Advances in bioinspired and biomedical materials 9780841232198, 0841232199, 9780841232211, 0841232210, 9780841232204, 9780841232228

Table of contents :
Content: Volume 1 --
Volume 2.

Citation preview

Advances in Bioinspired and Biomedical Materials Volume 2

ACS SYMPOSIUM SERIES 1253

Advances in Bioinspired and Biomedical Materials Volume 2 Yoshihiro Ito, Editor RIKEN Institute Wako-shi, Saitama, Japan

Xuesi Chen, Editor Chinese Academy of Sciences Changchun, Jilin, China

Inn-Kyu Kang, Editor Kyungpook National University Daegu, South Korea

American Chemical Society, Washington, DC Distributed in print by Oxford University Press

Library of Congress Cataloging-in-Publication Data Names: Ito, Yoshihiro, 1959- editor. Title: Advances in bioinspired and biomedical materials / editors, Yoshihiro Ito, RIKEN Institute, Wako-shi, Saitama, Japan, Xuesi Chen, Chinese Academy of Sciences, Changchun, Jilin, China, Inn-Kyu Kang, Kyungpook National University, Daegu, South Korea. Description: Washington, DC : American Chemical Society, [2017] | Series: ACS symposium series ; 1252, 1253 | Includes bibliographical references and index. Contents: Volume 1 -- Volume 2. Identifiers: LCCN 2017044797| ISBN 9780841232204 (volume 1) | ISBN 9780841232228 (volume 2) Subjects: LCSH: Biomimetic materials. | Biomedical materials. Classification: LCC R857.M3 A379 2017 | DDC 610.28/4--dc23 LC record available at https://lccn.loc.gov/2017044797

The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ANSI Z39.48n1984. Copyright © 2017 American Chemical Society Distributed in print by Oxford University Press All Rights Reserved. Reprographic copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Act is allowed for internal use only, provided that a per-chapter fee of $40.25 plus $0.75 per page is paid to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. Republication or reproduction for sale of pages in this book is permitted only under license from ACS. Direct these and other permission requests to ACS Copyright Office, Publications Division, 1155 16th Street, N.W., Washington, DC 20036. The citation of trade names and/or names of manufacturers in this publication is not to be construed as an endorsement or as approval by ACS of the commercial products or services referenced herein; nor should the mere reference herein to any drawing, specification, chemical process, or other data be regarded as a license or as a conveyance of any right or permission to the holder, reader, or any other person or corporation, to manufacture, reproduce, use, or sell any patented invention or copyrighted work that may in any way be related thereto. Registered names, trademarks, etc., used in this publication, even without specific indication thereof, are not to be considered unprotected by law. PRINTED IN THE UNITED STATES OF AMERICA

Foreword The ACS Symposium Series was first published in 1974 to provide a mechanism for publishing symposia quickly in book form. The purpose of the series is to publish timely, comprehensive books developed from the ACS sponsored symposia based on current scientific research. Occasionally, books are developed from symposia sponsored by other organizations when the topic is of keen interest to the chemistry audience. Before agreeing to publish a book, the proposed table of contents is reviewed for appropriate and comprehensive coverage and for interest to the audience. Some papers may be excluded to better focus the book; others may be added to provide comprehensiveness. When appropriate, overview or introductory chapters are added. Drafts of chapters are peer-reviewed prior to final acceptance or rejection, and manuscripts are prepared in camera-ready format. As a rule, only original research papers and original review papers are included in the volumes. Verbatim reproductions of previous published papers are not accepted.

ACS Books Department

Contents Preface .............................................................................................................................. ix

Polymeric Materials 1.

From Biomaterial, Biomimetic, and Polymer to Biodegradable and Biocompatible Liquid Crystal Elastomer Cell Scaffolds ...................................... 3 M. Prévôt and E. Hegmann

2.

Development of Redox Nanomedicine for Gastrointestinal Complications via Oral Administration Route ............................................................................. 47 Long Binh Vong and Yukio Nagasaki

3.

Glycoglycan Mimic by Synthetic Polymers ......................................................... 69 Yoshiko Miura, Tomohiro Fukuda, and Yu Hoshino

4.

Plant Cell-Inspired Hydrogel Composites with High Mechanical Strength ..... 79 Naozumi Teramoto, Keisuke Wakayama, Mitsuru Harima, Toshiaki Shimasaki, and Mitsuhiro Shibata

5.

Synthesis and Temperature-Responsiveness of Poly(ethylene glycol)-like Biodegradable Poly(ether-ester)s .......................................................................... 93 Yuichi Ohya, Akihiro Takahashi, Hiroki Takaishi, and Akinori Kuzuya

Inorganic/Hybrid Materials 6.

Synthesis of Calcium Phosphate Microspheres Using an Ultrasonic Spray–Pyrolysis Technique and Their Application as Novel Anti-Angiogenic Chemoembolization Agents for Cancer Treatment .......................................... 107 Mamoru Aizawa, Michiyo Honda, and Makoto Emoto

7.

Bioinspired Design and Engineering of Functional Nanostructured Materials for Biomedical Applications .............................................................. 123 Xin Ting Zheng, Hesheng Victor Xu, and Yen Nee Tan

8.

Construction of Bio-Inspired Composites for Bone Tissue Repair ................. 153 Junchao Wei, Lina Wang, Lan Liao, Jiaolong Wang, Yu Han, and Jianxun Ding

9.

CNT-Based and MSN-Based Organic/Inorganic Hybrid Nanocomposites for Biomedical Applications ................................................................................ 169 Jiemei Zhou, Jiaoyang Li, Decheng Wu, and Chunyan Hong

vii

Micro/Nano-Fabricated Devices 10. In Vitro Design of Nanoparticles Using an Artificial 3D-Blood Vessel Wall Model for Atherosclerosis Treatment ................................................................ 195 Michiya Matsusaki and Mitsuru Akashi 11. A Novel Cell Fusion Method for Direct Cytoplasmic Transfer Using a Microfluidic Device .............................................................................................. 227 Ken-Ichi Wada, Kazuo Hosokawa, Yoshihiro Ito, and Mizuo Maeda 12. Biodegradable Polymeric Esophagus Stents ..................................................... 237 Divia Hobson, Arvind Dhinakar, Nianyuan Shi, Le Zhang, Wenjing Wu, Lifeng Hou, and Wenguo Cui 13. Three-Dimensional Printing Technology Combined with Materials Drives Meniscal and Cartilaginous Regeneration ......................................................... 253 Zhu-Xing Zhou, Zheng-Zheng Zhang, Shao-Jie Wang, Dong Jiang, and Jia-Kuo Yu 14. Advances in Enhancing Mechanical Performance of Ultrahigh Molecular Weight Polyethylene Used for Total Joint Replacement .................................. 273 Yan-Fei Huang, Jia-Zhuang Xu, and Zhong-Ming Li

Indexes Author Index ................................................................................................................ 297 Subject Index ................................................................................................................ 299

viii

Preface Bioinspired concepts are becoming increasingly integrated into materials and devices intended for medical applications. Biological organisms evolve within specific environmental constraints, giving rise to elegant and efficient strategies for fabricating materials that often outperform man-made materials of similar composition. A main goal of the interdisciplinary field of bioinspired materials is to unlock the secrets of this process — the composition, processing, self-assembly, hierarchical organization, and properties of biological materials — and use this information to synthesize and engineer novel functional materials for a variety of practical applications. In consideration of the rapid advances in this area, we organized an international symposium on “Advanced in Bioinspired and Medical Materials” at the PacifiChem in Honolulu, Hawaii in December 2015. The symposium was successful, with a total of 41 papers and active participation and discussions among the leading researchers. In view of the success of the PacifiChem symposium, and the fact that this field is multidisciplinary where publications tend to be spread out over journals in different disciplines, we decided to edit this book in order to gather the information on the latest developments in one place. The authors are from a variety of scientific disciplines, including biology, biochemistry, chemistry, physics, materials science, mechanical engineering, and bioengineering. The book readers should be interested in the cross-disciplinary fertilization of new ideas in this emerging field. The content of the book is composed of six sections: (i) DNA/Protein/Peptide Assemblies, (ii) Polypeptides and Engineered Proteins, (iii) Catechols/ Polydopamine Derivatives, (iv) Polymeric Materials, (v) Inorganic/Hybrid Materials, and (vi) Micro/Nano-Fabricated Devices and divided into two volumes. Volume 2 includes sections (iv), (v), and (vi) and focuses on the bioinspired approaches using synthetic organic and inorganic materials, and on the fabrication using the materials for medical devices. Both volumes cover the interdisciplinary fields of biological, synthetic, and the hybrid materials, and describe their medical applications ranging from molecular to cellular levels. The book must attract the readers who are interested in the field of chemistry and biomedical/biomaterials science and engineering. We appreciate the efforts of the authors to submit their manuscript and their cooperation during the peer review process. We are also grateful to our many anonymous reviewers for their hard work. Thanks are also due to the staff of ACS Books.

ix

Yoshihiro Ito Nano Medical Engineering Laboratory RIKEN Institute Japan

Xuesi Chen Changchun Institute of Applied Chemistry Chinese Academy of Sciences China

Inn-Kyu Kang Department of Polymer Science & Engineering Kyungpook National University South Korea

x

Polymeric Materials

Chapter 1

From Biomaterial, Biomimetic, and Polymer to Biodegradable and Biocompatible Liquid Crystal Elastomer Cell Scaffolds M. Prévôt1 and E. Hegmann1,2,3,* 1Liquid

Crystal Institute, Kent State University, Kent, Ohio 44242-0001, United States 2Chemical Physics Interdisciplinary Program, Kent State University, Kent, Ohio 44242-0001, United States 3Department of Biological Sciences, Kent State University, Kent, Ohio 44242-0001, United States *E-mail: [email protected]

3D scaffolds are no longer simply physical templates for cell growth and tissue formation; they also have to provide chemical, biomolecular, mechanical and geometrical signals to cells. Liquid crystals (LCs) can be of significant importance in tissue engineering because they are able to report anisotropic growth of expanding cell lines back to the observer with an easily discernable optical response such as a change in birefringence or via the alignment of the LC molecules. The real challenge is to design materials that can direct or guide the behavior of biological materials. After a brief survey of several materials classes intensively investigated for the use as 3D cell scaffolds, we summarize recent research on a particular class of LC materials, liquid crystal elastomers (LCEs), that can serve as unique, longitudinal and multi-responsive cell scaffolds suitable for cell attachment, cell proliferation and cell alignment. Several types of biocompatible, biodegradable LCE scaffolds with specially engineered porous architectures will be introduced and their advantages discussed.

© 2017 American Chemical Society

Introduction Tissue, in a multicellular organism, is defined as an aggregate of similar cells forming a definite kind of structural material with a specific function, whereas an organ is the product of several similar tissues grouping together. There are in principle four types of tissue: connective, muscle, nervous, and epithelial tissue. Connective tissue binds tissues and organs in the body, providing cohesion and internal support. Connective tissue is mainly composed of fibrous tissues formed from non-living material secreted by living cells called the extracellular matrix (1) (ECM). The ECM is composed of two classes of macromolecules: fibrillar proteins (e.g., collagen, elastin) (2), bestowing mechanical and bioactive properties to the matrix (3), and glycosaminoglycans, that correspond to long unbranched polysaccharides (e.g., hyaluronic acid) (4), as shown schematically in Figure 1.

Figure 1. Description of extracellular matrix. Reproduced from reference (1). Copyright 2016, ACS Biomater. Sci. & Eng. Lett. Tissue engineering consists in the formation of functional substitutes for the therapeutic reconstruction of damaged or diseased tissues by the careful and controlled stimulation of selected target cells through a systematic combination of molecular and mechanical signals (5). Cells express functional responses by receptors on their surfaces that are mediated by the ECM. These responses convey developmental decisions, cell migration, cell maturation as well as differentiation, cell survival, tissue homeostasis and, in some cases, tumor cell invasion (6). Tissues are constantly in stress and must be resilient enough to deform reversibly without damage and still maintain their functions. Tissues must acquire specific mechanical properties to create an appropriate cell environment for the development of the ECM. The process of engineering tissue involves the design and use of a support that maintains tissue contour, particularly in the form of a 3D scaffold, implanted at the defective site. An ideal scaffold should then satisfy certain specific requirements (7, 8); it should: i.

be biocompatible (non-toxic) and should provide a framework for cells to attach and proliferate. Biocompatibility refers to the ability to co4

exist with human tissues without causing an inflammatory or a rejection response (9); ii. exhibit, during and after in vitro tissue culture tests, mechanical properties that can mimic native environments especially elasticity found in dynamic and structural tissues; iii. have a porous 3D architecture to allow cell growth, vascularization and transport of nutrients between the cells seeded within the matrix and the surroundings (10). As the properties of the ECM are characteristic of each tissue, so then engineering scaffolds must be designed to emulate the molecular and structural properties of the native ECM (11–13). Moreover, a well interconnected, highly porous structure is necessary to allow high cell seeding density and unimpeded tissue growth (14); and finally iv. degrade at a rate that matches the regeneration of new tissue and into non-toxic products that can easily be resorbed or excreted first by the cells, and finally by the body. Meaning that cells inside the scaffold can merge with the tissue at the implantation site and induce new tissue to infiltrate, while the scaffold material itself is gradually degraded in vivo (15). Biocompatibility Biocompatibility depends of the capability of a particular material to co-exist within body tissues without causing any considerable harm. Williams et al. defined biocompatibility as the “ability of a material to perform with an appropriate host response in a specific situation” (9). The biocompatibility of biomaterials is classified according to their ability to induce damage as described in Table 1. Thus, materials are expected to be non-toxic, non-immunogenic, non-thrombogenic, non-carcinogenic, and a non-irritant.

Table 1. Different types of biocompatibility problems (16). Name

Cause induced

Cytotoxicity

Cell or tissue death

Carcinogenicity

Cancer formation

Mutagenicity

Damage genes

Pyrogenicity, allergenicity

Immune response

Thrombogenicity

Blood clotting

According to Williams et al., biodegradability is not limited to immunological, toxic, or foreign body responses. Materials should be expected to passively allow or actively produce palpable beneficial effects in that particular host (9). Tissue constructs continuously interact with the body through the healing and 5

cellular regeneration process as well as during scaffold degradation. Because cell–matrix interactions influence a variety of important processes, it is essential that biomaterial scaffolds facilitate cell adhesion to the biomaterial surfaces to prevent apoptosis in cells. Cell-matrix anchorage is mediated through membrane proteins. The tensile strength of the ECM is a result of collagen; and fibronectin is responsible for adhesion of cells to the matrix (16). However, including both chemical and topographical characteristics, surface properties also control and affect cellular adhesion (17) as for example, hydrophilicity, geometrical features of the porous structure (pore size, pore size distribution), micro-architecture, heterogeneity, (an)isotropy, and interconnectivity of the porosity within the scaffold (9). Performance of biomaterials can be enhanced by surface modification such as, inclusion of short peptide motifs like the well-known Arg-Gly-Asp (RDG, tripeptide composed of L-arginine, glycine, and L-aspartic acid, a recognition system for cell adhesion) (18).

Mechanical Properties Before analyzing some functional mechanical attributes of materials, we need to clarify the definitions of stiffness and strength. Stiffness measures the ability of an elastic solid material to resist deformation to an applied stress. It is represented by the elastic modulus, also named Young’s modulus (YM). Strength corresponds to the maximal amount of tensile stress before failure. Among mechanical properties used to fully characterize tissue, toughness, extensibility, spring efficiency, durability, and spring capacity should all be considered (19). Table 2 summarizes functional attributes of materials and give the corresponding property and its unit. It should also be considered that growing cells cannot be exposed to more than 30% tissue extensibility without risk of damaging the cellular membrane. Stiffness and strength become then the main parameters to study scaffold materials. The mechanical properties of the ECM vary from tissue to tissue. Table 3 gives the YM of common soft body tissues. McKee et al. (20) have shown that the YM values depend on the method used to measure them. On the one hand, the YM can be obtained by applying a force to a section of tissue and measuring the change of length or strain (Figure 2 right). An alternative method involves controlled poking by indenters, including atomic force microscopy tips (Figure 2 left). The YM obtained by indentation measurements are usually quite different from those acquired by tensile measurements, (i.e., in the bulk of material). For example, liver and kidney tissues possess a much lower tensile YM compared to muscle tissues (10 MPa vs. 480 MPa) but show higher indentation YM (190 kPa vs. 7kPa) (20). Thus, to properly interpret cellular responses to biophysical stiffness there is a need to characterize both the indentation and tensile values of the YM, especially for non-homogeneous tissues (20). Table 3 summarizes indentation and tensile measurements of the moduli of common soft body tissues. 6

Table 2. Some functional attributes of materials and the material properties and units used to quantify these attributes. (taken from Gosline et al. (19)) Functional attribute

Material property

Units

Stiffness

Modulus of elasticity, Eint

Nm−2 (Pa)

Strength

Stress at fracture, σmax

Nm−2 (Pa)

Toughness

Energy to break work of fracture

Jm−3, Jm−2

Extensibility

Strain at failure, εmax

No units

Spring efficiency

Resilience

%

Durability

Fatigue lifetime

s to failure or cycles to failure

Spring capacity

Energy storage capacity, Wout

Jkg-1

Table 3. Indentation vs. tensile measurements of Young’s moduli. Data taken from McKee et al. (20) Tissue

Indentation (kPa)

Tensile (MPa)

Skin

~85

~30

Liver & kidney

~190

~10

Spinal cord & gray matter

~3

~2

Muscle

~7

~480

Tendon

No values

~560

Breast tissue

~8

No values

Artery & vein

~125

~2

Sclera

No values

~2.7

Cornea

~29

~3.0

Figure 2. Illustration of the two methods used to measure the anisotropy of the YM: indentation (left) and tensile (right).

Three-Dimensional (3D) Scaffolds Scaffolds play a pivotal role in tissue engineering and a porous architecture within these scaffold must allow for sufficient transport of oxygen, nutrients, metabolites, and cellular signals (among other regulatory factors) (21). Different 7

types of scaffolds have been studied by many groups including hydrogel-based scaffolds, microsphere-based scaffolds, fibrous scaffolds, and polymeric porous scaffolds. For these main scaffolds, fabrication techniques include a variety of methods such as salt leaching, gas foaming, phase separation, electrospinning, using of free-form, and lithography approach. A comparison of these fabrication techniques has been presented in a review-article by Seunarine et al. (22). Hydrogels are crosslinked hydrophilic polymers capable of swelling and retaining a large amount of water within their 3D network without dissolution. They possess a degree of flexibility very similar to natural tissue due to their water content, allowing minimally invasive procedures, and exhibit excellent biocompatibility (23). Synthetic polymers used in tissue engineering include especially poly(ethylene glycol) (PEG) (24–27), poly(vinyl alcohol) (PVA), and polyacrylates. Hydrogels acts as bulk 3D gel framework to which cells are adhered or suspended (28). Hydrogels can also be designed to incorporate bioactive agents (growth factors, proteins), however, they commonly do not guide cells to alignment and differentiation by applied external stimulus (29, 30). Microspheres can be fabricated using a variety of different biodegradable polymers and with surface porosity to improve diffusion of nutrients and oxygen (31). Microspheres as scaffold building blocks offer several benefits: (i) their fabrication, morphology, and physicochemical characteristics are simple to control and (ii) they allow simple control of the release kinetics of encapsulated factors (32). A 3D nano-fibrous scaffold can be generated by electrospinning, molecular self-assembly, or thermally induced phase separation (TIPS) of a polymer mixture. Phase separation has shown high potential to meet the needs of 3D tissue regeneration due to its ability to incorporate any pore shape and size or any overall 3D geometry (33). The high surface-area-to-volume ratio of the nanofibers combined with their microporous structure favors cell adhesion, proliferation, migration, and differentiation. Nano-fiber technology, however, must place the cells within the nano-fibrillar structure and form the porous network in situ without cellular damage, resulting in difficult processability (34). 3D polymeric porous scaffolds have been widely used for tissue engineering. The typical scaffold design includes sponge or foam morphologies. These designs promote a homogeneous, interconnected pore networks. These porous scaffolds can be manufactured with specific pore sizes, porosity, and surface-area-to-volume ratios (17). Moreover, they offer versatility of the employed chemistry. Later in this chapter, we will focus only on porous scaffold. Scaffolds, in general, should be morphological and architecturally similar to the host environment. They should also promote vascularization; permitting new blood vessels to grow implies appropriate porosity and pore size. Additionally, a high surface-area-to-volume ratio in interconnected porous scaffolds favors cell attachment and proliferation. Thus, scaffolds must possess a highly porous structure with an open and fully interconnected geometry (17) with high structure porosity (28). If the pores are too small, pore occlusion by the cells can occur, whereas larger pores disturb the stability of the scaffold (35). According to Dhandayuthapani et al. (17), and Chang et al. (36) optimal pore size depends on the nature of the cell (see Table 4). Pore interconnectivity must be sufficient to 8

ensure that all cells are within 200 µm from the blood supply for mass transfer of oxygen and nutrients (17).

Table 4. Optimal pore size for cell infiltration according to Dhandayuthapani et al. (17) and Chang et al. (36) Type of cell

Pore size (μm)

Red blood cell

5

Hepatocytes

20

Osteogenic cell

100-150

Adult mammalian skin cell

20-125

Smooth muscle cells

60-150

Endothelial cells

24

60-65

PDLLA

1200

44

12-16

55-60

+ biodegradable

- local reactions caused by acid degradation products - low mechanical properties

Elastomers PLA (74–76, 81, 82, 84, 111)

Tensile YM (MPa)

Tensile strength (MPa)

Degradation (month)

Tg (°C)

Advantages

Drawbacks

PGA (74, 76, 81, 82, 84, 111, 131)

6900

70

Several weeks

35-40

+ biodegradable

- low mechanical properties

PLGA (16, 84, 111)

1400-2800

41.4-55.2

1-12 (adjustable)

45-55

+ biodegradable + good degradation rate

- low mechanical properties

PCL (4, 16, 84–86, 111, 131)

0.21-0.34

20-34.5

>24

-72

+ good mechanical properties

- low degradation rate - low cell affinity

-60--6

+ good degradation + good processability

- low biostability - low cell affinity - carcinogenic - creep deformation

0.5

Several weeks

21

PU (16, 93–98)

Several months

Figure 7. Schematic of (a) end-on main chain, (b) side-on side chain, and (c) end-on side chain LCEs. Abbott et al. reported the effect of elastic stresses and defects of LC materials on the organization of certain lipids. They emphasized the connection between the mechanical properties of LC materials, LC ordering, and how this affected cell behavior (alignment) (134). Synthetically, the mechanical properties of LC materials can be tuned to match those of a range of cell types, and suitable approaches presented thus far include colloids in LC gels or LC gels derived from hydrogen-bonded molecular networks (135). Liquid crystal elastomers (LCEs) are soft materials that combine both the order of liquid crystals (orientational order and as a result anisotropic optical properties) (136) and the elasticity of elastomers (137, 138). Cross-linking allows LCEs to act as solids providing more support than gels or hydrogels. Nevertheless, the mesogenic groups within LCEs maintain considerable mobility because the network backbone formed by distant adjacent cross-links retains its rubbery state. Of the many potential applications for LCEs, sensors and actuators have been the most promising due to the LCEs ability to undergo significant shape change in response to a range of external stimuli (light, pH, temperature, among others) (139–144). The combination of biocompatibility, biodegradability and mechanical properties makes LCEs capable of satisfying most of the requirements of adaptable scaffolds, as we will discuss later. Also, the physical properties of LCEs can be tuned in a way to withstand mechanical tasks such as strain, stress, and impacts because they are soft, deformable (145), and can be functionalized with cell growth promoting moieties. They have also been found suitable as carriers in drug delivery applications. LCEs with their distinctive properties (146) 22

have been introduced as artificial muscles (liquid crystalline materials provide the greatest shape change with the least amount of energy) (147, 148), sensors (149, 150) and actuators (141, 151, 152), as tunable lasers (153), and as light-driven motors (154). LCEs are classified according to their structure. In a LCE, the mesogenic moieties may be attached into the main-chain (backbone), being part of the main chain, or be linked as a side (pendant) group on the main chain (see Figure 7). While LCEs can also show nematic or smectic order, the anisotropic arrangement (without an external stimuli) appears in form of a polydomain organization. Among the most promising properties of LCEs is the possibility to macroscopically orient the sample by mechanical stress, i.e., uniformly aligned the director and giving place to a monodomain organization within the structure. A stable monodomain arrangement can be reached by fixing the order by an extra cross-linking step as mentioned by Fleischmann et al. (146). Some of the factors that affect LCE behavior under stress (i.e., their Young’s modulus) are the ordering of the LC phase (e.g., nematic or smectic), the type of connectivity between LC molecules with respect to the polymer backbone (Figure 7), and the mono- or polydomain nature fixed during cross-linking. The LCEs’ anisotropic response to external stimuli (e.g., at the transition from the LC to the isotropic liquid phase) set them apart from conventional elastomer materials. Their soft and malleable nature allows them to conform to different shapes and surfaces (155, 156). The first nematic fixed monodomain elastomer was synthesized by Bergmann et al. (157). Recently, Yakacki et al. (158, 159) and Kim et al. (160) develop programmable monodomain LCEs. A potential advantage of LCEs is the often straightforward synthetic access to many structural variations and elastic properties from commercially available starting materials, which permits scale-up and high reproducibility. Several reports indicate that the contractile and expansion properties of these materials are associated with Young’s moduli (measured by tensile tests) between tens of kPa to several MPa; as such, LCEs can be regarded as artificial muscles (161–164) or used as biological actuators (146, 165). This range of elasticity perfectly matches those of several soft tissues in the human body. Considering both transverse Young’s moduli (obtained by indentation measurements) as well as Young’s moduli obtained by tensile measurements, the values range from 0.2 kPa for gray matter (darker tissue of brain and spinal cord) measured by indentation to ~480 and ~560 MPa for muscle and tendons, respectively. For LCEs to be considered as new materials for cell scaffolds, in addition to their optical and mechanical properties, they should also have well-defined porosity and surface properties to provide support for cell adherence, growth, and mass transport in and out of the scaffold under physiological conditions. Elastomer porosity should stimulate 3D cell–elastomer interactions, space for extracellular matrix (ECM) formation (6), and provide opportunities for linking molecular entities that allow binding of proteins, growth factors or proteins to enhance cellular adhesion. Considering all necessary requirements for LCEs as responsive cell scaffolds, we aimed at taking full advantage of the exceptional properties of the scaffolds’ liquid crystalline properties to guide cell attachment, proliferation, 23

and alignment. Additional advantages result from built-in morphological cues (porous architecture) and their anisometric response to external stimuli. To do so, we prepared series of smectic and nematic LCEs that have proven to be non-cytotoxic to soft tissue cell lines. We have tested several selected cell lines (myoblasts, fibroblasts, and neuroblastomas) to create novel dynamic and spatial in vitro systems to simulate and study the development of new tissue as well as the complex cell-elastomer interplay.

Smectic-A (Sm-A) Liquid Crystalline Elastomers (LCEs) As a proof-of-concept, we reported the first synthesis, characterization and use of biocompatible, biodegradable, and porous LCEs (128) using a modular, convergent solvent-free synthesis (no impurities or toxic solvents), that allowed us to adjust porosity, degradation rate, and hydrophilic/hydrophobic balance. To prepare these new LCEs we combined three monomers, two lactone-based monomers and (D,L)-lactide, into a cross-linkable star block copolymer (SBC). Glycerol was chosen as central node due to its known nontoxicity and multifunctional reactive nature of the hydroxyl groups. ε-Caprolactone (ε-CL) and (D,L)-lactide are polymerized from this central node giving random chains fused by easily hydrolysable ester bonds known to produce biocompatible six-carbon fragments (highly important for biodegradation (166)). At a later stage, other polyols (building blocks of triglycerides) were chosen as central nodes offering access to series of three-, four- and six-arm SBCs (126, 167). Following a slightly modified method, involving ring opening polymerization (ROP), developed by Amsden et al. (126), the method modifications allowed for further chemical manipulation of the SBC via halogen atom substitution on the ε-caprolactone segments in two positions (α or γ) permitting a later attachment of LC pendant groups via alkyne-azide Huisgen’s cycloaddition reaction (“click” reaction) (128). Cholesterol groups were selected as LC pendants since cholesterol is known to be non-toxic and biocompatible. In fact, cholesterol is found in the organization of DNA (e.g., mitochondrial DNA-protein complexes attach to cholesterol-rich membrane segments) (168), the shell of certain shellfish and insects, and in cell walls of plants. Cholesterol is one of the first compounds found to be liquid crystalline; its liquid crystalline behavior was first described by Friedrich Reinitzer in 1888 (169, 170). Changing the position of the cholesterol groups in the substituted ε-caprolactone from the α- to the γ-position increases their mobility within the elastomer chain by reducing steric constrains (Figure 8). A bis-caprolactone was selected as a cross-linker to maintain identical, hydrolysable chemical bonds throughout the entire LCE as those forming the polymer backbone. LCEs prepared using this method can be casted or molded, and present an inherent porosity (spherical voids) due to the cross-linking process. The size and density of spherical voids is affected by position of the cholesterol pendants (α or γ), by the degree of cross-linking, and by the degree of cholesterol functionalization (ranging from 10 to 40%) with respect to the ε-caprolactone block. 24

Figure 8. (a) Chemical structure of 3-arm, 4-arm, and 6-arm initiators (central nodes), (b) synthesis pathway to star block copolymer-cholesterol liquid crystal (SBCα-CLC) and SBCγ-CLC, and (c) Crosslinking with bis-caprolactone (BCP) to obtain a 3-LCE-α or 3-LCE-γ. For better figure description see publication (167). Reproduced from reference (167). Copyright 2017, Macromol. Biosci., Wiley-VCH Verlag GmbH & Co. KGaA.

We have further adjusted the hydrophilic-hydrophobic balance (171) required for different cell types. This is easily achieved by introducing additional hydrophilic poly(ethylene oxide) (PEO) segments (Figure 9) instead of the 3-, 4- and 6-arm polyol central nodes. In contrast to the polyol nodes, introduction of PEO results in linear block-co-polymers (LBCs). Adjusting the hydrophobic-hydrophilic balance of the LCEs is important because certain cells 25

types prefer more hydrophilic surfaces than others. Myoblasts, for example, prefer more hydrophobic and neuroblastomas more hydrophilic LCE scaffolds (172).

Figure 9. (A) Cross-Linking of the polycaprolactone-based LC block-co-polymers using bis-caprolactone (CL = cross-linker) resulting in the formation of liquid crystal elastomers (LCEs), and (B) procedure for fabricating LCE foams with primary porosity (LCEFPP, Path 1) and LCE foams with both primary and secondary porosity (LCEFP+SP, Path 2). Reproduced from reference (172). Copyright 2016, ACS Macro. Lett.

26

Following the requisite of adjustable porosity, another type of porosity was accomplished using commercially available metal (Ni) foams as templates. This was performed during thermal cross-linking to create more porous LCEs with interconnected channels (tubular morphology) (172). These templates can be used in two ways, either fully immersed during cross-linking (tubular morphology), or quickly dipped into the pre-elastomer mixture prior to thermal cross-linking (see Figure 9). The latter produces a more complex architecture leading to LCE foams with primary and secondary porosity once the Ni template is etched away. These foams can be shaped in countless ways to obtain rolls, films, and folded constructs, among others.

Nematic Liquid Crystalline Elastomers (LCEs) De Gennes proposed the idea to think of nematic gels and LCEs as artificial muscles (147, 173). However there were no reports on the study of nematic LCEs as muscle cell scaffolds. To address this, we reported the first nematic LCEs as scaffolds for muscle cells and other tissues (174, 175). Nematic LCEs were prepared featuring a porous morphology using micro-emulsion photopolymerization with commercially available reactive mesogens (phenylbenzoate, PhBz) (176, 177). The reactive mesogens were confined within surfactant micelles, and the photopolymerization reaction captured the building blocks in a globular structure. Once photopolymerization was complete, the water–toluene solvent mixture was removed, and then the resulting LCEs scrupulously washed and rinsed to entirely eliminate the surfactant, which results in voids between the LCE globules. The resulting nematic LCE-PhBz exhibited a globular morphology with a porosity that proved suitable for seeding, growth, and proliferation of C2C12 myoblasts and other cell lines (see Figure 10). The overall LCE-PhBz porous morphology is suitable as 3D cell scaffold for several cell lines and allowed for satisfactory management of mass transport (i.e., nutrients, waste, and gases) even during longitudinal cell studies.

Mechanical and Thermal Properties The thermal properties for all LCEs were tested by differential scanning calorimetry (DSC) and thermal gravimetric analysis (TGA). All LCEs appeared amorphous lacking endothermic peaks indicative of melting. Predictably all Tg values were found to be considerably below physiological temperatures (128, 167), and LC phase formation and associated anisotropic mechanical properties were found from well below room to slightly above physiological temperatures (25 – 45 °C). 27

Figure 10. Chemical structure, phase transition temperatures of components, and ratio of monomer (M), crosslinker (L), and photoinitiator (PI; 2,2-dimethoxy-2-phenylacetophenone) used for the synthesis of the globular nematic LCE cell scaffolds. The cartoon depicts the synthesis (microemulsion photopolymerization) and the internal structure of the resulting fused globular LCEs. Reproduced from reference (174). Copyright 2015, ACS Appl. Mat. & Interfaces.

Substituting the 3-, 4- and 6-arm polyol central nodes considerably changes both the elasticity and the cell viability. The polyol central nodes can be seen as cross-linkers, where the 3-arm, and by analogy, the 6-arm LCEs have shown nearly identical elasticity. Contrary, the 4-arm SBCs lead to LCEs with higher Young’s moduli (~4 MPa). This can be related to previous theoretical and experimental studies where tetra-arm polymer hydrogel systems have extremely high homogeneous packing and suppressed heterogeneity, explaining the higher stiffness of the 4-arm SBC-based LCEs (LCE-4γ). Most of the studied LCEs have presently lower moduli (~2.0 – 4.0 MPa) than those of the selected tissues for study (~30 MPa for skin and ~480 MPa for muscle; see Table 3) (20), and some widely used biodegradable polymers. This is most likely due to the low molecular weight and nonlinear star-block structure (20, 167). However, the preliminary results of mechanical tests show an encouraging future for the use of LCEs as scaffolds. Cells were not only able to expand and proliferate, they aligned on and within the LCE scaffolds before any use of external stimuli. Figure 11 shows a comparative scale of the Young’s moduli of several soft tissues and of our LCEs measured by indentation and tensile tests. Figure 11 displays 28

that structural variety coupled with the choice of morphology allows our LCEs to cover a broad range of target tissue elasticity in both sets of measurements.

Figure 11. YM values obtained: (a) by indentation, and (b) by tensile measurements. Comparison of these values with measured YM values for specific tissues.

Recent synchrotron SAXS studies confirmed an increase in the degree of ordering of our LCE scaffolds (i.e., reorientation and alignment of the mesogenic units) upon applying mechanical stress. Further experiments to support this are now performed using immersion tensile stage measurements to study anisotropic changes of the LCE scaffolds immersed in media. The next step will then be to analyze the cellular response to the external mechanical stress applied when embedded in the LCE scaffold. In the case of LCE foams, there are several means to apply external stimuli, such as mechanical stress and mechanical compression as shown for the films. Other opportunities necessitate slight changes during preparation of the LCE foams such as adding magnetic nanoparticles (for magnetic stimuli), the addition of acrylate ends for an additional cross-linking step after mechanical induced stress, or pH-dependent moieties that could trigger a shape change.

3D Morphology The LCE scaffolds were prepared with three distinct morphologies: molded scaffolds or films, foams, and globular. Molded or film scaffolds were characterized by a morphology similar to that of a “Swiss Chesse”, foams (with two types of porosity: primary porous and primary as well as secondary porous), and globular (Figure 12). 29

Figure 12. SEM images of the internal porosity of: smectic LCE-γ (A) before, (B) and after 16 weeks of biodegradation in PBS, (C-D) of globular, porous microstructure of nematic LCE (E) LCEPP, (F) LCEPP, (G) Optical image of LCEFP+SP. Scheme of type de foams used to fabricate: (H) LCEPP , and (I) LCEFP+SP. Scanning electron microscopy (SEM) shows that pore size of the LCE films ranges from 10 to 30 µm in the dry state, and can potentially increase to about 20 to 60 µm once immersed in buffer or cell culture media (cells range in size from 5 to 150 µm in diameter, see Table 4) (17, 36). The second type of porosity achieved using commercially available metal (Ni) foams resulted in LCEs with interconnected channels or LCEs foams with primary and secondary porosity once the Ni template was etched away. The third morphology achieved using micro-emulsion photopolymerization with commercially available reactive mesogens featured spherical microparticles bonded together. These nematic LCEs exhibit a globular morphology (Figure 12 C and D). Thus far, these morphologies can be ranked with an increasing void volume as follows: the Swiss-cheese-type morphology ranges from 10 to 25%, the globular morphology from ~20 to 30%, the tubular morphology from 30 to 40%, and the LCE foams well above 70% as determined by Brunauer-Emmett-Teller adsorption isotherms (BET analysis). Since we can adjust this synthetically, this range of void volume can be tailored for specific tissue target microenvironments. For example, tubular morphologies can mimic microvessels in the brain. More open foam morphologies would be better for cell co-culturing of neural and glial cells or for engineering muscle tissue by co-culturing of myoblasts, fibroblasts and endothelial cells. The Swiss-cheese morphology is of interest for tougher implantable scaffolds. 30

Mesomorphic Properties LCEs are in general highly viscous and the preparation of thin-film polarized optical microscopy (POM) samples (sandwiched between untreated glass slides) proved often to be difficult. SAXD on the other hand allowed to make clear phase assignment. The presence of cholesterol groups in polymer side chains is known to promote cholesteric (chiral nematic) phases. However, all LCEs, except the nematic globular morphology, showed the presence of two scattering maxima with q1:q2 = 1:2 ([100] and [200]) values in the medium angle region denoting an ordered layer structure (Figure 13) with an interdigitated arrangement where all cholesterol molecules are packed side-by-side. This type of arrangement corresponds to an interdigitated smectic-A phase (SmA).

Figure 13. SAXD patterns of: A) LCE-γ, and B) LCE-α, C) model of the molecular arrangement of the pendant cholesteric LC groups matching the experimentally observed layer spacing. Figure 13C, reproduced from reference (128). Copyright 2015, Macromol. Biosci., Wiley-VCH Verlag GmbH & Co. KGaA. Cell-Elastomer Confocal Imaging Three cell lines were selected to test the capability of the LCE materials as cell scaffolds. Since LCEs have been thought of as artificial muscles, myoblasts (C2C12s) (178) were the first cell line to test; other cell lines were neuroblastomas (SH-Sy5Y) and primary dermal fibroblasts (hDF). Neuroblastomas (SH-SY5Y) derived from bone marrow differentiates into neural-like cells via retinoic acid, and primary human dermal fibroblasts are isolated from adult skin fibroblasts (hDF). These cell lines exemplify various soft tissues in the human body. LCEs films were obtained (representing a “Swiss-cheese” like morphology) from molding into ampoules, then carefully microtomed to obtain 60-, 100-, or 31

200-μm thin slices. The LCEs with tubular or foam morphology were used as prepared. Globular LCEs were briefly soaked in poly-D-lysine to promote cell adherence and temporarily mask the elastomer’s hydrophobic surface. Prior to cell culture studies, all LCEs were first sterilized by washing with 70% ethanol (or exposed to O2 plasma as an alternative). The LCE casted as thin films or foams were inserted in cell culture wells, and cells were seeded on and within the LCE scaffolds using the appropriate cell culture medium. After seeding, the cells were allowed to attach and grow for an initial period of 24 to 48 hrs inside an incubator (at 37 °C and 5% CO2). Proliferation (and potentially alignment) was monitored and periodically imaged by fluorescence confocal microscopy (FCM) for extended periods (days, weeks). All LCE cell-elastomer cultures were subjected to cell viability, proliferation, and cytotoxicity tests and assays (PrestoBlue cell viability, CyQuant cell proliferation, and Cyto Tox-Fluor cytotoxicity assay). All results from the combined assays and tests demonstrated that all LCE scaffolds clearly promoted cellular viability and proliferation without any inherent cytotoxicity (128, 167, 174). All LCE cell cultures (after staining with appropriate fluorescent dyes) were imaged over specific time intervals by FCM to analyze cell proliferation, density, and alignment. LCE-cell cultures showed remarkable features such as spontaneous cell alignment that largely depended on the morphology as well as the mechanical properties of a given LCE scaffold. For example, hDF cells grown on the 3-, 4-, and 6-arm LCEs showed highly anisotropic orientational behavior and proliferation on the 4- and 6-arm LCEs but not on the 3-arm LCEs (Figure 14). Fibroblasts are known to adhere better to substrates with contact angles between 60° and 80°, which are values matched only by the 4- and 6-arm LCEs (see Figure 14) (179, 180). Also the 4-arm LCEs usually have a higher Young’s moduli than the 3-arm LCEs (~4 MPa vs. ~2 MPa), with the 6-arm LCEs showing Young’s moduli somewhere in between (167). However, cells grown on the 6-arm LCEs also show extended cell nuclei that are characteristic of a cellular response to substrate strain (181) and/or topographical cues (182). The alignment of cell nuclei and its deviation from a spherical shape was analyzed using NIH’s ImageJ (183, 184).

Figure 14. Directionality analysis of primary human dermal fibroblast (hDF) cells grown on a) 3LCEα , b) 4LCEα , and c) 6LCEα elastomer films. The insets in each of the images show a photoimage and value from contact angle measurements. Reproduced from reference (167). Copyright 2017 Macromol. Biosci., Wiley-VCH Verlag GmbH & Co. KGaA. 32

The LCE foams shown in Figure 12 feature a 3-dimensional scaffold and while C2C12 cells commonly prefer more hydrophobic scaffold surfaces, they attached and proliferated on the inside walls of even the more hydrophilic LCE foam scaffolds with interconnected hollow channels. More importantly, cells within straight channel sections tended to spontaneously align parallel to the channel walls. This alignment tendency of the cells along channels, clearly indicated by elongated nuclei (184, 185), is also retained in curved channel sections (Figure 15). In contrast, in the center of the channel junctions, cells appeared randomly oriented, as indicated by round cell nuclei. C2C12 cells within this foam LCE scaffold seemed to thrive in the more open 3D microenvironment; they can freely interact with neighboring cells without randomly growing on top of one another (commonly observed in 2D cell scaffolds) or facing any spatial constraints.

Figure 15. Fluorescence confocal microscopy images of myoblast cells (C2C12) cultured in SBC-based LCEFP+SP using DMEM as a cell culture medium and stained with DAPI (for cell nuclei): (A) 2D images stacked in z direction and (B−D) 3D views from different angles. The image in (B) is plotted with the original fluorescence intensity value (original colors can be seen on publication). Reproduced from reference (172). Copyright 2016, ACS Macro. Lett.

