Tissue Engineering and Novel Delivery Systems edited by
Michael J. Yaszemski Mayo Clinic Rochester; Minnesota, U.S.A.
...
95 downloads
2696 Views
8MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Tissue Engineering and Novel Delivery Systems edited by
Michael J. Yaszemski Mayo Clinic Rochester; Minnesota, U.S.A.
Debra J. Trantolo Cambridge Scientific, Inc. Cambridge, Massachusetts, U.S.A.
KaiHUwe Lewandrowski Cleveland Clinic Cleveland, Ohio, U.S.A.
Vasif Hasirci Middle East Technical University Ankara, Turkey
David E. Altobelli DEKA Research and Development Corporation Manchester; New Hampshire, U.S.A.
Donald L. Wise Northeastern University Boston, and Cambridge Scientijic, h e . Cambridge, Massachusetts, U.S.A.
MARCEL
MARCELDEKKER, INC. DEKKER
NEWYORK BASEL
The first edition of this book and its companion volume, Biomaterials in Orthopedics, were published as Biomaterials and Bioengineering Handbook, edited by Donald L. Wise (Marcel Dekker, Inc., 2000). Although great care has been taken to provide accurate and current information, neither the author(s) nor the publisher, nor anyone else associated with this publication, shall be liable for any loss, damage, or liability directly or indirectly caused or alleged to be caused by this book. The material contained herein is not intended to provide specific advice or recommendations for any specific situation. Trademark notice: Product or corporate names may be trademarks or registered trademarks and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress. ISBN: 0-8247-4786-0 This book is printed on acid-free paper. Headquarters Marcel Dekker, Inc., 270 Madison Avenue, New York, NY 10016, U.S.A. tel: 212-696-9000; fax: 212-685-4540 Distribution and Customer Service Marcel Dekker, Inc., Cimarron Road, Monticello, New York 12701, U.S.A. tel: 800-228-1160; fax: 845-796-1772 Eastern Hemisphere Distribution Marcel Dekker AG, Hutgasse 4, Postfach 812, CH-4001 Basel, Switzerland tel: 41-61-260-6300; fax: 41-61-260-6333 World Wide Web http://www.dekker.com The publisher offers discounts on this book when ordered in bulk quantities. For more information, write to Special Sales/Professional Marketing at the headquarters address above. Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming, and recording, or by any information storage and retrieval system, without permission in writing from the publisher. Current printing (last digit): 10 9 8 7 6 5 4 3 2 1 PRINTED IN THE UNITED STATES OF AMERICA
Preface
Tissue Engineering and Novel Delivery Systems contains cutting-edge chapters by leading practitioners who deal with critical issues concerning medical applications of biomaterials. The topics range from synthetic biopolymers used in controlled-release delivery systems to natural equivalents used in tissue reconstruction. Each chapter derives its targeted discussion from basic science and engineering exposure, as well as clinical experience. The text offers a wealth of valuable data and experience that will be of use to all bioengineers, materials scientists, and clinicians concerned with the properties, performance, and use of biomaterials as delivery or reconstructive vehicles—from research engineers faced with designing biomaterials to physicians and surgeons charged with shepherding the use of the biomaterial into the applied clinical settings. The chapters provide rich insights into our experiences today with a broad spectrum of modern biomaterials applications from forever challenging delivery systems to exciting tissue replacements. The book features discussion of the following: Basic science, engineering, and medical applications of biomaterials The properties, performance, and use of engineered biopolymers for delivery systems Developments in tissue-engineered biomaterials Emerging concepts in biomaterials Biomaterials are the subject of intense and demanding development. New challenges brought forth from the frontiers of a society that embraces the concept of premium healthcare drive intellectual curiosity and experimentation. No longer do engineered tissues and targeted delivery vehicles seem untenable. This book focuses on materials in or on the human body—materials that define the world of ‘‘biomaterials’’ and cover a wide range of biomaterials both natural and synthetic. The diversity of the field necessitates multidisciplinary contributions from science, engineering, and practical medical approaches. As a result, scientists, engineers, and physicians are among the authors. They provide a full and detailed accounting of the state of the art in this rapidly growing area and reflect the diversity of the field. The users of this book will represent a broad base of backgrounds ranging from the basic sciences (e.g., polymer chemistry and biochemistry) to the more applied disciplines (e.g., mechanical and chemical engineering, pharmaceutics, and medicine). To meet varied needs, each chapter provides clear and fully detailed discussions. This in-depth but practical coverage should also assist recent inductees to the biomaterials circle. This volume conveys the intensity of this fast-moving field in an enthusiastic presentation. Michael J. Yaszemski Debra J. Trantolo Kai-Uwe Lewandrowski Vasif Hasirci David E. Altobelli Donald L. Wise iii
Contents
Preface
iii
PART I: BIOCOMPATIBILITY AND THE BIOMATERIAL–TISSUE INTERFACE 1.
Role of Extracellular Matrix Remodeling in Advanced Biocompatibility M. Cannas, M. Bosetti, M. Sabbatini, and F. Reno`
2.
Elastic Protein-Based Biomaterials: Elements of Basic Science, Controlled Release, and Biocompatibility Dan W. Urry, T. Copper Woods, Larry C. Hayes, Jie Xu, David T. McPherson, Masamichi Iwama, Masakazu Furutan, Toshio Hayashi, Mitsuhiro Murata, and Timothy M. Parker
3.
4.
Noninvasive Measurement of the Transient Adhesion of Cells to BloodContacting Surfaces Jun Li, M. Kurt Sly, Robert Chao, Anca Constantinescu, Padmakar Kulkarni, Michael Jessen, and Robert Eberhart Clinical Properties and Healing Characteristics of e-PTFE Vascular Grafts and Their Effects on Long-Term Patency Glenn C. Hunter, Hana Holubec, Alex Westerband, David A. Bull, Kenneth J. Woodside, and Charles W. Putnam
5.
Biomaterials and Stent Technology Kytai T. Nguyen, Shih-Horng Su, Meital Zilberman, Pedram Bohluli, Peter Frenkel, Richard Timmons, Liping Tang, and Robert Eberhart
6.
Long-Term Evaluation of a Novel Tissue Adhesive (BioGlue) for Use in Surgery Nancy Perlman, Charles W. Hewitt, Steve D. Lenz, K. Umit Yuksel, Steven W. Marra, Jean-Luc V. Tran, Jonathan H. Cilley, Vincent A. Simonetti, and Anthony J. DelRossi
1
31
55
75
107
131
v
vi
Contents
PART II: TISSUE ENGINEERING EQUIVALENTS 7.
Bioengineered Skin Reconstituted from Cultured Epidermis and Cryopreserved Dermis: Converting Cryopreserved Skin Allografts into a Permanent Skin Substitute Hanna Ben-Bassat
8.
Cutaneous Gene Therapy with Cultured Skin Substitutes Dorothy M. Supp and Steven T. Boyce
9.
Surface Properties of Polymeric Biomaterials and Their Modification for Tissue Engineering Applications Ays¸en Tezcaner and Vasif Hasirci
141
155
173
10.
Principles of Living Organ Reconstruction by Tissue Engineering Lucie Germain, Franc¸ois Berthod, Ve´ronique Moulin, Francine Goulet, and Franc¸ois A. Auger
197
11.
Recombinant Protein Scaffolds for Tissue Engineering Jerome A. Werkmeister, Paul R. Vaughan, Yong Peng, and John A. M. Ramshaw
229
12.
Auricular Cartilage Tissue Engineering Doreen Rosenstrauch, Kamuran Kadipasaoglu, Harnath Shelat, Pierre Zoldhelyi, and O. H. Frazier
253
13.
Rheology of Biological Fluids and Their Substitutes Assunta Borzacchiello, Luigi Ambrosio, Paolo Netti, and Luigi Nicolais
265
PART III: DELIVERY SYSTEMS 14.
Production of Microbial Polyesters and Their Application in the Construction of Biodegradable, Controlled Antibiotic Release Systems Vasif Hasirci, F. Korkusuz, N. Akkas, E. Bayramli, N. Hasirci, F. Severcan, M. Timucin, and G. Alaeddinoglu
281
15.
Microencapsulation of Protein Drugs: A Novel Approach Yoon Yeo and Kinam Park
305
16.
Polymeric Gene Delivery Systems Goldie Kaul and Monsoor Amiji
333
17.
Bioactive Molecules and Biodelivery Systems ¨ ner and H. Su¨heyla¨ Kas Filiz O
369
18.
Biodegradable Nanoparticles as Drug Delivery Systems for Parenteral Administration Michael Chorny, Hagit Cohen-Sacks, Ilia Fishbein, Haim D. Danenberg, and Gershon Golomb
393
Contents
19.
Biodegradable Hydrogels as Drug Controlled Release Vehicles C. C. Chu
20.
Growth Factor Delivery by a Cloned Multipotent Cell from Bone Marrow Stroma Potentiates Bone Repair Nigel M. Azer, Quanjun Cui, Ing-Lin Chang, T. Lisle Whitman, Chang Hahn, Gwo-Jaw Wang, and Gary Balian
21.
22.
New Synthetic Biodegradable Polymers for Bone Morphogenetic Protein Delivery Systems Naoto Saito, Hiroshi Horiuchi, Narumichi Murakami, Jun Takahashi, Takao Okada, Kazutoshi Nozaki, and Kunio Takaoka Nanosized Biosensors and Delivery Vehicles Agnes E. Ostafin, Hiroshi Mizukami, and Joel P. Burgess
vii
423
463
475
483
PART IV: EMERGING CONCEPTS IN BIOMATERIALS 23.
Relationships Between Biomaterials and Biosensors Kirk J. Bundy
505
24.
Collagen: The Next Generation of Resorbable Biodevices in Surgery Frank DeLustro, Louis Sehl, Trudy Estridge, and Donald Wallace
521
25.
Polyhydroxybutyrate and Its Copolymers: Applications in the Medical Field Fatma Kok and Vasif Hasirci
543
26.
From Polymer Chemistry and Physicochemistry to Nanoparticulate Drug Carrier Design and Applications C. Vauthier, E. Fattal, and D. Labarre
27.
Index
Hyperspectral Analysis of Collagen Infused with BisGMA-Based Polymeric Adhesive Paulette Spencer, J. Lawrence Katz, Massood Tabib-Azar, Yong Wang, Ajay Wagh, and Tsutomu Nomura
563
599
633
1 Role of Extracellular Matrix Remodeling in Advanced Biocompatibility M. Cannas, M. Bosetti, M. Sabbatini, and F. Reno` University of Eastern Piedmont Medical School Novara, Italy
I. DEFINITIONS For many years the extracellular matrix (ECM) was thought to serve only as a structural support for tissues. However, as early as 1966, Hauschka and Konigsberg [1] showed that interstitial collagen promoted the conversion of myoblasts to myotubes, and shortly thereafter it was shown that both collagen [2] and glycosaminoglycans [3] play a crucial role in salivary gland morphogenesis. Based on these (and other) pieces of indirect evidence, Hay [4] put forth the idea that the ECM was an important component in embryonic inductions, a concept which implicated the presence of binding sites (receptors) for specific matrix molecules on the surface of cells. By this time, the stage was set to begin investigating in detail the mechanisms by which extracellular matrix molecules influence cell behavior. Bissell et al. [5] proposed the model of ‘‘dynamic reciprocity’’ between the ECM on the one hand and the cytoskeleton and nuclear matrix on the other. In this model, ECM molecules interact with receptors on the surface of cells which then transmit signals across the cell membrane to molecules in the cytoplasm; these signals initiate a cascade of events through the cytoskeleton into the nucleus, resulting in the expression of specific genes, whose products, in turn, affect the ECM in various ways [6,7]. Cell–ECM interactions participate directly in promoting cell adhesion, migration, growth, differentiation, and programmed cell death; in modulation of the activities of cytokines and growth factors, and in directly activating intracellular signaling. All these activities are connected in some way to biological compatibility. The molecular complex called ECM has as basic components collagens and other glycoproteins implicated also in the resistance to tensile and compressive mechanical forces. The macromolecular components of the polymeric assemblies of the ECM are in many cases secreted by cells as precursor molecules that are significantly modified (proteolysis processed, sulfated, oxidized, and cross-linked) before they assemble with other components onto functional polymers [8]. The formation of matrix assemblies in vivo is therefore in most instances a unidirectional, irreversible process, and the disassembly of the matrix is not a simple reversal of assembly, but involves multiple, highly regulated processes. One consequence of this is that polymers reconstituted in the laboratory with components extracted from extracellular matrices do not have all the properties they have when assembled by cells in vivo. 1
2
Cannas et al.
The ECM in vivo is also modified by cells as they proliferate, differentiate, and migrate, and cells in turn continuously interact with the matrix and communicate with each other through it (Fig. 1) [9]. The ECM is therefore not an inert product of secretory activities, but influences cellular shape, fate, and metabolism in ways that are as important to tissue and organ structure and function as the effects of many cytoplasmic processes. In the past the ECM was primarily appreciated for its challenge to biochemists interested in protein and complex carbohydrate structure; a detailed characterization of ECM constituents is now considered essential for understanding cell behavior in the context of tissue and organ development and function. Some of these constituents are obviously important for their structural properties (as collagens, elastin, and fibrillin), while others (matrix-bound FGFs, TGF-, BMPs) are signaling molecules. In a third category are multidomain molecules (as fibronectin, laminin, thrombospondin, tenascin, syndecans, and other proteoglycans) that are both structural constituents as well as regulators of cell behavior. These molecules and their organization vary with tissue type (mesenchymal cells are immersed in an interstitial matrix, whereas epithelial and endothelial cells contact a basement membrane only through their basal surfaces), and their distribution is not static, but rather varies from tissue to tissue and during development from stage to stage [10–13]. This diversity of composition, organization, and distribution of ECM results not only from differential gene expression for the various molecules in specific tissues, but also from the existence of differential splicing and post-translational modifications of those molecules. For example, differential splicing may change the binding potential of proteins to each other [11,14]
Figure 1 General organization and functions of some receptions involved in interactions between the cell and extracecullar matrix.
Extracellular Matrix Remodeling
3
or to their receptors [15]; variations in the glycosylation can lead to changes in cell adhesion [16]. In addition, the presence of divalent cations such as Ca2Ⳮ [17,18] can affect matrix organization and influence molecular interactions that are important in the way ECM molecules interact with cells [19]. The ECM of bone is a complex structure composed primarily of type I collagen, but also containing other constituents, including proteoglycans (biglycan and decorin), glycoproteins (osteonectin, osteopontin, fibronectin, thrombospondin and bone sialoprotein), and factors that may be essential for the mineralization process, such as bone protein [20]. The remodeling of bone ECM–hardened connective tissue involves coupling of the degradation of the extracellular matrix with the synthesis of new matrix components. In certain bone diseases, such as osteoporosis, an imbalance occurs such that a disproportionate amount of matrix is degraded compared to the amount synthesized [21]. Growth factors and cytokines interact with the ECM in a variety of ways which allows them to mutually affect each other [22,23]. On the one hand, ECM can serve as a reserve by binding growth factors and cytokines and protecting them from being degraded [24], by presenting them more efficiently to their receptors [16,25] , or by affecting their synthesis [26]. In this way the ECM can affect the local concentration and biological activity of these factors. For example, when neutrophils adhere to fibronectin, they produce higher levels of tumor necrosis factor (TNF) [22]. On the other hand growth factors and cytokines can stimulate cells to alter the production of ECM molecules, their inhibitors, and/or their receptors [27]. TGF- up-regulates the expression of matrix molecules [28] and of inhibitors of enzymes that degrade ECM molecules [29]. In a number of cases, only specific forms of these growth factors and cytokines bind to specific ECM molecules. Platelet-derived growth factor (PDGF) [30] and the 9E3 protein [31,32] (a chicken cytokine that is overexpressed during wound repair) fit in this category. The latter is secreted as a 9-kDa protein but can be processed to 6 kDa by plasmin; both forms are found in association with interstitial collagen, but only the smaller form binds to laminin, and neither form binds to fibronectin [33]. Importantly, binding of specific forms of these factors to specific ECM molecules can lead to their localization to particular areas and affect their biological activities. Many cell types undergo cell death when deprived of adhesion to the appropriate extracellular matrix. For example, endothelial and epithelial cells die by apoptosis when detached from substrate or when the attachment is prevented by growing them in suspension. This phenomenon is named anoikis from the Greek word for homelessness. It is involved in a wide diversity of tissue-homeostatic, developmental, and oncogenic processes [34]. Normal cell and tissue homeostasis reflects a dynamic balance of cell proliferation, differentiation, and apoptosis; in this context anoikis maintains the correct cell number of high-turnover epithelial tissues. The clearest evidence for this is that the breakdown of anoikis contributes to neoplasia [35,36]. The substrate-dependent cell growth is mediated by surface receptors belonging to the integrin family. Meredith et al. [37] demonstrated that detachment-induced apoptosis in human endothelial cells was blocked by plating cells on an immobilized antibody to l integrin.
II. COMPONENTS OF THE EXTRACELLULAR MATRIX The extracellular matrix is an intricate arrangement of glycoproteins, collagens, proteoglycans, and growth factors that act not only as a physical scaffold for the attachment and organization
4
Cannas et al.
of cellular structures, but also as a mediator of intracellular signaling through cell surface receptors that recognize these ECM molecules. Most ECM glycoproteins promote cell adhesion and cause cytoskeletal reorganization, leading to signals that direct differentiation and promote cell survival. ECM is composed of fibrillar protein primarily important for cell adhesion and structural characterization of the matrix, represented by collagen, laminin, and fibronectin. These proteins are accompanied by other proteins primarily involved in the dynamic relationship proprieties of ECM (growth factor activity modulation, cellular growth, differentiation, and migration), represented by proteoglycan, vitronectin, and several type of glycoprotein that mediate cell–matrix or cell–cell interaction. All these protein elements are immersed in an amorphous environment represented by glycosaminoglycan chains that form a matrix gel structure [38]. The ability of cells to adhere to the ECM is a critical determinant of cytoskeletal organization and thus of cellular morphology (Fig. 2). In addition to regulating cell shape, cell–ECM interactions also regulate the ability of a cell to proliferate, migrate, and differentiate. Furthermore, cell–matrix interactions that support cytoskeletal organization of focal adhesions are essential for survival of anchorage-dependent, nontransformed cells [39]. This wide range of activities suggests that the ECM is a key contributor to overall cellular physiology. Correspondingly, the ability of matricellular proteins to modulate cell adhesion and cytoskeletal organization suggests an important role for these proteins in essential processes. A. Fibrillar Proteins The collagens are a superfamily of extracellular fibrillar proteins that plays a dominant role in maintaining the integrity of various tissues and also has a number of other important functions. The superfamily actually includes more than 20 collagen types with altogether at least 38 distinct polypeptide chains and more than 15 additional proteins that have collagen-like domains. The collagens are characterized by the presence of one or several domains termed ‘‘triple helix’’ that are made of three polypeptide chains folded around each other, all differing by their molecular structure and by the way helical and globular domains are arranged. In any case, however, at least one triple helical domain exists. It is formed by the association of three polypeptide chains, each of them containing a glycine every three residues and many proline
Figure 2 Schematic drawing of connective tissue. Variation in the organization, amount, and composition of the ECM depends, on local physiological requirements. The ECM can be mineralized in bone and dental tissue, elasticized in skin, and rendered transparent in the lens of the eye.
Extracellular Matrix Remodeling
5
or hydroxyproline residues, this specific molecular characteristic attests for the belonging of the protein to the collagen group. Probably these domains are useful for the association of peptide chains in register prior to their folding; however, they participate in the transport of the elementary molecules from the synthesizing cells to their final place in the connective tissue, and they contribute to insertion of the molecules into their specific place inside the growing fibrils [40],[41]]. Most collagens form polymeric assemblies, such as fibrils, networks, and filaments, and the superfamily can be divided into several families based on these assemblies and other features. Most notable are (1) fibrils that are found in most connective tissues and are made by alloys of fibrillar collagens (types I, II, III, V, and XI) and (2) sheets constituting basement membranes (type IV collagen) and Descemet’s membrane (type VIII collagen). Collagen fibers as they are evidenced by histological methods, for instance in tendons, are of complex structure. Most of their constituting subunits are type I tropocollagen molecules, but they also contain in their center a filament of type V collagen that seems to serve as a guide during their edification. On the surface of the fibers are molecules of type III collagen that limit the growth in diameter and also type XII molecules that serve to bind the fibers to the surrounding substances. The collagen type multiplicity is explained by their various functions (mechanical role for tendons and ligaments, functions of wrapping around muscle cells, basement membrane role as a support for endothelial cells, function of glomerular filter, etc.) [42]. Collagens are not only essential for the mechanical resistance and resilience of multicellular organisms, but are also signaling molecules defining cellular shape and behavior. The communication between collagens and cells is achieved by cell surface receptors. Three types of cell surface receptors for collagen are known: integrins, discoidin domain receptors, and glycoprotein VI. All three types independently trigger a variety of signaling pathways upon collagen binding. Besides regulating numerous cellular responses, both integrin and discoidin domain receptors monitor the integrity of the collagenous extracellular matrix by triggering matrix degradation and renewal. Some recently discovered mechanisms of locally controlled expression of collagen, collagen-binding receptors, and collagen-degrading proteases in the cellular microenvironment have been described [43]. The laminins are a family of glycoproteins that provides an integral part of the structural scaffolding of basement membranes in almost every animal tissue. Each laminin is a heterotrimer assembled from alpha, beta, and gamma chain subunits, secreted and incorporated into cellassociated extracellular matrices. A different genetic regulation leads to the expression of different laminin forms and determines the formation of extracellular matrices with variable laminin composition and thereby different biological properties [44]. The laminins can self-assemble, bind to other matrix macromolecules, and have unique and shared cell interactions mediated by integrins, dystroglycan, and other receptors. The many interactions of laminins are mediated by binding sites, often contributed by single domains, which may differ between different forms of laminin. By virtue of their receptor interactions, they initiate intracellular signaling events that regulate cellular organization and differentiation. Through these interactions, laminins critically contribute to cell differentiation, cell shape and movement, maintenance of tissue phenotypes, and promotion of tissue survival [45]. Fibronectin is a high molecular glycoprotein present in the blood, in connective tissue, and at cell surface. It is a dimer of subunits of around 250 kDa, each composed of a series of independently folding modular domains. It is composed of multiple homologous repeats and contains many functional domains. Because of its ability to interact with many ligands—including cells, heparin, fibrin, collagen, DNA, and immunoglobulin—fibronectin can play its role in a variety of biological processes. As circulating glycoprotein may function as a nonspecific opsonin design to facilitate the uptake of tissue debris by phagocytic cells (Fig. 1) [46].
6
Cannas et al.
Fibronectin is synthesized by many types of differentiated cells and is believed to be involved in the attachment of cells to the surrounding extracellular matrix. It appears in different isoforms due to alternative mRNA splicing and subsequent post-translational modification. This glycoprotein interacts with cell surfaces as shown by the fact that fibronectin–collagen complexes, or fibronectin alone when insolubilized on a surface such as plastic, enhances the attachment of various types of cells to such surfaces. It seems that fibronectin, through its binding to collagen and to the cell surface, forms a bridge between the cell and its surrounding matrix. The abundance of fibronectin in basement membrane structures and the developmental changes observed in its expression lead one to believe that the attachment (or lack of it) of cells to fibronectin plays a significant role in morphogenetic events or in normal development [46,47]. Proteoglycan consists of a core protein and an associated glycosaminoglycan (GAG) chain of heparan sulfate, chondroitin sulfate, dermatan sulfate or keratan sulfate, which are attached to a serine residue. Proteoglycans may be found adhering to cell surface or Matrigel. The core proteins of cell surface proteoglycans may be transmembrane (e.g., syndecan) or GPI anchored (e.g., glypican). Many different cell surface and matrix proteoglycan core proteins are expressed in several tissues, some of them have tissue specificity expression as neurocan. The level of expression of these core proteins, the structure of their GAG chains, and their degradation are regulated by many of the effectors that control the development and function of tissue (Fig. 1) [48]. Regulatory proteins bind GAG including many growth factors and morphogens (fibroblast growth factors, hepatocyte growth factor/scatter factor, members of the midkine family), matrix proteins (collagen, fibronectin, and laminin), enzymes (lipoprotein lipase), and microbial surface proteins [49]. Structural diversity within GAG chains ensures that each protein–GAG interaction is as specific as necessary. The GAG–protein interactions serve to regulate the signal output of growth factor receptor tyrosine kinase and hence cell fate as well as the storage and diffusion of extracellular protein effectors. In addition, GAGs clearly coordinate stromal and epithelial development, and they are active participants in mediating cell–cell and cell–matrix interactions. Since a single proteoglycan, even if it carries a single GAG chain, can bind multiple proteins, proteoglycans are also likely to act as multireceptors which promote the integration of cellular signals [50]. B. Classic (Non fibrillar) Proteins The syndecans are a family of transmembrane heparan sulfate/chondroitin sulfate proteoglycans involved in the control of cell growth and differentiation. The biological activities of syndecan involve interactions with a variety of extracellular ligands, such as growth factors and matrix components, that are mainly mediated by the heparan sulfate moieties. The core proteins of the syndecan family proteoglycans are involved in signaling [51]. In particular the cytoplasmic tail of the ubiquitously expressed syndecan-4 is distinct from the other syndecans in its capacity to bind phosphatidylinositol 4,5-bisphosphate (PIP2) and to activate protein kinase C (PKC) alpha. These properties may confer on syndecan-4 specific and unique signaling functions (Fig. 1) [52]. The glypicans are a family of heparin sulfate proteoglycans (HPSGs) that are linked to the cell surface by a glycosylphospatidylinositol (GPI) anchor. They play a critical role in developmental morphogenesis and seems to regulate growth factor distributions in extracellular space. In general glypicans are expressed predominantly during development. Expression levels change in a stage- and tissue-specific manner, suggesting that glypicans are involved in the regulation of morphogenesis [53].
Extracellular Matrix Remodeling
7
Neurocan is a chondroitin sulfate proteoglycan of the lectican family and a component of the extracellular matrix of the central nervous system. It is mainly expressed during modeling and remodeling stages of this tissue. Neurocan can bind to various structural extracellular matrix components, such as hyaluronan, heparin, tenascin, and growth and mobility factors, and can interact with several cell surface molecules [54]. Vitronectin is a multifunctional adhesive glycoprotein—a plasma serum component—that can also function as a matrix component. The extracellular matrix protein vitronectin is recognized as an adhesive substrate by cells expressing specific vitronectin receptors of the integrin family (alpha v beta 1, alpha v beta 3, alpha v beta 5 or alpha IIb beta 3). Osteoclasts to osteopontin exposed on the bone surface via the classic vitronectin receptor alpha v beta 3, may be crucial to their bone resorption activity [55]. Cell interaction with vitronectin may induce spreading and migration and represents an important component in the tissue provisional matrix. In the matrix vitronectin also has an important regulatory role with respect to growth factors (e.g., vitronectin-bound IGF binding protein-5 (IGFBP-5) by modifying the responsiveness to insulin-like growth factor I [56]). As plasma serum is well known, the primary role of vitronectin is the regulation of vascular cell function [57]. In the ECM, vitronectin has an important role in cell growth and differentiation in specific processes that involve repair, as in promoting keratinocytes migration after lesion or in regulating the tissue integrity of bone matrix and osteoblast proliferation [58]. The galectins are a family of carbohydrate-binding proteins that are distributed widely in metazoan organisms. Each galectin exhibits a specific pattern of expression in various cells and tissues, and expression is often closely regulated during development. Although these proteins are found mainly in the cell cytoplasm, some are secreted from cells and interact with appropriately glycosylated proteins at the cell surface or within the extracellular matrix. These receptors include cell-adhesion molecules such as integrins and matrix glycoproteins such as laminin and fibronectin isoforms. Recent studies have increased understanding of the roles of the galectins in regulating cell–cell and cell–matrix adhesion. These interactions are critically involved in modulation of normal cellular motility and polarity as well as during tissue formation, and loss of adhesive function is implicated in several disease states including tumour progression, inflammation, and cystic development in branching epithelia such as kidney tubules [59–61]. Agrin is an extracellular matrix protein identified and named based on its involvement in the aggregation of acetylcholine receptors (AChRs) during synaptogenesis at the neuromuscular junction. Recent studies have demonstrated that agrin is a large extracellular heparan sulfate proteoglycan, with a molecular mass in excess of 500 kDa and a protein core of 220 kDa. Emerging evidence indicates that agrin’s function is not limited to its role in AChR aggregation during synaptogenesis, as the majority of agrin expression occurs in the developing central nervous system, especially in developing axonal tracts [62,63]. Osteopontin (OPN) is a phosphorylated acidic glycoprotein that has been implicated in a number of physiological and pathological events, including maintenance or reconfiguration of tissue integrity during inflammatory processes. As such, it is required for stress-induced bone remodeling and certain types of cell-mediated immunity. It also acts in dystrophic calcification, coronary restenosis, and tumor cell metastasis. An RGD-containing protein, OPN exists both as an immobilized ECM molecule in mineralized tissues and as a cytokine in body fluids; it is not a significant part of typical nonmineralized ECM. Several studies have demonstrated that OPN delivers a prosurvival, antiapoptotic signal to the cell. OPN influences cellular functions in a unique manner by mimicking key aspects of an ECM signal outside the confines of the ECM [64].
8
Cannas et al.
C. Matricellular Proteins These proteins are structurally unrelated glycoproteins that function as adaptors and modulators of cell–matrix interactions and are associated with remodeling, morphogenesis, and vascular growth. These protein are expressed primarily during development, during growth, and in response to injury, and they are not abundant in the normal adult animal, except in tissues with continued turnover, such as bone [65,66]. One key feature of matricellular proteins is that they function as both soluble and insoluble proteins. As substrates, these proteins are only capable of supporting the initial and intermediate stages of cell adhesion, attachment, and spreading. Focal adhesion and stress fiber formation, characteristics of strong cell adhesion, are rarely observed when cells are plated on these substrates. When presented in mixed substrata, the matricellular proteins can also antagonize the proadhesive activities of other matrix proteins. Interestingly, these matricellular proteins actually have deadhesive effects when presented as soluble proteins to cells in a strong adhesive state. These structurally diverse proteins include thrombospondins (TSPs), the tenascins, and SPARC (secreted protein, acidic and rich, in cysteine), all of which exhibit highly regulated expression during development and following cellular injury. TSP1, tenascin-C, and SPARC stimulate reorganization of actin stress fibers and disassembly of focal adhesion complexes but have only minimal or negligible effects on cell shape [65]. Probably the regulatory role of the matricellular components is manifested primarily in the capacity of responses to injury, as opposed to normal development, when the organization of copious amounts of secreted, structural proteins becomes necessary over a relatively short period of time. Trombospondins are a small family of secreted, modular glycoproteins consisting of TSP1 and TSP2. It can interact with specific cell surface receptors, cytokines, growth factors, and proteases, and the availability of each of these diverse molecules may help to define their function in a given environment. TSP1 was first identified as a thrombin-sensitive protein that was released in response to activation of platelets by thrombin. It is incorporated into fibrin clots and binds to a number of plasma proteins including fibrinogen, plasminogen, and histidine-rich glycoprotein. Instead TSP2 seems to be required for the generation of normal platelets from megakaryocytes. TSPs are induced in response to injury; they are able to modulate cell function interacting in the matrix environment as proadhesive or deadhesive factors [67,68]. Tenascins (TN) are a family of large extracellular matrix glycoproteins that comprise five known members. They display highly restricted and dynamic patterns of expression in the embryo, particularly during neural development, skeletogenesis, and vasculogenesis. These molecules are reexpressed in the adult during normal processes such as wound healing, nerve regeneration, and tissue involution, as well as in pathological states including vascular disease, tumorigenesis, and metastasis. In concert with a multitude of associated ECM proteins and cell surface receptors, TN proteins impart contrary cellular functions, depending on their mode of presentation (i.e., soluble or substrate bound) and the cell types and differentiation states of the target tissues. Expression of tenascins is regulated by a variety of growth factors, cytokines, vasoactive peptides, ECM proteins, and biomechanical factors [69,70]. SPARC (secreted protein, acidic and rich in cysteine) is a multifunctional glycoprotein that modulates cellular interaction with the extracellular matrix by its binding to structural matrix proteins, such as collagen and vitronectin, and by its abrogation of focal adhesions, features contributing to a counteradhesive effect on cells [71]. SPARC inhibits cellular proliferation and regulates the activity of growth factors, such as platelet-derived growth factor, fibroblast growth factor (FGF) 2, and vascular endothelial growth factor (VEGF). The expression of SPARC in adult animals is limited largely to remodeling
Extracellular Matrix Remodeling
9
tissue, such as bone, gut mucosa, and healing wounds, and it is prominent in tumors and in disorders associated with fibrosis. SPARC is a prototypical matricellular protein that functions to regulate cell–matrix interactions associated with development, remodeling, cell turnover, and tissue repair; thereby it is able to influence many important physiological and pathological processes [71,72].
III. CELL–EXTRACELLULAR MATRIX INTERACTIONS Interactions of cells with extracellular matrix molecules play a crucial role during development and wound healing. It is the continuous crosstalk between cells and the surrounding matrix environment that leads to the formation of patterns, the development of form (morphogenesis), and the acquisition and maintenance of differentiated phenotypes during embryogenesis. Similarly, during wound healing these interactions contribute to the processes of clot formation, inflammation, granulation tissue development, and remodeling. Many different lines of experimental evidence have shown that the basic cellular mechanisms that result in these events involve cell adhesion/deadhesion, migration, proliferation, differentiation, and programmed cell death. A. Receptors Cells ultimately dictate the location and composition of regional matrices. These matrices in turn communicate with cells and regulate their attachment, movement, growth, and gene expression. Integrins encompass a family of cell surface transmembrane glycoproteins molecules which play a crucial role in cell–cell and cell–extracellular matrix interaction in order to create and maintain tissue architecture. Of these heterodimeric transmembrane glycoproteins (consisting of an alpha and beta chain) as yet at least 20 different types have been described, all with a different pattern of reactivity with extracellular matrix components. The same integrin heterodimer can recognize several ECM proteins, and a particular ECM ligand may be recognize by more than one integrin [73,74],[35]. It has become clear that through integrin–ligand interaction cell function is also modulated. Furthermore, in pathological conditions integrins play a role of some significance. Otherwise integrins mediate leucocyte traffic in developing inflammatory processes and function in neoplastic growth when it comes to invasion and metastasis. As a result of the ability of integrins to specifically bind various ligands, they mediate specific binding of cells to each other and to the extracellular matrix. The subunit of integrin transduces intracellular signals as a consequence of binding of the cytoplasmic tail to components of the cytoskeleton, including actinin and actin, as well as with components of several intracellular signaling pathways, including focal adhesion kinase (FAK), src-family kinases, ras, and phosphoinositol-3kinase (PI-3K). Overall these data indicate that cells need to adhere to ECM via integrins and undergo some minimal degree of cytoskeleton organization to survive (Fig. 1) [75]. These receptors mediated cell adhesion to several structural proteins in ECM, and the several types of integrin receptors are often associated with cell or tissue specificity (e.g., beta 1 integrins appear to be the predominant adhesion receptor subfamily utilized by human osteoblast-like cells to adhere to collagen and laminin and in part to fibronectin [76]). This broad spectrum of activity is achieved by combining the ability to create mechanically functional junctions (cell–matrix and cell–cell) and signal-transducing capabilities. Osteoblasts and osteoclasts express specific integrin receptors, and the pattern of expression varies depending on the stage of cell differentiation. Interactions of integrins with bone–matrix adhesive proteins are thought to be important for regulating the tissue integrity and may provide a local, responsive regulatory system of osteoblastic differentiation as well [77].
10
Cannas et al.
Integrin-mediated binding to the matrix is required for growth cells and survivor, because prevention or disruption of integrin-mediated attachment or functional blocking with antagonists leads to a form of apoptosis termed anoikis [78,79]. Dystroglycan connects the extracellular matrix and cytoskeleton. It was originally identified as the extracellular and transmembrane constituents of a large oligomeric complex of sarcolemmal proteins associated with dystrophin, the protein product of the Duchenne muscular dystrophy (DMD) gene. During the last few years, dystroglycan has been demonstrated to be a novel receptor of not only laminin, but also agrin, two major proteins of the extracellular matrix having distinct biological effects. Dystroglycan plays a critical role in organizing extracellular matrix molecules on the cell surface and in basement membranes, and at least two human pathogens utilize dystroglycan to gain access to host cells [80,81]. As a receptor of laminin/ agrin, it has been implicated in such diverse and specific developmental processes as epithelial morphogenesis, synaptogenesis, and myelinogenesis. These findings point to the fundamental role of dystroglycan in the cellular differentiation process shared by many different cell types (Fig. 1) [82]. Discoidin domain receptors (DDRs) are a large family of ECM receptors that transmit signals through the use of an intrinsic tyrosine kinase function. The subgroup DDRs is distinguished from other members of the receptor tyrosine kinase family by a discoidin homology repeat in their extracellular domains such as typically found in a variety of other transmembrane and secreted proteins. Recently, various types of collagen have been identified as the ligands for the two mammalian discoidin domain receptor tyrosine kinases, DDR1 and DDR2. Both receptors display several potential tyrosine phosphorylation sites that are able to relay the signal by interacting with cytoplasmic effector proteins [83]. Glycoprotein VI (GPVI) is a receptor glycoprotein that has received particular attention. It is expressed on platelets in association with a signaling adapter, the Fc receptor gamma chain (Fc R-␥). The platelet response to collagen is a primary event in hemostasis and thrombosis, but the precise roles of the numerous identified platelet collagen receptors remain incompletely defined. The GPVI expression confers both adhesive and signaling responses to collagen in a graded fashion that is proportional to the GPVI receptor density. These results resolve some of the conflicting data regarding GPVI-collagen interactions and demonstrate that (1) GPVI-Fc R␥ expression is sufficient to confer both adhesion and signaling responses to collagen and (2) GPVI-mediated collagen responses are receptor density dependent at the receptor levels expressed on human platelets [84]. Cadherins are the receptors involved in cell–cell adhesion, as mediated by the cadherin–catenin system; this is a prerequisite for normal cell function and the preservation of tissue integrity. With recent progress in our understanding, beta-catenin as a component of a complex signal transduction pathway may serve as a common switch in central processes that regulate cellular differentiation and growth. The function of the cadherin–catenin system in cell adhesion as well as in intracellular signaling appears to be subjected to multifactorial control by a variety of different mechanisms, and data on a hormonal control of these signaling pathways suggest an important regulatory influence in many cellular systems (Fig. 1) [85]. B. Growth Factors Growth factors play a key role in regulation of their activity. However, growth factor signaling can also be regulated outside of cells by extracellular matrix proteins and proteolytic enzymes. The ability of extracellular proteins to process complex information in the absence of new protein synthesis is illustrated in blood clotting and complement pathways. An increasing number of growth factors, including insulin-like growth factors, fibroblast growth factor, transforming
Extracellular Matrix Remodeling
11
growth factor beta (TGF-), and hepatocyte growth factor, have been found to associate with the extracellular matrix proteins or with glycosaminoglican heparan sulfate. Rapid and localized changes in the activity of these factors can be induced by release from matrix storage and/or by activation of latent forms. These growth factors, in turn, control cell proliferation, differentiation, and synthesis and remodeling of the extracellular matrix. It is therefore likely that much of the information processing necessary for construction of complex multicellular organisms occurs in the extracellular environment. This suggests that the extracellular matrix plays a major role in the control of growth factor signaling. A direct interaction of growth factors with the matrix environment is based on the presence of growth factor binding proteins that may regulate the kinetic release and storage of growth factors in accordance with the immediate needs of cells [86]. An interesting example of interaction among growth factor and matrix may be observed in bone ECM, where a major mechanism for storage of TGF- is via its association with latent TGF- binding protein1 (LTBP1). LTBP1 proteolysis by plasmin, elastase, and metalloproteinase (MMP) may be a physiological mechanism for release of TGF- from ECM-bound stores, potentially the first step in the pathway by which matrix-bound TGF- is rendered active [87]. Growth factor interaction with ECM can involve the single molecular component such as glycoproteins, matricellular proteins, and proteoglycans that together with specificity of the matrix inside the different tissutal district contribute to regulate the action of several growth factors on cell physiology. The multifunctional adhesive glycoprotein vitronectin (VN), which is found in the circulation and widely distributed throughout different tissues, has been implicated in the regulation of vascular cell functions, and these activities could be related to interactions with various growth factors. In vitro, soluble VN interfered with TGF- binding to isolated extracellular matrix and was found to associate with TGF-1 and TGF-2 as well as with other growth factors such as vascular endothelial growth factor, epidermal growth factor, or basic fibroblast growth factor in a saturable manner. In particular, binding of TGF- was maximal for the heparin-binding multimeric isoform of VN, whereas VN in a ternary complex with thrombin and antithrombin or plasma VN exhibited weaker binding. Plasminogen activator inhibitor 1 (PAI-1) or heparin interfered with binding of VN to TGF-, and soluble PAI-1 was able to dissociate VN-bound TGF- [88]. The receptors important in adesive interaction with the cell–matrix are also involved in growth factor activity, both through interactions with the growth factor receptors, as well as direct interaction with the growth factor itself. A crosstalk between integrins and growth factor receptors was evidenced as an important signaling mechanism to provide specificity during normal development and pathological processes. Evidence from several model systems demonstrates the physiological importance of the coordination of signals from growth factors and the extracellular matrix to support cell proliferation, migration, and invasion in vivo. Several examples of crosstalk between these two important classes of receptors indicate that integrin ligation is required for growth factor–induced biological processes. Furthermore, integrins can directly associate with growth factor receptors, thereby regulating the capacity of integrin/growth factor receptor complexes to propagate downstream signaling [89]. Syndecans interact with growth factors, such as fibroblast growth factor, insulin-like growth factor, and epidermal growth factor. These interactions are required for biological activity, because these factors must first interact with the heparan sulfate chains of the syndecans before they can interact with their high-affinity signaling receptors (Fig. 1) [90]. C. Cytoskeleton and Signal Transduction Cell matrix adhesion occurs at many specialized sites termed focal contacts or focal adhesions. They consist of multimolecular protein complexes of transmembrane adhesion receptors anchor-
12
Cannas et al.
ing intracellular cytoskeletal structural proteins such as talin, vinculin, and ␣-actinin and signal transduction molecules including c-Src, FAK, p130cas, and paxillin. Moreover, many of these components and their variants are expressed in a cell type–restricted fashion, introducing a high level of complexity (Fig. 1) [91,92]. At the cytoplasmic faces of the adhesion site more than 50 proteins have been reported to be associated with focal contacts and related to ECM adhesions. The major transmembrane ECM receptors in these sites belong to the integrin family of proteins. Most of these contain multiple domains through which they can interact with different molecular partners, potentially forming a dense and heterogeneous protein network. The molecular and structural diversity of this ‘‘submembrane plaque’’ is regulated by a wide variety of mechanisms, including competition between different partner proteins for the same binding sites, interactions triggered or suppressed by tyrosine phosphorylation, and conformational changes in component proteins, which can affect their reactivity. Indeed changes can also be driven by mechanical force generated by the actin- and myosin-containing contractile machinery of the cells, or by external forces applied to the cells, and regulated by matrix rigidity [93]. Recent advances reveal that components of cell adhesion complexes display multiple interactions and functions, which cooperate to mediate both cell adhesion and signaling. Cell–matrix and cell–cell adhesions can serve as both recipients and generators of signaling information, using hierarchical and synergistic molecular interactions regulated by aggregation, conformational changes, phosphorylation, and tension, to act as signaling centers from which numerous intracellular pathways emanate to regulate cell growth, survival, and gene expression in normal and pathological conditions [35,94]. The readily apparent differences between the cytoskeletal structures of attached versus suspended cells suggest that survival signaling in anoikis is likely to be extensively regulated by the cytoskeleton. Such regulation may be affected by the multiple cytoskeletal changes apparent in transformed cells. Indeed, substantial evidence now exists showing that both signaling molecules and apoptosis regulators are associated with the cytoskeleton, and as such may together regulate anoikis by serving as sensors of cytoskeletal integrity [95].
IV. EXTRACELLUAR MATRIX REMODELING Biomaterials implanted into the human body participate in the process of wound healing. Healing is a complex and long-lasting process of tissue repair and remodeling in response to injury or implant. The wound response is aimed at reconstituting a tissue closely similar to the original one and can be divided into several distinct but overlapping phases such as blood coagulation, inflammation, cellular proliferation, and ECM deposition and remodeling (Fig. 3). The optimal wound healing process brings the complete tissue integration of the implant. Normal wound healing starts blood coagulation and coagulation factors like factor XII [96,97] and thrombin [98] modulate wound healing by acting as mitogens and chemoattractants. Thrombin also stimulates procollagen production by fibroblasts [99]. The presence of an artificial foreign surface may alter the wound healing process by selective protein adsorption to the material [100–102] , thus amplifying or down-regulating subsequent cell reactions to the adhering proteins [103–105]. In this context the chain of cellular reactions (adhesion receptors, intracellular signaling pathways, release of intercellular signal mediators, and the effect of such mediators on surrounding tissue) is of prime importance for the understanding of the effect of implanting foreign materials into the body. Knowledge of the signaling pathways from material to tissue is also important for the possibility of engineering wound healing in desired directions, e.g., healing with differentiated tissue rather than scar formation [106]. During the initial contact between implant materials and whole blood (Fig. 4), proteins adsorb within fractions of a second
Extracellular Matrix Remodeling
13
Figure 3 Wound healing phases after biomaterial implantation.
and platelets adhere to the surface-adsorbed proteins within the first few seconds of blood material contact [104]. Polymorphonuclear leukocytes (PMNLs) are found at the material surface after l0 min of blood exposure, and the cells are activated within 30 min of exposure [104], starting periprostetic inflammation. The PMNLs may become activated either directly, through some adhesion receptors, or via platelet-derived mediators (serotonin, platelet factor 4, lysophosphatidic acid, P-selectin, von Willebrand factor), and they produce reactive oxygen species as a part of the activation. Recently, phagocyte-derived oxygen species have received increasing recognition for their role in host defense and tissue injury [107]. The oxygen metabolites, notably O2 and reaction products of this molecule, play an important role in the intracellular and extracellular killing of microorganisms [108] and may also serve as mediators of the immune system and modulators of cellular activities such as cell adhesion [109], phagocytosis [110], and signal transduction [111]. Many aspects of acute and chronic inflammatory processes seem to be mediated by oxidants released by phagocytes through their ability to cause cellular production of cytokines [112,113]. In macrophages, the transcription factor NF-kB can be activated by H2O2 generated by the respiratory burst [114]. An immediate response to oxygen radicals has also been demonstrated for macrophages adhering to plastic surfaces where exposure to H2O2 lead to spreading via MAP kinase–dependent intracellular signaling pathways [115]. The superoxide anion may also activate phospholipase A2 (PLA2), resulting in membrane damage and generation of lipid inflammatory mediators [116]. Platelet functions, of importance in the early contact between material and blood, can be significantly altered by exposure to reactive oxygen species (ROS), including eicosanoid biosynthesis. Eicosanoids such as PGE2 have diverse effects on the regulation and activity of T lymphocytes present in the wound site. Eicosanoids and cytokines (TNF␣, IL-1) can regulate the proliferation and apoptosis of T cells that are cleared at the end of the immune response and therefore regulate the duration and intensity of inflammation [117]. Fibroblasts are the main effector cells in wound healing after wound formation; fibroblasts migrate to the wound site where they produce extracellular matrix and proinflammatory cytokines [118]. The regulation of extracellular matrix deposition is a key event in many physiological
14
Cannas et al.
Figure 4 Initial interaction between whole blood and biomaterial surface. PLP, plasmatic proteins; PLT, platelets; PMN, polymorphonuclear leukocytes; M/M, monocytes/macrophage
and pathological conditions. During wound healing ECM molecules need to be rapidly synthestized during the formation of early granulation tissue and also during the final replacement by mature connective tissue and tissue remodeling. At the end of wound healing, fibroblast are cleared by apoptosis [119]. Cell migration, angiogenesis, degradation of provisional matrix, and remodeling of newly formed granulation tissue all require controlled degradation of the extracellular matrix. A tight balance between connective tissue synthesis and breakdown is required for the normal functioning of all tissues. In fact an excessive deposition of connective tissue is a hallmark of fibrosis [120], while an excessive matrix degradation is the main cause of tissue reasorption and prosthesis aseptic loosening [121]. One way to control this balance is by the release of mediators from inflammatory cells or connective cells which can influence collagen and matrix metalloproteinase production very effectively in both paracrine and autocrine fashions, as demonstrated for cytokines and growth factors such as the trasforming growth factor  family, the interleukins, tumor necrosis factor, platelet-derived growth factor, and others [122]. A. The ECM as Peri-Implantar Site Periprosthetic bone loss can occur as a result of a reduction in the load transmitted to bone, socalled stress shielding. Periprosthetic bone loss also occurs as a result of an inflammatory reaction to small particles, such as those produced by the various wear modes. To varying degrees, both processes occur simultaneously in complex mechanical/biological systems such as joint replacements, and the adverse effects can be additive. Bone with decreased density secondary to stress shielding may be more susceptible to osteolysis. Relative motion between an implant and bone can cause bone loss through both mechanical and biological mechanisms [123].
Extracellular Matrix Remodeling
15
Active coupling of bone formation and resorption and increased osteocytes with abundant bone canalicular projections were found combined with the presence of immature bone matrices (osteoid and low-mineralized bone areas) in periprosthetic bones from loose hip joints. These results indicated that active osteoclastic bone resorption and/or defective bone formation are coupled with monocyte/macrophage-mediated foreign body type granuloma in the synovial-like interface membrane of loose hip joints. Thus, this unique high-turnover periprosthetic bone remodeling with bad bone quality probably is caused by the result of cellular host response combined with inappropriate cyclic mechanical loading. The fragile periprosthetic bone may contribute to hip prosthesis loosening [124]. The tissue adjacent to total hip and knee prostheses consists of synovial tissue, ‘‘fibrous membrane’’ (variably organized and variably vascularized fibrous tissue), lymphocytes (occasionally), and foreign body inflammatory cells (macrophages and giant cells) that are present roughly in proportion to the number of small particles [124]. Prosthetic particles elicit a cascade of responses at the cellular and tissue levels. The cell whose function is central to the biological reaction to prosthetic wear particles appears to be the macrophage. The mononuclear stem cell, which originates in the bone marrow, is the progenitor for both mononuclear macrophages and osteoclasts. Macrophages phagocytose small wear particles and may fuse to form foreign body multinucleated giant cells, usually in association with larger particles (Fig. 5) [126]. Osteoblasts and fibroblasts also may be important in the response to wear particles, resulting in altered formation of bone and connective tissue [127]. Although lymphocytes are occasionally present, their role in the inflammatory reaction is unclear. The histological features of the fibrous membrane coincide with the context of its formation and evolution. The quiescent membrane is composed of a thin layer of fibrous tissue and its occurrence is compatible with the biofunctionality of the implant.
Figure 5 TEM micrograph of macrophages incubated in the presence of biomaterial. (A) Control macrophages in absence of biomaterials; (B, C, and D) marcophages exposed to different biomaterials. The biomaterial is fragmented in small particles inside the vacuoles (B) or appears as large solid clumps inside irregularly shaped vacuoles (C) or the scarce cytoplasm has plenty of vacuoles containing lipid particles. Image at 3500⳯. TEM observations showed that only smaller particles were actively ingested by cells, while larger particles elicited no visible reaction.
16
Cannas et al.
The aggressive or lytic membrane develops when tissue-irritating, small, irregularly shaped and edgy breakdown products are deposited at the interface. The thick lytic membrane consists of an inflamed fibrous tissue, scattered within which are myriad granulomas, and its surface facing the implant displays a synovial-like aspect (Fig. 6) [128]. The mono- and polykaryonic macrophages, constituting the granulomatous response, ingest and abut on the wear particles. Among the intermediary substances of inflammation elaborated by the lymphocytes and macrophages of the lytic membrane, factors which stimulate the osteoclasts play the pivotal role in as much as progressive bone resorption is associated with progressive growth of the quiescent membrane and, hence, with incremental interfacial motion, interfacial deposition of wear particles, and inflammatory granulomatous response. The ensuing vicious circle culminates in aseptic loosening of the arthroplasty. The morphological features of the lytic membrane, though characterized by a stereotypical reaction pattern, are in their details closely linked with the nature of the diverse components of the composite joint replacement. The histological appearances of the bone–implant interface of stable and loose arthroplasties; the tissular reactions to polymethylmethacrylate, polyethylene, polyacetal, metals, and hydroxyapatite; as well as the characteristics of cemented and cementless porous-coated, press-fit, and hydroxyapatite-coated prostheses are described in many works [129,130]. Most periprosthetic bone resorption is effected by osteoclasts, but there is evidence that macrophages and foreign body giant cells are capable of direct, low-grade bone resorption. In vitro studies have indicated that activated macrophages release cytokines, including interleukins and prostaglandins, which play a role in the recruitment and differentiation of cells and stimulate bone resorption, but the specificity of the cytokine response and the regulatory mechanisms has not been defined. Under certain conditions, macrophages appear to directly release interleukin 1 beta and tumor necrosis factor. Several cytokines, including interleukin 1 beta, stimulate
Figure 6 Immunoistochemistry of the new tissue formed at the implant material interface. Lymphocytes were identified using a monoclonal mouse antibody anti-human leukocyte common antigen (LCA), Figure shows strongly labeled lymphoid cells (generally small lymphocytes) that were numerous in the granulation tissue of the perimplantar membrane. Light microscopy at 250⳯.
Extracellular Matrix Remodeling
17
osteoclast maturation. Although cytokines released by macrophages may directly stimulate bone resorption by osteoclasts, other effects may be mediated by intermediary cells such as fibroblasts or osteoblasts [131]. Matrix metalloproteinases (collagenase, gelatinase, and stromelysin), which are capable of effecting bone resorption, are also produced by interfacial membrane tissue around failed total hip and knee replacements. On the basis of the knowledge of such cellular and biochemical mechanisms of bone resorption, there has been increasing interest in and investigation of pharmacological agents that may modify these cellular responses. Other work indicates that, in addition to bone resorption, there is also a decrease in bone formation in association with periprosthetic osteolysis. It appears that all of the materials used in total joint replacement are capable of inducing an inflammatory foreign body reaction if the particles are within a certain size range and there are enough of them. The bone resorbing ability of macrophages in vitro is a function of the size, shape, and composition of the particles, and it is dose dependent. It has been previously recognized that there is an upper size limit for particle reactivity, but there may also be a lower size limit. Analyses of matrix metalloproteinase and tissue inhibitors of metalloproteinase interaction show imbalance between the enzymes and the endogenous inhibitors in favor of matrix metalloproteinase in peri-implantar tissue [132]. This induces pathologic connective tissue remodeling in the interface. The data suggest that matrix metalloproteinase and tissue inhibitors of the metalloproteinase system participate in the extracellular matrix degradation and tissue remodeling in artificial hip joints, and may contribute to the periprosthetic weakening, implant loosening, and osteolysis around implants. More evidence for their active involvement is sought by intervention studies with type-specific matrix metalloproteinase inhibitors [133]. Matrix metalloproteinases have been shown to play a role in aseptic loosening of total hip replacement (THR). Extracellular matrix metalloproteinase inducer (EMMPRIN) can upregulate expression of several MMPs but has little effect on their tissue inhibitor (TIMP). EMMPRIN expression is up-regulated in interface tissue, and that locally accumulated EMMPRIN may modulate MMP-1 expression. An imbalance in the activity of MMPs and TIMP may lead to tissue destruction and periprosthetic osteolysis. These biological responses, combined with mechanical stress caused by micromotion and oscillating fluid pressure, may eventually cause aseptic loosening of THR [134]. B. ECM and Matrix Metalloproteinase The timely breakdown of ECM is a key process for embryonic development, morphogenesis, and tissue resorption and remodeling. The matrix metalloproteinases play a central role in these processes. The expression of most matrices is transcriptionally regulated by growth factors, hormones, cytokines, and cellular transformation [135,136]. The proteolytic activities of MMPs are controlled during activation from their precursors and inhibition by endogenous inhibitors, ␣-macroglobulins, and tissue inhibitors of metalloproteinases (TIMPs). Almost 28 different MMPs have been individuated in vertebrates (21 in humans, Table 1) [137], and in most cases MMPs are synthesized as proenzymes and secreted as inactive pro-MMPs. The MMPs’ primary structure comprises several domain motifs (Fig. 7). The propeptide domain (about 80 amino acids) presents a highly conserved PRCG(V/n)PD sequence. The cys in the sequence binds the catalytic zinc to maintain the latency of pro-MMP [138]. The catalityc domain (about 170 amino acids) contains a zinc binding motif, HEXXHXXGXXH, and a conserved methionine, which forms a ‘‘Met-turn’’ structure [139]. The catalytic domains of MMPs have an additional structural zinc ion and 2–3 calcium ions, which are required for the stability and the expression of enzymatic activity. MMP-2 and MMP-9 have three repeats of fibronectin type II domain inserted
18
Cannas et al.
in the catalytic domain. These repeats interact with collagens and gelatins [140,141]. The CterminaI hemopexin-like domain (about 210 amino acids) has an ellipsoidal disk shape with a four-bladed -propeller structure; each blade consists of four antiparallel -strands and an ␣helix [142]. The hemopexin domain is an absolute requirement for collagenases to cleave triple helical interstitial collagens [143], although the catalytic domains alone retain proteolytic activity toward other substrates [144]. The hemopexin domain of MMP-2 is also required for the cell surface activation of proMMP-2 by MTl-MMP [145]. A transmembrane domain is found in the MT-MMPs which anchors those enzymes to the cell surface. Metalloproteinase genes are inducible by many stimuli such as growth factors, cytokines, chemical agents, and physical stress. MMP gene expression is down-regulated by trasforming growth factor beta, retinoic acids, and glucocorticoids [135,136]. Recent studies emphasize also that cell-matrix and cell-cell interactions can modulate the MMP gene expression in fibroblasts [146,147] , endothelial cells [148], lymphocytes [149], macrophages [150], and neoplastic cells [151]. Certain signaling pathways lead to expression of a particular MMP genes. For example, addition of soluble antibody to ␣5/1 integrin causes disruption of the actin cytoskeleton and an increased expression of MMP-1 in rabbit synovial fibroblasts [152]. This is because of activation of the GTP-binding protein Racl, which generated reactive oxygen species and induced activation of NF-KB [153]. This leads to induction of IL-l, an autocrine inducer of MMP-1 expression. Inflammatory cytokines, such as TNF-␣ and IL-1, also trigger a ceramide-dependent expression of MMP-1 in human skin fibroblast mediated by three distinct MAP kinase pathways, i.e., ERK1/2, stress-activated protein kinase SAPK/JNK, and p38 [154]. Apart from a few members activated by furin, most MMPs are secreted from the cell as inactive zymogens. Secreted pro-MMPs are activated in vitro by proteinases and by nonproteolytic agents such as SH-reactive agents, mercurial compounds, reactive oxygen, and denaturants. In all cases activation requires the disruption of the Cys-Zn2Ⳮ (cysteine switch) interaction and the removal of the propeptide [155]. In vivo, most pro-MMPs are likely to be activated by tissue or plasma proteinase or bacterial proteinases. It has been suggested that the urokinase type plasminogen activator (uPA)/plasmin system is a significant activator of pro-MMPs [156]. In-
Figure 7 Basic domain structure of MMPs.
Extracellular Matrix Remodeling
19
Table 1 Human MMPs, Substrate and Exogenous Activators Enzyme Collagenases Collagenase-1 (MMP-1) Collagenase-2 (MMP-8) Collagenase-3 (MMP-13)
Stromelysins Stromelysin-1 (MMP-3)
Stromelysin-2 (MMP-10) Stromelysin-like MMPs Stromelysin-3 (MMP-11) Metalloelastase (MMP-12)
Matrilysins Matrilysin (MMP-7) Matrilysin-2 (MMP-26) Gelatinases Gelatinase A (MMP-2)
Gelatinase B (MMP-9)
Membrane-type MMPs MT1-MMP (MMP-14)
MT2-MMP (MMP-15) MT3-MMP (MMP-16) MT4-MMP (MMP-17) MT5-MMP (MMP-24) MT6-MMP (MMP-25) Other MMPs MMP-19 Enamelysin (MMP-20) MMP-23 MMP-28
Substrates
Activated by
Collagen I, II, III, VII, VIII, X, aggregan, serpins, 2M Collagen I, II, III, aggregan, serpins, 2M Collagen I, II, III, VII, VIII, X, XIV, gelatin, FN, laminin, large tenascin osteonectin, serpins
MMP-3, -7, -10, plasmin kallikrein, chymase MMP-3, -10, plasmin MMP-2, -3, -10, -14, -15, plasmin
Collagen IV, V, IX, X, FN, elastin, gelatin, laminin, aggrecan, nidogenfibrillin, osteonectin, 1PI, myelin basic protein, OP, E-cadherin As MMP-3
Plasmin, kallikrein, chymase, tryptase
Serine protease inhibitor, 1PI Collagen IV, gelatin, FN, laminin, vitronectin, elastin, fibrillin, 1PI, myelin basic protein, apolipoprotein A
Furin ND
Elastin, FN, laminin, nidogen, collagen IV, tenascin, versican, 1-IP, OE-cadherin, TNFGelatin, 1-IP, syntetic MMP- substrates, TACE-substrate
MMP-3, plasmin
Gelatin, collagen I, IV, V, VII, X, FN, tenascin, fibrillin, osteonectin, monocyte chemoattractant protein 3 Gelatin, collagen IV, V, VII, IX, XIV, elastin, fibrillin, osteonectin 2
MMP-1, -13, -14, -15, -16, -tryptase?
Collagen I, II, III, gelatin, FN, laminin, vitronectin, aggrecan, tenascin, nidogen, perlecan, fibrillin, 1-IP, 2M, fibrin FN, laminin, aggrecan, tenascin, nidogen, perlecan Collagen III, FN, gelatin, casein, cartilage proteoglicans, laminin-1, 2M Fibrin, fibrinogen, TNF precursor Proteoglycan Collagen IV, gelatin, FN, fibrin
Plasmin, furin
Gelatin, aggrecan, COMP, collagen IV, laminin, nidogen, large tenas Amelogenin, aggrecan, COMP McaPLGLDpaARNh2 (synthetic MMPsubstrate Casein
Trypsin
Elastase, cathepsin G
ND
MMP-2, -3, -7, -13, plasmin, trypsin, chymotrypsin, cathepsin D
ND ND ND ND ND
ND ND
FN, fibronectin; 2M, 2-macroglobulin; 1PI, 1-proteinase inhibitor; COMP, cartilage oligomeric matrix protein; ND, not determinated; TACE, TNF-converting enzyme; OP, osteopontin.
20
Cannas et al.
stead, the activation of pro-MMP-2 is thought to take place on the cell surface thanks to the action of membrane-anchored MMP (MT1-MMP) [157]. Tissue inhibitors of metalloproteinases are small (21–30 kDa) endogenous regulators of MMP activiy in the tissue, and four homologous TIMP (TIMP-1 to 4) have been identified [158]. Tissue inhibitors of MMPs inhibit cell invasion in vitro, tumorigenesis, metastasis in vivo, and angiogenesis [158]. TIMPs exhibit additional biological functions. T1MP-1 and T1MP-2 have mitogenic activities on a number of cell types, whereas overexpression of these inhibitors reduces tumor cell growth [158]. These biological activities of TIMPs are independent of MMPinhibitory activities [159,160]. TIMPs seem to be therefore regulators not only in matrix turnover, but also in cellular activities. Although the main function of MMPs is remova1 of ECM during tissue resorption and progression of many diseases, it is notable that MMPs also alter biological functions of ECM macromolecules by specific proteolysis. For example, MMP-2 released by growth cones promotes neurite outgrowth by inactivating neurite-inhibtory chondroitin sulfate proteoglycans, thereby unmasking the neurite-promoting activity of laminin [161]. Specific cleavage of the Ala586–Leu587 bond in the ␣2 chain of laminin 5 by MMP-2 induces migration of normal breast epithelial cells by exposing a cryptic promigratory site [162]. The cleaved form of laminin 5 was found in tumors and in tissues undergoing remodeling but not in quiescent tissues [162]. Cleavage of type I collagen by MMP-1 and by MMP-13 initiates keratinocyte migration during reepithelialization [163] and osteoclast activation [164], respectively. Further insights into the biological and pathological function of MMPs have been provided by the use of transgenic animals and gene transfer techniques [165]. The expression of MMP-9 appears to be critical in later embryonic skeletal tissue development [166] and MMP-9–deficient (MMP-9-/-) mice exhibited phenotypic defects with a delayed, long bone growth associated with an abnormally thickened growth plate, which was accompanied by delayed apoptosis of hypertropic chondrocytes, vascularization, and ossification [167].
V. ECM AND ADVANCED BIOCOMPATIBILITY The ECM cells, components, and enzymes form a dinamic microenvironment that interact with implanted biomaterials. Biomaterial can alter the normal ECM turnover or remodeling causing a reaction in the periprosthetic tissue that leads to the prosthesis loosening or it can promote a ‘‘positive’’ interaction that leads to prosthesis integration and optimal wound healing. Therefore biomaterial interaction with ECM is a key aspect in biocompatibility as demonstrated by a growing amount of evidence. One of the most striking indications of the importance of ECM remodeling in the biocompatibility is the aseptic loosening of joint implants. Aseptic loosening is a current problem of major concern in patients with osteoarthritis and rheumatoid arthritis who had total hip replacements for the end-stage destructive arthritis. The problem has been increasing in recent years and pathologists observe the tissues obtained at revision arthroplasties from the femoral and/or cup components. It is well recognized that bone lysis occurs around the loose cemented prosthesis, where the fibrous membranes containing a histiocytic reaction to cement and polyethylene debris are formed at the bone–cement interface [168]. Various factors may contribute to the enhanced osteolysis around the cemented hip prostheses: they include localized mechanical stresses [169], micromotion between the cement and bone [170], abrasion particulate from the artificial joint surfaces [171], fragmentation of the cement [172], and hypersensitivity to metal [173].
Extracellular Matrix Remodeling
21
A synovial-like membrane is reported to occur often at the bone–cement interface, and it releases prostaglandin E2 (PGE2) and collagenase [174,175] , both of which may be involved in bone resorption [175]. Prostheses that can be inserted without cement were developed to diminish the adverse effects of cement. However, aseptic loosening of the uncemented prostheses due to osteolysis has also been observed [176]. Examination of tissue from the cementless bone–implant interface in individuals with clinical evidence of loosening has demonstrated the formation of a membrane characterized by the presence of foreign body giant cells containing polyethylene debris and mononuclear histiocytic infiltrates within a fibrous tissue stroma [177]. Different studies [132,178–182] have shown that the cementless interface membranes as well as cemented ones produce significantly higher levels of collagenase, gelatinolytic activity, PGE2, and interleukin-1 than the control tissue. These studies indicate that cells in the interface membranes produce matrix MMPs such as MMP-1, MMP-2, and MMP-9 as well as TIMPs. However, information is limited on the identification of the MMPs responsible for the enzymic activities or the source of the MMPs in the membranes [183,184]. Recently also an important role for MMP-1 and MMP-13 has been suggested in the loosening of artificial hip joints [185]. However the production of PGE2 and IL-1 suggests the indirect role of the membrane in bone resorption through stimulation of osteoclasts by the factors at the local site. On the other hand, recent studies using an in vitro bone resorption assay have shown that macrophages and macrophage polykaryons derived from the joint capsule of hip arthroplasties or granulomas induced by bone cement can directly resorb bone [186]. Another mechanism inducing periprosthetic tissue remodeling is the wear of implant materials, release of implant particles and debris into periprosthetic tissues with the formation of reactive granulation tissue against a foreign body, and the activation of cells to produce cytokines and enzymes [187,188]. Some studies demonstrated that granulomatous tissue cells such as infiltrated macrophages have the ability to produce inflammatory cytokines and substances (i.e., IL-1 and TNF [189,190] ) and activate osteoclasts. Osteoclasts were believed to cause bone resorption around implants and a recent study reported a strong expression of MMP-9 mRNA in the osteoclasts [191] as observed in the destructive joint of rheumatoid arthritis [192]. It was pointed out that the balance between MMPs and TIMPs was important to avoid the catastrophic effects of MMPs in various tissues [165]. In fact it has been hypothesized that an imbalance between MMPs and TIMPs at the edge of the interface granulation tissue and local overexpression of MMPs may cause tissue destruction around implants [193]. It is noteworthy that the cemented interface tissue has the ability to produce TIMPs [183]. It has been also observed that polymer surface chemical carachteristics can modulate the expression of MMPs in human osteoblast-like cells [194], fibroblasts [195], and peripheral blood cells [196] (Fig. 8) by direct contact, possibly through an autocrine mechanism or a surface differential adsorption. Recently a new intriguing role for MMP and TIMP espression at the biomaterial–bone interface has been suggested in the absence of any kind of disease [193]. Osteointegration at the site of orthopedic implants is dependent on the recruitment, attachment, and differentiation of osteogenic cells following implantation. The presence of fibrocartilage tissue and calcified bone within the interface, together with the presence of MMP-1 and MMP2 and their inhibitors (TIMP-1 and TIMP-2) within fibroblasts, chondrocytes, and osteoblasts, indicates a possible important role of these enzyme in the osteo-integrative phenomenon. In conclusion, ECM forms a dynamic site of interactions with biomaterials; its role is not a structural ‘‘static’’ support for tissues, but rather a very complex entity rich in factors that mediate the relation of intra- and extracellular signals. The study of the ECM environment can help in the understanding and improvment of biocompatibility.
22
Cannas et al.
Figure 8 Representative gelatin zymograpy obtained at 24 (A) and 48 (B) h. PBMCs conditioned medium in the presence of polystyrene (CT), UHMWPE (PE), and UHMWPE-oxidized (PEOx) MMP-2 (gelatinase A) and both inactive and active MMP-9 (gelatinase B). Both inactive (pro-MMP-2 and-9) and active (MMP-2 and-9) forms are indicated along with molecular weight markers for inactive forms. (From Ref. 196.)
REFERENCES
1. Hauschka SO, Konigsberg IR. The influence of collagen on the development of muscle colonies. P.N.A.S. 1966; 55:119–126. 2. Wessells NK, Cohen JH. Effects of collagenase on developing epithelia in vitro: lung, ureteric bud, and pancreas. Dev. Biol. 1968; 18:294–309. 3. Bernfield MR, Cohn RH, Banerjee SO. Glycosaminoglycans and epithelial organ formation. Amer. Zool. 1973; 13:1067–1083. 4. Hay ED. Interaction between the cell surface and extracellular matrix in corneal development In: Lash JW , Burger MM, eds. Cell and Tissue Interactions. New York: Raven Press, 1977:115–137. 5. Bissell MJ, Hall HG, Parry G. How does the extracellular matrix direct gene expression?. J. Theor. Bio. 1982; 99:31–68. 6. Ingber O. Extracellular matrix and cell shape: potential control points for inhibition of angiogenesis. J. Cell Biochem. 1991; 47:236–241. 7. Boudreau N, Myers C, Bissell MJ. From laminin to lamin: regulation of tissue-specific gene expression by the ECM. Trends Cell Biol. 1995; 5:1–4. 8. Schwartz N. Biosynthesis and regulation of expression of proteoglycans. Front. Biosci. 2000; 5: D649–D655. 9. Hay EDEd. Cell Biology of Extracellular Matrixnd. New York: Plenum Press, 1991. 10. Laurie GW, Horikoshi S, Killen PO, Segui-Realy B, Yamada Y. In situ hybridization reveals temporal and spatial changes in cellular expression of mRNA for a laminin receptor, laminin, and basement membrane (type IV) collagen in the developing kidney. J. Cell Biol. 1989; 109:1351–1362.
Extracellular Matrix Remodeling
23
11. French-Constant C, Hynes RO. Alternative splicing of fibronectin is temporally and spatially regulated in the chicken embryo. Dev. 1989; 106:375–388. 12. Sanes JR, Engvall E, Butkowski R, Hunter DD. Molecular heterogeneity of basal laminae: isoforms of laminin and collagen IV at the neuromuscular junction and elsewhere. J. Cell Biol. 1990; 111: 1685–1699. 13. Ferguson JE, Schor AM, Howell A. Tenascin distribution in the normal human breast is altered during the menstrual cycle and in carcinoma. Differentiation 1990; 42:199–207. 14. Chiquet-Ehrismann R, Matsuoka Y, Hofer U. Tenascin variants: differential binding to fibronectin and distinct distribution in cell cultures and tissues. Cell Reg. 1991; 2:927–938. 15. Hunter DD, Murphy MO, Olsson CV, Brunken WJ. S-Laminin expression in adult and developing retinae: a potential cue for photoreceptor morphogenesis. Nueron 1992; 8:399–413. 16. Oean JW, Chanrasekaran S, Tanzer ML. A biological tale of the carbohydrate moieties of laminin. J. Biol. Chem. 1990; 265:12553–12562. 17. Paulsson M. The role of Ca2Ⳮ binding in the self–aggregation of laminin-nidogen complexes. J. Biol. Chem. 1988; 263:5425–5430. 18. Ekblom P, Ekblom M, Fecker L. Role of mesenchymal nidogen for epithelial morphogenesis in vitro. Dev. 1994; 120:2003–2014. 19. Juliano RL, Haskill S. Signal transduction from the extracellular matrix. J. Cell Biol. 1993; 120: 577–585. 20. Blair HC, Zaidi M, Schlesinger PH. Mechanisms balancing skeletal matrix synthesis and degradation. Biochem. J. 2002; 364(Part 2):329–341. 21. Bateman JF, Golub SB. Deposition and selective degradation of structurally abnormal type I collagen in a collagen matrix produced by osteogenesis imperfecta fibroblasts in vitro. Matrix Biol. 1994; 14(3):251–262. 22. Nathan C, Sporn M. Cytokines in context. J. Cell Biol. 1991; 113:981–986. 23. Adams JC, Watt FM. Regulation of development and differentiation by the extracellular matrix. Dev. 1993; 117:1183–1198. 24. Flaumenhaft R, Moscatelli D, Rifkin OB. Heparin and heparan sulfate increase the radius of diffusion and action of basic fibroblast growth factors. J. Cell Biol. 1990; 111:1651–1659. 25. Roberts R, Gallagher J, Spooncer E. Heparan sulphate bound growth factors: a mechanism for stromal cell mediated haemopoiesis. Nature 1988; 332:376–378. 26. Flaumenhaft R, Rifkin OB. The extracellular regulation of growth factor action. Mol. Biol. Cell. 1992; 3:1057–1065. 27. Streuli CH, Schmidhauser C, Kobrin M. Extracellular matrix regulates expression of the TGF-1 gene. J. Cell Biol. 1993; 120:253–260. 28. Ignotz RA, Massague J. Transforming growth factor- stimulates the expression of fibronectin and collagen and their incorporation into the extracellular matrix. J. Biol. Chem. 1986; 261:4337–4345. 29. Edwards OR, Murphy G, Reynolds T. Transforming growth factor-3 modulates the expression of collagenase and metalloproteinase inhibitor. EMBO J. 1987; 6:1899–1904. 30. La Rochelle WJ, May-Shiroff M, Robbins KC, Aaronson SA. A novel mechanism regulating growth factor association with the cell surface: identification of a PDGF retention domain. Genes Dev. 1991; 5:1191–1199. 31. Martins-Green M, Bissell MJ. Localization of 9E3/CEF4 in avian tissues: expression is absent in RSV-induced tumors but is stimulated by injury. J. Cell Biol. 1990; 110:581–595. 32. Martins-Green M, Aotaki-Keen A, Hjelmeland L, Bissell MJ. The 9E3 protein: immunolocalization in vivo and evidence for multiple forms in culture. J. Cell Science 1992; 101:701–707. 33. Martins-Green M, Stoeckle M, Wimberly S. The 9E3/CEF4 cytokine: kinetics of secretion, processing by plasmin, and interaction with extracelluiar matrix. Cytokine 1996; 8(6):448–459. 34. Frisch SM, Francis H. Disruption of epithelial cell–matrix interactions induces apoptosis. J. Cell Biol. 1994; 124:619–626. 35. Schwartz MA, Schaller MD, Ginsberg MH. Integrins: emerging paradigms of signal transduction. Annu. Rev. Cell Dev. Biol. 1995; 11:549–599.
24
Cannas et al.
36. Zhu X, Ohtsubo M, Bohemer RM, Roberts JM, Assoian RK. Adhesion-dependent cell cycle progression linked to the expression of cyclin D1, activation of cyclin E-cdk2, and phosphorylation of the retinoblastoma protein. J. Cell Biol. 1996; 133:549–599. 37. Meredith J, Fazeli B, Schwartz M. The extracellular matrix as a survival factor. Mol. Cell Biol. 1993; 6:3470–3480. 38. Piez KA. History of extracellular matrix: a personal view. Matrix Biol. 1997; 16:85–92. 39. Gumbiner BM. Cell adhesion: the molecular basis of tissue architecture and morphogenesis. Cell 1996; 84:345–357. 40. Myllyharju J, Kivirikko KI. Collagens and collagen-related diseases. Ann. Med. 2001; 33:7–21. 41. Brown JC, Timpl R. The collagen superfamily. Int. Arch. Allergy Immunol. 1995; 107:484–490. 42. Van der Rest M, Garrone R. Collagen family of proteins. FASEB J. 1991; 5:2814–2823. 43. Vogel WF. Collagen-receptor signaling in health and disease. Eur. J. Dermatol. 2001; 11:506–514. 44. Colognato H, Yurchenco PD. Form and function: the laminin family of heterotrimers. Dev. Dyn. 2000; 218:213–234. 45. Tunggal P, Smyth N, Paulsson M, Ott MC. Laminins: structure and genetic regulation. Microsc. Res. Tech. 2000; 51:214–27. 46. Ruoslahti E. Fibronectin. J. Oral Pathol. 1981; 10:3–13. 47. Imoto E, Kakuta S, Hori M, Yagami K, Nagumo M. Adhesion of a chondrocytic cell line (USAC) to fibronectin and its regulation by proteoglycan. J. Oral Pathol. Med. 2002; 31:35–44. 48. Hardingham TE, Fosang AJ. Proteoglycans: many forms and many functions. FASEB J. 1992; 6: 861–870. 49. Delehedde M, Lyon M, Sergeant N, Rahmoune H, Fernig DG. Proteoglycans: pericellular and cell surface multireceptors that integrate external stimuli in the mammary gland. J. Mammary Gland Biol. Neoplasia 2001; 6:253–273. 50. Templeton DM. Proteoglycans in cell regulation. Crit. Rev. Clin. Lab. Sci. 1992; 29:141–184. 51. Viklund L, Loo BM, Hermonen J, El-Darwish K, Jaalkanen M, Salmivirta M. Expression and characterization of minican, a recombinant syndecan-1 with extensively truncated core protein. Biochem. Biophys. Res. Comm. 2002; 290:146–152. 52. Simons M, Horowitz A. Syndecan-4 mediated signalling. Cell Signal 2001; 13:855–862. 53. Filmus J, Selleck SB. Glypicans: proteoglycans with a surprise. J. Clin. Invest. 2001; 108:497–501. 54. Rauch U, Feng K, Zhou XH. Neurocan: a brain chondroitin sulfate proteoglycan. Cell Mol. Life Sci. 2001; 58:1842–1856. 55. Cheng SL, Lai CF, Blystone SD, Avioli LV. Bone mineralization and osteoblast differentiation are negatively modulated by integrin alpha(v)beta3. J. Bone Miner. Res. 2001; 16:277–288. 56. Nam T, Moralez A, Clemmons D. Vitronectin binding to IGF binding protein-5 (IGFBP-5) alters IGFBP-5 modulation of IGF-I actions. Endocrinology 2002; 143:30–36. 57. Scatena M, Giachelli C. The alpha(v)beta3 integrin, NF-kappaB, osteoprotegerin endothelial cell survival pathway. Potential role in angiogenesis. Trends Cardiovasc. Med. 2002; 12:83–88. 58. Felding-Habermann B, Cheresh DA. Vitronectin and its receptors. Curr. Opin. Cell Biol. 1993; 5: 864–868. 59. Hughes RC. Galectins as modulators of cell adhesion. Biochimie 2001; 83:667–676. 60. Barondes SH, Cooper DN, Gitt MA, Leffler H. Galectins. Structure and function of a large family of animal lectins. J. Biol. Chem. 1994; 269:20807–20810. 61. Kasai K, Hirabayashi J. Galectins: a family of animal lectins that decipher glycocodes. J. Biochem. 1996; 119:1–8. 62. Cole GJ, Halfter W. Agrin: an extracellular matrix heparan sulfate proteoglycan involved in cell interactions and synaptogenesis. Perspect Dev. Neurobiol. 1996; 3:359–371. 63. Deyst KA, Ma J, Fallon JR. Agrin: toward a molecular understanding of synapse regeneration. Neurosurgery 1995; 37:71–77. 64. Denhardt DT, Noda M, O’Regan AW, Pavlin D, Berman JS. Osteopontin as a means to cope with environmental insults: regulation of inflammation, tissue remodelling, and cell survival. J. Clin. Invest. 2001; 107:1055–1061. 65. Sage EE. Regulation of interactions between cells and extracellular matrix: a command performance on several stages. J. Clin. Invest. 2001; 107:781–783.
Extracellular Matrix Remodeling
25
66. Murphy-Ullrich JE. The de-adhesive activity of matricellular proteins: is intermediate cell adhesion an adaptative state?. J. Clin. Invest. 2001; 107:785–790. 67. Adams JC. Thrombospondins: multifunctional regulators of cell interactions. Annu. Rev. Cell Dev. Biol. 2001; 17:25–51. 68. Bornstein P. Thrombospondins as matricellular modulators of cell function. J. Clin. Invest. 2001; 107:929–934. 69. Jones FS, Jones PL. The tenascin family of ECM glycoprotein: structure, function, and regulation during embryonic development and tissue remodelling. Develop. Dyn. 2000; 218:235–259. 70. Joester A, Faissner A. The structure and function of tenascins in the nervous system. Matrix Biol. 2001; 20:13–22. 71. Brekken RA, Sage EH. SPARC, a matricellular protein: at the crossroads of cell–matrix communication. Matrix Biol. 2000; 19:569–580. 72. Yan Q, Sage EH. SPARC, a matricellular glycoprotein with important biological functions. J. Histochem. Cytochem. 1999; 47:1495–1506. 73. Bosman FT. Integrins: cell adhesives and modulators of cell function. Histochem. J. 1993; 25: 469–477. 74. Hynes RO. Integrins: versatility, modulation and signalling in cell adhesion. Cell 1992; 69:11–25. 75. Critchley DR, Holt MR, Barry ST, Priddle H, Hemmings L, Norman J. Integrin-mediated cell adhesion: the cytoskeletal connection. Biochem. Soc. Symp. 1999; 65:79–99. 76. Gronthos S, Stewart K, Graves SE, Hay S, Simmons PJ. Integrin expression and function on human osteoblast-like cells. J. Bone Miner. Res. 1997; 12:1189–1197. 77. Grzesik WJ. Integrins and bone—cell adhesion and beyond. Arch. Immunol. Ther. Exp. (Warsz) 1997; 45:271–275. 78. Meredith JE, Jere E, Schwartz MA. Integrins, adhesion and apoptosis. Trends Cell Biol. 1997; 7: 146–150. 79. Frisch SM, Ruoslahti E. Integrins and anoikis. Curr. Opin. Cell Biol. 1997; 9:701–706. 80. Winder SJ. The complexities of dystroglycan. Trends Biochem. Sci. 2001; 26:118–124. 81. Henry MD, Campbell KP. Dystroglycan inside and out. Curr. Opin. Cell. Biol. 1999; 11:602–607. 82. Matsumura K, Yamada H, Saito F, Sunada Y, Shimizu T. The role of dystroglycan, a novel receptor of laminin and agrin, in cell differentiation. Histol. Histopathol. 1997; 12:195–203. 83. Vogel W. Discoidin domain receptors: structural relations and functional implications. FASEB J. 1999; 13(Suppl):S77-S82. 84. Chen H, Locke D, Liu Y, Liu C, Kahn ML. The platelet receptor GPVI mediates both adhesion and signaling responses to collagen in a receptor density-dependent fashion. J. Biol. Chem. 2002; 277:3011–3019. 85. Potter E, Bergwitz C, Brabant G. The cadherin–catenin system: implications for growth and differentiation of endocrine tissues. Endocr. Rev. 1999; 20:207–239. 86. Taipale J, Keski-Oja J. Growth factors in the extracellular matrix. FASEB J. 1997; 11:51–59. 87. Dallas SL, Rosser JL, Mundy GR, Bonewald LF. Proteolysis of latent-TGF beta binding protein1 by osteoclasts—a cellular mechanism for release of TGF beta from bone matrix. J. Biol. Chem. 2002; 277:21352–21360. 88. Schoppet M, Chavakis T, Al-Fakhri N, Kanse SM, Preissner KT. Molecular interactions and functional interference between vitronectin and transforming growth factor-beta. Lab. Invest. 2002; 82: 37–46. 89. Eliceiri BP. Integrin and growth factor receptor crosstalk. Circ. Res. 2001; 89:1104–1110. 90. Woods CK, Couchman JR. Syndecans: synergistic activators of cell adhesion. Trends Cell Biol. 1998; 8:189–192. 91. Abercrombie M, Dunn GA. Adhesion of fibroblasts to substratum during contact inhibition observed by interference reflection microscopy. Exp. Cell Res. 1975; 92:57–62. 92. Izzard CS, Lochner LR. Formation of cell-to-substrate contacts during fibroblast motility: an interference-reflection study. J. Cell Sci.; 42:82–116. 93. Zamir E, Geiger B. Molecular complexity and dynamics of cell–matrix adhesion. J. Cell Sci. 2001; 114:3583–3590.
26
Cannas et al.
94. Howe A, Aplin A, Alahari S, Juliano RL. Integrin signal and cell growth control. Curr. Opin. Cell Biol. 1998; 10:220–231. 95. Sastry SK, Burridge K. Focal adhesion: a nexus for intracellular signaling and cytoskeletal dynamics. Exp. Cell Res. 2000; 261:25–36. 96. Muszbek L, Yee VC, Hevessy Z. Blood coagulation factor XIII: structure and function. Thromb. Res. 1999; 94:271–305. 97. Cario E, Goebell H, Dignass AU. Factor XIII modulates intestinal epithelial wound healing in vitro. Scand J Gastroenterol. 1999; 34:485–490. 98. Gorodetsky R, Vexler A, An J, Mou X, Marx G. Haptotactic and growth stimulatory effects of fibrin(ogen) and thrombin on cultured fibroblasts. J Lab Clin Med. 1998; 131:269–280. 99. Chambers RC, Dabbagh K, McAnulty RJ, Gray AJ, Blanc-Brude OP, Laurent GJ. Thrombin stimulates fibroblast procollagen production via proteolytic activation of protease-activated receptor 1. Biochem J. 1998; 333:121–127. 100. Brash JL. The fate of fibrinogen following adsorption at the blood–biomaterial interface. Ann NY Acad Sci. 1987; 516:206–222. 101. Andrade JD, Hlady V. Plasma protein adsorption: the big twelve. Ann NY Acad Sci. 1987; 516: 158–172. 102. Kanagaraja S, Lundstrom I, Nygren H, Tengvall P. Platelet binding and protein adsorption to titanium and gold after short time exposure to heparinized plasma and whole blood. Biomaterials. 1996; 17: 2225–2232. 103. De La Cruz C, Haimovich B, Greco RS. Immobilized IgG and fibrinogen differentially affect the cytoskeletal organization and bactericidal function of adherent neutrophils. J Surg Res. 1998; 80: 28–34. 104. Nygren H, Eriksson C, Lausmaa J. Adhesion and activation of platelets and polymorphonuclear granulocyte cells at TiO2 surfaces. J Lab Clin Med. 1997; 129:35–46. 105. Nygren H, Broberg M. Specific activation of platelets by surface-adsorbed plasma proteins. J Biomater Sci Polym Ed. 1998; 9:817–831. 106. Stocum DL. Limb regeneration: re-entering the cell cycle. Curr Biol. 1999; 9:644–666. 107. Miller RA, Britigan BE. The formation and biologic significance of phagocyte-derived oxidants. J Invest Med. 1995; 43:39–49. 108. Hyslop PA, Hinshaw DB, Scraufstatter IU, Cochrane CG, Kunz S, Vosbeck K. Hydrogen peroxide as a potent bacteriostatic antibiotic: implications for host defense. Free Radical Biol Med. 1995; 19:31–37. 109. Blouin E, Halbwachs-Mecarelli L, Rieu P. Redox regulation of beta2-integrin CD11b/CD18 activation. Eur J Immunol. 1999; 29:3419–3431. 110. Salmon JE, Millard SS, Brogle NL, Kimberly RP. Fc gamma receptor IIIb enhances Fc gamma receptor IIa function in an oxidant-dependent and allele-sensitive manner. J Clin Invest. 1995; 95: 2877–2885. 111. Kamata H, Hirata H. Redox regulation of cellular signalling. Cell Signal. 1999; 1:1–14. 112. Kasama T, Kobayashi K, Fukushima T, Tabata M, Ohno I, Negishi M, Ide H, Takahashi T, Niwa Y. Production of interleukin 1–like factor from human peripheral blood monocytes and polymorphonuclear leukocytes by superoxide anion: the role of interleukin 1 and reactive oxygen species in inflamed sites. Clin Immunol Immunopathol. 1989; 53:439–448. 113. Metinko AP, Kunkel SL, Standiford TJ, Strieter RM. Anoxia-hyperoxia induces monocyte-derived interleukin-8. J Clin Invest. 1992; 90:791–798. 114. Baeuerle PA, Henkel T. Function and activation of NF-kappa B in the immune system. Annu Rev Immunol. 1994; 12:141–179. 115. Oclura M, Kitamura M. Oxidant stress incites spreading of macrophages via ERK/p38-SRE signalling pathways. In: Tisher CC, ed. American Society of Nephrology 31st Annual Meeting. Philadelphia. PA: Williams & Wilkins, 1998:429A. 116. Madesh M, Balasubramanian KA. Activation of liver mitochondrial phospholipase A2 by superoxide. Arch Biochem Biophys. 1997; 346:187–192. 117. Muller M, Sorrell TC. Oxidative stress and the mobilisation of arachidonic acid in stimulated human platelets: role of hydroxyl radical. Prostaglandins 1997; 54:493–510.
Extracellular Matrix Remodeling
27
118. Eckes B, Zigrino P, Kessler D, Holtkotter O, Shephard P, Mauch C, Krieg T. Fibroblast–matrix interactions in wound healing and fibrosis. Matrix Biol. 2000; 19:325–332. 119. Desmouliere A, Redard M, Darby I, Gabbiani G. Apoptosis mediates the decrease in cellularity during the transition between granulation tissue and scar. Am J Pathol. 1995; 146:56–66. 120. Tredget EE, Nedelec B, Scott PG, Ghahary A. Hypertrophic scars, keloids, and contractures. The cellular and molecular basis for therapy. Surg Clin North Am. 1997; 77:701–730. 121. Takei I, Takagi M, Santavirta S, Ida H, Ishii M, Ogino T, Ainola M, Konttinen YT. Messenger ribonucleic acid expression of 16 matrix metalloproteinases in bone–implant interface tissues of loose artificial hip joints. J Biomed Mater Res. 2000; 52:613–620. 122. Mauch C, Oono T, Eckes B, Krieg T. Cytokines and wound healing. In: Luger TA , Schwarz T, eds(). Epidermal growth factors and cytokines. New York: Marcel Dekker, 1994:325–344. 123. Jacobs JJ, Roebuck KA, Archibeck M, Hallab NJ, Glant TT. Osteolysis: basic science. Clin. Orthop. 2001; 393:71–77. 124. Takagi M, Santavirta S, Ida H, Ishii M, Takei I, Niissalo S, Ogino T, Konttinen YT. High-turnover periprosthetic bone remodeling and immature bone formation around loose cemented total hip joints. J. Bone Miner. Res. 2001; 16(1):79–88. 125. Bosetti M, Ottani V, Kozel D, De Pasquale V, Raspanti M, Ruggeri A, Cannas M. Macrophages response to hydroxyapatite, natural apatite and alumina: a comparative evaluation in vitro. Biomaterials 1999; 20(4):363–370. 126. Bosetti M, Masse` R, Navone R, Cannas M. Biochemical and histological evaluation of human synovial-like membrane around failed total hip replacement prostheses during in vitro mechanical loading. J Mat. Sci., Materials in Medicine; 12(8):693–698. 127. Pioletti DP, Leoni L, Genini D, Takei H, Du P, Corbeil J. Gene expression analysis of osteoblastic cells contacted by orthopedic implant particles. J Biomed. Mater. Res. 2002; 61(3):408–420. 128. Bosetti M, Navone R, Rizzo E, Cannas M. Histochemical and morphometric observations on the new tissue formed around mammary expanders coated with pyrolytic carbon. J. Biomed. Mater. Res. 1998; 40:307–313. 129. Davies JE. In vitro modeling of the bone/implant interface. Anat. Rec. 1996; 245(2):426–445. 130. Schmalzried TP, Maloney WJ, Jasty M, Kwong LM, Harris WH. Autopsy studies of the bone–cement interface in well-fixed cemented total hip arthoplasties. J. Arthoplasty 1993; 8(2):179–188. 131. Glant TT, Jacobs JJ, Mikecz K, Yao J, Chubinskaja S, Williams JM, Urban RL, Shanbhag AS, Lee SH, Sumner DR. Particulate-induced, prostaglandin- and cytokine-mediated bone resorption in an experimental system and in failed joint replacements. Am. J. Ther. 1996; 3(1):27–41. 132. Yokohama Y, Matsumoto T, Hirakawa M, Kuroki Y, Fujimoto N, Imai K, Okada Y. Production of matrix metalloproteinases at the bone–implant interface in loose total hip replacements. Lab. Invest. 1995; 73(6):899–911. 133. Takagi M, Santavirta S, Ida H, Ishii M, Mandelin J, Konttinen YT. Matrix metalloproteinases and tissue inhibitors of metalloproteinases in loose artificial hip joints. Clin. Orthop. 1998; 352:35–45. 134. Li TF, Santavirta S, Virtanen I, Kononen M, Takagi M, Konttinen YT. Increased expression of EMMPRIN in the tissue around loosened hip prostheses. Acta Orthop. Scand. 1999; 70(5):446–451. 135. Nagase H. in Zinc Metalloproteases in Health and Disease( Hoopei NM, ed). London: Taylor & Francis, 1996:153–204. 136. Fini ME, Cook JR, Mohan R, Brinckerhoft CE. Regulation of matrix metalloproteinase gene expression In Matrix Metalloproteinases( Parks WC , Mecham RP, eds). San Diego: Academic Press, 1998:299–351. 137. Vihinen P, Kahari VM. Matrix metalloproteinases in cancer: prognostic markers and therapeutic targets. Int. J. Cancer 2002; 99:157–166. 138. Van Wart HE, Birkedal-Hansen H. The cysteine switch: a principle of regulation of metalloproteinase activity with potential applicability to the entire matrix metalloproteinase gene family. Proc Natl Acad Sci USA. 1990; 87:5578–5582. 139. Bode W, Gomis-Ruth FX, Stockler W. Astacins, serralysins, snake venom and matrix metalloproteinases exhibit identical zinc-binding environments (HEXXHXXGXXH and Met-turn) and topologies and should be grouped into a common family, the ‘metzincins’. FEBS Lett. 1993; 331:134–140.
28
Cannas et al.
140. Allan JA, Docherty AJ, Barker PJ, Huskisson NS, Reynolds JJ, Murphy G. Binding of gelatinases A and B to type-I collagen and other matrix components. Biochem J. 1995; 309:299–306. 141. Steffensen B, Wallon UM, Overall CM. Extracellular matrix binding properties of recombinant fibronectin type II–like modules of human 72-kDa gelatinase/type IV collagenase. High affinity binding to native type I collagen but not native type IV collagen. J Biol Chem. 1995; 270: 11555–11566. 142. Gomis-Ruth FX, Gohlke U, Betz M, Knauper V, Murphy G, Lopez-Otin C, Bode W. The helping ˚ crystal structure of its C-terminal haemopexin-like domain. hand of collagenase-3 (MMP-13): 2.7 A J Mol Biol. 1996; 264:556–566. 143. Bode W. A helping hand for collagenases: the haemopexin-like domain. Structure 1995; 3:527–530. 144. Clark IM, Cawston TE. Fragments of human fibroblast collagenase. Purification and characterization. Biochem J. 1989; 263:201–206. 145. Strongin AY, Collier I, Bannikov G, Marmer BL, Grant GA, Goldberg GI. Mechanism of cell surface activation of 72-kDa type IV collagenase. Isolation of the activated form of the membrane metalloprotease. J Biol Chem. 1995; 270:5331–5338. 146. Guo H, Zucker S, Gordon MK, Toole BP, Biswas C. Stimulation of matrix metalloproteinase production by recombinant extracellular matrix metalloproteinase inducer from transfected Chinese hamster ovary cells. J Biol Chem. 1997; 272:24–27. 147. Seltzer JL, Lee AY, Akers KT, Sudbeck B, Southon EA, Wayner EA, Eisen AZ. Activation of 72kDa type IV collagenase/gelatinase by normal fibroblasts in collagen lattices is mediated by integrin receptors but is not related to lattice contraction. Exp Cell Res. 1994; 213:365–374. 148. Haas TL, Davis SJ, Madri JA. Three-dimensional type I collagen lattices induce coordinate expression of matrix metalloproteinases MT1-MMP and MMP-2 in microvascular endothelial cells. J Biol Chem. 1998; 273:3604–3610. 149. Aoudjit F, Potworowski EF, St-Pierre Y. Bi-directional induction of matrix metalloproteinase-9 and tissue inhibitor of matrix metalloproteinase-1 during T lymphoma/endothelial cell contact: implication of ICAM-1. J Immunol. 1998; 160:2967–2973. 150. Xie B, Laouar A, Huberman E. Fibronectin-mediated cell adhesion is required for induction of 92kDa type IV collagenase/gelatinase (MMP-9) gene expression during macrophage differentiation. The signaling role of protein kinase C-beta. J Biol Chem. 1998; 273:11576–11582. 151. Gilles C, Polette M, Seiki M, Birembaut P, Thompson EW. Implication of collagen type I–induced membrane–type 1–matrix metalloproteinase expression and matrix metalloproteinase-2 activation in the metastatic progression of breast carcinoma. Lab Invest. 1997; 76:651–660. 152. Werb Z, Tremble PM, Behrendtsen O, Crowley E, Damsky CH. Signal transduction through the fibronectin receptor induces collagenase and stromelysin gene expression. J Cell Biol. 1989; 109: 877–889. 153. Kheradmand F, Werner E, Tremble P, Symons M, Werb Z. Role of Rac1 and oxygen radicals in collagenase-1 expression induced by cell shape change. Science. 1998; 280:898–902. 154. Reunanen N, Westermarck J, Hakkinen L, Holmstrom TH, Elo I, Eriksson JE, Kahari VM. Enhancement of fibroblast collagenase (matrix metalloproteinase-1) gene expression by ceramide is mediated by extracellular signal–regulated and stress-activated protein kinase pathways. J Biol Chem. 1998; 273:5137–5145. 155. Nagase H. Activation mechanisms of matrix metalloproteinases. Biol Chem. 1997; 378:151–160. 156. Carmeliet P, Moons L, Lijnen R, Baes M, Lemaitre V, Tipping P, Drew A, Eeckhout Y, Shapiro S, Lupu F, Collen D. Urokinase-generated plasmin activates matrix metalloproteinases during aneurysm formation. Nat Genet. 1997; 17:439–444. 157. Sato H, Takino T, Okada Y, Cao J, Shinagawa A, Yamamoto E, Seiki M. A matrix metalloproteinase expressed on the surface of invasive tumour cells. Nature. 1994; 370:61–65. 158. Gomez DE, Alonso DF, Yoshiji H, Thorgeirsson UP. Tissue inhibitors of metalloproteinases: structure, regulation and biological functions. Eur J Cell Biol. 1997; 74:111–122. 159. Hayakawa T, Yamashita K, Ohuchi E, Shinagawa A. Cell growth–promoting activity of tissue inhibitor of metalloproteinases-2 (TIMP-2). J Cell Sci. 1994; 107:2373–2379.
Extracellular Matrix Remodeling
29
160. Chesler L, Golde DW, Bersch N, Johnson MD. Metalloproteinase inhibition and erythroid potentiation are independent activities of tissue inhibitor of metalloproteinases-1. Blood. 1995; 86: 4506–4515. 161. Zuo J, Ferguson TA, Hernandez YJ, Stetler-Stevenson WG, Muir D. Neuronal matrix metalloproteinase-2 degrades and inactivates a neurite-inhibiting chondroitin sulfate proteoglycan. J Neurosci. 1998; 18:5203–5211. 162. Giannelli G, Falk-Marzillier J, Schiraldi O, Stetler-Stevenson WG, Quaranta V. Induction of cell migration by matrix metalloprotease-2 cleavage of laminin-5. Science. 1997; 277:225–228. 163. Pilcher BK, Dumin JA, Sudbeck BD, Krane SM, Welgus HG, Parks WC. The activity of collagenase1 is required for keratinocyte migration on a type I collagen matrix. J Cell Biol. 1997; 137: 1445–1457. 164. Holliday LS, Welgus HG, Fliszar CJ, Veith GM, Jeffrey JJ, Gluck SL. Initiation of osteoclast bone resorption by interstitial collagenase. J Biol Chem. 1997; 272:22053–22058. 165. Shapiro SD. Matrix metalloproteinase degradation of extracellular matrix: biological consequences. Curr Opin Cell Biol. 1998; 10:602–608. 166. Reponen P, Sahlberg C, Munaut C, Thesleff I, Tryggvason K. High expression of 92-kD type IV collagenase (gelatinase B) in the osteoclast lineage during mouse development. J Cell Biol. 1994; 124:1091–1102. 167. Vu TH, Shipley JM, Bergers G, Berger JE, Helms JA, Hanahan D, Shapiro SD, Senior RM, Werb Z. MMP-9/gelatinase B is a key regulator of growth plate angiogenesis and apoptosis of hypertrophic chondrocytes. Cell. 1998; 93:411–422. 168. Harris WH, Schiller AL, Scholler JM, Freiberg RA, Scott R. Extensive localized bone resorption in the femur following total hip replacement. J Bone Joint Surg Am. 1976; 58:612–618. 169. Carlsson AS, Gentz CF, Linder L. Localized bone resorption in the femur in mechanical failure of cemented total hip arthroplasties. Acta Orthop Scand. 1983; 54:396–402. 170. Scott WW, Riley LH, Dorfman HD. Focal lytic lesions associated with femoral stem loosening in total hip prosthesis. AJR Am J Roentgenol. 1985; 144:977–982. 171. Revell PA, Weightman B, Freeman MA, Roberts BV. The production and biology of polyethylene wear debris. Arch Orthop Trauma Surg. 1978; 91:167–181. 172. Maloney WJ, Jasty M, Rosenberg A, Harris WH. Bone lysis in well-fixed cemented femoral components. J Bone Joint Surg Br. 1990; 72:966–970. 173. Evans EM. Metal sensitivity as a cause of bone necrosis and loosening of the prosthesis in total joint replacement. J Bone Joint Surg Br. 1974; 56:626–642. 174. Goldring SR, Schiller AL, Roelke M, Rourke CM, O’Neil DA, Harris WH. The synovial-like membrane at the bone–cement interface in loose total hip replacements and its proposed role in bone lysis. J Bone Joint Surg Am. 1983; 65:575–584. 175. Goldring SR, Jasty M, Roelke MS, Rourke CM, Bringhurst FR, Harris WH. Formation of a synoviallike membrane at the bone–cement interface. Its role in bone resorption and implant loosening after total hip replacement. Arthritis Rheum. 1986; 29:836–842. 176. Eskola A, Santavirta S, Konttinen YT, Hoikka V, Tallroth K, Lindholm TS. Cementless revision of aggressive granulomatous lesions in hip replacements. J Bone Joint Surg Br. 1990; 72:212–216. 177. Santavirta S, Hoikka V, Eskola A, Konttinen YT, Paavilainen T, Tallroth K. Aggressive granulomatous lesions in cementless total hip arthroplasty. J Bone Joint Surg Br. 1990; 72:980–984. 178. Kim KJ, Rubash HE, Wilson SC, D’Antonio JA, McClain EJ. A histologic and biochemical comparison of the interface tissues in cementless and cemented hip prostheses. Clin Orthop. 1993; 287: 142–152. 179. Kim KJ, Chiba J, Rubash HE. In vivo and in vitro analysis of membranes from hip prostheses inserted without cement. J Bone Joint Surg Am. 1994; 76:172–180. 180. Chiba J, Rubash HE, Kim KJ, Iwaki Y. The characterization of cytokines in the interface tissue obtained from failed cementless total hip arthroplasty with and without femoral osteolysis. Clin Orthop. 1994; 300:304–312. 181. Ishiguro N, Ito T, Kurokouchi K, Iwahori Y, Nagaya I, Hasegawa Y, Iwata H. mRNA expression of matrix metalloproteinases and tissue inhibitors of metalloproteinase in interface tissue around implants in loosening total hip arthroplasty. J Biomed Mater Res. 1996; 32:611–617.
30
Cannas et al.
182. Takei I, Takagi M, Santavirta S, Ida H, Hamasaki M, Ishii M, Fukushima S, Ogino T, Konttinen YT. Matrix metalloproteinases and tissue inhibitors of metalloproteinases in joint fluid of the patients with loose artificial hip joints. J Biomed Mater Res. 1999; 45:175–83. 183. Takagi M, Konttinen YT, Lindy O, Sorsa T, Kurvinen H, Suda A, Santavirta S. Gelatinase/type IV collagenases in the loosening of total hip replacement endoprostheses. Clin Orthop. 1994; 306: 136–144. 184. Takagi M, Konttinen YT, Santavirta S, Sorsa T, Eisen AZ, Nordsletten L, Suda A. Extracellular matrix metalloproteinases around loose total hip prostheses. Acta Orthop Scand. 1994; 65:281–286. 185. Syggelos SA, Eleftheriou SC, Giannopoulou E, Panagiotopoulos E, Aletras AJ. Gelatinolytic and collagenolytic activity in periprosthetic tissues from loose hip endoprostheses. J Rheumatol. 2001; 28:1319–1329. 186. Quinn J, Joyner C, Triffitt JT, Athanasou NA. Polymethylmethacrylate-induced inflammatory macrophages resorb bone. J Bone Joint Surg Br. 1992; 74:652–658. 187. Amstutz HC, Campbell P, Kossovsky N, Clarke IC. Mechanism and clinical significance of wear debris–induced osteolysis. Clin Orthop.; 1992; 276:7–18. 188. Maloney WJ, Smith RL. Periprosthetic osteolysis in total hip arthroplasty: the role of particulate wear debris. Instr Course Lect. 1996; 45:171–182. 189. Yu Y, Walsh WR, Sonnabend DH, Yang JL, Bonar F, Markovic B, Bruce W, Kohan L, Neil M. Cytokines and matrix metalloproteinases mRNA expression in archival human tissues from failed total hip arthroplasty using in situ hybridization and color video image analysis. Bull Hosp Jt Dis. 1998; 57:23–29. 190. Ishiguro N, Kojima T, Ito T, Saga S, Anma H, Kurokouchi K, Iwahori Y, Iwase T, Iwata H. Macrophage activation and migration in interface tissue around loosening total hip arthroplasty components. J Biomed Mater Res.; 1997; 35(3):399–406. 191. Okada Y, Naka K, Kawamura K, Matsumoto T, Nakanishi I, Fujimoto N, Sato H, Seiki M. Localization of matrix metalloproteinase 9 (92-kilodalton gelatinase/type IV collagenase ⳱ gelatinase B) in osteoclasts: implications for bone resorption. Lab Invest. 1995; 72:311–322. 192. Kaneko M, Tomita T, Nakase T, Ohsawa Y, Seki H, Takeuchi E, Takano H, Shi K, Takahi K, Kominami E, Uchiyama Y, Yoshikawa H, Ochi T. Expression of proteinases and inflammatory cytokines in subchondral bone regions in the destructive joint of rheumatoid arthritis. Rheumatology (Oxford) 2001; 40:247–255. 193. al-Saffar N. The osteogenic properties of the interface membrane at the site of orthopedic implants: the impact of underlying joint disease. J Long Term Eff Med Implants. 1999; 9:23–45. 194. Bosetti M, Reno` F, Bracco P, Costa L, Cannas M. UHMWPE oxidation reduces MMP-2 secretion in human osteoblast-like cells in vitro: a mechanism of modulation for ECM. J Biomed Mat Res. 195. Reno` F, Costa L, Cannas M. UHMWPE oxidation increases matrix metalloproteinase 2 (MMP-2) release in human fibroblast. J Mat Sci Mat Med 2002; 13:403–406. 196. Reno` F, Bracco P, Costa L, Cannas M. UHMWPE oxidation induces cytotoxicity and MMP-9 release in PBMCs. Biomaterials 2002; 23:3645–3650.
2 Elastic Protein-Based Biomaterials: Elements of Basic Science, Controlled Release, and Biocompatibility Dan W. Urry Biotechnology Institute, University of Minnesota, St. Paul, Minnesota, U.S.A., and Bioelastics Research, Ltd., Birmingham, Alabama, U.S.A.
T. Copper Woods, Larry C. Hayes, Jie Xu, David T. McPherson, and Timothy M. Parker Bioelastics Research, Ltd., Birmingham, Alabama, U.S.A.
Masamichi Iwama Bioelastics Japan, Yokohama, Japan
Masakazu Furutan and Toshio Hayashi Research Institute of Advanced Science and Technology, Osaka Prefecture University Osaka, Japan
Mitsuhiro Murata Japan Synthetic Rubber Co., Ltd. Tokyo, Japan
I. INTRODUCTION A. The Unique Opportunity Provided by Elastic Protein-Based Materials The basic model protein of focus here, (Gly-Val-Gly-Val-Pro)n, [or simply (GVGVP)n], exhibits both dominantly entropic elasticity and the energy conversions of thermomechanical and chemomechanical transduction. It achieves these functional capacities with only mildly hydrophobic (Val and Pro) and neutral (Gly) side chains, that is, without any functional groups, save the peptide moiety itself. These fundamental functions achieved with such simplicity of composition become remarkably compounded due to the capacity of protein biosynthesis to position precisely any one of 20 different amino acid side chains at any location along the protein sequence. The result is a family of elastic, and even plastic, protein-based materials of near endless functional versatility. The application of a wide range of experimental methodologies, of thermodynamic and statistical mechanical analyses, and of molecular mechanics and dynamics calculations to protein-based polymers of such demonstrated functional diversity provides a unique opportunity 31
32
Urry et al.
to gain a level of understanding and utilization afforded by few, and possibly no other, chain molecules. All of this potential can be turned toward biomaterials applications. This review provides one glimpse into this potential. B. Elastic Protein Mechanics Characterized by Two Interlinked but Distinct Physical Processes The most effective use of elastic protein-based biomaterials requires an understanding of the molecular mechanics responsible for two interlinked but distinct physical processes—the development of entropic elastic force and contraction achievable by diverse energy sources[1]. The physical basis of desirable ideal (entropic) protein elasticity has been a concern for half a century and has been intensely contested for the last three decades. There are three contending mechanisms for protein elasticity: the classical (random chain network) theory of rubber elasticity [2,3], the solvent entropy theory [4,5], and the damping of internal chain dynamics on extension [6,7]. Data and analyses, briefly reviewed below, argue that entropic elastic force arises due to a damping of internal chain dynamics on extension [1,8,9]. Observation of entropic elasticity in atomic force microscopic (AFM) single-chain force-extension studies eliminates the random chain network mechanism, because its formulation requires a random network of chains with a Gaussian distribution of end-to-end chain lengths [10]. Isometric contraction by inverse temperature transition, that is, the development of entropic elastic force at fixed length whether due to thermally or chemically driven hydrophobic association, eliminates the solvent entropy mechanism as a source of entropic elastic force [1]. Versatile protein contractility arises due to hydrophobic association of protein sequences with the consequence of extension of interconnecting chain segments resulting in an increase in entropic elastic force due to damping of internal chain dynamics within the interconnecting chain segments. Versatility of energy inputs that drive contraction arises out of the many variables that control hydrophobic association/disassociation in aqueous systems [11,12]. C. Proposed Basis of Remarkable Biocompatibility of Elastic ProteinBased Materials The basic elastic protein-based polymer sequence, (GVGVP)n, originally observed in the mammalian elastic protein, elastin [13,14], exhibits a remarkable biocompatibility [15]. For example, it has not been possible to obtain monoclonal antibodies, even with very weak titers, to this elastic protein-based polymer [16]. When adequately purified, the phase-separated (hydrophobically associated) state of this elastic protein-based polymer appears to be simply ignored by the host. As will be argued herein, the dynamic nature of the phase-separated state at physiological temperatures constitutes a barrier to interaction that would otherwise be required for identification as a foreign material. Without eliciting an apparent inflammatory response, the hydrophobically disassociated (solution) state proteolytically degrades within the host. Thus, to control the state is to control the rate of degradation. D. Hydration-Mediated Apolar–Polar Repulsion Provides for Special Controlled Release Devices There exists a competition for hydration between apolar (hydrophobic) and polar (e.g., charged) groups constrained to coexist along the protein chain sequence [11]. In 1937 Butler reported the key realization that hydration of hydrophobic groups is a favorable exothermic reaction [17], but that solubility ultimately becomes lost due to the increase in order of water on going from
Elastic Protein-Based Biomaterials
33
bulk water to hydrophobic hydration. In short, a limited amount of hydrophobic hydration occurs, but too much hydrophobic hydration drives insolubility on association of hydrophobic groups, obviously, with essentially complete loss of hydrophobic hydration. Charged species, such as carboxylates (—COOⳮ) and amino (—NH3Ⳮ) groups, when sufficiently proximal to hydrophobic groups, destructure hydrophobic hydration in the process of achieving adequate hydration of their own [11,18]. This has two consequences: increased solubility of hydrophobic groups as their potential for hydrophobic hydration decreases and pKa shifts for the ionizable species. When there is abundant bulk water available, charged species form with a characteristic pKa, but when ionizable species must destructure hydrophobic hydration to achieve adequate hydration, they are at a higher free energy, as reflected by significant pKa shifts and associated positive cooperativity apparent in the degree of ionization versus pH curve. In the process of attempting to minimize their free energies of hydration, hydrophobic groups and charged groups each seek out water unperturbed by the other. This constitutes an apolar–polar repulsive free energy of hydration, ⌬Gap. Neutralization, for example by ion pairing, of a charged group with a hydrophobic-induced pKa shift lowers the free energy of the charged group, relaxes the need for hydration, and can allow for reconstitution of so much hydrophobic hydration that hydrophobic association results. For elastic protein-based polymers hydrophobic association is observed as a phase separation, called an inverse temperature transition because it is a transition to a more ordered state for the protein part of the system as the result of raising the temperature. When an oppositely charged drug neutralizes polymer charge, the drug provides the chemical energy for phase separation, that is, for formation of the drug delivery vehicle. As will be shown below, the phase-separated state of drug plus elastic protein-based polymer becomes a drug delivery vehicle. The level of release depends on several factors—the surface area of the drug delivery vehicle available to the surrounding medium, the ionic composition of the medium, and, most fundamentally, the decrease in free energy on ionpairing of drug with polymer. Of course, lower release levels correlate with larger decreases in free energy on ionpairing of drug with polymer. The decrease in free energy on ionpairing is proportional to the hydrophobic-induced pKa shift exhibited by the ionizable groups of the polymer in the absence of drug. When the polymer is not cross-linked, the polymer, which is the drug delivery vehicle, disperses as the drug is released [16], unless ion exchange occurs to maintain the phaseseparated state. The special cases of the loading and release of dexamethasone phosphate and betamethasone phosphate, reported below, demonstrate the effectiveness of this design of drug delivery vehicle using positively charged elastic protein-based polymers. E. The Formation of Nanoparticles Under selected conditions of concentration of poly(GVGVP) and temperature, quasi–elastic light scattering demonstrated the occurrence of stable nanoparticles [19]. As will also be reported here, quite uniform nanoparticles can be stably formed and fixed by cross-linking using the compositions—(GVGVP)251, glutamic acid residue–containing polymer with increased hydrophobicity exhibiting large pKa shifts, and lysine-containing polymers with pKa-shifted amino functions. This means that controlled release of the type described above becomes possible with cross-linked nanoparticles. II. MECHANICS OF ELASTIC PROTEIN-BASED BIOMATERIALS Two elements of mechanics—the basis of ideal elasticity and the process of hydrophobic association—dominate consideration of elastic protein-based materials. Proper design and effective use
34
Urry et al.
of elastic protein-based polymers as biomaterials requires clarification of these two elements of mechanics. Furthermore, elastic protein-based polymers become model chain molecules with which to determine the molecular basis of entropic elasticity and with which to develop the comprehensive hydrophobic effect for chain molecules in general. A. Mechanism of Entropic Elasticity As noted in the section I, there are basically three proposed mechanisms for entropic elasticity. The key arguments delineating these mechanisms are briefly considered in the immediately following text. 1. The Classical Rubber (Random Chain Network) Theory of Entropic Elasticity. In this argument, stretching of a random network of chains causes the chains to become more aligned and shifted from a Gaussian distribution of end-to-end chain lengths between crosslinks [10]. In this mechanism the chains are random; the chains do not exhibit regular structure. So the presence of a regular structure, as revealed by the presence of mechanical resonances in the dielectric relaxation spectra of the elastic protein-based polymers (see Figs. 1A and B), eliminates this mechanism. Furthermore, the ideal (entropic) elasticity exhibited by a single chain instead of requiring a network of random chains, as demonstrated previously [8,9], also eliminates this mechanism. Nonetheless, calculations of the distribution of end-to-end chain lengths, as the result of rotations about one bond at a time, do provide approximate estimates of entropic elastic force. The fundamental flexibility of the chain (the chain entropy) is reasonably evaluated in this way. The natural constraint of the elasticity experiment, however, requires consideration of the coupling of rotations about two or more bonds at a time. In the elasticity experiment, the ends of an individual chain are fixed in space and then extended by a fixed amount with the result of an increase in force. So the difference in elastic force occurs without an experimental sampling of a distribution of end-to-end chain lengths. The computation reasonably becomes one of determining the difference in chain entropies calculated at two different extensions of the same chain length. The way that this happens of course is by rotation about one bond while allowing compensating rotation about another bond or bonds in such a way that the ends remain fixed in space. This, in fact, becomes a representation of the mechanism referred to as the damping of internal chain dynamics on extension [6,7], as discussed below. Simply stated, as the distance between the ends of a chain of a given number of repeating units becomes greater, rotation about bonds becomes more limited. In terms of statistical mechanics the volume in phase space becomes less, and volume in phase space is the definition of entropy. So increase in elastic force occurs with a decrease in chain entropy of chain segments that sustain the force. Some argue that it is the decrease in entropy of chain and surrounding water, as considered below. 2. The solvent entropy mechanism (the decrease in entropy on stretching due to bulk water forming lower entropy hydrophobic hydration around newly exposed hydrophobic groups). Weis-Fogh and Andersen [4] proposed that a decrease in solvent entropy occurred on extension of elastin as the result of exposure of hydrophobic groups to water. On stretching, an exothermic hydration of hydrophobic groups occurs with a decrease in the entropy of water. This mechanism maintains computational adherents up to the present time [5,20]. Experimental studies, wherein a chosen solvent mixture largely removes the solvent entropy change, demonstrate an increase in entropic elastic force rather than a decrease [21]. Such a result seriously questions the validity of this source of negative entropy for the development of entropic elastic force (see Fig. 2A). The isometric contraction experiment, the development
Elastic Protein-Based Biomaterials
35
Figure 1 Dielectric relaxation spectra of the (GVGIP)n member of the family of elastic protein–based polymers showing the development of mechanical resonances on formation of the hydrophobically associated, phase-separated, and more ordered state achieved by raising the temperature. (A) The low frequency range of the spectrum (0.1 to 100 kHz) with the acoustic frequency range being in the lower two-thirds of the frequency range. As the temperature is raised from below to above the temperature range of the inverse temperature transition of hydrophobic association, there develops an intense relaxation centered near 3 kHz. (From Ref. 9.) (B) The 1-MHz to 1-GHz frequency range of the dielectric relaxation spectrum showing the development of an intense relaxation centered near 5 MHz that continues to grow in intensity on raising the temperature up to 60⬚C. (From Ref 33.)
of entropic elastic force at fixed length, however, eliminates the solvent entropy mechanism for any polymer in water that exhibits a hydrophobic association transition during the isometric contraction. For amphiphilic polymers in water such as protein, isometric contraction results when hydrophobic association due to an inverse temperature transition occurs at fixed length. This can be shown both with thermally driven and chemically driven hydrophobic association when the protein or other amphiphilic polymer in water is kept at fixed length. In particular, the thermoelasticity experiment used to assess the fraction of entropic elastic force is carried out under conditions of fixed length. As the temperature is raised from below to above the onset temperature for hydrophobic association, the elastic force develops until the hydrophobic association transition is complete and a plateau results in a plot of ln(force/T⬚K) versus temperature
36
Urry et al.
Figure 2 (A) Differential scanning calorimetry data on the endothermic heat of the transition of poly(GVGVP) in water and in increasing amounts of ethylene glycol (EG). The EG is increased until the heat of the transition approaches zero. Since the heat of the transition divided by the temperature at which the transition occurs is the entropy of the transition and since the heat of the transition is almost entirely due to the heat required to destructure hydrophobic hydration to form bulk water, the result is to remove the solvent entropy contribution to the transition from the hydrophobically associated to the disassociated state. (B) Thermoelasticity study of an elastic matrix of poly(GVGVP) formed on ␥-irradiation cross linking. The sample is equilibrated at 40⬚C and stretched to a fixed length. The temperature is reduced, while keeping the sample at the extended length, equilibrated at a temperature below that of the phase transition; then, continuing to maintain a constant length, the temperature is slowly raised. The development of force plotted as the log[f/T], f divided by the temperature in degrees Kelvin ⬚K, is given as a function of temperature. The near zero slope reached at higher temperatures indicates that the developed force is 90% or more entropic in origin. Very significantly the rapid increase in force as the hydrophobic folding transition proceeds occurs while the solvent entropy change must be positive. During the hydrophobic folding of the inverse temperature transition, more ordered, low-entropy hydrophobic hydration becomes less ordered, higher-entropy bulk water, as the 90% entropic elastic force develops due to a decrease in entropy. This eliminates solvent entropy change as the source of the entropic elastic force. (From Ref. 21.)
Elastic Protein-Based Biomaterials
37
(see Fig. 2B). From the slope of the plateau in Fig. 2B, the developed force calculates to be 90% or more entropic for cross-linked poly(GVGVP) [21]. During a hydrophobic association transition, low entropy (more ordered) hydrophobic hydration becomes higher entropy (less ordered) bulk water. In the above experiment, an elastic force that is 90% entropic, due to a decrease in entropy, develops during the hydrophobic association transition as hydrophobic hydration becomes higher entropy bulk water. Chemically driven isometric contraction gives the same result [1,22]. Under conditions of isometric contraction due to hydrophobic association, the solvent entropy change is of the wrong sign, positive instead of negative, and could not contribute to the negative entropy change in the development of entropic elastic force. Therefore, one concludes that a change in solvent entropy does not contribute directly to the entropic component of elastic force. Even at the outset, it would seem apparent that the continuous backbone of a polymer chain sustains force, not the surrounding water molecules [1,3]. The change in entropy of water, however, is central to hydrophobic association that determines polymer structure and function, as occurs in energy conversion [1,11], and the hydrophobic association of certain chain segments results in the extension of interconnecting chain segments that do sustain the force [11]. 3. Decrease in Entropy on Extension Due to the Damping of Internal Chain Dynamics Development of a molecular structure for (GVGVP)n represented in Fig. 3 utilized the physical methods of proton and carbon-13 nuclear magnetic resonance, of Raman, infrared, circular dichroism, and ultraviolet absorption spectroscopies; of transmission electron microscopy with negative staining and optical diffraction; and of x-ray diffraction of cyclic analogs, and utilized molecular mechanics calculations constrained by allowed ranges of torsion angles and hydrogen bonding determined from NMR [23,24]. The molecular structure represented in Fig. 3E immediately suggested an entropic elastic mechanism of the damping of internal chain dynamics on extension [6,23] and just as abruptly pointed to a set of chemical tests in which the Gly residues were replaced by D-alanine and Lalanine residues [25–28]; to physical tests involving NMR relaxation and dielectric relaxation methodologies [29–33]; and to computational tests [6,7] of the mechanism. Furthermore, using molecular mechanics to calculate the net dipole moment of the (VPGVG) permutation of the pentamer and of a D-alanine analog of the pentamer reproduced the dielectric relaxation results [34]. Thus, the concept of the damping of internal chain dynamics on extension was placed on a solid experimental and theoretical foundation. Nonetheless, so entrenched was the random chain network theory that it took the AFM single-chain force-extension results and the acoustic absorption data on the elastic protein-based polymers in direct comparison with random chain elastomers in combination with thermodynamic and statistical mechanical analyses [8,9] to be convincing to many parties. In addition, one hopes recent analyses of isometric contractions [1], reviewed above, will put to rest the proposition that the solvent entropy change makes a direct contribution to entropic elastic force.
B. The Comprehensive Hydrophobic Effect: Process of Hydrophobic Association 1. Thermodynamics of Solubility of Hydrophobic Groups in Water The remarkable feature of the solubility of hydrophobic groups is that their dissolution in water is a favorable exothermic process [17]. In particular, Butler measured the heat of dissolution in
38
Urry et al.
Figure 3 Molecular structure of poly(GVGVP), also referred to as poly(VPGVG). (A) Schematic representation as a series of -turns inserted by the Pro-Gly sequence with interconnecting Val-Gly-Val sequences. (B) Detailed structure of the -turn as determined from the crystal structure of the cyclopentadecapeptide cyclo(GVGVP)3, showing the single secondary structural feature of a ten-atom hydrogen-bonded ring involving the Val1 ⳮCO ●●● HN-Val4 residues. The residues are so numbered as to enclose and identify the -turn structure. (C) Schematic of helical structure that forms on raising the temperature with optimization of intramolecular hydrophobic contacts. (D) Schematic of helical structure showing the series of -turns to function as hydrophobic spacers between turns of what is called a -spiral. (E) Stereo views (side and axial) of detailed computed structure of the -spiral derived using NMR structural constraints of coupling constants and nuclear Overhauser effects to limit allowed torsion angle ranges and of the Val1 ⳮCO ●●● HN-Val4 hydrogen bond and molecular mechanics calculations. (F) Association of -spirals to form twisted filaments as observed in the transmission electron micrographs of negatively stained incipient aggregates undergoing a phase separation.
water of the methanol to n-pentanol series of linear alcohols, and determined the average heat released on addition of each CH2 to be a favorable ⌬H ⳱ ⳮ1.4 kcal/mol-CH2, whereas the average change in free energy due to formation of structured hydrophobic hydration from bulk water for the series was (ⳮT⌬S) ⳱ Ⳮ 1.7 kcal/mol-CH2 [17]. As the change in Gibbs free energy for solubility, ⌬G(solubility) ⳱ ⌬H ⳮ T⌬S, the ⌬G(solubility) of methanol is a large negative number that shifts positively by 0.3 kcal/mol for each of the four added CH2 groups as the series progresses from methanol to n-pentanol. Even though the heat released on dissolution increases with each CH2 group added, solubility is entirely lost once seven CH2 groups have been added to methanol, that is, n-octanol is insoluble at 25⬚C. Obviously, the solubility of n-hexanol, for example, would decrease with increases in temperature, because the (ⳮT⌬S) term would become more positive by the magnitude of the increase in temperature. This is the thermodynamic basis for the inverse temperature transition of hydrophobic association on raising the temperature.
Elastic Protein-Based Biomaterials
39
2. Phase Diagrams for Inverse Temperature Transitions of Elastic Protein-Based Polymers The traditional way to characterize phase transitions is by phase diagrams that plot temperature of the transition as a function of polymer concentration in a solvent. Figure 4 gives the phase diagrams in water for two elastic protein-based polymer compositions, poly(GVGVP) and poly(GGVP), as well as partial plots for additional compositions, e.g., poly(GVGIP), poly[0.8(GVGVP),0.2(GEGVP)]. Interestingly, the curvatures of the plots are inverted from those of conventional petroleum-based polymers dissolved in organic solvents. The solubilities are also inverted, that is, the elastic protein-based polymers are soluble below the binodal or coexistence line and insoluble above, rather than the usual increase in solubility with increase in temperature. Characterization of the inverse temperature transition of hydrophobic association involves determination of the onset temperature for the phase separation as the temperature is raised (see Fig. 5). The onset temperature for the inverse temperature transition has been designated as Tt, which is equivalent to temperature of the binodal or coexistence line. For discussion of inverse temperature transitions, therefore, we refer to the binodal or coexistence line as the Tt-divide, because it is the line that divides the lower temperature soluble state from the higher temperature insoluble state. 3. The Change in Free Energy of Hydrophobic Association, ⌬GHA The change in Gibbs free energy of hydrophobic association resulting from any variable that changes the value of Tt has been derived as [1] (1) ∆G HA(χ) = [Tt(χ) - Tt (reference)] ∆S t (reference )
Figure 4 Phase diagrams of poly(GVGVP) and poly(GGVP) showing an inverted curvature with respect to the x-axis and an inverted solubility with the soluble state below the binodal or coexistence line (called the Tt-divide). The inverted curvature and inverted solubility, when compared to the more common phase diagrams of petroleum-based polymers, is part of the reason why the phase transition for these elastic protein–based polymers is called an inverse temperature transition. (Adapted from Ref. 19.)
40
Urry et al.
⌬St(reference) is the entropy change for the transition before introduction of the variable, , where ⌬St(reference) is obtained from the differential scanning calorimetry data by dividing the increment of heat absorbed over an incremental change in temperature and summed over all of the increments of the transition. Thus, the change in the onset temperature for the transition due to a perturbation ⌬Tt() times a constant for the reference polymer provides an approximate measure of the change in Gibbs free energy of hydrophobic association for these elastic proteinbased polymers. Since the energy of deformation on stretching the chain is less than 20% of the heat of the hydrophobic hydration that attends the extension [3], neglecting ⌬Ht(chain) gives one aspect of the approximation. The derivation also takes the chemical potential of the hydrophobically associated state at the Tt-divide to be equal to the chemical potential of the disassociated state. This approximation becomes less satisfactory as the width of the transition increases. 4. The Competition for Hydration Between Hydrophobic and Charged Moieties From Fig. 4 the presence of four carboxylates, COOⳮ, of the glutamate (Glu, E) residue per 100 residues in poly[0.8(GVGVP),0.2(GEGVP)], increases the value of the Tt-divide by 45⬚C in the presence of phosphate-buffered saline (PBS), 0.15 N NaCl, and 0.01 M phosphate. In the absence of PBS, the Tt-divide for poly(GVGVP) increases by no more than 2 to 3⬚C, whereas the Tt-divide for poly[0.8(GVGVP),0.2(GEGVP)] extrapolates toward 200⬚C. The ion pairing COOⳮ NaⳭ dramatically lowers the Tt-divide to 70⬚C. From the above discussion of ⌬G(solubility), the fact that the carboxylate in the absence of cations raised the Tt-divide so high suggested that the carboxylate did so by destructuring hydrophobic hydration in the process of obtaining its own hydration. Indeed using microwave dielectric relaxation methods, the loss of hydrophobic hydration on formation of less than two carboxylates per 100 residues has been directly observed and quantified [11,18]. Thus, ion paring allows reconstitution of hydrophobic hydration with the lowering of the Tt-divide. This lowering of the Tt-divide on ion pairing is used to follow the loading profile of positively charged polymers with negatively charged drugs, as seen below in Figs. 6A and B. 5. Apolar–Polar Repulsive Free Energy of Hydration, ⌬Gap The other side of the coin of the competition for hydration between hydrophobic and charged moieties is the increase in free energy of the charged state as seen in hydrophobic-induced pKa shifts. Prior to the work on these protein-based polymers it was thought that any significant pKa shift arose either due to charge–charge repulsion or due to a conformation that forced the ionizable function into a medium of low dielectric constant. We reject the latter explanation and show that charge–charge repulsion can be a few tenths of a pH unit with one charged residue every fifth residue in the absence of PBS, but even that is relaxed in the presence of 0.15 N NaCl [35,36,37]. On the other hand hydrophobic-induced pKa shifts as much as 6 pH units have been observed [38]. Thus, we have that the change in Gibbs free energy due to the apolar–polar repulsive free energy of hydration, ⌬Gap, can be written as [11] ∆Gap = 2.3 RT ∆ pKa
(2)
When charge–charge repulsion is negligible and the competition for hydration dominates, ⌬Gap 艐 ⌬GHA. This provides an understanding of the energetics involved in the loading and release of drugs by appropriately designed elastic protein-based polymers.
Elastic Protein-Based Biomaterials
41
Figure 5 Characterization of the phase transition of (GVGVP)251 on raising the temperature. (A) Tubes containing (GVGVP)251 and water showing a clear solution turn cloudy due to aggregation on raising the temperature above 25⬚C. On standing the aggregates settle out and form a phase-separated state. (B) Following the temperature elicited aggregation by light scattering with the development of tubidity being used to define the value of Tt. (C) Differential scanning calorimeter curve of (GVGVP)251 showing the relationship between Tt and the onset and breadth of the phase transition. Note that the curve has been inverted from the plots of Fig. 4. (Adapted from Ref. 11.)
42
Urry et al.
III. ELEMENTS OF THE ION-PAIRED CONTROLLED RELEASE DEVICE A. Design of Elastic Protein-Based Polymers for Loading and Release of Anionic Drugs When considering the design of elastic protein-based polymers for controlled release of anionic drugs with the above-presented background information, a series of lysine-containing polymers with a range of hydrophobic-induced pKa shifts become an obvious choice. The following family of polymers were designed, prepared by means of recombinant DNA technology, and purified by means of the inverse temperature transition [39]: Polymer Polymer Polymer Polymer Polymer
i′ : (GVGVP GVGVP GKGVP GVGVP GVGVP GVGVP)22(GVGVP) K/0F ii′ : (GVGVP GVGFP GKGFP GVGVP GVGVP GVGVP)22(GVGVP) K/2F iii′ ’: (GVGVP GVGVP GKGVP GVGVP GVGFP GFGFP)22(GVGVP) K/3F iv′ : (GVGVP GVGFP GKGFP GVGVP GVGFP GVGFP)21(GVGVP) K/4F v′ : (GVGVP GVGFP GKGFP GVGVP GVGFP GFGFP)21(GVGVP) K/5F
The pKa values for the above polymers i′ through v′ are 10.0 (K/0F), 9.8 (K/2F), 9.5 (K/3F), 9.0 (K/4F), and 8.5 (K/5F) (unpublished results). These elastic protein-based polymers were designed for systematic nonlinear hydrophobic-induced pKa shifts. The expectation is that affinity for anionic drug will parallel the pKa shifts and that release rates will be inversely related to the to the pKa shifts. Larger hydrophobic-induced pKa shifts equate to lower release rates. B. Profiles for Loading Anionic Drugs into Cationic Elastic Protein-Based Polymer Release Devices 1. Determination of Tt Values for Elastic Protein-Based Polymers As shown in Fig. 5A, the elastic protein-based polymer (GVGVP)251 is completely soluble in water at temperatures below 25⬚C. On raising the temperature above 25⬚C, the solution becomes cloudy, and on standing phase separation occurs. Our standard determination of the Tt value utilizes a solution of 40 mg/mL polymer with a molecular weight of approximately 100 kDa. At a sufficiently low temperature the polymer is completely dissolved as a clear solution exhibiting no scattering of light. The temperature is raised at a rate of 30⬚C per hour while monitoring light scattering at 300 to 400 nm. There occurs an abrupt development of light scattering that soon reaches a maximum. The temperature at 50% turbidity is defined as the Tt for the sample, as indicated in Fig. 5B. In Fig. 5C the differential scanning calorimetry curve for an equivalent sample is reported with the onset of the endothermic reaction for the destructuring of hydrophobic hydration corresponding to the value of Tt. The Tt values for the five polymers, polymers i′ through v′ , listed above are plotted at 0 on the x axes of Fig. 6A and B for a 0.01 M phosphate solution at pH 7.5. 2. Determination of the Profile for Loading Drug into a Phase-Separated State The profile for loading drug into the release device again uses a 40-mg/mL solution of polymer at pH 7.5, which is approximately 16 mM in lysine ε-amino (—NH3Ⳮ) side chains. When a less than equimolar quantity of dexamethasone phosphate (DMP) or betamethasone phosphate (BMP) is dissolved with the particular polymer at a sufficiently low temperature for complete dissolution and the temperature is increased, the value of Tt is plotted as a function of the ratio of [DMP]/2[Lys residues], as the phosphate of the DMP is expected to bridge between two lysine ε-amino (—NH3Ⳮ) groups. The drug loading profiles are given in Fig. 6A and B. The set of loading profiles are essentially identical for the two drugs, as might be expected since they differ only in the orientation of the methyl at position 16 in the D ring of the steroid nucleus, being 16 ␣-methyl for dexamethasone and 16 -methyl for betamethasone.
Elastic Protein-Based Biomaterials
43
Figure 6 Loading profiles for dexamethasone phosphate (A) and betamethasone phosphate (B) followed by the decrease in the value of Tt on formation of an ion-pair between anionic drug and cationic polymers with one lysine (Lys, K) residue per thirty residues but with 0, 2, 3, 4, and 5 Phe (F) residues having replaced Val (V) residues in each 30 mer. The affinity for ion-pairing formation increases in parallel with the hydrophobic-induced pKa shifts. The two steroids differ only by the orientation of a methyl substituent at position 16 in the D-ring. Because the basis for the affinity is the relaxation of the apolar-polar repulsive free energy of hydration on ion-pairing and does not involve the selective binding of the steroid, the loading profiles are essentially indistinguishable.
At a temperature above Tt the drug is in the phase-separated state of the polymer. The drugcontaining phase-separated state becomes the drug delivery device. In fact, the drug provides the chemical energy to drive phase separation for its own packaging. In particular, if the polymer were cross-linked into a swollen elastic matrix suspending a weight, the addition of drug would drive contraction with performance of the mechanical work of lifting the weight. C. Release Profiles of Anionic Drug Release from Cationic Protein-Based Polymeric Devices The drug release profiles for dexamethasone phosphate (DMP) and betamethasone phosphate are given in Figs. 7A and B, respectively, for the polymers indicated above as polymers i′ through v′ . Also the release data are tabulated in Tables 1 and 2.
44
Urry et al.
Figure 7 Release profiles for dexamethasone phosphate (A) and betamethasone phosphate (B) from a constant surface area of phase separated state formed due to ion-pairing of anionic drug with cationic polymer. Again the profiles of the two steroids are essentially identical. Interestingly, the release profile for betamethasone phosphate phase-separated on ion-pairing with (GVGVP GVGFP GKGFP GVGVP GVGFP GVGFP)21(GVGVP) appears to exhibit zero order release kinetics for a period of over 100 days.
Table 1 Dexamethasone Release Polymer K/2F K/3F K/4F K/5F
mg loaded
mg released
Days of release
Days remaining
Last release,
35.3 34.5 18.2 13.7
31.4 17.4 13.7 3.4
216 169 235 94
111 444 277 953
.035 .039 .016 .011
Elastic Protein-Based Biomaterials
45
Table 2 Betamethasone Release Polymer K/2F K/3F K/4F
mg loaded
mg released
Days of release
Days remaining
Last release,
41.4 33.5 18.2
37.6 28.8 10.5
236 236 236
70 83 314
.0487 .0566 .0245
Clearly the principle is established for lysine-containing elastic protein-based polymers that the extent of hydrophobic-induced pKa shifts can be used to control the relative rates of release. Because of the phosphate in the medium and its capacity to exchange for the drug used in this case, however, not all compositions result in a zero order release profile. The K/4F composition, polymer iv′ , does appear to give a near zero order release profile for both drugs.
D. Release Profiles of Cationic Drug Release from Anionic Protein-Based Polymeric Devices Zero order release profiles are possible when using a cationic drug with an anionic, carboxylatecontaining, protein-based polymer. Of course, the level of release can be controlled by design of the desired hydrophobic-induced pKa shift. This is demonstrated in Fig. 8 for Leu-enkephalin amide and a series of glutamate-containing elastic protein-based polymers [16]. In this case we have loaded with a 50% excess of drug over ion-paired binding sites, such that there is a burst release. This would not occur if the device began with an exact equivalency of drug to binding site. Also, as indicated in Fig. 8B, the drug delivery device disperses with the drug. The length of time that the constant, zero order release profiles last depends on the size of the depot, and the release remains zero order for a constant surface area. The release rates for a constant surface area of 0.55 cm2 are seen to vary from 4.7 to 1.85 M per day, with the constant release in the latter case demonstrated for three months.
IV. NANOPARTICLE OF ELASTIC PROTEIN-BASED POLYMERS A. Formation of Nanoparticles from Elastic Protein-Based Polymers Nanoparticles of the protein-based polymers, (GVGVP)251 and [(GVGVP GVGFP GEGFP GVGVP GVGFP GFGFP)32 (GVGVP)], have been successfully prepared. Polymer solutions of (GVGVP)251 (25 mg/mL) and of [(GVGVP GVGFP GEGFP GVGVP GVGFP GFGFP)32(GVGVP)] (7.2 mg/mL) were cross-linked at above 40 and 50 ⬚C, respectively, by cobalt-60 ␥-radiation with an exposure rate of 1.5 Mrad per hour. The transmission electron micrographs of Figs. 9A and B show the size and relative homogeneity of the nanoparticles obtained from (GVGVP)251 and from [(GVGVP GVGFP GEGFP GVGVP GVGFP GFGFP)32 (GVGVP)]. Particle sizes with diameters in the range of 50 to 200 nm were achieved for (GVGVP)251, whereas a narrower range of particle sizes with diameters of 50 to 100 nm were obtained for [(GVGVP GVGFP GEGFP GVGVP GVGFP GFGFP)32 (GVGVP)]. It may be noted that the distribution of nanoparticles of (GVGVP GVGFP GEGFP GVGVP GVGFP GFGFP)n containing Glu and Phe residues, being more hydrophobic and controllable by pH, are much more homogeneous than those of (GVGVP)251.
46
Urry et al.
Figure 8 Release profiles for cationic Leu-enkephalin amide ion-paired with anionic, carboxylate-containing protein-based polymers. For this configuration of cationic drug and anionic polymer zero order release profiles are obtained once the excess drug is released. The release levels are inversely proportional to the hydrophobic-induced pKa shifts; they differ by a factor of 2.5, and remain at constant release levels for one to three months. (From Ref. 16.)
Elastic Protein-Based Biomaterials
47
Figure 9 Nanoparticles of elastic protein-based polymers fixed by cross-linking with 20 Mrads of ␥irradiation. (A) Composed of (GVGVP)251. (B) Composed of [(GVGVP GVGFP GEGFP GVGVP GVGFP GFGFP)32(GVGVP)].
48
Urry et al.
B. Advantage of Nanoparticles Made from Elastic Protein-Based Polymers Nanoparticles made from elastic protein-based polymers have the usual advantages of being able to escape from the circulation into diseased tissues and be taken up by the liver; aspirated into the lungs; and injected, as desired, intramuscularly, subcutaneously, intrathecally, etc. Being transductional, however, nanoparticles made from elastic protein-based polymers exhibit a number of special advantages. They can be designed for a range of different release rates, as seen in Figs. 7 and 8, and as they are depleted of ion-paired drug they swell and become proteolytically degradable. In addition, it becomes a simple matter to introduce into the protein-based polymer desirable biologically active peptide sequences such as cell attachment sites, proteolytic cleavage sites, selective kinase sites, etc. Another specific advantage of elastic protein-based polymers is the basic biocompatibility of these dynamic polymeric structures from which they derive their entropic elastic property. V. BASIS OF BIOCOMPATIBILITY OF ELASTIC PROTEIN-BASED POLYMERS The preceding reviews our understanding of the mechanism of elasticity and the comprehensive hydrophobic effect, which enable the use of elastic protein-based polymers as biomaterials. Also, the particular way in which they achieve their entropic elasticity provides insight into their remarkable biocompatibility. Information establishing the outstanding biocompatibility is briefly reviewed followed by the proposed entropic explanation of the unique biocompatibility. A. Studies Establishing Biocompatibility of the Basic Elastic Sequences 1. Standard Battery of Eleven Biocompatibility Tests on (GVGVP) and (GGVP) The full battery of 11 biocompatibility tests performed at North American Science Associates (NAmSA) have been obtained on the basic elastic protein-based polymer sequences poly(GVGVP) and poly(GGAP) [15,40] and their ␥-irradiation cross-linked matrices. A listing of the tests, their description, the test system, and the results is given in Table 3 for poly(GVGVP) and its crosslinked matrix. The results are the same for poly(GGAP). 2. In Vitro Assay for the Response of Human Macrophage (Monocyte) Cell Lines to Biomaterials Macrophages fulfill the role of identifying and destroying foreign materials in the body. On identification of a foreign object they give off bursts of superoxide and hydrogen peroxide that oxidize and thereby begin the degradation and disposal of materials recognized as foreign. Grace Picciolo and coworkers [41] at the Center for Devices and Radiological Health of the Food and Drug Administration (FDA) developed an instrument to evaluate the reaction of human monocytes to biomaterials. The usual biomaterials elicit substantial oxidative bursts. They determined, however, that poly(GVGVP) and the related polymer-containing cell attachment sequences, poly[40(GVGVP),(GRGDSP)], slightly reduce the background oxidative activity of the monocytes. Simply stated these elastic protein-based polymers are not seen as foreign by human macrophage cell lines. 3. Efforts to Obtain Monoclonal Antibodies Perhaps most remarkably, using sensitive and sensitized Balb/c mice, a primary immunization using 200 g of poly(GVGVP) in emulsion with complete Freund’s adjuvant, followed by a
Elastic Protein-Based Biomaterials
49
Table 3 Summary of Biological Test Results for Poly(GVGVP) and Its ␥-irradiation Cross-Linked Matrix Test (1) Ames (Mutagenicity) (2) Cytoxicity (3) Systemic Toxicity (4) Intracutaneous Toxicity (5) Muscle Implantation (6) I.P. Implantation (7) Systemic Antigenicity (BPAT) (8) Sensitization (Kligman Test) (9) Pyrogenicity (10) Clotting Study (11) Hemolysis
Description
Test System
Results
Determine reversion rate to wild type of histidine-dependent mutants Agarose overlay determine cell death and zone of lysis Evaluate acute systemic toxicity from an I.V. or I.P. injection Evaluate local dermal irritant or toxic effects by injection Effect on living muscle tissue Evaluate potential systemic toxicity Evaluate general toxicology
Salmonella typhimurium
non-mutagenic
L-929 mouse fibroblast Mice
non-toxic
Rabbit
non-toxic
Rabbit Rat
favorable favorable
Guinea Pigs
non-antigenic
Dermal sensitization potential
Guinea Pigs
non-sensitizing
Determine febrile reaction Whole blood clotting times Level of hemolysis in the blood
Rabbit Dog Rabbit blood
non-pyrogenic normal clotting time non-hemolytic
non-toxic
* Reports from North American Science Associates (NAmSA)
secondary immunization using 200 g of poly(GVGVP) in emulsified incomplete Freund’s adjuvant, and finally boosted with a third immunization repeating the conditions of the second failed to produce antigen-specific hybridomas to poly(GVGVP). The use of two additional strains of mice (DBA and CB6 F1) similarly failed. As stated in the report of the results [16], ‘‘The preceding results suggest that certain protein-based (bioelastic) polymers, such as the elastic poly(GVGVP), appear to be incapable of eliciting a significant immune response.’’ 4. Guinea Pig Subcutaneous Injection Model for Evaluation of Inflammatory and Pyrogenic Responses A substantial body of work established chemically synthesized elastic protein-based polymers to have remarkable biocompatibility. On developing recombinant DNA technology to produce the elastic protein-based polymers in E. coli [39], a strong inflammatory response was obtained in the guinea pig subcutaneous injection model on using several cycles of phase separation for purification. This occurred even though by the Western immunoblot technique impurities were less than 1 ppm [42]. This is in spite of the fact that a level of less than 10 ppm impurities was the standard for use of insulin produced by E. coli. The difference is that a drug such as insulin is commonly used in microgram quantities or less. On the other hand the use of an elastic protein-based polymer as a biomaterial involves quantities of one thousand or even one million times greater. In particular in the guinea pig subcutaneous injection model 30 mg of (GVGVP)251 was injected. The gold standard for biocompatibility would then become the subcutaneous injection in the guinea pig of 30 mg of an elastic protein-based polymer that would release all of its impurities within a few days time without eliciting a significant inflammatory response. This was achieved
50
Urry et al.
with the microbially produced elastic protein-based polymer (GVGVP GVGVP GEGVP GVGVP GVGVP GVGVP)32(GVGVP). As seen in Fig. 10, at 2 weeks after subcutaneous injection of 30 mg of this polymer there was no evidence of the polymer having been present. Out of five test sites three showed traces of no more than a half dozen inflammatory cells [42]. Accordingly, the desired level of purification was achieved, and most remarkably in some sites despite a thorough search no tissue reaction whatever was discernible. 5. Glu- and Phe-Containing Polymers Also Biocompatible in Guinea Pig Subcutaneous Injection Model It had been our expectation, once the elastic protein-based polymer contained a polar group such as a carboxylate in combination with more hydrophobic groups as in (GVGVP GVGFP GEGFP GVGVP GVGFP GFGFP)n(GVGVP) and (GVGVP GVGVP GEGVP GVGVP GVGFP GFGFP)n(GVGVP) that significant epitopes would be present and that a much more significant inflammatory response would be found. From the anecdotal evidence to date using the guinea pig subcutaneous injection model, the expected increase has not been seen. The question becomes, therefore, whether there might be a fundamental reason why these elastic protein-based polymers and the mammalian elastic fiber itself elicit such meager immuno-
Figure 10 One of four subcutaneous sites in the guinea pig for injection of 30 mg of 兵GVGVP GVGVP GEGVP GVGVP GVGVP GVGVP其36(GVGVP). At two weeks no trace of the polymer was observed at any of the four sites. On thoroughly searching through all of the slides of the four sites, two of the four sites (one of which is represented here) exhibited no evidence of any kind that this large injection of the elastic protein-based material had ever occured. In two sites there was only a single trace at one spot one spot in each of a half dozen to a dozen inflammatory cells. This provides the ultimate test for biocompatibility and for purification of E. coli-produced elastic protein-based polymer. (From Ref. 42.)
Elastic Protein-Based Biomaterials
51
genic responses. An answer emerges below when considering further the biological implications of the mechanical resonances indicated by Fig. 1.
B. Proposed Basis for the Remarkable Biocompatibility of Elastic Protein-Based Polymers 1. The Presence of Mechanical Resonances. As seen in Figs. 1A and B, on raising the temperature from below to above the inverse temperature transition observed in the dielectric relaxation experiment, noticeable are the development of strong relaxations localized near 3 kHz and again near 5 MHz. Therefore, during the hydrophobic association transition the pentamers fold into a regularly repeating dynamic structure. If there were no regular structure, the responsible rotation about bonds would occur throughout the frequency spectrum. On the contrary for this family of elastic protein-based polymers, more than 90% of the relaxation can be fit by a single relaxation frequency [29]. This means that the relaxations may be considered as describing mechanical resonance where the entire pentamer moves at the frequency of the relaxation. As argued below, such low frequency motions make large contributions to lowering the free energy of the structure, and an interaction that would interfere with or stop such motion would occur only on overcoming the stabilization that the motion represents. 2. The Dynamic Entropic Shield Against Identification as Foreign Insight into the energetics that the coordinated oscillations about backbone bonds indicated by the mechanical resonances at 3 kHz and 5 MHz comes from the harmonic oscillator model approximation for the relevant partition function. At the top of Fig. 10A is the expression for the entropy of a harmonic oscillator as a function of the frequency of the oscillation. Using this expression, a plot of the entropy as a function of log(frequency) is given in Fig. 11A. The salient feature is that as the frequency decreases, the entropy increases. On the right-hand ordinate is plotted the TSi term. For a frequency of 5 MHz the contribution, based on this idealized model, to lowering the free energy would be about 9 kcal/mol-pentamer, whereas the mechanical resonance near 3 kHz would calculate to lower the free energy by some 14 kcal/mol-pentamer. One should not take the harmonic oscillator approximation as quantitative in any way. Nonetheless, it is informative and allows the perspective that presence of such resonances very significantly lowers the free energy of the hydrophobically associated phase-separated state. The next point to consider is that the barrier to such motions is very low, as demonstrated in Fig. 11B. This means that an oscillating electric field is not required to set the resonances in motion, nor for that matter is an acoustic wave required to producem motion, as the nominal 3 kHz resonance is in the acoustic frequency range. It would seem that the low frequency motions that lower the free energy of the phase-separated, -spiral–containing state would be present under physiological conditions. Now in order to identify elastic protein-based polymers exhibiting these low frequency motions would require that the elements that constitute an epitope would have to be stopped. Just as the motions lower the free energy of the -spiral–containing state, stopping the motions to identify an epitope would require raising the free energy of the elastic protein-based polymer. The need to raise the free energy in order to identify an elastic protein-based polymer as foreign constitutes an entropic shield that would give the impression that the elastic protein-based material was being ignored by the host.
52
Urry et al.
Figure 11 (A) Plot of the contribution to the entropy of a harmonic oscillator as a function of frequency from 1013 to 103 Hz(cycles/sec). The relaxation frequencies of the two mechanical resonances of Fig. 1 were approximately 3 ⳯ 103 Hz (3 kHz) and 5 ⳯ 106 Hz (5 MHz). By this approximation these mechanical resonances would contribute 9 and 14 kcal/mol-pentamer to stabilization of the regular, but dynamic, structure of the elastic protein-based polymer. Of course, plotting the entropy calculated using the harmonic oscillator partition function to such low frequencies cannot be considered numerically accurate, but it does show the trend and the trend is to increasing stabilization with decreasing frequency. (B) Plot of the temperature dependence of the 5 MHz mechanical resonance to determine the barrier to mobility. The barrier is 1.2 to 1.3 kcal/mol-pentamer, which means that the motion occurs without excitation by either an oscillating electric field as in Fig. 1, or an acoustic wave as would be relevant to the 3 kHz relaxation. See text for the implication with respect to biocompatibility. (From Ref. 33.)
To the extent that this is a significant factor in the remarkable biocompatibility, it would be unwise and self-defeating to compound this family of elastic protein-based polymers to other protein-based materials without such beneficial properties.
ACKNOWLEDGMENTS The authors wish to acknowledge the support of the office of Naval Reearch, under contracts N0001–00-C-0404 and N00014–00-C-0178, and the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Under SBIR Grant 5R43AI49005–01.
Elastic Protein-Based Biomaterials
53
REFERENCES 1. Urry DW, Parker TM. Mechanics of elastin: molecular mechanism of biological elasticity and its relevance to contraction. Special Issue: Mechanics of Elastic Biomolecules. J. Muscle Res. Cell Motility, 2002. 2. Hoeve CAJ, Flory PJ. The elastic properties of elastin. J. Am. Chem. Soc 1958; 80:6523–6526. 3. Hoeve CAJ, Flory PJ. Elastic properties of elastin. Biopolymers 1974; 13:677–686. 4. Weis-Fogh T, Andersen SO. New molecular model for the long-range elasticity of elastin. Nature 1970; 227:718–721. 5. Alonso LB, Bennion BJ, Daggett V. Hydrophobic hydration is an important source of elasticity in elastin-based polymers. J. Am. Chem. Soc 2001; 123:11991–11998. 6. Urry DW, Venkatachalam C, Long MM, Prasad KU. Dynamic -spirals and a librational entropy mechanism of elasticity. Conformation in Biology: The Festschrift celebrating the Sixtieth Birthday of G.N. Ramachandran, F.R.S. Srinivasan R, Sarma RH, eds. New York: Adenine Press, 1982:11–27. 7. Chang DK, Urry DW. Polypentapeptide of elastin: damping of internal chain dynamics on extension. J. Computational Chem 1989; 10:85–855. 8. Urry DW, Hugel T, Seitz M, Gaub H, Sheiba L, Dea J, Xu J, Parker T. Elastin: a representative ideal protein elastomer. Phil. Trans. R. Soc. Lond. B. Vol. 357, 2002:169–184. 9. Urry DW, Hugel T, Seitz M, Gaub H, Sheiba L, Dea J, Xu J, Hayes L, Prochazka F, Parker T. Ideal Protein Elasticity: The Elastin Model. Elastomeric Proteins. Shewry P, Bailey A, eds: Cambridge University Press, 2002: In press. 10. Flory PJ. Molecular interpretation of rubber elasticity. Rubber Chemistry and Technology 1968; 41: G41–G48. 11. Urry DW. Physical chemistry of biological free energy transduction as demonstrated by elastic protein-based polymers. J. Phys. Chem. B 1997; 101:11007–11028. 12. Urry DW. Molecular Machines: How motion and other functions of living organisms can result from reversible chemical changes. Angew. Chem. (German). Angew. Chem. Int. Ed. Engl 1993; 32: 819–841. 13. Sandberg LB, Soskel NT, Leslie JG. Elastin structure, biosynthesis, and relation to disease states. N. Engl. J. Med 1981; 304:566–79. 14. Sandberg LB, Leslie JG, Leach CT, Alvarez VL, Torres AR, Smith DW. Elastin covalent structure as determined by solid phase amino acid sequencing. Pathol. Biol. (Paris) 1985; 33:266–274. 15. Urry DW, Parker TM, Reid MC, Gowda DC. Biocompatibility of the bioelastic materials, poly(GVGVP) and its ␥-irradiation cross-linked matrix: summary of generic biological test results. J. Bioactive Compatible Polym 1991; 6:263–282. 16. Urry DW, Pattanaik A, Accavitti MA, Luan C-X, McPherson DT, Xu J, Gowda DC, Parker TM, Harris CM, Jing N. Transductional elastic and plastic protein-based polymers as potential medical devices. Handbook of Biodegradable Polymers Domb AJ, Kost J, Wiseman DM, eds;, Harwood Academic Publishers 1997:367–386. 17. Butler JAV. The energy and entropy of hydration of organic compounds. Trans. Faraday Soc 1937; 33:229–238. 18. Urry DW, Peng S-Q, Xu J, McPherson DT. Characterization of waters of hdrophobic hydration by microwave dielectric relaxation. J. Am. Chem. Soc 1997; 119:1161–1162. 19. Sciortino F, Palma MU, Urry DW, Prasad KU. Nucleation and accretion of bioelastomeric fibers at biological temperatures and low concentrations. Biochem. Biophys. Res. Commun 1988; 157: 1061–1066. 20. Wasserman ZR, Salemme FRA. molecular dynamics investigation of the elastomeric restoring force in elastin. Biopolymers 1990; 29:1613–1631. 21. Luan C-H, Jaggard J, Harris RD, Urry DW. On the source of entropic elastomeric force in polypeptides and proteins: backbone configurational vs. side chain solvational entropy. Int. J. Quant. Chem.: Quant. Biol. Symp 1989; 16:235–244. 22. Urry DW, Haynes B, Zhang H, Harris RD, Prasad KU. mechanochemical coupling in synthetic polypeptides by modulation of an inverse temperature transition. Proc. Natl. Acad. Sci. USA 1988; 85:3407–3411.
54
Urry et al.
23. Urry DW. Characterization of soluble peptides of elastin by physical techniques. Methods in Enzymology Cunningham LW, Frederiksen DW, eds. Vol. 82. New York: Academic Press, 1982:673–716. 24. Urry DW. Thermally driven self-assembly, molecular structuring and entropic mechanisms in elastomeric polypeptides. Molecular Conformation and Biological Interactions Balaram P, Ramaseshan S, eds;. Bangalore. India: Indian Academy of Sciences, 1991:555–583. 25. Urry DW, Trapane TL, Long MM, Prasad KU. Test of the librational entropy mechanism of elasticity of the polypentapeptide of elastin: effect of introducing a methyl group at residue-5. J. Chem. Soc., Faraday Trans. I 1983; 79:853–868. 26. Urry DW, Trapane TL, Wood SA, Walker JT, Harris RD, Prasad KU. D-Ala5 analog of the elastin polypentapeptide. Physical characterization. Int. J. Pept. Protein Res 1983; 22:164–175. 27. Urry DW, Trapane TL, Wood SA, Harris RD, Walker JT, Prasad KU. D-Ala3 analog of elastin polypentapeptide: an elastomer with an increased Young’s modulus. Int. J. Pept. Protein Res 1984; 23:425–434. 28. Urry DW, Jaggard J, Prasad KU, Parker T, Harris RD. Poly(Val1-Pro2-Ala3-Val4-Gly5): A Reversible, Inverse Thermoplastic. Biotechnology and Polymers Gebelein CG, ed;, Plenum Press. 1991:265–274. 29. Henze R, Urry DW. Dielectric relaxation studies demonstrate a peptide librational mode in the polypentapeptide of elastin. J. Am. Chem. Soc 1985; 107:2991–2993. 30. Urry DW, Henze R, Redington P, Long MM, Prasad KU. Temperature dependence of dielectric relaxations in ␣-elastin coacervate: evidence for a peptide librational mode. Biochem. Biophys. Res. Commun 1985; 128:1000–1006. 31. Urry DW, Trapane TL, Iqbal M, Venkatachalam CM, Prasad KU. Carbon-13 NMR relaxation studies demonstrate an inverse temperature transition in the elastin polypentapeptide. Biochemistry 1985; 24:5182–5189. 32. Urry DW, Trapane TL, McMichens RB, Iqbal M, Harris RD, Prasad KU. Nitrogen-15 NMR relaxation study of inverse temperature transitions in elastin polypentapeptide and its cross-linked elastomer. Biopolymers 1986; 25:S209–S228. 33. Buchet R, Luan C-H, Prasad KU, Harris RD, Urry DW. Dielectric relaxation studies on analogs of the polypentapeptide of elastin. J. Phys. Chem 1988; 92:511–517. 34. Venkatachalam CM, Urry DW. Calculation of dipole moment changes due to peptide librations in the dynamic -spiral of the polypentapeptide of elastin. Int. J. Quant. Chem.: Quant. Biol. Symp 1986; 12:15–24. 35. Urry DW, Peng S-Q, Parker TM. Delineation of electrostatic-and hydrophobic-induced pKa shifts in polypentapeptides: the glutamic acid residue. J. Am. Chem. Soc 1993; 115:7509–7510. 36. Urry DW, Peng S-Q, Parker TM, Gowda DC, Harris RD. Relative significance of electrostatic- and hydrophobic-induced pKa shifts in a model protein: the aspartic acid residue. Angew. Chem. (German). Angew. Chem. Int. Ed. Engl 1993; 32:1440–1442. 37. Urry DW, Peng S-Q, Gowda DC, Parker TM, Harris RD. Comparison of electrostatic- and hydrophobic-induced pKa shifts in polypentapeptides: the lysine residue. Chem Phys Lett 1994; 225:97–103. 38. Urry DW, Gowda DC, Peng S-Q, Parker TM. Non-linear hydrophobic-induced pKa shifts: implications for efficiency of conversion to chemical energy. Chem. Phys. Lett 1995; 23(9):67–74. 39. McPherson DT, Xu J, Urry DW. Product purification by reversible phase transition following E. coli expression of genes encoding up to 251 repeats of the elastomeric pentapeptide GVGVP. Protein Expression and Purification 1996; 7:51–57. 40. Urry DW, Nicol A, McPherson DT, Xu J, Shewry PR, Harris CM, Parker TM, Gowda DC. Properties, preparations, and applications of bioelastic materials. In Encyclopedic Handbook of Biomaterials and Bioengineering. Part A: Materials, Marcel Dekker. 1995; 2:1619–1673. 41. Picciolo GL, Kaplan DS, Batchelder KF, Kapur R, Kotz RM. Biotechnology-derived biomaterials modulate host cell reactive oxygen production as measured by chemiluminescence, 19th Annual Meeting, Society for Biomaterials. 1993. 42. Urry DW, Pattanaik A, Xu J, Woods TC, McPherson DT, Parker TM. Elastic protein-based polymers in soft tissue augmentation and generation. J. Biomater. Sci. Polym. Edn. 1998; 9:1015–1048. Also In Polymers for Tissue Engineering. Shoichet MS, Hubbell JA, eds, VSP BV, The Netherlands. 1998: 19–52.
3 Noninvasive Measurement of the Transient Adhesion of Cells to Blood-Contacting Surfaces Jun Li, M. Kurt Sly, Robert Chao, Anca Constantinescu, Padmakar Kulkarni, Michael Jessen, and Robert Eberhart The University of Texas Southwestern Medical Center at Dallas Dallas, Texas, U.S.A.
I. INTRODUCTION Contact with synthetic polymer and metal surfaces of blood-contacting equipment (pumpoxygenators, dialyzers, ventricular assist devices, vascular prostheses, etc.) activates host defenses, including the coagulation and inflammatory pathways, and the fibrinolytic cascade [1]. In largesurface contact situations, especially with cardiopulmonary bypass (CPB) equipment, the end result of these reactions includes reduced platelet function and survival and increased postperfusion bleeding times [2]. The resulting organ dysfunction may be viewed as a direct complication of CPB. The generalized inflammatory response is multifactorial and includes activation of complement, the initiation of fibrinolytic, kallikrein, kinin, and coagulation cascades within the body. It also includes the activation of platelets and neutrophils. The activation of platelets may render them less effective for hemostasis, leading to postoperative bleeding problems. The activation of neutrophils may lead to sequestration of neutrophils within the pulmonary vasculature, contributing to postoperative complications, including capillary leak syndrome and microvascular lung injury. Oxygenator membranes are believed to represent the greatest challenge to host inflammatory systems during CPB, due to the large blood-contacting surface area (0.6–2 m2) and relatively slow blood flow (⬍3 cm/s). How to improve the biocompatibility of CPB circuits becomes one of the critical factors that affect the performance of CPB circuits. Underlying the response of the organism to the CPB equipment is the initial protein adsorption and subsequent cell interactions at the foreign material–blood interface. Numerous studies have shown these interactions to be transient events, which are analyzed in detail in numerous laboratories. However, the focus on these interactions is generally lost when one moves to performance evaluation of full-scale clinical devices in vivo or ex vivo. Most works report at best the changes in formed elements between the inlet and outlet of the device under study. These changes represent a small difference of two large numbers, e.g., inlet and outlet whole blood platelet count. The large variances in each of these numbers preclude the accurate measurement of the transient uptake and release of cells from the device surface. These transient measurements are theoretically feasible by 55
56
Li et al.
noninvasive gamma scintigraphy. In fact, platelet responses to devices have been studied by various applications of noninvasive gamma counting, ranging from simple detectors to scintigraphic imaging systems [3–6]. However, these methods have not discriminated between those cells passing through the device without surface interaction and those cells that attach to the surface. We have modified a scintigraphy system and developed an algorithm to permit this discrimination, including the visualization of the kinetics of blood cell distribution shifts in both the oxygenator circuits and internal organs. We present it here in the context of CPB system analysis of platelet and neutrophil adhesion and release from the device surfaces. This analysis is used to demonstrate the quite different responses of these cells to different surface treatments of the oxygenators. The regional distribution of platelets and neutrophils in internal organs, during CPB and postperfusion, is also feasible with this system, although better characterized with a tomographic system, rather than a 2D system. Inspection of the simpler 2D application presented herein will provide the reader with the knowledge necessary for the tomographic 3D system application. The quantitative measurement system is composed of gamma scintigraphy combined with circulating cell counting. Additional measurements of cell activity, e.g., by flow cytometry or whole blood platelet or neutrophil impedance aggregometry, could be used if assessment of cell function were to be combined with that of cell adhesion and release. However, the focus of this chapter will be the modified gamma scintigraphy and analysis system. Dual-channel images of 111 In-labeled platelets and 99mTc-labeled neutrophils were acquired periodically by a GE 400T gamma camera in a 90-min CPB focusing on oxygenator circuits studies, followed by a 120min post-CPB focusing on organ studies. The images were digitized, analyzed, and carefully corrected for energy overlap; blood pool subtraction; gamma decay; attenuation from plastic phantom, tissue, and distance; ratio change of labeled to total cells; and isotope concentration change in plasma. Platelet and neutrophil distributions were determined on the microporous polypropylene membrane surfaces of two different types of oxygenators, the Cobe sheet membrane oxygenator and the Medtronic hollow fiber membrane oxygenator. Furthermore, these systems were evaluated with different types of surface treatments. First, that of a surface modifying additive (SMATM) coating, which is composed of polycaprolactone (PCL) and polydimethylsiloxane (PDMS) functional blocks, aiming to lower interfacial free energy, is one of the polymeric surfactant techniques. SMA coating on CPB circuits may reduce activation of cellular and protein blood components. SMA is licensed for coatings on Cobe DUOTM oxygenator circuits by Thoratec Laboratories [7]. Second we evaluated a popular heparin coating technique, using the CarmedaTM process: end-point attachment of heparin by covalent binding developed by Larm et al. [8] as applied to the Medtronic MaximaTM oxygenator circuit. For this report we also present another promising method which does not involve surface treatment, the infusion of trace amounts of nitric oxide (NO) in the oxygenator inlet gas port.
II. CELL SEQUESTRATION AND ISOTOPE LABELING A. Platelet Labeling The 111In-oxine method proposed by Thakur et al. [9] made it possible to label and quantify platelets for studies of acute and chronic phase thrombosis, and atherosclerosis and thrombosis. The interest in the use of 111In stemmed from its efficient incorporation and the ability of cells to retain the label after they have been introduced back into the subject. The relatively long half-life (2.8 days) allows studies to be performed for up to 5 days and the efficient gamma photons (173 keV 84% and 247 keV 94%) permit excellent images to be obtained with an
Adhesion Measurement at Blood-Contacting Surfaces
57
administered dose of less than 500 Ci [10]. Chelates are used as isotope carriers. The two most commonly used carriers are oxine (8-hydroxyquinoline) and tropolone [11,12]. Other chelates have been compared with these two agents. Tropolone and mercaptopyridine-N-oxide resulted in a high labeling efficiency in a comparison of oxine, oxine-sulfate, tropolone, and mercaptopyridine-N-oxide in human and rabbit platelets [11]. Labeling efficiency was not significantly different for any of these when a labeling medium free of plasma was used. The platelet life span was significantly longer when labeling was performed with tropolone in plasma compared with oxine in ACD-saline. In a similar study, however, labeling of human platelets with either 111In oxine in saline or 111In tropolone in plasma did not result in any difference in platelet kinetics [12]. 111 In-labeled platelet imaging, with its ability to noninvasively localize and quantitate platelet uptake, has considerably expanded the evidence supporting a platelet role in the thrombotic response to prosthetic materials. Multiple studies have demonstrated consistent 111In platelet accumulation on prosthetic surface in vivo. The experimental types of platelet–prosthetic material interactions that have been assessed using labeled platelets in animal and in human studies include arterial and venous prosthetic materials, autologous vein coronary bypass grafts, vascular catheters, mechanical heart valves, acutely placed grafts, long-term implanted grafts (⬎1 month), drug effects on graft platelet deposition, cardiopulmonary bypass, etc. [13]. Hope et al. [14] studied the kinetics and sites of sequestration of 111In-labeled platelets during CPB in patients. They found that the mean loss of 111In platelets was 13 Ⳳ 4%, with 10.8 Ⳳ 1.3% of administered platelets lost in the pump and oxygenator. The survival of the remaining platelets was markedly shortened to 58 Ⳳ 8 h. In the 48 hr following surgery most of the senescent platelets localized in the liver. To obtain platelets in as viable a condition as possible (not influenced by anesthesia), platelet harvesting is generally done from awake pigs, which are calmed by gentle handling and small amounts of food. Platelet activation must be inhibited during separation procedures, either by acidification of the labeling medium or addition of inhibitory prostaglandins. Platelet aggregation as measured by aggregometry was unaltered in the labeled platelet suspension compared with platelet aggregation in the unlabeled PRP. The aggregometry curve was similar, with the same aggregation velocity and aggregation maximum, which is an indication of intact platelet function [15]. When reinjected, labeled platelets retained radioactivity throughout their life spans of 158 Ⳳ 25 h. B. Neutrophil Labeling In evaluating a neutrophil-labeling technique, the most important criteria are satisfactory migration and recovery of the labeled cells in the circulation after reinjection. Diisopropyl fluorophosphonate (DFP-32), 111In-oxine, and 99mTc are commonly used isotopes for neutrophil labeling. The gamma emission recovery for 111In-oxine (or 111In-tropolone) [16], and 99mTc [17] in the circulating granulocyte pool is 35 and 10%, respectively. The labeling efficiency with 111In for neutrophil labeling is 80 to 90% for tropolone [18], but only 36% for oxine. The labeling efficiency with 99mTc by classical phagocytosis procedure is about 30 to 40%, compared to about 90% for human and canine blood with revised method [19]. The 6-h half-life of 99mTc was clearly too short for platelet survival and long-term imaging studies; however, it is fine for short-term studies like cardiopulmonary bypass. As a commonly accepted conclusion, 99mTc is inferior to 111In when considered for neutrophil labeling. The half-life of 111In-labeled neutrophils reported is 5 to 12 h [16,20]. The half-life for 99mTc-labeled neutrophils is not available. Dual radiotracer methods, with platelets labeled with 111In and red blood cells labeled 99m with Tc, were introduced by Powers et al. [21]. We used 111In-labeled platelets and 99mTc-
58
Li et al.
labeled neutrophils to study the platelet and neutrophil dynamic accumulations on oxygenator circuits. C. Platelet Distribution The distribution of viable platelets in the body is best assessed shortly after reinjection of the labeled platelets, once equilibrium has been reached. The attainment of equilibrium between injected labeled platelets and circulating platelets is influenced by a number of events: the mixing of the labeled platelets with circulating platelets; the size of the exchangeable splenic platelet pool; sequestration of irreversibly damaged platelets from the circulation in the reticuloendothelial system; and reentry of reversibly damaged platelets from the reticuloendothelial system to circulation. [22] In human beings [23,24], following intravenous injection of the labeled platelets as a bolus, the equilibrium of radioactivity was reached in the circulation within 30 to 60 min, with the different internal organs’ radioactivity expressed as of whole body radioactivity stable at the following percentages: heart 8%; spleen 34%; liver 12%. There are quite large variations between species due to size differences of internal organs size and the blood flow fraction. D. Platelet Sequestration Platelet sequestration is defined as the final irreversible accumulation of nonviable platelets in an organ or region. Normal platelets have a life span of 9 to 11 days in humans, 5 to 7 days in dogs, 5 to 6 days in baboon, 3 days in rabbits, and 3 to 4 days in rats. Basically, the number of labeled platelets sequestered in an organ can thus only be determined by quantification of organ radioactivity at the end of the platelet life span. 111In-labeled platelets are principally sequestered in the liver and spleen in most species, while other organs contribute smaller amounts (Table 1). For example, 111In-labeled platelet sequestration in kidney accounts for less than 3%, and heart and lung both contribute less than 1% of 111In-labeled platelet sequestration in baboons and rabbits [25,26]. In our study, the total time from reinjection of 111In-labeled platelets to the end of the experiment was only 8 h. Thus platelet sequestration occurs far from the end of the platelet life span. We can anticipate that there are large differences between the platelet sequestration patterns of our study and the cited studies, which are based on the investigation at the end of platelet life span. E. Neutrophil Distribution The distribution of neutrophils (granulocytes) is more complex than that of platelets. Following reinjection, gamma camera images showed marked sequestration of 99mTc-labeled neutrophil
Table 1 Sequestration Sites of 111In-Labeled Platelets Distribution (%) Species
Spleen
Liver
Humans Baboons Dogs Rabbits
33 ⫾ 13 34 ⫾ 3 45 ⫾ 10 14 ⫾ 4
39 ⫾ 9 35 ⫾ 7 33 ⫾ 6 40 ⫾ 7
Source: Ref. 22.
Remainder 28 ⫾ 14 30 ⫾ 5 22 ⫾ 13 41 ⫾ 0
Adhesion Measurement at Blood-Contacting Surfaces
59
in the lung, spleen, and liver [27,28]. Following injection of autologous radiolabeled neutrophils in intact normal subjects, there is prominent retention of neutrophils in the lungs, observed with 111 In,211 99mTc,212 and 51Cr. [29]. Clearance of the activity has been reported to occur over the ensuring 1 to 4 h. This early lung activity might be regarded as evidence in favor of the lung as a major site of granulocyte margination. The extensive lung sequestration is indicative of cell damage and activation. The progressive rise in cell-bound activity in peripheral blood for up to 4 h after injection of 111In-labeled neutrophils is related to the release of cells from the lungs [30]. Saverymuttu et al. also showed that granulocytes separated in plasma-enriched media and labeled in plasma with 111In-tropolone were sequestered in the lung to a much lesser extent than granulocytes isolated and labeled in saline. The spleen is another important organ regarding neutrophil margination. Lung sequestration and spleen 111In uptake are complementary, i.e., when lung sequestration was marked, the rate of uptake of activity by the spleen was reduced, and vice versa when transit was rapid. The liver also plays an important role in neutrophil margination due to the cell damage and cell activation. The absolute quantitation of the 111In activity present in the spleen (36%), liver (25%), and circulating granulocyte pool (37%) about 1 h after injection accounted for almost 100% of the injected dose [20].
III. GAMMA SCINTIGRAPHY The physical properties of 111In and 99mTc allow external imaging of their in vivo distributions and areas of localization. The two most important classes of imaging systems used for these purposes are the rectilinear scanner and the Anger scintillation gamma camera. Scintigraphic quantitation of platelet deposition in vivo with a gamma camera requires one to take into account the variations in geometry, attenuation, and sensitivity as well as the errors introduced by Compton scatter and superimposition of structures. Furthermore, due to the poor spatial resolution (2 mm at best) of the gamma camera, radiolabeled platelets circulating in the blood adjacent to a thrombus cannot be reliably distinguished from platelets incorporated into a thrombus and therefore often contribute to the total radioactivity measured for the thrombus. In our study, scintigraphic images were obtained with a large field of view gamma camera equipped with a mediumenergy, parallel-hole collimator and interfaced to a digital computer. Single channel analyzers on the gamma camera were set for the 247-KeV photopeak of 111In and the 140-KeV photopeak of 99mTc. The gamma camera portrays three-dimensional structures as two-dimensional images, actually determining excess platelet deposition relative to blood for the entire volume of tissue within the region of interest, not just for the vessel under study. 111In and 99mTc emit gamma photos with sufficiently different energies to be measured separately by the pulse height spectrometer in a gamma camera. However, 111In also emits a second gamma photon with an energy of 173 KeV that is sufficiently close to the energy of the 99mTc photon as to make discrimination between the two difficult [21]. In animal studies, the timing of platelet injection is important in determining the magnitude and time course of deposition.
A. Quantitative Analysis The quantitative analysis procedure is shown in Fig. 1. The total cell adhesion on the surface of the oxygenator can be evaluated by subtracting the total cells in the pool of the oxygenator from the total cell in the oxygenator. The total cells in the oxygenator can be obtained by comparing the gamma counts of ROI to a 5-mL whole blood calibrating sample drawn before
60
Li et al.
CPB from the dual channel images, while the total cells in the pool of the oxygenator can be obtained by cell counts multiplying the volume of the oxygenator in ROI. The cell sequestration in the internal organs can be evaluated in the similar way, except that the cell numbers in the blood pool are not subtracted due to the unavailability of the accurate parameters of the blood volume in the internal organs. The results need to be corrected for the following factors: 1. 2. 3. 4. 5. 6. 7.
Gamma decay Energy overlap Ratio of labeled to total cells Plastic phantom attenuation Distance attenuation Tissue attenuation Plasma radioactivity
where factor 6 is not applied to cell adhesion on the oxygenator and the pump, and factor 4 is not applied to cell sequestration in the internal organs. In the study to be described, the labeled/ total cell ratios used in this study actually contain the information of plasma radioactivity, so correction for factor 7 is not actually done. There was a difficult problem to be solved before this analysis could be called quantitative. The ratios of the labeled–to–total cell populations obtained from the previous section are actually the label ratios in the blood pool, rather than the ratios on the surface of the synthetic surfaces. Due to the occurrence of large changes in the labeled–to–total cell ratios during CPB and postCPB, especially for neutrophils, different models of correction for this factor were studied. For cell adhesion in the oxygenator circuits, an advancing average model was applied in numerical analysis. This model considers all the ratios to the current time point, with more emphasis on
Figure 1 Block diagram of quantitative cell adhesion analysis procedure.
Adhesion Measurement at Blood-Contacting Surfaces
61
the present than on the previous history. This characteristic matches the cells attachment and detachment dynamic process, in which the ratios of the labeled to total cells on the synthetic surface are more affected by the current ratios of the labeled to total cells in the blood pool. The advancing average model used in this study is R (t) = 1/2 R (t ) + 1/2 ( R(t − 15) + 1/2 {R (t − 30 ) + ... + 1/2 [R (5) + R (−20)]})
(1)
where R(t) is the ratio, at time t, of labeled to total cells on the surface, and R(t) is the ratio in the blood pool. The time step used in this study is 15 min. The model was used in correcting for the labeling ratios in cell adhesion on the surfaces on both the oxygenators and the pumps in Cobe studies. The Medtronic studies were not corrected for this factor due to the unavailability of the parameters in the protocol in the early research stage. The total number of labeled and unlabeled platelets (NP.Ttl) accumulated in each region of interest (ROI) for the oxygenator and blood pump were calculated as indicated in Eq. 2. (The complete set of equations and definitions are given in Appendix A of this chapter.)
NP.Tt1=[CB.1 *C Tc.s td.2 − CB.2*CTc.std.1] ÷ [C RO I1*C Tc.s td.2 − C ROI.2*C Tc.std.1] * η P.B (t= − 20)*VB*Ki R P.Sur f.*1
(2)
where CROI.1 and CROI.2 are the counts in the ROI for the 111In and 99mTc channels, respectively. CTc.std.1 and CTc.std.2 are the counts of the 99mTc point source in the field of view for the 111In and 99mTc channels, respectively, and CB.1 and CB.2 are the counts for the calibration blood sample in the field of view for the 111In and 99mTc channels, respectively. The whole blood platelet count for this sample, drawn 20 min prior to CPB, is P,B (t⳱ⳮ20). VB denotes the volume of blood in the sample tube (5mL). Ki represents the attenuation coefficients for the plastic masses of the oxygenator and blood pump, and includes the camera distance from the centroid of the device K i = (1 − A ij)−1
(3)
Aij is the attenuation for the ith label and jth device, summed over all j sources of attenuation for each label. RP.Surf is the ratio of labeled to unlabeled platelets on the device surface, determined by sequential cell and gamma activity counts for circulating blood samples acquired during the imaging period; it also compensates for the hemodilution effect on platelets.
IV. ANIMAL STUDIES Adult male farm pigs 35–45 kg in weight were fasted 24 h before being premedicated with ketamine hydrochloride (15–20 mg/kg IM) and atropine (0.05 mg/kg IM) and anesthetized with isofluorane (1–5%) by mask. A catheter was placed in the femoral vein for blood sampling. A 43-mL whole blood sample was drawn from the femoral vein into a sterile syringe containing 7 mL acid-citrate-dextrose anticoagulation formula and processed for platelet isolation and labeling. Isolated platelets were labeled with 111In-oxine, employing a modification of the technique of Thakur et al. [9]. Labeled platelets were resuspended in platelet-poor plasma. No platelet clumps were noted in the suspension, and quality control steps were taken to ensure that more than 80% incorporation of 111In-oxine into cells was obtained in all cases. Labeled platelets were reinfused through the femoral vein and allowed to circulate for 3 h before CPB was initiated. The radioactivity of labeled platelets varied between 1 and 3 mCi per 4 mL of injectate.
62
Li et al.
The animal was intubated and maintained on a ventilator with oxygen enrichment and isofluorane anesthesia. ECG, blood pressure, and body temperature were monitored throughout the procedure, maintaining mean blood pressure at 100 Ⳳ 20 mmHg and temperature at 37 Ⳳ 1⬚C. The animal was then prepared for CPB using a circuit including the Cobe DuoTM oxygenator with incorporated heat exchanger, a centrifugal pump (BiomedicusTM), and a closed collapsible venous return reservoir (Cobe VRBTM). An angled tip arterial cannula was inserted into the ascending aorta and a two-stage venous cannula was inserted in the right atrium. Systemic heparin (300 U/kg IV) was administered, and blood samples were drawn periodically for activated clotting time (ACT), hematocrit, and circulating platelet measurements (simultaneous Unopette and Coulter methods). Once proper ACT levels were verified, animals were placed on CPB for 90 min. The CPB blood flow rate was maintained at 75–100 mL/kg/min. ACT levels were initially quite high, but decreased throughout the CPB period. Supplemental heparin was administered as needed in order to maintain the ACT values greater than 3⳯ the pre-CPB control value. Following 90 min of CPB, the animal was weaned from the pump-oxygenator circuit and maintained an additional 120 min. One study group comprised circuits with an amphophilic silicone–caprolactone oligomer (SMATM) coating on all elements except the centrifugal pump. This oligomer and the membrane coating technique have been described previously [31]. The other groups employed uncoated pump-oxygenator circuits. For the NO study group, 30 mL polypropylene syringes were purged with N2 and filled with nitric oxide gas. The syringes were placed in a syringe pump, attached to 20-ga. needle tipped polyethylene tubing, and connected to the oxygenator gas inlet port via a rubber septum. The NO flow rate was 1 ml/min (500 ppm) for the first 60 min of CPB and 2 mL/min (1000 ppm) for the next 20 min; NO flow was stopped for the final 10 min of CPB. Selection of the oxygen entry port obviated the need for development of a second NO–blood mass exchanger at a proximal site. The experimental setup shown in Fig. 2 utilizes a 2D gamma camera (GE Maxi 400T General Purpose Scintillation Camera System, General Electric, Milwaukee, WI) fitted with a medium energy parallel hole collimator. Gamma analog intensity output is fed to a Technicare 560 image acquisition computer for digitizing, display, and storage. Intensity distributions were ported to a Macintosh IICI computer for further image, analysis, employing NIH Image, version 1.47. Flood field calibration of the gamma sensitivity distribution was done before each experiment. Simultaneous dual channel images of 111In-labeled platelets and 99mTc-labeled neutrophils (not described further in this report) were acquired. The computer was programmed for data acquisition with 64 ⳯ 64 word matrix, on an 8-inch floppy diskette. All images were corrected for 111In-99mTc energy overlap, gamma absorption by the device, emission from circulating blood in the device and the camera–device distance effect [32,33]. Note that this computer has since been upgraded to a modern PC employing LabView software. Control and treatment group NP.Ttl (Eq. 2) were compared for each device, at each time sampling point, by the paired Student’s t test. The circulating platelet counts at the same time points were also compared by the paired Student’s t test.
V. RESULTS There were no between-group differences in weight, pre-CPB hematocrit, heparin dosage, or total fluid administration [33]. There were also no differences in mean blood pressure, heart rate, or temperature. A hemodilution effect due to CPB was observed for all groups, demonstrated by the decrease in hematocrit from pre-CPB control values. Activated clotting times on CPB were significantly elevated from pre-CPB controls due to systemic heparinization. ACT values tended to be much higher than the 3⳯ control value at the beginning of CPB, and declined
Adhesion Measurement at Blood-Contacting Surfaces
63
Figure 2 Schematic of the pump oxygenator imaging layout in relation to the animal. The camera field of view is indicated by the dashed circle. Regions of interest (ROI) within the field of view were selected by software to separately analyze the oxygenator and centrifugal pump head images.
thereafter. Post-CPB ACT values continued to decline, but did not reach pre-CPB values in the 2-h postoperative monitoring period. Figure 3a shows the platelet accumulations on the oxygenator surfaces for the Cobe control and SMA groups. Significant differences in the patterns of adhesion are observed with time and SMA treatment. Maximal platelet adhesion for the control group is observed 20 to 50 min following initiation of CPB. About 2.5 ⳯ 1011 platelets (⬎20% of the pre-CPB circulating cells) adhere to the untreated surfaces during this period. Following an initial transient which mirrors the control group results, platelet adhesion to the SMA-treated surfaces decreases significantly, to 6% or less of the pre-CPB control circulating cell count. Figure 3b depicts the trends in circulating platelet counts for the control and SMA treatment groups. Upon initiation of CPB, platelet counts decrease to values below those predicted by the hemodilution effect (the dashed line). Thereafter, there is a tendency for circulating platelets to increase for the SMA treatment group, which becomes statistically significant with reference to the corresponding control value at several sampling times in the 35 to 75 min CPB period. Comparison of the decrease in oxygenator-adherent platelets for the SMA group (Fig. 3a) with the corresponding increase in circulating platelets (Fig. 3b) suggests that some adherent platelets may be released to the circulation during the CPB period. In contrast, there is no statistically significant reduction in adherent platelets below the maximum values at 20–50 min of CPB for the control group, and no corresponding increase in circulating platelet count. Figures 4a and b show the adherent and circulating platelet counts, respectively, for the control and NO infusion groups. NO treatment provides an even larger reduction in adherent platelets than that observed for the SMA treatment group, without the initial transient platelet adhesion. NO did not produce a statistically significant elevation of circulating platelets in this point-by-point comparison. Nitric oxide reacts avidly with oxyhemoglobin to form methemoglobin (metHb). Control experiments indicated that metHb rose slightly during CPB, remaining below 1%. In contrast, metHb rose to 4% of total hemoglobin during the first hour of NO administration (500 ppm); when the NO dose was increased to 1000 ppm the metHb level rose to 8.6%; the metHb level decreased upon cessation of NO administration.
64
Li et al.
Figure 3 Platelet responses to CPB with the SMA-treated Cobe oxygenator system. (a) Adhesion to the surfaces within the oxygenator. (b) Circulating platelet counts (Unopette method), normalized to the 20min pre-CPB values; similar values were obtained by the Coulter method. Untreated: controls (n⳱8); SMA: all circuit elements except the pump were coated with the surface-modifying additive (n⳱8). CPB begins at time 0 in both panels. The theoretical hemodilution effect is indicated by the dashed line. * p⬍0.05. Note: error bars denote the standard error of the mean in all figures. (From Ref. 33.)
Platelet adhesion to the Biomedicus pump for the control and SMA series is shown in Fig. 5. A monotonic rise and stabilization at approximately 1.4 ⳯ 1010 cells is observed, with no difference between control and treatment groups. Note that the pump surface was not treated with SMA for this series. Steady accumulations of neutrophils occur on the Cobe oxygenator membranes throughout the period of CPB (Fig. 6). This monotonic increase in neutrophil concentration contrasts with the initially rapid accumulation of platelets on the untreated membrane, followed by periods of stabilization, then decrease, or the very rapid increase and prompt decrease in platelet adhesion on the SMA-treated Cobe membrane surface. No significant differences in neutrophil concentrations are seen between the untreated and the SMA-treated surfaces. About 1.5 ⳯ 1010, or 40%, of the circulating neutrophils were found to adhere to both untreated and treated oxygenator surfaces by the end of the 90 min period of CPB. The corresponding surface concentrations are three neutrophils per 1000 nm2 microns for both groups. This contrasts with 47 platelets per 1000 nm2 for the untreated surface and 11 platelets per nm2 for the SMA-treated surface at the same CPB termination point.
Adhesion Measurement at Blood-Contacting Surfaces
65
Figure 4 Platelet responses to CPB with the nitric oxide treatment of the Cobe oxygenator system. (a) Adhesion to the surfaces within the oxygenator. (b) Circulating platelet counts (Unopette method), normalized to the 20-min pre-CPB values; similar values were obtained by the Coulter method. Untreated: controls (n⳱10); NO: nitric oxide administration (n⳱10). The NO infusion regimen is indicated on both figures. * p⬍0.05. (From Ref. 33.)
Platelet and neutrophil retentions on the Cobe oxygenator surfaces following a gentle saline rinse post-CPB were measured. Significantly lower platelet retention percentages were observed for the SMA-treated surfaces compared with the untreated surfaces. More than 10 ⳯ 1010, or 50%, of the adherent platelets before rinse remained on the untreated group surface, whereas less than 4 ⳯ 1010, or 20%, of the previously adherent platelets remained on the SMA group surfaces. There were no differences in neutrophil retention on the two types of surface following saline rinse. Approximately 1.1 ⳯ 1010, or 60–70%, of the adherent cells were retained following rinse. Thus in terms of both accumulations during blood contact and retention following rinse, SMA treatment appears to influence only platelet behavior. There were no significant differences in platelet accumulations on the treated and untreated Medtronic oxygenator (Fig. 7). In contrast to the common platelet adhesion behavior on the treated and untreated Medtronic oxygenator membranes, there were significant differences in neutrophil accumulations (Fig. 8). On untreated surfaces there were 3 ⳯ 109 cells, or 12%,
66
Li et al.
Figure 5 Platelet adhesion on the Biomedicus pump surfaces during CPB. Untreated: controls (n⳱8); SMA: all circuit elements except the pump were coated with the surface-modifying additive (n⳱8). (From Ref. 33.)
of the prebypass circulating neutrophil population, whereas on CarmedaTM heparin treated surfaces there were less than 109, or 3%, of the pre-CPB neutrophil population. The corresponding cell surface concentrations at the termination of CPB are 23 platelets per 1000 nm2 for both treated and untreated surfaces, and only 0.5 neutrophil per 1000 nm2 for the Carmeda heparin treatment group versus 1.5 per 1000 nm2 for the control group. Platelet retention post-CPB on Carmeda-treated Medtronic oxygenator membranes is only half that on untreated surfaces, despite the lack of difference in ion accumulations during blood contact. Thus the cells may have been less strongly adherent and more capable of release following a suitable perturbing event (fluctuation in blood flow or pressurization). The differences in absolute neutrophils adherent following post-CPB rinse are significant, with approximately 1.5 ⳯ 108 neutrophils measured on the Carmeda treatment group surfaces, versus 4 ⳯ 108 cells on the control group surfaces. However, in terms of percent retention there are no significant differences: about 45% of the neutrophils adherent before rinse are retained on both treated and
Figure 6 Neutrophil adhesion on Cobe oxygenator surfaces during CPB. Untreated (n⳱8); SMA: all circuit elements except the pump were coated with the surface modifying additive (n⳱8). (From Ref. 32.)
Adhesion Measurement at Blood-Contacting Surfaces
67
Figure 7 Platelet adhesion on Medtronic oxygenator surfaces during CPB. Carmeda heparin-treated group (n⳱4) versus untreated group (n⳱4). (From Ref. 32.)
untreated group surfaces. Statistics were not computed to quantify platelet and neutrophil retention differences on Cobe and Medtronic oxygenator membranes. However, there appears to be at least a qualitative difference: 45% retention for the Medtronic class versus 60% retention for the Cobe class. This may reflect subtle differences in washing efficiency, although as noted there was only a small difference in the wall shear at equal flow rate.
VI. DISCUSSION This study is designed to measure platelet adhesion and release kinetics in a realistic model of extracorporeal oxygenation. The pig is the animal of choice since (1) resting cardiac output is similar to that of man, permitting adult cardiopulmonary bypass equipment, priming, and perfu-
Figure 8 Neutrophil adhesion on Medtronic oxygenator surfaces during CPB. Carmeda heparin-treated group (n⳱4) versus untreated group (n⳱4). (From Ref. 32.)
68
Li et al.
sion procedures to be employed; (2) cardiac and major blood vessel dimensions are similar to those of man, enabling clinical aortic and venous cannulation techniques and equipment to be used; and (3) there are no major differences in platelet numbers and anticoagulation treatment between pig and man. The SMA surface modification is designed to render CPB circuits more biocompatible, thereby enhancing recovery of normal homeostasis following periods of circulatory support. In the present study, circuits treated by the SMA process were found to accumulate fewer platelets and maintain higher circulating platelet counts during CPB. Two clinical studies have recently confirmed the platelet protective effect of SMA-coated CPB circuits, which supports our observations [34,35]. In addition to increased circulating platelet counts, lower prothrombin activation and lower platelet glycoprotein IIIa antibody binding on washed CPB tubing surfaces was observed with SMA treatment [34]. Higher platelet counts with SMA treatment, and lower indices of platelet activation, thrombin generation and plasmin-␣2-antiplasmin complex formation were observed in the other study [35]. We did not measure platelets sequestered within internal organs in this study, a large and volatile pool, which might reasonably be expected to account for a substantial portion of the released platelets. As reported elsewhere, we observed large decreases in this platelet organ pool following CPB for the untreated study group and partial restorations to pre-CPB control values with SMA treatment [36]. We did not study the function of released platelets, which might have been altered by interactions with the foreign surfaces. Studies employing flow cytometry and platelet membrane receptor antibodies or other appropriate methods are recommended to resolve this question. The platelet surface densities observed on the oxygenator at 60 min of CPB (36 and 8 platelets per 1000 m2 for the control and SMA surfaces, respectively) are in the range observed at similar times in a gamma counter–based quantitative study of PVC arteriovenous shunt tubing [3]; this experiment employed a ‘‘stop flow and wash’’ technique prior to gamma emission measurement which may have removed cells adherent during blood contact. We performed a separate calibration experiment, washing and fixing the oxygenator post-CPB and comparing SEM- and scintigraphy-derived platelet densities. The observed platelet densities agreed within Ⳳ25% for these washed surfaces. Infusion of nitric oxide into blood oxygenators has been shown to decrease platelet consumption [6,37,11]. The platelet-sparing effect of NO treatment is compared in this study to that of the SMA siloxane–caprolactone oligomer. In contrast to the SMA series results, no transient platelet deposition was observed on the oxygenators with NO administration, suggesting that the treatment adequately blunted the platelet response throughout the period of CPB (Figs. 3a and 4a). Despite this improvement, and in contrast with the SMA series results, there was no significant increase in circulating platelet counts with NO administration (Fig. 4b). This may be due in part to the higher variances in circulating platelet count for the NO group and may reflect the more difficult control of ACT levels in the NO series. Relatively high doses of NO were employed in this study (500–1000 ppm), which would not be clinically relevant. We selected the high dose in order to ensure that sufficient NO was delivered across the oxygenator membrane in an active form to provide an antiplatelet effect. We have evidence (unpublished results) that a lower NO dose (⬃100 ppm) is also effective. Furthermore, marked potentiation of the antiplatelet effect of NO has been observed in vitro with Zaprinast, a phosphodiesterase V inhibitor39, suggesting that even lower NO doses may be employed during CPB. Our scintigraphic studies demonstrated initial transient adhesion of platelets to oxygenators in the control and SMA groups but not the NO group. An interesting hypothesis of the control mechanism is suggested from studies of platelet membrane receptors. As shown by Kieffer, GPIb (von Willebrand factor) receptor expression accounts for early platelet adhesion to surfaces [40]. Furthermore, Kieffer suggests that up-regulation of the platelet membrane GPIIb/IIIa (fi-
Adhesion Measurement at Blood-Contacting Surfaces
69
brinogen) receptor is controlled by a signaling linkage from GPIb expression. The lack of free calcium inhibits upregulation of GPIb from the open canalicular system, thus inhibiting platelet adhesion. NO promotes cytoplasmic Ca2Ⳮ sequestration in mitochondria, and sequestration is extended by Zaprinast [39]. Thus NO may interfere with the membrane-signaling events required for platelet activation. Mellgren et al. reported, in a small patient series, that low dose (40 ppm) NO administration did not improve platelet count nor did it reduce ADP-induced platelet aggregation; however, it did lower expression of GPIb [41]. Furthermore, there was a significant decrease in GPIIb/IIIa expression for retained washed platelets in the Gu study of SMA-treated oxygenators [34]. These observations support the notion that critical signaling events in the platelet pathway to the surface are blunted by SMA and NO treatment. SMA may inhibit GPIIb/ IIIa expression but not GPIb expression, and thus may not block initial platelet binding to the oxygenator. In contrast, NO may inhibit GPIb expression sufficient to prevent the initial binding of the platelet. This argument does not account for the role of fibrinogen binding. The siloxane–caprolactone oligomer might interfere with fibrinogen binding, or alter its bound conformation, confounding the platelet GPIIb/IIIa membrane receptor. The oxygenator is not the only potentially traumatic device in the CPB circuit; the pump, venous reservoir, heat exchanger, and other accessories may contribute to cell activation and sequestration. Venous reservoirs, in particular, may promote platelet sequestration and thrombus formation, especially at low blood flow rate. The centrifugal pump operates at high fluid wall shear stresses, in the range of 400–800 dyne/cm2. Platelet surface densities on the Biomedicus pump increased monotonically during CPB to a steady state of approximately 216/1000 m2, which is higher than the maximum platelet surface density on the oxygenator surface (approximately 51/1000 m2). The fluid shear rate in the oxygenator is much lower, approximately 36 dyne/cm2, according to a packed bed calculation (courtesy of M. Voorhees). One might expect the higher shear rate in the centrifugal pump to induce detachment of cells; this was not observed, and is in keeping with reports of higher platelet adhesion on denuded blood vessels at high shear rate [42]. No differences were observed in platelet adhesion to the centrifugal pump surfaces between the control and treatment groups during the CPB period. However, there was no SMA coating on the pump surfaces, the only element of the oxygenator system not treated in this study group. Thus one may infer that the influence of the SMA coating is limited to the immediate vicinity of the foreign surface. There are several limitations to the animal study protocol which warrant mention. These experiments were conducted under normothermic conditions, with no aortic cross-clamping. Myocardial protection was not employed. The Cobe Duo oxygenator is comprised of 150 membrane-mesh sandwich structures with a total polypropylene surface area of 4.9 m2 (2.6 m2 microporous membrane and 2.3 m2 mesh in a plane-parallel configuration). The distribution of adherent platelets between the mesh and the microporous membrane was not determined. A quadrant-by-quadrant analysis of the oxygenator indicated uniform platelet density distributions were observed within the exchanger, but significant variations were observed within the manifolds (P. Shastri, unpublished observations). Sporadic release of manifold-entrapped cells into the exchanger section was not evaluated in this study. We employed a 2D imaging technique with a gamma camera spatial resolution of 7 mm. Lumped corrections for gamma absorption within the devices were employed which do not account for spatial variations in absorption. These comments notwithstanding, statistically significant differences were observed in platelet accumulations in the pump and oxygenator as a function of time and device treatment. VII. CONCLUSION The gamma scintigraphy system permitted the visualization and quantification of platelet and neutrophil mass distributions within intact pump oxygenators. The transient adhesion of large
70
Li et al.
numbers of platelets, up to 24% of the circulating platelet mass, was observed on the surfaces of a microporous polypropylene sheet membrane oxygenator during a typical 90-min period of normothermic cardiopulmonary bypass. Much lower platelet accumulations, but higher platelet surface densities, were observed on the high fluid shear rate centrifugal pump. SMA treatment significantly decreased platelet adhesion in the oxygenator, following a short transient accumulation period; these released platelets increased the circulating platelet count. SMA treatment did not alter the monotonic increase of adherent neutrophils on the oxygenator surfaces. Nitric oxide infusion significantly reduced platelet but not neutrophil accumulations in the oxygenator. Both SMA treatment and NO administration may prevent the exhaustion of platelets during cardiopulmonary bypass in this system. Somewhat lower but still significant platelet adhesion was observed on the surfaces of a microporous polypropylene tube/shell blood oxygenator. A heparin surface treatment (Carmeda) did not lower but in fact raised the platelet surface density in this oxygenator. The same treatment decreased neutrophil surface density. Employing similar methods, it should be possible to measure, simultaneously, the accumulations of blood proteins and cells on pump-oxygenator and other medical device systems in vivo. Finally, the ability to measure local transients in cell activity by noninvasive gamma scintigraphy may lead to improved understanding of cell signaling events, in medical devices and in internal organs.
ACKNOWLEDGMENTS This work was supported by a grant from the Texas Advanced Technology Program and contributions from Cobe Laboratories. The assistance of many workers is gratefully acknowledged, including Rohan Bhujle, Peter Ching, James Clift, Mike Dollar, Debbie Douglass, Kevin Esau, Keith Est, Jeff Floyd, John Gaffke, Fred Harris, Nicklett Johnston, Kim Jones, Helen Leonhardt, Prasad Sistla, Dorothy Smith, and Rui Xiong.
APPENDIX A A. Cell Surface Adhesion (Oxygenator/Pump) Step 1 Images of the circuitry are produced by the gamma camera, and the region of interest (ROI) for the oxygenator (or pump) is defined. Simultaneous counts are acquired from two channels: one for 111In and one for 99mTc. The counts are corrected for gamma decay, with the half-life times of 4046 and 360 min for 111In and 99mTc, respectively. The following calculation is done to correct for energy overlap: CRO I.1 = A Ind .RO I *
CTc .S td .1 CInd .S td .1 + A Tc .RO I * ATc .St d AInd .St d
(A-1)
C RO I.2 = A Ind .RO I *
CTc .St d .2 CInd .S td .2 + ATc .RO I * A Tc .St d AInd .St d
(A-2)
Step 2 The corrected counts from step 1 are then compared with counts from standard isotope point sources (both 111In and 99mTc) placed inside the camera field of view. The following equations convert the counts within the ROI to the activity of these isotopes (in Ci). Solving for AInd.Roi and ATc.Roi, we have
A Ind .RO I =
A Ind .S td (C ROI.1 *CTc .St d .2 − C ROI.2 *CTc .S td .1) C Ind .S td .1 *CTc .S td .2 − CInd .S td .2 *C Tc .S td .1
(A-3)
Adhesion Measurement at Blood-Contacting Surfaces
A Tc .ROI =
A Tc .S td (CRO I.1 *CInd .St d .2 − CRO I.2 *C Ind .S td .1) CTc .S td .1 *CInd .St d .2 − CTc .St d .2 *CInd .S td .1
71
(A-4)
Step 3 A 5-mL sample of blood drawn from the animal after circulation of labeled cells and before the onset of cardiopulmonary bypass is also placed inside the camera field of view. Calculations similar to those in steps 1 and 2 are performed to calculate activity of each isotope (in Ci) within this blood sample. Similarly, we repeat steps 1 and 2 for the calibrating blood tube, thus
A Ind .B =
A Ind .S td (CB .1 *CTc .S td . 2 − CB .2 *CTc .St d .1) CInd .St d .1 *C Tc .St d .2 − C Ind .S td .2 *CTc .St d .1
(A-5)
A Tc .B =
A Tc .S td (CB .1 *C Ind .S td .2 − CB .2 *CInd .St d .1) C Tc .St d .1 *C Ind .S td .2 − CTc .S td .2 *C Ind .S td .1
(A-6)
Step 4 Knowing the activity within ROI, and both activity and cell counts within the calibrating blood tube, the following equations calculate the total blood cell counts within the ROI. Plastic and distance attentions of gamma energy are corrected. The ratios of labeled to total cells are also corrected: N P .Ttl =
A Ind .RO I 1 *ηP .B (t = −20 )*V B *KInd .Attn * A Ind .B R Su rf .Plt (t)
(A-7)
NN .Ttl =
A Tc .RO I 1 * ηN .B (t = −20)*V B *KTc .Attn * A Tc .B R S urf .Ne ut (t )
(A-8)
where
KInd . Attn =
1 1 * 1 − B Oxy .Ind 1 − B Dist .Ind
KTc .Attn =
1 1 * 1 − B Oxy .Tc 1 − B Dist .Tc
Further simplifying (A-7) and (A-8) gives
N P .Ttl =
CRO I.1 *CTc .S td .2 − CRO I.2 *CTc .St d .1 1 *ηP .B (t = −20)*VB *KInd . Attn * CB .1 *CTc .S td .2 − CB .2 *CTc .S td .1 R surf .Plt (t )
NN .Ttl =
CRO I.1 *CInd .S td . 2 − CRO I.2 *CInd .St d .1 1 *ηN .B (t = −20)*V B *KTc .Attn * R S urf .Ne ut (t ) CB .1 *CInd .S td . 2 − CB .2 *CInd .St d .1
(A-9)
(A-10)
Step 5 Additional studies determined the volume of blood within the oxygenator (or pump). The final calculation subtracts calculated platelet and neutrophil counts within this circulating blood, leaving values for adherent platelets and neutrophils on the surface of the ROI. NP .Dep = N P .Ttl − ηP . B ( t)*V Oxy
(A-11)
NN .Dep = NN .Ttl − ηN .B (t )*V Oxy
(A-12)
72
Li et al.
Steps 1–5 are then repeated for each time point and ROI.
SYMBOLS AInd.ROI 111In activity of ROI ATc.ROI 99mTc activity of ROI AInd.Std 111In activity of the standard 111In point source ATc.Std 99mTc activity of the standard 99mTc point source AInd.B 111In activity of the calibrating blood sample tube ATc.B 99mTc activity of the calibrating blood sample tube AInd.w/o 111In activity measured without material attenuatin ATc.w/o 99mTc activity measured without material attenuation AInd.Oxy 111In activity measured with material attenuation ATc.Oxy 99mTc activity measured with material attenuation CROI.1 Image counts in ROI, 111In channel CROI.2 Image counts in ROI, 99mTc channel CInd.Std.1 Image counts of the standard 111In point source, 111In channel CInd.Std.2 Image counts of the standard 111In point source, 99mTc channel CTc.Std.1 Image counts of the standard 99mTc point source, 111In channel CTc.Std.2 Image counts of the standard 99mTc point source, 99mTc channel Ht(t) Hematocrit at time t KInd.Attn 111In activity attenuation compensation KTc.Attn 99mTc activity attenuation compensation NP.Ttl Total platelets in ROI (surfaceⳭblood pool) NN.Ttl Total neutrophils in ROI (surfaceⳭblood pool) NP.Dep Total platelets adherent on the surface of ROI NN.Dep Total neutrophils adherent on the surface of ROI nP.B (t)⳱ⳮ20 Platelet count in the calibrating blood sample tube (t ⳱ ⳮ20 min) nN.B(t)⳱ⳮ20 Neutrophil count in the calibrating blood sample tube (t ⳱ ⳮ20 min) VB Volume of blood in the sample tube VOxy Volume of the ROI (oxygenator volume: 400 mL for Cobe, 260 mL for Medtronic; pump volume: 90 mL for Biomedicus in both cases) BDis.Ind Distance attenuation for 111In BDis.Tc Distance attenuation for 99mTc BOxy.Ind Material (oxygenator or pump) attenuation for 111In BOxy.Tc Material (oxygenator or pump) for 99mTc BOrg.Ind Tissue attenuation for 111In; different values for different organs BOrg.Tc Tissue attenuation for 99mTc; different values for different organs Rsurf.plr(t) Normalized ratio of labeled to total platelets on surface (at time t) RSurf.neut(t) Normalized ratio of labeled to total neutrophils on surface (at time t) ROrg.plr(t) Normalized ratio of labeled to total platelets in organs (at time t) ROrg.neut(t) Normalized ratio of labeled to total neutrophils in organs (at time t)
REFERENCES
1. Clagett GP, Eberhart RC. Artificial devices in clinical practice. In: Colman RW , Hirsh J , Marder VJ (Editors), Salzman EW, eds. Hemostasis and Thrombosis: Basic Principles and Clinical Practice. 3rd Edition. Philadelphia: J.B. Lippincott, 1994:1486–1505.
Adhesion Measurement at Blood-Contacting Surfaces
73
2. Edmunds LH. Why cardiopulmonary bypass makes patients sick: strategies to control the bloodsynthetic surface interface. Adv. Cardiac Surg. 1995; 6:131–167. 3. Ihlenfeld JV, Mathis TR, Barber TA, Mosher DF, Riddle LM, Hart AP, Hart SJ, Cooper SL. Transient in-vivo thrombus deposition onto polymeric biomaterials: role of plasma fibronectin. Trans. ASAIO 1978; 24:727–732. 4. Dewanjee MK, Vogel SR, Peterson KA, Lim MF, Kaye MP. Trans. ASAIO 1981; 27:197–201. 5. Megerman J, Christenson JY, Hanel KC, Strauss HW, Abbott WM. Imaging vascular grafts in vivo with indium-111–labeled platelets: influence of timing on image interpretation. Ann. Surg. 1983; 198:178–184. 6. Sly MK, Prager MD, Li J, Harris FB, Shastri P, Bhujle R, Chao RY, Kulkarni PV, Constantinescu A, Jessen ME, Eberhart RC. Platelet and neutrophil distributions in pump-oxygenator circuits. III. Influence of nitric oxide gas infusion. J. ASAIO 1996; 42:M494–499. 7. Ward RW. Surface-modifying additives for biomedical polymers. IEEE Eng. Med. Biol. 1989; 8: 2–6. 8. Larm O, Larsson R, Olsson P. Surface immobilized heparin. In: Lane DA (Editors), Lindahl U, eds. Heparin, Chemical and Biological Properties: Clinical Applications. Boca Raton: CRC Press, 1989: 597–608. 9. Thakur ML, Welch L, Malech HL, Gottschalk A. 111Indium-labeled human platelets: improved method, efficacy and evaluation. J. Nucl. Med. 1981; 22:381–385. 10. Thakur ML. Radionuclides and chelates used for platelet labeling. In: Heyns AP , Badenhorst PN (Editors), Lotter MG, eds. Platelet Kinetics and Imaging, CRC Press; I:23–33. 11. Hill-Zobel RL, Gannon S, McCandless B. Effects of chelates and incubation media on platelet labeling with indium-111. J. Nucl. Med. 1987; 28:223–228. 12. Kotze HF, Heyyns AduP, Lotter MG. Comparison of oxine and tropolone methods for labeling platelets with indium-111. J. Nucl. Med. 1991; 32:62–66. 13. Stratton JR. Platelet kinetics and imaging of prosthetic materials. In: Heyns AduP , Badenhorst PN (Editors), Lotter MG, eds. Platelet Kinetics and Imaging. Vol. II. Boca Raton: CRC Press:21–43. 14. Hope AF, Heyns AP, van Reenen OR, de Kock F. Kinetics and sites of sequestration of indium111–labeled human platelets during cardiopulmonary bypass. J. Thor. Cardiovasc. Surg. 1981; 81: 880–886. 15. Softeland E, Hervig T, Framstad T, Holmsen H. Simple method for labeling of porcine platelets with indium-111 oxine. Lab. Animal Sci. 1995; 45:195–198. 16. Weiblen J, Forstrum L, McCullough J. Studies of the kinetics of In-111 granulocytes. J. Lab. Clin. Med. 1979; 94:246–252. 17. Linhart-Colas N, Meignan M, Bok B. ‘‘In vivo’’ kinetics of technetium-99m–labeled leukocytes in dogs and the effect of an abcess. Biomedicine 1980; 32:133–139. 18. McAfee JG, Subramanian G, Gagne G. Present trends and future directions in ‘‘leukocyte labeling’’. In: Thakur ML, ed. Radiolabeled Cellular Blood Elements: Pathophysiology, Techniques and Scintigraphic Applications. New York: Plenum Press, 1985:265–284. 19. Schroth HJ, Oberhausen E, Berberich R. Cell labeling with colloidal substances in whole blood. Eur. J. Nucl. Med. 1981; 6:469–478. 20. Peters AM, Saverymuttu SH, Lavender JP. Granulocyte kinetics. In (Editor) Thakur ML, ed. Radiolabeled Cellular Blood Elements: Pathophysiology, Techniques and Scintigraphic Applications. New York: Plenum Press, 1985:285–303. 21. Powers WJ, Hopkins KT, Welch MJ. Validation of the dual radiotracer method for quantitative In111 platelet scintigraphy. Thromb. Res. 1984; 34:135–145. 22. Kotze HF, Wessels P, Pieeters H. Kinetics, redistribution and sites of sequestration of normal platelets. In: Heyns AP , Badenhorst PN (Editors), Lotter MG, eds. Platelet Kinetics and Imaging: vol. I Clinical Applications. Boca Raton: CRC Press, 1985:107–123. 23. Heyns AP, Lotter MG, Badenhorst PN, Reene V. Kinetics, distribution and sites of destruction of indium-111–labeled platelets. B. J. Haematol. 1980; 44:269–292. 24. Klonizakis I, Peters AM, Fitzpatrick ML, Kensett MJ. Radionuclide distribution following injection of Indium-111–labelled platelets. B. J. Haematol. 1980; 46:595–599. 25. Heyns AP, Lotter MG, Kotze HF, Pieters H. Quantitation of in vivo distribution of platelets labeled with 111In-oxine. J. Nucl. Med. 1982; 23:943–950.
74
Li et al.
26. Hill-Zobel RL, Scheffel U, McIntyre PA, Tsan MF. 111In-oxine–labeled platelets: in vivo distribution and sites of sequestration. Blood 1983; 61:149–153. 27. Jemionek JF, Contreras TJ, French JE. Technique for increased granulocyte recovery from human whole blood by counterflow centrifugation-elutriation. I. In vivo analysis. Transfusion 1979; 19: 120–131. 28. Schroth HJ, Oberhausen E, Berberich R. Cell labeling with colloidal substances in whole blood. Eur. J. Nucl. Med. 1981; 6:469–475. 29. Harvey WC, Silva J. Detection of abcess with Cr-51–labeled leukocytes. J. Nucl. Med. 1974; 15: 375–382. 30. Saverymuttu SH, Peters AM, Danpure HJ, Reavy HJ. Lung transit of In-111–labeled granulocytes: relationship to labeling techniques. Scand. J. Haematol. 1983; 30:151–161. 31. Tsai CC, Deppisch RM, Forrestal LM, Ritzau GM, Oram AD, Gohl HJ, Voorhees ME. Surface modifying additives for improved device–blood compatibility. J. ASAIO 1994; 40:M619–624. 32. Li J. Quantitative evaluation of platelet and neutrophil adhesion in membrane oxygenator circuits and internal organs, 1996. 33. Li J, Sly MK, Chao RY, Constantinescu A, Kulkarni PV, Wians FH, Jessen ME, Eberhart RC. Transient adhesion of platelets in pump-oxygenator systems: influence of SMA and nitric oxide treatments. J. Biomater. Sci.: Polym Ed. 1999; 10:235–246. 34. Gu YJ, Boonstra PW, Rijnsburger AA, Haan J, van Oeveren W. Cardiopulmonary bypass circuit treated with surface-modifying additive: a clinical evaluation of biocompatibility. Ann. Thorac. Surg. 1998; 65:1342–1347. 35. Rubens FD, Labow RS, Lavallee GR, Watson MI, Robblee JA, Voorhees ME, Nathan HJ. Hematologic evaluation of cardiopulmonary bypass circuits prepared with a novel block copolymer. Ann. Thor. Surg. 1999; 67:689–696. 36. Bhujle R, Li J, Shastri P, Gaffke JN, Clift JE, Ye Y-W, Dollar ML, Ching P, Chao RY, Constantinescu A, Kulkarni PV, Cheng Q-C, Wians F, Jessen ME, Eberhart RC. Influence of cardiopulmonary bypass on platelet and neutrophil accumulations in internal organs. J. ASAIO 1997; 43:M739–744. ˚ , Wadenvik H. Nitric oxide in the 37. Mellgren K, Friberg LG, Mellgren G, Hedner T, Wennmalm A oxygenator sweep gas reduces platelet activation during experimental perfusion. Ann Thorac Surg 1996; 61:1194–1198. 38. Keh D, Gerlach M, Kurer I, Falke KJ, Gerlach H. Reduction in platelet trapping in membrane oxygenators by transmembraneous application of gaseous nitric oxide. Int. J. Artif. Organs 1996; 19:291–293. 39. Sly MK, Eberhart RC, Prager MD. Anti-platelet action of nitric oxide and selective phosphodiesterase inhibitors. Shock 1997; 8:115–118. 40. Kieffer N. Adhesive platelet glycoproteins and platelet function. In: (Editor) Horton MA, ed, CRC Press. 1996:139–169. ˚ , Wadenvik H. Ann Thor. Surg.() 1998; 65:1335. 41. Mellgren K, Mellgren G, Lundin S, Wennmalm A 42. Turitto VT, Baumgartner HR. Platelet interaction with subendothelium in flowing rabbit blood: effect of blood shear rate. Microvasc. Res. 1979; 17:38–54.
4 Clinical Properties and Healing Characteristics of e-PTFE Vascular Grafts and Their Effects on Long-Term Patency Glenn C. Hunter and Kenneth J. Woodside University of Texas Medical Branch Galveston, Texas, U.S.A.
Hana Holubec, Alex Westerband, and Charles W. Putnam University of Arizona Health Sciences Center Tucson, Arizona, U.S.A.
David A. Bull University of Utah Salt Lake City, U.S.A.
I. INTRODUCTION Approximately 400,000 coronary artery bypass and 150,000 femoral popliteal bypass grafting procedures are performed in the United States annually [1,2]. Saphenous vein is the preferred conduit for bypassing femoropopliteal occlusive and aneurysmal lesions and together with the internal mammary artery (IMA) is used for coronary artery bypass grafting. IMA grafts have higher patency rates (96 and 90% vs. 81 and 53% than saphenous vein as aortocoronary bypass grafts at 1 and 10 years, respectively. However, their supply is limited and bilateral harvesting of the IMAs may be associated with a higher incidence of sternal wound complications [2,3]. In a significant number of individuals, saphenous veins are unavailable or unsuitable for usage because they have been previously removed, are varicosed, or have been the site of thrombophlebitis. The increasing longevity of the population and complexity of operative vascular procedures being performed as well as the development of intimal hyperplasia and accelerated atherosclerosis in vein grafts [1,4,5] have increased the demand for an effective arterial substitute for both aortocoronary and distal limb bypass grafts. Few alternative materials exist for bypass grafting in patients who do not have a suitable saphenous vein. Expanded polytetrafluoroethylene (e-PTFE), Dacron, polyurethane, human umbilical vein, and, more recently, cryopreserved saphenous vein have all been used to bypass stenotic and aneurysmal lesions of the femoral, popliteal and tibial arteries [6–10]. e-PTFE is the most commonly used small diameter (ⱖ6 mm) arterial prosthesis and is often used preferentially by some surgeons for above-knee femoropopliteal bypass grafting, reserving the saphenous vein for later use [11]. Herein, we discuss the physical properties, 75
76
Hunter et al.
clinical use, mechanisms of healing and graft failure, and the methods currently employed to improve the long-term patency of e-PTFE grafts.
II. PHYSICAL CHARACTERISTICS OF e-PTFE First discovered by Plunkett in 1938, the basic e-PTFE molecule is comprised of a carbon atom – carbon atom (C—C) bond with four attached fluorine atoms [12]. When polymerized, the ePTFE molecule forms chains of repeating units with molecular weights ranging from 400,000 to 10 million Da. The negatively charged fluorine atoms form a protective sheath over the chain of carbon atoms, and are responsible for the stability and chemical inertness of the polymer and, by lowering surface energy, its resultant nonstick properties [13]. Polytetrafluoroethylene grafts consist of solid nodes of e-PTFE from which extend longitudinally oriented fibrils varying in length from 1 to 100 m depending on the extrusion process. The node–fibril structure of e-PTFE vascular grafts occupies only 15–20% of its total volume, the void space being filled with air (Fig. 1). e-PTFE has many qualities that make it suitable for use as a vascular conduit. In addition to being chemically inert, it is resistant to dilation, has good suture retention properties, is compatible with blood, can be implanted without preclotting, is readily incorporated into the surrounding tissues, and is not altered by repeated gas or steam sterilization [12]. Although there are now several commercially available e-PTFE grafts, Gore-Tex 30 m average internodal distance (IND) supported by an external wrap of e-PTFE and Impra with a similar internodal distance are the two most intensively studied and commonly used e-PTFE grafts. The grafts are manufactured with a wall thickness of either 0.4 (thin wall) or 0.6 m (standard wall).
Figure 1 Scanning electron micrograph demonstrating the node–fibril structure of an e-PTFE graft.
Characteristics of e-PTFE Vascular Grafts
77
III. CLINICAL USE OF e-PTFE GRAFTS Polytetrafluoroethylene grafts have many clinical applications, including repair of abdominal and chest wall defects, replacement of knee ligaments, and hernia repair, in addition to its use as a vascular conduit or patch material. This discussion will be limited to the use of e-PTFE as a conduit for bypass grafting in cardiovascular surgery. A. Femoropopliteal and Tibial Bypass Grafts Reversed or in situ saphenous vein is the conduit of choice for revascularization of ischemic limbs. Primary patency rates for femoropopliteal saphenous vein grafts range from 60 to 80% at 5 years [14,15]. It is evident from these data that a significant number of patients will undergo more than one revascularization procedure, thereby increasing the need for durable arterial substitutes. Because of lower patency rates in the majority of patients, the use of e-PTFE and other prosthetic or biological grafts is limited to patients who do not have usable saphenous vein or have a shortened life expectancy. The 5-year primary patency rates for above-knee e-PTFE femoral popliteal bypass grafts ranges from 38 to 61% [10,16]. Patency rates are significantly lower (12 to 59%) for below-knee grafts at 3 years [11,17]. The major cause of failure of these grafts is occlusion due to anastomotic intimal hyperplasia, most often at—but not confined
Figure 2 Peripheral angiogram demonstrating intimal hyperplasia (arrow) at the distal anastomosis of a below-knee femoropopliteal e-PTFE bypass graft.
78
Hunter et al.
to—the distal anastomosis (Fig. 2). Less frequently, fibrous stenotic lesions occur at proximal anastomoses needle puncture or at graftomy sites. Primary occlusion of e-PTFE grafts as a result of low cardiac output or hypercoagulability is uncommon and is usually manifest within the first 30 days after implantation. B. Dialysis Access Grafts Approximately 220,000 patients are currently undergoing chronic hemodialysis in the United States [18]. A Brescia-Cimino arteriovenous (A-V) fistula constructed at the wrist or Kaufman A-V fistula at the elbow, with patency rates of 81, 75, and 71% at 1, 2, and 3 years, respectively, is the preferred autogenous method of vascular access in patients requiring chronic hemodialysis [19]. Veins of marginal diameter, venous damage from intravenous infusion or drug use, and diabetic calcification of the radial artery are the most important factors that preclude the use of autogenous grafts and are usual causes of early graft failure. Late complications, including aneurysmal dilation, segmental stenoses, and thrombophlebitis, may further limit their use. Arteriovenous grafts constructed between the radial or brachial arteries and the cephalic or basilic veins using e-PTFE in a straight or looped configuration are the most commonly used prosthetic A-V fistula for chronic hemodialysis. Patency rates range from 30 to 70% at 1 year and are approximately 20% at 3 years [20–23]. The long-term patency of e-PTFE dialysis access grafts is punctuated by repeated episodes of thrombosis. There are four possible sites of stenosis that can contribute to the occlusion of dialysis grafts, including the graft–vein interface, midgraft sites, and the artery–graft interface, as well as central vein (e.g., subclavian or innominate vein) sites. Contributing factors include the inherent thrombogenicity of the graft, hypercoagulable states, and injury and thrombosis of outflow veins resulting from placement of dialysis cannulae or intravenous infusions. In addition, damage incurred at puncture sites contribute to the failure of these grafts (Fig. 3). Acute thrombosis of A-V grafts is usually associated with kinks, anastomotic
Figure 3 Histologic section of a e-PTFE dialysis graft showing granulation tissue ingrowth along a needle tract (arrow).
