limiting side effects [3]. The application of a systems biology approach in understanding how intertwined molecular-level networks impact a macro-environment has renewed an exciting frontier within biological sciences. The forefront of the biological sciences lies at the systems-level and has the potential to revolutionize scientific research. This chapter outlines the importance of the cellular environment as it affects various cellular processes involved in cancers. The signaling cascades utilized by cells for tumor cell mestastasis are complex as they integrate communications from distant intracellular processes and extracellular stimulation. The studies that have been performed to unveil the contributions of independent and collective variables on cancer cell migration have not only provided invaluable information, but also show that such cellular actions are extremely sophisticated. Systems-level investigations probing into cellular signaling utilized in tumor cell metastsis, such as those used for cellular proliferation, adhesion, environmental remodeling, cytoskeletal rearrangement, and apoptosis, are presented.

7.1.1

Systems Biology: An Applied Approach for Cancer Research

Research within the cancer field has concentrated on identifying the proteins, interactions, and discrepancies that exist between healthy and cancerous cells [4, 5]. At a molecular level, multiple signaling pathways are known to exist, many of which cross paths to elicit subsequent signaling effects within the pathway circuitry [6]. Not only is cross-talk between signaling pathways difficult to predict, but the multitude of molecular effectors that influence these regulatory circuits also complicates the ability to anticipate cellular behavior.

126

J. Srivastava and M.H. Zaman

Over the past century, the field of cancer research has identified key mutations that confer either tumor suppressor loss of function or oncogenes with gain of function [5, 7]. These discoveries subsequently piloted investigations into the signaling processes affected by such mutations [8]. Though such research has revealed invaluable information regarding cellular signaling, the true impact of biochemical signaling processes must be investigated at a systems-level. The idea that the process of tumorigenesis occurs as a multistep process has motivated investigations that examine the influence of macro-level stimuli on micro-level cancer cell behavior [5, 9]. The capabilities of cancer cells to proliferate and metastasize is summarized in six critical cellular transformations. These transformations allow cells to invade tissue and metastasize, support angiogenesis, evade apoptosis, replicate indefinitely, elude antigrowth signals, and produce and sustain growth signals [7]. Each of these transformations implicates the interruption of normal signal processes exhibited by healthy cells. The aforementioned signatures of cancerous cells merely classify the complex interactions involved in cellular migration and proliferation within a system. Cellular migration contributes to a multitude of physiological processes that include wound healing and tissue growth [10]. The regulation of cell migration is imperative to the function of a system, and the impaired regulation of this process is known to be a driving force in the progression of cancer invasion and metastasis [4, 5]. Research that probes cancer cell migration has progressed to the stage of investigating the regulation of signaling pathways from a spatial perspective that relates them to dynamic processes such as microtubule polymerization, actin reorganization, the assembly of adhesions, and the formation of polar structures [7, 9–11]. Studies performed in 3D in vitro environments have provided valuable information regarding the importance of mechanical perturbations as they impact cellular processes [1, 12–14]. Such investigations provide a system-level perspective of the integral cellular events that have the potential to provide a new understanding of cell behavior. The following sections profile the major determinants that influence cancer cell migration upon perturbation of a system.

7.2

Environmental Impact

Cellular invasion of a tissue involves multiple intracellular activities that are cued by extracellular effectors. The investigation of these interactions often occurs in vitro in the absence of physiological conditions. Though difficult to recapitulate exact conditions, investigations within systems-level cancer research attempt to mimic in vivo-like environmental conditions using 3D in vitro models. The extracellular matrix (ECM) and the basement membrane are composed of polysaccharides and proteins, such as proteoglycans, collagen, laminin, and fibronectin. These components provide structural support to cells and bear environmental mechanical stresses. In vivo, cells are known to migrate through three types of ECM that are often recapitulated in vitro to study cell behavior. These three

7 Systems Biology of 3D Tumor Cell Migration

127

cellular environments consist of loose connective tissue, dense connective tissue, and thin acellular layers of the basement membrane [6]. In a typical epithelial tissue, the tissue layer consists of two parts: one which contains closely-neighboring epithelial cells on a basement membrane and the other that holds an ECM-rich stroma that surrounds lymphatic and blood vessels, immune cells, and fibroblasts [15]. Characteristics of the ECM components influence the ability of structural molecules to bind growth factors, enzymes, and other signaling molecules. These characteristics underscore the importance of including ECM-like mimics in systemlevel studies of cancer cell migration. Concordantly, multiple cancer-cell migration studies have shifted from using 2D strata to utilizing 3D substrates, such as collagen type I or type IV, and Matrigel™, a basement membrane mimic composed of laminin, collagen type IV and proteoglycans [16]. Such studies on various mammary epithelial cell (MEC) lines in 3D environments have shown that expression of key migratory proteins and cellular adhesion molecules, such as integrins and focal contacts, promote tumor initiation and cell migration [12, 14, 17, 18]. In addition to relaying signaling factors, the ECM provides a medium for which cells can exert stresses to promote tissue remodeling, differentiation, and morphogenesis [15]. Proteoglycans, for example, regulate hydration of the ECM and thus influence the contribution of hydrostatic tension to the overall compressive forces that are born by cells. The basement membrane is secreted by epithelial and endothelial cells; it acts not only to separate cells from vascular tissue, but also increases the overall mechanical stiffness of the tissue. The arrangement and density of ECM fibers contributes to cell migration and adhesion by determining matrix pore size and ECM sensitivity to cellularly secreted proteolytic enzymes. The cytoskeletal network allows cells to respond to changes in mechanical properties of the surrounding matrix by varying intracellular tension. In response to changes in cellular tension, the ECM consequently undergoes mechanical remodeling (Fig. 7.2) [15, 18]. Alterations of cellular tension not only influence cellular behavior, but also induce modifications to the mechanical properties of the ECM. These mechanical changes manifest as compressive, shear, or tensile stress [15, 18]. In addition to stress, the ECM experiences changes in strain, or normalized deformation, in response to cell movement. The degree of a material’s deformation in response to a load is determined by its composition and organization [18]. For example, elastic solids easily return to their original state following deformation under stress, whereas viscous fluids exhibit strain and stress as a function of time and resist shear flow. Soft biological tissues are considered viscoelastic materials and deform directly in response to an applied stress [18]. Most importantly, however, the changes in ECM can induce changes of intracellular mechanical properties. Systems-level studies have provided contrasting views of the cellular response to 2D and 3D environments. Investigations utilizing collagen type I as both a 2D substrate and 3D matrix show that the actin cytoskeleton of cells from the PC-3 prostate cancer cell line differs dramatically when cells are cultured on 2D or 3D environments [19]. Additionally, intracellular rheology experiments of such cells demonstrate that the culture environment impacts the intracellular mechanical state such that cells cultured in 3D matrices of increasing stiffness and decreased

128

J. Srivastava and M.H. Zaman

Fig. 7.2 Importance of ECM composition and mechanical forces. The mechanical properties of the extracellular matrix (ECM) are determined by the architecture, degree of fiber cross-linking, and composition of the ECM, which imposes mechanical stresses that are experienced by cells within the ECM. Black arrows indicate the direction of mechanical stresses applied to a cell body; dark gray arrows represent the reaction forces exerted by the cell body. Shear, compressive, and tensile stresses are forms of stress experienced by soft biological tissues. Stress is computed as a force per unit area that supports a load and is typically reported in units of newton per square meter (N/m2), or equivalently, Pascals (Pa)

pore size exhibit increased intracellular effective creep compliance, Je, and decreased relative intracellular stiffness, G’p, whereas the 2D environments showed elevated values for both variables [19]. The viscoelasticity of the ECM is affected by its porosity, stiffness, and material composition, which have been shown to affect cell movement [13]. Experiments conducted in 3D Matrigel™ matrices indicate that the composition of the ECM, namely ligand density and matrix stiffness, impacts the migration speed and proteolytic activity of HT-1080 fibrosarcoma cells and DU-145 prostate carcinoma cell lines. With a smaller matrix pore size due to increased matrix stiffness and ligand density, prostate cancer cell migration was found to be dependent on proteolytic activity of enzymes, particularly matrix metalloproteinases (MMPs), secreted by the cells. Experiments performed in like environments with inhibition of integrin b1 mediated adhesion forces compromise the migration of cells due to a lack of cell signaling via adhesion complexes [13, 20]. In addition, cell migration speed has been shown to have a dependence on cell traction and adhesion forces, which vary with matrix composition and density [13]. Systems-level studies conducted with 3T3 fibroblast cells cultured in acrylamide substrates, show that cells preferentially exhibited accelerated protrusion formation on more rigid 2D substrates; more rigid environments promoted the spreading area of a single cell to 25% more than that observed on less rigid substrates [21]. The cell migratory preference for rigid substrates is referred to as “durotaxis” [21]. During these same studies, cells showed bias of migration, detected by podosome formation, towards the direction of tension of the matrix [21]. The rigidity of the matrix

7 Systems Biology of 3D Tumor Cell Migration

129

not only accelerates cell migration, but also alters cell traction forces that promote angiogenesis, an upregulated feature of many cancers [7, 22]. Studies conducted with human umbilical vein endothelial cells (HUVEC) embedded within 3D fibrin gels suggest that cell traction forces that are dependent on ECM stiffness affect the initiation and maintenance of capillary morphogenesis as fewer capillary-like processes form in stiffer matrices [22]. Collectively, these various studies implicate that cell migration is not only impacted by forces exerted on and exerted by cells, but is also dependent on ECM stiffness and porosity. ECM pore architecture greatly impacts cell migration but also acts as a binding site for a variety of effector molecules and soluble growth factors such as hepatocyte growth factor (HGF), transforming growth factor-b (TGF-b) and vascular endothelial-cell growth factor (VEGF). The porosity of the matrix affects the ability of such growth factors to diffuse through the environment; therefore, the arrangement and assembly of structural components of the ECM influences the local concentration gradients of such molecules [15]. This may also promote cell proliferation. Growth factors tethered within the ECM are often bound in their inert forms; upon cellular stimulation, growth factors are cleaved to their active state, enabling binding and stimulation of signaling pathways in neighboring cells [5]. Through increased local binding of an autocrine growth factor to the matrix, controlled amounts of the signaling molecule can be accessed for more efficient signaling. Conversely, paracrine factors that bind to the matrix can diffuse slowly to adjacent cells to optimize signaling fluxes [15].

7.3

Signaling

Signaling cascades act as information highways to deliver data to the power-center of the cell so that information obtained from other parts of the cell can be integrated into an executable order. A systems biology view of this system integrates both mechanical and chemical cues from the surrounding environment. Systems-level studies have shown that ECM proteins regulate cancer cell migration through interactions with transmembrane proteins and extracellular effector molecules [23–25]. In this section, an introduction to the signaling cascades most notoriously associated with cancer-cell behavior are presented. For proteins with conventional abbreviated nomenclature, a reference list of proteins discussed in this chapter is provided in Table 7.1. The signaling processes involved in metastasis are complex; those most investigated in 3D in vitro studies include those transmitted by integrins and Growth factor receptors (GFRs) (Fig. 7.3). Integrins transmit ECM cues intracellularly by recruiting focal adhesion kinase (FAK) and the Src family of kinases (SFKs), such as Src, to the cytoplasmic side of the b-subunit to activate phosphatidylinositol 3-kinase (PI3-K). Actin stress fibers of the cytoskeleton, such as paxillin and talin, recruit FAK to focal contacts upon extracellular stimulation of integrins.

130

J. Srivastava and M.H. Zaman Table 7.1 Table of proteins with conventionally abbreviated names. These are introduced throughout this chapter Abbreviation Name ECM EGF EGFR FAK FGF GAP GEFs HGF IGF-1 JAK Lef MAPK mDIA MLC MLCK MLCP MMP MT-MMP PI3K PDGF ROCK SFK SOCS STAT TGF-b TIMPs uPa VEGF WASP WAVE

Extracellular matrix Epidermal growth factor Epidermal growth factor receptor Focal adhesion kinase Fibroblast growth factor GTPase activating protein Guanine-nucleotide exchange factors Hepatocyte growth factor Insulin-like growth factor I Janus kinase Lymphoid enhancer factor Mitogen-activated protein kinase Mammalian diaphanous Myosin light chain Myosin light chain kinase Myosin light chain phosphatase Matrix metalloproteinase Membrane-type matrix metalloproteinase Phosphatidylinositol 3-kinase Platelet derived growth factor Rho kinase Sre-family kinase Suppressors of cytokine signaling Signal transducers and activators of transcription Transforming growth factor-b Tissue inhibitor of metalloproteinases Urokinase plasminogen activator Vascular endothelial-cell growth factor Wiskott–Aldrich syndrome protein WASP family verprolin homologous protein

GFRs are activated by extracellular stimulation via growth factors (GFs) to activate intracellular signaling cascades. The crosstalk between GFR and integrin signaling ensures that the extracellular signaling mediated by these proteins is optimal. Activation of Src leads to the recruitment of the Crk-DOCK complex that activates Rac. Rac subsequently recruits a number of proteins to ultimately activate NF-kB and the Jun proteins, JUN and JNK. Phosphorylated FAK activates a number of other proteins including growth-factor-receptor bound-2 (Grb-2), which transduces signals to activate the extracellular-signal related kinase/mitogen-activated protein kinase (ERK/MAPK) signaling cascade [12, 26]. The PI3K can also activate the small GTPases Rac, Rho and Cdc42 through mediator proteins such as guaninenucleotide exchange factor (GEF) and phosphatidylinositol-3,4,5-triphosphate (PIP3) to induce changes in the cell’s cytoskeleton, migration, and gene invasion

7 Systems Biology of 3D Tumor Cell Migration

131

Fig. 7.3 Generalized scheme of the signaling cascades relevant to cancer cell migration in 3D environments. Extracellular stimulation of transmembrane receptors such as integrins and growth factor receptors transmit signals into the cell. Intracellular signaling cascades enable the cell to respond in a number of fashions including replication of cellular material, protein synthesis, cell migration, and cell proliferation [12, 18–23]

capacity [27]. The Wiskott–Aldrich syndrome protein (WASP) and Arp 2/3, an actin polymerization protein, contribute to the reorganization of the cytoskeleton to induce cell migration [23–25, 28].

7.3.1

Outside-In Signaling

Growth factors have been shown both to positively and negatively regulate cell migration. The stimulation of migration in two human breast cancer cell lines, MCF-7 and MDA-231, by insulin-like growth factor-I (IGF-I) was shown to be attributed to IGF-I triggered chemotaxis [29]. However, cells derived from the aforementioned cell lines did not elicit a notable migratory response to epidermal growth factor (EGF), platelet derived growth factors (PDGFs), or fibroblast growth factor (FGF), indicating that MCF-7 and MDA-231 migration is responsive to specific growth factors [29]. This notion that cancerous cells, which vary in origin, are responsive to different growth factors has been demonstrated in a number of studies. A variety of squamous cell carcinomas derived from the head and neck show elevated levels of EGF receptor (EGFR) expression, and consequently, increased proliferation; however, cells from glioblastoma, osteoblastoma, and T24 human

132

J. Srivastava and M.H. Zaman

bladder cancer cells that show elevated levels of PDGF and PDGF-like peptides form larger tumor masses [30]. Intracellular signaling processes are often a secondary cascade to the initial extracellular stimulation of transmembrane receptors. The Ras pathway is activated by extracellular signaling molecules such as EGF, FGF, and PDGF growth factors, which permit translocation of Erk to the nucleus where it regulates the activity of transcription factors [12]. This regulation occurs through the activation of the Erk/ MAPK signaling pathway. Erk, a MAPK, is known to regulate cell migration as it regulates the turnover of focal adhesions and membrane protrusions through the activation of myosin light chain kinase (MLCK) [31]. MLCK phosphorylation promotes the disassembly of focal adhesions via the activation of calpains, Ca2+ regulated proteolytic enzymes that are engaged during cell migration. Stimulated by MLCK phosphorylation, calpains degrade cytoskeletal proteins within focal adhesions to accelerate disassembly [31]. The phosphorylation of FAK at Ser910 by Erk is believed to interrupt paxillin-FAK interactions to promote migration. Studies of MECs cultured within 3D collagen type-I matrices showed that the mechanical features of the ECM influence the MAPK pathway to promote cell growth. In environments of higher stiffness, phosphorylation at key residues, FAK (Y397) and (Y925), was shown to increase and co-precipitate with Shc and Grb2, upstream effectors of Ras [12]. Such results suggest that Erk activation is regulated by FAK via the Ras pathway. Thus, an elevated FAK-Rho loop causes increased activity of the Ras-activated Erk/MAPK pathway that promotes a hyper-proliferative cell state [12]. Other systems-level studies of Ras proteins have established important links between protein signaling and cellular migration through matrices. Rab proteins, a class within the subfamily of Ras proteins, are GTPases that are responsible for the cycling of internalized membrane-associated functional groups to the cell surface [32, 33]. Observations in both clinical and murine models of highly metastatic ovarian and breast tumors have shown that Rab25 contributes to the aggressiveness of tumorigenic cells. Such observations, combined with studies implicating the involvement of other Rab proteins in tumorigenicity, implicated a key role of Rab proteins in the progression of cell migration and metastasis [34]. For example, in the absence of Rab11, integrin-dependent cell migration is impaired [35]. Rab4 and Rab11 have been previously characterized as regulators of recycling a host of integrins, including a5b1, a6b1, a6b4, and avb3 [32]. The interaction between a5b1 and fibronectin contribute to a number of adhesion complexes that contribute to the migratory capacity of cells through a 3D environment [36, 37]. Studies using a human ovarian cancer cell line, A2780, established a relationship between Rab25 and the cytoplasmic tail of b1 integrins that promotes the cellular invasion through 3D matrices. Rab25 allows for the localization and maintenance of a5b1 at the leading edge of pseudopodia. The interaction between a5b1 and Rab25 is mediated by the GTP-dependent association of Rab25 and the cytoplasmic tail of integrin b1 [32]. This invasion mechanism of Rab25 is dependent on the fibronectin concentration within the ECM, the ability of Rab25 to interact with the cytoplasmic tail of b1

7 Systems Biology of 3D Tumor Cell Migration

133

integrin, and the presence of a5b1 heterodimers. The ability of Rab25 to localize and recycle a5b1 integrin subunits promotes the formation and disassembly of focal adhesions at the protruding ends of filopodia to accelerate cell migration [32].

7.3.2

Inside-Out Signaling

The complexities of signaling pathways are but highlighted through proteins that appear to have dual roles in the cell. For example, TGF-b has been shown to act as a positive regulator of proliferation and differentiation neoplastic cells, but also as a potent inducer of growth inhibition in non-neoplastic cells, such as epithelial cells [30, 38]. TGF-b activated receptors are known to initiate a signaling effector pathway by phosphorylation of the carboxy-terminus of proteins in the Smad family. A number of Smads, such as Smad3 and Smad4, translocate to the nucleus upon TGF-b activation to serve as transcriptional activators of target genes. Smad6 and Smad7 act as inhibitors to prevent the activation of Smad3 and Smad4. Interestingly, expression of Smad7 is induced by TGF-b, implicating that TGF-b provides itself with a negative-feedback loop. Smad activation by TGF-b allows for interaction with other signaling pathways that influence cell migration, such as the mitogen activated protein kinase (MAPK), Jun, a transcription factor, and p38 MAP kinase [38]. Briefly, the p38 sub-family of the MAPK family of kinases are stimulated by growth factors to promote cell migration. The phosphorylation of p38 allows for downstream signaling of various MAPK proteins, most of which influence actin-remodeling dynamics to promote cell migration [31]. In epithelial cells, TGF-b inhibits cell growth, whereas in tumor cells, perturbation of the TGF-b signaling pathway by epigenetic factors promotes metastasis. For example, activation of the Ras pathway in a particular cancer cell can result in a direct increase of TGF-b and thus attenuate tumorgenic activities such as invasion and migration [38]. The aforementioned pathways are but a few of the many signaling processes that occur within a cell. However, these pathways have been implicated in metastasis. Though 2D studies have revealed valuable information regarding the proteins involved in these signaling pathways, 3D studies are imperative as they provide a clarified idea of cell behavior on at a systems-level.

7.4

Cell Movement

Cellular movement through and around the terrain of animal tissue is an integral component to metastasis. A motile cell compounds signals from the external environment, cellular cytoskeleton, and nucleus. In this section, the regulatory protein complexes involved in cell movement, such as actin rearrangement and podosome formation, in a 3D system are presented.

134

J. Srivastava and M.H. Zaman

The reorganization of the actin cytoskeleton during cellular migration produces intracellular forces that propagate the cell through the ECM [39]. Cell migration is a critical feature of multiple pathological and physiological processes that are triggered by chemoattractants [8]. For example, during wound healing, PDGF interacts with cell surface receptors to trigger intracellular signaling cues that elicit cell migration [39]. The process of actin reorganization is complex and involves a number of processes that allow for the binding of the leading edge of the cell, translocation of the cellular mass, and release of the lagging end of the cell [8, 9, 40]. The actin rearrangements that control these processes are regulated by the Rho family of small GTPases, the WASP family of proteins, cofilins, and cortactin [5, 10, 11, 41–43].

7.4.1

Pseudopodia

During cell migration and invasion into tissues, cells form protrusive structures, such as lamellipodia, invadopodia, and filopodia, which are regulated by actin polymerization and depolymerization. These structures are named based on their functional, morphological, and structural characteristics [11]. Lamellipodia are sheet-like protrusions that form at the leading edge of a migrating cell and are believed to have a key role in the induction of cell migration. Lamellipodia are composed of a dendritic network-like array of actin filaments that likely assist in protrusion formation. In short, lamellipodia attach to a substrate and generate an initiating force to pull the cell mass forward – this event has been observed in carcinoma cells from a mammary tumor [11]. Upon stimulation by a chemoattractant, the carcinoma cells crawl along ECM fibers toward blood vessels of a primary tumor via extension of lamellipodia-like pseudopod extensions at the leading edge of the cell [44]. In contrast, filopodia are thin rod-like bundles of actin which are believed to be sensitive to extracellular chemical cues [10, 11]. The dynamic reorganization of filopodia is believed to increase intracellular pressure at the leading edge of a migrating cell to effectively produce the stiffness necessary to advance the filopodia protrusion through the matrix [10]. Invadopodia form ventral protrusions at the membrane. These protrusions are composed of a variety of active proteins including adhesion molecules, such as N-cadherin, actin regulatory proteins, such as WASP, and proteases with matrix degrading functions, such as MMPs [10, 45–47]. These proteins have been found to be overexpressed in malignant cancer cells, supporting the observations that invadopodia are frequently utilized in highly aggressive carcinoma lines in order to invade and migrate through the tumor stroma for intravasation into blood vessels [11, 12, 27, 42]. Cofilin is an essential regulator of actin disassembly during cell migration. This small protein binds to both monomeric and filamentous forms of actin in order to sever actin filaments. During cancer cell migration, cofilin activation by upstream effectors has been shown to play a key role in establishing cell migration directionality by inducing lamellipodia formation [11]. Additionally, in experiments

7 Systems Biology of 3D Tumor Cell Migration

135

that suppressed cofilin expression via siRNA, the assembly and maintenance of invadopodia were compromised, reducing the migratory capacity of carcinoma cells [43]. Cofilin is negatively regulated by a number of signaling proteins such as LIM1, LIM2, TES1, TES2, and phosphatidylinositol-4,5-bisphosphate (PIP2) [11]. Considering that most metastatic cell lines, including HeLa (cervix), Jurkat (T-lymphoma), and COS1 (kidney), have been found to express elevated levels of cofilin, it can be assumed that the aforementioned negative regulators experience suppression of activity in these carcinomas [11]. The activation of cofilin occurs transiently upon cellular stimulation by EGF. Upon binding of EGF to the cell surface receptor, EGFR, a number of events occur that permit the activation of both cofilin and LIM kinases, allowing for precise control over actin polymerization at the leading edge of the cell such that actin polymerization occurs in a controlled, transient fashion [11, 48]. The action and activation of cofilin is carefully regulated intracellularly; however, it is unknown if cofilin is transiently regulated by ECM factors, such as stiffness. Investigations using a systems-level approach are necessary to elucidate these possibilities.

7.4.2

Actin and Rho GTPases

The Rho family of GTPases, or guanine triphosphate binding proteins, are conformationally regulated via binding of GDP or GTP. The binding of GTP activates these proteins to permit interaction with downstream effector proteins [10, 42]. The ability of Rho GTPases to hydrolyze GTP are catalyzed by GTPase activating proteins (GAPs) and guanine-nucleotide-exchange factors (GEFs) [42]. Of the Rho GTPases, RhoG, Rac, and Cdc42 are required for the protrusion formations of filopodia and lamellipodia. Briefly, Rac and Cdc42 target the WASP family of proteins, which influence Arp2/3 activity, to promote actin polymerization at the leading edge of a migrating cell [23]. In mammalian cells, there are two subfamilies of WASP proteins, WASP and WAVE, or the WASP family verprolin homologous protein. There are two known WASP proteins, WASP and n-WASP, a neural WASP, and three WAVE proteins, WAVE1, WAVE2, and WAVE3 [11]. Regardless of the designation, the WASP family of proteins integrate upstream protein signals, such as those from Cdc42, to elicit activity of Arp2/3, an actin-associated complex that permits actin filament nucleation and branching [10, 11]. In tumorigenic phenotypes, the expression of WASP proteins and Arp2/3 complexes has been implicated in malignancy [11, 49]. Additionally, the presence of both Cdc42 and Rac are required for the migration and invasion of tumorigenic cells in 3D collagen matrices [23, 50]. Cdc42 is active at the leading edge of the cell and regulates cell polarity by regulating where and when lamellipodia form [10]. This particular polarization by Cdc42 is associated with phosphoinositides (PI3Ks), PtdIns(3,4,5)P3, or (PIP3), and PtdIns(3,4)P2, or (PI(3,4)P2), as well as the deregulating phosphatase, PTEN [10]. The PI3K molecules are highly polarized at the leading edge of cells that have been exposed to a chemoattractant, such as

136

J. Srivastava and M.H. Zaman

PDGF, platelet derived growth factor, or EGF. These extracellular factors bind to cell-surface receptors to elicit an intracellular signaling cascade. Proteins within this cascade, such as FAK and RTK, are often overexpressed in carcinogenic cells [23]. Thus, tumorigenic cells utilize extracellular effectors to regulate Cdc42 and PI3K activation in order to stimulate lamellipodia formation and cell migration. Alternatively, Rho proteins are known to contract and promote the assembly of actomyosin fibers to effectively pull the trailing edge of the cell in the forward direction of migration [23]. Actomyosin is a complex of actin and myosin, where one head of a myosin molecule binds to an actin filament so that it is organized towards the barbed end of actin. The other head of myosin binds to a second actin filament to produce a resilient tug-of-war system that elicits a contraction when signaled [51]. The formation of the actomyosin fibers requires the activity of two additional Rho effector molecules, Rho kinase (ROCK) and mammalian diaphanous (mDIA). ROCK promotes the phosphorylation of myosin light-chain protein (MLC) by inhibiting MLC phosphatase (MLCP); this phosphorylation of MLC promotes actomyosin contraction, a crucial event for cell movement [23]. The Rho-ROCK regulation of actin polymerization and activity has too been implicated as having an active role in promoting migration of tumorigenic cells. Namely, the expression and activity of both Rho and ROCK are required for tumor cell migration via amoeboid movement through 3D matrices [23, 27, 52]. Additionally, ROCK expression, independent of Rho activity, has been shown to have a facilitating role in carcinoma cell penetration of the host mesothelium [53]. The regulation of structural proteins is integral in mediating cell migration through the ECM. Cortactin is a cytoskeletal protein that has been observed to be overexpressed in a number of cancers and confers highly vigorous migratory capacity to tumorigenic cells including those of esophageal squamous carcinomas, head and neck squamous carcinoma, breast, gastric, colorectal, and ovarian cancers [41]. Cortactin contains two distinct binding sites, one for Arp 2/3 binding and a second for regulatory binding proteins such as those from the SFKs [41]. The importance of the src-kinase binding domain of cortactin has been implicated through cortactin overexpression studies that increase metastasis of breast cancer, esophageal squamous carcinoma, and hepatocellular carcinoma cells [41]. The Arp 2/3 complex mediates cytoskeletal actin assembly by nucleating branched actin filament networks that confer structural support to the cytoplasm and generate forces involved in cell movement [41, 54]. Via interaction with WASP proteins, cortactin is believed to promote Arp 2/3 nucleation and stabilize branched actin networks. Regarding cell migration, cortactin has been identified as an intermediate regulator of lamellipodia formation induced by growth factors. Cortactin is an imperative factor in propelling cell migration as it promotes cellular adhesion formation at the leading edge of cells and is able to maintain established lamellipodia structures [41, 55]. Cortactin localization to the leading edge of migratory cells occurs immediately prior to the secretion of matrix-degrading proteases, as visualized by live-cell imaging [41, 56]. In particular, trafficking of a number of MMPs relies on cortactin in a several cancers, including head and neck squamous carcinoma cells [57].

7 Systems Biology of 3D Tumor Cell Migration

7.5

137

Cellular Adhesion

Cell–cell and cell–ECM adhesions are important interactions that regulate important processes necessary for cell movement. In this section, the proteins involved with such interactions are analyzed as they contribute to intracellular signaling and cell motility, are analyzed. A notorious trait of cancerous cells is their ability to rapidly turn over cell adhesions. In normal cells, cell–cell junctions are formed between cells via cellsurface receptors, such as cadherin and integrin adhesion molecules. Most notably in epithelial cells, the adherins junctions mechanically connect the cytoplasmic actin cytoskeletons of neighboring cells as well as anchoring cells to the ECM [5]. In order to metastasize, tumorigenic cells must migrate through the ECM, and thus, rapidly assemble and disassemble adhesions such that sufficient mechanical traction is produced to propagate the cell through the ECM without impeding migration.

7.5.1

Cadherins

Cadherins are Ca2+ dependent transmembrane proteins that mediate cell-cell adhesion molecules. The epithelial cadherin, E-cadherin has been observed to be downregulated in a number of epithelial and breast carcinomas to promote tumor invasion. In contrast, N-cadherin, usually expressed in neurons, are shown to be upregulated in tumor cells. Studies conducted with breast carcinoma cells indicate that exogenous N-cadherin expression promotes cancer cell motility and invasiveness in cells with basal expression levels of E-cadherin [58]. The increase in N-cadherin expression promotes adhesion of tumorigenic cells to endothelial cells, which facilitates cell migration. In addition, cellular motility and invasion of the ECM are coordinated in N-cadherin expressing cells with enhanced FGF-2 expression and MMP-9 secretion, which promotes ECM degradation [58]. The cadherins act in complex with catenins to efficiently link cadherins to the cytoskeleton and to signaling pathways involved in regulation of transcription factors. a-catenin mediates the linkage of cadherin-catenin complexes via actin binding proteins to the actin filament network, while b-catenin interacts directly with the cytoplasmic domain of cadherins to mediate cell-cell adhesion functions of cadherins [59]. In addition, b-catenin serves a cadherinindependent function as a signal transducer that regulates gene expression. Particularly, the b-catenin pathway can be activated by Wnt proteins, which causes b-catenin to accumulate in the cytoplasm and subsequently translocate to the nucleus where it can activate b-catenin/lymphoid enhancer factor (Lef) dependent gene expression [59]. The Wnt family of proteins are cysteine-rich molecules that are associated with their ability to regulate transformation of MEC lines [60]. Aberrant activation of the Wnt proteins has been implicated in initiating tumor formation and cell migration in a number of cancers, including gastric and breast cancers [60, 61].

138

J. Srivastava and M.H. Zaman

The presence of Wnt proteins is critical to the action of b-catenin; without stimulation by Wnt, the Axin complex phosphorylates and ubiquitinates b-catenin to signal its degradation by the proteasome. However, in the presence of Wnt proteins, Wnt interacts with Frizzled, a transmembrane receptor complexed with LNT, a lipoprotein receptor related protein. The interaction protects b-catenin from proteasomal degradation and allows it to translocate to the nucleus where b-catenin interacts with transcription factor T-cell factor (Tcf) and Lef to stimulate gene expression [59, 60]. A number of human cancers exhibit abnormalities in the regulation of the b-catenin pathway that are implicated in cell migration, however the mechanism of this activation is unclear. Studies investigating the cell migration in rat kidney epithelial cells indicate stimulation of cell motility upon HGF and EGF mediated induction of b-catenin signaling [59]. Additional studies implicate that stable b-catenin expression contribute to tumor formation in nude mice, perhaps identifying a new role for b-catenin in the epithelial-to-mesenchymal cell transition that is imperative for tumorigenicity [59]. Investigations probing the role of Wnt-5a, a non-transforming member of the Wnt family, established a regulatory function for Wnt-5a in the migration and invasion of gastric cancer cells. In cells with suppressed Wnt-5a activity, turnover of paxillin, membrane ruffling, and cell migration were inhibited. Key proteins involved in cell migration, FAK and Rac, required Wnt-5a activation for full function. Though the importance of Wnt-5a activity and b-catenin regulation during cancer cell growth and migration have been identified, the relationship between these pathways remains elusive.

7.5.2

Integrins

Integrins, a family of cell-surface receptors, transmit both chemical and mechanical signals to efficiently mediate adhesion of cells to the ECM, activate intracellular signaling pathways, and organize the cytoskeleton [20, 23]. Upon binding to the ECM, integrins become clustered at adhesion sites to associate with signaling and cytoskeletal complexes. This association promotes the assembly and reorganization of actin filaments, which further promotes integrin–matrix interactions. A unique property of integrins is their ability to relay signals in two directions: signals are transmitted via outside-in signaling when binding of the ECM induces signals for transport into the cell, whereas inside-out signaling transmits intracellular signals to the exterior of the cell [28]. The multi-faceted regulatory role of this family of transmembrane proteins implicates the importance of integrins in cell migration and proliferation. Integrins are composed of two transmembrane units, an a and a b unit. To date, 18a and 8b subunits have been identified in mammals, which combine to form 24 unique integrins that bind different types of ECM ligands. The integration of signaling between the ECM and the cell, as mediated by integrins, imparts strict control over cell proliferation through the activation of signaling pathways that are necessary for tumor cell migration [23].

7 Systems Biology of 3D Tumor Cell Migration

7.5.3

139

Focal Adhesion Kinase

FAK is one of the most commonly integrin-activated intracellular membrane-associated proteins. Upon activation, FAK subsequently activates proteins within the src-family kinases (SFKs), which are associated with a multitude of signaling pathways associated with migration, survival, and transcription [23]. The adhesion-dependent and growth factor-dependent signals transmitted by FAK to the cell interior have been extensively studied with respect to their influence on cell migration. However, investigations of FAK regulation at a systems biology level have recently implicated its multidimensional role in providing a migratory advantage to tumorigenic cells [62]. The dexterity of FAK allows this tyrosine kinase to act as an adapter for multiple signaling pathways. In one study, growth factor receptor and integrin mediated signaling were shown to be linked by FAK. Particularly, integrin, EGF, and PDGF-stimulated cell migration require signaling by activated FAK. The presence of PDGF and subsequent activation of PDGFR, the corresponding receptor, are necessary for FAK interaction with Src tyrosine kinases [63]. A particularly important residue for FAK activity is the FAK tyrosine (Y) 397 site. Numerous studies have shown that autophosphorylation of FAK Y397 is a critical mechanistic connector for signaling processes initiated by growth factors, integrins, and other extracellular effectors such as protein tyrosine phosphatases [23, 26, 63]. Upon autophosphorylation at FAK Y397, additional FAK residues become available for activation by Src and other tyrosine kinases. These FAK tyrosine residues include 407, 576, 577, 861, and 925 [62]. The expression and enzymatic activity of FAK have been previously identified as key regulatory components of cell migration and proliferation. FAK has been implicated as a mediator of migratory invasion of the ECM in both breast and prostate cancer cell models [14, 64]. In prostate cancer cell line models, inhibition of FAK expression by a FAK and src-kinase specific inhibitor greatly impaired cell migration through a fibronectin substrate. Additionally, in highly metastatic prostate cancer cell lines, DU-145 and PC-3, phosphorylated FAK Y397 levels were significantly higher than that observed in a poorly metastatic prostate cancer cell line, LNCaP [64]. Systems level studies have been performed to correlate matrix density with gene expression. Such studies show that breast cancer cells and fibroblasts cultured in high density collagen matrices induce an increased level of phosphorylation of FAK Y397 as compared to cells cultured in low density collagen matrices. Thus, increased activation of FAK leads to an elevated signaling loop that connects focal adhesions to Rho and ultimately to the MAPK signaling pathway. These observations implicate an important ability of the cell to sense the microenvironment and elicit corresponding chemical cues for cellular proliferation and motility [12]. FAK contains additional activating phosphorylation sites that are known to be involved in cellular processes, such as motility. Phosphorylation of FAK Y576 and Y577 by Src stimulates FAK such that FAK catalytic activity is maximized. Tyrosine 925 is a known Src-specific binding site, which upon phosphorylation, enables the binding of GRB2 to FAK Y925. GRB2 subsequently

140

J. Srivastava and M.H. Zaman

signals a number of other proteins, including those involved in the MAPK pathway, to promote cell migration and epithelial to mesenchymal transition, a hallmark event of tumorigenic cells [26, 62]. A number of studies have been conducted to investigate additional involvement of Y925 in cell adhesion and migration. During focal contact turnover, FAK Y925 activation and consequent GRB2 binding are thought to promote the dissociation of FAK from focal contacts by displacing paxillin. The dissociation of FAK then allows for the turnover of focal contacts [26]. The mechanism responsible for this action has yet to be discovered. The role of FAK and its associated signaling tyrosine kinases have identified its importance in signaling pathways that mediate cellular migration, growth, and survival. The remodeling of focal adhesions are regulated by FAK in a variety of mechanisms. Through systems-level studies, the links between FAK and matrix components are slowly being established. For example, the organization of fibronectin and fibrillar adhesions of the matrix are known to be regulated by FAK. Cells that lack FAK in focal adhesions have limited ability to translocate complexes that are necessary for the formation of fibrillar adhesions [65]. These complexes include focal adhesion proteins and integrin-bound fibronectin, which are translocated along actin filaments to form fibrillar adhesions in the presence of FAK. FAK plays a key regulatory role in the organization of fibronectin within a matrix [65]. Using the observation that the vascular organization of epithelial cells in FAK deficient mice is impaired, cellular level studies were conducted to show that FAK localization to focal adhesions and FAK kinase activity at focal adhesions are required for adequate formation of fibrillar adhesions, fibronectin organization, and allocation. In addition, the organization of actins stress fibers within the matrix are dependent on FAK [65]. FAK also influences actin remodeling through its interaction through Rho GTPases, which stimulate lamellipodia and filopodia to enable forward movement of the cell via membrane protrusion and polarization [62].

7.5.3.1

Janus Kinase/Signal Transducers and Activators of Transcription

From systems-level studies, it is clear that FAK contributes to a number of processes in the cell. However, there are other signaling cascades that transduce signals external of the cell to modulate gene expression and regulate other ongoing intracellular signaling. The JAK/STAT pathway is dynamic in its ability to target genes and regulate proteins, such as FAK. For efficient cell migration and proliferation, target genes within a cell must be activated for transcription. The JAK/STAT pathway has been established as a key signal transduction pathway that transmits stimuli from extracellular moieties and transmembrane receptors to the cell nucleus in order to regulate gene expression. Upon binding of growth factors or cytokines to membrane-associated receptors, the Janus kinase (JAK), a nonreceptor tyrosine kinase, is localized to the cytoplasmic face of the receptor where it creates sites for phosphotyrosine protein binding. JAK proteins subsequently phosphorylate any localized signal transducers and activators of transcription (STAT) at tyrosine residues within the SH2 domains of

7 Systems Biology of 3D Tumor Cell Migration

141

STAT, causing the STAT transcription factors to dimerize, translocate to the nucleus, and initiate transcription of target genes [5, 66]. The JAK/STAT signaling cascade is regulated by cellular suppressors, namely the suppressors of cytokine signaling (SOCS), which interfere with JAK binding to membrane-associated receptors or with STAT phosphorylation to effectively terminate signal transduction initiated by growth factors or cytokines [67]. The SOCS-1 and SOCS-3 genes have been implicated in inhibition of cell growth and differentiation as they promote the degradation of FAK in healthy kidney cells [66]. However, investigations of hepatocellular carcinoma cell lines identified a suppressor of SOCS-3. Such suppression promotes the phosphorylation of STAT3 to encourage cell growth and increases the phosphorylation of FAK, yielding enhanced cell migration [66]. These studies highlight the complexities in identifying clear relationships between integrated signaling cascades and the stimuli residing in the ECM with respect to cell migration. It is clear that cell migration is an extremely complex process; unveiling a clear mechanism for such actions awaits future systems-level studies.

7.6

Proteases

Cellular migration through the ECM is reliant on a number of processes that occur both intra- and extracellularly. Tumorigenic cells are efficient in sensing the architecture of the ECM scaffold to elicit intracellular signaling so that they can overcome the migratory restraints of the matrix [68]. In this section, two classes of proteases, MMPs and urokinase plasminogen activator (uPa), that active in the remodeling of the ECM are introduced.

7.6.1

Matrix Metalloproteinases

In almost every type of human cancer, MMPs are upregulated [69]. The family of MMPs are Zn2+ and Ca2+ dependent endopeptidases that are known to degrade ECM fibrils to permit efficient migration of cells. MMPs are synthesized by cells as pro-MMPs, or inactive zymogens, which are activated extracellularly by serine proteases or other active MMPs [47, 69]. The MMPs are divided into subfamilies based on their specificity for ECM ligands (Table 7.2) [69, 70]. The MMP subfamilies include collagenases, gelatinases, matrilysins, and stromelysins. More recently, MMPs have also been divided into classes based on their structure; thus far, three membrane-type MMPs (MT-MMPs) and five secreted MMPs classes have been identified [69]. MMP activity is regulated by tissue inhibitors of metalloproteinases (TIMPs) that are also secreted by cells. Increased levels of TIMPs have been implicated in poor prognosis of cancer patients as they likely contribute to matrix remodeling during tumor cell migration. Of the four TIMPs identified, TIMP-1 and TIMP-2 are

142

J. Srivastava and M.H. Zaman

Table 7.2 Matrix metalloproteinase families MMP gene Name

Classification based on substrate specificity

MMP-1 Interstitial collagenase Collagenase MMP-2 Gelatinase A (72 kDA) Gelatinase MMP-3 Stromelysin 1 Stromelysin MMP-8 Neutrophil collagenase Collagenase MMP-9 Gelatinase B (92 kDA) Gelatinase MMP-10 Stromelysin 2 Stromelysin MMP-11 Stromelysin 3 Stromelysin MMP-13 Collagenase 3 Collagenase MMP-14 MT1-MMP Membrane associated MMP-15 MT2-MMP Membrane associated MMP-16 MT3-MMP Membrane associated MMP-17 MT4-MMP Membrane associated Matrix metalloproteinases (MMPs) have been classified both by their substrate specificity and by their structure. The most studied MMPs in regards to cancer cell migration, are listed above based on their substrate specificity [59, 61]

implicated in aiding tumor cells to evade apoptosis and, conversely, inhibiting tumor growth. Additional evidence constructs a role for TIMP-1 in promoting angiogenesis via upregulation of VEGF, a key factor involved in angiogenesis [69, 71]. TIMP-3 and TIMP-4 are believed to serve as markers for cell differentiation and as active regulators of ECM homeostasis [71]. 7.6.1.1

MMP Activity During Cell Migration

System-levels studies of these MMPs are needed to assign function during cell migration. In addition to promoting cell migration, MMPs are also instrumental in indirect acceleration of migration by cleaving inactive growth factor precursors from cells and by activating survival factors [68]. For example, the cleaving of insulin growth factor-binding protein (IGF-BP) by MMP increases the cellular concentration of free IGF, which is associated with tumor cell proliferation and migration [5, 29, 69]. As previously mentioned, cell-adhesion molecules such as cadherins and integrins are substrates of MMPs. The cleavage of adhesion molecules is directly correlated to the epithelial to mesenchymal cell transition as well as the migration and invasion of the ECM by cells of most identified cancer types [5, 10, 58, 59, 69]. MMP-3 and -7 have been identified as active cleaving enzymes of E-cadherins. Fragmented cadherins are associated with increasing tumor cell migration via interference with full length cadherins or by triggering the hallmark epithelial-to-mesenchymal transition of highly migratory tumorigenic cells [72, 73]. In a number of tumors, MMPs are expressed at aberrantly high levels as compared to normal tissue [69]. Systems-level studies have identified the contribution of MMPs to migration of cancerous cells. Migration is the first step of cancer cell invasion, which requires the detachment of cancer cells from the matrix and

7 Systems Biology of 3D Tumor Cell Migration

143

from neighboring cells. Particular MMPs are secreted by cancer cells, such as MMP-7, whereas others are expressed by stromal cells of the tumor. Stromal cells, which often consist of endothelial cells, inflammatory cells, fibroblasts, and myofibroblasts, secrete MMP-2 and -9 in response to ECM and intracellular stimuli [69]. Such signals include secretion of growth factors, interleukins, and interferons by tumorigenic cells that act as paracrine stimulators of stromal cells to promote MMP secretion. Particularly, xenograft studies using animal models show that highly malignant cells have reduced migratory capacity when MMP-9 expression is reduced whereas normally benign cells adopt an invasive phenotype when MMP expression is upregulated [74]. To initiate migration and invasion, cells produce adhesions and alter their morphology to produce a leading edge, often characterized by the formation of invadopodia. The recruitment of MMPs to the site of invadopodia formation is necessary for surface protrusions to invade the surrounding environment as the proteases enable local degradation of the ECM [69]. The presence of cortactin in invodopodia implicated its association with MMPs. In 3D studies, the expression levels of MMP-2, -9, and plasma level expression of MT1-MMP were directly correlated with cortactin expression levels in cells cultured on fibronectin matrices. In these MMP knockdown studies performed on cells of head and neck squamous cell carcinoma, the migratory capacity of the cells was inhibited, much like that observed in cortactin knockdown cells, implicating the coupling of MMP expression with actin assembly and disassembly during ECM degradation and cell migration [57]. Multiple studies have identified MMP-2, -9 and -14 as proteases localized to sites of invadopodia formation; the interaction of MMPs with adhesion proteins, such as integrins, is known to occur, but the mechanism of interaction remains unclear in system-level studies [75, 76]. These proteases are not limited in function, rather, they appear in multiple activating events associated with cell migration. MMP-2 and -14 have been identified as cleaving enzymes of laminin-5 [77]. This cleaving event exposes cryptic sites of laminin that are affiliated activators of cell motility [77, 78]. The importance of the event is supported by evidence of MMP-14 co-localization with laminin in a number of human cancers. MMP-14 interacts with CD44, a hyaluronan receptor, so that its extracellular domain is cleaved, an event that promotes cancer cell migration. CD44 has been identified as a binding partner of MMP-9 to enable localization of the protease to the cell surface for promotion of cellular migration and angiogenesis [79]. In total, CD44 regulates the activity and localization of MMPs, suggesting that regulation of MMPs during cell migration occurs from the cycling of the proteases as opposed to their level of expression. The investigation of MMP activity is imperative to understanding the mechanism of cell migration within a system. However, only a number of studies have addressed the significance of MMP expression in 3D environments. The secretion of soluble MMPs, such as MMP-1,-2, -3, and -9, were found to minimally influence the migration of human microvascular endothelial cells through a 3D collagen type-I matrix. However, the increased expression level of MT1-MMP, a membrane associated protease, was found to be directly associated with the promotion of cell migration [69, 80]. Conversely, in vivo studies with lung cancer cells of various

144

J. Srivastava and M.H. Zaman

metastatic capabilities show similar levels of MT1-MMP expression in cell lines of varying migratory abilities whereas active MMP-2 increases at the leading edge of the most highly aggressive tumor cells [6, 81]. It is clear that the activity of MMPs and TIMPs influence the migratory capacity of tumorigenic cells, but a correlation remains to be clarified at the systems-level.

7.6.2

Urokinase Plasminogen Activator

Systems-level investigations of cancer cell migration have adopted a unique technique of multicellular spheroid culture to mimic some of the in vivo conditions experienced by a tumor in a 3D environment. Using breast cancer cell lines, a distinction between 2D planar culture and 3D multicellular spheroid culture was established, particularly in the expression of uPa, a serine protease [82]. uPa is significant in cell migration as it is an initiator of proteolytic cascades that contribute to the remodeling of the ECM and a mediator of signaling events that promote cell motility [83]. Cell migration experiments of HT1080 fibrosarcoma cells conducted in collagen type I matrices have visualized the localization of uPa proteases to lamellipodia and invadopodia, identifying a role for these enzymes in a 3D environment [40]. In 3D spheroid cultures of MCF-7 breast cancer cells, the expression levels of various uPa types were significantly elevated as compared to cells cultured as a 2D monolayer, exposing the ability of these tumorigenic cells to adapt to environmental variations and the importance of establishing cancer cell migratory mechanisms at a systems-level [82]. In aggressive carcinomas, it has been observed that primary tumor masses secrete aberrant amounts of proteases, such as MMPs, that may degrade components of the basement membrane to mechanically construct new motility pathways or mobilize various growth factors [70]. Studies of proteolytic molecules at a systems-level have provided an interesting link between cellular communication with the external environment, revealing the importance and complexities of such molecules in cancer cell migration.

7.7

Apoptosis

For many cell types, adherence to the ECM is critical for survival. Interruption of adherence renders the cell susceptible to various degradative processes that result in apoptosis. Cellular adherence to the matrix may be disturbed by a number of physiological processes that include inflammation, tumorgenesis, and wounding [84]. Through protease activity, signaling and structural proteins are degraded into small units that are removed by professional phagocytes, white blood cells that phagocytose foreign or decaying cell matter [85]. Apoptosis elicited by cellular detachment from the matrix is referred to as anoikis [84, 85]. Key signaling molecules that are involved in adhesion formation, maintenance, and turnover have

7 Systems Biology of 3D Tumor Cell Migration

145

been implicated in triggering anoikis. These proteins include FAK, PI3-kinase, and the proteins of the Ras/MAPK cascade [70, 85]. As cancer cell metastasis is correlated to increased cell growth and decreased apoptosis, investigations of the latter event have begun in 3D systems. In one study, the activated form of Erb2, an oncogene amplified in a number of breast cancers, was shown to suppress apoptosis and promote cancer cell proliferation in tumor tissues [86]. Other studies demonstrate the influence of the 3D ECM on apoptotic signaling by malignant MECs; since the anchoring of cells to the ECM is reliant on integrin receptors that attach to laminin of the basement membrane, the signaling of the proteins involved in attachment was investigated. These investigations revealed that cellular attachment to the membrane decreased cellular sensitivity to proapototic stimuli [84, 87]. The avoidance of proapototic factors, such as TNFa and FAS, can be contributed to mechanical barriers of the ECM and activation of transcriptional signaling by NF-kB, whose activation is dependent on integrin binding to laminin [45]. Cancer models that probe into the effect of apoptotic factors on cells in 3D environments are in their infancy. As multiple studies have investigated the role of apoptosis regulatory proteins, such as Bcl2, Bim, and the caspase family of proteins, the roles of these regulatory molecules have yet to be elucidated in 3D environments [88].

7.8

Summary

Despite the numerous investigations into the molecular mechanisms of cancer, many questions remain regarding the multi-step process involved in this type of disease. From studies performed in 2D environments, the presence and involvement of multiple proteins has been unveiled; however, systems-level studies are beginning to highlight the complexities of such pathways that were not apparent in the 2D studies. By adding just 1D to the in vitro culture environment, researchers are able to investigate links between the external and internal cellular environments. As such systems-level investigations progress, we are provided with a unique view of the integrative processes utilized by cells, one that will contribute immensely to the progression of field.

References 1. Sauer, U., M. Heinemann, and N. Zamboni, Getting closer to the whole picture. Science, 2007. 316(5824): p. 550–551. 2. Ideker, T., T. Galitski, and L. Hood, A new approach to decoding life: Systems biology. Annual Review of Genomics and Human Genetics, 2001. 2(1): p. 343–372. 3. Kitano, H., Systems biology: a brief overview. Science, 2002. 295(5560): p. 1662. 4. Hornberg, J.J., et al., Cancer: a systems biology disease. Biosystems, 2006. 83(2–3): p. 81–90. 5. Weinberg, R.A., The Biology of Cancer 2007, New York: Garland Science. 313.

146

J. Srivastava and M.H. Zaman

6. Even-Ram, S. and K.M. Yamada, Cell migration in 3D matrix. Current Opinion in Cell Biology, 2005. 17(5): p. 524–532. 7. Hanahan, D. and R.A. Weinberg, The hallmarks of cancer. Cell, 2000. 100(1): p. 57–70. 8. Friedl, P., et al., Migration of highly aggressive MV3 melanoma cells in 3-dimensional collagen lattices results in local matrix reorganization and shedding of {alpha} 2 and {beta} 1 integrins and CD44. Cancer Research, 1997. 57(10): p. 2061. 9. Wolf, K., et al., Multi-step pericellular proteolysis controls the transition from individual to collective cancer cell invasion. Nature Cell Biology, 2007. 9(8): p. 893–904. 10. Ridley, A.J., et al., Cell migration: integrating signals from front to back. Science, 2003. 302(5651): p. 1704. 11. Yamaguchi, H. and J. Condeelis, Regulation of the actin cytoskeleton in cancer cell migration and invasion. Biochimica et Biophysica Acta, 2007. 1773(5): p. 642–652. 12. Provenzano, P.P., et al., Matrix density-induced mechanoregulation of breast cell phenotype, signaling and gene expression through a FAK-ERK linkage. Oncogene, 2009. 28(49): p. 4326–4343. 13. Zaman, M.H., et al., Migration of tumor cells in 3D matrices is governed by matrix stiffness along with cell-matrix adhesion and proteolysis. Proceedings of the National Academy of Sciences, 2006. 103(29): p. 10889. 14. Wozniak, M.A. and P.J. Keely, Use of three-dimensional collagen gels to study mechanotransduction in T47D breast epithelial cells. Biological Procedures Online, 2005. 7(1): p. 144–161. 15. Griffith, L.G. and M.A. Swartz, Capturing complex 3D tissue physiology in vitro. Nature Reviews. Molecular Cell Biology, 2006. 7(3): p. 211–224. 16. Kleinman, H.K., et al., Isolation and characterization of type IV procollagen, laminin, and heparan sulfate proteoglycan from the EHS sarcoma. Biochemistry, 1982. 21(24): p. 6188–6193. 17. Erler, J.T. and V.M. Weaver, Three-dimensional context regulation of metastasis. Clinical and Experimental Metastasis, 2009. 26(1): p. 35–49. 18. Butcher, D.T., T. Alliston, and V.M. Weaver, A tense situation: forcing tumour progression. Nature Reviews. Cancer, 2009. 9(2): p. 108–122. 19. Baker, E.L., R.T. Bonnecaze, and M.H. Zaman, Extracellular matrix stiffness and architecture govern intracellular rheology in cancer. Biophysical Journal, 2009. 97(4): p. 1013–1021. 20. Baker, E.L. and M.H. Zaman, The biomechanical integrin. Journal of Biomechanics, 2010. 43(1): p. 38–44. 21. Lo, C.M., et al., Cell movement is guided by the rigidity of the substrate. Biophysical Journal, 2000. 79(1): p. 144–152. 22. Kniazeva, E. and A.J. Putnam, Endothelial cell traction and ECM density influence both capillary morphogenesis and maintenance in 3-D. American Journal of Physiology. Cell Physiology, 2009. 297(1): p. C179. 23. Guo, W. and F.G. Giancotti, Integrin signalling during tumour progression. Nature Reviews. Molecular Cell Biology, 2004. 5(10): p. 816–826. 24. Larsen, M., et al., The matrix reorganized: extracellular matrix remodeling and integrin signaling. Current Opinion in Cell Biology, 2006. 18(5): p. 463–471. 25. Mitra, S.K., D.A. Hanson, and D.D. Schlaepfer, Focal adhesion kinase: in command and control of cell motility. Nature Reviews. Molecular Cell Biology, 2005. 6(1): p. 56–68. 26. Mitra, S.K. and D.D. Schlaepfer, Integrin-regulated FAK-Src signaling in normal and cancer cells. Current Opinion in Cell Biology, 2006. 18(5): p. 516–523. 27. Provenzano, P.P., et al., Contact guidance mediated three-dimensional cell migration is regulated by Rho/ROCK-dependent matrix reorganization. Biophysical Journal, 2008. 95(11): p. 5374–5384. 28. Giancotti, F.G. and E. Ruoslahti, Integrin signaling. Science, 1999. 285(5430): p. 1028.

