Organelle-Specific Pharmaceutical Nanotechnology [1 ed.] 0470631651, 9780470631652

Citation preview

ORGANELLE-SPECIFIC PHARMACEUTICAL NANOTECHNOLOGY Edited by

Volkmar Weissig Gerard G. M. D’Souza

A JOHN WILEY & SONS, INC., PUBLICATION

ORGANELLE-SPECIFIC PHARMACEUTICAL NANOTECHNOLOGY

ORGANELLE-SPECIFIC PHARMACEUTICAL NANOTECHNOLOGY Edited by

Volkmar Weissig Gerard G. M. D’Souza

A JOHN WILEY & SONS, INC., PUBLICATION

Copyright © 2010 by John Wiley & Sons, Inc. All rights reserved Published by John Wiley & Sons, Inc., Hoboken, New Jersey Published simultaneously in Canada No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com. ISBN 978-0-470-63165-2 Library of Congress Cataloging-in-Publication Data is available. Printed in the United States of America 10

9

8

7

6

5

4

3

2

1

CONTENTS

Preface

ix

Contributors

xi

1. An Introduction to Subcellular Nanomedicine: Current Trends and Future Developments

1

Gerard G. M. D’Souza and Volkmar Weissig

2. Delivery of Nanonsensors to Measure the Intracellular Environment

15

Paul G. Coupland and Jonathan W. Aylott

3. Cytoplasmic Diffiusion of Dendrimers and Dendriplexes

35

Alexander T. Florence and Pakatip Ruenraroengsak

4. Endocytosis and Intracellular Trafficking of Quantum Dot– Ligand Bioconjugates

55

Tore-Geir Iversen, Nadine Frerker, and Kirsten Sandvig

5. Synthesis of Metal Nanoparticle-Based Intracellular Biosensors and Therapeutic Agents

73

Neil Bricklebank

6. Subcellular Fate of Nanodelivery Systems

93

Dusica Maysinger, Sebastien Boridy, and Eliza Hutter

7. Intracellular Fate of Plasmid DNA Polyplexes

123

Kevin Maier and Ernst Wagner

8. Intracellular Trafficking of Membrane Receptor-Mediated Uptake of Carbon Nanotubes

143

Bin Kang and Yaodong Dai

9. Real-Time Particle Tracking for Studying Intracellular Transport of Nanotherapeutics

161

Clive Chen and Junghae Suh

10. Tracking Intracellular Polymer Localization Via Fluorescence Microscopy

177

Simon C. W. Richardson v

vi

CONTENTS

11. Can QSAR Models Describing Small-Molecule Xenobiotics Give Useful Tips for Predicting Uptake and Localization of Nanoparticles in Living Cells? And If Not, Why Not?

193

Richard W. Horobin

12. Self-Unpacking Gene Delivery Scaffolds

207

Millicent O. Sullivan

13. Cellular Trafficking of Dendrimers

231

Yunus Emre Kurtoglu and Rangaramanujam M. Kannan

14. Endolysosomolytically Active pH-Sensitive Polymeric Nanotechnology

247

Han Chang Kang and You Han Bae

15. Uptake and Intracellular Dynamics of Proteins Internalized by Cell-Penetrating Peptides

263

Arwyn T. Jones

16. Cargo Transport by Teams of Molecular Motors: Basic Mechanisms for Intracellular Drug Delivery

289

Melanie J. I. Müller, Florian Berger, Stegan Klumpp, and Reinhard Lipowsky

17. The Potential of Photochemical Internalization (PCI) for the Cytosolic Delivery of Nanomedicines

311

Kristian Berg, Anette Weyergang, Anders Høgset, and Pål Kristian Selbo

18. Peptide-Based Nanocarriers for Intracellular Delivery of Biologically Active Proteins

323

Seong Loong Lo and Shu Wang

19. Organelle-Specific Pharmaceutical Nanotechnology: Active Cellular Transport of Submicro- and Nanoscale Particles

337

Galya Orr

20. Subcellular Targeting of Virus-Envelope-Coated Nanoparticles

357

Jia Wang, Mohammad F. Saeed, Andrey A. Kolokoltsov, and Robert A. Davey

21. Mitochondria-Targeted Pharmaceutical Nanocarriers

385

Volkmar Weissig and Gerard G.M. D’Souza

22. Cell-Penetrating Peptides for Cytosolic Delivery of Biomacromolecules

403

Camilla Foged, Xiaona Jing, and Hanne Moerck Nielsen

23. Therapeutic Nano-object Delivery to Subdomains of Cardiac Myocytes Valeriy Lukyanenko

433

CONTENTS

24. Design Parameters Modulating Intracellular Drug Delivery: Anchoring to Specific Cellular Epitopes, Carrier Geometry, and Use of Auxiliary Pharmacological Agents

vii

449

Silvia Muro and Vladimir R. Muzykantov

25. Uptake Pathways Dependent Intracellular Trafficking of DNA Carrying Nanodelivery Systems

475

Ikramy A. Khalil, Yuma Yamada, Hidetaka Akita, and Hideyoshi Harashima

26. Cellular Interactions of Plasmon-Resonant Gold Nanorods

507

Qingshan Wei and Alexander Wei

27. Quantum Dot Labeling for Assessment of Intracellular Trafficking of Therapeutically Active Molecules

535

Diane J. Burgess and Mamta Kapoor

Index

569

PREFACE

Pharmaceutical nanotechnology as applied to the level of cell organelles, that is, to the subcellular level, is emerging as a new field in biomedical research. While efforts aiming at the targeted delivery of biologically active molecules date back to the early 1970s, the successful delivery of such agents to particular cell organelles has only in recent times gained broader recognition. Without exaggeration it can be stated that the subcellular organelle-specific targeting of therapeutic and diagnostic agents has become the new frontier for drug delivery. A variety of pharmaceutical nanocarrier platforms like liposomes, carbon nanotubes, quantum dots, micelles, and dendrimers have undergone testing for their ability to control the subcellular disposition of drugs. Potential improvements in therapy through the use of organelle-targeted nanocarriers have been demonstrated. Two major strategies are the basis for achieving organelle-specific targeting of pharmaceutical nano carriers. The first is based on the inherent predisposition of the nanocarrier for a particular cellular compartment and the second is based on an appropriate surface modification of the carrier with organelle-specific ligands. Despite encouraging preliminary findings, or maybe even because of them, many questions begin to emerge or still remain unanswered. Principally, our knowledge about the cellular internalization of nanocarriers and about the impact of intracellular morphology and dynamics on the fate and disposition of nanocarriers is very limited. Do nanocarriers remain intact upon cell entry and subsequent disposition? Naturally, significant differences between solid nanoparticles and liposomes can be expected. What is the impact of size and architecture of nanoparticles on their intracellular trafficking and distribution? Can the drug release from nanocarriers be controlled in a timely and spatial manner? Can or do cells actively transport nanocarriers in cell-membrane-derived vesicles? Can intracellular membrane trafficking be utilized for organelle-specific drug delivery? How do nanocarriers interact with different organelles? Do liposomal phospholipids become part of cellular vesicles? If so, what then is the fate of the lipophilic drugs originally incorporated into liposomal membranes? Can liposomes fuse with organellar membranes such as the mitochondrial outer or inner membrane? How do nonbiodegradable nanocarriers affect cellular transport and metabolism? Our ability to answer those and many more questions largely depends on technological advances in imaging technology. It can be hoped that improvements in real-time fluorescence confocal microscopy of ix

x

PREFACE

live cells as well as the emergence of new imaging techniques together with the design of novel nanoprobes will help to more satisfactorily answer some of the questions raised above. This book is the result of our efforts to bring together the best of current knowledge in this new and emerging field of subcellular pharmaceutical nanotechnology. All chapters were written by leading experts in their particular fields, and we are extremely grateful to them for having spent part of their valuable time to contribute to this book. It is our hope that together we have succeeded in providing an essential source of state-of-the-art knowledge for any investigator, young and seasoned alike, whose research area involves the application of new nanotechnology to the subcellular level. VOLKMAR WEISSIG Glendale, Arizona GERARD G. M. D’SOUZA Boston, Massachusetts

CONTRIBUTORS

Hidetaka Akita, Laboratory for Molecular Design of Pharmaceutics, Faculty of Pharmaceutical Sciences, Hokkaido University, Hokkaido, Japan and CREST, Japan Science and Technology Agency (JST), Japan Jonathan W. Aylott, School of Pharmacy, University of Nottingham, Nottingham, United Kingdom You Han Bae, Department of Pharmaceutics and Pharmaceutical Chemistry, University of Utah, Salt Lake City, Utah Kristian Berg, Department of Radiation Biology, Institute for Cancer Research, Norwegian Radium Hospital, Oslo University Hospital, Montebello, Oslo, Norway Florian Berger, Department of Theory & Bio-Systems, Max Planck Institute of Colloids and Interfaces, Potsdam, Germany Sebastian Boridy, Department of Pharmacology & Therapeutics, McGill University, Montreal, Quebec, Canada Neil Bricklebank, Faculty of Health and Wellbeing, Biomedical Research Center, Sheffield Hallam University, Sheffield, United Kingdom Diane Burgess, School of Pharmacy, University of Connecticut, Storrs, Connecticut Clive Chen, Department of Bioengineering, Rice University, Houston, Texas Paul G. Coupland, School of Pharmacy, University of Nottingham, Nottingham, United Kingdom Gerard G. M. D’Souza, Department of Pharmaceutical Sciences, Massachusetts College of Pharmacy and Health Sciences, Boston, Massachusetts Yaodong Dai, Nanjing University of Aeronautics and Astronautics, Nanjing, Peoples Republic of China Robert A. Davey, Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas Alexander T. Florence, School of Pharmacy, University of London, London, United Kingdom xi

xii

CONTRIBUTORS

Camilla Foged, Department of Pharmaceutics and Analytical Chemistry, Faculty of Pharmaceutical Sciences, University of Copenhagen, Copenhagen, Denmark Nadine Frerker, Centre for Cancer Biomedicine, Faculty Division, Norwegian Radium Hospital, University of Oslo and Department of Biochemistry, Institute for Cancer Research, Norwegian Radium Hospital, Oslo University Hospital, Oslo, Norway Hideyoshi Harashima, Laboratory for Molecular Design of Pharmaceutics, Faculty of Pharmaceutical Sciences, Hokkaido, University, Hokkaido, Japan and CREST, Japan Science and Technology Agency (JST), Japan Anders Hgset, PCI Biotech AS, Oslo, Norway Richard W. Horobin, Division of Integrated Biology, FBLS, The University of Glasgow, Glasgow, Scotland, United Kingdom Eliza Hutter, Department of Pharmacology and Therapeutics, McGill University, Montreal, Quebec, Canada Tore-Geir Iversen, Centre for Cancer Biomedicine, Faculty Division, Norwegian Radium Hospital, University of Oslo, and Department of Biochemistry, Institute for Cancer Research, Norwegian Radium Hospital, Oslo University Hospital, Oslo, Norway Xiaona Jing, Department of Pharmaceutics and Analytical Chemistry, Faculty of Pharmaceutical Sciences, University of Copenhagen, Copenhagen, Denmark Arwyn T. Jones, Welsh School of Pharmacy, Cardiff University, Cardiff, Wales, United Kingdom Bin Kang, Nanjing University of Aeronautics and Astronautics, Nanjing, Peoples Republic of China Han Chang Kang, Department of Pharmaceutics and Pharmaceutical Chemistry, University of Utah, Salt Lake City, Utah Rangaramanujam M. Kannan, Department of Chemical Engineering and Materials Science, NICHD Perinatology Research Branch, Wayne State University, Detroit, Michigan Mamta Kapoor, School of Pharmacy, University of Connecticut, Storrs, Connecticut Ikramy A. Khalil, Laboratory for Molecular Design of Pharmaceutics, Faculty of Pharmaceutical Sciences, Hokkaido, University, Hokkaido, Japan and CREST, Japan Science and Technology Agency (JST), Japan Stefan Klumpp, Department of Theory & Bio-Systems, Max Planck Institute of Colloids and Interfaces, Potsdam, Germany

CONTRIBUTORS

xiii

Andrey A. Kolokoltsov, Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas Yunus Emre Kurtoglu, Department of Chemical Engineering and Materials Science, NICHD Perinatology Research Branch, Wayne State University, Detroit, Michigan Reinhard Lipowsky, Department of Theory & Bio-Systems, Max Planck Institute of Colloids and Interfaces, Potsdam, Germany Seong Loong Lo, Institute of Bioengineering and Nanotechnology, Singapore Valeriy Lukyanenko, Medical Biotechnology Center, University of Maryland Biotechnology Institute, Baltimore, Maryland Kevin Maier, Pharmaceutical Biotechnology, Munich Center for SystemBased Drug Research, and Center for NanoScience, Ludwig-MaximiliansUniversität, Munich, Germany Dusica Maysinger, Department of Pharmacology & Therapeutics, McGill University, Montreal, Quebec, Canada Melanie J. I. Müller, Department of Theory & Bio-Systems, Max Planck Institute of Colloids and Interfaces, Potsdam, Germany Silvia Muro, Center for Biosystems Research, University of Maryland Biotechnology Institute and Fischell Department of Bioengineering, University of Maryland, College Park, Maryland Vladimir R. Muzykantov, Department of Pharmacology and Targeted Therapeutics Program of Institute of Translational Medicine and Therapeutics, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania Hanne Moerck Nielson, Department of Pharmaceutics and Analytical Chemistry, Faculty of Pharmaceutical Sciences, University of Copenhagen, Copenhagen, Denmark Galya Orr, Chemical and Materials Sciences Division, Pacific Northwest National Laboratory, Richland, Washington Simon C. W. Richardson, University of Greenwich School of Science, Kent, England Pakatip Ruenraroengsak, Lung Cell Biology, Respiratory Medicine, National Heart and Lung Institute, Imperial College, London, United Kingdom Mohammad F. Saeed, Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas Kirsten Sandvig, Centre for Cancer Biomedicine, Faculty Division, Norwegian Radium Hospital, University of Oslo, and Department of Biochemistry,

xiv

CONTRIBUTORS

Institute for Cancer Research, Norwegian Radium Hospital, Oslo University Hospital, Oslo, Norway Pål Kristian Selbo, Department of Radiation Biology, Institute for Cancer Research, Norwegian Radium Hospital, Oslo University Hospital, Montebello, Oslo, Norway Junghae Suh, Department of Bioengineering, Rice University, Houston, Texas Millicent O. Sullivan, Department of Chemical Engineering, University of Delaware, Newark, Delaware Ernst Wagner, Pharmaceutical Biotechnology, Munich Center for SystemBased Drug Research, and Center for NanoScience, Ludwig-MaximiliansUniversität, Munich, Germany Jia Wang, Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas Shu Wang, Institute of Bioengineering and Nanotechnology, and Department of Biological Sciences, National University of Singapore, Singapore Alexander Wei, Department of Chemistry, Purdue University, West Lafayette, Indiana Qingshan Wei, Department of Chemistry, Purdue University, West Lafayette, Indiana Volkmar Weissig, Department of Pharmaceutical Sciences, Midwestern University College of Pharmacy, Glendale, Arizona Anette Weyergang, Department of Radiation Biology, Institute for Cancer Research, Norwegian Radium Hospital, Oslo University Hospital, Montebello, Oslo, Norway Yuma Yamada, Laboratory for Molecular Design of Pharmaceutics, Faculty of Pharmaceutical Sciences, Hokkaido University, Hokkaido, Japan and CREST, Japan Science and Technology Agency (JST), Japan

CHAPTER 1

An Introduction to Subcellular Nanomedicine: Current Trends and Future Developments GERARD G. M. D’SOUZA Department of Pharmaceutical Sciences, Massachusetts College of Pharmacy and Health Sciences, Boston, Massachusetts

VOLKMAR WEISSIG Department of Pharmaceutical Sciences, Midwestern University College of Pharmacy Glendale, Glendale, Arizona

Drug therapy is based largely on the paradigm that an ideal drug will selectively exert a desired pharmacological activity free from negative side effects to modulate either the symptoms or the underlying biochemical cause of a disease to provide a benefit to the patient. In order to have such selective action, the drug molecule should ideally interact with only the disease-associated biochemical pathway but have no activity with respect to any normal biochemical pathway. This principle was famously explored by Paul Ehrlich in his search for agents with selective toxicity toward bacteria. Ehrlich’s work is widely accepted to have given rise to the concept of the ideal drug molecule as a “magic bullet,” a term that he used for the first time in his Harben Lectures [1]. Finding such selective molecules is relatively easy when there are significant differences between the disease-causing process and normal human biochemical pathways, as in the case of infectious diseases. Not surprisingly, in the decades since Ehrlich’s work infectious diseases have become much easier to treat but it does bear consideration that the lack of activity in nondisease cells is dose dependent and not absolute. Most drugs that are considered to be selectively toxic to invading pathogens are in fact toxic to human cells as well but just at higher doses. However, given that the new challenges in drug therapy lie in the treatment of diseases associated with malfunctions Organelle-Specific Pharmaceutical Nanotechnology, Edited by Volkmar Weissig and Gerard G. M. D’Souza Copyright © 2010 John Wiley & Sons, Inc.

1

2

AN INTRODUCTION TO SUBCELLULAR NANOMEDICINE

of normal human biochemical pathways in certain tissues, the concept of the magic bullet perhaps needs to be redefined or at least clarified. In the infectious disease example it is essential to understand that the so called magic bullet did not necessarily have to home in on the disease agent but could in fact accumulate to the same level in both host and pathogen cells. Just as long as the agent was toxic only to the pathogen it was considered by many to be a magic bullet. In fact, Ehrlich’s Nobel Prize Lecture from December 11, 1908 speaks of “outlining the principles of selective toxicity” rather than selective accumulation. Therefore one could argue that for a molecule to be a magic bullet, it does not have to selectively accumulate at its intended site of action but just that it should not exert its action anywhere but at that intended site of action. This is different from what is now referred to as drug targeting. The term “targeting” is most often meant to imply that the molecule is in some way able to selectively accumulate at an intended site of action and that the selective accumulation is associated with its selective action as a magic bullet. Unfortunately (perhaps due to the widespread use of the noun target to describe a potential molecular site of action), there is often the misconception that a drug that is believed to act at a molecular target (noun) is by default also able to target (verb) or “home in” on that target (noun). It would therefore be more appropriate to define a true magic bullet as a drug molecule that is specific in its activity for a molecular target but that is also able to selectively accumulate at this molecular target and exert a selective therapeutic action by virtue of both its specific activity and its selective accumulation. This distinction is important when we consider the daunting challenge of developing magic bullets for diseases like cancer and neurodegenerative diseases like Alzheimer’s, as well as hormone imbalance diseases like diabetes that are becoming more widespread. Unless unique molecular targets found exclusively (or at sufficiently higher levels) in the diseased state and not in normal state are discovered, magic bullets by the traditional definition or the compromise of dose-dependent activity at the site of action may not be feasible. Strategies to effectively control the disposition of drug molecules thus represent an important tool for successful therapy. At a very basic level, selective accumulation is influenced by bioavailability and subsequent biodistribution. In the context of drug molecules, biodistribution is primarily related to physicochemical properties. Many potent drug candidates exhibit low bioavailability due to their limited water solubility. On the other hand, water-soluble compounds display a very limited ability to cross biological membranes, which essentially can exclude them from the cell interior. To overcome the limitations that a compound’s physicochemical nature can impose on its potential pharmaceutical application, the process of largescale screening of chemical libraries has been extended beyond just identifying desired bioactivity. Screening approaches routinely incorporate selection for desirable physicochemical properties that might confer high bioavailability as well. On the down side, this approach leads to many potent molecules being

AN INTRODUCTION TO SUBCELLULAR NANOMEDICINE

3

excluded from further development because they aren’t true magic bullets; that is, while they may have a potent pharmacological action at a desired molecular target they aren’t able to find their way exclusively to that target. There is most certainly a growing list of such molecules that are in essence potential drugs if only a delivery strategy can be devised to get them to their molecular target in the human body. As the search for the perfect magic bullet continues there is a significant effort to improve the action of currently available molecules by using targeted delivery approaches. It is not surprising that the field of drug targeting has grown significantly in the effort to develop better therapy. Based on the experience over the decades since Ehrlich first introduced the magic bullet concept, it seems more reasonable to separate the functions of pharmacological action and selective accumulation into properties desired in a drug molecule and in a delivery system, respectively, rather than the traditional expectation that the drug molecule alone possess both properties. Generally, drug disposition may be modulated via three broad approaches First, the drug molecule might be modified subtly to change its physicochemical properties without adversely affecting its inherent pharmacological action. This is essentially the intent of medicinal chemistry approaches and the concept of structure–activity relationship (SAR) studies that have now become standard practice and often aren’t even considered a means of achieving targeting. The second approach might be considered to be an extension of the first but is different in that it involves using chemistry to conjugate ligands that are often larger than simple organic functional groups to change the biodistribution of a molecule. Again this approach works as long as the conjugation does not adversely affect the desired pharmacological activity of the molecule. Conjugation using selectively cleavable linkers is an extension of this strategy. The third strategy involves the use of some sort of delivery system or a carrier system and does not involve chemical modifications to the pharmacologically active molecule. Pharmaceutical nanocarriers fall into this category and several such technologies are being developed that are fast becoming applicable to a variety of pharmacologically active molecules. Pharmaceutical nanocarriers offer what might be viewed as a nonchemical approach to modify the disposition of drug molecules. All chemistry can be performed on the components of the nanocarrier system that can then be loaded with the drug to afford targeted delivery [2–5]. Most pharmaceutical nanocarriers can be modified for some level of targeting to specific tissues if not specific cell types. Long circulating liposomes and nanoparticles are able to passively target areas of leaky vasculature by virtue of the EPR effect and can additionally be modified with antibodies or other targeting ligands to afford cell specific recognition [6–10]. However, despite such advances, the improvement in drug action is not always dramatic. This is likely because many drugs act at molecular targets inside the cells and these molecular targets are often in well-organized subcellular structures inside the mammalian cell.

4

AN INTRODUCTION TO SUBCELLULAR NANOMEDICINE

The interior of a cell is very different from an aqueous buffer solution, in which small drug molecules can freely diffuse and randomly interact with potential cosolutes. In addition to the presence of the cytoskeletal network and various dispersed organelles, the cytoplasm contains a large amount of dissolved macromolecules. The concentration of dissolved macromolecules in the nucleoplasm and cytoplasm of living cells has been determined to be between 50 and 400 g/L [11, 12]. Subsequently, transport or diffusion events in such a crowded solution cannot be expected to be the same as those in buffer solutions. Generally, intracellular diffusion has been characterized as hindered diffusion, reflecting among other factors the high level of molecular crowding [13, 14]. Additionally, the fluid-phase viscosity of the cytoplasm and binding to intracellular components are believed to influence the diffusion of solutes inside a cell [15, 16]. While efforts aimed at thoroughly understanding cellular material properties such as cytoplasmic viscosity are currently underway [17], it is generally accepted that the physicochemical properties of the drug also play a major role in determining the subcellular fate of the drug molecule. Consequently, the ability to predict the influence of various properties of the drug molecule on the likely site of accumulation within the cell could prove to be a powerful tool in drug design to either select molecules with a desired subcellular accumulation or identify molecules that would benefit from subcellular targeting strategies. There is apparent fractal symmetry between the case of drug delivery to a cell and drug delivery to a molecular target inside a subcellular compartment. The cell could be viewed as being a small, slightly simpler but nonetheless highly organized “body” with “organs” (organelles) and “cells” (defined structures and molecular arrangements) within these organs. It should therefore stand to reason that controlling drug disposition within the cell might also be necessary for optimal drug action [18–27]. Consequently, the next logical step in the development of targeted nanocarriers would be to extend our control over nanocarrier distribution to the subcellular level as well. At least two major schemes can be imagined to be useful in the design of nanocarriers with the potential for subcellular targeting: the first based on the inherent predisposition of the nanocarrier for a particular compartment and the second based on attaching subcellular targeting ligands to the surface of nanocarriers to redirect their accumulation to the desired compartment. Essential to the latter of these approaches is the use of a subcellular targeting ligand. Such ligands could be, as in the case of leader sequences, derived from normal cellular trafficking processes or, as in the case of triphenyl phosphonium, based on observations of a predisposition of an organic compound for subcellular compartments. The availability of a wide range of subcellular stains is proof enough that there are several molecules with an inherent ability to accumulate in a particular subcellular compartment. Based on the intracellular distribution of a large variety of fluorophores, a quantitative structure–activity relationship (QSAR) model for predicting cellular uptake and intracellular distribution of low molecular weight compounds

AN INTRODUCTION TO SUBCELLULAR NANOMEDICINE

5

has been proposed [28]. This QSAR approach was recently applied to identify potential common chemical features of molecules that are known to selectively accumulate at or inside mammalian mitochondria within living cells [29]. The QSAR approach has additionally proved useful for the modeling of cationic transfection lipids [30] and could therefore be applicable to predicting the subcellular disposition of a potential therapeutic molecule and even to design molecules with a desired subcellular affinity for the development of subcellular targeting approaches. Approaches to nanocarrier-mediated subcellular delivery are based on the principle that the subcellular destination of a drug is the same as that of the nanocarrier. Nanocarriers by virtue of their particulate nature are believed to be subject to endocytic cell entry mechanisms and subsequent endolysosomal processing. As such, nanocarriers could be considered to be ideally suited for delivering bioactive molecules to the endolysosomal system. Indeed, directing nanomedicine complexes to the endolysosomal system has increasingly gained attention, as pathological conditions associated with endosomes and lysosomes could potentially benefit from therapies targeting these pathways [31– 34]. Although endocytosis is a common mechanism that almost all cells possess for the internalization of macromolecules, a wide array of such vesicular internalization mechanisms exist [31]. For example, nanoscale drug carrier systems taken up by clathrin-dependent receptor-mediated endocytosis (RME) are most likely to undergo lysosomal degradation, while clathrin-independent RME may lead to endosomal accumulation [31]. Consequently, the type of targeting moiety displayed by the nanocarrier system will determine whether the carrier delivers its cargo to either endosomes or lysosomes. Well-characterized endocytic targeting moieties potentially useful for nanocarrier-mediated drug delivery are folic acid, low-density lipoprotein, cholera toxin B, mannose-6-phosphate, transferrin, riboflavin, the tripeptide RGD, ICAM-1 antibody, and nicotinic acid, as recently reviewed by Bareford and Swaan [31]. The cellular internalization mechanisms utilized by these ligands involve clathrin-dependent RME, caveolin-assisted endocytosis, lipid raft associated endocytosis, and cell adhesion molecule (CAM) directed cellular uptake [31]. In addition to several approaches to exploiting the inherent tendency of nanoparticles to accumulate in the endolysosomal compartment for possible therapeutic purposes, there is a growing body of work that suggests the feasibility of modifying nanocarriers to redirect delivery of their cargo to other subcellular compartments as well. Liposomes modified with mitochondriotropic ligands have been shown to improve the efficacy of an anticancer drug both in vitro and in vivo [35]. AuNPs have already been targeted to the nucleus using the adenoviral nuclear localization signal (NLS) and integrin binding domain [36]. Such an approach has been reported to be useful in the development of probes for cell tracking by surface-enhanced Raman scattering [37]. Modification with a leader sequence peptide has also been applied to creating delivery systems for mitochondria. A mitochondrial leader peptide

6

AN INTRODUCTION TO SUBCELLULAR NANOMEDICINE

(MLP), derived from the nucleocytosol expressed but mitochondria localized ornithine transcarbamylase, has been reported to render polyethylene imine (PEI) mitochondriotropic and represents a potential approach for mitochondrial DNA delivery [38]. It is interesting to note that the examples discussed above share a common assumption. Nanocarriers are assumed to have a predisposition for the endolysosomal pathway by virtue of their nanometer size and, without a subcellular targeting ligand, all nanocarriers would remain in the endolysosomal compartment. However, it is interesting to also consider the disposition of a nanocarrier made exclusively of a molecule with a predisposition for a subcellular compartment. A good example is the mitochondriotropic amphiphile dequalinium chloride. A serendipitous discovery while screening mitochondriotropic drugs potentially able to interfere with the mitochondrial DNA metabolism in Plasmodium falciparum [39] revealed this self-associating tendency of dequalinium chloride and its ability to form vesicles. At the time of their discovery, these unusual vesicles were termed DQAsomes (pronounced dequasomes), that is, dequalinium (DQA) based liposome-like vesicles [40]. Based on the fact that these carriers were composed exclusively of mitochondriotropic molecules and that they were able to bind and protect DNA, DQAsomes were explored as potential mitochondria-specific DNA delivery vehicles for direct mitochondrial gene therapy [41–44]. DQAsomes have also been explored as a mitochondria-targeted nanocarrier system for small drug molecules, in particular, for anticancer drugs known to trigger apoptosis via direct action on mitochondria [45, 46]. It would therefore appear that, for now, a basic proof of concept for an alternative strategy toward the design of subcellular targeting nanocarriers seems to have been established. It is also obvious that in order to design similar carriers for other subcellular compartments it would be necessary to first find self-assembling molecules with an affinity for the intended subcellular compartment. To this end, recent work on the subcellular distribution of micelle-forming agents offers some interesting insights [47–51]. Based on the examples discussed so far, it would seem that there is indeed hope that nanocarrier systems could be designed to achieve true molecular level targeting inside cells. However, to say that these systems will be available soon is perhaps premature given what little we know about the subcellular dynamics associated with nanoparticle trafficking. There are in our opinion several unanswered questions. For example, do all nanocarriers remain intact upon cell entry and subsequent disposition? Are there differences in the disposition of vesicles in comparison to particles? What is the true influence of size on the intracellular disposition of various nanocarriers? Most important, however, is the question of the mechanism by which the nanocarrier is able to achieve selective uptake and delivery into the subcellular compartment. All the strategies described so far report observations of altered or improved subcellular accumulation that appears to result in improved activity, but how exactly this happens is still unclear. Do the nanocarriers remain intact upon

AN INTRODUCTION TO SUBCELLULAR NANOMEDICINE

7

internalization and then get trafficked as intact structures? If so, how is the therapeutic cargo released to the correct subcellular compartment? Alternatively, it could be imagined that once taken up into the early endosomal vesicle, the nanocarrier components undergo a redistribution to become part of the endosomal vesicle. There is some evidence to suggest that in fact cells actively traffic nanocarriers in cell membrane-derived vesicles [52]. Assuming the targeting ligand was able to redistribute to the surface of the endosomal vesicle, it might be possible then that the vesicle would have an altered subcellular fate that could involve transport to and association with a target compartment other than the lysosome. While this may seem to be farfetched speculation, there has already been some work along similar lines toward the development of nanocarrier systems for delivery of molecules to the nucleus and even the mitochondria. A strategy that involved stepwise membrane fusion was devised based on the premise that, to efficiently deliver DNA to the nucleus, a delivery system must penetrate through the plasma membrane, nuclear envelope, prior to DNA release in the nucleus. Using a multilayered nanoparticle called a Tetralamellar Multifunctional Envelopetype Nano Device (T-MEND) and consisting of a DNA-polycation condensed core coated with two nuclear membrane-fusogenic inner envelopes and two endosome-fusogenic outer envelopes, which are shed in stepwise fashion, transgene expression in nondividing cells was reported to be dramatically increased [53]. A similar approach in designing a mitochondria-specific delivery sytem has been reported as well. Liposomal carriers called MITO-Porters, which carry octaarginine surface modifications to stimulate their entry into cells as intact vesicles (via macropinocytosis) were prepared with lipid compositions that were identified in various experiments to promote both fusion with the mitochondrial membrane and the release of liposomal cargo to the intramitochondrial compartment in living cells. Using GFP protein as a model cargo, it was shown that MITO-Porter liposomes are able to selectively deliver their cargo to mitochondria [54, 55]. It is also interesting to note that changes in nanoparticle architecture result in changes in subcellular disposition [56]. Fluorescein isothiocyanate labeled layered double hydroxide (LDH) nanoparticles were prepared from Mg2Al under conditions that yielded either hexagonal sheets (50–150 nm wide and 10–20 nm thick) or nanorods (30–60 nm wide and 100–200 nm long). A comparison of the subcellular distribution of these two types of preparations revealed that the nanorods trafficked to the nucleus but the hexagonal sheets remained in the cytoplasm [56]. Not surprisingly, an active microtubule mediated transport process is hypothesized to be responsible for the observed rapid nuclear accumulation of the nanorods [56]. As discussed so far, various nanocarrier platforms have already undergone preliminary investigation for their ability to control the subcellular disposition of drugs and a potential improvement in therapy through the use of nanocarriers for subcellular targeting of bioactive molecules. A common paradigm that currently seems to apply to most of these approaches is the use of

8

AN INTRODUCTION TO SUBCELLULAR NANOMEDICINE

subcellular targeting ligands to control subcellular distribution. Given the relative ease of modifying nanocarriers with various surface functionalities and the fact that such approaches are already in use to achieve targeting at the cellular and organ levels, the ligand-based approach does seem to be a logical extension of current technology. It helps that new tools are concurrently being investigated to understand some of the physicochemical aspects of how small molecules [28, 29] as well as proteins [57] are able to selectively accumulate in certain subcellular compartments to afford the rational design of a wide repertoire of subcellular targeting ligands. However, one must be cautious in the knowledge that the approaches described so far are based only on current understanding of subcellular trafficking. Current knowledge of subcellular trafficking processes is based largely on studies with solid nanoparticles and on quantum dots [58–64]. Whether these observations can be extended to vesicular carriers like liposomes and micelles remains in question and in our opinion is due in large part to current limitations in imaging technology. We are however hopeful that technological advances in real-time fluorescence confocal imaging of live cells [65–68], as well as the emergence of new imaging techniques like total internal reflection microscopy [69] and label-free approaches like Raman microscopy [70–72], will allow some of the questions raised to be more satisfactorily answered. This book is our attempt to bring the best of current knowledge together to provide a comprehensive resource for anyone interested in this emerging area of drug delivery. The subsequent chapters discuss in full detail the current state of the art in the various approaches to nanocarrier-mediated bioactives to cell organelles as well as emerging research methods for the identification of subcellular targeting ligands and the study of subcellular transport processes.

REFERENCES 1. Ehrlich, P. Experimental Researches on Specific Therapy. On Immunity with Special Reference to the Relationship Between Distribution and Action of Antigens. London: Royal Institute of Public Health, 1908, p. 107. 2. Ganta, S., Devalapally, H., Shahiwala, A., and Amiji, M. A review of stimuliresponsive nanocarriers for drug and gene delivery. J. Control. Release 126: 187– 204 (2008). 3. Jabr-Milane, L., Van Vlerken, L., Devalapally, H., Shenoy, D., Komareddy, S., Bhavsar, M., and Amiji, M. Multi-functional nanocarriers for targeted delivery of drugs and genes. J. Control. Release 130: 121–128 (2008). 4. Mishra, B., Patel, B. B., and Tiwari, S. Colloidal nanocarriers: a review on formulation technology, types and applications toward targeted drug delivery. Nanomedicine 6(1): 9–24 (2010). 5. Torchilin, V. P. Targeted pharmaceutical nanocarriers for cancer therapy and imaging. AAPS J. 9: E128–147 (2007).

REFERENCES

9

6. Ma, L., Liu, Y. L., Ma, Z. Z., Dou, H. L., Xu, J. H., Wang, J. C., Zhang, X., and Zhang, Q. Targeted treatment of choroidal neovascularization using integrinmediated sterically stabilized liposomes loaded with combretastatin A4. J. Ocul. Pharmacol. Ther. 25: 195–200 (2009). 7. Hatakeyama, H., Akita, H., Ishida, E., Hashimoto, K., Kobayashi, H., Aoki, T., Yasuda, J., Obata, K., Kikuchi, H., Ishida, T., Kiwada, H., and Harashima, H. Tumor targeting of doxorubicin by anti-MT1-MMP antibody-modified PEG liposomes. Int. J. Pharm. 342: 194–200 (2007). 8. Nallamothu, R., Wood, G. C., Kiani, M. F., Moore, B. M., Horton, F. P., and Thoma, L. A. A targeted liposome delivery system for combretastatin A4: formulation optimization through drug loading and in vitro release studies. PDA J. Pharm. Sci. Technol. 60: 144–155 (2006). 9. Frisch, B., Carriere, M., Largeau, C., Mathey, F., Masson, C., Schuber, F., Scherman, D., and Escriou, V. A new triantennary galactose-targeted PEGylated gene carrier, characterization of its complex with DNA, and transfection of hepatoma cells. Bioconjug. Chem. 15: 754–764 (2004). 10. Torchilin, V. P., Levchenko, T. S., Lukyanov, A. N., Khaw, B. A., Klibanov, A. L., Rammohan, R., Samokhin, G. P., and Whiteman, K. R. p-NitrophenylcarbonylPEG-PE-liposomes: fast and simple attachment of specific ligands, including monoclonal antibodies, to distal ends of PEG chains via p-nitrophenylcarbonyl groups. Biochim. Biophys. Acta. 1511: 397–411 (2001). 11. Minton, A. P. How can biochemical reactions within cells differ from those in test tubes? J. Cell Sci. 119: 2863–2869 (2006). 12. Ellis, R. J. and Minton, A. P. Cell biology: join the crowd. Nature 425: 27–28 (2003). 13. Sanabria, H., Kubota, Y., and Waxham, M. N. Multiple diffusion mechanisms due to nanostructuring in crowded environments. Biophys. J. 92: 313–322 (2007). 14. Goins, A. B., Sanabria, H., and Waxham, M. N. Macromolecular crowding and size effects on probe microviscosity. Biophys. J. 95: 5362–5373 (2008). 15. Lukacs, G. L., Haggie, P., Seksek, O., Lechardeur, D., Freedman, N., and Verkman, A. S. Size-dependent DNA mobility in cytoplasm and nucleus. J. Biol. Chem. 275: 1625–1629 (2000). 16. Seksek, O., Biwersi, J., and Verkman, A. S. Translational diffusion of macromolecule-sized solutes in cytoplasm and nucleus. J. Cell Biol. 138: 131–142 (1997). 17. Weiss, M. Probing the interior of living cells with fluorescence correlation spectroscopy. Ann. N. Y. Acad. Sci. 1130: 21–27 (2008). 18. Duvvuri, M. and Krise, J. P. Intracellular drug sequestration events associated with the emergence of multidrug resistance: a mechanistic review. Front. Biosci. 10: 1499–1509 (2005). 19. Breunig, M., Bauer, S., and Goepferich, A. Polymers and nanoparticles: intelligent tools for intracellular targeting? Eur. J. Pharm. Biopharm. 68: 112–128 (2008). 20. Torchilin, V. P. Recent approaches to intracellular delivery of drugs and DNA and organelle targeting. Annu. Rev. Biomed. Eng. 8: 343–375 (2006). 21. Li, S. D. and Huang, L. Pharmacokinetics and biodistribution of nanoparticles. Mol. Pharm. 5: 496–504 (2008). 22. Kaufmann, A. M. and Krise, J. P. Lysosomal sequestration of amine-containing drugs: analysis and therapeutic implications. J. Pharm. Sci. 96: 729–746 (2007).

10

AN INTRODUCTION TO SUBCELLULAR NANOMEDICINE

23. Panyam, J. and Labhasetwar, V. Biodegradable nanoparticles for drug and gene delivery to cells and tissue. Adv. Drug Deliv. Rev. 55: 329–347 (2003). 24. Weissig, V. Mitochondrial-targeted drug and DNA delivery. Crit. Rev. Ther. Drug Carrier Syst. 20: 1–62 (2003). 25. Weissig, V. Targeted drug delivery to mammalian mitochondria in living cells. Expert Opin. Drug Deliv. 2: 89–102 (2005). 26. Weissig, V., Boddapati, S. V., Jabr, L., and D’Souza, G. G. Mitochondria-specific nanotechnology. Nanomedicine 2: 275–285 (2007). 27. Weissig, V., Cheng, S. M., and D’Souza, G. G. Mitochondrial pharmaceutics. Mitochondrion 3: 229–244 (2004). 28. Horobin, R. W. Uptake, distribution and accumulation of dyes and fluorescent probes within living cells: a structure-activity modelling approach. Adv. Colour. Sci. Technol. 4: 101–107 (2001). 29. Horobin, R. W., Trapp, S., and Weissig, V. Mitochondriotropics: a review of their mode of action, and their applications for drug and DNA delivery to mammalian mitochondria. J. Control. Release 121: 125–136 (2007). 30. Horobin, R. W. and Weissig, V. A QSAR-modeling perspective on cationic transfection lipids. 1. Predicting efficiency and understanding mechanisms. J. Gene Med. 7: 1023–1034 (2005). 31. Bareford, L. M. and Swaan, P. W. Endocytic mechanisms for targeted drug delivery. Adv. Drug Deliv. Rev. 59: 748–758 (2007). 32. Gregoriadis, G. and Ryman, B. E. Liposomes as carriers of enzymes or drugs: a new approach to the treatment of storage diseases. Biochem. J. 124: 58P (1971). 33. Castino, R., Demoz, M., and Isidoro, C. Destination “lysosome”: a target organelle for tumour cell killing? J. Mol. Recognit. 16: 337–348 (2003). 34. Tate, B. A. and Mathews, P. M. Targeting the role of the endosome in the pathophysiology of Alzheimer’s disease: a strategy for treatment. Sci. Aging Knowledge Environ. 2006(10): 2 (2006). 35. Boddapati, S. V., D’Souza, G. G., Erdogan, S., Torchilin, V. P., and Weissig, V. Organelle-targeted nanocarriers: specific delivery of liposomal ceramide to mitochondria enhances its cytotoxicity in vitro and in vivo. Nano. Lett. 8: 2559–2563 (2008). 36. Tkachenko, A. G., Xie, H., Liu, Y., Coleman, D., Ryan, J., Glomm, W. R., Shipton, M. K., Franzen, S., and Feldheim, D. L. Cellular trajectories of peptide-modified gold particle complexes: comparison of nuclear localization signals and peptide transduction domains. Bioconjug. Chem. 15: 482–490 (2004). 37. Xie, W., Wang, L., Zhang, Y., Su, L., Shen, A., Tan, J., and Hu, J. Nuclear targeted nanoprobe for single living cell detection by surface-enhanced Raman scattering. Bioconjug. Chem. 20: 768–773 (2009). 38. Lee, M., Choi, J. S., Choi, M. J., Pak, Y. K., Rhee, B. D., and Ko, K. S. DNA delivery to the mitochondria sites using mitochondrial leader peptide conjugated polyethylenimine. J. Drug Target. 15: 115–122 (2007). 39. Rowe, T. C., Weissig, V., and Lawrence, J. W. Mitochondrial DNA metabolism targeting drugs. Adv. Drug Deliv. Rev. 49: 175–187 (2001).

REFERENCES

11

40. Weissig, V., Lasch, J., Erdos, G., Meyer, H. W., Rowe, T. C., and Hughes, J. DQAsomes: a novel potential drug and gene delivery system made from Dequalinium. Pharm. Res. 15: 334–337 (1998). 41. D’Souza, G. G., Boddapati, S. V., and Weissig, V. Mitochondrial leader sequence— plasmid DNA conjugates delivered into mammalian cells by DQAsomes colocalize with mitochondria. Mitochondrion 5: 352–358 (2005). 42. D’Souza, G. G., Rammohan, R., Cheng, S. M., Torchilin, V. P., and Weissig, V. DQAsome-mediated delivery of plasmid DNA toward mitochondria in living cells. J. Control. Release 92: 189–197 (2003). 43. Weissig, V., D’Souza, G. G., and Torchilin, V. P. DQAsome/DNA complexes release DNA upon contact with isolated mouse liver mitochondria. J. Control. Release 75: 401–408 (2001). 44. Weissig, V., Lizano, C., and Torchilin, V. P. Selective DNA release from DQAsome/ DNA complexes at mitochondria-like membranes. Drug Deliv. 7: 1–5 (2000). 45. Cheng, S. M., Pabba, S., Torchilin, V. P., Fowle, W., Kimpfler, A., Schubert, R., and Weissig, V. Towards mitochondria-specific delivery of apoptosis-inducing agents: DQAsomal incorporated paclitaxel. J. Drug Deliv. Sci. Tech. 15: 81–86 (2005). 46. D’Souza, G. G., Cheng, S. M., Boddapati, S. V., Horobin, R. W., and Weissig, V. Nanocarrier-assisted sub-cellular targeting to the site of mitochondria improves the pro-apoptotic activity of paclitaxel. J. Drug Target. 16: 578–585 (2008). 47. Bae, Y., Nishiyama, N., Fukushima, S., Koyama, H., Yasuhiro, M., and Kataoka, K. Preparation and biological characterization of polymeric micelle drug carriers with intracellular pH-triggered drug release property: tumor permeability, controlled subcellular drug distribution, and enhanced in vivo antitumor efficacy. Bioconjug. Chem. 16: 122–130 (2005). 48. Maysinger, D., Lovric, J., Eisenberg, A., and Savic, R. Fate of micelles and quantum dots in cells. Eur. J. Pharm. Biopharm. 65: 270–281 (2007). 49. Savic, R., Azzam, T., Eisenberg, A., and Maysinger, D. Assessment of the integrity of poly(caprolactone)-b-poly(ethylene oxide) micelles under biological conditions: a fluorogenic-based approach. Langmuir 22: 3570–3578 (2006). 50. Savic, R., Azzam, T., Eisenberg, A., Nedev, H., Rosenberg, L., and Maysinger, D. Block-copolymer micelles as carriers of cell signaling modulators for the inhibition of JNK in human islets of Langerhans. Biomaterials 30(21): 3597–3604 (2009). 51. Xiong, X. B., Mahmud, A., Uludag, H., and Lavasanifar, A. Multifunctional polymeric micelles for enhanced intracellular delivery of doxorubicin to metastatic cancer cells. Pharm. Res. 25: 2555–2566 (2008). 52. Ruan, G., Agrawal, A., Marcus, A. I., and Nie, S. Imaging and tracking of tat peptide-conjugated quantum dots in living cells: new insights into nanoparticle uptake, intracellular transport, and vesicle shedding. J. Am. Chem. Soc. 129: 14759– 14766 (2007). 53. Akita, H., Kudo, A., Minoura, A., Yamaguti, M., Khalil, I. A., Moriguchi, R., Masuda, T., Danev, R., Nagayama, K., Kogure, K., and Harashima, H. Multilayered nanoparticles for penetrating the endosome and nuclear membrane via a step-wise membrane fusion process. Biomaterials 30: 2940–2949 (2009).