Characteristics of e-PTFE Vascular Grafts
79
stenosis, hypotension, and damage as a result of early cannulation. In contrast, late failure is usually the result of anastomotic stenosis due to intimal hyperplasia, aneurysmal dilatation, or thrombophlebitis. C. Aortofemoral Bypass Grafts Woven and knitted Dacron grafts are the most frequently used prosthetic material for aortic reconstructive procedures. Aortofemoral bypass grafts maintain function well, with patency rates ranging between 85 to 90% at 5 years [24,25]. However, knitted Dacron grafts have a tendency to dilate over time. In addition, the edges of woven Dacron grafts tend to fray, predisposing them to anastomotic aneurysms or even aneurysmal dilation of the graft itself. [26] Thrombectomy of an occluded Dacron graft limb is often difficult, as thrombus adheres to the velour of the inner surface. Several potential advantages of e-PTFE include that offers it does not require preclotting, it does not dilate significantly over time, it can readily be thrombectomized, and it may be somewhat more resistant to infection. There are, however, few studies comparing Dacron and e-PTFE. Cintora et al. and Friedman et al. have reported long-term patency rates of 95–97% for e-PTFE, which is comparable to that of Dacron (86–90%) [27,28]. Lord, in a study comparing Dacron and e-PTFE used for aortic replacement, found no difference in the performance of these two graft materials in patients observed for up to 24 months [29]. More recently, Prager et al. reported 91% primary and 97% secondary patency rates at 5 years [25]. The early standardwall e-PTFE bifurcated grafts were stiff, difficult to handle, and often complicated by significant intraoperative suture line bleeding. The new stretch e-PTFE bifurcation grafts are easier to suture with less needle hole suture line bleeding [30]. When compared to Dacron, e-PTFE bifurcated grafts have similar primary (89 vs. 91%) and secondary (97 vs. 97%) patency rates, but a slightly lower incidence of infection (0 vs. 3%) [25]. D. Axillofemoral and Femorofemoral Grafts Extraanatomic grafts are used to bypass occlusive and aneurysmal lesions of the abdominal aorta, and iliac vessels in patients believed to be poor operative candidates because of severe coexisting cardiac or pulmonary disease. This procedure is also used in patients who have undergone multiple operations or have prosthetic graft infection. Externally supported ringed e-PTFE grafts have been used for the construction of axillofemoral and femorofemoral bypass grafts. Primary patency rates of 47–71% at 3–5 years for axillofemoral grafts have recently been reported by Johnson et al. and Taylor et al. [31,32]. Because of the length, the axillofemoral portion of the graft is subject to kinking that results in repeated episodes of thrombosis and may require multiple thrombectomies to maintain longterm patency. Unrecognized, stenosis of the proximal subclavian artery, avulsion of the axillary artery graft anastomosis, anastomotic intimal hyperplasia, and progression of distal disease are frequent causes of failure. E. Carotid and Subclavian Grafts Polytetrafluoroethylene grafts are used to bypass lesions of the innominate, common and internal carotid arteries and also to bypass stenotic or aneurysmal lesions of the subclavian and axillary arteries. Although many surgeons profess strong preferences for the use of either autogenous or prosthetic grafts, there have been no prospective studies comparing these two types of grafts. One concern with the use of autogenous saphenous vein grafts to bypass carotid and proximal
80
Hunter et al.
subclavian lesions is the frequent size mismatch and a tendency to aneurysmal dilatation. Patency rates of 95% at 5 years have been reported for e-PTFE used to bypass lesions of vessels originating from the aortic arch [33]. F. Cardiac Uses of e-PTFE The current trend in management of infants with congenital heart defects is to undertake primary one-stage corrective repair. However, systemic-pulmonary artery shunts still continue to play a vital role in the management of infants with a single ventricle and various forms of tetralogy of fallot with pulmonary atresia [34–36]. Such shunts are used as a staged approach to repair underlying congenital defects and on rare occasions as permanent palliation. The shunts used include the direct Blalock–Taussig (BT) subclavian to pulmonary artery end-side anastomosis, which allows construction of the correct size shunt that will grow with the child to allow just enough pulmonary blood flow, and the modified Blalock–Taussig shunt, which consists of an interposition graft of e-PTFE between the subclavian and pulmonary arteries. The modified BT shunt is used almost exclusively at the present time because of its excellent patency, low infection rates, ease of implantation, absence of arm ischemia, and more predictable pulmonary blood flow. The lack of growth potential has become less important because of the increasing use of primary repair at an earlier age. The Waterston (ascending) and Potts (descending) aortic-pulmonary arterial shunts are of historical interest only. The size of the shunt is crucial—a 1-mm deviation from the optimal diameter may result in either too much or too little pulmonary blood flow and can be fatal. The more proximal the origin of the shunt, the smaller it needs to be. Generally accepted shunt diameters range from 4–6 mm when arising from the subclavian artery and 3–4 mm, depending on the infant’s weight, for shunts originating from the innominate artery [3 mm (2.0–2.4 kg), 3.5 mm (2.5–3.9 kg) and 4 mm (4–5 kg)]. A central shunt arising from the aorta is generally 3–3.5 mm in diameter, regardless of infant size. Opie et al., in a study using shunts of three different internal diameters (4, 5, and 6 mm) with follow-up ranging from 3 to 27 months, reported a patency rate of 89% (42 of 47 grafts) [34]. The patency of these grafts appeared to be related to graft diameter. Sixtyone percent of 4-mm, 53% of 5-mm and, 100% of 6-mm grafts were functioning during the follow-up period. The use of e-PTFE in acquired coronary artery disease has been limited to those patients where no autogenous saphenous veins are available. These cases generally are the second or third coronary artery reoperation, with poor target vessels measuring less than 1 mm in diameter [37]. G. Venocaval and Portal Vein Grafts The use of e-PTFE for venous replacement was first suggested by Soyer et al., who reported a patency rate of 87% for 25 e-PTFE grafts implanted into veins in dogs for intervals ranging from 2 weeks to 26 months [38]. More recently, e-PTFE has been used as a venous conduit to treat patients with obstruction of the superior and inferior cava and iliac veins as a result of trauma, thrombophlebitis, or intrinsic or extrinsic compression by neoplasm [39]. The successful use of e-PTFE for replacement of the major veins is limited by low venous distending pressures, slow flow, and the resulting increased tendency to thrombus deposition, which is the most common cause of failure. Externally supported ring grafts have been used to reduce the effects of low distending pressure. Warfarin, by increasing the thrombotic threshold, and construction of an adjunctive distal arteriovenous fistula to increase flow through the graft have been used to enhance patency of prosthetic venous grafts. Despite the advances in the use
Characteristics of e-PTFE Vascular Grafts
81
of e-PTFE for venous replacement, assessment of its efficacy is limited due to the small number of patients. Patency rates of 33 to 57% for bypass of benign lesions in the IVC have been reported [39]. Primary and secondary 5-year patency rates of 53 and 74%, respectively, have been reported for SVC bypass grafting procedures [40].
IV. HEALING CHARACTERISTICS OF e-PTFE A. Human Implants There have been few studies documenting the healing characteristics of e-PTFE implanted into humans. The systematic in vivo evaluation of functioning grafts has only recently been routinely undertaken with the introduction of high quality color duplex imaging. Specimens retrieved from patients with patent grafts demonstrate an outer fibrous capsule of varying thickness and a thick fibrin sheath lining the inner surface of the grafts. The intervening matrix of the graft is usually filled with a relatively acellular proteinaceous material. In 30 m e-PTFE grafts, multinuclear giant cells can be observed at the interface between the e-PTFE wrap and the adjacent host perigraft tissue (Fig. 4). Anastomotic changes are of two varieties: intimal hyperplasia (AIH), extending for approximately 1–2 cm from the adjacent artery across the anastomosis onto the e-PTFE graft, and consisting of pannus ingrowth thrombus invaded by granulation tissue. Anastomotic intimal thickening occurs most often at the distal anastomosis, but can be observed at the proximal anastomosis. Occasionally stenotic lesions occur at graftomy or arteriography puncture sites. Needle puncture sites in dialysis access grafts are characterized by damage to the inherent structure of the grafts at these sites. The marked cellular infiltration and collagen deposition along the needle tracks may act as a nidus for thrombus formation and ultimately contribute to occlusion of the graft [23,41].
Figure 4 Histologic section of a 30-m e-PTFE graft showing the accumulation of inflammatory cells and foreign body giant cells along the e-PTFE wrap with only a few cells infiltrating the internodal spaces.
82
Hunter et al.
In a clinical study comparing the thrombogenicity of composite wrapped e-PTFE grafts with 30- and 60-m IND, Kohler et al. found no difference in the uptake of 111In-labeled platelets between 30- and 60-m segments of the grafts at 1 and 3 months. Of the two 60 m segments available for histologic examination, capillary ingrowth was present, but rarely extended for more than half the distance of the wall of the grafts. No smooth muscle or endothelial cells were visible on the luminal surface [42]. In an analysis of 73 e-PTFE explants, Formichi et al. observed an external fibrous capsule of varying thickness that was absent or vestigial in the presence of infection. Cellular penetration was only evident on 19% of specimens examined. Scanning electron microscopy of the luminal surface revealed the presence of bacteria, leukocytes, lipids, cholesterol, adipocytes, and endothelial cells. The authors also observed kinks in regions of flexion and anastomosis. Once formed, these indentations persist and may contribute to the ultimate failure of the graft. It should be emphasized that the grafts evaluated in this study were mainly retrieved at the time of reoperation for graft failure. While such studies provide some evidence of the behavior of these grafts in humans, the major limitation of such studies is biased sampling [43]. In summary, the healing of e-PTFE grafts in man is quite variable as evidenced by a wide range of patency rates at 1 and 5 years. What can be stated, however, is that endothelialization of the graft is limited to within the 1–2 cm immediately adjacent to the anastomoses and does not occur to any significant extent in the mid-portion of the graft. The long-term patency of ePTFE grafts implanted in man, except for perhaps dialysis grafts where there is in addition inherent damage to the graft structure, is limited by anastomotic intimal hyperplasia. B. Experimental Implants The healing characteristics of e-PTFE grafts have been studied in a number of different animal species including rodents, dogs, sheep, pigs, and primates. In addition to variations in animal models, grafts have been implanted into different vascular beds, thereby making interpretation of the data difficult. The extrapolation of results from grafts implanted into the abdominal or thoracic aorta, for example, may not be applicable to the behavior of femoral or carotid grafts. Furthermore, the rapidity of healing varies between the different species, with very rapid healing in the pig, calf, and baboon; intermediate-rate healing in the dog; and slow healing in man [44–50]. In both mongrel dogs and baboons, the healing of small diameter (4-mm internal diameter) 30-m e-PTFE grafts is characterized by slow endothelial ingrowth across the anastomosis on the surface of the graft (Fig. 5) [47,50]. The extent of the anastomotic endothelial cell ingrowth appears to be dependent on the length of the graft. The mid-portion of the graft is usually devoid of endothelium and covered with a thin layer of fibrin with focal aggregations of thrombus and leukocytes. The outer capsule is of varying thickness with multinucleate giant cells accumulating on the outer surface of the graft immediately adjacent to the outer wrap. The graft wall is usually filled with proteinaceous material, fibrin, collagen, erythrocytes, and fibroblasts (Fig. 6). Antiplatelet agents are usually necessary to ensure patency of e-PTFE grafts in dogs. Selecting animals with a low thrombotic potential can further enhance patency. Graft failure is rapid in animals with a high thrombotic potential in the absence of antiplatelet agents. Thrombus instead of pannus is evident at the anastomosis and, when organized, results in the accumulation of acellular tissue at the anastomosis. C. Mechanisms of Failure of e-PTFE Grafts 1. Initiating Events The long-term patency of a prosthetic bypass graft is determined by the velocity of blood flow through the graft, the resistance of the arterial bed into which it is being implanted, compatibility
Characteristics of e-PTFE Vascular Grafts
83
Figure 5 (A) Histologic section of an e-PTFE graft showing pannus ingrowth at the graft–artery interface. (B) High power view of the pannus showing a rather acellular matrix covered by a layer of endothelial cells.
with cells and proteins in the blood, and hemodynamic changes at the proximal and distal anastomoses. Clinically, flow rates through e-PTFE grafts can be optimized by ensuring adequate inflow, either by an additional proximal bypass graft or transluminal angioplasty. Inserting a graft into a distal vessel with adequate outflow is essential if the graft is to remain patent and is the basic principle of vascular surgery. Once a graft has been successfully implanted, the long-term patency is determined by interactions at the blood–graft surface interface and by the development of AIH. Graft failure in humans is a biphasic phenomenon. The initial early failure (within 30 days) is usually due to technical defects or the inherent thrombogenicity of the host. The major reason for failure resulting in occlusion after the first 30 days is almost invariably AIH, usually at but not confined to the distal anastomosis. The etiology of AIH lesions is unknown. However, several contributing factors, such as increased intraluminal pressure, vessel wall ischemia, trauma, shear stress, hyperlipidemia, immunologic reactions, and compliance mismatch, have all been implicated in the development of anastomotic changes [51–58]. Irrespective of the mode of injury, endothelial injury is believed to be the initiating event in the development of AIH. Understanding the factors that influence host–graft compatibility and the mechanisms underlying the development of AIH are essential if patency rates of small diameter prosthetic
84
Hunter et al.
Figure 6 Histologic section of an e-PTFE vein-wrapped (30-m) graft segment showing cellular ingrowth into the graft through the interstices of the wrap. The remainder of the graft wall is filled with acellular proteinaceous material.
grafts are to approximate those of autogenous arterial and venous grafts. The biochemical and molecular events implicated in the development of AIH are complex and only a brief summary of these events is provided below. The reader is referred to a more detailed discussion of the cellular and molecular mechanisms involved in the development of AIH in recent reviews of the subject [52–54, 58]. The damage to the vessel wall incurred during construction of an anastomosis activates the clotting cascade resulting in the formation of mural thrombi. The interactions between platelets, fibrin, thrombin-activated macrophages, endothelial and smooth muscle cells at the site of injury result in the release of cytokines and other growth regulatory molecules. These molecules induce the proliferation and migration of endothelial and smooth muscle cells and the synthesis of extracellular matrix. These growth regulatory molecules implicated in the healing response at an anastomosis have many divergent effects: they may stimulate or inhibit cell proliferation and also act as chemoattracts depending on the local environment. Platelet-derived growth factor (PDGF), basic fibroblast growth factor (bFGF), heparin-binding epidermal-like growth factor (Hb-EGF), insulin-like growth factor-1 (IGF-1), interleukin-1 (IL-1), and transforming growth factor  (TGF-) are among several growth regulatory molecules that act in an autocrine or paracrine fashion to modulate cell proliferation. Several of these mitogens also involve chemotaxis of smooth muscle cells (PDGF, IGF-1, bFGF) and endothelial cells (bFGF and TGF) [52].
Characteristics of e-PTFE Vascular Grafts
85
The mitogenic influence of these molecules is counterbalanced by interferon ␥ (IF␥) TGF, and other cytokines which in high doses inhibit cellular proliferation [52,59,60]. Heparin produced by the endothelium at the site of injury may also exert an inhibitory effect on smooth muscle proliferation and alter the composition of the extracellular matrix [61]. The presence of a prosthetic graft and the resulting mismatch in compliance or hemodynamic flow disturbances may also influence the healing response at an anastomosis. Macrophages and foreign body giant cells are commonly observed at the interface between the perigraft tissue of the host and the graft in both humans and experimental animals, suggesting that the prosthetic material induces an inflammatory response. These activated macrophages release cytokines that may contribute to the development of anastomotic intimal thickening. Contrary to observations in some animal experiments of prosthetic vascular implants, such as those in the baboon model, transanastomotic cellular ingrowth merely extends 1 to 2 cm from the cut edge in humans. The progressive nature of AIH suggests a state of chronic healing involving all of the cellular events described, including cellular proliferation migration and matrix production [62–64]. These responses likely contribute to the ultimate failure of prosthetic vascular implants.
V. ADJUNCTIVE METHODS FOR IMPROVED PATENCY OF SMALL DIAMETER e-PTFE GRAFTS A number of approaches have been used to improve the long-term patency of e-PTFE grafts, including modifying the configuration of the distal anastomoses with vein patches and cuffs, varying the porosity, altering the characteristics of the flow surface, and endothelial cell seeding. A. Surface Properties 1. Vein Collars and Patches There are presently a number of techniques used to modify hemodynamic stress at the graft–arterial interface. These include the use of a vein patch or collar of different configurations as described by Linton, Taylor, Miller, and Tyrell and their coworkers (Fig. 7) and hoaded grafts such as the Distaflo graft [65–68]. Using the Linton patch technique, a segment of vein is sewn into the diseased artery as a patch [65]. The prosthetic graft is then sewn into a venotomy in the middle of the vein patch. Similarly, Taylor [66]completes the residual portion of the distal anastomosis with a vein patch. Taylor has reported patency rate of 71% for popliteal and 54% for infrapopliteal grafts at 5 years. The Miller cuff technique utilizes a segment of vein sutured to the circumference of arteriotomy [67]. The prosthetic graft is then sewn into the venous cuff. In a retrospective study, Jakobsen et al. reported primary patency rates of 62.5 and 50% at 1 and 2 years, respectively, using a Miller cuff. [69] Data from the European prospective study showed a significant benefit using cuffed anastomoses to the below-knee popliteal artery (80 vs. 65% and 52 vs. 29%) at 1 and 2 years, respectively) [70]. Based on the experimental studies of Harris and How demonstrating an increase in intimal hyperplasia in areas of low shear stress at anastomoses, distally widened e-PTFE grafts were constructed [71]. Panneton et al. [72], reporting for the North American Prospective Multicenter Trial comparing Distaflo e-PTFE to standard wall e-PTFE, reported short-term patency rates of 93 and 67% for Distaflo and 92 and 79% for standard e-PTFE at 6 and 12 months, respectively. A few comments regarding the use of a vein cuff or collar to modify compliance mismatch appear warranted. First, it is extremely difficult to measure compliance at anastomoses in vivo
86
Hunter et al.
Figure 7 Currently used vein patch and cuff techniques.
using presently available techniques. Second, and probably more important, changes in compliance at the anastomosis occur once a graft is implanted. Unless infection is present, a dense fibrous capsule is almost invariably present at both arterial to prosthetic and autogenous anastomoses. To date, no method exists which truly assess anastomotic compliance beyond 30 days. Whether the benefits of a collar or patch are mechanical or merely delay the effects of anastomotic intimal hyperplasia by enlarging the outflow remains to be determined. 2. Endothelialization of e-PTFE Grafts An intact endothelial cell lining is essential for maintaining the fluidity of the blood and plays a major role in hemostasis and thrombosis. The antithrombogenic nature of endothelium is maintained by a balance between antithrombotic and prothrombotic factors. The effects of procoagulants, including tissue factor, factor VIII, thrombospondin, plasminogen activator inhibitors, and collagen, are balanced by anticoagulant factors such as heparin-like glycosaminoglycans, prostacyclin, plasminogen activators and thrombomodulin [66]. The ability of the endothelial lining to ensure a nonthrombogenic flow surface has application to prosthetic grafts. Theoretically, the application of endothelial cells to the luminal surface of a prosthetic graft should result in a nonthrombogenic surface, limit AIH, and enhance the long-term patency of such grafts [73,74]. The highly electronegative surface charge of e-PTFE, which is responsible for its relatively nonthrombogenic surface, is also a barrier to adhesion of large numbers of seeded endothelial cells. There are three potential endogenous sources of endothelial cells to line prosthetic grafts. First, endothelial cells originating from the divided ends of an artery at an anastomosis may migrate onto the graft [64]. Second, microvascular endothelial cells migrating through the inter-
Characteristics of e-PTFE Vascular Grafts
87
stices of the graft may contribute to endothelialization of e-PTFE grafts [44,50]. Finally, Shi et al. [75] have shown that progenitor cells present in the circulation may differentiate into endothelial cells and contribute to the endothelialization of prosthetic grafts. Endothelial cells harvested from exogenous sources (vein, artery, omentum, adipose tissue) have been seeded onto the flow surface of dialysis access and femoropopliteal prosthetic grafts. Cell retention remains a problem despite the use of fibronectin or bFGF. Most procedures used for exogenous seeding remain cumbersome and have therefore not received widespread application [76–80]. 3. Anastomotic Ingrowth Endothelial cell ingrowth from the artery adjacent to an anastomosis is a characteristic finding in both experimental and human implants. Why this ingrowth is limited to within 1–2 cm of the anastomosis in humans and some experimental animals and whether it can be modulated to cover a greater area of the graft surface are presently unknown. 4. Transmural Ingrowth Commercially available e-PTFE grafts have a node–fibril distance of 20 to 30 m and permit very little transmural cellular ingrowth in either experimental animals or humans. The application of an external e-PTFE wrap to Gore-Tex grafts, even though porous, may further compromise the ultimate porosity by overlapping with the nodes and fibrils of the primary graft structure [81]. Although the node–fibril structure of e-PTFE appears to provide a suitable matrix for cellular ingrowth, the optimal internodal diameter that would permit cellular ingrowth is not fully established. While cellular ingrowth and endothelialization of the luminal surface of ePTFE grafts occurs readily in experimental animals, healing is much slower and less complete in humans. In other tissue implants (e.g., sponge foams with continuing channels) the optimal porosity for cellular ingrowth ranges between 60 and 80 m. Porosity ⬎120 m is associated with reduced ingrowth, presumably because of the smaller surface area available for cellular adhesion and locomotion. The relationship between porosity and cellular ingrowth in e-PTFE grafts used to replace segments of portal vein and vena cava was first explored by Soyer et al. [38]. They found that cellular ingrowth was absent in grafts with porosity of 0.5–2.5 m in size. Campbell et al. reported higher patency rates (88 vs. 53%) and less neointimal thickening in 4 m internal diameter grafts with an average IND ⱕ 22 m compared to grafts with a porosity of ⱖ 34 m [82]. Recent experimental studies by Clowes et al. [50], Kogel et al. [83,84], and Holubec et al. [44] have shown the transmural tissue ingrowth and the rate of healing of small diameter ePTFE grafts can be significantly altered by slight modifications in graft porosity (from 30 to 60 m IND). Clowes et al. found that endothelialization of 4-mm internal diameter 30-m ePTFE grafts implanted into the iliac arteries of baboons originated predominantly from the arterial anastomosis and extended onto the flow surface of the grafts to cover 60% of 6 to 9cm lengths at 12 months. In contrast, the healing of 60-m e-PTFE grafts occurred predominantly from capillary ingrowth from the surrounding perigraft tissue [50]. Once they reach the flow surface of the grafts, these microvessels occur at distances ranging from 100 to 500 m, undergo phenotypic changes, and form an endothelial coverage of the luminal surface. Endothelial cell outgrowth from this source is rapid and appears to be complete by 14 days. Endothelial cell proliferation, as determined by 3H-thymidine labeling, is markedly increased at these early time intervals and declines at 1 and 3 months, but does not reach the background levels observed in the adjacent artery. Furthermore, they noted that the anastomotic intimal thickening typically found in the 30-m grafts was absent in the 60-m grafts. In a similar study in dogs, Kogel et al. demonstrated complete endothelialization in 44% of 90 m grafts and 50% of 60 m grafts
88
Hunter et al.
at 12 months. There was some degree of pannus ingrowth limited to a distance within 1–2 cm of each anastomosis. In contrast, the neointima and endothelial lining of polyurethane grafts remained incomplete and restricted to the anastomotic region with only a few islets of endothelial cells in the mid-third of the prosthesis in two of the 11 grafts [83,84]. Although there is convincing evidence of the importance of porosity in determining the healing of e-PTFE grafts, the ideal porosity that would permit cellular migration, adhesion, and ingrowth and yet maintain the structural integrity of e-PTFE grafts is not clearly defined. It is evident that there is almost no transmural tissue ingrowth in 10 m grafts (Fig. 8), and only limited ingrowth in 30-m e-PTFE grafts implanted into canine femoral or carotid arteries (Fig. 9). The available data suggest that the optimal IND necessary for cellular adhesion and migration appears to range between 60 and 90 m. While cellular ingrowth and endothelialization will occur more readily in grafts with larger pore sizes (ⱖ90 m), cellular adhesion and retention may be more tenuous. Golden et al. have clearly demonstrated intimal damage and endothelial cell loss in 60% of 90-m grafts implanted in baboons at 3 months [85]. One of the remaining questions is whether transmural ingrowth can be enhanced beyond that originating from perigraft tissue. Providing another potential source of endothelial cells (i.e., vein or omental wrap) or the application of growth factors such as vascular endothelial or fibroblast growth factor could increase transmural ingrowth and the rate of healing.
Figure 8 Histologic section of a 10-m e-PTFE subcutaneous implant in a rat showing the nearly complete absence of cells within the structure of the graft.
Characteristics of e-PTFE Vascular Grafts
89
Figure 9 Histologic section of a 30-m e-PTFE-wrapped subcutaneous implant showing the absence of cellular ingrowth.
Because of the difficulties inherent in endothelial cell seeding, we and other investigators have postulated that the application of a source of endothelial cells’ on the outer surface of e-PTFE grafts would permit transluminal cellular ingrowth via the interstices of the prosthesis utilizing the cells’ own adhesive properties. In order to test this hypothesis, we developed a technique in which the e-PTFE graft was inserted into a sleeve of external jugular vein and anastomosed end-to-side to the carotid arteries of greyhound dogs (Fig. 10). A contralateral unwrapped graft served as control. e-PTFE grafts with 4-mm internal diameter of 30 and 60 m IND were evaluated. Minimal ingrowth was observed in 30- m grafts, despite the wrap. Transluminal tissue ingrowth was greatest in 60-m grafts, and the rate of ingrowth was increased approximately twofold from control grafts by the application of a vein wrap (Figs. 11 and 12). All the 60- m grafts implanted in greyhound dogs were endothelialized by 35 days by microvessel endothelial cells migrating from the perigraft tissues through the interstices of the grafts (Table 1). In contrast, the luminal surface of nonwrapped grafts was filled with proteinaceous material (Fig. 13). Table 1 Healing Characteristics of 40 and 60 m Grafts at 5 Weeks 40 m Internodal distance N Patency (%) Endothelialization (%) Microvessels (mm2) Cellular ingrowth * p⬍0.05
60 m
Control
Vein wrap
Control
Vein wrap
2 100 16.9⫾7.8 1.2⫾0.5 2.0⫾1.4
2 100 33.1⫾3.5* 6.8⫾2.1* 4.0⫾1.4*
2 100 20.4⫾4.3 5.6⫾5.8 3.0⫾1.4
2 100 64.3⫾12.5* 16.4⫾6.5* 4.5⫾0.7*
90
Hunter et al.
Figure 10 The vein wrap technique.
Figure 11 Histologic sections of 30-m grafts implanted in the carotid artery of greyhound dogs with (A) and without (B) an external vein wrap as a source of endothelial cells. Despite the increase in the number of cells on the external surface of the vein-wrapped graft (B), there was no significant difference in the number of cells in the interstices of the graft wall.
Characteristics of e-PTFE Vascular Grafts
91
Figure 12 Histologic section of a 60-m e-PTFE carotid interposition graft. Note the greater number of cells infiltrating the vein-wrapped graft (A) than the unwrapped graft (B).
Figure 13 Histologic section of the luminal surface of a 60-m e-PTFE graft showing acellular proteinaceous material without a recognizable endothelial layer in the unwrapped graft (A) with a cellular infiltrate and endothelial lining of the vein-wrapped graft (B).
92
Hunter et al.
Hazama et al., using an omental wrap as a potential source of microvascular endothelial cells with 4-mm internal diameter 60 m e-PTFE grafts, observed no differences in wrapped and unwrapped grafts at 4 weeks. However, three of five wrapped grafts were patent, compared to one of five unwrapped ones, at 12 weeks. There was also a significant difference in thrombinfree surface (86.4 vs. 62.4%) and capillary density between the two techniques at 12 weeks. In contrast, 60- and 90-m grafts showed considerable cellular and microvessel ingrowth. Deformation of the nodal–fibril structure by a variety of surgical instruments may have further inhibited cell ingrowth. The time course studies of transmural ingrowth in 60-m grafts harvested at 3, 5, 7, 14, 21, and 35 days demonstrated that tissue ingrowth occurred from at least two directions: from the arterial anastomosis and through the interstices of the graft. Endothelial cell ingrowth was first visible at approximately 7 days. In grafts wrapped with an external vein wrap, graft endothelialization was both more rapid and more extensive. The endothelial cells in the interstices of the grafts were tubular and extended perpendicularly to reach the lumen, where these cells formed a monolayer oriented in the direction of the blood flow (Figs. 14 and 15). Basic FGF was detected by immunohistochemistry during the early time periods (within 14 days). In order to identify those growth regulatory molecules that may influence endothelial cell-migration, we studied the migration of bovine pulmonary artery endothelial cells in a 48- well micro-chemotaxis chamber. Growth factors, including epidermal growth factors, transforming growth factors, basic fibroblast growth factor followed by polysulfate plasma, fibronectin, fibrinogen, granulocyte macrophage colony stimulating factor (G-CSF), heparin, adenosine, and magnesium sulfate stimulus, were the most potent. Platelet-derived growth factor BB and platelet-activating factor inhibited cell migration [86]. Because of its electronegative charge, bonding growth factors to flow surface of e-PTFE grafts remains a difficult undertaking. In preliminary studies, using endothelial cell growth factor in varying concentrations, we were able to demonstrate enhanced endothelial cell ingrowth; however, the bonding of the growth factor to the graft surface was quite unpredictable. Greisler et al. using acidic FGF impregnated into fibrinogen and inspissated under pressure to 60-m grafts, were able to demonstrate improved cellular migration [87]. The porosity of e-PTFE grafts can be altered adversely in a number of ways. Although the external wrap applied to 30-m Gore-Tex graft has an identical pore size as the graft, the effective porosity is considerably reduced by the overlapping of the pores and fibrils of the wrap with those of the graft. The porosity of e-PTFE grafts may further be perturbed by handling with
Figure 14 Histologic section of a 60-m vein-wrapped e-PTFE graft showing a microvessel.
Characteristics of e-PTFE Vascular Grafts
93
Figure 15 Histologic section of 60-m vein-wrapped e-PTFE graft after 35 days in situ demonstrating endothelialization of the luminal surface.
surgical instruments. Tissue ingrowth into the compressed graft segments was considerably reduced (2.5- to 7-fold) in subcutaneous implants in rats (Figs. 16–19). In a series of implants in greyhound dogs, we observed a significant relationship between in vivo porosity and anastomotic changes. It would therefore seem prudent to preserve the in vivo porosity to minimize damage to material and as a result impair healing (Table 2). There is evidence in the literature suggesting that porosity influences both cellular ingrowth and thrombogenicity and that both variables should be considered when selecting the optimally porous graft. Using e-PTFE grafts with porosities ranging from 10 to 80 m suspended in the venous circulation of dogs, we demonstrated an inverse relationship between surface thrombus formation and porosity. Ten-micrometer e-PTFE grafts accumulated significantly less thrombus
Figure 16 Scanning electron micrographs of a e-PTFE graft demonstrating compression of the nodefibril structure of the graft when cut by scissors.
94
Hunter et al.
Figure 17 Scanning electron micrograph of a e-PTFE graft showing the effects of a hemostat on the node-fibril structure of the graft.
Figure 18 Histologic section of a 30-m subcutaneous e-PTFE implant demonstrating the effects of compression of the internodal structure on cellular ingrowth. Note the uniform infiltration of cells into the nondeformed graft (A) compared with the architectural distortion and scant cellular infiltrate in the deformed graft (B).
Characteristics of e-PTFE Vascular Grafts
95
Figure 19 Influence of porosity on cellular ingrowth. The significant decrease in cell density in the compressed graft segments is readily seen.
than 80-m grafts (Table 3 and Figure 20). However, our studies were criticized because the graft segments were suspended in the venous system rather than the arterial circulation. Kogel et al. observed greater amounts of thrombus deposition on 90-m grafts compared to 60-m [83,84]. In order to test the hypothesis that alterations in porosity influence the thrombogenicity of the flow surface, a series of experiments were undertaken using grafts with dual porosity (60 m pore size on the luminal surface, and 20 m on the outer surface). In another series of implants, the porosity was reversed (20 m on the luminal, 60 m on the outer surface). Implanted as carotid interposition grafts with a porosity of 20 m on the luminal surface favored endothelial ingrowth from the arterial anastomosis and microvessel ingrowth in the perigraft tissue. When the dual porosity was reversed (60 m luminal surface, 20 m perigraft), however, thrombus formed on the flow surface became more evident with no ingrowth from the perigraft tissue (Figs. 21–23). 5. Fall Out Endothelialization There is increasing evidence that the deposition of stem cells from the bloodstream may be an additional source of endothelial cells. Stump et al. [88] were among the first to demonstrate that endothelial cells adhered to Dacron hubs suspended in the thoracic aorta of pigs. Subsequently, Sbarbati et al. [89] demonstrated an increase from baseline in the number of endothelial cells in the circulation of patients undergoing coronary angioplasty. Frazer et al. [90] and Rafii et al. [91] have demonstrated endothelial cells on the surface of ventricular assist devices. Shi
Table 2 Changes in IND After Implantation and Effects on in Anastomotic Healing Initial porosity (m) 60 60, 40 30
Reduction in porosity (%)
Anastomotic lesions
68.0 68.0, 72.0 75.0
No lesion IH Pl
IH = intimal hyperplasia; PI = Pannus in growth.
96
Hunter et al.
Figure 20 Histologic section of a 80-m e-PTFE carotid interposition graft demonstrating thrombus accumulation on the graft surface.
et al. [75] demonstrated endothelial cells on the flow surface of grafts excised from patients as well as grafts implanted in the descending thoracic aorta and vena cava of dogs. Possible sources of fallout endothelialization include mature endothelial cells detached from vessel walls by surgical manipulation, rheologic or pathologic factors, endothelial progenitor cells in the circulation, and progenitor cells derived from bone marrow. Mature endothelial cells have been identified in the blood of patients with cardiovascular disease and cancer and following
Table 3 Effect of Porosity on Thrombogenicity (Non-Heparinized Dogs) Porosity (m) 10 30 60 80
Thrombus weight (% increase) 72 ⫾ 30 108 ⫾ 13 189 ⫾ 31 816 ⫾ 148
Characteristics of e-PTFE Vascular Grafts
97
Figure 21 Histologic section of a dual porosity e-PTFE graft (20 m luminal, 60 m outer IND) showing a uniform cellular infiltrate throughout the wall of the graft and the absence of thrombus on the luminal surface.
Figure 22 Histologic section of a dual porosity e-PTFE graft (20 m luminal, 60 m outer surface) carotid interposition graft harvested after 35 days in situ that demonstrates cellular deposition in the midportion of the graft that is suggestive of endothelial fallout from the circulation.
98
Hunter et al.
Figure 23 Histologic section of an anastomosis of a dual porosity graft showing minimal deformity of the graft structure at the anastomosis with uniform pannus extending on to the graft.
cardiac catheterization. Sbarbati et al. [89] have demonstrated an increase from baseline in the number of endothelial cells in patients undergoing cardiac catheterization. It is presently unknown to what extent fallout endothelialization contributes to the endothelialization of grafts implanted into humans. It appears that endothelialization of porous graft occurs as a result of anastomotic ingrowth, fallout endothelialization, and transmural ingrowth from perigraft tissue. B. Bonding of Antithrombogenic Agents to Graft Surfaces 1. Heparin Heparin is the agent most fully studied. Two forms of heparin bonding are currently being—investigated—ionic bonding, which provides an excellent initial concentration of heparin but is rapidly eluted from the surface, and covalent bonding, which prevents rapid heparin loss but is less effective in preventing platelet adhesion [92]. Ionic bonding has been successfully used to bind heparin to polyurethane grafts, but similar attempts using e-PTFE have to date been uniformly unsuccessful. Recently, Iwai reported that thermal crosslinking results in a high concentration of heparin on the graft surface [93]. The ideal amount of heparin applied to the surface of prosthetic grafts should be sufficient to prevent thrombus formation yet permit tissue ingrowth. In a comparison of heparin-bonded Dacron and e-PTFE grafts used for below-knee femoropopliteal bypass, Devine et al. reported 3-year patency rates of 55% for Dacron and 42% for e-PTFE [94]. 2. Carbon Coating Carbon-impregnated e-PTFE grafts have been advocated in an attempt to decrease the thrombogenicity of such grafts. In a comparison between carbon coated and standard wall e-PTFE, Bacourt found no statistical difference in primary and secondary patency rates (45 and 53% for carboncoated and 36 and 47% for standard wall e-PTFE) [95]. Similarly, Groegler et al. found no
Characteristics of e-PTFE Vascular Grafts
99
differences in patency between carbon-impregnated and standard e-PTFE grafts at 36 months [96]. 3. Pharmacological Agents The inhibition of platelet function has been the prime focus of attempts to modulate intimal thickening and graft thrombosis and as a consequence enhance graft patency. The rationale for the use of antiplatelet and antithrombotic agents to enhance graft patency is the ability of these agents to inhibit prostaglandin synthesis and as a consequence prevent platelet activation, adhesion and aggregation. Aspirin (ASA) and dipyridamole, two of the earlier drugs used, exerted their influence by inhibiting cyclooxygenase and phosphodiesterase, respectively [97,98]. Inhibition of cyclooxygenase inhibits the syntheses of both prostacyclin (PGI2) and thromboxane A2 (TxA2) from arachidonic acid. Hunter et al. have suggested that the balance between PGI2 and TxA2 may be a determinant of graft patency [48]. More selective inhibition of thromboxane synthetase has not significantly improved graft patency rates or reduced intimal hyperplasia when compared to ASA [99]. It is not surprising that neither selective nor nonselective inhibition of prostaglandin synthesis ensures inhibition of platelet function. The accumulation of prostaglandin and leukotriene intermediates through alternative pathways may facilitate platelet aggregation. Although experimental use of these agents has demonstrated improved patency of prosthetic grafts, their clinical use has not been uniformly successful. The most frequently used agents are sodium warfarin, ASA, and, more recently, ticlopidine and clopidogrel. [97, 98, 100–102] Recent studies examining the use of warfarin sodium in patients undergoing e-PTFE femoropopliteal bypass grafting have produced conflicting results [112–113]. Green, in a prospective study of patients with femoral popliteal bypass grafts, reported improved patency in above-knee but not below-knee grafts in patients treated with ASA alone or in combination with dipyridamole [98]. In experimental studies in our laboratory, we found that ibuprofen, a cyclooxygenase inhibitor, improved early—but not late—patency of small diameter e-PTFE grafts implanted into the carotid arteries of sheep [48]. Landymore was able to demonstrate inhibition of intimal hyperplasia in cholesterol-fed animals receiving cod liver oil, [103,104] and Casali et al. demonstrated improved 6-month patency in small diameter prosthetic grafts of 86% compared to 44% of control grafts in animals treated with omega-3 fatty acids [105]. Fish oil inhibits platelet aggregation and synthesis of PDGF, resulting in decreased intimal hyperplasia in experimental animals. A number of other pharmacologic and biologic agents have been and are currently under evaluation for the prevention of AIH. Corticosteroids, angiopeptin, heparin alone and in combination with ACE inhibitors, platelet glycoprotein IIb/IIIa inhibitors, and even gene therapy have shown experimental promise but have not proved uniformly beneficial in humans. [106–111] With the increasing number of patients undergoing coronary artery and peripheral bypass grafting procedures, the need for a durable prosthetic substitute for autogenous conduits seems evident. Modification of e-PTFE ultrastructure and porosity, in addition to improved pharmacological manipulation of the interface between the blood and the prosthetic graft surface, may ultimately improve long-term patency. REFERENCES
1. Motwani JG, Topol EJ. Aortocoronary saphenous vein graft disease: pathogenesis, predisposition, and prevention. Circulation 1998; 97(9):916–931.
100
Hunter et al.
2. Grondin CM, Campeau L, Lesperance J, Enjalbert M, Bourassa MG. Comparison of late changes in internal mammary artery and saphenous vein grafts in two consecutive series of patients 10 years after operation. Circulation 1998; 70(3 Pt 2):I208–I212. 3. Loop FD, Lytle BW, Cosgrove DM, Stewart RW, Goormastic M, Williams GW. Influence of the internal-mammary-artery graft on 10-year survival and other cardiac events. N Engl J Med 1998; 314(1):1–6. 4. Hutchins GM. Pathological changes in aortocoronary bypass grafts. Annu Rev Med 1998; 31: 289–301. 5. Szilagyi DE, Elliott JP, Smith RF, Reddy DJ, McPharlin MA. thirty-year survey of the reconstructive surgical treatment of aortoiliac occlusive disease. J Vasc Surg 1998; 3(3):421–436. 6. Abbott WM, Green RM, Matsumoto T, Wheeler JR, Miller N, Veith FJ. Prosthetic above-knee femoropopliteal bypass grafting: results of a multicenter randomized prospective trial. Above-knee femoropopliteal study group. J Vasc Surg 1998; 25(1):19–28. 7. Dardik H, Wengerter K, Qin F, Pangilinan A, Silvestri F, Wolodiger F. Comparative decades of experience with glutaraldehyde-tanned human umbilical cord vein graft for lower limb revascularization: an analysis of 1275 cases. J Vasc Surg 1998; 35(1):64–71. 8. Dereume JP, van Romphey A, Vincent G, Engelmann E. Femoropopliteal bypass with a compliant, composite polyurethane/Dacron graft: short-term results of a multicentre trial. Cardiovasc Surg 1998; 1(5):499–503. 9. El-Massry S, Saad E, Sauvage LR, Zammit M, Smith JC, Davis CC. Femoropopliteal bypass with externally supported knitted Dacron grafts: a follow-up of 200 grafts for one to twelve years. J Vasc Surg 1998; 19(3):487–494. 10. Johnson WC, Lee KK. A comparative evaluation of polytetrafluoroethylene, umbilical vein, and saphenous vein bypass grafts for femoral-popliteal above-knee revascularization: a prospective randomized Department of Veterans Affairs cooperative study. J Vasc Surg 1998; 32(2):268–277. 11. Harris M, O’Brien-Irr M, Ricotta JJ. Long-term assessment of cryopreserved vein bypass grafting success. J Vasc Surg 1998; 33(3):528–532. 12. Plunkett RJ. Tetrafluoroethylene polymers, Assigned to Kinetic Chemicals, Inc., Wilmington, February 4, 1941. 13. Plunkett RJ. The history of polytetrafluoroethylene: discovery and development. Polym Preprints 1998; 27(1):485–487. 14. Taylor LM, Porter JM. Clinical and anatomic considerations for surgery in femoropopliteal disease and the results of surgery. Circulation 1998; 83(2 Suppl):I63–I69. 15. Allen BT, Reilly JM, Rubin BG, Thompson RW, Anderson CB, Flye MW. Femoropopliteal bypass for claudication: vein vs. e-PTFE. Ann Vasc Surg 1998; 10(2):178–185. 16. Woratyla SP, Darling RC, Chang BB, Paty PS, Kreienberg PB, Leather RP. The performance of femoropopliteal bypasses using polytetrafluoroethylene above the knee versus autogenous vein below the knee. Am J Surg 1998; 174(2):169–172. 17. Quinones-Baldrich WJ, Prego AA, Ucelay-Gomez R, Freischlag JA, Ahn SS, Baker JD. Long-term results of infrainguinal revascularization with polytetrafluoroethylene: a ten-year experience. J Vasc Surg 1998; 16(2):209–217. 18. . USRDS 2001 Annual Data ReportBethesda JD, National Institute of Diabetes and Digestive Kidney Diseases, National Institutes of Health (NIH), DHHS. 2001. 19. Pherwani AD, Reid JA, Connolly JK. Patency and survival of primary arteriovenous fistulae. In: ed. Henry ML-. Chicago: W.L. Gore & Associates, and Precept Press, 2001:47–53. 20. Hakaim AG, Scott TE. Durability of early prosthetic dialysis graft cannulation: results of a prospective, nonrandomized clinical trial. J Vasc Surg 1998; 25(6):1002–1005; discussion 5–6. 21. Enzler MA, Rajmon M, Lachat M, Largiader F. Long-term function of vascular access for hemodialysis. Clin Transplant 1998; 10(6 Pt 1):511–515. 22. Palder SB, Kirkman RL, Whittemore AD, Hakim RM, Lazarus JM, Tilney NL. Vascular access for hemodialysis. Patency rates and results of revision. Ann Surg 1998; 202(2):235–239. 23. Anderson JM, Bennert KW, Johnson JM. The pathology and healing responses of expanded polytetrafluoroethylene vascular access grafts. In: ed. Wilson SE. Vascular Access Surgery. ed. 2nd: Year Book Medical Publishers, 1988:213–31.
Characteristics of e-PTFE Vascular Grafts
101
24. Malone JM, Moore WS, Goldstone J. The natural history of bilateral aortofemoral bypass grafts for ischemia of the lower extremities. Arch Surg 1998; 110(11):1300–1306. 25. Prager M, Polterauer P, Bohmig HJ, Wagner M, Fugl A, Kretschmer G. Collagen versus gelatincoated Dacron versus stretch polytetrafluoroethylene in abdominal aortic bifurcation graft surgery: results of a seven-year prospective, randomized multicenter trial. Surgery 1998; 130(3):408–414. 26. Downs AR, Guzman M, Formichi M, Courbier R, Jausseran JM, Branchereau M. Etiology of prosthetic anastomotic false aneurysms: pathologic and structural evaluation in 26 cases. Can J Surg 1998; 34(1):53–58. 27. Friedman SG, Lazzaro RS, Spier LN, Moccio M, Tortolani AJ. A prospective randomized comparison of Dacron and polytetrafluoroethylene aortic bifurcation grafts. Surgery 1998; 117(1):7–10. 28. Cintora M, Pearce DE, Cannon JA. A clinical survey of aortobifemoral bypass using two inherently different graft types. Ann Surg 1998; 208(5):625–630. 29. Lord RS, Nash PA, Raj BT, Stary DL, Graham AR, Hill DA. Prospective randomized trial of polytetrafluoroethylene and Dacron aortic prosthesis. I. Perioperative results. Ann Vasc Surg 1998; 2(3):248–254. 30. Chiesa M, Melissano G, Castellano R, Frigerio S. Extensible expanded polytetrafluoroethylene vascular grafts for aortoiliac and aortofemoral reconstruction. Cardiovasc Surg 1998; 8(7):538–544. 31. Taylor LM, Moneta GL, McConnell M, Yeager RA, Edwards JM, Porter JM. Axillofemoral grafting with externally supported polytetrafluoroethylene. Arch Surg 1998; 129(6):588–595. 32. Johnson WC, Lee KK. Comparative evaluation of externally supported Dacron and polytetrafluoroethylene prosthetic bypasses for femorofemoral and axillofemoral arterial reconstructions. Veterans Affairs Cooperative Study 噛141. J Vasc Surg 1998; 30(6):1077–1083. 33. Law MM, Colburn MD, Moore WS, Quinones-Baldrich WJ, Machleder HI, Gelabert HA. Carotidsubclavian bypass for brachiocephalic occlusive disease. Choice of conduit and long-term followup. Stroke 1998; 26(9):1565–1571. 34. Opie JC, Traverse M, Hayden RI, Ho CY, Culham JA, Ashmore PG. Experience with polytetrafluoroethylene grafts in children with cyanotic congenital heart disease. Ann Thorac Surg 1998; 41(2): 164–168. 35. Alexi-Meskishvili V, Ovroutski S, Ewert P, Dahnert I, Berger F, Lange PE. Optimal conduit size for extracardiac Fontan operation. Eur J Cardiothorac Surg 1998; 18(6):690–695. 36. Gold JP, Violaris M, Engle MA, Klein AA, Ehlers KH, Lang SJ. A five-year clinical experience with 112 Blalock–Taussig shunts. J Card Surg 1998; 8(1):9–17. 37. Weyand M, Kerber S, Schmid C, Rolf N, Scheld HH. Coronary artery bypass grafting with an expanded polytetrafluoroethylene graft. Ann Thorac Surg 1998; 67(5):1240–1245. 38. Soyer M, Lempinen M, Cooper P, Norton L, Eiseman B. A new venous prosthesis. Surgery 1998; 72(6):864–872. 39. Sarkar M, Eilber FR, Gelabert HA, Quinones-Baldrich WJ. Prosthetic replacement of the inferior vena cava for malignancy. J Vasc Surg 1998; 28(1):75–81; discussion 82–83. 40. Alimi YS, Gloviczki M, Vrtiska TJ, Pairolero PC, Canton LG, Bower TC. Reconstruction of the superior vena cava: benefits of postoperative surveillance and secondary endovascular interventions. J Vasc Surg 1998; 27(2):287–301. 41. Guidoin M, Chakfe N, Maurel S, How T, Batt M, Marois M. Expanded polytetrafluoroethylene arterial prostheses in humans: histopathological study of 298 surgically excised grafts. Biomaterials 1998; 14(9):678–693. 42. Kohler TR, Stratton JR, Kirkman TR, Johansen KH, Zierler BK, Clowes AW. Conventional versus high-porosity polytetrafluoroethylene grafts: clinical evaluation. Surgery 1998; 112(5):901–907. 43. Formichi MJ, Guidoin RG, Jausseran JM, Awad JA, Johnston KW, King MW. Expanded e-PTFE prostheses as arterial substitutes in humans: late pathological findings in 73 excised grafts. Ann Vasc Surg 1998; 2(1):14–27. 44. Holubec M, Hunter GC, Chvapil M, Chvapil TA, Bernhard VM, Misiorowski RL. The relationship between e-PTFE graft ultrastructure and cellular ingrowth: the influence of an autologous jugular vein wrap. In: Kambic HE , Yokobori ATJ, eds’. Philadelphia: American Society for Testing and Materials, 1994:53–64.
102
Hunter et al.
45. Holubec M, Hunter GC, Putnam CW, Bull DA, Rappaport WD, Chvapil M. Effect of surgical manipulation of polytetrafluoroethylene grafts on microstructural properties and healing characteristics. Am J Surg 1998; 164(5):512–516. 46. Kelly GL, Eiseman B. Development of a new vascular prosthetic: lessons learned. Arch Surg 1998; 117(10):1367–1370. 47. Kusaba M, Fischer CR, Matulewski TJ, Matsumoto T. Experimental study of the influence of porosity on development of neointima in Gore-Tex grafts: a method to increase long-term patency rate. Am Surg 1998; 47(8):347–354. 48. Hunter GC, Carson SN. Arterial wall thromboxane: dominance after surgery predisposes to thrombosis. J Vasc Surg 1998; 1(2):314–319. 49. Parsson M, Jundzill W, Hallberg E, Thorne J, Norgren L. Acute thrombogenicity and 4 weeks healing properties of a new stretch-ePTFE graft. Eur J Vasc Surg 1998; 7(1):63–70. 50. Clowes AW, Kirkman TR, Reidy MA. Mechanisms of arterial graft healing. Rapid transmural capillary ingrowth provides a source of intimal endothelium and smooth muscle in porous e-PTFE prostheses. Am J Pathol 1998; 123(2):220–230. 51. Abbott WM, Megerman M, Hasson JE, L’Italien G, Warnock DF. Effect of compliance mismatch on vascular graft patency. J Vasc Surg 1998; 5(2):376–382. 52. Davies MG, Hagen PO. Pathobiology of intimal hyperplasia. Br J Surg 1998; 81(9):1254–1269. 53. How TV, Rowe CS, Gilling-Smith GL, Harris PL. Interposition vein cuff anastomosis alters wall shear stress distribution in the recipient artery. J Vasc Surg 1998; 31(5):1008–1017. 54. Chervu M, Moore WS. An overview of intimal hyperplasia. Surg Gynecol Obstet 1998; 171(5): 433–447. 55. Deng M, Marois Y, How T, Merhi Y, King M, Guidoin R. Luminal surface concentration of lipoprotein (LDL) and its effect on the wall uptake of cholesterol by canine carotid arteries. J Vasc Surg 1998; 21(1):135–145. 56. Dangas M, Fuster V. Management of restenosis after coronary intervention. Am Heart J 1998; 132(2 Pt 1):428–436. 57. O’Malley MK. Intimal hyperplasia. Eur J Vasc Surg 1998; 6(4):343–345. 58. Mattsson EJ, Kohler TR, Vergel SM, Clowes AW. Increased blood flow induces regression of intimal hyperplasia. Arterioscler Thromb Vasc Biol 1998; 17(10):2245–2249. 59. Massague M, Polyak K. Mammalian antiproliferative signals and their targets. Curr Opin Genet Dev 1998; 5(1):91–96. 60. Hosang M. Recombinant interferon-gamma inhibits the mitogenic effect of platelet-derived growth factor at a level distal to the growth factor receptor. J Cell Physiol 1998; 134(3):396–404. 61. Taylor M, Fletcher JP, Ao PY. Inhibition of fibro-intimal hyperplasia in a polytetrafluoroethylene vascular graft with standard heparin and low molecular weight heparin. Aust N Z J Surg 1998; 66(11):764–776. 62. Clowes AW, Reidy MA, Clowes MM. Kinetics of cellular proliferation after arterial injury. I. Smooth muscle growth in the absence of endothelium. Lab Invest 1998; 49(3):327–333. 63. Tsuchida M, Cameron BL, Marcus CS, Wilson SE. Modified polytetrafluoroethylene: indium111–labeled platelet deposition on carbon-lined and high-porosity polytetrafluoroethylene grafts. J Vasc Surg 1998; 16(4):643–650. 64. Kuwano M, Hashizume M, Yang Y, Kholoussy AM, Matsumoto T. Patterns of pannus growth of the expanded polytetrafluoroethylene vascular graft with special attention to the intimal hyperplasia formation. Am Surg 1998; 52(12):663–666. 65. Batson RC, Sottiurai VS, Craighead CC. Linton patch angioplasty. An adjunct to distal bypass with polytetrafluoroethylene grafts. Ann Surg 1998; 199(6):684–693. 66. Taylor RS, Loh M, McFarland RJ, Cox M, Chester JF. Improved technique for polytetrafluoroethylene bypass grafting: long-term results using anastomotic vein patches. Br J Surg 1998; 79(4): 348–354. 67. Miller JH, Foreman RK, Ferguson M, Faris I. Interposition vein cuff for anastomosis of prosthesis to small artery. Aust NZ J Surg 1998; 54(3):283–285.
Characteristics of e-PTFE Vascular Grafts
103
68. Wolfe M, Tyrell M. Venous patches, collars, and boots improve the patency rates of polytetrafluoroethylene grafts. Adv Vasc Surg 1998; 3:134–143. 69. Jakobsen HL, Baekgaard M, Christoffersen JK. Below-knee popliteal and distal bypass with ePTFE and vein cuff. Eur J Vasc Endovasc Surg 1998; 15(4):327–330. 70. Stonebridge PA, Prescott RJ, Ruckley CV. Randomized trial comparing infrainguinal polytetrafluoroethylene bypass grafting with and without vein interposition cuff at the distal anastomosis. The Joint Vascular Research Group. J Vasc Surg 1998; 26(4):543–550. 71. Harris PL, How TV. Haemodynamics of cuffed arterial anastomoses. Critical Ischaemia 1998; 9(1): 20–26. 72. Panneton JM. Randomized prospective evaluation of the distally widened (Distaflo姟) e-PTFE graft. 27th Global Vascular Endovascular Issues Techniques Horizons Symposium 2000:IA2.1. 73. Levi M, ten Cate H, van der Poll T. Endothelium: interface between coagulation and inflammation. Crit Care Med 1998; 30(5 Suppl):S220–S224. 74. Krishnaswamy M, Kelley J, Yerra L, Smith JK, Chi DS. Human endothelium as a source of multifunctional cytokines: molecular regulation and possible role in human disease. J Interferon Cytokine Res 1998; 19(2):91–104. 75. Shi M, Rafii S, Wu MH, Wijelath ES, Yu M, Ishida A. Evidence for circulating bone marrowderived endothelial cells. Blood 1998; 92(2):362–367. 76. Jarrell BE, Williams SK, Stokes M, Hubbard FA, Carabasi RA, Koolpe M. Use of freshly isolated capillary endothelial cells for the immediate establishment of a monolayer on a vascular graft at surgery. Surgery 1998; 100(2):392–399. 77. Visser MJ, van Bockel JH, van Muijen GN, van Hinsbergh VW. Cells derived from omental fat tissue and used for seeding vascular prostheses are not endothelial in origin. A study on the origin of epitheloid cells derived from omentum. J Vasc Surg 1998; 13(3):373–381. 78. Herring M, Baughman S, Glover J, Kesler K, Jesseph J, Campbell J. Endothelial seeding of Dacron and polytetrafluoroethylene grafts: the cellular events of healing. Surgery 1998; 96(4):745–755. 79. Graham LM, Burkel WE, Ford JW, Vinter DW, Kahn RH, Stanley JC. Expanded polytetrafluoroethylene vascular prostheses seeded with enzymatically derived and cultured canine endothelial cells. Surgery 1998; 91(5):550–559. 80. Bearn PE, Seddon AM, McCollum CN, Marston A. Mesothelial seeding of knitted Dacron. Br J Surg 1998; 80(5):587–591. 81. Boyce B. Physical characteristics of expanded polytetrafluoroethylene grafts. In: ed. Stanley JC. Biologic and Synthetic Vascular Prostheses. New York: Grune & Stratton, 1982:553–561. 82. Campbell CD, Goldfarb M, Roe R. A small arterial substitute: expanded microporous polytetrafluoroethylene: patency versus porosity. Ann Surg 1998; 182(2):138–143. 83. Kogel M, Vollmar JF, Stenzenberg M. Morphologic analysis of artificial blood conduits in the short-term carotid artery test. Vasc Surg 1998; 24:297–306. 84. Kogel M, Amselgruber W, Frosch D, Mohr W, Cyba-Altunbay S. New techniques of analyzing the healing process of artificial vascular grafts, transmural vascularization, and endothelialization. Res Exp Med (Berl) 1998; 189(1):61–68. 85. Golden MA, Hanson SR, Kirkman TR, Schneider PA, Clowes AW. Healing of polytetrafluoroethylene arterial grafts is influenced by graft porosity. J Vasc Surg 1998; 11(6):838–845. 86. Bull DA, Seftor EA, Hendrix MJ, Larson DF, Hunter GC, Putnam CW. Putative vascular endothelial cell chemotactic factors: comparison in a standardized migration assay. J Surg Res 1998; 55(5): 473–479. 87. Greisler HP, Cziperle DJ, Kim DU, Garfield JD, Petsikas M, Murchan PM. Enhanced endothelialization of expanded polytetrafluoroethylene grafts by fibroblast growth factor type 1 pretreatment. Surgery 1998; 112(2):244–255. 88. Stump MM, Jordan GLJ, DeBakey ME, Halpert B. Endothelium grown from circulating blood on isolated intravascular Dacron hub. Am J Pathol 1998; 43:361–367. 89. Sbarbati M, de Boer M, Marzilli M, Scarlattini M, Rossi G, van Mourik JA. Immunologic detection of endothelial cells in human whole blood. Blood 1998; 77(4):764–769.
104
Hunter et al.
90. Frazier OH, Baldwin RT, Eskin SG, Duncan JM. Immunochemical identification of human endothelial cells on the lining of a ventricular assist device. Tex Heart Inst J 1998; 20(2):78–82. 91. Rafii M, Shapiro F, Rimarachin J, Nachman RL, Ferris M, Weksler B. Isolation and characterization of human bone marrow microvascular endothelial cells: hematopoietic progenitor cell adhesion. Blood 1998; 84(1):10–19. 92. Mohamed MS, Mukherjee M, Kakkar VV. Thrombogenicity of heparin and non-heparin bound arterial prostheses: an in vitro evaluation. J R Coll Surg Edinb 1998; 43(3):155–157. 93. Iwai Y. Development of a thermal cross-linking heparinization method and its application to small caliber vascular prostheses. Asaio J 1998; 42(5):M693–M697. 94. Devine M, Hons B, McCollum C. Heparin-bonded Dacron or polytetrafluoroethylene for femoropopliteal bypass grafting: a multicenter trial. J Vasc Surg 1998; 33(3):533–539. 95. Bacourt F. Prospective randomized study of carbon-impregnated polytetrafluoroethylene grafts for below-knee popliteal and distal bypass: results at 2 years. The Association Universitaire de Recherche en Chirurgie. Ann Vasc Surg 1998; 11(6):596–603. 96. Groegler FM, Kapfer M, Meichelbock W. Crural prosthetic revascularization: randomized, prospective, multicentric comparison of standard and carbon impregnated ePTFE grafts. 27th Global Vascular Endovascular Issues Techniques Horizons Symposium 2000:I2.1–I2.3. 97. Harker LA, Bernstein EF, Dilley RB, Scala TE, Sise MJ, Hye RJ. Failure of aspirin plus dipyridamole to prevent restenosis after carotid endarterectomy. Ann Intern Med 1998; 116(9):731–736. 98. Green RM, Roedersheimer LR, DeWeese JA. Effects of aspirin and dipyridamole on expanded polytetrafluoroethylene graft patency. Surgery 1998; 92(6):1016–1026. 99. Graham LM, Brothers TE, Darvishian M, Harrell KA, Vincent CK, Burkel WE. Effects of thromboxane synthetase inhibition on patency and anastomotic hyperplasia of vascular grafts. J Surg Res 1998; 46(6):611–615. 100. Windus DW, Santoro SA, Atkinson M, Royal HD. Effects of antiplatelet drugs on dialysis-associated platelet deposition in polytetrafluoroethylene grafts. Am J Kidney Dis 1998; 29(4):560–564. 101. Jackson MR, Johnson WC, Williford WO, Valentine RJ, Clagett GP. The effect of anticoagulation therapy and graft selection on the ischemic consequences of femoropopliteal bypass graft occlusion: results from a multicenter randomized clinical trial. J Vasc Surg 1998; 35(2):292–298. 102. Waksman M, Ajani AE, Pinnow M, Cheneau E, Leborgne L, Dieble R. Twelve versus six months of clopidogrel to reduce major cardiac events in patients undergoing gamma-radiation therapy for in-stent restenosis: Washington Radiation for In-Stent Restenosis Trial (WRIST) 12 versus WRIST PLUS. Circulation 1998; 106(7):776–778. 103. Landymore RW, Kinley CE, Cooper JH, MacAulay M, Sheridan B, Cameron C. Cod-liver oil in the prevention of intimal hyperplasia in autogenous vein grafts used for arterial bypass. J Thorac Cardiovasc Surg 1998; 89(3):351–357. 104. Landymore RW, MacAulay MA, Cooper JH, Sheridan BL. Effects of cod-liver oil on intimal hyperplasia in vein grafts used for arterial bypass. Can J Surg 1998; 29(2):129–131. 105. Casali RE, Hale JA, LeNarz M, Faas F, Morris MD. Improved graft patency associated with altered platelet function induced by marine fatty acids in dogs. J Surg Res 1998; 40(1):6–12. 106. Colburn MD, Moore WS, Gelabert HA, Quinones-Baldrich WJ. Dose responsive suppression of myointimal hyperplasia by dexamethasone. J Vasc Surg 1998; 15(3):510–518. 107. Hoepp LM, Elbadawi M, Cohn M, Dachelet R, Peterson C, DeWeese JA. Steroids and immunosuppression. Effect on anastomotic intimal hyperplasia in femoral arterial Dacron bypass grafts. Arch Surg 1998; 114(3):273–276. 108. Emanuelsson M, Beatt KJ, Bagger JP, Balcon M, Heikkila J, Piessens J. Long-term effects of angiopeptin treatment in coronary angioplasty. Reduction of clinical events but not angiographic restenosis. European Angiopeptin Study Group. Circulation 1998; 91(6):1689–1696. 109. Clowes AW, Clowes MM, Vergel SC, Muller RK, Powell JS, Hefti M. Heparin and cilazapril together inhibit injury-induced intimal hyperplasia. Hypertension 1998; 18(4 Suppl):II65–II69. 110. Moliterno DJ, Yakubov SJ, DiBattiste PM, Herrmann HC, Stone GW, Macaya M. Outcomes at 6 months for the direct comparison of tirofiban and abciximab during percutaneous coronary revascularisation with stent placement: the TARGET follow-up study. Lancet 1998; 360:355–360.
Characteristics of e-PTFE Vascular Grafts
105
111. Ohno M, Gordon D, San H, Pompili VJ, Imperiale MJ, Nabel GJ. Gene therapy for vascular smooth muscle cell proliferation after arterial injury. Science 1998; 265(5173):781–784. 112. Jackson MR, Johnson WC, Williford WO, Valentine RJ, Clagett GP. The effect of anticoagulation therapy and graft selection on the ischmic consequences of femoropopliteal bypass graft occlusion: results from a multicenter randomized clinical trial. J Vasc Surg; 35:292–298. 113. Dutch Bypass Oral Anticoagulants or Aspirin (BOA) Study Group. Efficacy of oral anticoagulants compared with aspirin after infrainguinal bypass surgery: a randomized trial. Lancet 1998; 355: 346–351.
5 Biomaterials and Stent Technology Kytai T. Nguyen, Shih-Horng Su, Meital Zilberman, Pedram Bohluli, Peter Frenkel, Liping Tang, and Robert Eberhart University of Texas Southwestern Medical Center at Dallas Dallas, Texas, U.S.A.
Richard Timmons University of Texas at Arlington Arlington, Texas, U.S.A.
I. INTRODUCTION The term stent, originally defined in dental reconstruction practice at the turn of the 20th century, now commonly denotes a short metal or plastic filamentous tube that is inserted into the lumen of a vessel (artery, bile duct, ureter, trachea, etc.), especially to keep a formerly blocked passageway open [1]. More specifically, a stent is used to provide radial support to the vascular wall while healing takes place after a minimally invasive or open procedure. As early as 1912, endovascular support devices were placed in the canine aorta to maintain luminal patency under various conditions. Vascular stents similar in design and function to current devices were introduced in the 1960s and have since received increasingly heavy attention. The use of coronary stents in interventional procedures increased from 5 to 10% in 1994 to over 80% in 2000. It is estimated that more than 1.3 million percutaneous transluminal coronary angioplasty (PTCA) procedures were performed worldwide in 1999 [2–4] and stents were employed in about 70% of these cases [5,6]. Current indications for stenting are given in Table 1. A minimally invasive procedure, PTCA opens blocked or severely narrowed coronary arteries, thereby allowing blood to better circulate to the heart muscle. A thin flexible guidewire is used as a rail to direct a balloon-tipped catheter to the vessel target site with the aid of fluoroscopic imaging. The balloon is straddled over the plaque/thrombus obstruction, then inflated to both compress this material against the wall and irreversibly deform the artery. The balloon is then deflated and withdrawn so that blood can once again flow freely through the expanded vessel. Coronary stents were developed to overcome some of the limitations of conventional balloon angioplasty: dissection, acute wall recoil, and restenosis. A stent is premounted on a second balloon catheter and delivered and positioned over the expansion site, with the aid of radiopaque markers on the balloon catheter. Inflation of the balloon results in the expansion of the stent, in apposition to the newly expanded lumenal surface of the coronary artery [6–10]. In recent practice the initial balloon expansion and stent expansion steps are often combined in a single deployment maneuver. 107
108
Nguyen et al.
Table 1 Indications and Contraindications for Coronary Arterial Stenting Proven indications for stenting
Unproven indications for stenting
Contraindications
Acute MI [10] Chronic total occlusion [7,12] Improvement of outcomes of PTCA [10] Native coronaries 3.0 mm in diameter [7,12] Restenotic lesions after nonstent interventions [7,12] Acute or threatened vessel closure [7,10] Bifurcational lesion [7,12] Long lesion and diffuse disease [7,12] Multivessel stenting (omit vs. CABG) [7,12] Ostial lesion [7,12] Small vessel [7,12] Coronary artery spasm absent significant stenosis Patients in whom antiplatelet and anticoagulation therapy is contraindicated [7,12] Patients judged to have a lesion which prevents complete inflation of an angioplasty balloon [7,12] Unprotected left main coronary artery
Stent-grafts are related devices for remodeling of aneurysmal peripheral vessels. They generally consist of wire-link structures covered with a fabric of woven or knitted polyester or other fabric. The metal/fabric assembly is compressed within a vascular sheath, then delivered via catheter to a peripheral vessel site where the sheath is removed and the stent expands to a preset diameter. Balloon-tipped catheters are employed in some procedures to aid expansion. The expanded stent-graft defines the conduit and, when properly deployed, excludes the dissected, aneurysmal vascular tissue from flowing blood and the pressure which might otherwise expand the lesion. The stent-graft attaches to the arterial wall by barbed hooks or other means. A neointima is created in association with the fabric, as is the case for vascular prostheses. A. Stent Characteristics and Materials Stents are classified as balloon-expandable or self-expandable (Fig. 1). Balloon-expandable stents are delivered by catheters with rigid or compliant balloons incorporated with the catheter tip. Rigid balloons expand to a fixed diameter over a range of balloon pressures, while the diameter of compliant balloons depends on the applied pressure. The expanded stent dimensions depend both on balloon characteristics and the applied pressure. Another important feature of the stent-balloon combination is the seating of the compressed (furled) stent in furrows fashioned in the deflated balloon. The furrows anchor the stent during maneuver to the deployment site so as not to dislodge the furled stent from the catheter. Self-expanding metal stents employ a shape memory alloy, principally Nitinol威. The shape memory property allows stent expansion upon mild temperature change (cold and hot saline techniques [3,11]) or release from a compresing sheath. Nitinol has good elasticity, an advantage for stent practice. Recently shape memory polymers have been introduced which may offer the possibility of self-expanding polymeric stents. Ideally, stents should have the following features: flexibility during maneuver; expandability, conformability, and minimal balloon overhang during deployment; radial strength, low compliance, low recoil, and no slippage following expansion; low vessel crossing profile at branches,
Biomaterials and Stent Technology
109
Figure 1 Endovascular stent categorization.
low lumenal profile, smooth surfaces, and biocompatibility; optimal radio-opacity during placement and beyond [2,3,6,7]. Stents must be available in a variety of lengths and diameters to accommodate differing applications and anatomies [2,3,6,7,12]. Balloon-expandable metal stent designs (slotted-tubular or helical coil) provide mechanical characteristics ranging from excellent radial strength and low flexibility to higher compliance and flexibility. Stents with coil designs have excellent flexibility and trackability in tortuous vessels, but relatively poor radial strength. Slotted-tubular stents, on the other hand, have good radial strength and relatively poor flexibility. Stents are fabricated from materials ranging from uncoated metal to polymer-clad metal to polymer, especially biodegradable polymer [5]. Stainless steel and tantalum are commonly used materials for both balloon- and self-expandable designs [11]. The inherent radio-opacity of tantalum is a benefit in comparison with stainless steel, however, it is a more brittle material. A listing of recent metal, polymer-coated metal, and polymeric stents is given in Table 2. Discussion of polymeric stents is reserved for a later section.
II. METAL STENTS A. Balloon-Expandable The operation of balloon-expandable stents is based on the plastic deformation of metal beyond the elastic limit. Examples of several stent designs are given in Fig. 2. Balloon inflation enlarges precut openings in slotted-tube stents to diamond or other large-aperture patterns. These stents offer two advantages over coiled wire stents: higher expansion ratios and larger spaces between the wire components. Both features limit potential sites for intimal growth and provide
110
Nguyen et al.
Table 2 Metal, Polymer-Coated, and Polymeric Endovascular Stents Stent Metal stents Gianturco-Roubin
Gianturco Z HARTS
Palmaz BEIS
Strecker
Wallstent Wiktor
Composition Stainless steel suture wire wrapped around a cylinder Zig-zag tempered stainless steel wire Nitinol
Staggered rows of rectangular stainless steel slots Tantalum or stainless steel wire mesh
Braided spring-steel filaments Tantalum wire wrapped around PTCA balloon
Polymer-coated metal stents Polymer-coated Wallstent Biogold stent Boston Scientific Cleveland Clinic/ Mayo/Thoraxcenter bioabsorbable stent Fibrin sleeve stent
Polymer-coated metal PLGA, PCL, and PHBVcoated tantalum wire
Guidant/Cook Cordis (Johnson & Johnson)
Polymer-coated metal Polymer-coated metal
Polymer stents Duke University PET PGA PLLA:PCL
Resin UT Southwestern Collagen I
Fibrin-coated Wiktor stent
Open mesh Poly-L-lactic acid PET mesh Poly-glycolic acid 75 : 25 poly-L-lactic acid: poly-caprolactone Cyanoacrylate PLLA Charged type I collagen formed into a solid tube
Properties
References
Approved for “bail outs” Flexible with low expansion ratio Strong radial force
[115–117]
Removable by heating to thermoelastic transition temperature (theoretical) Applicable to a wide range of cases by varying size; strong radial force Radio-opaque with tantalum; effective in treating femoropopliteal atherosclerosis Strong radial force in vessels; ⬎ 3 mm Flexible, designed for tortuous vessels
[119]
Strong radial force; less thrombogenic Coating delivers drug Flexible, moderate strength; bioresorbable coating for drug delivery Flexible, strong, less thrombogenic Coating delivers drug Delivers rapamycin
[124]
Bioresorbable, moderate strength Rigid, nonresorbable Flexible, bioresorbable Extremely porous, swellable, bioresorbable, flexible Strong, nonthrombogenic Strong, flexible Flexible, bioresorbable, biocompatible; can bind some agents directly
[130,131]
[118]
[118]
[117,118,120]
[5] [121–123]
[125] [126]
[127] [128] [129]
[132–135] [136] [137]
[138] [11,24,29,30] [139]
Biomaterials and Stent Technology
111
Figure 2 Various metal stent designs.
lower hydraulic resistance at branch crossover points. However, slotted-tube stents have lower flexibility in general than wire pattern design stents [2,3,7,12]. Uncoated metal stents initiate an inflammatory process, and a neointimal proliferation that may be more robust than the neointimal thickening after a nonstented PTCA procedure. This so-called in-stent restenosis has unique histological properties that may relate to a local tissue reaction on the presence of foreign material. Low dose, beta-emitting radioisotopes loaded into the bulk metal and drug-impregnated surface coatings of metal stents have been explored to limit in-stent restenosis. Recently the enthusiasm for isotope-loaded stents has diminished, owing in part to collateral radiation-induced damage; this category will not be further treated. Three types of drug delivery systems were developed to prevent restenosis of metal stents. One method employs an intraluminal catheter, deployed in the vicinity of the stent, to deliver a drug bolus. Alternatively, the stent is coated with a drug-impregnated polymer; the drug is eluted over periods varying from days to several weeks. Stent-based drug delivery enables direct, sustained, more precise delivery to the target tissue than is possible with parenteral delivery. Smaller doses and higher concentrations of the drug can be administered [13,14].
III. POLYMERIC STENTS A. Biostable Polymer Stents Several reports of biostable and bioresorbable polymeric stents have recently appeared [15–21]. The rationale for the nondegradable polymer stent is more convenient drug loading and a larger reservoir than is possible with a metal stent. Improved biocompatibility is often claimed, although studies supporting this view in the context of stenting are lacking. Nonresorbable polymers being investigated for stent use include polyethylene terephthalate, polyurethane urea, and polydimethyl siloxane. B. Bioresorbable Stents The rationale for the bioresorbable stent is support of the vascular wall only during vessel healing, with transfer of the mechanical load to the wall as stent mass and strength decrease over time; longer-term delivery of drug and/or gene therapy to the vessel wall from an internal reservoir; and no need for a second surgery to remove the device. The principal objections to a bioresorbable stent are an increased inflammatory and proliferative response and fragmentation and embolization of the polymer as degradation proceeds. Recent developments give reason to believe that these objections can be overcome.
112
Nguyen et al.
Bioresorbable polymers under investigation include aliphatic polyesters, polyorthoesters, and polyanhydrides. Bioresorbable, linear, multiblock copolymers with shape memory capability have recently been introduced [22]. Controlled incremental heating of this thermoplastic material has been used to shrink sutures, making graded tissue approximations feasible in minimally invasive surgery applications. The same concept is also valuable for bioresorbable stent applications. Following balloon expansion, heat (⬃5 ⬚C temperature change) applied to shape memory elements in the stent could reinforce designs that might not otherwise have sufficient recoil resistance; alternatively, the copolymer might permit stent expansion without balloon pressurization. Poly-L-lactic acid (PLLA), poly-D,L-lactic acid (PDLA), poly- ε-caprolactone (PCL), and polyglycolic acid (PGA), all aliphatic polyesters, are the most frequently used materials for bioresorbable stents [15–17]. PLLA and PDLA have high tensile strength, permitting robust mechanical design but requiring long degradation times. PGA and PCL have less strength, but faster degradation rates. Useful combinations of these materials (copolymers and blends) can be made to improve flexibility. These materials degrade principally by simple hydrolysis of the ester bond in the polymer backbone. Partial chain scission degrades the polymer to 10–40 m particles, capable of being phagocytized and metabolized to carbon dioxide and water, which are of course fully resorbed. The degradation time is a function of the chemical structure of the polymer and its molecular weight. In typical formulations, PGA degrades over a time period of 6 to 12 months, while PLLA degrades over several years (Table 3). The long-term behavior of biodegradable polymers in blood vessels has not been well established. Van der Giessen et al. [23], testing strips of five different biodegradable polymers (PGA/PLA, PCL, polyhydroxybutyrate valerate, polyorthoester, and polyethyleneoxide/ polybutylene terephthalate) found extensive inflammatory responses within the coronary arterial wall. The observed tissue responses might be due the parent polymer compound, additives to the polymer, intermediate biodegradation products, implant geometry, or combinations thereof. On the other hand, the authors noted the implants were cleaned but not sterilized; therefore, bacterial or nonbacterial contamination might also have accounted for the inflammatory response. We have also observed a similar inflammatory response to sterilized PLLA stents implanted in the porcine femoral artery [24]. This may have been due to the original polymer formulation, which was not intended for medical applications and contained an epoxide functionality. Recently, in our laboratory, purified formulations have induced a much less intense inflammatory response, as determined both by 1- to 4-week implantations of PLLA fibers in the rat aortic wall, and by 2-week PLLA stent implantation in the pig femoral artery. A thin (50–100 m) intima, with
Table 3 Characteristics of Bioresorbable Polymers Used in Endovascular Stents Polymer PGA PLLA PDLA PCL
Melting point (°C)
Glass transition temperature (°C)
Modulus (Gpa)
Degradation time (months)
225–230 173–178 Amorphous 58–63
35–40 60–65 55–60 (⫺65) to (⫺60)
7.0 2.7 1.9 0.4
6–12 ⬎24 12–16 24–48
PGA, poly(glycolic acid); PLLA, poly(L-lactic acid); PDLA, poly(D,L-lactic acid); PCL, poly(-caprolactone). Source: Adapted from “Synthetic Biodegradable Polymers as Medical Devices”, MPB Archive, March 1998).
Biomaterials and Stent Technology
113
almost no inflammatory cells, formed around the fiber. While this is only preliminary data, it suggests that control of proliferative inflammatory processes and containment of PLLA fragments upon breakup of the fiber is possible. Similar observations were made in a study of polylactide copolymer, PLLA/PDLA (PLA96), stents in a rabbit abdominal aorta model found stent degradation proceeded with minimal tissue response over 24 months, with suitable encapsulation of polymer fragments in a thin neointima, leading the authors to suggest PLA96 as a promising stent core material [25]. Several early designs of expandable bioabsorbable stents have been developed as alternatives for metallic stents [15–17]. These include bioresorbable stents from Duke University, [26] Tianjin/Beijing University, [27] Kyoto University, [20] Igaki/Tamai Corp., [28] and the University of Texas at Arlington/Southwestern Medical School (UTA/UTSW) [24]. The first biodegradable stent was developed and investigated at Duke University. This PLLA stent was based on a slotted polymer fiber design and was reported to withstand up to 1000 mmHg compression pressure; in vivo studies demonstrated minimal thrombosis and inflammatory responses and moderate neointimal growth. The Tianjin/Beijing stent, made of PDLA/PCL copolymer with an inner heparin layer, was deployed with a balloon catheter, employing heating and pressurization. This stent produced mild neointimal proliferation in swine carotid artery models at 2 months. The Kyoto University PGA coil stent exhibited thrombus deposition in canine implant studies, but no subacute closure. The Igaki/Tamai stent, a bioresorbable PLLA zig-zag coil thought to be derived from the Kyoto design, was studied in the first clinical report of a bioresorbable stent in the human coronary artery. This stent also required a combination of heating and pressurization for expansion. The preliminary (6-month) results suggest this stent is safe and effective for human use. Long-term studies are anticipated. C. The Multiple-Lobe Stent The UTA/UTSW PLLA stent employs a linear, continuous coil array principle, by which multiple furled lobes (four in the present design) convert to a single large lobe upon balloon expansion (Fig. 3) [11,29]. Melt-extruded PLLA fibers (drawn 6:1) with a diameter of 0.14 mm and ultimate stress of 350 Ⳳ 40 MPa are woven continuously around a four-mandrel array (one central, three peripheral) into a four-lobe configuration. Three longitudinal fibers are interwoven and glued to the coil for mechanical support. After fabrication, a conventional angioplasty balloon catheter is inserted in the central lobe, and the stent can be deployed at the target site. The structure of the fully expanded stent is that of a helical coil with three longitudinal reinforcing fibers. The initial and final diameters of stents are adjustable by various combinations in sizes of central and peripheral rod mandrels. Stents with furled diameters ranging from 1.6 to 2.4 mm were fully expanded by 3 atm pressure to 2.3 to 4.7 mm; the corresponding expansion ratios ranged from 1.4 to 1.9. Collapse pressure under radial compression was adequately high, ranging from 0.4 to 2.4 atm, depending on fiber ply and other design parameters. Preliminary results from various in vitro and in vivo studies of this expandable bioresorbable stent suggested that the design principles and fabrication technique were sufficiently robust and versatile, thus justifying further investigation [11,24,29,30]. In 1- and 2-week implant studies in the porcine common femoral artery, stents did not migrate; however, the vessel lumen was markedly reduced at 2 weeks due to a strong inflammatory response. Stents, as other implants, elicit a range of host responses which interfere with the patency of the device [23,31]. Various approaches have been investigated to improve the biocompatibility of these stents, including surface plasma treatment and drug incorporation. Pulsed RF plasma treatment with di(ethylene glycol) vinyl ether significantly reduced platelet adhesion in a 1-hour porcine arteriovenous shunt model, to less than 10% of untreated control values [29]. Curcumin (diferoyl methane),
114
Nguyen et al.
Figure 3 Prototype multiple-lobe coil endovascular stent. (From Ref. 29.)
a nonsteroidal anti-inflammatory drug, was melt extruded with PLLA to generate curcuminloaded PLLA fiber (C-PLLA) [30]. Curcumin was uniformly distributed within the fibers, and a stable curcumin release rate for 36 days was observed. In vitro studies of mouse peritoneal phagocytes indicated significant reductions in adhesion of these cells to C-PLLA than to PLLA controls. These results suggested that C-PLLA has anti-inflammatory properties which may benefit the implants. Other nonsteroidal anti-inflammatory agents with sufficiently high melting points can be introduced into the polymer bulk in the same way. We have also investigated bulk loading of aqueous drugs that cannot tolerate melt extrusion, using a wet spinning technique that permits incorporation of large amounts of drug (up to 20 wt%) in the PLLA fiber [32]. In addition, hollow PLLA fiber spinning processes that allow loading of drugs, genetic vectors, or radioisotopes for cancer treatment into bioresorbable stents have been examined.
D. Design and Properties of a High Molecular Weight PLLA Stent Bioresorbable fibers were made from a relatively high molecular weight PLLA (RESOMER姟 L21, Boehringer Ingelheim, Germany) with an inherent viscosity of 3.6 dL/g in CHCl3 at 30 ⬚C. This polymer was melt spun at 190 ⬚C and drawn 8:1 at 80 ⬚C to create a fiber with the following tensile properties: ultimate strength, 974 MPa, Young’s modulus 4.9 Gpa, and maximal strain 0.88. This fiber combines high strength and modulus with good ductility and flexibility. Four-lobe expandable stents were prepared from these PLLA fibers, generally following the design of Fig. 4 [33]. Stents of 15-mm length, 3.0-mm final (dilated) diameter, and 1.8-mm predilated diameter were used. Both single and double fiber stent windings were investigated. Each dilated coil stent contained 12 loops, each of them bonded to three longitudinal support fibers, i.e., 36 binding points per each single fiber stent and 72 binding points per each double fiber stent. The initial radial compression strength of the dilated form of both types of stents was more than 200 KPa (the pressure limit of our radial compression chamber).
Biomaterials and Stent Technology
115
Figure 4 Bioresorbable stent radial compression strength. (䊏) Single and (●) double fiber stents, as a function of immersion time in PBS at 37 ⬚C. The number of rupture points are indicated. (From Ref. 33.)
The stents were immersed in phosphate-buffered saline at 37 ⬚C, for certain periods of time in order to investigate the effect of their invitro degradation on mechanical properties. The mode of failure observed was rupture of binding points, where the longitudinal support fibers were glued to the coil. The radial compression pressure needed in order to create a rupture of at least one binding point, in both types of stents, is presented as a function of immersion time in Fig. 4. The number of ruptures (binding points that failed) is indicated in parentheses for each immersion time. Neither type of stent underwent any failure after applying 200 KPa throughout the first 8 weeks. Then the radial compression pressure required to create a rupture at binding points showed a linear decrease with time, and the number of ruptures increased with time. Since the double fiber design has more binding points than the single fiber one, each binding point of the former is exposed to a smaller pressure; thus a total higher pressure is needed to fail the double fiber stent. In general, these stents showed good radial compression endurance. They resisted at least 150 KPa (approximately 75% of the initial strength) and exhibited only several rupture points, while most of the binding points remained intact. The combination of the suggested design and the relatively high molecular weight PLLA was concluded to be applicable for supporting blood vessels for at least 20 weeks and was chosen for further studies.
IV. STENTS AS RESERVOIRS FOR LOCAL DRUG AND GENE THERAPY A. Metal Stent Surface Treatments Abundant evidence shows that restenosis following stent placement entails a series of interrelated host responses: thrombosis, inflammation, and smooth muscle cell (SMC) proliferation. Stent treatments have addressed each of these three responses. The earliest view held that metal stent strut–induced thrombogenesis ultimately resulted in restenosis. Thus major efforts were made to reduce the thrombogenic property of metal stents. As thrombosis was controlled, it became clear that other (inflammatory) processes governed the restenosis rate. Broader strategies, employing both passive and active surface modification, were called for to reduce the host response. The so-called passivation methods alter selected properties of the stent surface without delivery of drugs or other agents. These techniques are expected to remove surface initiators of the host
116
Nguyen et al.
response, although less intense cellular responses to the surface may continue. In contrast, active surface modification provides a drug release matrix to block or otherwise alter cellular responses. Significant reduction of platelet aggregation, inflammation, and SMC proliferation are achieved by some of these means. B. Surface Passivation The passivation approach is frequently employed to generate a thromboresistant coating. Both organic and inorganic compounds are used. Among many nonpolymeric organic chemicals, phosphorylcholine (PC) has had the most attention as a stent passivation method. PC mimics phosphatidyl choline, a major component of the outer membrane layer of the red blood cell which inhibits the activation of formed and unformed elements of blood. There is good evidence to show that PC-coated stents reduce thrombotic complications. A recent clinical trial of the PC-coated BiodivYsio stent yielded encouraging results. In 6-month implantations, angiographic evidence of stenosis decreased from 89 to 5.6% [34,35]. However, the PC coating is subject to erosion. Optimization of the PC coating and longer-term studies are needed to prove its superiority. Organic polymeric materials are also used for stent surface modification. Both biostable and biodegradable polymer coatings have been studied. Most polymer-coated stents are in the developmental stage and frequently have been tested only in experimental animal models. Biostable polymer coatings include expanded polytetrafluoroethylene (ePTFE) [36], polyester (Dacron) [37], a blend of methylmethacrylate (MMA) and 2-hydroxyethyl methacrylate (HEMA) [38], fluorine-acryl-styrene-urethane-silicone (FASUS) copolymer [39], parylene [40], and amphiphilic polyurethane [41,42]. Variable success rates in reducing thrombus formation or neointimal growth have been reported. Biodegradable polymer-coated stents of poly-(organo)phosphazene [42] and poly-(L-lactic acid) (PLLA) [43] have also been reported to be less thrombogenic. Despite the short-term successes, a major concern for all these polymer coatings is the longerterm tissue–materials response. As noted above, mild to severe foreign body responses associated with biodegradable and biostable polymer coatings have been observed. In some cases, an overwhelming inflammatory response has been closely linked to later restenosis, an unacceptable outcome [44]. Inorganic compounds, including gold, silicon carbide, hydrogenated carbon, and titanium nitride/oxide, have been tested as thromboresistant surfaces for metal stents. Gold is believed to be a biocompatible material [45,46]. However, gold-coated stents revealed poor thromboresistance in an ex vivo human stasis model [47]. The 12-month outcome from a randomized trial of a gold-coated stent yielded less favorable neointimal proliferation [48]. Silicon carbide, postulated as a less thrombogenic semiconducting material, was shown to inhibit IgG, fibrinogen, and fibronectin adsorption [49]. Fibrinogen is easily activated when it comes in contact with a metallic surface, and activated fibrinogen in the presence of thrombin can be converted into fibrin to start the clotting process. Patients implanted with silicon carbide–coated stents showed less acute and subacute stent thrombosis, even when only antiplatelet medication was received postprocedure [50]. This finding was confirmed in a larger scale randomized trial. However, the long-term benefit of silicon carbide coating was not confirmed in the same study [51]. Hydrogenated diamondlike carbon and titanium nitride oxide are other inert materials that have been applied as stent coatings. By means of plasma or electronic beam techniques, diamondlike nanocomposite and titanium nitride/oxide coatings were deposited onto stent surfaces. Significant reductions of adherent thrombus on both coated stents was demonstrated in a short-term, porcine model study [52,53]. But reduction of neointimal proliferation was observed only for the titanium nitride/oxide coating.
Biomaterials and Stent Technology
117
In general, metal stent surface modifications by these passive approaches achieve only a certain degree of thromboresistance, without significant reduction of neointimal hyperplasia. Thus more aggressive approaches must be taken in order to reach a satisfactory outcome. C. Active Surface Modification The active approach aims for tissue responses beyond thrombosis, since thromboresistant surfaces are achievable with some surface passivation methods. A number of active surface modification approaches are based on the successful development of stent coating procedures. These stent coatings function as reservoirs for specific therapeutic agents. The rationale is that drugs are gradually and locally released from the coating matrix and taken up by surrounding tissues. A reduction in neointimal hyperplasia is expected, at least for the period that the beneficial drug actions remain in effect and perhaps beyond, if healing processes are sufficiently advanced. Optimal combination of drugs and coating matrix maximizes the local therapeutic benefit, thereby maintaining the patency of the stented vessel. A variety of pharmacologic agents have been utilized to prevent thrombosis and restenosis (Table 4). The agents are loaded in bulk by physical entrapment or by chemically bonding to the polymer chain [18]. Four types of drugs have been studied in experimental animal models or human clinical trials: anticoagulants and, antiplatelet, antiproliferative, and immunosuppressive agents. The rationale for immobilized heparin, the most intensively investigated stent coating agent, is to inhibit thrombin-dependent positive feedback reactions at the stent surface [54].
Table 4 Studies of Local Drug Delivery Using a Stent-Based System Type Antiplatelet
Agents/drugs Sodium nitroprusside (SNP), NO donor Abciximab, GPIIb/IIIa inhibitor AZ1, GP IIb/IIIa receptor antibody Eptifibatide, GPIIb/IIIa inhibitor L-703081, GPIIb/IIIa antagonist
Antiproliferation
Immunosuppressant
Source: Ref. 11.
Argatroban, inhibitor of thrombininduced platelet activation Tyrosine kinase inhibitor Angiopeptin, inhibitor of SMC proliferation Paclitaxel and 7-hexanoytaxol (QP2), microtubular inhibitor Methylprednisolone, anti-inflammation Dexamethasone, anti-inflammation Sirolimus (rapamycin), immunosuppressant and antiproliferative
Trial/animal model
References
Pig coronary arteries
[61]
TARGET, ADMIRAL Rabbit iliac arteries CRUISE, ESPRIT Canine coronary arteries Swine coronary arteries Pig coronary arteries Pig coronary arteries
[140,141] [62] [142–144] [63] [64] [20] [67]
TAXUS I, ASPECT; rabbit iliac arteries Pig coronary arteries
[68–70,145,146]
Pig coronary arteries RAVEL, US trials
[72] [76–78,148,149]
[71,147]
118
Nguyen et al.
Heparin-coated stents have been used in a primary antirestenosis strategy (Benestant-II, MENTOR) and in the context of primary angioplasty for acute myocardial infarction (PAMI stent pilot trial) and total coronary occlusion (TOSCA) [55–58]. Although thrombogenicity was reduced in the Benestent-II study, investigators doubted the importance of the heparin coating. In the longer term (6-month follow-up), the heparin coating did yield improved angiographic evidence of restenosis [59]. Stents coated with the anticoagulant hirudin have been shown to have reduced neointima formation [60]. Despite these findings, it is now generally believed that anticoagulant coatings are unnecessary when proper stent deployment techniques are coupled with effective antiplatelet medication, e.g., aspirin and ticlopidine. Antiplatelet agents employed in drug-eluting stents include nitric oxide donors, glycoprotein IIb/IIIa receptor antagonist/antibodies, and thrombin inhibitors. Significant reductions in platelet deposition and thrombosis have been demonstrated for several of these agents in animal studies. Unfortunately, no indication was revealed that the palliation of thrombosis by these means resulted in the inhibition of intimal growth [61–64]. Antiproliferative drug-eluting stents combat restenosis by directly inhibiting vessel smooth muscle cell proliferation. Drugs designed to inhibit SMC migration or to arrest the SMC cell cycle at the G2/M transition have been studied [65,66]. Stents coated with such drugs proved to reduce neointimal growth in animal studies [20,67,68]. Paclitaxel-eluting stents attracted particular attention. A paclitaxel-eluting stent was the first device reported to have 6-month inhibitory efficacy on neointimal growth [68]. This 6-month efficacy was reproduced in a human clinical trial, although the antiproliferative efficacy was not maintained in 12-month follow-up as measured by angiographic restenosis [69]. Postponed thrombosis and uncompleted wound healing were later reported in more thorough animal examinations [70]. The paclitaxel-eluting stent experience nevertheless implies that the optimal combination of drug and coating matrix may yield a long-term, restenosis-free therapy. A third, exciting area of progress concerns stents coated with immunosuppressive drugs. These coatings attack neointimal growth in a different way, possessing both potent anti-inflammatory and immunosuppressive effects. Such drugs have been shown to decrease the migration and functional capabilities of inflammatory cells. Some of these drugs prevent cell cycle progression in the G1 stage, and consequently inhibit cell proliferation [71–73]. In a pig coronary artery model, methylprednisolone-eluting stents showed promise in reducing the foreign body reaction and restenosis at 6 weeks [71]. In 1-month porcine and rabbit models, a rapamycin (sirolimus)eluting stent inhibited restenosis [74,75]. Moreover, the long-term behavior of the sirolimuseluting stent is as encouraging as its short-term property. Multiple experiences in human clinical trials show that the antirestenotic efficacy of the sirolimus-eluting stents is sustained for 12 months [76–78]. In contrast to their opinion of paclitaxel-eluting stents, interventional cardiologists are optimistic about the 12-month follow-up angiographic outcome with this device. As of this writing, it is uncertain whether or not improper dosage of a sirolimus-eluting stent would initiate localized hypercholesterolemia, a risk factor for coronary disease [79,80]. The concern is being addressed in long-term studies with more thorough histopathological outcome evidence. D. Stent-Based Gene Therapy In addition to local drug delivery, stents can also serve as carriers for gene therapy delivery. Stents seeded with cells transfected with the desired gene, stents loaded with recombinant adenovirus gene transfer vectors, and stents loaded with naked DNA impregnated in various matrices have been proposed [21,81–84]. The introduction of a gene of interest into the arterial wall can be achieved either by in vitro genetic manipulation of cells before their seeding onto stents or by direct in vivo gene transfer. Cell-based gene transfer via a stent platform has been shown to
Biomaterials and Stent Technology
119
be advantageous in terms of site-specific gene expression. However, cell-based gene delivery has several limitations, including removal or injury of cells from the stent after balloon expansion and significant time delay, required for cell harvest, expansion, gene transfer, and subsequent selection prior to stent seeding. Yet seeding with genetically engineered endothelial cells (ECs) producing tPA was shown to inhibit SMC proliferation [85]. A recent study has shown that a mesh-stent coated with fibronectin is an excellent platform for adherence of transduced SMC [86]. Site-specific gene therapy via direct gene release from stents is an attractive concept. Genes that encode enzymes of the prostacyclin synthetic pathway, nitric oxide synthase, thrombin inhibitor, and thrombomodulin have been demonstrated to significantly reduce thrombosis and restenosis in nonstented animal models [21,84,87–89]. However, in spite of promising results in animal models, no effective human gene therapy has yet been found to prevent restenosis [21,84,90]. Another promising strategy for gene therapy delivery involves the introduction of antisense oligonucleotides into cells in order to inactivate the mRNA encoding proteins important in the restenotic process [91]. The use of synthetic oligonucliotides to suppress proto-oncogenes, including c-myb and c-myc, proliferating cell nuclear antigen, and cell cycle specific proteins cdc2 and cdk2 kinases, reduced protein expression and cell proliferation [15,91]. We successfully demonstrated local gene transfer and expression from PLLA stents. We implanted stents loaded with a recombinant adenovirus carrying a nuclear localizing Gal reporter gene into the carotid and renal arteries in the rabbit. Local expression of copious amounts of Gal was demonstrated from intima to adventitia, at both sites. Liver transfection was negligible in both cases, suggesting that the gene was not convected to remote sites to a significant degree [82]. The potential side effects of the gene therapy approach are of concern. Malignant transformation due to oncogene activation via retroviral gene vectors and subsequent gene expression in other organs is only one problem that requires careful evaluation. Stents have been used to combat another potentially dangerous problem, the distal spread of viral vectors. Antiviral antibodies have been tethered onto the collagen-coated surface of a stent, creating a suitable platform for local gene delivery [92]. E. Drug Delivery from Bioresorbable Stents Bioresorbable stents can serve both to mechanically support blood vessels and as drug delivery platforms. We have studied this concept with steroid- loaded stent fibers. Dexamethasone was successfully incorporated into PLLA fibers during melt processing. However, only small drug quantities (less than 10 wt%) could be loaded before loss of mechanical properties was observed. Notable exceptions to this restriction were obtained with melt-processed curcumin and paclitaxel [24]. Most other drugs, and all proteins will be destroyed when exposed to the elevated melt processing temperatures. Bulk loading by wet spinning or porous coatings, hollow microspheres, or other drug reservoir techniques is required in this case (Fig. 5). In this regard, PLLA fibers with a bonded microsphere reservoir have been investigated in our laboratory. These microspheres can be loaded with aqueous or nonaqueous biologically active molecules. Since mild materials and processing steps are used, these microspheres can be also loaded with proteins or gene transfer vectors. To study the concept of microsphere-coated stents as aqueous drug delivery systems, 75/25 poly(DL-lactide-co-glycolide) (PDLGA) microspheres containing bovine serum albumin (BSA) were prepared through a water/oil/water double emulsion process. In this technique, a small volume of aqueous phase is first emulsified with a polymeric (polymer Ⳮ solvent) phase (W/O), and this emulsion is reemulsified in a large volume of aqueous phase to generate a (W/
120
Nguyen et al.
Figure 5 Techniques for reservoir formation on bioresorbable stents. (a) Incorporation of curcumin in melt spun PLLA fiber (from Ref. 29). (b) Porous PLLA coating on PLLA fiber. (c) PDLGA microspheres attached to melt-spun PLLA fiber (from Ref. 33).
Biomaterials and Stent Technology
121
O)/W emulsion. The volatile solvent is removed as it diffuses into the aqueous medium and evaporates at the air–water interface during the second emulsion step. Removal of the solvent precipitates the polymer, thereby immobilizing and encapsulating the inner aqueous phase within a spherical polymeric matrix. The last stage of this process is separation of the microspheres from the aqueous medium by centrifugation. A water-soluble drug can be incorporated in the microsphere by dissolving it in the first (small) aqueous phase. A non-water-soluble drug can be incorporated in the microsphere by dissolving it the solvent of the polymer. Two kinds of 75/25 PDLGA with differing intrinsic viscosities were used: (i.v.) ⳱ 0.35 dL/g (⬃43,000 Da) and 0.69 dL/g (⬃118,000 Da). The ‘‘oil’’ phase contained 20 wt% polymer in methylene chloride and the internal water phase contained 20 wt% BSA. The quantity of the ‘‘oil’’ phase was constant, and two different internal water/polymer (W/P) ratios were tried: 1 (large BSA content) and 0.2 (small BSA content). BSA-loaded microspheres of 10–60 m diameter were obtained in both cases. These had smooth surface and a core/shell (aqueous phase/ polymer) structure. Approximately 85% of the BSA was taken up in these microspheres. The 75/ 25 PDLGA shell was dense and did not show any porous structure. The W/P ⳱ 1 microspheres, containing high water content, exhibited a relatively thin polymer shell, while the W/P ⳱ 0.2, containing low water content, exhibited a relatively thick polymer shell. The PLLA fibers were immersed in a 70/30 methylene chloride/ethanol solution for 3 s in order to slightly dissolve their surface layers, then the microspheres were bound to the fiber (see Fig. 5). The same process was used to coat stents with microspheres. The initial radial compression strength of the four types of ‘‘microsphere-coated’’ stents was higher than 200 KPa. The cumulative BSA release profiles from the four types of microsphere-loaded stents are presented in Figure 6. In general, a burst effect is accompanied by a linear cumulative release profile, as expected for a ‘‘reservoir’’ system, such as these core/shell microspheres. After 6 weeks of incubation, the relatively low molecular weight (i.v. ⳱ 0.35) microspheres exhibited rough surface features with cracks and ruptures, and a porous shell structure, due to erosion processes. In contrast, the higher molecular weight microsphere (i.v. ⳱ 0.69) exhibited only very thin cracks on the surface. Therefore the release rate from the i.v. ⳱ 0.35 microsphere stents was higher than that from the i.v.⳱ 0.69 ones, and the burst effect of the former is higher than that of the latter. Also, the low W/P ratio, leading to thick polymer shell, is effective in reducing the burst effect.
Figure 6 Cumulative in vitro BSA release from different types of 75/25 PDLGA microspheres. (앩) i.v. ⳱ 0.35 dL/g, W/P ⳱ 1; (●) i.v. ⳱ 0.35 dL/g, W/P ⳱ 0.2; (왍) i.v. ⳱ 0.69 dL/g, W/P ⳱ 1; (䊏) i.v. ⳱ 0.69 dL/g, W/P ⳱ 0.2. (From Ref. 33.)
122
Nguyen et al.
In summary, novel expandable bioresorbable stents were prepared, with excellent initial radial compression strength and good invitro degradation resistivity. Microspheres bonded to these stents enabled effective protein (BSA) loading, without reducing stent mechanical strength. The protein release profile from the microsphere-loaded stent occurs by diffusion and is determined by the polymer erosion rate (determined by its chemical structure), its initial molecular weight, and the microsphere structure. The release rate from the stent can be controlled by choosing appropriate bioresorbable polymer and microsphere parameters.
V. NONVASCULAR USES OF STENTS The range of stent applications has expanded with increased experience and encouraging results in the treatment of vascular diseases. Stents have been used for the treatment of urethral obstruction from benign prostatic hyperplasia; for treatment of tracheobronchial obstruction of benign or malignant origin; for treatment of benign and malignant strictures of the esophagus, the GI tract, and the bile duct; and for treatments (stents and stent-grafts) of arterial dissections, aneurysms, and various neurovascular diseases.
A. Stents in Urology Stents have been used to prevent postoperative urine retention following thermal treatment of benign prostatic hyperplasia (BPH) by various means, including direct vision laser ablation of the prostate and transurethral microwave therapy. Several stent designs, including the Nissenkorn, Barnes, Finnish biodegradable self-reinforced polyglycolic acid (SR-PGA) spiral, and Trestle were shown to prevent obstruction of the prostatic urethra and restricture of the anterior urethra [19,93,94]. Biodegradable stents have been studied clinically in the treatment of benign prostatic hyperplasia and are claimed to provide superior results to suprapubic catheters [95–99]. Self-reinforced PLLA bioresorbable spiral stents are also undergoing evaluation for use in the anterior urethra, posterior urethra, and upper urinary tract to prevent urinary retention and repair of local ureteral trauma or defects [100,101]. Surface modification of these biodegradable stents, by grafting with hydroxyethylmethacrylate or by incorporation of biologically active compounds, is claimed to be an efficient approach to improve biocompatibility and cell adhesion properties [102,103].
B. Stents for Management of Tracheobronchial Obstruction Tracheobronchial obstruction from either benign or malignant disease causes significant morbidity and mortality. Metal stents, developed originally for the vascular system, have been adapted for lesions involving the tracheobronchial tree. These include the Palmaz (Johnson & Johnson), Strecker (Boston Scientific), Gianturco-Z (William Cook Europe), Wallstent (Boston Scientific), and Ultraflex (Boston Scientific) stents [104]. These stents were successfully used to treat patients with inoperable bronchogenic cancer, esophageal tumors, primary tracheal tumors, and metastatic malignancy. Bioresorbable tracheal stents have been investigated in the setting of pediatric tracheal malacia to solve the problem of limited tracheal growth in children with rigid external fixation and to avoid the necessity of a second procedure to remove the synthetic material [104–106]. The general results from these studies suggest that stenting is a promising method to treat tracheal obstruction.
Biomaterials and Stent Technology
123
C. Stents in the Esophagus and Gastrointestinal Tract Many malignant and benign strictures in the esophagus and GI tract can be treated by minimally invasive alternatives to surgery, including the use of stents. Most commonly used in the esophagus and gastrointestinal tract are the Wallstent (Boston Scientific), Ultraflex (Boston Scientific), Gianturco-Z (William Cook Europe), Esophacoil (Instent), and Flamingo stents (Boston Scientific). In general, these stents have been shown to be effective in relieving esophageal dysphagia [107–109]. This success has led to the employment of stents to manage lesions of the GI tract, including the stomach, pylorus, upper small intestine, duodenum, and colon [107,108]. The use of bioresorbable material is currently being explored for the esophageal stent. First results in the placement of a PLLA stent (Instent) for the management of benign esophageal stricture suggest that a bioresorbable stent offers a new treatment modality [110]. In addition, bioresobable stents were recently used in pancreaticojejunal anastomoses [111]. D. Stents in the Management of Neurovascular Disease Stents and stent-grafts have been used for the management of arterial and venous sinus stenosis, arterial dissection, arterial aneurysm, and arteriovenous fistulae [112,113]. A number of case reports have been published describing the significant reduction of carotid stenosis with the use of stents in the treatment of carotid stenosis, recurrent carotid stenosis, vertebral artery stenosis, and venous sinus stenosis [113]. As cited in this review, Shawl et al. reported a series of 124 stented vessels in which carotid stenosis was reduced from 86 Ⳳ 7% to 2 Ⳳ 2%; the major postprocedural stroke rate was 0.8%, and the minor stroke rate was 2%. Three cases of basilar artery stenosis have been successfully treated with coronary stents at our institution [114]. Other unpublished reports from our institution have demonstrated the effectiveness of stents in bridging side-wall aneurysmal ostia, suggesting stents are a promising means for the management of arterial dissection and pseudoaneurysm. Unfortunately, no large studies have yet been published, so the effectiveness of stents in this application remains to be determined.
ACKNOWLEDGMENTS This work was supported in part by USPHS Grants RO1 HL53225 and F32 HL010380.
REFERENCES
1. http://webster.com/about/ack.htm. 2. Ashby DT, Dangas G, Mehran R, Leon MB. Coronary artery stenting. Catheter Cardiovasc. Interv 2002; 56:83–102. 3. Ratner BD, Hoffman AS, Schoen FJ, Lemons JE, eds. Biomaterials Science: Academic Press, 1996. 4. Lincoff AM. Stent scrutiny. JAMA 2000; 284:1839–41. 5. Sigwart U. Coronary stents. Z. Kardiol 1995; 84 Suppl 2:65–77. 6. www.medtronic.com/newsroom/news_19991116114343.html. 7. www.guidant.com. 8. www.nlm.nih.gov/medlineplus/ency/presentations/100160_6.htm. 9. www.endovascularone.com. 10. McConachie I, Roberts DH. Handbook of Cardiac Emergencies: Greenwich Medical Media Limited, 2000.
124
Nguyen et al.
11. Eberhart RC, Su S-H, Nguyen KT, Zilberman M, Tang L, Nelson K, Frenkel P. Bioresorbable polymeric stents: current status and future promise. J. Biomater. Sci. Polym. Ed 2002: In press. 12. www.jnjgateway.com. 13. www.biocompatibles.co.uk. 14. Wise DL. Biomaterials and Bioengineering Handbook. New York: Marcel Dekker, 2000. 15. Tanguay JF, Zidar JP, Phillips HR, Stack RS. Current status of biodegradable stents. Cardiol Clin 1994; 12:699–713. 16. Labinaz M, Zidar JP, Stack RS, Phillips HR. Biodegradable stents: the future of interventional cardiology? J. Int. Cardiol 1995; 8:395–405. 17. Colombo A, Karvouni E. Biodegradable stents. Fulfilling the mission and stepping away. Circulation 2000; 102:371–373. 18. Peng T, Gibula P, Yao K, Goosen MFA. Role of polymers in improving the results of stenting in coronary arteries. Biomaterials 1996; 17:685–694. 19. Talja M, Valimaa T, Tammela T, Petas A, Tormala P. Bioabsorbable and biodegradable stents in urology. J Urol 1997; 11:391–397. 20. Yamawaki T, Shimokawa H, Kozai T, Miyata K, Higo T, Tanaka E, Egashira K, Shiraishi T, Tamai H, Igaki K, Takeshita A. Intramural delivery of a specific tyrosine kinase inhibitor with biodegradable stent suppresses the restenotic changes of the coronary artery in pigs in vivo. J. Am. Coll. Cardiol 1998; 32:780–786. 21. Indolfi C, Coppola C, Torella D, Arcucci O, Chiariello M. Gene therapy for restenosis after balloon angioplasty and stenting. Cardiol Rev 1999; 7:324–331. 22. Lendlein A, Langer R. Biodegradable, elastic shape-memory polymers for potential biomedical applications. Science 2002; 296:1673–1676. 23. van der Giessen WJ, Lincoff M, Schwartz RS, van Beusekom HMM, Serruys PW, Holmes DR, Ellis SG, Topol EJ. Marked inflammatory sequelae to implantation of biodegradable and nonbiodegradable polymers in porcine coronary arteries. Circulation 1996; 94:1690–1697. 24. Su S-H, University of Texas at Arlington, 2000. 25. Hietala E, Salminen U, Sta¨hls A, Va¨limaa T, Maasilta P, To¨rma¨la¨ P, Nieminen MS, Harjula ALJ. Biodegradation of the copolymeric polylactide stent. Long-term follow-up in a rabbit aorta model. J Vasc Res 2001; 38:361–369. 26. Agrawal CM, Clark HG. Deformation characteristics of a bioabsorbable intravascular stent. Invest Radiol 1992; 27:1020–1024. 27. Gao R, Shi R, Qiao S, Song L, Li Y. A novel polymeric local heparin delivery stent: initial experimental study. J Am Coll Cardiol 1996; 27:85A. 28. Tamai H, Igaki K, Kyo E, Kosuga K, Kawashima A, Matsui S, Komori H, Tsuji T, Motohara S, Uehata H. Initial and 6-month results of biodegradable poly-L-lactic acid coronary stents in humans. Circulation 2000; 102:399–404. 29. Su S-H, Chao RY, Landau CL, Nelson KD, Timmons RB, Meidell RS, Eberhart RC. An expandable bioresorbable endovascular stent: 1 Fabrication and properties (in review). 30. Su S-H, Landau CL, Chao RY, Timmons RB, Meidell RS, Tang L, Eberhart RC. Expandable, bioresorbable endovascular stent with anti-platelet and anti-inflammation treatments. Circulation 2001; 104:II–507. 31. Bertrand OF, Sipehia R, Mongrain R, Rode´s J, Tardif J, Bilodeau L, Coˆte´ G, Bourassa MG. Biocompatibility aspects of new stent technology. J. Am. Coll. Cardiol 1998; 32:562–571. 32. Alikacem N, Yoshizawa T, Wilson C, Nelson KD. Quantitative MR imaging study of intravitreal sustained release of VEGF in rabbits. Invest. Ophthalmol. Vis. Sci 2000; 41:1561–1569. 33. Zilberman M, Su S-H, Schwade ND, Eberhart RC. Protein-loaded bioresorbable fibers and expandable stents: mechanical properties and protein release. J. Biomater. Sci. In press. 34. Atalar E, Haznedaroglu I, Aytemir K, Aksoyek S, Ovunc K, Oto A, Ozmen F. Effects of stent coating on platelets and endothelial cells after intracoronary stent implantation. Clin. Cardiol 2001; 24:159–164. 35. Grenadier E, Roguin A, Hertz I, Peled B, Boulos M, Nikolsky E, Amikam S, Kerner A, Cohen S, Beyar R. Stenting very small coronary narrowings (⬍ 2 mm) using the biocompatible phosphorylcholine-coated coronary stent. Cath. Cardiovasc. Interv 2002; 55:303–308.
Biomaterials and Stent Technology
125
36. Woody JD, Nishanian G, Kopchok GE, Kim NI, Donayre CE, White RA. Healing response of normal canine aorta and iliac artery to a Nitinol stent encapsulated in carbon-lined ePTFE. J. Endovasc. Ther 2001; 8:274–281. 37. Avino A, Johnson B, Bandyk D, Back M, Roth S, Kudryk B, Cantor A. Does prosthetic covering of Nitinol stents alter healing characteristics or hemodynamics?. J. Endovasc. Ther 2000; 7:469–478. 38. Bar FW, van der Veen H, Benzina A, Habets J, Koole LH. New biocompatible polymer surface coating for stents results in a low neointimal response. J. Biomed. Mater. Res 2000; 52:193–198. 39. Matsuhashi T, Miyachi H, Ishibashi T, Sakamoto K, Yamadera A. In vivo evaluation of a fluorineacryl-stylene-urethane-silicone antithrombogenic coating material copolymer for intravascular stents. Acad. Radiol 1996; 3:581–588. 40. Fontaine AB, Koelling K, Passos SD, Cearlock J, Hoffman R, Spigos DG. Polymeric surface modifications of tantalum stents. J. Endovasc. Surg 1996; 3:276–283. 41. Fontaine AB, Dos Passos S, Spigos D, Cearlock J, Urbaneja A. Use of polyetherurethane to improve the biocompatibility of vascular stents. J. Endovasc. Surg 1995; 2:255–265. 42. De Scheerder IK, Wilczek KL, Verbeken EV, Vandorpe J, Lan PN, Schacht E, Piessens J, De Geest H. Biocompatibility of biodegradable and nonbiodegradable polymer-coated stents implanted in porcine peripheral arteries. Cardiovasc. Intervent. Radiol 1995; 18:227–232. 43. Schellhammer F, Berlis A, Bloss H, Pagenstecher A, Schumacher M. Poly-lactic-acid coating for endovascular stents. Preliminary results in canine experimental arteriovenous fistulae. Invest. Radiol 1997; 32:180–186. 44. Fischell TA. Polymer coatings for stents. Can we judge a stent by its cover? Circulation 1996; 94: 1494–1495. 45. Ignatius MJ, Sawhney N, Gupta A, Thibadeau BM, Monteiro OR, Brown IG. Bioactive surface coatings for nanoscale instruments: effects on CNS neurons. J. Biomed. Mater. Res 1998; 40: 264–274. 46. Zitter H, Plenk H. The electrochemical behavior of metallic implant materials as an indicator of their biocompatibility. J. Biomed. Mater. Res 1987; 21:881–896. 47. Herrmann R, Schmidmaler G, Ma¨rkl B, Resch A, Ha¨hnel I, Stemberger A, Alt E. Antithrombogenic coating of stents using a biodegradable drug delivery technology. Thromb. Haemost 1999; 82: 51–57. 48. Kastrati A, Schomig A, Dirschinger J, Mehilli J, von Welser N, Pache J, Schuhlen H, Schilling T, Schmitt C, Neumann FJ. Increased risk of restenosis after placement of gold-coated stents: results of a randomized trial comparing gold-coated with uncoated steel stents in patients with coronary artery disease. Circulation 2000; 101:2478–2483. 49. Takami Y, Yamane S, Makinouchi K, Otsuka G, Glueck J, Benkowski R, Nose Y. Protein adsorption onto ceramic surfaces. J. Biomed. Mater. Res 1998; 40:24–30. 50. Heublein B, Ozbek C, Pethig K. Silicon carbide–coated stents: clinical experience in coronary lesions with increased thrombotic risk. J. Endovasc. Surg 1998; 5:32–36. 51. Kalnins U, Erglis A, Dinne I, Kumsars I, Jegere S. Clinical outcomes of silicon carbide coated stents in patients with coronary artery disease. Med. Sci. Monit 2002; 8:PI16–PI20. 52. De Scheerder I, Szilard M, Yanming H, Ping XB, Verbeken E, Neerinck D, Demeyere E, Coppens W, Van de Werff F. Evaluation of the biocompatibility of two new diamond-like stent coatings (Dylyn) in a porcine coronary stent model. J. Invasive Cardiol 2000; 12:389–394. 53. Windecker S, Mayer I, De Pasquale G, Maier W, Dirsch O, De Groot P, Wu YP, Noll G, Leskosek B, Meier B, Hess OM. Stent coating with titanium-nitride-oxide for reduction of neointimal hyperplasia. Circulation 2001; 104:928–933. 54. Blezer R, Cahalan L, Cahalan PT, Lindhout T. Heparin coating of tantalum coronary stents reduces surface thrombin generation but not factor IXa generation. Blood Coag. Fibrinolysis 1998; 9: 435–440. 55. Serruys PW, Emanuelsson H, van der Giessen W. Heparin-coated Palmaz–Schatz stents in human coronary arteries: early outcome of the Benestent-II pilot study. Circulation 1996; 93:412–422. 56. Vrolix MC, Legrand VM, Reiber JH, Grollier G, Schalij MJ, Brunel P, Martinez-Elbal L, GomezRecio M, Bar FW, Bertrand ME, Colombo A, Brachman J. Heparin-coated Wiktor stents in human coronary arteries (MENTOR trial). MENTOR Trial Investigators. Am. J. Cardiol 2000; 86:385–389.
126
Nguyen et al.
57. Stone GW, Brodie BR, Griffin JJ, Costantini C, Morice MC, St. Goar FG, Overlie PA, Popma JJ, McDonnell J, Jones D, O’Neill W, Grines CL. Clinical angiographic follow-up after primary stenting in acute myocardial infarction: the primary angioplasty in myocardial infarction (PAMI) stent pilot trial. Circulation 1999; 99:1548–1554. 58. Buller CE, Dzavik V, Carere RG, Mancini GB, Barbeau G, Lazzam C, Anderson TJ, Knudtson ML, Marquis JF, Suzuki T, Cohen EA, Fox RS, Teo KK. Primary stenting versus balloon angioplasty in occluded coronary arteries: the total occlusion study of canada (TOSCA). Circulation 1999; 100: 236–242. 59. Wohrle J, Al-Khayer E, Grotzinger U, Schindler C, Kochs M, Hombach V, Hoher M. Comparison of the heparin coated vs the uncoated Jostent—no influence on restenosis or clinical outcome. Eur. Heart J 2001; 22:1808–1816. 60. Alt E, Haehnel I, Beiharz C, Prietzel K, Preter D, Stemberger A, Fliedner T, Erhardt W, Scho¨mig A. Inhibition of neointima formation after experimental coronary artery stenting: a new biodegradable stent coating releasing hirudin and the prostacyclin analogue iloprost. Circulation 2000; 101: 1453–1458. 61. Yoon JH, Wu CJ, Homme J, Tuch RJ, Wolff RG, Topol E, Lincoff AM. Local delivery of nitric oxide from an eluting stent to inhibit neointimal thickening in a porcine coronary injury model. Yonsei Med. J 2002; 43:242–251. 62. Aggarwal R, Ireland D, Azrin M, Ezekowitz M, De Bono D, Gershlick A. Antithrombotic potential of polymer-coated stents eluting platelet glycoprotein IIb/IIIa receptor antibody. Circulation 1996; 94:3311–3317. 63. Santos RM, Tanguay JF, Crowley JJ, Kruse KR, Sanders-Millare D, Zidar JP, Phillips HR, Merhi Y, Garcia-Cantu E, Bonan R, Cote´ G, Stack RS. Local administration of L-703,081 using a composite polymeric stent reduces platelet deposition in canine coronary arteries. Am. J. Cardiol 1998; 82: 673–675, A8. 64. Kruse KR, Crowley JJ, Tanguay JF, Santos RM, Millare DS, Phillips HR, Zidar JP, Stack RS. Local drug delivery of argatroban from a polymeric–metallic composite stent reduces platelet deposition in a swine coronary model. Catheter Cardiovasc. Interv 1999; 46:503–507. 65. Mooradian DL, Fernandes B, Diglio CA, Lester BR. Angiopeptin (BIM23014C) inhibits vascular smooth muscle cell migration in vitro through a G-protein-mediated pathway and is associated with inhibition of adenylyl cyclase and cyclic AMP accumulation. J. Cardiovasc. Pharmacol 1995; 25: 611–618. 66. Sindermann JR, Skaletz-Rorowski A, Bartels A, Hohage H, Plenz G, Schmidt A, Breithardt G. Paclitaxel and cyclosporine A show supra-additive antiproliferative effects on smooth muscle cells by activation of protein kinase C. Basic Res. Cardiol 2002; 97:125–131. 67. Armstrong J, Gunn J, Arnold N, Malik N, Chan KH, Vick T, Stratford P, Cumberland DC, Holt CM. Angiopeptin-eluting stents: observations in human vessels and pig coronary arteries. J. Invasive Cardiol 2002; 14:230–238. 68. Drachman DE, Edelman ER, Seifert P, Groothuis AR, Bornstein DA, Kamath KR, Palasis M, Yang D, Nott SH, Rogers C. Neointimal thickening after stent delivery of paclitaxel: change in composition and arrest of growth over six months. J. Am. Coll. Cardiol 2000; 36:2325–2332. 69. Liistro F, Stankovic G, Di Mario C, Takagi T, Chieffo A, Moshiri S, Montorfano M, Carlino M, Briguori C, Pagnotta P, Albiero R, Corvaja N, Colombo A. First clinical experience with a paclitaxel derivate-eluting polymer stent system implantation for in-stent restenosis: immediate and long-term clinical and angiographic outcome. Circulation 2002; 105:1883–1886. 70. Farb A, Heller PF, Shroff S, Cheng L, Kolodgie FD, Carter AJ, Scott DS, Froehlich J, Virmani R. Pathological analysis of local delivery of paclitaxel via a polymer-coated stent. Circulation 2001; 104:473–479. 71. de Scheerder I, Wang K, Wilczek K, van Dorpe J, Verbeken E, Desmet W, Schacht E, Piessens J. Local methylprednisolone inhibition of foreign body response to coated intracoronary stents. Coron. Artery Dis 1996; 7:161–166. 72. Lincoff AM, Furst JG, Ellis SG, Tuch RJ, Topol EJ. Sustained local delivery of dexamethasone by a novel intravascular eluting stent to prevent restenosis in the porcine coronary injury model. J. Am. Coll. Cardiol 1997; 29:808–816.
Biomaterials and Stent Technology
127
73. Kelly PA, Gruber SA, Behbod F, Kahan BD. Sirolimus, a new, potent immunosuppressive agent. Pharmacotherapy 1997; 17:1148–1156. 74. Klugherz BD, Llanos G, Lieuallen W, Kopia GA, Papandreou G, Narayan P, Sasseen B, Adelman SJ, Falotico R, Wilensky RL. Twenty-eight-day efficacy and phamacokinetics of the sirolimuseluting stent. Coron. Art. Dis 2002; 13:183–188. 75. Suzuki T, Kopia G, Hayashi S, Bailey LR, Llanos G, Wilensky R, Klugherz BD, Papandreou G, Narayan P, Leon MB, Yeung AC, Tio F, Tsao PS, Falotico R, Carter AJ. Stent-based delivery of sirolimus reduces neointimal formation in a porcine coronary model. Circulation 2001; 104: 1188–1193. 76. Morice MC, Serruys PW, Sousa JE, Fajadet J, Ban Hayashi E, Perin M, Colombo A, Schuler G, Barragan P, Guagliumi G, Molnar F, Falotico R. A randomized comparison of a sirolimus-eluting stent with a standard stent for coronary revascularization. N. Engl. J. Med 2002; 346:1773–1780. 77. Sousa JE, Costa MA, Abizaid AC, Rensing BJ, Abizaid AS, Tanajura LF, Kozuma K, Langenhove G, Sousa AG, Falotico R, Jaeger J, Popma JJ, Serruys PW. Sustained suppression of neointimal proliferation by sirolimus-eluting stents: one-year angiographic and intravascular ultrasound followup. Circulation 2001; 104:2007–2011. 78. Serruys PW, Degertekin M, Tanabe K, Abizaid A, Sousa JE, Colombo A, Guagliumi G, Wijns W, Lindeboom WK, Ligthart J, de Feyter PJ, Morice MC. Intravascular ultrasound findings in the multicenter, randomized, double-blind RAVEL (RAndomized study with the sirolimus-eluting VElocity balloon-expandable stent in the treatment of patients with de novo native coronary artery Lesions) trial. Circulation 2002; 106:798–803. 79. Trotter JF, Wachs ME, Trouillot TE, Bak T, Kugelmas M, Kam I, Everson G. Dyslipidemia during sirolimus therapy in liver transplant recipients occurs with concomitant cyclosporine but not tacrolimus. Liver Transpl 2001; 7:401–408. 80. Podder H, Stepkowski SM, Napoli KL, Clark J, Verani RR, Chou TC, Kahan BD. Pharmacokinetic interactions augment toxicities of sirolimus/cyclosporine combinations. J. Am. Soc. Nephrol 2001; 12:1059–1071. 81. Bult H. Restenosis: a challenge for pharmacology. Ti.P.S 2000; 21:274–279. 82. Ye YW, Landau C, Willard JE, Rajasubramanian G, Moskowitz A, Aziz S, Meidell RS, Eberhart RC. Bioresorbable microporous stents deliver recombinant adenovirus gene transfer vectors to the arterial wall. Ann. Biomed. Eng 1998; 26:398–408. 83. Klugherz BD, Jones PL, Cui X, Chen W, Meneveau NF, DeFelice S, Connolly J, Wilensky RL, Levy RJ. Gene delivery from a DNA controlled-release stent in porcine coronary arteries. Nat. Biotechnol 2000; 18:1181–1184. 84. Feldman MD, Sun B, Koci BJ, Wu CC, Kneller JR, Borovetz HS, Watkins S, Nadeem A, Weiss LE, Reed ML, Smith AJ, Rosenblum WD. Stent-based gene therapy. J. Long Term Eff. Med. Implants 2000; 10:47–68. 85. Dichek DA, Neville RF, Zwiebel JA, Freeman SM, Leon MB, Anderson WF. Seeding of intravascular stents with genetically engineered endothelial cells. Circulation 1989; 80:1347–1353. 86. Panetta CJ, Miyauchi K, Berry D, Simari RD, Holmes DR, Schwartz RS, Caplice NM. A tissueengineered stent for cell-based vascular gene transfer. Hum. Gene Ther 2002; 13:433–441. 87. Vermeersch P, Nong Z, Stabile E, Varenne O, Gillijns H, Pellens M, Pelt N, Hoylaerts M, De Scheerder I, Collen D, Janssens S. l-Arginine administration reduces neointima formation after stent injury in rats by a nitric oxide–mediated mechanism. Arterioscler. Thromb. Vasc. Biol 2001; 21: 1604–1609. 88. Akyurek LM, Nallamshetty S, Aoki K, San H, Yang ZY, Nabel GJ, Nabel EG. Coexpression of guanylate kinase with thymidine kinase enhances prodrug cell killing in vitro and suppresses vascular smooth muscle cell proliferation in vivo. Mol Ther 2001; 3:779–786. 89. Varenne O, Sinnaeve P. Gene therapy for coronary restenosis: a promising strategy for the new millennium?. Curr. Interv. Cardiol. Rep 2000; 2:309–315. 90. Gruberg L, Waksman R, Satler LF, Pichard AD, Kent KM. Novel approaches for the prevention of restenosis. Expert Opin. Invest. Drugs 2000; 9:2555–2578. 91. Kipshidze N, Moses J, Shankar LR, Leon M. Perspectives on antisense therapy for the prevention of restenosis. Curr. Opin. Mol. Ther 2001; 3:265–277.
128
Nguyen et al.
92. Klugherz BD, Song C, DeFelice S, Cui X, Lu Z, Connolly J, Hinson JT, Wilensky RL, Levy RJ. Gene delivery to pig coronary arteries from stents carrying antibody-tethered adenovirus. Hum. Gene Ther 2002; 13:443–454. 93. Kapoor R, Liatsikos EN, Badlani G. Endoprostatic stents for management of benign prostatic hyperplasia. Curr. Opin. Urol 2000; 10:19–22. 94. Kletscher BA, Oesterling JE. Prostatic stents. Current perspectives for the management of benign prostatic hyperplasia. Urol. Clin 1995; 22:423–430. 95. de la Rosette JJMCH, Beerlage HP, Debruyne FMJ. Role of temporary stents in alternative treatment of benign prostatic hyperplasia. J. Endourol 1997; 11:467–472. 96. Kemppainen E, Talija M, Riihela¨ M, Pohjonen T, To¨rma¨la¨ P, Alfthan OA. Bioresorbable urethral stent. Urol. Res 1993; 21:235–238. 97. Pe´tas A, Talja M, Tammela T, Taari K, Lehtoranta K, Va¨limaa T, To¨rma¨la¨ P. A randomized study to compare biodegradable self-reinforced polyglycolic acid spiral stents to suprapubic and indwelling catheters after visual laser ablation of the prostate. J. Urol 1997; 157:173–176. 98. Pe´tas A, Isotalo T, Talja M, Tammela TLJ, Va¨limaa T, To¨rma¨la¨ P. A randomised study to evaluate the efficacy of a biodegradable stent in the prevention of postoperative urinary retention after interstitial laser coagulation of the prostate. Scand. J. Urol. Nephrol 2000; 34:262–266. 99. Pe´tas A, Ka¨rkka¨inen P, Talja M, Taari K, Laato M, Va¨limaa T, To¨rma¨la¨ P. Effects of biodegradable self-reinforced polyglycolic acid, polylactide acid and stainless-steel spiral stents on uroepithelium after Nd:YAG laser irradiation of the canine prostate. Br. J. Urol 1997; 80:903–907. 100. Lumiaho J, Heino A, Pietila¨inen T, Ala-Opas M, Talja M, Va¨limaa T, To¨rma¨la¨ P. The morphological, in situ effects of a self-reinforced bioabsorbable polylactide (SR-PLA 96) ureteric stent; an experimental study. J. Urol 2000; 164:1360–1363. 101. Lumiaho J, Heino A, Tunninen V, Ala-Opas M, Talja M, Va¨limaa T, To¨rma¨la¨ P. New bioabsorbable polylactide ureteral stent in the treatment of ureteral lesions: an experimental study. J. Endourol 1999; 13:107–112. 102. Brauers A, Jung PK, Thissen H, Pfannschmidt O, Michaeli W, Hoecker H, Jakse G. Biocompatibility, cell adhesion, and degradation of surface-modified biodegradable polymers designed for the upper urinary tract. Tech. Urol 1998; 4:214–220. 103. Brauers A, Thissen H, Pfannschmidt O, Bienert H, Foerster A, Klee D, Michaeli W, Ho¨cker H, Jakse G. Development of a biodegradable ureteric stent: surface modification and in vitro assessment. J. Endourol 1997; 11:399–403. 104. Rafanan AL, Mehta AC. Stenting of the tracheobronchial tree. Int. Chest Radiol 2000; 38:395–408. 105. Robey TC, Eiselt PM, Murphy HS, Mooney DJ, Weatherly RA. Biodegradable external tracheal stents and their use in a rabbit tracheal reconstruction model. Laryngoscope 2000; 110:1936–1942. 106. Lochbihler H, Hoelzi J, Dietz H. Tissue compatibility and biodegradation of new absorbable stents for tracheal stabilization: an experimental study. J. Pediatr. Surg 1997; 32:717–720. 107. Lamber R. Treatment of esophagogastric tumors. Endoscopy 2000; 32:322–330. 108. Morgan R, Adam A. Use of metallic stents and balloons in the esophagus and gastrointestinal tract. J. Vasc. Interv. Radiol 2001; 12:283–297. 109. Nicholson T. Other uses of non-vascular stents. Hosp. Med 2000; 61:97–102. 110. Fry SW, Fleischer DE. Manegement of a refractory benign esophageal stricture with a new biodegradable stent. Gastrointest. Endosc 1997; 45:179–182. 111. Parviainen M, Sand J, Harmoinen A, Kainulainen H, Va¨limaa T, To¨rma¨la¨ P, Nordback I. A new biodegradable stent for the pancreaticojejunal anastomosis after pancreaticoduodenal resection: in vitro examination and pilot experiences in humans. Pancreas 2000; 21:14–21. 112. Gray HA, Crane AW. Endovascular repair of abdominal aortic aneurysms: current status and future directions. Am. J. Rad 2000; 175:289–302. 113. Horowitz MB, Purdy PD. The use of stents in the management of neurovascular disease: a review of historical and present status. Neurosurgery 2000; 46:1335–1343. 114. Horwitz MB, Pride GL, Graybeal DF, Purdy PD. Percutaneous transluminal angioplasty and stenting of midbasilar stenoses: three technical case reports and literature review. Neurosurgery 1999; 45: 925–931.
Biomaterials and Stent Technology
129
115. Roubin GS, King SB, Douglas JS, Lembo NJ, Robinson KA. Intracoronary stenting during percutaneous transluminal coronary angioplasty. Circulation– 1990; 81. 116. de Jaegere PP, de Feyter PJ, van der Giessen WJ, Serruys PW. Endovascular stents: preliminary clinical results and future developments. Clin. Cardiol 1993; 16:369–378. 117. Hirshfeld JW, Schwartz JS, Jugo R, MacDonald RG, Goldberg S, Savage MP, Bass TA, Vetrovec G, Cowley M, Taussig AS. Restenosis after coronary angioplasty: a multivariate statistical model to relate lesion and procedure variables to restenosis. The M-HEART Investigators. J. Am. Coll. Cardiol 1991; 18:647–656. 118. Becker GJ. Intravascular stents. General principles and status of lower-extremity arterial applications. Circulation 1991; 83:I122–I136. 119. Lambert T, Dev V, Litvack F, Forrester JS, Eigler NL. Localized arterial drug delivery from a polymer coated removable metallic stent: kinetics and bioactivity of Forskolin. Circulation 1993; 88(Suppl 1):I–310. 120. Strecker EP, Liermann D, Barth KH, Wolf HR, Freudenberg N, Berg G, Westphal M, Tsikuras P, Savin M, Schneider B. Expandable tubular stents for treatment of arterial occlusive diseases: experimental and clinical results. Work in progress. Radiology 1990; 175:97–102. 121. Buchwald AB, Unterberg C, Nebendahl K, Grone HJ, Wiegand V. Low-molecular-weight heparin reduces neointimal proliferation after coronary stent implantation in hypercholesterolemic minipigs. Circulation 1992; 86:531–537. 122. Itoh A, Hall P, Maiello L, Di Mario C, Moussa I, Blengino S, Ferraro M, Martini G, Di Francesco L, Finci L, Colombo A. Intracoronary stent implantation in native coronary arteries and saphenous vein grafts: a consecutive experience with six types of stents without prolonged anticoagulation. Mayo Clin. Proc 1997; 72:101–111. 123. Schrader SC, Beyar R. Evaluation of the compressive mechanical properties of endoluminal metal stents. Cath. Cardiovasc. Diag 1998; 44:179–187. 124. van der Giessen WJ, Strauss BH, van Beusekom HM, van Loon H, van Woerkens LJ. Self-expandable mesh stents: an experimental study comparing polymer coated and uncoated Wallstent stents in the coronary circulation of pigs. Circulation 1990; 82:III–542. 125. www.angiotech.com/investor_relations/2002_press_releases/020523a.php. 126. Susawa TK, Shiraki K, Shimizu Y. Biodegradable intracoronary stents in adult dogs. J Am. Coll. Cardiol 1993; 21(Suppl. 1):483A. 127. Holmes DR, Camrud AR, Jorgenson MA, Edwards WD, Schwartz RS. Polymeric stenting in the porcine coronary artery model: differential outcome of exogenous fibrin sleeves versus polyurethanecoated stents. J Am Coll Cardiol 1994; 24:525–531. 128. www.cookgroup.com/news/082101.html. 129. www.seas.upenn.edu/⬃seider/Rapamycin.pdf. 130. Stack RS, Califf RM, Phillips HR, Pryor DB, Quigley PJ, Bauman RP, Tcheng JE, Greenfield JC. Interventional cardiac catheterization at Duke Medical Center. Am. J. Cardiol 1988; 62:3F–24F. 131. Tenaglia AN, Buller CE, Kisslo KB, Stack RS, Davidson CJ. Mechanisms of balloon angioplasty and directional coronary atherectomy as assessed by intracoronary ultrasound. J. Am. Coll. Cardiol 1992; 20:685–691. 132. Murphy JG, Schwartz RS, Edwards WD, Camrud AR, Vlietstra RE, Holmes DR. Percutaneous polymeric stents in porcine coronary arteries. Initial experience with polyethylene terephthalate stents. Circulation 1992; 86:1596–1604. 133. van Beusekom HM, van der Giessen WJ, van Ingen Schenau D, Slager CJ. Synthetic polymers as an alternative to metal in stents? In vivo and mechanical behaviour of polyethylene-terephthalate. Circulation 1992; 86(Suppl 1):I–731. 134. Lincoff AM, van der Giessen WJ, Schwartz RS, van Beusekom HM, Serruys PW, Holmes DRJ, Ellis SG, Topol EJ. Biodegradable and biostable polymers may both cause vigorous inflammatory responses when implanted in the porcine coronary artery. J. Am. Coll. Cardiol 1993; 21:179A. 135. Schwartz RS, Holmes DR, Topol EJ. The restenosis paradigm revisited: an alternative proposal for cellular mechanisms. J. Am. Coll. Cardiol 1992; 20:1284–1293. 136. Susawa T, Shiraki K, Shimizu Y. Biodegradable intracoronary stents in adult dogs. J. Am. Coll. Cardiol 1993; 21:483A.
130
Nguyen et al.
137. Ye YW, Landau C, Meidell RS, Willard JE, Moskowitz A, Aziz S, Carlisle E, Nelson K, Eberhart RC. Improved bioresorbable microporous intravascular stents for gene therapy. J. ASAIO 1996; 42:M823–M827. 138. Sonobe T, Naganuma S, Yambe T, Kobayashi SI, Sizuka K, Katahira Y, Nitta K, Nitta SI. Development of intracoronary local adhesive delivery technique. Int. J. Artif. Organs 1997; 20:319–326. 139. Bier JD, Zalesky P, Li ST, Sasken H, Williams DO. A new bioabsorbable intravascular stent: in vitro assessment of hemodynamic and morphometric characteristics. J. Interv. Cardiol 1992; 5: 187–194. 140. Swierkosz TA, Kapoor S, Tardiff DC, Hirshfeld JW, Klugherz BD, Kolansky DM, Magness K, Valettas N, Wilensky RL, Herrmann HC. Does greater platelet inhibition explain Abciximab’s superiority in the TARGET trial? A randomized clinical comparison. Circulation 2001; 104:II–385. 141. Montalescot G, Barragan P, Wittemberg O, Pinton P, Elhadad S, Deboisgelin X, Akesbi A, Jouve B, Amor M, Funck F. Abciximab reduces clinical restenosis in diabetic patients undergoing primary stenting for acute myocardial infarction (the ADMIRAL study). Circulation 2001; 104:II–386. 142. Amoroso G, van Boven AJ, van Veldhuisen DJ, Tio RA, Balje-Volkers CP, Petronio AS, van Oeveren W. Eptifibatide and abciximab exhibit equivalent antiplatelet efficacy in an experimental model of stenting in both healthy volunteers and patients with coronary artery disease. J. Cardiovasc. Pharmacol 2001; 38:633–641. 143. Bhatt DL, Lincoff AM. Combined use of Eptifibatide and Enoxaparin in patients undergoing percutaneous coronary intervention: the results of the CRUISE trial. Circulation 2001; 104:II–384. 144. Rebeiz AG, Pieper KS, O’Shea JC, Gilchrist IC, Chandler B, Slater J, Muhlestein JB, Loenz TJ, Joseph D, Kitt MM, Tcheng JE. 16- to 18-hour infusions of Eptifibatide seem optimal in reducing ischemic complications following percutaneous coronary intervention. Circulation 2001; 104: II–386. 145. Grube E, Silber SM, Hauptmann KE. Taxus I: Prospective, randomized, double-blind comparison of NIRxTM stents coated with Paclitaxel in a polymer carrier in de-novo coronary lesions compared with uncoated controls. Circulation 2001; 104:II–463. 146. Park S, Shim WH, Ho DS, Raizner AE, Park SW, Kim JJ, Hong MK, Lee CW, Cho SY, Jang YS, Choi DH, Lau CP, Lam R, Wang Y. The clincal effectiveness of Paclitaxel-coated coronary stents for the reduction of resenosis in the ASPECT trial. Circulation 2001; 104:II–464. 147. Huang Y, Wang L, Vermeire I, Verbeken E, Schacht E, Scheerder ID. Methylprednisolone coated stents decrease neointimal hyperplasia in a porcine coronary model. Circulation 2001; 104:II–665. 148. Moses JW, Leon MB, Popma JJ, Kuntz RE. The U.S. multicenter, randomized, double-blind study of the Sirolimus-eluting stent in coronary lesions: Early (30-day) safety results. Circulation 2001; 104:II–464. 149. Sousa JE, Morice M, Serruys PW, Fajadet J, Perin M, Hayashi EB, Colombo A, Schuler G, Barragan P, Bode C. The RAVEL study: a randomized study with the Sirolimus coated BX velocity balloonexpandable stent in the treatment of patients with de novo native coronary artery lesions. Circulation 2001; 104:II–463.
6 Long-Term Evaluation of a Novel Tissue Adhesive (BioGlue) for Use in Surgery Nancy Perlman, Charles W. Hewitt, Steven W. Marra, Jean-Luc V. Tran, Jonathan H. Cilley, Vincent A. Simonetti, and Anthony J. DelRossi University of Medicine and Dentistry of New Jersey, and Cooper Hospital/University Medical Center Camden, New Jersey, U.S.A.
Steve D. Lenz Auburn University of Medicine Auburn, Alabama, U.S.A.
K. Umit Yuksel CryoLife Inc. Kennesaw, Georgia, U.S.A.
I. INTRODUCTION Thoracic aortic surgery for the repair of aneurysms, tissue injury, and other disorders can result in a disseminated intravascular coagulopathy. This coagulopathy is caused by a dysfunction of the normal coagulation cascade. Factors that lead to this dysfunction include aortic cross-clamping, hypothermia, multiple blood transfusions, acidosis, and excessive hemorrhaging secondary to rupture or traumatic injury [1–4]. Additionally, a large percentage of the patients undergoing thoracic aortic surgery are taking aspirin, NSAIDS, and/or other anticoagulants for preexisting medical conditions. These medications further predispose the patient to excessive bleeding and therefore increase the risk for the occurrence of a coagulopathy during surgical repair. Additionally, another source of bleeding in these patients is through the synthetic graft used for the aortic repair. These grafts are not self- sealing; therefore, bleeding from the creation of needle holes and suturing inevitably occurs. This dilemma cannot be corrected with additional suturing, as each new suture creates an additional needle hole, thus exacerbating the bleeding [5–7]. Efficacy of various surgical tissue adhesives has been investigated in an attempt to resolve this problem. We studied a new tissue adhesive, BioGlue威 Surgical Adhesive (CryoLife, Inc., Kennesaw, GA), for the control of needle hole bleeding and bleeding from anastamoses of native thoracic aorta to synthetic bypass grafts in the presence of coagulopathy using a novel sheep model. We have published the use of this unique coagulopathic sheep model previously [8,9]. We hypothesized that the induction of coagulopathy in sheep would model clinical needle hole bleeding and surgical bleeding from synthetic graft anastamoses, and that this new tissue bioadhesive would control intra- and postoperative blood loss for surgical repair of the thoracic aorta. 131
132
Perlman et al.
The BioGlue Surgical adhesive consists of two components: glutaraldehyde, a bifunctional aldehyde, and albumin, which contain the ε-amino groups of lysine residues. These components are kept separate until the time of application. The mechanism of action for BioGlue is based on the well-known chemical reaction between aldehydes and amines. Glutaraldehyde bridges the ε-amino groups of lysine residues of the albumin to the extracellular matrix proteins in the tissue creating a covalent bond between the tissues and the adhesive.
II. MATERIALS AND METHODS All experiments were performed under a research protocol approved by the University of Medicine and Dentistry of New Jersey/Robert Wood Johnson Medical School, Camden Institutional Animal Care and Use Committee (IACUC). All animals received humane care in compliance with the ‘‘Guide for Care and Use of Laboratory Animals’’ published by the National Institutes of Health (NIH publication 85–23, revised 1985). A. Operative Technique Our method of inducing systemic coagulopathy in a sheep model has been previously published in detail [8,9]. Sheep were given a 600-mg aspirin suppository (Raddok Laboratories, Minneapolis, MN) daily for two days prior to surgery, with one further dose on call to surgery. The animals were endotracheally intubated and put under general anesthesia. Intravenous doses of gentamycin (2 mg/kg, Elkins-Sinn, Inc Cherry Hill, NJ) and ampicillin sodium (7 mg/kg, Bristol Myers Squibb Princeton, NJ) were administered. A preoperative coagulation profile was obtained, which included activated clotting time (ACT), prothrombin time (PT), activated partial thromboplastin time (aPTT), fibrinogen, and complete blood count (CBC). The descending thoracic aorta was dissected and exposed via a left lateral thoracotomy. An intravenous bolus of unfractionated heparin sodium (400 IU/kg, Elkins-Sinn) was administered, and after 5 min the descending aorta was partially cross-clamped both distal to the subclavian artery and supradiaphragmatically, using atraumatic partially occluding vascular clamps. A woven, gelatin-impregnated synthetic tube graft (Gelweave威 [15-mm diameter], Sulzer Vascutek, Austin, TX) was then anastamosed end-to-side to the descending aorta utilizing a clampand-sew technique [10,11], obviating the need for cardiopulmonary bypass. Intravenous heparin boluses (400 IU/kg) were repeated every 30 min following the initial dose, and additional coagulation profiles were obtained. Following completion of the anastamoses, the aortic crossclamps were released, the native aorta was ligated, and the bypass established. After ensuring adequate hemostasis, a chest tube was inserted through a separate stab incision. Once adequate hemostasis was achieved, the thoracotomy incision was closed with interrupted sutures and the chest tube was connected to a graded receptacle (Atrium, Hudson, NH) with continuous suction applied for hourly determination of blood output. B. Experimental Design Adult sheep weighing 35–65 kg were utilized. Experimental anastamoses in the sheep (n ⳱ 9, EXP) were treated with BioGlue. Surgicel威 (Ethicon, Somerville, NJ) was utilized in control animals (n ⳱ 5, CON). All animals received a synthetic tube bypass graft anastamosed to the descending thoracic aorta using a clamp-and-sew technique as detailed above. Intra- and postoperative blood loss were recorded and compared between the two groups. Postoperative bleeding was measured by hourly chest tube output. Statistics were analyzed by the Student’s
A Novel Tissue Adhesive (BioGlue) in Surgery
133
t test on normal or log-transformed data or using the posthypothesis Mann–Whitney rank sum test, as appropriate. Comparisons were considered significant if p ⱕ 0.05. Data were presented as either mean or median Ⳳ standard error of the mean or median (SEM). C. BioGlue Surgical Adhesive BioGlue was supplied in a carton containing the delivery pack and separate adhesive solutions. The adhesive solutions were composed of two separate vials containing 9 mL of 45% bovine serum albumin solution and 2.5 mL of 10% glutaraldehyde solution, respectively. The delivery device was handled much like a pistol to apply the adhesive to surgical tissues. The solutions were mixed together as they traveled through a tortuous pathway through the applicator tip prior to direct application onto dry surgical tissues (Fig. 1).
D. Post-Operative Care Once awake, the sheep were extubated and all intravascular lines were removed. The chest tube was removed no less than 6 h postoperatively when the ACT had returned to baseline, drainage had fallen to an acceptable level, and there was no air leak with the chest tube to water-seal. For postoperative analgesia, meperidine (2–6 mg/kg IM, every 2–3 h, Wyeth Laboratories, Philadelphia, PA) and buprenorpherine (0.005 mg/kg IM, every 12 h, Rickette and Coleman, Richmond, VA) were given for the first 3–4 days. Following this period, narcotics were withdrawn unless sheep demonstrated signs of discomfort or pain. All animals received a 10-day postoperative course of ampicillin (7 mg/kg, SQ) and gentamycin (2 mg/kg, SQ), twice daily.
Figure 1 Application of BioGlue to synthetic aortic graft; gross view.
134
Perlman et al.
Figure 2 Mean perioperative activated partial thromboplastin time.
Any animal that demonstrated excessive perioperative blood loss (⬎50% of total body fluid volume), postoperative paraplegia, or ischemic changes in the lower extremities was euthanized.
III. RESULTS A. Surgical Outcomes Detailed data concerning the use of BioGlue for thoracic aortic surgery under coagulopathic conditions has previously been reported [8,9]. In summary, these results showed that the surviving experimental animals were found to have no major thrombotic, ischemic, or intracerebral hemorrhagic complications. Also, rupture of the graft and/or exsanguinations did not occur in any animal. There were no significant differences between the CON and EXP groups with respect to mean perioperative fibrinogen, platelet counts, PT, aPTT, or ACT levels (Figs 2–6). The mean cross-clamp times were not significantly different between EXP and CON groups (26.0 Ⳳ 4.9 min vs. 26.1 Ⳳ 3.6 min, p ⬎ 0.9). Total surgical bleeding was significantly and dramatically reduced in the experimental group compared to controls. Bleeding was determined by chest tube output recorded at hourly time points. The data can be broken down into four separate categories: intraoperative total,
Figure 3 Mean perioperative activated clotting time.
A Novel Tissue Adhesive (BioGlue) in Surgery
Figure 4 Mean perioperative fibrinogen levels.
Figure 5 Mean perioperative platelet levels.
Figure 6 Mean perioperative prothrombin time.
135
136
Perlman et al.
intraoperative after cross-clamp release, postoperative, and total. Intraoperative blood loss was reduced in the experimental group compared to the control group (CON mean 422 mL vs. EXP mean 329 mL), though this was not significant (p ⳱ 0.3). The amount of blood loss after aortic cross-clamp release was also less in the EXP group (CON mean 431 Ⳳ 164 mL vs. EXP mean 241 Ⳳ 100 mL, p⳱0.3). This also was not statistically significant. Both postoperative and total blood loss were found to be significantly reduced in the EXP group (postoperative: CON mean 955 mL vs. EXP mean 470 mL, p ⬍ 0.003; total: CON mean 1377 mL vs. EXP mean 799 mL, p ⬍ 0.008) (Figure 7). B. Histopathology Animals were sacrificed at various times, ranging from 6 h (for humane reasons) to greater than 3 months postsurgery. At the time of sacrifice, excision of the native aorta and synthetic graft was performed. BioGlue deposits were identified in all EXP animals. Grossly, these deposits were solitary and variable in size. Upon closer examination, they were dark brown, firm, and pliable. Microscopic evaluation of hematoxylin- and eosin-stained sections was performed. BioGlue deposits were identified in all EXP animals. Deposits were composed of a homogeneous, eosinophilic material surrounded by a thin discontinuous zone of mature, fibrous connective tissue. In the tissue explanted after 3 months the histologic examination was strikingly notable for a relative paucity of an inflammatory response (Fig. 8). Chronic granulomatous inflammation was occasionally observed. At this time there was little or no fibrotic response, and multinucleated giant cells were not seen in any of the samples. The overall acute inflammatory response was minimal and inconsistent as well. In a few instances, it was observed that there was tissue necrosis immediately adjacent to BioGlue deposits. There was an increase in fatty connective tissue in the tunica adventitia of certain specimens. It was evident that the BioGlue bonded firmly and was adherent to both surgical tissues and the synthetic graft material (Fig. 9). In occasional specimens, loss of elastin fibers and degeneration of the tunica media was observed. The relationship of this finding to the presence of BioGlue was not clear. IV. DISCUSSION Thoracic aortic surgical repair of aneurysms, trauma, or congenital anomalies are often associated with increased risks of morbidity and mortality predominantly due to excessive and uncontrol-
Figure 7 Mean perioperative blood loss.
A Novel Tissue Adhesive (BioGlue) in Surgery
137
Figure 8 Hematoxylin- and eosin-stained specimen demonstrating paucity of inflammatory response in Response to BioGlue adjacent to aortic adventitia. This sample was explanted 3 months after operation (H&E 200⳯).
lable bleeding because of disseminated intravascular coagulopathy. This sheep model reliably and consistently simulated the coagulopathic state that is frequently observed in the clinical setting that is well known to occur [8,9]. The success of bioadhesives for the repair of thoracic aortic dissection is well established. Nguyen reported that for acute type A aortic dissection, open repair using a gelatin-resorcinformadehyde (GRF) glue effect repair substantially decreases the incidence of reoperation and persistent false-lumen perfusion [13]. Others reported their 20-year experience with the GRF glue, with results showing greater ease of initial emergency surgery for acute type A aortic dissection and safety as a beneficial influence on late results [14]. Besides application in aortic surgery, Nomori has reported that a gelatin-resorcinol-formaldehyde- glutaraldyhyde (GRFG) glue seals air leaks during thorascopic procedures [15,16].
Figure 9 Bonding of BioGlue to synthetic aortic graft material, approximately 3.5 h after operation (H& E 200⳯).
138
Perlman et al.
The major concern of the GRF glue involves toxicity due to the formaldehyde component [17]. Ennker and associates reported experimental use of a similar glue; the formaldehyde component was replaced with pentanedial and ethanedial in order to decrease the formaldehyde-associated toxicity [8]. Of note, despite widely established use in Europe, a comparative study of biological glues in dogs showed the GRF glue resulted in inferior clot-forming characteristics than both cryoprecipitate and two-component fibrin sealant [19]. In the areas of abdominal and cardiothoracic procedures, the use of various other surgical bioadhesives has been widely accepted. Of all these available products, the most widely used has been two-component fibrin glue [20,31]. A sutureless technique using an ethyl-1–2-cyanoacrylate monomer–based synthetic glue in conjunction with a Teflon patch has been reported by Lijoi to treat left ventricular free wall rupture following acute myocardial infarction [32]. With the ever-increasing concern of potential toxicity of formaldehyde in the GRF glue promoted in Europe, BioGlue Surgical Adhesive represents a new entry into the surgical armamentarium. When compared to oxidized regenerated cellulose hemostats (i.e., Surgicel姟) the use of BioGlue in our coagulopathic sheep model effectively reduced the time needed to achieve hemostasis, the bleeding rate, total blood loss from suture lines, and the number of needle holes to secure the synthetic graft material to the native thoracic aorta. The mechanism of action which BioGlue, i.e. protein cross-linking, is unique among other glue products and it has demonstrated superior intra- and postoperative hemostasis in our model. The surgeon today is faced with an aging, more debilitated population of patients that are more prone to bleeding and its associated complications. BioGlue, as we have demonstrated in our model, should prove to be very beneficial in reducing bleeding-associated morbidity and mortality in these difficult patients as well as others with medically associated bleeding concerns.
ACKNOWLEDGMENT This work was supported in part by CryoLife, Inc., Kennesaw, Georgia.
REFERENCES
1. Cohen JR, Angus L, Asher A, Chang JB, Wise L. Disseminated intravascular coagulation as a result of supraceliac clamping: implications for thoracoabdominal aneurysm repair. Ann Vasc Surg 1987; 1(5):552–557. 2. Crawford ES, Crawford JL, Safi HJ. Thoracoabdomimal aortic aneurysms: preoperative and intraoperative factors determining immediate and long–term results of operations in 605 patients. J Vasc Surg 1986; 3(3):389–404. 3. Mattox HJ, Mattox KL, Moore KL, Moore EE, Felicano EE, Felicano DV, eds. Trauma. East Norwalk, CT: Appleton & Lange, 1988:809–819. 4. Moylan JA, ed. Surgical Critical Care. St. Louis, MO: Mosby Year Book Inc., 1994:291–324. 5. Quint LE, Francis IR, Williams DM. Synthetic interposition grafts of the thoracic aorta: postoperative appearance on serial CT studies. Radiology 1999; 211(2):317–324. 6. Jeschke MG, Hermanutz V, Wolf SE, Koveker GB. Polyurethane vascular prostheses decrease neointimal formation compared with expanded polytetrafluoroethylene. J Vasc Surg 1999; 9(1):16–76. 7. Walpoth BH, Rogulenko R, Tikhvinskaia E. Improvement of patency rate in heparin–coated small synthetic vascular grafts. Circulation 1998; 98(9 Suppl):II319–II323. 8. Tran HS, Chrzanowski FA, Puc MM. A sheep model for thoracic aortic surgery in the presence of systemic coagulopathy. J. Invest. Surg 2000; 13:111–116.
A Novel Tissue Adhesive (BioGlue) in Surgery
139
9. Hewitt CW, Marra SW, Kann BR. BioGlue surgical adhesive for thoracic aortic repair during coagulopathy: Efficacy and histopathology. Ann Thorac Surg 2001; 71:1609–1612. 10. DelRossi AJ, Cernaianu AC, Madden LD. Traumatic disruptions of the thoracic aorta: treatment and outcome. Surgery 1990; 108(5):864–870. 11. Puc MM, Marra SW, Tran HS. A novel technique in a sheep model for evaluating prosthetic heart valve performance. 2001; 14(1):55–61. 12. Chen K, Varon J, Wenker OC. Acute thoracic aortic dissection: the basics. J Emerg Med 1997; 15(6): 859–867. 13. Nguyen B, Muller M, Kipfer B. Different techniques of distal aortic repair in acute type A dissection: impact on late aortic morphology and reoperation. Eur J Cardiothorac Surg 1999; 15(4):496–500. 14. Bachet J, Goudot B, Dreyfus G. Surgical treatment of acute dissection of the ascending aorta (20 years experience). Chirurgie 1998; 123(3):229–237. 15. Nomori H, Horio H. Gelatin–resorcinol–formaldehyde–glutaraldehyde glue-spread stapler prevents air leakage from the lung. Ann Thorac Surg 1997; 63(2):352–355. 16. Nomori H, Horio H, Morinaga S, Suemasu K. Gelatin–resorcinol-formaldehyde–glutaraldehyde glue for sealing pulmonary air leaks during thoracoscopic operation. Ann Thorac Surg 1999; 67(1): 212–216. 17. Fukunaga S, Karck M, Harringer W. The use of gelatin–resorcin–formalin glue in acute aortic dissection type A. Eur J Cardiothorac Surg 1999; 15(5):564–569. 18. Ennker IC, Ennker J, Schoon D. Formaldehyde-free collagen glue in experimental lung gluing. Ann Thorac Surg 1994; 57(6):1622–1627. 19. Basu S, Marini CP, Bauman FG. Comparative study of biological glues: cryoprecipitate glue, twocomponent fibrin sealant, and ‘‘French’’ glue. Ann Thorac Surg 1995; 60(5):1255–1262. 20. Koveker G, de Vivie ER, Hellberg KD. Clinical experience with fibrin glue in cardiac surgery. Thorac Cardiovasc Surg 1981; 29(5):287–289. 21. Thetter O. Fibrin adhesive and its application in thoracic surgery. Thorac Cardiovasc Surg 1981; 29(5):290–292. 22. Wolner E. Fibrin gluing in cardiovascular surgery. Thorac Cardiovasc Surg 1982; 30(4):236–237. 23. Rousou JA, Engelman RM, Breyer RH. Fibrin glue: an effective hemostatic agent for nonsuturable intraoperative bleeding. Ann Thorac Surg 1984; 38(4):409–410. 24. Dresdale A, Malm JR. Hemostatic effectiveness of fibrin glue derived from single-donor fresh frozen plasma. Ann Thorac Surg 1985; 40(4):385–387. 25. Lupinetti FM, Stoney WS, Alford WC. Cryoprecipitate-topical thrombin glue. Initial experience in patients undergoing cardiac operations. J Thorac Cardiovasc Surg 1985; 90(4):502–505. 26. Spotnitz WD, Dalton MS, Baker JW, Nolan SP. Reduction of perioperative hemorrhage by anterior mediastinal spray application of fibrin glue during cardiac operations. Ann Thorac Surg 1987; 44(5): 529–531. 27. Rousou J, Levitsky S, Gonzalez–Lavin L. Randomized clinical trial of fibrin sealant in patients undergoing resternotomy or reoperation after cardiac operations. A multicenter study. J Thorac Cardiovasc Surg 1989; 97(2):194–203. 28. Grunenwald D. Intraoperative use of fibrin sealant in pulmonary surgery. A prospective study on a series of 124 procedures. Ann Chir 1989; 43(2):147–150. 29. Garcia–Rinaldi R, Simmons P, Salcedo V, Howland C. A technique for spot application of fibrin glue during open heart operations. Ann Thorac Surg 1989; 47(1):59–61. 30. Kjaergard HK, Fairbrother JE. Controlled clinical studies of fibrin sealant in cardiothoracic surgery—a review. Eur J Cardiothorac Surg 1996; 10(9):727–733. 31. Spotnitz WD. As originally published in 1990: Four years’ experience with fibrin sealant in thoracic and cardiovascular surgery. Ann Thorac Surg Updated in 1998; 65(2):592–593. 32. Lijoi A, Scarano F, Parodi E. Subacute left ventricular free wall rupure complicating acute myocardial infarction. Successful surgical repair with a sutureless technique. J Cardiovasc Surg (Torino) 1996; 37(6):627–630. 33. Schalm OW, Jain NC, Carrol EJ. Veterinary Hematology. 3rd ed. Philadelphia, PA: Lea & Febiger, 1975.
7 Bioengineered Skin Reconstituted from Cultured Epidermis and Cryopreserved Dermis: Converting Cryopreserved Skin Allografts into a Permanent Skin Substitute Hannah Ben-Bassat Hadassah University Hospital Jerusalem, Israel
I. INTRODUCTION A. Skin Substitutes 1. Preservation of Skin Allografts for Application in Burn Therapy, Skin Disorders, and Plastic and Reconstructive Surgery One of the fundamental problems in burn therapy is accomplishing a permanent replacement of skin in full and partial thickness burns. The most superior graft used in the treatment of burn wounds is the autograft [1]. However when autografts may not be available, other tissue sources have to be sought [2,3]. Fresh cadaver allografts are still considered the ‘‘gold standard’’ skin substitute for coverage of large, full-thickness burns [4–5]. Unfortunately the use of fresh allografts is severely impeded by their inadequate availability. Human skin stored at 40⬚C, in appropriate medium (McCoy’s 5) maintains its efficacy for several weeks, but viability and take decrease with time [6]. Thus skin graft preservation for the purpose of delayed application is still a basic tool in burn treatment and plastic and reconstructive surgery [7,8]). Indeed extensive research has resulted in improved methods of processing, storage, and evaluation of preserved skin [9–10]. Skin banks differ in their mode of operation and techniques for preservation and storage. Presently, it is well recognized by the American Association of Tissue Banks (AATB) that the viability of skin is an essential prerequisite for the functional closure of wounds [5–11]. Preservation of skin by programmed freezing at a controlled rate (1⬚C/min), followed by storage at low temperature, has been shown to protect the structure of skin and to maintain a level of skin metabolism equal to about 80% of that present in fresh skin [12]. However, in the Netherlands and to some extent in other Western European countries the glycerolization method has enjoed popularity [13]. The principal indications for glycerol preserved allografts (GPA) are scald burns and an overlay for wide meshed autograft in extensive full-thickness burns [13]. This nonviable skin is considered less rejectable. Its performance is satisfactory when it is combined with autografts widely expended and meshed either 1:6 or 1:9 with the micrograft 141
142
Ben-Bassat
technique [14]. With most procedures the results showed consistent graft take, resulting in ⬎95% epithelialization at 6 weeks and less than 2% graft failure [13]. It has also been reported that allografts preserved in this way function satisfactory in providing allodermal bed for cultured epidermal autografts [15]. Recent results have compared the in vitro immunogenicity of glycerolized skin to untreated skin by coculture of peripheral T-cells with allogeneic treated skin cells. The results indicated that the observed inflammatory process is mediated by infiltrating host monocytes, rather than by a rejection process mediated by T-cells [16]. As the demand for skin allografts has increased, the responsibility for processing, storage, and evaluation of graft performance of preserved skin has become an important issue of banking organizations. The Israel National Skin Bank (INSB) cryopreserves cadaver skin allografts by programmed freezing and storage in liquid nitrogen [17]. A mouse recipient model developed by us has been applied to evaluate the transplantation performance of preserved skin [18]. Briefly, human preserved skin is grafted on Balb/c mice, and primary take evaluated by gross observation and predetermined histology criteria after 7 days. This model assesses graft performance before the rejection process takes place. Its in vivo design has inherent clinical relevance that is especially appealing, since it attempts to meet the demand that the ultimate criterion for effective skin preservation is its transplantation performance. With this model we demonstrated that skin grafts preserved by programmed freezing, and stored in liquid nitrogen performed better than did glycerolized skin. However, both methods of preservation provided a less successful product then did viable fresh cadaver skin [18]. The quality of the final product, i.e., the cryopreserved skin is determined by many factors that quite often interact. Highly significant is an interaction factor that represents the combined effect of sample variability with the method of cryopreservation or with the storage period. Nevertheless cadaver skin preserved by programmed freezing at a controlled rate (1⬚C/min) and stored for up to 10 months in liquid nitrogen (ⳮ130 to ⳮ150⬚C) maintained adequate graft performance [19]. 2. How Long Can Cryopreserved Skin Allografts Be Stored? To address the issue of how long cryopreserved cadaver skin can be stored to maintain adequate graft performance, we applied our in vivo model and ‘‘first take’’ analysis to evaluate transplantation performance of skin banked for 5, 6, and 7 years. We presented first evidence that skin allografts cryopreserved for up to 5 years maintain adequate graft performance [17]. Further, graft performance of cryopreserved skin decreased with time, as reflected in the lower percent age of samples with high scores of separate histology criteria after prolonged storage. Nevertheless, paired comparison analysis between cryopreserved and fresh skin indicated that this decrease was not significant for storage of 5 years, whereas it was highly significant for 6 years of storage. Linear regression analysis indicated that there was no correlation between the score of the histology criteria and storage period for up to 65 months (⬍6 years), in line with the paired comparison analysis [17]. This information is of value to storage policies of skin banks abroad and to the Israel National Skin Bank, which is the largest in the world because of Israel’s unique security problems [17]. II. RESULTS FROM BIOENGINEERED FUNCTIONAL SKIN SUBSTITUTES A. Dermis from ‘‘Outdated’’ Cryopreserved Allografts for Bioengineered Reconstituted Skin A major field of biotechnology is the of bioengineering of tissues and organs to repair the human body. The loss or failure of an organ/tissue is one of the major problems in human health care [20]. The development of functional substitutes for damaged skin is an important issue in wound healing
Bioengineered Skin from Cryopreserved Allografts
143
and burn care [21–23]. Indeed, there are already tissue engineered skin products that could eventually replace the use of allografts when they because safe, effective, and not rejected [1,22,24,25]. Based on our previous and present results we raised the possibility of examining ‘‘outdated’’ cryopreserved skin (which has a less satisfactory graft performance) as a source of dermis for tissue engineered/reconstituted skin [17–19]. We have previously reconstituted skin with dermis separated from cryopreserved skin stored for up to 1 year with cultured epidermis and demonstrated its efficacy in transplantation performance in vivo [26]. The dermis was separated from cryopreserved skin either by short and mild trypsinization or by prolonged incubation in medium. The two products were compared with fresh dermis obtained by the same methods. All types of dermis were shown to retain normal ultrastructure and topographic organization, as detected by scanning and transmission electron microscopy and immunofluorescence analysis (Figs. 1–3, respectively). The dermis prepared by both methods was acellular; when cultured with no keratinocytes on top for 3 weeks, it showed neither residual epidermal cells nor epidermalization. However, the cell layers in fresh skin were more firmly attached, mechanical separation was more difficult, and residual epidermis often remained attached to the dermis. Keratinocytes cultured on trypsinized dermis attached better, began replication earlier, and generally reached higher cell numbers than those cultured on medium-treated dermis. Thereafter, the transplantation performance of several modifications in the reconstitution and grafting procedures of these composite skin grafts were evaluated in athymic mice: cultured epidermis combined onto trypsinized or medium-treated whole and meshed dermis, dermis pregrafted and allowed to take before transplanting epidermis on top, and keratinocytes grown into multiple epithelia on top of trypsinized meshed or whole dermis prior to grafting. The best grafting results were obtained with the ‘‘instant reconstituted skin,’’ namely keratinocytes grown in vitro to multiple epithelia that were combined onto trypsinized meshed dermis just before grafting. The transplantation performance of this modification was significantly better than that of all the other modifications, including that of keratinocytes initially grown into multiple epithelia on trypsinized dermis prior to grafting [26].
Figure 1 Surface appearance of human dermis prepared from cryopreserved skin by trypsinization: (A) 480 ⳯; (B) 5400 ⳯.
144
Ben-Bassat
Figure 2 Transmission electron micrograph of dermis showing intact lamina densa and collagen fibrils along the edge prepared from cryopreserved skin by trypsinization Original magnification (A) and (B) 13,000 ⳯.
Figure 3 Immunohistochemical staining of (A) collagen type IV and (B) laminin distribution along the surface of dermis prepared from cryopreserved skin by trypsinization, Original magnification, 200 ⳯.
Bioengineered Skin from Cryopreserved Allografts
145
B. Biochemical and Ultrastructural Features of Basement Membrane Components of Cryopreserved Human Skin The performance of currently bioengineered skin is still suboptimal, and its preparation and handling complex [24,25]. Thus the demand for skin allografts is increasing, and graft performance of preserved skin has become an important issue [5,10,21]. The ultimate criterion for effective skin preservation is its transplantation performance, evaluated by gross observation and by histology criteria; among these graft adherence to the wound bed and the integrity of the dermal–epidermal junction (DEJ) are of special clinical relevance [4,12,19]. The basement membrane (BM) is a highly specialized structure located between the epidermis and the dermis (also named dermal–epidermaljunctions) (Fig. 4) [27,28]. Both the epidermal keratinocytes and dermal fibroblasts contribute protein components to the BM zone [29,30]. The BM and the DEJ provide adhesion, restrict transit of molecules, permit cell passage, and support the epidermis. The BM influences the overall the integrity of the skin, such as keratinocyte behavior and their proliferation, migration, and differentiation [31,32]. Several classes of molecules are involved in cell–cell and cell–matrix interactions. Among these molecules a superfamily of transmembrane receptors=Mintegrins=Mplays a direct role, not only as a mechanical support but also in controlling keratinocyte growth and function [33–35]. Integrins are also involved in hemidesmosomes formation as well as in transducing intracellular signals [36–38] and are directly associated with transmembrane tyrosine kinases [35]. To evaluate biological and biochemical changes in the BM of human cryopreserved skin, we analyzed BM components by Western blot and immunochemistry methods. We examined the possibility of using Integrin 4 [35], laminin 5 [39–41], and collagen VII [42,43] as markers
Figure 4 Illustration of the proposed structure of the dermal– epidermal junction (DEJ). This model attempts to depict the relative locations of the molecules contained within the DEJ basement membrane and the interactions believed to occur between them. (From Ref. 27).
146
Ben-Bassat
Figure 5 Western blot analysis of the state of integrin 4 and laminin 5 in cryopreserved skin allografts stored for 4, 5, and 6 years.
because of their important role in the attachment of epidermis to dermis. Laminin5 (a heterodimer of laminin ␣3, 3, and 2 chains) is a laminin isoform which strongly promotes adhesion, migration and scattering of cells through binding to integrins ␣31, ␣61, and ␣64 [40] [44]. We hypothesized that changes in their expression might reflect and correlate with storage period, and further that their different location in the BM layers could suggest the separation site of the DEJ. Thus, skin samples cryopreserved and stored for 4, 5, and 6 years were analyzed for changes in the BM components. Western blot results showed that the amounts of integrin 4 and laminin 5 decreased significantly in cryopreserved skin in correlation with storage period, as indicated by box-plot statistical analysis (Fig. 5 and 6, respectively). Laminin 5 expression in the BM of skin stored for 4, 5, and 6 years was also detected by immunohistochemical analysis (Fig. 7). It is of interest to note that in cryopreserved skin stored for 6 years there were areas of detachment between the epidermis/ermis that seemed to occur between integrin 4 and laminin 5; a stronger staining was detected at the dermal region. The areas of detachment between the epidermis/dermis were also detected in the DEJ by TEM (transmission electron microscopy) (Fig. 8). The Western blot and immunohistochemical analyses for collagen VII were not clear
Figure 6 Box-plot analysis of integrin 4 and laminin 5 of cryopreserved skin allografts stored for 4, 5 and 6 years. Ten samples per storage period were Western blot analyzed.
Bioengineered Skin from Cryopreserved Allografts
147
Figure 7 Immunohistochemical staining for laminin 5 of skin allografts stored for (A) 4, (B) 5, and (C) 6 years, 100 ⳯).
enough (unsuitable/not strong enough antibody) and therefore inconclusive. The areas of detachment between the epidermis/dermis might suggest a possible mechanism for the decreased graft performance of cryopreserved skin after long storage (e.g., ⬎6 years). To this point, it should be emphasized again that sample variability among the skin allografts seems to be of major importance in determining the quality of cryopreserved skin [45].
Figure 8 Transmission electron micrograph of cryopreserved skin hemidesmosomes (structures that join the basal surface of keratinocytes to the underlying basal lamina): (A) Disrupted hemidesmosome from skin after long storage showing widening areas at the lamina lucida and fussiness with fewer and disappearing anchoring fibrils at the lamina densa. (B) Intact ‘‘normal’’ hemidesmosome showing dense, 100,000 ⳯.
148
Ben-Bassat
C. Keeping BM Components intact We raised the possibility of using ‘‘outdated’’ cryopreserved skin (with a less satisfactory graft performance) as a source for dermis of tissue engineered reconstituted skin [18,19,45]. Therefore, we examined several methods for separating dermis from cryopreserved skin without affecting the integrity of BM components. The experiments were undertaken in order to better understand the involvement and the role of BM components in defining BM integrity and transplantation performance of our bioengineered skin model. Western blot (WB) analysis using laminin 5 and integrin 4 as markers showed that the separation method to obtain dermis did affect BM components. After trypsinization at 36⬚C for 4 h WB analysis detected integrin 4 only in the epidermis and laminin 5 only in the dermis. Whereas after incubation in medium at 27⬚C for 6 days (until complete separation was obtained) integrin 4 was detected in the epidermis and in the dermis and laminin 5 only in the dermis (Fig. 9). D. Bioengineered Skin with Dermis from ‘‘Outdated’’ Cryopreserved Allografts and Keratinocytes Cultured on Top Various methods for composite skin grafts have been reported [46–48]. We reconstituted skin with dermis separated from ‘‘outdated’’ cryopreserved allografts (stored for ⬎6 years) and keratinocytes cultured on top and grown into multiple epithelia. In this set of experiments the dermis was obtained by trypsinization at 36⬚C for 4h. Previously we have shown that the dermis was acellular, since when cultured with no keratinocyres on top for 3 weeks neither residual epidermal cells nor epidermalization could be detected [26]. The keratinocytes were seeded onto the dermis and grown into multiple epithelia to reconstitute a bioengineered skin substitute (Fig. 10). A representative experiment shows samples of reconstituted skin examined after 3, 7, and 14 days in culture (Fig. 11). Keratinocyte growth was determined by WB analysis using  actin as a marker. The results indicated progressive increase in cell protein that suggests cell growth. WB analysis of laminin 5 and integrin 4 showed progressive increase of protein amounts, indicating functioning keratinocytes and formation of BM/DEJ components (Fig. 12). Ultrastruc-
Figure 9 The effect of various separation methods of dermis from epidermis on basement membrane (BM) components.
Bioengineered Skin from Cryopreserved Allografts
149
Figure 10 Electron microscopic appearance of keratinocytes cultured on dermis. (A) Keratinocytes have formed several cell layers. (B) Desmosomes between the cells with typical intracellular dense lines. (C) Basal cell layer, intact lamina densa beneath the cells, dermis with undisrupted collagen bundles. (Original magnification (A), 500 ⳯; (B), 16,000 ⳯; (C) 8300 ⳯.)
tural observation with TEM and SEM indicated that the cells had anchored, attached, and spread onto the dermis by 3 days after seeding (Fig. 13). With continued culture, the cells formed a multilayered sheet, looked healthy, retained their typical morphology, and exhibited good culture organization on the dermis (Fig. 13). The present results suggest a successfully bioengineered skin reconstituted with human cultured epidermis formed with keratinocytes plated on dermis obtained from cryopreserved skin (for ⬎6 years) concomitant with formation of BM components.
III. CONCLUSIONS The long-term goal of this research is to develop a procedure(s) for bioengineered skin to be used in burn therapy, skin disorders, and plastic and reconstructive surgery. We propose to reconstitute skin from dermis of ‘‘outdated’’ cryopreserved skin with keratinocytes grown into multiple epithelia on top, and evaluate its structure, function, and transplantation performance. The present results suggest a successfully bioengineered skin reconstituted with keratinocytes that formed epidermis on dermis separated in vitro from ‘‘outdated’’ cryopreserved skin. The plated keratinocytes attached and spread onto the dermis and with continued culture formed
150
Ben-Bassat
Figure 11 A representative experiment of keratinocytes cultured on dermis from skin cryopreserved for 6 years. (A) Western blot analysis demonstrating time course increase of integrin 4, laminin 5 (BM components), and  actin (a cellular protein) suggesting formation of basement membrane components and cell growth. (B) Histology of a sample of the dermis on demonstrating that the dermis is acellular and has normal microscopic appearance.
a multilayer sheet. The cells looked ‘‘healthy’’, retained their typical morphology, and exhibited good culture organization. In parallel we analyzed potential changes in the ultrastructural organization of BM components and DEJ after cryopreservation and prolonged storage. Specifically, the superfamily of transmembrane receptors=Mintegrins are considered important for graft adherence and hence transplantation performance of skin. We applied the ␣64 laminin receptor, which binds with a higher affinity to laminin 5 and is found both in vivo and in cultured stratified epithelia, and laminin 5 as markers. Indeed we identified structural and biochemical changes in basement
Bioengineered Skin from Cryopreserved Allografts
151
Figure 12 Relative amounts of integrin 4, laminin 5, and  actin of the representative experiment illustrated in Fig. 11. The results indicate that there were no detectable changes in the proteins of the control dermis (with no cultured keratinocytes on top) and confirm the time course increase of the examined proteins [integrin 4, laminin 5 (BM components), and  actin (a cellular protein)]
membrane components of cryopreserved skin, in correlation with storage, crucial for the attachment of the epidermis to the dermis and skin integrity. Successful reconstitution will make use of ‘‘outdated’’ cryopreserive skin unsuitable for grafting skin, thus converting cryopreserved allografts into permanent skin substitutes.
ACKNOWLEDGMENTS These studies were partially supported by contract 09343682–01 from the Ministry of Defense, LIBI—The Fund for Strengthening Israel Defense, and the MJF Foundation. The assistance, cooperation, and good will of M. Chaouat, M.Sc., is greatly appreciated.
152
Ben-Bassat
Figure 13 Human keratinocytes cultured on dermis cryopreserved for 6 years. (A) Anchored, attached, and well-spread cell, looking healthy, 3 days after plating. (B) Cell spread on the dermis displaying flat configuration and looking healthy; the ultrastructure of the dermis and the anchoring areas of the cell onto the dermis visualized, 17 days after plating. (C) Well-formed sheath of confluent keratinocytes, 17 days after plating (Original magnification by scanning electron microscopy for (A) 2200 ⳯; (B) 1500 ⳯; (C) 1100 ⳯.)
Bioengineered Skin from Cryopreserved Allografts
153
REFERENCES
1. Hansbrough JF, Franco ES. Skin replacements. Clin Plast Surg 1998; 25(3):407. 2. Ninnemann JL, Fisher JC, Frank HA. Clinical skin banking: a simplified system for processing, storage retrieval of human allografts. J Trauma 1978; 18(10):723. 3. May SR. The future of skin banking [editorial]. J Burn Care Rehabil 1990; 11(5):484. 4. Bondoc CC, Burke JF. Clinical experience with viable frozen human skin and a frozen skin bank. Ann Surg 1971; 174(3):371. 5. Greenleaf G, Hansbrough JF. Current trends in the use of allograft skin for patients with burns and reflections on the future of skin banking in the United States. J Burn Care Rehabil 1994; 15(5):428. 6. DeBono R, Rao GS, Berry RB. The survival of human skin stored by refrigeration at 4⬚C in McCoy’s 5A medium: does oxygenation of the medium improve storage time?. Plast Reconstr Surg 1998; 102(1):78. 7. Aggarwal SJ, Baxter CR, Diller KR. Cryopreservation of skin: an assessment of current clinical applicability. J Burn Care Rehabil 1985; 6(6):469. 8. Kearney JN, Wheldon LA, Gowland G. Cryopreservation of skin using a murine model: validation of a prognostic viability assay. Cryobiology 1990; 27(1):24. 9. Kalter ES, de By TM. Tissue banking programmes in Europe. Br Med Bull 1997; 53(4):798. 10. Kearney JN. Quality issues in skin banking: a review. Burns 1998; 24(4):299. 11. Baxter CR. Skin banking in the United States [editorial]. J Burn Care Rehabil 1985; 6(4):322. 12. Blondet R, Gibert-Thevenin MA, Pierre C, Ehrsam A. Skin preservation by programmed freezing. Br J Plast Surg 1982; 35(4):530. 13. Mackie DP. The Euro Skin Bank: development and application of glycerol-preserved allografts. J Burn Care Rehabil 1997; 18(1 Pt 2):S7. 14. Kreis RW, Vloemans AF, Hoekstra MJ, Mackie DP, Hermans RP. The use of non-viable glycerolpreserved cadaver skin combined with widely expanded autografts in the treatment of extensive thirddegree burns. J Trauma 1989; 29(1):51. 15. McKay I, Woodward B, Wood K, Navsaria HA, Hoekstra H, Green C. Reconstruction of human skin from glycerol-preserved allodermis and cultured keratinocyte sheets. Burns 1994; 20(Suppl 1): S19. 16. Richters CD, Hoekstra MJ, van Baare J, du Pont JS, Kamperdijk EW. Immunogenicity of glycerolpreserved human cadaver skin in vitro. J Burn Care Rehabil 1997; 18(3):228. 17. Ben-Bassat H, Chaouat M, Zumai E, Segal N, Cinamon U, Ron M, Wexler MR, Eldad A. The Israel National Skin Bank: quality assurance and graft performance of stored skin. Cell and Tissue Banking 2000; 1:303–312. 18. Cinamon U, Eldad A, Chaouat M, Wexler MR, Israeli A, Zagher U, Ben-Bassat H. A simplified testing system to evaluate performance after transplantation of human skin preserved in glycerol or in liquid nitrogen. J Burn Care Rehabil 1993; 14(4):435. 19. Ben-Bassat H, Strauss N, Ron M, Chaouat M, Breiterman S, Israeli A, Wexler MR, Eldad A. Transplantation performance of human skin cryopreserved by programmed or stepwise freezing and stored at ⳮ80⬚C or ⳮ180⬚C. J Burn Care Rehabil 1996; 17(5):421. 20. Malinin TI, Perry VP. A review of tissue and organ viability assay. Cryobiology 1967; 4(3):104. 21. Glaser V. U.S. skin replacement market could heat up [news]. Biotechnology (N Y) 1995; 13(9): 933. 22. Harris PA, Leigh IM, Navsaria HA. Pre-confluent keratinocyte grafting: the future for cultured skin replacements? [editorial]. Burns 1998; 24(7):591. 23. Gibran NS, Heimbach DM. Current status of burn wound pathophysiology. Clin Plast Surg 2000; 27(1):11. 24. Navsaria HA, Myers SR, Leigh IM, McKay IA. Culturing skin in vitro for wound therapy. Trends Biotechnol 1995; 13(3):91. 25. Pomahac B, Svensjo T, Yao F, Brown H, Eriksson E. Tissue engineering of skin. Crit Rev Oral Biol Med 1998; 9(3):333.
154
Ben-Bassat
26. Ben-Bassat H, Eldad A, Chaouat M, Livoff A, Ron N, Ne’eman Z, Wexler MR. Structural and functional evaluation of modifications in the composite skin graft: cryopreserved dermis and cultured keratinocytes. Plast Reconstr Surg 1992; 89(3):510. 27. Burgeson RE, Christia AM. The dermal=Nepidermal junction. Curr Opin Cell Biol 1997; 9(5):651. 28. Borradori L, Sonnenberg A. Structure and function of hemidesmosomes: more than simple adhesion complexes. J Invest Dermatol 1999; 112(4):411. 29. Briggaman RA. Epidermal–dermal interactions in adult skin. J Invest Dermatol 1982; 79(Suppl 1): 21s. 30. Smola H, Stark HJ, Thiekotter G, Mirancea N, Krieg T, Fusenig NE. Dynamics of basement membrane formation by keratinocyte=Nfibroblast interactions in organotypic skin culture. Exp Cell Res 1998; 239(2):399. 31. Timpl R. Macromolecular organization of basement membranes. Curr Opin Cell Biol 1996; 8(5): 618. 32. Jost M, Huggett TM, Kari C, Rodeck U. Matrix-independent survival of human keratinocytes through an EGF receptor/MAPK-kinase=Ndependent pathway. Mol Biol Cell 2001; 12(5):1519. 33. De Luca M, Pellegrini G, Zambruno G, Marchisio PC. Role of integrins in cell adhesion and polarity in normal keratinocytes and human skin pathologies. J Dermatol 1994; 21(11):821. 34. Marchisio PC, Trusolino L, De Luca M. Topography and biological role of integrins in human skin. Microsc Res Tech 1997; 38(4):353. 35. Biffo S, Sanvito F, Costa S, Preve L, Pignatelli R, Spinardi L, Marchisio PC. Isolation of a novel 4 integrin-binding protein (p27(BBP)) highly expressed in epithelial cells. J Biol Chem 1997; 272(48):30314. 36. Schaapveld RQ, Borradori L, Geerts D, van Leusden MR, Kuikman I, Nievers MG, Niessen CM, Steenbergen RD, Snijders PJ, Sonnenberg A. Hemidesmosome formation is initiated by the 4 integrin subunit, requires complex formation of 4 and HD1/plectin involves a direct interaction between 4 and the bullous pemphigoid antigen 180. J Cell Biol 1998; 142(1):271. 37. Jones JC, Hopkinson SB, Goldfinger LE. Structure and assembly of hemidesmosomes. Bioessays 1998; 20(6):488. 38. Nievers MG, Schaapveld RQ, Sonnenberg A. Biology and function of hemidesmosomes. Matrix Biol 1999; 18(1):5. 39. Kainulainen T, Hakkinen L, Hamidi S, Larjava K, Kallioinen M, Peltonen J, Salo T, Larjava H, Oikarinen A. Laminin-5 expression is independent of the injury and the microenvironment during reepithelialization of wounds. J Histochem Cytochem 1998; 46(3):353. 40. Mizushima H, Koshikawa N, Moriyama K, Takamura H, Nagashima Y, Hirahara F, Miyazaki K. Wide distribution of laminin-5 gamma 2 chain in basement membranes of various human tissues. Horm Res 1998; 50(Suppl 2):7. 41. Nishiyama T, Amano S, Tsunenaga M, Kadoya K, Takeda A, Adachi E, Burgeson RE. The importance of laminin 5 in the dermal=Nepidermal basement membrane. J Dermatol Sci 2000; 24(Suppl 1):S51. 42. Myllyharju J, Kivirikko KI. Collagens and collagen-related diseases. Ann Med 2001; 33(1):7. 43. Wetzels RH, Robben HC, Leigh IM, Schaafsma HE, Vooijs GP, Ramaekers FC. Distribution patterns of type VII collagen in normal and malignant human tissues. Am J Pathol 1991; 139(2):451. 44. Jones JC, Dehart GW, Gonzales M, Goldfinger LE. Laminins: an overview. Microsc Res Tech 2000; 51(3):211. 45. Ben-Bassat H, Chaouat M, Segal N, Zumai E, Wexler MR, Eldad A. How long can cryopreserved skin be stored to maintain adequate graft performance?. Burns 2001; 27(5):425. 46. Krejci NC, Cuono CB, Langdon RC, McGuire J. In vitro reconstitution of skin: fibroblasts facilitate keratinocyte growth and differentiation on acellular reticular dermis. J Invest Dermatol 1991; 97(5): 843. 47. Langdon RC, Cuono CB, Birchall N, Madri JA, Kuklinska E, McGuire J, Moellmann GE. Reconstitution of structure and cell function in human skin grafts derived from cryopreserved allogeneic dermis and autologous cultured keratinocytes. J Invest Dermatol 1988; 91(5):478. 48. Cuono CB, Langdon R, Birchall N, Barttelbort S, McGuire J. Composite autologous-allogeneic skin replacement: development and clinical application. Plast Reconstr Surg 1987; 80(4):626.
8 Cutaneous Gene Therapy with Cultured Skin Substitutes Dorothy M. Supp and Steven T. Boyce Shriners Hospital for Children, Cincinnati Burns Hospital, and The University of Cincinnati College of Medicine Cincinnati, Ohio, U.S.A.
I. INTRODUCTION Recovery from very large skin injuries such as burns requires timely closure of open wounds. The historical standard of care for healing large excised wounds has been grafting of splitthickness autologous skin. In patients with massive skin loss, wound closure is problematic due to the lack of donor sites for skin autografting. Donor sites can be reharvested several times, but this increases the number of surgical procedures and length of hospitalization required for definitive wound closure. Delay in wound coverage can increase the likelihood of infection, which is a major cause of mortality from burn injury [1]. Wound closure can also be problematic for patients suffering from chronic wounds. Nonhealing wounds such as decubitus, venous, or pressure ulcers are relatively common, especially in the elderly population [2], and thus have large medical and economic impacts. As with acute burn wounds, the standard method for closing small nonhealing wounds has been split-thickness or full-thickness skin grafting, but these therapies can lead to painful donor sites that are slow to heal due to underlying deficiencies in wound healing [3]. The need for rapid closure of extensive skin wounds has led to the development of a number of alternatives to split-thickness skin autograft, including tissue engineered skin substitutes (Table 1). These skin substitutes vary in structure and composition, but are all designed to meet the primary goal of restoration of barrier function. Though none of the currently available models can fully replace all the functions of uninjured skin, several can provide temporary coverage, facilitate wound repair, or act as permanent skin replacements. In addition, skin substitutes containing cultured cells can be utilized as gene therapy vehicles for wound healing applications or in the treatment of cutaneous or systemic genetic disorders.
II. FUNCTION AND COMPOSITION OF SKIN SUBSTITUTES Native skin performs a wide range of protective, perceptive, and regulatory functions, but its role in providing a barrier to fluid loss and microbial contamination is most critical for survival. The barrier function of skin is performed by the epidermis, which is comprised mainly of 155
156
Supp and Boyce
Table 1 Engineered Skin Substitutes Skin substitute
Source
Alloderm® [4]
LifeCell Corp.
Apligraf® [5,6] Biobrane™ [7,8] Cultured skin substitutes (CSS) [9–11] Dermagraft® [12,13]
Epicel™ [14] Fibrin–cell suspensions [15,16]
Dermal component
Allogeneic acellular human dermis Collagen gel containing Organogenesis, Inc. allogeneic fibroblasts Dow Hickam/Bertek Collagen-coated nylon mesh Pharmaceuticals Not yet commercially Collagen–glycosaminoglycan sponge containing available autologous fibroblasts Polymer mesh scaffold Advanced Tissue containing allogeneic Sciences human fibroblasts None Genzyme Tissue Repair Corp. Not yet commercially None available
Integra® [17–20]
Integra Life Science Corp.
Laserskin™ [21]
Fidia Advanced Biopolymers
Orcel™ [22]
Ortec International, Inc.
Transcyte® [23,24]
Advanced Tissue Sciences
Epidermal component Can be used with thin split-thickness autograft Allogeneic keratinocytes Silicone layer Autologous keratinocytes
Can be used with thin split-thickness autograft
Autologous keratinocyte sheets Autologous keratinocytes applied with fibrin net or in suspension with fibrin glue Collagen–glycosaminogly- Silastic coating (can be can membrane replaced with thin splitthickness autograft) Allogeneic or autologous Laser-perforated fibroblasts hyaluronic acid membrane containing autologous keratinocytes Allogeneic keratinocytes Collagen sponge containing allogeneic fibroblasts Biobrane™ nylon mesh None populated with fibroblasts, then frozen
This list gives examples of several types of skin substitutes and is not intended to be all-inclusive.
keratinocytes. Other cells of the epidermis include Langerhans cells, which function in immune regulation, and melanocytes, which produce skin pigment. The keratinocytes form a stratified epithelium with basal, proliferating cells at the innermost layer and the keratinized, relatively impermeable outer stratum corneum layer at the skin surface [25]. Beneath the epidermal layer, the dermis provides structural integrity and elasticity and contains blood vessels that nourish the skin. The cellular components of dermis include fibroblasts, endothelial cells, smooth muscle cells, and mast cells, but the bulk of dermis is comprised of extracellular matrix. Skin appendages, such as hair follicles and sweat glands, span the dermal and epidermal layers. There are currently no engineered skin substitutes that completely restore the anatomy and physiology of native human skin. However, this does not diminish the importance of skin substitutes in the treatment of burns and chronic wounds. Among the most widely used skin substitutes is Integra威 artificial skin, a collagen–glycosaminoglycan (GAG) membrane with a
Cutaneous Gene Therapy with Cultured Skin Substitutes
157
silastic coating [17–19,26]. This acellular product is used primarily as a dermal replacement, as its porous structure promotes ingrowth of fibroblasts and endothelial cells from the wound bed. The silastic layer acts as a barrier until the dermal replacement is vascularized; it is then removed and replaced with thin split-thickness autograft. Other skin substitutes containing either epidermal keratinocytes or dermal fibroblasts or both are being increasingly utilized to facilitate wound healing. These can be used alone or in combination with other skin substitutes [27–32]. The cells may be allogeneic (e.g., Apligraf威, OrcelTM), usually isolated from neonatal foreskin, or autologous (e.g., EpicelTM), isolated from biopsies of the patient’s own skin (Table 1). Allogeneic cells are beneficial in that they are readily obtained and can be immediately available when needed for grafting, and they secrete cytokines that can facilitate healing. However, they do not persist on the patient after healing is complete and are usually replaced by the patient’s own cells within 1 to 6 weeks after grafting [33–36]. For permanent skin replacement, autologous cells must be used. Favorable clinical results have been obtained with cultured skin substitutes (CSS) comprised of collagen-GAG membranes containing autologous fibroblasts and keratinocytes (Fig. 1) [11,32,37,38]. CSS can provide permanent replacement of both dermal and epidermal layers. After healing, they resemble split-thickness autograft and provide a satisfactory cosmetic outcome. CSS are prepared using patient-derived fibroblasts and keratinocytes that are isolated
Figure 1 Cultured skin substitute for healing of pediatric burn wounds. A Acellular collagen–glycosaminoglycan (GAG) substrate (CGAG) prior to inoculation of cultured cells. B Cultured skin substitute 1 week after inoculation of human keratinocytes (HK) and fibroblasts (HF) cultured from a pediatric burn patient. The keratinocytes form a stratified epidermal layer with a keratinized stratum corneum analog at the air-exposed surface. The fibroblasts fill the dermal compartment, where they begin to degrade the collagen-GAG substrate and synthesize new extracellular matrix. The graft was biopsied 2 days prior to patient grafting. Collagen and graft biopsies were embedded in glycol-methacrylate and stained with toluidine blue. The scale bar shown in panel A is for both sections.
158
Supp and Boyce
from small split-thickness skin biopsies using standard techniques [39–41]. Because cultured fibroblasts and keratinocytes have population doubling times of 1 day or less, very large populations of cells can be generated within a few weeks [39]. Fibroblasts and keratinocytes are serially inoculated onto collagen-GAG substrates at high cell densities (0.5–1.0 ⳯ 106/cm2) [9,10,42]. Culture at the air–liquid interface provides a polarized environment, with nutrient medium contacting the dermal substitute and air contacting the epidermal substitute [43–45]. This results in stratification and cornification of the keratinocyte layer. In the dermal layer, fibroblasts fill the biopolymer substrate, begin to degrade it, and generate new extracellular matrix (Fig. 1). Because CSS grafted to clinical patients for closure of excised burns [11,32,38] or chronic wounds [37] contain only two cell types, fibroblasts and keratinocytes, they cannot replace all of the functions of native skin. Thus, despite favorable results, limitations in anatomy remain which can influence engraftment and functional and cosmetic outcome. Preclinical studies have investigated the preparation of engineered skin containing additional cell types to increase homology to native human skin and enhance functional outcome (Fig. 2). For example, incorporation of endothelial cells has been utilized to attempt to initiate angiogenesis in engineered skin grafts in vitro. Because cultured skin lacks a vascular plexus, it is vascularized more slowly than splitthickness skin autograft after transplantation. This can contribute to graft failure by protracting the time of nutrient deprivation of grafted cells and increasing susceptibility to microbial contamination. This has been addressed clinically, in part, by the use of nutrient and antimicrobial dressing fluids for several days after graft application [46–48]. The dressing fluids nourish and protect the grafts until vascularization occurs, usually within 1 week after grafting. In preclinical studies, human umbilical vein endothelial cells (HUVEC) [49,50] or human dermal microvascular endothelial cells (HDMEC) [51] have been incorporated into cultured skin grafts, resulting in formation of vascular analogs after grafting to athymic mice. HUVEC are readily available but can only be utilized in allogeneic skin substitutes. For clinical application of autologous endothelialized skin substitutes, use of dermal endothelial cells is optimal, and these cells should be isolated from the same skin biopsy used for preparation of fibroblast and keratinocyte cultures. These criteria have been met in preclinical studies [51], but enhanced vascularization due to inclusion of endothelial cells has not yet been demonstrated. A current practical limitation to the inclusion of endothelial cells in CSS is the slower growth in primary culture of HDMEC compared to dermal fibroblasts, which could be expected to delay preparation of endothelialized
Figure 2 Cells used for preparation of cultured skin substitutes (CSS). Dermal cells, combined with a matrix (not diagrammed), serve as a dermal substrate for epidermal cell inoculation. CSS prepared with keratinocytes and fibroblasts have been used clinically as adjunctive therapies for burns and chronic wounds. CSS prepared with cultured melanocytes and endothelial cells have been evaluated in preclinical studies.
Cutaneous Gene Therapy with Cultured Skin Substitutes
159
CSS for grafting to patients. In addition, studies have shown that only a small proportion of HDMEC persists during culture of engineered skin grafts [51], which has been proposed to be due to apoptosis of the endothelial cells [50,52,53]. Thus, though the addition of endothelial cells to engineered skin is feasible, technical hurdles still need to be overcome before progression to clinical application. Another limitation of grafted cultured skin is absent or irregular pigmentation. In normal skin, pigmentation results from the appropriate distribution and function of epidermal melanocytes. These cells are important physiologically, for protection from ultraviolet irradiation [54,55], and psychologically, influencing a patient’s body image and personal identity. It has long been recognized that ‘‘passenger’’ melanocytes can persist in cultures of epidermal keratinocytes, resulting in foci of pigmentation after grafting to mice [56,57] or patients [32,58]. Selective cultivation of human melanocytes and quantitative addition to CSS in preclinical studies showed that uniform pigmentation can be achieved [59], though the intensity of pigment was not regulated.
III. GENETIC MODIFICATION OF SKIN CELLS The gene expression profile of keratinocytes can be regulated by the transfer of recombinant genes [60,61], and the genetically modified cells retain their ability to differentiate into a stratified epidermis [60]. Hypothetically, genetic modification of cells within CSS can be used to overcome limitations in anatomy and physiology, resulting in skin substitutes with greater homology to native human skin. Genetic modification can be used to ectopically express cytokines not normally expressed in a particular cell type, which may compensate for deficiencies of engineered skin compared to native skin. Alternatively, CSS can be genetically engineered to overexpress growth factors with roles in wound healing to enhance their therapeutic value for wound repair. Cultured skin cells are amenable to genetic modification by a variety of methods. Some of the more commonly used techniques are outlined in Table 2. One of the most efficient and widely used techniques for modification of keratinocytes is retroviral transduction. This method has also been used for other primary skin cells, including fibroblasts [79–81], endothelial cells [82,83], and melanocytes [84]. Currently, over one-third of clinical gene therapy trials involve retroviral-mediated gene transfer [85]. The retroviral vectors utilized for gene delivery are replication-incompetent, which is a safety measure designed to prevent unintentional spread of infectious viruses. In replication-incompetent retroviruses, the genes necessary for retroviral transcription and packaging (gag, pol, and env) have been deleted [86,87]. These genes have been separately transfected into a cell line known as a packaging cell line, which is often derived from NIH 3T3 cells. The retroviral vector contains the sequences necessary for viral integration, known as the long terminal repeats (LTRs), and viral RNA packaging ( sequence) [86,87]. The two LTRs generally flank a multiple cloning site that can be used to insert the investigator’s gene of interest. The upstream LTR can act as a promoter to drive expression of the inserted gene or alternative promoters (for example, tissuespecific promoters) can be inserted along with the gene’s coding sequence [87]. The retroviral vector sequences are contained on a plasmid that can be stably transfected into the packaging cell line. The proteins encoded by the gag, pol, and env genes, which are located elsewhere in the genome of the packaging cell, enable reverse transcription of the vector DNA into RNA. The retroviral RNA genome is packaged and released from the packaging cell in the form of infectious viral particles which can be used to genetically modify (i.e., transduce) target cells. Because the infectious particles contain retroviral RNAs without the gag, pol, and env genes, they can infect target cells without further spread of the virus. Infectious particles are only
160
Supp and Boyce
Table 2 Methods for the Genetic Modification of Cultured Mammalian Cells Method Nonviral DEAE-dextran [62,63]
Description
Pros
Cationic reagent, associates with negatively charged DNA; enters cell via endocytosis pathway Coprecipitate of calcium phosphate and DNA; enters cell via endocytosis pathway Cationic lipid–DNA complex with net neutral or positive charge to allow close contact with cell membrane High voltage pulses produce pores in cell membrane that DNA can pass through “Gene gun”; highvelociy delivery of DNA-coated microprojectiles
Simple and inexpensive
Enveloped singlestranded RNA viruses; random integration
Efficient, stable integration; potential for long-term expression
Adenovirus [74,75]
Double-stranded DNA virus; nonintegrating
Adeno-associated virus [76]
Single-stranded DNA virus; site-specific integration
Can infect broad range of cells; infects dividing and nondividing cells; large capacity for inserted sequences Infects broad range of cells; infects dividing and nondividing cells
Lentivirus [77,78]
Based on human immunodeficiency virus type I (HIV)
Calcium phosphate [64,65]
Liposome [66,67]
Electroporation [68,69]
Biolistic [70,71]
Viral Retrovirus [72,73,60]
Simple and inexpensive; can be used for stable transfection
Cons Primary cells not easily transfected; inefficient for stable integration; can be cytotoxic to some cells Some cells (i.e., keratinocytes) sensitive to calcium exposure; variable efficiency
Simple, effective for a wide variety of cell types
Optimization required for different cell types; variable efficiency
Can be used for transient or stable transfection of several cell types
High cell mortality; must be used on cells in suspension; requires specialized equipment Requires specialized equipment that can be very costly
Efficient; can be used for cultured cells or cells in vivo
Can infect nondividing cells; efficient
Only infects dividing cells; can rarely cause insertional mutation leading to cell transformation Episomal; not suited for long-term expression studies; can trigger host immune response Low capacity for inserted sequences; difficult to produce—requires adenovirus as helper for replication and packaging Relatively new; safety issues need to be studied further
This list is intended as an overview of some of the more popular methods for gene delivery in cultured cells and is not meant to be all-inclusive.
Cutaneous Gene Therapy with Cultured Skin Substitutes
161
produced in cells encoding the viral structural genes in trans. Thus, the retrovirus gets trapped in the target cell, where it randomly incorporates into the genome. The gene encoded by the retroviral vector can be transcribed, but not packaged, preventing unintentional horizontal spread [86,87]. Genes of up to approximately 8 kilobases can be cloned into retroviral vectors [87], making this system very useful for most gene delivery applications, although it requires cell replication for retroviral integration [88].
IV. GENE THERAPY WITH CULTURED SKIN GRAFTS Skin has been recognized as an appealing target organ for gene therapy. This is mainly due to the skin’s accessibility, which facilitates the administration of therapy [89]. Skin is amenable to both in vivo gene therapy, in which cells are modified while on the patient, as well as ex vivo gene therapy, in which skin is removed from the patient, cells are isolated and genetically modified in culture, and then transplanted back to the patient [90]. For either in vivo or ex vivo gene therapy, genetically modified skin can be readily monitored and removed if necessary [89], an advantage not available for many other organs. Techniques for the culture of skin cells are well established [39–41,91], and several methods have been described for their genetic modification [61,92] (Table 2). These factors, combined with the potential to transplant modified cells using a variety of constructs (Table 1), make ex vivo gene therapy with genetically modified cultured skin substitutes an attractive approach for improved wound healing and treatment of cutaneous or systemic diseases (Table 3). A. Enhanced Performance of Cultured Skin Using Genetically Modified Cells Preclinical studies have evaluated the use of retroviral transduction to enhance the performance of cultured skin grafts containing only keratinocytes. Retroviral vectors have been used to overex-
Table 3 Ex Vivo Gene Therapy with Cultured Skin Therapeutic goal
Technical strategy
Condition(s)
Example(s) PDGF-A [93,94] or VEGF overexpression [95,96] in skin substitutes LAMB3 gene transfer for JEB [97,98]; TGase1 gene transfer for LI [99,100] Correction of tyrosinase mutation in albino melanocytes [101,102]
Enhance performance of cultured skin graft Replace deficient or absent gene
Overexpress therapeutic gene
Deficient wound healing
Add a wild type version of defective or missing gene
Recessive cutaneous disorder
Correct mutated gene by gene conversion
Homologous recombination with DNA/RNA oligo hybrids Overexpress circulating protein
Dominant or recessive cutaneous disorders
Treat systemic disorder
Deficiency of blood factor or hormone
Delivery of clotting factor IX to treat hemophilia B [103–105]
PDGF-A, platelet-derived growth factor A; VEGF, vascular endothelial growth factor; LAMB-3, laminin 5  3 chain; JEB, junctional epidermolysis bullosa; TGase1, transglutaminase 1; LI, lamellar ichthyosis.
162
Supp and Boyce
press cytokines involved in the process of wound healing, particularly those that have been shown to improve healing upon topical application [106,107]. An example is platelet-derived growth factor (PDGF), a mitogen for cells of mesenchymal origin that is expressed in normal and wounded skin [108] and has been shown to stimulate wound healing [109]. In one study, retroviral transduction was used to increase expression of PDGF-A in human keratinocytes [93]. The modified cells were transplanted as epithelial sheets under skin flaps in athymic mice. Compared to controls containing unmodified cells, the connective tissue beneath modified grafts was thicker and contained more blood vessels [93]. In a subsequent study, composite grafts prepared with PDGF-A–modified keratinocytes combined with acellular human dermis were evaluated [94]. Retroviral transduction leading to PDGF-A overexpression was found to enhance graft performance in athymic mice by increasing dermal cellularity and collagen production and decreasing wound contraction [94]. Interestingly, different results were obtained with PDGF-A overexpression in a composite cultured skin substitute containing both keratinocytes and fibroblasts prior to grafting [81]. In this study, retroviral transduction was used to overexpress PDGFA in either keratinocytes or fibroblasts. Unlike the results of previous studies, which involved skin substitutes that did not contain fibroblasts, no differences were observed between control or PDGF-A modified grafts, either in vitro or after grafting to athymic mice [81]. The observations from that study suggested that overproduction of PDGF-A, a mitogen for dermal fibroblasts, was not required for satisfactory performance of cultured skin containing both keratinocytes and fibroblasts. In the prior studies, overexpression of PDGF-A most likely compensated for the absence of fibroblasts in vitro, leading to enhanced performance in vivo [93,94]. Thus, these studies demonstrate the utilization of gene modification to improve wound healing with cultured keratinocyte grafts. Despite satisfactory performance in preclinical and clinical studies, CSS containing both keratinocytes and fibroblasts are not without limitations. For example, absence of a vascular plexus results in slower vascularization of CSS after grafting compared to split-thickness skin autograft. Studies were performed to assess the feasibility of overexpressing an angiogenic cytokine in genetically modified CSS for improved vascularization. Vascular endothelial growth factor (VEGF) was chosen for these studies because it is a specific and potent mitogen for microvascular endothelial cells [110]. VEGF is expressed by epidermal keratinocytes, and its expression is increased during the angiogenic phase of skin wound healing [111,112]. Several preclinical and clinical trials of VEGF gene therapy have been published, describing improved perfusion of ischemic heart tissue [113], enhanced vascularization of ischemic limbs [114], and normalization of vascularization levels in diabetic tissues [115]. Retroviral transduction was used to overexpress VEGF in keratinocytes in CSS, leading to increased VEGF expression after grafting to athymic mice [95,96]. No differences were observed between control and VEGFmodified CSS in vitro. However, enhanced and accelerated vascularization was observed in VEGF-modified CSS after grafting [95,96]. VEGF-modified CSS had greater levels of engraftment and decreased wound contraction compared to controls, suggesting improved performance due to VEGF overexpression. These studies demonstrate compensation for an anatomical deficiency of engineered skin through the use of gene therapy and indicate that genetic modification of cultured skin cells can be used to improve wound healing with CSS. Furthermore, CSS genetically engineered to secrete an angiogenic cytokine can prospectively offer an improved method for treatment of chronic wounds, such as diabetic ulcers, in which ischemia hinders the normal wound healing process. B. Gene Therapy to Treat Cutaneous Disease There are several hereditary skin diseases that are amenable to gene therapy with genetically modified cultured skin cells. One example is lamellar ichthyosis (LI), a disease characterized
Cutaneous Gene Therapy with Cultured Skin Substitutes
163
by abnormal epidermal differentiation and barrier function [116]. This disfiguring disease has been associated with loss of activity of the enzyme transglutaminase 1 (TGase1), which is required for normal cornified barrier formation [117]. The gene encoding TGase1 was cloned, and a retroviral vector constructed for gene transfer to keratinocytes. Transduction of keratinocytes isolated from LI patients with the TGase1 expression vector was able to restore enzyme activity to normal levels [118]. Composite grafts of LI patient–derived keratinocytes seeded on acellular dermis and grafted to athymic mice displayed a phenotype characteristic of the human disease, but restoration of TGase1 activity through gene transfer led to a normalization of epidermal barrier [117]. A subsequent study compared the efficiency of ex vivo versus in vivo gene delivery for correction of the TGase1 defect in LI keratinocytes [90]. Skin substitutes prepared with LI patient–derived keratinocytes were grafted to athymic mice and were directly injected in vivo with plasmid DNA encoding TGase1. Expression of TGase1 was detected, but was nonuniform and was insufficient to elicit phenotypic correction of the disease [90]. This study underscores a principal advantage to ex vivo gene therapy: the ability to efficiently modify most or all cells in the target population. Another cutaneous disease amenable to treatment with genetically modified keratinocytes is epidermolysis bullosa (EB). The phenotype ranges from mild to very severe blistering and erosions of the skin [119]. There are several types of EB, resulting from mutations in genes encoding structural proteins of the basement membrane zone of the skin. One of the more severe forms, junctional EB (JEB), can result from mutation of genes encoding subunits of laminin 5, a component of anchoring filaments in the basement membrane zone [119]. Preclinical studies have shown correction of the JEB phenotype by gene transfer of the beta 3 subunit of laminin 5, encoded by the LAMB3 gene. Using a retroviral gene transfer vector, the LAMB3 gene was introduced with high efficiency into epidermal stem cells from patients with laminin 5–dependent JEB [120]. Organotypic cultures prepared with LAMB3-transduced keratinocytes showed normal assembly of the dermal–epidermal attachment structures that are missing in JEB skin, indicating correction of the mutant phenotype by LAMB3 gene transfer [121]. Animal studies further confirmed the ability of LAMB3 to correct the JEB defect. Cultured skin prepared using JEB patient–derived keratinocytes was grafted to mice. Cells without genetic modification showed an abnormal phenotype characteristic of human JEB, while the cells transduced with the LAMB3 gene were phenotypically normal [122]. Related studies targeted correction of another gene mutated in JEB: BP180, a structural component of the dermal–epidermal junction. Retroviral gene transfer was used to restore BP180 expression to primary keratinocytes from JEB patients [123]. Cultured skin regenerated from BP180-transduced cells showed normal BP180 expression at the dermal–epidermal junction [123]. These studies demonstrate the potential for correction of recessive cutaneous defects using genetically modified cultured skin. C. Genetically Modified Cultured Skin for Treatment of Systemic Deficiencies The transplantation of genetically modified skin cells raises the possibility of systemic gene therapy with cultured skin grafts. This requires that proteins expressed by keratinocytes, or other cells within engineered skin grafts, enter systemic circulation and reach levels sufficient to elicit a physiological response. The first demonstration that a protein secreted by epidermal keratinocytes could reach the systemic circulation involved apolipoprotein E (apoE), a lipoprotein involved in reverse cholesterol transport [124]. Normal human skin and unmodified cultured keratinocytes both secrete apoE. Human apoE protein could be detected in the serum of rats grafted with split-thickness human skin in a sandwich-flap model, and athymic mice grafted with human keratinocyte sheets [124]. These experiments showed that a protein produced by
164
Supp and Boyce
grafted keratinocytes could traverse the basement membrane of the dermal–epidermal junction and reach systemic circulation. A subsequent study examined apoE secretion by genetically modified keratinocytes, transfected with a recombinant apoE gene [125]. When grafted to athymic mice, both endogenous and recombinant apoE were detected in serum, demonstrating the feasibility of systemic gene therapy with modified keratinocytes [125]. Other studies examined the possibility of engineering transplantable keratinocytes to secrete non-native proteins with prospective therapeutic value for systemic protein delivery. For example, keratinocytes were modified to express the human growth hormone (hGH) gene by retroviral transduction [60,126]. When hGH-modified keratinocyte sheets were grafted subcutaneously to athymic mice, they formed a stratified epidermis that continued to produce localized hGH at low levels, but circulating hGH could not be detected [60]. In a subsequent study, hGHmodified keratinocytes were grafted to chamber-enclosed full-thickness wounds in pigs, and hGH was detected in the wound fluid [126]. Other investigators used an Epstein-Barr virus–based expression system combined with lipofection for transfer of the hGH gene to primary human keratinocytes [127]. When transplanted to athymic mice, hGH was detected in mouse serum for 4 days, although the grafts persisted for much longer [127]. To test the possibility that a tissue-specific promoter could enhance expression and secretion from keratinocytes, transgenic mice were generated using a keratinocyte-specific keratin 14 (K14) promoter to drive expression of hGH. The hGH gene was highly expressed in the skin of K14-hGH transgenic mice, and hGH was detected at physiological levels in transgenic mouse serum [128]. Furthermore, grafts of K14-hGH transgenic mouse skin transplanted to nontransgenic hosts resulted in hGH in the serum of grafted mice [128]. These studies demonstrated the importance of a strong, specific promoter for driving expression of the exogenous gene in keratinocytes, and demonstrated that systemic delivery of an exogenous protein from grafted keratinocytes could be achieved. Hemophilia B is an X-linked bleeding disorder that has served as a model for systemic gene therapy with genetically modified keratinocytes. This hereditary disease is caused by mutations in the clotting factor IX gene that lead to absence of circulating factor IX. Primary human keratinocytes transduced with a retroviral vector encoding factor IX, driven by the retroviral LTR promoter, were transplanted to athymic mice, and human factor IX was detected in the serum of mice for up to 1 week after grafting [129]. Subsequent studies utilized an improved vector containing keratinocyte-specific enhancer elements and a silicone chamber for transplantation of modified cells [130]. This model resulted in extended expression of factor IX and detectable levels in the serum of grafted mice for up to 5 weeks [130]. When factor IX–modified keratinocytes were transplanted as part of a bilayered skin equivalent, human factor IX was detected in the serum for over 1 year [131]. This long-term expression suggested that keratinocyte stem cells had been transduced, and indicated the importance of the cultured skin design in addition to vector construction for extended expression and protein circulation [131]. More recently, a genetic defect in mice was corrected using grafts of genetically modified keratinocytes. Leptin, a hormone that regulates food intake and body weight, is associated with obesity in humans and is deficient in the genetically obese ob/ob mutant mouse [132]. Retroviral transduction was used for gene transfer of human leptin to primary human keratinocyte cultures. Composite cultured skin was prepared using leptin-modified keratinocytes and fibrin–fibroblast gels, and the grafts were transplanted to leptin-deficient ob/ob mice. Human leptin was detectable in the serum of grafted mice, and significant reductions in food intake and body weight were observed [132]. This study represents the first clear demonstration of replacement of a deficient circulating protein and phenotypic correction by genetically modified human keratinocyte grafts [132].
Cutaneous Gene Therapy with Cultured Skin Substitutes
165
V. CONCLUSIONS AND FUTURE DIRECTIONS In 1975, Rheinwald and Green described a method for the culture and serial propagation of human keratinocytes [91]. This technical advance allowed for the development of tissue engineered skin substitutes, which are now used for treatment of acute and chronic wounds. Progress in cell and molecular biology has made genetic engineering of skin cells possible. Combined, these developments have paved the way for ex vivo cutaneous gene therapy, which promises new approaches for treatment of hereditary disease and improved wound healing. Preclinical studies have suggested enormous potential therapeutic benefits of genetically modified cultured skin substitutes, but many obstacles will need to be overcome. First and foremost, safety issues will need to be addressed. As with any new therapy, the potential benefit to the patient must be weighed against the risks involved with treatment. For example, for patients with extensive full-thickness burns, cultured skin substitutes offer the benefits of increasing material available for grafting and reducing donor site utilization [38], but it is currently unclear if inclusion of genetically modified cells will enhance the therapeutic value of CSS such that the risks of the modified cells will be justified. Although replicationincompetent retroviral vectors are commonly used for gene delivery in epidermal cells, the theoretical risk of replication competent virus production still exists. Improved design of retroviral vectors, the introduction and evaluation of new vectors such as the lentivirus-based system, and improved nonviral gene delivery methods will greatly advance the safety of the gene transfer process. Among other safety issues to be addressed is the long-term risk of grafted genetically modified cells, which is not easily determined solely from animal studies. Improved design of gene transfer vectors to enhance efficiency and safety will undoubtedly shift the risk/ benefit balance and facilitate clinical application. Advances in vector design that permit regulation of gene expression, such as the gene-switch system for inducible skin gene expression [133], will reduce risks associated with extended cytokine gene expression, further enhancing the safety of genetically modified skin cells. For treatment of wounds, the use of genetically modified cells can prospectively lead to production of cultured skin grafts with greater homology to native human skin, enhancing their clinical efficacy. For conditions where only palliative treatments currently exist, such as some hereditary cutaneous or systemic disorders, gene therapy with cultured skin offers the promise of permanent replacement of missing or defective genes, and thus, curative therapies. Improvements in efficiency of gene transfer and stability of gene expression will be necessary for treatment of cutaneous disorders. For treatment of systemic deficiencies, expression level, duration of expression, and secretion into systemic circulation must also be optimized. Preclinical studies have demonstrated the feasibility of cutaneous gene therapy for the treatment of a multitude of disorders. Technical advances will continue to promote the translation of these therapies from the laboratory to the clinic.
REFERENCES
1. Berthod F, Damour O. In vitro reconstructed skin models for wound coverage in deep burns. Br. J. Dermatol 1997; 136:809–816. 2. Falanga V. Chronic wounds: pathophysical and experimental considerations. J. Invest. Dermatol 1993; 100(5):721–725.
166
Supp and Boyce
3. Phillips TJ, Gilchrest BA. Cultured epidermal grafts in the treatment of leg ulcers. Adv. Dermatol 1990; 5:33–48. 4. Wainwright D, Madden M, Luterman A, Hunt J, Monafo W, Heimbach D, Kagan R, Sittig K, Dimick A, Herndon D. Clinical evaluation of an acellular allograft dermal matrix in full-thickness burns. J. Burn Care Rehabil 1996; 17:124–136. 5. Falanga V, Margolis DJ, Alvarez O, Auletta M, Maggiacomo F, Altman M, Jensen J, Sabolinski M, Hardin-Young J. Rapid healing of venous ulcers and lack of clinical rejection with an allogeneic cultured human skin equivalent. Arch. Dermatol 1998; 134:293–300. 6. Eaglstein WH, Iriondo M, Laszlo K. A composite skin substitute (Graftskin) for surgical wounds: a clinical experience. Dermatol. Surg 1995; 21:839–843. 7. Tavis MN, Thornton NW, Bartlett RH, Roth JC, Woodroof EA. A new composite skin prosthesis. Burns 1980; 7:123–130. 8. Purdue GF, Hunt JL, Gillespie RW, Hansbrough JF, Dominic WJ, Robson MC, Smith DJ, MacMillan BG, Waymack JP, Heradon DN. Biosynthetic skin substitute versus frozen human cadaver allograft for temporary coverage of excised burn wounds. J. Trauma 1987; 27:155–157. 9. Hansbrough JF, Boyce ST, Cooper ML, Foreman TJ. Burn wound closure with cultured autologous keratinocytes and fibroblasts attached to a collagen–glycosaminoglycan substrate. J.A.M.A 1989; 262:2125–2130. 10. Boyce ST, Greenhalgh DG, Kagan RJ, Housinger T, Sorrell JM, Childress CP, Rieman M, Warden GD. Skin anatomy and antigen expression after burn wound closure with composite grafts of cultured skin cells and biopolymers. Plast. Reconstr. Surg 1993; 91:632–641. 11. Boyce ST, Goretsky MJ, Greenhalgh DG, Kagan RJ, Rieman MT, Warden GD. Comparative assessment of cultured skin substitutes and native skin autograft for treatment of full-thickness burns. Ann. Surg 1995; 222(6):743–752. 12. Cooper ML, Hansbrough JF, Spielvogel RL, Cohen R, Bartel RL, Naughton G. In vivo optimization of a living dermal substitute employing cultured human fibroblasts on a biodegradable polyglycolic or polygalactin mesh. Biomaterials 1991; 12:243–248. 13. Hansbrough JF, Dore C, Hansbrough WB. Clinical trials of a living dermal tissue replacement placed beneath meshed, split-thickness skin grafts on excised wounds. J. Burn Care Rehabil 1992; 13:519–529. 14. Carsin H, Ainaud P, Le Bever H, Rives J, Lakhel A, Stephanazzi J, Lambert F, Perrot J. Cultured epithelial autografts in extensive burn coverage of severly traumatized patients: a five year singlecenter experience with 30 patients. Burns 2000; 26:379–387. 15. Hunyadi J, Farkas B, Bertenyi C, Olah J, Dobozy A. Keratinocyte grafting: a new means of transplantation for full-thickness wounds. J Dermatol. Surg. Oncol 1998; 14:75–78. 16. Kaiser HW, Stark GB, Kopp J, Balcerkiewicz A, Spilker G, Kreysel HW. Cultured autologous keratinoyctes in fibrin glue suspension, exclusively and combined with STS-allograft (preliminary clinical and histological report of a new technique). Burns 1994; 20:23–29. 17. Yannas IV, Burke JF. Design of an artifical skin. I Basic design principles. J. Biomed. Mater. Res 1980; 14:65–81. 18. Yannas IV, Burke JF, Gordon PL, Huang C, Rubenstein RH. Design of an artificial skin, II. Control of chemical composition. J. Biomed. Mater. Res 1980; 14:107–131. 19. Heimbach D, Luterman A, Burke JF, Cram A, Herndon D, Hunt J, Jordon M, McManus W, Solem L, Warden G, Zawacki B. Artifical dermis for major burns; a multi-center randomized clinical trial. Ann. Surg 1988; 208:313–320. 20. Sheridan RL, Hegarty M, Tompkins RG, Burke JF. Artificial skin in massive burns — results to ten years. Eur. J. Plast. Surg 1994; 17:91–93. 21. Lam PK, Chan ES, To EW, Lau CH, Yen SC, King WW. Development and evaluation of a new composite Laserskin graft. J. Trauma 1999; 47:918–922. 22. Stephens R, Wilson K, Silverstein P. A premature infant with skin injury successfully treated with bilayered cell matrix. Ostomy Wound Manage 2002; 48:34–38. 23. Hansbrough JF, Mozingo DW, Kealey GP, Davis M, Gidner A, Gentzkow GD. Clinical trials of a biosynthetic temporary skin replacement, Dermagraft–transitional covering, compared with cryopre-
Cutaneous Gene Therapy with Cultured Skin Substitutes
24.
25. 26. 27.
28. 29.
30.
31.
32.
33.
34. 35. 36. 37. 38.
39.
40. 41.
42.