7 Systems Biology of 3D Tumor Cell Migration

147

29. Doerr, M.E. and J.I. Jones, The roles of integrins and extracellular matrix proteins in the insulin-like growth factor I-stimulated chemotaxis of human breast cancer cells. Journal of Biological Chemistry, 1996. 271(5): p. 2443. 30. Sporn, M.B. and A.B. Roberts, Autocrine growth factors and cancer. Nature. 1985. 313(6005): p. 745–747. 31. Huang, C., K. Jacobson, and M.D. Schaller, MAP kinases and cell migration. Journal of cell science, 2004. 117(20): p. 4619. 32. Caswell, P.T., et al., Rab25 associates with a5b1 integrin to promote invasive migration in 3D microenvironments. Developmental Cell, 2007. 13(4): p. 496–510. 33. Zerial, M. and H. McBride, Rab proteins as membrane organizers. Nature Reviews. Molecular Cell Biology, 2001. 2(2): p. 107–117. 34. Gebhardt, C., et al., c-Fos-dependent induction of the small ras-related GTPase Rab11a in skin carcinogenesis. American Journal of Pathology, 2005. 167(1): p. 243. 35. Caswell, P.T. and J.C. Norman, Integrin trafficking and the control of cell migration. Traffic, 2006. 7(1): p. 14–21. 36. Cukierman, E., et al., Taking cell-matrix adhesions to the third dimension. Science, 2001. 294(5547): p. 1708. 37. Cukierman, E., R. Pankov, and K.M. Yamada, Cell interactions with three-dimensional matrices. Current Opinion in Cell Biology, 2002. 14(5): p. 633–640. 38. Derynck, R., R.J. Akhurst, and A. Balmain, TGF-beta signaling in tumor suppression and cancer progression. Nature Genetics, 2001. 29(2): p. 117–130. 39. Yamazaki, D., S. Kurisu, and T. Takenawa, Regulation of cancer cell motility through actin reorganization. Cancer Science, 2005. 96(7): p. 379–386. 40. Wolf, K. and P. Friedl, Mapping proteolytic cancer cell-extracellular matrix interfaces. Clinical and Experimental Metastasis, 2009. 26(4): p. 289–298. 41. Weaver, A.M., Cortactin in tumor invasiveness. Cancer Letters, 2008. 265(2):157–166 42. Ridley, A.J., Rho GTPases and cell migration. Journal of Cell Science, 2001. 114(15): p. 2713. 43. Yamaguchi, H., et al., Molecular mechanisms of invadopodium formation: the role of the N-WASP-Arp2/3 complex pathway and cofilin. Journal of Cell Biology, 2005. 168(3): p. 441. 44. Condeelis, J. and J.E. Segall, Intravital imaging of cell movement in tumours. Nature Reviews. Cancer, 2003. 3(12): p. 921–930. 45. Weaver, V.M., et al., Reversion of the malignant phenotype of human breast cells in threedimensional culture and in vivo by integrin blocking antibodies. Journal of Cell Biology, 1997. 137(1): p. 231. 46. Satpathy, M., et al., Tissue transglutaminase regulates MMP-2 in ovarian cancer by modulating CREB activity. Journal of Biological Chemistry, 2009. 284(23):15390–15399. 47. Woessner, J.F., MMPs and TIMPs: A historical perspective. Molecular Biotechnology, 2002. 22(1): p. 33–49. 48. Mouneimne, G., et al., Phospholipase C and cofilin are required for carcinoma cell directionality in response to EGF stimulation. Journal of Cell Biology, 2004. 166(5): p. 697. 49. Wang, W., et al., Identification and testing of a gene expression signature of invasive carcinoma cells within primary mammary tumors. Cancer Research, 2004. 64(23): p. 8585. 50. Keely, P.J., et al., Cdc42 and Rac1 induce integrin-mediated cell motility and invasiveness through PI (3) K. Nature, 1997. 390(6660): p. 632–636. 51. Bray, D., Cell Movements: From Molecules to Motility. 2nd ed. 2001, New York: Garland 52. Sahai, E. and C.J. Marshall, Differing modes of tumour cell invasion have distinct requirements for Rho/ROCK signalling and extracellular proteolysis. Nature Cell Biology, 2003. 5(8): p. 711–719. 53. Itoh, K., et al., An essential part for Rho-associated kinase in the transcellular invasion of tumor cells. Nature Medicine, 1999. 5(2): p. 221–225. 54. Goley, E.D. and M.D. Welch, The ARP2/3 complex: an actin nucleator comes of age. Nature Reviews. Molecular Cell Biology, 2006. 7(10): p. 713–726.

148

J. Srivastava and M.H. Zaman

55. Bryce, N.S., et al., Cortactin promotes cell motility by enhancing lamellipodial persistence. Current Biology, 2005. 15(14): p. 1276–1285. 56. Artym, V.V., et al., Dynamic interactions of cortactin and membrane type 1 matrix metalloproteinase at invadopodia: defining the stages of invadopodia formation and function. Cancer Research, 2006. 66(6): p. 3034. 57. Clark, E.S., et al., Cortactin is an essential regulator of matrix metalloproteinase secretion and extracellular matrix degradation in invadopodia. Cancer Research, 2007. 67(9): p. 4227. 58. Hazan, R.B., et al., Exogenous expression of N-cadherin in breast cancer cells induces cell migration, invasion, and metastasis. Journal of Cell Biology, 2000. 148(4): p. 779. 59. M€uller, T., et al., Regulation of epithelial cell migration and tumor formation by catenin signaling. Experimental Cell Research, 2002. 280(1): p. 119–133. 60. Kurayoshi, M., et al., Expression of Wnt-5a is correlated with aggressiveness of gastric cancer by stimulating cell migration and invasion. Cancer research, 2006. 66(21): p. 10439. 61. Polakis, P., Wnt signaling and cancer. Genes & Development, 2000. 14(15): p. 1837. 62. McLean, G.W., et al., The role of focal-adhesion kinase in cancer: a new therapeutic opportunity. Nature Reviews. Cancer, 2005. 5(7): p. 505–515. 63. Sieg, D.J., et al., FAK integrates growth-factor and integrin signals to promote cell migration. Nature Cell Biology, 2000. 2(5): p. 249–256. 64. Slack, J.K., et al., Alterations in the focal adhesion kinase/Src signal transduction pathway correlate with increased migratory capacity of prostate carcinoma cells. Oncogene, 2001. 20(10): p. 1152. 65. Ilic, D., et al., FAK promotes organization of fibronectin matrix and fibrillar adhesions. Journal of Cell Science, 2004. 117(2): p. 177. 66. Niwa, Y., et al., Methylation silencing of SOCS-3 promotes cell growth and migration by enhancing JAK/STAT and FAK signalings in human hepatocellular carcinoma. Oncogene, 2005. 24(42): p. 6406–6417. 67. O’Shea, J.J., M. Gadina, and R.D. Schreiber, Cytokine signaling in 2002 new surprises in the JAK/STAT Pathway. Cell, 2002. 109(2S1): p. 121–131. 68. Comoglio, P.M. and L. Trusolino, Cancer: the matrix is now in control. Nature Medicine, 2005. 11(11): p. 1156–1157. 69. Egeblad, M. and Z. Werb, New functions for the matrix metalloproteinases in cancer progression. Nature Reviews. Cancer, 2002. 2(3): p. 163–176. 70. MacDougall, J.R. and L.M. Matrisian, Contributions of tumor and stromal matrix metalloproteinases to tumor progression, invasion and metastasis. Cancer and Metastasis Reviews, 1995. 14(4): p. 351–362. 71. Gomez, D.E., et al., Tissue inhibitors of metalloproteinases: structure, regulation and biological functions. European Journal of Cell Biology, 1997. 74(2): p. 111. 72. Sternlicht, M.D., et al., The stromal proteinase MMP3/stromelysin-1 promotes mammary carcinogenesis. Cell, 1999. 98: p. 137–146. 73. Lochter, A., et al., Matrix metalloproteinase stromelysin-1 triggers a cascade of molecular alterations that leads to stable epithelial-to-mesenchymal conversion and a premalignant phenotype in mammary epithelial cells. Journal of Cell Biology, 1997. 139(7): p. 1861. 74. Coussens, L.M., et al., MMP-9 supplied by bone marrown˜derived cells contributes to skin carcinogenesis. Cell, 2000. 103(3): p. 481–490. 75. Mueller, S.C., et al., A novel protease-docking function of integrin at invadopodia. Journal of Biological Chemistry, 1999. 274(35): p. 24947. 76. Nakahara, H., et al., Transmembrane/cytoplasmic domain-mediated membrane type 1-matrix metalloprotease docking to invadopodia is required for cell invasion. Proceedings of the National Academy of Sciences of the United States of America, 1997. 94(15): p. 7959. 77. Giannelli, G., et al., Induction of cell migration by matrix metalloproteinase-2 cleavage of laminin-5. Science, 1997. 277: p. 225–228. 78. Giannelli, G., et al., Expression of matrix metalloprotease-2-cleaved laminin-5 in breast remodeling stimulated by sex steroids. American Journal of Pathology, 1999. 154(4): p. 1193.

7 Systems Biology of 3D Tumor Cell Migration

149

79. Kajita, M., et al., Membrane-type 1 matrix metalloproteinase cleaves CD44 and promotes cell migration. Journal of Cell Biology, 2001. 153(5): p. 893. 80. Koike, T., et al., MT1-MMP, but not secreted MMPs, influences the migration of human microvascular endothelial cells in 3-dimensional collagen gels. Journal of Cellular Biochemistry, 2002. 86(4): p. 748–758. 81. Hofmann, U.B., et al., Expression of matrix metalloproteinases in the microenvironment of spontaneous and experimental melanoma metastases reflects the requirements for tumor formation. Cancer Research, 2003. 63(23): p. 8221. 82. Faute, M.A., et al., Distinctive alterations of invasiveness, drug resistance and celln˜cell organization in 3D-cultures of MCF-7, a human breast cancer cell line, and its multidrug resistant variant. Clinical and Experimental Metastasis, 2002. 19(2): p. 161–167. 83. Chandrasekar, N., et al., Downregulation of uPA inhibits migration and PI3k/Akt signaling in glioblastoma cells. Oncogene, 2003. 22(3): p. 392–400. 84. Jacks, T. and R.A. Weinberg, Taking the study of cancer cell survival to a new dimension. Cell, 2002. 111(7): p. 923–925. 85. Reddig, P.J. and R.L. Juliano, Clinging to life: cell to matrix adhesion and cell survival. Cancer and Metastasis Reviews, 2005. 24(3): p. 425–439. 86. Slamon, D.J., et al., Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science, 1987. 235(4785): p. 177. 87. Weaver, V.M., et al., b4 integrin-dependent formation of polarized three-dimensional architecture confers resistance to apoptosis in normal and malignant mammary epithelium. Cancer Cell, 2002. 2(3): p. 205–216. 88. Vander Heiden, M.G. and C.B. Thompson, Bcl-2 proteins: regulators of apoptosis or of mitochondrial homeostasis? Nature Cell Biology, 1999. 1(8): p. E209–E216.

Chapter 8

Development of Three-Dimensional Tumor Models for the Study of Anti-Cancer Drug Effects Wei Sun, Raj Rajagopalan, and Chwee Teck Lim

This chapter is part of Section III: Mechano-pathology of Disease

Abstract Cell-based assays can be used prior to in vivo tests to enable quantitative examinations of the responses of cells and, on the tissue scale, tumor-stromal mimics to drugs. However, the challenge is always to reproduce as closely as possible the native-state tumor microenvironment so as to accurately predict in vivo drug responses. In this chapter, we will list the advantages of 3D models over conventional 2D culture, review the current progresses in developing anti-cancer drug tests based on 3D in vitro culture systems, and discuss the importance of mechanics in designing these 3D models. We will highlight various techniques that have been used to construct 3D models that reflect the native physiology/pathology of a solid tumor as closely as possible in order to evaluate these drug effects, and detail specific research questions addressed using these 3D models. We will conclude this review with an outlook of the future of 3D models for improved in vitro drug assays. Keywords Extracellular matrix Three-dimensional In vitro Drug resistance Multicellular organizations Cancer

8.1

Introduction

In the course of drug development and discovery, a necessary step after molecular screening and before in vivo animal model-based testing will be to perform cellbased assay testing. This is because the biological activities of molecules in living

C.T. Lim (*) NUS School of Integrative Sciences and Engineering, National University of Singapore, Singapore and Division of Bioengineering and Department of Mechanical Engineering, National University of Singapore, Singapore and Mechanobiology Institute, National University of Singapore, Singapore e-mail: [email protected] A. Wagoner Johnson and Brendan A.C. Harley (eds.), Mechanobiology of Cell-Cell and Cell-Matrix Interactions, DOI 10.1007/978-1-4419-8083-0_8, # Springer Science+Business Media, LLC 2011

151

152

W. Sun et al.

cells are often so complex that cell-specific responses cannot be studied through target-oriented approaches [1]. While in vivo evaluations are often limited by the techniques of high resolution imaging and real-time assays, there are more established and easier methods for detecting the drug-induced changes from in vitro assays than from animal models. If drug testing involves an in vitro pre-screening step using cell-based assays before in vivo tests, it may help to cut down on the number of expensive animal studies [2]. Moreover, it allows for a comprehensive and detailed examination of the drug responses from cells and, on tissue scale, from tumor-stromal mimics. However, the challenge is to reproduce as much as possible the native-state tumor microenvironment so as to predict the in vivo drug responses. To reconstruct a tumor-associated microenvironment in vitro, cell-based drug assays using three-dimensional (3D) cell culture systems are beginning to gain popularity in recent years. As will be discussed in this chapter, cell-cell and cellmatrix interactions determine not only cell fate but also the sensitivity of cells to drug treatment. In vivo, cells are manifested in a 3D environment which implies that 3D cell culture systems have higher potential in restoring tissue-level organizations and functions, thus providing more reliable indications of the drug effects. These new assay systems, once validated, will find applications in pre-clinical trial drug testing and drug delivery techniques and may lead to novel cancer therapies. In this chapter, we will list the advantages of 3D models over conventional 2D culture, review the current progresses in developing anti-cancer drug tests based on 3D in vitro culture systems, and discuss the importance of mechanics in designing these 3D models. We hope to highlight the various techniques for constructing 3D models that can reflect as closely as possible the physiology/pathology of a solid tumor and for evaluating drug effects, and see how recent studies addressed specific research questions using these 3D models. We will conclude this review with an outlook into the future of 3D models for improved in vitro drug assays.

8.2

Background: Why 2D Cell Culture is Not Ideal for Anti-Cancer Drug Testing

The extracellular matrix (ECM) supports cells with structural anchorage and also plays important roles in a variety of cell activities including motility, proliferation, differentiation, and apoptosis. Since cells are in contact with neighboring cells and ECM from all three dimensions in vivo, artificial and flat 2D culture is not sufficient in recreating the tissue-specific microenvironment. Under the treatment of drugs or radiations, cancer cells grown on 2D showed responses differing from those found in vivo, in terms of uptake dosage [3], cytotoxicity [4, 5], metabolism [2] and cell proliferation rate [6]. There are a few underlying reasons for the shortcomings of 2D drug assays, among which, tumor heterogeneity and tissue mechanics will be discussed in the following sections. Tumor tissues are geometrically and mechanically heterogeneous, with uneven distributions of insoluble factors that can bind to cell-surface receptors and mediate

8 Three‐Dimensional Tumor Models

153

signaling. A study conducted by Murray and colleagues in 1954 revealed that drug responses can vary in different regions in the same tumor [7]. In tumor tissues with a high cell density, there are gradients of oxygen, nutrients, catabolites as well as pH [2]. None of these gradients is found in monolayer culture conditions where cells live in uniform liquid media. In turn, the traditional 2D models are incapable of presenting localized physical or biochemical cues to cancer cells. Unlike 2D models, a proper 3D model allows cell-cell and cell-ECM interactions to mimic those in a real tumor, and also allows the central hypoxia region and peripheral proliferating regions to develop [8]. Figure 8.1 schematically illustrates the spatial heterogeneity in a tumor model. Recent findings in the field show that the molecular expressions and enzymatic activities of cancer cells are influenced by the stressful microenvironment associated with solid tumors [9]; therefore, the molecular targets of anti-tumor drugs are affected [2]. Another interesting study of oral squamous cell carcinoma found that, after having been cultured in a 3D scaffold, cells showed phenotype changes and became more fibroblast-like, expressing more of a5b1 integrin, a fibronectin receptor, as well as N-cadherin [8]. These findings suggest that it is critical for an in vitro drug assay to capture the cancer microenvironment which often regulates tumorigenesis and metastasis. Also, another topic of interest is matrix mechanics. There has been an increasing number of findings relating matrix mechanics to cancer progression and metastasis [10–13] as well as angiogenesis [14], which is a critical factor contributing to tumor malignancy development and metastasis. With the potential to better recapitulate the pathological features of cancers, fine-tuning the mechanical microenvironment

Fig. 8.1 Schematic presentation of the similarities of a tumor (in vivo) and a multilayered postconfluent cell culture and spheroids (in vitro) (adapted from Padron et al. [2] with permission from Elsevier)

154

W. Sun et al.

created by these 3D in vitro cancer models will add to their advantages for pre-clinical drug testing. Consequently, putting cells in a properly reconstructed tumor environment, which is often 3D, is required for evaluating the effectiveness of anti-tumor therapies. In some cases, when the drug-targeted cellular function has a 3D geometry in nature, such as angiogenesis and tumor invasion, a 3D model is necessary for such functional studies.

8.3 8.3.1

3D Drug Assay Techniques Building 3D Culture Systems

To recapitulate the 3D features of solid tumors, multicellular spheroids could be developed from hanging drops of single cell suspensions, from spinning-created low gravity cultures [15, 16], or from postconfluent growth in V-bottomed microtiter plates [2]. Natural ECM components or their simulates, such as collagens and Matrigel, synthetic biocompatible polymer gels (e.g., RADA and PEG), scaffolds (PLG) and nanofibers [17] all have found applications in supporting cells to form 3D tumor models. Embedding a tumor biopsy in hydrolyzed meshes to form a histoculture shows potential of examining the patient-dependent drug responses ex vivo [18]. Fibrin-thrombin gel is a substrate widely used for simulating angiogenesis [19, 20]. In search of anti-metastasis treatments, the in vitro simulation of malignant cell invasion into surrounding tissues requires confronting the latter with the former. This configuration is achieved with a variety of models, using embryonic chick heart-derived 3D cell culture system [21], multicellular aggregates of tumor and fetal brain cells co-culture [22], as well as hydrogel-based cell invasion models [23, 24]. For more comprehensive discussion of 3D culture systems, there are many other good reviews available [1, 25].

8.3.2

Issues and Parameters for Drug Evaluation

As illustrated in Fig. 8.2, two issues must be addressed in order to reproduce and then evaluate the drug responses using in vitro models and to achieve high clinical relevance. First, as the bottom part of the chart shows, the malignant development of the tumor should be fully manifested, as a baseline. Next, the changes induced by drug treatment must be readily detectable and further, be measured qualitatively or quantitatively. There are a few key parameters established for evaluating drug effects under various contexts. For cytotoxicity evaluations, cell percentage of growth (PG) is the golden standard, which is determined through dosages corresponding to 50% growth inhibition (IC50), total growth inhibition (TGI) and 50% cell death (LC50).

8 Three‐Dimensional Tumor Models

155

Drug evaluaon

Change in tumor malignancy

Pharmacodynamics

Drug uptake, retenon and metabolites

Protein expression/ transcripts level, enzyme acvity, cell viability, morphology, proliferaon, migraon and ssue morphogenesis

Angiogenesis

Malignancy manifestaon (base-line)

Fig. 8.2 Flowchart showing the parameters for malignancy and drug effect evaluation. Altered protein expression/transcripts level, enzyme activity, cell viability, morphology, proliferation, migration, and tissue morphogenesis as well as angiogenesis are all parameters of interest for the evaluation of malignancy manifestation. Effectiveness of drugs is evaluated not only by the changes of the above parameters, but also through the time-dependent levels of drug uptake, retention and metabolites. The selection of endpoint indices depends on objectives of the study and specific targets of the drugs

Alternatively, an index indicating the concentration of drugs that can reduce the cell number by 50% is recommended for 3D culture [2]. For morphogenesis/angiogenesis evaluations, histochemistry and histofluorescence can reveal the molecular and cellular details deep inside a 3D tissue. With advanced microscopy techniques, including confocal and multi-photon, imaging 3D-reconstructed tissue structures is becoming more of routine work. In particular, light-sheet based illumination microscopy methods such as selective plane illumination microscopy (SPIM) [26] and its enhanced and more optically efficient model, digital scanned laser light sheet fluorescence microscopy (DSLM) [27], have improved resolution in thick samples and meanwhile reduced photo damage in the tissues. As specifically developed to visualize 3D structures, optical coherence tomography enabled the analysis of 3D scaffolds micro-architecture [28] and cell-induced collagen gel remodeling [29]. In addition, live microscopy allows for

156

W. Sun et al.

tracking and quantification of the 3D morphology dynamics and cell migratory behaviors in real-time [30, 31], and enables the elucidation of dynamic cell-cell and cell-ECM interactions in 3D [32].

8.4

3D Assay Related Research

Investigating drug effects in 3D systems is primarily driven by (1) understanding the mechanisms that cause discrepancies between drug screening outcome of monolayer cell assays and clinical trials; (2) screening for anti-cancer drugs that are affected by cell-tissue organization; (3) searching for drugs that target at pathological processes of intrinsic 3D geometrical nature, such as tumor morphogenesis, angiogenesis and tissue invasion; and other specific research questions including drug delivery vehicle tests [33] and preservation of hepatic-specific functions for drug metabolizing tests [1, 34].

8.4.1

Mechanistic Studies

Research into mechanisms underlying drug resistance has involved issues such as drug transportation and multicellular-specific phenomena. Meanwhile, searching for agents that have the potential in countering drug resistance has also become a hotspot.

8.4.1.1

Multi-Layer Drug Transport

It is speculated that the transport of drugs through multilayer of cells may contribute to the differences from drug screening using monolayer cell culture and that from clinical trials. Conventional monolayer culture exposes cells directly to drugcarried media, thus it is not representative of the physical environment for in vivo drug transport. In contrast, multi-cellular spheroids models and histocultures improve the simulation of pharmacokinetics in solid tumors, which form a better basis to analyze drug penetration, accumulation and retention. On such a basis, studies found that drug transportation is not limited by diffusion through the collagen matrix supporting the histocultures, but by cell density in the tumor [35], because the drug consumption by outer-layer cells reduced the dose received by cells in the core [36]. One way to improve anti-tumor treatment is to find drugs that reach adequately deep into tumors, a tumor tissue-disk of about 150 mm thick was examined by Kyle et al. [37] for drug penetration from a specific direction. The authors observed a tenfold dosage decrease with exposure to doxorubicin and epirubicin, and 30-fold or more decrease for daunorubicin and mitoxantrone.

8 Three‐Dimensional Tumor Models

157

3D culture systems also provided a platform to test the biocompatibility and efficiency of novel drug delivery methods. Such 3D models proved that adenoviral vectors were able to deliver some fusion proteins to infected cells, and the protein was able to traffic into uninfected cells and, similarly, cause apoptosis [33]. For advanced drug delivery, the biocompatibility of microcapsules fabricated through layer-by-layer self-assembly was examined with glioma cells and fibroblasts cells, by evaluating cell invasion and tumor spheroids growth in 3D collagen culture [38]. 8.4.1.2

Multicellular/Tissue-Dependent Drug Resistance

Some acquired anti-tumor drug resistance is found in vivo but not detected in monolayer culture. However, growing tumor cells as 3D multicellular spheroids, as shown in Fig. 8.3, reproduced the drug resistance as proven in vivo [39]. The drug resistance has been found to be related to increased compaction and intercellular adhesion in these 3D cell aggregates. On such basis, multicellular spheroids have been used in mechanistic research to understand what causes chemotherapy resistance and, subsequently, find measures to enhance drug effect. First of all, large multicellular spheroids often result in hypoxia in the tumor center, besides impeding transcellular drug transport. Using spheroids of breast carcinoma cells [40], it is found that higher resistance to anti-cancer agents such as doxorubicin are probably due to reduced nitric oxide (NO) signaling. In addition, interactions between cells and ECM may also alter the cellular responses to drugs. A multi-component 3D matrix was investigated by Harisi et al. [41] in terms of cell proliferation, apoptosis and DNA damage based on an osteosarcoma cell line. ECM-associated p53 function deficiency might explain the doxorubicin resistance. In particular, ECM gel components, primarily heparin sulfate proteoglycan and fibronectin, were responsible for such drug resistance. Collagen IV showed chemosensitizing effect, while laminin and nidogen did not show any effect [41]. The same type of models facilitated the search for mechanisms and possible agents, termed “chemosensitizers”, to alleviate such drug resistance related to multicellular nature of tumors [42]. For example, it was proved that hyaluronidase was able to disrupt cell-cell adhesion, thus sensitizing the tumors to chemotherapy [43]. In the above-mentioned study of NO signaling, NO mimetics partially reversed multicellular resistance, suggesting of the potential to chemosensitize solid tumors [40].

8.4.2

Drug Testing Using Multicellular 3D Models

Multicellular models have shown strengths in the research into cancer therapeutics in a number of special contexts. One such domain is radio therapy. As cells in a 3D structure more effectively absorb radiation energy which causes DNA damage, multi-cellular spheroids are good models to develop radioisotope concentrator gene

158

W. Sun et al.

Fig. 8.3 Drug-resistant variant of EMT-6 mammary tumor cells formed spherical multicellular aggregates after 3–5 days of culture in agarose gel and when the drugs were administered. There was a necrotic core in the spheroid and viable cells distributed on the periphery (a) is of lower magnification (scale bar ¼ 100 am) and (b) is of higher magnification (scale bar ¼ 40 am). (Courtesy of Proc Natl Acad Sci USA (Kobayashi et al. [37])) (c) and (d), The parental tumor cells produced sparsely distributed cell-aggregates, instead of compact multi-cellular spheroids

therapy. The same dose of radioisotope achieved complete killing of cancer cells in 3D multicellular spheroids, comparing to a 50–70% killing effect of monolayer prostate cancer cells [3]. In another case, the outcome of photodynamic therapy (PDT) to treat brain tumor is associated with the size of tumor. Such size effect was investigated using a glioma 3D spheroid model, based on apoptosis, necrosis and morphological changes after the treatment of drug or laser or combination of the two. The growth

8 Three‐Dimensional Tumor Models

159

kinetics could be monitored for 8 days, finding partial or complete cell death after drug-laser combined PDT [44]. To develop patient-specific cancer therapy, 3D culture systems can be extended from cell lines to clinical samples. Led by the studies in Hoffman laboratory in San Diego, the incorporation of a small scale tumor into collagen sponge gel approximately reproduced tumor tissues in native status, in which cancer cells, stromal cells, including fibroblasts and immune cells, together with ECM formed a heterogeneous microenvironment. This histoculture system has tested the effectiveness of many new anti-cancer drugs [18]. A clinical application of histoculture is to predict chemotherapy responses, through testing drug sensitivity of biopsies from individual patients. In a well controlled in vitro setting, histoculture can retain 3D tissue structure and organization up to a 100 days [45], and showed enhanced clinical relevance/predictability of pharmacologic studies.

8.4.3

Searching Drugs Against 3D Pathological Processes

8.4.3.1

Tissue Morphogenesis

Cell and tissue-level morphological changes are important indicators for cancer malignancy development, as well as the effects of anti-cancer drugs. 3D interactions between cells and the ECM allow malignant morphologies to develop in a more-realistic context, and to reproduce the cancer-induced generation/degeneration of the multicellular organizations. For example, the number of microvilli is an important measure of 3D cytomorphologic differentiation of thyrocyte, indicating the progression of anaplastic thyroid carcinoma. Tumor necrosis factor alpha (TNF-alpha) was found to enhance the differentiation of anaplastic thyroid carcinoma cells, through NF-kappa B activation, thus becoming a potential differentiation therapy [46]. Carcinoma progression may destroy the polarized epithelial structures, and sometimes develop into malignant organizations with 3D structures that are tumor-specific. As reported by Debnath et al. [47], 3D on-Matrigel culture model of benign breast cancer cells enabled the investigation of the morphogenesis of epithelia to elucidate the role of serine/threonine kinase Akt. Through inhibiting Akt effector, an anti-cancer drug reduced morphological disruption and also limited Akt cooperatively induced promotion of proliferation [47]. Yao et al. [48] observed that ovarian cancers under hypoxia formed blood vessel-like structures in Matrigel-based 3D culture. The study also identified an inhibitor that limited the formation of tumor channels and networks, through blocking hypoxia-inducible factor (HIF)-1alpha at transcription level. The histomorphological analysis of 3D culture was able to reveal necrosis, nuclei shrinkage, and other tumor-killing effect of drugs, along with cell viability assays. For example, in a study by Wang et al. [49], by growing 1 mm3 tissue biopsies from gastric cancer patients on ready-made 3D collagen sponge, the cytotoxicity of anti-cancer agent combined with hyperthermia was examined.

160

8.4.3.2

W. Sun et al.

Angiogenesis

The spread of new blood vessels into tumor microenvironment is not only important for supplying oxygen and nutrients to facilitate tumor growth, it may also create channels for tumor cell dissemination [50]. The 3D culture allows endothelial network to develop [19], and in addition, to show changes in angiogenesis in a tumor-endothelia co-culture model [51]. Anti-angiogenesis drug effects on preventing vascular network expansion can thus be examined in these systems. Moreover, some tumor cells secrete pro-angiogenic factors, and the phenomenon is only recapitulated by properly designed 3D culture systems, but not by 2D monolayer culture [8]. 3D models thus have multifaceted advantages in anti-angiogenesis drug research. For example, 3D vascularization assays can reveal unique angiogenesis characteristics as compared to other assays. For instance, certain drug treatment was found to degenerate vascular capillary networks formed in 3D, but the treatment did not decrease endothelial cells (ECs) migration/chemotaxis in collagen-coated Boyden Chamber. The results suggest that the potential of vascularization in complex 3D geometry is not able to be reflected in whether cells can migrate through vertical channels, which are usually a few microns in diameter [52]. Collagen-I, Matrigel and fibrin-thrombin gels were most commonly applied in supporting 3D angiogenesis, which will be discussed with examples in the following paragraphs. Collagen-I models provide sufficient mechanical support for prolonged vascular generation, and thus are widely used in angiogenesis modeling. An agarose well-set, cylindrical collagen-I gel was maintained in floating condition in drug-dissolved or control medium, to culture segments of rat aorta for 2 weeks. The study evaluated the dose- and time-dependent effect of a prodrug in 3D angiogenesis, with or without the drug activation enzyme [53]. The same group also demonstrated anti-angiogenesis effects of a panel of drugs [54]. Another collagen 3D gel-supported EC culture model was adopted in testing the function of a cholesterol-lowering agent in inhibiting vessel growth. The drug worked through Rho-A pathways, and limited the vascular tubular structure formation of ECs [55]. Likewise, the drug Avastin was found to inhibit the growth and induce apoptosis of lung microvascular ECs in a 3D culture based on rat-tail collagen-I. Measured by the size of vessel-free area, micro-vascularization was impeded in a dose-dependent manner [56]. Matrigel-based angiogenesis models are also popular. To study the interactions between human breast epithelial cells (HBECs) and ECs, a 3D model was developed to recapitulate estrogen-induced angiogenesis as in vivo. Meanwhile, with HBEC – EC co-culture on Matrigel, the proliferation of ECs and branching ductal-alveolar morphogenesis of HBECs was co-localized. The cooperative growth-promotion effects were further enhanced by estrogen, but inhibited by anti-estrogens [48]. Similarly, Matrigel model supported ECs to form 3D capillary tubes. Based on such a model, it was found that suppressing certain matrix-degradation enzyme decreased the viability and retraction of ECs [57]. Formed from fibrinogen and thrombin, fibrin gels are commonly adopted in the evaluation of tubular structure formation of ECs. With a fibrin-gel based

8 Three‐Dimensional Tumor Models

161

3D angiogenesis model, the effect from three individual angiogenic factors and a combination of nine angiogenic factors was tested. Enhanced endothelial tubulogenesis was found in the co-culture with glioma tumor cells, but not with normal epithelial cells. The function of a panel of metalloproteinases was also examined, and MT1-MMP, in particular, could become a therapeutic target against angiogenesis [20]. Studies have also started to build co-culture systems involving ECs, tumor cells and even smooth muscle cells (SMCs). In a 3D fibrin gel coculture assay, a drug produced from berry extracts was found to prevent ECs and SMCs from forming capillary-like tubes [58]. In a way similar to culturing cancer biopsy, fresh human tissues are embedded into hydrogel to allow for vessel network development. Models using human placental vein discs cultured in 3D fibrin-thrombin clots revealed the dose-dependent angiogenesis inhibitory effect of heparin, steroid [19] among a number of drugs [59, 60]. This model can sustain long term study for up to 15 days [19].

8.4.3.3

Tumor Invasion

3D culture systems are essential for anti-metastasis drug assays, because it is risky to extrapolate drug effects on cancer cells in 2D migration to that on the same cells undergoing 3D invasion. For example, for a drug that showed reversible migrationinhibitory effect on 2D cell migration failed to show anti-invasion capacity in a 3D model, where cell aggregates confronted precultured fragments of embryonic chick heart [61]. Similarly, inconsistent drug effects were detected between 2D migration and 3D tissue invasion of glioma [22]. 3D multicellular culture models enabled the examination of malignant tissue invasion process. Gent University in Belgium has pioneered anti-invasion studies by mimicking the process that tumor cells infiltrate a neighboring normal tissue in 3D. Mareel et al. [21] studied how fibrosarcoma cell aggregates invaded into an embryonic chick heart-derived 3D cell culture system, with or without several antimigratory drugs. The drugs were tested in two dosages, to achieve minimal effect and complete inhibition of tumor invasion, respectively. Cell proliferation, directional migration and microtubules assembly were also monitored during drug treatment. They discovered positive correlations between microtubules assembly and invasiveness, which is in line with the results from subcutaneous implantation using mouse models [56]. Encouraged by the findings from modeling tumor invasion using animal organ fragments, organotypic cultures of human tissues are built in a similar way. One of such organotypic cultures was achieved by confronting lung cancer cell clusters with non-malignant tissues including normal epithelium, normal stroma or tumor stroma. Since both tissue blocks were a few hundred micrometers in diameter, the initial contact between the two blocks were often supported with an agar gel [57]. These complex 3D organotypic cultures recreated tumor-stroma co-invasion for in vitro anti-invasion drug assays. Plating spheroids of avascular brain tumor cells

162

W. Sun et al.

and human cerebral endothelial cells (HCEC) spheroids next to each other on agar is such a 3D co-culture model. Both tumor cell and EC spheroids were reported to invade into each other in control. But the EC migration was completely halted after the drug treatment [58]. Some more fundamental studies with a focus on the mechanisms of 3D cancer cell migration may facilitate the identification of potential anti-metastasis drugs. So far, there have been reports on the role of MMPs [59,60], cell adhesive receptors [61] and cytoskeletal contractility mediator such as Rho kinase [62].

8.5

Comparing 3D Assays of Different Structural and Physical Properties

In working with 3D assays, an important question to ask is: Are the mechanisms contributing to cancer progression/metastasis found in 3D assays universal, or will they be dependent on the engineered 3D model? To improve the reliability of these relatively new drug assays, scientists are now driven to investigate the effects of the structural and physical properties of 3D culture models. Synthetic hydrogel material and natural ECM protein may function differently. 3D culture systems built with a self-assembling peptide, RADA16, was compared with collagen I and Matrigel regarding how they recapitulated the malignancy phenotype of human breast-cancer cells. The differences in cell morphology, proliferation rate and migration were attributed to the components and structural properties of the biomaterials [63]. There are also differences between types of natural ECM proteins. An ovarian cancer 3D invasion assay found that matrices built from collagen I but not Matrigel provided an MMP-dependent barrier to prevent cell penetration. The results were consistent with that from 2D Transwell assay [64]. Even in the same type of hydrogel containing collagen-I, the differences in matrix structure-mechanics also have major influences on cells. As illustrated in Fig. 8.4, angiogenesis is sensitive to the density of collagen networks, as indicated by the endothelial cells penetration depth and network thickness [14]. The sources of the hydrogel materials further complicate the issue because some of the gel properties vary with sources. Depending on how the collagen was extracted from animals [59], whether the collagen was secreted by healthy or cancer-associated fibroblasts [61] and the collagen network micro-architecture and mechanical properties [65], cell migrated differently. Some suggested reasons are that the contribution of some cell migration mechanisms may vary with the tissue environment, with enzymatic collagen degradation being a typical example. MMP activity was essential for cancer cell invasion into acid-extracted rat tail collagen I containing impaired telopeptide regions, but not for matrices made from or bovine pepsin-digested Vitrogen [59, 64]. It further implies that the interventions of these molecular mechanisms will have inconsistent effectiveness depending on the microenvironment in collagen models.

8 Three‐Dimensional Tumor Models

163

Fig. 8.4 3D endothelial network structures formed in flexible (a, c) and rigid (b, d) collagen-I gels at day 7 of culture. Confocal images were recorded starting from cells on the gel surface downwards at 5 mm intervals. In (a, b) the images of the 3D networks were reconstructed using images taken at depths more than 15 am into the gel to avoid the cell monolayer (c, d) were lateral views of the 3D networks. Scale bars, 100 mm (a, c). (Courtesy of Tissue Engineering [14])

In vitro tissue mechanics has been found to determine the fate of stem cell differentiation [66, 67], intercellular interactions [68, 69], malignancy development [11, 70] and migration [13, 71]. Drug sensitivity has also shown certain dependence on the stiffness of 2D substrates [72]. However, whether mechanics plays similarly significant roles in drug testing based on 3D models remains to be discovered.

8.6

Conclusions and Future Outlook

With many features that are distinctly different, in vitro 3D tumor models fare better than 2D cultures in recapitulating cell-cell and cell-ECM interactions as that found in vivo. These interactions may largely interfere with the molecular targets, uptake of cells and cellular metabolism of drugs, and thus play an important role in drug assays. Morphogenesis, angiogenesis and tissue invasion occur exclusively in 3D, thus 3D cellular assays become a must to reproduce these processes in vitro. As reflected

164

W. Sun et al.

in the recent literature, 3D models show advantages in testing the sensitivity of anti-neoplastic drugs, in searching for radio- and chemo-sensitizers, for preventing angiogenesis/metastasis and in selecting effective treatments for individual patients. Some organotypic 3D culture models with substantial complexity make efficient drug assays, because these models are versatile in monitoring invasion, angiogenesis, and stromal activation associated with the malignancy development of cancers. There are a number of issues to be resolved in order to improve these 3D tumor models for further applications in cancer therapeutics. For example, maintaining a steady oxygen / nutrients supply and sustaining a micro-scale fluid flow in the newly formed vessels are now limiting factors for prolonged 3D culture [34] and yet are critical for studying drug effects on malignant morphology development and angiogenesis. In addition, reproducibility needs be improved through standardized protocol/reagents although this may not completely be achievable [25]. Since the structure and physical property of 3D substrates can determine cell fate and elicit some cell behaviors such as adhesion and migration, properly modulated 3D ECM models are particularly important for developing anti-migration therapies [30,73]. In the same way, mechanical conditions of the synthesized tissues should also be carefully tuned for more accurate anti-cancer drug assays using 3D tumor models.

References 1. Mazzoleni G, Di Lorenzo D, Steimberg N. 2009. Modelling tissues in 3D: the next future of pharmaco-toxicology and food research? Genes Nutr 4(1):13–22. 2. Padron JM, van der Wilt CL, Smid K, Smitskamp-Wilms E, Backus HH, Pizao PE, Giaccone G, Peters GJ. 2000. The multilayered postconfluent cell culture as a model for drug screening. Crit Rev Oncol Hematol 36(2–3):141–57. 3. Mitrofanova E, Hagan C, Qi J, Seregina T, Link C, Jr. 2003. Sodium iodide symporter/ radioactive iodine system has more efficient antitumor effect in three-dimensional spheroids. Anticancer Res 23(3B):2397–404. 4. Boyd M, Cunningham SH, Brown MM, Mairs RJ, Wheldon TE. 1999. Noradrenaline transporter gene transfer for radiation cell kill by 131I meta-iodobenzylguanidine. Gene Ther 6(6):1147–52. 5. Kim BW, Lee ER, Min HM, Jeong HS, Ahn JY, Kim JH, Choi HY, Choi H, Kim EY, Park SP and others. 2008. Sustained ERK activation is involved in the kaempferol-induced apoptosis of breast cancer cells and is more evident under 3-D culture condition. Cancer Biol Ther 7(7):1080–9. 6. Horning JL, Sahoo SK, Vijayaraghavalu S, Dimitrijevic S, Vasir JK, Jain TK, Panda AK, Labhasetwar V. 2008. 3-D tumor model for in vitro evaluation of anticancer drugs. Mol Pharm 5(5):849–62. 7. Murray MR, Peterson ER, Hirschberg E, Pool JL. 1954. Metabolic and chemotherapeutic investigation of human glioblastoma in vitro. Ann N Y Acad Sci 58(7):1147–71. 8. Fischbach C, Chen R, Matsumoto T, Schmelzle T, Brugge JS, Polverini PJ, Mooney DJ. 2007. Engineering tumors with 3D scaffolds. Nat Methods 4(10):855–60. 9. Ingber DE. 2008. Can cancer be reversed by engineering the tumor microenvironment? Semin Cancer Biol 18(5):356–64.

8 Three‐Dimensional Tumor Models

165

10. Levental KR, Yu H, Kass L, Lakins JN, Egeblad M, Erler JT, Fong SF, Csiszar K, Giaccia A, Weninger W and others. 2009. Matrix crosslinking forces tumor progression by enhancing integrin signaling. Cell 139(5):891–906. 11. Paszek MJ, Zahir N, Johnson KR, Lakins JN, Rozenberg GI, Gefen A, Reinhart-King CA, Margulies SS, Dembo M, Boettiger D and others. 2005. Tensional homeostasis and the malignant phenotype. Cancer Cell 8(3):241–54. 12. Pedersen JA, Swartz MA. 2005. Mechanobiology in the third dimension. Ann Biomed Eng 33(11):1469–90. 13. Zaman MH, Trapani LM, Sieminski AL, Mackellar D, Gong H, Kamm RD, Wells A, Lauffenburger DA, Matsudaira P. 2006. Migration of tumor cells in 3D matrices is governed by matrix stiffness along with cell-matrix adhesion and proteolysis. Proc Natl Acad Sci U S A 103(29):10889–94. 14. Yamamura N, Sudo R, Ikeda M, Tanishita K. 2007. Effects of the mechanical properties of collagen gel on the in vitro formation of microvessel networks by endothelial cells. Tissue Eng 13(7):1443–53. 15. Kohno T, Matsutani M, Hoshino T, Takakura K. 1985. Effects of anticancer drugs on multicellular spheroid of 9L rat brain tumor. No To Shinkei 37(10):991–7. 16. Sutherland RM. 1988. Cell and environment interactions in tumor microregions: the multicell spheroid model. Science 240(4849):177–84. 17. Kim YJ, Bae HI, Kwon OK, Choi MS. 2009. Three-dimensional gastric cancer cell culture using nanofiber scaffold for chemosensitivity test. Int J Biol Macromol 45(1):65–71. 18. Hoffman RM. 1991. Three-dimensional histoculture: origins and applications in cancer research. Cancer Cells 3(3):86–92. 19. Jung SP, Siegrist B, Wade MR, Anthony CT, Woltering EA. 2001. Inhibition of human angiogenesis with heparin and hydrocortisone. Angiogenesis 4(3):175–86. 20. Lafleur MA, Handsley MM, Knauper V, Murphy G, Edwards DR. 2002. Endothelial tubulogenesis within fibrin gels specifically requires the activity of membrane-type-matrix metalloproteinases (MT-MMPs). J Cell Sci 115(Pt 17):3427–38. 21. Mareel MM, Storme GA, De Bruyne GK, Van Cauwenberge RM. 1982. Vinblastine, Vincristine and Vindesine: anti-invasive effect on MO4 mouse fibrosarcoma cells in vitro. Eur J Cancer Clin Oncol 18(2):199–210. 22. Terzis AJ, Fiskerstrand T, Refsum H, Ueland PM, Arnold H, Bjerkvig R. 1993. Proliferation, migration and invasion of human glioma cells exposed to antifolate drugs. Int J Cancer 54(1):112–8. 23. Grinnell F, Rocha LB, Iucu C, Rhee S, Jiang H. 2006. Nested collagen matrices: a new model to study migration of human fibroblast populations in three dimensions. Exp Cell Res 312(1):86–94. 24. Nakagawa H, Yoshioka K, Miyahara E, Fukushima Y, Tamura M, Itoh K. 2005. Intrathecal administration of Y-27632, a specific rho-associated kinase inhibitor, for rat neoplastic meningitis. Mol Cancer Res 3(8):425–33. 25. Lee J, Cuddihy MJ, Kotov NA. 2008. Three-dimensional cell culture matrices: state of the art. Tissue Eng Part B Rev 14(1):61–86. 26. Huisken J, Swoger J, Del Bene F, Wittbrodt J, Stelzer EH. 2004. Optical sectioning deep inside live embryos by selective plane illumination microscopy. Science 305(5686):1007–9. 27. Keller PJ, Stelzer EH. 2010. Digital scanned laser light sheet fluorescence microscopy. Cold Spring Harb Protoc 2010(5):pdb top78. 28. Chen C, Betz MW, Fisher JP, Paek A, Chen Y. 2010. Macroporous hydrogel scaffolds and their characterization by optical coherence tomography. Tissue Eng Part C Methods. doi:10.1089/ten.tec.2010.0072 29. Levitz D, Hinds MT, Ardeshiri A, Hanson SR, Jacques SL. 2010. Non-destructive label-free monitoring of collagen gel remodeling using optical coherence tomography. Biomaterials 31(32):8210–7.

166

W. Sun et al.

30. Decaestecker C, Debeir O, Van Ham P, Kiss R. 2007. Can anti-migratory drugs be screened in vitro? A review of 2D and 3D assays for the quantitative analysis of cell migration. Med Res Rev 27(2):149–76. 31. Sahai E. 2007. Illuminating the metastatic process. Nat Rev Cancer 7(10):737–49. 32. Wolf K, Friedl P. 2009. Mapping proteolytic cancer cell-extracellular matrix interfaces. Clin Exp Metastasis 26(4):289–98. 33. Green KL, Southgate TD, Mulryan K, Fairbairn LJ, Stern PL, Gaston K. 2006. Diffusible VP22-E2 protein kills bystander cells and offers a route for cervical cancer gene therapy. Hum Gene Ther 17(2):147–57. 34. Griffith LG, Swartz MA. 2006. Capturing complex 3D tissue physiology in vitro. Nat Rev Mol Cell Biol 7(3):211–24. 35. Kuh HJ, Jang SH, Wientjes MG, Weaver JR, Au JL. 1999. Determinants of paclitaxel penetration and accumulation in human solid tumor. J Pharmacol Exp Ther 290(2):871–80. 36. Hicks KO, Fleming Y, Siim BG, Koch CJ, Wilson WR. 1998. Extravascular diffusion of tirapazamine: effect of metabolic consumption assessed using the multicellular layer model. Int J Radiat Oncol Biol Phys 42(3):641–9. 37. Kyle AH, Huxham LA, Chiam AS, Sim DH, Minchinton AI. 2004. Direct assessment of drug penetration into tissue using a novel application of three-dimensional cell culture. Cancer Res 64(17):6304–9. 38. An Z, Kavanoor K, Choy ML, Kaufman LJ. 2009. Polyelectrolyte microcapsule interactions with cells in two- and three-dimensional culture. Colloids Surf B Biointerfaces 70(1):114–23. 39. Kobayashi H, Man S, Graham CH, Kapitain SJ, Teicher BA, Kerbel RS. 1993. Acquired multicellular-mediated resistance to alkylating agents in cancer. Proc Natl Acad Sci U S A 90(8):3294–8. 40. Muir CP, Adams MA, Graham CH. 2006. Nitric oxide attenuates resistance to doxorubicin in three-dimensional aggregates of human breast carcinoma cells. Breast Cancer Res Treat 96(2):169–76. 41. Harisi R, Dudas J, Nagy-Olah J, Timar F, Szendroi M, Jeney A. 2007. Extracellular matrix induces doxorubicin-resistance in human osteosarcoma cells by suppression of p53 function. Cancer Biol Ther 6(8):1240–6. 42. Kerbel RS, St Croix B, Florenes VA, Rak J. 1996. Induction and reversal of cell adhesiondependent multicellular drug resistance in solid breast tumors. Hum Cell 9(4):257–64. 43. Croix BS, Rak JW, Kapitain S, Sheehan C, Graham CH, Kerbel RS. 1996. Reversal by hyaluronidase of adhesion-dependent multicellular drug resistance in mammary carcinoma cells. J Natl Cancer Inst 88(18):1285–96. 44. Zelenkov P, Baumgartner R, Bise K, Heide M, Meier R, Stocker S, Sroka R, Goldbrunner R, Stummer W. 2007. Acute morphological sequelae of photodynamic therapy with 5-aminolevulinic acid in the C6 spheroid model. J Neurooncol 82(1):49–60. 45. Wientjes MG, Pretlow TG, Badalament RA, Burgers JK, Au JL. 1995. Histocultures of human prostate tissues for pharmacologic evaluation. J Urol 153(4):1299–302. 46. Wang CY, Zhong WB, Chang TC, Lai SM, Tsai YF. 2002. Tumor necrosis factor alpha induces three-dimensional cytomorphologic differentiation of human anaplastic thyroid carcinoma cells through activation of nuclear factor kappaB. Cancer 95(9):1827–33. 47. Debnath J, Walker SJ, Brugge JS. 2003. Akt activation disrupts mammary acinar architecture and enhances proliferation in an mTOR-dependent manner. J Cell Biol 163(2): 315–26. 48. Yao LQ, Feng YJ, Ding JX, Xu CJ, Jin HY, Yin LH. 2005. [Primary study of vasculogenic mimicry induced by hypoxia in epithelial ovarian carcinoma]. Zhonghua Fu Chan Ke Za Zhi 40(10):662–5. 49. Wang XG, Fu J, Zheng H, Li GX, Luo RC. 2006. Effects of pirarubicin chemotherapy combined with hyperthermia on gastric cancer in vitro. Nan Fang Yi Ke Da Xue Xue Bao 26(10):1487–90.

8 Three‐Dimensional Tumor Models

167

50. Folkman J. 1971. Tumor angiogenesis: therapeutic implications. N Engl J Med 285(21): 1182–6. 51. Shekhar MP, Werdell J, Tait L. 2000. Interaction with endothelial cells is a prerequisite for branching ductal-alveolar morphogenesis and hyperplasia of preneoplastic human breast epithelial cells: regulation by estrogen. Cancer Res 60(2):439–49. 52. Mesri M, Morales-Ruiz M, Ackermann EJ, Bennett CF, Pober JS, Sessa WC, Altieri DC. 2001. Suppression of vascular endothelial growth factor-mediated endothelial cell protection by survivin targeting. Am J Pathol 158(5):1757–65. 53. Stevenson DP, Collins WP, Farzaneh F, Hata K, Miyazaki K. 1998a. Thymidine phosphorylase activity and prodrug effects in a three-dimensional model of angiogenesis: implications for the treatment of ovarian cancer. Am J Pathol 153(5):1573–8. 54. Stevenson DP, Milligan SR, Collins WP. 1998b. Effects of platelet-derived endothelial cell growth factor/thymidine phosphorylase, substrate, and products in a three-dimensional model of angiogenesis. Am J Pathol 152(6):1641–6. 55. Park HJ, Kong D, Iruela-Arispe L, Begley U, Tang D, Galper JB. 2002. 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors interfere with angiogenesis by inhibiting the geranylgeranylation of RhoA. Circ Res 91(2):143–50. 56. Liu JX, Wang XG, Fu J, Luo RC. 2006. Bevacizumab (Avastin) inhibits lung cancerinduced pulmonary microvascular angiogenesis. Nan Fang Yi Ke Da Xue Xue Bao 26(7): 1027–9; 1043. 57. Jiang C, Ganther H, Lu J. 2000. Monomethyl selenium – specific inhibition of MMP-2 and VEGF expression: implications for angiogenic switch regulation. Mol Carcinog 29(4):236–50. 58. Lamy S, Beaulieu E, Labbe D, Bedard V, Moghrabi A, Barrette S, Gingras D, Beliveau R. 2008. Delphinidin, a dietary anthocyanidin, inhibits platelet-derived growth factor ligand/ receptor (PDGF/PDGFR) signaling. Carcinogenesis 29(5):1033–41. 59. Hornick CA, Myers A, Sadowska-Krowicka H, Anthony CT, Woltering EA. 2003. Inhibition of angiogenic initiation and disruption of newly established human vascular networks by juice from Morinda citrifolia (noni). Angiogenesis 6(2):143–9. 60. Stafford SJ, Schwimer J, Anthony CT, Thomson JL, Wang YZ, Woltering EA. 2005. Colchicine and 2-methoxyestradiol Inhibit Human Angiogenesis. J Surg Res 125(1):104–8. 61. Schroyens W, Tchao R. 1985. Migration inhibition of an epithelial cell line by s-Con A and the effect on its invasiveness. Virchows Arch B Cell Pathol Incl Mol Pathol 49(2):175–81. 62. Meyvisch C, Storme GA, Bruyneel E, Mareel MM. 1983. Invasiveness and tumorigenicity of MO4 mouse fibrosarcoma cells pretreated with microtubule inhibitors. Clin Exp Metastasis 1(1):17–28. 63. Fjellbirkeland L, Laerum OD, Eide GE, Bjerkvig R. 1998. Invasiveness by lacZ transfected non-small-cell lung cancer cells into human bronchial tissues in vitro. Lung Cancer 21(1):7–19. 64. Wagner S, Fueller T, Hummel V, Rieckmann P, Tonn JC. 2003. Influence of VEGF-R2 inhibition on MMP secretion and motility of microvascular human cerebral endothelial cells (HCEC). J Neurooncol 62(3):221–31. 65. Sabeh F, Shimizu-Hirota R, Weiss SJ. 2009. Protease-dependent versus -independent cancer cell invasion programs: three-dimensional amoeboid movement revisited. J Cell Biol 185(1):11–9. 66. Wolf K, Alexander S, Schacht V, Coussens LM, von Andrian UH, van Rheenen J, Deryugina E, Friedl P. 2009. Collagen-based cell migration models in vitro and in vivo. Semin Cell Dev Biol 20(8):931–41. 67. Castello-Cros R, Khan DR, Simons J, Valianou M, Cukierman E. 2009. Staged stromal extracellular 3D matrices differentially regulate breast cancer cell responses through PI3K and beta1-integrins. BMC Cancer 9:94.

168

W. Sun et al.

68. Provenzano PP, Inman DR, Eliceiri KW, Trier SM, Keely PJ. 2008. Contact guidance mediated three-dimensional cell migration is regulated by Rho/ROCK-dependent matrix reorganization. Biophys J 95(11):5374–84. 69. Mi K, Wang G, Liu Z, Feng Z, Huang B, Zhao X. 2009. Influence of a self-assembling peptide, RADA16, compared with collagen I and Matrigel on the malignant phenotype of human breast-cancer cells in 3D cultures and in vivo. Macromol Biosci 9(5):437–43. 70. Sodek KL, Brown TJ, Ringuette MJ. 2008. Collagen I but not Matrigel matrices provide an MMP-dependent barrier to ovarian cancer cell penetration. BMC Cancer 8:223. 71. Friedl P, Wolf K. 2010. Plasticity of cell migration: a multiscale tuning model. J Cell Biol 188(1):11–9. 72. Alcaraz J, Xu R, Mori H, Nelson CM, Mroue R, Spencer VA, Brownfield D, Radisky DC, Bustamante C, Bissell MJ. 2008. Laminin and biomimetic extracellular elasticity enhance functional differentiation in mammary epithelia. EMBO J 27(21):2829–38. 73. Engler AJ, Sen S, Sweeney HL, Discher DE. 2006. Matrix elasticity directs stem cell lineage specification. Cell 126(4):677–89. 74. Guo WH, Frey MT, Burnham NA, Wang YL. 2006. Substrate rigidity regulates the formation and maintenance of tissues. Biophys J 90(6):2213–20. 75. Winer JP, Oake S, Janmey PA. 2009. Non-linear elasticity of extracellular matrices enables contractile cells to communicate local position and orientation. PLoS One 4(7):e6382. 76. Kumar S, Weaver VM. 2009. Mechanics, malignancy, and metastasis: the force journey of a tumor cell. Cancer Metastasis Rev 28(1–2):113–27. 77. Harley BA, Freyman TM, Wong MQ, Gibson LJ. 2007. A new technique for calculating individual dermal fibroblast contractile forces generated within collagen-GAG scaffolds. Biophys J 93(8):2911–22. 78. Rehfeldt F, Engler AJ, Eckhardt A, Ahmed F, Discher DE. 2007. Cell responses to the mechanochemical microenvironment–implications for regenerative medicine and drug delivery. Adv Drug Deliv Rev 59(13):1329–39. 79. Raeber GP, Lutolf MP, Hubbell JA. 2005. Molecularly engineered PEG hydrogels: a novel model system for proteolytically mediated cell migration. Biophys J 89(2):1374–88.