12

AN INTRODUCTION TO SUBCELLULAR NANOMEDICINE

54. Yamada, Y. and Harashima, H. Mitochondrial drug delivery systems for macromolecule and their therapeutic application to mitochondrial diseases. Adv. Drug Deliv. Rev. 60: 1439–1462 (2008). 55. Yamada, Y., Akita, H., Kamiya, H., Kogure, K., Yamamoto, T., Shinohara, Y., Yamashita, K., Kobayashi, H., Kikuchi, H., and Harashima, H. MITO-Porter: a liposome-based carrier system for delivery of macromolecules into mitochondria via membrane fusion. Biochim. Biophys. Acta. 1778: 423–432 (2008). 56. Xu, Z. P., Niebert, M., Porazik, K., Walker, T. L., Cooper, H. M., Middelberg, A. P., Gray, P. P., Bartlett, P. F., and Lu, G. Q. Subcellular compartment targeting of layered double hydroxide nanoparticles. J. Control. Release 130: 86–94 (2008). 57. Nakai, K. and Horton, P. Computational prediction of subcellular localization. Methods Mol. Biol. 390: 429–466 (2007). 58. Yacobi, N. R., Malmstadt, N., Fazlollahi, F., Demaio, L., Marchelletta, R., HammAlvarez, S. F., Borok, Z., Kim, K. J., and Crandall, E. D. Mechanisms of alveolar epithelial translocation of a defined population of nanoparticles. Am. J. Respir. Cell. Mol. Biol. 42(5): 604–614 (2009). 59. Hillaireau, H. and Couvreur, P. Nanocarriers entry into the cell: relevance to drug delivery. Cell. Mol. Life Sci. 66(17): 2873–2896 (2009). 60. Smirnov, P. Cellular magnetic resonance imaging using superparamagnetic anionic iron oxide nanoparticles: applications to in vivo trafficking of lymphocytes and cell-based anticancer therapy. Methods Mol. Biol. 512: 333–353 (2009). 61. Harush-Frenkel, O., Altschuler, Y., and Benita, S. Nanoparticle-cell interactions: drug delivery implications. Crit. Rev. Ther. Drug Carrier Syst. 25: 485–544 (2008). 62. Huser, T. Nano-biophotonics: new tools for chemical nano-analytics. Curr. Opin. Chem. Biol. 12: 497–504 (2008). 63. Vasir, J. K. and Labhasetwar, V. Quantification of the force of nanoparticle–cell membrane interactions and its influence on intracellular trafficking of nanoparticles. Biomaterials. 29: 4244–4252 (2008). 64. Rajan, S. S., Liu, H. Y., and Vu, T. Q. Ligand-bound quantum dot probes for studying the molecular scale dynamics of receptor endocytic trafficking in live cells. ACS Nano. 2: 1153–1166 (2008). 65. Perrine, K. A., Lamarche, B. L., Hopkins, D. F., Budge, S. E., Opresko, L. K., Wiley, H. S., and Sowa, M. B. High speed method for in situ multispectral image registration. Microsc. Res. Tech. 70: 382–389 (2007). 66. Rabut, G. and Ellenberg, J. Automatic real-time three-dimensional cell tracking by fluorescence microscopy. J. Microsc. 216: 131–137 (2004). 67. Sunaguchi, M., Nishi, M., Mizobe, T., and Kawata, M. Real-time imaging of green fluorescent protein-tagged beta 2-adrenergic receptor distribution in living cells. Brain Res. 984: 21–32 (2003). 68. Jester, J. V., Andrews, P. M., Petroll, W. M., Lemp, M. A., and Cavanagh, H. D. In vivo, real-time confocal imaging. J. Electron Microsc. Tech. 18: 50–60 (1991). 69. Byrne, G. D., Pitter, M. C., Zhang, J., Falcone, F. H., Stolnik, S., and Somekh, M. G. Total internal reflection microscopy for live imaging of cellular uptake of sub-micron non-fluorescent particles. J. Microsc. 231: 168–179 (2008).

REFERENCES

13

70. Matthaus, C., Kale, A., Chernenko, T., Torchilin, V., and Diem, M. New ways of imaging uptake and intracellular fate of liposomal drug carrier systems inside individual cells, based on Raman microscopy. Mol. Pharm. 5: 287–293 (2008). 71. Fujita, K. and Smith, N. I. Label-free molecular imaging of living cells. Mol. Cells 26: 530–535 (2008). 72. Freudiger, C. W., Min, W., Saar, B. G., Lu, S., Holtom, G. R., He, C., Tsai, J. C., Kang, J. X., and Xie, X. S. Label-free biomedical imaging with high sensitivity by stimulated Raman scattering microscopy. Science 322: 1857–1861 (2008).

CHAPTER 2

Delivery of Nanosensors to Measure the Intracellular Environment PAUL G. COUPLAND and JONATHAN W. AYLOTT School of Pharmacy, University of Nottingham, Nottingham, United Kingdom

2.1

INTRODUCTION

Nanosensors can be internalized in viable cells, providing an optical signal to report on the intracellular environment. This allows an understanding of intracellular function to be gained in real time. Nanosized sensors are especially desirable as they provide an output with minimal cellular disruption, which enhances the quality of data generated from each cell. Chemical perturbation is also minimized by the structural design of a nanosensor; it is these characteristics of nanosized sensing particles that give the potential for observation that is minimally invasive [1–5]. In recent years, new nanosized sensing particles have been demonstrated and reported widely in the literature. Borisov and Klimant [6] provide a good summary for the state of the art of optical nanosensor technology and its increasing use in scientific fields, including biology, biotechnology, and clinical medicine. In this chapter we will focus our discussion on nanosensors of a single polymeric matrix, for example, polyacrylamide or silica sol-gel, and the delivery methods used for intracellular translocation of these nanoparticles. The delivery principles discussed here are transferable to many types of nanoparticle systems currently being researched. The most common technique for intracellular research utilizes fluorescent dyes in combination with confocal microscopy [7]. The small size of the free sensing molecule is a key advantage, providing high spatial resolution and allowing information throughout the cell to be collected en masse. A problem faced when using free dye in direct contact with the intracellular environment of the cell is chemical perturbation due to any cytoxic action of the dye molecules [5]. Equally, the intracellular environment can impact on the efficacy Organelle-Specific Pharmaceutical Nanotechnology, Edited by Volkmar Weissig and Gerard G. M. D’Souza Copyright © 2010 John Wiley & Sons, Inc.

15

16

DELIVERY OF NANOSENSORS TO MEASURE THE INTRACELLULAR ENVIRONMENT

of the sensing molecules [8, 9], where random protein binding and intracellular sequestration are major complications [10] as well as simply retaining the dye within the cell during the lifetime of an experiment [11]. To overcome the analytical difficulties of using free sensing dyes, an alternative approach evolved—optochemical sensors. These are typically a pulled optical fiber with a modified polymeric tip, to entrap sensing dye molecules, inserted directly into the cell [12, 13]. Pulled fiber sensors provide a biocompatible platform, protecting the dye and the cell from one another, and the polymer matrix of the modified tip facilitates complex sensing schemes to be devised by closely packing elements that interact complementarily [14, 15]. However, the large size of the sensor relative to a single cell causes damaging physical perturbation to all but the largest cells. This is due to the physical pathway required to transmit the signal generated at the tip of the sensor to the detector. Additionally, only low spatial resolution is achievable as few optodes can be inserted into a cell before it is damaged irreversibly [10, 16]. Fiber-optic sensors and free-dye molecules have advantages over each other, yet both have the limitations described that prevents them from being ideally suited to single cell intracellular measurement. A new technique that retained the benefits of free dyes (small size, good spatial resolution) and optochemical sensors (biocompatible dye entrapment, complex sensing schemes) that prevented chemical and physical perturbation of the cell was required. The development of PEBBLE (probes encapsulated by biologically localized embedding) nanosensors by Kopelman [1] was the next step in miniaturization of pulled fiber optochemical sensors and removed the fiber-optic component used for signal transmission. Understanding that the tip of the fiber-optic sensor was all that was needed for sensing, and recognizing that fluorescence microscopy was ideally suited to record the fluorescence output, Kopelman undertook to fabricate the same polymer tip as a single unit. Typically sub 200 nm in diameter, these sensors occupy approximately 1 ppb of a typical mammalian cell and cause negligible physical perturbation [4]. Nanosensors resemble the tip of an optode, in that they consist of a polymer matrix surrounding chemical sensing elements; there is no longer a physical signal path—thus being more akin to free dyes—but the protective capacity of a biocompatible polymer matrix is retained. Chemical perturbation to the cell is minimized, as is the effect of the intracellular environment upon the efficiency of the reporter molecule; in essence, the polymer matrix protects the cell from the dye and the dye from the cell [1]. Fluorophores usually affected through random protein binding respond characteristically when entrapped within the polymer matrix of a nanosensor [17, 18]. The matrix also allows for enhanced longevity of the sensing components (by protecting the dye from protein binding) enabling fully reversible sensors to be produced [3]. The nanosensor matrix is a discrete domain in which components making up a more complex sensing regime can be located. This allows for highly selective sensors incorporating ratiometric sensing regimes that use more than one dye [4], or electron transfer regimes using optically silent ionophores, in conjunc-

INTRODUCTION

17

tion with fluorophore partners [15]; significantly, the use of a sensor matrix increases the scope of analysis. Nanosensors retain the spatial resolution advantage of free dyes rather than being limited to the immediate locality of a single optode tip, because many nanosensors can be loaded into a single cell. In brief, a nanosensor is a nanometer-size polymeric spherical matrix in which chemical sensing elements/fluorophores are entrapped. The polymer matrix is porous, allowing diffusion of analytes to interact with the entrapped fluorophores, giving a fluorescent response that is captured with an optical system [19, 20]. By minimizing the chemical effects of free dyes and the physical effects of bulky optodes, a more natural state of the intracellular environment can be investigated [21]. Nanosensors are most commonly fabricated from polyacrylamide [2], silicate sol-gels [5], or liquid polymer matrices [22], each being porous, photostable, optically transparent, and chemically inert. Functionalization of the nanosensor matrix is possible when required, to provide chemical functional groups on the nanosensor surface for subsequent modification and/or conjugation. The variety of matrices available for nanosensor fabrication and their chemical diversity allow for a flexible and modular approach to nanosensor design (Figure 2.1). O NH

2

O

O

O

N

N H

O N H

O

R functional matrix, chemical groups

lon+ polymer matrix, (silica sol-gel, polyacrylamide, liquid polymer)

CPP attachment

NH

O

O

O

N O NO H

O O

O N

N HO

S

NH 2

O

ionic matrix, comonomer additives +/–

SP* Lipid Multi-layer multifunction surface chemistry

O 2 NH

S

Fluorophore(s) (incl. FRET pairing), ionophores, enzymes hv

O

S

hv

O 2

S

O

N

Nanosensor function

PEG CPP *Signal Peptide

Figure 2.1. Nanosensor schematic. (See color insert.)

Encapsulation in liposomes

18

DELIVERY OF NANOSENSORS TO MEASURE THE INTRACELLULAR ENVIRONMENT

The surface of the nanosensor is important due to its involvement in a number of biophysical interactions. Modification of the nanosensor surface with chemical moieties to provide a specific function allows for application to different cell types, delivery methods, and intracellular tasks. Modification of the nanosensor matrix can be achieved, for example, by the inclusion of an amine-modified acrylamide into a polyacrylamide matrix, providing primary amines throughout the nanosensor matrix [17]. Various molecules, from heterobifunctional linkers to phospholipids, can then be conjugated to the nanosensor, providing additional functionality and capability. As with all emerging technologies, continued development and improvements to the processes and protocols are ongoing. For nanosensors, one approach to improve their reporting of intracellular function and condition is by increasing the brightness of the signal emitted from each individual nanosensor. This is important in the realm of delivery and analytical performance, as brighter nanosensors will generate the same amount of signal from a cell with a lower nanosensor loading. This of course reduces the potential for physical disruption to the cell during and after delivery. Alternative strategies for attachment of fluorophores to the nanosensor surface, or entrapping them in the polymer matrix, to yield much brighter nanosensors are being developed. Historically, the most common approach is to physically entrap fluorophores into the polymer matrix. This method typically requires that fluorophores are attached to a large, inert molecule such as dextran, which will prevent subsequent leaching of the entrapped fluorophore from the polymer matrix. An alternative methodology exists for incorporating fluorophores in nanosensors, whereby the fluorophores are covalently attached directly to the backbone of the polymer matrix [23]. This eliminates the requirement of dextran altogether and allows for the potential incorporation of greater numbers of fluorophores by several orders of magnitude. However, indiscriminate attachment to the functional groups in the polymer matrix includes those on the nanosensor surface and as the nanosensor matrix provides a protective role, attachment of fluorophores on the nanosensor surface will contradict the idea of the protective capacity of a nanosensor. To prevent unwanted interaction between the fluorophore and its environment, additional steps are needed to remove or mask these surface-located fluorophores. Steps such as enzyme-mediated fluorophore removal would be required, where the fluorophores are attached via a cleavable linker allowing surface bound fluorophores to be easily removed [24]. Approaches to enhance the brightness of nanosensing particles have been investigated by other groups including the Wiesner group at Cornell University, who introduced superbright C-dots with brightness levels and enhanced photostability approaching those reported for semiconductor quantum dots [25]. C-dots are core-shell particles based on a dye-rich silica core surrounded by a pure silica shell. The shell acts as a protective layer similar to the polyacrylamide or sol-gel matrix of the nanosensors described here. Also, there are investigations into using quantum dots as intracellular probes and sensors [26–28]. The enhancement

INTRACELLULAR DELIVERY

19

in brightness and functionality of nanosensors and other types of nanosensing particles is varied and promising but for each type of nano(sensor)particle the challenges of delivery remain.

2.2

INTRACELLULAR DELIVERY

The nanosensor is a tool that has been designed to observe and analyze intracellular domains with minimal cellular perturbation. However, one of the major challenges to the widespread uptake of nanosensor technology has been the lack of a generic delivery system to translocate the nanosensor across the cell membrane and to the intracellular domain of choice. There are several techniques by which nanoparticles can be delivered into cells in culture; historically, many have been developed for the delivery of plasmids and oligonucleotides in molecular biology, or for the uptake of molecules packaged within organic polymer structures developed for drug delivery. The various delivery techniques investigated can be designated into two main categories: mechanical and membranal. In the mechanical category each delivery technique uses a mechanical and physical force to insert the nanosensors into the cells, for example, using a needle to inject them directly or a pulse of electricity to disrupt the cellular membrane. Membranal techniques are based on an interaction at the biological level between the cellular membrane and a chemical group on the nanosensor surface that aids the nanosensor in its translocation to the intracellular environment. 2.2.1 Mechanical Methods 2.2.1.1 Gene Gun Bombardment One of the most commonly used nanosensor delivery methods to date is gene gun delivery. Originally developed in molecular biology research for the delivery of DNA into living cells, it was easily adapted for nanosensor delivery [2]. The gene gun method can be described as a shotgun method in which nanosensors are literally fired into cells. Nanosensors are layered onto a disk or the inner surface of a small length of tube (to create a bullet) depending on the model of gene gun, by applying a thin layer of suspended nanosensors in solution and allowing it to dry. Pressurized helium is then used to propel the nanosensors off the disk or from the prepared bullet into the cell culture. It is reported that by controlling the original concentration of nanosensors in the suspension when layered onto the carrier it is possible to deliver a single sensor or thousands of sensors into the cell culture sitting a short distance away. It is also argued [2, 5] that certain levels of spatial delivery are possible by altering the helium pressure and the distance from carrier disk to cell culture. As the nanosensors are intrinsically fired in a random pattern, this only serves to selectively choose the distance the nanosensors will travel through the cell—with a certain amount of control

20

DELIVERY OF NANOSENSORS TO MEASURE THE INTRACELLULAR ENVIRONMENT

and practice (and luck!), this will allow one to deliver sensors to the nucleus or have them remain in the cytosol. Great care is needed to achieve the fine window of conditions that accomplishes cellular penetration without dislodging the cells from the culture plate. 2.2.1.2 Picoinjection Picoliter injection is a technique that has been applied to nanosensor internalization; physical damage and cellular distress are the main problems encountered during delivery. Needles are formed using a standard glass pipette puller and are used to inject a nanosensor solution directly into the cell. Generally, cell viability is good although normal cellular activity is likely affected and there is always the chance the cell becomes irreparably damaged. The deliverable range of nanosensor concentration is broad as this method dopes the cell with increasing numbers of nanosensorsolution aliquots until the desired level of concentration is reached. This method requires a high level of operator skill and is a batch process, meaning only one cell at a time can be loaded [9]. These limitations have led to this technique being used mostly in specialized applications where cell-by-cell analysis of a small number of cells is required, for example, embryology [21]. 2.2.1.3 Electroporation and Sonication These two methods have been used successfully for the delivery of various molecules into cells, such as plasmid DNA, for many years [29, 30]. One study suggests specifically that electroporation primarily transports molecules across the plasma membrane, because its mechanism is specific to lipid bilayer disruption, whereas sonication transports molecules across both the plasma membrane and cell walls, for example, in algae and plants, because it nonspecifically disrupts cell-surface barriers [31]. In electroporation an electrical pulse temporarily permeabilizes the phospholipid bilayer of the plasma membrane. As this technique bypasses the endocytic pathway, nanosensors can be delivered directly to the cytoplasm with no need for subsequent endosomal escape, which is a clear benefit when, for example, considering CPP mediated delivery, described later in this chapter. However, as the cells are both permeabilized and subjected to very strong electrical pulses, there is usually a high rate of cellular mortality, which is often not reported [28]. 2.2.1.4 Patch Clamp and Scanning Ion Conductance Microscopy (SICM) Both these methods require a pulled micro/nanopipette to be held in contact, or at least in very close proximity, to the cell membrane. The notion of using these techniques as methods for nanosensor delivery is simple; with slight modification to each technique, a nanosensor-loaded solution is contained within the pipette as it is brought to the cell membrane. During patch clamping transient cell membrane disruption caused by creating an electrical current allows a temporary route for nanosensors to internalize. Similarly, in SICM the scanning tip can be used to provide a disruptive electrical field, which again disrupts the phospholipid bilayer, allowing the surrounding

INTRACELLULAR DELIVERY

21

nanosensors intracellular access. A major drawback to these techniques is the batch one-cell-at-a time nature of the technique. 2.2.1.5 Scrape Loading This procedure is used to transiently disrupt cell membranes and allow direct cell loading. This is achieved by drawing a tool (e.g., a glass rod with rubber end known as a rubber policeman) across an adherent population of cells, which scrapes them off the dish; the disrupted membranes temporarily allow various molecules [32], macromolecules [33], and nanoparticles [34] to cross this usually highly selective barrier. The method has not found great favor, however, due to the impracticality of severely reduced cell viability, approximately 50%, and because the method results in loaded cells that are in suspension. They must be replated and allowed to recover and spread, which is a serious disadvantage, as during this stage the labile probes/particles may be subjected to cytoplasmic redistribution and degradation. 2.2.2 Membranal Techniques The dynamic nature of the endomembranal system is a consistent theme in different modes of intracellular delivery: specifically, sequestration through phagocytosis and pinocytosis in macrophages, liposomal delivery, and cell penetrating peptide mediated delivery. 2.2.2.1 Phagocytosis (Macrophages) and Pinocytosis Phagocytosis or “cell-eating” is the uptake of solid material by a cell. Only a few kinds of cells display this behavior, including protists (e.g., amebas and ciliates) and the phagocytic white blood cells of animals. The protists use it for feeding while the white blood cells use it as means of capturing bacteria and cleaning cell debris from infection sites. In phagocytosis the cellular membrane actively surrounds the solid object to be taken up, eventually pinching off internally to form a phagocytic vesicle, or phagosome. In general, this fuses with a lysosome whereupon the contents are digested or recycled. Macrophages have been shown to take up nanosensors from a surrounding solution, providing an easy delivery method for these types of cells [10]. The number of nanosensors internalized is a factor of the original concentration of nanosensors in the solution and how long the macrophages remain in solution. Cell viability is excellent. The disadvantages are that this technique is highly selective for macrophages, the sensors end up in certain cellular regions, as directed by the cell, and these specialized immune system cells can be difficult to culture. Pinocytosis or “cell-drinking” is the uptake of fluids from the outside of the cell. Here the plasma membrane tends to fold inwards, forming small vesicles containing the material. As in phagocytosis, the vesicle generally fuses with a lysosome and digestion of the contents takes place. Unlike receptor-mediated endocytosis, however, pinocytosis is nonspecific with regard to the materials taken up. Cell biologists first demonstrated pinocytosis by exposing amebas

22

DELIVERY OF NANOSENSORS TO MEASURE THE INTRACELLULAR ENVIRONMENT

to concentrated protein solutions and observing protein uptake. Effective uptake of nanoparticles via pinocytosis has been demonstrated in mouse macrophages [35] and for nanosensors with mouse embryos [21]. 2.2.2.2 Liposomal Delivery Commercially available liposomes, such as Lipofectamine 2000, Fugene, and Escort, can be used to produce nanosensor– liposome complexes. The technique relies on the natural fusion of liposomes with certain cells, which occurs due to the structural similarities between liposome or lipid micelle and the phosholipid bilayer of the cell membrane. The nanosensor–liposome complex fuses with the cell membrane and nanosensors, internalized or bound to the wall of the liposomes, can become internalized within the cell. With cell-specific tailoring of cell to nanosensor to liposome concentration a wide range of nanosensor concentrations can be delivered in a similar fashion to the macrophage system, specifically the concentration of the original nanosensor solution and of nanosensor–liposome complexes incubated with the cells is important for achieving defined final concentrations [9]. Delivery to the cytosol is possible, which is a great benefit of this technique; this can be demonstrated with pH measurements of the cytoplasmic environment of approximately pH 7 (cytoplasmic pH is similar to cell culture media pH [36, 37]), indicating that sensors are not sequestered into acidified vesicles (endosomes/lysosomes). 2.2.2.3 Cell-Penetrating Peptides Cell-penetrating peptides (CPPs) are individual short peptides able to translocate across cell membranes. They are being studied widely because of their ability to aid translocation and drive the uptake of varied cargos into mammalian cells. There are several reviews on the potential and advantages of using CPPs as delivery vectors, in particular, therapeutic agents for human treatment [38–41]. One example illustrates how the functional efficacy of therapeutic agents would be greatly enhanced through the controlled import and intracellular targeting to specific organelles [41]. The value of using CPPs as delivery agents became more appealing as it emerged that CPPs could offer targeted delivery and that the short peptide structure of a CPP could be tailored for improved delivery strategies. CPP endomembranal delivery may prove to be an ideal method that is near silent— for both chemical and physical disruption. The effective cell translocation of the CPP nature was observed prior to the isolation of an individual peptide; this is because CPP sequences in nature form part of a larger protein. It was the observation that some proteins were able to translocate across cellular membranes by a process known as protein transduction that initiated the search for the element of each protein that was responsible. Several proteins are well known for this behavior, for example, human immunodeficiency virus (HIV-1) Tat protein [42], herpes simplex virus (HSV-1) VP22 protein [43], and Drosophila Antennapedia homeoprotein [44]. A portion of each protein associated with intracellular translocation was found through cleavage or site-directed mutagenesis, identifying the amino

INTRACELLULAR DELIVERY

23

acids required for membrane translocation. This part of the protein was referred to as the “protein-transduction domain” (PTD) of each protein. The approach revealed PTDs that translocated the entire protein but also retained their activity in isolation, when either removed from the native protein or synthesized in the laboratory. In isolation, therefore, the PTD of a protein becomes a CPP [45]. Derossi et al. [46] working on the Drosophila Antennapedia homeoprotein revealed the PTD through site-directed mutagenesis, discovering the driving force behind the internalization of this protein; the third helix was essential for membrane translocation. Importantly, however, it was observed that the third helix was also sufficient for membrane translocation alone, and from this work it was possible to develop a 16 amino acid long CPP, referred to as penetratin (pAntp) [44, 47]. In a similar approach, Green and Loewenstein had identified an 86 amino acid section of the HIV-1 Tat protein that was internalized readily by viable cells [42]. Currently, CPPs are defined as peptides with a maximum of 30 amino acids, which are able to enter cells in a seemingly energy-independent manner, thus being able to translocate across membranes in a nonendocytotic fashion [48], although the exact biochemical and biophysical mechanisms involved in membrane translocation are yet to be established and debate is ongoing in this area of CPP research. It was thought that endocytosis did not play a part due to the evidence that peptide internalization occurred at 4 °C—because all active transport mechanisms involving endocytosis do not function at this low temperature. Also, CPPs being taken up by different cells and tissue types suggests a common internalization method, which suggests binding to conserved cell membrane determinants [49]. However, data from several groups argues for energydependent processes [50]. One regular feature of all CPPs and PTDs is the higher than average constitute of cationic amino acids [51]. This suggests that in some way at least the initiation of cellular uptake comes from ionic interactions between the cationic CPP and the anionic lipid membrane, although it is known that charge alone is not a sufficient driving force to induce translocation [40]. Recent literature describes a principal role for arginine in the CPP [52] and studies into the various permutations of Tat and poly-R-PTD (arginine-rich protein transduction domains) suggested an essential role of the guanidinium head-groups of the arginine residues for cellular uptake [52]. A good review of the delivery of quantam dots to cells, which describes the use of CPP in some detail, is provided by Delehanty et al. [28]. In our research we conjugated Tat peptide to the surface of aminefunctionalized polyacrylamide nanosensors for delivery to a variety of different cell lines, including CHO-K1, GH4 pituitary lactotrope, A172 human glioblastoma, and human embryonic stem cells [53] (Figure 2.2). It was demonstrated that the hydrodynamic radius of the cargo played a significant role in affecting the route of internalization. The function of the nanosensors was to determine pH and we showed that the nanosensors resided in vesicles with a measured pH of 5, postdelivery. These were presumed to be lysosomes, as they could be co-stained with LysoTracker Red® as shown in Figure 2.3. In

24

DELIVERY OF NANOSENSORS TO MEASURE THE INTRACELLULAR ENVIRONMENT

Figure 2.2. Confocal fluorescence microscopy images of Tat-conjugated nanosensors delivered to (from left): CHO cells incubated with Ca2+-sensitive nanosensors (green) and DRAQ5 nuclear stain (blue); GH4 pituitary lactotrope cells loaded with Tat conjugated Ca2+-sensitive nanosensors; A172 human glioblastoma cell loaded with fluorescein nanosensors; human mesenchymal stem cells (hMSCs) loaded with rhodamine B nanosensors (red) and co-stained with CD105:FITC membrane stain (green) and DRAQ5 nuclear stain (blue). (See color insert.)

Figure 2.3. Confocal fluorescence microscopy images of Chinese hamster ovary K1 (CHO-K1) cells containing: (left) Tat-FITC conjugate (green) which do not show colocalized fluorescence with endosomes/lysosomes stained with LysoTracker Red® when loaded into CHO-K1 cells; (middle) Tat-conjugated nanosensors and LysoTracker Red® showing colocalized fluorescence (orange color) indicating that the nanosensors are residing in acidified endosomes: (right) Tat-FITC conjugate (green) and Tat-TRITC-nanosensors (red) loaded into a single cell. The overlay of these two channels shows no colocalization of fluorescence of the two different cargo sizes mediated by Tat peptide delivery. (See color insert.)

contrast to this, fluorescently labeled Tat, which had a much smaller hydrodynamic radius, localized to different vesicles in the cytoplasm and did not colocalize with the LysoTracker Red® dye. It was apparent that the difference in size of cargo that the Tat peptide was conjugated to seemed to directly affect the mechanism of translocation into the cell. The precise mechanisms involved following the first ionic interaction will continue to be debated; they may consist of currently unknown mechanisms or, simply, endocytosis. It is

INTRACELLULAR TRAFFICKING, NANOSENSOR FATE, AND STRATEGIES

25

feasible, considering the physicochemical diversity of CPPs, that diverse mechanisms of uptake account for cellular uptake of different peptides. Even so, CPPs continue to be a useful research tool, having been demonstrated as delivery vectors for a variety of molecules and nanosized particles [54–57].

2.3 INTRACELLULAR TRAFFICKING, NANOSENSOR FATE, AND STRATEGIES FOR TARGETED DELIVERY After delivery, indicating the crossing of the cell membrane, the final location of nanosensors is often directed by cellular pathways that deal with the movement and trafficking of internalized particles and vesicles. As shown in Figure 2.3, CPP delivered nanosensors were found to locate to lysosomes within a few hours after incubation. This therefore means we have to carefully consider mechanisms by which the nanosensors can be directed to specific locations within the cell after internalization. Frequently, the required location of the delivered nanosensors is to reside freely in the cytoplasm without being sequestered into endosomal vesicles, for example, when wishing to monitor ion flux throughout the cell. Alternatively, the ultimate target may be a specific organelle. Either way, controlling the fate of the nanosensor is a detail that needs attention when developing a delivery system. 2.3.1 Cytochalasin-D Cytochalasin-D is a cell-permeable fungal toxin from Zygosporium mansonii and an inhibitor of actin microfilament function. It binds to the barbed end of actin filaments, inhibiting the association and dissociation of subunits and causing disruption to the actin filaments and inhibition of actin polymerization. This action of cytochalasin-D is known to disrupt the fusion of the endosomes with lysosomes in late stage endocytosis [58], which is interesting with regard to nanoparticle delivery as internalization into lysosomes after translocating the membrane is limiting. Preincubation with cytochalasin-D has been shown to prevent nanoparticles being ultimately digested in lysosomes by blocking the fusion of endosomes with lysosomes [59]. For some delivery methods (e.g., CPP), this approach will only halt the endocytic pathway at the endosome stage; but this at least provides a point at which efforts can be focused on working to release the nanosensors from the endosome, before they are recycled out of the cell. 2.3.2 Targeted Delivery For targeted delivery to specific intracellular organelles, like mitochondria, targeting molecules can be located on the nanosensor surface. These targeting molecules, often signaling peptides, are used ubiquitously in the cell to allow cell machinery to correctly transport proteins within the cellular environment

26

DELIVERY OF NANOSENSORS TO MEASURE THE INTRACELLULAR ENVIRONMENT

[60]. Many examples have been found in cytology for protein translocation, including the sorting of proteins to peroxisomes, the endoplasmic reticulum, mitochondria, and chloroplasts. If these signal peptides were conjugated to the nanosensor surface, then there is potential for directing the nanosensor to an organelle of choice. Other surface-bound chemistries that may be useful in nanosensor surface chemistry are those that would enable endosomal escape. A route of investigation may be to accept the inevitability of endocytosis and concentrate on devising mechanisms for breaking the delivered payload (nanosensors) out of the endosome once inside the cell. Utilizing the fusion of nanosensor-bound lipids with the endosomal membrane, which can then be cleaved by cellular enzymes, may provide a means for endosomal escape and cytoplasmic localization. 2.3.3 Multilevel, Multifunctional Surface Chemistry The nanosensor surface is a valuable locale that should be used to provide biophysical properties for enhanced delivery methodology. A theoretical nanosensor surface could carry out multiple functions in a specific sequence that chaperones the nanosensor through the various membranes and on to a specific intracellular microdomain. Consider a CPP-mediated delivery method that triggers endocytosis. In the outer level of the molecular coat the CPP is presented to the cell membrane; directly beneath this is a layer of PEG that protects other functional molecules closer to the nanosensor surface. The CPP triggers translocation across the outer cell membrane via receptor-mediated endocytosis that ultimately leads to the nanosensor being located in a lysosome. Importantly, the CPP and PEG layer are attached to the nanosensor by a linker that is cleaved by the lysosomal environment—be that a lower pH or lysosome-specific enzymes. Removal of the CPP and PEG layers then reveals a lipid layer on the nanosensor that fuses with, and breaks through, the lysosomal lipid membrane, gaining the nanosensor access to the cytoplasmic environment. Here the nanosensor may be available to take measurement or perhaps, when required, a photocleavable linker is illuminated that removes the lipids and reveals signal peptides that direct the nanosensor to a cellular organelle. Alternatively, the linker could simply be susceptible to cytoplasmic enzymes such as esterases.

2.4

CELL BEHAVIOR DIAGNOSTICS

What is vital in all delivery techniques is the minimal level of cellular disruption. It is worthy of reiteration here that developing a delivery method that is successful, in terms of the numbers of internalized sensors, is of little value if cellular perturbation (physical or chemical) is such that normal cellular processes are disrupted. Methods for assessing cellular normality/viability must

CELL BEHAVIOR DIAGNOSTICS

27

therefore be carried out to appraise each delivery technique. This feature of nanosensor delivery can be referred to as “postdelivery cell behavior diagnostics,” which is the investigation that confirms and ensures a delivery method is as least disruptive as possible. A cell is considered viable if it has the ability to grow and develop [61, 62]; we extend our definition to include any morphogenic or phenotypic change that has occurred because of nanosensor delivery. For in vitro cultures of cells, assays to test viability can be based on the physical characteristics of the cells such as cell membrane integrity, cytoplasmic streaming in plant cells, deformability as monitored with optical tweezers [63], or granulation measured with flow cytometry. Several dyes are used for the assessment of cell membrane integrity, including trypan blue, methylene blue, and neutral red. These dyes are able to cross the disrupted membranes of dead cells and stain the intracellular contents. This is commonly referred to as a live : dead assay. The dead cells can be accounted for microscopically or with spectrometric estimation. Additionally, assays monitoring the metabolic activity, for example, the hydrolysis of fluorogenic compounds, are commonly used as indicators of cellular health. These assays often rely on the function of cellular enzymes (e.g., reductases or esterases) capable of altering the structure of an applied substrate to give an observable result. One such example is the administration of fluorescein diacetate, which is able to transfer across intact membranes. Inside the cell esterases cleave off the diacetate group yielding fluorescein; therefore the intracellular accumulation of this fluorescent compound correlates directly to the health of the cell. Again, fluorescent cells can be monitored with fluorescent microscopy, spectroscopy, or flow cytometry. The kinetic response to a chemoattractant could be monitored as another demonstration of normal cell function. Other methods based on the assessment of structural, morphological, and phenotypic characteristics can be used as additional or more convenient features of natural cell behavior including adherence and cell proliferation. For example, dielectrophoresis could be used to investigate changes in cell permittivity, and patterned electrode pair structures, used to measure impedance, can provide inferential measurement of proliferation. A method we employed to assay the well-being of a cell after delivery was to assess the degree of cellular identity. By this we refer to the unique identity of a cell, in terms of it being able to carry out a particular function or the type of receptors on the cell surface. Other proteinomic or genomic assessments may be used to identify the expressed biomolecular landscape of the cell and compare this with control populations. Human embryonic mesenchymal stem cells (hMSCs) were loaded with nanosensors prior to assessment with flow cytometry. A benefit of flow cytometry is the ability to investigate far greater numbers of cells than microscopy-based techniques; here each plot of data represents 50,000 cells. Flow cytometry can provide information on cellular size and granulation, as well as cell surface markers when the cells are previously incubated with relevant CD markers. In Figure 2.4 the dual use of flow cytometry data and confocal microscopy images show hMSC, containing

28

DELIVERY OF NANOSENSORS TO MEASURE THE INTRACELLULAR ENVIRONMENT

104

104

104

A

B

C PE

103

PE

103

PE

103

102

102

102

101

24.0 73.4

0 1.77

100 0 10

101

nic

102

103

104

101

20.3 60.5

0.01 1.43

100 0 10

101

2 nic 10

103

104

101

94.4 0.50

4.01 0.17

100 0 10

101

2 nic 10

103

104

Figure 2.4. Use of cell surface markers to confirm cellular identity. (Top) Simultaneous detection of CD29 expression on hMSC containing Tat-functionalized FITCnanosensors. (A) Isotype control showing only FITC fluorescence coming from internalized nanosensors. (B) CD29 second antibody control showing only FITC fluorescence coming from internalized nanosensors. (C) hMSC containing Tat-functionalized FITC nanosensors and stained with CD29-PE. Both signals are detected, as evidenced by the shift along the y-axis of the cell population. (Bottom) hMSC cells loaded with rhodamine B nanosensors (red) and counterstained with mouse anti-human CD105:FITC antibody (green). (See color insert.)

internalized Tat nanosensors, that retain the expression of CD29 and CD105, respectively. The hMSCs were positive for cell surface markers CD29 and CD105, and negative for CD34 and CD45 (control experiments; data not shown), demonstrating a population of undifferentiated hMSCs of nonhemopoietic lineage as defined by Pittenger et al. [64]. The combination of flow cytometry and fluorescence microscopy can be a powerful tool for cell behavior diagnostics.

2.5

CHALLENGES AND FUTURE PERSPECTIVES

The ultimate goal of nanosensor research is to have the ability to measure the analyte of choice at a cellular location of choice in real time. Further advantages accrue if the measurement system can be multiplexed to measure multiple analytes simultaneously. There is currently a rich stream of investigation to develop new measurement technologies at the nanoscale level to provide

REFERENCES

29

devices capable of probing the intracellular environment, concisely summarized by Borisov and Klimant [6]. New nanostructures, including “smart” sensors, and theranostics can be envisaged with increasing complexity and functionality being built into the devices. However, without improved delivery methods and, importantly, better understanding of how nanoparticles enter the cell and cross cell membranes the potential of these devices for intracellular analysis will not be realized. Each of the existing delivery methods described in this chapter has disadvantages when looked at in isolation. Most notable are the distress and physical perturbation to the cell from the “mechanical” loading techniques; gene gun delivery, picoliter injection, and scrape loading. Patch clamp and SICM methods have not yet found wide usage probably due to the cell-by-cell nature of loading, similar to picoinjection. Electroporation and sonication are thought to cause considerable disruption of the cellular membranes, which presents difficulties if trying to make measurements of “normal” cells. Of the membranal techniques phagocytosis and pinocytosis are limited to specific cell types, but liposomal delivery is used routinely for delivery of oligonucleotides and has been shown to work for nanoparticles too. It remains, however, that specific conditions need to be developed for each cell line and general applicability is notably limited. The attachment of cell-penetrating peptides has been investigated and shows promise as a technique for enabling nanosensor internalization without disruption to the cell wall or cellular biochemical activity of the cell. With nanosensors particularly, and most other nanosized particles destined for intracellular translocation, the surface of the particle provides the potential to be utilized for delivery strategies capable of traversing complicated routes through the cellular environment. Some of the exciting possibilities, in terms of intracellular targeting and fabricating multifunctional nanodevices were outlined. A final thought concerns the quality of fluorescence imaging as demand leads to the requirement of nanoscale resolution. A limitation at present is that the fluorescent output from nanosensors has to be assumed to arise from an ensemble of nanosensors as the standard optical microscopy resolution limit of λ/2 does not allow individual nanosensors to be imaged. The rapidly expanding field of superresolution microscopy, including the techniques of STED, PALM, STORM, and RESOLFT (see Hell [65] for a comprehensive review) will present opportunities to overcome the resolution limit; this may prove to be as transformative to intracellular imaging and analysis as the development and application of confocal microscopy has been to cellular analysis in the past 20 years.

REFERENCES 1. Buck, S. M., et al. Optochemical nanosensor PEBBLEs: photonic explorers for bioanalysis with biologically localized embedding. Curr. Opin. Chem. Biol. 8(5): 540–546 (2004).

30

DELIVERY OF NANOSENSORS TO MEASURE THE INTRACELLULAR ENVIRONMENT

2. Clark, H. A., et al. Optical nanosensors for chemical analysis inside single living cells. 1. Fabrication, characterization, and methods for intracellular delivery of PEBBLE sensors. Anal. Chem. 71(21): 4831–4836 (1999). 3. Clark, H. A., et al. Optical nanosensors for chemical analysis inside single living cells. 2. Sensors for pH and calcium and the intracellular application of PEBBLE sensors. Anal. Chem. 71(21): 4837–4843 (1999). 4. Park, E. J., et al. Ratiometric optical PEBBLE nanosensors for real-time magnesium ion concentrations inside viable cells. Anal. Chem. 75(15): 3784–3791 (2003). 5. Xu, H., et al. A real-time ratiometric method for the determination of molecular oxygen inside living cells using sol-gel-based spherical optical nanosensors with applications to rat C6 glioma. Anal. Chem. 73(17): 4124–4133 (2001). 6. Borisov, S. M. and Klimant, I. Optical nanosensors—smart tools in bioanalytics. Analyst 133(10): 1302–1307 (2008). 7. Srivastava, A. and Krishnamoorthy, G. Time-resolved fluorescence microscopy could correct for probe binding while estimating intracellular pH. Anal. Biochem. 249(2): 140–146 (1997). 8. Clark, H. A., et al. Subcellular optochemical nanobiosensors: probes encapsulated by biologically localised embedding (PEBBLEs). Sensors and Actuators B—Chemical 51(1–3): 12–16 (1998). 9. Clark, H. A., et al. Optochemical nanosensors and subcellular applications in living cells. Mikrochim. Acta. 131(1–2): 121–128 (1999). 10. Buck, S. M., et al. Nanoscale probes encapsulated by biologically localized embedding (PEBBLEs) for ion sensing and imaging in live cells. Talanta 63(1): 41–59 (2004). 11. Sumner, J. P. and Kopelman, R. Alexa Fluor 488 as an iron sensing molecule and its application in PEBBLE nanosensors. Analyst 130(4): 528–533 (2005). 12. Tan, W. H., Shi, Z. Y., and Kopelman, R. Development of submicron chemical fiber optic sensors. Anal. Chem. 64(23): 2985–2990 (1992). 13. Vo-Dinh, T. and Kasili, P. Fiber-optic nanosensors for single-cell monitoring. Anal. Bioanal. Chem. 382(4): 918–925 (2005). 14. Vo-Dinh, T., et al. Antibody-based nanoprobe for measurement of a fluorescent analyte in a single cell. Nature Biotechnol. 18(7): 764–767 (2000). 15. Brasuel, M., et al. Production, characteristics and applications of fluorescent PEBBLE nanosensors: Potassium, oxygen, calcium and pH imaging inside live cells. Sensors and Materials 14(6): 309–338 (2002). 16. Tan, W. H., et al. Ultrasmall for cellular. Anal. Chem. 71(17): 606A–612A (1999). 17. Webster, A., Compton, S. J., and Aylott, J. W. Optical calcium sensors: development of a generic method for their introduction to the cell using conjugated cell penetrating peptides. Analyst 130(2): 163–170 (2005). 18. Sumner, J. P., et al. A fluorescent PEBBLE nanosensor for intracellular free zinc. Analyst 127(1): 11–16 (2002). 19. Koo, Y. E. L., et al. Photonic explorers based on multifunctional nanoplatforms for biosensing and photodynamic therapy. Appl. Optics 46(10): 1924–1930 (2007). 20. Peterson, J. I., et al. Fiber optic pH probe for physiological use. Anal. Chem. 52(6): 864–869 (1980).

REFERENCES

31

21. Webster, A., et al. The delivery of PEBBLE nanosensors to measure the intracellular environment. Biochem. Soc. Trans. 035(3): 538–543 (2007). 22. Brasuel, M., et al. Fluorescent nanosensors for intracellular chemical analysis: decyl methacrylate liquid polymer matrix and ion exchange-based potassium PEBBLE sensors with real-time application to viable rat C6 glioma cells. Anal. Chem. 73(10): 2221–2228 (2001). 23. Sun, H., et al. Synthesis and characterization of ratiometric, pH sensing nanoparticles with covalently attached fluorescent dyes. Chem. Materials 18(15): 3381–3384 (2006). 24. Ulijn, R. V. Enzyme-responsive materials: a new class of smart biomaterials. J. Materials Chem. 16(23): 2217–2225 (2006). 25. Fuller, J. E., et al. Intracellular delivery of core-shell fluorescent silica nanoparticles. Biomaterials 29(10): 1526–1532 (2008). 26. Hild, W. A., Breunig, M., and Goepferich, A. Quantum dots—nano-sized probes for the exploration of cellular and intracellular targeting. Eur. J. Pharm. Biopharm. 68(2): 153–168 (2008). 27. Zorov, D. B., et al. Examining intracellular organelle function using fluorescent probes—from animalcules to quantum dots. Circ. Res. 95(3): 239–252 (2004). 28. Delehanty, J. B., Mattoussi, H., and Medintz, I. L. Delivering quantum dots into cells: strategies, progress and remaining issues. Anal. Bioanal. Chem. 393(4): 1091– 1105 (2009). 29. Hartikka, J., et al. Electroporation-facilitated delivery of plasmid DNA in skeletal muscle: plasmid dependence of muscle damage and effect of poloxamer 188. Mol. Ther. 4(5): 407–415 (2001). 30. Wyber, J. A., Andrews, J., and Demanuele, A. The use of sonication for the efficient delivery of plasmid DNA into cells. Pharm. Res. 14(6): 750–756 (1997). 31. Azencott, H. R., Peter, G. F., and Prausnitz, M. R. Influence of the cell wall on intracellular delivery to algal cells by electroporation and sonication. Ultrasound Med. Biol. 33: 1805–1817 (2007). 32. McNeil, P. L., et al. A method for incorporating macromolecules into adherent cells. J. Cell Biol. 98(4): 1556–1564 (1984). 33. McKeon, J. and Khaledi, M. G. Quantitative nuclear and cytoplasmic localization of antisense oligonucleotides by capillary electrophoresis with laser-induced fluorescence detection. Electrophoresis 22(17): 3765–3770 (2001). 34. Hosseinkhani, H. DNA nanoparticles for gene delivery to cells and tissue. Int. J. Nanotechnol. 3(4): 416–461 (2006). 35. Buono, C., et al. Fluorescent pegylated nanoparticles demonstrate fluid-phase pinocytosis by macrophages in mouse atherosclerotic lesions. J. Clin. Invest. 119(5): 1371–1381 (2009). 36. Chu, G. L., et al. The role of intracellular pH and its variance in low pH sensitization of killing by hyperthermia. Radiat. Res. 122(3): 288–293 (1990). 37. Kneen, M., et al. Green fluorescent protein as a noninvasive intracellular pH indicator. Biophys. J. 74(3): 1591–1599 (1998). 38. Foged, C. and Nielsen, H. M. Cell-penetrating peptides for drug delivery across membrane barriers. Expert Opin. Drug Deliv. 5(1): 105–117 (2008).

32

DELIVERY OF NANOSENSORS TO MEASURE THE INTRACELLULAR ENVIRONMENT

39. Vives, E., Schmidt, J., and Pelegrin, A. Cell-penetrating and cell-targeting peptides in drug delivery. Biochim. Biophys. Acta Rev. Cancer 1786(2): 126–138 (2008). 40. Futaki, S., et al. Arginine-rich peptides—an abundant source of membranepermeable peptides having potential as carriers for intracellular protein delivery. J. Biol. Chem. 276(8): 5836–5840 (2001). 41. Gariepy, J. and Kawamura, K. Vectorial delivery of macromolecules into cells using peptide-based vehicles. Trends Biotechnol. 19(1): 21–28 (2001). 42. Green, M. and Loewenstein, P. M. Autonomous functional domains of chemically synthesized human immunodeficiency virus tat trans-activator protein. Cell 55(6): 1179–1188 (1988). 43. Elliott, G. and Ohare, P. Intercellular trafficking and protein delivery by a herpesvirus structural protein. Cell 88(2): 223–233 (1997). 44. Derossi, D., Chassaing, G., and Prochiantz, A. Trojan peptides: the penetratin system for intracellular delivery. Trends Cell Biol. 8(2): 84–87 (1998). 45. Richard, J. P., et al. Cell-penetrating peptides—a reevaluation of the mechanism of cellular uptake. J. Biol. Chem. 278(1): 585–590 (2003). 46. Derossi, D., et al. The 3rd helix of the Antennapedia homeodomain translocates through biological-membranes. J. Biol. Chem. 269(14): 10444–10450 (1994). 47. Derossi, D., et al. Cell internalization of the third helix of the Antennapedia homeodomain is receptor-independent. J. Biol. Chem. 271(30): 18188–18193 (1996). 48. Hansen, M., Kilk, K., and Langel, U. Predicting cell-penetrating peptides. Adv. Drug Deliv. Rev. 60(4–5): 572–579 (2008). 49. Silhol, M., et al. Different mechanisms for cellular internalization of the HIV-1 Tat-derived cell penetrating peptide and recombinant proteins fused to Tat. Eur. J. Biochem. 269(2): 494–501 (2002). 50. Vives, E. Cellular uptake of the Tat peptide: an endocytosis mechanism following ionic interactions. J. Mol. Recognition 16(5): 265–271 (2003). 51. Futaki, S., Goto, S., and Sugiura, Y. Membrane permeability commonly shared among arginine-rich peptides. J. Mol. Recognition 16(5): 260–264 (2003). 52. Lundberg, P. and Langel, U. A brief introduction to cell-penetrating peptides. J. Mol. Recognition 16(5): 227–233 (2003). 53. Coupland, P. G., et al. Internalisation of polymeric nanosensors in mesenchymal stem cells: analysis by flow cytometry and confocal microscopy. J. Control. Release 130(2): 115–120 (2008). 54. Mae, M., et al. Design of a tumor homing cell-penetrating peptide for drug delivery. Int. J. Pept. Res. Ther. 15(1): 11–15 (2009). 55. Chen, B., et al. Transmembrane delivery of the cell-penetrating peptide conjugated semiconductor quantum dots. Langmuir 24(20): 11866–11871 (2008). 56. Torchilin, V. P. Cell penetrating peptide-modified pharmaceutical nanocarriers for intracellular drug and gene delivery. Biopolymers 90(5): 604–610 (2008). 57. Palm-Apergi, C. and Hallrink, M. A new rapid cell-penetrating peptide based strategy to produce bacterial ghosts for plasmid delivery. J. Control. Release 132(1): 49–54 (2008). 58. Qualmann, B., Kessels, M. M., and Kelly, R. B. Molecular links between endocytosis and the actin cytoskeleton. J. Cell Biol. 150(5): F111–F116 (2000).