LCE foams offer the possibility to engineer any free macroscopic form by molding the template into stripes, rolls, or other shapes, which allows for the construction of specific spatial cell environments to study not only cell-material but also cell-cell interactions, critical to emulating endogenous tissue. 33

The globular LCEs permitted cell adhesion directly onto the globular LCE surfaces and cells continued to grow and proliferate within the bulk of the globular elastomer matrix (Figure 16) that was co-stained to distinguish between cells and LCE matrix. It appears that the surface roughness and the porosity allowed the cells to attach better and permeate into the elastomer matrix. These nematic globular LCEs were also found to be excellent scaffolds for other cell lines such as hDF and SHSY5Y. Both types of SHSY5Y cells can be seen to have attached throughout the globular LCE scaffolds and stretching across several microspheres highlighting their expansion within the matrix and showing morphological phenotypes indicative of matured cells (186). Numerous neuritic extensions can be seen that are typical of these cell types (Figure 17).

Figure 16. Confocal microscopy images (x,y-, and x,z-plane) of C2C12 myoblasts 3, 5, and 7 days after seeding, costained with DAPI for cell nuclei and with rhodamine for the LCE scaffold. Reproduced from reference (174). Copyright 2015, ACS Appl. Mat. & Interfaces. 34

Figure 17. SEM images of: (A, ii–iv) C2C12 and (B, ii–iv) SHSY5Y after 7 and 5 days seeding, respectively. The cells (see arrows, for colored view see reference 169) can be seen as extending fibers directly attaching to the matrix for expansion and proliferation. Reproduced from reference (175). Copyright 2016, Frontiers, Loop (nature publishing group – npg) open-access Creative Commons License (CC-BY). Combining LCEs as scaffolds with 3D imaging techniques; fluorescence confocal microscopy and CARS (coherent anti-stokes Raman spectroscopy) will allow for real-time imaging and spectroscopic characterization of cells in 3D as they proliferate and differentiate under the various environmental/scaffold conditions. Additionally, the techniques will offer a solution to track multi-cellular interactions for extended periods of time. This in turn will help assess cell culture integrity by monitoring cell number, layering, longevity, and expansion into the scaffold to quantify cell extensions, developmental stages, and maturation markers. Degradation All LCEs based on ε-caprolactone-D,L-lactide start to fully in vitro biodegrade (under physiological pH) after a period of about 10 weeks (128). LCEs degrade at a faster rate when exposed to acidic or basic media (pH 3 or pH 11). The degradation process follows a bulk acid-catalyzed hydrolysis-long enough to permit cell alignment, and tissue formation. The degradation is autocatalytic, due to the formation of oligocarboxylic acids as the hydrolysis products. The LCEs first experience an increase in weight in the first week due to swelling (water absorption) into the bulk (Figure 18). It is known that water diffusion into caprolactone-lactide based elastomers decreases (at temperatures above Tg) as the elastomer cross-linking density increases (145). It is expected that the weight gain throughout the degradation studies will increase periodically due to the formation of degradation products within the matrix. This is turn will enhance the degradation and draw more water into the polymer matrix 35

(via osmosis) before the resulting degradation products finally leach out of the elastomer slab. Interestingly, LCE-α degraded faster than the non-LC elastomers, showing that the biodegradation rate is also affected by LC modification in the polymer backbone.

Figure 18. Biodegradation plots (time vs. % weight change) for: (A) LCE-α and (B) unmodified elastomer. Reproduced from reference (128). Copyright 2015, Macromol. Biosci., Wiley-VCH Verlag GmbH & Co. KGaA. During the degradation process the LCE scaffolds lose their integrity, some of its mechanical properties, and LC characteristics as cells continue to grow, mature, and proliferate, i.e., overtake the space once occupied by the LCE scaffold. Most of the LCE scaffold properties such as chemical composition, morphology, and mechanical properties affect cell fate. The scaffold modifications already made allow us to control chemical and morphological parameters to design responsive cell scaffolds that promote guided cell attachment, proliferation, and alignment, all the way to differentiation into specific cell lineages solely by morphological or mechanical cues native to the LCE scaffold.

Final Remarks and Chapter Conclusions LCE scaffolds containing a build-in anisotropic response to external stimuli together with the synthetic freedom to be able to adjust almost every chemical, mechanical and morphological parameters allow us to enhance our fundamental understanding of 3D cell microenvironments and cell response to scaffolds mimicking native tissues. Careful selection of elastic properties within modified LCEs will define the intended cell type to be grown to provide the most optimal proliferation rates. Looking forward into the future of LCEs as scaffolds there are several aspects that still need attention, such as of the temperature-dependent and 36

milieu-dependent mechanical properties of our LCE scaffolds. This includes measuring the Young’s moduli at physiological temperatures and conditions. Particular emphasis must be placed on conditions existing in vitro in cell culture media and in vitro in soft tissues (immersion biodegradability studies under stress), and in longitudinal tests at various stages of biodegradation, in parallel with cell culture. Regarding other morphologies, electro-spinning or melt-spinning are currently being tested and depend on the molecular structure of the LCE. In that case acrylate-based LCEs appear suitable and can undergo reactive electro-spinning with in situ photo-cross-linking by UV irradiation, not only adding to different size porosities but also interesting elastomer anisotropies. The success of a tissue-engineered material depends on many parameters, each of which must be carefully adjusted and controlled (i.e., elastic properties) to ensure cellular growth and optimal proliferation rates. Also, any 3D scaffold implanted at a defective or malfunctioning site must comprise the basic requirements of being biocompatible, occasionally biodegradable, appropriately porous, and be the location of cell attachment and proliferation. Overall, our 3D-LCE combine all those requirements together with the properties of LC materials providing a unique test platform that opens the path to improve current and develop new future cell therapies, as well as improved scaffolds for tissue regeneration. 3D LCE scaffold properties will impact molecular/cellular research and deliver new ways to understand and provide fundamental knowledge on how local microenvironments affect, for example, cell structure and protein expression. Our LCEs will help provide new insights into plastic/developmental changes occurring between certain types of cells, help us to determine which compounds cause these effects, and elucidating a mechanism to monitor these events. We envision incorporating other cell types and constituents into the scaffold including macrophages, endothelial cells, mast cells and microglia. There is sufficient pre-clinical evidence that stem cell therapies can result in major breakthroughs to cure or arrest the progression of many disease states and help prolong life (171). There are devastating degenerative diseases that cannot simply be cured by the use of therapeutic drugs. Stem cells have great potential to replace or repair damaged tissue, and can reverse degenerative diseases (diabetes, heart, lung, and neurological disorders) without the risk of side effects, as long as ways to implant these stem cells are created using scaffolds that mimic various affected tissues.

Dedication Je voudrais dédier ce chapitre à Patrick Prévôt, mon père – MP

References 1.

Xing, Q.; Qian, Z.; Jia, W.; Ghosh, A.; Tahtinen, M.; Zhao, F. ACS Biomater. Sci. Eng. 2016, DOI: 10.1021/acsbiomaterials.6b00235. 37

2. 3.

4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.

Barker, T. H. Biomaterials 2011, 32, 4211–4214. Alberts, B.; Johnson, A.; Lewis, J.; Morgan, D.; Raff, M.; Roberts, K.; Walter, P. In Molecular Biology of the Cell, 6th ed.; Alberts, B.; Johnson, A.; Lewis, J.; Morgan, D.; Raff, M.; Roberts, K.; Walter, P., Eds.; Garland Science: New York, NY, 2015. Divya, P.; Krishnan, L. K. J. Tissue. Eng. Regener. Med. 2009, 3, 377–388. Williams, D. F. Biomaterials 2009, 30, 5897–5909. Novak, U.; Kaye, A. H. J. Clin. Neurosci. 2000, 7, 280–290. Chung, H. J.; Park, T. G. Adv. Drug Delivery Rev. 2007, 59, 249–262. Rahaman, M. N.; Day, D. E.; Bal, B. S.; Fu, Q.; Jung, S. B.; Bonewald, L. F.; Tomsia, A. P. Acta Biomater. 2011, 7, 2355–2373. Williams, D. F. Biomaterials 2008, 29, 2941–2953. Martin, Y.; Vermette, P. Biomaterials 2005, 26, 7481–7503. Scott, J. E. J. Anat. 1995, 187, 259–269. Rosso, F.; Giordano, A.; Barbarisi, M.; Barbarisi, A. J. Cell. Physiol. 2004, 199, 174–180. Badylak, S. F. Biomaterials 2007, 28, 3587–3593. Hollister, S. J. Nat. Mater. 2006, 5, 590–590. Jayakumar, R.; Prabaharan, M.; Muzzarelli, R. A. A. In Chitosan for Biomaterials II; Jayakumar, R., Prabaharan, M., Muzzarelli, R. A. A., Eds.; Springer-Verlag Berlin Heidelberg: Berlin, 2011; Vol 244. Chen, Q. Z.; Liang, S. L.; Thouas, G. A. Prog. Polym. Sci. 2013, 38, 584–671. Dhandayuthapani, B.; Yoshida, Y.; Maekawa, T.; Kumar, D. S. Int. J. Polym. Sci. 2011, 2011, 1–19. Shin, Y. C.; Lee, J. H.; Jin, L.; Kim, M. J.; Kim, C.; Hong, S. W.; Oh, J. W.; Han, D. W. J. Nanosci. Nanotechnol. 2015, 15, 7907–7912. Gosline, J.; Lillie, M.; Carrington, E.; Guerette, P.; Ortlepp, C.; Savage, K. Philos. Trans. R. Soc., B 2002, 357, 121–132. McKee, C. T.; Last, J. A.; Russell, P.; Murphy, C. J. Tissue Eng., Part B 2011, 17, 155–164. Lovett, M.; Lee, K.; Edwards, A.; Kaplan, D. L. Tissue Eng., Part B 2009, 15, 353–370. Seunarine, K.; Gadegaard, N.; Tormen, M.; O Meredith, D.; O Riehle, M.; Wilkinson, C. D. W. Nanomedicine 2006, 1, 281–296. Ahmed, E. M. J. Adv. Res. 2015, 6, 105–121. Lewis, K. J. R.; Anseth, K. S. MRS Bull. 2013, 38, 260–268. Alge, D. L.; Azagarsamy, M. A.; Donohue, D. F.; Anseth, K. S. Biomacromolecules 2013, 14, 949–953. McCall, J. D.; Luoma, J. E.; Anseth, K. S. Drug Delivery Transl. Res. 2012, 2, 305–312. DeForest, C. A.; Anseth, K. S. Annu. Rev. Chem. Biomol. 2012, 3, 421–444. Slaughter, B. V.; Khurshid, S. S.; Fisher, O. Z.; Khademhosseini, A.; Peppas, N. A. Adv. Mater. 2009, 21, 3307–3329. Mathias, E. V.; Aponte, J.; Kornfield, J. A.; Ba, Y. Colloid Polym. Sci. 2010, 288, 1655–1663. 38

30. Olsen, B. D.; Kornfield, J. A.; Tirrell, D. A. Macromolecules 2010, 43, 9094–9099. 31. Qutachi, O.; Vetsch, J. R.; Gill, D.; Cox, H.; Scurr, D. J.; Hofmann, S.; Muller, R.; Quirk, R. A.; ShakeSheff, K. M.; Rahman, C. V. Acta Biomater. 2014, 10, 5090–5098. 32. Berkland, C.; Kim, K.; Pack, D. W. Pharm. Res.-Dord. 2003, 20, 1055–1062. 33. Smith, L. A.; Liu, X. H.; Ma, P. X. Soft Matter. 2008, 4, 2144–2149. 34. Lutolf, M. P.; Hubbell, J. A. Nat. Biotechnol. 2005, 23, 47–55. 35. Levenberg, S.; Langer, R. Advances in Tissue Engineering. In Current Topics in Developmental Biology; Schatten, G. P., Ed.; Elsevier: New York, NY, 2004; Vol. 61, pp 113−134. 36. Chang, H. I.; Wang, Y. Cell Responses to Surface and Architecture of Tissue Engineering Scaffolds. In Regenerative Medicine and Tissue Engineering Cells and Biomaterials; Eberli, D., Ed.; InTech: Rijeka, Croatia, 2011; pp 569−588. 37. Zhang, C. Elastic Degradable Polyurethanes for Biomedical Applications, Ph.D thesis, Clemson University, Clemson, SC, 2006. 38. Ward, I. M.; Pinnock, P. R. Br. J. Appl. Phys. 1966, 17, 3–32. 39. Balla, V. K.; Bodhak, S.; Bose, S.; Bandyopadhyay, A. Acta Biomater. 2010, 6, 3349–3359. 40. Dabrowski, B.; Swieszkowski, W.; Godlinski, D.; Kurzydlowski, K. J. J. Biomed. Mater. Res., Part B: Appl. Biomater. 2010, 95B, 53–61. 41. Matassi, F.; Nistri, L.; Paez, D. C.; Innocenti, M. Clin. Cases Miner. Bone Metab. 2011, 8, 21–24. 42. Karageorgiou, V.; Kaplan, D. Biomaterials 2005, 26, 5474–5491. 43. Mastrogiacomo, M.; Scaglione, S.; Martinetti, R.; Dolcini, L.; Beltrame, F.; Cancedda, R.; Quarto, R. Biomaterials 2006, 27, 3230–3237. 44. Wilson, C. E.; de Bruijn, J. D.; van Blitterswijk, C. A.; Verbout, A. J.; Dhert, W. J. A. J. Biomed. Mater. Res. A 2004, 68A, 123–132. 45. Xu, H. H. K.; Weir, M. D.; Burguera, E. F.; Fraser, A. M. Biomaterials 2006, 27, 4279–4287. 46. Ginebra, M. P.; Delgado, J. A.; Harr, I.; Almirall, A.; Del Valle, S.; Planell, J. A. J. Biomed. Mater. Res. A 2007, 80, 351–361. 47. del Real, R. P.; Wolke, J. G. C.; Vallet-Regi, M.; Jansen, J. A. Biomaterials 2002, 23, 3673–3680. 48. Yuan, H. P.; de Bruijn, J. D.; Zhang, X. D.; van Blitterswijk, C. A.; de Groot, K. J. Biomed. Mater. Res. 2001, 58, 270–276. 49. Chen, Q. Z. Z.; Thompson, I. D.; Boccaccini, A. R. Biomaterials 2006, 27, 2414–2425. 50. LeGeros, R. Z. Clin. Orthop. Relat. Res. 2002, 395, 81–98. 51. Kenny, S. M.; Buggy, M. J. Mater. Sci. Mater. Med. 2003, 14, 923–938. 52. Wheeler, D. L.; Stokes, K. E.; Park, H. M.; Hollinger, J. O. J. Biomed. Mater. Res. 1997, 35, 249–254. 53. Wheeler, D. L.; Stokes, K. E.; Hoellrich, R. G.; Chamberland, D. L.; McLoughlin, S. W. J. Biomed. Mater. Res. 1998, 41, 527–533. 39

54. Boccaccini, A. R.; Ma, P. X. In Tissue Engineering Using Ceramics and Polymers, 2nd ed.; Boccaccini, A. R., Ma, P. X., Eds.; Elsevier Science: Oxford, U.K., 2014. 55. Habraken, W.; Wolke, J. G. C.; Jansen, J. A. Adv. Drug Delivery Rev. 2007, 59, 234–248. 56. Kim, I. Y.; Seo, S. J.; Moon, H. S.; Yoo, M. K.; Park, I. Y.; Kim, B. C.; Cho, C. S. Biotechnol. Adv. 2008, 26, 1–21. 57. Kean, T.; Thanou, M. Adv. Drug Delivery Rev. 2010, 62, 3–11. 58. Zilla, P. P.; Greisler, H. P. In Tissue Engineering of Vascular Prosthetic Grafts, 1st ed.; Zilla, P. P., Greisler, H. P., Eds.; R.G. Landes Company: Austin, TX, 1999. 59. Vaissiere, G.; Chevallay, B.; Herbage, D.; Damour, O. Med. Biol. Eng. Comput. 2000, 38, 205–210. 60. Suh, J. K. F.; Matthew, H. W. T. Biomaterials 2000, 21, 2589–2598. 61. Madihally, S. V.; Matthew, H. W. T. Biomaterials 1999, 20, 1133–1142. 62. Khor, E.; Lim, L. Y. Biomaterials 2003, 24, 2339–2349. 63. Lewandowska, K.; Sionkowska, A.; Kaczmarek, B.; Furtos, G. Mol. Cryst. Liq. Cryst. 2014, 590, 193–198. 64. Lu, Q. J.; Ganesan, K.; Simionescu, D. T.; Vyavahare, N. R. Biomaterials 2004, 25, 5227–5237. 65. Urry, D. W. Proc. Natl. Acad. Sci. U.S.A. 1971, 68, 810–814. 66. Glicklis, R.; Shapiro, L.; Agbaria, R.; Merchuk, J. C. Biotechnol. Bioeng. 2000, 67, 344–353. 67. Chung, T. W.; Yang, J.; Akaike, T.; Cho, K. Y.; Nah, J. W.; Kim, S. I.; Cho, C. S. Biomaterials 2002, 23, 2827–2834. 68. Rowley, J. A.; Mooney, D. J. J. Biomed. Mater. Res. A. 2002, 60, 217–223. 69. Ning, L. Q.; Xu, Y. T.; Chen, X. B.; Schreyer, D. J. J. Biomater. Sci., Polym. E 2016, 27, 898–915. 70. Peter, S. J.; Miller, M. J.; Yasko, A. W.; Yaszemski, M. J.; Mikos, A. G. J. Biomed. Mater. Res. 1998, 43, 422–427. 71. Lu, L.; Peter, S. J.; Lyman, M. D.; Lai, H. L.; Leite, S. M.; Tamada, J. A.; Uyama, S.; Vacanti, J. P.; Langer, R.; Mikos, A. G. Biomaterials 2000, 21, 1837–1845. 72. Oh, S. H.; Kang, S. G.; Kim, E. S.; Cho, S. H.; Lee, J. H. Biomaterials 2003, 24, 4011–4021. 73. Rowlands, A. S.; Lim, S. A.; Martin, D.; Cooper-White, J. J. Biomaterials 2007, 28, 2109–2121. 74. Guo, B. L.; Ma, P. X. Sci. China-Chem. 2014, 57, 490–500. 75. Drumright, R. E.; Gruber, P. R.; Henton, D. E. Adv. Mater. 2000, 12, 1841–1846. 76. Garlotta, D. J. Polym. Environ. 2001, 9, 63–84. 77. Jing, D. Y.; Wu, L. B.; Ding, J. D. Macromol. Biosci. 2006, 6, 747–757. 78. Wu, L. B.; Zhang, H.; Zhang, J. C.; Ding, J. D. Tissue Eng. 2005, 11, 1105–1114. 79. Pan, Z.; Ding, J. D. Interface Focus 2012, 2, 366–377. 80. Yang, J.; Shi, G. X.; Bei, J. Z.; Wang, S. G.; Cao, Y. L.; Shang, Q. X.; Yang, G. G.; Wang, W. J. J. Biomed. Mater. Res 2002, 62, 438–446. 40

81. Ratner, B. D.; Hoffman, A. S.; Schoen, F. J.; Lemons, J. E. In Biomaterials Science: An Introduction to Materials in Medicine, 2nd ed.; Ratner, B. D., Hoffman, A. S., Schoen, F. J., Lemons, J. E., Eds.; Elsevier Science: Oxford, U.K., 2004. 82. Kaplan, D. In Biopolymers from Renewable Resources, 1st ed.; Kaplan, D., Ed.; Springer-Verlag Berlin Heidelberg: Berlin, 1998. 83. Liu, X. H.; Ma, P. X. Biomaterials 2010, 31, 259–269. 84. Gunatillake, P.; Mayadunne, R.; Adhikari, R. Biotechnol. Annu. Rev. 2006, 12, 301–347. 85. Mondal, D.; Griffith, M.; Venkatraman, S. S. Int. J. Polym. Mater. Polym. 2016, 65, 255–265. 86. Fields, R. D.; Rodriguez, F.; Finn, R. K. J. Appl. Polym. Sci. 1974, 18, 3571–3579. 87. Woodward, S. C.; Brewer, P. S.; Moatamed, F.; Schindler, A.; Pitt, C. G. J. Biomed. Mater. Res. 1985, 19, 437–444. 88. Patricio, T.; Domingos, M.; Gloria, A.; Bartolo, P. Procedia CIRP 2013, 5, 110–114. 89. Vieira, A. C.; Vieira, J. C.; Guedes, R. M.; Marques, A. T. Mater. Sci. Forum. 2010, 636−637, 825–832. 90. Yeong, W. Y.; Sudarmadji, N.; Yu, H. Y.; Chua, C. K.; Leong, K. F.; Venkatraman, S. S.; Boey, Y. C. F.; Tan, L. P. Acta Biomater. 2010, 6, 2028–2034. 91. Alvarez-Perez, M. A.; Guarino, V.; Cirillo, V.; Ambrosio, L. Biomacromolecules 2010, 11, 2238–2246. 92. Zhang, Y. Z.; Venugopal, J.; Huang, Z. M.; Lim, C. T.; Ramakrishna, S. Biomacromolecules 2005, 6, 2583–2589. 93. Krol, P. Prog. Mater. Sci. 2007, 52, 915–1015. 94. Martina, M.; Hutmacher, D. W. Polym. Int. 2007, 56, 145–157. 95. Kucinska-Lipka, J.; Gubanska, I.; Janik, H.; Sienkiewicz, M. Mater. Sci. Eng., C 2015, 46, 166–176. 96. Guelcher, S. A. Tissue Eng., Part B 2008, 14, 3–17. 97. Wise, S. G.; Liu, H. J.; Yeo, G. C.; Michael, P. L.; Chan, A. H. P.; Ngo, A. K. Y.; Bilek, M. M. M.; Bao, S. S.; Weiss, A. S. Tissue Eng., Part A 2016, 22, 524–533. 98. Barrioni, B. R.; de Carvalho, S. M.; Orefice, R. L.; de Oliveira, A. A. R.; Pereira, M. D. Mater. Sci. Eng., C 2015, 52, 22–30. 99. Zhou, L.; He, H.; Jiang, C.; He, S. J. Appl. Polym. Sci. 2015, 132, 42196. 100. Wang, Y. D.; Kim, Y. M.; Langer, R. J. Biomed. Mater. Res. A 2003, 66A, 192–197. 101. Wang, Y. D.; Ameer, G. A.; Sheppard, B. J.; Langer, R. Nat Biotechnol 2002, 20, 602–606. 102. Li, X. D.; Hong, A. T. L.; Naskar, N.; Chung, H. J. Biomacromolecules 2015, 16, 1525–1533. 103. Patel, A.; Gaharwar, A. K.; Iviglia, G.; Zhang, H. B.; Mukundan, S.; Mihaila, S. M.; Demarchi, D.; Khademhosseini, A. Biomaterials 2013, 34, 3970–3983. 41

104. Li, Y.; Cook, W. D.; Moorhoff, C.; Huang, W. C.; Chen, Q. Z. Polym. Int. 2013, 62, 534–547. 105. Lendlein, A.; Kelch, S. Shape-memory polymers. Angew. Chem., Int. Ed. 2002, 41, 2034–2057. 106. Neuss, S.; Blomenkamp, I.; Stainforth, R.; Boltersdorf, D.; Jansen, M.; Butz, N.; Perez-Bouza, A.; Knuchel, R. Biomaterials 2009, 30, 1697–1705. 107. Chen, M. C.; Tsai, H. W.; Chang, Y.; Lai, W. Y.; Mi, F. L.; Liu, C. T.; Wong, H. S.; Sung, H. W. Biomacromolecules 2007, 8, 2774–2780. 108. Gall, K.; Yakacki, C. M.; Liu, Y. P.; Shandas, R.; Willett, N.; Anseth, K. S. J. Biomed. Mater. Res., Part A 2005, 73a, 339–348. 109. Yakacki, C. M.; Shandas, R.; Safranski, D.; Ortega, A. M.; Sassaman, K.; Gall, K. Adv. Funct. Mater. 2008, 18, 2428–2435. 110. Kim, H. W.; Knowles, J. C.; Kim, H. E. Biomaterials 2004, 25, 1279–1287. 111. Rezwan, K.; Chen, Q. Z.; Blaker, J. J.; Boccaccini, A. R. Biomaterials 2006, 27, 3413–3431. 112. Mark, J. E.; Erman, B.; Eirich, F. R. Science and Technology of Rubber I, 2nd ed.; Mark, J. E., Erman, B., Eirich, F. R., Eds.; Elsevier: San Diego, CA, 1994. 113. Lee, S. H.; Kim, B. S.; Kim, S. H.; Choi, S. W.; Jeong, S. I.; Kwon, I. K.; Kang, S. W.; Nikolovski, J.; Mooney, D. J.; Han, Y. K.; Kim, Y. H. J. Biomed. Mater. Res., Part A 2003, 66, 29–37. 114. Zdrahala, R. J. J. Biomater. Appl. 1996, 11, 37–61. 115. Cardy, R. H. J. Natl. Cancer Inst. 1979, 62, 1107–1116. 116. Guan, J. J.; Sacks, M. S.; Beckman, E. J.; Wagner, W. R. Biomaterials 2004, 25, 85–96. 117. Alperin, C.; Zandstra, P. W.; Woodhouse, K. A. Biomaterials 2005, 26, 7377–7386. 118. Chen, Q. Z.; Bismarck, A.; Hansen, U.; Junaid, S.; Tran, M. Q.; Harding, S. E.; Ali, N. N.; Boccaccini, A. R. Biomaterials 2008, 29, 47–57. 119. Lee, K. W.; Wang, Y. D. J. Vis. Exp. 2011, 50, e2691. 120. Zhang, X. L.; Jia, C. L.; Qiao, X. Y.; Liu, T. Y.; Sun, K. Polym. Test. 2016, 54, 118–125. 121. Masoumi, N.; Johnson, K. L.; Howell, M. C.; Engelmayr, G. C. Acta Biomater. 2013, 9, 5974–5988. 122. Hiob, M. A.; Crouch, G. W.; Weiss, A. S. Curr. Opin. Biotechnol. 2016, 40, 149–154. 123. Kerativitayanan, P.; Gaharwar, A. K. Acta Biomater. 2015, 26, 34–44. 124. Gerecht, S.; Townsend, S. A.; Pressler, H.; Zhu, H.; Nijst, C. L. E.; Bruggeman, J. P.; Nichol, J. W.; Langer, R. Biomaterials 2007, 28, 4826–4835. 125. HiljanenVainio, M. P.; Orava, P. A.; Seppala, J. V. J. Biomed. Mater. Res. 1997, 34, 39–46. 126. Younes, H. M.; Bravo-Grimaldo, E.; Amsden, B. G. Biomaterials 2004, 25, 5261–5269. 127. Storey, R. F.; Warren, S. C.; Allison, C. J.; Puckett, A. D. Polymer 1997, 38, 6295–6301. 42

128. Sharma, A.; Neshat, A.; Mahnen, C. J.; Nielsen, A. D.; Snyder, J.; Stankovich, T. L.; Daum, B. G.; LaSpina, E. M.; Beltrano, G.; Gao, Y.; Li, S.; Park, B.-W.; Clements, R. J.; Freeman, E. J.; Malcuit, C.; McDonough, J. A.; Korley, L. T. J.; Hegmann, T.; Hegmann, E. Macromol. Biosci. 2015, 15, 200–214. 129. Bostman, O.; Pihlajamaki, H. Biomaterials 2000, 21, 2615–2621. 130. Nishino, T.; Matsui, R.; Nakamae, K. J. Polym. Sci.,Polym. Phys. 1999, 37, 1191–1196. 131. Serrano, M. C.; Chung, E. J.; Ameer, G. A. Adv. Funct. Mater. 2010, 20, 192–208. 132. Stewart, G. T. Liq. Cryst. 2003, 30, 751–751. 133. Mitov, M.; Dessaud, N. Nat. Mater. 2006, 5, 361–364. 134. Lowe, A. M.; Abbott, N. L. Chem. Mater. 2012, 24, 746–758. 135. Agarwal, A.; Huang, E.; Palecek, S.; Abbott, N. L. Adv. Mater. 2008, 20, 4804–4809. 136. Ohm, C.; Brehmer, M.; Zentel, R. Applications of Liquid Crystalline Elastomers. In Liquid Crystal Elastomers: Materials and Applications; DeJeu, W. H., Ed.; Springer-Verlag Berlin Heidelberg: Berlin, 2012; Vol. 250, pp 49−93. 137. de Jeu, W. H.; Obraztsov, E. P.; Ostrovskii, B. I.; Ren, W.; McMullan, P. J.; Griffin, A. C.; Sanchez-Ferrer, A.; Finkelmann, H. Eur. Phys. J. E 2007, 24, 399–409. 138. Hashimoto, S.; Yusuf, Y.; Krause, S.; Finkelmann, H.; Cladis, P. E.; Brand, H. R.; Kai, S. Appl. Phys. Lett. 2008, 92, 181902. 139. Sanchez-Ferrer, A.; Fischl, T.; Stubenrauch, M.; Albrecht, A.; Wurmus, H.; Hoffmann, M.; Finkelmann, H. Adv. Mater. 2011, 23, 4526–4530. 140. Palffy-Muhoray, P.; Meyer, R. B. Nat. Mater. 2004, 3, 139–140. 141. Camacho-Lopez, M.; Finkelmann, H.; Palffy-Muhoray, P.; Shelley, M. Nat. Mater. 2004, 3, 307–310. 142. Broemmel, F.; Kramer, D.; Finkelmann, H. Preparation of Liquid Crystalline Elastomers. In Liquid Crystal Elastomers: Materials and Applications; DeJeu, W. H., Ed.; Springer-Verlag Berlin Heidelberg: Berlin, 2012; Vol. 250, pp 1−48. 143. Martinoty, P.; Stein, P.; Finkelmann, H.; Pleiner, H.; Brand, H. R. Eur. Phys. J. E 2004, 14, 311–321. 144. Stein, P.; Assfalg, N.; Finkelmann, H.; Martinoty, P. Eur. Phys. J. E 2001, 4, 255–262. 145. El-Laboudy, H.; Shaker, M. A.; Younes, H. M. Soft Matter 2011, 9, 409–428. 146. Fleischmann, E. K.; Zentel, R. Angew. Chem., Int. Ed. 2013, 52, 8810–8827. 147. deGennes, P. G.; Hebert, M.; Kant, R. Macromol. Symp. 1997, 113, 39–49. 148. Woltman, S. J.; Jay, G. D.; Crawford, G. P. Nat. Mater. 2007, 6, 929–938. 149. Herzer, N.; Guneysu, H.; Davies, D. J. D.; Yildirim, D.; Vaccaro, A. R.; Broer, D. J.; Bastiaansen, C. W. M.; Schenning, A. P. H. J. J. Am. Chem. Soc. 2012, 134, 7608–7611. 150. Ohm, C.; Brehmer, M.; Zentel, R. Adv. Mater. 2010, 22, 3366–3387. 151. Artal, C.; Ros, M. B.; Serrano, J. L.; Pereda, N.; Etxebarria, J.; Folcia, C. L.; Ortega, J. Macromolecules 2001, 34, 4244–4255. 43

152. Fleischmann, E.-K.; Liang, H.-L.; Kapernaum, N.; Giesselmann, F.; Lagerwall, J.; Zentel, R. Nat. Commun. 2012, 3. 153. Finkelmann, H.; Kim, S. T.; Munoz, A.; Palffy-Muhoray, P.; Taheri, B. Adv. Mater. 2001, 13, 1069–1072. 154. Yamada, M.; Kondo, M.; Mamiya, J. I.; Yu, Y.; Kinoshita, M.; Barrett, C. J.; Ikeda, T. Angew. Chem., Int. Ed. 2008, 47, 4986–4988. 155. Agrawal, A.; Yun, T. H.; Pesek, S. L.; Chapman, W. G.; Verduzco, R. Soft Matter 2014, 10, 1411–1415. 156. Agrawal, A.; Chipara, A. C.; Shamoo, Y.; Patra, P. K.; Carey, B. J.; Ajayan, P. M.; Chapman, W. G.; Verduzco, R. Nat. Commun. 2013, 4. 157. Bergmann, G. H. F.; Finkelmann, H.; Percec, V.; Zhao, M. Y. Macromol. Rapid. Commun. 1997, 18, 353–360. 158. Yakacki, C. M.; Saed, M.; Nair, D. P.; Gong, T.; Reed, S. M.; Bowman, C. N. RSC Adv. 2015, 5, 18997–19001. 159. Saed, M. O.; Torbati, A. H.; Nair, D. P.; Yakacki, C. M. J. Vis. Exp. 2016. 160. Kim, H.; Zhu, B. H.; Chen, H. Y.; Adetiba, O.; Agrawal, A.; Ajayan, P.; Jacot, J. G.; Verduzco, R. J. Vis. Exp. 2016. 161. Li, M. H.; Keller, P. Philos. Trans. R. Soc., A 2006, 364, 2763–2777. 162. Buguin, A.; Li, M. H.; Silberzan, P.; Ladoux, B.; Keller, P. J. Am. Chem. Soc. 2006, 128, 1088–1089. 163. Finkelmann, H.; Shahinpoor, M. Proc. SPIE 2002, 4695, 459–464. 164. Thomsen, D. L.; Keller, P.; Naciri, J.; Pink, R.; Jeon, H.; Shenoy, D.; Ratna, B. R. Macromolecules 2001, 34, 5868–5875. 165. Mayer, S.; Zentel, R. Liquid crystalline polymers and elastomers. Curr. Opin. Solid State Mater. Sci. 2002, 6, 545–551. 166. Cheng, H.; Hill, P. S.; Siegwart, D. J.; Vacanti, N.; Lytton-Jean, A. K. R.; Cho, S. W.; Ye, A.; Langer, R.; Anderson, D. G. Adv. Mater. 2011, 23, H95–H100. 167. Sharma, A.; Mori, T.; Mahnen, C. J.; Everson, H. R.; Leslie, M. T.; Nielsen, A. d.; Lussier, L.; Zhu, C.; Malcuit, C.; Hegmann, T.; McDonough, J. A.; Freeman, E. J.; Korley, L. T. J.; Clements, R. J.; Hegmann, E. Macromol. Biosci. 2017, 17, 1600278. 168. Gerhold, J. M.; Cansiz-Arda, Ş.; Lõhmus, M.; Engberg, O.; Reyes, A.; van Rennes, H.; Sanz, A.; Holt, I. J.; Cooper, H. M.; Spelbrink, J. N. Sci. Rep. 2015, 5, 15292. 169. Reinitzer, F. Ann. Phys. Berlin 1908, 27, 213–224. 170. Reinitzer, F. Monatsh. Chem. 1888, 9, 20. 171. Nottelet, B.; Di Tommaso, C.; Mondon, K.; Gurny, R.; Moeller, M. J. Polym. Sci., Polym. Chem. 2010, 48, 3244–3254. 172. Gao, Y. X.; Mori, T.; Manning, S.; Zhao, Y.; Nielsen, A. D.; Neshat, A.; Sharma, A.; Mahnen, C. J.; Everson, H. R.; Crotty, S.; Clements, R. J.; Malcuit, C.; Hegmann, E. ACS Macro Lett. 2016, 5, 14–19. 173. deGennes, P. G. C. R. Acad. Sci. II 1997, 324, 343–348. 174. Bera, T.; Freeman, E. J.; McDonough, J. A.; Clements, R. J.; Aladlaan, A.; Miller, D. W.; Malcuit, C.; Hegmann, T.; Hegmann, E. ACS Appl. Mater. Interfaces 2015, 7, 14528–35. 44

175. Bera, T.; Malcuit, C.; Clements, R. J.; Hegmann, E. Front. Mater. 2016, 3, 31–39. 176. Lub, J.; Broer, D. J.; Wegh, R. T.; Peeters, E.; van der Zande, B. M. I. Mol. Cryst. Liq. Cryst. 2005, 429, 77–99. 177. Harris, K. D.; Cuypers, R.; Scheibe, P.; van Oosten, C. L.; Bastiaansen, C. W. M.; Lub, J.; Broer, D. J. J. Mater. Chem. 2005, 15, 5043–5048. 178. Peltzer, J.; Colman, L.; Cebrian, J.; Musa, H.; Peckham, M.; Keller, A. Dev. Dyn. 2008, 237, 1412–23. 179. Goddard, J. M.; Hotchkiss, J. H. Prog. Polym. Sci. 2007, 32, 698–725. 180. Uchida, E.; Uyama, Y.; Ikada, Y. Langmuir 1993, 9, 1121–1124. 181. Tremblay, D.; Andrzejewski, L.; Leclerc, A.; Pelling, A. E. Cytoskeleton (Hoboken, N.J.) 2013, 70, 837–48. 182. Chalut, K. J.; Kulangara, K.; Giacomelli, M. G.; Wax, A.; Leong, K. W. Soft Matter 2010, 6, 1675–1681. 183. Schneider, C. A.; Rasband, W. S.; Eliceiri, K. W. Nat. Methods 2012, 9, 671–5. 184. Charest, J. L.; García, A. J.; King, W. P. Biomaterials 2007, 28, 2202–10. 185. Marshall, J. E.; Gallagher, S.; Terentjev, E. M.; Smoukov, S. K. J. Am. Chem. Soc. 2014, 136, 474–479. 186. Teppola, H.; Sarkanen, J. R.; Jalonen, T. O.; Linne, M. L. Neurochem. Res. 2016, 41, 731–747.

45

Chapter 2

Development of Redox Nanomedicine for Gastrointestinal Complications via Oral Administration Route Long Binh Vong1,2 and Yukio Nagasaki1,3,4,* 1Department

of Materials Science, Graduate School of Pure and Applied Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8573, Japan 2Department of Biochemistry, Faculty of Biology and Biotechnology, University of Science, Vietnam National University Ho Chi Minh City (VNU-HCM), Ho Chi Minh City 702500, Vietnam 3Master’s School of Medical Sciences, Graduate School of Comprehensive Human Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8575, Japan 4Satellite Laboratory, International Center for Materials Nanoarchitectonics (WPI-MANA), National Institute for Materials Science (NIMS), University of Tsukuba, Tsukuba, Ibaraki 305-8573, Japan *E-mail: [email protected]; Phone: +81-29-853-5749; Fax: +81-29-853-5749

Oral drug administration is most widely used route and convenient for patients, although numerous challenges for this administration route are required to overcome. Nanoparticle-based drug delivery systems have been studied as novel therapeutics for treating numerous human diseases. These systems can improve drug bioavailability and specifically deliver therapeutic agents to diseased tissues, enhancing treating efficacy and minimizing unwanted adverse effects. In this chapter, we describe recent researches on treating oxidative stress-associated diseases including gastrointestinal disorders, using traditional medications and nanoparticle-based drug delivery systems via oral administration route. We also describe the development and recent research of our newly designed redox nanoparticle as a novel nanotherapeutics. The preparation of redox nanoparticle, the evaluation of its toxicity in vitro and in vivo, as well as the therapeutic application for

© 2017 American Chemical Society

inflammatory bowel disease and colon cancer will be also discussed in this chapter.

Oxidative Stress and Inflammatory Bowel Disease Reactive oxygen species (ROS) are highly reactive molecules containing oxygen including superoxide anion, hydroxyl radical, hydrogen peroxide, singlet oxygen. Mitochondria are well-known as important source to generate ROS in most mammalian cells (1). Under normal physiological conditions, ROS play critical functions as redox messengers for intracellular signaling and regulation (Figure 1) (2, 3). Many important intracellular transcription factors and heat shock proteins are activated by ROS. In addition, ROS regulate ion transporter, cellular pH, and immune systems (4). In the living organisms, endogenous antioxidants such as superoxide dismutase, glutathione, and catalase contribute to consistently maintain the balance between pro- and anti-oxidation. However, under oxidative stress conditions, excessive generation of ROS causes oxidative damage to biomolecules including nucleic acids, lipids, and proteins. ROS-mediated injuries are strongly related to numerous human diseases (Figure 1) such as inflammatory, stroke, myocardial infarction, diabetes, neurodegenerative diseases, aging, cancer, ect… (5–8)

Figure 1. Roles of reactive oxygen species (ROS) in physiological and pathogenesis pathways (see color insert)

Inflammatory bowel diseases (IBD) comprise mainly two types of chronic relapsing intestinal disorders: Crohn’s disease (CD) and ulcerative colitis (UC). While CD involves any party of gastrointestinal (GI) tract, UC occurs only in the inner lining of colon and rectum. Millions of patients are affected by IBD worldwide, and the prevalence and incidence of IBD is increasing (150–250/100,000 population), especially in US and Western countries (9, 10). Importantly, after 30 years of living with these diseases, 18–20% of UC and 8% of CD patients develop colitis-associated colon cancer (CAC), the third most common malignancy and one of the major causes of cancer-related death (5, 6). 48

For many years, there were only 2 treatment options for IBD: corticosteroids and mesalamine (11, 12). Although they are effective in treating IBD to some extent, their severe side effects have raised significant concerns among both physicians and patients, and limited their use. In addition, anti-tumor necrosis factor alpha (TNF-α) antibody is employed to suppress inflammation of IBD, which works well though it is cost-oriented therapy with multiple side effects (13). Although the etiology and pathogenesis of IBD are not exactly understood, environmental factors, genetic susceptibility, immune response, and luminal microbial antigens are considered as main factors contributing to the onset of IBD (14, 15). These factors support an important redox equilibrium between oxidant and antioxidant for intestinal homeostasis. Recently, many studies have focused on the imbalance of redox equilibrium as an etiologic factor of IBD. In fact, oxidative stress has been proposed as a critical mechanism underlying pathophysiology of IBD (16). The intestinal mucosa of patients with IBD is characterized by overproduction of ROS and an imbalance of important antioxidants, leading to oxidative damage. Self-sustaining cycles of oxidant production may amplify inflammation and mucosal injury (17–20). Deficiency of important endogenous antioxidant enzymes causes the development of spontaneous colitis in mice (21). On the other hand, a number of antioxidants and free radical scavengers have been studied and exhibited some efficacies in animal models of IBD (22). For examples, conventional antioxidant such as tocopherols, polyphenols, glutathione, and N-acetylcystein have been investigated to treat murine model of colitis (23–25). However, these compounds are not completely effective due to poor solubility, low bioavailability, non-specific drug distribution, low retention in the GI tract, and severe adverse side effects. Maintaining the balance of redox equilibrium as normal physiological level is a critical issue for effective treatment of IBD. If the antioxidant compounds are specifically targeted to only the diseased sites in GI tract and effectively scavenge excessively generated ROS without any disturbance to healthy tissues, they represent a safe and highly effective therapy for IBD patients.