Chapter 9

Mechanobiology of Epidermal Keratinocytes: Desmosomes, Hemidesmosomes, Keratin Intermediate Filaments, and Blistering Skin Diseases John C. Selby This chapter is part of Section III: Mechano-pathology of Disease

Abstract Despite continuous advances in our understanding of the human epidermis and its principal cell type, the keratinocyte, the biophysical mechanisms that endow these cells with such remarkable mechanobiological properties remain largely unknown. Towards this end, this chapter serves as an eclectic but didactic introduction to the biology, pathology, and rheology of human epidermal keratinocytes. The initial discussion is primarily morphological in content, including a histological illustration of the human epidermis and an ultrastructural description of the various cell-cell and cell-matrix anchoring junctions that are expressed by human epidermal keratinocytes in vivo. The associations between intercellular adhesion, intracellular resistance to mechanical deformation, and keratinocyte differentiation are reviewed. Next, the pathophysiologies of several different autoimmune, genetic, and infectious blistering skin diseases are explored, including epidermolysis bullosa simplex, pemphigus vulgaris, and the staphylococcal scalded skin syndrome, among others. These clinical disorders serve as poignant examples as to how mechanical failures within the cytoarchitectural system of keratin intermediate filaments, desmosomes, and hemidesmosomes can give rise to human disease states. The chapter concludes with a brief accounting of past and current experimental attempts to characterize the mechanobiology of human epidermal keratinocytes with rheological measurements across a large range of experimental length scales. Keywords Keratinocyte Keratin intermediate filament Desmosome Hemidesmosome Adherens junctions Focal adhesion Pemphigus Epidermolysis bullosa

J.C. Selby (*) Department of Mechanical Science and Engineering, and the College of Medicine, University of Illinois at Urbana-Champaign, 611 West Park Street, Urbana, IL 61801, USA e-mail: [email protected]

A. Wagoner Johnson and Brendan A.C. Harley (eds.), Mechanobiology of Cell-Cell and Cell-Matrix Interactions, DOI 10.1007/978-1-4419-8083-0_9, # Springer Science+Business Media, LLC 2011

169

170

9.1

J.C. Selby

Introduction

Skin, or integument, represents the largest human organ, serving critical roles in fluid homeostasis, thermoregulation, immune surveillance, and vitamin D synthesis. In addition to these functions with distinctive biological connotations, skin is also a robust mechanical organ. The outermost component of skin, the epidermis, routinely undergoes large strains while maintaining a physiological barrier against noxious, infectious, and physical insults. We have all encountered, to some degree, the painful consequences associated with failure of cutaneous barrier function: a scraped knee, a peeling sunburn, or a fingertip blister. Of more clinical significance, consider people who have endured traumatic burns or those who experience recurring skin ulcerations secondary to a chronic health condition like diabetes mellitus. Perhaps most devastating of all, consider those individuals who suffer from inherited or acquired skin fragility diseases – their epidermis cannot withstand even the most benign of mechanical deformations before barrier function becomes compromised. Despite continuous advances in our understanding of the molecular biology and clinicopathology of the epidermis and its principal cell type, the keratinocyte, from a standpoint of mechanics, it remains largely unknown how these cells proliferate, stratify, and differentiate into such an active mechanical tissue. This chapter is posited as an eclectic but didactic introduction to the biology, pathology, and rheology of human epidermal keratinocytes. Rather than a comprehensive review of these topics, the work is organized as a cursory overview intended to stimulate interest in an area of interdisciplinary research that spans molecular biology, clinical medicine, biophysics, and mechanics. The information presented here is geared towards undergraduate, graduate, post-doctoral, and medical students with limited background knowledge in any of the three major areas of discussion. In Sects. 9.2 and 9.3, the topic of keratinocyte biology is presented from a morphological perspective, with the discussion focusing on a histological illustration of the human epidermis and an ultrastructural description of the cytoskeletal protein networks and anchoring junctions deterministic of keratinocyte cytoarchitecture. The known associations between intercellular adhesion, intracellular resistance to mechanical deformation, and keratinocyte differentiation are explored. In Sect. 9.4, the pathophysiologies of several different types of blistering skin disease are reviewed as they pertain to structural failures in cell-cell and cell-matrix anchoring junctions. These clinical disorders serve as poignant examples as to why continued investigation of keratinocyte mechanobiology is both interesting and relevant. Lastly, in Sect. 9.5, the rheological behavior of epidermal keratinocytes is examined from an engineering perspective. Biomechanical techniques probing keratinocyte load-deformation behavior are surveyed at macroscopic, microscopic, mesoscopic, and nanoscopic length scales. The chapter concludes with a brief summary given in Sect. 9.6.

9 Mechanobiology of Epidermal Keratinocytes

9.2

171

Human Skin – Structure and Function

As “the usual sequence of biological knowledge is from the anatomical to the physiological [1],” let us begin this discourse with a morphological description of the integumentary system. Integument is the largest and heaviest single organ of the human body, consisting of skin and its related epidermal derivatives: nails, hair, sweat glands, sebaceous glands, and the mammary gland [2–6]. In an average adult male, the integument covers roughly 2.0 m2 of surface area and accounts for nearly 16% of total body weight [7]. All vital functions of skin are regulated by cells and structures localized within a two-layer anatomy, the dermis and the epidermis [5, 6] (cf. Fig. 9.1). Depending on the relative thickness of the epidermis, human skin can

Fig. 9.1 Human skin has two major layers: dermis and epidermis. Both layers are present in this hematoxylin and eosin (H&E) stained specimen of thick human skin found near the nail bed of the thumb. The dermis is composed of dense irregular collagenous connective tissue; the epidermis is a stratified squamous keratinized epithelium. Note the presence of the interdigitating dermalepidermal junction, as partially indicated by the dashed line. Figure created from images available in ref. [8], with permission

172

J.C. Selby

be classified into two types. With a 400–600 mm thick epidermis, thick skin covers the palms of the hands and soles of the feet [7]. Thin skin – with a 75–150 mm thick epidermis – covers the remainder of the body [7]. Between the dermis and the epidermis, one finds the dermal-epidermal junction, otherwise known as the basement membrane zone. Technically, the term basement membrane refers exclusively to the thin mat of extracellular matrix material that separates the two skin layers. In Fig. 9.1, note that the dermal-epidermal junction is not a flat plane, but instead conforms to a series of aperiodic undulations comprised of upward projections of the dermis called papillae and downward evaginations of the epidermis known as rete ridges (sometimes called interpapillary pegs). Three-dimensional variations in the frequency and depth of these microscopic interdigitations create regional differences in the mechanical adhesion between the epidermis and dermis. The macroscopic pattern of ridges and furrows that can be viewed grossly on the external surface of the skin give rise to fingerprints (or dermatoglyphics). Histologically, the dermis is classified as a dense irregular connective tissue, consisting mainly of extracellular matrix material. As its classification implies, the dermis connects the epidermis to the hypodermis, a subcutaneous tissue overlying the deep fascial layers that envelop and compartmentalize other gross anatomical structures of the body. The primary structural material of the dermal extracellular matrix is type I collagen. Elastin, fibronectin, and type III collagen are also present, though in smaller amounts. Proteoglycan aggregates fill the interstitial volumes between these fibrous proteins. Because they can absorb large quantities of water, these proteoglycans not only help to maintain the hydration state of the dermis, they also provide the tissue with an innate ability to resist compressive loads. Fibroblasts are the predominant cellular constituent. Responsible for the continuous synthesis, degradation, and remodeling of extracellular matrix material, fibroblast activities are critical to both normal skin growth and wound repair. In Fig. 9.2, observe how dermal fibroblasts sparsely populate a sea of collagen fiber bundles. Especially note how fibroblasts within the dermis are mostly in contact with extracellular matrix material. Under most circumstances, fibroblasts are not mechanically coupled to neighboring cells, i.e., multicellular aggregates of fibroblasts are not observed in normal dermis [9–12]. Additionally, note the large network of blood vessels coursing throughout the specimen. The dermal vasculature serves the entire integumentary system, and plays a significant role in thermoregulation [3]. Small numbers of macrophages and mast cells resident within the dermis are involved in immune surveillance and the inflammatory response, respectively. The epidermis, as can be viewed in Fig. 9.3, is histologically classified as a stratified squamous keratinized epithelium, composed principally of cells known as keratinocytes. Although they collectively account for only one-fortieth of the overall thickness of skin, epidermal keratinocytes provide the all-important cutaneous barrier function. Keratinocytes also produce various matrix metalloproteinases that aid in the turnover of dermal extracellular matrix during wound healing. Additional cell types found within the epidermis (not shown in Fig. 9.3) include melanocytes, Langerhans cells, and Merkel cells. These cells play important physiological roles in melanin production, immune surveillance, and tactile sensation, respectively.

9 Mechanobiology of Epidermal Keratinocytes

173

Fig. 9.2 In this view of dermis from an H&E stained specimen of human thick skin, several fibroblast nuclei, highlighted with block arrows, are scattered among an abundance of intensely stained bundles of collagen fibers. Most dermal fibroblasts are not in direct contact with neighboring fibroblasts – they are mechanically coupled to extracellular matrix protein(s). Figure created from images available in ref. [8], with permission

Within the epidermis, keratinocytes are categorized as belonging to one of five distinct layers: the stratum basale, stratum spinosum, stratum granulosum, stratum lucidum, and stratum corneum. Upon closer examination of Fig. 9.3, observe that most keratinocytes are intimately surrounded by other keratinocytes. In contrast to the dermis, the epidermis is nearly void of extracellular matrix material. As one might expect, these two critical differences in histology distinguish the mechanical behavior of the epidermis from that of the dermis. Keratinocytes within the stratum basale represent an exception. In addition to adjacent cells, basal keratinocytes are also mechanically coupled to the basement membrane at the dermal-epidermal junction. Given this rudimentary background information on the structure and function of human skin, consider how epidermal morphology relates to keratinocyte biology. As previously mentioned, keratinocytes can be histologically classified as belonging to one of five discrete layers of the epidermis (cf. Fig. 9.3). In terms of the underlying cell biology, each layer actually represents a different state of keratinocyte differentiation. Also termed cornification or keratinization, the process of keratinocyte differentiation encompasses all the biochemical and morphological events related to epidermal barrier formation and renewal. The process begins with the proliferation of keratinocyte stem cells that reside within the stratum basale. Following cell division, committed daughter cells migrate to layers increasingly superficial to the dermis, undergoing major changes

174

J.C. Selby

Fig. 9.3 Four layers can be identified within this H&E stained specimen of epidermis from thick human skin: stratum basale, stratum spinosum, stratum granulosum, and stratum corneum. A distinct stratum lucidum is not present. Individual keratinocytes are in direct contact with neighboring keratinocytes, while intercellular space is nearly void of extracellular matrix. Block arrow indicates keratohyalin granules present within the stratum granulosum. Figure created from images available in ref. [8], with permission

in protein expression within the stratum spinosum, stratum granulosum, stratum lucidum, and stratum corneum. The corneocyte, a terminally differentiated keratinocyte, eventually reaches the outermost layer of the epidermis, the stratum corneum, and is sloughed from the surface by intercellular fracture of individual or small aggregates of cells [13]. The normal process of keratinocyte differentiation requires 15–30 days, depending on body region [7]. Note the dramatic morphological changes that accompany keratinocyte differentiation, as can be viewed in Fig. 9.3.1 During cornification, an average keratinocyte begins as a high cuboidal cell in the stratum basale, roughly 6.0 mm in diameter, 15–30 mm tall, and eventually flattens to an anucleated squame in the 1

Specific biochemical changes underlying the process of keratinocyte differentiation will be discussed in Sect. 9.3.3.

9 Mechanobiology of Epidermal Keratinocytes

175

stratum corneum, 30 mm in diameter, 0.8 mm thick [1]. Prior to desquamation, each corneocyte projects an area that covers approximately 25 basal cells [1]. At some anatomical sites, e.g., the inner arm, thorax, and abdomen, the horny layer of thin skin appears to be a highly ordered structure, often described as a lattice of polyhedral squames arranged as interdigitating vertical columns that form a coherent membrane [14]. The stratum corneum of calloused thick skin is thought to be much less organized [13]. For a more elaborate introduction to the basic structure and function of human skin, see refs. [4–6].

9.3

Keratinocyte Biology

Today, many clinicians, scientists, and engineers would argue that the macroscopic load-bearing capacity of human skin is dominated by the dermis and the complex interactions that take place between collagen, elastin, proteoglycan aggregates, and fibroblasts when undergoing an applied deformation [13, 15]. Moreover, many of these researchers would also attest to the idea that epidermal corneocytes serve as the principal physical barrier of skin, providing resistance to chemical, mechanical, and thermal insults [5, 6, 13]. But what about keratinocytes in the stratum spinosum and stratum basale, i.e., the basal and subcorneal layers of the epidermis? These cells are also robustly capable of withstanding large mechanical strains during normal physiological function. What characterizes their deformation response? To answer this question, one needs a thorough understanding of the proteins that constitute and regulate the keratinocyte cytoskeleton as well as those proteins responsible for the formation of cell-cell and cell-matrix anchoring junctions. “The question of what keeps cells together is one of the most interesting in all biology [1].”

9.3.1

Cytoskeletal Network Proteins

Keratinocytes, like most nucleated human cells, possess a cytoskeleton composed of actin microfilaments, intermediate filaments, and microtubules. Within individual keratinocytes, these protein filament networks are intimately involved with a diverse array of cellular processes indicative of their inherent mechanical functions: whole cell locomotion, organelle arrangement, intracellular transport, and cell division. Whereas microfilaments and microtubules are the most well-known cytoskeletal components, intermediate filaments are perhaps the least studied, and many of their suggested functions remain purely speculative [16, 17]. Furthermore, as we will see in Sect. 9.4, dysfunctional intermediate filament networks within epidermal keratinocytes (and/or their corresponding anchoring junctions) can cause various types of human blistering skin diseases. Thus, a short introduction to the intermediate filament system has been included here to supplement our forthcoming discussions of keratinocyte pathology and rheology. Interested readers are urged to locate references [18–21] for more in-depth reviews of microfilaments and microtubules.

176

J.C. Selby

Under both the light and electron microscope, microfilaments, intermediate filaments, and microtubules are expressed as a seemingly disparate array of cellular structures. Yet, with respect to their molecular architecture, all three major cytoskeletal proteins share the common organizational principle that “large, complex structures are built from small, simple components [22].” The basic building block of an intermediate filament is a small monomeric fibrous protein with distinct globular head and tail regions separated by an extended rodlike a-helical domain [3, 16]. Filament polymerization initiates as two parallel monomers associate sideby-side to form a coiled-coil dimer, following which two staggered anti-parallel dimers associate to form a tetramer. As tetramers assemble into a helical array 16 dimers or 32 monomers in cross-section, a filament 10 nm in diameter is formed [16, 23]. Intermediate filaments were named so because their filament diameter was found to be smaller than a microtubule but larger than a microfilament. In this arrangement, intermediate filament networks are by definition, polymers, but they differ from most synthetic polymers in that they are wholly noncovalent assemblies. The most common morphological form of intermediate filaments in cells that express them is that of a diffuse three-dimensional network of rope-like fibers that extend from the nucleus to various degrees throughout the cell periphery [24]. In some cells, accessory proteins like filaggrin and plectin are known to facilitate the formation of large bundles of intermediate filaments within the cytoplasm [25]. Plectin is also thought to mediate cross-links between intermediate filaments and other cytoskeletal networks [26]. In comparison to microtubules and microfilaments, perhaps the most distinguishing feature of intermediate filaments is that they are not known to associate with any force-generating motor proteins. As a consequence, intermediate filaments are not involved in cellular functions associated with active mechanical processes. Once thought to be the most stable of the three cytoskeletal protein networks, intermediate filaments have recently been shown to be a dynamic element of the cytoskeleton [27–30]. More specific details regarding the intrinsic and extrinsic modulators of intermediate filament organization can be found elsewhere [25, 31]. In human cells, actin and tubulin (the monomeric precursors of microfilaments and microtubules, respectively) represent protein products derived from mostly conserved genes. Closely related, if not identical isotypes of microfilament and microtubules are expressed in epidermal keratinocytes and dermal fibroblasts. In stark contrast, intermediate filament proteins exhibit a great deal of sequence variation. The human genome is thought to contain over 65 different functional genes encoding fibrous intermediate filament protein monomers [17, 32–34]. Biochemically, this superfamily of intermediate filament genes can be subcategorized into five major classes: type I and type II keratins, type III vimentin-like filaments, type IV nuclear lamins, and type V axonal filaments [16, 32]. Of these major classes, genes encoding keratins are the most abundant. There exists at least 25 keratins genes for each type I and type II category, including keratins specific for hair and nails [35, 36]. In vivo, most intermediate filaments proteins can self-polymerize into homodimeric filament structures with a hierarchical organization as previously discussed [32].

9 Mechanobiology of Epidermal Keratinocytes

177

Keratin intermediate filaments represent an exception to this rule as they are obligate equimolar heterodimers of type I and type II proteins [32]. Given such a large family of structurally diverse proteins, it is not surprising that intermediate filaments are expressed in a tissue-type and differentiation-specific manner. In other words, not all cells express the same classes of intermediate filament proteins. Consider our ongoing comparison between the epidermal keratinocyte and the dermal fibroblast. Both cells express nuclear lamins that strengthen the inner nuclear membrane, but while the fibroblast expresses vimentin intermediate filaments in its cytoplasm, the keratinocyte expresses keratin intermediate filaments. The difference in intermediate filament expression between the keratinocyte and the fibroblast is critical: it affects the expression of cell-cell and cell-matrix anchoring junctions within these cells, and consequently, the histological characteristics of the tissues in which they reside [37–40].

9.3.2

Anchoring Junctions

The plasma membrane that bounds a typical eukaryotic cell is an inherently fragile structure. By itself, the phospholipid bilayer cannot sustain large external tractions nor is it capable of distributing applied loads to adjacent cells or extracellular matrix in a manner that would be conducive to the morphogenesis of complex multicellular organisms [16]. On the other hand, the cytoskeleton is a robust mechanical structure with a large load-bearing capacity. But if it were to be completely enveloped by the plasma membrane, the cytoskeleton would be of little functional significance in the structuring of stable tissue architectures. Evolution has found a solution to this apparent paradox in the form of specialized aggregates of transmembrane and cytoplasmic proteins that co-localize at the interface between the cell membrane and the cytoskeleton. In modern terminology, these contacts are collectively referred to as cell-cell and cell-matrix anchoring junctions. To date, there exist four histologically recognized and morphologically distinct types of anchoring junctions, namely, focal adhesion complexes, adherens junctions, desmosomes, and hemidesmosomes (cf. Table 9.1). As their descriptive name implies, anchoring junctions serve to mechanically couple the cytoskeleton to its extracellular environment at sites of well-developed adhesive contact. By most standards, anchoring junctions are primarily classified as either cell-cell or cellmatrix (or cell-ECM) adhesions, and secondarily classified by the type of cytoskeletal protein network associated with the junction.2 Examination of the molecular 2

Extensive biochemical studies have identified several classes of transmembrane proteins involved in cell-cell and cell-matrix adhesive contacts: the cadherin superfamily of proteins, integrins, selectins, and the immunoglobulin superfamily of proteins. Section 9.3.2 will focus on cadherins and integrins, since they are components of histologically defined anchoring junctions. Selectins and proteins from the immunoglobulin superfamily are involved in more transient adhesive contacts like those observed during embryological development and leukocyte extravasation. Although of great physiological significance, they are not relevant to our discussion of epidermal keratinocytes.

178

J.C. Selby

Table 9.1 Anchoring junction classification Adhesion Cytoskeletal Anchoring junction type proteina Focal adhesion Cell-matrix MF

Adherens junction Hemidesmosome

Cell-cell

MF

Cell-matrix

IF

Desmosome

Cell-cell

IF

Transmembrane protein(s) ab integrin heterodimers Cadherin homodimers ab integrin heterodimers, BP180b Desmocollin and desmoglein heterodimers

Linker protein(s) Talin, filamin, a-actinin, vinculin, paxilliln, tensin a-Catenin, b-catenin, a-actinin, vinculin Plectin, BP230c

Plakoglobin, plakophilin, desmoplakin

a

MF microfilament; IF intermediate filament BP180: 180-kDa bullous pemphigoid antigen c BP230: 230-kDa bullous pemphigoid antigen b

ultrastructure allows further characterization, namely through identification of the specific transmembrane and linker proteins that localize to the adhesive site. Cartoon molecular schematics of model cell-cell and cell-matrix anchoring junctions can be found in Fig. 9.4. Other types of cell-cell contacts expressed in epithelial tissues, specifically, tight junctions and gap junctions, are not usually categorized as anchoring junctions per se [16, 41–46].3 In the following paragraphs, we present a general overview of each of the four major types of cell-cell and cellmatrix anchoring junctions, along with a more detailed discussion of their unique manifestations within epidermal keratinocytes. Focal adhesion complexes, also called focal adhesion contacts, focal contacts, or focal adhesions, represent one of two types of anchoring junctions that serve to connect the intracellular microfilament network of a cell to its local extracellular environment. In particular, focal adhesions are used to tether microfilaments, usually in the form of a stress fiber, to an explicitly targeted component of the extracellular matrix. Members of the integrin superfamily of transmembrane proteins, in the form of heterodimers, are responsible for making the adhesive attachment to these matrix ligands [16]. The specificity of integrin binding is determined by the combination of 3

Tight junctions are expressed as anastomosing networks of sealing strands where the plasma membranes of adjacent cells in an epithelium become tightly apposed [16]. Tight junctions induce epithelial cell polarity, endowing epithelial tissues with the physiological property of selective permeability. Although microfilaments [41–43] and microtubules [44] have been implicated in the formation and regulation of tight junctions, the junctions have not been associated with any known load-bearing function. As such, tight junctions are usually not classified as a type of anchoring junction. Gap junctions are cell-cell junctions that allow the passage of small metabolites and ions between the cytoplasm of adjacent cells [16]. In cell types that are not electrically excitable, e.g., epidermal keratinocytes, gap junctions have been postulated to play a role in coordinating the epithelial differentiation and renewal process [6, 45, 46]. Gap junctions, like tight junctions, are presumed to have negligible load-bearing function.

9 Mechanobiology of Epidermal Keratinocytes

a cytoskeletal

intercellular space

proteins

transmembrane proteins

b

179

linker proteins

cytoskeletal proteins linker proteins

transmembrane proteins

extracellular matrix proteins

extracellular space α integrin

β integrin

Fig. 9.4 The diagrams above depict the generic molecular assembly of (a) cell-cell and (b) cellmatrix anchoring junctions. Both types of junctions consist of cytoskeletal, transmembrane, and linker proteins, the latter more commonly referred to as intracellular or cytoplasmic plaques. Extracellular domains of the transmembrane proteins are responsible for the adhesive contact

a and b subunits that constitute the heterodimer pair, e.g., a5b1 binds fibronectin, whereas a2b1 binds collagen [47]. Intracellularly, integrins are mechanically coupled to the actin microfilament network through a submembranous plaque of linker proteins [48]. These plaques are known to consist of more than 50 proteins, including talin, a-actinin, filamin, vinculin, paxillin, and tensin [49]. Because multiple signal transducing molecules activated by integrins have been observed to physically associate with the microfilament network near focal adhesion contacts, they are often presumed to serve a critical function in mechanochemical signal transduction [47, 49, 50]. Adherens junctions4 represent the second type of anchoring junction associated with the microfilament network, but rather than provide a cell with an adhesive

4

To avoid confusion, note that some researchers use the term adherens junction to refer to either a microfilament-associated cell-cell junction or a cell-matrix focal adhesion contact.

180

J.C. Selby

contact to an extracellular matrix protein, adherens junctions provide a mechanical coupling to neighboring cells. Although similar to a focal adhesion complex, adherens junctions form a continuous peripheral adhesion belt on the basolateral sides of cells in a simple epithelium, as opposed to discrete contacts located solely on the ventral (basal) surface. Hence, adherens junctions, or zonula adherens, are also commonly referred to as belt desmosomes. Classical calcium-dependent cadherin proteins, in the form of paired homodimers, comprise the major transmembrane component of a prototypical adherens junction [51]. Intracellular linker proteins like a-catenin, b-catenin, vinculin, and a-actinin provide a molecular bridge between the cytoplasmic domains of the cadherins and the microfilament network [16, 52]. While adherens junctions and focal adhesion complexes provide attachment sites for microfilaments, desmosomes and hemidesmosomes are junctional assemblies intracellularly coupled to intermediate filament networks. Desmosomes, or macula adherens, appear as punctate cell-cell adhesive contacts in epithelial tissues [53], and are often referred to as spot desmosomes. Molecularly, desmosomes are organized as heterodimers of the non-classical calcium-dependent cadherin transmembrane proteins desmocollin and desmoglein, which in turn are connected to a linker plaque comprised of desmoplakin, plakoglobin, and plakophilin [54–57]. Within most epithelial cells, intermediate filaments do not terminate on linker proteins, but rather appear to “loop through” the intracytoplasmic plaque [54]. Hemidesmosomes form localized cell-matrix contacts that mediate adhesion between the basal layer of cells in an epithelial tissue and the underlying basement membrane [58]. Although similar to a desmosome when viewed through the electron microscope, hemidesmosomes differ biochemically, employing a unique set of transmembrane and linker proteins. Like focal adhesions, hemidesmosomes use an integrin heterodimer as a transmembrane protein, namely the a6b4 integrin that possesses a high affinity for the matrix protein laminin 5 [58]. In addition to a6b4 integrins, the protein referred to as bullous pemphigoid antigen II (or BP180 or collagen XVII) also functions as a transmembrane protein in hemidesmosomes [58–60]. Bullous pemphigoid antigen I (or BP230) and plectin are the major constituents of the cytoplasmic plaque that bind intermediate filaments to transmembrane proteins. In the classification shown in Table 9.1, microtubules are not listed as a cytoskeletal component for any of the four major types of anchoring junctions. In wake of recent studies of cytoskeletal network crosstalk, especially interactions mediated by the plakin family of proteins [61, 62], this omission is misleading. For example, microtubules have been shown to localize at cell-matrix focal contacts [63] and cell-cell contacts [64] representative of immature adherens junctions. The putative function of microtubules at these adhesive sites has been proposed to be both structural and biochemical in nature [65–69]. Microtubules have not been observed to participate in the reorganization of intermediate filaments, linker plaques, or transmembrane proteins associated with desmosome assembly [70, 71], and microtubule involvement with hemidesmosomes is, at present, unclear. In light of these findings, one should acknowledge that the standard histological classification of anchoring junctions described here does not tell a complete story. As biologists continue to

9 Mechanobiology of Epidermal Keratinocytes

181

elucidate new structural and biochemical interconnections within the cytoskeleton, one thing remains certain: microfilament, microtubule, and intermediate filament networks do not always function as independent entities. In spite of this fact, we will proceed with the working assumption that microfilaments and intermediate filaments are the predominant cytoskeletal elements associated with the four major types of cell-cell and cell-matrix anchoring junctions. Given general background knowledge of desmosomes, focal adhesions, hemidesmosomes, and adherens junctions, let us now address some more specific details concerning their expression within epidermal keratinocytes. Figure 9.5 serves to complement the discussion. As with most adherent eukaryotic cells, focal adhesion complexes are used by keratinocytes to attach and spread on substrates in vitro. Such focal contacts are easily observed along the ventral surface of cultured cells, appearing as spots a few square micrometers in area where the normal 50 nm gap between the plasma membrane and the substrate is reduced to 15 nm [72]. However, in vivo, only keratinocytes within the stratum basale are thought to express focal contacts. Their existence is marked by the localization of the linker protein talin to cell-matrix contacts with the basement membrane at the dermal-epidermal junction [73]. Focal adhesions that employ a2b1 (collagen- and laminin-specific)

SC

Dsg1 Dsg4 Dsc1

K1/10

SG

SS

SB

Dsg2 Dsg3 Dsc2 Dsc3 K5/14

DEJ AJ

D

FAC

HD

MF

KIF

KG

Fig. 9.5 In this cartoon representation of the human epidermis, all four major classifications of anchoring junctions can be observed: focal adhesion complexes (FAC), adherens junctions (AJ), desmosomes (D), and hemidesmosomes (HD). DEJ dermal-epidermal junction; SB stratum basale; SS stratum spinosum; SG stratum granulosum; SC stratum corneum; MF microfilament; KIF keratin intermediate filament; Dsg desmoglein; Dsc desmocollin; K keratin; KG keratohyalin granule. Gradients for plakophilin, P-Cadherin, E-Cadherin, and K9 are not shown

182

J.C. Selby

and a3b1 (laminin- and epiligrin/kalinin-specific) integrins seem to preponderate within the population of keratinocyte stem cells at this level of the epidermis [74]. Like focal adhesions, hemidesmosomes are also observed in vitro, but their formation is apparently limited to cultures that utilize substrates coated with components of the extracellular matrix [75, 76]. In vivo, hemidesmosomes are exclusively localized to keratinocytes of the stratum basale within the basement membrane zone. In human skin, the basement membrane is an extraordinarily complex structure, consisting of an epidermally-derived basal lamina and a lamina reticularis of dermal origin [2, 6, 77]. Ultrastructurally, the basal lamina comprises a 50–100 nm thick composite structure that includes a distinct lamina lucida and lamina densa [7, 78]. In terms of biochemical composition, the lamina lucida is made from several different types of laminins, but the lamina densa is constructed from mostly type IV collagen [3, 7, 77]. In a structure unique to skin, hemidesmosome attachments to the lamina lucida are connected to type VII collagen fibrils that extend through the lamina densa to join with anchoring plaques entangled with connective tissue proteins of the lamina reticularis (cf. Fig. 9.6, Sect. 9.4) [6, 79]. Together, this fibrous anchoring complex is thought to reinforce the adhesive interface between the dermis and the epidermis at the dermal-epidermal junction [6, 77]. Of the four types of anchoring junctions expressed in epidermal keratinocytes [6], the assembly of cell-cell desmosomes and adherens junctions has been the most actively investigated [54, 70, 80–83]. In vitro, adherens junctions initiate their formation as streak-like attachments of contact between radially oriented microfilament

Fig. 9.6 As an artifact of tissue preparation for H&E staining, desmosomes appear to circumscribe keratinocytes of the stratum spinosum. Desmosomes are indicated by block arrows. Figure created from images available in ref. [8], with permission

9 Mechanobiology of Epidermal Keratinocytes

183

bundles that extend from a peripheral band of contractile bundles running parallel to the plasma membrane [40, 70, 83].5 As the assembly process continues, radially oriented actin bundles decrease in size up to the point of their virtual disappearance, concomitant with the creation of a continuous adhesion belt. In contrast, desmosome formation is marked by the development of prominent button-like contacts between adjacent cell membranes that join keratin intermediate filaments reaching out from a cage-like network that surrounds the nucleus [54, 70, 80, 81]. Interestingly, the processes related to desmosome and adherens junction assembly have been observed to share a close temporal and spatial coordination of activity [54, 82]. Unfortunately, the biophysical mechanisms that drive the process are not well known. In vivo, adherens junctions have been observed in all layers of the human epidermis, often in close association or even alternating with neighboring desmosomes [84, 85] (cf. Fig. 9.5). Classical cadherin proteins such as E-cadherin and P-cadherin are the major transmembrane component of these junctions [86, 87]. Desmosomes are also known to exist in all layers of the epidermis, either in a complete or partially degraded form. In the stratum spinosum, they are particularly prominent (cf. Fig. 9.6).

9.3.3

Keratinocyte Differentiation

Major changes in the expression of cytoskeletal and anchoring junction proteins accompany the morphological alterations in cell shape experienced by epidermal keratinocytes during the process of keratinocyte differentiation. An excellent review can be found in ref. [88]. First, consider keratinocytes within the deepest layers of the epidermis. Intermediate filament expression in the stratum basale is dominated by the K5/K14 keratin heterodimer. Keratin filaments are anchored to both desmosomes and hemidesmosomes within this layer of cells (cf. Fig. 9.5), but these junctions are usually outnumbered by microfilament-associated adherens junctions. As keratinocytes migrate to the stratum spinosum and beyond, K5/K14 keratins are downregulated in favor of K1/K10 heterodimers. In the epidermis of the palms and soles, K9 is also expressed in these cells [6, 23, 88]. Independent of body region, desmosomes are the predominant anchoring junction in suprabasal keratinocytes [89, 90]. With regards to desmosomal transmembrane and linker proteins, the expression of desmoglein 1, desmocollin 1, and plakophilin 1 occurs in a gradient. The relatively low levels of these proteins present in stratum basale gradually increase within keratinocytes as they approach the stratum corneum. The opposite profile has been observed for desmocollin 3, desmoglein 3, and 5

Under normal physiological conditions, dermal fibroblasts do not form cell-cell adherens junctions or desmosomes in vivo [9, 12]. However, following fibroblast-to-myofibroblast differentiation that occurs during wound repair, adherens junctions are formed between myofibroblasts that participate in granulation tissue contraction [11]. Fibroblasts cultured in vitro are capable of forming filopodial-like adherens junctions [40].

184

J.C. Selby

plakophilin 2 [91]. With respect to adherens junctions, both E-cadherin and P-cadherin are strongly expressed in the stratum basale, but only E-cadherin is found in suprabasal keratinocytes [87]. Other proteins associated with adherens junctions and desmosomes show little, if any, changes in expression during the process of terminal differentiation. As keratinocytes enter the stratum granulosum, protein expression becomes dominated by the production of protofilaggrin, the major protein constituent of the large keratohyalin granules that accumulate within cells of this layer. Proteolytic cleavage of a profilaggrin molecule results in multiple peptides of filaggrin, a protein that bundles keratin intermediate filaments into aggregates termed macrofibrils [90, 92]. Lamellar granules also become abundant. These granules eventually merge with plasma membrane and empty their lipid content into the intercellular space. In the final stages of keratinocyte differentiation occurring in the stratum lucidum and stratum corneum, the nucleus degenerates, the cytoplasm becomes filled with keratin macrofibrils, and transglutaminases become activated. These enzymes crosslink the proteins involucrin and loricrin, in addition to other small proline-rich proteins (SPRs) in the formation of a reinforcing structure known as the cornified cell envelope [3, 89]. Within the stratum corneum, degraded desmosomes (or corneosomes) and intercellular lipids are responsible for corneocyte cell-cell adhesion. To summarize the cornification process, keratinocyte differentiation can be likened to a series of biochemical reactions and morphological restructuring events in which an individual keratinocyte transforms itself from a vibrant, but relatively fragile cell of the stratum basale to a lifeless, though nearly indestructible squame of the stratum corneum. With regards to cell-cell adhesion, the reverse process occurs. As desmosomes and adherens junctions are replaced by corneosomes and intercellular lipids, the forces of cell-cell cohesion are weakened, resulting in fracture and desquamation [1]. In closing this section on keratinocyte biology, consider the hypothesis that cell-cell and cell-matrix anchoring junctions regulate the mechanical integrity of keratinocytes residing within the stratum basale and stratum spinosum of the human epidermis. Given the coincident expression of adherens junctions, focal adhesions, desmosomes, and hemidesmosomes within these deepest layers of the epidermis, which is the more mechanobiologically important structure: an interconnected network of microfilaments or keratin intermediate filaments? The answer to this question can be found in what follows next: a discussion of keratinocyte pathology and blistering skin diseases.

9.4

Keratinocyte Pathology: Blistering Skin Diseases

Morphological descriptions of desmosomes, hemidesmosomes, and their associated intermediate filament networks preceded an in-depth understanding of their physiological significance [93]. Early microscopists noted that although desmosomes

9 Mechanobiology of Epidermal Keratinocytes

185

and hemidesmosomes are present in epithelial tissues, lymph nodes, and meninges, they are most prominent in tissues subjected to mechanical stress, i.e., the epidermis and the heart. A turning point in our understanding of desmosomehemidesmosome-intermediate filament function came to light in the early 1990s when the link between keratin proteins and inherited human skin fragility disorders was established [94]. Since that time, it has been shown that non-traumatic mechanical failures of the epidermis, categorically described here as blistering skin diseases,6 can manifest not only from inherited mutations of keratin intermediate filaments, desmosomes, and hemidesmosomes, but also from infectious or autoimmune disease processes that target the transmembrane and/or extracellular matrix proteins associated with these anchoring junctions [23, 59, 95–97]. Dermatologists were among the first to observe and document human disease states in which the skin forms blisters, i.e., fluid-filled regions of localized tissue separation, in response to an otherwise benign mechanical deformation [98]. Building from a foundation of detailed patient case reports and meticulous histological/immunohistochemical studies, the molecular pathologies underlying many of these blistering skin diseases are now universally recognized. For some readers, the clinical lexicon associated with the names and classifications of these diseases can be confusing. To simplify matters, in the following discussion, consider only (a) whether the disease is acquired or inherited, and (b) whether the target protein is associated with a hemidesmosome or desmosome (cf. Table 9.2). Molecular views of representative blistering skin pathologies associated with these two anchoring junctions can be viewed in Figs. 9.7 and 9.8, respectively. Inherited blistering skin diseases are hallmarked by genetic mutations that result in the diminished expression of a target protein, the production of a mutant target protein with impaired functionality, or the total absence of the target protein. Blisters arise due to the presence of a “missing link” within the cytoarchitecture of desmosomes, hemidesmosomes, and keratin intermediate filaments that is normally expressed in epidermal keratinocytes. During the past several decades, both autosomal dominant and recessive mutations have been reported for the desmosomal proteins desmogleins 1 and 4, plakoglobin, plakophilins 1 and 2, and desmoplakin [95]. Mutations in K5, K14, and K9 are associated with keratin intermediate filament dysfunction. Mutations have also been discovered for the hemidesmosomal proteins collagen VII, BP180, plectin, a6b4 integrins, and laminin 5 [99]. Together with the K5 and K14 mutations, this collection of hemidesmosomal protein mutations gives rise to the family of blistering skin disorders known as epidermolysis bullosa. Autoimmune blistering skin disorders, also referred to as immunobullous diseases, are the most common type of acquired disorder, typically associated with the production of autoantibodies through an aberrant immune response to

6

Here, the term blistering skin disease (or disorder) is used as a general reference to bullous skin diseases of autoimmune, genetic, or infectious etiologies. The term is not indicative of any clinically relevant, dermatological classification or diagnosis of skin disease.

186

J.C. Selby

Table 9.2 Blistering skin diseases Disease Epidermolysis bullosa with muscular dystrophy Junctional epidermolysis bullosa with pyloric atresia Dystrophic epidermolysis bullosa Epidermolysis bullosa simplex Bullous pemphigoid

Pathology Inherited

Anchoring Target protein junctiona HD Plectin

Inherited

HD

b integrin

[101]

Inherited Inherited Acquired

HD D, HD HD

[102] [102–105] [99, 106]

HD D D

Collagen VII K5 or K14b BP180,c BP230d Collagen VII Desmoplakin K9b

[106] [107] [108]

D D D

Desmoglein 3 Desmoglein 1 Desmoglein 1

[91, 99, 109] [91, 99, 109] [109]

Epidermolysis bullosa acquisita Acquired Striate palmoplantar keratoderma Acquired Epidermolytic palmoplantar Acquired keratoderma Pemphigus vulgaris Acquired Pemphigus foliaceus Acquired Bullous impetigo and Acquired staphylococcal scalded skin syndrome a HD hemidesmosome; D desmosome b K: keratin c BP180: 180-kDa bullous pemphigoid antigen d BP230: 230-kDa bullous pemphigoid antigen

References [100]

self-antigens. Autoimmune diseases associated with hemidesmosomes constitute what is referred to as the pemphigoid group of disorders, whereas the pemphigus group of disorders comprises the set of immunobullous diseases associated with desmosomes [110]. Self-antigens related to pemphigoid disorders include collagen VII, BP180, BP230, plectin, a6b4 integrin, and laminin 5. Pemphigus group antigens include desmocollin 1, desmoglein 1, desmoglein 3, plakoglobin, and desmoplakin [99]. E-cadherin has also been found to be an immunological target for pemphigus autoantibodies, but its functional relevance in the pathogenesis of these diseases remains to be determined [111]. Autoantibodies generated during the course of immunobullous diseases are thought to mediate their pathogenic effects through a variety of mechanisms, including steric interference of the extracellular domains of transmembrane adhesive proteins, transduction of unwanted biochemical signals that disrupt cell function, and/or concurrent activation of the complement system [96]. As a specific example, autoantibodies to desmoglein 3 generated during the course of pemphigus vulgaris are thought to induce blistering via the activation of apoptotic executioner caspases that sequentially lead to (a) the cleavage and retraction of keratin filaments within basal keratinocytes, (b) tearing of whole desmosomes from suprabasal keratinocytes that renders them incapable of preventing apoptotic cell death, and (c) basal cell shrinkage with concomitant cell-cell separation and intraepidermal fracture in the stratum spinosum [112, 113]. In essence, the basal cell shrinkage

9 Mechanobiology of Epidermal Keratinocytes

187

normal hemidesmosome K5/K14 keratin plectin intermediate filaments

dystrophic epidermolysis bullosa

BP230

αβ integrin

BP180 laminins collagen IV

collagen VII

loss of collagen VII expression

anchoring plaque collagen I (dermis)

epidermolysis bullosa - muscular dystrophy

bullous pemphigoid

loss of plectin expression

autoantibodies to BP180

complement activation C3

epidermolysis bullosa acquisita

autoantibodies to collagen VII

C3

C3

junctional epidermolysis bullosa - pyloric atresia

mutant β integrin expression

Fig. 9.7 Ultrastructural views of normal and pathological hemidesmosomes located along the basement membrane zone of the dermal-epidermal junction (cf. Table 9.2)

hypothesis [112] promotes the idea that the mechanism of blistering in pemphigus is truly an integrated mechanobiological phenomenon. Although few in number, microbial infections can also initiate certain types of acquired blistering skin disorders. Some strains of the gram-positive bacteria

188

J.C. Selby

normal desmosome intercellular space plakophilin

plakoglobin desmoplakin keratin intermediate filaments

desmocollin desmoglein

striate palmoplantar keratoderma

reduced expression of desmoplakin

epidermolytic palmoplantar keratoderma

keratin K9 mutation results in dysfunctional filament assembly

bullous impetigo and staphylococcal scalded skin syndrome

exfoliative toxin produced during S. aureus infection cleaves desmoglein-1

Fig. 9.8 Ultrastructural views of normal and pathological desmosomes observed between epidermal keratinocytes of the stratum spinosum and stratum granulosum (cf. Table 9.2)

9 Mechanobiology of Epidermal Keratinocytes

189

Staphylococcus aureus are known to produce exfoliative toxins that proteolytically cleave desmoglein 1 within their extracellular domain [114]. Cleavage in this domain seems to impair the ability of the truncated protein to form stable adhesive bonds within the desmosome, resulting in blister formation due to intraepidermal separation at the level of the stratum granulosum, i.e., the layer of the epidermis where desmoglein 1 is predominantly expressed (cf. Fig. 9.5). Localized infection of the epidermis with these specific toxin-producing strains of S. aureus produces the blisters of bullous impetigo [109, 114, 115]. In infants and small children, localized Staph infections complicated by toxin release into the systemic circulation results in widespread blistering and a much more serious condition referred to as staphylococcal scalded skin syndrome [114, 115]. To better appreciate the clinicopathologic differences between the different types blistering skin diseases, consider the diagnostic comparison of epidermolysis bullosa simplex (EBS) and pemphigus vulgaris (PV) presented in Fig. 9.9. At first glance, the gross anatomical features of the blisters in either patient appear to be quite similar (cf. Fig. 9.9a, b). However, the more astute observer would note the various body regions to which the lesions are confined, as well as whether or not lesions at different sites of the body appear to be at the same stage of development. The next step in distinguishing between the underlying pathologies is to analyze a biopsy of each lesion in order to determine at what histological level the blisters are forming, i.e., is the plane of fracture subcorneal, intraepidermal, dermal-epidermal, or sub-epidermal? The biopsy could also be used gain insight into the mechanobiological mechanisms associated with the mechanical failure. Was blistering related to keratinocyte cytolysis, i.e., the fragmentation and disintegration of individual keratinocytes, or was the failure acantholytic, i.e., related to keratinocyte cell-cell separation? Observations such as these invariably facilitate the creation of a list of suspect target proteins and corresponding differential diagnoses. In the case of epidermolysis bullosa simplex, epidermal failure occurs by a cytolytic cleavage of individual keratinocytes within the stratum basale due to the cytoplasmic clumping of dysfunctional keratin intermediate filaments. This clumping phenomenon is thought to be associated with the intracellular vacuoles (white arrows) and eosinophilic homogenizations (black arrows) sometimes observed in the histopathology of specific subtypes of epidermolysis bullosa simplex (cf. Fig. 9.9c). In contrast to cytolytic failure, skin blisters formed during an outbreak of pemphigus vulgaris are typically acantholytic, marked by an intraepidermal separation localized to the stratum spinosum. The acantholytic cleavage of pemphigus vulgaris often generates a histopathological “tombstone” row of basal keratinocytes that remain attached at the basement membrane zone (cf. Fig. 9.9d). In both of these case examples, identification of the exact target protein(s) underlying the disease in question would require further testing. In addition to relevant observations made during clinical examination, factors such as patient age and family history must also be taken into account when selecting more

190

J.C. Selby epidermolysis bullosa simplex

a

pemphigus vulgaris

b posterior thorax 55-year old male head and scalp 5-day old male

c

d

e

plectin

mutant K5/K14 keratin intermediate filaments

intercellular space

BP230

desmoplakin plakoglobin

αβ integrin

BP180 laminins collagen IV

collagen VII anchoring plaques

f

plakophilin

keratin intermediate filaments

autoantibodies desmocollin 3 to desmoglein 3

collagen I (dermis)

Fig. 9.9 Gross anatomy (a, b), histology (c, d), and molecular ultrastructure (e, f), of the inherited blistering skin disease, epidermolysis bullosa simplex, and the acquired disorder, pemphigus vulgaris. Image (a) adapted from ref. [116], with permission. Images (b, d) adapted from refs. [117, 118], respectively, with permission. Image (c) exhibits histopathology of the Dowling-Meara subtype of epidermolysis bullosa simplex, adapted from ref. [104], with permission

advanced diagnostic investigations. For the patient with epidermolysis bullosa simplex, immunofluorescence antigenic mapping, transmission electron microscopy, and/or genetic screening could be used to identify mutations present in keratins K5 or K14 that impair intermediate filament assembly. In the patient

9 Mechanobiology of Epidermal Keratinocytes

191

with pemphigus vulgaris, immunofluorescence analysis of the biopsy specimen and a separate serum sample could be used to identify and quantify the titer of autoantibodies to desmoglein 3 (among others). In recent years, further evidence in support of the putative mechanobiological functions of keratin intermediate filaments, desmosomes, and hemidesmosomes has been generated in the laboratory using transgenic and knockout mice to control the expression of specific proteins associated with these cytoskeletal components. In one remarkable example, transgenic mice were engineered to express truncated K14 keratins [103]. In these animals, the resultant murine blistering disorder was observed to resemble the human disease epidermolysis bullosa simplex Dowling-Meara. The severity of blistering was shown to be associated with the degree of perturbed keratin filament assembly [103, 105]. In knockout animals involving keratins other than K14, disease phenotypes have been shown to be confined to the epithelial tissues in which the knockout keratin filament is normally expressed [23, 119]. With respect to hemidesmosomes, mice with b4 integrin knockouts were observed to have severe blistering and tissue separation at the dermal-epidermal junction as a result of the mechanical trauma induced during parturition [60]. Plectin-null mice died 2–3 days following birth, exhibiting bullous epidermal lesions correlated to a reduction in the number of hemidesmosomes present at the basal lamina [120, 121]. These mice were also observed to have severe skeletal and cardiac muscle abnormalities. With respect to desmosomes, desmoplakin-null mice were noted to have cytolytic separation within the stratum basale and stratum spinosum in skin areas that contacted during limb movement [122]. In animal models of acquired blistering skin diseases, it has been shown that a murine model of human pemphigus vulgaris can be generated in an immunodeficient (Rag 2-null) mouse by passive transfer of B and T cells isolated from a desmoglein 3 knockout mouse that has been previously immunized with recombinant desmoglein 3 [123, 124]. Without question, evidence stemming from both clinically documented human blistering skin diseases and their corresponding animal models of disease has led to the conviction that keratin intermediate filaments, desmosomes, and hemidesmosomes are responsible for maintaining the mechanical integrity of keratinocytes that reside within the deepest layers of the epidermis. Despite the large volume of clinicopathological evidence, this assertion continues to be supported by mostly observational data, as opposed to direct mechanical measurements. The inability to transmit load within and between keratinocytes seems to give rise to most blistering skin diseases, but does any rheological data exist concerning the load-deformation behavior of desmosomes, hemidesmosomes, or keratin intermediate filaments that can be used to affirm or refute this notion?

9.5

Keratinocyte Rheology

The correlation between the morphology and mechanical properties of the integument has intrigued men of scientific bent throughout history. No less than three centuries worth of archival literature exists on the subject, beginning with De externo tactus

192

J.C. Selby

organo anatomica observatio, an epistle containing the first comprehensive histological observations of human skin, as reported by Marcello Malpighi (1628–1694) [125]. Since that time, countless others have attempted to explain the mechanical attributes of human skin in relation to its anatomical constitution. In the following sections, some of the many different experiments used to assess the rheological behavior of skin, and more specifically, epidermal keratinocytes, are discussed at macroscopic, mesoscopic, microscopic, and nanoscopic length scales. Despite the categorical presentation, the purpose of this section is not to provide an exhaustive review of skin biomechanics. Rather, the intention is to differentiate what is known from what is not known with regard to keratinocyte rheology as it relates to the function of keratin intermediate filaments, desmosomes, and hemidesmosomes in the context of blistering skin diseases. Throughout this section, the term mechanical behavior – a measured relationship between an applied load and its corresponding deformation, or a measured relationship between an applied deformation and its corresponding reactive load – will be used synonymously with the terms mechanical response, load-deformation behavior, rheological behavior, and mechanical properties.

9.5.1

Macroscale: Mechanics of Skin, Dermis, and Epidermis

Over the past 100 years, researchers have developed a variety of mechanical apparatus to characterize, both in vitro and in vivo, the infinitesimal and finite deformation responses of human, pig, rat, mouse, and rabbit skin. Inherently macroscopic techniques, like uniaxial tension tests, biaxial membrane-extension tests, torsion tests, and spherical-tip indentation tests, have demonstrated particular utility. The interested reader is encouraged to review references [13, 15, 126] as they are rich with bibliographic information pertaining to the development and utilization of these measurement modalities. Using one or more of these techniques, the mechanical properties of full-thickness skin [127–131], dermis [132], and epidermal stratum corneum [133–135] have been explored. More recently, variations of these experimental modalities have been used to assess the mechanical response of tissue-engineered dermal [136, 137] and epidermal [138] equivalents. Despite the abundance of work in this area, studies of the rheological behavior of epidermal layers other than the stratum corneum have been mostly neglected [139, 140]. Moreover, the mechanical properties of skin manifest under pathological conditions have not been widely investigated, with the noted exceptions of work on connective tissue disorders like Ehlers-Danlos syndrome [141] and work on the etiology of skin pressure ulcers [142]. In crude qualitative summary, the mechanical responses of full-thickness human skin, dermis, stratum corneum, and tissue-engineered dermal/epidermal equivalents can be said to exhibit five distinct behaviors: a nonlinear stress-strain relationship, loading-unloading cycle hysteresis, stress relaxation at constant strain, creep at constant stress, and preconditioning. Perhaps most remarkable of all,

9 Mechanobiology of Epidermal Keratinocytes

193

skin and its various structural components are characterized by an ability to undergo large deformations with near complete elastic recovery following removal of the applied load. With respect to studies of the epidermis, moisture and lipid content of corneocytes seem to represent the two most critical mediators of the mechanical behavior of the stratum corneum [135, 143]. Elasticity of the stratum corneum is not well understood, but it is thought to be related to the intracellular macromolecular bundling of keratin filaments. Mechanical failure of the stratum corneum is thought to occur by intercellular cleavage, i.e., fracture of intercellular lipids and corneosomes [133, 135]. In addition to the traits detailed above, a few other important mechanical attributes of skin warrant discussion. Foremost, human skin in vivo displays anisotropy and heterogeneity in its macroscopically observable properties. These characteristics are a strong function of body region, as expected from basic anatomical considerations (cf. Sect. 9.2). Moreover, human skin, even in an unloaded state, possesses finite levels of residual stress in vivo. Also referred to as resting tension, knowledge of these natural lines of tension (Langer’s lines) has important implications with respect to the proper placement of surgical incisions that minimize scar formation during wound healing [6, 13, 144]. Although the majority of the resting tension in skin is maintained within the dermis, it is likely that the epidermis also possesses a finite tensile residual stress in its undeformed state, as suggested by the culture of sheets of epidermal keratinocytes in vitro [138, 145].

9.5.2

Microscale: Isolated Keratinocyte Mechanics

For many years, research in cell mechanics was geared towards describing the highly specialized cellular processes of locomotion, phagocytosis, and mitosis within select populations of eukaryotic cells. However, the past two decades has witnessed a resurgent interest in cell mechanics since it has become apparent that most all cell types – regardless of their stereotypical tissue function – are capable of sensing, transducing, and responding to changes in their mechanical environment [146, 147]. How do cells accomplish such tasks? In a belief held by many, the cytoskeleton and its associated anchoring junctions are thought to play multiple roles in the regulation of mechanical homeostasis [47, 69, 148], with the most critical role being the maintenance of structural stability. Thus, it seems reasonable that “understanding how cells stabilize their structure and shape may help explain how cells sense and respond to mechanical signals [47].” Fittingly, intense efforts have been made to uncover the mechanisms by which cytoskeletal networks develop and distribute intracellular reactive loads when subject to mechanical deformations. A brief review of this work is included here. The more interested reader is advised to locate references [149–151] for more in-depth assessments of the rapidly evolving field of cell mechanics.

194

J.C. Selby

Microscopic rheological investigations of individual cells are almost exclusively carried out in vitro. Experimental techniques developed for interrogating the loaddeformation behavior of single cells include: micropipette aspiration [152], magnetic twisting cytometry [153, 154], magnetic tweezers [155, 156], magnetic microneedles [157], microfabricated silicon cantilever beams [158, 159], elastomeric microneedles [160], micropatterned elastic substrates [161], glass microplates [162, 163], the optical stretcher [164], and atomic force microscopy [165–167]. Most of these techniques are predicated on the use of probes that act as mechanical grips at sites of cell-matrix focal adhesion contacts. As a consequence, corresponding loaddeformation measurements tend to be dominated by the rheology of the microfilament network. Although laser trapping techniques have been developed to characterize the adhesive forces associated with the desmosomal proteins desmocollin and desmoglein [168–172], no methodology has been demonstrated capable of assaying the rheology of intermediate filament networks as probed through a fully formed desmosome or hemidesmosome. In contrast to the vast array of modalities available for exploring the rheology of single cells, relatively few different types of cells have been actively investigated. Rheological experiments on erythrocytes, leukocytes, fibroblasts, smooth muscle cells, and endothelial cells are the most popular. Epithelial cells have received far less attention. More importantly, to the best knowledge of this author, no formal biomechanical exploration of the single-cell rheology of a human epidermal keratinocyte has ever been reported. In rheological terms, most eukaryotic cell types can be summarily described as nonlinear viscoelastic entities. Single cell measurements have shown that most cells possess both fluid-like and solid-like behaviors, depending on the time scale of interest [151]. Cells have been observed to exhibit creep at constant stress, stress-relaxation, and loading-unloading cycle hysteresis [150]. The dynamic response of cells to oscillatory perturbations is marked by a wide spectrum of time constants [173]. In response to increasing external load, cell stiffness has been observed to increase, a characteristic often referred to as strain hardening [150]. Even in the absence of external loads, adherent cells are also known to carry a tensile residual stress (or prestress or cytoskeletal tension) analogous to the macroscopic property of resting skin tension (cf. Sect. 9.5.1). Under normal physiological conditions, the existence of a cellular prestress seems to endow cells with the property of prestress-induced stiffening [150] where an increase in the magnitude of prestress is directly related to an increase in the observed cell stiffness [174]. Admittedly, it is tempting to suggest that individual epidermal keratinocytes would demonstrate several, if not all of the mechanical characteristics exhibited by other adherent eukaryotic cell types. However, observations of cell rheology, especially when based on experiments of a specific cell type, should not be generalized as model behavior for all cells [151]. This argument aside, a more important question to be asked here is whether or not we should investigate the rheology of single epidermal keratinocytes at all. As illustrated in the preceding sections, isolated keratinocytes do not exist in vivo – they exist only in the context of the multicellular stratified squamous keratinized epithelium that is the human epidermis. Due to the extensive

9 Mechanobiology of Epidermal Keratinocytes

195

integration of keratin intermediate filaments and microfilaments by cell-cell and cell-matrix anchoring junctions, the cytoarchitectures exhibited by epidermal keratinocytes in vivo are perhaps best thought of as mesoscopic structures, as opposed to microscopic entities associated with individual cells.