REFERENCES

33

59. Morgan, T. T., et al. Encapsulation of organic molecules in calcium phosphate nanocomposite particles for intracellular imaging and drug delivery. Nano Lett. 8(12): 4108–4115 (2008). 60. Emanuelsson, O., et al. Locating proteins in the cell using TargetP, SignalP and related tools. Nature Protocols 2(4): 953–971 (2007). 61. Pegg, D. E. Viability assays for preserved cells, tissues, and organs. Cryobiology 26(3): 212–231 (1989). 62. Palta, J. P., Levitt, J., and Stadelmann, E. J. Plant viability assay. Cryobiology 15(2): 249–255 (1978). 63. Lincoln, B., et al. High-throughput rheological measurements with an optical stretcher. In: Cell Mechanics. Elsevier Academic Press, San Diego, CA, 2007, pp. 397–423. 64. Pittenger, M. F., et al. Multilineage potential of adult human mesenchymal stem cells. Science 284(5411): 143–147 (1999). 65. Hell, S. W. Microscopy and its focal switch. Nature Methods 6(1): 24–32 (2009).

CHAPTER 3

Cytoplasmic Diffusion of Dendrimers and Dendriplexes ALEXANDER T. FLORENCE School of Pharmacy, University of London, London, United Kingdom

PAKATIP RUENRAROENGSAK Lung Cell Biology, Respiratory Medicine, National Heart and Lung Institute, Imperial College, London, United Kingdom

3.1

INTRODUCTION

Diffusion of nanoparticulate carriers within the cytoplasm is one of the controlling factors in the delivery of active therapeutic agents not only to the cell nucleus but also to a large extent to adjacent cells. Delivery to the nucleus is the goal of gene therapy but the overall flux of particles in tissues is determined cell by cell. The cytoplasm may be an aqueous environment but it is a complex one, crowded molecularly and far from isotropic. Hence the diffusion process of particles that have gained access to the individual cell in a free state, or escape from organelles in which they have been snared, and then diffuse, is complex. The flux of particles is determined by several factors: (1) by the properties of the medium in which the particles move; (2) by the properties of the particle, such as its radius, surface properties, shape, and flexibility; (3) by the concentration gradients that drive diffusion; and (4) by a range of other influences such as physical obstructions, which increase the effective path length that particles must traverse, and binding of nanoparticles to components within the cytoplasm. Convection currents may enhance or retard diffusion. The problem in both modeling and exact mathematical description is that the cytoplasm is a dynamic medium.

Organelle-Specific Pharmaceutical Nanotechnology, Edited by Volkmar Weissig and Gerard G. M. D’Souza Copyright © 2010 John Wiley & Sons, Inc.

35

36

CYTOPLASMIC DIFFUSION OF DENDRIMERS AND DENDRIPLEXES

In this chapter we describe some of the work conducted in our laboratories on the diffusion of dendrimers and dendriplexes, and discuss it in the light of findings in the literature. The bulk of the work discussed here centers on a novel synthetic 6.5-nm dendrimer based on branched lysine residues, which we found serendipitously to be autofluorescent. This feature of the dendrimer allowed us to investigate nanoparticle movement within the cytoplasm using confocal microscopy and other techniques. We also studied the diffusion of the dendrimer in aqueous media (water and glycerol solutions) and in aqueous gels for comparison with the much more complex cytoplasmic transport. One complication, which may not be unique to our dendrimer, is the finding that, at least in vitro, it has a biphasic effect on actin polymerization. Dendrimers are three-dimensional hyperbranched macromolecules [1]. They have been employed mainly in experimental applications encompassing diagnostic tools and as drug carriers; some have intrinsic biological activity. Despite this, elucidation of the uptake mechanisms and the cellular pathways of dendrimer carrier systems and dendrimers per se has been slow. Before diffusion can occur in the cytoplasm, the particles have to be taken up; the mode of uptake may influence subsequent events, depending on whether the particle emerges in the cytoplasm as a free entity or encapsulated within vesicles. The uptake of nanoparticles and dendrimers has perforce been studied by attachment of fluorescent probes either by conjugation [2] or by physical interactions [3]. Scission or loss of the fluorescent moiety will result in difficulty with interpretation of resulting data [4]. The use of an intrinsic fluorescent particle can overcome these problems as there are no chemical changes required and surface properties are the original. In our work, for example, with dendrimer–DNA complexes (dendriplexes), their diameter can increase by up to 20–25 nm on fluorescently labeling the DNA; this is accompanied by a twofold increase in the zeta potential of the complex. Artifacts are also created by fixing cells; thus experiments in live cells are preferable for estimating cytoplasmic transport [5, 6]. The synthetic self-fluorescent sixth generation amino-terminated polyamide polylysine dendrimer (Figure 3.1A), namely [(Gly)(Lys63)(NH2)64] with a molecular weight of 8149 Da, has been reported previously by us [7]. The use of this compound as a nanoscopic probe is discussed in this chapter, using confocal laser scanning microscopy (CLSM). The results relate both to the native dendrimer and to a dendrimer–DNA complex and are discussed in terms of diffusion, obstruction, and binding effects. We can only speculate on its effects on actin polymerization, which we found in vitro. From a naïve physicochemical point of view, if such effects occur in vivo, depolymerization would ease the passage of the dendrimer and polymerization would complicate it. However, actin polymerization actually provides the propulsive force for intracellular transport of endosomes and some microorganisms, so this is a complex subject [8].

DIFFUSION WITHIN CELLS

A

Lys Lys Lys Lys Lys Lys Lys Lys Lys Lys

Lys Lys

Lys

Lys Lys

Lys Lys

Lys

Lys Lys

Lys

Lys

Lys

Lys

Lys

Lys Lys

Lys Lys

Lys Lys

Lys

Lys

Gly

Lys

NH2

Lys

Lys Lys

Lys

Lys Lys Lys Lys

Lys

Core Lys Lys Branch Surface

Lys

Lys Lys

Lys Lys

Lys Lys

Lys Lys

C

B

Lys

Lys

Lys

Lys Lys

Lys Lys Lys

Lys

Control D

10 µm

10 µm

G

t=1h

10 µm

10 µm

t=2h

10 µm

t=1h

10 µm

I

H

t=2h

t=15min F

E

t=30min

37

10 µm

t=4h

10 µm

Figure 3.1. The structure (A) of the sixth generation polylysine dendrimer used in this study and a time-dependent study of the interaction between the dendrimer (green fluorescence) and Caco-2 cells (B–I): the cells were incubated with complete media containing 0.1% (w/v) of dendrimers for 15 min (C), 30 min (D), 1 h (E, F), 2 h (G, H), and 4 h (I). The cells were observed under confocal microscopy. In the control (A), the cells were exposed to the complete medium without dendrimer. The dendrimer is found to attach to the cell membrane periphery (red arrows) as a result of electrostatic attraction (C, D). At 1 h the dendrimer distributes toward the cell nuclei in all regions of the cells (E, F). At longer incubation times (G–I), the dendrimer appears to be concentrated on the surface of the cells. This technique cannot prove internalization. (See color insert.)

3.2

DIFFUSION WITHIN CELLS

Diffusion of spherical nanoparticles in a liquid medium obeys the Stokes– Einstein equation relating diffusion coefficient (D) to the radius of the particle (r) and the viscosity (η) of the medium in which the particle moves as

38

CYTOPLASMIC DIFFUSION OF DENDRIMERS AND DENDRIPLEXES

D = kT 6πηr One can demonstrate in binary aqueous mixtures (e.g., glycerol–water) that the diffusion coefficient reduces with increase in bulk viscosity, as these are simple fluids composed of small molecules. In macromolecular systems, the viscosity parameter in the Stokes–Einstein equation is not the bulk viscosity but in fact the intrinsic or “microscopic” viscosity. Movement of particles through some gels is rapid because the “pore” sizes are large and the medium in which the particles move is akin to bulk water. We have calculated the intrinsic viscosity of the interior of HPMC gels to be very close to that of water, using a variety of nanoparticle probes. In nonionic viscous vehicles such as glycerol solutions and HPMC gels, there is no or little influence of particle charge [9]. As other chapters in this book discuss, the cell cytoplasm is concentrated and crowded, occupied as it is by a variety of macromolecules between 5% and 40% of the total cell volume or 400 mg/L [10]. The cellular architecture provides a complex non-Newtonian fluid comprising an aqueous phase filling the spaces between an entangled mesh of filamentous cytoskeleton and other macromolecular structures, resembling a gel-like structure [11–14]. This phenomenon is well known as “macromolecular crowding”, which plays a decisive role on several levels of cellular organization, particularly in the diffusion and binding of matter inside cells [13]. Translocation of nanosystems occurs in this three-dimensional forest. The Stokes–Einstein equation may not be entirely applicable to diffusion phenomena within the constantly changing interior of cells. Diffusion of matter in cells relies on fluid-phase viscosity (F1), solute binding to macromolecules (F2), and collisions between solutes and macromolecules (F3) [15, 16]. Diffusion theory of solutes in the cell cytoplasm has been described by Kao et al. [16], who assumed that the cytoplasm was composed of an aqueous fluid-phase compartment bathing a matrix of mobile and static macromolecules/particles much larger than water molecules and small solutes. They identify three main factors involved in the reduction in diffusion coefficient of a small solute in cytoplasm (Dcyto) relative to that in water (D0): Dcyto = F1 ( η) × F2 ( Du , {Db,i , fb,i }) × F3 ({ni , Vi }) D0

(3.1)

The function F1(η) represents the deceleration of net solute translational diffusion because of an increase in true fluid-phase cytoplasmic viscosity, an increase due to solute-induced perturbation in solvent structure, which can be written F1 ( η) =

η0 ηcyto

(3.2)

DIFFUSION WITHIN CELLS

39

where η0 represents the viscosity of water and ηcyto is the true fluid-phase microviscosity of the cytoplasm. The function F2(Du,{Db,i, fb,i}) indicates the total reduction of solute translational diffusion due to the transient binding of solute molecules to cytoplasmic structures as addressed in Equation 3.3 below [16]. F2 represents the ratio of the weighted diffusion coefficient of bound and unbound solute to the diffusion coefficient of the unbound solute. In Equation 3.3 Du and Db,i are, respectively, the diffusion coefficients of unbound and ith bound solute, and fb,i is the fraction of net solute bound to component i. F2 ( Du , {Db,i , fb,i }) = fu + ∑ ( Db,i Du ) fb,i

(3.3)

i

where fu + ∑ fb,i = 1

(3.4)

i

The function F3({ni,Vi}) illustrates the deceleration of the total reduction of solute translational diffusion owing to collisional interaction with cytoplasmic structures assuming there are ni structures of type i, each of volume Vi. The “volume exclusion” by mobile obstacles has been used as a model (a stretchedexponential [SE] model) to explain this function [17–19]. This SE model is applicable to the diffusion of Brownian particles (or solutes), which are relatively large compared to the solvent molecules. v F3,SE ( ni , Vi ) = exp ⎡⎣ −α ( ni , Vi ) ⎤⎦

(3.5)

Here niVi represents the volume fraction occupied by the occluding molecule and the prefactor α and ν represent scaling parameters obtained by fitting Equation 3.5 to data, but can be calculated independently from theory [16]. Viscosities of the cytoplasm of various cell types have been calculated. Values ranged from 2 to 20 cP [16, 20–22]. By using time-resolved fluorescence anisotropy, Verkman and colleagues found that the cytoplasmic solvent measured in different types of cell lines may be significantly different from that of bulk water [23]. This has also been demonstrated [24] using a different approach in two mammalian tissue culture cell lines. Luby-Phelps et al. [25] characterized the nature of the diffusion barrier by comparing the diffusion of 3.2–25.8 nm Ficoll probe molecules. The results showed that relative to water the diffusion rate in the cytoplasm, which is already low with the smallest (3.2-nm) probe, reduces with an increase of the probe size. To describe this relationship, diffusion of various sizes of FITCFicoll was measured in model solutions [19]. The relative diffusion coefficient of FITC-Ficoll in concentrated solutions of Ficoll is low, but it is independent of the size of the FITC-Ficoll probe. In a tangle of F-actin filaments, on the

40

CYTOPLASMIC DIFFUSION OF DENDRIMERS AND DENDRIPLEXES

other hand, relative diffusion of the 3.2-nm probe is not impeded, but its diffusion decreases with the larger FITC-Ficoll probes. Taken together, it was noticed that the effects of Ficoll (10%) and F-actin filaments (5 mg/mL) resemble the pattern seen in the cytoplasm. The solution conditions calculated to most closely mimic the size dependency of FITC-Ficoll diffusion in the cells are 120 mg/mL of bovine serum albumin dissolved in a fluid matrix containing 37 mg/mL F-actin fibers. Based on these results, a model of cytoplasm as a densely entangled filament network interpenetrated by a fluid phase crowded with globular macromolecules was developed. This is in good agreement with cell structures proposed by Medalia et al. [14]. Uncoated GFP-M6tail-associate vesicles within cells [26], found during the process of clathrin-dependent endocytosis, are trapped in an actin-rich region of the cytoplasm of ARPE-19 cells. These vesicles exhibited Brownian-like motion. They could exit from the actin-rich region by a slow diffusion-based mechanism. The diffusion coefficient of the vesicles (n = 250) was found to be 1.42 × 10−12 cm2/s (SD = 1.24). Small molecules traverse cell membranes by passive transport, whereas macromolecules cross membranes by endocytosis, the process of membranebound vesicles originated from the invagination and pinching off of the membrane [27]. Phagocytosis occurs in mammalian cells, while pinocytosis may occur in all cell types. It has been reported that some dendrimers may create holes in cell membranes ranging between 15 and 40 nm, allowing passive diffusion of dendriplexes [28–32]. In addition, polylysine polymer was found to initiate holes on the cell membrane with a diameter ranging from 1 to 10 nm [29]. This suggests that, as we believe, the polylysine dendrimers used in our work are taken up intact and are not sequestered. The use of cell-penetrating peptides to increase the bioavailability of drug/gene/siRNA delivery has been seen to overcome both extracellular and intracellular barriers [33]. Peptide carried MPG and Pep was found to form a stable nanoparticle complex with the active macromolecules. The particles were found to enter the cells independently of endosomal pathways. Both lipoplexes and dendriplexes possessing positive charges first adhere to the cell membrane as a result of electrostatic interactions. Endocytosis has been invoked depending on types of carrier systems and cell lines [34, 35]. The nature of the carrier surface, in fact, regulates specific or nonspecific binding of the complexes onto the cell membrane and triggers different endocytic pathways such as clathrin- and caveolin-dependent endocytosis, and clathrinand caveolin-independent endocytosis. Clathrin- and caveolin-dependent endocytosis has been proposed as the major pathway of internalization of lipoplexes and dendriplexes [36–38]. Interestingly, a difference in uptake pathway was also found between surface-engineered dendrimers with amine and carboxyl functional groups. The carboxyl-modified dendrimers were taken up by caveolin-dependent endocytosis, whereas the amine-modified dendrimers were taken up by the clathrin- and caveolin-independent pathway [2].

DIFFUSION AND UPTAKE OF DENDRIMER AND DENDRIPLEXES

41

3.3 METHODS USED TO ASSESS THE UPTAKE AND CELLULAR DIFFUSION OF NANOPARTICLES The self-fluorescent sixth generation amino-terminated polyamide polylysine dendrimer (Figure 3.1A), with a molecular weight of 8149 Da, was used in this study and the details of the synthesis method can be found elsewhere [7]. For the dendriplexes, the same dendrimer was complexed with plasmid DNA (pDNA), pDsRed2-N1 (4.689 kb, Clonetech, USA), at 10 : 1 molar charge ratio (+/−). As the fluorescence of the dendrimer decreases after the complexation, the pDNA was fluorescently labeled using Label IT® Tracker™ Fluorescein kit (Mirus Bio Coporation, USA) before forming the complex. Dynamic uptake of the dendrimer and dendriplexes in all types of cells was followed every 5 min before and after adding the dendrimer and dendriplexes using confocal microscopy and data analyzed using custom procedures written in IgorPro software (Wavemetric Inc). The z-stack image at time t = 5 min before the dendrimer or the dendriplexes was combined and used as the control image. The fluorescent intensity detected in the green channel in this stack represents the background fluorescent intensity. The regions of interest (ROIs) were drawn and saved as the reference file. The fluorescent intensity values of the dendrimer/dendriplexes inside each z-stack image were averaged and standardized by subtraction of the background fluorescent intensity. The diffusion coefficients (D) and relative diffusion coefficients (D/D0) of the dendrimer/dendriplexes were calculated assuming that the cytoplasm is homogeneous. The focus was on the transport of the dendrimer and dendriplexes from the plasma membrane to the nucleus [39]. The lag time (tL) for the dendrimer/dendriplexes to develop a uniform concentration gradient within cytoplasm allows calculation of a diffusion coefficient where the thickness (h) of the diffusion layer is known, as D = h2/6tL. The thickness is the mean distance between the plasma membrane and nuclear membrane, termed the “cytoplasmic radius.” The cytoplasmic radius was measured using TEM and confocal microscopy. The value of tL was obtained by measuring the fluorescence intensity of the dendrimer over time from uniformly sized regions of interest placed equidistantly across the cell using custom procedures written in IgorPro software. The diffusion coefficient (D) of the dendrimer in the cell cytoplasm was then calculated, using the experimental tL value. The set fluorescent intensity of ROIs in each time point (F) was then plotted against time (minutes).

3.4

DIFFUSION AND UPTAKE OF DENDRIMER AND DENDRIPLEXES

This topic is discussed here as uptake may be a rate-limiting step in subsequent events in the cytoplasm. There is of course a concentration dependency of both flux and diffusion coefficient [40]; hence rates and extents of uptake are important.

42

CYTOPLASMIC DIFFUSION OF DENDRIMERS AND DENDRIPLEXES

3.4.1 Adhesion of Dendrimer to Different Cell Lines The adhesion of the dendrimer to Caco-2 cells is time dependent as illustrated in Figure 3.1B–I. At 15 min, the dendrimer (seen by the green fluorescence) concentrates at the periphery of cell membrane (Figure 3.1C). Within 30 min, the dendrimer is found to accumulate in the cytoplasm (Figure 3.1D). The relatively rapid uptake into the Caco-2 cells and specifically the nucleus can be attributed first to the electrostatic attraction between the amino surface groups of the dendrimer and the negative charge of membrane proteins, allowing attachment prior to endocytosis. Within an hour, the dendrimer covers all parts of the cell (Figure 3.1E–F). Distribution of the dendrimer toward the cell nucleus is observed in dividing (D) and nondividing cells (E). The dendrimer was found to cover the cells and apparently concentrate in the cytoplasm region (Figure 3.1G–H). The cells after exposure to the dendrimer for 4 h are seen in Figure 3.1I; the higher number of areas of dendrimer coverage are found on the cells. Different results were found in SKMES-1 cells. The dendrimer appeared to randomly adhere to the cell membrane. At longer incubation periods a more typical trend was found: the dendrimer was found to cover the cell surface. The reasons for this variation remain unclear but may result from differences in uptake mechanisms. Xia et al. [41] found such diferences in uptake mechanisms for 60-nm diameter amine-modified polystyrene nanoparticles when studying macrophage (RAW 264.7) and endothelial cells (BEAS2B). The uptake of the particles in the latter was found to be caveolin-dependent endocytosis whereas that in the former was independent of caveolin. We have demonstrated differences in the rate of binding of our dendrimer to three different cell types, and also the reduction of cellular adhesion when the medium bathing the cells is flowing (results not shown here). The extent of reduction is dependent on the force of adhesion (unpublished data).

3.4.2 Diffusion and Uptake of Dendrimer and Dendriplexes in Living Cells At low concentrations of the dendrimer, the colocalization of the dendrimer may not be very obvious, but increased levels of the fluorescent signal of the dendrimer are found inside the cell nucleus after background subtraction. Interestingly, the same pattern was noticed in the nucleus after the dendrimer was added into the system (Figure 3.2A–C). Hoechst33342 was used for locating the nuclear compartment. The dynamic uptake of the dendrimer into the nucleus was examined by overlaying images of the dendrimers (in green channel (Ch2)) over images of the nucleus (in blue channel (Ch1)) as can be seen in Figure 3.2. The fluorescent signal in the green channel, representing the dendrimers in each ROI, was normalized by background fluorescence subtraction (white boxes no. 14

DIFFUSION AND UPTAKE OF DENDRIMER AND DENDRIPLEXES

43

Fluorescent Intensity

60 50 40 † Boxes in both channels are regions of interest (ROIs)

30

ROI 5 ROI 6 ROI 7 ROI 8 ROI 9 ROI 10

20 10 0 0

20

40 60 Time (min)

80

100

Figure 3.2. An overlay of the compressed z-stack images of the dendrimer uptake process in living Caco-2 cells. The cells were incubated with medium containing 0.1% dendrimer (represented in green) and images collected every 5 min before (A) and after addition of dendrimer (B–F). Images presented are at time t = 5 (B), 15 (C), 30 (D), 40 (E), and 90 (F) min of incubation. Data were analyzed over the time and plotted in (G); the inserts show the positions of regions of interest (ROIs, white boxes) for data analysis. The mean fluorescence signal in each ROI was measured from the projected z-stack images at different time points with background subtraction. The dendrimer was found in the cytoplasmic compartment 5–35 min after exposure to the dendrimer (ROI 9 and 10). The concentration of the dendrimer declined after 30 min, whereas the relative fluorescent intensity of the dendrimer inside the nucleus was detected (ROI 5–8). The cell nuclei were stained with Hoechst33342 represented in blue. The scale bar is 20 µm. (See color insert.)

44

CYTOPLASMIC DIFFUSION OF DENDRIMERS AND DENDRIPLEXES

in the insert pictures of Figure 3.2G). The relative fluorescent signal of the dendrimer was then plotted (F) against time (t, min) as in Figure 3.2G. The results (n = 5) demonstrated that the dendrimer reached the nucleus of Caco-2 cells within 35–45 min (tL) of incubation (Figure 3.2G), compared to 25–30 min in SKMES-1 cells (data not shown). This might be due to the differences of the intracellular architecture within each cell line, but not cell size as the cell radius is taken into account in the calculation of D. Average cell radii were 12.09 (± 6.02 (SD)) and 8.01 (± 3.21 (SD)) µm for the Caco-2 cells and SKMES cells, respectively. Dcyto values of the dendrimer in Caco-2 cells and SKMES-1 cells were found to be 9.82 (± 0.98) × 10−11 cm2/s and 5.99 (± 0.16) × 10−11 cm2/s, respectively. Cellular components such as secretory granules have diffusion coefficients of 1.9 × 10−11 cm2/s, lower than that of the dendrimers [42] as might be expected from their larger dimensions. The diffusion coefficients relative to those in water (Dcyto/D0) are, respectively, 1.24 (± 0.12) × 10−4 and 0.76 (± 0.05) × 10−4 (D0 = 79.54 × 10−8cm2/s). These values imply that macromolecular crowding and obstacles in the cytoplasm provide a formidable barrier to dendrimer transport. They also emphasize that differences between cell lines can be significant. In work on transfection with dendriplexes, we have found that there can be a 1000-fold difference in efficiencies when using the same DNA construct with a range of cell lines [43]. The cytoplasm hinders the mobility of the dendrimer by some 1000-fold compared to that in water. This may be explained in loose terms by exclusion and obstruction effects, the former due to the loss of free water by hydration of macromolecules and the latter due to the tortuous pathway that the dendrimer must travel in the cytoplasm, avoiding organelles and other structural features, such as actin fibers. The diffusion of globular proteins in muscle cells approximates to zero when the hydrodynamic radius of the proteins increased from 1 nm (Dcyto = 16.3 ± 3 µm2/s) to 7 nm (Dcyto = 4.0 (± 0.7) × 10−3 µm2/s) [44]. This slower-than-predicted diffusion of the proteins or nanoparticles may be explained by an elasticity of F-actin/α-actinin networks in the cell cytoplasm [45]. 3.4.3 Dendrimer–Actin Interactions We have evidence of an interaction due to the electrostatic attraction between the dendrimer and actin cytoskeleton in vitro, and we have determined the diffusion coefficient of the dendrimer in the presence of actin gel using the fluorescent recovery after photobleaching (FRAP) technique. The ratio Dactin/D0 of the dendrimer was found to be 0.26, at a concentration of actin gel of 1 mg/mL, much lower than the concentration inside cells (∼10 mg/mL). Although the medium in which particles diffuse may approximate to that of pure water [16, 46], the auxiliary reduction of the translation diffusion of the dendrimers in cytoplasm can be attributed to microscopic barriers, enhanced tortuosity, and binding. Further studies on this are needed.

DIFFUSION AND UPTAKE OF DENDRIMER AND DENDRIPLEXES

A

B

C

D

45

Relative fluorescent intensity (F)

E

90 80 70 60 50 40 30 20 10 0 0

20

40

60

Time (min)

80

100

ROI 2 ROI 3 ROI 4 ROI 5 ROI 6 ROI 7 ROI 8 ROI 9 ROI 10 ROI 11 ROI 12 ROI 13

White boxes are regions of interest (ROIs)

Figure 3.3. An overlay of the compressed z-stack images of the dendriplex uptake process in living Caco-2 cells. The cells were incubated with complete medium containing dendriplexes (represented in green) and images were collected every 5 min before (A) and after adding the dendrimer (B–D). Images presented here are at time t = 5 (B), 30 (C), and 60 (D) min of incubation period. Data were analyzed and plotted in (E) and the insert pictures show the positions of regions of interest (ROIs, white boxes) for data analysis in both channels. The mean fluorescent signal in each ROI was measured from projected z-stack images taken at different time points, with background subtraction. The dendriplexes were found to be taken up into the cytoplasmic compartment (ROI 2–3, 7, 9, 12–13) but there is no uptake of the dendriplexes into the nuclear compartment within 2 h of incubation (ROI 4, 8, 10). (See color insert.)

46

CYTOPLASMIC DIFFUSION OF DENDRIMERS AND DENDRIPLEXES

3.4.4 Dendriplex Diffusion The diffusion coefficients of the dendriplexes were expected to be lower than that of the parent dendrimer as they consist of many dendrimer molecules complexed with the DNA. In the dynamic uptake study (Figure 3.3) they were found after 2 h only in the cytoplasmic compartment throughout the stack images (A–D). Although dendriplexes were first found attached to the cell membrane after 5 min incubation, they remained in the cytosol until the end of the experiment (A–D). This was confirmed by the quantification of the fluorescent signal of the dendriplexes as plotted in Figure 3.3E (ROI 2–3, 7, 9, 12–13). There is no fluorescence in the nuclear compartment after 2 h. Changes in cell morphology were found in the cells after 2 h under the conditions of the experiment. Further work was carried out using living cells incubated with the dendrimer at longer time points; cells were removed to view at the end of each selected time point. The dendriplexes were found to be taken up into the nuclear compartment after 2.5 h of incubation and uptake was seen to be widespread in nearly every cell after 3 h. As this lag time is long, the calculated diffusion coefficient of the dendriplexes is lower than that of the parent dendrimer, as shown in Table 3.1. The lag time (tL) here was found to be between 2.5 and 3 h (n = 5), the diffusion coefficient of the dendriplexes in cytoplasm (Dcyto) was calculated to be 2.36 (± 0.34 (SD)) × 10−11 cm2/s (2.36 × 10−3 µm2/s) and the diffusion coefficients (Dcyto) relative to that in water (4.53 × 10−8 cm2/s (D0)) was 5.21 (± 0.75) × 10−4. In Caco-2 cells the cytoplasmic diffusion coefficient, Dcyto, of the dendriplexes relative to the dendrimer was found to be 0.24. The cytoplasm hinders the mobility of the dendriplexes up to 2000 times that in water. This might be explained by the interaction between the cationic dendriplexes and the cell organelles, which impede the mobility of the dendriplexes apart from the effect of particle size, sieving effects, and obstruction of the crowded condition of the cells. The larger unit should be more affected by the obstruction effect. Whether or not binding of the dendriplex, say, to actin filaments occurs is not known; it may be that dissociation of the complex occurs and the diffusion coefficients are mixed values. TABLE 3.1 Diffusion Coefficients and Relative Diffusion Coefficients of the Parent Dendrimers and the Dendriplexes in Cell Cytoplasm Compared to Other Media Diffusion Coefficients in Various Media (cm2/s) SKMES-1 Cells

Dactin D0

Dcyto D0

Ddendriplexes Ddendrimer

Dendrimer (6.5 nm)

79.54 × 10−8 23.75 × 10−8 9.82 × 10−11 5.99 × 10−11

0.26

—

Dendriplexes (107.3 nm)

4.53 × 10−8

1.24 × 10−4 (Caco-2) 0.76 × 10−4 (SKMES-1) 5.21 × 10−4

Nanoparticles

Water

Actin Gel

—

Caco-2 Cells

Relative Diffusion Coefficients (cm2/s)

2.36 × 10−11

—

—

0.24

MECHANISMS OF PARTICLE UPTAKE INTO THE CELL NUCLEUS

47

The differences are caused by two main factors: the physicochemical properties (size, charge, and mass) of the species and the uptake processes. The hydrodynamic diameter and zeta potential of the dendriplexes were, respectively, 107.3 nm and 39.80 mV compared to values for the parent dendrimer of 6.5 nm and 47.27 mV. The positive surface charge of both compounds facilitates electrostatic attraction with the cell membrane as a prelude to being taken up by endocytosis. Concomitantly, the dendrimer may be taken up by transient nanopores, which can be created after adherence to the cell membrane, allowing also passive diffusion and resulting in the shorter uptake process [28–32]. The larger size of the dendriplexes would attenuate their diffusion inside the cytoplasm. In comparison to the native dendrimer the diffusion coefficient of the dendriplex is reduced by a factor of 4. One might have expected a larger decrease simply because the radius has increased by ∼15-fold from 6.5 to 107 nm (D0 of dendrimer and dendriplexes are 79.54 × 10−8 and 4.53 × 10−8 cm2/s, respectively) in accord with the Stokes–Einstein equation. The diffusion coefficient in water of the dendriplexes is 20 times lower than that of the dendrimer. Although particle size precisely controls the diffusion of the particle in bulk or aqueous solution, it was found to also play a more complex role in the diffusion of the two particle types discussed here. Dcyto values of dendriplexes and dendrimer were reduced from 9.82 × 10−11 to 2.36 (± 0.34 (SD)) × 10−11 cm2/s (fourfold reduction). This may be related somehow to uptake pathways. In the cells, diffusion coefficients generally can be explained by the obstruction effect and sieving effects, which are a function of particle size (Figure 3.4A) [13, 25, 47–49]. Janson and Luby-Phelps proposed the pore slit model having an average cutoff radius between 15 and 50 nm. Following this model the diameter of the dendriplexes is beyond the maximum cutoff size and thus cannot pass through the cellular meshwork, which presents as the slow mobility of the dendriplexes at the rim of the plasma membrane, and their ability to traverse the cytoplasm. 3.5 MECHANISMS OF PARTICLE UPTAKE INTO THE CELL NUCLEUS The cell nucleus is the ultimate target to achieve efficient gene transfection. The exact process by which lipoplexes deliver their plasmid DNA to nuclei after escaping from endosomes has remained elusive. The nuclear pore complex (NPC) responsible for nucleocytoplasmic transport [51, 52] plays a fundamental role in human and other eukaryotic cells. The NPC possesses remarkable transport competencies with two distinct modes: passive and facilitated (or active) transport [53, 54]. Passive transport is a nonspecific process that involves ordinary diffusion of matter through the nuclear pore with a cutoff at about 10 nm in diameter [50] and/or molecular mass less than 25 kDa [53]. Facilitated transport is the highly specific process acting against concentration gradients [55] with cutoff diameters of approximately 50 nm [50] and a molecular mass (m) range of 25 < m < 75 kDa [53]. This limit easily includes

48

CYTOPLASMIC DIFFUSION OF DENDRIMERS AND DENDRIPLEXES

the size of most protein transport substrates and macromolecules, for which the molecular weight of the cargo–receptor complex should be in the low hundreds of kilodaltons. In addition, Thachenko et al. [56] have suggested a cutoff size for particles for targeted nuclear delivery; particles should have diameters less than 100 nm to enter the cell membrane and less than 30 nm to be imported through the NPC, respectively. The passive diffusion of the DNA through nuclei in nonmitotic cells has been reported [57, 58]. Salman et al. [57] found that the uptake of DNA, 2 nm in linear dimensions, is independent of ATP or GTP hydrolysis, indicating linear diffusion without the need for conformation change or specific biochemical interaction with nuclear pore complexes (NPCs). The kinetic results, however, show diffusion to be much slower than would be estimated from purely hydrodynamic considerations. This perhaps suggested that the DNA may possible disassemble from its carrier before it passively diffuses through the NPC into the nucleus. Once through the NPC the journey of the DNA is not over. The nucleus is crowded. Lukacs et al. [59] have found that DNA diffusion in the nucleus is negligible—and, unlike diffusion in the cytoplasm, independent of DNA size, in agreement with our data shown in Figure 3.2. Active transport entails nuclear localization signal (NLS) peptides that can be linked to the plasmid DNA; this peptide triggers the transport of the DNA through the NPC by forming a complex with nuclear membrane importers, but such a strategy does not always produce higher transfection levels and can indeed reduce the effect, possibly because the added peptide changes the physical properties of the complex, not least its size or susceptibility to aggregation [60, 61]. The dendrimers and the dendriplexes traverse the cell cytoplasm to the nucleus as can be proposed in the model in Figure 3.4. The 1000- and 2000 fold reduction in diffusion coefficient of the dendrimer and the dendriplexes in the cell cytoplasm can be dissected into two main steps: diffusion within cell cytoplasm and passage through the nuclear pore complex (NPC). The factors involved in the first step are the obstruction and exclusion effects regarding the molecular crowding and the sieving effect mentioned earlier. The cell cytoplasm is depicted here acting as a meshwork, providing a sieving effect with maximum cutoff size at 50 nm in radius (Figure 3.4A). Likewise for a gel, the meshwork structure and crowded state appear as an insoluble compartment through which the dendrimer and the dendriplexes cannot diffuse (Figure 3.4B). Thus the dendrimer and the dendriplexes have to detour in the nonexcluded aqueous volume, until they reach their destination. After endocytosis or passive diffusion through transient nanopores ( – ), the dendrimer (red arrow) and the dendriplexes (green arrow) diffuse toward the nuclear compartment. Some of the dendrimer may interact with cell organelles ( ) and the interaction can be found after endosomal escape of the dendriplexes ( ). Particle size and type of carrier system seem to play a decisive role in the diffusion of the particles in this first step. After endosomal escape, the debate is still whether or not DNA disassembles from the dendrimer and how the DNA reaches the nucleus. Knowledge

MECHANISMS OF PARTICLE UPTAKE INTO THE CELL NUCLEUS

49

A

Cell radius (h) B 4 1

Dendriplexes Dendrimer

2 3

Cell membrane

Nuclear membrane

FG surface

C 1

Cytoplasmic filaments Threading area

2

Passive transport

Selectivity filter Nuclear filaments

3

Active transport

Figure 3.4. An uptake model for dendrimer and dendriplexes: (A) demonstrates the sieving effect of the cytoplasmic meshwork with a cutoff size of 100 nm in diameter. The crowded condition and the obstruction effect cause the slow diffusion of the dendrimer and dendriplexes after endocytosis ( – ) or passive diffusion ( ) through transient nanoholes, 1–10 nm in diameter, illustrated in (B). The dendrimer and dendriplexes have to detour in the aqueous compartment excluding crowded volumes. Some of the dendrimer and dendriplexes may interact with cell organelles or cytoplasmic proteins ( ). Passing through the NPC, the DNA released from the endosome ( ) may cross the NPC by passive diffusion ( ). As small as 6.5 nm, the dendrimer can diffuse through the central pore having a cutoff size of 8–10 nm ( ), whereas the larger of the dendriplexes, 107 nm, may employ active transport ( ). (Part C – is modified from Peters [50]; other parts are original.) (See color insert.)

50

CYTOPLASMIC DIFFUSION OF DENDRIMERS AND DENDRIPLEXES

of transport mechanisms through the nuclear pore complexes (NPCs) might help to make this clear. A vertebrate somatic cell usually contains between 1000 and 10,000 NPCs [62]. Transport of the DNA through the NPC is assumed to be the rate-limiting step for gene therapy, a process divided into two strategies: passive and active pathways (Figure 3.4C). Particles with diameters between 8 and 10 nm can be transported by passive diffusion, whereas particles of larger size and particles with molecular masses between 25 and 75 kDa [53] utilize active transport processes. Recent models of the NPC have been proposed as a hydrogel [60, 63, 64], as seen in Figure 3.4C. The selectivity is supplied by the meshwork construction of the phenylalanine glycine (FG) motif and the nonselectivity or passive transport is acquired from the aqueous tube of the channel center, 8–10 nm in diameter. DNA and dendrimers may passively diffuse though this channel (Figure 3.4C – ), while the dendriplexes may have to interact as transport complexes in the perinuclear membrane. These transport complexes are searching for the FG motif in the “threading area,” seen in Figure 3.4C , and they will only selectively passage via FG active transport pathways through the NPC (Figure 3.4C ).

3.6

CONCLUSION

The intrinsically fluorescent lysine-based dendrimer used in our work has proved to be useful as a nanofluorescent probe for tracking uptake and translocation within the cells, as it possesses a fluorescent signal detected by both spectrofluorimetry and confocal microscopy. The uptake of the dendrimer and dendriplexes formed from it depend on the cell type and naturally incubation periods. The dendrimer is taken up into the nuclear compartment. Dendriplexes, because of their size, do so also at a slower rate. One uptake mechanism investigated here was found to be due to endocytic-mediated endocytosis. Other uptake mechanisms have also been proposed. A model of the uptake pathway of the dendrimer and the dendriplexes from the cytoplasm throughout the nuclear pore complexes (NPCs) has been proposed here. Macromolecular crowding in the cytoplasm presents a significant barrier to diffusion of dendrimer and dendriplexes and this has been discussed here in terms of exclusion and obstruction effects. Finally, the uptake of the dendrimer was found to be impeded by the influence of fluid medium flow. The diffusion of the dendrimer and the dendriplexes within the cell nucleus may be expected to be dependent on similar key factors obtained inside the cell cytoplasm, but the nucleus is even more crowded than the cytoplasm. Whether or not the dendriplexes release their DNA before or after passing through the NPC is not known. The transport of the parent dendrimer through the NPC into the cell nucleus is expected to be dependent on passive diffusion. Further investigation to clarify this issue is indispensable so that more successful gene delivery systems can be developed. The main challenge is to determine why there are differences in transfection and diffusion in different cell

REFERENCES

51

types, and more work needs to be done on tissues in vivo, with the challenges of getting sufficiently high concentration gradients of nanoparticles such as dendimers and dendriplexes. The physical nature of the cytoplasm in so far as it allows translocation of particles requires further definition and investigation. In conclusion, an understanding of movement of delivery vectors within cells and in the nucleus is essential for the development of effective systems. Good models of the physical structure of cell interiors are required along with probes such as the one described here to advance the design of new delivery vectors. REFERENCES 1. Tomalia, D. A. Birth of a new macromolecular architecture: dendrimers as quantized building blocks for nanoscale synthetic polymer chemistry. Prog. Polym. Sci. 30: 294–324 (2005). 2. Perumal, O. P., Inapagolla, R., Kannan S., and Kannan, R. M. The effect of surface functionality on cellular trafficking of dendrimers. Biomaterial 29: 3469–3476 (2008). 3. El-Sayed, M., Kiani, M. F., Naimark, M. D., Hikal, A. H., and Ghandehari, H. Extravasation of poly(amidoamine) (PAMAM) dendrimers across microvascular network endothelium. Pharm. Res. 18: 23–28 (2001). 4. Preston, R. A., Murphy, R. F., and Jones, E. W. Apparent endocytosis of fluorescein isothiocyanate-conjugated dextran by Saccharomyces cerevisiae reflects uptake of low molecular weight impurities, not dextran. J. Cell Biol. 105: 1981–1987 (1987). 5. Lundberg, M. and Johansson, M. Positively charged DNA-binding proteins cause apparent cell membrane translocation. Biochem. Biophys. Res. Commun. 291: 367–371 (2002). 6. Richard, J. P., Melikov, K., Vives, E., Ramos, C., Verbeure, B., Gait, M. J., Chernomordik, L. V., and Lebleu, B. Cell-penetrating peptides: a reevaluation of the mechanism of cellular uptake. J. Biol. Chem. 278: 585–590 (2003). 7. Al-Jamal, K. T., Ruenraroengsak, P., Hartell, N., and Florence, A. T. An intrinsically fluorescent dendrimer as a nanoprobe of cell transport. J. Drug Target. 14: 405–412 (2006). 8. Dickinson, R. B. and Purich, D. I. Diffusion rate limitations in actin-based propulsion of hard and deformable particles. Biophys. J. 91: 1548–1563 (2006). 9. Ruenraroengsak, P. and Florence, A. T. The diffusion of latex nanospheres and the effective (microscopic) viscosity of HPMC gels. Int. J. Pharm. 298: 361–366 (2005). 10. Ellis, R. J. and Minton, A. P. Cell biology: join the crowd. Nature 425: 27–28 (2003). 11. Fulton, A. B. How crowded is the cytoplasm? Cell 30: 345–347 (1982). 12. Goodsell, D. S. Inside a living cell. TIBS 16: 203–206 (1991). 13. Luby-Phelps, K. Cytoarchitecture and physical properties of cytoplasm: volume, viscosity, diffusion, intracellular surface area. Int. Rev. Cytol. 192: 189–221 (2000).

52

CYTOPLASMIC DIFFUSION OF DENDRIMERS AND DENDRIPLEXES

14. Medalia, O., Weber, I., Frangakis, A. S., Nicastro, D., Gerisch, G., and Baumeister, W. Macromolecular architecture in eukaryotic cells visualized by cryoelectron tomography. Science 298: 1209–1213 (2002). 15. Zimmerman, S. B. and Minton, A. P. Macromolecular crowding: biochemical, biophysical, and physiological consequences. Annu. Rev. Biophys. Biomol. Struct. 22: 27–65 (1993). 16. Kao, H. P., Abney, J. R., and Verkman, A. S. Determinants of the translational mobility of a small solute in cell cytoplasm. J. Cell Biol. 120: 175–184 (1993). 17. Furukawa, R., Arauz-Lara, J. L., and Ware, B. R. Self-diffusion and probe diffusion in dilute and semidilute aqueous solutions of dextran. Macromolecules 24: 599–605 (1991). 18. Phillies, G. D. J. The hydrodynamic scaling model for polymer self-diffusion. J. Phys. Chem. 93: 5029–5039 (1989). 19. Hou, L., Lanni, F., and Luby-Phelps, K. Tracer diffusion in F-actin and Ficoll mixtures. Toward a model for cytoplasm. Biophys. J. 58: 31–43 (1990). 20. Clegg, J. S. Properties and metabolism of the aqueous cytoplasm and its boundaries. Am. J. Physiol. 246: R133–R151 (1984). 21. Mastro, A. M., Babich, M. A., Taylor, W. D., and Keith, A. D. Diffusion of a small molecule in the cytoplasm of mammalian cells. Proc. Natl. Acad. Sci. U.S.A. 81: 3414–3418 (1984). 22. Nicolay, K., van der, T. A., and Dijkhuizen, R. M. In vivo diffusion spectroscopy: an overview. NMR Biomed. 8: 365–374 (1995). 23. Verkman, A. S., Armijo, M., and Fushimi, K. Construction and evaluation of a frequency-domain epifluorescence microscope for lifetime and anisotropy decay measurements in subcellular domains. Biophys. Chem. 40: 117–125 (1991). 24. Luby-Phelps, K., Mujumdar, S., Mujumdar, R. B., Ernst, L. A., Galbraith, W., and Waggoner, A. S. A novel fluorescence ratiometric method confirms the low solvent viscosity of the cytoplasm. Biophys. J. 65: 236–242 (1993). 25. Luby-Phelps, K., Castle, P. E., Taylor, D. L., and Lanni, F. Hindered diffusion of inert tracer particles in the cytoplasm of mouse 3T3 cells. Proc. Natl. Acad. Sci. U.S.A. 84: 4910–4913 (1987). 26. Aschenbrenner, L., Naccache, S. N., and Hasson, T. Uncoated endocytic vesicles require the unconventional myosin, myo6, for rapid transport through actin barriers. Mol. Biol. Cell 15: 2253–2263 (2004). 27. Conner, S. D. and Schmid, S. L. Regulated portals of entry into the cell. Nature 422: 37–44 (2003). 28. Hong, S., Bielinska, A. U., Mecke, A., Keszler, B., Beals, J. L., Shi, X., Balogh, L., Orr, B. G., Baker, J. R., Jr., and Banaszak Holl, M. M. Interaction of poly(amidoamine) dendrimers with supported lipid bilayers and cells: hole formation and the relation to transport. Bioconjug. Chem. 15: 774–782 (2004). 29. Hong, S., Leroueil, P. R., Janus, E. K., Peters, J. L., Kober, M. M., Islam, M. T., Orr, B. G., Baker, J. R., Jr., and Banaszak Holl, M. M. Interaction of polycationic polymers with supported lipid bilayers and cells: nanoscale hole formation and enhanced membrane permeability. Bioconjug. Chem. 17: 728–734 (2006). 30. Mecke, A., Lee, D. K., Ramamoorthy, A., Orr, B. G., and Banaszak Holl, M. M. Membrane thinning due to antimicrobial peptide binding: an atomic force microscopy study of MSI-78 in lipid bilayers. Biophys. J. 89: 4043–4050 (2005).

REFERENCES

53

31. Mecke, A., Lee, D. K., Ramamoorthy, A., Orr, B. G., and Holl, M. M. Synthetic and natural polycationic polymer nanoparticles interact selectively with fluid-phase domains of DMPC lipid bilayers. Langmuir 21: 8588–8590 (2005). 32. Mecke, A., Majoros, I. J., Patri, A. K., Baker, J. R., Jr., Holl, M. M., and Orr, B. G. Lipid bilayer disruption by polycationic polymers: the roles of size and chemical functional group. Langmuir 21: 10348–10354 (2005). 33. Morris, M. C., Deshayes, S., Heitz, F., and Divita, G. Cell-penetrating peptides: from molecular mechanisms to therapeutics. Biol. Cell 100: 201–217 (2008). 34. Mislick, K. A. and Baldeschwieler, J. D. Evidence for the role of proteoglycans in cation-mediated gene transfer. Proc. Natl. Acad. Sci. U.S.A. 93: 12349–12354 (1996). 35. Manunta, M., Izzo, L., Duncan, R., and Jones, A. T. Establishment of subcellular fractionation techniques to monitor the intracellular fate of polymer therapeutics II. Identification of endosomal and lysosomal compartments in HepG2 cells combining single-step subcellular fractionation with fluorescent imaging. J. Drug Target. 15: 37–50 (2007). 36. Woodle, M. C. and Scaria, P. Cationic liposomes and nucleic acids. Curr. Opin. Colloid Interface Sci. 6: 78–84 (2001). 37. Manunta, M., Nichols, B. J., Tan, P. H., Sagoo, P., Harper, J., and George, A. J. Gene delivery by dendrimers operates via different pathways in different cells, but is enhanced by the presence of caveolin. J. Immunol. Methods 314: 134–146 (2006). 38. Elouahabi, A. and Ruysschaert, J. M. Formation and intracellular trafficking of lipoplexes and polyplexes. Mol. Ther. 11: 336–347 (2005). 39. Ruenraroengsak, P., Al-Jamal, K. T., Hartell, N., Braeckmans, K., De Smedt, S. C., and Florence, A. T. Cell uptake, cytoplasmic diffusion and nuclear access of a 6.5 nm diameter dendrimer. Int. J. Pharm. 331: 215–219 (2007). 40. Sagidullin, A. I., Muzafarov, A. M., Krykin, M., Ozerin, A. N., Skirda, V. D., and Ignat’eva, G. M. Generalized concentration dependence of self-diffusion coefficients in poly(allylcarbosilane) dendrimer solution. Macromolecules 35: 9472–9479 (2002). 41. Xia, T., Kovochich, M., Liong, M., Zink, J. I., and Nel, A. E. Cationic polystyrene nanosphere toxicity depends on cell-specific endocytic and mitochondrial injury pathways. ACS Nano 2: 85–96 (2008). 42. Kaether, C. and Gerdes, H. H. Monitoring of protein secretion with green fluorescent protein. Methods Enzymol. 302: 11–19 (1999). 43. Bayele, H. K., Sakthivel, T., O’Donell, M., Pasi, K. J., Wilderspin, A. F., Lee, C. A., Toth, I., and Florence, A. T. Versatile peptide dendrimers for nucleic acid delivery. J. Pharm. Sci. 94: 446–457 (2005). 44. Arrio-Dupont, M., Foucault, G., Vacher, M., Devaux, P. F., and Cribier, S. Translational diffusion of globular proteins in the cytoplasm of cultured muscle cells. Biophys. J. 78: 901–907 (2000). 45. Tseng, Y. and Wirtz, D. Mechanics and multiple-particle tracking microheterogeneity of alpha-actinin-cross-linked actin filament networks. Biophys. J. 81: 1643– 1656 (2001). 46. Fushimi, K. and Verkman, A. S. Low viscosity in the aqueous domain of cell cytoplasm measured by picosecond polarization microfluorimetry. J. Cell Biol. 112: 719–725 (1991).