167
served human cadaver skin for temporary coverage of excised burn wounds. J. Burn Care Rehab 1997; 18:43–51. Purdue GF, Hunt JL, Still JM, Law EJ, Herndon DN, Goldfarb IW, Schiller WR, Hansbrough JF, Hickerson WL, Himel HN, Kealey GP, Twomey J, Missavage AE, Solem LD, Davis M, Totoritis M, Gentzkow GD. A multicenter clinical trial of a biosynthetic skin replacement, Dermagraft-TC, compared with cryopreserved human cadaver skin for temporary coverage of excised burn wounds. J. Burn Care Rehab 1997; 18:52–57. Elias PM. Epidermal lipids, barrier function, and desquamation. J. Invest. Dermatol 1983; 80: 44s–49s. Sheridan RL, Hegarty M, Tompkins RG, Burke JF. Artificial skin in massive burns—results to ten years. Eur. J. Plast. Surg 1994; 17:91–93. Orgill DP, Butler CE, Regan JF, Barlow MS, Yannas IV, Compton CC. Vascularized collagen–glycosaminoglycan matrix provides a dermal substrate and improves take of cultured epithelial autografts. Plast. Reconstr. Surg 1998; 102:423–429. Pandya AN, Woodward B, Parkhouse DM. The use of cultured autologous keratinocytes with Integra in the resurfacing of acute burns. Plast. Reconstr. Surg 1998; 102:825–830. Medalie DA, Eming SA, Tompkins RG, Yarmush ML, Krueger GG. Evaluation of human skin reconstituted from composite grafts of cultured keratinocytes and human acellular dermis transplanted to athymic mice. J. Invest. Dermatol 1996; 107(1):121–127. Compton CC, Butler CE, Yannas IV, Warland G, Orgill DP. Organized skin structure is regenerated in vivo from collagen-GAG matrices seeded with autologous keratinocytes. J. Invest. Dermatol 1998; 110:908–916. Horch RE, Bannasch H, Stark GB. Cultured human keratinocytes as a single cell suspension in fibrin glue combined with preserved dermal grafts enhance skin reconstitution in athymic mice fullthickness wounds. Eur. J. Plast. Surg 1999; 22:237–243. Boyce ST, Kagan RJ, Meyer NA, Yakuboff KP, Warden GD. The 1999 Clinical Research Award. Cultured skin substitutes combined with Integra to replace native skin autograft and allograft for closure of full–thickness burns. J. Burn Care Rehabil 1999; 20:453–461. Brain A, Purkis P, Coates P, Hackett M, Navsaria H, Leigh I. Survival of cultured allogeneic keratinocytes transplanted to deep dermal bed assessed with probe specific for Y chromosome. Br. Med. J 1989; 298:917–919. Phillips TJ, Bhawan J, Leigh IM, Baum HJ, Gilchrest BA. Cultured epidermal autografts and allografts: a study of differentiation and allograft survival. J. Am. Acad. Dermatol 1990; 23:189–195. Thivolet J, Faure M, Demidem A, Mauduit G. Long-term survival and immunological tolerance of human epidermal allografts produced in culture. Transplantation 1986; 42(3):274–280. Zhao YB, Zhao XF, Li A, Lu SZ, Wang X, Huang SZ, Zhuo XT. Clinical observations and methods for identifying the existence of cultured epidermal allografts. Burns 1992; 18:4–8. Boyce ST, Glatter R, Kitzmiller WJ. Treatment of chronic wounds with cultured cells and biopolymers. Wounds 1995; 7(1):24–29. Boyce ST, Kagan RJ, Yakuboff KP, Meyer NA, Rieman MT, Greenhalgh DG, Warden GD. Cultured skin substitutes reduce donor skin harvesting for closure of excised, full-thickness burns. Ann. Surg 2002; 235:269–279. Boyce ST, Ham RG. Calcium-regulated differentiation of normal human epidermal keratinocytes in chemically defined clonal culture and serum-free serial culture. J. Invest. Dermatol 1983; 81(Suppl 1):33S–40S. Boyce ST, Ham RG. Cultivation, frozen storage, and clonal growth of normal human epidermal keratinocytes in serum-free media. J. Tiss. Cult. Meth 1985; 9:83–93. Boyce ST. Methods for serum-free culture of keratinocytes and transplantation of collagen-GAG based composite grafts. In: Morgan JR, Yarmush M, eds. Methods in Tissue Engineering. Totowa, NJ: Humana Press, 1998:365–389. Boyce ST, Hansbrough JF. Biologic attachment, growth, and differentiation of cultured human epidermal keratinocytes on a graftable collagen and chondroitin-6–sulfate substrate. Surgery 1988; 103:421–431.
168
Supp and Boyce
43. Boyce ST, Williams ML. Lipid supplemented medium induces lamellar bodies and precursors of barrier lipids in cultured analogues of human skin. J. Invest. Dermatol 1993; 101:180–184. 44. Prunieras M, Regnier M, Woodley DT. Methods for cultivation of keratinocytes at the air–liquid interface. J. Invest. Dermatol 1983; 81(1):28S–33S. 45. Ponec M, Kempenaar J, Weerheim A, de Lannoy L, Kalkman I, Jansen H. Triglyceride metabolism in human keratinocytes cultured at the air–liquid interface. Arch. Dermatol. Res 1995; 287(8): 723–730. 46. Boyce ST, Holder IA. Selection of topical antimicrobial agents for cultured skin for burns by combined assessment of cellular cytotoxicity and antimicrobial activity. Plast. Reconstr. Surg 1993; 92(4):493–500. 47. Boyce ST, Harriger MD, Supp AP, Warden GD, Holder IA. Effective management of microbial contamination in cultured skin substitutes after grafting to athymic mice. Wound Rep. Reg 1997; 5:191–197. 48. Boyce ST, Supp AP, Harriger MD, Greenhalgh DG, Warden GD. Topical nutrients promote engraftment and inhibit wound contraction of cultured skin substitutes in athymic mice. J. Invest. Dermatol 1995; 104(3):345–349. 49. Black AF, Berthod F, L’Heureux N, Germain L, Auger FA. In vitro reconstruction of a human capillary-like network in a tissue-engineered skin equivalent. FASEB J 1998; 12:1331–1340. 50. Schechner JS, Nath AK, Zheng L, Kluger MS, Hughes CCW, Sierra-Honigmann MR, Lorber MI, Tellides G, Kashgarian M, Bothwell ALM, Pober JS. In vivo formation of complex microvessels lined by human endothelial cells in an immunodeficient mouse. Proc. Natl. Acad. Sci. USA 2000; 97:9191–9196. 51. Supp DM, Wilson-Landy K, Boyce ST. Human dermal microvascular endothelial cells form vascular analogs in cultured skin substitutes after grafting to athymic mice. FASEB J 2002; 16:797–804. 52. Pollman MJ, Naumovski L, Gibbons GH. Endothelial cell apoptosis in capillary network remodeling. J. Cell Physiol 1999; 178:359–370. 53. Nor JE, Christensen J, Mooney DJ, Polverini PJ. Vascular endothelial growth factor (VEGF)mediated angiogenesis is associated with enhanced endothelial cell survival and induction of Bcl2 expression. Am. J. Pathol 1999; 154:375–384. 54. Abdel-Malek ZA. Endocrine factors as effectors of integumental pigmentation. Dermatol. Clin 1988; 6(2):175–183. 55. Nordlund JJ, Abdel-Malek ZA, Boissy RE, Rheins LA. Pigment cell biology: an historical review. J. Invest. Dermatol 1989; 92:53S–60S. 56. Boyce ST, Supp AP, Harriger MD, Pickens WL, Wickett RR, Hoath SB. Surface electrical capacitance as a noninvasive index of epidermal barrier in cultured skin substitutes in athymic mice. J. Invest. Dermatol 1996; 107(1):82–87. 57. Supp AP, Wickett RR, Swope VB, Harriger MD, Hoath SB, Boyce ST. Incubation of cultured skin substitutes in reduced humidity promotes cornification in vitro and stable engraftment in athymic mice. Wound Rep. Reg 1999; 7:226–237. 58. Harriger MD, Warden GD, Greenhalgh DG, Kagan RJ, Boyce ST. Pigmentation and microanatomy of skin regenerated from composite grafts of cultured cells and biopolymers applied to full-thickness burn wounds. Transplantation 1995; 59:702–707. 59. Swope VB, Supp AP, Cornelius JR, Babcock GF, Boyce ST. Regulation of pigmentation in cultured skin substitutes by cytometric sorting of melanocytes and keratinocytes. J. Invest. Dermatol 1997; 109:289–295. 60. Morgan JR, Barrandon Y, Green H, Mulligan RC. Expression of an exogenous growth hormone gene in transplantable human epidermal cells. Science 1987; 237:1476–1479. 61. Fenjves ES. Approaches to gene transfer in keratinocytes. J. Invest. Dermatol 1994; 103:70S–75S. 62. McCutchan JH, Pagano JS. Enchancement of the infectivity of simian virus 40 deoxyribonucleic acid with diethylaminoethyl-dextran. J. Natl. Cancer Inst 1968; 41:351–357. 63. Schenborn ET, Goiffin V. DEAE-dextran transfection of mammalian cultured cells. Methods Mol. Biol 2000; 130:147–153.
Cutaneous Gene Therapy with Cultured Skin Substitutes
169
64. Loytner S, Scangos GA, Ruddle FH. Mechanisms of DNA uptake by mammalian cells: fate of exogenously added DNA monitored by the use of fluorescent dyes. Proc. Natl. Acad. Sci. USA 1982; 79:422–426. 65. Chen CA, Okayama H. Calcium phosphate–mediated gene transfer: a highly efficient transfection system for stably transforming cells with plasmid DNA. Biotechniques 1988; 6:632–638. 66. Gao X, Huang L. Cationic liposome–mediated gene transfer. Gene Ther 1995; 2:710–722. 67. Zellmer S, Gaunitz F, Salvetter J, Surovoy A, Reissig D, Gebhardt R. Long-term expression of foreign genes in normal human epidermal keratinocytes after transfection with lipid/DNA complexes. Histochem. Cell Biol 2001; 115:41–47. 68. Andreason GL, Evans GA. Introduction and expression of DNA molecules in eukaryotic cells by electroporation. Biotechniques 1988; 6:650–660. 69. Shigekawa K, Dower WJ. Electroporation of eukaryotes and prokaryotes: a general approach to the introduction of macromolecules into cells. Biotechniques 1988; 6:751. 70. Biewenga JE, Destree OH, Schrama LH. Plasmid-mediated gene transfer in neurons using the biolistics technique. J. Neurosci. Methods 1997; 71:67–75. 71. Thompson TA, Gould MN, Burkholder JK, Yang N-S. Transient promoter activity in primary rat mammary epithelial cells evaluated using particle bombardment gene transfer. In Vitro Cell Dev. Biol 1993; 29A:165–170. 72. Eglitis MA, Anderson WF. Retroviral vector for introduction of genes into mammalian cells. Biotechniques 1988; 6:608–614. 73. Flowers MED, Stockschlaeder MAR, Schuening FG, Niederwieser D, Hackman R, Miller AD, Storb R. Long-term transplantation of canine keratinocytes made resistant to G418 through retrovirus–mediated gene transfer. Proc. Natl. Acad. Sci. USA 1990; 87:2349–2353. 74. Kuroki T, Kashiwagi M, Ishino K, Huh N-H, Ohba M. Adenovirus-mediated gene transfer to keratinocytes—a review. J. Invest. Dermatol. Symp Proc 1999; 4:153–157. 75. Kovesdi I, Brough DE, Bruder JT, Wickham J. Adenoviral vectors for gene transfer. Curr. Opin. Biotechnol 1997; 8:583–589. 76. Rabinowitz JE, Samulski J. Adeno-associated virus expression systems for gene transfer. Curr. Opin. Biotechnol 1998; 9:470–475. 77. Dull T, Zufferey R, Kelly M, Mandel RJ, Nguyen M, Trono D, Naldini L. A third-generation lentivirus vector with a conditional packaging system. J. Virol 1998; 72:8463–8471. 78. Kuhn U, Terunuma A, Pfutzner W, Foster RA, Vogel JC. In vivo assessment of gene delivery to keratinocytes by lentiviral vectors. J. Virol 2002; 76:1496–1504. 79. Chen BF, Chang WC, Chen ST, Chen DS, Hwang LH. Long-term expression of the biologically active growth hormone in genetically modified fibroblasts after implantation into a hypophysectomized rat. Hum. Gene Ther 1995; 6:917–926. 80. Breitbart AS, Mason JM, Urmacher C, Barcia M, Grant RT, Pergolizzi RG, Grande DA. Geneenhanced tissue engineering: applications for wound healing using cultured dermal fibroblasts transduced retrovirally with the PDGF-B gene. Ann. Plast. Surg 2002; 43:632–639. 81. Supp DM, Bell SM, Morgan JR, Boyce ST. Genetic modification of cultured skin substitutes by transduction of keratinocytes and fibroblasts with platelet derived growth factor A. Wound Rep. Reg 2000; 8:26–35. 82. Inaba M, Toninelli E, Vanmeter G, Bender JR, Conte MS. Retroviral gene transfer: effects on endothelial cell phenotype. J. Surg. Res 1997; 78:31–36. 83. Zheng L, Dengler TJ, Kluger MS, Madge LA, Schechner JS, Maher SE, Pober JS, Bothwell ALM. Cytoprotection of human umbilical vein endothelial cells against apoptosis and CTL-mediated lysis provided by caspase-resistant Bcl-2 without alterations in growth or activation response. J. Immunol 2000; 164:4665–4671. 84. Schiaffino MV, Dell’Ambra E, Cortese KI, Baschirotto C, Bondanza S, Clementi M, Nucci P, Ballabio A, Tacchetti C, De Luca M. Effective retrovirus-mediated gene transfer in normal and mutant human melanocytes. Hum. Gene Ther 2002; 13:947–957. 85. Yao F, Eriksson E. Gene therapy in wound repair and regeneration. Wound Rep. Reg 2000; 8: 443–451.
170
Supp and Boyce
86. Danos O, Mulligan RC. Safe and efficient generation of recombinant retroviruses with amphotropic and ecotropic host ranges. Proc. Natl. Acad. Sci. USA 1988; 85:6460–6464. 87. Riviere I, Sadelain M. Methods for the construction of retroviral vectors and the generation of hightiter producers. In: Robbins PD, ed. Gene Therapy Protocols. Totowa, N.J.: Humana Press, 1997: 59–78. 88. Eming SA, Morgan JR. Methods for the use of genetically modified keratinocytes in gene therapy. In: Robbins PD, ed. Gene Therapy Protocols. Totowa, N.J.: Humana Press, 1997:265–280. 89. Greenhalgh DA, Rothnagel JA, Roop DR. Epidermis: an attractive target tissue for gene therapy. J. Invest. Dermatol 1994; 103:63S–69S. 90. Choate KA, Khavari PA. Direct cutaneous gene delivery in a human genetic skin disease. Hum. Gene Ther 1997; 8:1659–1665. 91. Rheinwald JG, Green H. Serial cultivation of strains of human epidermal keratinocytes: the formation of keratinizing colonies from single cells. Cell 1975; 6:331–343. 92. Jiang C-K, Connolly D, Blumenberg M. Comparison of methods for transfection of human epidermal keratinocytes. J. Invest. Dermatol 1991; 97:969–973. 93. Eming SA, Lee J, Snow RG, Tompkins RG, Yarmush ML, Morgan JR. Genetically modified human epidermis overexpressing PDGF-A directs the development of a cellular and vascular connective tissue stroma when transplanted to athymic mice–Implications for the use of genetically modified keratinocytes to modulate dermal regeneration. J. Invest. Dermatol 1995; 105:756–763. 94. Eming SA, Medalie DA, Tompkins RG, Yarmush ML, Morgan JR. Genetically modified human keratinocytes overexpressing PDGF-A enhance the performance of a composite skin graft. Hum. Gene Ther 1998; 9:529–539. 95. Supp DM, Supp AP, Bell SM, Boyce ST. Enhanced vascularization of cultured skin substitutes genetically modified to overexpress Vascular Endothelial Growth Factor. J. Invest. Dermatol 2000; 114:5–13. 96. Supp DM, Boyce ST. Overexpression of vascular endothelial growth factor accelerates early vascularization and improves healing of genetically modified cultured skin substitutes. J. Burn Care Rehabil 2002; 23:10–20. 97. Vailly J, Gagnouz-Palacios L, Dell’Ambra E, Romero C, Pinola M, Zambruno G, De Luca M, Ortonne JP, Meneguzzi G. Corrective gene transfer of keratinoyctes from patients with junctional epidermolysis bullosa restores assembly of hemidesmosomes in reconstructed epithelia. Gene Ther 1998; 5:1322–1332. 98. Dellambra E, Vailly J, Pellegrini G, Bondanza S, Golisano O, Macchia C, Zambruno G, Meneguzzi G, De Luca M. Corrective transduction of human epidermal stem cells in laminin-5–dependent junctional epidermolysis bullosa. Hum. Gene Ther 1998; 9:1359–1370. 99. Choate KA, Medalie DA, Morgan JR, Khavari PA. Corrective gene transfer in the human skin disorder lamellar ichthyosis. Nature Med 1996; 2:1263–1267. 100. Choate KA, Kinsella TM, Williams ML, Nolan GP, Khavari PA. Transglutaminase 1 delivery to lamellar ichthyosis keratinocytes. Hum. Gene Ther 1996; 7:2247–2253. 101. Alexeev V, Igoucheva O, Domashenko A, Cotsarelis G, Yoon K. Localized in vivo genotypic and phenotypic correction of the albino mutation in skin by RNA-DNA oligonucleotide. Nat. Biotechnol 2000; 18:43–47. 102. Alexeev V, Yoon K. Gene correction by RNA-DNA oligonucleotides. Pigment Cell Res 2000; 13: 72–79. 103. Gerrard AJ, Hudson DL, Brownlee GG, Watt FM. Towards gene therapy for haemophilia B using primary human keratinocytes. Nature Gen 1993; 3:180–183. 104. Page SM, Brownlee GG. An ex vivo keratinocyte model for gene therapy of hemophilia B. J. Invest. Dermatol 1997; 108:139–145. 105. White SJ, Page SM, Margaritis P, Brownlee GG. Long–term expression of human clotting factor IX from retrovirally transduced primary human keratinocytes in vivo. Hum. Gene Ther 2002; 9: 1187–1195. 106. Lynch SE, Colvin RB, Antonaides HN. Growth factors in wound healing: single and syneristic effects on partial thickness porcine skin wounds. J. Clin. Invest 1989; 84:640–646.
Cutaneous Gene Therapy with Cultured Skin Substitutes
171
107. Bennett NT, Schultz GS. Growth factors and wound healing: Part II. Role in normal and chronic wound healing. Am. J. Surg 1993; 166:74–81. 108. Beer H-D, Longaker MT, Werner S. Reduced expression of PDGF and PDGF receptors during impaired wound healing. J. Invest. Dermatol 1997; 109:132–138. 109. Heldin CH, Westermark B. Role of platelet-derived growth factor in vivo. In: Clark RAF, ed. The Molecular and Cellular Biology of Wound Repair. New York: Plenum Press, 1996:249–273. 110. Detmar M, Yeo KT, Nagy JA, Van de Water L, Brown LF, Berse B, Elicker BM, Ledbetter S, Dvorak HF. Keratinocyte-derived vascular permeability factor (vascular endothelial growth factor) is a potent mitogen for dermal microvascular endothelial cells. J. Invest. Dermatol 1995; 105:44–50. 111. Weninger W, Uthman A, Pammer J, Pichler A, Ballaun C, Lang IM, Plettenberg A, Bankl HC, Sturzl M, Tschachler E. Vascular endothelial growth factor production in normal epidermis and in benign and malignant epithelial skin tumors. Lab. Invest 1996; 75:647–657. 112. Brown LF, Yeo KT, Berse B, Yeo TK, Senger DR, Dvorak HF, Van de Water L. Expression of vascular permeability factor (vascular endothlial growth factor) by epidermal keratinocytes during wound healing. J. Exp. Med 1992; 176:13375–13379. 113. Losordo DW, Vale PR, Symes JF, Dunnington CH, Esakof DD, Maysky M, Ashare AB, Lathi K, Isner JM. Gene therapy for myocardial angiogenesis: initial clinical results with direct myocardial injection of phVEGF165 as sole therapy for myocardial ischemia. Circulation 1998; 98:2800–2804. 114. Baumgartner I, Pieczek A, Manor O, Blair R, Kearney M, Walsh K, Isner JM. Constitutive expression of phVEGF165 after intramuscular gene transfer promotes collateral vessel development in patients with critical limb ischemia. Circulation 1998; 97:1114–1123. 115. Rivard A, Silver M, Chen D, Kearney M, Magner M, Annex B, Peters K, Isner JM. Rescue of diabetes-related impairment of angiogenesis by intramuscular gene therapy with adeno-VEGF. Am. J. Pathol 1999; 154:355–363. 116. Bale SJ, Doyle SZ. The genetics of ichthyosis: a primer for epidemiologists. J. Invest. Dermatol 1994; 102:49S–50S. 117. Choate KA, Medalie DA, Morgan JR, Khavari PA. Corrective gene transfer in the human skin disorder lamellar ichthyosis. Nature Med 1996; 2:1263–1267. 118. Choate KA, Kinsella TM, Williams ML, Nolan GP, Khavari PA. Transglutaminase 1 delivery to lamellar ichthyosis keratinocytes. Hum. Gene Ther 1996; 7:2247–2253. 119. Uitto J, Pulkkinen L. Molecular genetics of heritable blistering disorders. Arch. Dermatol 2001; 137:1458–1461. 120. Dellambra E, Vailly J, Pellegrini G, Bondanza S, Golisano O, Macchia C, Zambruno G, Meneguzzi G, De Luca M. Corrective transduction of human epidermal stem cells in laminin-5–dependent junctional epidermolysis bullosa. Hum. Gene Ther 1998; 9:1359–1370. 121. Vailly J, Gagnouz-Palacios L, Dell’Ambra E, Romero C, Pinola M, Zambruno G, De Luca M, Ortonne JP, Meneguzzi G. Corrective gene transfer of keratinoyctes from patients with junctional epidermolysis bullosa restores assembly of hemidesmosomes in reconstructed epithelia. Gene Ther 1998; 5:1322–1332. 122. Robbins PB, Lin Q, Goodnough JB, Tian H, Chen X, Khavari PA. In vivo restoration of laminin 5 beta 3 expression and function in junctional epidermolysis bullosa. Proc. Natl. Acad. Sci. USA 2001; 98:5193–5198. 123. Seitz CS, Giudice GJ, Balding SD, Marinkovich MP, Khavari PA. B P180 gene delivery in junctional epidermolysis bullosa. Gene Ther 1999; 6:42–47. 124. Fenjves ES, Gordon DA, Pershing LK, Williams DL, Taichman LB. Systematic distribution of apolipoprotein E secreted by grafts of epidermal keratinocytes: implications for epidermal function and gene therapy. Proc. Natl. Acad. Sci. USA 1989; 86:8803–8807. 125. Fenjves ES, Smith J, Zaradic S, Taichman LB. Systemic delivery of secreted protein by grafts of epidermal keratinocytes: prospects for keratinocyte gene therapy. Hum. Gene Ther 1994; 5(10): 1241–1248. 126. Vogt PM, Thompson S, Andree C, Liu P, Breuing K, Hatzis D, Brown H, Mulligan RC, Ericksson E. Genetically modified keratinocytes transplanted to wounds reconstitute the epidermis. Proc. Natl. Acad. Sci. USA 1994; 91(20):9307–9311.
172
Supp and Boyce
127. Jensen UB, Jensen TG, Jensen PK, Rygaard J, Hansen BS, Fogh J, Kolvraa S, Bolund L. Gene transfer into cultured human epidermis and its transplantation onto immunodeficient mice: an experimental model for somatic gene therapy. J. Invest. Dermatol 1994; 103:391–394. 128. Wang X, Zinkel S, Polonsky K, Fuchs E. Transgenic studies with a keratin promoter-driven growth hormone transgene: prospects for gene therapy. Proc. Natl. Acad. Sci. USA 1997; 94:219–226. 129. Gerrard AJ, Hudson DL, Brownlee GG, Watt FM. Towards gene therapy for haemophilia B using primary human keratinocytes. Nature Gen 1993; 3:180–183. 130. Page SM, Brownlee GG. An ex vivo keratinocyte model for gene therapy of hemophilia B. J. Invest. Dermatol 1997; 108:139–145. 131. White SJ, Page SM, Margaritis P, Brownlee GG. Long-term expression of human clotting factor IX from retrovirally transduced primary human keratinocytes in vivo. Hum. Gene Ther 2002; 9: 1187–1195. 132. Lonnqvist F, Nordfors L, Schalling M. Leptin and its potential role in human obesity. J. Intern. Med 1999; 245:643–652. 133. Cau T, Wang X-J, Roop DR. Regulated cutaneous gene delivery: the skin as a bioreactor. Hum. Gene Ther 2000; 11:2297–2300.
9 Surface Properties of Polymeric Biomaterials and Their Modification for Tissue Engineering Applications Ays¸en Tezcaner and Vasif Hasirci Middle East Technical University Ankara, Turkey
I. INTRODUCTION Polymers have been used in fields as diverse as biomaterials, adhesion, protective coatings, composites, microelectronic devices, thin film technology, and scaffold development in tissue engineering. The success in all these fields depends on the surface properties (i.e., chemical composition, hydrophilicity, roughness, crystallinity, conductivity, and lubricity) of the polymer used. Surface chemistry and/or topography affects the type and the strength of interactions taking place at the biomaterial/biological environment (e.g., water and ion sorption; protein adsorption; cell adhesion, spreading and proliferation). Tissue engineering requires the reformation of an extracellular matrix generally from synthetic or biological polymers. In order to select appropriate substrates or matrices, it is necessary to understand the influence of the polymer on cell viability, growth, and function. The properties of natural tissues play a major role in the design and development of tissue engineering product development. It is critical to choose a biomaterial that does not elicit a cellular response. Depending on the requirements of targeted tissue either nondegradable or degradable biomaterials are chosen. Among these, synthetic and natural polymers are important elements of producing novel engineered tissue [1,2]. Among the biomaterials, several classes of polymers, (polyesters, polyanhydrides, poly(orthoesters), etc.) [3–6] have proven to be useful in biomedical applications. It has been known for a long time that cells differ in their ability to grow and differentiate, depending on both the chemistry and mechanics of their ECM substratum. Epithelial cells commonly express more differentiated functions when grown on immobilized ECM molecules (e.g., collagen, fibronectin, and laminin) compared to tissue culture plastic alone. These cells function even better when maintained on specialized artificial basement membrane substrata such as Matrigel威 (a complex of different basement membrane molecules that commonly supports cell differentiation) [7]. The control of cell adhesion to synthetic polymers is a key factor in biomedical applications, where such materials are used to support and stimulate tissue integration, tissue reconstruction, or cellular colonization [1]. Numerous in vitro experiments have shown that cell behavior 173
174
Tezcaner and Hasirci
is influenced by physicochemical properties of polymer surfaces. There are several driving forces for this physicochemical interaction. The most important of these are entropic, associated with the protein acting as a surfactant to release the highly ordered water near a hydrophobic interface. Of somewhat less importance, depending on the characteristics of the biomaterial, are electrostatic interactions at long protein distances from the surface and van der Waals interactions in short range [8]. Rates of cell migration on a polymer surface have been shown to depend on the concentration of preadsorbed adhesive proteins as well as the presence of soluble inhibitors to protein-mediated cell adhesion [9]. Substratum properties influencing cell behavior in culture include chemical group expression, hydrophilicity/hydrophobicity, physical and chemical anisotropy, and substratum contractility [10–14]. Surface energy of the substratum was shown to influence the adhesion of mammalian cells: high energy surfaces were generally reported to promote cell adhesion as opposed to low energy surfaces [15]. Cell adhesion appears to be maximized on surfaces with intermediate wettability [11–12,16,17]. van Wachem and colleagues [12] studied the adhesion of human endothelial cells onto a series of well-characterized methacrylate polymer surfaces with varying wettabilities and surface charges either in serum-containing or in serum-free culture media. Complete cell spreading was only observed on the positively charged copolymers. Copolymers containing negatively charged monomers showed cell adhesion only when the second monomer had a hydrophobic character.
II. SURFACE MODIFICATION METHODS The orientations and the morphology of the cells and the arrangement of intercellular material in the tissue-engineered construct are of vital importance for success. Surface modification techniques have been actively studied for achieving optimal tissue regeneration (Table 1). Surfaces of polymers can be modified to alter the chemical and physical properties without affecting bulk properties. These treatments have been applied to 1. 2. 3. 4. 5. 6. 7. 8.
Produce functional groups for specific interactions with other functional groups Change surface energy Change chemical inertness Introduce surface crosslinkages Modify surface morphology (i.e., roughness) Modify surface crystallinity Change surface electrical conductivity Change surface lubricity
Modification of surfaces is sometimes needed for biocompatibility concerns. The biocompatibility of a polymer is determined largely by specific interactions between the adsorbed proteins on the polymer surface and the receptors on the surface of cells. Other important parameters that are influential on biocompatibility, which are also controlled by controlling the surface, are the porosity and texture of the material. Efforts to improve biocompatibility have focused on reducing, enhancing, or selecting these interactions depending on the demands of intended application [8]. For example, reducing protein adsorption or selectively adsorbing proteins that lack cell-adhesive activities in order to produce surfaces that are more biologically inert than others improves blood compatibility of the surface with limited cell–surface interactions. Cell adhesion can be enhanced by increasing adsorption or by selectively adsorbing proteins responsible for adhesion to improve cell–material interactions and promote cell growth on the biomaterial in soft tissue engineering applications.
Surface Properties of Polymeric Biomaterials
175
Table 1 Approaches for Modification of Surfaces and Their Characterization for Biomedical Applications Material Poly (3-hydroxybutyrate-cohydroxyvalerate) Poly(3-hydroxybutyrate-cohydroxyvalerate) film and foams Polystyrene (PS) and poly(ethylene terephthalate) (PET)
Segmented PEOT/PBT copolymer Polypropylene, Polystyrene, polytetrafluoroethylene
Poly(D-lactic acid) film (PDLA)
Polycarbonate, poly(ethylene terephthalate), poly(vinylidine difluoride) Tissue culture grade polystyrene Glass
PLA-PEG monomethyl ether diblock copolymer Blends of pure soluble collagen with poly(acyrlic acid) or sodium vinylsulfonate)
Modification technique
Surface characterization
Ref.
Insulin immobilization with acrylic acid grafting via oxygen plasma discharge Oxygen plasma
ATR-FTIR, ESCA, water contact angle
33
SEM, AFM, water contact angle
29,30
Radiofrequency plasma deposition (methanol, acetone, formic acid, ethyleneoxide, allyl alcohol) Radiofrequency discharge of ethylene oxide, perfluoropropane, acetone, air, methanol, or water vapors Carbon dioxide plasma
ESCA
27
ESCA
13
SEM, water contact angle
6
Hyaluronic acid immobilization with argon and ammonia plasma treatment Plasma treatment followed by collagen modification Micropatterning from silicon and quartz substrates Oxygen plasma followed by type I collagen coating
FTIR
24
ESCA, SEM, water contact angle SEMO, cell culture functionality study ESCA, AFM, water contact angle
28
Coating with cyclic RGD peptides with UV laser lithography Micropatterning with UV laser lithography followed by chemical modification with synthetic peptides Copolymer synthesis
Cell culture functionality study (microscopical visualization) ESCA, AFM, fluorescence microscopy
77
AFM, ESCA,
35
Blending
ESCA
34
54 110
52
176
Tezcaner and Hasirci
Most conventional materials do not meet both the surface and bulk properties necessary to be considered as biomaterials. An effective approach for developing a clinically applicable biomaterial is to modify the surface of a material which has excellent bulk properties. Surface modifications fall into two categories: (1) chemically or physically altering the atoms, compounds, or molecules in the existing surface (e.g., etching, chemical modification) or (2) covering the existing surface with a material with different composition (e.g., coating, grafting, thin film deposition) [18,19]. Nondegradable and degradable synthetic biocompatible polymeric matrices can be used in tissue engineering. Although they have the advantage of being manufactured in complex, predetermined forms reproducibly, they are moderately suitable for cell attachment. Polymers contain few functional chemical groups suitable for protein immobilization. Techniques have been developed to render the polymers suitable for protein adsorption and cell adhesion. A variety of approaches have been used toward this end, including plasma modification; graft polymerization; grafting, and adsorption of biologically active or biomimetic molecules [16,20,24]. ˚ ), are Thin surface modifications, altering only the outermost molecular layers (3–10 A most desirable. However, in practice, thicker films are necessary to ensure complete and uniform coverage of the surface. Also, extremely thin layers may be more prone to surface reversal and mechanical erosion. Besides, surface modified layers should resist delamination in aqueous and proteinaceous solutions [18]. The major challenge in surface modification is to attain precise control over functional groups. Many surface modification methods produce a spectrum of functional groups such as hydroxyl, ether, carbonyl, and carboxyl, contrary to the intention of producing one specific functional group. The structure and chemical properties of polymeric surfaces vary with time and environment. Functional groups of homopolymers and block copolymers tend to reorient in response to different environments (Fig. 1). The low energy component tends to migrate to the polymer–air interface as a result of the thermodynamic drive to minimize the surface energy. In the aqueous environment, however, the hydrophilic groups are attracted to the surface. Atomic or molecular mobility must allow surface changes to occur in reasonable periods of time [18]. Such compositional reversal should be prevented or minimized by crosslinking, by sterically blocking the ability of surface structures to move, or by incorporating a rigid, impermeable layer between the substrate and the modified surface layer [18]. A variety of methods used in surface modifications are presented below.
Figure 1 Orientation of molecules with different polarities in different environments.
Surface Properties of Polymeric Biomaterials
177
A. Flame and Corona Treatment The equipment for corona and flame treatments is very simple and cost effective. These treatments can be used in continuous operation. Flame treatment, producing oxidized polymer surface due to high temperature (1000–2000⬚C) or excited species in the flame, has been used to treat polyolefin and other polymeric products for improved bondability with an adhesive and imprintability. The highly portable setup (a burner and a fuel tank) is suitable for surface treatment at any site (Fig. 2). The important parameters are air-to-gas ratio, gas flow rates, the distance between the tip of the flame and object, the nature of the gas, and treatment time. Corona discharge is one of the most widely used industrial surface treatments of continuous polyolefin films. A corona treatment system consists of a high voltage and high frequency generator, an electrode, and a grounded metal roll covered by an insulating material to prevent direct arc between two electrodes. A high voltage is applied across the electrodes to cause ionization of air. When the plasma is formed, a light blue color can be observed in the air gap. This atmospheric pressure plasma is called a corona discharge. Power applied and film speed determining the total amount of energy available for treatment are among the important parameters. Among the other variables are air-gap thickness, relative humidity, and treatment temperature. The nature of the surface-treated polymer should be taken into account when determining the optimal setting for each parameter. B. Glow Discharge (Plasma) Deposition Radiofrequency (rf) plasma etching and deposition has been used to modify the polymer surfaces. The technique utilizes an ionized gas environment to deposit thin polymer films, resulting in uniform, tightly adherent surface coatings with unique properties. Plasma thin films can be laid down on most solid substrates with microwave, radiofrequency, or acoustic activation. The energetic species in gas plasma include ions, electrons, radicals, metastables, and photons in the short-wave ultraviolet range. Different types of gases such as argon, oxygen, nitrogen, fluorine, and carbon dioxide have been used to produce unique surface properties. For example, oxygen plasma treatment can increase the surface energy of polymers, whereas fluorine plasma treatment can decrease the surface energy and can improve chemical inertness. Crosslinking at a polymer surface can be introduced by inert gas plasma [26]. The technique offers several advantages for the preparation of improved tissue culture substrates.
Figure 2 Schematic of a corona discharger.
178
Tezcaner and Hasirci
One major drawback in the plasma processing has been the difficulty in predicting the surface chemistry of the final polymer from the composition of deposited monomer. However, this problem was overcome by depositing at low substrate temperatures, which minimizes molecular fragmentation and results with surface structures that resemble the monomer. Other strategies involve the use of pulsed plasmas or the use of polymerizable double bonds [18]. Thin polymer films with unique chemical and physical properties are produced by plasma polymerization. There are several advantages of plasma deposited films. The coatings which could be applied to polymers, ceramics, metals, carbons, and glasses are extremely thin (20–200 ˚ ) and highly resistant to delamination. An extremely wide range of chemistries can be achieved A through deposition of hundreds of organic compounds. The processing is rapid and the plasmadeposited films exhibit no toxicity in soft tissue and cell culture studies. The plasma films demonstrated unique biological reactivity [13,14,27]. Chinn and his coworkers [13] used a radiofrequency plasma discharge of ethylene oxide or perfluoropropane vapor; their mixture; or vapors of acetone, air, methanol, and water to modify poly(ethylene terephthalate) (PET) and polystyrene (PS) surfaces (hydrophobic surfaces). Electron spectroscopy for chemical analysis (ESCA) showed that ethylene oxide plasma treatment of both PS and PET resulted in a modified, delamination-resistant surface that showed enhanced fibronectin (FN) adsorption from serum and that supported cell growth better than untreated tissue culture plastics. The participation of hydroxyl groups in FN adsorption was found to be consistent with ESCA results. Chinn and colleagues [13] found a good correlation between cell spreading and FN adsorption. It therefore shows that the polymer surfaces might be optimized for different cell types by altering the surface modification type and extent via plasma treatment (e.g., rf power, reaction time, pressure, gas, or gas flow rate). The surface of hydrophobic poly(lactic acid) (PLA) is difficult to modify by common chemical methods because there are no functional groups in the backbone. However, plasma technique is a convenient method for modifying the surface properties of such materails via introduction of desired groups or chains onto the surface. Yang et al. [28] modified poly(D,Llactide) by combining plasma treatment with tight collagen anchorage. The ammonia plasma pretreatment improved collagen anchorage, thereby also increasing cell affinity toward the surface. Hasirci and his group [17,29,30] used oxygen plasma treatment to change surface composition and to render the surface of poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV8)based matrices for different tissue engineering applications. Calcium phosphate–loaded PHBV8 foams were treated with rf oxygen plasma at 100 W for 10 min prior to osteoblast seeding. After 29 days of incubation, large size (300–500 m) sucrose-loaded PHBV8 (6 w/w%) foams treated with rf oxygen plasma at 100 W for 10 min were found to be the most suitable matrices for osteoblast growth [30]. PHBV8 film was also chosen as a temporary substrate for growing retinal pigment epithelium cells as an organized monolayer for their subretinal transplantation [17]. Power and duration were changed during plasma treatment. The effect of these two parameters on surface hydrophilicity, morphology, topography, surface composition, in vitro degradation, cell attachment, and cell growth rate were studied. As the total power applied increased, the water contact angle decreased, from 58.3 Ⳳ 7.8 to 43.3 Ⳳ 3.3. The effect of oxygen plasma treatment was pronounced on the texture of the films as observed by SEM and AFM. The O/C atomic ratio derived from XPS increased as a function of plasma power and duration [29]. The PHBV8 films treated at 100 W for 10 min were found to be the most suitable for reattachment of D407 cells in 24 h, and the were grown to confluency as an organized monolayer.
Surface Properties of Polymeric Biomaterials
179
C. Grafting Poly(ethylene terephthalate) films were modified by grafting of acrylic acid to subsequently allow immobilization of collagen types I and III to serve as a suitable template for urothelial cells [31]. Collagen adsorption on poly(acrylic acid)–grafted polymers not only rendered the matrix suitable for cell adhesion and proliferation, but also induced the stratification of urothelial cells to show cell differentiation. Such a cell–matrix could be applied in reparative surgery of the urinary tract. Previously, the same group showed the importance of extracellular matrix protein adsorption on surface-modified poly(acrylic acid)–grafted polymers on human smooth muscle cell adhesion and proliferation in vitro [32]. Radiation grafting and related methods have been used to modify the surfaces of biomaterials. Within this category, three types of reactions can be used: grafting using ionizing radiation sources (most commonly cobalt-60 gamma radiation source), grafting using UV radiation (photografting), and grafting using high energy electron beams. In all cases, the radiation breaks chemical bonds in the material to be grafted, forming free radicals, peroxides, or other reactive species. These reactive surface groups are then exposed to a monomer. The monomer reacts with the free radicals at the surface and propagates as a free radical chain reaction, incorporating other monomers in a surface-grafted polymer [19,20,23]. Kang et al. [33] modified the surface of PHBV (Biopol D600P) with insulin for tissue engineering applications through acrylic acid grafting prior to exposure to oxygen plasma. The PHBV films were exposed to an oxygen plasma discharge to produce peroxides on its surfaces which were then used as catalysts for the polymerization of acrylic acid (AA) for introduction of carboxyl group onto the surface. The carboxylic acid groups on the surface were activated, and the poly(ethylene oxide) was grafted onto the surface for coupling insulin as a final step. In the cell culture studies, the proliferation of human fibroblasts in the presence of serum was significantly accelarated on the insulin-immobilized PHBV.
D. Polymer Blends and Block Copolymers Multiphase polymers, including block and graft copolymers as well as blends, are known to exhibit distinct chemical and physical properties [6,34,35]. One approach to obtain biodegradable polymers that allow surface modification is the attachment of a poly(ethylene glycol) (PEG) to a biodegradable polymer such as PLA or PLGA. The hydrophilic PEG chains control protein and peptide adsorption, thereby allowing the control of cell behavior on the surface. To obtain biodegradable polymers with variable surface properties for tissue culture applications, Lucke et al. attached poly(ethylene glycol) blocks to poly(lactic acid) blocks in a variety of combinations. The resulting poly (D,L-lactic acid)–poly(ethylene glycol)–monomethyl ether (ME-PEGPLA) diblock copolymers were evaluated in terms of their suitability for drug delivery applications as well as for the manufacture of scaffolds in tissue engineering. The suppressive effect on the adsorption of two model peptides, namely calcitonin and human atrial natriuretic peptide, on poly(D,L-lactic acid)–poly(ethylene glycol)–monomethyl ether diblock copolymer was observed, suggesting a suitable scaffold for tissue engineering applications [35]. Deschamps and his coworkers [6] designed and evaluated segmented poly(ether ester) materials for the tissue engineering of bone. Their mechanical properties, degradability, and ability to sustain bone marrow cell growth make such segmented poly(ether esters) excellent materials for bone tissue engineering. PEOT/PBT multiblock copolymers were prepared by a two-step polycondensation in the presence of titanium tetrabutoxide as catalyst. PEOT/PBT
180
Tezcaner and Hasirci
scaffolds with varying porosities and pore sizes were prepared by molding and freeze-drying techniques in combination with particulate leaching. Surface modification by gas plasma treatment showed improvement in cell adhesion on all compositions of the copolymer. E. Surface Micropatterning It is claimed that the most effective way of adhering to a substrate is to mimic the natural environment of the cell at the submicron level and create the cell’s own adhesion mechanism: the focal contact. Surface microfabrication techniques involve strategies that employ variations in surface charge, hydrophilicity, and topography for regulating cellular attachment and functions, including growth, differentiation, and apoptosis via surface biomolecule immobilization [36,37]. The central hypotheses of biomimetic surface engineering are that the monolayer coatings of biologically active peptides affect cell attachment to the materials through biomolecular recognition events and that surfaces of three-dimensional structures modified with these peptides induce tissue formation preferentially with a specific cell type. Macromolecular assemblies are the key processes of life (i.e., nucleic acid synthesis, energy transduction, and control of cell growth and division). Attempts in biomimetics involve Simulation of surface topography (polymer patterning, positional assembly) to generate micro- and nanoenvironments of tissue for significant effects on cell adhesion, migration, function, and tissue integration Protein patterning and other macromolecular patterning to create chemical microenvironments via surface functionalization of biomaterials Mechanophysical patterning (micro- and nanoscale mechanical stresses generated by cellbiomaterial interactions) Cell patterning via precise positioning of cells on biomaterials to generate in vivo biological microenvironments which enable controlled homotypic and heterotypic interactions Nanostructured materials should involve surface patterned molecular arrays, nanoscale synthetic scaffolding mimicking the cell–extracellular matrix microenvironment, precise positioning of molecules with specific signals to provide microheterogeneity, composites of bioorganic and organic molecules, molecular layering (coating), and supramolecular self-assembly and selforganization (template-directed) assembly [38]. Molecular assembly is the spontaneous assortment of molecules under equilibrium conditions into stable, well-defined aggregates joined by noncovalent bonds such as ionic bonds, hydrogen bonds, hydrophobic interactions, and van der Waals interactions [39]. For their end use, highly functionalized biomaterials with an inherent capability (i.e., natural polymers such as proteins, polysaccharides, or some synthetic analogs) can be allowed to assemble into structures of a given size and geometry (positional assembly) [38]. This process is mainly driven by local geometry and molecular forces. The molecular layering (surface coating) resulting from repeated and alternating processing of the surface, often with oppositely charged molecules, is called template-directed assembly, and the template prepared by such method is used to create a complex initial pattern for subsequent self-assembly [40]. F. Different Approaches in Biomimetics 1. Polymer Patterning Patterning of polymers at micro- and nanoscale via self-assembly is carried out with microlithographic chemical processing and film deposition, etching methods that find use in both electronic device manufacturing and in other areas such as biology and medicine. Microfabrication tech-
Surface Properties of Polymeric Biomaterials
181
niques make it possible to produce patterns with varying geometries on the same matrix with a micrometric accuracy. Microislands of cells, either in homotypic with same or heterotypic interaction with another type of cell in a given proportion and geometrical arrangement, can be obtained. Micropatterned substrates prevent cells from migrating and proliferating outside the microdomains created and hence can keep adherent cells in well-defined and stable positions. The lower limit for cell culture pattern size may be explained by their need for proximity to the same or other cell types, sharing of diffusive substances, and cell–cell contact. For the first time, Bohanon et al. [41] created a patterned surface on poly(N-isopropylacrylamide) by coupling with either surface-bound aminosilanes or copolymerization with surfacebound methacrylsilane. The polymer-grafted regions were shown to inhibit the adhesion of neuroblastoma glioma hybrid cells. Several groups immobilized the thermoresponsive polymer on polystyrene [42,43]. Nanofibrous extracellular matrix generation is a nanoscale patterning process and is based on phase separation of the polymer in a solvent, followed by leaching of a porogen leaching from the matrix. The matrices prepared by this method have high surface-to-volume ratios [44]. Thermally induced gelation followed by solvent exchange and freeze-drying is another technique used for the preparation of such nanofibrous extracellular matrices [45]. 2. Protein Patterning Protein patterning can be defined as protein immobilization at specific locations in a two- or three-dimensional space [46]. The field of protein patterning is a critical technology that is finding applications in the creation of complex miniaturized devices such as biosensors with multiple sensing regions [47]. This method is also employed in the creation of engineered tissue and organ grafts by promoting patterned cell growth. Protein patterning is in fact protein immobilization on template surfaces. The simplest (albeit reversible) method for protein immobilization onto surfaces is physical adsorption. Proteins are large, amphipathic molecules which are intrinsically surface active. Protein adsorption results from attractive forces such as ionic, hydrophobic, or van der Waals interactions [48]. The competitive adsorption of proteins is important in the many interfacial phenomena that occur in the presence of mixtures of proteins. These phenomena include blood and tissue biocompatibility of biomaterials, the mechanisms of cell adhesion and cell culture on solid supports, fouling of contact lenses in tear fluids, etc. Competitive protein adsorption affects the cell adhesion mechanisms because it is known that the competition between fibronectin and vitronectin in the serum are different on different substrates [49,50]. Depending on the substrate to which the protein is expected to adsorb, variations in the ability of adsorbed adhesion proteins to influence cell adhesion arise possibly due to conformational or orientational changes in the adsorbed proteins that modulate the availability and potency of their binding domains. Surface properties of the biomaterial have a significant effect on the rate, extent, and the mechanism of adsorption. The most widely investigated surface property is hydrophobicity. Generally the more hydrophobic the surface, the greater the adsorption of the proteins [48]. Substrate-mediated physical and chemical guidance on the growth and alignment of cells (i.e., neurons, Schwann cells, epithelial cells, bone-derived cells) are among the interest areas of many researchers [51–54]. Initial experiments analyzing cell behavior on biomimetic surfaces were carried out with surfaces modified with peptides using mostly organosilane chemistry. Several groups have worked with photoresist technology of silanes to control cellular growth [55–60]. There are several approaches to this end. Photolithography is an expensive technique that has been used most extensively for patterning proteins and cells on planar substrates. Photolithography could be used to generate patterns
182
Tezcaner and Hasirci
by photoablating proteins preadsorbed to a silicon or glass surface (Hammarback et al., 1985). The other and most frequently applied approach is the immobilization of proteins on thiolterminated siloxane films that have been patterned by irradiation with UV (Fig. 3). In this approach, the substrate is covered by a spun cast photoresist. The photoresist is then exposed to ultraviolet irradiation with a mask over the substrate. Depending on the selected resist, the exposed regions of the photoresist will dissolve away or get fixed due to crosslinkage. The amino-terminated silanes come into contact with the patterned photoresist when they become bound to exposed regions. Finally the remaining photoresist is removed by sonicating in acetone, thus making the other regions available for exposure to hydrophobic, nonadhesive silanes to obtain a mixed monolayer interface. The self-assembled monolayers bound to the surface might have reactive end groups that can be activated for specific protein binding. EDS (N-(2-aminoethyl)-3-aminopropyl-trimethoxysilane) and DMS (dimethyldichlorosilane) patterned surfaces using photoresist technology have also been extensively used for studying cell growth and patterning of proteins [56,57]. Micropatterning the outgrowth of neurons and their morphological evaluation would provide a new approach to study neuron–neuron interactions in vitro. Knoll and his coworkers [52] fabricated a coplanar (alkane/amine) surface on a glass substrate via a deep UV photolithographic technique using aminopropyl dimethylethoxysilane. The patterned coplanar surfaces were then used to selectively attach synthetic peptides of laminin to amine surface regions. On the chemically defined surfaces, cerebellar neurons attached their cell bodies and grew on the modified surface by extending their neuritic processes along the boundaries of modified pathways. Soft lithography, a microfabrication technology not involving exposure to radiation, includes a set of related techniques, each of which uses stamps or channels fabricated in an elastomeric material for protein transfer [62,63]. Microcontact printing, patterning using microfluidic channels and laminar flow patterning, are inexpensive and can be used to pattern a variety of planar and nonplanar substrates. The key elements of soft lithography are the elastomeric stamps or masks. Elastomeric stamps or molds are prepared by casting a prepolymer of an elastomer that has the desired pattern structure. Most of the researchers use the durable elastomer poly(dimethylsiloxane) (PDMS). For the fabrication of masters with a desired pattern high resolution laser printing technology can be employed [64]. Model substrates with desired chemical micropatterns could be prepared using microcontact printing. The technique uses the pattern on the surface of an elastomeric PDMS stamp to form patterns of self-assembled monolayers (SAMs) on the surface of substrates. Compared to photolithography, microcontact printing is more suitable for biomedical applications because it involves fewer chemical treatments of the surface (Fig. 3) [65]. First, the stamp is inked with a solution of alkanethiol in ethanol and dried and brought into contact with gold block substrate for 10–20 s. During this short contact, the alkanethiol is transferred to the gold substrate. Subsequent exposure with another alkanethiol would produce a surface patterned into regions presenting different terminal groups. Most of the studies involving the patterning of proteins and cells using microcontact printing have used self-assembled monolayers. Mikos and his group [66] fabricated organized arrays of circular glass domains with a diameter of either 10 or 50 m surrounded and separated by regions modified with octadecyltrichlorosilane (OTS) self-assembled monolayers using a microcontact printing technique for in vitro retinal pigment epithelial (RPE) cell culture. RPE cells on the micropatterned surfaces had the typical RPE cell morphology and expressed the characteristic keratin intermediate filaments. It was shown that differentiated cell morphology was maintained throughout the in vitro cell culture period. This can be regarded as an important improvement over cells cultured on plain substrates such as glass, where cell dedifferentiation occured prior to confluency.
Surface Properties of Polymeric Biomaterials
183
Figure 3 Schematic of micropatterning techniques. (A) Conventional lithographic techniques with SAMs; (B) fabrication of micropatterned polymeric films; (C) microcontact printing onto gold substrates with SAMs.
The patterning of proteins by direct microcontact (microCP) relies on the adsorption to a substrate. Ideally, the proteins need to be firmly anchored onto a surface without adversely affecting their activity. Patel et al. [67] has exploited the high-affinity avidin–biotin receptor–ligand interaction to form arrays of avidin molecules onto a polymeric substrate expressing biotin moieties. Thereby any biotinylated species can be subsequently immobilized into defined patterns. For example, the micropatterned sample supported cell adhesion when biotin-(G)11GRGDS was bound to the avidin-bearing arrays [68]. Scholl et al. [69] have produced rectangular networks of functional rat hippocampal neurons on functionalized silicon oxide surfaces with a geometrical grid pattern of adhesion peptide
184
Tezcaner and Hasirci
PA22 (which matches in sequence a part of the A chain of laminin). PA22-2 was applied by contact printing onto the silicon oxide surface using a polydimethylsiloxane (PDMS) stamp and was immobilized by hetero-bifunctional crosslinking with sulfo-GMB. Attachment and network formation of the neurons were guided by the geometrical grid pattern. The immobilized neurons showed resting membrane potentials comparable with controls and, after 1 day of culture, were capable of eliciting action potentials. Two-dimensional neuronal model networks were formed on glass substrates as well as on microelectronic devices using microcontact printing by Offenhauser and his group [60]. The pattern structure consisted of lines with 3-, 5-, 8-, and 10-m widths crossing at 50 and 100 m in one and the other direction. The substrates were coated with aminosilane and a crosslinker. A PDMS stamp was used to transfer the peptides. Patterned biofunctionalization of the glass surfaces induced the hippocampal neurons to form a neuronal network of the same geometry. Patel et al. [68] used the microfluidic channel method to generate micron-scale patterns of any biotinylated ligand on the surface of a biodegradable block copolymer, polylactide– poly(ethylene glycol). The technique achieved control of biomolecule deposition with nanometer precision. Spatial control over cell development has been observed when using these templates to culture bovine aortic endothelial cells and PC12 nerve cells. Furthermore, neurite extension on the biodegradable polymer surface was observed to be directed with pattern features created. The patterns were composed of peptides containing the IKVAV sequence, suggesting achievement of directional control over nerve regeneration on biodegradable biomaterials [68]. Mrksich and his coworkers used microcontact printing and micromolding in capillaries to prepare tissue culture substrates in which both the topography and molecular structure of the interface could be controlled [70]. The method combined optically transparent contoured surfaces with self-assembled monolayers of alkanethiolates on gold to control interfacial characteristics. These tailored interfaces, in turn, controlled the adsorption of proteins and the attachment of cells. The technique used replica molding in polydimethylsiloxane molds having micrometerscale patterns on their surfaces to form a contoured film of polyurethane supported on a glass slide. Evaporation of a thin film of gold on this surface-contoured polyurethane provided an optically transparent substrate, on which SAMs of terminally functionalized alkanethiolates could be formed. In one application, a flat polydimethylsiloxane stamp was used to form a SAM of hexadecanethiolate on the raised plateaus of the contoured surface by contact printing hexadecanethiol; a SAM terminated in triethylene glycol groups was subsequently formed on the bare gold in the grooves by immersing the substrate in a solution of a second alkanethiol.Then this patterned substrate was immersed in a solution of fibronectin; the protein was adsorbed only on the methyl-terminated plateau regions of the substrate.The triethylene glycol–terminated regions resisted the adsorption of protein. Bovine capillary endothelial cells attached only on the regions that adsorbed fibronectin. Photochemical protein patterning methods use chemically labile species. For example, aryl azide photochemistry/UV irradiation of aryl azide results in an active nitrene which can insert C–H, C–C, C⳱C, N–H, O–H, or S–H bonds. Aryl dizirine chemistry/UV irradiation, on the other hand, results in an active carbene which can insert into C–H, C–C, C⳱C, N–H, O–H, or S–H bonds [71] or be deactivated by UV irradiation (such as conversion of thiol groups to sulfonates) [72]. This protein patterning on surfaces can be used for controlled cellular growth. Due to the potency of this approach, several investigators have used peptide or protein grafting to photochemically patterned substrates creating cell-reactive and cell-inert regions. Based on the photochemistry of the phenylazido group, Matsuda and Sugawara [73] prepared striped patterns of cell adhesive (nonirradiated regions of tissue culture dish) and nonadhesive domains (fixed photoreactive copolymer poly(N,N-dimethylacrylamide-co-3-azidostyrene) on tissue cul-
Surface Properties of Polymeric Biomaterials
185
ture plates using chemical microprocessing. Endothelial cells adhered, migrated, and proliferated on cell adhesive domains forming a stripe-patterned cellular sheet [73]. Integrins (cell surface receptors) bind to specific extracellular proteins, in particular to amino acid sequences comprising the adhesive regions. For example, the Tyr-Ile-Gly-Ser-Arg (YIGSR) sequence on the B1 chain of laminin promotes neural cell adhesion and outgrowth, while the Arg-Gly-Asp (RGD) sequence on both laminin and fibronectin influences numerous cell types. The most extensively used cell-binding domain to enhance cell adhesion onto biomaterial surfaces is the RGD sequence. However, other non-RGD-containing cell-binding domains such as YIGSR and Ile-Lys-Ala-Val (IKVAV) in laminin, Arg-Glu-Asp-Arg-Val (REDVR) and Leu-Asp-Val (LDV) in fibronectin, and Asp-Gly-Glu-Ala (DGEA) in collagen I and various heparin-binding domains are also present. Using surface modification techniques, it is possible to design a material for a specific cellular interaction or selectivity. Immobilization of active peptides along patterns was carried out by several researchers to regulate cell adhesion [5,42,59,74–78]. Matsuda and his coworkers [74] prepared a patterned octapeptide surface via immobilizing an octapeptide containing the RGD sequence on polyvinyl alcohol surface to culture endothelial cells. Endothelial cells only proliferated within the localized RGD regions [54,74]. These immobilized oligopeptides are surprisingly potent, and a very low number of these sequences are sufficient to dramatically influence cell behavior. Jeschke et al. coated the cell culture plates with tailor-made cyclic RGD peptides (bound to specific integrin receptors). Their aim was to mimic the physiological adhesion process of chondrocytes to the extracellular matrix by amplifying the cells for subsequent three-dimensional cartilage tissue engineering without any sign of dedifferentiation. Human chondrocytes seeded onto cyclic RGD peptide–coated tissue culture plate expressed integrins during a cultivation period of 20 weeks, and the receptors were proven to be functionally active as human and pig chondrocytes attached to RGD-coated surfaces [77]. In designing vascular grafts one of the biggest challenges is to permit attachment of endothelial cells but not blood platelets. Toward this end the tetrapeptide ligand REDV (ArgGlu-Asp-Val) for the receptor present only on endothelial cell but not on platelet was grafted to PEO-modified polymer via the terminal hydroxyl groups of PEO chains, and the designed graft was selective for endothelial cells [79]. In addition to specific peptide ligands, less specific peptide sequences in adhesion proteins with regions of uninterrupted cationic sequences (bound to cell surface proteoglycans through electrostatic interactions) have been used to promote cell adhesion [80]. Rezania et al. [81] used coupling of an aminofunctional organosilane [N-(2-aminoethyl)3-aminopropyl-trimethoxysilane] to the metal oxide surface and derivatizing the terminal amine to a maleimide surface by coupling a heterofunctional crosslinker [4-(N-maleimidomethyl) cyclohexane-1-carboxylate]. The maleimide-terminated surfaces were then used to couple a cellbinding domain of bone sialoprotein (BSP). This methodology ensured the free interaction of the molecule with the cell surface receptors [81,82]. It was shown that heterogenous mimetic peptide surfaces containing both the RGD (cellbinding) and heparin-binding domains enhanced adhesion, spreading, formation of discrete focal contacts, and organized cytoskeletal assembly [83,84]. Rezania and his coworkers modified the surfaces with both RGD (cell-binding) and FHRRIKA (putative heparin-binding) peptides unique to BSP, in the ratios of 75:25 and 50:50 [85]. These modified surfaces enhanced primary rat calvaria osteoblast-like cells and long-term events such as mineralization of the extracellular matrix compared to homogenous peptide surfaces and controls [85]. The poly(␣-hydroxyacid)s such as poly(lactic acid) (PLA) are one of the most preferable biodegradable polymer choices for scaffold preparation for tissue engineering. Due to the absence of suitable functionality for covalent grafting, many researchers have investigated methods for
186
Tezcaner and Hasirci
modifying the surface chemistry of these polymers. Common approaches involved complicated synthetic pathways and achieved limited densities of grafted biomolecules. Quirk et al. modified the surface of PLA with poly(L-lysine–GRGDS) conjugate (as a biomimetic surface modifier). They observed that large quantities of poly(L-lysine) (PLL) were attached on PLA surfaces via physical adsorption and remained attached even after repeated washings. This simple approach of biomolecule modification improved the initial cell–biomaterial interactions that are required for successful tissue engineered construct [78]. Apart from the initial adhesion and subsequent proliferation, the amount of matrix produced by cells grown on biodegradable scaffolds is particularly important in tissue engineering applications where scaffolds are expected to degrade over time. In order for the tissue to remain viable, the cells must replace the scaffold. While the initial adhesion was greatest on the higher density peptide surface, all cell types exhibited decreased matrix production on the more adhesive surfaces [86]. This study pointed out limitations of the use of bioactive materials in tissue engineering scaffolds. The proposed approach for overcoming this limitation was the use of exogenous factors such as TGF- for smooth muscle and endothelial cells on highly adhesive surfaces. 3. Polysaccharide Patterning In addition to synthetic polymers and proteins, various types of biological macromolecule derivatives (e.g., polysaccharide derivatives) can be immobilized to regulate cell functions [87–91]. Chen and his group [87] immobilized sulfated hyaluronic acid on polystyrene plate via coupling with azidoaniline. Hyaluronic acid–immobilized regions on the plate showed reduced platelet adhesion in comparison to unmodified sulfated hyaluronic acid. Two types of polysaccharides, sulfated hyaluronic acid and heparin, were immobilized on poly(ethylene terephthalate) and polystyrene film, respectively, in a specific pattern by lithography. The polysaccharides coupled with azidoaniline were cast on a film from an aqueous solution. After drying, the film was photoirradiated in the presence or absence of a photomask. It was shown that thrombus formation was reduced when platelet cells were cultivated on patterned sulfated hyaluronic acid areas. On the other hand, the presence of fibroblastic growth factor (FGF) enhanced the proliferation of fibroblast STO cells on the heparin-immobilized regions. This study indicated that the immobilization of heparin in the presence of FGF enhanced cell growth. Many tissue engineering applications require a three-dimensional scaffold or template conducive to cell attachment and maintenance of cellular functions. Polymeric foams with different porosities, densities, and bulk and surface chemistries have been fabricated for different tissue engineering applications. A polymer derivatized with covalent attachment of cell-specific ligands or extracellular signaling molecules would have the advantages of being interactive and having a geometry conducive to cell-to-cell interactions. Gutsche et al. [92] engineered porous carbohydrate-derivatized polystyrene foams for hepatocyte culture. Polystyrene foams fabricated by phase separation were derivatized with lactose and heparin, both of which are known to promote hepatocyte attachment and maintenance of differentiated functions. An increase in albumin secretion was observed within 3 days of culture and then decreased to initial values by the end of 7 days. Although the culture formulations used were more stringent and devoid of many hormones such as glucagon, epidermal growth factor, or dexamethasone (generally used other hepatocyte cultures), P450 activity could be maintained in the construct [92]. 4. Mechanochemical Patterning Cells in vital tissues align to form the most efficient configuration. Mechanochemical signaling at both micro- and nanometer levels can be used for bioartificial organ applications. It was
Surface Properties of Polymeric Biomaterials
187
hypothesized that an oriented cellular tissue for incorporation into vital, functioning, hybrid artificial organs can be prepared by periodically applying mechanical stresses on a hybrid tissue. The effect of cyclic stretching on the two-dimensional (2D) orientation response of arterial smooth muscle cells (SMCs) was studied by Kanda et al. [93]. Smooth muscle cells derived from bovine aortas were seeded onto transparent elastomeric membranes made of polyurethane and subjected to periodic stretching with various amplitudes from 5 to 20% at frequencies of 15 to 120 rpm for up to 24 h. It was shown that an applied mechanical stress induced a significant orientation response without morphologic alteration of the SMCs. Three-dimensional elastomeric scaffold with micrometric heterogeneity was used by Mulder and coworkers for terminal differentiation of skeletal myoblast culture. Porous threedimensional (3D) cell scaffolds were fabricated in a variety of shapes, thicknesses, and porosities by an immersion-precipitation method [94]. Mechanical analysis revealed that the constructs were elastomeric, recovering their original length following 100% elongation. The 3D substrates were seeded with muscle precursors. Following several weeks in culture, histological studies revealed the presence of multinucleated myotubes within the elastomeric material. In addition, immunohistochemical analysis indicated that the myotubes expressed the myosin heavy chain protein, suggesting that the myotubes had reached a state of terminal differentiation. 5. Topographical Patterning Both chemical and topographical properties of the biomaterial surface play crucial roles in determining the cellular response. Topographical effects on cell behavior were observed by R. G. Harrison in the last century when he cultured nerve cells on spider webs and coverslips [95]. Some of the responses of cells to substrate topography are cell orientation, cell adhesion, cell movement, gene expression, and orientation of cytoskeleton [37]. Microfabrication technology provides tools for producing well-defined and precisely controlled topographical features on various materials. Researchers have shown that the shape (such as ridges/grooves, spikes, holes, and spirals), the dimension, and the distribution of the features can have significant effects on cell behavior [96–99]. Some types of cells can react to features as small as 10 nm [100,101]. Nanotopographical features are usually made by electron beam lithography. Fused silica is a good choice creating patterned surfaces with grooves and ridges. The patterned substrates in the work of WojciakStothard et al. [100] were made as follows. First a photomask was created using electron beam lithography. Following the preparation of the photomask, the substrate was coated with the photoresist and exposed to UV through patterned photomask. After developing the resist the desired pattern on the substrate was obtained. At this stage the substratum is ready for etching. After etching, the photoresist was removed by a strong solvent. Laser ablation could be used to remove materials either on a point-to-point basis with a moving beam or through a mask. Embossing or casting methods into polymers and perhaps into metallic surfaces can carry out the replication of such structures. In the work of Tan and Saltzman, parallel ridges/grooves were prepared where widths of 2 m and lengths of 400 m were kept constant, and height and repeat spacing were varied [99]. They were applied on glass surfaces using photosensitive polyimide and were used to investigate the migratory behavior of human neutrophils on the patterned surfaces. The neutrophils moved in the direction of the long axis of ridges/grooves regardless of the topographical geometry and composition, which was consistent with the phenomenon known as contact guidance. However, it was observed that the rate of cell movement was strongly dependent on the microgeometry of the ridges. From this study, it could be concluded that parallel ridges/grooves can be used to control the direction and rate of cell migration.
188
Tezcaner and Hasirci
Duncan et al. [102] used a laser photoablation technique for microsculpturing polymer surfaces using a laser excimer KrF beam coupled with a microphotolithographic projection technique. The laser beam was focused onto the polymer surface through a mask at the micron level. Reproducible submicron three-dimensional structures without any sign of thermal effects during photoablation were produced on model poly(ethylene terephthalate) surfaces. The use of laser excimer technique for fabrication of two- and three-dimensional model surfaces for biomaterials presents several advantages over conventional microfabrication techniques. First, it is a rapid, extremely versatile, and biocompatible technique. It excludes the need for a photoresist and the drawbacks of thin film resist coating (i.e., time consuming steps, imperfect coating, removal of photoresist, elimination of solvents that can alter polymer surface morphology or denature proteins). These microfabricated model surfaces were used to study the effects of microtopography on cell-preferential cell deposition (osteoprogenitor cells from human bone marrow) and orientation. Preferential cell deposition was observed on surfaces with smooth microtopographical transitions (i.e., minimal microgroove width and depth).
II. SURFACE ANALYSIS The structure and chemistry of surfaces differ from those of the bulk for virtually all materials. Surfaces control the immediate interactions with the local environment. Biological response to biomaterials that do not contain extractable or leachable compounds is, largely, the direct response of proteins and cells adsorbed to the few molecular layers of the surface. The most important parameters for characterization of biomaterial surface are roughness, wettability, surface mobility, chemical composition, and crystallinity. There are various surface characterization techniques with different depths of analysis. For complete characterization, more than one method should be used [103,104]. Many of the effects of surface modification appear to be secondary to increased adsorption of cell attachment proteins such as fibronectin (Fn) and vitronectin (Vn) to the surface [50]. Complete characterization of the polymer including both bulk and surface properties is critical to understanding the nature of cell–polymer interactions. In general ‘‘surface’’ is not a well-defined term. The meaning of surface changes according to the surface analysis technique used because both the mechanism and the size of sampling depth of each technique differ. Ion-scattering spectroscopy (ISS) and static secondary ion mass spectrometry (SIMS) have a sampling depth of a few tenths of a nanometer, in which low energy ions are scattered or stopped by the first few layers of surface atoms. The sampling depth for techniques such as x-ray photoelectron spectroscopy (XPS) and Auger electron spectroscopy (AES), determined by the kinetic energy of emitted electrons, is typically 1 to 10 nm. However, the sampling depth can be increased up to a micrometer in other spectroscopic methods such as attenuated total internal reflection (ATIR) and energy dispersive x-ray (EDX) [105]. Apart from the sampling depth, surface information obtained, analysis environment, and sample suitability are among the important parameters that should be kept in mind while choosing the surface analysis technique. Each technique supplies different and often complementary information. When high-resolution, three-dimensional images of surfaces are needed, atomic force microscopy (AFM), scanning electron microscopy (SEM), and scanning tunnelling microscopy (STM) are among the surface analysis techniques used frequently. However, if there is a high demand for a surface-sensitive probe, contact angle measurements and SIMS are good choices. Meanwhile quantification and chemical state information are given by XPS.
Surface Properties of Polymeric Biomaterials
189
A. Contact Angle Measurement Contact angle measurement is also used extensively in studying the changes in polymer surface composition caused by various surface treatment techniques, aging characteristics of modified surfaces, and migration of hydrophobic and hydrophilic functional groups in aqueous and nonaqueous environments and also for determining optimal treatment times for various surface treatment processes. The principle of surface characterization depends on estimation of surface ˚ . When energy by determining liquid wetting of surfaces. The depth of analysis is within 3–20 A a drop of liquid is placed on a solid surface, it makes a definite angle between the liquid and solid phases owing to its larger surface tension compared to the surface tension of the solid. The contact angle observed on smooth, chemically homogenous, and rigid surfaces, either by the addition of a small amount of water on the surface or receded by substraction, indicates the thermodynamic equlibrium state (the lowest energy state). If the same liquid is placed on surfaces of increasing surface tension, the contact angle decreases as the surface tension of the solid increases. Drop and bubble methods are the two commonly used direct methods for measuring contact angles. The methods involve measurement of the profile of a liquid drop (sessile drop method) and of a bubble resting on a solid surface (captive bubble method). The Wilhelmy plate method is a simple method. The system comprises an electrobalance which measures the wetting force at the solid–liquid interface as a function of time. Contact angle measurements are frequently used with other surface analysis techniques such as x-ray photoelectron spectroscopy and static secondary ion mass spectrometry to monitor the changes in chemical composition of polymer surfaces. In particular, water contact angle has been used to study plasma-treated and corona discharge–treated surfaces [13,29,33]. Oxygen or nitrogen plasma treatments generally produce polar functional groups on the surface. The increase in the concentration of these polar groups tends to decrease the water contact angle [13–17,29]. There are several concerns in contact angle measurements. The measurement is operation dependent. Surface heterogeneity and roughness influence the results. The liquids used can reorient the surface structure, which affects the results. However, contact angle measurements provide a ‘‘first line’’ of characterization of materials. B. Atomic Force Microscopy The atomic force microscope can produce three-dimensional images of solid surfaces at high resolution. It can be used in the imaging of nonconducting samples such as polymers and ceramics. A typical commercial AFM apparatus consists of a piezoelectric scanner, which controls the scanning motion and an optical head that senses cantilever deflection signal [106]. It operates by scanning across the surface with a sharp tip mounted on a soft cantilever spring, which are microfabricated from silicon, silicon oxide, or silicon nitride. The cantilever deflects as a result of sample surface features moving under the tip. An optical system is used to sense the position of the tip relative to the sample. The atomic force microscope is capable of sampling small areas. There is no need for fixation of the biological specimen. The biological specimens in their alive state can be visualized and analyzed in an aqueous environment down to the molecular level without causing any damage. However, the only limiting parameter for resolution is the radius of the AFM tip, which is normally in the range of 10–20 nm. The depth of analysis ˚ , with a spatial resolution of 1 A ˚. is within 10–250 A C. Scanning Tunnelling Microscopy Scanning tunnelling microscopy measures the tunnelling current between the probe and the sample when the tip is scanned over the surface, revealing three-dimensional pictures of a
190
Tezcaner and Hasirci
conductive sample. The STM method measures electrical current and is therefore suited for ˚. conductive and semiconductive surfaces. The depth of analysis is 5 A The scanning tunnelling microscope can be operated in two modes: constant current and constant height. Either a thin conducting layer should be deposited when studying the polymer surface structures, or polymer samples can be solvent cast on the surface of highly oriented pyrolytic graphite as a very thin layer. D. Scanning Electron Microscopy Scanning electron microscopy words (SEM) by focusing and rastering a relatively high energy electron beam on a specimen. Low energy secondary electrons are emitted from each spot where ˚ with a spatial resolution of 40 the focused electron beam impacts. The depth of analysis is 5 A ˚ . The detectable intensity of secondary electron emission is a function of the atomic composition A of the sample and the geometry of the surface. Nonconductive materials are typically coated with gold prior to observation in order to minimize negative charge accumulation from the electron beam. The specimen preparation for biological systems requires fixation; therefore, the image produced does not reflect the natural state of the specimen. Moreover, working under high vacuum state also affects the specimen under investigation. SEM, in spite of these limitations in providing true surface information, is an important complementary method for other surface characterization techniques. Surface roughness and texture influence data gathered from ESCA, SIMS, and water contact angle determinations; SEM provides important information in the interpretation of these data. Environmental scanning electron microscopy (ESEM) is a novel technique that has appeared in recent years. It can be applied to observe the topography [107,108] or to study the surface structures and defects in crystals [109]. The most outstanding advantage of ESEM is that it can operate at pressures 104 –105 times higher than conventional SEM, which avoids surface charge accumulation. This allows the examination of uncoated and insulating samples. Low voltage SEM has been used to study platelets and phase separation in polymers. Also, environmental SEM permits wet, uncoated specimens to be studied. E. X-Ray Photoelectron Spectroscopy X-ray photoelectron spectroscopy, also known as electron spectroscopy for chemical analysis (ESCA), is the most widely used technique in characterization of polymer surfaces. A beam of x-rays irradiates a sample. The interaction between an x-ray photon and the inner shell electron causes a complete transfer of photon energy to the electron. The kinetic energy of the emitted electron is measured by an electron energy analyzer. The difference between kinetic energy of the photoelectron and the x-ray photon energy of the inner shell electron allows identification of the element. The binding energy of the inner shell electron is also sensitive to the electronic environment of the atom. When an atom is bonded to atoms of different elements of differing electronegativity, the binding energy is called chemical shift, which can be used to provide structural information for a molecule. A sampling depth of 3–5 nm is typical for XPS. The ESCA instrument should provide for x-ray excitation of a specimen and subsequent electron detection as a function of energy. This must be done in ultrahigh vacuum. The ESCA x-ray source must deliver a high-intensity monoenergetic photon flux to the analysis surface. Electron energy analyzers are used to measure the distribution of photoelectron energies [103]. The ESCA technique has many advantages and a few disadvantages for studying biomaterials. The speed of analysis, high information content, low damage potential, and ability to analyze without specimen preparation are among the advantages. The disadvantages include the need
Surface Properties of Polymeric Biomaterials
191
of vacuum compatibility, the possibility of sample damage during possible long analysis times, the need for expertise, and the cost of analysis. F. Static Secondary Ion Mass Spectrometry The surface is bombarded with a primary ion beam of low current density (typically 10ⳮ-11 to 10ⳮ8 A/cm2) sputtering largely secondary neutral species and a small fraction of secondary positive and negative ions in SIMS. A mass analyzer measures the mass of the ions. The detailed qualitative analysis of positive and negative ion spectra gives structural and chemical information about the surface of the polymer. The sampling depth for SSIMS is approximately 1 nm, and it is sensitive enough to detect amounts less than a monomolecular layer. SSIMS can differentiate among polymers that have similar XPS spectra and therefore complements XPS analysis. Like ESCA, it requires complex instrumentation and an ultra high vacuum for operation.There are two types of SSIMS, depending on the ion dose used: dynamic and static. Dynamic SSIMS uses high ion doses. The primary ion beam sputters so much material from the surface that the surface erodes at an appreciable rate. A depth profile can be constructed with information from the outermost atoms to a micron or more into the specimen. Static SSIMS, on the other hand, causes minimal surface destruction. During the period of analysis less than one monolayer of surface atoms is sputtered. Since extensive degradation, and rearrangements do not take place, large fragments can be ejected into the vacuum for measurement.
CONCLUSION Among the criteria for selection of suitable biomaterials for tissue engineering scaffolds, surface properties of the biomaterial are the most important. The biomaterial–cell interactions take place at the interface and influence cellular behavior. Most of the biomaterials with adequate bulk properties and insufficient surface characteristics require surface treatments. These modifications enhance biocompatibility and maximize cellular attachment, spreading, proliferation, and differentiation. Surface engineered materials mimicking the natural environment of the cell at the submicron level would provide a significant leap toward achieving viable bioartificial organs. Different approaches (chemical, physical, and microfabrication techniques) are employed to modify the surfaces of biomaterials to meet the needs for a scaffold. The analysis and characterization of engineered surfaces are of utmost importance. The information gathered from complementary surface analysis techniques together with cell culture functionality studies should yield the essential information about the most appropriate surfaces.
REFERENCES
1. Hubbel JA. Biomaterials in tissue engineering. Biotechnology 1995; 13:565–576. 2. Vacanti JP, Vacanti CA. The challenge of tissue engineering. In: Lanza RP, Langer R, Chick WL, eds. Principles of Tissue Engineering. New York: Academic Press, 1997:1–6. 3. Vozzi Christopher G, Flaim J, Bianchi F, Ahluwalia A, Bhatia S. Microfabricated PLGA scaffolds: a comparative study for application to tissue engineering. Mater. Sci. Eng 2002; 31:43–47. 4. Landers R, Hu¨bner U, Schmelzeisen R, Mu¨lhaupt R. Rapid prototyping of scaffolds derived from thermoreversible hydrogels tailored from application in tissue engineering. Biomaterials 2002; 23: 4437–4447.
192
Tezcaner and Hasirci
5. Burdick JA, Anseth KS. Photoencapsulation of osteoblast in injectable RGD-modified hydrogels for bone tissue engineering. Bioamterials 2002; 23:4315–4323. 6. Deschamps AA, Claase MB, Sleijster WJ, Brujn JD, Grijpma DW, Feijen J. Design of segmented poly(ether ester) materials and structures for the tissue engineering of bone. J. Contr. Rel 2002; 78: 175–186. 7. Ingber D. Extracellular matrix. In Cellular Mechanics and Tissue Engineering Eugene Bell. Basel: Birkhauser, 1993:69–83. 8. Hubbel JA. Chemical modification of polymer surfaces to improve biocompatibility. TRIP 1994; 2(1):20–25. 9. Saltzman WM. Cell interactions with polymers. In: Lanza R, Langer R, Chick W, eds. Principles of Tissue Engineering. New York: Academic Press, 1997:225–246. 10. Lydon MJ, Minett TW, Tighe BJ. Cellular interactions with synthetic polymer surfaces in culture. Biomaterials 1985; 6:396–402. 11. Horbett TA, Schway MB. Correlations between mouse 3T3 cell spreading and serum fibronectin adsorption on glass and hydroxyethylmethacrylate–ethylmethacrylate copolymers. J. Biomed. Mater. Res 1988; 22:763–793. 12. Van Wachem PB, Hogt AH, Beugeling T, Feijen J, Bantjes A, Detmers JP, Van Aken WG. Adhesion of cultured human endothelial cells onto methacrylate polymers with varying surface wettability and charge. Biomaterials 1987; 8:323–328. 13. Chinn JA, Horbett TA, Ratner BD, Schway MB, Haque Y, Hauschka SD. Enhancement of serum fibronectin adsorption and the clonal plating efficiencies of Swiss mouse 3T3 fibroblast and MM14 mouse myoblast cells on polymer substrates modified by radiofrequency plasma deposition. J. Colloid. Interfac. Sci 1989; 127(1):67–86. 14. Ertel SI, Ratner BD, Horbett TA. Radiofrequency plasma deposition of oxygen containing films on polystyrene and poly(ethylene terephthalate) substrate improves endothelial cell growth. J. Biomed. Mater. Res 1990; 24:1637–1659. 15. Lee JH, Lee SJ, Khang G, Lee HB. Interaction of fibroblasts on polycarbonate membrane surfaces with different micropore sizes and hydrophilicity. J. Biomater. Sci. Polym Ed 1999; 10(3):283–294. 16. Ikada Y. Surface modification of polymers for medical applications. Biomaterials 1994; 15:725–736. 17. Tezcaner A. Tissue engineered retinal pigment epithelium. PhD Thesis, METU (Turkey). 2001. 18. Ratner BD. Surface modification of polymers: chemical, biological and surface analytical challenges. Biosensors and Bioelectronics 1995; 10:797–804. 19. Ratner BD, Hoffman AS. Thin films, grafts and coatings, in Biomaterials Science: An Introduction to Materials in Medicine. Ratner BD, Hoffman AS, Schoen FJ, Lemons JE, eds. San Diego: Academic Press, 1996:105–118. 20. Elbert DL, Hubbel JA. Surface treatments of polymers for biocompatibility. Annu. Rev. Mater. Sci 1996; 26:365–394. 21. Holland J, Hersh L, Bryhan M, Onyiriuka E, Ziegler L. Culture of human vascular endothelial cells on an RGD-containing synthetic peptide attached to a starch-coated polystyrene surface—comparison with fibronectin-coated tissue grade polystyrene. Biomaterials 1996; 17:2147–2156. 22. Chen G, Ito Y, Imanishi Y. Regulation of growth and adhesion of cultured cells by insulin conjugated with thermoresponsive polymers. Biotechnol. Bioeng 1997; 53:339–344. 23. Ito Y, Chen G, Imanishi Y. Photoimmobilization of insulin onto polystyrene dishes for protein-free cell culture. Biotechnol Prog 1996; 12(5):700–702. 24. Mason M, Vercruysse KP, Kirker KR, Frisch R, Marecak DM, Prestwich GD, Pitt WG. Attachment of hyaluronic acid to polypropylene, polystyrene, and polytetrafluoroethylene. Biomaterials 2000; 21:31–36. 25. Ratner BD, Yoon SC. Polyurethane surfaces: Solvent and temperature induced structural rearrangements. In: Andrade JD, ed. Polymer Surface Dynamics. New York: Plenum Press, 1988:137–152. 26. Yasuda H. Plasma Polymerization. New York: Academic Press, 1985. 27. Ertel SI, Chilkoti A, Horbett TA, Ratner BD. Endothelial cell growth on oxygen containing films deposited by radiofrequency plasmas: the role of surface carbonyl groups. J. Biomater. Sci. Polym Ed 1991; 3:163–183.
Surface Properties of Polymeric Biomaterials
193
28. Yang J, Bei J, Wang S. Enhanced cell affinity of poly(D,L-lactide) by combining plasma treatment with collagen anchorage. Biomaterials 2002; 23:2607–2614. 29. Hasirci V, Tezcaner A, Hasirci N, Su¨zer S¸. Oxygen plasma modification of poly(3-hydroxybutyrateco-3-hydroxyvalerate) film surfaces for tissue engineering purposes. J. Appl. Polym Sci 2002; 87: 1285–1289. 30. Ko¨se-Torun G. PhD Thesis. Bone tissue generation on biodegradable polymeric scaffold. METU (Turkey), 2002. 31. Bisson I, Hilborn J, Wurm F, Meyrat B, Frey P. Human urothelial cells grown on collagen adsorbed to surface modified polymers. Urology 2002; 60:176–180. 32. Bisson I, Bisson I, Kosinski M, Ruault S. Acrylic acid–grafting and collagen immobilisation on poly(ethylene terephthtalate) surfaces for adherence and growth of human bladder smooth muscle cells. Biomaterials 2002; 15:3149–3158. 33. Kang IK, Choi SH, Shin DS, Yoon SC. Surface modification of polyhydroxyalkanoate films and their interaction with human fibroblasts. Int. J. Biol. Macromolecules 2001; 28:205–212. 34. Cascone MG, DiPasquale G, La Rosa AD, Cristallini C, Barbani N, Recca A. An ESCA study of new collagen-based bioartificial polymeric materials. Polymer 1998; 39:6357–6361. 35. Lucke A, Tebmar J, Schnell E, Schmeer G, Go¨pferich A. Biodegradable poly(D,L-lactic acid)–poly(ethylene glycol)–monomethyl ether (Me-PEG-PLA) diblock copolymers: structures and surface properties relevant to their use as biomaterials. Biomaterials 2000; 21:2361–2370. 36. Ito Y. Surface micropatterning to regulate cell functions. Biomaterials 1999; l20:2333–2342. 37. Curtis A, Riehle M. Tissue engineering: the biophysical background. Phys. Med. Biol 2001; 46: R47–R65. 38. Prokop A. Bioartificial organs in the twenty-first century: nanobiological devices. In: Hunkeler D, Cherrington A, Prokop A, Rajotte R, eds. Annals of the New York Academy of Sciences. 2001; 944:472–487. 39. Whitesides GM, Mathias JP, Seto CT. Molecular self-assembly and nanochemistry: a chemical strategy for the synthesis of nanosytems. Science 1991; 254:1312–1319. 40. LaVal JM, Chopineau J, Thomas D. Nanotechnology: R&D challenges and opportunities for application in biotechnology. Trends Biotechnol 1995; 13:474–481. 41. Bohanon T, Elender G, Knoll W, Koberle P, Lee JS, Offenhauser A, Ringsdorf H, Sackmann E, Simon J, Tovar G, Winnik FM. Neural pattern formation on glass and oxidized silicon surfaces modified with poly(N-isopropylacrylamide). J. Biomater. Sci. Polym. Ed 1996; 8:19–39. 42. Ito Y, Chen G, Guan Y, Imanishi Y. Patterned immobilization of thermoresponsive polymer. Langmuir 1997; 13:2756–2759. 43. Chen G, Imanishi Y, Ito Y. Effect of protein and cell behavior on pattern grafted thermoresponsive polymer. J. Biomed. Mater. Res 1998; 42:38–44. 44. Widmer MS, Gupta PK, Lu L. Manufacture of porous biodegradable polymer conduits. Biomaterials 1998; 19:1945–1955. 45. Zhang R, P.X. Ma. Synthetic nanofibrillar extracellular matrices with predesigned macroporous architectures. J. Biomed. Mater. Res 2000; 52:430–438. 46. Blawas AS, Reichert WM. Protein patterning. Biomaterials 1998; 19:595–609. 47. Lowe CR, Early FGP. Device incorporating a biochemical ligand. US Patent: 4,562,157, 1985. 48. Brash JL, Horbett TA. Proteins at interfaces. ACS Symposium Series Proteins at Interfaces II: Fundamentals and Applications. 1995:1–23. 49. Horbett TA, Schway MB. Correlations between mouse 3T3 cell spreading and fibronectin adsorption on glass and hydroxylethymethacrylate–ethylmethacrylate copolymers. J. Biomed. Mater. Res 1988; 22:763–793. 50. Steele JG, Dalton BA, Johnson G, Underwood PA. Adsorption of fibronectin and vitronectin onto Primaria culture polystyrene and relationship to the mechanism of initial attachment of human vein endothelial cells and BHK-21 fibroblasts. Biomaterials 1995; 16:1057–1067. 51. Clark P, Connolly P, Dow JAT, Wilkinson CDW. Cell guidance by ultrafine topography in vitro. J. Cell Sci 1991; 99:73–77. 52. Matsuzawa M, Umemura K, Beyer D, Sugioka K, Knoll W. Micropatterning of neurons using organic substrates in culture. Thin Solid Films 1997; 305:74–79.
194
Tezcaner and Hasirci
53. McFarland CD, Thomas CH, DeFilippis C, Steele JG, Healy KE. Protein adsorption and cell atachment to patterned surfaces. J. Biomed. Mater. Res 2000; 49:200–210. 54. Miller C, Shanks H, Witt A, Rutkowski G, Mallapragada S. Oriented Schwann cell growth on micropatterned biodegradable polymer substrates. Biomaterials 2001; 22:1263–1269. 55. Kleinfeld D, Kahler KH, Hockberger PE. Controlled outgrowth of dissociated neurons on patterned substrates. Neuroscience 1988; 8:4098–4120. 56. Lom B, Healey K, Hockberger P. A versatile technique for patterning biomolecules onto glass coverslips. J. Neurosci. Methods 1993; 50:385–397. 57. Britland S, Perez-Arnaud E, Clark P, McGinn B, Connolly P, Moores G. Micropatterning proteins and synthetic peptides on solid supports: a novel application for microelectronics fabrication technology. Biotechnol Prog 1992; 8(2):155–160. 58. Masuda T, Inouke K, Tani N. Control of cell arrangement. US Patent 5,202,227, 1993. 59. Sugawara T, Matsuda T. Photochemical surface derivatization of a peptide containing Arg-GlyAsp. (RGD). J. Biomed. Mater. Res 1995; 29:1047–1052. 60. Spro¨sler C, Scholl M, Denyer MC, Krause M, Nkajima K, Maelicke A, Knoll W, Offenhauser A. Model network architectures in vitro on extracellular recording systems using microcontact printing. Synthetic Metals 2001; 117:281–283. 61. Hammarback JA, Palm SL, Furcht LT, Letourneau PC. Guidance of neurite outgrowth by pathways of substratum adsorbed laminin. J. Neurosci. Res 1985; 13:213–220. 62. Jackman RJ, Wilbur JL, Whitesides GM. Fabrication of submicrometer features on curved substrates by microcontact printing. Science 1995; 269:664–665. 63. Kane RS, Takayama S, Ostuni E, Ingber DE, Whitesides GM. Patterning of proteins and cells using soft lithography. Biomaterials 1999; 20:2363–2376. 64. Qin D, Xia Y, Whitesides GM. Rapid prototyping of complex structures with feature sizes larger than 20 m. Adv. Mater 1996; 8:917–919. 65. Kumar A, Whitesides GM. Features of gold having micrometer to centimeter dimensions can be formed through a combination of stamping with an elastomeric stamp and an alkanethiol ink followed by chemical etching. Appl. Phys. Lett 1993; 63:2002–2004. 66. Lu L, Kam L, Hasenbein M, Nyalakonda K, Bizios R, Gopferich A, Young JF, Mikos AG. Retinal pigment epithelial cells on substrates with chemically micropatterned surfaces. Biomaterials 1999; 20:2351–2361. 67. Patel N, Bhandari R, Shakesheff K M, Cannizzaro S M, Davies M C, Langer R, Roberts C J, Tendler S J, Williams P M. Printing patterns of biospecifically adsorbed protein. J. Biomater Sci. Polym Ed 2000; 11:319–331. 68. Patel N, Padera R, Sanders G H, Cannizzaro S M, Davies M C, Langer R, Roberts C J, Tendler S J, Williams P M, Shakesheff K M. Spatially controlled cell engineering on biodegradable polymer surfaces. FASEB J 1998; 12:1447–1454. 69. Scholl M, Spro¨ssler C, Denyer M, Krause M, Nakajima K, Maelicke A, Knoll W, Offenha¨usser A. Ordered networks of rat hippocampal neurons attached to silicon oxide surfaces. J. Neurosci. Methods 2000; 104:65–75. 70. Mrksich M, Chen C S, Xia Y, Dike L E, Ingber D E, Whitesides G M. Controlling cell attachment on contoured surfaces with self assembled monolayers of alkanethiolates on gold. Proc Nat. Acad. Sci. USA 1996; 93:10775–10778. 71. Das M, Fox CF. Chemical cross-linking in biology. Ann Rev. Biophys. Bioeng 1979; 8:165–193. 72. Bhatia SK, Shriver-Lake LC, Prior KJ, Georger JH, Calvert JM, Bredehorst R, Ligler FS. Use of thiol-terminated silanes and heterobifunctional crosslinkers for immobilisation of antibodies on silica surfaces. Anal. Biochem 1989; 178:408–413. 73. Matsuda T, Sugawara T. Control of cell adhesion, migration, and orientation on photochemically microprocessed surfaces. J. Biomed. Mater. Res 1996; 32:165–173. 74. Matsuda T, Inouke K, Sugawara T. Development of micropatterning technology for cultured cells. Trans. Am. Soc. Artif. Int. Organs 1990; 36(3):M559–M563. 75. Holland J, Hersh L, Bryhan M, Onyiriuka E, Ziegler L. Culture of human vascular endothelial cells on an RGD-containing synthetic peptide attached to a starch-coated polystyrene surface—comparasion with fibronectin-coated tissue grade polystyrene. Biomaterials 1996; 17:2147–2156.
Surface Properties of Polymeric Biomaterials
195
76. Grzesiak JJ, Pierschbacher MD, Amodeo MF, Malaney TI, Glass JR. Enhancement of cell interactions with collagen/glycosaminoglycan matrices by RGD derivatization. Biomaterials 1997; 18: 1625–1632. 77. Jeschke B, Meyer J, Jonczyk A, Kessler H, Adamietz P, Meenen NM, Kantlehner M, Goepfert C, Nies B. RGD-peptides for tissue engineering of articular cartilage. Biomaterials 2002; 23: 3455–3463. 78. Quirk RA, Chan WC, Davies MC, Tendler SJB, Shakesheff KM. Poly(L-lysine)–GRGDS as a biomimetic surface modifier for poly(lactic acid). Biomaterials 2001; 22:865–872. 79. Hubbell JA. Chemical modification of polymer surfaces to improve biocompatibility. TRIP 2 1994; 1:20–25. 80. Massia SP, Hubbel JA. Vascular endothelial cell adhesion and spreading promoted by the peptide REDV of the IIICS region of plasma fibronectin is mediated by integrin ␣4 1. J. Biol. Chem 1992; 267:10133. 81. Rezania A, Thomas CH, Branger AB, Waters CM, Healy KE. The detachment strength and morphology of bone cells contacting materials modified with a peptide sequence found within bone sialoprotein. J. Biomed. Mater. Res 1997; 37:9–19. 82. Rezania A, Johnson R, Lefkow AR, Healy KE. Bioactivation of metal oxide surfaces: I. Surface characterization and cell response. Langmuir. 1999; 15:6931–6939. 83. Woods AJR, Couchman S, Johansson S, Hook M. Adhesion and cytoskeletal organization of fibroblasts in response to fibronectin fragments. EMBO J 1986; 5:665–670. 84. Dalton BA, McFarland CD, Underwood PA, Steele JG. Role of the heparin-binding domain of fibronectin in attachment and spreading of human bone–derived cells. J. Cell Sci 1995; 108: 2083–2092. 85. Rezania AR, Healy KE. Biomimetic peptide surfaces that regulate cell adhesion, spreading, cytoskeletal organization, and mineralization of matrix deposited by osteoblast-like cells. Biotechnol. Prog 1999; 15:19–32. 86. Mann BK, Tsai AT, Scott-Burden T, West JL. Modification of surfaces with cell adhesion peptides alters extracellular matrix deposition. Biomaterials 1999; 20:2381–2386. 87. Chen G, Ito Y, Imanishi Y, Magnani A, Lamponi S, Barbucci R. Photoimmobilization of sulphated hyaluronic acid for antithrombogenecity. Bioconj. Chem 1997; 8:730–734. 88. Taguchi T, Kishida A, Sakamoto N, Akashi M. Preparation of a novel functional hydrogel consisting of sulfated glucose-bearing polymer activation of basic fibroblastic growth factor. J. Biomed. Mater. Res 1998; 41:386–391. 89. Griffith IG, Lopina S. Microdistribution of substratum-bound ligands affects cell function: hepatocyte spreading on PEO-tethered galactose. Biomaterials 1998; 19:979–986. 90. Blanquaert F, Barritault D, Caruelle JP. Effects of heparin-like polymers associated with growth factors on osteoblast proliferation and phenotype expression. J.Biomed. Mater. Res 1999; 44:63–72. 91. Bos GW, Scharenborg NM, Poot AA, Engbers GH, Beugeling T, Aken WG, Feijen J. Proliferation of endothelial cells on surface-immobilised albumin–heparin conjugate loaded with basic fibroblastic growth factor. J.Biomed. Mater. Res 1999; 44:330–340. 92. Gutsche AT, Lo H, Zurlo J, Yager J, Leong KW. Engineering of a sugar derivatized porous network for hepatocyte culture. Biomaterials 1996; 17:387–393. 93. Kanda K, Matsuda T, Oka T. Two-dimensional orientational response of smooth muscle cells to cyclic stretching. ASAIO J 1992; 38(3):M382–385. 94. Mulder MM, Hitchock RW, Tresco PA. Skeletal myogenesis on elastomeric substrates: implications for tissue engineering. J. Biomater. Sci. Polym Edn 1998; 9:731–748. 95. Harrison RG. The cultivation of tissues in extraneous media as a method of morphogenetic study. Anat. Rec 1912; 6:181–193. 96. Curtis A, Wilkinson C. Topographical control of cells. Biomaterials 1997; 18:1573–1583. 97. Brunette DM, Chehroudi B. The effects of surface topography on micromachined titanium substrata on cell behavior in vitro and in vivo. J. Biomech. Eng 1999; 121:49–57. 98. Tan J, Shen H, Carter KL, Saltzman WM. Controlling human polymorphonuclear motility using microfabrication technology. J.Biomed. Mater. Res 2000; 51:694–702.
196
Tezcaner and Hasirci
99. Tan J, Saltzman WM. Topographical control of human neutrophil motility on micropatterned materials with various surface chemistry. Biomaterials 2002; 23:3215–3225. 100. Wojciak-Stothard B, Curtis A, Monaghan W, MacDonald K, Wilkinson C. Guidance and activation of murine macrophages by nanometric scale topography. Exp. Cell Res 1996; 233:426–435. 101. Rajnicek AM, Britland S, McCaig CD. Contact guidance of CNS neurites on grooved quartz: influence of groove dimensions, neuronal age and cell type. J. Cell Sci 1997; 110:2905–2913. 102. Duncan AC, Weisbuch F, Rouais F, Lazare S, Baquey C. Laser microfabricated model surfaces for controlled cell growth. Biosensors and Bioelectronics 2002; 17:413–426. 103. Ratner BD, McElroy BJ. Electron spectroscopy for chemical analysis: applications in the biomedical sciences. In: Gendreau RM, ed. Spectroscopy in the Biomedical Sciences, CRC Press. 1986: 107–140. 104. Ratner BD. Surface characterisation of biomaterials by electron spectroscopy for chemical analysis. Ann Biomed. Eng 1983; 11:313–336. 105. Briggs D, Seah MP, Eds. Practical Surface Analysis. New York: Wiley, 1990. 106. Rugar D, Hansma P. Atomic force microscopy. Physics Today 1990; 43:23–30. 107. Hanein Y, Pan YV, Ratner BD, Denton DD, Bo¨hringer KF. Micromachining of non-fouling coatings for Bio-MEMs application. Sensors and Actuator 2001; 81:49–54. 108. Dalton PD, Flynn L, Shoichet MS. Manufacture of poly(2-hydroxyethylmethacrylate-co-methylmethacrylate) hydrogel tubes for use as nerve guidance channels. Biomaterials 2002; 23(18): 3843–3885. 109. Xiao J, Shao M, Yin S. Environmental SEM investigation on surface defects in 0.92 Pb(Zn1/3 Nb1/3)O3 –0.08 PbTiO3 single crystal. J. Crystal Growth 2002; 240:521–525.
10 Principles of Living Organ Reconstruction by Tissue Engineering Lucie Germain, Franc¸ois Berthod, Ve´ronique Moulin, Francine Goulet, and Franc¸ois A. Auger Hoˆpital du Saint-Sacrement and Laval University Que´bec City, Que´bec, Canada
I. INTRODUCTION Tissue engineering is a novel sector arising from the biomaterial field, which is developing rapidly as a result of the dramatic clinical need for organ replacement, since there is unfortunately an ever-growing lack of organs for transplantation. Various approaches are presently being developed in different laboratories and companies based on the utilization of biomaterials, extracellular matrix components, and cells to produce substitutes to allow the replacement of wounded or diseased tissues [1–4]. The organ reconstruction by tissue engineering presented in this chapter are of living tissues. This concept entails that the various cells incorporated in our constructs or tissues are not only readily dividing, but also metabolically active. Thus, mesenchymal cells (fibroblasts, smooth muscle cells) incorporated into the stromal component of these substitutes are also significantly involved in the reorganization of the extracellular matrix. Furthermore, the interactions between the mesenchymal cells and the epithelial cells improve the very nature, structure, and function of the resulting organ. Lastly, the presence of living cells, within the in vitro engineered tissues, adds the benefit of tissue remodeling and healing after transplantation in vivo. The source of cells that can be used for tissue reconstruction is dictated by the foreseen application. Autologous cells will be necessary for the production of living tissue substitutes when striving for permanent replacement of organs in order to prevent any histocompatibility mismatch and the ensuing predictable rejection (e.g., skin grafting for full-thickness burns). However, the rejection process has been shown to vary with the type of cells involved, and it may be possible to graft allogeneic engineered tissue under some appropriate conditions. But in such cases as keratinocytes, dentritic cells and endothelial cells that are privileged targets for rejection, autologous cells are necessary to permanently replace tissues encompassing these cells. In sharp contrast, when the living tissue substitute is destined to improve wound healing, such as in the case of ulcers, allogeneic cells are sufficient since they act as a temporary coverage, enhancing the natural healing process, and will be replaced over time by cells from the receiver. The first step in reconstructing a living organ by tissue engineering in vitro is the isolation and culture of each cell type. The most stringent conditions must be met during this step since it has a direct impact on the quality of the desired tissue engineered product. 197
198
Germain et al.
The ideal cell source for tissue reconstruction should provide cells with extensive proliferation potential (self-renewal capacity) and appropriate differentiation abilities (able to give rise to a differentiated progeny). Each cell culture method must be characterized in such a manner to ensure that the isolation method and culture conditions (e.g., culture medium and growth factors) during the growth as well as during the maturation period are the most appropriate to conserve cell purity and phenotype. This chapter will focus on the various approaches developed over the years by the Laboratoire d’Organoge´ne`se Expe´rimental (LOEX) (Hoˆpital du Saint-Sacrement, Chauq, Quebec) to obtain three-dimensional tissues such as reconstructed epidermis, skin, blood vessel, cornea, bronchi, and ligament [2,5–13]. II. RECONSTRUCTED EPIDERMIS The epidermis is the first living tissue to have been reconstructed in vitro and used for clinical applications. Thus, tissue engineering of epidermis has been introduced as a life-saving procedure for severely burned patients [14,15]. Since extensive third-degree burns must absolutely be grafted in order to heal, the necessity and life-saving potential of cultured epidermis becomes obvious as such patients are affected over a large surface area. For these patients, the only previous classic therapy was the cropping of the patient’s own skin from spared sites (called donor sites) and their transplantion to cover their burn wound. Thus, a more severely burned patient has much fewer donor sites for long-term remedy [16–19]. Eventually, the surface of potential donor sites can then be insufficient to ensure patient treatment. Consequently, burn survival is, among other parameters, directly related to the total wound surface [17,20,21]. Although the growth and expansion of epidermal cells has been possible for almost 40 years, it was the improvement in cell isolation, culture conditions, and control of fibroblast overgrowth that led to the clinical application of cultured autologous epidermal sheets for the treatment of burned patients [22–27]. It is now possible to grow colonies from single cells and to produce a few square meters of cultured epidermis from a small biopsy (1 to 6 cm2) harvested from adult patients [3,28–31]. Moreover, much larger surface areas (hundreds of square meters) can be generated from 1 cm2 of newborn skin [29,32]. The differences in the total surface area produced is directly linked to the keratinocyte growth potential that has been proven to diminish after birth. This potential is many orders of magnitude higher in newborns than in adults [25,33–35]. This decreasing capacity corresponds to a reduction in the percentage of stem cells after birth and to a lower healing rate in adults as compared to children [36]. The epidermal cells form an organized and stratified epithelium when grown under the culture conditions described below. Upon grafting, the autologous cultured epidermal sheets differentiate into a functional epidermis; the barrier and protection against infection are thus restored. These results lead the way into the very active field of tissue engineering [37,38]. The adaptation of the methodology first described by Green et al. [39] for epidermal cell isolation, culture, and epidermal sheet production will be described in detail below. We have shown that the addition of a thermolysin incubation step and epidermis separation before the trypsin digestion of the skin will reduce the contamination of the epidermal cultures by human fibroblasts, increase the colony-forming efficiency level, and consequently reduce the time necessary to obtain confluent cultures [26]. A. Methodology for Epidermal Cell Isolation, Culture, and Epidermal Sheet Production The keratinocyte culture is initiated from a small biopsy of the patient’s skin. For burn patients, it is necessary to harvest the skin biopsy from a spared site, preferably as early as possible after
Living Organ Reconstruction
199
their admission to the burn unit. A 1- to 6-cm2 full-thickness skin biopsy is excised; placed aseptically into a sterile transport medium (Dulbecco-Vogt modification of Eagle’s medium with Ham’s F-12 (InVitrogen, Mississauga, Canada) in a 3:1 proportion (DME-F12); and supplemented with 24.3 g/mL adenine, 10% fetal calf serum (FCS, Hyclone, PDI BioScience, Aurora, Canada), 100 IU/mL penicillin G (PEN, Sigma), 25 g/mL gentamicin (GEN, Schering Canada Inc.), and 0.5 g/mL amphotericin B (Squibb Canada, Montreal). This biopsy should be sent to the culture laboratory and processed immediately or kept at 4⬚C. The cell isolation procedure is preferably initiated within 8–10 h following harvesting. The following procedures are performed under a sterile air flow hood. The cell isolation procedure was designed to cleave the dermo-epidermal junction with thermolysin (Fig. 1) and remove the dermis prior to epidermal cell dissociation with trypsin. The skin is trimmed in order to remove the fat. Skin fragments (1 cm ⳯ 3 mm) are then floated on 500 g/mL of thermolysin (Sigma Chemicals, St. Louis, MO) in 4-(2-hydroxyethyl)-1-piperazineethane–sulfonic acid (HEPES) buffer (6.7 mM KCl; 142 mM NaCl; 10 mM HEPES; 1 mM CaCl2; 0.45 mM NaOH; pH 7.4) overnight at 4⬚C. The epidermis is separated from the dermis with forceps and further incubated with 20 mL of 0.05% trypsin/0.1% EDTA in PBS containing PEN/GEN within a trypsinization unit (Single Sidearms Celstir, Wheaton 356533, Millville, NJ). The trypsinization unit is placed onto a magnetic stirrer for agitation for 30 min (120 rev/min at 37⬚C). Twenty milliliters of medium containing serum is added to inhibit trypsin, and an aliquot is taken for cell enumeration and cells. After centrifugation (300⳯ g, for 10 min), cells are ready for analysis or culture. After centrifugation, epidermal cells are resuspended in culture medium, counted, and plated at 2 to 5 ⳯ 106 cells per 75 cm2 culture flasks in 20 mL of culture medium; the culture flasks were preseeded with 2 ⳯ 106 irradiated fibroblasts as feeder cells (see below). Then the flasks are transferred and kept in an incubator with an atmosphere containing 8% CO2, at 37⬚C and 100% humidity. The resulting cell population is free of human fibroblasts [26]. The feeder layer is necessary to increase the keratinocyte ability to form colony and to inhibit human fibroblast overgrowth [40]. A mouse NIH Swiss 3T3 fibroblast cell line is maintained in culture in DME containing 10% CS and antibiotics and routinely passaged by trypsinization. When needed, 3T3 cells are irradiated directly in culture flasks or in suspension (after trypsinization) with 6000 rads (60Co source) to stop further growth and plated in 75 cm2 flasks 1 to 24 h before keratinocyte seeding. Alternatively, irradiated human fibroblasts can be used as feeder cells [41,42].
Figure 1 (A) Macroscopic view of the human skin separation with forceps after thermolysin digestion. (B) Histological staining of human skin showing the clear separation at the dermo-epidermal junction after thermolysin digestion. Scale bar (B): 50 m. (Courtesy of Danielle Larouche, LOEX).
200
Germain et al.
The keratinocyte culture medium consists of DME-F12 (see transport medium), supplemented with 24.3 g/mL adenine, 5 g/mL crystallized insulin, 0.4 g/mL hydrocortisone (Calbiochem, La Jolla, CA), 10 ng/mL recombinant human epidermal growth factor (EGF, Austral Biologicals, San Ramon, CA), 0.2 g/mL isoproterenol (SABEX, Boucherville, Canada), 10% FetalClone II serum (HyClone), PEN, and GEN. The culture medium is changed after 3 days and every 2 days thereafter. The cell culture is regularly observed under phase contrast microscopy. These culture conditions favor keratinocyte proliferation and differentiation [25,29,33,39,43]. Keratinocyte colonies begin to be visible about 4 days after seeding, and confluent cultures are produced within 7 to 10 days. These flasks are then trypsinized for subculturing as follows: after removal of the culture medium, the flasks are rinsed with 2 mL of trypsin-EDTA. Eight milliliters of trypsin-EDTA are then added to each flask, and they are incubated at 37⬚C until keratinocytes detach (10 to 20 min depending on the level of confluence). The enzyme is then inhibited with 8 mL of culture medium added directly to the flasks. The cells are well suspended by pipetting and then centrifuged at 300⳯ g, 10 min. The keratinocyte pellet is resuspended in keratinocyte culture medium and seeded on a 3T3 feeder layer at a concentration of 2 to 15 ⳯ 105 cells per 75 cm2 flask. For the subcultures, 5% serum is used. After trypsinization, one flask yields 15 to 25 ⳯ 106 cells that can be further divided into 30 to 125 flasks. Confluent keratinocyte cultures are obtained after 7 to 10 days, depending on the seeding concentration used. The subculturing process can be repeated several times. The number of passages that can be done varies inversely with the age of the donor [36]. Epidermal sheets for transplantation are produced from keratinocyte cultures, 24 to 48 h postconfluence. Usually, 50 to 70 flasks of 75 cm2 are prepared together. Two days before grafting, the sterility of each flask is ascertained by seeding 1 mL of the culture medium into thioglycollate broths (Que´lab, Canada). These cultures are kept at 37⬚C for 48 h. On transplantation day, the thioglycollate broths are verified and any presumptive microbial contamination is then indicated by an abnormal turbidity. If this is the case, the corresponding epidermal culture flasks are then discarded. The keratinocyte sheets are detached from the flasks with an enzyme (dispase or thermolysin) that severs the attachment between cell and plastic but not any cell–cell links (desmosomes) [27,39]. This enzymatic specificity allows the whole culture to be obtained en bloc as a tissue sheet. Each flask is rinsed twice with 10 mL of serum-free DME medium containing PEN, GEN. Ten milliliters of the same medium containing 2.5 mg/mL Dispase is then added to the flasks which are incubated at 37⬚C for 35 to 45 min, until the sheets detach from the flask’s border. The Dispase is then removed and the flasks are rinsed twice with 10 mL of medium. To prevent desiccation, 2 mL of medium are left upon the sheet, and the top of the flask is removed. A vaseline-impregnated gauze (nonadhering dressing), previously cut into a slightly smaller rectangle than the sheet’s dimension (35 cm2), is deposited to facilitate the ensuing manipulations. The gauze is smoothly pressed on the epidermal sheet with a rubber policeman. The sheet borders are then folded back onto the gauze using the same policeman. The epithelial sheet, now attached to the gauze, is turned over in the dish. Ten ligaclips (Ligaclip Extra, Ethicon, Inc., Somerville, NJ) are added to secure the sheet onto the gauze. The gauze sheets are then transferred with forceps into sterile petri dishes (100 mm in diameter) and 15 mL of serum-free medium are added to cover each sheet, which is facing up. The petri dishes are placed into a mobile incubator filled with an air atmosphere containing 8% CO2 and transported to the surgery room. The cultured epithelium comprises three to five cell layers (Fig. 2). For transplantation, the gauze sheet is taken with forceps from the petri dishes and deposited onto the wounds, with the epithelium facing downward. The basal cell layer which was attached to the culture flask is now in direct contact with the wound bed, while the differentiated layers are more superficial and the gauze protects them.
Living Organ Reconstruction
201
Figure 2 Histological staining of cultured human epidermal sheets after detachment with dispase. Scale bar: 25 m. (Courtesy of Annie Beauparlant and Danielle Larouche, LOEX.)
The engraftment level of these grafts is dependent on many factors. The rates of success vary widely between various sites in the same patient, different patients, and different burn centers [14,15,28,30,31,44–55]. The exact reason for these discrepancies are beyond the scope of this chapter, but they are similar to any classical skin graft. These parameters include the quality of the wound bed, adequacy of the surgical preparation, level of microbial wound colonization or infection, nutritional status of the patient, and various systemic imbalances [54,55]. Furthermore, displacement of grafts over the wound certainly has some deleterious effect on the basal cells. Desiccation will ultimately destroy all grafts. In our own institution, the take level has an average of 60% (with a range of 10–100%). The best results were obtained on a clean wound with some preserved dermis in the clinical setting of a patient who is recovering nicely; the take level is then approximately 80–90%. These grafts, after a successful take, steadily evolve to present a full-thickness epidermis with a normal stratum corneum. These wounds are then efficaciously and permanently covered with a self-regenerating living tissue [14,15,56–58]. The self-repair properties of these grafts have also been demonstrated; they will heal again after wounding [52]. Alternatively, the epidermal sheets we produced have been grafted onto graft donor sites that had been harvested [59]. It was shown that donor site grafted with cultured sheet can be recropped much earlier and for an increased number of times without significant sequella. The presence of a dermal component has some advantages related to the more rapid and physiological tissue repair which ensues.
III. RECONSTRUCTED SKIN BY THE SELF-ASSEMBLY APPROACH A. Principles of the Self-Assembly Approach The concept of the self-assembly approach is to reconstruct an organ in a fashion resembling its formation in vivo in which the use of appropriate culture and mechanical conditions induce cells to secrete significant amounts of extracellular matrix as during organogenesis. Cells play
202
Germain et al.
a central role since there is no addition of exogenous extracellular matrix or synthetic material. This method takes full advantage of the various intrinsic properties of cells when adequately cultured. The development of this process entailed the determination of the particular culture conditions and medium supplementation as well as the appropriate mechanical conditioning according to our previous experience. For example, epidermal cells reform a thin but wellorganized tissuelike structure by assembling together with desmosomes under adequate culture conditions (Fig. 2). Mesenchymal cells influence epithelial cells via secreted factors while also secreting and organizing extracellular matrix components [60–62]. Moreover, an adequate mechanical stimulus allows the alignment of cells and their surrounding extracellular matrix [63]. The first step in the formation of mesenchymal tissues is to coax mesenchymal cells cultured on a plastic substrate to secrete abundantly their own extracellular matrix and organize it into a relatively thick sheet. These living sheets are then assembled in three-dimensional construct by rolling them on a mandrel or stacking them. The epithelial portion, or endothelial lining, can then be added at the appropriate structural position. Success obtained with this technique to reconstruct skin and blood vessels is presented in Sections III.B and VI. The tissues generated by the self-assembly approach present very good histological characteristics and impressive mechanical strength. Moreover, the absence of exogenous extracellular matrix provides the advantage of accelerated remodeling after grafting. This simplified healing phenomenon advantageously compares to the phase of biomaterials resorption that at times leads to a slow inflammatory reaction.
B. Methodology for Skin Reconstruction by the Self-Assembly Approach To produce the living mesenchymal tissue sheet, fibroblasts were cultured from the dermal portion obtained by thermolysin digestion of skin biopsies (see Section II) and used in passages 4 to 10. Dermal fibroblasts were seeded and cultured in DME, supplemented with 50 g/mL of sodium ascorbate (Sigma), 10% fetal calf serum (Gibco BRL, Burlington, Canada), and antibiotics (penicillin and gentamicin). Twenty-eight to 35 days later, the living sheets, comprising fibroblasts and the matrix they synthesized, were peeled off from the flasks. Three layers were superimposed, and the surface area of the construct was maintained by a stainless anchoring ring for one additionnal week of culture. To induce the formation of the epidermal layer, human keratinocytes were then seeded at a density of 2 ⳯ 105 cells/cm2 in keratinocyte medium (see Section II). They reached confluence after 8 days of submerged culture and were then raised at an air–liquid interface. Keratinocyte medium with 5% serum, 50 g/mL sodium ascorbate, and without EGF was changed three times a week. Reconstructed skin was analysed after 21 days of culture at the air–liquid interface [5,64].
C. Characteristics of the Reconstructed Skin Macroscopically, the reconstructed skin presents a tissuelike appearance and texture (supple and resistant). The culture at the air–liquid interface induces the whitening of the surface of the epidermis (Fig. 3A). This is coincident with the detection of the stratum corneum on histological sections (Fig. 3B). The epidermis is well organized with cuboidal cells in the basal layer, more elongated cells in the suprabasal layers, and the presence of granules in the upper layers below the stratum corneum. Keratin 10 (Fig. 3C), filaggrin (Fig. 3D), and involucrin are expressed in the upper layers, indicating the high level of differentiation obtained in vitro under these culture conditions. The dermal compartment comprises a dense network of collagen (Fig. 3E), leading to a higher mechanical resistance of this tissue. The dermo-epidermal junction is particularly
Living Organ Reconstruction
203
Figure 3 (A) Macroscopic view of reconstructed human skins. (B) Histological staining of reconstructed human skin. (C) Keratin 10. (D) Filaggrin immunofluorescence staining of reconstructed skin. Ultrastructural analysis (E) shows the presence of keratin intermediate filaments (if), hemidesmosomes (arrows), and the lamina lucida (ll) and lamina densa (ld) of the dermo-epidermal junction of the reconstructed skin. (E,F) Collagen fibers (arrowheads) are observed in the dermal portion. Scale bars (B–D): 50 m; scale bars (E,F): 200 nm. (Courtesy of Roxane Pouliot and Danielle Larouche, LOEX.)
well organized with numerous hemidesmosomes, lamina lucida, and a complete lamina densa, as can be visualized by electron microscopy (Fig. 3E). The grafting of this human skin reconstructed by the auto-assembly approach on athymic mice confirmed the integrity of the dermo-epidermal junction since these grafts can be pinched with forceps without breaking apart [65]. The particular elasticity and suppleness of this substitute is conserved after grafting (Fig. 4A) and is thought to be associated with the significant presence and organization of collagen and elastic fibers (Figs. 4B and C). These results are promising for clinical applications. These skin substitutes could be produced from the cells of a patient to provide autologous grafts when necessary such as in the case of burn patients. Alternatively, it could be produced from allogeneic cells and thus be used as a specialized dressing to accelerate wound healing. Other skin substitutes have been applied with success on chronic skin ulcers [47,52,66–74]. The coverage is then temporary since allogeneic keratinocyte cells are always rejected, as we and others have shown [75–77]. Such a model also represents an excellent tool for various in vitro studies in physiology and in pharmacotoxicology, as we have previously shown [5,9].
IV. IN VITRO MODEL OF WOUND HEALING The first aim of tissue engineering is to replace traumatized or deficient organ and tissues in vivo. However, reconstruction of living tissues can also be used in vitro for various purposes such as the screening of new drugs before utilizing them in vivo or the elucidation of several fundamental physiological mechanisms. Until now, studies of mammalian skin wound healing were limited by the thickness and opacity of the dermis combined with the difficulty of adequately processing wounds with scabs
204
Germain et al.
Figure 4 (A) Macroscopic view of the reconstructed human skin after grafting on athymic mice. (B,C) Electron microscopy showing the presence of a well organized basement membrane between the epidermis and the dermis—hemidesmosomes (arrows), lamina lucida (ll), lamina densa (ld)—and a network of collagen (arrowheads) and elastin (e). Scale bars: 200 nm. (Courtesy of Roxane Pouliot and Danielle Larouche, LOEX.)
in order to carry out immunohistological and ultrastructural analyses. We have developed a new in vitro tissue engineered wound-healing model [78] in order to provide a tool which will then allow the clarification of the wound closure phenomenon. Indeed, despite the fact that wound closure of epithelial tissues does occur efficiently, rapidly restoring their barrier function, the exact mechanisms of wound reepithelialization remains unclear. The present model of reconstructed human skin (RHS) has been obtained by using the self-assembly approach described previously. The resulting RHS was then wounded with a 6mm-punch biopsy and placed over a third fibroblast sheet to allow the migration of keratinocytes. From the surrounding epidermis, the reepithelialization progressed toward the center. The increasing surface area of the ring formed by epidermal cells can be observed macroscopically (Fig. 5). As in vivo, features of human wound healing, such as fibroblast and keratinocyte migration into the wound (Fig. 6) and reorganization of a dermo-epidermal junction over a regenerating epidermis (Fig. 7), can be observed. This model presents numerous advantages: a three-dimensional and completely biological cellular environment; the major morphological and ultrastructural features of human skin; controlled and reproducible in vitro conditions; no physical (e.g., biomaterial) deterring proper tissue processing; and obvious landmarks for locating the wound margins. The histological observation of the spatial distribution of keratinocytes as distinguished by their particular features suggests two complementary mechanisms of reepithelialization, these
Living Organ Reconstruction
205
Figure 5 Tissue engineering of the wound healing model. Two human fibroblast sheets (a) were piled up and seeded with human keratinocytes. After differentiation (b), a 6-mm wound (W) was performed (c), and the wounded reconstructed human skin (RHS) was placed over a third fibroblast sheet to allow the migration of keratinocytes (d). (from Ref. 9.)
mechanisms being distinct and dependant of the location in the wound. The first one occurs near the wound margin where we observed a passive displacement of the superficial layers of the epidermis, which should quickly regenerate a barrier function over the wound (Figs. 6A and 7B). Our observations showing that the cornified layer had advanced further in relationship with the wound center than the granular layer indicates that the passive sliding of these superficial layers results from a pushing rather than from a dragging force. We raised the hypothesis that this pushing force originates from the mitotic pressure of proliferative keratinocytes. Because of the presence of scabs in wounds [79,80] and wound contraction [81], this phenomenon would be difficult to demonstrate in vivo, but nonetheless could significatively contribute to close small wounds (around 30% of the 6-mm wound studied after 3 days of healing). The second mechanism takes place at the tip of the neoepidermis. In agreement with the leap-frog model of migration [79,82–84], we showed that suprabasal keratinocytes migrate individually over each other to reach the connective tissue and the basal layer (Figs. 6D and E). This mechanism ensures an effective reepithelialization in the wound center based on the proliferation and migration of keratinocytes present locally once the wound margin has become distant. These two complementary mechanisms act to quickly regenerate a functional epidermal barrier over the shortest possible time period. In addition to its relevance for fundamental studies of the wound healing process, this model should be very useful to screen drugs with any potential activity in wound healing. By following wound closure macroscopically, it offers a tool to compare under standardized conditions the effect of various factors on the rate of reepithelialization. We have demonstrated that this model could correlate in vivo phenomenon by observing an acceleration of wound closure in the presence of exogenous factors such as fibrin (the main constituent of blood clot) or
206
Germain et al.
Figure 6 Histology of the wound healing model. (a) Broad view of the wound healing model after 3 days of healing. The RHS, composed of a differentiated epidermis lying over two fibroblast sheets (F1 and F2), was wounded and placed over a third sheet (F3) to allow cell migration over a natural matrix. The cut produced in the fibroblast sheets served as a landmark for locating the wound margin. Keratinocytes of the RHS formed a differentiated epidermis (b) composed of stratum basale (SB), stratum spinosum (SS), stratum granulosum (SG), and stratum corneum (SC). Suprabasal keratinocytes at the tip of the migrating epithelial tongue (MET) (*) migrated over the foremost basal one to make contact with the fibroblast sheet (d) or the fibrin matrix (e). Fibroblasts (arrows) had invaded the fibrin matrix in the 3–day wounds (c and e). Hematoxylin, phloxine and saffron staining (a–d); Masson’s trichrome staining (e) Scale bars: 50 m. (From Ref. 9.)
platelet-rich plasma (containing a mixture of cytokines and growth factors trapped in fibrin clot) (Fig. 8). Both treatments significantly accelerated the reepithelialization rate (0.75 mm/day) compared with the control without any treatment (0.4 mm/day) [85]. The completely biological human wound-healing model we developed has thus allowed us to bring a new perspective to long-standing questions concerning the mechanisms of wound closure of pluristratified epithelial tissue. It constitutes another significant example of the various applications of human tissues reconstructed in vitro by tissue engineering.
V. IN VITRO RECONSTRUCTION OF CAPILLARIES IN ENGINEERED TISSUES One of the main limitations in the transplantation of tissue engineered organs is the vascularization process, which must colonize the entire thickness of the graft in a timely fashion to prevent tissue necrosis. It is of evident clinical interest to note that full-thickness cadaver skin transplants in burn patients do recover capillary blood flow within a few days, leading to a rather unique ‘‘revitalization and take’’ of these grafts. The true nature of this intriguing phenomenon was explicated rather recently considering that cadaver skin has been used as temporary cover for over 40 years in these severely traumatized patients. This phenomenon is due to inosculation between the preexisting vessels of the cadaver skin and the wound vasculature [86]. The lack
Living Organ Reconstruction
207
Figure 7 Indirect immunofluorescence of the dermo-epidermal junction protein laminin 5 (a) and a keratinocyte differentiation marker filaggrin (b and insert). Arrowheads: abrupt end of the granular layers (SG);*: newly labeled granular cells. Scale bars: 50 m. (From Ref. 9.)
of a vascular plexus in tissue engineered skin substitutes necessitates vascularization to occur de novo rather than through inosculation. Thin tissues (thickness less than 5 mm) can be nourished and oxygenated by diffusion of molecules from the wound bed. Indeed, we demonstrated that a 5-mm-thick tissue engineered skin made from a collagen sponge transplanted on nude mice was not fully vascularized until 15 days, when the vascular network of the wound had completely ingressed into the tissue engineered substitute. Despite this long delay, the graft survived [87]. Such graft survival was permitted by imbibition of nutrients from the wound. To overcome this limitation of tissue thickness, nutrients can be topically applied on the graft [88]. But this method is restricted to very few easily accessible organs such as skin, could increase risk of microbial contamination, and is very demanding in a clinical setting. The best alternative to enhance vascularization in tissue engineered organs after transplantation is to reconstruct a vascular plexus in the tissue prior to graft, thus ensuring a rapid connection between the receiver and the grafted tissue vessels. Transplantation surgery with its all important step of vascular anastomosis of complex grafts is a testimony to this principle.
1. Development of an Endothelialized Reconstructed Skin for Clinical Application We developed the first model of tissue engineered organ in which a capillary-like network was reconstructed in vitro to facilitate and accelerate revascularization through inosculation with the patient’s own blood vessels [89].
208
Germain et al.
Figure 8 Exogenous molecules increase reepithelialization rate. *: p⬍0.05 vs. control using Wilcoxon rank sum statistical test. (From Ref. 9.)
The aim was to produce in vitro an endothelialized reconstructed skin (ERS) designed to be transplanted on deep and extensive wounds. The most evident application being targeted toward burn patients. The initial hypothesis at the inception of our experiments was that the addition of human endothelial cells in this reconstructed skin would possibly lead to a spontaneous capillary-like tube formation. The process of this tissue engineered reconstruction is as follows. Keratinocytes and dermal fibroblasts were isolated from human skin biopsies following breast reductive surgeries as previously described [90,91]. Human umbilical vein endothelial cells (HUVEC) were obtained from healthy newborns by enzymatic digestion of veins with collagenase, as previously described [7]. The ERS was prepared by seeding a suspension of 1:1 ratio of fibroblasts and HUVEC on a collagen–chitosan sponge. This cell-populated sponge was then cultured for 10 days in a medium containing 50 g/mL ascorbic acid and DME supplemented with 10% FCS and M199 supplemented with 20% NCS in a 1:1 ratio [89]. Human keratinocytes were then plated on the sponge and cultured in complete medium with 10% NCS and 50 g/mL ascorbic acid under submerged conditions for 7 days. It is important to note that after addition of keratinocytes, 10 days following the initiation of the culture, all specific growth factors for endothelial cells (contained in the M199 medium) were then removed. Afterward, the ERS was lifted to the air–liquid interface to enhance the epidermal differentiation and was cultured 14 additional days in DME-Ham’s F12, supplemented with 10% NCS, 0.4 g/mL hydrocortisone, 5 g/mL insulin, 50 g/mL ascorbic acid, and antibiotics. The ERS was then processed for histology, immunohistochemistry, and electron and confocal microscopy after 31 days of in vitro maturation.
Living Organ Reconstruction
209
We observed, on histological cross-sections of ERS (Fig. 9A), that the coculture of HUVEC with fibroblasts in the collagen sponge promoted the formation of capillary-like structures. These structures were characterized by a striking organization of endothelial cells into well-defined tubular constructs in which a lumen was observed. Capillary-like tubes were absent when each cell type was cultured separately in the sponge [89]. The capillary-like tubes were observed from days 15 to 31 in culture. The endothelial nature of the cells was demonstrated by immunohistochemistry with a positive labeling with the specific marker: the von Willebrand factor (Fig. 9B). Laminin and type IV collagen, two basement membrane components, were detected around these tubules but not inside the lumen, demonstrating that cells were adequately polarized [89]. In addition, the ultrastructural analysis of ERS showed that these capillary-like tubes exhibited tight junction forming endothelial cells, which contained the specific Weibel-Palade bodies. The tubes featured a lumen in which cell remnants were detected, but no trace of extracellular component, indicating a closed lumen (Fig. 9C). Thus, these capillary-like tubes mimicked most features of a native endothelium, except that they did not contain blood. We have shown that the spontaneous capillary-like formation that occurred in the ERS was certainly due to the presence of fibroblasts. Fibroblasts could promote the endothelial cell organization in a tubular structure by paracrine secretion of growth factors and/or by cell–matrix interactions. Indeed, fibroblasts have been proven to display angiogenic properties and increase both the number and life span of microvessels in vitro [92]. In addition, keratinocytes secrete the most powerful angiogenic factor, the vascular endothelial growth factor [93]. Furthermore, fibroblasts also produce high amounts of a well-organized human ECM when cultured in a three-dimensional porous structure [94,95]. Extracellular matrix (ECM) is known to play a major role in angiogenesis. The close contact of endothelial cells with ECM is critical to promote survival and protect cells from apoptosis [96–99], while ECM components stabilize growing microvessels [92,100]. The main feature of our model of reconstructed skin is the deposition in high amounts of a newly formed ECM by human fibroblasts cultured in the sponge. This ECM is highly differentiated since specific molecules, such as types XII and XIV collagen that are lost in conventional culture are reexpressed in this model [91]. We believe that the promotion by fibroblasts of endothelial cell organization in capillary-like tubes is due to a combination of growth factors secreted by fibroblasts and to the deposition of ECM, within which some growth factors are probably stocked. For a clinical application, the capillary-like network formed in the ERS must be of autologous origin to prevent rejection. Such a prerequisite entails that the three cell types needed in order to prepare an autologous ERS must be extracted from a single biopsy from the receiving patient. Several methods were proposed to isolate endothelial cells from human skin or from fat biopsies, either using magnetic beads or a panning technique in combination with either lectin, Ulex europaeus agglutinin-1, or a monoclonal antibody specific to endothelial cells, such as CD31 [101–104]. The extraction, culture, and expansion of autologous microvascular endothelial cells is a demanding but quite feasible method, as recently demonstrated by Supp et al. [105]. We are presently conducting experiments to demonstrate that ERS is vascularized faster after transplantation on mice compared with a conventional reconstructed skin substitute. 2. Pharmacotoxicological Application of the ERS The ERS can be also a highly valuable tool in predicting the anti-angiogenic or pro-angiogenic potential of drugs in vitro. The anti-angiogenic molecule approach to treat solid tumor has stirred a tremendous level of interest. The need for efficient models to assess the anti-angiogenic potential of such drugs thus arose. Several models have been developed to screen angiogenesis modulating agents [106]. The most common in vitro models consist of culture of endothelial
210
Germain et al.
Figure 9 Capillary-like formation in the endothelialized reconstructed dermis. (A) Spontaneous capillarylike tube formation (arrows) was observed by histology after 30 days of in vitro maturation. Magnification 400⳯. (B) The capillary-like tubes were made of endothelial cells as demonstrated by their specific staining with antibodies against CD31 (PECAM-1). Magnification 400⳯. (C) When observed by transmission electron microscopy, endothelial cells formed a closed lumen in which cell rubbles were observed and were surrounded by an abundant extracellular matrix. Scale bar: 1 m.
Living Organ Reconstruction
211
cells in gels made of various extracellular matrix compounds such as type I collagen [107–109], fibrin [110,111], fibronectin [100], basement membrane extracts, or Matrigel姟 [112,113]. Most of these models are produced with animal endothelial cells [106,112,114,115] which are known to have a much lower growth factor requirement to form tubes compared with human cells. When human endothelial cells are used, the addition of growth factors and tumor promoting agents are required to elicit capillary-like tube formation [108,110,116]. In contrast to the other in vitro angiogenesis models produced with human endothelial cells [106], our model of ERS generates an angiogenesis process with human endothelial cells without such an addition of specific growth factors or tumor promoting agents like PMA (phorbol 12-myristate 13-acetate). Since this model mimics more closely the physiology of a human connective tissue, such as the dermis, it should allow a finer and less confounding analysis of the pro- and anti-angiogenic potential of various drugs. In addition, these studies could then be performed on a long-term basis, up to 31 days, instead of the 3- to 4-day limit of other current human in vitro systems [110,112].
VI. RECONSTRUCTED BLOOD VESSELS BY THE SELF-ASSEMBLY APPROACH The self-assembly approach was initially devised and applied to reconstruct small-diameter, tissue engineered blood vessels since the long-term patency of synthetic or classically engineered prostheses has not been deemed to be satisfactory under the particular hemodynamic conditions present in arteries smaller than 5 mm in diameter [117–119]. Accordingly, there is a need for new alternatives to the autologous arteries and saphenous veins that are presently transplanted. This situation is due to their drawbacks such as limited availability, which is also compounded by the number of bypasses a patient could necessitate during a lifetime. The advantage of a functional endothelium lining the internal lumen of different types of vascular prostheses is now readily recognized and proven [120–122]. Therefore, our aim was to produce by tissue engineering a reconstructed living blood vessel as close as possible to its native counterpart. Our method was based on the exclusive use of human cells and their culture in the absence of any exogenous collagen or synthetic material. All three layers of the blood vessel—the media, the adventitia, and the intima—were sequentially added to an acellular inner membrane to form a living tubular structure, leading to a true tissue engineered blood vessel.
A. Methodology for Blood Vessel Reconstruction by the Self-Assembly Approach The endothelial cells and the smooth muscle cells were isolated from human umbilical cords by the method of Jaffe [123] and Ross [124], respectively. The fibroblasts were isolated from dermis and cultured as described above. For each cell type, the phenotype and purity of the cultures were assessed [125]. Living tissue sheets were obtained from smooth muscle cells and fibroblasts by culturing them in medium supplemented with serum and sodium ascorbate (see Section IIIB). The sheets were then detached from the plastic flasks and sequentially wrapped around a tubular mandrel. A smooth muscle cell sheet was rolled over an acellular inner membrane (dehydrated tubular tissue formed with a fibroblast sheet) to form the media. One week later, a fibroblast sheet was added in order to mature, during 7 weeks in culture, into a living reconstructed adventitia. Then, the endothelial cells were seeded in the lumen that was created
212
Germain et al.
after the removal of the mandrel. These cells then attached to the inner membrane to form a confluent endothelium (Fig. 10A). B. Characteristics of the Reconstructed Tissue Engineered Blood Vessel Macroscopically, the resconstructed tissue engineered blood vessel (TEBV) had a tubular form with an open lumen (Fig. 10B). Histologic and immunohistologic observations revealed a confluent endothelium at the inner surface (Fig. 10A). The endothelial cells expressed von Willebrand factor and incorporated acetylated low-density lipoproteins indicating that they were functional. In the media and adventitia, the cells were surrounded by a dense extracellular matrix comprising collagen and elastin. A reexpression of desmin was observed specifically in the media, not in the intima or adventitia. This intermediate filament, specific for smooth muscle cells, is lost when human vascular smooth muscle cells are cultured as monolayers on plastic substrates. They then harbor a proliferative phenotype. The reexpression of desmin clearly demontrates that smooth muscle cells become quiescent in this three-dimensionnal environment, as they are in normal blood vessels. This is a distinct advantage of the present construct. As we strived to better characterize the functionality of the reconstructed blood vessel, its mechanical resistance was assessed in vitro by testing its burst strength. This parameter has always been considered to be crucial for future in vivo applications of TEBV since the prosthesis must readily sustain the blood pressure after transplantation. The mechanical resistance of the adventitia regularly increased during the maturation in culture. This phenomenon was associated with an initial burst, followed by a lower and constant gelatinase secretion along the 2 weeks of maturation that led to a dense extracellular matrix network. The 2500 mmHg resistance
Figure 10 (A) Microscopic and (B) macroscopic views of the mature tissue-engineered blood vessel. Reconstructed human blood vessels stained with Masson’s trichrome present a confluent endothelium attached to the inner membrane (IM), reconstructed media (M), and adventitia (A). (C) Angiogram of the lower limbs 7 days after implantation showing two patent reconstructed human blood vessels providing normal blood flow in both legs. (From Ref. 7.)
Living Organ Reconstruction
213
displayed by our reconstructed blood vessel is about 20 times higher than the mean systolic blood pressure and is beyond the known burst level of the saphenous vein. This latter vessel is the conventional choice for bypass. Because of its contractile properties, the media layer of the blood vessel is responsible for the control of blood flow in response to physiological stimuli. The vasoactive response of our reconstructed media was evaluated in vitro in isolated organ bath. It contracts under stimuli such as histamine and relaxes when cAMP is added, indicating contractile/relaxation responses comparable to those of the normal vessel. Therefore, this reconstructed media provides a new human model with applications in pharmacological research. The implantation of human reconstructed blood vessels in dogs demonstrated that they could be handled and sutured by conventional surgical techniques. A patency level of 50% was obtained. These results are very significant considering that the endothelium was not added to avoid the hyperacute rejection and also because of the intrinsic xenogeneic situation of the experiment. The grafts did not tear or dilate during this week of transplantation (Fig. 10C). In conclusion, this TEBV reconstructed blood vessel is a promising substitute for clinical applications in vascular surgery of small-diameter arteries (e.g., coronary heart vessels). Other in vitro applications comprise its use as a model to understand the development of pathologies such as atherosclerosis or to shed new light on specific pharmacological pathways.
VII. RECONSTRUCTED CORNEA The tissue engineered reconstructed cornea is being developped as an in vitro model to study physiological processes such as wound repair by our research group. However, our long-term goal is the clinical application of these corneal substitutes for their transplantation in individuals afflicted with severe corneal wounds or scars. Therefore, we focused our experimental protocols on the use of normal human cells (nontransformed) for the three-dimensional reconstruction. A. Methodology for the Production of Reconstructed Cornea Human epithelial cells were isolated from the limbus or central cornea by a dispase digestion and cultured with a 3T3 feeder cell layer as described for epidermal cells (see Section II and Ref. 126). Corneal keratocytes were obtained from explants of the corneal stroma. When cultured in DME medium with 10% fetal calf serum, they resemble fibroblasts and can be serially subcultured for more than 9 passages. The anchored collagen gel method was used to produced the reconstructed stroma. Briefly, this approach consists of mixing fibroblasts in a human or bovine collagen solution that gel in a petri dish containing an anchorage. The anchorage allows the surface area to be kept constant during the culture period. This aspect is not trivial since without anchorage, the final surface would always be useless for any future purpose because of the contraction. Then epithelial cells are seeded and cultured on the surface to reconstruct the corneal epithelium (for details see Ref. 126). B. Characteristics of the Reconstructed Cornea Human epithelial cells in primary cultures form colonies that contain both proliferative and differentiated cells. Cells isolated from the limbus grow much better in culture than cells from the central cornea [126–128]. This is consistent with the identification of the limbus as the stem-
214
Germain et al.
cell repository in the cornea [13,129]. Therefore, limbal cells were thereafter chosen as the starting material for the production of our reconstructed cornea. Macroscopically, the reconstructed cornea appears as a transparent tissue. Microscopically, fibroblasts are dispersed in the collagen gel (Fig. 11). The epithelial cells form a multilayered epithelium with a basal layer containing cuboidal cells. As they progress in the suprabasal layers, cells become elongated. Similar results were obtained for cornea reconstructed with bovine and human collagen (Fig. 12). Basement membrane components characteristic of the stromal–epithelial junction (laminin, type VII collagen, and fibronectin) were expressed in our reconstructed cornea. Integrins were studied since they are involved in cell–cell as well as cell–extracellular matrix interactions, these in turn being of paramount importance in tissue formation, cell migration, and wound healing [130,131]. The epithelial cells of the reconstructed cornea expressed the 1, ␣3, ␣5, and ␣6 integrin subunits (Fig. 13) but not the ␣4 integrin subunit that has been identified in the cornea [132–134]. This reconstruction of human cornea in vitro from normal human cells can be considered as a stepping-stone achievement toward the production of a transplantable cornea. Other threedimensional models have been proposed using animal or transformed human cells [135–137]. The future challenge for our team is now to further improve the model while conserving the transparency necessary for its appropriate function. The formation of a thick stroma organized to reproduce the regular order of the collagen fibers that insures the transparency of this tissue is a closely related challenge. The endothelium and the Bowman’s and the Descemet’s membranes were not added in the present model. The availability of three-dimensional corneal models will also provide tools for the understanding of the wound-healing process of the cornea. Further knowledge in this field is of major importance since scar formation impairs the function of the cornea.
Figure 11 Masson’s trichrome staining of reconstructed human stroma produced with bovine collagen. Note the presence of keratocytes and extracellular matrix. Magnification 243⳯. (From Ref. 13.)
Living Organ Reconstruction
215
Figure 12 Masson’s trichrome staining of reconstructed human corneas produced with either bovine (A) or human (B) collagen and normal human cornea in situ. (C) Histological sections of reconstructed corneas cultured under submerged conditions for 3 days. Magnification 243⳯. (From Ref. 126.)
VIII. RECONSTRUCTED LIGAMENT The anterior cruciate ligament (ACL) is one of the most common wounded structures after significant knee trauma. It is one of the four stabilizing ligaments of the knee. It prevents abnormal anterior displacement and rotation of the lower leg with respect to the thigh. The ACL is injured when the knee is twisted under load, angled to one side or hyperextended. This can occur during a fall, rapid deceleration, pivot stress, or collision. The ligament usually ruptures or tears in its mid portion, which leads immediately to loss of function and knee instability. Unfortunately, the torn ACL has a very poor healing potential and cannot be surgically repaired [138–142]. Either the torn ACL is replaced or the patient must do without it, but at some high
216
Germain et al.
Figure 13 Immunofluorescence staining of reconstructed cornea cultured 3 days under submerged conditions (A,C,E,G) and limbus in situ (B,D,F,H). Expression of 1 (A,B), ␣3 (C,D), ␣5 (E,F), and ␣6 (G,H) integrin subunits. Magnification 250⳯. (Taken from Ref. 126.)
Living Organ Reconstruction
217
cost since then knee function is quite restricted. Some criteria are taken into account when making such a surgical decision: (1) age of the patient, (2) activity goals, (3) associated injuries, and (4) degree of knee instability. Various types of matrices have been used to replace torn ACL. Knee inflammation and synovitis have often been reported following implantation of ACL synthetic substitutes (Dacron, Gore-Tex, and Leeds-Keio) [143]. Chemically crosslinked collagen matrices have also shown limitations [144,145]. The use of cadaver ACL (allografts) [146–149] requires safe sterilization procedures to avoid disease transmission. Unfortunately, some microbial contaminants such as hepatitis C virus may resist several methods of tissue sterilization [150]. The most popular reconstructive option for ACL replacement is the use of the central portion of the autologous patellar tendon [151–153]. Other autologous tendons such as the quadriceps tendon [154,155] and Achilles tendon [156] can also be used. However, these approaches involve tissue morbidity and deleterious functional consequences. Tissue engineered ACL may provide a radically new alternative for ACL replacement. It aims at producing tissue substitutes in vitro, blending cell culture and mechanical engineering concepts with classical surgical skills. The main matrix component of ligaments and tendons is type I collagen. The ultrastructural organization of the collagen fiber network provides the strength to the ligaments and the support to the various cellular, vascular, and nervous structures required for the maintenance of its functional state and regeneration in vivo. The collagen fibers (70–150 m diameter) are organized in a parallel fashion to the long axis of the ACL. Fibroblasts (60–100 m diameter) are the main cells populating the ACL. These cells are located between the collagen fibers and are capable of regenerating the surrounding matrix. Ligament tissue engineering must take into account these histological observations if one desires to produce ACL substitutes which will share several features of the native/natural ACL in culture. Goulet et al. [10,11,157] have developed tissue engineered ACL substitutes based on several hypotheses supporting the use of natural matrix to support the tissues (Fig. 14). If the drawbacks associated with synthetic prostheses are considered as also the steps involved in tissue regeneration in situ, then a living matrix appears to be the key factor for a successful integration of ACL substitutes in vivo. Cells need a proper environment to attach, grow, and express their functions in culture and in vivo. The use of biocompatible, biodegradable, and natural matrix fibers appears to be an essential criterion to produce tissue engineered ACL substitutes. Furthermore, the role played by the synovial membrane in ACL regeneration and remodeling in vivo seems to be extremely important [158,159]. In fact, vascularization and innervation of ACL substitutes are essential to carry nutrients to the implant and insure the long-term regeneration of proprioceptive structures [158,160–162]. The tissue engineered ACL substitutes produced in our laboratory are developed in agreement with these hypotheses. To insure permanent integration of any tissue engineered substitute, living cells are expected to find the proper microenvironment to recreate the complete tissue structures. We postulate that the synovial membrane must regenerate postsurgery and reach the ACL substitute implanted to insure its integration in the knee host, since it is rich in blood vessels and nervous structures [158–164]. To reach this goal, the matrix of the implant must attract and provide the support to synoviocytes, endothelial cells, and other colonizing cell types. In addition, the ACL substitute must be able to resist to the biomechanical constraints induced during knee joint movements. However, we also postulate that the biomechanical constraints and joint shear stresses are essential to provoke adequate cell and collagen fibers alignment in situ. We believe that the cells present in the ACL substitute can and must respond to these signals to synthesize matrix constituents and play their functional role in collagen remodeling in vivo, like it is observed in vitro in ligament [10,156] and cartilage constructs [165]. In the near future, gene therapy may be combined with tissue engineering skills and concepts, notably in the field of orthopedic surgery [166], to provide competent ligament substitutes containing engineered cells which could contribute to
218
Germain et al.
Figure 14 A tissue engineered ACL substitute produced in vitro by seeding autologous ACL fibroblasts in a type I collagen matrix anchored by two bone plugs.
enhance connective tissue healing and bone formation or to control inflammatory reactions in in situ postgrafting [166]. These new therapeutic strategies are highly promising and their successful development is in progress.
IX. RECONSTRUCTED BRONCHI Another bioengineered tissue, the human bronchial equivalent (hBE), has been developed at LOEX. Such three-dimensional hBE is the first ever produced that includes both epithelial and fibroblastic cells isolated from single human bronchial biopsies. The human respiratory system is the target of several diseases, notably asthma, cystic fibrosis, and cancer. In vitro studies of these disorders led to the development of new models produced with bronchial cells of different species [167–175]. In normal subjects, bronchial basement membrane is mainly composed of laminin and type IV collagen, to which epithelial cells are attached. The human bronchial epithelium is mucociliated and pseudostratified in vivo [176]. Among epithelial cell layers, goblet cells secrete mucus, which protects the epithelium and is slowly cleared by the ciliary beats on the apical pole of specialized cells [176]. The bronchial epithelium is supported by a mesenchymal tissue
Living Organ Reconstruction
219
populated with several cell types, including fibroblasts, mucus glands, and endothelial cells [176]. Several groups have reported the isolation of bronchial cells from various species [169,170,172–174]. We reported the new enzymatic isolation and serial culture of human bronchial epithelial cells (HBEC) and fibroblastic cells (HBFC) obtained from small biopsies [170]. This was the first demonstration that mesenchymal and epithelial human bronchial cell populations could be isolated from the same donors biopsies. In absence of extracellular matrix, cells grown in monolayers or in presence of other cell types in coculture cannot adopt an organization comparable to the native tissue. In contrast, a three-dimensional matrix allows cell interactions to occur through growth factor exchanges and provides the microenvironment to promote tissue organization in vitro. Indeed, deepidermized dermis, collagen membranes, and collagen-coated dishes can provide good matricial support to epithelial cell growth in vitro [167–169,172,177–189]. However, the collagen matrix must have a minimal thickness to maintain tracheal cells in culture and promote optimal mucociliary cell differentiation [186]. Unfortunately, none of these models included living human mesenchymal and epithelial cells seeded within a three-dimensional collagenous structure. Tissue architecture seems to influence the morphology, orientation, and organization of various cell types as well as their functional properties [10,11,190,191]. The bioengineering technology developed for skin culture in our laboratory was adapted to produce the first bilayered disc-shaped model (Fig. 15), constructed with epithelial and mesenchymal bronchial cells isolated from the same donor biopsies [12]. Under the culture conditions established, our bronchial equivalent allowed epithelial cells to differentiate into ciliary and gobletlike cells (Fig. 16). The pseudostratified organization of the epithelial layers was also observed upon culture time [12]. Our approach was also used by other research groups later on [192]. This bronchial equivalent could be very useful as an alternative for the animal use for pharmacological and toxicological studies in vitro. Tissue engineered bronchi can be produced with cells of different origin, isolated from very small biopsies. They provide excellent tools to study the cellular interactions which occur in normal, asthmatic, or atypic human bronchial mucosa.
X. CONCLUSION The field of tissue engineering is progressing presently at an impressive rate. Many related fields, such as stem cell research, are adding to this impetus. Wound coverage by reconstructed epidermis and reconstructed skin happens to be the most advanced clinical project in this field.
Figure 15 Experimental procedure developed in our laboratory to produce tissue engineered bronchial mucosa in vitro.
220
Germain et al.
Figure 16 Histological section of our human tissue engineered bronchial mucosa in vitro.
Furthermore, the scope and complexity of the present clinical studies are enlarging into other domains of regenerative medicine. Tissue engineering definitely has an everincreasing role in the therapeutic armamentarium belong to the the clinical arena. We have also stressed the significant role that tissue engineered substitutes have gained in many diverse in vitro applications. Obviously the coming years shall be both demanding and very exciting.
ACKNOWLEDGMENTS The authors are grateful to their collaborators on the development of reconstructed living substitutes by tissue engineering—Drs. Raymond Labbe´, Denis Rancourt, Richard Cloutier, and Sylvain Gue´rin, and all members of LOEX—for their kind help, advice, and technical assistance in relation to the work presented in this chapter. The authors also acknowledge the collaboration of the plastic surgeons at the burn unit of the Centre Hospitalier Affilie´ Universitaire de Que´bec.
REFERENCES
1. Langer R, Vacanti JP. Tissue engineering. Science 1993; 260:920–926. 2. Berthod F, Germain L, Guignard R, Lethias C, Garrone R, Damour O, van der Rest M, Auger FA. Differential expression of collagens XII and XIV in human skin and in reconstructed skin. J Invest Dermatol 1997; 108:737–742. 3. Damour O, Braye F, Foyatier J, Fabreguette A, Rousselle P, Vissac S, Petit P. Cultured autologous epidermis for massive burn wounds: 15 years of practice. In: Rouabhia M, ed. Skin Substitute Production by Tissue Engineering: Clinical and Fundamental Applications. Austin: Landes, 1997: 23–45. 4. Auger FA, Rouabhia M, Goulet F, Berthod F, Moulin V, Germain L. Tissue-engineered human skin substitutes developed from collagen- populated hydrated gels: clinical and fundamental applications. Med Biol Eng Comput 1998; 36:801–812. 5. Michel M, L’Heureux N, Pouliot R, Xu W, Auger FA, Germain L. Characterization of a new tissueengineered human skin equivalent with hair. In Vitro Cell Dev Biol Anim 1999; 35:318–326. 6. Black AF, Berthod F, L’Heureux N, Germain L, Auger FA. In vitro reconstruction of a human capillary-like network in a tissue-engineered skin equivalent. FASEB J 1998; 12:1331–1340. 7. L’Heureux N, Paquet S, Labbe R, Germain L, Auger FA. A completely biological tissue-engineered human blood vessel. FASEB J 1998; 12:47–56.
Living Organ Reconstruction
221
8. L’Heureux N, Stoclet JC, Auger FA, Lagaud GJ, Germain L, Andriantsitohaina R. A human tissueengineered vascular media: a new model for pharmacological studies of contractile responses. FASEB J 2001; 15:515–524. 9. Laplante A, Germain L, Auger F, Moulin v. Mechanisms of wound reepithelialization: hints from a tissue-engineered reconstructed skin to long-standing questions. FASEB J 2001; 15:2377–2389. 10. Goulet F, Germain L, Caron C, Rancourt D, Normand A, Auger F. Tissue-engineered ligament. In: Yahia L, ed. Ligaments and Ligamentoplasties. Berlin: Springer-Verlag, 1997:367–377. 11. Goulet F, Germain L, Rancourt D, Caron C, Normand A, Auger FA. Tendons and ligaments. In: Lanza R, Langer R, Chick W, eds. Textbook of Tissue Engineering. San Diego: Academic Press, 1997:633–644. 12. Paquette J-S, Goulet F, Boulet L-P, Laviolette M, Tremblay N, Chakir J, Germain L, Auger FA. Three-dimensional production of bronchi in vitro. Can. Resp. J 1998; 5:43. 13. Germain L, Carrier P, Auger FA, Salesse C, Guerin SL. Can we produce a human corneal equivalent by tissue engineering?. Prog Retin Eye Res 2000; 19:497–527. 14. O’Connor N, Mulliken J, Banks-Schlegel S, Kehinde O, Green H. Grafting of burns with cultured epithelium prepared from autologous epidermal cells. Lancet 1981; 1:75–78. 15. Gallico GG, O’Connor NE, Compton CC, Kehinde O, Green H. Permanent coverage of large burn wounds with autologous cultured human epithelium. N Engl J Med 1984; 311:448–451. 16. Tanner JC, Vandeput J, Olley JF. The mesh skin graft. Plast. Reconstr. Surg 1964; 34:466–478. 17. Demling RH, Lalonde C. Burn Trauma. In: Blaidsdell C, Trunkey C, eds. Trauma Management. Vol. 4. New York: Thieme Medical Publishers, 1989:42–65. 18. Monafo WW. Tangential excision. Clin. Plast. Surg 1974; 1:591–601. 19. Shuck JM. Preparing and closing the burn wound. Clin. Plast. Surg 1974; 1:577–590. 20. Zoch G, Meissl G, Bayer S, Kyral E. Reduction of the mortality rate in aged burn patients. Burns 1992; 18:153–156. 21. Scott-Conner CEH, Meydrech E. Quantification of rate of wound closure and the prediction of death following major burns. Burns 1988; 14:373–378. 22. Cruickshank C, Cooper J, Hooper C. The cultivation of cells from adult epidermis. J. Invest. Dermatol 1960; 92:122–125. 23. Karasek MA, Charlton ME. Growth of postembryonic skin epithelial cells on collagen gels. J Invest Dermatol 1971; 56:205–210. 24. Karasek MA. Dermal factors affecting epidermal cells in vitro. J Invest Dermatol 1972; 59:99–101. 25. Rheinwald JG, Green H. Serial cultivation of strains of human epidermal keratinocytes: the formation of keratinizing colonies from single cells. Cell 1975; 6:331–343. 26. Germain L, Rouabhia M, Guignard R, Carrier L, Bouvard V, Auger FA. Improvement of human keratinocyte isolation and culture using thermolysin. Burns 1993; 19:99–104. 27. Germain L, Guignard R, Rouabhia M, Auger FA. Early basement membrane formation following the grafting of cultured epidermal sheets detached with thermolysin or Dispase. Burns 1995; 21: 175–180. 28. Auger FA. The role of cultured autologous human epithelium in large burn wound treatment. Transplantation/Implantation Today 1988; 5:21–26. 29. Green H, Barrandon Y. Cultured epidermal cells and their use in the generation of epidermis. News Physiol. Sci 1988; 3:53–56. 30. De Luca M, Bondanza S, Cancedda R, Tamisani AM, Di Noto C, Muller L, Dioguardi D, Brienza E, Calvario A, Zermani R. Permanent coverage of full skin thickness burns with autologous cultured epidermis and reepithelialization of partial skin thickness lesions induced by allogeneic cultured epidermis: a multicentre study in the treatment of children. Burns 1992; 18:516–519. 31. Donati L, Magliacani G, Bormioli M, Signorini M, Preis FW. Clinical experiences with keratinocyte grafts. Burns 1992; 18:519–526. 32. Fradette J, Michel M, Germain L. Culture d’e´piderme humain in vitro: des ke´ratinocytes normaux posse´dant un potentiel de prolife´ration exceptionnel. In: 37ie`me re´union annuelle du Club de Recherches Cliniques du Que´bec. Vol. 11: Me´decine Sciences, 1995:19. 33. Rheinwald J, Green H. Epidermal growth factor and the multiplication of cultured human epidermal keratinocytes. Natures 1977; 265:421–424.
222
Germain et al.
34. Gilchrest BA. In vitro assessment of keratinocyte aging. J Invest Dermatol 1983; 81:184S–189S. 35. Germain L, Michel M, Fradette J, Xu W, Godbout M-J, Li H. Skin stem cell identification and culture: a potential tool for rapid epidermal sheet production and grafting. In: Rouabhia M, ed. Skin Substitute Production by Tissue Engineering: Clinical and Fundamental Applications. Austin. TY: Landes Bioscience, 1997:177–210. 36. Michel M, Torok N, Godbout MJ, Lussier M, Gaudreau P, Royal A, Germain L. Keratin 19 as a biochemical marker of skin stem cells in vivo and in vitro: keratin 19 expressing cells are differentially localized in function of anatomic sites, and their number varies with donor age and culture stage. J. Cell Sci 1996; 109:1017–1028. 37. Langer R, Vacanti JP. Tissue engineering. Science 1989; 3:53–56. 38. Auger FA, Re´my-Zolghadri M, Grenier G, Germain L. The self-assembly approach for organ reconstruction by tissue engineering. e-biomed 2000; 1:75–85. 39. Green H, Kehinde O, Thomas J. Growth of cultured human epidermal cells into multiple epithelia suitable for grafting. Proc Natl Acad Sci USA 1979; 76:5665–5668. 40. Badylak SF, Lantz GC, Coffey A, Geddes LA. Small intestinal submucosa as a large diameter vascular graft in the dog. J. Surg. Res 1989; 47:74–80. 41. Braye F, Pascal P, Black A, Venet E, Damour O. Use of cultured epidermis for the treatment of extensive burns in children. The Edouard Herriot Burn Center Experience, Lyon, France. In: Horch R, Munster A, Achauer B, eds. Cultured Human Keratinocytes and Tissue Engineered Skin Substitutes. Stuttgart: Georg Thieme Verlag, 2001:175–182. 42. Limat A, Hunziker T, Boillat C, Bayenther K, Noser F. Post mitotic human dermal fibroblasts efficiently support the growth of human follicular keratinocytes. Proc. Natl. Acad. Sci. USA 1985; 82:5390–5394. 43. Simon M, Green H. Enzymatic cross-linking of involucrin and other proteins by keratinocyte particulates in vitro. Cell 1985; 40:677–683. 44. Weekley R, Klein R. Clinical nursing experience with cultured epidermal autografts. J Burn Care Rehabil 1992; 13:138–141. 45. Herzog SR, Meyer A, Woodley D, Peterson HD. Wound coverage with cultured autologous keratinocytes: use after burn wound excision, including biopsy followup. J. Trauma 1988; 28:195–198. 46. Kumagai N, Nishina H, Tanabe H, Hosaka T, Ishida H, Ogino Y. Clinical application of autologous cultured epithelia for the treatment of burn wounds and burn scars. Plast. Reconstr. Surg 1988; 82: 99–108. 47. Teepe RGC, Koebrugge EF, Ponec M, Vermeer BJ. Fresh versus cryopreserved cultured allografts for the treatment of chronic skin ulcers. Br. J. Dermatol 1990; 122:81–89. 48. Hunyadi J, Farkas B, Bertenyi C, Olah J, Dobozy A. Keratinocyte grafting: a new means of transplantation for full-thickness wounds. J Dermatol Surg Oncol 1988; 14:75–78. 49. Clugston PA, Snelling CFT, Macdonald IB, Maledy HL, Boyle JC, Germann E, Courtemanche AD. Cultured epithelial autografts:three years of clinical experience with eighteen patients. J. Burn Care Rehabil 1991; 12:533–539. 50. Pittelkow M, Shipley G. Serum-free culture of normal human melanocytes: growth kinetics and growth factor requirements. J. Cell. Physiol 1989; 140:565–576. 51. De Luca M, Albanese E, Bondanza S, Megna M, Ugozzoli L, Molina F, Cancedda R, Santi PL, Bormioli M, Stella M, Magliacani G. Multicentre experience in the treatment of burns with autologous and allogenic cultured epithelium, fresh or preserved in a frozen state. Burns 1989; 15:303–309. 52. Carver N, Leigh IM. Keratinocyte grafts and skin equivalents. Int. J. Dermatol 1991; 30:540–551. 53. Desai MH, Mlakar JM, McCauley RL, Abdullah KM, Rutan RL, Waymack JP, Robson MC, Herndon DN. Lack of long-term durability of cultured keratinocyte burn-wound coverage: a case report. J. Burn Care Rehabil 1991; 12:540–545. 54. Blight A, Mountford EM, Cheshire IM, Clancy JMP, Levick PL. Treatment of full skin thickness burn injury using cultured epithelial grafts. Burns 1991; 17:495–498. 55. Woodley DT. Covering wounds with cultured keratinocytes. JAMA 1989; 20:2140–2141. 56. Compton CC, Gill JM, Bradford D, Regauer S, Gallico GG, O’Connor NE. Skin regenerated from cultured epithelial autografts on full-thickness burn wounds from 6 days to 5 years after grafting. Lab. Invest 1989; 60:600–612.
Living Organ Reconstruction
223
57. Mommaas AM, Teepe RGC, Leigh IM, Mulder AA, Koebruge EJ, Vermeer BJ. Ontogenesis of the basement membrane zone after grafting cultured human epithelium: a morphologic and immunoelectron microscopic study. J. Invest. Dermatol l992; 99:71–77. 58. Aihara M. Ultrastructural study of grafted autologous cultured human epithelium. Br. J. Plast. Surg 1989; 42:35–42. 59. Guay N, Daigle J, Beauregard G, Grossenbacher R, Papillon J. An alternative use of culture grafts in extensive burn injuries. In 10th Congress International Society for Burn Injuries, 1998. 60. Goulet F, Poitras A, Rouabhia M, Cusson D, Germain L, Auger FA. Stimulation of human keratinocyte proliferation through growth factor exchanges with dermal fibroblasts in vitro. Burns 1996; 22:107–112. 61. Coulomb B, Lebreton C, Dubertret L. Influence of human dermal fibroblasts on epidermalization. J. Invest. Dermatol 1989; 92:122–125. 62. Mackenzie I, Fusenig N. Regeneration of organized epithelial structure. J. Invest. Dermatol 1983; 81:189S–194S. 63. Lopez Valle CA, Auger FA, Rompre P, Bouvard V, Germain L. Peripheral anchorage of dermal equivalents. Br. J. Dermatol 1992; 127:365–371. 64. Michel M, Auger FA, Germain L. Anchored skin equivalent cultured in vitro: a new tool for percutaneous absorption studies. In Vitro Cell Dev Biol Anim 1993; 29A:834–837. 65. Pouliot R, Larouche D, Auger FA, Juhasz J, Xu W, Li H, Germain L. Reconstructed human skin produced in vitro and grafted on athymic mice. Transplantation 2002; 73:1751–1757. 66. Phillips T, Kehinde O, Green H, Gilchrest B. Treatment of ulcers with cultured epidermal allografts. J. Am. Acad. Dermatol 1989; 21:191–199. 67. Beele H, Naeyaert JM, Goeteyn M, De Mil M, Kint A. Repeated cultured epidermal allografts in the treatment of chronic leg ulcers of various origins. Dermatologica 1991; 183:31–35. 68. Leigh IM, Purkis PE, Navsaria HA, Phillips TJ. Treatment of chronic venous ulcers with sheets of cultured allogenic keratinocytes. Br. J. Dermatol 1987; 117:591–597. 69. Bogensberger G, Eaglstein WH, Kirsner RS. Chronic leg ulcers in a patient with combined arterial and venous disease successfully treated with a human skin equivalent: a case report. Wounds 2000; 12:118–121. 70. Falanga V. Apligraf treatment of venous ulcers and other chronic wounds. J. Dermatol 1998; 25: 812–817. 71. Naughton G, Mansbridge J, Gentzkow G. A metabolically active human dermal replacement for the treatment of diabetic foot ulcers. Artif. Organs 1997; 21:1203–1210. 72. Sabolinski M, Alvarez O, Auletta M, Mulder G, Parenteau N. Cultured skin as a ’smart material’ for healing wounds: experience in venous ulcers. Biomaterials 1996; 17:311–320. 73. Sibbald R. Apligraf姟 living skin equivalent for healing venous and chronic wounds. Cutaneous Med Surg 1998; 3:S1-24–S1-28. 74. Veves A, Falanga V, Armstrong DG, Sabolinski ML. Graftskin, a human skin equivalent, is effective in the management of noninfected neuropathic diabetic foot ulcers: a prospective randomized multicenter clinical trial. Diabetes Care 2001; 24:290–295. 75. Rouabhia M, Germain L, Belanger F, Auger FA. Cultured epithelium allografts: Langerhans cell and Thy-1Ⳮ dendritic epidermal cell depletion effects on allograft rejection. Transplantation 1993; 56:259–264. 76. Aubo¨ck J, Irschick E, Romani N, Kompatscher P, Ho¨pfl R, Herold M, Schuler G, Bauer M, Huber C, Fritsch P. Rejection, after a slightly prolonged survival time, of Langerhans cell-free allogeneic cultured epidermis used for wound coverage in humans. Transplantation 1988; 45:730–737. 77. Phillips TJ. Cultured epidermal allografts—a temporary or permanent solution?. Transplantation 1991; 51:937–941. 78. Laplante A, Germain L, Auger F, Moulin V. Mechanisms of wound reepithelialization: hints from a tissue-engineered reconstructed skin to long-standing questions. FASEB J 2001; 15:2377–2389. 79. Winter GD. Epidermal regeneration studied in domestic pig. In: Maibach HI, Rovee DT, eds. Epidermal Wound Healing. Chicago: Year Book Medical Publishing, 1972:71–113.
224
Germain et al.
80. Rovee DT, Kurowsky CA, Labun J, Downes A. Effect of local wound environment on epidermal healing. In: Maibach HI, Rovee DT, eds. Epidermal Wound Healing. Chicago: Year Book Medical Publishing, 1972:159–181. 81. Cohen IK, Diegelmann RF, Crossland MC. Wound care and wound healing. In: Schwartz SI, Shires GT, Spencer FC, Cowles Husser W, eds. Principles of Surgery. New-York: McGraw-Hil, 1994: 279–303. 82. Krawczyk W. A pattern of epidermal cell migration during wound healing. J. Cell Biol 1971; 49: 247–263. 83. Gibbins JR. Epithelial migration in organ culture. Morphological and time lapse cinematographic analysis of migrating stratified squamous epithelium. Pathology 1978; 10:207–218. 84. Ortonne JP, Loning T, Schmitt D. Immunomorphological and ultrastructural aspects of keratinocyte migration in epidermal wound healing. Virchows Arch 1981; 392:217–230. 85. Drukala J, Bandura L, Cieslik K, Korohoda W. Locomotion of human skin keratinocytes on polystyrene, fibrin, and collagen substrata and its modification by cell-to-cell contacts. Cell Transplant 2001; 10:765–771. 86. Young DM, Greulich KM, Weier HG. Species-specific in situ hybridization with fluorochromelabeled DNA probes to study vascularization of human skin grafts on athymic mice. J. Burn Care Rehabil 1996; 17:305–310. 87. Berthod F, Germain L, Li H, Xu W, Damour O, Auger FA. Collagen fibril network and elastic system remodeling in a reconstructed skin transplanted on nude mice. Matrix Biol 2001; 20:463–73. 88. Boyce ST, Supp AP, Harriger DM, Greenhalgh DG, Warden GD. Topical nutrients promote engraftment and inhibit wound contraction of cultured skin substitutes in athymic mice. J. Invest. Dermatol 1995; 104:345–349. 89. Black A, Berthod F, L’Heureux N, Germain L, Auger FA. In vitro reconstruction of a human capillary-like network in a tissue-engineered skin equivalent. FASEB J 1998; 12:1331–1340. 90. Germain L, Rouabhia M, Guignard R, Carrier L, Bouvard V, Auger FA. Improvement of human keratinocyte isolation and culturing using thermolysin. Burns 1993; 19:99–104. 91. Berthod F, Germain L, Guignard R, Lethias C, Garrone R, Damour O, van der Rest M, Auger FA. Differential expression of collagen XII and XIV in human skin and in reconstructed skin. J. Invest. Dermatol 1997; 108:737–742. 92. Villaschi S, Nicosia RF. Paracrine interactions between fibroblasts and endothelial cells in a serumfree coculture model. Modulation of angiogenesis and collagen gel contraction. Lab. Invest 1994; 71:291–299. 93. Ballaun C, Weninger W, Uthman A, Weich H, Tschachler E. Human keratinocytes express the three major splice forms of vascular endothelial growth factor. J. Invest. Dermatol 1995; 104:7–10. 94. Berthod F, Sahuc F, Hayek D, Damour O, Collombel C. Deposition of collagen fibril bundles by long-term culture of fibroblasts in a collagen sponge. J. Biomed. Mater. Res 1996; 32:87–94. 95. Berthod F, Hayek D, Damour O, Collombel C. Collagen synthesis by fibroblasts cultured within a collagen sponge. Biomaterials 1993; 14:749–754. 96. Chavakis E, Dimmeler S. Regulation of endothelial cell survival and apoptosis during angiogenesis. Arterioscler. Thromb. Vasc. Biol 2002; 22:887–893. 97. Re F, Zanetti A, Sironi M, Polentarutti N, Lanfrancone L, Dejana E, Colotta F. Inhibition of anchorage-dependent cell spreading triggers apoptosis in cultured human endothelial cells. J. Cell Biol 1994; 127:537–546. 98. Meredith JE, Fazeli B, Schwartz MA. The extracellular matrix as a cell survival factor. Mol. Biol. Cell 1993; 4:953–961. 99. Ingber DE. Fibronectin controls capillary endothelial cell growth by modulating cell shape. Proc. Natl. Acad. Sci. USA 1990; 87:3579–3583. 100. Ingber DE, Folkman J. How does extracellular matrix control capillary morphogenesis?. Cell 1989; 58:803–805. 101. Kader KN, Moore LR, Saul JM, Zborowski M, Ziats NP, Bellamkonda RV. Isolation and purification of canine adipose microvascular endothelial cells. Microvasc. Res 2001; 61:220–226. 102. Marelli-Berg FM, Peek E, Lidington EA, Stauss HJ, Lechler RI. Isolation of endothelial cells from murine tissue. J. Immunol. Methods 2000; 244:205–215.
Living Organ Reconstruction
225
103. Conrad-Lapostolle V, Bordenave L, Baquey C. Optimization of use of UEA-1 magnetic beads for endothelial cell isolation. Cell Biol. Toxicol 1996; 12:189–197. 104. Marx M, Perlmutter RA, Madri JA. Modulation of platelet-derived growth factor receptor expression in microvascular endothelial cells during in vitro angiogenesis. J. Clin. Invest 1994; 93:131–139. 105. Supp DM, Wilson-Landy K, Boyce ST. Human dermal microvascular endothelial cells form vascular analogs in cultured skin substitutes after grafting to athymic mice. FASEB J 2002; 16:797–804. 106. Cockerill GW, Gamble JR, Vadas MA. Angiogenesis: models and modulators. Int. Rev. Cytol 1995; 159:113–160. 107. Kanda S, Tomasini-Johansson B, Klint P, Dixelius J, Rubin K, Claesson-Welsh L. Signaling via fibroblast growth factor receptor-1 is dependent on extracellular matrix in capillary endothelial cell differentiation. Exp. Cell Res 1999; 248:203–213. 108. Yang S, Graham J, Kahn JW, Schwartz EA, Gerritsen ME. Functionnal roles for PECAM-1 (CD31) and VE-cadherin (CD144) in tube assembly and lumen formation in three-dimensional collagen gels. Am. J. Pathol 1999; 155:887–895. 109. Montesano R, Orci L. Phorbol ester induces angiogenesis in vitro from large vessel endothelial cell. J. Cell. Physiol 1987; 139:284–291. 110. Bayless KJ, Salazar R, Davis GE. RGD-dependent vacuolation and lumen formation observed during endothelial cell morphogenesis in three-dimensional fibrin matrices involves the alpha(v)beta(3) and alpha(5)beta(1) integrins. Am. J. Pathol 2000; 156:1673–1683. 111. Bach T, Barsigian C, Chalupowicz D, Busler D, Yaen C, Grant D, Martinez J. VE-Cadherin mediates endothelial cell capillary tube formation in fibrin and collagen gels. Exp. Cell Res 1998; 238: 324–334. 112. Troyanovsky B, Levchenko T, Ma¨nsson G, Matvijenko O, Holmgren L. Angiomotin: an angiostatin binding protein that regulates endothelial cell migration and tube formation. J. Cell Biol 2001; 152: 1247–1254. 113. Kubota Y, Kleinman HK, Martin GR, Lawley TJ. Role of laminin and basement membrane in the morphological differentiation of human endothelial cells in capillary-like structure. J. Cell Biol 1988; 107:1589–1598. 114. Chen CS, Toda KI, Maruguchi Y, Matsuyoshi N, Horiguchi Y, Imamura S. Establishment and characterization of a novel in vitro angiogenesis model using a microvascular endothelial cell line, F-2C, cultured in chemically defined medium. In Vitro Cell. Dev. Biol. Anim 1997; 33:796–802. 115. Vernon RB, Sage EH. A novel, quantitative model for study of endothelial cell migration and sprout formation within three-dimensional collagen matrices. Microvasc. Res 1999; 57:118–133. 116. Davis GE, Camarillo CW. An ␣ 2  1 integrin-dependent pinocytic mechanism involving intracellular vacuole formation and coalescence regulates capillary lumen and tube formation in three-dimensional collagen matrix. Exp. Cell Res 1996; 224:39–51. 117. Stephen M, Loewenthal J, Little JM, May J, Sheil AG. Autogenous veins and velour dacron in femoropopliteal arterial bypass. Surgery 1977; 81:314–318. 118. O’Donnell TF, Mackey W, McCullough JL, Maxwell SL, Farber SP, Deterling RA, Callow AD. Correlation of operative findings with angiographic and noninvasive hemodynamic factors associated with failure of polytetrafluoroethylene grafts. J. Vasc. Surg 1984; 1:136–148. 119. Sayers RD, Raptis S, Berce M, Miller JH. Long-term results of femorotibial bypass with vein or polytetrafluoroethylene. Br. J. Surg 1998; 85:934–938. 120. Zilla P, von Oppell U, Deutsch M. The endothelium: a key to the future. J. Card. Surg 1993; 8: 32–60. 121. Cines DB, Pollak ES, Buck CA, Loscalzo J, Zimmerman GA, McEver RP, Pober JS, Wick TM, Konkle BA, Schwartz BS, Barnathan ES, McCrae KR, Hug BA, Schmidt AM, Stern DM. Endothelial cells in physiology and in the pathophysiology of vascular disorders. Blood 1998; 91:3527–3561. 122. Zilla P, Deutsch M, Meinhart J. Endothelial cell transplantation. Semin. Vasc. Surg 1999; 12:52–63. 123. Jaffe EA, Nachman RL, Becker CG, Minick CR. Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria. J. Clin. Invest 1973; 52: 2745–2756. 124. Ross R. The smooth muscle cell. II. Growth of smooth muscle in culture and formation of elastic fibers. J. Cell Biol 1971; 50:172–186.
226
Germain et al.
125. Germain L, Remy-Zolghadri M, Auger F. Tissue engineering of the vascular system: from capillaries to larger blood vessels. Med Biol Eng Comput 2000; 38:232–240. 126. Germain L, Auger FA, Grandbois E, Guignard R, Giasson M, Boisjoly H, Guerin SL. Reconstructed human cornea produced in vitro by tissue engineering. Pathobiology 1999; 67:140–147. 127. Pellegrini G, Golisano O, Paterna P, Lambiase A, Bonini S, Rama P, Luca M. Location and clonal analysis of stem cells and their differentiated progeny in the human ocular surface. J. Cell Biol 1999; 145:769–782. 128. Lindberg K, Brown ME, Chaves HV, Kenyon KR, Rheinwald JG. In vitro propagation of human ocular surface epithelial cells for transplantation. Invest. Ophthalmol. Vis. Sci 1993; 34:2672–2679. 129. Cotsarelis G, Cheng SZ, Dong G, Sun TT, Lavker RM. Existence of slow-cycling limbal epithelial basal cells that can be preferentially stimulated to proliferate: implications on epithelial stem cells. Cell 1989; 57:201–209. 130. Hynes RO. Integrins: versatility, modulation, and signaling in cell adhesion. Cell 1992; 69:11–25. 131. Ruoslahti E. RGD and other recognition sequences for integrins. Annu Rev Cell Dev Biol 1996; 12:697–715. 132. Larjava H, Peltonen J, Akiyama SK, Yamada SS, Gralnick HR, Uitto J, Yamada KM. Novel function for beta-1 integrins in keratinocyte cell–cell interactions. J. Cell Biol 1990; 110:803–815. 133. Stepp MA, Spurr-Michaud S, Gipson IK. Integrins in the wounded and unwounded stratified squamous epithelium of the cornea. Invest. Ophthalmol. Vis. Sci 1993; 34:1829–1844. 134. Lauweryns B, van den Oord JJ, Volpes R, Foets B, Missotten L. Distribution of very late activation integrins in the human cornea. An immunohistochemical study using monoclonal antibodies. Invest. Ophthalmol. Vis. Sci 1991; 32:2079–2085. 135. Minami Y, Sugihara H, Oono S. Reconstruction of cornea in three-dimensional collagen gel matrix culture. Invest. Ophthalmol. Vis. Sci 1993; 34:2316–2324. 136. Zieske JD, Mason VS, Wasson ME, Meunier SF, Nolte CJ, Fukai N, Olsen BR, Parenteau NL. Basement membrane assembly and differentiation of cultured corneal cells: importance of culture environment and endothelial cell interaction. Exp. Cell Res 1994; 214:621–633. 137. Griffith M, Osborne R, Munger R, Xiong X, Doillon CJ, Laycock NL, Hakim M, Song Y, Watsky MA. Functional human corneal equivalents constructed from cell lines. Science 1999; 286: 2169–2172. 138. Frank CB, Jackson DW. The science of reconstruction of the anterior cruciate ligament. J. Bone Joint Surg. Am 1997; 79:1556–1576. 139. Frank CB. Ligament healing: current knowledge and clinical applications. J Am Acad Orthop Surg 1996; 4:74–83. 140. Johnson RJ, Beynnon BD, Nichols CE, Renstrom PA. The treatment of injuries of the anterior cruciate ligament. J. Bone Joint Surg. Am 1992; 74A:140–151. 141. Lyon RM, Akeson WH, Amiel D, Kitabayashi LR, Woo SL. Ultrastructural differences between the cells of the medical collateral and the anterior cruciate ligaments. Clin. Orthop 1991:279–286. 142. Fox RJ, Harner CD, Sakane M, Carlin GJ, Woo SL. Determination of the in situ forces in the human posterior cruciate ligament using robotic technology. A cadaveric study. Am. J. Sports Med 1998; 26:395–401. 143. Koski JA, Ibarra C, Rodeo SA. Tissue-engineered ligament: cells, matrix, and growth factors. Orthop. Clin. North Am 2000; 31:437–452. 144. Dunn MG, Liesch JB, Tiku ML, Zawadsky JP. Development of fibroblast-seeded ligament analogs for ACL reconstruction. J. Biomed. Mater. Res 1995; 29:1363–1371. 145. Dunn MG, Trai AJ, Kato YP, Bechler JR, Ochner RS, Zawadsky JP, Silver FH. Anterior cruciate ligament reconstruction using a composite collagenous prosthesis. Am. J. Sports Med 1992; 20: 507–514. 146. Jackson DW, Simon TM. Tissue engineering principles in orthopaedic surgery. Clin. Orthop 1999: S31–S145. 147. Sterling JC, Meyers MC, Calvo RD. Allograft failure in cruciate ligament reconstruction. Followup evaluation of eighteen patients. Am. J. Sports Med 1995; 23:173–8.
Living Organ Reconstruction
227
148. Shino K, Kawasaki T, Hirose H, Gotoh I, Inoue M, Ono K. Replacement of the anterior cruciate ligament by an allogeneic tendon graft. An experimental study in the dog. J. Bone Joint Surg. Br 1984; 66:672–681. 149. Miller CM, Tibone JE, Hewitt M, Kharrazi FD, Elattrache NS. Interference screw divergence in femoral tunnel fixation during endoscopic anterior cruciate ligament reconstruction using hamstring grafts. Arthroscopy 2002; 18:510–514. 150. Yamakado K, Kitaoka K, Nakamura T, Yamada H, Hashiba K, Nakamura R, Tomita K. Histologic analysis of the tibial bone tunnel after anterior cruciate ligament reconstruction using solvent-dried and gamma-irradiated fascia lata allograft. Arthroscopy 2001; 17:32. 151. Jackson DW, Grood ES, Goldstein JD, Rosen MA, Kurzweil PR, Cummings JF, Simon TM. A comparison of patellar tendon autograft and allograft used for anterior cruciate ligament reconstruction in the goat model. Am. J. Sports Med 1993; 21:176–185. 152. Clancy WG, Narechania RG, Rosenberg TD, Gmeiner JG, Wisnefske DD, Lange TA. Anterior and posterior cruciate ligament reconstruction in rhesus monkeys. J. Bone Joint Surg. Am 1981; 63: 1270–1284. 153. Abe S, Kurosaka M, Iguchi T, Yoshiya S, Hirohata K. Light and electron microscopic study of remodeling and maturation process in autogenous graft for anterior cruciate ligament reconstruction. Arthroscopy 1993; 9:394–405. 154. Fulkerson JP, Langeland R. An alternative cruciate reconstruction graft: the central quadriceps tendon. Arthroscopy 1995; 11:252–254. 155. Grana WA, Egle DM, Mahnken R, Goodhart CW. An analysis of autograft fixation after anterior cruciate ligament reconstruction in a rabbit model. Am. J. Sports Med 1994; 22:344–351. 156. Linn RM, Fischer DA, Smith JP, Burstein DB, Quick DC. Achilles tendon allograft reconstruction of the anterior cruciate ligament-deficient knee. Am. J. Sports Med 1993; 21:825–831. 157. Goulet F, Rancourt D, Cloutier R, Germain L, Poole A, Auger FA. Tendons and ligaments. In: Lanza R, Langer R, Vacanti J, eds. Principles of Tissue Engineering. San Diego: Academic Press, 2000:711–722. 158. Halata Z, Baumann KI. Sensory nerve endings in the hard palate and papilla incisiva of the rhesus monkey. Anat Embryol (Berl) 1999; 199:427–437. 159. Haus J, Refior HJ. A study of the synovial and ligamentous structure of the anterior cruciate ligament. Int. Orthop 1987; 11:117–124. 160. Shimizu T, Takahashi T, Wada Y, Tanaka M, Morisawa Y, Yamamoto H. Regeneration process of mechanoreceptors in the reconstructed anterior cruciate ligament. Arch. Orthop. Trauma Surg 1999; 119:405–409. 161. Zimny ML. Mechanoreceptors in articular tissues. Am J Anat 1988; 182:16–32. 162. Buma P, Elmans L, Van Den Berg WB, Schrama LH. Neurovascular plasticity in the knee joint of an arthritic mouse model. Anat. Rec 2000; 260:51–61. 163. Grigg P. Properties of sensory neurons innervating synovial joints. Cells Tissues Organs 2001; 169: 218–225. 164. Denti M, Monteleone M, Berardi A, Panni AS. Anterior cruciate ligament mechanoreceptors. Histologic studies on lesions and reconstruction. Clin. Orthop 1994; 308:29–32. 165. Quinn TM, Schmid P, Hunziker EB, Grodzinsky AJ. Proteoglycan deposition around chondrocytes in agarose culture: construction of a physical and biological interface for mechanotransduction in cartilage. Biorheology 2002; 39:27–37. 166. Musgrave DS, Fu FH, Huard J. Gene therapy and tissue engineering in orthopaedic surgery. J Am Acad Orthop Surg 2002; 10:6–15. 167. Baeza-Squiban A, Boisvieux-Ulrich E, Guilianelli C, Houcine O, Geraud G, Guennou C, Marano F. Extracellular matrix–dependent differentiation of rabbit tracheal epithelial cells in primary culture. In Vitro Cell. Dev. Biol. Anim 1994; 30A:56–67. 168. Davenport EA, Nettesheim P. Regulation of mucociliary differentiation of rat tracheal epithelial cells by type I collagen gel substratum. Am. J. Respir. Cell Mol. Biol 1996; 14:19–26. 169. de Jong D, Prins F, van Krieken HH, Mason DY, van Ommen GB, Kluin PM. Subcellular localization of bcl-2 protein. Curr. Top. Microbiol. Immunol 1992; 182:287–292.
228
Germain et al.
170. Goulet F, Boulet L-P, Chakir J, Tremblay N, Dube´ J, Laviolette M, Boutet M, Xu W, Germain L, Auger FA. Morphologic and functional properties of bronchial cells isolated from normal and asthmatic subjects. Am. J. Respir. Cell Mol. Biol 1996; 15:312–318. 171. Gray TE, Guzman K, Davis CW, Abdullah LH, Nettesheim P. Mucociliary differentiation of serially passaged normal human tracheobronchial epithelial cells. Am. J. Respir. Cell Mol. Biol 1996; 14: 104–112. 172. Kaartinen L, Nettesheim P, Adler KB, Randell SH. Rat tracheal epithelial cell differentiation in vitro. In Vitro Cell Dev Biol Anim 1993; 29A:481–492. 173. Kondo M, Finkbeiner WE, Widdicombe JH. Cultures of bovine tracheal epithelium with differentiated ultrastructure and ion transport. In Vitro Cell Dev Biol 1993; 29A:19–24. 174. Niles R, Kim KC, Hyman B, Christensen T, Wasano K, Brody J. Characterization of extended primary and secondary cultures of hamster tracheal epithelial cells. In Vitro Cell Dev Biol 1988; 24:457–463. 175. Whitcutt MJ, Adler KB, Wu R. A biphasic chamber system for maintaining polarity of differentiation of cultured respiratory tract epithelial cells. In Vitro Cell Dev Biol 1988; 24:420–428. 176. Colby T, Yousem S. Histology for pathologists. In:. Sternberg S, ed. Lungs New York: Raven Press, 1992:479–497. 177. Cambrey AD, Kwon OJ, Gray AJ, Harrison NK, Yacoub M, Barnes PJ, Laurent GJ, Chung KF. Insulin-like growth factor I is a major fibroblast mitogen produced by primary cultures of human airway epithelial cells. Clin Sci (Lond) 1995; 89:611–617. 178. Clark AB, Randell SH, Nettesheim P, Gray TE, Bagnell B, Ostrowski LE. Regulation of ciliated cell differentiation in cultures of rat tracheal epithelial cells. Am. J. Respir. Cell Mol. Biol 1995; 12:329–338. 179. de Jong D, Prins FA, Mason DY, Reed JC, van Ommen GB, Kluin PM. Subcellular localization of the bcl-2 protein in malignant and normal lymphoid cells. Cancer Res 1994; 54:256–260. 180. Emery N, Place GA, Dodd S, Lhermitte M, David G, Lamblin G, Perini JM, Page AM, Hall RL, Roussel P. Mucous and serous secretions of human bronchial epithelial cells in secondary culture. Am. J. Respir. Cell Mol. Biol 1995; 12:130–141. 181. Infeld MD, Brennan JA, Davis PB. Human tracheobronchial epithelial cells direct migration of lung fibroblasts in three-dimensional collagen gels. Am. J. Physiol 1992; 262:L535–541. 182. Jetten AM, Shirley JE, Stoner G. Regulation of proliferation and differentiation of respiratory tract epithelial cells by TGF . Exp. Cell Res 1986; 167:539–549. 183. Jetten AM. Multistep process of squamous differentiation in tracheobronchial epithelial cells in vitro: analogy with epidermal differentiation. Environ. Health Perspect 1989; 80:149–160. 184. Nakamura Y, Tate L, Ertl RF, Kawamoto M, Mio T, Adachi Y, Romberger DJ, Koizumi S, Gossman G, Robbins RA. Bronchial epithelial cells regulate fibroblast proliferation. Am. J. Physiol 1995; 269:L377–387. 185. Ostrowski LE, Nettesheim P. Inhibition of ciliated cell differentiation by fluid submersion. Exp. Lung Res 1995; 21:957–970. 186. Robinson CB, Wu R. Mucin synthesis and secretion by cultured tracheal cells: effects of collagen gel substratum thickness. In Vitro Cell Dev Biol Anim 1993; 29A:469–477. 187. Shoji S, Ertl RF, Linder J, Romberger DJ, Rennard SI. Bronchial epithelial cells produce chemotactic activity for bronchial epithelial cells. Possible role for fibronectin in airway repair. Am Rev Respir Dis 1990; 141:218–225. 188. Shoji S, Rickard KA, Ertl RF, Robbins RA, Linder J, Rennard SI. Bronchial epithelial cells produce lung fibroblast chemotactic factor: fibronectin. Am. J. Respir. Cell Mol. Biol 1989; 1:13–20. 189. Shoji S, Rickard KA, Takizawa H, Ertl RF, Linder J, Rennard SI. Lung fibroblasts produce growth stimulatory activity for bronchial epithelial cells. Am Rev Respir Dis 1990; 141:433–439. 190. Bellows CG, Melcher AH, Aubin JE. Association between tension and orientation of periodontal ligament fibroblasts and exogenous collagen fibres in collagen gels in vitro. J. Cell Sci 1982; 58:125–138. 191. Bouvard V, Germain L, Rompre P, Roy B, Auger FA. Influence of dermal equivalent maturation on the development of a cultured skin equivalent. Biochem. Cell Biol 1992; 70:34–42. 192. Chakir J, Page N, Hamid Q, Laviolette M, Boulet LP, Rouabhia M. Bronchial mucosa produced by tissue engineering: a new tool to study cellular interactions in asthma. J Allergy Clin Immunol 2001; 107:36–40.
11 Recombinant Protein Scaffolds for Tissue Engineering Jerome A. Werkmeister, Paul R. Vaughan, Yong Peng, and John A. M. Ramshaw CSIRO Molecular Science Victoria, Australia
I. INTRODUCTION It is postulated that approximately 1 in 5 people who reach the age of 65 will require some form of organ or tissue replacement [1]. The direct and indirect costs associated with these surgeries are an enormous burden to the health care providers. In addition, there are severe limitations to supply and use of conventional biomaterials and medical devices, as well as traditional cadaver organ/tissue replacements. Tissue engineering is an emerging multidisciplinary alternative solution to this shortfall that creates new functional and viable tissues/organs from autologous cells and scaffolds. The ultimate aim of tissue engineering is to regenerate exactly the natural tissue architecture and function, rather than to produce an inferior scarlike tissue that may have the desired shape, but is biologically inadequate. This can be achieved directly by generating a tissue-equivalent implant or indirectly through use of an implant containing the cells, signals, and scaffold that are subsequently remodeled to provide correct tissue and function (Fig. 1). It is assumed in this technology that the tissue structure is understood and easy to replicate. For some tissues the composition and structure of the natural tissue is not well defined. For example, for heart valve leaflets the distribution of the different collagen types and associated components has not been well described; type VI collagen, for example, shows a very distinct asymmetric distribution in these leaflets (unpublished results, J.F. White). The use of specific monoclonal antibodies to different tissue components [2–5] will be important for defining the key elements of the natural tissue and for examining if the tissue engineered product—or the final tissue formed in vivo if the implant is an intermediate—matches these requirements. A successfully engineered tissue will require selection of the appropriate cell type for the tissue being repaired. These could be adult autologous cells, for example, chondrocytes for cartilage repair, osteoblasts for bone repair, or a combination of fibroblasts, smooth muscle cells, and endothelial cells for blood vessel regeneration. Alternatively, adult (for example, tissue-derived or bone marrow mesenchymal) or embryonic stem cells can be used and selectively directed down the desired differentiation pathway to the appropriate cell type. This process of cell selection and proliferation will require development of a specific set of signals, such as growth factors, for each individual cell type. In all cases, the balance between cell proliferation and phenotype retention is an important issue if the correct tissue is to be formed. For example, 229
230
Werkmeister et al.
Figure 1 The process of tissue engineering.
if adult chondrocytes are expanded in monolayer culture, they rapidly change their macroscopic round phenotype toward a flat elongated fibroblast cell type that is accompanied by a switch in collagen synthesis from type II to type I collagen (Fig. 2). The selection of an appropriate scaffold also presents a significant challenge. The present polymers that are commonly used, such as poly(lactide-co-glycolide) (PLGA), may need improvement or replacement with custom-designed polymers that are more suitable mechanically and biologically to the damaged tissue; for example, the acidity that arises during degradation may prove detrimental in certain situations. It may be that multicomponent systems are preferable, where each component has a different turnover rate and provides selective signals for cell attachment and growth, so that the shape of the scaffold may be retained by a minor slow turnover component while new tissue forms and replaces a major faster turnover component. The tissue that forms must have the correct biological and biomechanical characteristics. The scaffolds may provide signals, distinct from soluble signals, as well as provide shape and form to the construct. This may require further modification of polymer components by signals known to enhance cell attachment, for example, by grafting peptides such as RGD to the polymer, or using more cell-specific signals that can activate the desired cell type. Additional signals, such as oxygen tension or force, either as minimal force in a microgravity bioreactor or specific pressure to maintain phenotype such as with chondrocytes, may also be important.
Figure 2 Chondrocytes isolated from sheep articular cartilage lose their rounded phenotype: (A) partly after 3 days in monolayer culture and (B) almost completely after 6 days in culture, giving elongated fibroblast cell morphology, characterized by type I collagen rather than type II collagen synthesis.
Recombinant Protein Scaffolds
231
Natural polymers, proteins, and carbohydrates may prove to be very useful as scaffolds and preferable to the synthetic polymeric materials. Carbohydrates, however, are likely to be less attractive scaffolds, generally lacking specific biological signals that can affect cell behavior, and some may not always have appropriate turnover rates. For example, although chitosan and alginates are natural materials, they are not natural to man and the absence of appropriate degradation pathways may limit their turnover characteristics. Proteins, on the other hand, should prove much more useful. For example, collagen and fibrinogen can be readily prepared from natural tissues and have well-defined properties that can be tailored to suit recruitment of cells. When proteins such as collagens are derived from the extra cellular matrix of animal tissues, they contain the natural binding domains for cells and for proteins as an integral part of their structure, as well as have their own defined functionalities. However, when these proteins are isolated from these natural tissues there is a risk of contamination by infectious agents. The availability of recombinant protein sources minimizes this risk of contamination. The use of recombinant proteins also provides several other advantages. For example, it is possible to make sufficient quantities of minor tissue components that may have specific functions, for example, certain of the minor collagens, that could not be purified from tissue in sufficient quantity for commercial applications. Also the technology provides the opportunity to make novel molecules, such as those based on repeats of a particular functional element, or chimeric molecules where different functions are combined, such as collagen and a growth factor. The recombinant products will be available in reproducible form, without the batch-to-batch variation that can occur for tissue extracts. The degradation rates can be controlled, for example, by either adding or removing specific proteinase sites, and the degradation products are readily cleared, not leading to particulates or localized pH changes. The use of human derived sequences minimizes any risk of immunogenicity, which is an issue when bovine or other heterologous proteins are used. This chapter considers a range of potential recombinant proteins that could be used as scaffolds in tissue engineering. In addition, we explore the opportunities for recombinant manufacture of novel protein structures based on natural structural motifs.
II. COLLAGEN AND GELATIN The collagen family of proteins is the most abundant protein group of the extracellular matrix, being the principal component, for example, of skin, cartilage, blood vessels, ligaments, and tendons [6]. The most abundant collagens are the interstitial, fibril-forming collagens, particularly type I collagen. These collagens form the major tissue structures through forming fiber bundle networks that are stabilized by specific crosslinks to give stability and strength to the tissues [7]. However, there is also a range of other collagen types, many forming network structures, which are present in low quantities [6]. However, in a particular tissue location the minor collagen may be a significant and critical component, for example the type X collagen in hypertrophic cartilage [8] or the type IV collagen in basement membranes [9]. The essential role of collagens in tissue, and their numerous interactions with cells and other molecules, has made collagen a widespread choice as a biomaterial for many clinical applications [10], and this is now being extended to their use as a scaffold for tissue engineering [11,12]. Each collagen type has a genetically distinct amino acid sequence and a characteristic chain composition. Collagen triple helical domains are characterized by a (Gly-X-Y)n repeating sequence, where X and Y can be any amino acid, but are frequently (⬃20%) proline (Pro) (X position) and hydroxyproline (Hyp) (Y position) [13]. The Hyp is derived by secondary modification of Pro residues. The essential Gly and the high Pro/Hyp content is a requirement for the formation of a stable triple helix, which is the characteristic structural feature of collagens, in
232
Werkmeister et al.
which three individual chains are wound in a coiled-coil ropelike structure [14]. When collagens are denatured they produce gelatin. The mixture of individual chains does not reform into native collagen, but rather zones of mismatched triple helix form, providing the junctional domains of the gelatin gel [15]. Because of the repetitive structure of collagens, and their long chain length of over 1000 amino acids for fibrillar collagens (Fig. 3) [16], the initial amino acid sequence determination work was difficult. However, Fietzek, Kuhn, and others made significant progress that allowed the full determination of several complete chains [17]. Subsequently, the use of DNA sequencing approaches allowed the confirmation of these protein data and allowed extension of the data available to all collagen chains and types, including those in low natural abundance [18–20]. These data further showed the complexity of collagen structures, including the various noncollagenous domains that are found for each collagen and the interruptions to the triple helix that are found in certain collagen types. The availability of the clones derived for sequence studies provided a library for subsequent expression studies of recombinant protein. Collagen can be readily prepared for tissue engineering scaffolds. It can be in the form and shape of natural tissue, where other molecular and cellular components have been removed by biochemical treatments, while retaining the crosslinked structure of the tissue. Alternatively, it can be made as a soluble, purified product by enzymatic removal of the crosslinks of the intact tissue and solubilization in dilute acid, and then reconstituted and stabilized in the shape and form required of the scaffold [21]. However, in both types of scaffold material, the use of animal-derived materials may evoke an immunological reaction in some patients [22], and for both animal- and human-sourced collagen the risk of disease transfer is ever present. Hence the availability and use of recombinant human collagen would be a significant advantage and would also allow the production in quantity of minor collagen types. The biosynthesis of collagens is complex and involves a large number of secondary modification steps, many of which are unique to collagen and collagen-like molecules [6,21]. A key feature of any system that is used for production of recombinant collagen is that it must be able to provide active prolyl-4-hydroxylase (P4H), either endogenous or introduced, in order to convert Pro residues in the Y position of the Gly-X-Y repeat to 4-hydroxyproline (Hyp) [23]. The P4H enzyme is a tetramer, comprising two ␣ and two  chains. This conversion step is essential as the presence of Hyp increases the thermal stability of the collagen such that it is stable at body temperature [24,25]. A range of other secondary modification steps also occur during collagen synthesis, such as hydroxylation and glycosylation of some lysine residues [26,27], but these steps are not as critical for stability and function. Initial attempts to produce recombinant collagen-like material were described using an Escherichia coli system [28]. However, it has not proved possible to effectively introduce P4H into E. coli so these products would be gelatin like and less likely to form suitable stable
Figure 3 Schematic diagram for an ␣ chain of an interstitial collagen, for example, type II or type III collagen, showing the arrangement of the triple helical (䊏) and procollagen (왓) domains. The procollagen domains are removed in the fully processed product. (From Ref. 18.)
Recombinant Protein Scaffolds
233
scaffolds. An alternative approach for introducing Hyp into the collagen product has also been described through incorporation from Hyp-enriched media [29]. However, this approach would not restrict the Hyp to the Y position of the triple-helical domain, but would also insert it in the X position, which could lead to instability. Also, Hyp would be incorporated into any noncollagenous segments that were present. Subsequently, recombinant collagens that include correct hydroxylation were described from human HT1080 cells that contain endogenous P4H [30], although the yields were low and would not be sufficient for scaffold construction. Insect cell systems using Baculovirus vectors also contain endogenous P4H that can be used to produce hydroxylated recombinant collagen [31,32], again in low yields. This expression system has proved useful, however, in optimizing the requirements of a recombinant system [32]. The use of yeast expression systems that allow high yields of protein has now become the current method of choice for production of recombinant collagens [33–39]. However, it is possible that the use of plant systems [40–42] or transgenic animals [43,44] may also grow in importance. Hydroxylated collagen has been successfully produced in Saccharomyces systems [35–37], for example, by using nonintegrating, stable vectors. The collagen gene can be introduced under control of an ADH promoter, while the two genes for each of the P4H subunits can be introduced on a separate vector using a single bidirectional GAL promoter. Collagen secretion can be initiated using yeast signal sequences [37]. Other arrangements of the three genes on these vectors are also effective [45]. Alternatively, the three genes can be introduced as multiple copies integrated into the yeast genome, all under the control of GAL promoters [36]. This approach leads to excellent yields of collagen that has a high degree of prolyl hydroxylation, as shown by a thermal stability comparable to native, tissue-derived collagen [36]. The approach of chromosomal integration has also been used successfully in other yeasts, including Pichia and Hansenula [34,46–48]. In Hansenula, secretion can occur using a native collagen signal sequence, but this approach is less effective in Pichia, where yeast signal sequences are preferred [48]. These systems produce fully hydroxylated collagens in commercial yields [32,46]. Biochemically, the recombinant collagens mimic natural collagens, showing good hydroxylation of Pro residues, and an ability to form fibrillar aggregates. The use of recombinant technology allows the production of all types and combinations of natural human collagens that are of high purity and disease-free [49]. This approach also makes it possible to modify the structures of these natural collagens to produce molecules with altered functionality (see Section VII). Products based on this approach will take longer to reach the market, but are expected to be particularly useful in areas such as tissue engineering.
III. ELASTIN Elastin is one of the more important ‘‘elastic proteins’’ found in mammalian tissues, being a significant component of tissues such as blood vessels, bladder, uterus, skin, and lung. It is also a significant component of certain elastic cartilage tissues such as ear cartilage. Elastin is the molecular component that provides extensibility and elastic recoil to these tissues. Many structural proteins can be considered as elastic proteins even though they show a diverse range of mechanical properties. In some cases, the term elastic implies the ability to deform reversibly, without loss of energy, requiring high resilience; whereas in other cases it is taken to mean stretchy, with an ability to extend to large strains with little applied force. Elastin is a rubberlike protein that exhibits the combination of high resilience, low stiffness, and large strain and functions in storage of elastic strain energy [50]. The biochemistry of elastin has been extensively studied [51].
234
Werkmeister et al.
Elastic fibers in tissue consist primarily of an amorphous elastin core associated with microfibrils, about 10–12 nm in diameter, that contain fibrillins and microfibril-associated glycoproteins [52], together forming a lattice of microfibrils [53]. These other proteins are also elastic and comprise the elastic component of tissues in more primitive organisms where elastin protein is absent [54]. Elastin is synthesized as a monomer, tropoelastin, which subsequently assembles into a stable polymeric structure that is stabilized by specific covalent crosslinks. The crosslinks, desmosine and isodesmosine are multivalent and arise from oxidative deamination of specific Lys residues by the copper-dependent enzyme lysyl oxidase [55]. These crosslinks mean that mature elastin is insoluble in all reagents except those which hydrolyze the peptide bonds. For example it is resistant to hot 0.1 M NaOH. A most unusual property of elastin is its extreme durability. For most tissues it appears not to be replaced during the lifetime of an animal [56], the exception being the elastin in the uterus [57]. Thus, for example, in blood vessels the elastin in an older person was present at birth and had remained intact despite the tens of millions of cycles of extension and recoil. Thus elastin-based scaffolds may be highly durable and long-lived when included in tissue engineered devices. Another property that is important for scaffolds is that elastin and fragments of elastin are chemotactic for fibroblasts and monocytes [58], suggesting the presence of cell recognition sites. It was shown that the elastin-derived peptides Gly-Phe-Gly-Val-Gly and Val-Gly-ValAla-Pro-Gly, are chemotactic, with the latter being frequent in the elastin sequence, suggesting multiple cell binding sites [59]. Peptide Val-Gly-Val-Ala-Pro-Gly is also a ligand for a 67-kDa elastin cell surface receptor. Subsequently it was shown that removal of these sites in elastin did not eliminate fibroblast attachment, suggesting that further binding site(s) are present [60]. Poly(Gly-Val-Gly-Val-Pro), one of the common repeating sequence blocks of human elastin, has also been synthesized and shown to have similar properties to natural elastin in vitro and in vivo [61]. This base repeating unit is nonadhesive to cells, but can be engineered to promote cell attachment by incorporation of an additional peptide block, Gly-Arg-Gly-Asp-Ser-Pro [62]. Yet another atypical property of elastin is its capacity for coacervation. Coacervation occurs when hydrophobic proteins self-aggregate and come out of solution as a second phase when the temperature is increased or the ionic strength or pH is altered. This is a property of elastin fragments obtained by hydrolysis—of polypeptide analogs [63] as well as of the intact monomeric protein [64]. Coacervation is a rapid and reversible process that takes place through multiple intermolecular interactions of its hydrophobic domains [65]. Although coacervation is an intrinsic property of pure elastin, in vivo it is likely that glycosaminoglycans and other molecules mediate the coacervation through interactions with the Lys side chains [66]. The structural characterization of elastin was severely hampered by its lack of solubility. Indeed elastin could be prepared as the residue when chemical and biochemical steps had dissolved everything else. For example, there are no methionine residues in elastin so CNBr does not cleave it. However, the isolation of the soluble precursor that lacks any crosslinks, tropoelastin, from copper-deficient pigs [67] led to a better understanding of the protein. Structural studies suggested that the molecule was built of two types of alternating domains (Fig. 4). One was hydrophobic and represented the elastic region of the molecule, while the other was hydrophilic, rich in Lys residues involved in the crosslinks. Subsequently, cDNA cloning and sequencing of human and other elastins have given a detailed picture of the protein structure [68,69]. This has confirmed that elastin does not have pro-domains, as found in collagens, that are needed for alignment of adjacent chains during elastic fiber formation. Elastin is useful as a biomaterial, but has not become as widely used and accepted as collagen. Elastin has potential particularly for devices where compliance is important, for example, in the construction of bladder prostheses [70].
Recombinant Protein Scaffolds
235
Figure 4 Schematic diagram for tropoelastin, showing the repeating hydrophobic ( ) domains, and the intervening hydrophilic ( ) crosslinking domains. Those parts of the molecule that occur in alternatively spliced forms are indicated ( ). (From Ref. 18.)
Elastin-based materials, derived from natural tissue such as aorta, have been used in repair of esophageal injury [71] and for duodenal repair [72]. Soluble elastin has been shown as potentially useful as a coating for certain synthetic polymers such as polyethylene glycol terephthalate and polyhexamthylene diamine–adipic acid, where it is able to promote endothelial cell growth while maintaining phenotype; it is, however, not suitable for coating polyurethanes [73,74]. Elastin has also been successfully combined with other materials. For example, an elastin fibrin composite has proved successful for periodontal treatment [75]. These applications clearly indicate the potential value of elastin as a scaffold material for tissue engineering. The principal concern, however, has been the availability of a reproducible source of elastin or tropoelastin in sufficient quantity. The development of recombinant elastin is, therefore, seen as an encouraging milestone to meet this requirement. The initial production of recombinant human elastin was achieved in a bacterial system using the lysogenic E. coli host AR120 [76]. It was found that the bacterial cell readily degraded the product, so production as a fusion product with influenza virus NS1 protein with an intervening Met residue was preferred. Since Met is absent in elastin, CNBr cleavage could be used to release the elastin product, which could then be highly purified by reverse-phase chromatography. However, in this system there was modest expression with a yield of only 2–4 mg/L [76]. Biochemical tests, including N-terminal sequencing and immunological evaluation, demonstrated identity with natural protein. Also, the expressed recombinant protein was shown to be chemotactic to fibroblasts [76] and to promote cell binding [77]. Recombinant elastin was also a suitable substrate for lysyl oxidase–based crosslinking, giving a product that was insoluble in hot 0.1 M NaOH [78]. The modest yield obtained by this initial recombinant approach would limit the usefulness of the product as a scaffold. However, an alternative approach has been developed which gives a considerably better yield. Rather than using the native coding sequence for the elastin gene, a fully synthetic gene of 2100bp was constructed that coded for the same protein sequence but had a different, host-preferred, codon distribution. This synthetic gene was then expressed in E. coli [79]. The key feature of this approach was to match the synthetic gene codon usage pattern to the E. coli host, since in the human sequence at least one-third of the codons are rare expression-limiting codons in E. coli. Also, elastin is a highly repetitive sequence where four amino acids account for 75% of the sequence [79]. In systems where expression is severely limited there is also an increased propensity for in-frame translational hopping as the synthetic machinery tries to cope [80]. This is turn leads to product purification difficulties due the microheterogeneity of the product. In the initial trial of this approach the elastin was successfully expressed as a C-terminal fusion product with glutathione-S transferase, giving a yield of about 30% of total cellular protein. However, release of elastin by thrombin cleavage led to significant proteolysis of the elastin product [79]. The alternative approach of preparing an unfused product gave a yield of
236
Werkmeister et al.
around 20% of cell protein but without the same preparative losses. The product was easily purified through its high solubility in solutions of short chain alcohols [81], giving better than 90% purity. Further purification could be achieved by gel permeation chromatography. The initial yield from this system, without optimization, was over 30 mg/L. The recombinant protein, derived from the synthetic gene, also showed the biochemical and biophysical characteristics of native tropoelastin. Thus, circular dichroism indicated a high -structure, with 41% -sheet and 21% -turn, both consistent with naturally extracted tropoelastin [64]. Studies on its coacervation showed that it was reversible and was dependent on temperature, ionic strength, concentration, and pH. Of particular interest, it was found that optimal coacervation occurred at 37⬚C, 150 mM NaCl, and pH 7 to 8 [64]. This form of recombinant tropoelastin was also suitable for incorporation into matrix. It was shown that nonelastogenic fibroblasts could be used as a source for lysyl oxidase, which is also necessary for collagen crosslinking and 12% of the tropoelastin was incorporated into insoluble matrix with formation of desmosine and isodesmosine crosslinks [55] Recombinant technology also allows fragments of elastin to be constructed [65,82–85]. The fragments are useful for further elucidation of the mechanisms of assembly, coacervation, and cell binding. The can also be used to form novel molecules with repeating structures or chimeric components (see Section VII).
IV. FIBRINOGEN–FIBRIN One of the earliest events in wound healing is the formation of a fibrin clot, which arrests bleeding that occurs during the initial tissue injury. The fibrin also serves as a provisional matrix during this initial phase of tissue repair, and allows infiltration of inflammatory cells and mesenchymal cells for progression of the healing response. Fibrin has been used commonly as a tissue sealant or glue during surgery, and as a hemostatic agent in the United States and Europe. The efficacy of the sealant and/or glue normally requires the action of three primary components: fibrinogen, thrombin, and factor XIII, which will result in clot (fibrin) formation in the presence of calcium. The sealant is manufactured commercially and is available as a two-part system (e.g., Baxter BioScience). The first component contains primarily human fibrinogen as well as small quantities of plasma fibronectin, factor XIII, plasminogen, and bovine aprotinin, a plasmin inhibitor that prevents fibrinolysis. The second component contains human thrombin, a serine protease that initiates the polymerization. Fibrinogen is a 340-kDa plasma protein that is cleaved by thrombin to produce an insoluble fibrin clot. Fibrinogen is a hexamer containing two pairs of nonidentical chains (A␣, B, and ␥) (Fig. 5) that are linked together in an antiparallel arrangement by disulfide bonds. The molecule is symmetrical and contains two globular D nodules that are attached to a central E nodule by a coiled coil region [86]. The central nodule consists of the N termini of all six polypeptide chains. The D nodules primarily consist of the C termini of the  and ␥ chains that are folded into a globular domain. Polymerization begins when thrombin cleaves four specific Arg-Gly bonds at the N termini of first the A␣ and then the B chain, resulting in the release of fibrinopeptide A (FpA) and fibrinopeptide B (FpB). As the FpA is released, initial polymerization occurs leading to the formation of half-staggered double-stranded protofibrils through exposure of new sites to allow interactions between the ␣ and ␥ chains [87]. In a similar fashion, it is assumed that the release of FpB also exposes sites to allow lateral propagation [88], although the precise
Recombinant Protein Scaffolds
237
Figure 5 Schematic diagram for the structure of fibrinogen, showing the three separate chains and their coiled-coil heptad repeat (ⵦ) and extension regions ( ). Those parts of the molecule that occur in alternatively spliced forms are indicated ( ). (From Ref. 18.)
mechanism is unclear. The result of this cleavage and stabilization is an intricate fibrous network that will control bleeding and act as a provisional scaffold for tissue regeneration. Since fibrinogen/fibrin is a natural provisional scaffold in wound healing, it has attracted attention as a choice material in tissue engineering applications. It has a number of important attributes that make it suitable for use as a scaffold. It can be prepared by relatively simple fractionation methods from animal or autologous human sources, as well as from recombinant systems; the scaffold can be used as a gel or molded into a defined shape to mimic the tissue being reconstructed; and the degradation can be controlled by crosslinking or by use of aprotinin and tranexamic acid [89]. Fibrin has been used as a delivery vehicle for cells, growth factors, and genes. For instance, urothelial cells, tracheal epithelial cells, and preadipocytes can be grown in culture and reimplanted into animals with fibrin [90]. Fibrin can be mixed with other components like alginates to enhance proliferation and differentiation of chondrocytes, periosteal cells, and nucleus pulposus cells and to stabilize tissue transplants [91], or it can be used to design more complex structures including small blood vessels and heart valve prostheses [92]. Fibrin can also be fabricated to include growth factors like TGF-1 to augment recruitment of mesenchymal cells [93], and can be used as a transfection system to carry genes like human EGF to accelerate wound healing in tissue constructs [94]. Cloning and sequencing of the three chains of human fibrinogen was carried out in the early 1980s. Cross-species hybridization using bovine probes was used to screen a human liver cDNA library to isolate each of the fibrinogen chains [95]. The ␣ chain contains 2224 base pairs, including noncoding regions at check the 5′- and 3′-ends, a 19 amino acid signal peptide, and a mature protein of 625 amino acids [96]. The  chain contains 461 amino acids with a 30 residue signal sequence [97], while the ␥ chain comprises 437 amino acids with a leader sequence of 26 residues [98]. There is also a variant ␥B form of the ␥ chain that results from alternative splicing, substituting the terminal Ala-Gly-Asp-Val sequence with a 20 amino acid sequence. Full details of the sequence data for each of the chains, along with thrombin cleavage sites, the FpA and FpB, phosphorylation site, cross-linking sites, and calcium binding sites can be accessed in common databases [18]. Fibrinogen, or at least the individual chains, has been expressed in various cells. For example, intact human ␣ and ␥ chains have been expressed at high levels, around 13 g/mL, in E. coli using IPTG induction and promoters and signal sequences from bacterial plasmids. The expressed proteins were largely insoluble and present in inclusion bodies, but could be resolubilized in 6 M guanidine HCl or 6 M urea with retention of functionality [99]. Expression
238
Werkmeister et al.
of the  chain in COS cells demonstrated an association with BiP and degradation in the rough endoplasmic reticulum, indicating a reliance on other chains for optimal secretion [100]. The most common expression systems have used Chinese hamster ovary (CHO) cells using a twostep cloning strategy that permits efficient synthesis of engineered variant fibrinogens [101]. Variant recombinant proteins can be grown in serum-free media for several months, purified using protamine-Sepharose chromatography to produce around 2–4 g/mL culture media, and can be compared to control recombinant and plasma fibrinogen [102]. Using this CHO expression system, Lord and colleagues have been able to systematically dissect the mechanism of thrombincatalyzed fibrinogen polymerization [101]. In addition to understanding the relevant amino acid sequences and related structural regions involved with cleavage, protofibril formation, and subsequent lateral aggregation, an understanding of the cell binding domains is clearly important for a scaffold that is to be engineered for cell propagation and tissue formation. The ␣ chain contains two Arg-Gly-Asp potential cell binding domains, although the pentapeptide Gln-Ala-Gly-Asp-Val located near the C terminus of the ␥ chain is the main adhesion site for the platelet integrin receptor ␣IIb3 [18]. More recent studies have shown that there are additional functional platelet binding domains in the ␥ chain — blood from patients with dysfibrinogen Vlissingen/Frankfurt have variant sequences at position ␥318–320 and ␥408–411, again near the C terminus, that are associated with reduced or absence of platelet adhesion [103]. Most tissue engineering applications will require promotion of vascularization into the newly forming tissue. Interaction of fibrin with endothelial cells will stimulate capillary tube formation via the endothelial cell receptor VE-cadherin and the  chain 15–42 region [104]. More recently, it was found that this thrombin-treated 15–66 fragment required dimerization via Cys65 to mimic the natural dimeric structure of the  chains in fibrin, and that His16 and Arg17 were essential for cell binding [105].
V. SILKS Silks are externally spun fibrous protein secretions that are produced by a range of organisms, including many insects, particularly Lepidoptera, of which the domesticated silkworm Bombyx mori is the best known. They are also made by spiders, including over 30,000 species from the class Arachnida, and by scorpions, mites, and flies. Overall, silks have diverse properties varying with source and the different functions they fulfill. Since silks are not of mammalian origin, they would not be expected to have the same cell and molecular interaction sites nor the same turnover mechanisms, such as matrix metalloproteinase susceptibility, as mammalian protein materials. Nevertheless, silk-based materials have been shown to be useful in various biomaterial applications. For example, biodegradable cast membranes with oxygen and water vapor permeability have been described [106]. It was shown that fibroblast attachment and growth on silk fibroin was comparable to that observed on a collagen substrate [107], and that the cell attachment properties could be modified, for example, with a phosphorylcholine to reduce platelet adhesion [108], or with integrin recognition sequences and parathyroid hormone fragments to enhance cell binding [109]. Most recently a role for silks as a scaffold in tissue engineering has been demonstrated. A silk-fiber matrix was shown to be a suitable material for tissue engineering of an anterior cruciate ligament device, which matched the complex and demanding mechanical requirements of a native human ACL, including adequate fatigue performance. This protein matrix supported the attachment, expansion, and differentiation of adult human progenitor bone marrow stromal cells [110]. Thus silks will be of particular interest where strength is required in a tissue engineering scaffold.
Recombinant Protein Scaffolds
239
Sericulture, the feeding of silkworms with mulberry leaves for silk textile production, has been practiced in China for over 4000 years. The secreted silk comprises two fibers, up to 1 km long, held together by a family of adhesive proteins, the sericins. The other well-characterized silk is the dragline silk from the major ampullate gland of the orb-weaving spider Nephila clavipes. Spiders produce up to seven different types of silk, using various glands [111], but due to culture difficulties none are commercially produced. The mechanical properties of silks are what make them remarkable fibers. Some silks exhibit up to 200% elongation before break, and silks are unmatched in terms of energy absorption prior to break by any natural or synthetic fiber. While the B. mori silk is of lower strength than spider dragline, it still shows considerable strength properties. These silks are also thermally stable, show glass transitions between 170 and 210⬚C and such stabilities around 230 to 250⬚C. Structurally, the insoluble nature of spun silks has made determination of amino acid sequence data difficult. Initial studies indicated that B. mori silk comprised two structural proteins, the fibroin heavy (325 kDa) and light (25 kDa) chains. Various consensus sequences have been described, including a crystalline 59 residue repeat, GlyAlaGlyAlaGlySerGlyAlaAlaGly[SerGlyAlaGlyAlaGly]8Tyr, which is found with some variations [112], and extensive GlyAlaGlyAlaGlySer repeats and less well conserved GlyAlaGlyAlaGlyTyr repeats [113]. Shorter nonrepetitive domains surround all the repeated elements. In the spider N. clavipes a major protein MAGS1 (spidroin 1, 275 kDa) contains amino acid repeats that are shorter and less well conserved than those from B. mori and which frequently contain polyalanine, [Ala]6–9, domains and a 15 residue repeat based on a GlyGlyXaa framework. These repeating sequences provide crystalline domains that form antiparallel -sheet structures parallel to the fiber axis and which crosslink the structure. The amorphous rubberlike phase separates these semicrystalline structures. In B. mori, the crystalline phase is about 60% of the structure, while in N. clavipes the polyalanine phase is 30 to 50%. The remarkable mechanical properties of these silks are a direct result of the size and orientation of the crystalline domains and their linkages. To date, the main biomaterial and tissue engineering applications of silks have been based on natural materials. Initial biotechnology studies enable the structure of fibroin to be defined [114–116], and these revealed a clear link between protein sequence and structure–property relationships [117,118]. There has been some interest in the expression of recombinant silk. For example, genetic engineering can be used to mix different modules in specific proportions so that proteins with defined strength and elasticity can be designed that have many potential medical and engineering uses (see Section VII) [119]. The process of spider silk production has been mimicked by expressing dragline silk genes in mammalian cells, leading to soluble recombinant dragline silk proteins with molecular masses of 60 to 140 kDa [120]. Monofilaments could be spun from a concentrated aqueous solution of soluble recombinant protein that were water insoluble and exhibited toughness and modulus values comparable to those of native dragline silks but with lower tenacity [120]. However, it is the understanding of how specific modules contribute to the physical properties of silks that enables novel structures based around repeats of consensus structures to be designed for biomaterials applications (see Section VII).
VI. OTHER NATURAL PROTEIN MATERIALS A range of other recombinant human proteins may potentially find use in scaffolds for tissue engineering. These include various globular as opposed to the fiber forming proteins that have been discussed in the preceding sections. For example, serum albumin has been cloned and expressed as a recombinant protein using various host systems, including E. coli [121], Pichia pastoris [122], and transgenic potato and tobacco plants [123].
240
Werkmeister et al.
Of more interest, however, are fibronectin and other adhesive molecules. Fibronectin is a glycoprotein that is widespread and is present at high concentration in most intracellular matrix tissue. It is also present at high concentration, 300 g/mL, in plasma and other body fuilds. Fibronectin plays an important role as an adhesive protein, where it is involved in cell–matrix interactions, particularly in relation to cell migration, regulation of cell growth and differentiation, and in hemostasis [18]. Fibronectin comprises a dimer of two nonidentical protein chains (Fig. 6) that differ due to alternative splicing and which are linked by disulfide bonds located near the C termini of the chains. Each chain is constructed of a series of functional domains linked by flexible polypeptide chain. These domains are built by replication of three types of distinct repeating structural motifs (Fig. 6). The different functional domains have specific binding characteristics for other matrix proteins and receptor complexes. Thus the central region interacts with most adherent cells via a range of different integrin receptors, while other regions of the molecule display other interactions, for example, with collagen, heparin and fibrin (Fig. 6). The gene structure for fibronectin is well established [124,125], allowing easy production of recombinant fragments. Fibronectin in body fluids exists as a soluble dimer, whereas in tissues fibronectin can exist as an oligomeric disulfide-bonded fibrous form. This fibrous format may have some application as a scaffold for tissue engineering. However, the main interest in fibronectin, and other adhesion molecules, lies in the use of the various binding modules in fibronectin in combination with other components in the design of novel scaffolds with unique properties (see Section VII). A range of nonmammalian proteins may also prove of interest in the construction of scaffolds for tissue engineering, being adhesives that work in wet conditions which are durable and potentially biodegradable. These include, for example, the various components of the adhesive proteins that have been isolated from various marine species, such as Mytilus edilus. The Byssal thread of M. edilus consists of functional-specific zones, each with their own specific protein components [126]. Also, the adhesive plaque, which can bond to various materials including synthetics such as plastic and natural tissues such as bone, is a mixture of four protein families [127]. It forms a trabecular, solid foamlike structure [128] which could potentially be adapted as a scaffold for cell infiltration. The Byssal thread consists of a gradient of molecules that are composites of silklike and collagen domains in the distal region and elastin and collagen domains in the proximal region [129]. These composite molecules, while not necessarily being directly useful in scaffolds, clearly illustrate that in nature different functional modules can be combined into novel structures that can provide unique mechanical properties. A clear advantage of recombinantly derived protein materials for tissue engineering scaffolds is that this technology
Figure 6 Schematic diagram for a single chain of fibronectin, showing the repeating domains, termed FNI ( ), FNII ( ), and FNIII ( ). Those parts of the molecule that occur in alternatively spliced forms are indicated ( ). The regions of specific binding activities are indicated. (From Ref. 18.)
Recombinant Protein Scaffolds
241
can also be readily used to make novel composite assemblies that extend what is directly available from nature.