Development of Nanomedicine for IBD via Oral Administration Route Barriers for Oral Drug Administration While oral administration of drugs are preferred by patient due to its convenience and compliance, these low-molecular-weight (LMW) drugs are not always effective due to instability, nonspecific drug distribution, low retention in the GI tract, and absorption in the bloodstream, causing undesired adverse side effects in the entire body. Oral drug delivery faces big challenge due to many barriers in GI tract such as hash acidic pH in stomach, mucus layer, and intestinal microflora (Figure 2), preventing the therapeutic efficacy of medication. 49

Figure 2. Barriers in GI tract for oral drug delivery. The wide range of pH values (left), the thickness of mucus layer (middle), and bacteria population (right) in GI tract (see color insert)

pH GI Tract Orally administered drugs are strongly affected by wide pH range in human GI tract (Figure 2). After administration, drugs pass quickly through the esophagus to reach to the stomach, where the pH values lie within 1–2.5 and up to pH 5 after feeding (26). This gastric acid in the stomach has the function to digest food and to catalytically cleave zymogen such as pepsinogen to produce its corresponding active pepsin, which can breaks down ingested protein. After passing stomach, orally administered drugs enter small intestine, where the majority of drug absorption occurs. The pH values of small intestine lie in the range of pH 6.15–7.35 in the proximal region, rising to pH 6.80–7.88 in the distal area. After passage through the small intestine the drugs reach the large intestine at which point the pH lowers slightly to pH 5.26–6.72 in the ascending colon, and pH 5.20–7.02 in the descending colon (27, 28).

Mucus Barrier In the GI tract, epithelial cells are covered by a thick mucus layer, which is composed of glycoproteins, lipids, and others macromolecules. The mucus layer has the function to create environment for colonization by commensal and beneficial bacteria and provide a barrier to pathogenic bacteria and other potentially damaging compounds reaching epithelial surface (29, 30). The thickness of mucus layer varies and increases from upper GI part to lower GI tract (Figure 2) (30). Due to rapid secretion and shedding, mucus layer in GI tract enable to trap pathogens and foreign particulates to effectively remove them. In the pathogenic conditions of GI tract such as intestinal inflammation and UC, the thin mucus layer and crypt loss were observed (31, 32), resulting in the penetration of pathogenic bacteria and foreign compounds to epithelial cells. It remains big challenge for oral drug delivery via mucus layer in GI tract. 50

Intestinal Bacteria GI tract is a main source of commensal bacteria with more than 500 different bacterial species living together in a state of balance. The distribution of bacteria population generally increases from 102–103/mL in the stomach to 104–107/mL in the small intestine and up to 1012/mL in the large intestine (Figure 2) (33). The intestinal microflora play the important role in regulating host inflammatory response and in maintaining the immunological homeostasis (34, 35). It has been reported that the disruption of the normally stable microflora is predicted to result in the pathological conditions of IBD (36, 37), as well as associated with many different diseases (38). The GI surface and bacteria regularly contact with many food components and chemical substances, which may positively or negatively affect the balance of intestinal microflora (39). Intestinal bacteria play many important roles in the absorption and metabolism of ingested compounds as well as in the energy regulation (40). Nanomedicine for Inflammatory Bowel Disease Nanoparticle-based drug delivery systems such as liposome and polymeric micelles have gained worldwide attention as a new medical technology, because they change biodistribution of drugs to result in increasing therapeutic effect of drugs significantly, and suppressing their adverse side effect (41, 42). In particular, the intratumoral microdistribution of nanoparticle has been studied for over two decades that nanoparticles can accumulate in sites of tumor due to the increased vascular permeability or enhanced permeability and retention effect (43–45). For IBD treatment, although oral administration is the most common method for drug administration, many commercially available drugs exhibit low bioavailability and stability in GI tract for achieving therapeutic efficiency. Pharmaceutical strategies using nano-carriers as a drug vehicle are promising approach that can protect drugs from degradation in GI tract and enhance the specific drug delivery to diseased sites, resulting in more efficient therapy (46, 47). Wilson et al. developed thioketal nanoparticles (TKNs) for oral delivery of small interfering RNAs (siRNAs) to intestinal inflammatory sites (48). The TKNs are composed of ROS-sensitive polymers [poly-(1,4-phenyleneacetone dimethylene thioketal)], that are stable to acid-, base- and protein-catalysed degradation. Orally administered siRNA-loaded TKNs undergo degradation at highly generated ROS levels at intestinal inflamed sites and release siRNA suppressing TNF-α and inflammation. Laroui et al. encapsulated anti-inflammatory tripeptides-loaded nanoparticles (NP-KPV) into a chitosan/alginate hydrogel, exhibiting different swelling degrees against pH values in digestive tract (49). This hydrogel is stable under pH of a stimulating gastric fluid (pH 1 to 3), while this biomaterial exposed to intestinal and colonic pH solutions (pH 5 to 7) completely collapses and releases NP-KPV reducing inflammation in colitis mice. However, the physicochemical entrapments of bioactive compounds into nanoparticle cause an early uncontrollable release, reducing the therapeutic efficacy, and increasing the risk of undesired adverse effects. In order to reduce drug leakage and uncontrollable release, Pertuit et al. conjugated 5-aminosalicylic acid (5-ASA) to 51

poly(ε-caprolactone) (PLC)-based nanoparticles, that slowed down drug release and minimized early drug release in upper parts of intestine, a prerequisite for oral colonic drug delivery (50). However, this carrier system exhibited extremely low drug encapsulated capacity due to limited number of conjugating sites of PLC.

Design of Nitroxide Radicals Containing Nanoparticles (RNPO) for Oral Drug Delivery Nitroxide radicals are stable chemical compounds containing the tertiary amine (R2N-O•) functional group with an unpaired electron, providing them with paramagnetic properties and the ability to participate in redox reactions. Nitroxide radicals have been utilized for many years as biophysical tools including electron spin resonance (ESR) spectroscopic probes, ESR and magnetic resonance imaging agents (51), and oxidizing agents (52). The antioxidant activity of nitroxide radicals is reported as a superoxide dismutase mimic to scavenge superoxide radicals, while they also possess peroxidase mimic and hydroxyl radial scavenging activities. In addition, unlike other antioxidants that protect through stoichiometric reactions, nitroxide radicals can inhibit lipid peroxidation by participating in redox reactions at every step (53, 54). Due to their unique ROS scavenging property, many of interesting biochemical interactions and therapeutic applications of nitroxide radicals have recently been studies for biomedical use such as radiation protector (55), myocardial ischemia and reperfusion (56), inflammation (57, 58), and cancer (59, 60). However, the clinical use of these LMW nitroxide radicals as medications is limited due to the preferential renal clearance, low stability in vivo conditions, as well as the adverse side effects such as mitochondrial dysfunction and antihypertensive activity (61, 62). To improve stability and targeting delivery of nitroxide radical with minimizing adverse effects, we have developed a novel redox nanoparticle (RNPO), which prepared by self-assembly of amphiphilic redox polymer, poly(ethylene glycol)-b-poly[4-(2,2,6,6-tetramethylpiperidine-1-oxyl)oxymethyl styrene] (MeO-PEG-b-PMOT) (Figure 3). Our newly designed amphiphilic redox polymer MeO-PEG-b-PMOT is composed of highly biocompatible hydrophilic poly(ethylene glycol) (PEG) shell and biodegradable hydrophobic poly(chloromethylstyrene) (PCMS), at which nitroxide radicals are easily introduced at the side chain via ether linkage (63, 64). Redox polymer MeO-PEG-b-PMOT is easily obtained via a simply two-step synthesis (Figure 3). At first, MeO-PEG-b-PCMS is synthesized by the free-radical telomerization of chloromethylstyrene monomer by using methoxy-PEG-sufanyl (MeO-PEG-SH) as a telogen. Then, MeO-PEG-b-PCMS interacts with nitroxide radical TEMPOL in the presence of sodium hydride in dry N,N-dimethylformamide (DMF), to obtain redox polymer MeO-PEG-b-PMOT. Finally, MeO-PEG-b-PMOT is dissolved in DMF, followed by dialysis against water to form core–shell-type polymeric nanoparticles (Figure 3).

52

Figure 3. Synthetic scheme of redox polymer (MeO-PEG-b-PMOT) and preparation of redox nanoparticle (RNPO). MeO-PEG-b-PCMS copolymer was synthesized by the radical telomerization of CMS using MeO-PEG-SH as a telogen in the presence of a radical initiator (AIBN). Then, redox polymer MeO-PEG-b-PMOT was synthesized by converting chloromethyl groups to TEMPOs via a Williamson ether synthesis of benzyl chloride in the MeO-PEG-b-PCMS block copolymer with the alkoxide of TEMPOL. Redox nanoparticles, RNPO was prepared by dialysis method, whereas MeO-PEG-b-PMOT was dissolved in DMF and dialyzed against water. (see color insert)

The size of RNPO is approximately 40 nm in diameter, determined by dynamic light scattering (DLS), and it shows unique electron spin resonance (ESR) spectra. In contrast with the TEMPOL signal, a clear triplet ESR signal corresponding to an interaction between 14N nuclei and the unpaired electron in the dilute solution, the ESR signals of RNPO become broaden after dialysis, which is attributable to the restricted mobility and exchange interaction of the nitroxide radicals in the solid hydrophobic core of RNPO (Figure 4) (65). This is one of the proofs of the core–shell-type structure of RNPO. High colloidal stability owing to the PEG shell layer maintains the micelle form under physiological environments without their aggregation. While LMW nitroxide radicals retained in blood only within several minutes after intravenous injection, the blood circulation of RNPO prolonged for a day, indicating the long-term stability of RNPO in bloodstream (66). The stable character also improves the accumulation tendency and ROS scavenging activity of RNPO in GI tract via oral administration route (see below) (64).

53

Figure 4. ESR spectra of LMW TEMPOL, redox polymer (MeO-PEG-b-PMOT), redox nanoparticle (RNPO), and its size. TEMPOL and RNPO were dissolved in MilliQ water, while MeO-PEG-b-PMOT was dissolved in DMF, and the ESR signal intensities were measured by an X-band ESR spectrometer at room temperature. The size of RNPO was measured by DLS using a Zetasizer Nano ZS. Reproduced with permission from ref. (70). Copyright 2015 Elsevier B.V. (see color insert)

The Safety of Redox Nanoparticles RNPO LMW nitroxide radicals are also known to show a dose-related antihypertensive action accompanied by reflex tachycardia, increased skin temperature, and seizures (62, 67). The entrapment of nitroxide radicals in the hydrophobic core of polymeric micelles via covalent conjugation is one of the important strategies for suppressing the toxicity and adverse effects of nitroxide radicals. Many types of physically drug-loaded nanoparticles have been reported and studied in the field of drug delivery systems because drug delivery by nanoparticles such as polymeric micelles and liposomes can alter the pharmacokinetics of drugs. However, as the leakage of drug from the nanoparticle occurs, it leads to adverse side effects. Since nitroxide radicals in the hydrophobic core are conjugated to polymer backbone via covalent linkage, there is no leakage of nitroxide radicals in vivo and the toxicity of the nitroxide radical is suppressed. We have so far investigated the toxicity of RNPO in vitro and in vivo using several models such as zebrafish larvae (68), chicken eggs (69), and mice (70).

Evaluation of the Toxicity of RNPO in Zebrafish The adequate assessment of toxicity and safety of RNPO must be addressed for pre-clinical and clinical applications. For the past decades, the zebrafish (Danio rerio) has been widely utilized as a correlative and predictive model for evaluation of nanoparticle toxicity (71, 72). Many studies have been reported that the bio-applicable nanomaterials including metal, silica, and polymeric 54

nanoparticles induced oxidative damages to zebrafish embryos (73–75). In a previous study, we evaluated the toxicity of RNPO on development and survival rate of zebrafish embryos as compared to LMW TEMPOL (68). We found that LMW antioxidant TEMPOL exhibited a severe toxicity to reduce survival rate of zebrafish embryos even at low concentration treatments (1 and 3 mM). Furthermore, all of the zebrafish embryos were dead just after 12 h incubation with high concentrations of LMW TEMPOL (10 and 30 mM) (Figure 5A and B). In contrast, only few dead embryos were observed even after 5 d incubation with RNPO at the same high concentrations (up to 20 mg/mL polymer concentration) (68). In a previous report, cationic polymeric nanoparticles exhibited high toxicity to zebrafish embryos compared to neutral nanoparticles (76). For example, highly cationic poly(ethylenimine) caused an abnormal development and mortality of zebrafish embryos at concentration as low as 0.01 mg/mL, while neutral poly(N-2-hydroxypropyl) methacrylamide started to affect the development of zebrafish embryos at concentrations of 1 mg/mL. We investigated that the number of healthy mitochondria in the zebrafish larvae treated with LMW TEMPOL significantly decreased compared to that of untreated zebrafish larvae. It is interesting to note that RNPO-treated larvae did not display a significantly reduced number of healthy mitochondria compared to untreated zebrafish larvae (Figure 5C). In fact, no remarkably abnormal morphological changes were observed in zebrafish embryos treated with RNPO even at a high polymer concentration (10 mg/mL), indicating an extremely low toxicity of RNPO to zebrafish embryos.

Figure 5. Evaluation of the toxicity of RNPO on zebrafish embryo. After RNPO and TEMPOL exposure, the survival of larvae was measured under light microscope at 12-h intervals throughout the 5 d of exposure time (n = 30). (A) 3 mM; (B) 10 mM (RNPO: green lines, and TEMPOL: blue lines). The data were presented as Kaplan-Meier plots and analyzed with a log-rank test (from 3 independent experiments). (C) Zebrafish larve mitochondria were stained with Mitotracker and analyzed using a fluorescent confocal microscope system. Scale bars: 100 μm. The data were expressed mean ± standard deviation (n = 6) and statistical significance was examined using 1-way analysis of variance, followed by Tukey’s post hoc test (* p < 0.05). Reproduced with permission from ref. (68). Copyright 2016, American Chemical Society. (see color insert) 55

Evaluation of the Toxicity of RNPO in Mice Since RNPO is used as an oral drug delivery, it is important to confirm the toxicity of RNPO in GI tract and other organs. Its toxicity was also confirmed in mice model. In order to investigate the toxicity of RNPO in the GI tract, healthy mice were treated long-term with free drinking RNPO (5 mg/mL polymer concentration) for one month (70). There were no remarkable differences in the hematological analysis of RNPO-treated mice as compared to healthy mice. Additionally, there were no noticeable toxicities in tissues from the GI tract and other organs, even in mice treated with a high concentration of RNPO (5 mg/mL) for a month. These results indicate that, during long-term oral administrations, RNPO causes almost no toxicity to mice although it highly accumulates in the intestinal mucosa (see below). It is also confirmed that oral administration of RNPO does not show the significant changes in intestinal microflora in healthy mice, indicating the innocuousness of this oral nanotherapeutics against healthy intestinal microflora (77). Low toxicity of RNPO can be explained by several reasons. (1) It was investigated in vitro and in vivo that the uptake of RNPO in healthy cells is much lower than in inflamed and cancer cells (77–79). Due to nitroxide radicals confined in solid core of RNPO via covalent bonds, it prevents the leakage of nitroxide radicals out of nanoparticles. (2) Although orally administered RNPO presents high accumulation in intestinal mucosa, it does not internalize into bloodstream (64), avoiding possible adverse side effects in other organs. (3) Since it has been reported that ingested nanoparticles induce oxidative damage and inflammation to epithelial cells by overproduction of ROS (80, 81), the antioxidant character of nitroxide radicals might also contribute to the reduced toxicity of RNPO. Taken together, covalent conjugations of nitroxide radicals into the core of RNPO prevent non-specific distribution, avoiding their uptake into healthy cells, resulting in an extremely low toxicity of this nanotherapeutics.

The Specific Accumulation of Orally Administered RNPO in Colon The accumulation of nanoparticles in the colon area is one of the most important features for an effective nanomedicine against IBD. Because we introduced nitroxide radicals into the particles, their accumulation could be quantitatively monitored by ESR measurements. In order to quantify the accumulation of nanoparticles in the colon area, we compared RNPO with different sizes of commercial available polystyrene latex particles and LMW TEMPOL. When LMW TEMPOL was orally administered to mice, almost no ESR signal was observed in the colon. TEMPO-installed polystyrene latex particles with different sizes (40 nm, 100 nm, 0,5 µm, and 1 µm) were confirmed to accumulate size-dependently in colon (64). For example, polystyrene latex particles with 40 nm showed highest accumulation among particles investigated in this study, which is consistent with previous reports (82, 83). Interestingly, when RNPO was administered orally to mice, a considerable high accumulation of RNPO in colon was observed, as compared to polystyrene latex particles with the same size (40 56

nm) (Figure 6A). The area under the concentration-time curve (AUC) of RNPO was 50 times higher that that of LWM TEMPOL (64). In the harsh environment of the GI tract, which includes gastric juices with strong acid, digestive enzymes, and bile acid, the polystyrene nanoparticles do not always stably maintain their sizes. High colloidal stability of RNPO to inhibit nanoparticle aggregation in hash environments in GI tract due to the PEG tethered chains on the surface of RNPO might be effective to accumulate in colonic mucosa as compared to polystyrene latex particles. In addition, the hydrophilic and uncharged property of PEG shell prevents the interaction of nanoparticles with intestinal environment and mucus constituents (84, 85). Contrary to the rapid absorption of LMW TEMPOL into the bloodstream through the GI tract, no blood uptake of orally administered RNPO was observed, suggesting a lack of systemic side effects of this oral nanotherapy (64). The extremely high accumulation of RNPO in colonic mucosa without internalization into bloodstream is one of the most important points for high performance efficiency as a colitis therapy (see below).

Figure 6. The accumulation of RNPO in colon and its specific internalization in inflamed tissue. (A) Accumulation of LMW TEMPOL, RNPO and polystyrene latex particles (with different sizes from 40 nm to 1000 nm) in the colon. After oral administration of LMW TEMPOL, RNPO or polystyrene latex particles with equivalent nitroxide radicals (1.33 mg; 7.5 μM), the amount of nitroxide radicals were measured by ESR. The data are expressed as mean ± SEM, n = 3. (B) The accumulations of rhodamine-labeled RNPO in a normal colon and a DSS-induced inflamed colon 4 h after oral administration were observed under fluorescent microscope (n = 3–4; L: lumen and Sm: submucosa), and the ESR spectra of RNPO. Reproduced with permission from refs. (64) and (78). Copyright 2012 and 2015 Elsevier B.V. (see color insert) Specific cellular internalizations of RNPO in inflamed and cancer colon tissues were analyzed in vivo using mice with DSS-induced colitis and azoxymethane (AOM)/DSS-induced CAC, and compared to healthy mice. After oral gavage of RNPO, colon tissues were collected from these mice and isolated cells were oxidized for ESR assays. It is interesting to note that the total ESR intensity of RNPO is significantly higher in inflammatory and cancer tissues compared to normal tissues (64, 70, 78). Alternatively, RNPO remarkably surrounded mucosa 57

of inflammatory and tumor tissues due to the defective structure of mucus layer in these sites (86), and exhibition of abnormal tight junction (87), resulting in the facile penetration of the nanoparticles to mucosa in inflammatory and cancer tissues. This result demonstrates that RNPO tends to accumulate in inflammatory and cancer cells, where large amounts of ROS and pro-inflammatory cytokines are produced. Interestingly, a triplet peak on ESR spectrum was observed inside inflammatory and cancer cells (Figure 6B), indicating exposure of the nitroxide radicals after disintegration of RNPO inside these cells (70, 78). In contrast, no internalization of RNPO was observed in normal cells (64, 70). The intracellular internalization mechanism of RNPO in inflamed and cancer cells was confirmed in vitro, suggesting the uptake of RNPO via both endocytosis pathway and simply diffusion due to the leaky cellular membranes of these damaged cells (70, 78). Higher accumulation in colonic mucosa, specific internalization in cancer cells, and low uptake in normal cells are the most important characteristics of RNPO, which are anticipated for high therapeutic efficiency with extremely low adverse effects.

Therapeutic Efficacy of Redox Nanoparticles in GI Tract Diseases GI Tract Inflammation DSS-Induced Colitis Mice Since orally administered RNPO highly accumulated in the colonic mucosa of DSS-injured mice and was not absorbed into the bloodstream, it is anticipated as an ideal nanomedicine for UC treatment. DSS-induced colitis mice model was used to investigate the therapeutic efficacy of RNPO, which was orally administered daily to DSS-injured mice for 7 days (64). Additional DSS-injured mice were treated with LMW TEMPOL, commercial anti-ulcer mesalamine and micelle without nitroxide radicals as controls. RNPO-treated mice showed much lower disease activity index and preserved colon length compared to DSS-treated mice and other treated groups (Figure 7A) (64). Additionally, histological analyses showed that mucosal structures of DSS- and micelle-treated mice were significantly damaged, viz., destruction of crypts and high levels of neutrophil invasion were observed in these mice. LMW TEMPOL- and mesalamine-treated mice showed moderately damaged mucosal structures. Contrary to those treatments, RNPO-treated mice showed almost similar to that of control mice, indicating the significant therapeutic effect of RNPO on DSS-induced colitis in mice (64). It should be noted that as compared to mice treated with LWM TEMPOL, RNPO-treated mice showed a significant suppression of ROS level and pro-inflammatory mediators including myeloperoxidase and IL-1β in colonic tissue of DSS-injured mice (Figure 7B) (64). These results indicate that high accumulation and ROS scavenging capacity in inflammatory sites, orally administered RNPO more effectively suppressed the inflammation and mucosal damage than LMW drugs treated. 58

Figure 7. The therapeutic effect of RNPO on DSS-induced colitis. (A) Changes in disease activity index. Disease activity index is the summation of the stool consistency index (0–3), fecal bleeding index (0–3), and weight loss index (0–4). The data are expressed as mean ± SEM, * p < 0.05, **p < 0.01 and ***p < 0.001 vs. control group; ‡p < 0.05 and ¶p < 0.001 vs. DSS groups, n = 6–7, two-way ANOVA, followed by Bonferroni post-hoc test. (B) Generation of superoxide in colon homogenates was measured by dihydroethidium fluorescence. The data are expressed as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, n = 6. (C) The survival rate of mice was determined after 15 days of 3% (wt/vol) DSS treatment. Starting on day 5, test drugs were orally administered daily until day 15. The number of surviving mice was counted until day 15, n = 6. Reproduced with permission from ref. (64). Copyright 2012 Elsevier B.V.

The effect of orally administered RNPO was investigated on the survival rate of mice with colitis. After 15 days of treatment, orally administered LMW TEMPOL and mesalamine slightly increased the survival rate (33.3% and 50%, respectively) compared with DSS- and micelle-treated mice (16.7%) (Figure 7C). On the other hand, RNPO treatment significantly increased the survival rate of DSS-treated mice to 83.3% (64). In addition, by using different doses from 50 mg/kg to 300 mg/kg, we further confirmed that oral administration of RNPO induced a robust dose-response efficacy in DSS-induced colitis mice (Figure 8) (78). At low treatment dose (50 mg/kg), the efficacy RNPO was almost similar to positive treatment of commercially available anti-inflammatory drug, mesalamine. In contrast, we observed significant improvement of treatment efficiency when higher doses of RNPO (100 mg/kg and 300 mg/kg) were used (Figure 8), suggesting a pharmacological evidence of RNPO in preclinical model. We also investigated the effect of RNPO on intestinal bacteria in DSS-induced colitis mice. By scavenging overproduced ROS and suppress intestinal inflammation to protect the intestinal mucosa, oral administration of RNPO inhibited the increase in pathogenic bacteria in the inflamed mucosa, and maintained the balance of intestinal bacteria (77). 59

Figure 8. Dose-dependent therapeutic effect of RNPO on DSS-induced colitis. RNPO was administered daily via oral gavage at low (50 mg/kg), medium (100 mg/kg), and high (300 mg/kg) doses, denoted RNPO(50), RNPO(100), and RNPO(300), respectively. The positive control 5-ASA was suspended in 0.5% (wt/vol) carboxymethyl cellulose and administered via oral gavage at a dose of 35 mg/kg, which is a dose equivalent of 200 mg/kg RNPO (35 mg/kg of nitroxide radicals). (A) Histology analysis was assessed using 7-μm-thick colon sections staining with hematoxylin and eosin (H&E). Scale bars, 500 μm. (B) Pro-inflammatory cytokine (IL-1β) were measured using enzyme-linked immunosorbent assay kits. Protein content was determined using a bicinchoninic acid kit. The data are expressed as mean ± SEM, *p < 0.05, **p < 0.01, n = 6. Reproduced with permission from ref. (78). Copyright 2015 Elsevier B.V. (see color insert)

Non-Steroidal Anti-Inflammatory Drugs (NSAIDs)-Induced Small Intestinal Inflammation It has been reported that overproduction of ROS in the intestine of patients receiving repeated doses of NSAIDs such as aspirin and indomethacin (Indo) (88–90). The efficacy of orally administered of RNPO was also investigated in NSAIDs-induced intestinal injury model in mice (91). The accumulation of RNPO in small intestine was measured using ESR, and confirmed that the AUC of orally administered RNPO in small intestine was 40 times higher than that of LMW TEMPOL. Consequently, as compared to LMW TEMPOL treatment, orally administered RNPO significantly suppressed ROS and lipid peroxidation levels in Indo-treated mice, resulting in an increase in the survival rate of mice treated daily with Indo. Another strategy to improve the low bioavailability of Indo is encapsulation of Indo into hydrophobic core of RNPO (92). By this entrapment, the enhancement of Indo uptake into bloodstream and suppression 60

of small intestinal inflammation could be achieved. It should be noticed that Indo-encapsulated control nanoparticle, which has no ROS scavenging capacity, caused severe intestinal inflammation (92). Based on these results, oral administration of RNPO is a promising nanotherapeutics and drug delivery nanocarrier for treatment of inflammation not only in intestine and colon but also whole GI tract.

Colitis-Associated Colon Cancer It has been reported that oxidative stress and inflammation pathways are critically associated to cancer development (93). The inhibition of overproduced ROS and inflammation in the tumor microenvironments effectively works to suppress the tumor progression and the resistance against chemotherapy (94, 95). Since oral administration of RNPO effectively scavenged ROS and significantly suppressed the intestinal and colonic inflammation, it is anticipated to exhibit a therapeutic potential in treating cancer in GI tract. In this study, we used AOM/DSS to chemically induced CAC in mice, and confirmed the efficacy of orally administered RNPO as a nanomedicine and combination therapy with irinotecan (70). As anticipated, we found that oral administration of RNPO along with DSS treatment clearly suppressed inflammation in the colon, significantly preventing carcinoma progression in the CAC mouse model. Because DSS is chemical agent to induce inflammation to promote cancer development, simultaneous administration of RNPO with DSS protected again the generation of inflammation, resulting in suppression of carcinoma propagation. It should be rather noted that administration of RNPO after DSS treatment also effectively suppressed tumor progression in mice given free drinking RNPO (70). Because RNPO clearly suppressed inflammation around tumor microenvironment, a combination treatment with RNPO and conventional cancer drugs is a robust strategy. In this study, we investigated the combination of oral RNPO with irinotecan and found that the anticancer efficacy was significantly enhanced with the combination compared to treatment with irinotecan alone (Figure 9A and B) (70). It is also interesting to notice that co-treatment with RNPO significantly suppressed the adverse effects in GI tract such as body weight loss, diarrhea, and intestinal inflammation caused by irinotecan treatment, indicating that a synergistic effect was successfully achieved (Figure 9C and D) (70). In another study, combination of RNPO with intravenously injected doxorubicine also significantly suppressed tumor development in CAC mice as compared to mice treated with doxorubicin alone (96). In addition, administration of RNPO effectively ameliorated doxorubicin-induced cardiac, hepatic, and intestinal adverse effects in mice by scavenging overproduced ROS. On the basis of these results, this redox nanotherapeutics, RNPO, exhibits great potential and clinically applicable property for treating IBD patients, colon cancer, as well as other ROS-related diseases. 61

Figure 9. Oral administration of RNPO enhances the anticancer effect of irinotecan (Iri) and reduces its side effect on the GI tract. (A) The scheme of AOM/DSS-induced CAC and administration of Iri and RNPO. RNPO (2.5 mg/mL) was given to mice in drinking water, while of Iri (5 mg/kg) were given daily by oral gavage 5 times per week for 4 weeks. (B) Combination effect of Iri and RNPO against CAC development was evaluated by assessment of tumor scores. The data are expressed as mean ± SEM, *p < 0.05, n = 6 mice. (C) and (D) The effect of co-treatment with RNPO to reduce Iri-induced GI toxicity. (C) The diarrhea score and (D) histological assessment of small intestine and colon sections by H&E staining. The data are expressed as mean ± SEM, *p < 0.05 and ***p < 0.001, n = 5 mice. Representative sections are shown for n = 3 mice. Scale bars = 100 μm. Reproduced with permission from ref. (70). Copyright 2015 Elsevier B.V. (see color insert)

Summary Oral drug delivery using nanoparticles seem to be a highly promising in treatment of GI tract disorders. Nanomedicines have recently been developed to carry active biomolecules to overcome the barriers and to target to diseased sites of GI tract. In this article, we described the development and application of our novel redox nanoparticle RNPO for GI tract disorders via oral administration route. By confinement of nitroxide radicals in the hydrophobic core of RNPO via covalent linkage, we could inhibit the leakage of nitroxide radicals from nanoparticles, minimizing non-specific distribution and adverse side effects of nitroxide radicals. High accumulation of orally administered RNPO in GI tract and specific internalization in inflammatory and cancer tissues are important characters to effectively scavenge overproduced ROS to suppress the intestinal 62

inflammation and cancer progression. Importantly, this redox nanotherapeutics exhibits extremely low toxicity as compared to LWM nitroxide radicals. Taken together, ROS scavenging nanoparticle RNPO is anticipated as a highly promising nanotherapeutics in treating not only GI tract disorders but also other oxidative stress-related diseases.

Acknowledgments A part of this work was supported by Grant-in-Aid for Scientific Research A (21240050) and Grant-in-Aid for Research Activity Start-up (22800004), Young Scientist B (16K16397), and the World Premier International Research Center Initiative on Materials Nanoarchitronics of the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Conflicts of interest The authors declare that they have no competing financial interests.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

14. 15. 16. 17. 18. 19. 20.

Murphy, M. P. Biochem. J. 2009, 417, 11–13. Ray, P. D.; Huang, B. W.; Tsuji, Y. Cell Signal. 2012, 24, 981–990. D'Autréaux, B.; Toledano, M. B. Nat. Rev. Mol. Cell Biol. 2007, 8, 813–824. Bartosz, G. Biochem. Pharmacol. 2009, 77, 1303–1315. Cooke, M. S.; Evans, M. D.; Dizdaroglu, M.; Lunec, J. FASEB J. 2003, 17, 1195–1214. Barnham, K. J.; Masters, C. L.; Bush, A. I. Nat. Rev. Drug Discovery 2004, 3, 205–214. Brieger, K.; Schiavone, S.; Miller, F. J., Jr.; Krause, K. H. Swiss Med. Weekly 2012, 142, w13659. Klaunig, J. E.; Kamendulis, L. M. Annu. Rev. Pharmacol. Toxicol. 2004, 44, 239–267. Kaplan, G. G. Nat. Rev. Gastroenterol. Hepatol. 2015, 12, 720–727. Loftus, E. V., Jr. Gastroenterology 2004, 126, 1504–1517. Friend, D. R.; Sellin, J. Adv. Drug Delivery Rev. 2005, 57, 215–216. Stephen, B. H. Gastroenterology 2004, 126, 1582–1592. Singh, K.; Chaturvedi, R.; Barry, D. P.; Coburn, L. A.; Asim, M.; Lewis, N. D.; Piazuelo, M. B.; Washington, M. K.; Vitek, M. P.; Wilson, K. T. J. Biol. Chem. 2011, 286, 3839–3850. Khor, B.; Gardet, A.; Xavier, R. J. Nature 2011, 474, 307–317. Xavier, R. J.; Podolsky, D. K. Nature 2007, 448, 427–434. Zhu, H.; Li, Y. R. Exp. Biol. Med. 2012, 237, 474–480. Simmonds, N. J.; Rampton, D. S. Gut 1993, 34, 865–868. Fournier, B. M.; Parkos, C. A. Mucosal Immunol. 2012, 5, 354–366. Babbs, C. F. Free Radic. Biol. Med. 1992, 13, 169–182. McCord, J. M. Am. J. Med. 2000, 108, 652–659. 63

21. Esworthy, R. S.; Aranda, R.; Martín, M. G.; Doroshow, J. H.; Binder, S. W.; Chu, F. F. Am. J. Physiol. Gastrointest. Liver Physiol. 2001, 281, G848–G855. 22. Moura, F. A.; de Andrade, K. Q.; dos Santos, J. C.; Araújo, O. R.; Goulart, M. O. Redox Biol. 2015, 6, 617–639. 23. Ju, J.; Hao, X.; Lee, M. J.; Lambert, J. D.; Lu, G.; Xiao, H.; Newmark, H. L.; Yang, C. S. Cancer Prev. Res. 2009, 2, 143–152. 24. You, Y.; Fu, J. J.; Meng, J.; Huang, G. D.; Liu, Y. H. Dig. Dis. Sci. 2009, 54, 1643–1650. 25. Oz, H. S.; Chen, T. S.; McClain, C. J.; de Villiers, W. J. J. Nutr. Biochem. 2005, 16, 297–304. 26. Evans, D. F.; Pye, G.; Bramley, R.; Clark, A. G.; Dyson, T. J.; Hardcastle, J. D. Gut 1988, 29, 1035–1041. 27. Press, A. G.; Hauptmann, I. A.; Hauptmann, L.; Fuchs, B.; Fuchs, M.; Ewe, K.; Ramadori, G. Aliment. Pharmacol. Ther. 1998, 12, 673–678. 28. Cook, M. T.; Tzortzis, G.; Charalampopoulos, D.; Khutoryanskiy, V. V. J. Controlled Release 2012, 162, 56–67. 29. Corfield, A. P.; Carroll, D.; Myerscough, N.; Probert, C. S. Front. Biosci. 2001, 6, D1321–1357. 30. Atuma, C.; Strugala, V.; Allen, A.; Holm, L. Am. J. Physiol. Gastrointest. Liver Physiol. 2001, 280, G922–G929. 31. Johansson, M. E.; Gustafsson, J. K.; Sjöberg, K. E.; Petersson, J.; Holm, L.; Sjövall, H.; Hansson, G. C. PLoS One 2010, 5, e12238. 32. Zhou, F. X.; Chen, L.; Liu, X. W.; Ouyang, C. H.; Wu, X. P.; Wang, X. H.; Wang, C. L.; Lu, F. G. World J. Gastroenterol. 2012, 18, 2344–2356. 33. Mowat, A. M.; Agace, W. W. Nat. Rev. Immunol. 2014, 14, 667–685. 34. Sartor, R. B. Res. Immunol. 1997, 148, 567–576. 35. Strauch, U. G.; Obermeier, F.; Grunwald, N.; Gürster, S.; Dunger, N.; Schultz, M.; Griese, D. P.; Mähler, M.; Schölmerich, J.; Rath, H. C. Gut 2005, 54, 1546–1552. 36. Takaishi, H.; Matsuki, T.; Nakazawa, A.; Takada, T.; Kado, S.; Asahara, T.; Kamada, N.; Sakuraba, A.; Yajima, T.; Higuchi, H.; Inoue, N.; Ogata, H.; Iwao, Y.; Nomoto, K.; Tanaka, R.; Hibi, T. Int. J. Med. Microbiol. 2008, 298, 463–472. 37. Abraham, C.; Medzhitov, R. Gastroenterology 2011, 140, 1729–1737. 38. Belkaid, Y.; Hand, T. W. Cell 2014, 157, 121–141. 39. Mowat, A. M.; Agace, W. W. Nat. Rev. Immunol. 2014, 14, 667–685. 40. Krajmalnik-Brown, R.; Ilhan, Z. E.; Kang, D. W.; DiBaise, J. K. Nutr. Clin. Pract. 2012, 27, 201–214. 41. Kim, B.; Rutka, J.; Chan, W. N. Eng., J. Med. 2010, 36, 2434–2443. 42. Otsuka, H.; Nagasaki, Y.; Kataoka, K. Adv. Drug Delivery Rev. 2003, 55, 403–419. 43. Maeda, H.; Fang, J.; Inutsuka, T.; Kitamoto, Y. Int. Immunopharmacol. 2003, 3, 319–328. 44. Weis, S. M.; Cheresh, D. A. Nat. Med. 2011, 17, 1359–1370. 45. Davis, M. E.; Chen, Z. G.; Shin, D. M. Nat. Rev. Drug Discovery 2008, 7, 771–782. 64

46. Xiao, B.; Merlin, D. Expert Opin. Drug Delivery 2012, 9, 1393–1407. 47. Collnot, E. M.; Ali, H.; Lehr, C. M. J. Controlled Release 2012, 161, 235–246. 48. Wilson, D. S.; Dalmasso, G.; Wang, L.; Sitaraman, S. V.; Merlin, D.; Murthy, N. Nat. Mater. 2010, 9, 923–928. 49. Laroui, H.; Dalmasso, G.; Nguyen, H. T.; Yan, Y.; Sitaraman, S. V.; Merlin, D. Gastroenterology 2010, 138, 843–853. 50. Pertuit, D.; Moulari, B.; Betz, T.; Nadaradjane, A.; Neumann, D.; Ismaïli, L.; Refouvelet, B.; Pellequer, Y.; Lamprecht, A. J. Controlled Release 2007, 123, 211–218. 51. Matsumoto, K.; Fuminori, H.; Matsumoto, A.; Koretsky, A. P.; Sowers, A. L.; Mitchell, J.; Krishna, C. M. Clin. Cancer Res. 2006, 12, 2455–2462. 52. Adam, W.; Saha-Möller, C. R.; Ganeshpure, P. A. Chem. Rev. 2001, 101, 3499–3548. 53. Soule, B. P.; Hyodo, F.; Matsumoto, K.; Simone, N. L.; Cook, J. A.; Krishna, M. C.; Mitchell, J. B. Free Radic. Biol. Med. 2007, 42, 1632–1650. 54. Miura, Y.; Utsumi, H.; Hamada, A. Arch. Biochem. Biophys. 1993, 300, 148–156. 55. Hahn, S. M.; Tochner, Z.; Krishna, C. M.; Glass, J.; Wilson, L.; Samuni, A.; Sprague, M.; Venzon, D.; Glatstein, E.; Mitchell, J. B. Cancer Res. 1992, 52, 1750–1753. 56. McDonald, M. C.; Zacharowski, K.; Bowes, J.; Cuzzocrea, S.; Thiemermann, C. Free Radic. Biol. Med. 1999, 27, 493–503. 57. Cuzzocrea, S.; McDonald, M. C.; Mota-Filipe, H.; Mazzon, E.; Costantino, G.; Britti, D.; Mazzullo, G.; Caputi, A. P.; Thiemermann, C. Arthritis Rheum. 2000, 43, 320–328. 58. Cuzzocrea, S.; McDonald, M. C; Mazzon, E.; Dugo, L.; Lepore, V.; Fonti, M. T.; Ciccolo, A.; Terranova, M. L.; Caputi, A. P.; Thiemermann, C. Eur. J. Pharmacol. 2000, 406, 127–137. 59. Mitchell, J. B.; Xavier, S.; DeLuca, A. M.; Sowers, A. L.; Cook, J. A.; Krishna, M. C.; Hahn, S. M.; Russo, A. Free Radic. Biol. Med. 2003, 34, 93–102. 60. Soule, B. P.; Hyodo, F.; Matsumoto, K.; Simone, N. L.; Cook, J. A.; Krishna, M. C.; Mitchell, J. B. Antioxid. Redox Signal. 2007, 9, 1731–1743. 61. Monti, E.; Supino, R.; Colleoni, M.; Costa, B.; Ravizza, R.; Gariboldi, M. B. J. Cell. Biochem. 2001, 82, 271–276. 62. Wilcox, C. S.; Pearlman, A. Pharmacol. Rev. 2008, 60, 418–469. 63. Yoshitomi, T.; Hirayama, A.; Nagasaki, Y. Biomaterials 2011, 32, 8021–8028. 64. Vong, L. B.; Tomita, T.; Yoshitomi, T.; Matsui, H.; Nagasaki, Y. Gastroenterology 2012, 143, 1027–1036.e3. 65. Nagasaki, Y. Ther. Delivery 2012, 3, 165–179. 66. Chonpathompikunlert, P.; Fan, C. H.; Ozaki, Y.; Yoshitomi, T.; Yeh, C. K.; Nagasaki, Y. Nanomedicine 2012, 7, 1029–1043. 67. Hahn, S. M.; Sullivan, F. J.; DeLuca, A. M.; Bacher, J. D.; Liebmann, J.; Krishna, M. C.; Coffin, D.; Mitchell, J. B. Free Radic. Biol. Med. 1999, 27, 529–535. 65

68. Vong, L. B.; Kobayashi, M.; Nagasaki, Y. Mol. Pharm. 2016, 13, 3091–3097. 69. Abe, C.; Uto, Y.; Kawasaki, A.; Noguchi, C.; Tanaka, R.; Yoshitomi, T.; Nagasaki, Y.; Endo, Y.; Hori, H. J. Controlled Release 2014, 182, 67–72. 70. Vong, L. B.; Yoshitomi, T.; Matsui, H.; Nagasaki, Y. Biomaterials 2015, 55, 54–63. 71. Fako, V. E.; Furgeson, D. Y. Adv. Drug Delivery Rev. 2009, 61, 478–486. 72. Zhu, X.; Zhu, L.; Duan, Z.; Qi, R.; Li, Y.; Lang, Y. J. Environ. Sci. Health, Part A: Toxic/Hazard. Subst. Environ. Eng. 2008, 43, 278–84. 73. van Aerle, R.; Lange, A.; Moorhouse, A.; Paszkiewicz, K.; Ball, K.; Johnston, B. D.; de-Bastos, E.; Booth, T.; Tyler, C. R.; Santos, E. M. Environ. Sci. Technol. 2013, 47, 8005–8014. 74. Asharani, P. V.; Lian Wu, Y.; Gong, Z.; Valiyaveettil, S. Nanotechnology 2008, 19, 255102. 75. Zhao, X.; Wang, S.; Wu, Y.; You, H.; Lv, L. Aquat. Toxicol. 2013, 136-137, 49–59. 76. Rizzo, L. Y.; Golombek, S. K.; Mertens, M. E.; Pan, Y.; Laaf, D.; Broda, J.; Jayapaul, J.; Möckel, D.; Subr, V.; Hennink, W. E.; Storm, G.; Simon, U.; Jahnen-Dechent, W.; Kiessling, F.; Lammers, T. J. Mater. Chem. B 2013, 1, 3918–3925. 77. Vong, L. B.; Yoshitomi, T.; Morikawa, K.; Saito, S.; Matsui, H.; Nagasaki, Y. J. Gastroenterol. 2014, 49, 806–813. 78. Vong, L. B.; Mo, J.; Abrahamsson, B.; Nagasaki, Y. J. Controlled Release 2015, 210, 19–25. 79. Shimizu, M.; Yoshitomi, T.; Nagasaki, Y. J. Clin. Biochem. Nutr. 2014, 54, 166–173. 80. Khanna, P.; Ong, C.; Bay, B. H.; Baeg, G. H. Nanotoxicity: An Interplay of Oxidative Stress, Inflammation and Cell Death. Nanomaterials 2015, 5, 1163–1180. 81. Bergin, I. L.; Witzmann, F. A. Nanoparticle toxicity by the gastrointestinal route: evidence and knowledge gaps. Int. J. Biomed. Nanosci. Nanotechnol. 2013, 3 (1-2). 82. Lamprecht, A.; Schafer, U.; Lehr, C. M. Pharm. Res. 2001, 18, 788–793. 83. Francis, M. F.; Cristea, M.; Winnik, F. M. Pure Appl. Chem. 2004, 76, 1321–1335. 84. Cu, Y.; Saltzman, W. M. Mol. Pharm. 2008, 6, 173–181. 85. Lai, S. K.; Wang, Y. Y.; Hanes, J. Adv. Drug Delivery Rev. 2009, 61, 158–171. 86. Johansson, M. E.; Sjovall, H.; Hansson, G. C. Nat. Rev. Gastroenterol. Hepatol. 2013, 10, 352–361. 87. Morin, P. J. Cancer Res. 2005, 65, 9603–9606. 88. Bjarnason, I.; Hayllar, J.; MacPherson, A. J.; Russell, A. S. Gastroenterology 1993, 104, 1832e47. 89. Matsui, H.; Shimokawa, O.; Kaneko, T.; Nagano, Y.; Rai, K.; Hyodo, I. J. Clin. Biochem. Nutr. 2011, 48, 107–111. 90. Higuchi, K.; Umegaki, E.; Watanabe, T.; Yoda, Y.; Morita, E.; Murano, M.; Tokioka, S.; Arakawa, T. J. Gastroenterol. 2009, 44, 879–888. 66

91. Sha, S.; Vong, L. B.; Chonpathompikunlert, P.; Yoshitomi, T.; Matsui, H.; Nagasaki, Y. Biomaterials 2013, 34, 8393–8400. 92. Yoshitomi, T.; Sha, S.; Vong, L. B.; Chonpathompikunlert, P.; Matsui, H.; Nagasaki, Y. Ther. Delivery 2014, 5, 29–38. 93. Reuter, S.; Gupta, S. C.; Chaturvedi, M. M.; Aggarwal, B. B Free Radic. Biol. Med. 2010, 49, 1603–1616. 94. Grivennikov, S. I.; Greten, F. R.; Karin, M. Cell 2010, 140, 883–899. 95. Pelicano, H.; Carney, D.; Huang, P. Drug Resist. Updates 2004, 7, 97–110. 96. Vong, L. B.; Nagasaki, Y. Mol. Pharm. 2016, 13, 449–455.