54

CYTOPLASMIC DIFFUSION OF DENDRIMERS AND DENDRIPLEXES

47. Janson, L. W., Ragsdale, K., and Luby-Phelps, K. Mechanism and size cutoff for steric exclusion from actin-rich cytoplasmic domains. Biophys. J. 71: 1228–1234 (1996). 48. Luby-Phelps, K. Effect of cytoarchitecture on the transport and localization of protein synthetic machinery. J. Cell Biochem. 52: 140–147 (1993). 49. Provance, D. W. Jr., McDowall, A., Marko, M., and Luby-Phelps, K. Cytoarchitecture of size-excluding compartments in living cells. J. Cell Sci. 106(2): 565–577 (1993). 50. Peters, R. Introduction to nucleocytoplasmic transport: molecules and mechanisms. Methods Mol. Biol. 322: 235–258 (2006). 51. Allen, T. D., Cronshaw, J. M., Bagley, S., Kiseleva, E., and Goldberg, M. W. The nuclear pore complex: mediator of translocation between nucleus and cytoplasm. J. Cell Sci. 113(10): 1651–1659 (2000). 52. Rout, M. P., Aitchison, J. D., Suprapto, A., Hjertaas, K., Zhao, Y., and Chait, B. T. The yeast nuclear pore complex: composition, architecture, and transport mechanism. J. Cell Biol. 148: 635–651 (2000). 53. Lusk, C. P., Blobel, G., and King, M. C. Highway to the inner nuclear membrane: rules for the road. Nat. Rev. Mol. Cell Biol. 8: 414–420 (2007). 54. Bickel, T. and Bruinsma, R. The nuclear pore complex mystery and anomalous diffusion in reversible gels. Biophys. J. 83: 3079–3087 (2002). 55. Dingwall, C., Sharnick, S. V., and Laskey, R. A. A polypeptide domain that specifies migration of nucleoplasmin into the nucleus. Cell 30: 449–458 (1982). 56. Tkachenko, A. G., Xie, H., Liu, Y., Coleman, D., Ryan, J., Glomm, W. R., Shipton, M. K., Franzen, S., and Feldheim, D. L. Cellular trajectories of peptide-modified gold particle complexes: comparison of nuclear localization signals and peptide transduction domains. Bioconjug. Chem. 15: 482–490 (2004). 57. Salman, H., Zbaida, D., Rabin, Y., Chatenay, D., and Elbaum, M. Kinetics and mechanism of DNA uptake into the cell nucleus. Proc. Natl. Acad. Sci. U.S.A. 98: 7247–7252 (2001). 58. de Gennes, P. G. Passive entry of a DNA molecule into a small pore. Proc. Natl. Acad. Sci., U.S.A. 96: 7262–7264 (1999). 59. Lukacs, G. L., Haggie, P., Seksek, O., Lechardeur, D., Freedman, N., and Verkman, A. S. Size-dependent DNA mobility in cytoplasm and nucleus. J. Biol. Chem. 275: 1625–1629 (2000). 60. Sakthivel, T., Toth, I., and Florence, A. T. Synthesis and physicochemical properties of lipophilic polyamide dendrimers. Pharm. Res. 15: 776–782 (1998). 61. Toth, I., Sakthivel, T., Wilderspin, A. F., Toth, I., Bayele, H. K., O’Donnell, M., Perry, D. J., Pasi, K. J., Lee, C. A., and Florence, A. T. Novel cationic lipidic peptide dendrimer vectors—in vitro gene delivery. STP Pharm. Sci. 9: 93–99 (1999). 62. Burke, B. Cell biology. Nuclear pore complex models gel. Science 314: 766–767 (2006). 63. Frey, S., Richter, R. P., and Gorlich, D. FG-rich repeats of nuclear pore proteins form a three-dimensional meshwork with hydrogel-like properties. Science 314: 815–817 (2006). 64. Elbaum, M. Materials science. Polymers in the pore. Science 314: 766–767 (2006).

CHAPTER 4

Endocytosis and Intracellular Trafficking of Quantum Dot–Ligand Bioconjugates TORE-GEIR IVERSEN, NADINE FRERKER, and KIRSTEN SANDVIG Centre for Cancer Biomedicine, Faculty Division, Norwegian Radium Hospital, University of Oslo, and Department of Biochemistry, Institute for Cancer Research, Norwegian Radium Hospital, Oslo University Hospital, Oslo, Norway

4.1

INTRODUCTION

Nanoparticles (NPs) are gaining increasing relevance for use in biological and biomedical applications both within diagnostic imaging and therapeutic drug delivery systems. These applications require the targeting of specific tissues and cells, and important for several of the actions is the uptake of NPs into cells. In vivo and in vitro labeling with NPs allow detection and tracing of molecules into cells as well as cell fractionation. The entry of NPs into a targeted cell is determined and restricted by the endocytic pathways operating. Not all endocytic mechanisms have been identified and characterized so far. The following paragraphs give a short overview of the most studied pathways and introduce important factors and observations relevant in endocytosis and trafficking of NPs.

4.2

DIVERSITY OF ENDOCYTIC MECHANISMS

Endocytosis is a process in which molecules are internalized into the cell interior without passing through the membrane. There are different types of Organelle-Specific Pharmaceutical Nanotechnology, Edited by Volkmar Weissig and Gerard G. M. D’Souza Copyright © 2010 John Wiley & Sons, Inc.

55

56

ENDOCYTOSIS AND INTRACELLULAR TRAFFICKING

pinocytosis

phagocytosis clathrin-mediated endocytosis

clathrin-independent endocytosis

dynamin-dependent endocytosis

dynamin-independent endocytosis

RhoA = dynamin = clathrin

caveolae

Cdc42

Arf6

Rac ruffling

Figure 4.1. Endocytic pathways. Phagocytosis (left: enclosure of large particle) and different forms of endocytosis—clathrin-mediated endocytosis and clathrinindependent forms such as macropinocytosis, caveolae, RhoA-, Cdc42-, and Arf6mediated endocytosis.

endocytosis—each involving the formation of intracellular vesicles by means of invagination of the plasma membrane and membrane fission. In general, a rough division can be made between phagocytosis (“cell eating”) and pinocytosis (“cell drinking”). Large particles such as bacteria are taken up via phagocytosis. The particle is engulfed by extrusions from the cell membrane (a “phagocytic cup”), which then mediates the formation of a phagosome. Phagocytosis occurs mainly in specialized mammalian cells such as macrophages, monocytes, and neutrophils but also to a lesser extent in nonprofessional phagocytes (i.e. fibroblasts, endothelial and epithelial cells). Phagocytosis constitutes the initial step for the degradation of particles larger than 0.5 µm. Pinocytosis is used to internalize fluid simultaneously with whatever substance is found within the area of invagination. There are multiple types of endocytotic pathways, which can be classified in part based on their requirement for the coat protein clathrin, the GTPase dynamin, or the formation of caveolae. Thus a general division into clathrindependent/independent and dynamin-dependent/independent endocytic pathways can be made (Figure 4.1). 4.2.1 Clathrin-Dependent Endocytosis The clathrin-dependent endocytosis is one of the best characterized endocytic pathways and constitutes a major route for selective receptor internalization

DIVERSITY OF ENDOCYTIC MECHANISMS

57

in eukaryotic cells [1]. The formation of a clathrin-coated pit starts with adaptor protein-mediated recruitment of clathrin to the plasma membrane, where the adaptor proteins bind transmembrane receptor molecules with cargo. Finally, the coated pit invaginates and, regulated by the GTPase dynamin and GTP, pinches off to form a clathrin-coated vesicle. In general, clathrin-mediated endocytosis (CME) is a rapid uptake process ( 0; CBN < 40. No binding to cells requires that if pKa >7 and CBN < 40, then log P < 0; or if pKa >7 and CBN > 40, then log P < −10; or if pKa ∼ 7 and CBN < 40, then log P (most ionized species) log P > 0; Z = 0. Endoplasmic reticulum membranes require that 6 > AI > 3.5; 6 > log P > 0; Z > 0. Fat droplets require that AI < 3.5; HGS < ∼20; log P > 5.0; pKa (if relevant) < 6. Generic biomembranes, for reversible uptake, require that 8 > AI or log P > 5. Golgi membranes require that 5 > log P > 3. Lysosomes/acidic organelles, weak base ion-trapping requires that 0 > log P (cation) > −5; 10 > pKa > 6; Z > 0. Lysosomes/acidic organelles, weak acid precipitation trapping requires that log P (free acid) > 0; pKa = 7 ± 3; Z(most ionized species) < 0. Lysosomes and endosomes, membranes, lipid domains require that AI or log P > 8. Lysosomes and endosomes, membranes, protein domains require that CBN > 40. Mitochrondria, potential driven and/or cardiolipid complexation require that 5 > log P > 0; pKa > 12; Z > 0. Mitochrondria, ion trapping of weak acids requires that 5 > log P (least ionized species) > 0; pKa = 7 ± 3; Z(most ionized species) < 0. Mitochondria, outer membrane, require that 6 > AI > 3.5; log P > 0; Z > 0. Nuclear DNA, chromatin requires that LCF > 24; −2 > log P(cation) > 0; pKa > 10; Z(ionized species) > 0. Nuclear histones require that if Z(most ionized species) < 0, and CBN > 40, then log P > −10; or if Z < 0 and CBN > 16, then 8 > log P(anion) > 0. Nucleoli, nucleic acids require that LCF > 17; 8 > log P(least ionized species) > 0; pKa > 10; Z > 0. Phagosomes require particles with a “diameter” greater than ∼500 nm. Plasma membrane uptake, fluid lipid domains (e.g., in apoptotic cells) require that ∼5.5 > AI > 3.5; log P < 5.0; HGS > 400. Plasma membrane uptake, lipid domains require that AI or log P > 8 or AI > 5; HGS > 400. Plasma membrane uptake, protein domains require that log P < −10; CBN > 40. Ribosomal RNA requires that LCF > 17; 8 > log P(least ionized species) > 0; pKa > 10; Z > 0.

198

CAN QSAR MODELS FOR SMALL-MOLECULE XENOBIOTICS GIVE USEFUL TIPS?

11.4 MORE BACKGROUND: WHAT FACTORS CONTROL UPTAKE AND INTRACELLULAR LOCALIZATION OF NANOPARTICLES? First, a disclaimer: this chapter, as acknowledged above, does not discuss the major approaches to the targeting of nanoparticles adopted to date. Most investigators have been focused on the development and application of “smart targeting” mediated by various entities. Thus a recent review of the cellular and intracellular targeting of quantum dots gave considerable attention to work involving targeting by antibodies, by ligands for cell-surface receptors such as folate, and by peptide ligands for various tumor cell markers, with some attention also directed at ligands binding to receptors for angiotensin and neurotransmitters [1]. Other interactions of nanoparticles and cells were discussed under the heading of “nonspecific particle uptake,” which is precisely the type of phenomenon we will be considering in the present chapter. Given the focus on smart targeting, it is perhaps not surprising that there is a paucity of rigorous investigations into such “nonspecific” mechanisms of uptake and localization of nanoparticles within living cells, not only because of the issues noted above, but also because preparation of novel nanoparticles is within the technical capabilities of many chemical and related laboratories. Indeed, a wait-and-see approach to such a complex problem as nanoparticle uptake and targeting may be a valid, if not the only, way to successfully solve pressing problems of drug delivery and toxicology. However, we will now briefly review some of the available rigorous studies, first concerning uptake into cells, and then the post-uptake localization within cells. It is immediately apparent, and of course no surprise, that the most dramatic particle-associated variable influencing cellular uptake is the size of the nanoparticle. Of course, nanoparticles are (depending on definitions, see above) very large objects to pass through the plasma membrane in the same manner that is easily possible for small molecules entering a cell by passive diffusion. Instead by far the most common mode of entry is by endocytosis, whether the particles are lipoplexes [21], liposomes [22], quantum dots [23], or some other variety. At the extreme, it was observed as long ago as the early twentieth century—when India ink was used as the probe—that particles exceeding approximately 0.5 µm in diameter were extensively taken up by phagocytes [24]; for more modern accounts see Besterman and Low [25] and Haas [26]. However, for “true” nanoparticles, a discriminating and critical investigation has been carried out more recently by Hoekstra’s group [2]. This involved exposing nonphagocytic eukaryotic cells to fluorescent latex beads ranging in size from 50 to 1000 nm. All except the largest (1000-nm) particles were taken up, by energy-dependent processes. In the size range up to 200 nm this involved a clathrin-mediated process, with the particles finally arriving in late endosomal/lysosomal compartments. In the larger size range, above 200 nm to below 1000 nm, uptake involved a caveolae-mediated process, with the final destination not being lysosomal. This latter process was distinct from

MORE BACKGROUND: WHAT FACTORS CONTROL UPTAKE

199

macropinocytosis, which can also favor larger particles [27]. These size criteria have been found applicable to a variety of cell types (e.g., epithelial [28] and fibroblast [29]) and to diverse types of particle (e.g., lipoplexes [30] and porphyrin aggregates [31]. Endocytic uptake, however, is modulated by other factors. One such is a particle’s surface electric charge. Thus endocytotic uptake of nanoparticles of very widely varying types is often facilitated by a positive surface charge, as reviewed, for instance, by Blau et al. [32]. This is generally considered to be due to an initial adsorption to surface negative charges, usually on the plasma membrane. Indeed, this does often result in an observed increase (short or long term) of the surface adhesion of the nanoparticles at that stage [33]. However, a positive surface charge does not always result in entry into the cell, since if a cell has an external coating of an anionic glygosaminoglycan, for instance, then a cationic nanoparticle can be trapped in that layer [34]. Another surface property dramatically modifying endocytic uptake is the lipophilicity or hydrophilicity of the particle’s surface. In vivo, it was observed many years ago that after intravenous administration nanoparticles were rapidly removed from the circulation as a result of being phagocytosed by macrophages, especially in bone marrow, liver, lungs, and spleen [35]. It was shown that this was largely due to adsorption of proteins onto the particle surface resulting in opsonization [36]. In keeping with this, occluding the hydrophobic surface with hydrophilic moieties reduces protein binding and consequently phagocytic endocytosis, as demonstrated quantitatively by Chang’s group [37]. Hydrophilic materials as varied as polyethylene glycol [38] and polysaccharides [39] have been used in this way. Of course, this reduction of competition from macrophages enhances uptake into other cell types by other endocytic mechanisms. Before moving on to consider the intracellular localization of nanoparticles, it should be noted that their entry into cells is not entirely restricted to endocytosis. For instance, membrane fusion can also enable internalization, albeit largely that of liposomes. However, even for those structures it is not the most common mechanism (although it has given rise to some wonderful electron micrographs; e.g., see Dini et al. [40], Figure 5d). Crucially, membrane fusion can result in direct delivery of hydrophilic materials into the cytosol [41]. In addition, some nanoparticles do cross the plasma membrane. Least surprising is to see this occuring with small (∼1 nm) fullerene derivatives, which include lipophilic species [42, 43]. More surprisingly, it has also been reported for particles with diameters above 10 nm. Thus Jablonski et al. [44] have developed a system involving coating the particle with a cationic species with a subsequent application of a hydrophobic counterion, pyrenebutyrate, in the published example. This is not a permeabilization process, and the negative surface charge of the particle means that the plasma membrane potential strongly influences uptake. Another process involves coating quantum dots with amphiphilic polymers [45].

200

CAN QSAR MODELS FOR SMALL-MOLECULE XENOBIOTICS GIVE USEFUL TIPS?

Post-uptake intracellular localization of nanoparticles will now be considered. As described above, particle size strongly influences which particular endocytotic mechanism is involved. Consequently, particle size substantially controls whether the nanoparticle will end up in a lysosome or an acidified late endosome, or instead in a nonacidified vesicle. Once again, however, a particle’s electric charge modulates this: for instance, in directing down which intracellular pathway the nanoparticles travel. A recently reported example involves the application of various anionic and cationic derivatives of poly(ethylene glycol)-d,l-polylactide nanoparticles, all with diameters between 50 and 100 nm, to cultured epithelial cells. Most were taken up via the clathrin pathway. However, whereas many anionic particles transited through the lysosomal route, cationic particles typically did not but instead diverted to the lateral plasma membrane [46]. Ionization can also play a role in escape from endosomes and related vesicles. For instance, weak bases such as protonsponge dendrimers can become ionized within acidified vesicles, thus dramatically raising the local osmotic pressure, favoring vesicle rupture [47]. Amphilicity, and the molecular features influencing this property, also has an impact on membrane rupture and so on escape of particles from endosomes. For instance, nanoparticles will escape more readily if their surfaces carry amphiphils with surfactant properties. In particular, this effect is influencial in lipoplexes, where chain length effects have been discussed using the structure parameter AI [48]. A related, but distinct, property influencing membrane rupture and fusion is lipid shape. In this case it is membrane curvature that is the key property. This is controlled both by the relative sizes of the head group and hydrocarbon tail, and by the cross-sectional area of the head group; for an accessible account of which Israelachvili’s [49] formulation is still to be recommended.

11.5 SO CAN SMALL-MOLECULE QSAR MODELS TELL US ANYTHING ABOUT NANOPARTICLE UPTAKE AND LOCALIZATION? As discussed above, the factors found to be of significance do include those considered in the small-molecule models—namely, amphilicity, electric charge, head group size, lipophilicity, and pKa. However, it is hard to see how the small-molecule models can be applied directly to such vast structures as nanoparticles. So is that all? Are there are no regularities to which the smallmolecule QSAR models can be applied, or even any useful tips? It does seem that there are two distinct ways in which numerical structure–localization rules may be usefully applied to nanoparticle uptake into cells. First, consider nanostructures at the lower end of the size range, such as C60 fullerene derivatives. Such, if lipophilic, do penetrate the plasma membrane directly rather than by endocytosis. In one study, a dicarboxylic C60 fullerene was found to localize in biomembranes including the plasma mem-

SIMPLE NUMERICAL SUMMARIES OF NANOPARTICLE UPTAKE

201

brane and, most markedly, in mitochondria [43]. This fullerene is small enough for structure parameters to be estimated, and it was found to be a lipophilic weak acid, with some species being extremely lipophilic. Plugging the parameters into the relevant QSAR models, accumulation of the weak acid by iontrapping in mitochondria, and more general partitioning of the strongly lipophilic species into membranes, with superlipophilic species favoring the plasma membrane, matched observation. However, when a hexacarboxylic acid C60 fullerene was investigated it was observed to still accumulate in mitochondria, with cytoskeletal keratin fiber binding also occuring [42]. Again, estimation of structure parameters and application of the QSAR models gave congruent predictions, again of lipophilic weak acid ion trapping and, in this case, protein binding by the hydrophilic carboxylate anions. Morover, some larger nanoparticles do behave somewhat as suggested by the small-molecule models. Examples are provided by mitochondrial targeting of nanoparticles with lipophilic and cationic surfaces, noting that small molecules with lipophilic and cationic character are typically mitochondrially targeted. One such instance involved DQAsomes derived from dequalinium and its derivatives in intact cultured cells [16]. Another involved linking triphenylphosphonium to hydroxypropylmethacrylamide polymers [50], which nanoparticles appeared to show localization to mitochondria, but only with the smallest ( 1), complexation enhances cellular internalization of DNA by promoting electrostatic interactions between the positively charged complexes and negatively charged glycosaminoglycan residues displayed on the extracellular surfaces of cells. Finally, the condensing polymers or lipids provide a scaffold onto which a variety of ligands can be attached, with the goal of enhancing the in vivo properties or appropriate trafficking of the complexes.

210

SELF-UNPACKING GENE DELIVERY SCAFFOLDS

For example, cell-binding and other targeting ligands have been grafted to free amine residues within the vehicle to stimulate vehicle uptake and appropriate subcellular trafficking by target cells.

12.3

VEHICLE TRANSPORT BARRIERS

Despite promising developments in vitro, nonviral vehicles have had limited success in vivo and have rarely reached clinical application. The relatively poor in vivo gene transfer efficiency of this class of vehicles has been attributed to a variety of factors, including their inadequate tissue and cellular targeting and poorly controlled intracellular trafficking and processing. One significant and understudied issue is the lack of control over the timing and location of DNA release. Polyplexes and lipoplexes can be prematurely unpackaged by exposure to extracellular components such as heparan sulfate proteoglycans (HSPGs), leading to DNA degradation by serum nucleases [19–25]. Inefficient unpackaging is also problematic, as condensed DNA has a limited ability to interact with the transcriptional machinery [26]. The following subsections will highlight major barriers in the extra- and intracellular gene delivery pathways and will briefly review material design strategies that address these barriers. The subsequent section will include a detailed discussion of the factors that determine polyplex/lipoplex stability and unpackaging, as well as new material-based strategies for programming site-targeted DNA release and activation. 12.3.1 Serum Stability and Extracellular Transport The gene delivery pathway presents formidable challenges to the delivery vehicle immediately following the introduction of the vehicle into the extracellular environment (Figure 12.2A). For in vivo gene transfer via systemic administration, the initial challenge to the vehicle is survival in the bloodstream. In this environment, the half-life of a vehicle depends on its ability to avoid aggregation and protein adsorption following contact with serum components. Charged particles like polyplexes are destabilized by electrostatic screening in the 150 mM physiological saline environment, which can cause their rapid aggregation, reduced activity, and significant adverse consequences to the host such as particle embolization in the lung [8]. Interactions between polyplexes and lipoplexes and serum proteins have been shown to initiate rapid vehicle elimination by phagocytic cells and the reticuloendothelial system [22, 27, 28]. To avoid these issues, poly(ethylene glycol) (PEG) and other hydrophilic and uncharged polymers, proteins, or oligosaccharides are frequently grafted as a brush onto free amine residues within the condensing polycation to provide a steric barrier against salt-induced aggregation [22, 27, 29–31]. These molecules can also prevent protein- or complement-induced inactivation [29, 32–34].

VEHICLE TRANSPORT BARRIERS

211

(A)

(B)

Figure 12.2. Barriers to DNA transport. (A) Extracellular barriers. Poly- or lipoplex vehicles (black circles) travel through the host vasculature following administration, and must then extravasate (1) from the vasculature within the target tissue. Following extravasation, the vehicles must traverse the ECM containing collagen, proteoglycans, GAGs, and other ECM proteins to reach the target cell (2a). In some cases, interactions between the vehicle and the ECM can cause vehicle unpackaging and lead to the subsequent degradation of the DNA (2b). Inset image shown in (B). (B) Intracellular barriers. Vehicles enter cells by receptor-mediated endocytosis (3a) or macropinocytosis (3b). Subsequently, the early endocytic vesicles containing the vehicles are either recycled (4a) or mature into late endosomes (4b). Vehicles trapped within late endosomes might escape (5a) or be retained during endolysosomal maturation (5b); retention typically leads to DNA degradation. Released vehicles must travel through the cytosol to the vicinity of the nucleus (5a). Vehicles might enter the nucleus intact (6a) and subsequently unpackage (7a); alternatively, the vehicles might unpackage in the cytosol (6b), and the unpackaged DNA might enter the nucleus (7b) or be degraded by cytosolic nucleases (7c). (Author thanks John D. Larsen for assistance with figure preparation.)

212

SELF-UNPACKING GENE DELIVERY SCAFFOLDS

Provided the vehicle maintains sufficient serum stability, it must then begin the process of navigation to its target cell, which for nonhematopoietic and nonendothelial targets requires vascular extravasation and transport through the interstitial extracellular matrix (ECM) and/or lymphatics. As a general rule, methods to predictably control or promote vascular extravasation are not well established, given the low capillary permeability in most organs. For select tissues and organs, passive targeting strategies have been developed that take advantage of the unique physical properties of the target tissue. For example, nanoparticles such as polyplexes and lipoplexes tend to passively accumulate in the liver due to phagocytic clearance by the reticuloendothelial system [35–38]. A similar passive accumulation effect has been observed as a result of the aberrant architecture and increased vascular permeability and hydraulic conductivity of tumor endothelium [39–42]. This occurrence of tumor-specific accumulation, termed the enhanced permeability and retention (EPR) effect [43], is strongly size and molecular weight dependent. The average vascular pore sizes in various human and rodent tumors have been estimated as ∼400–500 nm in diameter [40], and macromolecules with sufficient serum stability and molecular weights greater than ∼50 kDa accumulate in tumor tissues. Transport of polyplexes and lipoplexes through dense ECM is mediated by a combination of convection and diffusion and depends on both their size and surface chemistry [44–46]. For example, in vitro studies employing hydrogels and multicellular spheroids have estimated the influence of nanoparticle characteristics on the efficiency of transport through simplified ECM. These studies suggest that particles with diameters of ∼100 nm or less can penetrate model matrices, but that penetration was relatively limited even for relatively small particles (∼20 nm) [47–49]. The surface chemistry of the nanoparticles was also critical. For example, PEGylated nanoparticles were found to move more rapidly through model ECM than unPEGylated nanoparticles, an effect attributed to reduced interactions between the PEGylated vehicles and fibrin networks within the ECM [50]. In another study, liposomes with neutral or low surface charges were found to penetrate to the central regions of spheroids, whereas charged liposomes were restricted [51]. A related issue with respect to the penetration of dense ECM by polyplexes and lipoplexes is the innate stability of the polyplexes and lipoplexes, and their tendency to unpackage and release DNA following contact with ECM components. This issue is covered in more detail in the following section on vehicle unpackaging. 12.3.2 Cellular Internalization and Intracellular Transport Upon arrival at their target cell, lipoplexes and polyplexes must navigate cellular entry and subcellular trafficking, ideally resulting in the delivery of intact cargo DNA to the nucleus (Figure 12.2B). Gene delivery vehicles typically enter the cell by either nonspecific macropinocytosis or receptor-mediated endocytosis, depending on their surface functionalization and size. Positively

VEHICLE TRANSPORT BARRIERS

213

charged lipoplexes and polyplexes are generally internalized via macropinocytosis as a result of their nonspecific electrostatic interactions with the cell surface, as noted previously. Neutral or negatively charged complexes containing cell-binding ligands typically traffic through the relevant receptor-mediated pathway. The size of a lipoplex or polyplex is an important determinant of its internalization rate, and the optimal size for endocytosis has been investigated for various receptor-mediated pathways. As a general rule, previous experimental and computational studies have suggested that particles with diameters 2000 bp and that of nanoscale particles [63–65]. Microtubule-dependent active transport of PEI polyplexes has been reported for a fraction of the internalized polyplexes (∼17%) [64]. Recently, anionic plasmid carriers formed from poly(lactic-coglycolic acid) and from 1,2-dioleoyl-sn-glycero-3-phosphocholine liposomes were conjugated to actin comet tails and shown to move through cytoplasmic extracts much more rapidly than their unconjugated counterparts [66]. Polyplex redistribution has also been proposed as a result of the mixing that occurs during mitosis [8]. The final intracellular transport barrier for gene delivery is the nuclear membrane. Transport into the nucleus is mediated by the nuclear pore complex, a multiprotein transmembrane assembly that permits the passage of small molecules but prohibits the free diffusive transport of macromolecules larger than 10–20 kDa [8]. The upper limit for active intranuclear transfer of proteins has been estimated as ∼50 kDa [67, 68], making it nearly prohibitive

214

SELF-UNPACKING GENE DELIVERY SCAFFOLDS

for DNA and nanoscale particles. Thus the majority of gene delivery vehicles are thought to accumulate in the nucleus during cell division [69], a theory that is supported by the findings that transfection of nondividing cells occurs at only very low levels, and that transfection of dividing cells is 30- to 500-fold more efficient when transfection is performed immediately prior to cell division as opposed to at the beginning of the cell cycle [8, 70].

12.4

VEHICLE UNPACKAGING

The final critical step in the gene delivery pathway for lipoplexes and polyplexes is the release of DNA from the complex, given that condensed DNA does not allow efficient access of the transcriptional machinery to the underlying DNA sequences [26]. This self-disassembly, or self-unpackaging, presents an important challenge in materials design, in that the same beneficial characteristics (i.e., a high density of positive charge [71–73]) that promote tight association of a polycation with DNA during gene transport eventually inhibit the activity of the DNA within the nucleus. A change in complex packaging state requires a change in the underlying structure or charge state of the condensing polycation. Unfortunately, there is a lack of obvious intracellular triggers (i.e., ionic strength, temperature, or pH) to stimulate this transition, given that the local pH drop in the endosome does not deprotonate the cationic amine residues that constitute the basis for many condensing agents. Indeed, the stability of various polyplexes to intracellular dissociation has been noted [26, 64, 74]. Inefficient polyplex disassembly has been identified as a critical rate-limiting barrier for gene delivery [26, 64, 74–78]. While unpackaging is a clear requirement for efficient transcription, the desired time and place for DNA release are less clear [71]. For example, premature unpackaging can be as problematic as inefficient unpackaging: interactions between lipoplexes or polyplexes and various extracellular components have been shown to cause destabilization, DNA release, and DNA degradation by serum nucleases. Intracellular components have also been shown to mediate lipo- or polyplex destabilization and DNA release, although in many cases, higher levels of intracellular release are correlated with improved transfection, despite the presence of cytoplasmic nucleases. Thus DNA packaging materials face opposing requirements, as they must protect DNA by tight packaging during transport, and release DNA prior to transcription. Materials designed with mechanisms to initiate the site-specific release of DNA in the vicinity of the cell’s nucleus would be expected to significantly enhance the efficient utilization of administered DNA. It is worth note that despite the development of numerous nonviral methods for achieving high levels of in vitro cell transfection, including those that employ lipoplexes and polyplexes, no method exists that could truly be considered efficient. For example, for a typical in vitro Lipofectamine-mediated transfection (Lipofectamine is a product of Invitrogen Corporation, Carlsbad, CA), millions of copies of

VEHICLE UNPACKAGING

215

plasmid DNA must be administered per cell (this assumes that 10 µg of plasmid DNA would be used to transfect 106 cells). Given that the accumulation of uncomplexed intranuclear DNA appears to very minimal, even when the levels of overall intracellular lipo- or polyplex accumulation are high [26, 64], it is clear that intracellular trafficking and unpackaging remain critical and unsolved barriers with respect to gene delivery. In this section, the parameters determining the innate stability of lipo- and polyplexes will be addressed, and known mechanisms that promote the unpackaging of these complexes in the extra- and intracellular environments will be explored. Subsequently, several classes of new materials that have been designed for the explicit purpose of site-specific unpackaging and DNA release within the cell will be reviewed, with emphasis on identifying the most promising new directions for further development. 12.4.1

Complex Structure and Stability

Both lipids and polymers have demonstrated the capacity to compact and deliver DNA payloads in vitro. Polycation structure plays an important role in the efficiency of DNA binding and condensation. For example, branched PEI has been shown to more efficiently compact DNA then linear PEI [79], perhaps as a result of the higher density of primary amine groups on branched PEI relative to linear PEI. In general, primary amine groups within cationic lipids and polymers have been found to compact DNA more efficiently than higher order amines within corresponding molecules, presumably due to the reduced steric interference around the primary amine groups [80, 81]. The number and density of cationic groups on the polycation have also been shown to affect condensation efficiency. With polypeptide structures, several groups have found that a minimum of six to eight charged residues per polypeptide were necessary for tight polyplex packaging (i.e., packaging at a level that would reduce ethidium bromide staining of the complexed DNA) [26, 82, 83]. Different structural requirements for packaging were identified by these same groups, where Plank and co-workers found that the inclusion of tryptophan residues did not affect DNA binding affinity but reduced DNA compaction [82], and Wadhwa and co-workers found that the inclusion of tryptophan increased DNA binding efficiency [83]. Differences in the structures (branched vs. linear) and charge density of the studied polypeptides could underlie these different findings. Correlations between polycation length and packaging efficiency have also been identified for chitosan-based carriers [84, 85]. Furthermore, other work to systematically address the impact of various characteristics of polycationic polymers on DNA binding and condensation suggested that polymers with a low density of cationic residues were unable to efficiently compact DNA [86]. Random copolymers formed from neutral monomers and monomers containing a single quaternary amine residue bound to DNA inefficiently when the ratio of neutral monomer was 50% or greater [86].

216

SELF-UNPACKING GENE DELIVERY SCAFFOLDS

The functionalization of lipoplexes and polyplexes with stabilizing or targeting groups has also been shown to influence the efficiency of DNA compaction. For example, hydrophilic polymers such as PEG are often grafted to DNA condensing agents via reaction with primary amine groups along the backbone of the polycation [29, 31, 32, 87]. When a polycation is functionalized with PEG or another hydrophilic polymer prior to complexation, the charge density of the agent is reduced and DNA compaction is adversely affected [28, 88–91]. Furthermore, despite the fact that this surface grafting is typically performed for the purpose of reducing polyplex self-interactions and polyplex interactions with serum components [92], efficiently compacted polyplexes grafted with hydrophilic polymers have been found to be less stable to competing ions in blood than their ungrafted counterparts [19, 24, 32, 93, 94]. This effect has been attributed to thermodynamic destabilization caused by hydrophilic chains dispersed within the polyplex [24, 32]. However, even when grafting is performed on surface-exposed free amines postcomplexation, grafted polyplexes are often less stable to competitive ions [19, 23], and PEGylated polyplexes formed from lower molecular weight polycations can be destabilized by the addition of salt [28]. 12.4.2

Extracellular Unpackaging

Depending on the innate stability of a lipo- or polyplex, a variety of components within the extracellular environment can drive premature unpackaging prior to cellular entry. Immediately following introduction into the bloodstream, PEG–PEI polyplexes have been shown to dissociate significantly [32, 93]. Polypeptide–DNA complexes have also been shown to dissociate within minutes upon introduction into the blood, allowing the degradation of DNA by serum endonucleases; when these same peptide-based polyplexes were crosslinked with glutaraldehyde prior to in vivo administration, their in vivo half-life was improved significantly [95]. Serum-induced destabilization may result from the enhanced electrostatic screening of DNA–polycation interactions at physiological salt levels. Alternatively, charged serum components such as albumin may destabilize these polyplexes by competitive DNA or polycation displacement. However, despite the fact that negatively charged model polyanions have been shown to displace DNA from polyplexes, albumin has been shown to bind to N-(2-hydroxypropyl)-methacrylamide/ 2-(trimethylammonio)ethyl methacrylate polyplexes without significant polyplex destabilization and only minor increases in the overall polyplex size [24]. Thus albumin-mediated polyplex clearance has been proposed to result from enhanced interactions between albumin-bound polyplexes and serum opsonins, as opposed to polyplex dissociation [24]. Following vascular extravasation, interactions with ECM proteins can also cause rapid dissociation of lipo- and polyplexes and the subsequent degradation of the DNA. For example, extracellular glycosaminoglycans (GAGs) such as heparin have been shown to cause polyplex destabilization/unpackaging,

VEHICLE UNPACKAGING

217

resulting in the degradation of the freed DNA by extracellular nucleases [19–21, 96, 97]. Burke and Pun demonstrated that both soluble and insoluble ECM GAGs and proteoglycans facilitated polyplex unpackaging in a fashion that correlated with the density of negatively charged residues, supporting the hypothesis of DNA displacement by competing anions in the ECM [19]. 12.4.3

Intracellular Unpackaging

Interactions between lipo- and polyplexes and intracellular anions have also been proposed to stimulate polyplex unpackaging. For example, several groups have demonstrated the release of plasmid DNA from DNA-condensing lipids and polymers by cytosolic components [98–102]. In the case of lipoplexes, anionic HSPGs on the cell surface have been shown to interact with polycationic lipids within the lipoplex to initiate receptor-mediated endocytosis [103–106]. These interactions have been proposed to destabilize the endosomal membrane, cause the membrane lipids to flip to the cytoplasmic face of the membrane, and displace the DNA from the lipoplex into the cytoplasm [100, 102]. Polyplexes have been proposed to unpackage in the cytoplasm as a result of displacing interactions with other cytoplasmic polyanions. For example, Okuda et al. [99] demonstrated that the cytoplasmic component causing the release of plasmid DNA from dendritic poly(l-lysine) (PLL) and jetPEI was protease sensitive, as proteinase K-treated cytosolic fractions were inactive at DNA release. This effect was only observed at low PEI : DNA ratios, prompting speculation that other factors such as cytoplasmic RNA might be involved in cytosolic DNA displacement at high polycation : DNA ratios [101]. Huth et al. [101] demonstrated that the release of plasmid DNA from linear and branched PEI and PLL in cytoplasmic extracts was RNA-dependent, as RNase-treated cytoplasmic fractions did not release DNA. Furthermore, these authors demonstrated that in the absence of cytoplasmic extracts, the addition of RNA at levels similar to intracellular levels of RNA also resulted in DNA release from polyplexes formed at high polycation : DNA ratios. Chromosomal DNA has also been suggested to play a role in the displacement and release of DNA from lipo- and polyplexes [26, 107]. For example, fluorescence imaging analyses with PLL–DNA polyplexes demonstrated that a fraction of the polyplexes reached the nucleus, and that the amount of intranuclear polyplex dissociation was correlated with the length of the PLL chain [26]. Whereas long PLL chains (180 residues) remained closely associated with plasmid DNA, shorter PLL chains (19 and 36 residues) were found to be completely separated from the plasmid within the nucleus. In subsequent experiments, Schaffer et al. [26] demonstrated that double-stranded DNA could more easily dissociate the shorter PLLs from the plasmid than the longer PLL, a result consistent with cation exchange to the surrounding chromatin. Other investigators have identified a more minor role for chromosomal DNA in unpackaging. For example, Zabner et al. [103] demonstrated that lipoplexes

218

SELF-UNPACKING GENE DELIVERY SCAFFOLDS

injected into the nucleus showed low transfection efficiency, implying that chromosomal DNA plays only a small role in unpackaging [102]. Together, these experiments suggest that polycation structure and the innate stability of the lipo- or polyplex are important determinants of the site, mechanism, and extent of unpackaging, and support roles for multiple components in the unpackaging process. Although lipo- and polyplex materials capable of intracellular DNA release often have high in vitro gene transfer activities in comparison with their more stably packaged counterparts [26, 58, 84, 98, 108–112], cytoplasmic release has also been suggested to correlate with lower gene transfer activity [113]. For example, Oupicky et al. [113] demonstrated that cytoplasmic microinjection of free DNA resulted in lower levels of transfection than cytoplasmic microinjection of polyplexes, whereas nuclear microinjection of the same samples had the opposite effect. These findings were attributed in part to the enhanced resistance of the polyplex-contained DNA to cytoplasmic nucleases, although the authors also noted that the polyplexes might have improved cytoplasmic mobility and transport to the nucleus in comparison with the free DNA. In general, it is probable that cytoplasmic degradation of at least a portion of the released DNA occurs even with materials that exhibit high in vitro transfection efficiencies. Presumably, only a portion of the released DNA is sacrificed, and the remaining intact free DNA is utilized relatively efficiently.

12.5

SELF-UNPACKAGING MATERIALS

Several strategies have been proposed with respect to the design of materials capable of packaging and releasing DNA for efficient in vivo gene delivery. For example, the utilization of materials with an intermediate level of packaging has been suggested; ideally, these materials would be sensitive to the existing intracellular release mechanisms but still stable in serum and during contact with the ECM. However, given the similar structures and charge densities of the extracellular and intracellular GAGs and other anions that have been shown to initiate complex unpackaging, it would seem difficult to strike such a balance. Thus a variety of materials containing active DNA-releasing strategies have recently been explored. These include materials that release DNA by carrier degradation, carrier charge reduction, and crosslinker degradation (Figure 12.3). These major categories of active materials will be explored in the following subsections. 12.5.1 Release by Carrier Degradation Given that a correlation between the length of a polycationic carrier and its tendency to dissociate from DNA has been identified by several investigators [26, 82–85], as discussed above, one strategy that has been pursued for targeted unpackaging involves the use of high molecular weight degradable polycations

SELF-UNPACKAGING MATERIALS

219

(A)

(B)

(C)

Figure 12.3. Strategies for DNA release by active self-unpackaging. DNA within lipoor polyplexes can be released by hydrolytic or reductive degradation of the cationic carrier (A), charge reversal of the carrier (B), or degradation of surface crosslinkers on the carrier (C). Strategy (C) is often combined with carriers that degrade as in (A).

(Figure 12.3A). These species are typically either hydrolytically [109, 114–120] or reductively [75, 77, 110, 121] degradable, and thus are designed to gradually release DNA as they transition into low molecular weight (and fast-dissociating) species following exposure to aqueous media or reducing environments such as the endosome [108]. For example, Langer and co-workers have synthesized libraries of poly(β-aminoesters) and related biodegradable materials that condense DNA at physiological pH, but are hydrolytically degradable into nontoxic fragments [114, 115, 119, 120]. These investigators identified several candidate polymers with high in vitro gene transfection efficiencies. Hydrolytically degradable PEI derivatives have also been created and shown to have DNA-binding properties similar to high molecular weight (25-kDa) PEI with comparably low cytotoxicity and high gene transfer efficiency [109]. Forrest et al. [109] speculated that enhanced DNA release might be one determinant of the enhanced in vitro transfection efficiency of these materials. Jain and co-workers have synthesized pH-sensitive linear poly(amido amines)

220

SELF-UNPACKING GENE DELIVERY SCAFFOLDS

whose hydrolysis rate is significantly enhanced at pH values correspondent to the lysosomal pH [116]. Because of their capacity to selectively degrade in the intracellular environment, as opposed to in a sustained fashion, these new materials would be expected to have desirable properties for targeted DNA release. With a similar goal of targeted intracellular degradation and DNA release, reductively degradable DNA-packaging materials have also been synthesized. For example, Pichon et al. [77] synthesized high molecular weight PLL polymers for gene delivery by the reductive crosslinking of low molecular weight PLL via disulfide linkages. In a similar strategy, Gosselin et al. [75] created crosslinked PEI polymers from low molecular weight PEI linked by disulfide bonds. These reductively linked polymers had high in vitro gene transfer efficiencies relative to the corresponding uncrosslinked low molecular weight species, and were shown to release DNA upon incubation with reducing agents such as glutathione [75, 77]. More recently, reductively degradable poly(amido amines) and poly(amido ethylenimines) were synthesized and shown to exhibit high levels of in vitro gene transfection in the presence of serum, even when serum incubation times were extended for several hours [110, 121]. These results suggest that the intracellular targeted release mechanism employed by these materials had good potential for in vivo activity. 12.5.2 Release by Carrier Charge Reduction Other strategies for the targeted release of DNA have also been pursued, including the use of polycationic carriers designed to undergo controlled reductions in their net charge when exposed to physiologically relevant environments [71, 108, 111, 112, 122–125] (Figure 12.3B). For example, Hennink and co-workers have created charge-reducing polycations from poly(methacrylates) whose amine-containing side chains are attached to the polymer backbone via hydrolyzable carbonate bonds; thus hydrolytic cleavage of these polymers results in the gradual loss of cationic amine groups [123–125]. These polymers were shown to release DNA in aqueous media in a fashion correlated with the expected time-dependent changes in electrostatic interactions, and mediated high levels of in vitro cell transfection. Others have also created materials whose positive character is reduced by hydrolysis. For example, Lynn and co-workers have created charge-shifting cationic polymers whose hydrolysis results not only in the loss of cationic groups, but also in the introduction of anionic residues [71, 108, 112]. The DNA-releasing capacities of these materials have been demonstrated for surface-mediated [25] and polyplex-mediated [71, 108] delivery. Prata et al. [111] have designed chargereversal amphiphiles that transition from positively to negatively charged within cells. These amphiphiles bind to DNA to form supramolecular lipoplex complexes, but are cleaved by endosomal esterases to become anionic and release DNA. The supramolecular lipoplexes had high in vitro activity that was dependent upon esterase-mediated lipid cleavage.

REFERENCES

12.5.3

221

Release by Crosslinker Degradation

Although the above approaches for targeted release are promising, the lack of evaluation (to date) of these systems in vivo combined with the increasing concerns about the stability of lipo- and polyplexes during prolonged exposure to the extracellular environment have led to recent interest in new methods to reversibly stabilize polyplexes. For example, several approaches have focused on the development of methods to covalently crosslink the surfaces of polyplexes with degradable linkages [76, 126] (Figure 12.3C). Oupicky et al. [76] explored the potential of polyplexes formed from reducible polycations and stabilized by surface crosslinking with hydrophilic polymers. Surface coating was shown to suppress the unpackaging of the polyplexes by poly(acrylic acid)-mediated displacement. The stabilized particles were capable of moderate levels of cellular transfection when cell-binding ligands were conjugated to their surfaces. In another approach, PEI–DNA polyplexes were crosslinked with disulfide bonds [126]. These crosslinked polyplexes were stable against dissociation by polyanions and high ionic strength and had improved biocompatibility against albumin and erythrocyte interactions in comparison with uncrosslinked PEI–DNA complexes, suggesting the potential applicability for systemic application. 12.6

CONCLUSIONS

Gene delivery represents the most significant hurdle to the realization of nucleic acid-based therapies, and despite years of effort directed at the design of new nonviral delivery materials, successes have been limited. Many pervasive problems with these materials are related to the lack of correlation between their in vitro and in vivo performance. With respect to lipo- and polyplex-mediated delivery, the ability to protect DNA within the extracellular environment but release it within the cell has presented a particular challenge. Many new materials are now being developed with promising potential for the targeted intracellular release of DNA. These materials offer possible solutions to a significant problem in gene delivery, and improved understanding in this area may enable the synthesis of nonviral materials with efficacies that rival those of viral delivery systems, without the related concerns of toxicity and immunogenicity. The demonstration of these new releasing materials in vivo will improve understanding of the mechanisms that control gene delivery and will pave the way for their clinical translation. REFERENCES 1. Mulligan, R. C. The basic science of gene therapy. Science 260: 926–932 (1993). 2. Huang, H., Yee, J., Shew, J., Chen, P., Bookstein, R., Friedmann, T., Lee, E., and Lee, W. Suppression of the neoplastic phenotype by replacement of the RB gene in human cancer cells. Science 242: 1563–1566 (1988).