VII. DESIGNED MOLECULES The various proteins discussed in the preceding sections represent a wide range of secondary structure motifs, typically as fibrous structures: the triple helix of collagen, the  structures of the elastin and silks, and the ␣-helical structure within fibrinogen. These represent the major secondary structures that are found in proteins. While the ‘‘rules’’ that define the tertiary folding of proteins are not understood, there is a considerable understanding of the key features that define these secondary structure motifs. Recombinant approaches create opportunities for the production of novel, non-natural molecules based on these secondary structures. For example, products could be made by selecting a specific region and making a multiple repeat structure, or they could contain a selection of specific regions making alternating and repeating structures. Also novel products can be designed that mix these structures with other distinct elements, such as the binding motifs described from fibronectin or other proteins such as growth factors. Each of these constructs would aim to modify the cell–scaffold interaction so as to improve the quality of the tissueengineered construct. In designing a novel construct certain features can be considered. These include codon selection. For many proteins, as noted for elastin [79], the codon usage pattern in the human gene is quite distinct from that of the microbial host, to the extent that inefficient synthesis and potential deletions or other errors in the product may occur. Synthetic genes with selected codon usage patterns allow this problem to be minimized. Another potential problem which can be minimized, particularly for repetitive structures, is the risk of loss of message integrity through looping out of sequence. Also, in silk constructs the yield and homogeneity of higher molecular weight silk proteins were found to be limited by truncated synthesis, probably as a result of ribosome termination errors [130]. The extent of this problem was related to the host and was an issue in E. coli but not in P. pastoris [130]. A further opportunity that arises from production of recombinant proteins as scaffolds is the opportunity to incorporate unusual amino acids into the product. It has long been known that several amino acid analogs can be substituted for the natural amino acids during protein synthesis [131]. These include, for example, selenomethionine and other derivatives for methionine, adding bulk or reactive groups [132,133], fluorinated derivatives such as p-fluorophenylalanine for phenylalanine and trifluoroleucine or hexafluoroleucine for leucine, which will change the properties of the products to more resemble fluoropolymers [134]. Inclusion of 3,4-dehydroproline for proline can be used to introduce alkene groups into a structure that may then be used for site-specific chemical modifications. Most recently the range of these opportunities has been extended by modification of tRNA molecules to give novel specificities [135]. A. Novel Scaffolds Based on the Triple Helix Collagen molecules have now been recognized to have important binding and interaction functions as well as to provide strength and stability to tissues, and a wide range of specific binding sites have now been identified [19]. These include the recognition sites for enzymatic cleavage, such as for the MMP-1 (collagenase) cleavage site [136]. The collagenase cleavage site is well defined for the major, fibril-forming collagens, and cleavage at this site is a key step in the turnover of collagen [136]. Therefore, modification or removal of this site would be expected to reduce the turnover of recombinant collagen-based biomaterials. Other binding sites include,
242
Werkmeister et al.
for example, the ␣21 integrin binding site in different collagen types [21,137,138]. Further understanding of the nature of these sites and their specificity, for example, to different cells types, will assist in the development of novel scaffolds. As well as designing modification to natural structures, new scaffolds can be designed based on the triple helix motif, but not equivalent to a natural collagen [139]. These could include structures based on multiple copies of a single function zone, such as an integrin binding site [139] or a polymer or repeating polymers containing more than one functional domain. Alternatively, constructs that are non-natural chimeras between a collagen fragment or a designed triple helical structure and another protein, such as a growth factor, can be made [139]. For example, constructs of collagen with EFG [140] and interleukin-2 [141] have been described and shown to have mitogenic activity [140]. Novel triple helical products should also have adequate thermal stability so that they are stable at body temperature. This can be achieved through sufficient Hyp incorporation as part of the expression system. However, consideration of the amino acid sequence of the Gly-XaaYaa repeat structure can also assist. For example, Arg in the Yaa position can also give stability to the triplehelix [142]. Selection of the nonproline residues in the Xaa and Yaa positions can also affect overall stability. A comprehensive study using a host–guest peptide strategy has enabled the effects of all amino acids on stability to be determined [143]. These data have shown, for example, that Gly-Gly sequences are destabilizing, as are Phe and Tyr, particularly in the Yaa position. Asp residues in the Yaa position are also destabilizing. However, in selecting the Xaa and Yaa residues, a structure that is very stable and rigid may not be optimal, and a degree of flexibility may be important to enable proper binding [21,144]. B. Novel Scaffolds Based on  Structures The possibility of designing a protein with specific characteristics from knowledge of the roles of different amino acids in the stability of a secondary structure was demonstrated for a -sheet structure [145,146]. The roles of different amino acid substitutions in these structures were able to provide further information on chain packing [147]. Further studies have shown that the characteristics of these designer molecules can be further adapted, for example to give hydrogels that undergo reversible gelation in response to changes in pH or temperature [148] or by including a range of non-natural amino acids [132,133] Several groups have examined new structures based around the consensus repeats found in various silks, particularly N. clavipes spider dragline silk with its remarkable mechanical properties [120,149–154]. The many different silks that a spider makes allow a comparison of structure with function, linking mechanical properties to the peptide modules that confer those properties. By using genetic engineering to mix the modules in specific proportions, proteins with defined strength and elasticity can be designed which have many potential medical and engineering uses [119]. Synthetic genes encoding recombinant spider silk proteins have been constructed, cloned, and expressed by various groups [149–151]. In general, short monomer sequences, based on observed consensus regions, have been made into multimers [149]. For example, a glycine-rich sequence, GlyLeuGlyGlyGlnGlyGlyGlyAlaGlyGlnGlyGlyTyrGly, designated SCAP(1) from spidroin 1, was formed into a range of repeats from 4 up to 13 [152], while others made constructs with up to 32 contiguous units of a consensus repeat sequence [150] and polymer constructs which yield proteins of up to 163 kDa [155]. Escherichia coli has been the general choice for expression [149,150,155], and codon usage patterns have been used to enhance expression [149]. However, for larger constructs, yield and homogeneity of the product was limited by premature termination of synthesis [155]. Various strategies have been used to purify the products, including, chelation chromatography
Recombinant Protein Scaffolds
243
with subsequent removal of the poly-His tail. Yields of product from E. coli have been 5.2 mg of lyophilized uncleaved polypeptides per liter of fermentation [152]. An alternative host, the yeast P. pastoris, has proved to have several advantages. For example, genes of 3000 codons in length or longer could be expressed with no evidence of the truncated synthesis [151]. Also, the silk analogs could be secreted by P. pastoris when fused to both the signal sequence and N terminal prosequence of the Saccharomyces cerevisiae alphamating factor gene. [130]. The best yields, however, appear to be obtained from plant constructs. Thus, recombinant proteins of up to 100 kDa, which exhibit homology of ⬎90% compared to native protein models, could be expressed at up to a level of at least 2% of total soluble protein in the endoplasmic reticulum (ER) of tobacco and potato leaves and potato tubers [154]. Spider dragline proteins with molecular masses of 60 to 140 kDa have also been expressed in mammalian cells [120]. It was possible to spin silk monofilaments from a concentrated aqueous solution of soluble product (60 kDa) under modest shear and coagulation conditions. The spun fibers were water insoluble and exhibited toughness and modulus values comparable to those of native dragline silks but with lower tenacity. It was possible to improve these properties, for example, by postspinning draw [120]. Control of the solubility of the recombinant products can be difficult. Therefore, further modifications to the structure have been made that allow control of the silk protein properties. For example, incorporation of a methionine redox trigger into consensus sequences has been shown [156]. When in the oxidized state, the methionine prevented the formation of the sheets, while the reduced state allowed -sheet formation and, despite the extra bulk of the amino acid, did not cause any sterical limitations nor disrupt the normal macromolecular assembly behavior of the protein [156,157]. A further example is the use of enzymatic phosphorylation and dephosphorylation reactions, where an enzymatic phosphorylation event can be used to control the solution structure [158]. It is also possible to make chimeric proteins based on the silk framework. Thus, Baculovirus-mediated transgenesis of the silkworm allowed a chimeric protein that consisted of fibroin and green fluorescent protein to be expressed in silkworms. The gene product was spun into the cocoon layer, illustrating that gene targeting will enable effective modification and enhancement of silk protein [153]. C. Novel Scaffolds Based on the ␣-Helix and Coiled Coils The coiled-coil motif, which is the key element of the fibrinogen structure, can also be used as a design feature in novel recombinant structure. This can include the development of composite molecules. For example, recombinant human serum albumin can be polymerized and conjugated to fibrinogen using N-succinimidyl 3-(2-pyridyldithio)propionate (SPDP) to enhance platelet binding via the ␣IIb3 receptor [159]. Incorporation of a ligand, L1Ig6, that binds to endothelial cells through the ␣V3 integrin receptor can markedly influence the angiogenic activity of polymerized fibrin gels [160]. More complex constructs, made entirely as recombinant proteins can also be considered. The basic structural requirements of the coiled coil motif have been examined by a number of groups, leading to an excellent understanding of how individual amino acids contribute to the structure, principally when a heptad repeat is present, and how these side chains also detmined the oligomeric state (dimer, trimer, etc.) of the coiled coil [161–164]. These groups have also produced a range of designed structures to test and validate the predictions. These designed peptides can form the basis of a fibrous material that could be a useful tissue engineering scaffold [165]. In this case two complimentary 28 residue synthetic peptides were made by chemical methods. They comprised 14 residue sections in alternate orders, AB and BA, where the individ-
244
Werkmeister et al.
ual A and B sequences form coiled-coils with themselves but not the other. These ‘‘sticky’’ ends promote the formation of fibers [165]. In this case chemical synthesis was use. However, the approach is well suited to recombinant techniques, where additional structural element(s) could also be readily included adjacent to the coiled coil–forming elements.
VIII. CONCLUSION Recombinant proteins offer significant opportunities for scaffolds for tissue engineering, for example, by providing better control of the phenotype and biosynthetic activity of a specific cell type. Once regulatory issues have been clarified, these products would have many advantages, including: Ready availability of product Ease of isolation and purification Uniformity in quality Freedom from infectious agents Availability of otherwise rare proteins Novel variations on natural protein structures Novel structures based on repeating elements Novel structures based on chimeric constructs
REFERENCES
1. Lysaght MJ, Reyes J. The growth of tissue engineering. Tissue Eng 2001; 7:485–493. 2. Werkmeister JA, Ramshaw JAM, Ellender G. Characterisation of a monoclonal antibody against native human type I collagen. Eur. J. Biochem 1990; 187:439–443. 3. Werkmeister JA, Ramshaw JAM. Monoclonal antibodies to type V collagen for immunohistological examination of new tissue deposition associated with biomaterial implants. J. Histochem. Cytochem 1991; 39:1215–1220. 4. Werkmeister JA, Ramshaw JAM. Multiple antigenic determinants on type III collagen. Biochem. J 1991; 274:895–898. 5. Werkmeister JA, Tebb TA, White JF, Ramshaw JAM. Monoclonal antibodies to type VI collagen demonstrate new tissue augmentation of a collagen-based biomaterial implant. J. Histochem. Cytochem 1993; 41:1701–1706. 6. Bateman JF, Lamande S, Ramshaw JAM. Collagen superfamily. In: Comper WD, ed. Extracellular Matrix: Vol. 2, Molecular Components and Interactions. Amsterdam: Harwood Academic Publishers, 1995:22–67. 7. Yamauchi M, Mechanic G. Cross-linking of collagen. In: Nimni ME, ed. Collagen. Boca Raton: CRC Press, 1988:157–172. 8. Kielty CM, Kwan AP, Holmes DF, Schor SL, Grant ME. Type X collagen, a product of hypertrophic chondrocytes. Biochem. J 1985; 227:545–554. 9. Glanville RW. Type IV collagen. In: Mayne R, Burgesson RE, eds. Structure and Function of Collagen Types. Orlando: Academic Press, 1987:43–79. 10. Ramshaw JAM, Werkmeister JA, Glattauer V. Collagen-based biomaterials. Biotechnol. Genet. Eng. Rev 1996; 13:335–382. 11. Booth C, Korossis SA, Wilcox HE, Watterson KG, Kearney JN, Fisher J, Ingham E. Tissue engineering of cardiac valve prostheses. I: Development and histological characterization of an acellular porcine scaffold. J. Heart Valve Dis 2002; 11:457–462.
Recombinant Protein Scaffolds
245
12. Kuberka M, von HD, Schoof H, Heschel I, Rau G. Magnification of the pore size in biodegradable collagen sponges. Int. J. Artif. Organs 2002; 25:67–73. 13. Ramshaw JAM, Shah NK, Brodsky B. Gly-X-Y tripeptide frequencies in collagen: a context for host–guest triple-helical peptides. J. Struct. Biol 1998; 122:86–91. 14. Brodsky B, and Ramshaw JAM. The collagen triple-helix structure. Matrix Biol 1997; 15:545–554. 15. Keller RCA, de Wolf FA. Characterization of helical junction zones in gelatin networks: a spectroscopic approach. Wiley Poly. Networks Group Rev. Ser 1998:107–117. 16. Brodsky B, Shah NK. Protein motifs. 8. The triple-helix motif in proteins. FASEB J 1995; 9: 1537–1546. 17. Hofmann H, Fietzek PP, Kuhn K. Comparative analysis of the sequences of the three collagen chains ␣1(I), ␣2 and ␣1(III). Functional and genetic aspects. J. Mol. Biol 1980; 141:293–314. 18. Ayad S, Boot-Handford RP, Humphries MJ, Kadler KE, Shuttleworth CA. The Extracellular Matrix: Facts Book. London: Academic Press, 1994. 19. Kadler K. Extracellular matrix. 1: Fibril-forming collagens. Protein Profile 1994; 5:519–638. 20. Ricard-Blum S, Dublet B, van der Rest M. Unconventional Collagens. Oxford: Oxford University Press, 2000. 21. Glattauer V, Werkmeister JA, Kirkpatrick A, Ramshaw JAM. Identification of the epitope for a monoclonal antibody that blocks platelet aggregation induced by type III collagen. Biochem. J 1997; 323:45–49. 22. Werkmeister JA, Ramshaw JAM. Immunology of collagen-based biomaterials. In: Wise DL, ed. Handbook of Biomaterials Engineering. New York: Marcel Dekker, 2000:739–759. 23. Kivirikko KI, Myllyharju J. Prolyl 4-hydroxylases and their protein disulfide isomerase subunit. Matrix Biol 1998; 16:357–368. 24. Berg RA, Prockop DJ. The thermal transition of a non-hydroxylated form of collagen. Evidence for a role for hydroxyproline in stabilizing the triple-helix of collagen. Biochem. Biophys. Res. Commun 1973; 52:115–120. 25. Rosenbloom J, Harsch M, Jimenez S. Hydroxyproline content determines the denaturation temperature of chick tendon collagen. Arch. Biochem. Biophys 1973; 158:478–484. 26. Kivirikko KI, Myllyla R. Collagen glycosyltransferases. Int. Rev. Connect. Tissue Res 1979; 8: 23–72. 27. Kivirikko KI, Myllyla R. Posttranslational enzymes in the biosynthesis of collagen: intracellular enzymes. Methods Enzymol 1982; 82:245–304. 28. Ferrari FA, Richardson C, Chambers J, Causey SC, Pollock TJ, Capello J, Crissman JW. DNA composition encoding peptide containing oligopeptide repeating unit(s). US Patent 5,243,038. 29. Connolly K, Buechter DD, Gruskin EA, Zhang G. Incorporation of hydroxyproline into polypeptide(s). US Patent 5,821,089. 30. Fertala A, Sieron AL, Ganguly A, Li SW, Ala-Kokko L, Anumula KR, Prockop DJ. Synthesis of recombinant human procollagen II in a stably transfected tumour cell line (HT1080). Biochem. J 1994; 298:31–37. 31. Lamberg A, Helaakoski T, Myllyharju J, Peltonen S, Notbohm H, Pihlajaniemi T, Kivirikko KI. Characterization of human type III collagen expressed in a baculovirus system. Production of a protein with a stable triple helix requires coexpression with the two types of recombinant prolyl 4hydroxylase subunit. J. Biol. Chem 1996; 271:11988–11995. 32. Myllyharju J. Recombinant collagen trimers from insect cells and yeast. Methods Mol. Biol 2000; 139:39–48. 33. Prockop DJ, Ala-Kokko L, Fertala A, Sieron A, Kivirikko KI, Geddis A, Pihlajaniemi T. Synthesis of human procollagens and collagens in recombiant DNA systems. US Patent 5,593,859. 34. Nokelainen M, Tu H, Vuorela A, Notbohm H, Kivirikko KI, Myllyharju J. High-level production of human type I collagen in the yeast Pichia pastoris. Yeast 2001; 18:797–806. 35. Olsen DR, Leigh SD, Chang R, McMullin H, Ong W, Tai E, Chisholm G, Birk DE, Berg RA, Hitzeman RA, Toman PD. Production of human type I collagen in yeast reveals unexpected new insights into the molecular assembly of collagen trimers. J. Biol. Chem 2001; 276:24038–24043. 36. Toman PD, Chisholm G, McMullin H, Giere LM, Olsen DR, Kovach RJ, Leigh SD, Fong BE, Chang R, Daniels GA, Berg RA, Hitzeman RA. Production of recombinant human type I procollagen
246
37.
38.
39. 40.
41.
42.
43.
44. 45. 46. 47.
48.
49. 50. 51. 52. 53. 54. 55.
56. 57.
Werkmeister et al. trimers using a four-gene expression system in the yeast Saccharomyces cerevisiae. J. Biol. Chem 2000; 275:23303–23309. Vaughan PR, Galanis M, Richards KM, Tebb TA, Ramshaw JA, Werkmeister JA. Production of recombinant hydroxylated human type III collagen fragment in Saccharomyces cerevisiae. DNA Cell Biol 1998; 17:511–518. Werten MW, Wisselink WH, Jansen-Van Den Bosch TJ, de Bruin EC, de Wolf FA. Secreted production of a custom-designed, highly hydrophilic gelatin in Pichia pastoris. Protein Eng 2001; 14:447–454. Werten MW, van den Bosch TJ, Wind RD, Mooibroek H, de Wolf FA. High-yield secretion of recombinant gelatins by Pichia pastoris. Yeast 1999; 15:1087–1096. Ruggiero F, Exposito JY, Bournat P, Gruber V, Perret S, Comte J, Olagnier B, Garrone R, Theisen M. Triple helix assembly and processing of human collagen produced in transgenic tobacco plants. FEBS Lett 2000; 469:132–136. Perret S, Merle C, Bernocco S, Berland P, Garrone R, Hulmes DJ, Theisen M, Ruggiero F. Unhydroxylated triple helical collagen I produced in transgenic plants provides new clues on the role of hydroxyproline in collagen folding and fibril formation. J. Biol. Chem 2001; 276:43693–43698. Merle C, Perret S, Lacour T, Jonval V, Hudaverdian S, Garrone R, Ruggiero F, Theisen M. Hydroxylated human homotrimeric collagen I in Agrobacterium tumefaciens–mediated transient expression and in transgenic tobacco plant. FEBS Lett 2002; 515:114–118. Toman PD, Pieper F, Sakai N, Karatzas C, Platenburg E, de Wit I, Samuel C, Dekker A, Daniels GA, Berg RA, Platenburg GJ. Production of recombinant human type I procollagen homotrimer in the mammary gland of transgenic mice. Transgenic Res 1999; 8:415–427. John DC, Watson R, Kind AJ, Scott AR, Kadler KE, Bulleid NJ. Expression of an engineered form of recombinant procollagen in mouse milk. Nat. Biotechnol 1999; 17:385–389. Toman PD, Berg RA, Daniels GA, Hitzeman RA, Chisholm GE. Production of mammalian collagen or pro-collagen using recombinant yeast cells. Patent Application WO 9714431. Myllyharju J, Nokelainen M, Vuorela A, Kivirikko KI. Expression of recombinant human type I–III collagens in the yeast Pichia pastoris. Biochem. Soc. Trans 2000; 28:353–357. Vuorela A, Myllyharju J, Nissi R, Pihlajaniemi T, Kivirikko KI. Assembly of human prolyl 4hydroxylase and type III collagen in the yeast Pichia pastoris: formation of a stable enzyme tetramer requires coexpression with collagen and assembly of a stable collagen requires coexpression with prolyl 4-hydroxylase. EMBO J 1997; 16:6702–6712. de Bruin EC, de Wolf FA, Laane NC. Expression and secretion of human ␣1(I) procollagen fragment by Hansenula polymorpha as compared to Pichia pastoris. Enzyme Microb. Technol 2000; 26: 640–644. Kivirikko KI, Pihlajaniemi T. Synthesis of human procollagens and collagens in recombinant DNA systems. Patent Application WO 9738710. Gosline J, Lillie M, Carrington E, Guerette P, Ortlepp C, Savage K. Elastic proteins: biological roles and mechanical properties. Philos. Trans. Roy. Soc. Lond. B. Biol. Sci 2002; 357:121–132. Vrhovski B, Weiss AS. Biochemistry of tropoelastin. Eur. J. Biochem 1998; 258:1–18. Jensen SA, Reinhardt DP, Gibson MA, Weiss AS. Protein interaction studies of MAGP-1 with tropoelastin and fibrillin-1. J. Biol. Chem 2001; 276:39661–39666. Ross R, Fialkow RJ, Altman LK. The morphogenesis of elastic fibers. Adv. Exp. Med. Biol 1977; 79:7–17. Davison IG, Wright GM, DeMont ME. The structure and physical properties of invertebrate and primitive vertebrate arteries. J. Exp. Biol 1995; 198:2185–2196. Stone PJ, Morris SM, Griffin S, Mithieux S, Weiss AS. Building Elastin. Incorporation of recombinant human tropoelastin into extracellular matrices using nonelastogenic rat-1 fibroblasts as a source for lysyl oxidase. Am. J. Respir. Cell. Mol. Biol 2001; 24:733–739. Davis EC. Stability of elastin in the developing mouse aorta: a quantitative radioautographic study. Histochemistry 1993; 100:17–26. Starcher B, Percival S. Elastin turnover in the rat uterus. Connect. Tissue Res 1985; 13:207–215.
Recombinant Protein Scaffolds
247
58. Senior RM, Griffin GL, Mecham RP, Wrenn DS, Prasad KU, Urry DW. Val-Gly-Val-Ala-Pro-Gly, a repeating peptide in elastin, is chemotactic for fibroblasts and monocytes. J. Cell Biol 1984; 99: 870–874. 59. Long MM, King VJ, Prasad KU, Urry DW. Chemotaxis of fibroblasts toward nonapeptide of elastin. Biochim. Biophys. Acta 1988; 968:300–311. 60. Grosso LE, Scott M. PGAIPG, a repeated hexapeptide of bovine tropoelastin, is a ligand for the 67-kDa bovine elastin receptor. Matrix 1993; 13:157–164. 61. McPherson DT, Morrow C, Minehan DS, Wu J, Hunter E, Urry DW. Production and purification of a recombinant elastomeric polypeptide, G-(VPGVG)19-VPGV, from Escherichia coli. Biotechnol. Prog 1992; 8:347–352. 62. Nicol A, Gowda DC, Urry DW. Cell adhesion and growth on synthetic elastomeric matrices containing Arg-Gly-Asp-Ser-3. J. Biomed. Mater. Res 1992; 26:393–413. 63. Urry DW. Entropic elastic processes in protein mechanisms. I. Elastic structure due to an inverse temperature transition and elasticity due to internal chain dynamics. J. Protein Chem 1988; 7:1–34. 64. Vrhovski B, Jensen S, Weiss AS. Coacervation characteristics of recombinant human tropoelastin. Eur. J. Biochem 1997; 250:92–98. 65. Toonkool P, Jensen SA, Maxwell AL, Weiss AS. Hydrophobic domains of human tropoelastin interact in a context-dependent manner. J. Biol. Chem 2001; 276:44575–44580. 66. Wu WJ, Vrhovski B, Weiss AS. Glycosaminoglycans mediate the coacervation of human tropoelastin through dominant charge interactions involving lysine side chains. J. Biol. Chem 1999; 274: 21719–21724. 67. Foster JA, Bruenger E, Gray WR, Sandberg LB. Isolation and amino acid sequences of tropoelastin peptides. J. Biol. Chem 1973; 248:2876–2879. 68. Indik Z, Yeh H, Ornstein-Goldstein N, Sheppard P, Anderson N, Rosenbloom JC, Peltonen L, Rosenbloom J. Alternative splicing of human elastin mRNA indicated by sequence analysis of cloned genomic and complementary DNA. Proc. Natl. Acad. Sci. USA 1987; 84:5680–5684. 69. Fazio MJ, Olsen DR, Kauh EA, Baldwin CT, Indik Z, Ornstein-Goldstein N, Yeh H, Rosenbloom J, Uitto J. Cloning of full-length elastin cDNAs from a human skin fibroblast recombinant cDNA library: further elucidation of alternative splicing utilizing exon-specific oligonucleotides. J. Invest. Dermatol 1988; 91:458–464. 70. Urry DW, Pattanaik A, Xu J, Woods TC, McPherson DT, Parker TM. Elastic protein-based polymers in soft tissue augmentation and generation. J. Biomater. Sci. Polym. Ed 1998; 9:1015–1048. 71. Kajitani M, Wadia Y, Hinds MT, Teach J, Swartz KR, Gregory KW. Successful repair of esophageal injury using an elastin based biomaterial patch. ASAIO J 2001; 47:342–345. 72. Kajitani M, Wadia Y, Xie H, Hinds MT, Shalaby SW, Swartz KR, Gregory KW. Use of a new elastin patch and glue for repair of a major duodenal injury. ASAIO J 2000; 46:409–414. 73. Dutoya S, Lefebvre F, Deminieres C, Rouais F, Verna A, Kozluca A, Le Bugle A, Rabaud M. Unexpected original property of elastin derived proteins: spontaneous tight coupling with natural and synthetic polymers. Biomaterials 1998; 19:147–155. 74. Dutoya S, Verna A, Lefebvre F, Rabaud M. Elastin-derived protein coating onto poly(ethylene terephthalate). Technical, microstructural and biological studies. Biomaterials 2000; 21:1521–1529. 75. Da Costa-Noble R, Soustre EC, Cadot S, Lauverjat Y, Lefebvre F, Rabaud M. Evaluation of bioabsorbable elastin–fibrin matrix as a barrier in surgical periodontal treatment. J. Periodontol 1996; 67: 927–934. 76. Indik Z, Abrams WR, Kucich U, Gibson CW, Mecham RP, Rosenbloom J. Production of recombinant human tropoelastin: characterization and demonstration of immunologic and chemotactic activity. Arch. Biochem. Biophys 1990; 280:80–86. 77. Grosso LE, Parks WC, Wu LJ, Mecham RP. Fibroblast adhesion to recombinant tropoelastin expressed as a protein A–fusion protein. Biochem. J 1991; 273:517–522. 78. Bedell-Hogan D, Trackman P, Abrams W, Rosenbloom J, Kagan H. Oxidation, cross-linking, and insolubilization of recombinant tropoelastin by purified lysyl oxidase. J. Biol. Chem 1993; 268: 10345–10350. 79. Martin SL, Vrhovski B, Weiss AS. Total synthesis and expression in Escherichia coli of a gene encoding human tropoelastin. Gene 1995; 154:159–166.
248
Werkmeister et al.
80. Bonekamp F, Andersen HD, Christensen T, Jensen KF. Codon-defined ribosomal pausing in Escherichia coli detected by using the pyrE attenuator to probe the coupling between transcription and translation. Nucleic Acids Res 1985; 13:4113–4123. 81. Sandberg LB, Wolt TB. Production and isolation of soluble elastin from copper-deficient swine. Methods Enzymol 1982; 82:657–665. 82. Jensen SA, Vrhovski B, Weiss AS. Domain 26 of tropoelastin plays a dominant role in association by coacervation. J. Biol. Chem 2000; 275:28449–28454. 83. Toonkool P, Regan DG, Kuchel PW, Morris MB, Weiss AS. Thermodynamic and hydrodynamic properties of human tropoelastin. Analytical ultracentrifuge and pulsed field-gradient spin-echo NMR studies. J. Biol. Chem 2001; 276:28042–28050. 84. Bellingham CM, Woodhouse KA, Robson P, Rothstein SJ, Keeley FW. Self-aggregation characteristics of recombinantly expressed human elastin polypeptides. Biochim. Biophys. Acta 2001; 1550: 6–19. 85. Keeley FW, Bellingham CM, Woodhouse KA. Elastin as a self-organizing biomaterial: use of recombinantly expressed human elastin polypeptides as a model for investigations of structure and self-assembly of elastin. Philos. Trans. Roy. Soc. Lond. B. Biol. Sci 2002; 357:185–189. 86. Brown JH, Volkmann N, Jun G, Henschen-Edman AH, Cohen C. The crystal structure of modified bovine fibrinogen. Proc. Natl. Acad. Sci. USA 2000; 97:85–90. 87. Ferry JD, Morrison PR. Preparation and properties of serum and plasma proteins. J. Am. Chem. Soc 1947; 69:388–400. 88. Everse SJ, Spraggon G, Veerapandian L, Doolittle RF. Conformational changes in fragments D and double-D from human fibrin(ogen) upon binding the peptide ligand Gly-His-Arg-Pro-amide. Biochemistry 1999; 38:2941–2946. 89. Meinhart J, Fussenegger M, Hobling W. Stabilization of fibrin-chondrocyte constructs for cartilage reconstruction. Ann. Plast. Surg 1999; 42:673–678. 90. Wechselberger G, Russell RC, Neumeister MW, Schoeller T, Piza-Katzer H, Rainer C. Successful transplantation of three tissue-engineered cell types using capsule induction technique and fibrin glue as a delivery vehicle. Plast. Reconstr. Surg 2002; 110:123–129. 91. Perka C, Arnold U, Spitzer RS, Lindenhayn K. The use of fibrin beads for tissue engineering and subsequential transplantation. Tissue Eng 2001; 7:359–361. 92. Jockenhoevel S, Zund G, Hoerstrup SP, Chalabi K, Sachweh JS, Demircan L, Messmer BJ, Turina M. Fibrin gel — advantages of a new scaffold in cardiovascular tissue engineering. Eur. J. Cardiothorac. Surg 2001; 19:424–430. 93. Huang Q, Goh JC, Hutmacher DW, Lee EH. In vivo mesenchymal cell recruitment by a scaffold loaded with transforming growth factor 1 and the potential for in situ chondrogenesis. Tissue Eng 2002; 8:469–482. 94. Andree C, Voigt M, Wenger A, Erichsen T, Bittner K, Schaefer D, Walgenbach KJ, Borges J, Horch RE, Eriksson E, Stark GB. Plasmid gene delivery to human keratinocytes through a fibrinmediated transfection system. Tissue Eng 2001; 7:757–766. 95. Chung DW, Rixon MW, Que BG, Davie EW. Cloning of fibrinogen genes and their cDNA. Ann. NY Acad. Sci 1983; 408:449–456. 96. Rixon MW, Chan WY, Davie EW, Chung DW. Characterization of a complementary deoxyribonucleic acid coding for the alpha chain of human fibrinogen. Biochemistry 1983; 22:3237–3244. 97. Chung DW, Que BG, Rixon MW, Mace M, Jr Davie EW. Characterization of complementary deoxyribonucleic acid and genomic deoxyribonucleic acid for the beta chain of human fibrinogen. Biochemistry 1983; 22:3244–3250. 98. Chung DW, Chan WY, Davie EW. Characterization of a complementary deoxyribonucleic acid coding for the gamma chain of human fibrinogen. Biochemistry 1983; 22:3250–3256. 99. Bolyard MG, Lord ST. High-level expression of a functional human fibrinogen gamma chain in Escherichia coli. Gene 1988; 66:183–192. 100. Danishefsky K, Hartwig R, Banerjee D, Redman C. Intracellular fate of fibrinogen B beta chain expressed in COS cells. Biochim. Biophys. Acta 1990; 1048:202–208. 101. Lord ST, Gorkun OV. Insight from studies with recombinant fibrinogens. Ann. NY Acad. Sci 2001; 936:101–116.
Recombinant Protein Scaffolds
249
102. Binnie CG, Hettasch JM, Strickland E, Lord ST. Characterization of purified recombinant fibrinogen: partial phosphorylation of fibrinopeptide A. Biochemistry 1993; 32:107–113. 103. Remijn JA, Ijsseldijk MJ, van Hemel BM, Galanakis DK, Hogan KA, Lounes KC, Lord ST, Sixma JJ, de Groot PG. Reduced platelet adhesion in flowing blood to fibrinogen by alterations in segment ␥316–322, part of the fibrin-specific region. Br. J. Haematol 2002; 117:650–657. 104. Bach TL, Barsigian C, Yaen CH, Martinez J. Endothelial cell VE-cadherin functions as a receptor for the 15–42 sequence of fibrin. J. Biol. Chem 1998; 273:30719–30728. 105. Gorlatov S, Medved L. Interaction of fibrin(ogen) with the endothelial cell receptor VE-cadherin: mapping of the receptor-binding site in the NH2-terminal portions of the fibrin beta chains. Biochemistry 2002; 41:4107–4116. 106. Minoura N, Tsukada M, Nagura M. Physico-chemical properties of silk fibroin membrane as a biomaterial. Biomaterials 1990; 11:430–434. 107. Minoura N, Aiba S, Higuchi M, Gotoh Y, Tsukada M, Imai Y. Attachment and growth of fibroblast cells on silk fibroin. Biochem. Biophys. Res. Commun 1995; 208:511–516. 108. Furuzono T, Ishihara K, Nakabayashi N, Tamada Y. Chemical modification of silk fibroin with 2methacryloyloxyethyl phosphorylcholine. II. Graft-polymerization onto fabric through 2-methacryloyloxyethyl isocyanate and interaction between fabric and platelets. Biomaterials 2000; 21: 327–333. 109. Sofia S, McCarthy MB, Gronowicz G, Kaplan DL. Functionalized silk-based biomaterials for bone formation. J. Biomed. Mater. Res 2001; 54:139–148. 110. Altman GH, Horan RL, Lu HH, Moreau J, Martin I, Richmond JC, Kaplan DL. Silk matrix for tissue engineered anterior cruciate ligaments. Biomaterials 2002; 23:4131–4141. 111. Denny MW. Silks—their properties and functions. Symp. Soc. Exp. Biol 1980; 34:247–272. 112. Mita K, Ichimura S, Zama M, James TC. Specific codon usage pattern and its implications on the secondary structure of silk fibroin mRNA. J. Mol. Biol 1988; 203:917–925. 113. Strydom DJ, Haylett T, Stead RH. The amino-terminal sequence of silk fibroin peptide CP—a reinvestigation. Biochem. Biophys. Res. Commun 1977; 79:932–938. 114. Ohshima Y, Suzuki Y. Cloning of the silk fibroin gene and its flanking sequences. Proc. Natl. Acad. Sci. USA 1977; 74:5363–5367. 115. Xu M, Lewis RV. Structure of a protein superfiber: spider dragline silk. Proc. Natl. Acad. Sci. USA 1990; 87:7120–7124. 116. Arcidiacono S, Mello C, Kaplan D, Cheley S, Bayley H. Purification and characterization of recombinant spider silk expressed in Escherichia coli. Appl. Microbiol. Biotechnol 1998; 49:31–38. 117. Hayashi CY, Shipley NH, Lewis RV. Hypotheses that correlate the sequence, structure, and mechanical properties of spider silk proteins. Int. J. Biol. Macromol 1999; 24:271–275. 118. Gosline JM, Guerette PA, Ortlepp CS, Savage KN. The mechanical design of spider silks: from fibroin sequence to mechanical function. J. Exp. Biol 1999; 202:3295–3303. 119. Hinman MB, Jones JA, Lewis RV. Synthetic spider silk: a modular fiber. Trends Biotechnol 2000; 18:374–379. 120. Lazaris A, Arcidiacono S, Huang Y, Zhou JF, Duguay F, Chretien N, Welsh EA, Soares JW, Karatzas CN. Spider silk fibers spun from soluble recombinant silk produced in mammalian cells. Science 2002; 295:472–476. 121. Lawn RM, Adelman J, Bock SC, Franke AE, Houck CM, Najarian RC, Seeburg PH, Wion KL. The sequence of human serum albumin cDNA and its expression in E. coli.. Nucleic Acids Res 1981; 9:6103–6114. 122. Qiu RD, Li SY, Chen JG, Wu XF, Yuan ZY. High expression and purification of recombinant human serum albumin from Pichia pastoris. Sheng Wu Hua Xue Yu Sheng Wu Wu Li Xue Bao (Shanghai) 2000; 32:59–62. 123. Sijmons PC, Dekker BM, Schrammeijer B, Verwoerd TC, van den Elzen PJ, Hoekema A. Production of correctly processed human serum albumin in transgenic plants. Biotechnology 1990; 8:217–221. 124. Kornblihtt AR, Umezawa K, Vibe-Pedersen K, Baralle FE. Primary structure of human fibronectin: differential splicing may generate at least 10 polypeptides from a single gene. EMBO J 1985; 4: 1755–1759.
250
Werkmeister et al.
125. Umezawa K, Kornblihtt AR, Baralle FE. Isolation and characterization of cDNA clones for human liver fibronectin. FEBS Lett 1985; 186:31–34. 126. Waite JH, Qin XX, Coyne KJ. The peculiar collagens of mussel byssus. Matrix Biol 1998; 17: 93–106. 127. Inoue K, Takeuchi Y, Miki D, Odo S. Mussel adhesive plaque protein gene is a novel member of epidermal growth factor–like gene family. J. Biol. Chem 1995; 270:6698–6701. 128. Benedict CV, Waite JH. Location and analysis of byssal structural proteins of Mytilus edulis. J. Morphol 1986; 189:171–181. 129. Qin X, Waite JH. Exotic collagen gradients in the byssus of the mussel Mytilus edulis. J. Exp. Biol 1995; 198:633–644. 130. Fahnestock SR, Yao Z, Bedzyk LA. Microbial production of spider silk proteins. J. Biotechnol 2000; 74:105–119. 131. Cowie DB, Cohen GN. Biosynthesis by Escherichia coli of active altered proteins containing selenium instead of sulfur. Biochim. Biophys. Acta 1957; 26:252–261. 132. Kiick KL, Weberskirch R, Tirrell DA. Identification of an expanded set of translationally active methionine analogues in Escherichia coli. FEBS Lett 2001; 502:25–30. 133. Kiick KL, Saxon E, Tirrell DA, Bertozzi CR. Incorporation of azides into recombinant proteins for chemoselective modification by the Staudinger ligation. Proc. Natl. Acad. Sci. USA 2002; 99:19–24. 134. Tang Y, Tirrell DA. Biosynthesis of a highly stable coiled-coil protein containing hexafluoroleucine in an engineered bacterial host. J. Am. Chem. Soc 2001; 123:11089–11090. 135. Datta D, Wang P, Carrico IS, Mayo SL, Tirrell DA. A designed phenylalanyl-tRNA synthetase variant allows efficient in vivo incorporation of aryl ketone functionality into proteins. J. Am. Chem. Soc 2002; 124:5652–5653. 136. Fields GB. A model for interstitial collagen catabolism by mammalian collagenases. J. Theor. Biol 1991; 153:585–602. 137. Emsley J, King SL, Bergelson JM, Liddington RC. Crystal structure of the I domain from integrin ␣21. J. Biol. Chem 1997; 272:28512–28517. 138. Emsley J, Knight CG, Farndale RW, Barnes MJ, Liddington RC. Structural basis of collagen recognition by integrin ␣21. Cell 2000; 101:47–56. 139. Vaughan PR, Galanis M, Ramshaw JAM, Werkmeister JA. Stable expression of hydroxylated triple helical proteins. Patent Application WO 9818918. 140. Hayashi M, Tomita M, Yoshizato K. Production of EGF-collagen chimeric protein which shows the mitogenic activity. Biochim. Biophys. Acta 2001; 1528:187–195. 141. Hayashi M, Tomita M, Yoshizato K. Interleukin-2–collagen chimeric protein which liberates interleukin-2 upon collagenolysis. Protein Eng 2002; 15:429–436. 142. Yang W, Chan VC, Kirkpatrick A, Ramshaw JAM, Brodsky B. Gly-Pro-Arg confers stability similar to Gly-Pro-Hyp in the collagen triple-helix of host–guest peptides. J. Biol. Chem 1997; 272: 28837–28840. 143. Persikov AV, Ramshaw JAM, Kirkpatrick A, Brodsky B. Amino acid propensities for the collagen triple-helix. Biochemistry 2000; 39:14960–14967. 144. Shah NK, Sharma M, Kirkpatrick A, Ramshaw JAM, Brodsky B. Gly-Gly–containing triplets of low stability adjacent to a type III collagen epitope. Biochemistry 1997; 36:5878–5883. 145. Krejchi MT, Atkins ED, Waddon AJ, Fournier MJ, Mason TL, Tirrell DA. Chemical sequence control of beta-sheet assembly in macromolecular crystals of periodic polypeptides. Science 1994; 265:1427–1432. 146. McGrath KP, Tirrell DA, Kawai M, Mason TL, Fournier MJ. Chemical and biosynthetic approaches to the production of novel polypeptide materials. Biotechnol. Prog 1990; 6:188–192. 147. Cantor EJ, Atkins ED, Cooper SJ, Fournier MJ, Mason TL, Tirrell DA. Effects of amino acid sidechain volume on chain packing in genetically engineered periodic polypeptides. J. Biochem. (Tokyo) 1997; 122:217–225. 148. Petka WA, Harden JL, McGrath KP, Wirtz D, Tirrell DA. Reversible hydrogels from self-assembling artificial proteins. Science 1998; 281:389–392. 149. Prince JT, McGrath KP, DiGirolamo CM, Kaplan DL. Construction, cloning, and expression of synthetic genes encoding spider dragline silk. Biochemistry 1995; 34:10879–10885.
Recombinant Protein Scaffolds
251
150. Lewis RV, Hinman M, Kothakota S, Fournier MJ. Expression and purification of a spider silk protein: a new strategy for producing repetitive proteins. Protein Expr. Purif 1996; 7:400–406. 151. Fahnestock SR, Bedzyk LA. Production of synthetic spider dragline silk protein in Pichia pastoris. Appl. Microbiol. Biotechnol 1997; 47:33–39. 152. Fukushima Y. Genetically engineered syntheses of tandem repetitive polypeptides consisting of glycine-rich sequence of spider dragline silk. Biopolymers 1998; 45:269–279. 153. Mori H, Tsukada M. New silk protein: modification of silk protein by gene engineering for production of biomaterials. J. Biotechnol 2000; 74:95–103. 154. Scheller J, Guhrs KH, Grosse F, Conrad U. Production of spider silk proteins in tobacco and potato. Nat. Biotechnol 2001; 19:573–577. 155. Fahnestock SR, Irwin SL. Synthetic spider dragline silk proteins and their production in Escherichia coli. Appl. Microbiol. Biotechnol 1997; 47:23–32. 156. Winkler S, Szela S, Avtges P, Valluzzi R, Kirschner DA, Kaplan DL. Designing recombinant spider silk proteins to control assembly. Int. J. Biol. Macromol 1999; 24:265–270. 157. Szela S, Avtges P, Valluzzi R, Winkler S, Wilson D, Kirschner D, Kaplan DL. Reduction-oxidation control of beta-sheet assembly in genetically engineered silk. Biomacromolecules 2000; 1:534–542. 158. Winkler S, Wilson D, Kaplan DL. Controlling beta-sheet assembly in genetically engineered silk by enzymatic phosphorylation/dephosphorylation. Biochemistry 2000; 39:12739–12746. 159. Takeoka S, Teramura Y, Okamura Y, Handa M, Ikeda Y, Tsuchida E. Fibrinogen-conjugated albumin polymers and their interaction with platelets under flow conditions. Biomacromolecules 2001; 2:1192–1197. 160. Hall H, Baechi T, Hubbell JA. Molecular properties of fibrin-based matrices qfor promotion of angiogenesis in vitro. Microvasc. Res 2001; 62:315–326. 161. Parry DAD. Coiled-coils in alpha-helix–containing proteins: analysis of the residue types within the heptad repeat and the use of these data in the prediction of coiled-coils in other proteins. Biosci. Rep 1982; 2:1017–1024. 162. Zhou NE, Kay CM, Hodges RS. Synthetic model proteins. Positional effects of interchain hydrophobic interactions on stability of two-stranded alpha-helical coiled-coils. J. Biol. Chem 1992; 267: 2664–2670. 163. Woolfson DN, Alber T. Predicting oligomerization states of coiled coils. Protein Sci 1995; 4: 1596–1607. 164. Betz S, Fairman R, O’Neil K, Lear J, Degrado W. Design of two-stranded and three-stranded coiledcoil peptides. Philos. Trans. Roy. Soc. Lond. B. Biol. Sci 1995; 348:81–88. 165. Pandya MJ, Spooner GM, Sunde M, Thorpe JR, Rodger A, Woolfson DN. Sticky-end assembly of a designed peptide fiber provides insight into protein fibrillogenesis. Biochemistry 2000; 39: 8728–8734.
12 Auricular Cartilage Tissue Engineering Doreen Rosenstrauch, Kamuran Kadipasaoglu, Harnath Shelat, Pierre Zoldhelyi, and O. H. Frazier University of Texas Health Sciences Center at Houston, and Texas Heart Institute at St. Luke’s Episcopal Hospital, Houston, Texas, U.S.A.
I. INTRODUCTION The persistent need for cartilage replacement material in fields such as head and neck surgery, urology, and orthopedics has led to the development of numerous methods for the tissue engineering of cartilage and to a variety of laboratory and clinical research programs. One particular area of research on cartilage tissue engineering relates to cartilage tissue derived from the ear (auricular cartilage). An overview of developments in this area is provided in this chapter.
II. BACKGROUND A. Embryology A fertilized egg, or zygote, differentiates into an embryo, which then develops into a mature organism. Adult mammalian organisms consist of more than 200 kinds of cells [1]. All cells of the mature organism derive from one of three embryonic germ layers: ectoderm (external layer), mesoderm (middle layer), or endoderm (internal layer). Cells derived from the ectoderm include skin (epithelium); nerve cells (neurons); pigment cells; tooth enamel; cells of the adrenal medulla and pituitary, mammary, and subcutaneous glands; and sensory epithelia of the eye, ear, and nose. Cells derived from the mesoderm include cells of the spleen, adrenal cortex, lymphatic vessels, and urogenital system (kidney tubule cells); blood vessel (vascular) cells; cardiac, smooth, and skeletal muscle cells (myocytes); bone marrow and blood cells (erythrocytes, monocytes, and lymphocytes); cells of the genital ducts, ovaries, and testes; bone cells (osteocytes); serous membranes of the body cavities (pericardium, pleura, and peritoneum); and cartilage cells (chondrocytes). Cells derived from the endoderm include cell of the thymus, thyroid, tonsils, parathyroid glands, larynx, trachea, and vagina; cells of the gastrointestinal organs (liver, pancreas); and epithelial cells lining the gastrointestinal and respiratory tracts (alveolar cells), tympanic cavity, urinary bladder, and most of the urethra. Cells that can give rise to all three embryonic germ layers are called pluripotent (Latin, plures potentia, many power). Pluripotent cells can also give rise to any type of cell in the mature 253
254
Rosenstrauch et al.
organism. Cells that can differentiate along only one lineage are called unipotent (Latin, unus potentia, one power). The adult cells in many differentiated, undamaged tissues are typically unipotent and give rise to just one cell type under normal conditions. This theoretically allows for a steady state of tissue self-renewal [2]. Cartilage harvested from a mature organism is unipotent. B. Definition of Cartilage Cartilage is connective tissue that is free of blood vessels (avascular), is free of nerves, and has supportive functions. Since cartilage is relatively avascular, nutrients are received by passive diffusion from blood vessels of the outer cell layer, or perichondrium. Cartilage forms the skeleton of mammalian embryos. The embryonic skeleton is found only in the earlier stages of development; later, as the embryo grows, it is gradually replaced by bone. This process, called ossification, is not complete until the organism nears adulthood. The regeneration of cartilage is limited because of its avascularity. Some appositional regeneration (i.e., addition of perichondrial cells) from chondrocyte precursors (chondroblasts) can occur around the periphery. C. Histology of Cartilage 1. Composition of Cartilage In general, cartilage is composed of a combination of chondrocytes and extracellular matrix. The extracellular matrix consists of fibers and ground substance. Most cartilage, except for the articular surfaces of long bones, is covered by perichondrium, which is dense connective tissue. Perichondrium contains blood vessels, which provide nutrients to the cartilage, and chondroblasts, which have the potential to (1) differentiate into chondrocytes, (2) begin the synthesis of matrix material, and (3) wall themselves off into lacunae (Latin, little caves). Chondrocytes are small cells with an oval nucleus and one or two nucleoli. Because they are derived from the same cell lineage as fibroblasts, chondrocytes greatly resemble fibroblasts. Chondrocytes are located in lacunae within the extracellular matrix. They are rich in endoplasmic reticulum and produce extracellular matrix components such as chondronectin, which increases the adhesiveness of chondrocytes. Fibers are small (10–20 nm in diameter), elongated, threadlike structures each consisting of several strands of type II collagen. The fibers provide shape, retention, and tensile strength. Other fibers are type I collagen and elastin. Ground substance is a homogeneous material principally composed of proteoglycans noncovalently linked to hyaluronic acid. Proteoglycans are macromolecules made of proteins and carbohydrates. The carbohydrates are glycosaminoglycans, primarily chondroitin sulfate and keratan sulfate. Hyaluronic acid is a gelatinous mucopolysaccharide that binds the proteoglycans together into large aggregates. The matrix is relatively basophilic because of the presence of chondroitin-4-sulfate and chondroitin-6-sulfate. The ground substance fills the spaces in the meshwork of type II collagen fibers. The viscous, hydrated matrix absorbs compression forces. 2. Types of Cartilage Some cartilage persists in the adult organism in areas that require resilient but flexible stiffening. Different histological types of cartilage answering to different bodily mechanical demands can be found in the mature mammalian organism. There are three basic types of cartilage: hyaline, fibrous, and elastic. Hyaline cartilage is the most common type. It is found in articular surfaces of the joints, trachea, larynx, costae,
Auricular Cartilage Tissue Engineering
255
and tip of the nose (e.g., between the ribs and breastbone; and between the ends of bone where they meet and form joints). Perichondrium is surrounding hyaline cartilage. The cartilage matrix is smooth and basophilic and divided into two regions (the territorial and the matrix). Chondrocytes are enclosed in lacunae. Fibers are mainly type II collagen. Fibrous cartilage is found in places where high stress occurs such as in the intervertebral disks, between the pubic bones in front of the pelvic girdle, and around the edges of the articular cavities (e.g., the glenoid cavity in the shoulder joint). It never occurs alone. It is closely associated with either dense connective tissue or with hyaline cartilage. Perichondrium is usually absent. The matrix component is minimal; the fibrous component (mainly type I collagen) predominates. The chondrocytes are less numerous and much more widely separated than in other types of cartilage, but most are still enclosed in lacunae. Elastic cartilage is found in the pinna of the ear (auricula), the walls of the eustachian tube, and the epiglottis. These are all places in which the maintenance of a specific shape is important for proper function. Perichondrium surrounds the elastic cartilage. Elastic cartilage contains more fibers than does hyaline cartilage. The majority of those fibers are made of elastin, though some are made of type II collagen. The elastin fibers form a network whose properties give elastic cartilage its ability to be deformed and to spring back into shape immediately. The chondrocytes in elastic cartilage are more tightly packed together than in hyaline cartilage [3].
III. TISSUE ENGINEERING OF CHONDROCYTES FROM AURICULAR CARTILAGE A. Harvesting and Culture of Cartilage 1. Biopsy 1. Harvest a 2-mm-diameter sample from the ear of a calf by punch biopsy under local anesthesia and sterile conditions (Fig. 1). 2. Immediately place the sample in sterile cell culture medium at 37⬚C containing antibiotics (Table 1) and transfer to a sterile hood. 2. Isolation of Cartilage 1. Under the sterile hood, isolate auricular elastic cartilage by surgically dissecting the dermal tissue surrounding the biopsy specimen. 2. Place the isolated cartilage in a 10- ⳯ 10-mm tissue culture dish with 2% antimycotic–antibiotic phosphate-buffered saline (Table 2) and incubate at 37⬚C for 10 min. 3. Place small pieces of the cartilage in six-well tissue culture plates containing 500 L of modified RPMI cell culture medium in each well. Make sure that the cartilage pieces are in contact with the bottom of the plate and do not float in the cell culture medium. The adherence to the bottom of the plate will ease the attachment of the chondrocytes to the surface of the plate. 4. Place the loaded culture plates in a humidified 5% CO2 incubator and incubate at 37⬚C. 3. Growth of Chondrocytes in Culture 1. Take the chondrocytes growing from the auricular cartilage (Fig. 2) and culture under sterile conditions in a humidified 5% CO2 incubator at 37⬚C.
256
Rosenstrauch et al.
Figure 1 Two-mm-diameter biopsy sample of auricular cartilage harvested from the ear of a calf using a hole puncher.
2. Every 12 h, add two to three drops of modified RPMI to each culture well so that the cartilage pieces are wet but not floating. 3. After 5–7 days, when chondrocytes become visible and adherent to the culture well, remove cartilage pieces. 4. When the bottom of the culture well is covered with chondrocytes, passage the chondrocytes through the tissue culture as follows. 5. Trypsinize the chondrocytes in a solution of 0.5 g trypsin and 0.2 g EDTA. 6. Incubate mixture at 37⬚C for approximately 5 min. 7. When the cells become detached, add fetal bovine serum to inactivate the trypsin. 8. Add 5 mL of modified RPMI and mix well to get a single cell solution. 9. Transfer the solution into sterile tubes and centrifuge at 1500 rpm for 5 min.
Table 1 Composition of Modified RPMI 1640 Cell Culture Medium RPMI 1640 medium (90%) Fetal bovine serum (10%) (Equitech-Bio, Inc., Kerrville, TX) Sodium pyruvate (1X), (100 mM MEM sodium pyruvate solution, GIBCO-BRL, Rockville, MD) Minimum essential amino acids (1X) L-Glutamine (2 mmol/L) Penicillin G sodium (10,000 U/mL) Streptomycin sulfate (10,000 g/mL) HEPES (1X)
Auricular Cartilage Tissue Engineering
257
Table 2 Composition of 2% Antimycotic–Antibiotic Phosphate-Buffered Saline Solution Dulbecco’s phospate-buffered saline (1X) (Sigma, St. Louis, MO) Penicillin G sodium (10,000 U/mL) Streptomycin sulfate (10,000 g/mL) Amphotericin B (25 g/mL)
10. Discharge the supernatant, which will contain dead cells and a residue of trypsin. 11. Resuspend the resulting cell pellet in modified RPMI. 12. Transfer the resulting single cell suspension to tissue culture plates to establish a monolayer of chondrocytes (Fig. 3). 13. Thereafter, change the cell culture medium every 48 h.
Figure 2 Chondrocytes growing out from a piece of cartilage in tissue culture at zero passage. No staining; magnification 100⳯.
258
Rosenstrauch et al.
Figure 3 Bovine chondrocytes in tissue culture at fourth passage. No staining; magnification 40⳯.
Chondrocytes can be passaged several times using this method. Vitamin C (ascorbic acid) and growth hormones stimulate the synthesis and formation of extracellular matrix components such as collagen. Hydrocortisone has been shown to inhibit proteoglycan synthesis. 4. Histology and Immunocytochemistry of Chondrocytes Histology. 1. Slice paraffin-embedded sections of pure elastic cartilage tissue into 1-m-thick sections. 2. Stain sections with Verhoff–van Gieson stain to visualize elastic fibers, Masson’s
Auricular Cartilage Tissue Engineering
259
trichrome stain to visualize collagen fibers, and hematoxylin and eosin to assess tissue and cell morphology (Fig. 4). Immunocytochemistry. 1. Plate chondrocytes of any passage onto slides and place in a tissue culture chamber. 2. When the chondrocytes have completely covered the surface of the slides, remove the chamber. 3. Fix the chondrocyte monolayer of each slide in 1% formalin. 4. To identify collagen type II fibers, apply a double-antibody labeling technique. The recommended primary antibody is antibody to collagen type II (NCL-Coll-Iip; Novocastra); the recommended secondary antibody is biotinylated immunoglobulin specific for the primary antibody (Vectastain Elite IgG; Vector Laboratories, Burlington, CA). Other antibodies may be applied, as needed, to identify the area of interest. 5. Incubate hydrated chondrocytes in 3% hydrogen peroxide for 10 min to quench endogenous peroxidase activity. 6. Incubate the specimens with normal mouse serum for 20 min to block nonspecific binding sites. 7. Expose the specimens to a 1:50 diluted solution of the primary antibody for 1 h. 8. After repeated washes, incubate all specimens with the secondary antibody for 30 min. 9. Expose all specimens to avidin DH and biotinylated enzyme (Vectastain Elite ABC reagent) for 30 min and then to diaminobenzidine as a peroxidase substrate for 30 s. 10. Counterstain all specimens with hematoxylin for 1 min. Control incubations are useful in the absence of primary antibody. The same immunohistochemi-
Figure 4 Chondrocytes (Cc) growing from a piece of elastic cartilage (Ec) at zero passage and stained for elastic and collagen fibers. Hematoxylin and potassium iodine (Verhoeff’s elastin) stain; magnification 100⳯.
260
Rosenstrauch et al.
cal technique described here can be applied to fine (1-m-thick) paraffin-embedded sections of pure elastic cartilage tissue after the cartilage sections are deparaffinized.
IV. EXPERIMENTAL APPLICATIONS OF CARTILAGE TISSUE ENGINEERING A. Lining of Left Ventricular Assist Devices The method of harvesting, isolating, and culturing auricular chondrocytes described above was recently applied in a study evaluating the feasibility of lining left ventricular assist devices (LVADs) with auricular chondrocytes and exposing the chondrocyte lining to in vivo blood flow conditions [4,15]. In brief, auricular cartilage was harvested from the anesthetized ear of a calf, isolated, and cultured in vitro on tissue culture dishes. Primary chondrocytes were typed by immunocytochemistry, transferred into culture medium, passaged twice, and seeded onto the blood-contacting luminal surfaces of four LVADs (HeartMate; Thoratec Corporation, Woburn, MA). The HeartMate has two luminal artificial surfaces: a flexible diaphragm made of a ‘‘biomer’’ (i.e., polyurethane flocked with polyester microfibrils) and a metal housing made of sintered titanium microspheres. Seeded cell linings were preconditioned under simulated flow conditions to promote cell adhesion to luminal surfaces. Seeding efficiency and cumulative cell loss under flow conditions were quantitated. One of the four autologous chondrocyte-lined LVADs was ultimately implanted into the tissue donor calf under cardiopulmonary bypass using a standard procedure. Seven days later, the calf was euthanized and necropsied. The LVAD was disassembled, inspected, photographed, and subjected to scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The efficiency of seeding chondrocytes onto the luminal surfaces of the four LVADs was 95.11% Ⳮ/ⳮ4.23% (n ⳱ 4). Cumulative cell loss during preconditioning under flow conditions in vitro did not exceed 12% (n ⳱ 4). After 7 days of in vivo implantation, the luminal surfaces of the implanted LVAD demonstrated an intact, strongly adherent cellular lining. SEM revealed an extensive amount of extracellular matrix components and an intact, well-incorporated cellular lining on the sintered titanium and polyurethane surfaces of the implanted LVAD. TEM revealed a well-established monolayer of chondrocytes. No endothelial cells were seen. The authors concluded from this feasibility study that auricular elastic cartilage is a ready and easily accessible source of chondrocytes whose ability to produce collagen II, elastin, and other important extracellular matrix constituents allows them to adhere strongly to the luminal surfaces of LVADs. They also concluded that their simple method of isolating and expanding auricular chondrocytes in tissue culture might be used to provide strongly adherent autologous cell linings for LVADs and other cardiovascular devices. If and when chondrocytes can be genetically engineered to produce antithrombogenic factors and then used to line the luminal surfaces of LVADs or other cardiovascular prostheses, they may be able to improve the hemocompatibility of the blood–biomaterial interface in such devices [4]. B. Auricular Plastic Reconstruction Auricular cartilage derived from juvenile human and juvenile pig ears have been used to reconstruct the entire auricle [5]. In brief, de Chalain et al. isolated chondrocytes enzymatically, agitated them in suspension to form chondronlike aggregates, and then embedded the aggregates in molded hydrogel scaffolds made of alginate and type I collagen augmented with kappaelastin. The scaffolds were then implanted in nude mice and harvested 4 and 12 weeks later.
Auricular Cartilage Tissue Engineering
261
The resulting neocartilage closely resembled native auricular cartilage at the gross, microscopic, and ultrastructural levels. C. Reflux and Incontinence Auricular chondrocytes have been used to treat vesicoureteral reflux in pediatric patients and urinary stress incontinence in adults [6]. In brief, autologous chondrocytes were suspended in a crosslinked alginate hydrogel, which was then injected submucosally at critical sites surrounding the ureteral orifice and the urethra and bladder neck, respectively. In the case of reflux, the cell–alginate matrix serves as a bulking agent to prevent retrograde flow of urine from the bladder to the ureter and kidney. In the case of incontinence, the bulking agent allows proper closure of the urethral sphincter. D. Penile Prostheses Autologous human auricular cartilage tissue has been used as a cell source for engineering cartilaginous penile prostheses [7]. In brief, derived chondrocytes were seeded onto preformed, rod-shaped, biodegradable polymer scaffolds and maintained in a cell culture bioreactor for 1 month. The cell-seeded scaffolds were then implanted into the subcutaneous space of athymic rats. Specimens were retrieved 3 months after initial seeding for histomorphological and biomechanical analyses. Biomechanical properties of the cartilage prostheses, including tension, bending, and elasticity, were compared with those of silicone penile prostheses (AMS 700 inflatable and Jonas silicone rod prostheses). Chondrocytes seeded on polymer scaffolds formed cartilaginous rods that maintained their size and shape. Histological analyses revealed mature and wellformed chondrocytes in implants retrieved 2 months after implantation. Tensile tests showed that the engineered rods were readily elastic and withstood high tensile force. Dynamic bending tests showed that the rods were flexible and durable. No rod ruptured during the mechanical testing. E. Growth of Auricular Cartilage on Endoskeletal Scaffolding Tissue engineered autologous cartilage has been grown on nonreactive, permanent endoskeletal scaffolds [8]. In brief, auricular chondrocytes from Yorkshire swine were suspended in a hydrogel and then incorporated into nonbiodegradable endoskeletal scaffolds made of high-density polyethylene, soft acrylic, polymethylmethacrylate, extra-purified Silastic, or conventional Silastic. When implanted subdermally and harvested after 8 weeks, all showed healthy new cartilage. F. In Utero Tracheal Augmentation in Ovine Model Auricular chondrocytes from fetal lambs have been used to show the feasibility of using prenatal tracheoplasty to treat severe congenital tracheal malformations [9]. In brief, the chondrocytes were seeded dynamically onto biodegradable scaffolds and then maintained in a rotating bioreactor for 6–8 weeks. The seeded scaffolds were then implanted into fetal tracheas. Cartilage in the treated tracheas was found to resemble normal hyaline cartilage and to display epithelium derived from the native trachea. In addition it has been shown that mechanical properties of tissue engineered cartilage can be quantified. For this purpose, auricular chondrocytes from rabbit and humans were seeded onto a template and implanted for 8 or 16 weeks. Ultimate tensile strength for cartilage, stiffness,
262
Rosenstrauch et al.
and resilience were assessed and revealed that the tissue engineered cartilage was significantly different from controls [10]. G. Optimization of Tissue Culture of Auricular Chondrocytes Using Growth Factors Investigations are underway to optimize and enhance the growth and expansion of auricular chondrocytes in tissue culture. Growth factors appear to play a major role in this process, and platelet supernatant has been used as an inexpensive autologous source of multiple growth factors [11]. In brief, auricular chondrocytes of rabbits cultured in low-density monolayers are exposed to different concentrations of human platelet supernatant. Proliferation and matrix synthesis is stimulated by incubating auricular chondrocytes in low (0.5 or 1%) concentrations of platelet supernatant. Basic fibroblast growth factor (bFGF) has been shown to positively influence the in vitro and in vivo growth of tissue engineered pediatric human auricular chondrocytes [12]. Therefore, bFGF might be useful for quickly generating from a small piece of donor ear cartilage large amounts of chondrocytes that are similar in quality to native human elastic cartilage. Transforming growth factor- (TGF-) is secreted by malignantly transformed cells. It is also known to stimulate the rate of chondrocyte proliferation. Since aneuploidy has been found to occur in human cartilaginous tumors, at least one group has investigated the theoretical risk of malignant transformation associated with growth factor stimulation of chondrocytes [13]. In brief, tissue engineered cartilage was cultured in medium containing either bFGF or TGF- and then implanted in athymic mice. After 8 weeks, the cartilage cultured in bFGF-containing medium most resembled normal, native cartilage, whereas the cartilage cultured in TGF-containing medium had a suboptimal morphology. Flow cytometry revealed evidence of aneuploidy, which suggested that the tissue engineered chondrocytes had maintained their normal diploid state.
V. RESEARCH DIRECTIONS A. Genetic Engineering of Chondrocytes It may be possible to genetically engineer auricular chondrocytes so that they remain functional and produce antithrombogenic factors (e.g., nitric oxide, prostacycline) even after seeding onto artificial surfaces. This would result in the desired combination of tissue availability, strong surface adhesion, effective production of antithrombogenic factors at a blood–biomaterial interface, and increased possibility of fallout healing (i.e., the deposition of circulating endothelial cells or endothelial cell precursors on the seeded surface). It may also be possible to transfect auricular chondrocytes with angiogenic factors. If so, the transfected chondrocytes might be useful for revitalizing ischemic myocardium when administered directly into the heart muscle. This would result in the desired combination of a cell type that strongly adheres to the surrounding myocardium and that can induce the formation of new blood vessels (angiogenesis) in ischemic tissue. B. Transfection of Chondrocytes A viable method for transfecting hyaline chondrocytes has already been published [14]. The method described is applicable to bovine auricular chondrocytes. Auricular chondrocytes are transfectable with plasmids containing DNA that encodes proteins of interest (e.g., angiogenic
Auricular Cartilage Tissue Engineering
263
factor, antithrombogenic factor) at a DNA/transfection reagent ratio of 2:3. The procedure is as follows: 1. Plate chondrocytes onto six-well plates 24 h before transfection. 2. Gently mix 96 L of sterile modified RPMI 1640 medium with 3 L of transfection reagent FuGENE6 (Boehringer Mannheim, Mannheim, Germany). 3. Incubate solution for 10 min at room temperature. 4. Add DNA (2 g/L) to the solution. 5. Gently mix the solution and incubate for 10 min more at room temperature. 6. After incubation, add the solution to 2 mL of modified RPMI 1640 medium. 7. Place the prepared solution in one well of a six-well plate containing chondrocytes and incubate for 24 h. 8. Replace the solution with 2 mL of modified RPMI 1640 medium for an additional 24 h. 9. Determine the transfection efficiency by using a staining kit or by isolating RNA.
VI. SUMMARY Auricular cartilage used in tissue engineering has been shown to be useful and efficient in a variety of medical disciplines. Since auricular cartilage is abundantly available, easily accessible, and quickly harvested and isolated, it has great potential for use in tissue engineering in the future. In addition, the persistent need for cartilage and the potential need for transfected and nontransfected chondrocytes in many medical disciplines might lead to the development of other methods for tissue engineering auricular cartilage.
ACKNOWLEDGMENT We express our gratitude to Mr. Jude Richard for his help in editing this chapter. We also acknowledge the collaborative efforts between the following institutions. (1) Cullen Cardiovascular Surgical Research Laboratories at the Texas Heart Institute/St. Luke’s Episcopal Hospital (Drs. Rosenstrauch, Kadipasaoglu, and Frazier); (2) The University of Texas Health Science Center at Houston, Department of Internal Medicine/Cardiology (Drs. Rosenstrauch and Zoldhelyi); and (3) The Gene Therapy Laboratories at the Texas Heart Institute (Mr. Shelat and Drs. Rosenstrauch and Zoldhelyi). This work was supported by a grant from the Roderick MacDonald Research Fellowship (Dr. Rosenstrauch).
REFERENCES
1. Stem Cells: Scientific Progress and Future Research Directions. Department of Health and Human Services. June 2001. http://www.nih.gov/news/stemcell/scireport.htm. Accessed 08/28/2002. 2. Slack JM. Stem cells in epithelial tissues [review]. Science 2000; 287:1431–1433. 3. Caceci T. Connective tissues III: cartilage VM8054 veterinary histology, exercise 7. In: Gartner LP, Hiatt JL, eds. Color Atlas of Histology. 3rd edition. Philadelphia: Lippincott Williams & Wilkins, 2000.
264
Rosenstrauch et al.
4. Scott-Burden T, Bosley JP, Rosenstrauch D, Henderson KD, Clubb FJ, Eichstaedt HC, Eya K, Gregoric I, Myers T, Radovancevic B, Frazier OH. Use of autologous auricular chondrocytes for lining artificial surfaces: a feasibility study. Ann Thorac Surg 2002; 73:1528–1533. 5. de Chalain T, Phillips JH, Hinek A. Bioengineering of elastic cartilage with aggregated porcine and human auricular chondrocytes and hydrogels containing alginate, collagen, and kappa-elastin. J Biomed Mater Res 1999; 44:280–288. 6. Gentile FT, Borland KM, Nugent HM, Omstead DR. Application of autologous allogenic cell therapy. Presented at Symposium FF, Biomedical Materials. San Francisco, 1999. 7. Kim B, Yoo JJ, Atala A. Physical characteristics of tissue-engineered prosthesis for the treatment of erectile dysfunction. Presented at American Urological Association Annual Meeting, Miami, May 24–30, 2002. 8. Are´valo-Silva CA, Eavey RD, Cao Y, Vacanti M, Weng Y, Vacanti CA. Internal support of tissueengineered cartilage. Arch Otolaryngol Head Neck Surg 2000; 126:1448–1452. 9. Fuchs JR, Terada S, Ochoa ER, Vacanti JP, Fauza DO. Fetal tissue engineering: in utero tracheal augmentation in an ovine model. J Pediatr Surg 2002; 37:1000–1006. 10. Park SS, Chi DH, Lee AS, Taylor SR, Iezzoni JC. Biomechanical properties of tissue-engineered cartilage from human and rabbit chondrocytes. Otolaryngol Head Neck Surg 2002; 126:52–57. 11. Yang SY, Ahn ST, Rhie JW, Lee KY, Choi JH, Lee BJ, Oh GT. Platelet supernatant promotes proliferation of auricular chondrocytes and formation of chondrocyte mass. Ann Plast Surg 2000; 44:405–411. 12. Arevalo-Silva CA, Cao Y, Vacanti M, Weng Y, Vacanti CA, Eavey RD. Influence of growth factors on tissue-engineered pediatric elastic cartilage. Arch Otolaryngol Head Neck Surg 2000; 126:1234–1238. 13. Kamil SH, Aminuddin BS, Bonassar LJ, Silva CA, Weng Y, Woda M, Vacanti CA, Eavey RD, Vacanti MP. Tissue-engineered human auricular cartilage demonstrates euploidy by flow cytometry. Tissue Eng 2002; 8:85–92. 14. Stove J, Fiedler J, Huch K, Gunther KP, Puhl W, Brenner R. Lipofection of rabbit chondrocytes and long lasting expression of a lacZ reporter system in alginate beads. Osteoarthritis Cartilage 2002; 10:212–217. 15. Rosenstrauch D, Scott-Burden T, Frazier OH. ‘The use of tissue engineering to improve the biocompatibility of the left ventricular assist device. 20th Annual Meeting of the International Society of Heart Lung Transplantation, Osaka, Japan, April 2000.
13 Rheology of Biological Fluids and Their Substitutes Assunta Borzacchiello, Luigi Ambrosio, Paolo Netti, and Luigi Nicolais Institute of Composite and Biomedical Materials, C.N.R., and University of Naples Federico II Naples, Italy
I. INTRODUCTION The primary physiological functions of biological fluids such as synovial fluid, mucus, saliva, aqueous humour, and tears are to lubricate, absorb shock, and regulate the transport of fluids and macromolecules. From a biochemical and physical point of view, biological fluids are polymeric solutions in which the solvent is water, present in large amounts (⬎70%), and the main macromolecular components are proteins (collagens), noncollagenous glycoproteins, proteoglycans, and polysaccharides. The physiochemical properties and structure of the macromolecular components and therefore the macroscopic characteristics are tissue specific, i.e., change with the specific fluid. To perform their functions these fluids must possess specific rheological properties. For example, the rheological properties of mucus gel regulate the efficiency of mucociliary clearance, and the lubrication ability of synovial fluids is related to its rheological behavior. Therefore to understand the rheological properties of biofluids and to appreciate their functions, it’s necessary to explore their structure and composition. Nature, in fact, modulates the rheological functions of biological fluids through a fine-tuning of their structure, i.e., the biochemical properties of the macromolecular components, such as molecular weight and charge. Through the study of rheology it is possible to investigate the relationship between structure and properties. The understanding of this relationship can help to engineer macromolecular solutions able to simulate biological fluids [1,2]. The study of the rheological properties also provides a useful diagnostic tool by correlating the pathological conditions with the rheological behavior. Indeed it is known that in the case of osteoarthritic synovial fluids [3], there is a loss of physiological structure and function of synovial fluids that can lead to the complete loss of the cartilage. Rheological properties have also been used in clinics for monitoring the efficacy of treatment. In many obstructive pulmonary diseases, for instance, there is a strong modification in the rheological properties of the mucus secretion. Many therapeutic strategies in clinics today are aimed at the reduction of the mucus secretion viscosity [4,5]. In this chapter a review of the rheological properties of biological fluids will be presented and discussed in terms of the structure–function relationship of biological fluids, along with the rheological properties of biosubstitutes. 265
266
Borzacchiello et al.
II. RHEOLOGICAL PROPERTIES OF BIOLOGICAL FLUIDS From a rheological point of view biofluids show non-Newtonian behavior, that is, they have characteristic of both solids and fluids. An elastic solid follows Hooke’s law that states the power of any spring is in the same proportion with the tension thereof. This can be expressed as
σ = Gγ
(1)
where G is the elastic modulus, is the stress, and ␥ is the deformation. On the other hand, given a Newtonian fluid placed between two parallel plates, each of area A, with distance d between the plates, and the upper plate moving with relative velocity U, the force per unit area required to produce the motion is F/A, denoted by , and is proportional to the velocity gradient or shear rate, U/d, also indicated as ␥⬚. Newton’s law can be expressed as
σ=
ηU d
(2)
The constant of proportionality, , the viscosity, is a measure of resistance to flow and of the energy dissipated during the motion. Non-Newtonian fluids have both elastic and viscous components, and the stress within the fluid depends both on the actual deformation and its gradient. This characteristic confers to these fluids particular properties. To test the rheological properties of a biofluid, or more generally a viscoelastic fluid, there are two main kinds of tests: small deformation from an equilibrium condition and measurement in a steady flow. In the first kind of experiments the material is thought of as a solid and it is measured by the relationship between stress and strain history. In general the ratio between the stress and strain may depend both on time and stress. The ratio will depend only on time if the deformations are kept small in order to obtain a liner material behavior. The properties evaluated through this kind of experiment are also said to be linear viscoelastic properties. In the second kind of experiment the material is thought as a fluid, and the resistance to flow is evaluated by the aid of Eq. (2). In these experiments the dependence of viscosity on the shear rate is evaluated. A. Viscoelastic Properties One of the simplest ways to experimentally determine the viscoelastic properties of biological fluids is to subject the material to periodic oscillation. Moreover, the stresses imposed on biofluids, such as in the case of joint motion, being unsteady and time dependent, are better represented by dynamic experiments. In a small dynamic experiment the material is subjected to a sinusoidal shear strain:
γ = γ0 sin (ω t)
(3)
where ␥0 is the shear strain amplitude, is the oscillation frequency (which can be also expressed as 2 f, where f is the frequency in Hz), and t is the time. The mechanical response, expressed as shear stress of viscoelastic materials, is intermediate between an ideal pure elastic solid (in phase with the deformation) and an ideal pure viscous fluid (90⬚ out of phase with the deformation) and therefore is out of phase with respect to the imposed deformation as expressed by
σ = G * (ω ) γ0 sin (ω t + δ )
(4)
Rheology of Biological Fluids and Their Substitutes
267
and consequently
σ = G * (ω ) γ0 sin (ω t) cos (δ ) + G "(ω ) γ0 cos (ω t)sin (δ )
(5)
and if it is defined
G '(ω ) = G * cos δ G "(ω ) = G * sin δ
(6)
σ = G '(ω ) γ0 sin (ω t ) + G "(ω ) γ0 cos (ω t)
(7)
then
where G′() is the shear storage modulus and G⬙() is the shear loss modulus. G′ gives information about the elasticity or the energy stored in the material during deformation, whereas G⬙ describes the viscous character or the energy dissipated as heat. The combined viscous and elastic behavior is given by the absolute value of complex shear modulus G*: G * (ω ) = G ' 2 + G "2
(8)
or by the absolute value of complex viscosity, *, defined as
η*(ω ) =
G ' 2+ G " 2 ω
(9)
which is usually compared with the steady shear viscosity in order to evaluate the effect of large deformations and shear rates on the material structure. The ratio between the viscous modulus and the elastic modulus is expressed by the loss tangent: tan δ =
G" G'
(10)
where ␦ is the phase angle, The loss tangent is a measure of the ratio of energy lost to energy stored in the cyclic deformation [6]. To determine the region of linearity, preliminary strain sweep tests at a fixed oscillation frequency are performed. Specifically, the dynamic moduli are monitored while logarithmically varying the strain amplitude ␥0. According to their viscoelastic properties, most biological fluids can be divided into three main categories: entangled networks, weak gels, and strong gels. The entangled network behavior is observed when the macromolecules entangle with each other, forming a continuous temporary network. The entangled condition depends upon the molecular weight and/or the concentration. When these variables are sufficiently high the molecules start to entangle with each other forming a network structure. At low frequency, the molecular chains can release stress by disentanglement and molecular rearrangement during the period of oscillation, and hence the solution presents viscous behavior (G⬙⬎G′). While at high frequency molecular chains cannot disentangle during the short period of oscillation, and therefore they behave as a temporary crosslinked network and elastic behavior (G′⬎G⬙) is observed. The transition between the viscous and the elastic behavior, indicated by the crossing of the G′ and G⬙ curves as a function of frequency (mechanical spectrum), occurs at a given value of the frequency (f*) corresponding to the intrinsic rate of disentanglement of polymer chains. Since the disentanglement rate depends upon the mobility
268
Borzacchiello et al.
of the polymer chains, the crossover frequency (f*) will depend on both molecular weight and concentration of the polymer solution. In weak gel system, G′ and G⬙ are almost constant with frequency and are parallel to each other within all frequency ranges. Similar behavior is observed for many biological gels (e.g., collagen, agarose) [7,8]. For these systems the overall rheological response is due to the contributions of physical crosslinks among the macromolecules, such as electrostatic interactions, van der Waal bonds, hydrogen bonds, hydrophobic interactions, and to the entanglements. The intermolecular physical crosslinks bring about a reduction of intrinsic mobility of the polymer chains leading to a decreases in f*. Furthermore, since the molecular weight of pendant chains is widely dispersed, there will not be a unique value of disentanglement rate but rather a distribution. This would explain the weak dependence of moduli upon frequency. When chemical crosslinks (covalent bonds) are present among the chains, a strong gel behavior is observed. The typical strong gel spectrum, over the frequency range 10ⳮ2 –102Hz, consists of two nearly horizontal straight lines. G′ is typically one or two orders of magnitude greater than G⬙, and both may show some slight increase at higher frequency. B. Flow Properties For Newtonian fluids the viscosity is constant with the shear rate, but most biofluids show a shear thinning behavior, that is, a reduction of the viscosity with increasing rate of shear in steady flow. The shear viscosity decreases with the shear rate in an S-shaped fashion. At low shear rate the viscosity is almost constant, decreasing slightly with the shear rate (Newtonian plateau). At greater shear rate the viscosity sharply drops with the shear rate (thinning). At higher shear rate, a second plateau is generally observed. The shear thinning behavior can be explained by the dynamics of entanglements. At low shear, the rate of molecular entanglements (i.e., the time required to a given chain to relax one entanglement and to form the next) is higher than the rate of shear, and therefore the average entanglement density is constant, resulting in almost constant viscosity. As the shear rate increases, the rate of entanglement disruption becomes predominant, leading to the thinning.
III. SYNOVIAL FLUID The cavities of the joints that allow various degrees of relative motion of the bone are called diarthrodial or synovial joints. They are formed by a surface of cartilage, that is a dense connective tissue, and by the synovium, a metabolically active tissue, and are filled by the synovial fluid (SF) (Fig. 1). The synovial fluid is secreted into the joint cavity by the synovium. Normally it is a clear and colorless or slightly yellowish fluid, slightly alkaline, with a pH ranging between 7.3 and 7.64. The water content ranges between 960 and 988 g/kg, the dry residue is about 34 g/kg. In normal human knee joint the SF volume is about 2 mL, but due to knee pathologies can be 10–30 mL. Biochemically, synovial fluid is a dialysate of blood plasma, similar to the plasma in the electrolytes and low molecular weight organic molecules, but different in protein composition (one-third that of plasma) and the presence of macromolecules such as glycosaminoglycan. The LS total protein content is about 12 mg/mL. The proteins contained in SL and their relative amounts are reported in Table 1. The glycosaminoglycans are present as chondroitin-4-sulfate (2%), the remaining 98% is hyaluronic acid (HA). The hyaluronic acid is associated with the protein forming a complex known as hyaluronic acid protein complex (HAP). The hyaluronic acid or hyaluronan is linear
Rheology of Biological Fluids and Their Substitutes
269
Figure 1 Schematic of human knee joint.
polysaccharide composed of repeating disaccharide units of glucoronic acid and N-acetyl glucosamine linked. The average molecular weight of the HAP complex is about 2 million [9–14]. The synovial fluid regulates the transport of nutrients to articular tissue and mechanically assists the joint by acting as a lubricant and wear inhibitor and by protecting the articular cartilage and soft tissue surfaces from mechanical stress during joint function [15–17]. Synovial fluid provides a cushion for the synovial lining by filling the crevices where synovial tissue cannot
Table 1 Proteins Contained in Synovial Fluid Protein Albumin ␣1-Antripsin Ceruloplasmin Aptoglobin ␣2 – Macroglobulin Lactoferrin IgG IgA IgM IgE
Concentration (g/L) 8 0.78 43 ⫻ 10⫺3 0.09 0.31 0.44 ⫻ 10⫺3 2.62 0.84 0.14 14 ⫻ 10⫺6
270
Borzacchiello et al.
reach. Synovial lining, in fact, cannot fill the space between cartilage surfaces perfectly in all positions. The mechanical function of the SF can be ascribed to its rheological properties, in particular, to its viscoelastic properties. The viscoelasticity of SF was demonstrated by Ogston and Stainer in 1953 [9]. They placed a drop of fluid on an optically flat surface and a convex lens was set on top. The lens was pressed down and the interface pattern known as Newton’s ring was used to measure the distance between the lens and the optical flat. When the experiment was done with water, the lens was easily pressed into contact with the optical flat. With SF, the lens stopped moving while it was still some distance above the optical flat. The distance depended on the load pressing it down. Moreover, when the load was removed, the lens moved slightly up again due to elastic recoil. These experiments elucidated the role of the viscoelasticity, and in particular of the elastic component of SF, in regulating the biomechanics of the joint. The mechanical response of this material may be largely ascribed to the physical characteristics of hyaluronic acid (concentration, molecular weight, and molecular weight distribution) and to the physical and noncovalent interactions within the molecule itself and with other molecules, especially proteins and ions present in solution (HAP) [18]. In solution, hyaluronic acid behaves as an expanded random coil molecule and occupies a large hydrodynamic volume because of its chain backbone stiffness, the presence of intramolecular hydrogen bonding, and the electrostatic repulsion between its charged groups [19–21]. At physiological concentration and pH, HA molecules overlap and interact through physical ‘‘entanglements’’ or temporary crosslinks and interactions with ions and other link proteins due to the polyelectrolyte nature. These interactions determine the formation of the transient network structure that is responsible for the viscoelasticity of synovial fluid. The synovial fluid occupies the narrow channel between the soft tissue of the joint, and it is sandwiched between the two cartilage surfaces. When the joint is flexed slowly, corresponding to the low deformation rates (rest and walk), the synovial fluid moves like a viscous liquid in the channel and behaves as a lubricant. On the other hand, under movement with high deformation rate (run) the synovial fluid stays in the channels and its elasticity prevails determining a rapid elastic response to mechanical stress and preventing articular tissues from compressive damage. In pathological joints, the synovial fluid does not have such elastic properties, but it is always viscous and the synovial fluid might not stay in the channel nor store the energy mechanically against quick movement. A. Synovial Fluid Viscoelastic Properties The SF viscoelastic behavior is well represented by the elastic and viscous dynamic moduli as a function of oscillation frequency. The qualitative behavior of G′ and G⬙ for typical normal and pathological synovial fluids is represented in Fig. 2 (derived from Balazs [15]). Normal SF presents prevalent viscous behavior at low frequency (G⬙ ⬎ G′) and prevalent elastic behavior at high frequency (G′ ⬎ G⬙), the limit between the two regions being represented by frequency c, which typically moves to higher frequencies as solutions become more dilute and structure disappears [22]. The degeneration of synovial fluid is evident from its rheological properties. Indeed, in general, the SF from a young subject has lower c and higher absolute values of G′ and of G⬙ (with the exception of a small range at higher frequencies), over the entire frequency range, than those from an old subject. This is due to the fact that SF shows a reduction in HA concentration or molecular weight with age, even if the overall functionality of SF is substantially not affected. Moreover pathological SF presents reduced absolute values of G′ and G⬙ and the loss modulus is higher than the storage modulus over the whole frequency range under investigation, indicating that there are few interactions among HA molecules and the structure disappears, the rheological
Rheology of Biological Fluids and Their Substitutes
271
Figure 2 Qualitative dynamic modulus dependence upon frequency for synovial fluids of young normal, old normal, and osteoarthritic knee joints. (Adapted from Ref. 15.)
properties being now essentially controlled by intramolecular conformational changes affecting hydrodynamic molecular size. In this way, the mechanical response of SF cannot adequately sustain the compression and tangential forces arising in everyday life, allowing cartilage–cartilage contact and increasing wear of the surfaces. The reduction of SF viscoelasticity is due to the decrease of HA concentration and its degree of polymerization in the course of joint pathologies; this may be due to superoxide ion action and reduced synthesis by synoviocytes [3,23,24]. Different pathologies lead to different SF viscoelastic properties, and the difference in viscoelasticity has been used to clearly identify the different disease states [25] B. Synovial Fluid Flow Properties Synovial fluid shows highly non-Newtonian flow properties: the resistance to shearing of SF decreases as the shear rate increases; as synovial fluid is sheared, a force perpendicular to the flow is generated and energy may be elastically stored by deforming the HAP complex. In detail, both normal and pathological SF show shear thinning (pseudoplastic) behavior, with a pseudo-Newtonian plateau at low shear rates, characterized by the zero-shear viscosity (o) and a subsequent decrease that is well described by the power law model: η = K γ n−1
(11)
where k and n are empirical parameters and k is generically defined as consistency. In general, pathologies of SF is evidenced by a reduction of viscosity. Indeed pathological synovial fluids have zero-shear viscosity o within the range 0.01–1 Pa s, significantly reduced compared to that of normal SF, which is 10 Pa s. Analogously the consistency K of the power-law model varies from 10ⳮ2 to 10ⳮ1 Pa sn (normal value ⬃ 1 Pa sn) and the exponent n in the range 0.5 to ⳮ0.1 (normal value ⬃ ⳮ0.5) [24,26]. The decrease of the SF viscosity, due to aging or pathological conditions, is in agreement with the analogous reduction of the viscoelastic function.
272
Borzacchiello et al.
Joint diseases as well as aging effects cause the decrease of hyaluronic acid molecular weight and/or concentration, and a consequent reduction of the inter and intramolecular interactions resulting in a lower viscosity.
IV. VITREOUS BODY The vitreous body is a transparent mass that fills the posterior cavity of the eye in vertebrates. It is composed almost completely of water (97%) and three main components: collagens, proteoglycans (PG), and hyaluronic acid (HA). The gel-like structure is formed by an ordered network of fine collagen fibrils (with a diameter of about 10 nm) immersed in a viscoelastic matrix composed mainly of highly hydrated HA macromolecules [27,28]. The collagen fibrils are tied together by PG bridges and run in approximately parallel bundles. The PG bridges can interact through their glycosaminoglycan (GAG) chains with HA matrix via noncovalent bonds [29]. Gel vitreous composition changes with age and species, because of variations in collagen content and in HA concentration, molecular weight, and Molecular weight distribution. In fact there is a significant variation among different species in collagen fibril concentration and hence in their network density, because all the species have about the same fibril thickness. Human gel vitreous has a low-density network of collagen fibrils, while rabbit gel vitreous (a model widely used in ophthalmic research) has a very dense collagen network. The molecular weight of hyaluronic acid varies with species and age. It is reported that in adult human HA molecular weight is about 3 ⳯ 106 Da; in rabbit vitreous it is about 2 ⳯ 106; in newborn calf vitreous HA has a an average molecular weight of about 3 ⳯ 106, while this value drops to about 5 ⳯ 105 in old bovine [30]. The particular viscoelastic properties of the vitreous body enable it to resist sudden compression shocks, offering the best protection for the retina [31]. It is generally accepted that the viscoelastic properties of the vitreous body are mainly controlled by hyaluronan. It is, indeed, known that certain important ophthalmic diseases, such as diabetic proliferative retinopathy and retinal detachment, are strictly related to alterations in hyaluronan [32,33]. A. Vitreous Body Viscoelastic Properties The results of viscoelastic analyses indicate that vitreous body behaves as a solidlike material and its rheological behavior is typical of a weak gel [34–36]. The dependence of dynamic moduli of different animal vitreous (pig, sheep, and rabbit ) on the frequency is shown in Fig. 3. The dynamic elastic modulus G′ is always higher than the dynamic viscous modulus G⬙ over the whole frequency range investigated (0.05–10 Hz), and the moduli are almost parallel to each other. Neither modulus varies significantly with frequency. Moreover tan ␦ for the different species ranged from 0.2 to 0.5 at 1 Hz [36]. The moduli are species dependent, specifically the elastic one ranges from 3.7 Pa (rabbit) to 130 Pa (male goat). Male goat vitreous body has viscoelastic parameters one order of magnitude higher than those of female goat (Fig. 4) and of the other animal species. Rabbit, pig, and sheep vitreous properties are not influenced by sex, while the goat ones are sensitive. Since the most important biological function of the vitreous body is to serve as a shock absorbing transparent medium these differences can be understood considering that the vitreous of different animal species is subjected to different stress magnitudes [37]. It is thought that the main component controlling the rheological properties of vitreous body is HA. This is because HA is highly hydrophilic and has a large excluded volume which results in an increased friction coefficient and thus an increased viscosity and bulk modulus of the vitreous. Moreover, high molecular weight HA chains, as well as PGs, can entangle together
Rheology of Biological Fluids and Their Substitutes
273
Figure 3 Mechanical spectra of animal (pig, rabbit, sheep) vitreous body.
to form macromolecular aggregates with gel-like behavior [38,39]. Nevertheless, the importance of the collagen network on the mechanical properties of connective tissues such as the vitreous has long been emphasized in the literature [40]. From the viscoelastic properties on native tissues, it is not possible to separate the contribution of HA from collagen networks since both had the same qualitative behaviour in vitro. The viscoelastic behavior of vitreous seems to depend neither only on hyaluronic acid nor on collagen gel, but rather on a synergistic interaction [41].
Figure 4 Mechanical spectra of vitreous body of male and female adult goat.
274
Borzacchiello et al.
B. Vitreous Body Flow Properties The steady shear viscosity and the complex viscosity * of animal vitreous bodies decrease linearly with the imposed shear rate showing a typical shear thinning behavior. Values of * are one to two orders of magnitude higher than those of , indicating that vitreous structure is very sensitive to deformations and that simple viscosity is not a correct measure of energy dissipation for this system [36].
V. MUCUS In the human body, mucus layers cover many epithelia such as the respiratory, the gastrointestinal, and the genital tracts; the eyes and the urinary bladder to protect them from environmental insults and from the effect of enzymes or other chemical agents. Functions that have been attributed to the mucus are listed in Table 2. The main constituents of mucus, irrespective of its origin, are water, glycoproteins, lipids, sloughed epithelial cells, electrolytes, and bacteria. Overall, mucus contains water (⬃95%), glycoproteins and proteins (0.5–5 %), mineral salt (0.5–1%), and free proteins (1%) [42]. The main constituents of mucus are mucins that are a family of highly modified, unusually large glycoproteins (⬃7000 kD, ⬍6 m in length) composed of approximately 10–30 % by weight of peptide core linked via O-glycosidic bonds to oligosaccharide chains that constitute the remaining 70–80% of their weight. It appears that mucins macromolecules are well adapted to binding and trapping inhaled particles for clearance from the lung, at least in part because of the extraordinary diversity of their carbohydrate side chains. Because they provide in effect a combinatorial library of carbohydrate sequences, mucins can bind to virtually all particles that land on airways epithelia and can thus clear them from the lung. Mucins form viscous gels due to their high molecular weights and their ability to form intermolecular interaction and in particular disulfide bridges [43]. The most important and extensively studied physicochemical propert of mucus is its viscoelasticity. The efficiency of mucociliary clearance depends strictly on the viscoelastic properties of mucous gel [44]. In fact it has been reported that the basal rate of particle clearance depends strictly on viscoelastic properties of the mucus. As a consequence of inflammatory processes, mucostasis is a characteristic shared by several airways diseases, including chronic bronchitis, cystic fibrosis, and asthma [45,46]. Is widely reported in the literature that in these diseases there is an increase in the viscoelastic properties of mucus secretion [47,48]. In cystic fibrosis, for instance, the mucus is denser and more highly glycosylated than normal mucus, thus resulting in a more viscoelastic one. The effect of mucolytic agent is to reduce mucus viscoelasticity by disrupting its gel structure [49–51]. Agents such urea, sodium dodecyl sulfate can break down the hydrogen
Table 2 Functions of Mucus in the Human Body Site of secretion
Functions
Respiratory tract
Clearance of mucosal insults, water balance, ion transport, and regulation Clearance of cellular debris, water balance Lubrication, water balance, antimicrobial Cytoprotection, lubrication, water balance, diffusion barrier Regulation of sperm transport, sperm reservoir, energy source, antimicrobial, water balance
Middle ear Salivary glands Gastrointestinal tract Cervical tract
Rheology of Biological Fluids and Their Substitutes
275
bonding, thus disrupting the self-association of the mucin macromolecules and then the mucus gel structure [52]. Cysteine-containing compounds reduce the viscoelasticity of the mucus by reducing the molecular weight due to their ability to disrupt the disulfite linkages [53]. The rheological analysis of mucus secretion after mucolytic therapy is an accepted method to evaluate the effectiveness of such treatments [4,5]. A. Mucus Viscoelastic Properties The viscoelasticity of mucus is its most important and extensively studied property. As already mentioned, mucus is a non-Newtonian fluid, and its rheological properties change with the magnitude and frequency of the applied force. In this context it is important to study the frequency dependence of mucus, since it is important in terms of mucus transport, either by ciliary movement that occurs at low frequency deformation, or by cough that occurs at high frequency deformation. Lutz et al. [1] have studied the viscoelastic properties of tracheal dog mucus. They show that the normal mucus of the respiratory tract is heterogeneous and shows considerable variation in dry weight/ volume and pH from sample to sample from the same animal from day to day, as well as from animal to animal. The mucus shows a solidlike behavior with the storage modulus higher than the loss modulus over the entire frequency range (10ⳮ1 –102 rad/s). At the lowest pH (5.4) the G′ curve is almost flat, while increasing the pH the curve results in a more marked frequency dependence. Lowering the pH there is an increase in both moduli, even if the effect of pH on G⬙ is smaller. These results suggest that mucus behaves as an entanglement network or weak gel depending on pH. Ionic strength changes can also bring about significant changes of G′ and G⬙. The rheological properties of mucus strictly depend upon the molecular configuration of mucin, which is strictly related to pH and ionic strength. Tam and Verdugo [54] have proposed that ionic strength and pH modulate the spontaneous swelling of mucins, and consequently of mucus, in a manner according to the Donnan equilibrium properties. Several studies of mucus dynamic properties have indicated that the rheological parameters more significantly associated with mucus clearance are not only viscosity and elasticity per se, but also the ratio between viscosity and elasticity (tan ␦) and the dynamic complex moduli [55–57]. The results of these studies agree in showing that mucus transportability is inversely related to the elastic modulus of the sample. The importance of tan ␦ in contributing the mucus clearance transport is controversial, since there is experimental evidence that mucus relaxation times are quite long (up to 30 s) [44]. In this situation, mucus would be seen by cilia as an elastic structure capable of accepting efficient energy transfer from cilia and relaxing very little between successive beats [44]. If this description were valid, mucus relaxation time would be of minor importance in determining mucus transport, and the relevant rheological parameter would be elasticity. However, studies that used mucus samples collected in situations of chronic airway inflammation [55] or when the range of tan ␦ is increased by the use of mucolytic [57] suggest that rheological optimum is dependent not only on mucus overall moduli, but also on the relationship between viscosity and elasticity (tan ␦). A sample that is too rigid (i.e., with a large elastic modulus) may impede the proper penetration of cilia tips in the mucus gel during the effective stroke, while a gel with a low recoil factor would not provide an adequate transmission of the mechanical input of cilia to the mucus blanket. B. Mucus Flow Properties Mucus is rarely tested in the steady flow condition because one fears that structure of mucus would be different in steady flow as compared with in vivo. There are same data obtained [58] by subjecting the mucus sample to a uniform acceleration, followed by a deceleration. In this kind of test, the stress shows a large hysteresis loop. There exists a yield point on the loading
276
Borzacchiello et al.
curve at which the mucus structure apparently breaks. After the completion of the strain cycle there is a residual stress left in the sample.
VI. EXAMPLES OF BIOLOGICAL FLUID SUBSTITUTES To augment or substitute the biological fluids, research has been focused on natural, semisynthetic, and synthetic polymers solutions, since biological fluids are mainly constituted by water. Among the natural polymers, hyaluronic acid has been widely investigated to be used as a biofluid substitute. Hyaluronic acid (HA) is a naturally occurring, biocompatible, and biodegradable linear polysaccharide composed of repeating disaccharide units of glucoronic acid and Nacetyl glucosamine linked by -(1–3) and -(1–4) glycosidic bonds. HA is present in all soft tissues of higher organisms and in particular concentrations in the synovial fluids, in the extracellular matrix of connective tissues, and in the vitreous humour of the eyes [38]. Commercially available HA is obtained from different sources, mainly extracted from umbilical cord, rooster comb, synovial fluid, or vitreous humor. Recently there are also biotechnological production of HA from Streptococcus. In aqueous solution, a HA molecule is stiffened by inter-residue hydrogen bonds. Consequently, a hyaluronan molecule behaves as an expanded random coil occupying a large hydrodynamic volume overlapping with other HA molecules at concentrations above 1 mg/cm3 (MW about 1–2 MDa) [59]. Extensive investigations on rheological properties of HA solutions have been made by several workers [19,20,38,60,61]. It has been reported that HA solution can behave as viscous liquid at low molecular weight and concentration, whereas it can behave as an entanglement network and weak gel at higher molecular weight and concentration due to chain entanglement. Increasing the molecular weight is more effective than increasing concentration in promoting entanglement of molecular chains of HA [38]. In Fig. 5 is reported the mechanical spectra of high molecular weight HA (1200 kDa) solutions
Figure 5 Mechanical spectra of high molecular weight hyaluronic acid solutions at different concentrations.
Rheology of Biological Fluids and Their Substitutes
277
at different concentrations. The behavior of these solution is typical of an entangled network with a crossover between the curve of G′ and G⬙. The crossover frequency is shifted toward the lower frequency, increasing the concentration. Concerning the flow properties, low molecular weight hyaluronic acid (150 KDa) exhibits essentially Newtonian characteristics with a viscosity constant with the shear rate range. Pseudoplastic behavior (shear thinning) is observed for high molecular weight HA (1.2 MDa) [38]. The typical flow curves are reported in Fig. 6. Hyaluronic acid solutions are widely used as an aid in viscosurgery (ophthalmological surgery) for they have adequate rheological properties. In cataract surgery, for example, the role of HA solutions is to facilitate the procedures and to protect the corneal endothelium [62]. HA solutions are also extensively used in viscosupplementation. Intra-articular hyaluronic acid injection has benefited patients with rheumatoid or osteoarthritis partly by supplementing the lubricating characteristics of synovial fluids [63,64]. Increase in synovial fluid viscosity and hyaluronic acid concentration or molecular weight after intra-articular injection of exogenous hyaluronic acid 1% (5 to 8 ⳯ 105 Da) have been beneficial in some patients with osteoarthritis or rheumatoid arthritis [26]. There is experimental evidence in humans, in vitro [65] and in vivo [26], demonstrating that injected hyaluronic acid (⬎5 ⳯ 105 Da) may stimulate synoviocyte production of endogenous HA. Balazs and Denlinger have hypothesis that viscosupplementation with exogenous HA restores the normal physiological homeostasis of joint cells [66]. The clinical indications for HA are limited by the short residence time [59]. Therefore, a mounting research effort has been devoted to explore ways of chemically modifying HA by crosslinking [67] or coupling reactions such as esterification reactions [59]. These derivatives show different physical-chemical properties, such as longer residential times, compared with native HA but maintained the inherent biocompatibility of HA; for this reason they are currently used in viscosurgery and viscosupplementation. Some of them have also been exploited as vitreous substitutes [41]. Other polysaccharides [68–70] such as dextran, alginic acid, chondroitin sulfate, methyl cellulose and carboxymethyl cellulose, and hydroxypropyl methyl cellulose have been used as biofluid substitutes, in particular as potential vitreous substitute. The problem with using those materials as vitreal
Figure 6 Flow curves of high molecular weight hyaluronic acid solutions at different concentrations.
278
Borzacchiello et al.
substitute is connected to their opacity. Among the synthetic polymers, aqueous solutions of polyvinyl pyrrolidone (PVP) has been used extensively as a blood substitute. PVP has also been tested as a potential vitreous substitute [71,72]. In spite of numerous attempts using synthetic polymers such as polyacrylamide [73], polyglyceryl methacrylate [74], and polyvinyl alcohol [75], a permanent vitreal substitute that can fully replace the vitreous body functions has not yet been found.
VII. CONCLUSIONS In this chapter the rheological properties of biological fluids in terms of the structure–function relation have been presented. Biological fluids can perform their functions because they possess viscoelastic properties. For this reason, in order to design biomaterials able to help or to substitute the functions of biological fluids it is necessary to understand their viscoelastic properties. The viscoelastic properties of biological fluid substitutes have also been discussed.
REFERENCES
1. Kocevar-Nared J, Kristal J, Smid-Kobar J. Comparative rheological investigation of crude gastric mucin and natural gastric mucus. Biomaterials 1997; 18:677–681. 2. Lutz RJ, Litt M, Chakrin LW. Physical-chemical factors in mucus rheology. In: Gabelnick HL, Litt M, eds. Rheology of Biological Systems. Springfield. IL: Charles C. Thomas, 1973:119–157. 3. Gardner DL. The nature and causes of osteoarthrosis. Br Med J (Clin Res Ed.) 1983; 286(6363): 418–424. 4. Moretti M, Lopez-Vidriero MT, Pavia D, Clarke SW. Relationship between bronchial reversibility and tracheobronchial clearance in patients with chronic bronchitis. Thorax 1997; 52:176–180. 5. Rhee C-S, Majima Y, Cho J-S, Arima S, Min Y-G, Sakura Y. Effects of mucokinetic drugs on rheological properties of reconstituted human nasal mucus. Otolaryngology—Head and Surgery 1999; 125:101–105. 6. Ferry JD. Viscoelastic Properties of Polymers. New York: Wiley, 1970. 7. Ross-Murphy SB. In Polymer Gels. Fundamentals and Biomedical Applications. De Rossi D, Kajiwara K, Osada Y, Yamauchi A, eds. New York: Plenum Press, 1991:21. 8. Clark A, Ross-Murphy SB. Adv. Polym Sci 1987; 83:61. 9. Ogston AG, Stainer JE. Further observation on the preparation and composition of hyaluronic acid complex of ox synovial fluid. Biochem. J 1952; 52:149–156. 10. Sundblad L. Studies of hyaluronic acid in synovial fluids. Acta Soc. Med. Upsal 1953; 58:113–238. 11. Sundbland L. Glycosaminoglycans and glycoproteins in synovial fluid. In. Balazs EA, Jeanloz RW(), eds. The Amino Sugars. Vol. 2A. New York: Academic Press, 1965:229–250. 12. Balazs EA, Watson D, Duff IF, Roseman S. Hyluronic acid in synovial fluid: I. Molecular parameters of hyaluronic acid in normal and arthritic human fluids. Arthritis Rheum 1967; 10:357–376. 13. Sandoson J. Human synovial fluid: detection of a new component. Science 1967; 155:839–841. 14. Swann DA. Macromolecules of synovial fluid, in. Sokoloff L(), ed. The Joints and Synovial Fluids. Vol. 1. New York: Academic Press, 1978:407–435. 15. Balazs EA. Disorders of the knee, 63. Helfeet J, Helfeet J, eds. Philadelphia: T.B. Lippincoat, 1974. 16. McCutchen CW. Joint lubrication. Bulletin of the Hospital for Joint Diseases. Orthopaedic Institute 1983; 43(2):118–29. 17. Hou JS, Mow VC, Lai WM, Holmes MH. An analysis of the squeeze-film lubrication mechanism for articular cartilage. J Biomechanics 1992; 25(3):247–259. 18. Atkins EDT, Phelps CF, Sheehan JK. Conformation of the mucopolysaccharides Hyaluronates. Biochem. J 1972; 128(5):1255–63.
Rheology of Biological Fluids and Their Substitutes
279
19. Morris ER, Rees DA, Welsh EJ. Conformation and dynamic interactions in hyaluronate solutions. J Molec Biol 1980; 138(2):383–400. 20. Bothner H, Wik O. Rheology of hyaluronate. Acta Oto-Laryngologica. (Suppl) 1987; 442:25–30. 21. Graessley WW. The entanglement concept in polymer rheology. Adv Polym Sci 1974; 16:179. 22. Rochefort WE, Middleman S. Rheology of xanthan gum: salt, temperature, and strain effects in oscillatory and steady shear experiments. J. Rheol 1987; 31(4):337–369. 23. Seller PC, Dowson D, Wright V. Rheology of synovial fluid. Rheol. Acta 1971; 10(1):2–7. 24. Altmann S, Zeidler H, Gaffga R, Kulicke WM. Rheology of pathological joint fluids. II. Structural rheology of hyaluronic acid. Rheol. Acta 1980; 19(5):642–659. 25. Gomez GE, Thurston GB. Comparisons of the oscillatory shear viscoelasticity and composition of pathological synovial fluids. Biorheology 1993; 30:409–427. 26. Mensitieri M, Ambrosio L, Iannace S, Perbellini A, Nicolais L. Viscoelastic evaluation of different knee osteoarthritis therapies. J Mater Sci: Mater Med 1993; 6:130–137. 27. Scott JE. The chemical morphology of the vitreous. Eye 1992; 6:553–55. 28. Balazs EA, Denlinger JL. In The Eye, Vol. Ia: Vegetative Physiology and Biochemistry. Davson H, ed. 3d ed. New York: Academic Press:1984. 29. Bishop P. The biochemical structure of mammalian vitreous. Eye 1996; 10:664–670. 30. Laurent UBG, Granath KA. The Molecular Weight of Hyaluronate in the aqueous humour and vitreous body of rabbit and cattle eyes. Exp. Eye Res 1983; 36:481–492. 31. Bettelheim FA, Wang TJY. Dynamic viscoelastic properties of bovine vitreous. Exp. Eye Res 1976; 23:435–411. 32. Sebag J, Balazs EA. Human vitreous fibres and vitreoretinal disease. Trans. Ophthalmol. Soc. U.K 1985; 104:123–127. 33. Larsson L, Osterlin S. Posterior vitreous detachment: a combined clinical and physicochemical study. Graefes Arch. Clin. Exp. Ophthalmol 1985; 223:92. 34. Tokita M, Fujiya Y, Hikichi K. Dynamic viscoelasticity of bovine vitreous body. Biorheology 1984; 21:751–756. 35. Lee B, Litt M, Buchsbaum G. Rheology of vitreous body. Part I: Viscoelasticity of human vitreous. Biorheology 1992; 29:522–533. 36. Mensitieri M, Ambrosio L, Nicolais L, Balzano L, Lepore D. The rheological behaviour of animal vitreous and its comparison with vitreal substitutes. J Mater Sci: Mater Med 1994; 5:743–747. 37. Borzacchiello A, Netti PA, Ambrosio L, Nicolais L. Viscoelastic properties of vitreous body: role of hyaluronic acid, Cambridge, UK, 2000. 38. Ambrosio L, Borzacchiello A, Netti PA, Nicolais L. Rheological study on hyaluronic acid and its derivatives solutions. J. Macromolec Sci Pure Appl Chem 1999; A36(7&8):991–1000. 39. Soby L, Jamieson AM, Blackwell J, Choi HU, Rosenberg LC. Viscoelastic and rheological properties of concentrated solutions of proteoglycan subunit and proteoglycan aggegate. Biopolymers 1990; 29: 1587–1592. 40. Scott JE. Proteoglycan–fibrillar collagen interactions. Biochem. J 1988; 252:313–323. 41. Borzacchiello A, Netti PA, Ambrosio L, Nicolais L. Hyaluronic acid derivatives mimic the rheological properties of vitreous body. In New Frontiers in Medical Sciences: Redefining Hyaluronan(. Abatangelo G, Weigel PH), eds: Elsevier Science, 2000:195–202. 42. Rathbone MJ, Hadgraft J. Absorption of drugs from the human oral cavity. Int. J. Pharm 1991; 74: 9–24. 43. Verdugo P, Tam PY, Butler J. Conformational structure of respiratory mucus studied by laser correlation spectroscopy. Biorheology 1983; 20:223–230. 44. Sleigh MA, Blake JR, Liron N. The propulsion of mucus by cilia. Am. Rev. Respir. Dis 1998; 137: 726–741. 45. Knowles MR, Boucher RC. Mucus clearance as a primary innate defense mechanism for mammalian airways. J Clin Invest 2002; 109:571–577. 46. Sade` J, Eliezer N, Silberberg A, Nevo AC. The role of mucus in transport by cilia. Am. Rev. Respir. Dis 1970; 102:48–52. 47. Majima Y, Sakakura Y, Hattore M, Hirata K. Rheologic properties of nasal mucus from patients with chronic sinusitis. Am J Rhinol 1993; 7:217–221.
280
Borzacchiello et al.
48. Glass GBJ, Slomiany GL. ‘Derangements of biosynthesis, production, and secretion of mucus in gastrointestinal injury and disease’. In:. Elstein M, Parke DV(), eds. Mucus in Health and Disease. New York: Plenum Press, 1977:311–347. 49. Ziment I. Mucokinetic agents. In Respiratory Pharmacology and Therapeutics. Philadelphia. Pa: WB Saunders, 1978:60–104. 50. Marriot C. Modification of the rheological properties of mucus by drugs. Adv Exp Med Biol 1982; 144:75–84. 51. Will P, Wodehouse T, Corkery K, Mallon K, Wilson R, Cole PJ. Short-term recombinant human DNase in bronchiectasis. Am J Respir Crit Care Med 1996; 154:413–417. 52. Khan MA, Wolf DP, Litt M. Effect of mucolytic agents on the rheological properties of tracheal mucus. Biochim. Biophys. Acta 1976; 444:369–373. 53. Currie DC, Greenstone M, Pavia D, Agnew JE, Pellow P, Clarke SW, Hendry WF, Cole PJ. Efficacy of mucoregulatory agents in Young’s syndrome. Thorax 1988; 43:480–481. 54. Tam PY, Verdugo P. Control of mucus hydration as a donnen equilibrium process. Nature 1981; 292:340–342. 55. Puchelle E, Girard F, Zahm JM. Rhe`ologie des secretions bronchiques et transport muco-ciliaire. Bull. Eur. Physiopathol. Respir 1976; 12:771–779. 56. King M. Relationship between mucus viscoelasticity and ciliary transport in guaran gel/frog palate model system. Biorheology 1980; 17:249–254. 57. Lorenzi G, Bohm GM, Guiramaraes ET, Costa Vaz MA, King M, Saldiva PHN. Correlation between rheologic properties and in vitro ciliary transport of rat nasal mucus. Biorheology 1992; 29:433–440. 58. Davis S. In Rheology of Biological Systems. Gabelnick HL, Litt M, eds. Springfield. IL: Charles C. Thomas:158–194. 59. Borzacchiello A, Ambrosio L. Network formation of low molecular weight hyaluronic acid derivatives. J Biomater Sci Polym Ed 2001; 12(3):307–316. 60. Gibbs DA, Merrill EW, Smith KA. Rheology of hyaluronic acid. Biopolymers 1968; 6:777–791. 61. De Smedt SC, Dekeyser SC, Ribititsch V, Lauwers PA, Demeester J. Viscoelastic and transient network properties of hyaluronic acid as a function of the concentration. Biorheolog 1993; 30:31–41. 62. Goa KL, Benfield P. Drugs 1994; 47:536. 63. Iwata H. Pharmacologic and clinical aspects of intraarticular injection of hyaluronate. Clin Orthop Rel Res 1993; 289:285–291. 64. Strachan RK, Smith P, Gardner DL. Hyaluronate in rheumatology and orthopaedics: is there a role?. Ann Rheumatic Dis 1990; 49:949–952. 65. Smith MM, Gosh P. The synthesis of hyaluronic acid by human synovial fibroblasts is influenced by the nature of the hyaluronate in the extracellular environment. Rheumatol Int 1987; 7:113–122. 66. Balazs EA, Denlinger JL. Viscosupplementation: a new concept in the treatment of osteoarthritis. J Rheumatol 1993; 20(Suppl. 39):3–9. 67. Mensitieri M, Ambrosio L, Nicolais L, Bellini D, O’Regan M. Viscoelastic properties’ modulation of a novel autocrosslinked hyaluronic acid polymer. J Mater Sci Mater Med 1996; 7:695–698. 68. Kishimoto M, Yamanouchi U, Mori S, Nakamori F. Experimental study on the substitute of the vitreous body [in Japanese]. Nippon Ganka Gakkai Zasshi 1964; 68:1145–1158. 69. Liesegang TJ. Viscoelastic substances in pphthalmology. Surv. Ophthalmol 1990; 34:268–293. 70. Liesegang TJ. Viscoelastics. Int. Ophthalmol. Clin 1993; 33:127–147. 71. Scuderi G. Ricerche Sperimentali sul Trapianto del Vitreo (Tentativi di Sostituzione Parziale con Vitreo Omologo, con Liquor Eterologo, con Soluzione di Polivinilpirrolidone). Ann. Ophthalmol. Clin. Ocul 1954; 80:213–220. 72. Hayano S, Yoshzino T. Local application of polyvinylpyrrolidone (PVP) for some ocular disease [in Japanese]. Rinsho Ganka, J. Clin. Ophthalmol 1959; 13:449–453. 73. Muller-Jensen K, Klein J, Kohler H. Versuch eines Glaskorperersatzes durk Poliacrylamid. Ber. Deut. Ophthalmol. Ges 1968; 68:181–184. 74. Refojo NF. Glyceryl Metacrylate Hydrogels. J. Appl. Polym. Sci 1965; 9:3161–3170. 75. Yamauchi A, Matsuzawa Y, Nishioka K, Hara Y, Kamiya S. PVA hydrogel for vitreous replacement [in Japanese]. Kobunshi Robunshu 1977; 34:261–266.
14 Production of Microbial Polyesters and Their Application in the Construction of Biodegradable, Controlled Antibiotic Release Systems Vasif Hasirci, F. Korkusuz, E. Bayramli, N. Hasirci, F. Severcan, M. Timucin, and G. Alaeddinoglu Middle East Technical University Ankara, Turkey
N. Akkas Baskent University Ankara, Turkey
I. INTRODUCTION There is a great need for an approach that can maintain sufficient local doses of antibiotics at osteomyelitic sites because the wound areas are almost impenetrable. The inaccessability of the diseased site is due to the decrease of the vasculature at the infected site, and also the site is surrounded by a thick, collageneous material, making access and therefore therapy more difficult. Osteomyelitis is a dramatic disease with a mortality of around 20% and a morbidity three times that amount. Secondary infection due to trauma or iatrogenic infection following orthopedic surgery where implants are used are the main causes of the disease in developed countries, whereas hematogenous osteomyelitis is still a problem in developing countries. Eighty-five percent of the people who are involved in traffic accidents have one or more fractures in their long bones, and at least 50% of them undergo a surgery related to that fracture. Open fractures caused by other events (accidents, old age, etc.) are also sources of infection. In addition there is the possibility of infection during surgeries such as total hip implantation. When metal implants are used (and at present there are practically no orthopedic implants of other materials), the risk of infection increases due to the adherence of bacteria to the metal. A systemic route of antibiotic administration is inefficient because only low levels of antibiotic reach the target site and persist there for a short duration. The use of nonresorbable controlled release systems containing antibiotics to prevent the establishment of the disease or to treat an already established disease by delivering the drug locally has been practiced for quite a number of years. Gentamicin-loaded polymethylmethacrylate (PMMA) [1–6] is the pioneer of these systems and is still in clinical use; however, it has its disadvantages. The nonresorbability and thus the need for removal upon depletion of the drug content and the low rates of antibiotic release from these hydrophobic, 281
282
Hasirci et al.
acrylic beads are the major problems [7,8]. A resorbable, controlled release implant would diminish the need for a second surgery and would also positively influence the outcome of fracture fixation. As alternatives to PMMA, biodegradable and biocompatible materials such as degradable polymers [9–15], bioceramics [16–22], polymer–ceramic composites [23], and, recently, calcium phosphate–gelatin composites [24] were developed. The major advantages of these implants are decreased morbidity due to one-stage surgery, osteoconductivity and/or osteoinductivity, steady and extended release providing longer and higher antimicrobial agent availability, and a wider selection of antibiotics including thermolabile ones that cannot be used with PMMA microspheres. Polyhydroxyalkanoates (PHA) are of microbial origin, and their chemical structure is very similar to polylactides but they degrade at a much slower rate. Their use in drug release systems started in recent years [25,26]. Poly(3-hydroxybutyrate) and its copolymers with hydroxyvalerate (PHBV) with varying hydroxyvalerate (HV) ratios are the most widely existing members of this biopolymer group [25]. These biodegradable and biocompatible polyesters, because of their unique physicochemical properties such as piezoelectricity which is claimed to induce bone reformation at load-bearing sites (i.e., fracture and bone gap healing), are regarded as specialty materials and are the center of growing interest [27,28]. As a result, there has been a tremendous increase in the application of PHAs in the biomedical field [28,29]. In the experimental study described in this chapter, various polyhydroxyalkanoates were produced using basically two microorganisms under various growth conditions using different carbon sources, and the polymers were characterized. Antibiotic-loaded rods were prepared using two different, wide spectrum antibiotics for drug release studies in vitro and in vivo in rabbits. The efficiency of these antibiotic-loaded biodegradable polyester rods in the treatment of rabbit experimental implant related osteomyelitis (IRO) was studied.
II. MATERIALS AND METHODS Carbon substrates, valerolactone (valeric acid methyl ester), 4-hydroxybutyric acid sodium salt, fructose, sucrose, and Nile red were purchased from Sigma Co. (USA). Bromopropionic acid was obtained from Aldrich Co. (Germany). All growth media were from Oxoid (UK), and the reagents and solvents were from Merck Co. (Germany) and were of analytical grade. The antibiotics Sulperazone威 (S; Sulbactam-Cefoperazone 1:1) and Duocid威 (D; Sulbactam-Ampicilin 1:1) were gifts from Pfizer (Istanbul, Turkey). Stainless steel implant (ca. 0.2 cm3, from a medical grade K-wire) was a gift from Hipokrat A.S. (Izmir, Turkey). Bone wax was from Ethicon (USA). The Staphylococcus aureus strain (coagulase positive; phage type 52/52b) was obtained from a chronic osteomyelitis patient undergoing surgical treatment at the Third Department of Orthopaedic Surgery of Numune State Hospital, Ankara. Phage typing of the bacteria was performed at the Department of Infectious Diseases, Ankara University School of Medicine. All further microbiological studies were performed at the METU Medical Center. The standard inoculum (0.5 mL of 5 ⳯ 106 CFU/mL) was directly delivered into the medullar cavity of the rabbits in the in vivo studies. Alcaligenes latus (ATCC 29713) and Ralstonia eutropha (ATCC 17699) (previously known as Alcaligenes eutrophus, AE) were purchased from American Type Culture Collections (ATCC) (Manassas, VA).
Polyesters in Controlled Release Systems
283
A. Production of Polyhydroxyalkanoates 1. Shake-Flask Studies Homopolymer Production. The polymer production conditions for the two organisms differed. When R. eutropha cells were used, the biosynthesis was realized in two steps. First, the growth of the cells was promoted since a high number of cells was needed for high polymer yield. Following the growth of bacteria for 25 h, which was determined to correspond to the late logarithmic phase, the cells were harvested aseptically, and wet weights were determined, and transferred into production medium. In that step, the polymer production and deposition was encouraged by a stress condition (nitrogen deficiency in this case) and by excess fructose as carbon source. Polymer production by A. latus cells was performed at a single step because A. latus is a growth-associated producer, i.e., it can produce and store PHAs during log phase. The polymer production was realized by the addition of the extra carbon source, sucrose, after 19 h of incubation, which was determined to be the midpoint of the logarithmic phase. Both A. latus and R. eutropha cells were cultivated in 250, 500, and 1000 mL media. Temperature (30⬚C) and shaking speed (150–200 rpm) were kept constant. The A. latus and R. eutropha cells were harvested after 48 and 72 h, respectively. Copolymer Production. With R. eutropha cells, copolymer production was achieved by the addition of cosubstrates to the production medium together with the main carbon source. In the case of A. latus, the cosubstrate addition was realized at the hour 19 together with the excess carbon source addition. Valerolactone (VL) and 4-hydroxybutyrate (4-HBA) were the cosubstrates used for the production of copolymers poly(3-hydroxybutyrate-co-3-hydroxyvalerate) [P(3HB-co-3HV)] and poly(3-hydroxybutyrate-co-4-hydroxybutyrate) [P(3HB-co-4HB)], respectively. Bromopropionic acid (BrPPA) was also used as a cosubstrate with R. eutropha in order to be able to prepare a halogenated copolymer. The effect of different carbon/cosubstrate ratios on the polymer production yield and characteristics were investigated using five different ratios—9:1, 8:2, 7:3, 6:4, and 5:5 (w/ w)—of sucrose and 4-HBA. 2. Fermenter Studies Fermentative production of poly(3-hydroxybutyrate) (PHB) and its copolymer P(3HB-co-4HB) was carried out in a 2-L LH fermentation system (500 series, Bucks, UK) with a running volume of 1.5 L. Three different fermentation runs were performed using A. latus. They were preferred for the fermentative polymer production since they do not require different growth and production media due to their being a growth-associated producer. Temperature, oxygen concentration, pH, and foam controls were automatic. Sugar, nitrogen, dry cell weight (DCW), and polymer content determinations were made at 2-h intervals. For all runs, temperature and agitation rates were kept constant at 30⬚C and 500 rpm, respectively. Twenty milliliters of samples were taken from the fermenter for analysis. Sucrose was used as the carbon source in the production of PHB. Cosubstrate 4-hydroxybutyrate was added together with sucrose (sucrose/4-HBA 24:1 w/w) at the end of hour 12 if the production of the P(3HBco-4HB) copolymer was desired. Part of the sample (10 mL) was filtered from a preweighed filter paper (Whatman cellulose acetate filters, pore size 0.45 m). The filter paper was then dried in an oven (50⬚C, 2 h) and weighed for DCW determination. Ammonium nitrogen and sucrose determinations were carried out in the filtrate. Ammonium nitrogen concentration was measured using a Spectroquant威 14752 Ammonium kit (Merck, Darmstadt, Germany). Sucrose
284
Hasirci et al.
detection was carried out via enzymatic assay. Briefly, the sample was treated with glucose oxidase and peroxidase after acid hydrolysis and H2O2 production were assessed spectrophotometrically at 530 nm, using o-dianisidine dye [30]. B. Determination of PHAs Two methods were used in the determination of the PHA content of the cells. Crotonic acid assay necessitated extraction of the polymer from the cell mass, while the Nile red staining method used live intact cells in the determination. 1. Crotonic Acid Assay The polymer was extracted with a modified procedure of Williamson and Wilkinson [31] by using sodium hypochloride (NaOCl) for the destruction of cell membrane and release of the polymer granules. For this purpose a known amount of freeze-dried cells was dissolved in glass tubes with 10 mL of 1% NaOCl solution and incubated in a water bath (50⬚C) after vigorously shaking for 2 h. After centrifugation for 10 min at 10,000 g, the obtained pellet was washed with 10 mL distilled water and dissolved in chloroform. To determine the polymer amount, precipitation was done in cold ethanol, and polymer was collected by filtration, dried in a vacuum oven, and weighed. To carry out the crotonic acid assay, the ethanol precipitation step was not carried out. Instead, chloroform solutions were maintained in boiling water until all the solvent evaporated and then concentrated H2SO4 (10 mL) was added to each tube to convert hydroxybutyric acid into crotonic acid. The samples were incubated in a boiling water bath for 10 min. The cooled tubes were vortexed and the absorbances at 235 nm were recorded against a concentrated H2SO4 blank. A standard curve constructed with 3-hydroxybutyrate was used in the calculation of PHA content. 2. Nile Red Staining Ralstonia eutropha or A. latus cell suspensions (1 mL) and Nile red solution (1 mL, 0.2 mg/ mL in ethanol) were mixed, and the volume was brought to 5 mL with 70% ethanol. Escherichia coli cells were used as negative control. The solution was incubated at room temperature for 2 h, followed by centrifugation. The concentration of the polymer in the stained cells was determined through spectrofluorimetry at excitation and emission wavelengths of 515 nm and 569.6 nm, respectively. C. Characterization of PHAs 1. Mechanical Properties Mechanical tests were performed on dog bone and rod-shaped specimens at 20 Ⳳ 1⬚C. Tensile tests were performed using a tensile tester (Lloyd LR30K and Lloyd LS500) at a constant rate (1 mm/min for Lloyd LR30K and 2 mm/min for Lloyd LS500). The data were collected at a rate of 60 readings per second. Microhardness tests were performed on sample rods (sanded to yield flat, smooth samples) using a microhardness tester (Wilson Tukon Series 200) with load of 100 gf. A Knoop indenter was used for indentation, and the dwell time was selected as 30 s. 2. Degradation of Polyhydroxyalkanoates Polymeric films were obtained by solvent evaporation; disks (diameter 0.8 cm) were punched out and placed in test tubes containing physiological saline solution (pH 7.4, 5 mL). The air
Polyesters in Controlled Release Systems
285
was removed from the tubes under vacuum, heat sealed after freezing in liquid nitrogen, and incubated at 120⬚C for 1, 7, and 15 days. The samples were examined by FTIR (Nicolet DX), SEM, and DSC (DuPont 2000 DSC system, nitrogen atmosphere, 10⬚C/min heating rate). 3. Surface Modification with Rf Plasma Membrane samples were treated with argon plasma to create neutral surfaces or treated with oxygen plasma to create hydrophilic surfaces. In order to achieve this the samples were placed in the inductively coupled plasma system (Advanced Plasma System Inc., USA), the inner pressure was decreased to 10 mtorr, and electrical discharge was applied at 50 W for 30 min. 4. Molecular Weight Determination Through Viscometry Solution viscosities of the produced polymers were measured in chloroform with an Ubbelohde type viscometer at 25⬚C. Molecular weights (MW) of the polymers were determined by using the Mark-Houwink equation: [η] = K Mα
(1)
where K and ␣ are Mark-Houwink constants, is intrinsic viscosity, and M is the molecular weight (g/mol). K and ␣ for the polymers in chloroform at 25⬚C are 1.5 ⳯ 10ⳮ4 (dL/g) and 0.756, respectively [28]. 5. Nuclear Magnetic Resonance A Bruker-Spectrospin Avance DPX 400 MHz multinuclear magnetic resonance (NMR) detector was used to determine the structure and the composition of the PHA produced. The samples were dissolved in deuterated chloroform. For NMR spectra, 5–10% w/w of the polymer solution was prepared and added to a 20-cm-long NMR tube. NMR spectra were obtained with tetramethylsilane as the reference. 6. Fourier Transform Infrared Spectroscopy Infrared spectra of P(3-HB), P(3HB-co-3HV), and P(3HB-co-4HB) were obtained in the range of 4000–400 cmⳮ1 by a Fourier transform infrared (FTIR) spectrometer (Bomem 157D, Canada). Samples were dissolved in chloroform, and a solution cell with KBr windows and 0.1-mm spacers was used for the measurements at room temperature. Infrared spectra of solvent-cast films which were obtained by solvent evaporation were also studied. D. Antibiotic-Loaded Controlled Release Rods: Preparation and in Vitro Antibiotic Release A homogeneous paste was prepared by mixing polymer (2 g), antibiotic (Sulperazon or Duocid) (2 g), and chloroform (5 mL) to yield a polymer/drug ratio of 1:1 (w/w). The paste was introduced into a glass mold (20 ⳯ 0.3 ⳯ 0.3 cm3). Following the complete solvent evaporation, controlled release rods (1.0 cm long) were cut and stored in sealed pouches until use. The same procedure was used in the preparation of rods with polymer/drug ratios of 2:1 and 5:1 (w/w). Some rods were then coated by dipping in a chloroform solution of the same polymer (100 mg/mL) and allowing to dry at room temperature. Release studies were carried out in triplicate in a shaking water bath (rate of 70 cycles/ min) at 37⬚C in sterile PBS solution (0.1 M, pH 7.4, 100 mL). At certain intervals aliquots were
286
Hasirci et al.
removed, absorbances were measured at max 224 nm using a UV spectrophotometer (Shimadzu, Model 1201, Japan), and then were reintroduced to the flasks. E. In Vivo Studies 1. Experimental Groups and Surgical Procedures Fifty local albino rabbits (Institute of Farming Animals, Ankara, Turkey) aged 3 months and weighing 2090 Ⳳ 390 g (range 2000–2500 g) were used. The rabbits were divided into four groups and monitored for 1, 3, and 6 weeks. The rabbits were caged individually at room temperature allowing free movement and were on a 12-h day/night (light/dark) cycle. Standard laboratory food and water were provided constantly. All procedures were in full compliance with Turkish Law 6343/2, Veterinary Medicine Deontology Regulation 6.7.26, and with the Helsinki Declaration of Animal Rights. 1. Group 1: Sulperazone-Loaded Implant, Placebo Implant; No infection, n ⳱ 12. Anesthesia was induced with an intramuscular injection of ketamine (100 mg/kg) and xylazine (5 mg/ kg). An aperture 1.0 cm long was created in the proximal cortex of the right tibia. Sulperazoneloaded poly(3-hydroxybutyrate-co-4-hydroxybutyrate) (polymer 1) or poly(3-hydroxybutyrateco-3-hydroxyvalerate) (polymer 3) rods (with antibiotic/polymer ratio 1:1) were implanted into the intramedullary area through this aperture. On the left tibia of the same animal the same type of rod with no antibiotic was implanted. Six of the implants were subjected to x-ray radiography, histology, and macroscopic examination. The other six were subjected to scanning electron microscopy (SEM), in vitro release kinetic determination, and macroscopic examination. 2. Groups 2–4: Drug-Loaded Implant, Placebo Implant; Infection. Rabbit Model for Implant-Related Experimental Osteomyelitis (IRO). Following anesthesia, a 1.2-mm aperture was created in the proximal cortex of the right tibia. Staphylococcus aureus was inoculated through this aperture together with a stainless steel implant (ca. 0.2 cm3). This aided bacterial attachment, establishment of glycocalix, and exponential growth of bacteria. Following implantation and bacterial inoculation, the aperture was sealed with bone wax. The establishment of IRO in 3 weeks was confirmed by macroscopy, radiography, and swab cultures. Sulperazone/P(3HB-co4HB) rods of 1.2 ⳯ 0.2 ⳯ 0.2 cm3 size were implanted in the infection sites in groups 2 and 3, whereas Duocid-loaded P(3HB-co-4HB) polymers of the same size were implanted in group 4. Soft tissue debridement was not performed at the time of implantation, and the stainless steel implants were left in place. 2. Evaluation Methods of in Vivo Tests Macroscopy. Local signs of infection were graded between 0 and 3 according to swelling, local warmth, and drainage, where 0 ⳱ none and 3 ⳱ grossly swollen. Radiography. Conventional x-rays (Siemens model Multix-C) were obtained at a setting of 44 kV and 3.2 mAs/s. The distance of the x-ray source to the bones was 100 cm. X-ray films (Agfa Gevart Curix) were developed in an Agfa Curix 60 automatic developer. Signs of disease were graded according to a scale where 0 ⳱ none and 4 ⳱ severe. Microbiology. Swab cultures were obtained directly from the bone at the end of the 3week incubation period prior to treatment with PHBV antibiotic-loaded rods. In some cases (i.e., soft tissue abscess formation), swab cultures were also taken from the soft tissues. The swabs were cultured on blood-agar and EMB (eosin methylene blue agar) plates within ca. 15–30 min of sampling. Antibiotic sensitivity of the isolated bacteria was determined by commercially available Sulperazone and Duocid paper disks using Mueller-Hinton agar (Difco, USA).
Polyesters in Controlled Release Systems
Statistical Analysis.
287
Two-tailed heteroscedentic t test was performed for statistical anal-
ysis.
F. Production and Characterization of Calcium Hydroxyapatite Hot (80⬚C) calcium nitrate solution (1 M) and diammonium hydrogen phosphate solution (1 M), in the Ca/P ratio of 1.667, as in Ca10(PO4)6 (OH)2, were mixed by adding dropwise while stirring vigorously at a pH of 9–9.5 (maintained with 25% ammonia). The system was allowed to age for 24 h, centrifuged, precipitate washed twice with deionized water, and finally dried at 120⬚C. Calcium and phosphorous contents of the obtained hydroxyapatite (HAP) powder were determined by atomic absorption spectroscopy and by colorimetry, respectively. The effect of calcination temperature on powder morphology and on the sintering tendency of HAP was studied by calcining for 20 h at temperatures ranging from 150 to 1300⬚C in 100⬚C increments. The powder x-ray diffraction (XRD) patterns of the calcined precipitates were obtained using a Rigaku D-MAX/B powder diffractometer with monochromatic Cu-K␣ radiation. Hydroxyapatite powder was examined under a JEOL 6400 scanning electron microscope. The crystallographic features were studied by powder XRD collected at 20 scan rate of 0.01 degree/min. High purity silicon metal (grade A10 of Hermann C. Starck, Inc.) was used as an external standard to eliminate systematic operational errors. The lattice parameters were evaluated by well-known cell refinement methods.
III. RESULTS A. Production of Polyhydroxyalkanoates 1. Shake-Flask Studies PHA Production by R. eutropha. Ralstonia eutropha accumulated almost pure PHB when fructose was provided as the sole carbon source. Polyester content obtained with chloroform extraction showed that the polymer yield [PHA per dry cell weight (DCW), w/w%] of ca. 43.5% is possible. Five different substrate-to-cosubstrate ratios were used to determine the effect of cosubstrate (4-HBA) introduction on (1) growth, (2) PHA yield, (3) extent of polymer formation, (4) mole fraction of polymer and (5) molecular weight and intrinsic viscosity of polymers. The results are summarized in Table 1. It can be seen that the DCWs and polymer recovery displayed similar trends. The best composition was found to be an 8:2 ratio with a 43.5% PHA recovery. By then, as the cosubstrate ratio increased, a gradual decrease was observed not only in dry cell amount, but also in the polymer recovery. In addition, 8:2 composition gave the highest copolymer content. When BrPPA was used as the cosubstrate, PHB was produced instead of copolymer. In addition, cell and polymer yield were low. When used in a 19:1 (F/BrPPA) ratio, polymer yield was 29%, while a 9:1 ratio resulted in only trace amounts of polymer. PHA Production by A. latus. Polyester contents determined by chloroform extraction show that A. latus cells grown on sucrose (15 g/L) accumulated PHB at a level of 44.2% of dry biomass. When sucrose was used at a higher concentration (25 g/L), the biomass increased but there was a significant decrease in the polyester deposition capability of these cells (18.4% yield), probably due to the catabolite repression of the producer cells under this condition.
288
Hasirci et al.
Table 1 Properties of Polymers Produced by R. eutropha with Different Fructose/4-HBA Ratios in the Medium F/4-HBA ratios 9:1 8:2 7:3 6:4 5:5
Dry cell weight (g/L)
Polymer yield (g/L)
Polymer/dry cell weight (% w/w)
Copolymer content (mol %)
MW ⫻ 105 (g/mol)
Intrinsic viscosity (dL/g)
6.04 6.27 6.21 5.33 4.72
2.34 2.73 1.77 1.39 1.23
39.16 43.53 28.41 26.01 25.99
6.00 9.86 9.60 7.79 5.27
1.2 1.7 0.9 2.2 2.5
1.07 1.34 0.87 1.63 1.82
With sucrose as the main carbon source and 4-HBA as cosubstrate, the results summarized in Table 2 were obtained. It was seen that the increasing 4-HBA concentration decreased viability and consequently DCW. The gradual decrease of the cell mass finally resulted in nearly half of the initial value, confirming the toxic effect of 4-HBA. Polymer yield was also decreased. On the other hand, highest polymer recovery was reached at a 6:4 ratio (31.31%), although the difference with the other ratios is quite minimal. The maximum mol fraction of 4-HB (about 7.3%) was achieved with 8:2 and 7:3 S/4-HBA ratios. 2. Fermenter Studies The experimental conditions and results of the fermentation studies are summarized in Table 3. It can be seen that three different runs were performed. Run 1. PHB Production with A. latus Without pH Control. When the carbon source was sucrose, PHB deposition started at the time of nitrogen depletion (12 h). Between 12 h and 21 h the polymer production rate was at its maximum (54.7 mg PHB per gram DCW per hour). A second stage was observed where the production rate was about one-tenth of the first stage (5.5 mg PHB/g DCW/h) (Fig. 1). At the end of 42 h the carbon source was depleted. The resultant dry cell weight and polymer content were 13.70 and 4.61 g/L, respectively, which corresponds to a yield of 33.6% (based on dry weight) (Table 3). Run 2. PHB Production with A. latus with pH Control. Fermenter was initiated with 20 g/L sucrose as the sole carbon source, and, as was done in the previous run, sucrose (15 g/
Table 2 Properties of Polymers Produced by A. latus with Different Sucrose/4-HBA Ratios in the Medium S/4-HBA ratios 9:1 8:2 7:3 6:4 5:5
Dry cell weight (g/L)
Polymer yield (g/L)
Polymer/dry cell weight (% w/w)
Copolymer content (mol %)
MW ⫻ 105 (g/mol)
Intrinsic viscosity (dL/g)
7.16 6.29 5.04 4.57 3.70
1.86 1.72 1.46 1.43 0.95
25.99 27.33 29.00 31.31 25.67
4.59 7.30 7.26 6.31 6.55
1.8 1.6 3.2 2.8 2.4
1.43 1.27 2.16 1.97 1.75
Polyesters in Controlled Release Systems
289
Table 3 PHA Production of Fermentation Using A. latus
Run 1 Run 2 Run 3
Sucrosea (g/L)
Nitrogen (g/L)
Oxygenation (%)
pH control
Dry cell weight (g/L)
15 : 15 20 : 15 20 : 15b
0.460 0.410 0.264
20 30 20
— ⫹ —
13.70 13.60 7.91
Polymer Polymer dry cell yield (g/L) weight (%) 4.61 6.43 2.83
33.60 47.30 35.80
a
Carbon source was added twice. The second addition was done at the end of hour 12, which corresponded to the nitrogen limitation. b Cosubstrate, 4-hydroxybutyrate (4-HBA) was added together with sucrose (24 : 1 w/w sucrose/4-HBA) at the end of hour 12.
L) was introduced at the onset of nitrogen depletion (12 h). In this run the pH was adjusted to 7.0 Ⳳ 0.2. Sucrose was totally metabolized in the fermenter at the end of 36 h and the cells were harvested. The polymer production rate was found to be 7.1 mg PHB/g DCW/h. The DCW was 13.6 g/L. The polymer constituted 47.3% of DCW and was the highest yield obtained. Run 3. P(3HB-co-4HB) Production with A. latus. The fermenter was initiated with 20 g/L sucrose, and the rest of the parameters were adjusted according to the optimizations in the previous run. At nitrogen depletion, cosubstrate 4-hydroxybutyrate was added (15 g/L; sucrose/ 4-HBA 24:1). Following this, a sharp decrease was observed in the pH of the medium and even
Figure 1 Fermentative production of PHB by A. latus (RUN 1).
290
Hasirci et al.
though several attempts were made to raise the pH back to neutral it was not possible. Due to this, DCW (7.9 g/L) was far below the input carbon (35 g/L). The polymer was ca. 36% of DCW. B. PHA Quantification via Spectrofluorimetry Using Nile Red A calibration curve was plotted for R. eutropha and the extent of staining was compared with the negative control (E. coli). The fluorescence intensities were quite different for E. coli and R. eutropha. The comparatively small amounts of staining in E. coli could be attributed to the nonspecific adsorption on lipids and on trace amounts of polyester found in the membrane of E. coli. The Nile red staining method yielded a polymer content of 42.8% PHA in DCW, whereas the usual extraction methods yield 42.6%. This shows that the on-line PHA detection method employed here is rapid, reliable, and more sensitive than the conventional and tedious spectrophotometric crotonic acid assays. C. Characterization of PHAs Different PHA polymers that were produced in this study were characterized with various methods. In order to be able to obtain polymers with different compositions, the bacteria type, carbon sources, and production conditions were changed. Properties of the resultant polymers are presented in Table 4. In this table it is observed that the input ratios of the carbon sources do not always match those of the products obtained. For example, with polymer 1 the starting composition was sucrose/4-HB 24:1 (4% 4-HB in the total), and the product had almost twice that amount of 4-HB (as determined by NMR). But then again valerolactone content of polymer 3 had a starting composition of 10%, and the product had 11.3% VL in the composition. Thus, it was not always predictable. This could be one of the reasons why the commercial products of a given specification are expensive. Another common observation is that the products had a molecular weight on the order of 105, considered low by industral polymer users. Some relatively high molecular weight polymers such as polymer 5, which was mainly poly(3-hydroxybutyrate), could be preferred for high load-carrying applications, whereas the lower molecular weight ones would be more useful in higher degradation rate applications. It was thus possible to prepare polymers for specific purposes using the approach of the present study. 1. Mechanical Testing The yield stress and failure stress results show that polymers that contained valerolactone (polymers 3 and 6) and BrPPA (polymer 5) had higher strengths (Table 5). Polymer 5 showed relatively high yield and failure strains, implying that it has a high energy absorption capacity. This behavior was clearly seen in toughness and resilience data. Polymer 2 had the highest resilience but the lowest toughness value. This was because the yield and failure occurred at the same stress, indicating that it is very sensitive to plastic deformations. In the elastic range, it had the highest energy absorption capacity. Polymer 4 had a relatively high standard deviation in yield stress, yield strain, and failure stress, and also it had the lowest yield stress, yield strain, and resilience average values. Tensile tests were also performed on Sulperazone-loaded (1:1 w/w) rods of PHBV22—polymer 1, P(3HB-co-4HB, 7.7% 4-HB and polymer 3, P(3HB-co-3-HV, 11.3% 3HB)—to determine their mechanical properties (Table 6). The PHBV22-based rod had the lowest modulus of
AL
AL
AE
AE
AE
AE
AL
AL AL
1
1⬘
2
3
4
5
6
7 8 S (15) S/BrPPA (24 : 1)
S/VL (24 : 1)
F/BrPPA (19 : 1)
F/4-HB (18 : 2)
F/VL (18 : 2)
S/4-HB (24 : 1) HYPO treated F
S/4-HB (24 : 1)
Commercial Aldrich
45.3 ⫾ 1.4
42.4 ⫾ 8.0
47.8 ⫾ 1.3
45.0 ⫾ 5.6
(185) 173.46 (Tg⬃5 ° C)
176.3
163.75 167, 142.9 (small)
160.82
169.32
165.11, 138, 58.10
162.21, 138.06, 58.41
149.01, 139.17
DSC
2.42.105
3.84.105
6.41.105
2.91.105
3.31⫻105
3.53⫻105
1.45⫻105 U. T. Str. 12.67 MPa Ef 0.022 U. T. Str. 28.82 MPa Ef 0.135 U. T. Str. 20.96 MPa Ef 0.083 U. T. Str. 22.56 MPa Ef 0.2111 U. T. Str. 20.62 MPa Ef 0.167
U. T. Str. 6.42 MPa Ef 1.034 U. T. Str. 16.18 MPa Ef 0.12
1.26⫻105 2.58⫻105
Mechanical testing
MW
S, sucrose; F, fructose; VL, valerolactone; 4-HB, 4-hydroxybutyric acid; BrPPA, bromopropionic acid; AL, Alcaligenes latus; AE, R. eutropha.
Alcaligenes
0
Yield (%)
NMR
P(3-HB)
P(3-HB)
11.3% 3-HV 88.7% 3-HB 9.5% 4-HB 90.5% 3-HB P(3-HB)
Pure P(3-HB)
16.5% 3-HV 83.5% 3-HB 7.7 % 4-HB 92.3 % 3-HB P(3-HB)
Microorganism
Sample
Growth medium
Summary Information About the Production of Various PHAs by R. eutropha and A. latus and the Characterization of the Produced Polymers
Table 4
Polyesters in Controlled Release Systems 291
292
Hasirci et al.
Table 5 Mechanical Properties of Dog Bone Specimens Obtained from Tensile Tests
Property Toughness (MPa) Resilience (kPa) Failure stress (MPa) Failure strain Yield stress (MPa) Yield strain UTS (MPa)
Polymer 1, P(3HB-co4HB)–AL
Polymer 2, P(3-HB)– AE
Polymer 3, P(3HB-co3HV)–AE
Polymer 4, P(3HB-co4HB)–AE
Polymer 5, P(3HB-co3BrHB)–AE
Polymer 6, P(3HB-co3HV)–AL
1.67⫾0.73 93.5⫾23.5 12.5⫾3.7 0.124⫾0.048 10.9⫾1.5 0.019⫾0.002 16.18⫾1.42
0.95⫾0.63 950.9⫾629.8 12.7⫾1.9 0.021⫾0.003 12.7⫾1.9 0.022⫾0.003 12.67⫾1.85
2.60⫾0.94 177.6⫾22.9 23.8⫾4.7 0.135⫾0.043 18.5⫾1.2 0.022⫾0.003 28.82⫾1.09
1.23⫾0.57 79.8⫾53.1 17.1⫾5.8 0.083⫾0.016 10.6⫾4.7 0.014⫾0.005 20.96⫾6.92
3.18 285.3 12.7 0.211 16.8 0.045 22.56
2.80⫾0.49 169.8⫾16.3 13.2⫾3.7 0.167⫾0.020 15.5⫾1.6 0.025⫾0.003 20.62⫾0.70
Values are given as averages ⫾ standard deviations.
elasticity and failure strain among the three materials considered. The material with the highest modulus of elasticity and the largest failure strain was polymer 3. The PHBV22 and polymer 1 showed a brittle failure, while polymer 3 possessed some ductility. The modulus of elasticity values for these three materials were all of the same order of magnitude. The tensile properties of the drug-loaded rods are not very important for the application at hand, implantation into rabbit tibia. 2. In vitro PHA Degradation Studies Polymers 1 (7.7% 4-HB and 92.3% 3-HB produced by A. latus from sucrose and 4-HB), 2 (pure 3-HB produced by R. eutrophus from fructose), and 4 (9.5% 4-HB and 90.5% 3-HB produced by R. eutrophus from fructose and 4-HB) were aged in vitro at 120⬚C for 15 days and the changes caused by this treatment were investigated with scanning electron microscopy (Fig. 2), Fourier transform infrared spectroscopy, and differential scanning calorimetry (DSC). With polymer 1, the FTIR peaks at 1530–1540 cmⳮ1 were altered, and the 927 cmⳮ1 peak of the original sample disappeared upon aging. In polymer 2, the peak intensities at 1539 cmⳮ1 and 674 cmⳮ1 decreased upon aging, and the peak at 927 cmⳮ1 shifted to 933 cmⳮ1. With polymer 4, the intensity of the peak at 3437 cmⳮ1 increased upon aging indicating oxidation in the structure. The rest of the spectra was practically the same.
Table 6 Mechanical and Physical Properties of Rod Specimens Obtained by Tensile Testing Property Toughness (MPa) Yield stress (MPa) Yield strain Failure stress (MPa) Failure strain UTS (MPa) Modulus of elasticity (MPa) Molecular weight (daltons ⫻105)
PHBV22/ Sulperazone (1 : 1)
Polymer 1/ Sulperazone (1 : 1)
Polymer 3/ Sulperazone (1 : 1)
0.364 4.52 0.157 4.519 0.157 4.52 27.9 1.26
0.635 7.02 0.174 7.516 0.189 7.65 41.9 2.90
4.523 13.02 0.268 14.388 0.422 15.48 52.9 3.31
Polyesters in Controlled Release Systems
293
Figure 2 Polymers 1 (above) and 2 (below) before (left) and after (right) degradation.
The samples were examined by SEM to check the changes on the surface topography. All the samples demonstrated significant pore formation and fragmentation upon aging, implying significant levels of degradation (Fig. 2). The final method employed was DSC. The DSC of untreated polymer 1 had three endothermic peaks, at 72, 88, and 161⬚C. After aging, the first two disappeared, the third one shifted to 165⬚C, and a shoulder appeared at 170.5⬚C. It is possible that the first two peaks belonged to the solvents used in the preparation of the PHA. The shift of the third peak to higher temperatures indicates conformational changes. Untreated polymer 2 also had three endothermic peaks, at 72, 91, and 168⬚C. After aging, three new peaks at different locations (113, 126, and 162⬚C) were observed. This implied a significant change in the structure, possibly due to hydrolysis. The untreated polymer 4 had three endothermic peaks, at 47, 147, and 161⬚C. The aged samples retained these peaks, and a group of new peaks at lower temperatures (18, 31, and 38⬚C) were observed. It could thus be concluded that high temperature and aqueous saline medium caused considerable changes in the structure of PHAs, possibly through hydrolysis of the ester groups that form the backbone. 3. Surface Modification Studies Certain samples were modified by applying argon plasma to create a neutral surface and others were modified with oxygen plasma to create hydrophilic surfaces. As can be seen from the initial data (Table 7) the extent of modification was not high enough for observation with viscometry. In neither of the samples did NMR reveal a change. Differences could only be observed in the FTIR analysis, revealing bond breakage and carbonyl and/or OH formation (Fig. 3). Finally, the SEM of rods revealed a smoother surface upon treatment. These all implied that the changes caused by the application of rf plasma were all located superficially, on the exterior of the implants, thus not affecting the bulk.
294
Hasirci et al.
Table 7 Influence of Oxygen Plasma Treatment on Polymer Properties Polymer Polymer 1 Polymer 1 Polymer 7 Polymer 7
Plasma treatment
Viscosity [] (dL/g)
Molecular weight (105) (g/mol)
— ⫹ — ⫹
1.9530 1.9562 1.7310 1.7270
2.270 2.275 1.937 1.930
Figure 3 The FTIR spectra of PHBV 22 for (a) plasma untreated and (b) plasma treated.
Polyesters in Controlled Release Systems
295
4. FTIR Spectroscopy of PHB, P(3HB-co-3HV), and P(3HB-co-4HB) A majority of the bands in the infrared spectra of polymers are associated with both the crystalline and amorphous regions. In general when the crystallinity of a polymer is altered, intensities and frequencies of some of the IR bands also vary. The FTIR spectra of poly(3-hydroxybutyrate) (PHB), poly(3-hydroxybutyrate-co-3-hydroxyvalerate), and poly(3-hydroxybutyrate-co-4-hydroxybutyrate) in chloroform solution and a solution-cast film are given in Fig. 4. A comparison of these spectra shows significant differences between the two phases. The IR bands at 1279, 1228, and 1185 cmⳮ1and C—H stretching modes of the CH3 and CH2 groups were found to be sensitive to changes in the degree of crystallinity. Other spectral differences between polymers were observed at asymmetric C—H stretching frequencies of the CH3 and CH2 groups. The carbonyl stretching band is also sensitive to the conformational changes. In crystalline solids the C⳱O frequencies are lowered due to lattice field effect. With PHBV in chloroform solution the C⳱O band is at 1737 and 1731 cmⳮ1 and in the solution cast film at 1724 cmⳮ1.
D. In Vitro Antibiotic Release from PHA Rods In vitro release of antibiotics from the implants was studied in PBS (pH 7.4, 37⬚C). In the release studies from rods constructed of polyester and drug at a weight ratio of 1:1 (uncoated samples) using different PHA it was observed that the drug contents were completely liberated within less than 3 days. Upon coating of the rods with polymer, the release profiles became more linear (approaching zero-order), and duration of complete release was extended to about 2 weeks (Fig. 5). It was found in general that the release of Sulperazone from the coated rods was slower than that of Duocid, and this trend did not change with the polyester type or composition (Figs 6 and 7). The release rate was almost inversely proportional to the coating thickness in all the samples tested, and this also was not influenced by the polymer type (Tables 8 and 9).
E. In Vivo Results 1. In Vivo Antibiotic Release The rods implanted into rabbit tibia were retrieved after 1, 3, and 6 weeks, and their drug contents were determined after extraction. The drug in the uncoated rods was released almost completely within the first week. This could have been prolonged if the coated rods (which released longer in vitro) were used instead of the uncoated ones. One interesting observation, however, is that the therapy was very effective even when the release rate was this high. Thus, further optimization could only make the results better (Tables 10 and 11). 2. Macroscopic and Radiological Findings Macroscopic and radiological findings of the in vivo experiment are summarized in Tables 12 and 13. Group I. This was the uninfected control. Macroscopic evaluation at follow-up revealed minimal swelling and increase in local warmth (most probably due to the surgery rather than the reaction toward the implant) but no drainage. The average scores for swelling, local warmth, and drainage were 0.6/4, 0.2/3, and 0/4, respectively. There was no difference between the macroscopic results of 1, 3, and 6 weeks.
296
Hasirci et al.
Figure 4 The FTIR of the samples in CHCI3 solution (dashed line) and as solution cast film (solid line). (a) PHB homopolymer, (b) PHBV copolymer, and (c) P(3HB-4HB) copolymer (at room temperature).
Polyesters in Controlled Release Systems
297
Figure 5 Release behavior of P(3HB-4HB)/sulperazone (1:1 w/w, polymer to drug ratio) uncoated and coated rods. Mean coating thickness: 10.7 mg.cmⳮ2.
The overall scores for radiographical findings were 0.8/5 for the right and 1.1/5 for the left limbs. There was no statistical difference between the antibiotic-loaded and antibiotic-free polymeric rods. There also was no statistically valid difference between polymers 1 and 3 based on macroscopic and radiological examination. A significant observation was that the rod made from polymer 1 was more rigid and less porous than that produced from polymer 3, and in the later studies polymer 3 was used.
Figure 6 Release behavior of P(3HB-4HB)/Sulperazone (1:1 w/w, polymer to drug ratio) uncoated and coated rods. Mean coating thickness: 10.9 mg.cmⳮ2.
298
Hasirci et al.
Figure 7 Release behavior of P(3HB-4HB)/Duocid (1:1 w/w, polymer to drug ratio) uncoated and coated rods. Mean coating thickness: 7.17 mg.cmⳮ2.
Table 8 In Vitro Release from PHA Rods Loaded with Sulperazone (1:1, w/w) Sample Polymer 1/UC Polymer 1C Polymer 3/UC Polymer 3C Polymer 4/UC Polymer 4/C Polymer 6/UC Polymer 6/C
Uncoated rod weight (mg)
Sulperazone content (mg)
Polymer coat weight (mg)
Cumulative release (%) per duration (days)
92.0 85.7 85.5 100.5 102.9 94.3 117.6 92.9
46.0 42.8 42.8 50.2 51.4 47.1 58.8 46.4
— 15.1 — 14.8 — 21.5 — 9.7
100.8 (3) 72.9 (12) 113.5 (1) 69.8 (12) 98.7 (3) 53.5 (13) 111.3 (3) 95.7 (9)
UC, uncoated; C, coated. Properties of polymers 1, 3, 4, and 6 can be found in Table 4.
Table 9 In Vitro Release from PHA Rods Loaded with Duocid (1:1, w/w) Sample Polymer 1/UC Polymer 3/UC Polymer 3/C Polymer 3/C
Initial rod weight (mg)
Duocid content (mg)
Polymer coat weight (mg)
Cumulative release (%) per duration (days)
95.9 101.7 79.0 111.3
47.9 50.9 39.5 55.6
— — 11.8 11.0
99.8 (6) 100.3 (6) 97.0 (7) 93.7 (12)
UC, uncoated; C, coated. Properties of polymers 1 and 3 can be found in Table 4.
Polyesters in Controlled Release Systems
299
Table 10 In Vivo Release (in Uninfected Rabbits) of Sulperazone from polymers 1 and 3 Loaded 1:1 with Sulperazone Polymer type
Sacrifice time (weeks postimplantation)
Initial rod weight (mg)
Weight of Sulperazone (mg)
Weight of retrieved rod (mg)
The amount of Sulperazone remained (mg)
Sulperazone released (%)
1 3 6 1 3 6
68.2 90.4 92.0 67.8 91.9 80.7
34.1 45.2 46.0 33.9 45.9 40.3
41.6 42.1 48.7 59.3 53.1 41.6
1.4 0.8 0.2 1.3 1.4 0.7
96.0 98.2 99.5 96.0 97.0 96.8
Polymer 1
Polymer 3
Table 11 In Vivo Release from Polymer 1 Loaded with Sulperazone (1:1, w/w) in Rabbits Infected with S. aureus Time postimplantation (weeks) 1 3 6
Initial weight (mg)
Weight of Sulperazone (mg)
Weight of retrieved rod (mg)
Amount of Sulperazone remained (mg)
Sulperazone released (%)
92.9 92.1 80.8
46.4 46.0 40.4
54.0 43.0 45.9
3.9 1.4 2.2
91.5 96.9 94.6
Table 12 Overall Macroscopic Findings Macroscopy (weeks) 1 3 6
Swelling
Local warmth
Drainage
2.8 to 1.2 (p⫽0.001) 3.0 to 0.3 (p⬍0.0001) 2.0 to 0 (p⬍0.0001)
2.0 to 0 (p⬍0.0001) 2.0 to 0 (p⬍0.0001) 1.6 to 0 (p⬍0.0001)
0.6 to 0.8 (p⫽0.760) 0.8 to 0 (p⬍0.0001) 0.4 to 0 (p⬍0.0001)
Table 13 Overall Radiological Findings Radiology (weeks)
Periosteal elevation
Architectural deformation
Widening of bone shaft
New bone formation
Soft tissue deformation
1
2.2 to 2.2 (p⫽1.0) 3.3 to 2.3 (p⫽0.304) 2.8 to 1.4 (p⫽0.156)
2.0 to 1.4 (p⫽0.172) 2.8 to 2.0 (p⫽0.730) 2.4 to 1.0 (p⫽0.008)
1.8 to 1.2 (p⫽0.359) 3.3 to 2.0 (p⫽0.041) 2.6 to 1.2 (p⫽0.083)
2.2 to 1.6 (p⫽0.421) 2.8 to 2.3 (p⫽0.654) 2.4 to 1.4 (p⫽0.203)
3.0 to 2.0 (p⫽0.296) 3.5 to 1.8 (p⫽0.020) 3.2 to 0.4 (p⬍0.001)
3 6
300
Hasirci et al.
Group II. In group II, IRO was established and Sulperazone-loaded rods were used in the treatment. Infection rate in the rabbits following 3 weeks of inoculation was 100%. Following implantation of the PHBV–Sulperazone at the infection site, bacteria could not be isolated on weeks 1, 3, or 6, revealing complete eradication. The macroscopic findings of swelling, local warmth, and drainage decreased from 2.0/4 to 0.8/4, 1.6/3 to 0.1/3, and 0.4/4 to 0.1/4, respectively, between the time before and 6 weeks after implantation. Group III. In group III, IRO was established and treatment with Sulperazone-loaded PHBV rods was carried out. Infection rate was 100% following 3 weeks of inoculation. Following treatment with implantation of the Sulperazone–PHBV at the infection site, bacteria could not be isolated on weeks 1 or 6. On the third week, however, there was an anomaly. A small amount of S. aureus was detected in the swab cultures. Macroscopy revealed obvious swelling of the soft tissues and increase in local warmth as a result of infection. Drainage was present in four of the 18 animals. At the end of the treatment, two of the draining animals were completely cured, but two other rabbits, which originally had no drainage, developed mild drainage following treatment. The score of drainage before and after treatment, therefore, remained constant (0.4/4). In the radiographical findings the most significant changes were the loss of soft tissue and the deformation followed by the widening of the bone shaft. Although statistically not significant, architectural deformation was also substantially influenced. Upon implantation of PHBV–Sulperazone rods the overall score of radiological findings decreased from 2.6/5 to 1.7/ 5, indicating the effectiveness of treatment. Group IV. In group IV, Duocid was used as the antibiotic. Microbiological studies showed that the infection rate was 100% following 3 weeks of inoculation. Following treatment with implantation of the PHBV–Duocid rods at the infection site, bacteria could not be isolated on weeks 1, 3, or 6. In one of the animals sacrificed at 3 weeks, S. aureus was present in the adjacent soft tissue swelling site and not in the bone. F. Production of HAP Powder Chemical compositions of oven-dried precipitates indicated that all were virtually pure HAP with a Ca/P ratio of 1.667. The typical XRD pattern for HAP is reproduced in Fig. 8. This particular HAP was the precipitate calcined at 900⬚C for 20 h. The peaks of the diffractogram were compared with the corresponding JCPDS card number 9–432. The excellent match was a proof of the fact that the HAP produced was stoichiometric and phase pure. Lattice parameter determinations confirmed that the crystal structure of the HAP was hexagonal and that the pertinent lattice parameters were a ⳱ 0.94163 nm and c ⳱ 0.68815 nm. These unit cell dimensions agreed quite well with the data available in JCPDS files. The HAP powders obtained from the calcination process described above were examined under a JEOL 6400 scanning electron microscope. The powder calcined at 500⬚C seemed to have a flaky structure with length-to-diameter ratio of about 5 and a typical diameter less than 0.1 m. As the calcination temperature was raised spherical morphology developed gradually. The micrograph for the sample calcined at 700⬚C showed that the powder size was still very small (ca. 0.1 m). The coarsening of the powder started at 800⬚C; the micrograph at this temperature revealed well-defined individual powder particles. The average particle size of the HAP precipitate calcined at 900⬚C was about 0.2 m. At this temperature, however, there was considerable coalescence between powder particles indicating the start of partial sintering. It was thus observed that calcination temperature could be used as a tool to produce HAP in various sizes and shapes and could be of great value in the production of implant materials for strength requiring biomedical applications.
Polyesters in Controlled Release Systems
301
Figure 8 XRD trace for pure hydroxyapatite.
IV. CONCLUSION In conclusion, it was possible to design and produce polyesters of known composition by using different bacteria, growth conditions, and carbon source. Their characterization through NMR, viscometry, FTIR, and DSC was without any complications. Their mechanical properties were found to change with the polymer tested, indicating that it is possible to obtain implants of varying mechanical properties. These properties could further be improved by introduction of the HAP produced in the study. Surface properties on the other hand could be modified by using a rf-plasma treatment approach. P(3HB-co-4HB)–Sulperazone and – Duocid rods were macroscopically and radiologically effective in treating experimental IRO. Swelling diminished gradually in 6 weeks. Local warmth subsided in the first week. Drainage, however, was present in the first week. It was a nonpurulant sterile discharge, probably due to the presence of the polymer itself. This drainage disappeared completely at 3 and 6 weeks. The surgery site was without infection, the implant could hardly be detected, and bone healing was almost completed at 6 weeks. Radiological scores confirmed the findings of macroscopy and microbiology. Although improvement of the radiological data was not significant in the first week, the results on weeks 3 and 6 were improved. At 6 weeks the most apparent improvement was recorded in the soft tissue deformation and then in the architectural deformation. The improvement of periosteal elevation and new bone formation were not that significant, as it is almost impossible to differentiate the findings of infection and bone healing itself caused by the implantation of rods.
302
Hasirci et al.
ACKNOWLEDGMENTS We acknowledge the support by NATO Scientific and Environmetal Affairs Division and the Scientific and Technical Research Council of Turkey for the Grant TU-Polyesters to V.Hasirci and colleagues. We also thank the Middle East Technical University for making this study possible. The authors also wish to acknowledge the significant contributions of the following colleagues for conducting various stages of the experiments, analyses, and preparation for publica¨ z, P. tion of the final form: I. Gursel, F. N. Kok, F. Turesin, Y. Arica, K. Ulubayram, M. K. O Korkusuz, N. Koc¸, C. Balcik, and S. Bayari.
REFERENCES
1. Wahling H, Dingeldein E, Bergmann R, Reuss K. The release of gentamicin from polymethylmethacrylate beads. J. Bone Joint Surg 1978; 60-B:270–275. 2. Vecsei V, Barquet A. Treatment of chronic osteomyelitis by necrectomy and gentamicin–PMMA beads. Clin. Orthop. Rel. Res 1981; 159:201–207. 3. Bayston R, Milner RDG. The sustained release of antimicrobial drugs from bone cement. J. Bone Joint Surg 1982; 64-B:460–464. 4. Murray WR. Use of antibiotic-containing bone cement. Clin. Orthop. Rel. Res 1984; 190:89–95. 5. Bucholz HW, Elson RA, Heinert K. Antibiotic-loaded acrylic cement: current concepts. Clin. Orthop. Rel. Res 1984; 190:96–108. 6. Majid SA, Lindberg LT, Gunterberg B, Siddiki MS. Gentamicin-PMMA beads in the treatment of chronic osteomyelitis. Acta Orthop. Scand 1985; 56:265–268. 7. Calhoun J, Daum W, Mader JT, Koulisis C, Evans EB. Failure of bipolar hip arthroplasty secondary to retained antibiotic-impregnated polymethylmethacrylate beads. J. Bone Joint Surg 1988; 70-A: 1246–1248. 8. Fiddian NJ, Sudlow RA, Browett JP. Ruptured femoral vein. A complication of the use of gentamicin beads in an infected excision arthroplasty of the hip. J. Bone Joint Surg 1984; 66-B:493–494. 9. Jacob E, Setterstrom JA, Bach DE, Heath III JR, McNiesh LM, Cierney III G. Evaluation of biodegredable ampicillin anhydrate microcapsules for local treatment of experimental Staphylococcal osteomyelitis. Clin. Orthop. Rel. Res 1991; 267:237–244. 10. Jacob E, Cierny III G, Fallon MT, McNeill JF, Siderys GS. Evaluation of biodegradable cefalozin sodium microspheres for the prevention of infection in rabbits with experimental open tibial fractures stabilized with internal fixation. J. Orthop. Res 1993; 11:404–411. 11. Jacob E, Cierny III G, Zorn K, McNeill JF, Fallon MT. Delayed local treatment of rabit tibial fractures with biodegradable cefalozin microspheres. Clin. Orthop. Rel. Res 1997; 336:278–285. 12. Wei G, Kotoura Y, Oka M, Yamamuro T, Wada R, Hyon S, Ikada Y. A bioabsorbable delivery system for antibiotic treatment of osteomyelitis. J. Bone Joint Surg 1991; 73-B:246–252. 13. Sampath SS, Garvin K, Robinson DH. Preparation and characterisation of biodegradable poly(Llactic acid) gentamicin delivery systems. Int. J. Pharm 1992; 78:165–174. 14. Teupe C, Meffert R, Winckler S, Ritzerfeld W, To¨rma¨la¨ P, Brug E. Ciprofoxacin-impregnated polyL-latic acid drug carrier. Arch. Orthop. Trauma. Surg 1992; 112:33–35. 15. Garvin KL, Miyano JA, Robinson D, Giger D, Novak J, Radio S. Polylactide/polyglycolide antibiotic implants in the treatment of osteomyelitis. J. Bone Joint Surg 1994; 76-A:1500–1506. 16. Otsuka M, Matsuda Y, Yu D, Wong J, Fox JL, Higuchi WI. A novel skeletal drug delivery system for anti-bacterial drugs using self-setting hydroxyapatite cement. Chem. Pharm. Bull 1990; 38: 3500–3502. 17. Yamamura K, Iwata H, Yotsuyanagi T. Synthesis of antibiotic-loaded hydroxyapatite beads and in vitro drug release testing. J. Biomed. Mater. Res 1992; 26:1053–1064.
Polyesters in Controlled Release Systems
303
18. Shinto Y, Uchida A, Korkusuz F, Araki N, Ono K. Calcium hydroxyapatite ceramic used as a delivery system for antibiotics. J. Bone Joint Surg 1992; 74-B:600–604. 19. Korkusuz F, Uchida A, Inoue K, Shinto Y, Araki N, Ono K. Experimental implant-related osteomyelitis treated by antibiotic–calcium hydroxyapatite ceramic composites. J. Bone Joint Surg 1993; 75B:111–114. 20. Itokazu M, Matsunaga T, Kumazawa S, Wenyi Y. A novel drug delivery system for osteomyelitis using porous hydroxyapatite blocks loaded by centrifugation. J. Appl. Biomater 1995; 6:167–169. 21. Radin S, Campbell JT, Ducheyne P, Cuckler JM. Calcium phosphate ceramic coatings as carriers of vancomycin. Biomaterials 1997; 18:777–782. 22. Takano I, Ishi Y. Experimental study of apatite cement containing antibiotics. J. Orthop. Sci 1997; 2:98–105. 23. Ikada Y, Hyon SH, Jamshidi K, Higashi S, Yamamuro T, Katutani Y, Kitsugi T. Release of antibiotic from composites of hydroxyapatite and poly(lactic acid). J. Control. Rel 1985; 2:179–186. ¨ rs U ¨ , Korkusuz F, Hasirci V. Development of a calcium phosphate–gel24. Yaylaoglu MB, Korkusuz P, O atin composite as a bone substitute and its use in drug release. Biomaterials 1999; 20:711–719. 25. Gu¨rsel I. Use of microbial polyhydroxyalkanoates in the construction of biomedical drug release systems, Ph.D. Thesis. Ankara, Turkey: Middle East Technical University, 1995. ¨ rs U ¨ , Hasirci V. Sulbactum-cefoperazone polyhy26. Yagmurlu MF, Korkusuz F, Gu¨rsel I, Korkusuz P, O droxybutyrate-co-hydroxyvalerate (PHBV) local antibiotic delivery system: in vivo effectiveness and biocompatibility in the treatment of implant-related experimental osteomyelitis. J Biomed Mater Res 1999; 46:494–503. 27. Lee SY. Plastic bacteria? Progress and prospects for polyhydroxyalkanoate production in bacteria. Trends Biotechnol 1996; 14:431–438. 28. Pouton CW, Akhtar S. Biosynthetic polyhydroxyalkanoates and their potential in drug delivery. Adv. Drug Del. Rev 1996; 18:133–162. 29. Gu¨rsel I, Hasirci V. Microcapsules of PHB and PHBV: morphology and release characteristics. J. Microencapsulation 1995; 12:185–193. 30. Arica MY, Hasirci V. Bioreactor applications of glucose oxidase covalently bonded on HEMA membranes. Biomaterials 1993; 14(11):803–808. 31. Williamson DH, Wilkinson CF. The isolation and estimation of poly--hydroxybutyrate inclusions in Bacillus species. J. General Microbiology 1958; 19:198–209.
15 Microencapsulation of Protein Drugs: A Novel Approach Yoon Yeo and Kinam Park Purdue University West Lafayette, Indiana, U.S.A.
I. INTRODUCTION Microencapsulation technologies have advanced significantly during the last few decades, and the current technologies are at such a level that drugs can be delivered at predetermined rates for days or years, depending on the application. These advances, however, usually apply only to low molecular weight drugs. Microencapsulation of high molecular weight drugs such as peptides and proteins is still complicated due to the intricate nature of the physical and chemical properties of protein. In fact, almost all attempts to deviate from traditional parenteral dosage forms have suffered from the same kind of problems. In many cases, developing a traditional dosage form is not an easy task either [1,2]. Since the early promise of sustained protein delivery [3], studies on protein microencapsulation have increased exponentially (Fig. 1A). In fact, some of the early efforts succeeded in bringing the first microparticle product for peptide delivery (Lupron Depot威) onto the market [48]. Not long after the commercial success of Lupron Depot威, however, it was recognized that susceptibility of most proteins to the environmental stresses would pose much harder barriers in development of microparticle systems for proteins. Despite all the obstacles, protein microencapsulation is still an attractive approach, especially when protein pharmaceuticals cannot be delivered via oral routes and the need for infusions or frequent injections demands development of long-term delivery systems. Success in protein microencapsulation relies on maintaining the protein stability during the preparation process and throughout the lifetime of the product. To this end, it is imperative to have clear understanding of the current issues in the protein microencapsulation processes and the previous efforts devoted to overcome the obstacles. Actually, studies on stabilization of encapsulated proteins are increasing at the same speed as those on protein delivery itself (Fig. 1B). This chapter reviews several sources of protein instability within microparticles that have been identified thus far and the strategies to counteract these causes, including the most recent progress. A novel microencapsulation technology, called the solvent exchange method, will be introduced in the last part of this review. Since most experiences on protein microencapsulation in the literature have been gained with the double emulsion–solvent extraction/evaporation method using poly(lactic-co-glycolic acid) (PLGA), discussions are made mostly in reference to PLGA microparticles produced by the double emulsion method unless specified otherwise. 305
306
Yeo and Park
Figure 1 The number of research articles on microparticle systems for protein delivery (A) and protein stability in the microparticle systems (B) published from 1975 to 2001. (SciFinder威 Scholar content is reproduced with the permission of CAS, a division of the American Chemical Society.)
II. BASIC PROTEIN CHEMISTRY A. Forces Influencing Protein Structure Protein stability issues are among the most critical issues in the development of biodegradable microparticle formulations for pharmaceutical proteins. Before discussing specific cases of protein instability in microparticle formulations, it would be profitable to briefly review basic information important to protein stability. Proteins are involved in almost all biological processes with specialized functions. The specific function of a protein is normally determined by its three-dimensional conformation. The native conformation of a protein is dictated by the primary protein structure (the amino acid sequence) and the higher-order structures which are maintained by the subtle balance of noncovalent interactions [9]. Forces influencing the protein structures include covalent bonds, such as peptide bonds and disulfide bonds, and noncovalent interactions such as hydrogen bonds, hydrophobic interactions, electrostatic interactions, and van der Waals interactions. When any of these forces is disrupted, protein is likely to lose its functionality. On the other hand, the free energy of stabilization of the biologically active folded state is only about 50 kJ/mol for most globular proteins [10,11]. While the balance of several noncovalent interactions is of vital importance in maintaining an active protein structure, it does not require a large input of energy to be disrupted. Most protein drugs are particularly susceptible to either chemical or physical inactivation that may occur during formulation processes and storage [1]. Moreover, the problems are not limited only to the processes, but are also found during in vitro and in vivo release stages [12–17]. B. Relevant Routes of Protein Inactivation Protein instability can be broadly classified into chemical and physical instability (Table 1). Chemical instability is induced by breaking and/or reforming covalent bonds, whereas physical instability refers to changes in the secondary or higher structures due to disruption of the noncova-
Microencapsulation of Protein Drugs
307
Table 1 Major Routes of Protein Inactivation During Microencapsulation Chemical inactivation 1. Deamidation Hydrolysis of the side chain amide of asparagine (Asn) or glutamine (Gln) Induced nonenzymatically by water and functional groups within the protein Facilitated by both acidic and alkaline pH and high temperature 2. Oxidation Occurs usually in methionine (Met), tryptophan (Trp), tyrosine (Tyr), and cysteine (Cys) Induced by atmospheric oxygen Catalyzed by metal ions 3. Peptide bond hydrolysis Catalyzed by the protein itself or a contaminating protease Favored in acidic pH 4. -Elimination Destruction of disulfide bonds in cystine residues Favored by alkaline pH and high temperature 5. Disulfide scrambling Incorrect intra- and intermolecular disulfide bond formation Occurs by various pathways at all pH conditions Catalyzed by sulfenium ion (at acidic conditions: disulfide exchange) or by thiolate ion (at neutral to alkaline conditions: thiol–disulfide interchange, thiol catalyzed disulfide exchange) 6. Acylation [22, 23] Peptide acylation due to lactic and glycolic acid accumulated in degrading PLA or PLGA microspheres Occurs with the N-terminal amine group, lysine (Lys), tyrosine (Tyr), or serine (Ser) Physical inactivation 7. Unfolding Usually reversible Induced by elevated temperature, extremes in pH, or denaturants such as hydrophobic organic solvents, salts, and detergents Exposes hydrophobic moieties originally buried in the protein interior 8. Adsorption to hydrophobic surface Induced by binding of unfolded proteins to hydrophobic polymer [24] 9. Aggregation Often irreversible Driven by association of hydrophobic moieties in unfolded proteins Results in a large molecular weight aggregate with low solubility
lent interactions inherent in the protein. Many reviews on the inactivation pathway of proteins are available in the literature [1,2,13,18–21]. Major routes of protein inactivation often encountered in the protein microparticle system are summarized below.
III. PROTEIN INSTABILITY IN PLGA MICROPARTICLES The protein instability issues have two aspects: (1) incomplete and little release of the native protein and (2) immunogenicity or toxicity concern over the degraded or aggregated proteins even with the high recovery of native proteins [15]. Since the earlier recognition of the instability problem in microparticle formulations [13,19], a myriad of studies have focused on identification
308
Yeo and Park
of possible mechanisms of protein inactivation and prevention of such inactivation. Representative examples are provided in the previous reviews [12–15,17]. Although the instability issues may be specific for each polymer and each protein used, the sources described below are generally considered as major causes for the instability of the encapsulated protein. A. Preparation 1. Exposure to Water/Organic Solvent Interfaces Since the successful commercialization of Lupron Depot威 [4–8], the solvent evaporation method utilizing water oil-water (w/o/w) double emulsion became the most widely used technique for encapsulation of most water-soluble drugs. Although this method is thus employed for most proteins and peptides, their behaviors at water/organic solvent (w/o) interfaces, especially in the primary emulsion stage, often cause problems [25–38]. For example, when an aqueous solution of carbonic anhydrase was subjected to vortex mixing in the presence of methylene chloride [30], about 40% of carbonic anhydrase was recovered at the interface after centrifugation of the emulsion. Most of the protein present at the interface was aggregated. Further evidence of protein aggregation at the interface is found in the literature [26,27]. Three proteins [bovine serum albumin (BSA), lysozyme, ovalbumin] were tested under a condition simulating the primary emulsion. Emulsification resulted in considerable changes in the composition of water-soluble protein species and formation of both water-soluble and waterinsoluble aggregates in all three proteins [26]. Susceptibility to the aggregation varied with protein species: BSA was relatively robust, whereas lysozyme and ovalbumin appeared more vulnerable to the interface-induced aggregation. Based on this result, the author questioned the validity of BSA as a model protein for stability studies in the microencapsulation processes. The magnitude of denaturation was also dependent on the species of organic solvents (36.2% of total soluble ovalbumin recovery with methylene chloride, 86.5% with ethyl acetate). The protein aggregation at the interface can be explained by the amphipathic nature of proteins. Formation of an emulsion requires an ordered structure of water molecules around the dispersed solvent droplets, which is thermodynamically unfavorable. Protein molecules arrange themselves at the water/organic solvent interface and expose the hydrophobic moieties (which are originally buried inside the three-dimensional structure in a natural aqueous milieu) toward the hydrophobic organic solvent. In other words, protein, which in a sense is a surfactant, stabilizes the interface. The dependence of stability profiles on the hydrophilicity of the organic solvent can be interpreted in two ways. Partitioning of methylene chloride into an aqueous phase may have facilitated protein unfolding. Although the solubility of methylene chloride in water is negligible as compared with ethyl acetate, the hydrophobicity might have made it a more powerful denaturant. Alternatively, methylene chloride may have provided more hydrophobic interfaces than ethyl acetate, where the proteins undergo the interface-induced conformational changes. The structural fate of lysozyme upon exposure to the water/methylene chloride (w/o) interface was later systemically studied using FTIR spectroscopy [36,37]. Lysozyme aggregates accumulating at the interface, which amounted to about 40% of the total protein, were non-covalent in nature and contained intermolecular -sheets. 2. Exposure to Hydrophobic Organic Solvent Protein denaturation can also occur upon direct exposure to the organic solvent. Denaturation may take place when a protein is dissolved in organic solvents and allowed to construct a structure different from a native one in aqueous solutions [39]. When the polarity of the aqueous phase decreases with the increasing amounts of the organic solvent dissolved in water, internal
Microencapsulation of Protein Drugs
309
hydrophobic moieties of the protein are likely to be exposed, leading to disruption of the hydration shell in its natural state [18]. Problems can also occur when protein particles are suspended within organic solvents during anhydrous preparation. In general, proteins are known to be stable in pure organic solvents due to restrictions on the conformational mobility [40]. However, even in this case, a small amount of moisture (e.g., initial moisture after lyophilization of the protein) available on the surface of protein particle can increase the mobility of a protein molecule, facilitating the protein unfolding and hydrophobic interactions with the organic solvent [41,42]. 3. Physical Stresses During Preparation Shear Stress. Shear stresses produced by the emulsification processes have also been suspected as a cause of protein denaturation. It was reported that although the shear stress itself would not cause any serious denaturation event, once it was coupled with the above factors, an increase in the shear rate facilitated protein aggregation [26]. It appears that the extent of aggregation varies with the emulsification method [34]. According to the previous studies, aggregation increased in the following order: high speed homogenization ⬍ vortex mixing ⬍ ultrasonication [34,35]. In addition, the activity of the encapsulated protein was significantly reduced with an increase in sonication time [34]. Elevated Temperature. Although there is no well-defined mechanism to describe the effect of temperature on the structure and function of proteins due to their structural complications, a general rule exists: the higher the temperature, the lower the protein stability [18]. In particular, significant changes may occur when the temperature is increased above the glass transition temperature (Tg) of the suspended (solid) protein during anhydrous processes [21]. In this regard, heat produced during spray-drying is often considered undesirable. Some proteins undergo cold denaturation as well as a heat induced one [18], and this means that low temperature processes such as freeze-drying and the cryogenic encapsulation process that are employed with an intention to protect the protein integrity are not necessarily safe in certain cases. B. Protein Release Numerous in vitro studies on protein release from microspheres have shown an initial burst release for the first few days and little or incomplete release during subsequent days in spite of increasing porosity resulting from polymer degradation. This release pattern is unique to proteins, as opposed to the biphasic release profile of small molecular weight nonprotein drugs [43]. The incomplete release is mainly attributed to the instability of the encapsulated protein. Denaturation events that may occur during the release period are chemical inactivation, covalent or noncovalent aggregation, and nonspecific adsorption to the PLGA matrix [15]. Adverse conditions for the encapsulated proteins during release experiments include moisture, acidification of microenvironment, and adsorption of protein to the polymer surface. These are generally accepted as major sources for incomplete release and denaturation [14,17]. 1. Moisture Moisture content in solid proteins is an important component that determines protein stability. The dependence of protein stability on the moisture level is well represented by the bell-shaped relationship shown in Fig. 2. This bell-shaped relationship between the water content and protein aggregation is often attributed to the effect of water on the molecular mobility and concentration of proteins [21]. At low moisture levels, the inactivation event proceeds slowly because mobility
310
Yeo and Park
Figure 2 Bell-shaped relationship between the water content and the aggregation level. (Adapted from Ref. 44–46.)
of the protein is restricted. At high moisture levels, the aggregation rate decreases due to dilution of the protein by water [46,47]. On the other hand, as the moisture level increases from the dried state, water with a very low Tg (ⳮ134⬚C) increasingly reduces the Tg of the protein and thus enhances conformational mobility of the protein molecules (Fig. 3.) [48]. It is at the intermediate molecular levels that protein molecules possess enough conformational freedom while being highly concentrated and thus become liable to aggregation and other degenerative reactions. Thus, maintaining low moisture levels during storage is critical for protein integrity in most dried protein formulations. The rate and extent of water sorption during rehydration of microspheres is often suspected as a denaturation source for the encapsulated proteins. As dried microparticles are rehydrated, the encapsulated protein is exposed to increasing amounts of water. The time scale of this event, however, may be substantially longer as compared with direct reconstitution, due to the low wettability of the polymer matrix [13]. Concerns over aggregation of the encapsulated proteins upon rehydration arise from their longer residence time at the intermediate moisture levels. Few
Figure 3 Plasticization effect of water on protein Tg. (Adapted from Ref. 48.)
Microencapsulation of Protein Drugs
311
studies exist on the effect of moisture on the aggregation of encapsulated proteins separated from other degenerative events occurring during the release period. However, much evidence found in other solid formulations [44–47,49,50] suggest that slow rehydration of microspheres should be responsible for denaturation of proteins to a certain level. A recent study showed that one of the formaldehyde-treated antigens [e.g., formalinized BSA (f-BSA), formalinized ribonuclease A (f-RNase), or tetanus toxoid], which is highly liable to covalent aggregation via a formaldehyde-mediated aggregation pathway (FMAP) in the presence of the intermediate level of moisture, underwent formaldehyde-mediated aggregation within 1 week during the release experiment [51]. Complete release of monomeric proteins was obtained when the aggregation pathway was efficiently inhibited by coencapsulating nucleophile stabilizers such as histidine and glass formers like trehalose. 2. Acidification of Microenvironment Through the years, pH reduction within PLGA microspheres has been one of the major suspects for the instability of encapsulated proteins. Recent trends in research on acidification of the microenvironment and corresponding stabilization strategies were thoroughly reviewed elsewhere with many useful examples [14,17]. PLGA polymers degrade by hydrolytic deesterification to their constituent monomers. An acidic pH environment within the microspheres can be created when these degradation products bearing carboxylic acid end groups are entrapped and accumulated within the polymer matrix (Fig. 4). Acidic pH can cause a number of undesirable events: First, hydrolysis of peptide bonds is often facilitated in acidic conditions. It is known that peptide bonds of aspartic acid are readily cleaved in dilute acid [2]. Second, disulfide exchange can take place by forming a sulfenium cation [2]. Third, some proteins (e.g., BSA) can undergo conformational transition, providing a driving force for noncovalent aggregation via hydrophobic interactions at low pH [52]. Fourth, a low pH environment developed within microspheres can contribute to deamidation and covalent dimerization of insulin [53]. Fifth, it was reported recently that peptides were acylated by lactic acid and glycolic acid units accumulated within degrading microspheres [22,23]. Direct and indirect evidence for acidification of the microenvironment within PLGA microspheres can be found in numerous articles [30,52–63]. For instance, Brunner et al. measured the pH within microspheres loaded with a pH-sensitive spin probe using electron paramagnetic resonance (EPR) and determined the pH to be less than 4.7 [61]. Shenderova et al. used a confocal microscope to visually demonstrate the acidic pH with a pH-sensitive dye. The result showed that the interior of the microspheres became as acidic as pH 2, when the buffer was
Figure 4 Potential pathway of protein denaturation due to acidification of the microenvironment.
312
Yeo and Park
replaced every 3 to 4 days [62]. Similarly, Shao and Bailey encapsulated a pH indicator, bromophenol blue, in PLGA microspheres to show that the pH decreased to ⬃ 3.8 after 3 weeks [53]. Zhu et al. performed experiments to identify the most responsible condition for denaturation of encapsulated protein during in vitro release [52]. Three conditions (elevated moisture, acidic pH, and polymer surface) were individually simulated to induce denatured protein products. The patterns of fragmentation and aggregation of the model protein observed at the acidic pH were comparable to those from the PLGA microenvironment, which suggested that high acidity around pH 2 might be a major source of protein denaturation. Despite the increasing number of publications regarding acidification, there has been a controversy on this issue. Burke [64] reported that the microenvironmental pH within PLGA microspheres prepared by the ProLease威 method (see below) remained constant at neutral pH as measured by 31P NMR spectroscopy. Moreover, Cleland et al. claimed that the pH within the microspheres was comparable to that in the surrounding buffer due to the continuous influx of fresh buffer into the microspheres in reference to indirect evidence: no significant degradation [65] and no significant change in deamidation rate and solubility of recombinant human growth hormone (rhGH) that should have been induced by acidic pH [66]. There is yet another controversy about whether or not this inactivation event experienced in vitro is directly applicable to the in vivo situation. In some way, the local pH at the injection site of microspheres in vivo can be as acidic as in vitro, if fibrous tissue and other inflammatory responses impede the efficient efflux of the polymer degradation products out of the site, as is often the case [54]. On the other hand, the possibility that the acidic oligomers are readily cleared out in vivo cannot be excluded [57]. In fact, when the acidic degradation products were efficiently removed using a dialysis bag and the pH of the release medium was maintained constant at pH of 7.4, denaturation of released and unreleased proteins was not as severe as when the microspheres were incubated in a plastic tube with timed replenishment [54]. 3. Interactions Between Protein and Polymer Proteins tend to adsorb to the polymer surface and undergo structural perturbations [24,67]. Microencapsulated proteins are also expected to have the same behavior leading to protein aggregation [30,35,43,68–73]. Major interactions involved in the protein adsorption can be classified as hydrophobic interactions, hydrogen bonding, and electrostatic interactions [74]. Hydrophobic Interactions. It was suggested in a model study that adsorption of individual protein molecules to the hydrophobic surface could induce aggregation (Fig. 5) [24]. Proteins adsorbed to the surface may undergo conformational changes, exposing their internal hydrophobic moieties. Interactions between these partially denatured monomeric protein molecules would remove the exposed hydrophobic groups before the native conformation could be restored and eventually lead to (noncovalent) aggregation. As evidence of surface adsorption and aggregation, Park et al. showed that sodium dodecyl sulfate (SDS) enhanced the release of encapsulated proteins, which were otherwise unreleasable [32,70,72]. When rhGH-loaded PLGA microspheres were incubated in the medium containing guanidine HCl (GuHCl) or SDS, the extent of extra release was about the same within 1 h in both cases, but greater with SDS after 19-day incubation [32]. Considering that GuHCl is mainly responsible for dissociation of noncovalent aggregation, whereas SDS can reduce both nonspecific adsorption and noncovalent aggregation, the difference in protein release in the later stage of incubation could be explained by the nonspecific adsorption of the protein to the polymer surface. This interpretation was substantiated from the evidence that the surface area of degrading microspheres increased as a function of time [43,75]. Hydrogen Bonding. Hydrogen bonding may account for interactions between proteins (e.g., BSA) and the polymer surface bearing carboxyl groups on the surface [74]. For instance,
Microencapsulation of Protein Drugs
313
Figure 5 Proposed mechanism of insulin aggregation on the polymer–water interface. N, native; U, unfolded. (From Ref. 24.)
the high encapsulation efficiency and incomplete release obtained with the PLGA having free carboxyl end groups (—COOH, Resomer威 RG503H, Boehringer Ingelheim) [71] can support the presence of interactions between protein and free carboxyl moieties in the polymer possibly via hydrogen bonding. Although Resomer RG503H showed a higher degradation rate than Resomer RG503 (having an esterified end group), BSA release was slower from Resomer RG503H. The authors explained that the affinity of the BSA for this polymer hindered its release into the aqueous medium. Electrostatic Interactions. Electrostatic interactions were proposed to explain the difference in the encapsulation efficiencies and release rates obtained with polymers of different molecular weight [76]. It is possible that ionic interactions that may be present between positively charged amino groups in proteins and negatively charged carboxyl end groups in PLGA have increased the encapsulation efficiency and decreased the burst effect and the release rate. The high encapsulation efficiency and the low burst effect obtained with low molecular weight polymer can be explained by the fact that the carboxyl end groups are more abundant in low molecular weight PLGA. Other evidence for ionic interactions could be found from the observation that lysozyme release was significantly increased with an increase in NaCl concentration within the release medium [72]. The lysozyme release enhanced by addition of NaCl was attributed to the disruption of ionic interaction existing between lysozyme and PLGA. However, it was the author’s opinion that the ionic interaction might not fully account for the incomplete protein release, since the addition of high concentrations of NaCl to the medium did not contribute to the elution of unreleased protein, which otherwise should have been removed from PLGA microspheres. Among the above three possible interactions that may exist between encapsulated proteins and PLGA polymer, hydrophobic interactions are generally regarded as the main driving force of adsorption and aggregation causing incomplete release. On the other hand, two results obtained by Zhu et al. provoke an interesting argument that it is the acidic pH rather than adsorption that results in protein denaturation during the release experiment [52]. In a simulation study of BSA adsorption, BSA solutions were incubated with blank microspheres at various pH, but negligible adsorption was recorded. In addition, (acid-induced) aggregates were almost eliminated when
314
Yeo and Park
a basic component was coencapsulated to neutralize the acidic microclimate in the PLGA microspheres. Another question is often raised whether such polymer–protein interactions will indeed play an important role in vivo, where the components of biological fluids may compete with the protein for interaction with the polymer [15]. Despite these controversies, it is safe to assume that rehydration, acidification, and surface adsorption are highly potent sources for protein inactivation. Also, it is likely that these destructive conditions contribute cooperatively to protein instability. For instance, the disorder of the conformational structure caused by the acidic environment may promote nonspecific adsorption of the protein to the hydrophobic polymer surface. The extent to which each condition delivers its effect may depend on the characteristics of the proteins of interest and their liability to the damage.
IV. STABILIZATION APPROACHES A. Preparation 1. Reducing or Avoiding Denaturation at the Water/Organic Solvent Interface Protective Excipients. In previous studies [27,30], it was found that the degree of the interface-induced protein aggregation is influenced by the protein concentrations and the presence of another protein. It indicates that the w/o interface is a saturable space, and thus additional proteins that can compete for adsorption may be effective in prevention of the aggregation. In this context, approaches were made to shield the w/o interface by coencapsulating high concentration of protein [25] or employing protective proteins such as serum albumin or gelatin [29–31,34,57,77,78], although some of them were implemented without clear appreciation of the mechanism. However, one cannot completely rule out the possibility of an immune response toward the aggregates of protective proteins themselves. Alternative approaches devised to avoid the medical complication due to the aggregate of protective proteins include the use of nonproteinic amphiphatic exicipients [37] or sugar excipients [25,78]. Although surfactant approaches have not proved quite successful in practice, coencapsulation of sugars along with proteins appeared to be effective in some cases [25]. For instance, coencapsulation of sugars such as mannitol and trehalose increased stability of encapsulated rhGH [25]. The authors attributed this protective effect to formation of a hydration layer around the protein (preferential hydration or preferential exclusion) that might have reduced the interaction between the protein and the adverse environment. A similar opinion can be found elsewhere [78]. However, this hypothesis does not apply to a very wide range of proteins in microencapsulation [29,34]. It was suggested that the reason should be the absence of surface activity of sugars [29]. Preferential hydration is an experimental phenomenon observed in an aqueous solution of protein and stabilizing excipients, which suggests that there is a deficiency of excipients in the immediate vicinity of the protein relative to the bulk solution and the protein is preferentially hydrated [79]. Conceptually, proteins are stabilized in their native structures as surrounded by the ordered water layers in the presence of sugars, because unfolding of the protein would lead to a greater contact surface between the protein and the solvent and thus a further exclusion event which is entropically more unfavorable (Fig. 6 top) [79]. On the other hand, emulsification creates a huge w/o interface where the surrounding water molecules should be highly ordered (which is also thermodynamically unfavorable). Spontaneous adsorption of protein on this hydrophobic surface of organic solvent (e.g., methylene chloride) can be justified by the entropy gain resulting from displacement of ordered water molecules neighboring both the surface of the organic solvent and the hydrophobic patches
Microencapsulation of Protein Drugs
315
Figure 6 (Top) protein stabilizing mechanism by preferential hydration in an aqueous solution. (Bottom) Surface dehydration mechanism describing the spontaneous protein adsorption to the methylene chloride/ water interface. (From Ref. 29.)
of proteins (Fig. 6. bottom) [29]. Sugars have no surface activity to interfere with this interaction between protein and the organic solvent as opposed to other proteinic excipients. As a result, they are often ineffective in protecting proteins from adsorption and denaturation at the w/o interface. Coencapsulation of protective excipients proved to be effective to some extent; however, complete preservation of the protein integrity was seldom achieved only by those excipients, mostly due to the denaturation events occurring during the release experiment. Semiaqueous or Anhydrous Microencapsulation. Another approach used to avoid destructive w/o interfaces occurring in the primary emulsion stage of the double emulsion method is to employ semiaqueous or anhydrous microencapsulation processes. The rationale and examples of this encapsulation method are excellently reviewed in a recent publication [17]. Briefly, the rationale behind this strategy is that the primary emulsion can be avoided by suspending solid protein in an organic phase, and proteins would be more stable in the absence of water (due to the decreased mobility) [17]. One representative example is to use the solid oil water (s/o/w) [25,80] or the solid oil oil (s/o/o) double emulsion method [33,81–84]. Although some of them showed desirable in vitro release profiles without any stability complications [80,84], residual amounts of the solvent or the continuous phase and the initial burst effect due to the proteins present in the surface are considered as issues in the s/o/o method. Disadvantages of the s/o/w method are the low encapsulation efficiency and the loss of stabilizing excipients due to their diffusion into the water phase in the o/w stage [17]. The cryogenic solvent extraction method (the ProLease technology developed by Alkermes, Inc.) is another example of the anhydrous encapsulation technique. In this method, solid protein was suspended in a PLGA solution in methylene chloride and sprayed over liquid nitrogen using an ultrasonic nozzle to form microparticles. Methylene chloride was then removed by extraction in cold ethanol [85–89]. The success of this approach led to the first commercial
316
Yeo and Park
microparticle product of a protein drug (Nutropin Depot威 by Genentech, Inc., developed for controlled release of rhGH from PLGA microspheres). Another approach to eliminate the w/o emulsion step is the in situ forming polymer depot system. In this case, a biodegradable polymer is dissolved in a pharmaceutically acceptable water-miscible solvent, and solid proteins are suspended in it. This polymer solution is solidified into the matrix type depot when it comes in contact with water from an aqueous buffer or physiological fluid. Dimethyl sulfoxide (DMSO), N-methyl-2-pyrrolildone (NMP) [90,91], triacetin, ethyl benzoate, and benzyl benzoate [92–94] are water-miscible solvents used for this purpose. Sustained in vivo efficacy up to 3 months was achieved [91]; however, studies on protein stability in this system are virtually nonexistent. 2. Alternative Solvents It has been known that more hydrophilic organic solvents are in general less destructive to protein integrity [39,95]. For example, denaturation of lysozyme dissolved in organic solvents increased with an increase in solvent hydrophobicity [39]. Similar observations were made in the investigation of the catalytic activity of hydroxynitrile lyase (Hnl) at the water/organic solvent interface [95]. Hnl adsorption to the interface was stronger with more apolar solvents, and the loss of its catalytic activity was faster despite higher initial activity. Evidence has been found in the protein microencapsulation area as well. For instance, many proteins were more stable and bioactive when encapsulated with ethyl acetate instead of methylene chloride [25,26,96,97]. In addition to the potential instability issue, the use of conventional chlorinated solvents such as methylene chloride is not desirable from the viewpoint of environmental and human safety considerations [98]. Since methylene chloride is a suspected carcinogen and mutagen, the concern and strict regulation over the residual solvent from the regulatory authority may pose a considerable challenge in process development. FDA guidelines for the residual solvents limit the permitted daily exposure (PDE) to 6 mg per day for methylene chloride [99]. On the other hand, ethyl acetate is considered to be safer and more protein friendly than methylene chloride, and thus has received increasing attention as an alternative solvent [25,26,96–98,100–107]. Most cases reported higher recovery of native protein using ethyl acetate as compared to methylene chloride [25–26,96,97,100]. 3. Reducing Physical Stresses The cryogenic solvent extraction method is a good example of minimizing the shear stress, by eliminating the emulsification step and the temperature effect by maintaining very low temperatures during encapsulation [85–89]. In combination with the anhydrous encapsulation strategy, this technology resulted in successful preservation of rhGH integrity throughout the process (Table 2). Table 2 Integrity of rhGH Before and After Encapsulation by the ProLease Process Protein characterization assay Size exclusion (%) Reverse-phase HPLC (%) Anion-exchange HPLC (%) SDS-PAGE (%) Specific bioactivity (IU/mg) Source: Ref. 89.
Before encapsulation
After encapsulation
99.4 97.0 99.0 ⱖ 99 2.8
97.5 ⫾ 0.4 98.9 ⫾ 1.2 97.7 ⫾ 0.7 ⱖ 99 2.7 ⫾ 0.2
Microencapsulation of Protein Drugs
317
Recently the supercritical fluid precipitation method has been used for microencapsulation. This method can be broadly categorized as two processes: rapid expansion of supercritical solutions (RESS) [108–111] and supercritical antisolvent crystallization (SAS) [112,113]. Taking advantage of two distinctive properties of supercritical fluids (i.e., high compressibility and liquidlike density), the method employs supercritical fluid like CO2 as a solvent for the polymer (RESS) or as an antisolvent that causes polymer precipitation (SAS). Relatively mild process temperatures (35–45⬚C) [111] and absence of the emulsification step are ideal features that make this method attractive for protein encapsulation; however, protein delivery application of this technology is still in its infancy. Recent reviews on applications of the supercritical technology for particle production are available in the literature [114,115].
B. Protein Release 1. Modifying Moisture Level A rational strategy to avoid moisture-induced aggregation is to manipulate the water content in the encapsulated protein to be outside the moisture window, causing maximal aggregation during rehydration [14,44]. Hydrophobic additives, such as poloxamer L101, L121, or ethyl stearate, were coencapsulated along with tetanus toxoid to lower the water content of the microparticles during release [78]. Another similar approach is to reduce the aqueous solubility of protein using ‘‘salting-out salts’’ such as ammonium sulfate or forming a metal complex [116]. In fact, rhGH was effectively stabilized during a month-long release when encapsulated as an insoluble zinc complex [85–89]. In addition, the reduced solubility contributed to minimizing the initial burst and achieving the target duration of release. Once this approach proved to be successful, recombinant human insulin-like growth factor-I (rhIGF-I) [117] and recombinant human nerve growth factor (rhNGF) [105] were stabilized and microencapsulated in the same manner. The water-insoluble zinc–protien complex and zinc carbonate, which were coencapsulated as a zinc reservoir, might have delayed hydration of the PLGA microparticles [89,117] and thus maintained the moisture content at a relatively low level. Perez et al. suggested increasing the hydrophobicity of the polymer to limit the moisture uptake in the microsphere [17]; however, the overall efficiency is questionable because more hydrophobic polymer would result in stronger adsorption of the protein to the polymer surface. Increasing water uptake in the polymer matrix can also be considered for the same purpose. A recent publication reported that 10% sucrose, when encapsulated along with BSA within a PLGA microcylinder, nearly doubled the water content in the polymer (i.e., 46% water with sucrose vs. 24% without sucrose after 1 week of incubation) and improved BSA stability (51% aggregates with sucrose and 81% without sucrose after 4 weeks of incubation at 37⬚C) [118]. It is likely that the increased water content has contributed to diluting the reactive species, such as unfolded proteins or acidic degradation products, and facilitating their efflux. The stabilization effect of Mg(OH)2, presumably afforded by neutralizing the acidic pH in degrading PLGA polymer (to be introduced in the next section), can be ascribed in part to the enhanced water uptake [118,119]. 2. Counteracting Acidification Recent trends in counteracting acidification developed within degrading microparticles can be classified into three basic strategies: (1) coencapsulating excipients to neutralize the acids formed by PLGA hydrolysis, (2) facilitating the escape of the water-soluble hydrolytic products of
318
Yeo and Park
PLGA polymer by increasing its permeability, and (3) modifying degradation characteristics of the polymer [14]. Coencapsulating Antacid Excipients. As a direct relief to the pH drop, inorganic basic salts such as Mg(OH)2 have been exploited and were found to be effective in stabilization of BSA, basic fibroblast growth factor (bFGF), and bone morphogenetic protein-2 (BMP-2) within PLGA microcylinders [52,119]. BSA release increased in proportion to the amount of added Mg(OH)2, and the aggregation was virtually eliminated in the presence of 3% Mg(OH)2 [52]. Improvement of the antigenic stability of toxoids achieved by coencapsulation of BSA can also be attributed to the role of BSA as a proton scavenger or a proton sink [59,78]. However, this conclusion is rather controversial since the other study reported no sign of buffering effect from coencapsulated BSA [52]. Increasing the Permeability of the Polymer Matrix. Increasing the permeability or porosity of the microparticles can enhance the efflux of the degradation products and avoid the build-up of acids. A blend of poly(D,L-lactic acid) (PLA) and water-soluble poly(ethylene glycol) (PEG) was used to make relatively permeable microparticles [83]. Stability of the protein remaining in the microparticles was significantly enhanced with the increasing amount of PEG in the blend, although the release control function of the polymer was somewhat compromised as a result of the increased permeability [83]. In fact, the controversy on the acidity of the microenvironment stems in part from enhanced diffusion of degradation products out of highly porous microparticles produced by the ProLease威 method [14]. Modifying the Polymer Degradation Characteristics. Another strategy to avoid the acid builid-up is to reduce production of acidic species by using slowly degrading PLA [83,120]. PLA has a much slower degradation rate than PLGA due to its higher hydrophobicity and the steric hindrance to the water attack on the ester bond. Indeed, a relatively neutral pH was maintained in the release medium for PLA formulations, as opposed to PLGA formulations, which caused a dramatic pH drop after a 4-week incubation [83]. Alternatively, poly(ε-caprolactone) (PCL), which is also a biocompatible and biodegradable polyester undergoing delayed degradation, may not generate an environment as acidic as PLGA polymers [121,122]. In addition to these polyesters, alternative polymer candidates include polyanhydrides and poly(ortho esters) (POE), which degrade by surface erosion and thus may cause less changes in the internal pH [12,123]. 3. Preventing Adsorption Modifying the Polymer Surface Hydrophilicity. Surface adsorption of proteins via hydrophobic interactions can be reduced by altering the hydrophilicity of the polymer surface [124]. Early efforts in minimizing protein adsorption to the polymer surface include covering hydrophobic polymer surface with a layer of hydrophilic polymer such as polyethylene oxide (PEO) [125] to attain steric resistance to protein adsorption. A similar approach is physical blending of PLGA with other hydrophilic polymers (e.g., PEO) [126,127]. More recently, block copolymers including hydrophilic segments [e.g., monomethoxy poly(ethylene oxide) (MPEO)PLA, PLGA-PEO-PLGA, POE-PEO-POE] were introduced to make amphiphilic microspheres [123,124,128–133]. A few approaches resulted in significant reduction in protein adsorption and increase in the amount of the released protein [124,128,129]. The advantages of the amphiphilic copolymers may also be due to the formation of a swollen hydrogel-like structure with high water content that could enhance the material exchange between the release buffer and degrading microparticles [128,130,133,134]. Pre-entrapment of the Protein in Hydrophilic Core. It was proposed to protect the protein by entrapping them in hydrogel particles prior to PLGA encapsulation [81,82,135–138].
Microencapsulation of Protein Drugs
319
One of the advantages of this approach is claimed to be that the hydrophilic layer can isolate proteins from the surrounding polymer matrix and thus prevent deleterious denaturation events caused by the hydrophobic microenvironment. Subscribing to this strategy, Wang et al. have shown that PLGA–poly(vinyl alcohol) (PVA) composite nanoparticles provided prolonged BSA release over 50 days without any complications related to protein structure [137]. Adsorption Competitor. Protein release from microparticles was increased as Poloxamer 188 was coencapsulated [71]. The enhanced release was attributed to the reduced interaction between the protein (e.g., BSA) and PLGA in the presence of the surfactant.
V. FURTHER CONSIDERATIONS Table 3 summarizes common inactivation sources of encapsulated proteins and representative stabilization strategies discussed in the previous section. Although remarkable improvement in the microencapsulation of proteins has been achieved, there is still ample room for improvement. The issues for further consideration are described below. First, since a large amount of protein is enclosed within the polymeric microparticles for long-term delivery, encapsulated protein would exist as a highly concentrated solution from the encapsulation stage and/or after hydration during release incubation. Proteins are in general prone to aggregation at high concentrations [18]. Moreover, the protein concentration can influence the chemical stability in some cases. For example, it was shown that concentrated BSA (100 mg/ mL) suffered from rapid covalent aggregation through thiol-disulfide interchange [70]. The need for carrying relatively high concentrations of protein at the physiological temperature for a long period is a major challenge in the development of protein-loaded microparticle systems.
Table 3 Common Inactivation Sources of Encapsulated Proteins and Stabilization Strategies Inactivation source Preparation Exposure to w/o interfaces
Exposure to the hydrophobic organic solvent Physical stresses during preparation Protein release Intermediate moisture level during hydration
Acidification of the microenvironment
Interaction between protein and polymer
Stabilization strategy Reduce or avoid denaturation at w/o interfaces Protective excipients Anhydrous microencapsulation process Alternative solvents Reduce physical stresses Modify moisture level Reduce water content in the microparticles Increase water uptake Counteract acidification Coencapsulate antacid excipients Increase permeability of the polymer matrix Modify degradation characteristics of the polymer Prevent adsorption Modify hydrophilicity of the polymer surface Preentrapment of protein in the hydrophilic core Adsorption competitor
320
Yeo and Park
Second, most of the protective excipients encapsulated along with proteins for various purposes are hydrophilic and of low molecular weight (e.g., sugars). They can be lost to the outer aqueous phase during the hardening process [139] or preferably removed in the early stage of release. As a result, these excipients may not be able to afford protein stabilization during process and/or throughout the release experiments [13,17,34]. In this regard, it may be necessary to control the release of the protective excipients as well as the protein. In the ProLease威 technique for instance, an insoluble zinc salt (zinc carbonate) was coencapsulated as a reservoir for zinc in order to provide additional zinc needed to maintain the protein in a zinc complex form [87,89]. Proteins released faster when the zinc carbonate salt was not included in the microspheres [87,105]. This suggests that the release of coencapsulated excipients should also be controlled to exert any protective effect on the protein throughout the release period. Third, it is highly possible that several mechanisms listed above may cooperatively participate in the denaturation events of the encapsulated protein. As mentioned previously, the shear stress could facilitate the protein aggregation to a greater extent when coupled with other adverse conditions [26]. In another example, it was shown that the antigenicity loss of the tetanus toxoid was synergistically increased when two factors (temperature and acidity) were combined [59]. The improvement that can be expected from bypassing one or two sources may be therefore limited.
VI. A NOVEL APPROACH: THE SOLVENT EXCHANGE METHOD In preparation of microspheres by conventional microencapsulation processes, such as solvent extraction/evaporation techniques, a number of factor are known to affect the stability of loaded protein drugs, loading efficiency, and release properties. In particular, the exposure of protein drugs to the w/o interface, the acidification of the microenvironment inside the PLGA microspheres, and adsorption of protein drugs to the PLGA matrices are known to be critical. In an attempt to provide an alternative microencapsulation method that can accommodate most stabilization strategies described in the previous sections, recently we have developed a novel microencapsulation method, which we call the solvent exchange method [140–142]. A. Background 1. Hypotheses The solvent exchange method is a simple way to produce mononuclear microcapsules in which a single hydrophilic core containing drug molecules is surrounded by a thin biodegradable polymer membrane (Fig. 7). The rationale behind this novel approach is that the integrity of the encapsulated proteins can be preserved by the following strategies.
Figure 7 Diagram of the mononuclear microcapsule produced by the solvent exchange method.
Microencapsulation of Protein Drugs
321
Preparation. 1. Microparticles can be produced without emulsification by a variety of methods (e.g., ink-jet nozzle, atomization in the air). Thus, invasive w/o interfaces or prolonged exposure of proteins to organic solvents can be avoided. 2. Complications associated with hydrophobicity of the organic solvents (e.g., methylene chloride) can be overcome by using more hydrophilic solvents which are known to be rather benign to protein stability. 3. Physical stresses created during emulsification may be avoided using alternative systems that can produce microdroplets with low impact. Protein release. 1. Hydrophilicity of the drug-loaded core may allow easy hydration of the encapsulated drug and quickly increase the moisture content inside the microcapsules above the intermediate level. 2. The hydrophilic core leaves sufficient room for protective excipients including antacids or buffer substances. In addition, the PLGA polymer remains as a layer around the aqueous core; therefore the degradation products of the polymer are likely to diffuse into the release medium rather than to invade the structured aqueous core where protein is loaded. 3. In mononuclear microcapsules, only a minimal amount of protein drugs are in contact with the hydrophobic polymer, and thus irreversible protein adsorption to the polymer surface can be minimized. 2. Concept It is common to observe that a solid polymer layer forms at the interface between two solutions when the polymer solution comes in contact with the aqueous surface (Fig. 8A). The solvent exchange method is based on this simple phenomenon that a water-insoluble polymer such as PLGA dissolved in the organic solvent is phase separated from the solvent by only a small exchange of solvents (i.e., solvent exchange), leading to a decrease in the solvent quality. Mononuclear microcapsules can be produced by localizing the solvent exchange on the surface of aqueous microdroplets (Fig. 8(B)). To this end, the organic solvent should be able to favorably spread out on the aqueous surface to cover the entire microdroplet instantly, and be miscible with water to a certain degree to allow instant phase separation of the polymer thereafter. Any kind of polymer can be used to make a release-controlling shell, but PLGA is the preferable one considering its wide applications in drug delivery formulations. The aqueous core serves as a drug reservoir and plays an important role in maintaining the physical integrity of finished microcapsules. One of the methods to apply the polymer solution to the aqueous droplets is to produce droplets of aqueous and PLGA solutions and induce their collision in the air. Microdroplets of aqueous and PLGA solutions can be generated by various methods, e.g., ink-jet nozzles and ultrasonic nozzles [140–142]. B. The Solvent Exchange Method 1. Selection of the Organic Solvent Since the key attribute of the method is the solvent exchange at the water solvent interface in relation to spreading of the solvent over the aqueous surface, the selection of an appropriate solvent is of utmost importance. The organic solvent for the solvent exchange method should meet the following requirements:
322
Yeo and Park
Figure 8 Schematic of the formation of the PLGA membranes through the solvent exchange at the interface between the aqueous and the PLGA solution: (A) on an aqueous film; (B) on an aqueous droplet. (From Ref. 140.)
1. The solvent should possess the right solvency for the polymer (e.g., PLGA). 2. The solvent should be able to favorably spread out on the aqueous surface. 3. The solvent should be miscible with water to a certain degree, thus allowing fast phase separation of the polymer. Solvent screening can begin with the Hildebrand solubility parameters of the organic solvents. The solubility parameter of PLGA (lactic acid/glycolic acid 58:42) is known to be 16.8–18.7 MPa1/2, depending on how it was determined [143]. In our study, it was observed that solubility parameters of good solvents for PLGA (lactic acid/glycolic acid 50:50) are roughly in the range of 16–24 MPa1/2. However, it was not always successful to predict the solvency based on the Hildebrand parameter only. Among the 60 solvents in this range we have tested, only half of them were able to produce clear PLGA solutions, and the rest were poor solvents or marginally able to swell the PLGA. For better prediction of the polymer solubility, Hansen’s solubility parameters [144] were used. Hansen’s multicomponent solubility parameters are described as
δ 2 = δ d2 + δ p2 + δ h2
(1)
where ␦ is the total solubility parameter and ␦d, ␦p, and ␦h are the contributions from dispersion forces, polar interactions, and hydrogen bonding, respectively. A triangular plotting technique of the Hansen parameters (which are available in the literature [146,147]) developed by Teas [145] provides a graphic tool for the solvency prediction. In the Teas method, fractional parameters are calculated as
fi =
100 δ i δd + δp + δ h
where i = d, p, or h.
(2)
where i ⳱ d, p, or h. When solvents were plotted by using fractional parameters on a triangular graph, good solvents for PLGA were localized in a region of relatively low fp and fh but of high
Microencapsulation of Protein Drugs
323
fd, as shown in Fig. 9. This triangular graph can be used as a quick selection tool for solvents meeting the first requirement. The second and third qualities of the solvent can be simultaneously examined by a simple screening tool called the hydrogel layer method. Briefly, a library of PLGA solutions was constructed using the good solvents selected above. A drop of each solution was placed on an aqueous gel layer containing 0.5% agarose. The diameter of the polymer film that formed on the gel was measured, and the turbidity of the film developed after 1 min was measured at 620 nm. The diameter of the film reflects the degree of spreading of each solvent, and the turbidity allows evaluation of the quality of the polymer membrane (Fig. 10). A solvent which forms a membrane of relatively large diameter is preferable, since it means that the polymer solution spreads spontaneously. In addition, polymer membranes displaying a relatively low optical density are preferred since the high turbidity reflects formation of a rough and discontinuous precipitate, which is likely to fail in controlling the mass transfer across the membrane. According to the results obtained to date, the spreading capability of the polymer solution is related to the surface tension of the solvent, and the quality of the membrane is closely linked to the solubility of the solvent in water. For instance, solvents that have medium solubility in water (5–60 % w/w) result in a transparent and dense membrane, whereas highly water-miscible solvents (ⱖ100 % w/w) make a turbid and discontinuous membrane. Therefore, it may be possible to make a rough judgment in solvent selection by looking at the surface tension and the water solubility of the solvent. When the solvents were screened with the selection rules described above, ethyl acetate was chosen as one of the best solvents for the solvent exchange method. The basic parameters examined in the screening are listed in Table 4. Ethyl acetate was used as a polymer solvent in the subsequent studies. As a matter of fact, relatively hydrophilic organic solvents such as ethyl acetate have been used with increasing popularity as an alternative to methylene chloride
Figure 9 Plot of various solvents as a function of the fractions of their dispersion force (Fd), polarity (Fp), and hydrogen bonding (Fh). (●) Good solvent; (䉱, 왕) intermediate solvent; and (⳯) poor solvent.
324
Yeo and Park
Figure 10 Optical density as a measure of the solvent quality (left) and the film diameter reflecting the degree of spreading (right).
[25,26,96–98,100–107]. However, its relatively high solubility in water often made it difficult to produce dense and regular microspheres by the conventional emulsion methods [98,101,106], and sometimes required doping the continuous phase with additional ethyl acetate in order to delay diffusion of the solvent out of the discontinuous phase [104,106]. When the polymer solution in ethyl acetate is emulsified into an aqueous medium, ethyl acetate diffuses out immediately until it attains equilibrium with the external phase and leads to instant formation of porous microspheres. For this reason, the morphology of microspheres is far more dependent on the phase ratio (the volume ratio between the organic and aqueous phase) when ethyl acetate is used in the conventional emulsion method. The hydrophilicity of the solvent caused the same problem in defining regular microspheres when methyl ethyl ketone was used [148,149]. On the other hand, the solvent exchange method takes advantage of the hydrophilicity of the organic solvent to produce microcapsules. 2. Selection of the Aqueous Core The aqueous core is another component that initiates the phase separation of the polymer and a reservoir that houses the drug molecules and the protective excipients (if necessary). Moreover, it can play a major role in maintaining the mechanical strength of the microcapsules. In this
Table 4 Parameters Used for Solvent Screening—Ethyl Acetate Example Ethyl acetate (CH3COOC2H5) Hildebrand solubility parameter [147] Hansen multicomponent parameters [146] Solubility in water [147] Surface tension [147] Optical density of the PLGA film (at 620 nm) Film diameter
18.6 MPa1/2 fd/fp/fh ⫽ 56/19/25 8% w/w 23.8 dynes/cm 0.0438 16.3 mm
Microencapsulation of Protein Drugs
325
regard, it is preferred to use an aqueous base which is capable of the sol-to-gel transformation. This unique property present in some polysaccharides and hydrophilic polymers makes it possible to process the drug formulation in a liquid state until particle formation and to solidify the particles later by providing appropriate conditions. For example, we have been using sodium alginate as an aqueous base which forms an ionotropic hydrogel in the presence of divalent cations such as calcium. 3. Microencapsulation Another key aspect of this method is to apply the polymer solution on the surface of the aqueous microdroplets in the air causing an initial phase separation of the polymer membrane and to collect the microcapsules in a liquid bath to complete the phase separation. The polymer solution can be provided as microdroplets in the vicinity of the aqueous microdroplets. Upon collision of the two droplets, a mononuclear microcapsule (Fig. 7) is produced as a function of their physicochemical properties described above. Any kind of nozzle that can produce microdroplets in the air can be utilized for this purpose; we have opted to use a two ink-jet nozzle assembly [140–142] or ultrasonic nozzles considering their mildness. Thus far we have confirmed that these nozzle systems are able to produce microcapsules with high efficiency.
VII. CONCLUSIONS The first part of this review focused on understanding the common inactivation sources of encapsulated proteins and different approaches for stabilizing the encapsulated proteins. Incomplete release, protein denaturation, and the failure to control the initial burst have been the most commonly encountered problems in protein microencapsulation. Physical and chemical sources for these undesirable phenomena have been identified in many aspects of conventional microencapsulation processes. Efforts to address these problems led to the first commercialization of the protein microsphere product in recent years [89], and recent publications began to report that desirable release kinetics without serious stability complications could be achieved at least for in vitro studies [65,80,105,107,117,120,150]. However, complete release of fully bioactive proteins is still a difficult task in most cases for the following reasons. First, individual proteins have unique stability profiles and thus often require tailoring of the stabilization strategy. Second, several denaturation pathways can cooperatively participate in denaturation of the encapsulated proteins. There is a great need to design a new microencapsulation technique that can address many sources of the problems. The solvent exchange method is a new approach that can potentially eliminate many problems associated with the conventional microencapsulation methods.
REFERENCES
1. Nail SL. Formulation of proteins and peptides. In: Park K, ed. Controlled Drug Delivery Challenges and Strategies. Washington DC: ACS Professional Reference Book, 1997:185–203. 2. Manning MC, Patel K, Borchardt RT. Stability of protein pharmaceuticals. Pharm. Res 1989; 6: 903–918. 3. Langer R, Folkman J. Polymers for the sustained release of proteins and other macromolecules. Nature 1976; 263:797–800.
326
Yeo and Park
4. Ogawa Y, Yamamoto M, Takada S, Okada H, Shimamoto T. Controlled release of leuprolide acetate from PLA or PLGA microcapsules: influence of molecular weight and copolymer ratio of polymer. Chem. Pharm. Bull 1988; 36:1502–1507. 5. Ogawa Y, Yamamoto M, Okada H, Yashiki T, Shimamoto T. A new technique to efficiently entrap leuprolide acetate into microcapsules of polylactic acid or copoly(lactic/glycolic) acid. Chem. Pharm. Bull 1988; 36:1095–1103. 6. Ogawa Y, Okada H, Yamamoto M, Shimamoto T. In vivo release profiles of leuprolide acetate from microcapsules prepared with polylactic acids or copoly(lactic/glycolic) acids and in vivo degradation of these polymers. Chem. Pharm. Bull 1988; 36:2576–2581. 7. Okada H, Heya T, Ogawa Y, Shimamoto T. One-month release injectable microcapsules of luteinizing hormone-releasing hormone agonist (leuprolide acetate) for treating experimental endometriosis in rats. J. Pharmacol. Exp. Ther 1988; 244:744–750. 8. Okada H, Inoue Y, Heya T, Ueno H, Ogawa Y, Toguchi H. Pharmacokinetics of once-a-month injectable microspheres of leuprolide acetate. Pharm. Res 1991; 8:787–791. 9. Garrett RH, Grisham CM. Biochemistry. 2nd ed.: Saunders College Publishing, 1999. 10. Pace CN. Conformational stability of globular proteins. Trends Biochem. Sci 1990; 15:14–17. 11. Jaenicke R. Protein folding: local structures, domains, subunits, and assemblies. Biochemistry 1991; 30:3147–3161. 12. Putney SD, Burke PA. Improving protein therapeutics with sustained-release formulations. Nat. Biotechnol 1998; 16:153–157. 13. Schwendeman SP, Cardamone M, Klibanov A, Langer R, Brandon MR. Stability of proteins and their delivery from biodegradable polymer microspheres. In: Cohen S, Bernstein H, eds. Microparticulate systems for the delivery of proteins and vaccines. New York: Marcel Dekker, 1996:1–49. 14. Schwendeman SP. Recent advances in the stabilization of proteins encapsulated in injectable PLGA delivery systems. Crit. Rev. Ther. Drug Carrier Syst 2002; 19:73–98. 15. Weert M, Hennink WE, Jiskoot W. Protein instability in PLGA microparticles. Pharm. Res 2000; 17:1159–1167. 16. Yeo Y, Baek N, Park K. Microencapsulation methods for delivery of protein drugs. Biotechnol. Bioprocess Eng 2001; 6:213–230. 17. Perez C, Castellanos IJ, Costantino HR, Al-Azzam W, Griebenow K. Recent trends in stabilizing protein structure upon encapsulation and release from bioerodible polymers. J. Pharm. Pharmacol 2002; 54:301–313. 18. Wang W. Instability, stabilization, and formulation of liquid protein pharmaceuticals. Int. J. Pharm 1999; 185:129–188. 19. Cleland JL, Langer R. Formulation and delivery of proteins and peptides: design and development strategies. In: Cleland JL, Langer R, eds. Formulation and Delivery of Proteins and Peptides. Washington, D.C.: ACS, 1994. 20. Schwendeman SP, Constantino HR, Gupta RK, Langer R. Peptide, protein, and vaccine delivery from implantable polymeric systems: progress and challenges. In: Park K, ed. Controlled Drug Delivery Challenges and Strategies. Washington DC: ACS Professional Reference Book, 1997: 229–267. 21. Lai MC, Topp EM. Solid-state chemical stability of proteins and peptides. J. Pharm. Sci 1999; 88: 489–500. 22. Lucke A, Kiermaier J, Gopferich A. Peptide acylation by poly(␣-hydroxy esters). Pharm. Res 2002; 19:175–181. 23. Lucke A, Fustella E, Tessmar J, Gazzaniga A, Gopferich A. The effect of poly(ethylene glycol)–poly(d,L-lactic acid) diblock copolymers on peptide acylation. J. Controlled Release 2002; 80:157–168. 24. Sluzky V, Klibanov AM, Langer R. Mechanism of insulin aggregation and stabilization in agitated aqueous solutions. Biotechnol. Bioeng 1992; 40:895–903. 25. Cleland JL, Jones AJS. Stable formulations of recombinant human growth hormone and interferongamma for microencapsulation in biodegradable microspheres. Pharm. Res 1996; 13:1464–1475. 26. Sah H. Protein instability toward organic solvent/water emulsification: implications for protein microencapsulation into microspheres. PDA J. Pharm. Sci. Technol 1999; 53:3–10.