67

Chapter 3

Glycoglycan Mimic by Synthetic Polymers Yoshiko Miura,*,1 Tomohiro Fukuda,2 and Yu Hoshino1 1Department

of Chemical Engineering, Graduate School of Engineering, Kyushu University, 744 Motooka, Nisi-ku, Fukuoka 819-0395, Japan 2National Institute of Technology, Toyama College, 13 Hongo-machi, Toyama city, Toyama 939-8630, Japan *E-mail: [email protected]

Glycosaminoglycans (GAGs) are important polysaccharides in the living system, but the availability of GAGs is limited due to the complicated structures. In this investigation, GAGs were re-organized by synthetic glycopolymers. GAGs mimic glycopolymers were synthesized by polymerization of acrylamide sugar derivatives carrying sulfated N-acetyl glucosamine (GlcNAc). The interaction with proteins were investigated in view of inhibition of Alzheimer disease. The glycopolymer were prepared as molecular library with changing molecular weight, sugar structure and sugar ratios. The inhibitory activity of Alzheimer disease was studied by inhibitions of protein amyloidosis of amyloid beta (Aβ) peptides. The biological activities depended on the chemical structure of glycopolymers. We found the correlation of glycopolymer activities to the native GAGs.

1. Introduction The saccharides exist as glycolipids, glycoproteins and polysaccharides, and play important roles in the living system (1). Specially, polysaccharides, glycosaminoglycans (GAGs), the saccharide moiety of proteoglycans, are paid much attention due to the biological function such as cell adhesion, growth factor activation and anti-thrombogenicity (2, 3). GAGs are polysaccharides having alternating saccharides of amino-sugars (N-acetyl glucosamine (GlcNAc) or N-acetyl galgactosamine) and uronic acids (iduronic acid (IdoA) or glucuronic acid (GlcA). Most of the GAGs are highly sulfated except hyaluronic acid, and © 2017 American Chemical Society

the sulfation patters and saccharide structures are specific to each GAG. There are several famous GAGs such as heparin, heparan and chondroithin sulfate. Since GAGs have important in biology, many researchers have investigated the preparation of GAGs by organic and biological synthesis (5). Actually, total organic synthesis of GAGs has enthusiastically been studied for various synthetic chemists (4–7). The synthesis of oligosaccharides of GAGs have been attained. For example, the synthesis of disaccharide and of heparin like αIdoA(2S)αGlcNS(6S) have been reported (8). The disaccharide showed the function of GAGs, but the biological activity of anti-thrombogenicity was much weaker than the original GAGs (9). It is considered that the macromolecular structure of GAGs is important to exhibit strong biological activities. The biological synthesis of GAGs with enzymes are another solution for GAGs preparation. However, enzymes and sugar nucleic acids are expensive and difficult to control, and so the biological synthesis was also difficult. Considering the saccharide-protein interaction, the key of the interaction is multivalency. We have reported the sugar multivalent compounds with polymer, so-called glycopolymers (10). We have reported the various glycopolymers, which exhibit large biological interaction. Suda group have reported glyco-dendrimer (dendritic polymer) with GAGs oligosaccharides (11). Hsieh-Wilson group has reported the glycopolymer with GAGs oligosaccharide. Those glycopolymers with GAGs oligosaccharides showed the strong biological activities like the original GAGs (12, 13). We took a hint to re-organize GAGs with glycopolymers. GAGs structure was divided into the functional saccharide component with polymerizable sugars. The sugar monomers were polymerized by radical initiator. The glycopolymers can exhibit the strong interaction to the target proteins. The advantage of this polymer method is the facile preparation of GAGs mimic molecules (Figure 1).

2. Glycopolymer with Sulfonated GlcNAc The specific feature of GAGs is sulfated sugars in the polysaccharides. We focused on the sulfated GlcNAc in GAGs. In order to simplify the synthetic procedure, p-nitrophenyl glucosides were used for the starting material. Acrylamide phenyl 6-sulfo-GlcNAc was prepared as the monomer for GAGs mimic, because 6-sulfo-GlcNAc was reported to play important roles in the living systems (4). GAGs mimic polymers were prepared with radical initiator (14). The sugar content was changed from 10% to 60%. The biological activities of glycopolymer with sulfated GlcNAc was investigated by interaction with proteins. Since GAGs have related the protein amyloidosis, the glycopolymers with sulfo-GlcNAc was incubated with amyloid β (Aβ) peptide. Aβ peptides ((1-42) and (1-40)) are known as the pathogen of Alzheimer disease, and spontaneously formed protein amyloid. The aggregation of Aβ(1-42) spontaneously induced the aggregation which was monitored by fluorescence intensity of thioflavin T(ThT) (Figure 2). 70

Figure 1. Concept of GAGs mimic glycopolymers. Original GAGs was re-constructed with GAGs derived sugar and multivalency.

The glycopolymer was incubated with Aβ(1-42). The increase of ThT fluorescence was effectively inhibited in the presence of glycopolymers. The inhibitory effect of protein aggregation was depended on the sugar structure of glycopolymer. When Aβ was incubated with glycopolymer without sulfo-GlcNAc, the aggregation behavior of Aβ was similar to that without additives. The inhibitory effect of Aβ aggregation was depended on the sugar content. The glycopolymer with higher sugar content (100% and 65 %) didn’t show the inhibitory effect on protein aggregation. On the other hand, sulfo-GlcNAc glycopolymer with lower sugar content (20% and 10%) showed the good inhibitory effect. Glycopolymer having 6-sulfo-GlcNAc electrostatically interacted with Aβ via cationic residues such as His13. However, at the same time, the net charge of Aβ was negative. Not only glycopolymer can electrostatically interact with Aβ with the cationic residues, but the polymer induced the electron repulsion due to the negative Aβ net charge. The balance of interaction and repulsion determined the degree of inhibitory effect, and so the glycopolymer partially substituted with 6-sulfo-GlcNAc showed the strong inhibitory effect. The morphology of Aβ was measured by atomic force microscope (AFM). Aβ without additive formed nanofiber after incubation. The addition of glycopolymer with low sugar content (10-20%) changed the morphology of Aβ. It showed the interaction of glycopolymer with Aβ induced the different morphology from nanofibril. 71

Figure 2. The inhibitory effect on glycopolymer on Aβ(1-42) aggregation. (a) Time course of the fluorescence change in ThT with Aβ(1-42) and polymer additives. (b) Chemical structure of glycopolymers with 6-sulfo-GlcNAc (1) and GlcNAc (2) and polyacrylamide (3) (14). The neutralization of Aβ cytotoxicity was also investigated with Hela cell. The addition of Aβ induced the cytotoxicity, and the cell survival rate was decreased. The addition of glycopolymer with 6-sulfo-GlcNAc recovered the cell survival rate due to the neutralization of Aβ. The glycopolymer with 6-sulfo-GlcNAc itself did not induced cytotoxicity. These investigations showed that the glycopolymer with 6-sulfo-GlcNAc could be a polymer nanomedicine based on GAGs mimetics.

3. GAGs Mimic Polymers with Controlled Molecular Weight The biological ability of GAGs depends on the molecular weight due to the physical property and cell permeability (15). It has been reported that heparin with low molecular weight and high molecular weight show the different biological activities. It is difficult to prepare and investigate the synthetic GAGs with different molecular weight, but it is easy to prepare GAGs mimic polymers with different molecular weight. Controlled polymerization provides the facile method to prepare the molecules with different defined molecular weight. We tried to prepare the glycopolymer with different molecular weight using living radical polymerization with reversible addition fragmentation chain transfer (RAFT) reagent (Figure 72

3) (16), because RAFT living radical polymerization is advantageous of bulky monomer like sugar. The glycopolymers were prepared with different molecular weights, 105, 104 and 103 order (17). Sugars used were 6-sulfo-GlcNAc and GlcA from GAGs structure.

Figure 3. Chemical structure of GAGs mimic polymers with different molecular weights via RAFT living radical polymerization (17).

The function of glycopolymer with different molecular weights were investigated with Aβ(1-40). The kinetics of amyloidosis was investigated with the polymer additives using ThT fluorescence. The glycopolymer with 6-sulfo-GlcNAc showed the inhibitory effect on Aβ, but the glycopolymer with GlcA did not. Interestingly, the glycopolymer having both 6-sulfo-GlcNAc and GlcA showed the strongest inhibitory effect. The molecular weight effect of glycopolymer on Aβ protein aggregation was clear. Though the monomers didn’t not show the inhibitory effect on protein aggregation, the glycopolymers with low molecular weight showed the better activity than that with high molecular weight. Since the sugar density in the polymer was same, the multivalent effect of glycopolymer to Aβ exhibited in a similar degree. The mobility of glycopolymer in solution was dependent on the molecular weight, and the glycopolymer with low molecular weight. The kinetics of protein aggregation was studied by time course of ThT fluorescence. The glycopolymer with 6-sulfo-GlcNAc inhibited the nucleation of protein aggregation, and the glycopolymer with lower molecular weight exhibited the stronger inhibitory effect on nucleation (Figure 4). The glycopolymer 73

with GlcA showed the weak inhibitory effect on protein aggregation, but the glycopolymer with GlcA inhibited the elongation of protein fibril. The elongation inhibitory effect was related to molecular weight, and the glycopolymer with GlcA with low molecular weight showed the better inhibitory effect on elongation.

Figure 4. The schematic image of inhibition of Aβ(1-40) with GAGs mimic polymer. AFM image of Aβ(1-40) (a) without glycopolymer and (b) in the presence of glycopolymer of poly(AAm/6S-GlcNAc/GlcA) (17). Reproduced with permission from reference (17). The results suggested that the sugar structure in GAGs relates the nucleation and elongation of Aβ. The addition of glycopolymer with 6-sulfo-GlcNAc and GlcA inhibited the nucleation and elongation of Aβ aggregates, respectively. The glyco-ter-polymer with 6-sulfo-GlcNAc and GlcA showed the strong inhibitory effect on Aβ aggregation. Among them, the glyco-ter-polymer with 6-sulfo-GlcNAc and GlcA showed the strongest inhibitory effects. The natural GAGs have various sugar structures which interacted with proteins and control activities. The Aβ kinetics experiments with glycopolymer can clarify the role of each saccharide in GAGs. The glycopolymer with lower molecular weight showed the better activities in this chapter’s experiment. Though the multivalent effect of polymer is essential, but the polymer with lower molecular weight was advantageous on inhibitory effect of protein aggregation due to the better mobility.

4. Other GAGs Mimic Polymers Preparation of the GAGs mimic polymer is a unique method to investigate the GAGs with different sugar structures and molecular weights. One problem of the synthetic polymer is inhomogeneous structure based on the polydispersity. Most of the polymer has polydispersity and inhomogeneous structure. On the other hand, dendrimers have uniform structure (18). Dendrimers with 6-sulfo-GlcNAc were synthesized (19). The dendrimers were prepared by click chemistry of Huisgen reaction, which enables facile synthesis without protective groups. The glycodendrimer was synthesized by divergent method. Glycodendrimers with three different generations (G0, G1 and G2) were synthesized (Figure 5). The inhibitory effect of glycodendrimer was studied with Aβ(1-42). The 74

glycodendrimer with 6-sulfo-GlcNAc in high generation (G2) showed the good inhibitory effect like GAGs mimic linear polymers.

Figure 5. Chemical structure of GAGs mimic dendrimers of G0, G1 and G2 (19). Reproduced with permission from reference (19). Copyright (2012) MDPI. The glycodendrimers were also utilized as microarray to analyze sugar-protein interaction in detail (20). Fan-type glycodendrimers with 6-sulfo-GlcNAc were synthesized by click chemistry as shown in Figure 6. The glycodendrimers were immobilized onto gold substrate by click chemistry. he interaction with Aβ was studied by SPR and AFM. The interaction with Aβ was amplified in trimer and dimer of 6-sulfo-GlcNAc. The multivalent structure of 6-sulfo-GlcNAc was indispensable to interaction with Aβ. But at the same time, multivalent 6-sulfo-GlcNAc induced the electrostatic repulsion due to the negatively charged protein and saccharides. The interaction and electric repulsion between Aβ and 6-sulfo-GlcNAc determined the morphology of Aβ aggregates. The weak 75

interaction induced the nano fibrils, and the strong interaction and repulsion of multivalent sugar induced the spherical objects that exhibited cytotoxicity.

Figure 6. A representative image of glycodendrimer array with 6-sulfo-GlcNAc (20). The dendrimer with 6-sulfo-GlcNAc provided the defined interaction to Aβ. The investigation with glycodendrimer and glycodendrimer array was useful to understand the mechanism of amyloidosis with GAGs.

5. Conclusion We investigated the preparation and investigation of GAGs mimic polymer using sulfonated saccharides. GAGs mimic polymer was facile way to prepare GAGs’ mimic libraries. The glycopolymer with 6-sulfo-GlcNAc interacted with Aβ based on GAGs function. The GAGs mimic polymer libraries can change the various factors such as the polymer structure, sugar structure, sugar content and molecular weight. The detailed study on Aβ with GAGs mimic polymer was useful to clarify the mechanism of Aβ aggregation with GAGs.

References 1.

Taylor, M. E.; Drickamaer, K. Introduction to Glycobiology, 3rd ed.; Oxford Press: London, 2011, pp 3−16. 76

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

Rudd, T. R.; Skidomore, M. A.; Guerrini, M.; Hricovini, M.; Powell, A. K.; Siligardi, G.; Yates, E. A. Curr. Opin. Struct. Biol. 2010, 20, 567–574. Miura, Y.; Fukuda, T.; Seto, H.; Hoshino, Y. Polym. J. 2016, 48, 229–237. Tamura, J. Trends Glycosci. Glycotechnol. 2001, 13, 65–89. Koshida, S.; Suda, Y.; Sobel, M.; Ormsby, J.; Kusumoto, S. Bioorg. Med. Chem. Lett. 1999, 9, 3127–3132. Hu, Y. P.; Lin, S. Y.; Huang, C. Y.; Zulueta, M. M. L.; Liu, J. Y.; Chang, Y.; Hung, S. C. S. Nat. Chem. 2011, 3, 557–563. de Paz, J. L.; Noti, C.; Seeberger, P. H. J. Am. Chem. Soc. 2006, 128, 2766–2767. van Boeckel, C. A. A.; Peitou, M. Angew. Chem., Int. Ed. 1993, 32, 1671–1818. Suda, Y.; Marques, D.; Kermode, J. C.; Kusumoto, S.; Sobel, M. Thromb. Res. 1993, 69, 501–50. Miura, Y.; Hoshino, Y.; Seto, H. Chem Rev. 2016, 116, 1673–1692. Suda, Y.; Arano, A.; Fukui, Y.; Koshida, S.; Wakao, M.; Nishimura, T.; Kusumoto, S.; Sobel, M. Bioconjugate Chem. 2006, 17, 1125–1135. Rawat, M.; Gama, C. I.; Matson, J. B.; Hsieh-Wilson, L. C. J. Am. Chem. Soc. 2008, 130, 2959–2961. Lee, S. G.; Brown, J. M.; Rogers, C. J.; Matson, J. B.; Krishnamurthy, C.; Rawat, M.; Hsieh-Wilson, L. C. Chem Sci. 2011, 1, 322–325. Miura, Y.; Yasuda, K.; Yamamoto, K.; Koike, M.; Nishida, Y.; Kobayashi, K. Biomacromolecules 2007, 8, 2129–2134. Hirsh, J.; Raschke, R. Chest J. 2004, 126, 188S–203S. Chong, Y.; Le, T.; Moad, G.; Rizzardo, E.; Thang, S. H. A. Macromolecules 1999, 32, 2071–2074. Miura, Y.; Mizuno, H. Bull Chem Soc. Jpn 2010, 83, 1004–1009. Mignani, S.; Kazzouli, S. E.; Bousmina, M.; Majoral, J. P. Prog. Polym. Sci. 2013, 38, 993–1008. Miura, Y.; Onogi, S.; Fukuda, T. Molecules 2012, 17, 11877–11896. Fukuda, T.; Matsumoto, E.; Onogi, S.; Miura, Y. Bioconjugate Chem. 2010, 21, 1079–1086.

77

Chapter 4

Plant Cell-Inspired Hydrogel Composites with High Mechanical Strength Naozumi Teramoto,* Keisuke Wakayama, Mitsuru Harima, Toshiaki Shimasaki, and Mitsuhiro Shibata Department of Applied Chemistry, Faculty of Engineering, Chiba Institute of Technology, 2-17-1 Tsudanuma, Narashino, Chiba 275-0016, Japan *E-mail: [email protected]

We have proposed a novel strategy for reinforcing hydrogels with polymer foam based on the inspiration from the framework of plant cells, and we found that the mechanical property of hydrogels are significantly improved by the reinforcement with polyurethane foam. In the present study, we prepared poly(ethylene glycol) (PEG) hydrogels reinforced by open-cell type polyurethane foam. A cut piece of polyurethane foam was impregnated with an aqueous solutions of poly(ethylene glycol) diacrylate, followed by polymerization using 2,2′-azobis(2-methylpropionamidine) dihydrochloride (V-50) radical initiator. Without polyurethane foam, the PEG hydrogel was broken at the compression stress of around 40 kPa. On the other hand, the PEG gel combined with polyurethane foam did not show the evident break point, and the hydrogel composites endured very high compression stress >2 MPa. We used two types of polyurethane foams with different cell sizes. The compression modulus was influenced by the cell size. When the polyurethane foam with a small cell size was used for preparation of the hydrogel composite, the compression modulus of the hydrogel composite was higher than that of the hydrogel composite with the polyurethane foam with a large cell size. We also observed the cyclic recovery from the 70% compression. The cyclic compression test revealed that large hysteresis exists at the first compression cycle of the hydrogel composite. The mechanical properties of our hydrogel

© 2017 American Chemical Society

composite in the compression test are comparable to tough hydrogels developed recently.

1. Introduction Hydrogel is one of promising materials as a biomaterial for regenerative medicine (1, 2). Hydrogels show solid properties, though they contain large amount of water. The distinguished advantages of hydrogels are biocompatibility and diffusibility of solutes in hydrogel due to its hydrophilicity and large content of water (2, 3). Though hydrogels are attractive for biomaterials, most traditional hydrogels are mechanically weak and brittle. Many trials to improve mechanical properties of hydrogels have been carried out, and some successful cases are being reported in this decade (4–13). Among these cases, double network gels (4), topological gels (5), nanocomposite gels (6) and tetra-PEG gels (7) are prominent examples of high-performance hydrogels. Other researches have been following them to produce high-performance hydrogels. Huang et al. (8) developed a composite hydrogel containing macromolecular microspheres as a novel type of composite hydrogel. The microspheres play a role as both an initiator and a crosslinker. The composite gel showed very high mechanical strength. Sun et al. (9) found that a double-network hydrogel prepared from ionically crosslinked alginate and covalently crosslinked polyacrylamide had extremely stretchable and tough mechanical properties. The hydrogel kept the stretchability even if it contains a notch. Zhang et al. (10) succeeded to enhance mechanical properties of a biodegradable hydrogel composed of poly(ethylene glycol) and oligo(trimethylene carbonate) by controlling the crosslinked density and hydrophilic-hydrophobic balance. A poly(vinyl alcohol) (PVA) hydrogel prepared by freezing-thawing cyclic processing has been known to show good mechanical properties (14). Tong et al. (11) prepared carbon nanotube (CNT)-reinforced PVA hydrogels and found that addition of CNT at 0.5% of weight of PVA improved mechanical strength by 94%. Zhang et al. (12) also improved the mechanical strength of PVA hydrogel using graphene. Li et al. (13) reported the very high mechanical strength of hybrid hydrogels prepared from PVA and poly(acrylamide) (PAAm) with long stability in water. The hybrid gel contains a physical network of PVA and chemical network of PAAm. It shows very high toughness compared to PVA hydrogels. We have proposed a novel strategy for reinforcing hydrogels using polymer foams based on the inspiration from the framework of plant cells, and we found that the mechanical property of poly(sodium acrylate) (Na-PAA) hydrogel was significantly improved by the reinforcement with polyurethane foam in the previous study (15). Plant cells are surrounded by the strong cell walls mainly composed of cellulose. Under water shortage conditions, green plants wilt and fall to the ground, which means that the water in cells also plays an important role to help maintain the mechanical strength of the plant body. This combination of cell walls and water with in the walls is important and gave good suggestion to us. Inspired from the similarity of the plant body framework and the cell structure of polymer foam, we proposed a new type of hydrogel composites (Figure 1). In our 80

hydrogel composite, polymer foam plays a role of cell walls, and hydrogel plays a role of water in a plant cell. From the different concepts, two research groups prepared and evaluated the composites from polyurethane foam and hydrogels (16, 17).

Figure 1. Schematic concept of plant cell-inspired hydrogel composites with high mechanical strength. In the previous study (15), we prepared the composite hydrogel from polyurethane foam and Na-PAA. Since Na-PAA has very high water absorbability, the close packing of the hydrogel in the polyurethane foam was expected. Our subsequent study, however, shows the additional water absorption of the hydrogel did not affect the mechanical properties largely (unpublished data). Therefore we intended to investigate the possibility of reinforcement of hydrogel with polyurethane foam using another polymer material having no ionic groups. In the present study, we prepared hydrogel composites from PEG and polyurethane foam (PUF). PEG is a nonionic water soluble polymer constructed from repeating units of oxyethylene. It is a non-toxic biocompatible polymer often used in biomedical applications (18–20). PEG is also soluble in organic solvents and ionic liquids, and is also one of promising candidates for the polymer gel electrolyte of ion-conductive materials, which is expected to apply to lithium-ion batteries (21, 22). Polyurethane is also non-toxic and often used for artificial organs (23, 24). For industrial applications, polyurethane was mainly used as elastomer and polymer foam. The mechanical properties of polyurethane are fascinating due to its flexibility and high toughness (25). PUF used here is an open-cell type foam so that it can be impregnated with aqueous solution of reactive PEG. Melamine foam (MF) with fibrous morphology was used to compare the effect of foam morphology on mechanical properties of the composite gel. Here we prepared the PEG hydrogel/PUF composites and found its extremely toughness in the compression test. 81

2. Sample Preparation Hydrogel composites were prepared by a simple method. The aqueous solutions of poly(ethylene glycol) diacrylate (PEGDA) (number average molecular weight ~700, Product No. 455008, Sigma-Aldrich, MO) were prepared at the concentration of 10%. To the solution was added 0.2wt% of a thermolabile radical initiator, 2,2′-azobis(2-methylpropionamidine) dihydrochloride (V-50, Wako Pure Chemical Industries, Ltd., Osaka, Japan). A piece of polymer foam with a columnar shape (20 mm diameter × 10 mm height) is immersed in the PEGDA solution poured into a small Teflon® cylindrical container with 20 mm inner diameter equipped with a screw cap. The piece of polymer foam with a columnar shape was prepared by clipping using a steel punch cutter with a diameter of 20 mm. We used three types of polymer foams. Two are polyurethane foams (PUFs), CF-S with a small cell size (cell size: 68 cells/25 mm) and CFH-30 with a large cell size (cell size: 30 cells/25 mm) kindly provided from INOAC Corp. (Aichi, Japan). The other is a melamine foam (MF) (Basotect®) produced by BASF SE (Ludwigshafen am Rhein, Germany) purchased from Strider Co. Ltd. (Toyohashi, Japan). Figure 2 shows scanning electron microscope (SEM) photographs of these foams. MF did not show the cell structure but fibrous network. After degassing the polymer foam impregnated with the PEGDA solution for 30 min, the foam was heated at 70°C for 24 h in the screw-capped container to obtain the hydrogel composite. Calculated from the cell size, the averaged diameters of hydrogel in cells of PUFs are 370 μm (CF-S) and 830μm (CFH-30).

Figure 2. SEM photographs of foams used for hydrogel composites: (a) PUF CF-S, (b) PUF CFH-30, and (c) melamine foam (MF) SEM observation of hydrogels and their composites were carried out after lyophilization. The cross-sectional surface morphology of lyophilized samples was observed by a Hitachi S-4700 field emission scanning electron microscope (FE-SEM) (Hitachi High-Technologies Corp., Tokyo, Japan); the accelerating voltage was 1 kV or 5 kV, and the samples were coated with gold prior to the observation.

3. Compression Test of Samples Hydrogels and hydrogel composites with a columnar shape (20 mm diameter × 10 mm height) were prepared for the compression test. At least six samples 82

were prepared for one condition. The compression test was carried out by a Shimadzu EZ-S tabletop universal tester (Shimadzu Corp., Kyoto, Japan) with a 100 N load cell or a Shimadzu AG-I tabletop universal tester with a 5 kN load cell at a crosshead speed of 1 mm/min. The former tester was used for weaker samples such as non-reinforced hydrogels and polymer foams alone. The latter tester was used for stronger samples such as hydrogel composites.

4. Results and Discussion 4.1. Appearance of the Hydrogel Composites The hydrogel composites were easily prepared by our very simple method by which PEGDA was polymerized using thermolabile radical initiator in the aqueous solution in the presence of an impregnated polymer foam piece. The number average molecular weight of PEGDA was ~700. The length of PEG moiety influences hydrophilicity and crosslinking density. While the repeating units of oxyethylene are hydrophilic, the main chains which were formed by radical polymerization and composed of polymerized acrylate units are hydrophobic. Though the PEGDA aqueous solution was transparent, the hydrogel of PEGDA was white and turbid. Some poly(methyl methacrylate)s with short PEG side chains are known to show lower critical solution temperature (LCST) (26). We consider that the turbidity of PEGDA hydrogel was due to the phase separation of PEGDA during polymerization at a high temperature (70°C). Figure 3a shows the SEM photograph of lyophilized 10% PEGDA hydrogel (PEGDA-10). Our PEGDA hydrogel showed partial porous morphology without interconnected pores. Figure 3b and 3c shows SEM photographs of lyophilized 10% PEGDA hydrogel composites with PUF CF-S (PEGDA-10-CF-S) and PUF CFH-30 (PEGDA-10-CFH-30), respectively. We observed the cell structure of PUF and small lyophilized gel beads of PEGDA in each cell. We consider that the interconnection of hydrogel was broken during lyophilization, since the gel shrank during lyophilization. On the other hand, the fibrous melamine network of MF was almost covered by PEGDA gel (Figure 3d). The cleft observed in the SEM photograph may have appeared during lyophilization. Table 1 shows the hydrogel content of each composite. The hydrogel content was calculated from the weight of the hydrogel composite sample and the weight of the foam piece used for preparation of the hydrogel composite. The results revealed that the hydrogel composite was composed of at least 90% of hydrogel. This result suggests that the main component of the hydrogel composites is water, and it is expected that the composites keep many characteristics of hydrogels. The hydrogel content also related with the bulk density of the foam. When the bulk density of foam was higher, the hydrogel content became lower since the substantial volume of the foam polymer was larger. On the other hand, the apparent density of the hydrogel composites was 1.00-1.01 g cm-3 and close to that of the PEGDA hydrogel (1.02 g cm-3), when measured by Archimedes’ principle. The bulk density of foam did not influence the apparent density of the hydrogel composites, because the real density of the foam polymer is not so different from that of hydrogel. 83

Figure 3. SEM photographs of lyophilized hydrogel and hydrogel composites: (a) PEGDA-10, (b) PEGDA-10-CF-S, (c) PEGDA-10-CFH-30, and (d) PEGDA-10-MF

Table 1. The hydrogel content of each hydrogel composite.

a

Hydrogel composite sample

Hydrogel content (wt%)

Foam densitya (g/cm3)

PEGDA-10-CF-S

91.1

0.072±0.005

PEGDA-10-CFH-30

96.3

0.030±0.003

PEGDA-10-MF

98.8

0.010±0.003

Foam density is the value of bulk density published by the vender.

4.2. Mechanical Properties of Hydrogel Composites The mechanical properties of hydrogel composites were investigated comparing with those of the PEGDA hydrogel. Figure 4 shows the stress-strain curves of PEGDA hydrogel and the composite hydrogels of PEGDA with each foam in the compression tests. The strength of PEGDA hydrogel (PEGDA-10) was lower than 50 kPa and the gel was broken at the strain less than 60%. The slope of the curve was low, implying that the modulus of the PEGDA hydrogel was also low. On the other hand, the PEGDA hydrogel/PUF composites (PEGDA-10-CF-S and PEGDA-10-CFH-30) did not show an evident break point, and the composite gel material did not break after loading high stress >2 MPa and high strain ~95%. Secondly, the mechanical properties of hydrogel composites were compared with those of PUF. Figure 5 shows the stress-strain 84

curves of the hydrogel/PUF composites comparing with PUF. The slope of the stress-strain curve of PUF without hydrogel was very low. The stress on PUF kept very low throughout the test, though PUF was not broken even at high strain ~95%. These results reflect our concept of plant cell-inspired hydrogel composites. When there is no hydrogel, the foam does not bear the force of compression, as if a green plant body wilts and falls down from lack of water. The counteracting force of liquid surrounded by cell walls is important in the plant body. Furthermore hydrogel in each cell was protected by the cell wall of PUF. The compression force is converted to the force of expanding horizontally at the high compression strain and counteracting force (Figure 6). Though the strength of PUF is considered to influence the mechanical strength of the hydrogel composite at the breaking strain, we could not measure the stress and strain at break of hydrogel composites. The tensile strengths of CF-S PUF and CFH-30 PUF are reported as 196 kPa and 147 kPa by the manufacturer, respectively. On the other hand, MF did not show a significant reinforcing effect in the hydrogel composite (Figure 4). The modulus of the PEGDA hydrogel/MF composite (PEGDA-10-MF) increased in comparison with PEGDA-10, while its brittleness increased. This tendency is typical for fiber-reinforced materials (27, 28). The compression modulus of the hydrogel composite with the small cell-size PUF (PEGDA-10-CF-S) was higher than that of the hydrogel composite with the large cell-size PUF (PEGDA-10-CFH-30). These results imply that the reinforcing effect of foams depends on the morphology and cell size. It is noteworthy that the highest stress of the hydrogel composite with the small cell-size PUF was >5 MPa, and that the hydrogel composite did not broken after the test (Figure 7), though some small gel pieces have come out from the foam.

Figure 4. Stress-strain curves of hydrogel and hydrogel composites: PEGDA-10 (open triangle), PEGDA-10-CF-S (open diamond), PEGDA-10-CFH-30 (open square), and PEGDA-10-MF (open circle). The arrow shows the break point of PEGDA-10-MF. 85

Figure 5. Stress-strain curves of hydrogel composites comparing with PUF: PEGDA-10-CF-S (open diamond), PEGDA-10-CFH-30 (open square), CF-S (closed circle), and CFH-30 (closed triangle).

Figure 6. Schematic explanation of the counteracting force at the compression of foam and hydrogel/foam composite.

86

Figure 7. Photographs of hydrogel and its composites after the tensile test.

Figure 8. Reproducibility of the stress-strain curve of different PEGDA-10-CF-S samples.

87

We also checked the reproducibility of the stress-strain curve of the PEGDA hydrogel/PUF composite. Usually, hydrogels are very sensitive to the small cracks of their surface and inhomogeneity of crosslinking. So we must pay careful attention at handling hydrogels, and the stress-strain curves of conventional hydrogels are less reproducible. Figure 8 shows the stress-strain curves of PEGDA-10-CF-S in three compression tests using three different specimens. These curves overlapped each other with high reproducibility. This reproducibility is very important for the commercial use. The mechanical properties of PEGDA hydrogel and its composites with foams were summarized in Table 2. Standard deviations of PEGDA hydrogel/PUF composites were very low relative to their high stress values. The compression tests were stopped at the strain of 95% and the stress at 95% strain was described for PEGDA hydrogel/PUF composites, because the clearance of the upper compression plate and the lower compression plate was very small at 95% strain. The number of specimens was 6 for each test.

Table 2. The Mechanical properties of PEGDA hydrogel and hydrogel composites.

a

Hydrogel or composite sample

Stress at break or at 95% straina (MPa)

Strain at break (%)

PEGDA-10

0.034±0.013

52.5±5.4

PEGDA-10-CF-S

5.68±0.11a

N.D.a

PEGDA-10-CFH-30

2.67±0.16a

N.D.a

PEGDA-10-MF

0.189±0.015

46.8±3.8

The test was stopped at the strain of 95%. The strain at break was not determined.

4.3. Cyclic Compression Tests of Hydrogel Composites We investigated the hysteresis in the cyclic compression test of PEGDA-10CF-S and PEGDA-10-CFH-30. Each sample underwent three cycle compression and decompression at a crosshead speed of 1 mm/min without intervals. The maximum strain was set at 70%. Figure 9 and 10 shows the stress-strain curves of the cyclic compression tests for PEGDA-10-CF-S and PEGDA-10-CFH-30, respectively. These hydrogel composites showed large hysteresis especially at the first cycle and relatively small hysteresis at the second and third cycles. The curve of the third cycle was overlapped with that of the second cycle. This result indicates an interesting behavior of our hydrogel composites, suggesting that some irreversible event occurred in the first cycle. We consider that this behavior is caused by the breaking or weakening of the interconnection between each hydrogel particle in each cell of the foam at the first cycle. The deformation of the composite with the force toward the compression axis changes the shape of hydrogel particles in each cell to the disk shape. At the point where the hydrogel particle contacts the cell wall, the part of the hydrogel is protected from breaking as well as plant cells 88

are protected with cell walls. Therefore the stress for breaking is considered to be concentrated at the interconnection part. The similar behavior was observed also in the mechanical test of the double network hydrogels (29). Detailed analysis for this behavior will be required as a future study.

Figure 9. Stress-strain curves of PEGDA-10-CF-S at the cyclic compression test: The first cycle (open circle), the second cycle (open diamond), and the third cycle (open square)

Figure 10. Stress-strain curves of PEGDA-10-CFH-30 at the cyclic compression test: The first cycle (open circle), the second cycle (open diamond), and the third cycle (open square) 89

5. Conclusions Hydrogel composites of PEG and polymer foams were prepared and evaluated by compression tests. PEG hydrogel was synthesized by the radical polymerization of PEGDA using thermolabile radical initiator in the presence of polymer foam. The mechanical strength and toughness of the PEG hydrogel was significantly improved by combination with PUF. The hydrogel composites with PUF endured very high compression stress >2 MPa and high strain of 95%. When the polyurethane foam with a small cell size (68 cells/25 mm, 370 μm on average) was used, the compression stress went beyond 5 MPa at 95 % strain without an apparent break of the composites. Despite of this significant reinforcing effect, the content of the foam was only less than 10% by weight. In contrast, the PEG hydrogel was not reinforced with MF at such high level. In the cyclic test up to 70% strain, PEG hydrogel/PUF composites showed large hysteresis especially at the first cycle and relatively small hysteresis at the second and third cycles. Hydrogel composites with PUF also showed very good reproducibility in the compression properties, suggesting that they are suited for industrial and commercial uses in terms of quality stability. Since PEG hydrogel and polyurethane are both known as a good candidate for biomaterials, our composite is expected to be a new biomaterial. We expect that our concept of “plant cell-inspired hydrogel composites” can be applied not only to the combination of PEG and polyurethane but also to various combinations of hydrogel and foam, such as hydrogel from biopolymers and biodegradable foams. Future researches will seek detailed mechanical properties and biomaterial applications such as artificial cartilage, tendon, and ligament.

Acknowledgments The authors would like to acknowledge support on the presentation in the ACS fall meeting 2016 with Sasagawa Grants for Science Fellows by the Japan Science Society.

References 1. 2. 3. 4. 5. 6. 7.

8.

Hunt, J. A.; Chen, R.; van Veen, T.; Bryan, N. J. Mater. Chem. B 2014, 2, 5319–5338. Kirschner, C. M.; Anseth, K. S. Acta Mater. 2013, 61, 931–944. Seliktar, D. Science 2012, 336, 1124–1128. Gong, J. P.; Katsuyama, Y.; Kurokawa, T.; Osada, Y. Adv. Mater. 2003, 15, 1155–1158. Okumura, Y.; Ito, K. Adv. Mater. 2001, 13, 485–487. Haraguchi, K.; Takehisa, T. Adv. Mater. 2002, 14, 1120–1124. Sakai, T.; Matsunaga, T.; Yamamoto, Y.; Ito, C.; Yoshida, R.; Suzuki, S.; Sasaki, N.; Shibayama, M.; Chung, U.-i. Macromolecules 2008, 41, 5379–5384. Huang, T.; Xu, H. G.; Jiao, K. X.; Zhu, L. P.; Brown, H. R.; Wang, H. L. Adv. Mater. 2007, 19, 1622–1626. 90

9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.

Sun, J.-Y.; Zhao, X.; Illeperuma, W. R. K.; Chaudhuri, O.; Oh, K. H.; Mooney, D. J.; Vlassak, J. J.; Suo, Z. Nature 2012, 489, 133–136. Zhang, C.; Aung, A.; Liao, L.; Varghese, S. Soft Matter 2009, 5, 3831–3834. Tong, X.; Zheng, J.; Lu, Y.; Zhang, Z.; Cheng, H. Mater. Lett. 2007, 61, 1704–1706. Zhang, L.; Wang, Z.; Xu, C.; Li, Y.; Gao, J.; Wang, W.; Liu, Y. J. Mater. Chem. 2011, 21, 10399–10406. Li, J.; Suo, Z.; Vlassak, J. J. J. Mater. Chem. B 2014, 2, 6708–6713. Peppas, N. A.; Stauffer, S. R. J. Controlled Release 1991, 16, 305–310. Teramoto, N.; Shigehiro, O.; Ogawa, Y.; Maruyama, Y.; Shimasaki, T.; Shibata, M. Polym. J. 2014, 46, 592–597. Liu, K.; Ovaert, T. C.; Mason, J. J. J. Mater. Sci.: Mater. Med. 2008, 19, 1815–1821. Huang, Y.; He, K.; Wang, X. Mater. Sci. Eng. C 2013, 33, 3220–3229. Molineux, G. Cancer Treat. Rev. 2002, 28, 13–16. Lin, C.-C.; Anseth, K. S. Pharm. Res. 2009, 26, 631–643. Zhu, J. Biomaterials 2010, 31, 4639–4656. Manuel Stephan, A. Eur. Polym. J. 2006, 42, 21–42. Liao, C.; Sun, X.-G.; Dai, S. Electrochim. Acta 2013, 87, 889–894. Zdrahala, R. J.; Zdrahala, I. J. J. Biomater. Appl. 1999, 14, 67–90. Janik, H.; Marzec, M. Mater. Sci. Eng. C 2015, 48, 586–591. Akindoyo, J. O.; Beg, M. D. H.; Ghazali, S.; Islam, M. R.; Jeyaratnam, N.; Yuvaraj, A. R. RSC Adv. 2016, 6, 114453–114482. Lutz, J.-F. J. Polym. Sci., Part A: Polym. Chem. 2008, 46, 3459–3470. Ku, H.; Wang, H.; Pattarachaiyakoop, N.; Trada, M. Composites, Part B 2011, 42, 856–873. Teramoto, N.; Hayashi, A.; Yamanaka, K.; Sakiyama, A.; Nakano, A.; Shibata, M. Materials 2012, 5, 2573–2585. Webber, R. E.; Creton, C.; Brown, H. R.; Gong, J. P. Macromolecules 2007, 40, 2919–2927.

91

Chapter 5

Synthesis and Temperature-Responsiveness of Poly(ethylene glycol)-like Biodegradable Poly(ether-ester)s Yuichi Ohya,*,1,2 Akihiro Takahashi,2 Hiroki Takaishi,1 and Akinori Kuzuya1,2 1Department

of Chemistry and Materials Engineering, Faculty of Chemistry, Materials and Bioengineering, Kansai University, 3-3-35 Yamate, Suita, Osaka 564-8680, Japan 2Organization for Research and Development of Innovative Science and Technology (ORDIST), Kansai University, 3-3-35 Yamate, Suita, Osaka 564-8680, Japan *E-mail: [email protected]

Poly(ethylene glycol) (PEG) is widely used as a biomedical material because it possesses the highest level of biocompatibility among synthetic polymers. However, PEG is water-soluble and non-biodegradable. To provide biodegradable water-insoluble bulk materials possessing the biocompatibility of PEG, water-insoluble poly(ether-ester)s were synthesized by thiol-ene polyaddition of poly(ethylene glycol) diacrylates (PEGDA) and alkyl dithiols (ADTs). Among them, P(PEG-DDT) prepared from PEGDA and decanedithiol (DDT) film cast on a glass substrate possessed temperature-responsive changes only when immersed in water. The P(PEG-DDT) film was transparent in a water bath at 0 ºC, but become opaque after subsequent soaking in a hot water bath at 40–70 ºC. The mechanism was examined by measuring transmittance changes in response to temperature, differential colorimetry, and air contact angle of the film in water. Water-insoluble poly(ether-ester)s exhibiting temperature-responsiveness have potential, not only as new bulk-state biomaterials with PEG-like biocompatibility but

© 2017 American Chemical Society

also as new environmentally friendly temperature-responsive materials.