9.5.3

Mesoscale: Epithelial Sheet Mechanics

Although a mesoscopic length scale does not correlate with any conventional subdivision of tissue histology, its potential utility for the rheological exploration of epidermal keratinocytes should not be overlooked. For the sake of discussion, a mesoscopic biological specimen is defined here as any assembly of cells and/or extracellular matrix material configured to resemble a specific tissue type, but geometrically constrained such that the critical dimensions of the specimen span at least two physically disproportionate length scales, i.e., one macroscopic dimension and one microscopic dimension. Assuming they reconstitute some essential characteristics of their model tissues, mesoscopic biological specimens engineered in vitro offer a unique opportunity for performing mechanobiological investigations at a phenomenological length scale that is more experimentally accessible in the laboratory. As detailed in Fig. 9.10, consider the formation of an epithelial sheet in vitro that initiates with the culture of primary keratinocytes at calcium concentrations of 1–2 mM [70, 80–83, 175, 176]. Through the formation of cell-cell adherens junctions and desmosomes, the microfilament and keratin intermediate filament networks are reorganized and integrated into a cytoarchitectural structure with a characteristic dimension larger than a single keratinocyte, but smaller than the largest dimension of the cell layer. In this sense, the keratinocyte sheet can be said to constitute a mesoscopic specimen with an average thickness of 3–5 mm (or more if the keratinocytes are stratified) and a lateral dimension that can vary in size up to 1 cm and beyond. By most accounts, the keratinocyte sheet is likely to exhibit at least some of the essential mechanobiological properties of cells resident within the stratum basale and stratum spinosum in vivo. At present, there are two major types of experiments that have been utilized to investigate the mechanobiological properties of keratinocyte sheets reconstituted in vitro: substrate-distension cell mechanostimulus experiments [177] and composite diaphragm inflation experiments [178–180]. Both types of experiment employ freestanding elastomeric membranes as culture substrates, and both experiments use these membranes to subject an attached keratinocyte sheet to cyclic uniaxial or biaxial stretch deformations. However, due to engineered differences in the mechanical properties of the elastomeric membranes, only composite diaphragm inflation experiments can be used to assess the rheological response of a keratinocyte sheet [178]. Substrate-distension cell mechanostimulus experiments, though typically limited to experiments assaying mechanotransduction in keratinocytes

196

J.C. Selby

stress fibers

peripheral microfilaments

keratin intermediate filaments desmosome

adherens junction

[Ca2+] ~ 0.1mM

[Ca2+] ~ 1.2 mM calcium switch

Fig. 9.10 The cartoon above depicts the cytoarchitectural reorganization of keratin intermediate filaments and microfilaments observed in epidermal keratinocytes during the formation of an epithelial sheet in vitro. The illustration was compiled from studies that utilized immunofluorescence, phase contrast, and differential interference contrast microscopy to simultaneously observe cell margins, cytoskeletal network morphology, and anchoring junction localization before, during, and after the calcium switch [70, 80, 81, 83, 175, 176]

[181, 182], have been used to investigate morphological changes in keratinocyte cytoarchitecture in response to applied mechanical deformations [183, 184]. In one particularly relevant series of substrate-distension cell mechanostimulus experiments, immortalized keratinocytes expressing the K14 mutation of epidermolysis bullosa simplex-Dowling-Mera exhibited severe fragmentation of their keratin intermediate filament network in response to repeated stretch and relaxation cycles (12% nominal biaxial stretch at 4 Hz) [183]. Keratin filament fragmentation was also found to be associated with a progressive disassembly of cell-cell desmosomes. Together, these observations were used to suggest the idea that the maintenance of functional hemidesmosomes and desmosomes within living epidermal keratinocytes requires some amount of intrinsic tension within the keratin filament network [175]. In a second substrate-distension study conducted by the same researchers, sheets of immortalized keratinocytes transfected with non-mutant GFP-K14 keratin were quasi-statically stretched (0.006 s 1 strain rate) in a uniaxial configuration to maximum strains of ~133% [184]. The fluorescently tagged keratin networks sustained only minor damage at strains as large as 100%, concomitant with only minor losses in cell viability. Electron microscopy demonstrated that increasing keratinocyte stretch was accompanied by progressive keratin filament elongation, suggesting that some individual filaments within the network might be subjected to direct tensile loading in response to the applied deformation [184]. Furthermore, buckling of individual keratin filaments was noted to occur when the keratinocytes

9 Mechanobiology of Epidermal Keratinocytes

197

were restored to their undeformed configurations [184]. The possibility that the observed buckling was related to a plastic deformation occurring within the keratin filament bundles could not be experimentally excluded [184]. With respect to sheet rheology, data generated from a series of composite diaphragm inflation experiments was recently used to suggest that epidermal keratinocyte sheets exhibit a combined viscoelastic-plastic stress response when subject to large quasi-static biaxial strains (~50% nominal stretch at 0.05 Hz) [178–180]. In these experiments, keratinocyte sheet specimens that were returned to quiescent culture following a sequence of stretch tests recovered roughly 80% of their initial ability to store elastic strain energy – a result suggestive that keratinocyte sheets undergo a mechanobiological adaptation and recovery process in response to large deformations [180]. Regrettably, no attempt was made to isolate the cytoarchitectural elements and deformation mechanisms contributing to these observed responses.

9.5.4

Nanoscale: Mechanics of Cytoskeletal Proteins and Cell-Cell Adhesion Molecules

Macroscopic mechanical measurements of organs and tissues as well as microscopic measurements of individual cells are related in that characteristic load-deformation behaviors arise as a collective response of many different interacting structures. Nanoscopic rheological measurements, as defined in this work, are distinguished by the fact that that they can be used to evaluate the rheological behavior of isolated cytoarchitectural components. Chronologically, nanoscale mechanical testing is a relative newcomer to the field of biomechanics. As such, there is less archival literature available for review. However, with regards to the present topic of epidermal keratinocytes, two areas merit discussion, namely, the rheological testing of cytoskeletal filaments and networks, and the mechanical testing of cadherinmediated cell-cell adhesive bonds.

9.5.4.1

Cytoskeletal Filaments and Networks

In a landmark work, the first quantitative rheological comparison between the three major classes of cytoskeletal protein networks was published in 1991. In this study, a specially designed torsion pendulum was used to collect viscometric measurements of equal weight concentrations of purified microfilament, microtubule, and vimentin intermediate filament suspensions [185, 186]. In response to applied shear stresses, microtubules possessed the greatest deformability, i.e., small stresses produced large network strains. However, at strains exceeding ~50%, microtubule networks ruptured and began to flow like a viscous fluid. In contrast to microtubules, microfilament networks were extremely stiff, undergoing small deformations in response to increasing magnitudes of shear stress. At strains of ~25%, microfilament networks ruptured and flowed. Vimentin intermediate

198

J.C. Selby

filaments, like microtubules, were easily deformed at low stresses, but their deformability gradually decreased with increasing levels of applied stress. This stiffening phenomenon was particularly prominent at strains of ~40%. In a characteristic unique from the other two cytoskeletal networks, vimentin intermediate filaments were able to sustain extremely large deformations without rupturing [185]. In the years following this comparative study, researchers have arduously worked to expand our knowledge of the rheological properties of cytoskeletal proteins reconstituted as isolated networks. The torsion pendulum has been replaced by more technically advanced cone-plate and parallel-plate viscometers, diffusing wave spectroscopy, and single particle tracking microrheology. Improved techniques for protein purification and suspension preparation have likewise been developed. These enhancements in methodology ultimately enabled more in-depth studies of both microfilaments [187–189] and intermediate filament networks [25, 190]. Of particular relevance to epidermal keratinocytes, viscometric studies have demonstrated differences in rheological behavior between naturally occurring heterotypic keratin filaments found in simple and complex epithelia, and between these natural keratin filaments and synthetic filaments constructed from intentionally mismatched pairs [191, 192]. The K5/K14 keratin heterodimer has been intensely scrutinized due to its known association with the blistering skin disease epidermolysis bullosa simplex [193]. Viscometric investigations have shown that isolated suspensions of K5/K14 filament networks reconstitute as gels with mechanical properties resembling a viscoelastic solid. When subjected to large shear deformations, K5/K14 suspensions exhibit strain hardening and yield. Remarkably, however, K5/K14 suspensions also demonstrate a unique ability to rapidly recover their preshear properties once a deformation has terminated. Further viscometric testing showed that a specific amino acid sequence within the K14 keratin is critical to the filament bundling process [194]. More recently, as an alternative to the characterization of purified cytoskeletal protein networks, some investigators have attempted to measure the mechanical properties of isolated protein filaments using atomic force microscopy. F-actin [195] and microtubule [196] filaments were the first to be studied. With regards to intermediate filaments, vimentin [197, 198], desmin [199], and K5/K14 keratins [200] have also been explored. Uniaxial tension tests of keratin-like slime threads [201] have also been used as fiducial representatives of keratin intermediate filament behavior. Perhaps the most important rheological observation that can be extracted from these collective studies is that single intermediate filaments, directly loaded in tension, are remarkably extensible, flexible, and tough [202, 203]. Most wildtype intermediate filaments have been observed to withstand 250% strains without rupture [200, 202, 203]. Single filaments loaded in tension have also been noted to exhibit strain hardening and yield [203]. Additionally, molecular modeling of the filament deformation process has provided evidence that an a-helix to b-sheet conformational change within the rodlike portion of the coiled-coil dimer

9 Mechanobiology of Epidermal Keratinocytes

199

(cf. Sect. 9.3.1) is at least partly responsible for the extensibility observed in these experiments [201, 203]. Based on these rheological findings, it has been proposed that intermediate filaments function as intracellular shock absorbers, capable of dissipating large quantities of mechanical energy via plastic deformations that preclude filament rupture [201, 202, 204]. In turn, if the cell were able to recognize the conformational changes induced in these filaments as a consequence of the deformation, it is also plausible that intermediate filaments could simultaneously function as mechanotransducers [201, 202]. In this regard, the role of intermediate filaments as mediators of intracellular signaling processes is only just beginning to be explored [202–204].

9.5.4.2

Cell-Cell Adhesion

During our previous discussion of blistering skin diseases (cf. Sect. 9.4), it was observed that many bullous pathologies seem to manifest because of a “missing link” in the intracellular and intercellular connectivity created by desmosomes, hemidesmosomes, and the keratin intermediate filament network. But what forces mediate momentum transfer between the molecular “links” of this cytoarchitecture? As one might expect, the complexity of molecular binding forces that exist within a living cell prevents their rigorous understanding in terms non-specific electrostatic and van der Waals forces; steric, hydration, and hydrophobic interactions; and entropically-derived forces [205]. Nevertheless, recent experimental advances have increased our understanding of the nanoscale physics of biological association and organization, and in particular, the mechanics governing the formation of cell-cell and cell-matrix contacts [205]. A few of these pioneering methodologies include the atomic force microscope [206], the surface force apparatus [207–209], optical tweezers [210], the bioforce probe [211], and the molecular force probe [212]. The classical cadherins have been extensively studied with these techniques [213]. The results of these investigations have refined our understanding of the intermolecular bond specificity, compliance, and rupture of cadherin-based cell-cell junctions, including E-cadherin adherens junctions of the human epidermis [214–219]. Classical cadherins are now known to consist of an extracellular domain (or ectodomain), a single-pass transmembrane domain, and a cytoplasmic domain. The ectodomain is composed of five distinct beta-barrel domains referred to as EC1 to EC5, separated by four distinct calcium binding sites localized to each of the EC domain junctions [213]. The N-terminus of EC1 is known to mediate homophilic cadherin selectivity [213]. Cadherin adhesion has been proposed to initiate with the formation of parallel cis-dimers on the same cell, followed by the sequential development of trans-interaction adhesive bonds with an anti-parallel dimer originating from a neighboring cell [213]. The homophilic trans-interactions between EC1, EC2, and EC3 are thought to provide the majority of the adhesive forces, with the strongest force mediated by the EC3 domain [213]. Cadherin ectodomains have also been observed to exhibit interdomain and intradomain cooperativity that

200

J.C. Selby

allows the transmission of conformational perturbations over large distances in the protein. This latter finding suggests that cadherins might be capable of inside-out signaling via a mechanobiological coupling between the cytoplasmic and ectodomains of the molecule [213]. Adhesive force measurements of the non-classical cadherins desmocollin and desmoglein have also received a great deal of attention due to their role in the pathology of pemphigus disorders (cf. Sect. 9.4). Atomic force microscopy and optical laser trapping have been used to make both morphological and rheological assessments of these cell-cell junctional proteins [168–172, 220]. Experiments have demonstrated that desmosomal adhesion is mediated in part by calciumdependent, homophilic trans-interactions of desmoglein 1, desmoglein 3, and desmocollin 3, in addition to heterophilic trans-interactions between desmocollin 3 and desmoglein 1 [168–172, 220]. With “all or nothing” calcium dependence, the natural calcium gradient that exists within the human epidermis could potentially regulate desmoglein/desmocollin-mediated adhesion in vivo [169, 221]. More importantly, it has been shown that the autoantibodies (IgG) generated in pemphigus foliaceus and pemphigus vulgaris do not directly inhibit adhesive desmoglein 1 homophilic trans-interactions, whereas pemphigus vulgaris autoantibodies do directly interfere with desmoglein 3 homophilic trans-interactions in vitro [169–171]. Pemphigus autoantibodies targeting desmoglein 1 may also contribute to the loss of keratinocyte adhesion as mediated by heterophilic trans-interactions between desmocollin 3 and demoglein 1 [172]. The mechanics of desmosomal adhesion loss in epidermal keratinocytes as mediated by autoantibody-triggered intracellular signaling pathways has only just begun to be explored [171]. Despite the growing interest in nanoscopic rheological experiments, one must be cautious when attempting to extrapolate the results of these measurements to the behavior of cell-cell adhesive proteins and cytoskeletal proteins in vivo. The structural organization and biological function of these proteins depends not only on their intrinsic properties, but also on their interactions with a host of accessory and regulatory molecules. Without question, studies of pure cytoskeletal protein filaments and networks in vitro “can only offer an incomplete picture of how they are put to work in a cell [192].”

9.6

Summary

Undeniably, advances in molecular biology and medicine have revealed much about the organization and regulation of cytoskeletal filaments and anchoring junctions within the human epidermis. Despite these advances, the current morphological, biochemical, clinicopathological, and rheological knowledge base surrounding keratinocyte cytoarchitecture is not, in and of itself, complete. Collectively, we do not understand how interconnected microfilament and intermediate filament networks regulate the mechanobiological behaviors of these cells. We do not understand how the dysfunction of keratin intermediate filaments,

9 Mechanobiology of Epidermal Keratinocytes

201

desmosomes, and hemidesmosomes gives rise to the diverse array of phenotypes observed in human blistering skin diseases. In essence, the biophysical mechanisms by which basal and subcorneal keratinocytes are able to withstand large strains during normal physiological function have yet to be fully revealed. Consider the following list of unanswered questions. If desmosomes distribute tensile loads between keratinocytes undergoing large deformations, what purpose do adherens junctions serve? Do adherens junctions assume a different mechanobiological function or is there some measure of redundancy between an adherens junction and a desmosome? Why are there no known blistering skin disorders associated with a dysfunctional adherens junction and/or its associated microfilament network? If the maintenance of functional hemidesmosomes and desmosomes requires some amount of intrinsic tension in the keratin intermediate filament network, where does this tension come from? How do intermediate filaments develop reactive loads in response to an applied deformation, i.e., is the rheology of intermediate filaments governed by enthalpic forces, entropic forces, or some combination thereof? Do desmosomes and/or adherens junctions play a role in cellular mechanotransduction? Do intermediate filaments themselves act as intrinsic mechanotransducers or as platforms for mediating signaling cascades involved in the process of mechanotransduction? Arguably, answers to these questions will require the development of more synergistic research programs that combine the experimental and analytical tools of molecular biology, clinical medicine, biophysics, and mechanics. Perhaps one day, working together, we will paint a more complete picture of the physiology that governs the human epidermis in states of both health and disease.

References 1. Kligman AM (1964) The biology of the stratum corneum. In: Montagna W, Lobitz Jr WC (eds) The Epidermis. Academic Press, New York. 2. Gartner LP, Hiatt JL (2000) Color Atlas of Histology, 3rd edn. Lippincott Williams & Wilkins, Philadelphia. 3. Kierszenbaum AL (2002) Histology and Cell Biology: An Introduction to Pathology. Mosby, St. Louis. 4. Montagna W, Parakkal PF (1974) The Structure and Function of Skin, 3rd edn. Academic Press, New York. 5. Burns T, Breathnach S, Cox N, Griffiths C (2010) Rook’s Textbook of Dermatology. WileyBlackwell, Oxford. 6. Fitzpatrick TB, Freedberg IM, Austen KF, et al (2003) Fitzpatrick’s Dermatology in General Medicine, 6th edn. McGraw-Hill, New York. 7. Junqueira LC, Carneiro J, Kelley RO (1998) Basic Histology, 9th edn. Appleton & Lange, Stamford. 8. Williams BD, Kokko-Cunningham A, Cvetko S, et al (2003) Internet Atlas of Histology. https://histo.life.illinois.edu/histo/atlas/index.php. Accessed June 2003. 9. Welch MP, Odland GF, Clark RA (1990) Temporal relationships of F-actin bundle formation, collagen and fibronectin matrix assembly, and fibronectin receptor expression to wound contraction. J Cell Biol 110:133–145.

202

J.C. Selby

10. Gabbiani G, Chaponnier C, Huttner I (1978) Cytoplasmic filaments and gap junctions in epithelial cells and myofibroblasts during wound healing. J Cell Biol 76:561–568. 11. Hinz B, Pittet P, Smith-Clerc J, et al (2004) Myofibroblast development is characterized by specific cell-cell adherens junctions. Mol Biol Cell 15:4310–4320. 12. Gabbiani G, Rungger-Brandle E (1981) The fibroblast. In: Glynn LE (ed) Handbook of Inflammation, Vol 3: Tissue Repair and Regeneration. Elsevier, Amsterdam. 13. Wilkes GL, Brown IA, Wildnauer RH (1973) The biomechanical properties of skin. CRC Crit Rev Bioeng 1:453–495. 14. Mackenzie IC (1983) The cellular architecture of the stratum corneum. In: Marks R, Plewig G (eds) Stratum Corneum. Springer, Berlin. 15. Kenedi RM, Gibson T, Evans JH, Barbenel JC (1975) Tissue mechanics. Phys Med Biol 20:699–717. 16. Alberts B, Johnson A, Lewis J, et al (2002) Molecular Biology of the Cell, 4th edn. Garland Science, New York. 17. Goldman RD (2001) Worms reveal essential functions for intermediate filaments. Proc Natl Acad Sci U S A 98:7659–7661. 18. Mitchison TJ, Cramer LP (1996) Actin-based cell motility and cell locomotion. Cell 84:371–379. 19. Bretscher A (1991) Microfilament structure and function in the cortical cytoskeleton. Annu Rev Cell Biol 7:337–374. 20. Nogales E (2000) Structural insights into microtubule function. Annu Rev Biochem 69:277–302. 21. Gelfand VI, Bershadsky AD (1991) Microtubule dynamics – mechanism, regulation, and function. Annu Rev Cell Biol 7:93–116. 22. Lodish H, Baltimore D, Berk A, et al (1995) Molecular Cell Biology, 3rd edn. Scientific American Books, New York. 23. Fuchs E, Cleveland DW (1998) A structural scaffolding of intermediate filaments in health and disease. Science 279:514–519. 24. Lazarides E (1980) Intermediate filaments as mechanical integrators of cellular space. Nature 283:249–256. 25. Coulombe PA, Bousquet O, Ma LL, et al (2000) The “ins” and “outs” of intermediate filament organization. Trends Cell Biol 10:420–428. 26. Svitkina TM, Verkhovsky AB, Borisy GG (1996) Plectin sidearms mediate interaction of intermediate filaments with microtubules and other components of the cytoskeleton. J Cell Biol 135:991–1007. 27. Goldman RD, Khuon S, Chou YH, et al (1996) The function of intermediate filaments in cell shape and cytoskeletal integrity. J Cell Biol 134:971–983. 28. Goldman RD, Clement S, Khuon S, et al (1998) Intermediate filament cytoskeletal system: dynamic and mechanical properties. Biol Bull 194:361–362. 29. Helfand BT, Chang L, Goldman RD (2004) Intermediate filaments are dynamic and motile elements of cellular architecture. J Cell Sci 117:133–141. 30. Yoon KH, Yoon M, Moir RD, et al (2001) Insights into the dynamic properties of keratin intermediate filaments in living epithelial cells. J Cell Biol 153:503–516. 31. Parry DAD, Steinert PM (1999) Intermediate filaments: molecular architecture, assembly, dynamics and polymorphism. Q Rev Biophys 32:99–187. 32. Fuchs E, Weber K (1994) Intermediate filaments – structure, dynamics, function, and disease. Annu Rev Biochem 63:345–382. 33. Hesse M, Magin TM, Weber K (2001) Genes for intermediate filament proteins and the draft sequence of the human genome: novel keratin genes and a surprisingly high number of pseudogenes related to keratin genes 8 and 18. J Cell Sci 114:2569–2575. 34. Szeverenyi I, Cassidy AJ, Chung CW, et al (2008) The human intermediate filament database: comprehensive information on a gene family involved in many human diseases. Hum Mutat 29:351–360.

9 Mechanobiology of Epidermal Keratinocytes

203

35. Langbein L, Rogers MA, Winter W, et al (1999) The catalog of human hair keratins I. Expression of the nine type I members in the hair follicle. J Biol Chem 274:19874–19884. 36. Coulombe PA, Tong XM, Mazzalupo S, et al (2004) Great promises yet to be fulfilled: defining keratin intermediate filament function in vivo. Eur J Cell Biol 83:735–746. 37. Kirfel J, Peters B, Grund C, et al (2002) Ectopic expression of desmin in the epidermis of transgenic mice permits development of a normal epidermis. Differentiation 70:56–68. 38. Kartenbeck J, Schwechheimer K, Moll R, et al (1984) Attachment of vimentin filaments to desmosomal plaques in human meningiomal cells and arachnoidal tissue. J Cell Biol 98:1072–1081. 39. Kowalczyk AP, Stappenbeck TS, Parry DAD, et al (1994) Structure and function of desmosomal transmembrane core and plaque molecules. Biophys Chem 50:97–112. 40. Yonemura S, Itoh M, Nagafuchi A, et al (1995) Cell-to-cell adherens junction formation and actin filament organization – similarities and differences between non-polarized fibroblasts and polarized epithelial-cells. J Cell Sci 108:127–142. 41. Fanning AS (2001) Organization and regulation of the tight junction by the actin-myosin cytoskeleton. In: Cereijido M, Anderson J (eds) Tight Junctions, 2nd edn. CRC Press, Boca Raton. 42. Madara JL (1987) Intestinal absorptive cell tight junctions are linked to cytoskeleton. Am J Physiol C 253:171–175. 43. Kovbasnjuk ON, Szmulowicz U, Spring KR (1998) Regulation of the MDCK cell tight junction. J Membr Biol 161:93–104. 44. Yap AS, Stevenson BR, Abel KC, et al (1995) Microtubule integrity is necessary for the epithelial barrier function of cultured thyroid-cell monolayers. Exp Cell Res 218:540–550. 45. Salomon D, Saurat JH, Meda P (1988) Cell-to-cell communication within intact human-skin. J Clin Invest 82:248–254. 46. Hall JE, Zampighi GA, Davis RM (1993) Progress in Cell Research, Volume 3: Gap Junctions. Elsevier, Amsterdam. 47. Ingber DE (1997) Tensegrity: the architectural basis of cellular mechanotransduction. Annu Rev Physiol 59:575–599. 48. Zamir E, Geiger B (2001) Components of cell-matrix adhesions. J Cell Sci 114:3577–3579. 49. Shemesh T, Geiger B, Bershadsky AD, et al (2005) Focal adhesions as mechanosensors: a physical mechanism. Proc Natl Acad Sci U S A 102:12383–12388. 50. Bershadsky AD, Balaban NQ, Geiger B (2003) Adhesion-dependent cell mechanosensitivity. Annu Rev Cell Dev Biol 19:677–695. 51. Adams CL, Nelson WJ (1998) Cytomechanics of cadherin-mediated cell-cell adhesion. Curr Opin Cell Biol 10:572–577. 52. Geiger B, Volk T, Volberg T (1985) Molecular heterogeneity of adherens junctions. J Cell Biol 101:1523–1531. 53. Garrod DR, Merritt AJ, Nie ZX (2002) Desmosomal adhesion: structural basis, molecular mechanism and regulation (Review). Mol Membr Biol 19:81–94. 54. Kowalczyk AP, Bornslaeger EA, Norvell SM, et al (1999) Desmosomes: intercellular adhesive junctions specialized for attachment of intermediate filaments. Int Rev Cytol 185:237–302. 55. Smith EA, Fuchs E (1998) Defining the interactions between intermediate filaments and desmosomes. J Cell Biol 141:1229–1241. 56. Bornslaeger EA, Corcoran CM, Stappenbeck TS, et al (1996) Breaking the connection: displacement of the desmosomal plaque protein desmoplakin from cell-cell interfaces disrupts anchorage of intermediate filament bundles and alters intercellular junction assembly. J Cell Biol 134:985–1001. 57. Kouklis PD, Hutton E, Fuchs E (1994) Making a connection: direct binding between keratin intermediate filaments and desmosomal proteins. J Cell Biol 127:1049–1060.

204

J.C. Selby

58. Jones JCR, Hopkinson SB, Goldfinger LE (1998) Structure and assembly of hemidesmosomes. Bioessays 20:488–494. 59. Borradori L, Sonnenberg A (1996) Hemidesmosomes: roles in adhesion, signaling and human diseases. Curr Opin Cell Biol 8:647–656. 60. Dowling J, Yu QC, Fuchs E (1996) Beta-4 integrin is required for epidermal cell adhesion, organization, and survival. Mol Biol Cell 7:3401. 61. Karakesisoglou I, Yang YM, Fuchs E (2000) An epidermal plakin that integrates actin and microtubule networks at cellular junctions. J Cell Biol 149:195–208. 62. Jefferson JJ, Leung CL, Liem RKH, (20004) Plakins: Goliaths that link cell junctions and the cytoskeleton. Nat Rev Mol Cell Biol 5:542–553. 63. Kaverina I, Rottner K, Small JV (1998) Targeting, capture, and stabilization of microtubules at early focal adhesions. J Cell Biol 142:181–190. 64. Stehbens SJ, Paterson AD, Crampton MS, et al (2006) Dynamic microtubules regulate the local concentration of E-cadherin at cell-cell contacts. J Cell Sci 119:1801–1811. 65. Ingber DE (2003) Tensegrity II. How structural networks influence cellular information processing networks. J Cell Sci 116:1397–1408. 66. Lee TYJ, Gotlieb AI (2003) Microfilaments and microtubules maintain endothelial integrity. Microsc Res Tech 60:115–125. 67. Rodriguez OC, Schaefer AW, Mandato CA, et al (2003) Conserved microtubule-actin interactions in cell movement and morphogenesis. Nat Cell Biol 5:599–609. 68. Horwitz AR, Parsons JT (1999) Cell biology – cell migration – movin’ on. Science 286:1102–1103. 69. Ingber DE (2003) Tensegrity I. Cell structure and hierarchical systems biology. J Cell Sci 116:1157–1173. 70. Zamansky GB, Nguyen U, Chou IN (1991) An immunofluorescence study of the calcium-induced coordinated reorganization of microfilaments, keratin intermediate filaments, and microtubules in cultured human epidermal keratinocytes. J Invest Dermatol 97:985–994. 71. Pasdar M, Li Z, Krzeminski KA (1992) Desmosome assembly in MDCK epithelial-cells does not require the presence of functional microtubules. Cell Motil Cytoskeleton 23:201–212. 72. Zamir E, Geiger B (2001) Molecular complexity and dynamics of cell-matrix adhesions. J Cell Sci 114:3583–3590. 73. Kaiser HW, Ness W, Offers M, et al (1993) Talin – adherens junction protein is localized at the epidermal-dermal interface in skin. J Invest Dermatol 101:789–793. 74. Jones PH, Harper S, Watt FM (1995) Stem-cell patterning and fate in human epidermis. Cell 80:83–93. 75. Chapman SJ, Eady RAJ (1985) Blistering in keratinocyte cultures: a regular phenomenon associated with differentiation. Eur J Cell Biol 39:352–359. 76. Mann PR, Constable H (1977) Induction of basal lamina formation in epidermal cell cultures in vitro. Br J Dermatol 96:421–426. 77. Burgeson RE, Christiano AM (1997) The dermal-epidermal junction. Curr Opin Cell Biol 9:651–658. 78. Timpl R, Brown JC (1996) Supramolecular assembly of basement membranes. Bioessays 18:123–132. 79. Keene DR, Sakai LY, Lunstrum GP, et al (1987) Type-VII collagen forms an extended network of anchoring fibrils. J Cell Biol 104:611–621. 80. Watt FM, Mattey DL, Garrod DR (1984) Calcium-induced reorganization of desmosomal components in cultured human keratinocytes. J Cell Biol 99:2211–2215. 81. Jones JCR, Goldman RD (1985) Intermediate filaments and the initiation of desmosome assembly. J Cell Biol 101:506–517. 82. Vasioukhin V, Bauer C, Yin M, et al (2000) Directed actin polymerization is the driving force for epithelial cell-cell adhesion. Cell 100:209–219.

9 Mechanobiology of Epidermal Keratinocytes

205

83. Okeefe EJ, Briggaman RA, Herman B (1987) Calcium-induced assembly of adherens junctions in keratinocytes. J Cell Biol 105:807–817. 84. Kaiser HW, Ness W, Jungblut I, et al (1993) Adherens junctions – demonstration in human epidermis. J Invest Dermatol 100:180–185. 85. Haftek M, Hansen MU, Kaiser HW, et al (1996) Interkeratinocyte adherens junctions: immunocytochemical visualization of cell-cell junctional structures, distinct from desmosomes, in human epidermis. J Invest Dermatol 106:498–504. 86. Blaschuk OW, Rowlands TM (2002) Plasma membrane components of adherens junctions (Review). Mol Membr Biol 19:75–80. 87. Jensen PJ, Wheelock MJ (1996) The relationships among adhesion, stratification and differentiation in keratinocytes. Cell Death Differ 3:357–371. 88. Fuchs E, Byrne C (1994) The epidermis: rising to the surface. Curr Opin Genet Dev 4:725–736. 89. Reichert U, Michel S, Schmidt R (1993) The cornified envelope: a key structure of terminally differentiating keratinocytes. In: Darmon M, Blumenberg M (eds) Molecular Biology of the Skin: The Keratinocyte. Academic Press, San Diego. 90. Fuchs E (1990) Epidermal differentiation. Curr Opin Cell Biol 2:1028–1035. 91. Green KJ, Gaudry CA (2000) Are desmosomes more than tethers for intermediate filaments? Nat Rev Mol Cell Biol 1:208–216. 92. Smith FJD, Irvine AD, Terron-Kwiatkowski A, et al (2006) Loss-of-function mutations in the gene encoding filaggrin cause ichthyosis vulgaris. Nat Genet 38:337–342. 93. Farquhar MG, Palade GE (1963) Junctional complexes in various epithelia. J Cell Biol 17:375–412. 94. Lane EB, McLean WHI (2004) Keratins and skin disorders. J Pathol 204:355–366. 95. McGrath JA (2005) Inherited disorders of desmosomes. Australas J Dermatol 46:221–229. 96. Sitaru C, Zillikens D (2005) Mechanisms of blister induction by autoantibodies. Exp Dermatol 14:861–875. 97. Omary MB, Coulombe PA, McLean WHI (2004) Mechanisms of disease: intermediate filament proteins and their associated diseases. New Engl J Med 351:2087–2100. 98. McGrath JA (2001) Keratinocyte adhesion and the missing link: from Dowling-Meara to Hay-Wells – St John’s Hospital Dermatological Society Annual Oration 2000. Clin Exp Dermatol 26:296–304. 99. Fassihi H, Wong T, Wessagowit V, et al (2006) Target proteins in inherited and acquired blistering skin disorders. Clin Exp Dermatol 31:252–259. 100. Smith FJD, Eady RAJ, Leigh IM, et al (1996) Plectin deficiency results in muscular dystrophy with epidermolysis bullosa. Nat Genet 13:450–457. 101. Vidal F, Aberdam D, Miquel C, et al (1995) Integrin beta-4 mutations associated with junctional epidermolysis-bullosa with pyloric atresia. Nat Genet 10:229–234. 102. Uitto J, Richard G (2004) Progress in epidermolysis bullosa: genetic classification and clinical implications. Am J Med Genet C 131:61–74. 103. Vassar R, Coulombe PA, Degenstein L, et al (1991) Mutant keratin expression in transgenic mice causes marked abnormalities resembling a human genetic skin-disease. Cell 64:365–380. 104. Bergman R, Harel A, Sprecher E (2007) Dyskeratosis as a histologic feature in epidermolysis bullosa simplex-Dowling Meara. J Am Acad Dermatol 57:463–466. 105. Coulombe PA, Hutton ME, Vassar R, et al (1991) A function for keratins and a common thread among different types of epidermolysis bullosa simplex diseases. J Cell Biol 115:1661–1674. 106. Zillikens D (1999) Acquired skin disease of hemidesmosomes. J Dermatol Sci 20:134–154. 107. Armstrong DK, McKenna KE, Purkis PE, et al (1999) Haploinsufficiency of desmoplakin causes a striate subtype of palmoplantar keratoderma. Hum Mol Genet 8:143–148. 108. Reis A, Hennies HC, Langbein L, et al (1994) Keratin-9 gene-mutations in epidermolytic palmoplantar keratoderma (EPPK). Nat Genet 6:174–179.

206

J.C. Selby

109. Payne AS, Hanakawa Y, Amagai M, et al (2004) Desmosomes and disease: pemphigus and bullous impetigo. Curr Opin Cell Biol 16:536–543. 110. Habif TP (2004) Clinical Dermatology: A Color Guide to Diagnosis and Therapy, 4th edn. Mosby, Edinburgh. 111. Evangelista F, Dasher DA, Diaz LA, et al (2008) E-cadherin is an additional immunological target for pemphigus autoantibodies. J Invest Dermatol 128:1710–1718. 112. Bystryn J-C, Grando SA (2006) A novel explanation for acantholysis in pemphigus vulgaris: the basal cell shrinkage hypothesis. J Am Acad Dermatol 54:513–516. 113. Grando SA, Bystryn J-C, Chernyavsky AI, et al (2009) Apoptolysis: a novel mechanism of skin blistering in pemphigus vulgaris linking the apoptotic pathways to basal cell shrinkage and suprabasal acantholysis. Exp Dermatol 18:764–770. 114. Ladhani S (2001) Recent developments in staphylococcal scalded skin syndrome. Clin Microbiol Infect 7:301–307. 115. Stanley JR, Amagai M (2006) Pemphigus, bullous impetigo, and the staphylococcal scaldedskin syndrome. New Engl J Med 355:1800–1810. 116. Eliades A (2010) eb_1_020819.jpg. #Dermatology Image Atlas. http://dermatlas.org. Accessed June 2010. 117. Cohen B (2010) pemphigus_7_050806.jpg. #Dermatology Image Atlas. http://dermatlas. org. Accessed June 2010. 118. Hosler G (2010) pemphigus_2_051121.jpg. #Dermatology Image Atlas. http://dermatlas. org. Accessed June 2010. 119. Galou M, Gao J, Humbert J, et al (1997) The importance of intermediate filaments in the adaptation of tissues to mechanical stress: evidence from gene knockout studies. Biol Cell 89:85–97. 120. Andr€a K, Lassmann H, Bittner R, et al (1997) Targeted inactivation of plectin reveals essential function in maintaining the integrity of skin, muscle, and heart cytoarchitecture. Gene Dev 11:3143–3156. 121. Andr€a K, Kornacker I, Jorgl A, et al (2003) (2003) Plectin-isoform-specific rescue of hemidesmosomal defects in plectin (-/-) keratinocytes. J Invest Dermatol 120:189–197. 122. Vasioukhin V, Bowers E, Bauer C, et al (2001) Desmoplakin is essential in epidermal sheet formation. Nat Cell Biol 3:1076–1085. 123. Amagai M, Tsunoda K, Suzuki H, et al (2000) Use of autoantigen knockout mice in developing an active autoimmune disease model for pemphigus. J Clin Invest 105:625–631. 124. Hashimoto T (2003) Recent advances in the study of the pathophysiology of pemphigus. Arch Dermatol Res 295:S2–S11. 125. Adelmann HB (1966) Marcello Malpighi and the Evolution of Embryology. Cornell University Press, New York. 126. Vogel HG (2002) Mechanical properties of human skin: Animal models. In: Elsner P, Berardesca E, Wilhelm K-P, et al (eds) Bioengineering of the Skin: Skin Biomechanics. CRC Press, Boca Raton. 127. Lanir Y, Fung YCB (1974) Two-dimensional mechanical properties of rabbit skin – II. Experimental results. J Biomech 7:171–182. 128. Manschot JFM, Brakkee AJM (1986) The measurement and modelling of mechanical properties of human skin in vivo – I. The measurement. J Biomech 19:511–515. 129. Manschot JFM, Brakkee AJM (1986) The measurement and modelling of mechanical properties of human skin in vivo – II. The model. J Biomech 19:517–521. 130. Lanir Y, Fung YCB (1974) Two dimensional mechanical properties of rabbit skin – I. Experimental system. J Biomech 7:29–34. 131. Mansour JM, Davis BR, Srour M, et al (1993) A method for obtaining repeatable measurements of the tensile properties of skin at low strain. J Biomech 26:211–216. 132. Daly CH (1982) Biomechanical properties of dermis. J Invest Dermatol 79:17s–20s. 133. Koutroupi KS, Barbenel JC (1990) Mechanical and failure behavior of the stratum corneum. J Biomech 23:281–287.

9 Mechanobiology of Epidermal Keratinocytes

207

134. Wu KS, van Osdol WW, Dauskardt RH (2006) Mechanical properties of human stratum corneum: effects of temperature, hydration, and chemical treatment. Biomaterials 27:785–795. 135. Wildnauer RH, Bothwell JW, Douglass AB (1971) Stratum corneum biomechanical properties. I. Influence of relative humidity on normal and extracted human stratum corneum. J Invest Dermatol 56:72–78. 136. Lafrance H, Guillot M, Germain L, et al (1995) A method for the evaluation of tensile properties of skin equivalents. Med Eng Phys 17:537–543. 137. Lafrance H, Yahia LH, Germain L, et al (1995) Study of the tensile properties of living skin equivalents. Biomed Mater Eng 5:195–208. 138. Chistolini P, De Angelis G, De Luca M, et al (1999) Analysis of the mechanical properties of in vitro reconstructed epidermis: preliminary results. Med Biol Eng Comput 37:670–672. 139. Hendriks FM, Brokken D, Oomens CWJ, et al (2004) Influence of hydration and experimental length scale on the mechanical response of human skin in vivo using optical coherence tomography. Skin Res Technol 10:231–241. 140. Hendriks FM, Brokken D, Oomens CWJ, et al (2006) The relative contributions of different skin layers to the mechanical behavior of human skin in vivo using suction experiments. Med Eng Phys 28:259–266. 141. Dobrev HP (2002) Mechanical properties in other dermatological diseases. In: Elsner P, Berardesca E, Wilhelm K-P, et al (eds) Bioengineering of the Skin: Skin Biomechanics. CRC Press, Boca Raton. 142. Bronneberg D, Bouten CVC, Oomens CWJ, et al (2006) An in vitro model system to study the damaging effects of prolonged mechanical loading of the epidermis. Ann Biomed Eng 34:506–514. 143. Wolfram MA, Wolejsza NF, Laden K (1972) Biomechanical properties of delipidized stratum corneum. J Invest Dermatol 59:421–426. 144. Wilhelmi BJ, Blackwell SJ, Phillips LG (1999) Langer’s lines: to use or not to use. Plast Reconstr Surg 104:208–214. 145. Tomson AM, Scholma J, Blaauw EH (1995) The effect of detachment of cultured epithelial sheets on cellular ultrastructure and organization. Epithelial Cell Biol 4:43–51. 146. Humphrey JD (2003) Continuum biomechanics of soft biological tissues. Proc Roy Soc Lond Ser A 459:3–46. 147. Vogel V, Sheetz M (2006) Local force and geometry sensing regulate cell functions. Nat Rev Mol Cell Biol 7:265–275. 148. Ingber DE (2003) Mechanobiology and diseases of mechanotransduction. Ann Med 35:564–577. 149. Zhu C, Bao G, Wang N (2000) Cell mechanics: mechanical response, cell adhesion, and molecular deformation. Annu Rev Biomed Eng 2:189–226. 150. Stamenovic D, Ingber DE (2002) Models of cytoskeletal mechanics of adherent cells. Biomech Model Mechanobiol 1:95–108. 151. Humphrey JD (2002) On mechanical modeling of dynamic changes in the structure and properties of adherent cells. Math Mech Solids 7:521–539. 152. Hochmuth RM (2000) Micropipette aspiration of living cells. J Biomech 33:15–22. 153. Wang N, Butler JP, Ingber DE (1993) Mechanotransduction across the cell surface and through the cytoskeleton. Science 260:1124–1127. 154. Wang N, Ingber DE (1995) Probing transmembrane mechanical coupling and cytomechanics using magnetic twisting cytometry. Biochem Cell Biol 73:327–335. 155. Bausch AR, Moller W, Sackmann E (1999) Measurement of local viscoelasticity and forces in living cells by magnetic tweezers. Biophys J 76:573–579. 156. Bausch AR, Ziemann F, Boulbitch AA, et al (1998) Local measurements of viscoelastic parameters of adherent cell surfaces by magnetic bead microrheometry. Biophys J 75:2038–2049.

208

J.C. Selby

157. Matthews BD, Overby DR, Alenghat FJ, et al (2004) Mechanical properties of individual focal adhesions probed with a magnetic microneedle. Biochem Biophys Res Commun 313:758–764. 158. Saif MTA, Sager CR, Coyer S (2003) Functionalized biomicroelectromechanical systems sensors for force response study at local adhesion sites of single living cells on substrates. Ann Biomed Eng 31:950–961. 159. Yang S, Saif T (2005) Reversible and repeatable linear local cell force response under large stretches. Exp Cell Res 305:42–50. 160. Tan JL, Tien J, Pirone DM, et al (2003) Cells lying on a bed of microneedles: an approach to isolate mechanical force. Proc Natl Acad Sci U S A 100:1484–1489. 161. Balaban NQ, Schwarz US, Riveline D, et al (2001) Force and focal adhesion assembly: a close relationship studied using elastic micropatterned substrates. Nat Cell Biol 3:466–472. 162. Thoumine O, Ott A, Cardoso O, et al (1999) Microplates: a new tool for manipulation and mechanical perturbation of individual cells. J Biochem Biophys Meth 39:47–62. 163. Suresh S, Spatz J, Mills JP, et al (2005) Connections between single-cell biomechanics and human disease states: gastrointestinal cancer and malaria. Acta Biomater 1:15–30. 164. Guck J, Ananthakrishnan R, Mahmood H, et al (2001) The optical stretcher: a novel laser tool to micromanipulate cells. Biophys J 81:767–784. 165. Radmacher M (1997) Measuring the elastic properties of biological samples with the AFM. IEEE Eng Med Biol 16:47–57. 166. Bowen WR, Hilal N, Lovitt RW, et al (1998) Direct measurement of the force of adhesion of a single biological cell using an atomic force microscope. Colloid Surf A 136:231–234. 167. Hoh JH, Schoenenberger CA (1994) Surface morphology and mechanical properties of MDCK monolayers by atomic force microscopy. J Cell Sci 107:1105–1114. 168. Waschke J, Bruggeman P, Baumgartner W, et al (2005) Pemphigus foliaceus IgG causes dissociation of desmoglein 1-containing junctions without blocking desmoglein 1 transinteraction. J Clin Invest 115:3157–3165. 169. Waschke J, Menendez-Castro C, Bruggeman P, et al (2007) Imaging and force spectroscopy on desmoglein 1 using atomic force microscopy reveal multivalent Ca2+-dependent, lowaffinity trans-interaction. J Memb Sci 216:83–92. 170. Heupel W-M, Zillikens D, Drenckhahn D, et al (2008) Pemphigus vulgaris IgG directly inhibit desmoglein 3-mediated transinteraction. J Immunol 181:1825–1834. 171. Heupel W-M, Engerer P, Schmidt E, et al (2009) Pemphigus vulgaris IgG cause loss of desmoglein-mediated adhesion and keratinocyte dissociation independent of epidermal growth factor receptor. Am J Pathol 174:475–485. 172. Spindler V, Heupel W-M, Efthymiadis A, et al (2009) Desmocollin 3-mediated binding is crucial for keratinocyte cohesion and is impaired in pemphigus. J Biol Chem 284:30556–30564. 173. Fabry B, Maksym GN, Butler JP, et al (2001) Scaling the microrheology of living cells. Phys Rev Lett 87:148102. 174. Wang N, Tolic-Norrelykke IM, Chen JX, et al (2002) Cell prestress. I. Stiffness and prestress are closely associated in adherent contractile cells. Am J Physiol Cell Physiol 282: C606–C616. 175. Hennings H, Holbrook KA (1983) Calcium regulation of cell-cell contact and differentiation of epidermal cells in culture – an ultrastructural-study. Exp Cell Res 143:127–142. 176. Green KJ, Geiger B, Jones JCR, et al (1987) The relationship between intermediate filaments and microfilaments before and during the formation of desmosomes and adherens-type junctions in mouse epidermal keratinocytes. J Cell Biol 104:1389–1402. 177. Brown TD (2000) Techniques for mechanical stimulation of cells in vitro: a review. J Biomech 33:3–14. 178. Selby JC, Shannon MA (2008) A method to fabricate mesoscopic freestanding polydimethylsiloxane membranes used to probe the rheology of an epithelial sheet. J Biochem Biophys Meth 70:932–944.

9 Mechanobiology of Epidermal Keratinocytes

209

179. Selby JC, Shannon MA (2007) Apparatus for measuring the finite load-deformation behavior of a sheet of epithelial cells cultured on a mesoscopic freestanding elastomer membrane. Rev Sci Instrum 78:094301. 180. Selby JC, Shannon MA (2007) Mechanical response of a living human epidermal keratinocyte sheet as measured in a composite diaphragm inflation experiment. Biorheology 44:319–348. 181. Yano S, Komine M, Fujimoto M, et al (2004) Mechanical stretching in vitro regulates signal transduction pathways and cellular proliferation in human epidermal keratinocytes. J Invest Dermatol 122:783–790. 182. Kippenberger S, Bern A, Loitsch S, et al (2000) Signaling of mechanical stretch in human keratinocytes via MAP kinases. J Invest Dermatol 114:408–412. 183. Russell D, Andrews PA, James J, et al (2004) Mechanical stress induces profound remodelling of keratin filaments and cell junctions in epidermolysis bullosa simplex keratinocytes. J Cell Sci 117:5233–5243. 184. Fudge D, Russell D, Beriault D, et al (2008) The intermediate filament network in cultured human keratinocytes is remarkably extensible and resilient. PLoS ONE doi:10.1371/journal. pone.0002327. 185. Janmey PA, Euteneuer U, Traub P, et al (1991) Viscoelastic properties of vimentin compared with other filamentous biopolymer networks. J Cell Biol 113:155–160. 186. Janmey PA (1991) A torsion pendulum for measurement of the viscoelasticity of biopolymers and its application to actin networks. J Biochem Biophys Meth 22:41–53. 187. Xu JY, Tseng Y, Wirtz D (2000) Strain hardening of actin filament networks – regulation by the dynamic cross-linking protein alpha-actinin. J Biol Chem 275:35886–35892. 188. Xu JY, Palmer A, Wirtz D (1998) Rheology and microrheology of semiflexible polymer solutions: actin filament networks. Macromolecules 31:6486–6492. 189. Tseng Y, Schafer BW, Almo SC, et al (2002) Functional synergy of actin filament crosslinking proteins. J Biol Chem 277:25609–25616. 190. Ma LL, Xu JY, Coulombe PA, et al (1999) Keratin filament suspensions show unique micromechanical properties. J Biol Chem 274:19145–19151. 191. Yamada S, Wirtz D, Coulombe PA (2003) The mechanical properties of simple epithelial keratins 8 and 18: discriminating between interfacial and bulk elasticities. J Struct Biol 143:45–55. 192. Yamada S, Wirtz D, Coulombe PA (2002) Pairwise assembly determines the intrinsic potential for self-organization and mechanical properties of keratin filaments. Mol Biol Cell 13:382–391. 193. Ma LL, Yamada S, Wirtz D, et al (2001) A “hot-spot” mutation alters the mechanical properties of keratin filament networks. Nat Cell Biol 3:503–506. 194. Bousquet O, Ma LL, Yamada S, et al (2001) The nonhelical tail domain of keratin 14 promotes filament bundling and enhances the mechanical properties of keratin intermediate filaments in vitro. J Cell Biol 155:747–753. 195. Liu XM, Pollack GH (2002) Mechanics of F-actin characterized with microfabricated cantilevers. Biophys J 83:2705–2715. 196. Kis A, Kasas S, Babic B, et al (2002) Nanomechanics of microtubules. Phys Rev Lett 89:248101. 197. Mucke N, Kreplak L, Kirmse R, et al (2004) Assessing the flexibility of intermediate filaments by atomic force microscopy. J Mol Biol 335:1241–1250. 198. Guzman C, Jeney S, Kreplak L, et al (2006) Exploring the mechanical properties of single vimentin intermediate filaments by atomic force microscopy. J Mol Biol 360:623–630. 199. Kiss B, Karsai A, Kellermayer MSZ (2006) Nanomechanical properties of desmin intermediate filaments. J Struct Biol 155:327–339. 200. Kreplak L, Bar H, Leterrier JF, et al (2005) Exploring the mechanical behavior of single intermediate filaments. J Mol Biol 354:569–577.

210

J.C. Selby

201. Fudge DS, Gardner KH, Forsyth VT, et al (2003) The mechanical properties of hydrated intermediate filaments: insights from hagfish slime threads. Biophys J 85:2015–2027. 202. Kreplak L, Fudge D (2006) Biomechanical properties of intermediate filaments: from tissues to single filaments and back. Bioessays 29:26–35. 203. Qin Z, Buehler MJ, Kreplak L (2010) A multi-scale approach to understand the mechanobiology of intermediate filaments. J Biomech 43:15–22. 204. Herrmann H, Bar H, Kreplak L, et al (2007) Intermediate filaments: from cell architecture to nanomechanics. Nat Rev Mol Cell Biol 8:562–573. 205. Leckband D, Israelachvili J (2001) Intermolecular forces in biology. Q Rev Biophys 34:105–267. 206. Florin EL, Moy VT, Gaub HE (1994) Adhesion forces between individual ligand-receptor pairs. Science 264:415–417. 207. McGuiggan PM, Israelachvili JN (1990) Adhesion and short-range forces between surfaces 2. Effects of surface lattice mismatch. J Mater Res 5:2232–2243. 208. Israelachvili JN, McGuiggan PM (1990) Adhesion and short-range forces between surfaces 1. New apparatus for surface force measurements. J Mater Res 5:2223–2231. 209. Helm CA, Knoll W, Israelachvili JN (1991) Measurement of ligand receptor interactions. Proc Natl Acad Sci U S A 88:8169–8173. 210. Svoboda K, Block SM (1994) Biological applications of optical forces. Annu Rev Biophys Biomol Struct 23:247–285. 211. Evans E, Ritchie K, Merkel R (1995) Sensitive force technique to probe molecular adhesion and structural linkages at biological interfaces. Biophys J 68:2580–2587. 212. Hanley W, McCarty O, Jadhav S, et al (2003) Single molecule characterization of P-selectin/ ligand binding. J Biol Chem 278:10556–10561. 213. Leckband D, Prakasam A (2006) Mechanism and dynamics of cadherin adhesion. Annu Rev Biomed Eng 8:259–287. 214. Sivasankar S, Gumbiner B, Leckband D (2001) Direct measurements of multiple adhesive alignments and unbinding trajectories between cadherin extracellular domains. Biophys J 80:1758–1768. 215. Leckband D, Sivasankar S (2000) Mechanism of hemophilic cadherin adhesion. Curr Opin Cell Biol 12:587–592. 216. Sivasankar S, Brieher W, Lavrik N, et al (1999) Direct molecular force measurements of multiple adhesive interactions between cadherin ectodomains. Proc Natl Acad Sci U S A 96:11820–11824. 217. Panorchan P, Thompson MS, Davis KJ, et al (2006) Single-molecule analysis of cadherinmediated cell-cell adhesion. J Cell Sci 119:66–74. 218. Leckband D (2004) Molecular mechanisms of cell adhesion: new perspectives from surface force measurements. J Adhes 80:409–432. 219. Bayas MV, Leung A, Evans E, et al (2006) Lifetime measurements reveal kinetic differences between homophilic cadherin bonds. Biophys J 90:1385–1395. 220. Fung CKM, Seiffert-Sinha K, Lai KWC, et al (2010) Investigation of human keratinocyte cell adhesion using atomic force microscopy. Nanomedicine: NMB 6:191–200. 221. Elias PM, Ahn SK, Brown BE, et al (2002) Origin of the epidermal calcium gradient: regulation by barrier status and role of active vs. passive mechanisms. J Invest Dermatol 119:1269–1274.

Chapter 10

Quantifying Cell-Matrix Deformations in Three Dimensions Christian Franck and Stacey A. Maskarinec

This chapter is part of Section IV: Tools for Exploring Mechanobiology

Abstract In recent years, the importance of mechanical forces in directing cellular function has been recognized as a significant factor in biological and physiological processes. A complete quantification of cell tractions during cell-material interactions can lead to a deeper understanding of the fundamental role these forces play in cell biology. Previous research has contributed significant descriptions of celltissue interactions by quantifying cell tractions in two-dimensional environments; however, most physiological processes are three-dimensional in nature. This chapter presents a full-field imaging technique capable of quantitatively measuring cell tractions in all three spatial dimensions, and hence addresses the need of a threedimensional quantitative imaging technique to gain insight into the fundamental role of physical forces in biological processes. In this chapter we will explain and demonstrate the working principles of the technique and how it can be used to quantify the three-dimensional mechanical interactions of cells and their extracellular matrix by measuring cell-induced displacement and traction fields during the migration of individual fibroblast cells on polyacrylamide gels.

10.1

Introduction and Rationale

The exchange of physical forces in cell-cell and cell-matrix interactions plays a significant role in regulating a variety of physiological and pathological processes including wound healing, angiogenesis, metastasis and embryogenesis [1–3]. For example, in tumor cells benign to malignant phenotype transformation has been shown to be strongly influenced by the surrounding extracellular microstructure and its mechanical signature, particularly in three-dimensional environments [4–6]. Hence, quantification and understanding of the nature of cell-ECM interactions and regulation within three-dimensional environments become an important part for the development of new biomaterials and clinical diagnostics and treatments.

C. Franck (*) School of Engineering, Brown University, Providence, RI 02912, USA e-mail: [email protected] A. Wagoner Johnson and Brendan A.C. Harley (eds.), Mechanobiology of Cell-Cell and Cell-Matrix Interactions, DOI 10.1007/978-1-4419-8083-0_10, # Springer Science+Business Media, LLC 2011

211

212

C. Franck and S.A. Maskarinec

Within the last few decades, studies began to quantify surface tractions on two-dimensional substrates that are developed by migrating cells through a variety of techniques. For example, in 1980 Harris et al. demonstrated that cellular forces could be visualized by tracking the wrinkling formation of thin elastic silicone rubber substrates due to applied cell stresses [7]. However, since wrinkle formation is an intrinsically nonlinear and unstable process, the quantitative characterization using this technique is difficult. In 1995 Oliver et al. and Dembo et al. developed a quantitative, non-wrinkling technique called traction force microscopy (TFM) to study fibroblast migration on two-dimensional substrate surfaces [8–10]. While other experimental techniques, such as micropillars and embedded force sensors have made significant contributions in quantifying cell-matrix interactions [1, 11], traction force microscopy currently remains one of the most widely used methods in determining cellular tractions during cell-cell and cell-matrix processes [12–15]. Traction force microscopy utilizes optical phase and wide-field microscopy to track substrate surface displacements due to cellular tractions through the spatial correlation of fluorescent particles embedded in a polyacrylamide or similar polymer based substrate. Polyacrylamide gels are among the most commonly used polymer-based substrate materials in studying cell force responses due to their mechanical tunability, optical translucency, and elastic material behavior [16]. By controlling the incorporation percentage or mole fraction of added crosslinker N, N-methylene-bis-acrylamide (BIS), the Young’s modulus of each particular polyacrylamide gel can be modified with a typical modulus ranging from ~1 to 30 kPa, which is in the range of physiologically relevant moduli [12–14, 17]. To record cell surface deformations, cells are initially seeded on the substrate material and allowed to spread or migrate. After some time, a first image is captured optically, where typically both the cell and tracker particles are recorded simultaneously. Then, cells are chemically detached from the surface through trypsinization or similar chemical treatment. A second image is captured (without moving the microscope’s objective) to serve as the undeformed or reference configuration. Cell-induced substrate displacements are then determined from the two images by using either a single particle tracking or digital image correlation algorithm. The resulting gel displacements are converted into tractions using the inverse Boussinesq formulation, where the Boussinesq theory describes the displacement equilibrium solutions inside a semi-infinite elastic half-space with applied forces at its free boundary [18]. However, since the Boussinesq formulation needs to be utilized inversely to compute cell tractions, it has the complication that the solution is no longer unique and the computation itself can become expensive. Hence, additional regularization and iteration algorithms are needed to provide a stable solution [12, 13, 19]. Although previously published traction force microscopy reports have contributed much to describing cell behavior and local cell-ECM interactions, their experimental data and interpretation remain inherently restricted to two dimensions. This has implication in many of our currently existing cell motility and

10

Quantifying Cell-Matrix Deformations in Three Dimensions

213

mechanotransduction models that are based on those experimental findings [20, 21]. To gain deeper insights into cell-matrix interactions and to extract quantitative physical information from cellular physiological processes, new three-dimensional imaging techniques are needed. The three-dimensional quantitative imaging technique presented in this chapter provides a unique way to determine cellular tractions in all three dimensions and thus opens new avenues for quantifying cell-matrix interactions in more realistic three-dimensional environments.

10.2

Development of Three-Dimensional Traction Force Microscopy (3D TFM)

This section describes the development of a newly developed three-dimensional traction force microscopy technique [22–24]. This method has the capability of determining the three-dimensional cellinduced full-field displacements and tractions inside transparent biomaterials, such as polyacrylamide and collagen gels. Figure 10.1 provides a schematic overview of this technique showing how cell-induced traction forces are computed in all three spatial dimensions.

Fig. 10.1 Schematic overview of the quantitative three-dimensional traction force microscopy technique illustrating the methodology to compute the full-field tractions of migrating cells in all three spatial dimensions

214

C. Franck and S.A. Maskarinec

The method presented here is based on laser scanning confocal microscopy (LSCM) to acquire three-dimensional volumetric images, while a digital volume correlation (DVC) algorithm is used to determine the full-field displacements. The next sections will explain each component of the technique in more detail.