Microencapsulation of Protein Drugs
327
27. Sah H. Protein behavior at the water/methylene chloride interface. J. Pharm. Sci 1999; 88: 1320–1325. 28. Sah H. Connection between interfacial tension kinetcs and protein instability at water/organic solvent interface&, San Diego, CA, 2001. 29. Sah H. Stabilization of proteins against methylene chloride/water interface-induced denaturation and aggregation. J. Controlled Release 1999; 58:143–151. 30. Lu W, Park TG. Protein release from poly(lactic-co-glycolic acid) microspheres: protein stability problems. PDA J. Pharm. Sci. Technol 1995; 49:13–19. 31. Uchida T, Shiosaki K, Nakada Y, Fukada K, Eda Y, Tokiyoshi S, Nagareya N, Matsuyama K. Microencapsulation of hepatitis B core antigen for vaccine Preparation. Pharm. Res 1998; 15: 1708–1713. 32. Kim HK, Park TG. Microencapsulation of human growth hormone within biodegradable polyester microspheres: protein aggregation stability and incomplete release mechanism. Biotechnol. Bioeng 1999; 65:659–667. 33. Iwata M, Tanaka T, Nakamura Y, McGinity JW. Selection of the solvent system for the preparation of poly(D,L-lactic-co-glycolic acid) microspheres containing tumor necrosis factor-alpha (TNF-␣). Int. J. Pharm 1998; 160:145–156. 34. Morlock M, Koll H, Winter G, Kissel T. Microencapsulation of rh-erythropoietin, using biodegradable PLGA: protein stability and the effects of stabilizing excipients. Eur. J. Pharm. Biopharm 1997; 43:29–36. 35. Zambaux MF, Bonneaux F, Gref R, Dellacherie E, Vigneron C. Preparation and characterization of protein C–loaded PLA nanoparticles. J. Controlled Release 1999; 60:179–188. 36. Perez C, Griebenow K. Improved activity and stability of lysozyme at the water/methylene chloride interface: enzyme unfolding and aggregation and its prevention by polyols. J. Pharm. Pharmacol 2001; 53:1217–1226. 37. Weert M, Hoechstetter J, Hennink WE, Crommelin DJA. The effect of a water/organic solvent interface on the structural stability of lysozyme. J. Controlled Release 2000; 68:351–359. 38. Sah H, Choi S-K, Choi H-G, Yong C-S. Relation of dynamic changes in interfacial tension to protein destabilization upon emulsification. Arch. Pharmacal Res 2002; 25:381–386. 39. Knubovets T, Osterhout JJ, Klibanov AM. Structure of lysozyme dissolved in neat organic solvents as assessed by NMR and CD spectroscopies. Biotechnol. Bioeng 1999; 63:242–248. 40. Griebenow K, Klibanov AM. On protein denaturation in aqueous-organic mixtures but not in pure organic solvents. J. Am. Chem. Soc 1996; 118:11695–11700. 41. Zaks A, Klibanov AM. The effect of water on enzyme action in organic media. J. Biol. Chem 1988; 263:8017–8021. 42. Griebenow K, Vidal M, Baez C, Santos AM, Barletta G. Nativelike enzyme properties are important for optimum activity in neat organic solvents. J. Am. Chem. Soc 2001; 123:5380–5381. 43. Crotts G, Park TG. Protein delivery from poly(lactic-co-glycolic acid) biodegradable microspheres: release kinetics and stability issues. J. Microencapsulation 1998; 15:699–713. 44. Schwendeman SP, Constantino HR, Gupta RK, Siber GR, Klibanov AM, Langer R. Stabilization of tetanus and diphtheria toxoids against moisture-induced aggregation. Proc. Natl. Acad. Sci. USA 1995; 92:11234–11238. 45. Costantino HR, Langer R, Klibanov AM. Aggregation of a lyophilized pharmaceutical protein, recombination human albumin: effect of moisture and stabilization by excipients. Bio/Technology 1995; 13:493–496. 46. Liu WR, Langer R, Klibanov AM. Moisture-induced aggregation of lyophilized proteins in the solid state. Biotechnol. Bioeng 1991; 37:177–184. 47. Costantino HR, Langer R, Klibanov AM. Solid-phase aggregation of proteins under pharmaceutically relevant conditions. J. Pharm. Sci 1994; 83:1662–1669. 48. Ahlneck C, Zografi G. The molecular basis of moisture effects on the physical and chemical stability of drugs in the solid state. Int. J. Pharm 1990; 62:87–95. 49. Costantino HR, Schwendeman SP, Griebenow K, Klibanov AM, Langer R. The secondary structure and aggregation of lyophilized tetanus toxoid. J. Pharm. Sci 1996; 85:1290–1293.
328
Yeo and Park
50. Costantino HR, Langer R, Klibanov AM. Moisture-induced aggregation of lyophilized insulin. Pharm. Res 1994; 11:21–29. 51. Jiang W, Schwendeman SP. Stabilization of a model formalinized protein antigen encapsulated in poly(lactide-co-glycolide)-based microspheres. J. Pharm. Sci 2001; 90:1558–1569. 52. Zhu G, Mallery SR, Schwendeman SP. Stabilization of proteins encapsulated in injectable PLGA. Nat. Biotechnol 2000; 18:52–57. 53. Shao PG, Bailey LC. Porcine insulin biodegradable polyester microspheres: stability and in vitro release characteristics. Pharm. Dev. Technol 2000; 5:1–9. 54. Park TG, Lu W, Crotts G. Importance of in vitro experimental conditions on protein release kinetics, stability and polymer degradation in protein encapsulated PLGA microspheres. J. Controlled Release 1995; 33:211–222. 55. Mader K, Gallez B, Liu KJ, Swartz HM. Non-invasive in vivo characterization of release processes in biodegradable polymers by low-frequency electron paramagnetic resonance spectroscopy. Biomaterials 1996; 17:457–461. 56. Xing DK-L, Crane DT, Bolgiano B, Corbel MJ, Jones C, Sesardic D. Physicochemical and immunological studies on the stability of free and microsphere-encapsulated tetanus toxoid in vitro. Vaccine 1996; 14:1205–1213. 57. Chen L, Apte RN, Cohen S. Characterization of PLGA microspheres for the controlled delivery of IL-1␣ for tumor immunotherapy. J. Controlled Release 1997; 43:261–272. 58. Igartua M, Hernandez RM, Esquisabel A, Gascon AR, Calvo MB, Pedraz JL. Stability of BSA encapsulated into PLGA microspheres using PAGE and capillary electrophoresis. Int. J. Pharm 1998; 169:45–54. 59. Johansen P, Merkle HP, Gander B. Physico-chemical and antigenic properties of tetanus and diphtheria toxoids and steps towards improved stability. Biochim. Biophys. Acta 1998; 1425:425–436. 60. Shao PG, Bailey LC. Stabilization of pH-induced degradation of porcine insulin in biodegradable polyester microspheres. Pharm. Dev. Technol 1999; 4:633–642. 61. Brunner A, Mader K, Gopferich A. pH and osmotic pressure inside biodegradable microspheres during erosion. Pharm. Res 1999; 16:847–853. 62. Shenderova A, Burke TG, Schwendeman SP. The acidic microclimate in poly(lactide-co-glycolide) microspheres stabilizes camptothecins. Pharm. Res 1999; 16:241–248. 63. Fu K, Pack DW, Klibanov AM, Langer R. Visual evidence of acidic environment within degrading poly(lactic-co-glycolic acid) (PLGA) microspheres. Pharm. Res 2000; 17:100–106. 64. Burke PA. Determination of internal pH in PLGA microspheres using 31P NMR spectroscopy. International Symposium on Controlled Release of Bioactive Materials. 1996:133–134. 65. Cleland JL, Duenas ET, Park A, Daugherty A, Kahn J, Kowalski J, Cuthbertson A. Development of poly(D,L-lactide-co-glycolide) microsphere formulations containing recombinant human vascular endothelial growth factor to promote local angiogenesis. J. Controlled Release 2001; 72:13–24. 66. Cleland JL, Mac A, Boyd B, Yang J, Duenas ET, Yeung D, Brooks D, Hsu C, Chu H, Mukku V, Jones AJS. The stability of recombinant human growth hormone in PLGA microspheres. Pharm. Res 1997; 14:420–425. 67. Haynes CA, Norde W. Structures and stabilities of adsorbed proteins. J. Colloid Interface Sci 1995; 169:313–328. 68. Tsai T, Mehta RC, DeLuca PP. Adsorption of peptides to poly(D,L-lactide-co-glycolide). 1. Effect of physical factors on the adsorption. Int. J. Pharm 1996; 127:31–42. 69. Tsai T, Mehta RC, DeLuca PP. Adsorption of peptides to poly(D,L-lactide-co-glycolide). 2. Effect of solution properties on the adsorption. Int. J. Pharm 1996; 127:43–52. 70. Crotts G, Park TG. Stability and release of bovine serum albumin encapsulated within PLGA microparticles. J. Controlled Release 1997; 44:123–134. 71. Blanco MD, Alonso MJ. Development and characterization of protein-loaded poly(lactide-co-glycolide) nanospheres. Eur. J. Pharm. Biopharm 1997; 43:287–294. 72. Park TG, Yong Lee H, Sung Nam Y. A new preparation method for protein loaded poly(D,L-lacticco-glycolic acid) microspheres and protein release mechanism study. J. Controlled Release 1998; 55:181–191.
Microencapsulation of Protein Drugs
329
73. Gaspar MM, Blanco D, Cruz MEM, Alonso MJ. Formulation of l-asparaginase-loaded poly(lactideco-glycolide) nanoparticles: influence of polymer properties on enzyme loading, activity and in vitro release. J. Controlled Release 1998; 52:53–62. 74. Yoon J-Y, Kim J-H, Kim W-S. The relationship of interaction forces in the protein adsorption onto polymeric microspheres. Colloids Surf., A 1999; 153:413–419. 75. Crotts G, Sah H, Park TG. Adsorption determines in-vitro protein release rate from biodegradable microspheres: quantitative analysis of surface area during degradation. J. Controlled Release 1997; 47:101–111. 76. Boury F, Marchais H, Proust JE, Benoit JP. Bovine serum albumin release from poly(␣-hydroxy acid) microspheres: effects of polymer molecular weight and surface properties. J. Controlled Release 1997; 45:75–86. 77. Chang A-C, Gupta RK. Stabilization of tetanus toxoid in poly(D,L-lactic-co-glycolic acid) microspheres for the controlled release of antigen. J. Pharm. Sci 1996; 85:129–132. 78. Johansen P, Men Y, Audran R, Corradin G, Merkle HP, Gander B. Improving stability and release kinetics of microencapsulated tetanus toxoid by co-encapsulation of additives. Pharm. Res 1998; 15:1103–1110. 79. Arakawa T, Prestrelski SJ, Kenney WC, Carpenter JF. Factors affecting short-term and long-term stabilities of proteins. Adv. Drug Delivery Rev 1993; 10:1–28. 80. Castellanos IJ, Carrasquillo KG, De Jesus Lopez J, Alvarez M, Griebenow K. Encapsulation of bovine serum albumin in poly(lactide-co-glycolide) microspheres by the solid-in-oil-in-water technique. J. Pharm. Pharmacol 2001; 53:167–178. 81. Schwendeman SP, Tobio M, Joworowicz M, Alonso MJ, Langer R. New strategies for the microencapsulation of tetanus vaccine. J. Microencapsulation 1998; 15:299–318. 82. Tobio M, Schwendeman SP, Guo Y, McIver J, Langer R, Alonso MJ. Improved immunogenicity of a core-coated tetanus toxoid delivery system. Vaccine 2000; 18:618–622. 83. Jiang W, Schwendeman SP. Stabilization and controlled release of bovine serum albumin encapsulated in poly(D,L-lactide) and poly(ethylene glycol) microsphere blends. Pharm. Res 2001; 18: 878–885. 84. Carrasquillo KG, Carro JCA, Alejandro A, Toro DD, Griebenow K. Reduction of structural perturbations in bovine serum albumin by non-aqueous microencapsulation. J. Pharm. Pharmacol 2001; 53: 115–120. 85. Johnson OL, Jaworowicz W, Cleland JL, Bailey L, Charnis M, Duenas E, Wu C, Shepard D, Magil S, Last T, Jones AJS, Putney SD. The stabilization and encapsulation of human growth hormone into biodegradable microspheres. Pharm. Res 1997; 14:730–735. 86. Lee HJ, Riley G, Johnson O, Cleland JL, Kim N, Charnis M, Bailey L, Duenas E, Shahzamani A, Marian M, Jones AJ, Putney SD. In vivo characterization of sustained-release formulations of human growth hormone. J. Pharmacol. Exp. Ther 1997; 281:1431–1439. 87. Johnson OL, Cleland JL, Lee HJ, Charnis M, Duenas E, Jaworowicz W, Shepard D, Shahzamani A, Jones AJ, Putney SD. A month-long effect from a single injection of microencapsulated human growth hormone. Nature Med 1996; 2:795–799. 88. Cleland JL, Johnson OL, Putney S, Jones AJS. Recombinant human growth hormone poly(lacticco-glycolic acid) microsphere formulation development. Adv. Drug Delivery Rev 1997; 28:71–84. 89. Tracy MA. Development and scale-up of a microsphere protein delivery system. Biotechnol. Prog 1998; 14:108–115. 90. Lambert WJ, Peck KD. Development of an in situ forming biodegradable poly-lactide-co-glycolide system for the controlled release of proteins. J. Controlled Release 1995; 33:189–195. 91. Ravivarapu HB, Moyer KL, Dunn RL. Sustained suppression of pituitary-gonadal axis with an injectable, in situ forming implant of leuprolide acetate. J. Pharm. Sci 2000; 89:732–741. 92. Brodbeck KJ, Pushpala S, McHugh AJ. Sustained release of human growth hormone from PLGA solution depots. Pharm. Res 1999; 16:1825–1829. 93. Brodbeck KJ, DesNoyer JR, McHugh AJ. Phase inversion dynamics of PLGA solutions related to drug delivery. Part II. The role of solution thermodynamics and bath-side mass transfer. J. Controlled Release 1999; 62:333–344.
330
Yeo and Park
94. Cleland JL. Injectable gels for local and systemic delivery of proteins. Proceedings of the 28th International Symposium on Controlled Release of Bioactive Materials and 4th Consumer & Diversified Products Conference, San Diego, CA, 2001. 95. Hickel A, Radke CJ, Blanch HW. Role of organic solvents on Pa-hydroxynitrile lyase interfacial activity and stability. Biotechnol. Bioeng 2001; 74:18–28. 96. Alonso MJ, Gupta RK, Min C, Siber GR, Langer R. Biodegradable microspheres as controlledrelease tetanus toxoid delivery systems. Vaccine 1994; 12:299–306. 97. Sturesson C, Carlfors J. Incorporation of protein in PLG microspheres with retention of bioactivity. J. Controlled Release 2000; 67:171–178. 98. Sah H. Microencapsulation techniques using ethyl acetate as a dispersed solvent: effects of its extraction rate on the characteristics of PLGA microspheres. J. Controlled Release 1997; 47: 233–245. 99. ;;. 100. Blanco D, Alonso MJ. Protein encapsulation and release from poly(lactide-co-glycolide) microspheres: effect of the protein and polymer properties and of the co-encapsulation of surfactants. Eur. J. Pharm. Biopharm 1998; 45:285–294. 101. Freytag T, Dashevsky A, Tillman L, Hardee GE, Bodmeier R. Improvement of the encapsulation efficiency of oligonucleotide-containing biodegradable microspheres. J. Controlled Release 2000; 69:197–207. 102. Hildebrand GE, Tack JW. Microencapsulation of peptides and proteins. Int J Pharm 2000; 196: 173–176. 103. Soppimath KS, Kulkarni AR, Aminabhavi TM. Development of hollow microspheres as floating controlled-release systems for cardiovascular drugs: preparation and release characteristics. Drug Dev. Ind. Pharm 2001; 27:507–515. 104. Soppimath KS, Kulkarni AR, Aminabhavi TM. Encapsulation of antihypertensive drugs in cellulosebased matrix microspheres: characterization and release kinetics of microspheres and tableted microspheres. J. Microencapsulation 2001; 18:397–409. 105. Lam XM, Duenas ET, Cleland JL. Encapsulation and stabilization of nerve growth factor into poly(lactic-co-glycolic) acid microspheres. J. Pharm. Sci 2001; 90:1356–1365. 106. Soppimath KS, Aminabhavi TM. Ethyl acetate as a dispersing solvent in the production of poly(D,Llactide-co-glycolide) microspheres: effect of process parameters and polymer type. J. Microencapsulation 2002; 19:281–292. 107. Reslow M, Gustafsson N-O, Jonsson M, Laakso T. Polymer coated starch microspheres as protein delivery system. Proceedings of the 29th International Symposium on Controlled Release of Bioactive Materials, Seoul, Korea, 2002. 108. Debenedetti PG, Tom JW, Yeo S, Lim G. Application of supercritical fluids for the production of sustained delivery devices. J. Controlled Release 1993; 24:27–44. 109. Debenedetti PG, Tom JW, Yeo S. Supercritical fluids: a new medium for the formulation of particles of biomedical interest. International Symposium on Controlled Release of Bioactive Materials. 1993: 141–142. 110. Mishima K, Matsuyama K, Tanabe D, Yamauchi S, Young TJ, Johnston KP. Microencapsulation of proteins by rapid expansion of supercritical solution with a nonsolvent. AIChE J 2000; 46: 857–865. 111. Santos I, Richard J, Pech B, Thies C, Benoit JP. Microencapsulation of protein particles within lipids using a novel supercritical fluid process. Int. J. Pharm 2002; 242:69–78. 112. Randolph TW, Randolph AD, Mebes M, Yeung S. Sub-micrometer-sized biodegradable particles of PLA via the gas antisolvent spray precipitation process. Biotechnol. Prog 1993; 9:429–435. 113. Young TJ, Johnston KP, Mishima K, Tanaka H. Encapsulation of lysozyme in a biodegradable polymer by precipitation with a vapor-over-liquid antisolvent. J. Pharm. Sci 1999; 88:640–650. 114. Jung J, Perrut M. Particle design using supercritical fluids: literature and patent survey. J. Supercrit. Fluids 2001; 20:179–219. 115. Sunkara G, Kompella UB. Drug delivery applications of supercritical fluid technology. Drug Delivery Technol 2002; 44:46–50.
Microencapsulation of Protein Drugs
331
116. Zale SE, Burke PA, Bernstein H, Brickner A. Composition for sustained release of non-aggregated erythropoietin, 1997. 117. Lam XM, Duenas ET, Daugherty AL, Levin N, Cleland JL. Sustained release of recombinant human insulin-like growth factor-I for treatment of diabetes. J. Controlled Release 2000; 67:281–292. 118. Kang J, Schwendeman SP. Comparison of the effects of Mg(OH)2 and sucrose on the stability of bovine serum albumin encapsulated in injectable poly(D,L-lactide-co-glycolide) implants. Biomaterials 2001; 23:239–245. 119. Zhu G, Schwendeman SP. Stabilization of proteins encapsulated in cylindrical PLGA implants: mechanism of stabilization by basic additives. Pharm. Res 2000; 17:351–357. 120. Lavelle EC, Yeh M-K, Coombes AGA, Davis SS. The stability and immunogenicity of a protein antigen encapsulated in biodegradable microcapsules based on blends of lactide polymers and polyethylene glycol. Vaccine 1999; 17:512–519. 121. Jameela SR, Suma N, Jayakrishnan A. Protein release from poly(ε-caprolactone) microspheres prepared by melt encapsulation and solvent evaporation techniques: a comparative study. J. Biomater. Sci., Polym. Ed 1997; 8:457–466. 122. Baras B, Benoit MA, Gillard J. Influence of various technological parameters on the preparation of spray-dried poly(ε-caprolactone) microparticles containing a model antigen. J. Microencapsulation 2000; 17:485–498. 123. Yang YY, Wan JP, Chung TS, Pallathadka PK, Ng S, Heller J. POE-PEG-POE triblock copolymeric microspheres containing protein. I. Preparation and characterization. J. Controlled Release 2001; 75:115–128. 124. Bouillot P, Ubrich N, Sommer F, Duc TM, Loeffler JP, Dellacherie E. Protein encapsulation in biodegradable amphiphilic microspheres. Int. J. Pharm 1999; 181:159–172. 125. Jeon SI, Lee JH, Andrade JD, Gennes PG. Protein–surface interactions in the presence of polyethylene oxide. I. Simplified theory. J. Colloid Interface Sci 1991; 142:149–158. 126. Yeh M-K, Jenkins PG, Davis SS, Coombes AGA. Improving the delivery capacity of microparticle systems using blends of poly(D,L-lactide-co-glycolide) and poly(ethylene glycol). J. Controlled Release 1995; 37:1–9. 127. Cleek RL, Ting KCG, Eskin S, Mikos AG. Microparticles of poly(D,L-lactic-co-glycolic acid)/ poly(ethylene glycol) blends for controlled drug delivery. J. Controlled Release 1997; 48:259–268. 128. Kissel T, Kissel Y, Kissel XL, Volland C, Gorich S, Koneberg R. Parenteral protein delivery systems using biodegradable polyesters of ABA block structure, containing hydrophobic poly(lactide-coglycolide) A blocks and hydrophilic poly(ethylene oxide) B blocks. J. Controlled Release 1996; 39:315–326. 129. Luck M, Pistel K, Li Y, Blunk T, Muller RH. Plasma protein absorption on biodegradable microspheres consisting of PLGA, PLA or ABA triblock copolylmers containing PEO: influence of production method and polymer composition. J. Controlled Release 1998; 55:107–120. 130. Jung T, Koneberg R, Hungerer K-D, Kissel T. Tetanus toxoid microspheres consisting of biodegradable poly(lactide-co-glycolide) and ABA-triblock copolymers: immune response in mice. Int. J. Pharm 2002; 234:75–90. 131. Pistel KF, Bittner B, Koll H, Winter G, Kissel T. Biodegradable recombinant human erythropoietin loaded microspheres prepared from linear and star-branched block copolymers: influence of encapsulation technique and polymer composition on particle characteristics. J. Controlled Release 1999; 59:309–325. 132. Bittner B, Witt C, Mader K, Kissel T. Degradation and protein release properties of microspheres prepared from biodegradable poly(lactide-co-glycolide) and ABA triblock copolymers: influence of buffer media on polymer erosion and bovine serum albumin release. J. Controlled Release 1999; 60:297–309. 133. Wan J-P, Yang Y-Y, Chung T-S, Tan D, Ng S, Heller J. POE-PEG-POE triblock copolymeric microspheres containing protein. II. Polymer erosion and protein release mechanism. J. Controlled Release 2001; 75:129–141. 134. Ma¨der K, Bittner B, Li Y, Wohlauf W, Kissel T. Monitoring microviscosity and microacidity of the albumin microenvironment inside degrading microparticles from poly(lactide-co-glycolide)
332
135. 136. 137.
138.
139.
140. 141.
142.
143. 144. 145. 146. 147. 148. 149. 150.
Yeo and Park (PLG) or ABA-triblock polymers containing hydrophobic poly(lactide-co-glycolide) A blocks and hydrophilic poly(ethylene oxide) B blocks. Pharm. Res 1998; 15:787–793. Lee HK, Park JH, Kwon KC. Double-walled microspheres for single shot vaccine. J. Controlled Release 1997; 44:283–293. Wang N, Wu XS. A novel approach to stabilization of protein drugs in PLGA microspheres using agarose hydrogel. Int. J. Pharm 1998; 166:1–14. Wang N, Wu XS, Li JK. A heterogeneously structured composite based on poly(lactic-co-glycolic acid) microspheres and poly(vinyl alcohol) hydrogel nanoparticles for long-term protein drug delivery. Pharm. Res 1999; 16:1430–1435. Li JK, Wang N, Wu XS. A novel biodegradable system based on gelatin nanoparticles and poly(lactic-co-glycolic acid) microspheres for protein and peptide drug delivery. J. Pharm. Sci 1997; 86: 891–895. Castellanos IJ, Cuadrado WL, Griebenow K. Prevention of structural perturbations and aggregation upon encapsulation of bovine serum albumin into poly(lactide-co-glycolide) microspheres using the solid-in-oil-in-water technique. J. Pharm. Pharmacol 2001; 53:1099–1107. Yeo Y, Chen AU, Basaran OA, Park K. Solvent exchange method: a novel microencapsulation technique. The 29th Annual Meeting of the Controlled Release Society, Seoul, Korea, 2002. Yeo Y, Kim BY, Kim JD, Chen AU, Basaran OA, Park K. Solvent exchange method: variables for process optimization. The 29th Annual Meeting of the Controlled Release Society, Seoul, Korea, 2002. Yeo Y, Park K. Solvent exchange method: a novel microencapsulation process for protein delivery. Proceedings of the 28th International Symposium on Controlled Release of Bioactive Materials and 4th Consumer & Diversified Products Conference, San Diego, CA, 2001. Wise DL, Trantolo DJ, Altobelli DE, Yaszemski MJ, Gresser JD, Schwartz ER, eds. Encyclopedic Handbook of Biomaterials and Bioengineering. New York: Marcel Dekker, 1995. Hansen CM. The three dimensional solubility parameter—key to paint component affinities: II and III. J. Paint Technol 1967; 39:505–514. Teas JP. Graphic analysis of resin solubilities. J. Paint Technol 1968; 40:19–25. Barton AFM. Solubility parameters. Chem Rev 1975; 75:731–753. Wypych G. Solvents Database: ChemTec Publishing, 2000. Sah H, Smith M, Chern R. A novel method of preparing PLGA microcapsules utillizing methylethyl ketone. Pharm. Res 1996; 13:360–367. Sah H, Smith MS, Chern RT. The use of methyl ethyl ketone as a solvent for the preparation of PLGA microcapsules. Proc. Int. Symp. Control. Rel. Bioact. Mater 1995; 22:408–409. Carrasquillo KG, Stanley AM, Aponte-Carro JC, Jesus PD, Constantino HR, Bosques CJ, Griebenow K. Non-aqueous encapsulation of excipient-stabilized spray-freeze dried BSA into poly(lactide-coglycolide) microspheres results in release of native protein. J. Controlled Release 2001; 76:199–208.
16 Polymeric Gene Delivery Systems Goldie Kaul and Mansoor Amiji Northeastern University Boston, Massachusetts, U.S.A.
I. INTRODUCTION The correction of a gene defect which directly underlies a disease or disorder is the primary goal of gene therapy [1]. This also includes all treatment regimens that directly employ or target genetic material or correct a defect associated with a particular disease and therapies which introduce genes that can combat disease in a less direct but not necessarily a less effective manner. Gene therapy approaches, in addition, include efforts to cure diseases by introducing genetically modified cells (e.g., by transfection so that they may produce products of therapeutic benefit), DNA vaccination, and antisense therapy [1,2]. Gene therapy is formally defined as an approach to treat, cure, or prevent diseases by changing the expression of a gene. Given that genes regulate all the basic physiological processes in the body, there is tremendous potential for them to be employed as therapeutic agents [2]. The two broad divisions of gene therapy include the somatic and the germ line approaches. Although the somatic approach causes changes in the individual’s genome, these changes are not conserved and passed on to future generations. Most of the approaches have been only successful in gene therapy via the somatic approach [2]. Exogenously administered DNA rarely integrates with the chromosomes of the host genome. Instead it persists as an extrachromosomal element (episome) capable of expressing gene products for a period of time before it is eliminated from the host cell by nuclease degradation [1,2]. Thus, in the most general sense gene therapy is a method of providing somatic cells with the genetic information needed to produce specific therapeutic proteins for a transient period of time. In vivo gene therapy necessitates the formulation and evaluation of vectors to deliver therapeutic genes to a specific population either locally or systemically in order to efficiently express encoded proteins at the target site. The principle underlying gene delivery is that DNA, or more commonly the plasmid, is a particulate material with a net negative charge, a negatively charged surface, and a hydrodynamic diameter of approximately 100 nm. Target cells in the body usually have one or more barriers that would preclude penetration by the DNA. For instance, DNA cannot cross barriers like the endothelium, gastrointestinal tract, keratinized epithelium, or the blood brain barrier. In addition, because of its particulate nature and a net negative charge, DNA is likely to be opsonized from the body by the cells of the reticuloendothelial system [2]. Current gene medicine comprises optimizing a multicomponent-based system that includes a gene encoding a specific therapeutic protein, a plasmid-based gene expression system that is aimed at controlling the functioning of a gene within a target cell, and a gene delivery system 333
334
Kaul and Amiji
that controls the delivery of the gene expression plasmid to specific locations in the body [3]. The vectors for gene delivery must have qualities of safety for repeated use and provide reproducible levels of the gene product. Gene delivery systems should ideally control the location of a gene in the body by influencing the distribution and access of a gene expression system to the target cell and/or recognition by a cell surface receptor followed by intracellular trafficking and nuclear translocation. They should also protect a gene expression system from premature degradation in the extracellular milieu and to effect nonspecific or cell-specific delivery to the target cell [3]. Strategies to deliver genes efficiently to specific tissue and cell populations is a priority that was addressed in the 1995 Report and Recommendation of the Panel to Assess the NIH Investment in Research on Gene Therapy [4]. The first recommendation of the Panel was as follows: In order to confront the major outstanding obstacles to successful somatic gene therapy, greater focus on basic aspects of gene transfer, and gene expression within the context of gene transfer approaches, is required. Such efforts need to be applied to improving vectors for gene delivery, enhancing and maintaining high level of expression of genes transferred to somatic cells, achieving tissue-specific and regulated expression of transferred genes, and directing transfer to specific cells. Viral vectors, although very efficient in transfection, are plagued by issues of integration with the host genome to permanently alter its genetic structure, self-replication capability, recombination potential, and the possibility of complement activation (immunogenecity). The safety issue was brought into focus by the recent death of Jesse Gelsinger, an 18-year-old, in a clinical trial of gene therapy experiment to correct for ornithine transcarbamylase deficiency [5–7]. Viral vectors may also be unable to bypass the immune defense mechanism of the host and be limited in the amount of genetic information that they can carry. They are relatively expensive to manufacture too. In the last decade, there has been greater focus on the development of nonviral gene delivery vectors. An ideal vector, in order to successfully transfect cells at a remote site, needs to be inert and stable while in circulation, yet once it reaches the target site should release its load and effect an efficient and specific transfection of the target cells [2]. Specific viruslike characteristics that must be included into nonviral vectors are small size; stability against aggregation in blood, serum, or extracellular fluid; the ability to be efficiently internalized by the target cells; and the ability to disassemble and release the DNA in the cell once internalized. The design of an optimal synthetic gene carrier is still a limiting step for effective nonviral gene therapy [8]. Some of the nonviral vectors employed in gene therapy include cationic lipids, protein/peptide, and polymer based systems. Also of some importance to mention are the alternative approaches of using naked DNA, gene gun, jet injections, electroporation, and ultrasound [9–12]. This review focuses on the methods of nonviral gene delivery, with an emphasis on the use of synthetic and natural polymers as gene or intracellular delivery vehicles. The review will also address the various alternatives to polymeric delivery systems, discuss barriers (extracellular and intracellular) to gene therapy with polymeric vectors, and provide the advantages and disadvantages of the currently used polymeric systems.
II. NONVIRAL GENE DELIVERY SYSTEMS Nonviral gene delivery in many ways is similar to conventional drug delivery in general in that a single formulation cannot be used to target all sites in the body. Instead formulations need to be tailored for each target on the basis of physiological and biological characteristics. A variety
Polymeric Gene Delivery Systems
335
of methods of nonviral gene delivery are summarized in Table 1 that allow for targeting to different cell types in vitro and tissues in vivo. Briefly, these system include administration of the following techniques. A. Naked DNA The use of plasmid on its own, so-called naked DNA, was initially devised as a simplest delivery approach. Direct DNA injection has been found to efficiently transfect skeletal muscle [13] when administered by the intramuscular route. Naked DNA administration happens to be the method of choice for DNA-based vaccines. The main reason for the relatively good success of naked DNA in the muscle is attributed to the low levels of serum nucleases, which in serum and other organs clear the various topoisoforms of plasmid DNA from the plasma in minutes and reduce its efficacy [14,15]. Occluding efferent vessels followed by injection of naked DNA in afferent liver vessels has been found to achieve good transfection results as well [16]. Rapid injection of naked DNA into large volumes, hydrodynamic transfection, also allows efficient transfection of the liver on IV administration, as the pressure in the capillary bed temporarily dilates the fenestrae of the endothelial cells. It can additionally cause dilution of serum nucleases, thus protecting the DNA from degradation [17,18]. B. Jet Injections Some studies have also shown that the administration of DNA through jet injections may lead to enhanced uptake and expression. This involves projecting a column of liquid into tissues under pressure. Expression has been observed in skin and muscle [19,20]. C. Hydrogels Not widely used as a nonviral vector, delivery of pure DNA to the arterial endothelium has been done using a balloon catheter coated with a hydrogel impregnated with DNA [21,58].
Table 1 Delivery Options and Targets for Nonviral Gene Therapy Cell target Delivery system
In vitro
In vivo
Naked DNA Hydrogel Gene gun Jet injection Cationic lipids
None None Many None Many
Liposomes Virosomes Ligand-mediated asialoproteins Transferrin Surfactant
Some Many Hepatocytes
Skeletal muscle, cardiac muscle, thyroid, and synovium Endothelium Epidermis Epidermis and muscle Lung (IV, intratracheal, and arterial), tumors, and endothelium Liver (Kupffer cells) Liver (Kupffer cells, hepatocytes) and kidney Liver (hepatocytes)
Many Alveolar cells
Not done Not done
336
Kaul and Amiji
D. Gene Gun High-pressure bombardment of DNA bound to microparticles (Biolistic威) by a gene gun is a method for injecting directly into the nucleus of the target cells. This entails precipitating DNA onto microparticles like gold, which are then projected into the body by an explosive or gasdriven ballistic device. Once implanted into the cell the particles release DNA gradually, which is then expressed. This is mainly used in cell cultures of epithelial cells, endothelial cells, fibroblasts, lymphocytes, and monocytes. The main disadvantage of this method is that it is not capable of deep tissue or organ penetration, but it holds promise for DNA vaccines [22]. E. Cationic Lipids and Liposomes With the limitations of the viral vectors, several groups in academic institutions and industry are developing nonviral approaches to systemic gene delivery. A majority of the protocols for gene delivery using nonviral vectors have utilized lipid-DNA complexes (lipoplexes) and liposomes (DNA encapsulated in lipid bilayer vesicle). Philip Felgner, currently at Gene Therapy Systems (San Diego, CA), pioneered the development of cationic lipid-based DNA delivery systems [23]. Lipofectin姟 and other cationic lipids have been tested in various stages of clinical trials. Lipofectin is a cationic lipid composed of 1:1 (w/w) ratio of DOTMA lpar;N[1-(2,3dioleyloxy)propyl]-N,N,N-trimethylammonium chloride) and DOPE (dioleylphosphatidyl ethanolamine). Other commonly used cationic lipids include (2,3-dioleyloxy-N-[2-(sperminecarboxamido)ethyl]-N,N-dimethyl-1-propanaminium trifluoroacetate (DOSPA), (dioctamido-decylamidoglycyl spermine (DOGS), 1,2-dimyristyloxypropyl-3-dimethylhydroxyethyl ammonium bromide (DMRIE), 1,2-bis(oleyloxy)-3-(trimethylammonio)propane (DOTAP), and (3-N,Ndimethylaminoethane)carbamoyl cholesterol (DC-chol) [24–27]. Liposome-DNA complexes were used for local cytokine delivery in human tumor xenografts [28]. Delepine et al. [29] administered -galactosidase plasmid/phosphonolipid (cationic lipid) complexes to mice by intratracheal and aerosol route, and luciferase plasmid/phosphonolipid by intravenous route. This led to the transgene expression mainly in the lungs. Farhood et al. [30] provided early evidence for the use of lipoplexes for in vivo DNA delivery in murine tumor models. These authors report on the toxicity profile and biodistribution of lipoplexes delivered intravenously and intratumorally. Although the lipoplexes failed to localize in the tumor after intravenous injection, significant DNA accumulation was observed after intratumoral administration. Reimer et al. [31] recently reported on the intraperitoneal administration of cationic lipidbased gene delivery system for chloramphenicol acetyl transferase (CAT) gene in B16/BL6 murine tumor models. Gene expression in the B16/BL6 tumors was highly variable, with values ranging from greater than 2,000 mU/g to less than 100 mU/g tumor. In small tumors (⬍250 mg), however, almost 18% of the administered DNA dose as lipoplexes was in the tumor mass 2 h after intraperitoneal injection. The initial clinical results of lipoplex-mediated gene therapy examined the delivery of human leukocyte antigen gene HLAB7 to tumors in order to enhance both the presentation of tumor-specific antigen and stimulation of CTL response [32]. Five patients in this study exhibited successful uptake and expression of the plasmid in tumors as well as the generation of antibodies against HLAB7. The main disadvantages of lipoplexes for DNA delivery are poor efficiency of transfection, lack of selectivity of the target, and in some cases poor stability of the entrapped plasmid. In spite of the early excitement with cationic lipids, there are serious limitations similar to those of viral vectors. The cationic lipids were found to be highly toxic on repeated use and induced a potent inflammatory reaction in vivo. Although poly(ethylene glycol) (PEG) modification was found to attenuate the levels of complement activation and plasma protein binding, it was found to adversely affect the transfection potential.
Polymeric Gene Delivery Systems
337
F. Virosomes Virosomes are liposomes that contain viral proteins such as the Sendai virus that helps in hemagglutinin-mediated fusion of the liposome bilayer with the target cell membrane. Transfection efficiency can also be increased by encapsulating DNA with nuclear proteins in a liposome interior as demonstrated by Kaneda et al. [33]. The presence of these proteins is believed to improve trafficking by virtue of nuclear targeting. Kaneda and Kato [34–36] have demonstrated that direct injection of these into the liver leads to high levels of transgene expression in hepatocytes. G. Cationic Polymers Cationic polymer–based DNA delivery systems offer significant promise in gene therapy. Early work in this area concentrated on the use of nondegradable polymers such as poly(L-lysine) (PLL), poly(ethyleneimine) (PEI), diethylaminoethy1 (DEAE) methacrylate, DEAE-dextran, and others [2]. Research from Jean-Paul Behrs group at the Universits Louis Pasteur in France showed very efficient transfection of PEI-DNA complex both in vitro and in vivo [37]. These investigators postulated the ‘‘proton sponge effect’’ as the mechanism of transfection, which is based on the ability of PEI to be fully protonated in the acidic environment of the endosomes [37]. The buffering action of PEI in the endosome leads to an increase in the pH and osmotic pressure and a subsequent rupture and release of the DNA complex into the cytoplasm. By minimizing the fusion of endosomes with lysosomes and subsequent degradation of DNA through the rapid endosomal escape, efficient tranfection was observed. Subsequently, Mikos and coworkers at Rice University elegantly studied the intracellular uptake and trafficking of PEI-DNA complex in tumor cells by dual fluorescence labeling [38,39]. They found the PEI-DNA complex to be internalized by endocytosis after 2 h and subsequent release from the endosome in about 4 h posttransfection. After an additional half hour, the complex was localized in the nucleus. Interestingly, Mikos’s group observed that PEI-DNA complex remains stable in the nucleus and the unpacking of PEI and DNA was not necessary for transfection. Although PEI by itself has been found to be cytotoxic, the DNA complex is significantly less toxic [39]. Langer’s group has embarked on synthesis of DNA-binding biodegradable polymers with inherent endoplasmic escape properties and low toxicity profile [40–42]. Lynn and Langer [89] describe the synthesis of poly(-amino esters) using different secondary amines and dioldiacrylates. These synthetic biodegradable cationic polymers were able to condense DNA to form nanocomplexes of about 150 nm [90]. From initial observations with poly(-amino ester), the polymeric nanoparticle dissolves in low pH medium and forms soluble complexes with DNA and other negatively charged macromolecules. Leong’s group has developed nanospheres of DNA by complexation and precipitation with cationic biopolymers like gelatin and chitosan [43]. Nanospheres offer enhanced protection against degradation of DNA and could be tailored for targeted delivery to specific cells by attachment of recognition elements on the surface. H. Ligand-Mediated Gene Delivery Ligand-mediated gene delivery is an effective method of gene transfer to cells in vitro. Protein ligands can be coupled to PLL and then incorporated into a ligand-DNA complex by ionic interactions between the PLL and DNA [44–49]. The ligand may alternatively be coupled to intercalating agents including bis-acridine or ethidium dimers and complexed with DNA [50,51]. These complexes retain their ability to interact specifically with receptors on the target cell
338
Kaul and Amiji
leading to internalization of the complex within the cell. Transfection studies with transferrinPLL-DNA complexes have been performed in hematopoietic and pulmonary epithelium cells [51,53,54]. Gene transfer has also been done using tris-galactosyl compounds and folate-PLLDNA complexes [52]. In reporter gene studies asialoorosomucoid-PLL-DNA complexes have been shown to deliver genes directly to the liver in vivo [49]. This approach has been used to replace low density lipoprotein (LDL) receptors in rabbits with LDL deficiency [55]. I. Adenovirus-Coupled Polymers and Peptides It was observed that adenoviral viruses can cause endosomal lysis during infection and this fact can be utilized for facilitating endosomal release of a plasmid DNA molecule coupled to a carrier, a strategy that would prevent degradation during the late endosome–lysosome fusion of particles undergoing vesicular transport. In vitro studies using adenovirus coupled to transferrinpolylysine-DNA, asialoglycoprotein-polylysine-DNA or folate-polylysine-DNA enhanced gene delivery 100- to 1000-fold [56–58]. This is most efficient if the DNA molecule is covalently coupled to the adenoviral particles, which can be done by either coupling polylysine on the surface of the adenovirus [59,60] or biotinylating the adenoviral surface and attaching DNA via the streptavidin-polylysine-DNA complexes [61]. A major disadvantage of this method is again the large size and the immunogenic potential of these systems.
III. POLYMERIC VECTORS Based on their physical, chemical, and biological properties several polymer-based gene delivery systems are under investigation. There has been a fair amount of success in reducing immunogenicity and cytotoxicity with the concomitant enhancement in the level of transgene expression associated with the polymers in pipeline. Among the polymeric systems, the cationic polymers have been studied the most as they offer the dual advantage of being able to condense DNA and also mask the negative charge. The polymers used in gene delivery are summarized in Table 2. A. Biodegradable Polymers 1. Chitosan Chitosan is a nontoxic biodegradable and biocompatible polysaccharide polymer composed of D-glucosamine and N-acetyl– D-glucosamine linked together by a -(1,4) glycosidic linkage [62–64]. General lysosomes in the body degrade chitosan to N-acetyl glucosamine, which is taken up into the synthetic pathway of glycoproteins and subsequently excreted as carbon dioxide [65]. These amino groups of D-glucosamine residues have an intrinsic pKa of 6.0–6.5, and thus chitosan behaves as a polycation at acidic and neutral pH and is able to effectively condense DNA [63,66]. The electrostatic interactions of chitosan with negatively charged polyions like indomethacin, sodium hyaluronate, pectin, and polysaccharides are well characterized. It is found to behave in a similar manner for condensing plasmid DNA [62]. Hydrophobically modified chitosan with deoxycholic acid residues was shown to condense plasmid DNA by gel retardation experiments [67]. Roy et al. [68] prepared 200- to 300-nm DNA-chitosan nanoparticles by a complex coacervation process and used transfected HEK-293 (human embryonic kidney) cells in vitro at lower levels than with Lipofectamine威. Mao et al. [69] prepared stable, nonaggregating chitosan nanoparticles, which on freeze-drying had a shelf
Polymeric Gene Delivery Systems
339
Table 2 Summary of Polymeric Systems Used for Gene Delivery Polymer type Biodegradable polymers
Example Natural
Synthetic
Nondegradable polymers
DNA-condensing systems
Noncondensing systems
Thermosensitive polymers
Chitosan Collagen and atelocollagen Alginate Gelatin Poly(-amino esters) Poly(D, L-lactic acid-co-glycolic acid) Poly[␣-(4-aminobutyl)-L-glycolic acid] Poly(4-hydroxy-1-proline) ester Polyanhydride Copolymers of fumaric and sebacic acid Gene activated matrix (GAM) Imidazole-containing polymers Poly(amino acids) like poly(L-lysine) Poly(ethylenimine) Poly(2-dimethylaminoethyl) methacrylate Starburst dendrimers Poly(ethylene-co-vinyl acetate) Poly(vinyl alcohol) and poly(Nvinylpyrrolidone) Pluronic® and Tetronic® copolymers Poly(N-isopropylacrylamide)
life of 1 month. Chitosan condensation of DNA was also found to protect it from DNAse degradation in vitro. Surface PEGylation of chitosan-DNA nanoparticles was also studied to improve their physicochemical properties on storage and to yield a formulation that could be lyophilized without loss in transfection ability [70]. Ouchi et al. [71] also synthesized a quaternary chitosan derivative coupled to galactose residues to exploit the receptor-mediated endocytosis in targeting Hep G2 cells in vitro. The physicochemical properties of chitosan, like the colloidal and the surface properties, depend on the molecular weight, the plasmid-to-polymer ratio, and the medium. MacLaughlin et al. [72] studied the effect of a pH-sensitive endosomolytic peptide that was found to enhance the levels of a reporter gene expression in Cos-1 cells four fold, wherein the combination of a strong complex stability and low in vivo expression levels suggest that uptake and/or decomplexation, but not endosomal release, were the critical rate-limiting steps in the uptake process. Lee et al. [73] have also studied the complexation of low molecular weight chitosan, with plasmid DNA in physiological buffers. The plasmid DNA was retarded completely at a weight ratio of 1:2 (plasmidchitosan) in a 1.0% agarose gel and completely protected from DNase-1 degradation over a 1-h period. This approach was found to have significantly higher transfection efficiency than naked DNA or when complexed with PLL. Erbacher et al. [74] partially substituted chitosan with lactosyl residues to target cells expressing galactose-binding membrane lectin. This approach, however, did not enhance transfection in BNL, CL2, or Hep G2 cells at concentrations of 3, 10, and 20% lactosylation. The authors suggested that lactosylation of chitosan reduced the surface charge of the complexes, leading to aggregation, and also reduced the affinity of DNA for chitosan. Being a mucoadhesive polymer, chitosan is a good potential carrier for sustained interaction of the plasmid DNA with mucus-secreting membrane epithelia. The polymer is considered
340
Kaul and Amiji
a good candidate for transfecting gastrointestinal epithelia, nasal airway epithelia, and immune cells in the gut/mucosal-associated lymphoid tissue (MALT). Roy et al. [68] have demonstrated that oral peanut allergen gene therapy with chitosan/DNA nanoparticles could be one of the better ways to modulate anaphylactic reactions in rats. 2. Collagen and Atelocollagen Collagen is a fibrous protein of connective tissue (such as the dermis of the skin) that plays an important role in the maintenance of the structural integrity and morphology of tissues and organs [75,76]. Collagen (tropocollagen) is a helical molecule with three polypeptide chains. It is a rodlike molecule with a length of 300 nm and diameter of 1.5 nm. The amino acid that confers antigenic properties to the molecule is an amino acid sequence called the telopeptide on both its N- and C-terminals. Atelocollagen, a decomposition product of type I collagen, is derived from the dermis of cattle [76]. Atelocollagen has a low antigenic potential as it is obtained by peptic digestion of collagen and is free from telopeptides [77]. Atelocollagen is soluble at a lower temperature, but solidifies when temperature is raised to over 30⬚C. Therefore, atelocollagen can be used as a fluid or gel, either of which is injected locally, or as a solid material in the form of beads, sponges, membranes, or cylinders. This means that atelocollagen can fill the contour spaces in the body and around the blood vessels when it is used as an internal implant. When used for gene delivery, the DNA can be released gradually and with appropriate control of the concentration, shape, and dose of the gene as well as of the site of injection [78]. Ochiya et al. [79] injected an atelocollagen minipellet into the femoral muscle of an ICR mouse and then examined the plasmid DNA release and gene expression regulation from peripheral blood and various organs. In a control experiment using plasmid DNA alone, the release of DNA in the blood lasted 7 days after administration. With the minipellet, on the other hand, the release persisted for over 40 days. Moreover, intact plasmid DNA was detected in the muscle 60 days later in the case of the minipellet. These facts suggest that intramuscular administration of the minipellet might result in a sustained release of DNA and protect it from biodegradation. These results suggest that atelocollagen can play an important role in gene therapy as a biocompatible material, increasing the effectiveness of plasmid DNA in vivo 3. Alginate Sodium alginate is a naturally occurring biopolymer extracted from brown algae (kelp) that can be easily polymerized into a solid matrix to form gels and microspheres [80]. Mittal et al. [81] prepared biodegradable alginate microspheres containing the bacterial -galactosidase (LacZ) gene under the control of either the cytomegalovirus (CMV) immediate-early promoter or the Rous sarcoma virus (RSV) early promoters for use as a delivery vehicle for DNA-based mucosal vaccines. LacZ-coding DNA was encapsulated in microspheres of alginate to which a significant immune response was observed on mucosal immunization. Mice inoculated orally with microspheres containing plasmid DNA expressed LacZ in the intestine, spleen, and liver. Inoculation of mice with microspheres containing both the plasmid DNA and bovine adenovirus type 3 (BAd3) resulted in a significant increase in LacZ expression compared to those inoculated with microspheres containing only the plasmid DNA. These results implied that the adenoviruses are capable of augmenting transgene expression by plasmid DNA both in vitro and in vivo. 4. Gelatin Gelatin is a proteinaceous biopolymer, obtained by hydrolysis of collagen, with proven safe use in pharmaceutical and food products. It is commercially available by hydrolytic degradation of
Polymeric Gene Delivery Systems
341
naturally occurring porcine or bovine collagen. It is most importantly nonirritating, biocompatible, and biodegradable [82,83]. It is a natural polyampholyte that gels below 35–40⬚C. At pH below 5, positively charged gelatin can form complexes with DNA [82,83]. Truong et al. [84] prepared and characterized nanoparticles made by salt-induced complex coacervation of DNA and gelatin. Injection of this preparation in mice footpad and the tibialis muscle resulted in enhanced expression of LacZ gene product compared to equal doses of naked DNA and Lipofectamine-DNA. Truong et al. [85] also prepared DNA-gelatin nanoparticulate coacervate with chloroquine as an endosomolytic agent, calcium, and transferrin bound to the gelatin molecule. These nanoparticles offered limited protection against DNase I enzyme in vitro. Transfection of 9HTEo (human tracheal epithelial cells) with the gelatin nanoparticle system containing the cystic fibrosis transport regulator (CFTR) gene resulted in 50% transfection of the cells in vitro. Leong et al. [86] also prepared DNA-polycation nanospheres as a potential gene delivery vehicle. Kaul and Amiji [87,88] developed and characterized long-circulating nanoparticulate formulation of gelatin by PEG modification of the biopolymer. PEG-gelatin nanoparticles had a decreased degradation in the presence of protease enzyme due to steric repulsion. Also a large fraction of the nanoparticles had localized in the perinuclear region of BT-20 (human breast cancer) cells after 12 h. 5. Poly(-Amino Esters) Choice of a cationic polymer for gene therapy is basically dependent on a compromise between the transfection efficiency and short- or long-term cytotoxicity. For in vivo delivery, it is also very important that the carrier is biocompatible. As many nondegradable polymers are not biocompatible, one approach used by Langer’s group is to synthesize cationic polymers that can gradually degrade in vivo into nontoxic by products [40–42]. Lynn and Langer [89], for instance, describe the synthesis of poly(-amino esters) as biodegradable polymers bearing cationic side chains. By selecting appropriate primary or secondary diamines and diol-diacrylate, poly(-amino ester) can be made with tailored properties. For instance, they can be hydrophilic or hydrophobic, be insoluble or soluble at physiologic pH, and have controllable biodegradation kinetics. These synthetic biodegradable cationic polymers were found to condense DNA to form nano-complexes of about 150 nm [90]. From initial observations with poly(-amino ester), the polymeric nanoparticle dissolves in low pH medium and form soluble complexes with DNA and other negatively charged macromolecules. The nanometer-sized dimensions and reduced cytotoxicities of these DNA-polymer complexes suggest that they may be useful as degradable polymeric gene transfer vectors. 6. Poly(D,L-lactic-co-glycolic acid) Poly(D,L-lactic-co-glycolic acid) (PLGA) is one of the most widely used biocompatible and biodegradable polymers. It also happens to have an extensive record of safe use in medicine [91]. PLGA microparticles of size less than 10 m in diameter are actively taken up by the macrophages of Peyer’s patches in the gastrointestinal tract [92]. Antigens encapsulated in PLGA microspheres have been shown to be effective in eliciting systemic and mucosal immunity [93]. Jones et al. [94] administered encapsulated insect protein luciferase DNA under the control of a human cytomegalovirus immediate-early promoter (CMV promoter) to mice by intraperitoneal injection or oral gavage. This injection was found to elicit good serum IgG and IgM responses and a modest IgA response on intraperitoneal administration. It was concluded from those studies that PLGA encapsulation protected DNA against degradation and allowed the DNA to reach immune cells to elicit both systemic and local mucosal antibody response. Gebrekidan et al. [95] studied the stability, in vitro release, and cell transfection with PLGA microspheres contain-
342
Kaul and Amiji
ing free and polylysine (PLL)-complexed DNA. These microspheres were prepared by a wateroil-water solvent extraction/evaporation method, where encapsulation was found to enhance the retention of the supercoiled DNA, and PLL complexation increased the stability of both the supercoiled and the encapsulated form. PLGA-encapsulated DNA was released in a sustained fashion for up to 42 days. Bioactivity of the encapsulated DNA was retained to a greater extent with the PLL-complexed formulation. Hsu et al. [96] prepared PLGA microspheres by the waterin-oil-in-water emulsion and solvent evaporation method, which was found to protect DNA against DNase degradation. They also showed that when the DNA was condensed with cationic polymer prior to encapsulation, it had better physical stability, especially during processing steps. To increase the transfection efficiency with PLGA-based DNA delivery systems, Maruyama et al. [97] conjugated dextran-grafted PLL and mixed it with PLGA (Mr 10,000 Da) to prepare nanoparticles by the solvent evaporation method. Adsorption of DNA on the nanoparticle surface was found to depend on the dextran content in the graft copolymer. Mooney’s group [98,99] reported on the in vivo delivery of plasmid DNA encoding a platelet-derived growth factor gene using a PLGA matrix for tissue engineering applications. 7. Poly[␣-(4-Aminobutyl)-L-Glycolic Acid] Lim et al. [100,101] described the synthesis of (poly[␣-(4-aminobutyl)-L-glycolic acid] (PAGA), a polyester that can be degraded by the esterases in vivo. When the degradation kinetics were studies in aqueous solution by MALDI-TOF-MS, PAGA was found to degrade very quickly into an L-oxylysine monomer. Formation of self-assembling complexes between DNA and PAGA was confirmed by gel band shift assay at a charge ratio of 1:1. These complexes were found to show two fold higher transfection efficiency using a -galactosidase reporter gene than DNA-PLL complexes. The authors concluded that PAGA would be safe in gene delivery because of its biocompatibility and biodegradability. Lee et al. [102] delivered interleukin-4 (IL-4) plasmid in PAGA to prevent autoimmune insulitis in transgenic NOD mice. The plasmid/PAGA complex was transfected into 293T cells and the expression of IL-4 measured by ELISA. In vitro transfection assays showed that the PAGA enhanced the expression of IL-4 in 293T cells. When plasmid/PAGA complexes were injected intravenously in NOD mice, IL-4 mRNA levels in the liver were found to be highest 4 weeks after administration. Single-dose administration also was found beneficial in the treatment of autoimmune insulitis in mice. 8. Poly(4-Hydroxy-L-Proline Ester) Poly(4-hydroxy-L-proline) (PHP) ester was synthesized from CBZ–4-hydroxy-1-proline by melting condensation polymerization of N-CBZ–4-hydroxy-L-proline followed by subsequent protection by palladium on activated carbon as a catalyst [103–104]. Alternatively it can also be prepared by dicyclohexylcarbodiimide pyridine (DCC/DMAP)-activated polycondensation of N-carboxy-4-hydroxy-L-proline. PHP ester effectively condenses DNA, is biodegradable, and has low cytotoxicity. Its transfection efficiency is similar to that of PLL [39]. 9. Polyanhydride Copolymers of Fumaric and Sebacic Acid The efficient delivery of growth-promoting genes locally in a sustained manner is important for effective tissue regeneration. Orally administered biodegradable poly(FA:SA) (20:80), whose copolymers have polyanhydride copolymers of fumaric and sebacic acid highly adhesive properties, promoted gene delivery after oral administration. In vivo experiments using poly(FA:SA)
Polymeric Gene Delivery Systems
343
(20:80) microspheres with pCMV/-galactosidase plasmid resulted in positive expression of galactosidase activity [106,107]. 10. Gene Activated Matrix Plasmid DNA carrying a fragment of the human parathyroid hormone gene was carried into target tissue for its regeneration by a polymer matrix sponge called a gene-activated matrix (GAM) [108]. GAM implantation at bone injury sites was found to increase the retention and expression of plasmid DNA for a longer period and resulted in reproducible new bone tissue regeneration. Wound healing involves proliferation of fibroblasts of granulating tissue and migration from viable tissue at the margin into the wound bed. Once there, fibroblasts take up and transiently express DNA. GAM typically consists of a DNA vector and a structural matrix carrier. The GAM matrix has two functions: it holds vector DNA in the wound site (until cells arrive) and it acts as scaffolding that promotes fibroblast ingrowth and accumulation near the DNA. While in the matrix, transfected fibroblasts act as local in vivo bioreactors, producing vector-encoded proteins that promote wound repair [108–111]. 11. Imidazole-Containing Polymers Pack et al. synthesized imidazole-containing polymers, which are intrinsically endosomolytic and allow for efficient release of the DNA-polymer complex from endocytotic vehicles into the cytoplasm.
B. Nonbiodegradable Polymers 1. DNA Condensing Systems DNA is capable of being condensed into small mono- or polynuclear particles with an excess of polycations in aqueous solutions into polymer-DNA complexes, commonly known as polyplexes [112,113]. The cationic polymers spontaneously form DNA complexes because of the ionic interactions between the positively charged amine groups of the polycation and the negatively charged phosphate groups in the DNA backbone. This interaction effectively neutralizes the negative charge on the DNA molecule and is found to facilitate DNA uptake by cells and thus increase transfection efficiency [112]. Figure 1 shows the chemical structure of some common cationic condensing polymers. Cationic polymers generally have an excess of protonable amines. They can either be branched or linear [114]. Diethylaminoethyl dextran (DEAE-dextran) was the first investigated molecule in this category [115]. To improve the solubility and stability characteristics and to reduce nonspecific interactions with biomolecules, PEG-grafted analogs of polyplexes have been prepared [116–121]. Many cationic polymers are also amenable to conjugation with targeting ligands [122–124]. Plasmid DNA undergoes condensation with the cationic polymers to reduce the hydrodynamic size of the DNA molecule from hundreds of nanometers to tens of nanometers. The condensation reduced DNA volume by 1000- to 10,000fold from that of the naked plasmid DNA [125]. Prokop et al. [126] developed and evaluated a series of cationic polymers for in vivo delivery of plasmid expression vectors and compared them with soluble DNA. Weakly cationic polymers with a moderate negative charge after DNA complexation were found to be more efficient in transfection. Also these systems were significantly less cytotoxic as compared to cationic polymers with high charge density. Thus for an in vivo expression system, a compromise must reached between cytotoxicity and charge density
344
Kaul and Amiji
to allow for efficient transfection. Some of the commonly used gene-condensing polymers are as follows. Poly(L-Lysine). PLL is a polycation that is found to condense DNA under various salt conditions. PLL possesses in its backbone an adequate number of positively charged primary amino groups that interact with the phosphate groups in DNA [120,127]. The intensity of binding between the polymer and DNA determines the degree of condensation and intracellular release
Figure 1 Chemical structures of cationic polymers used for DNA complexation.
Polymeric Gene Delivery Systems
345
Figure 1 Continued.
efficiency. This intensity depends on the number of available ε-amino groups. The binding requires more than 20–80 units to yield transfection competent complexes, but has an upper limit for high molecular weight PLL [128,129]. Also it is important to mention that higher molecular weight PLLs are significantly more toxic [130,131]. PLL-DNA polyplexes tend to be cleared rapidly from the systemic circulation with a plasma half-life of only 5 min [132]. Rapid clearance of these polyplexes was attributed to charge-mediated interactions with serum albumin and other proteins. PLL-DNA complexes were also found to aggregate under physiological conditions. When PLL was modified with hydrophilic dextran by a reductive amination reaction between the ε-amino groups of PLL and reductive ends of dextran, the aggregation problem was significantly decreased. Also, in order to decrease aggregation and toxicity, PLL was linked covalently to PEG to create a steric barrier and increase the plasma half-life. The PEG-graft-PLL was found to be relatively nontoxic to cells in culture compared to PLL alone, and the transfection efficiency was 30-fold higher than that of PLL-DNA complexes [133].
346
Kaul and Amiji
Additionally, PLL has been reacted to have pendant hydroxypropyl methacrylate (HPMA) groups to reduce cytotoxicity. PLL has also been modified with lipophilic compounds to create ‘‘lipopolylysine’’ [134,135]. Kim et al. [136], for instance, have prepared stearyl-PLL conjugates that form 100-nm complexes with low density lipoprotein that were transfection competent. Addition of chloroquine, as an endosomolytic/buffering agent, has been found to be necessary for transfection with unmodified PLL [137–140]. Brown and Schatzlein [130] prepared neutral, relatively less toxic, lipo- and hydrophilically modified conjugates of PLL by palmitoyl-PEG-PLL amphiphilic graft copolymer that could complex with DNA into vesicles in the presence of cholesterol. These systems were able to transfect in the absence of an endosomolytic and/or buffering agent. Poly(L-lysine) transfection is greatly improved using endosomolytic peptides and adenofection. Chloroquine and membrane-active peptides have been found to exert synergistic effect on endosomal lysis. PLL modifications with histidyl residues that buffer the endosomal pH by virtue of the proton sponge effect are also effective in improving transfection efficiencies [141]. PLL has also been coupled to targeting moieties like the asialoorsomucoid/asialoglycoproteins receptors to target liver cells, transferrin, and its receptor to rapidly growing cells [142–144]. Monoclonal antibodies, sugars, peptides, and epidermal growth factor modifications of PLL have also been described by Zauner et al. [145]. Although a majority of PLL applications in gene therapy have been limited to in vitro studies, some studies with galactosylated PLL polyplexes with a low negative zeta potential have been described to target hepatocytes via the asialoglycoprotein receptor in vivo [146,147]. Poly(Ethyleneimine). PEI is an excellent example of a cationic polymer. It has a 1:2: 1 ratio of primary/secondary/tertiary amines in its structure, with every third nitrogen being protonable. This results in a polymer with a very high charge density [148]. Protonation of the amine groups is also a function of pH, and changes from 20 to 40% when the pH is lowered from 7.0 to 5.0 [149]. PEI is also a classic example of a cationic polymer with endosome buffering capability [148]. Studies have shown that PEI in the endosome results in the accumulation of protons brought in by the endosomal ATPase enzyme, which also results in an influx of chloride anions. The highly charged proton-dense environment of the endosome then results in swelling of the PEI molecule due to charge repulsion that ultimately leads to osmotic swelling of the organelle. Due to high internal pressure in the early endosome or endosone/lysosome complex, the released DNA remains stable from degradation by acidity and enzymes in the organelle [150]. PEI-DNA complexes have been shown to produce high levels of expression in a mature mouse brain. Delivery of the antiapoptotic gene bcl-XL, with a reporter luciferase, was found to significantly increase expression of both these genes at 1 week after administration [151,152]. As is the case with many cationic polymers, PEI is significantly cytotoxic. In a recent study, 25 to 50% cell death was observed after administration of PEI-DNA complexes at a ratio of 1: 1 w/w and 100% cell death was reported at a ratio of 10:1 w/w [153–156]. Poly(ethyleneimine) has been conjugated to carbohydrate moieties [157–159], integrinbinding peptides [160], transferrin [156], and monoclonal antibodies [161] to improve polyplex targeting and cellular uptake. A lower ratio of polymer and/or inactivated adenovirus to plasmid is recommended because receptor-mediated endocytosis is sensitive to interference from nonspecific interactions [153]. The nearly neutral PEI-DNA complexes were found to be more efficient than the highly positive ones, which were also unacceptably toxic. Higher molecular weight PEI complexes were found to be more efficient in vitro. The smaller linear PEI complexes were more efficient than either the branched or the higher molecular weight polymers on systemic administration in vivo [153,162–164]. In a biodistribution study following tail-vein injection in rats, PEI-DNA complexes were predominantly trapped in the lung [151,154,156,163,165]. Lung transfection studies alone have also been done by conjugating PEI to anti-PECAM antibody,
Polymeric Gene Delivery Systems
347
which showed 20-fold higher expression than a control PEI-only polyplex [156]. PEI coupled to nonionic poly(ethylene oxide) altered this preference for the lung, resulting in a rank order of expression of liver ⬎ spleen ⬎ lung [166]. Cationic Starburst Dendrimers. Starburst polyamidoamine (PAMAM) dendrimers are highly branched spherical polymers. The surface charge and the diameter of PAMAM dendrimer are determined by the number of steps (or generations) in its synthesis [167]. PAMAM dendrimers are prepared with either ammonia or ethylenediamine as the core. The overall shape, density, and surface charge are determined by the choice of core and the monomers used for dendrimer synthesis [168]. Dendrimers condense DNA through electrostatic interactions of the terminal primary amines with the negatively charged phosphates on the DNA molecule. Dendrimer concentration in the complexes has been found to influence the resulting polyplex particle size, surface charge, and gene transfer efficiency. Transfection efficiency of the dendrimer-based systems into cells depends on the size, the shape, and the number of primary amino groups on the surface of the polymer. Dendrimers have been used to mediate nonspecific but efficient transfection to eukaryotic cells [169]. Epstein-Barr virus (EBV)–based polymer vector conjugated with PAMAM dendrimer was investigated and found to be effective in suicide gene herpes simplex virus type-1 thymidine kinase (HSV-1 TK) transfer in both in vitro into Ewing’s sarcoma A4573 cells and in vivo in subcutaneous tumors grown in immunocompetent mice [170]. Poly(2-Dimethylaminoethyl) Methacrylate. pDMAEMA is a hydrophilic cationic polymer that has the potential to bind and condense DNA and mediate transfection into a variety of cells [171]. Zuidam et al. [171] compared pDMAEMA with its quaternary ammonium analog poly(2-trimethylamino)ethyl methacrylate (pTMAEMA) to evaluate the effects of tertiary groups versus quaternary ammonium groups on transfection. Based on the results of transfection and cellular interaction, they postulated that pTMAEMA did not have the intrinsic endosome escape properties as seen with proton sponge effect of pDMAEMA. This was explained by the tighter and irreversible complexation of the quaternary analog that did not dissociate within the cell to release the DNA [172]. A 3:1 w/w ratio of pDMAEMA to DNA gave polyplexes of approximately 150 nm diameter that were able to transfect even in the presence of serum [173]. Additionally, the pDMAEMA-DNA complexes were found to be stable for over 10 months at 4⬚C and 20⬚C, but became unstable at higher temperatures [174]. 2. Noncondensing Polymers Poly(Ethylene-co-Vinyl Acetate). Mathiowitz et al. [104] showed that DNA was released from poly(ethylene-co-vinyl acetate) (EVAc) without degradation and retained the ability to transfect cells in vitro. Luo et al. [105] also showed that both small and large DNA molecules were encapsulated and successfully released from EVAc matrices. However, the studies ended up only in a model experiment in vitro, and the dynamic phase in vivo remained unexplored. Poly(Vinyl Alcohol) and Poly(N-Vinylpyrrolidone). The use of protective, interactive, and noncondensing (PINC) polymers by Gene Medicine (The Woodlands, TX) and other companies is a novel trend in gene delivery. (Poly(Vinyl alcohol) (PVA) and Poly(N-vinylpyrrolidine (PVP) are amphiphilic molecules having a hydrophilic and a lipophilic portion. Hydrophilic portions of these polymers are known to interact with the plasmid DNA by hydrogen bonding through hydrogen bond acceptors or donor interactions. The hydroxyl groups of PVA serve as hydrogen bond donors, while the nitrogen atom in PVP acts as an acceptor. The hydrophobic monomer units form a hydrophobic coating of vinyl backbone around the DNA molecule [175]. Poly(N-vinylpyrrclidone) formulations result in improved dispersion of plasmid through the extracellular matrix of solid tissue and are hyperosmotic in nature. A comparative study of
348
Kaul and Amiji
unmodified plasmid versus PVP-entrapped plasmid showed a greater number (ten fold enhancement) and distribution of cells transfected with the entrapped plasmid. Also PVP was found to protect the DNA from nuclease degradation and facilitate cellular uptake by hydrophobic interactions with the cell membranes [176]. Addition of PEO chains to PVP-plasmid was found to facilitate cellular uptake. Pluronic威 and Tetronic威 Copolymers. Manufactured by BASF Corp. (Parsipanny, NJ), Pluronic and Tetronic are copolymers of PEO and poly(propylene oxide) arranged either in ABA or star configuration. These also fall into the category of PINC as these systems are neutral at physiological pH. Astafieva et al. [178] prepared poly(N-ethyl-4-vinylpyridium bromide) (PEVP – plasmid) complexes with Pluronic-85. Administration of PEVP-plasmid complexes in the presence of 0.1% Pluronic-85 increased uptake two to three fold in NIH-3T3, MDCK, and Jurkat cell lines, probably through an additional endosomolytic effect of Pluronic-85 in eukaryotic cells. Additionally, polyethylene oxide–block–poly(N-methyl-4-vinylpyridinium sulfate) (PEO-b-PVP) synthesized by sequential anionic polymerization followed by quaternization with dimethyl sulfate selectively binds after recognizing tertiary structure of DNA in both linear and supercoiled DNA [177]. 3. Thermosensitive Polymers Poly(N-Isopropylacrylamide). pNIPAAm is a random copolymer of N′,N-dimethylacrylamide (DMA or DMAEMA) and N-isopropylacrylamide with and other comonomers like butyl methacrylate. Temperature changes can alter the swelling behavior of certain polymers. pNIPAAm is a neutral, water soluble thermosensitive polymer that is soluble at temperatures below 31–32⬚C, but undergoes a phase transition to a solid precipitate at lower critical solution temperature (LCST) above this range [179]. The LCST of these copolymers can be controlled from lower to higher temperatures by altering the composition of NIPAAm, and by incorporating other comonomers like butyl methacrylate (BMA), N′,N-dimethylacrylamide (DMA or DMAEMA) composition [179]. DMAEMA, which is a hydrophilic comonomer, causes an increase in LCST of this copolymer. Hinrichs et al. [180] formulated and evaluated ⬃200-nm sized complexes of plasmid DNA with poly(DMAEMA-co-NIPAAm). It was observed that the size of the complexes increased with increase in the NIPAAm concentration. Formulations with higher molecular weight poly(DMAEMA-co-NIPAAm) or lower ratios of NIPAAm with plasmid formulations at 37⬚C were comparatively stable compared to other weight fractions of poly(DMAEMA-co-NIPAAm). The zeta potential of these formulations was dependent on the copolymer/pDNA ratios and NIPAAm content of the copolymer. Poly(Ethylene Oxide) Copolymers. Poloxamers (or Pluronics) are ABA-type nonionic copolymers of PEO/poly(propylene oxide)/PEO. Some of the poloxamers such as poloxamer407, have unique properties of temperature-induced reversible ‘‘gel-sol’’ transition [181]. These systems are found to significantly increase gene expression in skeletal muscle, probably by offering plasmid protection from enzymes and enhancing distribution of plasmid DNA in the tissue [182]. Poloxamer concentration in a formulation is of importance as studied in a PC: CHO and a PC: CHOL: PEG DSPE based liposomal formulations. A dispersion of liposomes with 2% concentration of the poloxamer resulted in leakage of the oligonucleotide from the liposome, whereas increasing the concentration to 27% poloxamer gel significantly decrease the oligonucleotide diffusion. Since poloxamers are nonbiodegradable and may cause hypersensitivity reactions, Jeong et al. [184,185] have synthesized a series of diblock and triblock biodegradable thermosensitive polymers comprising of PEG or PEO, and polyesters like poly(D,Llactic acid) or PLGA. Aqueous solutions of these copolymers showed temperature-dependent sol-gel transition in the 30–35⬚C range. These transitions were functions of polymer composition,
Polymeric Gene Delivery Systems
349
concentration, and molecular weight of the PEG. A gel phase was found to exist at concentrations above 16% by weight. This formulation with desired properties of biodegradability and sol-gel transition is ideal for development of injectable systems.
IV. SPECIALIZED POLYMER-BASED GENE DELIVERY SYSTEMS The specialized polymeric systems use a complex multicomponent approach to increase efficiency of gene delivery with less toxicity. Summary of the specialized polymeric systems is shown in Table 3. A. Lipopolyplexes Cationic liposomes are prepared using a combination of a neutral phospholipid and one of the cationic lipids like DOTMA, DC-Chol, DOSPA, and so on. Among the many studies carried out with combinations of plasmid DNA with cationic lipids, it was observed that interactions of DC-Chol liposomes with pDNA produced large particles with poor transfection capabilities [186]. On the other hand, small (⬍100 nm) particles were formed when PLL was used to precondense DNA prior to mixing with the lipid. The PLL-DNA-lipid complexes (lipopolyplexes) also had much higher transfection efficiency as compared to lipoplexes. Protamine is another example of a cationic system used to generate reduced-sized liposome/DNA complexes [187]. Another study also reported six times higher transfection efficiency when DOPE was mixed with lipo-PLL as compared to lipo-PLL-DNA complexes without DOPE [188]. Histone H1 (nuclear binding protein) can condense DNA well and be used as a DNA carrier. These are unfortunately prone to aggregation under physiological conditions. Thus anionic lipids such as phosphatidylserine (PS), phosphatidylglycerol (PG), or phosphatidic acid (PA) have been added to histone H1/DNA complexes [189]. Neutral lipid like DOPE has also been added to H1/DNA complexes to increase efficiency. B. Encapsulation Systems: Nanoparticles and Microspheres Nanoparticulate and microparticulate systems are attractive methods of DNA delivery because of their versatility, ease of preparation, and protection of the encapsulated plasmid DNA. These
Table 3 Specialized Polymeric Gene Delivery Systems Polymeric System Lipopolyplexes Encapsulated systems such as nanospheres and microspheres Polymeric micellar systems Polymers with multi functional moieties Polymers with targeting ligands Polymers with endosomolytic systems Polymers with nuclear localization signal (NLS) Polymeric matrices for gene delivery in tissue engineering Polymeric vaccine delivery systems
350
Kaul and Amiji
carrier systems can efficiently encapsulate the DNA and protect it during transit in the systemic circulation. They can also be targeted to reach specific tissues and cells in the body and avoid uptake by the mononuclear phagocytic system after systemic administration through the use of cell-specific ligands and attachment of PEG chains on the nanoparticle surface. Nanoparticles usually have a high surface area–to–volume ratio and thus are able to efficiently encapsulate DNA even without a precondensing step. Nanoparticles can also be made to reach a target site by virtue of their size and charge. Microspheres can be used to direct DNA to specific cells in the body such as for the delivery of DNA vaccines to professional antigen presenting cells like macrophages [227]. Lastly, for industrial production, these systems are amenable to scale-up and manufacturing under the GMP guidelines [87]. DNA-containing microspheres and nanoparticules have been prepared with natural polymers like gelatin [83–87], chitosan [64,66–69,73], and alginates [79,80,106] as well as synthetic polymers like poly(-amino esters) [89] and PLGA [95]. C. Polymer-Based Micellar Systems Micelles are assemblies of amphiphilic molecules with a hydrophilic exterior and a hydrophobic interior in aqueous solution. Block copolymers made up of hydrophobic and hydrophilic segments have micelle forming potential when they are dissolved above their critical micelle concentration (CMC) values. Polymeric micelles are formed by electrostatic interactions between positively charged block copolymers and the negatively charged DNA molecule [192]. Although the DNA-micelle complexes are efficient in transfecting cells in culture, they are poor in diffusing though a multicellular layer or escaping from blood vessels [192–194]. Cationic polymers or detergents that can form micelles can also condense DNA into discrete particles of single nucleic molecule. Harada-Shiba et al. [193] tested nanometer-sized polyionic complex micelles of PLL and PEG in an in vivo turnover study. Southern blot analysis confirmed that the supercoiled DNA was observed for 30 min and the linear form for 3 h following intravenous administration. The polymeric micelles with shorter PLL chain length showed better stability in vivo, probably because the DNA remained intact. These ligand-free micelles were able to transfect Hep G2 cells at charge ratios of 1:2 and 1:4 Preincubation with free copolymer was found to inhibit the expression of the reporter gene, suggesting that the block copolymer adsorption to the cell surface blocked the interaction site of the polymeric micelles. When injected through the supramesentric vein, these polymeric micelles exhibited gene expression in the liver for up to 3 days. The assembly of plasmid DNA with PEO-b-PLL into a core shell structure takes place at equimolar ratios of lysine to phosphate groups, forming highly ordered structures with a spherical shape and nanometer size. To stabilize these core structures, Kataoka et al. [194] crosslinked the pDNA with the thiols in the lysine side chains of PEG-PLL. These micelles had higher transfection efficiencies in cultured cells compared to either free or PLL-associated pDNA. Polymeric micelles are also made using lipopolyamine–plasmid DNA complexes. Lipopolyamines are amphiphiles with a self-aggregating hydrocarbon tail linked to a polycationic head group. Enhanced transfection using lipopolyamine–plasmid DNA micelles was demonstrated by Pitard et al. [195]. Addition of pDNA to lipopolyamine solutions formed hexagonal structures of tubular micelles of approximately 5 nm diameter. The transfection efficiency of the micellar structures was 400 times higher than DNA complexes formed with cationic liposomes of lipopolyamine/DOPE or DMRIE/cholesterol. D. Gene Delivery Systems Utilizing Polymers with Multifunctional Moieties In order to enhance transfection efficiency with nonviral systems, a number of investigators have prepared polymers with multifunctional groups with specific functions. Some of these
Polymeric Gene Delivery Systems
351
functions include ability to condense DNA and protect it from nucleases, target the cell population, enhance cell uptake, lyse the endosome/lysosome compartment, and allow for efficient translocation into the nucleus [196]. Polymer/DNA complexes can be attached to the cell membrane by receptor proteins, escape from endosomal degradation by endosomolytic peptides or osmosis, and possibly obtain nuclear entry via an active mechanism. Pinocytosis, adsorptive endocytosis, receptor mediated endocytosis, and phagocytosis are some of the ways that uptake of pDNA/polymer complexes occurs [197]. Cell-specific-targeting ligand-conjugated polymeric gene carriers are capable of delivering gene to the specific receptors via receptor-mediated endocytosis. Endosomal transit may degrade DNA in the lysosomes; hence endosomal destabilizing molecules like influenza hemagglutinin protein (HA 2) and buffering substances that cause osmotic disruption like chloroquine are employed [198,199]. Nuclear localization signal (NLS) – conjugated polymers are also in use to enhance nuclear entry via protein recognition [200,201]. 1. Polymeric Carriers with Targeting Ligands Cell-specific-targeting ligands when conjugated to polymer backbone provide receptor binding to a specific cell type. Transferrin; monoclonal antibodies; sugars like mannose, galactose, lactose, and folic acid; low density lipoproteins; and RGD peptide are some examples of the targeting moieties that have been conjugated to polymers [202,203]. Galactosylated PLL was synthesized for delivery and expression of genes into mouse hepatocytes [204]. Mannose-linked PLL has also been demonstrated to express genes in murine macrophages isolated from peritoneal exudates in vitro and macrophages in the liver and spleen of adult animals [205]. Lactose-PEGgrafted-PLL was able to overcome undesirable aggregation of PLL/DNA complexes and enhance cell targeting [206,207]. The lactose-PEG-PLL pDNa complexes were nontoxic, compact structures of 100–200 nm diameter and protected the DNA from enzymatic degradation. These lactosylated carriers also exhibited better solubility and were able to specifically target asialoglycoprotein receptors on hepatocytes. They were seven fold more efficient in transfection relative to the PLL-DNA complexes in Hep G2 cell line. Stearyl-PLL, low density lipoprotein (LDL), and PLL were combined into a system for gene delivery [206]. Stearyl-PLL synthesized from PLL (⬃50 kDa) and stearyl bromide in the presence of triethylamine formed a supramolecular ‘‘terplex’’ system that balanced out the charge interactions between stearyl-PLL, LDL, and pDNA. Adjusting the ratio between charged stearyl-PLL and hydrophobic LDL could optimize transfection results [207]. This terplex system showed higher transfection efficiency and lower cytotoxicity compared to liposome-mediated gene transfer in Hep G2 and A7R5 cell lines. Yu et al. [208] showed the ability of the terplex system to deliver luciferase reporter gene as well as therapeutic genes like hrVEGF165 cDNA into bovine aortic artery walls by receptor-mediated endocytosis. The human T-cell thymocyte-differentiation antigen JL-1 is present on most T-cell-derived acute lymphoblastic lymphoma, like the Molt 4 cell line. Suh et al. [209] demonstrated that the transfection efficiency of anti-JL1 antibody g–PLL/pDNA complexes in Molt 4 cells was significantly higher than that of PLL/pDNA complexes. 2. Polymers with Endosomolytic Systems After attachment to the cell membrane, the polymer/DNA complexes are internalized into structures called the coated pits through the endocytotic pathway [210]. Hemagglutinin, a membrane fusion protein of the influenza virus, mediates pH-independent fusion between the viral envelope and the endosome membrane on receptor-mediated endocytosis [211] Also used are synthetic peptides like the HA-2 [212], a 20–amino acid peptide (Gly-Leu-Phe-Gly-Ala-Ile-Ala-Gly-Phe-
352
Kaul and Amiji
Ile-Glu-Asn-Gly-Trp-Glu-Gly-Met-Ile-Asp-Gly) that can act as a potential endosome releasing agent. A synthetic peptide that resembles the fusion peptide of the influenza virus has a structure of an amphiphilic ␣-helix and induces membrane fusion under acidic conditions of pH 5.0–5.5 [213,214]. Other endosomolytic peptides include GALA and KALA [215], both of which comprise 30 amino acid residues. Poly(-amino esters) synthesized by Lynn and Langer [88] are also pH responsive biodegradable polymers that dissolve at early endosomal pH and have the potential for use as endosomolytic polymers. Murthy et al. [216] synthesized pH-sensitive polymers like poly(ethyl acrylic acid) (PEEAc) that disrupts red blood cells (RBCs). PEAAcs hemolytic activity increases rapidly as the pH falls from 6.3 to 5.0, with no hemolytic activity observed at pH 7.4. Increasing the hydrophobicity of this polymer by adding a methene group in the side chain as in poly(propyl acrylic acid) was found to disrupt RBCs 15 times more efficiently. Poly(propyl acrylic acid) was also inactive at pH 7.4 and showed a pH-dependent hemolytic activity. The imidazole group of histidine has a pKa of ⬃6.0 and thus acquires a positive charge in a slightly acidic medium. This cationic charge confers endosome-disrupting properties to this poly(amino acid) by virtue of endosomal buffering capacity [141,217,218]. Conjugation of poly(L-histidine) to PLL shows higher transfection in 293 T-cells than PLL alone [219]. PEI can also enhance endosomal disruption by osmotic swelling of the endosomes [220]. 3. Polymers with Nuclear Localization Signal Nuclear membranes of most eukaryotic cells are only permeable to solutes with a diameter of up to 9 nm. However, active transport can allow for large molecules to cross the nuclear membrane. The small peptide sequences that are recognized by the proteins on the nuclear membrane and bind to receptors in the cytoplasm are termed the nuclear localization signals (NLS) [221]. The most basic peptide derived from simian virus SV-40 large tumor antigen (Pro-Lys-LysLys-Arg-Lys-Val) is a NLS that helps in binding of the karyophilic protein [221]. NLS sequences containing peptides that contain this sequence, such as P-101 (Cys-Gly-Pro-Gly-Ser-Asp-AspGlu-Ala-Ala-Ala-Asp-Ala-Gln-His-Ala-Ala-Pro-Pro-Lys-Lys-Lys-Arg-Lys-Val-Gly-Tyr) and its mutant P-101T (Cys-Gly-Pro-Gly-Ser-Asp-Asp-Glu-Ala-Ala-Ala-Asp-Ala-Gln-His-Ala-AlaPro-Pro-Lys-t-Lys-Lys-Arg-Lys-Val-Gly-Tyr), have been covalently linked to PLL [201]. This has been found to enhance transfection efficiency over PLL or P-101T-linked PLL alone. NLSbinding importin units recognize these complexes and translocate them to the nucleus [222]. Because the electrostatic interactions of the NLS polymer with the pDNA is weak under physiological conditions, peptide nucleic acids (PNAs) that have amide bonds in their backbone instead of phosphate bounds have been used. They form very strong bonds with the NLS sequences under physiological conditions. Linking SV-40 NLS to PNA followed by hybridization with DNA allows NLS to mediate nuclear transport of DNA. Zanta et al. [223] have demonstrated that a single NLS peptide mediates better transport to the nucleus than multiple NLS peptides. Branden et al. [224] have shown that sequence-specific PNA with a NLS peptide can hybridize DNA and increase transfection efficiency eight fold. E. Polymer Matrices for Gene Delivery in Tissue Engineering Gene-based tissue engineering involves the incorporation of genes into polymeric matrices that can be injected or implanted to promote tissue regeneration. Three-dimensional polymeric matrices are frequently being used to transplant cells, create potential space for tissue development, and provide mechanical support for the developing tissues. Major limitations of this therapy is the need to isolate and multiply the cells of interest in culture in order to form a 3D matrix and
Polymeric Gene Delivery Systems
353
the poor survival of cells following transplantation. One approach to enhance cell survival on the tissue engineered scaffold is to locally deliver inductive growth factors, like bone morphogenic protein (BMP). In many instances, however, the encapsulated protein undergoes degradation, either during incorporation or release, and is affected by changes in temperature, pH, and/ or polymeric degradation products [225]. Thus to address some of these issues, gene therapy approaches for local protein delivery are being investigated for tissue regeneration. Plasmid DNA incorporated into biodegradable polymers like collagen, PLGA, alginates, and PINC matrices have been developed [225]. Collagen type 1 was initially used to deliver plasmids encoding for BMP and fragments of human parathyroid hormone genes. The collagen-DNA mixture termed a gene activated matrix (GAM) demonstrated repairs of femur and shaft defects at 9 weeks. Shea et al. [99] incorporated plasmid DNA encoding platelet-derived growth factor (PFGF) directly into PLGA matrices and found that the pDNA was released from the matrices over a period ranging from days to a month in vitro and led to matrix deposition and blood vessel formation in developing tissues in vivo. Bonadio et al. [108,109] found that the implantation of polymeric matrices at the site of bone injury was followed by retention and expression of human parathyroid hormone for at least 6 weeks and induction of a normal new bone in a stable, reproducible, and dose- and timedependent manner. Ochiya et al. [79] have developed a pDNA system using biocompatible atelocollagen for both local and systemic release of pDNA with HST-1/FGF-4 cDNA. This minipellet device released the pDNA for up to 6 h postinjection. Adequate amounts of the HST1 protein were detected at the site of the muscle with some in the brain, lung, liver and kidney. F. Polymeric Vaccine Delivery Systems DNA vaccine systems are generally particulate in nature such as emulsions, microparticles, and liposomes. Their function is to present the DNA they carry to antigen-presenting cells like the macrophages and dendritic cells that play a major role in antigen presentation and activation of specific immunity. A novel formulation containing a fluorescent-labeled plasmid in biodegradable PLGA microparticles containing a cationic surfactant cetyltrimethylammonium bromide (CTAB) was prepared and tested both in vitro and in vivo for DNA delivery into the macrophages and dendritic cells [226].
V. SYSTEMIC AND CELLULAR BARRIERS TO POLYMERIC GENE DELIVERY SYSTEMS A. Systemic Barriers Cationic lipid and polymer formulations form DNA complexes through charge–charge interactions and thus protect the DNA from enzymatic and hydrolytic degradation [227]. As the cationic systems are usually used in excess, the resulting complex is usually positively charged. Although this charge facilitates endocytosis in vitro, it also renders these complexes prone to nonspecific interactions with cellular components such as lipids and proteins (Fig. 2) [227,231–233]. Adsorption of serum proteins onto lipoplexes can cause charge reversal. Charge reversal can be minimized by either removing or neutralizing the adsorbed proteins or by including cholesterol in the lipoplex formulation to reduce plasma protein binding [228–230]. Proteins that can strongly interact with lipoplexes are albumin, lipoproteins, and macroglobulin. Albumin can also release DNA from PLL polyplexes [234]. RBCs and platelets can interact with DNA complexes as well. PEI polyplexes can cause extensive RBC aggregation in vitro. If formulated in high ionic strength media, polyplexes can cause lung embolism because of aggregation [155]. Particulate
354
Kaul and Amiji
Figure 2 Barriers to intracellular delivery of nonviral vectors. (1) Adsorption of vector onto the plasma membrane followed by fusion, leading to complete vector/pDNA separation. (2) Endocytosis in the early endosome. (3) Late endosome at subsequently lower pH. (4) Lysosomal degradation. (5) Endosomal escape that may lead to cytosolic degradation. (6) Translocation into nucleus.
polymeric vectors (⬃ 500–1000 nm) can only cross over from the systemic circulation into organs that are fenestrated to a large degree like the liver or certain tumors with abnormal vasculature [235–240]. Because of the leaky vasculature in tumors, the extravasation is employed to passively target the particulate polymeric vectors into perivascular space. This effect is termed the enhanced permeability and retention effect of solid tumors [241]. Whereas most of the intravenously administered naked DNA ends up in lung, liver, and blood, complexation of DNA to polymers results in the lung being the primary organ of accumulation. The transgene expression correlates well with the biodistribution profile of the complexed plasmid and thus is usually higher in the lung [239]. For minimizing nonspecific interactions, one of the promising approaches appears to be that of creating a steric barrier by incorporating hydrophilic polymers like PEG onto the carrier to minimize protein adsorption. The coating of polyplexes with PEG and similar hydrophilic polymers has been found to reduce nonspecific interactions, reduce accumulation in blood, and increase levels of transgene expression in solid tumors [242]. Since the reduction of nonspecific interactions with cells also reduces the overall uptake of polyplexes, polymeric carriers have been modified using receptor-specific ligands such as transferrin for specific targeting to rapidly growing cells [155]. Another approach to circumvent problems of systemic administration is to administer the polyplexes directly in or near the region of interest. For instance, intratumoral administration of polyplexes has been used for suicide gene therapy with moderate success.
B. Cellular Barriers and Intracellular Trafficking Cellular barriers to polyplexes are generally a series of steps that involve initial binding, internalization through endosomal uptake, degradation the endosome/lysosome compartment, escape from the endosome/lysosome compartment, and translocation to the nucleus.
Polymeric Gene Delivery Systems
355
1. Cellular Binding and Internalization A positive correlation has been found to exist between the excess positive charge on the polyplex and cellular uptake by electrostatic interactions between the polyplexes and the negatively charged membrane-associated proteoglycans [244,245,257]. The nonspecific interactions tend to overpower the ligand-receptor binding and thus interfere with specific cell type targeting both in vitro and in vivo. 2. Endosomal Processing It is important for the DNA or DNA-polymer complex to escape from the endosome before the endosomal–lysosomal fusion occurs. This enables the DNA to remain stable from degradation in the acidic and enzymatic milieu. Unlike lipoplexes, where dissociation of the pDNA into the cytosol from the lipid is essential, release from the complex does not seem to be a precondition for polyplexes. For endosomal escape, polyplexes may employ the use of PEI-based polymers that cause osmotic lysis of the endosome by the proton sponge effect or substitute PLL side groups with histidyl groups to confer the buffering property at low pH [141,149,199,249]. Endosomolytic peptides and chloroquine have also been used to enhance gene expression in combination with PLL-based polyplex systems, where they act together through a synergistic effect [199,248]. Adenoviral peptides (e.g., HA 2) have also been found to be useful as they can be made to selectively lyse a membrane at a specific pH [252–254]. 3. Trafficking in Cytosol In contrast to lipoplexes, polyplexes do not dissociate in the endosome and are thus released as a complex. Therefore, during cytosol trafficking, the DNA does not undergo appreciable nucleaseinduced degradation. Improvement of gene expression after cytoplasmic injection of polyplexes has been reported with high molecular weight PLL (199). PEI was also found to be better than PLL in terms of nuclease protection of the pDNA in the cytosol [248]. 4. Nuclear Translocation The exogenous DNA in polyplexes has to gain access to nuclear transcription machinery in order to effectively express the protein. The probability of a plasmid being expressed is almost 1 in 104 –105 plasmids that were taken up by the cell [255–257]. The two routes for cytoplasmic DNA to gain ingress to the nucleus are 1. Breakdown of nuclear membrane during mitosis. PLL polyplexes were found to enter the nucleus during mitosis [258]. This method of entry is more dependent on the cell cycle. 2. Nuclear pore transport. PLL and PEI have both been reported to promote nuclear transport. PLL temporarily stabilizes the plasmid in the cytoplasm and may increase the total amount of intact plasmid at the nuclear membrane during mitosis. PEI accumulates around nucleus to enhance transport of complexed DNA [259].
VI. CONCLUDING REMARKS The gene transfer efficiency of polymeric vectors still remains significantly lower than the viral vectors. It is still important to overcome the various barriers of intracellular trafficking such as endosomal release, cytoplasmic transit, and nuclear translocation of polyplexes. Biocompatibil-
356
Kaul and Amiji
ity, immunological considerations, and other toxicity issues are still important factors that need to be taken into account while developing both new polymers and modifying the old ones for gene delivery applications. The in vitro transfection studies with polymeric vectors need to critically examine such to ensure there is a positive correlation between in vitro and in vivo gene expression. A structure – activity relationship needs to be charted out for relating the polymer structures and their transfection properties. Finally, scale-up and industrial manufacturing of polyplexes under regulatory guidelines is still an important task to be considered for the near future. Despite all of these hurdles, polymeric systems offer a versatility that is unmatched by any other type of gene delivery method. Using a combinatorial synthesis approach, many investigators have developed polymers that satisfy the physical, chemical, and biological requirements for effective gene delivery. Through rational approaches, polymeric systems can be used to create multifunctional DNA delivery systems for cell and tissue targeting and efficient transfection. In addition, different types of polymeric carriers such as nanoparticles, microspheres, scaffolds for tissue engineering, and hydrogel systems can be designed to improve delivery properties and localize the DNA at the target site. Overall, the future of polymeric gene delivery systems looks very bright. REFERENCES 1. Ledley FD. Non-viral gene therapy. Curr. Opin. Biotechnol 1994; 5:626–636. 2. www.haverford.edu/biology/HHMI/Definition.html. 3. Fenske BD, MacLachlan I, Cullis PR. Long-circulating vectors for the systemic delivery of genes. Curr. Opin. Mol. Ther 2001; 3:153–158. 4. Orkin SH, Motulsky AG;. 5. Lehrman S. Virus treatment questioned after gene therapy death. Nature 1999; 401:517–518. 6. Marshall E. Gene therapy on trial. Science 2000; 288:951–956. 7. Fox JL. Gene therapy safety issues come to fore. Nat. Biotechnol 1999; 17:1153. 8. Blessing T, Remy JS, Behr JP. Monomolecular collapse of plasmid DNA into stable virus-like particles. Proc. Natl. Acad. Sci. USA 1998; 95:1427–1431. 9. Hui KM, Chia TF. Eradication of tumor growth via biolistic transformation with allogeneic MHC genes. Gene Ther 1997; 4:762–767. 10. Sawamura D, Ina S, Itai K. In vivo gene introduction into keratinocytes using jet injection. Gene. Ther 1999; 6:1785–1787. 11. Oshima Y, Sakamoto T, Yamanaka I, Nishi T, Ishibashi T, Inomata H. Targeted gene transfer to corneal endothelium in vivo by electric pulse. Gene Ther 1998; 5:1347–1354. 12. Lauer U, Burgelt E, Squire Z, Hans P, Gregor M, Delius M. Shockwave permeabilization as new gene transfer method. Gene Ther 1997; 4:710–715. 13. Wolff JA, Malone RW, Williams P, Chong W, Acsadi G, Jani A, Felgner PL. Direct gene transfer into mouse muscle in vivo. Science 1990; 247:1465–1468. 14. Niven R, Pearlman R, Wedking T, Mackeigan J, Noker P, Simpson-Herren L, Smith JG. Biodistribution of radiolabeled lipid-DNA complexes and DNA in mice. J Pharm. Sci 1998; 87:1292–1299. 15. Houk BE, Hochhaus G, Hughes JA. Kinetic modeling of plasmid DNA degradation in rat plasma. AAPS PharmSci 1999; 1(3): article 9. http://www.pharmsci.org/scientificjournals/pharmsci/journal/ 99_9.html. 16. Zhang G, Vargo D, Budker V, Armstrong N, Knechtle S, Wolff JA. Expression of naked plasmid DNA injected into the afferent and efferent vessels of rodent and dog livers. Hum. Gene Ther 1997; 8:1763–1772. 17. Liu F, Song YK, Liu D. Hydrodynamics-based transfection in animals by systemic administration of plasmid DNA. Gene Ther 1999; 6:1258–1266.
Polymeric Gene Delivery Systems
357
18. Zhang G, Budker V, Wolff JA. High levels of foreign gene expression in hepatocytes after tail vein injections of naked plasmid DNA. Hum. Gene Ther 1999; 10:1735–1737. 19. Davis HL, Michel ML, Mancini M, Schleef M, Whalen RG. Direct gene transfer in skeletal muscle: plasmid DNA-based immunization against the hepatitis B virus surface antigen. Vaccine 1994; 12: 1503–1509. 20. Furth PA, Shamay A, Wall RJ, Hennighausen L. Gene transfer into somatic tissues by jet injection. Anal. Biochem 1992; 205:365–368. 21. Rosengart TK, Lee LY, Patel SR, Sanborn TA, Parikh M, Bergman GW, Hachamovitch R, Szulc M, Kligfield PD, Okin PM. Angiogenesis gene therapy: phase I assessment of direct intramyocardial administration of an adenovirus vector expressing VEGF121 cDNA to individuals with clinically significant severe coronary artery disease. Circulation 1999; 100:468–474. 22. Rakhmilevich AL, Timmins JG, Janssen K. Gene gun–mediated IL-12 gene therapy induces antitumor effects in the absence of toxicity: A direct comparison with systemic IL-12 protein therapy. J. Immunother 1999; 22:135–144. 23. Felgner PL, Gadek TR, Holm M, Roman R, Chan HW, Wenz M, Northrop JP, Ringold GM, Danielsen M. Lipofection: a highly efficient, lipid-mediated DNA-transfection. Proc. Natl. Acad. Sci 1987; 84:7413–7417. 24. Schwartz B, Benoist C, Abdallah B, Scherman D, Behr JP, Demeneix BA. Liposmermin-based gene transfer into the newborn mouse brain is optimized by a low lipospermine/DNA charge ratio. Hum. Gene Ther 1995; 6:1515–1524. 25. Felgner JH, Kumar R, Sridhar CN, Wheeler CJ, Tsai YJ, Border R, Ramsey P, Martin M, Felgner PL. Enhanced gene delivery and mechanism studies with a novel series of cationic lipid formulations. J. Biol. Chem 1994; 269:2550–2561. 26. McLachalan G, Davidson H, Davidson D, Dickinson P, Dorin J, Porteus D. DOTAP as a vehicle for efficient gene delivery system in vitro and in vivo. Biochemistry 1994; 11:19–21. 27. Gao X, Huang L. A novel cationic liposome reagent for efficient transfection of mammalian cells. Biochem. Biophys. Res. Commun 1991; 179:280–285. 28. Egilmez NK, Cuenca R, Yokota SJ, Sorgi F, Bankert RB. In vivo cytokine gene therapy of human tumor xenografts in SCID mice by liposome-mediated DNA delivery. Gene Ther 1996; 3:607–614. 29. Delepine P, Guillaume C, Floch V, Loisel S, Yaouanc JJ, Clement JC, Des Abbayes H, Ferec C. Cationic phosphonolipids as nonviral vectors: in vitro and in vivo applications. J. Pharm. Sci 2000; 89:629–638. 30. Farhood H, Gao X, Son K, Yang YY, Lazo JS, Huang L. Cationic liposomes for direct gene transfer in therapy of cancer and other diseases. Ann. NY Acad. Sci 1994; 716:23–34. 31. Riemer DL, Kong S, Monck M, Wyles J, Tam P, Wassan EK, Sally MB. Liposomal lipid and plasmid DNA delivery to B16/BL6 tumors after intraperitoneal administration of cationic liposome DNA aggregates. J. Pharmacol. Exp. Therap 1999; 289:807–815. 32. Plautz GE, Yang ZY, Wu BY, Gao X, Huang L, Nabel G. Immunotherapy of malignancy by in vivo gene transfer into tumors. Proc. Natl. Acad. Sci. USA 1993; 90:4645–4649. 33. Kaneda Y, Iwai K, Uchida T. Increased expression DNA cointroduced with nuclear protein in adult rat liver. Science 1989; 243:375–378. 34. Kato K, Kaneda Y, Sakurai M, Nakanishi M, Okada Y. Direct injection of hepatitis B virus DNA into liver induced hepatitis in adult rats. J Biol. Chem 1991; 268:22071–22074. 35. Kaneda Y, Iwai K, Uchida T. Introduction and expression of the human insulin gene in adult rat liver. J Biol Chem 1989; 64:12126–12129. 36. Kato K, Nakanishi M, Kaneda Y, Uchida T, Okada Y. Expression of hepatitis B virus surface antigen in adult rat liver: cointroduction of DNA and nuclear protein by a simplified liposome method. J Biol. Chem 1991; 266:3361–3364. 37. Boussif O, Lezoualch F, Zanta MA, Mergny MD, Scherman D, Demeneix B, Behr JP. A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: polyethyleneimine. Proc. Natl. Acad. Sci. USA 1995; 92:7297–7301. 38. Godbey WT, Wu KK, Mikos AG. Tracking the intracellular path of poly(ethyleneimine)/DNA complexes for gene delivery. Proc. Natl. Acad. Sci. USA 1999; 96:5177–5181.
358
Kaul and Amiji
39. Godbey WT, Wu KK, Mikos AG. Poly(ethyleneimine) and its role in gene delivery. J. Control. Rel 1999; 60:149–160. 40. Putnam D, Langer R. Poly(4-hydroxyl-L-proline ester): low temperature polycondensation and plasmid DNA complexation. Macromolecules. 1999; 32:3658–3663. 41. Pack DW, Putnam D, Langer R. Design of imidazole-containing endosomolytic biopolymers for gene delivery. Biotechnol. Bioeng 2000; 67:217–223. 42. Putnam D, Gentry CA, Pack DW, Langer R. Polymer-based gene delivery with low cytotoxicity by a unique balance of side-chain termini. Proc. Natl. Acad. Sci., USA 2001; 98:1200–1205. 43. Leong KW. Biopolymer-DNA nanospheres. In Huang L, Hung MC, Wagner E, eds. Nonviral Vectors for Gene Therapy. San Diego. CA.: Academic Press, 1999:267–287. 44. Wagner E, Zenke M, Cotten M, Beug H, Birnstiel ML. Transferrin–polycation conjugates as carriers for DNA uptake into cells. Proc. Natl. Acad. Sci., USA 1990; 87:3410–3414. 45. Wagner E, Cotton M, Foisner R, Birnsteil ML. Transferrin–polycation DNA complexes: the effect of polycations on the structure of the complex and DNA delivery to cells. Proc. Natl. Acad. Sci., USA 1991; 88:4255–4259. 46. Wu GY, Wu CH. Receptor-mediated in vitro gene transformation by a soluble DNA carrier system. J.Biol.Chem 1987; 262:4429–4432. 47. Wu GY, Wu CH. Evidence of targeted gene delivery to Hep G2 hepatoma cells in vitro. Biochemistry 1988; 27:887–892. 48. Wu CH, Wilson JM, Wu GY. Targeting genes: delivery and persistent expression of a foreign gene driven by mammalian regulatory elements in vivo. J. Biol. Chem 1989; 264:16985–16987. 49. Wu GY, Wu. CH. Receptor mediated gene delivery and expression in vivo. J. Biol. Chem 1988; 263:14621–14624. 50. Haensler J, Szoka FC. Synthesis and characterization of a trisgalactosylated bisacridine compound to target DNA to hepatocytes. Bioconjug. Chem 1993; 4:85–93. 51. Wagner E, Cotten M, Mechtler K, Kirlappos H, Birnstiel ML. DNA-binding transferrin conjugates as functional gene-delivery agents: synthesis by linkage of polylysine or ethidium homodimer to the transferrin carbohydrate moiety. Bioconjug. Chem 1991; 2:226–231. 52. Zenke M, Steinlein P, Wagner E, Cotten M, Beug H, Birnstiel ML. Receptor-mediated endocytosis of transferrin-polycation conjugates: an efficient way to introduce DNA into hematopoietic cells. Proc. Natl. Acad. Sci. USA 1990; 87:3655–3659. 53. Harris CE, Agarwal S, Hu P, Wagner E, Curiel DT. Receptor mediated gene transfer to airway epithelial cells in primary culture. Am. J. Respir. Cell Mol. Biol 1993; 9:441–447. 54. Trubetskoy VS, Torchilin VP, Kennel S, Huang L. Use of N-terminal modified poly-L-lysine–antibody conjugate as a carrier for targeted gene delivery in mouse lung endothelial cells. Bioconjug. Chem 1992; 3:323–327. 55. Wilson JM, Grossman M, Wu CH, Chowdhury NR, Wu GY, Chowdhury JR. Hepatocyte-directed gene transfer in vivo leads to transient improvement of hypercholesterolemia in low-density lipoprotein receptor-deficient rabbits. J. Biol. Chem 1990; 267:963–967. 56. Wu GY, Zhan P, Sze LL, Rosenberg AR, Wu CH. Incorporation of adenovirus into a ligand-based DNA carrier system results in retention of original receptor specificity and enhances targeted gene expression. J Biol. Chem 1994; 269:11542–11546. 57. Cristiano RJ, Smith LC, Woo SLC. Hepatic gene therapy: adenovirus enhancement of receptormediated gene delivery and expression in primary hepatocytes. Proc. Natl. Acad. Sci. USA 1993; 90:2122–2126. 58. Riessen B, Rahlmzadeh H, Blessing E, Takeshita S, Barry JJ, Isner JM. Arterial gene transfer using pure DNA applied directly to a hydrogel coated angioplasty balloon. Hum. Gen. Ther 1993; 4: 749–758. 59. Cristiano RJ, Smith LC, Kay MA, Brinkley B, Woo SLC. Hepatic gene therapy: efficient gene delivery and expression in primary hepatocytes utilizing a conjugated adenovirus-DNA complex. Proc. Natl. Acad. Sci. USA 1993; 90:11548–11552. 60. Fisher KJ, Wilson JM. Biochemical and functional analysis of an adenovirus-based ligand complex for gene transfer. Biochem. J 1994; 299:49–58.
Polymeric Gene Delivery Systems
359
61. Curiel DT. High-efficiency gene transfer employing adenovirus-polylysine-DNA complexes. Nat. Immunol 1994; 13:141–164. 62. Mumper RJ, Wang JJ, Claspell JM, Rolland AP. Novel polymeric condensing carriers for gene delivery. Proc Int Symp Control Rel Bioactive Mater 1995; 22:178–179. 63. Henriksen I, Vagen SR, Sande SA, Smistad G, Karlsen J. Interactions between liposomes and chitosan II: effect of selected parameters on aggregation and leakage. Int. J. Pharm 1997; 146: 193–204. 64. Ilum L, Farraj NF, Davis SS. Chitosan as a novel nasal delivery system for peptide drugs. Pharm. Res 1994; 11:1186–1189. 65. Chandy T, Sharma CP. Chitosan as a biomaterial. Biomaert. Artif. Cells Artif. Org 1990; 18:1–24. 66. Schipper NGM, Varum KM, Artursson P. Chitosans as absorption enhancers for poorly soluble drugs. 1. Influence of molecular weight and degree of acetylation on drug transport across human intestinal epithelial (Caco-2) cells. Pharm. Res 1996; 13:1686–1692. 67. Lee KY, Kwon IC, Kim YH, Jo WH, Jeong SY. Preparation of chitosan self-aggregates as a gene delivery system. J. Control. Rel 1998; 51:213–220. 68. Roy K, Mao HQ, Huang SK, Leong KW. Oral gene delivery with chitosan-DNA nanoparticles generates immunologic protection in a murine model of peanut allergy. Nat Med 1999; 5:387–391. 69. Mao HQ, Roy K-, Le VL, August JT, Leong KW. DNA-chitosan nanospheres: derivatization and storage stability. Proc Int Symp Control Rel Bioactive Mater 1997; 24:671–672. 70. Richardson SC, Kolbe HV, Duncan R. Potential of low molecular mass chitosan as a DNA delivery system: biocompatibility, body distribution and ability to complex and protect DNA. Int. J. Pharm 1999; 178:231–243. 71. Ouchi T, Murata J, Ohya Y. Possibility of application of quaternary chitosan conjugate having pendant galactose residues as gene delivery tool. In: Domard A, Roberts GAF, Roberts K, Varum M, eds. Advances in Chitin Science: Jacques Andre Publisher, 1998:870–877. 72. MacLaughlin FC, Mumper RJ, Wang J, Tagliaferri JM, Gill I, Hinchcliffe M, Roland AP. J. Control. Rel 1998; 56:259–272. 73. Lee M, Nah JW, Kwon Y, Koh JJ, Ko KS, Kim SW. Water-soluble and low molecular weight chitosan-based plasmid DNA delivery. Pharm. Res 2001; 18:427–431. 74. Erbacher P, Zou S, Bettinger T, Steffan AM, Remy JS. Chitosan-based vector/DNA complexes for gene delivery: biophysical characteristics and transfection ability. Pharm. Res 1998; 15:1332–1339. 75. Ramirez F, Wet W. Molecular biology of the human fibrillar collagen genes. Ann. N Y Acad. Sci 1988; 543:109–116. 76. Stenzel KH, Miyata T, Rubin AL. Collagen as a biomaterial. Annu Rev Biophys Bioeng 1974; 3: 231–253. 77. DeLustro F, Condell RA, Nguyen MA, McPherson JM. A comparative study of the biologic and immunologic response to medical devices derived from dermal collagen. J. Biomed. Mater. Res 1986; 20:109–120. 78. Fujioka K, Takada Y, Sato S, Miyata T. Novel delivery system for proteins using collagen as a carrier material: the minipellet. J. Control. Rel 1995; 33:307–315. 79. Ochiya T, Takahama Y, Nagahara S, Sumita Y, Hisada A, Itoh H, Nagai Y, Terada M. New delivery system for plasmid DNA in vivo using atelocollagen as a carrier material: the minipellet. Nat. Med 1999; 5:707–710. 80. Aggarwal N, HogenEsch H, Guo P, Suckow MA, Mittal SK. Biodegradable alginate microspheres as a delivery system for naked DNA. Can. J. Veter. Res 1999; 63:148–152. 81. Mittal SK, Aggarwal N, Sailaja G, van Olphen A, Hogenesch H, North A, Hays J, Moffatt S. Immunization with DNA, adenovirus or both in biodegradable alginate microspheres: effect of route of inoculation on immune response. Vaccine 2000; 19:253–263. 82. Ward AG, Courts A. The Science and Technology of Gelatin. New York: Academic Press, 1977. 83. Veis A. The Macromolecular Chemistry of Gelatin. New York: Academic Press, 1964. 84. Truong VL, August JT, Leong K. Gene delivery by DNA-gelatin nanospheres. Hum. Gene Ther 1998; 9:1709–1717. 85. Truong VL, Walsh SM, Schweibert E, Mao HQ, Guggino JT, Leong KW. Gene transfer by DNAgelatin nanospheres. Arch. Biochem. Biophys 1999; 361:47–56.
360
Kaul and Amiji
86. Leong KW, Mao HQ, Truong-Le VL, Roy K, Walsh SM, August JT. DNA-polycation nanospheres as non-viral gene delivery vehicles. J. Control. Rel 1998; 53:183–193. 87. Kaul G, Amiji M. Long-circulating poly(ethylene glycol)-modified gelatin nanoparticles for intracellular delivery. Pharm. Res 2002; 19 (7):1061–1067. 88. Kaul G, Potineni A, Lynn DM, Langer R, Amiji M. Surface-modified polymeric nanoparticles for tumor-targeted delivery. SurFACTs in Biomaterials 2002; 7(2):1–6. 89. Lynn DM, Langer R. Degradable poly(-amino esters): synthesis, characterization, and self-assembly with plasmid DNA. J. Am. Chem. Soc 2000; 122:10761–10768. 90. Lynn DM, Amiji MM, Langer R. pH-responsive biodegradable polymer microspheres: rapid release of encapsulated material within the range of intracellular pH. Angew Chemie. Int. Ed 2001; 40: 1707–1710. 91. Eldridge JH, Staas JK, Meulbroek JA, McGhee JR, Tice TR, Gilley RM. Biodegradable microspheres as a vaccine delivery system. Mol. Immunol 1991; 28:287–294. 92. Rafati H, Coombes AGA, Adler J, Holland J, Davis SS. Protein loaded poly(DL-lactide-co-glycolide) microparticles for oral administration: formulation, structural and release characteristics. J. Control. Rel 1997; 43:89–102. 93. Hedley ML, Curley JL, Urban RG. Microspheres containing plasmid encoded antigens elicit cytotoxic T cell responses. Nat. Med 1998; 4:365–368. 94. Jones DH, Corris S, McDonald S, Clegg JCS, Farrar GH. Poly(DL-lactide-co-glycolide) -encapsulated plasmid DNA elicits systemic and mucosal antibody responses to encoded protein after oral administration. Vaccine 1997; 15:814–817. 95. Gebrekidan S, Woo BH, DeLuca PP. Formulation and in vitro transfection efficiency of poly (D,Llactide-co-glycolide) microspheres containing plasmid DNA for gene delivery. AAPS PharmSciTech 2000; 1(4). 96. Hsu S, Hao T, Hedley ML. Parameters effecting DNA encapsulation in PLGA based microspheres. J. Drug Target 1999; 7:313–323. 97. Maruyama T, Ishihara JS, Kim SW, Kim W, Akaike T. Nanoparticle DNA carrier with poly(L-lysine) grafted polysaccharide copolymer and poly(D,L-lactic acid). Bioconjug. Chem 1994; 8:735–742. 98. Murphy WL, Mooney DJ. Controlled delivery of inductive proteins, plasmid DNA and cells from tissue engineering matrices. J. Periodontal Res 1999; 34:413–419. 99. Shea LD, Smiley E, Bonadio J, Mooney DJ. DNA delivery from polymer matrices for tissue engineering. Nat. Biotechnol 1999; 17:551–554. 100. Lim YB, Han SO, Kong HU, Lee Y, Park JS, Jeong B, Kim SW. Biodegradable polyester, poly[alpha(4-aminobutyl)-L-glycolic acid], as a non-toxic gene carrier. Pharm. Res 2000; 17:811–816. 101. Lim YB, Kim CH, Kim K, Kim SW, Park JS. Development of safe gene delivery system using biodegradable polymer, poly[␣-(4-aminobutyl)-L-glycolic acid]. J. Am. Chem. Soc 2001; 122: 6524–6525. 102. Lee M, Koh JJ, Han SO, Ko KS, Kim SW. Prevention of autoimmune insulitis by delivery of interleukin-4 plasmid using a soluble and biodegradable polymeric carrier. Pharm. Res 2002; 19: 246–249. 103. Lim YB, Choi YH, Park JS. A self-destroying polycationic polymer: biodegradable poly(4-hydroxyL-proline ester). J. Am. Chem. Soc 1999; 121:5633–5639. 104. Mathiowitz E, Jacob JS, Jong YS, Carino GP, Chickering DE, Chaturvedi P, Santos CA, Vijayaraghavan K, Montgomery S, Bassett M, Morrell C. Biologically erodable microspheres as potential oral drug delivery systems. Nature 1997; 386:410–414. 105. Luo D, Woodrow-Mumford K, Belcheva N, Saltzman WM. Controlled DNA delivery systems. Pharm. Res 1999; 16:1300–1308. 106. Chickering DE, Jacob JS, Mathiowitz E. Poly(fumaric-co-sebacic) microspheres as oral drug delivery systems. Biotechnol. and Bioeng 1996; 52:96–101. 107. Chickering DE, Jacob JS, Desai TA, Harrison M, Harris WP, Morrell CN, Chaturvedi P, Mathiowitz E. Bioadhesive microspheres. 3. An in vivo transit and bioavailability study of drug-loaded alginate and poly(fumaric-co-sebacic anhydride) microspheres. J. Control. Rel 1997; 48:35–46. 108. Bonadio J, Smiley E, Patil P, Goldstein S. Localized, direct plasmid-gene delivery in vivo: prolonged therapy results in reproducible tissue regeneration. Nat. Med 1999; 5:753–759.
Polymeric Gene Delivery Systems
361
109. Bonadio J. Tissue engineering via local gene delivery: update and future prospects for enhancing the technology. Adv. Drug Del. Rev 2000; 44:185–194. 110. Langer R. Selected advances in drug delivery and tissue engineering. J. Control. Rel 1999; 62:7–11. 111. Langer R. Tissue engineering. Science 1993; 260:920–926. 112. Trubetskoy VS, Loomis A, Hagstrom JE, Budker VG, Wolff JA. Layer-by-layer deposition of oppositely charged polyelectrolytes on the surface of condensed DNA particles. Nucleic Acids Res 1999; 27:3090–3095. 113. Felgner PL, Barenholz Y, Behr JP, Cheng SH, Cullis P, Huang L, Jessee JA, Seymour L, Szoka F, Thierry AR, Wagner E, Wu G. Nomenclature for synthetic gene delivery systems. Hum. Gene. Ther 1997; 8:511–512. 114. De Smedt SC, Demeester J, Hennink WE. Cationic polymer based gene delivery systems. Pharm. Res 2000; 17:113–126. 115. Takai T, Ohmori H. DNA transfection of mouse lymphoid cells by the combination of DEAEdextran–mediated DNA uptake and osmotic shock procedure. Biochim. Biophys. Acta 1990; 1048: 105–109. 116. Kabanov AV, Kabanov VA. Interpolyelectrolyte and block ionomer complexes for gene delivery: physico-chemical aspects. Adv. Drug Del. Rev 1998; 30:49–60. 117. Wolfert MA, Schacht EH, Toncheva V, Ulbrich K, Nazarova O, Seymour LW. Characterization of vectors for gene therapy formed by self-assembly of DNA with synthetic block co-polymers. Hum. Gene Ther 1996; 7:2123–2133. 118. Toncheva V, Wolfert MA, Dash PR, Oupicky´ D, Ulbrich K, Seymour LW, Schacht E. Novel vectors for gene delivery formed by self-assembly of DNA with poly(L-lysine) grafted with hydrophilic polymers. Biochim. Biophys. Acta 1998; 1380:354–368. 119. Maruyama A, Katoh M, Ishihara T, Akaike T. Comb-type polycations effectively stabilize DNA triplex. Bioconjug. Chem 1997; 8:3–6. 120. Maruyama A, Watanabe H, Ferdous A, Katoh M, Ishihara T, Akaike T. Characterization of interpolyelectrolyte complexes between double-stranded DNA and polylysine comb-type copolymers having hydrophilic side chains. Bioconjug. Chem 1998; 9:292–299. 121. Katayose S, Kataoka K. Remarkable increase in nuclease resistance of plasmid DNA through supramolecular assembly with poly(ethylene glycol)–poly(L-lysine) block copolymer. J. Pharm. Sci 1998; 87:160–163. 122. Zauner W, Ogris M, Wagner E. Polylysine-based transfection systems utilizing receptor-mediated delivery. Adv. Drug Del. Rev 1998; 30:97–114. 123. Kabanov AV, Kabanov VA. DNA complexes with polycations for the delivery of genetic material into cells. Bioconjug. Chem 1995; 6:7–20. 124. Kichler A, Zauner W, Morrison C, Wagner E. Ligand–polylysine mediated gene transfer. In: Felgner PL, Heller MJ, Lehn P, Behr JP, Szoka FC, eds. Artificial Self-Assembling Systems for Gene Delivery. Washington: ACS, 1996:120–128. 125. Bloomfield VA. DNA condensation. Curr. Opin. Struct. Biol 1996; 6:334–341. 126. Prokop A, Kozlov E, Moore W, Davidson JM. Maximizing the in vivo efficiency of gene transfer by means of nonviral gene delivery vehicles. J. Pharm. Sci 2002; 91:67–76. 127. Gonsho A, Irie K, Susaki H, Iwasawa H, Okuno S, Sugawara T. Tissue-targeting ability of saccharide-poly(L-lysine) conjugate. Biol. Pharm 1994; 17:275–282. 128. Kwoh DY, Lollo CP. Stabilization of Poly-L-lysine/DNA polyplexes for in vivo gene delivery to the liver. Biochim. Biophys. Acta 1999; 1444:171–190. 129. Plank MX, Tang AR, Wolff JA, Szoka FC. 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 1999; 10:319–332. 130. Brown MD, Scha¨tzlein A, Brownlie A, Jack V, Tetley L, Gray AI, Uchegbu IF. Preliminary characterization of novel amino acid based polymeric vesicles as gene delivery agents. Bioconjug. Chem 2000; 11:880–891. 131. Wadhwa MS, Collard WT, Adami RC, McKenzie DL, Rice KG. Peptide mediated gene delivery: influence of peptide structure on gene expression. Bioconjug. Chem 1997; 8:81–88.
362
Kaul and Amiji
132. Boussif O, Delair T, Brua C, Veron L, Pavirani A, Kolbe HVJ. Synthesis of polyallylamine derivatives and their use as gene transfer vectors in vitro. Bioconjug. Chem 1999; 10:877–883. 133. Choi YH, Liu F, Kim JS, Choi YK, Park JS, Kim SW. Polyethylene glycol–grafted poly-L-lysine as polymeric gene carrier. J. Control. Rel 1998; 54:39–48. 134. Zhou X, Klibanov A, Huang L. Lipophilic polylysines mediate efficient DNA transfection in mammalian cells. Biochim. Biophys. Acta 1991; 1065:8–14. 135. Zhou X, Huang L. DNA transfection mediated by cationic liposomes containing lipopolylysine: characterization and mechanism of action. Biochim. Biophys. Acta 1994; 1189:195–203. 136. Kim JS, Maruyama A, Akaike T, Kim SW. A new non-viral DNA delivery vector: the terplex system. J. Control. Rel 1998; 53:175–182. 137. Robinson P, Hawtrey AO, Ariatti M, Govender P, Jagganath D. Effect of polyethylene glycol conjugated to DNA-transfecting complexes targeted at the transferrin receptor of HeLa cells. Drug Del 1997; 4:115–119. 138. Coll JL, Wagner E, Combaret V, Metchler K, Amstutz H, Iacano-di-Cacito I, Simon N, Favrot MC. In vitro targeting and specific transfection of human neuroblastoma cells by chCE7 antibody–mediated gene transfer. Gene Ther 1997; 4:156–161. 139. Shewring L, Collins L, Lightman SL, Hart S, Gustafsson K, Fabre JW. A non-viral vector system for efficient gene transfer to corneal endothelial cells via membrane integrins. Transplantation 1997; 64:763–769. 140. Pouton CW, Lucas P, Thomas BJ, Uduehi AN, Milroy DA, Moss SH. Polycation-DNA complexes for gene delivery—a comparison of the biopharmaceutical properties of cationic polypeptides and cationic lipids. J. Control. Rel 1998; 53:289–299. 141. Midoux P, Monsigny M. Efficient gene transfer by histidylated polylysine/pDNA complexes. Bioconjug. Chem 1999; 10:406–411. 142. Wu GY, Wu CH. Receptor-mediated in vitro gene transformation by a soluble DNA carrier system. J. Biol. Chem 1987; 262:4429–4432. 143. Wu CH, Wilson JM, Wu GY. Targeting genes: delivery and persistent expression of a foreign gene driven by mammalian regulatory elements in vivo. J.Biol.Chem 1989; 264:16985–1698. 144. Cotten M, Langle-Rouault F, Kirlappos H, Wagner E, Mechtler K, Zenke M, Beug H, Birnstiel ML. Transferrin-polycation mediated introduction of DNA into human leukemic cells: stimulation by agents that affect the survival of transfected DNA or modulate transferrin receptor levels. Proc. Natl. Acad. Sci. USA 1990; 87:4033–4037. 145. Zauner W, Ogris M, Wagner E. Polylysine-based transfection systems utilizing receptor-mediated delivery. Adv. Drug Del. Rev 1998; 30:97–114. 146. Nishikawa M, Takemura S, Takakura Y, 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. Pharmacol. Exp. Ther 1998; 287: 408–415. 147. Hashida M, Takemura S, Nishikawa M, Takakura Y. Targeted delivery of plasmid DNA complexed with galactosylated poly(L-lysine). J. Control. Rel 1998; 53:301–310. 148. Boussif O, Lezoualch F, Zanta MA, Mergny MD, Scherman D, Demeneix B, Behr JP. Versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: polyethyleneimine. Proc. Natl. Acad. Sci. USA 1995; 92:7297–7301. 149. Suh J, Paik HJ, Hwang BK. Ionization of poly(ethylenimine) and poly(allylamine) at various pHs. Bioorg. Chem 1994; 22:318. 150. Behr JP, Demeneix BA. Gene delivery with polycationic amphiphiles and polymers. Curr. Res. Mol. Therap 1998; 1:5–12. 151. Goula D, Lemkine G, Becker N, Normandy P, Rodrigues J, Mantero S, Levi G, Demeneix BA. Rapid crossing of the pulmonary endothelial barrier by polyethyleneimine/DNA complex. Gene Ther 2000; 7:499–504. 152. Lemkin G, Goula D, Becker N, Paura L, Levi G, Demeneix BA. Optimization of polyethyleneiminebased gene delivery to mouse brain. J. Drug Target 1999; 7:305–312. 153. Diebold SS, Kursa M, Wagner E, Cotten M, Zenke M. Mannose polyethyleneimine conjugates for targeted DNA delivery into dendritic cells. J. Biol. Chem 1999; 274:19087–19094.
Polymeric Gene Delivery Systems
363
154. Goula D, Benoist C, Montero S, Merlo G, Levi G, Demeneix BA. Polyethyleneimine-based intravenous delivery of transgenes to mouse lung alveolar cells. Gene Ther 1998; 5:1291–1295. 155. Ogris M, Brunner S, Schuller S, Kircheis R, Wagner E. PEGylated DNA/transferrin-PEI complexes: reduced interaction with blood components, extended circulation in blood and potential for systemic gene delivery. Gene Ther 1999; 6:595–605. 156. Li S, Tan Y, Viroonchatapan E, Pitt BR, Huang L. Targeted gene delivery to the lung by antiPECAM antibody. Am. J. Physiol 2000; 278:L504–L511. 157. Zanta MA, Boussif O, Adib A, Behr JP. In vitro gene delivery to hepatocytes with galactosylated polyethyleneimine. Bioconjug. Chem 1997; 8:839–844. 158. Bettinger T, Remy JS, Erbacher P. Size reduction of galactosylated PEI/DNA complexes improves lectin mediated gene transfer into hepatocytes. Bioconjug. Chem 1999; 10:558–561. 159. Diebold SS, Lehrmann H, Kursa M, Wagner E, Cotten M, Zenke M. Efficient gene delivery into human dendritic cells by adenovirus polyethyleneimine and mannose polyethyleneimine transfection. Hum. Gene Ther 1999; 10:775–786. 160. Erbacher P, Remy JS, Behr JP. Gene Transfer with synthetic virus-like particles via the integrinmediated endocytosis pathway. Gene Ther 1999; 6:138–145. 161. Kircheis R, Kichler A, Wallner G, Kursa M, Ogris M, Felzmann T, Buchberger M, Wagner E. Coupling of cell-binding ligands to polyethyleneimine for targeted gene delivery. Gene Ther 1997; 4:409–418. 162. Goula D, Remy JS, Erbacher P, Wasowicz M, Levi G, Abdallah B, Demeneix BA. Size, diffusibility and transfection performance of linear PEI/DNA complexes in the mouse central nervous system. Gene Ther 1998; 5:712–717. 163. Bragonzi A, Boletta A, Biffi A, Muggia A, Sersale G, Cheng SH, Bordignon C, Assael BM, Conese M. Comparison between cationic polymers and lipids in mediating systemic gene delivery to the lungs. Gene Ther 1999; 6:1995–2004. 164. Abdallah B, Hassan A, Benoist C, Goula D, Behr JP, Demeneix BA. A powerful non-viral vector for in vivo gene-transfer in the adult mammalian brain: polyethyleneimine. Hum. Gene Ther 1996; 7:1947–1954. 165. Coll JL, Chollet P, Brambilla E, Desplanques D, Behr JP, Favrot M. In vivo delivery to tumors of DNA complexed with linear polyethyleneimine. Hum. Gene Ther 1999; 10:1659–1666. 166. Nguyen HK, Lemieux P, Vinogradov S, Gebhart CLe´, rin N, Paradis G, Bronich T, Alakhov V, Kabanov AV. Evaluation of polyether–polyethyleneimine graft copolymers as gene transfer agents. Gene Ther 2000; 7:126–138. 167. Haensler J, Szoka FC. Polyamidoamine cascade polymers mediate efficient transfection of cells in culture. Bioconjug. Chem 1993; 4:372–379. 168. Tang M, Redemann CT, Szoka FC. In vitro gene delivery by degraded polyamidoamine dendrimers. Bioconjug. Chem 1996; 7:703–714. 169. Kukowska-Latallo JF, Bielinska AU, Johnson J, Spindler R, Tomalia DA, Baker JR. Efficient transfer of genetic material into mammalian cells using starburst polyamidoamine dendrimers. Proc. Natl. Acad. Sci. USA 1996; 93:4897–4902. 170. Maruyama-Tabata H, Harada Y, Matsumura T, Satoh E, Cui F, Iwai M, Kita M, Hibi S, Imanishi J, Sawada T, Mazda G. Effective suicide gene therapy in vivo by EBV-based plasmid vector coupled with polyamidoamine dendrimer. Gene Ther 2000; 7:53–60. 171. Zuidam NJ, Posthuma G, Vries ET, Cromelin DJ, Hennink WE, Storm G. Effects of physicochemical characteristics of poly[2-(dimethylamino)ethyl methacrylate]-based polyplexes on cellular association and internalization. J. Drug Target 2000; 8:51–66. 172. Arigita C, Zuidam NJ, Crommelin DJA, Hennink WE. Association and dissociation characteristics of polymer/DNA complexes used for gene delivery. Pharm. Res 1999; 16:154–1541. 173. Cherng JY, Van de Wetering P, Talsma H, Crommelin DJ, Hennink WE. Effect of size and serum proteins on transfection efficiency of poly(2-dimethylamino)-plasmid nanoparticles. Pharm. Res 1996; 13:1038–1042. 174. Cherng JY, Talsma H, Crommelin DJ, Hennink WE. Long term stability of poly((2-dimethylamino)ethyl methacrylate)-based gene delivery systems. Pharm. Res 1999; 16:1417–1423.
364
Kaul and Amiji
175. Mumper RJ, Rolland AP. Plasmid delivery to muscle: recent advances in polymer delivery systems. In: Non-Viral Gene Delivery Systems for Controllable Gene Therapy. P Felgner and AP Rolland (Eds.). Adv. Drug Del. Rev 1998; 30:151–172. 176. Mumper RJ, Wang J, Nitta H, Anwer K, Tagliaferri F, Rolland AP. Protective interactive noncondensing (PINC) polymers for enhanced plasmid distribution and expression in rat skeletal muscle. J. Control. Rel 1998; 52:191–203. 177. Bronich TK, Nguyen HK, Eisenberg A, Kabanov A. Recognition of DNA topology in reactions between plasmid DNA and cationic copolymers. Polym. Prepr 2000; 41:1006–1007. 178. Astafieva I, Maksimova I, Lukanidin E, Alakhov V, Kabanov A. Enhancement of the polycationmediated DNA uptake and cell transfection with Pluronic P85 block copolymer. FEBS Lett 1996; 389:278–280. 179. Heskins M, Guillet JE, James E. Solution properties of poly(N-isopropylacrylamide). J. Macromol. Sci. Chem 1968; A2:1441. 180. Hinrichs WLJ, Schuurmans-Nieuwenbroek NME, van de Wetering P, Hennink WE. Thermosensitive polymers as carriers for DNA delivery. J. Control. Rel 1999; 60:249–259. 181. Gao YL, Lach E, Kim TH, Rodriguez A, Arevalo CA, Vacanti CA. Tissue-engineered nipple construction. Plast. Reconstr. Surg 1998; 102:2293–2298. 182. Lemieux P, Guerin N, Paradis G, Proulx R, Chistyakova L, Kabanov A, Alakhov V. A combination of poloxamers increases gene expression of plasmid DNA in skeletal muscle. Gene Ther 2000; 7: 986–991. 183. Bochot A, Fattal E, Gulik A, Couarraze G, Couvreur P. Liposomes dispersed within a thermosensitive gel: a new dosage form for ocular delivery of oligonucleotides. Pharm. Res 1999; 15:1364–1369. 184. Jeong BM, Bae YH, Lee DS, Kim SW. Biodegradable block copolymers as injectable drug delivery systems. Nature 1997; 388:860. 185. Jeong BM, Bae YH, Kim SW. Drug release from biodegradable injectable thermosensitive hydrogel of PEG-PLGA-PEG triblock copolymer. J. Control. Rel 2000; 63:155–167. 186. Gao X, Huang L. Potentiation of cationic liposome mediated gene delivery by polycations. Biochem 1996; 35:1027–1036. 187. Whitmore M, Li S, Huang L. LPD lipopolyplex initiates a potent cytokine response and inhibits pulmonary metastatic tumor growth. Gene Ther 1999; 6:1867–1875. 188. Zhou X, Huang L. DNA transfection mediated by cationic liposomes containing lipopolylysine: characterization and mechanism of action. Biochim. Biophys. Acta 1994; 1189:195–203. 189. Koiv A, Palvimo J, Kinnunen PKJ. Evidence for ternary complex formation by histone H1, DNA, and liposomes. Biochem 1995; 34:8018–8027. 190. Hagstrom JE, Sebestyen MG, Budker V, Ludtke JJ, Fritz JD, Wolff JA. Complexes of non-cationic liposomes and histone H1 mediate efficient transfection of DNA without encapsulation. Biochim. Biophys. Acta 1996; 1284:47–55. 191. Fritz JD, Herweijer H, Zhang G, Wolff JA. Gene transfer into mammalian cells using histonecondensed plasmid DNA. Hum. Gene Ther 1996; 7:1395–404. 192. Kataoka K. Nanoscopic vehicles with core-shell structure for drug targeting. In: Ogata N, Kim SW, Feijen J, Okano T, eds. Advanced Biomaterials in Bomedical Engineering and Drug Delivery Systems. Tokyo: Springer-Verlag, 2000:96–100. 193. Harada-Shiba M, Yamauchi K, Harada A, Shimokado K, Kataoka K. Polyion complex micelles as a vector for gene therapy—pharmacokinetics and in vivo gene transfer. Gene Ther 2002; 9:407–414. 194. Kataoka K. Tailor-made block copolymer micelles and nanoparticles for drug delivery. 59th Prague Meeting on Macromolecules, IU PAC. p. ML3, 2000. 195. Pitard B, Aguerre O, Airiau M, Lachages AM, Boukhnikachvili T, Byk G, Dubertret C, Herviou C, Scherman D, Mayaux JF, Crouzet J. Virus-sized self-assembling lamellar complexes between plasmid DNA and cationic micelles promote gene transfer. Proc. Natl. Acad. Sci. USA 1997; 94: 14412–14417. 196. Mahato RI. [editorial] Plasmid-based gene therapy: opportunities and challenges knock the millennium. J. Drug Target 1999; 7:241–244.
Polymeric Gene Delivery Systems
365
197. Pouton CW, Uduehi AN, Milroy DA, Lucas P. Cellular internalization and fate of polycation-DNA complexes. In: Kabanov AV, Seymour LW, Felgner PL, eds. Self-Assembling Complexes for Gene Delivery: From Laboratory to Clinic. New York: John Wiley, 1998:255–274. 198. Lear JD, DeGrado WF. Membrane binding and conformational properties of peptides representing the NH2 terminus of influenza HA-2. J. Biol. Chem 1987; 262:6500–6505. 199. Wolfert MA, Seymour LW. Chloroquine and amphipathic peptide helices show synergistic transfection activity in vitro. Gene Ther 1998; 5:409–414. 200. Chan CK, Jans DA. Enhancement of polylysine-mediated transferrinfection by nuclear localization sequences (NLSs): polylysine does not function as an NLS. Hum. Gene Ther 1999; 10:1695–1702. 201. Sebestyen MG, Ludtke JJ, Bassik MC, Zhang G, Budker V, Lukhtanov E, Hagstrom JE, Wolff JA. DNA vector chemistry: the covalent attachment of signal peptides to plasmid DNA. Nat. Biotechnol 1998; 16:80–85. 202. Kim JS, Maruyama A, Akaike T, Kim SW. Terplex DNA delivery system as a gene carrier. Pharm. Res 1998; 15:116–121. 203. Ferkol T, Perales JC, Mularo F, Hanson RW. Receptor-tissue-mediated gene transfer into macrophages. Proc. Natl. Acad. Sci. USA 1996; 93:101–105. 204. Choi Y, Liu F, Park J-S, Kim SW. Lactose-poly(ethylene glycol)-grafted poly-L-lysine as hepatoma cell-targeted gene carrier. Bioconjug. Chem 1998; 9:708–718. 205. Choi YH, Liu F, Choi JS, Kim SW, Park J-S. Characterization of Lactose-poly(ethylene glycol)grafted poly-L-lysine gene carrier. Hum. Gene Ther 1999; 10:2657–2665. 206. Kim J-S, Maruyama A, Akaike T, Kim SW. In vitro gene expression on smooth muscle cells using a terplex delivery system. J. Control. Rel 1997; 47:51–59. 207. Benns JM, Kim SW. Tailoring New Gene Delivery Designs for Specific Target. J. Drug Target 2000; 8:1–12. 208. Yu L, Nielsen M, Kim SW(). Terplex DNA gene carrier system targeting to artery wall cells. J. Control. Rel. 209. Suh W, Chung JK, Park SH, Kim SW. Anti-JL1 antibody conjugated poly(L-lysine) for targeted gene delivery to leukemia T Cells. J. Control. Rel 2001; 72:171–178. 210. Simo˜es S, Slepushkin V, Gaspar R, Pedroso de Lima MC, Du¨zgu¨nes N. Mechanisms of gene transfer mediated by lipoplexes associated with targeting ligands of pH-sensitive peptides. Gene Ther 1999; 6:1798–1807. 211. Schoen P, Chonn A, Cullis PR, Wilschut J, Scherrer P. Gene transfer mediated by fusion protein hemagglutinin reconstituted in cationic lipid vesicles. Gene Ther 1999; 6:823–32. 212. Mechtler K, Wagner E. Gene transfer mediated by influenza virus peptides: the role of peptide sequences. New J. Chem 1997; 21:105–111. 213. Ishiguro R, Matsumoto M, Takahashi S. Interaction of fusogenic synthetic peptide with phospholipid bilayers: orientation of the peptide alpha-helix and binding isotherm. Biochemistry 1996; 35: 4976–4983. 214. Subbarao NK, Parente RA, Szoka FC, Nadasdi L, Pongracz K. pH-dependent bilayer destabilization by an amphipathic peptide. Biochemistry 1987; 26:2964–2972. 215. Parente RA, Nir S, Szoka FC. Mechanism of leakage of phospholipid vesicle contents induced by the peptide GALA. Biochemistry 1990; 29:8720–8728. 216. Murthy N, Robichaud JR, Tirrell DA, Stayton PS, Hoffman AS. The design and synthesis of polymers for eukaryotic membrane disruption. J. Control. Rel 1999; 61:137–143. 217. Midoux P, Kichler A, Boutin V, Maurizot JC, Monsigny M. Membrane permeabilization and efficient gene transfer by a peptide containing several histidines. Bioconjug. Chem 1998; 9:260–267. 218. Gnatenko DV, Korneliuk AI, Matsuka GK. Tyrosyl-tRNA-synthetase from bovine liver. Functional role of histidine residues. [Russian]. Bioorganicheskaia Khimiia 1991; 17:1033–1037. 219. Benns JM, Choi JS, Mahato RI, Park JS, Kim SW. pH sensitive cationic polymer gene delivery vehicle: N-Ac-poly(L-histidine)-graft-poly(L-lysine) comb shaped polymer. Bioconjug. Chem 2000; 11:637–645. 220. Ferrari S, Moro E, Pettenazzo A, Behr JP, Zacchello F, Scarpa M. ExGen 500 is an efficient vector for gene delivery to lung epithelial cells in vitro and in vivo. Gene Ther 1997; 4:1100–1106.
366
Kaul and Amiji
221. Pouton CW. Nuclear import of polypeptides, polynucleotides and supramolecular complexes. Adv. Drug Del. Rev 1998; 34:51–64. 222. Van Dijk-Wolthuis WN, Van de Wetering P, Hinrichs WL, Hofmeyer LJ, Liskamp RM, Crommelin DJ, Hennink WEA. Versatile method for the conjugation of proteins and peptides to poly[2-(dimethylamino)ethyl methacrylate]. Bioconjug. Chem 1999; 10:687–692. 223. Zanta MA, Belguise-Valladier P, Behr JP. Gene delivery: with single nuclear localization signal peptide is sufficient to curry DNA to the cell nucleus. Proc. Natl. Acad. Sci. USA 1999; 96:91–96. 224. Branden LJ, Mohamed AJ, Smith CIE. A peptide nucleic acid nuclear localization signal fusion that mediates nuclear transport of DNA. Nat. Biotechnol 1999; 17:784–787. 225. Madsen SK, Mooney DJ. Delivering DNA with polymer matrices: applications in tissue engineering and gene therapy. PSTT 2000; 3:381–384. 226. Cleland JL. Single-administration vaccines: controlled release technology to mimic repeated immunizations. Trends biotechnol 1999; 17:25–29. 227. Schatzlein AG. Non-viral vectors in cancer gene therapy: principles and progress. Anti-Cancer Drugs 2001; 12:275–304. 228. Yang JP, Huang L. Overcoming the inhibitory effects of serum on lipofection by increasing the charge ratio of cationic liposome and DNA. Gene Ther 1997; 4:950–960. 229. Li S, Rizzo MA, Bhattacharya S, Huang L. Characterization of cationic lipid-protamine-DNA (LPD) complexes for intravenous gene delivery. Gene. Ther 1998; 5:930–937. 230. Crook K, Stevenson B, Dubouchet M, Porteous DJ. Inclusion of cholesterol in DOTAP transfection complexes increases the delivery of DNA to cells in vitro in the presence of serum. Gene. Ther 1998; 5:137–143. 231. Semple SC, Chonn A, Cullis PR. Influence of cholesterol on the association of plasma proteins with liposomes. Biochem 1996; 35:2521–2525. 232. Oja CD, Semple SC, Chonn A, Cullis PR. Influence of dose on liposome clearance: critical role of blood proteins. Biochim. Biophys. Acta 1996; 1281:31–37. 233. Zelphati O. PNA-dependent gene chemistry: stable coupling of peptides and oligonucleotides to plasmid DNA. Biotechnology 2000; 28:304–316. 234. Dash PR, Read ML, Barrett LB, Wolfert M, Seymour LW. Factors affecting blood clearance and in vivo distribution of polyelectrolyte complexes for gene delivery. Gene Ther 1999; 6:643–650. 235. McLean JW, Fox EA, Baluk P, Bolton PB, Haskell A, Pearlman R, Thurston G, Umemoto EY, McDonald DM. Organ-specific endothelial cell uptake of cationic liposome-DNA complexes in mice. Am. J. Physiol 1997; 273:H387–H404. 236. Jain RK. Delivery of molecular and cellular medicine to solid tumors. Adv. Drug Del. Rev 1997; 26:71–90. 237. Hashizume H, Baluk P, Morikawa S, McLean JW, Thurston G, Roberge S, Jain RK, McDonald DM. Openings between defective endothelial cells explain tumor vessel leakiness. Am. J. Pathol 2000; 156:1363–1380. 238. Griesenbach U, Chonn A, Cassady R, Hannam V, Ackerley C, Post M, Tanswell AK, Olek K, O’Brodovich H, Tsui LC. Comparison between intratracheal and intravenous administration of liposome/DNA complexes for cystic fibrosis lung gene therapy. Gene Ther 1998; 5:181–188. 239. Liu F, Qi H, Huang L, Liu D. Factors controlling the efficiency of cationic lipid-mediated transfection in vivo via intravenous administration. Gene Ther 1997; 4:517–523. 240. Hofland HEJ, Nagy D, Liu JJ, Spratt K, Lee YL, Danos O, Sullivan SM. In vivo gene transfer by intravenous administration of stable cationic lipid/DNA complex. Pharm. Res 1997; 14:742–749. 241. Maeda H, Wu J, Sawa T, Matsumura Y, Hori K. Tumor vascular permeability and the EPR effect in macromolecular therapeutics: a review. J. Control. Rel 2000; 65:271–284. 242. Blume G, Cevc G. Liposomes for the sustained drug release in vivo. Biochim. Biophys. Acta 1990; 1029:91–97. 243. Kircheis R, Schuller S, Brunner S, Ogris M, Heider KH, Zauner W, Wagner E. Polycation-based DNA complexes for tumor-targeted gene deliver in vitro. J. Genet Med 1999; 1:111–120. 244. Mislick KA, Baldeschwieler JD. Evidence for the role of proteoglycans in cation-mediated gene transfer. Proc. Natl. Acad. Sci. USA 1996; 93:12349–12354.
Polymeric Gene Delivery Systems
367
245. Zabner J, Fasbender AJ, Moninger T, Poellinger KA, Welsh MJ. Cellular and molecular barriers to gene transfer by a cationic lipid. J. Biol. Chem 1995; 270:18997–19007. 246. Pires P, Simoes S, Nir S, Gaspar R, Duzgunes N, Lima MC. Interaction of cationic liposomes and their DNA complexes with monocytic leukemia cells. Biochim. Biophys. Acta 1999; 1418:71–84. 247. Schaffer D, Lauffenburger DA. Optimization of cell surface binding enhances efficiency and specificity of molecular conjugate gene delivery. J. Biol. Chem; 273:28004–28009. 248. Pollard H, Remy JS, Loussouarn G, Demolombe S, Behr JP, Escande D. Polyethylenimine but not cationic lipids promotes transgene delivery to the nucleus in mammalian cells. J. Biol. Chem 1998; 273:7507–7511. 249. Suh J, Lee SH, Kim SM, Hah SS. Conformational Flexibility of poly(ethylenimine) and its derivatives. Bioorg. Chem 1997; 25:221. 250. Boussif O, Zanta MA, Behr JP. Optimized galenics improve in-vitro gene transfer with cationic molecules up to 1000-fold. Gene Ther 1996; 3:1074–1080. 251. Vogt TC, Bechinger B. The interactions of histidine-containing amphipathic helical peptide antibiotics with lipid bilayers. J. Biol. Chem 1999; 274:29115–29121. 252. Wagner E, Plank C, Zatloukal K, Cotten M, Birnstiel ML. Influenza virus hemagglutinin HA-2 Nterminal fusogenic peptides augment gene transfer by transferrin-polylysine/DNA complexes: towards a synthetic virus-like gene transfer vehicle. Proc. Natl. Acad. Sci. USA 1992; 89:7934–7938. 253. Rabinowitz J, Samulski RJ. Building a better vector: the manipulation of AAV virions. Virology Minireview 2000; 278:301–308. 254. Loimas S, Pasanen T, Gomes A, Bizarro S, Morgan RAa¨, nne J, Wahlfors J. Viral vectors carrying a marker-suicide fusion gene (TK-GFP) as tools for TK/GFP-mediated cancer gene therapy. Gene Ther. Mol. Biol 2000; 5:165–174. 255. Zelphati O, Liang XW, Hobart P, Felgner PL. Gene chemistry: functionally and conformationally intact fluorescent plasmid DNA. Hum. Gene Ther 1999; 10:15–24. 256. Tseng W, Haselton FR, Giorgio TD. Transfection by cationic liposomes using single cell measurements of plasmid delivery and transgene expression. J. Biol. Chem 1997; 272:25641–25647. 257. Capecchi MR. High efficiency transformation by direct microinjection of DNA into cultured mammalian cells. Cell 1980; 22:479–488. 258. Zauner W, Brunner S, Buschle M, Ogris M, Wagner E. Differential behavior of lipid based and polycation-based gene transfer systems in transfecting primary human fibroblasts: a potential role of polylysine in nuclear transport. Biophys. Biochim. Acta 1999; 1428:57–67. 259. Godbey WT, Wu KK, Mikos AG. Tracking the intracellular path of poly(ethylenimine)/DNA complexes for gene delivery. Proc. Natl. Acad. Sci. USA 1999; 96:5177–5181. 260. Han SO, Mahato RI, Sung YK, Kim SW. Development of biomaterials for gene therapy. Mol. Ther 2000; 2:302–317.
17 Bioactive Molecules and Biodelivery Systems ¨ ner and H. Su¨heyla˙ Kas Filiz O Hacettepe University Ankara, Turkey
I. INTRODUCTION Biomaterials are defined as substances or combination of substances, natural or synthetic in origin, that are biocompatible, safe, effective, biodegradable, nontoxic and nonallergenic and that can be used to treat, augment, or replace any tissue, organ, or function of the body by mimicking biological phenomena. Applications of biomaterials have generated a great deal of interest recently and include uses as biologics, drugs, or medical devices. Among biocompatible biomaterials such as metals, ceramics, composites, and polymers, natural polymers (biopolymers) are becoming more attractive due to their biological characteristics. Biopolymers are composed of macromolecules including proteins, polysaccharides, and nucleic acids formed by living organisms [1]. Medical applications of biopolymers, which can be used both as bioactive substances or biodelivery systems, range from artificial organs to gene therapy. Biological products, which are specific biotechnology products, are differentiated from traditional small molecule drugs of classic pharmaceutical technology by the U.S. Food and Drug Administration (FDA). The Center for Biologics Evaluation and Research (CBER) regulates biological products in order to protect and enhance the public health through these regulations to ensure safety and effectiveness in the delivery of biologicals to the patients. Related products are blood and products derived from blood, vaccines including AIDS vaccines, tissue for transplantation, allergenics and antitoxins as well as biological therapeutics including biotechnologically derived products. These product categories cover all bioactive substances which are derived from living sources and may offer effective means to treat a number of illnesses that have no effective treatments yet. Specific biotechnological products are therapeutic plasmid DNA products, therapeutic synthetic peptide products, monoclonal antibody products for in vivo use, and therapeutic recombinant DNA derived products. Enormous amounts of data created from the human genome project will be used for the development of new drugs and therapy methods [2,3]. Protein microarrays are developed for protein expression profiling, functional proteomics, and pharmacoproteomics. Bioinformatics, high-throughput target validation, chemical genomics, and pharmacogenomics are approaches which will play major roles in speeding up drug discovery and development [4]. Regulatory issues for biotechnologically derived agents have focused on product quality, safety, and consistency. To bring a biological product from investigational to marketable status, licensing of the product is necessary. Manufacturers must provide data demonstrating that the 369
¨ ner and Kas O
370
product is safe and effective and then must go through the licencing process including an investigational new drug (IND) phase, premarket approval phase, and postlicensure phase. In the United States CBER regulates approval and inspection of biological products. Product license applications must provide a detailed description of the product including quality control, stability, labeling, safety, and efficacy data. Regulatory standarts for biotechnologically derived pharmaceuticals have been comparable among the EU, Japan, and United States. In order to cooperate authotorities of these three major regions that manufacture biotechnology products have held International Conferences for Harmonization (ICH) since the 1990s. Specifications for biotechnological/biological products are well documented in the ICH guidelines, which include the control of raw materials, inprocess testing, and process validation [5]. Guidance was developed to provide a document regarding standards for good manufacturing practices for active pharmaceutical ingredients (APIs) [6]. Types of manufacturing listed in the guidance are Chemical manufacturing API derived from animal sources API extracted from plant sources, herbal extracts used as API, and powdered herbs Biotechnological processes including fermentation/cell culture Classical fermentation to produce API Vaccines, whole cells, whole blood and plasma, plasma fractions and gene therapeutics are excluded from these guidelines. Products having biological activity are classified in the EC guidelines [7] as vaccines, hormones, immunosera, enzymes, cytokines, other fermentation products such as monoclonal antibodies, and recombinant DNA technology products.
II. BIOACTIVE SUBSTANCES Therapeutic activities of the biomolecules expressed naturally in the body have been understood since the 1950s, and the first bioactive therapeutic materials were isolated from biological sources. Bioactive molecules are produced naturally within living organisms in small amounts, limiting their large-scale production due to the difficulties arising from classical extraction methods depending upon natural sources. Medical applications also cannot be widespread because of the insufficient amounts and the safety problems of these traditional biomolecules. Developments in biotechnology and materials science facilitated the production of bioactive materials. At the moment the bulk of existing materials can be produced by using alternative production systems due to the results of the human genome project and developments in genetic engineering, cell biology, and immunology. Biotechnology is one of these novel production systems which integrates the natural sciences and engineering sciences in order to achieve the application of organisms, cells, parts thereof, and molecular analogs for products and services [1]. Considering these definitions and classifications, bioactive substances obtained by classical or novel biotechnological methods are classified in Table 1. Regulatory standards were developed for bioactive substances, and several guidelines, regulations, and points-to-consider documents were issued by regulatory authorities [8]. For the safety evaluation of active biotechnology products—which include proteins and peptides, their derivatives, and products arising from their components (e.g., derived from cell cultures or produced by rDNA technology), principles have been outlined to provide guidance for industry.
Bioactive Molecules and Biodelivery Systems
371
Table 1 Bioactive Molecules Obtained by Different Methods Classical Methods Chemical synthesis
Isolation and purification from plants, animal organs, or body fluids
From microorganisms by traditional fermentation
Peptides/proteins Polysaccharides Lipids
Antibiotics
Small drug molecules Synthetic peptides Lipids
Biotechnological Methods Recombinant DNA technology Small molecules Amino acids Vitamins Antibiotics Peptides/proteins Polypeptides Vaccines Monoclonal antibodies Nucleic acids DNA RNA Oligonucleotides
Hybridoma technology
Transgenic animals
Transgenic Plants
Monoclonal antibodies
Peptides/proteins
Vaccines
In this chapter the authors focus only on natural bioactive polymeric materials such as peptides/proteins, polysaccharides, lipids, and nucleic acids and delivery systems derived from these materials. A. Peptides and Proteins Substances derived from amino acid polymers differ from one another according to their molecular sizes and sequences. Peptides are amides derived from two or more amino carboxylic acid molecules by formation of a covalent bond from the carbonyl carbon of the nitrogen atom of another with loss of water, which usually applies to structures formed from alpha amino acids [1]. Polypeptides are peptides containing ten or more amino acid residues. Proteins are naturally occurring and synthetic polypeptides can have molecular weights greater than about 10,000 Da. The limit is not precise; in some references it is given as 5000 or 6000 Da [9]. Proteins are large complex molecules composed of interacting polypeptides. The sequence of the amino acids, and thus the function of the protein, is determined by the sequence of the base pairs in the gene that encodes it. Proteins are essential to the structure, function, and recognition mechanisms of the body such as peptide hormones, enzymes, and antibodies. Since the 1970s protein engineering methods have been used to produce peptides and proteins with altered or novel amino acid compositions. Therapeutic proteins were obtained from human or animal organs and blood until recombinant DNA technology was developed that could produce bioactive proteins [10]. Risk of unwanted serious side effects is a major problem due to contaminants such as viruses or immunogenic materials in the protein products extracted
¨ ner and Kas O
372
from animal organs and blood. Some examples of the proteins isolated from animal sources are insulin, human growth hormone, blood coagulation factors, human and bovine serum albumin, peptide hormones, and trypsin. At the moment recombinant DNA technology, transgenic animal pharming, and hybridoma technology are the methods used to produce therapeutic proteins, vaccines, monoclonal antibodies, and DNA products [11]. Proteins are also produced by chemical modification of normal proteins or by solid-state polypeptide synthesis [12]. Site-specific mutagenesis is an approach for altering physical and chemical properties of proteins, but with these mutations new substances are generated with their own activities and side effects [13]. In 1982 the first recombinant human protein for human health (insulin) came to market, and others followed [14]. Development of recombinant proteins as bioactive molecules has accelerated in the last 10 years. More than 60 protein products are commercially available and approximately $10 billion of revenue was garnered in 2000. More than 400 drug products came to the end of clinical studies and are waiting for approval from regulatory bodies. The first modern techniques used for producing bioactive materials were recombinant DNA technology and hybridoma technology [15,16]. By using recombinant DNA technology desired genetic material of an organism can be combined with the genetic material of another organism. By transforming a host organism with the new gene construction, bioactive proteins can be expressed in large amounts to be used as pharmaceutical active substances. Bioreactor design, scale-up of production, protein expression systems, upstream processing, product purification, and waste treatment are important elements of the bioprocess industry [17]. High density feed batch fermentation, low density continuous processing, two-phase extraction, and expanded bed chromatography are methods used to optimize production yield. Immobilization technologies applied for long-term use have increased productivity of bioreactors [18]. A recent trend has been an integrated bioprocessing system in which proteases are continuously removed from the fermentation medium [19]. Speeding up of the recovery process leads to high yield and high quality product. In situ recovery processes such as adsorption chromatography, membrane chromatography, extraction, and flotation are expensive methods for protein studies. Optimization studies for the expression systems to improve yield and efficiency are expanding. Complementary expression of lyzozyme or coexpression of endonuclease is suitable for rapid recovery and improved product yield. Scalability of the production method is an important issue. Processes used for small-scale production cannot meet clinical and commercial demands—as in the case of transgenic plant cells, which is not satisfactory for the commercial production of active substances. The majority of therapeutic proteins have been produced either in microbial cells (E. coli) or mammalian cell culture systems. Hosts for protein expression include Prokaryotic systems (E. coli, B. Subtilis) [20,21] Yeasts (Saccharomyces cerevisiae, Pichia pastoris) [20] Animal cell lines (CHO, BHK) [22] Plant cells (tobacco) [23] Insect cells [20] 1. Microbial Expression Systems Escherichia coli is a widely used host organism for the production of pharmaceutical recombinant proteins. However, deficiency of posttranslational modifications such as glycosylation still remains as a problem to be overcome [20]. Secretion of correctly folded large amounts of proteins in E. coli is difficult. In the expression medium, soluble active protein production is desired instead of inclusion body formation. Expression vectors, promoters, and transcriptional factors are parameters affecting the yield of protein production. [24]. Useful protein expression systems by using promoters such as araE, which uses arabinose, ramnose, and microaerobic conditions,
Bioactive Molecules and Biodelivery Systems
373
have been developed recently. Cell-free protein synthesis is a new method in which cell extracts of E. coli are used instead of the microorganism [25]. In this approach extract is prepared following cell growth and lysis; substrates and salts are than added to the extract, and protein synthesis is initiated by adding the template. By using this method cell-toxic proteins can be produced and protein synthesis and folding can be manipulated. This method has high yields, but it is an expensive system. 2. Mammalian Expression Systems Mammalian cells are the most desirable choice for pharmaceutical proteins Cell line selection and bioreactor design are critical factors for the large-scale production. In order to maximize cell culture productivity for proteins, culture duration and viable cell density must be optimized, and benefits and risks of the method must be evaluated. Chinese hamster ovary cells (CHO), baby hamster kidney cells (BHK), and NSO murine myeloma cells are the systems used most for mammalian expression [27]. Environmental conditions are important for controlling glycosylation, carbon metabolism, cell growth, and cell death. Glycosylation of recombinant proteins produced in NSO and CHO cultures can be controlled by adding culture supplements such as nucleotide sugar precursors (glycosamine and N-acetylmannosamine). Protein products containing galactose and sialic acid are obtained due to the overexpression of galactocyltransferase and sialyltransferase in CHO cells. Recombinant protein production has been dependent upon the growth state of the culture [27]. Apoptosis, the programmed cell death process, is a critical factor for optimizing productivity of mammalian cell cultures. Bioreactor conditions can trigger apoptosis in mammalian cells. Families of intracellular proteases known as caspases are responsible from the mechanisms regulating apoptosis [28]. Inhibition of apoptosis by chemicals extends the cell culture viability. Bcl-2 is a human proto-oncogene located on chromosome 18, and high levels of the Bcl-2 family proteins have been reported to inhibit apoptotic death in some mammalian cell cultures [29]. Development of host cell lines that express trans or cis activating elements on expression materials are also useful techniques for protein production [30]. 3. Yeast Systems Yeast systems such as Saccharomyces cerevisiae and Pichia pastoris are major production systems for proteins [20]. Recombinant hepatitis B surface antigen was the first protein product developed in yeast systems. Recombinant human chitinase, single chain Fv (scFv) antibody fragment and an insulin precursor are proteins recently generated in yeast systems [31]. 4. Insect Expression Systems Insect cells have also been used for protein expression. Due to the overexpression of galactosyltransferases and sialyltransferases, sialiated oligosaccharides on insect-derived proteins can be generated [32]. 5. Other Expression Systems Transgenic plants, animals, and plant cell cultures are under investigation as a safe means for production of bioactive proteins [20,33]. The goal of transgenic plant studies is to develop bioengineered plants for producing large amounts of bioactive materials at low cost and reduced risk. Posttranslational modifications such as glycosylation, stability, and environmental controls are problems that need to be solved in transgenic plant biotechnology. The current techniques are not fully applicable to large-scale industrial production. In some studies attempts were found
¨ ner and Kas O
374
Table 2 Bioactive Peptides Proteins Produced by Recombinant DNA Technology Biological activity
Peptides proteins
Hormones Cytokines
Hemopoietic growth factors Growth factors Blood clotting factors Enzymes Vaccine antigens Thrombolytic agents Monoclonal antibodies (therapeutically active)
rh insulin, rh growth factor, rh leutinizing hormone, rh follicle stimulating hormone, rh corionic gonadotrophin Interferons: rhIFN-␥1b, rhIFN-␣2b, rhIFN-␣2a, rhIFN-␣n3, rhIFN-1a, Interleukins: rhIL-2, rhIL-3, rhIL-4, rhIL-6, rhIL-10, rhIL-12 Tumor necrosis factors: TNF-␣, TNF- Erythropoietin, thrombopoietin EGF, CSF, rhG-CSF, rhGM-CSF, TGF, FGF, NGF, BDNF, PDGF, rh BMP-2, Fibronectin, vitronectine, VEGF rhAHF, rhFVIII, rhFIX, rhFVIIa Asparaginase, DNase, glucocerebrosidase, superoxide dismutease rhHBsAg rht-PA Anti-CD3 anticor (Orthoclone OKT3®)
IFN: interferon; EGF: epidermal growth factor; CSF: colony stimulating factor; G-CSF: granulosyte colony stimulating factor; GM-CSF: granulocyte macrophage colony stimulating factor; BNDF: brain derived neurotropic factor; TGF: transforming growth factor; FGF: fibroblast growth factor; NGF: nerve growth factor; PDGF: platelet derived growth factor; BMP: bone morphogenic protein; TNF: tumor necrosis factor; IL: interleukin; AHF: antihemophilic factor; t-PA: tissue plasminogen activator; FVIII: factor VIII; FIX: factor IX; FVIIa: factor VIIa; HBsAg: Hepatitis B surface antigen r:recombinant; h: human
successful to produce fast, efficient and economical protein purification by using fusion protein beta-glucuronidase (GUS)–calmodulin (CaM) [34]. Recombinant DNA techniques produce a large number of therapeutic proteins. Only a few examples having major importance are included in this Chapter. Bioactive peptides and proteins produced by recombinant DNA technology are classified in Table 2. Biological response modifiers play an important role in regulating immune and inflammatory reactions against viruses and tumors. Interferons are classified as the major families of proteins in the cytokines group [35]. They are potent molecules exerting their activities at very low concentrations. Cytokines can initiate signal transduction by binding receptors, which are transmembrane glycoproteins. Lymphocytes, macrophages, leucocytes, somatic cells, and tumor cells display interferon receptors on their surfaces. A number of cytokines are available commercially and some of them are under clinical investigation for numerous indications (Tables 3 and 4). Interleukin molecules are another group of cytokine proteins produced by T-lymphocytes, macrophages, eosinophils, fibroblasts, or keratinocytes in different structures.
Table 3 Approved Indications of Interferons and Interleukins Protein
Indication
IFN-␣ IFN- IFN-␥ IL-2
Hairy cell leukemia, genital warts Multiple sclerosis Chronic granulomatous disease Renal cell carcinoma
Bioactive Molecules and Biodelivery Systems
375
Table 4 Indications of Interferons and Interleukins in Clinical Trials Protein IFN-␣ IFN- IFN-␥ IL-1␣,  IL-11 IL-2 IL-3 IL-4 IL-6
Indication AIDS, hepatitis B and C, bladder cancer, basal cell carcinoma Hepatitis B and C, genital warts, basal cell carcinoma Rheumatoid arthritis, warts, cancers Cancer immunotherapy, vaccine adjuvant wound healing, prevention of bone marrow suppression Thrombocytopenia AIDS, infectious diseases, rheumatoid arthritis Platelet and bone marrow deficiencies Cancers, vaccine adjuvant Thrombocytopenia
Growth factors are also an important group of biotechnology products; these can induce growth and differentiation of bone marrow. Several families of growth factors have been produced and their important role in the regulation of both normal and cancer cell growth have been shown [36]. Vascular endothelial growth factor (VEGF) is an endothelial cell specific protein which plays a role in angiogenesis during normal embryonic and neonatal development, but it is also a critical mediator of blood vessel growth in tumors. Inhibitors of VEGF such as blocking antibodies or small molecule receptor antagonists may be useful for treatment of tumors and intraocular neovascular syndromes [37]. 6. Monoclonal Antibodies Antibodies represent an important class of protein biotherapeutics produced for use in transplants, cancer therapies, infections and cardiovascular diseases [38]. A specific B cell clone produces monoclonal antibodies (MoAbs), which recognize a specific epitope on the surface of an antigen, for the purpose of destroying antigens. Y-shaped four chain structures of MoAbs are composed of two heavy (H) and two light (L) chains held together by disulfide linkages and noncovalent interactions. Monoclonal antibodies can be administered as intact, fragmented, or conjugated with radioisotopes or toxins. Antigen binding sites are located on the variable regions, and the constant region is the antibody effector part of the molecule. Polyclonal antibodies are developed by immunizing animals with a desired antigen (bacterial, viral, bacterial toxin, or venom) and by purification from plasma, serum, or placental tissues of donors . Hundreds of monoclonal antibodies are undergoing preclinical and clinical studies for the purpose of therapy and in vivo imaging [39]. Immunogenicity of these molecules is limiting their use. In 1984 known hybridoma technology was modified to produce chimeric [40] and in 1990s to produce humanized [41] and fully human [42] antibodies due to the immunogenicity of mouse MoAbs. Differences from 100% mouse to 100% human antibody are as follows;
100% murine → 66% human → 90 − 95% human → 100% human mou se c himeric humanized fully human
(1)
Bacterial fermentation is offered as a suitable method for the expression of antibody and antibody fragments. Mammalian cell cultures and transgenic animals and plants are also used as expression systems for antibody production. Among these mammalian systems is the choice for large-scale,
¨ ner and Kas O
376
safe, and effective production of monoclonal antibodies which mimic natural antibodies [43]. Composition and temperature of the culture medium are effective factors on the cell viability. 7. Vaccine Antigens Vaccines are pharmaceutical products containing antigenic components of a pathogen that activate the immune system by humoral or cell-mediated immune mechanisms [44]. Conventional vaccines are sterile preparations of live or dead attenuated bacteria, live or attenuated viruses, toxoids, and surface antigens. More than 70 biotechnologically derived vaccines are now in development. The nature of the immune response itself and sustained release of the antigen from the formulation are factors affecting immune response [45]. Synthetic small peptides, which have immunological protection against pathogens, are also used for vaccination by coupling with carrier proteins [46]. Novel subunit vaccines have been prepared by using recombinant antigens, and DNA vaccines have been developed containing gene constructs encoding surface antigens or adjuvants. These vaccine formulations have advantages over conventional vaccines due to their purity and safety characteristics. Some approved vaccines are listed in Table 5. B. Polysaccharides Polysaccharides are carbohydrates found in nature consisting of a large number of monosaccharides linked glycosidically. The term commonly used for those containing more than ten monosaccharides is glycans [1]. The major forms of stored carbohydrate in animals are glycogen and in plants are starch. Glycogen stores large amounts of energy in a small volume due to its compact structure [9]. 1. Lipopolysaccharides Lipopolysaccharides are natural compounds consisting of a trisaccharide repeating unit (two heptose units and octulosonic acid) with oligosaccharide side chains and 3-hydroxytetradecanoic acid units (they are a major constituent of the cell walls of gram-negative bacteria [1]. C. Lipids Lipids are biological molecules that are soluble in nonpolar solvents that have several important functions in living organisms [9]. They are found in the biological membranes as structural components, they have vitamin and hormone effects, and they act as energy reserves. They can be found in the form of saponifiable glycerides (fats and oils), phospholipids, sphingolipids, nonsaponifiable steroids (cholesterol), and bile salts [1].
Table 5 Vaccines Developed by Biotechnology Vaccine Recombinant hepatitis B surface antigen (Recombivax HB®) Recombinant hepatitis B surface antigen (Engerix-B®) Hemophilus influenza b conjugate vaccine (Comvax®)
Use
Approval
Prevention of hepatitis B virus infections
1986
Prevention of hemophilus influenza b infections
1996
Prevention of hepatitis B virus infections
1989
Bioactive Molecules and Biodelivery Systems
377
Sphingolipids are a unique form of phospholipids that generates a family of molecules named as ceramides. Phospholipids are the main lipids found in biological systems; they contain phosphoric acid as mono- or di-esters and include phosphatidic acids and phosphoglycerides. Lipoproteins are large amphiphilic molecules composed of lipids and proteins. These are clathrate complexes consisting of a lipid wrapped in a protein host without covalent binding. The complex has a hydrophilic outer surface consisting of all the protein and the polar ends of any phospholipids. They play an important role in the lipid metabolism of the body. D. Nucleic Acids and Oligonucleotides Until the last decade all biotechnologically based products were in protein structure, but now a wide range of nucleic acid pharmaceuticals are taking their place in gene therapy trials. Single gene defects can cause serious genetic disorders; therefore nucleic acids are considered good candidates for correcting these conditions [47]. Using bioengineering methods DNA and RNA are produced in sufficiently large amounts to be used as gene therapeutics or vaccines. For protein expression, DNA should be delivered to the nucleus and RNA should be delivered to the cytoplasm of the cells within the living organisms. Antisense oligonucleotides are short single-stranded DNA or RNA molecules [11]. Protein production is decreased by using antisense oligonucleotides, which interrupt the genes responsible for the expression of this protein by binding to target mRNA and preventing the transcription and translation processes. Use of plasmid DNA (pDNA) for vaccine formulations is a new approach in vaccine technology. Plasmid DNA vaccines are purified preparations of plasmid DNA designed to contain a gene or genes for the intended vaccine antigen as well as genes incorporated into the construct to allow for production in a suitable host system [48]. Plasmid DNA vaccines are biological products within the meaning of the Public Health Service Act and regulated by CBER (42 U.S.C. 262). Regulations that govern the use of biologicals apply to pDNA vaccines (21 CFR, Parts 600, 601, and 610).
III. BIODELIVERY SYSTEMS Delivery of bioactive substances to target cells is one of the most important aspects of bioactive drug formulation, and a wide variety of strategies and materials have been investigated. Issues to be considered for each protein delivery method include purity, integrity of the protein during storage and after administration to the patient, scalability of the manufacturing methods, bioavailability of the protein, and safety and toxicity of the formulation [49,50]. Molecular recognition, signal transduction, enzymatic specificity, immunomodulation, and bioactivity of the peptide/ protein molecules depend on their native three-dimensional structure, which must be stabilized against the distruptive conditions during processing [51]. In Table 6 natural and synthetic biomaterials that have been used as carriers for bioactive protein substances are listed [52–56]. In this chapter we are focusing on natural biodelivery systems for delivering bioactive substances. Biomaterials such as proteins, polysaccharides, lipids, and nucleic acids have been attached to active drug molecules via degradable chemical linkages or have been loaded with them as carrier systems. Active substances are released from the delivery systems by different mechanisms, which include [57]. Diffusion Erosion
¨ ner and Kas O
378
Table 6 Biomaterials Used for Delivering Bioactive Substances Synthetic biomaterials Natural biomaterials Proteins: albumin, gelatine, collagen gluten, kazein, fibrinogen, fibronectin, antibodies Polysaccharides: alginates, dextran, chitin, chitosan, starch, cellulose, pectin Lipids: stearic acid, ethyl stearate, tristearine, hydrogenated vegetable oils, phospholipids (soybean or egg) Nucleic acids: plasmid DNA Others: Calcium phosphate, ceramides
Biodegradable
Nonbiodegradable
poly(alcylcyanoacrilates) (PACA), poly(␣-hydroxyacids), poly(lactic acid) (PLA), poly(glycolic acid) (PLG), poly(lactic-coglycolic acid) (PLGA), poly(orthoesters), poly(amino acids), poly(caprolactone), poly(urethane)
Hydrogels: poly(hydroxyethyl methacrylate) (PHEMA), poly(vinyl acetate), poly(methoxyethyl methacrylate), poly(vinyl alcohol) (PVA), poly(ethylene oxide) (PEO), poly(ethylene glycol) (PEG) Silicones: poly(dimethysiloxane) (PDMS), ethylene vinyl acetate copolymer (PEVAc) Poloxamers
Chemical (pH or hydrolysis) Biological (chemical or enzymatic) Solvent activation Osmotic effects Swelling Novel delivery approaches include chemical modification or entrapment of macromolecules in microcarriers such as lipid vesicles, microspheres, nanospheres, microcapsules, emulsions, or minipumps. [58]. Natural biomaterials are attracting more interest due to their biodegradable, biocompatible, less toxic, and less immunogenic properties. A. Protein-Based Delivery Systems Proteins are bioactive macromolecules which can also be used to develop microparticulate carrier systems for synthetic chemical active substances, proteins, and nucleic acids. Nanoparticles or microparticles are prepared through denaturation of the proteins by using crosslinking agents, heating, or coacervation [59,60]. The main goal of biopharmaceutical formulations is to prepare stable and effective drug delivery systems by enhancing stability and biological activity of the bioactive substances. A large number of protein delivery systems has been investigated but only a few of them have reached human health care applications yet. There are many hurdles to be overcome in the production and formulation of polypeptide structures. In vitro and in vivo stability of the peptides and proteins is an important aspect of the bioavailability and is divided into two forms, chemical and physical stability [61]. Types of chemical instability are oxidation, reduction, deamidation, hydrolysis, arginine conversion, beta elimination and racemization, isomerization, deglycosilation or proteolytic degradation. Physical instability is observed as denaturation, aggregation, sedimentation, or adsorption, which result from alteration of the 3D shape of the protein along with the effects of pH, temperature, and shearing stress [58]. A strategy for enhancing protein stability and prolonging efficacy would be to decrease protein dynamics during storage and transportation [49]. Excipients such as albumin, amino acids, polysaccharides, alcohols, or surface-active agents may be added to bio-
Bioactive Molecules and Biodelivery Systems
379
pharmaceutical products as stabilizing agents. Lower temperatures and dehydration, which are expected to decrease protein dynamics, are not sufficient to stabilize proteins. Sucrose and trehalose are efficient in the forming of hydrogen bonds with proteins, thus replacing the water of hydration that is to be removed [62]. Bioactive proteins required being absorbed rapidly to prevent themselves from proteolytic degradation. In vivo barriers for protein delivery include the permeability of proteins across cell membranes (e.g., epithelial barriers of the GI tract and lungs or stratum corneum of the skin), proteolysis at the site of administration, and immunogenic reactions [63]. Polyethylene glycol conjugation (PEGylation) is an irreversible chemical modification for improving in vivo stability and solubility of proteins, lipids, nucleic acids, and prolonged effects of microparticles and liposomes. By the use of PEGylation, RES uptake, drug leakage, immunogenicity, and hemolytic toxicity can be overcome [64,65]. PEG-ADA (adenosine deaminase) and PEG-ALL (asparaginase) are two commercially available formulations which provide low dose exposure and reduced side effects in the administration of enzymes [66]. PEGylated interferon maintained both antiviral and immunotherapeutic activity and PEG conjugated rhTNF-␣ displayed high antitumor activity as long circulating targeted systems [67,68]. Hydrophobic ion pairing is a reversible method used to purify and prepare protein dosage forms in polymeric matrices [69]. Proteins such as albumin, gelatin, collagen, and fibrin are formulated to develop parenteral, oral, pulmonary, or nasal microparticulate carriers for bioactive materials. Protein-based polymers are also produced by using genetic engineering techniques, showing improved biodegradability and controlled release characteristics [70]. Delivery method may affect clinical toxicology or pharmacology of the protein due to the administration teachnique or the matrix material. In controlled release protein delivery systems low dose proteins can be administered with minimal side effects. According to desired exposure, a suitable delivery method (pulsatile or sustained, local or systemic) can be selected. Initially pharmacokinetics of the protein in the delivery system are studied in animals or primates. When the desired profile is achieved toxicology and pharmacology studies are performed and initial human trials for pharmacokinetic, pharmacodynamic, and safety data are organized. If bioavailability of a protein is low, production costs will be higher due to the high protein amount. The traditional route for protein application is parenteral, but more convenient alternative routes for protein delivery are under investigation [71,72]. Oral delivery of peptides proteins, vaccines and nucleic acid drugs is a major challenge in the pharmaceutical industry, but is preferable due to the ease of administration, patient compliance, and reduced costs. Noninvasive promising routes for protein delivery are lung and skin. Pulmonary delivery has greater promise than oral and transdermal routes due to the large surface area and higher bioavailability. Gelatin microspheres have been shown to be useful vehicles for pulmonary delivery of salmon calcitonin (SCT). The hypocalcemic effects after im administration of SCT in gelatin microspheres were examined, and they were found to be useful systems for nasal or im delivery of SCT [73]. Crosslinked gelatin nanospheres prevented DNA degradation in serum and allowed prolonged gene expression [74]. Albumin is a blood protein secreted by the liver and is a widely used protein to prepare microparticles due to its nontoxic and biodegradable characteristics [75]. Typical smooth surfaced albumin microspheres are shown in Fig. 1. Biologically active compounds such as insulin, heparin, progesterone, and labeled leutinizing hormone were entrapped in glutaraldehyde crosslinked bovine serum albumin (BSA) and human serum albumin (HSA) microspheres. Effects of proteolytic enzymes such as trypsin, chymotrypsin, papain, and pronase-E on microspheres were studied in order to understand the biodegradability of the crosslinked proteins [76].
380
¨ ner and Kas O
Figure 1 SEM photographs of albumin microspheres.
Collagen is a fibrous protein and as a biomaterial it has several advantages in the field of tissue engineering. Collagen is a component of extracellular matrix and interacts with adhesins to promote cellular adhesion and growth. Collagen gels are available in the form of injections, surgical materials, corneal implants, and tissue sealants and implants. Bioactive proteins are embedded in a collagen rod with a diameter and length of 1.0 mm and 1.0 cm, respectively. Interferon released from these collagen rods for 1 week after implantion into the hypodermis [77]. Delivery of osteotropic biomolecules directly to the bone–implant interface can alter initial interactions between tissue and biomaterial. Type I collagen coatings containing a model molecule lysozyme were deposited. Larger amounts of collagen in the coatings allowed incorporation of more lysozyme. Release and retention of lysozyme were monitored over a 7-day period [78]. Atelocollagen is decomposition product of type I collagen and is derived from dermis of cattle. Atelocollagen minipellets increased the effectiveness of plasmid DNA in vivo for two months and inhibited nucleotide digestion in the serum [79]. Stimuli-sensitive intelligent biomaterials are self-regulated drug delivery systems. Protein (enzyme and antigen)-sensitive hydrogels can respond to large biomolecules such as proteins by swelling [80]. Peptide-based gene delivery has focused on the design of short synthetic peptides that overcome both extracellular and intracellular limitations of other gene delivery systems by binding and condensing DNA, targeting cells, rapidly releasing plasmids into the cytoplasm, and mediating efficient nuclear targeting [81]. Poly(amino acids) such as poly-L-lysine and poly-L-arginine can make complexes with DNA [82]. Viruses have proteins which play a role in membrane fusion and disruption. Transfection efficiency markedly increased with viral vectors incorporating whole virus to the gene delivery system, but this system has some risks. For this reason amphipathic and fusogenic peptides of the viruses have been investigated as components of gene delivery systems [83]. Encapsulated recombinant adenovirus can be released for longer than 10 days. Liquid mixture of atelocollagen and adenovirus is solidified at body temperature in the injection site and viruses are released gradually [79]. B. Monoclonal Antibodies as Delivery Systems Monoclonal antibodies can be used as drug delivery systems by coupling with toxins or radioactive substances to destroy tumor cells [84]. Tumor-specific antibodies are conjugated with drugs
Bioactive Molecules and Biodelivery Systems
381
to develop cancer therapeutics. Carcinoembryonic antigen (CEA), Alpha feta protein (AFP), and CA 125 are tumor antigens associated with some cancer cells. Elevated levels of these antigens indicate cancers of gastrointestinal tract, liver, or ovarian carcinomas. These antigens have value as tumor markers, and monoclonal antibodies capable of selectively binding to these tumor-specific antigens have been prepared for the purpose of recognizing and destroying tumor cells [85]. Approved monoclonal antibodies for therapeutic purposes are listed in Table 7. Targeting of drugs to desired cells involves conjugation of bioactive agents to biodelivery systems such as monoclonal antibodies. Required dose and toxic effects on the healthy cells would be reduced by using these conjugates [86]. Fab’ targeted conjugates showed higher antitumor efficacy than nontargeted ones [87]. C. Vaccine Adjuvants Adjuvants are used to enhance immune response to many antigens by influencing antibody titer and cell-mediated immunity [88]. Adjuvants are selected by considering characteristics of the antigen such as size and charge. Conventional adjuvants are found in the form of emulsions or suspensions. Recombinant hepatitis B vaccine is formulated with alum (aluminum phosphate suspension). New adjuvants with fewer side effects and higher antibody response are being studied extensively. Immunoadjuvant carriers such as niosomes, iscoms, squalene emulsions, microparticulate systems, and immune stimulant molecules such as interleukins, saponins and surfactants can be formulated in the form of vaccines having strong adjuvant effects [89]. Carrier adjuvants, which release antigen slowly from a depot system, interact with immune system cells and enhance macrophage activity. The nature of the immune response itself and sustained release of the antigen from the formulation are factors affecting immune response [90,91]. The ultimate goal in vaccination is to prepare single-shot, controlled release vaccine formulations that elicit prolonged and higher immune responses [91]. Interferon-␣-2b was administered as systemic adjuvant for 1 year for the treatment of high-risk melanoma, and promising results were obtained [92]. D. Polysaccharides as Biodelivery Systems A wide variety of strategies and materials have been developed to carry bioactive substances to target sites. In recent studies natural polymers have been preferred to prepare biocompatible, Table 7 Approved Monoclonal Antibodies for Therapeutic and Diagnostic Use Therapeutic monoclonal antibodies Anti-CD3 anticor (Orthoclone OKT3®) ReoPro® Panorex® In vivo diagnostics Myoscint® Oncoscint CR® Oncoscint OV® Indimacis 125® Prostacint® CeaScan® Tecnemab K-1®
Use To prevent organ rejection in kidney transplantations Treatment after coronary angioplasty Colorectal cancers Cardiac imaging agent Diagnosis and follow-up of colorectal cancers Diagnosis and follow-up of of ovary cancers Imaging of ovary adenocarcinoma Imaging of prostate cancers Recognizing carcinoembryonic antigen (CEA) Diagnosis and follow-up of melanoma
382
¨ ner and Kas O
nontoxic, effective, and stable nanoparticles. Polysaccharide chitosan–poly(ethylene oxide) nanoparticles were loaded with BSA (80% loading capacity) and released the protein in one week [93]. Alginate is a naturally found biopolymer produced from brown algae and used as a matrix for delivering bioactive materials. Lac–Z coding DNA encapsulated in alginate microparticles yielded significant immune response achieved by the mucosal route [94]. Insulin was encapsulated in calcium alginate beads coated with chitosan. This study has demonstrated that alginate–chitosan and alginate–chitosan–glutaraldehyde beads may be used for diabetes therapy [95]. Alginate polycation or alginate poly-L-lysine microcapsules have been successfully applied for encapsulating bioactive agents such as proteins [60]. Hemoglobin, BSA, and dextrans have been encapsulated in calcium alginate beads coated with chitosan [96]. Urease was immobilized within polyanionic carboxymethylcellulose–alginate microspheres coated with cationic polysaccharide–chitosan. Coating with chitosan improved the mechanical durability of the polyanionic microspheres as well as increased enzyme immobilization yield [97]. A natural nontoxic, nonimmunogenic polysaccharide chitosan is produced from deacetylation of chitin. Chitosan is a strong candidate for a nonviral effective gene delivery system for DNA and RNA, which strongly bind to cationic chitosan. Chitosan microparticles can be prepared by dispersing acidic solutions of the polymer in alkaline solutions to form droplets or by crosslinking with glutaraldehyde; then they are filtered and dried [59]. Chitosan-based vector/ DNA complexes are promising gene delivery systems due to the system characteristics such as the ability to condense DNA and form nanometer-sized particles [98]. Biomaterial-sensitive hydrogels have attracted considerable attention as intelligent materials. They can sense environmental changes and respond. Glucose-sensitive hydrogels exhibit swelling changes in response to glucose concentration [99] Dextran is another polysaccharide used as biodelivery system. Recombinant hIL-2 has been released from dextran-based hydrogels with satisfactory biological activity [100]. Gellan gum is an anionic polysaccharide produced by Pseudomonas elodea and could be a useful delivery agent for encapsulating bioactive substances [101]. E. Lipid Based Biodelivery Systems Micelles, liquid crystals, and liposomes are self-assembling colloids in which amphiphiles selfaggregate in the form of ordered structures. Polar lipids and DNA form liquid crystalline phases in the aqueous medium. Lipid membranes and colloidal DNA are important biological colloids which form particulate structures such as liposomes and DNA plasmids[102]. The use of nonviral plasmid-based gene therapeutics represents a promising in vivo gene therapy method which is easier and safer than viral systems. Gene delivery systems are designed to control proper targeting and distribution of the gene within the body by recognizing cell surface receptors and transfecting cells [103]. Liposomes are vesicles formed by double layers of amphiphilic phospholipids. Their particle size is small enough to form a relatively stable suspension in aqueous media. [104]. Szebeni and Alving showed that liposome-encapsulated hemoglobin indicated that this potential blood substitute could activate the component system of rats, pigs, and man [105]. Nishiya et al. prepared liposomes carrying both recombinant platelet membrane glycoproteins and fibrinogen. In this study, the authors showed that these liposomes interact with activated platelets to enhance the formation of platelet aggregates [106]. Liposomal plasmid DNA formulations can retain their structural integrity, and in vitro transfection efficiency is dependent on the size and surface charge of the vesicles [107].
Bioactive Molecules and Biodelivery Systems
383
Lipid vectors have advantages of being less toxic and less immunogenic than viral vectors. Cationic lipids are widely used as delivery systems for DNA. DNA-polycation complexes protect condensed DNA from nuclease digestion. Fusogenic materials such as DOPE have generally been added to the carrier systems due to the fusogenic properties [108]. Polyanionic DNA can be encapsulated in cationic liposomes through the charge and hydrophobic interactions between them. Cationic lipids are capable of effecting transfection. Association complexes formed between the cationic surfactants and plasmid DNA retain a stranded structure and interchalate appropriate dye molecules in organic solvents. These complexes can protect DNA from DNase I and serum digestion. In Fig. 2 and 3 gel retardation photographs of the cationic surfactantpDNA-oil emulsion complexes are seen protecting DNA from endonucleases [109,110]. In Fig. 4. a cationic carrier complex formation is represented [111]. Human erythrocyte ghosts as natural liposomes can be loaded with proteins and nucleotides to be used as blood substitutes or delivery systems [112]. Lipospheres are water-insoluble spherical microparticles consisting of hydrophobic triglycerides. Solid-lipid nanoparticles (SLN) are prepared by homogenizing melted natural lipids in aqueous surfactant solutions [113]. SLN formulations exhibited immune response as a potential vaccine adjuvant. Lipid emulsions are composed of oils, water, phospholipids, and are surfactants and used for nutritional or drug delivery purposes. Incorporation of SCT (salmon calcitonin) in the inner aqueous phase of a w/o/w emulsion protected the peptide from enzymatic degradation, and SCT was further protected by incorporating the protease inhibitor aprotinin in the outer aqueous phase [114]. By employing a new emulsifying technology, a novel perfluorocarbon emulsion (NeoPFC) with a higher ability for oxygen delivery in vivo compared with others was obtained. It is thought to be a candidate for blood substitute [115].
Figure 2 Protection of complexed pDNA from DNase I degradation. (Lane 1) Naked pDNA; (Lane 2) pDNA incubated with DNase I; (Lane 3) pDNA emulsion complex (1:0.23); (Lane 4) pDNA emulsion complex (1:0.23) incubated with DNase I:(Lane 5) pDNA emulsion complex (1:2.74); (Lane 6) pDNA emulsion complex (1:2.74) incubated with DNase I. CTAB: cetyl trimethyl ammonium bromide pDNA: CTAB ratios are given as g/nmol. (From ref. 109)
384
¨ ner and Kas O
Figure 3 Gel electrophoresis of pDNacationic emulsion complex incubated with human serum. Lane 1) pDNA cut with Bgl II; (Lane 2) uncut pDNA; (Lane 3–6) pDNA cationic emulsion complex incubated with human serum for 10 min, and 7, 10, and 24 h, respectively. (From ref. 109).
Figure 4 Illustration of a carrier system for DNA by using positively charged lipids or surfactants. (From Ref. 111.)
Bioactive Molecules and Biodelivery Systems
385
F. Nucleic Acid Based Delivery Systems The transfer of genes from one organism to another requires a cloning vector. Plasmids are small, circular pieces of DNA that are able to replicate independently within bacterial cells. They meet the two requirements of a good cloning vector: they are capable of carrying significant portions of donor DNA, and certain cloning hosts readily accept them. Plasmids are small DNA structures and are easily manipulated, but they have certain limitations. They are not good vectors for eukaryotic genes, and are unable to carry extremely large genome sequences. Cosmids and phages are also used to carry longer gene sequences into cells [11]. Gene therapy has focused on the delivery of genes and oligonucleotides to targeted organs and tissues. In order for gene therapy to succeed clinically, controlled release of the required amount of DNA should be performed in the planned period. Two strategies for gene therapy are ex vivo and in vivo methods. The first strategy is a difficult operation in which cells of the patient are taken out of the body and after transfection with the correct gene they are implanted back into the body. In the in vivo approach deficient gene is administered to the patient in a pharmaceutical delivery system via ordinary administration routes similar to an ordinary pharmaceutical application. In this procedure DNA must be delivered to the nucleus of the desired cells without degradation, and expression of the gene should be correctly controlled in the correct tissues [55]. In Fig. 5 A possible mechanism of gene therapy is summarized schematically [111].
Figure 5 Schematic of gene therapy. (a) Recognition of the plasmid delivery system by the cell to be transfected. (b) Entry of the system to the cell by endocytosis. (c) Integration of the foreign gene to the cellular DNA following endosomal release. (d) Protein expression and release. (From ref. 111.)
386
¨ ner and Kas O
The majority of the gene therapy experiments utilize viruses for delivering genes due to the high transfection efficiency of viral systems. But potential safety risks of viruses limit their utility as therapeutic agents. Safe and efficient nonviral delivery systems are receiving more attention than viral delivery systems due to their less immunugenic, highly targetable, and nontoxic properties [116]. Problems related with formulation, effectiveness, and stability need to be solved. Low in vivo stability and transfection efficiency of the nonviral gene delivery systems are major concerns in their development. Studies on gene delivery systems are increasing rapidly, and a large number of investigational new drug applications for gene therapy have started. Cystic fibrosis, cancers, cardiovascular diseases, and hemophilia are target diseases for gene therapy applications. Main targets for gene therapy are inherited single-gene disorders and cancers [117]. There is no gene therapy product on the market, but clinical phase studies are ongoing. Especially for cancers, phase I/II clinical trials provide new treatment opportunities [129]. Suitable biomaterials for gene transfer should be biocompatible, protective, biodegradable and safe and exhibit controlled release, prolonged action. Extracellular nonspecific interactions and intracellular trafficking to nucleus are also major concerns [59]. Plasmid vectors including transcription factor binding sites were constructed and nuclear entry can be enhanced resulting in higher gene expression [118]. Intracellular synthesis of single-stranded DNA is a new method for chromosomal triplex formation. Transfection efficiency of this method is higher than synthetic triplex-forming oligonucleotides [119].
G. Cells, Organs and Tissues as Bioactive Biodelivery Systems Whole cells, organs, and tissues used as human therapeutic systems have possibilities for new therapeutic applications. Stem cell replacement, hematopoetic stem cell transplantation, or stem cell–based gene therapy may provide exciting alternatives for the immune-based therapies [120]. Prakash and Chang reported the use of artificial cells microencapsulated in genetically engineered E. coli DH5 cells for lowering plasma creatinine in vitro and in vivo [121]. Xenotransplantation means to use animal organs, tissues, or cells as human therapeutic products-Xenografts are live cells, tissues, and organs used in xenotransplantation procedures. Limitations in the use of human organs has led to the use of transgenic animals to produce human-compatible xenoplantation products. In this application the major problem is organ rejection. During 1999–2000, three guidances on xenoplantation were published in the United States: one by Public Health Service (PHS) and two by FDA. Gene delivery systems for cardiovascular diseases can be prepared by using arterial transfection [122]. Sustained release of DNA from coated stents resulted with arterial transfection. By using tissue engineering techniques biological substitutes to organs are developed to maintain and improve human tissue functions. Biological substitutes are cells or combinations of different cells to be implanted on a scaffold such as natural collagen or synthetic biocompatible materials to form a tissue [123].
IV. CONCLUSION Growing knowledge in the field of bioactive molecules and biodelivery systems, with detailed insight in peptides and proteins, vaccines, lipids, polysaccharides, nucleic acids, and oligonucleotides has opened new fields in pharmaceutical biotechnology. Besides the approved therapeutics
Bioactive Molecules and Biodelivery Systems
387
in the market, there are hundreds undergoing preclinical and clinical studies. Future developments in this growing field will find exciting applications in the health care area.
REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.
18. 19. 20. 21. 22. 23. 24. 25.
IUPAC. Compendium of Chemical Terminology. 2nd Ed., 1997. Initial sequencing and analysis of the human genome. Nature 2001; 409:860–921. The sequence of the human genome. Science 2001; 291:1305–1351. Setti EWL, Micetich R. Modern drug discovery and lead discovery: an overview. Curr. Med. Chem 1996; 3:317–324. International Conference on Harmonization. Q6B: Specifications, test procedures and acceptance criteria for biotechnological/biological products. Federal Register, August 18, 1999. International, Conference on Harmonization. Guidance for industry, Q7A GMP guidance for active pharmaceutical ingredients, USDHHS, FDA, CDER, CBER, August 2001, ICH. European Union. The rules governing medicinal products in the European Community. Official Publications of the European Community Vol.II, 1989, Vol III, 1992. Murano G. International conference on harmonization–critical discussion of the biotech specifications document. Curr.Opin.Biotechnol 2000; 11(3):303–308. Kuchel PW, Ralston GB. Shaum’s Outline of Theory and Problems of Biochemistry. NY: Mc GrawHill, 1998. Walsh G. Blood products and therapeutic enzymes, In: Biopharmaceuticals: Biochemistry and Biotechnology. New York: John Wiley and Sons, 1998:387–406. Glick BR, Pasternak JJ. Molecular Biotechnology. Principles and Applications of Recombinant DNA, ASM Press. 1994:5–15, 56–64, 331–353, 359–380. Plaue S, Muller S, Briand JP, Van Regenmortel HM. Recent advances in solid-phase peptide synthesis and preparation of antibodies to synthetic peptides. Biologicals 1990; 18:147–157. Landt O, Grunert HP, Hahn U. A general method for rapid site-directed mutagenesis using the polymerase chain reaction. Gene 1990; 96:125–128. Johnson I. Human insulin from recombinant DNA technology. Science 1983; 219:632–637. Cohen S, Chang ACY, Boyer H, Helling CB. Construction of biologically functional bacterial plasmid in vivo. Proc. Natl. Acad. Sci. USA 1973; 70:3240–3244. Kohler G, Milstein C. Continuous culture of fused cells secreting antibody of pre-defined specificity. Nature 1975; 256:495–497. Hoffmann FK. Scaleup of production and purification of cell-expressed proteins. In: Klegerman ME, Groves MJ, eds. Pharmaceutical Biotechnology: Fundamentals and Essentials. IL: Interpharm Press, 1992:138–164. Pringle AT. Fermentation and mammalian cell culture. In: Klegerman ME, Groves MJ, eds. Pharmaceutical Biotechnology:Fundamentals and Essentials, Interpharm Press. 1992:115–138. Schu¨ger K. Integrated processing of biotechnology products. Biotechnol. Adv 2000; 18(7):581–599. Andersen DC, Krummen L. Recombinant protein expression for therapeutic applications. Curr. Opin. Biotechnol 2002; 13(2):117–123. Swartz JR. Advances in Escherichia coli production of therapeutic proteins. Curr. Opin. Biotechnol 2001; 12:188–194. Werner RG, Noe W, Kopp K, Schulter M. Appropriate mammalian expression systems for biopharmaceuticals. Drug Res 1998; 48:870–880. Giddings G, Allison G, Brooks D, Carter A. Transgenic plants as factories for biopharmaceuticals. Nat. Biotechnol 2000; 18:1151–1155. Guzman M, Belin D, Carson MJ, Beckwith J. Tight regulation, modulation, and high level expression by vectors containing the arabinose PBAD promoter. J. Bacteriol 1995; 177:4121–4130. Nakano H, Yamane T. Cell-free protein synthesis systems. Biotechnol. Adv 1998; 16:367–384.
388
¨ ner and Kas O
26. Baker KN, Rendall MH, Hills AE, Hoare M, Freedman RB, James DC. Metabolic control of recombinant protein N-glycan processing in NSO and CHO cells. Biotechnol. Bioeng 2001; 73: 188–202. 27. Weikert S, Papae D, Briggs J, Cowfer D, Tom S, Gawlitzek M, Lofgren J, Mehta S, Chisholm V, Modi N. Engineering chinese hamster ovary cells to maximize sialic acid content of recombinant proteins. Nat. Biotechnol 1999; 17:1116–1121. 28. Reed JC, Tomaselli KJ. Drug discovery opportunities from apoptosis research. Curr. Opin.Biotechnol 2000; 11(6):586–592. 29. Kim NS, Lee GM. Overexpression of Bcl-2 inhibits sodium butyrate–induced apoptosis in chinese hamster ovary cells resulting in enhanced humanized antibody production. Biotechnol. Bioeng 2001; 71:184–193. 30. Zahn-Zabal M, Kobr M, Girod PA, Imhof M, Chatellard P, Jesus M, de Wurm F, Mermod N. Development of stable cell lines for production or regulated expression using matrix attachment regions. J. Biotechnol 2001; 87:29–42. 31. Goodrick JC, Xu M, Finnegan R, Schilling BM, Chavi SS, Hoppe H, Wan NC. High-level expression and stabilization of recombinant human chitinase produced in a continuous constitutive Pichia pastoris expression system. Biotechnol. Bioeng 2001; 74:492–497. 32. Marshal I, Jarvis DL, Cacan R, Verbert A. Glycoproteins from insect cells: sialyated or not?. Biol. Chem 2001; 382:151–159. 33. Larrick JW, Thomas DW. Producing proteins in transgenic plants and animals. Curr. Opin. Biotech 2001; 12(4):411–418. 34. Desi UA, Sur G, Daunert S, Babbitt R, Li Q. Expression and affinity purification of recombinant proteins from plants. Protein Expr. Purif 2002; 25(1):195–202. 35. Kruth SA. Biological response modifiers: interferons, interleukins, recombinant products, liposomal products. Vet. Clin. North Am. Small Anim. Pract 1998; 28(2):269–295. 36. Vose JM, Armitaje JO. Clinical applications of hematopoietic growth factors. J. Clin. Oncol 1995; 13:1023–1035. 37. Kim KJ, Li B, Winner J, Armanini M, Gillett N, Philips HS, Ferrera N. Inhibition of vascular endothelial growth factor–induced angiogenesis supresses tumor growth in vivo. Nature 1993; 362: 841–844. 38. Waldmann T. Monoclonal antibodies in diagnosis and therapy. Science 1991; 252:1657–1662. 39. Adair JR, Zivin RA, Guzman NA, Iqbal K. Monoclonal antibody–based pharmaceuticals. In: Crommelin DJ, Sindelar RD, eds. Pharmaceutical Biotechnology: An Introduction for Pharmacists and Pharmaceutical Scientists. Amsterdam: Overseas Publishers Association, 1997:279–287. 40. Morrison SL, Johnson MJ, Herzenberg LA, Oi VT. Chimeric human antibody molecules:mouse antigen-binding domains with human constant region domains. Proc. Natl. Acad. Sci. USA 1984; 81:6851–6855. 41. Padlan EA. A possible procedure for reducing the immunogenicity of antibody variable domains while preserving their ligand binding properties. Mol. Immunol 1991; 28:489–498. 42. Green L. Antibody engineering via genetic engineering of the mouse: xenomouse strains are a vehicle for the facile generation of therapeutic human monoclonal antibodies. J. Immunol. Methods 1999; 231:11–23. 43. Chadd HE, Chamow SM. Therapeutic antibody expression technology. Curr. Opin. Biotechnol 2001; 12(1):188–194. 44. Powell MF. Drug delivery issues in vaccine development. Pharm. Res 1996; 13(12):1777–1785. 45. Brown F. Recombinant subunit and peptide viral vaccines. In: Gregoriadis G, Allison AC, Poste G, eds. Immunological Adjuvants and Vaccines. NY: Plenum Press, 1989:163–174. 46. Davis HL. Immune responses with direct gene transfer, DNA vaccines and implications for gene therapy. In: Gregoriadis G, McCormack B, eds. Targeting of Drugs 5: Strategies for Oligonucleotide and Gene Delivery in Therapy. NY: Plenum Press, 1995:21–27. 47. Rolland AP. From genes to gene medicines: recent advances in nonviral gene delivery. Crit. Rev in Ther.Drug Carr. Sys 1998; 15:143–198.
Bioactive Molecules and Biodelivery Systems
389
48. FDA. 1996. Points to consider on plasmid DNA vaccines for preventive infectious disease indications, 1996. Food and Drug Administration Center for Biologics Evaluation and Research, Office of Vaccine Research and Review. 49. Lee JC. Biopharmaceutical formulation. Curr. Opin. Biotechnol 2000; 1(1):81–84. 50. Cleland JL, Daugherty A, Mrsny R. Emerging protein delivery methods. Curr. Opin. Biotechnol 2001; 12(2):212–219. 51. Hanson MA, Rouan SKE. Introduction to formulation of protein pharmaceuticals. In: Borchardt RT, Ahern TJ, Manning MC, eds. Stability of Pharmaceutics. Part B: In Vivo Pathways of Degradation and Strategies for Protein Stabilization. New York: Plenum Press, 1992:236–257. 52. Heller J. Polymers for controlled parenteral delivery of peptides and proteins. Adv. Drug Del. Rev 1993; 10:163–204. 53. Whitely MA, Chang M, Park E, Langer R. Coated alginate microspheres; factors influencing the controlled delivery of macromolecules. J. Appl. Polym. Sci 1991; 43:2123–2135. 54. Hennink WF, Franssen O, van Dijk-Wolthuis WNE, Talsma H. Dextran hydrogels for controlled release of proteins. J. Control. Rel 1997; 48:107–116. 55. Peppas NA. Hydrogels and drug delivery. Curr. Opin. Colloid Interface Sci 1994; 2:531–537. 56. Damge C, Vrancky H, Balschmidt P, Couvreur P. Poly(alkylcyanoacrylate) nanospheres for oral administration of insulin. J. Pharm. Sci 1997; 86:1403–1409. 57. Langer R. New methods of drug delivery. Science 1990; 249:1527–1533. 58. Burgess DJ. Drug delivery aspects of biotechnology products. In: Pezzuto JM, Johnson ME, Manasse HR, eds. Biotechnology and Pharmacy. NY: Chapman and Hall, 1993:116–131. 59. Kas HS. Chitosan: properties, preparations and applications to microparticulate systems. J. Microencap 1997; 14:689–711. 60. Kumar MNVR. Nano- and microparticles as controlled drug delivery devices. J. Pharm. Pharmaceut. Sci 2000; 3(2):234–258. 61. Powell MF. A compendium and hydropathy/flexibility analysis of common reactive sites in proteins: reactivity at Asn, Asp, Gln, and Met motifs in neutral pH solution. In: Pearlman R, John Wang Y, eds. Formulation, Characterization, and Stability of Protein Drugs, Plenum Press. 1996. 62. Reddy IK. Stability of peptide and protein drugs. In: Zito WS, ed. Pharmaceutical Biotechnology. Basel: Technomic Publishing, 1997:137–157. 63. Borchardt RT. Special biotechnology problems with biotechnology-derived drugs. In: Blume H, Midha KK, eds. Bio-international 2:Bioavailability, Bioequivalence and Pharmacokinetic Studies. Stuttgart: Medpharm SP, 1995:59–160. 64. Inada Y, Furukawa M, Sasaki H, Kodera Y, Hiroto M, Nishimura H, Matsushima A. Biomedical and biotechnological applicatios of PEG- and PM-modified proteins. Trends Biotechnol 1995; 13(3): 86–91. 65. Bhadra D, Bhadra S, Jain P, Jain NK. Pegnology: a review of PEGylated systems. Pharmazie 2002; 57(1):5–29. 66. Francis GE, Delgado C, Fisher D. PEG-modified proteins. In: Borchardt RT, Ahern TJ, Manning MC, eds. Stability of Protein Pharmaceuticals Part B: In Vivo Pathways of Degradation and Strategies for Protein Stabilization, Plenum Press. 1992:236–257. 67. Grace M, Yougster S, Gitlin G, Sydor W, Xie L, Westreich L, Jacobs S, Brassard D, Bausch J, Bordens R. Structural and biologic characterization of pegylated recombinant IFN-␣ 2b. J. Interferon Cytokine Res 2001; 21(12):1103–1115. 68. Li YP, Pei YY, Zhou ZH, Zhang XY, Gu ZH, Ding J, Gao XJ, Zhu JH. PEGylated recombinant human tumor necrosis factor alpha: pharmacokinetics and anti-tumor effects. Biol. Pharm. Bull 2001; 24(6):666–670. 69. Meyer JD, Manning MC. Hydrophobic ion pairing: altering the solubility properties of biomolecules. Pharm. Res 1998; 15(2):188–193. 70. Nagarsekar A, Ghandehari H. Genetically engineered polymers for drug delivery. J. Drug Target 1999; 1:11–32. 71. Sood A, Panchagnula R. Per oral route: an opportunity for protein and peptide drug delivery. Chem. Rev 2001; 101(11):3275–3303.
390
¨ ner and Kas O
72. Squier CA, Wertz PW, Kremer M. Methods for studying protein and peptide delivery accross oral ¨ ner F, eds. Peptide and Protein Delivery. (buccal) mucosa in Recent Advances. In: Hincal AA, O Paris: Editions de Sante´, 1998:32–44. 73. Morimoto K, Katsumata H, Yabuta T, Iwanaga K, Kakemi M, Tabata Y, Ikada Y. Gelatin microspheres as a pulmonary delivery system: evaluation of salmon calcitonin absorption. J. Pharm. Pharmacol 2000; 52:611–617. 74. Truong-Le VL, August JT, Leong KW. Controlled gene delivery by DNA-gelatin nanospheres. Hum. Gene Ther 1998; 9(12):1709–1717. ¨ ner L, Akalin N, Sargon MF. In vivo/in vitro characterization of bovine serum 75. Eroglu H, Kas HS, O albumin microspheres containing dexamethazone sodium phosphate. STP Pharma. Sci 2000; 10(4): 303–308. 76. Tulsani NB, Kumar A, Pasha Q, Kumar H, Sharma UP. Immobilization of hormones for drug targeting. Art. Cells Blood Subs, and Immob. Biotech 2000; 28:503–519. 77. Fujioka K, Takada Y, Sato S, Miyata T. Novel delivery system for proteins using collagen as a carrier material: the minipellet. J. Control Rel 1995; 33:307–315. 78. Pueo DA. Release and retention of biomolecules in collagen deposited on orthopedic biomolecules. Art. Cells, Blood Subs. Immob. Biotech 1999; 27:65–75. 79. Ochiya T, Takahama Y, Nagahara S, Sumita Y, Hisada A, Itoh H, Nagai Y, Terada M. New delivery system for plasmid DNA in vivo using atelocollagen as a carrier material: the minipellet. Nat. Med 1999; 5:707–710. 80. Miyata T, Uragami T, Nakamae K. Biomolecule sensitive hydrogels. Adv. Drug Del. Rev 2002; 54(1):79–98. 81. Mahato RI, Monera OD, Smith LC, Rolland A. Peptide based gene delivery. Curr. Opin. Mol. Ther 1999; 1:226–243. 82. Garnett MC. Gene-delivery systems using cationic polymers. Crit. Rev. Ther. Drug Car. Sys 1999; 16(2):147–207. 83. Takahashi S. Conformation of membrane fusion active 20-residue peptides with or without lipid bilayers. Implication of alpha helix formation for membrane fusion. Biochemistry 1990; 29: 6257–6260. 84. Groves DJ. Monoclonal antibodies. In: Pharmaceutical Biotechnology: Fundamentals and Essentials Klegerman ME, Groves MJ, eds, Interpharm Press. 1992:104–114. 85. Rapley R. The biotechnology and applications of antibody engineering. Mol. Biotechnol 1995; 3(2): 139–154. 86. Conti S, Polonelli L, Frazzi R, Artusi M, Bettini R, Cocconi D, Colombo P. Controlled delivery of biotechnological products. Curr. Pharm. Biotech 2000; 1:313–323. 87. Lu ZR, Shiah JG, Kopeckova P, Kopecek J. Preparation and biological evaluation of polimerizable antibody Fab’ fragment targeted polymeric delivery systems. J. Control. Rel 2001; 74(1–3):263–8. 88. Gupta R, Siber G. Adjuvants for human vaccines: current status, problems and future prospects. Vaccine 1995; 13(14):1263–1274. 89. Newman MJ, Powell MF. Immunological and formulation design considerations for subunit vaccines. In: Powell MF, Newman MJ, eds. Vaccine Design: The Subunit and Adjuvant Approach: Plenum Press, 1995:1–42. 90. Fitzgerald D, Mrsny RJ. New approaches to antigen delivery. Crit.Rev. in Ther.Drug Carr. Sys 2000; 17:165–248. 91. Schijns VE. Induction and direction of immune responses by vaccine adjuvants. Crit. Rev. Immunol 2001; 21(1–3):75–85. 92. Kirkwood JM. Systemic adjuvant treatment of high-risk melanoma: the role of interferon-␣ 2b and other immunotherapies. Eur. J. Cancer 1998; 34(Suppl 3):S12–17. 93. Calvo P, Remunan-Lopez C, Vila-Jato JL, Alonso MJ. Chitosan and chitosan/ethylene oxide block copolymer nanoparticles as novel carriers for proteins and vaccines. Pharm. Res 1997; 14(10): 1431–1436. 94. Mittal SK, Aggarwal N, Sailaja G, van Olphen A, HogenEsch H, North A, Hays J, Moffatt S. Immunization with DNA, adenovirus or both in biodegradable alginate microspheres: effect of route of inoculation on immune response. Vaccine 2000; 19:253–263.
Bioactive Molecules and Biodelivery Systems
391
95. Onal S, Zihniog˘lu F. Encapsulation of insulin in chitosan-coated alginate beads: oral therapeutic peptide delivery. Art. Cells Blood Subs. Immob. Biotech 2002; 30:229–237. 96. Huguet ML, Dellacherie E. Calcium alginate beads coated with chitosan: effect of structure encapsulated materials on their release. Process Biochemistry 1996; 31:745–751. 97. Elcin AE, Elcin YM. Polycation-coated polyanion microspheres of urease for urea hydrolysis. Art. Cells Blood Subs. Immob. Biotech 2000; 28:95–111. 98. Erbacher P, Zou S, Bettinger T, Steffan A, Remy J. Chitosan based vector/DNA complexes for gene delivery: biophysical characteristics and transfection ability. Pharm. Res 1998; 15(9):1332–1339. 99. Trehan A, Ali A. Recent approaches in insulin delivery. Drug Dev. Ind. Pharm 1998; 24:589–597. 100. Cadee JA, de Groot CJ, Jiskoot W, den Otter W, Hennink WE. Release of recombinant interleukin2 from dextran based hydrogels. J Control Release 2002; 78(1–3):1–13. 101. Chilvers GR, Morris VJ. Coacervation of gelatin gellan gum mixtures and their use in microencapsulation. Carbohydr. Polym 1987; 7:111–120. 102. Lasic DD, Templeton NS. Bioorganic colloids: macromolecules, DNA, self-assemled particles, and their complexes. In: Gene Therapy: Therapeutic Mechanisms and Strategies Templeton NS, ed, Marcel Dekker. 2000:241–265. 103. Anwer K. Targeted gene delivery. Two pronged approach. Crit. Rev. Ther. Drug Carr. Syst 2000; 17:377–424. 104. Gregoriadis G. Engineering liposomes for drug delivery: progress and problems. Trends Biotechnol 1995; 13:527–537. 105. Szebeni J, Alving CR. Complement-mediated acute effects on liposome-encapsulated hemoglobin. Art. Cells Blood Subs. Immob. Biotech 1999; 27:23–41. 106. Nishiya T, Kainoh M, Murata M, Handa M, Ikeda Y. Platelet interactions with liposomes carrying recombinant platelet membrane glycoproteins or fibrinogen approach to platelet substitutes. Art. Cells Blood Subs. Immob. Biotech 2001; 29:453–464. 107. Gregoriadis G, Saffie R, Hart SL. High yield incorporation of plasmid DNA within liposomes: effect of DNA integrity and transfection efficiency. J. Drug Target 1996; 3:469–475. 108. Lee RJ, Huang L. Lipidic vector systems for gene transfer. Crit. Rev. Ther. Drug Carr. Syst 1997; 14(2):173–206. ¨ ner F. Preparation and characterization of a cationic emulsion as 109. Barut KD, Cos¸kun-Ari FF, O potential DNA carrier system. Proceedings of the 4th World Meeting on Pharmaceutics, Biopharmaceutics and Pharmaceutical Technology, 2002:225–226. 110. Bhattacharya S, Mandal SS. Interaction of surfactant with DNA. Role of hydrophobicity and surface charge on intercalation and melting. Biochim. Biophys. Acta 1997; 1323:29–44. ¨ ner F. Kontrollu Salim sistemlerinin biyoteknoloji alaninda uygulamalari. In: Gu¨rsoy AZ, ed. 111. O Kontrollu Salim Sistemleri. Istanbul: Elma Bilgisayar Basim, 2002. 112. Zolla L, Lupidi G, Marcheggiani M, Falconi G, Brunori M. Encapsulation of proteins into human erythrocytes:a kinetic investigation. Biochim. Biophys. Acta 1990; 1024:5–9. 113. Mu¨ller RH, Mader K, Gohla S. Solid lipid nanoparticles (SLN) for controlled drug delivery—a review of the state of the art. Eur. J. Pharm. Biopharm 2000; 50(1):161–167. ¨ ner F. Oral w/o/w multiple emulsion formulation of salmon calcitonin: in vitro// 114. Dog˘ru T, C¸alis¸ S, O in vivo evaluation. J. Clin. Pharm. Ther 2000; 25:435–443. 115. Sakanoue J, Tamura M, Fukushima S, Takeuchi Y, Sakuma I, Kitabake A. Assessment of newly developed perfluorocarbon emulsion: oxygen carrying capacity as the blood substitute in vivo. Art. Cells. Blood Subs. Immob. Biotech 2001; 29:389–397. 116. Smith A. Viral vectors in gene therapy. Annu. Rev. Microbiol 1995; 49:807–838. 117. Mhashilkar A, Chada S, Roth JA, Ramesh R. Gene therapy, therapeutic approaches and implications. Biotech. Adv 2001; 19(4):279–297. 118. Mesika A, Grogoreva I, Zohar M, Reich Z. A regulated, NFB-assisted import of plasmid DNA into mammalian cell nuclei. Mol. Ther 2001; 5:653–657. 119. Datta HJ, Glazer PM. Intracellular generation of single-stranded DNA for chromosomal triplex formation and induced recombination. Nucleic Acids Res 2001; 29:5140–5147. 120. Gearhart J. New potential for human embryonic stem cells. Science 1998; 282:1061–1062.
392
¨ ner and Kas O
121. DHHS. 2001 Stem cells: scientific progress and future directions, Department of Health and Human Services, June 2001. 122. Klughertz BD, Jones PL, Cui X, Chen W, Meneveau NF, Felice S, Connolly J, Wilensky RL, Levy R. Gene delivery from a DNA controlled release stent in porcine coronary arteries. Nat. Biotech 2000; 18:1181–1184. 123. Prakash S, Chang TMS. Artificial cells microencapsulated genetically engineered E. coli DH5 cells for the lowering of plasma creatinine in vitro and in vivo. Art. Cells Blood Subs. Immob. Biotech 2000; 28:397–408.
18 Biodegradable Nanoparticles as Drug Delivery Systems for Parenteral Administration Michael Chorny, Hagit Cohen-Sacks, Ilia Fishbein, Haim D. Danenberg, and Gershon Golomb The Hebrew University of Jerusalem Jerusalem, Israel
I. INTRODUCTION The pharmacological efficacy of parenterally administered drugs is dependent on their distribution throughout the body as well as the elimination route and kinetics. More effective and safe drug action may be achieved by accumulating the active substance at the target site while reducing its systemic levels. Among several strategies for controlling the pharmacokinetics of parenterally administered therapeutic substances, drug incorporation into submicroscopic biodegradable drug carriers has been one of the most extensively investigated [1–9]. These carriers include oil-in-water nanoemulsions, liposomes, and nanoparticles. Nanoparticles are solid bodies in the submicronial size range, usually made of polymers. Depending on their composition nanoparticles may be divided in two major classes: nanospheres and nanocapsules. Nanospheres are composed of a solid matrix, whereas nanocapsules have a characteristic core-shell structure. The drug may be chemically bound to the particle-forming polymer, adsorbed to the nanoparticle surface or entrapped in the nanoparticle. In the latter case, it may be either dissolved in the nanocapsule liquid core or dispersed in the nanosphere matrix. Recently drug nanosuspensions showing potential as a new class of injectable nanoparticulate formulations have been developed and described [10]. To date the intensive research in the field of polymer-based particulate drug carriers for parenteral drug delivery has yielded only a few microparticulate products, whereas nanoparticulate formulations for therapeutic uses have not yet appeared on the market, apparently reflecting the difficulty in developing a formulation process capable of producing nanoparticles acceptable from a regulatory standpoint. This chapter presents methods most commonly employed for preparation of biodegradable nanoparticles. The methods are discussed in terms of nanoencapsulation efficiency and basic requirements applying to injectable formulations. Comprehensive reviews on carrier biodistribution and fate as well as advances in targetable and long circulating carrier technology can be found in the literature [5,8,11–15], therefore this subject lies largely beyond the scope of this review. 393
394
Chorny et al.
II. NANOPARTICLES FOR PARENTERAL DRUG ADMINISTRATION Biodegradable nanoparticles historically emerged as a subclass of microparticulate implantable devices. This is the reason why the nanoparticle formulation technology, having been subject to extensive development in the last 15 years, basically relies on the principles of microparticle preparation methods developed in the 1970s. These particulate systems also share a number of common features when applied for delivery of therapeuticals. With appropriate modifications polymer-based particulate systems may be suitable for encapsulation of practically any kind of pharmacological agent, including biological macromolecules, such as proteins and DNA. A hydrophobic, water-insoluble substance may be formulated and administered in aqueous dispersion of solid particles composed of a lipophilic polymer. Entrapment in the polymeric particles protects chemically labile agents from fast degradation in biological milieu. Adjustment of the particle manufacturing protocol makes it possible to achieve the release kinetics optimal for the pharmacological action of a therapeutic compound. However, a number of important formulation properties related to safety and efficacy appear to be carrier size dependent [16–18], making the use of nanoparticles especially advantageous in several respects. These include reduced inflammatory reaction in tissues due to submicronial carrier size [19,20], tissue targeting via modifications of carrier size and surface chemistry [3,5], and the possibility of cellular localization, which is critical for intracellularly acting agents [21–23]. In preparation of injectable nanoparticles the requirements stated by the parenteral administration route should be addressed along with the treatment efficacy. The requirements referring to the safety of the final dosage form may be regarded as obligatory and include particle size, sterility [24], and use of biocompatible components (otherwise special purification steps should be included in the preparation procedure). The U.S. Pharmacopeia states strict limits on presence of particles with sizes exceeding 10 and 25 m in parenteral preparations [25]. The particle size of particular concern in preparations intended for intravenous administration has not been clearly delineated, but it has been suggested that since erythrocytes have a diameter of approximately 4.5 m, particles of more than 5 m should be avoided [24]. Particles exceeding this size will primarily be entrapped by the capillary bed of the lungs forming emboli; therefore it is recommended to aim the particle size in the formulation below this limit and preferably below 1 m [17,26]. Cases in which the particle disposition in the lungs thus achieved may be utilized to deliver radio- and chemotherapeuticals for the treatment of malignant disorders present a notable exception. However, since it is commonly agreed that submicronial particle size is a prerequisite for parenterally delivered particulate pharmaceuticals, this chapter will further focus on formulation methods suitable for the production of nanoparticles.
III. POLYMER TYPE AND FORMULATION TECHNIQUES Nanoparticles may be formulated using natural and synthetic polymers. Choosing natural polymers, such as proteins (albumin and gelatin) and polysaccharides (dextran, alginate, and chitosan), would apparently seem beneficial from the formulation biocompatibility standpoint [27]. However, until recently methods available for preparing nanoparticles from natural macromolecules relied on heat denaturation and crosslinking agents, which might be harmful to chemically unstable drugs and result in formulations with considerable immunogenicity and toxicity [28]. In addition, the hydrophilic polymers are not capable of providing protracted release kinetics for small drug molecules, thus limiting possible applications to delivery of biological macromolecules or drugs, for which immediate action is desirable. Obviously, for this reason reports on the use of these polymers for preparing injectable nanoparticles are scarce in the literature.
Biodegradable Nanoparticles as Drug Delivery Systems
395
Chitosan-based nanoparticles formulated by ionotropic gelation and complex coacervation, two recently developed methods, and spongelike alginate/polylysine nanospheres present valuable exceptions.
IV. NANOPARTICLES FORMULATED WITH NATURAL MACROMOLECULES Ionotropic gelation method for formulation of chitosan-based nanoparticles has the advantage of avoiding use of crosslinking agents and employing mild preparation conditions [29]. The method was applied successfully for nanoencapsulation of DNA [29] and hydrophilic high and low molecular weight proteins, such as tetanus toxoid and insulin [30], as well as a hydrophobic peptide, Cyclosporine A [31]. In the latter case the authors hypothesized that the drug dispersed as nanocrystals was entrapped in the core of the polysaccharide particles. The technique was also found applicable for incorporation of low molecular weight drugs, such as doxorubicin [32]. The entrapment efficacy of doxorubicin was significantly enhanced by incorporating dextran sulfate in the formulation, and sustained in vitro release kinetics was demonstrated. Leong et al. [33] used complex coacervation to entrap DNA in gelatin or chitosan nanospheres. Using this method no crosslinking was required to reproducibly obtain chitosan nanospheres stable for a few hours with encapsulation efficacy as high as 95%. The chitosan-DNA nanoparticles modified with polyethylene glycol (PEG) could be lyophilized without particle aggregation; they demonstrated improved formulation stability in buffer and extended residence in the circulation compared to uncoated nanoparticles [34]. Recently spongelike alginate nanospheres were developed by Aynie et al. [35] as potential carriers for oligonucleotides (ODN). The nanosponges are formed from calcium alginate pre-gel following addition of polylysine solution. The alginate nanospheres provide protection to the associated ODN and appear to be promising carriers for ODN delivery to lungs, liver, and spleen.