222

SELF-UNPACKING GENE DELIVERY SCAFFOLDS

3. Blaese, R. M., Culver, K. W., Miller, A. D., Carter, C. S., Fleisher, T., Clerici, M., Shearer, G., Chang, L., Chiang, Y., Tolstoshev, P., Greenblatt, J. J., Rosenberg, S. A., Klein, H., Berger, M., Mullen, C. A., Ramsey, W. J., Muul, L., Morgan, R. A., and Anderson, W. F. T lymphocyte-directed gene therapy for ADA–SCID: initial trial results after 4 years. Science 270(5235): 475–480 (1995). 4. Cavazzana-Calvo, M., Hacein-Bey, S., de Saint Basile, G., Gross, F., Yvon, E., Nusbaum, P., Selz, F., Hue, C., Certain, S., Casanova, J.-L., Bousso, P., Le Deist, F., and Fischer, A. Gene therapy of human severe combined immunodeficiency (SCID)-X1 disease. Science 288(5466): 669–672 (2000). 5. Kay, M. A., Manno, C. S., Ragni, M. V., Larson, P. J., Couto, L. B., McClelland, A., Glader, B., Chew, A. J., Tai, S. J., Herzog, R. W., Arruda, V., Johnson, F., Scallan, C., Skarsgard, E., Flake, A. W., and High, K. A. Evidence for gene transfer and expression of factor IX in haemophilia B patients treated with an AAV vector. Nat. Genet. 24: 257–261 (2000). 6. Khuri, F. R., Nemunaitis, J., Ganly, I., Arseneau, J., Tannock, I. F., Romel, L., Gore, M., Ironside, J., MacDougall, R. H., Heise, C., Randlev, B., Gillenwater, A. M., Bruso, P., Kaye, S. B., Hong, W. K., and Kirn, D. H. A controlled trial of intratumoral ONYX-015, a selectively-replicating adenovirus, in combination with displatin and 5-fluorouracil in patients with recurrent head and neck cancer. Nat. Med. 6: 879–885 (2000). 7. Gene Therapy Clinical Trials Worldwide [online]. March 2009. Available at: http:// www.wiley.co.uk/genetherapy/clinical/ (accessed July 2009). 8. Pack, D. W., Hoffman, A. S., Pun, S., and Stayton, P. S. Design and development of polymers for gene delivery. Nat. Rev. Drug Discov. 4(7): 581–593 (2005). 9. Verma, I. M., Naldini, L., Kafri, T., Miyoshi, H., Takahashi, M., Blömer, U., Somia, N., Wang, L., and Gage, F. H. Gene therapy: promises, problems, and prospects. In: V. Boulyjenkov, K. Berg, and Y. Christen, eds. Genes and Resistance to Disease. Springer, New York, 2000. 10. Verma, I. M. and Somia, N. Gene therapy—promises, problems and prospects. Nature 389: 239–242 (1997). 11. Bakeev, K. N., Izumrudov, V. A., Kuchanov, S. I., Zezin, A. B., and Kabanov, V. A. Kinetics and mechanism of interpolyelectrolyte exchange and addition-reactions. Macromolecules 25: 4249–4254 (1992). 12. Manning, G. S. Limiting laws and counterion condensation in polyelectrolyte solutions. I. Colligative properties. J. Chem. Phys. 51: 924–938 (1969). 13. Wagner, E., Cotten, M., Foisner, R., and Birnstiel, M. L. Transferrin–polycation– DNA complexes: the effect of polycations on the structure of the complex and DNA delivery to cells. PNAS 88: 4255–4259 (1991). 14. Hansma, H. G., Golan, R., Hsieh, W., Lollo, C. P., Mullen-Ley, P., and Kwoh, D. DNA condensation for gene therapy as monitored by atomic force microscopy. Nucleic Acids Res. 26(10): 2481–2487 (1998). 15. Lai, E. and van Zanten, J. H. Monitoring DNA/poly-l-lysine polyplex formation with time-resolved multiangle laser light scattering. Biophys. J. 80: 864–873 (2001). 16. Bartlett, D. W. and Davis, M. E. Physicochemical and biological characterization of targeted, nucleic acid-containing nanoparticles. Bioconjug. Chem. 18: 456–468 (2007).

REFERENCES

223

17. Bloomfield, V. A. DNA condensation by multivalent cations. Biopolymers 44: 269–282 (1997). 18. Kabanov, A. V., Astafyeva, I. V., Chikindas, M. L., Rosenblat, G. F., Kiselev, V. I., Severin, E. S., and Kabanov, V. A. DNA interpolyelectrolyte complexes as a tool for efficient cell transformation. Biopolymers 31(12): 1437–1443 (1991). 19. Burke, R. S. and Pun, S. H. Extracellular barriers to in vivo PEI and PEGylated PEI polyplex-mediated gene delivery to the liver. Bioconjug. Chem. 19(3): 693– 704 (2008). 20. Ruponen, M., Ronkko, S., Honkakoski, P., Pelkonen, J., Tammi, M., and Urtti, A. Extracellular glycosaminoglycans modify cellular trafficking of lipoplexes and polyplexes. J. Biol. Chem. 276(36): 33875–33880 (2001). 21. Ruponen, M., Ylä-Herttuala, S., and Urtti, A. Interactions of polymeric and liposomal gene delivery systems with extracellular glycosaminoglycans: physicochemical and transfection studies. Biochim. Biophys. Acta Biomembranes 1415(2): 331–341 (1999). 22. Dash, P., Read, M., Barrett, L., Wolfert, M., and Seymour, L. Factors affecting blood clearance and in vivo distribution of polyelectrolyte complexes for gene delivery. Gene Ther. 6: 643–650 (1999). 23. Oupicky, D., Howard, K. A., Konak, C., Dash, P. R., Ulbrich, K., and Seymour, L. W. Steric stabilization of poly-l-lysine/DNA complexes by the covalent attachment of semitelechelic poly[N-(2-Hydroxypropyl)methacrylamide]. Bioconjug. Chem. 11(4): 492–501 (2000). 24. Oupicky, D., Konak, C., Dash, P. R., Seymour, L. W., and Ulbrich, K. Effect of albumin and polyanion on the structure of DNA complexes with polycation containing hydrophilic nonionic block. Bioconjug. Chem. 10(5): 764–772 (1999). 25. Eliyahu, H., Joseph, A., Schillemans, J. P., Azzam, T., Domb, A. J., and Barenholz, Y. Characterization and in vivo performance of dextran-spermine polyplexes and DOTAP/cholesterol lipoplexes administered locally and systemically. Biomaterials 28: 2339–2349 (2007). 26. Schaffer, D. V., Fidelman, N. A., Dan, N., and Lauffenburger, D. A. Vector unpacking as a potential barrier for receptor-mediated polyplex gene delivery. Biotechnol. Bioeng. 67(5): 598–606 (2000). 27. Nishikawa, M., Takemura, S., Takakura, Y., and Hashida, M. Targeted delivery of plasmid DNA to hepatocytes in vivo: optimization of the pharmacokinetics of plasmid DNA/galactosylated poly(l-lysine) complexes by controlling their physicochemical properties. J. Pharm. Exp. Ther. 287(1): 408–415 (1998). 28. Pun, S. H. and Davis, M. E. Development of a nonviral gene delivery vehicle for systemic application. Bioconjug. Chem. 13: 630–639 (2002). 29. Ogris, M., Brunner, S., Schuller, S., Kircheis, R., and Wagner, E. PEGylated DNA/ transferrin-PEI complexes: reduced interaction with blood components, extended circulation in blood and potential for systemic gene delivery. Gene Ther. 6(4): 595–605 (1999). 30. Toncheva, V., Wolfert, M. A., Dash, P. R., Oupicky, D., Ulbrich, K., Seymour, L. W., and Schacht, E. H. Novel vectors for gene delivery formed by self-assembly of DNA with poly(l-lysine) grafted with hydrophilic polymers. Biochim. Biophys. Acta 1380(3): 354–368 (1998).

224

SELF-UNPACKING GENE DELIVERY SCAFFOLDS

31. Kircheis, R., Blessing, T., Brunner, S., Wightman, L., and Wagner, E. Tumor targeting with surface-shielded ligand–polycation DNA complexes. J. Control. Release 72(1–3): 165–170 (2001). 32. Merdan, T., Kunath, K., Petersen, H., Bakowsky, U., Voigt, K. H., Kopecek, J., and Kissel, T. PEGylation of poly(ethylene imine) affects stability of complexes with plasmid DNA under in vivo conditions in a dose-dependent manner after intravenous injection into mice. Bioconjug. Chem. 16(4): 785–792 (2005). 33. Ogris, M., Steinlein, P., Carotta, S., Brunner, S., and Wagner, E. DNA/ polyethylenimine transfection particles: influence of ligands, polymer size, and PEGylation on internalization and gene expression. AAPS PharmSci. 3(3): E21 (2001). 34. Petersen, H., Fechner, P. M., Martin, A. L., Kunath, K., Stolnik, S., Roberts, C. J., Fischer, D., Davies, M. C., and Kissel, T. Polyethylenimine-graft-poly(ethylene glycol) copolymers: influence of copolymer block structure on DNA complexation and biological activities as gene delivery system. Bioconjug. Chem. 13(4): 845–854 (2002). 35. Illum, L., Davis, S. S., Muller, R. H., Mak, E., and West, P. The organ distribution and circulation time of intravenously injected colloidal carriers sterically stabilized with a block copolymer—Poloxamine 908. Life Sci. 40: 367–374 (1987). 36. Bazile, D., Robert, C., Huve, P., Verrecchia, T., Marlard, M., Frydman, A., Veillard, M., and Spenlehauer, G. Body distribution of fully biodegradable 14C-poly(lactic acid) nanoparticles coated with albumin after parenteral administration to rats. Biomaterials 13: 1093–1102 (1992). 37. Allemann, E., Gurny, R., Doelker, E., Skinner, F. S., and Schutz, H. Distribution, kinetics and elimination of radioactivity after intravenous and intramuscular injection of savoxepine loaded poly(d,l-lactic acid) nanospheres to rats. J. Control. Release 29: 97–104 (1994). 38. Kreuter, J., Tauber, U., and Illi, V. Distribution and elimination of poly(methyl2–14C-methacrylate) nanoparticles radioactivity after injection in rats and mice. J. Pharm. Sci. 68: 1443–1447 (1979). 39. Gerlowski, L. E. and Jain, R. K. Microvascular permeability of normal and neoplastic tissues. Microvasc. Res. 31(3): 288–305 (1986). 40. Yuan, F., Dellian, M., Fukumura, D., Leunig, M., Berk, D. A., Torchilin, V. P., and Jain, R. K. Vascular permeability in a human tumor xenograft: molecular size dependence and cutoff size. Cancer Res. 55(17): 3752–3756 (1995). 41. Yuan, F., Leunig, M., Berk, D. A., and Jain, R. K. Microvascular permeability of albumin, vascular surface area, and vascular volume measured in human adenocarcinoma LS174T using dorsal chamber in SCID mice. Microvasc. Res. 45(3): 269–289 (1993). 42. Yuan, F., Leunig, M., Huang, S. K., Berk, D. A., Papahadjopoulos, D., and Jain, R. K. Microvascular permeability and interstitial penetration of sterically stabilized (stealth) liposomes in a human tumor xenograft. Cancer Res. 54(13): 3352–3356 (1994). 43. Maeda, H. The enhanced permeability and retention (EPR) effect in tumor vasculature: the key role of tumor-selective macromolecular drug targeting. Adv. Enzyme Regul. 41: 189–207 (2001).

REFERENCES

225

44. Goodman, T. T., Ng, C. P., and Pun, S. H. 3-D tissue culture systems for the evaluation and optimization of nanoparticle-based drug carriers. Bioconjug. Chem. 19(10): 1951–1959 (2008). 45. Netti, P. A., Berk, D. A., Swartz, M. A., Grodzinsky, A. J., and Jain, R. K. Role of extracellular matrix assembly in interstitial transport in solid tumors. Cancer Res. 60(9): 2497–2503 (2000). 46. Pluen, A., Boucher, Y., Ramanujan, S., McKee, T. D., Gohongi, T., di Tomaso, E., Brown, E. B., Izumi, Y., Campbell, R. B., Berk, D. A., and Jain, R. K. Role of tumor–host interactions in interstitial diffusion of macromolecules: cranial vs. subcutaneous tumors. PNAS 98(8): 4628–4633 (2001). 47. Goodman, T. T., Olive, P. L., and Pun, S. H. Increased nanoparticle penetration in collagenase-treated multicellular spheroids. Int. J. Nanomed. 2(2): 265–274 (2007). 48. Kuhn, S., Hallahan, D., and Giorgio, T. Characterization of superparamagnetic nanoparticle interactions with extracellular matrix in an in vitro system. Ann. Biomed. Eng. 34(1): 51–58 (2006). 49. Enger, P. Ø., Thorsen, F., Lønning, P. E., Bjerkvig, R., and Hoover, F. Adenoassociated viral vectors penetrate human solid tumor tissue in vivo more effectively than adenoviral vectors. Hum. Gene Ther. 13(9): 1115–1125 (2002). 50. Valentine, M. T., Perlman, Z. E., Gardel, M. L., Shin, J. H., Matsudaira, P., Mitchison, T. J., and Weitz, D. A. Colloid surface chemistry critically affects multiple particle tracking measurements of biomaterials. Biophys. J. 86: 4004–4014 (2004). 51. Saga, T., Neumann, R. D., Heya, T., Sato, J., Kinuya, S., Le, N., Paik, C. H., and Weinstein, J. N. Targeting cancer micrometastases with monoclonal antibodies: a binding-site barrier. PNAS 92: 8999–9003 (1995). 52. Gao, H., Shi, W., and Freund, L. B. Mechanics of receptor-mediated endocytosis. PNAS 102(27): 9469–9474 (2005). 53. Aoyama, Y., Kanamori, T., Nakai, T., Sasaki, T., Horiuchi, S., Sando, S., and Niidome, T. Artificial viruses and their application to gene delivery. Size-controlled gene coating with glycocluster nanoparticles. J. A. Chem. Soc. 125(12): 3455–3457 (2003). 54. Nakai, T., Kanamori, T., Sando, S., and Aoyama, Y. Remarkably size-regulated cell invasion by artificial viruses. Saccharide-dependent self-aggregation of glycoviruses and its consequences in glycoviral gene delivery. J. Am. Chem. Soc. 125(28): 8465–8475 (2003). 55. Osaki, F., Kanamori, T., Sando, S., Sera, T., and Aoyama, Y. A quantum dot conjugated sugar ball and its cellular uptake. On the size effects of endocytosis in the subviral region. PNAS 126(21): 6520–6521 (2004). 56. Desai, M. P., Labhasetwar, V., Walter, E., Levy, R. J., and Amidon, G. L. The mechanism of uptake of biodegradable microparticles in Caco-2 cells is size dependent. Pharm. Res. 14(11): 1568–1573 (1997). 57. Prabha, S., Zhou, W.-Z., Panyam, J., and Labhasetwar, V. Size-dependency of nanoparticle-mediated gene transfection: studies with fractionated nanoparticles. Int. J. Pharm. 244(1–2): 105–115 (2002). 58. Boussif, O., Lezoualc’h, F., Zanta, M. A., Mergny, M. D., Scherman, D., Demeneix, B., and Behr, J.-P. A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: polyethylenimine. PNAS 92(16): 7297–7301 (1995).

226

SELF-UNPACKING GENE DELIVERY SCAFFOLDS

59. Akinc, A., Thomas, M., Klibanov, A. M., and Langer, R. Exploring polyethylenimine-mediated DNA transfection and the proton sponge hypothesis. J. Gene Med. 7(5): 657–663 (2005). 60. Behr, J. P. The proton sponge: a trick to enter cells the viruses did not exploit. Chimia 51: 34–36 (1997). 61. Sonawane, N. D., Szoka, F. C. Jr., and Verkman, A. S. Chloride accumulation and swelling in endosomes enhances DNA transfer by polyamine–DNA polyplexes. J. Biol. Chem. 278(45): 44826–44831 (2003). 62. Wagner, E. Application of membrane-active peptides for nonviral gene delivery. Adv. Drug Deliv. Rev. 38(3): 279–289 (1999). 63. Luby-Phelps, K., Castle, P. E., Taylor, D. L., and Lanni, F. Hindered diffusion of inert tracer particles in the cytoplasm of mouse 3T3 cells. PNAS 84: 4910–4913 (1987). 64. Suh, J., Wirtz, D., and Hanes, J. Efficient active transport of gene nanocarriers to the cell nucleus. PNAS 100: 3878–3882 (2003). 65. Lukacs, G. L., Haggie, P., Seksek, O., Lechardeur, D., Freedman, N., and Verkman, A. S. Size-dependent DNA mobility in cytoplasm and nucleus. J. Biol. Chem. 275(3): 1625–1629 (2000). 66. Ng, C. P., Goodman, T. T., Park, I.-K., and Pun, S. H. Bio-mimetic surface engineering of plasmid-loaded nanoparticles for active intracellular trafficking by actin comet-tail motility. Biomaterials 30(5): 951–958 (2009). 67. Görlich, D. and Kutay, U. Transport between the cell nucleus and the cytoplasm. Annu. Rev. Cell Dev. Biol. 15: 607–660 (1999). 68. Macara, I. G. Transport into and out of the nucleus. Microbiol. Mol. Biol. Rev. 65: 570–594 (2001). 69. Subramanian, A., Ranganathan, P., and Diamond, S. L. Nuclear targeting peptide scaffolds for lipofection of nondividing mammalian cells. Nat. Biotechnol. 17: 873–877 (1999). 70. Brunner, S., Sauer, T., Carotta, S., Cotten, M., Saltik, M., and Wagner, E. Cell cycle dependence of gene transfer by lipoplex, polyplex and recombinant adenovirus. Gene Ther. 7: 401–407 (2000). 71. Liu, X., Yang, J. W., Miller, A. D., Nack, E. A., and Lynn, D. M. Charge-shifting cationic polymers that promote self-assembly and self-disassembly with DNA. Macromolecules 38: 7907–7914 (2005). 72. Kircheis, R., Wightman, L., and Wagner, E. Design and gene delivery activity of modified polyethylenimines. Adv. Drug Deliv. Rev. 53(3): 341–358 (2001). 73. Wagner, E. Strategies to improve DNA polyplexes for in vivo gene transfer: Will “artificial viruses” be the answer? Pharm. Res. 21(1): 8–14 (2004). 74. Godbey, W. T., Wu, K. K., and Mikos, A. G. Tracking the intracellular path of poly(ethylenimine)/DNA complexes for gene delivery. PNAS 96: 5177–5181 (1999). 75. Gosselin, M. A., Guo, W. and Lee, R. J. Efficient gene transfer using reversibly cross-linked low molecular weight polyethylenimine. Bioconjug. Chem. 12(6): 989–994 (2001). 76. Oupicky, D., Parker, A. L., and Seymour, L. W. Laterally stabilized complexes of DNA with linear reducible polycations: strategy for triggered intracellular activation of DNA delivery vectors. J. Am. Chem. Soc. 124(1): 8–9 (2002).

REFERENCES

227

77. Pichon, C., LeCam, E., Guerin, B., Coulaud, D., Delain, E., and Midoux, P. Poly[Lys(AEDTP)]: a cationic polymer that allows dissociation of pDNA/cationic polymer complexes in a reductive medium and enhances polyfection. Bioconjug. Chem. 13(1): 76–82 (2002). 78. Forrest, M. L., Meister, G., Koerber, J., and Pack, D., Partial acetylation of polyethylenimine enhances in vitro gene delivery. Pharm. Res. 21(2): 365–371 (2004). 79. Dunlap, D. D., Maggi, A., Soria, M. R., and Monaco, L. Nanoscopic structure of DNA condensed for gene delivery. Nucleic Acids Res. 25(15): 3095–3101 (1997). 80. Savvaa, M., Aljaberia, A., Feiga, J., and Stolzb, D. B. Correlation of the physicochemical properties of symmetric 1,3-dialkoylamidopropane-based cationic lipids containing single primary and tertiary amine polar head groups with in vitro transfection activity. Colloids Surf. B Biointerfaces 43(1): 43–56 (2005). 81. Wu, D. C., Liu, Y., Jiang, X., He, C. B., Goh, S. H., and Leong, K. W. Hyperbranched poly(amino ester)s with different terminal amine groups for DNA delivery. Biomacromolecules 7(6): 1879–1883 (2006). 82. Plank, C., Tang, M. X., Wolfe, A. R., and Szoka, F. C. Branched cationic peptides for gene delivery: role of type and number of cationic residues in formation and in vitro activity of DNA polyplexes. Hum. Gene Ther. 10(2): 319–332 (1999). 83. Wadhwa, M. S., Collard, W. T., Adami, R. C., McKenzie, D. L., and Rice, K. G. Peptide-mediated gene delivery: influence of peptide structure on gene expression. Bioconjug. Chem. 8(1): 81–88 (1997). 84. Koping-Hoggard, M., Varum, K. M., Issa, M., Danielsen, S., Christensen, B. E., Stokke, B. T., and Artursson, P. Improved chitosan-mediated gene delivery based on easily dissociated chitosan polyplexes of highly defined chitosan oligomers. Gene Ther. 11(19): 1441–1452 (2004). 85. MacLaughlin, F. C., Mumper, R. J., Wang, J., Tagliaferri, J. M., Gill, I., Hinchcliffe, M., and Rolland, A. P. Chitosan and depolymerized chitosan oligomers as condensing carriers for in vivo plasmid delivery. J. Control. Release 56(1–3): 259–272 (1998). 86. Wolfert, M. A., Dash, P. R., Nazarova, O., Oupicky, D., Seymour, L. W., Smart, S., Strohalm, J., and Ulbrich, K. Polyelectrolyte vectors for gene delivery: influence of cationic polymer on biophysical properties of complexes formed with DNA. Bioconjug. Chem. 10(6): 993–1004 (1999). 87. Mishra, S., Webster, P., and Davis, M. E. PEGylation significantly affects cellular uptake and intracellular trafficking of non-viral gene delivery particles. Eur. J. Cell Biol. 83(3): 97–111 (2004). 88. Garrett, S. W., Davies, O. R., Milroy, D. A., Wood, P. J., Pouton, C. W., and Threadgill, M. D. Synthesis and characterisation of polyamine-poly(ethylene glycol) constructs for DNA binding and gene delivery. Bioorg. Med. Chem. 8(7): 1779–1797 (2000). 89. Kaminskas, L. M., Karellas, P., Krippner, G. Y., Lessene, R., Kelly, B., and Porter, C. J. The impact of molecular weight and PEG chain length on the systemic pharmacokinetics of PEGylated poly l-lysine dendrimers. Mol. Pharm. 5: 449–463 (2008). 90. Rimann, M., Luhmann, T., Textor, M., Guerino, B., Ogier, J., and Hall, H. Characterization of PLL-g-PEG-DNA nanoparticles for the delivery of therapeutic DNA. Bioconjug. Chem. 19: 548–557 (2008).

228

SELF-UNPACKING GENE DELIVERY SCAFFOLDS

91. Choi, Y., Liu, F., Kim, J.-S., Choi, Y., Park, J., and Kim, S. Poly(ethylene glycol)grafted poly-l-lysine as polymeric gene carrier. J. Control. Release 54: 39–48 (1998). 92. Moghimi, S. M. Mechanisms regulating body distribution of nanospheres conditioned with pluronic and tetronic block co-polymers. Adv. Drug Deliv. Rev. 16 (2–3): 183–193 (1995). 93. Fischer, D., Osburg, B., Petersen, H., Kissel, T., and Bickel, U. Effect of poly(ethylene imine) molecular weight and pegylation on organ distrubition and pharmacokinetics of polyplexes with oligodeoxynucleotides in mice. Drug Metab. Dispos. 32: 983–992 (2002). 94. Mullen, P. M., Lollo, C. P., Phan, Q.-C., Amini, A., Banaszczyk, M. G., Fabrycki, J. M., Wu, D., Carlo, A. T., Pezzoli, P., Coffin, C. C., and Carlo, D. J. Strength of conjugate binding to plasmid DNA affects degradation rate and expression level in vivo. Biochim. Biophys. Acta Gen. Subjects 1523(1): 103–110 (2000). 95. Collard, W. T., Yang, Y., Kwok, K. Y., Park, Y., and Rice, K. G. Biodistribution, metabolism, and in vivo gene expression of low molecular weight glycopeptide polyethylene glycol peptide DNA co-condensates. J. Pharm. Sci. 89(4): 499–512 (2000). 96. Moret, I., Esteban Peris, J., Guillem, V. M., Benet, M., Revert, F., Dasí, F., Crespo, A., and Aliño, S. F. Stability of PEI-DNA and DOTAP-DNA complexes: effect of alkaline pH, heparin and serum. J. Control. Release 76(1–2): 169–181 (2001). 97. Ruponen, M., Honkakoski, P., Rönkkö, S., Pelkonen, J., Tammi, M., and Urtti, A. Extracellular and intracellular barriers in non-viral gene delivery. J. Control. Release 93(2): 213–217 (2003). 98. Itaka, K., Harada, A., Yamasaki, Y., Nakamura, K., Kawaguchi, H., and Kataoka, K. In situ single cell observation by fluorescence resonance energy transfer reveals fast intra-cytoplasmic delivery and easy release of plasmid DNA complexed with linear polyethylenimine. J. Gene Med. 6(1): 76–84 (2004). 99. Okuda, T., Niidome, T., and Aoyagi, H. Cytosolic soluble proteins induce DNA release from DNA-gene carrier complexes. J. Control. Release 98(2): 325–332 (2004). 100. Bhattacharya, S. and Mandal, S. S. Evidence of interlipidic ion-pairing in anioninduced DNA release from cationic amphiphile; DNA complexes. Mechanistic implications in transfection. Biochemistry 37(21): 7764–7777 (1998). 101. Huth, S., Hoffmann, F., Gersdorff, K.V., Laner, A., Reinhardt, D., Rosenecker, J., and Rudolph, C. Interaction of polyamine gene vectors with RNA leads to the dissociation of plasmid DNA-carrier complexes. J. Gene Med. 8(12): 1416–1424 (2006). 102. Xu, Y. and Szoka, F. C. Mechanism of DNA release from cationic liposome/DNA complexes used in cell transfection. Biochemistry 35(18): 5616–5623 (1996). 103. Zabner, J., Fasbender, A. J., Moninger, T., Poellinger, K. A., and Welsh, M. J. Cellular and molecular barriers to gene transfer by a cationic lipid. J. Biol. Chem. 270(32): 18997–19007 (1995). 104. Zhou, X. and Huang, L. DNA transfection mediated by cationic liposomes containing lipopolylysine: characterization and mechanism of action. Biochim. Biophys. Acta Biomembranes 1189(2): 195–203 (1994).

REFERENCES

229

105. Legendre, J.-Y. and Szoka, F. C. Jr. Delivery of plasmid DNA into mammalian cell lines using pH-sensitive liposomes: comparison with cationic liposomes. Pharm. Res. 9(10): 1235–1242 (1992). 106. Felgner, P. L., Gadek, T. R., Holm, M., Roman, R., Chan, H. W., Wenz, M., Northrop, J. P., Ringold, G. M., and Danielsen, M. Lipofection: a highly efficient, lipidmediated DNA-transfection procedure. PNAS 84: 7413–7417 (1987). 107. Remy, J. S., Kichler, A., Mordvinov, V., Schuber, F., and Behr, J. P. Targeted gene transfer into hepatoma cells with lipopolyamine-condensed DNA particles presenting galactose ligands: a stage toward artificial viruses. PNAS 92: 1744–1748 (1995). 108. Liu, X., Yang, J. W., and Lynn, D. M. Addition of “charge-shifting” side chains to linear poly(ethyleneimine) enhances cell transfection efficiency. Biomacromolecules 9: 2063–2071 (2008). 109. Forrest, M. L., Koerber, J. T., and Pack, D. W. A degradable polyethylenimine derivative with low toxicity for highly efficient gene delivery. Bioconjug. Chem. 14(5): 934–940 (2003). 110. Lin, C., Zhong, Z., Lok, M. C., Jiang, X., Hennink, W. E., Feijen, J., and Engbersen, J. F. J. Novel bioreducible poly(amido amine)s for highly efficient gene delivery. Bioconjug. Chem. 18(1): 138–145 (2007). 111. Prata, C. A. H., Zhao, Y., Barthelemy, P., Li, Y., Luo, D., McIntosh, T. J., Lee, S. J., and Grinstaff, M. W. Charge-reversal amphiphiles for gene delivery. J. Am. Chem. Soc. 126(39): 12196–12197 (2004). 112. Zhang, J. and Lynn, D. M. Ultrathin multilayered films assembled from chargeshifting cationic polymers: extended, long-term release of plasmid DNA from surfaces. Adv. Mater. 19(23): 4218–4223 (2007). 113. Oupicky, D., Konák, C., Ulbrich, K., Wolfert, M. A., and Seymour, L. W. DNA delivery systems based on complexes of DNA with synthetic polycations and their copolymers. J. Control. Release 65(1–2): 149–171 (2000). 114. Akinc, A., Lynn, D. M., Anderson, D. G., and Langer, R. Parallel synthesis and biophysical characterization of a degradable polymer library for gene delivery. J. Am. Chem. Soc. 125(18): 5316–5323 (2003). 115. Anderson, D. G., Lynn, D. M., and Langer, R. Semi-automated synthesis and screening of a large library of degradable cationic polymers for gene delivery. Angew. Chem. Int. Ed. 42(27): 3153–3158 (2003). 116. Jain, R., Standley, S. M., and Frechet, J. M. J. Synthesis and degradation of pHsensitive linear poly(amidoamine)s. Macromolecules 40(3): 452–457 (2007). 117. Lim, Y.-B., Kim, S.-M., Lee, Y., Lee, W.-K., Yang, T.-G., Lee, M.-J., Suh, H., and Park, J.-S. Cationic hyperbranched poly(amino ester): a novel class of DNA condensing molecule with cationic surface, biodegradable three-dimensional structure, and tertiary amine groups in the interior. J. Am. Chem. Soc. 123(10): 2460–2461 (2001). 118. Lim, Y.-B., Kim, S.-M., Suh, H., and Park, J.-S. Biodegradable, endosome disruptive, and cationic network-type polymer as a highly efficient and nontoxic gene delivery carrier. Bioconjug. Chem. 13(5): 952–957 (2002). 119. Lynn, D. M., Anderson, D. G., Putnam, D., and Langer, R. Accelerated discovery of synthetic transfection vectors: parallel synthesis and screening of a degradable polymer library. J. Am Chem. Soc. 123(33): 8155–8156 (2001).

230

SELF-UNPACKING GENE DELIVERY SCAFFOLDS

120. Lynn, D. M. and Langer, R. Degradable poly(beta-amino esters): synthesis, characterization, and self-assembly with plasmid DNA. J. Am. Chem. Soc. 122(44): 10761–10768 (2000). 121. Christensen, L. V., Chang, C.-W., Kim, W. J., Kim, S. W., Zhong, Z., Lin, C., Engbersen, J. F. J., and Feijen, J. Reducible poly(amido ethylenimine)s designed for triggered intracellular gene delivery. Bioconjug. Chem. 17(5): 1233–1240 (2006). 122. Shim, M. S. and Kwon, Y. J. Controlled delivery of plasmid DNA and siRNA to intracellular targets using ketalized polyethylenimine. Biomacromolecules 9(2): 444–455 (2008). 123. Funhoff, A., van Nostrum, C., Janssen, A., Fens, M., Crommelin, D., and Hennink, W. Polymer side-chain degradation as a tool to control the destabilization of polyplexes. Pharm. Res. 21(1): 170–176 (2004). 124. Jiang, X., Lok, M. C., and Hennink, W. E. Degradable-brushed pHEMApDMAEMA synthesized via ATRP and click chemistry for gene delivery. Bioconjug. Chem. 18(6): 2077–2084 (2007). 125. Luten, J., Akeroyd, N., Funhoff, A., Lok, M. C., Talsma, H., and Hennink, W. E. Methacrylamide polymers with hydrolysis-sensitive cationic side groups as degradable gene carriers. Bioconjug. Chem. 17(4): 1077–1084 (2006). 126. Neu, M., Sitterberg, J., Bakowsky, U., and Kissel, T. Stabilized nanocarriers for plasmids based upon cross-linked poly(ethylene imine). Biomacromolecules 7(12): 3428–3438 (2006).

CHAPTER 13

Cellular Trafficking of Dendrimers YUNUS EMRE KURTOGLU and RANGARAMANUJAM M. KANNAN Department of Chemical Engineering and Materials Science, NICHD Perinatology Research Branch, Wayne State University, Detroit, Michigan

13.1

DENDRITIC ARCHITECTURE

Dendrimers are polymeric molecules made of multiple monomers branching radially from a central core. The monomers used determine the dendrimer subfamilies, while the number of branching points between the core and the surface of dendrimers determine their generation number. The monomer used for the synthesis, generation number, and surface groups determines the chemical, physical, and biological characteristics. For biological applications various dendrimer structures have been synthesized, which include polyamidoamines [1], poly(aryl ethers) [2], polyamines [3], polyesters [4, 5], nucleic acids [6, 7], polypeptides [8], and carbohydrates [9]. These dendrimers have shown great promise in diverse biological applications such as gene and drug delivery, magnetic resonance imaging (MRI), antivirals, and antibacterials, as well as tissue scaffolds [10]. In gene and drug delivery applications, usually the aim is a combination of drug solubility enhancement, tissue targeting, increasing blood circulation time, reduction in drug/gene metabolism rate, or overcoming biological barriers. Gene delivery strategies often involve complexation of nucleic acids to positively charged dendrimers, whereas most small drug delivery approaches involve covalent attachment of active molecules to dendrimer surface groups. Alternatively, encapsulation of drug molecules to the interior space of a dendrimer has also been investigated. In general, dendrimers have lower polydispersity indices compared to linear polymers due to their stepwise synthesis schemes and highly symmetrical branching patterns. The low polydispersity index is reflected by welldefined number of surface functional groups as compared to hyperbranched Organelle-Specific Pharmaceutical Nanotechnology, Edited by Volkmar Weissig and Gerard G. M. D’Souza Copyright © 2010 John Wiley & Sons, Inc.

231

232

CELLULAR TRAFFICKING OF DENDRIMERS

and linear polymers synthesized by statistically controlled polymerization reactions. As a result of the branching, higher generation dendrimers have more surface end groups and higher molecular weights. On the other hand, the surface groups get more densely packed as the generation numbers go up, which may increase the charge density on the surface depending on the end groups. Most dendrimers are not synthesized over generation 10 (G10), since the surface crowding does not leave any room for further branching and typical sizes for dendrimers vary between 2 and 10 nm. The lower generation dendrimers (up to G4) are generally ellipsoidal, whereas higher generations are more spherical in shape. Well-defined globular structure and the multivalency at their surfaces give dendritic structures some advantages over linear and hyperbranched polymers for drug delivery applications. While the well-defined structure can provide more uniform biodistribution and pharmacokinetics, the functional surface groups enable the attachment of multiple drugs, targeting ligands, and imaging agents. Due to their globular structures and surface charge densities the biological characteristics of dendrimers differ from linear polymers. This macromolecular architecture of dendrimers creates opportunities for cellular targeting and fine-tuning their uptake mechanisms and rates. The specific aim of this chapter is to discuss how the surface properties of these threedimensional (3D) globular polymers govern their cellular transport characteristics. Polyamidoamine (PAMAM) dendrimers have been most widely investigated among the other dendritic molecules perhaps due to their commercial availability and the various sizes and surface functional groups readily available. Therefore, in order to understand how the globular 3D structure and surface functionality affects the cellular uptake of dendritic structures, it is imperative to discuss how various types of PAMAM dendrimers and other dendrimers behave in a comparative manner. Mechanism and rate of cellular uptake on particular cell types and their fate after internalization are important considerations for drug delivery applications and these properties are closely related to surface characteristics. Cellular uptake of dendrimers is mainly affected by their size, the type and charge of surface groups, and the attachment of drugs or other moieties to their functional groups on the periphery. The mechanisms for transport of molecules through biological barriers include endocytosis, passive diffusion, and carrier-mediated and paracellular transport. Like other polymeric macromolecules, endocytotic mechanisms are responsible for cellular uptake of most dendrimers even though smaller and hydrophobic dendrimers may be able to enter the cells by diffusion through the cellular membrane [11]. Endocytosis is an active cellular uptake mechanism to internalize small and large macromolecules and particles. Two types of endocytotic processes have been defined: phagocytosis and pinocytosis. Phagocytosis involves the uptake of large particles and is a characteristic mechanism of white blood cells, whereas pinocytosis follows clathrin-coated pits, caveolae, or clathrin- and

DENDRITIC VERSUS LINEAR AND HYPERBRANCHED POLYMERS

233

caveolae-independent pathways. The pathway by which a macromolecule is taken up can vary with the cell type and the proteins involved in the intracellular vesicles and size of these vesicles, as well as the intracellular fate of the macromolecule. Typically, after internalization polymers are transported to early endosomes followed by late endosomes and later in catalytic environment of lysosomes. The cellular uptake rates and residence times in intracellular compartments are important for designing dendrimer-based therapeutics since they often rely on release of covalently linked drugs intracellularly.

13.2 DENDRITIC VERSUS LINEAR AND HYPERBRANCHED POLYMERS The rate and mechanism of cellular uptake of G2, G3, and G4 amine terminated (cationic) PAMAM dendrimers were investigated in comparison to linear and branched polyethylene imine (PEI) polymers, which are also cationic with amine functional groups [12]. The polymers were labeled by Oregon green (through conjugation), and their cell membrane binding, endocytosis, and exocytosis characteristics were studied on B16F10 melanoma cells. While the cellular uptake mechanism of all the polymers was determined to be endocytosis, the hyperbranched PEI and the PAMAM dendrimers had cholesteroldependent pathways, whereas linear PEI uptake was cholesterol and clathrin independent. The rate of uptake was the highest for G4 PAMAM followed by branched PEI, linear PEI, and G3 and G2 PAMAM. The globular structure and the cationic surface charge density of G4 PAMAM is believed to enhance the cellular uptake by adsorptive endocytosis compared to lower generation dendrimers and PEI polymers. Particularly, the surface functional groups of the dendrimer can produce a highly localized charge density, which can have a significant influence on its interactions with the negatively charged proteoglycans in the cell membrane. All the polymers showed significant membrane binding, which suggests the cationic charge may give rise to nonspecific absorption. This type of adsorptive endocytosis usually follows curvilinear cell uptake kinetics due to saturation of the membrane binding sites. Cellular entry dynamics of amine-terminated G4 and G3 PAMAM dendrimers and hydroxyl-terminated PAMAM dendrimers were studied with human lung epithelial carcinoma cells (A549) in comparison to hyperbranched polyol polymer (Figure 13.1) [13]. Additionally, the amine-terminated G3 PAMAM dendrimer surface was modified with poly(ethylene glycol) PEG to investigate the PEG attachment effects on cellular uptake kinetics. Approximately 90% of amine-terminated PAMAM dendrimers (G3 and G4) entered the cells within 1 hour while the initial rate of cell entry of G4-NH2 was faster compared to that of G3-NH2 (Figure 13.2). In comparison, approximately 55% of G3 PEG and 75% of polyol were inside the cells in 1 hour. G4 PAMAM dendrimers were taken up by cells more rapidly when compared to G3 dendrimers with or without PEG modification and polyol. The higher cell

234 H2N H2N H2N

H2N

NH O NH O N

OH

H2N NH2

NH O

NH2 NH NH2 O NH NH2 O NH NH2 N O O NH NH NH2 O NH N NH NH O O N O N NH O O NH NH NH OO N N HN NH O O N

NH H2N O NH NH H2N O N O N NH N HO O N NH H2N N O NH O H2N HO N N O N NO H H NH N NO O H N NHO O NHNH2 N H2N O NH O O O O O H2N NH N O N NH N NH N H N NH N N H N NH O NHO O O NH OO O NH NH2 NH NH NH H2N O NH2 N N O N N O N NH H2N H O N O NH O NH NH O O NH O N H2N NH NH NH O NH H2N NH NH 2 N OO O 2 N NH O NH NH2 NH O NH H2N NH2 NH H2N NH2 H2N

PAMAM Dendrimer

HO O

O

HO OH HO

OH O

O

O

O

O

O

O O

O

OH OH

OH

HO OH HO

O

O

O O

OH OH O O O OH

O O O

HO

O

HO HO

O

O

O O

HO O O HO O HO HO

O

O

O O O

O

O O

O

O O

O

O

O O

O

O

O

O O O

OO O

O O

OO

O

OH HO

OH

O

O

OH OH O

OH OH OH

O O

O HO OH O OH HO HO OH

O

HO

OH O

O

Hyperbranched polyol

Figure 13.1. Schematic representation of the PAMAM dendrimer (left) and the polyol hyperbranched polymer (right). The “imperfect” branching pattern in the hyperbranched polymer can be seen [13].

DENDRITIC VERSUS LINEAR AND HYPERBRANCHED POLYMERS

235

% Change in absorbance

100 80

G3 PEG Polyol G4 OH G4-NH2

60 40 20 0

0

1

2

3 4 TIme (hours)

5

6

7

6

7

% Change in absorbance

100 80 60 G4-OH Polyol G3 PEG G4-NH2

40 20 0

0

1

2

3 4 TIme (hours)

5

Figure 13.2. Effect of end groups on cellular entry profile: (top) supernatant analysis, (bottom) cell lysate analysis [13]. A549 cells were treated separately with each of the dendrimers and hyperbranched polymer was tagged with fluorescent dye. The supernatant was removed at times 0, 30, 60, 120, 240, and 360 min. The amount of tagged dendrimer in the supernatant was estimated by UV/Vis absorbance at 492 nm. The cells were then washed with PBS, trypsinized, and centrifuged to obtain a cell pellet. The cell pellet was then lysed with cell lysate buffer and the amount of tagged dendrimer in the lysate was quantified. An increase in the fluorescent activity in the cell lysate along with a corresponding decrease in activity in the supernatant indicates the intracellular entry of the tagged dendrimer.

entry rate of amine-terminated G4 PAMAM compared to G3 PAMAM was explained by the larger amount of surface charges on the G4 dendrimer. The other polymers studied, namely, hydroxyl-terminated G4 PAMAM, polyol, and G3 PAMAM-PEG, do not carry cationic surface charges and their cellular uptake is associated with nonspecific adsorption to cell membrane. The hydroxyl-terminated G4 PAMAM entered the cells more rapidly compared to other uncharged polymers, suggesting that the globular dendrimer structure is contributing to the enhanced cellular uptake rates. PEG modification of amine-terminated G3 PAMAM reduced its cellular uptake rates due to reduced ionic interaction with the cell membrane and therefore reduced

236

CELLULAR TRAFFICKING OF DENDRIMERS

surface adsorption that would lead to endocytosis. When the noncationic polymers were compared for their cellular uptake rates, the dendritic structure enhances the cellular uptake compared to hyperbranched polyol. These studies must be viewed as somewhat qualitative and relative, since quantification would require more detailed work.

13.3

SURFACE CHARGE

G4 PAMAM dendrimers with 64 (calculated) amine (cationic), carboxyl (anionic), or hydroxyl (neutral) surface groups were studied for their cellular uptake mechanisms on A549 lung epithelial cells, which are known to have a negative surface charge [14]. The three types of PAMAM dendrimers were tagged with fluoroisothiocyanate (FITC) and their cellular uptake was investigated by flow cytometry and fluorescence microscopy analysis. The cellular uptake rates of the three dendrimers were analyzed in comparison to each other (Figure 13.3). Rate of cell uptake was highest for the cationic dendrimer, followed by anionic and neutral dendrimers. In order to identify specific uptake mechanisms, inhibitors of specific endocytotic pathways were used in each group. Regardless of their surface charge, all PAMAM dendrimers studied were endocytosed by energy-dependent, fluid-phase endocytosis while their specific endocytosis uptake mechanisms varied. Anionic dendrimers were partly taken up by caveolae, while the cationic and anionic dendrimers were 100

%Cell entry

80 60 40 COOH OH NH

20 0 0

1

2 Time (h)

3

4

Figure 13.3. A549 cell entry profile of G4-NH2, G4-OH, and G3.5-COOH PAMAM dendrimers [14]. Rate of cell entry is determined as percent of (cell entry at time t/cell entry at 3 h). The cellular uptake was quantified by flow cytometry using fluorescence. A comparison of the time-dependent fluorescence intensity levels indicates that rate of cell uptake was highest for the cationic dendrimer, followed by anionic and neutral dendrimers. The cell uptake for cationic dendrimer plateaus off after 1 hour, unlike with the other two dendrimers where the cell entry increased more or less linearly with the treatment time, as shown in the graph. (See color insert.)

SURFACE CHARGE

237

(A)

(B)

(C)

Figure 13.4. Confocal microscopic images (63×) of A549 lung epithelial cells after treatment with different dendrimers and LysoTracker for 5 min followed by chase for 30 min in serum-free medium devoid of any dendrimers. Images indicate cells treated with (A) FITC-labeled PAMAM-G4-NH2 dendrimer and LysoTracker. (B) FITClabeled PAMAM-G4-OH dendrimer and LysoTracker. (C) FITC-labeled PAMAMG3.5-COOH dendrimer and LysoTracker. The left-hand panel shows the green fluorescence from the dendrimer, the middle panel shows the red fluorescence from the LysoTracker, and the right panel shows the yellow fluorescence due to the colocalization of green dendrimer–FITC and red LysoTracker in the lysosomes. LysoTracker dye preferentially partitions to the lysosome, giving an intense red color in the acidic environment of the lysosomes. From the colocalization studies, it is evident that the dendrimers are in the lysosomes within 30 min of their transport from the cell membrane [14]. (See color insert.)

taken up by a non-clathrin- and non-caveolae-mediated endocytosis mechanism in A549 cells. Furthermore, the surface charge of PAMAM dendrimers seems to play a role in their intracellular trafficking. After cellular uptake, extent of cationic dendrimer that resides in the lysosomes was less than that of anionic and neutral dendrimers, whereas the cationic dendrimer was found in greater extent in peripheral vesicles (Figure 13.4). This finding is

238

CELLULAR TRAFFICKING OF DENDRIMERS

significantly important for dendrimer–drug conjugate therapies that rely on lysosome enzyme activity for detachment of drug from its carrier. The greater lysosomal residence time for the anionic and neutral dendrimer can provide longer release times, whereas cationic dendrimer conjugates may have less time in the lysosome compartment, thus requiring faster releasing covalent linkages. Therefore the surface charge of the dendrimers can be utilized to modulate the cell entry kinetics, mechanisms, and residence time in intracellular compartments.

13.4

SURFACE MODIFICATION

Human lung epithelial carcinoma cells were also used to investigate the efficacy of PAMAM dendrimer conjugates of ibuprofen (Ibu) [15] and methylprednisolone (MP) [16]. The cellular uptake along with their in vitro efficacy were reported. MP was conjugated to carboxyl terminated G2.5 PAMAM and hydroxyl-terminated G4-OH PAMAM dendrimers while the conjugates were also tagged with FITC. The conjugates were rapidly internalized within minutes, while the uptake continued for up to 4 hours. The localization of the conjugates was studied using fluorescence and confocal microscopy. The in vitro efficacy of MP-PAMAM conjugates was studied by inhibition of prostaglandin secretion. The efficacy of the conjugated MP was comparable to free MP treatment, suggesting that the dendritic delivery approach was successful. Even though the dendrimer may enhance the uptake of MP, the drug has to be released from the dendrimer for prostaglandin suppression. The fact that the conjugate shows comparable efficacy at short times (4 hours), when the drug release is expected to be very small, suggests that the delivery is significantly better with the dendrimer. Ibuprofen complexes of G4 PAMAM dendrimers were also studied for their cell entry. Compared to free Ibu the PAMAM complexes had higher cell entry rates, and the complexes were taken up to significant levels within 1 hour. When ibuprofen was attached to hydroxyl-terminated G4 PAMAM dendrimer and the cellular uptake rates were studied, it was determined that the conjugates entered the cells rapidly with ∼30% uptake in 15 minutes. The punctuated distribution of fluorescence in the cytoplasm determined by confocal and fluorescence microscopy images suggested that the cellular uptake was through endocytotic pathways. Intracellular trafficking of G3 PAMAM dendrimers with amine functionality in the human colon adenocarcinoma HT-29 cell line was also studied. PAMAM dendrimer surface was modified with lauroyl chains, propranolol molecules, or both among with FITC tagging [17]. Cellular uptake of unmodified and propranolol-modified G3 PAMAM dendrimers was determined to be by both caveolae-dependent endocytosis and macropinocytosis pathways, whereas when both surface modifications were present internalization of

PERMEABILITY THROUGH BIOLOGICAL BARRIERS

239

dendrimer was caveolae dependent, and possibly clathrin-dependent. Cell entry pathways of lauroyl chain-modified G3 PAMAM were caveolae dependent, clathrin dependent, and macropinocytosis. All the modified G3 PAMAM dendrimers were trafficked to endosomes and lysosomes after internalization. Modification of G3 PAMAM dendrimer with lauroyl moieties increased the cellular uptake rates. Furthermore, lauroyl modification of dendrimer surface reduced the extent of lysosomal accumulation. The impact of PEG modification on cellular uptake kinetics of carboxylterminated PAMAM dendrimers was also studied with Caco-2 cells with focus on transepithelial transport [18–20]. Cellular uptake and transport of G3.5 and G4.5 PAMAM dendrimers with three levels of PEG surface modification were studied. PEG modification of the surface significantly decreased the transepithelial transport for both generations, whereas the cellular uptake showed generation dependency. G3.5 dendrimer cellular uptake was largely reduced by attachment of PEG to the surface in contrast to a strong reduction in uptake for G4.5 PAMAM with PEG modification. The degree of PEG modification did not create a significant difference in cellular uptake of G3.5 dendrimers and regardless of surface groups modified the cellular uptake was decreased. On the other hand, for G4.5 PAMAM dendrimers the lowest extent of PEG surface modification showed the most improvement over the unmodified dendrimer while further PEG modification actually reduced the cellular uptake compared to less surface modification. The conformation of flexible PEG chains on the surface are believed to be determining parameters for the different characteristics for the two generations studied. The reduction in transepithelial permeability was about 65% for both G3.5 and G4.5 PAMAM dendrimers upon PEG modification as compared to permeability of unmodified G3.5 and G4.5 PAMAM dendrimers.