10.2.1 Laser Scanning Confocal Microscopy (LSCM) Confocal microscopy has emerged as a powerful imaging technique owing to the optical sectioning capability enabling construction of three-dimensional images. In conventional wide-field microscopy, light is collected from the entire sample volume, including the focal plane as well as all other planes, whereas, in confocal microscopy, light is generally collected from the focal plane only. This is achieved by using a pinhole in front of a photomultiplier tube (PMT) detector that blocks the incoming light from all other planes. As illustrated in Fig. 10.2, the solid line represents light reflected or emitted from the focal plane, while the dashed line represents light from the out-of-focus plane. The overall contrast and resolution of the image is significantly increased as compared to conventional wide-field microscopy where the image is blurred by out-of-plane light. Furthermore, the inherent optical sectioning of the specimen in confocal microscopy allows the assembly of three dimensional image volumes by stacking together individually acquired planar slices. In a LSCM system, a laser with a single-diffraction limited spot size is used to sequentially scan a selected focal plane. Thus, the image is not formed using a CCD camera as in conventional microscopy, but rather the image is a result of the light’s interaction with successive areas

Fig. 10.2 Illustration of the confocal imaging principle (solid lines ¼ in-focus light; dashed lines ¼ out-of-focus light). Reprinted with the permission of Franck et al. [24] and the Society for Experimental Mechanics

10

Quantifying Cell-Matrix Deformations in Three Dimensions

215

of the specimen, i.e. the image is recorded pixel by pixel analogous to a scanning electron microscope. The resulting image is generally superior in resolution to images recorded by conventional optical microscopy. To further improve the resolution of laser scanning confocal microscopy deconvolution algorithms can be employed; however, such descriptions are beyond the scope of this chapter. Figure 10.3 shows an example of two three-dimensional volumetric images recorded using LSCM. The submicron-sized particles are shown in blue. A more detailed description of the confocal principle and the current applications of confocal microscopy are well documented and can be found elsewhere [25, 26]. Similar to traditional, two-dimensional traction force microscopy, fluorescent markers are added to the transparent materials of interest and they serve as image sources for constructing the three-dimensional images and as markers for performing DVC described in the next section.

Fig. 10.3 Schematic illustration of the principle of digital volume correlation (DVC). Reprinted with the permission of Franck et al. [24] and the Society of Experimental Mechanics

216

C. Franck and S.A. Maskarinec

10.2.2 Digital Volume Correlation (DVC) Laser scanning confocal microscopy provides discretized volume images visualizing three-dimensional structural patterns of fluorescent markers in a transparent sample. Then, the combination of digital volume correlation (DVC) and confocal images is used to achieve three-dimensional full-field deformation measurements as an extension of the vision-based surface deformation measurement techniques, well-known as digital image correlation (DIC) [27]. The basic principle of the DVC is schematically illustrated in Fig. 10.3. Two confocal volume images of a polyacrylamide gel with randomly dispersed fluorescent particles are obtained before and after mechanical loading. Then, the two images are subdivided into a set of subvolumes that are centered on the points of interest. Using each pair of corresponding subvolume images, the respective local displacement vector can be obtained from threedimensional volume correlation methods. Consider two scalar signals f(x) and g(x) which represent a pair of intensity patterns in a subvolume O before and after a continuous mapping, y^ðxÞ : x ! y, respectively. Assuming that the signal is locally invariant during the mapping, f(x)¼ g(y(x)), correlation matching by subvolume can be obtained by finding an optimal mapping that maximizes the cross-correlation functional defined as Z mð^ yÞ ¼

f ðxÞgðyðxÞÞdOx

(10.1)

The methodology is illustrated using a translational volume correlation, which is presented below. The continuous mapping is assumed to be a rigid body translation, y ¼ x + c, and the cross-correlation function is represented as a function of a displacement vector c as Z mðcÞ ¼

f ðxÞgðx þ cÞdOx

(10.2)

The cross-correlation function can be written using Fourier transforms as mðcÞ ¼ =1 =½f ðxÞ =½gðxÞ

(10.3)

where the Fourier transform of f(x) is defined as cross-correlation function can be written using Fourier transforms as Z =½f ðxÞ ¼

f ðxÞeik x dOx

(10.4)

and * denotes the complex conjugate. The discrete cross-correlation function can be computed efficiently by using the fast Fourier transform (FFT) algorithm. Then, the rigid body translation vector c can be estimated from the location of the

10

Quantifying Cell-Matrix Deformations in Three Dimensions

217

cross-correlation peak with respect to the origin. Finding the displacement vector c from the discrete cross-correlation function is straightforward and provides halfvoxel accuracy. While this particular translation-based correlation delivers accurate results for cell-material interactions where the applied local strains are small (<5%), higher order approximations of the deformation field within each subvolume are required for large deformation measurements in soft materials. These formulations are part of ongoing research and can be found elsewhere [23, 28]. Determining the displacement vector c within subvoxel accuracy generally requires fitting and interpolation of the correlation function near the peak. Generally, a three-dimensional quadratic polynomial fitting can be used to fit the correlation function near the peak and hence provides improved subvoxel accuracy over lower order fitting polynomials. Finally, the displacement gradients are computed by using a three-dimensional least-square fitting of each displacement component in a 3 3 3 grid of neighboring data points. Alternatively, more sophisticated smoothing or filtering algorithm can be employed before or during the gradient calculation to obtain smoother strain fields if needed. Once the displacement gradient fields are determined, either infinitesimal or finite strain values can be computed from the displacement gradient fields.

10.2.3 Calculation of Displacement and Tractions All displacement and strain fields are determined from three-dimensional LSCM volume stacks using described DVC algorithm. In order to establish the maximum resolution sensitivity of this technique, fibronectin-functionalized polyacrylamide samples were prepared without cells and imaged under identical conditions as the live cell measurements and the three dimensional displacement and strain fields were computed (zero load images). Following statistical error analysis, these measurements yielded that the 3D TFM technique is capable of detecting displacement changes greater than 0.12 mm under the described imaging conditions (see next section), representing subpixel or submicron accuracy. The typical grid spacing of the digital volume correlation calculations is 2 mm but can be decreased at an increase in computational cost. All of the calculated and presented displacements, tractions and surface normals are functions of the generalized Cartesian coordinates x1, x2, x3 (x, y, z).

10.2.3.1

Definition of the Three-Dimensional Displacement Vector

The three-dimensional displacement vector u is defined as follows, u ¼ ½u1 u2 u3 T , pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ with its magnitude given by juj ¼ u21 þ u22 þ u23 , where u1, u2, u3 are the components of the displacement vector.

218

10.2.3.2

C. Franck and S.A. Maskarinec

Dynamic Cell Traction Calculations

In order to calculate traction stresses including surface tractions, the stress tensor s needs to be determined first, and will need to be calculated based on the experimentally determined constitutive properties of the cell matrix or cell substrate, which in this case is polyacrylamide. As will be described later, the constitutive properties of polyacrylamide were determined to be linearly elastic and incompressible [29]. Hence, the material behavior for polyacrylamide can be mathematically expressed as s ¼ 2me

(10.5)

where e is the strain tensor and m is the shear modulus, which can be related to Young’s modulus E and Poisson’s ratio n by E ¼ 2mð1 þ vÞ

(10.6)

The stress (s) and strain (e) tensors can be written in matrix form, 0

s11 s ¼ @ s12 s13 0

e11 e ¼ @ e12 e13

s12 s22 s23

1 s13 s23 A; s33

e12 e22 e23

1 e13 e23 A: e33

Then, calculation of the tractions involves using the well-known Cauchy relation [29], T = sn

(10.7)

where T is defined as the three-dimensional traction vector, and n (components, n1, n2, n3) is the surface normal of an arbitrary plane on which T (components, T1, T2, T3) acts given by T ¼ ½T1 T2 T3 T ,

and

n ¼ ½n1 n2 n3 T :

For example, to compute the tractions directly underneath the cell on the planar geometry of the polyacrylamide gels the surface normal was taken to be [0 0 1]T. All samples were imaged via LSCM to ensure that their surfaces were microscopically flat such that the x–y plane below the cell could indeed be represented by a [0 0 1]T normal vector. The magnitude of the three-dimensional traction vector is then defined as jT j ¼

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ T12 þ T22 þ T32

10

Quantifying Cell-Matrix Deformations in Three Dimensions

Table 10.1 Baseline (no load) displacement and strain values of the 3D TFM technique

‐ u1 (voxel) u2 (voxel) u3 (voxel) e11 (%) e22 (%) e33 (%)

25 C 0.0605 0.0541 0.2106 6.39 103 9.80 103 0.26

219 37 C 0.0289 0.0282 0.187 6.59 102 6.04 102 0.526

All tractions are presented in units of pN/mm2 or Pascal (Pa). In this chapter the 1 (x1) and 2 (x2) components refer to the in-plane quantities, and the 3 (x3) component corresponds to the out-of-plane or normal quantities.

10.2.3.3

Resolution and Measurement Sensitivity

To assess the sensitivity of the 3D TFM technique experiments are performed using the same materials setup as in case for the migrating fibroblasts but without any cells present (control). Hence, the measured displacements and calculated tractions are solely due to thermal fluctuations and instrumental error, and thus establish the sensitivity of the displacement and strain. Table 10.1 shows typical displacement and strain values the 3D TFM technique can achieve, both at room temperature and 37 C. Using the above-described procedure of calculating tractions and the strain information calculated for the control samples, the traction resolution can be determined. For example, for a polyacrylamide gel of Young’s modulus of 9.64 kPa [22] the technique can accurately determine tractions that are greater than 80 Pa or 80 pN/mm2.

10.3

Experimental Design

This section describes the experimental procedures performed to determine the three-dimensional displacements and tractions of migrating fibroblasts on polyacrylamide gels.

10.3.1 Preparation of Activated Coverslips Glass coverslips (Gold-Seal coverslip No. 0, Electron Microscopy Sciences, Hatfield, PA) were chemically modified to allow covalent attachment of polyacrylamide films using previously established protocols with some modifications [10, 29]. Briefly, coverslips were rinsed with ethanol before being placed in a dish containing a solution of 0.5% (v/v) 3-aminopropyltrimethoxysilane (Gelest, Inc., Morrisville, PA) in ethanol for 5 min. The coverslips were removed from the dish

220

C. Franck and S.A. Maskarinec

and rinsed with a stream of ethanol before being immediately placed treated-side up in a solution of 0.1% (v/v) glutaraldehyde (Polysciences, Inc., Warrington, PA) in water for 30 min. Activated coverslips were rinsed with a stream of deionized water and left to dry for several hours at 60 C.

10.3.2 Preparation of Polyacrylamide Films Thin films of polyacrylamide were generated and fused to activated coverslips using a protocol adopted from previous reports with some modifications [13, 30]. Solutions of acrylamide (40% w/v, Bio-Rad, Hercules, CA) and N, N-methylene-bisacrylamide (BIS, 2.5% w/v, Bio-Rad, Hercules, CA) were mixed with distilled water to obtain the following concentrations used for tested samples: (1) 10% acrylamide and 0.015% BIS, and (2) 10% acrylamide and 0.0075% BIS. To these solutions, red fluorescent microparticles (0.5 mm, carboxylate-modified, Molecular Probes, Carlsbad, CA) were vortexed for 10 s and subsequently added in a volume ratio of 9:100. Crosslinking was initiated through the addition of ammonium persulfate (Sigma-Aldrich, St. Louis, MO) and N, N, N0 ,N0 -tetramethyl-ethylenediamine (Invitrogen, Carlsbad, CA). The samples were vortexed for 10 s, and 5–7 ml of the acrylamide solution was pipetted on the surface of a precleaned microscope slide (No. 1, 75 25 mm, Corning, Corning, NY). To generate thicker films, 20–40 ml of the solution was used. The activated surface of the coverslip was then placed on top of the acrylamide droplet, causing the solution to flatten under the weight of the coverslip. The entire assembly was left undisturbed for 5 min, and then placed in a Petri dish (100 mm diameter, VWR, West Chester, PA) containing distilled water for 10–30 min. The bonded sample was then peeled from the microscope slide using forceps and thoroughly rinsed with several volume changes of water.

10.3.3 Functionalization of Polyacrylamide Films with Fibronectin In order to promote cell attachment to polyacrylamide films, a saturating density of fibronectin (FN) was conjugated to the gel surface using the heterobifunctional crosslinker, sulfo-SANPAH (Pierce Chemicals, Rockford, IL). Adopting procedures outlined previously [13, 30], polyacrylamide gel samples were briefly dried in air to rid the surface of excess water before 200 ml of sulfo-SANPAH in water (1.0 mg/ml) was applied. The surface of the sample was then exposed to unfiltered UV light from a high-pressure mercury lamp (Oriel Q 100 W at 5 A, >10 in warm up time) at a distance of 10 in. for 7.5 min. The darkened sulfo-SANPAH solution was removed from the surface of the gel and replaced with another 200 ml aliquot

10

Quantifying Cell-Matrix Deformations in Three Dimensions

221

and irradiated for another 7.5 min for a total of 15 min of UV exposure. The samples were then rinsed vigorously with water for 5 min, and adhered to the bottom of 60 mm Petri dishes (Becton Dickinson, Franklin Lakes, NJ) by applying a thin layer of vacuum grease (Dow Corning, Midland, MI) around the perimeter of the unmodified side of the coverslip. The samples were rinsed twice with phosphate buffered saline (PBS, pH 7.4), covered with a solution of fibronectin (FN, 0.2 mg/ml, Millipore, Billerica, MA) and left undisturbed overnight at 4 C. Following overnight incubation, the substrates were rinsed three times with PBS.

10.3.4 Cell Culture Prior to depositing cells, FN-modified gel samples were equilibrated in growth media at 37 C for 15 min. Swiss 3T3 fibroblasts transfected with a GFP-actin vector (a gift from Professor Scott Fraser, California Institute of Technology) were cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum, 50 mg/ml streptomycin, and 50 U/ml penicillin (Invitrogen, Carlsbad, CA). For all experiments, cells were first treated with Mitotracker Deep Red (Molecular Probes, Carlsbad, CA) for 45 min before passaging with trypsinEDTA (0.05%, Invitrogen, Carlsbad, CA). Mitotracker dyes accumulate in actively respiring mitochondria providing an additional method for tracking cells and determining cell viability. Cells were plated at density of ~40,000 cells/coverslip, and incubated on samples for 8–12 h before imaging.

10.3.5 Confocal Microscopy and Time-Lapse Imaging Three-dimensional image stacks were acquired using a Nikon C-1 confocal system mounted on a TE-2000-U inverted optical microscope. A 40 CFI planar fluor air objective with a numerical aperture of 0.6 was used in all experiments. Three laser lines were used to image the cells and the fluorescent microparticles: Argon (488 nm) laser for the GFP-actin, a green HeNe (543 nm) for the microparticles in the polyacrylamide gels, and a red HeNe (633 nm) for the Mitotracker Deep Red for mitochondrial labeling. Confocal stacks were acquired every 35 min for several hours at a resolution of 512 512 H (X Y Z) pixels3, where H ranges from 120 to 250 pixels. In this study, image stacks were acquired every 35 min to minimize phototoxic cell death (significant cell death was observed when time intervals were decreased to 15 and 20 min). It is anticipated that with the development of faster (e.g. resonant or line-scanning or spinning-disc) confocal microscopes the acquisition time of the three-dimensional volume stacks will be significantly reduced. This should allow the user to shorten the observation intervals (<35 min) and to capture cell-induced deformation for relatively fast moving cells and/or cellular processes of shorter timescales.

222

C. Franck and S.A. Maskarinec

Typical imaging areas were ~150 150 mm; images with a larger field of view were captured before and after experiments to ensure that measured displacements were not the result of contributions from neighboring cells. A cell was considered “isolated” if no other cell could be visualized within 80–100 mm of the selected cell. Visually mitotic cells, as evidenced by cell rounding and the formation of two daughter cells, were excluded from analysis. Physiological conditions were maintained during all times by housing the entire confocal microscope inside a custombuilt temperature controlled chamber. The temperature was controlled using a feedback-controlled heater, Air-Therm ATX Air Heater Controller (World Precision Instruments, Sarasota, FL), and an arterial blood gas mixture (5% CO2, 20% O2, 75% N2) was injected into the chamber in order to maintain appropriate culturing conditions.

10.3.6 Mechanical Characterization Using Confined and Unconfined Compression Testing The mechanical properties of polyacrylamide specimens were determined by performing both unconfined and confined compression using a custom-built compression setup [23]. Figure 10.4 shows the raw data for an incremental loading cycle highlighting negligible time-dependent material behavior, and Fig. 10.5 shows the loading-unloading stress strain curve for a typical sample following unconfined compression. The Young’s modulus for the polyacrylamide samples was calculated from each stress-strain curve using the relationship: E¼

Fig. 10.4 Compressive loading increments of polyacrylamide samples highlighting the negligible time-dependent relaxation behavior of the material

s F=A ¼ e Dh=h

10

Quantifying Cell-Matrix Deformations in Three Dimensions

223

Fig. 10.5 Representative stress-strain curves of loading and unloading cycles on cylindrical hydrated polyacrylamide samples demonstrating a linear elastic material response with negligible hysteresis

Table 10.2 Mechanical characterization results for polyacrylamide substrates

Crosslinker volume fraction 0.015% BIS 0.0075% BIS

Young’s modulus (kPa) 9.64 1.12 0.82 0.23

where s and e denote the uniaxial stress and strain, and can be expressed as the applied force per sample contact area, and change in sample height over its original height. Table 10.2 summarizes the unconfined compression test results for two different crosslinker volume fractions. Using the determined Young’s modulus value of the unconfined test case and observing that further compression beyond an initial compression strain of ~0.25% was not possible (due to the Poisson effect) during confined compression tests, Poisson’s ratio was determined to be ~0.48–0.5 according to the following equation: s 1v E ¼ ¼ E e ð1 þ vÞð1 2vÞ where E denotes the measured confined compression modulus, u is the Poisson’s ratio, and E is the Young’s modulus as determined from unconfined compression test. From this set of experiments, Poisson’s ratio was taken to be 0.5, and the material behavior is described as a linearly elastic, isotropic, and incompressible for all traction calculations. In addition, dynamic mechanical shear (DMA) measurements measuring the loss and storage modulus of the polyacrylamide substrata were performed. The results from the DMA measurements showed both moduli, G0 (storage) and G00 (loss), to be time-independent over a frequency range from 0.01 to 100 rad/s, and the ratio G00 /G0 1, which is consistent with the mechanical characterization results from the uniaxial compression experiments.

224

10.4

C. Franck and S.A. Maskarinec

Quantifying 3D Cell-Matrix Interactions of Migrating Fibroblast Cells

10.4.1 Measuring 3D Cell-Applied Displacements and Tractions Full field displacement measurements were carried out using the 3D TFM technique applied to migrating 3T3 fibroblast cells on polyacrylamide gels with a Young’s modulus of E 9.64 kPa. Confocal volume stacks were recorded at 35 min time increments capturing the locomotion of individual cells over several hours. The results depict dynamic cell-ECM interactions of the same cell over several hours without chemical intervention, such as trypsinization to completely detach the cells from the substrate surface. This allows time-lapse observations of a single 3T3 fibroblasts and their dynamic interactions with the substrate during locomotion, and thus provides more temporal details than the traditional one-byone image comparisons between events of cell activity before and after trypsin (chemical detachment) treatment [12, 13, 30]. The general motion of each tracked cell can be best described as a random walk, with changes in direction every 2–4 time frames with an average cell speed of 8 mm/h. The nomenclature of leading and trailing edge describes the frontal and posterior region of the cell during a particular time increment in which a net direction is clearly visible by tracking the cell nucleus. Time t0 denotes the start point of each confocal imaging series, whereas t1 describes the first 35 min time increment (i.e. time frame). Although the initial cell spreading time history was not recorded, the results presented here display dynamic snapshots of single fibroblast cell-matrix interactions since cells were not chemically detached in order to capture the matrix deformations. The substrate thickness for all results is 40 mm. In order to reduce the effects of phototoxicity and photobleaching during cell imaging, the imaged confocal volume size was limited to 48% of the total substrate thickness. As the subsequent results show this volume size is sufficient to capture most of the cellinduced deformation field within the resolution limits of the technique. Each cell is visualized simultaneously with the displacement of the fluorescent particles inside the PA gels using two separate photodetectors. This procedure allows correlating the position of the cell determined by the GFP-actin fluorescent vector construct with the substrate displacement field. GFP-actin highlights the actin filaments of the cell, which are one of the main structural cell filaments. Therefore, GFP-actin can be used to visualize the three-dimensional shape of the cell during the migration increments. However, due to the finite life-time and degradation of the GFP, parts of the cell are occasionally not visible at locations where considerable deformations are observed. Figure 10.6 shows the time evolution of the three-dimensional displacement field beneath a single motile cell on the polyacrylamide substrate between two arbitrary time increments over a time span of 70 min. The color contour plots display the

10

Quantifying Cell-Matrix Deformations in Three Dimensions

225

Fig. 10.6 Surface contour (a, b) and depth contour (c, d) plots of the magnitude of the threedimensional displacement vector at two particular imaging increments t1 and t2 (separated by a 35 min interval) during cell migration

magnitude of the three-dimensional displacement vector in mm. The confocal cell image shown by the intensity distribution of GFP-actin is superimposed on the displacements field to correlate the location of the cell with the detected displacement fields. Figure 10.6 shows a highly polarized cell with locations of displacement concentrations in particular at the leading and trailing edge of the cell. The linear dimension of the cell in all of the plots is approximately 80–100 mm. The direction of cell locomotion is from the lower left to the upper right, and the average cell speed as determined by tracking the nucleus of the cell is 8 mm/h. Figure 10.6a, b display the distribution of the magnitude of the threedimensional displacement vector on the surface plane directly underneath the cell at two different time increments (t1 and t2) with a 35 min time interval in-between recordings. Figure 10.6c, d display the magnitude of the three-dimensional displacement vector along an arbitrarily chosen plane through the thickness of the

226

C. Franck and S.A. Maskarinec

polyacrylamide substrate at the same time increments. The displacement contour slices highlight the dynamic interaction of the cell with its substrate characterized by changes in magnitudes and location of the observed displacements, both along the surface plane (x1-x2) and through the substrate thickness (x1-x3). Since the three-dimensional full-field displacements are determined as shown in Fig. 10.6, the complete three-dimensional strain and stress tensors can be calculated without any a priori assumptions regarding the stress state (e.g. plane stress) and geometry (e.g infinite substrate thickness). This removes previously necessary conditions to determine tractions from two-dimensional data sets, such as an infinite substrate requirement or inverse calculations as within the framework of the Boussinesq theory (13, 15, 20). Here, since the three-dimensional stress state is known, the calculations of cell tractions is straight-forward using (10.7). Figure 10.7 shows the traction fields calculated from the displacement data in Fig. 10.6. The color contours display the magnitude of the three-dimensional traction vector. Both time increments, t1 and t2, show highly localized traction fields with a peak magnitude directly underneath the cell near the cell edge, most

Fig. 10.7 Corresponding surface contour (a, b) and depth contour (c, d) plots of the magnitude of the three-dimensional traction vector to the displacement fields presented in Fig. 10.6

10

Quantifying Cell-Matrix Deformations in Three Dimensions

227

likely corresponding to local force transmission at focal adhesions at both the trailing and leading edge of the cell. Figure 10.7c, d highlight the depth distribution of the cell-applied tractions at an arbitrarily chosen plane underneath the cell. The dynamic mechanical interaction of the cell and its substrate is highlighted in both time frames by a change in location and magnitude of the tractions. As in Fig. 10.6 the outline of the cell is shown in green and superimposed on the traction field data. While the cell’s outline is not always visible directly above some of the traction concentration locations due to the earlier described degradation of the incorporated GFP-actin, the cell is still transmitting force there locally, as has been confirmed through multiple experiments where GFP-actin was clearly visible. Figure 10.8 shows line profiles of the traction distribution at a particular location in Fig. 10.6 for time increment t1. Figure 10.8b displays a one-dimensional profile of the magnitude (absolute value) of each surface traction component (|T1|, |T2|, |T3|), including the magnitude of the total three-dimensional traction vector |T| along the highlighted line in Fig. 10.8a. The line profile is drawn across one particular area of stress concentrations at the leading edge of the cell likely corresponding to a focal adhesion site. The magnitude of the total traction vector follows a Gaussian-like profile with a peak value of 1.22 nN/mm over an area of 1 mm. For this particular location and time increment, the in-plane shear tractions |T1| and |T2| dominate the out-of-plane (thickness) tractions with magnitude 2–3 times larger than |T3|. Figure 10.8d displays the decay in the magnitude of the three-dimensional traction vector |T| and its components (|T1|, |T2|, |T3|) through the thickness of the substrate at the trailing edge of the fibroblast. The surface plane where the cell is adhering is located at x3 ¼ 24 mm, and shows the highest value of cellmediated tractions. On the surface, the ratio of the out-of-plane (normal) traction component |T3| to both the in-plane components |T1| and |T2| is 1.6 and 5, respectively, and thus |T3| is the major contributing traction component at that location during the particular time increment.

10.4.2 Time Evolution of Cell-Mediated Displacement and Traction Fields Comparing series of confocal stacks affords insight into the time evolution of the displacement and traction profiles produced during cell migration. A planar slice through the top surface of the sample shows how the pattern of displacements changes as a cell moves along the substrate surface (Fig. 10.9). The set of images in Fig. 10.9a represents 3D data collapsed into 2D images, and four successive images of this type are displayed for a 140-min time course of cell migration. The middle panel of parallel images (Fig. 10.9b) shows the total displacements (3D) of the surface plane as colored contours, while the vectors indicate only in-plane (2D) displacements. These displacements are then transformed into tractions in the last

228

C. Franck and S.A. Maskarinec

Fig. 10.8 Cell-induced surface tractions contour plot (a) at a particular time increment (t1 ¼ 35 min) during migration. The pink line depicts the location of the generated one-dimensional plot shown in (b), illustrating the distribution of the magnitude of the total traction vector and its in-plane (T1, T2) and normal (T3) components along the selected line-cut highlighted in (a); (c, d) display the cell-induced traction contour and line plot profiles as a function of depth (x3) through the thickness of the gel

panel of complementary images (Fig. 10.9c) using the experimentally determined displacement field and material properties as described above. A polarized cell extending processes in the form of a leading edge generates localized, contractile stresses that change as it propels itself forward (Fig. 10.9c, 35–70 min). Stresses near the rear of the cell are also observed during migration and have been shown in previous studies to correspond to focal complexes that transmit significant cellular forces [31]. Notable forces detected near the rear of the cell during migration in this study may explain previously reported observations which demonstrated that on surfaces of high and intermediate adhesiveness, cells can become so tightly bound to the substrate that a portion of their trailing edge can be

10

Quantifying Cell-Matrix Deformations in Three Dimensions

229

Fig. 10.9 Time evolution of displacement and traction contours during cell migration for a single cell. (a) A series of four successive confocal images recorded during time-lapse imaging experiment. The images shown were captured every 35 min for a total of 140 min. (b) Consecutive displacement contours and vectors resulting from the motile cell shown in (a). The color bar represents the magnitude of the 3D displacement vectors, and the arrows represent the in-plane displacement vectors. (c) Complementary traction results calculated directly from displacements in Fig. 10.3b>. The scale bar represents 3D tractions in pN/mm2, and the arrows represent the inplane forces. Figure reprinted with the permission of Maskarinec et al. (2009) and the Proceedings of the National Academy of Sciences

fractured during migration, leaving behind adhesion receptors (e.g., integrins) on the surface [32]. It is important to note that the magnitudes of the forces determined here lie within the range reported for fibroblasts [13, 33], and that these measurements only present the distribution of cellular forces that occur during movement and thus cannot be directly compared to a total detachment force. Confirmation of the cellular origins of the measured displacements can be accomplished by replicating time-lapse experiments followed by treatment with

230

C. Franck and S.A. Maskarinec

blebbistatin [24]. Addition of this myosin-II-specific blocker [34] resulted in a gradual inhibition of cell-generated displacements, showing that the localized matrix deformations observed were caused by contraction of actin bundles by myosin-II activity. At certain time points (e.g., see Fig. 10.9c: 105–140 min) significant displacements are measurable beneath the cell body near the nucleus, suggesting that a large portion of the cell’s contact area participates in force exchange with the substrate. These observations are consistent with two-dimensional cell traction studies showing similar results [11, 13–15].

10.5

Conclusions

This chapter describes a recent method [24] to track and quantify cell-matrix interactions and cell-applied tractions in three dimensions by using laser scanning confocal microscopy and digital volume correlation. This combination of techniques allows dynamic interrogation of the complex process of cell migration and yields further insight into the interactions of cells with their extracellular environments. The approach can be used to correlate local force generation with the concentration of focal adhesions during migration and to analyze the interplay of competing force-fields generated by neighboring cells and sheets of cells. Furthermore, the technique can be extended to elucidate changes in tractions due to malignant transformation, force profiles of encapsulated cells, and the effect of soluble factors on force production. Acknowledgments The authors thank Guruswami Ravichandran and David Tirrell for their advice, support and guidance with the project. We also gratefully thank the National Science Foundation through the Center Science and Engineering of Materials (CSEM) at the California Institute of Technology for the support of this project.

References 1. Galbraith CG,Yamada KM, Sheetz MP (2002) The relationship between force and focal complex development. J Cell Biol 159(4):695–705. 2. Lo CM, Wang HB, Dembo M, Wang YL (2000) Cell movement is guided by the rigidity of the substrate. Biophys J 79(1):144–152. 3. Yang JT, Rayburn H, Hynes RO (1995) Cell-adhesion events mediated by alpha(4) integrins are essential in placental and cardiac development. Development 121:549–560. 4. Ingber DE (2008) Can cancer be reversed by engineering the tumor microenvironment? Semin Cancer Biol 18(5):356–364. 5. Paszek MJ, Zahir N, Johnson KR et al (2005) Tensional homeostasis and the malignant phenotype. Cancer Cell 8(3):241–254. 6. Paszek MJ, Weaver VM (2004) The tension mounts: Mechanics meets morphogenesis and malignancy. J Mammary Gland Biol Neoplasia 9(4):325–342.

10

Quantifying Cell-Matrix Deformations in Three Dimensions

231

7. Harris AK, Wild P, Stopak D (1980) Silicone-rubber substrata – new wrinkle in the study of cell locomotion. Science 208(4440):177–179. 8. Dembo M, Oliver T, Ishihara A et al (1996) Imaging the traction stresses exerted by locomoting cells with the elastic substratum method. Biophys J 70(4):2008–2022. 9. Lee J, Leonard M, Oliver T et al (1994) Traction forces generated by locomoting keratocytes. J Cell Biol 127(6):1957–1964. 10. Oliver T, Dembo M, Jacobson K (1995) traction forces in locomoting cells. Cell Motil Cytoskeleton 31(3):225–240. 11. Tan JL, Tien J, Pirone DM et al (2003) Cells lying on a bed of microneedles: An approach to isolate mechanical force. Proc Natl Acad Sci U S A 100(4):1484–1489. 12. Butler JP, Tolic-Norrelykke IM, Fabry B et al (2002) Traction fields, moments, and strain energy that cells exert on their surroundings. Am J Physiol Cell Physiol 282(3):C595–C605. 13. Dembo M, Wang YL (1999) Stresses at the cell-to-substrate interface during locomotion of fibroblasts. Biophys J 76(4):2307–2316. 14. Reinhart-King CA, Dembo M, Hammer DA (2003) Endothelial cell traction forces on RGDderivatized polyacrylamide substrata. Langmuir 19:1573–1579. 15. Schwarz US, Balaban NQ, Riveline D et al (2003) Measurement of cellular forces at focal adhesions using elastic micro-patterned substrates. Mat Sci Eng C 23(3):387–394. 16. Pelham RJ, Wang YL (1997) Cell locomotion and focal adhesions are regulated by substrate flexibility. Proc Natl Acad Sci U S A 94 (25):13661–13665. 17. Levental I, Georges PC, Janmey PA (2007) Soft biological materials and their impact on cell function. Soft Matter 1:299–306. 18. Landau LD, Lifshitz EM, Kosevich AM, Pitaevskii LP (1986) Theory of elasticity. 3rd ed. Pergamon Press, Oxford. 19. Schwarz US, Balaban NQ, Riveline D et al (2002) Calculation of forces at focal adhesions from elastic substrate data: The effect of localized force and the need for regularization. Biophys J 83(3):1380–1394. 20. Ananthakrishnan R, Ehrlicher A (2007) The forces behind cell movement. Int J Biol Sci 3(5):303–317. 21. Safran SA, Gov N, Nicolas A et al (2005) Physics of cell elasticity, shape and adhesion. Phys A 352:171–201. 22. Franck C (2008) Quantitative characterization of 3D deformations of cell interactions with soft biomaterials. California Institute of Technology, Pasadena. 23. Franck C, Hong S, Maskarinec SA et al (2007) Three-dimensional full-field measurements of large deformations in soft materials using confocal microscopy and digital volume correlation. Exp Mech 47(3):427–438. 24. Franck C, Maskarinec SA, Tirrell DA et al (2009) Quantifying cellular traction forces in three dimensions. Proc Natl Acad Sci U S A 106(52):22108–22113. 25. Corle TR, Kino GS (1996) Confocal scanning optical microscopy and related imaging systems. Academic Press, San Diego. 26. Sheppard C, Shotton D (1997) Confocal laser scanning microscopy, Royal Microscopical Society (Great Britain). Microscopy handbooks 38. UK BIOS Scientific Publishers, New York, xii, 106p. 27. Chu TC, Ranson WF, Sutton MA et al (1985) Applications of digital-image-correlation techniques to experimental mechanics. Exp Mech 25(3):232–244. 28. Smith TS, Bay BK, Rashid MM (2002) Digital volume correlation including rotational degrees of freedom during minimization. Exp Mech 42(3):272–278. 29. Fung YC (1994) A first course in continuum mechanics: for physical and biological engineers and scientists. 3rd ed. Prentice-Hall, Englewood Cliffs. 30. Sabass B, Gardel ML, Waterman CM et al (2008) High resolution traction force microscopy based on experimental and computational advances. Biophys J 94(1):207–220. 31. Ridley AJ, Schwartz MA, Burridge K et al (2003) Cell migration: Integrating signals from front to back. Science 302(5651):1704–1709.

232

C. Franck and S.A. Maskarinec

32. Palecek SP, Huttenlocher A, Horwitz AF et al (1998) Physical and biochemical regulation of integrin release during rear detachment of migrating cells. J Cell Sci 111:929–940. 33. Balaban NQ, Schwarz US, Riveline D et al (2001) Force and focal adhesion assembly: a close relationship studied using elastic micropatterned substrates. Nat Cell Biol 3(5):466–472. 34. Kovacs M, Toth J, Hetenyi C et al (2004) Mechanism of blebbistatin inhibition of myosin II. J Biol Chem 279(34):35557–35563.

Chapter 11

Tools for Studying Biomechanical Interactions in Cells Rebecca E. Taylor*, Vikram Mukundan*, and Beth L. Pruitt

This chapter is part of Section IV: Tools for Exploring Mechanobiology

Abstract Cells interact with their environment through forces that are generated and sensed by the cell. Forces generated by cells are in the few nanoNewton to several microNewton range and can change spatially over subcellular size scales. Transducing forces at such size and force scales requires development of platforms that can mechanically interface with cells. We describe several techniques that have been developed to study the role of mechanical forces in cellular processes. The measurement tools include those to measure the forces exerted by the cell on the extracellular environment, internal forces of contraction and the cytoskeletal properties.

11.1

Introduction

11.1.1 Cytoskeletal Structure The generation of force is integral to eukaryotic cell function and morphology. Unlike bacterial or plant cells that have a double membrane cell wall structure to provide structural integrity, eukaryotic cells rely on a meshwork of cross-linked proteins forming a dynamic cell cytoskeleton. This cell structure includes microfilaments of bundled actin fibers, intermediate filaments of proteins like desmin, vimentin, keratins, etc., and microtubules of polymerized tubulin (Fig. 11.1). These structural elements are dynamically and transiently tied together by various crosslinking proteins to govern cytoskeletal stiffness. In the cytoskeleton, actin filaments, intermediate filaments, and microtubules work together as rods and ropes to provide cellular structure and shape. Actin filaments act as tension-bearing ropes and when cross-linked together these filaments can also act as bundles, networks, and gels. Intermediate filaments and microtubules are *These authors contributed equally. B.L. Pruitt (*) Department of Mechanical Engineering and Cardiovascular Institute, Stanford University, Stanford, CA, USA e-mail: [email protected] A. Wagoner Johnson and Brendan A.C. Harley (eds.), Mechanobiology of Cell-Cell and Cell-Matrix Interactions, DOI 10.1007/978-1-4419-8083-0_11, # Springer Science+Business Media, LLC 2011

233

234

R.E. Taylor et al.

Extracellular matrix proteins Microtubules

Focal adhesion complex

Nucleus

Actin stress fiber Intermediate filaments

Actin tight bundle

Actin gel network

Fig. 11.1 Eukaryotic cells contain a number of dynamic structures. Microtubules, filamentous actin and intermediate filaments along with molecular motors drive and maintain this machine. (below left) Fimbrin crosslinks actin into tight bundles while (below right) filamin crosslinks actin into loose gel networks (After Figs. 11.17 and 11.18 in Lodish et al. [73])

compression-bearing rods that contribute bending and twisting stiffness to the cell. Under axial compression these filaments can be modeled as having pinned ends that are free to rotate. The Euler buckling forces for such filaments are proportional to the filament effective elastic modulus, E (N/m2), and area moment of inertia, I (m4), and inversely proportional to overall filament length squared L2 (m2) [113]. Fbuckling ¼

p2 EI 4L2

(11.1)

Actin filaments are soft filaments incapable of individually supporting significant compression. Individually, they typically act as ropes supporting tension. However, filaments can be crosslinked together to further increase cytoskeletal rigidity. Filamentous (F)-actin can be cross-linked by a-actinin (two actin binding sites with a linear 30 nm spacer between) into loose bundles, which allow motor protein participation, or by the smaller protein fimbrin (14 nm spacing between the binding sites)

11

Tools for Studying Biomechanical Interactions in Cells

235

into tight rigid bundles which exclude myosin (Fig. 11.1). Actin filaments can also be cross-linked at nearly right angles into 3D gel-like networks by filamin, which has two actin binding sites with a V-shaped linkage. Spectrin (300 nm spacer between binding sites) links actin filaments into loose two-dimensional networks characterized by greater compliance, e.g., in the membrane cytoskeleton connections of highly deformable erythrocytes. Electrochemical potential energy in the form of ion gradients and membrane potentials and stored energy in adenosine triphosphate (ATP) drive all of the transport mechanisms within the cell. The molecular motors dynein and kinesin walk along microtubules enabling cellular transport. Myosin motors actively “walk” along F-actin causing filaments to slide against each other. For example, a dense meshwork of Factin, crosslinkers, and myosin motors form the actomyosin cell cortex, which is a shelllike structure underneath the membrane that contracts to govern cell movement and shape. Translational and rotational molecular motors utilize electrochemical potential energy from ion gradients and ATP to apply linear forces and torques. These active elements control the mechanical properties of the cell cytoskeleton, and to understand the biomechanics of the cell, we need tools to investigate cytoskeletal mechanics.

11.1.2 Mechanotransduction All cytoskeletal protein structures are constantly polymerizing and depolymerizing and the dynamics and kinetics of cell “stiffness” and force generation in a cell at rest or in motion are thus dynamic properties. These properties are of interest in understanding fundamental questions in health, development, and the progression of disease. How cells rearrange their cytoskeletal structures to divide and migrate during gastrulation and the requisite signaling and molecular pathways are of great interest in developmental biology and regenerative medicine. Changes in cell adhesion, stiffness and motility are central to questions of disease progression in the field of cancer biology. Normal tissue and systemic homeostasis rely on cell sensing and response to changes in the mechanical environment, e.g., extracellular matrix (ECM) or tissue stiffness, shear flows, pressure, tension or compression. Such dynamic environments underlie processes like wound healing [23, 48], cancer metastasis [30, 66], as well as cardiovascular pressure and flow regulation through dilation/contraction of blood vessels and heart muscle contraction and output [45]. A number of cell signaling pathways are identified in the regulation of the cell cytoskeleton and concomitant functions of cell motility, mitosis, and cell adhesion via cell–cell contacts and cell-substrate contacts. During differentiation and renewal, cell phenotypes and morphologies may be characterized by cytoskeletal features, including stiffness and force generation, which are promoted through both biochemical and biomechanical cues. The density, type, and kinetics of cell–cell contacts and their selectivity to cell-type or ECM proteins characteristically vary with cell and tissue type. For example, endothelial and epithelial cells preferentially form barrier layers from sheets of two-dimensional layers of cells. These cell sheets

236

R.E. Taylor et al.

are bound in plane by tight junctions, which can also act like valves for extracellular transport, rivet-like desmosomes, and distributed calcium dependent adhesive sites modulated by cadherins. Organs and tissues, including the skin, use these barrier layers to control transport between the apical and basal surface. For example, the entire cardiovascular system is lined by a monolayer of endothelial cells; a healthy endothelium provides a low-adhesive interface to blood and a barrier to inflammatory and plaque forming proteins. Focal adhesion complexes (FACs) are dynamic macromolecular assemblies that connect the cell cytoskeleton to the ECM. FACs are characterized by integrin mediated receptors (Fig. 11.2). Integrins enable these cell–ECM bonds, because they are transmembrane proteins that bind the cytoskeletal actin filaments to specific amino acid sequences on ECM proteins on the outside of the cell. Integrins enable FACs to transmit force between the environment and the cell cytoskeleton. Migratory cell types like fibroblasts, keratocytes, or lymphocytes assemble and disassemble these FACs and redistribute force through FACs, cell–cell contacts and their dynamic microtubule and actin networks. While much is still unknown about the exact function of each protein involved in a cell–cell cadherins junction (Fig. 11.2), it is clear that cell–cell junctions sense and transmit force from one cell to another. This force transmission is critical for cell maintenance and morphogenesis [125]. Dysfunction in cell adhesivity is implicated in a range of development and health issues including birth defects, blistering diseases, cancer and even viral infection. Further, recent studies have shown that the crosstalk between cell–cell and cell–matrix adhesions is critical for tissue maintenance and cell motility [13, 31, 122]. In addition, the interested reader is directed to recent papers about FAC proteins as mechanosensors [21, 26, 42, 81, 117, 132] and to recent papers about cadherins junctions as mechanosensors [9, 58, 60, 72, 125]. Mechanical forces are an integral component of many physiological functions including contraction, motility, mitosis, reproduction and axonal transport. Measurement of the mechanics of cells encompasses measuring the interactions of cells with their environment, their biological response and its relationship to physiological and

Fig. 11.2 Integrin-based focal adhesions bind cells to the extracellular matrix, and cadherin-based adherens junctions bind cells to each other

11

Tools for Studying Biomechanical Interactions in Cells

237

diseased states. The migration and attachment of cells depend on mechanical cues not just from neighboring cells but also from the environment. Other cell types such as hair cells, heart muscle cells and bone cells function as mechanical transducers to generate or sense mechanical stresses and strains. Additionally, in many developmental processes, cells utilize mechanical forces to remodel and organize themselves into the appropriate environment. Examples of this are cell polarity during mitosis and cell motility and adhesion preference during gastrulation, which promotes germ layer organization in a developing embryo. Further, substrate stiffness has been shown to affect cell adhesion, contractility, and even differentiation [26, 32]. The relationship between structure and function of cells is also evident in the pathological organization of structural components in diseased cells. For example, changes in sarcomeric density, structure and organization in dilated and hypertrophic cardiomyopathy critically affect the contractile properties of cardiac myocytes [30, 67, 74]. Further, this remodeling of cardiomyocyte sarcomeric structure is in response to extended hypertensive overload or overpressure conditions. The mechanical property changes in red blood cells due to malarial infection have shed light on the molecular pathways involved in the cellular level pathology [106]. Similarly, studies of the mechanical properties of epithelial cancer cells have raised questions about the role of cell mechanics in cell mobility and metastasis [106]. The magnitude and localization of generated forces and how they correlate to cytoskeletal proteins such as actomyosin filaments and the various focal adhesion proteins can provide insight into cellular level pathogenesis. Understanding how the forces are generated, sensed and transmitted by the subcellular machinery is an important piece of the cell biology puzzle that could help elucidate what goes wrong when tissues fail to form or heal properly. Cells can both generate and sense forces in the extra cellular matrix and basement membranes that comprise their 3D in vivo environment. These forces can be transmitted through the ECM proteins or through direct cell-to-cell contact and sensed by neighboring cells. Different experiments have shown that transmitted forces alter both cell mechanics and chemical signaling. Indeed, two main paradigms exist for describing how forces alter cytoskeletal assembly and cell function: the force balance paradigm and the chemical signaling paradigm [93]. Research supporting the force-balance paradigm has shown that forces directly influence cellular level processes that regulate both normal and pathological states [25, 51, 52]. Tensegrity or tensional integrity models of cells can be used to describe how cytoskeletal filaments withstand and transmit cellular forces. In tensegrity systems the balance between tension-bearing and compression-bearing elements provide the structural integrity. Tensegrity models of the cytoskeleton are useful for capturing whole cell mechanical properties, including prestress, from simple filament models that contain rods under compression and ropes under tension. Cell contractility and the mechanosensitivity of FACs have been modeled computationally, and these models predict that on semi-infinite substrates focal adhesions will grow with increasing substrate stiffness up to some threshold [80, 127]. Continuum models have also been developed to describe the dynamic cytoskeletal reorganization; these models can predict the decrease in forces generated on increasingly compliant substrates as well as the high concentration of stress fibers at FACs [22].

238

R.E. Taylor et al.

Research supporting the chemical signaling paradigm has demonstrated the small molecules such as focal adhesion kinase (FAK) are mechanosensitive and integrin-transmitted forces cause FAK activation, which results in signals that trigger membrane protrusion, adhesion, migration, and even differentiation [55, 63, 120, 121]. It is likely that both of these paradigms in concert are responsible for cellular responses to transmitted forces. For example, recent studies have shown how forces present in the cellular environment play an important role in events such as wound healing, tissue morphogenesis and regeneration. Tissue morphogenesis is a complex process involving both spatial and temporal interplays of biochemical factors and mechanical tensions. Tension present during embryonic development can influence proper growth and organization [83, 90, 118]. Fibroblasts migrate into wound areas, and the tension generated by the cells closes the wound and reorganizes the matrix to form scar tissue [23, 48]. The tension generated in the ECM also acts as a feedback mechanism to regulate fibroblast function [23, 29, 54]. Force generation in cells occurs at the molecular level by converting chemical energy through the hydrolysis of ATP molecules. The forces are then regulated through a number of biochemical pathways that relate the molecular behavior to organization at the cellular to organ level. Advances in the last couple of decades have enabled measurements of molecular forces and polymer assemblies [5, 114]. The hierarchy of organization leads to complex and varied structures that relate to the functions of the cells and tissues. While the mechanical properties of the polymer and molecular network are significant to understand their organization, the cell body is far more dynamic, exhibiting complex properties that define cellular functions. Understanding this structure-function relationship is a large part of the puzzle relating molecular properties to tissue and organ level biological operation. As a result, researchers seek to examine the role of the mechanical environment and the forces in cell development and function, as well as to exploit mechanical assays of force generation, adhesivity, and stiffness as indicators of cell type and disease progression. Thus, a range of techniques and tools have been exploited to probe the biomechanics of tissues and isolated cells, cell adhesion, cell force generation and cell stiffness. Many of these methods provide powerful insights to this “mechanobiology” of a living system when coupled with biochemistry, molecular biology and imaging technologies.

11.2

Tools for Mechanobiology

In this review, we outline traditional tools for mechanobiology and then focus on several microscale or microfabricated platforms that have been used to study force production and sensing in a variety of cell types (see Table 11.1). These traditional and microfabricated platforms are suitable for studying various biological functions related to mechanics, namely traction forces and migration, biochemical pathways,

11

Tools for Studying Biomechanical Interactions in Cells

Table 11.1 Tools for mechanobiology and their various applications Section and technique Whole cell Cell–cell Tensile loading Yes [14, 70, 82] No Magnetic tweezers (MT) No Yes [57] Optical tweezers (OT) Yes [20] Yes [6, 112] Micropipette aspiration (MA) Yes [33, 49] No Atomic force microscopy (AFM) Yes [105] Yes [10] Traction force microscopy (TFM) Yes [38, 78, 79] Yes [13] Force posts and beam-based Yes [128] Yes [27, 58, 72] polymer transducers MEMS for Applying and Yes [76, 110, 123, 124] Yes, in theory Measuring Force

239

Cell–matrix No Yes [17] Yes [50, 111] No Yes [10] Yes [7, 40, 91] Yes [98, 108] Yes, in theory

focal adhesion measurements, contraction forces, and cytoskeletal stiffness. Finally we discuss a new trend in studying cell mechanics involving embedded sensors and actuators for applying controlled forces to cells. These devices provide new insight to how cells actively respond to applied forces. Excellent reviews of broader methods for assaying cell mechanics provide more detailed information about the traditional cell mechanics assay techniques [5, 114]. Finally, as we focus on measurement techniques, we will not discuss methods used in the macroscale mechanical stimulation of cells such as stretch, compression, and shear. For the interested reader thorough reviews of these techniques are available [16]. When choosing between various techniques for a given experiment, it is important to determine the range of forces or displacements that are needed. Further, these tools operate over bandwidths that may be appropriate for some studies and not others. The graphs in Fig. 11.3 show the force range, displacement range, and bandwidth of popular tools for mechanobiology alongside the force and displacement ranges of popular biological systems. Calibration is an important issue for devices that measure forces smaller than 105 N, because below that level the National Institute of Standards and Technology (NIST) does not have methods for establishing force measurement traceability to the Syste´me International d’Unite´s (SI) [94]. As a result most devices are calibrated through indirect force measurements. Tools like optical and magnetic tweezer (MTs) are typically calibrated by pulling beads through solutions of known viscosity [36, 64]. Micropipette aspiration measurements depend upon system geometry and require well-calibrated suction pressure sources. Elastomeric systems like TFM substrates are difficult to calibrate, and are typically indirectly calibrated through bulk substrate indentation. Tools microfabricated from silicon are easier to calibrate and can be calibrated in many ways. The stiffness of silicon beam-based tools such as AFM cantilevers and piezoresistive microcantilevers can be calculated from material properties and cantilever dimensions according to Euler–Bernoulli beam bending equations. Typically these cantilevers are calibrated using thermal noise or resonant frequency methods, but recent work by NIST has shown that they can also be calibrated by an absolute force standard, the electrostatic force balance [59, 94].

240

R.E. Taylor et al.

Fig. 11.3 Comparison of force, displacement and stiffness ranges. (a) Comparison of various devices for force delivery; AFM range is limited by optical detector and stage travel and the force range shown is accessible only by changing cantilevers (vertical lines) whereas piezoresistive devices, our example MEMS devices, can span 3–5 orders of magnitude depending on the design. Color gradient indicates bandwidth accessible and is related to cantilever size for piezoresistance. MT and OT have similar stiffness/position/force relations, but larger bead size reduces MT bandwidth. Notably, MT is readily parallelized to >10 s of simultaneous observations. (b) Force regimes for C. elegans whole animal studies, dissociated TRN studies, protein/probe calibration and other biological systems (Adapted with permission from Park et al. [87]. Copyright 2007 National Academy of Sciences, USA)

11

Tools for Studying Biomechanical Interactions in Cells

241

11.2.1 Tensile Loading Studies on tissue mechanics began with the use of traditional materials testing platforms such as tensile stress apparatuses [37, 53, 89]. These devices, however, only provide a macroscale, population-level cell study of the tissue. They neither allow isolation of the cellular components from the structural components of a tissue, such as the ECM proteins, nor do they allow for direct measurement of single cells. Single-cell platforms for tensile loading were first developed by scaling down two-point tensile-testing methodologies. For example, one of the most common methods for studying heart muscle cell mechanics employs drawn-glass pipettes with a cell glued between the ends. Cell contractile forces are then inferred from the deflection and stress in the pipettes. This method is very time consuming and allows for the study of a single cell in each test [14, 15, 115]. Carbon fibers confer an interesting advantage over micropipettes because for currently unknown reasons cells can adhere to carbon fibers without glue. Recent advances integrating feedback control of carbon fiber position have enabled advances in the investigation of contractility in primary rat cardiomyocytes [82]. A polysilicon microelectromechanical systems (MEMS) force sensor was developed to scale down contractile force measurements on cardiomyocytes (Fig. 11.4).

Fig. 11.4 Polysilicon beams act as force transducers. The contractile force of the cardiomyocyte clamped between the plates is measured from the deflection on the beams (Reproduced with permission from Lin et al. [70])

242

R.E. Taylor et al.

Lin et al. fabricated a 3D MEMS sensor from surface micromachined components [68, 69]. The 3D structures were manually assembled with micromanipulators after surface micromachining processes and held with the aid of hinges in polysilicon [70]. Cardiac cells were glued to the grips attached to cantilever force sensors and contractile forces were inferred from optically observed deflections of the calibrated polysilicon beam structures. The force sensors were miniaturized to a volume less than 1 mm3 and fit into fluidic chambers perfusing cell culture media. This device was used to measure the contractile forces of rat heart cells in the presence of CaCl2 in a range of Ca2+ concentrations. The average contractile force of rat heart cells in the presence of CaCl2 activating solution was 12 mN. These first MEMS-based measurements of cardiac cells agreed well with pipettebased measurements [109]. These studies were promising despite some deviations from physiological conditions such as lack of substrate adhesions, cell handling damage, and uncontrolled length variation during contraction.

11.2.2 Magnetic Micromanipulation MT experiments employ ligand-coated superparamagnetic beads that bind to cells. In this method a magnetic needle is used to apply pulling forces to the beads, which correspondingly apply forces to surface receptors (Fig. 11.5). Magnetic twisting cytometry (MTC) experiments are similar to MT, but employ ferromagnetic beads that can be used to apply torques to biological systems. Beads can be functionalized and bound to cell surfaces or engulfed by cells. The applied

Fig. 11.5 Magnetic pulling (left) and twisting (right) cytometry can be used to apply controlled torques or pulling forces. Surface beads functionalized with extracellular matrix proteins such as collagen or fibronectin will bind to surface integrins mimicking a cell–matrix adhesion (left), and beads functionalized with cadherins will bind to surface cadherins mimicking a cell–cell adhesion (not shown). An alternate method to investigate cell–cell adhesions is to pull or twist ferromagnetic beads that have been internalized by a cell that is bound to another surface-attached cell (right) (After figure from Ko et al. [57])

11

Tools for Studying Biomechanical Interactions in Cells

243

torque takes the following form where m0 is the permeability of free space, M is a vector representing the microbead magnetic moment, and Ha is a vector representing the external twisting field [64]. Tmag ¼ m0 M Ha

(11.2)

Since the functionalized bead-cell interface can mimic cell–cell or cell–matrix junctions, this technique is useful for studying whole cell elastic and viscoelastic properties as well as various adhesion mechanisms in whole cells. MTC has been used to twist integrin receptors and was shown to increase endothelin-1 gene expression twofold [17]. MTC has also been used on fibroblasts that have engulfed beads to apply controlled forces to intercellular adherens junctions between fibroblasts. At early stages of intercellular adhesion pulling on the junctions resulted in robust Ca2+ transients, supporting the claim that cadherins mediate intercellular mechanotransduction [17, 57].

11.2.3 Optical Tweezers A single beam optical laser trap is capable of pulling dielectric particles towards its center. This technique, invented in 1986 [2], is now known as optical tweezers (OT), and typically is used with functionalized dielectric beads that have been attached to cell membranes or individual molecules. Optical tweezers have been used to study the cell–cell adhesion strength of various cadherins. The adhesion strength of VE-cadherin-coated beads was shown to decrease when cells were treated with permeability-increasing compounds Cytochalasin D and Ca2+ ionophore A23187, and this decrease was blocked when F-actin inside the cell was stabilized [6]. Optical traps were also used to show that growth cones of neurons are reactive to N-cadherin and quickly captured N-cadherin coated beads and dragged them rearward [112]. Cell–matrix adhesion has also been studied using functionalized beads. The force of tether formation in OT-stretched chondrocytes was observed to increase significantly over the course of 1–6 h [50]. Using optical tweezers fibroblasts were held on fibronectin-coated surfaces for varying durations before being released. By studying the adhesion and detachment of fibroblasts, Thoumine et al. measured fibroblast-fibronectin association rates and dissociation rates and successfully modulated those rates by varying compressive force during attachment [111]. On the cellular level this technique has been used to measure forces required to stretch and manipulate whole red blood cells to quantify cell elasticity and viscoelasticity [20]. The application of this technique is particularly appropriate for red blood cells since their function in health and disease is directly tied to their mechanical properties. For example, stiffness of malaria-infected cells is drastically increased and this change reduces the ability of red blood cells to squeeze through capillaries (Fig. 11.6).

244

R.E. Taylor et al.

Fig. 11.6 Optical tweezers can apply controlled stretching to healthy red blood cells (top) and red blood cells infected with Plasmodium falciparum in the schizont infection stage (bottom). Three stages show cells prior to tensile stretching by optical tweezers (left column), at a constant force of 68 12 pN (middle column) and at a constant force of 151 20 pN (right column) (Reprinted and adapted with permission from Suresh et al. [106])

11.2.4 Micropipette Aspiration Micropipette aspiration is a technique used to study the effects of extracellular pressure on individual cells such as erythrocytes. Cells are slowly drawn into the tip of a micropipette. The time-dependent deformations of whole cells can be used to differentiate healthy and diseased cell states as well as cell types [33, 49]. According to the law of Laplace, the applied pressure, DP, can be related to the cortical tension, Tc, inner radius of the pipette, Rp, and the radius of the cell, Rc. 1 1 (11.3) DP ¼ 2Tc Rp Rc The length of extension of the cell into the pipette, Lp, is used to determine whether a cell acts like a solid or liquid drop. The law of Laplace holds until the critical pressure is reached when Lp/Rp ¼ 1. Beyond this pressure, cells like neutrophils that behave like liquid drops flow freely into the pipettes [49]. Micropipette aspiration studies have also been used to study the mechanics of individual organelles such as the cell nucleus [44, 97, 116]. The deformability of the nucleus may have important implications for cell functions (Fig. 11.7).