V. NANOPARTICLES FORMULATED WITH SYNTHETIC POLYMERS Synthetic biocompatible and biodegradable polymers were evaluated as an alternative to the natural macromolecules for preparation of injectable nanoparticles. The three most extensively studied classes of synthetic biodegradable polymers are poly(alkylcyanoacrylates), aliphatic polyesters, and polyanhydrides. The polymers of these families have a long history of use as medical materials and implantable devices (sutures and suture coatings, bioresorbable surgical glues, controlled drug delivery implants, etc.), and therefore having the advantage of wellestablished biocompatibility [36]. The lipophilicity of these polymers depends on their composition and molecular weight and may be precisely controlled, which is especially important for successful encapsulation and sustained release of low molecular weight therapeuticals, most of which are amphiphilic and lipophilic compounds. Moreover, it is possible to synthesize mixed polymers by combining different families, thereby obtaining materials with improved biodistribution and biodegradation characteristics [37–39]. A. Poly(alkylcyanoacrylates) Poly(alkylcyanoacrylate)-based nanoparticles are most commonly formulated by in situ polymerization of corresponding monomers. Cyanoacrylates are able to polymerize easily in various media including water, and in contrast to other acrylic derivatives, their polymerization does
396
Chorny et al.
not require external energy input. In vivo degradation of poly(alkylcyanoacrylates) (PACA) may involve both hydrolytic chain scission and enzymatic ester hydrolysis, the second mechanism shown to be dominant and leading to formation of metabolites with low toxicity. In the course of ester hydrolysis the polymer chain remains intact, but the molecule solubility in water gradually increases until it is completely soluble. The resulting low molecular weight polymeric acid is rapidly excreted from the body. The degradation rate of PACA is dependent on the length of the alkyl chain; the higher homologs degrade at a slower rate [40]. PACA nanospheres are prepared by emulsification polymerization, in which droplets of water-insoluble monomers are emulsified in an aqueous phase containing a stabilizer (dextran 70 or Poloxamer 188). Iⳮ, CH3COOⳮ, Brⳮ, or OHⳮ are often used as initiators, although even hydroxyl anions deriving from water dissociation are capable of initiating the reaction. Depending on the solubility of the monomers in water, the reaction may be initiated either in aqueous solution or in monomercomposed micelles formed above the critical micellar concentration. In the latter case the monomers reach micelles by diffusion from the emulsion droplets serving as a reservoir. In the course of the reaction the micelles solidify, eventually forming nanoparticles. Alternatively, when more water-soluble monomers react in the aqueous phase, the nanoparticles may form by phase separation of the polymer as its solubility in the aqueous medium decreases. In both cases the nanoparticles are composed of a large number of short polymeric molecules. Indeed, it was shown that PACA nanospheres are built by numerous small oligomeric subunits with uniform molecular weights ranging between 500 and 1000 rather than by several long polymeric chains [41]. 1. Drug Entrapment in PACA Nanospheres The polymerization rate determining the size and structure of the nanoparticles depends inversely on the acidity of the medium. Therefore the reaction is usually performed under acidic conditions (pH 2–4) in order to avoid rapid polymerization resulting in formation of large agglomerates. Optimizing the pH of the aqueous phase and the amount of the surfactant, it was possible to prepare nanoparticles less than 50 nm in size [42]. The adsorption capacity of the particles is a function of both the nature and amount of the monomer and the hydrophilicity of the drug. Higher loading is generally achieved by employing alkylcyanoacrylates with a longer alkyl chain. Illum et al. demonstrated the Langmuirian mechanism of adsorption of Rose Bengal, a model compound with environmentally sensitive light absorption spectrum, on poly(butyl 2cyanoacrylate) nanoparticles [43]. The nanoparticle surface adsorption approach was also shown to be applicable for binding ODN. Since ODN are negatively charged, their association with the hydrophobic PACA nanospheres was accomplished using ion pair complexation with quaternary ammonium salts, such as cetyltrimethylammonium bromide (CTAB). Increasing the carrier capacity and extending the release kinetics may potentially be achieved by adding the drug before or in the course of the polymerization due to drug entrapment in the particles. The encapsulation may be further enhanced by using auxiliary substances. Thus the encapsulation of progesterone used as a model lipophilic drug was shown to increase 50-fold when incorporated in a complex with cyclodextrins [44]. However, it should be noted that basic drugs may act as reaction initiators when added before the reaction has completed, resulting in covalent binding with the polymer [45]. Furthermore, the timing of the drug incorporation into the reaction medium is critical, with significant drug–polymer covalent binding or poor loading capacity resulting from the ‘‘too early’’ and ‘‘too late’’ drug addition, respectively [46]. 2. Drug Entrapment in Lipid or Aqueous Core PACA Nanocapsules The anionic polymerization process was adapted for production of lipid or aqueous core nanocapsules suitable for encapsulation of lipophilic and hydrophilic drugs, respectively. The form of
Biodegradable Nanoparticles as Drug Delivery Systems
397
the obtained nanocapsules is determined by the type of the emulsion (o/w vs. w/o) used in the formulation process. Lipid core nanocapsules are formed when nonvolatile oil is included along with the monomer and lipid-soluble drug in the dispersed phase of the o/w emulsion [47]. The nanocapsule formation mechanism is described in detail by Couvreur et al. [26]. Interestingly, nanocapsules are effectively obtained by this process only if water-miscible organic solvent, such as ethanol, is included in the dispersed phase. The spontaneous diffusion of ethanol to water causes formation of oily nanodroplets with monomers at the interface, which subsequently form the polymeric shell of a nanocapsule. Nanocapsules thus obtained may be separated from nanospheres forming as a byproduct by ultracentrifugation due to the difference in their density [48]. Recently a modification of the interfacial polymerization method suitable for production of aqueous core nanocapsules was reported [49,50]. In this method the emulsification of aqueous drug solution and ethanol mixture in an external organic phase is followed by incorporation of the monomer under stirring. After the reaction is complete nanocapsules are sedimented by ultracentrifugation from the oily phase in presence of water and subsequently washed and resuspended. This method was effectively applied for encapsulation of ODN. The entrapment of ODN in nanocapsules was shown to be superior to nanosphere surface-adsorption in terms of ODN in vitro stability: surface-adsorbed ODN were protected against degradation by an exonuclease of snake venom [51], but not in plasma or serum [49,52], whereas a significant fraction of nanoencapsulated ODN remained stable in fetal calf serum for as long as 60 min. 3. Approaches to Improve Formulation Stability and Biocompatibility Attempts have been made to develop nanoparticles avoiding acidic preparation conditions, potentially harmful to chemically labile drugs, and cytotoxicity associated with some of the PACA polymers. Nanoparticles formulated with methylidene malonate derivatives using the emulsification/polymerization method were investigated as an alternative to PACA [53,54]. The optimum reaction medium pH was found to be 5.5–6.0 and 7.6 for preparing nanoparticles composed of poly(diethylmethylidene malonate) (PDEMM) and poly(methylidene malonate 2.1.2) (PMM 2.1.2), respectively. While the PDEMM nanoparticles were shown to be practically not biodegradable both in vitro and in vivo [53], PMM 2.1.2 nanoparticles degraded by ester hydrolysis into glycolic acid and ethanol, noncytotoxic products. More than 80% of the polymeric acid remaining after the polymer degradation was excreted in urine and feces within 2 days following intravenous administration in rats [54]. The cytotoxicity of PACA nanoparticles may potentially result from in vivo polymeric overloading when the polymer degradation or metabolite elimination occurs at unfavorably slow rates, or due to the presence of toxic alkylcyanoacrylate monomers. The degradation and elimination rates may be controlled through adjusting the monomer chemical composition and the reaction conditions; however, the retention of the unreacted monomers in the formulation is still a concern, although their presence has not been revealed by studies of the PACA nanoparticle composition using gel permeation chromatography [41]. This potential drawback may be avoided when nanoparticles are formulated using a preformed polymer. For this purpose a novel PEGcyanoacrylate–hexadecylcyanoacrylate (PEG-PHDCA) copolymer was synthesized and evaluated [38]. The PEG-PHDCA-based nanoparticle formulation approach is advantageous in several additional respects: (1) simple and not time-consuming methods of nanoparticle preparation may be employed, (2) the drug may be incorporated into the polymeric matrix, which increases the carrier capacity, prevents the drug’s premature degradation, and extends its release, (3) use of stabilizers is minimized or avoided, since the dispersion of the nanoparticles is effectively stabilized due to the steric protection provided by the hydrophilic PEG chains, (4) spleen or brain
398
Chorny et al.
targeting may be achieved due to the unique biodistribution properties of the nanoparticulate drug carrier. The composition, cytotoxicity, biodegradability and biodistribution of PEG-PHDCA nanospheres prepared by polymer precipitation methods were reported in several studies [39,55,56]. The nanosphere size was shown to depend directly on the polymer amount, and smaller sized nanospheres were obtained using water-miscible organic solvents (THF, nanoprecipitation method) rather than solvents of low water solubility (dichloromethane/chloroform, emulsification/solvent evaporation method) [39]. The nanospheres showed low cytotoxicity and notable stability in circulation following intravenous injection in mice compared to non-PEGylated nanospheres. In addition, such nanospheres were able to escape the macrophages of the reticuloendothelial system (RES), but showed significant retention by splenic filtration. Recently Calvo et al. demonstrated a 2 to 2.5-fold increase in the brain distribution of PEG-PHDCA nanospheres after intravenous injection in scrapie-infected mice [57]. The increased brain penetration was not associated with modification of the blood brain barrier (BBB) permeability in rats in contrast with polysorbate 80 (TWEEN 80)–coated PHDCA nanospheres [58]. Finally, using acetone as organic solvent and Poloxamer 188 as a surface active agent in the external aqueous phase it was possible to prepare nanospheres as small as 110 nm, capable of extravasation across leaky tumoral endothelium and therefore potentially applicable for the treatment of solid tumors [55]. B. Polyesters and Polyanhydrides The biodegradable polyester family includes polylactide (PLA), polyglycolide (PGA), poly(εcaprolactone) (PCL), and the their copolymers, of which the most widely used is polylactideco-glycolide (PLGA), while the polyanhydrides developed to date are considerably more diverse [36]. The unique chemistry and biodegradation mechanism of biodegradable polyesters and polyanhydrides determine the in vitro and in vivo characteristic properties of injectable particulate formulations. In contrast to PACA the hydrolysis of these polymers leads to the breakdown of the polymeric chain and formation of the respective monomeric acids that may further be degraded intracellularly or excreted from the body. Low molecular weight of the polymers therefore is not a prerequisite for effective bioelimination (as is the case for PACA), although it is one of the most important determinants of the carrier degradation and drug release rate along with the crystallinity, hydrophobicity and composition of the polymer, particle size, and medium acidity [16,59–62]. It should be noted that formulation of submicronial particles made of polyanhydride for parenteral administration has not been reported, apparently due to the inability to provide sustained drug release with a polymer rapidly degrading in aqueous dispersion [59,63,64], although attempts have been made to formulate polyanhydride particles sufficiently small for intravenous injection [65]. However, the applicability of polyanhydride nanoparticles for parenteral administration may be revised in the future due to the rapidly increasing number of novel polyanhydrides differing in biodegradation kinetics [66]. On the other hand, the comparatively slow degradation rates of the aliphatic polyesters render polyester-based injectable nanoparticles useful for a wide variety of applications [67]. Interestingly, it was shown that an emulsification/solvent evaporation method (see Section V.C.1) might be applied to prepare double-walled particle dispersions using combinations of polyanhydrides and polyesters. The two-layered microparticles were obtained due to a phase separation in the course of the solvent elimination [68]. The difference in the decomposition rates of the polymers results in fast digestion of the particle inner core formed by a polyanhydride with the outer shell staying intact for several months [59,69]. Such a carrier degradation profile apparently might be advantageous in cases where biphasic release of the particle-entrapped drug is desirable. In addition, this technique
Biodegradable Nanoparticles as Drug Delivery Systems
399
might be helpful for effective coencapsulation of compounds possessing different affinities to these polymers [70]. C. Polymer Precipitation Methods The in situ polymerization in aqueous media is not generally applicable for preparation of polyanhydride or polyester-based particle dispersions. The most widely used formulation methods involve emulsification of the organic solution containing a preformed polymer in the external aqueous phase followed by polymer precipitation. These methods do not employ chemical reactions and extreme preparation conditions, which is an important advantage compared to emulsification/polymerization; however, the utilization of organic solvents necessitates their elimination from the formulation. The polymer precipitation methods are most conveniently classified according to the principle by which the solvents are displaced from the initially formed emulsion droplets. The precipitation of the polymer with subsequent nanoparticle formation may be effected by evaporation, extraction, or spontaneous diffusion of the organic solvents depending on their volatility and water miscibility. The formation mechanisms of nanoparticles formulated by these techniques were reviewed by Quintanar-Guerrero et al. [71]. 1. Emulsification/Solvent Evaporation Polymer precipitation may be accomplished by evaporation, when volatile organic solvents with relatively low water miscibility are used to form the dispersed polymer-containing phase. Chloroform, dichloromethane (DCM), and ethyl acetate (EA), commonly employed in emulsification/solvent evaporation, are excellent solvents for many polyanhydrides and polyesters. The polymeric solution is emulsified in an aqueous phase containing a stabilizer. The size of the resultant particles is determined by the precursory emulsion, the formation of which completely relies on external mechanical energy. Therefore in order to obtain formulations with particle size acceptable for parenteral administration route (below 5 m, and preferably below 1 m) the use of a high-energy source, such as microfluidizer, homogenizer, or ultrasonicator is essential. The polymeric dispersion is obtained by elimination of the organic solvent that may be facilitated by heat and/or vacuum. Alternatively, the extraction of the organic solvent with subsequent formation of nascent particles may be promoted by dilution in a large amount of water that may contain a stabilizer (usually in low concentration) [72–76] or a volatile amphiphilic solvent, such as isopropanol [77], prior to the final nanoparticle solidification by solvent evaporation. This procedure, often referred as solvent extraction/evaporation, is advantageous in two respects: the solvent diffusion rate determining the particle solidification and eventually the structure and morphology may be precisely controlled [72,78,79]; the dilution increases the distances between the nascent particles and thus prevents coalescence that may take place before their final hardening. Drug Encapsulation and Release. Lipid-soluble drugs may be incorporated in the organic phase along with the polymer. In some cases drugs that cannot be directly dissolved in the organic phase may be added in a suitable cosolvent [70,80]. The drug loadings for lipophilic drugs generally achievable by this method are high, ranging from about ten to several tens of percent [11,80–83]. The encapsulation efficiency of ionizable substances may be significantly improved by adjusting the pH of the external aqueous phase to minimize drug partitioning from the nascent particles during preparation [84,85]. However, this protocol is not directly applicable for encapsulation of water-soluble compounds. A modification, often termed double emulsion/ solvent evaporation, was adopted for this purpose. In this modification the water-soluble drug is first dissolved in water, optionally containing a w/o emulsion stabilizer, and the inner aqueous
400
Chorny et al.
phase is emulsified in the organic solution of the polymer. This w/o emulsion is further emulsified in an external aqueous phase to form double emulsion (w/o/w). The particle dispersion is produced by polymer precipitation after the evaporation of the organic solvent. Although originally applied for entrapping a hydrophilic drug in microspheres [86], this technique was shown to be suitable for production of protein-loaded nanoparticles as small as 200 nm [74,75,80,87] and DNA-loaded nanospheres with a size of 230 nm and even 70 nm [23,88]. Drug encapsulation yield and release rate appear to be significantly dependent on the rate of the organic solvent evaporation [75,89] as a result of the differences in polymeric matrix porosity and crystallinity [89] as well as the homogeneity of the drug distribution in the particle [78,79]. Liggins et al. [90] described a pronounced reduction of drug release rate from paclitaxelloaded microspheres with an increase in the polymer molecular weight. The higher molecular weight was shown to directly affect the degree of crystallinity of the polymer matrix, resulting in lower diffusion rates observed for the encapsulated drug. The rate and mechanism of the drug release from lidocaine-loaded nanospheres was shown to depend on their drug loading and size [91]. A higher drug loading led to a shift from diffusion-controlled to dissolution-controlled drug release with a more pronounced burst phase characteristic of the smaller-sized particles. Go¨rner et al. [82] further showed that the nanosphere size was a direct function of the polymer concentration and molecular weight. The authors also demonstrated an increase in lidocaine encapsulation efficiency with the nanosphere size. Particle Size and Morphology. The nanosphere size being largely determined by the energy source and input [92,93], was also shown to be dependent on the nature and amount of the stabilizer in the external aqueous phase. Considering the ability to produce smaller-sized nanospheres polyvinyl alcohol (PVA), (PVA) was shown to be a more effective stabilizer than human serum albumin [75] and Poloxamer 188 [94] and as effective as D-␣-tocopheryl PEG succinate [83]. Increasing the amount of PVA, Poloxamer 188, or albumin used as stabilizers resulted in nanosphere size reduction [75,87,94,95]. Zambaux et al. [75] further showed that a PVA concentration above 3% is necessary for the adequate emulsion stabilization, allowing for production of smaller sized, less polydisperse nanoparticles. Scholes et al. however reported that the nanosphere size miniaturization was reversed by increasing the PVA concentration above 8–10%, apparently due to a parallel increase in the external phase viscosity [96]. The water miscibility of the organic solvent is another important determinant of the particle size and morphology. Sah [72] studied the size and morphology of PLGA microspheres as a function of employed solvent and its extraction rate. The application of higher external aqueous phase volume resulted in quick leaching of EA from the nascent particles, which in turn led to formation of hollow microspheres with higher residual levels of the solvent. In contrast, compact, monolithic spheres were produced using lower external phase volume or, alternatively, employing DCM as an organic solvent for PLGA. The more rapid diffusion of EA to the external aqueous phase as a consequence of its higher water solubility resulted in nanoparticles significantly smaller than those obtained with DCM [97]. QuintanarGuerrero et al. suggested that the diffusion of partially water-miscible EA leads to formation of several solid nanospheres from each emulsion droplet, which obviously is not the case for DCM characterized by low diffusion rates in water [98]. Primary Emulsion Stability. Considering successful application of the double emulsion/solvent evaporation method, the stability of the primary w/o emulsion is a primary determinant of the drug encapsulation yield in the formulation as well as the particle size and morphology [92]. Whereas the effect of the primary emulsion stability on the properties of microparticles has been extensively studied [86,99,100], the relationship between the primary emulsion stability and the characteristics of nanoparticles prepared by this method has not been adequately addressed. Poloxamer 188 and bovine serum albumin (BSA) were shown to effec-
Biodegradable Nanoparticles as Drug Delivery Systems
401
tively stabilize w/o emulsion leading to formation of matrix type microspheres with a compact core, whereas multivesicular porous structure was characteristic of microparticles prepared with unstabilized inner emulsion [99]. Interestingly, the stabilizing effect of these surfactants notably deteriorated when used in combination, apparently because of their interfacial complexation. The addition of a stabilizer to the primary emulsion was also adopted for production of nanospheres. Lemoine and Pre´at [87], who applied the double emulsion/solvent evaporation method for encapsulation of hemagglutinin in PLGA-based nanoparticles, estimated the physical stability of the primary emulsion by visual observation of the phase separation. The authors report that the primary emulsion prepared with 2.8% w/v polyvinyl alcohol (PVA) in the internal aqueous phase remained stable for 4 h, whereas in the absence of the stabilizer the separation occurred in 10 min. Tabata and Langer showed that the size of polyanhydride-based microspheres was determined by the size of the inner emulsion droplets, depending on the mixing method used in its preparation [64]. Probe sonication, though effective for production of fine primary emulsion, was shown to adversely affect the activity of trypsin [63] and protein C [74] used as model proteins, although loss of protein activity may be prevented to a large extent by performing the sonication in an ice bath [75]. Considering use of double emulsion/solvent evaporation for encapsulation of DNA, Ando et al. suggested lowering the temperature below the freezing point of the aqueous inner phase, resulting in a solid particulate suspension prior to formation of the secondary emulsion [101]. This modification, termed cryoprotection, was shown to preserve the plasmid DNA in a supercoiled state with the highest level of bioactivity. Song et al. [80] used several protocol modifications in order to stabilize the primary emulsion and prevent the escape of a model protein from the nanoparticles: (1) the viscosity of the internal aqueous phase was increased by adding a high amount of BSA and cooling to 4⬚C; (2) Poloxamer 188 was incorporated into the organic phase; (3) acetone was admixed to the organic phase in order to initiate deposition of a polymeric wall around the primary emulsion droplets, thus preventing rapid protein escape to the external aqueous phase. These modifications when combined with use of a higher molecular weight polymer (PLGA) resulted in a severalfold increase in loading efficiency and more sustained release kinetics. We applied the double emulsion/solvent evaporation process to encapsulate plasmid DNA in PLGA nanoparticles [21]. Tris-EDTA buffer with plasmid was emulsified in an organic phase consisting of PLGA solution in chloroform. The primary emulsion was emulsified in an aqueous phase containing 2% PVA prepared with or without calcium chloride. The emulsification was performed on an ice bath using a homogenizer. The solvent was evaporated at 4⬚C under atmospheric pressure, and the obtained nanoparticles were washed and lyophilized. The nanoparticles were spherical in shape. Their size was not affected by calcium salt in the external aqueous phase and ranged from 538 to 644 nm with narrow size distribution. DNA loading in nanoparticles was found to be proportional to DNA formulation amount. The entrapment yield was significantly improved in the presence of calcium (28 Ⳳ 1 vs. 70 Ⳳ 17%), with a parallel increase in DNA loading (0.15 vs. 0.40% w/w). Intraparticulate localization of plasmid was confirmed by confocal fluorescence microscopy using labeling with ethidium bromide. Approximately 30% of encapsulated DNA was released in the first 4 h of release studies (Fig. 1). The DNA fraction remaining in the particles exhibited sustained release over more than 28 days. In a control experiment performed with blank nanoparticles incubated with free DNA the whole amount of plasmid was determined in release medium after 24 h, indicating that the sustained release observed with the loaded particles could not be due to DNA surface adsorption. The nanoparticle preparation procedure did not affect the integrity of DNA; however, the amount of plasmid in relaxed form having lower bioactivity increased about twofold. Nanoencapsulated DNA was effectively taken up by mouse fibroblasts in cell culture in comparison to only marginal cellular uptake of naked plasmid. The expression of DNA delivered in nanoparticles was demonstrated after 7 days in
402
Chorny et al.
vitro, in human endothelial cells and after 28 days in vivo, in rat tibialis muscle. Intramuscular injection of the nanoparticles in the latter case was not associated with local inflammation as evidenced by histological examination. Formulation Purification. The elimination of polyvinyl alcohol from the final formulation has to be considered. Following subcutaneous administration, PVA with molecular weight 120,000 Da exhibited notable cancerous reactions in rats [102]. A more recent study by Yamaoka et al. [103] demonstrated that both PVA and PEG, nonbiodegradable polymers (the latter approved by FDA for parenteral administration), exhibited little accumulation in tissues and were eliminated by urinary excretion in a similar manner with molecule size–dependent kinetics after intravenous injection in mice. Further it was shown that almost 80% of PVA with a molecular weight of 15,000 Da was excreted in urine within the first 30 min. However, the biocompatibility of PVA and its approval status for parenteral use have not been revised, and its presence in the formulation is considered unacceptable for safety reasons. Methods evaluated for PVA elimination from nanoparticulate formulations include ultrafiltration [70], multiple washing by sedimentation and resuspension in fresh medium [75,93,96,104,105], cross-flow filtration [106,107], and gel permeation chromatography [105]. All studies demonstrated significant retention of PVA in the nanoparticles following purification. The residual level of PVA was shown to directly depend on its initial concentration in the external aqueous phase [75,104] and vary to some extent with the molecular weight of the particle forming polymer [75]. The elimination of PVA was also shown to depend on the number of washing steps [75] or, alternatively, the volume of water used for purification by cross-flow filtration [107]. The residual level of PVA, 13,000–23,000, used at a concentration optimal for formation of the smallest sized nanoparticles (4%) was found to be as high as ⬃9% w/w [75] following nanoparticle triple washing. PVA
Figure 1 Cumulative release (percent of amount loaded) of DNA from DNA-loaded nanoparticles. Release was performed in Tris-EDTA buffer, pH 7.4, 37⬚C. (From Ref. 21.)
Biodegradable Nanoparticles as Drug Delivery Systems
403
retention in the formulation was also shown to be affected by the water miscibility of organic solvent in the dispersed phase [104], being highest for acetone and lowest for chloroform. Interestingly, a notably lower level of the residual PVA was demonstrated for PLA-PEG nanoparticles compared to PLA nanoparticles of similar diameter (0.6 and 8.8%, respectively) following purification by washing, apparently because the adsorption of PVA was hindered by the PEG chains protruding from the nanoparticle surface. Nanoparticle surface associated PVA appears to be composed of two fractions: washable (adsorbed) and washing-resistant (anchored). The presence of PVA irreversibly bound to the nanoparticle interface was shown to result in better formulation stability after purification [97,108] and nanoparticle size preservation after freezedrying [94,107], and also might potentially be utilized for postgrafting of targeting moieties [105]. However, since PVA is not acceptable for parenteral use, the stabilizing efficacy of other natural and synthetic biocompatible emulsifiers that might provide an alternative to PVA was evaluated. Successful examples include both ionic and nonionic surfactants, such as human serum albumin [74,95,109–111], phospholipids [81,112,113], Poloxamer 188 [94], polysorbate 80 [97], sodium cholate [11,73], and recently D-␣-tocopheryl PEG succinate [83]. The latter two emulsifiers exhibited good stabilizing properties and were also shown to be readily removed by washing from the nanosphere surface [73,83]. The amounts of residual solvents in the formulation are also subject to limitations because of their toxicity and potential carcinogenic effects. The guideline for residual solvents of the International Conference on Harmonization (ICH) states concentration limits of 60, 600, and 5000 ppm for chloroform, dichloromethane, and ethyl acetate, respectively. The volatility of these solvents allows for their comparatively easy elimination. However, it must be ascertained prior to formulation up-scaling that their levels do not exceed the limitations. Unfortunately, the retention of the organic solvents in nanoparticles prepared by emulsification/solvent evaporation has not been systematically studied. Characterizing the residual levels of volatile solvents in microspheres as a function of preparation method, Bitz and Doelker [114] showed that DCM was relatively easy to remove from PLA microspheres prepared by double emulsion/solvent evaporation. Although microspheres formulated with D,L-PLA revealed somewhat higher residual levels compared to L-PLA, the requirements stated for residual DCM in the formulation were met in both cases. The lower volatility of EA is obviously responsible for its significant residual levels (2.6% w/w) in vacuum-dried PLGA microspheres, as reported by Sah [72]. In another study [70] dimethyl sulfoxide (DMSO), the amount of which is also limited in formulations for parenteral use, was utilized as a cosolvent for amphotericin B encapsulated in PLGA nanoparticles. While DMSO could not be eliminated by evaporation at ambient temperature due to its low volatility, it was shown to be readily removed by ultrafiltration. 2. Salting-Out and Emulsification/Solvent Diffusion The extraction of organic solvent from emulsion nanodroplets may be achieved by evaporation (in cases where a volatile solvent of low water miscibility is employed) or, alternatively, by partitioning toward the continuous aqueous phase promoted by dilution. The accomplishment of polymer precipitation in the latter case does not pose the requirement of high volatility for the solvent to be eliminated; the selection of the solvent is instead based on its water miscibility properties. The principle of solvent extraction upon aqueous phase dilution is utilized in saltingout and emulsification/solvent diffusion methods. Whereas the organic phase of o/w emulsion in the emulsification/solvent diffusion is composed of partially water-miscible solvents, such as ethyl acetate, benzyl alcohol, propylene carbonate, 2-butanone, etc., the salting-out process makes use of reversible phase separation of a solvent completely water miscible under normal conditions, such as acetone, tetrahydrofuran, or isopropyl alcohol [115]. Use of organic solvents
404
Chorny et al.
with low toxic potential that are readily eliminated by cross-flow filtration is an important advantage of these methods. The dispersion of the organic phase with subsequent formation of submicronial particles is achieved due to the combined effect of intensive mechanical stirring and energy released in the process of organic solvent redistribution into its new equilibrium state following the external phase dilution [71]. In this sense these techniques are similar to the extraction/evaporation procedure mentioned above, the main difference being the method most commonly employed for elimination of the organic solvent from the formulation (filtration vs. evaporation). In fact, Quintanar-Guerrero et al. [98] recently demonstrated that controlled evaporation of partially water-miscible EA or 2-butanone might effectively be utilized to induce formation of submicronial particles by promoting the solvent diffusion from emulsion droplets. Such a modification of the original emulsification/diffusion technique makes it possible to avoid excessive dilution of the formulation in the preparation process. Nanosphere Formulation by a Salting-Out Process. Although different polymers, solvents, and salting-out agents were evaluated for the salting-out process [115,116], the nanoparticles prepared from D,L-PLA using acetone and magnesium chloride or magnesium acetate as salting-out agents have been most extensively characterized. Practically, an aqueous gel containing the magnesium salt and PVA used as a stabilizer is added on intensive stirring to the acetonic solution containing the polymer and the drug. The type of the liquid–liquid two-phase system changes from w/o to o/w emulsion in the course of the aqueous phase incorporation. The complete diffusion of acetone leading to polymer precipitation is promoted by adding pure water. The salt, acetone, and PVA are subsequently removed from the obtained nanoparticle dispersion by cross-flow filtration. Smaller sized nanoparticles were produced by using low molecular weight, partially hydrolyzed PVA in concentration above 10% [106,107,115], as well as by increasing polymer amount, stirring rate, and external/internal phase ratio [107,115,117]. The drug encapsulation yield may be improved by choosing a suitable salting-out agent. Magnesium chloride was shown to provide better entrapment for a hydrophobic peptidomimetic compound, CGP 57813 [4], an inhibitor of HIV-1 protease, whereas magnesium acetate producing higher pH in the aqueous phase was more suitable for encapsulation of savoxepine [106]. The method was shown to be applicable for production of PEG-coated nanoparticles with modified biodistribution and reduced in vitro uptake by human monocytes in plasma [12,118,119]. Diluting the o/w emulsion with PEG aqueous solution instead of pure water resulted in comparatively low levels of PEG associated with the nanoparticles [12]. However, a denser coating might be obtained by using PLA-PEG admixed to PLA in the nanoparticle preparation process. An increase in the nanoparticle size due to the addition of PLA-PEG was compensated by elevating the stirring rate during the phase mixing from 1600 to 1800 rpm [118]. When freeze-dried without a lyoprotectant, coated formulations with high PEG content exhibited a significant degree of aggregation that could be reduced in the presence of trehalose [118,120]. Interestingly, the particle size remained practically unchanged after lyophilization (⬃300 nm) in formulations prepared with plain PLA. This protective effect could apparently be attributed to the residual PVA present at the nanoparticle surface, as shown by Konan et al. [107]. Emulsification/Solvent Diffusion—Formulation of Nanospheres and Nanocapsules. High entrapment yields achievable by the salting-out process appear to be due in part to the increased drug partitioning in favor of the organic phase of the o/w emulsion because of the salting-out effect of the magnesium salts. The incompatibility of the salts with bioactive compounds states, however, considerable limits the method’s applicability [121]. The emulsification/solvent diffusion method is advantageous in avoiding the use of concentrated salts and in allowing for preparation of stable nanodispersions utilizing either PVA or Poloxamer
Biodegradable Nanoparticles as Drug Delivery Systems
405
188, a biocompatible emulsifier suitable for parenteral administration [108,122]. In the latter technique polymer is dissolved along with drug in a partially water-miscible solvent previously saturated with water, and this organic phase is emulsified upon intensive stirring in aqueous solution of a stabilizer (typically in 1:2 volume ratio). Depending on the volatility of the organic solvent employed, nanoparticle formation is achieved either by dilution with pure water or by controlled solvent evaporation. In contrast to the salting-out method, higher internal phase volume and lower polymer formulation amount were associated with reduction in the size of PLA nanoparticles prepared by emulsification/solvent diffusion [122]. The molecular weight of PVA had no effect on the nanoparticle size [122]. The smallest sized PLA nanoparticles were obtained above 5% w/v of PVA or Poloxamer 188 in the aqueous phase. The emulsification/solvent diffusion process was recently applied for encapsulation of ODN in PLA nanoparticles [121,123]. Negatively charged ODN are not soluble in organic solvents; therefore a neutral complex of ODN with CTAB was used for nanoparticle preparation. The complex was dissolved together with PLA in benzyl alcohol, and the nanoparticles were formed by slow dilution with water using PVA as a stabilizer [121]. The nanoparticles had an average size of 400 nm. Both ODN and polymer were shown to remain stable in the nanoparticle preparation process. The entrapment efficiency inversely depended on the ODN formulation amount ranging from 56 to 37% in formulations with theoretical loading of 2.5–10%. A size increase significantly enhanced in nanoparticles with higher ODN loading was observed during nanoparticle washing and after lyophilization, and was attributed to reversible nanoparticle aggregation. This aggregation is apparently due to a lower residual level of PVA displaced from the nanoparticles by a surface-associated fraction of the ODN-CTAB. Delie et al. [124] compared ODN nanoencapsulation by emulsification/solvent diffusion and double emulsion/solvent evaporation with DCM as a solvent. The ODN was encapsulated in complex with CTAB and in a free form by the former and the latter method, respectively. Although these methods were equally effective in terms of ODN entrapment yield and produced nanoparticles of similar size, the ODN release from the nanoparticles prepared by emulsification/solvent diffusion was complete in 24 h, whereas less than 40% was released after 8.5 days from the nanoparticles prepared by double emulsion/solvent evaporation. The extremely rapid release of the ODN encapsulated by the solvent diffusion process was explained by ODN adsorption to the nanoparticle surface rather than association with the PLA matrix, as appears to be the case for the nanoparticles prepared by solvent evaporation [124]. Inclusion of nonvolatile oil (mixed triglycerides or mineral oil) in the organic phase along with polymeric solution in EA resulted in formation of nanocapsules [125] with a density intermediate between those of nanoemulsion and nanospheres. High entrapment yields of lipophilic compounds were demonstrated for nanocapsules; however, these were not reported in comparison to nanospheres [125]. The authors also noticed a considerable fragility of the nanocapsules that might have an effect on their drug release properties and stability. Preparation of drug nanosuspensions is another interesting application of this method [126]. In this case the organic phase containing hydrophobic drug and partially water-miscible solvent is prepared without addition of polymer, and suspension of drug nanocrystals is formed from o/w emulsion in the presence of a stabilizer upon dilution with water. The nanocrystals of mitotane used as a model drug in this study were somewhat smaller than the emulsion droplets, which might be due to the generation of new globules in the solvent extraction process. Use of triacetin as a partially water-miscible solvent resulted in twofold larger nanocrystals and slower dissolution compared to formulations prepared with butyl lactate, the size and dissolution profile of nanocrystals prepared with benzyl alcohol being intermediate. All formulations exhibited slow crystal growth upon storage [126].
406
Chorny et al.
3. Nanoprecipitation Spontaneous formation of nanoparticles without external energy input may be achieved by using organic solvents completely water miscible under preparation conditions. Rapid diffusion of the solvent toward the aqueous phase leads to intensive spreading of the organic polymer solution and formation of emulsion droplets of submicronial size followed by polymer precipitation in the form of nanodispersion [71]. Since the emulsification and subsequent polymer precipitation in this process, termed nanoprecipitation or solvent displacement, are spontaneous and driven entirely by solvent diffusion, the physicochemical properties of the obtained nanoparticles are determined solely by phase composition and conditions immediately after the phase mixing [127]. Submicronial particle size with low polydispersity achievable in a highly reproducible fashion without external energy source is the main advantage of nanoprecipitation over the previously discussed methods. With appropriate modifications nanoprecipitation is suitable for the production of nanospheres and nanocapsules as well as nanosuspensions of hydrophobic drugs [128]; however, similarly to the emulsification/solvent diffusion and salting-out procedures, no established modification of this method exists for production of hydrophilic drugloaded nanoparticles, although attempts to encapsulate water-soluble substances have been made [129–132]. Nanoprecipitation may in principle be applied to a large variety of biodegradable and nonbiodegradable polymers; however, it was used most extensively for preparation of nanoparticles made of biodegradable polyesters [129–135] and more recently poly(alkylcyanoacrylates) [39,55,56]. Acetone, acetonitrile (MeCN), and tetrahydrofuran (THF) are the most commonly employed organic solvents. These solvents are water soluble, have a low toxic potential, and are readily eliminated by evaporation. PVA and Poloxamer 188, which is acceptable in formulations for parenteral administration and therefore does not require special purification steps for its elimination, are used as stabilizers in the nanoprecipitation method, although method modifications avoiding use of surfactant have also been reported [39,56,131,132,136]. Although sterilization of nanoparticles prepared by this process has not been systematically studied, protocol modifications have been reported suitable for production of sub-200-nm and even ultrasmall (⬍100 nm) nanoparticles [39,127,133,134,137,138] that can be sterilized by filtration without adversely affecting the formulation properties. Particle Size, Recovery Yield, and Stability. The effect of nanoprecipitation formulation variables on the nanoparticle size has been addressed in several studies [127,133,138–141]. Using central composite design Chaco´n et al. [133] evaluated the carrier size and encapsulation of Cyclosporine A used as a model hydrophobic drug in PLGA nanoparticles. The nanoparticles were formed by injection of polymer solution in acetone to an aqueous phase containing 0.6% PVA as a stabilizer. The injection rate was controlled by combining different needle diameters and pressures. Nanoparticles as small as 50 nm with a homogeneous size distribution could be produced by using lower polymer amount, higher injection rate, and smaller needle gauge selected as independent variables. The drug encapsulation yield ranging from 48 to 84% strongly depended on the polymer amount, but was not affected significantly by the other variables [133]. In another study by this group [141] a similar effect of these formulation variables on the properties of PCL nanoparticles stabilized with Poloxamer 188 was observed. In addition, smaller nanoparticles were produced by increasing the acetone volume that was identified as another important size-controlling variable. Wehrle et al. [127] studied the size of D,L-PLA nanoparticles as a function of formulation variables by means of factorial experimental design. The PLA nanoparticles were prepared by a modified protocol, where DCM, which has low water miscibility but is a good solvent to the polymer, was incorporated in the organic phase. The authors observed a considerable inverse effect of the acetone amount on the nanoparticle size in presence of DCM and practically no effect in its absence. This was explained by different processes
Biodegradable Nanoparticles as Drug Delivery Systems
407
determining the nanoparticle size: the diffusion of acetone resulting in formation of DCM emulsion and the rapid precipitation of PLA were the key factors in the first and the second cases, respectively. Poloxamer 188 did not significantly affect the size of PCL nanoparticles in the studied amount range [141]. Interestingly, Murakami et al. [142] reported an increase in PLGA nanoparticle size with PVA amount. Use of a highly hydrolyzed PVA grade resulted in smaller sized nanoparticles and in compromised recovery due to particle aggregation in the preparation process. The recovery yield of the nanoparticles was improved by replacing DCM with ethanol or methanol, which are nonsolvents to PLGA [143]. The increased recovery was apparently due to more rapid solidification of the nanoparticles achieved by this modification [143,144]. Stolnik et al. [136] prepared stable nanodispersions of poly(-malic acid-co-benzyl malate) (PLMABeH) and PLGA without use of a stabilizer by slowly adding a mixture of water and ethanol to the acetonic polymer solution. The PLGA nanoparticles were stabilized in aqueous dispersion electrostatically due to the presence of ionized carboxyl groups at the nanoparticle surface, as evidenced by intensive aggregation of nanoparticles observed at pH below 4 [136]. Chaco´n et al. [137] compared the stability of unloaded and cyclosporine A–loaded PLGA nanoparticles stored under various conditions and examined the possibility of the nanoparticles freeze-drying with or without a cryoprotectant. The authors observed aggregation of PLGA nanoparticles with initial size below 80 nm upon storage for 6 months in aqueous dispersion at room temperature. Both the mean size and the size distribution profile remained unchanged in larger-sized nanoparticles or, alternatively, in all formulations stored at 8⬚C. The initial drug encapsulation yields varied from 48 to 84% with the total surface area of the nanoparticles. Interestingly, a significant increment in drug loading was demonstrated in formulations stored at room temperature as a function of their total surface area. Formulations stored at 8⬚C exhibited a similar though less pronounced increase in their drug content. This was explained by higher drug solubility at low temperature associated with lower extent of adsorption to the nanoparticle surface. The authors report a significant polymer degradation, increasing with the storage temperature and decreasing with the nanoparticle size. In addition, the presence of drug enhanced the polymer degradation in the formulations stored at room temperature. Nanoparticles lyophilization without use of a cryoprotectant led to a 7- to 10-fold increase in the mean particle size. More satisfactory results were obtained in formulations freeze-dried with 5% glucose or trehalose, although those also revealed an increase in the particle size and the number of size populations [137]. Drug Encapsulation and Release. Some examples of drug-loaded nanoparticulate formulations prepared by nanoprecipitation are given in Table 1. Barichello et al. [129] examined the entrapment efficacy of several drugs varying in their hydrophobicity in PLGA nanoparticles. Highest entrapment yields were observed for cyclosporine A (84%) and indomethacin (94%), corresponding to approximately 3% w/w drug loading. High entrapment yields for these drugs in PLGA nanoparticles have also been reported by other authors [137,140,145,147]; however, it should be noted that these correspond to drug loadings ranging from 2.5 to 5.8% w/w that are inferior to those usually achievable by other polymer precipitation methods. Furthermore, it was shown that these drugs are localized predominantly at nanoparticle surfaces, which is in agreement with rapid indomethacin release characterized by a pronounced burst phase [140,147,148]. Guzman et al. [139] studied encapsulation of cyclosporine A in PCL by nanoprecipitation in comparison to poly(isobutyl-2-cyanoacrylate) nanoparticles formed by in situ polymerization. For both types of nanoparticles the drug loading depended directly on the drug formulation amount, with maximum drug loading being higher in PCL-based formulations (4.6 vs. 2.3% w/w). Brigger et al. [55] entrapped tamoxifen in nanoparticles made of preformed PEGylated poly(alkylcyanoacrylate), PEG-PHDCA, with encapsulation efficacy of 80Ⳳ10%. However, despite the good entrapment yields the maximum loading achievable by the method
b
a
Poloxamer 188 PVA PVA PVA Poloxamer 188 Poloxamer 188/PVA Poloxamer 188 Poloxamer 188 Poloxamer 188 Poloxamer 188 Poloxamer 188 Poloxamer 188 Poloxamer 188 Poloxamer 188 Poloxamer 188 — — Poloxamer 188
Acetone Chloroform: acetone DCM:acetone:methanol DCM:acetone DCM:acetone Acetone Acetone Acetone Acetone Acetone Acetone Acetone Acetone Acetone Acetone Acetonitrile Acetonitrile Acetone
Presented as mean ⫾ size distribution. Calculated approximately from data provided in the article.
Stabilizer
Organic phase
Drug-Loaded Nanospheres Prepared by Nanoprecipitation
PCL PLGA PLGA PLGA PLA PLGA PCL PLGA PLA PLGA PLGA PLGA PLGA PLGA PLGA PLGA PLA-PEG PEG-PHDCA
Polymer
Table 1
Size (nm) 95 ⫾ 25a 338–637 199–283 233–311 170 ⫾ 50a 46–146 198 126 119 187 157 166 168 169 167 124–210 28–175 110
Entrapment yield (%) 85 ⫾ 5 14.5–50.0 0.6–15.0 5.0–11.8 92.9 53.8–92.0 87.4 76.2–97.4 74.2 12.1 ⫾ 1.3 9.4 ⫾ 1.3 5.6 ⫾ 1.0 93.9 ⫾ 1.3 83.7 ⫾ 3.4 46.2 ⫾ 1.2 14.5 ⫾ 62.0 6.6 ⫾ 10.7 80 ⫾ 10
Drug loading (% w/w) 4.4 2.5–5.9 0.08–2.65 0.15–2.70 5.94 0.48–2.3b 1.4b 1.2–1.5b 1.2b 0.39 ⫾ 0.04b 0.30 ⫾ 0.04b 0.18 ⫾ 0.03b 3.13 ⫾ 0.04b 2.79 ⫾ 0.11b 1.54 ⫾ 0.04b 0.2–3.2 0.19–0.27 0.35–0.46
Drug Cyclosporine A Indomethacin 5-Fluorouracil Nafarelin acetate Indomethacin Cyclosporine A Isradipine Isradipine Isradipine Vancomycin Phenobarbital Valproic acid Indomethacin Cyclosporine A Ketoprofen Procaine HCl Procaine HCl Tamoxifen
139 145 145 146 140 133 135 135 135 129 129 129 129 129 129 130 132 55
Ref.
408 Chorny et al.
Biodegradable Nanoparticles as Drug Delivery Systems
409
modification applied in this study was only 0.46% w/w because of limited drug solubility in the organic phase. More than 60% of the drug was apparently adsorbed at the nanoparticle surface as evidenced by significant burst effect in the drug release. Low and intermediate entrapment yields were obtained for relatively hydrophilic drugs, vancomycin and phenobarbital, and amphiphilic ketoprofen, respectively [129]. The entrapment of valproic acid was surprisingly low (5.6%) [129], apparently because the escape of the ionizable drug to the aqueous phase occurred faster than the nanoparticle solidification. The entrapment efficacy of another ionizable drug, procaine hydrochloride, in PLGA nanoparticles was improved five-fold (corresponding to a drug loading change from 0.3 to 1.3% w/w) by increasing the pH of the aqueous phase without significantly affecting the nanoparticle morphology and size [130,132]. Although replacing the hydrochloride salt with procaine base also resulted in higher drug entrapment, this modification was associated with a size reduction from 160 to 20 nm and loss of nanoparticle sphericity [130]. Surprisingly, neither of these modifications resulted in improved drug loadings in nanoparticles made of PLA-PEG [132]. This is apparently because drug adsorption playing a significant role in drug carrier association was prevented by the PEG chains protruding from the nanoparticle surface. Kawashima et al. [144] studied entrapment of peptides, elcatonin, and thyrotropin releasing hormone (TRH). The peptides were loaded by nanoprecipitation with either aqueous or oily continuous phase, whereas the dispersed phase consisted of PLGA solution in acetone and methanol. The partitioning of the peptides into the aqueous medium was apparently responsible for entrapment yields inferior to those observed with the latter modification. On the other hand, Barichello et al. [129] demonstrated that by optimizing the composition of the aqueous medium it was possible to achieve significant loading of insulin in PLGA nanoparticles prepared by nanoprecipitation. However, the authors showed that a significant fraction of the peptide was surface adsorbed, and only 20% was entrapped in the nanoparticles. Recently modifications of the nanoprecipitation method have been applied for nanoencapsulation of plasmid DNA [149,150]. Hirosue et al. [149] encapsulated DNA in a complex with dimethyldioctadecylammonium bromide (DDAB). The dispersed phase was prepared by dissolving PLGA and DDAB in trifluoroethanol (TFE) and incorporating DNA dissolved in a small amount of water. Phase inversion followed by immediate polymer precipitation was achieved by rapidly pouring ethanol into this mixture. The nanoparticle suspension was diluted with water, and the solvents were eliminated by evaporation. The nanoparticles were separated from nonencapsulated DNA-DDAB complexes by density-based separation using sucrose gradient. DNA remained intact in the formulation process, although its bioactivity was significantly lower than that of the original plasmid. The nanoparticles were shown to be ⬃100 nm in diameter. The entrapment of DNA was affected by the initial plasmid amount, DNA/DDAB ratio, and the polymer composition and molecular weight, ranging from about 15 to 50%; however, the plasmid loading was significantly improved only by use of higher molecular weight PLGA [149]. Perez et al. [150] reported another procedure, where aqueous solution of DNA (with or without viscosing agents) was emulsified in DCM containing PLA-PEG. Nanoemulsion formed by brief sonication was poured in ethanol to induce immediate nanoparticle formation, and the obtained nanodispersion was diluted with water. The organic solvents were removed by evaporation. The authors also prepared DNA-loaded PLA-PEG nanoparticles by double emulsion/solvent evaporation for comparison. Sizes of 130 and 280 nm, and encapsulation efficiencies of 80–90 and 60–80% were demonstrated for nanoparticles prepared by nanoprecipitation and double emulsion/solvent evaporation, respectively; whereas PLA nanoparticles prepared by the latter method exhibited an encapsulation yield of only 25%. Complete conversion of supercoiled DNA to open circular and linear forms was observed, apparently due to exposure to sonication. The release of nanoprecipitation encapsulated DNA was extremely fast (80–90%
410
Chorny et al.
in 15 min) compared to protracted release over 28 days from nanoparticles prepared by emulsification/solvent evaporation. Recently we reported on preparation and characterization of PLA nanoparticles loaded with an antirestenotic compond, tyrphostin AG-1295 [19,138,151,152]. The nanoparticles were formulated by nanoprecipitation with some modifications. Three milligrams of tyrphostin AG1295, a highly hydrophobic compound with extremely low water solubility (⬍0.5 mg/L), was dissolved along with 200 mg D,L-PLA in an organic solution containing DCM in a low volume ratio to acetone (1:19–1:39). In some formulations ethanol was incorporated in the organic phase as a volatile water-miscible nonsolvent to the polymer. Poloxamer 188 was employed as a stabilizer. Nanoparticle formation was induced by mixing the organic and the aqueous phase (20 and 40 mL, respectively) on moderate stirring. The organic solvents were eliminated by controlled vacuum evaporation, and the formulation was concentrated to a final volume of 10 mL. The obtained nanoparticle dispersion was passed through a 0.45-m syringe filter to remove aggregates. The nanoparticles were spherical in shape with homogeneous size distribution. Their size was shown to depend on several formulation variables, of which the effect of PLA and PLA nonsolvent (ethanol) formulation amounts was the most significant. PLA amount reduction from 300 to 100 mg resulted in a nanoparticle size decrease from 143 to 70 nm accompanied by a smaller drug encapsulation yield (96 vs. 56%). Nanoparticle size exhibited inverse linear dependence on the ethanol amount in the studied range (0–4 mL). The inclusion of ethanol in the organic phase allowed for production of ultrasmall nanoparticles (70 nm), without a decrease in the drug encapsulation yield (⬃72%). The drug loss during filtration was studied as a function of nanoparticle size. No reduction in the drug amount was observed in filtered 65 nm-sized particles. However, an increasingly significant drug loss was revealed by larger sized nanoparticles with only 58% of encapsulated drug recovered following filtration in nanoparticles with a mean size 143 nm. The reduced drug recovery was associated with a considerable retention of the larger sized particles by the filter. The drug release rate was examined under perfect sink conditions by an in situ method based on environmental sensitivity of the drug emission spectrum, and by an external sink method [151], where the drug released from nanoparticles is continuously extracted and subsequently assayed in heptane forming a separate liquid phase above the aqueous nanodispersion. A higher drug release was observed in smaller sized nanoparticles (Fig. 2), obtained either by reducing the polymer formulation amount or by incorporation of ethanol in the organic phase. An extensive nanoparticle washing with water led to a drug content reduction of 8.5%, attributed to elimination of a loosely bound drug fraction. This hypothesis was supported by similarity of the drug release behavior of the washable drug fraction and drug adsorbed on blank nanoparticles. In both cases the observed drug release was considerably faster in comparison to encapsulated drug, suggesting that the drug encapsulation process resulted in formation of two particle-associated drug populations differing in their release properties and respective amounts: a notably larger drug fraction appears to be entrapped in the nanoparticle/polymer matrix, whereas a smaller fraction is localized on the nanoparticle surface. Analysis of drug release kinetics using models developed for a matrix-type sphere indicated that the release behavior of the encapsulated drug is governed by Fickian diffusion (Fig. 3). The release pattern of the drug changed in the course of the experiment. This change was explained by presence of two compartments in the polymer matrix differing in their permeability to the drug molecules. The rapid polymer precipitation at the nanodroplet interface induced by intensive acetone diffusion is apparently responsible for formation of a more permeable polymeric shell, whereas the nanoparticle core characterized by lower permeability to the drug is formed by slow polymer desolvation in the process of elimination of DCM-rich residual solvent. Considerably different elimination patterns were exhibited by AG-1295 formulated in smaller and larger sized nanoparticles (90 and 160 nm) following intraluminal delivery in bal-
Biodegradable Nanoparticles as Drug Delivery Systems
411
Figure 2 Nanosphere size and release profiles of AG-1295 from PLA nanospheres as a function of formulation variables: (left column) PLA amount; (right column) ethanol volume in organic phase. The hydrodynamic nanosphere size was obtained by photon correlation spectroscopy (A, B). The drug release was determined by the external sink (C, D) and in situ methods (E, F). (From Refs. 138 and 151.)
412
Chorny et al.
Figure 3 The in situ release data of AG-1295–loaded nanospheres (170 nm) fitted by the diffusional release model Mt/Mⴥ⳱ktn. A straight line is predicted for diffusion-controlled release presented as log (fractional release) vs. log (time). The diffusional exponent derived from the graph slope (n⳱0.46Ⳳ0.03) is in agreement with a Fickian diffusion controlled release from a homogeneous population of matrix-type spheres. (From Ref. 151.)
loon-injured rat carotid arteries (Fig. 4) [152]. The rates of the faster and slower drug elimination phases of the local elimination profile attributed to nanoparticle wash-out and drug release, respectively, were higher for the 90-nm nanoparticles. On the other hand, the smaller sized nanoparticles revealed superior tissue uptake and residence properties. Their administration resulted in a three-fold larger drug amount (1165 vs. 380 ng/mg) retrieved from the tissue at the earliest studied time point (5 min), as well as a more than eight-fold higher amount in nanoparticles remaining associated with the artery following the fast nanoparticle wash-out phase (31.5 vs. 3.8 ng/mg, unpublished data) as compared to 160-nm nanoparticles. In fact, the smaller sized nanoparticles showed advantage in terms of effective local drug levels for as long as 14 days following administration. In comparison to encapsulated drug, AG-1295 adsorbed on the surface of 160-nm-sized blank nanoparticles was not detectable in the tissue 24 h following delivery. The antirestenotic efficacy of these formulations was in accord with their drug elimination profiles: AG-1295 encapsulated in 90 nm nanoparticles and adsorbed on 160 nm nanoparticles exhibited the highest and the lowest therapeutic effects, respectively. Formulation of Nanocapsules by Nanoprecipitation. The nanoprecipitation procedure can be adapted to yield nanocapsules by incorporating non-volatile oil in the organic phase [153,154]. In this case the diffusion of an amphiphilic solvent induces polymer precipitation at the interface of two immiscible liquids, oil and water, that are nonsolvents to polymer. Therefore
Biodegradable Nanoparticles as Drug Delivery Systems
413
Figure 4 Local elimination kinetics of AG-1295 formulated in 160- and 90-nm PLA nanospheres (530 g drug/mL) delivered intraluminally to balloon-injured rat carotid arteries. (Abstracted in part from Ref. 152.)
this modification of the nanoprecipitation process is often termed interfacial polymer deposition. Similarly to nanospheres, nanocapsules may be formed with different biodegradable and nonbiodegradable polymers. Several lipophilic and amphiphilic drugs have been entrapped in nanocapsules with high efficiency by Fessi et al. [153] including paclitaxel, dexamethasone, progesterone and indomethacine. Use of nanocapsules is especially advantageous for encapsulation of liquid oily substances, such as vitamin K. Such substances can replace the oil in the standard procedure, thereby increasing their loading in the formulation to tens of percent [153]. The effect of formulation variables on nanocapsule size and morphology was extensively studied by Ammoury [154] and Mosqueira [155]. The amount of PLA did not alter the nanocapsule properties and physical stability in the studied amount range [155]. The nanocapsule size was shown to be dependent on the oil formulation amount as well as the oil–water interfacial tension, whereas the effect of oil viscosity was less pronounced. Miglyol 810 and ethyl oleate resulted in the smallest sized nanocapsules (200 and 170 nm, respectively), apparently because of their reduced interfacial tension. Nanocapsules could be stabilized by either Poloxamer 188 or lecithin. Higher lecithin amount in the formulation corresponding to 0.3% of soy lecithin in the organic phase under given conditions was associated with a nanocapsule size decrease from 260 to 180 nm. Interestingly, the nanocapsule size showed no clear correlation with the amount of Poloxamer 188. Transmission electron microscopic examination identified presence of nanospheres and multilamellar liposomes, which is in accord with findings by Ammoury et al. [154]. The presence of liposomes was notably increased with higher amounts of lecithin, although
414
Chorny et al.
being significant even at relatively low lecithin concentrations in the organic phase. The liposome contamination was reduced when lecithin was replaced with other surfactants; however, use of other surfactants resulted in compromised stability of the formulations [155]. Ammoury et al. [154] reported that maximal stability of nanocapsule dispersion required use of Poloxamer 188 and phospholipids in combination, whereas presence of at least one of them was needed to provide satisfactory stability. The stability of PCL nanocapsules coated with lecithin and Poloxamer 188 in aqueous suspension was examined by Calvo et al. [147]. The size of nanocapsules stored for 6 months increased from 200 to 300 nm. The increase in the mean size was paralleled by a moderate increase in the size distribution. In contrast, nanoemulsion prepared without polymer under the same conditions exhibited a size increase from 200 nm to more than 600 nm during the same storage period. The release behavior of nanocapsules prepared by interfacial deposition was studied by several authors [147,154,156] using indomethacine and clofibride as model drugs. The release of clofibride was complete in about 20 min under perfect sink conditions. No difference in the drug release rate was observed between nanocapsules and nanoemulsion [156]. Similarly, indomethacine release was shown to be equally fast for drug-loaded nanocapsules and nanoemulsion [147]. The results of these studies show that polymer coating, the thickness of which is estimated to be about 10 nm, although necessary to maintain the structural stability of a nanocapsule, cannot act as an efficient barrier capable of providing sustained drug release from a formulation.
VI. THE EFFECT OF A STERILIZATION PROCEDURE Although the impact of sterilization methods on polymer integrity and drug release rate has been well characterized in microparticulate formulations [157–160], only a few studies addressed carrier stability and the effectiveness of a sterilization process applied to nanoparticles. In a series of experiments Alle´mann et al. evaluated the effect of sterilization by ␥-irradiation on the properties of PLA nanoparticles [106,161]. Unloaded and CGP 57813–loaded nanoparticles were exposed to a dose of 25 kGy on dry ice to avoid degradation by heat [161]. Whereas the size and morphology of the nanoparticles remained unchanged, ␥-irradiation led to a considerable reduction in the polymer molecular weight in both formulations without affecting the molecular weight polydispersity. The release rate of savoxepine from PLA nanoparticles was shown to be enhanced by ␥-irradiation [106]. The faster drug release observed in this study was apparently due to the decrease in the polymer molecular weight, resulting in a more hydrophilic nanoparticle matrix that is more permeable to water. This finding, however, is in contrast to the results reported by Song et al. [80], who examined the effect of an equal dose on the stability of PLGA nanoparticles loaded with U-86983. The lack of change in the drug release profile following exposure to ␥-irradiation observed in the latter case is apparently explained by a different drug and/or polymer or, alternatively, by different drug release methods employed in these studies. Recently membrane filtration was successfully applied for sterilization of sub-200-nm nanoparticles prepared by the salting-out process [107]. This method, though only applicable to nanoparticles with a size significantly below 0.2 m (the filter cutoff), has the advantage of not adversely affecting the drug release properties and the stability of a formulation nor the chemical stability of ingredients.
Biodegradable Nanoparticles as Drug Delivery Systems
415
REFERENCES
1. Ahsan F, Rivas IP, Khan MA, Torres Sua´rez AI. Targeting to macrophages: role of physicochemical properties of particulate carriers—liposomes and microspheres—on the phagocytosis by macrophages. J Control Rel 2002; 79:29–40. 2. Allen TM, Moase EH. Therapeutic opportunities for targeted liposomal drug delivery. Adv Drug Deliv Rev 1996; 21:117–133. 3. Kreuter J. The influence of coating with surfactants on the body distribution of nanoparticles after intravenous injection to rats. Clin Mater 1993; 13:131–134. 4. Leroux JC, Cozens R, Roesel JL, Galli B, Kubel F, Gurny R. Pharmacokinetics of a novel HIV-1 protease inhibitor incorporated into biodegradable or enteric nanoparticles following intravenous and oral administration to mice. J Pharm Sci 1995; 84:1387–1391. 5. Moghimi SM, Hunter AC, Murray JC. Long-circulating and target-specific nanoparticles: theory to practice. Pharmacol Rev 2001; 53:283–318. 6. Nishioka Y, Yoshino H. Lymphatic targeting with nanoparticulate system. Adv Drug Deliv Rev 2001; 47:55–64. 7. Schroeder U, Sommerfeld P, Ulrich S, Sabel BA. Nanoparticle technology for delivery of drugs across the blood-brain barrier. J Pharm Sci 1998; 87:1305–1307. 8. Torchilin VP. Drug targeting. Eur J Pharm Sci 2000; 11(Suppl 2):S81–S91. 9. Yokoyama M, Okano T. Targetable drug carriers: present status and a future perspective. Adv Drug Deliv Rev 1996; 21:77–80. 10. Mu¨ller RH, Jacobs C, Kayser O. Nanosuspensions as particulate drug formulations in therapy. Rationale for development and what we can expect for the future. Adv Drug Deliv Rev 2001; 47: 3–19. 11. Gref R, Domb A, Quellec P, Blunk T, Mu¨ller RH, Verbavatz JM, Langer R. The controlled intravenous delivery of drugs using PEG-coated sterically stabilized nanospheres. Adv Drug Deliv Rev 1995; 16:215–233. 12. Leroux JC, Alle´mann E, Jaeghere FD, Doelker E, Gurny R. Biodegradable nanoparticles–from sustained release formulations to improved site specific drug delivery. J Control Rel 1996; 39: 339–350. 13. Moghimi SM. Mechanisms regulating body distribution of nanospheres conditioned with pluronic and tetronic block-copolymers. Adv Drug Deliv Rev 1995; 16:183–193. 14. Stolnik S, Illum L, Davis SS. Long circulating microparticulate drug carriers. Adv Drug Deliv Rev 1995; 16:195–214. 15. Torchilin VP, Trubetskoy VS. Which polymers can make nanoparticulate drug carriers long-circulating?. Adv Drug Deliv Rev 1995; 16:141–155. 16. Anderson JM, Shive MS. Biodegradation and biocompatibility of PLA and PLGA microspheres. Adv Drug Deliv Rev 1997; 28:5–24. 17. Kreuter J. Nanoparticle-based drug delivery systems. J Control Rel 1991; 16:169–176. 18. Rı´hova´ B. Biocompatibility of biomaterials: hemocompatibility, immunocompatibility and biocompatibility of solid polymers and soluble targetable polymeric carriers. Adv Drug Deliv Rev 1996; 21:157–176. 19. Fishbein I, Chorny M, Rabinovich L, Banai S, Gati I, Golomb G. Nanoparticulate delivery system of a tyrphostin for the treatment of restenosis. J Control Rel 2000; 65:221–229. 20. Labhasetwar V, Song C, Humphrey W, Shebuski R, Levy RJ. Arterial uptake of biodegradable nanoparticles: effect of surface modifications. J Pharm Sci 1998; 87:1229–1234. 21. Cohen H, Levy RJ, Gao J, Fishbein I, Kousaev V, Sosnowski S, Slomkowski S, Golomb G. Sustained delivery and expression of DNA encapsulated in polymeric nanoparticles. Gene Ther 2000; 7: 1896–1905. 22. Davda J, Labhasetwar V. Characterization of nanoparticle uptake by endothelial cells. Int J Pharm 2002; 233:51–59.
416
Chorny et al.
23. Panyam J, Zhou W-Z, Prabha S, Sahoo SK, Labhasetwar V. Rapid endo-lysosomal escape of poly(DL-lactide-co-glycolide) nanoparticles: implications for drug and gene delivery. FASEB J 2002; 16:1217–1226. 24. Schwartz JB. Pharmaceutical manufacturing. In: Gennaro AR, Der Manderosian AH, Hanson GR, Medwick T, Popovich NG, Schnaare RL, Schwartz JB, White SH, eds. Remington: The Science and Practice of Pharmacy. Philadelphia: Lippincott Williams & Wilkins, 2000:780–820. 25. The United States Pharmacopeia. Vol. 24. Philadelphia: National Publishing, 2000. 26. Couvreur P, Couarraze G, Devissaguet JP, Puisieux F. Nanoparticles: preparation and characterization. In: Benita S, ed. Microencapsulation: Methods and Industrial Applications. Vol. 73. New York: Marcel Dekker, 1996:183–211. 27. Soppimath KS, Aminabhavi TM, Kulkarni AR, Rudzinski WE. Biodegradable polymeric nanoparticles as drug delivery devices. J Control Rel 2001; 70:1–20. 28. Alonso MJ. Nanoparticulate drug carrier technology. In: Lieberman HA, Rieger MM, Banker GS, eds. Pharmaceutical Dosage Forms: Disperse Systems. Vol. 1. New York: Marcel Dekker, 1996: 203–242. 29. Janes KA, Calvo P, Alonso MJ. Polysaccharide colloidal particles as delivery systems for macromolecules. Adv Drug Deliv Rev 2001; 47:83–97. 30. Vila A, Sanchez A, Tobio M, Calvo P, Alonso MJ. Design of biodegradable particles for protein delivery. J Control Rel 2002; 78:15–24. 31. De Campos AM, Sanchez A, Alonso MJ. Chitosan nanoparticles: a new vehicle for the improvement of the delivery of drugs to the ocular surface. Application to cyclosporin A. Int J Pharm 2001; 224: 159–168. 32. Janes KA, Fresneau MP, Marazuela A, Fabra A, Alonso MJ. Chitosan nanoparticles as delivery systems for doxorubicin. J Control Rel 2001; 73:255–267. 33. Leong KW, Mao H-Q, Truong-Le VL, Roy K, Walsh SM, August JT. DNA-polycation nanospheres as non-viral gene delivery vehicles. J Control Rel 1998; 53:183–193. 34. Mao H-Q, Roy K, Truong-Le VL, Janes KA, Lin KY, Wang Y, August JT, Leong KW. ChitosanDNA nanoparticles as gene carriers: synthesis, characterization and transfection efficiency. J Control Rel 2001; 70:399–421. 35. Aynie I, Vauthier C, Chacun H, Fattal E, Couvreur P. Spongelike alginate nanoparticles as a new potential system for the delivery of antisense oligonucleotides. Antisense Nucleic Acid Drug Dev 1999; 9:301–312. 36. Domb AJ, Kost J, Wiseman DM. Handbook of Biodegradable Polymers. Vol. 7. Amsterdam: Harwood Academic Publishers, 1997:526. 37. Kumar N, Ravikumar MN, Domb AJ. Biodegradable block copolymers. Adv Drug Deliv Rev 2001; 53:23–44. 38. Peracchia MT, Dismae¨le D, Couvreur P, d’Angelo J. Synthesis of novel MePEGcyanoacrylatehexadecylcyanoacrylate amphiphilic copolymer for nanoparticle technology. Macromolecules 1997; 30:846–851. 39. Peracchia MT, Vauthier C, Desmae¨le D, Gulik A, Dedieu JC, Demoy M, d’Angelo J, Couvreur P. Pegylated nanoparticles from a novel methoxypolyethylene glycol cyanoacrylate-hexadecyl cyanoacrylate amphiphilic copolymer. Pharm Res 1998; 15:550–556. 40. Grislain L, Couvreur P, Lenaerts V, Roland M, Deprez-Decampeneere D, Speiser P. Pharmacokinetics and distribution of a biodegradable drug carrier. Int J Pharm 1983; 15:335–345. 41. Van Snick L, Couvreur P, Christiaens-Leyh D, Roland M. Molecular weights of free and drugloaded nanoparticles. Pharm Res 1985; 1:36–41. 42. Seijo B, Fattal E, Roblot-Treupel L, Couvreur P. Design of nanoparticles of less than 50 nm diameter: preparation, characterization and drug loading. Int J Pharm 1990; 62:1–7. 43. Illum L, Khan MA, Mak E, Davis SS. Evaluation of carrier capacity and release characteristics for poly(butyl 2-cyanoacrylate) nanoparticles. Int J Pharm 1986; 30:17–28. 44. da Silveira AM, Ponchel G, Puisieux F, Ducheˆne D. Combined poly(isobutylcyanoacrylate) and cyclodextrin nanoparticles for enhancing the encapsulation of lipophilic drugs. Pharm Res 1998; 15:1051–1055.
Biodegradable Nanoparticles as Drug Delivery Systems
417
45. Guise V, Drouin JY, Benoit J, Mahuteau J, Dumont P, Couvreur P. Vidarabine-loaded nanoparticles: a physicochemical study. Pharm Res 1990; 7:736–741. 46. Grangier JL, Puygrenier M, Gauthier JC, Couvreur P. Nanoparticles as carriers for growth hormone releasing factor (GFR). J Control Rel 1991; 15:3–13. 47. Al Khouri Fallouh N, Roblot-Treupel L, Fessi H, Devissaguet JP, Puisieux F. Development of a new process for the manufacture of polyisobutylcyanoacrylate nanocapsules. Int J Pharm 1986; 28: 125–132. 48. Rollot JM, Couvreur P, Roblot-Treupel L, Puisieux F. Physicochemical and morphological characterization of polyisobutylcyanoacrylate nanocapsules. J Pharm Sci 1986; 75:361–364. 49. Lambert G, Fattal E, Pinto-Alphandary H, Gulik A, Couvreur P. Polyisobutylcyanoacrylate nanocapsules containing an aqueous core as a novel colloidal carrier for the delivery of oligonucleotides. Pharm Res 2000; 17:707–714. 50. Lambert G, Fattal E, Pinto-Alphandary H, Gulik A, Couvreur P. Polyisobutylcyanoacrylate nanocapsules containing an aqueous core for the delivery of oligonucleotides. Int J Pharm 2001; 214:13–16. 51. Chavany C, Saison-Behmoaras T, Le Doan T, Puisieux F, Couvreur P, He´le`ne C. Adsorption of oligonucleotides onto polyisohexylcyanoacrylate nanoparticles protects them against nucleases and increases their cellular uptake. Pharm Res 1994; 11:1370–1378. 52. Nakada Y, Fattal E, Foulquier M, Couvreur P. Pharmacokinetics and biodistribution of oligonucleotide adsorbed onto poly(isobutylcyanoacrylate) nanoparticles after intravenous administration in mice. Pharm Res 1996; 13:38–43. 53. De Keyser J-L, Poupaert JH, Dumont P. Poly(diethyl methylidenemalonate) nanoparticles as a potential drug carrier: preparation, distribution and elimination after intravenous and peroral administration in mice. J Pharm Sci 1991; 80:67–70. 54. Lescure F, Seguin C, Breton P, Bourrinet P, Roy D, Couvreur P. Preparation and characterization of novel poly(methylidene malonate 2.1.2)-made nanoparticles. Pharm Res 1994; 11:1270–1277. 55. Brigger I, Chaminade P, Marsaud V, Appel M, Besnard M, Gurny R, Renoir M, Couvreur P. Tamoxifen encapsulation within polyethylene glycol–coated nanospheres. A new antiestrogen formulation. Int J Pharm 2001; 214:37–42. 56. Peracchia MT, Fattal E, Desmaele D, Besnard M, Noel JP, Gomis JM, Appel M, d’Angelo J, Couvreur P. Stealth PEGylated polycyanoacrylate nanoparticles for intravenous administration and splenic targeting. J Control Rel 1999; 60:121–128. 57. Calvo P, Gouritin B, Brigger I, Lasmezas C, Deslys J, Williams A, Andreux JP, Dormont D, Couvreur P. PEGylated polycyanoacrylate nanoparticles as vector for drug delivery in prion diseases. J Neurosci Methods 2001; 111:151–155. 58. Calvo P, Gouritin B, Chacun H, Desmaele D, D’Angelo J, Noel JP, Georgin D, Fattal E, Andreux JP, Couvreur P. Long-circulating PEGylated polycyanoacrylate nanoparticles as new drug carrier for brain delivery. Pharm Res 2001; 18:1157–1166. 59. (Pekarek) Leach KJ, Mathiowitz E. Degradation of double-walled polymer microspheres of PLLA and P(CPP:SA)20:80. I. In vitro degradation. Biomaterials 1998; 19:1973–1980. 60. Belbella A, Vauthier C, Fessi H, Devissaguet JP, Puisieux F. In vitro degradation of nanospheres from poly(D,L-lactides) of different molecular weights and polydispersities. Int J Pharm 1996; 129: 95–102. 61. Lemoine D, Francois C, Kedzierewicz F, Preat V, Hoffman M, Maincent P. Stability study of nanoparticles of poly(ⱖ-caprolactone), poly(D,L-lactide) and poly(D,L-lactide-co-glycolide). Biomaterials 1996; 17:2191–2197. 62. Santos CA, Freedman BP, Leach KJ, Press DL, Scarpulla M, Mathiowitz E. Poly(fumaric-co-sebacic anhydride). A degradation study as evaluated by FTIR, DSC, GPC and X-ray diffraction. J Controlled Release 1999; 60:11–22. 63. Tabata Y, Gutta S, Langer R. Controlled delivery systems for proteins using polyanhydride microspheres. Pharm Res 1993; 10:487–496. 64. Tabata Y, Langer R. Polyanhydride microspheres that display near-constant release of water-soluble model drug compounds. Pharm Res 1993; 10:391–399.
418
Chorny et al.
65. Lacasse FX, Filion MC, Phillips NC, Escher E, McMullen JN, Hildgen P. Influence of surface properties at biodegradable microsphere surfaces: effects on plasma protein adsorption and phagocytosis. Pharm Res 1998; 15:312–317. 66. Domb AJ, Elmalak O, Shastri VB, Ta-Shma Z, Masters DM, Ringel I, Teomim D, Langer R. Polyanhydrides. In: Domb AJ, Kost J, Wiseman DM, eds. Handbook of Biodegradable Polymers. Vol. 7. Amsterdam: Harwood Academic Publishers, 1997:135–159. 67. Perrin DA, English JP. Polyglycolide and polylactide. In: Domb AJ, Kost J, Wiseman DM, eds. Handbook of Biodegradable Polymers. Vol. 7. Amsterdam: Harwood Academic Publishers, 1997: 3–26. 68. Pekarek KJ, Jacobs JS, Mathiowitz E. One-step preparation of double-walled microspheres. Adv Mater 1994; 6:684–687. 69. Pekarek KJ, Dyrud MJ, Ferrer K, Jong YS, Mathiowitz E. In vitro and in vivo degradation of double-walled polymer microspheres. J Control Rel 1996; 40:169–178. 70. Venier-Julienne MC, Benoıˆt JP. Preparation, purification and morphology of polymeric nanoparticles as drug carriers. Pharm Acta Helv 1996; 71:121–128. 71. Quintanar-Guerrero D, Alle´mann E, Fessi H, Doelker E. Preparation techniques and mechanisms of formation of biodegradable nanoparticles from preformed polymers. Drug Dev Ind Pharm 1998; 24:1113–1128. 72. Sah H. Microencapsulation techniques using ethyl acetate as a dispersed solvent: effects of its extraction rate on characteristics of PLGA microspheres. J Control Rel 1997; 47:233–245. 73. Quellec P, Gref R, Dellacherie E, Sommer F, Tran MD, Alonso MJ. Protein encapsulation within poly(ethylene glycol)–coated nanospheres. II. Controlled release properties. J Biomed Mater Res 1999; 47:388–395. 74. Zambaux MF, Bonneaux F, Gref R, Dellacherie E, Vigneron C. Preparation and characterization of protein C–loaded PLA nanoparticles. J Control Rel 1999; 60:179–188. 75. Zambaux MF, Bonneaux F, Gref R, Maincent P, Dellacherie E, Alonso MJ, Labrude P, Vigneron C. Influence of experimental parameters on the characteristics of poly(lactic acid) nanoparticles prepared by a double emulsion method. J Control Rel 1998; 50:31–40. 76. Go¨pferich A, Alonso MJ, Langer R. Development and characterization of microencapsulated microspheres. Pharm Res 1994; 11:1568–1574. 77. Li X, Zhang Y, Yan R, Jia W, Yuan M, Deng X, Huang Z. Influence of process parameters on the protein stability encapsulated in poly-DL-lactide–poly(ethylene glycol) microspheres. J Control Rel 2000; 68:41–52. 78. Li W-I, Anderson KW, DeLuca PP. Kinetic and thermodynamic modeling of the formation of polymeric microspheres using solvent extraction/evaporation method. J Control Rel 1995; 37: 187–198. 79. Li W-I, Anderson KW, Mehta RC, DeLuca PP. Prediction of solvent removal profile and effect on properties for peptide-loaded PLGA microspheres prepared by solvent extraction/evaporation method. J Control Rel 1995; 37:199–214. 80. Song CX, Labhasetwar V, Murphy H, Qu X, Hunphrey WR, Shebuski RJ, Levy RJ. Formulation and characterization of biodegradable nanoparticles for intravascular local drug delivery. J Control Rel 1997; 43:197–212. 81. Feng S, Huang G. Effects of emulsifiers on the controlled release of paclitaxel (Taxol威) from nanospheres of biodegradable polymers. J Control Rel 2001; 71:53–69. 82. Go¨rner T, Gref R, Michenot D, Sommer F, Tran MN, Dellacherie E. Lidocaine-loaded biodegradable nanospheres. I. Optimization of the drug incorporation into the polymer matrix. J Control Rel 1999; 57:259–268. 83. Mu L, Feng SS. Vitamin E TPGS used as emulsifier in the solvent evaporation/extraction technique for fabrication of polymeric nanospheres for controlled release of paclitaxel (Taxol威). J Control Rel 2002; 80:129–144. 84. Ueda M, Kreuter J. Optimization of the preparation of loperamide-loaded poly (L-lactide) nanoparticles by high pressure emulsification–solvent evaporation. J Microencapsul 1997; 14:593–605. 85. Vandervoort J, Ludwig A. Preparation factors affecting the properties of polylactide nanoparticles: a factorial design study. Pharmazie 2001; 56:484–488.
Biodegradable Nanoparticles as Drug Delivery Systems
419
86. Herrmann J, Bodmeier R. Somatostatin containing biodegradable microspheres prepared by a modified solvent evaporation method based on w/o/wmultiple emulsions. Int J Pharm 1995; 126:129–138. 87. Lemoine D, Pre´at V. Polymeric nanoparticles as delivery system for influenza virus glycoproteins. J Control Rel 1998; 54:15–27. 88. Labhasetwar V, Bonadio J, Goldstein SA, Levy RJ. Gene transfection using biodegradable nanospheres: results in tissue culture and a rat osteotomy model. Colloids Surfaces B: Biointerfaces 1999; 16:281–290. 89. Izumikawa S, Yoshioka S, Aso Y, Takeda Y. Preparation of poly(L-lactide) microspheres of different crystalline morphology and effect of crystalline morphology on drug release rate. J Control Rel 1991; 15:133–140. 90. Liggins RT, Burt HM. Paclitaxel loaded poly(L-lactic acid) microspheres: properties of microspheres made with low molecular weight polymers. Int J Pharm 2001; 222:19–33. 91. Polakovic M, Go¨rner T, Gref R, Dellacherie E. Lidocaine loaded biodegradable nanospheres. II. Modelling of drug release. J Control Rel 1999; 60:169–177. 92. O’Donnell PB, McGinity JW. Preparation of microspheres by the solvent evaporation technique. Adv Drug Deliv Rev 1997; 28:25–42. 93. Gref R, Quellec P, Sanchez A, Calvo P, Dellacherie E, Alonso MJ. Development and characterization of CyA-loaded poly(lactic acid)–poly(ethylene glycol) micro- and nanoparticles. Comparison with conventional PLA particulate carriers. Eur J Pharm Biopharm 2001; 51:111–118. 94. Vandervoort J, Ludwig A. Biocompatible stabilizers in the preparation of PLGA nanoparticles: a factorial design study. Int J Pharm 2002; 238:77–92. 95. Verrecchia T, Huve P, Bazile D, Veillard M, Spenlehauer G, Couvreur P. Adsorption/desorption of human serum albumin at the surface of poly(lactic acid) nanoparticles prepared by a solvent evaporation process. J Biomed Mater Res 1993; 27:1019–1028. 96. Scholes PD, Coombes AGA, Illum L, Davis SS, Vert M, Davies MC. The preparation of sub-200 nm poly(lactide-co-glycolide) microspheres for site-specific drug delivery. J Control Rel 1993; 25: 145–153. 97. Birnbaum DT, Kosmala JD, Brannon-Peppas L. Optimization of preparation techniques for poly(lactic acid-co-glycolic acid) nanoparticles. J Nanoparticle Res 2000; 2:173–181. 98. Quintanar-Guerrero D, Alle´mann E, Fessi H, Doelker E. Pseudolatex preparation using a novel emulsion-diffusion process involving direct displacement of partially water–miscible solvents by distillation. Int J Pharm 1999; 188:155–164. 99. Nihant N, Schugens C, Grandfils C, Jerome R, Teyssie P. Polylactide microparticles prepared by double emulsion/evaporation technique. I. Effect of primary emulsion stability. Pharm Res 1994; 11:1479–1484. 100. Soriano I, Delgado A, Diaz RV, Evora C. Use of surfactants in polylactic acid protein microspheres. Drug Dev Ind Pharm 1995; 21:549–558. 101. Ando S, Putnam D, Pack DW, Langer R. PLGA microspheres containing plasmid DNA: preservation of supercoiled DNA via cryopreparation and carbohydrate stabilization. J Pharm Sci 1999; 88: 126–130. 102. Hueper WC. Carcinogenic studies on water-soluble and insoluble macromolecules. Arch Pathol 1959; 67:589–617. 103. Yamaoka T, Tabata Y, Ikada Y. Comparison of body distribution of poly(vinyl alcohol) with other water-soluble polymers after intravenous administration. J Pharm Pharmacol 1995; 47:479–486. 104. Sahoo SK, Panyam J, Prabha S, Labhasetwar V. Residual polyvinyl alcohol associated with poly(D,Llactide-co-glycolide) nanoparticles affects their physical properties and cellular uptake. J Control Rel 2002; 82:105–114. 105. Scholes PD, Coombes AG, Illum L, Davis SS, Watts JF, Ustariz C, Vert M, Davies MC. Detection and determination of surface levels of poloxamer and PVA surfactant on biodegradable nanospheres using SSIMS and XPS. J Control Rel 1999; 59:261–278. 106. Alle´mann E, Leroux JC, Gurny R, Doelker E. In vitro extended-release properties of drug-loaded poly(DL-lactic acid) nanoparticles produced by a salting-out procedure. Pharm Res 1993; 10: 1732–1737.
420
Chorny et al.
107. Konan YN, Gurny R, Alle´mann E. Preparation and characterization of sterile and freeze-dried sub200-nm nanoparticles. Int J Pharm 2002; 233:239–252. 108. Quintanar-Guerrero D, Ganem-Quintanar A, Alle´mann E, Fessi H, Doelker E. Influence of the stabilizer coating layer on the purification and freeze-drying of poly(D,L-lactic acid) nanoparticles prepared by an emulsion–diffusion technique. J Microencapsul 1998; 15:107–119. 109. Landry FB, Bazile DV, Spenlehauer G, Veillard M, Kreuter J. Release of the fluorescent marker Prodan威 from poly(D,L-lactic acid) nanoparticles coated with albumin or polyvinyl alcohol in model digestive fluids (USP XXII). J Control Rel 1997; 44:227–236. 110. Verrecchia T, Spenlehauer G, Bazile D, Murry-Brelier A, Archimbaud Y, Vieillard M. Non-stealth (poly(lactic acid albumin)) and stealth (poly(lactic acid-polyethylene glycol)) nanoparticles as injectable drug carriers. J Control Rel 1995; 36:49–61. 111. Bazile DV, Ropert C, Huve P, Verrecchia T, Marlard M, Frydman A, Veillard M, Spenlehauer G. Body distribution of fully biodegradable [14C]-poly(lactic acid) nanoparticles coated with albumin after parenteral administration to rats. Biomaterials 1992; 13:1093–1102. 112. Evora C, Soriano I, Rogers RA, Shakesheff KN, Hanes J, Langer R. Relating the phagocytosis of microparticles by alveolar macrophages to surface chemistry: the effect of 1,2-dipalmitoylphosphatidylcholine. J Control Rel 1998; 51:143–152. 113. Mu L, Feng SS. Fabrication, characterization and in vitro release of paclitaxel (Taxol威) loaded poly(lactic-co-glycolic acid) manospheres prepared by spray drying technique with lipid/cholesterol emulsifiers. J Control Rel 2001; 76:239–254. 114. Bitz C, Doelker E. Influence of the preparation method on residual solvents in biodegradable microspheres. Int J Pharm 1996; 131:171–181. 115. Alle´mann E, Gurny R, Doelker E. Preparation of aqueous polymeric nanodispersions by a reversible salting-out process: influence of process parameters on particle size. Int J Pharm 1992; 87:247–253. 116. Ibrahim H, Bindschaedler C, Doelker E, Buri P, Gurny R. Aqueous nanodispersions prepared by a salting-out process. Int J Pharm 1992; 87:239–246. 117. Zweers ML, Engbers GHM, Grijpma DW, Feijen J. Biodegradable nanospheres for local drug delivery to prevent restenosis after baloon angioplasty. J Control Rel 2001; 72:225–309. 118. De Jaeghere F, Alle´mann E, Leroux JC, Stevels W, Feijen J, Doelker E, Gurny R. Formulation and lyoprotection of poly(lactic acid-co-ethylene oxide) nanoparticles: influence on physical stability and in vitro cell uptake. Pharm Res 1999; 16:859–866. 119. De Jaeghere F, Alle´mann E, Feijen J, Kissel T, Doelker E, Gurny R. Cellular uptake of PEO surface-modified nanoparticles: evaluation of nanoparticles made of PLA:PEO diblock and triblock copolymers. J Drug Target 2000; 8:143–153. 120. De Jaeghere F, Alle´mann E, Feijen J, Kissel T, Doelker E, Gurny R. Freeze-drying and lyopreservation of diblock and triblock poly(lactic acid)–poly(ethylene oxide) (PLA-PEO) copolymer nanoparticles. Pharm Dev Technol 2000; 5:473–483. 121. Berton M, Alle´mann E, Stein CA, Gurny R. Highly loaded nanoparticulate carrier using an hydrophobic antisense oligonucleotide complex. Eur J Pharm Sci 1999; 9:163–170. 122. Quintanar-Guerrero D, Fessi H, Alle´mann E, Doelker E. Influence of stabilizing agents and preparative variables on the formation of poly(D,L-lactic acid) nanoparticles by an emulsification-diffusion technique. Int J Pharm 1996; 143:133–141. 123. Berton M, Benimetskaya L, Alle´mann E, Stein CA, Gurny R. Uptake of oligonucleotide-loaded nanoparticles in prostatic cancer cells and their intracellular localization. Eur J Pharm Biopharm 1999; 47:119–123. 124. Delie F, Berton M, Alle´mann E, Gurny R. Comparison of two methods of encapsulation of an oligonucleotide into poly(D,L-lactic acid) particles. Int J Pharm 2001; 214:25–30. 125. Quintanar-Guerrero D, Alle´mann E, Doelker E, Fessi H. Preparation and characterization of nanocapsules from preformed polymers by a new process based on emulsification–diffusion technique. Pharm Res 1998; 15:1056–1062. 126. Trotta M, Gallarate M, Pattarino F, Morel S. Emulsions containing partially water-miscible solvents for the preparation of drug nanosuspensions. J Control Rel 2001; 76:119–128. 127. Wehrle P, Magenheim B, Benita S. The influence of process parameters on the PLA nanoparticle size distribution, evaluated by means of factorial design. Eur J Pharm Biopharm 1995; 41:19–26.
Biodegradable Nanoparticles as Drug Delivery Systems
421
128. Ford J, Woolfe J, Florence AT. Nanospheres of cyclosporin A: poor oral absorption in dogs. Int J Pharm 1999; 183:3–6. 129. Barichello JM, Morishita M, Takayama K, Nagai T. Encapsulation of hydrophilic and lipophilic drugs in PLGA nanoparticles by the nanoprecipitation method. Drug Dev Ind Pharm 1999; 25: 471–476. 130. Govender T, Stolnik S, Garnett MC, Illum L, Davis SS. PLGA nanoparticles prepared by nanoprecipitation: drug loading and release studies of a water soluble drug. J Control Rel 1999; 57:171–185. 131. Govender T, Riley T, Stolnik S, Garnett MC, Illum L, Davis SS. PLA-PEG nanoparticles for site specific delivery: drug incorporation study. J Controlled Release 2000; 64:318–319. 132. Govender T, Riley T, Ehtezazi T, Garnett MC, Stolnik S, Illum L, Davis SS. Defining the drug incorporation properties of PLA-PEG nanoparticles. Int J Pharm 2000; 199:95–110. 133. Chaco´n M, Berges L, Molpeceres J, Aberturas MR, Guzman M. Optimized preparation of poly(D,Llactic-glycolic) microspheres and nanoparticles for oral administration. Int J Pharm 1996; 141: 81–91. 134. Panagi Z, Beletsi A, Evangelatos G, Livaniou E, Ithakissios DS, Avgoustakis K. Effect of dose on the biodistribution and pharmacokinetics of PLGA and PLGA-mPEG nanoparticles. Int J Pharm 2001; 221:143–152. 135. Verger MLL, Fluckiger L, Kim YI, Hoffman M, Maincent P. Preparation and characterization of nanoparticles containing an antihypertensive agent. Eur J Pharm Biopharm 1998; 46:137–143. 136. Stolnik S, Garnett MC, Davies MC, Illum L, Bousta M, Vert M, Davis SS. The colloidal properties of surfactant-free biodegradable nanospheres from poly(-malic acid-co-benzyl malate)s and poly(lactic acid-co-glycolide). Colloids and Surfaces A: Physicochem and Eng Aspects 1995; 97: 235–245. 137. Chaco´n M, Molpeceres J, Berges L, Guzman M, Aberturas MR. Stability and freeze-drying of cyclosporine loaded poly(D,L lactide-glycolide) carriers. Eur J Pharm Sci 1999; 8:99–107. 138. Chorny M, Fishbein I, Danenberg HD, Golomb G. Lipophilic drug loaded nanospheres prepared by nanoprecipitation: effect of formulation variables on size, drug recovery and release kinetics. J Control Rel 2002. 139. Guzman M, Molpeceres J, Garcia F, Aberturas MR, Rodriguez M. Formation and characterization of cyclosporine-loaded nanoparticles. J Pharm Sci 1993; 82:498–502. 140. Magenheim B, Levy MY, Benita S. A new in vitro technique for the evaluation of drug release profile from colloidal carriers—ultrafiltration technique at low pressure. Int J Pharm 1993; 94: 115–123. 141. Molpeceres J, Guzman M, Aberturas MR, Chaco´n M, Berges L. Application of central composite designs to the preparation of polycaprolactone nanoparticles by solvent displacement. J Pharm Sci 1996; 85:206–213. 142. Murakami H, Kawashima A, Niwa T, Hino T, Takeuchi H, Kobayashi M. Influence of degrees of hydrolyzation and polymerization of poly(vinyl alcohol) on the preparation and properties of poly(DL-lactide-co-glycolide) nanoparticles. Int J Pharm 1997; 149:43–49. 143. Murakami H, Kobayashi M, Takeuchi H, Kawashima Y. Preparation of poly(DL-lactide-co-glycolide) nanoparticles by modified spontaneous emulsification solvent diffusion method. Int J Pharm 1999; 187:143–152. 144. Kawashima Y, Yamamoto H, Takeuchi H, Hino T, Niwa T. Properties of a peptide containing DLlactide/glycolide copolymer nanospheres prepared by novel emulsion solvent diffusion methods. Eur J Pharm Biopharm 1998; 45:41–48. 145. Niwa T, Takeuchi H, Nino T, Kunou N, Kawashima Y. Preparations of biodegradable nanospheres of water-soluble and insoluble drugs with D,L-lactide/glycolide copolymer by a novel spontaneous emulsification solvent diffusion method, and the drug release behavior. J Controlled Release 1993; 25:89–98. 146. Niwa T, Takeuchi H, Hino T, Kunou N, Kawashima Y. In vitro drug release behavior of D,L-lactide/ glycolide copolymer (PLGA) nanospheres with nafarelin acetate prepared by a novel spontaneous emulsificationsolvent diffusion method. J Pharm Sci 1994; 83:727–732.
422
Chorny et al.
147. Calvo P, Vila-Jato JL, Alonso MJ. Comparative in vitro evaluation of several colloidal systems, nanoparticles, nanocapsules, and nanoemulsions, as ocular drug carriers. J Pharm Sci 1996; 85: 530–536. 148. Magenheim B, Benita S. Elucidation of the indomethacin in vitro release mechanism from polylactic acid nanoparticles. Pharm Res 1993; 10:S–187. 149. Hirosue S, Mu¨ller BG, Mulligan RC, Langer R. Plasmid DNA encapsulation and release from solvent diffusion nanospheres. J Control Rel 2001; 70:231–242. 150. Perez C, Sanchez A, Putnam D, Ting D, Langer R, Alonso MJ. Poly(lactic acid)–poly(ethylene glycol) nanoparticles as new carriers for the delivery of plasmid DNA. J Control Rel 2001; 75: 211–224. 151. Chorny M, Fishbein I, Danenberg HD, Golomb G. Study of the drug release mechanism from tyrphostin AG-1295 loaded nanospheres using in situ and external sink methods. J Control Rel 2002. 152. Fishbein I, Chorny M, Banai S, Levitzki A, Danenberg HD, Gao J, Chen X, Moerman E, Gati I, Goldwasser V, Golomb G. Formulation and delivery mode affect disposition and activity of tyrphostin-loaded nanoparticles in the rat carotid model. Atheroscler Thromb Vasc Biol 2001; 21: 1434–1439. 153. Fessi H, Puisieux F, Devissaguet JP, Ammoury N, Benita S. Nanocapsule formation by interfacial polymer deposition following solvent displacement. Int J Pharm 1989; 55:R1–R4. 154. Ammoury N, Fessi H, Devissaguet JP, Puisieux F, Benita S. In vitro release kinetic pattern of indomethacin from poly(D,L-lactide) nanocapsules. J Pharm Sci 1990; 79:763–767. 155. Mosqueira VCF, Legrand P, Pinto-Alphandary H, Puisieux F, Barratt G. Poly(D,L-lactide) nanocapsules prepared by a solvent displacement process: influence of the composition on physicochemical and structural properties. J Pharm Sci 2000; 89:614–626. 156. Magalhaes NS, Fessi H, Puisieux F, Benita S, Seiller M. An in vitro release kinetic examination and comparative evaluation between submicron emulsion and polylactic acid nanocapsules of clofibride. J Microencapsul 1995; 12:195–205. 157. Bittner B, Mader K, Kroll C, Borchert HH, Kissel T. Tetracycline-HCl–loaded poly(DL-lactide-coglycolide) microspheres prepared by a spray drying technique: influence of gamma-irradiation on radical formation and polymer degradation. J Control Rel 1999; 59:23–32. 158. Hausberger AG, Kenley RA, DeLuca PP. Gamma irradiation effects on molecular weight and in vitro degradation of poly(D,L-lactide-co-glycolide) microparticles. Pharm Res 1995; 12:851–856. 159. Montanari L, Cilurzo F, Valvo L, Faucitano A, Buttafava A, Groppo A, Genta I, Conti B. Gamma irradiation effects on stability of poly(lactide-co-glycolide) microspheres containing clonazepam. J Control Rel 2001; 75:317–330. 160. Shen ZR, Zhu JH, Ma Z, Wang F, Wang ZY. Preparation of biodegradable microspheres of testosterone with poly(D,L-lactide-co-glycolide) and test of drug release in vitro. Artif Cells Blood Substit Immobil Biotechnol 2000; 28:57–64. 161. Alle´mann E, Gurny R, Leroux JC. Biodegradable nanoparticles of poly(lactic acid) and poly(lacticco-glycolic acid) for parenteral administration. In: Liberman HA, Rieger MM, Banker GS, eds. Pharmaceutical Dosage Forms: Disperse Systems: Marcel Dekker, 1996:163–194.
19 Biodegradable Hydrogels as Drug Controlled Release Vehicles C. C. Chu Cornell University Ithaca, New York, U.S.A.
I. INTRODUCTION Hydrogels are three-dimensional, hydrophilic, polymeric networks capable of absorbing and retaining large amounts of water or biological fluids [1]. The hydrogel networks are formed by crosslinking polymer chains via covalent ionic, or hydrogen bonds or via physical entanglement. The crosslinked polymer chains provide the network structure and physical integrity [1,2]. The crosslinking density is one of the most important factors that affect the swelling of hydrogels. Highly crosslinked hydrogels have a tighter structure so that they hinder the mobility of the polymer chains, hence lowering their swelling ratios. The chemical structure of hydrogels also affects their swelling ratios. Hydrogels containing hydrophilic groups swell to a higher degree than those containing hydrophobic ones. There are numerous applications of hydrogels. Specifically, hydrogels are widely used in the medical and pharmaceutical industries for the following reasons [1,3,4]. First, hydrogels have low surface tension with surrounding biological fluids. This is due to the high water content of the materials, which contributes to their biocompatibility. Second, hydrogels stimulate hydrodynamic properties of cells and tissues in many ways. The high mobility of polymer chains at a hydrogel surface contributes to the prevention of protein absorption and cell adhesion. Third, the soft rubbery nature of swollen hydrogels can minimize their mechanical irritation to surrounding tissues [1,4]. In a drug delivery system, hydrogels should be designed to protect sensitive therapeutic agents until their release from the system and must maintain their integrity. However, there is a major disadvantage of hydrogels as biomaterials, i.e., poor mechanical properties after swelling [3,4]. Many researchers have investigated means to improve the mechanical strength of hydrogels. Crosslinking density is one of the important factors to mechanical integrity. Increasing crosslinking density will result in a stronger hydrogel. However, the approach of using higher crosslinking density has one major drawback—a lower swelling ratio. Contrary to conventional hydrogels, there is a special group, stimuli-sensitive, intelligent, or smart hydrogels, that can respond to the changes in environmental conditions, such as pH, temperature, light, electric field, magnetic field, metabolite, and pressure [5–10], and the extent of the response can be controlled. These smart hydrogels can exhibit dramatic changes in their swelling behavior, network structure, permeability, or mechanical strength in response to alterations in environmental conditions. The smart hydrogels have been used in various biomedical 423
424
Chu
fields, such as artificial muscle [11], immobilization of enzyme and cells [12,13], and drug delivery [14–18]. Among intelligent hydrogels, pH- and/or temperature-sensitive hydrogels have been subject to extensive studies. Temperature-sensitive hydrogels have gained considerable attention in the pharmaceutical field due to their ability to swell or deswell as a result of temperature change [5,14–18]. Many polymers exhibit a temperature-responsive phase transition property. The common characteristic of temperature-sensitive polymers is the presence of hydrophobic groups, such as methyl, ethyl and propyl groups. Poly(N-isopropylacrylamide) (PNIPAAm) is the most extensively studied, and poly(N,N-diethylacrylamide) (PDEAAm) has also been widely used among poly(N,N′-alkyl substituted acrylamide). The copolymer of NIPAAm and DEAAm having an increase amount of hydrophilic comonomers, leads to a raised lower critical solution temperature (LCST), the temperature that induces phase transitions like the collapse of gels. The block copolymers of poly(ethylene oxide) (PEO) and poly(propylene oxide) (PPO), such as Pluronic威 (or Poloxamer) and Tetronics威 [18,19] also show some temperature sensitivity. Kim et al. advanced the temperature-sensitive polymers into biodegradable temperature-sensitive polymers by integrating PEG into synthetic aliphatic polyesters like polyglycolic acid and polylactic acid, e.g., block copolymer poly(ethylene glycol-b-(DL-lactic acid-co-glycolic acid)-b-ethylene glycol) (PEG-PLGAPEG). This block PEG-PLGA-PEG copolymer is in aqueous solution form at room temperature but becomes a transparent hydrogel at body temperature [20,21]. Temperature-sensitive hydrogels can also be classified as positive or negative temperaturesensitive systems. A positive temperature-sensitive hydrogel has an upper critical solution temperature (UCST). Such hydrogels contract upon cooling below their UCST. Negative temperature-sensitive hydrogels have a lower critical solution temperature (LCST). These hydrogels collapse upon heating above their LCST [18,22]. This temperature-induced phase separation property of hydrogels makes them especially useful for biomedical and bioengineering applications, such as tissue engineering [4], protein/ligand recognition [23], on-off switches [24], and artificial organs [25]. Since the rapid response rate and large volume changes due to temperature variation are the essential requirements for the practical applications of intelligent hydrogels, there is a clear need for thermoresponsive hydrogels with improved response rate and larger volume changes upon an external temperature stimulus. The response and volume change rates of thermosensitive PNIPAAm hydrogels are a diffusion-controlled process and are determined by the collective diffusion coefficient of water through the hydrogel matrix. In this regard, several attempts [26–33] have been proposed to achieve fast responsive thermoresponsive PNIPAAm hydrogel. For instance, Kabra et al. [26] utilized the phase separation technique to yield fast responsive PNIPAAm hydrogels. Wu et al. [27] modified the phase separation technique: the polymerization was conducted at a temperature above the LCST (50⬚C) and the reactor was evacuated near the end of polymerization. This resulted in a macroporous hydrogel structure, and hence a fast response rate was achieved. Yoshida et al. [28] found that introducing the freely mobile and grafted polymer chain to the network of PNIPAAm gel could improve the deswelling rate of the gel matrix during the shrinking process owing to the increased channels for water diffusion. Recently, several strategies were also reported by Zhang et al. [29–33] that the improved response rate of the gels could be obtained through incorporating siloxane linkage, cold or vacuum polymerization, and using the pore-forming agent. Whether conventional or intelligent hydrogels, they are usually prepared from nonbiodegradable polymers. Although these nonbiodegradable hydrogels have some successful biomedical applications, they do have some drawbacks. First, the presence of these nonbiodegradable foreign materials inside human body is known to elicit permanent tissue reaction, which is undesirable, particularly in patients of compromised immune response, impaired wound healing,
Biodegradable Hydrogels as Drug Vehicles
425
malnutrition, and diabetes. Second, the proper release of the recently genetically engineered pharmacologically active peptides and proteins from nonbiodegradable hydrogels may be difficult because these molecularly large sized bioactive agents are very difficult to be completely released from nonbiodegradable hydrogels via diffusion mechanism only. In this chapter, the recently reported new developments of biodegradable hydrogels for drug control/release are reviewed. These newly developed technologies are largely based on polysaccharides, particularly dextran, and synthetic biodegradable polymers like polylactide. Dextran is one of the most abundant and naturally occurring biodegradable polymers and has been experimentally used for the delivery of pharmaceutically active drugs, peptides, and proteins [34–38] . Dextran consists mainly of (1 → 6) ␣-D-glucoside linkages with about 5–10% (1 → 3) ␣-linked branching. Dextran has chemically active functional groups (i.e., —OH groups) that can be used to provide greater flexibility in the formulation of hydrogels, and dextran is susceptible to enzymatic digestion (dextranase) in the human body [1]. Some examples of these newly developed biodegradable hydrogels include Kim et al.’s maleic anhydride–modified dextran-based hydrogel (Dex-Ma) that has a very unique swelling property that conventional hydrogels don’t have, i.e., increased swelling with an increase in crosslinking units [39,40]. Kim et al. also reported other types of dextran-based hydrogels as drug carriers [41–43]. Zhang et al. reported two types of hybrid biodegradable hydrogels that were prepared from modified dextrans and modified polylactide [45–51]. In addition to those reported new classes of biodegradable hydrogels, there are some emerging advanced technologies that appear to have a great potential as biodegradable hydrogels for biomedical applications. For example, Lang et al. very recently reported the synthesis and characterization of multiarm poly(ε-caprolactone) as the possible precursor for biodegradable hydrogels [52–54], and Wu et al. advanced the concept of such multiarm synthetic biodegradable polymers by combining them with modified dextran to form hybrid multiarm biodegradable hydrogels as potential drug carriers [55].
II. DEXTRAN–MALEIC ACID BIODEGRADABLE HYDROGEL In the synthesis of photocrosslinked hydrogels, the incorporation of photocrosslinkable unsaturated groups (e.g., vinyl groups) into a polymer is the primary requirement for preparing hydrogel precursors. In most cases, the unsaturated groups incorporated have been achieved at the expense of existing hydrophilic groups of a polymer, such as hydroxyl or carboxylic groups, which contribute water solubility and swelling to the polymer. Thus, as the amount of vinyl groups incorporated in a polymer increases, the available hydroxyl or carboxylic functional groups of that polymer decrease, i.e., the hydrophilicity of the resulting polymer and its swelling are adversely affected by the degree of substitution (DS) [41]. For example, poly(␥-glutamic acid) hydrogel exhibited a lower swelling degree, Q (defined as the volume of swollen gel by mL per mg of dried hydrogel), from 0.032 to 0.016 as the concentration of crosslinkers (various dihalogenoalkanes) increased from 2 to 10% [56]. Glycidyl methacrylate–derived dextran hydrogel also showed a reduction of swelling by 70% as DS increased from 4 to 11 [57]. This same relationship (i.e., lower swelling with a higher DS) was also observed by Kim et al. [41] who reported the swelling ratio of dextran hydrogel (acrylic group attached) was decreased by 752% with an increase in DS from 0.45 to 2.99. This classic relationship between swelling and DS of hydrogels is attributed to the fact that more hydrophilic functional groups (e.g., hydroxyl or carboxylic acid groups) of hydrogel precursors are replaced by relatively hydrophobic unsaturated vinyl groups. This consumption of hydrophilic functional groups of hydrogel precursors would increase the hydrophobicity of hydrogels and lower their swelling ratios in water.
426
Chu
Since the structural integrity, mechanical properties, and swelling behavior of hydrogels depend on the degree of crosslinking, which in turn depends on DS, it is desirable to synthesize hydrogels with an adequate DS. Unfortunately, as described above, a higher DS for better structural integrity and mechanical properties of hydrogels is frequently achieved at the expense of the solubility and hydrophilicity of the resulting hydrogel and its precursors. These seemingly mutually exclusive and contradictory requirements (higher swelling ratio vs. better structural integrity and mechanical properties) were recently achieved by Kim et al. in their reported synthesis of dextran–maleic acid (Dex-Ma) biodegradable hydrogels [39]. Kim et al. used maleic anhydride as a carrier for vinyl group segments. Maleic anhydride has been used in coatings, adhesives, detergent additives, cosmetic additives, dispersants, enzyme carriers, and ion exchange resins [58,59]. It is known that polymers having pendant hydroxyl functional groups like dextran can easily react with the carbonyl carbon of maleic anhydride to attach maleic acid groups onto the polymers, as evidented in the Jedlovcnik et al. preparation of a step-growth polymer by reacting maleic anhydride with 1,2-propylene glycol at high temperature (180–210⬚C) [60]. The method of using maleic anhydride as the carrier for photocrosslinkable vinyl groups in hydrogel precursors has several advantages over conventional methods of introducing vinyl groups. First, water or solvent solubility of the resulting hydrogel precursors would be enhanced with an increase in DS as opposed to the decrease in water solubility and swelling ratio of conventional hydrogels that are based on acrylic acid and its derivatives, i.e., an increase in hydrophilicity of maleic acid–based hydrogels would also be expected to increase their swelling ratio. This is because each substitution would open an anhydride ring of maleic anhydride and would result in one free carboxylic acid group available at the substituted chain end. This carboxylic group is able to enhance the solubility of a polymer. Second, functionality of a polymer will be preserved, irrespective of DS. This is because the consumption of one hydroxyl functional group of a polymer used for the attachment of a maleic acid group will be compensated by the generation of one free carboxylic functional group. The newly available carboxylic acid functional group can then be used in any further incorporation of drug or other bioactive agent. Third, hydrophilicity of hydrogels will be increased with degree of substitution due to the increasing incorporation of carboxylic acid. In Kim et al.’s study [39], dextran–maleic acid hydrogel was made via the substitution of the hydroxyl groups of dextran by maleic anhydride. The resulting dextran–maleic acid conjugate was then irradiated by a long wave UV lamp to form a hydrogel. Scheme 1 illustrates this synthesis route. They examined the effects of temperature, reaction time, concentrations of reactants, and catalyst on the DS of maleic acid group in dextran. A range of DS from 0.30 to 1.53 was achieved based on varying reaction temperature, time, and the composition ratio of maleic anhydride to the hydroxyl groups in dextran. The swelling behavior of Dex-Ma was also investigated as a function of pH, from 3 to 7, and a representative one at pH 7 is shown in Fig. 1. At all three pHs, Dex-Ma hydrogels exhibited very high swelling ratios with time, and their magnitudes depended on DS and pH of the media. Maximal swelling of the Dex-Ma hydrogel was achieved in a neutral pH medium, irrespective of DS, followed by an acidic medium, while an alkaline medium had relatively the smallest swelling ratio. Contrary to conventional hydrogels made from acrylic acid or its derivatives, a higher DS of Dex-Ma hydrogels always resulted in a higher swelling ratio over the entire pH range studied. As shown in Fig. 1, the Dex-Ma hydrogel of DS 0.90 showed an abrupt and rapid increase in swelling within 1 h (601%), and 91% of its equilibrium swelling was attained within 3 h (1,069%). An equilibrium swelling was reached after 5 h (1,171%) and maintained without any sign of structural disintegration for a few days. The Dex-Ma hydrogel of higher DS (DS 1.26) reached bulk of its maximal swelling ratio (1,332%) within 1 h, and
Biodegradable Hydrogels as Drug Vehicles
427
Scheme 1 Synthesis of dextran–maleic acid hydrogel from dextran and maleic anhydride. (From Ref. 39.)
its maximal swelling point (1,489%) was reached after 2 h. However, this hydrogel dissolved thereafter. This pH-dependent swelling behavior of Dex-Ma hydrogels was also reported by Liou et al. in their study of poly(vinyl alcohol)-co-maleic anhydride (PVA-MA) hydrogel [61] and it was attributed to the formation of hydrogen bond complexation. As illustrated in Scheme 2, the relatively lower swelling ratio of Dex-Ma hydrogel in an acidic vs. neutral pH was attributed
428
Chu
Figure 1 Swelling ratio of dextran–maleic acid (Dex-Ma) hydrogel in pH 7 buffer at 37⬚C. (From Ref. 39.)
to the hydrogen bond complexation between the free carboxylic acid group in the maleic acid segment and the free hydroxyl group in dextran. These carboxylic acid groups would become ionized in a neutral buffer solution, which would break hydrogen bonds and generate electrostatic repulsion among macromolecules. This repulsive force would push the network chain segments apart and attract more water into the hydrogel, i.e., a higher swelling ratio. At a higher pH, counterions would pair with the ionized carboxylic group and decrease electrostatic repulsion of carboxylic ions. As a result, a lower swelling ratio of Dex-Ma hydrogel was observed at pH 10. The rapid irreversible ester hydrolysis in an alkaline medium could also contribute to the observed lower swelling ratio of Dex-Ma hydrogel in an alkaline medium.
III. DEXTRAN–METHACRYLATE HYDROGEL FOR DOXORUBICIN威 RELEASE Among many anticancer drugs, doxorubicin (DOX) has shown a marked inhibitory effect on the growth of various tumors and a considerable improvement in survival rate of treated animals [62–66]. Doxorubicin, as a single agent, has been reported to give high response rate in advanced
Biodegradable Hydrogels as Drug Vehicles
429
Scheme 2 Molecular modeling of dextran–maleic acid hydrogel in various pH media. (From Ref. 39.)
breast cancer and favorable result in gastric carcinoma, a tumor on which only four drugs are known to be active [62]. However, doxorubicin has a narrow therapeutic index and hence is more prone to dose-dependent toxicity [67,68]. The major side effect in doxorubicin therapy is the development of severe cumulative dose-dependent cardiomyopathy, leading to congestive heart failure which is fatal in the majority of cases [62]. Total cumulative dose limitations of doxorubicin is at ⬃ 450–550 mg/m2 [62,67]. The effort to overcome this dose-limiting toxicity has been done in two directions. One is combination therapy with other chemotherapeutic agents [62]. The other is to use a drug delivery matrix to control the release of the drug influx to the body [69–71]. Human and animal erythrocytes as well as crosslinked albumin have been proposed as carrier vehicles for cytotoxic drugs for their slow release [69,70,72]; however, the rapid diffusion of doxorubicin from erythrocytes to plasma remained a major problem [70]. Treatment of the erythrocytes carrier with glutaraldehyde was done by Tonetti et al. to prevent the rapid efflux of doxorubicin [69,70]. They achieved a slower release of doxorubicin, but potential toxicity of glutaraldehyde and possible alteration of doxorubicin during glutaraldehyde treatment still remain as problems. Kim et al. [44] reported the use of a biocompatible polymer matrix, such as dextran–methacrylate as the carrier for doxorubicin, and its in vitro release profiles appear to provide a better alternative to minimize drug toxicity than the glutaraldehyde-treated erythrocyte carrier. Scheme 3 illustrates the synthesis of dextran–methacrylate hydrogel. As shown in Fig. 2, DOX-loaded dextran–methacrylate hydrogels showed an initial rapid release of DOX within 4–6 h after immersion in a buffer solution followed by a very slow steady release thereafter. The amounts of cumulative release of DOX within the initial 24 h ranged from as high as 90% to as low as 23%, depending on the degree of substitution of dextran–methacrylate hydrogels and the pH of the media. It appears that DS was the most critical factor to determine the amounts and rate of DOX release, irrespective of the pH of the media. The dextran–methacrylate hydrogel with the highest DS (DS 0.60) showed not only the slowest rate of release during the initial 24 h but also had the lowest cumulative release (40%) at the end of the release study period (i.e., 240 h) in pH 3 medium. The dextran–methacrylate hydrogel with a medium DS (DS 0.24) showed a cumulative
430
Chu
Scheme 3 Synthesis of dextran–methacrylate hydrogel from dextran. (From Ref. 41.)
release of 67% within the same study period and pH medium, while the lowest DS dextran–methacrylate hydrogel (DS ⳱ 0.09) had a cumulative release as high as 99% and reached the majority of its maximal cumulative release within 3–5 h. The same release profiles were observed in pH 7.4 medium. Although the DOX-loaded dextran-methacrylate hydrogels having lower DS appeared to have reached equilibrium release at the end of 24 h, the highest DS hydrogel showed continuous slow release of DOX beyond 240 h. The DOX-loaded dextran–methacrylate hydrogels showed a less profound pH-dependent release than DS. A higher release of DOX was observed in an acidic medium than the physiological pH, except for the dextran–methacrylate hydrogel with the highest DS, i.e., 0.60. The pH effect became far less pronounced in DS 0.09 and 0.60 dextran–methacrylate hydrogels. The DS 0.09 Hydrogel showed only 7% cumulative release difference of DOX between the pH 3 and 7.4 media. Dextran–methacrylate hydrogel of DS 0.60 virtually showed no pH effect (a cumulative ⬃ 40–41% of DOX released in both pH 3 and 7.4 media at the end of 240 h).
Biodegradable Hydrogels as Drug Vehicles
431
Figure 2 In vitro doxorubicin (DOX) release from dextran–methacrylate hydrogel in three buffers pH (pH 3, 7.4, and 10) buffers at 37⬚C. (From Ref. 44.)
The release profile of DOX from dextran–methacrylate hydrogels during the initial 24 h was found to be very closely related to the swelling behavior at the same period, as shown in Fig. 3, in which the release profiles of DOX from dextran–methacrylate hydrogel (DS 0.24) and its swelling ratio are plotted together. Similar shapes between the swelling and release profiles were observed. The swelling ratios of dextran–methacrylate hydrogel reached its equilibrium swelling in less than 5 h [43]. During the same initial period, there was a rapid release of DOX from dextran–methacrylate hydrogel, although the swelling profiles reached equilibrium faster than DOX release profiles. This observation supports the hypothesis that the rate of swelling of a hydrogel would control its rate of release of the impregnated drugs, particularly during the early stage. The cumulative release of DOX from all dextran–methacrylate hydrogels were found to be linearly proportional to the square root of time (t1/2) during the very early stage of release (5 h), as shown in Fig. 4. This linear relationship between release amount (Mt/Mo) and t1/2 indicates that DOX release from dextran–methacrylate hydrogel followed the simple Fickian diffusion controlled release mechanism in the early stage of release [73]. The DS 0.24 hydrogel showed the highest fit for linear relationship in both pH (R2 ⬎ 0.98), followed by DS 0.09 hydrogel (R2 ⳱ 0.89 in pH 3 and R2 ⳱ 0.65 in pH 7.4) and DS 0.60 hydrogel (R2 ⳱ 0.85 in pH 3 and R2 ⳱ 0.69 in pH 7.4). The diffusion coefficients (mm2/h) of DOX from dextran–meth-
432
Chu
Figure 3 Comparison between swelling and release behavior of doxorubicin from dextran–methacrylate hydrogel (DS 0.24). (From Ref. 44.)
acrylate hydrogels also depended on the DS of hydrogels and decreased as DS increased, e.g., 0.224 at DS 0.09 to 0.0377 at DS 0.24 and 0.0024 at DS 0.60. Surface and interior SEM morphological structures of dextran–methacrylate hydrogel in swollen and unswollen states are shown in Figs 5 and 6, respectively. The unswollen dextranmethacrylate hydrogel (Fig. 5A) showed a relatively smooth surface with many tiny random cracks without any porous feature, while the swollen hydrogel (Fig. 5B) exhibited a rugged and porous structure. The interior of the unswollen hydrogel (Fig. 6A) was very similar to its surface morphology and did not visually exhibit any porous structure. Upon swelling, however, the interior of the swollen dextran–methacrylate hydrogel (Fig. 6B) exhibited a highly porous honeycomb-like structure. The pores assumed either square, round, rectangular, or irregular shape with their long axis less than 5–10 m. The thickness of the pore wall was less than 0.01 m. The morphology and size of the pore structure observed were expected to be the maximal limit that the dextran–methacrylate hydrogel could reach under the specific swelling condition. This is because all dextran–methacrylate hydrogels reached their equilibrium swelling within 24 h. Different pore sizes and morphologies due to different DS in dextran–methacrylate hydrogels were also observed. The DS 0.24 dextran–methacrylate hydrogel showed larger pores
Biodegradable Hydrogels as Drug Vehicles
433
Figure 4 Cumulative release of doxorubicin as a function of square root of time (t1/2) from dextran–methacrylate hydrogels of various degrees of substitution (DS 0.09, 0.24, and 0.60) in pH 3.0 and 7.4 media. (From Ref. 44.)
and more porous structure than the one with a higher DS, e.g., the DS 0.60 hydrogel which showed ˚ ) even at a very high magnification (60,000⳯). Generally, pore size very tiny pores (50–1000 A decreased with an increase in DS of dextran–methacrylate hydrogels. Table 1 illustrates such a relationship between average pore characteristics of swollen dextran–methacrylate hydrogels and DS [42]. As shown in the table, both total pore area and average pore diameter (obtained from mercury intrusion porosimetry) of dextran–methacrylate hydrogels decreased with an increase in DS of methacrylate. A reduction of total pore area by 12% was observed as the DS increased from 0.09 to 0.60. The reduction in average pore diameter with an increase in DS was far more profound than the total pore area. For example, there was a near 52% reduction in average pore diameter as the DS of dextran–methacrylate hydrogel increased from 0.09 to 0.60. Bulk density also increased with an increase in DS of dextran–methacrylate hydrogel because a lower porosity (higher bulk density) would reduce the volume of hydrogels at a constant mass. The DS 0.60 hydrogel had the highest bulk density (0.8014 g/mL), followed by DS 0.24 hydrogel (0.5624 g/mL) and DS 0.09 hydrogel (0.5133 g/mL). The pore size distribution in dextran–methacrylate hydrogels showed a bimodal character, irrespective of their DS; one region with large pores, from 10 to 100 m, and the other with small pores from 0.01 to 0.1 m. The relative proportion
434
Chu
Figure 5 Surface morphology of dextran–methacrylate hydrogels: (A) unswollen and (B) swollen. (From Ref. 42.)
Biodegradable Hydrogels as Drug Vehicles
435
Figure 6 SEM photographs of the interior dextran–methacrylate hydrogels: (A) unswollen and (B) swollen. (From Ref. 42.)
436
Chu
Table 1 Average Pore Characteristics of Swollen Dextran–Methacrylate Hydrogels of Different Degrees of Methacrylate Substitution (DS) Measured by Mercury Intrusion Porosimetry
Total pore area (m2/g) Average pore diameter (4V/A) (m) Bulk density (g/mL)
DS 0.09
DS 0.24
DS 0.60
105 0.0514 0.5133
97 0.0280 0.5624
92 0.0243 0.8014
of macropores to micropores depended on the DS of dextran–methacrylate hydrogels. Both DS 0.09 and 0.24 dextran–methacrylate hydrogels had more macropores than micropores. The DS 0.60 dextran–methacrylate hydrogel, however, had more micropores than macropores. These quantitative findings were consistent with the qualitative SEM observations.
IV. DEXTRAN–POLYLACTIDE HYBRID HYDROGELS AND THEIR PROPERTIES One promising approach to modify the swelling property of hydrogels is to form a homogeneousstructured hydrogel consisting of both biodegradable hydrophilic and hydrophobic segments. The hydrophilic component would ensure swelling of the system in an aqueous medium, while the presence of the hydrophobic part would regulate the swelling property and mechanical strength of the system as well as the proper dispersion of hydrophobic drugs [74–76]. It is difficult to homogeneously disperse hydrophobic drugs within a totally hydrophilic hydrogel to achieve predictable drug release profiles. In certain applications, hydrogels with a balanced hydrophilicity to hydrophobicity are preferred over highly hydrophilic hydrogels [77]. The increasingly available new therapeutic proteins, peptides, and oligonucleotides that are mainly of a hydrophobic nature further demonstrate the need of a new class of hydrogels having a controlled wide range of hydrophobicity to hydrophilicity. Toward the goal of achieving better hydrogels having both hydrophilic and hydrophobic characteristics, there were a few reported studies of the heterogeneous-structured drug delivery hydrogel systems. In those reported studies, the hydrophilic polymers used were mostly based on synthetic polymers, like polyethylene glycol (PEG). The methods used for making hydrogels having both hydrophilic and hydrophobic segments were either the synthesis of copolymers from appropriate monomers or physically blending one polymer into another polymer network to make a semi-interpenetrated polymer network (semi-IPN) hydrogel. An IPN is defined as a network composed of two separately chemically crosslinked polymers [78]. The crosslinked nature of the IPN would not only enhance the compatibility of polymer components and prevent phase separation because the micro domains within the IPN structured polymers are less pronounced (i.e., more homogeneous) when compared with that of polymer blends, but also allow access to those properties that might be hybrids of the polymer components. Lee and Kim [74] reported that semi-IPNs composed of physically blending -chitin and PEG-co-poly(D,L-lactide) macromer showed an improved tensile strength at a high equilibrium water content. Sawhney et al. [79]used a copolymer of polylactide/PEG hydrogel as a biodegradable drug delivery system. However, the hydrophobic character of the PLA/PEG hydrogel was not adequate because its hydrophobic segments (PLA) were too low molecular weight (i.e., oligomers). The properties of this PLA/ PEG hydrogel were determined primarily by the water-soluble PEG segments. Huang and Onyari reported that the hydrogels based on polylactide and poly(2-hydroxyethyl methacrylate) could
Biodegradable Hydrogels as Drug Vehicles
437
be amorphous or semicrystalline, depending on the stereochemical composition, and the degradation of poly(LA-HEMA) copolymer could be controlled and tailored for target biomedical applications [80]. Very recently, Zhang et al. reported a totally biodegradable homogeneous–structured hydrogel with a controlled wide range of swelling and drug control/release properties [45–51]. Zhang et al. developed biodegradable hybrid hydrogels from two modified dextrans and one modified poly(D,L-lactic acid) (PDLLA). The purpose of the chemical modification of both dextran and PDLLA is to introduce photocrosslinkable groups into them for subsequent photoinduced hydrogel formation. PDLLA serves as the hydrophobic segment, and modified dextran services as the hydrophilic segment. PDLLA has been widely used as a biodegradable hydrophobic polymer for structural support and as a hydrophobic drug carrier due to its combination of biodegradability, biocompatibility, and good mechanical strength [81–83]. The in vitro and in vivo degradation of PDLLA is well understood, and the degradation products are natural metabolites that can readily be eliminated by the human body. A. Chemical Structure The two modified dextrans are dextran acrylate (Dex-AC) and dextran allyl isocyanate (Dex-AI), their chemical structures are shown in Scheme 4. In Dex-AC precursor, the photocrosslinkable unsaturated groups (i.e., vinyl groups) are attached to the anhydroglucose ring via ester bonds, while the photocrosslinkable unsaturated groups in Dex-AI are attached to the anhydroglucose ring via urethane linkages. Due to the difference in hydrolytic sensitivity between ester and urethane bonds, the Dex-AI–derived biodegradable hydrogels would be more resistant to hydrolytic degradation than the ones derived from Dex-AC. Scheme 4 also shows the modified PDLLA, PDLLA macromer (PDLLAM). A comparison between PDLLA macromer and DexAC (or Dex-AI) precursors indicates that the photocrosslinkable groups in the PDLLA macromer precursor are located at the two chain ends, while the Dex-AC and Dex-AI precursors have the photocrosslinkable groups along the anhydroglucose backbone, i.e., there are far more photocrosslinkable groups in Dex-AC and Dex-AI than PDLLA macromer. This difference in the location of photocrosslinkable groups would certainly predispose to different molecular weight effects of the precursors on their hydrogel swelling and hence drug control/release properties. The representative chemical structures of the Dex-AC/PDLLA and Dex-AI/PDLLA hybrid biodegradable hydrogels are shown in Scheme 5. B. Swelling Properties The swelling behaviors of Dex-AC/PDLLA hybrid hydrogels in pH 3, 7, and 10 buffer solutions at 37⬚C are shown in Fig. 7. Generally, the swelling behavior of the hybrid hydrogels depended on both the Dex-AC to PDLLA ratio and the pH of the media. In three different pH media from acid to alkaline, the 100% Dex-AC hydrogel always had the largest swelling ratio, while the 100% PDLLA one had the smallest. The swelling ratio of the 100% Dex-AC hydrogel decreased gradually with an increasing incorporation of PDLLA component in the hybrid hydrogels. Within the same hydrogel, its swelling behavior depended on the pH of the media. 100% Dex-AC and 100% PDLLA hydrogels had their largest swelling ratio in pH 10 buffer, followed by pH 7 and 3. For Dex-AC/PDLLA hybrid hydrogels, the lowest swelling ratio occurred in pH 7 medium. The incorporation of PDLLA into Dex-AC appeared to improve the mechanical property of the hydrogel because Dex-AC/PDLLA hybrid hydrogels retained structural integrity throughout the swelling procedure. The 100% Dex-AC hydrogel, however, started to break into fragments at about 10 h and was difficult to handle, especially in a pH 10 buffer solution.
438
Chu
Scheme 4 Chemical structure of the modified dextran and poly(D,L)-lactide (PDLLA) precursors: dextran acrylate (Dex-AC), dextran allyl isocyanate (Dex-AI), PDLLA, and PDLLA macromer (PDLLAM).
Swelling of hydrogels is similar to dissolution. The rate of swelling is governed by the solvation power and the diffusion rate of water molecules. The parameters that influenced the swelling ratio of hydrogels were examined by several investigators [6,84–87]. These parameters are hydrophilic and hydrophobic properties of the constituent polymers, hydrogen bond capability between polymers, and the pH-dependent hydrolysis rate of ester groups located in crosslinkers as well as polymer backbones. All these factors are affected by molecular weight, pH, and temperature. The pH effect on the swelling of Dex-AC/PDLLA hybrid hydrogels could be explained by the association and disassociation of hydrogen bonds between nitrogen, hydrogen, and oxygen atoms. The association and disassociation of hydrogen bonding could partially contribute to the different swelling ratios in different pH buffers. In alkaline media, the hydroxyl group was deprotonated. This would lead to the disruption of hydrogen bonding, a more open structure and a higher swelling ratio. The effect of pH on the swelling ratio was more significant for
Biodegradable Hydrogels as Drug Vehicles
439
Scheme 5 Representative chemical structures of the Dex-AC/PDLLAM and Dex-AI/PDLLA hybrid biodegradable hydrogels.
those hybrid hydrogels having a dominant Dex-AC component, which is understandable, because Dex-AC has many free hydroxyl groups that could be deprotonated in an alkaline medium. For the 100% Dex-AC hydrogel at a lower pH medium (pH 3 and 7), Dex-AC chains assumed a more compact structure because of a smaller degree of polymer–solvent interaction and a higher degree of hydrogen bonding among macromolecules, i.e., lower swelling ratio. In a pH 10 buffer solution, the hydrogen bonding capability would be reduced, while Dex-AC and water interaction increases. This would lead to a more open structure and higher swelling ratio. In addition, hydrolysis of the ester groups in the crosslinkers became so significant in a high
440
Chu
Scheme 5 Continued.
alkaline medium that the Dex-AC was totally dissolved after 3 days. This observed swelling behavior of the 100% Dex-AC hydrogel in three different buffer solutions (pH 3, 7, 10) was similar to the dextran-derived hydrogels in Dijk-Wolthuis et al.’s study [37] and in Zhang et al.’s study [45]. Yeh et al.’s pH-sensitive hydrogel synthesized by crosslinking of N,N-dimethyl acrylamide copolymer precursors [88] also showed similar swelling behavior in different pH buffer solutions, i.e., swelling ratio increased with an increase in pH value. In contrast to the hydrophilic 100% Dex-AC hydrogel, which prefers to interact with water, the 100% PDLLA hydrogel shows a fraction of the swelling of the 100% Dex-AC hydrogel. This is attributed to the hydrophobic nature of PDLLA polymers. It is known that hydrophobic
Biodegradable Hydrogels as Drug Vehicles
441
Figure 7 Swelling ratio of Dex-AC/PDLLA hybrid hydrogels as a function of Dex-AC to PDLLA composition and pH and after 11.5 h in phosphate buffer of pH 7.4 at 25⬚C. (From Ref. 46.)
polymers exhibit strong attraction among the hydrophobic segments [89]. This hydrophobic interaction is believed to play a central role in the observed very low swelling ratio of 100% PDLLA hydrogel. In addition, the degree of structural openness of the PDLLA depended on the pH of the buffer solution; a more open PDLLA structure would be expected in an alkaline pH medium and hence a higher swelling ratio. This is because intermolecular hydrogen bonds in PDLLA were retained in acidic and neutral media, but destroyed in an alkaline medium. PDLLA and its crosslinker would also degrade faster in an alkaline medium because the rates of hydrolysis of aliphatic polyesters like polyglycolide and its lactide copolymers have been known to be faster in a base medium [83]. This faster rate of hydrolytic degradation of ester linkages in both PDLLA backbone and its crosslinker would certainly facilitate a more open network structure, i.e., higher swelling ratio in alkaline media. In the case of Dex-AC/PDLLA hybrid hydrogels, their swelling behavior depended on the relative composition of Dex-AC to PDLLA and their pH dependence of structural openness. As PDLLA component increased, the hydrophobicity of the hydrogel increased, a more compact structure and subsequently a lower swelling ratio was obtained. For example, after 11.5 h immersion in a pH 7 buffer medium, 100% Dex-AC hydrogel had a swelling ratio of 540%. The incorporation of PDLLA macromer into the Dex-AC hydrogel decreased its swelling ratio to 260% for 80/20 Dex-AC/PDLLA hydrogel, 98% for 50/50 Dex-AC/PDLLA hydrogel and 30% for 100% PDLLA hydrogel. The impact of incorporating PDLLA component on the pH-dependent swelling ratio became less apparent as the PDLLA component increased to 50% of the total composition. The swelling behavior of Dex-AI/PDLLA hybrid hydrogels, however, is different from Dex-AC/PDLLA, chiefly due to the difference in dextran derivatives in the hybrid system. Figure 8 shows the different level of swelling of 100% Dex-AI vs. 100% Dex-AC at three different pH levels [45]. Dex-AC has a higher swelling ratio than Dex-AI over the pH range of 3–10, particularly at an alkaline pH. Although the degree of substitution in dextran by AC or AI may be one of the factors for such a difference in swelling, Dex-AC has an ester linkage in the crosslinker, while Dex-AI has a urethane linkage, which is more hydrolytically resistant than an ester linkage. At an alkaline environment, base catalyzed hydrolysis of the ester linkage
442
Chu
Figure 8 Swelling ratios of 100% Dex-AC (DS 3.70) and 100% Dex-AI (Dex-AI-3 of DS 5.03 and DexAI-5 of DS 8.36) hydrogels after 23 h immersion in pH 3, 7, and 10 buffer solutions. (From Ref. 45.)
in the Dex-AC hydrogel may lead to a more open network structure than Dex-AI, and hence Dex-AC would exhibit larger swelling than Dex-AI, as observed. This difference in swelling ratio between Dex-AI and Dex-AC is also reflected in the Dex-AI/PDLLA hybrid hydrogels as shown in Fig. 9. The data show the swelling ratios of the Dex-AI/PDLLA hydrogels with different Dex-AI/PDLLA composition ratios (100/0, 80/20, 60/ 40, 40/60, 20/80, and 0/100) in pH 3, 7, and 10 buffer solutions after a 23-h immersion. The Dex-AI component of the hydrogel system had a DS of 8.36. All hydrogels were crosslinked under UV for 3 h and tested after extraction with both water and acetone. Except for the 100% Dex-AI hydrogel, none of the Dex-AI/PDLLA hydrogels reached an equilibrium swelling ratio at the end of the study period (23 h). Although the magnitude of the swelling profiles varied with pH of the medium, the overall swelling behaviors in the same buffer solution were generally similar; that is, the swelling ratio decreased as the PDLLA component increased. The swelling ratio of the Dex-AI/PDLLA hydrogels depended on the hydrophilicity of the Dex-AI component and the hydrophobicity of the PDLLA component. As the PDLLA component increased, the hydrophobicity of the hydrogel increased because of the hydrophobic attraction among PDLLA segments, which led to a more compact structure and a suppressed swelling ratio, as observed. The composition dependence of the swelling behavior of the Dex-AI/PDLLA hydrogels in buffer solutions was generally consistent with other studies [84–88]. For example, Inoue et al. [87] studied the swelling behavior of a hydrophobically modified hydrogel by grafting methyl methacrylate to the backbone of poly(acrylic acid). The swelling ratios of Inoue et al.’s hydrogels decreased as the graft level of methyl methacrylate (the hydrophobic part) increased. The consistently higher swelling ratios of the Dex-AI/PDLLA hydrogels in an alkaline pH medium compared with those of acidic and neutral pH media could be partially attributed to the higher rate of hydrolytic degradation of ester groups in the PDLLA backbone in an alkaline medium. This alkaline-accelerated hydrolysis of aliphatic polyesters was also reported by Chu in his study of the effect of pH on the hydrolytic degradation of polyglycolide and poly(glycolideco-lactide) [90,91]. Similar findings were also reported by Karmalkar et al. in their study of the ester hydrolysis rate of p-nitro pendent groups from 2-hydroxyethyl methacrylate [92]. This faster hydrolysis of ester groups in the PDLLA segment would result in a more open hydrogel network structure in an alkaline medium, which would lead to a higher swelling ratio than those in acidic and neutral media. Zhang et al.’s preliminary hydrolytic degradation study of the DexAI/PDLLA hydrogels through the monitoring of their weight changes as a function of time in
Biodegradable Hydrogels as Drug Vehicles
443
Figure 9 Effect of Dex-AI/PDLLA composition ratio on the swelling ratios of Dex-AI/PDLLA hydrogels after 23 h in pH 3, 7, and 10 buffer solutions (Dex-AI-5 of DS 8.36, UV 3h after extraction). (From Ref. 45.)
a pH 7.4 buffer showed that hydrolytic degradation did occur; that is, the weights of the DexAI/PDLLA hydrogels decreased as the incubation time increased. Like Dex-AC/PDLLA hydrogels, the parameters that influenced the swelling ratio of the Dex-AI/PDLLA hydrogels in different pH media include hydrophilic and hydrophobic properties of the constituent polymers and the pH-dependent hydrolysis rate of ester linkages located in the PDLLA backbones.
C. Morphological Structure The three-dimensional porous network structures of these hybrid hydrogels are shown in Figs. 10–12 [48]. In all hybrid hydrogels of the Dex-AI/PDLLA systems, the unswollen ones (0 day) show smooth solid morphology without visible pores (Figs. 10A, 11A, and 12A). This common feature, however, disappears as the hybrid hydrogels start swelling and the 3 D pore structure and shape depend on the composition ratio of Dex-AI to PDLLA and the extent of swelling. The 3D pores in the hybrid hydrogels having higher Dex-AI amount exhibit far more regular cylindrical pore structure than those hybrid hydrogels having higher PDLLA component. A comparison between Fig. 10C (100% Dex-AI at 30 days) and Fig. 11C (100% PDLLA at 30 days) demonstrates this composition-dependent pore structure upon swelling. Since PDLLA can be degraded simply via water-based hydrolysis, the pore formation in the 100% PDLLA hydrogel is mainly due to the hydrolytic degradation of the PDLLA mass. On the other hand, Dex-AI or Dex-AC hydrogels cannott be degraded by water alone; their pore formation is mainly attributed to the swelling of the hydrogels rather than mass destruction. Thus, a combination of Dex-AI with PDLLA as in the case of hybrid hydrogels would exhibit their morphology between 100% Dex-AI and 100% PDLLA. A comparison of 2-day morphology among the three types of hybrid hydrogels (Figs. 10B, 11B, and 12B) illustrates this point of view. The 100% Dex-AI shows the most 3D porous structure development followed by 50/50 Dex-AI/PDLLA, and 100% PDLLA has no visible pore structure at all at the end of 2 days.
444
Chu
Figure 10 SEM photographs of 100/0 Dex-AI/PDLLA hydrogels after (A) 0 days; (B) 2 days; (C) 30 days; (D) 60 days; (E) 120 days immersion in pH 7.4 buffer solution (5kX). (From Ref. 48.)
Biodegradable Hydrogels as Drug Vehicles
445
Figure 11 SEM photographs of 0/100 Dex-AI/PDLLAM hydrogels after (A) 0 days, (B) 2 days; (C) 10 days; (D) 20 days; and (E) 30 days immersion in pH 7.4 buffer solution (5kX). (From Ref. 48.)
446
Chu
Figure 12 SEM photographs of 50/50 Dex-AI/PDLLA hydrogels after (A) 0 days; (B) 2 days; (C) 30 days; (D) 60 days; and (E) 120 days immersion in pH 7.4 buffer solution (5kX). (From Ref. 48.)
Biodegradable Hydrogels as Drug Vehicles
447
Table 2 Thermal Properties of the Dex-AI/PDLLA Hybrid Hydrogels and Their Components 100/0 Samples Undried dextran Dried dextran, PDLLA, or dextran/PDLLA mixture Dried Dex-AI, PDLLAM, or Dex-AI/PDLLAM mixture Dried 0-day Dex-AI/PDLLAM hydrogels Dried and hydrolyzed Dex-AI/PDLLAM hydrogels a b
50/50
0/100
Tg
Tw
T1g
T2g
Tg
Tm
222.9 222.7
116.6 —
— 15.9
— 197.2
— 9.6
— 152.1
172.5
—
21.3
138.4
12.4
159.3
179.0
—
38.1
—
16.2
—
173.3a
—
—
170.5a
7.3b
—
60-day incubation. 20-day incubation.
D. Thermal Properties The incorporation of photocrosslinkable units, such as acrylate (AC) or allyl isocyanate (AI), into dextran and the hydrogel network formation with PDLLA macromer, as expected were found to alter their thermal and mechanical properties of the hydrogel precursors and hydrogels (Tables 2 and 3). Generally, either the incorporation of AI onto dextran backbone (Dex-AI synthesis) or the hydrolytic degradation of the hydrogels reduced their corresponding glass transition temperature (Tg) values. However, the incorporation of acrylate units onto the two chain ends of PDLLA (PDLLA macromer, PDLLAM) and the photocrosslinking of the individual hydrogel precursors to form hydrogels (i.e., 100% Dex-AI or 100% PDLLA hydrogel) increased Tg. For example, the dried dextran had a Tg at 222.7⬚C, and this Tg was shifted to a lower temperature (172.5⬚C) after the incorporation of AI onto dextran backbone. These thermal data appeared to confirm Kakizaki et al.’s findings [94] that dextran is an amorphous biopolymer and its Tg is higher than 150⬚C. The strong hydrogen bonding among the dextran macromolecules is believed to be the cause of its high Tg. The reason that the Tg of Dex-AI was lower than dextran is due to the replacement of —OH group in dextran by an AI group that would reduce the number of hydroxyl groups available for hydrogen bonding and increase the distance between the macromolecules due to the bulky AI group. As a result, a smaller degree of intermolecular hydrogen bonding would be formed, i.e., lower Tg. Upon photocrosslinking of Dex-AI into a hydrogel, the Tg of the 100% Dex-AI hydrogel (179.0⬚C) was higher than that of the Dex-AI precursor (172.5⬚C). This is because crosslinking reduced the segmental movement of Dex-AI and hence increased its Tg. Table 3 Initial Compression Moduli (MPa) of Dex-AI/PDLLA Hybrid Hydrogels as a Function of in Vitro Degradation Time and Composition Ratio Time (days) 0 2 30 60 a
100/0
80/20
50/50
20/80
0/100
7.82 ⫾ 0.71 0.19 ⫾ 0.02 0.10 ⫾ 0.06 0.09 ⫾ 0.02
6.01 ⫾ 0.36 0.45 ⫾ 0.05 0.31 ⫾ 0.08 0.07 ⫾ 0.01
2.56 ⫾ 0.26 0.51 ⫾ 0.07 0.27 ⫾ 0.02 0.04 ⫾ 0.02
1.39 ⫾ 0.19 0.77 ⫾ 0.12 0.12 ⫾ 0.02
1.17 ⫾ 0.05 0.91 ⫾ 0.08 0.05 ⫾ 0.01
a
a
Completely degraded.
448
Chu
After hydrolytic degradation of the 100% Dex-AI hydrogel in buffer solutions at 37⬚C for 2 months, its Tg was shifted to a lower value (173.3⬚C) from that of the initial 0-day 100% Dex-AI hydrogel (179⬚C). Since the dextran polymer backbone is not hydrolytically degradable and the urethane linkage in the crosslinker is relatively stable in a buffer solution, the 100% Dex-AI hydrogel was structurally stable in buffers and its Tg did not decrease significantly with the duration of immersion (3% reduction over 60 days). The slight reduction in Tg after a 2month incubation might come from the decrease in crosslinking density of the hydrogel by the osmotic force due to the high water content in the hydrogel, which was found to be 67% at 69day incubation. The Tg of the PDLLA and PDLLA macromer were 9.6 and 12.4(C, respectively. Photocrosslinking of PDLLA macromer precursor into hydrogel form increased its Tg to 16.2(C, while the 20-day degradation of the 100% PDLLA macromer–based hydrogel lowered its Tg to 7.3⬚C. The crosslinking of Dex-AI with PDLLA macromer at a composition ratio of 50/50 resulted in only one Tg (38.1⬚C), indicating that no phase separation occurred in the 50/50 Dex-AI/PDLLA hydrogel. After the hydrolytic degradation of this hydrogel in PBS for 2 months, its Tg appeared around 170.5⬚C. A comparison of the thermal data of the 0-day 50/50 Dex-AI/PDLLA hydrogel with 0-day uncrosslinked 50/50 mixture of Dex-AI and PDLLA indicate that the uncrosslinked physical mixture of Dex-AI and PDLLA macromer had a melting peak at 138.4⬚C, while the corresponding crosslinked 50/50 Dex-AI/PDLLA hydrogel did not. The Tg observed in this uncrosslinked 50/50 Dex-AI/PDLLA mixture (21.3⬚C) was lower than the crosslinked 50/50 Dex-AI/PDLLA hydrogel and was suspected from the contribution of the PDLLA macromer component. Similar to the Tg data of the 100% Dex-AI hydrogel, the 100% PDLLA macromer hydrogel had a higher Tg (16.2(C) than the PDLLA macromer precursor (12.4⬚C) due to the reduction in free volume from crosslinking. The disappearance of the crystalline nature of the PDLLA macromer upon crosslinking shows that crosslinking reaction would prohibit crystallization. Because network structure increases the constraints of segmental mobility imposed by crosslinked PDLLA macromer, this phenomenon renders the system with no freedom for spatial rearrangement, i.e., no crystallinity and no Tm observed in the 100% PDLLA macromer–based hydrogel. Hydrolytic degradation of the 100% PDLLA macromer–based hydrogel reduced its Tg from 16.2⬚C at day 0 to 7.3⬚C at 20 days of immersion in a buffer solution, a 50% reduction over a period of 20 days. This behavior might be attributed mainly to the random chain scission of the ester linkages in the PDLLA macromer backbone as well as the reduction in the crosslinking density of the network structure. For example, due to hydrolytic degradation, 38% of the original weight of the 100% PDLLA macromer–based hydrogel was lost during a 20-day incubation period [50]. This is because the reduction in MW of polymer and crosslinking density would result in a lower Tg. The decrease in Tg as a result of hydrolytic degradation of poly-Llactic acid (PLLA) was also reported by Migliaresi et al. [95]. They found that the Tg of high MW PLLA decreased gradually from 70.6 to 46.7⬚C over a period of 779 days incubation in Ringer’s solution at 37⬚C. During this incubation period, the MW of the PLLA decreased from 103,000 to less than 4900. In the 50/50 Dex-AI/PDLLA hybrid hydrogel, there was only one Tg (38.1⬚C) observed. This clearly demonstrated that no phase separation occurred in this hybrid hydrogel. The calculated Tg of a totally miscible 50/50 Dex-AI/PDLLA blend using the following empirical equation is 29.7⬚C [96]:
1 W W = 1 + 2 Tg Tg1 Tg2
(1)
Biodegradable Hydrogels as Drug Vehicles
449
where w1 ⳱ w2 ⳱ 0.5 for 50/50 Dex-AI/PDLLA hydrogel, Tg1 ⳱ 179.0, Tg2 ⳱ 16.2. The experimentally observed Tg of 50/50 hydrogel (38.1⬚C) was 22% higher than the calculated 50/ 50 blend (29.7⬚C). This difference between experimental and calculated Tg must be attributed to the crosslinking nature in the 50/50 hydrogel that the 50/50 blend did not have, i.e., a crosslinked network structure would increase Tg of the miscible blends. For example, Eschbach and Hvang [97] synthesized a hydrophilic–hydrophobic binary system of poly (2-hydroxyethyl methacrylate) (PHEMA) and polycaprolactone (PCL). A crosslinked interpenetrating polymer network of PHEMA/PCL (90/10 w/w) had Tg-95⬚C that was 16⬚C higher than the Tg of the PHEMA/PCL (90/10 w/w) copolymer. Although the hydrolytic degradation of polymers is generally believed to decrease their Tg, the Tg of the 60-day hydrolyzed 50/50 Dex-AI/PDLLA hydrogel (170.5⬚C) was found to be 348% higher than the unhydrolyzed 50/50 Dex-AI/PDLLA hydrogel (38.1⬚C). This phenomenon can be explained as follows. Based on Zhang et al.’s studies of the degradation property of Dex-AI/PDLLA hydrogel [48,49], the 100% PDLLA macromer–based hydrogel was found to totally disappear in PBS at about 50 days. Therefore, after 2 months incubation in PBS, the PDLLA macromer component of the 50/50 Dex-AI/PDLLA hybrid hydrogel would degrade completely and only the Dex-AI component remained. Thus, the Tg of the 60-day hydrolyzed 50/50 Dex-AI/PDLLA hydrogel (170.5⬚C) reflected the glass transition of the 60-day hydrolyzed Dex-AI component, which had a Tg of 173.3⬚C, a negligible 1.6% difference. The broadening of the Tg of the 60-day hydrolyzed 50/50 Dex-AI/PDLLA hydrogel also indicated that the polydispersity of Dex-AI increased, which might come from the chain cleavage of Dex-AI. E. Mechanical Properties As shown in Table 3, the 100% 0-day Dex-AI hydrogel had the highest compression modulus (7.82 MPa) among all Dex-AI/PDLLA hydrogels. In this 0-day group, the moduli of the Dex-AI/ PDLLA hydrogels decreased gradually as the PDLLA macromer composition in the hydrogels increased, and the 100% PDLLA macromer–based hydrogel had the lowest modulus (1.17 MPa). This relationship might be attributed to the relative magnitudes of the Tg of the hydrogel precursors to the temperature at which the mechanical property was tested. The modulus of the 100% Dex-AI hydrogel tested in this study reflected its mechanical strength in a glassy state, i.e. the temperature of compression testing (25⬚C) was far below the Tg of the 100% Dex-AI hydrogel (179⬚C), but was far above the Tg of the PDLLA macromer (12.4⬚C). Therefore, the incorporation of the PDLLA macromer component into the Dex-AI hydrogel would reduce the Tg of the 100% Dex-AI hydrogel and shift the Tg of the hybrid hydrogels closer to the testing temperature for mechanical property, and hence reduce the moduli of the hybrid hydrogels. The level of reduction in modulus depended on the composition ratio and became larger with an increase in the PDLLA macromer composition. Upon swelling, there was a large reduction in the modulus of the swollen hydrogel (2day immersion) compared to its dried unswollen sample; and contrary to the 0-day mechanical data, the higher the Dex-AI (hydrophilic component) composition was, the larger the reduction in modulus became in swollen hybrid hydrogels. For example, the 100% Dex-AI hydrogel had the lowest 2-day swollen modulus (0.19 MPa) among all the hydrogels. As the PDLLA macromer composition increased, the 2-day swollen moduli increased; 0.19 MPa for 100/0, 0.45 MPa for 80/20, 0.51 MPa for 50/50, 0.77 MPa for 20/80, and 0.91 MPa for 0/100 Dex-AI/PDLLA hydrogels. The moduli of the swollen hydrogels decreased further as incubation time increased beyond 2 days, and the rate of loss of this swollen modulus increased with an increase in the PDLLA macromer composition (hydrolytic degradable component) in the hydrogels. For example, during a 2–60 day immersion period, the percentages of reduction in the swollen moduli
450
Chu
of the 100/0, 50/50, and 0/100 Dex-AI/PDLLA hydrogels were 53, 92, and 100%, respectively. In fact the 100% PDLLA macromer–based hydrogel disappeared in PBS after about a 50-day immersion period. The observed complete opposite dependence of swollen moduli of Dex-AI/PDLLA hybrid hydrogels on their composition from the 0-day dried Dex-AI/PDLLA hydrogels might be attributed to the composition-dependent swelling and the formation of various degrees of 3D porous network structure. According to Zhang et al.’s study of the Dex-AI/PDLLA hydrogels [49], the swelling ratios Dex-AI/PDLLA hydrogels reduced from 349% for 100/0 and 82% for 50/50 to 33% for 0/100 Dex-AI/PDLLA composition ratios during a 2-day immersion period, while the moduli of the same Dex-AI/PDLLA hydrogels increased from 0.19 MPa for 100/0 and 0.51 MPa for 50/50 to 0.91 MPa for 0/100 Dex-AI/PDLLA hydrogels. Therefore, Zhang et al. suggested that as the PDLLA macromer composition increased, the hydrophobicity of the hydrogel increased and swelling in the hydrogel decreased, thus the modulus increased accordingly. This composition–swelling relationship was also reflected in the surface morphology of these hydrogels described previously (Section IV, C). As shown in Fig. 10–12, at the end of 2-day immersion, the 100/0 Dex-AI/PDLLA hydrogel showed the most open 3D porous structure (Fig. 10B), followed by the 50/50 Dex-AI/PDLLA hydrogel (Fig. 12B), and the 0/100 Dex-AI/PDLLA hydrogel had the most compact and dense structure (Fig. 11B). Therefore, the number and size of swelling-induced pores were the highest in the 100/0 Dex-AI/PDLLA hydrogel and decreased as the PDLLA macromer composition increased. This composition-dependent morphological change at 2-day immersion led to the largest modulus reduction in the 100/0 Dex-AI/PDLLA hydrogel (97%), followed by the 50/50 Dex-AI/PDLLA hydrogel (80%), and the 0/100 DexAI/PDLLA hydrogel had the smallest reduction (22%) from its 0-day modulus. At a constant composition ratio, the moduli of the Dex-AI/PDLLA hydrogels decreased with incubation time. This loss in mechanical strength was attributed either to the swellinginduced formation of 3D porous network structure and or to the hydrolytic degradation of the PDLLA macromer via the chain scission of ester linkages located in the PDLLA macromer backbone [48,49]. The magnitude of swelling-induced loss of modulus was composition ratio dependent and occurred in the early stage of immersion; the higher the swelling was, the larger the loss of modulus became. The loss of mechanical strength after the early stage of immersion was attributed to the hydrolytic degradation of the PDLLA macromer component in the DexAI/PDLLA hydrogels; the higher the PDLLA macromer composition was, the larger the loss of modulus became. This PDLLA macromer composition dependence of modulus loss after 2 days immersion was consistent with the weight loss data from Zhang et al.’s study [50]. They found that the weight losses of the 100/0, 50/50, and 0/100 Dex-AI/PDLLA hydrogels were 10, 43, and 100% from the 2- to 60-day incubation periods, respectively. Their losses of modulus were 53, 92, and 100%, respectively. F. Control Release Profiles of Model Drugs The controlled release profiles and kinetics of three model drugs—indomethacin (IDM), insulin, and albumin (BSA)—from Dex-AI/PDLLA hybrid hydrogels and the effect of molecular weight changes in hydrogel precursors on release profiles were also reported by Zhang et al. [47–50,98]. Figure 13 summarizes the release profiles of these three model drugs as a function of the composition ratio of Dex-AI to PDLLA macromer precursors. The data in Fig. 13 suggest that, due to the wide range of chemical characteristics and sizes of these three model drugs, their release profiles from the same Dex-AI/PDLLA hybrid hydrogels were very different from each other. The smallest size model drug among the three tried, IDM, could totally be released from all Dex-AI/PDLLA hydrogels within 40 days and the release rate decreased as the PDLLA
Biodegradable Hydrogels as Drug Vehicles
451
Figure 13 Cumulative release profiles of indomethacin (IDM), insulin, and bovine serum albumin (BSA) from Dex-AI/PDLLA hybrid hydrogels over a wide range of Dex-AI/PDLLA macromer composition ratios (DS 6 for Dex-AI). (From Ref. 98.)
macromer composition in the hydrogel increased. There are two possible reasons for the reduction in IDM release with an increase in the PDLLA macromer composition. First, the hydrophobic interaction between IDM and PDLLA macromer increased as the PDLLA macromer composition increased, this would delay the IDM release. Second, the swelling-induced pores in the hydrogel were large enough for a relatively smaller size IDM to diffuse out freely. As the PDLLA macromer composition increased, the rate of formation of these swelling-induced pores became slower, and hence IDM release also slowed. The PDLLA macromer degradation induced pores and the subsequent formation of a more open and loose 3D network structure appeared not to have a significant effect on the release of small size drugs like IDM. On the contrary, the release characteristics of larger size model drugs like insulin and BSA were found to be significantly affected by the degradation property of the hydrolytically degradable PDLLA macromer component in the Dex-AI/PDLLA hydrogels. For example, large amounts of insulin and BSA were still entrapped inside the 100% Dex-AI hydrogel at the end of a 48-day study period because the pore size formed upon swelling of the 100% Dex-AI hydrogel alone was not large enough for these large size proteins to diffuse out freely. The incorporation of PDLLA macromer into the 100% Dex-AI hydrogel would increase the sustained release and the total amounts of release of both insulin and BSA because the hydrolytic degradation of the PDLLA macromer component would enlarge these swelling-induced porous structures into looser and more open 3D network structures. This benefit was not found in smaller size model drugs like IDM. The release profiles of these three model drugs from the 0/100 Dex-AI/PDLLA hydrogel also revealed that insulin had the highest initial burst release effect compared to IDM and BSA. In IDM and BSA, the 100% Dex-AI hydrogel had the largest initial burst release. This difference could be attributed to the incorporation efficiency of drugs as confirmed by the laser confocal scanning microscopic (LCSM) image data [47,98]. As the PDLLA macromer composition in
452
Chu
the hydrogel increased, more insulin accumulated preferentially near the surface of the hydrogel, while the opposite was found in BSA. The calculated release kinetics of these three drugs were summarized in (Table 4) to assess the effects of drug molecular weight and characteristics on their release properties from the DexAI/PDLLA hybrid hydrogels. The diffusion coefficient, D, of IDM (141.33–14.29⳯10ⳮ9 cm2/s) was substantially larger than those of insulin (0.15–5.40⳯10ⳮ9 cm2/s) and BSA (0.98–9.28⳯10ⳮ9 cm2/s) over all the Dex-AI/PDLLA macromer composition ratios, even though IDM has a very low solubility in buffer (about 0.1 mg/mL). This clearly illustrates that the drug molecular weight and size had a significant effect on the release kinetics. Based on the molecular weights of these three model drugs alone, we would expect that insulin should have a higher D than BSA. Zhang et al.’s experiment a findings, however, showed that the D of BSA was generally larger than that of insulin over the entire composition ratio range. This means that the release rate of BSA at an early immersion stage after initial burst (Mt/M⬀⬍0.6) was higher than that of insulin. The possible reason for these unexpected D values of insulin and BSA may be attributed to the higher solubility of BSA in water (10 mg/mL) than that of insulin (2.5 mg/mL). Therefore, the observed different release profiles of IDM, insulin, and BSA from the DexAI/PDLLA hydrogels might come from the following factors: molecular weight and size of drugs, their water solubility and incorporation efficiency, the hydrophobic interaction between drugs and the hydrogel network, and the swellability and degradability of hydrogels. The above factors that could affect the drug release behavior from delivery devices were also observed by others over a wide range of types and sizes of drugs and a variety of hydrogels [87,99–101]. For example, in Skarda et al.’s poly[N-(2-hydroxyethyl)-L-glutamine] hydrogel (PHEG) [99], they found that the difference in release characteristics of drugs definitely related to the molecular weight. For example, the release of a small size drug like sodium azide (MW 65.01) was the fastest (D ⬃4800⳯10ⳮ9 cm2/s) and was not influenced by the PHEG hydrogel network density, while the diffusion of trypsin-dallikrein inhibitor (TKI) and BSA was controlled by the hydrogel network density within the 7-day study period; i.e., as the hydrogel network density (correlated to pore size) increased from 0.040 to 0.226 mmol/cmⳮ3, the D decreased from 280⳯10ⳮ9 to 61⳯10ⳮ9 cm2/s for TKI and from 4.4⳯10ⳮ9 cm2/s to nondetectable for BSA. Inoue et al. prepared a hydrophobically modified bioadhesive polyelectrolyte hydrogel by grafting oligomers of methyl methacrylate (OMMA) to the backbone of a poly(acrylic acid) (PAAc) [87]. They found that the release rate of IDM was slowed down by the OMMA modification because of the hydrophobic interactions between the OMMA domains and IDM. Furthermore, lysozyme, a positively charged protein drug, was more slowly released from the hydrogel because of its large size and association with the negatively
Table 4 Comparison of the Diffusion Coefficient D (⫻10⫺9 cm2/s) of Three Model Drugs Released from Dex-AI/PDLLA Hybrid Hydrogels in pH 7.4 Buffers at 37°C Dex-AI/PDLLAM 100/0 80/20 50/50 20/80 0/100
Indomethacin
Insulin
BSA
141.33 82.49 54.67 39.58 14.29
0.15 0.53 1.50 1.71 5.40
0.98 1.43 1.94 4.69 9.28
D calculated by Mt/M∝ ⫽ 4 (Dt/␦2)1/2 with Mt/M∝⬍0.6. Mt is the drug released at time t, and M∝ is the estimated total amount of drug entrapped in the hydrogels. ␦ is the thickness of the hydrogel after 2-day immersion.
Biodegradable Hydrogels as Drug Vehicles
453
charged PAAc and the hydrophobic domains. Cheng et al. developed an azopolymer-coated capsule – containing model drug vitamin B12 or insulin, and found that the diffusion of the model drugs depended on the molar volume, which was correlated to the molecular weight [101]. The release of insulin was much slower than vitamin B12.
V. EMERGING BIODEGRADABLE HYDROGEL SYSTEMS One of the most unique emerging biodegradable hydrogel systems is the multiarm star-shaped biodegradable hydrogels. In recent years, there has been increased interest in star-shaped polymers [102–109], a kind of branched polymer having more than two linear polymeric arms attached to a center core. Since a star-shaped polymer has a higher segment density within the distance of its radius of gyration than a homologous linear polymer does under the same conditions, the excluded volume effects were more pronounced in star-shaped polymers. Compared to linear polymers, the characteristic properties of star-shaped polymers derived from their unique shape were observed in solution and in bulk. For example, star-shaped polymers exhibit a smaller hydrodynamic radius and lower solution viscosity compared to linear polymers of the same molecular weight and composition [110]. There are many methods to synthesize the star-shaped polymers. For example, amphiphilic star-shaped polymers of vinyl ethers with hydroxyl or carboxylic groups were synthesized via linking reactions of linear living polymers mediated by bifunctional vinyl compounds [111,112]. Many synthesis methods, such as chain coupling reactions [113], stable free-radical polymerization [114], living free-radical polymerization [115–117], chain transfer polymerization [118,119], have been reported recently. Star-shaped biodegradable polylactones, such as star-shaped poly(ε-caprolactone), appear to be very interesting hydrogel precursors and may have a great potential in biomedical applications because their higher polymer mass and functionality per unit volume may provide a unique means to regulate the release of both small and large size drugs. Lang et al. recently reported the synthesis and structural characterization of multiarm star-shaped poly(ε-caprolactone)s using either glycerol as the core for three arms or pentaerythritol as the core for four arms [52,53]. In their and subsequent studies by Wu et al. [52,53,55], photocrosslinkable maleic acid was incorporated into the chain ends of the poly(ε-caprolactone) multiarms for preparing star-shaped hydrogel precursors. The advantages of using maleic acid as the photocrosslinkable units were previously given in Kim et al.’s study of dextran-based biodegradable hydrogels [39,40]. In addition to those merits, the use of maleic acid could also provide the means to incorporate nitric oxide derivatives into the star-shaped biodegradable hydrogels via the free —COOH groups in maleic acid monoester located at the chain ends of the poly(ε-caprolactone) arms for providing nitric oxide biological functionality. Such a means is based on the recently reported patented and published method by Lee et al. [120–122]. Lang et al. also applied such a means to attach nitric oxide derivatives (4-amino-2,2,6,6-tetramethylpiperidine-1-oxy) into the carboxylic acid sites of the acrylic acid/lactide/ε-caprolactone copolymer [123]. Schemes 6 and 7 illustrate the chemical structure of three- and four- arm star-shaped poly(ε-caprolactone) hydrogel precursors and Scheme 8 shows the representative hydrogels from these precursors. Wu et al. [55] very recently advanced these star-shaped biodegradable hydrogel systems further by integrating the star-shaped poly(ε-caprolactone) hydrogel precursor (PCLMa) with dextran–maleic acid (Dex-Ma) precursor previously prepared by Kim et al. [39,40] via photocrosslinking. Such a hybrid star-shaped biodegradable hydrogel system exhibits a very wide range of swelling ratios from several hundreds to thousands of percent as shown in Fig. 14. It is expected that this unique hybrid star-shaped Dex-Ma/PCL-Ma hydrogel system may
454
Chu
Scheme 6 Chemical structures of hydroxyl terminated three-arm poly(ε-caprolactone) (III) (PGCL-OH) from ε-caprolactone (II) and glycerol core (I), and double bond functionalized three-arm poly(ε-caprolactone) hydrogel precursor (V) (PGCL-Ma) from PGCL-OH (III) and maleic anhydride (IV).
Scheme 7 Chemical structures of double bond–functionalized four-arm poly(ε-caprolactone) (IV) from ε-caprolactone monomer (I), pentaerythritol core (II), four-arm hydroxyl terminated poly(ε-caprolactone) (III) and maleic anhydride (IV).
Biodegradable Hydrogels as Drug Vehicles
455
Scheme 8 Representative of a four, arm poly(ε-caprolactone) hydrogel network.
Figure 14 Swelling ratios (%) of hybrid and star-shaped three-arm poly(ε-caprolactone)–maleic acid (PCL-Ma) and dextran–maleic acid (Dex-Ma DS 1.15) with composition ratios of PCL-Ma to Dex-Ma at 100/0, 60/40, 30/70, and 0/100. The Mn of PCL-Ma is 5600. (From Ref. 55.)
456
Chu
have a great potential for the controlled release of a wide range of sizes of hydrophilic to hydrophobic bioactive agents.
ACKNOWLEDGMENTS The author wishes to thank his former graduate students, Sim-Hi Kim, Chee-Youb Won, Yeli Zhang, and Sunny Namkung, and postdoctoral fellows, Drs. M. D. Lang, X. Z. Zhang, and D. Q. Wu. The content of this review chapter is based on their published research activities.
REFERENCES
1. Park K, Shalaby WSW, Park H. Biodegradable Hydrogels for Drug Delivery. Lancaster. PA: Technomic, 1993. 2. Hennink WE, Nostrum van CF. Novel crosslinking methods to design hydrogels. Adv. Drug Delivery Rev 2002; 54:13–36. 3. Hoffman AS. Hydrogels for biomedical applications. Adv. Drug Delivery Rev 2002; 43:2–12. 4. Lee KY, Mooney DJ. Hydrogels for tissue engineering. Chem. Rev 2001; 101:1869–1880. 5. Hoffman AS. Application of thermally reversible polymers and hydrogels in therapeutics and diagnostics. J. Control Rel 1987; 6:297–305. 6. Park K, Park H, Hydrogels Smart. In: Salomone JC, ed. Concise Polymeric Materials Encyclopedia. Boca Raton: CRC Press, 1999:1476–1478. 7. Chen GH, Hoffman AS. Graft copolymers that exhibit temperature-induced phase transitions over a wide range of pH. Nature 1995; 373:49–52. 8. Liu Z, Calvert P. Multilayer hydrogels as muscle-like actuators. Adv. Mater 2000; 12:288–291. 9. Miyata T, Asami N, Uragami T. A reversibly antigen-responsive hydrogel. Nature 1999; 399: 766–769. 10. Juodkazis S, Mukai N, Wakaki R, Yamaguchi A, Matsuo S, Misawa H. Reversible phase transitions in polymer gels induced by radiation forces. Nature 2000; 408:178–181. 11. Li Y, Hu Z, Chen Y. Shape memory gels made by the modulated gel technology. J. Appl. Polym. Sci 1997; 63:1173–1178. 12. Hiroya T, Tamura N, Yasui M, Fujimoto K, Kawaguchi H. Enzyme immobilization on thermosensitive hydrogel microspheres. Colloids Surf., B 1990; 4:24–32. 13. Chen JP, Sun YM, Chu DH. Immobilization of ␣-amylase to a composite temperature-sensitive membrane for starch hydrolysis. Biotechnol. Prog 1998; 14:473–478. 14. Ramkissoon-Ganokar C, Liu F, Baudys M, Kim SW. Modulating insulin release profile from pH/ thermoensitive polymeric beads through polymer molecular weight. J. Control Rel 1999; 59: 287–298. 15. Chung JE, Yokoma M, Okano T. Inner core segment design for drug delivery control of thermoresponsive polymeric micelles. J. Control Rel 2000; 65:95–103. 16. Hinrichs WLJ, Schuurmans-Nieuwenbrock NME, Van de Wetering P, Hennink WE. Thermosensitive polymers as carriers for DNA delivery. J. Control Rel 1999; 60:249–259. 17. Sutani K, Kaetu I, Uchida K, Matsubara Y. Stimuli responsive drug release from polymer gel. Controlled release of ionic drug from polyampholyte gel. Radiat. Phys. Chem 2002; 64:331–336. 18. Qiu Y, Park K. Environmental-sensitive hydrogels for drug delivery. Adv. Drug Delivery Rev 2001; 53:321–339. 19. Bardwaj R, Blanchard J. Controlled-release delivery system for the ␣-MSH melanptan-I using Poloxamer 407. J. Pharm. Sci 1996; 85:915–919. 20. Jeong B, Bae YH, Kim SW. Themoreversible gelation of PEG-PLGA-PEG triblock copolymer aqueous solutions. Macromolecules 1999; 31:7064–7069.
Biodegradable Hydrogels as Drug Vehicles
457
21. Jeong B, Bae YH, Kim SW. Drug release from biodegradable injectable thermosensitive hydrogel of PEG-PLGA-PEG triblock copolymers. J. Control Rel 2000; 63:155–163. 22. Bae YH, Okano T, Kim SW. Temperature dependence of swelling of crosslinked poly(N,N′-alkyl substituted acrylamide) in water. J. Polym. Sci.: Polym. Phys 1990; 28:923–936. 23. Stayton PS, Shimobji T, Long C, Chilkoti A, Chen G, Harris JM, Hoffman AS. Control of proteinligand recognition using a stimuli-responsive polymer. Nature 1995; 378:472–474. 24. Ramkissoon-Ganorkar C, Liu F, Baudysˇ M, Kim SW. Modulating insulin-release profile from pH/thermosensitive polymeric beads through polymer molecular weight. J. Control Rel 1999; 59: 287–298. 25. Osada Y, Okuzaki H, Hori H. A polymer gel with electrically driven motility. Nature 1992; 355: 242–244. 26. Kabra BG, Gehrke SH. Synthesis of fast response poly(n-isopropylacrylamide) gels. Polym. Commun 1991; 32:322–323. 27. Wu XS, Hoffman AS, Yager P. Synthesis and characterization of thermally reversible macroporous poly(n-isopropylacrylamide) hydrogels. J. Polym. Sci., Part A, Polym. Chem 1992; 30:2121–2129. 28. Yoshida R, Uchida K, Kaneko Y, Sakai K, Kikuchi A, Sakurai Y, Okano T. Comb-type grafted hydrogels with rapid de-swelling response to temperature-changes. Nature 1995; 374:240–242. 29. Zhang XZ, Zhuo RX. Novel method to prepare a fast responsive, thermosensitive poly(n-isopropylacrylamide) hydrogel. Macromol. Rapid Commun 1999; 4:229–231. 30. Zhang XZ, Zhuo RX. Preparation of fast responsive, temperature-sensitive poly(n-isopropylacrylamide) hydrogel. Macromol. Chem. Phys 1999; 200:2602–2605. 31. Zhang XZ, Zhuo RX. Dynamic properties of temperature sensitive poly(n-isopropylacrylamide) gel crosslinked through siloxane linkage. Langmuir 2001; 17:12–16. 32. Zhang XZ, Yang YY, Chung TS, Ma KX. Preparation and characterisation of the macroporous poly(n-isopropylacrylamide) hydrogel with fast response. Langmuir 2001; 17:6094–6099. 33. Zhang XZ, Wang FJ, Chu CC. Thermoresponsive hydrogels with rapid response dynamics. J. Mater. Sci., Mater in Med; (in press). 34. Brondsted H, Andersen C, Hovgaard L. Crosslinked dextran–a new capsule material for colon targeting of drugs. J. Controlled Release 1998; 53:7–13. 35. Hennink WE, Talsma H, Borchert JCH, De Smedt SC, Demeester J. Controlled release of proteins from dextran hydrogels. J. Control Rel 1996; 39:47–55. 36. Hovgaard L, Brondsted H. Dextran hydrogels for colon-specific drug delivery. J. Controlled Release 1995; 36:159–166. 37. Van Dijk-Wolthuis WNE, Hoogeboom JAM, Van Steenbergen MJ, Tsang SKY, Hennink WE. Degradation and release behavior of dextran-based hydrogels. Macromolecules 1997; 30: 4639–4645. 38. Won CY, Chu CC. Dextran–estrone conjugate: synthesis and in vitro release study. Carbohydr. Polym 1998; 36:327–334. 39. Kim S, Won CY, Chu CC. Synthesis and Characterization of Dextran–maleic acid based hydrogel. J. Biomed. Mater. Res 1999; 46(2):160–170. 40. Kim S, Won CY, Chu CC. Dextran maleic acid monoesters and hydrogels therefrom, U.S, Patent 6,476,204, Nov. 5, 2002. 41. Kim S, Won CY, Chu CC. Synthesis and characterization of dextran-based hydrogel by photocrosslinking method. Carbohydr. Polym 1999; 40(3):183–190. 42. Kim S, Chu CC. Pore structure analysis of swollen dextran–methacrylate hydrogels by scanning electron microscopy and mercury intrusion porosimetry. J. Biomed. Mater. Res., Appl Biomateri 2000; 53(3):258–266. 43. Kim S, Chu CC. Synthesis and characterization of dextran–methacrylate hydrogel and its structure study by SEM. J. Biomed. Mater. Res 2000; 49(4):517–527. 44. Kim S, Chu CC. In vitro release behavior of dextran–methacrylate hydrogel using doxorubicin and other model compounds. J. Biomater. Appl 2000; 15(1):23–46. 45. Zhang YL, Won CY, Chu CC. Synthesis and characterization of biodegradable hydrogels having polylactide and dextran components with controlled swelling properties. J. Polym. Sci., Polym. Chem 2000; 38:2392–2404.
458
Chu
46. Zhang YL, Won CY, Chu CC. Synthesis and characterization of biodegradable network hydrogels having both hydrophobic and hydrophilic components with controlled swelling properties. J. Polym. Sci., Polym. Chem 1999; 37:4554–4569. 47. Zhang YL, Chu CC. Biodegradable dextran–polylactide based hydrogel network and its controlled release of albumin. J. Biomed. Mater. Res 2001; 54(1):1–11. 48. Zhang YL, Chu CC. Biodegradable dextran–polylactide hydrogel networks: their swelling, morphology and controlled release of indomethacin. J. Biomed. Mater. Res 2002; 59(2):318–328. 49. Zhang YL, Chu CC. The effect of molecular weight of biodegradable hydrogel precursors on the in vitro indomethacin release from dextran and poly(D,L)-lactic acid hybrid hydrogel. J. Bioact. Compat. Polym 2002; 17:65–85. 50. Zhang YL, Chu CC. The effect of degradation of biodegradable hydrogel network on its release of albumin. J. Mater. Sci., Mater Med 2002; 13:667–676. 51. Won CY, Zhang YL, Chu CC. Hydrogel forming system with hydrophobic and hydrophilic components. US patent 6, 388, 047, May 14, 2002. 52. Lang MD, Wong RP, Chu CC. Synthesis and structural analysis of functionalized three-arm star poly(ε-caprolactone)s. J. Polym. Sci., Polym. Chem 2002; 40(8):1127–1141. 53. Lang MD, Chu CC. Functionalized multi-arm poly(ε-caprolactone)s: synthesis, structure analysis and network formation. J. Appl. Polym. Sci 2002; 86(9):2296–2306. 54. Lang MD, Chu CC. Biodegradable Polyhydroalcohol Esters, US non-provisional patent filed on June 15, 2002. 55. Wu DQ, Zhang XZ, Chu CC. Synthesis, characterization & drug release from multiarm poly(εcaprolactone) maleic acid poly(ethylene glycol) diacrylate hydrogels. J. Bimat. Sci Polym. Ed. (In press). 56. Gonzales D, Fan K, Sevoian M. Synthesis and swelling characterization of a poly(␥-glutamic acid) hydrogel. J. Polym. Sci., Part A: Polym. Chem 1996; 34:2019–2027. 57. Van Dijk-Wolthuis WNE, Hoogeboom JAM, Van Steenbergen MJ, Tsang SKY, Hennink WE. Degradation and release behavior of dextran-based hydrogels. Macromolecules 1997; 30: 4639–4645. 58. Trivedi BC, Culbertson BM. Maleic Anhydride. New York: Plenum Press, 1982. 59. Flett LH, Gardner WH. Maleic anhydride derivatives. New York: John Wiley & Sons, 1952. 60. Jedlovcnik R, Sebenik A, Golob J, Korbar J. Step-growth polymerization of maleic anhydride and 1,2-propylene glycol. Polym. Eng. Sci 1995; 35:1413–1417. 61. Liou FJ, Wang YJ. Preparation and characterization of crosslinked and heat-treated PVA-MA films. J. Appl. Polym. Sci 1996; 59:1395–1403. 62. Arcamone F. Doxorubicin; Anticancer Antibiotics. New York: Academic Press, 1981:16–47, 93–125. 63. Remers WA. The chemistry of Antitumor Antibiotics. New York: John Wiley & Sons, 1979: 108–127. 64. Jurczak J, Golebiowski A. Antibiotics and antiviral compounds. Weinheim: VCH, 1993:352–356. 65. Arcamone F. Anticancer agents based on natural product models. New York: Academic Press, 1980: 1–41. 66. Barabas K, Sizensky JA, Faulk WP. Transferrin conjugates of adriamycin are cytotoxic without intercalating nuclear DNA. J. Biol. Chem 1992; 267:9437–9442. 67. Arrington KA, Legendre AM, Tabeling GS, Frazier DL. Comparison of body surface area–based and weight-based dosage protocols for doxorubicin administration in dogs. Am. J. Vet. Res 1994; 55:1587–1592. 68. Boucek RJ, Olson RD, Brenner DE, Ogunbunmi EM, Inui M, Fleischer S. The major metabolite of doxorubicin is a potent inhibitor of membrane-associated ion pumps. J. Biol. Chem 1987; 262: 15851–15856. 69. Matherne CM, Satterfield WC, Gasparini A, Tonetti M, Astroff AB, Schmidt RD, Rowe LD, DeLoach JR. Clinical efficacy and toxicity of doxorubicin encapsulated in glutaraldehyde-treated erythrocytes administered to dogs with lymphosarcoma. Am. J. Vet. Res 1994; 55:847–853.
Biodegradable Hydrogels as Drug Vehicles
459
70. Tonetti M, Astroff AB, Satterfield W, Flora AD, Benatti U, DeLoach JR. Pharmacokinetic properties of doxorubicin encapsulated in glutaraldehyde-treated canine erythrocytes. Am. J. Vet. Res 1990; 52:1630–1635. 71. Arap W, Pasqualini R, Ruoslahti E. Cancer treatment by targeted drug delivery to tumor vasculature in a mouse model. Science 1998; 279:377–279. 72. Gupta PK, Hung CT. Comparative disposition of adriamycin delivered via magnetic albumin microspheres in presence and absence of magnetic field in rats. Life Sci 1990; 46:471–479. 73. Baker R. Controlled release of biologically active agents. New York: John Wiley & Sons, 1987: 1–105. 74. Lee YM, Kim SS. Hydrogels of poly(ethylene glycol)-co-poly(lactones) diacrylate macromers and -chitin. Polymer 1997; 38:2415–2420. 75. Kissel T, Li YX, Volland C, Gorich S, Koneberg R. Parenteral protein delivery systems using biodegradable polyesters of ABA block structure, containing hydrophobic poly(lactide-co-glycolide) A blocks and hydrophilic poly(ethylene oxide) B blocks. J. Control Rel 1996; 39:315–326. 76. West JL, Hubbell JA. Photopolymerized hydrogel materials for drug delivery applications. React. Polym 1995; 25:139–147. 77. Kamath KR, Park K. Biodegradable Hydrogels in Drug Delivery. In: Advanced Drug Delivery Reviews: Elsevier Science Publishers, 1993:Chapter 11, 60. 78. Kurisawa M, Terano M, Yui N. Double-stimuli-responsive degradation of hydrogels consisting of oligopeptide-terminated poly(ethylene glycol) and dextran with an interpenetrating polymer network. J. Biomater. Sci. Polym Ed 1997; 8:691–708. 79. Sawhney AS, Pathak CP, Hubbell JA. Bioerodible hydrogels based on photopolymerized poly(ethylene glycol)-co-poly(␣-hydroxy acid) diacrylate macromers. Macromolecules 1993; 26:581–587. 80. Huang SJ, Onyari J. Multicomponent polymers of poly(lactic acid) macomonomers with methacrylate terminal and copolymers of poly(2-hydroxyethylmethacrylate). J.M.S Pure Appl. Chem 1996; A33(5):571–584. 81. Mikos AG, Wald HL, Sarakinos G, Leite SM, Langer R. Biodegradable cell transplantation devices for tissue regeneration. Symposium Proceedings, 252: Materials Research Society, 1992:353–358. 82. Suganuma J, Alexander H, Traub J, Ricci JL. Biological response of intramedullary bone to poly(Llactic acid). Symposium Proceedings, 252: Materials Research Society, 1992:339–343. 83. Chu CC. Biodegration Property. In: Chu CC, Von Fraunhofer JA, Greisler HP, eds. Wound Closure Biomaterials and Devices. Boca Raton. Fla.: CRC Press, 1997:Chapter 7, 172. 84. Nakamura H, Tachi K. Effects of hydrophobic interactions on swelling of carboxylated core-shell structured latex particles. J. Appl. Polym. Sci 1997; 65:1933–1938. 85. Akala EO, Kopeckova P, Kopecek J. Novel pH-sensitive hydrogels with adjustable swelling kinetics. Biomaterials 1998; 19:1037–1047. 86. Chen X, Li WJ, Zhong W, Lu YH, Yu TY. pH sensitivity and ion sensitivity of hydrogels based on complex-forming chitosan/silk fibroin interpenetrating polymer network. J. Appl. Polym. Sci 1997; 65:2257–2262. 87. Inoue T, Chen G, Nakamae K, Hoffman AS. A hydrophobically modified bioadhesive polyelectrolyte hydrogel for drug delivery. J. Control Rel 1997; 49:167–176. 88. Yeh PY, Kopeckova P, Kopecek J. Biodegradable and pH sensitive hydrogels: synthesis by crosslinking of N,N-dimethylacrylamide copolymer precursors. J. Polym. Sci., Part A: Polym. Chem 1994; 32:1627–1637. 89. Israelachvili J. Special Interactions: Hydrogen-bonding, Hydrophobic, & Hydrophilic Interaction. In: Intermolecular and Surface Forces: Academic Press, 1985:Chapter 8, 97–106. 90. Chu CC. The Effect of pH on the in vitro degradation of poly(glycolide-lactide) co-polymer absorbable sutures. J. Biomed. Mater. Res 1982; 16:117–124. 91. Chu CC. In-vitro degradation of polyglycolic acid sutures: effect of pH. J. Biomed. Mater. Res 1981; 15:795–804. 92. Karmalkar RN, Kulkarni MG, Mashelkar RA. Pendent chain linked delivery systems. I. Facile hydrolysis through anchimeric effect. J. Control Rel 1996; 42:185–193. 93. Zhang YL, Chu CC. Thermal and mechanical properties of a new class of biodegradable hydrogel networks from polysaccharide and synthetic biodegradable polyesters. J. Mater. Sci., Mater Med().
460
Chu
94. Kakizaki M, Yamamoto H, Ohe T, Hideshima T. Molecular motions and dielectric relaxations in chitin, chitosan and related polymers. In: Skjak-Braek G, Anthonsen T, Sandford P, eds. Chitin and Chitosan. London: Elsevier Applied Science, 1989:512. 95. Migliaresi C, Fambri L, Cohn D. A study on the in vitro degradation of poly(lactic acid). J. Biomater. Sci. Polym Ed 1994; 5:591–606. 96. Rodriguez F. Principles of Polymer Systems. In: Physical States & Transitions. Washington, DC: Taylor & Francis, 1996:Chapter 3, 56. 97. Eschbach FO, Huang SJ. Hydrophilic–hydrophobic binary systems of poly(2-hydroxyethly methacrylate) and polycaprolactone. Part I. Synthesis and characterization. J. Bioact. Compat. Polym 1994; 9:29–54. 98. Zhang YL, Chu CC. Controlled release of insulin from biodegradable dextran–polylactide hydrogel network. J. Biomater. Appl(). 99. Skarda V, Rypacek F. Permeation of proteins through a biodegradable hydrogel. J. Bioact. Compat. Polym 1997; 12:186–195. 100. Fitzgerald JF, Corrigan OI. Investigation of the mechanisms governing the release of levamisole from poly-lactide-co-glycolide delivery systems. J. Control Rel 1996; 42:125–132. 101. Cheng CL, Gehrke SH, Ritschel WA. Development of an azopolymer based colonic release capsule for delivering proteins/macromolecules. Meth. Find. Exp. Clin. Pharmacol 1994; 16:271–278. 102. Takano A, Okada M, Nose T, Fujimoto T. Synthesis and characterization of star-shaped polymer with one labeled arm. Macromolecules 1992; 25:3596–3598. 103. Chujo Y, Sada K, Kawasaki T, Saegusa T. Synthesis of nonionic hydrogel from star-shaped polyoxazoline. Polym. J 1992; 24(11):1301–1306. 104. Kanaoka S, Sueoka M, Sawamoto M, Higashimura T. Star-shaped polymers by living cationic polymerization. 7. Amphiphilic graft polymers of vinyl ethers with hydroxyl groups—synthesis and host–guest interaction. J. Polym. Sci., Part A: Polym. Chem 1993; 31:2513–2521. 105. Raphael E, Pincus P, Fredrickson GH. Conformation of star polymers in high-molecular-weight solvents. Macromolecules 1993; 26:1996–2006. 106. Naka K, Konishi G, Kotera K, Chujo Y. Synthesis of a star-shaped polymer having tris(betadiketonato)chromium(III) at the center core. Polym. Bull 1998; 41:263–266. 107. Choi YK, Bae YH, Kim SW. Star-shaped poly(ether-ester) block copolymers: synthesis, characterization, and their physical properties. Macromolecules 1998; 31:8766–8774. 108. Ruckenstein E, Zhang HM. A novel breakable cross-linker and pH-responsive star-shaped and gel polymer. Macromolecules 1999; 32:3979–3983. 109. Deng F, Bisht KS, Gross RA, Kaplan DL. Chemoenzymatic synthesis of a multiarm poly(lactideco-ε-caprolactone). Macromolecules 1999; 32:5159–5161. 110. Choi YK, Bae YH, Kim SW. Protein release from microspheres of star-shaped PEO-PLA block copolymers. Proc. Int. Symp. Control. Relat. Bioact. Mater 1996; 23:349–350. 111. Kanaoka S, Sawamoto M, Higashimura T. Star-shaped polymers by living cationic polymerization. 2. Synthesis of amphiphilic star-shaped block polymers of vinyl ethers with hydroxyl groups. Macromolecules 1991; 24:5741–5745. 112. Kanaoka S, Sawamoto M, Higashimura T. Star-shaped polymers by living cationic polymerization. 5. Core-functionalized amphiphilic star-shaped polymers of vinyl ethers with hydroxylgroups—synthesis and host–guest interaction. Macromolecules 1993; 26:254–259. 113. Kurian P, Zschochi S, Kennedy JP. Synthesis and characterization of novel amphiphilic block copolymers di-, tri-, multi-, and star blocks of PEG and PIB. J. Polym. Sci., Part A: Polym. Chem 2000; 38:3200–3209. 114. Pasquale AJ, Long TE. Synthesis of star-shaped polystyrenes via nitroxide-mediated stable freeradical polymerization. J. Polym. Sci., Part A: Polym. Chem 2001; 39:216–223. 115. Ohno H, Wong B, Haddleton DM. Synthesis of well-defined cyclodextrin-core star polymers. J. Polym. Sci., Part A: Polym. Chem 2001; 39:2206–2214. 116. Percec V, Barboiu B, Bera TK, Van Der Sluis M, Grubbs RB, Frechet JMJ. Designing functional aromatic multisulfonyl chloride initiators for complex organic synthesis by living radical polymerization. J. Polym. Sci., Part A: Polym. Chem 2000; 38:4776–4791.
Biodegradable Hydrogels as Drug Vehicles
461
117. Tsoukatos T, Pispas S, Hadjichristidis N. Star-branched polystyrenes by nitroxide living free-radical polymerization. J. Polym. Sci., Part A: Polym. Chem 2001; 39:320–325. 118. Guo Y, Xu J, Pan C. Block and star block copolymers by mechanism transformation. IV. Synthesis of S-(PSt)2(PDOP)2 miktoarm star copolymers by combination of ATRP and CROP. J. Polym. Sci., Part A: Polym. Chem 2001; 39:437–445. 119. Stenzel-Rosenbaum M, Davis TP, Chen V, Fane AG. Star polymer synthesis via radical reversible addition-fragmentation chain-transfer polymerization. J. Polym. Sci., Part A: Polym. Chem 2000; 39:2777–2783. 120. Lee KH, Chu CC. Molecular design of biologically active biodegradable polymers for biomedical applications. Macromol. Symp 1998; 130:71–80. 121. Chu CC. Biodegradable biomaterials having nitric oxide biological activity. In: Dumitriu S. , ed. New York: Marcel Dekker, 2002:123–132. 122. Lee KH, Chu CC, Freed J. Aminoxyl-containing radical spin labeling in polymers and copolymers. U.S. patent 5,516,881, May 16, 1996. 123. Lang MD, Chu CC. Synthesis and chemical structural analysis of nitroxyl radical incorporated poly(acrylic acid/lactide-ε-caprolactone) copolymers. J. Polym. Sci., Part A: Polym. Chem 2001; 39(24):4214–4226.
20 Growth Factor Delivery by a Cloned Multipotent Cell from Bone Marrow Stroma Potentiates Bone Repair Nigel M. Azer, Quanjun Cui, Ing-Lin Chang, T. Lisle Whitman, Chang Hahn, Gwo-Jaw Wang, and Gary Balian University of Virginia School of Medicine Charlottesville, Virginia, U.S.A.
I. INTRODUCTION Bone repair is a complex process that involves the interaction of numerous cells and cell products in a carefully controlled milieu, regulated locally and systemically by hormones and other growth factors. Bone healing and repair are of paramount importance in orthopedic surgery, especially as they relate to fracture nonunion, delayed union, segmental defect repair, spinal fusion, osteoporosis, and bony ingrowth to prostheses. Autogenous bone grafting and allografts are commonly used to remedy these problems. In fact, 100,000 to 200,000 such operations are performed annually in the United States [1]. The results are good or excellent in most instances. However, they do not solve all cases of bone deficiency and are not free of complications. Failure rates as high as 13–30% have been reported in the literature with autologous bone grafting, not to mention the clinically significant donor and recipient site morbidity [2,3]. Allogeneic bone grafts have an even higher complication rate, including the risk of disease transmission. Thus, supplements and alternatives for bone grafting are being developed [4]. Recent advances in molecular and cellular biology coupled with an improved understanding of the mechanisms that regulate bone formation have had direct implications on orthopaedic surgery. Mesenchymal stem cells, which are distinct from the endothelial or hematopoetic cell lineages of the marrow, have the capability to become osteogenic, chondrogenic, or adipogenic depending upon the conditions and growth factors they are exposed to. The literature suggests that endogenous mesenchymal stem cells play a role in virtually all forms of musculoskeletal tissue regeneration, including endochondral ossification [5–12]. In fact, diminished capabilities for bone healing in the elderly and the pathology of osteoporosis have been attributed to lower quantities of mesenchymal stem cells [13]. Many cloned lineages of mesenchymal stem cells are being studied in various laboratories across the country, realizing their potential as vehicles of cellular and gene therapy [6,12,14]. However, details regarding their precise role in bone healing as well as their ultimate fate in vivo remains in question. The D1 population of cells is a lineage of cloned, murine, pluripotential stem cells which has been extensively studied and modified in our laboratory. It has osteogenic and adipogenic 463
464
Azer et al.
properties in vitro and in vivo in diffusion chambers, evidenced by increased alkaline phosphatase levels, increased osteocalcin mRNA expression, and increased cAMP after PTH administration [15]. This population has been modified to allow histological and molecular tracing of the cells, making them ideal for experimental study. These modifications did not alter the phenotype of the cells [16]. These cells have been shown to retain their osteogenic phenotype, to repopulate bone marrow, to localize to implanted titanium rods, and to localize to fracture sites in mice [17,18]. Bone healing relies not only on the cells involved, but also on the transient and localized need for enhancing factors such as growth factors and bone morphogenetic proteins. Growth factors in particular are important in the regulation and control of bone growth and repair. However, they must be delivered to the site of bone healing and have short half-lives in vivo which may curtail their clinical efficacy. Like all proteins, these growth factors are synthesized by the body using the genetic material that encodes them. When a gene encoding a growth factor is isolated and transferred in the appropriate fashion to a cell, the recipient cell begins to locally manufacture the particular growth factor in question [19]. Hence, efficient means of reproducibly delivering critical growth factors to specific sites of bone healing at the appropriate time are sought. Insulin-like growth (IGF), also known as somatomedin C, is a 7.5-kD protein synthesized in the liver which has been shown to be particularly important in bone formation and remodeling. It is the most abundant growth factor in the bone matrix and has a stimulatory effect on both osteoblasts and osteoclasts. It is an autocrine and paracrine regulator that is incorporated into the bone matrix [20,21]. Thus, during bone resorption it is liberated and recruits osteoblasts, thereby coupling bone resorption and formation. Several studies have illustrated the bone forming capabilities of IGF [22,23]. Increased bone formation and bone density after systemic treatment of IGF in osteopenic ovariectomized rats is noted [24]. Subcutaneous administration in postmenopausal females generated increased expression of type I collagen carboxy-terminal propeptide [25]. The D1 population of mesenchymal stem cells is further modified by transfection, using a CMV promoter, with cDNA encoding insulin-like growth factor-I and neomycin resistance, termed D1-IGF. These cells retain their osteogenic phenotype in vivo and are traceable by DNAPCR, but do not stain histologically with XGAL. In vitro studies have confirmed higher titers of IGF expression from these cells [26]. In the present study, we examine the effects of administration of genetically manipulated cloned mesenchymal stem cells in vivo to further elucidate the role of these cells as a vehicle for cellular and gene therapy in bone repair. We hypothesize that the cells will localize to defect sites, participate in bone healing, and potentiate that healing process.
II. MATERIALS AND METHODS A. Cells The D1 line of cloned murine pluripotential cells was isolated and cloned [15]. These cells were genetically labeled using a lacZ virus encoding -galactosidase and neomycin resistant gene, allowing these cells to be traced, and termed D1-BAG [16]. The D1 cells were further modified in our laboratory by electroporation using a CMV promotor with cDNA for rat osteocalcin promotor and murine IGF-I. They were stabilized in cell culture, selected for, and termed D1IGF (Fig. 1). Northern and Western blots confirmed that the cells transfected with CMV-IGFI yielded a 20-fold increase in production of IGF [26]. Both cell populations were separately maintained at near confluence in a monolayer in 75-cm2 flasks in Dulbecco modified Eagle’s
Growth Factor Delivery by a Cloned Multipotent Cell
465
medium with L-glutamine (Gibco, Gaithersburg, MD) containing 15% fetal bovine serum, 500 g of L-ascorbic acid per millilter and antibiotics (100 U/mL penicillin G and 100 mcg/mL streptomycin) in a humidified atmosphere of 5% carbon dioxide at 37⬚C (Forma Scientific, Marietta, OH). The cells were transferred to a new culture flask every 4 to 5 days as needed to maintain their exponential growth phase. Upon reaching a near-confluent state, the cells were lifted using a combination of ethylenediamine tetraacetic acid (EDTA) followed by trypsin (Difco, Detroit, MI), and agitation. The cells were centrifuged at 600 G on a Dynac benchtop centrifuge (Clay Adams, Pipsippany, NJ) for 10 min and suspended in 100 L of lactated Ringer’s solution at a concentration of 1 ⳯ 106 cells/mL as determined by multiple manual counts using a Hausser ultraplane hemacytometer. B. In Vivo Cell Delivery Sixty three 8-week-old balb/c mice (Hilltop Lab Animals, Scottdale, PA) were used in this study. Using sterile and ethical techniques, the mice were anesthetized with 0.3 cc of IM 1% xylazine and ketamine. The right distal femur of each mouse was surgically exposed. A 1-mm transverse, cortical defect was made in the medial femoral condyle using a hand drill. Figure 2 (top) represents a low power photomicrograph for orientation, which represents the defect site 4 weeks after injury. The mice were then divided into experimental groups: those that would receive D1-BAG cells, D1-IGF cells, or lactated Ringer’s solution as a control. Two groups of
Figure 1 IGF-1 mRNA expression stable transfection. D1 cells tranfected with IGF-1 cDNA were selected with G418.
466
Azer et al.
1 ⳯ 106 D1-BAG and D1-IGF-1 cells were lifted from cell culture, quantified, and suspended in 100 L of lactated Ringer’s solution. An equivalent volume of lactated Ringer’s solution without cells was used as a control. Using a 30-guage needle, the samples were administered into the ipsilateral femoral medullary canal. The wounds were closed in layers with 5–0 PDS and 4–0 nylon suture, and the mice were permitted to fully bear weight. The mice were euthanized in a CO2 chamber at 2, 4, and 8 weeks. Their distal femurs were surgically excised. One-half of the specimens were utilized for histological staining and examination, whereas the remaining were used for DNA extraction and DNA-PCR as described below.
C. Histological Tracing of Cells The extracted femurs were placed in 0.05% glutaraldehyde at 4⬚C for 24 h, rinsed in phosphate buffered saline (PBS) for 24 h, then decalcified in 0.25 M EDTA at 4⬚C for 72 h with daily EDTA changes. Each femur was then immersed in 10 mL of 1M 5-bromo-4-chloro-3-indolyl-D-galactopyranoside (XGAL) (Boehringer Mannheim, Indianopolis, IN) in PBS with 20 mM potassium ferricyanide, 20 mM potassium ferrocyanide, and 1 mM magnesium chloride at 37⬚C for 24 h. The femurs were then removed from the XGAL staining solution, rinsed with PBS for 24 h, and then embedded in paraffin for histologic analysis. Sagittal sections which were 5 m thick were cut using a microtome (American Optical/Leica, Deerfield, IL), mounted on slides amd stained with hematoxylin and eosin. The sections were examined manually and also by digital histomorphometric analysis using a Dell PC and ImagePro software. In a blind trial, the number of blue staining cells and the relative areas of cartilage, fibrous stroma, fat, and bone were quantified.
D. Molecular Detection of Cells DNA-PCR was performed to examine the presence of neomycin resistance at the site of the femoral defect to identify the presence of the D1-IGF cells. The femurs were placed in digestion buffer containing 100mM NaCl, 100 mM Tris-Cl, 25 mM EDTA, 0.5% sodium dodecyl sulfate, and 0.1 mg/mL of proteinase K (Fischer Scientific, Fair Lawn, NJ). The tissue was homogenized (T25 Basic Disperser, IKA-works, Wilmington, NC), and DNA was extracted by incubating in a 50⬚C waterbath for 1 h followed by the addition of phenolcholoroform mixture (1:1) to layer out the DNA to organic matter. The supernatant of the solution was combined with 10 M ammonium acetate and DNA was precipitated out with 7.3 mL of 100% ethanol per milliliter of sample. The sample was centrifuged and suspended in Tris:EDTA (TE) buffer. The amounts of material were quantified and normalized to allow subsequent quantitative analysis by establishing the purity of DNA between 260 and 280 nm on a Lambda 3A spectrophotometer (Perkin-Elmer, Foster City, CA). DNA was amplified by polymerase chain reaction using primers for the neomycin resistance gene (5′ ⳱ CTTTTTGTCAAGACCGACC; 3′ ⳱ GATGTTTCGCTTGGTGG). PCR conditions were 10 min at 95⬚C, followed by 45 cycles with 55⬚C for the annealing step and 72⬚C for polymerization, using AmpliTaq Gold DNA polymerase (Perkin-Elmer). The resultant 280bp segments at a concentration of 100 ng/ml were analyzed by agarose gel electrophoresis in 0.1 M Trizma base (Sigma):boric acid (Fisher):EDTA buffer. Multiple PCR runs were performed comparing each experimental group at each time point. The PCR gels were examined and digitized using a Hewlett-Packard scanner directly into a Dell PC. The relative optical density of the bands were quantified using ImagePro software. All quantified data were statistically and graphically evaluated using Microsoft Excel software.
Growth Factor Delivery by a Cloned Multipotent Cell
467
III. RESULTS A. Histological Analysis of Bone Healing Histological study with XGAL stain for -galactosidase activity was performed on specimens from each experimental group at 2, 4, and 8 weeks. Representative sections are shown in Fig. 2. Cells with blue staining have -galactosidase activity and represent the injected D1BAG cells. A discussion of the salient comparative histological findings ensues. At 2 weeks, the control group is characterized by abundant chondrocytes and fairly disorganized cellular arrangement. There are no blue staining cells evident. Overall, this represents the expected histology for 2 weeks postinjury. The D1-BAG group reveals a predominance of chondrocytes and cartilaginous matrix filling the injury repair site, but early bone formation is evident at the periphery of the injury repair tissue. Blue staining cells are evident and appear to be localized to the areas of new bone formation. The D1-IGF group demonstrates the predominance of early bone formation with visible lacunae and a relative paucity of chondrocytes and cartilage matrix. There are very few blue staining cells, representing endogenous galactosidase activity. Thus, the control consists mainly of cartilage, whereas the D1-BAG group consists of a mixture of cartilaginous and osseous tissue, and the D1-IGF group consists primarily of osseous tissue. This suggests that the experimental groups progress through the stages of bone healing at a faster rate to reach a more mature, and therefore more osseous, stage in less time.
Figure 2 Histology of repair tissue.
468
Azer et al.
At 4 weeks, there continues to be an abundance of chondrocytes and cartilaginous matrix, as well as no osseous tissue and no blue staining cells in the injury repair tissue of the control group. The D1-BAG injected specimens reveal bone formation with lacunae present. Blue staining cells are present in the injury repair tissue. Upon closer examination (Fig. 3), it appears that the blue cells are localized at the periphery of the injury site and are surrounding only the areas of new bone formation. The center of the injury repair tissue is filled with cartilage and does not reveal any blue staining cells, suggesting that not only do the D1-BAG cells localize to the injury site, but they are actively participating in endochondral ossification. The D1-IGF group consists of abundant osseous tissue with few chondrocytes and little cartilaginous matrix. At 8 weeks, the control specimens are characterized by the prevalence of osseous tissue and paucity of chondrocytes and cartilaginous matrix. The D1-BAG injected specimens also demonstrate predominantly osseous tissue with haversian canals present, filled with blue staining cells. The D1-IGF injected group demonstrates mature bone with haversian canals present.
Figure 3 Temporal changes in bone repair. (Top row 2 weeks, middle row 4 weeks, and bottom row 8 weeks.)
Growth Factor Delivery by a Cloned Multipotent Cell
469
Table 1 Temporal Decline in XGAL Positive Cells Number of XGAL positive cells/HPFa Time (weeks) 2 4 8
1
2
3
4
5
Average
SD
142 86 15
160 45 18
80 78 17
135 102 21
153 120 24
134* 86.2* 19*
31.7 28.1 3.5
a
HPF⫽high power field. * p⬍0.05.
Hence, comparative histological study indicates that the control and experimental groups consist predominantly of mature bone at 8 weeks postinjury. B. Quantification of Transplanted Cells The number of blue staining cells in the D1-BAG injected specimens was quantified using computerized histomorphometric analysis. The results are illustrated in Table 1. There is a statistically significant linear decline in the number of blue staining cells with time. This suggests that the injected cells do not continue to express their phenotype in the injury repair tissue, possibly because the cells’ existence in the injury repair site is transient. C. Molecular Analysis Molecular analysis of the injury repair tissue from each group reveals the presence of the neomycin-resistant band at 280bp for the mice injected with D1-BAG and D1-IGF cells at each time point. At all time points, there was no band present for the control group. Figure 4 represents the data acquired when optical densitometry studies were performed on the DNA-PCR bands. The relative optical densities, corresponding to the quantity of injected cells, were similar between the D1-BAG and D1-IGF groups at each time point, suggesting that the cells localized to the injury repair tissue in nearly equivalent amounts. There was a decline in the optical density measurements between each time period for the D1-BAG and D1-IGF injected specimens. The number of specimens in these analyses preclude the determination of the statistical significance. The data do however suggest that there is a decline in the relative neoexpression with time.
Figure 4 Percent bone in repair tissue.
470
Azer et al.
The histological and molecular examinations suggest that the injected cells are present at the injury site and visibly appear to be participating in the bone repair process. By the declining number of blue staining cells and concomitant decline in relative amounts of neo-DNA, it appears as though the injected cells do not continue to express their phenotype in the area of injury repair. Rather, it appears that their presence in the injury repair site is transient, with fewer cells present after osseous tissue has filled the defect site. D. Histomorphometric Analysis of Bone Healing To examine the effect that the injected cells had on bone formation, histomorphometric analysis was performed to quantify the relative percent of osseous tissue in the repair tissue. These data suggest that the experimental groups had a favorable effect on the defect repair process. The D1-BAG and D1-IGF groups were able to initiate and progress through the stages of bone healing at a faster rate to achieve a more mature, thus more osseous, stage in less time than controls. This implies that the mice injected with mesenchymal stem cells have enhanced healing of their cortical defects.
IV. DISCUSSION Our results demonstrate that cloned pluripotential mesenchymal stem cells will localize to the site of bone injury and participate in the bone healing process when administered systemically. The administration of these cells has a favorable effect on the histological appearance of the bone defect repair tissue. When these cells were modified to provide increased expression of insulin-like growth factor-I, they also localized to the bone defect site and had a favorable effect on the histological appearance of the bone defect repair tissue. In fact, the cells with IGF-1 appear to accelerate the bone healing process. This study adds to the extensive literature that supports the utility of mesenchymal stem cells in the treatment of musculoskeletal disorders including bone defects [14,28,29]. The cells themselves may have therapeutic implications by increasing the number of available progenitor cells, by creating a favorable hormonal and cell signaling milieu, or by functioning as a vehicle for localized gene delivery. A. Cell-Mediated Gene Delivery in Bone Repair Bone defect healing is a complex process which involves the accurate interaction of numerous cell types in coordinated intramembranous ossification, endochondral ossification, and remodeling. Osteoprogenitor cells are central to these processes [30]. There are numerous studies demonstrating the clinical utility of cloned mesenchymal stem cells in the treatment of bone defects and nonunions and as fracture healing enhancers [28,29]. Bruder showed that implantation of purified, culture expanded mesenchymal stem cells heals critically sized bone defects when delivered locally in porous hydroxyapatite/tricalcium phosphate implants [29]. Connolly has stimulated osteogenesis in bone grafts by using marrow osteoprogenitor cells aspirated from the iliac crest [28]. We found that the injected mesenchymal stem cells localized to the bone defect sites and participated in the bone healing. In fact, our histological study reveals that the transplanted cells are closely approximated to the new bone formation, participating in the endochondral ossification and that some become incorporated in the bone as osteocytes. Histomorphometric analysis suggests that mesenchymal stem cell transplantation accelerates bone healing providing more mature osseous tissue at earlier time points. Our research supports the literature describing the efficacy of mesenchymal stem cells, expanded in vitro, in localizing to, participat-
Growth Factor Delivery by a Cloned Multipotent Cell
471
ing in, and impacting upon the repair of bone defects. The declining number of cells with time suggests either that the cells are no longer present in the defect site as a result of humoral or cell-mediated immunity, apoptosis, or cell inhibition which shuts off the cells’ phenotypic expression once the bone healing process is well under way. Nevertheless, further research, using cellular labeling techniques and immunobiology are needed to define the ultimate fate of the transplanted cells before they may be used safely in a clinical setting. B. Clinical Implications of Gene Therapy Gene therapy has the potential for limitless clinical applications in orthopedic surgery [31]. Bone defect healing lends itself well to gene therapy because of the temporal and localized need for enhancing factors such as bone morphogenetic proteins, insulin-like growth factor-I, fibroblast growth factors, and transforming growth factor-beta [32]. Gene delivery not only provides a sustained higher concentration of key proteins locally in the defect, but also can lead to a greater biological responsiveness of the cells [20]. However, an efficient means of delivering such factors to the site where they are needed is sought [33,34]. At present, two strategies are incorporated, one where bone marrow cells are isolated and transfected in vitro and another which involves the in vivo infection of cells with viral vectors or activated gene matrices [20]. The ideal vehicle has yet to be identified. In our study, we found that cloned mesenchymal stem cells transfected in vitro with IGF-1 reliably homed to the area of the bone defect repair, participated in that repair, and appeared to be cleared from that area with time. We also found markedly accelerated bone healing where the defects were filled with more mature bone at earlier time points. Hence, the genetic information was transferred to the correct spatial and temporal location, suggesting that cloned stromal cells may be a viable option for gene delivery as further research evolves in this area. Our stromal cells not only provide a local source of growth factors, but also aid in bone healing by providing the cellular framework needed to support that repair. Further study is needed to confirm that the local titers of IGF-1 are increased at the defect site when D1-IGF cells are used to deliver insulin-like growth factor to the defect site. Nevertheless, accurate delivery in a spatial and temporal manner provides the framework for further research in the promising arena of cellular and molecular therapies for the significant clinical problem of bone defect healing. C. Biological Delivery Systems Having such a biological vehicle that reliably localizes to a defect site and participates and enhances bone formation while not persisting indefinitely is of tremendous clinical value. These mesenchymal cells are excellent delivery vehicles because they are reliable, are biological, and do not interfere with the natural healing processes, as do other substances such as poly(␣hydroxyacids) or foreign bodies [35]. This research will benefit the study of any condition that benefits from enhanced bone healing or the local delivery of any substance, such as genes, growth factors, or pharmaceuticals. It also provides further understanding of the effect of gene therapy on injury repair as well as further reiterating our understanding of the basic bone healing and repair process. A solid understanding of the basic behavior and characteristic of these cells paves the way for limitless further clinical studies. REFERENCES 1. Lane J, Sandhu H. Current approaches to experimental bone grafting. Orthop Clin North Am 1997; 18:213–225.
472
Azer et al.
2. Gregory CF. The current status of bone and joint transplants. Clin Orthop Rel Res 1972; 87:156–166. 3. Enneking WF. Observations on massive retreived human allografts. J. Bone Joint Surg 1991; 73A: 1123–1142. 4. Yaszemski MJ. Evolution of bone transplantation: molecular, cellular, and tissue strategies to engineer human bone. Biomaterials 1996; 17:175–185. 5. Caplan AI. Mesenchymal stem cells. J. Orthop Res 1991; 9:641–650. 6. Jaiswal N. Osteogenic differentiation of purified culture-expanded human mesenchymal stem cells in vitro. J. Cell Biochem 1997; 64:295–312. 7. Cui Q. Steroid-induced adipogenesis in a pluripotential cell line from bone marrow. J. Bone Joint Surg 1997; 79A:1054–1063. 8. Prockop DJ. Marrow stromal cells as stem cells for nonhematopoietic tissues. Science 1997; 276: 71–74. 9. Haynesworth SE. Characterization of cells with osteogenic potential from human marrow. Bone 1992; 13:81–88. 10. Wakitani S. Mesenchymal cell–based repair of large full thickness defects of articular cartilage. J. Bone Joint Surg 1994; 76:579–592. 11. Dennis JE. Osteogenesis in marrow derived mesenchymal cell porous ceramic composites transplanted subcutaneously: effect of fibronectin and laminin on cell retention and rate of osteogenic expression. Cell Transplantation 1992; 1:23–32. 12. Bruder SP. Mesenchymal stem cells in bone development, bone repair, skeletal regeneration therapy. J. Cell Biochem 1994; 56:283–294. 13. Bennett JH. Adipocytic cells cultured from marrow have osteogenic potential. J. Cell Sci 1991; 99: 131–139. 14. Lazarus HM. Ex vivo expansion and subsequent infusion of human bone marrow–derived stromal progenitor cells: implications for therapeutic use. Bone Marrow Transplantation 1995; 16:557–564. 15. Diduch DR. Two cell lines from bone marrow that differ in terms of collagen synthesis, osteogenic characteristics, and matrix mineralization. J. Bone Joint Surg 1993; 75A:92–105. 16. Dahir GA. Pluripotential mesenchymal cells repopulate bone marrow and retain osteogenic properties. Clin Ortho Rel Res 2000; 379S:S134–S145. 17. Su CC. Bone marrow osteoprogenitor cells adhere to titanium implants. Transactions of Orthopaedic Research Society 1996; 91:16. 18. Devine MJ. Transplanted bone marrow cells localize to fracture callus in a mouse model. J Orthop Res 2002; 20:1232–1239. 19. Niyibizi C. Potential role for gene therapy in enhancement of fracture healing. Clin Orthop Rel Res 1998; 355-S:148–153. 20. Mohan S. Bone growth factors. Clin Orth Rel Res 1991; 263:30–48. 21. Linkhart TA. Growth factors for bone growth and repair: IGF, TGF-B, and BMP. Bone 1996; 19S:1–12. 22. Andrew JG. Insulin-like growth factor gene expression in human fracture callus. Calcif. Tiss. Int 1993; 53:97–102. 23. Baylink DJ. Growth Factors to stimulate bone formation. J Bone Mineral Res 1993; 8:S565–S572. 24. Ammann P. IGF-1 and pamidronate increase bone mineral density in ovariectomized adult rats. Am J Physiol 1993; 265:770–776. 25. Ebeling PR. Short-term effects of recombinant human insulin-like growth factor I on bone turnover in normal women. J Clin Endocrinol Metab 1993; 77:1384–1387. 26. Shen F. Systematically administered mesenchymal cells transduced with insulin-like growth factor1 localize to a fracture site and potentiate healing. J Orthop Trauma 2000; 16(9):651–9. 27. Price J. Retroviruses and the study of cell lineage. Development 1987; 101:409–419. 28. Connolly JF. Clinical use of marrow osteoprogenitor cells to stimulate osteogenesis. Clin Orth Rel Res 1998; 355S:257–266. 29. Bruder SP. Mesenchymal stem cells in osteobiology and applied bone regeneration. Clin Orth Rel Res 1998; 355S:247–256. 30. Einhorn TA. The cell and molecular biology of fracture healing. Clin Orth Rel Res 1998; 355S: 7–21.
Growth Factor Delivery by a Cloned Multipotent Cell
473
31. Goldstein SA. Potential role for direct gene transfer in the enhancement of fracture healing. Clin Orth Rel Res 1998; 355S:154–162. 32. Rosier RN. Regional gene therapy. Clin Orth Rel Res 1998; 355S:361–363. 33. Felger PL. Nonviral strategies for gene therapy. Sci Am 1997; 6:102–106. 34. Firedmann T. Overcoming the obstacles of gene therapy. Sci Am 1997; 6:96–101. 35. Hollinger J. Poly-␣-hydroxyacids: carrier for bone morphogenetic proteins. Biomaterials 1996; 17: 187–194.
21 New Synthetic Biodegradable Polymers for Bone Morphogenetic Protein Delivery Systems Naoto Saito, Hiroshi Horiuchi, Narumichi Murakami, and Jun Takahashi Shinshu University School of Medicine Matsumoto, Japan
Takao Okada Taki Chemical Co., Ltd. Kakogawa, Japan
Kazutoshi Nozaki Yamanouch Pharmaceutical Co., Ltd. Tsukuba, Japan
Kunio Takaoka Osaka City University Graduate School of Medicine Osaka, Japan
I. INTRODUCTION The regenerating potential of human bone appears to be limited since repair of large bone defects, which are often associated with comminuted open fractures or resection of bone tumors, usually remain unrepaired. For the treatment of such cases, autogeneic or allogeneic bone grafting has been routinely indicated [1]. However, major problems associated with autogeneic bone grafting include a limited source of donor bone and high morbidity resulting from additional surgery for procurement of graft. In the case of allogeneic bone grafting, major concerns are potential risks of transfer of diseases, immunological reactions from hosts, poor osteogenic capacity of the transplanted bone, and high costs associated with a bone banking system [2]. In order to address these issues, new approaches using alternative techniques are needed. Such approaches may include use of molecules or genes involved in the osteoblastic differentiation process and the control of bone formation [3–6]. Since gene therapy is still in the early stages of development, the use of cytokines with the capacity to elicit new bone formation may currently be the most promising. This technique requires the supply of an adequate amount of cytokines produced by DNA recombination and an appropriate delivery system to achieve maximal biological activity of the cytokines at the implanted site. Currently, the most promising cytokines for the induction of new bone formation are bone morphogenetic proteins (BMPs) [7,8]. 475
476
Saito et al.
II. BONE MORPHOGENETIC PROTEINS Bone Morphogenetic Proteins are a group of proteins with the special biological ability to induce new bone formation under in vivo conditions. The biological activity of BMP was originally detected in decalcified bone matrix in an experimental model of ectopic bone formation [9–11]. Under physiological conditions, the BMP molecules are produced by osteoblasts and retained in bone matrix [12–14]. Production of BMPs by cells in the periosteum at the fracture site is thought to be up-regulated in the early stages of fracture repair, and therefore to contribute to new bone (callus) formation [11]. A common feature of the BMP molecules is the position of cysteine residues as determined from the carboxyl terminus. Since the positions of these seven cysteine residues are the same as those of TGF-, this indicates that BMP molecules are members of TGF- superfamily [15]. To date, several BMP molecules have been successfully synthesized by DNA recombination [16], and the protein products (rBMP) have been proven to retain the bone inducting effect of BMP. Thus, several human-type BMPs (BMP-2,-4,-7 and GDF-5) have become available for potential medical use. A. Delivery Systems for BMPs Bone formation induced by BMPs is basically a local biological event, but it has been found that injection of a small amount of BMP into muscles elicits no new bone formation in situ, probably due to rapid diffusion of the BMP molecules away from the injected site [16]. Therefore, a delivery system is essential to retain BMP at the implanted site for the required period of time and to release BMP at an optimal rate. Another important function of the delivery system is control of the induced bone mass by providing scaffolding for bone formation. Previous experiments have indicated that the shape and size of the BMP-induced bone mass depend on those of the carrier and are almost identical to those of the carrier when the carrier materials are optimized. The physicochemical and biological requirements for such an optimal carrier are (1) to be nontoxic and nonimmunogenic (not cause inflammation), (2) to remain insoluble under physiological conditions, (3) to be degradable in host tissue in a few weeks, and (4) to be easily molded. Animal-derived collagen was thought to meet the requirements and has been routinely utilized as a BMP carrier in experimental systems [16,18,19]. But in terms of clinical use, such collagen may pose potential risks of disease transfer or unexpected inflammation due to residual immunogenicity of the xenogeneic collagen molecules when implanted into humans [20–22]. In order to avoid these risks, synthetic biodegradable polymers would be preferable to tissuederived materials such as collagen [23]. 1. Synthetic Biodegradable Polymers for BMP-Delivery Systems Polylactic Acid. In a previous study, we tried using polylactic acid (PLA) polymers as a BMP carrier (Fig. 1A). However, the physicochemical characteristics as well as the biological reactions of PLA when implanted in animals varied significantly depending on the molecular size of the PLA polymers. Only one polymer with a low molecular weight of 650 (PLA 650) and the characteristics of a viscous liquid functioned as BMP carrier. However, its efficacy was not satisfactory because the BMP-induced ossicles were much smaller than the original PLA/ BMP composite implants, probably due to too rapid degradation and mild toxicity resulting from the marked acidity of the low molecular weight PLA [24].
New Polymers for Bone Morphogenetic Protein Delivery
477
Figure 1 Structural formula of three types of polymers. m, n, o, p, q: number of units. (A) Poly-D,Llactic acid homopolymer (PLA). (B) Poly-D,L-lactic acid–polyethylene glycol block copolymer (PLAPEG). (C) Poly-D,L-lactic acid–para-dioxanone–polyethylene glycol block copolymer (PLA-DX-PEG).
Poly-D,L-Lactic Acid–Polyethylene Glycol Block Copolymer. To overcome these problems, a hydrophilic polyethylene glycol (PEG) polymer with a molecular weight of 200 was linked to the carboxyl radical of PLA 650 (PLA 650/PEG 200 block copolymer), resulting in elimination of acidity and slowing down of degradation due to the larger molecular size, but without changing, the semiliquid nature of the polymer (Fig. 1B). Bone yield showed much improvement when this polymer was implanted together with BMP. Although this PLA-PEG block copolymer is highly viscous and thus injectable, it cannot be molded and keeps its threedimensional configuration [25]. We therefore developed a new BMP carrier polymer with plasticity at room temperature, which was obtained by synthesizing PLA-PEG-PLA block copolymers with various molecular sizes and PLA/PEG ratios (Fig. 2). The results were assessed in vivo, i.e., each polymer was mixed with rhBMP-2 and implanted into the back muscles of mice for 3 weeks to assess their capacity to generate new ectopic bone. The results showed the superiority of a PLA-PEG block copolymer with a total molecular weight of approximately 9500 and a PLA/PEG molar ratio of 59: 41 (PLA 6500/PEG 3000). Although this polymer worked well as a carrier for BMP, and new ectopic bone was consistently induced with BMP, the degradation rate of the polymer appeared to be a little slow and was seen to remain in the center of the BMP-induced ossicles [26,27]. Poly-D,L-Lactic Acid–para-Dioxanone–Polyethylene Glycol Block Copolymer. In order to optimize degradation of the polymer as carrier for BMP, para-dioxanone (DX) molecules were randomly inserted into the PLA segments of the PLA 6500/PEG 3000 polymer without changing the total molecular weight. The molecular weights of the components (PLA/DX/PEG) of the new polymer were 5000/1500/3000 (Fig. 1C). The use of this modified polymer as a carrier for BMP resulted in the complete replacement of the implants by new bone without any visible remnants of the carrier polymer, probably due to the optimal rate of degradation of the polymer (Fig. 3). An in vivo assay showed that the PLA-DX-PEG/rhBMP-2 composite implants can induce new bone formation more effectively than either the PLA-PEG/rhBMP-2 or the collagen/rhBMP-2 composites (Fig. 4) [28]. 2. Combination of the BMP/Polymer Composites with Biomaterials Due to the hydrophobic nature of the PLA-DX-PEG polymer, it swells on contact with water. This physical property provides additional advantage for practical use of the polymer in combina-
478
Saito et al.
Figure 2 Appearance of the PLA 6500–PEG 3000 implant: waxy and doughy at room temperature.
Figure 3 Loss of mass as a function of degradation time for PLA-DX-PEG and PLA-PEG polymers. Five hundred milligrams of the polymer samples was immersed in 25 mL of PBS and kept in a water bath at 37⬚C for 23 days. The polymers were then vacuum-dried, and the weight of the residual polymers was measured. The degradation rate was represented by the weight of the remainder as a percentage of the original weight. (From Ref. 28)
New Polymers for Bone Morphogenetic Protein Delivery
479
Figure 4 Soft x-ray photograph and light micrograph of new bone formed 3 weeks after implantation of the PLA-DX-PEG implant with 10 g of rhBMP-2. (A) A trabecular pattern can be observed in the radiopaque area. (B) New bone formation with hematopoietic marrow and bony trabeculae. Hematoxylin and eosin stain. (From Ref. 28.)
tion with porous biomaterials. When the solid implant with its pores filled with the BMP/polymer composite is implanted, the composite will swell, ooze out from the pores, and form a layer of BMP/polymer composite. This layer induces new bone that covers the surface of the biomaterials and encases the biomaterial. Figure 5 shows an example of bone induction on the surface of a block of porous hydroxyapatite combined with BMP-2. The special characteristics of this delivery system were thought to be useful for the repair of large bone defects if combined with biomaterials. Recombinant human-type BMP-2 (120 g) was therefore mixed in with the polymer (120 mg) and impregnated into titanium fiber mesh cylinders. Three 5-mm cylinders were placed end to end to fill a 15-mm defect created in the humeri of adult rabbits and stabilized with an intra-medullary rod. In controls, the titanium fibermesh cylinders were combined with the polymer but without BMP. Six weeks after implantation, new bone had formed on the surface of the implant and had bridged the defect (Fig. 6) [29]. These results provide strong evidence that these new composite implants, combining rhBMP, synthetic degradable polymers, and compatible biomaterials, provide enhanced regenerative potential for the repair of large bone defects.
III. CLINICAL APPLICATIONS The novel delivery system described here is expected to eventually be utilized for clinical purposes, although its safety as well as efficacy for humans must be rigorously tested. The safety of the polymer is not expected to be problematic because PLA, DX, and PEG fragments have already been accepted without difficulties in humans [30–32]. Neverthless further tests in large animals or primates are essential before this bone-inducing implant can be used in a clinical setting.
480
Saito et al.
Figure 5 Example of bone induction on the surface of a block of porous hydroxyapatite combined with rhBMP-2 and the plastic and swelling-prone carrier polymer. The composite was implanted into the dorsal muscles of a mouse for 3 weeks. A considerable volume of bone covers the porous HA block. (From Ref. 27.)
Figure 6 Soft x-ray photograph of the entire humerus 6 weeks after surgery. (A) In controls, a small amount of new bone was seen at both ends of the defect with no evidence of bridging. (B) The defect treated with implants containing 120 g of rhBMP-2 shows radiographic union. (From Ref. 29.)
New Polymers for Bone Morphogenetic Protein Delivery
481
REFERENCES
1. Einhorn TS. Enhancement of fracture healing. J Bone Joint Surg 1995; 77-A:940–56. 2. Malinin TI, Martinez OV, Brown MD. Banking of massive osteoarticular and intercalary bone allografts—12 years’ experience. Clin Orthop 1985; 197:44–57. 3. Fang J, Zhu YY, Smiley E, Bonadio J, Rouleau JP, Goldstein SA, McCauley LK, Davidson BL, Roessler BJ. Stimulation of new bone formation by direct transfer of osteogenic plasmid gene. Proc.Natl.Acad.Sci.USA 1996; 93:5753–5758. 4. Bonadio J, Smiley E, Patil P, Goldstein S. Localized, direct plasmid gene delivery in vivo: prolonged therapy results in reproducible tissue regeneration. Nat Med 1999; 5:753–9. 5. Caplan AI. Mesenchymal stem cells. J Orthop Res 1991; 9:641–650. 6. Haynesworth SE, Goshima J, Goldberg VM, Caplan AI. Characterization of cells with osteogenic potential from human marrow. Bone 1992; 13:81–88. 7. Reddi AH. Symbiosis of biotechnology and biomaterials: application in tissue engineering of bone and cartilage. J Cell Biochem 1994; 56:192–195. 8. Ripamonti U, Reddi AH, Tissue engineering. morphogenesis, and regeneration of the periodontal tissues by bone morphogenetic proteins. Crit Rev Oral Biol Med 1997; 8:154–163. 9. Urist MR. Bone formation by autoinduction. Science 1965; 150:893–899. 10. Urist MR, Iwata H. Preservation and biodegradation of the morphogenetic property of bone matrix. J Theor Biol 1973; 38:155–161. 11. Canalis E. Effect of growth factors on bone cell replication and differentiation. Clin Orthop 1985; 193:246–263. 12. Amitani K, Nakata Y, Stevens J. Bone induction by lyphilized osteosarcoma in mice. Calcif Tiss Res 1974; 16:305–313. 13. Takaoka K, Yoshikawa H, Hashimoto J, Miyamoto S, Masuhara K, Nakahara H, Matsui M, Ono K. Purification and characterization of a bone-inducing protein from a murine osteosarcoma (Dunn type). Clin Orthop 1993; 292:122–129. 14. Takaoka K, Yoshikawa H, Hashimoto J, Masuhara K, Miyamoto S, Suzuki S, Ono K, Matsui M, Oikawa S, Tsuruoka N, Tawaragi Y, Inuzuka C, Katayama T, Sugiyama M, Tsujimoto M, Nakanishi T, Nakazatio H. Gene cloning and expression of a bone morphogenetic protein derived from a murine osteosarcoma. Clin Orthop 1994; 294:344–352. 15. Celeste AJ, Iannazzi JA, Taylor RC, Hewick RM, Rosen V, Wang EA, Wozney JM. Identification of transforming growth factor- family members present in bone inductive protein purified from bovine bone. Proc Natl Acad Sci USA 1990; 87:9843–9847. 16. Takaoka K, Koezuka M, Nakahara H. Telopeptide-depleted bovine skin collagen as a carrier for bone morphogenetic protein. J Orthop Res 1991; 9:902–907. 17. Wang EA, Rosen V. Recombinant human bone morphogenetic protein induces bone formation. Proc Natl Acad Sci USA 1990; 87:2220–2224. 18. Aspenberg P, Turek T. BMP-2 for intramuscular bone induction: effect in squirrel monkeys is dependent on implantation site. Acta Orthop Scand 1996; 67:3–6. 19. Itoh H, Ebara S, Kamimura M, Tateiwa Y, Kinoshita T, Yuzawa Y, Takaoka K. Experimental spainal fusion with use of recombinant human bone morphogenetic protein 2. Spine 1999; 24:1402–1405. 20. Butler D, Wadman M, Lehrman S, Schiermeier Q. Last chance to stop and think on risks of xenotransplants. Nature 1998; 391:320–324. 21. Bach FH, Fishman JA, Daniels N, Proimos J, Anderson B, Carpenter CB, Forrow L, Robson SC, Fineberg HV. Uncertainty in xenotransplantation: individual benefit versus collective risk. Nat Med 1998; 4:141–144. 22. DeLustro F, Dasch J, Keefe J, Ellingsworth L. Immune responses to allogeneic and xenogeneic implants of collagen and collagen derivatives. Clin Orthop 1990; 260:263–279. 23. Hollinger JO, Leong K. Poly(␣-hydroxy acids): carriers for bone morphogenetic proteins. Biomaterials 1996; 17:187–194.