13.5

PERMEABILITY THROUGH BIOLOGICAL BARRIERS

For oral delivery of dendrimer drug conjugates, it is important to evaluate the permeability though epithelial barriers. For determination of absorption and cell entry characteristics human epithelial colorectal adenocarcinoma cells (Caco-2) are often employed [21, 22]. To demonstrate the influence of size and charge on transport of PAMAM dendrimers across Caco-2 cells, extensive studies were carried out using positive, neutral, and negative charged PAMAM dendrimers [23]. Permeability increased with an increase in the number of anionic surface groups, thus generation number in the carboxyl-terminated PAMAM dendrimers also increased. Cationic G2 PAMAM had greater permeability than hydroxyl-terminated G2 and carboxyl-terminated G1.5 and G2.5 dendrimers. Anionic dendrimer G3.5 PAMAM and cationic G4 PAMAM modified with FITC molecules showed the highest transport rates. Increase in the attachment of hydrophobic FITC label increased the permeability and

240

CELLULAR TRAFFICKING OF DENDRIMERS

reduced the toxicity of cationic dendrimers. The overall permeabilities of various PAMAM dendrimers through epithelial cells were determined as G3.5COOH > G2NH2 > G2.5COOH > G1.5COOH > G2OH. The enhanced PAMAM permeability was attributed to opening of tight junctions in cells via paracellular transport across the Caco-2 cells. Furthermore, depending on their surface properties, PAMAM dendrimers are transported across epithelial barriers to a higher extent than some of the conventional linear watersoluble polymers. Epithelial permeability of cationic dendrimers decreases with size and in contrast the permeability of anionic dendrimers increases with their size. The charged dendrimers had greater permeability than neutral dendrimers, which has no net surface charge at physiological pH to enable its interaction with cell monolayers and between the cationic and anionic, the cationic dendrimers exhibit higher permeability. Furthermore, the cationic dendrimers are transported by a combination of paracellular and endocytotic mechanisms. The trafficking of PAMAM dendrimers to endosomal and lysosomal compartments in Caco-2 cells is rapid and mediated by a clathrin-dependent endocytosis mechanism. This was confirmed from the uptake of amineterminated G4 PAMAM using endocytosis inhibitors such as brefeldin A, colchicine, filipin, and sucrose. In the presence of these inhibitors the uptake and permeability of the G4 PAMAM were significantly lowered [24, 25]. The anionic G4 PAMAM dendrimers are taken up by caveolae-mediated endocytosis in A549 lung epithelial cells and the cationic and neutral G4 PAMAM dendrimers are internalized by a non-clathrin-, non-caveolae-mediated mechanism involving electrostatic interactions or other nonspecific fluid-phase endocytosis. These studies revealed that PAMAM dendrimer internalization by endocytosis is largely dependent on the types of cells targeted in addition to the surface charge, molecular weight, and generation considerations. Therefore the differences in cell uptake mechanism of dendrimers may be cell type dependent. This dependency requires that dendritic devices may need to be tailored for each application and for ultimate cell type being targeted for optimal cellular uptake kinetics. In another study, 125I-labeled anionic G1.5 and G2.5 PAMAM dendrimers were transported across everted rat intestinal tissue much faster than other linear polymers such as polyvinylpyrrolidone (PVP), poly(N-vinylpyrollidoneco-maleic anhydride) (NVPMA), and N-(2-hydroxypropyl)methacrylamide (HPMA) copolymers [26]. While 80% of PAMAM radioactivity was transferred directly across to the serosal fluid, 20% of the anionic PAMAM dendrimer remained associated with the tissue. On the other hand, when the studies were carried out with G3 and G4 amine-terminated cationic PAMAM dendrimers, 60% of the dendrimer radioactivity was found in the tissue whereas 40% was transferred into the serosal fluid. This observation clearly demonstrates the increased interactions between the cationic dendrimers and the cell membranes.

INTRACELLULAR TRAFFICKING AND EFFICACY

13.6

SIRNA

241

OLIGONUCLEOTIDE DELIVERY

siRNA therapeutics have generated a lot of interest in drug delivery research. However, their low resistance against enzymatic degradation, poor cellular uptake, and rapid renal and liver clearance have limited the in vivo applications. In order to address these issues, PAMAM dendrimers were investigated as carriers for antisense and siRNA oligonucleotides. Amine-terminated G5 PAMAM dendrimer was complexed with antisense and siRNA oligonucleotides using the cationic charge of the dendrimer surface and the negative charge of the oligonucleotides [27]. Dendrimer–oligonucleotide complexes were further modified by covalent attachment of a cell-penetrating peptide that is cationic in nature. BODIPY fluorescent molecules were attached to dendrimers for visualization, whereas Cy-5 fluorescent molecules were used for oligonucleotide tracking. The efficacies and the cellular uptake of the complexes were studied on NIH 3T3 fibroblast cells transfected with a plasmid containing the human MDR1 gene. G5 PAMAM–oligonucleotide complexes were effectively internalized and accumulated in intracellular vesicles as determined by confocal microscopy and flow cytometry. Conjugation of cellpenetrating peptide to the G5 PAMAM increased the cellular uptake by 25%, whereas this enhanced uptake did not improve the effectiveness of therapy, which was studied by inhibition of P-glycoprotein expression. G5 PAMAM delivery systems, both with and without peptide modification, were moderately effective in intracellular delivery of antisense oligonucleotide. As delivery vehicles, the effectiveness of dendrimers was comparable to Lipofectamine 2000 as a positive control. The confocal microscopy images suggested that both G5 PAMAM with and without peptide modification, complexed with either siRNA or antisense oligonucleotide, had similar intracellular distribution with most of the material associated with intracellular vesicles while a small portion of siRNA and antisense oligonucleotides were found in the cytoplasm and nucleus. Therefore amine-terminated G5 PAMAM was equally effective in carrying the oligonucleotides into the cells, but overall effectiveness varied due to the differences in siRNA and antisense nucleotide characteristics.

13.7 INTRACELLULAR TRAFFICKING AND EFFICACY Efficacy of the dendrimer–drug conjugates is usually dependent on the release of the active molecules from the dendritic carrier. The activity of small drugs are often suppressed or completely blocked in their conjugated form. It is important to design systems that will release their payload intracellularly at predefined rates. Intracellular trafficking characteristics of dendrimer-based delivery systems such as cellular uptake rates, lysosomal residence time, exocytosis rates, and ability to exit lysosomes to reach the cytoplasm determine the overall extent of drug release. The drug release rates of covalently linked

242

CELLULAR TRAFFICKING OF DENDRIMERS

dendrimer–drug conjugates vary with the intracellular conditions, such as availability and activity of the lysosomal enzymes and acidity of the intracellular compartment, along with time spent under these conditions. Therefore the intracellular activation of the drugs being delivered as well as the efficacy of the dendrimer conjugate system will be in direct relation with its intracellular trafficking properties. The effects of dendrimer surface functional groups on the efficacy of the dendrimer–drug conjugates were investigated with PAMAM dendrimer– methotrexate (MTX) conjugates [28]. The efficacy of amine- and carboxylterminated PAMAM–MTX conjugates were evaluated in MTX-sensitive and MTX-resistant human acute lymphoblastoid leukemia (CCRF-CEM) and Chinese hamster ovary (CHO) cell lines. Two amide-bonded PAMAM dendrimer–MTX conjugates were prepared with carboxylic acid-terminated G2.5 dendrimer and the amine groups of the MTX and another between an amineterminated G3 dendrimer and the carboxylic acid group of the MTX. Carboxylic acid-terminated G2.5 dendrimer conjugate showed an increased drug activity compared to free MTX toward both sensitive and resistant cell lines, whereas amine-terminated dendrimer conjugate did not show significant activity on any of the cell lines. The successful enhancement in cellular entry of MTX by carboxyl-terminated dendrimer conjugation was evidenced by 8and 24-fold enhancement in IC50 values toward MTX-resistant CCRF-CEM and CHO cells, respectively, which otherwise had impaired MTX transport mechanisms. The difference in efficacy of these amide-bonded conjugates was associated with the intracellular drug release from the cationic dendrimer versus the anionic dendrimer, due to the differences in lysosomal residence times dictated by the surface functionality. Carboxyl-terminated dendrimer conjugates may have longer lysosomal residence time and longer period of drug release compared to cationic PAMAM conjugate. Dendrimer–drug conjugates that depend on cytoplasmic conditions rather than lysosomal drug release pathways were also designed. An amineterminated G4 PAMAM dendrimer-N-acetyl cysteine (NAC) conjugate that contains a disulfide linkage was synthesized and evaluated for its efficacy on activated BV-2 microglial cells [29, 30]. The drug release pathway of the conjugates were glutathione (GSH) dependent, which is the most abundant thiol species in the cytoplasm functioning as the major reducing agent in biochemical processes. G4 PAMAM–NAC conjugate rapidly internalized by the cells within 15 minutes and the internalization continued for up to 4 hours. The conjugates showed an order of magnitude increase in antioxidant activity compared to free NAC, which suggests that the conjugate was successfully exiting the lysosomes and releasing its payload in the reductive cytoplasmic environment. Since the antioxidant effect of NAC is associated with its thiol group, which is occupied when in conjugated form, the conjugate would have to release the NAC to have efficacy. After the conjugates are taken up by endocytosis, they will reside in the lysosomes for a period of time, where the release of NAC may be relatively slow, because of the lower thiol content

CONCLUSION

243

in lysosomes. As the conjugate escapes the lysosomal compartment, the NAC will be released into the cytoplasm. Therefore the lysosomal residence times of these types of dendrimer conjugates will ultimately affect their efficacy.

13.8

KEY ASPECTS IN DATA INTERPRETATION

Intracellular trafficking of dendrimers is often studied by attachment of some fluorescent molecules or other imaging agents followed by visualization of the uptake by confocal microscopy or quantification via flow cytometry techniques. Attachment of multiple copies of drugs to the functional surface groups of the dendrimers may alter their uptake characteristics. For a given dendrimer type, clearly the intracellular trafficking can be quite different if the dendrimer is conjugated to different drugs in two studies. The labeling of dendrimers with fluorescent molecules followed by in vitro visualization, identification of compartments, and colocalization studies via fluorescent or confocal microscopy techniques have some other limitations. Often, the fluorescent molecules have pH-dependent fluorescence or pH-dependent quenching properties. Concentration-dependent quenching is also observed in such studies. When considering the endocytotic uptake pathway of most dendrimers, such pH dependency can lead to artifacts, since the lysosomal pH is considerable different from that of the cytoplasm. Furthermore, the lysosomal accumulation could lead to a high enough concentration to introduce concentration-dependent quenching. Therefore it is important to evaluate such cellular uptake data with caution. The cellular uptake studies performed with such systems should be used as a guide for better delivery system designs, while efficacy of the ultimate therapeutic version of the dendritic delivery systems (without imaging agent) should be used for evaluating performance. Another consideration in studying cellular uptake and trafficking of dendrimers is the pathway inhibition strategies used for blocking certain pathways. These inhibition mechanisms are usually very complex and it is reported that one inhibitor can upregulate another cellular uptake mechanism under certain conditions. The experimental conditions can alter the uptake pathway of macromolecules completely even with the same cell type. The knowledge from a cell line that is used for the dendrimer uptake studies may not be readily transferable to other cell types due to variability in their membrane and intracellular trafficking characteristics.

13.9

CONCLUSION

Dendritic architecture creates opportunities for designing well-defined nanosized gene and drug delivery devices for tissue, cell, and even intracellular targeting. Due to their unique structure and advantages over linear or

244

CELLULAR TRAFFICKING OF DENDRIMERS

hyperbranched polymers in drug delivery applications, dendrimer research is receiving growing interest. In addition to cellular and intracellular targeting of dendritic devices, new applications are under investigation to overcome biological barriers such as gastrointestinal layers, amniotic membranes, and the blood–brain barrier. In order to create successful dendrimer-based designs, a solid understanding of intracellular trafficking and cellular uptake properties is required. The most important parameters to be optimized will be regarding surface properties of the dendrimers, such as size, overall charge, surface charge and surface charge density, the extent of drug or targeting moiety attachment, hydrophobicity, and hydrogen bonding capacity. Most of these parameters are interrelated with each other in every system; therefore researchers will have to gain strong fundamentals in engineering the surface of these globular macromolecules.

REFERENCES 1. Tomalia, D. A., Naylor, A. M., and Goddard, W. A. Starburst dendrimers: molecular-level control of size, shape, surface chemistry, topology, and flexibility from atoms to macroscopic matter. Angew. Chem. Int. Ed. Engl. 29: 138–175 (1990). 2. Hawker, C. J. and Fréchet, J. M. J. Preparation of polymers with controlled molecular architecture. A new convergent approach to dendritic macromolecules. J. Am. Chem. Soc. 112: 7638–7647 (1990). 3. de Brabander-van den Berg, E. M. M. and Meijer, E. W. Poly(propylene imine) dendrimers: large-scale synthesis by hetereogeneously catalyzed hydrogenations. Angew. Chem. Int. Ed. Engl. 32: 1308–1311 (1993). 4. Ihre, H., Hult, A., and Söderlind, E. Synthesis, characterization, and 1H NMR self-diffusion studies of dendritic aliphatic polyesters based on 2,2bis(hydroxymethyl)propionic acid and 1,1,1-tris(hydroxyphenyl)ethane. J. Am. Chem. Soc. 118: 6388–6395 (1996). 5. Grinstaff, M. W. Biodendrimers: new polymeric biomaterials for tissue engineering. Chemistry 8: 2838–2846 (2002). 6. Nilsen, T. W., Grayzel, J., and Prensky, W. Dendritic nucleic acid structures. J. Theor. Biol. 187: 273–284 (1997). 7. Li, Y., et al. Controlled assembly of dendrimer-like DNA. Nat. Mater. 3: 38–42 (2004). 8. Sadler, K. and Tam, J. P. Peptide dendrimers: applications and synthesis. J. Biotechnol. 90: 195–229 (2002). 9. Turnbull, W. B. and Stoddart, J. F. Design and synthesis of glycodendrimers. J. Biotechnol. 90: 231–255 (2002). 10. Svenson, S. and Tomalia, D. Dendrimers in biomedical applications—reflections on the field. Adv. Drug Deliv. Rev. 57: 2106–2129 (2005). 11. Duncan, R. and Izzo, L. Dendrimer biocompatibility and toxicity. Adv. Drug Deliv. Rev. 57: 2215–2237 (2005).

REFERENCES

245

12. Seib, F. P., Jones, A. T., and Duncan, R. Comparison of the endocytic properties of linear and branched PEIs, and cationic PAMAM dendrimers in B16f10 melanoma cells. J. Control. Release 117: 291–300 (2007). 13. Kolhe, P., et al. Drug complexation, in vitro release and cellular entry of dendrimers and hyperbranched polymers. Int. J. Pharm. 259: 143–160 (2003). 14. Perumal, O. P., et al. The effect of surface functionality on cellular trafficking of dendrimers. Biomaterials 29: 3469–3476 (2008). 15. Kolhe, P., et al. Preparation, cellular transport, and activity of polyamidoaminebased dendritic nanodevices with a high drug payload. Biomaterials 27: 660–669 (2006). 16. Khandare, J., et al. Synthesis, cellular transport, and activity of polyamidoamine dendrimer-methylprednisolone conjugates. Bioconjug. Chem. 16: 330–337 (2005). 17. Saovapakhiran, A., et al. Surface modification of PAMAM dendrimers modulates the mechanism of cellular internalization. Bioconjug. Chem. 20: 693–701 (2009). 18. Sweet, D. M., Kolhatkar, R. B., Ray, A., Swaan, P., and Ghandehari, H. Transepithelial transport of PEGylated anionic poly(amidoamine) dendrimers: implications for oral drug delivery. J. Control. Release 138: 78–85 (2009). 19. Kim, Y., et al. Systematic investigation of polyamidoamine dendrimers surfacemodified with poly(ethylene glycol) for drug delivery applications: synthesis, characterization, and evaluation of cytotoxicity. Bioconjug. Chem. 19: 1660–1672 (2008). 20. Kitchens, K. M., et al. Transport of poly(amidoamine) dendrimers across Caco-2 cell monolayers: influence of size, charge and fluorescent labeling. Pharm. Res. 23: 2818–2826 (2006). 21. Kolhatkar, R. B., et al. Surface acetylation of polyamidoamine (PAMAM) dendrimers decreases cytotoxicity while maintaining membrane permeability. Bioconjug. Chem. 18: 2054–2060 (2007). 22. Pisal, D. S., et al. Permeability of surface-modified polyamidoamine (PAMAM) dendrimers across Caco-2 cell monolayers. Int. J. Pharm. 350: 113–121 (2008). 23. Yang, H., et al. Stealth dendrimers for drug delivery: correlation between PEGylation, cytocompatibility, and drug payload. J. Mater. Sci. Mater. Med. 19: 1991–1997 (2008). 24. Kitchens, K. M., et al. Endocytosis inhibitors prevent poly(amidoamine) dendrimer internalization and permeability across Caco-2 Cells. Mol. Pharm. 5: 364–369 (2008). 25. Kitchens, K. M., et al. Transepithelial and endothelial transport of poly(amidoamine) dendrimers. Adv. Drug Deliv. Rev. 57: 2163–2176 (2005). 26. Malik, N., Wiwattanapatapee, R., Klopsch, R., Lorenz, K., Frey, H., Weener, J. W., Meijer, E. W., Paulus, W., and Duncan, R. Dendrimers: relationship between structure and biocompatibility in vitro, and preliminary studies on the biodistribution of 125I-labelled polyamidoamine dendrimers in vivo. J. Control. Release 65: 133–148 (2000). 27. Kang, H., DeLong, R., Fisher, M. H., and Juliano, R. L. Tat-conjugated PAMAM dendrimers as delivery agents for antisense and siRNA oligonucleotides. Pharm. Res. 2099–2106 (2005).

246

CELLULAR TRAFFICKING OF DENDRIMERS

28. Gurdag, S., et al. Activity of dendrimer–methotrexate conjugates on methotrexatesensitive and -resistant cell lines. Bioconjug. Chem. 17: 275–283 (2006). 29. Navath, R. S., et al. Dendrimer drug conjugates for tailored intracellular drug release based on glutathione level. Bioconjug. Chem. 19: 2446–2455 (2008). 30. Kurtoglu, Y. E., et al. Poly(amidoamine) dendrimer–drug conjugates with disulfide linkages for intracellular drug delivery. Biomaterials 30: 2112–2121 (2009).

CHAPTER 14

Endolysosomolytically Active pHSensitive Polymeric Nanotechnology HAN CHANG KANG and YOU HAN BAE Department of Pharmaceutics and Pharmaceutical Chemistry, University of Utah, Salt Lake City, Utah

14.1

SITE-SPECIFIC NANOTHERAPEUTICS

Interest in developing pharmaceutical nanotechnology has blossomed with the desire to develop effective systems for delivering various therapeutics (e.g., from low molecular weight chemical drugs and imaging agents to high molecular weight peptides, proteins, and genetic materials) to specific sites of interest (e.g., organs, tissues, cells, cytoplasm, mitochondria, perinuclear regions, and nucleus). Unlike traditional dosage forms, site-specific nanotherapeutics are designed to maximize the bioavailability of the delivered therapeutics at the target sites and have shown beneficial therapeutic efficacy in treating diseases with reduced side effects [1]. Delivering nanosized therapeutic carriers to their target sites can be achieved by exploiting differences in anatomy, pathology, or cellular events between target and nontarget sites. First, anatomical specificity of the organs or tissues of interest can passively drive major accumulation of therapeutics. Hydrodynamic injection for liver targeting [2] and the enhanced permeability and retention (EPR) effect for tumor accumulation [3] are known examples of passive organ/tissue targeting. Second, nanosystems can selectively target solid tumors using pathological differences. For example, the local pH of solid tumors is lower than the extracellular pH of normal tissues, creating the opportunity for pH targeted delivery [4]. (Please refer to Lee et al. [4] for more detailed information.) Third, targeting specific cells has been attained mostly by utilizing particular interactions between ligands and cell-specific or

Organelle-Specific Pharmaceutical Nanotechnology, Edited by Volkmar Weissig and Gerard G. M. D’Souza Copyright © 2010 John Wiley & Sons, Inc.

247

248

ENDOLYSOSOMOLYTICALLY ACTIVE PH-SENSITIVE POLYMERIC NANOTECHNOLOGY

overexpressed receptors located on the plasma membrane. These ligand– receptor interactions accelerate the internalization rates of therapeutics and their carriers into the target cells and enhance the intracellular accumulation [5]. However, if the final destination of therapeutics carried by nanocarriers is subcellular compartments (e.g., cytoplasm, mitochondria, perinuclear areas, and nucleus), the therapeutics or their carriers should be able to escape from the endolysosomal pathways in order to achieve maximum therapeutic effects [6].

14.2

DESTABILIZATION OF ENDOLYSOSOMAL COMPARTMENTS

Therapeutics or nanocarriers larger than 1 kDa rarely cross the plasma membrane without the aid of cell-penetrating peptides [7]. These systems use other cellular entry mechanisms such as endocytosis: that is, the cellular membrane invaginates to engulf therapeutics or pharmaceutical nanosystems and forms intracellular membrane-bound vesicles (or endosomes). The therapeuticsloaded vesicles mature from early endosomes (neutral pH to approximately pH 6) to late endosomes (approximately pH 5 to pH 6) as acidification occurs due to cytosolic protons (H+) pumped into the endosomes by vacuolar ATPase-H+ pumps. As endosomes acidify even further, late endosomes merge with lysosomes (approximately pH 4 to pH 5), which contain various lytic enzymes [1, 7]. In lysosomes, the sequestrated therapeutics may be degraded by lytic enzymes, resulting in reduced bioavailability at the target site. To avoid such a scenario, therapeutics-carrying nanosystems should recognize endolysosomal characteristics, which are easily distinguishable from the extracellular and intracellular environment. One unique trait of endolysosomal compartments is pH—endolysosomal pH is generally lower than extracellular pH and cytosolic pH. To target specific pH ranges, materials containing specific functional groups (i.e., amines/imines [8–11], carboxylic acids [12–15], and sulfonamides [16] as shown in Figure 14.1) have often been used because the functional groups enable the transition between protonation/deprotonation states. Disrupting the endolysosomal membrane may be induced by the proton buffering capacity and/or fusogenicity of the “endolysosomolytic materials.” The “proton sponge” effect suggested by Boussif et al. [8] is related to inhibiting or delaying the acidification process of endolysosomes. When materials having protonable groups are entrapped in the endosomes, protons transferred by vacuolar ATPase-H+ pumps are mostly consumed by the materials themselves and do not contribute much toward endosomal maturation/ acidification. With further proton influx into the endosomes, chloride ions (Cl−) also accumulate in the endosomes, leading to osmotic imbalances between the cytoplasm and the endosomal compartments. This imbalance causes continuous water influx, endosomal swelling, and finally leads to endosomal rupture [17–20].

DESTABILIZATION OF ENDOLYSOSOMAL COMPARTMENTS

249

Amine/Imine-containing materials H N

H N

N H

N

N H H N

NH2 O H N CH C O

NH N

N

CH2

NH2 N

NH

NH

HN Poly (L-histidine)

Branched PEI

Carboxylic acid-containing materials C3H7 H2 C C

C2H5 H2 C C COOH

COOH

Poly (2-ethylacrylic acid)

Poly (2-propylacrylic acid)

Sulfonamide-containing materials CH3

S

H2 CH3 C C C=O NH

N

N

N N

OCH3

Poly/oligo Poly/oligo sulfamethizole sulfadimethoxine R=

O=S=O NH R

OCH3

N

N

N

N

Poly/oligo sulfadiazine

CH3

Poly/oligo sulfamerazine

Figure 14.1. Examples of protonable oligomers/polymers that destabilize the endosomal membrane. (Copyright © 2008 Springer Science + Business Media, LLC [1].)

Fusogenic materials cause physical interactions between pH-sensitive membrane destabilizers and endosomal membranes. For example, positively charged polymers can interact with negatively charged phospholipid bilayers in the endosomal membrane, destabilizing the vesicle structure [17]. Polymers having pH-induced hydrophilic-to-hydrophobic transitions (i.e., anionic polyelectrolytes) show similar destabilization mechanisms, which were investigated using lipid bilayers (i.e., liposomes). That is, endosomal acidification increased the hydrophobicity of polymers, which could lead to interactions

250

ENDOLYSOSOMOLYTICALLY ACTIVE PH-SENSITIVE POLYMERIC NANOTECHNOLOGY

with the hydrophobic tail region of the endosomal phospholipid bilayer membrane. This interaction stimulates lateral compression and forms pores in the bilayer, allowing the release of therapeutics into the cytoplasm [21–23].

14.3

ENDOLYSOSOMOLYTIC POLYMERIC NANOCARRIERS

Various materials (e.g., polymers, liposomes, peptides, viral components, and their hybrids) can induce the destabilization of endolysosomal compartments for effectively delivering therapeutics. This destabilization process is mostly triggered by a “pH” stimulus via various physical and chemical transitions of swelling/deswelling, hydrophilicity/hydrophobicity, micellization/demicellization, complexation/decomplexation, ionic/neutral charge, and α-helix/random coils conformational changes [1]. Target pH for endolysosomal rupture can range from extracellular pH to lysosomal pH while traversing the endolysosomal pathway. However, extracellular pH can vary depending on pathology (e.g., pH 7.4 for normal blood; for solid tumors, average pH 6.8 at normoglycemia and average pH 6.4 at hyperglycemia [24]; for brain, pH 7.2 (normal) and pH 6.4 (ischemic) [25]; for heart, pH 7.5 (normal) and approximately pH 6.8 (ischemic) [26]). Thus if certain materials can disrupt the endolysosomal membranes within specific pH ranges, nanocarriers with membrane destabilizers can be designed to treat specific diseases or target certain cells. Although there are various types of endolysosomolytic materials, this chapter will focus primarily on synthetic polymer/oligomer-triggered endolysosomal disruption. A brief summary of recent endolysosomolytic strategies developed by the Bae research group is presented with representative examples of chemical drug delivery nanocarriers (for low molecular weight therapeutics) and gene delivery nanocomplexes (for high molecular weight therapeutics). (For more information regarding other systems, please refer to their extensive reviews [6, 7, 14, 17, 27].) 14.3.1 Endolysosomolytic Polymeric Anticancer Drug Nanocarriers 14.3.1.1 Poly(L-histidine)-Based Micelles Histidine is an attractive amino acid for endosomolysis because its side chain, an imidazole ring, has pKa 6.0 and sharp proton buffering at pH 6.0 [28]. Also, as shown in Figure 14.2A, its homopolymer poly(l-histidine) (polyHis; see Figure 14.1 for its chemical structure) and various copolymers (i.e., polyHis-b-poly(ethylene glycol) (PEG)) show proton buffering over a broad pH range (pH 4 to pH 9) [29]. The apparent pKb values for polyHis5kDa and polyHis3kDa were about 6.5, whereas their PEGylated copolymers (i.e., polyHis5kDa-b-PEG2kDa and polyHis3kDa-b-PEG2kDa) showed a slightly higher apparent pKb of 7.0. These differences in apparent pKb values may result from differences in water content because the hydrophilic PEG block can induce more hydration [29].

ENDOLYSOSOMOLYTIC POLYMERIC NANOCARRIERS

251

Additional evaluation of polyHis for endolysosomolysis ability was performed using hemolysis because the membranes of red blood cells (RBCs) are similar to endolysosomal membranes. As shown in Figure 14.2B, polyHis5kDab-PEG2kDa disrupted 5–20% of total RBC number with decreasing pH from 7.4 to 6.75, whereas its hemolytic activity at pH 6.5 sharply increased to nearly 90%. However, poly(l-lactic acid) (PLLA)3kDa-b-PEG2kDa, which is used for preparing polyHis5kDa-b-PEG2kDa-containing mixed micelles, caused pHindependent membrane rupture of less than 5% in erythrocytes because the polymer does not possess pH-sensitive or endosomolytic groups [30]. To create a carrier for therapeutics (here, mostly hydrophobic chemical drugs), a polyHis block was conjugated to a hydrophilic PEG block. PolyHisb-PEG can maintain its micellar structure at basic pH due to its amphiphilicity, whereas at acidic pH its hydrophilicity can destabilize the micelles. At pH 8, micelles made from polyHis5kDa-b-PEG2kDa were more stable than those prepared from polyHis3kDa-b-PEG2kDa because the critical micelle concentration (CMC) of polyHis5kDa-b-PEG2kDa was much lower than polyHis3kDa-b-PEG2kDa (2.3 µg/mL for polyHis5kDa-b-PEG2kDa vs. 62 µg/mL for polyHis3kDa-b-PEG2kDa). However, the destabilization of polyHis5kDa-b-PEG2kDa-based micelles started at pH 7.4 and the extent of destabilization increased with decreasing pH [29]. This fact indicates that micelles fabricated from polyHis5kDa-b-PEG2kDa only are not suitable for destabilizing endolysosomes. To improve micelle stability at pH 7.4, the hydrophobic copolymer PLLA3kDa-b-PEG2kDa was introduced when preparing polyHis5kDa-b-PEG2kDa mixed micelles [31]. In mixed micelles made from polyHis5kDa-b-PEG2kDa and PLLA3kDa-b-PEG2kDa, the increased hydrophobicity (i.e., increased weight fraction of PLLA3kDa-b-PEG2kDa) lowered the pH that triggered micelle destabilization. PLLA3kDa-b-PEG2kDa (10 wt%) in the mixed micelles slightly improved micelle stability compared to micelles prepared from polyHis5kDa-bPEG2kDa. More PLLA3kDa-b-PEG2kDa (25 wt%) considerably enhanced micelle stability at pH 7.4 and initiated micelle destabilization below pH 7.0. Mixed micelles fabricated from 60 wt% of polyHis5kDa-b-PEG2kDa and 40 wt% of PLLA3kDa-b-PEG2kDa had a slightly lower destabilization pH than micelles containing 25 wt% PLLA3kDa-b-PEG2kDa. These results strongly correlated with the trigger pH for the release of doxorubicin (DOX, an anticancer drug). For the mixed micelles, DOX release could be triggered to release from pH 6.6 to pH 7.2 by modulating the weight fraction of PLLA3kDa-b-PEG2kDa [31]. These findings suggest that mixed micelles containing higher fractions of PLLA3kDa-b-PEG2kDa had weaker pH sensitivity although the micelle destabilization pH remained the same. Following the discovery that the pH at which micelle destabilization is triggered is closely linked to enhanced drug release, Yin et al. [32] investigated whether it was possible to predict pH-triggered micelle destabilization by monitoring pH-dependent changes in micelle size distribution using dynamic light scattering. The mixed micelles (75 wt% of polyHis5kDa-b-PEG2kDa and 25 wt% of PLLA3kDa-b-PEG2kDa) formed a spherical shaped secondary

(a) 14

(b) 120

NaCl polyHis5kDa-b-PEG2kDa polyHis5kDa polyHis3kDa-b-PEG2kDa polyHis3kDa

10

100 Hemolysis (%)

12

pH

8 6 4

80 60 40 20

2 0

polyHis5kDa-b-PEG2kDa PLLA3kDa-b-PEG2kDa

0

200

400

600

800

0 7.4

7

1 N HCl (mL) (c)

6.75

6.5

pH

polyHis-b-PEG PLLA-b-PEG Proton

pH

pH

pH 7.4–7.0

pH 6.8–6.5

(d)

pH 6.0 (e)

7.6 DHPE

7.4

Lysotracker dye

Merged

7.2 pHt

PHIM-f

7.0 6.8 6.6 6.4

PHSM-f 0

10 20 30 40 50 PLLA3kDa-b-PEG2kDa fraction (wt.%) (g) Tumor volume (mm3)

(f) 120

Cell viability (%)

100 80 60 40 20 0 0.0001 0.001 0.01 0.1 1 DOX concentration (µg/ml)

10

3500 3000 2500 2000 1500 1000 500 0 0 3 6 9 12 15 18 21 24 27 30 Days after 1st i. v. injection

ENDOLYSOSOMOLYTIC POLYMERIC NANOCARRIERS

253

Figure 14.2. (A) Acid–base titration curves of polyHis and polyHis-b-PEG. (Reproduced from Ref. 29. Copyright © 2003, with permission from Elsevier B.V.) (B) Hemolytic activity of polyHis5kDa-b-PEG2kDa and PLLA3kDa-b-PEG2kDa. (Copyright © 2008 Springer Science + Business Media, LLC [30].) (C) Schematic illustrations of micelle destabilization. (Reproduced from Ref. 32. Copyright © 2008 with permission from Elsevier B.V.) (D) The effects of PLLA3kDa-b-PEG2kDa on micelle destabilizationtriggering pH (pHt) in mixed micelles made from polyHis5kDa-b-PEG2kDa and PLLA3kDab-PEG2kDa. (Reproduced from Ref. 32. Copyright © 2008 with permission from Elsevier B.V.) (E) In vitro endosomal disruption using fluorescein DHPE-encapsulated PHSM/f and PHIM/f and the LysoTracker® dye against A2780/DOXR MDR cells. (Copyright © 2008 Springer Science + Business Media, LLC [30].) (F) In vitro cytotoxicity of free DOX (䉬), DOX-loaded PHSM/f (䊉), and PHIM/f (䊏) against A2780/DOXR MDR cells after 48-h incubation. (Copyright © 2008 Springer Science + Business Media, LLC [30].) (G) In vivo tumor growth inhibition test of s.c. A2780/DOXR MDR cells-bearing BALB/c nude mice using control (䉬), free DOX (ⵧ), DOX-loaded PHSM/f (䊉), and DOX-loaded PHIM/f (䉱). Three intravenous doses of 10 mg/kg DOX equivalent dose were administered at Days 0, 3, and 6. (Copyright © 2008 Springer Science + Business Media, LLC [30].) (See color insert.)

structure in the range of pH 7.0 to pH 7.4 because individual core–shell micelles with relatively hydrophobic cores associated together. In the pH range 7.0–7.4, the mixed micelles had a narrow and unimodal size distribution. When pH was decreased from 6.8 to 6.5, there was an obvious increase in the size and aggregation number of the micelles caused by the destabilization of the micelle core. Their size distribution was broad and unimodal, but hydrophilic polyHis5kDa-b-PEG2kDa unimers were released from the micelles. Further decreases in pH (reaching pH 6.0) resulted in a bimodal distribution of micelle size because most polyHis5kDa-b-PEG2kDa unimers had dissociated from the micelle core (see Figure 14.2C for an illustration of pH-induced micelle destabilization) [32]. As shown in Figure 14.2D, when different amounts of PLLA3kDab-PEG2kDa were introduced into the mixed micelles, pH-triggered micelle destabilization decreased with increasing weight fraction of PLLA3kDa-bPEG2kDa (e.g., pHt 6.8 for 25 wt% of PLLA3kDa-b-PEG2kDa and pHt 6.5 for 40 wt% of PLLA3kDa-b-PEG2kDa) [32]. With endolysosomal acidification, the protonated polyHis5kDa-b-PEG2kDa can cause the endolysosomes to swell due to proton buffering of the polymer. In addition, according to the destabilization mechanism of the micelles suggested by Yin et al. [32], the polyHis5kDa-b-PEG2kDa unimers released from the micelles can interact with the endolysosomal membrane. The proton buffering and fusogenic characteristics of polyHis5kDa-b-PEG2kDa may disrupt the endolysosomal compartments as shown in Figure 14.2E. When mixed pH-sensitive micelles (PHSM) containing targeting folate (PHSM/f) (i.e., micelle systems prepared from 80% of polyHis5kDa-b-PEG2kDa and 20% of PLLA3kDa-bPEG2kDa-folate) and mixed pH-insensitive micelles (PHIM) with folate

254

ENDOLYSOSOMOLYTICALLY ACTIVE PH-SENSITIVE POLYMERIC NANOTECHNOLOGY

(PHIM/f) (i.e., micelle systems prepared from 80% of PLLA3kDa-b-PEG2kDa and 20% of PLLA3kDa-b-PEG2kDa-folate) were administered to DOX-induced multidrug-resistant (MDR) ovarian carcinoma cell lines (i.e., A2780/ DOXR MDR cells), PHSM/f caused an even distribution of green fluorescent DHPE (N-(fluorescein-5-thiocarbamoyl)-1,2-dihexadecanoyl-sn-glycero3-phosphoethanoamine triethylammonium salt) and red fluorescent LysoTracker® dye in the cytoplasm. However, LysoTracker® and DHPE delivered using PHIM/f were scattered locally in the cytoplasm. This difference in the intracellular localization of DHPE and LysoTracker dye indicated that polyHis5kDa-b-PEG2kDa in the PHSM/f is an important component for the endolysosomal escape of therapeutics. Endolysosomolytic polyHis5kDa-b-PEG2kDa-containing PHSM/f effectively delivered DOX and killed both drug-sensitive tumor cell lines and MDR tumor cell lines [10, 30]. As shown in Figure 14.2F, DOX-loaded PHSM/f effectively killed in the range of tested DOX concentrations without displaying the cytotoxic effects of free DOX and DOX-loaded PHIM/f. It is noteworthy that DOX-loaded PHSM/f killed approximately 60% of A2780/DOXR MDR cells at 1 µg/mL of DOX, whereas free DOX and DOX-loaded PHIM/f killed approximately 10% and 20% of the cells, respectively [30]. For in vivo models of A2780/DOXR MDR cells-bearing BALB/c nude mice, PHSH/f treatment effectively inhibited tumor growth for 1 month, whereas free DOX and PHIM/f treatments did not suppress tumor growth (Figure 14.2G). 14.3.1.2 Poly(L-histidine-co-L-phenylalanine)-Based Micelles When PHSM/f were exposed to acidic extracellular pH (i.e., pH 6.4–6.8 for solid tumors and ischemia), micelles with up to 40 wt% of PLLA3kDa-b-PEG2kDafolate became destabilized in extracellular environments. Thus to destabilize micelles containing endosomolytic materials at pH lower than the destabilization pH of PHSM/f, polyHis was modified with the hydrophobic amino acid phenylalanine (Phe). As summarized in Table 14.1, pH-sensitive copolymers of His and Phe (i.e., poly(His-co-Phe) (PHP)) with various apparent pKb were synthesized. Their molecular weights were approximately 5 kDa and their apparent pKb values ranged from 6.7 to 4.8 as the hydrophobic Phe mole% in the copolymer was increased [33].

TABLE 14.1

Characterization of Poly(His-co-Phe)

Polymer

Phe (mol%) in Polymer

Apparent pKb

Buffering pH Range

10 16 22 27

6.7 6.3 5.7 4.8

8.2–4.8 7.6–4.7 6.5–4.0 5.8–3.2

PHP10 PHP16 PHP22 PHP27

Source: Reproduced with permission from Ref. 33. Copyright © 2008 Wiley-VCH Verlag GmbH & Co.KGaA.

ENDOLYSOSOMOLYTIC POLYMERIC NANOCARRIERS

1200

100

1000

80

800

Size (nm)

120

60 40

600

Intensity (%)

(B)

(A)

Transmittance (%)

255

30 25 20 15 10 5 0 1

100 10000 Size (nm)

400

20

200

0 3

4

5

6

7 8 pH

9 10 11 12

0

8

7.4

6.5 pH

6

5.5

Figure 14.3. (A) pH-dependent transmittance transition of PHSMPhe10 (䉫), PHSMPhe16 (䉱), PHSMPhe22 (ⵧ), and PHSMPhe27 (䊉). (B) pH-dependent particle size of mPHSMPhe16(5%) (black), mPHSMPhe16(10%) (gray), and mPHSMPhe16(20%) (white). The inset shows a bimodal particle size distribution at pH 6.5 using mPHSMPhe16(5%). (Reproduced with permission from Ref. 33. Copyright © 2008 Wiley-VCH Verlag GmbH & Co.KGaA.)

Like polyHis, poly(His-co-Phe) (PHP) was conjugated with PEG for encapsulating hydrophobic drugs. The pH-sensitive micelles (PHSMPhe) prepared from PEGylated PHP (i.e., PHP-b-PEG2kDa) showed transmittance transition at pH 7.0 for PHSMPhe10, at pH 6.5 for PHSMPhe16, at pH 5.8 for PHSMPhe22, and at pH 5.2 for PHSMPhe27 (Figure 14.3A). To obtain finer tuning of the destabilization pH, PLLA3kDa-b-PEG2kDa was mixed with PHP-b-PEG to prepare mixed micelles (mPHSMPhe). In the initial trial to target early endosomal pH values, PHP16-b-PEG2kDa was selected because its micelles were stable at pH 7.0 but became destabilized at pH 6.5. The micelles “mPHSMPhe16” prepared from various weight ratios of PHP16-b-PEG2kDa and PLLA3kDa-bPEG2kDa showed pH-dependent changes in micelle size, and more dramatic size changes were seen with increasing PLLA3kDa-b-PEG2kDa content. As shown in Figure 14.3B, the size of mPHSMPhe16(5%) (i.e., mPHSMPhe16 prepared from 95 wt% PHP16-b-PEG2kDa and 5 wt% of PLLA3kDa-b-PEG2kDa) prepared at pH 8 increased as pH dropped below pH 6.5, and the sizes of mPHSMPhe16(10%) and mPHSMPhe16(20%) increased below pH 6 and sharply below pH 5.5, respectively. At pH 6.5, a bimodal size distribution of mPHSMPhe16(5%) indicated that ionized PHP blocks were not miscible with PLLA3kDa-b-PEG2kDa and dissociated from PLLA3kDa-b-PEG2kDa micelles [33]. In vitro endolysosomal escape of mPHSMPhe16(20%)/f (i.e., mPHSMPhe16 prepared from 80 wt% PHP16-b-PEG2kDa and 20 wt% of PLLA3kDa-b-PEG2kDafolate) was similar to that of the PHSM/f system as shown in Figure 14.2E. That is, mPHSMPhe16(20%)/f caused even intracellular distribution of DHPE and LysoTraker dyes. This result may be induced by the PHP16-b-PEG2kDa

256

ENDOLYSOSOMOLYTICALLY ACTIVE PH-SENSITIVE POLYMERIC NANOTECHNOLOGY

released from the mixed micelles. The micelle systems effectively killed both A2780 wild-type cells and A2780/DOXR MDR cells [33]. Based on the destabilization pH of mPHSMPhe16 with different amounts of PLLA3kDa-b-PEG2kDa, mPHSMPhe22 and mPHSMPhe27 prepared from PHP22 and PHP27 with PLLA3kDa-b-PEG2kDa could be designed to target late endosomes and lysosomes, respectively, because PHSMPhe22 and PHSMPhe27 fabricated from PHP22 and PHP27 showed micelle destabilization at pH 5.8 and pH 5.2, respectively. 14.3.1.3 Virus-Mimetic Nanogels Virus-mimetic (VM) nanogels, which can migrate from one cell to other cells, have been developed to effectively kill solid tumors. As shown in Figure 14.4A, the nanogels consist of a hydrophobic PHP core (i.e., poly(His32-co-Phe6) with nearly pKb 6.4) and two hydrophilic layers of PEG2kDa and bovine serum albumin (BSA). Like the capsid shell of a virus, the gel’s mimetic outer shell is comprised of the core and BSA bridged by multiple PEG chains. Interestingly, the nanogels have a reversible swelling/deswelling characteristic that is pH dependent. As shown in Figure 14.4B, the size of VM nanogels at pH 7.4 was approximately 55 nm with a narrow distribution. At pH 6.8, VM nanogels still showed similar size and size distribution to those at pH 7.4. However, upon exposure to pH 6.4 (close to early endosomal pH), the nanogels became bigger up to approximately 355 nm with a broader distribution due to the ionization and swelling of the PHP core. This pH-induced swelling/deswelling transition of VM nanogels modulated DOX release rates. As shown in Figure 14.4C, DOX release rates from the nanogels at pH 7.4 and pH 6.8 were similar because the solidified core at pH 7.4 and pH 6.8 entrapped hydrophobic drugs. However, pH 6.4 induced the ionization and swelling of the PHP core and allowed more DOX to release. The release rate at pH 6.4 was approximately threefold faster than at pH 7.4 and pH 6.8. In addition, acidic pH-induced volumetric expansion of VM nanogels and known proton buffering capacities of histidine-based polymers (i.e., poly(His16-co-Phe6)) induced endolysosomal disruption as shown in Figure 14.4D. Control nanopaticles (NPs) without endolysosomolytic characters were localized with LysoTracker dye, whereas VM nanogels showed even intracellular distribution. Acidic pH-induced DOX release profiles and the endolysosomolytic function of VM nanogels effectively killed A2780 wild-type cells and A2780/ DOXR MDR cells compared to free DOX and control NP treatments. As shown in Figure 14.4E, the antitumor activity of the nanogels in pH 6.8 medium demonstrated that the nanogels were stable at pH 6.8 and released DOX to effectively kill tumor cells. Interestingly, after killing tumor cells, the nanogels that still contained DOX migrated to kill other tumor cells (Figure 14.4F). This concept of a “nanogel” that is stable in an acidic extracellular environment and can be activated to release drugs and to disrupt endolysosomes

257

ENDOLYSOSOMOLYTIC POLYMERIC NANOCARRIERS

pH 7.4 pH 6.4 DOX released F: folate F F F PEG Inner shell F F F F F F F pH F F F F F F F F F FF FF F F F BSA outer shell F FF 355 nm 55 nm

(A)

F F F F

F

F

(B) pH 7.4

pH 6.8

pH 7.4

0 100

0 100

pH 7.4

pH 6.4

pH 6.8

0 100

0 100

pH 7.4

pH 6.4

0 100 0 200 400 600 800 0 100 0 200 400 600 800

Particle size (nm)

(C) 20

FITC

Lyso Tracker

Merged

Control NPs VM-nanogels

pH pulses relative rate of DOX (µg/hour)

(D) 15 10 5 0

pH 7.4 pH 6.8 pH 6.4

: appled pH for 1 hour each

Cell viability / %

80

**

**

60 40 20 0

**

**

A2780 A2780/ A2780 A2780/ DOXR DOXR MDR cells MDR cells

Blank DOX-loaded DOX-loaded VM-nanogels Free DOX control NPs VM-nanogels

(F) pH 6.8

pH 7.4

(E) 100

Infected

not Infected

Subsequent Infections

A

B-0

B-1

B-2

B-3

A

B-0

B-1

B-2

B-3

A

B-0

B-1

B-2

B-3

A

B-0

B-1

B-2

B-3

Figure 14.4. VM nanogel: (A) Schematic presentation. (B) pH-modulating size change: A (pH 7.4), B (pH 6.8), and C (pH 6.4). (C) pH-dependent DOX release rate from DOX-loaded VM nanogels. (D) In vitro endosomal escaping activity. (E) pHdependent cytotoxicity of DOX-loaded VM nanogel (white), DOX-loaded control nanoparticles (gray), and free DOX (black) against A2780 wild-type cells and A2780/ DOXR MDR cells after 48-h incubation (DOX dose = 1 µg/mL). (F) Migration of DOXloaded VM nanogels in A2780/DOXR MDR cells. (Reproduced with permission from Ref. 38. Copyright © 2008 Wiley-VCH Verlag GmbH & Co.KGaA.) (See color insert.)

while contained in the endolysosomal pathway could be expanded to treat ischemia in the brain and heart. In this case, although migration of the nanogel toward other cells might be difficult, the pH-dependent drug release profile of the nanogel could effectively prevent the progress of ischemia over a long time period because ischemic states are initiated by drops in pH.