Introduction Poly(ethylene glycol) (PEG) is an amphiphilic polymer that is soluble in common organic solvents and in water, and is one of the most popular polymers used in biomaterials. It is a non-toxic, non-immunogenic polymer, and can be eliminated through renal and hepatic pathways. In addition, PEG demonstrates the lowest level of interaction with proteins or cells among synthetic and natural polymers (1–3). These PEG characteristics permit its use in a wide variety of established and emerging applications in pharmaceutics to improve the pharmacokinetic properties of drugs and drug carriers (4–16). Chemical modification using PEG (PEGylation) is effective in prolonging blood circulation times and the half-life in the body of drugs and drug carriers by preventing protein absorption and entrapment by reticuloendothelial systems. However, PEG is not biodegradable. Yamaoka et al. conducted a detailed study on the distribution and tissue uptake of PEGs with different molecular weights after intravenous (i.v.) administration into mice (17). The renal clearance of PEGs decreased with an increase in molecular weight, with the most dramatic change occurring at approximately 30,000 Da. This indicates that high-molecular-weight (>30,000 Da) PEG is difficult to excrete from the body, despite its biocompatibility. To provide suitable implantable biomaterials for temporal use, polymers must have a molecular weight high enough to escape renal clearance until their function is fulfilled and then degrade in the body. Thus, polymers possessing both PEG-like biocompatibility and the ability to biodegrade under physiological conditions are the best candidates for temporal implant biomaterials. Several groups have reported biodegradable poly(ether-ester)s composed of functional amino acids and PEG (18, 19). Another report described the synthesis of an alternative copolymer of PEG and L-aspartic acid, poly(L-Asp-alt-PEG), through polycondensation of PEG and an L-aspartic acid derivative. The poly(L-Asp-alt-PEG) obtained with alkyl side chains possessed both biodegradability and temperature-responsiveness, but did not have a molecular weight high enough to prevent renal clearance (20). High-molecular-weight poly(ether-ester)s are difficult to obtain via polycondensation. Wang et al. reported the synthesis of biodegradable and temperature-responsive PEG analogs (DPEGs) by thiol-ene polyaddition of dithiols and poly(ethylene glycol) diacrylates (PEGDA) or di(meth)acrylates. In aqueous solution, the DPEGs had a lower critical solution temperature (LCST), and the response temperature could be adjusted by changing the type of dithiol and the molecular weight of the PEG derivative (21, 22). Temperature-responsive polymers often contain PEG because it undergoes temperature-dependent dehydration in aqueous solution (23–26). Previously, PEGs have been used as a water-soluble polymer, solubilizer, and hydrophilic brush for a substrate surface in aqueous solution. However, water-insoluble “bulk” materials having excellent biocompatibility are also important as implantable biomedical materials. The present study describes 94

the preparation of water-insoluble “bulk” materials possessing PEG-like biocompatibility and degradation to low-molecular-weight compounds that could be excreted from the body by employing a poly(ether-ester) backbone containing PEGDA and alkyldithiols (ADT). Several types of water-insoluble poly(ether-ester)s with relatively short PEG chains and various length of alkyl chains were synthesized. The process accidentally revealed that cast films of the water-insoluble copolymers obtained possessed temperature-responsive transmittance changes in hot water. Therefore, the synthesis of PEG-like biodegradable poly(ether-ester)s and their temperature-responsive properties were investigated. Poly(PEG-alkyl dithiol)s [P(PEG-ADT)s] composed of low-molecularweight PEGs and alkyl chains of various length were prepared via thiol-ene polyaddition of PEGDA and α,ω-alkyl dithiols. The temperature-responsiveness of the P(PEG-ADT) was evaluated by transmittance changes, calorimetric analysis, and contact angle measurements. The temperature-responsiveness and PEG-based biocompatibility of these water-insoluble poly(ether-ester)s are expected to be useful biocompatible biodegradable biomaterials.

Experimental Materials 1,6-Hexanedithiol (HDT), 1,10-decanedithiol (DDT), and triethylamine were purchased from Wako Pure Chemical Ind., Ltd. (Osaka, Japan). The PEGDA (molecular weight = 700 Da), 1,4-butanedithiol (BDT), and 1,8-octanedithiol (ODT) were purchased from Sigma-Aldrich (St. Louis, USA). 1,2-Ethanedithiol (EDT) was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Water was purified using a reverse osmotic membrane method (Milli-Q). Other solvents and reagents were commercial grade and used without further purification. Measurements 1H-

and 13C-nuclear magnetic resonance (NMR) spectra were recorded on a JNM-GSX-400 (Jeol, 400 MHz) instrument using deuterated chloroform (CDCl3) as a solvent. The chemical shifts were calibrated against tetramethylsilane (TMS) and the CDCl3 solvent signal. The weight-average molecular weight (Mw), number-average molecular weights (Mn), and polydispersity indices (Mw/Mn) of the copolymers were determined by size exclusion chromatography (SEC) (column: TSKgel Multipore HXLM × 2, detector: RI). Measurements were obtained using dimethylformamide (DMF) as eluent at a flow rate of 1.0 mL/min at 40 ºC and a series of PEG standards. Transmittance changes of the copolymer films were measured using a V-650 UV-Vis spectrophotometer (Jasco). Thermal analysis of the copolymers was conducted using a differential scanning calorimeter (DSC-60, Shimadzu) with sealed aluminum pans. Contact angles of the films were recorded using a contact angle meter (Kyowa Interface Science, Co., Ltd., DM-550). 95

Synthesis of P(PEG-ADT)s A series of P(PEG-ADT)s having various alkyl chain lengths were synthesized via thiol-ene polyaddition according to Scheme 1. Typical procedures for preparation of P(PEG-DDT) (y = 10 in Scheme 1) are as follows. A mixture of PEGDA (896 mg, 1.28 mmol) and DDT (264 mg, 1.28 mmol) was dried in vacuo and dissolved in anhydrous DMSO (8 mL). Triethylamine (73 mg, 0.60 mmol) was added to the solution and stirred at room temperature (r.t.) for 36 h. Then the reaction mixture was poured into diethylether/methanol (10/1, v/v) and washed three times. The precipitate obtained was dried in vacuo at r.t. to give P(PEG-ADT). Yield: 72%.

Scheme 1. The synthesis of P(PEG-ADT)s. 1H NMR (400 MHz, CDCl3): δ (ppm) = 1.20-1.31 (br, -CH2CH2CH2-), 1.31-1.42 (br, -SCH2CH2CH2-), 1.52-1.62 (m, -SCH2CH2CH2-), 2.47-2.55 (t, -SCH2CH2CH2-), 2.59-2.81 (m, -OC(=O)CH2CH2S-), 3.55-3.68 (br, -OCH2CH2O-), 3.68-3.73 (t, -COOCH2CH2O-), 4.23-4.28 (t, -COOCH2CH2O-). 13C NMR (100 MHz, CDCl3): δ (ppm) = 26.7, 28.8, 29.1, 29.3, 29.4, 32.0, 34.7, 63.7, 69.0, 70.5, 70.6, 172.1. Other P(PEG-ADT)s, P(PEG-ODT) (y=8), P(PEG-HDT) (y=6), P(PEGBDT) (y=4), P(PEG-EDT) (y=2), were also synthesized by the similar procedures.

Temperature-Responsiveness of Copolymer Films Cast copolymer films were prepared using a spincoater (1HD7, Mikasa Co. Ltd.). The P(PEG-DDT) was dissolved at a concentration of 100 mg/mL in CHCl3, spin-coated onto a flat glass substrate, and dried in vacuo at r.t. to give a transparent film ca. 500 μm thick. The temperature-responsive behavior of P(PEG-DDT) film on a glass substrate was evaluated in the presence and absence of water. The P(PEG-DDT) film on a glass-substrate was placed in ice-cold water, and then moved into a water bath at various temperatures. The appearance of the films was observed by the naked eye, and the film transmittance at 500 nm was measured. Transmittance changes of the films on a glass plate in the air (without water) also were observed. 96

Results and Discussion Synthesis of the Copolymers The P(PEG-ADT)s containing alkyl chains of different lengths were successfully synthesized via thiol-ene polyaddition using PEGDA and ADTs according to Scheme 1. Results are shown in Table 1. The Mn, Mw, and Mw/Mn values were in the ranges of 22,000–35,000, 37,000–58,000, and 1.62–2.11, respectively, and yields were greater than 67%. A trend toward a higher degree of polymerization (DP) for longer alkyl chain ADTs was observed. Elution profiles of the P(PEG-ADT)s are shown in Figure 1, and suggest successful purifications. The polydispersity indices (Mw/Mn) were relatively high, but unimodal distributions were observed for all P(PEG-ADT)s. Typical examples of 1H-NMR spectra of P(PEG-DDT) and the starting materials, PEGDA and DDT, are shown in Figure 2. The acryl group signals from PEGDA at 5.8–6.6 ppm were not present in the P(PEG-DDT) spectrum, suggesting that the terminal acryl groups completely reacted. New peaks at 2.6–2.8 ppm also were observed and assigned to –S–CH2CH2–C(=O) (reaction product of the acryl group). Other P(PEG-ADT)s gave similar results. All of the P(PEG-ADT)s obtained were insoluble in water. The P(PEG-EDT), P(PEG-BDT), P(PEG-HDT), and P(PEG-ODT) were viscous pastes in the dry state at r.t. and the viscosity of the copolymer increased with of alkyl chain length. The P(PEG-DDT) was solid after drying. However, preparing self-standing films from P(PEG-ADT)s, including P(PEG-DDT), was not possible.

Table 1. Results of synthesis of P(PEG-ADT)s Name of P(PEG-ADT)

ya

Mn (Da)b

Mw (Da) b

Mw/Mn b

DP c

Yield (%)

P(PEG-EDT)

2

23,000

49,000

2.11

29

73

P(PEG-BDT)

4

22,000

37,000

1.70

26

69

P(PEG-HDT)

6

26,000

46,000

1.75

31

76

P(PEG-ODT)

8

27,000

44,000

1.62

31

67

P(PEG-DDT)

10

35,000

58,000

1.66

39

a

Number of alkyl chain of ADTs (y in Scheme 1). polymerization estimated from Mn.

97

b

Estimated by SEC.

72 c

Degree of

Figure 1. SEC profiles for P(PEG-EDT), EDT (a), P(PEG-BDT), BDT (b), P(PEG-HDT), HDT (c), P(PEG-ODT), ODT (d), P(PEG-DDT), DDT (e), and PEGDA. (see color insert)

98

Figure 2. 1H-NMR spectra of P(PEG-DDT) (top), PEGDA (middle) and DDT (bottom) in CDCl3. (see color insert)

Temperature Responsiveness The P(PEG-DDT) was solid in dry state at r.t. and insoluble in water at 0–100 ºC, but exhibited temperature-responsiveness when placed in water. As shown in Figure 3, P(PEG-DDT) film cast on a glass-substrate was transparent at 0–100 ºC in air. In contrast, despite the transparency of P(PEG-DDT) film in water at 0 ºC, it become opaque after placement in a water bath at 40–70 ºC. The turbidity of the film at 70 ºC appeared to be slightly greater than that at 40 ºC. These results indicate that the P(PEG-DDT) film underwent a temperature-responsive phase transition only in the presence of water. The turbidity change may be caused by phase separation of the copolymer on hydration/dehydration of the PEG segments. In addition, film kept in water at 40 ºC for 25 min was still translucent, but turbidity was decreased slightly compared to that before incubation. The turbid P(PEGDDT) film created by placement in hot water became transparent after placement in an ice-cold water bath again. These results demonstrate that the temperatureresponsive transparency change was reversible.

99

Figure 3. Photographs of P(PEG-DDT) films in the air or water at various temperature and history.

The temperature-responsive transmittance changes of P(PEG-DDT) films also were investigated by UV-Vis spectrophotometry. Transmittance changes of P(PEG-DDT) films placed in an ice water bath and then transferred to warmer water were measured. Figure 4a shows plots of film transmittance vs. duration of time in the warm water bath. When P(PEG-DDT) film was transferred to a water bath at 10 ºC, almost no transmittance change was observed. However, the transmittance of the film decreased dramatically to 20% within 1 min after transfer into a water bath at 40 ºC. The transmittance decreased to approximately 10% within 30 sec when the film was placed in a water bath at 70 ºC. Interestingly, the decreased transmittance values recovered slowly upon continued incubation in the warm water bath. These results indicate that the films exhibited temperature-responsive transmittance changes in the presence of water, but the change was temporal. Figure 4b shows plots of minimum transmittance value and transmittance value after 23 min in the warm water bath vs. water bath incubation temperature. The minimum transmittance value decreased dramatically when water temperature was 20–50 ºC, with a transition point of approximately 35 ºC. These results suggest that temperature-triggered phase separation of the films occurred by dehydration of PEG segments of the copolymer, which was accelerated with an increase in incubation temperature. In addition, the dehydration-induced phase separation was temporal and gradually reached an equilibrium state at the incubation temperature.

100

Figure 4. a) Time course of transmittance change at 500 nm after temperature change of P(PEG-DDT) film in water. The films were soaked in ice-cooled water (0 ºC) and subsequently moved to 10 ºC (dotted black line), 40 ºC (dashed blue line), 70 ºC (solid red line) water bath. b) Plots of minimum transmittance (●) and transmittance after 23 min (▲) vs. incubation temperature change for P(PEGDA-DDT) film. (see color insert)

To investigate the phase transition behavior, calorimetric analysis of the films was conducted using DSC. Figure 5 shows DSC plots of P(PEG-DDT) in the presence and absence of water. The P(PEG-DDT) without water showed endothermic peaks corresponding to a glass-transition temperature (Tg) and melting temperature (Tm) of -61 and 6 ºC, respectively. In contrast, DSC of P(PEG-DDT) with water showed several new peaks in addition to Tg and Tm. The Tm overlapped a large free water peak. An endothermic peak from -40 to -20 ºC was attributed to eutectic composition between the PEG segment of P(PEG-DDT) and water. Reports have demonstrated that PEG shows an endothermic eutectic peak with water from - 40 to -8 ºC, with the temperature dependent on molecular weight (27–31). The endothermic shoulder peak observed from -10 to 0 ºC was attributed to melting of bound water (30, 31). In addition, a small broad endothermic peak was observed from 25 to 75 ºC, which was assigned to dehydration of P(PEG-DDT), because this peak was observed only in the presence of water.

101

Figure 5. DSC charts for P(PEG-DDT) in water (red), P(PEG-DDT) (blue), and water (black, dotted line) from -100ºC to 130 ºC. (see color insert) To investigate the change in surface properties upon the temperature change, the water contact angle and air contact angle in water were investigated. Figure 6 shows the time course of water contact angle in the air at 20 and 40 ºC. The contact angles decreased gradually from approximately 110º to approximately 65º in 2 min. No significant difference was observed between 20 and 40 ºC. A gradual decrease in water contact angle is typical behavior for hydrophilic-hydrophobic copolymers due to conformational changes in the polymer chains at the air-water interface. Therefore, this phenomenon was not likely to be closely related to the temperature-responsiveness of P(PEG-DDT). Figure 7 shows the time course of air bubble contact angle in water during the temperature change from 0 to 40 ºC. No obvious contact angle change was observed. Therefore, no meaningful information was obtained from the contact angle measurements, and the surface hydrophilicity change was not critical for the temperature-responsive transmittance changes of this polymer. Although the mechanism for the temperature-responsive transmittance change of this polymer remains unclear, the interaction of PEG with water causing processes such as dehydration must play a critical role.

Figure 6. Time course of water contact angle for P(PEG-DDT) at 20 ºC (○) and 40 ºC (◆). (see color insert) 102

Figure 7. Time course of air bubble contact angle for P(PEG-DDT) in water after the temperature change from 4 ºC to 40 ºC. The films were soaked in cooled water (4 ºC) and subsequently moved to 40 ºC water bath.

Conclusions P(PEG-ADT)s, including P(PEG-DDT), were prepared successfully by thiol-ene polyaddition of PEGDA and ADTs. The Mw values of the polymers obtained were greater than 37,000. The P(PEG-DDT) film cast on glass possessed interesting temperature-responsive behavior. The transparency of the film did not change upon heating in air. However, in a water bath, the transmittance of the P(PEG-DDT) film changed from transparent to translucent upon an increase in temperature. This phenomenon was reversible with another temperature change: the film returned to transparency on cooling. The P(PEG-DDT) showed a broad endothermic peak from 25 to 75 ºC due to dehydration of the PEG segments, which was related to the polymer’s temperature-responsive behavior. The mechanism for the temperature-responsive transmittance change remains unclear at present, but this water-insoluble, temperature-responsive poly(ether-ester) containing PEG possesses great potential, not only as a new bulk-state biomaterial with PEG-like biocompatibility and biodegradability but also as a temperature-responsive environmental-friendly material for new applications.

Acknowledgments This work was financially supported in part by a Grant-in-Aid for Scientific Research (No. 25282147, and 16H01854) from the Japan Society for the Promotion of Science (JSPS), and by the Kansai University Subsidy for Supporting Young Scholars, 2013 and Kansai University Outlay Support for Establishing Research Centers, 2016.

References 1. 2. 3.

Lee, J. H.; Lee, H. B.; Andrade, J. D. Prog. Polym. Sci. 1995, 20, 1043–1079. Park, K.; Shim, H. S.; Dewanjee, M. K.; Eigler, N. L. J. Biomater. Sci., Polym. Ed. 2000, 11, 1121–1134. Xia, B.; Xie, M.; Yang, B. J. Biomed. Mater. Res., Part A 2013, 101A, 54–63. 103

4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.

Greenwald, R. B.; Choe, Y. H.; McGuire, J.; Conover, C. D. Adv. Drug Delivery Rev. 2003, 55, 217–250. Elvira, C.; Gallardo, A.; San Roman, J.; Cifuentes, A. Molecules 2005, 10, 114–125. Banerjee, S. S.; Aher, N.; Patil, R.; Khandare, J. J. Drug Delivery 2012, 103973. Veronese, F. M. Biomaterials 2001, 22, 405–417. Hoffman, A. S. J. Controlled Release 2008, 132, 153–163. Kolate, A.; Baradia, D.; Patil, S.; Vhora, I.; Kore, G.; Misra, A. J. Controlled Release 2014, 192, 67–81. Pasut, G.; Veronese, F. M. Adv. Drug Delivery Rev. 2009, 61, 1177–1188. Pasut, G.; Sergi, M.; Veronese, F. M. Adv. Drug Delivery Rev. 2008, 60, 69–78. Filpula, D.; Zhao, H. Adv. Drug Delivery Rev. 2008, 60, 29–49. Fontana, A.; Spolaore, B.; Mero, A.; Veronese, F. M. Adv. Drug Delivery Rev. 2008, 60, 13–28. Caliceti, P.; Veronese, F. M. Adv. Drug Delivery Rev. 2003, 55, 1261–1277. Knop, K.; Hoogenboom, R.; Fischer, D.; Schubert, U. S. Angew. Chem., Int. Ed. 2010, 49, 6288–6308. Greenwald, R. B. J. Controlled Release 2001, 74, 159–171. Yamaoka, T.; Tabata, Y.; Ikada, Y. J. Pharm. Sci. 1994, 83, 601–606. Braunová, A.; Pechar, M.; Laga, R.; Ulbrich, K. Macromol. Chem. Phys. 2007, 208, 2642–2653. Wona, C.-Y.; Chua, C.-C.; Lee, J. D. Polymer 1998, 39, 6677–6681. Nagahama, K.; Hashizume, M.; Yamamoto, H.; Ouchi, T.; Ohya, Y. Langmuir 2009, 25, 9734–9740. Wang, N.; Dong, A.; Radosz, M.; Shen, Y. J. Biomed. Mater. Res., Part A 2008, 84A, 148–157. Wang, N.; Dong, A.; Tang, H.; Van Kirk, E. A.; Johnson, P. A.; Murdoch, W. J.; Radosz, M.; Shen, Y. Macromol. Biosci. 2007, 7, 1187–1198. Lutz, J.-F.; Hoth, A. Macromolecules 2006, 39, 893–896. Lutz, J.-F.; Akdemir, Ö.; Hoth, A. J. Am. Chem. Soc. 2006, 128, 13046–13047. Hua, F.; Jiang, X.; Li, D.; Zhao, B. J. Polym. Sci., Part A: Polym. Chem. 2006, 44, 2454–2467. Li, D.; Jones, G. L.; Dunlap, J. R.; Hua, F.; Zhao, B. Langmuir 2006, 22, 3344–3351. Bhatnagar, B. S.; Martin, S. M.; Teagarden, D. L.; Shalaev, E. Y.; Suryanarayanan, R. J. Pharm. Sci. 2010, 99, 4510–4524. Takei, T.; Kurosaki, K.; Nishimoto, Y.; Sugitani, Y. Anal. Sci. 2002, 18, 681–684. Takei, T.; Sugitani, Y. Anal. Sci. 2010, 26, 337–341. Tanaka, M.; Mochizuki, A. J. Biomater. Sci., Polym. Ed. 2010, 21, 1849–1863. Hatakeyama, T.; Tanaka, M.; Hatakeyama, H. J. Biomater. Sci., Polym. Ed. 2010, 21, 1865–1875. 104

Inorganic/Hybrid Materials

Chapter 6

Synthesis of Calcium Phosphate Microspheres Using an Ultrasonic Spray–Pyrolysis Technique and Their Application as Novel Anti-Angiogenic Chemoembolization Agents for Cancer Treatment Mamoru Aizawa,*,1 Michiyo Honda,1 and Makoto Emoto2 1Department

of Applied Chemistry, School of Science and Technology, Meiji University, 1-1-1, Higashimita, Tama-ku, Kawasaki, Japan 2Department of Health and Welfare and Division of Preventive Medicine, Fukuoka Sanno Hospital, International University of Health and Welfare, 1-7-4, Momochihama, Sawara-ku, Fukuoka, Japan *E-mail: [email protected]

The purpose of the present study was to develop a novel process for chemoembolization to improve the therapeutic effectiveness and safety profile of cancer treatment. A chemoembolization approach was designed for treating human solid tumors using biodegradable calcium phosphate microspheres (hereafter, CPMs) prepared using an ultrasonic spray–pyrolysis technique combined with an anti-angiogenic agent (TNP-470; Takeda, Japan) that inhibits tumor vasculature formation in vivo. FU-MMT-3 human uterine sarcoma cells were used in this study because this type of tumor is aggressive and responds poorly to radiotherapy and currently used chemotherapy agents. In this chapter, preparation of biodegradable CPMs and their powder properties, including drug release characteristics, will be reviewed. We performed biological evaluations both in vitro and in vivo using CPMs loaded with TNP-470. The results of these tests indicated that microspheres loaded with TNP-470 inhibit i) the proliferation of FU-MMT-3 cells in a tumor model and ii) tumor enlargement in a model of nude mice injected with FU-MMT-3 cells. Biological evaluations demonstrated that

© 2017 American Chemical Society

CPMs loaded with TNP-470 exhibit excellent anti-tumorigenic effects. In addition to the above, this chapter will discuss i) the effects of particle size and CPM distribution on anti-angiogenic chemoembolization, ii) the creation of porous advanced carrier CPMs with nano-pores on the surface prepared via salt-assisted ultrasonic spray–pyrolysis, and iii) our future work.

Introduction Hydroxyapatite (Ca10(PO4)6(OH)2; HAp) and β-tricalcium phosphate (β-Ca3(PO4)2; β-TCP) are widely used as biomaterials substituting for human hard tissues (1). We previously synthesized HAp and other apatite-family compounds using an ultrasonic spray–pyrolysis (USSP) technique and examined the properties of the resulting powders (2–5). USSP techniques can be used to prepare powders via the liquid phase. USSP has the advantage of enabling the instantaneous stoichiometric preparation of homogeneous compounds by spraying solutions containing desired amounts of cations into the hot zone of an electric furnace (4, 5). Moreover, USSP is simple, provides a narrow particle size distribution, and can be used to prepare hollow spherical particles. Using USSP, we have thus far synthesized hollow biphasic calcium phosphate microspheres (hereafter, CPMs) consisting of β-TCP and calcium-deficient hydroxyapatite (Ca10-x(PO4)6-x(OH)2-x•nH2O; DAp) (6). In the present investigation, we explored the use of the resulting hollow biphasic microspheres as carriers for a drug delivery system (DDS) applicable to the medical treatment of cancers. Cancer is a serious worldwide problem. Since 1980, cancer has been the leading cause of mortality in Japan. Although uncommon, sarcomas (including carcinosarcomas) are the most aggressive neoplasms among uterine malignancies (7, 8). These tumors respond poorly to radiotherapy and any of the currently used chemotherapeutic agents with substantial toxic effects (7–10). The overall 5-year survival rate for all stages of uterine sarcomas is less than 40%, which is significantly lower than that for other uterine cancers (7). Except for the early stages, no standard treatment for these tumors has been developed; thus, new therapeutic strategies are urgently needed. Recent studies, including those of the authors, have shown that rapid growth and early metastasis of uterine sarcomas might be associated with high angiogenic potential in comparison with other uterine malignancies (10–12). The recent introduction of anti-angiogenesis therapies into clinical practice represented a turning point in cancer therapy (13). Emoto et al. reported that the agent TNP-470 (Takeda, Japan) is effective for anti-angiogenic therapy against highly aggressive human uterine carcinosarcomas both in vitro and in vivo (14–16). TNP-470 is a low-molecular-weight synthetic analogue of fumagillin, a natural compound secreted by Aspergillus fumigatus (17), and inhibits angiogenesis via endothelial cell cycle arrest in the late G1 phase (18). Methionine aminopeptidase2 has been identified as a molecular target of TNP-470 (19). TNP-470 exhibits antitumor activity against various human malignancies both in vitro and in vivo (20–24). 108

Radiotherapy is currently one of the most effective means of treating cancers. For deep-seated cancers, however, external irradiation provides only small doses and often damages healthy tissues. Therefore, chemoembolization approaches that employ biomaterials loaded with anti-angiogenic agents may be effective for treating deep-seated solid tumors. β-TCP is a well-known biodegradable calcium phosphate ceramic material. Some studies have shown that TCP is an ideal biomaterial because it dissolves gradually without any cytotoxic or allergenic effects and elicits no immune response (25–27). Thus, we selected biodegradable, hollow CPMs as carriers for a DDS suitable for cancer treatment. Our goal is to develop novel chemoembolization processes to improve the therapeutic effectiveness and safety profile of cancer treatments. This chapter will discuss a chemoembolization approach designed for use in treating human solid tumors. The approach involves combining biodegradable, hollow CPMs prepared using a USSP technique and the anti-angiogenic agent TNP-470, which can inhibit tumor vasculature formation in vivo (28, 29). Figure 1 illustrates the process using a modification of the image reported by Kawashita et al. (30). In addition, we will discuss the effect of CPM particle size and distribution on anti-angiogenic chemoembolization (31) and the production of porous CPMs with nano-pores on the surface prepared via a salt-assisted USSP (SAUSP) approach for use as an advanced carrier (32–34), together with our future work.

Figure 1. Our novel chemoembolization approach involved combining biodegradable hollow CPMs prepared using an ultrasonic spray–pyrolysis technique with an anti-angiogenic agent (TNP-470) that inhibits tumor vasculature formation. 109

Preparation of CPMs Containing an Anti-Angiogenic Agent and Their Application in Chemoembolization CPMs were prepared for anti-tumorigenesis experiments as previously reported (6, 28, 29). The starting solution, with a Ca/P ratio of 1.50, was prepared by mixing Ca(NO3)2, (NH4)2HPO4, and HNO3 to final concentrations of 0.60, 0.40, and 0.10 mol·dm−3, respectively. As shown in Figure 2(a), the lower-furnace temperature for drying the generated droplets was fixed at 300ºC, and the upper-furnace temperature for pyrolysis of the dried droplets was set at 850ºC. The solution was sprayed into the heating zone using an ultrasonic vibrator operated at a frequency of 2.4 MHz, and then the sprayed droplets were dried and pyrolyzed to form the CPMs, as shown in Figure 2(b). The resulting microspheres were washed with pure water and freeze dried to prepare the “washed powder” for analyses of anti-tumorigenic effects.

Figure 2. Synthesis of CPMs and their powder properties. (a) Overview of the USSP apparatus, (b) model diagram of CPM formation, (c) SEM image, and (d) TEM image.

The results of X-ray diffraction (XRD) analyses demonstrate that the resulting CPMs were biphasic, composed of β-TCP (~50 mass%) and CDAp (~50 mass%). Both β-TCP and DAp are well-known biodegradable ceramics. The Ca/P molar ratio, as determined by X-ray fluorescence spectrometry, was 1.49, in good agreement with the nominal composition of the starting solution. Although the powder before washing contained NO3− groups derived from Ca(NO3)2 in the starting material, the results of Fourier-transform infrared spectroscopy analyses indicated that the NO3− groups were removed by washing. The specific surface area (SSA) of the powder increased from ~10 m2·g−1 before washing to ~20 m2·g−1 after washing. Figure 2(c) and (d) show the particle morphology of the washed 110

powder as determined by scanning electron microscopy (SEM) and transmission electron microscopy (TEM), respectively. Figure 2(c) shows that the powder was composed of microspheres with a diameter of about 1 µm. When observed by TEM, the microspheres appeared transparent, suggesting that they were hollow. We then performed biological evaluations of the resulting hollow CPMs. The anti-angiogenic agent TNP-470 was dissolved in ethanol at 100, 500, and 1000 ppm (control, 0 ppm). TNP-470 was loaded by adding microspheres (0.05 g) to the ethanol solutions (1 cm3), followed by freeze drying. Microspheres loaded with TNP-470 were used as sample powders for in vitro and in vivo evaluations of anti-tumorigenic activity. In order to maintain the stability of TNP-470 in the medium used for cell culture, we added a medium-chain triglyceride (MCTG, ncaprylic acid) to some of the sample powders (35). The drug release profile of the TNP-470–loaded TCP microspheres was examined over a period of 25 h. The amount of TNP-470 in the supernatant was determined by high-performance liquid chromatography (HPLC). Seventy-five percent of the total TNP-470 was released relatively rapidly from the microspheres, within 30 min, and the remaining 25% was released more slowly up to 25 h following immersion, after which no TNP-470 was detected in the supernatant. Five samples of CPMs without TNP-470 and loaded with TNP-470 were evaluated, with a polystyrene cell culture plate serving as a control: i) microspheres (washed powder), ii) microspheres loaded with 1000 ppm TNP-470, iii) microspheres loaded with 500 ppm TNP-470/1 v/v% MCTG, iv) microspheres loaded with 100 ppm TNP-470, and v) microspheres loaded with 100 ppm TNP-470/1 v/v% MCTG. The anti-angiogenic activity of the microspheres was examined using FU-MMT-3 cells, which were established by Emoto et al. as a tumor model (36). A total of 5 × 105 cells were seeded in a 12-well plate and cultured for 1 day, and then the five different powder samples were set on Transwell® (Corning) membranes dipped in D-MEM/F12 (10% FBS) and cultured at 37°C in a 5% CO2 atmosphere for an additional 1 or 3 days, as illustrated in Figure 3(a). Cell proliferation was examined by counting cells using an erythrocytometer, and cell morphology was observed by phase-contrast microscopy. Powder compacts were prepared using microspheres loaded with various concentrations of TNP-470, and then the cell-culture test was performed. FU-MMT-3 cells were seeded in a 12-well plate and cultured for 1 d, and then the powder compacts were placed on the Transwell® membranes and cultured for up to 3 additional days. The results are shown in Figure 3(b). The cells proliferated in the control and powder compacts without TNP-470; however, cell proliferation was inhibited by the powder compacts of microspheres containing 100 ppm TNP-470 (Figure 3(b)). The number of dead cells increased with increasing TNP-470 concentration. Notably, the number of proliferating cells decreased drastically in the case of powder compacts of microspheres containing 500 and 1000 ppm TNP-470. The observed inhibition of cell growth is believed to have been due to the release of TNP-470 into the medium. Figure 3(c) and (d) show the morphologies of cells cultured with control and CPMs loaded with 1000 ppm TNP-470 after 3 days, respectively. In Figure 3(d), we can observe many the rounded-shaped cells. The excess TNP-470 may express the cytotoxicity. 111

Figure 3. In vitro evaluation of CPMs loaded with TNP-470. (a) Testing of cytotoxicity using Transwell® membranes, (b) cell proliferation, and (c, d) morphologies of cells after cell culture for 3 days; (c) Control, (d) CPMs loaded with 1000 ppm TNP-470.

Next, we examined the in vivo anti-tumorigenic activity of the microspheres using nude mice harboring tumors resulting from injection of FU-MMT-3 cells (37). To prepare the animal model, 6 × 104 FU-MMT-3 cells were injected subcutaneously into nude mice (weight ~20 g each), and the mice were then bred for 1 week. Washed powder and microspheres loaded with 1000 ppm TNP-470 (each 25 mg) were suspended in physiologic saline (1 cm3), and then 0.4 cm3 of the resulting suspension was injected around the tumor sites. The size of the tumors was measured using calipers over a period of 8 weeks and compared with control mice (no treatment) and mice injected with TNP-470 only. Based on the in vitro findings, we performed in vivo evaluations using microspheres prepared with 1000 ppm TNP-470. First, FU-MMT-3 cells were injected subcutaneously into nude mice (weight ~20 g) to induce the formation of tumors. After 1 week, microspheres loaded with TNP-470 were suspended in physiologic saline and then injected around the tumors. Figure 4(a) shows the change in tumor size over time. The control in Figure 4(a) refers to the case of no treatment after injection of FU-MMT-3 cells. The tumor size increased only minimally in the mice treated with microspheres loaded with TNP-470, in clear contrast to the cases of the control, injection of TNP-470 alone, and injection of empty microspheres alone. These results indicate that CPMs containing TNP-470 exhibit anti-tumorigenic effects. 112

Figure 4. In vivo evaluation of CPMs loaded with TNP-470. (a) Change in tumor volume over a period of 8 weeks. (b, c) Typical histologic images.

Histologic evaluations revealed that the CPMs were embolized in the feeding arteries of the xenografts in both treatment groups, with or without TNP-470. Fibrosis with granuloma formation and lymphocytic infiltration was observed in the lumen of the feeding arteries; no injury caused by microspheres was found in the vessel walls (Figure 4(b)). The TCP microspheres appeared as amorphous basophilic material upon hematoxylin and eosin staining and exhibited marked embolization in tumor microvessels in all mice of the CPM treatment groups (Figure 4(c)). Destruction of tumor microvessels and areas of coagulative necrosis were observed in tumors treated with TCP microspheres, with or without TNP-470. No significant loss of body weight was observed in any of the mice treated with TNP-470–loaded TCP microspheres compared with mice treated with TNP-470 alone. The mean body weight at 56 days in mice treated with TNP-470 alone, however, was 16.9 ± 2.55 g, which was significantly lower than that of mice treated with TNP- 470–loaded microspheres (25.17 ± 0.68 g) (p = 0.01). As indicated above, CPMs consisting of biphasic biodegradable β-TCP and DAp were used in the present study both as drug carriers and as an embolization material. In vitro, TNP-470 may be released in a sustainable two-step manner from the internal space and external surface of the hollow microspheres. Although the powder compacts of CPMs alone did not inhibit the proliferation of FU-MMT-3 cells in vitro (Figure 3), treatment with microspheres alone significantly inhibited the growth of the FU-MMT-3 xenografts in vivo, compared with the control (Figure 4). These results suggest that although CPMs alone do not inhibit sarcoma cell proliferation, a direct anti-tumor effect associated with this material occurs in vivo as a result of embolization. This is supported by histopathologic analyses, 113

which revealed marked embolization of microspheres in the feeding arteries of the xenografts, as well as in many tumor microvessels (Figure 4). This strong embolization effect exhibited by the CPMs alone is potentially useful for cancer treatment. In summary, a novel chemoembolization approach was developed for treating human solid tumors by combining biodegradable CPMs with an anti-angiogenic agent to inhibit tumor vasculature formation in vivo.

Effect of CPM Particle Size and Distribution on Anti-Angiogenic Chemoembolization In the previous section, we described the initial design of a novel biomaterial composed of resorbable hollow ceramic microspheres loaded with TNP-470 to target the tumor vasculature. We demonstrated the usefulness and safety of this approach using a highly aggressive human uterine sarcoma xenograft model. In the next step in the development of this new chemoembolization approach using CPMs, we designed a new type of anti-angiogenic microspheres as an advanced model to target (or measure) vascular heterogeneity in highly aggressive and highly angiogenic solid tumors. In this section, we describe the effect of CPM particle size and distribution on the anti-tumorigenic effect of CPMs loaded with TNP-470.

Table 1. Preparation and powder properties of CPMs of different sizes loaded with TNP-470.a Sample name 1510(-) 1510(+) 6040(-) 6040(+) 9060(-) 9060(+)

CPM (Ca: 0.15, PO4: 0.10)

(Ca: 0.60, PO4: 0.40)

(Ca: 0.90, PO4: 0.60)

Mix(-) Mix(+)

TNP-470 Non-loaded Loaded Non-loaded Loaded Non-loaded Loaded Non-loaded

(Ca: 0.15, PO4: 0.10) (Ca: 0.60, PO4: 0.40) (Ca: 0.90, PO4: 0.60)

Loaded

Crystalline phases

Particle sizec

46% CDApb + 54% β-TCP

1.8 µm

33% CDApb + 67% β-TCP

2.6 µm

26% CDApb + 74% β-TCP

3.0 µm

Mixed powder prepared from the above three CPMs (1:1:1 [w/w/w])

1.8 µm + 2.6 µm + 3.0 µm

(Ca:0.60, PO4:0.40) → Ca2+: 0.60 mol/dm3, PO43-: 0.40 mol/dm3. deficient hydroxyapatite. c Particle size: median size.

a

b

CDAp: Calcium-

CPMs with different particle sizes (i.e., 1510[−], 6040[−], and 9060[−]), were prepared by varying the concentration of Ca2+ (0.15, 0.60, 0.90 mol·dm−3) and PO43− (0.10, 0.40, 0.60 mol·dm−3) ions in the starting solution, as previously reported (6, 28, 29). The resulting microspheres were washed with pure water and 114

then freeze dried. Next, TNP-470 was dissolved in ethanol to a concentration of 2000 ppm then loaded by adding the microspheres (0.25 g) to the above-mentioned ethanol solution (1 cm3), followed by freeze drying. Microspheres loaded with TNP-470 (i.e., 1510[+], 6040[+], and 9060[+]) were used as sample powders for in vivo evaluations of anti-tumorigenic effects. Table 1 summarizes the preparation of the CPMs with and without TNP-470, together with their powder properties: crystalline phases and median diameter. Figure 5 shows the particle morphology of the CPMs, together with the release profile of TNP-470 from the CPMs over a period of 120 h. The SEM micrographs show that the resulting powders were composed of spherical particles. The diameter of the particles decreased with decreasing concentration of the starting solution. The spherical particles were formed via the following process: (i) removal of the solvent from the droplet surface, (ii) formation of microcrystalline calcium phosphates, and (iii) growth of calcium phosphate crystals. The diameter of the resulting spherical particles is dependent upon the starting solution droplet size. The droplet size can be decreased by decreasing the concentration of the starting solution, which in turn reduces the diameter of the spherical particles. The release profile data showed that 90% of the total TNP-470 was released from the CPMs slowly over a period of 20 h. The amount of TNP-470 released from the 1510(+) powder was the highest among the examined sample powders.

Figure 5. Profile of drug release from CPMs loaded with TNP-470, together with typical CPMs composed of various particle sizes. 115

Figure 6 shows the weekly change in mean tumor volume during CPM injection using the FU-MMT-3 xenograft model. Figure 6(a) illustrates the effect of CPM particle size on the anti-tumorigenic effect, with and without TNP-470. Microspheres with a smaller particle size inhibited tumor growth more effectively. Figure 6(b) illustrates the effect of particle size distribution on the anti-tumorigenic effect of CPMs with and without TNP-470. The particle morphology of Mix(+) is inserted into Figure 6(b). Both the Mix(+) and 6040(+) treatments significantly inhibited the growth of FU-MMT-3 xenografts in comparison with the controls (p < 0.05). In addition, the Mix(+) treatment significantly suppressed tumor growth in comparison with the 6040(+) and 9060(+) treatments (p < 0.05). No significant loss of body weight was observed in any mouse treated with any size of TNP-470–loaded CPMs. Histopathologic analyses indicated marked embolization of the CPMs in the feeding arteries in the peripheral areas of the xenografts in vivo, as well as in many of the tumor microvessels.

Figure 6. Change in tumor volume over a period of 8 weeks after injection of CPMs of different sizes loaded with TNP-470; the effect of (a) CPM particle size and (b) particle size distribution on the anti-tumorigenic effect with and without TNP-470 agent, together with particle morphology of “Mix(-)” powder. In order to adapt our biomaterial for chemoembolization, in the present study, the biodegradable hollow CPMs were further refined to target morphologic heterogeneity in the vasculature. The in vitro results showed that the amount of TNP-470 released from the 1510(+) powder was the highest among the examined sample powders. The 1510(−) powder consists of microspheres of the smallest diameter; therefore, this powder has the largest specific surface area (20 m2/g) in comparison with the 6040(−) (14.4 m2/g) and 9060(−) (16.3 m2/g) powders. Thus, the 1510(−) powder can accommodate the greatest amount of TNP-470. As the 1510 powder consists of microspheres of the smallest diameter, the 1510(+) solution would have the largest microsphere surface area; thus, this solution could 116

incorporate a greater amount of TNP-470 in comparison with the 6040(+) and 9040(+) powders. As TNP-470 is released from the internal space and external surface of the hollow CPMs in a two-step sustainable fashion in vitro, the smaller microspheres would theoretically be the most useful drug carriers. However, the in vivo results of the present study showed that the Mix(+) treatment tended to achieve a better suppression of FU-MMT-3 tumor growth compared with the 1510(+) treatment. This result suggests that as the feeding arteries and tumor microvessels vary in diameter (they exhibit vascular morphologic heterogeneity), these vessels might have been more effectively embolized by the different sizes of CPMs or their aggregates in the Mix(+) treatment. Additionally, the diameter of the CPMs was approximately 0.5-3 μm, and there was no evidence of blood vessel–related injury, such as a marked hemorrhaging or hematoma formation, in any of the treated mice. Thus, this new DDS using a mixture of CPMs of different sizes might be a useful approach for chemoembolization in the treatment of many solid tumors.