482
Saito et al.
24. Miyamoto S, Takaoka K, Okada T, Yoshikawa H, Hashimoto J, Suzuki S, Ono K. Evaluation of polylactic acid homopolymers as carriers for bone morphogenetic protein. Clin Orthop 1992; 278: 274–285. 25. Miyamoto S, Takaoka K, Okada T, Yoshikawa H, Hashimoto J, Suzuki S, Ono K. Polylactic acid–polyethylene glycol block copolymer; a new biodegradable synthetic carrier for bone morphogenetic protein. Clin. Orthop 1993; 294:333–343. 26. Saito N, Okada T, Toba S, Miyamoto S, Takaoka K. New synthetic absorbable polymers as BMPcarriers: plastic properties of poly-D,L-lactic acid–polyethylene glycol block copolymers. J Biomed Mater Res 1999; 47:104–10. 27. Saito N, Okada T, Horiuchi H, Murakami N, Takahashi J, Nawata M, Ota H, Miyamoto S, Nozaki K, Takaoka K. Biodegradable poly-D,L-lactic acid–polyethylene glycol block copolymers as a BMP delivery system for inducing bone. J Bone Joint Surg Am 2001; 83(Suppl 1):92–98. 28. Saito N, Okada T, Horiuchi H, Murakami N, Takahashi J, Nawata M, Ota H, Nozaki K, Takaoka K. A biodegradable polymer as a cytokine delivery system for inducing bone formation. Nat. Biotechnol 2001; 19:332–335. 29. Murakami N, Saito N, Horiushi H, Okada T, Nozaki K, Takaoka K. Repair of segmental defects in rabbit humeri with titanium fiber mesh cylinders containing recombinant human bone morphogenetic protein-2 (rhBMP-2) and a synthetic polymer. J Biomed Mater Res(). 30. Zhu KJ, Xiangzhou L, Shilin Y. Preparation, characterization and properties of polylactide (PLA)–poly(ethylene glycol) (PEG) copolymers: a potential drug carrier. J Appl Polym Sci 1990; 39:1–9. 31. Jeong B, Bae YH, Lee DS, Kim SW. Biodegradable block copolymers as injectable drug-delivery systems. Nature 1997; 388:860–862. 32. Ray JA, Doddi N, Regula D, Williams JA, Melveger A. Polydioxanone (PDS), a novel monofilament synthetic absorbable suture. Surg Gynecol Obstet 1981; 153:497–507.
22 Nanosized Biosensors and Delivery Vehicles Agnes E. Ostafin and Joel P. Burgess University of Notre Dame Notre Dame, Indiana, U.S.A.
Hiroshi Mizukami Wayne State University Detroit, Michigan, U.S.A.
I. INTRODUCTION The explosion of nanoscience as a scientific discipline has brought about revolutionary changes in the way we think about materials and devices for use in all aspects of science and engineering. As we become able to synthesize, control, and deliver nanosized doses of drugs, and target them to specific sites in the human body, we will discover better, more reliable and cost effective treatments and diagnoses. These advances in medical science and pharmacology are sure to improve the quality and lengthen the span of our lives for years to come. For example, nanoparticles loaded with highly fluorescent dyes or paramagnetic compounds may travel to specific target sites in the human body to detect diseases or enhance MRI and photonic images. In this chapter we discuss a novel process of making nanosized particles containing highly fluorescent dyes. Fluorescent nanoprobes used in life sciences are divided into two major types based on whether they are protected or unprotected in the body. Unprotected nanoprobes include dye-labeled antibodies, proteins, and cells which are exposed to biodegradative and immune system assault in living organisms. Aside from doubts about the survival of these molecules and cells in the body, only a few of them are safe for use in vivo, and often a trade-off between the diagnostic needs and patient discomfort must be made. On the other hand, protected nanoprobes are both safe and have prolonged life in vivo. As a group, they vary widely in their level of sophistication in eluding the body’s natural defenses. Advances in protein engineering and bioinformatics have enabled researchers to alter parts of macromolecules in order to hide an antigenic site [1] or to conjugate the probe with another protein to inhibit the host’s immune response [2–4]. Polymeric fluorescent nanoparticles formed via sol-gel methods from biocompatible materials, such as polyethylene glycol (PEG) [5–16], and liposomes [17–21] serve to protect fluorophores and drugs within the particle. Finally, genetic engineering of both prokaryotic and eukaryotic cells has led to the development of tracer cells that overexpress fluorescent proteins such as green fluorescent protein [22–27] or take advantage of cellular bioluminescence, such as the luciferase system [28–31]. 483
484
Ostafin et al.
With the availability of such a wide range of fluorescent nanoprobes the need for new fluorescent nanoparticles and delivery vehicles may be questioned. Some fluorescent nanoproducts suffer from low sensitivity, immunogenicity, or toxicity that limits their use in biological systems. More problematic is that there may be variability in the luminescence or molecule delivery properties of the particles when used in different cell types, tissues, or organs. Particles may have very short residence times in certain tissues or may aggregate or rupture, losing their utility. The chromophore may leach out or fluorescence quenchers leach in during use. New frontiers in biomedical research, such as rapid sensitive detection/treatment of biowarfare agents, detection/treatment of infectious and metastatic disease, biophotonic imaging and long-term biological monitoring, will require robust, selectively sensitive nanoprobes that avoid these problems. The aim of this chapter is to describe new developments in templated hollow fluorescent nanospheres that may be used for in vivo and in vitro detection and targeted delivery. Templated nanospheres are grown around a pre-existing nanosized particle that can be removed once the shell is formed (Fig. 1). Also called core-shell materials, numerous examples of nanosized inorganic composite materials [32–39] and several sol-gel organic–inorganic composites have been reported [40–43]. Unlike their solid sol-gel counterparts [44], hollow dye-filled inorganic nanospheres have the advantage of higher sensitivity per particle because the hollow interior can be filled with large numbers of chromophores whose luminescence remains unquenched despite a high local concentration [45]. These chromophores will then emit light in an environment protected by the material of the shell so that less material can be used to achieve the same results. The particles are also fluid filled and, depending on the porosity of the surrounding shell, interact with the environment and report on local conditions. The material of the shell is chemically stable in physiological conditions, and the porosity, survivability, and surface chemistry can be tailored. More than one type of indicator can be loaded into the same particle to enable different types of imaging using the same particles. The work described involves the synthesis and surface modification of dye-filled, hardened nanospheres made of silicate. Other types of nanospheres are also described and include highly porous silicate and also calcium phosphate–based particles. All the nanospheres introduced are strong, yet porous and could be packed, annealed, layered, or machined to produce devices such as sensors and implants.
II. METHODS The synthesis of fluorophore-filled nanosized particles starts with formation of colloidal gold. Gold colloids may be synthesized over a limited size range between about 10 and 50 nm in diameter using sodium citrate [46], and these particles may be grown to larger sizes using a seeded growth method [47]. Representative examples of both methods are described. The colloidal gold will then be coated with fluorescent dyes and crosslinkers, which allows silicate to enclose the gold–dye complex. The colloidal gold may be dissolved, and highly fluorescent nanosized particles with internal diameters determined by the colloidal gold used are the result. A number of variations are also described. A. Materials To construct the nanospheres described in this chapter, the following materials are needed: hydrogen tetrachloroaurate (III) hydrate (HAuCl4-䡠3H2O) (Sigma), sodium citrate (Na3C6H5O7䡠2H2O) (Fisher), hydroxylamine hydrochloride (NH2OH-HCl) (Sigma), Phosphatidylcholine (PC) (XVI-E) from egg yolk 99% pure (Sigma), 1,2 -dioleoyl-SN-glycero-3-phosphate
Nanosized Biosensors and Delivery Vehicles
485
(DOPA) (Avanti), 3-aminopropyl trimethoxysilane (APS) (Sigma), mercaptopropyltrimethoxysilane (MPTS) (Sigma), octyltrimethoxysilane (Gelest), sodium silicate solution [Na2O(SiO2)3, (27 wt% SiO2)] (Sigma), tetraethylorthosilicate (TEOS) (Sigma), tetramethylorthosilicate (TMOS) (Sigma), calcium nitrate [Ca(NO3)2] (Fisher), diammonium phosphate [(NH4)2HPO4] (Fisher), analytical grade ethanol (Sigma), analytical grade methanol (Fisher), reagent grade ammonium hydroxide (NH4OH, 29.7 wt% in H2O) (Fisher), sodium cyanide (NaCN) (Sigma), sodium chloride (Fisher), Cascade Blue hydrazide威 [(3,6,8-trisulfo-1-pyrenyl)oxy]-1-hydrazide, trilithium salt (CB) (Molecular Probes), fluorescein isothiocyanate (FITC) (Sigma), pyranine (Sigma), pyrene (Aldrich), cetyltrimethylammonium bromide (CTAB, 99.8 wt%) (Calbiochem, 3-(trieproxysilyl) propionitrile (Aldrich). All may be used as received without further purification. Bovine serum albumin (BSA), para-nitrophenyl phosphate (PNPP), and alkaline phosphatase (ALP)–conjugated donkey antigoat IgG may be obtained from Sigma. Deionized water (18 MOhm) (E-pure) is used for all preparations. Anodisc membrane filters (pore size 25 nm) may be obtained from Whatmann. Sephadex G-25 is obtained from Sigma. Polypropylene microcentrifuge and 50-mL conical tubes obtained from Fisher are used throughout the procedure except where indicated otherwise.
B. Synthesis of Colloidal Gold 1. 15-nm-Diameter Colloidal Gold The synthesis method for 15-nm-diameter colloidal gold has been described previously [46]. Briefly, 5 mL of 1% H2AuCl4 solution are added to 200 mL of water and heated to boiling while stirring in a thoroughly cleaned round bottom glass flask. Twenty-five milliliters of 1% sodium citrate solution at room temperature is added to the boiling solution while stirring vigorously. The stirring is continued until the solution becomes a deep red color, indicating that gold salt is reduced from its soluble Ⳮ3 oxidation state to form colloidal particles. The red color arises from the surface plasmons, whose optical absorbance is closely coupled with the overall particle size [48]. Once the suspension is cooled to room temperature, the total volume is adjusted to 250 mL using deionized water. Unreacted citrate or gold ions are removed by dialysis overnight against deionized water using Snake Skin威 Pleated Dialysis Tubing of 10,000 MWCO (Pierce). The average size of colloidal particles resulting from this synthesis is about 15 Ⳳ 5 nm in diameter as confirmed with a transmission electron microscope (TEM). In this manner, approximately 3.56 ⳯ 1015 particles per liter, or 5.93 ⳯ 10ⳮ9 M, colloidal gold will be obtained. The colloidal gold suspension should be stored in the dark at 4⬚C and warmed to room temperature before use. This suspension is stable for several weeks in the absence of any contamination. The negatively charged colloidal gold particles are stabilized by a layer of citrate in ionized form which helps to maintain particle–particle charge repulsion. The true number of citrate molecules on the surface of each gold particle is not known. 2. 65-nm-Diameter Colloidal Gold Formation of 65-nm-diameter colloidal gold with seeded growth requires the synthesis of the 15-nm-diameter colloidal gold described above. In brief, 0.83 mL of 0.2M NH2OH-HCl and 0.92 mL of 1% HAuCl4 solution are added under vigorous magnetic stirring at room temperature to 10 mL of the aqueous 15-nm colloidal gold stock solution that has been diluted to 100 mL. Continue to stir the solution for 15 min, then remove 25 mL of this material (25-nm-sized colloid), dilute to 100 mL with water, and add 0.562 mL of 0.2M NH2OH-HCl while stirring. Five minutes later, 0.95 mL of 1% HAuCl4 is added dropwise under vigorous stirring, and the
486
Ostafin et al.
solution stirred for 15 min. Next, dilute 55 mL of this solution (45-nm-sized colloids) to 105 mL with water, and add 0.375 mL of 0.2M NH2OH-HCl while stirring. Five minutes later, add 0.95 mL of 1% HAuCl4 dropwise under vigorous stirring, and then stir the solution for 15 min. After 20 min of stirring in a thoroughly cleaned glass beaker, the sample is dialyzed against deionized water at pH 7.0 to remove unreacted gold and hydroxylamine. The size of the particles may be verified using TEM to be 65 Ⳳ 10 nm. This colloidal gold suspension can also be stored in the dark at 4⬚C, but should be warmed to room temperature before use. C. Colloidal Gold Templated Silica Nanoshells 1. Adsorption of Dye onto the Colloidal Gold To reduce degradation of fluorescent dyes, all manipulations should be carried out in dim light and the containers covered with aluminum foil. A 3 ⳯ 10ⳮ4M solution of dye such as CB or FITC is prepared in deioinized water, and its pH adjusted to 7.0 using 0.1 M NaOH or 0.1M HCl. About 600 L of dye solution are added to 50 mL of 15-nm-diameter colloidal gold suspension at room temperature under vigorous stirring. After 20 min, 150 L of 1.1 ⳯ 10ⳮ3 M APS solution is added. Not all added dye will be bound to colloidal particles during this procedure. It is possible to ascertain the actual amount of dye bound to the colloidal particles by centrifuging the suspension to pellet the colloidal gold and measuring the absorbance of the dye remaining in the supernatant. This procedure has been described elsewhere for the case of CB dye on 15-nm-diameter colloidal gold [45], and the reagent amounts given here should yield a colloidal gold suspension with approximately 200 CB dye molecules per particle. An upper limit to the number of dyes that can be attached to each 15-nm colloidal gold particle is about 480; however, room must be left for the APS crosslinker if silicate is to be deposited later. If too little APS is added, then no shells will form. If too much APS is added, then in the subsequent steps solid silicate particles rather than silicate-coated gold will appear. Different dye molecules have different binding interaction strengths with the gold and the maximum number of dyes that can be bound will vary. The chemical nature of the surface of 15-nm-diameter colloidal gold has been discussed by others [49]. In the case of CB or FITC it is not necessary to dialyze the suspension again to remove unbound dye and APS before the silicate is deposited. For 65-nm-diameter colloidal gold, the particle’s surface is stabilized with NH2OH instead of citrate, and it is slightly more susceptible to flocculation if too much dye or APS is added rapidly to the suspension. A typical procedure used to attach pyranine dye to these larger colloids is as follows: to 100 mL of as-synthesized 65-nm-diameter gold colloid (dialyzed against Epure water to 4.7 S), 1 mL of 5 mM pyranine in water is added with moderate stirring. After 30 min of continuous stirring, the pH of the colloid–dye suspension should be adjusted to pH 8.5 using 0.1M NaOH, and 22 L of 0.5 mM APS or MPTS (prepared by dilution of a 10%v/ v solution of APS or MPTS in methanol into E-pure water) added drop-wise with moderate stirring. It is not necessary to dialyze the suspension again to remove unbound dye and APS before the silicate is deposited. 2. Silicate Nanosphere Formation Using the Dye/APS Coated Gold Colloid As soon as 20 min after the dye and APS have been added, silicate can be deposited onto the 15-nm colloidal gold particles using Na2O(SiO2)3 [50–52] or onto 65-nm colloidal gold particles using TMOS. For 15-nm colloidal gold, adjust the suspension pH slowly to about 10.5 using 0.1M NaOH, and then add 5 mL of 0.27% sodium silicate solution while stirring rapidly. Keep the vessel covered with parafilm and aluminum foil, let the particles age while stirring for 24 h at room temperature. After this, add 200 mL of 100% ethanol and continue to stir at room
Nanosized Biosensors and Delivery Vehicles
487
temperature for another 24 h. If needed, the suspension may be concentrated to about 12 mL using a pressurized ultrafiltration cell (Amicon) using argon gas (Mettler) or by centrifugation for 30 min or longer at 9000 RPM. For 65-nm colloidal gold suspension prepared as above, add 400 mL of 1% TMOS dropwise while stirring. Add 100 mL of 1M TRIS at pH 8.0 drop-wise. Adjust pH to 9.0 with 1M NH3 then add an additional 400 mL of 10% TMOS. Stir the suspension for several minutes and adjust with NH3 as needed to keep pH near 9.0. Keep the vessel covered with parafilm and aluminum foil and stir for 2 h. Add an additional 400 mL of 10% TMOS, again adjusting the pH to 9.0. After 22, and again after 48 h, add 400 mL of 10% TMOS and adjust the pH to 9.0. Twenty fours hours after the final addition, centrifuge the suspension for at least 40 min at 4200 RPM. Resuspend the pellet in the same volume of E-pure water, using mild sonication if needed. The centrifugation is repeated and the particles resuspended into 10 mL of E-pure water. Particles are stored at 4⬚C in the dark until used. The thickness of the silicate coating may be increased if the described procedure is repeated. A series of silica-coated 65-nm diameter gold particles coated with one through eight layers of silicate are shown in Fig. 2. 3. Removing Colloidal Gold from the Inside of Silicate Nanospheres The following procedures must be performed in a fume hood. To dissolve the gold core, 1–3 mL of slightly acidic 0.1M NaCN are added to about 50 mL of the gold–silicate nanocomposite suspension and stirred gently. The dissolution of gold colloid from the shells may be confirmed by the loss of color in the suspension. Once the suspension is colorless, dialyze against deionized water overnight. Alternatively, 10 mL of concentrated gold–silicate suspension may be dialyzed against 400 mL of 4 mM slightly acidic NaCN changing the dialysis solution often until the particle suspension becomes colorless. Confirm the presence of nanosized silicate shells using TEM (Fig. 1A) and the dye spectrophotometrically (Fig. 3). By removing gold, the fluorescent dye regains its fluorescence. The fluorescence intensity is pH dependent, suggesting that the silicate shell is permeable to hydronium ions. 4. Fluorescence Amplification with the Nanospheres The absorption spectra of 8 ⳯ 10ⳮ6M CB in solution and 1 ⳯ 10ⳮ7M in the shell suspension at pH 2.1 are shown in Fig. 3. In Fig. 3, the actual absorbance of the dye in solution is divided
Figure 1 Transmission electron microscope image of three types of templated luminescent nanospheres discussed in this chapter. (A) Gold templated; (B) liposome templated; (C) micelle templated.
488
Ostafin et al.
Figure 2 Transmission electron microscope image of a series of nanospheres with increasing shell thickness prepared following (A) one round of TMOS treatment as described, (B) two rounds of TMOS treatment, (C) four rounds of TMOS treatment, (D) eight rounds of TMOS treatment. The 65-nm gold nanoparticle template had not yet been removed.
by 80 to match the absorbance of the dye in the suspension. The absorption spectrum of CB in solution is similar to that in the nanosphere suspension except at shorter wavelengths where scattered light is observed. A high affinity of CB to the surface of colloidal gold results in unexpectedly high concentration of encapsulated dye in nanosized porous hollow silicate shells. Despite this high concentration, at neutral to basic pH little fluorescence quenching is observed (Fig. 4A). In solution, the fluorescence intensity at pH 1.8 is sharply decreased to about half of its original value at pH 4.0, reflecting the ionization of SO3ⳮ groups whose charge status has significant effect on fluorescence yields. From these results, the estimated pKa of SO3ⳮ is about 2.5 and is close to the reported range for sulfate at 1.99 [53]. Another inflection in the pH titration of the fluorescence intensity is seen above pH 8.0 and is likely due to the titratable amine. It is only after removing the colloidal gold that the complex pH dependence of the fluorescence yields of CB in the shells becomes evident (Fig. 4A). Inside the shells, the fluorescence intensity depends on the degree of ionization of CB and interactions between CB and the amino
Nanosized Biosensors and Delivery Vehicles
489
Figure 3 The absorption spectra of Cascade Blue威 in solution and inside nanospheres prepared starting with 15-nm colloidal gold.
group of APS and the siloxane interior of the nanosphere. The siloxane group is believed to be complexed with the silicate shell interior, with the amine extending to the interior of the shell. Fluorescence from an excited state dimer (excimer), often observed in a high concentration of pyrene-based dyes in nonpolar solvent [54] is not observed in CB in shells, despite the high internal concentration [45]. At a pH below 2, all the ionizable groups are protonated. The high concentration of dye creates a viscous environment, but the protonated dye is unlikely to be bound to the shell. As the pH increases beyond the pKa of SO3ⳮ, the ionized sulfate group reduces the fluorescence intensity. At pH 4, perhaps the highly packed ionized dye molecules experience dipole–dipole and ionic interactions with the charged amino group of APS leading to nearly complete fluorescence quenching. As the pH is increased further, the negative charge density of the silicate inner surface should also increase. The ionization will reduce the dipole–dipole interaction, as the negative silicate charge competitively interacts with the charged amino group of APS, thus breaking the ionic bond between CB and APS, and leading to an increased fluorescence at its fluorescence maximum in the nanosphere (shifted from 420 nm in solution to 430 nm). The outcome of these processes leads to a silicate shell in which highly concentrated dye demonstrates little fluorescence quenching at neutral solution pH. The pH-dependent fluorescence yields of another dye, FITC, inside silicate nanospheres is shown in Fig. 4B. For FITC, the intensity of fluorescence emission is greater at basic pH, as seen in solution, but is less intense in the neutral pH range. Hydrogen bonding interactions between the FITC and silicate have shifted the apparent pKa of the molecule to more acidic. The primary objective of forming dye-entrapped nanospheres is to synthesize a fluorescent nanoprobe with extremely high intensity and with relatively linear concentration dependence. In order for this to be a viable alternative to existing products, linearity in the observed fluores-
490
Ostafin et al.
Figure 4 The pH-dependent relative fluorescence intensity of Cascade Blue and fluoresce in isothiocyanate in 15-nm silicate nanospheres.
cence signal over a wide range of concentration needs to be demonstrated. The concentrationdependent relative fluorescence intensities of CB in solution and in nanospheres at pH 7.9 are compared in Fig. 5. Since the concentration of the dye in each sphere is unchanged, the number of nanospheres determines the total concentration of dye in the suspension. In the low total concentration range, the fluorescence intensities are linearly dependent on the dye concentration in both solution and in nanospheres over a wide range of dye concentrations. Note that the concentration scale for the dyes in the spheres should be read as the concentration of nanospheres in the suspension, rather than the overall concentration of the dye. By doing so, the fluorescence intensity arising from each nanosphere that contains about ⬃200 dye molecules can be determined, providing a fluorescence amplification factor of ⬃200 relative to the fluorescence from a single molecule. The amplification factor can be made even larger by increasing the number
Nanosized Biosensors and Delivery Vehicles
491
Figure 5 Log-log plot of the concentration-dependent relative fluorescence intensities of CB in solution and in nanoshells at pH 7 plotted against the log of the molar concentrations of dye and nanospheres in suspension.
of entrapped dye molecules in the sphere. Since the fluorescence intensity is pH dependent, the dye-entrapped nanospheres can also be used as a highly sensitive pH indicator. 5. Stabilization of Silicate Nanospheres with 3-Trihydroxysilyl Propionic Acid Silicate nanospheres often aggregate at low pH and at all the pHs in high salt concentration. The aggregation may be significantly decreased by modifying the silicate surface with 3-trihydroxysilyl propionic acid (Fig. 6). In a 250-mL round bottom flask equipped with a stir bar and water-cooled condensor, mix 5 g of 3-(triethoxysilyl) propio-nitrile with 4.2 g of KOH (approximately 1:4 molar ratio K dissolved in 25 mL of E-pure water, and heat to boiling while stirring. When the liquid in the flask begins to boil, the heating should be adjusted to maintain a steady reflux of water in the condensor tube. Ammonia gas is produced as the reaction progresses, and after 24 h of refluxing no further gas evolution should be observable. The resulting liquid is extracted against diethyl ether saturated with water and the ether fraction discarded. The aqueous fraction containing the carboxylic acid derivative should be saved, and excess ether and water removed in a rotary evaporator at 60⬚C. The resulting clear syrupy liquid is diluted to 10 mL, and based on FTIR studies should contain approximately 2.4 M 3-trihydroxylsilypropionic acid in the form of the potassium salt. Silicate nanospheres are incubated with 3trihydroxylsily-propionic acid, and the treated nanospheres are significantly stabilized against aggregation even under physiological salt concentrations for a wide pH range, except above pH 10. Although not discussed here, antibody crosslinking to the now carboxylic acid–decorated surface is possible using appropriate crosslinking agents.
492
Ostafin et al.
Figure 6 Attachment of 3-trihydroxysilane propionic acid to nanosphere surfaces.
6. Attachment of Antibodies to the Silicate Nanospheres The value of a fluorescent nanosphere is greatly enhanced if it can bind exclusively to a specific antigen or antibody [55–60]. Nanospheres with a reactive surface that readily binds antibody or antigen can specifically label the target with intense fluorescence. The ALP-conjugated donkey antisheep IgG is attached to either 65-nm dye filled nanospheres as follows (Fig. 7): One milliliter of 40 times concentrated nanospheres (approximately 1.4 ⳯ 1012 particles) is added to 40 mL of 0.01M borate-sodium at pH 9. Add 100 L of 1 ⳯ 10ⳮ4 M APS in water while stirring and incubate for 1 h at room temperature. The spheres are washed twice with 0.01M borate-sodium at pH 9 by centrifugation at 4200 RPM for 40 min and resuspended with mild sonication into 20 mL of the same buffer. Two hundred fifty microliters of the particle suspension is added to an antibody solution of known concentration in Hank’s phosphate buffered saline at pH 7.4 to make a total volume of 1 mL. The nanospheres and antibody are incubated in the dark for 1 h at room temperature, then resuspended by vortexing if settling has occurred. The antibodytreated nanospheres may be transferred to a new 1.5-mL centrifuge tube that has been treated overnight with 0.1% BSA (bovine serum albumin), and then centrifuged at 8000 RPM for 10 min in a microcentrifuge, and the pellet resuspended in 0.01M borate buffer containing 0.1% BSA. The wash procedure is repeated twice, and if desired the washes from the highest concentrations may be saved for later analysis. Finally the particles may be resuspended into 1 mL of 0.01M borate buffer using mild sonication if needed. For the ALP assay, the above ALP-coated nanospheres are mixed with 50 L of freshly prepared 10 mM PNPP in PBS. As a calibration standard, use 10 L of 1:1000 antibody in PBS, 1 mL of borate, and 50 L of 10 mM PNPP. The absorbance at 405 nm resulting from the ALP enzyme activity can be monitored after 30-min to 2-h incubations. Typical results obtained are shown in Fig. 8. A twofold improvement in attachment efficiency of antibody to the silicate nanospheres is observed over nanospheres not treated with APS. If APS is replaced with ANB-NOS, a threefold improvement in attachment is observed. To use ANB-NOS add 2 mL of 1 mM ANB-NOS in DMF (di-methyl formamide) to the nanospheres and add antibody as described above. Incubate in the dark for 1 h at room temperature, resuspending periodically if needed. Before proceeding with the remaining steps, expose the ANB-NOS-treated suspension
Nanosized Biosensors and Delivery Vehicles
493
Figure 7 (A) Attachment of APS to nanosphere surface. (B) Attachment of ANB-NOS to nanosphere surface. (C) Binding of ALP-conjugated donkey antigoat to silicate nanosphere templated with 65-nm gold nanparticles and treated with ALP crosslinker.
494
Ostafin et al.
to 300–360 nm UV light for 3 min at an intensity of approximately 200 W/m2, or longer for lower intensities. D. Other Types of Templated Silicate Nanospheres While colloidal gold is an excellent template for the formation of silicate nanospheres, it is also possible to use surfactant mesophases in a similar manner. 1. Synthesis of Silicate Nanospheres Using Phosphatidylcholine Liposomes The use of liposomes to template silicate nanospheres was first described by Hubert and coworkers [61–62]. Others have applied the method to the formation of hollow polymeric capsules as well [38–39]. The procedure that is described below is a modification of their procedure. Small size and excellent dye binding are the primary advantages of using colloidal gold as a core for nanospheres. However, it is relatively unstable and requires NaCN for its removal. Phosphatidylcholine liposomes may accommodate many types of molecules by either embedding them within the bilayer or trapping them within the aqueous core of the liposomes. Numerous studies on the fluorescence behavior of dyes trapped within liposomes have been reported [63–65]. In our laboratory, phosphatidylcholine is prepared as a 50 mM solution in chloroform. A 250 L portion is dried overnight at room temperature, and 10 mL of E-pure water added. Next, using a clean glass beaker, vortex the suspension for 10 min and sonicate at low power for 20 min using a Fisher FS30 bath sonicator to produce a solution of unilamellar liposomes. The average size of the liposome obtained by light scattering will be 100 Ⳳ 20 nm.
Figure 8 Relative binding of ALP-labeled donkey antigoat antibody attached to silica nanospheres for given amount added for (A) untreated, APS-treated, (B) ANB-NOS-treated and (C) untreated silicate nanospheres.
Nanosized Biosensors and Delivery Vehicles
495
2. Synthesis of Luminescent Silicate Nanospheres Using Phosphatidylcholine Liposomes Fluorescent dye–filled liposomes may be prepared by the same method described above except the initial reconstitution of the dried lipids is done using a 0.3 mM aqueous solution of the dye in place of water. The unencapsulated dye may be removed using size exclusion chromatography. To prepare the column, 20 g of Sephadex G-25 is allowed to swell overnight in 200 mL of Epure water, and 4 mL of the Sephadex solution poured into a glass wool–stoppered 5-mL syringe and allowed to settle. Excess water is drained from the column, and 1 mL of the dye-containing liposome suspension is gradually added to the top of the column, followed by a small amount of water. The dye-containing vesicles are eluted first, while the free dye is retained in the column. This process may be repeated as many times as necessary. The pH of purified liposomes (10 mL) is adjusted to 8.5 with 1M ammonium hydroxide before adding silicate. Three hundred microliters of 8% TMOS in water are added drop-wise to the suspension with continuous stirring at room temperature, and the suspension is aged for 24–48 h. Unreacted silicate is removed by dialysis against water (pH 8). 3. Fluorescence Studies Relative fluorescence of Sulforhodamine 101, 6-ROX, and Sulforhodamine G in liposomes and those in solutions are compared, and the results are summarized in Fig. 9. Quite contrary to the previously described encapsulation of dyes directly in silicate shells (Fig. 4), the pH dependences of all three dyes are similar between those in solution and in liposomes. This trend suggests that ionic interactions between the dyes and liposomes are negligible, except that the fluorescence intensities are decreased in the liposomes in different degrees. Since the overall concentration of dyes are similar between those in solution and in liposomes, the loss of fluorescence in the liposomes may arise from the pH independent interaction between the dyes and liposomes. 4. Synthesis of Silicate Nanospheres Using Organosilane-Stabilized Microemulsion Highly hydrophobic chromophores may be encapsulated in an organosilane-stabilized oil-inwater (o/w) microemulsion prior to silicate coating (Fig. 1C) For this, 4.8 mg of pyrene is dissolved in 1.2 mL of toluene and mixed with 20 mL of deionized water in a clean glass beaker with vigorous stirring or ultrasonication for 30 min. Add 26 L of octyltrimethoxysilane while stirring, and continue to stir or ultrasonicate for an additional 10 min. Adjust the pH to 8–9 using 0.1 M ammonia, add 800 L of 1.34 mM TMOS in methanol under mild agitation form the silica nanospheres and continue stirring for 24 hr or longer. Remove large aggregates with a 0.45-m filter and dialyze against deionized water at pH 8 for 24 h. Remove free dye by extracting with 10 mL of toluene, and repeat until no more pyrene fluorescence is detected in the organic phase. The fluorescence spectrum of the nanosphere-encapsulated pyrene is noted to be different in water and hydrocarbon solvent [67]. Figure 10 indicates that the molecules in the nanosphere suspension remain in a hydrocarbon environment, as would be expected if it were trapped inside the silica-coated microemulsion oil phase. E. Other Types of Templated Nanospheres 1. Synthesis of Mesoporous Nanospheres The silicate nanospheres already discussed at great length are semiporous, allowing small ionic species to pass through but excluding larger molecules. This property is advantageous for highly luminescent nanospheres used as sensors because it allows limited interactions with the surroundings. At the same time, the species trapped inside is protected from reactions that reduce the
496
Ostafin et al.
Figure 9 The pH-dependent relative fluorescence intensity of rhodamine-based dyes in solution and inside liposome-templated silica nanospheres.
probe’s sensitivity and usefulness. On the other hand, the speed at which material is transferred in and out of the particles may be too slow for a nanosized drug delivery vehicle, or microdevice targeted to a specific site in the body. If this is the case, then a more porous silicate nanosphere may be formed by using CTAB, ammonia, and TEOS (Fig. 11). In the standard procedure, 512 L of 0.5 mM MPTS in ethanol is added with rapid stirring to 317 mL of 65-nm-diameter colloidal gold prepared as described above. After stirring for 15 min, 317 mL of MPTS-treated gold is diluted to 421 mL using deionized water, and warmed to 40⬚C. To this solution, 43.3 g of 26 wt% NH3 in water is added, and the temperature raised to 50⬚C before 1.12 g of CTAB is added. After 5 min, 5.21 g of TEOS is added slowly with rapid stirring. The final molar ratio is H2O/NH3/SiO2/CTAB (1005:30:2:1:0.123). The sol is aged for 2 h with stirring at room
Nanosized Biosensors and Delivery Vehicles
497
Figure 10 Fluorescence spectrum of pyrene inside a toluene-filled silica nanosphere made using octadecyltrimethoxysilane/toluene microemulsion as the nanosphere template. Pyrene fluorescence spectrum is sensitive to water, and this result clearly indicates that it remains in a hydrophobic environment inside the nanospheres.
temperature. The solution pH during this period increases from 10.1 to the final value of 10.7. Nanospheres are recovered by filtration over a sintered glass frit (pore size 0.2 m). To remove the CTAB template, the material is calcined at 540⬚C for 8 h in air or is extracted using alcohol. Because of the additional step, if calcination is planned, then it may be necessary to load the particle with the organic dye after the pores have been cleared. The particle shown in Fig. 11 has no dye inside. The synthesis of mesoporous silicate nanospheres around colloidal gold is discussed elsewhere [68] and proceeds via three main stages: (1) the hydrolysis of monomeric silicon esters and subsequent condensation reactions to form silica oligomers; (2) formation of silica/CTAB primary particles; and (3) mesopore growth via either aggregation of primary particles or deposition of monomeric silica and CTAB molecules. A scheme of the self-assembly mechanism is presented in Fig. 12. The first stage in assembly follows basic silica chemistry where the ratelimiting step is the hydrolysis of silicon esters. Monomeric silica species, in the form of Si(OR)4-nOnⳮn, where R is the organic ester, form silica oligomers via rapid condensation reactions. In the second stage, silica oligomers and CTAB would likely interact via very strong multidentate linkages between the ammonium head group and siloxide ions to form primary particles made of several silicate and CTAB molecules. A silica/CTAB primary particle behaves like a ‘‘super surfactant,’’ enabling self-assembly at concentrations well below the CMC1 of pure CTAB. The third stage proceeds via aggregation, where primary particles agglomerate together in an ordered fashion because CTAB/silica primary particles have very low dissociation constants and the silica structures screen the electrostatic repulsion between adjacent primary particles, increasing the negative free energy for self-assembly to proceed.
498
Ostafin et al.
Figure 11 Transmission electron microscope image of 15-nm gold nanoparticles coated with a mesoporous silicate shell.
2. Synthesis of Biocompatible Calcium Phosphate Nanospheres While silicate is nontoxic, chemically inert, and suitable for use in most in vitro applications, a truly biocompatible nanosphere, such as one made from calcium phosphate, could be more useful in the human body. Calcium phosphate ceramics are well known for their uses in biotechnology [69–71]: as a biocompatible barrier between titanium and aluminum bone implants [72–73]; in orthopedics, maxillofacial surgery, and in periodontics [74]; and in drug delivery [75–76]. Recently, we have demonstrated that the liposome-templated method for constructing silicate nanospheres can be adapted to the synthesis of calcium phosphate nanospheres also loaded with fluorescent dye [77]. In contrast to silicate condensation, the reaction of calcium and phosphate to form insoluble crystals is very fast. Thus in our procedure, a small amount of the CaNO3 is added first, in proportion to the surface area of the liposome, to provide a monolayer that covers the periphery of the liposome, and then calcium and phosphate are added in small alternating aliquots over time. Throughout the process, two reactions will be competing—the deposition of calcium phosphate onto the liposomal nucleation site and the reaction of calcium and phosphate ions in solution—to form solid calcium phosphate particles and crystals. To prepare liposomes for subsequent coating with calcium phosphate, 25 mg of lyophilized DOPA lipid is added to 5 mL of deionized and degassed water to make a 6.9 mM solution. The suspension is extruded through a 100-nm pore size filter, and light scattering is used to confirm that the size of the liposomes is between 100 Ⳮ/ⳮ 20 nm in diameter. Typical nanospheres obtained via this method are shown in Fig. 13. For calcium phosphate shells on DOPA liposomes, 50 mL of water and 2.5 mL of liposome stock solution are mixed in a four-necked, roundbottom 250-mL flask, and then the solutions adjusted to pH 11 using ammonium hydroxide. Two 10-mL burets attached to the flask are each filled with 10 mL of water, and 150 L 0.29 M calcium nitrate is added to one and 105 L of 0.24 M diammonium phosphate is added to the other. A third port is used for pH monitoring using an Orion (420A) ATC probe. The fourth opening is used to flow nitrogen gas through the reaction chamber. To form nanospheres, 400L portions of calcium and phosphate solutions from the burets are added in alternating fashion at 10 min intervals with rapid stirring. After the third portion is added, the intervals between additions is decreased to 2 min. Following the addition of all the material, the suspension is stirred for an additional 15 min and aged at room temperature for 3 days.
Nanosized Biosensors and Delivery Vehicles
499
Figure 12 Schematic of mesoporous silicate nanosphere chemistry in aqueous solution.
III. CONCLUSIONS Highly fluorescent silicate nanospheres may find wide areas of application in the fields of science and engineering and, when the spheres are made immuno-specific, in clinical imaging and diagnosis. Silicate and calcium phosphate-coated liposomal and micellar templates allow the encapsulation and protection of biologically active molecules and their delivery. Mesoporous silicate shells will find a niche where the exchange of large molecules and ions between encapsulated compounds and the environment is vital.
500
Ostafin et al.
Figure 13 Transmission electron micrograph of calcium phosphate nanospheres produced using 1,2dioleoyl-SN-glycero-3-phosphate (DOPA) liposome template.
ACKNOWLEDGMENTS This work could not have been accomplished without the tireless and enthusiastic work of other postdoctorates, graduate students, and undergraduates in the group: Dr. Q. Dai, Dr. I. Banerjee, Mr. F. Liu, Mr. Q. Wang, Mr. H. Schmidt, Ms. N. Kohrt, Ms. Cintyu Wong, Mr. J. MacDonald, Mr. P. Henning, and Mr. M. Siegel.
REFERENCES
1. Watkins S, Geng XH, Li LH, Papworth G, Robbins PD, Drain P. Imaging secretory vesicles by fluorescent protein insertion in propeptide rather than mature secreted peptide. Traffic 2002; 3: 461–471. 2. Conway RR, Minor LK, Xu JZ, D’Andrea MR, Ghosh RN, Demarest KT. Quantitative analysis of agonist dependent parathyroid hormone receptor trafficking in whole cells using a functional green fluorescent protein conjugate. J. Cell. Physiol 2001; 3:341–355. 3. Kozlowski DJ, Murakami T, Ho RK, Weinberg ES. Regional cell movement and tissue patterning in the zebrafish embryo revealed by fate mapping with caged fluorescein. Biochem Cell Biol 1997; 75:551–562. 4. Hougland R. Handbook of Fluorescent Probes and Research Products. Molecular Probes. 6th Ed, 1996. 5. Emoto K, Van Alstine JM, Harris JM. Stability of poly(ethylene glycol) graft coatings. Langmuir 1998; 14:2722–2729.
Nanosized Biosensors and Delivery Vehicles
501
6. Men Y, Audran R, Merkle HP, Gander B. Immunological responses with biodegradable polymer microspheres (review). Drug Targeting and Delivery. Vol. 8, 1997:191–205. 7. Benedetti L, Bertucco A, Pallado P. Production of microparticles of a biocompatible polymer using supercritical carbon dioxide. Biotechnol Bioeng 1997; 53:232–237. 8. Cohen H, Levy RJ, Gao J, Fishbein I, Kousaev V, Sosnowski S, Slomkowski S, Golomb D. Sustained delivery and expression of DNA encapsulated in polymeric nanoparticles. Gene Therapy 2000; 7: 1896–1905. 9. Sumner JP, Aylott JW, Monson E, Kopelman R. A fluorescent PEBBLE nanosensor for intracellular free zinc. Analyst 2002; 127:11–16. 10. Xu H, Aylott JW, Kopelman R, Miller TJ, Philbert MA. 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. Analytical Chemistry 2001; 73:4124–4133. 11. Chien GL, Anselone CG, Davis RF, Van Winkle DM. Fluorescent vs. radioactive microsphere measurement of regional myocardial blood flow. Cardiovasc Res 1995; 30:405–412. 12. Zhang YZ, Kemper C, Bakke A, Haugland RP. Novel flow cytometry compensation standards: internally stained fluorescent microspheres with matched emission spectra and long-term stability. Cytometry; 33:244–248. 13. Yang WJ, Trau D, Renneberg R, Yu NT, Caruso F. Layer-by-layer construction of novel biofunctional fluorescent microparticles for immunoassay applications. J. Colloid Interface Sci 2001; 234:356–362. 14. Jalon EG, Blanco-Prieto MJ, Ygartua P, Santoyo S. PLGA microparticles: possible vehicles for topical drug delivery. Int. J. Pharm 2001; 226:181–184. 15. Brasuel M, Kopelman R, Miller TJ, Tjalkens R, Philbert MA. 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 2001; 73: 2221–2228. 16. Clark HA, Kopelman R, Tjalkens R, Philbert MA. Optical nanosensors for chemical analysis inside single living cells. 2. Sensors for pH and calcium and the intracellular application of PEBBLE sensors. Anal Chem 1999; 71:4837–4843. 17. Roy B, Arruda A, Mallik S, Campiglia A. Polymerizable fluorescent liposomes incorporating lanthanide ions. Am. Chem. Soc 2001; 222:561–565. 18. Roy BC, Fazal MA, Arruda A, Mallik S, Campiglia AD. Polymerized fluorescent liposomes incorporating lanthanide ions. Organic Lett 2000; 2:3067–3070. 19. Araujo A, Khoobehi B, Pouyadou S, Peyman GA, Pironis T, Shoelson B. Quantification of blood flow in retina using fluorescent liposomes, microspheres, or fluorescent cells. Inv. Ophthal. Vis. Sci 1999; 40:2577–2582. 20. Nguyen T, Rosenzweig Z. Calcium ion fluorescence detection using liposomes containing Alexalabeled calmodulin. Anal Bioanal Chem 2002; 374:69–74. 21. Singh AK, Harrison SH, Schoeniger JS. Gangliosides as receptors for biological toxins: development of sensitive fluoroimmunoassays using ganglioside-bearing liposomes. Anal Chem 2000; 72: 6019–6024. 22. Liu WT, Sun JR, Lin CH. Transgenic expression of the jellyfish green fluorescent protein in the cone photoreceptor of the mouse. Vis. Neurosci 2001; 18:615–623. 23. Grasela JJ, McIntosh AH, Goodman CL. An indicator cell assay for detection of human cytomegalovirus based on enhanced green fluorescent protein. J. Virol. Methods 2001; 96:85–92. 24. Grasela JJ, McIntosh AH, Goodman CL, Wilson LE, King LA. Expression of the green fluorescent protein carried by Autographa californica multiple nucleopolyhydrovirus in insect cell lines. In Vitro Cell Devel Biol Animal 2000; 36:205–210. 25. Fu J, Wang W, Bai X, Lu D, Ruan C, Chen Z. The enhanced green fluorescent protein as a marker for human tumor cells labelled by retroviral transduction. Chinese J Cancer Res 2002; 14:126–130. 26. Kehlenbach RH, Berace L. Analysis of nuclear protein import and export in vitro using fluorescent cargoes. Methods Molec Biol 2002; 189:213–245. 27. Miraglia S, Swartzman EE, Mellentin-Michelotti J, Evangelista L, Smith C, Gunawan I, Lohman K, Goldberg EM, Manian B, Yuan PM. Homogeneous cell- and bead-based assays for high throughput screening using fluorometric microvolume assay technology. J. Biomolec. Screen 1999; 4:193–204.
502
Ostafin et al.
28. Masuo A. Protein engineering for bioanalytical science, Seibutsu Shiryo Bunseki. 1999; 22:295–305. 29. Champiat D, Roux A, Lhomme O, Nosenzo G. Biochemiluminescence and biomedical applications. Cell Biol Toxicol 1994; 10:345–51. 30. Yang J, Thomason DB. An easily synthesized, photolyzable luciferase substrate for in vivo luciferase activity measurement. Biotechniques 1993; 15:848–850. 31. Unge A, Tombolini R, Molbak L, Jansson JK. Simultaneous monitoring of cell number and metabolic activity of specific bacterial populations with a dual gfp-luxAB marker system. Appl Env. Microbiol 1999; 65:813–821. 32. Zhu JM, Li P. Novel synthesis of Protein-PMMA core-shell nanoparticles via self-activating graft polymerization with hydroperoxide. Abs. Papers Am Chem. Soc 2002; 223:240–250. 33. Sershen SR, Westcott SL, Halas NJ, West JL. Independent optically addressable nanoparticle–polymer optomechanical composites. Appl. Phys. Lett 2002; 80:4609–4611. 34. Zhou CL, Zhao Y, Jao TC, Winnik MA, Wu C. Photoinduced aggregation of polymer nanoparticles in a dilute nonaqueous dispersion. J. Phys. Chem. B 2002; 106:1889–1897. 35. Thunemann AF, General S. Poly(ethylene oxide)-b-poly(ethylene imine) dodecanoate complexes: lamellar-within-lamellar morphologies and nanoparticles. Macromolecules 2001; 34:6978–6984. 36. Choi JS, Lee EJ, Park SJ, Kim HJ, Park JS. Novel macromolecular self-organization of poly(ethylene glycol)-block-poly(L-histidine): pH-induced formation of core-shell nanoparticles in aqueous media. Bull. Korean. Chem. Soc 2001; 22:261–262. 37. Goller MI, Vincent B. Silica encapsulation of liquid PDMS droplets. Colloids Surfaces A 1998; 142: 281–285. 38. McKelvey CA, Kaler EW, Zasadzinski JA, Coldren B, Jung H-T. Templating hollow polymeric spheres from cationic equilibrium vesicles: synthesis and characterization. Langmuir 2000; 16: 8285–8290. 39. Marinakos SM, Novak JP, Brousseau LC, Blaine House A, Edeki EM, Feldhaus JC, Feldheim DL. Gold particle as templates for the synthesis of hollow polymer capsules. Control of capsule dimensions and guest encapsulation. J Am Chem Soc 1999; 121:8518–8522. 40. Huo Q, Feng J, Schuth F, Stucky GD. Preparation of hard mesoporous silica spheres. Chem. Mater 1997; 9:14. 41. Ozin G. Nanochemistry Synthesis in Diminishing Dimensions. Adv. Mater 1992; 4:612. 42. Bo¨chel C, Gro¨n M, Unger KK, Matsumoto A, Tsutsumi K. Tailored synthesis of nanostructured silicas. Control of particle morphology, particle size and poresize. Supramolecular Sci 1998; 5:253. 43. Sierra L, Guth J-L. Synthesis of mesoporous silica with tunable pore size from sodium silicate solutions and a polyethylene oxide surfactant. Microporous Mesoporous Mater 1999; 27:243–253. 44. Santra S, Wang K, Tapec R, and Tan W. Development of novel dye-doped silica nanoparticles for biomarker application. J. Biomed. Optics 2001; 6:160–166. 45. Ostafin A, Seigel M, Wang Q, Mizukami H. Enhanced emission of cascade blue within silicate nanoshells. Microporous Mesoporous Mater() 2001. 46. Turkevich J, Stevenson PC, Hillier JA. Study of the nucleation and growth processes in the synthesis of colloidal gold. J Colloid Sci 1951; 1:55–78. 47. Brown KR, Walter DG, Natan, Michael J. Seeding of colloidal Au nanoparticle solutions. 2. Improved control of particle size and shape. Chem. Mater 2000; 12:306–313. 48. Hodak JH, Henglein A, Hartland GV. Photophysics and spectroscopy of metal particles. Pure Appl. Chem 2000; 72:189–197. 49. Sandroff CJ, Herschbach DR. Kinetics of displacement and charge transfer reactions probed by SERS—Evidence for distinct donor and acceptor sites on colloidal gold surfaces. Langmuir 1985; 1:131–135. 50. Liz-Marzan LM, Giersig M, Mulvaney P. Synthesis of nanosized gold-silica core-shell particles. Langmuir 1996; 12:4329–4335. 51. Giersig M, Ung T, Liz-Marzan LM, Mulvaney P. Direct observation of chemical reactions in silicacoated gold and silver nanoparticles. Adv Mater 1997; 9:570–575. 52. Giersig M, Liz-Marzan LM, Ung T, Su D, Mulvaney P. Chemistry of nano-sized silica-coated metal particles—EM study. Berichte der Bunsengesellschaft fur Physikalische Chemie 1997; 11: 1617–1627.
Nanosized Biosensors and Delivery Vehicles
503
53. Weast RC, ed. CRC Handbook of Chemistry and Physics. 66th Ed., CRC Press. 1985:D157–D160. 54. Birks JB, Christophorou LG. Excimer formation in polycyclic hydrocarbons and their derivatives. Nature 1963; 197:1064–1065. 55. Thomas RN, Guo CY. Nanosphere-antibody conjugates with releasable fluorescent probes. Fres. J. Anal. Chem 2001; 369:477–482. 56. Huang SC, Caldwell KD, Lin JN. Site-specific immobilization of monoclonal antibodies using spacermediated antibody attachment. Langmuir 1996; 12:4292–4298. 57. Oates MR, Clarke W, Marsh EM. Kinetic studies on the immobilization of antibodies to highperformance liquid chromatographic supports. Bioconjugate Chem 1998; 9:459–465. 58. Sanford MS, Charles PT, Commisso SM. Photoactivatable crosslinked polyacrylamide for the siteselective immobilization of antigens and antibodies. Chem Mater 1998; 10:1510–1520. 59. Piehler J, Brecht A, Geckeler KE. Surface modification for direct immunoprobes. Biosensors and Bioelectronics 1996; 11:579–590. 60. Collino R, Therasse J, Chaput F. Biological activity of functionalized SiO2 thin films prepared by sol-gel method. J Sol-Gel Sci Technol 1996; 7:81–85. 61. Hubert DHW, Jung M, Frederick PM, Bomans PHH, Meuldijk J, German AL. Vesicle-directed growth of silica. Adv. Mater 2000; 12:1286. 62. Hubert DHW, Jung M, German AL. Vesicle templating. Adv. Mater 2000; 12:1291. 63. Takagi K, Sawaki Y. Photochemical reactions in organized and semi-organized media. Crit Rev Biochem Molec Biol 1993; 28:323–367. 64. Duportial G, Merola F, Lianos P. Fluorescence energy transfer in lipid vesicles. A time-resolved analysis using stretched exponentials. J Photochem Photobiol A 1995; 89:135–140. 65. Hong-Ru L, Fang G, Chen-Ho T, Li-Zhu W. Energy transfer of ionic dyes in mixed surfactant vesicles. 2000; 26:575–585. 67. Kalyanasundam K, Thomas JK. Environmental effects on vibronic band intensities in pyrene monomer fluorescence and their application in studies of micellar systems. J. Am. Chem. Soc 1997:2039–2044. 68. Nooney RI, Thirunavukkarasu D, Chen Y, Josephs R, Ostafin AE. Synthesis of nanoscale mesoporous silica spheres with controlled particle size. Chem Mater:in press. 69. Hench L. Bioceramics from concept to clinic. J. Am. Ceram. Soc 1991; 74:1487–1510. 70. Barnes P. The myth becomes reality. Ceram. Ind 1987; 96:14–20. 71. Jiang G, Shi. D. Coating of hydroxyapatite on highly porous AL2O3 substrate for bone substitutes. J. Biomed. Res 1998; 43:77–81. 72. Doremus RH. Bioceramics. J. Mater. Sci 1992; 27:285–297. 73. Thomas KA. Hydroxyaptite coatings. Orthopedics 1994; 17:267–278. 74. Salahi E, Hashjin MS, Hossainia A, Torkman M. Hydroxyapatite Bioceramics for Ridge Augmentation. Proceedings of 2nd Iranian Ceramics Congress Oct. 1995:265–270. 75. Paul W, Sharma CP. Development of porous spherical hydroxyapatite granules: applications towards protein delivery. J. Mater. Sci: Mater Med 1999; 10:383–388. 76. Yamamura K, Iwata H, Yotsuyanagi T. Synthesis of antibiotic-loaded hydroxyapatite beads and in vitro drug release testing. J. Biomed. Mater. Res 1992; 26:1054–1064. 77. Schmidt HT, Ostafin AE. Calcium phosphate nano-shell Materials. Adv Mater 2002; 14:532–535.
23 Relationships Between Biomaterials and Biosensors Kirk J. Bundy Tulane University New Orleans, Louisiana, U.S.A.
I. INTRODUCTION Considered broadly, the use of sensors is one of the most potent weapons in the arsenal of clinicians in their efforts to diagnose and fight disease. Physical sensors can monitor many important parameters that provide information concerning health status and physiological function. Specific examples of such parameters include temperature, pressure, flow rate, volume, voltage, etc. The sensors considered in this chapter, however, are chemical sensors. These are devices yielding signals that allow detection of a specific analyte (or analyte group) of interest. A biosensor can be defined as a chemical sensor that incorporates biological materials (acting as sensing agents) coupled to appropriate transducers. More specifically, tissues, microorganisms, whole cells, cell receptors, enzymes, antibodies, organelles, nucleic acids, biologically derived materials, and biomimics can all be used for sensing (based on biomolecular recognition affinity interactions and other principles). The idea that the biological element acts as a sensing agent is a key one. This means that it interacts with a specific chemical in its environment in such a manner as to produce a measurable signal that is (in one way or another) proportional to the concentration of the chemical of interest. Optical, electrochemical, thermometric, gravimetric, piezoelectric, and magnetic transduction effects can be used for signal measurement. Thus, for example, a bacterium that produces less light in response to a toxicant could be the basis for a biosensor, while the canary that workers used to take with them into mines to sense carbon monoxide would not be. The canary is an example of what is sometimes known as a bioindicator. Though the field of biosensors has been the focus of widespread attention only comparatively recently, Clark and Lyons described the first use of a biosensor over 40 years ago. They developed an enzyme electrode for the sensing of glucose [1]. Since then, there has been an extensive proliferation in the application of the biosensor concept. This chapter reviews biomedical applications of biosensors.
II. USES OF BIOSENSORS As can be inferred from the description in the prior section, there are a great many uses for biosensors in a number of different application areas. For example, in the field of environmental 505
506
Bundy
science there is considerable interest in employing biosensors to measure the concentration of pesticides and other pollutants in a variety of media (air, water, soil, sediments, etc.). Biosensors can significantly aid in the detection of chemical and biological weapon (CBW) agents, a major priority for military defense and antiterrorism efforts. These uses have received heightened attention due to recent historical events that have focused conern on the need for means to counter CBW threats. The major focus of this chapter, however, is to consider the biomedical uses of biosensors, another major area of interest in this field. A. Agents That Can Be Detected by Biosensors As shown in Table 1, a vast assortment of chemical and biological entities either influence health status or reflect it. As one proceeds down the listing of the items in the table, they exhibit increasing molecular weight and/or structural complexity. The smaller and simpler molecules (HⳭ and Clⳮ ions or dissolved O2, for example) can be detected with chemical sensors that are based on conventional physical means, i.e., potentiometric and amperometric measurement, absorption of light, etc. Instrumentation and techniques such as pH meters, selective ion electrodes, colorimetry, polarographic analysis, and cyclic voltammetry can be employed to detect many simple molecules. Detection and measurement of the more ‘‘complex’’ entities in the table are generally more involved. In some situations, separation, counting, and microscopic techniques may be used for quantification. However, for other conditions, implementation of these methods may not be possible, or else using them may be inconvenient due to expense or time constraints. Here, an alternative may be to detect the entity of interest by physical means, modulated by means of its interaction with a biological agent. In other words, a biosensor may be used. The analytes of biomedical interest for which biosensors are currently commercially available [2] are listed in Table 2. The technical feasibility of detection of many additional substances
Table 1 Chemical and Biological Entities Whose Concentration or Presence Are Indicators of or Influences on Health Status pH Ions Dissolved gases Drugs Steroids Hormones Neurotransmitters Proteins Enzymes Antibodies Cytokines Other ligands for cell receptors Others Viruses Bacteria Parasites Tumors
Relationships Between Biomaterials and Biosensors
507
Table 2 Analytes of Biomedical Significance that Can Be Detected with Commercially Available Biosensors Glucose Lactate Alcohols Lactose Sucrose Galactose Uric acid Choline Alpha amylase L-Lysine Urea L-Lactic acid L-Ascorbic acid
of biomedical interest has also been demonstrated, and the results of this research have been discussed in various works, e.g., Refs. 3–10. Many of these substances are listed in Table 3. A wide variety of chemical classes are of concern in terms of pollution of the environment. Among those of interest include heavy metals, pesticides, phosphates, phenols, nitrates, volatile organic carbons, semivolatile organic carbons, polyaromatic hydrocarbons, and petroleum-related compounds. Many studies have explored the contributions that biosensors can make in terms of measuring analytes of interest to environmental science and technology, e.g., Refs. 11–16. Besides the chemicals of enviromental interest, many additional substances are of military and antiterrorism concern. Some classes of substances here include explosives; immune, nervous, and endocrine system toxins; other toxins; and agents potentially usable in CBW applications (e.g., bacteria, viruses, fungi, recombinant organisms/molecules, nerve agents, and blistering agents). Paddle has provided an extensive list of over 30 toxic materials potentially usable in weapons that are currently able to be sensitively detected with biosensors [17]. Biosensors also play a significant role in the food industry [18]. B. Types of Biosensors and Their Deployment Modalities Most biosensors of biomedical medical interest are based on biomolecular recognition principles (BRPs). This means that there is a very specific structural similarity between the sensing agent and its particular target ligand. In other words, there is an active site on the sensing molecule into which a portion of the structure of the target will fit such that a strong bonding interaction between the two will take place. ‘‘Hand and glove’’ or ‘‘lock and key’’ analogies are sometimes used to describe the interaction between the sensing agent and target analyte for BRP-based biosensors. Specific types of this kind of biosensor include those based on enzyme/substrate, antibody/antigen, or cell receptor/ligand interactions. Generically this sensor type is termed either a ‘‘catalytic’’ or an ‘‘affinity biosensor.’’ Other biosensors may respond to a broad spectrum of chemicals. For example, a probe utilizing bioluminescent bacteria as the sensing agent may emit a diminished light intensity
508
Bundy
Table 3 Analytes of Biomedical Significance Where Technical Feasibility of Detection with Biosensors Has Been Shown Glycolate Creatine kinase D-Gluconate Dopamine Substrates of NAD-dependent dehydrogenases Glycerophosphate Bilirubin Creatinine Glutamic acid Ammonia Oxalate Acetylcholine Phosphatidylcholine Salicylate Acetic acid Choline Cholesterol and its esters Urate Glutamate pyruvate transaminase Glutamate oxaloacetate transaminase Orosomucoid (EIA) Digoxin (EIA) IgG (EIA) Lidocaine (EIA) 2,4-Dimethyl-aminocaproic acid (EIA) Phenytoin (EIA) 2,4-DNP/anti-DNP HCG/anti-HCG Riboflavin/aporiboflavin Protein MRP 14/anti-MRP 14 17-Estradiol/anti-17-estradiol HBsAg/anti-HBsAg 12 amino acids besides lysine (which is listed in Table 2) Various metallic ions Glutamate GABA Fluoride and sulfite Adenosine 5⬘-AMP
Chloroform Flavine adenine dinucleotide Maltose Oxaloacetate Phenol Linoleic acid Various enzymes Various hydroxy bile acids Chenodiol Penicillin ATP/ADP Formate Thrombin Cephalosporins L-Malate Cellobiose Phospholipids Triglycerides Xanthine Hypoxanthine Nicotinic acid Vitamin B1 Glycerol Certain hormones Thiol compounds Formic acid Oxalic acid Various mutagens Anti–bovine serum albumin Candida albicans Salmonella typhimurium and its DNA Other microbes Human albumin Human tranferrin IgA, IgE, and IgM Polynucleotide hybridizations Cocaine Formaldehyde Triolein
EIA, via an enzyme immunoassay technique; HCG, human chorionic gonadotropin; HBsAg, hepatitis B surface antigen.
when exposed to any number of hazardous substances. Such a biosensor generally cannot be used for concentration monitoring, but it does form the basis for various quantitative toxicity assessment methods. Three basic deployment modes are pertinent to biosensors used for biomedical purposes. The first is in a laboratory setting, where a tissue sample or body fluid removed from the patient is analyzed. The entity of interest in some circumstances may be assayed with an appropriate
Relationships Between Biomaterials and Biosensors
509
biosensor on-site, or alternatively it may be sent to a remote clinical laboratory. In either case, before the test can be conducted, the sample may need to be subjected to various preparation steps (separation, extraction, exposure to certain reagents, etc.). All things being equal, on-site measurement is preferred for reasons of convenience and expense considerations if it is technically possible and practical to implement. A third possibility is the deployment of biosensors in vivo. In principle, this can be accomplished in two ways. First, the biosensor can invasively be placed in contact with the fluid containing the analyte. In this case, the probe is removed shortly after the measurements have been completed. The most demanding in vivo deployment modality is that of the in situ biosensor. Here, the device is implanted within the body for an extensive period, and many successive measurements are performed over time. The signals produced may either be monitored telemetrically (if measurement per se is the purpose of the biosensor) or they could be used as feedback signals in a larger system being used therapeutically to deliver drugs or other pharmaceutical agents.
C. Advantages and Limitations of Biosensors An important consideration in chemical analysis is the ability of the analytical method to quantify the concentration of the molecule of interest even in the presence of much larger concentrations of extraneous substances that potentially create interferences (i.e., in effect, chemical noise as opposed to chemical signal). Fortunately for the biosensor field, many natural physiological mechanisms have evolved to deal with such needle in the haystack problems. For example, the insulin receptor can bind to its target even in the presence of a 100,000-fold excess of other proteins [19]. Since affinity biosensors also employ such natural mechanisms, they offer the possibility of excellent specificity in sensing the molecule of interest, while providing resistance to interferences. They are thus of particular interest for situations requiring trace level detection of complex biomolecules. There are several potential disadvantages associated with biosensors that should also be mentioned. For example, maintaining the molecular functionality of the sensing agent while incorporating it into a practical sensing device may be problematic. Denaturation of proteins can occur when they are either immobilized by chemically binding to a substrate (such as a polymer membrane) or incorporated into a carrier matrix. Such denaturation can structurally distort the active site of the sensing agent, rendering it unable to interact with its target biomolecule. Similarly, if the site where the sensing agent is attached to its carrier is too close to the active site of the molecule, the target ligand may be sterically restricted from interacting. In either case, the bioactivity of the sensing agent is effectively restricted. Another difficulty is that, once attached to the biosensor, the sensing agent may need to be kept moist or refrigerated before use, which can pose an inconvenience in some situations. Saturation of the sensing molecules by high concentrations of the analyte of interest (or interfering molecules) can also pose problems. A final difficulty that should be mentioned is that in situations where there is irreversible binding of the analyte by the sensing agent, the biosensor is nonrenewable. It cannot be refreshed and is suitable for one-time use only. Besides the problems discussed above related to making measurements with biosensors, difficulties associated with obtaining sensing agents in suitable forms can also present obstacles for sensor development. Some agents may be readily commercially available. Many enzymes fall into this category. Other agents, the nicotinic acetylcholine receptor (nAChR), for example, are more difficult to obtain in a usable form. According to one approach for obtaining the nAChR, the source of the receptors is from the electric ray Torpedo californica. The receptor
510
Bundy
proteins are extracted from the carrier tissue, separated from extraneous proteins using affinity chromatography, and reconstituted according to rather intricate procedures [20–22]. Beyond the potential limitations and disadvantages mentioned above, in vivo use of biosensors can pose additional problems. Any time a biomaterial comes in contact with a body fluid, there is a tendency for proteins to adsorb onto the surface. For a biosensor, this represents biofouling and may prevent contact between the sensing agent and its target ligand. Another difficulty is that if the sensing agent leaches out from the biosensor, this will reduce its effectiveness. In addition, biocompatibility is a potential concern that must govern the selection of all portions of the biosensor-carrier biomaterial as well as the sensing agent. Possible biocompatibility problems could thus result from leaching out of the bioactive sensing molecules. These concerns are considered in more detail in the next section. D. In Vivo Operation of Biosensors Performance goals for biosensors used in different fields have been described previously, e.g., by Rogers and Lin [23]. Important concerns for biosensors to be used for biomedical purposes are given in Table 4. Measurement under in vivo conditions presents the most challenges. One of the most important considerations is that the surface of the sensing element of the biosensor device should resist fouling. When a material is brought into contact with an in vivo fluid, its surface will generally rather quickly start to become covered with adsorbed proteins. Here the biomaterials field may be able to offer some assistance in solving this biosensor problem. Polyethylene glycol (PEG), also known as PEO (polyethylene oxide), surface coatings have been used as an effective means for preventing protein adsorption in various biomaterials applications [24–26]. This approach has been suggested as applicable to the biosensor fouling problem [27] and could be a strategy for enhancing the service lifetime of in vivo biosensors. For biosensor use, though, it should be kept in mind that PEG antifouling coatings would also repel any proteins from contacting the sensing agent. Thus biosensors with PEO-covered surfaces would not be suitable for detecting such high molecular weight analytes. This PEG coating approach could be very useful when the analyte of interest is a low molecular weight substance, however. This form of sensor might be effective for detection of ions of various kinds for instance. In any biomaterials application where the materials are placed in contact with body tissues and fluids for any significant period of time (including biosensors used biomedically), the materi-
Table 4 Engineering Performance Requirements for Biosensors Used Biomedically Reasonable length of time for performing an assay Reasonable cost per analysis Threshold detection limit sensitivity (should be in the ppm to ppb region) Adequate dynamic range High specificity Resistance to interferences Ease of operation Minimal preparation of test samples (ex vivo operation) Antifouling surface (to allow in vivo operation for significant time periods) Resistant to release of sensing agents into the body Fabrication from biocompatible materials
Relationships Between Biomaterials and Biosensors
511
als must be biocompatible. In other words, released degradation products from the biomaterial due to interaction with the surrounding environment should not elicit untoward local effects (e.g., inflammation) or systemic effects (e.g., allergic reactions). This restriction limits the materials used to construct biosensors that are placed in vivo. The matrix carrying the sensing agents would be an example. Additionally for biosensors, however, consideration must be given to possible interaction of the sensing agents with in vivo fluids since these agents have active biofunctionality. The agents should be sufficiently stable in their attachment to the carrier so that they do not leach out into the environment in appreciable quantities. This will avoid two problems mentioned above. First, maintaining the amount of sensing agent present retains the effectiveness and sensitivity of the sensor. Second, and potentially more important, minimizing release of such agents into the body will help to avoid biocompatibility problems that may be caused by the bioactive nature of the molecules themselves. E. Operating Principles of Biosensors There are various types of measurement principles that can be used in biosensors to generate the signal associated with detection of analytes of interest. The main possibilities are given in Table 5. Various classes of sensing agents can produce detectable signals based on each of these principles. Examples of such measurement principle/biological element biosensor combinations are provided in Table 6. The interactions between sensing agents and target analytes can produce a rather diverse set of responses—production of heat, light photons, and electrical currents; shifts in electrical potential; and mass changes (that influence resonance frequencies and stress wave propagation velocities of sensing elements excited by vibratory waves). In some cases the reader may not be familiar with the operating principles listed in the table, so the basic ideas behind some of them are briefly sketched out below. The amperometric and potentiometric methods utilize electrochemical oxidation/reduction reactions associated with the interaction of the sensing agent and its target. They are based on measurement of current flow and voltage shifts, respectively. These electrochemical measurement techniques are discussed in more detail in the next section, where specific biomedical applications are considered.
Table 5 Physical Effects that Can Be the Basis for Measurements with Biosensors Electrical Amperometric Potentiometric Capacitance Optical Luminescence Fluorescence Evanescent wave Surface plasmon resonance Mechanical Surface acoustic wave Bulk acoustic wave Thermometric
512
Bundy
Table 6 Examples of Viable Biosensor Measurement Effect/Sensing Agent Combinations Measurement principle Amperometric Potentiometric Optical Luminescence Fluorescence Surface plasmon resonance Evanescent wave
Types of sensing agents used Enzymes, microorganisms, enzyme-labeled antibodies, plant and animal tissues Same as for amperometric methods; also cell receptors Microorganisms, enzymes Receptors, antibodies Antibodies, antigens, nucleic acids, enzymes Receptors, antibodies
The description of the first two optical techniques is fairly straightforward. For luminescence, visible light production is influenced by the interaction of the sensing agent (e.g., a bioluminescent bacterium) and the analyte. The intensity of the light emission is monitored photometrically. In fluorescence, the interaction is reported by a label which, when illuminated by light of the appropriate wavelength (often in the ultraviolet region), will produce visible (fluorescent) light. Explanation of the latter two optical measurement principles listed in Table 5 is somewhat more involved. When light is transmitted along a fiber optic element, there is a light wave located just outside of the element’s outer surface, known as the evanescent wave. If the surface of the optical fiber is coated (e.g., with immobilized antibodies), these will bind their specific antigen targets, which may be fluorescently labeled. The light transmitted in the fiber optical element can excite the fluorescence. This fluorescent light can in turn be collected by the optical fiber for measurement. This evanescent wave technique is specific as well as effective even in situations where fluorescence in the bulk solution would be quenched by absorbers or where other molecules in the bulk solution would extraneously fluoresce and interfere with the signal. The surface plasmon resonance method also makes use of an evanescent wave, but in a manner that is somewhat different than the method described above. The surface plasmon resonance technique is based upon total internal reflection of p-polarized monochromatic light in the optical fiber that is incident at the critical angle to the interface between the glass and the outside aqueous environment. The electric field of the light penetrates into the surrounding environment to a distance in the range of tens of nanometers in an exponentially attenuated fashion. In an actual surface plasmon resonance measurement, this evanescent wave is reflected off from a thin gold coating on the surface of the optical element. When the light is incident to the gold film at a certain angle, its intensity will be much reduced, producing, in effect, a sharp shadow. This phenomenon is termed surface plasmon resonance and refers to a resonant energy transfer between the evanescent wave and surface plasmons. A plasmon is a cooperative collective excitation of the quantized oscillatory vibrations of the electrons in a metallic material. This effect is useful as the basis for a biosensor since the resonance conditions will be affected by material that adsorbs onto the gold surface or that reacts (via biomolecular recognition interactions) with biomolecules immobilized on the coating. The surface plasmon resonance effect can be used to measure concentrations of proteins, DNA, and sugars as well association/ dissociation reactions between ligands and their analytes.
Relationships Between Biomaterials and Biosensors
513
Mechanical principles can also form the basis for biosensing techniques [2]. The surface acoustic wave (SAW) and bulk acoustic wave (BAW) methods take advantage of this fact. Binding of material to the surface of a sensor (typically a vibrating quartz crystal), either specifically through biomolecular recognition interactions or nonspecifically due to adsorption, may modify both stress wave propagation velocities and resonant frequencies due to this mass change. This technique is sometimes also called the quartz microbalance method. This method works best when the analyte is in a gas phase. Its use in an in vivo liquid is more problematic, since the mechanics of the oscillatory motions are more complex, and the mass change could arise from both specific and nonspecific sources. Biosensors based on thermometric methods make use of the fact that biomolecular recognition interactions can involve reactions where chemical bonds may be disrupted or formed, compounds may enter into solution, etc. Many of these reactions involve significant heat production. For example, one listing of the heats of enzymatic reactions [2] ranges from 28 kJ/mol for the hexokinase/glucose reaction to 225 kJ/mol for the NADH dehydrogenase/NADH reaction (where NADH refers to the reduced form of nicotinamide adenine dinucleotide, NAD). Release of this thermal energy into the environment can raise its temperature by a measurable amount, which is sufficient to be useful for biosensor applications. Such thermometric sensors will work best under laboratory conditions but may prove to be problematic for in vivo applications due to the thermal noise within the body. Biosensors may work in a direct or indirect fashion. A sensor that works in the direct mode produces a signal whose magnitude is proportional to the concentration of the target analyte. Examples include the glucose sensors described in the next section. This section focuses on indirect detection. Biosensors that work indirectly have a more complicated mode of measurement than do sensors that work directly. Enzyme biosensors that work according to an inhibition measurement principle are examples in the indirect category. As a specific case, consider an enzyme-based biosensor with butyrylcholinesterase (BuChE) as its sensing agent. When BuChE is exposed to its substrate, butyryl choline (BC), it dissociates it according to the following reaction:
BC + H2 O − − → choline + butyric acid (BA) BuChE
(1)
BA dissociates further, producing hydrogen ions, thus changing the pH at the sensor surface as the reaction proceeds. The output of a pH meter used in the sensing system is an electrical potential, E, proportional to the pH. For a given substrate concentration and active sensing agent density, let ⌬E1 be the change in potential for the reaction above over the time interval from when the substrate first comes in contact with the enzyme until the pH finally reaches a plateau value. When a biosensor that works by enzyme inhibition comes in contact with its target analyte, a fraction of the total number of enzyme sites becomes blocked by the analyte. The reaction above will then not proceed at those sites and fewer HⳭ ions will be produced, resulting in a smaller pH shift and a smaller magnitude of the potential change, denoted ⌬E2. The degree of inhibition of the reaction by the analyte, DI, quantifies this effect and is defined as
DI =
∆E1 − ∆E 2 ∆E1 × 100%
(2)
DI is proportional to the concentration of the analyte Can. An unknown Can value can be determined by measuring DI for known concentrations to determine a calibration curve from which the unknown concentration can be determined. Various enzymatic biosensors besides those based on BuChE also utilize inhibition effects.
514
Bundy
F. Specific Examples of Biomedically Useful Biosensors The commercially available biosensors used to detect the agents listed in Table 2 are all of the catalytic type. The specific kind of BRP used for these biosensors are enzyme–substrate interactions. All the sensors are based on amperometric detection of (1) oxygen consumption or (2) H2O2 production due to electrochemical mechanisms. Depending on the pH of the medium, these reactions are as follows: Oxygen consumption: O2 + 2H2O + 4e− O2 + 4H+ + 4e−
4OH− 2H2O
(3) (4)
Hydrogen peroxide production: 2H2O
H2O2 + 2H+ + 2e−
O2 + 2H+ + 2e−
H2O2
(5) (6)
As a specific example of an application area where enzyme-based biosensors are used biomedically, we will consider the glucose sensor. A glucose biosensor can work according to various principles. In one version, the sensing agent is the enzyme glucose oxidase (GOD). The basis for the sensor is either hydrogen peroxide production or oxygen depletion, which are proportional to the glucose concentration. The relevant chemical reactions are [28]
O2 + GL − − → H2 O2 + GA (7)
GOD H2 O2 − − → 2H+ + 2e−
(8)
where GL and GA are glucose and gluconic acid, respectively. In some versions of glucose biosensor systems, the role of O2 is replaced by a chemical mediator. Ferrocine (biscyclopentadienyliron) derivatives represent one such approach [29]. In one variation of the biosensor based on the ferrocine/ferricinium (Fer/FerⳭ) redox couple, GOD is covalently bonded to a carbon foil that has been modified with 1,1′-dimethylferrocine. The ferrocine production reaction is
GL + 2Fer + + H2 O − − → GA + 2Fer GOD The amount of ferrocine is amperometrically measured at Ⳮ160 mV (SCE) from the following reaction: 2Fer − − → 2Fer+ + 2e− For the reader not familiar with amperometric detection techniques, the basic idea is the following. Amperometric methods utilize a redox substrate upon which oxidation/reduction reactions occur. The term ‘‘substrate’’ in this context refers to an electroactive surface having an interface with an appropriate electrolyte.
Relationships Between Biomaterials and Biosensors
515
In some amperometric techniques (e.g., polarography and cyclic voltammetry) the potential is swept, and the current versus potential response of the electrode is measured. In other cases, the potential is held at a constant potential, and the oxidation or reduction current is measured. In either situation, what is of most interest is the determination of current I at a specific value of potential E1/2, termed the half wave potential. For a specific analyte the E1/2 value is unique in a given electrolyte. Amperometric measurements are performed under diffusion control conditions (i.e., where the rate of reaction is limited by the rate of mass transport of the reactant to the substrate). This means that I is directly proportional to the concentration of the analyte of interest. Another interesting biomedical biosensor approach is used for the analysis of dopamine [30,31], a catechol derivative and an important chemical related to brain function. Though assays for dopamine using conventional amperometric techniques exist, ascorbic acid can represent a significant interference [4]. This can be minimized by using a biosensor based on polyphenol oxidase (PPO). PPO catalyzes the oxidation of the dihydroxy dopamine form to the quinone form. Electrochemical reduction back to the dihydroxy form produces a measurable current proportional to the dopamine concentration. Interestingly, the source of the PPO used in the biosensor is from banana pulp, which has earned the device the nickname of bananatrode. Besides being used for dopamine assays, another interesting feature of the bananatrode is that it can be employed to determine the catechol content of beer as well. As a final example of a category of biomedically useful biosensors we will consider toxicity biosensors. Unlike most of the rest of the biosensors described, these do not utilize BRPs to detect specific analytes. Rather, they provide a means for quantifying the relative toxicity of substances (or mixtures of different substances). A few concepts from toxicology are required to understand the basic ideas behind toxicity biosensors. A classic measure of the relative toxicity of substances is the LD50. This refers to the dosage of a substance that is lethal to 50% of an animal test group. Usually the toxicant is administered orally, most commonly to laboratory rats, and the LD50 is expressed in terms of milligrams per kilogram of body weight. Differing substances can display a tremendous LD50 range. According to one source [32], the range can span about ten orders of magnitude, from DEHP, or bis(2-ethylhexyl)phthalate (LD50 ⬎ 1.5⳯104 mg/kg) to botulinum toxin (LD50 ⳱ 10ⳮ5 mg/kg). Many factors can affect the toxicity of a substance or mixture. Some of these are listed in Table 7. For implants in vivo, release of toxic substances, at least when the amounts exceed threshold limits, will result in biocompatibility problems, some of which may be severe and require revision surgery. To avoid such problems implant materials must be subjected to standard procedures designed to spot any potential difficulties before new implant materials and devices are allowed on the market. Though with current technology such procedures of necessity involve testing both in animal models and human patients, biosensors can also play a crucial initial role in the screening of newly developed materials to demonstrate that such elaborate animal and clinical testing is needed and worthwhile. There is increasing interest in using easily cultured bioluminescent marine bacteria for this purpose. This approach has been used to monitor toxicity of environmental pollutants for quite some time, and species found to be suitable include Vibrio fischeri (also known as Photobacterium phosphoreum), Vibrio harveyi, and Photobacterium leiognathi. These bacteria naturally produce light as a consequence of their metabolism and respiration. Exposure to a toxicant reduces the amount of light in proportion to the toxicity as either bacteria are killed or else their health is impaired sufficiently to reduce the amount of light they can produce. The light production is catalyzed by luciferase. The details of this mechanism have been described by Blum and Gautier [33]. For studies of biomaterials, the main microorganism that has been used is Vibrio fischeri, and the approach used has been the Microtox威 test. Here the bacterial strain is exposed
516
Bundy
Table 7 Factors Affecting the Toxicity of Chemicals or Mixtures of Chemicals Concentration Dose Duration, rate, and frequency of exposure Site of exposure and route of administration Whether exposure is acute or chronic Local and systemic responses to toxicant Chemical form (speciation and bioavailability) Matrix effects (potentiation) Mixture effects (synergism or antagonism) Clearance rate
to various concentrations of the toxicant of interest to determine the concentration needed to extinguish the light intensity by 50%. This value is called the EC50. This assay can be used for soluble or precipitated or otherwise solid toxicants, and versions to gauge both chronic and acute toxicity are available. Exposure time, pH, and ionic strength conditions must be standardized for these tests in order to obtain EC50 values that are comparable between substances. If these conditions are met, though, the Microtox assay is known for sensitivity to a broad spectrum of toxicants (particularly those with an octanol/water partition coefficient below 105) and its good correlation with toxicity assays based on nonmicrobial species. Several studies have used bioluminescent bacterial assays to investigate various aspects of the behavior of biomaterials and their biocompatibility [34–40]. These studies have involved screening to assess the relative toxicity of various polymeric and metallic biomaterials and their degradation products as well as more mechanistic studies. Research in the latter category, for example, has been related to synergism and antagonism among ions in corrosion product mixtures, the influence of chelation of degradation products by proteins on toxicity, how selective leaching influences toxicity, the influence of corrosion product speciation on toxicity (e.g., triand hexavalent chromium), potentiation (i.e., enhancement of the toxicity of degradation products by naturally present components of the body’s chemistry), the relative toxicity of solid and dissolved corrosion products, and the influence of temperature on degradation product toxicity. Microtox methods can be expected to provide additional insight into various biocompatibility questions in the future as well as to gauge the effectiveness of antimicrobial treatments. G. Contribution of Biomaterials to the Biosensor Field Much of the prior discussion has dealt with how biosensors have aided the field of biomaterials and indeed the fields of biomedical engineering and medicine as a whole. This section, however, is concerned with how biosensors for certain applications might be improved by employing materials that have first been used as biomaterials. The hydrogels represent one such class of materials that seems promising in this regard. Hydrogels are extremely hydrophilic, porous threedimensional polymer networks that imbibe large quantities of water (up to 98%) when placed in an aqueous environment. Their uses as biomaterials have been discussed by many researchers, e.g., Peppas and coworkers [41,42] and Refojo [43]. Main uses are in soft contact lenses; scleral buckling materials; blood contact applications; artificial skin, cartilage, and tendons; drug delivery systems; wound-healing bioadhesives; maxillofacial reconstruction; vocal cord replacement biomaterials; and artificial kidney membranes.
Relationships Between Biomaterials and Biosensors
517
In the context of the present chapter, though, hydrogels can be used as carriers for sensing agents when these are coupled through various means to the polymer structure. So far this concept has been exploited only to a limited extent, but hydrogel-based biosensors (HBBs) that use the biomaterial polyhydroxyethylmethacrylate (PHEMA) for the detection of the pesticide Malathion have been described [44–46]. The use of a 6-acryloyl--O-methyl galactopyranoside hydrogel to attach multiple proteins for simultaneous detection of a number of analytes has also been discussed [47,48]. Despite the restricted experience with HBBs so far, they appear in principle to have a number of potential advantages, as enumerated in Table 8. A number of structural and fabrication parameters of hydrogels can be adjusted depending on the details of their synthesis. These include molecular weight, the nature of the monomer(s), crosslink density, the molecular weight between crosslinks, charge, and the degree of crystallinity. These parameters can have strong influences on engineering properties that are pertinent to biomaterial applications, such as strength, optical quality, diffusion characteristics, inertness, biocompatibility, and sterilizability. They also can be used to control properties that are very important for hydrogel-based biosensor applications, such as pore size, percent H2O, and physical form (sheet, membrane, etc.). In principle, HBBs should be able to detect a wide range of analytes of interest for biomedical applications, environmental pollution monitoring, military and antiterrorism defense purposes, and food industry applications. The possibility of using new types of sensing agents such as cell receptors that have seen only restricted prior application for biosensors is particularly exciting. Among some of the more interesting possibilities for novel sensing agents for HBBs are the dopaminergic D2 receptor for detection of dopamine; the nAChR for its ability to bind to many types of chemicals such as insectides (including organophosphates, similar in structure to some chemical warfare agents), and various toxins (e.g., curare and botulinum toxin); and the hrER (human recombinant estrogen receptor) for its ability to bind to natural estrogens, certain metals, and a wide variety of ecoestrogens. In the latter category a compound of particular interest due to its presence in various biomaterials is BIS-GMA. The fact that, depending upon how they are synthesized, hydrogels can be formulated to respond to stimuli present in the environment or in vivo (pH, ionic strength, temperature, etc.) adds an added dimension of versatility to their use in biosensors.
III. CONCLUSION Biosensors play an important role in many different fields, including biomedical applications. They can be used for in vivo and in vitro detection of many different analytes that affect human
Table 8 Potential Advantages of Hydrogel-Based Biosensors High percentage of H2O can aid retention of biofunctionality of sensing agent. Biomolecules can be coupled to hydrogel structure without denaturation. Biomolecules can be coupled to hydrogel structure without steric limitations. Detection of airborne contaminants may be facilitated by ease of diffusion into HBB. Pore size control may allow exclusion of some agents (serving to enhance selectivity and reduce interferences). Low detection limits. Treatments to renew/refresh HBB may be possible (forming the basis for renewable sensors).
518
Bundy
health and for quantification of their concentrations. The technology associated with biosensors is very diverse. A wide variety of biological agents can be used as active sensing elements utilizing either extremely specific biomolecular recognition interactions, as is the case for affinity and catalytic biosensors, or else broad sensitivity to a spectrum of chemicals, as is the situation with toxicity biosensors. A larger number of physical effects can be used for signal transduction. These include, for example, electrochemical, optical, thermal, and mechanical phenomena. The field of materials science has contributed a great deal toward the development and improvement of biosensors. The biomaterials field may now be able to do this as well, since certain hydrogels originally used in biomedical applications may enhance the performance of biosensors in various ways that could have applications in a number of different biomedical, environmental, and defense areas.
ACKNOWLEDGMENTS Financial support from NSF-EPSCoR/LEQSF contract (2001-04)-RII-02 (Micro/Nano Technologies for Advanced Physical, Chemical, and Biological Sensors Consortium) and from NASA Glenn and Langley Research Centers for the TIMES (Tulane Institute for Macromolecular Engineering and Science) grant is gratefully acknowledged. The assistance of Rich Coller for searching of the literature pertinent to biosensors is also gratefully appreciated.
REFERENCES
1. Clark LC, Lyons C. Electrode systems for continuous monitoring in cardiovascular surgery. Ann. NY Acad. Sci 1962; 102:29ff. 2. Yager P. Biomedical sensors and biosensors. In: Ratner BD, Hoffman AS, Schoen FJ, Lemons JE, eds. Biomaterials Science—An Introduction to Materials in Medicine. San Diego: Academic Press, 1996:375–388. 3. Blum LJ, Coulet PR. Biosensor Principles and Applications. NY: Marcel Dekker, 1991. 4. Eggins BR. Biosensors: An Introduction. Chichester. UK: Wiley Teubner, 1996. 5. Kress-Rogers E. Handbook of Biosensors and Electronic Noses. Boca Raton. FL: CRC Press, 1997. 6. Cass AEG. Biosensors—A Practical Approach. Oxford: Oxford University Press, 1990. 7. Alcock SJ, Turner APF. In Vivo Chemical Sensors: Recent Developments. Bedford. UK: Cranfield Press, 1993. 8. Fraser DM. Biosensors in the Body—Continuous in Vivo Monitoring. Chichester. UK: John Wiley, 1997. 9. Ramsay G. Commercial Biosensors—Applications to Clinical, Bioprocess, and Environmental Samples. New York: John Wiley, 1998. 10. Yang VC, Ngo TT. Biosensors and Their Applications. New York: Kluwer Academic/Plenum Publishers, 2000. 11. Ga¨berlein S, Knoll M, Spener F, Zaborosch C. Disposable potentiometric enzyme sensor for direct determination of organophosphorous insecticides. Analyst 2000; 125:2274–2279. 12. Sanstro¨m KJM, Turner APF. Biosensors in air monitoring. J. Environ. Monit 1999; 1:293–298. 13. Evtugyn GA, Budnikov HC, Nikolskaya EB. Sensitivity and selectivity of electrochemical enzyme sensors for inhibitor determination. Talanta 1998; 46:465–484. 14. Pancrazio JJ, Whelan JP, Borkholder DA, Ma W, Stenger DA. Development and Application of Cell-Based Biosensors. Ann. Biomed. Eng 1999; 27:697–711. 15. Zhang S, Zhao H, John R. A theoretical model for immobilized enzyme inhibition biosensors. Electroanalysis 2001; 13(18):1528–1534.
Relationships Between Biomaterials and Biosensors
519
16. Arhypova VN, Dzadevych SV, Soldatkin AP, El’skaya AV, Jaffrezic-Renault N, Jaffrezic H, Martelet C. Multibiosensor based on enzyme inhibition analysis for determination of different toxic substances. Talanta 2001; 55:919–927. 17. Paddle BM. Biosensors for chemical and biological agents of defence interest. Biosensors and Bioelectronics 1996; 11(11):1079–1113. 18. Scott AO. Biosensors for Food Analysis. Cambridge: Royal Society of Chemistry, 1998. 19. Darnell J, Lodish H, Baltimore D. Molecular Cell Biology. New York: Scientific American Books, 1990:710–752. 20. Klausner RD, van Renswoude J, Blumenthal R, Rivnay B. Reconstitution of membrane receptors. In: Venter JC, Harrison LC, eds. Molecular and Chemical Characterization of Membrane Receptors. New York: Alan Liss, 1984:209–239. 21. Eldefrawi ME, Eldefrawi ET. Purification and molecular properties of the acetylcholine receptor from Torpedo electroplax. Arch. Biochem. Biophys 1973; 159:362–373. 22. Miller C. Ion Channel Reconstitution. New York: Plenum Press, 1986:186–197. 23. Rogers KR, Lin JN. Biosensors for environmental monitoring. Biosensors and Bioelectronics 1992; 7:317–321. 24. Gregonis D, Van Wagonen R, Andrade JD. Poly(ethylene glycol) surfaces to minimize protein adsorption. Transactions of the 2nd World Biomaterials Congress 1984:266. 25. Harris JM. Introduction to biotechnical and biomedical applications of polyethylene glycol. In: Harris JM, ed. Poly(Ethylene Glycol) Chemistry: Biotechnical and Biomedical Applications. New York: Plenum Press, 1992:127–136. 26. Kim SW. Nonthrombogenic treatments and strategies. In: Ratner BD, Hoffman AS, Schoen FJ, Lemons JE, eds. Biomaterials Science—An Introduction to Materials in Medicine. San Diego: Academic Press, 1996:297–308. 27. Tiefenauer L, Grubelnik A, Padeste C. Prevention of protein adsorption to active gold electrodes. Transactions of the 6th World Biomaterials Congress, 2000. 28. White SF, Turner APF. Enzymes, cofactors, and mediators. In:E. Kress-Rogers Turner APF, ed. Handbook of Biosensors and Electronic Noses. Boca Raton. FL.: CRC Press, 1997:43–57. 29. Bardeletti G, Se´chaud F, Coulet PR. Amperometric enzyme electrodes for substrate and enzyme activity determinations. In: Blum LJ, Coulet PR, eds. Biosensor Principles and Applications. New York: Marcel Dekker, 1991:7–45. 30. Sidwell JS, Rechnitz GA. ‘‘Bananatrode’’—an electrochemical biosensor for dopamine. Biotechnol. Lett 1985; 7:419ff. 31. Wang J, Lin MS. Mixed plant tissue–carbon paste electrode. Anal. Chem 1988; 60:1545ff. 32. Manahan SE. Environmental Chemistry. 6th Ed.. Boca Raton. FL.: Lewis Publishers, 1994:654. 33. Blum LJ, Gautier SM. Bioluminescence- and chemiluminescence-based fiberoptic sensors. In: Blum LJ, Coulet PR, eds. Biosensor Principles and Applications. New York: Marcel Dekker, 1991:213–247. 34. Burton SA, Petersen RV, Dickman SN, Nelson JR. Comparison of in vitro bacterial bioluminescence and tissue culture bioassays and in vivo tests for evaluating acute toxicity of biomaterials. J. Biomed. Mater. Res 1986; 20:827–837. 35. Nalecz-Jawecki G, Rudz B. Evaluation of toxicity of medical devices using Spirotox and Microtox tests. J. Biomed. Mater. Res 1997; 35:101–105. 36. Shettlemore MG, Bundy KJ. Toxicity measurements of orthopedic implant alloy degradation products using a bioluminescent bacterial assay. J. Biomed. Mater. Res 1999; 45:395–403. 37. Taylor MS, Daniels AU, Andriano KP, Heller J. Six bioabsorbable polymers: in vitro toxicity of accumulated degradation products. J. Appl. Biomater 1994; 5(18):151–157. 38. Shettlemore MG, Bundy KJ. Examination of in vivo influences on bioluminescent microbial assessment of corrosion product toxicity. Biomaterials 2001; 22:2215–2228. 39. Bulich A, Tung K, Scheibner G. The luminescent bacteria toxicity test: its potential as an in vitro alternative. J. Biolumin. Chemilumin 1990; 5:71–77. 40. Shettlemore MG, Bundy KJ. Assessment of dental material degradation product toxicity using a bioluminescent bacterial assay. Dental Mater 2002; 18:445–453. 41. Peppas NA, Bures P, Leobandung W, Ichikawa H. Hydrogels in pharmaceutical formulations. Eur. J. Pharmaceut. Biopharmaceut 2000; 50:27–46.
520
Bundy
42. Peppas NA. Hydrogels. In: Ratner BD, Hoffman AS, Schoen FJ, Lemons JE, eds. Biomaterials Science—An Introduction to Materials in Medicine. San Diego: Academic Press, 1996:60–64. 43. Refojo M. Ophthalmic applications. In: Ratner BD, Hoffman AS, Schoen FJ, Lemons JE, eds. Biomaterials Science—An Introduction to Materials in Medicine. San Diego: Academic Press, 1996: 328–335. 44. Bundy KJ, Mowat FS, Shettlemore MG, Jacob JT. Feasibility of hydrogel biosensors for environmental contaminant monitoring. Abstract Book 20th Annual SETAC Meeting, 1999. 45. Bundy KJ, Shettlemore MG, Mowat FS, Jacob JT, Bi J. Technology transfer between the fields of biomaterials and environmental science. Transactions of the 6th World Biomaterials Congress, 2000. 46. Bundy KJ, Jacob JT, Shettlemore MG, Bi J, Mowat FS. Biosensor hydrogels based on biomolecular recognition principles, Proceeding of the Annual Conference of the Institute of Biological Engineering, Vol. 3 Eiteman M, Cuello J, Nokes S, Richard T, eds. Cortland. NY: IBE Publications, 2002: 89. 47. Martin BD, Gaber BP, Patterson CH, Turner DC. Direct protein fabrication using a hydrogel ‘‘stamper.’’. Langmuir 1998; 14(18):3971–3975. 48. Gaber BP, Martin BD, Turner DC. Create a protein microarray using a hydrogel ‘‘stamper.’’. Chemtech March 1999:20–24.
24 Collagen: The Next Generation of Resorbable Biodevices in Surgery Frank DeLustro Global Biodevice Development North Charleston, South Carolina, U.S.A.
Louis Sehl Cohesion Technologies, Inc. Palo Alto, California, U.S.A.
Trudy Estridge Estridge Biomedical Consulting Fremont, California, U.S.A.
Donald Wallace Menlo Park, California, U.S.A.
I. INTRODUCTION Collagen has long been pursued in the development of medical devices and drug delivery vehicles due to its biocompatibility and diverse structural potential [1]. Indeed collagen types abound and their highly differentiated fibrillar structure conveys exquisitely different physical and biological properties [2,3]. This wide diversity in structural format and complexity provides an unlimited number of formulation and product development options. As the major structural protein in the body, collagen is responsible for most natural tissue design and matrix organization. The prevailing collagen used for medical applications is the abundant, interstitial type I collagen, usually from xenogeneic (bovine or porcine) sources. Type I collagen and, to a lesser extent, type III collagen dominate the in vivo tissue repair and regeneration process [4,5]. Indeed the tissue repair process requires complex interactions of a multitude of proteins (matrix, growth factors, receptors, and other proteins in signal transduction cascades) to achieve success. At the end of this process, fibrillar type I collagen is responsible for the resulting strength, integrity, and functionality. It is this fairly rapid healing process with deposition of type I collagen that allows the individual to recover from tissue damage in short order and return to biological function. In addition, once freed of other contaminating proteins, purified type I collagen has long been known to be of low immunogenicity and can be manipulated through a variety of crosslinking and process technologies to create a core component of an ideal biological matrix [6]. Due to the ubiquitous and voluminous presence of type I collagen in nature, the search for 521
522
DeLustro et al.
surgical biomaterials from animal tissue resulted in its historical dominance in surgical suture, bone, patches, and tissue repair implants (Table 1). The recent movement of surgery toward minimally invasive procedures has driven the need for a spectrum of tools which can be used in liquid forms for ease of administration through narrow openings and trocars, and which are resorbable to provide elimination from the patient without the need for surgical removal. As laparoscopic procedures have supplanted open ones in abdominal surgery for many applications, it has been limited by the absence of tools to allow adequate hemostasis, to attach tissues together using adhesives instead of suturing in confined areas, and to obtain reliable fluid-tight seals. Indeed the need to obtain adequate hemostasis in laparoscopic procedures is one of the major reasons for conversion to open procedures during surgery [7]. In the last few years we have begun to see the advent of liquid technologies, many based on collagen, to permit effective hemostasis, sealing, and tissue closure in surgery and thereby advancing the surgical advent of minimally invasive surgery in the abdomen and the chest.
II. DISCUSSION A. Hemostasis In surgery the control of bleeding is of paramount importance and must be achieved rapidly to minimize patient morbidity and mortality. Type I collagen in tissue provides the natural substrate
Table 1 Collagen in Medical Device Use: A Representative Sampling Application Artificial skin
Artificial dura Bone graft substitute Corneal shield Hemostasis
Infection control Injectable for asthetic surgery Urinary incontinence Vascular access closure Vascular implant Wound dressing
Device
Source
Dermagraft Graftskin OrCel DuraGen Collagraft Healos Bio-Cor CollaCote Avitene CoStasis FloSeal Helistat Instat BioPatch VitaCuff Zyderm Zyplast Contigen Angio-Seal VasoSeal Hemashield InterGard Apligraf
Advanced Tissue Sciences Organogenesis Ortec International Integra LifeSciences NeuColl Orquest Bausch & Lomb Sulzer Dental Davol Cohesion Technologies Fusion Medical Integra LifeSciences Ethicon Ethicon Arrow Int’l and Bard Access McGhan Medical McGhan Medical C.R. Bard St. Jude Medical Datascope Boston Scientific Datascope Organogenesis
Collagen Resorbable Biodevices in Surgery
523
for the initiation of coagulation due to platelet adhesion, aggregation, and degranulation [8]. This initiates the release of coagulation factors as a first step in hemostasis following compromise of the vascular integrity through trauma or surgery. It is only logical then that the prevalent medical devices devised for hemostasis are based on collagen (Table 1). Collagen powders, pledgets, and sponges have been used in surgery for decades as a staple in the armamentarium for the control of bleeding. Comparing gelatin and native collagen preparations, native collagen has proven to be superior with its helical structure and platelet binding sites [9,10]. These collagen-based devices have become excellent additions to the surgeon’s use of cautery, staples, and sutures to stop difficult bleeding, especially the diffuse bleeding from the myriad small vessels that are often hard to pinpoint, access, and effectively control. The biological reactions to these collagen hemostatic devices (including the cellular infiltration, implant resorption, and tissue repair processes) have been well characterized. The chronology of cellular infiltration into these collagen matrices at sites of hemostasis mirrors the pattern observed in typical wound healing responses [4,5], with polymorphonuclear cells being replaced in the first few days by a mononuclear cell infiltrate and subsequently resulting in fibroblast infiltration and new connective tissue deposition [11,12]. In assessing the immunologic profiles of these agents, the most significant responses observed have been against noncollagen contaminants in the xenogeneic collagen devices [13]. Medical implants of highly purified collagen produce minimal immunologic sensitization and, more importantly, those responses are most often of no clinical consequence. The natural properties of autologous collagen in wound healing and tissue repair are significant contributors to the properties of these xenogeneic collagen hemostatic agents in leading to tissue deposition as the implant is resorbed. The need to treat diffuse and difficult bleeding with easy-to-use, fluid-based biodevices in surgery has led more recently to the development and commercialization of a next generation of collagen-based hemostatic agents. The role for such agents was first filled by fibrin/thrombin composites, which provided a ready source of coagulation with the application of fibrinogen and thrombin to the bleeding tissue [14]. These products included those derived from human pooled blood as well as the ‘‘home brew’’ versions made in the operating room from cryoprecipitate and bovine thrombin. As an advance of this concept, both liquid- and solid-based collagencontaining products evolved to combine the hemostatic effectiveness of collagen and thrombin, with the matrix bioregenerative properties of collagen and fibrin. The first of these products was TachoComb威 hemostatic sponge, which utilized the composite of collagen/thrombin/fibrinogen to enhance the action of existing technologies in control of difficult bleeding [15,16]. This ready-to-use product resulted in a more effective and predictable hemostasis under the most difficult of circumstances. However the limitations of this product have been the difficulty in treating irregular tissue surfaces, hard-to-reach anatomical locations where tamponade cannot readily be achieved, and in use in minimally invasive surgery. As a result, the next product to use these biological components, Floseal威 surgical sealant (a gelatin/ thrombin combination), solved some of these limitations [17]. This product was an advance in providing a pastelike material which could be used to control bleeding in unusually problematic sites, but it requires a few minutes to prepare in the operating room. In addition this approach still often utilizes tamponade and is not easily adapted to minimally invasive surgery or laparoscopic surgery. It was as an offshoot of these limitations and findings that the next generation of liquidbased biodevices, CoStasis威 surgical hemostat, was introduced containing collagen, thrombin, platelets, and fibrinogen [18]. This product takes advantage of the historical biological assets and proven safety of highly purified bovine collagen and thrombin and also of the safety of the use of the patient’s own plasma as a source of additional clotting agents including fibrinogen and platelets. The utilization of autologous plasma eliminates the safety issues and infectious
524
DeLustro et al.
disease concerns associated with pooled blood derivatives. Simplified preparation instruments reduce the total preparation time of this surgical assist device to approximately 6–8 min, and the resulting surgical tool is readily usable in minimally invasive surgery or even through catheter delivery. This flowable material can easily be applied to irregular or difficult-to-reach bleeding surfaces and does not require tamponade. In preclinical studies, this easily applied hemostatic device has proven effectiveness in many bleeding tissue sites and in hosts with severely impaired hemostatic potential [19,20]. The biological properties of hemostasis, tissue repair, and healing were superior to control agents, including collagen sponge and fibrin glue. In controlled human clinical trials, CoStasis has been reported to provide superior results in the control of difficult and life-threatening bleeding in very diverse tissue sites [21]. Early application in cardiothoracic surgery demonstrated a life-preserving uniqueness of the robust hemostatic potential of CoStasis [22]. Later published studies demonstrated superiority to existing traditional collagen hemostats in hepatic surgery [23], plastic surgery [24], orthopedic surgery [25], and trauma [26,27], as indicated in Table 2. In addition to these studies, the ability of this hemostatic agent to be used in difficult-to-reach sites through catheter delivery has been described by Mikes et al. [28] in patients with gastrointestinal bleeding cancers. Although no results have yet been published in human patients, Robertson et al. [29] have reported the use of CoStasis in spinal laminectomy procedures in a porcine model with excellent results of hemostasis and dural sealing. This latest development in surgical devices for hemostasis exemplifies easy application of liquid barriers of collagen, thrombin, platelets, and fibrin at the bleeding site, providing highly robust control of bleeding through multiple mechanisms of action.
Table 2 Surgical Hemostasis with a Flowable Collagen-Based Hemostatic Agent Surgical application Controlled clinical trials in multiple surgical specialties Aortic homograft root replacement and aortic valve replacement Hepatic surgery: hemihepatectomy and segmental resection Plastic surgery: bleeding at muscle-flap donor sites Orthopedic surgery: sternotomy and iliac crest bone donor sites Acute trauma: perforation by an inferior vena cava filter Acute trauma: complex liver injuries Gastrointestinal surgery: severe bleeding from metastatic cancer Neurosurgery: dural sealant and hemostasis
Comments
Ref.
Improved hemostasis with collagen/thrombin/plasma suspensions Critical control after failure of conventional techniques Controlled clinical trial results of solid versus liquid (collagen/thrombin/plasma) formulations Controlled clinical study in large and moderate sized flaps with significant bleeding Advantageous use of liquid collagen/thrombin/plasma in bone bleeding clinical trial Uncontrolled case studies of emergency trauma application of collagen/thrombin/plasma Further studies in trauma surgery of collagen/thrombin/plasma Endoscopic applications of fluid collagen/thrombin/plasma hemostatic devices Studies of fluid collagen/thrombin/plasma in a porcine laminectomy model
21 22 23
24 25
26 27 28
29
Collagen Resorbable Biodevices in Surgery
525
B. Adhesives While the clinical pursuit of a tissue adhesive to replace or minimize suture use has gone on for decades, the advent of minimally invasive surgery has accelerated the search and amplified the need. As surgery is performed through small access sites and trocars, the value of adhesives to reduce the reliance on sutures for tissue closure has increased proportionally as a means of reducing time in the operating room and minimizing complications. Although we have seen the movement to laparoscopic procedures for cholecystectomies and hysterectomies, major abdominal procedures involving bowel and off-bypass cardiac procedures on the beating heart remain largely as open operative procedures and have demands of reliable fluid-tight seals and strength to guarantee closure until adequate tissue repair has ensued. The earliest tissue adhesives relied on the strength and permanence of cyanoacrylate chemistry or aldehyde crosslinking. However, these approaches have the disadvantages of inflexible structures which do not mimic the elasticity of the natural soft tissues of the body and can result in erosion at the juncture of the rigid adhesive and more pliable natural host tissue. In addition to these physical limitations of biocompatibility, the chemistry of current cyanoacrylate technologies is toxic to tissue [30]. More recent adhesive technologies, such as gelatin-resocin-formalin (GRF) and Bioglue威 surgical adhesive, have taken advantage of formaldehyde and glutaraldehyde crosslinking of proteins, respectively [31,32]. Despite its strength, this approach has proven to have the same tissue incompatibilities seen using cyanoacrylates [33]. These strong chemical crosslinking approaches have proven of great value in certain critical surgical procedures such as aortic reconstruction and repair, but are of limited value in more widespread surgical utilization due to their biological incompatibilities. To better achieve tissue compatibility and appropriate compliance, collagen has been used for some time to assist in tissue closure procedures. Besides the tissue grafts used in surgery, bovine collagen tissue has been applied as an adjunct to sutures and staples to help reinforce the suture line in tissue closure procedures and provide a tissue biocompatibility and compliance similar to the surrounding host tissue structures [34]. To eliminate the need for sutures or staples, recent literature describes the creation of denatured collagen (gelatin) sheets which have been crosslinked with several chemistries to provide pendant crosslinking sites available to react with the host tissue on application to the site in surgery [35,36]. This approach simplifies the surgical use of collagen sheets for tissue repair and reduces the time of application by removing the need for suture or staple application. However, these solid formulations do not allow flexible application to irregular tissue surfaces or use in minimally invasive surgery, so fluid adhesives using collagen technology have continued to evolve. With regard to collagen-based flowable adhesives, as early as 1955 [37], F. O. Schmitt and colleagues observed that liquid-soluble collagen could create gels, requiring from 15 min to several hours to form. Such gels can be rather firm yet friable [38], and they possess poor adhesive bonding to surrounding tissue. While they may be useful as an injectable formulation for anchoring cells in a specific tissue site, their mechanical performance is inadequate as a surgical adhesive. In a more sophisticated utilization of denatured collagen or gelatin, Bowyer and Jeffrey [39] describe a mixture of noncrosslinked and crosslinked gelatins which can be used to bond tissue. Gelatin is crosslinked by irradiation or by heating dry gelatin above 100⬚C for controlled lengths of time. Their formulations showed bonding to tissue, but the tests were apparently not performed under physiological conditions, e.g., room temperature and hydration at 37⬚C. Dapper et al. [40] showed that gelatin alone could be used to make bonds that are stable in physiological saline at 37⬚C, but only when the gelatin was heated above 80⬚C (preferably between 95 and 110⬚C). Presumably some kind of dehydrothermal linkage can be achieved
526
DeLustro et al.
under these conditions. Reich et al. [41] demonstrated that gels formed from high concentration gelatin solutions (40–55%, w/v) did not survive in physiological saline at 37⬚C for more than 24 h before dissolving. Using molten gelatin, one can effect a bond with biological tissue, by extruding it onto incisions or by melting gelatin films on tissue. Nevertheless unless the temperature becomes very high (near 100⬚C), or unless chemical crosslinkers such as glutaraldehyde are present, it is unlikely that stable closures suitable for surgical use can be achieved. Turning to chemically crosslinked formulations, as tissue adhesives in diverse surgical applications, several collagen formulations have been developed which use native collagen or gelatin with a variety of reactive agents [42,43]. Perhaps the most developed of these collagenbased glues employs native bovine type I collagen and traditional fibrin sealants. Described by Sierra [44], fibrillar collagen suspensions add strength and elasticity to fibrin networks that have been formed through activation of fibrinogen by thrombin. Much like the CoStasis hemostat previously described, this composite takes advantage of the structural and physical properties of fibrillar collagen and the biocompatibility of this natural structural protein. Improvements were readily seen using both fibrillar and afibrillar preparations of collagen with fibrin glues [44,45]. Although strength was still not adequate for a robust tissue adhesive, this approach provides the flowable properties needed in laparoscopic and thorascopic surgery, as well as the tissue regeneration and repair enhancements afforded by native type I collagen. Recent technological development [46] has permitted a highly biocompatible crosslinking at the tissue site using derivatives of polyethylene glycol (PEG). The advent of CoSeal威 surgical sealant for use in cardiothoracic and vascular surgeries permits effective sealing for prevention of blood loss and leakage at sites of anastomosis, incision, or trauma with a rapid preparation time of 1–2 min. This liquid biodevice resorbs in 7–10 days and is a safe synthetic composition. However, this composition lacks the strength to be a reliable tissue adhesive and cannot provide resistance to degradation in enzymatically active tissue sites. The next generation of tissue adhesive, which combines the advantages of collagen biocompatibility and the polymerization properties of this latest modified PEG technology, [47] has recently been described. PEG matrices tend to be very elastic and relatively weak and to swell, degrade relatively rapidly, and resist cellular attachment. Introduction of chemically modified collagen into a PEG matrix in a unique polymer gel yields substantially greater strength and persistence than PEG alone (Figs. 1 and 2). The polymerization rate of the gel is adjustable, from instant setting (amenable to sealant uses) to setting over several minutes (allowing apposition of surfaces for adhesive uses). This combination results in an effective answer to the needs of the surgeon seeking closure with minimization of staples or sutures: easy use, high tensile strength, tissue compliance, and excellent tissue regenerative and repair properties. The collagen and PEG chemistries are readily adaptable to differing surgical requirements for immediate or slow polymerization at the surgical site, resorption rate from weeks to months, and strength/rigidity of the final gelled composition. Early animal models demonstrated the ability of this unique composite [collagen, PEG backbone (thiol PEG), and PEG crosslinker] to effectively adhere to sites mimicking mastectomies and to discourage the formation of seromas [48]. Further experimental use of this biodevice in surgical models with high enzymatic activity resulted in good surgical handling and excellent clinical results. Wise et al. [49] examined closure of bile duct anastomoses and used these collagen/PEG compositions as an adjunct to incomplete choledochocholedochostomy. They reported significant improvements in reducing bile leakage and recommended clinical evaluation to prevent this problematic complication in human biliary reconstruction. Later experiments by Rosen et al. [50] were conducted in pancreatic surgery to examine the ability of these collagen/ PEG compositions to prevent enzyme-rich fluid leakage. Their results suggest impressive strength, durability, and adherence of this fast-polymerizing tenacious biodevice. The authors indicate that it would be ideal for use in pancreatic repair in open or laparoscopic procedures.
Collagen Resorbable Biodevices in Surgery
527
Figure 1 Stress–strain curves for PEG hydrogels Ⳳ modified collagen. PEG Ⳳ collagen material strength was clearly increased with the incorporation of chemically modified collagen into PEG gels. Tensile strength and elastic modulus were much greater for the matrix containing collagen.
Taken together these data indicate that this collagen-based liquid adhesive can broach the next step in surgery by providing the necessary strength and tissue regenerative properties to be used effectively in most open and minimally invasive procedures. This unique combination of properties opens the door to many uses, including tissue sealant, tissue adhesive, matrix for tissue engineering, and vehicle for local delivery of cells or biologically active molecules. To fully understand and compare the various approaches to surgical adhesives, it is useful to discuss briefly the methods for measuring their performance. Such formulations need to bond to tissue (exhibit adhesive properties) and have integrity within the glue itself, i.e., not rupture (exhibit cohesive properties). Furthermore, the glue should be mechanically compatible with the adjacent bonded tissue. An adhesive that is rigid and is bonded to a deformable tissue often fails, either by detaching from the tissue or tearing it. In contrast, when there is a good match of tissue and adhesive elastic properties, the two will respond to deformations together. Measurement of adhesive, cohesive, and elastic properties of materials is time consuming and sometimes complex; this has rarely been done for the bioadhesives mentioned above. Nevertheless taking the measurements which have been published [18,51–63], it is possible to get an idea of performance of current bioadhesive materials (Table 3). In Table 3, the elastic modulus and tensile strength (to failure) are measures of the glue’s cohesive strength, and the peel strength is a measure of adhesive strength. Depending on the mode of failure, the burst and lap shear data can give information about cohesive or adhesive properties. The elongation to failure is important in matching to bonded tissue, since it indicates the ease of deformability of the glue. The effect
528
DeLustro et al.
Figure 2 Swelling profiles for PEG hydrogels Ⳳ modified collagen. Persistence in vitro and in vivo was also greater, while swelling was decreased, by the presence of modified collagen.
of different tissues on bond strength was demonstrated by data provided in references 56, 57, and 63. Some of the low values shown in the table were obtained in bonds to lung tissue, which is quite friable [56,57]. Despite missing information, formulations shown in the Table 3 can be grouped into three broad classes. The first group includes collagen (no. 1) and collagen/plasma (no. 2), which seem to have relatively weak adhesive properties and are intended for other purposes such as hemostasis. The second class includes the various hydrogels such as crosslinked gelatin, which have good adhesive properties but relatively low cohesive strength. As a group, these hydrogels are suitable as sealants (nos. 3–10) but unreliable in strength for suture and staple replacement. Fibrin glue, no. 13, seems to fit in this latter category, not because of low cohesive strength, but because of low adhesive bonding. Material no. 11, containing reactive PEG and modified collagen, appears to be the best performer of this group. Finally, the fiber-reinforced formulation, no. 12, and cyanoacrylate, no. 14, form a third distinct class which has relatively high adhesive and cohesive strength. These materials are adequate to bear substantial tensile loads. The use of PEG crosslinkers in no. 12 can achieve the necessary strength for suture and staple replacement without the drawbacks attendant with the use of cyanoacrylates (no. 14) and aldehydes in surgery (e.g., cellular toxicity or long-term persistence of the bond in vivo). The data of Table 3 again highlight the high cohesive and adhesive performance that can be achieved through combinations of modified collagen networks with the hydrogel properties of new PEG composites as embodied in formulations (nos. 11 and 12) which are polymerized in situ and exhibit excellent biocompatibility and resorption properties. C. Tissue Engineering In the last decade, the evolution of tissue repair and regeneration has morphed into the realm of tissue engineering as replacement of lost structures with metals, plastics, and matrix materials
Collagen Resorbable Biodevices in Surgery
529
Table 3 Evaluation of Surgical Adhesives: Key Performance Properties
Material
Elastic modulus (Pa)
1. Fibrillar collagen 2. Collagen/thrombin/ plasma/ 3. Collagen/thrombin/ fibrinogen 4. Gelatin/genipin/ carbodiimide 5. Gelatin-PEG diacrylate 6. Gelatin/succinimidylpoly(L-GA) 7. Gelatin/poly(L-GA)/ carbodiimide
1 ⫻ 103 34
8. Gelatin/ dialdehyde starch
8 ⫻ 106
9. Aminothiol-grafted gelatin 10. Periodate-oxidized collagen 11. Modified collagen/ succinimidylThiol-PEG 12. Modified collagen/ succinimidyl-thiolPEG/vicryl/PLLA 13. Fibrinogen/ thrombin
14. Cyanoacrylate
Tensile strength (MPa)
Elongation to failure (%)
Burst strength (MPa) ⫻ 100
Lap shear (MPa) ⫻ 100
8
Ref. 51 18
0.1
200–330
52
0.05–0.16
30
2.0
0.49
0.6–1.6
53
1.2
54
2.0–2.4
55
1.0–2.0, 0.3
0.5–0.9
56 (lap), 57 (burst, lap) 58 (burst), 59 (lap shear) 60
0.3–0.5
61
⬎0.27
0.20
1.6–2.7
2.0
62
0.84
2.0–4.6
6.0
63
0.4, 0.68, 0.4
0.15, 1.4, 0.2, 0.9
2.0–3.3
10–16, 6.8
18, 54, 57 (burst); 54–57; (lap shear) 63 (tensile, burst, lap), 54 (lap)
9.4 ⫻ 103
2.0
1 N/cm2 ⫽ 0.01 MPa; 100 mmHg burst pressure ⫽ 0.013 MPa; 1 Pa ⫽ 10 dynes/cm2.
has begun to focus more acutely on tissue or organ regeneration [64]. Indeed numerous reviews of the issues and challenges facing pioneering efforts to achieve artificial tissues and organs have emphasized the relationships between cells (or growth factors) and delivery matrices as being of paramount importance [65,66]. As this new era of tissue repair has begun, collagen has established itself as a key component due to its biological and physical properties [1–6]. The earliest uses of collagen as tissue replacement structures for dural repair [67], soft tissue augmentation [68,69], and wound dressings [4] are evolving into the easily modified delivery
530
DeLustro et al.
matrices to ensure excellent cell and tissue biocompatibility in the tissue engineering era. The use of collagen is widely cited in the tissue engineering literature. For example, it has been used in artificial skin for burn treatments [70,71], in oral mucosa [72], for corneal constructs [73], for urethral stricture repair [74], in vascular grafts [75], and as a seed bed for mesenchymal stem cells [76,77]. Clearly the pathway for use of collagen matrices as cellular highways and biological blueprints calls for adaptability to the requirements of local tissue structures and chemistry, as well as surgical procedures. Crosslinking with glutaraldehyde [78] or modified chemically active PEG [79] can permit the use of solid or liquid injectable matrices for cell delivery and can change the nature of native collagen from being hemostatic to thromboresistant. In addition the already minimal immunogenicity seen with xenogeneic purified type I collagen implants is reduced still further through the action of many of these chemical crosslinking approaches [80,81]. As a result, collagen structures with predetermined rates of resorbability and biological properties of cell attachment have been constructed in solid scaffolds [82,83] and injectable (or liquid) forms [47,79]. While other chapters in this book deal with solid tissue and organ structure implants, we focus on uniquely fluid collagen matrices for tissue replacement and regeneration. Perhaps one of the early tissue engineering biodevices using a pastelike material is Collagraft威 bone graft substitute for use in orthopedic surgery [84]. In controlled clinical trials in bone grafting of major long bone fractures, these authors found that Collagraft was able to provide a bone repair substitute that was equivalent in performance to autologous bone grafts. These bone regenerative properties were a combination effect: type I bovine collagen matrix (providing a skeleton for cell differentiation and matrix deposition), autologous bone marrow (providing progenitor cells and growth factors for osteoblast differentiation), and HA-TCP (hydroxyapatite/tricalcium phosphate, for structural strength and integrity). Later studies in animal models have demonstrated the ability of these biodevices to perform as well as autograft in spinal fusion procedures as well [85]. These bone-inducing collagen formulations can be used in solid strip forms or in a paste which can be produced as an injectable device. Using safe and effective crosslinking of native or denatured collagen suspensions in situ with active PEG technology (as discussed above), one can prepare injectable formulations for open or minimally invasive surgery, forming local depositions of collagen which can be chemically bonded to the local tissue if desired, or can form solid matrices for implantation with cells or DNA (Figs. 3 and 4). Previous investigations have used formed collagen implants impregnated with cells for a variety of tissue replacement applications, most notably skin substitutes (Table 1). In addition Estridge has described injectable preparations with properties ranging from gels to in situ polymerizing suspensions which can permit an approach to delivery of cells and matrix via injection, laparoscopy, endoscopy, and other minimally invasive approaches [86,87]. The application of these technologies combines the biologically inductive properties of native type I collagen and the localization potential of crosslinking agents such as modified PEG [46,47]. Modification of the collagen fiber size and network compactness can control the porosity and density of the resulting implants, while the PEG chemistry and concentrations can regulate the dissolution times and the binding properties. Tissue engineering involves the use of living cells, manipulated through their extracellular environment, to develop biological substitutes for the repair, replacement, maintenance, or enhancement of a particular tissue or organ (Fig. 3). While tissue engineered constructs have been proposed for an array of agents and physiological functions (Fig. 4), the ability to deliver living cells for these applications is a challenging problem with demands on the biomaterial during implantation into the host and occasionally during preimplantation in vitro culture for cell amplification. One continuing problem is the inability to effectively combine cells and material in
Collagen Resorbable Biodevices in Surgery
531
Figure 3 Application of cells in tissue engineering. Methods for current and future development of cell delivery applications.
Figure 4 Matrices in tissue engineering. Multiple diverse opportunities in cell, DNA, and drug delivery applications to meet the biological and surgical needs.
532
DeLustro et al.
such a fashion that the final composition can easily be delivered in vivo. Although collagen alone has shown remarkable flexibility and inherent biological value in this regard, the use of PEG in combination with collagen imparts unique biochemical and physical properties to the implant. Previous in vivo studies have shown collagen crosslinked with PEG to be biocompatible, and they demonstrated the ability to use this PEG crosslinker to entrap living cells in the matrix [88,89]. Chemically modified collagens and multifunctional polyethylene glycol (mPEG) can create crosslinked collagen formulations which have good biocompatibility and physical presentation for most tissue engineering requirements (Fig. 4). Furthermore the growth of human dermal fibroblasts, human aortic smooth muscle cells, and human aortic endothelial cells grown on the surface of Zyderm威 Collagen Implant, Vitrogen威 collagen, and three crosslinked matrices was significantly impacted by the matrix material [88–91]. The crosslinked matrices were produced by using fibrillar collagen with different amounts of difunctional polyethylene glycol (dPEG). Using PEG/collagen at ratios of 10:1, 1:1, and 1:10, cell proliferation on day 7 revealed that the highest proliferation of fibroblasts was on the surface of matrices with the highest concentration of dPEG. In contrast, smooth muscle cells proliferated best on collagen matrices, and endothelial cell proliferation was similar for both collagen and crosslinked collagen. The effect of collagen and mPEG concentration was reported as assessed by cell morphology, viability, proliferation, outgrowth, and histological examination of the material [91]. However, with cells mixed into the structure of the mPEG crosslinked collagen matrices, results (Fig. 6) showed that all matrix materials supported cell growth, but collagen alone was the superior matrix for fibroblasts [90]. Histology showed fibroblasts homogeneously distributed in all matrices at days 7 and 14. Estridge et al. [92] showed the in vitro cell growth on and in the collagen/polymer matrix was similar across various PEG crosslinkers. Five PEG crosslinkers were tested with fibroblasts which were grown on these collagen/PEG matrices, or alternatively fibroblasts were mixed in situ into the collagen/PEG matrix. Both methods produced an injectable matrix that was able to deliver viable cells (Fig. 5). Cell proliferation on collagen or collagen/PEG matrices was significantly higher than proliferation of cells in collagen or collagen/PEG matrices. Increasing the concentration of mPEG in the collagen matrix reduced the proliferation of the fibroblasts at all collagen concentrations (Fig. 6). Previously, researchers have reported that collagen concentration could play a role in cell ingrowth in vivo. Cell outgrowth from the mPEG matrices was initially reduced; this appeared to be associated with fewer proliferating cells. Not surprisingly, cell proliferation on collagen or collagen/PEG matrices was significantly higher than proliferation of cells in collagen or collagen/ PEG matrices. Additionally, crosslinking of collagen with glutaraldehyde has been shown to reduce the remodeling of the material. In 1999, Estridge [86] reported on the use of collagen multifunctional polyethylene glycol crosslinked collagen matrices (collagen/mPEG) and mPEG/PEG hydrogels to support and affect cell growth in and on the matrices of fibrillar collagen, collagen-mPEG, and mPEG/PEG. Four cell lines were used as model cells in these experiments: human dermal fibroblasts (HDF), normal human liver hepatocytes (MG3), and normal human endothelial cells. Growth of cells seeded on or mixed into the matrices and variation of the matrix for each cell type was assessed through cell morphology, viability, and proliferation. The mPEG crosslinking of collagen matrices produced a significant increase in fibroblast proliferation for cells grown on matrices, but mPEG/PEG matrices did not support cell attachment, and therefore cells could not proliferate (Fig. 7). Similarly endothelial cells and MG3 cells grew on both collagen and collagen-mPEG matrices, but not on mPEG/PEG matrices. Hepatocytes (MG3) were able to proliferate well on all collagen-containing matrices (Fig. 8). However, they grew as spheroid groups, not in monolayer, loosely attached to the mPEG/PEG hydrogels. As expected anchorage-dependent cells (fibro-
Collagen Resorbable Biodevices in Surgery
533
Figure 5 Fibroblast growth on the surface of collagen and collagen/PEG matrices. Optical density (OD) is an indirect measure of living cell number as measured from an oxidative reduction reaction. ZI⳱collagen, SG⳱difunctional succinymidal glutarate PEG, mSG⳱multifunctional SG, SE⳱difunctional PEG with ether linkage, mSE⳱multifunctional SE, SC⳱difunctional PEG with a carbonate linkage.
blasts, MG63, and endothelial cells) were not able to attach or proliferate on the mPEG/PEG materials but did well on collagen and PEG/collagen matrices. Collagen has played a central role in tissue engineering to date as shown by its use in commercial devices (Table 1). The use of biocompatible PEG to crosslink collagen vastly expands its potential application in future tissue engineered products (Fig. 4) by providing a robust matrix system which can be engineered for resorption time, physical form, and delivery requirements from DNA to cells. This future is especially promising as these materials allow the use of injectable formulations which can polymerize in situ as needed in minimally invasive surgery or nonsurgical applications. D. Drug and Biologics Delivery The value of collagen in drug and growth factor delivery, as well as matrices to facilitate the delivery of therapeutic gene products [93,94], has been evident since its earliest use as a medical device and surgical hemostatic agent. The encouragement of tissue regeneration and release properties of the fibrillar collagen structure are exceptional assets to the value of collagen in this application (Fig. 4), particularly in wound healing and tissue repair procedures [95]. A few examples will illustrate the scope of applications which have been attempted. Miyata and colleagues [96] described the use of methylated and succinylated collagen membranes to deliver ophthalmic drugs, while Chvapil [97] prepared collagen sponges to deliver zinc compounds for a contraceptive application. Miyata and colleagues [98] developed collagen pellets for sustained release of interferon. Finally, Miyata and colleagues [99] again proposed
534
DeLustro et al.
Figure 6 Fibroblast proliferation within collagen and collagen/mPEG matrices. Zyderm collagen matrix concentration is listed in mg/mL and mPEG crosslinker concentration is listed as low and high concentrations. OD is an indirect measure of living cell number as measured from an oxidative reduction reaction.
an antiadhesive collagen membrane which slowly released heparin. Other approaches utilized commercially available collagen-based biomaterials such as the use of the hemostatic device TachoComb to control bleeding and release antibiotics at sites of wound healing and trauma [100]. However the physical diversity of collagen readily lends itself toward use in a variety of liquid and solid forms [1–3,95]. This flexible characteristic is accompanied by the ability to change the implant density, persistence, and tissue regenerative properties with appropriate crosslinking technologies. In the above examples, active agents of greatly different molecular weights, including inorganic compounds, small organics, and macromolecules such as interferon, were employed. Injectable collagen has been used to deliver and localize drugs for a variety of applications, especially in the treatment of cancer. Luck and Brown [101] used flowable fibrillar collagen suspensions to target chemotherapeutic drugs, such as vincristine, to tumors. Clinical utilization of glutaraldehyde crosslinked collagen has been successfully used to achieve both embolization of major blood vessels feeding metastatic tumors of the liver and to deliver chemotherapeutic agents to the site for timed release [102]. Rosenblatt et al. [103] carried out a fundamental investigation of drug release by fibrillar and nonfibrillar collagen suspensions. These authors showed that only macromolecules would be expected to be retarded in release by diffusion.
Collagen Resorbable Biodevices in Surgery
535
Figure 7 Fibroblast proliferation on various matrices. Growth from 3 to 14 days on matrices of collagen (35 mg/mL or Vitrogen), PEG gels, and plastic. OD is an indirect measure of living cell number as measured from an oxidative reduction reaction.
This is due to the relatively open nature of the collagen fibrillar and nonfibrillar network. To effect sustained release of small molecules (⬍10,000 Da), other mechanisms, such as ionic or hydrophobic binding to collagen, must be present. Available collagen technologies are amenable to meet most, if not all, of these diverse requirements. Another means to retard release of active agents from collagen is to tether them by covalent bonds (Fig. 4). In prototypic publications on delivery of growth factors from collagen matrices, transforming growth factor-beta2 (TGF-2) was demonstrated to be effectively released from fluid compositions of fibrillar collagen through diffusion and chemical release mechanisms. The former work [104] utilizes compositions of fibrillar suspensions of collagen and heparin to release entrapped TGF-2 with good in vitro release kinetics and demonstrable in vivo tissue regeneration promotion. Alternatively, Bentz et al. [105] showed that TGF-2 could remain active after crosslinking to collagen fibrillar suspensions using PEG crosslinking technology. Although covalently crosslinked to fibrillar collagen via PEG bridges, the TGF-2 remained fully active and demonstrated activity in vitro and in vivo.
536
DeLustro et al.
Figure 8 Hepatocyte proliferation on various matrices. Hepatocyte growth from 3 to 14 days on matrices of collagen (35 mg/mL or Vitrogen), PEG gels, and plastic. OD is an indirect measure of living cell number as measured from an oxidative reduction reaction.
III. CONCLUSIONS Collagen technology provides a broad constellation of benefits for tissue engineering opportunities. Leading among these is the biological advantage of collagen as a matrix which promotes extensive tissue healing (repair and regeneration). In addition collagen chemistry provides a flexible diversity of substrate formulations from solid to liquid forms, and from rapidly resorbed gelatin to densely crosslinked, long-term implants. These properties afford a basis for recent developments in biosurgical devices to achieve free-flowing tools for use in open and, especially, minimally invasive surgery for hemostasis, sealing surgical incision sites and anastomoses, and adhering tissues together to replace or minimize use of sutures or staples. These procedures are particularly difficult at surgical sites of limited access and in tissues with significant risk of failure. Thus besides use in current surgical procedures in almost all specialties, these advanced biodevices promote the opportunity to successfully adopt minimally invasive procedures for complex abdominal surgery of the bowel and liver, and off-bypass cardiac surgery. As we move into the next decade, the use of such fluid collagen devices will be combined with traditional drugs, novel growth factors, cells, and/or gene therapy for more advantageous synergies in the tissue repair process.
Collagen Resorbable Biodevices in Surgery
537
ACKNOWLEDGMENTS The authors would like to extend their appreciation to the technical staff of Cohesion Technologies, Inc., for their contribution to these studies and to the many contributing groups who have facilitated the flow of these promising technologies into the surgical suite. REFERENCES
1. Pachence JM. Collagen-based devices for soft tissue repair. J. Biomed. Mater. Res. 1996; 33:35–40. 2. Piez K. Collagen types: a review. In: Sen A, Thornhill T, eds. Development and Diseases of Cartilage and Bone Matrix: Alan R. Liss, 1987:1–19. 3. Piez K. Collagen. In: Encyclopedia of Polymer Science and Engineering. 2nd Ed. Vol. 3. New York: John Wiley & Sons, 1985:699–727. 4. McPherson JM, Piez K. Collagen in dermal wound repair. In: Clark RA, Henson PM, eds. The Molecular and Cellular Biology of the Wound Repair: Plenum Publishing, 1988:471–496. 5. Lawrence WT. Physiology of the acute wound. Clin. Plast. Surg. 1998; 25:321–340. 6. Miyata M, Taira K. Collagen engineering for biomedical use. Clin. Mater. 1992; 9:139–148. 7. McGinnis DE, Strup S, Gomella LG. Management of hemorrhage during laparoscopy. J. Endourol. 2000; 14:915–920. 8. Sixma JJ, van Zanten H, Huizinga EG, van der Plas RM, Verkley M, Wu YP, Gros P, de Groot PG. Platelet adhesion to collagen: and update. Thrombosis and Haemostasis 1997; 78:434–438. 9. Coln D, Horton J, Ogden ME, Buja LM. Evaluation of hemostatic agents in experimental splenic lacerations. Am. J. Surg. 1983; 145:256–259. 10. Wagner WR, Pachence JM, Ristich J, Johnson PC. Comparative in vitro analysis of topical hemostatic agents. J. Surg. Res. 1996; 66:100–108. 11. Anselme K, Bacques C, Charriere G, Hartmann DJ, Herbage D, Garrone R. Tissue reaction to subcutaneous implantation of a collagen sponge: a histological, ultrastructural, and immunological study. J. Biomed. Mater. Res. 1990; 24:689–703. 12. Smith KJ, Skelton HG, Barrett TL, Welch M, Beard J. Histologic and immunohistochemical features in biopsy sites in which bovine collagen matrix was used for hemostasis. J. Am. Acad. Dermatol. 1996; 34:434–438. 13. DeLustro F, Condell RA, Nguyen MA, McPherson JM. A comparative study of the biologic and immunologic response to medical devices derived from dermal collagen. J. Biomed. Mater. Res. 1986; 20:109–120. 14. Thompson DF, Letassy NA, Thompson GD. Fibrin glue: review of its preparation, efficacy, and adverse effects as a topical hemostat. Drug Intell. Clin. Pharm. 1988; 22:946–952. 15. Schiele U, Kuntz G, Riegler A. Haemostyptic preparations on the basis of collagen alone and as a fixed combination with fibrin glue. Clin. Mater. 1992; 9:169–177. 16. Agus GB, Bono AV, Mira E, Olivero S, Peilowich A, Homdrum E, Benelli C. Hemostatic efficacy and safety of TachoComb in surgery. Ready to use and rapid hemostatic agent. Int. Surg. 1996; 81: 316–319. 17. Oz MC, Cosgrove DM, Badduke BR, Hill JD, Flannery MR, Palumbo R. Controlled clinical trial of a novel hemostatic agent in cardiac surgery. The Fusion Matrix Study Group. Ann. Thorac. Surg. 2000; 69:1376–1382. 18. Prior JJ, Wallace DG, Harner A, Powers N. A sprayable hemostat containing fibrillar collagen, bovine thrombin, and autologous plasma. Ann. Thorac. Surg. 1999; 68:479–485. 19. Prior JJ, Powers N, DeLustro F. Efficacy of a novel hemostatic agent in animal models of impaired hemostasis. J. Biomed. Mater. Res. (Appl. Biomater.) 2000; 53:252–257. 20. Turner AS, Parker D, Egbert B, Maroney M, Armstrong R, Powers N. Evaluation of a novel hemostatic device in an ovine parenchymal organ bleeding model of normal and impaired hemostasis. J. Biomed. Mater. Res. (Appl. Biomater.) 2002; 63:37–47.
538
DeLustro et al.
21. The CoStasis Multi-Center Collaborative Writing Committee. A novel collagen-based composite offers effective hemostasis for multiple surgical indications: results of a randomized controlled trial. Surgery 2001; 129:445–450. 22. Turley K. Aortic homograft root replacement for failed freehand homograft aortic valve: effectiveness of a collagen/thrombin/plasma composite hemostat in the setting of technically complicated homograft to root anastomosis. The Heart Surgery Forum 1999; 2:143–146. 23. Chapman WC, Clavien PA, Fung J, Khanna A, Bonham A. Effective control of hepatic bleeding with a novel collagen-based composite combined with autologous plasma. Results of a randomized controlled trial. Arch. Surg. 2000; 135:1200–1204. 24. Buncke G, Sherman K. CoStasis provides superior control of diffuse bleeding at muscle-flap donor sites, compared to manual compression. J. Recontruct. Microsurg. 2000; 16:557–561. 25. Sherman R, Chapman WC, Hannon G, Block JE. Control of bone bleeding at the sternum and iliac crest donor sites using a collagen-based composite with autologous plasma: results of a randomized controlled trial. Orthopedics 2001; 24:137–141. 26. Bochicchio GV, Scalea KM. Acute caval perforation by an inferior vena cava filter in a multitrauma patient: hemostatic control with a new surgical hemostat. J. Trauma 2001; 51:991–992. 27. Bochiccio G. The management of complex liver injuries. Trauma Quarterly 2002; 15:556–76. 28. Mikes DE, Friedland S, Lin OS, Reid TR, Soetikno R. A novel method to control severe upper GI bleeding from metastatic cancer with a hemostatic sealant: CoStasis surgical hemostat. Gastrointestinal Endoscopy J. 2002; 55:735–740. 29. Robertson JT, Powers N, Nelson P. The evaluation of a new device with dural sealant and hemostasis properties in a porcine laminectomy model. Presented at the Congress of Neurological Surgeons. 51st Annual Meeting. CA: San Diego, 2001. 30. Evans CE, Lees GC, Trail IA. Cytotoxicity of cyanoacrylate adhesives to cultured tendon cells. J. Hand Surg. 1999; 6:658–661. 31. Neri E, Massetti M, Capannini G, Carone E, Sassi C. Glue containment and the anastomosis reinforcement in the repair of aortic dissection. Ann. Thorac Surg. 1999; 67:1510–1511. 32. Raanani E, Latter DA, Errett LE, Bonneau DB, Leclerc Y, Salasidis GC. Use of ‘‘Bioglue’’ in aortic surgical repair. Ann. Thorac Surg. 2001; 72:638–640. 33. LeMaire SA, Schmittling ZC, Coselli JS, Undar A, Deady BA, Chubb FJ, Fraser CD. BioGlue surgical adhesive impairs aortic growth and causes anastomotic strictures. Ann. Thorac Surg. 2002; 73:1500–1506. 34. Hazelrigg SR, Boley TM, Naunheim KS, Magee MJ, Lawyer C, Henkle JQ, Keller CN. Effect of bovine pericardial strips on air leak after stapled pulmonary resection. Ann. Thorac Surg. 1997; 63: 1573–1575. 35. Matsuda S, Iwata H, Se N, Ikada Y. Bioadhesion of gelatin films crosslinked with glutaraldehyde. J. Biomed. Mater. Res. 1999; 45:20–27. 36. Sung HW, Huang DM, Chang WH, Huang RN, Hsu JC. Evaluation of gelatin hydrogel crosslinked with various crosslinking agents as bioadhesives: in vitro study. J. Biomed. Mater. Res. 1999; 46: 520–530. 37. Gross J, Highberger JH, Schmitt FO. Extraction of collagen from connective tissue by neutral salt solutions. Proc. US Nat. Acad. Sci. 1955; 41:1–7. 38. Rosenblatt J, Devereux B, Wallace DG. Dynamic rheological studies of hydrophobic interactions in injectable collagen biomaterials. J. Appl. Polym. Sci. 1993; 50:953–963. 39. Bowyer BL, Jeffrey K. A collagen medical adhesive and its uses. European Patent 0–466–383–A1, February 7, 1991. 40. Dapper G, Wallace DG, Yamamoto R, Barrett S, Ly D, Nguyen M, Moravcsik P, Lifrieri J, Tran H, Reich C, Sawyer PN. Attachment of gelatin films to tissue using argon beam coagulator. J. Biomed. Mater. Res. 1998; 43:89–98. 41. Reich C, Wallace DG, Tammishetti S. Compositions and methods for sealing tissue and preventing post-surgical adhesions. International Patent No. WO 97/29715, August 21, 1997. 42. Mo X, Iwata H, Matsuda S, Ikada Y. Soft tissue adhesive composed of modified gelatin and polysaccharides. J. Biomater. Sci. Polym. Ed. 2000; 11:341–351.
Collagen Resorbable Biodevices in Surgery
539
43. Sekine T, Nakamura T, Shimizu Y, Ueda H, Matsumoto K, Takimoto Y, Kiyotani T. A new type of surgical adhesive made from porcine collagen and polyglutamic acid. J. Biomed. Mater. Res. 2001; 54:305–310. 44. Sierra DS. Fibrin–collagen composite tissue adhesive. In sierra DH, Saltz R, eds. Surgical Adhesives and Sealants: Current Technology and Applications: Technomic Publishing Basel, 1996:29–39. 45. Nomori H, Horio H, Suemasu K. Mixing collagen with fibrin glue to strengthen the sealing effect for pulmonary air leakage. Ann. Thorac Surg. 2000; 70:1666–1670. 46. Wallace DG, Cruise GM, Rhee WM, Schroeder JA, Prior JJ, Ju J, Maroney M, Duronio J, Ngo MH, Estridge T, Coker G. A tissue sealan based on reactive multifunctional polyethylene glycol. J. Biomed. Mater. Res. (Appl. Biomater.) 2001; 58:545–555. 47. Schroeder JA, Chu G, Prior J, Estridge T, Ngo MH, Trollsa˚s M, Sehl LC. A self-polymerizing, strong, flexible, adherent and biocompatible matrix formed by combining collagen with functionalized polyethylene glycol. Presented at the 28th Annual Meeting of the Society for Biomaterials, 2002. 48. Powers JN, Nelson P, Schroeder JA, Sehl LC. Sealant versus adhesive forms of a novel selfpolymerizing PEG-collagen gel for seroma prevention in a rat lymphadenectomy model. Presented at the 28th Annual Meeting of the Society for Biomaterials, 2002. 49. Wise PE, Wudel LJ, Belous AE, Allos TM, Kuhn SJ, Feurer ID, Washington MK, Pinson WC, Chapman WC. Biliary reconstruction is enhanced with a collagen–polyethylene glycol (PEG) sealant. J. Am. Surg. 2002; 68:553–561. 50. Rosen M, Walsh RM, Goldblum JR. Application of a new collagen based sealant for the treatment of pancreatic injury. Digestive Disease Week, Abstracts. 2002. 51. Rosenblatt J, Devereux B, Wallace DG. Dynamic rheological studies of hydrophobic interactions in injectable collagen biomaterials. J. Appl. Polym. Sci. 1993; 50:953–963. 52. Sierra DH, Luck EE, Brown DM. Surgical adhesive material. US Patent 5.290,552, Mar. 1, 1994. 53. Sung H-W, Huang D-M, Chang W-H, Huang R-N, Hsu J-C. Evaluation of gelatin hydrogel crosslinked with various cross-linking agents as bioadhesives: In vitro study. J. Biomed. Mater. Res. 1999; 46:520–530. 54. Nakayama Y, Matsuda T. Photocurable surgical tissue adhesive glues composed of photoreactive gelatin and poly(ethylene glycol) di-acrylate. J. Biomed. Mater. Res. (Appl. Biomater) 1996; 48: 511–521. 55. Iwata H, Matsuda S, Mitsuhashi K, Itoh E, Ikada Y. A novel surgical glue composed of gelatin and N-hydroxysuccinimide activated poly(L-glutamic acid). Biomaterials 1998; 19:1869–1876. 56. Otani Y, Tabata Y, Ikada Y. A new biological glue from gelatin and poly(L-glutamic acid). J. Biomed. Mater. Res. 1996; 31:157–166. 57. Otani Y, Tabata Y, Ikada Y. Sealing effect of rapidly curable gelatin–poly(L-glutamic acid) hydrogel glue on lung air leak. Ann. Thorac. Surg. 1999; 67:922–926. 58. Reich C, Wallace D, Tammishetti S. Compositions and methods for sealing tissue and preventing post-surgical adhesions. International Patent WO 97/29715, Aug. 21, 1997. 59. Tammishetti S, Wallace DG, Reich C. Dialdehyde starch cross-linked gelatin: a new tissue sealant. Society for Biomaterials 23rd Annual Meeting Abstracts, 1997:275. 60. Constancis A, Barc G, Tiollier J. COLCYS as surgical adhesives. Society for Biomaterials Annual Meeting Abstracts, 1995:258. 61. DuPont D, Rossin D, Lathuliere S, Tardy M, Gravagna P, Tiollier J. New surgical collagen gel sealant (glue) based on controlled oxidized collagen. Society for Biomaterials Annual Meeting Abstract, 1995:242. 62. Wallace DG, Cruise GM, Rhee WM, Schroeder JA, Coker GT, Maroney M. Rapid gelling biocompatible polymer composition. US Patent 6,312,725 B1, Nov. 6, 2001. 63. Wallace DG. Compositions that form interpenetrating polymer networks for use as high strength sealants. International Patent WO 01/16210 A1, Mar. 8, 2001. 64. O’Reilly S. Opportunities in tissue and organ regeneration. Biomaterials and Surgery 2000; 2: 127–136. 65. Langer R, Vacanti KP. Tissue engineering. Science 1993; 260:920–926. 66. Minuth WW, Sittinger M, Kloth S. Tissue engineering: generation of differentiated artificial tissues for biomedical applications. Cell Tiss. Res. 1998; 291:1–11.
540
DeLustro et al.
67. Narotam P, van Dellen JR, Bhoola KD. A clinicopathological study of collagen sponge as a dural graft in neurosurgery. J. Neurosurg. 1995; 82:406–412. 68. Wallace DG, McPherson JM, Ellingsworth L, Cooperman L, Armstrong R, Piez KA. Injectable collagen for tissue augmentation. In: Nimni ME, ed. Collagen Vol. III Biotechnology. Boca Raton. FL: CRC Press, 1988:117–144. 69. Keefe J, Wauk L, Chu S, DeLustro F. Clinical use of bovine collagen: a decade of experience. Clin. Mater. 1992; 9:155–162. 70. Ojeh NO, Frame JD, Navsaria HA. In vitro characterization of an artificial dermal scaffold. Tiss. Eng. 2001; 7:457–472. 71. Horch RE, Debus M, Wagner G, Stark GB. Cultured human keratinocytes on type I collagen membranes to reconstitute the epidermis. Tiss. Eng. 2000; 6:53–67. 72. Moriyama T, Asahina I, Ishii M, Oda M, Ishii Y, Enomoto S. Development of composite cultured oral mucosa utilizing collagen sponge matrix and contracted collagen gel. Tiss. Eng. 2001; 7: 415–427. 73. Orwin EJ, Hubel K. In vitro culture characteristics of corneal epithelial, endothelial, and keratocyte cells in a native collagen matrix. Tiss. Eng. 2000; 6:307–319. 74. El Kassaby K, El Kassaby AEW, Yoo JJ, Retik AB, Atala A. A novel collagen matrix for urethral stricture repair. Presented at the 3rd Annual Meeting of the Tissue Engineering Society, 2000. 75. Berglund JD, Nerem RM, Sambanis A. Towards development of an entirely biological vascular graft: incorporating exogenous elastins into collagen-based constructs. Presented at the 3rd Annual Meeting of the Tissue Engineering Society, 2000. 76. Harris MT, Schantz EJ, Butler DL. Lateral contraction of mesenchymal stem cell–seeded type I collagen gels in a deformable culture dish for mechanical conditioning of cell-delivery constructs for tendon repair. Presented at the 3rd Annual Meeting of the Tissue Engineering Society, 2000. 77. Nakamura I, Hori Y, Itoh S, Fukuda S, Toba T, Ueda H, Yoshitani M, Inoue M, Nakahara T, Onishi K, Shimizu Y. Collagen sponge scaffold for in situ tissue engineering with mesenchymal stem cell seeding. Presented at the 3rd Annual Meeting of the Tissue Engineering Society, 2000. 78. Nimni ME, Cheung D, Strates B, Kadama M, Sheikh K. Chemically modified collagen: a natural biomaterial for tissue replacement. J. Biomed. Mater. Res. 1987; 21:741–771. 79. Rhee W, Carlino J, Chu S, Higley H. Bovine collagen modified by PEG. In: Harris JM, ed. Poly(Ethylene Glycol) Chemistry: Biotechnical and Biomedical Applications: Plenum Press, 1992:183–198. 80. DeLustro F, Dasch J, Keefe J, Ellingsworth L. Immune responses to allogeneic and xenogeneic implants of collagen and collagen derivatives. Clin. Orthop. 1990; 260:263–279. 81. DeLustro F, Mackinnon V, Swanson N. Immunology of injectable collagen in human subjects. J. Dermatol. Surg. Oncol. 1988; 14(Suppl. 1):49–55. 82. Doillon CJ, Cote MF, Pietrucha K, LaRoche G, Gaudreault RC. Porosity and biological properties of polyethylene glycol–conjugated collagen materials. J. Biomater. Sci. Polym. Edn. 1994; 6:715–728. 83. Itoh H, Aso Y, Furuse M, Noishiki Y, Miyata T. A honeycomb collagen carrier for cell culture as a tissue engineering scaffold. Artif. Organs 2001; 25:213–217. 84. Chapman MW, Bucholz R, Cornell C. Treatment of acute fractures with a collagen–calcium phosphate graft material. J. Bone Joint Surg. 1997; 79-A:495–502. 85. Walsh WR, Brown M, Loefler A, Harrison J, Martin T, Sonnabend DH. Mechanical and histologic evaluation of Collagraft in an ovine lumber fusion model. Clin. Orthop. 2000; 375:258–266. 86. Estridge TD. Collagen and polyethylene glycol materials for cell delivery. Presented at the 25th Annual Meeting of the Society for Biomaterials. Trans. Soc. Biomater. 1999; 25:361. 87. Estridge TD, Tefft J, Lowings E, Jones K. The effect of crosslinked collagen on the prolifieration of human cell lines in vitro. Presented at the 20th Annual Meeting of the Society for Biomaterials. Trans. Soc. Biomater. 1994; 20:312. 88. Tefft J, Estridge T, Hansen S, Lowings E. Fibroblast proliferation on collagen–polyethylene glycol crosslinked gels. 4th Annual Wound Healing Society Meeting, 1994. 89. Estridge T, Tefft J, Lowings E, Jones K. The effect of crosslinked collagen on the proliferation of human cell lines in vitro. 20th Annual Meeting of the Society for Biomaterials. Trans. Soc. Biomater. 1994; 20:312.
Collagen Resorbable Biodevices in Surgery
541
90. Estridge T. Novel collagen-based biomaterials for cell delivery. 5th World Congress. Trans. Soc. Biomater. 1996; 5:130. 91. Estridge T, Condon J. Cell delivery via crosslinked collagen matrix. Keystone Symposia Conference on Tissue Engineering. Taos. NM, January 1996. 92. Estridge T, Rhee W. Effect of PEG crosslinker on collagen matrices for cell delivery. 23rd Annual Meeting of the Society for Biomaterials. Trans Soc. Biomater. 1997; 23:473. 93. Gu D, Chandler LA, Printz M, Nguyen T, Gonzalez AM, D’Andrea M, Phillips L, Pierce GF, Sosnowski B. Tissue repair using matrix-dependent gene delivery. Tissue Engineering Society 3rd. Annual Meeting Abstracts, Dec. 2000:223. 94. Hollinger JO, Winn S, Bonadio J. Options for tissue engineering to address challenges of the aging skeleton. Tiss. Eng. 2000; 6:341–350. 95. McPherson JM. The utility of collagen-based vehicles in delivery of growth factors for hard and soft tissue wound repair. Clin. Mater. 1992; 9:225–234. 96. Miyata T, Rubin AL, Stenzel KH, Dunn MW. Collagen drug delivery device. US Patent 4,164,559, Aug. 14, 1979. 97. Chvapil M. Intravaginal contraceptive and drug release device and method for making and using same. US Patent 4,274,410, June 23, 1981. 98. Fujioka K, Sato S, Sasaki Y, Miyata T, Furuse M, Naito H. Method for producing a sustained formulation. European Patent 0 230 647 A2, Dec. 23, 1986. 99. Miyata T, Fususe M, Yamane Y, Noishiki Y. A biodegradable antiadhesion collagen membrane with slow release heparin. Trans. Am. Soc. Artif. Intern. Organs 1988; 34:687–691. 100. Parker SJ, Brown D, Hill PF, Watkins PE. Fibrinogen-impregnated collagen as a combined haemostatic agent and antibiotic delivery system in a porcine model of splenic trauma. Eur. J. Surg. 1999; 165:609–614. 101. Luck EE, Brown DM. Treatments employing drug-containing matrices for introduction into cellular lesion areas. US Patent 4,619,913, Oct. 28, 1986. 102. Borner M, Castiglione M, Triller J, Baer HU, Soucek M, Blumgart L, Brunner K. Considerable side effects of chemoembolization for colorectal carcinoma metastatic to the liver. Ann. Oncol. 1992; 3:113–115. 103. Rosenblatt J, Rhee W, Wallace DG. The effect of collagen fiber size distribution on the release rate of proteins from collagen matrices by diffusion. J. Control Rel. 1989; 9:195–203. 104. Schroeder-Tefft JA, Bentz H, Estridge TD. Collagen and heparin matrices for growth factor delivery. J. Control Rel. 1997; 49:291–298. 105. Bentz H, Schroeder JA, Estridge TD. Improved local delivery of TGF-2 by binding to injectable fibrillar collagen via difunctional polyethylene glycol. J. Biomed. Mater. Res. 1998; 39:539–548.
25 Polyhydroxybutyrate and Its Copolymers: Applications in the Medical Field Fatma Kok and Vasif Hasirci Middle East Technical University Ankara, Turkey
Advances in biotechnology, biomedicine, polymer chemistry, and molecular biology have made available a variety of materials of biological origin for use specifically in medical and other fields. These products have become attractive in recent years, primarily due to their great potential for use in the construction of biodegradable controlled release systems and other temporary implants for in vivo use. The increasing awareness of the environmental pollution arising from the nondegradable products of petroleum origin and chemical industry has also contributed to their demand. Polyhydroxybutyrate (PHB) and its copolymers are members of the polyhydroxyalkanoate (PHA) family and are among the best known polyesters of biological origin. Their biosynthesis has attracted attention due to their unique advantages, such as optical activity, piezoelectricity, biodegradability, versatility, thermoplasticity, biocompatibility, and nontoxicity. PHAs are synthesized by various microorganisms (Rhodospirillum rubdum, Ralstonia eutropha, Pseudomonas oleovorans, Pseudomonas aeruginosa, etc.) under conditions of nutrient limitation (e.g., nitrogen, phosphate, magnesium, sulfate, and oxygen), as intracellular carbon and energy storage compounds [1,2]. The polymer accumulates in discrete, membrane-bound granules in the bacterial cell. Due to their low water solubility and high molecular weight, PHAs do not cause an increase in osmotic pressure, and they are therefore ideal storage compounds. Owing to its natural origin, PHB has an exceptional stereochemical regularity; it has a perfectly isotactic structure with only the (R)-configuration. Although it has several advantages as a biomaterial, its usefulness is limited by its brittleness [3]. Different approaches can be used to improve the properties of PHB.
I. APPROACHES TO THE DESIGN OF PHAs WITH THE DESIRED PROPERTIES A. Bulk Properties In order to change the bulk properties mainly two approaches, copolymerization and blending, are used. 543
544
Kok and Hasirci
Figure 1 General chemical structure of PHAs.
1. Copolymers The addition of different monomers [e.g., hydroxyvalerate (HV)] to the PHB polymer chains can change its physical and thermal properties and improve the ductility and processability of the polymer [2,4]. The PHA family of polyesters offers a wide variety of polymeric materials exhibiting a range of properties, from hard crystalline plastics to elastic rubbers. The PHA materials behave as thermoplastics with melting temperatures of 50–180⬚C. Detailed information regarding thermal properties, crystallinity, and enzymatic degradation of PHAs can be found elsewhere [5]. Random copolymers containing (R)-3HB as a constituent along with other HA units of chain lengths ranging from three to fourteen carbon atoms have been produced from various carbon substrates by a variety of bacteria (Fig. 1, Table 1). Poly(hydroxybutyrate-co-hydroxyvalerate) (PHBV) with varying molar ratios of HV has been the material of extensive, ongoing research for biomedical applications [6]. Being thermoplastic polyesters, PHB and PHBV copolymers can be processed by conventional techniques such as extrusion, injection, or compression molding. PHBV polymers are known to exhibit piezoelectricity [7]. Since electrical stimulation is thought to promote bone healing and repair, polymers of PHA type have been proposed for use as fracture fixation devices, pins, and plates [8]. 2. Blends and Grafts In the literature, blends of PHB with various polymers, namely poly(vinyl phenol) [9], poly(methylene oxide) [10], poly(vinyl alcohol) [11], poly(butylene succinate-co-butylene adipate); and poly(butylene succinate-co-caprolactone) [12], poly(lactide) [13], chitin and chitosan [14], poly(epichlorohydrin) [15], and sucrose acetate isobutyrate (SAIB) [16], have been reported. Blending with other polymers was especially useful to modify the degradation rate of PHAs. Apart from blending, the properties of PHB and its copolymers could be changed by grafting with another polymer. Styrene grafted by radiation, for example, was used to ameliorate the thermal stability of PHB and its copolymers to inhibit their degradation during molding
Table 1 PHAs Produced by Microorganisms Monomer -Hydroxybutyrate (HB) -Hydroxyvalerate (HV) -Hydroxycaproate (HC) -Hydroxyheptanoate (HH) -Hydroxyoctanoate (HO) -Hydroxynonanoate (HN) -Hydroxydecanoate (HD) Source: Ref. 4.
R group CH3 (methyl) CH2CH3 (ethyl) CH2CH2CH3 (n-propyl) CH2(CH2)2CH3 (n-butyl) CH2(CH2)3CH3 (n-pentyl) CH2(CH2)4CH3 (n-hexyl) CH2(CH2)5CH3 (n-heptyl)
Polyhydroxybutyrate and Its Copolymers
545
[17]. A low degree of grafting (9%) retarded thermal degradation; time necessary for 50% weight loss at 220⬚C changed from 76 to 110 min. Preparation of composites of PHAs with hydroxyapatite (HA) is another hot topic since the incorporation of HA improves the bone regenerative properties of PHAs and are thus useful in hard tissue replacement applications [3,18–21]. B. Surface Properties Surface properties of PHAs, on the other hand, can be altered by plasma treatment and ion implantation. Carboxyl ions, for example, can be used to bombard the surface of polyhydroxybutyrate and poly(hydroxybutyrate-co-hydroxyhexanoate) (PHBHHx) films to change the wettability characteristics [22]. Ion implantation resulted in a decrease in the contact angle of both kinds of polymers. For PHB, it gradually decreased from 68.5⬚ as the fluence increased and abruptly decreased to zero at a fluence of 1⳯1015 ions/cm2. For PHBHHx, the contact angle did not decrease as rapidly as for PHB (decreased from 91.5⬚ to 68.5⬚ at the fluence of 1⳯1015 ions/ cm2). Oxygen plasma treatment and attachment of different groups by grafting can also be used to change surface properties [23–25]. II. PRODUCTION Production of PHB and its copolymers is carried out by biofermentation or ring-opening polymerization [26]. For the production of the different copolymers, different carbon substrates can be added together with the main carbon source in biofermentation such as -bromoalkanoic acid [1], valerolactone, and 4-hydroxybutyrate [27,28]. Chemical modifications of the PHAs can also be carried out to yield functional groups affecting the properties of the polymer. Formation of hydroxyl groups, for example, can be useful to produce hydrophilic PHAs [29]. Genetic engineering techniques are also being used to achieve better polymer yields, easier recovery, and production of new copolymers. The use of recombinant Escherichia coli and transgenic plants as well as the other potential PHA production systems was reviewed by Sudesh et al. [5]. A family of PHBV copolymers was first commercialized by Imperial Chemical Industries (ICI, U.K.) under the trade name Biopol威 in the late 1980s. In the early 1990s, Zeneca Ltd. (U.K.) and in 1996, Monsanto Company (St. Louis, MO) acquired Biopol威. Monsanto stopped their research program and commercial Biopol business at the end of 1998 [30]. In 2001, Biopol production was finally transferred to Metabolix, Inc. (Cambridge, MA), which is currently developing transgenic techniques to make PHA production more favorable at a commercial scale. They are now working on optimizing expression of PHA genes in plants and target polymer synthesis to easily processable tissues like the seeds or tubers [31]. Tepha Inc. (Cambridge, MA), related with Metabolix, has recently submitted a Device Master File to the U.S. Food and Drug Administration for its first biomaterial, PHA4400. The approval of this biomaterial would be a milestone in increasing the faith of medical device manufacturers in PHA-based implants. The development of new products with Tepha 4400 including sutures, surgical meshes, nerve and bone regeneration devices, ligament, stents, vascular grafts, heart valves, and sealants are also in progress (www.tepha.com). III. DEGRADATION Many types of microorganisms secreting the extracellular PHA depolymerases are known to degrade PHAs in the biological environment [32–36]. Degradation of PHAs by microorganisms
546
Kok and Hasirci
is affected by the polymer properties. Stereoregularity, crystallinity, molecular mass, and side groups modify degradation patterns and rates [37]. Degradation of PHAs in the body is, however, more important for the medical field. Degradation of PHB and PHBV, the most abundantly used members of the PHA family, was recently reviewed [38]. Degradation studies of cast films of PHB in buffer (37⬚C) indicated that the degradation of the polymer occurred by surface hydrolysis and the degradation rate was increased when the pH of the hydrolytic medium was increased from pH 7.4 to 13 [39]. When in vivo degradation rate of PHB and PHBV was compared with that of polylactides, it was seen that the latter degraded significantly more rapidly (56–99% vs. 15–43% in 6 months) [40]. Among the PHBVs, generally the polymer with the higher valerate content (19 and 22%) had the higher degradation rate. In vitro degradation of PHB is relatively slower compared to PLGA, but the degradation rate of PHAs can be tailored according to need by the addition of different substances in the polymer matrix. Natural polysaccharides such as amylose, dextran, dextrin, and sodium alginate are examples of these substances [41]. Incorporation of these polysaccharides altered the hydrolytic degradation rate (37⬚C, pH 7.4) of the blends extensively. For example, for 20% PHBV/ polysaccharide (10%) blends, the time necessary for 10% weight loss decreased from 600 h to 431, 76, and 44 h in the cases of amylose, dextran and alginate, respectively. The use of low molecular weight additives, such as dodecanol, lauric acid, tributyrin, and trilaurin, were also reported toward that goal [42]. It was observed that a fairly small amount (1 %, w/w) of these additives act as an accelerator for the enzymatic degradation of PHB, while a larger amount (9 %, w/w) acts as a retardant.
IV. BIOCOMPATIBILITY When biocompatibility of an implant is under question, not only the material itself but also its processing conditions are important. The surface characteristics—porosity, degradation rates and products, implant geometry—change according to the method used and therefore highly affect the biocompatibility. Biocompatibility is also closely related with polymer degradation. PHB degrades in vivo to 3-hydroxybutyric acid, a normal constituent of human blood [2] and subcutaneous administration of monomer and polymer (PHB) shows neither toxic nor inflammatory reaction [43]. During degradation of a rapidly biodegradable 3-hydroxybutyric acid polymer that contains crystalline domains of PHB blocks, however, the PHB domains were transformed in a first step into small crystalline particles of short-chain PHB (PHB-P) [44]. Particles of (PHB-P) (Mn 2300) were investigated as possible degradation products. It was shown that phagocytosis of PHB-P at high concentrations (⬎ 10 g/mL) is dose dependent and associated with cell damage in macrophages but not in fibroblasts. Gogolewski et al. [40] studied the tissue response of injection-molded polylactides (PLAs), PHB, and PHBV (5–22% HV content) and saw that all polymers were well tolerated by the tissue. No acute inflammation, abscess formation, or tissue necrosis were observed in tissues adjacent to the implanted materials. In the case of the PHBV, the number of inflammatory cells increased with increasing content of the valerate unit in the polymer chain. For the first 3 months, there was slightly more tissue response to the PHB and PHBV polymers than to PLA, probably because of the presence of leachable impurities and a low molecular weight–soluble component in the poly(hydroxybutyrate-co-hydroxyvalerate). At 6 months, the extent of tissue reaction was similar for both types of polymers. Gursel et al. [45] detected no drainage and only minimal
Polyhydroxybutyrate and Its Copolymers
547
swelling and increase in local warmth after implantation of p(3HB–4HB) rods in rabbit tibia infected with Staphylococcus aureus. The medical use of PHAs necessitates the production of medical grade PHAs. Commercially available PHAs and their purification procedures were reported to be not sufficient for regulatory approval in the medical field [46]. Any residual protein, surfactant, or endotoxin in the polymer inducing immune reaction and fever could lead to wrong conclusions about the biocompatibility of the material. Surface characteristics are of utmost importance, especially in tissue culture applications where it is desirable to promote cell attachment and proliferation. It is also very important because the surface is the region of the implant which is in direct contact with the biological medium. Treatment of the surface with oxygen plasma, with chemicals and enzymes such as NaOH and lipase, appears to improve the biocompatibility of the polymer surface [24,47].
V. MEDICAL APPLICATIONS OF PHAS A. Applications in Bioactive Agent Delivery Hydrolytic degradation of PHAs occurs by surface erosion which makes it an attractive material for controlled release applications [8]. The homopolymer PHB has a relatively high melting point and crystallizes rapidly, making entrapment of drug technically difficult. The related copolymers with 3-hydroxyvalerate, PHBVs, have similar semicrystalline properties though their slower rates of crystallization result in matrices with different properties. Release of low molecular weight drugs from PHB and PHBV structures tends to proceed by penetration of water and pore formation, when loaded with approximately 5% (w/w) drug. Release from such structures is predominantly independent of polymer erosion; though at lower loadings it is possible to trap drug more effectively. PHB and PHBV matrices lose mass very slowly when compared to bulkdegrading poly(lactide-co-glycolide) systems. Therefore, applications of these materials in drug delivery are likely to depend on the formulation of suitable blends with other biocompatible polymers. Porosity, erosion rate, and hence drug release rate can be controlled by blending techniques. At a more fundamental level there is considerable potential for design and bioengineering of other PHAs for applications in drug delivery. For example, medium length PHAs are rubbery materials with low melting points that may be much more suitable as matrices for applications in drug delivery. Various drugs, hormones, and vaccines were reported to be used in bioactive agent delivery studies using PHAs as polymer matrix [2,48–52]. 1. Unmodified PHB- and PHBV-Based Systems Microencapsulation procedures using PHAs are mostly carried out by single or double emulsion techniques. Processing parameters (e.g., polymer precipitation, surfactant, solvent, stirring and solvent evaporation rate) affect the morphology of the produced microparticles [52]. Larger microparticles (100–250 m), for example, were produced at low stirring rates, while very small microparticles (5–10 m) were formed at high stirring rates. It was observed that microparticles obtained by o/w emulsions were monolithic (microspheres) while w/o/w emulsions resulted in microcapsules [51,52]. Kawaguchi et al. [50] prepared microspheres containing prodrugs of 5-fluoro-2′-deoxyuridine with poly(3-hydroxybutyrate) of three molecular weights (65,000, 135,000, and 450,000). The release rates from the spheres depended on both the lipophilicity of the prodrug and the molecular weight of the polymer. The release of prodrugs from the spheres consisting of low molecular weight polymer (65,000 Da) was faster than that from the spheres of higher molecular
548
Kok and Hasirci
weight (135,000 or 450,000 Da). A single intraperitoneal injection of spheres of the highest molecular weight polymer containing prodrug resulted in a higher antitumor effect against P388 leukemia in mice than did free prodrug given over a period of five consecutive days. The polymer sphere itself showed minimal reaction and good biocompatibility with mice and rats. Lu et al. [53] used the in-liquid drying method to prepare sustained release drug delivery microspheres of levo-norgestrol-poly(3-hydroxybutyrate). They examined the appearance, particle size and distribution, residual chloroform, drug content, drug release characteristics in vitro, stability, and anticonceptive effect on mice of the microspheres. Microspheres of average particle size around 64 m was obtained with the residual chloroform lower than 0.001%. The drug release behavior in vitro was described by the Higuchi equation and the drug release half-life was prolonged by 1.8 times, compared with the original drug. The microspheres were stable for 3 months and showed significant sustained release and anticonceptive effect in mice and lower toxicity compared with the original drug. The molecular weight and HV content are also very influential on the preparation of controlled release microspheres. Embleton and Tighe [54] prepared reservoir-type microcapsules from nine different PHB-based polymers in which both molecular weight and hydroxyvalerate content were varied, and they observed that shape and microporosity of the microcapsules were significantly influenced by this variation. Gangrade and Price [49] investigated PHB and PHBV for use as sustained delivery carriers of a model drug, progesterone. Microspheres were prepared by an emulsion/solvent-evaporation method with gelatin as an emulsifier. They found that methylene chloride yielded smoother microspheres than chloroform as the polymer solvent, and the surface texture was dependent upon the temperature of the preparation and polymer used. Surface crystals were observed when the drug loading was increased beyond 5% (w/w). The amount of residual solvent in the microspheres ranged from 3.4 to 58.4 ppm and was dependent on the processing temperature, concentration of the polymer in the solvent, and the polymer composition. In vitro release of the drug was slowest from microspheres made of the copolymer containing 9% HV (less porous microsphere matrix). Tetracycline hydrochloride (TC) – or neutralized TC (TCN)–loaded PHBV microspheres and microcapsules of various valerate contents (7, 14, and 22%) were prepared for the construction of a controlled release system for the treatment of periodontal diseases [51]. TC-loaded microcapsules were prepared by using a double emulsion (w/o/w), solvent evaporation technique described earlier [55]. The method was slightly modified to encapsulate TCN. Drug was introduced into PHBV solution in crystalline form (rather than dissolving in water) forming a single emulsion (o/w) leading to microspheres rather than microcapsules (Fig. 2). It was observed that the increase in PVA (surfactant) and gelatin (second water phase) concentrations caused a bellshaped curve in both encapsulation efficiency and the loading. Maximum efficiency was obtained when 2% gelatin and 1% PVA were used. No loss of drug activity was detected in bioassays. In vitro release studies were carried out in PBS (0.1M, pH 7.4), and it was observed that both the properties of the polymer and the properties of the drug significantly affect the release rate (Fig. 3). Antibiotic [Sulperazone威 (sulbactam-cefoperazone)]-loaded PHBV (22% HV) rods were prepared by Yagmurlu et al. [56], and their in vivo evaluation was carried out in rabbits. Radiological and histological studies showed that symptoms of Staphylococcus aureus infection were substantially improved on day 15 in rod-implanted rabbits, while osteomyelitis was still present in the drug-free controls. A similar study was done with antibiotic-loaded P(3HB–4HB) rods [50% w/w Sulperazone or Duocid威 (ampicillin-sulbactam)] [45]. The rods were implanted in rabbit tibia in which implant-related osteomyelitis (IRO) had been induced with S. aureus. Following the application of Sulperazone-P(3HB–4HB) rods, no infective agents could be isolated from
Polyhydroxybutyrate and Its Copolymers
549
Figure 2 Scanning electron micrograph of a tetracycline-loaded PHBV7 microcapsule.
the infection site within the 6-week test period, indicating complete treatment of the infection. Macroscopic evaluation at follow-up revealed no drainage, minimal swelling, and increase in local warmth, most probably due to the surgery rather than to a reaction toward the implant. In vivo drug release was almost complete within the first week. One interesting observation, however, was that the therapy was still very effective even when the release rate was very high. The morphology of the implant was significantly modified within 6 weeks postimplantation.
Figure 3 Effect of drug type on release from PHBV 7 microcapsules.
550
Kok and Hasirci
Since a substantial degree of the in vivo drug release was complete within 1 week, dissolution of the drug must be the predominant mechanism through which the drug release was controlled. Delivery of high molecular weight compounds like proteins and DNA is important in treatment of various diseases and in gene therapy. Encapsulation of these macromolecules could provide the opportunity for their efficient delivery since this increases the half-life and decrease the antigenic properties. Incorporation of these macromolecules to polymeric systems, however, poses several problems, such as interaction of polymer matrix with the macromolecule (polymer–polymer compatibility) or inability of biological macromolecules to withstand the solvent and temperature used during the process [57]. Encapsulation of model enzymes (L-asparaginase, catalase, glucose oxidase) and bovine serum albumin in PHBV nanocapsules by a double emulsion/solvent-evaporation procedure (w/o/w) is an interesting example for encapsulation of macromolecules [58]. Asparaginase-encapsulated PHBV nanocapsules are presented in Fig. 4. The major challenge in such a case is to increase the encapsulation efficiency and thus the activity of the encapsulated enzyme. It was observed that use of low molecular weight PHBV (297,000 vs. 21,000 Da) did not cause a significant increase in the amount of encapsulated protein, but increased the enzyme activity, observed in the tests probably because of an increase in permeability (reactants and products) that made more enzyme available. The adjustment of the second water phase to the isoelectric point of the proteins increased the encapsulation efficiency for catalase, L-asparaginase, and BSA. In addition, polyethylene glycol (PEG) was coupled with catalase and L-asparaginase to decrease the hydrophilicity of these proteins, thus to increase entrapment efficiencies. The results showed an increase both in the entrapment efficiency and the activity. A combination of optimal conditions led to a six-fold increase in entrapped catalase activity (Fig. 5). 2. Release Systems with PHBV Blends and Coats In order to improve or modify the release profile, different approaches are used. Preparation of blends with different polymers and coating of the drug release system with a polymer are two such methods. Chen and Davis [59], for example, used gelatin coating to alter the release profile of the incorporated drug. They prepared PHBV (630,000 Da, 21% mol HV) microspheres loaded with diazepam using different emulsion/solvent-evaporation processes. The mean diameter of microspheres was 30–40 m. Drug release from the microspheres over a 30-day period showed
Figure 4 Asparaginase-loaded PHBV nanocapsules.
Polyhydroxybutyrate and Its Copolymers
551
Figure 5 Preparation conditions and activity of catalase encapsulated in PHBV nanocapsules.
a characteristic, triphasic release pattern. The initial burst effect was not observed when gelatin was used as a coating agent. Coating can also be achieved by using the same polymer used in release system preparation. In a study where PHBV and P(3HB–4HB) were employed in controlled release of antibiotics (Sulperazone and Duocid), the coating was done with the same polymers [60]. Implantable rods were prepared and in vitro release results showed that drug loading, type of active agent, and coating of the implant all have significant contributions on the release profile. The rate of drug dissolution was found to be substantially higher than that of polymer degradation. Therefore, the release phenomenon was more dependent on the rate of drug dissolution rather than on the rate of polymer degradation or diffusion. In the scanning electron micrographs, it is possible to see the antibiotic crystals before release (Fig. 6a) and the voids formed after release (Fig. 6b). Coating the rods with the same type of polymer substantially reduced the initial burst effect observed with the uncoated rods and significantly decreased the release rate so that the release kinetics became almost zero order (Fig. 7). Antibiotic release from these coated rods could be sustained at a constant rate for over a period of 2 weeks, whereas uncoated rods released their contents in less than a week. Release of anesthetic (bupivacaine) and analgesics (hydromorphone, codeine, and morphine) from PHBV-coated and uncoated interpenetrating networks (IPNs) of PHBV and poly(hydroxyethyl methacrylate) (PHEMA) was studied by Sendil [61]. In vitro release from IPNs with different PHBV/PHEMA ratios showed that the increase in the PHEMA content led to an increase in drug release. Dip-coating of drug-loaded membranes by PHBV altered the release profile of membranes significantly. In the case of morphine, the amount of drug released in 25 h dropped from 80 to 25% upon coating (Fig. 8). Release of macromolecules cannot be achieved by simple diffusion through nonporous capsule membranes. Polymer degradation and diffusion through pores and channels generated by previously dissolving polymer or through well-defined pores created during fabrication could be alternative ways for the release of large molecules. To overcome these difficulties, Atkins and Peacock [62] used a double emulsion technique with solvent evaporation by using BSA as a model protein. They fabricated spherical, microporous, reservoir-type microcapsules composed of P(HB-HV) (10.8% HV/20% polycaprolactone (PCL)) containing BSA-loaded agarose. High
552
Kok and Hasirci
Figure 6 (a) SEM of a cross-section of a paste-molded, uncoated P(3HB-co-3HV) rod before release. Sulperazone (S) crystals are observed at various locations. (b) SEM of a cross-section of a paste-molded uncoated p(3HB-co-3HV) rod after release. The void (V) left by dissolution of Sulperazone is clearly visible.
yield (⬎ 75%, w/w) was achieved in microcapsule preparation, and BSA incorporation had no significant effect on microcapsule size distribution (21–200 m). The loss of BSA both by partitioning into the aqueous continuous phase, and through micropores of BSA-loaded agarose during the precipitation of the fabrication polymer concomitant with solvent evaporation, resulted in a low encapsulation efficiency (12%). PCL eluted at room temperature from hardened microcapsule membrane generating pores. The amount and duration of BSA release was influenced as much by micropore numbers and diameter as by the extent of reservoir loading. Detectable levels of BSA release could be monitored for up to 24 days. Atkins [57] constructed microspheres with a range of BSA loadings using poly(ethylene adipate) (PEAD)/PHBV blends with polycaprolactone by a single emulsion technique with solvent evaporation. Polymer blends of 80:20 PEAD/PCL, 80:20 PHBV/PCL, and 40:40:20 PEAD/ PHBV/PCL were prepared. The major problem was the low encapsulation efficiency of BSA in all systems. Rate of BSA release from PHBV/PEAD/PCL microcapsules was significantly greater than that from PHBV/PCL and PEAD/PCL microcapsules. A gradual increase in release with time was observed in PHBV/PCL microcapsules, whereas in the case of PEAD/PCL microcapsules, there was a 9 to 10-days of lag phase that was followed by a marked increase.
Polyhydroxybutyrate and Its Copolymers
553
Figure 7 Influence of coating on Sulperazone release. Higuchi plot for coated (䊏) and uncoated (⽧) p(3HB-co-3HV) rods.
B. Hard Tissue (Bone) Applications Incorporation of particulate hydroxyapatite (HA) into PHA matrices is usually used in applications of hard tissue replacement and regeneration. PHA and HA could form partially biodegradable composites in which the bioactivity of HA and the biodegradability of PHA are combined. The slow degradation of PHA is an advantage since it avoids particle migration from the implant
Figure 8 Morphine (M) release from coated (䊏) and uncoated (⽧) PHBV/PHEMA interpenetrating networks (ratio 2:1).
554
Kok and Hasirci
region before growth of new tissue, and it can preserve its mechanical properties for long periods. The mechanical properties of a composite material comprising polyhydroxybutyrate with hydroxyapatite added in proportions varying from 0 to 50% was studied by Boeree et al. [63]. Among the three methods of production, injection molding was found to result in the most satisfactory mechanical properties. The most important finding was that the tensile and compressive strength and the modulus of elasticity of composite produced in this way were within the range of fresh human bone from different anatomical sites. With the additional advantages of biocompatibility, biodegradability, and the potential for piezoelectric stimulation of new local bone formation, they concluded that the injection-molded composite material had considerable potential for use in orthopedic surgery, both as a material to construct certain orthopedic implants and as an alternative to corticocancellous bone grafts. Doyle et al. [18] studied the degradation and biological properties of polyhydroxybutyrate and its composites reinforced with particulate hydroxyapatite. They demonstrated that materials based on PHB produce a consistent, favorable bone tissue adaptation response after implantation periods of up to 12 months. Bone was rapidly formed close to the material and subsequently became highly organized, with up to 80% of the implant surface lying in direct apposition to new bone. Similar results were obtained by Luklinska and Bonfield [19]; greater than 90% of bone apposition was detected for PHB pins (with or without HA) at 6 months after implantation. Ni and Wang [21] incorporated HA into polyhydroxybutyrate to form a bioactive and biodegradable composite. HA/PHB composite containing 10, 20, and 30% (v/v) of HA was made for in vitro evaluation in acellular simulated body fluid (SBF). It was seen that a biologically active apatite layer forms within a short period on HA/PHB composite after its immersion in SBF, demonstrating high in vitro bioactivity of the composite. Different HA contents led to different bioactivity and mechanical properties. They concluded that HA/PHB composite had a potential as a bioactive and biodegradable material for applications in hard tissue replacement and regeneration. In another study, some PHAs were characterized by different techniques (GPC, DSC, DMA, FTIR, DRX) for use in hydroxyapatite composites [20]. The composite P(HB-co-8%HV)/ HA (30%, w/w) was found to have the highest modulus. In addition, this composite had a compression strength of 62 MPa, which is about the same order of magnitude of several human bones, and thus it has a potential as a biomaterial for use in fracture fixation. Chen and Wang [3] incorporated particulate HA and tricalcium phosphate into a PHBV copolymer separately to produce new bioactive and biodegradable composites for potential medical applications. HA/PHBV and TCP/PHBV composites contained up to 30% by volume of homogeneously distributed bioceramic particles. They showed that the modulus and microhardness of the composites increased with an increase in the proportion of the bioceramic particles. Another important result was that both series of composites showed enhanced ability to induce the formation of bonelike apatite, indicating bioactivity of the composites. They concluded that both composites have potential for various medical applications due especially to the biodegradability of their constituents. C. Applications in Tissue Culture In tissue engineering, the most important aspect is promotion of tissue formation. The cellcarrying scaffold is expected to provide a highly biocompatible medium to enable cell adhesion, migration, proliferation, and differentiated function. The cells must first attach to the scaffold to start proliferating and developing into a tissuelike structure. Furthermore, the cell–polymer construct must allow gas and nutrient exchange. Finally, degradation of the scaffold with no toxic products is desired once it fulfills its function. There are a number of studies on the
Polyhydroxybutyrate and Its Copolymers
555
suitability of PHAs in tissue engineering. PHAs offer scientists additional benefits in that the polymer properties can be easily tailored by a variety of building blocks and treatments, and a wide range of fabrication techniques can be applied [46]. Primary nerve repair is a challenge for the clinicians. Crushed nerve ends result in either suboptimal repair or the need for nerve grafting. Functional results after nerve surgery are relatively poor, leading to major sensory deficits, which may be due to the death of primary sensory neurons that follows the nerve injury. Silicon nerve conduits have been widely studied and used both experimentally and clinically, but the lack of degradation makes silicon unattractive. Resorbable nerve conduits could be an alternative to primary nerve repair. For that purpose, Ljungberg et al. [64] compared the use of polyhydroxybutyrate conduits by epineurally suturing of the nerve. They observed no statistically significant differences between the two methods, evidence for the future possibility of using PHB as a synthetic nerve graft. Hazari et al. [65] compared the level of regeneration in PHB conduits with nerve autografts in a 10-mm nerve gap of the rat sciatic nerve. Good angiogenesis was detected at the nerve ends and through the walls of the conduit. No failure was observed in any of the implanted conduits. Long reabsorption time of PHB conduits ensured the regeneration and maturation of the nerve, thus enabling the nerve to withstand the stress of mobilization. Transplantation of allogeneic Schwann cells (SC) is another approach to nerve gap reconstruction. For that purpose, Mosahebi et al. [66] used PHB conduits filled with alginate matrix containing SC to bridge a gap of 1 cm in the sciatic nerve. Little fibrotic reaction was observed toward the PHB conduits and there was a significant increase in regeneration distance after 3 weeks in the presence of SC. Another study described a biodegradable implant using polyhydroxybutyrate (PHB) fibers as a carrier scaffold for matrix components and cell lines supporting neuronal survival and regeneration after spinal cord injury [67]. A cervical spinal cord injury was created in adult rats, and then a graft consisting of PHB fibers coated with alginate hydrogel and fibronectin was implanted. The results showed that implants using PHB as a carrier scaffold and containing alginate hydrogel, fibronectin, and Schwann cells could support neuronal survival and regeneration after spinal cord injury. They also found that alginate and/or fibronectin fail to support cell survival in the absence of PHB support. They hypothesized that addition of PHB scaffold to the alginate–fibronectin gel reduces the wash-out of the latter components by the circulating cerebrospinal fluid and thus prolongs their ‘‘lifespan’’ and trophic interactions with the injured neurons. Copolymers of hydroxybutyric acid with different monomers, especially with hydroxyvalerate, are also used in tissue engineering applications. PHBV (9% HV) foams prepared by salt leaching/solvent casting were tested for their potential in fibroblast cell culture [68]. The results were compared with those of collagen foams. They observed that the matrices sustained similar cell growth, while fibroblasts in PHBV foam synthesized almost twice as much total proteins. In addition, the structural integrity of PHBV foams was preserved after 5 weeks, unlike collagen sponges. In another application, the surface of PHBV films was modified by different methods and the interaction of this material with human fibroblast was tested [3]. Oxygen plasma–treated films were grafted by acrylic acid (PHA-C); they were activated with water-soluble carbodiimide; poly(ethylene oxide) (PEO) was attached; and finally insulin was immobilized to the matrices with or without PEO. The researchers observed that human fibroblast cells adhered to PHAand PEO-treated PHA surfaces almost at the same level (ca. 32%). Insulin-attached surfaces gave almost the same result (36%). Cell proliferation, on the other hand, was largely accelerated by insulin presence (ca. 2.5 times).
556
Kok and Hasirci
The use of PHBV was also studied in bone tissue engineering using osteoblast cells [69]. PHBV (8% HV) foams were prepared and their surfaces were treated with oxygen plasma. The evidence of osteoblast proliferation was provided by Cell Titer 96TM Nonradioactivity Cell Proliferation Assay (MTS) and alkaline phosphatase activity. The effect of oxygen plasma treatment was more obvious after 29 days (Fig. 9) [25]. Cell growth was found to favor larger-poresized foams (made by leaching of 300 to 500-m sucrose crystals instead of 75 to 300-m sucrose crystals). A range of retinal diseases and degenerations require retinal pigment epithelium (RPE) or retina transplantation. Construction of an RPE/PHBV graft for correct orientation of RPE cells when implanted subretinally could be a promising alternative for use in the correction of this type of defects [24]. Solvent cast films of PHBV (8% HV) were treated with oxygen plasma at various extents. The optimal attachment of cells were observed with films treated with oxygen plasma at 50 W for 20 min (Fig. 10). Doi et al. [70] reported that poly(hydroxybutyrate-co-hydroxyhexanoate) has much better mechanical properties than PHB and PHBV. Yang et al. [47] evaluated in vitro the biocompatibility of PHB and PHBHHx. The study showed that PHBHHx could support cell growth. In order to improve the hydrophilicity of the polymer films, lipases and NaOH were employed. The improved hydrophilicity of the films allowed easier attachment of the cells on the polymer films compared to that on the untreated ones. Interestingly, the PHBHHx film showed little surface change upon lipase treatment. It was clear that increasing PHBHHx content contributed to increasing growth of viable cells on the PHB and PHBHHx blends. Deng et al. [71] used the same polymer, PHBHHx, for the fabrication of PHBHHx/PHB three-dimensional porous scaffolds by the salt-leaching method and investigated these scaffold systems for possible application as a matrix for the three-dimensional growth of chondrocytes.
Figure 9 Alkaline phosphatase (ALPase) activity of osteoblasts cultured inside the PHBV foams (6%, w/w) foams for 7, 14, 21, and 29 days: (shaded) untreated PHBV; (unshaded) rf-oxygen plasma treated (100 W, 10 min) PHBV.
Polyhydroxybutyrate and Its Copolymers
557
Figure 10 Phase contrast micrograph of rat retinal pigment epithelial cells on oxygen plasma–treated PHBV8 film after 24 h incubation (20⳯).
Chondrocytes were seeded on the scaffolds and incubated over 28 days. Results showed that chondrocytes proliferated better on the PHBHHx/PHB scaffolds than on PHB (control) ones. In addition, cells grew better on scaffolds consisting of PHBHHx/PHB in ratios of 2:1 and 1: 2 than they did at a PHBHHx/PHB ratio of 1:1. It was also seen that large quantities of chondrocytes grew initially on the surface of the scaffold and after 7 days they further spread into the open pores. In addition, chondrocytes proliferated on the scaffold and preserved their phenotype for up to 28 days. The superiority of blends over PHB was explained by differences in crystallinity; PHB crystallizes to form the crystalline domains that act as physical crosslinkers and fillers, and the PHBHHx forms the amorphous domain. It was already demonstrated that the arrangement of crystalline domains and amorphous domains could affect the evaporation process of the solvent, chloroform, resulting in new plastic properties which affect the oxygen permeability of the scaffold [72]. They claimed that this could be the possible explanation of the better compatibility of cells and the blend scaffolds of PHBHHx/PHB. An interesting application was the creation of a tissue engineered heart valve [73] where an elastomer of PHA origin was used (Metabolix, Inc.). To create a porous, three-dimensional structure suitable for tissue engineering, the salt leaching technique was used. Fabricated scaffolds were seeded with vascular cells and tested in a pulsatile flow bioreactor. One advantage of that material was said to be the ability to mold a complete trileaflet heart valve scaffold without the need for suturing leaflets to the conduit. A second advantage was the use of only one polymer material (PHA) as opposed to hybridized polymer scaffolds. Furthermore, the mechanical properties of PHA, such as elasticity and mechanical strength, exceeded those of the previously utilized materials. Dacron and polytetrafluoroethylene (PTFE) manufactured in various forms are commercially available and clinically used in patients without alternative vascular conduits. While these vascular substitutes are satisfactory in large-diameter applications, they perform poorly in smaller caliber vessel replacement. Shum-Tim et al. [74] evaluated whether a new PGA-PHA copolymer, which had a much longer degradation time, could withstand systemic pressure and be used to create a vascular autograft for use in the aortic position. The scaffold was formed using a similar PGA inner layer, which was shown to promote cell attachment and tissue formation, with an
558
Kok and Hasirci
outer, nonporous polyhydroxyoctanoate (PHO) layer. Autologous ovine carotid arteries were seeded onto 7-mm-diameter tubular scaffolds. Constructs were used to replace 3 to 4 cm abdominal aortic segments in lambs. The inner layer was made of randomly arrayed fibers of nonwoven PGA mesh. Ten days after implantation, the tissue engineered aortic graft showed evidence of tissue formation with both PHA-PGA still intact. At 3 and 5 months postoperatively, the inner PGA layer was completely replaced by engineered neoaortic tissue. The mechanical characteristics tended to resemble and further approached the native aorta over time. D. Other Applications Malm et al. [75] prepared an absorbable pericardial patch from polyhydroxybutyrate and investigated its potential as a temporary scaffold for regeneration of pericardial tissue by focusing on characterization of the regenerated surface cells. Their results indicated regeneration of a mesothelial layer with many of the important functions of native mesothelial cells. Clinical testing of PHB patches as pericardial substitutes was warranted in cardiac surgery when pericardial closure is desired. Freier et al. [76] developed a resorbable gastrointestinal patch from poly(3-hydroxybutyrate). Solution-cast PHB films were fabricated by dipping a metal core into a PHB/chloroform solution (film thickness up to ca. 100 m). To obtain a porous surface, metal core was dipped into the solution, unsieved NaCl (crystal size 300–500 m in length tenfold higher than PHB by weight) was added to the polymer/chloroform solution; and then the dipping process was repeated. To accelerate the relatively slow hydrolysis of PHB, atactic PHB was added. In vitro studies in the presence of pancreatin indicated a participation of enzymes on PHB hydrolysis (threefold increase in comparison to simple hydrolysis). However, they claimed that enzymatic catalysis is questionable due to the observed degradation behavior with molecular weight decrease but constant mass. It was thought that secondary effects caused by enhanced enzymatic activity could be involved. Both in vitro studies and in vivo experiments in rats for repair of bowel defects showed that PHB/atactic PHB was the most promising material combination. The material was sufficiently flexible to adapt and to be sutured over the defect. Its degradation characteristics were such that the temporary patch material resists the intestinal secretions and closes the bowel defect for a sufficiently long time to allow complete healing of the defect.
VI. CONCLUSION The extensive research on PHB and its copolymers shows the significant potential of these biodegradable and biocompatible materials for use in the medical field. These polyesters can be processed in various ways, and their properties could be tailor-made to the application. Introduction of several medical products is expected once the use of PHAs in the medical field is approved by the health and regulatory authorities.
REFERENCES
1. Kim YB, Lenz RW, Fuller RC. Poly(-hydroxyalkanoate) copolymers containing brominated repeating units produced by Pseudomonas oleovorans. Macromolecules 1992; 25(7):1852–1857. 2. Nobes GAR, Marchessault RH, Maysinger D. Polyhydroxyalkanoates: materials for delivery systems. Drug Delivery 1998; 5:167–177.
Polyhydroxybutyrate and Its Copolymers
559
3. Chen LJ, Wang M. Production and evaluation of biodegradable composites based on PHB-PHV copolymer. Biomaterials 2002; 23(13):2631–2639. 4. Doi Y. Microbial Polyesters. New York: VCH Publishers, 1990. 5. Sudesh K, Abe H, Doi Y. Synthesis, structure and properties of polyhydroxyalkanoates: biological polyesters. Prog Polym Sci 2000; 25(10):1503–1555. 6. Avella M, Martuscelli E, Raimo M. Properties of blends and composites based on poly(3-hydroxybutyrate) (PHB) and poly(3-hydroxybutyrate–hydroxyvalerate) (PHBV) copolymers. J Mater Sci 2000; 35:523–545. 7. Fukada E, Ando Y. Piezoelectric properties of poly--hydroxybutyrate and copolymers of -hydroxybutyrate and -hydroxyvalerate. Int J Biol Macromol 1986; 8:361–366. 8. Pouton CW, Akhtar S. Biosynthetic polyhydroxyalkanoates and their potential in drug delivery. Adv Drug Delivery Rev 1996; 18(2):133–162. 9. Iriondo P, Iruin JJ, Fernandez-Berridi MJ. Thermal and infra-red spectroscopic investigations of a miscible blend composed of poly(vinyl phenol) and poly(hydroxybutyrate). Polymer 1995; 36(16): 3235–3237. 10. Avella M, Martuscelli E, Orsello G, Raimo M, Pascucci B. Poly(3-hydroxybutyrate)/poly(methyleneoxide) blends: thermal, crystallization and mechanical behaviour. Polymer 1997; 38(25):6135–6143. 11. Ikejima T, Yoshie N, Inoue Y. Influence of tacticity and molecular weight of poly(vinyl alcohol) on crystallization and biodegradation of poly(3-hydroxybutyric acid)/poly(vinyl alcohol) blend films. Polym Degrad Stab 1999; 66(2):263–270. 12. He Y, Asakawa N, Masuda T, Cao A, Yoshie N, Inoue Y. The miscibility and biodegradability of poly(3-hydroxybutyrate) blends with poly(butylene succinate-co-butylene adipate) and poly(butylene succinate-co-caprolactone). Eur Polym J 2000; 36(10):2221–2229. 13. Ohkoshi I, Abe H, Doi Y. Miscibility and solid-state structures for blends of poly[(S)-lactide] with atactic poly[(R,S)-3-hydroxybutyrate]. Polymer 2000; 41(15):5985–5992. 14. Ikejima T, Inoue Y. Crystallization behavior and environmental biodegradibility of the blend films of poly(3-hydroxybutyric acid) with chitin and chitosan. Carbohydr Polym 2000; 41:351–356. 15. Miguel O, Egiburu JL, Iruin JJ. Blends of bacterial poly(3-hydroxybutyrate) with synthetic poly(3hydroxybutyrate) and poly(epichlorohydrin): transport properties of carbon dioxide and water vapour. Polymer 2001; 42(3):953–962. 16. Tezcaner A. Tissue engineered retinal pigment epithelium, PhD Thesis, Ankara, Turkey, 2001. 17. Bahari K, Mitomo H, Enjoji T, Yoshii F, Makuuchi K. Degradability of poly(3-hydroxybutyrate) and its copolymer grafted with styrene by radiation. Polym Degrad Stab 1998; 61(2):245–252. 18. Doyle C, Tanner ET, Bonfield W. In vitro and in vivo evaluation of polyhydroxybutyrate and of polyhydroxybutyrate reinforced with hydroxyapatite. Biomaterials 1991; 12(9):841–847. 19. Luklinska ZB, Bonfield W. Morphology and ultrastructure of the interface between hydroxyapatite–polyhydroxybutyrate composite implant and bone. J Mater Sci: Mater Med 1997; 8:379–383. 20. Galego N, Rozsa C, Sa´nchez R, Fung J, Va´zquez A, Toma´s JS. Characterization and application of poly(hydroxyalkanoates) family as composite biomaterials. Polym Test 2000; 19(5):485–492. 21. Ni J, Wang M. In vitro evaluation of hydroxyapatite reinforced polyhydroxybutyrate composite. Mater Sci Eng, C 2002; 20(1–2):101–109. 22. Zhang DM, Cui FZ, Luo ZS, Lin YB, Zhao K, Chen GQ. Wettability improvement of bacterial polyhydroxyalkanoates via ion implantation. Surf Coat Techno l 2000; 131(1–3):350–354. 23. Kang I-K, Choi S-H, Shin D-S, Yoon SC. Surface modification of polyhydroxyalkanoate films and their interaction with human fibroblasts. Int J Biol Macromol 2001; 28:205–212. 24. Tezcaner A, Hasirci V. Soft tissue engineering with ophthalmological applications, In: Lewandrowski K, Wise DL, Trantolo DJ, Gresser JD, Yaszemski MJ, Altobelli DE, eds. Tissue Engineering and Biodegradable Equivalents: Scientific and Clinical Applications. New York: Marcel Dekker, 2002: 545–560. 25. Kose GT, Ber S, Korkusuz F, Hasirci V. (In press). Poly(3-hydroxybutyric acid-co-3-hydroxyvaleric acid) (PHBV) based tissue engineering matrices. J Mater Sci Polym Ed 2003; 14:121–126. 26. Jedlinski Z, Kowalczuk M, Adamus G, Sikorska W, Rydz J. Novel synthesis of functionalized poly(3hydroxybutanoic acid) and its copolymers. Int J Biol Macromol 1999; 25:247–253.
560
Kok and Hasirci
27. Hasirci V. The production and application of microbial polyesters in the construction of biodegradable controlled antibiotic release systems and bone fracture repair rods (TU polyesters). NATO Scientific and Environmental Affairs Division, Science for Stability Project Final Report, 1998. 28. Turesin F, Gumusyazici Z, Kok FN, Gursel I, Alaeddinoglu G, Hasirci V. Biosynthesis of polyhydroxybutyrate and its copolymers and their use in controlled drug release. Turk J Med Sci 2000; 30: 535–541. 29. Lee MY, Park WH, Lenz RW. Hydrophilic bacterial polyesters modified with pendant hydroxyl groups. Polymer 2000; 41:1703–1709. 30. Asrar J, Gruys KJ. Biodegradable polymer (Biopol威) In: Steinbu¨chel A, Doi Y, eds. Biopolymers;. Vol. 3: Wiley-VCH, 2002:53–87. 31. Snell KD, Peoples OP. Polyhydroxyalkanoate polymers and their production in transgenic plants. Metab Eng 2002; 4(1):29–40. 32. Mergaert J, Webb A, Anderson C, Wouters A, Swings J. Microbial degradation of poly(3-hydroxybutyrate) and poly(3-hydroxybutyrate-co-hydroxyvalerate) in soils. J Appl Environ Microbiol 1993; 59:3233–3238. 33. Kasuya K, Doi Y, Yao T. Enzymatic degradation of poly[(R)-3-hydroxybutyrate] by Comamonas testosteroni ATSU of soil bacterium. Polym Degrad Stab 1994; 45:379–386. 34. Oda Y, Asari H, Urakami T, Tonomura K. Microbial degradation of poly(3-hydroxybutyrate) and polycaprolactone by filamentous fungi. J Ferment Bioeng 1995; 80(39):265–269. 35. Abou-Zeid D-M, Mu¨ller R-J, Deckwer W-D. Degradation of natural and synthetic polyesters under anaerobic conditions. J. Biotechnol 2001; 86(2):113–126. 36. Mabrouk MM, Sabry SA. Degradation of poly(3-hydroxybutyrate) and its copolymer poly(3-hydroxybutyrate-co-3-hydroxyvalerate) by a marine Streptomyces sp. SNG9. Microbiol Res 2001; 156(4): 323–335. 37. Jendrossek D, Schirmer A, Schlegel HG. Biodegradation of polyhydroxyalkanoic acids. Appl Microbiol Biotechnol 1996; 46:451–463. 38. Hasirci V. Biodegradable biomedical polymers: review of degradation of and in vivo responses to polylactides and polyhydroxyalkanoates. In: Wise DL, ed. Biomaterials and Bioengineering Handbook. New York: Marcel Dekker, 2000:141–155. 39. Majid MIA, Ismail J, Few LL, Tan CF. The degradation kinetics of poly(3-hydroxybutyrate) under non-aqueous and aqueous conditions. Eur Polym J 2002; 38(4):837–839. 40. Gogolewski S, Jovanovic M, Perren SM, Dillon JG, Hughes MK. Tissue response and in vivo degradation of selected polyhydroxyacids: polylactides (PLA), poly(3-hydroxybutyrate) (PHB), and poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHB/VA). J Biomed Mater Res 1993; 27(9): 1135–1148. 41. Holland SJ, Yasin M, Tighe BJ. Polymers for biodegradable medical devices. VII. Hydroxybutyrate–hydroxyvalerate copolymers: degradation of copolymers and their blends with polysaccharides under in vitro physiological conditions. Biomaterials 1990; 11(3): 206–215. 42. Yoshie N, Nakasato K, Fujiwara M, Kasuya K, Abe H, Doi Y, Inoue Y. Effect of low molecular weight additives on enzymatic degradation of poly(3-hydroxybutyrate). Polymer 2000; 41(9):3227–3234. 43. Koosha F, Muller RH, Davis SS. Polyhydroxybutyrate as a drug carrier. Crit Rev Ther Drug Carrier Syst 1989; 6:117–130. 44. Saad B, Ciardelli G, Matter S, Welti M, Uhlschmid GK, Neuenschwander P, Suter UW. Characterization of the cell response of cultured macrophages and fibroblasts to particles of short-chain poly[(R)3-hydroxybutyric acid]. J Biomed Mater Res 1996; 30(4):429–439. 45. Gursel I, Korkusuz F, Turesin F, Alaeddinoglu NG, Hasirci V. In vivo application of biodegradable controlled antibiotic release systems for the treatment of implant-related osteomyelitis. Biomaterials 2001; 22(1):73–80. 46. Williams SF, Martin DP, Horowitz DM, Peoples OP. PHA applications: addressing the price performance issue. I. Tissue engineering. Int J Biol Macromol 1999; 25(1–3):111–121. 47. Yang X, Zhao K, Chen G-Q. Effect of surface treatment on the biocompatibility of microbial polyhydroxyalkanoates. Biomaterials 2002; 23(5):1391–1397. 48. Kubota M, Nakano M, Juni K. Mechanism of enhancement of the release rate of aclarubicin from poly-hydroxybutyric acid microspheres by fatty acid esters. Chem Pharm Bull 1988; 36(1):333–337.
Polyhydroxybutyrate and Its Copolymers
561
49. Gangrade N, Price JC. Poly(hydroxybutyrate–hydroxyvalerate) microspheres containing progesterone: preparation, morphology and release properties. J Microencapsulation 1991; 8(2):185–202. 50. Kawaguchi T, Tsugane A, Higashide K, Endoh H, Hasegawa T, Kanno H, Seki T, Juni K, Fukushima S, Nakano M. Control of drug release with a combination of prodrug and polymer matrix: antitumor activity and release profiles of 2′,3′-diacyl-5-fluoro-2′-deoxyuridine from poly(3-hydroxybutyrate) microspheres. J Pharm Sci 1992; 81(6):508–512. 51. Sendil D, Gursel I, Wise DL, Hasirci V. Antibiotic release from biodegradable PHBV microparticles. J Control Rel 1999; 59:207–217. 52. Martin MA, Miguens FC, Rieumont J, Sanchez R. Tailoring of the external and internal morphology of poly-3-hydroxybutyrate microparticles. Colloids Surf B: Biointerfaces 2000; 17:111–116. 53. Lu B, Wang ZR, Yang H. Long-acting delivery microspheres of levo-norgestrol-poly(3-hydroxybutyrate): their preparation, characterization and contraceptive tests on mice. J Microencapsulation 2001; 18(1):55–64. 54. Embleton JK, Tighe BJ. Polymers for biodegradable medical devices. IX: Microencapsulation studies; effects of polymer composition and process parameters on poly-hydroxybutyrate–hydroxyvalerate microcapsule morphology. J Microencapsulation 1992; 9(1):73–87. 55. Gursel I, Hasirci V. Properties and drug release behavior of poly(3-hydroxybutyric acid) and various poly(3-hydroxybutyrate–hydroxyvalarate) copolymer microcapsules. J Microencapsulation 1995; 12(2):185–193. 56. Yagmurlu MF, Korkusuz F, Gu¨rsel I, Korkusuz P, Ors U, Hasirci V. Sulbactam-cefoperazone polyhydroxybutyrate-co-hydroxyvalerate (PHBV) local antibiotic delivery system: in vivo effectiveness and biocompatibility in the treatment of implant-related experimental osteomyelitis. J Biomed Mater Res 1999; 46(4):494–503. 57. Atkins TW. Fabrication of microcapsules using poly(ethylene adipate) and a blend of poly(ethylene adipate) with poly(hydroxybutyrate–hydroxyvalerate): incorporation and release of bovine serum albumin. Biomaterials 1996; 18(2):173–180. 58. Baran ET, Ozer N, Hasirci V. Poly(hydroxybutyrate-co-hydroxyvalerate) nanocapsules as enzyme carriers for cancer therapy: an in vitro study. J Microencapsulation 2002; 19(3):363–376. 59. Chen J, Davis SS. The release of diazepam from poly(hydroxybutyrate–hydroxyvalerate) microspheres. J Microencapsulation 2002; 19(2):191–201. 60. Turesin F, Gursel I, Hasirci V. Biodegradable polyhydroxyalkanoate implants for osteomyelitis therapy: in vitro antibiotic release. J Biomater Sci Polym Ed 2001; 12(2):195–207. 61. Sendil D. Controlled release of pain relievers for the alleviation of chronic pain, Ankara, Turkey, 2001. 62. Atkins TW, Peacock J. The incorporation and release of bovine serum albumin from poly-hydroxybutyrate–hydroxyvalerate microcapsules. J Microencapsulation 1996; 13(6):709–717. 63. Boeree NR, Dove J, Cooper JJ, Knowles J, Hastings GW. Development of a degradable composite for orthopaedic use: mechanical evaluation of an hydroxyapatite–polyhydroxybutyrate composite material. Biomaterials 1993; 14(10):793–796. 64. Ljungberg C, Johansson-Ruden G, Bostro¨m KJ, Novikov L, Wiberg M. Neuronal survival using a resorbable synthetic conduit as an alternative to primary nerve repair. Microsurgery 1999; 19(6): 259–264. 65. Hazari A, Wiberg M, Johansson-Rude´n G, Green C, Terenghi G. A resorbable nerve conduit as an alternative to nerve autograft in nerve gap repair. Br J Plast Surg 1999; 52(8):653–657. 66. Mosahebi A, Fuller P, Wiberg M, Terenghi G. Effect of allogeneic Schwann cell transplantation on peripheral nerve regeneration. Exp Neurology 2002; 173(2):213–223. 67. Novikov LN, Novikova LN, Mosahebi A, Wiberg M, Terenghi G, Kellerth J-O. A novel biodegradable implant for neuronal rescue and regeneration after spinal cord injury. Biomaterials 2002; 23(16): 3369–3376. 68. Rivard CH, Chaput CJ, DesRosiers EA, Yahia LH, Selmani A. Fibroblast seeding and culture in biodegradable porous substrates. J Appl Biomater 1995; 6:65–68. 69. Kose GT, Hasirci V. Polymeric biodegradable hard tissue engineering applications, In: Lewandrowski K, Wise DL, Trantolo DJ, Gresser JD, Yaszemski MJ, Altobelli DE, eds. Tissue Engineering and
562
70. 71.
72. 73.
74.
75.
76.
Kok and Hasirci Biodegradable Equivalents: Scientific and Clinical Applications. New York: Marcel Dekker, 2002: 301–316. Doi Y, Kitamura S, Abe H. Microbial synthesis and characterization of poly(3-hydroxybutyrate-co3-hydroxyhexanoate). Macromolecules 1995; 28:4822–4828. Deng Y, Zhao K, Zhang X-F, Hu P, Chen G-Q. Study on the three-dimensional proliferation of rabbit articular cartilage–derived chondrocytes on polyhydroxyalkanoate scaffolds. Biomaterials 2002; 23(20):4049–4056. Zinn M, Witholt B, Egli T. Occurrence, synthesis and medical application of bacterial polyhydroxyalkanoate. Adv Drug Delivery Rev 2001; 53:5–51. Sodian R, Sperling JS, Martin DP, Egozy A, Stock U, Mayer JE, Vacanti JP. Fabrication of a trileaflet heart valve scaffold from a polyhydroxyalkanoate biopolyester for use in tissue engineering. Tiss Eng 2000; 6(2):183–188. Shum-Tim D, Stock U, Hrkach J, Shinoka T, Lien J, Moses MA, Stamp A, Taylor G, Landis W, Langer R, Vacanti JP, Mayer JE. Tissue engineering of autologous aorta using a new biodegradable polymer. Ann Thorac Surg 1999; 68(6):2298–2304. Malm T, Bowald S, Bylock A, Saldeen T, Busch C. Regeneration of pericardial tissue on absorbable polymer patches implanted into the pericardial sac. An immunohistochemical, ultrastructural and biochemical study in the sheep. Scand J Thoracic Cardiovasc Surg 1992; 26(1):15–21. Freier T, Kunze C, Nischan C, Kramer S, Sternberg K, Sa M, Hopt UT, Schmitz K-P. In vitro and in vivo degradation studies for development of a biodegradable patch based on poly(3-hydroxybutyrate). Biomaterials 2002; 23(13):2649–2657.
26 From Polymer Chemistry and Physicochemistry to Nanoparticulate Drug Carrier Design and Applications C. Vauthier, E. Fattal, and D. Labarre University of Paris South Chatenay-Malabry, France
I. INTRODUCTION Design and manufacture of drug carrier systems has been a challenging area of research over the last 30 years which has led to a choice of suitable colloidal systems, including liposomes [1–3], nanoparticles [4–11], and block copolymer micelles [12,13]. The main advantage of using such colloidal systems is given by their submicronic size, allowing them to be administered intravenously without any risk of embolization and promoting their diffusion through capillary vessels [14] and even through mucosae [15]. This particle size range is also desirable for subcutaneous and intramuscular administration [16] and offers new functions such as large specific surface areas, considerable surface properties, improved solubility, and adhesion to tissues [17]. This chapter focuses on the technology of nanoparticles which are made of polymers of synthetic or natural origin. Nanoparticle is a collective name for nanospheres and nanocapsules. Nanospheres have a matrix type structure, whereas nanocapsules have a liquid core surrounded by a polymeric shell (Fig. 1). The active substances are usually dissolved in the core of the nanocapsules and dispersed in the matrix of the nanospheres. Occasionally the drug may also be adsorbed on the surface of the nanoparticles [4]. Historically, the first nanoparticles were proposed by Birrenbach and Speiser [18]. They were first developed as drug delivery systems at the time when liposomal formulations were still lacking stability and suffered from relatively poor drug loading capacity [19]. Since then the technology for making such polymeric nanoparticles has been expanded to adapt for the specific requirements for a drug carrier, including biocompatibility, compatibility with the type of drug to be carried, drug loading efficacy, defined drug releasing properties, and in vivo targeting [8]. Initial expectations were dampened by the fact that following intravenous administration, nanoparticles were quickly removed from the circulation by macrophages located in the organs of the mononuclear phagocytic system, thus hindering site-specific delivery of drugs to other organs or tissues in the body. In an attempt to reduce or to minimize particle interaction with opsonins, which facilitate this phagocytic process, the concept of steric stabilization of particles was introduced [20]. To this end, nanoparticles with prolonged blood circulation times have been designed. 563
564
Vauthier et al.
Figure 1 Schematic representation of the different types of nanoparticles that have been produced.
The purpose of this chapter is to provide the reader with a broad outline of nanoparticle technologies starting from polymer chemistry and polymer physicochemistry and thus providing an overview of the most relevant potential applications to date.
II. PREPARATION OF NANOPARTICLES The formulation and application of polymeric nanoparticles dispersed in a nonsolvent enjoys great popularity in academia and industry for various purposes. Emulsion polymerization is one of the most frequently employed methods of producing nanoparticles [21,22]. Early workers used a radical emulsion polymerization process to produce the first nanoparticles with a view to pharmaceutical applications [18]. However, to develop further the pharmaceutical applications of nanoparticles, these polymerization technologies were adapted to produce suitable nanoparticles, for instance, made of biodegradable polymers dispersed in an aqueous medium and with a limited amount of surfactant to achieve their stability. Alkylcyanoacrylates appeared to be optimal monomers for the development of nanoparticles by emulsion polymerization for human use [9,10,23]. Methods of anionic polymerizations in emulsion have been specifically designed for thoses monomers [10]. Alternative techniques were developed in parallel from preformed polymers to avoid polymerization reactions. The main advantage of these techniques is that the polymers entering the composition of the nanoparticles are well characterized and their intrinsic physicochemical characteristics would not depend on the conditions encountered during the preparation of the nanoparticles. Progress in nanoparticle technologies to circumvent the rapid clearance of nanoparticles from the systemic circulation has been made possible by new developments in polymer chemistry with the synthesis of biodegradable block copolymers [20,24–26]. These copolymers were mainly designed to prepare surface-modified nanoparticles by methods using preformed polymers.
Nanoparticulate Drug Carriers
565
Techniques for preparing the nanoparticles described in this chapter fall into two groups of methods: those based on polymerization reactions and those based on the use of preformed polymers (Fig. 2). This corresponds to the usual classification proposed by different authors [4,8,9,16,27]. However, the presentation will differ sligthly in the sense that it will consider the input of polymer chemistry for the methods based on polymerization reactions on one hand and, on the other hand, the input of polymer physicochemistry for the methods based on the use of preformed polymers. This new presentation has the advantage of introducing a new category devoted to the design of block copolymers which are now an active area of research promoting further progress for medical applications of nanoparticle technologies. A. Polymer Chemistry and Design of Nanoparticulate Drug Carriers 1. Radical Emulsion Polymerization The preparation of nanoparticulate systems through radical polymerization carried out in oilin-water emulsions has been developed for many years. The resulting colloidal system, i.e., a latex, can be either coagulated according to processes developed in the rubber industry or used as such, for instance, in the paint industry. In spite of their low to very low solubility into water, a large number of monomers, e.g., acrylic, vinylic, or dienic, can be polymerized in an aqueous emulsion. The whole mechanism is rather complex; the reaction is initiated in the aqueous phase by a water-soluble initiator, whereas at the beginning of the reaction the monomer is mainly present in large droplets acting as reservoirs, the size of these droplets depending on the rate of stirring [16,28]. Surfactants are usually added, resulting in formation of micelles which contain a part of the monomer. Polymerization takes place in nucleated micelles which grow progressively, resulting from displacement of the equilibrium between a pure monomer in droplets, a monomer in micelles or in aqueous solution at a very low concentration, and a monomer reacting
Figure 2 Summary of the different methods developed for the preparation of nanoparticles by polymerization and from polymers with corresponding types of nanoparticle produced.
566
Vauthier et al.
with the growing chains in the nucleated micelles. At the end of polymerization, the polymeric micelles swollen by the monomer and stabilized by the surfactants give nanoparticles, on which the surfactants remain adsorbed. The size of the particles depends on the experimental conditions and especially on the type and concentration of the surfactants. Emulsion polymerizations can also take place in the absence of added surfactants, providing that the initiator can ensure the growing polymeric chains with surfactant properties able to stabilize the particles. This is typically the case when charged water-soluble initiators such as persulfates are used. Thermal decomposition of the initiator results in formation of sulfate ion radicals SO4ⳮ● which initiate polymerization and remain bound as sulfate groups at least at one end of the polymeric chains. The resulting polymeric micelles bearing sulfate groups on their surface cannot coagulate, and a stable latex composed of almost monodispersed nanoparticles can be obtained. In the biomedical field, colloidal systems have been proposed as supporting surfaces for diagnostic tests, thanks to their high specific surface area, high stability, and ease of handling [29,30]. In the pharmaceutical field, nanoparticles made of poly(acrylamide) or poly(methyl methacrylate) were proposed during the 1970s as adjuvants for vaccines [18,31]. However, such polymers are not biodegradable even after months and are not acceptable for repeated human administration. Therefore, nanoparticles made of biodegradable polymers are preferred. It should be noted that monomers able to polymerize through a radical mechanism, i.e., containing one or two carbon–carbon double bonds, lead to polymers possessing carbon–carbon single bonds which are not usually biodegradable in vivo. But after erosion of the lateral groups leading to water-soluble polymers, and providing that their molecular weight is within the limits of the filtering capacity of the kidneys, some of them could be eliminated and be possible candidates for repeated human administration. Today the only monomers that could fulfill these requirements are alkylcyanoacrylates and their derivatives (Fig. 3A). However, these monomers are highly reactive compounds, and their anionic polymerization can be spontaneously and quickly initiated by even small amounts of a weak base such as water. Therefore, the radical polymerization of alkylcyanoacrylate monomers has hardly ever been described even in bulk conditions, i.e., pure monomer. 2. Anionic Emulsion Polymerization of Alkylcyanoacrylate Anionic emulsion polymerization of alkylcyanoacrylates was introduced by Couvreur et al. in 1979 [23] to design biodegradable polymer particles suitable for in vivo delivery of drugs. The nanoparticles prepared according to this technology are actually nanospheres. Interfacial polymerization methods were specifically designed to produce nanocapsules. The anionic polymerization of alkylcyanoacrylate was initiated in unusual conditions for such a polymerization mechanism by the hydroxyl ions of the water in media containing a large excess of water, i.e., usually more than 90% (Fig. 3B). Emulsions formulated to prepare poly(alkylcyanoacrylate) nanospheres to be used as drug carriers are usually very complex. For example, the monomer (100 L) is dispersed in acidified water containing a surfactant or a stabilizing agent (10 mL of a 0.5 to 1% solution of Pluronic威 F68 or dextran 70 at pH 2.5 with HCl) and the drug. This system is left to polymerize spontaneously for a few hours (3 to 4 h). The resulting colloidal polymer particles have a diameter ranging from 50 to 300 nm. At pH higher than 3, the polymerization is too fast and leads to polymer aggregates. In contrast with other monomers, it is noteworthy to point out that the emulsion polymerization of alkylcyanoacrylates could be initiated at a pH lower than 1 in the presence of strong acids, inhibiting the polymerization when it is carried out in an organic solution. This is due to the presence of additives such as surfactants or stabilizing agents which
Nanoparticulate Drug Carriers
567
Figure 3 Chemical structure of alkylcyanoacrylates and methylidenemalonates (A) and scheme of the anionic polymerization of alkylcyanoacrylate (B).
are dissolved in the polymerization medium and have nucleophilic functions able to initiate the polymerization of alkylcyanoacrylates [32–36]. The molecular weight of the polymers forming the nanospheres is usually low. It is affected by the pH of the polymerization medium, by the presence or the absence of surface-active agents, and by the presence or the absence of a drug [37–40]. The size of the nanospheres formed can be controlled by the amount of surface-active agent or by the molecular weight of a colloidal stabilizer like dextran [33,41,42] as well as by both the pH and the sulfur dioxide concentration, which affect the polymerization process [37,43,44]. Poly(alkylcyanoacrylate) nanospheres are stable as a suspension but can also be stored as a lyophilized powder [45,46]. In order to prepare sterile suspensions for clinical use, it is recommended to produce the nanospheres under aseptic conditions [47]. Many drugs have been entrapped with success in poly(alkylcyanoacrylate) nanospheres [10]. However, a few drugs were shown to initiate the polymerization reaction and lost their biological activity [38–40]. Cyclodextrines were found to be useful in improving the association of poorly water-soluble lipidic compounds with poly(alkylcyanoacrylate) nanospheres when they were added to the polymerization medium [48–50]. Side reactions occurring during the polymerization can advantageously be used to associate compounds by a covalent binding with nanospheres. This has been applied to naphthalocyanines, a photosensibilizer used in phototherapy of tumors [51] and to a series of molecules containing diethyltriaminepentacetic acid
568
Vauthier et al.
(DTPA), capable of complexing radioactive metals for radiolabelling of nanoparticles in medical imaging [52]. These reactions were also used to produce nanoparticles with modified surface properties allowing the covalent coupling of macromolecules on the nanoparticle surface [33,35,36,53,54]. Nanocapsules could be prepared by interfacial polymerization of alkylcyanoacrylates performed either in emulsions or in microemulsions [9,55–57]. Oil-containing nanocapsules are obtained by the polymerization of alkylcyanoacrylates at the oil/water interface of a very fine oil-in-water emulsion [55]. In practice, the oil (Miglyol威, 1 mL), the monomer (isobutyl-2cyanoacrylate, 0.125 mL), and the drug are dissolved in a water-miscible organic solvent (ethanol, 25 mL) to prepare the organic phase. This organic phase is injected in the aqueous phase (50 mL) containing a hydrophilic surfactant (Pluronic F68 0.25%) via a canula and under strong magnetic stirring. The nanocapsules form immediately to give a milky suspension. The organic phase is then removed under reduced pressure using a rotoevaporator. In such a system, the organic solvent, which is totally miscible with water, serves as a carrier for the monomer. The polymerization is believed to occur at the surface of the oil droplets when the monomer molecules encounter the water molecules [9,58–60]. To promote nanocapsule formation an ideal oil/ethanol ratio of 2% in the organic phase has been suggested [58] and the use of aprotic solvents such as acetone and acetonitrile has been recommended [61]. Nanocapsules prepared by this method contain more than 90% oil by weight [60]. Thus, this method is mainly suitable for the encapsulation of oil-soluble substances [57]. However, because of the extremely fast formation of the polymer shell of the nanocapsules, highly water-soluble macromolecules such as insulin could be encapsulated with high encapsulation yields (up to 97%) [9,59,62,63]. Water containing nanocapsules may be obtained by interfacial polymerization of alkylcyanoacrylate in water-in-oil microemulsions. In these systems, water-swollen micelles of surfactants of small and uniform size are dispersed in an organic phase. The monomer is added to the oily phase once the microemulsion is formed, and the anionic polymerization is initiated at the surface of the water-swollen micelles. The polymer which forms locally at the water/oil interface precipitates, allowing the formation of the nanocapsule shell [64–67]. The nanocapsules obtained by this method were shown to be of special interest for the encapsulation of watersoluble molecules such as peptides [65,67] and nucleic acids, including antisense oligonucleotides [66]. However, since they are dispersed in an oily continuous phase, they could have an application mainly as drug delivery systems for the oral route. For intravenous administrations, nanocapsules must be transfered into an aqueous continuous phase. This could be achieved by ultracentrifugation of the suspension over a layer of pure water [66]. Beside alkylcyanoacrylates, methylidene malonates (Fig. 3A) have been suggested as other monomers to produce biodegradable nanoparticles. This monomer could also undergo anionic emulsion polymerization to give nanospheres with a diameter ranging from 140 to 250 nm depending on the pH (6.7–8.7) of the polymerization medium [68–70]. The nanospheres could be freeze-dried and stored for at least 1 year [71]. However, their in vivo biodegradation remains a problem since at least 90% of the administered dose remained in the body for 90 days after intravenous injection into mice [68]. The very recent development of a new approach for the formulation of poly(methylidenemalonate) nanospheres could overcome this major drawback [72]. 3. Design of Copolymers The use of nanoparticles as carriers for drug delivery in humans requires that such a system remain well dispersed after administration in vivo in order to avoid the formation of aggregates
Nanoparticulate Drug Carriers
569
that may be responsible for emboli in the pulmonary capillary bed. In addition, the constituents of the systems and/or their degradation products have to be eliminated by physiological routes without inducing toxicity. Finally, if a long circulation time of the drug carrier in the bloodstream is required, the rate of uptake of the particles by phagocytes should be reduced sufficiently to make the delivery of the drug possible before elimination of the carrier. Stabilization of colloidal systems designed to be administered in vivo could be obtained by the addition of surfactants. However, many usual surfactants are not suitable for clinical use in humans. Furthermore, one of the main difficulties to be overcome is keeping the stability of colloids in a physiological medium such as blood after in vivo administration. Indeed, blood proteins can displace surfactant molecules from the nanoparticle surface and, in turn, adsorb onto it. Thus, the surfactant molecules should have a sufficiently high affinity for the nanoparticle surface to remain adsorbed in such a protein-rich medium. The nanoparticles which usually fulfill this requirement are not biodegradable, while surfactant-stabilizing particles made of biodegradable polymers are rapidly displaced by blood proteins after in vivo administration. This phenomenom leads to the formation of nanoparticle aggregates responsible for emboli and to the opsonization of the nanoparticles followed by a fast uptake by phagocytes. These problems could be overcome by incorporating the surfactant as a bound constituent of the nanoparticles, for instance, by preparing macromolecular surfactants consisting of amphiphilic block copolymers in which the hydrophilic and hydrophobic blocks are bound by a covalent linkage (Fig. 4). Such macromolecular surfactants were used to produce the so-called core-shell type nanospheres and nanocapsules (Fig. 1). These nanoparticles are usually constituted by a hydrophobic biodegradable core stabilized by a hydrophilic shell. The most popular systems designed for human use have been obtained by preparing block copolymers in which poly(ethylene oxide) (PEO) or poly(ethylene glycol) (PEG) were introduced as the hydrophilic blocks of the copolymers [20,24–26]. When not bound to the hydrophobic block, such polymers are highly water soluble. The hydrophobic blocks which were associated with PEO and PEG were composed of poly(lactide) (PLA), poly(lactide-co-glycolide) (PLGA) [20,25,73], and poly(alkylcyanoacrylates) [24,26,74]. PEG- or PEO-coated nanoparticles with a bioerodible poly(alkylcyanoacrylate) (PACA) core were obtained by initiating polymerization of alkylcyanoacrylates on either HO-PEG-OH, MeO-PEG-OH, MeO-PEG-triphenylphosphine, or triphenylphosphinePEG-triphenylphosphine, leading to triblock PACA-PEG-PACA or diblock MeO-PEG-PACA copolymers [24,35,36]. Another route for the synthesis of block copolymers containing poly(alkylcyanoacrylate) hydrophobic parts has been proposed [26]. This route, leading to branched copolymers, is based on the knoevenagel reaction. It is largely prefered as it avoids the use of the triphenylphosphine group, which may be toxic, and allows the production of very stable nanospheres [74,75]. Only a few other synthetic hydrophilic polymers have been suggested to produce core-shell type nanoparticles for therapeutic purposes. Sulfonated poly(vinyl alcohol)graft-PLGA has been used to produce negatively charged biodegradable nanospheres [76]. Currently, the synthesis of block copolymers for the formulation of drug carriers is a very active field of investigation. Most of these copolymers contain poly(ethylene glycol) blocks as the hydrophilic moiety and are actually designed to produce copolymer micelles, resulting from the self-aggregation of the copolymer molecules solubilized in water into molecular aggregates of much smaller size than nanoparticles [13,77]. (Fig. 1D). These rather new systems are exciting systems for the delivery of nucleic acids [12,78], but will not be further developed in this chapter. Special attention should be devoted to nanoparticles obtained from copolymers containing saccharides or polysaccharides as constituents of the hydrophilic shell. Sugars are actually present on the surface of cells and involved in many surface properties of the cells. Therefore, biomimetic strategies could be valuably developed taking advantage of the presence of sugars on the nanoparticle surface. A few polysaccharides are already administered to humans, for instance, dextran
570
Vauthier et al.