258

ENDOLYSOSOMOLYTICALLY ACTIVE PH-SENSITIVE POLYMERIC NANOTECHNOLOGY

14.3.2 Endolysosomolytic Polymeric Gene Nanocomplexes Kang and Bae have been interested in developing cell-customized endosomolytic agents for more effectively delivering therapeutics because endolysosomal characteristics such as endolysosomal pH [34, 35], endolysosomal membrane composition [36, 37], and acidification rate are cell specific. To investigate cell-specific endolysosomolytic agents, sulfonamides were selected. Sulfonamides have broad-range pKa values (3–11) and hydrophobicity conferred by various substituted groups, R (see Figure 14.1) [16]. For feasibility studies, sulfamethizole (SMT; pKa 5.45), sulfadimethoxine (SDM; pKa 6.1), sulfadiazine (SDZ; pKa 6.4), and sulfamerazine (SMZ; pKa 7.0) were selected because their pKa values fall within the endolysosomal pH range. Using chain transfer radical polymerization, the synthesized sulfonamide oligomers (designated as OSMT, OSDM, OSDZ, and OSMZ) had Mn = 1.8–2.5 kDa. Oligomeric sulfonamides (OSAs) showed different proton buffering and aqueous solubility transition within the endosomal/lysosomal pH range. In aqueous solubility transition studies of the OSAs (Figure 14.5A), OSMZ showed solubility transitions over a broad pH range, whereas OSMT, OSDM, and OSDZ showed relatively sharp transmittance changes within a narrow pH range. As shown in Figure 14.5B, OSMT and OSDZ demonstrated broad proton buffering over the pH ranges of 5.0–6.4 and 5.7–7.3, respectively, whereas OSDM and OSMZ buffered protons at a specific pH, 6.5 and 7.3, respectively. Their apparent pKa values were 5.7 (OSMT), 6.5 (OSDM and OSDZ), and 7.3 (OSMZ) and were slightly higher than their monomeric sulfonamides counterparts [16]. To understand the endolysosomolytic functions of anionic OSAs in polymeric nanocarriers, a well-known polymeric gene delivery system without endosomolytic function (here, poly(l-lysine) (PLL)-based polyplexes) was selected. Anionic OSAs cannot directly carry anionic genes, but OSAs can be incorporated within any polycation-based gene carriers. In intracellular distribution studies of polyplexes using pH-sensitive fluorescein pDNA (F-DNA), OSDZ-polyplexes (i.e., PLL/OSDZ/F-DNA) showed broad distribution in the cytoplasm, whereas PLL/F-DNA complexes were mostly localized (Figure 14.5C). These differences in fluorescence distribution demonstrate that OSA polyplexes caused more pDNA to escape from endosomal/lysosomal compartments than PLL/pDNA complexes. This faster endosomal release of OSDZ polyplexes resulted in faster gene expression than for PLL/pDNA complexes. In in vitro transfection studies using three different cell lines (i.e., human hepatoma HepG2 cells, human embryonic kidney HEK293 cells, and rat insulinoma RINm5F cells), OSA polyplexes showed 4–55-fold better transfection efficiency than control polyplexes (PLL/pDNA) as shown in Figure 14.5D. Interestingly, OSDM and OSDZ were more effective in transfecting HEK293 cells whereas OSMZ was the best for transfecting RINm5F cells. This study supports the need for cell-customized endosomolytic agents to achieve

CONCLUSION

8

PLL/OSDZ/F-DNA

(D)

150 mM NaCl OSMT OSDM OSDZ OSMZ OPAA

4

(C)

6 pH

PLL/F-DNA

7

20µm

150 mM NaCl OSMT OSDM OSDZ OSMZ OPAA

100 150 200 250 300 350 Added volume of 0.1 N HCI (µL)

80 70 60 50 40 30 20 10 0

HEK293 HepG2 RINm5F

PL

L PE O I SM O T SD M O SD O Z SM O Z PA A

20µm

5

Normalized Transfection Efficiency (Fold)

Transmittance (%)

100 90 80 70 60 50 40 30 20 10 0

(B) 11 10 9 8 7 6 5 4 3 2 50

pH

(A)

259

Polyplex type

Figure 14.5. (A) pH-dependent aqueous solubility transition of OSA and OPAA (oligomeric propylacrylic acid) solutions. (B) Acid–base titration curves of OSA and OPAA solutions. (C) In vitro intracellular localization studies of PLL/pDNA and PLL/ OSDZ/pDNA complexes using pH-sensitive F-DNA. (D) In vitro transfection studies for OSA-containing PLL/pDNA complexes (OSA polyplexes) containing a luciferase gene in HepG2, HEK293, and RINm5F cells. Dose of OSA and OPPA was 5 nmol (based on their monomeric units) per 1 µg pDNA. Normalized transfection efficiency was defined as (Absolute transfection efficiency of polyplexes)/(Absolute transfection efficiency of PLL/pDNA complexes) for a specific cell. Charge ratio (+/−) of polyplexes was 3 except for PEI/pDNA (+/− = 5). (Reproduced with permission from Ref. 16. Copyright © 2007 Wiley-VCH Verlag GmbH & Co.KGaA.)

clinically effective gene delivery [16]. In addition, these anionic materials could provide an opportunity to revisit the use of other biocompatible but less transfection-efficient gene carriers.

14.4

CONCLUSION

Endolysosomolytic materials are an indispensable component for effectively targeting subcellular compartments and maximizing therapeutic benefits. Because endolysosomal destabilization is mostly mediated by “pH,” understanding extracellular pH and endolysosomal pH of various pathological states

260

ENDOLYSOSOMOLYTICALLY ACTIVE PH-SENSITIVE POLYMERIC NANOTECHNOLOGY

is important. Knowledge about the pH in the body, organs, tissues, cells, and extracellular and intracellular environments for healthy and disease states would aid the design of effective endolysosomolytic polymeric nanocarriers.

ACKNOWLEDGMENTS This work is partially supported by NIH grants (CA101850 and GM82866). The authors appreciate the help of Deepa Mishra, Dr. Haiqing Yin, and Dr. Dongin Kim for their critical reading.

REFERENCES 1. Kang, H. C., Lee, E. S., Na, K., et al. Stimuli-sensitive nanosystems: for drug and gene delivery. In: V. P. Torchilin, ed. Multifunctional Pharmaceutical Nanocarriers. Springer, New York, 2008, pp. 161–199. 2. Herweijer, H. and Wolff, J. A. Gene therapy progress and prospects: hydrodynamic gene delivery. Gene Ther. 14: 99–107 (2007). 3. Maeda, H., Bharate, G. Y., and Daruwalla, J. Polymeric drugs for efficient tumortargeted drug delivery based on EPR-effect. Eur. J. Pharm. Biopharm. 71: 409–419 (2009). 4. Lee, E. S., Gao, Z., and Bae, Y. H. Recent progress in tumor pH targeting nanotechnology. J. Control. Release 132: 164–170 (2008). 5. Nishikawa, M. Development of cell-specific targeting systems for drugs and genes. Biol. Pharm. Bull. 28: 195–200 (2005). 6. Breunig, M., Bauer, S., and Goepferich, A. Polymers and nanoparticles: intelligent tools for intracellular targeting? Eur. J. Pharm. Biopharm. 68: 112–128 (2008). 7. Bareford, L. M., and Swaan, P. W. Endocytic mechanisms for targeted drug delivery. Adv. Drug Deliv. Rev. 59: 748–758 (2007). 8. Boussif, O., Lezoualc’h, F., Zanta, M. A., et al. A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: polyethylenimine. Proc. Natl. Acad. Sci. USA. 92: 7297–7301 (1995). 9. Pichon, C., Goncalves, C., and Midoux, P. Histidine-rich peptides and polymers for nucleic acids delivery. Adv. Drug Deliv. Rev. 53: 75–94 (2001). 10. Lee, E. S., Na, K., and Bae, Y. H. Doxorubicin loaded pH-sensitive polymeric micelles for reversal of resistant MCF-7 tumor. J. Control. Release 103: 405–418 (2005). 11. Yang, S. R., Lee, H. J., and Kim, J. D. Histidine-conjugated poly(amino acid) derivatives for the novel endosomolytic delivery carrier of doxorubicin. J. Control. Release 114: 60–68 (2006). 12. Jones, R. A., Cheung, C. Y., Black, F. E., et al. Poly(2-alkylacrylic acid) polymers deliver molecules to the cytosol by pH-sensitive disruption of endosomal vesicles. Biochem. J. 372: 65–75 (2003).

REFERENCES

261

13. Yessine, M. A., Lafleur, M., Meier, C., et al. Characterization of the membranedestabilizing properties of different pH-sensitive methacrylic acid copolymers. Biochim. Biophys. Acta. 1613: 28–38 (2003). 14. Yessine, M. A. and Leroux, J. C. Membrane-destabilizing polyanions: interaction with lipid bilayers and endosomal escape of biomacromolecules. Adv. Drug Deliv. Rev. 56: 999–1021 (2004). 15. Kiang, T., Bright, C., Cheung, C. Y., et al. Formulation of chitosan-DNA nanoparticles with poly(propyl acrylic acid) enhances gene expression. J. Biomater. Sci. Polym. Ed. 15: 1405–1421 (2004). 16. Kang, H. C. and Bae, Y. H. pH-tunable endosomolytic oligomers for enhanced nucleic acid delivery. Adv. Funct. Mater. 17: 1263–1272 (2007). 17. Cho, Y. W., Kim, J. D., and Park, K. Polycation gene delivery systems: escape from endosomes to cytosol. J. Pharm. Pharmacol. 55: 721–734 (2003). 18. Kang, H. C., Lee, M., and Bae, Y. H. Polymeric gene carriers. Crit. Rev. Eukaryot. Gene Expression 15: 317–342 (2005). 19. Pack, D. W., Hoffman, A. S., Pun, S., et al. Design and development of polymers for gene delivery. Nat. Rev. Drug Discov. 4: 581–593 (2005). 20. Sonawane, N. D., Szoka, F. C. Jr., and Verkman, A. S. Chloride accumulation and swelling in endosomes enhances DNA transfer by polyamine–DNA polyplexes. J. Biol. Chem. 278: 44826–44831 (2003). 21. Xie, A. F. and Granick, S. Phospholipid membranes as substrates for polymer adsorption. Nat. Mater. 1: 129–133 (2002). 22. Tirrell, D. A., Takigawa, D. Y., and Seki, K. pH sensitization of phospholipid vesicles via complexation with synthetic poly(carboxylic acid)s. Ann. N. Y. Acad. Sci. 446: 237–248 (1985). 23. Thomas, J. L. and Tirrell, D. A. Polymer-induced leakage of cations from dioeoylphosphatidylcholine and phosphatidylglycerol liposomes. J. Control. Release 67: 203–209 (2000). 24. Volk, T., Jahde, E., Fortmeyer, H. P., et al. pH in human tumour xenografts: effect of intravenous administration of glucose. Br. J. Cancer 68: 492–500 (1993). 25. Sarantopoulos, C., McCallum, B., Sapunar, D., et al. ATP-sensitive potassium channels in rat primary afferent neurons: the effect of neuropathic injury and gabapentin. Neurosci. Lett. 343: 185–189 (2003). 26. Hunjan, S., Mason, R. P., Mehta, V. D., et al. Simultaneous intracellular and extracellular pH measurement in the heart by 19F NMR of 6-fluoropyridoxol. Magn. Reson. Med. 39: 551–556 (1998). 27. Wattiaux, R., Laurent, N., Wattiaux-De Coninck, S., et al. Endosomes, lysosomes: their impication in gene transfer. Adv. Drug Deliv. Rev. 41: 201–208 (2000). 28. Gilbert, H. F. Basic Concepts in Biochemistry: A Student’s Survival Guide, 2nd ed. McGraw-Hill Health Professional Division, New York, 1999. 29. Lee, E. S., Shin, H. J., Na, K., et al. Poly(l-histidine)-PEG block copolymer micelles and pH-induced destabilization. J. Control. Release 90: 363–374 (2003). 30. Kim, D., Lee, E. S., Park, K., et al. Doxorubicin loaded pH-sensitive micelle: antitumoral efficacy against ovarian A2780/DOXR tumor. Pharm. Res. 25: 2074–2082 (2008).

262

ENDOLYSOSOMOLYTICALLY ACTIVE PH-SENSITIVE POLYMERIC NANOTECHNOLOGY

31. Lee, E. S., Na, K., and Bae, Y. H. Polymeric micelle for tumor pH and folatemediated targeting. J. Control. Release 91: 103–113 (2003). 32. Yin, H., Lee, E. S., Kim, D., et al. Physicochemical characteristics of pH-sensitive poly(l-histidine)-b-poly(ethylene glycol)/poly(l-lactide)-b-poly(ethylene glycol) mixed micelles. J. Control. Release 126: 130–138 (2008). 33. Kim, D., Lee, E. S., Oh, K. T., et al. Doxorubicin-loaded polymeric micelle overcomes multidrug resistance of cancer by double-targeting folate receptor and early endosomal pH. Small 4: 2043–2050 (2008). 34. Rybak, S. L., Lanni, F., and Murphy, R. F. Theoretical considerations on the role of membrane potential in the regulation of endosomal pH. Biophys. J. 73: 674–687 (1997). 35. Rybak, S. L. and Murphy, R. F. Primary cell cultures from murine kidney and heart differ in endosomal pH. J. Cell Physiol. 176: 216–222 (1998). 36. Alberts, B., Johnson, A., Lewis, J., et al. Molecular Biology of the Cell, 4th ed. Garland Science, New York, 2002. 37. Evans, W. H. and Hardison, W. G. Phospholipid, cholesterol, polypeptide and glycoprotein composition of hepatic endosome subfractions. Biochem. J. 232: 33– 36 (1985). 38. Lee, E. S., Kim, D., Youn, Y. S., et al. A virus-mimetic nanogel vehicle. Angew. Chem. Int. Ed. Engl. 47: 2418–2421 (2008).

CHAPTER 15

Uptake and Intracellular Dynamics of Proteins Internalized by Cell-Penetrating Peptides ARWYN T. JONES Welsh School of Pharmacy, Cardiff University, Cardiff, Wales, United Kingdom

15.1

INTRODUCTION

Many pathogens and therapeutic macromolecules need to enter cells to fulfill a requirement for, respectively, survival and replication or the treatment of disease. An ability to target specific cells and then overcome the barrier posed by the plasma membrane to reach the cytosol or a defined subcellular organelle, and to then mediate an effective biological response is, however, a feat limited to pathogens or pathogen proteins. An example is the HIV transactivator of transcription protein, HIV-Tat, that enters cells by endocytosis, escapes from the endolysosomal system to the cytosol, and then moves to the nucleus to mediate its effects on transcription. This protein highlights the extraordinary capacity of evolution to manufacture simple but highly efficient targeting systems. Essential for HIV-Tat to perform these feats is a basic protein transduction domain that can enter cells either alone or associated with cargo of various forms ranging from other proteins to nanoparticles. Hundereds of natural and synthetic protein transduction domains or cellpenetrating peptide (CPP) sequences have now been described and this chapter highlights the attempts that have been made to understand the cellular dynamics of proteins attached by various means to CPPs. An appreciation of the mechanisms by which these systems operate may hold answers to address a major pharmaceutical need for the design of more efficient systems for cellular delivery of therapeutic proteins. This chapter will describe endocytosis and endocytic pathways (Section 15.2), the need to deliver proteins into cells Organelle-Specific Pharmaceutical Nanotechnology, Edited by Volkmar Weissig and Gerard G. M. D’Souza Copyright © 2010 John Wiley & Sons, Inc.

263

264

UPTAKE AND INTRACELLULAR DYNAMICS OF PROTEINS INTERNALIZED

(Section 15.3), our current understanding of the mechanism by which the HIV-Tat enters cells and reaches the nucleus (Section 15.4), delivery of proteins linked to other CPP sequences (Section 15.5), alternative mechanisms for covalent and noncovalent attachment of CPPs to proteins (Section 15.6), delivery of CPP proteins to lysosomes (Section 15.7), and a comparison of the cellular dynamics of CPP proteins with CPP fluorophores (Section 15.8).

15.2

ENDOCYTIC PATHWAYS

The plasma membrane is a highly effective and selective boundary between the inside and outside of cells. All cells, however, have a physiological need to internalize membrane-impermeable molecules and to do this they invaginate portions of their plasma membrane to internalize membrane and accompanying extracellular fluid. This process is called endocytosis and the internalized membrane or vesicular structures are then delivered together with the fluid cargo to a sorting station that directs further membrane traffic to one of a number of cellular destinations. Only recently are we beginning to appreciate the enormous complexities of endocytosis and endocytic pathways [1]. This process, regulated by a network of proteins and lipids, is essential for several often unrelated cellular functions, and this may be why a single cell has the capacity to internalize material through several different endocytic pathways (Figure 15.1). These include those originating from clathrin-, and non-clathrin-coated structures such as caveolae through to larger structures called macropinosomes that are formed following plasma membrane ruffling. It is beyond the scope of this chapter to discuss endocytosis at great length and those wishing to gain a good foundation on the current knowledge of endocytic pathways should consult the reviews in Refs 1–9 and also this informative web site at http://endocytosis.org/. 15.2.1 Using Endocytic Pathways to Deliver Therapeutics The effectiveness of using endocytosis to deliver therapeutic macromolecules is in part constrained by the fact that the fate of the therapeutic within one of these pathways is predetermined by the cell’s trafficking mechanisms and biological barriers posed initially by the plasma membrane and then by the endolysosomal network [10–12]. Thus fragile molecules such as proteins and genes may rapidly be delivered to hostile environments such as lysosomes and inactivated before they have any chance to reach and interact with their intended targets. Pathogens and pathogen-proteins especially toxins have been shown to exploit and manipulate these pathways to reach defined subcellular destinations, thus suggesting that therapeutics can also be delivered by similar mechanisms [13, 14]. Endocytic features such as low pH and high degradative activities can, however, act as activators of bioresponsive molecules to enhance translocation of the active entity before the cell’s endocytic pathway

ENDOCYTIC PATHWAYS

265

Figure 15.1. Endocytic pathways and cell entry mechanisms available for proteins attached to cell-penetrating sequences or natural cell-penetrating proteins such as HIV Tat. Numbers 1–6 represent defined pinocytic pathways and not described here or in the text is phagocytosis.1, Clathrin-coated vesicles; 2, caveolae; 3, fluid-phase uptake; 4, flotillin-1-dependent pathway; 5, other clathrin- and caveolin-independent pathway(s) [5]; 6, macropinocytosis, and 7, direct penetration and entry through the plasma membrane. The named proteins represent those required for the pathways to operate and thus offer the opportunity for their mutation or depletion via siRNA. This together with pharmacological inhibitors will allow for more accurate requirement of particular pathways for entry of macromolecules designed for intracellular delivery and targeting. It is likely that an early or sorting endosome exists to accept cargo delivered through all these endocytic pathways and some have been defined. It still remains to be determined as to whether entry via any particular pathway is advantageous for allowing subsequent escape from the endolysosomal system to locate the protein in the cytosol.

traffics it to its final destination. Enhancing the delivery of therapeutics through endocytosis is therefore dependent on acquisition of a high level of understanding of specific endocytic pathways inherent in the target cell, the traffic and fate of the molecule within endocytic organelles, the effect of the macromolecule on the dynamics of endocytic pathways, the downstream effects of these on the integrity of the cell, and if required the mechanism by which escape from the endolysosomal system occurs.

266

UPTAKE AND INTRACELLULAR DYNAMICS OF PROTEINS INTERNALIZED

15.2.2 Identifying Endocytic Pathways Cell biologists studying endocytic pathways are consistently identifying new protein and lipid mediators and a goal for this remit that overlaps with that of drug delivery is to identify a protein that is only involved in regulating one distinct pathway. This then opens the possibility of either mutating this protein or inhibiting its manufacture to gain a better understanding of the pathway that it regulates and also the fate of the cargo that it helps to deliver. Much of the current information regarding the endocytic pathways utilized by drug delivery systems, including those using CPPs, comes from studies using microscopy and endocytic probes [15, 16] in conjunction with a library of pharmacological inhibitors of endocytic pathways [17]. Examples include methyl ß-cyclodextrin (MßCD), amiloride, chlorpromazine, and cytochalasins. At one time these agents were deemed to be specific for certain pathways but it is now clear that all suffer predominantly from a lack of specificity, and toxicity is also a common problem [17]. Increasingly, researchers are using molecular biology tools such as cell lines expressing dominant negative forms of endocytic proteins or siRNA-mediated depletion of proteins that regulate endocytic pathways. This siRNA approach has recently identified new pathways and roles in endocytosis for proteins such as flotillin-1 [18] and significantly added to the current knowledge on established endocytic proteins such as clathrin [19–21]. The use of siRNA depletion has its own drawbacks for those wishing to use the technology for analyzing endocytosis of their drug delivery systems and these include (1) the fact that a delivery system is first required to introduce the siRNA into the cytosol, (2) the possibility that the protein of interest regulates other membrane trafficking pathways, not necessarily endosomal (clathrin heavy chain and dynamin are good examples), and (3) the fact that depletion of the protein affects cell viability or dysregulates other endocytic pathways. Another important aspect of this approach is that a specific probe is required whose uptake is confined to one pathway; this allows researchers to identify whether expression of mutant proteins or siRNA depletion is having a biological response. Transferrin is a very well-established marker for uptake via clathrin-coated vesicles [22], both anti CD35 antibodies and poly(ethylenimine) complexes have been shown to be partially internalized in a flotillin-1-dependent manner [18, 23] and cholera toxin is often used as a marker for uptake via caveolae. A number of other poorly defined pathways exist and identifying specific endocytic probes for those not organized by clathrin has been difficult [5, 9]. For example, a number of studies show that cholera toxin can access through different pathways [9] and very careful analysis is required to ensure that uptake is confined to one route [24]. 15.3

DELIVERING PROTEINS INTO CELLS

Delivering a full-length protein that may have multiple domains, interactions, and effects is sometimes required to mediate a necessary biological response, whether for research or therapy. The realistic potential of delivering full-

HIV-TAT PROTEIN

267

length proteins was probably seeded when it was realized that many diseases were caused by malfunctioning or missing proteins and the technology became available for their purification, especially from transformed microorganisms such as Escherichia coli. Most proteins are membrane impermeable and a large fraction of them function inside cells; therefore systems are required for their delivery into their site of action, often the cytosol and in some cases subcellular organelles such as the nucleus and mitochondria. This therefore requires an intracellular delivery system and potential candidates for performing this function are short peptide sequences that have the capacity to enter cells either as single entities or attached to much larger cargo such as fulllength proteins. These are called protein transduction domains (PTDs) or cell-penetrating peptides. 15.3.1 Defining Cell-Penetrating Peptides and Protein Transduction Domains The term cell-penetrating peptide gained prominence around 20 years ago as did the use of the term protein transduction domain and there is still healthy discussion in meetings over the most appropriate terminology for what could generally be described as membrane active peptides. Peptides that interact with eukaryotic and prokaryotic membranes have, however, been described in the literature for decades before either of these now familiar terms gained prominence [25]. An excellent review written by Plank et al. [26] in 1998 described a number of peptides with potential for delivering macromolecules and examples include melittin, mastoparan, GALA, and KALA; these are more familiarly known as pore-forming, membrane active, or destabilizing peptides (Table 15.1). However, as previously noted, the distinction between these and what we now refer to as CPPs or PTDs can sometimes be difficult to define [27]. Hundereds of peptides have now been described that share a common ability to deliver cargo to cells and their heterogeneity with respect to sequence and physical and biological properties suggests that neither of these commonly used terms is appropriate for all. One of the best characterized is the cell-penetrating peptide from the protein transduction domain of the HIV-Tat protein.

15.4 HIV-TAT PROTEIN The genome of HIV encodes proteins that are important for structure, organization of its envelope, enzyme regulators, and also transactivators such as HIV-Tat. This 101 amino acid protein is manufactured inside infected cells and gains entry into the nucleus, where it binds RNA to activate viral transcription [28]. Residues 1–86 are transactivation active and a number of laboratories have used this truncated form for their studies, rather than the full-length protein. The protein also appears to be released from intact virus-infected cells and this fraction can enter other cells via endocytosis

268

UPTAKE AND INTRACELLULAR DYNAMICS OF PROTEINS INTERNALIZED

TABLE 15.1

Sequences of Membrane Active Peptidesa Described in This Chapter

Peptide HIV-Tat Bov-Tat JDV Tat Transportan TP10 pVEC Penetratin (pAntennapedia) MPG PEP-1 PTD-5 PDX-1 PTD VP22 GALA KALA Melittin Mastoparan HA2 HA2′

Sequence

Reference

GRKKRRQRRRPPQ RRRGTRGKGRRIRR RHDGRRKKRGTRGKGR GWTLNSAGYLLGKINLKALAALAKKIL AGYLLGKINLKALAALAKKIL LLIILRRRIRKQAHAHSK RQIKIWFQNRRMKWKK

127 54 53 83 106 84 78

GALFLGFLGAAGSTMGAWSQPKKKRKV KETWWETWWTEWSQPKKKRKV YARAARRAARR RHIKIWFQNRRMKWKK DAATATRGRSAASRPTERPRAPARSASRPRRPVE WEAALAEALAEALAEHLAEALAEALEALAA WEAKLAKALAKALAKHLAKALAKALKACEA IKITTMLAKLGKVLAHV INLKALAALAKKIL GDIMGEWGNEIFGAIAGFLG GLFEAIEGFIENGWEGMIDGWYG

89 88 81 62 128 92 92 25 25 58 64

a

Variations on these sequences are commonly used, especially the addition of terminal residues such as cysteine for conjugation to fluorophores. Bov, bovine immunodeficiency virus; JDV, Jembrane disease virus; PDX-1, pancreatic and duodenal homeobox factor-1 protein transduction domain.

[29–32]. But most importantly for those interested in drug delivery, the protein then escapes from the endolysosomal network to the cytosol and then on to the nucleus [33]. HIV-Tat has also been shown to interact with a number of other cell surface molecules to mediate biological effects, and its location on the plasma membrane and interaction with viral gp-120 may also aid in promoting viral entry [34, 35]. A crystal structure for HIV-Tat has not been published, but interestingly NMR studies reveal that the protein may be one of the many proteins that exist as an unfolded entity that may allow for multiple interactions with affectors and effectors [28, 36–38]. The basic domain is seen fully exposed on the surface (Figure 15.2) but it is likely that other parts of the molecule, perhaps the hydrophobic core, also aid in endosomal escape. 15.4.1 Cellular Dynamics HIV-Tat Protein The biology of HIV-Tat protein has been the subject of intense scrutiny and like other HIV proteins it is a target for anti-HIV drugs [28]. Its biology, either manufactured by the protein synthesis machinery inside a virus-infected cell or following uptake from the extracellular fluid, is extremely complex and over 250 HIV-Tat interacting partners have now been identified [39].

HIV-TAT PROTEIN

(A)

269

(B)

Figure 15.2. NMR structures of HIV-Tat 1–86. Both (A) [37] and (B) [36] show quite different structures but common is the exposure of the basic domain, residues 48–60, shown in blue. The deposited structures were obtained from the Protein Data Bank and were processed using Pymol software.

15.4.1.1 Binding to the Plasma Membrane A requirement for HIV-Tat endocytosis, endosomal escape, nuclear enrichment, and RNA binding is a short highly basic sequence spanning residues 47–57 (Table 15.1). This region interacts with negatively charged cell surface heparan sulfate proteoglycans and is thought to gain access to the cell as an ionic complex with these sugars [40–42]. The initial cellular uptake and endocytosis of HIV-Tat protein and of functional residues 1–72 were performed using E. coli expressed protein that was then purified and labeled with 125I [30, 31]. Uptake was partially characterized in a number of cell lines and found to be dependent on temperature and could be inhibited by competing anionic charge in the form of heparin. Of note is the fact that the protein at 1 nM (∼10 ng/mL) was able to promote transactivation—an amazing feat and one that bears consideration when looking at subsequent studies aiming to delineate in much more detail its uptake and intracellular dynamics and also of HIV-Tat fusion proteins and HIV-Tat peptide linked proteins. Immediate to the finding that HIV-Tat could enter cells were studies showing that major fractions of the protein or the basic region could be linked to the N or C terminus of proteins and promote their uptake as active entities both in vitro in a number of different cell lines, and also in vivo [43–45]. There was thus an obvious interest in elucidating the mechanism by which these proteins were able to overcome membrane bilayers. The creation of a HIVTat-Enhanced Green Fluorescent protein (EGFP) fusion protein confirmed that this domain interacted strongly with heparan sulfate proteoglycans and that attenuated uptake of the fusion protein and its capacity for transactivation was observed in Chinese hamster ovary (CHO) cells deficient in proteoglycan synthesis [42, 46]. Similar reduced uptake and transactivation was noted when

270

UPTAKE AND INTRACELLULAR DYNAMICS OF PROTEINS INTERNALIZED

soluble glycosaminoglycans, most notably heparin, were added to cell culture media during incubations with HIV-Tat fusion proteins and with HIV-Tat alone [31]. Recent data in primary brain cells also demonstrate a degree of positive correlation between glycosaminoglycan expression and the ability of GFP-HIV-Tat peptide fusion to gain cell entry [47]. 15.4.1.2 Endocytosis The most extensive characterization of endocytosis and translocation of native HIV-Tat (residues 1–86) was performed in Jurkat T lymphocytes [35]. Uptake of the protein at 50 nM was inhibited by approximately 50% in cells treated with chlorpromazine, a well-known inhibitor of clathrin-coated uptake and it was localized, using immunocytochemistry, in clathrin-coated pits. In support of this, the ability of Tat to escape from the endolysosomal system was inhibited in cells expressing mutant proteins that affected clathrin-coated uptake. A cell-free translocation assay was developed to monitor Tat release from endosomes and this revealed that escape was dependent on endosomal acidification. 15.4.1.3 Fixing Artifacts It is fair to say that the field of cell-penetrating peptides shuddered somewhat when Lundberg and co-workers published two reports regarding the uptake and subcellular localization of GFP extended with cell-penetrating sequences from HIV-Tat and herpes virus protein VP22, octalysine and octaarginine [48, 49]. The studies highlighted the different capacities of these four sequences to label CHO cells but also suggested that fixing cells for microscopy and flow cytometry had major effects leading to an overestimation of the capacity of CPPs to deliver cargo to the inside of cells and the nucleus rather than to label, respectively, the plasma membrane and endocytic vesicles. These studies and others investigating fluorescently labeled CPPs [50] had a huge effect on this field and awoke researchers to the realization that studies with CPPs, (especially the cationic forms) needed to be performed in unfixed cells, and that stringent washing procedures with enzymes and anionic molecules were required to remove surface label. 15.4.1.4 Endocytic Pathways Utilized by HIV-Tat Protein The aforementioned study of HIV-Tat endocytosis in Jurkat cells was performed in fixed cells using anti HIV-Tat antibodies [35], but to perform live cell imaging there was a requirement to tag the protein with a fluorophore or to extend its sequence with a fluorescent protein such as GFP. In these and other studies, additional tags are added to the protein to aid in purification and these include His6 and glutathione S-transferase tags [49, 51] that may or not be linked to the protein of interest via protease sensitive linkages [52]; this minimizes the possibility of tag-mediated effects. This technology has recently been used to investigate the uptake and subcellular distribution of Tat proteins from HIV 1 and more recently other viruses [53, 54]. Comparative analysis has been performed on the uptake and intracellular localization of GST-Tat (full length)-EGFP and GST-Tat(48–60)EGFP in a

HIV-TAT PROTEIN

271

HeLa cell derivative HL3T1 [51, 55]. Uptake of both was dependent on endocytosis and the actin cytoskeleton and fluorescence was shown to colocalize in vesicles labeled with caveolin-1 and cholera toxin that is often used as a marker of uptake via caveolae. The problems of using cholera toxin as an endocytic pathway marker have been described above. The cholesterol-sequestering agent MßCD inhibited the uptake of both fusion proteins, suggesting that lipid rafts were necessary for uptake, and very little colocalization was observed with fluorescent transferrin, a marker of uptake via clathrin-coated vesicles. This data was rather different from studies in T cells showing that uptake of HIV-Tat alone (i.e., without cargo) was through clathrin-coated pits [35]. One important aspect of the study on the uptake of HIV-Tat in the absence of cargo was the fact that it was performed in Jurkat T cells that do not express caveolin-1 [56], and thus uptake in this cell line could not have been mediated by caveolae. HeLa cells, on the other hand, express low levels of caveolin-1 but it is difficult to visualize caveolae in this cell line [57]. Overall, the studies demonstrate that caveolae are not absolutely required for cell entry and escape from endocytic pathways. A role of clathrin and caveolae was disputed when a Tat-Cre recombinase assay was used to investigate the capacity of the basic domain of Tat to deliver the 38-kDa Cre recombinase protein to mediate expression of the EGFP reporter gene [58]. The purified Cre protein was labeled with Alexa488 and its uptake into cells was shown to be inhibited after treatment with the cholesterol-interacting agent nystatin, that along with MßCD also inhibited recombination. The study showed that Tat-Cre entry and function were not dependent on caveolin expression and thus uptake via caveolae. Recombination, however, was inhibited following cytochalasin D and amiloride treatment, suggesting that macropinocytosis may be involved in peptide uptake. This endocytic process usually occurs in growth factor stimulated cells and involves extensive actin reorganization and engulfment of large volumes of extracellular liquid that are then enclosed to form internal macropinosomes [4, 6, 59]. Vesicles formed from other endocytic pathways have diameters in the range of 60–120 nm but macropinosomes can be micrometers in diameter. These enlarged structures were not observed but in support of a role for this pathway in Tat-Cre uptake was an increased fluid phase uptake of 70 kDa dextran. This was an important observation suggesting that the basic domain of Tat was interacting with the plasma membrane to cause cytoskeletal activation, ruffling, and enhanced uptake of fluid-phase markers, and possibly itself. It is important to note here that 70-kDa dextran (hydrodynamic radius 5.5 nm) is not a specific marker for macropinocytosis, as is sometimes mentioned in the literature, but rather a good probe to monitor enhanced fluid-phase uptake that is characteristic of this endocytic event. The studies did suggest, however, that entry of this protein and indeed cationic CPPs was not a passive mechanism but one that involved activation and reorganization at the plasma membrane.

272

UPTAKE AND INTRACELLULAR DYNAMICS OF PROTEINS INTERNALIZED

A later study using a similar Tat-Cre system also showed that energy and an intact actin cytoskeleton were required for cell entry [60]. The data also demonstrated the limitations of confocal microscopy for these kinds of analysis in so much that despite the fact that there was no visible (microscopical) evidence for cytosolic and nuclear Cre, the biological readout assay suggested otherwise. Conversely, confocal microscopy and laser-induced membrane damage and leakage can also potentially overestimate the fraction of an internalized entity that is released in the cytosol [61]. Macropinocytosis was also implicated in the uptake of the transcription factor pancreatic and duodenal homeobox factor 1 (PDX-1) [62] that has within its sequence a cell-penetrating domain (Table 15.1). Labeling recombinant PDX-1 with fluorescein isothiocyanate (FITC) and adding it to cells allowed for microscopical analysis of its distribution, and at 5-µM extracellular concentration the protein was found to label vesicles and the cytosol. Further analysis of the recombinant protein was not performed but amiloride was shown to inhibit uptake of the protein transduction domain of this protein. 15.4.2 Endocytosis of Tat from Other Viruses GFP fusion proteins of Jembrane disease virus and bovine immunodeficiency virus Tat have been generated and these also have basic rich sequences, and the Jembrane disease virus basic domain is longer than that of HIV-Tat (Table 15.1). Both of these proteins efficiently delivered GFP into cells but the uptake mechanism was not studied in detail. Interestingly, there was evidence that uptake of denatured Tat-EGFP constructs occurred via an energyindependent manner whereas uptake of the native proteins was inhibited in the absence of energy and therefore dependent on endocytosis. But this needs further analysis to reveal whether the proteins were actually located inside the cells rather than on the cell surface. 15.4.3 Escaping from the Endolysosomal Network If endosomal escape of HIV-Tat protein is dependent on acidification, then the possibility exists that the ultimate fate of the fraction that cannot escape will be degradation in lysosomes. One study utilized a mutant GFP (F64L, T203Y) that has a pH-dependent emission profile and its pH environment can then be monitored during endocytosis [63]. The mutant was extended with the HIV-Tat basic sequence and the protein was then added to cells to reveal that the fusion protein reached a compartment of pH 5.8 before the GFP fluorescence was quenched and the protein degraded. Extensive degradation had occurred within 8 hours. Degradation is a major issue if there is a requirement for the therapeutic to be delivered from the endolysosomal system, and beyond, and attempts have been made to enhance escape from endosomes. One study showed that coincubation of a membrane fusion peptide HA2 derived from the N terminus of influenza A hemagluttinin (Table 15.1) to

HIV-TAT PROTEIN

273

incubations with Tat-Cre enhanced its capacity for recombination [58]. A longer sequence that was also called HA2 (HA2 in Table 15.1) was shown to enhance the transcriptional activity of extracellularly delivered p53RRRRRRRRR fusion protein [64]. However, the approach here was quite different as in this case the CPP, fusion peptide, and delivered cargo were linked as one protein HA-p53-R9. Subsequent labeling of this protein with FITC and confocal microscopy analysis in live glioma cells showed it to be diffusely localized and enriched in the nucleus; this was 4 hours postincubation. The same fusion protein lacking the HA2 moiety was predominantly localized to the plasma membrane. The exact mechanism by which HIV-Tat protein in the absence of a defined fusion sequence escapes from the endolysosomal system is unknown but to aid viral replication a fraction must gain access to first the cytosol and then the nucleus. The previously mentioned GFP mutant study [63] demonstrated that the protein reached a compartment of pH 5.8 but whether the “active” fraction has already escaped before this pH is reached is unknown. It is technically difficult to analyze endosomal escape of a protein and there are many unanswered questions regarding the dynamics of this protein once it has gained access into cells. One interesting method employed fluorescent resonance energy transfer (FRET) to monitor the uptake and endosomal escape of the Tat peptide sequence conjugated to fluorescent protein mCherry that formed the acceptor fluorophore for FITC [65]. FRET between these two molecules was quenched by 4-(4′-dimethylaminophenylazo)benzoic acid (Dabcyl) and the fluorescence profile of the construct was dependent on whether it was intact or degraded in endosomes or the cytosol. Although this was a complex system, it offered an opportunity to monitor in real time the fate of delivered proteins fused with penetrating sequences. 15.4.4

Entry into the Nucleus

Passive diffusion should allow cytosol–nucleus exchange for molecules Fs, the motor walks backwards very slowly [25, 57]; we use the parametrization [13] v ( F ) = vb(1 − F Fs ) for Fs ≤ F with vb 1 motors, these N motors can share the force, so that the average cargo velocity remains relatively high even for large friction coefficient φfr; see Figure 16.5B. 16.3.5 Two Teams of Antagonistic Motors As discussed in Section 16.2.2, many cellular cargoes undergo fast transport both toward the plus end and toward the minus end of the microtubules. This bidirectional transport implies that these cargo particles must be attached to both plus-end and minus-end directed motors. 16.3.5.1 Tug-of-War Between Motors We now consider a cargo particle that is attached to N+ plus and N− minus motors. Because each motor unbinds from and rebinds to the filament in a stochastic manner, both the number of active plus motors and the number of active minus motors now fluctuate; see Figure 16.6. Three situations are possible: in the (+) states, only plus motors are active, which can pull the cargo into the plus direction without any opposition from the minus motors; in the (−) states, only minus motors are active and the cargo exhibits fast minus motion; and in the (0) states both types of motors are bound to the filament. In the latter case, the motors pull the cargo into opposite directions, leading to slow cargo movement. Since many bidirectional cargoes are observed to exhibit long periods of fast unidirectional motion, the motors must “cooperate” in some way to avoid the (0) states. Two mechanisms for bidirectional transport have been proposed [29, 30]: (1) coordination by a putative protein complex, which ensures that only one type of motors is active at any given time; and (2) tug-of-war between the motors that pull on each other until the stronger team wins and determines the direction of motion for a certain period of time. Previously, it was thought that a tug-of-war would lead to a prevalence of (0) states with slow cargo motion and thus would be inconsistent with the experimental observation of fast cargo motion [29, 30]. However, we have developed a realistic tug-of-war model, which is based on the single motor description as presented in Section 16.3.3 as well as on the assumptions that (1) each motor team exerts load forces on the other team and (2) motors of one team share the load force generated by the antagonistic motor team [12, 13]. Such a tug-of-war, which

TEAMWORK OF MOLECULAR MOTORS

(–)

(0)

(0)

(–)

(0)

(0)

(0)

(+)

(+)

301

Figure 16.6. Transport by two antagonistic teams of motors. The motors with blue heads are plus motors that pull to the right, the ones with yellow heads are minus motors that pull to the left. Both types of motors are firmly attached to the cargo particle (grey) via their long stalks. The total number of blue plus motors and yellow minus motors is N+ = 2 and N− = 2, respectively. The cargo particle exhibits (N+ + 1)(N− + 1) = 9 different states and undergoes transitions between these states arising from the unbinding and rebinding of single motors. (See color insert.)

does not require any coordination complex, is consistent with all experimental observations [13, 65]. 16.3.5.2 Patterns of Movements As shown in Refs. 12 and 13, a cargo transported by two teams of motors can exhibit different patterns of motility (or “motility states”) depending on the single motor parameters. For the single motor parameters given in Table 16.1, the cargo transport exhibits a dynamic instability, which effectively makes the (0) states very unlikely: if both types of motors are active simultaneously, they exert forces on each other and tend to pull each other off the filament. If, for example, a minus motor unbinds, each of the remaining minus motors has to sustain a larger force arising from the opposing team. Since the unbinding rate in Equation 16.4 increases exponentially with the force, the remaining minus motors now have a high unbinding probability, and undergo an unbinding cascade until all minus motors are detached. A similar unbinding cascade can happen for the plus motors. Because of these unbinding cascades, the cargo is unlikely to stay in a (0) state, in which both types of motors are attached to the filament, but is likely to attain a (+) or (−) state, in which only plus or only minus motors are bound to filament. Therefore the spatial displacement (or trajectory or kymograph) of a cargo pulled by several kinesin-1 and several dynein motors exhibits

302

CARGO TRANSPORT BY TEAMS OF MOLECULAR MOTORS

alternating periods of fast plus and fast minus motion at speeds of micrometers/ second (µm/s); see Figure 16.7A. Switching between the two directions happens when a series of motor binding and unbinding events leads to the predominance of the opposing team, which is then stabilized by another unbinding cascade. 16.3.5.3 Advantages for Cargo Transport The tug-of-war performed by two motor species has several advantages for the cell. First, fast bidirectional motion is possible without any coordination complex. Second, cargo transport can easily be regulated, since a change in a single motor parameter affects the competition of the two teams and can lead to net plus or net minus motion by increasing or decreasing the corresponding run lengths, as found experimentally [33–36]. Third, the transport by teams of motors again leads to larger force generation and to enhanced processivity [66]. In particular, the binding time of a cargo increases exponentially with the numbers N+ and N− of attached plus and minus motors, respectively. The tug-of-war model described here leads to the binding time Δtb as given by Δtb( N +, N − ) ≅

(1 + π 0 + ε 0 + ) N + + (1 + π 0 − ε 0 − ) N − − 2

(16.7)

N+ π0+ + N− π0−

where the indices “+” and “−” label plus and minus motor parameters, respectively. Fourth, the stochastic switching of bidirectional transport leads to enhanced diffusion on long time and large length scales; see Figure 16.7B. On short time scales, the cargo particle moves ballistically, that is, the mean square discplacement increases quadratically with time, while it increases linearly on long time scales, with a diffusion coefficient of about 1 µm2/s. This diffusion constant is similar to the diffusion constant of a 1-µm sized cargo particle in

10

MSD [µm2]

Distance [µm]

15 5 0 −5 0 (A)

10

20 30 Time t [s]

40

104 102 10−2 10−4 0.01

(B)

~t

1 ~t2 1 103 Time t [s]

104

Figure 16.7. Transport by two antagonistic teams of motors. (A) Spatial displacement (or trajectory or kymograph) of cargo particle as a function of time. The cargo is pulled by 4 dyneins and 3 kinesins with single motor parameters as in Table 16.1 and exhibits fast transport in both the plus direction and minus direction. (B) Mean square displacement (MSD) of cargo as a function of time. The cargo is pulled by N+ = 4 plus and N− = 4 minus motors, which are both characterized by kinesin parameters as in Table 16.1. The cargo’s MSD grows quadratically and linearly with time for short and long times, respectively.

TEAMWORK OF MOLECULAR MOTORS

303

water, and about 100–1000 times larger than the diffusion constant of such a particle in the cytoplasm. This enhanced diffusion should be especially useful for cargoes in search of their destination, or for cargo particles such as mitochondria or pigment granules that must be distributed over the whole cell [29, 36, 67]. Finally, one would intuitively expect that bidirectional transport helps to reduce jams in the dense traffic of cargo particles as found in eukaryotic cells. Indeed, if a jam builds up in one direction, for example, because of an obstacle, the cargo particles at the very end of the jam may then move in the opposite direction and, in this way, start to dissolve the jam. 16.3.6

Two Teams of Motors Working on Different Tracks

As discussed in Section 16.2.2, tracks for the long- and short-ranged transport of intracellular cargo particles are provided by microtubules and actin filaments, respectively. In order to switch between these two filament systems without interruptions, many cargo particles are attached to both microtubuleand actin-based motors [38, 39]. The probability of switching between the two filaments depends on the number and types of motors, on the cargo, as well as on cellular regulation [36, 40, 68]. 16.3.6.1 Processivity Enhancement During transport on one type of filament, both types of motors may be crosslinked to this filament. Indeed, myosin V, which walks along actin filaments, can also bind to microtubules and diffuse randomly on these latter filaments, whereas kinesin-1, which walks along microtubules, exhibits a weak affinity for actin filaments as well [49, 69]. This type of transport is illustrated in Figure 16.8A for a cargo particle that is

Cargo Myosin

Kinesin

(A)

Distance [µm]

6

+ Microtubule

4

2

0 (B)

0

3

6

9

12

Time t [s]

Figure 16.8. Cargo transport by one active and one “passive” motor. (A) A cargo particle (grey) is crosslinked to a microtubule by one kinesin-1 (blue) and one myosin V (red). The kinesin-1 motor is actively stepping whereas the myosin V motor is passively diffusing along the filament. (B) The trajectory of such a cargo exhibits fast plus motion interrupted by diffusive events. For this example, the total binding time was about 11 s, much longer than in the absence of the “passive” myosin V motor [14]. (See color insert.)

304

CARGO TRANSPORT BY TEAMS OF MOLECULAR MOTORS

crosslinked to a microtubule by both one kinesin-1 motor, which actively walks along the microtubule, and one myosin V, which passively diffuses along this filament. Such a cargo particle exhibits fast plus motion interrupted by diffusive events; see Figure 16.8B. It is then necessary to distinguish the run length, which the cargo particle exhibits during its plus motion, from the binding length of the composite cargo motion. The binding length of a cargo particle that is transported by one kinesin-1 motor and one myosin V motor is more than two times larger than that of a cargo particle transported by only one kinesin motor [49]. This processivity enhancement can be understood from the single motor properties of kinesin and myosin (compare Section 16.3.3), provided the myosin motor is characterized by its bound diffusion constant rather than by its velocity [14]. The increase in run length arises in a similar way as for cargo transport by a team of identical motors; see Section 16.3.4. If the kinesins unbind, the myosins still act as crosslinkers between cargo and microtubules, thereby preventing the cargo from diffusing away from the filament and giving the kinesins a chance to rebind. Furthermore, in contrast to the competition between kinesin and dynein as described in Section 16.3.5, the simultaneous crosslinking by kinesin-1 and myosin V does not lead to a tug-of-war: because of the relatively weak affinity of myosin V to microtubules, myosin V motors can easily be dragged along by kinesin motors toward the plus end of the microtubule [14].