11.2.5 Atomic Force Microscopy Atomic force microscopy (AFM) or scanning force microscopy is a technique for measuring sub-nanometer resolution topographies and mechanical properties. The AFM technology was invented in 1986 by Calvin Quate and Steve Gerber and was based on the previous scanning tunneling microscope principles [11, 12]. An AFM transducer typically employs a microfabricated silicon cantilever with a sharp tip or

11

Tools for Studying Biomechanical Interactions in Cells

245

Fig. 11.7 The isolated nucleus of a HeLa cell is shown during aspiration undergoing both small (a) and large (b) deformations. Scale bar, 5 mm (Reprinted with permission from Rowat et al. [97])

Fig. 11.8 Typical atomic force microscopy (AFM) measurements are made by positioning a sample such as a cell on a coverslip beneath the tip of the silicon cantilever. As the sample is brought in contact with the tip, the cantilever bends, and its deflection is measured by tracking the laser spot position on the position sensitive photodetector

probe at its end that is used via contact or tapping modes to study substrate topography and mechanical properties (Fig. 11.8). AFM measurements commonly yield maps of surface topography and stiffness. A cantilever functionalized with an antibody can be used to measure the binding affinities of antibody targets on a coverslip or even a cell. For example, using this technique Lee et al. mapped the locations and binding affinities of vascular endothelial growth factor receptor-2 on human microvascular endothelial cells [61].

246

R.E. Taylor et al.

The stiffness of the silicon cantilever is used to convert tip deflection, d, in [m] into force, F, in [N]. For silicon cantilevers the effective bending stiffness or F/d can be calculated using the classical Euler-Bernoulli equation for a fixed-free beam, where stiffness is a function of Young’s Modulus, E [kPa], area moment of inertia, I [m4], and cantilever length, L [m]. F¼

3EI d L3

(11.4)

Further, since cantilever beams have a rectangular section, the area moment of inertia equals wt3/12, where the width, w, and thickness, t, are given in meters. Substituting into the previous equation we obtain the following expression for applied force as a function of tip deflection. F¼

Ewt3 d 4L3

(11.5)

When used to investigate soft substrates such as tissues, spherical beads are often attached to the ends of the cantilevers to enable nondestructive studies similar to nanoindentation [28]. While often used on isolated molecules, AFM can also be used to study local cell properties and has been used to show that fibroblast internal stiffness can be tuned via the substrate’s stiffness [105]. The stiffness of extremely soft substrates like the polyacrylamide gels used in traction force microscopy (TFM) can also be measured using AFM [105]. Further, by functionalizing a microcantilever tip with cell–cell or cell–matrix adhesion molecules, either type of junction can be mimicked and its mechanical properties studied using AFM [10, 61].

11.2.6 Traction Force Microscopy Traction forces of cell migration were first investigated by observing wrinkles on a thin rubber sheet [47]. However, this method lacked the resolution to quantify the force distribution. Galbraith and Sheetz [38] demonstrated one of the first applications of MEMS in cell biology to measure the forces of cell migration. While the section on MEMS tools comes later the chapter, this device properly belongs under traction force tools, because it was one of the first to measure these traction forces. The device consisted of arrays of polysilicon beams with attachment pads at the ends. The traction forces exerted by the migrating cells as they passed over the pads were measured by optically tracking the beam deflections (Fig. 11.9). The force resolution was limited by the optical measurement to around 5 nN for a beam stiffness of 75 nN/mm. Calculation of beam deflection forces involves the same relationship used for AFM measurements (Eq. 4). Observations of fibroblast migration showed forces changed from backward to forward direction between the lamella and the tail of the cell with maximum forces at the tail (+100 nN) an order of magnitude

11

Tools for Studying Biomechanical Interactions in Cells

247

Fig. 11.9 (a–c) Silicon cantilever with binding pads deflects under the forces of cell migration. The forces are calculated from the measured deflection using the calibrated stiffness of the levers. The cantilevers are only sensitive to forces in the normal direction. (d) Measurement noise of the technique limits the force resolution to a few nanoNewtons (Reprinted with permission from Galbraith et al. [38]. Copyright 1997 National Academy of Sciences, USA)

greater than those in the lamellar regions (5 nN). Immunofluorescence imaging of b1-integrin suggested forces were generated at a small number of focal adhesions. One of the limitations of this method was that one had to wait for randomly migrating cells to crawl over the pads in a direction perpendicular to the cantilever for maximum sensitivity. Further, the spatial resolution was limited by the beam spacing, which obscured the subcellular distribution of forces. TFM, as we refer to it today, was developed to improve the spatial and force resolutions of cell traction force measurements [78, 79, 91]. Forces exerted by adherent cells deform polymer substrates such as polyacrylamide gel, gelatin and polydimethylsiloxane (PDMS). These deformations are tracked by observing fluorescent beads embedded in the thin polymer films (Fig. 11.10). The embedded particles allow tracking of strain fields with a spatial resolution of 4–6 mm, enabling measurement of subcellular distributions [7]. This technique has been used to provide evidence supporting the tensegrity models of cytoskeletal force transmission [119]. Trypsin is used after the experiment to detach the cells and relax the substrate to obtain the reference positions of the embedded particles. Time-lapse video microscopy and computational algorithms for determining the deformation field and shear of the substrate opened the way to quantitative measurements of traction forces generated while cells migrate over a surface. The determination of force maps is challenging in practice, and requires the use of computationally intensive algorithms. In early TFM studies of locomoting fibroblasts, researchers observed that the region in the front of a migrating cell plays a leading role, producing the greatest force [78, 79]. The rear regions of migrating fibroblasts exerted a much lower force implying they play a smaller role, possibly to keep the cell anchored (Fig. 11.11). This finding was contradictory to the behavior in locomotion studies using micromachined silicon cantilevers reported by Galbraith and Sheetz who observed greater traction forces concentrated

248

R.E. Taylor et al.

Fig. 11.10 (a) Fluorescent beads embedded in polyacrylamide gel with attached migrating fibroblast (arrow shows direction of migration). (b) Deformation vectors of the substrate. (c) Vector field and (d) color map showing local magnitude of traction stresses. The forces are highest at the leading edge of the migrating cell (Reprinted with permission from Munevar et al. [78, 79])

Fig. 11.11 Traction stresses across the cell front. (a) Immunofluorescence of paxillin (red-bright ellipses at distal edge), serine-19-phosphorylated myosin II light chain marking activated myosin II (blue-bright region from proximal edge to center), and phalloiden staining of F-actin (greenfibers spanning lamella region from proximal side). Locations of lamellipodium, lamellipodium base, and lamella are indicated; distal and proximal directions are defined. (b) GFP-paxillin (inverted contrast) with traction stress vectors superimposed (Video 2, available at http://www. jcb.org/cgi/content/full/jcb.200810060/DC1). (c) Heat-scale plot of traction stress magnitude; segmented FAs indicated by black outlines (Video 2). White lines delineate boundaries between lamellipodium (LP), FAs, and cell body (CB) (Reprinted with permission from Rockefeller University Press [40])

at the rear of a cell [38]. However, different fibroblast types were used in each study, and the device by Galbraith and Sheetz had limited spatial and directional capability, both of which could have contributed to the uncertainty.

11

Tools for Studying Biomechanical Interactions in Cells

249

Improvements to TFM were made by introducing a regular grid of patterns instead of randomly distributed particles in the substrate [4]. Such a design allows for easy measurement of applied forces by observing any deviations from the patterned array. This was achieved either by creating a regular array of features on the PDMS surface or by patterning fluorescent photoresist dots onto the surface. The fluorescent marker method has the advantage that it provides a greater contrast against the background for position measurement, thus yielding better force and displacement resolution. The spatial resolution of this technique is also improved over traditional TFM due to the control over the grid pattern. Such devices were used to measure the stress at a single focal adhesion. This FAC stress was constant at 5.5 nN/mm2 and the force of an average focal adhesion was around 10 nN. Further, the technique found a positive correlation between the area of the focal adhesion and the force over the area. A minimum force of about 1 pN was required for the formation of focal adhesions, which is believed to be the force of a single integrin molecule. TFM can be a powerful tool in elucidating the molecular pathways involved in traction force generation and cell migration. Gaudet et al. showed the influence of collagen concentration on fibroblast migration and found that force generation was dependent on cell area; they estimated the tension in an integrin-collagen bond at 100 pN [41]. Using the mapping ability of TFM, Munevar et al. found the front and rear cell-substrate adhesions play different roles during migration; frontal adhesions provide the major force transmission sites while rear adhesions play a passive role in anchoring the cell [78, 79]. They also observed that 3T3 fibroblasts transfected with the oncogene H-ras had abnormal migratory behavior – including poor directional stability, transient adhesions and increased migratory velocity – behavior associated with metastasis. TFM has been extended to study the effects of various biomolecules on the traction forces: a-smooth muscle actin (a-SMA) in myofibrils [18]; blebistattin, a myosin II inhibitor, on migrating fibroblasts [8]; a6b4 integrin in carcinoma cells [95, 96]; phosphorylation of the heat shock protein, HSP27 on the actin cytoskeleton [1]; role of electric fields in cellular repositioning [19]; and the effect of substrate compliance [43].

11.2.7 Microscale Tools for Mechanobiology Where single-cell, high-throughput platforms for studying cell mechanics are required, devices leveraging integrated circuit (IC) industry fabrication methods have been pursued. The scaling of such devices provides techniques that cover the force (1 nN–1 mN) and distance (1–100 mm) ranges suitable for single cells. Microfabricated platforms allow for controlled cell environments where the spatial arrangement of each cell can be defined through the surface topography [99, 126]. This has lead to over two decades of research on how the feature size, geometry and mechanical properties of a substrate can influence cell growth, apoptosis, organization, migration, and differentiation. A great advantage to cell

250

R.E. Taylor et al.

biologists, microsystems can be created at small scales with batch fabrication methods enabling hundreds or thousands of parallel devices on a single platform the size of a standard Petri dish. Such arrays allow data acquisition from each individual cell in an entire population and result in statistically powerful studies. We next review methods involving microfabricated devices for interacting with cells.

11.2.7.1

Force Posts and Other Beam-Based Polymer Transducers

Discrete arrays of force sensors or force posts based on beam bending are useful for measuring force generation at focal adhesions and for tuning the effective substrate stiffness presented to cells [107]. These detectors employ optically detected polymer cantilevers as force sensors. The cantilevers project from a polymer substrate such as PDMS. Unlike the device by Galbraith and Sheetz where the cantilever acts in the plane of cellular attachments, force posts project the cantilever out of plane. This offers significant advantages over the former technique by providing greater spatial resolution, only limited by the spacing between the posts and optical detector limits. A high force resolution can be easily achieved by varying the height of the polymer posts. Since the posts can be made cylindrical, they can sense forces in any direction in the plane. Significantly, the posts isolate the forces generated at focal adhesions from deformations of the substrate. In the experimental setup, cells lie on top of these beds of “deformable needles”. The adhesion and migration forces of cells cause the posts to deflect (Fig. 11.12). Post stiffness is calculated from classical beam bending theory, which relates the post deflections to the forces acting on them (Fig. 11.12b). The stiffness of the beam depends strongly on the length and the diameter of the cylindrical posts, thus either increasing the length or reducing the diameter would attain greater sensitivity. Post tip positions are measured optically to track the deflections. The discrete nature of the microposts decouples the generated forces at the point of attachment from the rest of

Fig. 11.12 (a) Smooth muscle cell attached to a field of microposts. Tension in the cell bends the posts. (b) Linear relationship between micropost bending and applied force where D is post diameter, L is length, E is the material’s Young’s modulus, X is the displacement, and F is the applied force. (c) Cell stained for visualization of actin (green fibers) and nucleus (blue ellipse highlighted in cell). Actin bundles are seen to terminate on the tops of the microposts (red circles) (Reproduced and adapted with permission from Sniadecki and Chen [103])

11

Tools for Studying Biomechanical Interactions in Cells

251

the substrate, thereby limiting influence from neighboring cells or posts. The posts can be calibrated by measuring their deflection against a known force applied with a glass micropipette. Numerical modeling of the device has also been used to validate the linear stiffness models, since variations from the designed geometry due to fabrication tolerances would greatly affect the stiffness values. Recent studies have also incorporated the effects of substrate deformation and viscoelasticity on beam stiffness to better capture force post deformation behavior [71, 100]. The stiffness of the sensor may be altered to study the effect of substrate compliance on the natural mechanical activity of the cell [32]. Focal adhesions on more compliant microposts are less stable and less defined than those on more rigid ones; stiffer microposts also elicit greater forces in epithelial cells [98]. Forces were found to be linearly proportional to total adhesion area when focal adhesions were greater than 1 mm2, but forces were much greater when focal adhesions were smaller [108]. Force posts have been applied to measure traction and adhesion forces of cells such as fibroblasts, smooth muscle cells and epithelial cells. The discrete nature of posts has enabled correlation of extracellular proteins with force generation by treating the surface of posts with specific proteins and measuring the magnitude of forces. Spatial correlation of proteins and microposts can be achieved through immunohistochemical staining. These studies determine which proteins are implicated in force production, how their expression levels change with level of force production, and how their arrangement influences direction of the forces (Fig. 11.12) [65, 103, 129]. The role of cadherins (cell–cell adhesion proteins) has been studied by attaching an N-cadherin chimera to the top of microposts [39]. Attachment to the microposts through cell-presented cadherins mimics cell–cell contacts. Force transmission through the cadherins was observed through the post deflections. Force transmitted through a single cadherin molecule was estimated at 10 pN. Study of cell–cell contacts in Madin–Darby canine kidney (MDCK) epithelial cells indicated that larger forces were observed in MDCK epithelial sheets than in isolated cells suggesting mechanical cooperation between contacting cells [27]. Force posts have also been used to study cell–cell adhesions. Ladoux et al. used cadherin coated force posts to show that larger traction forces and thus cellular tension are correlated with the formation of cadherin–cadherin contacts [58]. Force posts have been used to isolate and investigate forces at cell–cell junctions. These studies have shown that the size of adherens junctions increases with applied tugging force [72]. Force posts have also been employed to study active contractile cells like cardiomyocytes [128, 131]. Sarcomeric structure in heart cells, or cardiomyocytes, is intimately related to force production, which might be rapidly lost during in vitro culture [3]. Microchannels and large anchoring posts were first shown to maintain neonatal cardiomyocyte morphology and sarcomeric structure [75]. Zhao et al. then incorporated these features into force posts to maintain proper morphology and sarcomeric structure in isolated neonatal cardiomyocytes [129, 131]. Application of isoproterenol caused myocytes to increase their force of contraction, resulting in greater displacement of the microposts, thereby demonstrating force posts’ utility in observation of pharmacologically altered cell function [130].

252

R.E. Taylor et al.

Yet another variant of this technique is to measure forces in 3D culture [62]. This technique employs a 3D culture in a collagen matrix to measure the forces of contraction against a flexible post. The force sensors are fabricated in a fashion similar to those in force post arrays. Here a two layer SU-8 mold is used to create force posts with an anchoring point. The force posts are cast with PDMS and the cells are seeded within a collagen matrix, which is allowed to gel. The contractile forces are observed in about 3 h after seeding through the deflection of the force posts. The force posts are calibrated using a glass micropipette and the deflections of the posts are used to measure the contractile force of the 3D gel. The advantage of this technique is that besides measuring forces in 3D cultures, it can be combined with immunofluorescence imaging to observe the spatial distribution of cytoskeletal and ECM proteins. Immunofluorescence indicates increased development of actin stress fibers in regions of high stress gradients in the culture. Further, the thin tissue slices enable application of cytoskeletal inhibitors to study their effects on the distribution of forces and related proteins. For instance, blebbistatin, a Myosin ATP-ase inhibitor simultaneously causes a reduction in the forces and formation of stress fibers in the cell. This study once again highlights the close coupling between mechanosensation and cytoskeletal organization in the cell. Silicon devices require special design and handling when working with biological samples due to their stiffness mismatch and opacity. Polymer-based cantilevers however, offer platforms that have a compliance more closely matched to cells, and are optically transparent. In another approach to measuring contractile forces (Fig. 11.13), cardiomyocytes were attached to the surface of a PDMS cantilever [85]. As opposed to earlier techniques where the cellular forces were used to bend cantilevers, this apparatus utilizes the surface forces of contractility to produce bending. Key advantages of this method include cooperative studies of multiple cells on surfaces and studies on the role of surface properties. Determination of the forces that result in such deformation regimes is done through extensive image processing and analytical computation or finite element modeling (FEM) of the cantilever [56, 85]. Microgrooves can be created on the surface to control cell

Fig. 11.13 Neonatal rat cardiomyocytes cultured on PDMS cantilevers. Flat and grooved PDMS microcantilevers are shown side by side for comparison. The grooved cantilever shows greater deflection than the flat cantilever because more stress is applied along the long-axis of the cantilever (Reproduced with permission from Park et al. [86])

11

Tools for Studying Biomechanical Interactions in Cells

253

placement and organization on the cantilever (Fig. 11.13). These grooves allow for cells to organize along the long axis of the cantilever so that the majority of the generated force goes into bending the cantilever along its length. Using flat cantilevers, surface stresses generated by single cardiomyocytes were estimated to be 2–5 nN/mm2, which increased to 4–7 nN/mm2 on grooved substrates, again highlighting the role of the substrate [86].

11.2.7.2

MEMS for Applying and Measuring Force

The study of cell stiffness requires both sensing and actuation capabilities in the apparatus to apply known forces and measure the deformation of the cell. The deformations of the cells in response to external forces need to be monitored whereas traction and contractile forces can be extracted from deflection of the cellular environment measured in response to internally generated forces. Due to the constraints of liquid media required for cell culture, the force probes must be extended into the media by an actuator outside the liquid. Yang and Saif utilized a silicon force sensor coupled with a piezoelectric stage for studying the mechanics of single cells under large deformations [123, 124]. This technique was specifically developed to address some of the drawbacks of other methods, which were limited to measuring stiffness at small deformations. Force probes were fabricated from single crystal silicon and calibrated to provide a reliable reference. The force transducers were functionalized with fibronectin to promote adhesion to cells. The external stage was moved to stretch the cell and the deflection of the cantilever tip with respect to the stage was optically observed (Fig. 11.14). The cell force-deformation curves thus obtained were described by a truss model of cytoskeletal stiffness, with microtubule compression struts and actin tensile fibers. The cell deformations were linear and repeatable for large deformations; linearity and reversibility were lost in the presence of cytochalasin-D, an actin polymerization inhibitor [123, 124]. The responses suggested a reversible polymerization mechanism maintains cell stiffness over many cycles. This technique had the specific advantage of measuring large cell deformations, making it useful for understanding cell response to large strain events or injuries. Such measurements can enable an understanding of the physiological processes involved in defining and maintaining the structure of cells. Siechen et al. conducted in vivo experiments on the embryonic Drosophila nervous system using microfabricated two-dimensional force sensors [102]. They studied vesicle clustering at the neuromuscular presynaptic terminal and found that this function is mediated by mechanical tension within the axons. Vesicle clustering was eliminated when the axon was severed from the cell body, but restored when mechanical tension was applied to the severed end of the axon (Fig. 11.15). Clustering also increased when intact axons were stretched by pulling the postsynaptic muscle. The rest tension of embryonic axons with a neuromuscular junction was inferred to be 1 nN. Under mechanical perturbation, the axons also restored the rest tension either by relaxing or by contracting on a timescale of about 15 min.

254

R.E. Taylor et al.

Fig. 11.14 (a) Schematic of cell stiffness measurement with cantilever deflection and cell deformation. The probe is pulled to a distance R by an external actuator, deforming the cell by D. The force is measured from the calibrated spring stiffness and deflection, w. (b) Force-deflection plot of a cell indicating reversible and repeatable deformation. (c–e) Images of measured cell after probe attachment (c), under small deformation (d) and large deformation (e) corresponding to data in (b) (Reprinted with permission from Yang and Saif [123, 124])

11

Tools for Studying Biomechanical Interactions in Cells

255

Fig. 11.15 Rest tension in the axon and its self-regulation. (a) Scanning electron micrograph of a micro mechanical force sensor (spring constant, k ¼ 3.5 nN/mm) used to measure the force response of embryonic Drosophila axons. An x–y–z piezo stage held the sensor and brought the probe into contact with the axon to form nonspecific adhesion and apply stretch. The force, F ¼ kx, on the probe was measured from the deflection, x, of the force sensing beams. The tension, T, in the axon was obtained from the force balance at the point of contact. (b) After embryo dissection, the probe was used to stretch the axon and measure its tension. The interval between the two data points was 50 s. Extrapolation of the force-stretch curve to 0 stretch point gave an estimate of the rest tension of about ~1 nN. (c) A probe pushed an axon at mid length in <1 s, and then held the stretch with time. The corresponding tension in the axon was measured as a function of time. (d) The probe was quickly released from a similarly stretched axon after its tension had relaxed to the rest value. The axon was overstretched as soon as the probe was removed. The time lapse images of the axon show that it shortened its length with time linearly with a velocity of 5 nm/s. The axon recovered its initial length in about 10 min. R2 values of the linear fits in (b) and (d) are shown (Reprinted with permission from Siachen et al. [102]. Copyright 2009 National Academy of Sciences, USA)

Based on results from these microfabricated tools, the authors propose neuromuscular synapses rely on mechanical tension to modulate vesicle accumulation and possibly synaptic plasticity. Piezoresistive cantilevers are another type of active MEMS device useful for biomechanical measures. Piezoresistive microcantilevers are similar to AFM cantilevers except their tip deflection can be measured electrically via the change of resistance of piezoresistors (strain gauges) embedded at the root of the cantilever. Piezoresistive microcantilevers can be used in the same manner as AFM cantilevers, but have two key advantages over AFM that make them well-suited

256

R.E. Taylor et al.

for biological study, due to their non-optical deflection output. First, allowable cantilever deflections can be much larger than the limited range allowed when using optical detection, and second, piezoresistive microcantilevers can be used in harsh, space-constrained environments. Park et al. have used piezoresistive microcantilever based displacement clamps to investigate the body wall mechanics and touch sensitivity of the nematode C. elegans [87, 88, 92]. Local probe assays provide measurements of local stiffness, and the response of cells to point loads can elucidate the role of surface receptors, FAC proteins, and cytoskeletal response to load at a FAC. Other methods are better suited to global assays of cell stiffness as an indicator of organization and response to loading. For example, microplate stretching can measure cytoskeletal properties of whole cells [24, 110]. A single cell is attached between parallel glass plates, where one is flexible while the other is rigid (Fig. 11.16). The plates are stretched by a

Fig. 11.16 A fibroblast is held between two coated microplates. The deformation of the flexible microplate gives the force F acting on the cell. The position of the laser beam emerging from the optical fiber, which is in contact with the tip of the flexible plate, is detected using a position sensitive detector. A personal computer reads the signal from the detector and controls the piezoelectric translator (Reprinted with permission from Fernandez et al. [35])

11

Tools for Studying Biomechanical Interactions in Cells

257

piezoelectric stage, applying the deformation to the cell. The force applied is inferred through the displacement of the flexible plate, which is precalibrated to measure forces. The key advantage of this technique is that it measures the stiffness properties of the entire cell. This setup enables combined measurement of the cortical stiffness of the cytoskeleton and the viscous properties of the cytoplasm [34, 35]. These data are obtained by measuring cell response at different timescales of stimulation. The technique allows for controlling strain or force across the cell sample over long periods of time, allowing the properties to be probed over large time scales. Recent advances have enabled the operation of electrostatically actuated microdevices in ionic media with ionic strengths up to 150 mM/L [76, 77]. These advances enable temporal force-displacement control for the investigation of single cell viscoelastic properties. Researchers demonstrated that MDCK cells that had spread across collagen-coated gold stretching pads could be dynamically stretched and their viscoelastic properties measured (Fig. 11.17). A key advantage of this technology is that cells can be imaged throughout testing; the device and cells can be viewed in plan view on an upright microscope stage. Such devices can be used to study more than whole cell mechanical properties; it could potentially be functionalized to mimic both cell–cell and cell–matrix junctions. Another microfabricated tensile tester of interest is the MEMS single cell measurement system developed by Panchawagh et al. [84, 101]. The viscoelastic properties of single fibroblasts were studied using this platform. Cells spread across the transparent cell platforms and thereafter a piezoelectric stage was used to pull on the cell. Pulling forces were measured using sensor beam deflections.

Fig. 11.17 This microfabricated device can be operated in ionic media to both apply and sense forces from single cells in a tensile test configuration for adherent cells. Electrostatic comb drives are used for both sensing and actuation (left). MDCK cells spread across collagen coated gold pads can be stretched (right) and their viscoelastic properties quantified with this testing device (Reprinted with permission from Mukundan and Pruitt [76])

258

R.E. Taylor et al.

These microfabricated tools allow scientists unprecedented access to time varying cellular properties such as adhesion and cytoskeletal remodeling. Using these tools scientists can investigate heretofore untenable questions in cellular biomechanics.

11.3

Summary Case Study: Fibroblast Mechanobiology

As we have shown, the set of tools for studying cell–cell and cell–matrix interactions is ever increasing, with numerous examples of microfabricated sensors and actuators designed with specificity to both cell type and function. As a summary of the breadth of tools available and advances in our understanding of cellular biomechanics, we offer a case study of recent work on fibroblasts. For biologists interested in cell and tissue mechanics, the fibroblast is an ideal subject of study. Fibroblasts are not only one of the easiest cell types to culture, but they also play a critical role in tissue maintenance and structure in nearly every tissue and organ throughout the body. The main function of a fibroblast cell is to excrete collagenous ECM proteins. Further, they are extremely versatile connective-tissue cells that can differentiate into numerous different cell types based on chemical and physical cues. Biomechanists looking to study cell–matrix and cell–cell interactions have naturally studied fibroblasts extensively (Fig. 11.18). Recent advances have been made using traditional and microfabricated tools such as AFM, TFM, and silicon. MEMS testing devices have been covered in this chapter. These advances offer insights into how cells dynamically interact with their environments and one another.

Fig. 11.18 The fibroblast is an extremely versatile connective tissue cell that plays a critical role in tissue maintenance. It is highly motile and is sensitive to environmental mechanical properties. Typical fibroblast morphology is shown here with immunostained actin filaments (left-red channel), myosin heavy chain IIA (middle-green channel), and nuclear DNA (rightmerged with DAPI blue channel, brightest in central region of nucleus). Scale bars, 20 mm (Reprinted with permission from PLoS One article by Christopher M. Hale and Denis Wirtz [46])

11 l

l

l

l

Tools for Studying Biomechanical Interactions in Cells

259

MEMS devices such as the micromachined cantilevers by Galbraith and Sheetz enabled the first MEMS-based measurement of force production in fibroblasts, and results from these studies suggest that different types of fibroblasts may generate different patterns of force during migration [38]. Thoumin and Ott utilized MEMS microplates to manipulate chick fibroblasts and investigate the contributions of passive viscoelasticity and active traction to cytoskeletal mechanical properties [110]. In addition to measuring elastic moduli and viscosity, they identified a dominant elastic response as well as viscous force relaxation behavior and a longer term contractile behavior in response to axial load. TFM [133] of fibroblasts on bead-embedded polyacrylamide gels has shown that fibroblasts exhibit durotaxis or stiffness-gradient-driven migration and align perpendicular to the direction of substrate stretch. TFM has been used to study the effects of collagen concentration [41] and the myosin II inhibitor blebbistatin on fibroblast motility, spreading, and contractility [8]. Studies with TFM have also provided local force generation information and have indicated that force production is maximal at the front edge of migrating fibroblasts [78, 79]. Force posts with embedded magnetic wires have been used to directly apply forces to NIH 3T3 fibroblast cells. These studies have shown that focal adhesion size increases with local application of force, and that multiple applications of force increase focal adhesion size more than single applications [104]. AFM has been used to show that fibroblasts tune their internal stiffness based on the stiffness of their substrate [105]. This mechanism may allow environmental rigidity to drive fibroblast migration and direct wound repair.

These tools enable scientists to study the effect of the mechanical environment on cell function, migration and differentiation, thereby answering critical questions in mechanobiology. Acknowledgements The authors acknowledge support from the National Science Foundation (EFRI-CBE 073555, CAREER ECS-0449400), the National Institutes of Health (R21 HL089027, R01 EB006745), the California Institute for Regenerative Medicine (CIRM RC1-00151-1), Stanford Center for Integrated Systems, and Stanford University (Bio-X Graduate Fellowship, Stanford Graduate Fellowship and a Stanford DARE Doctoral Fellowship).

References 1. An, SS, CM Pennella, et al. (2005). Hypoxia alters biophysical properties of endothelial cells via p38 MAPK- and Rho kinase-dependent pathways. Am J Physiol Cell Physiol 289(3): C521–C530. 2. Ashkin, A, DJ M., et al. (1986). Observation of a single-beam gradient force optical trap for dielectric particles. Opt Lett 11(5): 288. 3. Atherton, BT, DM Meyer, et al. (1986). Assembly and remodelling of myofibrils and intercalated discs in cultured neonatal rat heart cells. J Cell Sci 86: 233–248. 4. Balaban, NQ, US Schwarz, et al. (2001). Force and focal adhesion assembly: a close relationship studied using elastic micropatterned substrates. Nat Cell Biol 3(5): 466–472.

260

R.E. Taylor et al.

5. Bao, G and S Suresh (2003). Cell and molecular mechanics of biological materials. Nat Mater 2: 715–725. 6. Baumgartner, W, GJ Schutz, et al. (2003). Cadherin function probed by laser tweezer and single molecule fluorescence in vascular endothelial cells. J Cell Sci 116(6): 1001–1011. 7. Beningo, KA, M Dembo, et al. (2001). Nascent focal adhesions are responsible for the generation of strong propulsive forces in migrating fibroblasts. J Cell Biol 153(4): 881–888. 8. Beningo, KA, K Hamao, et al. (2006). Traction forces of fibroblasts are regulated by the Rhodependent kinase but not by the myosin light chain kinase. Arch Biochem Biophys 456(2): 224–231. 9. Benjamin, JM, AV Kwiatkowski, et al. (2010). Alpha E-catenin regulates actin dynamics independently of cadherin-mediated cell-cell adhesion. J Cell Biol 189(2): 339–352. 10. Benoit, M, D Gabriel, et al. (2000). Discrete interactions in cell adhesion measured by singlemolecule force spectroscopy. Nat Cell Biol 2(6): 313–317. 11. Binnig, G, H Rohrer, et al. (1982). Surface Studies by Scanning Tunneling Microscopy. Phys Rev Lett 49(1): 57. 12. Binnig, G, CF Quate, et al. (1986). Atomic force microscope. Phys Rev Lett 56(9): 930. 13. Borghi, N, M Lowndes, et al. (2010). Regulation of cell motile behavior by crosstalk between cadherin- and integrin-mediated adhesions. PNAS 107(30): 13324–13329. 14. Brady, AJ (1991). Mechanical properties of isolated cardiac myocytes. Physiol Rev 71(2): 413–428. 15. Brandt, PW, F Colomo, et al. (1998). Force regulation by Ca2+ in skinned single cardiac myocytes of frog. Biophys J 74(4): 1994–2004. 16. Brown, TD (2000). Techniques for mechanical stimulation of cells in vitro: a review. J Biomech 33(1): 3–14. 17. Chen, J, B Fabry, et al. (2001). Twisting integrin receptors increases endothelin-1 gene expression in endothelial cells. Am J Physiol Cell Physiol 280(6): C1475–C1484. 18. Chen, J, H Li, et al. (2007). Alpha-smooth muscle actin expression enhances cell traction force. Cell Motil Cytoskeleton 64(4): 248–257. 19. Curtze, S, M Dembo, et al. (2004). Dynamic changes in traction forces with DC electric field in osteoblast-like cells. J Cell Sci 117(13): 2721–2729. 20. Dao, M, CT Lim, et al. (2003). Mechanics of the human red blood cell deformed by optical tweezers. J Mech Phys Solids 51(11–12): 2259–2280. 21. del Rio, A, R Perez-Jimenez, et al. (2009). Stretching single talin rod molecules activates vinculin binding. Science 323(5914): 638–641. 22. Deshpande, VS, RM McMeeking, et al. (2006). A bio-chemo-mechanical model for cell contractility. Proc Natl Acad Sci USA 103(38): 14015–14020. 23. Desmouliere, A, C Chaponnier, et al. (2005). Tissue repair, contraction, and the myofibroblast. Wound Repair Regen 13(1): 7–12. 24. Desprat, N, A Richert, et al. (2005). Creep Function of a Single Living Cell. Biophys J 88(3): 2224–2233. 25. Dike, LE, CS Chen, et al. (1999). Geometric control of switching between growth, apoptosis, and differentiation during angiogenesis using micropatterned substrates. In Vitro Cell Dev Biol Anim 35(8): 441–448. 26. Discher, DE, P Janmey, et al. (2005). Tissue Cells Feel and Respond to the Stiffness of Their Substrate. Science 310(5751): 1139–1143. 27. du Roure, O, A Saez, et al. (2005). Force mapping in epithelial cell migration. Proc Natl Acad Sci USA 102(7): 2390–2395. 28. Ebenstein, DM and LA Pruitt (2006). Nanoindentation of biological materials. Nano Today 1(3): 26–33. 29. Eckes, B, MC Zweers, et al. (2006). Mechanical tension and integrin alpha 2 beta 1 regulate fibroblast functions. J Investig Dermatol Symp Proc 11(1): 66–72. 30. Ehler, E and JC Perriard (2000). Cardiomyocyte cytoskeleton and myofibrillogenesis in healthy and diseased heart. Heart Fail Rev 5(3): 259–269.

11

Tools for Studying Biomechanical Interactions in Cells

261

31. Ehrlich, JS, MDH Hansen, et al. (2002). Spatio-temporal regulation of Rac1 localization and lamellipodia dynamics during epithelial cell-cell adhesion. Dev Cell 3(2): 259–270. 32. Engler, AJ, S Sen, et al. (2006). Matrix elasticity directs stem cell lineage specification. Cell 126(4): 677–689. 33. Evans, E and A Yeung (1989). Apparent viscosity and cortical tension of blood granulocytes determined by micropipet aspiration. Biophys J 56(1): 151–160. 34. Fernandez, P and A Ott (2008). Single cell mechanics: stress stiffening and kinematic hardening. Phys Rev Lett 100(23): 238102. 35. Fernandez, P, PA Pullarkat, et al. (2006). A master relation defines the nonlinear viscoelasticity of single fibroblasts. Biophys J 90(10): 3796–3805. 36. Finer, JT, RM Simmons, et al. (1994). Single myosin molecule mechanics: piconewton forces and nanometre steps. Nature 368(6467): 113–119. 37. Frisen, M, M Magi, et al. (1969). Rheological analysis of soft collagenous tissue. Part II: experimental evaluations and verifications. J Biomech 2(1): 21–28. 38. Galbraith, CG and MP Sheetz (1997). A micromachined device provides a new bend on fibroblast traction forces. Proc Natl Acad Sci USA 94(17): 9114–9118. 39. Ganz, A, M Lambert, et al. (2006). Traction forces exerted through N-cadherin contacts. Biol Cell 98(12): 721–730. 40. Gardel, ML, B Sabass, et al. (2008). Traction stress in focal adhesions correlates biphasically with actin retrograde flow speed. J Cell Biol 183(6): 999–1005. 41. Gaudet, C, WA Marganski, et al. (2003). Influence of type I collagen surface density on fibroblast spreading, motility, and contractility. Biophys J 85(5): 3329–3335. 42. Geiger, B, JP Spatz, et al. (2009). Environmental sensing through focal adhesions. Nat Rev Mol Cell Biol 10(1): 21–33. 43. Ghosh, K, Z Pan, et al. (2007). Cell adaptation to a physiologically relevant ECM mimic with different viscoelastic properties. Biomaterials 28(4): 671–679. 44. Guilak, F, JR Tedrow, et al. (2000). Viscoelastic Properties of the Cell Nucleus. Biochem Biophys Res Commun 269(3): 781–786. 45. Hahn, C and MA Schwartz (2009). Mechanotransduction in vascular physiology and atherogenesis. Nat Rev Mol Cell Biol 10(1): 53–62. 46. Hale, CM, SX Sun, et al. (2009). Resolving the role of actoymyosin contractility in cell microrheology. PLoS One 4(9): e7054. 47. Harris, AK, P Wild, et al. (1980). Silicone rubber substrata: a new wrinkle in the study of cell locomotion. Science 208(4440): 177–179. 48. Hinz, B, SH Phan, et al. (2007). The myofibroblast: one function, multiple origins. Am J Pathol 170(6): 1807–1816. 49. Hochmuth, RM (2000). Micropipette aspiration of living cells. J Biomech 33(1): 15–22. 50. Huang, W, B Anvari, et al. (2003). Temporal effects of cell adhesion on mechanical characteristics of the single chondrocyte. J Orthopaed Res 21(1): 88–95. 51. Ingber, DE (2003). Tensegrity I. Cell structure and hierarchical systems biology. J Cell Sci 116(Pt 7): 1157–1173. 52. Ingber, DE (2003). Tensegrity II. How structural networks influence cellular information processing networks. J Cell Sci 116(Pt 8): 1397–1408. 53. Kempson, GE, H Muir, et al. (1973). The tensile properties of the cartilage of human femoral condyles related to the content of collagen and glycosaminoglycans. Biochim Biophys Acta 297(2): 456–472. 54. Kessler, D, S Dethlefsen, et al. (2001). Fibroblasts in mechanically stressed collagen lattices assume a "synthetic" phenotype. J Biol Chem 276(39): 36575–36585. 55. Khatiwala, CB, SR Peyton, et al. (2006). Intrinsic mechanical properties of the extracellular matrix affect the behavior of pre-osteoblastic MC3T3-E1 cells. Am J Physiol Cell Physiol 290(6): C1640–C1650. 56. Kim, J, J Park, et al. (2006). Realistic computational modeling for hybrid biopolymer microcantilevers. Conf Proc IEEE Eng Med Biol Soc 1: 2102–2105.

262

R.E. Taylor et al.

57. Ko, KS, PD Arora, et al. (2001). Cadherins mediate intercellular mechanical signaling in fibroblasts by activation of stretch-sensitive calcium-permeable channels. J Biol Chem 276: 35967–35977. 58. Ladoux, B, E Anon, et al. (2010). Strength dependence of cadherin-mediated adhesions. Biophys J 98(4): 534–542. 59. Langlois, ED, GA Shaw, et al. (2007). Spring constant calibration of atomic force microscopy cantilevers with a piezosensor transfer standard. Rev Sci Instrum 78(9): 093705–093710. 60. le Duc, Q, Q Shi, et al. (2010). Vinculin potentiates E-cadherin mechanosensing and is ¨ `ıdependent recruited to actin-anchored sites within adherens junctions in a myosin II‚A manner. J Cell Biol 189(7): 1107–1115. 61. Lee, S, J Mandic, et al. (2007). Chemomechanical mapping of ligand-receptor binding kinetics on cells. Proc Natl Acad Sci USA 104(23): 9609–9614. 62. Legant, WR, A Pathak, et al. (2009). Microfabricated tissue gauges to measure and manipulate forces from 3D microtissues. Proc Natl Acad Sci USA 106(25): 10097–10102. 63. Lehoux, S, B Esposito, et al. (2005). Differential regulation of vascular focal adhesion kinase by steady stretch and pulsatility. Circulation 111(5): 643–649. 64. Lele, TP, JE Sero, et al. (2007). Tools to study cell mechanics and mechanotransduction. Methods in Cell Biology. W. YuLi and E. D. Dennis, Academic. Vol. 83: 441, 443–472. 65. Lemmon, CA, NJ Sniadecki, et al. (2005). Shear force at the cell-matrix interface: enhanced analysis for microfabricated post array detectors. Mech Chem Biosyst 2(1): 1–16. 66. Levental, KR, H Yu, et al. (2009). Matrix crosslinking forces tumor progression by enhancing integrin signaling. Cell 139(5): 891–906. 67. LeWinter, MM (2005). Functional consequences of sarcomeric protein abnormalities in failing myocardium. Heart Fail Rev 10(3): 249–257. 68. Lin, G, KSJ Pister, et al. (1995). microscale force-transducer system to quantify isolated heart cell contractile characteristics. Sens Actuators A Phys 46(1–3): 233–236. 69. Lin, G, KSJ Pister, et al. (1995). Novel microelectromechanical system force transducer to quantify contractile characteristics from isolated cardiac-muscle-cells. J Electrochem Soc 142(3): L31-L33. 70. Lin, G, KSJ Pister, et al. (2000). Micromachined polysilicon heart cell force transducer. J Microelectromech S 9(1): 9–17. 71. Lin, IK, Y-M Liao, et al. (2008). Viscoelastic mechanical behavior of soft microcantileverbased force sensors. App Phys Lett 93(25): 251907–251903. 72. Liu, Z, JL Tan, et al. (2010). Mechanical tugging force regulates the size of cell-cell junctions. Proc Natl Acad Sci USA 107(22): 9944–9949. 73. Lodish, H, A Berk, et al. (2007). Molecular Cell Biology (Lodish, Molecular Cell Biology), W. H. Freeman. 74. Makarenko, I, CA Opitz, et al. (2004). Passive stiffness changes caused by upregulation of compliant titin isoforms in human dilated cardiomyopathy hearts. Circ Res 95(7): 708–716. 75. Motlagh, D, TJ Hartman, et al. (2003). Microfabricated grooves recapitulate neonatal myocyte connexin43 and N-cadherin expression and localization. J Biomed Mater Res A 67(1): 148–157. 76. Mukundan, V and BL Pruitt (2009). MEMS Electrostatic Actuation in Conducting Biological Media. J Microelectromech Systems 18(2): 405–413. 77. Mukundan, V, P Ponce, et al. (2009). Modeling and characterization of electrostatic combdrive actuators in conducting liquid media. J Micromech Microeng 19(6): 065008. 78. Munevar, S, Y Wang, et al. (2001). Traction force microscopy of migrating normal and H-ras transformed 3 T3 fibroblasts. Biophys J 80(4): 1744–1757. 79. Munevar, S, YL Wang, et al. (2001). Distinct roles of frontal and rear cell-substrate adhesions in fibroblast migration. Mol Biol Cell 12(12): 3947–3954. 80. Nicolas, A, B Geiger, et al. (2004). Cell mechanosensitivity controls the anisotropy of focal adhesions. Proc Natl Acad Sci USA 101(34): 12520–12525.

11

Tools for Studying Biomechanical Interactions in Cells

263

81. Nishimura, T and M Takeichi (2009). Remodeling of the adherens junctions during morphogenesis. Current Topics in Developmental Biology. L. Thomas, Academic Press. Volume 89: 33–54. 82. Nishimura, S, S-I Yasuda, et al. (2004). Single cell mechanics of rat cardiomyocytes under isometric, unloaded, and physiologically loaded conditions. Am J Physiol Heart Circ Physiol 287(1): H196–H202. 83. Oster, GF, JD Murray, et al. (1983). Mechanical aspects of mesenchymal morphogenesis. J Embryol Exp Morphol 78: 83–125. 84. Panchawagh, HV, D Serrell, et al. (2005). Design and characterization of a BioMEMS device for in-vitro mechanical stimulation of single adherent cells. ASME Conf Proc 2005(4224X): 19–25. 85. Park, J, J Ryu, et al. (2005). Real-time measurement of the contractile forces of selforganized cardiomyocytes on hybrid biopolymer microcantilevers. Anal Chem 77(20): 6571–6580. 86. Park, J, J Kim, et al. (2006). Fabrication of complex 3D polymer structures for cell-polymer hybrid systems. J Micromech Microeng 16: 1614–1619. 87. Park, SJ, MB Goodman, et al. (2007). Analysis of nematode mechanics by piezoresistive displacement clamp. Proc Natl Acad Sci USA. 88. Park, S-J, Petzold B, et al. (2009). Piezoresistive cantilever-based force clamp system for the study of mechanotransduction in C. elegans. Proc MEMS, Sorrento, Italy. 89. Parodi, EN, GA Kaiser, et al. (1972). Comparative study of the tensile strength of autogenous systemic veins and preserved venous homografts. J Surg Res 12(2): 99–104. 90. Paszek, MJ and VM Weaver (2004). The tension mounts: mechanics meets morphogenesis and malignancy. J Mammary Gland Biol Neoplasia 9(4): 325–342. 91. Pelham, RJ, Jr. and Y Wang (1999). High resolution detection of mechanical forces exerted by locomoting fibroblasts on the substrate. Mol Biol Cell 10(4): 935–945. 92. Petzold, BC, S-J Park, et al. (2009). The contribution of body wall muscles to C. elegans body mechanics determined using piezoresistive microcantilevers. Proc MicroTAS, Jeju, South Korea. 93. Peyton, S, C Ghajar, et al. (2007). The emergence of ECM mechanics and cytoskeletal tension as important regulators of cell function. Cell Biochem Biophys 47: 300–320. 94. Pratt, JR, DT Smith, et al. (2004). Progress toward Syste´me International d’Unite´s traceable force metrology for nanmechanics. J Mat Res 19(1): 366–379. 95. Rabinovitz, I and AM Mercurio (1996). The integrin alpha 6 beta 4 and the biology of carcinoma. Biochem Cell Biol 74(6): 811–821. 96. Rabinovitz, I, IK Gipson, et al. (2001). Traction forces mediated by alpha6beta4 integrin: implications for basement membrane organization and tumor invasion. Mol Biol Cell 12(12): 4030–4043. 97. Rowat, AC, J Lammerding, et al. (2006). Mechanical properties of the cell nucleus and the effect of emerin deficiency. Biophys J 91(12): 4649–4664. 98. Saez, A, M Ghibaudo, et al. (2007). Rigidity-driven growth and migration of epithelial cells on microstructured anisotropic substrates. Proc Natl Acad Sci USA 104(20): 8281–8286. 99. Sathuluri, RR, S Yamamura, et al. (2008). Microsystems technology and biosensing. Adv Biochem Eng Biotechnol 109: 285–350. 100. Schoen, I, W Hu, et al. (2010). Probing cellular traction forces by micropillar arrays: contribution of substrate warping to pillar deflection. Nano Lett 10(5): 1823–1830. 101. Serrell, D, J Law, et al. (2008). A uniaxial bioMEMS device for imaging single cell response during quantitative force-displacement measurements. Biomed Microdevices 10(6): 883–889. 102. Siechen, S, S Yang, et al. (2009). Mechanical tension contributes to clustering of neurotransmitter vesicles at presynaptic terminals. Proc Natl Acad Sci USA 106(31): 12611–12616. 103. Sniadecki, NJ and CS Chen (2007). Microfabricated silicone elastomeric post arrays for measuring traction forces of adherent cells. Methods Cell Biol 83: 313–328.

264

R.E. Taylor et al.

104. Sniadecki, NJ, A Anguelouch, et al. (2007). Magnetic microposts as an approach to apply forces to living cells. Proc Natl Acad Sci USA 104(37): 14553–14558. 105. Solon, J, I Levental, et al. (2007). Fibroblast Adaptation and Stiffness Matching to Soft Elastic Substrates. Biophys J 93(12): 4453–4461. 106. Suresh, S, J Spatz, et al. (2005). Connections between single-cell biomechanics and human disease states: gastrointestinal cancer and malaria. Acta Biomater 1(1): 15–30. 107. Tan, JL, J Tien, et al. (2003). Cells lying on a bed of microneedles: an approach to isolate mechanical force. Proc Natl Acad Sci USA 100(4): 1484–1489. 108. Tan, JL, W Liu, et al. (2004). Simple approach to micropattern cells on common culture substrates by tuning substrate wettability. Tissue Eng 10(5–6):865–872. 109. Tasche, C, E Meyhofer, et al. (1999). A force transducer for measuring mechanical properties of single cardiac myocytes. Am J Physiol 277(6 Pt 2): H2400–H2408. 110. Thoumine, O and A Ott (1997). Time scale dependent viscoelastic and contractile regimes in fibroblasts probed by microplate manipulation. J Cell Sci 110(17): 2109–2116. 111. Thoumine, O, P Kocian, et al. (2000). Short-term binding of fibroblasts to fibronectin: optical tweezers experiments and probabilistic analysis. Eur Biophys J 29(6): 398–408. 112. Thoumine, O, M Lambert, et al. (2006). Regulation of N-Cadherin dynamics at neuronal contacts by ligand binding and cytoskeletal coupling. Molecular Biology of the Cell 17(2): 862–875. 113. Timoshenko, SP and JM Gere (1963). Theory of Elastic Stability, McGraw-Hill. 114. Van Vliet, KJ, G Bao, et al. (2003). The biomechanics toolbox: experimental approaches for living cells and biomolecules. Acta Mater 51(19): 5881–5905. 115. Vannier, C, H Chevassus, et al. (1996). Ca-dependence of isometric force kinetics in single skinned ventricular cardiomyocytes from rats. Cardiovasc Res 32(3): 580–586. 116. Vaziri, A and MRK Mofrad (2007). Mechanics and deformation of the nucleus in micropipette aspiration experiment. J Biomech 40(9): 2053–2062. 117. Vogel, V and M Sheetz (2006). Local force and geometry sensing regulate cell functions. Nat Rev Mol Cell Biol 7(4): 265–275. 118. Voronov, DA, PW Alford, et al. (2004). The role of mechanical forces in dextral rotation during cardiac looping in the chick embryo. Dev Biol 272(2): 339–350. 119. Wang, N, K Naruse, et al. (2001). Mechanical behavior in living cells consistent with the tensegrity model. Proc Natl Acad Sci USA 98(14): 7765–7770. 120. Wehrle-Haller, B and BA Imhof (2002). The inner lives of focal adhesions. Trends Cell Biol 12(8): 382–389. 121. Wozniak, MA, K Modzelewska, et al. (2004). Focal adhesion regulation of cell behavior. Biochim Biophy Acta Mol Cell Res 1692(2–3): 103–119. 122. Yamada, S and WJ Nelson (2007). Localized zones of Rho and Rac activities drive initiation and expansion of epithelial cell-cell adhesion. J Cell Biol 178(3): 517–527. 123. Yang, S and T Saif (2005). Micromachined force sensors for the study of cell mechanics. Rev Sci Instrum 76(4): 044301–044308. 124. Yang, S and T Saif (2005). Reversible and repeatable linear local cell force response under large stretches. Exp Cell Res 305(1): 42–50. 125. Yonemura, S, Y Wada, et al. (2010). [alpha]-Catenin as a tension transducer that induces adherens junction development. Nat Cell Biol 12(6): 533–542. 126. Zaouk, R, BY Park, et al. (2006). Introduction to microfabrication techniques. Methods Mol Biol 321: 5–15. 127. Zemel, A, F Rehfeldt, et al. (2010). Optimal matrix rigidity for stress-fibre polarization in stem cells. Nat Phys 6(6): 468–473. 128. Zhao, Y and X Zhang (2006). Cellular mechanics study in cardiac myocytes using PDMS pillar array. Sensor Actuat A-Phys 125: 398–404. 129. Zhao, Y, CC Lim, et al. (2005). Cellular force measurements using single-spaced polymeric microstructures: isolating cells from base substrate. J Micromech Microeng 15: 1649–1656.

11

Tools for Studying Biomechanical Interactions in Cells

265

130. Zhao, Y, CC Lim, et al. (2006). Microchip for subcellular mechanics study in living cells. Sensor Actuat B-Chem 114: 1108–1115. 131. Zhao, Y, CC Lim, et al. (2007). Simultaneous orientation and cellular force measurements in adult cardiac myocytes using three-dimensional polymeric microstructures. Cell Motil Cytoskel 64(9): 718–725. 132. Zhu, C, G Bao, et al. (2000). Cell mechanics: mechanical response, cell adhesion and molecular deformation. Annu Rev Biomed Eng 2: 189–226. 133. Wang, H-B and M Dembo (2000). Substrate flexibility regulates growth and apoptosis of normal but not transformed cells. Am J Physiol - Cell Ph 279(5): C1345–C1350.

Chapter 12

Biomaterials for Studies in Cellular Mechanotransduction Ross DeVolder and Hyunjoon Kong

This chapter is part of Section IV: Tools for Exploring Mechanobiology

Abstract The conversion of mechanical stimuli into chemical signals is of the utmost importance for developmental and normal physiology. Mechanotransduction plays a pivotal role in regulating cellular function and, subsequent tissue maintenance and repair, apoptosis, and many other physiological functions, coupled with a broad array of soluble factors. The successful examination of how mechanotransduction effects cells’ function, in vitro, requires the ability to develop cell culture platforms that recapitulate extracellular environments in which the cells reside. Recently, significant progress in biomaterial design has allowed the examination of the effects mechanotransduction plays on a broad array of extracellular microenvironments. This chapter will review a series of biomaterials used for mechanotransduction studies specifically focusing on glass substrates, poly (dimethyl siloxane) (PDMS) and polymeric hydrogels, and also discuss strategies for designing advanced biomaterial systems.

12.1

Introduction

The conversion of mechanical stimuli into chemical signals is of the utmost importance for developmental and normal physiology [1]. Mechanotransduction plays a pivotal role in regulating cellular function and, subsequent tissue maintenance and repair, apoptosis, and many other physiological functions, coupled with a broad array of soluble factors [1–3]. For example, in cardiovascular tissue, the repeated contraction and relaxation of the heart affects the morphology and functionality of cardiac muscle, and shear stresses also influences the organization and permeability of the endothelial lining. The application of forces, whether from gravity or human activity, on bone stimulates remodeling to maintain optimal mechanical performance.

H. Kong (*) Department of Chemical and Biomolecular Engineering, University of Illinois at Urbana Champaign, Urbana, IL 61801, USA e-mail: [email protected]

A. Wagoner Johnson and Brendan A.C. Harley (eds.), Mechanobiology of Cell-Cell and Cell-Matrix Interactions, DOI 10.1007/978-1-4419-8083-0_12, # Springer Science+Business Media, LLC 2011

267

268

R. DeVolder and H. Kong

Fig. 12.1 Transduction of mechanical factors into physiological signals through biomechanical interactions of cytoskeleton, focal adhesion complexes, and extracellular matrix [6]

The successful examination of how mechanotransduction effects cells’ function, in vitro, requires the ability to develop cell culture platforms that recapitulate extracellular environments in which the cells reside. Extensive studies in matrix biology have revealed that a variety of ECM (extracellular matrix) variables act as insoluble signals that regulate cellular phenotypes [4, 5]. Furthermore, these ECM variables are known to mediate the effects of external mechanical stimuli on cellular activities [6] (Fig. 12.1). The properties and structure of ECMs are known to depend significantly on the tissue type, age, injury, and disease [4, 5, 7]. Therefore, it is critical to study the effects of mechanotransduction using cell adhesion matrices designed to present in vivo cellular microenvironments in a precise and controllable manner. Recently, significant progress in biomaterial design has allowed the examination of the effects mechanotransduction plays on a broad array of extracellular microenvironments. This chapter will review a series of biomaterials used for mechanotransduction studies specifically focusing on glass substrates, poly(dimethyl siloxane) (PDMS) and polymeric hydrogels, and also discuss strategies for designing advanced biomaterial systems.

12.2

Glass Substrates

Glass substrates are widely used for mechanotransduction studies because of their practicality and commercial availability for cell cultures. These substrates provide a 2D cell culture environment, which is easily examined through optical, fluorescence,

12

Biomaterials for Studies in Cellular Mechanotransduction

269

and mechanical microscopic techniques. Mechanotransduction studies involving the application of forces directly to cells, such as examining integrin-ligand binding forces using atomic force microscopy or cellular activities using Magnetic Twisting Cytometry, have extensively used glass substrates [8, 9]. Cells can not form focal adhesion complexes with glass or plastic substrates alone, because of their hydrophobic properties. In order to form these complexes and ultimately culture cells on glass substrates, specific proteins, such as collagen, fibronectin or laminin, are adsorbed on substrate’s surfaces allowing cell binding. Glass surfaces are often treated with gas plasma to reduce hyrophobicity and create rough surfaces that facilitate protein adsorption [10]. The type and degree of glass surface treatments is extremely important since the adsorption properties and conformation of adsorbed proteins is substrate-dependent, and because these variables can modulate cellular activities [11]. Alternatively, glass substrates can be modified to present chemically reactive groups, in order to covalently link binding peptides. The use of silane reagents is a common method for covalently cross-linking cell adhesion domains to glass substrates, through the formation of silanol bonds. This method, when combined with simple lithographic techniques, allows specific patterning of binding domains for controlled cell distributions [12]. Additional methods for cross-linking adhesive molecules include laser vapor deposition and deep Ultra Violet irradiation, which include the activation of substrates and bioactive molecules for chemical linking [10]. Even though glass is widely used as a substrate in mechanotransduction studies, because of their ease for culturing cells, they have several disadvantages. First, these substrates provide a 2-dimensional platform for cells, which is not indicative of a cell’s 3D native environment. Several studies have demonstrated that cellular activities are significantly different between cells cultured in 2D vs. 3D environments. Second, the mechanical properties of glass and plastic are dissimilar to normal tissues. The elastic moduli of glass substrates is several orders of magnitude larger than normal tissues, which can result in altered cellular signaling compared to lower moduli environments. Additionally, these substrates are non-elastic making it not applicable for their use in several mechanotransduction studies such as traction force microscopy.