Porous CPMs with Nano-Pores on the Surface Prepared via SAUSP as Advanced Carriers, and Our Future Work In a previous study, we proposed that CPMs loaded with TNP-470 would be effective DDS carriers for novel chemoembolization treatment of cancers, based on the results of in vivo experiments (28, 29, 31). However, in vitro, the microspheres used in those studies released 80% of the total TNP-470 within 1 h. In order to achieve controlled release of TNP-470, we prepared novel microspheres using a SAUSP approach with NaCl as the salt and attempted to form nano-pores on the surface of the microspheres (33, 38). As the SAUSP technique was originally developed to prepare nano-dispersive powders (39), we clarified the optimal concentration of salt at which nano-pores form on the surface of the microspheres. Unfortunately, the resulting microspheres exhibited several problems: (i) decreased solubility, and (ii) cytotoxicity due to the elution of chloride ions. In order to overcome these problems, in the present study, we prepared novel microspheres using a SAUSP approach with potassium nitrate (KNO3) serving as the salt instead of NaCl. In addition, we examined the efficacy of the resulting hollow microspheres as DDS carriers for the treatment of cancers using the anti-angiogenic agent TNP-470 as a model drug (37). We then compared the drug release profile of the novel microspheres with nano-sized pores on the surface with that of traditional microspheres lacking nano-sized pores. Starting solution with a Ca/P ratio of 1.50 was prepared by mixing Ca(NO3)2·4H2O, (NH4)2HPO4 and HNO3 to final concentrations of 0.60, 0.40, and 0.10 mol·dm−3. The concentration of KNO3 in the starting solution was in the range 0.20 to 1.20 mol·dm−3. The upper- and lower-furnace temperatures were fixed at 850 and 300°C, respectively. These solutions were sprayed into the heating zone using an ultrasonic vibrator operated at a frequency of 2.4 MHz, and then the sprayed droplets were dried and pyrolyzed to form CPMs incorporating a KNO3 phase. The resulting microspheres were washed with pure water and freeze dried to prepare the washed powders. 117

XRD analyses indicated that the unwashed powders with KNO3 at a concentration of 0.20 to 1.0 mol·dm−3 were composed of a mixture of KNO3 and HAp with low crystallinity, whereas the washed powders consisted only of an HAp phase with low crystallinity. This result shows that KNO3 used in SAUSP technique can be removed by the washing process. In the case of powder prepared with KNO3 at a concentration of 1.20 mol·dm−3, the XRD patterns indicated that the unwashed powder consisted of HAp with low crystallinity, KNO3, and KCaPO4. The washed powder, by contrast, was composed of HAp with low crystallinity and KCaPO4. Figure 7 shows the typical particle morphology of the washed powders, together with a schematic illustration of the formation of porous CPMs with nano-pores on the surface prepared using the SAUSP approach. In addition, SEM micrographs indicated that the washed powders were composed of microspheres with a diameter of 0.5-3.0 µm. Upper part in Figure 7 shows a high-magnification image of the particle morphology of powders prepared with KNO3 at a concentration of 1.00 mol·dm−3. Small pores with a sizes of ~50 nm were present on the surface of the microspheres. However, in the case of powders prepared without KNO3 addition, these small pores were not present on the surface of the microspheres, as shown in lower part in Figure 7.

Figure 7. Particle morphologies of CPMs with nano-sized pores on the surface (upper part), together with a model illustrating the formation process (lower one); 1) Formation of droplet, 2) decrease of droplet volume by evaporating the solve, 3) Formation of highly-viscus layer on the droplet surface, 4) Formations of calcium phosphate and added salt particles, 5) Formations of hollow spherical agglomerates (secondary particle), and 6) Formation of fragment particles by bursting the hollow particle. 118

TNP-470 was dissolved in ethanol to a concentration of 2000 µg/cm3 and loaded by adding microspheres (0.06 g) to the ethanol solution (12 cm3), followed by freeze drying for 24 h. The TNP-470 release profile was determined using HPLC. Figure 8 shows the drug release profile for microspheres loaded with TNP470. In the case of powders prepared without KNO3, 90% of the total amount of TNP-470 was released from the microspheres within 1 h, and total amount of TNP470 released from the microspheres was 20 µg·mg−1. For powders prepared with KNO3 at a concentration of 1.00 mol·dm−3 (Figure 7), 70% of the total TNP-470 was released from the microspheres within 1 h. The remaining 30% was slowly released up to 24 h following immersion, and the total amount of TNP-470 released from these microspheres was 42 µg·mg−1. The amount of drug loaded depended on the SSA; of the sample powders examined, the washed powders prepared using KNO3 at a concentration of 1.00 mol·dm−3 released the greatest amount of drug. Notably, microspheres with nano-sized pores on the surface that were prepared using KNO3 at a concentration of 1.00 mol·dm−3 exhibited a two-step release profile, perhaps due to release of the drug from the surface and interior of the microspheres, as illustrated in Figure 1.

Figure 8. Two-step release profiles of drug from TNP-470–loaded CPMs with nano-sized pores on the surface, together with typical particle morphologies. We evaluated the effect of CPMs loaded with TNP-470 described above upon human uterine sarcoma cells (FU-MMT-3 (36)) transplanted into nude mice. Although the detailed data will be reported elsewhere as a future work, we report here that compared with control mice, the size of the tumor increased only minimally in mice treated with the CPMs loaded with TNP-470. In addition, the lifespan of mice treated with TNP-470–loaded CPMs prepared with KNO3 increased relative to the control mice. Here, “control” refers to no treatment 119

after injection of FU-MMT-3 cells. In addition, histologic analyses indicated that many of the CPMs remained in the tumor blood vessels. Collectively, these results show that CPMs with nano-sized pores loaded with TNP-470 may be effective for advanced chemoembolization.

Conclusion Our purpose was to develop a novel process for chemoembolization to improve the therapeutic effectiveness and safety profile of cancer treatment. This study showed that calcium phosphate ceramic microspheres loaded with TNP-470 inhibit the growth of human uterine sarcoma in vivo via a physiochemical mechanism. This new chemoembolization method incorporating an anti-angiogenic agent may contribute to the development of more effective treatments for locally advanced or recurrent solid tumors.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

13.

14. 15.

Hench, L. L. J. Am. Ceram. Soc. 1991, 74, 1487–1510. Aizawa, M.; Howell, F. S.; Itatani, K. J. Ceram. Soc. Jpn. 1999, 107, 1007–1011. Aizawa, M.; Hanazawa, T.; Itatani, K.; Howell, F. S.; Kishioka, A. J. Mater. Sci. 1999, 34, 2865–2873. Aizawa, M.; Itatani, K.; Okada, I. Phosphorus Res. Bull. 2006, 20, 61–78. Nishikawa, H.; Honda, M.; Yokota, T.; Shimizu, Y.; Aizawa, M. J. Nanomater. 2016, 6753203. Ohno, T.; Aizawa, M. Key Eng. Mater. 2006, 309−311, 235–238. Salazar, O. M.; Bonfiglio, T. A.; Patten, S. F. Cancer 1978, 42, 1152–1160. Spanos, W. J., Jr.; Wharton, J. T.; Gomez, L.; Fletcher, G. H.; Oswald, M. J. Cancer 1984, 15, 311–316. Acharya, S.; Hensley, M. L.; Montag, A. C.; Flemming, G. F. Lancet Oncol. 2005, 6, 961–71. Rose, P. G.; Piver, M. S.; Tsukada, Y.; Lau, T. Cancer 1989, 63, 935–938. Emoto, M.; Iwasaki, H.; Ishiguro, M.; Kikuchi, M.; Horiuchi, S.; Saito, T.; Tsukamoto, N.; Kawarabayashi, T. Hum. Pathol. 1999, 30, 1232–1241. Emoto, M.; Charnock-Jones, D. S.; Licence, D.; Ishiguro, M.; Kawai, M.; Yanaihara, A.; Saito, T.; Hachisuga, T.; Iwasaki, H.; Kawarabayashi, T.; Smith, SK. Gynecol. Oncol. 2004, 95, 474–482. Willett, C. G.; Boucher, Y.; di Tomaso, E.; Duda, D. G.; Munn, L. L.; Tong, R. T.; Chung, D. C.; Sahani, D. V.; Kalva, S. P.; Kozin, S. V.; Mino, M.; Cohen, K. S.; Scadden, D. T.; Hartford, A. C.; Fischman, A. J.; Clark, J. W.; Ryan, D. P.; Zhu, A. X.; Blaszkowsky, L. S.; Chen, H. X.; Shellito, P. C.; Lauwers, G. Y.; Jain, R. K. Nat. Med. 2004, 10, 145–147. Miura, S.; Emoto, M.; Matsuo, Y.; Kawarabayashi, T.; Saku, K. Cancer Lett. 2004, 203, 45–50. Emoto, M.; Ishiguro, M.; Iwasaki, H.; Kikuchi, M.; Kawarabayashi, T. Gynecol. Oncol. 2003, 89, 88–94. 120

16. Emoto, M.; Tachibana, K.; Iwasaki, H.; Kawarabayashi, T. Cancer Sci. 2007, 98, 929–935. 17. Ingber, D.; Fujita, T.; Kishimoto, S.; Sudo, K.; Kanamaru, T.; Brem, H.; Folkman, J. Nature 1990, 348, 555–557. 18. Zhang, Y.; Griffith, E. C.; Sage, J.; Jacks, T.; Liu, J. O. Proc. Natl. Acad. Sci. U. S. A. 2000, 97, 6427–6432. 19. Sin, N.; Meng, L.; Wang, M. Q.; Wen, J. J.; Bornmann, W. G.; Crews, C. M. Proc. Natl. Acad. Sci. U. S. A. 1997, 94, 6099–6103. 20. Yanase, T.; Tamura, M.; Fujita, K.; Kodama, S.; Tanaka, K. Cancer Res. 1993, 53, 2566–2570. 21. Yoshida, T.; Kaneko, Y.; Tsukamoto, A.; Han, K.; Ichinose, M.; Kimura, S. Cancer Res. 1998, 58, 3751–3756. 22. Inoue, K.; Chikazawa, M.; Fukata, S.; Yoshikawa, C.; Shuin, T. Clin. Cancer Res. 2002, 8, 2389–2398. 23. Inoue, K.; Chikazawa, M.; Fukata, S.; Yoshikawa, C.; Shuin, T. Clin Cancer Res. 2003, 9, 886–899. 24. Nahari, D.; Satchi-Fainaro, R.; Chen, M. Mol. Cancer Ther. 2007, 6, 1329–1337. 25. Yuan, H.; Kurashina, K.; de Bruijn, J. D.; Li, Y.; de Groot, K.; Zhang, X. Biomater. 1999, 20, 1799–1806. 26. Nagayama, M.; Takeuchi, H.; Doi, Y. Dent. Mater. J. 2006, 25, 219–225. 27. Victor, S. P.; Sampath Kumar, T. S. J. Mater. Sci.: Mater. Med. 2008, 19, 283–290. 28. Aizawa, M.; Ohno, T.; Kanomata, N.; Yano, K.; Emoto, M. Key Eng. Mater. 2008, 361−363, 1215–1218. 29. Emoto, M.; Naganuma, Y.; Choijamts, B.; Ohno, T.; Yoshihisa, H.; Kanomata, N.; Aizawa, M. Cancer Sci. 2010, 101, 984–990. 30. Kawashita, M.; Shineha, R.; Kim, H. M.; Kokubo, T.; Inoue, Y.; Araki, N.; Nagata, Y.; Hiraoka, M.; Sawada, Y. Biomaterials 2003, 24, 2955–2963. 31. Emoto, M.; Yoshihisa, H.; Yano, K.; Choijamts, B.; Tsugu, H.; Tachibana, K.; Aizawa, M. Anticancer Res. 2015, 35, 4757–4764. 32. Matsueda, M.; Emoto, M.; Aizawa, M. Key Eng. Mater. 2012, 493-494, 672–677. 33. Kawanobe, Y.; Honda, M.; Konishi, T.; Mizumoto, M.; Habuto, Y.; Kanzawa, N.; Uchino, T.; Aizawa, M. J. Aust. Ceram. Soc. 2010, 46, 6–10. 34. Aizawa, M.; Honda, M.; Emoto, M. J. Jpn. Soc. Biomater. 2015, 33-4, 314–321 (in Japanese). 35. Kakinoki, S.; Yasuda, C.; Kaetsu, I.; Uchida, K.; Yukutake, K.; Nakayama, M.; Fujie, S.; Kuroda, D.; Kato, M.; Ohyanagi, H. Eur. J. Pharm. Biopharm. 2003, 55, 155–160. 36. Emoto, M.; Iwasaki, H.; Oshima, K.; Kikuchi, M.; Kaneko, Y.; Kawarabayashi, T. Virchows Arch. 1997, 431, 249–256. 37. Emoto, M.; Ishiguro, M.; Iwasaki, H.; Kikuchi, M.; Kawarabayashi, T. Anticancer Res. 2000, 20, 601–604. 38. Matsueda, M.; Yoshihisa, H.; Emoto, M.; Aizawa, M. Arch. BioCeram. Res. 2010, 10, 137–140. 39. Itoh, Y.; Okuyama, K. J. Ceram. Soc. Jpn. 2003, 111, 815–820. 121

Chapter 7

Bioinspired Design and Engineering of Functional Nanostructured Materials for Biomedical Applications Xin Ting Zheng,1,4 Hesheng Victor Xu,1,2,4 and Yen Nee Tan1,3,* 1Institute

of Materials Research and Engineering (IMRE), Agency for Science, Technology and Research (A*STAR), 2 Fusionopolis Way, Singapore 138634 2Division of Chemical and Biological Chemistry, School of Physical and Mathematical Sciences, Nanyang Technological University, 21 Nanyang Link, Singapore 637371 3Department of Chemistry, National University of Singapore, 3 Science Drive, Singapore 117543 *E-mail: [email protected] 4These authors contributed equally.

Nature has provided many ways to derive various functional materials with highly-ordered hierarchical structures and superb attributes from the sophisticated biological processes. Inspired by natural biomineralisation process, it has led to the emergence of four “bioinspired” strategies, i.e., bio-structure mimicking, bio-function anchoring, bio-templating and bio-assembling, to construct nanostructured materials with remarkable biomimetic properties. In this chapter, we will highlight the development of bioinspired approaches involving biomolecules and elucidate their roles in directing the bottom-up synthesis and programmable assembly of functional nanostructured materials. Their recent applications in diagnostics and therapeutic delivery will also be discussed. Finally, we will conclude this chapter with the challenges and future outlook of these bioinspired nanomaterials for the advanced biomedical applications such as theranostics.

© 2017 American Chemical Society

1. Introduction In today’s rapid advancing nanotechnological world, the synthesis of nanomaterials with superior qualities is highly imperative. Nonetheless, the preparation of nanoscale materials is very challenging. Fortunately, Nature has provided mankind with many hints and clues on how to develop complex functional materials. Biominerals, a type of organic-inorganic hybrid materials, possess a highly-ordered hierarchical structures with extraordinary physiochemical features including high flexibility, light-weight, exceptional mechanical strength and dynamic functions (1, 2). For instance, hydroxyapatite in bones and teeth of mammals, amorphous silica in diatoms and marine sponges, as well as as magnetite in chiton teeth (3–5), are the essential components in living organisms which support their important physiological functions. These biominerals are synthesized through an intelligent biological process called biomineralisation, whereby functional proteins facilitate the conversion of natural minerals into biominerals (6, 7). Although direct extraction to obtain these biominerals seems attractive, it involves tedious multistep processing such as identification, isolation and purification thus limiting its effectiveness (8, 9). Furthermore, these biominerals vary widely in their compositions and structures which might not be suitable for direct application. Instead, by studying the biological system and mimicking their forming mechanism, it would provide the guiding principles and further inspire us to design and develop nanomaterials with similar features. As such, this has led to the emergence of “bioinspired” approaches to fabricate nanomaterials with exceptional properties. Scheme 1 summarizes the general design strategies of bioinspired nanomaterials through bio-structure mimicking, bio-function anchoring, bio-templating and bio-assembling approaches. The “bio-structure mimicking” approach requires the profound understanding of the structure-function relationship of the natural materials as well as their physical and chemical principles, to enable the fabrication of bioinspired nanomaterials through replication of the nanoscale architecture of natural materials (10, 11). In this way, the synthetic nanostructured materials could mimic the functions of the nature made materials. Representative examples of bio-structure mimicking include the optical property of butterfly wing, adhesiveness of gecko foot and self-cleansing features of lotus leaves (12, 13). “Bio-function achoring” is another approach to render the nanostructured materials with desirable biological functions by chemically conjugating a specific biomolecule onto the as-synthesized nanomaterials. For example, specific proteins or carbohydrates could be anchored onto the nanomaterials to provide them with the unique bio-recognition and/or enzymatic ability that are inherited from the anchored biomolecular shell, and at the same time enhance the solubility and biocompatibility of the bio-nanocomposites (23, 24). “Bio-templating” is a synthetic approach of using biomolecule as a template to facilitate or direct the gowth of nanomaterials into various sizes and shapes (25–27). In a typical bio-templated synthesis of metal nanoparticles (NPs), biomolecules can be designed to act as a stabilizer and/or reducing agent to mediate the synthesis. It could also serve as a sacrificial template (or precursor) 124

to form the carbon-based nanomaterials as part of the hydrothermal synthesis process (28). Such approach is ideal as it does not require toxic reagents and harsh conditions, thus promoting a greener and more environmental friendly route for the NPs synthesis. Furthermore, the resulting biotemplated nanomaterials usually exhibit superb characteristics such as excellent biocompatibility, water solubility, rich chemical functional groups and tunable physiochemical properties (27, 29, 30). “Bio-assembling” is a self-assembly process of biomolecules, which has been recently explored to form different synthetic functional nanostructures. The underlying principle of this method is the self-assembling of biomolecules into stable and well-ordered structures through manipulating of their exceptional molecular recognition capabilities (31, 32). In particular, by manipulating the cognate Watson-Crick base pairing, single-stranded DNA (ssDNA) has been assembled into supramolecular structures such as two-dimensional (2D) DNA tiles (33, 34) and three-dimensional (3D) DNA origami (35, 36). Applying a similar principle, peptide sequences could be carefully designed to include both the hydrophilic and hydrophobic amino acid residues to form a peptide amphiphile (37). Through non-covalent interactions such as hydrogen bonding, ionic interaction, aromatic π – π stacking, hydrophobic and van der Waal’s forces, these peptide amphiphile would self-assemble into diverse 3D structures from nanofibrous network to hydrogel (38, 39). These bio-assembled nanomaterials represent an excellent platform for responsive loading and release of therapeutic agents (40).

Scheme 1. Bioinspired approaches to form functional nanostructured materials via bio-structure mimicking, bio-function anchoring, bio-templating and bio-assembling. Reproduced with permission from (14), Copyright 2013 Nature Publishing Group; Reproduced with permissions from (15–18), Copyright John Wiley and Sons; Reproduced with permissions from (19–21), Copyright 2011 American Chemical Society; Reproduced with permissions from (22), Copyright 2012 Elsevier. 125

In this book chapter, we will focus on the bioinspired materials design of using bio-templating synthesis and bio-assembling approaches to construct a variety of functional nanomaterials suitable for biomedical applications. By incorporating biomolecules, these bioinspired approaches promote not only a greener synthesis but also shortened the route of fabrication. Furthermore, they inherit the intrinsic characteristics possessed by biomolecules such as excellent biocompatibility, rich surface functional groups, good aqueous solubility and negligible toxicity that make them useful for biomedical applications. In addition, the composition and sequence of the biomolecules can be designed to fine-tune the morphology of inorganic nanostructures, and enable the programmable assembly of different higher-ordered biomolecular nanostructures with smart properties such as, dynamic control and stimuli responsiveness, for specific technological applications.

2. Bio-Templating Synthesis of Nanomaterials Bio-templating strategy, as the name suggests, uses biomolecules as templates to facilitate or direct the synthesis of nanomaterials. In the synthetic process, the biomolecules can either serve as preserved templates to be integrated into the final nanohybrids or will be sacrificed at the end of synthesis leaving the pure inorganic nanomaterial product with well-defined size, shape, and structure. Without the need to introduce toxic reagents and harsh condition, this approach represents a green, energy-efficient and eco-friendly way for nanomaterial synthesis (41, 42). In this section, the unique functionial properties of different biomolecules (e.g. nucleic acids, proteins and peptides) and their abilities to serve as a designable template to enable the synthesis of nanomaterials will be discussed. 2.1. Design of Preserved Bio-Template for Metallic Nanostructures Synthesis 2.1.1. Nucleic Acids as Bio-Templates Nucleic acids such as DNAs are natural biopolymers formed by long chain of nucleotides consisting of nucleobases such as Cytosine (C), Guanine (G), Adenine (A) and Thymine (T). The nucleobases are able to interact with their corresponding pair via Watson-Crick base pairing. Upon annealing with the complementary strand, the two ssDNAs would hybridise into a double helix structure with the negatively charged phosphate groups forming the backbone to minimise repulsion. As the phosphate backbones are highly negatively charged, they are able to bind cationic species such as metal ions via electrostatic interactions (43). Such interactions are able to stabilise the unstable intermediate cationic complexes. Likewise, the nucleobases such as C, G, A and T that contains the electron-rich nitrogenous group could potentially bind to and stabilise the cations and their intermediate complexes. Furthermore, they could also reduce electron-poor complexes through electron transfer mechanism (44). In a typical synthesis of metal NPs, metal precursors (e.g. Ag+) are first reduced to form nuclei and then grow into different nanostructures which are 126

stabilized by capping agents. Since the formation of metal NPs is driven by the capping and reducing capabilities of the ligands, these corresponding properties of nucleic acids are critical in the bio-template design for NP synthesis. For instance, the negatively charged phosphate groups are able to bind to and stabilise the positively charged metal ions such as Ag+ and Au+ via electrostatic interactions (45, 46). External stimuli such as UV light could be irradiated to induce a reducing environment allowing metal NP nucleation (47). Since double-stranded DNAs (dsDNAs) are rigid molecules with linear structures, the metal nuclei could subsequently grow along the dsDNA nanostructure and gradually forming a nanowire (48). It is worth noting that the stabilisation of Ag/Au intermediate complexes by the phosphate backbone is sufficient to prevent aggregation of the resultant NPs. Besides dsDNA, ssDNA has also demonstrated high efficiency in seed-mediated growth of metal NPs. Unlike dsDNA, ssDNA has exposed nucleobases which are strong electron donors. Thus, they are not only capable of binding to the electron-poor metal ions, but also able to reduce the metal ions for nucleation and growth. For example, it is found that the nucleobases in ssDNA could bind to Ag+ with different affinities (i.e. C > G > A > T) and different Ag nanostructures such as nanocubes, nano-octahedron and nanoflowers have been obtained by varying the sequences (49–51). As the molecular structures of DNA templates are preserved throughout the synthesis, some of these DNA-templated NPs retain the molecular recognition functions of the DNAs, endowing them with intrinsic sensing capabilities for diagnostic applications. This is especially true with plasmonic NPs such as AuNPs or AgNPs as they possess unique interparticle-distance dependent localized surface Plasmon resonance properties (53, 54). For instance, their high extinction coefficient are sensitive enough to induce noticeable colour change and detection by naked eyes (55, 56). With the target-specific binding of DNA preserved, these DNA-AuNPs or DNA-AgNPs could form aggregation or dispersion depending on their biomolecular interactions with target analytes, providing a basis for colorimetric sensing (57, 58). Particularly, Liu et al. have demonstrated the use of DNAzyme-AuNP as a highly sensitive and selective colorimetric assay for Pb(II) detection (59). Upon hybridisation, the AuNP aggregates, resulting in a blue-coloured solution. In the presence of Pb2+, hydrolytic cleavage occurs, preventing the aggregation, producing a red-coloured solution of well-dispersed AuNPs. In another report, Lin et al. have described the use of a DNA-AgNP as a platform for a simple fluorescence turn-on detection of dopamine (DA) (Figure 1) (52). Since DAs are able to form strong Ag-catechol bonds, DA addition released AgNPs from the DNA and a fluorescent signal was produced in the presence of intercalating DNA dyes. Using a similar strategy, detection of thiol-containing biomolecules were also reported (60, 61). More recently, nucleic acids were also exploited as templates to mediate the synthesis of metal nanoclusters (NCs), a new class of luminescent nanomaterial. These metal NCs exhibit ultrasmall size of < 2 nm, which lead to strong quantum confinement effect, thus bestowing them with bright photoluminescence (63). In particular, AgNCs are commonly synthesized using nucleic acids due to the strong binding affinity between Ag+ and nucleobase C, which could stabilise the intermediate complex and serve as a nucleation site for AgNCs (64, 65). 127

Compared to typical NP synthesis, these ultra-small fluorescent NCs require a more meticulous and stringent preparation. For instance, it was found that slight variation in the overall DNA length and/or sequences would alter the fluorescence properties of the AgNCs such as the emission wavelength and the photostability (66–68). As such, a careful design of DNA sequences is critical. Notably, Guo et al. reported a C6-loop with C-C mismatch to first entrap the Ag+ and then gradually reducing them to form AgNCs (69). By manipulation of the base pair mismatches and a basic sites to produce Ag binding sites, other DNA supramolecular structures such as the i-motif (70), DNA duplex (71) and the G-quadruplex (72) have also been explored to synthesize bio-templated AgNCs. Similar to DNA-templated metal NPs, these DNA-AgNCs could also function as an excellent sensing platform. Instead of colorimetric assay, they could be developed as fluorometric assays due to their intrinsic fluorescence characteristics. For example, Yeh et al. have observed that the fluorescence of DNA-AgNCs could be enhanced when comes in close proximity with G-rich DNA sequences (73). Based on this observation, they have further developed a NanoCluster Beacon (NCB) to detect an influenza sequence. Applying a similar strategy, Yin et al. prepared a system of DNA-AgNCs for cancer cell detection (Figure 2) (62). The system consists of two tailored DNA probes, one containing sequence for AgNCs template synthesis while the other comprises of G-rich DNA sequence at 5′-end and a cancer targeting aptamer sequence at 3′-end. Upon detecting CCRF-CEM cancer cell, the recognition probe would undergo conformation change, giving out a fluorescent signal.

Figure 1. (A) Schematic representation of using DNA-templated AgNPs and intercalating dye Genefinder (GF) for dopamine (DA) detection. (B) Fluorescence emission spectra of GF/dsDNA–silver nanohybrids in the presence of increasing DA concentrations (0-0.5μM). (C) Fluorescence intensity at 525nm as a function of the DA concentration (0-0.5μM). Reproduced with permission from (52), Copyright 2011 John Wiley and Sons Ltd. 128

Figure 2. (A) Schematic Illustration of an aptamer based AgNC assay for the label-free and fluorescent turn-on detection of cancer cell. The fluorescence responses of aptamer TD05 involved in recognition probe to assay target Ramos cancer cells by (B) flow cytometry and (C) confocal microscopy images. Reproduced with permission from (62), Copyright 2013 American Chemical Society.

Further on, photoinduced electron transfer of DNA-AgNCs was also explored. Wang’s group prepared a functional DNA-AgNCs that is capable of detecting hemin biomolecule using a parallel G-quadruplex and a hemin specific sensing sequence (74). Upon detection of complementary sequence, the G-quadruplex would be released and subsequently captures the hemin biomolecule, forming a stable G-quadruplex/hemin complex. This formation promotes electron transfer from DNA-AgNCs to the hemin complex, thus reducing the fluorescent intensity of the former. Other DNA templated NCs such as DNA-CuNCs have also shown potential to be used as a biosensor. For instance, Zhou et al. developed a label-free aptamer sensor for adenosine triphosphate (ATP) by controlling the 129

formation of DNA-CuNCs (19). Since CuNCs would only form in the presence of dsDNA but not ssDNA, the existence of ATP would bind strongly to one of the strands, inhibiting the growth of CuNCs. This simple detection for ATP has a preferable linear range (0.05-500 μM) and a high sensitivity with detection limit of 28 nM. Likewise, Wang’s group exploited this strategy to test for single nucleotide polymorphism (75). Compared to the perfectly matched DNA, it was found that the mismatch base pair site would provide an appropriate environment for CuNCs, altering their fluorescent intensity. The differences in fluorescence intensity were highly dependent on the mismatch sequence which allows the detection of more than one mismatches in a specific DNA sequence. Besides sensing of biomolecules, detections of other ions such as Pb2+ and S- have also been demonstrated (64, 76, 77). In general, nucleic acids are good stabilising and reducing agent which serves as an excellent designable bio-template to direct the synthesis of metal NPs. Such methodology promotes the green synthesis of metal NPs and renders inherent biocompatibility to the nanohybrid. Furthermore, as the bio-templates are preserved after the synthesis, it endows them with the unique biorecognition capability for direct biosensing and diagnostic applications without post functionalization.

2.1.2. Proteins as Bio-Templates

Proteins are a class of biomacromolecules with complex three-dimensional (3D) architecture. The protein structures which constitute of amino acids as basic building blocks, have diverse functional groups such as -NH2, -CO2H, -OH, -SH in their side chains for chemical synthesis and modications. Some of the functional groups of amino acids residues such as tyrosine (Tyr) (78), aspartic acids (Asp) (79, 80) trytophan (Trp) (81), lysine (Lys) (82) and cysteine (Cys) (83, 84) have been shown to be good reducing agents and/or stabilising agents, which could provide specific binding and nucleation sites for metal ions. For example, Tyr residues in the BSA protein template are found to be responsible for the reduction process of Au+, leading to the formation of spherical AuNPs (85, 86). Conversely, Casein protein which forms micelle structure in aqueous environment, uses its Asp residues to bind to and reduce Au+, producing anisotropic Au nanoplates (87). As mentioned previously, the formation of NCs involves a more rigorous requirement in the bio-template design. In particular, the bulky proteins could introduce steric hindrance, providing sufficient stabilizing effect to promote the formation of NCs. For instance, bulky proteins such as transferrin (88), lactoferrin (89), lysozyme type VI (90), horseradish peroxidase (91), trypsin (92) and insulin (93) which contain Cys residues could provide sufficient steric hindrance as well as forming Au+-thiolate intermediates to stabilise the growth of protein-templated AuNCs effectively (94, 95). 130

Similar to DNA-templated metallic nanostructures, the protein templates also retain their initial bio-recognition functions after the bio-templating synthesis, enabling them to act as diagnostic sensors. For instance, lysozyme-AuNCs were found to be able to retain their bioactivity, allowing them to label bacteria such as E.coli and inhibit their growth (96, 97). In another example, Wang et al. prepared a transferrin-AuNCs coupled with sheets of graphene oxide (GO) for cancer cells diagnosis and imaging (98). The nanocomposite has a “turn-on” fluorescent feature which could display near infrared (NIR) fluorescence upon identification of the transferrin receptor on the surface of the cancerous cell both in vitro and in vivo. In addition, Ranjita et al. utilised protein seeding of AuNPs via glycosylated haemoglobin to study the mechanism of glycosylation and sense the glycosylated end products by solution color changes due to the changes in AuNPs size (99). The size increase of the AuNPs could be observed through transmission electron microscopy (TEM) while the structural alterations of the proteins were investigated using infrared spectroscopy and circular dichroism. Besides the bio-recognition ability, BSA-AgNCs synthesized by Yu et al. displayed strong singlet oxygen generation capacity with a quantum yield of ~1.26, and exhibited excellent cellular uptake and biocompatibility which demonstrated a high anticancer efficacy via photodynamic therapy (100). Using similar strategies of reduction and stabilisation, protein cages such as apoferritin, viral capsid, heat shock protein and lumazine synthase have also been demonstrated as efficient templates for the preparation of nanomaterials (101–103). In particular, apoferritin has been used to synthesize magnetic NPs such as Fe (104, 105), while other metal NPs like Ni (106), Cr (107), Cu (108), Au (34), Ag (109), and semiconductor NPs such as CdS (110) and CdSe (111) have also been reported. Ferritin (Fn) has a nearly spherical structure with a negatively charged channels containing amino acids such as Asp and Glu, could bind to and transport positively charged metal ions to its hollow cavity (112, 113). The encapsulated metal ions could be reduced using UV light or heating and the resulting NPs would be stabilised by the neighbouring carboxylate groups in the cavity (114, 115). The as-synthesized Fn-NP could be used for diagnostic or therapeutic applications due to the distinctive properties of the cage-enclosed metal NPs. For example, Li et al. prepared a Fn-Fe3O4 nanostructures for tumor sensing and imaging (116). The surface of Fn was conjugated with cancer targeting RGD peptide and a green fluorescent protein, combined with the magnetic resonance imaging capability of Fe3O4 NPs, the nanoproduct could specifically target and track cellular uptake by tumor cells. On the other hand, Wang et al. prepared a CuS-Fn nanocage which achieved superior cancer therapeutic efficiency using photothermal therapy (Figure 3) (117). Moreover, the nanocages are also excellent positron emission tomography (PET) and photoacoustic imaging (PAI) agent which provide real-time monitoring and guidance of the nanocage in vivo. The theranostic potential of Fn-NP has also been illustrated recently by Ceci’s group (118). By decorating the surface of Fn with melanoma targeting peptide and poly(ethylene) glycol (PEG), and doping the Fe3O4 core with Co2+, the resulting magnetoferritin exhibit excellent targeting properties, outstanding in vivo stability, enhanced magnetic anisotropy and hyperthermic effect toward melanoma cancer cells. 131

Figure 3. (A) The preparation procedure of CuS−Fn NCs; TEM images of (B) iron free Fn and (C) CuS−Fn NCs stained with 1% uranyl acetate; (D)Temperature recording of U87 MG tumour mice upon 5 min laser exposure of different powers; (E) The variation of temperature in tumour area upon laser irradiation; (F) Images of U87MG tumour mice at various days after treatment. Reproduced with permission from (117), Copyright 2016 American Chemical Society.

2.1.3. Peptides as Bio-Templates In contrary to proteins, peptides compose of shorter chain of amino acids, which provide a more versatile platform for bio-templating synthesis due to the absence of complex secondary and tertiary structures. Through careful selection of amino acid residues, the peptide could be programmatically designed into an efficient template to direct the synthesis of metal NPs. For instance, Tan et al. conducted a systematic study to uncover the design rules for peptide synthesis of AuNPs (Figure 4) (119). They demonstrated that through combining amino acids such as Tyr and Trp with the shape-directing sequences respectively, the resulting 132

peptides (i.e. SEKLWWGASL and SEKLYYGASL) were able to synthesized Au nanoplates from AuCl4- precursors in one pot solution. Moreover, other peptides such as Ac-TLHVSSY-CONH2 and SSFPQPN were also designed to facilitate the formation of platinum nanocrystals (120, 121).

Figure 4. (A) Schematic illustration for the peptide mediated synthesis of AuNPs in aqueous solution. The peptides and chloroaurate ions were first interacted to form peptide-AuCl4 complexes, facilitating the reduction of Au ions to Au(0) in forming nuclei. It is followed by the growth of nuclei into crystalline particles induced by the addition of more Au atoms from the solutions or by fusion with other nuclei. (B) The molecular interactions between peptide and Au ion, as well as peptide and gold atoms which determined the reactivity of functional peptide template for metal NPs synthesis can be designed by the selection of amino acids and their sequences. TEM images of AuNPs synthesized from using multifunctional peptides (C) SEKLWWGASL and (D) SEKLYYGASL. Reproduced with permission from (119), Copyright 2010 American Chemical Society. Other than synthetic peptides, naturally occurring peptides such as glutathione (GSH) were utilised in the synthesis of ultrasmall AuNCs. Being a short chain peptide, the GSH peptides could facilitate the reduction of Au+ to Au0 while stabilising the Au+ intermediate by binding through carboxylate and thiol functional groups (122). The as-synthesized protein- and peptide-based metallic nanostructures could also be used as promising biosensing probes owing to the conserved biorecogntion functions of the template. This includes the detection of enzymes (123, 124), small molecules (125, 126) and metal ions (127–130). Remarkably, peptide-NCs are able to produce near-infrared emission wavelength upon excitation, allowing them to activate any photosensitizer to yield reactive oxygen species (ROS) such as singlet oxygen for potential photodynamic therapy (PDT). For instance, Zhang et al. fabricated a GSH-AuNCs functionalized with folic acids and PEG on the surface, to enable the entrapment of photosensitizer 133

chlorin e6 (131). The in vitro and in vivo studies show the enhanced cellular uptake and satisfactory PDT effectiveness toward cancerous MGC-803 cells. More recently, Vankayala et al. reported a TAT (cell penetrating peptide)-AuNCs that could perform simultaneous in vitro and in vivo fluorescence imaging, gene delivery, and NIR activated photodynamic therapy for effective anticancer therapy (132). The positively charged nanoprobe could load the negatively charged DNAs via electrostatic interactions and transport them into the nucleus for successive transfection. Furthermore, the nanoprobe could also generate ROS to induce cell apoptosis without the use of any organic photosensitizer. Proteins/peptides offer an excellent platform for the bioinspired design and engineering of metallic nanostructures of different size, shape and properties. Specifically the biomolecular shell of the metallic nanostructures could endow them with good biocompatibility and a diverse functional group such as NH2, COOH and SH for further conjugation of therapeutic and/or diagnostics agents for biomedical applications.

2.2. Design of Sacrificial Bio-Templates for Carbon Nanomaterials Synthesis Besides functioning as a bio-template that can be preserved throughout the bioinspired synthesis process, biomolecules such as nucleic acids, proteins, peptides, amino acids and carbohydrates could also be used as sacrificial templates in the synthesis of carbon nanomaterials. Particularly, “bio-dots”, the biomolecule-derived fluorescent nanodot, represent a new class of fluorescent nanomaterial synthesized using biomolecules as the sacrificial templates (133–135). Abundant with elements such as oxygen, nitrogen, phosphorous and sulphur, biomolecules are good doping agents in the preparation of the bio-dots. By introducing heteroatom doping, it provides various trapping sites of different series of energy levels in bio-dots (136, 137). This enables electronic transitions such as π→π* and n→π* transitions, allowing emission of photons with varying excitation energy. Therefore, the use of biomolecules could endow these bio-dots with interesting optical properties that are desirable for biomedical applications (138). In one example, Du et al. synthesized nitrogen doped nanodots using glucose and serine as precursor (139). The surface of the nanodots were passivated with functional groups such as C-O, C=O and O=C-OH which provided surface defects with different energy levels, granting the nanodots with excitation-dependent properties. Upon incubation with A549 cells, the nanodots could be readily uptaken by the cells, displaying multiple fluorescence emission wavelengths (i.e. blue, green and red). Correspondingly, Sun’s group prepared Asp derived bio-dots to diagnose brain cancer (Figure 5) (140). The Asp-dots (CD-Asp) not only exhibit tunable full-colour emission with a quantum yield of 7.5%, but also possess intrinsic selectivity and targeting affinity towards cancerous C6 glioma cells. Both in vitro and in vivo studies showed high biodistribution of Asp-dots located to the brain tumor indicating their potential application as an excellent bio-imaging and diagnostic agent. 134

Figure 5. (A) TEM, (B) High-resolution TEM images of CD-Asp; (C) In vivo imaging of glioma-bearing mice at different time points after injection with CD-Asp, CD-G, and CD-A. Reproduced with permission from (140), Copyright 2015 American Chemical Society. In addition to surface passivation, the biomolecules also supply the bio-dots with rich chemical functionalities for linking targeting moieties and/or drug molecules. For instance, Sharon and co-workers prepared a fluorescent dots from sorbital via microwave-assisted heating (141). The sorbital bio-dots were first attached onto BSA surface via electrostatic attractions and then further functionalized with cancer targeting ligand, folic acids and anticancer agent, doxorubicin (dox). The resulting complex could be applied as anticancer theranostics. On the other hand, Shen’s group utilised DNA from E.coli to 135

synthesize fluorescent bio-dots (142). The surface of DNA dots was grafted with functional groups such as C−OH, N−O, and N−P resulting in a negative zeta potential, enabling electrostatic dox loading. Due to the inherent fluorescence from DNA dots, the nanocomplex was able to induce cell apoptosis and at the same time allowed real-monitoring of the drug release process. On the whole, biomolecules could serve as an efficient precursor in the synthesis of carbon nanomaterials such as bio-dots. Although the overall secondary or tertiary structure of the biomolecule may be altered or completely destroyed in the process, their inherent properties such as aqueous solubility, rich chemical functional groups and excellent biocompatibility are still imparted to the bio-dots. This makes them potentially useful in various biomedical applications such as imaging, diagnostic and therapeutic delivery.

3. Bio-Directed Assembly of 3D Smart Nanostructures Molecular self-assembly mechanism underlies the organization of biological systems. Natural biopolymers such as nucleic acids and peptides can self-assemble into well-defined nanostructures in a programmable manner based on the information encoded into their primary structure - their sequences. In recent decades, the principle of molecular assembly is employed to achieve bottom-up synthesis of nanostructured materials, whereby material functionalities and properties depend on the “bio-directed assembly” of basic biomolecular building blocks. So far, a wide range of smart nanostructures has been formed based on the self-assembling capabilities of biopolymers. The nanostructures are either composed exclusively of biomolecules or consisting of inorganic NPs complex assembled by the biomolecular scaffold. In addition to the bio-assembly of static nanostructures, dynamic assembly of intelligent nanomaterials that is stimuli-responsive, error-free and reversible is also highly desirable to create molecular machines (143).

3.1. Nucleic Acid-Directed Assembly Nucleic acids not only carry the genetic information for heredity; but are also powerful nanotools for the bottom-up assembly of nanomaterials. Their minuscule size (diameter < 2 nm), short structural repeat of helix (~3.4-3.6 nm), inherent complementary base pairing property and the sequence programmability make them excellent linkers in constructing highly structured materials with precise spatial and dynamic control. Chemically modified DNAs are able to functionalize different nanomaterials such as AuNPs, quantum dots, and carbon nanotubes, leading to the formation of various hierarchical structures. In addition, specifically designed DNA motifs could self-assemble into one-dimensional (1D), two-dimensional (2D) and three-dimensional (3D) nanostructures, which could in turn act as scaffolds to direct the assembly of inorganic NPs. Furthermore, a special group of functional nucleic acids, for instance, DNAzymes with enzymatic 136

activity and aptamers with recognition properties are ideal for constructing the stimuli-responsive and reversible assembly of nanostructured materials (144).