16.4

CONCLUSION AND DISCUSSION

Cellular cargoes, including endosomes, RNAs, protein complexes, and filaments as well as whole organelles, travel through the cell with the help of molecular motors that pull them along cytoskeletal filaments: kinesin and dynein motors mediate long-ranged transport on microtubule filaments, while local delivery is accomplished by myosin motors walking on a meshwork of actin filaments. 16.4.1 Team Work of Motors As described in this chapter, molecular motors work in teams in order to meet the challenges of intracellular transport. Three types of teamwork have been identified and elucidated. First, one team of identical motors as shown in Figure 16.4 is able to transport large cargoes through the viscous cytoplasm at considerable speeds of micrometers/second over distances of many micrometers; see Figure 16.5. Second, two teams of antagonistic motors that walk into opposite directions, as in Figure 16.6, accomplish bidirectional transport along an isopolar cytoskeletal network of motor tracks as illustrated in Figure 16.7A. Third, two teams of motors that can walk on different filaments (see

CONCLUSION AND DISCUSSION

305

Figure 16.8) allow a smooth transit from the long-range microtubule traffic to the short-range delivery on actin filaments and vice versa. The presence of several motor teams on one cargo allows versatile transport and fast reaction to cellular regulation. A regulation cascade that targets the motor proteins can change the transport properties of the cargo “on the fly,” making it travel on a microtubule to the minus instead of the plus direction by downregulating kinesin, or switching to actin filaments by upregulating myosin. 16.4.2

Virus, Gene, and Drug Delivery

Many viruses hijack the intracellular transport systems and use them for their own purposes. These viruses recruit molecular motors from the host’s cytoplasm and, in addition, “know” how to activate these motors to ensure transport to the desired destination. Copying this viral strategy would dramatically increase the efficiency of gene therapy and drug delivery, because transport through the crowded and viscous cytoplasm would be strongly enhanced. Virus-based expression vectors should include all factors that allow the virus to use the host’s transport machinery during entry and transcription, but not those involved in virus replication and egress. Nonviral vectors and drug carriers should be able to recruit motor proteins from the cytoplasm, for example, by coating them with receptor proteins for molecular motors. Although many of the motor–cargo interactions still remain to be elucidated, an increasing catalog of receptor proteins is emerging [8]. Ideally, the appropriate receptors should be selected from such a catalog in order to bind the correct number and types of motors from the cytoplasm to the gene or drug carrier. The choice of motors should reflect the cargo destination, and possible responses to intracellular or externally applied regulatory stimuli. 16.4.3

Outlook: Open Questions and Possible Applications

In order to construct such carrier and delivery systems, a good understanding of molecular motor traffic is necessary. Although much progress has been made during the last couple of years, many open questions remain. These include the details of the molecular stepping mechanisms of the different motors, the understanding of motor cooperativity, as well as the understanding of overall cellular traffic. A particularly challenging topic is the coupling of motor transport to cellular regulation. Such an understanding of molecular motor traffic would also be useful for other types of applications in medicine and nanotechnology. One example is provided by antiviral therapies: efficiency of virus infection would be greatly reduced if the viruses were prevented from hijacking the cellular transport machinery in an appropriate way. Likewise, understanding of molecular motor traffic could be useful in order to treat diseases in which improper intracellular transport plays an important role, such as Alzheimer’s disease or lissencephaly

306

CARGO TRANSPORT BY TEAMS OF MOLECULAR MOTORS

[70, 71]. Finally, molecular motors are possible building blocks for nanotechnological applications. One example is provided by molecular motors that carry cargoes on lab-on-a-chip devices [72, 73].

REFERENCES 1. Cohen, R. N., Rashkin, M. J., Wen, X., and Szoka, F. C. Molecular motors as drug delivery vehicles. Drug Discov. Today 2: 111–118 (2005). 2. Vaughan, E. E., DeGiulio, J. V., and Dean, D. A. Intracellular trafficking of plasmids for gene therapy: mechanisms of cytoplasmic movement and nuclear import. Curr. Gene Ther. 6: 671–681 (2006). 3. Luby-Phelps, K. Cytoarchitecture and physical properties of cytoplasm: volume, viscosity, diffusion, intracellular surface area. Int. Rev. Cytol. Surv. Cell Biol. 192: 189–221 (2000). 4. Lukacs, G. L., Haggie, P., Seksek, O., Lechardeur, D., Freedman, N., and Verkman, A. S. Size-dependent DNA mobility in cytoplasm and nucleus. J. Biol. Chem. 275: 1625–1629 (2000). 5. Dowty, M. E., Williams, P., Zhang, G. F., Hagstrom, J. E., and Wolf, J. A. Plasmid DNA entry into postmitotic nuclei of primary rat myotubes. Proc. Natl. Acad. Sci. U.S.A. 92: 4572–4576 (1995). 6. Capecchi, M. R. High-efficiency transformation by direct micro-injection of DNA into cultured mammalian-cells. Cell 22: 479–488 (1980). 7. Campbell, E. M. and Hope, T. J. Gene therapy progress and prospects: viral trafficking during infection. Gene Ther. 12: 1353–1359 (2005). 8. Dohner, K., Nagel, C. H., and Sodeik, B. Viral stop-and-go along microtubules: taking a ride with dynein and kinesins. Trends Microbiol. 13: 320–327 (2005). 9. Greber, U. F. and Way, M. A superhighway to virus infection. Cell 124: 741–754 (2006). 10. Beeg, J., Klumpp, S., Dimova, R., Gracià, R. S., Unger, E., and Lipowsky, R. Transport of beads by several kinesin motors. Biophys. J. 94: 532–541 (2008). 11. Klumpp, S. and Lipowsky, R. Cooperative cargo transport by several molecular motors. Proc. Natl. Acad. Sci. U.S.A. 102: 17284–17289 (2005). 12. Müller, M. J. I., Klumpp, S., and Lipowsky, R. Motility states of molecular motors engaged in a stochastic tug-of-war. J. Stat. Phys. 133: 1059–1081 (2008). 13. Müller, M. J. I., Klumpp, S., and Lipowsky, R. Tug-of-war as a cooperative mechanism for bi-directional cargo transport by molecular motor. Proc. Natl. Acad. Sci. U.S.A. 105: 4609–4614 (2008). 14. Berger, F., Müller, M. J. I., and Lipowsky, R. Enhancement of the processivity of kinesin-transported cargo by myosin V. EPL 87: 28002 (2009). 15. Alberts, B., Bray, D., Johnson, A., Lewis, J., Raff, M., Roberts, K., and Walter, P. Essential Cell Biology. An Introduction to the Molecular Biology of the Cell. Garland, New York, 1998. 16. Brady, S. T. A novel brain ATPase with properties expected for the fast axonal transport motor. Nature 317: 73–75 (1985).

REFERENCES

307

17. Vale, R. D., Reese, T. S., and Sheetz, M. P. Identification of a novel force-generating protein, kinesin, involved in microtubule-based motility. Cell 42: 39–50 (1985). 18. Vale, R. D. The molecular motor toolbox for intracellular transport. Cell 112: 467–480 (2003). 19. Block, S. M. Kinesin motor mechanics: binding, stepping, tracking, gating, and limping. Biophys. J. 92: 2986 (2007). 20. Paschal, B. M., Sheptner, H. S., and Vallee, R. B. MAP-1C is a microtubule- activated ATPase which translocates microtubules invitro and has dynein-like properties. J. Cell Biol. 1057: 1273–1282 (1987). 21. Gennerich, A. and Vale, R. D. Walking the walk: how kinesin and dynein coordinate their steps. Curr. Opin. Cell Biol. 21: 59–67 (2009). 22. Howard, J. Mechanics of Motor Proteins and the Cytoskeleton. Sinauer Associates, Sunderland, MA, 2001. 23. Mallik, R. and Gross, S. P. Molecular motors: strategies to get along. Curr. Biol. 14: R971–R982 (2004). 24. Goldstein, L. S. B. and Yang, Z. Microtubule-based transport systems in neurons: the roles of kinesins and dyneins. Annu. Rev. Neurosci. 23: 39–71 (2000). 25. Carter, N. J. and Cross, R. A. Mechanics of the kinesin step. Nature 435: 308–312 (2005). 26. Toba, S., Watanabe, T. M., Yamaguchi-Okimoto, L., Toyoshima, Y. Y., and Higuchi, H. Overlapping hand-over-hand mechanism of single molecular motility of cytoplasmic dynein. Proc. Natl. Acad. Sci. U.S.A. 103: 5741–5745 (2006). 27. Gross, S. P., Vershinin, M., and Shubeita, G. T. Cargo transport: two motors are sometimes better than one. Curr. Biol. 17: R478–R486 (2007). 28. Lane, J. and Allan, V. Microtubule-based membrane movement. Biochim. Biophys. Acta 1376: 27–55 (1998). 29. Gross, S. P. Hither and yon: a review of bi-directional microtubule-based transport. Phys. Biol. 1: R1–R11 (2004). 30. Welte, M. A. Bi-directional transport along microtubules. Curr. Biol. 14: R525– R537 (2004). 31. Klumpp, S., Nieuwenhuizen, Th. M., and Lipowsky, R. Self-organized density patterns of molecular motors in arrays of cytoskeletal filaments. Biophys. J. 88: 3118– 3132 (2005). 32. Müller, M. J. I., Klumpp, S., and Lipowsky, R. Molecular motor traffic in a halfopen tube. J. Phys. Condens. Matter 17: S3839–S3850 (2005). 33. Morris, N. R. and Hollenbeck, P. J. The regulation of bi-directional mitochondrial transport is coordinated with axonal outgrowth. J. Cell Sci. 104: 917–927 (1993). 34. Gross, S. P., Welte, M. A., Block, S. M., and Wieschaus, E. F. Dynein-mediated cargo transport in vivo: a switch controls travel distance. J. Cell Biol. 148: 945–955 (2000). 35. Smith, G. A., Murphy, B. J., Gross, S. P., and Enquist, L. W. Local modulation of plus-end transport targets herpesvirus enry and egress in sensory axons. Proc. Natl. Acad. Sci. U.S.A. 101: 16034–16039 (2004). 36. Gross, S. P., Tuma, M. C., Deacon, S. W., Serpinskaya, A. S., Reilein, A. R., and Gelfand, V. I. Interactions and regulation of molecular motors in Xenopus melanophores. J. Cell Biol. 156: 855–865 (2002).

308

CARGO TRANSPORT BY TEAMS OF MOLECULAR MOTORS

37. Tuma, M. C. and Gelfand, V. I. Molecular mechanisms of pigment transport in melanophores. Pigment Cell Res. 12: 283–294 (1999). 38. Brown, S. S. Cooperation between microtubule- and actin-based motor proteins. Annu. Rev. Cell Dev. Biol. 15: 63–80 (1999). 39. Goode, B. L., Drubin, D. G., and Barnes, G. Functional cooperation between the microtubule and actin cytoskeletons. Curr. Opin. Cell Biol. 12: 63–71 (2000). 40. Snider, J., Lin, F., Zahedi, N., Rodionov, V., Yu, C. C., and Gross, S. P. Intracellular actin-based transport: How far you go depends on how often you switch. Proc. Natl. Acad. Sci. U.S.A. 101: 13204–13209 (2004). 41. Lipowsky, R. and Klumpp, S. “Life is motion”—multiscale motility of molecular motors. Physica A 352: 53–112 (2005). 42. Liepelt, S. and Lipowsky, R. Kinesin’s network of chemomechanical motor cycles. Phys. Rev. Lett. 98: 258102 (2007). 43. Lipowsky, R., Klumpp, S., and Nieuwenhuizen, Th. M. Random walks of cytoskeletal motors in open and closed compartments. Phys. Rev. Lett. 87: 108101 (2001). 44. Maly, I. V. A stochastic model for patterning of the cytoplasm by the salutatory movement. J. Theor. Biol. 216: 59–71 (2002). 45. Parmeggiani, A., Franosch, T., and Frey, E. Phase coexistence in driven one dimensional transport. Phys. Rev. Lett. 90: 086601 (2003). 46. Smith, D. A. and Simmons, R. M. Models of motor-assisted transport of intracellular particles. Biophys. J. 80: 45–68 (2001). 47. Kuznetsov, A. V., Avramenko, A. A., and Blinov, D. G. Numerical modeling of molecular-motor-assisted transport of adenoviral vectors in a spherical cell. Comput. Methods Biomech. Biomed. Eng. 11: 215–222 (2008). 48. Lagache, T., Dauty, E., and Holcman, D. Quantitative analysis of virus and plasmid trafficking in cells. Phys. Rev. E 79: 011921 (2009). 49. Ali, M. Y., Lu, H., Bookwalter, C. S., Warshaw, D. M., and Trybus, K. M. Myosin V and kinesin act as tethers to enhance each others’ processivity. Proc. Natl. Acad. Sci. U.S.A. 105: 4691–4696 (2008). 50. Campàs, O., Leduc, C., Bassereau, P., Casademunt, J., Joanny, J.-F., and Prost, J. Coordination of kinesin motors pulling on fluid membranes. Biophys. J. 94: 5009– 5017 (2008). 51. Diehl, M. R., Zhang, K., Lee, H. J., and Tirrell, D. A. Engineering cooperativity in biomotor protein assembly. Science 311: 1468–1471 (2006). 52. Leduc, C., Campàs, O., Zeldovich, K. B., Roux, A., Jolimaitre, P., Bourel, L., Bonnet, B. G., Joanny, J.-F., Bassereau, P., and Prost, J. Cooperative extraction of membrane nanotubes by molecular motors. Proc. Natl. Acad. Sci. U.S.A. 101: 17096–17101 (2004). 53. Leduc, C., Ruhnow, F., Howard, J., and Diez, S. Detection of fractional steps in cargo movement by the collective operation of kinesin-1 motors. Proc. Natl. Acad. Sci. U.S.A. 104: 10847–10852 (2007). 54. Vershinin, M., Carter, B. C., Razafsky, D. S., King, S. J., and Gross, S. P. Multiplemotor based transport and its regulation by Tau. Proc. Natl. Acad. Sci. U.S.A. 104: 87–92 (2007). 55. Badoual, M., Jülicher, F., and Prost, J. Bi-directional cooperative motion of molecular motors. Proc. Natl. Acad. Sci. U.S.A. 99: 6696–6701 (2002).

REFERENCES

309

56. Kunwar, A., Vershinin, M., Xu, J., and Gross, S. P. Stepping, strain gating, and an unexpected force-velocity curve for multiple-motor-based transport. Curr. Biol. 18: 1–11 (2008). 57. Nishiyama, M., Higuchi, H., and Yanagida, T. Chemomechanical coupling of the forward and backward steps of single kinesin molecules. Nat. Cell Biol. 4: 790–797 (2002). 58. Visscher, K., Schnitzer, M. J., and Block, S. M. Single kinesin molecules studied with a molecular force clamp. Nature 400: 184–189 (1999). 59. Schnitzer, M. J., Visscher, K., and Block, S. M. Force production by single kinesin motors. Nat. Cell Biol. 2: 718–723 (2000). 60. King, S. J. and Schroer, T. A. Dynactin increases the processivity of the cytoplasmic dynein motor. Nat. Cell Biol. 2: 20–24 (2000). 61. Kramers, H. A. Brownian motion in a field of force and the diffusion model of chemical reactions. Physica 7: 284–304 (1940). 62. Bell, G. I. Models for the specific adhesion of cells to cells. Science 200: 618–627 (1978). 63. Block, S. M., Goldstein, L. S. B., and Schnapp, B. J. Bead movement by single kinesin molecules studed with optical tweezers. Nature 348: 345–352 (1990). 64. Coy, D. L., Wagenbach, M., and Howard, J. Kinesin takes one 8-nm step for each ATP that it hydrolyzes. J. Biol. Chem. 274: 3667–3671 (1999). 65. Welte, M. A. and Gross, S. P. Molecular motors: a trafifc cop within? HFSP J. 2: 178–182 (2008). 66. Müller, M. J. I., Klumpp, S., and Lipowsky, R. Dynamics of bi-directional transport by molecular motors. Biophys. J. 98: 2610–2618 (2010). 67. De Vos, K. J., Sable, J., Miller, K. E., and Sheetz, M. P. Expression of phosphatidylinositol(4,5) bisphosphate-specific pleckstrin homology domains alters direction but not the level of axonal transport of mitochondria. Mol. Biol. Cell 14: 3636–3649 (2003). 68. Ross, J. L., Ali, M. Y., and Warshaw, D. M. Cargo transport: molecular motors navigate a complex cytoskeleton. Curr. Opin. Cell Biol. 20: 41–47 (2008). 69. Ali, M. Y., Krementsova, E. B., Kennedy, G. G., Mahafy, R., Pollard, T. D., Trybus, K. M., and Warshaw, D. M. Myosin Va maneuvers through actin intersections and diffuses along microtubules. Proc. Natl. Acad. Sci. U.S.A. 104: 4332–4336 (2007). 70. Aridor, M. and Hannan, L. A. Traffic jams II: an update of diseases of intracellular transport. Traffic 3: 781–790 (2002). 71. Duncan, J. E. and Goldstein, L. S. The genetics of axonal transport and axonal transport disorders. PLoS Genet. 2: e124 (2006). 72. Hess, H., Bachand, G. D., and Vogel, V. Powering nanodevices with biomolecular motors. Chem. Eur. J. 10: 2110–2116 (2004). 73. van den Heuvel, M. G. L. and Dekker, C. Motor proteins at work for nanotechnology. Science 317: 333–336 (2007).

CHAPTER 17

The Potential of Photochemical Internalization (PCI) for the Cytosolic Delivery of Nanomedicines KRISTIAN BERG and ANETTE WEYERGANG Department of Radiation Biology, Institute for Cancer Research, Norwegian Radium Hospital, Oslo University Hospital, Oslo, Norway

ANDERS HØGSET PCI Biotech AS, Oslo, Norway

PÅL KRISTIAN SELBO Department of Radiation Biology, Institute for Cancer Research, Norwegian Radium Hospital, Oslo University Hospital, Oslo, Norway

17.1

INTRODUCTION

Utilization of macromolecules in the therapy of cancer and other diseases is becoming increasingly important. Recent advances in molecular biology and biotechnology have made it possible to improve targeting and design of cytotoxic agents or DNA complexes for clinical applications. Macromolecules with therapeutic potential include proteins such as ribosome-inactivating protein toxins for treatment of cancer and other indications, antibodies and growth factors for cell surface targeting, peptides and mRNA for vaccination, DNA utilizing nonviral and viral vectors for gene therapy, and oligonucleotides (antisense oligonucleotides, ribozymes, peptide nucleic acids (PNAs), and siRNA for gene silencing) and nanoparticles for drug delivery [1]. There are many extracellular and intracellular barriers for these molecules to overcome before they can arrive at the target cells, enter the cell, and reach intracellular therapeutic targets. Degradation by serum enzymes and elimination by cells of the reticuloendothelial system (RES), penetration into the target tissues Organelle-Specific Pharmaceutical Nanotechnology, Edited by Volkmar Weissig and Gerard G. M. D’Souza Copyright © 2010 John Wiley & Sons, Inc.

311

312

THE POTENTIAL OF PHOTOCHEMICAL INTERNALIZATION (PCI)

through the endothelial lining, as well as transport limitations within the tissue are important hurdles to obtain sufficient biological effect of these macromolecules [2]. New delivery systems have improved the cellular uptake of macromolecules, but tissue penetration, cellular uptake, and efficient transfer of the molecules into the cytosol of the target cells through the plasma membrane or the membranes of the endocytic vesicles are still fundamental obstacles. In many cases the targets of macromolecular therapeutics are intracellular. However, the limited release from and degradation of macromolecules in endocytic vesicles after uptake by endocytosis are still major intracellular barriers for the therapeutic application of macromolecules having intracellular targets of action. These limitations may also in many cases cause suboptimal therapeutic effect of nanoparticle-based therapeutics. Photochemical internalization (PCI) is a novel technology for release of endocytosed macromolecules into the cytosol [3]. The technology is based on the use of photosensitizers located in endocytic vesicles that upon activation by light induce a release of macromolecules from their compartmentalization in endocytic vesicles. PCI has been shown to potentiate the biological activity of a large variety of macromolecules and other molecules that do not readily penetrate cellular membranes. PCI is a technology derived from photodynamic therapy (PDT). Thus the basis for PDT will be described as well as the basic principles and utilization of PCI for therapeutic purposes.

17.2

PHOTODYNAMIC THERAPY

PDT is a treatment modality that takes advantage of the phototoxic effects induced by a photosensitizer and light in the presence of oxygen [4]. PDT is approved for several cancer indications, as well as for age-related macular degeneration. In addition, fluorescence diagnosis (FD) based on the fluorescing properties of a PDT-type photosensitizer has recently been approved for detection of bladder dysplasia and cancer [5]. 17.2.1 The Photosensitizer A photosensitizer is defined as a chemical entity that, upon absorption of light, induces a chemical or physical alteration of another chemical entity. Most photosensitizers and all clinically approved photosensitizers (with the exception of methylene blue) used in photodynamic therapy (PDT) are based on or related to the tetrapyrrole macrocycle [6]. Porphyrins consist of four pyrrole subunits linked together by four methine bridges as shown in Figure 17.1. Porphyrins and porphyrin-related dyes used in PDT have substituents in the peripheral positions of the pyrrole rings, on the four methine carbons (mesopositions) and/or coordinated metals. These derivatives are synthesized to influence the water/lipid solubility, amphiphilicity, pKa, and stability of the compounds [7]. These parameters determine the biodistribution of the compounds, that is, the intracellular localization, tissue distribution, and

PHOTODYNAMIC THERAPY

N N

Porphin

N N

Chlorin

N N

Bacteriochlorin

O

O

O

Soret

Q-bands Phthalocyanines Bacteriochlorins

N

O O

Naphthalocyanines

N N Absorbance

N N

N N

O

O

N

N N

Phthalocyanine

IV III II I

N N N N

O

313

400

Chlorins

500 600 700 800 Wavelength (nm)

Naphthalocyanine

Figure 17.1. Basic structure (left side) and absorption spectra (right side) of some photosensitizers. A full spectrum is shown for a porphyrin-type photosensitizer and additionally the locations of the main peaks for photosensitizers developed for PDT are shown on the figure (not to scale).

pharmacokinetics. The photosensitizer must absorb the wavelengths emitted by the light source in order to induce a treatment effect. Clinically, 6–800-nm light is mainly used due to the optimal tissue penetration properties at these wavelengths. Porphyrins have low absorption of light in this wavelength region. Thus the porphyrin structures have been modified in order to enhance the absorption properties of the photosensitizers (Figure 17.1). Necrotic depth after PDT of solid tumors is typically 2–10 mm depending on the photosensitizer in use. 17.2.2

Mechanisms of PDT

The photosensitizer is usually administered systemically by means of intravenous injection. There is substantial evidence for a preferential retention of the photosensitizers in neoplastic lesions, usually with a tumour : (normal surrounding tissue) ratio of 2–3 : 1 [8]. At some time interval after photosensitizer administration, the target tissue is illuminated with light of an appropriate wavelength. The light source is often a laser (e.g., a diode laser), but noncollimated light sources may also be used. The photosensitizer must absorb the wavelengths emitted by the light source in order to induce a treatment effect (Figure 17.1). The cytotoxic effect is mediated mainly through the formation of singlet oxygen (1O2). This reactive intermediate has a very short lifetime in cells (logP>0; Z>0; 6>AI>3.5

Z>0; 0>logP>−5 pKa0; 5>logP>0

For non-specific uptake into biomembranes: 8>logP>5

0>logP>−5; Z>0; pKa>10... Mitochondrion

DNA

Cell membrane

Figure 21.2. Richard Horobin’s QSAR decision rules for predicting the intracellular distribution of low molecular weight compounds (for abbreviations see Table 21.1). This image was graciously provided by Richard W. Horobin.

selectively accumulate at or inside mammalian mitochondria within living cells [19]. The authors generated from the literature a nonbiased sample of more than 100 so-called mitochondriotropic compounds and examined this data set using physicochemical classifications, quantitative structure–activity relationship (QSAR) models, and a Fick–Nernst–Planck physicochemical model. The ability of the latter two approaches to predict mitochondriotropic behavior was assessed, and comparisons were made between methods and with current assumptions. All approaches provided instructive pictures of the nature of mitochondriotropics. Most interestingly, however, although delocalized lipophilic cations have been regarded as the most common structural type of mitochondriotropic molecules [20–22], only a third were such. Much the same proportion were acids, potentially or actually anions. Many mitochondriotropics were electrically neutral compounds. From Table 21.2 it can be seen that selective mitochondrial accumulation involves electric potential, ion-trapping, and complex formation with cardiolipin, while nonspecific accumulation involves membrane partitioning and nonspecific access requires only low lipophilicity. Using QSAR and the Fick– Nernst approaches the authors were able to predict the mitochondriotropic behavior of more than 80% of the selected data set and allowed them to specify in some detail the physicochemistry of mitochondriotropic molecules. Overall, this approach is expected to facilitate guided syntheses and selection

INTRACELLULAR BARRIERS

389

TABLE 21.2 Physicochemical Features Favoring Targeting or Accumulation of Xenobiotics in Mitochondria within Intact Cells

Accumulation & targeting processes

Selectivity

Cell line applicability

Physicochemical thumbnails, including parameter ranges (see also footnote)

Compounds with access to, but not necessarily accumulating in, mitochondria Membrane No selectivity All cell lines Lipophilic but not permeance superlipophilic: 8 > log P > 0 Compounds accumulating in mitochondria Reversible Low: uptake All cell lines Strongly but not partitioning into into all super-lipophilic: membranes biomembranes 8 > log P > 5 OR Strongly but not super-amphiphilic: 8 > AI > 5 Ion-trapping of Low All cell lines Weak acids with lipophilic weak lipophilic less-ionized acids within species: Z ≤ 0; log mitochondria Pless ionized > 0; pKa = 7 ± 3 [lysosomes & nuclei may also show uptake] Precipitation by High All cell lines Lipophilic to strongly cardiolipin & lipophilic cations: attraction by Z > 0; 5 > log P > 0 electrical potential Attraction by high High Cells with higher Moderately hydrophilic mitochondrial mitochondrial to moderately membrane membrane lipophilic cations: potential potential Z > 0; −2 > log P > 2 (Uptake by normal cell nuclei if log P < 0) Footnote: few pharmaceuticals have conjugated systems large enough, or are sufficiently amphiphilic or lipophilic, to be trapped in the plasmalemma. However, note that all above specifications have implied requirements for non-trapping by the plasmalemma, namely: AI < 8; CBN < 40; log < 8.

390

MITOCHONDRIA-TARGETED PHARMACEUTICAL NANOCARRIERS

of optimal mitochondriotropic structures. Moreover, the strategy developed by Horobin and Weissig [23] should generally be applicable to the future design of low molecular weight drugs aimed at acting on or inside all other cell organelles and has already proved useful for the modeling of cationic transfection lipids [23].

21.3

MITOCHONDRIA-TARGETED NANOCARRIERS

21.3.1 DQAsomes for Drug and DNA Delivery to Mitochondria Based on their resemblance to an old South American hunting weapon, a group of symmetric amphiphilic molecules, in which two hydrophilic residues are linked by a hydrophobic chain, have become widely known as “bolalipids.” Well-characterized bola-amphiphiles are archaebacterial lipids consisting of two glycerol backbones, which are connected by two hydrophobic chains. The self-assembly behavior of these archaeal lipids has thoroughly been investigated. It could be shown that such archaeal bola-lipids are able to self-associate into mechanically stable monolayer membranes [24]. While screening mitochondriotropic drugs potentially able to interfere with the mitochondrial DNA metabolism in Plasmodium falciparum [25], a serendipitous discovery revealed the tendency of dequalinium chloride, a bolaamphiphile (Figure 21.3A), to self-associate into colloidal structures. Electron microscopy (Figure 21.3C, right panel) as well as photon correlation spectroscopy confirmed the formation of particles with a diameter between about 70 and 700 nm. Freeze fracture electron microscopic images (Figure 21.3C, right panel C) showed both convex and concave fracture faces, thereby demonstrating the liposome-like aggregation of dequalinium. At the time of their discovery, these unusual vesicles were termed DQAsomes (pronounced dequasomes), that is, dequalinium-based liposome-like vesicles [26]. One striking structural difference between dequalinium and archaeal lipids, however, involves the number of hydrophobic chains between both polar head groups. In contrast to archaeal lipids, dequalinium possesses only one alkyl chain connecting both hydrophilic head groups. Therefore, theoretically, two different conformations within a self-assembled layer structure are imaginable. While the stretched conformation would give rise to the formation of a monolayer, assuming the horseshoe conformation would result in the formation of a bilayer (Figure 21.3B). Just like cationic liposomes, which are widely being explored as nucleartargeted nonviral transfection vectors [27], also DQAsomes were found to form complexes with DNA following mixing the positively charged vesicles with the negatively charged polynucleotide acid. This observation that DQAsomes, that is, cationic vesicles entirely composed of mitochondriotropic molecules, are able to efficiently bind DNA [26] and to protect the DNA from nuclease digestion [28], led to the first carrier-based strategy ever proposed

MITOCHONDRIA-TARGETED NANOCARRIERS

391

(a) CH3 NH2

CH3

N – CH2 – CH2 – CH2 – CH2 – CH2 – CH2 – CH2 – CH2 – CH2 – CH2 – N

NH2

Cl

Cl

(b)

OR

(c)

Figure 21.3. (A) Chemical structure of dequalinium chloride, with overlaid gray shades indicating the bola-like nature of the molecule. (B) Theoretical possible conformations of dequalinium chloride, that is, stretched versus horseshoe conformation, leading to either a monolayer or a bilayer membranous structure following the process of selfassembly. (C) Left panel: Monte Carlo computer simulations demonstrate the possible self-assembly of dequalinium chloride into vesicles. The first image represents a transverse section of the second image, which represents a complete spherical vesicle. (D) Right panel: Electron micrographs of vesicles (“DQAsomes”) made from dequalinium chloride; from left to right: negatively stained transmission electron micrograph, rotary shadowed transmission electron micrograph, and freeze fracture scanning electron micrograph. (Reproduced with permission from Ref. 67).

for direct mitochondrial gene therapy [20–22]. This approach requires the transport of a DNA-mitochondrial leader sequence (MLS) peptide conjugate to mitochondria using DQAsomes, the liberation of this conjugate from the cationic vector upon contact with the mitochondrial outer membrane, followed by DNA uptake via the mitochondrial protein import machinery. DQAsome/DNA complexes (“DQAplexes”) were shown to release their DNA cargo only upon contact with mitochondrial membranes [29, 30]. Furthermore, utilizing a novel protocol for selectively staining free pDNA inside the cytosol, it was demonstrated that DQAsomes selectively deliver pDNA to and release the pDNA exclusively at the site of mitochondria in living mammalian cells [31]. Finally, utilizing confocal fluorescence

392

MITOCHONDRIA-TARGETED PHARMACEUTICAL NANOCARRIERS

microscopy and using DNA–MLS peptide conjugates, it was shown that DQAsomes not only deliver oligonucleotides but are also able to deliver plasmid-sized DNA into mitochondria within living mammalian cells [32]. In addition to having been established as the first mitochondria-specific DNA delivery system, DQAsomes have also been explored as a mitochondriatargeted nanocarrier system for small drug molecules, in particular, for anticancer drugs known to trigger apoptosis via direct action on mitochondria [33, 34]. For these studies, paclitaxel has been chosen as a model drug. It is a wellknown antitumor agent used in the treatment of several cancers. Clinically, the therapeutic potential of paclitaxel is limited due to a very narrow span between the maximal tolerated dose and intolerable toxic level. In addition, its poor aqueous solubility requires the use of emulsion formulations containing Cremophor EL, a surfactant of considerable toxicity in itself [35]. Paclitaxel is largely believed to exert its action by stabilizing the microtubules of cells, but the precise mechanism of paclitaxel-induced apoptosis remains unclear [36, 37]. It has also become evident that paclitaxel has other targets inside the cell, most notable of which is the mitochondrial network. For example, it was shown that paclitaxel could act directly on isolated mitochondria from human neuroblastoma cells to induce the permeability transition pore (PTP)-dependent release of mitochondrial cytochrome c [38]. Very interestingly from the perspective of subcellular drug targeting, a 24-h delay has been observed between the paclitaxel-triggered release of cytochrome c in intact cells versus a cell-free system [39]. This delay has been attributed to the existence of several drug targets inside the cell, making only a subset of the drug molecules available for mitochondria [39]. Consequently, in order to increase the apoptotic activity of paclitaxel, its subcellular (i.e., mitochondrial) bioavailability would have to be increased. Hence paclitaxel appears to be a molecule whose action may be significantly improved by specific subcellular delivery to the mitochondrion. This premise has been perfectly confirmed. First, it was shown that DQAsomal encapsulation of paclitaxel changes the subcellular distribution of a labeled derivative of the drug. Confocal fluorescence microscopic images demonstrate that, in contrast to the free paclitaxel, the DQAsomal encapsulated drug at least partially colocalizes with mitochondria [34]. Subsequently, the metabolic cytotoxicity was compared with the apoptotic activity. As anticipated, it was found that encapsulation of the drug into DQAsomes did not significantly alter the dosedependent metabolic toxicity of paclitaxel, but it clearly improved the proapoptotic action [34]. Figure 21.4A shows fluorescence micrographs of the nuclear morphology of colo205 colon cancer cells treated with different preparations. The quantitative analysis of apoptotic nuclei detected by this assay shown in Figure 21.4B revealed a significant increase (p < 0.05) in apoptotic nuclei in cells treated with DQAsomal paclitaxel over free drug or over paclitaxel mixed with empty DQAsomes. The results of the nuclear morphology assay were further confirmed by a DNA fragmentation assay, which suggested that 10-nM DQAsomal encapsu-

MITOCHONDRIA-TARGETED NANOCARRIERS

(A)

c

(B) b

d

% Apoptotic Nuclei

a

393

25 20 15 10 5 0 20 nm DQA

20 nM DQA + 10 20 nm DQA w/10 nm paclitaxel nm encapsulated paclitaxel (mixed)

Figure 21.4. Nuclear morphology assay for determining apoptosis. (A) Representative fluorescence micrographs showing nuclear morphology of Hoechst-stained, colo205 colon cancer cells treated with (a) negative control, (b) 20 nM DQAsomes, (c) 10 nM free paclitaxel + 20 nM DQAsomes, and (d) DQAsomal-encapsulated paclitaxel (10 nM free paclitaxel + 20 nM DQAsomes) for 20 h. (B) Quantitative estimation of apoptotic nuclei based on 400 cells counted for each group. Inset: Representative image showing normal (open arrow) and apoptotic nuclei (solid arrow). (Reproduced with permission from Ref 34).

lated paclitaxel was comparable to 50-nM free paclitaxel in inducing DNA fragmentation characteristic of apoptosis: that is, DQAsomal encapsulation increased the apoptotic activity of paclitaxel approximately five times [34]. Three years earlier the same group had tested paclitaxel-loaded DQAsomes for their ability to inhibit the growth of human colon cancer tumors in nude mice and their data suggested that encapsulation of paclitaxel in DQAsomes leads to improved in vivo efficacy [33]. Taken together, these data demonstrate that DQAsomes are able to change the intracellular distribution of drugs and that the specific targeting of a drug to an appropriate subcellular target can potentiate the desired drug activity. The antitumor efficiency of DQAsomal encapsulated paclitaxel was most recently further enhanced by modifying the DQAsomal surface with folic acid (FA) [40]. The folate receptor is a folate high-affinity membrane binding protein, which is overexpressed in a large variety of human tumors [41–43]. FA conjugates have been shown to be internalized in a tumor cell specific manner by receptor-mediated endocytosis, resulting in an increased toxicity of the corresponding drug [44–46]. Vaidya and colleagues [40] have incubated EDC activated FA-PEG-COOH with preformed paclitaxel-loaded DQAsomes at different molar ratios resulting in DQAsomes in which up to 5.3% of all dequalinium molecules were conjugated to FA-PEG-COOH. Cell cytotoxicity studies using folate receptor expressing HeLa cells demonstrated that folic acid conjugated DQAsomes possess better antitumor activity as compared to plain paclitaxel-loaded DQAsomes, folic acid conjugated paclitaxel-loaded liposomes, and the free drug. The authors concluded from their data that folic

394

MITOCHONDRIA-TARGETED PHARMACEUTICAL NANOCARRIERS

acid conjugated DQAsomes delivered the drug not only to the cytosol but also to mitochondria, whereas folic acid conjugated liposomes delivered the drug into the cytosol only [40]. 21.3.2 Mitochondria-Targeted Polyethylene Imine A mitochondrial leader peptide (MLP), derived from the cytosolically expressed but mitochondrially localized ornithin transcarbamylase, was recently used to render polyethylene imine (PEI) mitochondriotropic [47]. PEI had been developed in the mid-1990s as a versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo [48, 49]. Lee et al. [47] conjugated the mitochondrial leader peptide to PEI via a disulfide bond and confirmed the complex formation of PEI-MLP with DNA by a gel retardation assay. In vitro delivery tests into living cells performed with rhodamin-labeled DNA demonstrated that PEI-MLP/DNA complexes were localized at mitochondrial sites in contrast to controls carried out with PEI/DNA complexes lacking MLP. The author’s data suggest that PEI-MLP can deliver DNA to the mitochondrial sites and may be useful for the development of “direct mitochondrial gene therapy,” a strategy for the cure of mitochondrial DNA diseases originally proposed by Seibel et al. [50] and by Weissig and Torchilin [20–22] as an alternative to allotopic expression. 21.3.3 Mitochondriotropic Liposomes To render liposomes mitochondria specific, Weissig’s group has modified the liposomal surface with triphenyl phosphonium (TPP) cations [51]. Methyltriphenylphosphonium (MTPP) cations were shown 40 years ago to be rapidly taken up by mitochondria in living cells [52] and Murphy’s group has extensively explored the marked mitochondriotropism of MTPP for the delivery of biological active molecules to and into mitochondria [53–56]. Weissig and co-workers replaced the methyl group in MTPP with a stearyl residue in order to moor mitochondriotropic and hydrophilic TPP cations to the liposomal surface. The intracellular distribution of such TPP surface-coated liposomes (stearyl triphenylphosphonium liposomes, STPP liposomes) was analyzed by confocal fluorescence microscopy using rhodamine-labeled phosphatidylethanolamine (Rh-PE) as a liposomal fluorescence marker [57]. For control, liposomes with incorporated positively charged yet nonmitochondriotropic dioleoyl trimethyl ammonium propane (DOTAP) were used. Both preparations (i.e., the control liposomes and STPP liposomes) displayed the same positive surface charge. Not unexpectedly, both types of surface-linked cations were found to enhance the association of liposomes with cells in an almost identical manner, as was shown by flow cytometric analysis. However, the intracellular distributions of STPP liposomes and DOTAP liposomes were quite different. While confocal fluorescence microscopy revealed an almost complete association of STPP liposomes with mitochondria, DOTAP liposomes displayed significantly less

MITOCHONDRIA-TARGETED NANOCARRIERS

395

mitochondrial association. The authors concluded from their study that a positive surface charge of liposomes enhances cell association per se, while for directing liposomes to specific intracellular target sites an appropriate organelle-specific ligand on the surface of liposomes is essential. Following the demonstration that STPP liposomes deliver a lipophilic fluorescence marker (i.e., Rh-PE) almost exclusively to mitochondria, the authors then investigated whether increased drug concentrations at the site of mitochondria would also result in increased drug efficiency. To this end, ceramide was incorporated into STPP liposomes. Ceramide mediates a large variety of intracellular biological responses to extracellular stimuli [58–60]. It has also been found that anticancer drugs can cause an increase of ceramide concentrations in the vicinity of mitochondria [61] and it has been hypothesized that such increase enables the formation of ceramide channels in the mitochondrial outer membrane [62, 63]. Boddapati et al. [57] incorporated ceramide into STPP liposomes and assessed the apoptotic activity of such formulations in vitro and in vivo. It was found that at the low drug concentration used only the mitochondria-targeted ceramide was able to elicit a strong apoptotic response. Subsequent in vivo studies were carried out with polyethylene glycol bearing STPP liposomes. Interestingly, it was revealed that the cationic TPP ligand did not significantly change the biodistribution of STPP-PEG5000 liposomes in comparison to conventional charge-neutral PEG liposomes. Even more important, the tumor accumulation of STPP-PEG liposomes was almost identical to their noncharged counterparts. Mice were inoculated subcutaneously with mouse mammary carcinoma cells and upon formation of palpable tumors, the animals were divided into three groups for treatment with either buffer or empty or ceramide-loaded STPP-PEG liposomes. Figure 21.5A shows the tumor volumes measured during the tumor growth inhibition study. In both control groups, half of the animals developed necrotic morbidity after 12 days and had to be euthanized. Strikingly, none of the animals treated with mitochondria-targeted ceramide showed any morbidity even after 18 days. Statistical analysis of tumor growth rate at the 12-day time point (n = 6) showed that the treatment with ceramide in STPP-PEG liposomes significantly inhibited tumor growth rate compared to sham treatment (Figure 21.5B). It should be emphasized that in Weissig’s study a significant tumor growth inhibition was achieved with ceramide doses as low as 6 mg/kg. For comparison, administering ceramide formulated in conventional (i.e., nontargeted) liposomes, Stover et al. [64] were able to observe tumor growth reduction only at ceramide doses at least six times that high (i.e., equal or above 36 mg/kg). 21.3.4 The MITO-Porter Concept: Mitochondrial Delivery Via Membrane Fusion A unique liposome-based nanocarrier system able to deliver its cargo to the mitochondrial interior via membrane fusion has been developed by Harashima’s group [65, 66]. Using fluorescence resonance energy transfer (FRET) analysis,

396

MITOCHONDRIA-TARGETED PHARMACEUTICAL NANOCARRIERS

(A)

Tumor Volume (mm3)

1000 800 600 400 200

Da y − Da 2 y Da 0 y Da 2 y Da 4 y Da 6 y Da 8 y Da 10 y Da 12 y Da 14 y Da 16 y 18

0

Tumor Volume Increase (mm3/day)

(B)

70 60 50 40 30 *

20 10 0 Sham

Empty Nanocarrier

Ceramide in Nanocarrier

Figure 21.5. Tumor growth inhibition. (A) Tumor volume (mm3) measured over time period of treatment in BALB/c mice bearing murine 4T1 mammary carcinoma tumors (n = 6); after treatment with buffer (䊏), empty STPP nanocarrier (䉱), and ceramide in STPP nanocarrier (䉬). (B) Tumor growth (in mm3/day) at day 12. (n = 6, error bars denote standard deviation); (asterisk indicates a Student t test p-value of 10 nm. The ultrastructure of cardiac myocytes suggests the existence of intracellular barriers significantly restricting the diffusion of nano-objects toward functionally important cellular compartments and structures such as the nucleus, mitochondria, and the space between the junctional sarcoplasmic reticulum (jSR) and the T tubules (so-called junctional cleft) (Figure 23.1C). Figure 23.2A shows the primary targets and possible barriers and pathways for nano-objects within a cardiac myocyte. Targeting the nucleus and mitochondria is important because of their role in protein expression and cell energetics. The TATS sarcolemma contains numerous transporting structures and receptors crucial for cardiac function. The structural organization of the space around the jSR is a matter of special interest because of its direct involvement in excitation–contraction coupling [8]. Electron micrographs show how tightly the intracellular microstructures are packed

STRUCTURE OF VENTRICULAR CELL

435

(a)

(b)

(c)

Figure 23.1. Cardiac myocyte. (A) Image of a ventricular cell in transmitted light. (B) Three-dimensional confocal imaging of transverse-axial tubular system (TATS). Optical cut through the center of the cell. The TATS tubules are marked with di-8ANEPPS. (Parts (A) and (B) reproduced with permission from Ref. 2.) (C) Electron micrograph; longitudinal ultrathin section. (Copyright © 2006 Biophysical Journal [7].) Oblique section shows the localization of junctional SR (jSR) in relation to T-tubules and mitochondria (M). MF, myofilaments; T, T tubule of TATS; Z, Z line.

in this peri-T-tubule region (Figure 23.1C) [7, 8]. Under pathological conditions cardiac cells undergo so-called ultrastructural remodeling, which increases the distance between the structures and makes them available for nano-objects.

A

mDNA

Anti-apoptotic

Sarcoplasmic reticulum JC

Antigens T-tubule

Mitochondrion Nucleus Antisense oligos, siRNA, shRNA, drugs, CeO2 DNA B Transendothelial channel ~75 nm

Nano-object Plasmalemmal vesicles (caveolae) ~75 nm

Intercellular space ~10 nm

Capillary lumen

~200 nm

Adherens junction

Endothelial cells (internal tunic)

Tight junction

Basement membrane (middle tunic) Cells and fibers (adventitial tunic) Interstitial (perivascular) space C

E

D M M

M Z

N Z TT

Z

Figure 23.2. Nano-objects and barriers for their delivery to subdomains of cardiac myocytes. (A) Nano-objects and their cardiac intracellular targets. Simplified schematic representation of intracellular targets for nano-objects in ventricular cell. (Reproduced with permission from Ref. 2.). JC, Junctional cleft; mDNA, mitochondrial DNA; siRNA, synthetic small interfering RNA. (B) The capillary wall in the heart. (Copyright © 2006 Biophysical Journal [7].) Diagram schematically represents three mechanisms of cardiac vascular permeability: (1) transport by plasmalemmal vesicles, (2) transendothelial channel made by the fused vesicles, and (3) intercellular space. (C–E) Distribution of gold nanoparticles in permeabilized ventricular myocytes. Representative electron micrographs show the distribution of 3- (C) and 6-nm (D,E) particles within ventricular cells. M, mitochondrion; N, nucleus; Z, Z line, TT, T tubule. Ovals mark particles located deeper in the section and found with digitally enhanced contrast.

CARDIAC VASCULAR PERMEABILITY

23.3

437

NANO-OBJECTS AND THEIR ANALYTICAL APPLICATIONS

All nanostructures with important analytical applications could be divided into two groups: (1) nano-objects (nanotubes, spherical nanoparticles, macromolecules, etc.) and (2) nanodevices (nanocapacitors, nanopores, nanocantilevers, etc.) [2]. Nano-objects can be used in a variety of bioanalytical formats [2]: (1) as quantitative tags, such as the optical detection of quantum dots and the electrochemical detection of metallic nanoparticles; (2) as substrates for multiplexed bioassays (encoded nanoparticles such as striped metallic nanoparticles); (3) as controllers of signal transduction (e.g., in colloidal goldbased aggregation assays); or (4) as catalyzers or inducers of biological processes (i.e., nonviral vectors). Chemotherapeutic and imaging nano-objects are usually conjugated to a chemotherapeutic drug (i.e., folic acid, paclitaxel, doxorubicin, etc.) and/or imaging agent (FITC, GFP, etc.). Nano-objects that have been suggested for biomedical research have diameters from 0.8 to 400 nm [2, 9]. Note that the two most popular groups of viral vectors for gene transfer are 20 nm (adeno-associated virus) and 60–90 nm (adenovirus) in diameter [10]. Although all nano-objects within the 2–100-nm size range were found to alter signaling processes essential for basic cell functions, 40- and 50-nm nanoparticles demonstrated the greatest effect [11]. Nano-objects could be injected directly into the tissue, but a pumping heart is not a convenient target for an injection. Therefore intravenous injection is the most appropriate way for delivery of nano-objects to the myocardium. Note that only nano-objects 6 nm in size cannot diffuse into the nucleus or mitochondrial intermembrane space (Figure 23.2D,E) due to the clearance within the NPC (∼5.5 nm) and porin (or VDAC; ∼3 nm) [2]. Certain parts of “free” intracellular space within cardiac myocytes are relatively unavailable for naniobjects >3 nm [7]. For instance, Figure 23.2E shows that after silver enhancement 6-nm particles can be found with much higher probability along Z-lines than along myofibrils. The 3-nm clearance excludes passive diffusion of nanoparticles into mitochondrial intermembrane space. Moreover, the 3-nm objects probably cannot enter the mitochondrial matrix, the zone of highest interest for nanomedicine [43, 44]. Therefore the use of the phenomenon of mitochondrial fusion was suggested for nanoparticle delivery [2, 23]. The most important target for contemporary bioresearch and therapy is the delivery of nano-objects to the nucleus. Cardiomyocytes do not undergo mitotic reorganization of the nuclear envelope, and for nano-objects the NPC is the only gate to the nucleus. There are two mechanisms of translocation across the NPC: passive diffusion and active transport. Although the NPC central channel has a limiting diameter of ∼25–30 nm [38], it is known that molecules