12.3

Polydimethylsiloxane (PDMS)

PDMS is widely used in mechanotransduction studies because of its elastomeric properties, optical transparency, and processing capabilities. It is a viscoelastic liquid that upon cross-linking, through addition of a curing agent, forms a solid elastic structure that is extremely hydrophobic. Its elastic properties are advantageous for mechanotransduction studies because forces are easily transferred, either from the cells to the substrate or vice versa, and are easily examined through various techniques such as traction force microscopy [13]. For instance, Harris et al. were

270

R. DeVolder and H. Kong

some of the first to examine the forces exerted by cells on their environment using vulcanized PDMS [14]. Additionally, PDMS is commonly used in soft lithographic techniques, which create microfluidic devices and bioreactors widely used for cellular studies involving applications of various forces [10, 15, 16 ]. Similar to glass substrates, cells cannot adhere directly to the hydrophobic surfaces of PDMS and require protein adsorption or surface modification. Numerous types of ECM and serum proteins are physically adsorbed to PDMS substrates to allow cell binding [17]. Additional surface modifications of PDMS, for improved cellular adhesion, typically involve treating them with a variety of chemical reagents [10, 18]. One of the more common methods for modifying PDMS substrates is through the chemical linking of silane molecules to the surface. This reaction requires an initial plasma treatment followed by the addition of silane molecules which are covalently linked to the PDMS. This method, using silonal, is commonly used to create more hydrophilic PDMS surfaces, especially for fabricating microfluidic devices and bioreactors. Additionally, this method can be used to attach a variety of cellular binding domains to the surface of PDMS substrates. Lateef et al. demonstrated this by attaching oligopeptides containing the Arg-Gly-Asp sequence, termed as RGD peptides, a cellular binding ligand, to PDMS substrates (Fig. 12.2) [18]. They attached 3-aminopropyltriethoxysilane to substrates through a silination reaction, which provided a reactive primary amine on the surface. The primary amine was then linked to RGD peptides using carbodiimide chemistry, creating a PDMS surface on which cells could adhere. By using these methods to control the specific ECM-cell binding interactions, and because of the elastic properties of PDMS, mechanotransduction studies can be performed examining the effects of specific binding domains on cellular responses to mechanical force. Many mechanotransduction studies using traction force microscopy, substratedistension cell mechano-stimulus, and composite diaphragm inflation require a substrate with well defined mechanical properties. However, surface modification of PDMS substrates has been shown to alter their mechanical properties compared to untreated samples. Selsby and Shannon demonstrated that treating PDMS with water-vapor plasma and ultraviolet (UV) radiation alters the mechanical properties of the substrate [19]. They demonstrated both water-vapor plasma and UV treatments, in a dose dependent manner, decreases the compliance (increased stiffness) of PDMS substrates. Since most covalent surface modifications require water-vapor plasma or UV treatments, there tends to be slight discrepancies in the mechanical properties of PDMS substrates unaccounted for through bulk characterization. Even though PDMS has numerous advantageous properties that allow the examination of forces applied to and by cells, it also has several concerns that should be addressed. For instance, the surface modification of PDMS through water-vapor plasma treatments is temporary preventing an effective platform for long term mechanotransduction studies [18]. Additionally, PDMS provides a 2D platform that is not representative of the native 3D cellular environment.

12

Biomaterials for Studies in Cellular Mechanotransduction

271

Fig. 12.2 Steps for the surface modification of PDMS: (a) native surface; (b) amine surface; (c) malemide surface; and (d) peptide surface [18]

12.4

Hydrogels

Hydrogels, commonly formed from the cross-linking of hydrophilic natural or synthetic polymers, are widely used in cellular mechanotransduction studies because of several advantageous features [20–22]. First, hydrogels contain a large amount of water similar to a natural extracellular matrix. Second, a variety of cross-linking reactions are available to control gelation kinetics and subsequent properties. For example, polymers linked with methacrylic groups can form a hydrogel in response to an external stimulus such as ultraviolet or other forms of electromagnetic radiation. Third, hydrogels are readily modified to present chemical and mechanical properties similar to natural ECMs. Additionally, hydrogels uniquely allows one to

272

R. DeVolder and H. Kong

culture cells in a 3D environment similar to in vivo cellular microenvironments. Due to these advantageous features, hydrogels modified to present various cell adhesion molecules have been extensively used to examine the role of mechanotransduction in regulating cellular phenotypes in both 2D and 3D microenvironments. This section will therefore briefly describe strategies to assemble hydrogels for 2D and 3D cell cultures, respectively.

12.4.1 Hydrogel Design for 2D Cell Culture Over the past few decades, hydrogels have been extensively used to examine the role of matrix stiffness on regulating diverse cellular phenotypes, because of their controllable ranges similar to tissues [23–26]. A broad array of natural polymers (matrigel, collagen, fibrinogen, sodium alginate, etc.) and synthetic polymers (poly (acryl amide), poly(ethylene glycol) diacrylates and their derivatives, etc.) have been widely used as gel-forming polymers. Specifically, these polymers are physically associated with various cell adhesion proteins or chemically linked with oligopeptides containing cell adhesion sequences. The elastic moduli of cell adherent hydrogels are commonly controlled through the number of cross-linking groups. Additionally, stiffness gradients can be created for a construct by inserting a mask between pre-gelled solutions and polymerizing using an irradiation light source. Recently, the incorporation of photolabile cross-linking units in hydrogel networks has emerged as a method to control hydrogel properties in a transient manner [27–29]. This allows the examination of a dynamic mechanical environment, and its role towards regulating cell function. In addition, other material design parameters such as polymer concentration and rigidity of the polymer chain have also been altered to regulate hydrogel stiffness. These hydrogels are further integrated with nano- or microfabrication technologies to present functional cues that enable the evaluation of cellular responses to gel stiffness or external mechanical force. For example, various fluorescent microspheres are incorporated into hydrogels, to which cells adhere, so the effects of gel stiffness on cellular forces are examined using traction force microscopy [30]. Microsphere dislocations due to cellular contractions are used to quantify cellular traction forces. Kong et al. modified cell adhesion oligopeptides linked to gel matrices with a pair of fluorescent probes to implement a fluorescence resonance energy transfer (FRET)-based assay [30]. Throughout this FRET characterization, the role of hydrogel stiffness in the cellular phenotypes could be explained in terms of the nanoscale reorganization of cell adhesion peptides and also cellular traction forces. Additionally, Kong et al. used a similar FRET-based assay with fluorescently labeled cells and oligopeptides to quantify integrin ligation in relation to ECM mechanical properties and the resulting cellular phenotypic characteristics (Fig. 12.3) [31].

12

Biomaterials for Studies in Cellular Mechanotransduction

273

Fig. 12.3 Cellular binding peptides linked with fluorescence probe to hydrogels encapsulating fluorescently labeled cells for FRET [31]

12.4.2 Hydrogel Design for 3D Cell Culture Recently, the use of hydrogels to encapsulate cells in a 3D microenvironment has been increasingly used in cellular studies because of their controllable mechanical properties and similarities to tissues environments [22, 27, 32]. A broad array of natural and synthetic polymers has been used to form hydrogels for 3D cell studies. The mechanical properties of these hydrogels are controlled using similar methods as described for 2D cell studies; however, different strategies are used to encapsulate cells. Typically, cells are mixed with buffered pre-gelled solutions and then cross-linked to encapsulate cells within the hydrogel network in situ. A variety of gelation methods are used to encapsulate cells, which are all dependent on the polymer type. Many natural polymers such as collagen or mitrogel form hydrogels through physical gelation, but extensive efforts have been made to modify the polymers to induce chemically cross-linked junctions and to stabilize the gel structure during long cell culture durations. Various synthetic polymers are also modified in various ways to from gels via chemical cross-linking reactions or physical association, such as methacrylic-aglinate or colloidal hydrogels. The development of hydrogel systems that can encapsulate cells have created controllable platforms that can readily examine cellular responses to various mechanical stimuli in 3D environments. For example, one can examine the cellular response of encapsulated cells in hydrogels with varied stiffness. Matsudaira et al. examined the role matrix stiffness plays on prostate cancer cell’s migration [33]. Cancer cells were encapsulated in hydrogels with varied concentrations of Mitrigel, resulting in differing hydrogel stiffness, and they subsequently determined that cell

274

R. DeVolder and H. Kong

migration was modulated by matrix stiffness. Interestingly, at constant matrix ligand and cell integrin receptor levels, maximal cell migration occurred in less stiff hydrogels, which is contradictory to previous 2D studies. This result manifests that 3D cell cultures are important for precisely understanding the role of mechanotransduction on cellular phenotypes. The importance of a 3D environment in understanding cellular mechanotransduction, compared to 2D substrates, is increasingly being reported [32, 34]. Current strategies to control the mechanical stiffness of hydrogels typically results in a decreased permeability, which can separately impact cellular viability and activities. Recently, Cha et al. developed a new approach to decouple the dependency between stiffness and permeability of a cell-encapsulating hydrogel by increasing the cross-linking density of poly(ethylene glycol) dimethacrylate hydrogels with an anionic polysaccharide modified to chemically react with PEGDA [35]. This new hydrogel design strategy is a useful tool to examine the role of mechanotransduction on cellular activities in a more controllable manner. Additionally, cell-encapsulated hydrogels can be exposed to external mechanical stimuli, including compression and contraction, to understand their role on regulating cellular activities. Grinnel et al. examined cellular morphologies in response to mechanical contractile forces in a 3D environment [36, 37]. Interestingly, fibroblasts encapsulated in a collagen gel expressed a dendritic morphology in a relaxed gel state (Fig. 12.4a), while the cells subject to contraction expressed a stretched morphology (Fig. 12.4b). However, the mode and amplitude of external stress that can be applied to a hydrogel is limited because of a lower elasticity and high fragility, as compared with PDMS. Therefore, it will be essential to incorporate newly developed technologies to improve hydrogel’s elasticity and toughness, including but not limited to interpenetrating hydrogel networks and double networks.

Fig. 12.4 Fibroblasts embedded in collagen matrices and stained with rhodamine-conjugated palloidin cultured in a relaxed state (a) and in a stressed sate (b) (scale bar 50 mm) [36]

12

Biomaterials for Studies in Cellular Mechanotransduction

12.5

275

Conclusion

Understanding how mechanical signals, in the form of matrix stiffness or cyclic mechanical strain, regulate cellular phenotypes and gene expression is of the utmost importance for developing new biomedical technologies, and for ultimately improving people’s health in general. Biomaterials are an invaluable tool to allow one to understand the role of mechanical signals on cellular activities in a more precise manner and further tune the amplitude of mechanical signals [20, 26]. Many of these materials have specific properties that make them useful for specific cell studies: glass slides can easily be modified to contain biologically active molecules and allow easy cellular examination; PDMS has unique elastomeric properties that allow extensive mechanical deformation and has processing capabilities for creating sophisticated assemblies; and hydrogels allow 2D or 3D examination of cells in mechanically controlled micro-environments with specific concentrations and types of biologically active molecules. These materials, besides presenting specific mechanical properties, can also be incorporated into cellular platforms or bioreactors presenting mechanical stimuli.

References 1. Orr AW, Helmke BP, Blackman BR, Schwartz MA. Mechanisms of mechanotransduction. Dev. Cell 2006;10(1):11–20. 2. Darling EM, Athanasiou KA. Articular cartilage bioreactors and bioprocesses. Tissue Eng. 2003;9(1):9–26. 3. Sawakami K, Robling AG, Ai MR, Pitner ND, Liu DW, Warden SJ, et al. The Wnt co-receptor LRP5 is essential for skeletal mechanotransduction but not for the anabolic bone response to parathyroid hormone treatment. J. Biol. Chem. 2006;281(33):23698–23711. 4. Gomez DE, Alonso DF, Yoshiji H, Thorgeirsson UP. Tissue inhibitors of metalloproteinases: structure, regulation and biological functions. Eur. J. Cell Biol. 1997;74(2):111–122. 5. Jones FS, Jones PL. The tenascin family of ECM glycoproteins: structure, function, and regulation during embryonic development and tissue remodeling. Dev. Dyn. 2000; 218(2):235–259. 6. Guilak F, Cohen DM, Estes BT, Gimble JM, Liedtke W, Chen CS. Control of stem cell fate by physical interactions with the extracellular matrix. Cell Stem Cell 2009;5(1):17–26. 7. Netti PA, Berk DA, Swartz MA, Grodzinsky AJ, Jain RK. Role of extracellular matrix assembly in interstitial transport in solid tumors. Cancer Res. 2000;60(9):2497–2503. 8. Wang N, Ingber DE. Probing transmembrane mechanical coupling and cytomechanics using magnetic twisting cytometry. Biochem. Cell Biol. 1995;73(7–8):327–335. 9. Merkel R, Nassoy P, Leung A, Ritchie K, Evans E. Energy landscapes of receptor-ligand bonds explored with dynamic force spectroscopy. Nature 1999;397(6714):50–53. 10. Makamba H, Kim JH, Lim K, Park N, Hahn JH. Surface modification of poly(dimethylsiloxane) microchannels. Electrophoresis 2003;24(21):3607–3619. 11. Garcia AJ, Vega MD, Boettiger D. Modulation of cell proliferation and differentiation through substrate-dependent changes in fibronectin conformation. Mol. Biol. Cell 1999; 10(3):785–798.

276

R. DeVolder and H. Kong

12. Lahann J, Balcells M, Lu H, Rodon T, Jensen KF, Langer R. Reactive polymer coatings: a first step toward surface engineering of microfluidic devices. Anal. Chem. 2003;75(9):2117–2122. 13. Balaban NQ, Schwarz US, Riveline D, Goichberg P, Tzur G, Sabanay I, et al. Force and focal adhesion assembly: a close relationship studied using elastic micropatterned substrates. Nat. Cell Biol. 2001;3(5):466–472. 14. Harris AK, Wild P, Stopak D. Silicone-rubber substrata – new wrinkle in the study of cell locomotion. Science 1980;208(4440):177–179. 15. Kamotani Y, Bersano-Begey T, Kato N, Tung YC, Huh D, Song JW, et al. Individually programmable cell stretching microwell arrays actuated by a Braille display. Biomaterials 2008;29(17):2646–2655. 16 . Leclerc E, David B, Griscom L, Lepioufle B, Fujii T, Layrolle P, et al. Study of osteoblastic cells in a microfluidic environment. Biomaterials 2006;27(4):586–595. 17. Ruiz SA, Chen CS. Emergence of patterned stem cell differentiation within multicellular structures. Stem Cells 2008;26(11):2921–2927. 18. Lateef SS, Boateng S, Hartman TJ, Crot CA, Russell B, Hanley L. GRGDSP peptide-bound silicone membranes withstand mechanical flexing in vitro and display enhanced fibroblast adhesion. Biomaterials 2002;23(15):3159–3168. 19. Selby JC, Shannon MA. A method to fabricate mesoscopic freestanding polydimethylsiloxane membranes used to probe the rheology of an epithelial sheet. J. Biochem. Biophys. Methods 2008;70(6):932–944. 20. Lee KY, Mooney DJ. Hydrogels for tissue engineering. Chem. Rev. 2001;101(7):1869–1879. 21. Hennink WE, van Nostrum CF. Novel crosslinking methods to design hydrogels. Adv. Drug Deliv. Rev. 2002;54(1):13–36. 22. Drury JL, Mooney DJ. Hydrogels for tissue engineering: scaffold design variables and applications. Biomaterials 2003;24(24):4337–4351. 23. Peyton SR, Kim PD, Ghajar CM, Seliktar D, Putnam AJ. The effects of matrix stiffness and RhoA on the phenotypic plasticity of smooth muscle cells in a 3-D biosynthetic hydrogel system. Biomaterials 2008;29(17):2597–2607. 24. Kisiday J, Jin M, Kurz B, Hung H, Semino C, Zhang S, et al. Self-assembling peptide hydrogel fosters chondrocyte extracellular matrix production and cell division: implications for cartilage tissue repair. Proc. Natl. Acad. Sci. U.S.A. 2002;99(15):9996–10001. 25. Paszek MJ, Zahir N, Johnson KR, Lakins JN, Rozenberg GI, Gefen A, et al. Tensional homeostasis and the malignant phenotype. Cancer Cell 2005;8(3):241–254. 26. Sia SK, Whitesides GM. Microfluidic devices fabricated in poly(dimethylsiloxane) for biological studies. Electrophoresis 2003;24(21):3563–3576. 27. Ifkovits JL, Burdick JA. Review: photopolymerizable and degradable biomaterials for tissue engineering applications. Tissue Eng. 2007;13(10):2369–2385. 28. Wong DY, Griffin DR, Reed J, Kasko AM. Photodegradable hydrogels to generate positive and negative features over multiple length scales. Macromolecules 2010;43(6):2824–2831. 29. Sniadecki N, Desai RA, Ruiz SA, Chen CS. Nanotechnology for cell-substrate interactions. Ann. Biomed. Eng. 2006;34(1):59–74. 30. Kong HJ, Polte TR, Alsberg E, Mooney DJ. FRET measurements of cell-traction forces and nano-scale clustering of adhesion ligands varied by substrate stiffness. Proc. Natl. Acad. Sci. U.S.A. 2005;102(12):4300–4305. 31. Kong HJ, Boontheekul T, Mooney DJ. Quantifying the relation between adhesion ligandreceptor bond formation and cell phenotype. Proc. Natl. Acad. Sci. U.S.A. 2006; 103(49):18534–18539. 32. Cukierman E, Pankov R, Stevens DR, Yamada KM. Taking cell-matrix adhesions to the third dimension. Science 2001;294(5547):1708–1712. 33. Zaman MH, Trapani LM, Siemeski A, MacKellar D, Gong HY, Kamm RD, et al. Migration of tumor cells in 3D matrices is governed by matrix stiffness along with cell-matrix adhesion and proteolysis. Proc. Natl. Acad. Sci. U.S.A. 2006;103(29):10889–10894.

12

Biomaterials for Studies in Cellular Mechanotransduction

277

34. Pedersen JA, Swartz MA. Mechanobiology in the third dimension. Ann. Biomed. Eng. 2005;33(11):1469–1490. 35. Cha C, Kohmon RE, Kong H. Biodegradable polymer crosslinker: independent control of stiffness, toughness, and hydrogel degradation rate. Adv. Funct. Mater. 2009; 19(19):3056–3062. 36. Grinnell F. Fibroblast-collagen-matrix contraction: growth-factor signalling and mechanical loading. Trends Cell Biol. 2000;10(9):362–365. 37. Grinnell F. Fibroblast biology in three-dimensional collagen matrices. Trends Cell Biol. 2003;13(5):264–269.

Chapter 13

Optical Sensing of Red Blood Cell Dynamics YongKeun Park, Catherine A. Best, and Gabriel Popescu

This chapter is part of Section IV: Tools for Exploring Mechanobiology

Abstract Human red blood cell membrane (RBC) has remarkable deformability, which is crucial for its oxygen transportation in the blood circulatory system. This deformability of the RBC membrane can be altered by several patho-physiological conditions. Here we present recent development of optical imaging techniques to measure dynamic fluctuations in the RBC membrane, from which RBC membrane mechanical properties are probed non-invasively.

13.1

Introduction

13.1.1 RBC Membrane Mechanics The red blood cell (RBC) deformability in microvasculature governs the cell’s ability to survive the physical demands of circulation, as well as the cell’s ability to transport oxygen in the body [1]. Interestingly, RBCs must periodically pass a deformability test by being forced to squeeze through narrow passages (sinuses) in the spleen; upon failing this mechanical assessment, the cell is destroyed and removed from circulation by macrophages [2]. Quantifying and understanding the mechanics of live RBCs requires sensitive probes of their structure at the nanoscale, and promises new insights into the etiology of a number of human diseases [3]. In the healthy individual, these cells withstand repeated, large-amplitude mechanical deformations as they circulate through the microvasculature. Throughout their 120 day life span, the RBCs mechanical integrity degrades and ultimately they are replaced by new circulating RBCs. Certain pathological conditions such as spherocytosis, malaria, and sickle cell disease cause changes in both the equilibrium shape

G. Popescu (*) Department of Electrical and Computer Engineering, Quantitative Light Imaging Laboratory, Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL 61801, USA e-mail: [email protected] A. Wagoner Johnson and Brendan A.C. Harley (eds.), Mechanobiology of Cell-Cell and Cell-Matrix Interactions, DOI 10.1007/978-1-4419-8083-0_13, # Springer Science+Business Media, LLC 2011

279

280

Y. Park et al.

and mechanics of RBCs, which impact their transport function. Thus, understanding the microrheology of RBCs is highly interesting both from a basic science and a clinical point of view. Lacking a conventional 3D cytoskeleton structure, RBCs maintain their shape and mechanical integrity through a spectrin-dominated, triangular 2D network associated with the cytosolic side of their plasma membrane. This semiflexible filament network confers shear and bulk moduli to the composite membrane structure [4]. The fluid lipid bilayer is thought to be the principal contributor to the bending or curvature modulus of the composite membrane. Little is known about the molecular and structural transformations that take place in the membrane and spectrin network during the cell’s response to patho-physiological conditions, which are accompanied by changes in RBC mechanics.

13.1.2 Methods of Investigation Several techniques have been used recently to study the mechanical properties of live red blood cells [3]. Micropipette aspiration [5], electric field deformation [6], and optical tweezers [3] provide quantitative information about the shear and bending moduli of RBC membranes in static conditions (see Fig. 13.1). However, dynamic, frequency-dependent knowledge of RBC mechanics is currently very limited with the exception of ref. [7]. RBC thermal fluctuations (“flickering”) have been studied for more than a century [8] to better understand the interaction between the lipid bilayer and the cytoskeleton [9–12]. Membrane fluctuation dynamics of RBCs can be influenced by physiological conditions and disease states. Fluctuations in phospholipid bilayer and attached spectrin network are known to be influenced by cytoskeletal defects, stress, and actin–spectrin dissociations arising from metabolic activity linked to adenosine-50 -triphosphate (ATP) concentration [5, 13–16]. Nevertheless, quantifying these motions is experimentally challenging, and reliable spatial and temporal data are desirable [10, 17, 18]. Existing optical methods, including phase contrast microscopy (PCM) [17], reflection interference contrast microscopy (RICM) [19], and fluorescence interference contrast (FLIC) [18], are limited in their ability to measure cell membrane

Fig. 13.1 Schematic representation of conventional methods to study mechanical properties of RBCs. (a) Micropipette aspiration. (b) Magnetic twisting cytometry. (c) Optical trapping technique

13

Optical Sensing of Red Blood Cell Dynamics

281

displacements. In PCM, the optical phase shifts in the light passing through thin transparent samples is converted into amplitude of the light [20]. PCM has been widely used for visualizing red blood cell membrane dynamics [17]. However, quantifying the out-of-plane membrane fluctuations in RBCs using PCM is challenging. It is well known that PCM provides phase shifts quantitatively only for samples that are optically much thinner than the wavelength of light, which is a condition hardly satisfied by any cell type. RICM measured membrane dynamics by recording interference pattern formed from the light reflected from the cell surface and from the cover glass. Similar to PCM, a single RICM measurement cannot provide the absolute cell thickness unless additional measurements or approximations are made [21]. FLIC relies on inferring the absolute position of fluorescent dye molecules attached to the cell membrane from the absolute fluorescence intensity, which may limit both the sensitivity and acquisition rate of the technique [18]. RBCs lack nuclei and organelles and can be assumed optically homogeneous, i.e. characterized by a constant refractive index. Therefore, measurement of the cell optical path-length via interferometric techniques can provide information about the physical topography of the membrane with sub-wavelength accuracy and without contact. In the following, we present a review of the emerging field of quantitative phase imaging (QPI).

13.2

Quantitative Phase Imaging

13.2.1 Review of the Field Quantifying the optical phase shifts associated with cells gives access to information about morphology and dynamics at the nanometer scale (for a review, see ref. [22]). Over the past decade, the development of quantitative phase imaging techniques has received increased scientific interest. The technology can be divided into singlepoint and full-field measurements, according to the experimental geometry employed. Several point-measurement techniques have been applied for investigating the structure and dynamics of live cells [23–29]. This type of measurement allows for fiber-optic implementation and also high-speed punctual phase measurement by using a single, fast photo-detector. Full-field phase measurement techniques, on the other hand, provide simultaneous information from a large number of points on the sample, which has the benefit of studying both the temporal and spatial behavior of the biological system under investigation [11, 30–41]. With the recent advances in two dimensional array detectors, full-field phase images can now be acquired at high speeds (i.e., thousands of frames per second). Various point-measurement techniques have been developed over the years for quantifying phase shifts at a given point through biological samples. This class of techniques can be described as an extension of optical coherence tomography (OCT) [42] to provide measurements of phase, phase dispersion and birefringence associated with biological structures. DeBoer et al. demonstrated depth-resolved birefringence

282

Y. Park et al.

measurements with a polarization sensitive OCT system [43]. Differential phase-contrast OCT images have also been generated with a polarization-sensitive OCT instrument [44]. Recently, polarization-sensitive OCT was used to quantify phase retardation in the retinal nerve fiber [45]. An instantaneous quadrature technique was proposed based on using a one dimensional fiber coupler and the inherent phase shift between different output fibers [46]. Electrokinetic [47] and thermorefractive [48] properties of tissue and tissue phantoms have been measured by differential phase OCT. Phase sensitive OCT-type measurements have also been performed for studying static cells [29], for monitoring electric activity in nerves [27, 28], and spontaneous beating in cardiomyocytes [25]. However, these methods rely on single point measurements which, for imaging purposes, require raster scanning. This raster scanning procedure is often time consuming, reducing the applicability range of the OCT techniques to measure the fast dynamic fluctuations in the RBC membrane. Typically, the temporal resolution of the RBC membrane dynamics is around a millisecond. However, recent ultrahigh-speed OCTs achieve up to ~300 kHz A-scans [49], which is now able to be employed to measure the RBC membrane dynamics. Recently, new full-field phase imaging techniques, which are suitable for spatially-resolved investigation of biological structures, have been developed to overcome these limitations. Combining phase shifting interferometry with Horn microscopy, DRIMAPS (digitally recorded interference microscopy with automatic phase-shifting) has been proposed as a new technique for quantitative biology [30, 50]. This quantitative phase imaging technique has been successfully used for measuring cell spreading [31], cell motility [32], cell growth and dry mass [51]. A full-field quantitative phase microscopy method was also developed by using the transport-of-irradiance equation [52, 53]. The technique is inherently stable against phase noise because it does not require using two separate beams as in typical interferometry experiments. This approach requires recording images of the sample displaced through the focus and subsequently numberically solving partial differential equations. Digital holography was developed a few decades ago [54] as a technique that combines digital recording with traditional holography [55]. Typically, the phase and amplitude of the imaging field are measured at an out-of-focus plane. By solving numerically the Fresnel propagation equation [56], one can determine the field distribution at various planes. For optically thin objects, this method allows for reconstructing the in-focus field and, thus, retrieving the phase map characterizing the sample under investigation. This method has been implemented in combination with phase shifting interferometry [57]. More recently, digital holography has been adapted for quantitative phase imaging of cells [36, 37, 58]. In recent years, new full-field quantitative phase imaging techniques have been developed for studying live cells. The principle of the quantitative phase imaging is depicted in Fig. 13.2. The advance of Fourier phase microscopy (FPM) [39, 59], Hilbert phase microscopy (HPM) [40, 41], and diffraction phase microscopy (DPM) [11, 60] came in response to the need for high phase stability over broad temporal scales.

13

Optical Sensing of Red Blood Cell Dynamics

283

Fig. 13.2 Schematic representation of optical phase delay when the light passes through a red blood cell. The difference between refractive index of cytoplasm, nc, and of medium, nm, results in the phase delay Df(x,y). The field containing the phase delay, AejDf(x,y), where A is amplitude of the light, can be quantitatively measured by using interferometry

13.2.2 Diffraction Phase Microscopy (DPM) DPM combines the single-shot feature of the spatially-modulated quantitative phase measurement with the common-path geometry [11]. DPM provides quantitative phase images that are inherently stable to the level of the sub-nanometer optical path length and at an acquisition speed limited only by the detector. The experimental setup is shown in Fig. 13.3. Any continuous (or pulsed, for that matter) wave laser can be used as an illumination source. We use either Ar2+ laser or He-Ne laser. A specimen, located on the sample stage of an inverted microscope, is projected to the image plane, IP1. The image is further magnified and delivered to the second image plane, IP2. A holographic grating placed at IP2 generates multiple diffraction orders. Each order contains the same sample-induced electric field (E-field). The 0th- and 1st-order beams are isolated to constitute a Mach-Zehnder configuration. The 0th-order beam is spatially low-pass filtered using a pinhole in a telescopic imaging system; the 0th-order beam becomes a clean plane wave, free of sampleinduced E-field, at the camera plane and can serve as the reference beam in the interferometer. The 1st-order beam travels through the same optical elements as the 0th-order beam except for the pinhole, such that it serves as the sample beam. Both beams interfere at the camera plane and generate a spatially modulated interference image. Since both beams share almost the same beam path, common-mode phase noise is cancelled out upon interference. The direction of the spatial modulation was chosen at an angle of 45 with respect to the x- and y-axes of the CCD, such that the total field at the CCD plane has the form Eðx; yÞ ¼ jE0 jejðu0 xþv0 yÞ þ jE1 ðx; yÞje jDfðx;yÞ

(13.1)

In (13.1), jE0 j and jE1 j are the amplitudes of the 0th- and 1st-orders of diffraction, respectively. Dfðx; yÞ is the relative phase delay induced by the sample compared to the reference 0th order wave. u0 ; v0 represent the spatial frequencies induced by the grating to the 0th-order with respect to x-, and y-axis, respectively.

284

Y. Park et al.

CCD

Laser

Optical Fiber Spatial Filter

Collimator

Sample Object Lens TL

Microscopy

G IP2

IP1

f1

f1

f2

Fig. 13.3 DPM experimental setup. TL, tube lens. IP1-2, image planes. G grating (Modified from ref. [61])

The interference image is captured by a CCD camera as shown in Fig. 13.4a, which has the form Iðx; yÞ ¼ E E ¼ jE0 j2 þ jE1 ðx; yÞj2 þ 2jE0 jjE1 ðx; yÞj cosðu0 x þ n0 y Dfðx; yÞÞ

(13.2)

We can rewrite (13.2) as follows, Iðx; yÞ ¼ bðx; yÞ þ cðx; yÞejðu0 xþn0 yÞ þ c ðx; yÞejðu0 xþn0 yÞ ;

(13.3)

where bðx; yÞ ¼ jE0 j2 þ jE1 ðx; yÞj2 and cðx; yÞ ¼ jE0 jjE1 ðx; yÞje jDfðx;yÞ . Thus, the 2D Fourier transform of the measured interferogram has the form (see Fig. 13.4b), Gðu; vÞ fft½Iðx; yÞ ¼ Bðu; vÞ þ Cðu u0 ; v v0 Þ þ C ðu þ u0 ; v þ v0 Þ; (13.4) where Bðu; vÞ and Cðu; vÞ are Fourier transformed images of bðx; yÞ and cðx; yÞ, respectively. C(u,v) can be chosen by selecting the area in G(u,v) with a circled mask centered at (u0, v0) with a specific radius corresponding to the numerical

13

Optical Sensing of Red Blood Cell Dynamics

285

Fig. 13.4 (a) Interference pattern measured by DPM. (b) 2D Fourier transform of the interference pattern in (a). (c) Numerical process to retrieve quantitative image information. (d) Retrieved phase image from the interferogram. Units for the graybar is mm. (e) Temporal path-length fluctuations for a point and an area in a no-sample image. Figure 13.4e is reproduced with permission from ref. [11]. Copyright (2006) OSA

aperture of the optical system, and shifting to the center (Fig. 13.4c) [62]. Then, the 2D inverse Fourier transform of C(u,v) will result in the following E-field fft1 ½Cðu; vÞ ¼ cðx; yÞ ¼ jE0 jjE1 ðx; yÞje jDfðx;yÞ :

(13.5)

286

Y. Park et al.

From the retrieved cðx; yÞ, one can extract the quantitative phase image of the sample Dfðx; yÞ (Fig. 13.4d). To quantify the stability of the DPM instrument and thus the sensitivity of cell topography to dynamical changes, we recorded sets of 1,000 no-sample images, acquired at 10.3 ms each and the temporal path-length trace of an arbitrary point (3 3 pixel average) showed a standard deviation of 0.53 nm as shown in Fig. 13.4e [11]. Such a sensitive quantitative phase imaging capability is of a particular interest to measuring the shape and dynamics of biological cells, especially red blood cell membranes. By extracting optical path-length shifts produced at each point across the cell, DPM quantitatively measures cell thickness with spatial and temporal resolutions of nanometer and millisecond, respectively. This optical path-length information can be readily translated into cell thickness, since mature RBCs are characterized by a spatially uniform refractive index. The instantaneous cell thickness map is obtained as hðx; y; tÞ ¼

l Dfðx; y; tÞ; 2pDn

(13.6)

where l is the wavelength of the laser, and Df(x,y,t) is the quantitative phase image measured by DPM as in (13.5). The refractive index contrast Dn between the RBC and the surrounding PBS is mainly contributed from hemoglobin (Hb), which is optically homogeneous in RBC cytosol. The typical value for Dn is 0.06 for intact RBCs (~31 g/dl of Hb concentration) for l ¼ 633 nm [63]. DPM has been combined with epi-fluorescence microscopy to simultaneously image both the nanoscale structure and dynamics, as well as specific functional information in live cells [61]. The common-path geometry of DPM matches the optical path lengths for the sample and reference arms such that the alignment is independent of the wavelength and temporal coherence of the illumination source. Recently, DPM has shown to measure the wavelength-dependent phase maps, obtained via dispersion, which can provide the concentration of a specific molecule [64].

13.3

RBC Membrane Dynamics

13.3.1 Measurement of Thermal Fluctuations in RBC Membranes and in Vesicles In order to investigate the physical difference between RBC membrane and free bilayer fluctuations, we measured the fluctuations of giant unilamellar vesicles (GUVs). The membrane composition we used was 100% SOPC (1-Stearoyl-2Oleoyl-sn-Glycero-3-Phosphocholine). The GUVs were electroformed for 2 h under a 1 V, 10 Hz AC field in a 300 mM glucose solution [65] and diluted in a

13

Optical Sensing of Red Blood Cell Dynamics

287

sucrose solution for increased optical contrast, and to slightly deflate the vesicles. The refractive index contrast with respect to the surrounding fluid, Dn ¼ 0.011, was obtained from the quantitative phase image by assuming a spherical profile for the vesicles. Sets of 1,000 time-resolved phase images of individual GUVs were recorded at 10.3 ms/frame. The vesicles under investigation ranged in diameter from 8 to 12 mm and their thickness profile was obtained from phase images as uðx; y; tÞ ¼ ðl=2pDnÞfðx; y; tÞ. Each 1,000-frame data set was then separated into groups of 128 frames, analyzed separately to obtain mean squared displacements, and finally averaged to provide statistically significant information for each vesicle. Using Fourier transformations both in time and space, we obtained the mean squared displacement as a function of spatial wave vector and temporal frequency, Du2 ðq; oÞ. Figure 13.5 summarizes the static (spatial) behavior of the thermal R fluctuations averaged over five vesicles, Du2 ðqÞ ¼ Du2 ðq; oÞdo, with o is the temporal angular frequency, and q ¼ jqj. The data can be fit very well over the spatial wave vector interval 1
kB T ; kq4 þ sq2

(13.7)

where kB is the Boltzmann constant and T ¼ 298 K is the absolute temperature for our experiments. The measured values at low-q deviate from this form; however, this is expected for wavelengths corresponding to the size of the vesicles. The only two parameters used for the fit are the tension coefficient, s, and the bending modulus, k.

Fig. 13.5 Mean squared displacements of giant unilamellar vesicles fitted with (13.7), as indicated. The solid lines indicate the asymptotic behavior. The inset shows the phase image of a vesicle, with the gray scale in radians. Reproduced with permission from ref. [66], Copyright (2006) APS

288

Y. Park et al.

The value for the bending modulus obtained from the fit is k ¼ ð0:7 0:12Þ 1020 J and the tension coefficient was s ¼ ð3:5 0:6Þ 107 J=m2. These values agree very well with what was measured on vesicles using pipette aspiration [67]. Blood samples were collected and centrifuged for 10 min at an acceleration of 2,000 g and temperature of 5 C to separate RBCs from plasma. The cells were washed three times with a saline solution and were ultimately resuspended in phosphate buffered saline (pH ¼ 7.4), as described in ref. [68]. The RBCs were then placed between cover slips and imaged without additional preparation. Our samples were primarily composed of RBCs with the typical discocytic shape (DCs), but also contained cells with abnormal morphology which formed spontaneously in the suspension, such as echinocytes (ECs) (cells with a spiculated shape), and spherocytes (SCs) (cells with a roughly spherical shape). By taking into account the free energy contributions of both the bilayer and cytoskeleton, these morphological changes have been successfully modeled [69]. Figure 13.6a–c show typical quantitative phase images of cells in these three groups. For comparison, we also analyzed the motions of RBCs fixed with 40 mM gluteraldehyde using a standard procedure [68]. The resultant mean squared displacements, Du2 ðqÞ, for each group of 4–5 cells are summarized in Fig. 13.6d. The fixed cells show significantly diminished fluctuations. The curves associated with the three untreated RBC groups exhibit a power-law behavior with an exponent a ¼ 2. As in the case of vesicles, this dependence is an indication of tension; however, the RBC tension is determined by the confinement of

Fig. 13.6 (a–c) Quantitative phase images of a discocyte (a), echinocyte (b), and spherocyte (c). The gray bar shows thickness in microns. (d) Mean squared displacements for the three RBC groups and for the gluteraldehyde (GA) fixed cells. Reproduced with permission from ref. [66]. Copyright (2006) APS

13

Optical Sensing of Red Blood Cell Dynamics

289

the bilayer by the cytoskeleton [9, 70]. Based on this model, we fitted the data to extract the tension coefficient for each individual cell. The average values obtained for the DCs, ECs, and SCs are, respectively, s ¼ ð1:5 0:2Þ 106 J=m2 , s ¼ ð4:05 1:1Þ 106 J=m2 , and s ¼ ð8:25 1:6Þ 106 J=m2 . Thus our data indicates the existence of positive tension in the membrane, which agree very well with the simulation by Discher et al. [71]. The tension coefficient of RBCs is 4–24 times larger than that of vesicles, which suggests that the contribution of the cytoskeleton might be responsible for this enhancement. Further, it is known that the cytoskeleton plays a role in the transitions from a normal red blood cell shape to abnormal morphology, such as EC and SC [69]. Therefore the consistent increase in tension we measured for the DC-EC-SC transition can be explained by loss of membrane, thus of surface area, and potential changes in the cytoskeleton which pins the bilayer.

13.3.2 Mathematical Model for the RBC Membrane Fluctuations As shown above, the standard description of a RBC treats the membrane as a flat surface; in reality the membrane of a RBC is curved and has the compact topology of a sphere. Our new model takes into account the geometrical effect and incorporates the curvature and topology of a RBC within the fluctuation analysis (for details, see ref. [72]). Thus we are able to analytically trace membrane elasticity and fluid hydrodynamics more accurately. Because of the membrane curvature, the bending and compression modes of a spherical membrane are coupled in the linear order in deformations. To understand the height fluctuation spectrum of a spherical membrane, we need to account for the full linear response of the membrane to applied forces. The geometric coupling of bending and compression generates undulatory dynamics consistent with the experiments, and does not require the postulate of a surface-tension-like term in the Hamiltonian of the composite membrane [9]. The coupling of the fluid to the membrane is done using the usual stick boundary conditions and the stress balance condition at the surface of the membrane. To quantitatively investigate the material properties of RBCs, we analyze the spatial and temporal correlations of the out-of-plane fluctuations of the RBC membrane and interpret them using the viscoelastic continuum model of the composite spectrin-network/lipid membrane. This model, due to Levine and co-workers [73], accounts for the linear coupling between the bending and compression modes of a curved membrane, and thus provides a better description of the dynamics of the RBC than theories based on a flat membrane. The undulatory dynamics of a RBC are probed experimentally by measuring the spatial and temporal correlations of the out-of-plane fluctuations of the membrane. The fluctuation-dissipation theorem provides a theoretical description of these correlations from the deformation energy of the spherical shell including a bending energy and an in-plane elastic energy [74]. The correlation of height fluctuations

Height-height Correlation (m2s)

290

Y. Park et al. x 10−16

a

3

x 10−16

Discocyte w = 6 rad /sec w = 50 rad /sec

2

b

3

1 0

x 10−16

Echinocyte

2

2

1

1

0 0

2

4 6 d (µm)

8

c

3

Spherocyte

0 0

2

4 6 d (µm)

8

10

0

2

4 6 d (µm)

8

Fig. 13.7 Height-height correlation function from experiments (faint lines, N ¼ 30 per each group) and calculation (thick lines). (a–c) Height-height correlation function as a projected distance for DCs (a), ECs (b), and SCs (c) for o ¼ 6 rad/s (gray) and o ¼ 50 rad/s (black ). Reproduced with permission from ref. [72]. Copyright (2010) NAS

at two points on the membrane separated by the projected distance d and time t is defined by Cðd; tÞ ¼ hDhðd; tÞDhð0; 0Þi;

(13.8)

where the angular brackets denote both spatial and temporal averaging. From the results of a previous theoretical work [75] and the fluctuation-dissipation theorem, we find that the Fourier transform of this function in the frequency domain, o, is given by d 2 2kB T X d2 (13.9) Cðd; oÞ ¼ 1 2 Im½www ðl; oÞPl 1 2 ; 4R 4R o l where Pl(x) is the Legendre polynomial of lth order, R is the radius of RBC. These height-height correlation functions are plotted in Fig. 13.7a–c. We show the spatial decay of the height-height correlations at two fixed frequencies: one corresponding to the elastic plateau (low frequency) and one to the viscously-dominated regime (high frequency). At lower frequencies, we see a pronounced oscillatory behavior in the correlation function. The negative correlations are due to the dominance of the small-l contributions to the response function at low frequencies. At higher frequencies, we see that there is a shorter-ranged and nearly monotonic decay of the height-height correlations. Remarkably, this mode coupling predicts a decay rate o / q, which is identical to the behavior by a model of the membrane with an effective surface tension proposed by Gov et al. [9, 66].

13.3.3 Morphology Effects We used the new model of RBC mechanics over the commonly occurring DC-ECSC shape transition (Fig. 13.6a–c). From measurements of dynamic fluctuations on

13

Optical Sensing of Red Blood Cell Dynamics

291

RBC membranes, we extract the mechanical properties of the composite membrane structure. Subtracting the instant thickness map by the averaged thickness map provides the instantaneous displacement maps of the RBC membranes. Over the morphological transition from DC to SC,pthe root-mean-squared amplitude of ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ﬃ equilibrium membrane height fluctuations hDh2 i decreases progressively from 134 nm (DCs) to 92 nm (ECs) and 35 nm (SCs), indicating an increase in cell stiffness. In order to extract the material properties of the RBC, we fit our model to the oÞ by adjusting the following parameters: measured correlation function Cðd; the shear m and bulk K moduli of the spectrin network, the bending modulus k of the lipid bilayer, the viscosities of the cytosol c and the surrounding solvent s, and the radius of the sphere R. We constrain our fits by setting R to the average radius of curvature of the RBC obtained directly from the data and fixing the viscosities for all data sets to be s ¼ 1.2 mPa s, s ¼ 5.5 mPa s [9, 72]. Finally, for a triangular elastic network we expect m ¼ l so we set K ¼ 2m [73]. The fitting parameter space is now reduced to two dimensions and spanned by the bending modulus k of the lipid bilayer and the shear modulus m of the spectrin network. These are obtained by fitting the correlation data for each RBC. The experimental data (thin lines) and the best fit of the average data (thick lines) are shown in Fig. 13.7a–c. The theory generates a very good fit to the data including at low frequencies, where anti-correlated motion is observed in DCs and ECs. Both at higher frequencies (gray curves) and for stiffer membranes (e.g., SC cells) these anti-correlations are strongly suppressed. The parameters extracted from the fit are shown in Fig. 13.8a–c. The extracted bending modulus increases significantly during the DC-EC-SC transition ( p < 107). Their mean values (and associated standard deviations) are 6.3 1.0 (DC), 11.9 2.5 (EC), and 23.8 4.1 (SC) in units of kBT (Fig. 13.8b). These values are in general agreement with those expected for a phospholipid bilayer (5–20) kBT [74]. The increase in bending modulus suggests changes in the composition of the lipid membrane. We measured directly the change in surface area of RBCs during the transition from DC to SC morphologies and found a 31% decrease in surface area (not accounting for surface area stored in fluctuations). This surface area decrease must be accompanied by a loss of lipids, via membrane blebbing and microvesiculation. Previous work indicates that, after vesiculation, RBCs indeed have a higher cholesterol/phospholipid ratio and a lower phosphatidylserine/phospholipid ratio compared to the exovesicles shed from the parent RBCs [75]. Thus, there is evidence that a significant change in lipid composition of the RBC bilayer accompanies the morphological changes from DC to EC and SC. It is thought that these changes in lipid composition generate the observed changes in the bending modulus. The shear modulus results are shown in Fig. 13.8c. The fitted shear moduli are 6.4 1.4 (DC), 10.7 3.5 (EC), and 12.2 3.0 (SC) in mN/m. These values are consistent with earlier work based on micropipette aspiration [76] and optical tweezers [77]. The magnitude of the measured shear modulus also agrees well with simple elastic models of the spectrin network. We calculated the shear modulus of a disorder-free triangular network of wormlike chain elastic elements [78].

292

a

Y. Park et al.

b

20

30

κ (xkBT)

μ (μNm−1)

15

10

10

5

0

20

Discocyte Echinocyte Spherocyte

0

Discocyte

Echinocyte Spherocyte

c

6 Discocyte 4 2 0 6

Echinocyte

4 2 0 6

Spherocyte

4 2 0

0

5

10

15

20

Fig. 13.8 Shear modulus and bending modulus. (a) Shear moduli of DCs, ECs, and SCs with their mean values represented by the horizontal lines. p-values verify that the differences in the shear moduli between morphological groups are statistically significant: p < 105 between DCs and ECs, and between DCs and SC; p < 104 between ECs and SCs. (b) Bending modulus of three groups. (c) Distributions of shear moduli. The Gaussian fits are overlapped. The centers of the Gaussian fits are 7.34 (DCs), 6.97 and 12.93 (ECs), and 7.40 and 13.80 (SCs). Reproduced with permission from ref. [72]. Copyright (2010) NAS

Taking typical values for the lattice constant of the network (90 nm) and persistence length (7.5 nm) [79], we find that the network shear modulus of 5 mN/m requires a spectrin contour of 197 nm. This contour length is consistent with the previously published value of 194 nm [80]. The shear modulus of SCs and ECs increased by roughly a factor of two compared to the DCs (p < 105). There is, however, significant cell-to-cell variation in the shear modulus. While the histogram of shear moduli of DCs can be fit by a single Gaussian distribution centered at 6.7 mN/m, the analogous shear moduli distributions for ECs and SCs are bimodal, with peaks at, respectively, 6.4 and

13

Optical Sensing of Red Blood Cell Dynamics

293

13.0 mN/m (ECs), and 6.8 and 12.9 mN/m (SCs) (Fig. 13.8c). These data suggest that there are essentially two independent conformations of the spectrin network: a soft configuration (m ﬃ 7 mN/m) and a stiff one (m ﬃ 13 mN/m). Essentially all DCs have the soft configuration, but the morphological transition to EC and then SC promotes the transition to the stiff network configuration. Given the well-known extension-hardening of a wormlike chain [81], one might hypothesize that the stiff state of the spectrin network results from simply stretching the individual filaments. The observed stiffening would require approximately a 50% extension of the lattice constant of the spectrin network during the DC-EC-SC transition. However, such a stretch is incompatible with the observed area decrease of the membrane during this morphological transition. From the topographical information measured by DPM, we calculate mean surface areas of 139.4 (DC), 143.4 (EC), and 96.3 (SC) in mm2. Thus, we propose that the observed morphological changes must be accompanied by modifications of the spectrin elasticity, the connectivity of the network, or its attachment to the lipid bilayer.

13.3.4 Osmotic Effects Biological cells are highly sensitive and responsive to their physiological environments; they tolerate limited temperature, and ion concentration ranges. Homeostasis of osmosis in the blood serum is crucial as either hypertonic or hypotonic conditions affect circulation. RBCs do not have a complex cell-volume regulatory system; they simply undergo remarkable changes in volume when exposed to nonphysiologic osmolarities. When exposed to a hypotonic solution, water enters into the RBC cytosol through the membrane, resulting in immediate swelling and eventual bursting of the cell. In hypertonic solutions, RBCs shrink and crenate in shape due to the water efflux. The cellular response at different osmolarities changes the RBC deformability as well, which determines their ability to transport oxygen through the microvasculature. The effects of osmotic stress on the mechanical properties of RBCs have not been systematically investigated and the overall consequences to RBC deformability remain unknown. Using DPM, we measured the membrane fluctuations of RBCs at different osmolarities [82]. To modulate the osmolarity of the medium, we prepared RBC suspensions (106 cell/ml) with 11 different osmolarities ranging from 100 to 600 mOsm/kg H2O. Fresh blood samples were collected and diluted 1:5 in Hank’s buffer saline solution (HBSS), and were then immediately centrifuged at 2,000 g at 5 C for 10 min to separate RBCs from plasma. The RBCs were washed three times and then resuspended in the given sodium chloride (NaCl) solution. The NaCl solutions contained increasing concentrations of NaCl (0.3–1.8%), which correspond to suspension osmolarities ranging from 100 to 600 mOsm/kg. Then, using DPM we extracted quantitatively optical phase shifts f(x,y,t) associated with the cells, at spatial and temporal resolutions of nanometer and millisecond, respectively [11, 63]. The cell thickness profile is obtained from the optical phase shift as

294

Y. Park et al.

h(x,y,t) ¼ (l/2pDn)f(x,y,t). Since the refractive index difference, Dn, is mainly contributed from homogeneous Hb solution in cytoplasm, the integration of optical phase shiftsRover cell area, i.e. the dry mass [83], is related to the volume of RBC as, Volume ¼ hfðx; yÞidA. The refractive index difference, Dn at different osmolarities can be calculated using the data in the literature [84, 85]. Thickness profiles and horizontal cross-sections of RBCs in hypotonic, isotonic, and hypertonic medium are shown in Fig. 13.9. Different osmolarities of extracellular medium result in significant changes in RBC shape. In hypotonic medium (100 mOsm/kg), RBCs are swollen due to water influx, but still maintain the dimpled region in the center. At osmotic pressures below 100 mOsm/kg, most of RBCs are lysed. In the hypertonic case (600 mOsm/kg) RBCs shrink due to water efflux. In media with lower osmolarity than normal (295 mOsm/kg), the projected area of RBCs is decreased, which is indicative of an increase in the lipid layer tension caused by cell swelling (Fig. 13.10a). In contrast, the projected area of RBCs in hypertonic medium did not show significant differences compared to the normal RBCs. We extracted the mean corpuscular hemoglobin (MCH), the total amount of Hb in the cell, as the product of the Hb concentration and cytoplasmic volume (Fig. 13.10b). The Hb concentration was retrieved using its known relationship with the refractive index [86]. MCH maintains constant values at different osmolarities (Fig. 13.10c), which is consistent with impermeability of RBC membrane to large Hb protein. In order to address the progressive alterations of RBC mechanical properties due to osmotic pressure, we quantitatively measured dynamic membrane fluctuations of [μm]

b

a

4

c

3 2

Isotonic 300

Hypotonic 100

h [μm]

4

d

3

100 mOsm/kg

4

e

3

Hypertonic 600 4

300 mOsm/kg

2

2

2

1

1

1

0

0

2

4

r [μm]

6

0

0

f

3

2

4

r [μm]

6

0

0

1 0

600 mOsm/kg

2

4

6

r [μm]

Fig. 13.9 Shapes of RBCs exposed to different osmolarities. (a–c) Topography of a RBC exposed to hypotonic (100 mOsm/kg) (a), isotonic (300 mOsm/kg) (b), and hypertonic (800 mOsm/kg) (c). Scale bar is 2 mm. (d–f) Averaged shape as a function of a distance from the center of cells exposed to hypotonic (100 mOsm/kg) (d), isotonic (300 mOsm/kg) (e), and hypertonic (800 mOsm/kg) (f). Thick lines show the average value and the areas represent standard deviation for 20 RBCs

Optical Sensing of Red Blood Cell Dynamics

b

40

20

1.40 120 100

Refractive index 1.38 Volume 1.36 1.34

80 0

100 200 300 400 500 600 Osmorality [mOsm/kg]

c MCH [pg]

1.42

140

60 Cell volume [fl]

Projected Area [μm2]

a

295

100 200 300 400 500 600 Osmorality [mOsm/kg]

d

80

30

60

20

40

10

20

0

100 200 300 400 500 600 Osmorality [mOsm]

Refractive index

13

0

100 200 300 400 500 600 Osmorality [mOsm/kg]

Fig. 13.10 (a) Projected areas of RBCs at different osmolarities. (b) Averaged refractive index of RBC cytoplasm. Volumes of RBCs are calibrated from refs. [84, 85]. (c) Mean corpuscular hemoglobin (MCH) at different osmolarities. (d) RMS fluctuations in RBC membrane. Each box represents standard deviation for 20 RBCs and wisker for min-max data

RBCs. The map of instantaneous displacement of cell membrane fluctuation was obtained by subtracting the time-averaged cell profile from each instantaneous topography map in the series such that Dhðx; y; tÞ ¼ hðx; y; tÞ hhðx; yÞi. The RMS is shown for different osmotic condition (Fig. 13.10d). These results show that the maximum membrane fluctuations occur around 300 mOsm/kg, which is the normal physiological blood osmolarity. The decreased deformability of RBCs in both hypo- and hypertonic conditions is consistent with a variety of experimental techniques including laser scattering experiment, cell elongation measurement, blood filtration experiment [87, 88]. We retrieved the mechanical properties of RBCs by analyzing the spatial correlations of the out-of-plane membrane fluctuations via the viscoelastic continuum model described above [73]. The retrieved three mechanical properties of RBCs in different osmolarities are shown in Fig. 13.11. Shear modulus of spectrin network m in hypotonic medium shows a significant increase compared to the normal and hypertonic cases. Interestingly, m does not change above 300 mOsm/kg. This suggests that the spectrin network becomes stiff (m 12 mN/m) when it stretches out due to the volume expansion of RBCs. On the other hand, the membrane bending modulus does not show a significant dependence on osmolarities (k 5 kBT). This indicates that the composition of the RBC bilayer does not change with osmotic stress.

296 15

60 50

10

40 30

5

η [mPa.s]

μ [μN/m], κ [xkBT]

Fig. 13.11 Shear modulus m, cytosol viscosity , and bending modulus k vs. different osmotic pressure. Error bar represents standard deviation for 20 RBCs. The shaded area is normal physiological osmotic pressure (270–305 mOsm/kg)

Y. Park et al.

20 μ κ η 100 200 300 400 500 600 700 800 Osmolarity [mOsm/kg]

10 0

We found that the viscosities of the cytosol c increase progressively with increasing osmolarity of extracellular medium. Exposed to the hyperosmotic medium, water molecules in RBC cytoplasm go out through aquaporin-1, a membrane embedded channel facilitating the movement of water molecules across lipid membrane. Thus, an increase in cytoplasmic hemoglobin (Hb) concentration results in an increase in cytosolic viscosity [89]. This suggests the RBCs may have a damping mechanism that yields protection from the high viscosity of cytoplasm when RBCs pass through the small capillaries in renal medulla, part of the kidney where the osmolarity can be as high as 1,200 mOsm/kg. The decreased cell volume at constant surface area would facilitate the passage of the RBCs through narrow capillaries. However, RBCs passing through the capillaries in renal medulla may lose elasticity and tend to become fragile. The mechanism of how RBCs maintain their elasticity in hypertonic conditions is non-trivial to explain. If the cytoplasm of RBC becomes more viscous, it may prevent RBCs from lysis by absorbing external forces effectively and making the RBCs less deformable.