3.1.1. Design of Chemically Modified DNA Linkers To Create Geometrical Nanostructures Mirkin et al. pioneered the use of chemically modified DNA oligomers to connect nanoparticles into nanoassemblies (145). In this approach, two thiolated non-complementary DNA sequences were functionalized onto two sets of colloidal AuNPs via Au-S chemistry. When a dsDNA with “sticky ends” that are complementary to the two grafted sequences, the two sets of AuNPs will self-assemble into large aggregates (145). A sticky end is a few unpaired nucleotides extending from the end of a dsDNA, which offers both excellent control of intermolecular interactions and predictable geometry at the point of association, and they are the key to the formation of various DNA nanostructures via ligation. Similarly, Alivasato et al. demonstrated a more discrete organization of AuNPs into dimers or trimers by mixing short ssDNA modified AuNPs with a well-designed complementary ssDNA template (146). Through careful design of dsDNA sequence, various well-defined nanostructures with nonlinear geometries such as triangles, pyramids, cubes and polyhedral have been developed through the years (147–152). For example, discrete DNA pyramids with AuNPs at the tips have been created, and biomimetic chiral nanostructures have been demonstrated using four sets of AuNPs with different diameters (153). This strategy of using DNA scaffold to control the placement of NPs has opened up many opportunities to enhance the functionality of pure DNA nanostructures. In addition to the assembly of single type of NPs, Yan et al. prepared a collection of DNA-assembled heteroparticle chiral pyramids from multiple metal and/or semiconductor NPs with an 80% yield. They were able to correlate the optical properties of these systematically assembled heteroparticle pyramids to their chirality (154). Willner and co-workers then introduced a new paradigm to construct tether-modified DNA scaffolds to organize left- or right-handed plasmonic helices of AuNPs (155). Recently, researchers started to utilize 3D DNA nanostructures for various biomedical applications. For example, Leong and co-workers developed a theranostic DNA nanoscaffold by decorating them with fluorescent AuNCs through covalent functionalization and intercalating therapeutic Actinomycin D for simultaneous detection and killing of E. coli and S. aureus (156). The same group also reported a bioinspired DNA nanosensor comprising of a molecular beacon module to a DNA nanoshell for real-time detection of mRNA in living cells (157). A DNA tetrahedron loaded with the chemotherapy drug and photosensitizer linked to a circulating tumor cells (CTC) targeting aptamer has been used to sense and treat CTCs effectively. In general, these DNA nanostructures show many advantages including enhanced cellular uptake, increased drug loading and synergistic therapeutic effect by combined loading of several therapeutic agents (158). 137

3.1.2. Design of DNA Tiles for Bottom-Up Assembly of Complex Nanostructures Seeman et al. are the first to design and synthesize DNA tiles which include various DNA motifs such as four-way branched junctions (159, 160), and double-cross-overs (DX) (161) to construct complex DNA nanostructures such as 2D arrays (162), cubes (147), and octahedrons (150, 159, 163). As shown in Figure 6A, the synthetic branched DNA junctions mainly consist of three or four arms of dsDNAs. Half of each ssDNA contributes to one arm, while the other half is paired with a neighbouring arms. Thus, all the arms of the branched junction are connected to each other at a central point (164). For example, a truncated octahedron (molecular weight ~ 790 kDa) was assembled from well-designed cyclic DNAs on a solid support. Six cyclic ssDNA molecules were used to form the six squares and the extra arms. The eight hexagons were constructed from another eight cyclic strands (148). In a similar way, the combined application of branched DNA junctions with sticky-end ligation has led to the successful construction of multiply connected objects, networks and devices. However, these branched DNA junctions are usually too flexible to form regular higher order structures.

Figure 6. (A) Self-assembly of branched DNA junctions into a 2D crystal. Reproduced with permission from (150), Copyright 2003 Springer Nature. (B) Typical DNA motifs including double-crossover (DX), triple crossover (TX) and paranemic crossover (PX) tiles. Reproduced with permission from (168), Copyright 2003 Elsevier. (C) Three-dimensional(3D) structure of DNA octrahedron. Three views of the 3D map generated from reconstruction of the DNA octahedron (Top, right). Raw images of individual DNA octrahedron and corresponding projections of the 3D map (Bottom, right). Reproduced with permission from (151), Copyright 2004 Springer Nature. 138

To tackle this rigidity problem, a stiffer motif called DNA double-crossover (DX) tile, which contains two dsDNAs connected to each other at two crossover points (Figure 6B), are used for constructing 2D arrays (150). Besides individual nanostructures, the assembly of periodic nanostructures have also been demonstrated by designing the sticky ends on one side of the DNA tile to be complementary to the other side (149). In addition, triple crossover (TX) tiles consisting of three in-plane DNA helices connected through crossover points and paranemic crossover (PX) tiles comprising of two parallel double helices fused at every possible cohesion point by reciprocal exchange have also been developed as basic building motifs (Figure 6B) (164). The replication/cloning of large 3D geometric DNA objects are very challenging. Shih and co-workers are the first to demonstrate a well-designed 3D DNA structure assembled from a readily amplifiable 1669-nucleotide ssDNA molecule. As illustrated in Figure 6C, this ssDNA was folded into an octahedron spontaneously with the addition of five 40-mer synthetic DNA (151). This DNA octahedron consists of five DX structures and seven PX (165) structures joined at six four-way junctions. Different from previous multistep assembly approaches, Erben et al. introduced a novel design principle which allows rapid assembly of DNA tetrahedra from multiple strands in one step with high yield (166). The same design principle has been extended to create a trigonal biypramid from six DNA strands with close to 40% yield (152). Instead of designing multiple DNA motifs to form different DNA structures, Mao’s group are the first to propose a three-point star tiles which could be assembled into various larger 3D hierarchical structures using the same basic building blocks. Tetrahedra, dodecanhedra or buckyballs have been also assembled in one pot from four, twenty or sixty tiles, respectively (167).

3.1.3. Design of DNA Origami as Addressable 2D/3D Scaffolds In 2006, Rothemund proposed “DNA origami”, a modern technology to construct larger and more complicated DNA nanostructures (163). DNA origami are arbitrarily shaped structures formed from the folding of a long ssDNA scaffold fixed by several smaller staple strands to create multiple double-stranded sites and consequently achieve a rigid 2D structure (170). Using this approach, several 2D shapes including rectangle, stars, and smiley faces have been created. Later on, the Shih’s group reported a breakthrough in DNA origami by building various 3D structures with excellent control. The key to the successful constructions of such 3D architecture is the optimized design of the staple strands to allow formation of Holliday junctions at specific locations. They further developed a computer program to expedite the design of complex DNA origami (170). Another notable contribution by Shih’s group is the creation of curved origami with delicate control of both the twist direction and the bending angle (171). DNA origami provides an excellent framework to create nanohybrid structures. Schreiber et al. have applied DNA origami as scaffolds to achieve hierarchical assembly of metal NPs, quantum dots as well as organic dyes in a planet-satellite construction with excellent control over both distance and stoichiometry (Figure 7) (169). 139

Figure 7. (A) A ssDNA scaffold (~8 kb) is annealed with ~200 synthetic oligonucleotide staples (~40-mer) to create various DNA origami structures of defined shape and size. (B) Satellite nanoparticles functionalized with multiple thiolated DNA strands are hybridized to the DNA origami structures. (C) The nanoparticle bearing DNA origami structures are hybridized to planet nanoparticle. (D) Au nano origami cluster with 60 nm AuNP planet and 10 nm AuNP satellites. (E) Ag–Au nano origami cluster with 80 nm AuNP planet and 20 nm AgNP satellites. Reproduced with permission from (169), Copyright 2014 Springer Nature.

3.1.4. Design of Functional Nucleic Acids for Dynamic 3D Assembly Certain nucleic acid sequences such as i-motifs undergo conformational transitions upon environmental stimulation thus making them especially suitable for creating smart molecular devices/machines. In addition, functional nucleic acids such as DNAzymes, aptamers are suitable for dynamic 3D assembly of novel structures from various NPs (AuNPs, quantum dots, carbon nanotubes, and iron oxide NPs). To achieve stimuli-responsive assembly, for example, assembly and disassembly of AuNP networks can be controlled by DNAzymes or aptamers that undergo dehybridization or conformation change in response to metal ions (172), small organic molecules (173) or even proteins. For instance, Lu et al. applied both DNAzymes with high metal selectivity and metallic NPs with strong distance-dependent optical properties in a ‘tail-to-tail’ configuration for fast colorimetric sensing of Pb2+ ions (172). Smart design of multiple aptamer or aptazyme sequences enable the selective responses to multiple stimuli just like an AND/OR logic gate (174). For assembly with error correction, proofreading and error removal can be achieved by combining both a cleavage and a ligation 140

DNAzyme in an enzyme cascade. Furthermore, Lu and co-workers have taken advantages of the different binding affinities between biotin and desthiobiotin toward streptavidin to construct a controlled reversible assembly of strepavidins to convey an encrypted message on DNA origami (175). DNA based smart assembly has led to the development of a variety of smart molecular machines such as molecular switches, molecular motors, molecular walkers, switchable materials/devices. Furthermore, the advancement in this field has recently revolutionized DNA computation and molecular programming.

3.2. Peptide-Directed Assembly Similar to nucleic acid, peptide which consists of amino acids, is another popular biomolecular building block for assembly of various supramolecular structures. Peptides are attractive because of its good biocompatibility, biodegradability, low immunogenicity, structural programmability, versatile functionality as well as cost-effectiveness of large scale production via standard solid-phase synthesis. It is worthy to note that an enormous opportunity arises from the combinatorial complexity of the peptide sequences since they comprise 20 naturally occurring amino acids with different properties as building blocks and exhibit a wide range of chemical functionalities. To illustrate, the diverse properties of amino acids include charged (D, E, H, R, and K), polar (S, T, Q and N), nonpolar (A, V, L, I and M), aromatic (F, Y and W) and other special residues (P, C and G). The properties of individual amino acid residue can contribute differently toward the overall properties of resulting peptides, which subsequently determine the final supramolecular structures (176). The assembly of ordered peptide nanostructures is a spontaneous thermodynamic and kinetic driven process, controlled by the synergy of various intermolecular non-covalent interactions including hydrogen bonding, π-π stacking, electrostatic, hydrophobic and van der Waals interactions (177). As a result, peptide self-assembly has created various architectures over a large length scale ranging from nanoscale to macroscale with various 2D conformations such as α-helix and β-sheet. Frederix et al. recently demonstrated a computational tool to screen the self-assembly possibility of 8,000 tripeptides in aquesous medium and develop important design rules for self-assembling sequences (178). Three main ways to achieve peptide self-assembly by using 1) short diphenylalanine based aromatic peptides, 2) peptide amphiphiles and 3) more complex long polypeptides (> 20 amino acids) are discussed in details in this section.

3.2.1. Diphenylalanine-Based Aromatic Peptides Short peptide motifs containing aromatic groups can self-assemble in aqueous medium. Reches and Gazit observed that the self-assembly of diphenylalanine (FF), a fragment of Alzheimer’s β-amyloid protein into discrete nanotubes (179). 141

It is know that amyloid fibrils are highly organized protein aggregates that play a physiological role in microorganisms and melanin-producing mammalian cells. They can readily form nanofibers and nanotubes. Similar to the naturally occurring amyloids, the ordered organization of the unique peptide nanostructures comes from the cooperated effects of hydrogen bonding and π-π stacking in the FF dipeptides. A pathway for FF dipeptides to form nanotubes has been proposed whereby FF dipeptides first stack up to form a 2D layer and subsequently this 2D layer closes up to form a tubular structure (Figure 8A). This peptide nanotube can further act as a bio-template to form silver wires with a micrometer persistence length. Other aromatic dipeptides such as WY, WF and WW were unable to form any nanotubular structure. For FW peptides, nanotubes and amorphous aggregates were formed at the same time. Later on, FF dipeptides have been assembled into a variety of complex architectures such as vesicles, hexagonal microtubes, ordered chains, nanowires (180) and nanofibers (177, 181). These peptide nanostructures could be used for bioimaging, biosensing, drug delivery, 3D cell culture and nanofabrication. The short FF-based peptides are very popular building blocks due to simple structure, versatile functionalities and cost-effectiveness. To improve their functionality, they could integrate with inorganic components to direct the synthesis of metal nanowires, nanoribbons and polymers.

3.2.2. Peptides Amphiphiles

In general, peptide amphiphiles consist of a short hydrophobic chain linked with a short hydrophilic peptide sequence. They possess both the structural feature of an amphiphilic surfactant and the biological activity of a peptide. In detail, a peptide amphiphile molecule consists of four domains including an aliphatic tail (I), an β-sheet forming sequence (II), several charged amino acids to offer solubility and induce crosslinking (III) and a signalling peptide for biological response (IV) (182). Modifications of the typical peptide amphiphiles have been developed to suit specific applications. For example, a fibronectin-mimetic peptide amphiphile sequence was design to compose of both RGD and PHSRN binding domain, which renders superior cell adhesion properties (Figure 8B) (183). In most cases, their linear hydrophilic head and hydrophobic tail structure determines whether spherical micelles, cylindrical micelles or lamellar structures are formed under physiological conditions, which leads to interesting applications in tissue engineering, regenerative medicine and drug delivery (184). An alternative approach is to rely on the synthetic aromatic groups such as fluorene, naphthalene, azobezene, pyrene or phenyl groups to confer the amphiphilicity required to initiate and drive self-assembly. This special class of peptide amphiphiles is commonly named as aromatic peptide amphiphiles. A typical aromatic peptide amphiphile is a short peptide sequence capped with a synthetic aromatic moiety at the N-terminus (176). Different from their aliphatic 142

counterparts, the assembly of these aromatic peptide amphiphiles are controlled by both the stacking interactions and the β-sheet forming hydrogen bondings. Thus, the assembly process is also affected by the planarity of the aromatic cap and the geometric restrictions of their stacking arrangement. The aromatic peptide amphiphiles have led to the successful formation of nanoscale spheres, worms, sheets, tapes and fibers/tubes (176).

Figure 8. (A) A proposed formation pathway of diphenylalanine (FF)-based nanotubes. Reprinted with permission from (181), Copyright 2010 Royal Society of Chemistry. (B) Self-Assembly of fibronectin mimetic peptide amphiphile nanofibers. TEM image of the assembled nanofiber is displayed on the right. Reprinted with permission from (183), Copyright 2010 American Chemical Society. (C) Design of the self-assembling polypeptide tetrahedron. Reprinted with permission from (185), Copyright 2013 Springer-Nature.

143

3.2.3. Polypeptides Synthetic polypeptides are polymers composed of many amino acids (>20 residues) and they typically assemble into random coil, α-helix and β-sheet. The conformation is correlated with the polypeptide solubility and rigidity in solutions. Higher solubility favours the random coil structure (177). These polypeptides can be developed into a diverse range of nanostructures including nanofibrils, lamellae with stimuli-responsive properties. For instance, the hydrogen bondings and hydrophobic interactions between polypeptides are highly temperature dependent. Thus, a small change in temperature can easily alter the secondary structures of polypeptides. Temperature treatment has been demonstrated to change the secondary structures of polypeptides from α-helices to β-sheets, consequently leading to an overall structural transition from micelles to nanoribbons (186). In one example, Padilla et al. presented a modular approach to design symmetrical peptide nanostructures. This general strategy enables the successful construction of tetrahedral cages from 12 subunits, including 6 pairs of coiled-coil forming peptides, 2 antiparallel dimers and 4 parallel dimers (187, 188). In another example, a strategy was develop to form self-assembling polypeptide polyhedron from orthogonal dimerizing segments (Figure 8C). Jerala and co-workers successfully constructed a tetrahedron from a single polypeptide chain comprising of 12 coiled coil-forming segments separated by flexible peptide hinges (185). This polypeptide design principle provides a foundation for self-assembly construction of novel polypeptide nanostructures. In comparison to DNA nanostructures that are based on complementary base pairing, the peptide-directed assembly of nanostructures is far more sophisticated due to the complexity in long-range cooperative interactions between different amino acids, which is not easily predictable from the primary structure. DNAs as nanostructure building blocks have demonstrated great success, peptides may provide larger conformational variability and consequently more versatile functionality, which require further research.

4. Conclusion and Perspectives In summary, biomolecules possess unique properties which allow the development biomimetic functional nanomaterials suitable for a wide range of biomedical applications from sensing, imaging, delivery to therapy. Their distinctive features enable them to either act as a template in facilitating the formation of nanomaterials via the “bio-templating” synthesis or self-organize into highly ordered and functional architectures via the “bio-assembling” strategy. In both cases, the formation of higher ordered materials involves the cooperation of multiple intermolecular interactions including non-covalent interactions such as π-π stacking, hydrogen bonding, electrostatic attraction, hydrophobic interaction and van der Waals’ forces. By manipulating their interplay, it could give rise to dynamic and responsive changes, allowing the construction of a multitude of structures and morphologies from 1D, 2D to 3D over a range of 144

length scales. Nonetheless, this requires profound and extensive understanding of these bioinspired approaches in the formation of functional nanostructures. This will include the detailed study of effectual self-arrangement of DNAs and peptides into hierarchical complex, and their interactions with other materials to promote growth of nanoparticles. When equipped with this knowledge, the future of this field would be expected to expand towards more complex and responsive structures that closely mimics the biological system. The recent successes in the design of complex nanostructures based on DNAs and peptides have demonstrated great promises of bio-templating and bio-assembly approaches to engineer next generation functional biomimetic nanomaterials for more advanced biomedical applications such theranostic and nanorobotics for surgery.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

Meyers, M. A.; McKittrick, J.; Chen, P.-Y. Science 2013, 339, 773–779. Cölfen, H. Nat. Mater. 2010, 9, 960–961. Dorozhkin, S. V.; Epple, M. Angew. Chem., Int. Ed. 2002, 41, 3130–3146. Kröger, N.; Deutzmann, R.; Sumper, M. Science 1999, 286, 1129–1132. Lowenstam, H. A. Geol. Soc. Am. Bull. 1962, 73, 435–438. Suzuki, M.; Saruwatari, K.; Kogure, T.; Yamamoto, Y.; Nishimura, T.; Kato, T.; Nagasawa, H. Science 2009, 325, 1388–90. Freeman, C. L.; Harding, J. H.; Quigley, D.; Rodger, P. M. Angew. Chem., Int. Ed. 2010, 49, 5135–5137. Sanghi, R.; Verma, P. Bioresour. Technol. 2009, 100, 501–504. Xie, J.; Lee, J. Y.; Wang, D. I.; Ting, Y. P. Small 2007, 3, 672–682. Chen, Q.; Pugno, N. M. J. Mech. Behav. Biomed. Mater. 2013, 19, 3–33. Prasad, A.; Mahato, K.; Chandra, P.; Srivastava, A.; Joshi, S. N.; Maurya, P. K. J. Mol. Eng. Mater., 04, 1640004. Hensel, R.; Finn, A.; Helbig, R.; Braun, H. G.; Neinhuis, C.; Fischer, W. J.; Werner, C. Adv. Mater. 2014, 26, 2029–2033. Buehler, M. J. Acta Mech. Solida Sin. 2010, 23, 471–483. Guo, C. X.; Xie, J.; Wang, B.; Zheng, X.; Yang, H. B.; Li, C. M. Sci. Rep. 2013, 3, 2957. Naleway, S. E.; Porter, M. M.; McKittrick, J.; Meyers, M. A. Adv. Mater. 2015, 27, 5455–5476. Heim, M.; Keerl, D.; Scheibel, T. Angew. Chem., Int. Ed. 2009, 48, 3584–3596. Latorre, A.; Somoza, Á. ChemBioChem 2012, 13, 951–958. Fan, R.; Chew, S. W.; Cheong, V. V.; Orner, B. P. Small 2010, 6, 1483–1487. Zhou, Z.; Du, Y.; Dong, S. Anal. Chem. 2011, 83, 5122–5127. Mei, Q.; Wei, X.; Su, F.; Liu, Y.; Youngbull, C.; Johnson, R.; Lindsay, S.; Yan, H.; Meldrum, D. Nano Lett. 2011, 11, 1477–1482. Bakota, E. L.; Wang, Y.; Danesh, F. R.; Hartgerink, J. D. Biomacromolecules 2011, 12, 1651–1657. Loo, Y.; Zhang, S.; Hauser, C. A. E. Biotechnol. Adv. 2012, 30, 593–603. 145

23. Narain, R.; Housni, A.; Lane, L. J. Polym. Sci., Part A: Polym. Chem. 2006, 44, 6558–6568. 24. Dohi, H.; Kikuchi, S.; Kuwahara, S.; Sugai, T.; Shinohara, H. Chem. Phys. Lett. 2006, 428, 98–101. 25. Huang, J.; Lin, L.; Sun, D.; Chen, H.; Yang, D.; Li, Q. Chem. Soc. Rev. 2015, 44, 6330–6374. 26. Patwardhan, S. V. Chem. Commun. 2011, 47, 7567–7582. 27. Xie, J.; Tan, Y. N.; Lee, J. Y. In Biomimetic and Bioinspired Nanomaterials; Kumar, C. S. S. R., Ed.; John Wiley & Sons: Weiheim, 2010; Vol. 7, pp 251−282. 28. Ruiz-Hitzky, E.; Darder, M.; Aranda, P. J. Mater. Chem. 2005, 15, 3650–3662. 29. Cao, H.; Zheng, H.; Liu, K.; Warner, J. H. ChemPhysChem 2010, 11, 489–494. 30. Lin, N.; Huang, J.; Dufresne, A. Nanoscale 2012, 4, 3274–3294. 31. Sun, W.; Gu, Z. Biomater. Sci. 2015, 3, 1018–1024. 32. Yu, Z.; Xu, Q.; Dong, C.; Lee, S. S.; Gao, L.; Li, Y.; D'Ortenzio, M.; Wu, J. Curr. Pharm. Des. 2015, 21, 4342–4354. 33. Wei, B.; Dai, M.; Yin, P. Nature 2012, 485, 623–626. 34. He, Y.; Su, M.; Fang, P. a.; Zhang, C.; Ribbe, A. E.; Jiang, W.; Mao, C. Angew. Chem. 2010, 122, 760–763. 35. Han, D.; Pal, S.; Liu, Y.; Yan, H. Nat. Nanotechnol. 2010, 5, 712–717. 36. Fu, Y.; Zeng, D.; Chao, J.; Jin, Y.; Zhang, Z.; Liu, H.; Li, D.; Ma, H.; Huang, Q.; Gothelf, K. V. J. Am. Chem. Soc. 2012, 135, 696–702. 37. Cavalli, S.; Albericio, F.; Kros, A. Chem. Soc. Rev. 2010, 39, 241–263. 38. Childers, W. S.; Ni, R.; Mehta, A. K.; Lynn, D. G. Curr. Opin. Chem. Biol. 2009, 13, 652–659. 39. Zhao, X.; Pan, F.; Perumal, S.; Xu, H.; Lu, J. R.; Webster, J. R. Soft Matter 2009, 5, 1630–1638. 40. Xu, H. V.; Zheng, X. T.; Mok, B. Y. L.; Ibrahim, S. A.; Yu, Y.; Tan, Y. N. J. Mol. Eng. Mater. 2016, 04, 1640003. 41. Velmurugan, P.; Anbalagan, K.; Manosathyadevan, M.; Lee, K.-J.; Cho, M.; Lee, S.-M.; Park, J.-H.; Oh, S.-G.; Bang, K.-S.; Oh, B.-T. Bioprocess Biosyst. Eng. 2014, 37, 1935–1943. 42. Darroudi, M.; Hakimi, M.; Sarani, M.; Oskuee, R. K.; Zak, A. K.; Gholami, L. Ceram. Int. 2013, 39, 6917–6921. 43. Chen, Y.; Cheng, W. Wiley Interdiscip. Rev.: Nanomed. Nanobiotechnol. 2012, 4, 587–604. 44. Ritchie, C. M.; Johnsen, K. R.; Kiser, J. R.; Antoku, Y.; Dickson, R. M.; Petty, J. T. J. Phys. Chem. C 2007, 111, 175–181. 45. Nishinaka, T.; Takano, A.; Doi, Y.; Hashimoto, M.; Nakamura, A.; Matsushita, Y.; Kumaki, J.; Yashima, E. J. Am. Chem. Soc. 2005, 127, 8120–8125. 46. Yan, H.; Park, S. H.; Finkelstein, G.; Reif, J. H.; LaBean, T. H. Science 2003, 301, 1882–1884. 47. Kundu, S. Phys. Chem. Chem. Phys. 2013, 15, 14107–14119. 146

48. Watson, S. M.; Wright, N. G.; Horrocks, B. R.; Houlton, A. Langmuir 2009, 26, 2068–2075. 49. Wang, Z.; Tang, L.; Tan, L. H.; Li, J.; Lu, Y. Angew. Chem., Int. Ed. 2012, 51, 9078–9082. 50. Wang, Z.; Zhang, J.; Ekman, J. M.; Kenis, P. J.; Lu, Y. Nano Lett. 2010, 10, 1886–1891. 51. Wu, J.; Tan, L. H.; Hwang, K.; Xing, H.; Wu, P.; Li, W.; Lu, Y. J. Am. Chem. Soc. 2014, 136, 15195–15202. 52. Lin, Y.; Yin, M.; Pu, F.; Ren, J.; Qu, X. Small 2011, 7, 1557–1561. 53. Zhao, W.; Brook, M. A.; Li, Y. ChemBioChem 2008, 9, 2363–2371. 54. Wei, H.; Chen, C.; Han, B.; Wang, E. Anal. Chem. 2008, 80, 7051–7055. 55. Ai, K.; Liu, Y.; Lu, L. J. Am. Chem. Soc. 2009, 131, 9496–9497. 56. Liu, X.; Atwater, M.; Wang, J.; Huo, Q. Colloids Surf., B 2007, 58, 3–7. 57. Tan, Y. N.; Su, X.; Liu, E. T.; Thomsen, J. S. Anal. Chem. 2010, 82, 2759–2765. 58. Song, J.; Li, Z.; Cheng, Y.; Liu, C. Chem. Commun. 2010, 46, 5548–5550. 59. Liu, J.; Lu, Y. J. Am. Chem. Soc. 2003, 125, 6642–6643. 60. Lin, Y.; Tao, Y.; Ren, J.; Pu, F.; Qu, X. Biosens. Bioelectron. 2011, 28, 339–343. 61. Lin, Y.; Tao, Y.; Pu, F.; Ren, J.; Qu, X. Adv. Funct. Mater. 2011, 21, 4565–4572. 62. Yin, J.; He, X.; Wang, K.; Xu, F.; Shangguan, J.; He, D.; Shi, H. Anal. Chem. 2013, 85, 12011–12019. 63. Yu, Y.; Mok, B. Y.; Loh, X. J.; Tan, Y. N. Adv. Healthcare Mater. 2016, 5, 1844–1859. 64. Chen, W.-Y.; Lan, G.-Y.; Chang, H.-T. Anal. Chem. 2011, 83, 9450–9455. 65. Petty, J. T.; Sergev, O. O.; Nicholson, D. A.; Goodwin, P. M.; Giri, B.; McMullan, D. R. Anal. Chem. 2013, 85, 9868–9876. 66. Ma, J.-L.; Yin, B.-C.; Ye, B.-C. Analyst 2016, 141, 1301–1306. 67. Schultz, D.; Gwinn, E. Chem. Commun. 2011, 47, 4715–4717. 68. Volkov, I. L.; Ramazanov, R. R.; Ubyivovk, E. V.; Rolich, V. I.; Kononov, A. I.; Kasyanenko, N. A. ChemPhysChem 2013, 14, 3543–3550. 69. Guo, W.; Yuan, J.; Dong, Q.; Wang, E. J. Am. Chem. Soc. 2009, 132, 932–934. 70. Dong, Y.; Yang, Z.; Liu, D. Acc. Chem. Res. 2014, 47, 1853–1860. 71. Huang, Z.; Pu, F.; Hu, D.; Wang, C.; Ren, J.; Qu, X. Chem.−Eur. J. 2011, 17, 3774–3780. 72. Li, W.; Liu, L.; Fu, Y.; Sun, Y.; Zhang, J.; Zhang, R. Photochem. Photobiol. Sci. 2013, 12, 1864–1872. 73. Yeh, H.-C.; Sharma, J.; Han, J. J.; Martinez, J. S.; Werner, J. H. Nano Lett. 2010, 10, 3106–3110. 74. Zhang, L.; Zhu, J.; Guo, S.; Li, T.; Li, J.; Wang, E. J. Am. Chem. Soc. 2013, 135, 2403–2406. 75. Jia, X.; Li, J.; Han, L.; Ren, J.; Yang, X.; Wang, E. ACS Nano 2012, 6, 3311–3317. 76. Chen, J.; Liu, J.; Fang, Z.; Zeng, L. Chem. Commun. 2012, 48, 1057–1059. 77. Liu, J.; Chen, J.; Fang, Z.; Zeng, L. Analyst 2012, 137, 5502–5505. 147

78. Bhargava, S. K.; Booth, J. M.; Agrawal, S.; Coloe, P.; Kar, G. Langmuir 2005, 21, 5949–5956. 79. Tan, Y. N.; Lee, J. Y.; Wang, D. I. J. Phys. Chem. C 2008, 112, 5463–5470. 80. Shao, Y.; Jin, Y.; Dong, S. Chem. Commun. 2004, 1104–1105. 81. Selvakannan, P.; Mandal, S.; Phadtare, S.; Gole, A.; Pasricha, R.; Adyanthaya, S.; Sastry, M. J. Colloid Interface Sci. 2004, 269, 97–102. 82. Selvakannan, P.; Mandal, S.; Phadtare, S.; Pasricha, R.; Sastry, M. Langmuir 2003, 19, 3545–3549. 83. Mandal, S.; Gole, A.; Lala, N.; Gonnade, R.; Ganvir, V.; Sastry, M. Langmuir 2001, 17, 6262–6268. 84. Zhang, F. X.; Han, L.; Israel, L. B.; Daras, J. G.; Maye, M. M.; Ly, N. K.; Zhong, C.-J. Analyst 2002, 127, 462–465. 85. Bakshi, M. S.; Kaur, H.; Khullar, P.; Banipal, T. S.; Kaur, G.; Singh, N. J. Phys. Chem. C 2011, 115, 2982–2992. 86. Au, L.; Lim, B.; Colletti, P.; Jun, Y. S.; Xia, Y. Chem.−Asian J. 2010, 5, 123–129. 87. Liu, Y.; Liu, L.; Yuan, M.; Guo, R. Colloids Surf., A 2013, 417, 18–25. 88. Le Guével, X.; Daum, N.; Schneider, M. Nanotechnology 2011, 22, 275103. 89. Xavier, P. L.; Chaudhari, K.; Verma, P. K.; Pal, S. K.; Pradeep, T. Nanoscale 2010, 2, 2769–2776. 90. Chen, T. H.; Tseng, W. L. Small 2012, 8, 1912–1919. 91. Wen, F.; Dong, Y.; Feng, L.; Wang, S.; Zhang, S.; Zhang, X. Anal. Chem. 2011, 83, 1193–1196. 92. Kawasaki, H.; Yoshimura, K.; Hamaguchi, K.; Arakawa, R. Anal. Sci. 2011, 27, 591. 93. Liu, C. L.; Wu, H. T.; Hsiao, Y. H.; Lai, C. W.; Shih, C. W.; Peng, Y. K.; Tang, K. C.; Chang, H. W.; Chien, Y. C.; Hsiao, J. K. Angew. Chem., Int. Ed. 2011, 50, 7056–7060. 94. Xu, Y.; Sherwood, J.; Qin, Y.; Crowley, D.; Bonizzoni, M.; Bao, Y. Nanoscale 2014, 6, 1515–1524. 95. Chaudhari, K.; Xavier, P. L.; Pradeep, T. ACS Nano 2011, 5, 8816–8827. 96. Liu, J.; Lu, L.; Xu, S.; Wang, L. Talanta 2015, 134, 54–59. 97. Chen, W.-Y.; Lin, J.-Y.; Chen, W.-J.; Luo, L.; Wei-Guang Diau, E.; Chen, Y.C. Nanomedicine 2010, 5, 755–764. 98. Wang, Y.; Chen, J.-T.; Yan, X.-P. Anal. Chem. 2013, 85, 2529–2535. 99. GhoshMoulick, R.; Bhattacharya, J.; Mitra, C. K.; Basak, S.; Dasgupta, A. K. Nanomed. Nanotechnol. Biol. Med. 2007, 3, 208–214. 100. Yu, Y.; Geng, J.; Ong, E. Y. X.; Chellappan, V.; Tan, Y. N. Adv. Healthcare Mater. 2016, 5, 2528–2535. 101. Liepold, L.; Anderson, S.; Willits, D.; Oltrogge, L.; Frank, J. A.; Douglas, T.; Young, M. Magn. Reson. Med. 2007, 58, 871–879. 102. Liepold, L. O.; Abedin, M. J.; Buckhouse, E. D.; Frank, J. A.; Young, M. J.; Douglas, T. Nano Lett. 2009, 9, 4520–4526. 103. Wörsdörfer, B.; Pianowski, Z.; Hilvert, D. J. Am. Chem. Soc. 2012, 134, 909–911.

148

104. Kostiainen, M. A.; Ceci, P.; Fornara, M.; Hiekkataipale, P.; Kasyutich, O.; Nolte, R. J.; Cornelissen, J. J.; Desautels, R. D.; van Lierop, J. ACS Nano 2011, 5, 6394–6402. 105. Martinez-Perez, M.; de Miguel, R.; Carbonera, C.; Martinez-Julvez, M.; Lostao, A.; Piquer, C.; Gomez-Moreno, C.; Bartolome, J.; Luis, F. Nanotechnology 2010, 21, 465707. 106. Gálvez, N.; Sánchez, P.; Domínguez-Vera, J. M.; Soriano-Portillo, A.; Clemente-León, M.; Coronado, E. J. Mater. Chem. 2006, 16, 2757–2761. 107. Okuda, M.; Iwahori, K.; Yamashita, I.; Yoshimura, H. Biotechnol. Bioeng. 2003, 84, 187–194. 108. Iwahori, K.; Takagi, R.; Kishimoto, N.; Yamashita, I. Mater. Lett. 2011, 65, 3245–3247. 109. Domínguez‐Vera, J. M.; Gálvez, N.; Sánchez, P.; Mota, A. J.; Trasobares, S.; Hernández, J. C.; Calvino, J. J. Eur. J. Inorg. Chem. 2007, 2007, 4823–4826. 110. Iwahori, K.; Yamashita, I. Nanotechnology 2008, 19, 495601. 111. Yamashita, I.; Hayashi, J.; Hara, M. Chem. Lett. 2004, 33, 1158–1159. 112. Chiancone, E.; Ceci, P. Biochim. Biophys. Acta, Gen. Subj. 2010, 1800, 798–805. 113. Haldar, S.; Bevers, L. E.; Tosha, T.; Theil, E. C. J. Biol. Chem. 2011, 286, 25620–25627. 114. Bhushan, B.; Kumar, S. U.; Matai, I.; Sachdev, A.; Dubey, P.; Gopinath, P. J. Biomed. Nanotechnol. 2014, 10, 2950–2976. 115. Klem, M. T.; Young, M.; Douglas, T. J. Mater. Chem. 2010, 20, 65–67. 116. Li, K.; Zhang, Z.-P.; Luo, M.; Yu, X.; Han, Y.; Wei, H.-P.; Cui, Z.-Q.; Zhang, X.-E. Nanoscale 2012, 4, 188–193. 117. Wang, Z.; Huang, P.; Jacobson, O.; Wang, Z.; Liu, Y.; Lin, L.; Lin, J.; Lu, N.; Zhang, H.; Tian, R. ACS Nano 2016, 10, 3453–3460. 118. Fantechi, E.; Innocenti, C.; Zanardelli, M.; Fittipaldi, M.; Falvo, E.; Carbo, M.; Shullani, V.; Di Cesare Mannelli, L.; Ghelardini, C.; Ferretti, A. M. ACS Nano 2014, 8, 4705–4719. 119. Tan, Y. N.; Lee, J. Y.; Wang, D. I. J. Am. Chem. Soc. 2010, 132, 5677–5686. 120. Li, Y.; Huang, Y. Adv. Mater. 2010, 22, 1921–1925. 121. Chiu, C.-Y.; Li, Y.; Ruan, L.; Ye, X.; Murray, C. B.; Huang, Y. Nat. Chem. 2011, 3, 393–399. 122. Luo, Z.; Yuan, X.; Yu, Y.; Zhang, Q.; Leong, D. T.; Lee, J. Y.; Xie, J. J. Am. Chem. Soc. 2012, 134, 16662–16670. 123. Gu, Y.; Wen, Q.; Kuang, Y.; Tang, L.; Jiang, J. RSC Adv. 2014, 4, 13753–13756. 124. Wen, Q.; Gu, Y.; Tang, L.-J.; Yu, R.-Q.; Jiang, J.-H. Anal. Chem. 2013, 85, 11681–11685. 125. Yuan, X.; Tay, Y.; Dou, X.; Luo, Z.; Leong, D. T.; Xie, J. Anal. Chem. 2013, 85, 1913–1919. 126. Zhou, T.; Rong, M.; Cai, Z.; Yang, C. J.; Chen, X. Nanoscale 2012, 4, 4103–4106. 127. Zhang, G.; Li, Y.; Xu, J.; Zhang, C.; Shuang, S.; Dong, C.; Choi, M. M. Sens. Actuators, B 2013, 183, 583–588. 149

128. Huang, C. C.; Yang, Z.; Lee, K. H.; Chang, H. T. Angew. Chem. 2007, 119, 6948–6952. 129. Chen, Z.; Lu, D.; Zhang, G.; Yang, J.; Dong, C.; Shuang, S. Sens. Actuators, B 2014, 202, 631–637. 130. Wang, C.; Xu, L.; Wang, Y.; Zhang, D.; Shi, X.; Dong, F.; Yu, K.; Lin, Q.; Yang, B. Chem.−Asian J. 2012, 7, 1652–1656. 131. Zhang, C.; Li, C.; Liu, Y.; Zhang, J.; Bao, C.; Liang, S.; Wang, Q.; Yang, Y.; Fu, H.; Wang, K. Adv. Funct. Mater. 2015, 25, 1314–1325. 132. Vankayala, R.; Kuo, C. L.; Nuthalapati, K.; Chiang, C. S.; Hwang, K. C. Adv. Funct. Mater. 2015, 25, 5934–5945. 133. Liu, J.-M.; Lin, L.-p.; Wang, X.-X.; Lin, S.-Q.; Cai, W.-L.; Zhang, L.-H.; Zheng, Z.-Y. Analyst 2012, 137, 2637–2642. 134. Jiang, J.; He, Y.; Li, S.; Cui, H. Chem. Commun. 2012, 48, 9634–9636. 135. Yang, Z.-C.; Wang, M.; Yong, A. M.; Wong, S. Y.; Zhang, X.-H.; Tan, H.; Chang, A. Y.; Li, X.; Wang, J. Chem. Commun. 2011, 47, 11615–11617. 136. Yuan, Y. H.; Liu, Z. X.; Li, R. S.; Zou, H. Y.; Lin, M.; Liu, H.; Huang, C. Z. Nanoscale 2016, 8, 6770–6776. 137. Shang, J.; Ma, L.; Li, J.; Ai, W.; Yu, T.; Gurzadyan, G. G. Sci. Rep. 2012, 2, 792. 138. Kwon, W.; Do, S.; Kim, J.-H.; Jeong, M. S.; Rhee, S.-W. Sci. Rep. 2015, 5. 139. Du, F.; Yuan, J.; Zhang, M.; Li, J.; Li, Z.; Cao, M.; Chen, J.; Zhang, L.; Liu, X.; Gong, A. RSC Adv. 2014, 4, 37536–37541. 140. Zheng, M.; Ruan, S.; Liu, S.; Sun, T.; Qu, D.; Zhao, H.; Xie, Z.; Gao, H.; Jing, X.; Sun, Z. ACS Nano 2015, 9, 11455–11461. 141. Mewada, A.; Pandey, S.; Thakur, M.; Jadhav, D.; Sharon, M. J. Mater. Chem. B 2014, 2, 698–705. 142. Ding, H.; Du, F.; Liu, P.; Chen, Z.; Shen, J. ACS Appl. Mater. Interfaces 2015, 7, 6889–6897. 143. Liu, D.; Cheng, E.; Yang, Z. NPG Asia Mater. 2011, 3, 109–114. 144. Tan, L. H.; Xing, H.; Lu, Y. Acc. Chem. Res. 2014, 47, 1881–1890. 145. Mirkin, C. A.; Letsinger, R. L.; Mucic, R. C.; Storhoff, J. J. Nature 1996, 382, 607–609. 146. Alivisatos, A. P.; Johnsson, K. P.; Peng, X.; Wilson, T. E.; Loweth, C. J.; Bruchez, M. P.; Schultz, P. G. Nature 1996, 382, 609–611. 147. Chen, J.; Seeman, N. C. Nature 1991, 350, 631–633. 148. Zhang, Y.; Seeman, N. C. J. Am. Chem. Soc. 1994, 116, 1661–1669. 149. Yang, X.; Wenzler, L. A.; Qi, J.; Li, X.; Seeman, N. C. J. Am. Chem. Soc. 1998, 120, 9779–9786. 150. Seeman, N. C. Nature 2003, 421, 427–431. 151. Shih, W. M.; Quispe, J. D.; Joyce, G. F. Nature 2004, 427, 618–621. 152. Erben, C. M.; Goodman, R. P.; Turberfield, A. J. J. Am. Chem. Soc. 2007, 129, 6992–6993. 153. Mastroianni, A. J.; Claridge, S. A.; Alivisatos, A. P. J. Am. Chem. Soc. 2009, 131, 8455–8459. 154. Yan, W.; Xu, L.; Xu, C.; Ma, W.; Kuang, H.; Wang, L.; Kotov, N. A. J. Am. Chem. Soc. 2012, 134, 15114–15121. 150

155. Cecconello, A.; Kahn, J. S.; Lu, C.-H.; Khosravi Khorashad, L.; Govorov, A. O.; Willner, I. J. Am. Chem. Soc. 2016, 138, 9895–9901. 156. Setyawati, M. I.; Kutty, R. V.; Tay, C. Y.; Yuan, X.; Xie, J.; Leong, D. T. ACS Appl. Mater. Interfaces 2014, 6, 21822–21831. 157. Tay, C. Y.; Yuan, L.; Leong, D. T. ACS Nano 2015, 9, 5609–5617. 158. Chen, N.; Qin, S.; Yang, X.; Wang, Q.; Huang, J.; Wang, K. ACS Appl. Mater. Interfaces 2016. 159. Seeman, N. C. J. Theor. Biol. 1982, 99, 237–247. 160. Kallenbach, N. R.; Ma, R.-I.; Seeman, N. C. Nature 1983, 305, 829–831. 161. Fu, T. J.; Seeman, N. C. Biochemistry 1993, 32, 3211–3220. 162. Winfree, E.; Liu, F.; Wenzler, L. A.; Seeman, N. C. Nature 1998, 394, 539–544. 163. Ma, Y.; Yang, X.; Wei, Y.; Yuan, Q. Chin. J. Chem. 2016, 34, 291–298. 164. Feldkamp, U.; Niemeyer, C. M. Angew. Chem., Int. Ed. 2006, 45, 1856–1876. 165. Zhang, X.; Yan, H.; Shen, Z.; Seeman, N. C. J. Am. Chem. Soc. 2002, 124, 12940–12941. 166. Goodman, R. P.; Schaap, I. A. T.; Tardin, C. F.; Erben, C. M.; Berry, R. M.; Schmidt, C. F.; Turberfield, A. J. Science 2005, 310, 1661–1665. 167. He, Y.; Ye, T.; Su, M.; Zhang, C.; Ribbe, A. E.; Jiang, W.; Mao, C. Nature 2008, 452, 198–201. 168. Seeman, N. C. Chem. Biol. 2003, 10, 1151–1159. 169. Schreiber, R.; Do, J.; Roller, E.-M.; Zhang, T.; Schuller, V. J.; Nickels, P. C.; Feldmann, J.; Liedl, T. Nat. Nano 2014, 9, 74–78. 170. Somoza, Á. Angew. Chem., Int. Ed. 2009, 48, 9406–9408. 171. Dietz, H.; Douglas, S. M.; Shih, W. M. Science 2009, 325, 725–730. 172. Liu, J.; Lu, Y. J. Am. Chem. Soc. 2004, 126, 12298–12305. 173. Liu, J.; Lu, Y. Angew. Chem., Int. Ed. 2006, 45, 90–94. 174. Liu, J.; Lu, Y. Anal. Chem. 2004, 76, 1627–1632. 175. Wong, N. Y.; Xing, H.; Tan, L. H.; Lu, Y. J. Am. Chem. Soc. 2013, 135, 2931–2934. 176. Fleming, S.; Ulijn, R. V. Chem. Soc. Rev. 2014, 43, 8150–8177. 177. Wang, J.; Liu, K.; Xing, R.; Yan, X. Chem. Soc. Rev. 2016, 45, 5589–5604. 178. FrederixPim, W. J. M.; Scott, G. G.; Abul-Haija, Y. M.; Kalafatovic, D.; Pappas, C. G.; Javid, N.; Hunt, N. T.; Ulijn, R. V.; Tuttle, T. Nat. Chem. 2015, 7, 30–37. 179. Reches, M.; Gazit, E. Science 2003, 300, 625–627. 180. Kim, J.; Han, T. H.; Kim, Y.-I.; Park, J. S.; Choi, J.; Churchill, D. G.; Kim, S. O.; Ihee, H. Adv. Mater. 2010, 22, 583–587. 181. Yan, X.; Zhu, P.; Li, J. Chem. Soc. Rev. 2010, 39, 1877–1890. 182. Silva, G. A.; Czeisler, C.; Niece, K. L.; Beniash, E.; Harrington, D. A.; Kessler, J. A.; Stupp, S. I. Science 2004, 303, 1352–1355. 183. Rexeisen, E. L.; Fan, W.; Pangburn, T. O.; Taribagil, R. R.; Bates, F. S.; Lodge, T. P.; Tsapatsis, M.; Kokkoli, E. Langmuir 2010, 26, 1953–1959. 184. Cui, H.; Webber, M. J.; Stupp, S. I. Biopolymers 2010, 94, 1–18.