Figure 4 Structure of the amphiphilic block copolymers: (A) PEG-triphenylphosphate–poly(alkylcyanoacrylate) diblock and triblock copolymers [24]; (B) poly(ethylene glycol)–poly(alkylcyanoacrylate) diblock and triblock copolymers [35,36]; (C) monomethoxy or monoamino poly(PEGCA-co-HDCA) copolymer [26]; (D) poly(ethylene glycol)–poly(lactide) copolymer [20].
and heparin. Heparin is well known for its anticoagulant activity, and it has been demonstrated to also act as a physiological inhibitor of complement activation. It is a known fact that the complement system plays a key role in the nonspecific recognition and uptake of foreign bodies including colloidal drug carriers by phagocytes. Thus, in order to mimic the behavior of cells and pathogens which normally escape recognition by complement and phagocytes, copolymers of heparin and poly(methyl methacrylate) were produced by Passirani et al. [79] and heparinecoated nanospheres were prepared. These nanospheres have been shown to be nonactivators of complement in vitro [80]. In vivo, after intravenous administration, these nanospheres could remain in the bloodstream and show long-term circulating properties [81]. In addition, the results of this work have shown that the conformation of the polysaccharide chains grafted on the nanosphere surface could play an important role to define the fate of the colloidal particle after intravenous administration. Indeed, dextran bound to nanospheres by one chain end has been demonstrated to be as low an activator of complement as soluble dextran, whereas crosslinked dextran (i.e., Sephadex威) is a strong activator. Unfortunately, these heparine-coated nanopheres are not biodegradable and cannot be used clinically. This issue has recently been overcome by Chauvierre et al. [82,83], who described a general method for the synthesis of core-shell type nanospheres with a poly(alkylcyanoacrylate) core coated with a polysaccharide shell of a different nature. Following the strategy of using saccharides to modify nanoparticle surface properties, Rouzes et al. [84] suggested new hydrophobically modified dextrans as stabilizers of surfacemodified poly(lactide) nanospheres obtained by the oil-in-water emulsion/evaporation technique.
Nanoparticulate Drug Carriers
571
Positively charged nanoparticles could be prepared from chitosan–poly(butyl cyanoacrylate) copolymers [54], whereas functionalized nanoparticles with a poly(ε-caprolactone) core have recently been obtained by ring-opening polymerization of ε-caprolactone with monosaccharides as transfer agents [85]. Finally, copolymers of polyesters such as poly(ε-caprolactone) and poly(lactide) with dextran or other polysaccharides have been synthesized after introduction and activation of carboxylic groups at the chain ends of the polyesters to allow coupling to the hydroxyl functions carried by the polysaccharides [86]. A rather large number of copolymers are now available to obtain core-shell nanoparticles possessing a degradable core. The linked hydrophilic shell stabilizes the suspension and ensures particles with special properties suitable for the physiological medium for which they have been designed. B. Polymer Physicochemistry and Design of Nanoparticulate Drug Carriers In this group of methods, the nanoparticles are obtained from a polymer which is prepared according to a totally independent method. The methods mainly take advantage of the physicochemical properties of the polymers in terms of their solubility or of their faculty to form a gel under certain conditions. One can distinguish between methods based on the spontaneous formation of colloidal particles and methods which derive from microencapsulation techniques (Fig. 2). 1. Methods Based on the Spontaneous Formation of the Nanoparticles Spontaneous formation of nanoparticles can be achieved by taking advantage of the solubility and gelling properties of a dissolved polymer. Usually the step allowing polymer colloid particles to form is reversible and it is necessary to complete the procedure by a second step required to stabilize the polymer nanoparticles. The general principle, based on the solubility properties of a polymer, is to prepare a solution of the polymer and to induce a phase separation by addition of a nonsolvent [87] or by a salting-out effect [88]. The occurrence of the phase separation can be followed by turbidimetric measurements [88] or investigated using ternary phase diagrams [89]. Phase separations leading to polymer colloid particles are usually obtained with diluted solutions of polymers. In a ternary phase diagram, this corresponds to a small area. It is recommended to use a nonsolvent which is totally miscible with the polymer solvent. When the polymer was a protein, the colloidal particles formed at the limit of the phase separation which can be followed by turbidimetric measurements. In both cases, the particles should form spontaneously and almost instantaneously. Once the proper conditions to form polymer colloids are identified, the particles must be stabilized either by the elimination of the polymer solvent by evaporation or by chemical crosslinking of the polymer [90]. This method, named nanoprecipitation, could be applied to numerous synthetic polymers [87,89]. Usually a polymer is dissolved in acetone and is allowed to precipitate, such as colloidal polymer particles in water. The acetone is then evaporated to complete the insolubilization of the polymer particles formed during the first step of the procedure. Surfactants are added to water for better stability of the resulting nanospheres. This easy technique was scaled up for large batch production. It leads to the formation of nanospheres for lipophilic drugs. Proteins used as a polymer to produce the nanospheres represent much better material for the encapsulation of hydrosoluble compounds. Nanospheres made of gelatin were proposed as a carrier system for gene delivery [91,92].
572
Vauthier et al.
Nanocapsules could be prepared by a slightly modified method in which a small amount of an organic oil is added in the polymer solution [93,94]. When the polymer solution is poured into the water phase, the oil is dispersed as tiny droplets in the solvent/nonsolvent mixture, and the polymer precipitates onto the oil droplet surface. This method, leading to the preparation of oil-containing nanocapsules, is suitable for the encapsulation of liposoluble drugs. Another approach for the production of nanospheres which form spontaneously is based on the gelation properties of a few polymers and requires the control of their gelation process. This approach has mainly been developed with alginate and chitosan that form hydrogels, i.e., gels greatly swollen by water. The gelling agents for these two natural polysaccharides are, respectively, calcium and tripolyphosphate [95,96]. With alginate, nanospheres can be prepared when the respective concentrations of alginate and calcium are comprised in the domain of the pregel stage of the alginate gelling process (alginate 0.06%, w/w, calcium chloride 0.9mM). In this composition, tiny particles of gels form, resulting from inter- and intramolecular aggregations of alginate molecules caused by the interaction with calcium. Polylysine is then used to stabilize these aggregates by the formation of a polyelectrolyte complex with the alginate gel. Chitosan formed gel nanospheres with tripolyphosphate in the presence of poly(ethylene glycol). Alginate and chitosan nanospheres were found to be interesting carriers for nucleic acid and protein delivery [5,10,97]. More recently, poly(ethylene glycol)/poly(ethylenimine) copolymers were formulated as gelified nanospheres indicating a promising future for the delivery of nucleic acids [98]. These gel nanospheres are a family of nanoscale materials of particular interest because they exhibit intrinsic gel properties combined with the properties of colloids, such as heterogenous structure, small size, and high surface-to-volume ratio [98]. Induction of spontaneous formation of nanoparticles can also be achieved by complex coacervation, producing a polyelectrolyte complex. Based on this approach, DNA-chitosan nanospheres could be obtained using sodium sulfate as desolvating agent [99]. This technique yields nanospheres with a diameter ranging from 200 to 500 nm in which plasmids of various size (5.1 to 11.9 Kb) could be incorporated [100]. A polyelectrolyte complex made of poly(ethylenimine) and DNA may also form upon mixing poly(ethylenimine) with DNA [101]. These systems may sometimes be considered as nanospheres by the authors, depending on the size of the complex formed. 2. Methods Derived from Microencapsulation Techniques Nanospheres and nanocapsules can be prepared by the same methods as those described for microparticles, except that manufacturing parameters are adjusted to obtain nanometer-sized particles. Microencapsulation techniques adapted for making nanoparticles require the formation of an emulsion as a first step of the procedure. Special equipment is needed to reduce the droplet size of the emulsion during dispersion of the polymer solution into the continuous aqueous phase. This equipment is a high pressure homogenizer and a microfluidizer in which the energy input produced is considerably high and mainly due to high turbulence and cavitation forces. Alternative techniques are based on the use of ultrasounds. Once the desired emulsion is prepared, the formation of the nanoparticles can be induced in two ways: the gelation of the polymer on the one hand and the precipitation of the polymer either by solvent displacement or by solvent evaporation on the other hand. The gelation of the polymer within the droplets of the emulsion can be induced in different ways according to the nature of the polymer, and this generates the formation of nanospheres. Gelation of chitosan is generally induced by increasing the pH, whereas the gelation of alginate is achieved by adding calcium. Agarose nanospheres can be obtained by decreasing the temperature of the emulsion [5].
Nanoparticulate Drug Carriers
573
In the solvent displacement technique, the emulsion is prepared with a solvent of the polymer which is partially soluble in water and with an aqueous phase saturated with the polymer solvent. Once the emulsion is formed, for example, with ethylacetate, it is diluted by the further addition of water to displace the ethylacetate from the dispersed phase toward the continuous aqueous phase. This induces the polymer precipitation [102–104], and the resulting particles are nanospheres. Another technique was developed based on an inverse salting-out method [105,106]. A solution of polymer in acetone was dispersed in an aqueous phase containing a high concentration of salt to keep the acetone nonmiscible with water. Just by diluting the emulsion with a large amount of pure water, the acetone was then extracted from the dispersed phase, inducing the polymer precipitation. The total elimination of the acetone can be achieved by evaporation. Finally, precipitation of the polymer solubilized in the dispersed phase of the emulsion can be induced by the removal of the solvent by evaporation, leading to the formation of nanospheres [107,108]. This method was named emulsification/solvent evaporation. In contrast with the previous techniques described, the particles form rather slowly. Indeed, the polymer solvent diffuses through the aqueous continuous phase and evaporates at the air/water interface. Therefore, the emulsion should remain under agitation during the time required for the total evaporation of the solvent. This method is mainly suited to the encapsulation of hydrophobic drugs [109,110]. To produce nanocapsules, small amounts of oil can be added in the dispersed phase of the emulsion that will form the nanocapsules by solvent displacement [102]. Nanocapsules can also be prepared using the solvent evaporation technique after a double water-in-oil-in-water emulsion with the proper size range has been obtained [111,112]. In this double emulsion, the removal of the solvent in the intermediate phase by evaporation causes the polymer to precipitate at the surface of the inner aqueous phase. Whereas nanocapsules made by the solvent displacement method are more suitable for the encapsulation of lipophilic compounds [113,114], this latter method allows the encapsulation of hydrosoluble compounds [112,115,117]. Among the various polymers which can be proposed to prepare nanoparticles with these methods, poly(esters) like poly(lactide), poly(glycolide), and poly(lactide-co-glycolide) have been widely used up to now because of their excellent biocompatibility and biodegradability [97]. At the moment, the newly synthesized block copolymers have generated tremendous interest for the development of long-circulating nanoparticles [25,74,86,116,118,119]. One of the main limitations of these technologies is the difficulty encapsulating a drug and particularly controlling its release once it has been efficiently encapsulated within the nanoparticles. A few years ago, Quintanar-Guerrero et al. [120] suggested applying the ion-paring idea to overcome this problem, which considerably limited further development of the nanoparticles prepared according to the methods which had been used up to then.
III. APPLICATIONS OF NANOPARTICLES Although a wide variety of polymers has generated an immense interest for producing nanoparticles for therapeutic purposes, only a few types of nanoparticles were actually tested on animal models for the treatment of several diseases. As described below, major developments were performed with poly(alkylcyanoacrylate) nanoparticles. The main other polymer nanoparticles tested on animals were made of poly(lactide) derivatives and chitosan. However, poly(ε-caprolactone) nanoparticles were found to be superior for ocular therapy. The notion of drug targeting is particularly important for drugs with a narrow therapeutic window and the potential danger of detrimental effects, such as cancer and intracellular infections. Another pertinent reason may be related to the lack of stability of the drug in biological media, for instance, proteins, peptides,
574
Vauthier et al.
and nucleic acids. Therefore, nanoparticles have been used as carriers for many kinds of molecules, and various routes of administration have been investigated (Table 1). In some cases, nanoparticles have been shown to be more active than liposomes thanks to their better stability.
A. Intravenous Administration 1. The Fate of Nanoparticles and Their Content After Intravenous Administration The main interest of nanoparticles is their ability to achieve tissue targeting and enhance the intracellular penetration of drugs. After intravenous administration, nanoparticles are taken up by the liver, spleen, and to a lesser extent the bone marrow [121]. Within these tissues, nanoparticles are mainly taken up by cells of the mononuclear phagocyte system [122]. The uptake occurs through an endocytosis process after which the particles end up in the lysosomal compartment [122] where they are degraded, producing low molecular weight–soluble compounds that are eliminated from the body by renal excretion [123]. As a result of the mononuclear phagocyte system site-specific targeting, the avoidance of some organs were made possible, thus reducing the side effects and toxicity of some active compounds. Because of their strong lysosomal localization, it would seem that nanoparticles are not suitable for targeting the cytoplasm. To avoid their being trapped within the lysosomal compartment, several compounds able to destabilize the lysosomal membrane were added to the nanoparticulate systems (e.g., cationic surfactant) [124], allowing some drugs to be delivered to the cytoplasm. The development of sterically stabilized nanoparticles, in order to avoid mononuclear phagocyte system uptake, has been made possible by the linkage of poly(ethylene glycol) derivatives to the nanoparticle surface [20,25,75]. This linkage results in a lower uptake of nanoparticles by the mononuclear phagocyte system and in a longer circulation time in the blood after intravenous administration. As a consequence, these so-called stealth nanoparticles would be able to extravasate across the endothelium that becomes permeable due to the local alteration of the tissue caused by the disease, for instance, the presence of solid tumours. 2. Applications in the Treatment of Intracellular Infections Intracellular infections were found to be a field of application for drug delivery by means of nanospheres [125]. Indeed, infected cells may constitute a ‘‘reservoir’’ for microorganisms which are protected from antibiotics inside lysosomes. The resistance of intracellular infections to chemotherapy is often related to the low uptake of commonly used antibiotics or to their reduced activity in the presence of the acidic pH of lysosomes. To overcome these difficulties, the use of ampicillin, a -lactam antibiotic, bound to nanospheres was proposed as an endocytozable formulation [126]. The effectiveness of poly(isohexylcyanoacrylate) nanospheres was tested in the treatment of two experimental intracellular infections. First, ampicillin-loaded nanospheres were evaluated in the treatment of experimental Listeria monocytogenes infection in congenitally athymic nude mice, a model involving chronic infection of both liver and spleen macrophages [127]. After adsorption of ampicillin onto nanospheres, the therapeutic activity of ampicillin was found to increase dramatically over that of the free drug. Bacterial counts in the liver were at least 20-fold reduced after linkage of ampicillin to poly(isohexylcyanoacrylate) nanospheres. Nanoparticulate ampicillin was capable of ensuring liver sterilization after two injections of 0.8 mg of a nanosphere-bound drug. However, reappearance of living bacteria in the liver after the end of the treatment was probably due to a secondary infection derived from other organs such as the spleen, which was not completly sterilized by the treatment [127].
Nanoparticulate Drug Carriers
575
Table 1 Therapeutic Application of the Nanoparticules According to Their Route of Administration Route of administration Intravenous
Oral route
Applications Intracellular infections
PACA nanospheres
Cancers
PACA nanospheres
Delivery of drugs to the brain
Oil-containing PLA nanocapsules Water-containing PACA nanocapsules Polysorbate 80-coated nanospheres
Delivery of peptides and proteins
Vaccines Other drugs
Ocular
Type of nanoparticles
Antiglaucomatous
Anti-inflammatory drug Peptides
Antiviral agent
Oil-containing PACA nanocapsules PACA nanospheres Water-containing PACA nanocapsules PACA nanospheres PACA nanospheres Oil-containing PACA nanocapsules Oil-containing PACA nanocapsules Oil-containing PLGA nanocapsules PACA nanospheres in PEG gel Oil-containing PCL nanocapsules Oil-containing PCL nanocapsules Oil-containing PCL nanocapsules Chitosan nanospheres PACA nanospheres in Poly(acrylic acid) gel PEG-coated PACA nanospheres
Drugs [references] Ampicillin [126,127], ciprofloxacin [132], gentamicin [133], primaquine [134,135], dihydroemetine [136] Doxorubicin [139,140,148,149], granulocyte colony stimulating factor [145], antisense–oligonucleotides (anti-HA-ras) [165] Muramyl tripeptide cholosterol [157,158] Antisense–oligonucleotides (anti-Ewing sarcoma) [168] Kytorphin [172], dalargin [173], loperamide [174], turbocurarine [175], NMDA receptor antagonist [176, 177], doxorubicin [178,179] Insulin [62,200,201], salmon calcitonin [200], octreoide [63] Insulin [203] Calcitonin [65], insulin [204] Antigen [206] Vicamin [213] Lipiodol [214], darodipin [215], indomethacin [216] Bataxolol [218–220] Carteolol [218–220] Metipranol [218–220] Betaxolol [219] Pilocarpine [230] Carteocol [218] Indomethacin [223] Cyclosporine A [226] Cyclosporine A [228] Cyclosporine A [229] Acyclovir [231]
PACA, poly(alkylcyanoacrylate); PLA, poly(lactide); PLGA, poly(lactide-co-glycolide); PCL, poly(-caprolactone); PEG; poly(ethylene glycol).
576
Vauthier et al.
Second, nanosphere-bound ampicillin was tested in the treatment of experimental salmonellosis in C57/BL6 mice, a model involving acute fatal infection [126]. All mice treated with a single injection of nanosphere-bound ampicillin survived, whereas all control mice and all those treated with unloaded nanospheres died within 10 days postinfection. The therapeutic index of ampicillin, calculated on the basis of mouse mortality, was increased 120-fold when the drug was bound to nanospheres. The mechanism by which nanospheres improved the antimicrobial efficacy of ampicillin has been investigated by several authors [128–130]. In vitro, ampicillin-loaded poly(isobutylcyanoacrylate) nanospheres appeared to be much more effective than free ampicillin for inhibiting intracellular growth of L. monocytogenes in peritoneal macrophages [128]. With in vitro Salmonella typhimurium–infected macrophages, the situation was slightly more complicated since the bactericidal effect of ampicillin-bound poly(isohexylcyanoacrylate) nanospheres was poor, although the intracellular capture of ampicillin was dramatically increased and its efflux in the extracellular medium reduced [129]. The intracellular traffic of ampicillin-bound poly(isohexylcyanoacrylate) nanospheres has been investigated by confocal microscopy and transmission electron microscopy [130]. Observations clearly demonstrated that ampicillin-bound poly(isohexylcyanoacrylate) nanospheres taken up by murine macrophages were localized in the same vacuoles as the infecting bacteria, but in a restrictive way [130]. Thus, it has been postulated that the limited bactericidal effect measured could be due to a resistance mechanism of S. typhimurium involving the inhibition of the phagosome–lysosome fusion [131], which leaves some bacteria in phagosomes free of nanospheres. If this hypothesis were correct, the dramatic efficiency observed in vivo would rather be due to the specific targeting of the infected tissues (rich in reticuloendothelial cells) than to efficient intracellular targeting. In order to eliminate both dividing and nondividing bacteria, a fluoroquinolone antibiotic, ciprofloxacin, was associated with poly(isobutylcyanoacrylate) and poly(isohexylcyanoacrylate) nanospheres. In an animal model of persisting salmonella infection, although an effect on the early phase of the infection was observed, neither free nor nanosphere-bound ciprofloxacin was able to truly eradicate the persisting bacteria [132]. Poly(butylcyanoacrylate) nanospheres were also shown to promote the uptake of gentamycin, another interesting antibiotic for the treatment of intracellular infections, by cultured mice intraperitoneal macrophages and rat hepatocytes [133]. The nature of the stabilizer used in the nanosphere preparation, the particle size, and the gentamicin concentration were all found to have an effect on the uptake of the nanospheres by the phagocytic cells. Since they accumulate in the mononuclear phagocyte system, nanospheres are also promising drug carriers for the treatment of visceral leishmaniosis [134]. Thus it has been suggested that poly(isobutylcyanoacrylate) nanospheres could be used as a carrier of primaquine. The activity of the drug was increased 21-fold against intracellular Leishmania donovani when associated with nanospheres [135]. Part of the antiparasitic activity was attributed to the respiratory burst induced by the phagocytosis of nanospheres by the macrophages [135]. Dehydroemetine is another drug candidate for this treatment; it does however have some side effects involving the heart which were reduced after linkage with nanospheres [136]. 3. Applications in the Treatment of Cancer When given intravenously, anticancer drugs are distributed throughout the body as a function of the physicochemical properties of the molecule. A pharmacologically active concentration reaches the tumor tissue at the expense of massive contamination of the rest of the body. For cytostatic compounds, this poor specificity raises a toxicological problem which is a serious obstacle to effective therapy. Indeed, alteration of the drug distribution profile by linkage to
Nanoparticulate Drug Carriers
577
nanospheres can, in some cases, considerably reduce the toxicity of a drug because of the reduced accumulation in organs where the most acute toxic effects are exerted. This phenomenon was illustrated in particular with doxorubicin, which displays severe acute and chronic cardiomyopathy. After intravenous administration to mice, plasma levels of doxorubicin were higher when the drug was adsorbed onto nanospheres, and at the same time the cardiac concentration of the drug was dramatically reduced [137]. In agreement with the observed distribution profile, doxorubicin associated with nanospheres was found to be less toxic than free doxorubicin [138]. Therefore, the use of colloidal drug carriers could be a more rational approach to specific cancer therapy. In addition, the possibility of overcoming multidrug resistance might be achieved by using cytostatic-loaded nanospheres. The antitumor efficacy of doxorubicin-loaded nanospheres was first tested using the lymphoid leukemia L-1210 as a tumor model. In this study, one intravenous injection of doxorubicin-loaded poly(isobutylcyanoacrylate) nanospheres was found to be more effective against L1210 leukemia than when the drug was administered in its free form following the same dosing schedule [139]. The effectiveness of doxorubicin-loaded poly(isohexylcyanoacrylate) nanospheres against L1210 leukemia was more pronounced than that of doxorubicin loaded onto poly(isobutylcyanoacrylate) nanospheres. The drug toxicity was markedly decreased when it was bound to this sort of nanosphere. Consequently, impressive results were obtained with this formulation at doses for which the therapeutic efficiency of free doxorubicin was completely masked by the overpowering toxicity of the drug [139]. Preliminary experiments suggested that one intravenous bolus injection of doxorubicin-loaded nanospheres was more effective in L1210bearing mice than perfusion of the free drug for 24 h. The superiority of doxorubicin targeted with the aid of poly(alkylcyanoacrylate) nanospheres was later confirmed in a murine hepatic metastase model (M5076 reticulosarcoma) [140]. Irrespective of the dose and the administration schedule, the reduction in the number of metastases was much greater with doxorubicin-loaded nanospheres than with free doxorubicin, particularly if treatment was given only when the metastases were well established. Histological examinations of the liver tissue confirmed the improved efficacy of the targeted drug [140]. In order to elucidate the mechanism behind the enhanced efficiency of doxorubicin-loaded nanospheres, doxorubicin measurements were made in both metastatic nodules and neighboring healthy hepatic tissue [138]. During the first 6 h after administration the exposure of the liver to doxorubicin was 18 times greater for nanosphere-associated doxorubicin and, unexpectedly, no special affinity for the tumor tissue was detected. Electron microscopy revealed that the nanospheres were inside the Ku¨pffer cells (phagocytes). At later time points, the amount of drug in the tumor tissue finally increased in nanosphere-treated animals to 2.5 times the level found in animals given free doxorubicin. Since direct uptake of nanospheres by neoplastic tissue is unlikely, Ku¨pffer cells in the healthy hepatic tissue could play the role of a drug reservoir from which prolonged diffusion of the free drug toward the neighboring malignant cells could occur. This hypothesis raises the question of the long-term effect of an 18-fold increase of doxorubicin concentration in the liver. Doxorubicin-loaded nanospheres were not significantly or unexpectedly toxic to the liver in terms of survival rate at high doses, body weight loss, or histological appearance [14]. However, as observed with doxorubicin-loaded liposomes [141], a reversible decline in the phagocytic capacity of the liver after repeated dosing with poly(alkylcyanoacrylate) nanospheres due to a temporary depletion in number of Kupffer cells was observed as well as a slight inflammatory response [142,143]. Nanosphere-associated doxorubicin also accumulated in bone marrow, leading to a myelosuppressive effect [144]. However, this tropism of carriers might be useful to deliver myelostimulating compounds such as granulocyte colony stimulating factor to reverse the suppressive effects of intense chemotherapy [145]. Nanospheres are also captured by splenic macrophages [146].
578
Vauthier et al.
The spleen architecture was shown to play a role in the localization of the nanospheres. In mice, uptake was mainly observed in the metallophilic macrophages of the marginal zone, whereas in rats, which have a sinusoidal spleen similar to that of humans, nanospheres were found in the red pulp macrophages. The ability of tumor cells to develop simultaneous resistance to multiple lipophilic compounds is a major problem in cancer chemotherapy. Cellular resistance to anthracyclines has been attributed to an active drug efflux from resistant cells linked to the presence of transmembrane Pglycoprotein, which was not detectable in the parental drug-sensitive cell line. Upon entering the cell by passive diffusion through the lipid bilayer, drugs such as doxorubicin bind to Pglycoprotein and are rapidly pumped out of the cells by an active process [147]. To overcome pleiotropic resistance, the use of competitive P-glycoprotein inhibitors, such as the calcium channel blocker verapamil has been proposed. However, since the adverse effects of verapamil are serious, its clinical use for this purpose is limited. Thus, nanospheres loaded with doxorubicin are an interesting alternative to overcome multidrug resistance due to the presence of P-glycoprotein in the membrane of resistant cells. The cytotoxicity of free doxorubicin, doxorubicin-loaded poly(isohexylcyanoacrylate) nanospheres (mean diameter 300 nm), and nanospheres without drug against sensitive and multidrugresistant human cancer cell lines was compared [148–150]. Doxorubicin resistance could be circumvented in the majority of the resistant cell lines tested when the drug was incubated as associated with poly(alkylcyanoacrylate) nanospheres. These results indicated that nanospheres provided an effective carrier for introducing a cytotoxic dose of doxorubicin into the pleiotropicresistant human cancer cell lines. Some encouraging results were also obtained in vivo in a P388 model growing as ascites [149]. The mechanism of action of poly(alkylcyanoacrylate) nanospheres has been investigated in vitro on cell cultures [151–153]. The reversion of resistance seems to be due both to the adsorption of nanospheres on the cell surface and to the formation of a doxorubicin–poly(cyanoacrylic acid)—i.e., degradation product of a poly(alkylcyanoacrylate)—ion pair which facilitates the transport of the drug across the cell membrane [153,154]. In the light of the results obtained with doxorubicin-loaded nanospheres in the liver metastases model described above [140], the role of macrophages as a reservoir for doxorubicin was tested in a two-compartment coculture system in vitro with both resistant and sensitive P388 cells [155]. Even after prior uptake by macrophages, doxorubicin-loaded poly(isobutylcyanoacrylate) nanospheres were able to partially overcome resistance. An extremely effective reversion of P388 resistance could be produced when both doxorubicin and cyclosporin A, a compound capable of inhibiting the P-glycoprotein, were bound to the same nanospheres [156]. The association of cyclosporin A with nanospheres would ensure that it reached the same sites as the anticancer drug at the same time and would also reduce its toxic side effects. Besides poly(alkylcyanoacrylate) nanospheres and as early as 1986, Al Khoury-Fallouh et al. [55] observed that, like other colloidal carriers, nanocapsules (here administered by the intravenous route in rabbits) were taken up rapidly by organs of the mononuclear phagocyte system. One application to treat metastatic cancer that takes advantage of this uptake concerns nanocapsules of MTP-Chol (muramyl tripeptide cholesterol). This immunostimulating agent is able to activate macrophages and induce toxicity toward tumor cells due to the production of nitric oxide and TNF-␣. It had been shown in in vitro models of rat alveolar macrophages and RAW 264.7 mouse monocyte macrophage line that nanocapsules based on poly(D,L-lactide) containing MTP-Chol are more efficient activators than the free drug [157,158]. This action could be due to an intracellular delivery of the immunomodulator encapsulated in nanocapsules after phagocytosis and to an intermediate transfer of the drug to serum proteins [159]. This system has also demonstrated its efficiency in vivo. In fact, Barratt et al. [160] reported antimeta-
Nanoparticulate Drug Carriers
579
static effects of nanocapsules containing MTP-Chol in a model of liver metastases. Some antimetastatic activity was also seen after oral administration. 4. Applications in the Delivery of Oligonucleotides Oligodeoxynucleotides are potentially powerful drugs because of their selectivity for particular gene products in both sense and antisense strategies. However, using oligonucleotides in therapeutics is a challenge to pharmaceutical technology because of their susceptibility to enzymatic degradation and their poor penetration across biological membranes. Nanoparticulate preparations might be an interesting alternative because of better stability in the presence of biological fluids. In the case of nanospheres made of synthetic polymers [poly(alkylcyanoacrylate), poly(lactide)], since oligonucleotides have no affinity for the polymeric matrix, association with nanoparticles has been achieved either by ion pairing with a cationic surfactant, cetyltrimethylammonium bromide (CTAB) adsorbed onto the nanoparticle surface [124,161], or by complexing with structured oligopeptides [113,162]. Antisense oligonucleotides bound to poly(alkylcyanoacrylate) nanospheres via CTAB were protected from nucleases in vitro [124] and their intracellular uptake was increased [161]. The nanospheres were able to concentrate intact oligonucleotides in the liver and in the spleen [164]. Antisense oligonucleotides formulated in this way were also able to specifically inhibit mutated Ha-ras–mediated cell proliferation and tumorigenicity in nude mice [165]. However, these nanospheres were not totally satisfactory because of their toxicity, mainly due to the presence of CTAB [166], and of the quick desorption of the oligonucleotides in the presence of serum, which results from surface erosion by serum esterases [164]. Thus nanocapsules with an aqueous core containing antisense oligonucleotides were developed [66,167]. The protection of oligonucleotides against degradation by serum nucleases was much more efficient with these nanocapsules than that obtained with CTAB-coated nanospheres [124,161,167]. Phosphorothioate oligonucleotides directed against EWS Fli-1 chimeric RNA were encapsulated within poly(alkylcyanoacrylate) nanocapsules and tested in vivo for their efficacy against the experimental Ewing sarcoma in mice after intratumoral administration [168]. Only intratumoral injection of antisense-loaded nanocapsules led to a significant inhibition of tumor growth at a cumulative dose of 14.4 nmol, whereas no antisense effect could be detected with the free oligonucleotide. It is noteworthy to point out that, using the same antisense sequence as a free drug, Tanaka et al. [169] demonstrated that inhibition of a tumor growth in a similar model required a cumulative dose of 500 nmol oligonucleotide. This dose was 35-fold higher than the one required to obtain a comparable effect with the nanocapsules. Therefore, nanocapsule technology would allow lower phosphorothioate doses to be used and thus avoid the toxicity and the loss of specificity resulting from phosphorothioates at higher doses [170]. The mechanism by which oligonucleotide in nanocapsules led to a significant effect on the tumor growth may be explained by the protection of the oligonucleotide afforded by the nanocapsules, which may also act as a controlled release system of the oligonucleotide within the tumor. Thus, the use of phosphorothioates at low doses combined with nanocapsules may represent a new and safe solution for the administration of antisense therapy in vivo. Nanospheres containing oligonucleotides have also been formulated from a naturally occurring polysaccharide, alginate, which forms a gel in the presence of calcium ions. In this case, the oligonucleotides penetrate into the gel matrix by reptation, thus providing a high loading yield and good protection against nucleases [171]. 5. Applications in the Passage of the Blood–Brain Barrier The blood–brain barrier is an insuperable obstacle for a large number of drugs, such as antibiotics, anticancer agents, and a variety of central nervous system–active drugs, especially neuropep-
580
Vauthier et al.
tides. One of the possibilities to overcome this barrier is a drug delivery to the brain using nanoparticles [172]. Several drugs have already been successfully transported into the brain using this carrier: the hexapeptide dalargin [173], the dipeptide kytorphin [172], loperamide [174], tubocurarine [175], the NMDA receptor antagonist MRZ 2/576 [176,177], and doxorubicin [178]. The nanoparticles may be especially helpful for the treatment of the disseminated and very aggressive brain tumors. Indeed, intravenously injected doxorubicin-loaded polysorbate 80–coated nanospheres were able to lead to a 40% cure in rats with intracranially transplanted glioblastomas 101/8 [178,179]. The mechanism of the nanosphere-mediated transport of the drugs across the blood–brain barrier is not fully elucidated at present. The most likely mechanism is endocytosis by the endothelial cells lining the brain blood capillaries. Indeed, observations based on fluorescence and confocal laser scanning microscopy demonstrated that the nanospheres were taken up by cells and that this uptake occurs via an endocytotic mechanism followed by a transcytosis [180,181]. Nanosphere-mediated drug transport to the brain depends on the overcoating of the particles with polysorbates, especially polysorbate 80, which seems to lead to the adsorption of apolipoprotein E from blood plasma onto the nanosphere surface [182]. The nanospheres then ‘‘labeled’’ with apolipoprotein E seem to mimic low density lipoprotein particles and could interact with the low density lipoprotein receptors of the blood–brain barrier endothelial cells resulting in their uptake by these cells. The drug may then be released from the nanospheres taken up in these cells and diffuses into the brain interior. Other processes such as transcytosis of the nanospheres, tight junction modulation, or P-glycoprotein inhibition may also be involved. These mechanisms may occur in parallel or may be cooperative, thus enabling an efficient drug delivery to the brain. Recently, Olivier et al. [183] have shown that a nonspecific permeabilization of the blood–brain barrier, probably related to the toxicity of the carrier, may account for the central nervous system drug penetration when associated with poly(butylcyanoacrylate) nanospheres and polysorbate 80. Calvo et al. [184,185] have evaluated the ability of long-circulating PEG-coated poly(cyanoacrylate) nanospheres made by the amphiphilic copolymer poly[methoxy poly(ethylene glycol) cyanoacrylate-co-hexadecyl cyanoacrylate] (PEGCA-PHDCA) to diffuse into the brain tissue after intravenous administration in mice and rats. Based on their long circulating characteristics, (PEGCA-PHDCA) nanospheres penetrated into the brain to a larger extent than all the other tested formulations. Particles were located in the ependymal cells of the choroid plexuses, in the epithelial cells of pia mater and ventricles, and to a lower extent in the capillary endothelial cells of brain–blood barrier. These phenomena occurred without any modification of brain–blood barrier permeability, whereas polysorbate 80–coated nanospheres owed their efficacy in part to brain–blood barrier permeabilization induced by the surfactant. Poloxamine 908–coated nanospheres, which also display long circulating properties, failed to increase brain concentration probably because of their inability to interact with cells [184]. The concentration of PEG-coated nanospheres in the central nervous system, especially in white matter, was shown to be greatly increased in comparison to conventional non-PEG-coated nanospheres. In addition, this increase was significantly higher in pathological situations where brain–blood barrier permeability is augmented and/or macrophages have infiltrated the brain. Passive diffusion and macrophage uptake in inflammatory lesions seems to be the mechanism underlying such brain penetration. In addition, these PEG-coated nanospheres showed comparatively to conventional nonPEG-coated nanospheres a higher uptake by the spleen and the brain, which are both the target tissues of PrPres accumulation in scrapie-infected animals [186]. B. Subcutaneous/Intramuscular Administration Subcutaneous administration of nanoparticles was achieved mainly for the delivery of peptides and vaccines. It allows slow release of the entrapped drugs, therefore reducing the number of
Nanoparticulate Drug Carriers
581
administrations, increasing blood half-life of the active drug, and in some cases reducing side effects. Poly(isobutylcyanoacrylate) nanospheres were injected subcutaneously to rats. Autoradiographic pictures obtained after using radiolabeled polymer have shown a progressive staining reduction in the muscular tissue, suggesting that nanospheres were slowly biodegraded [187]. In addition, these nanospheres were found to release a peptide (growth releasing factor) in a sustained manner. Comparison of the area under the curve (AUC) of free growth releasing factor and growth releasing factor–loaded nanospheres showed that in addition to the slow release process nanospheres were also able to improve the bioavailability of the peptide. This improvement could be attributed to the fact that free administered growth releasing factor is very quickly metabolized at the injection site, whereas it is partly protected from massive enzymatic degradation when it is administered associated with nanospheres [40,187]. A few examples of the use of nanospheres as adjuvant for antigens and allergens delivery were described in the literature. The main advantage of this approach is to design single-shot vaccines. In this case the drug carrier has to remain at the site of administration and deliver either continuously or pulsatively the antigen. The use of slowly degradable polymers [poly(lactide), poly(methylmethacrylate)] is suitable for this application, allowing the peptide or protein to be released more adequately. Poly(methylmethacrylates) were first investigated as adjuvants for injectable vaccines [31,188,189]. These nanospheres are not really biodegradable after subcutaneous or intramuscular injection, but they were shown to exhibit very powerful adjuvant properties for a number of antigens. The adjuvant properties were better when the antigen was incorporated during the preparation of the nanospheres by emulsion polymerization than when adsorbed onto the nanosphere surface [190]. The antibody response after immunization against influenza whole and split virus, bovine serum albumin, and HIV-2 split virus with different types of nanospheres was increased for nanospheres of the smallest size and made of the more hydrophobic polymers [191–193]. C. Oral Route There are numerous reports showing that uptake and translocation of nanoparticles and microparticles takes place after oral administration in experimental models [194–196]. Nanoparticle uptake from the gut is an important route of entry to the systemic circulation. There have been many studies in this field investigating how particle uptake from the gastrointestinal tract occurs [15]. The involvement of the specialized M-cells in Peyer’s patches of the capture process is now reasonably clear, but the detailed mechanisms of this uptake still are not. The process could be particle dependent. Indeed, poly(alkylcyanoacrylate) nanocapsules have been localized by different authors and using various techniques over Peyer’s patches, in M-cells and in intercellular spaces arround lymph cells [196,197]. Nanocapsules filled with an iodinized oil (lipiodol) in order to render them detectable using a scanning electron microscope equipped with an energydispersive x-ray spectrometer appeared as vesicle-associated with intraluminal mucus and were observed in intravillus capillaries after administration in an isolated segment of a dog jejunum. In these capillaries, they were found in close contact with red blood cells and adsorbed to the inner wall of the endothelial cells. In the case of poly(lactide) nanosphere uptake, only little discrimination between lymphoid and nonlymphoid tissue was reported, whereas other types of particles prefered the M-cell route [198]. Particles with a diameter of 100 nm were shown to diffuse throughout the submucosal layer better than larger size particles [199]. Although it might exist in certain situations, the passage of particles between the absorptive cells is rather likely if the tight junctions are not disrupted. The potential effect of gastrointestinal diseases may also alter the uptake pathways of nanoparticles delivered by the oral route. Independently of the way
582
Vauthier et al.
nanoparticles are absorbed by the gastrointestinal tract, the question remains open whether or not the extent of particle translocation is compatible with a strategy of drug administration with therapeutic perspectives. This will be discussed below. 1. Oral Delivery of Peptides, Proteins and Vaccines Studies concerning the oral delivery of peptides, proteins, and vaccines by means of colloid polymeric carriers have been predominantly performed with poly(alkylcyanoacrylate) nanoparticles. Fifteen years ago, poly(isobutylcyanoacrylate) nanocapsules were found to be able to encapsulate insulin and to increase its activity after oral administration as assessed by a reduction of glycemia [62]. Several aspects of this phenomenon are surprising: encapsulation of a hydrophilic drug in the oily core of nanocapsules; reduction of glycemia was only obtained with diabetic animals; and reduction of the glycemia occurred 2 days after a single administration and was maintained for up to 20 days depending on the insulin doses, although the amplitude of the pharmacological effect (minimum level of blood glucose) did not depend on the insulin dose. Damge´ et al. [196] and Lowe and Temple [200] suggested that nanocapsules could protect insulin from proteolytic degradation in intestinal fluids, based on the protection of encapsulated insulin, observed in the presence of different proteolytic enzymes in vitro. The capacity of insulin nanocapsules to reduce glycemia could be explained by their translocation through the intestinal barrier, as suggested by Damge´ et al. [196], for example by paracellular pathway or via M-cells in Peyer’s patches [201]. Recently, the use of Texas Red威labeled insulin allowed this translocation to be visualized more readily [202]. One hour after oral administration, nanocapsules reached the ileum. The presence of fluorescent areas within the mucosa and even in the lamina propria suggested that insulin-loaded nanocapsules could cross the intestinal epithelium. Although this passage is certainly an important factor, it does not explain the duration of the hypoglycemia. This prolonged action could be due to the retention of a part of the colloidal system in the gastrointestinal tract, but this mechanism is still not yet ellucidated. Interestingly, a prolonged hypoglycemic effect was also observed with insulin entrapped in poly(alkylcyanoacrylate) nanospheres only when these nanospheres were dispersed in an oily phase containing surfactant [203]. This suggests that some components of the nanocapsules could act as promoters of absorption. Recently, insulin has also been encapsulated in water-containing nanocapsules [67]. These nanocapsules dispersed in a biocompatible microemulsion could facilitate the intestinal absorption of the encapsulated peptide after oral administration, as suggested by the reduced blood glucose level highlighted in diabetic rats [204]. Damge´ et al. [63] showed that the incorporation of another peptide, octreotide (a somatostatin analog), in oil-containing poly(alkylcyanoacrylate) nanocapsules also improved and prolonged the therapeutic effect of this peptide after administration by the oral route. Calcitonin has been encapsulated in both oil-containing and water-containing nanocapsules [65,200,205]. Calcitonin-loaded oil-containing nanocapsules showed similar behavior to the insulin-loaded oil-containing nanocapsules [200]. Using water-containing nanocapsules, the effectiveness of the encapsulated peptide after oral administration to rats was estimated at 45% of the activity after intravenous administration of the same dose, whereas an absolute bioavailability of 40% was observed [65]. Even if the main limitation to oral administration of poly(alkylcyanoacrylate) nanoparticles is that their passage through the intestinal barrier is probably restricted and sometimes erratic, they represent an interesting tool for oral delivery of antigens. Indeed, it is usually accepted
Nanoparticulate Drug Carriers
583
that limited doses of antigen are sufficient for a mucous immunization. In fact, oral delivery of antigens may be considered as the most convenient means of producing an IgA antibody response. However, it is considerably limited by enzymatic degradation of antigens in the gastrointestinal tract and additionally by their poor absorption. Thus it has been postulated that the use of micro- or nanoparticles for the oral delivery of antigens should be efficient if those systems are able to achieve the protection of the antigenic molecule. Poly(alkylcyanoacrylate) nanospheres have been shown to enhance the secretory immune response after their oral administration in association with ovalbumin [206]. This result was not fully reproduced in the case of poly(acrylamide) nanoparticles loaded with the same antigen. It was postulated that in the case of poly(acrylamide) nanospheres much of the antigen was located at the surface of the polymer and could have been degraded during its passage through the gut. The relatively high surface concentrations of ovalbumin adsorbed onto poly(butylcyanoacrylate) nanospheres may have reduced the ability of the proteolytic enzymes in the gut to gain access to and to degrade the antigen, resulting in a greater antigen availability. 2. Bioadhesive Nanoparticles Nanoparticles have a certain ability to interact with mucosal surfaces [207]. Nonspecific bioadhesive particulate systems interact with the intestinal membrane through physicochemical interactions. Some polymers, either of natural or synthetic origin, have the ability to adhere to wet mucosal surfaces by means of hydrogen bonding or van der Waals forces [208]. With swellable hydrophilic polymers, adhesion is optimal when the mucosal contact is made with the dry polymer. Furthermore, the progressive hydration of the polymer leads to the formation of a hydrogel which is responsible for the development of a considerable mucosa adhesion strength [208]. However, in the case of colloidal particles, bioadhesion was achieved with nonswellable polymers such as poly(alkylcyanoacrylate) and this is mainly due to the inherent tendency of these small particles to develop intimate contact on large mucosal sites [209]. The strategy used to improve the bioadhesive properties of the nanoparticles consists in the modification of the nanoparticle surface properties using various polymers such as poloxamers and poly(vinyl alcohol) [207,209,210]. Other authors suggested grafting lectines on the nanoparticle surface [209]. Recently, Yang et al. [54] proposed a new way to synthestize bioadhesive poly(alkylcyanoacrylate) nanoparticles coated with chitosan. Bioadhesion was tested in vivo. After peroral administration of radiolabeled poly(hexylcyanoacrylate) nanoparticles to mice, whole body autoradiography showed that 30 min after administration the particles were exclusively localized in the stomach [210]. After 4 h, a large quantity of radioactivity was found in the intestine in the form of clusters without macroradiographic evidence of accumulation at specific intestinal sites. On the contrary, a persistent film of nanoparticles adhering to the stomach wall was observed. In this study, very little of the radioactivity was found to be absorbed. In a similar study, microautoradiographs confirmed the presence of radioactivity throughout the whole gut [211,212]. The amount of radioactivity dropped to 30–40% of the 90-min value within 4–8h and to 5% 24 h after dosing. Histological investigation showed radioactivity adjacent to the brush border, incorporated into the underlying cell layers and in goblet cells up to 6 days after administration. However, the exclusive use of a radioactive tracer in these experiments makes the presence of physically intact particles 6 days after administration questionable because of a possible degradation of the particles in the gastrointestinal tract. The pharmacokinetics of several drugs after oral administration have been improved by means of nanoparticles. Most of those studies have been carried out with conventional formulations, which means that the carriers were generally not specifically designed for improving the bioadhesion performances of the nanoparticles. For instance, the bioavailability of vincamine
584
Vauthier et al.
was about 25% when administered in an aqueous solution to rabbits, but it reached 40% with oral administration of vincamine adsorbed on poly(hexylcyanoacrylate) nanospheres. This increase of bioavailability was probably due to a prolonged period of contact of the drug delivery system with the mucosa [213]. Nanocapsules of poly(isobutyl cyanoacrylate) increased the bioavailability of iodine after administration of lipiodol, an iodized oil, to the jejunum of dogs [214]. With the nanocapsules, the blood level of iodine was prolonged from 75 to over 105 min. This was attributed to a prolonged time period of contact between the lipiodol drug and the mucus over the microvilli membrane [214]. Darodipine, a calcium flux inhibitor which causes strong vasodilatation, is characterized by a rather short half-life ranging from 2 to 4 h. The use of nanocapsules made it possible to reduce the intensity of the initial very strong hypotensive effect it produces and to prolong the pharmacological activity of the drug [215]. In the case of indomethacin, the side effects, i.e., stomach ulcerations, were reduced when the drug was delivered by nanocapsules compared to the free drug [216]. 3. Cyclodextrins Containing Nanoparticles Cyclodextrins were proposed in order to enhance the association of hydrophobic drugs with poly(alkylcyanoacrylate) nanospheres during the preparation process by emulsion polymerization [48]. The use of cyclodextrin–drug complexes to load the nanospheres with the drug lead to very encouraging results when the formation of the complex increased the apparent solubility of the drug in the preparation medium. Using a cyclodextrin–saquinavir complex, the drug loading of the nanospheres was increased 20-fold compared to nanospheres prepared in the absence of cyclodextrins. In this case, it was found that a large amount of the cyclodextrin remained associated with the nanospheres. These nanospheres are an interesting system for oral application, as very recently shown by Boudad et al. [217]. Indeed, they were able to improve significantly the amount and the kinetics of saquinavir transported from apical to basolateral site in the CACO-2 monolayer cell model. D. Ocular Delivery The anatomical structure and the protective physiological process of the eye make ocular drug delivery difficult. This is the reason why conventional ocular dosage forms exhibit extremely low bioavailability. Limited absorption of the drug through the lipophilic corneal barrier is mainly due to short precorneal residence time due to the tear turnover, rapid nasolacrimal drainage of instilled drug from the tear fluid, and nonproductive absorption through the conjunctiva. Only a small proportion (1 to 3 %) of the applied drug penetrates the cornea and reaches intraocular tissues. For these reasons, it is necessary to develop efficient and more acceptable ocular therapeutic systems. Different strategies can be carried out to improve the precorneal residence time and/or penetration ability of the active ingredient. First, studies of antiglaucomatous agents such as betaxolol, carteolol, and metipranolol encapsulated in nanocapsules only showed a reduction of the noncorneal absorption (systemic circulation), leading to reduced side effects as compared with the free drug [218–220]. These systemic side effects are due to a poor ocular retention of drugs which are directly absorbed into the systemic circulation by conjunctival and nasal blood vessels. In two cases (carteolol and betaxolol), encapsulation in nanocapsules produced an improved pharmacological effect (reduction of intraocular pressure) compared to free drug and unloaded nanocapsules (although the penetration of nanocapsules was not tested) and reduced cardiovascular systemic side effects [218,219]. Metipranolol showed the same activity alone and associated with nanocapsules, but, as in the case of carteolol and betaxolol, its side effects were reduced. When betaxolol was used, the nature of the polymer making up the nanocapsule
Nanoparticulate Drug Carriers
585
wall was found to be important. Indeed, betaxolol-loaded poly(ε-caprolactone) nanocapsules were more efficient than betaxolol-loaded poly(isobutylcyanoacrylate) or poly(lactide-co-glycolide) nanocapsules [218]. Calvo et al. [221] explored the mechanisms of interaction of nanocapsules with ocular tissues to better understand the pharmacological responses obtained with antiglaucomatous agents. By confocal microscopy, they showed that poly(ε-caprolactone) nanocapsules could specifically penetrate the corneal epithelium by an endocytic process without causing any damage to the cells. In contrast, the uptake of poly(isobutylcyanoacrylate) nanocapsules was associated with cellular lysis [222]. These results explained the improved therapeutic effect and the reduction of systemic side effects as a result of drug loss through the conjunctiva provided by poly(ε-caprolactone) nanocapsules by increasing corneal epithelium penetration of lipophilic drugs. Calvo et al. [223] also excluded the influence of the oily inner structure in the activity of the nanocapsules in the light of the absence of differences in penetration between nanospheres and nanocapsules. This is in contrast with a study by Marchal-Heussler et al. [219], who observed a better therapeutic effect with nanocapsules than with nanospheres. Moreover, Calvo et al. [223] demonstrated with indomethacin-loaded nanocapsules that the colloidal nature of the carrier was the main factor influencing its ocular bioavailability. These authors have also investigated the influence of the nature and the charge of the surface of nanocapsules on their physical stability and on their ocular bioavailability on the encapsulated drug [224,225]. They found that coating the negatively charged surface of poly(ε-caprolactone) nanocapsules with cationic polymers could prevent their degradation caused by the adsorption of lysozyme, a positively charged enzyme found in tear fluid [224]. They also pointed out that among the cationic polymers, chitosan adsorbed on the nanocapsule surface was able to provide the best corneal drug penetration without any local intolerance. This was achieved by a combination of effects: penetration of particles into the corneal epithelial cells, mucoadhesion of positively charged particles onto negatively charged membranes, and a specific effect on the tight junctions [224]. This effect of improvement of ocular absorption was also reported with the immunosuppressive peptide cyclosporin A [226]. The corneal level of the drug was increased five-fold as compared with an oily solution of the drug, owing to a highly loaded nanocapsule preparation also containing poly(ε-caprolactone). The efficacy of this topical formulation has been observed on a penetrating keratoplasty rejection model in the rat [227]. The major drawback with these systems is that they did not provide significant cyclosporin A at the ocular mucosae for an extended period of time [226]. Thus, cyclosporin A has been associated with chitosan nanospheres to investigate the potential of this new carrier for delivering the drug to the outer ocular structure [228]. With this system, therapeutic concentration of cyclosporin A could be maintained for at least 48 h in the external tissue (cornea and conjunctiva), while negligible or undetectable cyclosporin A levels were measured in inner ocular structures (iris, ciliary body, and aqueous humour), blood, and plasma. These results suggest that chitosan nanospheres may represent an interesting carrier to enhance the therapeutic index of cyclosporin A and other active drugs at the extraocular level. Le Bourlais et al. [229] proposed another approach by preparing cyclosporin A–loaded nanocapsules based on poly(alkylcyanoacrylate) and dispersing them in a poly(acrylic acid) gel. This system was also able to promote absorption of the drug, and it drastically reduced the toxicity of poly(alkylcyanocrylates) on the cornea. Based on the same approach, pilocarpineloaded poly(isobutylcyanoacrylate) nanocapsules have been dispersed in a pluronic gel [230]. Despite the low level of pilocarpin encapsulation, an increase in the intensity and duration of the ocular hypotensive effect of the drug, could be measured and the ocular bioavailability of the drug was also improved. In both cases, the gel used for dispersing the nanocapsules probably plays a major role for the retention of the nanoparticles at the cornea level. The gel also improved the corneal tolerance of the nanocapsules made of poly(alkylcyanoacrylate). Recently, Fresta et
586
Vauthier et al.
al. [231] showed that the tolerability of poly(alkylcyanoacrylate) nanospheres could be augmented using nanospheres coated with poly(ethylene glycol). To explain the increase tolerability and bioavailability of the encapsulated drug acyclovir measured with these nanospheres, it has been postulated that the outer poly(ethylene glycol) chains could hide the poly(alkylcyanoacrylate) from the ocular mucosa and improve the ocular mucoadhesion of the nanospheres, leading to a prolonged contact of the drug with the absorbing tissue.
IV. CONCLUSION As shown in this chapter, polymer chemistry and knowledge of polymer physicochemistry have allowed the development of many types of nanoparticles made of biodegradable polymers and suitable for drug delivery. Efficient encapsulation of various drugs can be achieved considering the physicochemical properties of both the drug and the polymer forming the nanoparticles. These systems have been shown to improve the drug activity and bioavailability by different routes of administration. Their potential for the treatment of cancer and of intracellular infections after intravenous administration is now well established. The debate over whether or not nanoparticular systems are a suitable carrier for the delivery of drugs by the oral route remains open, whereas for the ocular route further investigations are required, especially to strengthen the very recent promising results reported on the approach which combined the advantages of two technologies consisting of a gel in which drug-loaded nanoparticles are dispersed. A major problem with colloidal carriers encountered after intravenous administration is the nonspecific uptake by the phagocytes of the mononuclear phagocyte system. The ongoing research shows that the coating of nanoparticles with block copolymers is a powerful technique to achieve stability in the blood circulation and for less accumulation in the organs of the mononuclear phagocyte system. Indeed, new biodegradable polymers consisting of amphiphilic block copolymers were created to produce a novel generation of nanoparticulate drug carriers. For instance, results obtained with nanoparticles prepared with poly(ethylene glycol)–poly(alkylcyanoacrylate) copolymers and with poly(ethylene glycol)–poly(lactide) copolymers were remarkable in terms of the increase of the blood circulation time of those carriers. Coated nanoparticles with poly(ethylene glycol) chains also have interesting features for the delivery and the targeting of drugs to the brain. However, to fully achieve the drug targeting concept, a lot of work remains to be done. Indeed, it should be kept in mind that, from the point of view of targeting, these PEG-coated nanoparticles are simple and passive systems with no specific targeting ligands. They basically exploit both the difference in microvascular permeability between healthy and altered tissues and their long circulating properties. Thus the challenge to realize the targeting concept remains unresolved and opens new avenues for future work. A greater understanding of the different mechanisms of biological interactions, depending on the particle engineering method used, may be required for further advances. It also appears more and more clearly that it is essential to design the chemical structure of carriers rationally. A breakthrough may arise from biomimetic systems such as those involving the coating of the nanoparticles with polysaccharides and from the design and use of new block copolymers with various and fully controlled structures.
Nanoparticulate Drug Carriers
587
AKNOWLEDGMENTS The authors would like to thank Stella Ghouti for her help in preparing the manuscript.
REFERENCES
1. Oku N, Tokudome Y, Asai T, Tsukada H. Evaluation of drug targeting strategies and liposomal trafficking. Curr. Pharm. Design 2000; 6:1669–1691. 2. Gabizon AA. Stealth liposomes and tumor targeting: one step further in the quest for the magic bullet. Clin. Cancer Res 2001; 7:223–225. 3. Bendas G. Immunoliposomes: a promizing approach to targeting cancer therapy. Biodrugs 2001; 15:215–224. 4. De Jaeghere F, Doelker E, Gurny R. Nanoparticles. In: Mathiowitz E, ed. The Encyclopedia of Controlled Drug Delivery. New York: J. Wiley and Sons, 1999:641–664. 5. Vauthier C, Couvreur P. Development of polysaccharide nanoparticles as novel drug carrier systems. In: Wise DL, ed. Handbook of Pharmaceutical Controlled Release Technology. New York: Marcel Dekker, 2000:413–429. 6. Kumar NMVR. Nano and microparticles as controlled drug delivery devices. J. Pharm. Pharmaceut. Sci 2000; 3:234–258. 7. Soppimath KS, Aminabhavi TM, Kulkarni AR, Rudzinski WE. Biodegradable polymeric nanoparticles as drug delivery devices. J. Control. Rel 2001; 70:1–20. 8. Fattal E, Vauthier C. Nanoparticles as drug delivery systems. In: Swarbrick J, Boylan JC, eds. Encyclopedia of Pharmaceutical Technology. New York: Marcel Dekker, 2002:1874–1892. 9. Couvreur P, Barratt G, Fattal E, Legrand P, Vauthier C. Nanocapsule technology: a review. Crit. Rev. Drug Del. Systems 2002; 19:99–134. 10. Vauthier C, Couvreur P. Biodegradation of poly(alkylcyanoacrylates). In: Matsumara JP, Steinbuchel A, eds. Handbook of Biopolymers. Miscellanous Biopolymers and Biodegradation of Synthetic Polymers. Vol. Vol 9. New York: Wiley-VHC, 2002. 11. Brigger I, Dubernet C, Couvreur P. Nanoparticles in cancer therapy and diagnosis. Adv. Drug. Deliv. Rev 2002; 54:631–651. 12. Kataoka K, Harada A, Nagasaki Y. Block copolymer micelles for drug delivery: design, characterization and biological significance. Adv. Drug Del. Rev 2001; 47:113–131. 13. Ro¨sler A, Vandermeulen GWM, Klok HA. Advanced drug delivery devices via self-assembly of amphiphilic block copolymers. Adv. Drug Del. Rev 2001; 53:95–108. 14. Couvreur P, Grislain L, Lenaerts V, Brasseur F, Guiot P, Biornacki A. Biodegradable polymeric nanoparticles as drug carrier for antitumor agents. In: Guiot P, Couvreur P, eds. Polymeric Nanoparticles and Microparticles. FL: Boca Raton, 1986:27–93. 15. Florence AT, Hussain N. Transcytosis of nanoparticle and dendrimer delivery systems: evolving vistas. Adv. Drug Del. Rev 2001; 50:S69–S89. 16. Kreuter J. Nanoparticles. In: Kreuter J, ed. Colloidal Drug Delivery Systems. New-York: Marcel Dekker, 1994:219–342. 17. Kawashima Y. Nanoparticulate systems for improved drug delivery. Adv. Drug Del. Rev 2001; 47: 1–2. 18. Birrenbach G, Speiser P. Polymerized micelles and their use as adjuvants in immunology. J. Pharm. Sci 1976; 65:1763–1766. 19. Vauthier C, Seiller M, Weingarten C, Couvreur P. Contribution to the development of oral dosage form for insulin delivery. STP Pharma Sciences 1999; 9:391–396. 20. Gref R, Minamitake Y, Peracchia M-T, Trubetskoy V, Torchilin V, Langer R. Biodegradable longcirculating polymeric nanospheres. Science 1994; 263:1600–1603.
588
Vauthier et al.
21. Landfester K. Polyreactions in miniemulsions. Macromol. Rapid. Commun 2001; 22:896–936. 22. Lowell PA, El-Aasser MS. Emulsion polymerization and emulsion polymers. In: Lowell PA, ElAasser MS, eds. New York: Wiley, 1997:801. 23. Couvreur P, Kante B, Roland M, Guiot P, Baudhuin P, Speiser P. Poly(cyanoacrylate) nanoparticles as potential lysosomotropic carriers: preparation, morphological and sorptive properties. J. Pharm. Pharmacol 1979; 31:331–332. 24. Choi YK, Bae YH, Kim SW. Block copolymer nanoparticles of ethylene oxide and isobutylcyanoacrylate. Macromolecules 1995; 28:8419–8421. 25. Bazile D, Prud’homme C, Bassoulet M-T, Marlard M, Spenlehauer G, Veillard M. Stealth MePEG-PLA nanoparticles avoid uptake by the mononuclear phagocytes system. J. Pharm. Sci 1995; 84:493–498. 26. Peracchia MT, Desmae¨le D, Couvreur P, D’Angelo J. Synthesis of a novel poly(PEG-cyanoacrylateco-alkylcyanoacrylate) amphiphilic copolymer for the development of ‘‘stealth’’ PEG-coated nanoparticles. Macromolecules 1997; 30:846–851. 27. Vauthier-Holtzscherer C, Benabbou S, Spenlehauer G, Veillard M, Couvreur P. Methodology for the preparation of ultradispersed polymer systems. STP Pharma Sci 1991; 1:109–116. 28. Odian G. Principles of Polymerization. 3rd Ed.. New York: J. Wiley and Sons, 1991. 29. Pichot C, Delair T. Functional nanospheres by emulsion polymerization. In: Arshady R, ed. Microspheres, Microcapsules and Liposomes. London: Citus Book, 1999:125–164. 30. Charles M-H, Charreyre M-T, Delair T, Elaı¨ssari A, Pichot C. Oligonucleotides–polymer nanoparticle conjugates: diagnostic applications. STP Pharma Sci 2001; 11:251–263. 31. Kreuter J, Speiser P. New adjuvants on a poly(methylmethacrylate) base. Infec. Immunol 1976; 13: 204–210. 32. Costa G, Cronin JP, Pepper DC, Loonan C. Termination and transfer by acids in the pyridineinitiated polymerization of butyl cyanoacrylate. Eur. Polym. J 1983; 19:939–945. 33. Douglas SJ, Illum L, Davis SS. Particle size and size distribution of poly(butyl 2-cyanoacrylate) nanoparticles. II. Influence of stabilizers. J. Colloid Interface Sci 1985; 103:154–163. 34. Nehan N, Birkinshaw C. Clarke. Poly n-butyl cyanoacrylate nanoparticles: a mechanistic study of polymerization and particle formation. Biomaterials 2001; 22(11):1335–1344. 35. Peracchia MT, Vauthier C, Puisieux F, Couvreur P. Development of sterically stabilized poly(isobutyl 2-cyanoacrylate) nanoparticles by chemical coupling of poly(ethylene glycol). J. Biomed. Mater. Res 1997; 34:317–326. 36. Peracchia MT, Vauthier C, Popa M, Puisieux F, Couvreur P. Investigation on the formation of sterically stabilized poly(ethylene glycol/isobutylcyanoacrylate) nanoparticles by chemical grafting of poly(ethylene glycol) during polymerization of isobutylcyanoacrylate. STP Pharma Sci 1997; 7: 513–520. 37. Vansnick L, Couvreur P, Christiaens-Ley D, Roland M. Molecular weights of free and drug-loaded nanoparticles. Pharm. Res 1985; 1:36–41. 38. Gallardo MM, Roblot-Treupel L, Mahuteau J, Genin I, Couvreur P, Plat M, Puisieux F. Nanocapsules et nanospheres d’alkylcyanoacrylate. Interaction principe actif-polyme`re. Proceedings of the 5th International Congress of Pharmaceutical Technology, APGI, Paris, France, 1989. 39. Guize V, Drouin JY, Benoit J, Mahuteau J, Dumont P, Couvreur P. Vidarabine-loaded nanoparticles: a physicochemical study. Pharm. Res 1990; 7:736–741. 40. Grangier JL, Puygrenier M, Gautier JC, Couvreur P. Nanoparticles as carriers for growth hormone releasing factor. J. Control. Rel 1991; 15:3–13. 41. Seijo B, Fattal E, Roblot-Treupel L, Couvreur P. Design of nanoparticles of less than 50 nm in diameter. Preparation, characteriation and drug loading. Int. J. Pharm 1990; 62:1–7. 42. Alonso MJ, Sanchez A, Torres D, Seijo B, Vila-Jato JL. Joint effect of monomer and stabilizer concentrations on the physico-chemical characteristics of poly(butyl-2-cyanoacrylate) nanoparticles. J. Microencapsulation 1990; 7:517–526. 43. Lenaerts V, Raymond P, Juhasz J, Simard MA, Jolicoeur C. New method for the preparation of cyanoacrylic nanoparticles with improved colloidal properties. J. Pharm. Sci 1989; 78:1051–1052.
Nanoparticulate Drug Carriers
589
44. Lescure F, Zimmer C, Roy D, Couvreur P. Optimization of polycyanoacrylate nanoparticle preparation: influence of sulfur dioxide and pH on nanoparticle characteristics. J. Coll. Interface Sci 1992; 154:77–86. 45. Verdun C, Couvreur P, Vranckx H, Lenaerts V, Roland M. Development of a nanoparticle controlled release formulation for human use. J. Control. Rel 1986; 3:205–210. 46. Sommerfeld P, Schroeder U, Sabel BA. Long-term stability of PBCA nanoparticles suspensions suggests clinical usefulness. Int. J. Pharm 1997; 155:201–207. 47. Sommerfeld P, Sabel BA, Schroeder U. Sterilization of unloaded polybutylcyanoarylate nanoparticles. Int. J. Pharm 1998; 164:113–118. 48. Monza da Silveira A, Ponchel G, Puisieux F, Ducheˆne D. Combined poly(isobutylcyanoacrylate) and cyclodextrins nanoparticles for enhancing the encapsulation of lipophilic drugs. Pharm. Res 1998; 15:1051–1055. 49. Ducheˆne D, Wouessidjewe D, Ponchel G. Cyclodextrins and carrier systems. J. Control. Rel 1999; 62:263–268. 50. Boudad H, Legrand P, Lebas G, Cheron M, Ducheˆne D, Ponchel G. Combined hydroxypropylbeta-cyclodextrin and poly(alkylcyanoacrylate) nanoparticles intended for oral administration of saquinavir. Int. J. Pharm 2001; 218:113–124. 51. Labib A, Lenaerts V, Chouinard F, Leroux JC, Ouellet R, Lier JE. Biodegradable nanospheres containing phthalocyanines and naphthalocyanines for targeted photodynamic tumor therapy. Pharm. Res 1991; 8:1027–1031. 52. Ghanem GE, Joubran C, Arnould R, Lejeune F, Fruhling J. Labelled polycyanoacrylate nanoparticles for human in vivo use. Appl. Radiat. Isot 1993; 44:1219–1224. 53. Douglas SJ, Illum L, Davis SS. Poly(butyl-2-cyanoacrylate) nanoparticles with differing surface charges. J. Control Rel 1986; 3:15–23. 54. Yang SC, Ge HX, Jiang XQ, Yang CZ. Formation of positively charged poly(butyl cyanoacrylate) nanoparticles stabilized with chitosan. Colloid Polym. Sci 2000; 278:285–292. 55. Al Khoury-Fallouh N, Roblot-Treupel L, Fessi H, Devissaguet JP, Puisieux F. Development of a new process for the manufacture of poly(isobutylcyanoacrylate) nanocapsules. Int. J. Pharm 1986; 28:125–136. 56. Chouinard F, Kan FWF, Leroux JC, Foucher C, Lenaerts V. Preparation and purification of polyisohexylcyanoacrylate nanocapsules. Int. J. Pharm 1991; 71:211–217. 57. Fresta M, Cavallaro G, Giammona G, Wehrli E, Puglisi G. Preparation and characterization of polyethyl-2-cyanoacrylate nanocapsules containing antiepileptic drugs. Biomaterials 1996; 17: 751–758. 58. Gallardo MM, Couarraze G, Denizot B, Treupel L, Couvreur P, Puisieux F. Preparation and purification of isohexylcyanoacrylate nanocapsules. Int. J. Pharm 1993; 100:55–64. 59. Aboubakar M, Puisieux F, Couvreur P, Deyme M, Vauthier C. Study of the mechanism of insulin encapsulation in poly(isobutylcyanoacrylate) nanocapsules obtained by interfacial polymerization. J. Biomed. Mater. Res 1999; 47:568–576. 60. Wohlgemuth M, Ma¨chtle W, Mayer C. Improved preparation and physical studes of polybutylcyanoacrylate nanocapsules. J. Microencapsulation 2000; 17:437–448. 61. Puglisi G, Fresta M, Giammona G, Venture CA. Influence of the preparation conditions in poly(ethylcyanoacrylate) nanocapsules formation. Int. J. Pharm 1995; 125:283–287. 62. Damge´ C, Michel C, Aprahamiam M, Couvreur P. New approach for oral administration of insulin with polyalkylcyanoacrylate nanocapsules as drug carrier. Diabetes 1988; 37:246–251. 63. Damge´ C, Vonderscher J, Marbach P, Pinget M. Poly(alkyl cyanoacrylate) nanocapsules as a delivery system in the rat for octreotide, a long-acting somatostatin analogue. J. Pharm. Pharmacol 1997; 49:949–954. 64. Gasco M, Trotta M. Nanoparticles from microemulsions. Int. J. Pharm 1986; 29:267–268. 65. Vranckx H, Demoustier M, Deleers M. A new nanocapsule formulation with hydrophilic core: application to the oral administration of salmon calcitonin in rats. Aur. J. Pharm. Pharmacol 1996; 42:345–347.
590
Vauthier et al.
66. Lambert G, Fattal E, Pinto-Alphandary H, Gulik A, Couvreur P. Polyisobutylcyanoacrylate nanocapsules containing an aqueous core as a novel colloidal carrier for the delivery of oligonucleotides. Pharm. Res 2000; 17:707–714. 67. Watnasirichaikul S, Davies NM, Rades R, Tucker IG. Preparation of biodegradable insulin nanocapsules from biocompatible microemulsions. Pharm. Res 2000; 17:684–689. 68. De Keyser JL, Poupaert JH, Dumont P. Poly(diethyl methylidenemalonate) nanoparticles as a potential drug carrier: preparation, distribution and elimination after intravenous and peroral administration to mice. J. Pharm. Sci 1991; 80:67–70. 69. Breton P, Guillon X, Roy D, Lescure F, Riess G, Bru N, Roques-Carmes C. Physico-chemical characterization, preparation and performance of poly(methylidene malonate 2.1.2) nanoparticles. Biomaterials 1998; 19:271–281. 70. Lescure F, Seguin C, Breton P, Bourrinet P, Roy D, Couvreur P. Preparation and characterization of novel poly(methylidene malonate 2.1.2.)-made nanoparticles. Pharm. Res 1994; 11:1270–1277. 71. Roy D, Guillon X, Lescure F, Couvreur P, Bru N, Breton P. On shelf stability of freeze-dried poly(methylidene malonate 2.1.2) nanoparticles. Int. J. Pharm 1997; 148:165–175. 72. De´sire´ L, Mysiakine E, Bonnafous D, Sogodira S, Breton P, Bru-Magniez N, Couvreur P. Delivery of cell communication factors with VL4001, a (methylidene malonate)-based composition. Proceedings of the 29th Annual Meeting of the Controlled Release Society, Seoul, Korea, 2002. 73. Kissel T, Li Y, Unger F. ABA-triblock copolymers from biodegradable polyester A-blocks and hydrophilic poly(ethylene oxide) B-blocks as a candidate for in situ forming hydrogel delivery systems for proteins. Adv. Drug Del. Rev 2002; 54:99–134. 74. Peracchia MT, Desmae¨le D, Vauthier C, Labarre D, Fattal E, D’Angelo J, Couvreur P. Development of novel technologies for the synthesis of biodegradable pegylated nanoparticles. In: Gregoriadis B, McCormack B, eds. Targeting of Drugs. Strategies for Stealth Therapeutic Systems. Vol. 6. New York: Plenum Press, 1998:225–239. 75. Peracchia MT, Fattal E, Desmae¨le D, Besnard M, Noel JP, Gomis JM, Appel M, d’Angelo J, Couvreur P. Stealth PEGylated polycyanoacrylate nanoparticles for intravenous administration and splenic targeting. J Control. Rel 1999; 60:121–128. 76. Jung T, Breitenbach A, Kissel T. Sulfobutylated poly(vinyl alcohol)-graft-poly(lactide-co-glycolide)s facilitate the preparation of small negatively charged biodegradable nanospheres. J. Control. Rel 2000; 67:157–169. 77. Yang L, Alexandridis P. Physicochemical aspects of drug delivery and release from polymer-based colloids. Curr Opin Coll. Interface Sci 2000; 5:132–143. 78. Kabanov AV, Lemieux P, Vinogradov S, Alakhov V. Pluronic威 block copolymers: novel functional molecules for gene therapy. Adv. Drug Del. Rev 2002; 54:223–233. 79. Passirani C, Ferrarini L, Barratt G, Devissaguet J-P, Labarre D. Preparation and characterization of nanoparticles bearing heparin or dextran covalently bound to poly(methyl methacrylate). J. Biomat. Sci. Polym Ed 1999; 10:47–62. 80. Passirani C, Barratt G, Devissaguet J-P, Labarre D. Interactions of nanoparticles bearing heparin or dextran covalently bound to poly(methyl methacrylate) with the complement system. Life Sciences 1998a; 62:775–785. 81. Passirani C, Barratt G, Devissaguet J-P, Labarre D. Long-circulating nanoparticles bearing heparin or dextran covalently bound to poly(methyl methacrylate). Pharm. Res 1998b; 15:1046–1050. 82. Chauvierre C, Vauthier C, Labarre D, Couvreur P. Copolyme`re a` structure se´quence´e compose´ d’un segment saccharidique lie´ a` au moins un segment hydrophobe bioe´rodable, et particules correspondantes. French Patent 00 14900, 17.11.2000, PCT no. FR01/03619, 2000. 83. Chauvierre C, Labarre D, Couvreur P, Vauthier C. Radical polymerization of alkylcyanoacrylate to produce polysaccharide-coated nanoparticles. Proceedings of the 4th World Meeting on Pharmaceutics, Biopharmaceutics, Pharmaceutical Technology, ADRITELF, APGI, APV, Florence, 8–11 April, 2002. 84. Rouzes C, Gref R, Le´onard M, Delgado A, Dellacherie E. Surface modification of polylactic nanopheres using hydrophobically modified dextrans as stabilizers in an o/w emulsion/evaporation technique. J. Biomed. Mater. Res 2000; 50:557–565.
Nanoparticulate Drug Carriers
591
85. Hamaide T, Pantiru M, Fessi H, Boullanger P. Ring opening polymerization of ε-caprolactone with monosaccharides as transfert agents. A novel route to functionalized nanoparticles. Macromol. Rapid Comm 2001; 22:659. 86. Gref R, Ponchel G, Ducheˆne D, Couvreur P. Mate´riaux a` base de polyme`res biode´gradables et son proce´de´ de pre´paration. French Patent 00 06232, 16.03.2000, WO no. FR01/01496, 2000. 87. Fessi H, Devissaguet JP, Puisieux F. Proce´de´ de pre´paration de syste`mes colloı¨daux dispersibles d’une substance sous forme de nanoparticules. French Patent 2.608.988, 1986. 88. Marty JJ, Oppenheim RC, Speiser P. Nanoparticles a new colloidal drug delivery systems. Pharm. Acta Helv 1978; 53:17–23. 89. Stainmesse S. Etude gale´nique d’un nouveau proce´de´ d’obtention de vecteurs colloı¨daux submicroniques a` partir d’une prote´ine ou d’un polyme`re synthe´tique, Ph.D. Thesis. France: Universite´ de Paris Sud, 1990. 90. Weber C, Coester C, Kreuter J, Langer K. Dessolvation process and surface characterization of protein nanoparticles. Int. J. Pharm 2000; 194:94–102. 91. Leong KW, Mao HQ, Truong-Le VL, Roy K, Walsh SM, August JT. DNA-polycation nanospheres as non-viral gene delivery vehicles. J. Control. Rel 1998; 53:183–193. 92. Truong-Le VL, Walsh SM, Schweibert E, Mao HQ, Guggino WB, August JT, Leong KW. Gene transfer by DNA-gelatin nanospheres. Arch. Biochem. Biophys 1999; 361:47–56. 93. Fessi H, Devissaguet JP, Puisieux F. Proce´de´ de pre´paration des syste`mes colloı¨daux dispersibles sous forme de nanocapsules. French Patent 86.18.444, 1986. 94. Legrand P, Barratt G, Mosqueira V, Fessi H, Devissaguet JP. Polymeric nanocapsules as drug delivery systems. A review. STP Pharma Sci 1999; 9:411–418. 95. Rajaonarivony M, Vauthier C, Couarraze G, Puisieux F, Couvreur P. Development of a new drug carrier made from alginate. J. Pharm. Sci 1993; 82:912–918. 96. Calvo P, Remunan-Lopez C, Vila-Jato JL, Alonso MJ. Novel hydrophilic chitosan–polyethylene oxide nanoparticles as protein carriers. J. Appl. Polym. Sci 1997; 63:125–132. 97. Janes KA, Calvo P, Alonso MJ. Polysaccharide colloidal particles as delivery systems for macromolecules. Adv. Drug Del. Rev 2001; 47:83–97. 98. Vinogradov SV, Bronich TK, Kabanov AV. Nanosized cationic hydrogels for drug delivery; preparation, properties and interactions with cells. Adv. Drug Del. Rev 2002; 54:135–147. 99. Mao HQ, Roy K, Walsh SM, August JT, Leong KW. Chitosan-DNA nanospheres for gene delivery. Proc. Intern. Symp. Control. Rel. Bioact. Mater 1996; 23:401–402. 100. Mao HQ, Roy K, Troung-Le VL, Janes KA, Lin KY, Wang Y, August JT, Leong KW. Chitosan-DNA nanoparticles as gene carriers: synthesis, characterization and transfection efficiency. J. Control. Rel 2001; 70:399–421. 101. Kircheis R, Wightman L, Wagner E. Design and gene delivery activity of modified polyethyleneimines. Adv. Drug Del. Rev 2001; 53:341–358. 102. Quintanar-Guerrero D, Alle´man E, Doelker E, Fessi H. Preparation and characterization of nanocapsules from preformed polymers by a new process based on emulsification-diffusion technique. Pharm. Res 1998; 15:1056–1062. 103. Quintanar-Guerrero D, Alle´man E, Fessi H, Doelker E. Preparation techniques and mechanisms of formation of biodegradable nanoparticles from preformed polymers. Drug Dev. Ind. Pharm 1998; 24:1113–1128. 104. Quintanar-Guerrero D, Alle´man E, Fessi H, Doelker E. Pseudolatex preparation using a novel emulsion-diffusion process involving direct displacement of partially water-miscible solvents by distillation. Int. J. Pharm 1999; 188:155–164. 105. Alle´man E, Doelker E, Gurny R. Preparation of aqueous polymeric nanodispersions by a reversible salting-out process, influence of process parameters on particle size. Int. J. Pharm 1992; 87:247–253. 106. Alle´man E, Doelker E, Gurny R. Drug loaded poly(lactic acid) nanoparticles produced by a reversible salting out process: purification of an injectable dosage form. Eur. J. Pharm. Biopharm 1993; 39: 13–18. 107. Gurny R, Peppas NA, Harrington DD, Banker GS. Development of biodegradable and injectable lattices for controlled release potent drugs. Drug Dev. Ind. Pharm 1981; 7:1–25.
592
Vauthier et al.
108. Vanderhoff JW, El-Aasser MS, Ugelstad J. Polymer emulsification process. U.S. Patent 4.177.177, 1979. 109. Venier-Julienne MC, Benoit JP. Preparation, purification and morphology of polymeric nanoparticles as drug carrier. Pharm. Acta Helvetica 1996; 71:121–128. 110. Feng S, Huang S. G Effects of emulsifiers on the controlled release of paclitaxel (Taxol) from nanospheres of biodegradable polymers. J. Control. Rel 2001; 71:53–69. 111. Zambaux MF, Bonneauz F, Gref R, Maincent P, Dellacherie E, Alonso MJ, Labrude P, Vigneron C. Influence of experimental parameters on the characteristics of poly(lactic acid) nanoparticles prepared by a double emulsion method. J. Control. Rel 1998; 50:31–40. 112. Blanco MD, Alonso MJ. Development and characterization of protein-loaded poly(lactide-co-glycolide) nanospheres. Eur. J. Pharm. Biopharm 1997; 43:287–294. 113. Berton M, Alle´mann E, Stein CA, Gurny R. Highly loaded nanoparticulate carrier using an hydrophobic antisense oligonucleotide complex. Eur. J. Pharm. Sci 1999; 9:163–170. 114. Mosqueira VCR, Legrand P, Pinto-Alphandary H, Puisieux F, Barratt B. Poly(D,L-lactide) nanocapsules prepared by a solvent displacement process: Influence of the composition on physicochemical and structural properties. J. Pharm. Sci 2000; 89:614–626. 115. Gasper MM, Blanco D, Cruz ME, Alonso MJ. Formulation of L-asparaginase-loaded poly(lacticco-glycolide) nanoparticles: influence of polymer properties on enzyme loading, activity and in vitro release. J. Control. Rel 1998; 2:53–62. 116. Li YP, Pei YY, Zhang XY, Gu ZH, Zhou ZH, Yuan WE, Zhou JJ, Zhu KH, Gao XJ. PEGylated PLGA nanoparticles as protein carriers: synsthesis, preparation and biodistribution in rats. J. Control Rel 2001; 71:203–211. 117. Delie F, Berton M, Alle´man E, Gurny R. Comparison of two metods of encapsulation of an oligonucleotide into poly(D,L-lactic acid) particles. Int. J. Pharm 2001; 214:25–30. 118. Li YP, Pei YY, Zhou ZH, Zhang XY, Gu ZH, Ding J, Zhou JJ, Gao XJ. PEGylated polycyanoacrylate nanoparticles as tumor necrosis factor-␣ carrier. J. Control Rel 2001; 71:287–296. 119. Sakuma S, Hayashi M, Akashi M. Design of nanoparticles composed of graft copolymers for oral peptide delivery. Adv. Drug Del. Rev 2001; 47:21–37. 120. Quintanar-Guerrero D, Alle´man E, Fessi H, Doelker E. Application of the ion-pair concept to hydrophilic substances with special emphasis on peptides. Pharm. Res 1997; 14:119–127. 121. Grislain L, Couvreur P, Lenaerts V, Roland M, Deprez-Decampeneere D, Speiser P. Pharmacokinetics and distribution of a biodegradable drug-carrier. Int. J. Pharm 1983; 15:335–345. 122. Lenaerts V, Nagelkerke JF, Van Berkel TJ, Couvreur P, Grislain L, Roland M, Speiser P. In vivo uptake of polyisobutyl cyanoacrylate nanoparticles by rat liver Kupffer, endothelial, and parenchymal cells. J. Pharm. Sci 1984; 73:980–982. 123. Lenaerts V, Couvreur P, Christiaens-Leyh D, Joiris E, Roland M, Rollman B, Speiser P. Degradation of poly(isobutyl cyanoacrylate) nanoparticles. Biomaterials 1984; 5:65–68. 124. Chavany C, Ledoan T, Couvreur P, Puisieux F, Helene C. Polyalkylcyanoacrylate nanoparticles as polymeric carriers for antisense oligonucleotides. Pharm. Res 1992; 9:441–449. 125. Pinto-Alphandary H, Andremont A, Couvreur P. Targeting delivery of antibiotics using liposomes and nanoparticles: research and applications. Int. J. Antimicrob. Agents 2000; 13:155–168. 126. Fattal E, Youssef M, Couvreur P, Andremont A. Treatment of experimental salmonellosis in mice with ampicillin-bound nanoparticles. Antimicrob. Agents Chemother 1989; 33:1540–1543. 127. Youssef M, Fattal E, Alonso MJ, Roblot-Treupel L, Sauzieres J, Tancrede C, Omnes A, Couvreur P, Andremont A. Effectiveness of nanoparticle-bound ampicillin in the treatment of Listeria monocytogenes infection in athymic nude mice. Antimicrob. Agents Chemother 1988; 32:1204–1207. 128. Forestier F, Gerrier P, Chaumard C, Quero AM, Couvreur P, Labarre C. Effect of nanoparticlebound ampicillin on the survival of Listeria monocytogenes in mouse peritoneal macrophages. J. Antimicrob. Chemother 1992; 30:173–179. 129. Balland O, Pinto-Alphandary H, Pecquet S, Andremont A, Couvreur P. The uptake of ampicillinloaded nanoparticles by murine macrophages infected with Salmonella typhimurium. J. Antimicrob. Chemother 1994; 33:509–522.
Nanoparticulate Drug Carriers
593
130. Pinto-Alphandary H, Balland O, Laurent M, Andremont A, Puisieux F, Couvreur P. Intracellular visualization of ampicillin-loaded nanoparticles in peritoneal macrophages infected in vitro with Salmonella typhimurium. Pharm. Res 1994; 11:38–46. 131. Buchmeier NA, Heffron F. Inhibition of macrophage phagosome–lysosome fusion by Salmonella typhimurium. Infect. Immun 1991; 59:2232–2238. 132. Page-Clisson ME, Pinto-Alphandary H, Chachaty E, Couvreur P, Andremont A. Drug targeting by polyalkylcyanoacrylate nanoparticles is not efficient against persistent Salmonella. Pharm. Res 1998; 15:544–549. 133. Zhang Q, Liao G, Wei D, Nagai T. Increase in gentamicin uptake by cultured mouse peritoneal macrophages and rat hepatocytes by its bioding to polybutylcyanoacrylate nanoparticles. Int. J. Pharm 1998; 164:21–27. 134. Gaspar R, Opperdoes FR, Preat V, Roland M. Drug targeting with polyalkylcyanoacrylate nanoparticles—in vitro activity of primaquine-loaded nanoparticles against intracellular Leishmania donovani. Ann Tropical Med Parasitology 1992; 86:41–49. 135. Gaspar R, Preat V, Opperdoes FR, Roland M. Macrophage activation by polymeric nanoparticles of polyalkylcyanoacrylates: activity against intracellular Leishmania donovani associated with hydrogen peroxide production. Pharm. Res 1992; 9:782–787. 136. Fouarge M, Dewulf M, Couvreur P, Roland M, Vranckx H. Development of dehydroemetine nanoparticles for the treatment of visceral leishmaniasis. J. Microencapsulation 1989; 6:29–34. 137. Verdun C, Brasseur F, Vranckx H, Couvreur P, Roland M. Tissue distribution of Doxorubicin associated with polyisohexylcyanoacrylate nanoparticles. Cancer Chemother. Pharmacol 1990; 26: 13–18. 138. Chiannilkulchai N, Ammoury N, Caillou B, Devissaguet JP, Couvreur P. Hepatic tissue distribution of doxorubicin-loaded nanoparticles after i.v. administration in reticulosarcoma M 5076 metastasis–bearing mice. Cancer Chemother. Pharmacol 1990; 26:122–126. 139. Brasseur F, Verdun C, Couvreur P, Deckers C, Roland M. Evaluation expe´rimentale de l’efficacite´ the´rapeutique de la doxorubicine associe´e associe´e aux nanoparticules de polyalkylcyanoacrylate. Proceedings of the 4th International Conference on Pharmaceutical Technology. 1986. APGI, Paris, 1986. 140. Chiannilkulchai N, Driouich Z, Benoit JP, Parodi AL, Couvreur P. Doxorubicin-loaded nanoparticles: increased efficiency in murine hepatic metastases. Sel. Cancer Ther 1989; 5:1–11. 141. Daemen T, Hofstede G, Ten Kate MT, Bakker-Woudenberg IA, Scherphof GL. Liposomal doxorubicin-induced toxicity: depletion and impairment of phagocytic activity of liver macrophages. Int. J. Cancer 1995; 61:716–721. 142. Fernandez-Urrusuno R, Fattal E, Porquet D, Feger J, Couvreur P. Evaluation of liver toxicological effects induced by polyalkylcyanoacrylate nanoparticles. Toxicol. Appl. Pharmacol 1995; 130: 272–279. 143. Fernandez-Urrusuno R, Fattal E, Rodrigues JM, Feger J, Bedossa P, Couvreur P. Effect of polymeric nanoparticle administration on the clearance activity of the mononuclear phagocyte system in mice. J. Biomed. Mater. Res 1996; 31:401–408. 144. Gibaud S, Andreux JP, Weingarten C, Renard M, Couvreur P. Increased bone marrow toxicity of doxorubicin bound to nanoparticles. Eur. J. Cancer 1994; 6:820–826. 145. Gibaud S, Rousseau C, Weingarten C, Favier R, Douay L, and P. Polyalkylcyanoacrylate nanoparticles as carriers for granulocyte–colony stimulating factor (G-CSF). J. Control. Rel 1998; 52: 131–139. 146. Demoy M, Gibaud S, Andreux JP, Weingarten C, Gouritin B, Couvreur P. Splenic trapping of nanoparticles: complementary approaches for in situ studies. Pharm. Res 1997; 14:463–468. 147. Kartner N, Evernden-Porelle D, Bradley G, Ling V. Detection of P-glycoprotein in multidrugresistant cell lines by monoclonal antibodies. Nature 1985; 316:820–823. 148. Treupel L, Poupon MF, Couvreur P, Puisieux F. Vectorization of doxorubicin in nanospheres and reversion of pleiotropic resistance of tumor cells (translation). C. R. Acad. Sci. III 1991; 313: 171–174.
594
Vauthier et al.
149. Cuvier C, Roblot-Treupel L, Millot JM, Lizard G, Chevillard S, Manfait M, Couvreur P, Poupon MF. Doxorubicin-loaded nanospheres bypass tumor cell multidrug resistance. Biochem. Pharmacol 1992; 44:509–517. 150. Bennis S, Chapey C, Couvreur P, Robert J. Enhanced cytotoxicity of doxorubicin encapsulated in polyisohexylcyanoacrylate nanospheres against multidrug-resistant tumour cells in culture. Eur. J. Cancer 1994; 1:89–93. 151. Nemati F, Dubernet C, Colin de Verdie`re A, Poupon MF, Acar L, Puisieux F, Couvreur P. Some parameters influencing cytotoxicity of free doxorubicin loaded nanoparticles in sensitive and multidrug resistant leucemic murine cells: incubation time, number of particles per cell. Int. J. Pharm 1994; 102:55–62. 152. Colin de Verdie`re A, Dubernet C, Nemati F, Poupon MF, Puisieux F, Couvreur P. Uptake of doxorubicin from loaded nanoparticles in multidrug-resistant leukemic murine cells. Cancer Chemother. Pharmacol 1994; 33:504–508. 153. Colin de Verdie`re A, Dubernet C, Nemati F, Soma E, Appel M, Ferte J, Bernard S, Puisieux F, Couvreur P. Reversion of multidrug resistance with polyalkylcyanoacrylate nanoparticles: towards a mechanism of action. Br. J. Cancer 1997; 76:198–205. 154. Pepin X, Attali L, Domrault C, Gallet S, Metreau JM, Reault Y, Cardot PJ, Imalalen M, Dubernet C, Soma E, Couvreur P. On the use of ion-pair chromatography to elucidate doxorubicin release mechanism from polyalkylcyanoacrylate nanoparticles at the cellular level. J. Chromatogr. B: Biomed. Sci. Appl 1997; 702:181–191. 155. Soma CE, Dubernet C, Barratt G, Nemati F, Appel M, Benita S, Couvreur P. Ability of doxorubicinloaded nanoparticles to overcome multidrug resistance of tumor cells after their capture by macrophages. Pharm. Res 1999; 16:1710–1716. 156. Soma CE, Dubernet C, Barratt G, Benita S, Couvreur P. Investigation of the role of macrophages on the cytotoxicity of doxorubicin and doxorubicin-loaded nanoparticles on M5076 cells in vitro. J. Control. Rel 2000; 68:283–289. 157. Morin C, Barratt G, Fessi H, Devissaguet JP, Puisieux F. Improved intracellular delivery of a muramyl dipeptide analog by means of nanocapsules. Int. J. Immunopharmacol 1994; 16:451–456. 158. Seyler I, Appel M, Devissaguet JP, Legrand P, Barratt G. Relationship between NO-synthase activity and TNF-alpha secretion in mouse macrophage lines stimulated by a muramyl peptide entrapped in nanocapsules. Int. J. Immunopharmacol 1996; 18:385–392. 159. Seyler I, Appel M, Devissaguet JP, Legrand P, Barratt G. Macrophage activation by a lipophilic derivative of Muramyldipeptide. J. Nanoparticle Res 1999; 1:91–97. 160. Barratt G, Puisieux F, Yu WP, Foucher C, Fessi H, Devissaguet JP. Anti-metastatic activity of MDP-L-alanyl-cholesterol incorporated into various types of nanocapsules. Int. J. Immunopharmacol 1994; 16:457–461. 161. Chavany C, Saison-Behmoaras T, Doan T, Puisieux F, Couvreur P, Helene C. Adsorption of oligonucleotides onto polyisohexylcyanoacrylate nanoparticles protects them against nucleases and increases their cellular uptake. Pharm. Res 1994; 11:1370–1378. 162. Fattal E, Vauthier C, Aynie I, Nakada Y, Lambert G, Malvy C, Couvreur P. Biodegradable polyalkylcyanoacrylate nanoparticles for the delivery of oligonucleotides. J. Control Rel 1998; 53:137–143. 163. Emile C, Bazile D, Herman F, Helene C, Veillard M. Me.PEG-PLA50 nanoparticles by compexation with strucutred oligopeptides. Drug Delivery 1996; 3:187–195. 164. Nakada Y, Fattal E, Foulquier M, Couvreur P. Pharmacokinetics and biodistribution of oligonucleotide adsorbed onto poly(isobutylcyanoacrylate) nanoparticles after intravenous administration in mice. Pharm. Res 1996; 13:38–43. 165. Schwab G, Chavany C, Duroux I, Goubin G, Lebeau J, Helene C, Saison-Behmoaras T. Antisense oligonucleotides adsorbed to polyalkylcyanoacrylate nanoparticles specifically inhibit mutated Haras-mediated cell proliferation and tumorigenicity in nude mice. Proc. Natl. Acad. Sci. USA 1994; 91:10460–10464. 166. Lambert G, Fattal E, Brehier A, Feger J, Couvreur P. Effect of polyisobutylcyanoacrylate nanoparticles and lipofectin loaded with oligonucleotides on cell viability and PKC alpha neosynthesis in HepG2 cells. Biochimie 1998; 80:969–976.
Nanoparticulate Drug Carriers
595
167. Lambert G, Fattal E, Pinto-Alphandary H, Gulik A, Couvreur P. Polyisobutylcyanoacrylate nanocapsules containing an aqueous core for the delivery of oligonucleotides. Int. J. Pharm 2001; 214: 13–16. 168. Lambert G, Bertrand JR, Fattal E, Subra F, Pinto-Alphandary H, Malvy C, Auclair C, Couvreur P. EWS fli-1 antisense nanocapsules inhibits Ewing sarcoma–related tumor in mice. Biochem. Biophys. Res. Commun 2000; 279:401–406. 169. Tanaka K, Iwakuma T, Harimaya K, Sato H, Iwamoto Y. EWS-Fli1 antisense oligodeoxynucleotide inhibits proliferation of human Ewing’s sarcoma and primitive neuroectodermal tumor cells. J. Clin. Invest 1997; 99:239–247. 170. Stein CA. Phosphorothioate antisense oligodeoxynucleotides: questions of specificity. Biotechnol 1996; 14:47–149. 171. Aynie I, Vauthier C, Chacun H, Fattal E, Couvreur P. Spongelike alginate nanoparticles as a new potential system for the delivery of antisense oligonucleotides. Antisense Nucleic Acid Drug Dev 1999; 9:301–312. 172. Kreuter J. Nanoparticulate systems for brain delivery of drugs. Adv. Drug Del. Rev 2001; 47:65–81. 173. Kreuter J, Alyautdin RN, Kharkevich DA, Ivanov AA. Passage of peptides through the blood–brain barrier with colloidal polymer particles (nanoparticles). Brain Res 1995; 674:171–174. 174. Alyautdin RN, Petrov VE, Langer K, Berthold A, Kharkevich DA, Kreuter J. Delivery of loperamide across the blood–brain barrier with polysorbate 80–coated polybutylcyanoacrylate nanoparticles. Pharm. Res 1997; 14:325–328. 175. Alyautdin RN, Tezikov EB, Ramge P, Kharkevich DA, Begley DJ, Kreuter J. Significant entry of tubocurarine into the brain of rats by adsorption to polysorbate 80–coated polybutylcyanoacrylate nanoparticles: an in situ brain perfusion study. J. Microencapsulation 1998; 15:67–74. 176. Friese A, Seiller E, Quack G, Lorenz B, Kreuter J. Increase of the duration of the anticonvulsive activity of a novel NMDA receptor antagonist using poly(butylcyanoacrylate) nanoparticles as a parenteral controlled release system. Eur. J. Pharm. Biopharm 2000; 49:103–109. 177. Sopala M, Schweizer S, Schafer N, Nurnberg E, Kreuter J, Seiller E, Danysz W. Neuroprotective activity of a nanoparticulate formulation of the glycineB site antagonist MRZ 2/576 in transient focal ischaemia in rats. Arzneimittelforschung 2002; 52:168–174. 178. Gulyaev AE, Gelperina SE, Skidan IN, Antropov AS, Kivman GY, Kreuter J. Significant transport of doxorubicin into the brain with polysorbate 80–coated nanoparticles. Pharm. Res 1999; 16: 1564–1569. 179. Gelperina SE, Khalansky AS, Skidan IN, Smirnova ZS, Bobruskin AI, Severin SE, Turowski B, Kreuter J. Toxicological studies of doxorubicin bound to polysorbate 80–coated poly(butyl cyanoacrylate) nanoparticles in healthy rats and rats with intracranial glioblastoma. Toxicol. Lett 2002; 126: 131–141. 180. Ramge P, Unger RE, Oltrogge JB, Zenker D, Begley D, Kreuter J, Briesen H. Polysorbate 80 coating enhances uptake of polybutylcyanoacrylate (PBCA) nanoparticles by human and bovine primary brain capillary endothelial cells. Eur. J. Neurosci 2000; 12:1931–1940. 181. Alyaudtin RN, Reichel A, Lobenberg R, Ramge P, Kreuter J, Begley DJ. Interaction of poly(butylcyanoacrylate) nanoparticles with the blood–brain barrier in vivo and in vitro. J. Drug Target 2001; 9:209–221. 182. Araujo L, Lobenberg R, Kreuter J. Influence of the surfactant concentration on the body distribution of nanoparticles. J. Drug Target 1999; 6:373–385. 183. Olivier JC, Fenart L, Chauvet R, Pariat C, Cecchelli R, Couet W. Indirect evidence that drug brain targeting using polysorbate 80–coated polybutylcyanoacrylate nanoparticles is related to toxicity. Pharm. Res 1999; 16:1836–1842. 184. Calvo P, Gouritin B, Chacun H, Desmaele D, D’Angelo J, Noel JP, Georgin D, Fattal E, Andreux JP, Couvreur P. Long-circulating PEGylated polycyanoacrylate nanoparticles as new drug carrier for brain delivery. Pharm. Res 2001; 18:1157–1166. 185. Calvo P, Gouritin B, Villarroya H, Eclancher F, Giannavola C, Klein C, Andreux JP, Couvreur P. Quantification and localization of PEGylated polycyanoacrylate nanoparticles in brain and spinal cord during experimental allergic encephalomyelitis in the rat. Eur. J. Neurosci 2002; 15:1317–1326.
596
Vauthier et al.
186. Calvo P, Gouritin B, Brigger I, Lasmezas C, Deslys J, Williams A, Andreux JP, Dormont D, Couvreur P. PEGylated polycyanoacrylate nanoparticles as vector for drug delivery in prion diseases. J. Neurosci. Methods 2001; 111:151–155. 187. Gautier JC, Grangier JL, Barbier A, Dupont P, Dussosoy D, Pastor G, Couvreur P. Biodegradable nanoparticles for subcutaneous adminstration of growth hormone releasing factor (hGRF). J. Control. Rel 1992; 3:205–210. 188. Kreuter J, Speiser PP. In vitro studies of poly(methyl methacrylate) adjuvants. J. Pharm. Sci 1976; 65:1624–1627. 189. Kreuter J. Possibilities of using nanoparticles as carriers for drugs and vaccines. J. Microencapsulation 1988; 5:115–127. 190. Kreuter J, Liehl E. Long-term studies of microencapsulated and adsorbed influenza vaccine nanoparticles. J. Pharm. Sci 1981; 70:367–371. 191. Kreuter J, Haenzel I. Mode of action of immunological adjuvants: some physicochemical factors influencing the effectivity of polyacrylic adjuvants. Infect. Immun 1978; 19:667–675. 192. Kreuter J, Berg U, Liehl E, Soliva M, Speiser PP. Influence of the particle size on the adjuvant effect of particulate polymeric adjuvants. Vaccine 1986; 4:125–129. 193. Kreuter J, Liehl E, Berg U, Soliva M, Speiser PP. Influence of hydrophobicity on the adjuvant effect of particulate polymeric adjuvants. Vaccine 1988; 6:253–256. 194. Jani P, Halbert GW, Langridge J, Florence AT. The uptake and translocation of latex nanospheres and microspheres after oral administration to rats. J. Pharm. Pharmacol 1989; 41:809–812. 195. Le Fevre HE, Joel DD, Shidlousky G. Retention of ingested latex particles in Peyers patches of gerin-free and conventional mice. Soc. Exp. Biol. Med 1985; 179:522–528. 196. Damge´ C, Michel C, Aprahamian M, Couvreur P, Devissaguet JP. Nanocapsules as carriers for oral peptide delivery. J. Control. Rel 1990; 13:233–239. 197. Aprahamian M, Michel C, Humbert W, Devissaguet JP, Damge C. Transmucosal passage of polyalkylcyanoacrylate nanocapsules as a new drug carrier in the small intestine. Biol. Cell 1987; 61: 69–76. 198. McClean S, Prosser E, Meehan E, O’Malley D, Clarke N, Ramtoola Z, Brayden D. Binding and uptake of biodegradable poly-DL-lactide micro and nanoparticles in intestinal epithelia. Eur. J. Pharm. Sci 1998; 6:153–163. 199. Desai MP, Labhasetwar V, Amidon GL, Levy RJ. Gastrointestinal uptake of biodegradable microparticles: effect of particle size. Pharm. Res 1996; 13:1838–1845. 200. Lowe PJ, Temple CS. Calcitonin and insulin in isobutylcyanoacrylate nanocapsules: protection against proteases and effect on intestinal absorption in rats. J. Pharm. Pharmacol 1994; 46:547–552. 201. Michel C, Aprahamian M, Defontaine L, Couvreur P, Damge C. The effect of site of administration in the gastrointestinal tract on the absorption of insulin from nanocapsules in diabetic rats. J. Pharm. Pharmacol 1990; 43:1–5. 202. Aboubakar M, Couvreur P, Pinto-Alphandary H, Gouritin B, Lacour B, Farinotti R, Puisieux F, Vauthier C. Insulin-loaded nanocapsules for oral administration: in vitro and in vivo investigation. Drug Dev. Res 2000; 49:109–117. 203. Damge´ C, Vranckx H, Balschmidt P, Couvreur P. Poly(alkyl cyanoacrylate) nanospheres for oral administration of insulin. J. Pharm. Sci 1997; 86:1403–1409. 204. Watnasirichaikul S, Rades T, Tucker IG, Davies NM. In-vitro release and oral bioavailability in diabetic rats of insulin using nanocapsules dispersed in biocompatible microemulsion. J. Pharm. Pharmacol 2002; 54:473–480. 205. Torres-Lugo M, Peppas NA. Transmucosal delivery systems for calcitonin: a review. Biomaterials 2000; 21:1191–1196. 206. O’Hagan DT, Palin K, Davis SS. Poly(butyl-2-cyanoacrylate) particles as adjuvant for oral immunization. Vaccine 1989; 7:213–216. 207. Takeuchi H, Yamamoto H, Kawashima Y. Mucoadhesive nanoparticulate systems for peptide drug delivery. Adv. Drug Deliv. Rev 2001; 47:39–54. 208. Ponchel G, Touchard F, Duchene D, Peppas NA. Bioadhesive analysis of the controlled-release systems. III. Fracture and interpenetration analysis in poly(acrylic acid) containing systems. J. Control. Rel 1987; 5:129–141.
Nanoparticulate Drug Carriers
597
209. Ponchel G, Irache J. Specific and non-specific bioadhesive particulate systems for oral delivery to the gastrointestinal tract. Adv. Drug Deliv. Rev 1998; 34:191–219. 210. Pimienta C, Chouinard F, Labib A, Lenaerts V. Effect of various poloxamer coatings on in vitro adhesion of isohexylcyanoacrylate nanospheres to rat ileal segments under liquid flow. Int. J. Pharm 1992; 80:1–8. 211. Lenaerts V, Couvreur P, Grislain L, Maincent P. Nanoparticles as a gastrointestinal drug delivery system. In: Lenaerts V, Gurny R, eds. Bioadhesive Drug Delivery Systems. Boca Raton. FL: CRC Press, 1990:94–104. 212. Kreuter J, Muller V, Munz K. Quantitative and microautoradiographic study on mouse intestinal distribution of polycyanoacrylate nanoparticles. Int. J. Pharm 1989; 55:39–45. 213. Maincent P, Le Verge R, Sado P, Couvreur P, Devissaguet JP. Disposition kinetics and oral bioavailability of vincamine-loaded polyalkyl cyanoacrylate nanoparticles. J.Pharm.Sci 1986; 75:955–958. 214. Damge C, Aprahamian M, Balboni G, Hoeltzel A, Andrieu V, Devissaguet JP. Polyalkylcyanoacrylate nanocapsules increase the intestinal absorption of a lipophilic drug. Int. J. Pharm 1987; 36: 121–125. 215. Hubert B, Atkinson J, Guerret M, Hoffman M, Devissaguet JP, Maincent P. The preparation and acute antihypertensive effects of a nanocapsular form of darodipine, a dihydropyridine calcium entry blocker. Pharm. Res 1991; 8:734–738. 216. Ammoury N, Fessi H, Devissaguet JP, Dubrasquet M, Benita S. Jejunal absorption, pharmacological activity, and pharmacokinetic evaluation of indomethacin-loaded poly(D,L-lactide) and poly(isobutyl-cyanoacrylate) nanocapsules in rats. Pharm. Res 1991; 8:101–105. 217. Boudad H, Legrand P, Coconnier MH, Appel M, Duchene D, Ponchel G. Absorption of saquinavir from combined hydroxypropyl  cyclodextrin and poly(alkylcyanoacrylate) nanoparticles through CACO-2 cells monolayers. Proceedings of the 4th World Meeting on Pharmaceutics, Biopharmaceutics, and Pharmaceutical Technology, ADRITELF, APGI, APV, Florence, 8–11 April 2002, 2002. 218. Marchal-Heussler L, Sirbat D, Hoffman M, Maincent P. Poly(ε-caprolactone) nanocapsules in carteolol ophthalmic delivery. Pharm. Res 1993; 10:386–390. 219. Marchal-Heussler L, Fessi H, Devissaguet JP, Hoffman M, Maincent P. Colloidal drug delivery systems for the eye. A comparison of the efficacy of three different polymers: polyisobutylcyanoacrylate, poly(lactic-co-glycolic acid), poly(ε-caprolactone). S.T.P. Pharma 1992; 2:98–104. 220. Losa C, Marchal-Heussler L, Orallo F, Vila Jato JL, Alonso MJ. Design of new formulations for topical ocular administration: polymeric nanocapsules containing metipranolol. Pharm. Res 1993; 10:80–87. 221. Calvo P, Thomas C, Alonso MJ, Vila Jato JL, Robinson JR. Study of the mechanism of interaction of poly(ε-caprolactone) nanocapsules with the cornea by confocal laser scanning microscopy. Int. J. Pharm 1994; 103:283–291. 222. Zimmer A, Kreuter J, Robinson JR. Studies on the transport pathway of PBCA nanoparticles in ocular tissues. J. Microencapsulation 1991; 8:497–504. 223. Calvo P, Alonso MJ, Vila-Jato JL, Robinson JR. Improved ocular bioavailability of indomethacin by novel ocular drug carriers. J. Pharm. Pharmacol 1996; 48:1147–1152. 224. Calvo P, Vila-Jato JL, Alonso MJ. Effect of lysozyme on the stability of polyester nanocapsules and nanoparticles: stabilization approaches. Biomaterials 1997; 18:1305–1310. 225. Calvo P, Vila-Jato JL, Alonso MJ. Evaluation of cationic polymer-coated nanocapsules as ocular drug carriers. Int. J. Pharm 1997; 153:41–50. 226. Calvo P, Sanchez A, Martinez J, Lopez MI, Calonge M, Pastor JC, Alonso MJ. Polyester nanocapsules as new topical ocular delivery systems for cyclosporin A. Pharm. Res[. Pharm. Res;] 1996; 13(9):1423. 227. Juberias JR, Calonge M, Gomez S, Lopez MI, Calvo P, Herreras JM, Alonso MJ. Efficacy of topical cyclosporine-loaded nanocapsules on keratoplasty rejection in the rat. Curr. Eye Res 1998; 17: 39–46. 228. De Campos AM, Sa´nchez A, Alonso MJ. Chitosan nanoparticles: a new vehicle for the improvement of the delivery of drugs to the ocular surface. Application to cyclosporin A. Int. J. Pharm 2001; 224:159–168.
598
Vauthier et al.
229. Le Bourlais CA, Chevanne F, Turlin B, Acar L, Zia H, Sado PA, Needham TE, Leverge R. Effect of cyclosporin A formulations on bovine corneal absorption: ex-vivo study. J. Microencapsulation 1997; 14:457–467. 230. Desai SD, Blanchard J. Pluronic威 F127–based ocular delivery systems containing biodegradable polyisobutylcyanoacrylate nanocapsules of pilocarpine. Drug Delivery. J Delivery Targeting Therapeutic Agents 2000; 7:201–207. 231. Fresta M, Fontana G, Bucolo C, Cavallaro G, Giammona G, Puglisi G. Ocular tolerability and in vivo bioacailability of poly(ethylene glycol) (PEG)-coated polyethyl-2-cyanoacrylate nanospheresencapsulated acyclovir. J. Pharm. Sci 2001; 9:288–297.
27
Hyperspectral Analysis of Collagen Infused with BisGMA-Based Polymeric Adhesive Paulette Spencer and Yong Wang University of Missouri–Kansas City School of Dentistry Kansas City, Missouri, U.S.A.
J. Lawrence Katz University of Missouri–Kansas City, Missouri and Case Western Reserve University Cleveland, Ohio, U.S.A.
Massood Tabib-Azar Case Western Reserve University Cleveland, Ohio, U.S.A.
Ajay Wagh Case Western Reserve University, Cleveland, Ohio and University of Memphis Memphis, Tennessee, U.S.A.
Tsutomu Nomura Case Western Reserve University, Cleveland, Ohio, U.S.A., and Niigata University School of Dentistry Niigata, Japan
I. BACKGROUND A. Dental Amalgam and Posterior Composite Restorations The Health Care Financing Administration estimated that expenditures for dental treatment approached $60 billion in 2000 [1]. Most of this money was spent repairing teeth and periodontal tissues. In general dentistry practices, nearly 75% of all operative dentistry is devoted to replacement of failed restorations [2,3]. The emphasis on replacement therapy in general dentistry practices is only expected to grow as the public’s concern about dental amalgam forces dentists to select alternative restorative materials, such as composite resin. The failure rate for large to moderate posterior composite restorations can be 2–3 times that of high copper amalgam [4]. After 10 years, 86% of class I and II composite resin restorations placed in young adult patients had failed [5]. Based on the poor performance of these materials, previous authors have concluded that extended class II composite restorations must be regarded as a clinical compromise [6]. The reduced clinical lifetime of moderate to large class II composite restorations can be particularly detrimental for patients because removal of these restorations can lead to extensive 599
600
Spencer et al.
loss of sound tooth structure. For example, in comparison to the removal of amalgam, the removal of tooth-colored restorations produced significantly greater increases in cavity volume [7]. The increase in cavity volume and increased frequency of replacement means that significantly greater amounts of tooth structure will be lost with treatment and retreatment of class II composite restorations, i.e., sound tooth structure is inevitably removed with each replacement [7,8]. Over the lifetime of the patient, the additional loss of tooth structure with treatment and retreatment of these intracoronal composite fillings will translate to the need for an enlarged and more complex restoration. Thus, the reduced longevity of the class II composite restoration means increased frequency of replacement with concomitant loss of sound tooth structure and an increase in cost as compared to class II amalgams [7,9]. In spite of the differences in failure rates, dentists are facing increased pressure to use composite materials in place of dental amalgam for posterior restorations [10]. The focus on the health and environmental risks associated with the release of mercury from dental amalgam has escalated in recent months with members of congress asking the National Institutes of Health to study the safety of low-level medical and dental uses of mercury including dental amalgam (ADA online). Although no nation has banned the use of amalgam, several countries, including Germany, England, Canada, the Netherlands, Norway, and Sweden, have restricted the use of dental amalgam (ADA online). B. Composite Failure The premature failure of moderate to large composite restorations can be traced to a breakdown of the bond at the tooth surface/composite material interface [4–11,12] and increased levels of the cariogenic bacteria Streptococcus mutans at the perimeter of these materials [13,14]. The breakdown of the bond leads to the formation of gaps at the tooth/composite interface; the gaps act as a conduit for penetration of bacterial enzymes, bacteria, fluids, and ions. This exchange of fluids and other substances at the tooth surface/composite material interface is a major factor in the development of secondary caries, hypersensitivity, and pulpal inflammation [15–18]. The bond at the tooth surface/composite material interface actually involves two distinctly different substrates, i.e., dentin and enamel. The composition of the enamel is approximately 98% mineral and 2% protein, while the dentin is 50% mineral, 30% protein, and 20% water (by volume) [19,20]. Acid-etching of enamel and subsequent resin bonding to the etched enamel is a well-established and reliable procedure for decreasing marginal leakage at the enamel/ composite interface. Dentin bonding has not been as predictable, and to date dentin bonding does not provide a seal that can prevent the penetration of bacterial enzymes, oral fluids, or other damaging substances at the composite/adhesive/tooth interface. C. Composition of Dentin Dentin is a complex hydrated biological composite structure that is modified by physiological, aging, and disease processes [20]. Dentin mineral is a carbonate-rich, calcium-deficient apatite. The organic component is predominantly type I collagen with minor contribution from other proteins that can be categorized as phosphoproteins, glycoproteins and ␥-carboxyglutamatecontaining proteins [21] A unique feature of the dentin structure is the tubules that traverse the structure from the pulp cavity to the region just below the dentin–enamel junction (DEJ) or cementum. The tubules, which could be modeled as narrow tunnels a few microns or less in diameter, represent the tracks taken by the odontoblastic cells from the pulp chamber to the respective junctions. Tubule density and orientation vary from location to location; density is lowest at the DEJ and highest
Hyperspectral Analysis of Polymeric Adhesive
601
at the predentin surface at the junction to the pulp chamber. The content of the tubules includes fluid and odontoblast processes for all or part of their course. In contrast to root dentin, the tubules in coronal dentin are surrounded by a collar of highly mineralized peritubular dentin [22]. The composition of the peritubular dentin is carbonate apatite with very small amounts of organic matrix, whereas intertubular dentin, i.e., the dentin separating the tubules, is type I collagen matrix reinforced with apatite. The apatite crystals are approximately 5 ⳯ 30 ⳯ 100 nm and contain 4–5% carbonate as compared with hydroxyapatite [20]. The amount of collagenrich intertubular dentin gradually decreases from superficial to deep dentin [23]. D. Bonding to Dentin Components In order to understand and define the mechanisms that contribute to bond formation at the dentin/ adhesive interface, numerous studies have correlated morphologic findings with bond strength measurements [24–27]. Based on the results of these studies, it generally is accepted that the following factors are critical to the development of an adequate adhesive/dentin (a/d) bond: (1) removal of the mineral phase from the surface of the dentin substrate by acid-etching; (2) wetting of the dentin substrate by components of the adhesive; and (3) adhesive penetration and encapsulation of the collagen fibrils in the demineralized dentin. Morphologic evidence of resin penetration of the exposed collagen fibrils was first reported by Nakabayashi, and he called the distinct zone between the bulk adhesive and unaltered dentin the hybrid layer [28]. Current adhesive systems that acid-etch the dentin characteristically bond via hybridization [29]; optimal interfacial bond strengths with these adhesives depend on resin diffusion through the hybrid layer and into the unaltered, intact dentin [27–30–32]. During acid-etching the mineral phase is extracted from the top 1–8 m of the dentin surface [33,34]. The composition of the exposed substrate differs radically from mineralized dentin, which is 50% mineral, 30% collagen, and 20% water by volume [19,20]. Hydrated demineralized dentin is 30% collagen and 70% water [30,35]. Thus, with removal of the mineral phase the collagen fibers are literally suspended in water. If there is a substantial zone of demineralization and the water supporting the collagen network is removed either by air drying or the action of an air syringe, the collagen will collapse [31–32,35]. The collapsed collagen network reduces the porosity and inhibits adhesive resin penetration through the demineralized layer [35]. The dentin/adhesive bond is weakened because the collapsed collagen forms a barrier that prevents resin penetration throughout the decalcified layer and into the unaltered dentin substrate. It has been reported that as long as the dentin is kept fully hydrated, the surface morphology of the demineralized layer does not change, i.e., the water supporting the collagen matrix is not lost, and the matrix network does not shrink [36]. Results from previous studies with a ‘‘wet’’ bonding technique support these findings [37–39]; the higher bond strengths with this technique reflect the minimal collapse of ‘‘wet’’ dentin collagen as opposed to collagen that is dried with an air syringe [35]. It is speculated that moist dentin provides a more porous collagen network and thus greater infiltration of adhesive monomers [27–35,39]. The initial increase in bond strength of adhesive applied using wet bonding techniques is not, however, maintained over time [40,41]. The factors and mechanisms involved in the premature failure of this adhesive/ dentin bond remain unclear. The durability of the a/d bond is directly related to the quality of the hybrid layer that connects the bulk adhesive to the subjacent, intact dentin. Ideally, the adhesive monomers occupy all the space left by mineral and envelop the exposed collagen fibrils, but recent studies have shown that this objective is frequently not achieved [33,42–44]. Although these studies have contributed substantially to our understanding, they do not provide quantitative information on
602
Spencer et al.
resin infiltration at the dentin interface as compared to an optimal hybrid layer [23]. The optimal hybrid layer would be characterized as a three-dimensional structurally integrated polymer/ collagen network. In this study, optimal hybrid structures were represented by adhesive-infiltrated demineralized dentin (AIDD) produced under controlled conditions. Using histomorphology and microRaman spectroscopy, the quality and molecular structure of adhesive/dentin interfaces formed with wet bonding were compared to these AIDD. To understand the molecular nature, mechanics, and permittivity of these optimal hybrid structures, the samples were analyzed using complementary chemical, mechanical, and electrical scanning techniques.
II. MATERIALS AND METHODS A. Specimen Preparation 1. Preparation of Demineralized Dentin Collagen and Adhesive-Infiltrated Demineralized Dentin Six extracted unerupted human third molars stored at 4⬚C in 0.9% w/v NaC1 containing 0.002% sodium azide were used. The teeth were collected after the patients’ informed consent was obtained under a protocol approved by the UMKC adult health sciences institutional review board. The occlusal one-third of the crown was sectioned perpendicular to the long axis of the tooth by means of a water-cooled low-speed diamond saw (Buehler Ltd, Lake Bluff, IL). Two parallel cuts, 2 mm deep, were made perpendicular to this surface using the diamond saw. The final cut was about 1.5 mm below the flat surface. The final dimension of the dentin slab was 10 mm long, 2 mm high, and 1.5 mm wide. The adjacent fraction of the cut tooth was used for the a/d interface specimens. The dentin slabs were demineralized in 0.5M EDTA (pH 7.3) at 25⬚C for 7 days with shaking. The solution was changed on alternate days. At the end of 1 week, randomly selected dentin slabs were sectioned and Raman spectra were acquired. The absence of spectral features associated with the mineral (P—O band at 960 cmⳮ1) indicated complete demineralization. The demineralized specimens were rinsed thoroughly with water. These demineralized dentin specimens were submitted for analysis using scanning acoustic microscopy. Randomly selected demineralized dentin slabs were dehydrated for 12 h in each of the following: 70, 95, and 100% ethanol. Following dehydration the specimens were immersed in Single Bond (SB) adhesive (3M, St. Paul, MN). The demineralized dentin collagen/adhesive specimens were placed in a dark room for 72 h. After 72 h, three specimens were polymerized with visible light (Spectrum light, Dentsply, Milford, DE). 2. Preparation of Adhesive/Dentin Interface Specimens The adhesive/dentin specimen preparation has been detailed in previous publications [33,42,43]. For each of the six molars, the fraction of the cut tooth, adjacent to the slab prepared for adhesive infiltration, was used. A uniform smear layer was created by abrading with 600-grit sandpaper under water cooling, and the prepared dentin specimens were treated with SB adhesive following manufacturer’s instructions. Dentin was etched with 35% phosphoric acid gel (15 s) and rinsed with water; excess water was removed but the dentin surface remained visibly moist. The adhesive was applied in two consecutive layers, gently air-dried and polymerized for 30 s using visible light. These specimens were stored for a minimum of 24 h in water at 25⬚C before further sectioning. The treated dentin surfaces were sectioned perpendicular and parallel to the bonded surface with a water-cooled low-speed diamond saw.
Hyperspectral Analysis of Polymeric Adhesive
603
3. Light Microscopy: Differential Staining Technique The rectangular 10 ⳯ 2 ⳯ 1.5 mm slabs of both Single Bond adhesive-infiltrated demineralized dentin and adhesive/dentin were mounted on a methacrylate support, and 3-m-thick sections were cut from the face of the slab using a tungsten carbide knife mounted on a Polycut S ‘‘sledge’’ microtome. Following recovery of the microtomed sections, the remaining fraction of the AIDD and a/d interface slabs was used for micro-Raman spectroscopic analysis. Thus, the same slab was used for both light microscopic and micro-Raman spectroscopic analysis. Differential staining was accomplished with Goldner’s trichrome. Slides with adherent stained sections were dehydrated through ascending ethanol and xylene. The sections were cover-slipped with mounting media and examined using a Nikon E 800 light microscope.
B. Micro-Raman Spectroscopy for Chemical Characterization In the broadest sense the Raman effect involves the scattering of light as a result of its interaction with matter. When monochromatic light from a laser strikes a sample almost all of the light is scattered elastically. This is known as Rayleigh scattering and is certainly the strongest component of the scattered radiation. A small fraction of the light is scattered inelastically, i.e., there is an energy transfer between the incident light and the scattering molecules. This change in energy or frequency between the incident and scattered light corresponds to an excitation of the molecular system, most often in a vibrational mode. The Raman spectrum represents the intensity of the scattered photons as a function of the difference in frequency. It is similar to an infrared spectrum in that it provides a ‘‘fingerprint’’ of the molecules present within the sample and can be used for qualitative identification and quantitative determination [45,46]. By combining spectroscopy with microscopy, micro-Raman spectroscopy (RS) can be used to detect and quantify the molecular chemistry of microscopic samples. Using this technique, an investigator can detect compound-specific molecules from a sample without spectral interference from H2O and at a lateral spatial resolution of approximately 1 m [47]. Samples can be analyzed directly, in air or water, at room temperature and pressure, wet or dry, and without destroying the sample. It is an exceptional tool for investigating the chemistry of material/ tissue interfaces because it does not rely on homogenization, extraction, or dilution, but rather each structure is analyzed in situ. We have used RS to quantify the diffusion of single-bottle adhesives into the ‘‘wet’’ demineralized dentin matrix [33,43], to determine for the first time the molecular structure of acid-etched smear layers [48] and smear debris [34], and to quantitate demineralization in hydrated dentin specimens [48]. 1. Micro-Raman Spectroscopic Analysis The micro-Raman spectrometer consisted of an argon ion laser beam (514.5 nm) focused through a X60 Olympus Plan Neofluor water-immersion objective (NA 1.2) to a ⬃1.5-m beam diameter. Raman back-scattered light was collected through the objective and resolved with a monochromator. The spectra were recorded with a software-controlled CCD array. The laser power was approximately 3mW; an imaging system and high resolution monitor were incorporated to allow visual identification of the position at which the Raman spectrum was obtained. Spectra were Raman shift frequency calibrated using known lines of neon and silicon. Each a/d interface slab was placed at the focus of the objective and covered with distilled water in preparation for micro-Raman spectroscopic analysis. Spectra were acquired at positions corresponding to 1-m intervals across the a/d interface using the computer controlled x-y-z stage with a minimum step width of 50 nm. Multiple sites across the interface of each specimen
604
Spencer et al.
were examined spectroscopically. Overlap of the spectra from these sites confirmed the reproducibility of the technique. No postprocessing of the data was performed. Raman spectra of AIDD were acquired from a minimum of six to eight different sites on each sample. Spectra were obtained at a resolution of ⬃6 cmⳮ1 over the spectral region of 875–1785 cmⳮ1 and with an integration time of 60 s. A comparison of the spectra that were collected from the six to eight different sites indicated complete overlap, suggesting that there was a homogeneous infiltration of adhesive throughout these samples. These specimens were submitted for micromechanical analysis using scanning acoustic microscopy. C. Scanning Acoustic Microscopy Scanning acoustic microscopy (SAM) is a powerful experimental tool for examining the acoustic and thus the elastic properties of materials at resolutions on the order of microns [49–51]. One of the major advantages of SAM in the study of the elastic properties of biological tissues is that a liquid couplant must be used in order to transmit the acoustic signal from the lens system to the specimen being examined. Thus, such factors as drying and heating that may change the properties of biological tissues are avoided as the liquid inhibits desiccation and heat generation during measurements. This procedure makes it possible to obtain the micromechanical elastic properties of fresh as well as of embedded specimens. These types of samples have been used by a number of groups interested in measuring the micromechanical elastic properties of bone and bone/biomaterial interfaces [52–60]. The heart of the SAM is a focusing lens (Fig. 1). The transmitter generates a radio frequency (RF) signal that excites a piezoelectric transducer mounted at the top of a high quality single crystal sapphire buffer rod. Thus, the RF signal is converted into an acoustic wave that, upon convergence by the lens, propagates through the liquid couplant onto the sample that is mounted on a movable stage (Fig. 1). At the liquid/sample interface, a portion of the impinging acoustic signal is reflected back to the lens, now acting as a receiver, that transforms the acoustic signal back into a voltage, V, that is proportional to the reflected signal (Fig. 2). This voltage is both stored in memory and displayed as a pixel of a given gray level proportional to V on the instrument’s monitor screen. A combination of both lens and/or stage motion repeats the process pixel over a desired area of the specimen providing a varying gray level acoustic image of that area. The voltage and thus the gray level are determined by an acoustic parameter, the reflection coefficient, r, where r ⳱ (Z2ⳮZ1)/(Z2ⳭZ1), and Z is the acoustic impedance (AI) (39). Thus, r is a dimensionless number between 0 and 1. The acoustic impedance is the product of the longitudinal (dilatational) velocity, v, and the local density, , at the point on the material, i.e., Z ⳱ v (Fig. 2). AI is measured in Rayls (e.g., water has a value Z ⳱ 1.40 Mrayl, at 0⬚C, and a value of Z ⳱ 1.51 Mrayl at body temperature, 37⬚C). This difference is due to water’s acoustic properties varying with temperature. Dentin, the source of the collagen used in this combined study, has an AI of Z ⳱ ⬃7.5 Mrayl, resulting in a reflection coefficient r ⳱ 0.67 in water at 37⬚C. The r value measured at each point on the sample determines the shade of gray at the corresponding pixel on the monitor. Dark gray levels at the points being measured represent lower values of AI (and thus of Young’s modulus, E); bright gray levels represent higher AI values (and thus higher values of E). As the Olympus UH3 SAM does not provide a direct measurement of quantitative values of Z and thus eventually of Young’s modulus, E, a calibration method must be used. This requires obtaining the acoustic velocities, v, and densities, , for a set of known materials including: polyethylene, PMMA, Durango apatite, aluminum, titanium, and stainless steel. Both the AI, Z ⳱ v, and Young’s modulus, E ⳱ v2, for each material is then calculated. These same materials are also imaged on the SAM in order to obtain their r values. Voltages corresponding to
Hyperspectral Analysis of Polymeric Adhesive
605
Figure 1 Lens design for scanning acoustic microscopy. The method of operation is illustrated and specimen conditions described. (From Ref. 61.)
these r values are obtained directly from the instrument and are plotted on a graph of r vs. V. Both V vs. Z and Z vs. E for the known sample materials are then graphed as well. Values for the materials under study are then either interpolated or extrapolated in sequence on these graphs of the known materials in order to eventually obtain their Young’s moduli. In a previous study of the unprotected protein (collagen) at the dentin/adhesive interface [61], the unprotected protein exhibited gray levels darker than that of the lowest polymer. Unfortunately, interpolation is limited to the polymer with the lowest Young’s modulus of the known materials used in this calibration process, polyethylene. Therefore, it would not be possible to obtain a specific Young’s modulus for any material that exhibited gray levels darker than that
Figure 2 Relationship between acoustic impedance, Z, and reflection coefficient, r. I represents the incident ultrasonic wave, R the reflected wave, and T the transmitted wave. (From Ref. 61.)
606
Spencer et al.
of the lowest polymer, polyethylene, implying a lower r value. Thus, it was only possible to estimate the upper limit for the unprotected protein [61]. In order to be able to extrapolate below the lowest polymer value on the Z vs. E curve its necessary to validate the procedure. This was done by calculating both the bar (extensional longitudinal) acoustic wave and bulk (shear and dilatational longitudinal) acoustic waves, and Young’s moduli for 13 materials, including polymers, glasses, ceramics, and metals. This validation led to the experiments described herein, the results of which will be published in the Journal of Biomedical Materials Research [108]. 1. Scanning Acoustic Microscopy/Calibration Methods The same technique used in the previous study[61] was applied to both the specimens of collagen with and without adhesive infiltration. Each specimen, in turn, was mounted on the stage of the Olympus UH3 SAM and analyzed using the 400 MHz Burst Mode lens (120⬚ aperture angle; nominal lateral resolution 2.5 m). Values of voltages corresponding to the respective gray levels were measured at ten different points on the SAM micrographs and averaged for both the demineralized dentin collagen and the demineralized dentin collagen samples infiltrated with Single Bond adhesive. The same calibration curves—based on known materials (PMMA, polypropylene, polystyrene, Teflon, dentin, Pyrex, glass, enamel, brass, copper, steel)—used in our previous study[61] were used here as well. These calibration curves are used in sequence: the graph of reflection coefficient, r, vs. stored voltage, V first (Figs. 3); then the graph of r vs. the acoustic impedance, Z (Fig. 4); and finally the graph of Z vs. Young’s modulus, E (Fig. 5). The high correlations denoted by the strong R2 value for each graph indicates the validity of the interpolation procedure. However, the same concerns about extrapolation below the lowest known calibration value described in our previous study [61] arose here as well, i.e., the gray levels for both the demineralized dentin collagen and the collagen component in the demineralized dentin collagen samples infiltrated with Single Bond adhesive were below the lowest values on the three calibration curves (Figs. 3–5). An additional [108] calculation was introduced in order to test two hypotheses:(1) the calibration procedure described above, which is based on ultrasonic measurements at frequencies
Figure 3 The graph of reflection coefficient versus stored voltage for 11 standard materials.
Hyperspectral Analysis of Polymeric Adhesive
607
Figure 4 The graph of reflection coefficient, r, versus acoustic impedance, Z.
below the 400 MHz used herein, is valid for use at the higher SAM frequencies and (2) extrapolation is valid for determining the properties of unknown materials whose reflection coefficients fall below the lowest values on the graphs. This would permit a direct measurement of Young’s moduli rather than an upper bound as presented previously [61] for the collagen specimens in the respective SAM micrographs in this study. The extensional longitudinal velocity, vL, transverse (shear) velocity, vt, dilatational longitudinal velocity, vl, and density, , for 13 calibration materials, covering the range from low modulus polymers to high modulus metals, were obtained from the literature (Table 1). Young’s modulus then was obtained for each of the 13 materials by two different calculations: (1) the lower acoustic frequency (bar wave) measurement E(bar),
Figure 5 The graph of acoustic impedance, Z, versus Young’s modulus, E.
608
Spencer et al.
Table 1 Densities and Bar Longitudinal (Extensional), vL, Bulk Longitudinal (Dilatational), v1, and Transverse (Shear), vt, Velocities for 13 Known Materials Material Aluminum, rolled Brass (70% Cu, 30% Zn) Copper, rolled Nickel Steel, 347 stainless Titanium Zinc, rolled Fused silica Glass (Pyrex) PMMA (Lucite) Nylon (6.6) Polyethylene Polystyrene
Density, (gm/cc)
vL (m/sec)
v1 (m/sec)
vt (m/sec)
2.7 8.6 8.93 8.9 7.9 4.5 7.1 2.2 2.32 1.18 1.11 0.90 1.06
5000 3480 3750 4900 5000 5080 3850 5760 5170 1840 1800 920 2240
6420 4700 5010 6040 5790 6070 4210 5968 5640 2680 2620 1950 2350
3040 2110 2270 3000 3100 3125 2440 3764 3280 1100 1070 540 1120
was calculated as E(bar) ⳱ vL2 and (2) the higher acoustic frequency (bulk wave) measurement, E(bulk), as (bulk) ⳱ 9KG/(3KⳭG), where the shear modulus is G ⳱ vt2 and the bulk modulus is K ⳱ [vl 2 ⳮ (4/3)vt2] (Table 2). Figure 6 is the graph of E(bar) vs. E(bulk).
D. Evanescent Microwave Probe Microscopy We discuss a new family of local probes that use evanescent electromagnetic fields to nondestructively study electromagnetic properties of materials at 1–20 GHz. These probes are used to perform microwave microscopy and imaging with spatial resolution approximating atomic force microscopy. Such an unprecedented high spatial resolution with electromagnetic fields having relatively large wavelengths (free space ⬃1.5–30 cm) has been made possible by small spatial
Table 2 Elastic Modulus for Bulk Wave and Bar Wave Material Aluminum, rolled Brass (70% Cu, 30% Zn) Copper, rolled Nickel Steel, 347 stainless Titanium Zinc, rolled Fused silica Glass (Pyrex) Lucite Nylon 6,6 Polyethylene Polystyrene
E ⫽ v2t
E ⫽ 9KG/(3K ⫹ G)
67.5 104.1 125.6 213.7 197.5 116.1 105.2 73 62 4 3.6 0.76 5.3
67.6 105.2 126.1 214.1 197.2 115.9 105.5 73 62.2 4 3.55 0.76 3.5
Hyperspectral Analysis of Polymeric Adhesive
609
Figure 6 The graph of bar wave modulus versus bulk wave modulus for 13 standard materials based on data in Table 2.
decay constants of evanescent fields generated at the terminal end of a microwave resonator near a wire tip. Upon interaction with a sample placed near this tip, the reflection coefficient of the resonator shifts to lower frequencies, enabling the characterization of the microwave properties of the sample which are affected by various factors including density, moisture, polymerization, carrier mobility and concentration, impurities, oxidation state, and temperature. By using the evanescent microwave probe (EMP) a variety of organic and inorganic materials including metals, semiconductors, insulators, composites, ferromagnetic materials, tooth enamel, botanical, and agricultural samples have been imaged. Here we discuss principles of operation of the EMP, parameters affecting its spatial and permittivity/permeability/conductivity resolutions, and examples of its applications in organic and inorganic conductors, semiconductors, and insulators. Current trends and future work are also discussed. 1. Introduction Evanescent fields were first used by Bethe in calculating the coupling coefficient of microwave waveguides connected to each other through a hole much smaller than the microwave wavelength
610
Spencer et al.
[62]. Baez is credited as the first individual to image objects using sonic evanescent waves [63] followed by Soohoo, who performed similar experiments using microwaves [64]. Ash is credited with generating evanescent microwave fields in a very elaborate test rig to demonstrate that these fields can be used to resolve metallic features on the order of /100 [65]. Others have studied evanescent fields in the context of imaging as well [66,67]. Several research groups, including our own, have used microwave resonators in the characterization of semiconductors with /100 [68–75]. The evanescent microwave probe used in our work is a planar structure [71] that readily lends itself to implementation on a silicon cantilever beam. We started working on evanescent microwave probes for imaging applications in the late 1980s, and our initial resolution using a microstripline resonator was around 80 m at 1 GHz [71]. The EMP’s applications in a variety of areas were also explored by us [76–93]. Recently we achieved nearly atomic resolution by coupling the microwave probe to the atomic force microscope (AFM) and scanning tunneling microscope (STM) [78,88]. We have shown that a variety of materials ranging from conductors to insulators can be imaged using the EMP [71–76–93]. Our work in microwave super-resolution imaging of biological and botanical specimens, e.g., woods and fruits, has shown that interesting microwave resistivity maps can be obtained and used to study aging and other processes in these materials. For example, EMP can be used to monitor electrical activities in bone samples [76–77,86]. Resorption and remodeling of bone have been attributed to obsteocytic cells sensing changes in streaming potentials generated during deformation of the tissue. Thus, measurements of the electromagnetic properties of such tissues could prove crucial in the development of models to explain bone remodeling. EMP is sensitive to moisture and ionic mineral content such as variations in tooth enamel. EMP is capable of detecting the onset of caries — believed to be accompanied by minute surface blemishes with increased moisture and varying degrees of mineralization. Having a microwave power of less than nanowatts, this unique probe promises to significantly reduce the time to detection of cavity formation, enabling timely prevention [76]. The EMP has the following important characteristics: (1) The EMP uses coherent microwave sources that are readily and inexpensively available over a wide range of frequencies covering 100 MHz up to 100 GHz. (2) The EMP does not require conducting or optically transmitting/reflecting samples. (3) The spatial resolution of the EMP can be varied over a wide range, even at a fixed frequency, by using different probe configurations. (4) The EMP can image subsurface defects and nonuniformities within the microwave skin depth inside the sample. (5) The EMP does not require any coupling medium and it can be used in air, vacuum, or in a suitable liquid. (6) Operating at very high frequencies, and using the homodyne detection technique, the EMP can achieve very high scan rates (up to a few cm/s) over hot and cold samples. (7) Many different parallel EMPs can be used simultaneously to scan over large areas. 2. Principles of Operation Evanescent microwave probe microscopy is based on a planar waveguide resonator geometry schematically shown in Fig. 7 [71–78–80,85]. The reflection coefficient of this resonator in air and with a metallic sample is shown in Fig. 8. The metallic sample causes the reflection spectrum to shift to lower frequencies. As shown in Fig. 8 both the resonance frequency, f0, and the quality factor of the resonator, Q, are affected by the presence of the sample. The amount of change in the resonance (⌬f and ⌬Q) depends primarily on the resonant frequency, the microwave properties of the sample, and the distance between the resonator’s tip and the sample. The reflection spectra, shown in Fig. 8, were obtained using an automated network analyzer (HP 8720C). The network analyzer was also used in the characterization, design, and tuning of the microwave probe.
Hyperspectral Analysis of Polymeric Adhesive
611
Figure 7 A coplanar waveguide half-wavelength resonator used in evanescent microwave probe (EMP) microscopy. The exponential decay of the electric field is schematically shown near the probe’s tip. Other planar geometry such as microstripline and stripline waveguides, not shown here, are also used.
Figure 8 The reflection (S11), spectrum (A), and its modification (M) due to the presence of a metallic sample near the probe tip. The difference between these two spectra (MⳮA) is also shown. For this probe the maximal sensitivity was obtained at fx 艐917 MHz.
612
Spencer et al.
Figure 9 (a) Schematic of the experimental setup used in EMP imaging and measurements. (b) Optical image of the apparatus.
3. Experimental Setup The experimental setup used in a variety of EMP measurements and imaging applications is shown in Fig. 9. It consists of a microwave resonator coupled to a feedline, which is connected to a circulator. The circulator is also connected to an RF (0.01–15 GHz) signal generator and to a crystal microwave detector. The detector output is a DC voltage proportional to the magnitude of the reflected wave. This is fed to an amplifier and then to a lock-in amplifier. The probe is mounted vertically over a–x–y stage (Fig. 9b). The x–y–z stage and the frequency generator are controlled by a personal computer that also acquires data from the lock-in amplifier. The lock-in amplifier is synchronized with the amplitude modulation signal (frequency ⬃ 1 KHz) generated by the RF signal source. A computer controls the stage and
Hyperspectral Analysis of Polymeric Adhesive
613
Figure 9 Continued.
the lock-in amplifier and generates the EMP images or records the EMP as a function time or gas concentration. The presence of a sample, or changes in the sample’s electromagnetic properties when exposed to illumination, heat, various fields, or to gases or chemicals, usually shifts the reflection spectrum to lower frequencies (Fig. 10). Thus to obtain monotonously changing EMP signal as a function of these changes, the operation frequency, fx, should be selected above the resonant frequency, f0. Moreover, to obtain maximal probe sensitivity, fx should be as close to f0 as possible. These observations are schematically illustrated in Figs. 8 and 10. The experimental setup shown in Fig. 9 can also be used to perform infrared transmission imaging simultaneously with EMP imaging. It can also be used to illuminate the sample’s surface directly under the EMP tip to perform optical spectroscopy and to use EMP to monitor carrier recombination/generation transients when the illumination is interrupted. These measurements, not discussed here, are used to study the integrity of wafer bonding and the quality of wafers. Other detection methods include (1) vibrating the tip or the sample and (2) using frequency modulation/demodulation techniques. An amplitude modulation method can also be implemented by vibrating the tip or the sample. This process modulates the stand-off distance, and at fixed RF operation frequency it results in the amplitude modulation of the reflected signal. This technique is very similar to the amplitude modulation technique discussed above, and the best operation frequency in this case is obtained as shown in Figs. 8 and 11. The frequency modulation technique is schematically shown in Fig. 11. In this technique the incident RF signal has a carrier frequency that is usually chosen to coincide with the resonant frequency of the EMP resonator. This results in a 2fm component of the S11 that is proportional to the quality factor, Q, of the resonator. In some cases when the sample is conducting it can be shown that f0 is mainly affected by the stand-off distance and Q depends on the dissipation in the sample. In this case the component of S11 at fm may be used (by changing the carrier
614
Spencer et al.
Figure 10 (a) Amplitude modulation methods. The RF signal at fx2 yields the largest change in the EMP signal compared to the signal at fx1. The tip–sample stand-off distance modulation also yields a similar result. (b) Frequency modulation technique. If the carrier frequency is chosen to coincide with the resonant frequency of the probe, the reflection coefficient is modulated as twice the frequency of the modulation (fm). In this case the variation of S11 at fm can be used to find fo, and its variation at 2 fm may be used to find the width of the resonance curve which is proportional to I/Q, where Q is the quality factor.
frequency until the fm component is nearly zero) to detect f0, and the 2fm component is used to measure Q and to monitor or image the sample’s conductivity. In metals, defects and stresses can locally change the conductivity and hence can be detected by the EMP. In the case of semiconductors the probe output can be affected by variations in carrier density, interface trap density, defects, presence of mobile or fixed charges, grain boundaries, and variations in film thickness. In an earlier work, changes in the reflection coefficient amplitude as a function of microwave frequency and carrier concentration in Si have been
Hyperspectral Analysis of Polymeric Adhesive
615
Figure 11 The EMP S11 spectra obtained in air and with three different sample sheet resistances. A glass substrate plated with different thicknesses of gold was used in these measurements. The sheet resistance of gold was measured using the four-point probe technique.
reported [80]. Magnetic domains in ferromagnets can also be mapped using the EMP through the contribution of the permeability to the microwave properties of these samples. The microwave properties of a material are a function of its permittivity, permeability, and free carrier concentration [85]. In most biological tissues, moisture content and ionic species, such as NaⳭ and KⳭ, probably affect the conductivity significantly. These parameters along with the density variations, which affect the permittivity, can be mapped using the EMP [79]. 4. Calibration Like any other sensor the EMP can be calibrated to quantify its output and relate it to the microwave properties of the sample. Figure 12 shows the resistivity calibration of the EMP. A silicon sample illuminated by white light was used as the sample. At each illumination setting, the silicon resistivity was independently measured using a four-point probe. Moreover, for each measurement, a fiber optic distance sensor was used to adjust the probe–sample distance to be exactly 2 m to ensure reproducibility. The accuracy of these measurements was better than 6% for samples of low resistivity and became worse (50% off) for high resistivity samples because of the inability of the four-point probe to correctly measure high resistivity values. Instead of performing single frequency measurements, it is also possible to obtain the reflection, or S11, spectrum of the EMP, as shown in Fig. 11. These spectra can be used to extract other EMP parameters as a function of a sample’s microwave properties. Three important parameters of these spectra are resonant frequency, f0, bandwidth, ⌬f, and the depth of the reflection spectrum at f0 [given by l0 艐 S11(0) -S11(f0)]. The important aspects of the spectra shown in Fig. 11 are that all the above parameters (f0, ⌬f, and l0) change as a function of the sample’s resistivity. The amount of change in the S11 spectra depends on whether the EMP resonator is critically or under- or overcoupled and on the coupling between the EMP and the sample. The best performance (i.e., largest sensitivity) is usually obtained when the EMP is
616
Spencer et al.
Figure 12 The EMP output as a function of conductivity of silicon. The experiment was repeated four times to check the reproducibility of data. This curve also shows one of the unique characteristics of the EMP related to its ability to respond over six orders of magnitude change in the sample’s sheet resistance. An illumination source was used to change the sample’s resistivity, which was independently measured using a four-point probe.
critically coupled, when the probe-to-sample distance is as small as possible, and when the feed line impedance is approximately in the range of the sample’s impedance of interest. The dielectric calibration of the EMP is shown in Fig. 13. A sample with four different dielectric regions was used in this study. The stand-off distance was kept constant across the sample using a fiber optic distance sensor. The EMP calibration of permeability is similar to the dielectric calibration, but is complicated by the hysterisis effects in the ferromagnetic and high- samples. It is possible to model the response of the probe to sample resistivity. The sample can be modeled by a wave impedance, which is zero for perfect conductors and (1Ⳮj)R for semiconductors and imperfect conductors. The characteristic impedance is (/ε)0.5 for lossless dielectrics, and it can be approximated for lossy dielectrics:
3 ε" µ 1− ε 8 ε'
2
+j
ε" 2ε'
(1)
Here ε′-jε⬙ is the complex permittivity and ε⬙/ε′ is assumed much less than 1. It is also well known that the variations in the local conductivity of the sample can be due to variations in carrier concentrations, thickness, defects, interface traps, stresses, impurities, and moisture content. The permittivity can vary due to local variations in atomic polarizability as well. 5. EMP Output and Sample Parameters The EMP output is related to the scattering parameter of its resonator as EMP = S11 Vin
(2)
where Vin is the microwave input to the probe and S11 is related to the probe impedance through the well-known transmission line equation:
Hyperspectral Analysis of Polymeric Adhesive
617
Figure 13 Dielectric calibration curve of the EMP. A sample with four different dielectric regions was used in this study. A fiber optic position sensor was used to keep the stand-off distance constant.
S11 =
Ztotal − Z0 Z total + Z 0
(3)
where Z0 is the characteristic impedance of the microstripline (50⍀) and Ztotal is the total impedance of the resonator–sample system. For a metallic sample and near the resonance frequency, capacitances and inductances cancel each other out and Ztotal can be approximately written as Z0 Ⳮ R, where R (⬍50) is the combined resistance of the resonator, R0, and the sample, Rs. Thus the above equation becomes:
S11 =
R 2Z 0
(4)
Usually R0 ⬍⬍ Rs and Rs is given by the dissipation process in the sample within one skin depth from the surface, as discussed previously. Equation (4) is quite useful, but to apply it to the experimental data it is more useful to cast it in a differential form:
∆EMP ∆S11 Rs − Rs ref = = EMPref S11 ref Rs ref
(5)
where ⌬x⳱xⳮxref and x is any of the parameters above. EMPref or S11ref can be the value of EMP or S11 at the origin or it can be obtained intermittently at every point on the sample by illuminating the sample [80,85], by applying a DC electric field [85], etc. The important aspect of the Eq. (5) is that it is written completely in terms of the sample properties. This way the absolute values implicit in Eq. (4) need to be calculated once.
618
Spencer et al.
To perform EMP simultaneously with AFM [94], we have designed and fabricated special AFM-compatible tips [95,96] to use the AFM system to perform EMP [96,97]. The experimental setup is shown in Fig. 14.
III EXPERIMENTAL RESULTS A. Micro-Raman Spectroscopy A representative photomicrograph of a Goldner’s trichrome stained section of the a/d interface is presented in Fig. 15a. This photomicrograph provides a clear, detailed image of the interface. The mineralized dentin stained green, while the bulk adhesive stained a very pale yellow. A distinct red region, representing exposed collagen at the a/d interface that was available for reaction with the Goldner’s trichrome stain is visible at the a/d interface in the light micrograph; the pale color resin tags and dark red intertubular area in the a/d interface are clearly differentiated. The mean width of the entire zone that showed red color distinct from either pure SB or mineralized dentin was 8 Ⳳ 0.6 m. Representative photomicrographs of Goldner’s trichrome stained sections from AIDD specimens, which were polymerized before and after removal of ethanol solvent, are shown in Figs. 15b and c, respectively. The ⬃4-m-diameter adhesive tags appeared clear, while the intertubular regions stained light orange in the specimen polymerized before removal of ethanol solvent. The corresponding Raman spectra of the intertubular region of the a/d interface and intertubular demineralized dentin collagen infiltrated with adhesive are also presented in Fig. 15. The bands associated with the adhesive occur at 1720 cmⳮ1 (carbonyl), 1609 cmⳮ1 (phenyl C⳱C), 1453 cmⳮ1 (CH2 def), 1187 cmⳮ1 (gem-dimethyl), 1113 cmⳮ1 (C—O—C); while the bands associated with collagen occur at 1667 cmⳮ1 (amide I), 1454 cmⳮ1 (CH2 def), and 1245 cmⳮ1 (amide III). In comparing these spectra, the bands associated with adhesive are dominant in the AIDD specimens, while the features associated with collagen dominate the spectrum recorded from the a/d interface. The less intense bands (1113, 1187, 1609, and 1720 cmⳮ1) associated with adhesive within the interface indicate less diffusion of resin monomers into the zone of demineralized dentin at the a/d interface. Using a water immersion lens the same slabs that were analyzed by light microscopy were imaged and Raman spectra were acquired at 1-m intervals across the a/d interface (Fig. 16a).
Figure 14 Atomic force microscopy and EMP experimental arrangement.
Hyperspectral Analysis of Polymeric Adhesive
619
Figure 15 Light micrographs and Raman spectra of Single Bond adhesive/dentin interface (a) and Single Bond adhesive-infiltrated demineralized dentin (AIDD) without and with removal of ethanol before polymerization (b and c, respectively). These 3-m sections were stained with Goldner’s trichrome; corresponding Raman spectra were recorded from the intertubular area of the samples. (From Ref. 98).
As shown in Fig. 16b, in the series of micro-Raman mapping spectra collected from this a/d interface, the first spectrum was acquired from pure adhesive. Bands associated with the adhesive and collagen components of dentin are noted in the second spectrum. The Raman band of the P—O group (960 cmⳮ1) in the tenth spectrum suggests that this represents the bottom of the demineralized dentin layer. Dentin was demineralized to a depth of ⬃7 m, and the depth of partially demineralized dentin was ⬃3 m (Fig. 16b). The spectra recorded at the second, fourth, and sixth micrometer position of the a/d interface are presented in detail, and major spectral changes have been marked with arrows (Fig. 16c). The intensity of the Raman bands attributed to the adhesive (1113, 1187, 1454, 1609, and 1720 cmⳮ1) decreased as a function of depth, indicating the gradual decrease of adhesive infiltration into demineralized dentin. To determine the differences in composition of AIDD and the a/d interface at different depths, the ratios of the relative integrated intensities of bands associated with adhesive and the band associated with collagen (amide I) were calculated (Table 3). The composition of SB is 60–70 wt% BisGMA and 40–30 wt% HEMA [99]. The band at 1454 cmⳮ1, which is assigned to the CH2 group of both BisGMA and HEMA, and the band at 1113 cmⳮ1, which is assigned to the C—O—C group of BisGMA, were selected as measures of resin adhesive and BisGMA monomer, respectively. As listed in Table 3, the mean ratios of 1454 cmⳮ1/1667 cmⳮ1 and 1113 cmⳮ1/1667 cmⳮ1 for AIDD are 0.927 and 0.60. These ratios are 0.913 and 0.41 for the a/d interface at the first micrometer, and decrease to 0.162 and 0.052, respectively, at the depth of the demineralized dentin. Figures 17a and b show these relative ratios as a function of spatial position across the a/d interface. These ratios for AIDD are also shown in Fig. 17 a and b. As compared with the ratios for AIDD, the difference in the ratios of 1454/1667 and 1113/1667 of the a/d interface increased as a function of position, indicating less diffusion of resin monomers across the interface. The ratio of 1454/1667 shows a gradual decline, while the ratio of 1113/
620
Spencer et al.
Figure 16 Light micrograph (a) and corresponding Raman mapping spectra (b) acquired at 1-m intervals across Single Bond adhesive/dentin interface. The spectra marked with arrows (c) were recorded from sites corresponding to the demarcations noted on the light micrograph. (From Ref. 98.)
1667 (BisGMA/collagen) shows a dramatic decrease in the concentration of BisGMA monomer across the a/d interface. In order to quantify the penetration of the adhesive resin monomers, based on the results presented in Table 3 and 17, we defined the AIDD prepared under controlled conditions as an optim hybrid which has 100% adhesive infiltration. A comparison of the penetration of BisGMA monomer and BisGMA/HEMA resin as a function of position across the a/d interface is presented in Fig. 17c and Table 3. This comparison provided a quantitative representation of the percent of adhesive monomer penetration as a function of position. It was shown that ⬃98% resin monomers penetrated the first micrometer of the interface, but only ⬃68% BisGMA monomer penetrated the first micrometer of demineralized dentin. The percent of the BisGMA/HEMA resin monomers penetrating the demineralized dentin drops gradually from ⬃89% at the second micrometer to ⬃71% at the third micrometer, and at the depth of the demineralized dentin (⬃8 m), the percent of BisGMA/HEMA resin monomers is ⬃18%. In comparison, the percent of BisGMA monomer dropped very rapidly, from ⬃42% at the second micrometer to ⬃32% at the third micrometer, from this position to the bottom of the demineralized zone of dentin, the percent of BisGMA penetration is only ⬃8% as compared to the optimal hybrid.
Hyperspectral Analysis of Polymeric Adhesive
621
Table 3 The Raman Intensity Ratios and Percent of Monomer Penetration Raman intensity ratioa Sample and position AIDD a/d interface 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 10th
1454/1667
1113/1667
0.927 ⫾ 0.046
0.600 ⫾ 0.031
0.913 ⫾ 0.054 0.820 ⫾ 0.049 0.653 ⫾ 0.043 0.584 ⫾ 0.045 0.393 ⫾ 0.029 0.225 ⫾ 0.021 0.186 ⫾ 0.019 0.162 ⫾ 0.024 0.142 ⫾ 0.017 0.082 ⫾ 0.016
0.410 ⫾ 0.034 0.250 ⫾ 0.022 0.193 ⫾ 0.024 0.178 ⫾ 0.019 0.096 ⫾ 0.012 0.063 ⫾ 0.011 0.062 ⫾ 0.010 0.052 ⫾ 0.016 0.048 ⫾ 0.013 0 ⫾ 0.009
Percent of penetration (%) BisGMA/HEMA 100 98.5 88.5 70.5 63.0 42.4 24.3 20.1 17.5 15.3 8.85
BisGMA 100 68.3 41.7 32.2 29.7 16.0 10.5 10.3 8.7 8.0 0
Mean ⫾ standard deviation. Source: Ref. 98. a
B. Scanning Acoustic Microscopy/Calibration Analyses The SAM micrograph for the demineralized dentin collagen without adhesive infiltration is shown in Fig. 18. Where the collagen network is dense, the uniform gray level indicates that the properties are alike. The random spaces in the collagen network are responsible for the few bright (white) regions on the micrograph as these represent reflections from the glass slide on which the sample is mounted during the scanning process. The SAM micrograph of the demineralized dentin collagen infiltrated with Single Bond adhesive (AIDD) is shown in Fig. 19 [108]. On this micrograph there is a significant number of uniform brighter images interspersed in the spaces between the collagen fibers; these likely are the regions where the Single Bond material has infiltrated the pores in the collagen network. The higher Z (and correspondingly the higher E) value of Single Bond relative to collagen’s respective values is responsible for the brighter image, i.e., the greater the AI value, Z, the greater the reflection coefficient, r, and thus the brighter the image. Similarly, there are the black regions overlaying some of the tubular collagen structures. As the collagen fibril orientation and thickness are random, it is not possible during sample preparation to provide a specimen that is everywhere in the same focal plane. Thus, these black regions are artifacts due to differences in height relative to the rest of the specimen, i.e., these are the out-of-focus regions. Voltage values were measured at ten different points on the respective gray levels for both the collagen and the Single Bond infiltrated collagen samples [108]. The gray levels for the collagen on both samples were below the lowest values on the calibration curves based on known materials, consistent with our previous study [61]. Measurements were interpreted using the values represented in Table 2 and the relationship presented in Fig. 6 in order to obtain valid values for both the as-is and Single Bond infiltrated collagen samples from the SAM scans. The equality between Young’s modulus for the higher (bulk) acoustic waves and Young’s modulus for the lower (bar) acoustic waves for each calibration material, with the exception of PMMA, is shown in Fig. 6 as well as in Table 2. The interpretation is that these 12 materials are not exhibiting dispersion over this range of frequencies. Therefore, it is now possible to
622
Spencer et al.
Figure 17 Raman intensity ratios of 1454/1667 (a) and 1113/1667 (b) and penetration of BisGMA monomer and BisGMA/HEMA (c) as a function of spatial position across the adhesive/dentin interface. The comparable relative ratios calculated from the AIDD represent the optimal hybrid and are presented for comparison. (From Ref. 98.)
Hyperspectral Analysis of Polymeric Adhesive
623
Figure 18 400 MHz SAM image of the demineralized dentin collagen without adhesive infiltration (xscan width 250 m). Bright regions are reflections from the glass slide on which the sample is mounted.
extend the calibration process to extrapolation below the lowest polymer values in order to obtain good approximations of Young’s moduli for the collagen gray levels displayed on both SAM micrographs. Young’s modulus values for the collagen from both samples as well as for the Single Bond adhesive from the infiltrated sample are given in Table 4. C. Imaging Electromagnetic Properties of Biological and Organic Specimens Figure 20 shows the SEM and AFM images of demineralized dentin. Figures 21a and b show the AFM and EMP of an infiltrated dentin sample. The EMP clearly shows a region of high electrical activity and permittivity in certain regions of the sample. We are currently investigating the origin of this nonuniformity. EMP, owing to the penetration of its sensing signal inside the sample, is capable of detecting the properties of adhesive dentin interface. When used along with SAM and RS, important and critical information regarding bonding strength and uniformity is obtained. IV. DISCUSSION Dentin can be regarded as a biological composite of a collagen matrix which is highly filled with nanometer-sized apatite crystals. After extracting the mineral from the collagen fibrils, the
624
Spencer et al.
Figure 19 400 MHz SAM image of the demineralized dentin collagen infiltrated with Single Bond dentin adhesive (x-scan width 250 m). The black regions overlaying some of the tubule structures are artifacts due to differences in height relative to the rest of the specimen, i.e., these regions are out of the focal plane.
Table 4 Elastic Moduli of Dentin/Adhesive Interface Components Component Collagen Collagen infiltrated with adhesive Adhesive infiltrate Dentin Adhesive Source: Ref. 108.
Elastic modulus (GPa) 1.76 ⫾ 0.00 1.84 ⫾ 0.65 3.40 ⫾ 1.00 28* 5*
Hyperspectral Analysis of Polymeric Adhesive
625
Figure 20 (a) SEM of the demineralized dentin (scale bar 5 m). (b) 50 ⳯ 50 m noncontact AFM scan of the same sample. (c) SEM of the infiltrated dentin (scale bar 5 m2); scan size 100 ⳯ 80 m2).
Figure 21 (A) 9 ⳯ 9 m2 noncontact AFM scan of infiltrated dentin. (B) EMP scan of the same region (white areas have higher electrical conductivity).
626
Spencer et al.
voids can be filled with a resin, thus forming a new composite made up of resin matrix filled with a fibrous collagen. This new structure is a hybrid of resin and collagen [100]. As shown in Fig. 15, this optimal hybrid structure was achieved when adhesive resin infiltrated demineralized dentin prepared under controlled conditions (Figs. 15b and c), but not at the a/d interface formed using the wet bonding technique. Using the histomorphologic technique described here and reported previously [33,42], any collagen that is not encased in adhesive resin is available for reaction with the Goldner’s trichrome stain. The section of AIDD specimen without total removal of ethanol solvent stains slightly. The distinct red zone at the a/d interface (Fig. 15a) indicated that the adhesive did not penetrate the full depth of the demineralized layer, i.e., it did not encapsulate the collagen fibrils throughout the width of the demineralized dentin. This novel staining technique, which identifies exposed collagenous protein at the light microscopic level, provides a unique, clear representation of the extent and degree to which the adhesive resins envelope the collagen fibrils of the demineralized dentin matrix. Corresponding Raman spectra recorded from the intertubular area of the a/d interface and the AIDD sections confirmed the above observation. The spectrum of a/d interface has a lower adhesive contribution as compared to the spectrum of the AIDD, indicating limited diffusion of resin monomers into the wet a/d interface. The Raman spectral bands collected as a function of position across the a/d interface also showed a gradual decrease in adhesive concentration. Gradual decreases in intensity of Raman bands associated with adhesive have been reported by previous authors [101,102]. However, in these studies, the authors measured the amount of resin penetration based on the absolute intensity of adhesive bands [102], totally ignoring the collagen matrix, the key component of acid-etched dentin. The absolute intensity of backscattering Raman band is dramatically affected by many factors, such as the smoothness of the sample surface, the position of focusing, the depth of detection, the fluorescence of biological components and the stability of the instrument and laser power, etc [103,104]. Since it is very difficult to maintain all of the above conditions the same across the breadth of the sample, the band intensity may vary from one measurement to the next even at the same spot on the sample. In order to account for the effect of instrumental fluctuation and errors, the Raman band associated with the collagen matrix was used as internal standard throughout our studies [33–34,43,48]. The quantitative band intensity ratios of adhesive and collagen across the a/d interface, indicating the relative amount of adhesive, are presented in Fig. 17 and the corresponding Table 3. Although there is substantial evidence to suggest that the resin–dentin interdiffusion zone is porous [44,105,106], there is no technique available to measure the extent of adhesive infiltration at the a/d interface and to determine how adhesive penetration at the interface relates to the condition of complete infiltration. In our previous studies [33,43], the weight percent of adhesive at each micrometer as a function of spatial position across the a/d interface was determined. To facilitate the comparison of adhesive penetration across the breadth of the interface, we assumed that the adhesive fully penetrated the first micrometer of the a/d interface. However, to fully identify, characterize, and understand the weak links in the a/d bond it is important for us to know if the adhesive occupied all the space left by mineral after etching. By preparing optim hybrid specimens using the slab adjacent to the a/d interface, a new method which truly calculates the extent of adhesive penetration at the a/d interface as compared to a resin-infiltrated sample was proposed [98]. As shown in Fig. 17c, even at the first micrometer, only ⬃68% of the concentration of BisGMA in the original adhesive penetrated the demineralized dentin. In comparison, the resin components (including HEMA) diffused more readily into the demineralized dentin zone than the BisGMA component. Phase separation of the adhesive that infiltrates the demineralized dentin matrix would compromise the structural integrity of the resultant hybrid layer. In contrast to an impervious three-dimensional collagen/polymer network, adhesive phase separation would lead to a very
Hyperspectral Analysis of Polymeric Adhesive
627
porous hybrid layer characterized by hydrophobic BisGMA-rich particles distributed in a hydrophilic HEMA-rich matrix [99]. Because of its low crosslink density HEMA is unstable in aqueous environments, and thus this phase will degrade when exposed to oral fluids [107]. In addition, unprotected demineralized collagen fibrils are more susceptible to hydrolysis than are mineralized collagen fibrils. This may explain the decreased bond strength in aqueous environments over the long term. The results of the complementary morphologic and spectroscopic analyses suggest that under wet bonding the a/d interface is not an impervious collagen/polymer network but a porous web; the composition of this web is predominately collagen and HEMA with less contribution from the BisGMA component. The results of this study suggest that the critical dimethacrylate component (BisGMA), which contributes the most to the crosslinked polymeric adhesive, infiltrates a fraction of the total wet demineralized, intertubular dentin layer. As shown in our previous investigation, limited infiltration of the BisGMA component impacts the mechanical properties of this interface [61]. It is conceivable that a mismatch in the modulus of elasticity values of the collagen and infiltrating adhesive at this interface could impact the transfer of stress and lead to unequal load sharing during occlusal loading and/or function. A. Complementary Micromechanics of the Adhesive-Infiltrated Dentin Previously we had indicated that we could only provide an upper limit of 2 GPa for Young’s modulus, E, of the unprotected protein at the dentin/adhesive interface [61]. The lowest values of r vs. V, those for the several polymers on our calibration curves (Fig. 3–5), were above those measured for the unprotected protein at the interface [61]. There also was the concern of dispersion, i.e., the dependence of E on frequency for viscoelastic materials such as the polymers and the unprotected protein. It was these concerns that led to the present SAM and SEMP studies of the properties of the demineralized dentin collagen both with and without the infusion of the Single Bond adhesive. We also extended our calibration procedure to include the effects of frequency on Young’s modulus (Tables 1 and 2 and Fig. 6). The extensional longitudinal ultrasonic velocity, L, provides a direct measurement of the bar wave Young’s modulus, E ⳱ PL2, where n˜ is the density. This value is equivalent to that obtained in quasistatic mechanical stress–strain experiments, except for dispersive materials, where it will generally be higher. However, if both shear wave velocities, t, and dilatational longitudinal wave velocities, l, are measured at higher frequencies it is possible to calculate values of the two Lame´ constants, assuming an isotropic material, which then can be used to obtain the shear, G, and bulk, B, moduli. These can then be used to calculate Young’s modulus, E ⳱ 9KG/(3KⳭG); this is often designated the bulk wave modulus. The strong correlation between E(bar) and E(bulk), even including the polymers used in this calibration are shown in both Fig. 6 and Table 2. This does not indicate what dispersive effects there might be for moduli measured by the low strain rate quasistatic mechanical measurements and the bar wave ultrasonic measurements for these polymers or the interface material. However, at worst it does provide a reasonable measure of the values for these materials over a wide range of ultrasonic frequencies. This analysis does provide the rational basis for extrapolations below the lowest known measurable materials used in the calibration process in order to obtain more realistic values of the moduli for the collagen samples analyzed here as well as for the unprotected protein at the interface. In addition further support for this calibration procedure is provided by analyzing the relationship between the shear wave velocities, t, and the dilatational longitudinal wave velocities, vl, for a large number of disparate materials (Table 1); the average value of the ratio vt/vl for the 13 materials used in this calibration is 0.48 Ⳳ 0.09. This justifies correlating the bar wave values with the values obtained at the high frequency longitudinal velocities used in the SAM measure-
628
Spencer et al.
ments in the calibration procedure of obtaining Young’s modulus for the known materials. The values of Young’s moduli for both the as-is and infiltrated demineralized dentin collagen, as well as for the Single Bond adhesive infiltrate, were based on this extensive calibration procedure (Table 4). Finally, we used a new local probe that uses decaying or evanescent microwave to image nonuniformity in different materials with a resolution much better than its wavelength. This nonintrusive technique is capable of imaging nonuniformities in organic and inorganic conductors, semiconductors, and insulators. Its resolution is determined by its operation frequency and probe parameters, such as substrate thickness, taper angle, diameter of its wire tip, etc. We discussed some imaging examples of EMP to demonstrate its versatility and scope of its applications as well. The EMP technique can result in quantitative information and it can be scanned over samples at a very fast rate of 1 cm/s. EMPs operating at different frequencies can be used in conjunction with micro-Raman spectroscopy and scanning acoustic microscopy to perform hyperspectral imaging of biological tissues, biomaterials, and their interfaces.
ACKNOWLEDGMENTS A contribution of the UMKC Center for Research on Interfacial Structure and Properties (CRISP), this investigation was supported in part by USPHS Research Grant DE 12487 from the National Institute of Dental and Craniofacial Research, National Institutes of Health, Bethesda, MD 20892. The authors gratefully acknowledge 3M, Dental Products Division, for donating the dentin adhesive products used in this study. The work was partially supported by grants from WPAFM (through TMC and MICC) and by grants from NIST and NSF.
REFERENCES
1. Tabak LA. NIDCR Director’s Statements Regarding the FY 2002 Budget for House and Senate Appropriations Subcommittees(). 2001. 2. Kidd EAM, Toffenetti F, Mjor IA. Secondary caries. Int Dent J 1992; 42:127–138. 3. Kidd EAM, Beighton D. Prediction of secondary caries around tooth-colored restorations: a clinical and microbiological study. J Dent Res 1996; 75(12):1942–1946. 4. Collins CJ, Bryant RW, Hodge KLV. A clinical evalutation of posterior composite resin restorations: 8-year findings. J. Dent 1998; 26:311–317. 5. Raskin A, Michottetheall B, Vreven J, Wilson NHF. Clinical evaluation of a posterior composite 10-year report. J. Dent 1999; 27:13–19. 6. Schriever A, Becker J, Heidemann D. Tooth-colored restorations of posterior teeth in German dental education. Clin Oral Invest 1999; 3:30–34. 7. Hunter AR, Treasue ET, Hunter AJ. Increases in cavity volume associated with the removal of class 2 amalgam and composite restorations. Operative Dent 1995:2–6. 8. Van Meerbeek B, Vargas M, Satoshie I, Yoshida Y, Perdigao J, Lambrechts P, Vanherle G. Microscopy investigations. Techniques, results, limitations. Am J Dent 2000; 13(Special Issue, November): 3–18. 9. Tobi H, Kreulen CM, Vondeling H, van Amerongen WE. Cost-effectiveness of composite resins and amalgam in the replacement of amalgam class II restorations. Commun Dent Oral Epidemiol 1999; 27:137–143. 10. Beznos C. Microleakage at the cervical margin of composite class II cavities with different restorative techniques. Operative Dent 2001; 26:60–69.
Hyperspectral Analysis of Polymeric Adhesive
629
11. Mair LH. Ten-year clinical assessment of three posterior resin composites and two amalgams. Quintessence Int 1998; 29:483–490. 12. Nordbo H, Leirskar J, von der Fehr FR. Saucer-shaped cavity preparations for posterior approximal resin composite restorations: observations up to 10 years. Quintessence Int 1998; 29:5–11. 13. Dunne SM, Gainsford ID, Wilson NHF. Current materials and techniques for direct restorations in posterior teeth. Part 1: silver amalgam. Int Dent J 1997; 47:123–136. 14. Svanberg M, Mjor IA, Orstavik D. Mutans Streptococci in plaque from margins of amalgam composite, and glass-ionomer restorations. J Dent Res 1990; 69(3):861–864. 15. Chan MF, Glynn Jones JC. A comparison of four in vitro marginal leakage tests applied to root surface restorations. J Dent Res 1992; 20:287–293. 16. Ferdianakis K. Microleakage reduction from newer esthetic restorative materials in permanent molars. J. Clin. Pediatr. Dent 1998; 23:221–229. 17. Marshall SJ, Marshall GW, Kinney JH, Balooch M. Dental restorative material—tooth interfaces. Scripta Metallurgica et Materials 1994; 31:983–988. 18. Roulet JF. Benefits and disadvantages of tooth-coloured alternatives to amalgam. J. Dent 1997; 25: 459–473. 19. LeGeros RZ. Calcium Phosphates in Oral Biology and Medicine. Basel. Karger, 1991. 20. Marshall GW, Marshall SJ, Kinney JH, Balooch M. The dentin substrate: structure and properties related to bonding. J Dent 1997; 25:441–458. 21. Butler WT. Dentin extracellular matrix and dentinogenesis. Operative Dent 1992; 5(Suppl):18–23. 22. Wang R, Weiner S. Human root dentin: structure anisotropy and Vickers microhardness isotropy. Connective Tiss Res 1998; 39:269–279. 23. Nakabayashi N, Pashley DH. Hybridization of Dental Hard Tissues: Quintessence Publishing London, 1998. 24. Chappell RP, Eick JD. Shear bond strength and scanning electron microscopic observation of six current dentinal adhesives. Quintessence Int 1994; 25:359–368. 25. Eick JD, Cobb CM, Chappell RP, Spencer P, Robinson SJ. The dentinal surface: its influence on dentinal adhesion. Part III. Quintessence Int 1993; 24:571–582. 26. Perdigao J, Denehy GE, EJ Swift. Silica contamination of etched dentin and enamel surfaces: a scanning microscopic and bond strength study. Quintessence Int 1994; 25:327–333. 27. Tam LE, Pilliar RM. Fracture surface characterization of dentin-bonded interfacial fracture toughness specimens. J Dent Res 1994; 73(3):607–619. 28. Nakabayashi N, Kojima K, Masuhara E. The promotion of adhesion by the infiltration of monomers into tooth substrates. J. Biomed. Mater. Res 1982; 16:265–273. 29. Nakabayashi N, Saimi Y. Bonding to intact dentin. J Dent Res 1996; 75(9):1706–1715. 30. Eick JD, Gwinnet AJ, Pashley DH, Robinson SJ. Current concepts on adhesion to dentin. Crit Rev Oral Biol Med 1997; 8:306–335. 31. Gwinnett AJ. Quantitative contribution of resin infiltration hybridization to dentin bonding. Am. J. Dent 1993; 6:7–9. 32. Inokoshi S, Nakaoki Y, Shono T, Pereira PNR, Yamada T, Tagami J. Collagen layer collapse and inhibited resin penetration by dentin etching. J. Dent. Res噛 1996; 75:389. 33. Spencer P, Wang Y, Walker MP, Wieliczka DM, Swafford JR. Interfacial chemistry of the dentin/ adhesive bond. J. Dent. Res 2000; 79:1458–1463. 34. Wang Y, Spencer P. Analysis of acid-treated dentin smear debris and smear layers using confocal Raman microspectroscopy. J. Biomed. Mater. Res 2002; 60:300–308. 35. Pashley DH, Ciucchi B, Sano H, Horner JA. Permeability of dentin to adhesive agents. Quintessence Int 1993; 24:618–631. 36. Kinney JH, Balooch M, Marshall SJ, Marshall GW. Atomic force microscope study of dimensional changes in dentine during drying. Arch. Oral Biol 1993; 38:1003–1007. 37. Kanca J. Effect of resin primer solvents and surface wetness on resin composite bond strength to dentin. Am. J. Dent 1992; 5:213–215. 38. Kanca J. Improved bond strength through acid etching of dentin and bonding to wet dentin surfaces. J. Am Dent Assoc 1992; 123:235–43.
630
Spencer et al.
39. Gwinnett AJ. Dentin bond strength after air drying and rewetting. Am. J. Dent 1994; 7:144–148. 40. Burrow MF, Satoh M, Tagami J. Dentin durability after three years using a dentin bonding agent with and without priming. Dent Mater 1996; 12:302–307. 41. Hashimoto M, Ohno H, Kaga M, Endo K, Sano H, Oguchi H. In vivo degradation of resin–dentin bonds in humans over 1 to 3 years. J Dent Res 2000; 79:1385–1391. 42. Spencer P, Swafford JR. Unprotected protein at the dentin–adhesive interface. Quintessence Int 1999; 30:501–507. 43. Wang Y, Spencer P. Quantifying adhesive penetration in adhesive/dentin interface using confocal Raman microspectroscopy. J. Biomed. Mater. Res 2002; 59:46–55. 44. Sano H, Takatsu T, Ciucchi BJAH, Matthews WG, Pashley DH. Nanoleakage: leakage within the hybrid layer. Operative. Dent 1995; 20:18–25. 45. Chase B. A new generation of Raman instrumentation. Appl. Spectrosc 1994; 48:14A–18A. 46. Turrell G, Corset J. Raman Microscopy—Developments and Applications. New York: Academic Press, 1996:1–12. 47. Meier RJ. MicroRaman spectroscopy: application to polymeric and inorganic systems. Int J Electronics 1995; 78(2):285–288. 48. Spencer P, Wang Y, Walker MP, Swafford JR. Molecular structure of acid-etched dentin smear layers—in situ study. J. Dent. Res 2001; 80:1802–1807. 49. Lemons RA, Quate CF. Acoustic microscope-scanning version. Apply Phys Lett 1974; 24:163–165. 50. Lemons RA, Quate CF. Acoustic Microscopy. Physical Acoustics 1979; 14:1–92. 51. Briggs GAD. Acoustic Microscopy. Oxford: Clarendon Press, 1992. 52. Meunier A, Katz JL, Christel P, Sedel L. A reflection scanning acoustic microscope for bone and bone–biomaterials interface studies. J Orthop Res 1988; 6:770–775. 53. Meunier A, Riot O, Katz JL, Christel P, Sedel L. Inhomogeneities in anisotropic elastic constrants of cortical bone. IEEE Ultrasonic Symposium, IEEE. New York, 1989:1015–1018. 54. Zimmerman MC, Meunier A, Katz JL, Christel P. The evaluation of cortical bone remodeling with a new ultrasonic technique. Trans Biomed Eng 1990; 37(5):433–441. 55. Turner CH, Chandran A, Pidaparti RM. The anisotropy of osteonal bone and its ultrastructural implications. Bone 1995; 17:85–89. 56. Takano Y, Turner CH, Burr DB. Mineral anistropy in mineralized tissue is similar among species and mineral growth occurs independently of collagen orientation in rats: results form acoustic velocity measurements. J. Bone Miner. Res 1996; 11:1292–1301. 57. Katz JL, Meunier A. Scannning acoustic microscope studies of the elastic properties of osteons and osteon lamellae. J Biomechanical Eng 1993; 115:543–548. 58. Katz JL, Meunier A. Material properties of single osteons and osteonic lamellae using high frequency scanning acoustic microscopy. In: Odgaard A, Weinans H, eds. Bone Structure and Remodeling. Singapore: World Scientific, 1995:157–165. 59. Hasegawa K, Turner CH, Recker RR. Elastic properties of osteoporotic bone measured by scanning acoustic microscopy. Bone 1995; 16:85–90. 60. Bumrerraj S, Katz JL. Scanning acoustic microscopy study of human cortical and trabecular bone. Ann. Biomed. Eng 2001; 29:1–9. 61. Katz JL, Bumrerraj S, Dreyfuss J, Wang Y, Spencer P. Micromechanics of the dentin/adhesive interface. J. Biomed. Mater. Res. Appl Biomater 2001; 58:366–371. 62. Bethe HA. Theory of diffraction by small holes. Phys Rev 1944; 66(7):163–182. 63. Baez V. Is resolving power independent of wavelength possible? an experiment with a sonic macroscope. J Optical Soc Am 1956; 46(10):901. 64. Soohoo RF. A microwave magnetic microscope. J Appl Phys (Suppl) 1962; 33(3):1276–1277. 65. Ash EA, NIcholls G. Super-resolution aperture scanning microscope. Nature 1972; 237:510–512. 66. Pohl DW, Denk W, Lanz M. Optical stethoscopy: image recording with resolution 2J20. Appl. Phys. Lett 1984; 44(7):651–653. 67. Lukosz W. Imagery with evanescent waves. Workshop on Unconventional Imagery. 1984:73–74. 68. Weiss JA, Hawks DA. Dielectric constant evaluation of insulating materials: an accurate, practical measurement systems. IEEE MTT-S Digest. 1987:457–460.
Hyperspectral Analysis of Polymeric Adhesive
631
69. Gutmann RJ, Borrego JM, Chakrabarti P, Wang MS. Microwave scanning microscopy for planar structure diagnostics. IEEE MTT-S Digest. 1987:281–284. 70. Fee M, Chu S, Hansch TW. Scanning electromagnetic transmission line microscope with subwavelength resolution. Optics Commun 1989; 69(3,4):219–224. 71. Tabib-Azar M, Shoemaker N, Harris S. Non-destructive characterization of materials by evanescent microwaves. Measurement Sci Technol 1993; 4:583–590. 72. Stranick SJ, Weiss PS. Alternating current scanning, tunneling microscopy and nonlinear scpectroscopy. J. Phys. Chem 1994; 98:1762–64. 73. Gao X, Duewer F, Lu Y, Xiancy XD. Hi-h spatial resolution quantitative microwave impedance microscopy by a scanning tip microwave near-field microscope. Appl. Phys. Lett 1997; 71(13): 1872–1874. 74. Golosovksy M, Lann A, Davidov D. A millimeter-wave near-field scanning probe with an optical distance control. Ultramicroscopy 1998; 71:133–141. 75. Thanawalla S, Dutta SK, Vlahacos CP, Steinhauser DE, Feenstra BJ, Anlage SM, Wlllstood FC. Microwave near-field imaging of electric fields in a superconducting microstrip resonator. Appl. Phys. Lett 1998; 73(17):2491–2493. 76. Tabib-Azar M, Katz JL, LeClair S. Evanescent microwaves: a novel, super-resolution, noncontact and nondestructive imaging technique for biological applications. IEEE Trans. Instr Measurement 1999; 48(6):1111–1116. 77. Tabib-Azar M, Bumrerraj S, Katz JL, You SH. Evanescent electromagnetics: a novel, super-resolution, and non-intrusive imaging technique for biological applications. Biomed Microdevices 1999; 2(1):73–80. 78. Tabib-Azar M, Su DP, Pohar A, Leclair SR, Ponchak G. 0.4 m spatial resolution with 1 GHz evanescent microwave probe. Rev Sci Instr 1999; 70(3):1725–1729. 79. Tabib-Azar M, Pathak PS, Ponchak G, Leclair SR. Nondestructive superresolution imaging of defects and nonuniformities in metals, semiconductors, dielectrics, composites and plants using evanescent microwaves. Rev Scientific Instr 1999; 70(6):2783–2792. 80. Tabib-Azar M, Akinwande D, Ponchak G, Leclair SR. Evanescent microwave probes on highresistivity silicon and its application in characterization of semiconductors. Rev Sci Instr 1999; 70(7):3083–3086. 81. Tabib-Azar M, Akinwande D, Ponchak G, Leclair SR. Novel physical sensors using evanescent microwave probes. Rev Scientific Instr 1999; 70(8):3381–3386. 82. Tabib-Azar M, Ciocan R, Ponchak G, Leclair SR. Transient thermography using microwave microscope. Rev Sci Instr 1999; 70(8):3387–3391. 83. Tabib-Azar M, Sutapun B. Novel hydrogen sensors using evanescent microwave probes. Rev Sci Instr 1999; 70(9):3707–3714. 84. Tabib-Azar M, Ciocan R, Leclair SR. Transient thermography of semiconductors using microwave microscope. Microscale Thermophys Eng 1999; 3(4):321–327. 85. Tabib-Azar M-, Valenzuela A, Ponchak G. Evanescent Microwave Microscopy for High Resolution Characterization of Materials. Boston: Kluwer Academic Publishing, 2000. 86. Tabib-Azar M, Akinwande D. Real-time imaging of semiconductor space-charge regions using high-spatial resolution evanescent microwave microscope. Rev Sci Instr 2000; 71(3):1725–1730. 87. Tabib-Azar M, Leclair SR. Applications of evanescent microwave probes in gas and chemical sensors. Sensors and Actuators B 2000; 67:112–121. 88. Tabib-Azar M, Muller R, Leclair SR. Nondestructive imaging of grain boundaries in polycrystalline materials using evanescent microwave probes. Eng. Appl. Artif intelligence 2000; 13:549–564. 89. Tabib-Azar M, Wang R. Planar evanescent microwave imaging probes for nondestructive evaluation of materials with very high spatial resolution and scan rates. Rev Prog Quantitat Nondestr Eval 2000; 20A:446–453. 90. Tabib-Azar M. Microwave microscopy and its applications. Rev Prog Quantitat Nondestr Eval 2000; 20A:400–413. 91. Tabib-Azar M. Evanescent microwave microscope: a new nondestructive material evaluation tool with very high resolutions. Can Soc Nondestr Testing 2000(January).
632
Spencer et al.
92. Tabib-Azar M. Applications of ultra-high resolution evanescent microwave imaging probe in nondestructive evaluation of materials. Mater Eval 2001:70–78. 93. Katz JL, You HS, Tabib-Azar M-, Azar M, Bumrerraj S. Electro-Mechanical Studies of Tooth structure and Caries at High Resolution. In: Stookey GK, ed. Early Detection of Caries II. Bloomington. IN: Indiana University Press, 2000:443–449. 94. Wang Y, M Tabib-Azar. AFM-compatible microwave probes for nondestructive surface measurement. In: ASNT 11th Annual Research Symposium . Portland. OR, 2002. 95. Wang Y, Scott A, Zhang T, M Tabib-Azar. Microwave atomic force microscopy. In: ASNT 11th Annual Research Symposium. Portland. OR, 2002. 96. M Tabib-Azar. Recent progress in microwave microscopy and its applications. In: ASNT 11th Annual Research Symposium. Portland. OR, 2002. 97. M Tabib-Azar. Applications of microwave microscopy in materials devices and systems: an overview. In: ASNT 11th Annual Research Symposium. Portland. OR, 2002. 98. Wang Y, Spencer P. Hybridization efficiency of the adhesive dentin interface with wet bonding. J. Dent. Res 2003; 82(2):141–145. 99. Spencer P, Wang Y. Adhesive phase separation at the dentin interface under wet bonding conditions. J. Biomed. Mater. Res 2002; 62:447–456. 100. Nakabayashi N. The hybrid layer: a resin–dentin composite. Proc. Finn. Dent. Soc. 88(Suppl.) 1992; 1:321–329. 101. Van Meerbeek BHM, Celis JP, Roos JR, Braem M, Lambrechts P, Vanherle G. Chemical characterization of the resin–dentin interface by micro-Raman spectroscopy. J. Dent. Res 1993; 72:1423–1428. 102. Hashimoto M, Ohno H, Kaga M, Sano H, Endo K, Oguchi H. The extent to which resin can infiltrate dentin by acetone-based adhesives. J. Dent. Res 2002; 81(1):74–78. 103. Chase DB. Modern Raman instrumentation and techniques. In: Grasselli JG, Bulkin BJ, eds. Analytical Raman Spectroscopy. New York: John Wiley & Sons, 1991:21–43. 104. Walton ADMJ, Koenig JL. Raman spectroscopy of calcified tissue. Calc. Tiss. Res 1970; 6(2): 162–167. 105. Sano H, Shono T, Takateu T, Hosoda H. Microporous dentin zone beneath resin-impregnated layer. Operative Dent 1994; 19:59–64. 106. Tay FR, Gwinnett AJ, Pang KM, Wei SHY. Variability in microleakage observed in a total-etch wet-bonding technique under different handling conditions. J. Dent. Res 1995; 74:1168–1178. 107. Yourtee DM, Smith RE, Russo KA, Burmaster S, Cannon JM, Eick JD, Kostoryz EL. The stability of methacrylate biomaterials when enzyme challenged: kinetic and systematic evaluations. J Biomed Mater Res 2001; 57(4):523–531. 108. Katz JL, Spencer P, Nomura T, Wagh A, Wang Y. Micromechanical properties of demineralized dentin collagen with and without adhesive infiltration. J Biomed Mater Res, in press.
Index
Acidification, of microspheres, 311–312 Adenovirus-coupled polymers and peptides, 338 Adhesion measurement at blood-contacting surfaces, 55–74 Adhesive-infiltrated demineralized dentin (AIDD), 619–620, 626–628 Adhesives, 525–528, 599–632 Adjunctive methods of improved patency of ePTFE grafts, 85–99 bonding of antithrombogenic agents, 98–99 surface properties, 85–98 A. latus, 287–288, 289, 291 Albumin, 379, 394, 450–452 (see also Proteins) Alginate, 340 Angioplasty, percutaneous transluminal coronary, 107 Animal studies, 61–62, 82, 132–133, 476 Anionic emulsion polymerization, 566–568 Antacid excipients, coencapsulating, 318 Anterior cruciate ligament (ACL), 215 (see also Ligament reconstruction) Antibiotic release, and microbial polyesters, 281–303, 492–493, 548–550 antibiotic-loaded controlled release rods, 285–286 calcium hydroxyapatite, 287 characterization of PHAs, 284–285, 290–295 determination of PHAs, 284 PHA quantification, 290 production of HAP powder, 300 production of PHAs, 283–284, 287–290 in vivo studies, 286–287, 295–300 in vitro antibiotic release, 295 Anticancer drugs, 428–436 Antithrombogenic agents, bonding of, 98–99 Aortofemoral bypass grafts, 79
Applications, 75, 78–79, 80–81, 82–85, 122, 123, 141, 203–206, 211–213, 218–219, 260–261, 522, 530, 553–554 558, 573–586 (see also Polyhydroxybutyrate and its copolymers) aortofemoral, 79 auricular plastic reconstruction, 260–261 axillofemoral, 79 blood–brain barrier, 579–580 blood vessels, 211–213 bone graft substitute, 522 bowel defects, 558 bronchi, 218–219 burn therapy, 141, 530 cancers, 575, 576–579 cardiac, 80 carotid arteries, 79–80, 558 collagen, 522 (see also Collagen) corneal, 213–215, 522, 530 dental amalgam, 599–600 dialysis, 78–79 drug delivery systems, 575 endoskeletal scaffolding, 261 epidermis, 198–201 (see also Skin substitutes) femoropopliteal, 77–78 gastrointestinal, 123, 558 hard tissue, 553–554 hemostasis, 522 infection control, 522, 575 intracellular infections, 575 knee, 215–218 ligament, 215–218 medical device use, collagen, 522 neurovascular disease, 123 ocular, 575 oral mucosa, 530 633
634 [Applications] osteomyelitis, 282, 286 penile prostheses, 261 portal vein grafts, 80–81 reflux and incontinence, 261 retinal diseases, 556 skin substitutes, 141–154, 155–172 tibial bypass, 77–78 tissue formation, 554–558 tracheal, 122, 261–262 urinary incontinence, 522, 530 urology, 122 vaccines, 575 vascular grafts, and closure, 75–105, 522, 530 venocaval, 80–81 ventricular assist devices, 260 wound healing, 81–85, 155, 203–206, 522 (see also Skin substitutes) Aqueous core, selection of, 324–325 Aspirin, 99 Atelocollagen, 340 (see also Collagen) Atomic force microscopy, 189–190, 618, 623 Auricular cartilage tissue engineering, 253–264 applications, 260–262 auricular plastic reconstruction, 260–261 and endoskeletal scaffolding, 261 and growth factors, 262 penile prostheses, 261 reflux and incontinence, 261 tracheal augmentation, 261–262 ventricular assist devices, 260 backgound, 253–255 definition of cartilage, 254 embryology, 253–254 histology of cartilage, 254–255 chondrocytes from, 255–260 harvesting, 255–260 introduction, 253 research directions, 262–263 genetic engineering of chondrocytes, 262 transfection of chondrocytes, 262–263 summary, 263 Axillofemoral bypass grafts, 79
Baboons, and ePTFE, 82 Basement membrane (BM), keeping intact, 148 Basic fibroblast growth factor, 84, 262 Bioactive molecules and biodelivery systems, 369–392, 546–552, 554 (see also Biodegradable hydrogels as drug controlled release vehicles; Biodegradable nanoparticles as drug
Index [Bioactive molecules and biodelivery systems] delivery systems for parenteral administration; Biological delivery systems; Collagen; Nanoparticles; Nanosized biosensors and delivery vehicles; PHAs; Polymeric gene delivery systems) bioactive substances, 370–377 lipids, 376–377 nucleic acids and oligonucleotides, 377 peptides and proteins, 371–376 polysaccharides, 376 biodelivery systems, 377–386 cells, organs, and tissues, 386 lipid-based, 382–384 monoclonal antibodies, 380–381 nucleic acid based, 385–386 polysaccharides, 381–382 protein-based, 378–380 vaccine adjuvants, 381 conclusion, 386–387 introduction, 369–370 Bioadhesive nanoparticles, 583–584 Biochemical and ultrastructural features, of skin substitutes, 145–147 Biocompatibility, 20–23, 48–52, 397–398, 546–547 basis of, 48–52 proposed, 51–52 studies, 48–51 and extracellular matrix remodeling, 20–23 and PHAs, 546–547 Biodegradable hydrogels as drug controlled release vehicles, 423–461 (see also Bioactive molecules and biodelivery systems; Biodegradable nanoparticles as drug delivery systems for parenteral administration; Biological delivery systems; Collagen; Nanoparticles; Nanosized biosensors and delivery vehicles; PHAs; Polymeric gene delivery systems) dextran–maleic acid , 425–428 dextran–methacrylate, 428–436 dextran–polyactide hybrids, 436–453 chemical structure, 437 controlled release profiles of model drugs, 450–453 mechanical properties, 449–450 morphological structure, 443–446 swelling properties, 437–443 thermal properties, 447–449 emerging hydrogel systems, 453–456 introduction, 423–425
Index Biodegradable nanoparticles as drug delivery systems for parenteral administration, 395–422 (see also Bioactive molecules and biodelivery systems; Biodegradable hydrogels as drug controlled release vehicles; Biological delivery systems; Collagen; Nanoparticles; Nanosized biosensors and delivery vehicles; PHAs; Polymeric gene delivery systems) effect of a sterilization procedure, 414 formulated with natural macromolecules, 395 formulated with synthetic molecules, 395–414 poly(alkylcyanoacrylates), 395–398 polyesters and polyanhydrides, 398–399 polymer precipitation methods, 399–414 introduction, 393–394 polymer type and formation techniques, 394–395 Biodegradable polymers, 112, 281–303, 338–343, 475–482 (see also Nanoparticles; Nanosized biosensors and delivery vehicles; Nonbiodegradable polymers) bone morphogenetic protein delivery systems, 475–482 clinical applications, 479–480 polylactic acid, 476–479 Biodelivery systems, 369–392 (see also Bioactive molecules and biodelivery systems; Biodegradable hydrogels as drug controlled release vehicles; Biodegradable nanoparticles as drug delivery systems for parenteral administration; Biological delivery systems; Collagen; Nanoparticles; Nanosized biosensors and delivery vehicles; PHAs; Polymeric gene delivery systems) cells, organs, and tissues, 386 lipid-based, 382–384 monoclonal antibodies, 380–381 nucleic acid based, 385–386 polysaccharides, 381–382 protein-based, 378–380 vaccine adjuvants, 381 Bioengineered skin, 141–154 conclusions, 149–151 introduction, 141–142 skin substitutes, 141–142 results from functional skin substitutes, 142–149 basement membrane, keeping intact, 148 biochemical and ultrastructural features, 145–147
635 [Bioengineered skin] cultured on top, 148–149 dermis from ‘‘outdated’’ cryopreserved allografts, 142–144 Biological delivery systems, 471, 533–535 (see also Bioactive molecules and biodelivery systems; Biodegradable hydrogels as drug controlled release vehicles; Biodegradable nanoparticles as drug delivery systems for parenteral administration; Collagen; Nanoparticles; Nanosized biosensors and delivery vehicles; PHAs; Polymeric gene delivery systems) Biological fluids, rheology of, 265–280 mucus, 274–276 flow, 275–276 viscoelastic properties, 275 properties of biological fluids, 266–268 flow, 268 viscoelastic, 266–268 substitutes, 276–278 synovial fluid, 268–272 flow, 271–272 viscoelastic properties, 268–271 vitreous body, 272–274 flow, 274 viscoelastic properties, 272–273 Biomaterials, 107–130, 378, 505–520 and biodelivery systems, 378 and biosensors, 505–520 and stent technology, 107–130 Biomimetics, different approaches, 180–188 Biomolecular recognition principles, 507, 514 Bioresorbable stents, 111–113 Biosensors, 483–503, 505–520 (see also Nanosized biosensors and delivery vehicles) and delivery vehicles, nanosized, 483–503 relationship to biomaterials, 505–520 use of, 505–517 advantages and limitations, 509–510 agents, 506–507 contribution to the field, 516–517 deployment modalities, 507–509 operating principles, 511–513 specific examples, 514–516 in vivo operation, 510–511 Biostable stents, 111 Biosynthesis of collagen, 232 BisGMA-based polymeric adhesive, and collagen, 599–632 (see also Collagen) Block copolymers and polymer blends, 179–180
636 Blood–brain barrier, 579–580 Blood-contacting surfaces, adhesion measurement, 55–74 Blood vessel reconstruction, 211–213, 523 characteristics of the reconstructed blood vessel, 212–213 methods, 211–212 Bonding of antithrombogenic agents, 98–99 Bone marrow stroma, and growth factor delivery, 463–473 Bone morphogenetic protein delivery systems, 475–482 Bone repair, and growth factor delivery, 463–473 Bovine, 214, 256, 258 collagen, 214 fetal serum, 256 Bronchi reconstruction, 218–219 Bulk properties, of PHAs, 543–545 Burn therapy, 141, 198–199
Calcium hydroxyapatite, 287 Calcium phosphate nanospheres, 498 Calf, 82, 256 and e-PTFE, 82 Capillaries, in vitro reconstruction, 206–211 Carbon coating, 98–99 Cardiac uses of e-PTFE, 80 Cardiopulmonary bypass (CPB), 55–56, 61–69 Carotid and subclavian grafts, 79–80 Cartilage, auricular, tissue engineering, 253–264 backgound, 253–255 definition of cartilage, 254 embryology, 253–254 histology of cartilage, 254–255 chondrocytes from, 255–260 harvesting, 255–260 introduction, 253 research directions, 262–263 genetic engineering of chondrocytes, 262 transfection of chondrocytes, 262–263 Cationic lipids and liposomes, 336 Cationic polymers, 337 Cell–extracellular matrix interactions, 9–12 Cells, 55–74, 198, 253–254, 386, 464–465, 470–471 as biodelivery system, 386, 470–471 embryology, 253–254 ideal source, 198 measurement of transient adhesion, 55–74
Index Cell sequestration and isotope labeling, 56–58 neutrophil labeling, 57–58 platelet labeling, 56–57 Cell surface adhesion, 70–71 Cellular and systemic barriers, to gene delivery, 353–355 cellular and intracellular, 354–355 systemic, 353–354 Cellular binding, 355 Chitosan, 338–340 Chondrocytes, 254, 255–260 harvesting, 255–260 Cloned multipotent cell, and growth factor delivery, 463–473 Collagen, 231–233, 340–341, 476, 521–541, 599–632 (see also Bioactive molecules and biodelivery systems; Biodegradable hydrogels as drug controlled release vehicles; Biodegradable nanoparticles as drug delivery systems for parenteral administration; Biological delivery systems; Nanoparticles; Nanosized biosensors and delivery vehicles; PHAs; Polymeric gene delivery systems) biosynthesis, 232 infused with bisGMA-based polymeric adhesive, 599–632 background, 599–602 discussion, 623–628 experimental results, 618–623 materials and methods, 602–618 resorbable biodevices, 521–541 adhesives, 525–528 drug and biologics delivery, 533–535 hemostasis, 522–524 tissue engineering, 528–533 Collagen–glycosaminoglycan (GAG) membrane, 156–157 Colloidal gold templated silica nanoshells, 485–494 synthesis of colloidal gold, 485–486 Composite failure, 600 Contact angle measurement, 189 Controlled release profiles of model drugs, 450–453 Coronary angioplasty, 107 Corona discharge, 177 Corona treatment, 177 Cornea reconstruction, 213–215 characteristics of the reconstructed cornea, 213–215 methods, 213 Crotonic acid assay, 284
Index Cryopreserved epidermis and dermis, 141–154 Cultured skin substitutes, (see Skin substitutes) Cutaneous gene therapy, 155–172 function and composition of skin substitutes, 155–159 future directions, 165 gene therapy with cultured skin grafts, 161–164 and cutaneous disease, 162–163 enhanced performance, 161–162 and systemic deficiencies, 163–164 genetic modification of skin cells, 159–161 Cyclodextrins, 584 Cytoskeleton and signal transduction, 11–12 Cytosol, trafficking in, 355
Dialysis access, 78–79 Degree of inhibition, 513 Dental amalgam, 599–600 Dentin, 600–603 bonding to components, 601–602 composition of, 600–601 specimen preparation, 602–603 Dermal–epidermal junction, 145–146, 150 Dermis and epidermis, cryopreserved, 141–154 from ‘‘outdated’’ cryopreserved allografts, 142–144 Designed molecules, 241–244 based on ␣ structures and coiled coils, 243–244 based on  structures, 242–243 based on triple helix, 241–242 Design, of PHAs, 543–545 bulk properties, 543–545 surface properties, 545 Dextran–maleic acid , 425–428 Dextran–methacrylate, 428–436 Dextran–polyactide hybrids, 436–453 chemical structure, 437 controlled release profiles of model drugs, 450–453 mechanical properties, 449–450 morphological structure, 443–446 swelling properties, 437–443 thermal properties, 447–449 Dialysis access grafts, 78–79 DNA condensing systems, 343–347 naked, 335 Dogs, 82, 96, 138, 292 and e-PTFE, 82, 96 and tissue adhesives, 138
637 Doxorubicin, 428–436 Drug delivery systems, (see Bioactive molecules and biodelivery systems; Biodegradable hydrogels as drug controlled release vehicles; Biodegradable nanoparticles as drug delivery systems for parenteral administration; Biological delivery systems; Collagen; Nanoparticles; Nanosized biosensors and delivery vehicles; PHAs; Polymeric gene delivery systems) Drug entrapment in PACA nanospheres, 396–397 Drugs, and encapsulation, 305–332, 399–400
Elastic protein-based biomaterials, 31–54 (see also Proteins; Recombinant protein scaffolds) basis of biocompatibility, 48–52 proposed, 51–52 studies, 48–51 introduction, 31–33 formation of nanoparticles, 33 ion-paired controlled release device, 42–45 design of, 42 profiles for loading, 42–43 release profiles, 43–45 mechanics of, 33–41 comprehensive hydrophobic effect, 37–41 mechanism of entropic elasticity, 34–37 nanoparticles, 45–52 advantage of, 48 formation of, 45–48 Elastin, 233–236 Electron spectroscopy for chemical analysis, 178, 190 Electrostatic interactions, 313–314 Embryology, 253–254 Emulsification, 399, 400–402 Encapsulation systems, 349–350, 399–400 Endoskeletal scaffolding, 261 Endosomal processing, 355 Endothelialized reconstructed skin, 207–209 Endotheliazation of e-PTFE grafts, 86–87, 95–98 Endovascular stent categorization, 108–109 Enhanced permeability, 354 Entropic elasticity, mechanism of, 34–37 (see also Elastic protein-based biomaterials) Environmental pollutants, and monitoring, 515–516 Epidermal sheets, for transplantation, 200
638 Epidermis, 141–154, 198–201 cell isolation, culture, and epidermal sheet production, 198–201 cryopreserved, 141–154 Epidermolysis bullosa, 163 Escherichia coli system, 232–233, 235, 237, 241, 242, 290, 386 Esophagus and gastrointestinal tract, and stents, 123 Evanescent microwave probe microscopy, 608–618 Expanded polytetrafluoroethylene (e-PTFE) vascular grafts, 75–105 adjunctive methods of improved patency, 85–99 bonding of antithrombogenic agents, 98–99 surface properties, 85–98 clinical use of e-PTFE grafts, 77–81 aortofemoral, 79 axillofemoral, 79 cardiac uses, 80 carotid and subclavian, 79–80 dialysis access, 78–79 femoropopliteal and tibial bypass, 77–78 venocaval and portal vein, 80–81 healing characteristics, 81–85 experimental implants, 82 human implants, 81–82 mechanisms of failure, 82–85 physical characteristics, 76 Extracellular matrix remodeling, 1–30, 209 and advanced biocompatibility, 20–23 cell–extracellular matrix interactions, 9–12 cytoskeleton and signal transduction, 11–12 growth factors, 10–11 receptors, 9–10 components of, 3–9 classic proteins, 6–9 fibrillar proteins, 3–6 matricellular proteins, 8–9 metalloproteinases, 17–20 definitions, 1–3 peri-implantar site, 14–17
Failure, of e-PTFE grafts, 82–85 Femoral grafts, 79 Femoropopliteal and tibial bypass grafts, 77–78 Fibrinogen–fibrin, 236–238, 243 Fibronectin, 178, 240 Flow properties of biological fluids, 268, 271–272, 274, 275–276 Fluid substitutes, 276–278
Index Fluorescence amplification with nanospheres, 487–491 Fluorescent nanoprobes, 483 Formulation purification, 402–404 Fourier transform infrared spectroscopy, 285
Gamma scintigraphy, 59–61 quantitative analysis, 59–61 Gastrointestinal patch, 558 Gelatin and collagen, 231–233, 529 (see also Collagen) Gelatin-resorcinol-formaldehyde (GRF), 137–138 Gene activated matrix, 343 Gene therapy, 99, 118–119, 155–172, 262–263, 333–367, 385–386, 533 (see also Bioactive molecules and biodelivery systems; Polymeric gene delivery systems) of chondrocytes, 262, 263 cutaneous gene therapy, 155–172 gene therapy with cultured skin grafts, 161–164 and cutaneous disease, 162–163 enhanced performance, 161–162 and systemic deficiencies, 163–164 genetic modification of skin cells, 159–161 transfection of chondrocytes, 262–263 Glow discharge (plasma) deposition, 177–178 Growth factor delivery, 463–473 discussion, 470–471 biological delivery systems, 471 cell-mediated gene delivery in bone repair, 470–471 clinical implications, 471 introduction, 463–464 materials and methods, 464–466 cells, 464–465 histological tracing, 466 molecular detection, 466 in vivo delivery, 465–466 results, 467–470 histological analysis of bone healing, 467–469 histomorphometric analysis of bone healing, 470 molecular analysis, 469–470 quantification of transplanted cells, 469 Growth factors, 10–11, 84, 237, 262, 312, 316, 317, 463–473, 535 delivery, 463–473
Index Healing characteristics of e-PTFE vascular grafts, 81–85 experimental implants, 82 human implants, 81–82 mechanisms of failure, 82–85 Hemostasis, 522–524 Heparin, 98, 99, 118, 534, 570 Histological analysis of bone healing, 467–469 Histological tracing, 466 Histomorphometric analysis of bone healing, 470 Hyaluronic acid protein, 268–270, 300 Hydrogels, 335, 423–461, 517 (see also Biodegradable hydrogels as drug controlled release vehicles) emerging hydrogel systems, 453–456 Hydrogen bonding, 312–313 Hydrophobic effect, 37–41, 308–309, 312, 318 Hyperspectral analysis of polymeric adhesive, 599–632
Implant related osteomyelitis, 282, 286 Incontinence and reflux, 261 Indomethacin (IDM), 450–452 Insect expression systems, 373 Instability in PLGA microparticles, 307–319 preparation, 308–309 protein release, 309–314 stabilization approaches, 314–319 preparation, 314–317 protein release, 317–319 Insulin-like growth factor, 84, 317, 464, 471 (see also Growth factors) Interferons, 374 Interleukin–1, 84, 374 Internal mammary artery, 75 Ion-paired controlled release device, 42–45 design of, 42 profiles for loading, 42–43 release profiles, 43–45
Jet injections, for gene delivery, 335
Keratinocytes, 159, 161, 165, 197, 200 Knee, and ligament reconstruction, 215–218
Lamellar ichthyosis, 162–163 Ligament reconstruction, 215–218
639 Ligand-mediated gene delivery, 337–338 Lipids and liposomes, cationic, 336–337, 376–377, 382–384, 494–495 as a biodelivery systems, 382–384 Lipopolyplexes, 349 Liquid crystals, 382 Local drug delivery, and stents, 119–122
Macromolecules, and nanoparticles, 395–414 natural macromolecules, 395 synthetic molecules, 395–414 poly(alkylcyanoacrylates), 395–398 polyesters and polyanhydrides, 398–399 polymer precipitation methods, 399–414 Mammalian expression systems, 373 Matricellular proteins, 8–9 (see also Proteins) Matrix, gene activated, 343 Matrix remodeling, extracellular, 1–30 Measurement of cells to blood-contacting surfaces, 55–74 animal studies, 61–62 cell sequestration and isotope labeling, 56–58 neutrophil labeling, 57–58 platelet labeling, 56–57 cell surface adhesion, 70–71 conclusion, 69–70 discussion, 67–69 gamma scintigraphy, 59–61 quantitative analysis, 59–61 results, 62–67 Mechanical properties of dextran–polyactide hybrids, 436–453 of PHAs, 284 Mechanochemical patterning, 186–187 Medical applications of PHAs, 547–558 (see also Applications; PHAs) bioactive agent delivery, 546–552 hard tissue, 553–554 other applications, 558 tissue culture, 554–558 Mesenchymal stem cells, 202, 470, 530 Mesoporous nanospheres, synthesis of, 495–498 Metalloproteinases, 17–20 Metal stents, 109–111 (see also Stents) balloon-expandable, 109–111 Metal surface treatment, 115–116 Micellar systems, polymer-based, 350, 382 Microbial expression systems, 372–373 Microbial polyesters, 281–303 conclusions, 301 introduction, 281–282 materials and methods, 282–287
640 [Microbial polyesters] antibiotic-loaded controlled release rods, 285–286 calcium hydroxyapatite, 287 characterization of PHAs, 284–285 determination of PHAs, 284 production of PHAs, 283–284 in vivo studies, 286–287 results, 287–300 characterization of PHAs, 290–295 PHA quantification, 290 production of HAP powder, 300 production of PHAs, 287–290 in vitro antibiotic release, 295 in vivo results, 295–300 Microencapsulation of protein drugs, 305–332, 572 basic protein chemistry, 306–307 forces influencing structure, 306 relevant routes of inactivation, 306–307 further considerations, 319–320 instability in PLGA microparticles, 307–314 preparation, 308–309 protein release, 309–314 solvent exchange method, 320–325 background, 320–321 method, 321–325 stabilization approaches, 314–319 preparation, 314–317 protein release, 317–319 Micro-Raman spectroscopy, 603–604, 618–620 Molecular analysis, 469–470 Molecular detection, 466 Monoclonal antibodies, 375–376, 380–381 Morphology, and particle size, 400, 443–446 and dextran–polyactide hybrids, 443–446 Mucus, 274–276 flow, 275–276 viscoelastic properties, 275 Multifunctional moieties, for gene delivery systems, 350–352 Multiple-lobe stents, 113–114
Naked DNA, 335 Nanoparticles, 33, 45–52, 395–422, 563–598 (see also Applications; Bioactive molecules and biodelivery systems; Biodegradable hydrogels as drug controlled release vehicles; Biodegradable nanoparticles as drug delivery systems for parenteral administration; Biological delivery
Index [Nanoparticles] systems; Collagen; Nanosized biosensors and delivery vehicles; PHAs; Polymeric gene delivery systems) advantage of, 48 drug carrier design and applications, 563–598 applications, 573–586 preparation of nanoparticles, 564–573 formation of nanoparticles, 33, 45–48, 412–414, 571–572 Nanoshells, and colloidal gold templated silica, 486–494 Nanosized biosensors and delivery vehicles, 483–503 (see also Bioactive molecules and biodelivery systems; Biodegradable hydrogels as drug controlled release vehicles; Biodegradable nanoparticles as drug delivery systems for parenteral administration; Biological delivery systems; Collagen; Nanoparticles; PHAs; Polymeric gene delivery systems) conclusions, 499 introduction, 483–484 methods, 484–494 colloidal gold templated silicate nanoshells, 486–494 materials, 484–485 other types of templated nanospheres, 495–498 other types of templated slicate nanospheres, 494–495 synthesis of colloidal gold, 485–486 Natural protein materials, 239–241 (see also Proteins; Recombinant protein scaffolds) Neurovascular disease, and stents, 123 Neutrophil distribution, 58–59 Neutrophil labeling, 57–58 Nile red staining, 284 Nitric oxide, 56, 63–69 Nonbiodegradable polymers, 339, 343–349 (see also Biodegradable polymers) Nonvascular use of stents, 122–123 esophagus and gastrointestinal tract, 123 neurovascular disease, 123 tracheobronchial obstruction, 122 urology, 122 Nonviral gene delivery systems, 334–338 (see also Gene therapy; Polymeric gene delivery systems)
Index [Nonviral gene delivery systems] adenovirus-coupled polymers and peptides, 338 cationic lipids and liposomes, 336 cationic polymers, 337 hydrogels, 335 jet injections, 335 ligand-mediated, 337–338 naked DNA, 335 virosomes, 337 Nuclear localization signal, 352 Nuclear magnetic resonance, 285 Nuclear translocation, 355 Nucleic acids and oligonucleotides, 371, 377, 385–386 as a biodelivery system, 385–386 Ocular delivery of nanoparticles, 584–586 Oglionucleotides and nucleic acids, 377, 395, 579 Oral administration of nanoparticles, 581–584 Organic solvents, 321–324 Organ reconstruction by tissue engineering, 197–228 blood vessels, 211–213 characteristics of the reconstructed blood vessel, 212–213 methods, 211–212 bronchi, 218–219 cornea, 213–215 characteristics of the reconstructed cornea, 213–215 methods, 213 epidermis, 198–201 cell isolation, culture, and epidermal sheet production, 198–201 ligament, 215–218 self-assembly approach, 201–203 characteristics of the reconstructed skin, 202–203 methods, 202 principles, 201–202 in vitro model of wound healing, 203–206 in vitro reconstruction of capillaries, 206–211 Osteomyelitis, 282, 286 Oxygen plasma treatment, 178, 189, 294, 555, 557 Parenteral administration (see Biodegradable nanoparticles as drug delivery systems for parenteral administration) Particle size, and morphology, 400
641 Patency, long-term effects, of e-PTFE vascular grafts, 75–105 Penile prostheses, 261 Percutaneous transluminal coronary angioplasty (PTCA), 107 Peri-implantar site, 14–17 Permeability, enhanced, 354 Pharmacotoxicological application of ERS, 209–211 PHAs, 283–295, 543–562 (see also Bioactive molecules and biodelivery systems; Biodegradable hydrogels as drug controlled release vehicles; Biodegradable nanoparticles as drug delivery systems for parenteral administration; Biological delivery systems; Collagen; Nanoparticles; Nanosized biosensors and delivery vehicles; Polymeric gene delivery systems) approaches to the design of PHAs, 543–545 bulk properties, 543–545 surface properties, 545 biocompatibility, 546–547 characterization of, 284–285, 290–295 degradation, 545–546 determination of, 284 medical applications, 547–558 bioactive agent delivery, 546–552 hard tissue, 553–554 other applications, 558 tissue culture, 554–558 production of, 283–284, 287–290, 545 quantification, 290 in vitro antibiotic release, 295 in vivo studies, 286–287, 295–300 Physical characteristics of e-PTFE, 76 Pigs, 61–62, 82, 260 and auricular cartilage, 260 and e-PTFE, 82 Plasma deposition, 177–178 Platelet adhesion, 67 Platelet-derived growth factor (PDGF), 84, 162 Platelet distribution, 58 Platelet function, inhibition of, 99 Platelet labeling, 56–57 Platelet sequestration, 58 Poly(alkylcyanoacrylates), 395–398, 585, 586 Polyanhydrides, 398–399, 427 Polycaprolactone (PCL), 56, 112, 398, 585 Polyethylene, 178, 179, 318, 348, 510 Poly(D,L-lactic acid) (PDLA), 112, 437–452 Poly(D,L-lactide-co-glycolide) (PDLGA), 119–122
642 Poly(2–dimethylaminoethyl) methacrylate (pDMAEMA), 347, 348 Poly(dimethylsiloxane), 182 Polyesters and polyanhydrides, 398–399 Polyethylene glycol (PEG), 379, 395, 397, 402–404, 407–409, 436, 477–479, 483, 510, 526–528, 529, 532, 533, 569, 574, 580, 586 Poly(ethyleneimine) (PEI), 337, 346 Poly(glycolic acid) (PGA), 112, 398 Polyhydroxyalkanoates, (see PHAs) Polyhydroxybutyrate and its copolymers, 543–562 (see also PHAs) approaches to the design of PHAs, 543–545 bulk properties, 543–545 surface properties, 545 biocompatibility, 546–547 degradation, 545–546 medical applications, 547–558 bioactive agent delivery, 546–552 hard tissue, 553–554 other applications, 558 tissue culture, 554–558 production, 545 Polylactides (PLA), 178, 179, 185, 402–404, 410, 414, 476–479 Poly(L-lactic acid) (PLLA) stents, 112, 14–115, 116, 119, 120, 529 Polylactide-co-glycolides (PLGA), 179, 230, 305–332, 341–342, 398, 400–402, 406–407, 414, 529, 569 instability in PLGA microparticles, 307–314 preparation, 308–309 protein release, 309–314 solvent exchange method, 320–325 background, 320–321 method, 321–325 stabilization approaches, 314–319 preparation, 314–317 protein release, 317–319 Poly(L-lysine) (PLL), 337–338, 345–346 Polymer blends and block copolymers, 179–180 Polymeric biomaterials, surface properties, and tissue engineering, 173–196 Polymeric gene delivery systems, 333–367 (see also Bioactive molecules and biodelivery systems; Biodegradable hydrogels as drug controlled release vehicles; Biodegradable nanoparticles as drug delivery systems for parenteral administration; Biological delivery systems; Collagen; Nanoparticles; Nanosized biosensors and delivery vehicles; PHAs)
Index [Polymeric gene delivery systems] concluding remarks, 355–356 introduction, 333–334 nonviral gene delivery systems, 334–338 adenovirus-coupled polymers and peptides, 338 cationic lipids and liposomes, 336 cationic polymers, 337 hydrogels, 335 jet injections, 335 ligand-mediated, 337–338 naked DNA, 335 virosomes, 337 polymeric vectors, 338–349 biodegradable polymers, 338–343 nonbiodegradable polymers, 343–349 specialized polymer-based, 349–353 encapsulation systems, 349–350 lipopolyplexes, 349 multifunctional moieties, 350–352 polymer-based micellar systems, 350 polymer matrices, 352–353 systemic and cellular barriers, 353–355 cellular and intracellular, 354–355 systemic, 353–354 vaccine delivery systems, 353 Polymeric vectors, 338–349 biodegradable polymers, 338–343 nonbiodegradable polymers, 343–349 Polymer matrices, 352–353 Polymer precipitation methods, 399–414 Polymer type, and formation techniques, 394–395 Polymethylmethacrylate (PMMA), 281–282 Polysaccharides, 186, 376, 381–382 as a biodelivery systems, 381–382 Polyvinyl alcohol, 347–348, 400, 401, 402, 404 Portal vein grafts, 80–81 P. pastoris, 243 Properties of e-PTFE vascular grafts, 75–105 Prostacyclin, 99 Proteins, 3–9, 31–54, 181–186, 229–251, 305–332, 371–376, 378–380, 550, 582 basic protein chemistry, 306–307 forces influencing structure, 306 relevant routes of inactivation, 306–307 classic proteins, 6–9 as delivery system, 378–380, 582 fibrillar proteins, 3–6 instability in PLGA microparticles, 307–314
Index [Proteins] preparation, 308–309 protein release, 309–314 matricellular proteins, 8–9 microencapsulation of drugs, 305–332 patterning, 181–186 recombinant protein scaffolds, 229–251 stabilization approaches, 314–319 preparation, 314–317 protein release, 317–319
643 viscoelastic, 266–268 mucus, 274–276 flow, 275–276 viscoelastic properties, 275 synovial fluid, 268–272 flow, 271–272 viscoelastic properties, 268–271 vitreous body, 272–274 flow, 274 viscoelastic properties, 272–273 Rhodospirillum rubdum, 543
Quantitative analysis, 59–61, 469 of transplanted cells, 469
Rabbits, 261, 282, 578 and auricular cartilage, 261 and implant related osteomyelitis, 282 Radical emulsion polymerization, 565–566 Radiofrequency plasma, 177, 285 Ralstonia eutropha, 287, 288, 289, 291, 543 Rats and bowel defects, 558 and retinal pigment, 557 Recombinant human growth factor, 312, 316, 317 Recombinant protein scaffolds, 229–251, 479 collagen and gelatin, 231–233 designed molecules, 241–244 based on ␣ structures and coiled coils, 243–244 based on  structures, 242–243 based on triple helix, 241–242 elastin, 233–236 fibrinogen(fibrin, 236–238 introduction, 229–231 other natural protein materials, 239–241 silks, 238–239 Reflux and incontinence, 261 Regions of interest (ROI), 60, 63 Resorbable biodevices, 521–541 adhesives, 525–528 drug and biologics delivery, 533–535 hemostatis, 522–524 tissue engineering, 528–533 Retinal pigment epithelial cell culture, 182 Rheology of biological fluids, 265–280 fluid substitutes, 276–278 introduction, 265 properties of biological fluids, 266–268 flow, 268 [Rheology of biological fluids]
Scaffolds, recombinant protein, 229–251 Scanning acoustic microscopy (SAM), 604–608, 621–623 calibration analyses, 621–623 Scanning electron microscopy (SEM), 190 Scanning tunneling microscopy (STM), 189–190 Self-assembly approach, of reconstruction, 201–203 characteristics of the reconstructed skin, 202–203 methods, 202 principles, 201–202 Self-assembled monolayers, 182 Shear stress, of proteins, 309 Sheep, and tissue adhesives, 132–133, 138, 230, 558 Shunts, for cardiac use, 80 Silica and nanospheres, 486–495 other types of templated slicate nanospheres, 494–495 Silks, 238–239 Skin banks, 141 Skin substitutes, 141–154, 155–172 cryopreserved epidermis and dermis, 141–154 cutaneous gene therapy, 155–172 function and composition of skin substitutes, 155–159 future directions, 165 gene therapy with cultured skin grafts, 161–164 and cutaneous disease, 162–163 enhanced performance, 161–162 and systemic deficiencies, 163–164 genetic modification of skin cells, 159–161 Solubility, 234 Solvent exchange method in PLGA microparticles, 320–325 alternative solvents, 316
644 [Solvent exchange method in PLGA microparticles] background, 320–321 method, 321–325 Spiders, 238–239, 242 Stabilization approaches in PLGA microparticles, 307–319, 325, 397, 400–402 instability in PLGA microparticles, 307–314 preparation, 308–309 protein release, 309–314 stabilization approaches, 314–319 preparation, 314–317 protein release, 317–319 Stabilization of silicate nanospheres, 491 Star-shaped polymers, 453456 Static secondary ion mass spectrometry, 191 Stent technology, and biomaterials, 107–130 gene therapy, 118–119 metal surface treatment, 115–116 surface passivation, 116–117 introduction, 107–109 characteristics and materials, 108–109 metal stents, 109–111 balloon-expandable, 109–111 nonvascular use of stents, 122–123 esophagus and gastrointestinal tract, 123 neurovascular disease, 123 tracheobronchial obstruction, 122 urology, 122 polymeric stents, 111–115 bioresorbable, 111–113 biostable, 111 design of a high molecular weight PLLA stent, 114–115 multiple lobe, 113–114 as reservoirs for local drug and gene therapy, 115–122 active surface modification, 117–118 drug delivery, 119–122 Sterilization procedure, and nanoparticles, 414 Storage, of skin, 142 Streptococcus mutans, 600 Subclavian grafts, 79–80 Subcutaneous administration of nanoparticles, 580–581 Surface modifying additive (SMA), 56, 65, 68–69 Surface passivation, 116–117 Surface properties of polymeric biomaterials, and tissue engineering, 173–196, 545 introduction, 173–174 methods, 174–188
Index [Surface properties of polymeric biomaterials] biomimetics, different approaches, 180–188 flame and corona treatment, 177 glow discharge (plasma) deposition, 177–178 grafting, 179 polymer blends and block copolymers, 179–180 surface micropatterning, 180 of PHAs, 545 surface analysis, 188–191 atomic force microscopy, 189–190 contact angle measurement, 189 scanning electron microscopy, 190 static secondary ion mass spectrometry, 191 x-ray photoelectron spectroscopy, 190–191 Synovial fluid, 268–272 flow, 271–272 viscoelastic properties, 268–271 Systemic and cellular barriers, to gene delivery, 353–355 cellular and intracellular, 354–355 systemic, 353–354
Temperature, elevated, and proteins, 309 Thoracic aortic surgery, 131, 136–137 Tibial bypass grafts, 77–78 Tissue adhesive for use in surgery, 131–139 discussion, 136–138 introduction, 131–132 materials and methods, 132–134 experimental design, 132–133 operative technique, 132 post-operative care, 133–134 surgical adhesive, 133 results, 134–136 histopathology, 136 surgical outcomes, 134–136 Tracheal augmentation, 261–262 Tracheobronchial obstruction, and stents, 122 Transforming growth factor-, 84, 237, 262, 535 Transmural ingrowth of e-PTFE grafts, 87–95
Urology, and stents, 122
Vaccines, 353, 376, 381 (see also Bioactive molecules and biodelivery systems) as biodelivery system, 353
Index Vascular endothelial growth factor, 162 Vascular grafts, e-PTFE, 75–105 Vibrio fischeri, 515–516 Viscoelastic properties of biological fluids, 266–268, 268–271, 272–273, 275 Vein collars and patches, 85–86 Venocaval and portal vein grafts, 80–81 Ventricular assist devices, 260 Viral vectors, 334 Virosomes, 337 Viscometry, 285 Vitreous body fluid, 272–274 flow, 274 viscoelastic properties, 272–273
645 Water avoiding denaturation, 314–315 contact angle, 189 protein exposure to, 308 Wound healing, 155, 203–206 in vitro model of wound healing, 203–206
Xenografts, 386 X-ray photoelectron spectroscopy, 190–191
Yeast expression systems, 373