13.3.5 Malaria Effects When the malaria parasite Plasmodium falciparum invades the RBC and multiplies inside the RBC, it causes structural, biochemical, and mechanical changes to the host RBCs. Major structural changes include the growth of parasitophorous vacuole in cytoplasm of host RBCs, loss of RBC volume, and the appearance of small protrusions, called “knobs,” on the membrane surface [90]. A considerable amount of Hb is digested by parasites during intra-erythrocytic development and converted into insoluble polymerized forms of heme, known as hemozoin [91, 92]. Hemozoin appears as brown crystals in the parasitophorous vacuole in later maturation stages of P. falciparum invaded human RBCs (Pf-RBCs). Two major mechanical

13

Optical Sensing of Red Blood Cell Dynamics

297

modifications are loss of RBC deformability [93–95] and increased cytoadherence of the invaded RBC membrane to vascular endothelium and other RBCs [96]. These changes lead to sequestration of RBCs in microvasculature in the later stages of parasite development, which is linked to vital organ dysfunction in severe malaria. In the earlier stage, where deformability occurs, Pf-RBCs continue to circulate in the blood stream despite infection. In order to study these modifications, we investigated membrane fluctuations in Pf-RBCs. Proteins transported from invading organisms, such as the virulent malaria-inducing parasite P. falciparum, to specific binding sites in the spectrin network are considered to introduce significant alterations to RBC membrane dynamics and mechanical response [93, 95, 97–100]. These changes may ultimately provide insights into the pathogenesis of malaria, as the parasite alters the biophysical properties of RBCs during its intra-erythrocyte stage that lasts up to 48 h. Despite the broad realization that membrane fluctuations provide information on critical interactions among sub-cellular structures, mechanical stress, and biochemical links between the cell interior and the external environment, systematic experiments of cell membrane dynamics, over the physiologically relevant temperature range, have not been performed. To quantify the progressive alterations to RBC membrane fluctuations and mechanical responses due to parasitization by P. falciparum, we employed DPM. To investigate morphological changes of Pf-RBCs, we measured the instantaneous thickness profile, h(x,y,t0) of cells [63]. Figure 13.12a–d show topographic images of healthy and Pf-RBCs at all stages of development. The effective stiffness map of the cell, ke(x,y) is obtained at each point on the cell, assuming an elastic restoring force associated with the membrane:

Fig. 13.12 Topographic images and effective elastic constant maps of Pf-RBCs. (a, e) Healthy RBC. (b, f ) Ring stage. (c, g) Trophozoite stage. (d, h) Schizont stage. The topographic images in (a–d) are the instant thickness map of Pf-RBCs. The effective elastic constant maps were calculated from the root-mean-squared displacement of the thermal membrane fluctuations in the Pf-RBC membranes. Black arrows indicate the location of P. falciparum, and the gray arrows the location of hemozoin (scale bar, 1.5 mm). Reproduced with permission from ref. [63]. Copyright (2008) NAS

298

Y. Park et al.

D E ke ðx; yÞ ¼ kB T= Dhðx; yÞ2 ;

(13.10)

D E where Dhðx; yÞ2 is the mean-squared displacement. Representative ke maps of RBCs at the indicated stages of parasite development are shown in Fig. 13.12e–g. The map of instantaneous displacement of cell membrane fluctuation, Dh(x,y,t), was obtained by subtracting time-averaged cell shape from each thickness map in the series. A histogram showing membrane displacements for all parasite stages is shown in Fig. 13.13a. RBC deformability is sensitive to membrane stiffness. Our DPM experiments provide quantitative information from which in-plane shear modulus of RBC membrane with attached spectrin cytoskeleton could be determined. The in-plane shear modulus G can be obtained using the Fourier-transformed Hamiltonian (strain energy) and equipartition theorem [101]: Gﬃ

kB T lnðA=aÞ ; 3p Dh2t

(13.11)

where A is the projected diameter of RBC (~8 mm), and a is the minimum spatial wavelength measured by DPM (~0.5 mm). The tangential component of displacement in membrane fluctuations, Dh2t , was decoupled from axial membrane fluctuation Dh2 using the angle a between the direction of Dht and the normal direction of membrane as illustrated in Fig. 13.13b inset. The angle a is extracted from cell

b

0.014 0.012 0.01 0.008

Schizont

×1.5

Trophozoite Ring Healthy

FWHM fluctuation 0.006 0.004

Shear modulus (μN/m)

Probability density,P(Δh)

a

100 50

DPM Optical tweezers

10 Δ ht

5

Δh α

Δhn

0.002 0

−200 −150 −100 −50

0

50 100 150 200

1

Healthy

Ring Trophozoite

Schizont

Membrane fluctuation displacement, Δh (nm)

Fig. 13.13 Membrane fluctuations and in-plane shear modulus at different intra-erythrocytic stages of Pf-RBCs. (a) Histograms of cell-thickness fluctuation of Pf-RBCs. (Histogram of the schizont stage is scaled down by a factor of 1.5.) (b) In-plane shear modulus of the RBC membrane vs. developmental stage of Pf-RBCs. The in-plane shear modulus was calculated from the in-plane membrane displacement, Dhn over the rims of RBCs. Also shown for comparison are the estimated from optical tweezers (22). (Inset) Illustration of RBC and membrane fluctuations: Dh, thickness fluctuations measured by DPM; Dht, in-plane membrane displacement; Dhn, out-of-plane membrane displacement, and a, the angle between Dh and Dht. The measurements were performed at the room temperature (23 C) and for each group 20 samples were tested. Reproduced with permission from ref. [63]. Copyright (2010) NAS

13

Optical Sensing of Red Blood Cell Dynamics

299

topographical information measured by DPM. Dh2t was calculated and averaged along the circumference of cell, from which in-plane shear modulus G was calculated. We determine that G ¼ 5.5 0.8 mN/m for healthy RBCs (Fig. 13.13b), which compares well with independent modulus measurements, extracted for healthy RBCs from micropipette aspiration and optical tweezers [76, 97, 98]. The modulus for ring (G ¼ 15.3 5.4 mN/m), trophozoite (G ¼ 28.9 8.2 mN/m) and schizont (G ¼ 71.0 20.2 mN/m) stages is in good quantitative agreement with that inferred from large-deformation stretching with optical tweezers of Pf-RBCs over all stages of parasite maturation [97].

13.3.6 Adenosine Triphosphate (ATP) Effects Although RBC membrane dynamics has been explored extensively, no definitive experiment has determined whether flickering is a purely thermally driven phenomena or whether flickering requires active contributions. First observed a century ago, its origin is generally believed to stem from thermal forces [17, 102]. Different interference microscopic techniques have been employed to study membrane fluctuations and mechanical properties assuming Brownian dynamics [19, 66]. In contrast, a technique that qualitatively measured the local fluctuations of RBC membranes, reported a correlation between the ATP concentration and the fluctuation amplitude [15]. However, more recent experimental works found no relationship between ATP depletion and membrane fluctuations [103, 104]. Theoretically, RBC membrane fluctuations have been traditionally studied using models of thermally driven equilibrium systems [17, 19]. A more recent theoretical model [13, 105], validated by simulation [106, 107], showed that local breaking and reforming of the spectrin network can result in enhanced fluctuations. In the following section we show direct, full-field, and quantitative measurements of ATP effects on RBC membrane morphology and fluctuations through DPM (for details, see ref. [108]). By extracting the optical path-length shifts produced across the cell, we measured cell thickness with nanometer sensitivity and millisecond temporal resolution. RBC samples were prepared under four different conditions: healthy RBCs, and irreversibly ATP-depleted group, metabolically ATP-depleted group, and ATP-repleted group. After collection, a group of healthy RBCs was minimally prepared. For RBCs in the irreversibly ATP-depleted group, the cytoplasmic pool of ATP was depleted by inosine and iodoacetamide. For the metabolically ATP-depleted group, healthy RBCs were incubated in a glucose-free medium for 24 h. For RBCs in the ATP-repleted group, cytoplasmic ATP was first metabolically depleted, and then regenerated through the addition of D-glucose. We first address the effects of ATP on the morphologies of RBC membranes. From the measured cell thickness profiles at a given time t, h(x,y,t), we calculated time-averaged heights hhðx; yÞi and observed the characteristic biconcave shape for healthy RBCs (Fig. 13.14a–d). When ATP was depleted, for both the irreversibly and the metabolically depleted groups, we observed loss of

300

Y. Park et al.

Fig. 13.14 Correlation between biconcave shape and enhanced membrane fluctuations. (a–d) Averaged height as a function of the distance from the center of cells for healthy RBCs (a), for RBCs in the irreversibly ATP-depleted group (b), for RBCs in the metabolically ATP-depleted group (c), and for RBCs in which ATP was reintroduced to the metabolically ATP-depleted group (d), respectively. (e–h) Averaged squared height fluctuations as a function of the distance from the center of cells in Fig. 13.14a–d, respectively. Thick lines show the average value and the areas represent standard deviation for 40 RBCs. Reproduced with permission from ref. [108]. Copyright (2010) NAS

biconcave shape and echinocyte shape transformation. Reintroducing ATP resulted in the recovery of biconcave shape. This shows that ATP is crucial to maintaining biconcave shape of RBCs [109]. In order to probe dynamic membrane fluctuations, we employed DPM and analyzed the membrane displacement map by subtracting the averaged shape from the cell thickness map Dhðx; y; tÞ ¼ hðx; y; tÞ hhðx; yÞi. Compared to healthy RBCs, the fluctuation amplitudes were decreased in both ATP-depleted groups. Reintroducing ATP, however, increased the fluctuation amplitudes to healthy RBC levels. We calculated the RMS displacement of membrane fluctuations, which covers the entire cell area for 2 s at 120 frame/s (Figs. 13.14e–h and 13.15). The RMS displacement of healthy RBCs is 41.5 5.7 nm. Fluctuations significantly decreased to 32.0 7.8 nm and 33.4 8.7 nm in both the irreversibly and metabolically ATP-depleted groups, respectively. However, the fluctuations in the ATPrepleted group returned to the level of healthy RBCs (48.4 10.2 nm). This is in agreement with an earlier report using the point measurement technique [15]. Even though we showed that the membrane fluctuations indeed decrease in the absence of ATP, this result does not answer the question – whether ATP drives “active” non-equilibrium dynamics, or simply modifies mechanical properties of the membrane? Of course, the two different situations can give rise to fundamentally different dynamics: (a) fluctuations exhibit out-of-equilibrium, or (b) the equilibrium Gaussian statistics is preserved. In order to answer this question, we calculated the non-Gaussian parameter, k, for the membrane fluctuations (Fig. 13.16). Please note that in this chapter k is non-Gaussian parameter, which is different from bending modulus used in earlier section. Theoretically, k ¼ 2 for purely thermally driven Gaussian motion and increases above two for active nonequilibrium dynamics [14]. For healthy RBCs, the average value of k was 2.8, which shows that membrane fluctuations contain non-equilibrium dynamic

Optical Sensing of Red Blood Cell Dynamics

Fig. 13.15 RMS displacements of membrane fluctuations for different ATP conditions: healthy RBCs, irreversibly ATP-depleted RBCs, metabolically ATPdepleted RBCs, and RBCs in which ATP was reintroduced to metabolically ATP-depleted RBCs. Each symbol represents an individual RBC and the horizontal line is the mean value. Reproduced with permission from ref. [108]. Copyright (2010) NAS

301

80

* * *

60 √〈Δh2〉 (nm)

13

40

20

0

* p < 0.001 Healthy

−ATP −ATP (irreversible) (metabolic)

+ATP

Fig. 13.16 Non-equilibrium dynamic in RBC membranes. Averaged non-Gaussian parameter vs. a lag time, Dt, and a spatial frequency, q, for membrane fluctuation in healthy RBC (a), irreversible ATP depletion group (b), metabolic ATP depletion group (c), and after reintroducing ATP to metabolic depletion group (d), respectively. N ¼ 40 RBCs per each group. Reproduced with permission from ref. [108]. Copyright (2010) NAS

302

Y. Park et al.

components, particularly on short length and time scales (q > 5 rad/mm and Dt < 0.5 s). With depletion of ATP, k decreased to 2, as expected in purely thermally-driven dynamics (the average values of k were 2.06 and 2.19 for the irreversibly depleted and metabolically depleted ATP groups, respectively). Reintroducing ATP increased k to healthy RBC levels (average value k ¼ 2.98). Our data clearly illustrate that active, metabolic energy from ATP enhances the RMS displacements by 44.9%. This measured value is lower than predicted by a theoretical model, where an increase of at least 100% was expected [14]. However, this disparity can be explained by recognizing that the ATP effect is more significant at large qvalues, comparable with the size of the spectrin network [13]. For example, the ATPmediated RMS displacement at q ¼ 17 0.5 rad/mm showed an increase of 143% compared with the thermal components. Thus, in our overall assessment that includes all spatial frequencies, ATP enhancement is likely to be underestimated. In order to study further spatial aspects of active motions, we analyzed the morphologies and fluctuations for RBCs in a polar coordinate system with its origin at cell center. Assuming cylindrical symmetry, the average height of the RBC membrane,hhðrÞi, and the membrane mean-squared displacements, h2 ðrÞ , are shown as functions of the radial distance, r (Fig. 13.14). In healthy RBCs, the membrane fluctuations are enhanced and strongly localized at the outer region, while both ATP depletion groups showed little variation in membrane fluctuations over the cell surface. Remarkably, reintroducing ATP restores not only the biconcave shape, but also causes enhanced fluctuations in the outer area. This is striking because continuum models predict a stronger restoring force and a decreased fluctuation amplitude in regions of high membrane curvature [110]. This result shows that active contributions are spatially inhomogeneous, and that they are correlated with the maintenance of the biconcave shape. It also helps explain different mechanistic inferences reported in the literature from prior measurements of membrane fluctuations [15, 103, 111]. Probing the edge shape of RBCs alone does not capture ATP-dependent enhanced fluctuations [103] since they are localized on the outer cell. Dark-field microscopy, which qualitatively measures the averaged dynamics of the RBC surface, can however measure ATP-dependence [15, 111]. Other cytoskeleton models incorporating actin, microtubules, and motor proteins such as myosin have demonstrated active motion [112]. However, this cannot be the case for the ATP-enhanced fluctuations in RBC because motor proteins are absent here. Then, how can the RBC exhibit active dynamics? To address this question, we further analyzed the results in the context of RBC cytoskeletal structure. The nonGaussian parameter, k, at short time delays, was plotted as a function of spatial distance L ¼ 2p/q (Fig. 13.17). Interestingly, in the presence of ATP, k showed distinct peaks at specific distances (L ¼ 361, 512, 680, 860, and 1,030 nm); these peaks are equally spaced at 167 10 nm. These results indicate that ATP-dependent enhanced fluctuations are correlated with the network structure of the underlying cytoskeleton. Considering the roughly hexagonal lattice of spectrin network, these peaks can be related to the dynamic remodeling of spectrin network by ATP. Possible elements responsible for this remodeling are the junctional complexes of the spectrin

Optical Sensing of Red Blood Cell Dynamics

Fig. 13.17 Non-Gaussian parameter at short time delay (Dt <0.1 s) as a function of spatial wavelength. The presence of ATP lead to nonthermal fluctuations, especially at L ¼ 361, 512, 680, 860, and 1,030 nm. Reproduced with permission from ref. [108]. Copyright (2010) NAS

3.0 2.8

303

1

Healthy RBCs ATP-(irreversible)

2

ATP-(natural)

3

ATP+ (Glucose)

4 5

κ

13

2.6 2.4 2.2 2.0 0.0

0.5

1.0

Λ

1.5

2.0

2.5

network which consists of a complex of six spectrin polymers joined by short actin segments and protein-4.1. It was proposed that this ATP-induced remodeling takes the form of local associations and dissociations of spectrin filaments within the network, or between the cytoskeleton and the lipid membrane [9, 13]. Both processes result in the formation of a structural defect in the hexagonal network, of size ~2D,where D is a distance between neighboring junctions. This remodeling of the cytoskeletal attachment causes a local release of the cytoskeleton-induced membrane tension, and results in local bilayer deformation [9, 13]. The length-scale of this local, ATP-induced, bilayer deformation will therefore be in multiples of the junction spacing, 2D. From the data we find D ~ 83.5 5 nm, which is in good agreement with the separation of the junctions complexes measured by electron microscopy [113]. The question then arises, how can ATP cause this dynamic remodeling of the cytoskeletal attachment? This may be related to protein phosphorylation powered by ATP, which is one of the physiological processes that controls membrane stability. One possible candidate is the phosphorylation of the phosphoinositides (PI) because it consumes more ATP than the combined phosphorylation of all the other membrane proteins [114]. PIP2, phosphorylated from PI by ATP, is thought to play an important role in modulating the binding of the lipid bilayer to the cytoskeleton by altering the protein interactions that comprise the junctional complex at spectrin tetramer ends [115]. For example, PIP2 strengthens the binding affinity of protein-4.1 to glycophorin C (GP) [116]. Furthermore, PIP2 dephosphorylation results in a decreased affinity for GP binding, and a subsequent detachment from spectrin network; the latter can result in increased membrane fluctuations since tension applied to the bilayer by spectrin network is locally and transiently released. In the absence of ATP, this dynamic remodeling may not occur, and thus RBCs exhibit only thermally driven membrane fluctuations. Our results provide further experimental evidence for the metabolism-dependent shape transformation; we show that ATP-dependent transient binding of junctional

304

Y. Park et al.

complexes are localized over the cell outer area, and that the spectrin network should therefore exert a lower tension on the membrane. We also note that in the absence of ATP, the shapes of RBCs are similar to those found in patients with hereditary elliptocytosis, where GP does not properly interact with protein 4.1, resulting in the lack of biconcave shape and deformability [117, 118]. This dynamic remodeling of the spectrin network also offers a possible explanation for the observed metabolic dependence of red cell deformability [119]. Taken together, we have shown that the biconcave shape and nonequilibrium dynamics in the membrane are both consequences of the same biochemical activity: the dissociations of the cytoskeleton at the spectrin junctions, powered by ATP metabolism.

13.4

Summary and Outlook

In summary, we showed that, quantitative phase imaging is capable of measuring nanoscale motions in RBCs, which in turn report on the cell stiffness. Thus, we developed diffraction phase microscopy (DPM) to obtain highly sensitive membrane displacements motions at the millisecond scales and measured the RBC microrheology during various states in health and disease. We found that the cell rheology is affected by morphology, osmotic conditions, energetic (via ATP), and disease state (i.e. malaria). These recent results furthered our understanding of the interplay between the spectrin network and lipid bilayer. We believe that in the future, this measurement approach can even be applied for clinical problems, either as a blood screening procedure, or for fast, quantitative testing of various drugs. Acknowledgements The authors are grateful for the mentoring provided by the late Michael Feld. The authors acknowledge fruitful collaborations with the groups lead by Subra Suresh, Alex Levine, Nir Gov, and Sam Safran.

References 1. Mohandas N and Gallagher P G (2008) Red cell membrane: past, present, and future. Blood 112:3939–3948 2. Cotran R, Kumar V, Collins T et al (2004) Robbins pathologic basis of disease. Philadelphia: WB Saunders 3. Bao G and Suresh S (2003) Cell and molecular mechanics of biological materials. Nature Mat 2:715–725 4. Fournier J B, Lacoste D, and Rapha E (2004) Fluctuation spectrum of fluid membranes coupled to an elastic meshwork: jump of the effective surface tension at the mesh size. Phys Rev Lett 92:18102 5. Discher D E, Mohandas N, and Evans E A (1994) Molecular maps of red cell deformation: hidden elasticity and in situ connectivity. Science 266:1032–1035 6. Engelhardt H, Gaub H, and Sackmann E (1984) Viscoelastic properties of erythrocyte membranes in high-frequency electric fields. Nature 307:378–380

13

Optical Sensing of Red Blood Cell Dynamics

305

7. Puig-de-Morales-Marinkovic M, Turner K T, Butler J P et al (2007) Viscoelasticity of the human red blood cell. Am J Physiol Cell Physiol 293:597–605 8. Browicz T (1890) Further observation of motion phenomena on red blood cells in pathological states. Zbl med Wissen 28: 625–627 9. Gov N, Zilman A G, and Safran S (2003) Cytoskeleton confinement and tension of red blood cell membranes. Phys Rev Lett 90:228101 10. Zilker A, Ziegler M, and Sackmann E (1992) Spectral-analysis of erythrocyte flickering in the 0.3-4-mu-m-1 regime by microinterferometry combined with fast image-processing. Phys Rev A 46:7998–8002 11. Popescu G, Ikeda T, Dasari R R et al (2006) Diffraction phase microscopy for quantifying cell structure and dynamics. Opt Lett 31:775–777 12. Tuvia S, Levin S, Bitler A et al (1998) Mechanical fluctuations of the membrane-skeleton are dependent on F-Actin ATPase in human erythrocytes. Proc Natl Acad Sci USA 141:1551–1561 13. Gov N S and Safran S A (2005) Red blood cell membrane fluctuations and shape controlled by ATP-induced cytoskeletal defects. Biophys J 88:1859 14. Lawrence C L L, Gov N, and Brown F L H (2006) Nonequilibrium membrane fluctuations driven by active proteins. J Chem Phys 124:074903 15. Tuvia S, Levin S, Bitler A et al (1998) Mechanical fluctuations of the membrane-skeleton are dependent on F-actin ATPase in human erythrocytes. J Cell Biol 141:1551–1561 16. Li J, Dao M, Lim C T et al (2005) Spectrin-level moddquo;eling of the cytoskeleton and optical tweezers stretching of the erythrocyte. Biophys J 88:3707–3719 17. Brochard F and Lennon J F (1975) Frequency spectrum of the flicker phenomenon in erythrocytes. J Phys 36:1035–1047 18. Kaizuka Y and Groves J T (2006) Hydrodynamic damping of membrane thermal fluctuations near surfaces imaged by fluorescence interference microscopy. Phys Rev Lett 96:118101 19. Zilker A, Engelhardt H, and Sackmann E (1987) Dynamic reflection interference contrast (Ric-) microscopy – a new method to study surface excitations of cells and to measure membrane bending elastic-moduli. J Phys 48:2139–2151 20. Zernike F (1942) Phase contrast, a new method for the microscopic observation of transparent objects part II. Physica 9:974–986 21. Zidovska A and Sackmann E (2006) Brownian motion of nucleated cell envelopes impedes adhesion. Phys Rev Lett 96:048103 22. Popescu G (2008) Quantitative phase imaging of nanoscale cell structure and dynamics. In: B. P. Jena, ed., Methods in Cell Biology. San Diego: Elsevier 23. Yang C, Wax A, Hahn M S et al (2001) Phase-referenced interferometer with subwavelength and subhertz sensitivity application to the study of cell membrane dynamics. Opt Lett 26:1271–1273 24. Yang C H, Wax A, Georgakoudi I et al (2000) Interferometric phase-dispersion microscopy. Opt Lett 25:1526–1528 25. Choma M A, Ellerbee A K, Yang C H et al (2005) Spectral-domain phase microscopy. Opt Lett 30:1162–1164 26. Joo C, Akkin T, Cense B et al (2005) Spectral-domain optical coherence phase microscopy for quantitative phase-contrast imaging. Opt Lett 30:2131–2133 27. Fang-Yen C, Chu M C, Seung H S et al (2004) Noncontact measurement of nerve displacement during action potential with a dual-beam low-coherence interferometer. Opt Lett 29:2028–2030 28. Akkin T, Dave D P, Milner T E et al (2004) Detection of neural activity using phase-sensitive optical low-coherence reflectometry. Opt Express 12:2377–2386 29. Rylander C G, Dave D P, Akkin T et al (2004) Quantitative phase-contrast imaging of cells with phase-sensitive optical coherence microscopy. Opt Lett 29:1509–1511

306

Y. Park et al.

30. Zicha D and Dunn G A (1995) An image-processing system for cell behavior studies in subconfluent cultures. J Microsc 179:11–21 31. Dunn G A, Zicha D, and Fraylich P E (1997) Rapid, microtubule-dependent fluctuations of the cell margin. J Cell Sci 110:3091–3098 32. Zicha D, Genot E, Dunn G A et al (1999) TGF beta 1 induces a cell-cycle-dependent increase in motility of epithelial cells. J Cell Sci 112:447–454 33. Paganin D and Nugent K A (1998) Noninterferometric phase imaging with partially coherent light. Phys Rev Lett 80:2586–2589 34. Allman B E, McMahon P J, Tiller J B et al (2000) Noninterferometric quantitative phase imaging with soft x rays. J Opt Soc Am A Opt Image Sci Vis 17:1732–1743 35. Bajt S, Barty A, Nugent K A et al (2000) Quantitative phase-sensitive imaging in a transmission electron microscope. Ultramicroscopy 83:67–73 36. Mann C J, Yu L F, Lo C M et al (2005) High-resolution quantitative phase-contrast microscopy by digital holography. Opt Express 13:8693–8698 37. Marquet P, Rappaz B, Magistretti P J et al (2005) Digital holographic microscopy: a noninvasive contrast imaging technique allowing quantitative visualization of living cells with subwavelength axial accuracy. Opt Lett 30:468–470 38. Iwai H, Fang-Yen C, Popescu G et al (2004) Quantitative phase imaging using actively stabilized phase-shifting low-coherence interferometry. Opt Lett 29:2399–2401 39. Popescu G, Deflores L P, Vaughan J C et al (2004) Fourier phase microscopy for investigation of biological structures and dynamics. Opt Lett 29:2503–2505 40. Ikeda T, Popescu G, Dasari R R et al (2005) Hilbert phase microscopy for investigating fast dynamics in transparent systems. Opt Lett 30:1165–1168 41. Popescu G, Ikeda T, Best C A et al (2005) Erythrocyte structure and dynamics quantified by Hilbert phase microscopy. J Biomed Opt Lett 10:060503 42. Huang D, Swanson E A, Lin C P et al (1991) Optical coherence tomography. Science 254:1178–1181 43. deBoer J F, Milner T E, vanGemert M J C et al (1997) Two-dimensional birefringence imaging in biological tissue by polarization-sensitive optical coherence tomography. Opt Lett 22:934–936 44. Hitzenberger C K and Fercher A F (1999) Differential phase contrast in optical coherence tomography. Gastrointest Endosc 24:622–624 45. Park J, Kemp N J, Milner T E et al (2003) Analysis of the phase retardation in the retinal nerve fiber layer of cynomolus monkey by polarization sensitive optical coherence tomography. Lasers Surg Med 55:55 46. Choma M A, Yang C H, and Izatt J A (2003) Instantaneous quadrature low-coherence interferometry with 3 x 3 fiber-optic couplers. Opt Lett 28:2162–2164 47. Youn J I, Akkin T, Wong B J F et al (2003) Electrokinetic measurements of cartilage measurements of cartilage using differential phase optical coherence tomography. Lasers Surg Med 56:56 48. Kim J, Telenkov S A, and Milner T E (2004) Measurement of thermo-refractive and thermoelastic changes in a tissue phantom using differential phase optical coherence tomography. Lasers Surg Med 8:8 49. Wojtkowski M (2010) High-speed optical coherence tomography: basics and applications. Appl Opt 49:30–61 50. Dunn G and Zicha D (1998) Using the DRIMAPS system of interference microscopy to study cell behavior. In Cell biology: a laboratory handbook 44–53 J. Celis, Ed., Academic press 51. Dunn G A and Zicha D (1995) Dynamics of fibroblast spreading. J Cell Sci 108:1239–1249 52. Gureyev T E, Roberts A, and Nugent K A (1995) Phase retrieval with the transport-of-intensity equation – matrix solution with use of Zernike polynomials. J Opt Soc Am A Opt Image Sci Vis 12:1932–1941

13

Optical Sensing of Red Blood Cell Dynamics

307

53. Gureyev T E, Roberts A, and Nugent K A (1995) Partially coherent fields, the transportof-intensity equation, and phase uniqueness. J Opt Soc Am A Opt Image Sci Vis 12:1942–1946 54. Goodman J W and Lawrence R W (1967) Digital image formation from electronically detected holograms. Appl Phys Lett 11:77 55. Gabor D (1948) A new microscopic principle. Nature 161:777 56. Goodman J (2005) Introduction to Fourier optics. Englewood Cliffs: Roberts & Company 57. Yamaguchi I and Zhang T (1997) Phase-shifting digital holography. Opt Lett 22:1268–1270 58. Carl D, Kemper B, Wernicke G et al (2004) Parameter-optimized digital holographic microscope for high-resolution living-cell analysis. Appl Opt 43:6536–6544 59. Lue N, Choi W, Popescu G et al (2007) Quantitative phase imaging of live cells using fast Fourier phase microscopy. Appl Opt 46:1836 60. Park Y K, Popescu G, Badizadegan K et al (2006) Diffraction phase and fluorescence microscopy. Opt Exp 14:8263 61. Park Y K, Popescu G, Badizadegan K et al (2006) Diffraction phase and fluorescence microscopy. Opt Express 14:8263–8268 62. Takeda M, Ina H, and Kobayashi S (1982) Fourier-transform method of fringe-pattern analysis for computer-based topography and interferometry. J Opt Soc Am 72:156–160 63. Park Y K, Diez-Silva M, Popescu G et al (2008) Refractive index maps and membrane dynamics of human red blood cells parasitized by Plasmodium falciparum. Proc Natl Acad Sci U S A 105:13730 64. Park Y, Yamauchi Y, Choi W, Dasari R and Feld M (2009) Spectroscopic phase microscopy for quantifying hemoglobin concentrations in intact red blood cells. Optics Letters 34: 3668–3670 65. Angelova M I, Soleau S, Meleard P et al (1992) Preparation of giant vesicles by external AC electric fields. Kinetics and applications. Prog Colloid Polym Sci 89:122 66. Popescu G, Ikeda T, Goda K et al (2006) Optical measurement of cell membrane tension. Phys Rev Lett 97:218101 67. Evans E and Rawicz W (1990) Entropy-driven tension and bending elasticity in condensedfluid membranes. Phys Rev Lett 64:2094–2097 68. Best C A, Cluette-Brown J E, Teruya M et al (2003) Red blood cell fatty acid ethyl esters: a significant component of fatty acid ethyl esters in the blood. J Lipid Res 44:612–620 69. Lim H W G, Wortis M, and Mukhopadhyay R (2002) Stomatocyte-discocyte-echinocyte sequence of the human red blood cell: evidence for the bilayer-couple hypothesis from membrane mechanics. Proc Natl Acad Sci U S A 99:16766–16769 70. Gov N, Zilman A, and Safran S (2003) Cytoskeleton confinement of red blood cell membrane fluctuations. Biophys J 84:486A 71. Discher D E, Boal D H, and Boey S K (1998) Simulations of the erythrocyte cytoskeleton at large deformation. II. Micropipette aspiration. Biophys J 75:1584–1597 72. Park Y, Best C, Badizadegan K et al (2010) Measurement of red blood cell mechanics during morphological changes. Proc Natl Acad Sci U S A 107:6731 73. Kuriabova T and Levine A J (2008) Nanorheology of viscoelastic shells: applications to viral capsids. Phys Rev E 77:031921 74. Chaikin P M and Lubensky T C (1995) Principles of condensed matter physics. Cambridge: Cambridge University Press 75. Kuriabova T and Levine A (2008) Nanorheology of viscoelastic shells: applications to viral capsids. Phys Rev E 77:31921 76. Waugh R and Evans E A (1979) Thermoelasticity of red blood cell membrane. Biophys J 26:115–131 77. Dao M, Lim C T, and Suresh S (2003) Mechanics of the human red blood cell deformed by optical tweezers. J Mech Phys Solids 51:2259–2280 78. Tang W and Thorpe M F (1988) Percolation of elastic networks under tension. Phys Rev B 37:5539

308

Y. Park et al.

79. Bursac P, Lenormand G, Fabry B et al (2005) Cytoskeletal remodelling and slow dynamics in the living cell. Nat Mater 4:557–561 80. Siegel D (1985) Partial purification and characterization of an actin-bundling protein, band 4.9, from human erythrocytes. J Cell Biol 100:775–785 81. Marko J F and Siggia E D (1995) Stretching DNA. Macromolecules 28:8759–8770 82. Park Y, Best C, Kuriabova T, Henle M L, Feld M S, Levine A J and Popescu G, Measurement of the nonlinear elasticity of red blood cell membrane. Phys Rev Lett (under review) 83. Popescu G, Park Y, Lue N et al (2008) Optical imaging of cell mass and growth dynamics. Am J Physiol Cell Physiol 295:C538 84. Evans E and Fung Y C (1972) Improved measurements of the erythrocyte geometry. Microvasc Res 4:335–347 85. Savitz D, Sidel V W, and Solomon A K (1964) Osmotic Properties of human red cells. J Gen Physiol 48:79–94 86. Friebel M and Meinke M (2006) Model function to calculate the refractive index of native hemoglobin in the wavelength range of 250–1100 nm dependent on concentration. Appl Opt 45:2838–2842 87. Schmid-SchoNbein H, Wells R O E, and Goldstone J (1969) Influence of deformability of human red cells upon blood viscosity. Circ Res 25:131–143 88. Mohandas N, Clark M R, Jacobs M S et al (1980) Analysis of factors regulating erythrocyte deformability. J Clin Invest 66:563 89. Wells R and Schmid-Schonbein H (1969) Red cell deformation and fluidity of concentrated cell suspensions. J Appl Physiol 27:213–217 90. Kilejian A (1979) Characterization of a protein correlated with the production of knob-like protrusions on membranes of erythrocytes infected with Plasmodium falciparum. Proc Natl Acad Sci U S A 76:4650–4653 91. Sherman I W (1979) Biochemistry of Plasmodium (malarial parasites). Microbiol Rev 43:453 92. Goldberg D E, Slater A F G, Cerami A et al (1990) Hemoglobin degradation in the malaria parasite Plasmodium falciparum: an ordered process in a unique organelle. Proc Natl Acad Sci U S A 87:2931–2935 93. Nash G B, O’Brien E, Gordon-Smith E C et al (1989) Abnormalities in the mechanical properties of red blood cells caused by Plasmodium falciparum. Blood 74:855–861 94. Cranston H A, Boylan C W, Carroll G L et al (1984) Plasmodium falciparum maturation abolishes physiologic red cell deformability. Science 223:400–403 95. Paulitschke M and Nash G B (1993) Membrane rigidity of red blood cells parasitized by different strains of Plasmodium falciparum. J Lab Clin Med 122:581–589 96. Miller L H, Baruch D I, Marsh K et al (2002) The pathogenic basis of malaria. Nature 415:673–679 97. Suresh S, Spatz J, Mills J P et al (2005) Connections between single-cell biomechanics and human disease states: gastrointestinal cancer and malaria. Acta Biomater 1:15–30 98. Mills J P, Diez-Silva M, Quinn D J et al (2007) Effect of plasmodial RESA protein on deformability of human red blood cells harboring Plasmodium falciparum. Proc Natl Acad Sci U S A 104:9213–9217 99. Glenister F K, Coppel R L, Cowman A F et al (2002) Contribution of parasite proteins to altered mechanical properties of malaria-infected red blood cells. Blood 99:1060–1063 100. Pei X, Guo X, Coppel R et al (2007) The ring-infected erythrocyte surface antigen (RESA) of plasmodium falciparum stabilizes spectrin tetramers and suppresses further invasion. Blood 110:1036–1042 101. Lee J C M and Discher D E (2001) Deformation-enhanced fluctuations in the red cell skeleton with theoretical relations to elasticity, connectivity, and spectrin unfolding. Biophys J 81:3178–3192 102. Parpart A and Hoffman J (1956) Flicker in erythrocytes. Vibratory movements in the cytoplasm? J Cell Comp Physiol 47:295–303

13

Optical Sensing of Red Blood Cell Dynamics

309

103. Evans J, Gratzer W, Mohandas N et al (2008) Fluctuations of the red blood cell membrane: relation to mechanical properties and lack of ATP dependence. Biophys J 94:4134 104. Szekely D, Yau T, and Kuchel P (2009) Human erythrocyte flickering: temperature, ATP concentration, water transport, and cell aging, plus a computer simulation. Eur Biophys J 38:923–939 105. Gov N S (2007) Active elastic network: cytoskeleton of the red blood cell. Phys Rev E 75:11921 106. Li J, Lykotrafitis G, Dao M et al (2007) Cytoskeletal dynamics of human erythrocyte. Proc Natl Acad Sci U S A 104:4937 107. Zhang R and Brown F (2008) Cytoskeleton mediated effective elastic properties of model red blood cell membranes. J Chem Phys 129:065101 108. Park Y, Best C, Auth T et al (2010) Metabolic remodeling of the human red blood cell membrane. Proc Natl Acad Sci U S A 107:1289 109. Sheetz M and Singer S (1977) On the mechanism of ATP-induced shape changes in human erythrocyte membranes. I. The role of the spectrin complex. J Cell Biol 73:638–646 110. Auth T, Safran S, and Gov N (2007) Fluctuations of coupled fluid and solid membranes with application to red blood cells. Phys Rev E 76:51910 111. Tuvia S, Almagor A, Bitler A et al (1997) Cell membrane fluctuations are regulated by medium macroviscosity: evidence for a metabolic driving force. Proc Natl Acad Sci U S A 94:5045–5049 112. Mizuno D, Tardin C, Schmidt C et al (2007) Nonequilibrium mechanics of active cytoskeletal networks. Science 315:370 113. Liu F, Mizukami H, Sarnaik S et al (2005) Calcium-dependent human erythrocyte cytoskeleton stability analysis through atomic force microscopy. J Struct Biol 150:200–210 114. Muller E, Hegewald H, Jaroszewicz K et al (1986) Turnover of phosphomonoester groups and compartmentation of polyphosphoinositides in human erythrocytes. Biochem J 235:775 115. Patel V and Fairbanks G (1981) Spectrin phosphorylation and shape change of human erythrocyte ghosts. J Cell Biol 88:430–440 116. Agre P and Parker J (1989) Red blood cell membranes: structure, function, clinical implications. New York: CRC Press 117. Tchernia G, Mohandas N, and Shohet S (1981) Deficiency of skeletal membrane protein band 4.1 in homozygous hereditary elliptocytosis. Implications for erythrocyte membrane stability. J Clin Invest 68:454 118. Suresh S (2006) Mechanical response of human red blood cells in health and disease: some structure-property-function relationships. J Mater Res 21:1872 119. Fred M and Pickens M (1969) Metabolic dependence of red cell deformability. J Clin Invest 48:795

Index

A Adhesion-mediated signals biomedical engineering and diseases, 7–8 cell migration, 2 chemical factors, 1–2 cytoskeletal structures, 2 durotaxis, 2 extracellular signals, 2 gene expression, 2 mechanical forces FAK, 3 fibronectin molecules, 4 force-induced protein, 4 inside-out signaling, 4, 5 scaffold proteins, 3 Src kinase, 2–3 transmembrane signals, 2 motile adhesive cells, 2 non-chemical signals, 2 topography, 2 universal sensor cell shape, 5–6 focal adhesion size, 6 substrate elasticity, 4–6 Angiogenesis, 160–161 Anti-cancer drug testing 3D drug assay techniques 3D culture systems, 154 drug evaluation, issues and parameters, 154–156 hydrogel materials, 162–163 multi-layer drug transport, 156–157 tissue-dependent drug resistance, 157, 158 2D models, 152–153 3D models, 153–154 3D pathological processes angiogenesis, 160–161 multicellular organizations, 159

tissue morphogenesis, 159 tumor invasion, 161–162 matrix mechanics, 153 multicellular 3D models histoculture system, 159 photodynamic therapy, 158–159 Apoptosis anoikis, 144–145 cancer cell metastasis, 145 cellular adherence, 144 proapototic factors, 145 Atherosclerosis, cell-generated forces ECM deposition and vessel stiffness, 66 endothelial barrier integrity, 64–65 increased substrate and endothelial cell stiffness, 65 increased vessel stiffness, 64, 65 VSMCs, 66 Atomic force microscopy (AFM), 244–246 Autoantibodies, 186–187 Autoimmune blistering skin disorders autoantibodies, 186–187 blistering mechanism, 187 self-antigens, 185–186

B Basal cell shrinkage hypothesis, 187 Biomechanical interactions tools cytoskeletal structure actin filaments, 233–235 adenosine triphosphate, 235 cell structure, 233, 234 Euler buckling forces, 234 intermediate filaments, 233 microtubules, 233 mechanobiology, 239 (see also Mechanobiological tools) atomic force microscopy, 244–246

A. Wagoner Johnson and Brendan A.C. Harley (eds.), Mechanobiology of Cell-Cell and Cell-Matrix Interactions, DOI 10.1007/978-1-4419-8083-0, # Springer Science+Business Media, LLC 2011

311

312 Biomechanical interactions tools (cont.) calibration issues, 239 force, displacement and stiffness ranges, 239, 240 magnetic micromanipulation, 242–243 micropipette aspiration, 244 optical tweezers, 243–244 silicon beam-based tools, 239 tensile loading, 241–242 traction force microscopy, 246–249 mechanotransduction cell generated force, 237, 238 cell signaling pathways, 235–236 chemical signaling paradigm, 238 cytoskeletal protein structures, 235 endothelial and epithelial cells, 235–236 focal adhesion complexes, 236 force-balance paradigm, 237 integrin-based focal adhesions, 236 mechanical forces, 236–237 Bio-MEMS force sensor cell colonies, 111–112 force-retraction relation, 113 maximum interaction force, 113–114 minimum displacement, 111 probe-cell contact, 112–113 structural beams, 111, 112 Blistering skin diseases acquired, 185, 191 animal models, 191 autoantibodies, 186–187 autoimmune, 185–186 clinical manifestation, 184–185 diagnostic comparison, EBS and PV anatomical features, 189, 190 biopsy specimen, 189–191 disease phenotypes, 191 inherited, 185 keratinocyte mechanical integrity, 191 Staphylococcus infections, 189 ultrastructural views desmosomes, 188 hemidesmosomes, 187 Boussinesq theory, 212

C Cancer, 7 anti-cancer drug testing (see Anti-cancer drug testing) apoptosis anoikis, 144–145 cancer cell metastasis, 145

Index cellular adherence, 144 proapototic factors, 145 cell adhesion (see Cellular adhesion) cell movement (see Cellular movement) extracellular matrix components, 126 composition and mechanical forces, 127, 128 durotaxis, 128 growth factors, 129 pore architecture, 129 proteoglycans, 127 stiffness and porosity, 128–129 strain, 127–128 stress, 127 viscoelasticity, 128 metastasis (see Metastasis) proteases (see Proteases) systems biology approach, 124–126 tumor metastasis, signaling (see Signaling cascades, tumor metastasis) Cardiac fibrosis, tissue construct adherens junctions, 96 desmosomes, 96 engineered heart tissues, 93–95 fibroblasts, 93 gap junctions, 96 myofibroblasts, 96–97 Cell-cell adhesion, 199–200 Cell fate 2D culture (see 2D culture studies, cell fate) 3D culture (see 3D culture studies, MSC) synthetic extracellular matrix analogs anoikis, 24 chemical features, 24, 25 2D culture, 25 transplanted cells, 24 Cell-generated forces atherosclerosis ECM deposition and vessel stiffness, 66 endothelial barrier integrity, 64–65 increased substrate and endothelial cell stiffness, 65 increased vessel stiffness, 64, 65 VSMCs, 66 cell-cell interactions, 58 cell-matrix interactions, 56–58 cell proliferation, 62–63 cell sorting differential adhesion hypothesis, 48–49 extracellularmatrix protein, 50 substrate mechanics, 50–52 chemical factors, 66

Index collective cell migration, 61–62 embryonic development and stem cell differentiation cells stiffness, 52–53 Drosophila, 52 matrix rigidity, 52 Rho-ROCK pathway, 53 substrate stiffness and elasticity, 52 Xenopus laevis, 52 extracellular matrix remodeling collagen, 59 fibronectin, 59–60 tissue assembly, 60–61 liver fibrosis, 63–64 myocardial infarction, 64 single cell motility, 61 traction force quantification landmark experiments, 55–56 microfabricated post array detectors, 56 traction force microscopy, 56 traction forces, 54 tumor progression and metastasis, 63 Cell-matrix deformations non-wrinkling technique (see Three-dimensional traction force microscopy) wrinkling formation, 212 Cell-matrix interactions, 56–58 Cellular adhesion cadherins, 137–138 focal adhesion kinase activation, 139 cell migration and proliferation, 139–140 JAK/STAT pathway, 140–141 suppressors of cytokine signaling, 141 tyrosine residues, 139 vascular organization, 140 integrins, 138 Cellular movement actin and Rho GTPases actomyosin, 136 Cdc42, 135 cortactin, 136 PI3K molecules, 135–136 WASP proteins, 135 pseudopodia cofilin, 135 filopodia, 134 invadopodia, 134 lamellipodia, 134 Cofilin, 135

313 Collagen, 59 Cytoskeleton. See also Keratinocytes adhesion-mediated signals, 2 mechanics, tissue construct force relaxation curve, 90, 91 mechanical relaxation data, 90 myofibroblasts, 90–91 pre-stretching, 91–92 structure actin filaments, 233–235 adenosine triphosphate, 235 cell structure, 233, 234 Euler buckling forces, 234 intermediate filaments, 233 microtubules, 233

D 2D culture studies, cell fate cell-matrix mechanics, 28 cell migration speed, 27 stem cell, 29–30 DiMilla-Lauffenburger model, 27 elastic modulus, 27, 29 substrate rigidity, 27 mechanosensing, 34 stem cell, matrix mechanics chemical factors, 30 elasticity-dependent cell differentiation, 29 pre-osteoblasts differentiation, 29 soluble factors, 29–30 3D culture studies, mesenchymal stem cells cells sense matrix elasticity, 31 a5-integrins, 31 matrix dimensionality, 30–31 mechanosensing ECM polymers, 31–32 matrix elasticity, 34–36 peptide modifed synthetic and natural polymers, 32 scaffolds, 31 stem cell fate regulation, matrix elasticity, 32–34 Differential adhesion hypothesis (DAH), 48–49 Differential surface compaction hypothesis (DSCH), 49 Diffraction phase microscopy epi-fluorescence microscopy, 286 experimental setup, 284 interference image, 284, 295 numerical process, 284–285

314

Index

Diffraction phase microscopy (cont.) quantitative phase images, 283 retrieved and sensitive phase image, 286 spatial modulation, 283–284 Digital volume correlation (DVC) cross-correlation function, 216 displacement gradients, 217 Fourier transforms, 216 principle, 215, 216 Displacement and tractions activated coverslips preparation, 219–220 cell culture, 221 confocal microscopy and time-lapse imaging, 221–222 dynamic cell traction calculations, 218–219 fibroblast cell migration, 3D cell-matrix interactions cell-induced surface tractions, 227, 228 cell tracking, 224 GFP-actin fluorescent vector construct, 224 locations and highly polarized cell, 225–226 substrate thickness, 224 time evolution, 224–225, 227–230 traction fields, 226–227 polyacrylamide film fibronectin, functionalization, 220–221 preparation, 220 unconfined and confined compression, 222–223 Young’s modulus, 212 resolution and measurement sensitivity, 219 three-dimensional displacement vector, 217 Drug resistance multi-layer drug transport, 156–157 tissue-dependent drug resistance, 157, 158 Durotaxis, 2, 6, 7

Exfoliative toxins, 189 Extracellularmatrix protein, 50

E Engineered heart tissues (EHTs), 78 Epidermal keratinocyte. See Keratinocytes Epidermolysis bullosa simplex (EBS) cytolysis, 189 vs. pemphigus vulgaris anatomical features, 189, 190 biopsy, 189–191

H Hilbert phase microscopy (HPM), 282, 283 Hill model, skeletal muscle, 80 Human breast epithelial cells (HBECs), 160 Hydrogels advantages, 271–272 2D cell culture, 272 3D cell culture, 273–274

F Fibroblast cells mechanobiology AFM, 259 cell type and function, 258 force posts, 259 MEMS devices, 259 microfabricated tools, 258 tissue maintenance, 258 migration, 3D cell-matrix interactions cell-induced surface tractions, 227, 228 cell tracking, 224 GFP-actin fluorescent vector construct, 224 locations and highly polarized cell, 225–226 substrate thickness, 224 time evolution, 224–225, 227–230 traction fields, 226–227 tissue construct, cell-cell interactions crossover point, 89 force relaxation, 88 inter-related features and models, 90 passive, incremental tangent moduli variation, 88–89 tissue constructs remodel, 87–88 Fibroblast populated matrices (FPMs), 77–78 Fibronectin, 59–60 Filament polymerization, 176 Filopodia, 134 Fluorescence interference contrast (FLIC), 281 Focal adhesion kinase (FAK), 2, 3 Focal adhesions, 3–6 F€ orster resonance energy transfer (FRET), 38 Fourier phase microscopy (FPM), 282, 283 Fourier-space automated band-pass length estimation (FABLE), 90

Index I Immunobullous diseases. See Autoimmune blistering skin disorders Integrin-ECM bonds catch-bond mechanism, 37 cell-material interfaces biochemical analysis, 39 molecular fluorescence techniques, 38 traction forces, indirect effects, 40–41 matrix elasticity, 36 synthetic materials, cellular reorganization adhesive epitopes, 41 cell-mediated changes, 41–42 Intermediate filaments building block, 176 human blistering skin diseases, 175 vs. microtubules and microfilament, 176 morphological form, 176 plectin, 176 proteins, 176–177 Invadopodia, 134

K Keratinocytes. See also Skin anchoring junctions adherens junctions, 179–180 cell-cell and cell-matrix junctions, 177–179 classification, 178 desmosome, 180 focal adhesion complex, 178–180 generic molecular assembly, 179 hemidesmosomes, 180 integrin, 178–179 microtubules, 180–181 in vitro studies, 181–183 in vivo studies, 183 blistering skin diseases (see Blistering skin diseases) categories, 173 cornification, 174–175 cytoskeletal network proteins intermediate filaments (see Intermediate filaments) microfilaments and microtubules, 175 differentiation biochemical and morphological events, 173–174 protein expression, 183–184 matrix metalloproteinases, 172

315 rheology (see also Rheological behavior, keratinocyte) epithelial sheet mechanics, 195–197 integument morphology and mechanical properties, 191–192 isolated keratinocyte mechanics, 193–195

L Lamellipodia, 134 Laser scanning confocal microscopy (LSCM), 214–215 Liver fibrosis, 63–64

M Madin–Darby canine kidney (MDCK) epithelial cells, 251 Matrix metalloproteinases (MMPs), 32 cell migration, 142–144 families and subfamilies, 141, 142 regulation, 141–142 Mechanobiological tools, 239. See also Fibroblast cells atomic force microscopy, 244–246 calibration issues, 239 force displacement and stiffness range, 239, 240 force posts and beam bending cardiomyocytes, 251 3D culture, 252 deformation behavior, 250–251 PDMS cantilevers, neonatal rat cardiomyocytes, 252–253 traction and adhesion forces, 251 magnetic micromanipulation, 242–243 microelectromechanical systems cell deformation, 253 cell stiffness measurement, 253, 254 electrostatically actuated microdevices, 257 microfabricated two-dimensional force sensors, 253 piezoresistive cantilevers, 255–256 probe assays, 256–257 single cell measurement system, 257 micropipette aspiration, 244 optical tweezers, 243–244 silicon beam-based tools, 239 tensile loading, 241–242 traction force microscopy, 246–249

316 Mechanotransduction biomaterial design, 268 cell generated force, 237, 238 cell signaling pathways, 235–236 chemical signaling paradigm, 238 cytoskeletal protein structures, 235 endothelial and epithelial cells, 235–236 focal adhesion complexes, 236 force-balance paradigm, 237 glass substrates, 268–269 hydrogels 2D cell culture, 272–273 3D cell culture, 273–274 integrin-based focal adhesions, 236 matrix biology, 268 mechanical forces, 236–237 mechanical signals, 267, 275 polydimethylsiloxane advantages, 270 elastic properties, 269, 275 surface modifications, 270, 271 water-vapor plasma and UV radiation, 270 Membrane dynamics, RBC ATP effects ATP-depleted and repleted group, 299–300 biconcave shape and membrane fluctuations, 300 metabolism-dependent shape transformation, 303–304 non-equilibrium dynamics, 301–302 non-Gaussian parameter, 302–303 protein phosphorylation, 303 RMS displacement, 300, 301 malarial effects mechanical modulation, 296–297 Pf-RBCs, 297–299 structural changes, 296 mathematical model, 289–290 morphology effects correlation function, 291 DC-EC-SC shape trasition, 290–291 shear and bending modulus, 291–293 osmotic effects cytosolic viscosity, 296 hypertonic/hypotonic solutions, 293 osmolarities, 294–295 RBCs shapes, 294 refractive index, 294 shear modulus, 295, 296 thermal fluctuations blood sample preparation, 288

Index DC-EC-SC transition, 288–289 discocyte images, 288–289 giant unilamellar vesicles, 286–287 spatial wave vector and temporal frequency, 287–288 Mesenchymal stem cells (MSC) cell fate (see Cell fate) integrin-ECM bonds (see Integrin-ECM bonds) Metastasis. See also Signaling cascades, tumor metastasis adhesion, 106 altered mechanical phenotypes, 107 altered tumor cell behavior adhesion, 106 detachment, 106–107 low adhesivity, 107 cell adhesion bio-MEMS, 110–111 classic Coulter counter, 111 E-cadherin/N-cadherin homophilic bonds, 110 non-specific adhesion activity, 111 single-molecule force spectroscopy, 110 target organ microenvironment, 109–110 cell-cell vs. cell matrix interaction cell-cell interaction, 109 collagen crosslinking induction, 109 extracellular and intracellular mechanical forces, 109 integrins, 109 mammary epithelial cells, 108 signaling cascades, 109 uPAR inhibition, 108–109 tumor microenvironments, remodeling epithelial-mesenchymal transition, 108 secondary sites alteration, 108 stromal-epithelial interaction, 107 Micro-electrical-mechanical-system (MEMS) force sensor. See also Bio-MEMS force sensor cell deformation, 253 cell stiffness measurement, 253, 254 electrostatically actuated microdevices, 257 microfabricated two-dimensional force sensors, 253 piezoresistive cantilevers, 255–256 probe assays, 256–257 single cell measurement system, 257 Microfabricated post array detectors (mPADs), 56, 67

Index Mitogen activated protein kinase (MAPK) signaling pathway, 24 Myocardial infarction, 64

N Non-specific adhesion study Bio-MEMS force sensor cell colonies, 111–112 force-retraction relation, 113 maximum interaction force, 113–114 minimum displacement, 111 probe-cell contact, 112–113 structural beams, 111, 112 malignant cancer cells adhesion strengths, 114 HCT–8 cancer cells, 114–115 homotypic cell-cell adhesion rates, 115–116 materials and methods cell culture and experiments setup, 117–118 Coulter counter cell-cell adhesion measurement, 118

P Pemphigus vulgaris (PV) acantholytic blisters, 189 autoantibodies, 186–187 blistering mechanism, 187 vs. EBS anatomical features, 189, 190 biopsy, 189–191 immunofluorescence analysis, 190–191 self-antigens, 186 Phase contrast microscopy (PCM), 280–281 Photodynamic therapy, 158–159 Plasmodium falciparum invaded human RBCs (Pf-RBCs) elastic restoring force, 297–298 in-plane shear modulus, 298–299 topographic images and elastic constant, 297 Poisson ratio, 223 Polydimethylsiloxane (PDMS), 16 mechanotransduction advantages, 270 elastic properties, 269, 275 surface modifications, 270, 271 water-vapor plasma and UV radiation, 270 neonatal rat cardiomyocytes, 252–253

317 Proteases matrix metalloproteinases cell migration, 142–144 families and subfamilies, 141, 142 regulation, 141–142 urokinase plasminogen activator, 144 Pseudopodia, 134–135

R Red blood cell dynamics fluorescence interference contrast, 281 membrane dynamics (see Membrane dynamics, RBC) membrane fluctuation, 280 membrane mechanics, 279–280 micropipette aspiration, 280 optical tweezers, 280–281 phase contrast microscopy, 280–281 quantitative phase imaging diffraction phase microscopy (see Diffraction phase microscopy) digital holography, 282 full-field phase imaging technique, 281, 282 optical coherence tomography, 281–282 optical phase delay, 282–283 shear and bending moduli, 280 thermal fluctuations, 280 Rheological behavior, keratinocyte dermis, 192–193 epidermis, 192–193 mesoscopic specimen, 195–197 microscopic investigations, 193–195 nanoscopic measurements cell-cell adhesion, 199–200 cytoskeletal filaments and networks, 197–199 skin, 192–193 Rho-associated kinase (ROCK), 53

S Signaling cascades, tumor metastasis growth factor receptors, 129–130 inside-out signaling, 133 integrins, 129 outside-in signaling MCF–7 and MDA–231, 131 myosin light chain kinase activation, 132 Rab proteins, 132–133 Ras pathway, 132 proteins, 129, 130

318 Signal transducers and activators of transcription (STAT), 140–141 Single cell motility, 61 Skin anisotropy and heterogeneity, 193 mechanical properties, 192–193 resting tension, 193 structure and function basement membrane, 172 connective tissue, 172 dermis and epidermis, 171–172 fibroblasts, 172, 173 integument, 171 proteoglycan, 172 stratified squamous keratinized epithelium, 172, 174 type I collagen, 172 Smooth muscle cells (SMCs), 161 Staphylococcus aureus, 189 Substrate elasticity, 19 cell-cell adhesions cadherins, 15 F-actin filaments, physiological stiffness, 17 neonatal rat cardiomyocytes, 16–17 optimal stiffness, 17 quantitative measurements, 16 single cardiomyocytes, 16 T lymphocyte activation, 15–16 vesicle movements, drosophila, 15 cell-cell vs. cell-ECM junctions, 17–18 cell-matrix adhesions adhesion time and force, 15 integrins, 12–13 quantitative measurements, 15 shear stress, 14 stiffness-dependent cell spreading and stiffening, 13–14 substrate stiffness, 13 cell-matrix contacts, 11–12 cytoskeletal elements, 11 erythrocytes, 11 force transmission and transduction, 11–12 tissue formation and function, 19 Suppressors of cytokine signaling (SOCS), 141

T Three-dimensional traction force microscopy (3D TFM) digital volume correlation cross-correlation function, 216

Index displacement gradients, 217 Fourier transforms, 216 principle, 215, 216 displacement and tractions (see Displacement and tractions) laser scanning confocal microscopy, 214–215 Three-dimensional tumor model anti-cancer drug testing angiogenesis, 160–161 3D culture systems, 154 drug evaluation, issues and parameters, 154–156 histoculture system, 159 hydrogel materials, 162–163 multi-layer drug transport, 156–157 photodynamic therapy, 158–159 tissue-dependent drug resistance, 157, 158 tissue morphogenesis, 159 tumor invasion, 161–162 vs. 2D models, 152–153 matrix mechanics, 153 Tissue constructs, 97–98 bio-artificial, 76 cardiomyocyte constructs, 78 cell-cell interaction active effect, 83 cardiac fibrosis (see Cardiac fibrosis, tissue construct) cytoskeletal mechanics (see Cytoskeleton) dimensionless cell concentration, 84 effective cell length, 84 ellipsoids, 83 fibroblast cells (see Fibroblast cells) homogenization methods, 85–86 modulus ratio, 84–85 passive effect, 83 planar models, 83–84 cell mechanics quantification FPM rings, 78–79 homogenization methods, 79 periodic microfield models, 79 Zahalak model, 79 (see also Zahalak model) 2D vs. 3D cultures, 76 engineered heart tissues, 78 fibroblast populated matrices, 77–78 interactions, 76 interconnections, 76

Index quantitative 3D culture systems collagen and cell population, 77 engineered heart tissues, 78 fibroblast populated matrices, 77–78 Tissue morphogenesis, 159, 238 Traction force microscopy, 56, 246–249 Traction forces, 2–3 Tumor-associated macrophages, 109 Tumor metastasis cellular adhesion (see Cellular adhesion) cellular movement (see Cellular movement) proteases (see Proteases) signaling cascades (see Signaling cascades, tumor metastasis)

U Urokinase plasminogen activator, 144

319 V Vascular smooth muscle cells (VSMCs), 66

W Wiskott–Aldrich syndrome protein (WASP), 131

Z Zahalak model cell anisotropy tensors, 81 Hill model, skeletal muscle, 80 macroscopic vs. axial strain, 82 one-dimensional embodiment, 80 orientation distribution, 80–81 Piola Kirchhoff stress, 79–80 self-consistent-type approach, 82–83 slender contractile rods, 80 small-strain tensor, deformations, 80 statistical averaging, active and passive mechanics, 79