151

185. Gradišar, H.; Božič, S.; Doles, T.; Vengust, D.; Hafner-Bratkovič, I.; Mertelj, A.; Webb, B.; Šali, A.; Klavžar, S.; Jerala, R. Nat. Chem. Biol. 2013, 9, 362–366. 186. Shen, J.; Chen, C.; Fu, W.; Shi, L.; Li, Z. Langmuir 2013, 29, 6271–6278. 187. Padilla, J. E.; Colovos, C.; Yeates, T. O. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 2217–2221. 188. Gradišar, H.; Jerala, R. J. Nanobiotechnol. 2014, 12, DOI: 10.1186/14773155-12-4.

152

Chapter 8

Construction of Bio-Inspired Composites for Bone Tissue Repair Junchao Wei,*,1 Lina Wang,1,2 Lan Liao,3 Jiaolong Wang,3 Yu Han,4 and Jianxun Ding*,4 1College

of Chemistry, Nanchang University, Nanchang 330031, P. R. China of Science, Nanchang Institute of Technology, Nanchang 330029, P. R. China 3Department of Prosthodontics, Affiliated Stomatological Hospital of Nanchang University, Nanchang 330006, P. R. China 4Key Laboratory of Polymer Ecomaterials, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, P. R. China *E-mail: [email protected] (J. Wei); [email protected] (J. Ding) 2College

Materials with excellent mechanical properties and biofunctions are key points of bone tissue repair. With the increase of knowledge about bone tissue structure and the development of nanotechnology, tough materials have been designed to mimic the structure of bone. Based on the structure of bone and nacre, we briefly introduced the factors affecting the mechanical properties of composites and also introduced the most widely used techniques, such as, electrospinning, phase separation, and three-dimensional (3D) printing method, to acquire porous scaffolds. In addition, the biofunctionalization of scaffolds was also introduced in this chapter.

1. Introduction Due to various diseases, accidents and aging of population, bone defect has been a common problem. Autologous bone graft is the gold standard for treating bone defect, however, it also brings a lot of side effects to patients. Although Allograft and Xenograft has brought some promise, sometimes donors’ shortage and immune response limit their application, and thus much work has been carried out to design alternative bone graft materials, which may be composed of © 2017 American Chemical Society

polymers, metals, and ceramics. Tissue engineering, which applies the principles of engineering and life sciences towards the development of biological substitutes to restore, maintain or improve the tissue functions (1, 2), has been an exciting and promising method to repair the bone defect. During the process of tissue engineering, the tissue engineering scaffold is a key factor to affect the functional reconstruction of tissues. The general requirement for a tissue engineering scaffold is that the scaffold materials should not only fill the defect area, but also supply both structure and mechanical support (3), even bio-functions to realize the regeneration of bone (4). And thus, much attention has been paid to the materials used in bone tissue engineering (5, 6). The materials of bone tissue engineering scaffolds or bone tissue regenerative composites should have proper mechanical properties to support the growth of bone tissue, furthermore, as for the load-bearing place, the mechanical properties are the key factors, and this point is very important for bone substitutes or bone fixation devices. Secondly, the porous structure should also be realized for the supplement of the ample space to support the growth of new tissues. Thirdly, enough biofunctions such as biocompatibility, bioactivity, bone conductivity and bone inductivity are also critical factors for materials used in bone tissue engineering. So far, various methods have been developed to fabricate polymer composites used in bone tissue repair. Bio-inspired idea has been widely used to design new materials or scaffolds that can mimic the functions of native bone tissue. However, before designing ideal bio-inspired composites, it is vital to understand the structure of native tissue and also it is much helpful to understand the interactions between polymers, nanofillers, and cells. In this chapter, we firstly give a brief introduction of bone structure and the reason why it is tough and strong, and then materials used in bone tissue repair was introduced, finally, methods about how to construct bone-inspired composites were introduced. This short review may give an idea about how to construct strong and tough materials used in bone tissue engineering, and also give some suggestion on how to design biofunctionalized scaffolds.

2. Materials for Bone Tissue Repair 2.1. Natural Strong and Tough Materials—Bone & Nacre Bone is mainly composed of collagen (mostly type I) and hydroxyapatite (Ca10(PO4)6(OH)2, HA). Both the HA and collagen consists about 95% weight of the total bone and formed a tough, strong and low weight materials. Collagen is a kind of natural polymer, its mechanical strength is very low, while HA is a kind of inorganic bioceramic, its mechanical properties is poor, and much brittle. However, an interesting thing is that the natural bone is typically strong and tough, due to the hierarchical structure of HA and collagen, and especially that the HA and collagen arranged in an order way. Briefly, the HA nanoplates are mostly arranged along its c-axis and arranged parallel to the collagen fibrils, the arrangement repeat periodically (Figure 1). 154

Figure 1. Hierarchical structure of Bone. Reproduced with permission from ref. (7). Copyright 2014, Nature publishing group.

The scheme of bone’s resistance to external force is based on its multiple deformation scales, ranging from nanoscale protein to microscale physiological structure (7, 8). As for the cortical bone, the origins of its fracture resistance ability rise from both intrinsic mechanisms that promote ductility and extrinsic mechanisms that act to shield the growing cracks (8). The intrinsic ductility origins from the smallest length scales, which is mainly from the molecular uncoiling of mineralized collagen. Most importantly, when the load is added to the bone, the stress may transfer between the HA plates and the collagen fibrils, when the stress is too high, fibrillar sliding may happens, which will make the materials tough to resist to the tension. Besides, many other factors contribute to the bones, such as the collagen fibers structure, the phase interaction between HA and collagen, the intermolecular crosslinking, these factors realize the increased strength of bone. These factors make it possible to dissipate energy. So, in order to design strong composites, basic requirements are: realizing the ordered arrangement of fillers, and enhancing the phase compatibility between fillers and polymer matrix. Nacre is another kind of strong materials. It is a brick-and-mortar structure consisting of 95% vol. layered aragonite (CaCO3) plates and a thin layer of protein molecules (Figure 2). Generally, the mechanical properties of both CaCO3 and protein molecules are very poor. However, the mechanical toughness of nacre is three orders of magnitude higher than that of CaCO3. The fracture toughness of CaCO3 is 0.25 Mp·m1/2, while that of nacre is about 10 Mpa·M1/2, nearly 40 times that of CaCO3 (9). The mineral aragonite is brittle, and it provides the strength of nacre. Due to the existence of protein chains (Figure 2a), it is possible for nacre to realize elastic deformation when external load added. The organic molecules work as glue to connect with the plate aragonites and realize the stress distribution, so the nacre shows toughness (Figure 2b). 155

Figure 2. Hierarchical structure of nacre. Structure illusion of nanoparticle glued by protein chains (a), the schemes of its toughness (b) and the brick-and-mortar structure (c). Reproduced with permission from ref. (7). Copyright 2014, Nature publishing group. During recent years, much work has been carried out to prepare nacre or bonelike bioinspired materials. Great progresses have been carried out, most of which are focused on the structure-properties characteristic. The unique structure of bone is not only the arrangement of compositions, but also its complex formation process, in which cells function and involvement place an important role. The exact understanding of the biomineralization process and structure has contributed a lot to the design of polymer composites, although much more things need to be deciphered clearly. Nowadays, although it is difficult to prepare materials that can completely mimic the structure of bone or nacre, bioninspired idea has been used to prepare materials from different points, such mechanical properties, porous structure and biofunctions, which may realize their further application in bone regeneration. 2.2. Materials Used for Bone Tissue Repair There are three kinds of materials used in bone tissue repair: metals, ceramics and polymers (3). They can be used in different parts due to their properties. Metals, such as titanium alloy and stainless alloy which are biocompatible and have good mechanical properties, have been used as bone plates or bone screws, however, this kind of materials always need second operation. Ceramics are most inorganic materials exist in bioactive glass, HA, tricalcium phosphate (TCP) and so on, these materials with bone conductivity or bone inductivity are the most widely investigated inorganic materials, however, these materials are brittle and can not be used in load bearing parts. Polymers with good mechanical properties and excellent biocompatibilities may satisfy the requirements of tissue regeneration, and have been widely used in tissue engineering. According to the source of polymers, they can be classified as natural and synthetic polymers. Generally, both degradable and nondegradable polymers can be used in bone tissue repair. For example, ultra-high molecular 156

weight polyethylene and polyether ether ketone (PEEK) have been used as bone substitute or artificial kneel, however, when tissue engineering is considered, biodegradable polymers show much more potential, such as collagen, gelatin, poly(L-lactide) (PLLA), poly(lactide-co-glycolide) (PLGA) and so on, have been widely used. However, pure polymers can only mimic part functions of native tissues, they still lack of enough biofunctions to induce bone formation, or biomechanical properties to satisfy bone loading requirements, so polymer composites, especially polymers complexed with inorganic bioceramics have been widely designed. An ideal polymer composite may have hierarchical structure and possess the advantages of different composites, realizing a synergistic effect to put forward the applications in bone repair. Up to now, various polymer composites have been used in bone tissue repair field, such as polymer-polymer blends and polymer-inorganic nanocomposites. The most widely investigated bioceramic/polymer composites are hydroxyapatite/polymer composites. Collage, gelatin, PLLA, PLGA and their HA composites have been widely investigated or reviewed (10). Besides, multicomponent composites contains more than two kinds of polymers or two kinds of inorganic component are also well investigated due to their combination of multi-advantages of different composites and show much better synergistic effects. Although much progress has been achieved, there still need a long way to prepare ideal composites which can mimic the properties of natural tissue and realize the rapid bone substitute or rebuilt of bone tissue.

3. How To Mimic the Properties of Bone During the bone repair process, five special targets should be considered, osteogenesis, vascularization, growth factors, mechanical environment and osteoconductive scaffolds (3). To realize successful bone regeneration, at least three of the targets should be involved. Thus, it is vital to design materials with proper mechanical properties, especially for load-bearing parts. The excellent properties are inevitable, besides, the materials should also have a proper structure and bioproperties to support the growth of new tissues. 3.1. Construction of Composites with Excellent Mechanical Properties Bone tissue has excellent mechanical properties, thus many efforts have been focused on preparing bone inspired composites, for one purpose to obtain strong and light weight materials, and another purpose is to prepare bone regenerative materials or related medical devices, especially for load bearing bone repair, the mechanical properties are always the vital factors. Up to now, plenty polymer composites have been designed, however, mostly, the mechanical properties are far from their theoretical value. A critical challenge is to transfer the excellent mechanical properties from nanoscale to macroscale (11). As for filler-reinforced composites, the obtained properties are always far from their ideal results. The key problem is that they could not realize the homogeneous dipsersion of nanofillers and easily control their arrangement. 157

Another factor is that the phase interaction between the fillers and polymer matrix are not as strong as the natural bone composition (12). Nicholas Kotov’s research demonstrated that it is possible to produce composites with properties that can compare with the theoretical values by tuning the spatial and orientation of nanofillers (11). They used a bottom up method called layer-by-layer assembly to prepare a kind of poly(vinyl alcohol)/Montmorillonite (PVA/MTM) composites. With the LBL method, an interlayer structure was formed which can mimic the structure of nacre, due to the controlled structure organization, the clay platelets in polymer matrix arranged orderly, due to much hydroxyl groups of PVA chains and SiO4 groups in MTM, the phase interaction between PVA and MTM are strong. Besides, when the film was crosslinked with GA, the interaction will be much stronger, and the mechanical properties can arrive to its theoretical value. The final tensile strength of the crosslinked PVA/MTM was 400±40Mpa, and the modulus was 106±11Gpa. By tuning the arrangement of nanofillers, various strong bio-inspired composites have been designed. Recently, Robert O Ritchie has reported a kind of hydroxyapatite/poly(methyl methacryalate) (HA/PMMA) composites with layered structure (preparation scheme is shown in Figure 3) (13), which can mimic the nacre structure and work as a kind of tissue engineering scaffolds. The strength, elastic stiffness and work of fracture were 100 Mpa, 20 Gpa and 2075 J·m-2, respectively. These results are nearly two orders of magnitude than monolithic HA.

Figure 3. Schematic illustration of fabrication of HA/PMMA composite with nacre-mimetic structure. Reproduced with permission from ref (13). Copyright 2015, John wiley and Sons.

158

Another important factor to enhance the mechanical properties of polymer composites is to improve the phase compatibility of nanofillers and polymer matrix. Up to now, many works have been reported to tune the surface properties of nanoparticles by grafting polymer chains on the surface of nanofillers (14–18). As for bone tissue regenerative materials, bioceramics such as hydroxyapatite, bioactive glass are most important to endow polymers with bone conductivity or bone inductivity. However, pure bioceramic particles are brittle and their phase interaction with polymers are always too weak, and thus polymer grafted bioceramics have been prepared. For example, Chen’s group have used ring opening polymerization method to graft polymers on the surface of HA (19–21), PLLA was grafted on the surface of HA and the tensile strength of PLLA-g-HA/PLLA (75 Mpa) was much higher than that of pure HA/PLLA composite (less than 60 Mpa). Besides, PLLA-g-HA can also be used to prepare porous scaffold and showed excellent osteogenesis properties (22). Wei used poly(benzly-l-glutamate) to modify the surface of HA not only change its biocompatibility, but also increase its phase compatibility with Polymer matrix, and the results showed that only 0.3% content can make the mechanical properties increased a lot (18). Besides, Wei’s group also used PBLG to modify the surface of SiO2@GO hybrid and then prepared its PLLA composites, the results showed that the tensile strength of PLLA composites can arrive to 88.9 Mpa, much higher than that of pure PLLA, PLLA/GO or PLLA/SiO2 (23).

3.2. Construction of Porous Scaffolds for Bone Tissue Repair The natural bone is porous structure, while tissue engineering scaffolds also need porous structure to support the growth of tissues. Many methods, such as electrospinning, phase separation and 3-D printing have been used to prepare porous structure scaffolds to realize the regeneration of bone tissue. Here, we will give a brief introduction about these methods.

3.2.1. Electrospinning Electrospinning method has been widely used to construct fibrous porous scaffold to mimic the fibrillar architecture of extracellular matrix (ECM). These fibrous biomimetic scaffolds can supply microenvironment for the regenerative of bone tissue. The basic progress of electrospinning contains three parts (Figure 4a). Firstly, polymer solutions were extruded from a conductive spinneret, and then voltage was applied between the spinneret and grounded collector. When the electric potential in the polymer solution overcomes the surface tension of polymer solution droplet, the droplet will eject to the collector, during this period, the solvent will evaporate, and polymer fibers will be collected on the collector.

159

Figure 4. Schemes of basic electrospinning process (a) and fibers with different orientation, random fiber (b), parallel fiber (c), crossed fiber (d), patterned fiber net (e), 3D fibrous stack (f), wavy fiber (g), helical fiber (h), and twisted fibers (i). Reproduced with permission from ref. (26). Copyright 2014, Elsevier.

The electrospinning fibers have showed potential applications in tissue engineering, the morphologies of electrospun fibers, such as fiber size, porosity, fiber orientation may affect the attached cells behaviors, thus many efforts have been worked to tune the structure or morphologies of fibers (24). Firstly, the polymer concentration is a key factor, it has to exceed a critical concentration so that enough polymer chains entangle within the polymer solution, then polymer fibers can be formed via electrospinning. Otherwise, dilute polymer solutions will spray into beads or uniform polymer fibers with much more beads aggregate. It is also important to choose polymers with proper molecular weight. If the molecular weight is too low, the polymer chains can not entangle well and it is difficult to form fibers. If the molecular weight is too high, the polymer entangles a lot and increases the solution viscosity, which means the surface tension of droplet is much higher, and thus it is also difficult to form fibers, unless high pressure voltage was used. So it is vital importance to tune the polymer solution, some times other additive, such as surfactant or amphiphilic molecules should be added. Besides, the solvent can also be a critical factor to affect the solution viscosity, proper solvent, sometimes mix solution is a prerequisite to obtain designed fibers (25). In addition, with different kinds of jet or needle, and different kinds of collectors can realize different morphologies electrospun fibers, 160

fibers with different morphologies or arrangement such as random or orientation aligned fibers can be prepared via different technologies (26). Electrospinning as an important method to prepare new tissue engineering is not only used to prepare polymer materials, up to now, various polymer blend or polymer-inorganic composites have been prepared, Such as PLA/PCL (27), PLLA grafted hydroxyapatite/PLLA (28), gelatine/chitoasan/hydroxyapatite/graphene oxide. With the increasing requirements for tissue engineering, various functional molecules (growth factor, proteins) and drugs can be loaded with fibers and enrich the properties of tissue engineering scaffolds (29, 30) (see part 3.3).

3.2.2. Phase Separation Phase separation is a simple method to prepare porous or fibrous structure that can mimic the native structure of ECM and has been widely used in tissue engineering scaffolds. The basic principle of phase separation is based on the thermally unstability of polymer solutions and tend to separate into two or more phases under certain conditions (31–33), briefly, either exposuring the polymer solution to a immiscible solution or cooling down to a certain temperature, phase separation would happen and form a polymer-rich phase and a polymer-poor phase. The basic procedure of phase separation includes polymer dissolution, phase separation and gelation, solvent extraction, freezing and freeze drying (31). By tuning the parameters of phase separation, porous scaffolds with different pore sizes and shapes can be obtained (34), for example, by tuning the temperature below or above the polymer solutions, both closed and open pores can be obtained, respectively; Under lower temperature, the solvent can crystallize quickly and the crystal size will be smaller, when the solvent crystal was removed, the scaffolds, with smaller pores will be obtained, otherwise, scaffolds with large pores can be formed. even the cooling rate and freezing temperature may have also vital affection on the pore structure (35). For example, when frozen at -80 °C and 190 °C, PLLA scaffolds with different pore sizes 47±8 and 22±4μm were prepared (36), many other kinds of scaffolds, such as chitosan, PLLA/chitosan have also been prepared by phase separation method (37). Sometimes it is difficult to realize precise control of the microstructure, some modified methods or combination of phase separation methods and other methods, such as solvent casting, porogen leaching and supercritical method have been used (38), polymer solution can be cast around salt sugar and other porogen. By tunning the size of porogen agents, scaffolds with different porous structures can be obtained, and thus it is vital important in tissue engineering, due to that the pore size has an important effect on the cell behavior and bone formation (39).

3.2.3. 3D Printing Although various methods such as elecrospinning, phase separation solvent casting, and salt leaching and many other methods have been designed to prepare porous scaffolds, it is still a challenge to precise control the hierarchical structure 161

such as pore size, shape and pore interconnectivity. Inspired by the complex structure of biological materials, especially some native tissue structure such as bone or nacre, 3D printing has been used to fabricate various scaffolds with intricate microstructure (40, 41). 3D printing applies additive manufacturing approaches, combine computer assisted design (CAD) software and printing machine together to prepare products by a layer-by-layer method, and the basic procedure has been introduced in many references (42, 43). The basic characteristics of 3D printing can be described as follows (44): A) supply of building blocks or raw materials in a continuous or stepwise method. B) A programme that contains the structure information should be supplied to determine the assembly of materials. C) Mechanisms or equipments that can fulfill the programme and control the assembly of building blocks. D) A consolidation step to fix the deposited materials and make the printed structure as designed. Up to now, various technologies, such as deposition modeling, stereolithography, ink-jet printing have been used to prepare tissue engineering scaffolds and can easily realize on-demand fabrication of customized products with precise structures (44, 45). For bone tissue engineering, 3D printing is very convenient to construct scaffolds with desired structure. Meantime, with the development of 3D technology, various materials, such as ceramic, metallic, polymers and polymer composites can be used for 3D print (46, 47). Cho used sterolithography method and prepared three-dimensional (3D) porous scaffolds of poly(propylene fumarate)/diethyl fumarate (PPF/DEF) (48), which have sufficient mechanical stability and are non-toxic. After post-modification, the scaffold can enhance the adhesion and proliferiation of MC3T3-E1 preosteoblast cells, showing potential application in bone tissue repair. Polyester, such as PLA, PCL, PLGA have been widely used in bone tissue engineering, when combined with 3D technology, various scaffolds have been prepared with these biodegradable polymers (49). Inorganic materials, such as HA, α-TCP, β-TCP and bioactive glass have also been used as 3D printing materials to build bone tissue engineering scaffolds (50). In order to combine the bone inductivity of bioceramics and the biodegradability of polymers, bioceramics, HA, TCP, bioactive glass have been used as fillers to prepare polymer composites used for bone tissue engineering. For examples, PLLA/HA, chitosan/HA, collagen/HA and alginate/bioactive glass have been prepared with 3-D printing (51–53). Via 3-d printing method, scaffolds which can mimic the structure of native bone tissue can be prepared easily; however, the mechanical properties of porous scaffolds are still challenges. The porous structure always results in low mechanical properties, and always used in non-load bearing place. However, by tuning the delicate printing method, it is a good method to realize the requirements of complicated structures. For example, PolyJet 3-D printing method based on ink jet technology can realize deposition of multi-materials, which enables the possibility to prepare both strong and tough materials. This method would be much useful to prepare polymer composites that can mimic the bone (44, 54). Buelher had used multi-material 3D printing to print composites with bone-inspired topologies that exhibit superior fractural mechanical properties, and the computational model predictions of the fracture behaviors and trends in mechanical properties are in accordant with the experimental results, 162

demonstrating that it is possible to design composite materials and then use 3-D printing to synthesize the desired materials with expected mechanical performance (54). 3-D printing will have much more promising perspective in tissue engineering. However, many critic requirements, such as specific technical, material and cellar aspects of the printing process will bring more challenges (43). The increasing resolution requirements may make 3-D printed products mimic the tissue structure more exactly. The ideal materials should not only be biocompatible, but also can be easily manipulated by the printing technology to acquire complex 3-D structure and maintain its bioproperties. The cells used should be easily available, and can reproduce all the functions of tissue or organ system (43), so multidiscipline should forged together to meet the challenges and thus further improve the application of 3-D printing in tissue engineering. 3.3. Composites with Special Biofunctions Although most biocompatible polymers or polymer composites can be made into scaffolds that can mimic the structure of native tissue, but it is difficult to mimic the biofunctions. The cell functions and growth factors play critical roles in the process of tissue regeneration, so various methods have been used to prepare composites with special biofunctions. Firstly, the native biomacromolecues or polymers found in the ECM can be used as ideal scaffolds to mimic the biofucntions. Collagen, the most organic content of bone, has been used for a long time. It is not only used alone, but also blended with various polymers or inorganic particles. Other materials, such as hydroxyapatite, the inorganic component of bone, have also been widely used in tissue engineering or bone repair, due to its bone conductivity or bone inductivity properties. In order to further improve the biofunctions, increasing works have been focused on the combination of scaffolds and biomacromolecules, such as protein or growth factors (BMP, IGF). The growth factors can control osteogenesis, bone tissue regeneration and ECM formation via recruiting and differentiation osteoprogenitor cells to specific lineages (55). So it is critical to incorporate protein or growth factors in the tissue engineering scaffolds. Generally, there are two kinds of method to prepare biomacromolecules contained scaffolds. One method is pre-treat method, which means that biomolecules were added into the scaffolds while the preparation process. The other method can be called post-treatment method, which means the biomacromolecules were adsorbed or anchored on the scaffolds by physical or chemical interactions. For example, biomolecules contained electrospinning scaffolds can be prepared via different blend electrospinning, coaxial electrospinning and covalent immobilization methods (56), and these methods can realize the functional molecules be adsorbed or covalently anchored on the surface of fibers or encapsulated within the fibers. The configurations of biomolecules have key effects on their signal transduction activity, and thus it is critical to keep the structure of growth factors. Mussel-inspired method with 3,4-dihydroxyphenethylamine (DOPA) contained peptide has been widely used, recently Ito’s group designed a method to prepare 163

DOPA containing insulin-like growth factor-1(IGF-1) and anchor the IGF-1 on the surface titanium (Figure 5). This method can be used to prepare novel cell-growth enhancing materials and thus have much potential in tissue engineering (57). Zhang’s group also used the mussel-inspired method and immobilized collagen mimetic peptide and osteogenic growth peptide on the surface of L-lactic acid oligomer modified hydroxyapatite/Poly(lactide-co-glycoclide) composite film (58). The results demonstrated that it is a good method to immobilize biomacromolecues on the surface of implants with bioinspired method and realize their enhanced osteointegration of bone implants.

Figure 5. Preparation of DOPA contained IGF-1 derivatives and its immobilization on the surface of titanium. Reproduced with permission from ref (57). Copyright 2016, John wiley and Sons. To encapsulate special cells in the scaffolds is also a good method to improve the biofunction of scaffolds, for example bio-printing has been designed to prepare cells contained scaffolds (43), and it can combine the biocompatible materials, cells and other components into a functional living tissue and thus will have much more applications.

4. Perspective In this chapter, we briefly introduce the structure of bone and how to construct the bio-inspired composites used in the field of bone tissue repair. Due to the requirements of bone tissue regenerative, tough materials with excellent mechanical properties and scaffolds with porous structure and special biomolecules are needed in bone tissue repair. Although many progresses 164

have been achieved, there still need a long way to prepare ideal scaffolds or substitutes that can real mimic the bone structure or properties. However, by further understanding the structure of native bone and the interactions between materials and tissues, new hints maybe brought out to the materials design, and the challenges will be overcame via the combination of materials, engineering, biology, medicine, and others. To our opinion, in the next few years, more interests will be placed on the design of new functional materials with special biofunctions, or prepared new functional polymer composites, which may be used for bone implant medical devices or manufactured into scaffolds with desired structure. As a basic requirement, how to realize enhancements of both toughness and strength, and make the mechanical properties satisfied for bone tissue will still be an interesting point. New preparation method or modification method will be continuously investigated to realize the multi-functions of tissue engineering scaffolds, such as how to keep the long term stability of growth factors or realize its controlled release in the scaffolds, and how to control the scaffold structure exactly, especially in the nanoscale level. Furthermore, reproducing all the functions of living tissue or organs is still a huge challenge, by mimicking the structure of bone tissue, the structures of scaffolds and their affection on the biofunctions will be an interesting topic, scaffolds loaded with living cells will arouse more interest to mimic the biofunction of living tissue, and have much more potential application in regenerative medicine.

Acknowledgments This work was financially supported by the National Natural Science Foundation of China (Nos. 51463013, 51663017, 81660444, and 51673187), the Natural Science Foundation of Jiangxi Province of China (No. 20151BAB206011), the Health and Family Planning Commission Science Foundation of Jiangxi Province of China (No. 20161082), and Natural Science Foundation of Nanchang Institute of Technology (No. 2012KJ028).

References 1. 2. 3. 4. 5. 6. 7. 8.

Tian, H.; Tang, Z.; Zhuang, X.; Chen, X.; Jing, X. Prog. Polym. Sci. 2012, 37, 237–280. Langer, R.; Vacanti, J. P. Science 1993, 260, 920–926. Jahan, K.; Tabrizian, M. Biomater. Sci. 2016, 4, 25–39. Carrow, J. K.; Gaharwar, A. K. Macromol. Chem. Phys. 2015, 216, 248–264. Bose, S.; Tarafder, S. Acta Biomater. 2012, 8, 1401–1421. Pina, S.; Oliveira, J. M.; Reis, R. L. Adv. Mater. 2015, 27, 1143–1169. Wegst, U. G. K.; Bai, H.; Saiz, E.; Tomsia, A. P.; Ritchie, R. O. Nat. Mater. 2015, 14, 23–26. Launey, M. E.; Buehler, M. J.; Ritchie, R. O. Annu. Rev. Mater. Res. 2010, 40, 25–53. 165

9. 10. 11.

12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

Barthelat, F.; Tang, H.; Zavattieri, P. D.; Li, C. M.; Espinosa, H. D. J. Mech. Phys. Solids 2007, 55, 306–337. Wang, J.; Wang, L.; Zhou, Z.; Lai, H.; Xu, P.; Liao, L.; Wei, J. Polymers 2016, 8, 115. Podsiadlo, P.; Kaushik, A. K.; Arruda, E. M.; Waas, A. M.; Shim, B. S.; Xu, J.; Nandivada, H.; Pumplin, B. G.; Lahann, J.; Ramamoorthy, A.; Kotov, N. A. Science 2007, 318, 80–83. Bonderer, L. J.; Studart, A. R.; Gauckler, L. J. Science 2008, 319, 1069–1073. Bai, H.; Walsh, F.; Gludovatz, B.; Delattre, B.; Huang, C.; Chen, Y.; Tomsia, A. P.; Ritchie, R. O. Adv. Mater. 2016, 28, 50–56. Li, W.; Xu, Z.; Chen, L.; Shan, M.; Tian, X.; Yang, C.; Lv, H.; Qian, X. Chem. Eng. J. 2014, 237, 291–299. Kuila, T.; Bose, S.; Mishra, A. K.; Khanra, P.; Kim, N. H.; Lee, J. H. Prog. Mater. Sci. 2012, 57, 1061–1105. Wei, J.; He, P.; Liu, A.; Chen, X.; Wang, X.; Jing, X. Surface Macromol. Biosci. 2009, 9, 1237–1246. Wei, J.; Liu, A.; Chen, L.; Zhang, P.; Chen, X.; Jing, X. Macromol. Biosci. 2009, 9, 631–638. Wei, J.; Dai, Y.; Chen, Y.; Chen, X. Sci. China-Chem. 2011, 54, 431–437. Hong, Z. K.; Qiu, X. Y.; Sun, J. R.; Deng, M. X.; Chen, X. S.; Jing, X. B. Polymer 2004, 45, 6699–6706. Hong, Z. K.; Zhang, P. B.; He, C. L.; Qiu, X. Y.; Liu, A. X.; Chen, L.; Chen, X. S.; Jing, X. B. Biomaterials 2005, 26, 6296–6304. Qiu, X. Y.; Hong, Z. K.; Hu, J. L.; Chen, L.; Chen, X. S.; Jing, X. B. Biomacromolecules 2005, 6, 1193–1199. Zhang, P.; Hong, Z.; Yu, T.; Chen, X.; Jing, X. Biomaterials 2009, 30, 58–70. Guo-Wang, P. Y.; Ding, J. X.; Guo, W.; Wu, H. Y.; Wei, J. C.; Dai, Y. F.; Deng, F. J. RSC Adv. 2016, 6, 5688–5694. Lin, J.; Wang, X.; Ding, B.; Yu, J.; Sun, G.; Wang, M. Crit. Rev. Solid State Mater. Sci. 2012, 37, 94–114. Lin, J.; Ding, B.; Yu, J.; Hsieh, Y. ACS Appl. Mater. Interfaces 2010, 2, 521–528. Sun, B.; Long, Y. Z.; Zhang, H. D.; Li, M. M.; Duvail, J. L.; Jiang, X. Y.; Yin, H. L. Prog. Polym. Sci. 2014, 39, 862–890. Li, H. T.; Qiao, T. K.; Song, P.; Guo, H. L.; Song, X. F.; Zhang, B. C.; Chen, X. S. J. Biomater. Sci., Polym. Ed. 2015, 26, 420–432. Wei, J. C.; Guo-Wang, P.; Han, Q.; Ding, J. X.; Chen, X. S. J. Controlled Release 2015, 213, E62–63. Place, L. W.; Sekyi, M.; Taussig, J.; Kipper, M. J. Macromol. Biosci. 2016, 16, 371–380. Braghirolli, D. I.; Steffens, D.; Pranke, P. Drug Discovery Today 2014, 19, 743–753. Holzwarth, J. M.; Ma, P. X. Biomaterials 2011, 32, 9622–9629. Ma, P. X. Mater. Today 2004, 7, 30–40. Ma, P. X. Adv. Drug Delivery Rev. 2008, 60, 184–198. Akbarzadeh, R.; Yousefi, A.-M. J. Biomed. Mater. Res., Part B 2014, 102, 1304–1315. 166

35. Mannella, G. A.; Carfì Pavia, F.; Conoscenti, G.; La Carrubba, V.; Brucato, V. J. Polym. Sci., Part B: Polym. Phys. 2014, 52, 979–983. 36. Budyanto, L.; Goh, Y. Q.; Ooi, C. P. J. Mater. Sci.: Mater. Med. 2009, 20, 105–111. 37. Salehi, M.; Nosar, M. N.; Amani, A.; Azami, M.; Tavakol, S.; Ghanbari, H. Int. J. Polym. Mater. Polym. Biomater. 2015, 64, 675–682. 38. Salerno, A.; Fernandez-Gutierrez, M.; Roman del Barrio, J. S.; Domingo, C. J. Supercrit. Fluids 2015, 97, 238–246. 39. Karageorgiou, V.; Kaplan, D. Biomaterials 2005, 26, 5474–5491. 40. Martin, J. J.; Fiore, B. E.; Erb, R. M. Nat. Commun. 2015, 6, 8641. 41. Le Ferrand, H.; Bouville, F.; Niebel, T. P.; Studart, A. R. Nat. Mater. 2015, 14, 1172–1179. 42. Berman, B. Business Horizons 2012, 55, 155–162. 43. Murphy, S. V.; Atala, A. Nat. Biotechnol. 2014, 32, 773–785. 44. Studart, A. R. Chem. Soc. Rev. 2016, 45, 359–376. 45. He, C. L.; Tang, Z. H.; Tian, H. Y.; Chen, X. S. Acta Polym. Sin. 2013, 722–732. 46. Bose, S.; Vahabzadeh, S.; Bandyopadhyay, A. Mater. Today 2013, 16, 496–504. 47. Guvendiren, M.; Molde, J.; Soares, R. M. D.; Kohn, J. ACS Biomater. Sci. Eng. 2016, 2, 1679–1693. 48. Shin, J. H.; Lee, J. W.; Jung, J. H.; Cho, D.-W.; Lim, G. J. Mater. Sci. 2011, 46, 5282–5287. 49. Serra, T.; Planell, J. A.; Navarro, M. Acta Biomater. 2013, 9, 5521–5530. 50. Brunello, G.; Sivolella, S.; Meneghello, R.; Ferroni, L.; Gardin, C.; Piattelli, A.; Zavan, B.; Bressan, E. Biotechnol. Adv. 2016, 34, 740–753. 51. Lin, K.-F.; He, S.; Song, Y.; Wang, C.-M.; Gao, Y.; Li, J.-Q.; Tang, P.; Wang, Z.; Bi, L.; Pei, G.-X. ACS Appl. Mater. Interfaces 2016, 8, 6905–6916. 52. Li, X.; Cui, R.; Sun, L.; Aifantis, K. E.; Fan, Y.; Feng, Q.; Cui, F.; Watari, F. Int. J. Polym. Sci. 2014, 829145. 53. Yongxiang, L.; Chengtie, W.; Anja, L.; Michael, G. Biofabrication 2013, 5, 015005. 54. Dimas, L. S.; Bratzel, G. H.; Eylon, I.; Buehler, M. J. Adv. Funct. Mater. 2013, 23, 4629–4638. 55. Bose, S.; Roy, M.; Bandyopadhyay, A. Trends Biotechnol. 2012, 30, 546–554. 56. Ji, W.; Sun, Y.; Yang, F.; van den Beucken, J. J. J. P.; Fan, M.; Chen, Z.; Jansen, J. A. Pharm. Res. 2011, 28, 1259–1272. 57. Zhang, C.; Miyatake, H.; Wang, Y.; Inaba, T.; Wang, Y.; Zhang, P.; Ito, Y. Angew. Chem. 2016, 128 (38), 11619–11623. 58. Wang, Z.; Chen, L.; Wang, Y.; Chen, X.; Zhang, P. ACS Appl. Mater. Interfaces 2016, 8 (40), 26559–26569.

167

Chapter 9

CNT-Based and MSN-Based Organic/Inorganic Hybrid Nanocomposites for Biomedical Applications Jiemei Zhou,1 Jiaoyang Li,1 Decheng Wu,2 and Chunyan Hong1,* 1CAS

Key Laboratory of Soft Matter Chemistry, Department of Polymer Science and Engineering, University of Science and Technology of China, Hefei, Anhui, 230026, P. R. China 2Beijing National Laboratory for Molecular Sciences, State Key Laboratory of Polymer Physics and Chemistry, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, P. R. China *E-mail: [email protected]

The advancement of nanotechnology opens new and fantastic area of biomaterials for biomedical applications. The organic/inorganic hybrid nanocomposites have attracted growing attention of researchers due to their excellent properties. This review highlights and addresses issues related to recent research on carbon nanotubes (CNTs) based hybrid nanocomposites and mesoporous silica nanoparticles (MSNs) based hybrid nanocomposites for biomedical applications. In this review, the modification of CNTs and applications of CNT-based hybrid nanocomposites in cancer treatment and regenerative medicine are introduced. Then, the functionalization of MSNs and applications of MSN-based hybrid nanocomposites in drug delivery and bio-imaging are described. Prospects of CNT-based and MSN-based hybrid nanocomposites in biomedical field are also envisioned.

Introduction In the past decades, the rapid development of nanotechnology leads to the significant advancement of diverse nanoscale biomaterials, such as polymeric nanoparticles, polymer-protein conjugates, micelles, liposomes, inorganic © 2017 American Chemical Society

nanoparticles and hyperbranched polymers (1–15), which have aroused wide research interest for biomedical applications. Studies show that traditional drugs fail to meet the promising therapeutics in some cases, usually because of the disadvantages such as poor solubility, rapid clearance from body environments and nonselective treatment of diseased tissues in clinical application. While, the nanoscale drug delivery systems demonstrate higher accumulation in diseased tissues by virtue of better pharmacokinetics and the enhanced permeation and retention (EPR) effect (16, 17). Besides, appropriate surface modification of drug delivery systems with ligands which have specific interaction with diseased tissues can effectively promote the selectivity of treatment (18, 19). Nanobiomaterials have been extensively explored not only in drug/gene delivery systems, but also in lots of other biomedical-related fields, for example probes for diagnosis, contrast agents for bio-imaging, repair materials for tissue engineering, etc. (20–26). Nanomaterials can be roughly divided into three categories: organic, inorganic and organic/inorganic hybrid nanocomposites. The organic materials are relatively mature in clinical translations after decades of development, and several nanotechnology-based organic drug delivery systems (liposomes, virosomes, albumins, etc.) have been used in clinic for tumor chemotherapy, such as DaunoXome, Myocet, Doxil, Abroxane, etc. (27, 28) Polymeric micelles and polymer gels have also been widely investigated with great promise as intercellular delivery carriers as well as cellular imaging platforms (14, 15). However, the disadvantages of organic drug delivery systems, including structural instability and low drug-loading capacity, hinder the development and further clinical translations (29). Inorganic materials, such as MSNs, magnetic nanoparticles, metallic nanoparticles and CNTs, attract increasing attention of researchers, due to the unique properties of physicochemical stability, magnetism, fluorescence, plasma absorption, etc. (30–34). Nevertheless, the inorganic materials themselves can hardly meet the requirement of good dispersion in physiological media, targeted transportation, stimuli-responsive drug release, etc. Hence, the organic/inorganic hybrid nanocomposites were developed to combine the merits of both organic and inorganic nanomaterials for wider applications in biomedical filed (11). There are abundant kinds of organic/inorganic hybrid nanocomposites. MSNs and CNTs are popular to design the hybrid nanomaterials, because lots of reports have demonstrated the safety and biocompatibility of MSNs and CNTs for biomaterial application (35–37). Herein, this review focuses on MSN-based and CNT-based hybrid nanocomposites for the biomedical applications. The applications of CNT-based hybrid nanocomposites in cancer treatment and regenerative medicine are discussed. The applications of MSN-based hybrid nanocomposites in drug delivery and bio-imaging are also reviewed.

CNT-Based Hybrid Nanocomposites as Biomaterials CNTs have attracted great attention in various fields and generated significant influence in the materials research community since the discovery in 1991 by Sumio Iijima (38). As sp2 carbon nanomaterials, CNTs can be described as 170

cylindrical tubes by rolling up graphene. Single rolled layer of graphene forms single-walled carbon nanotube (SWNT) with a diameter between 0.4 and 2 nm, while rolled multi-layered graphene forms multi-walled carbon nanotube (MWNT) with a diameter between 2 and 100 nm (39). CNTs possess many unique properties, including high tensile strength, ultralight weight, excellent transport conductivity, physical and chemical stability, high aspect ratio, etc. (40) Owing to the large surface area, CNTs can be functionalized or conjugated with various molecules or polymers. The hollow structure of CNTs can result in excellent drug loading capacity. The small size is likely to facilitate the reaction with living organisms or living cells. All these characteristics make CNTs a promising material in various biomedical applications (41). A wide range of biomedical applications about CNTs have been investigated for therapeutic purposes. For example, cancer diagnosis and treatment, regenerative medicine and implant materials (42–50). It is likely that the CNTs in biological applications would be functionalized-CNTs or CNT-based hybrid composites (34). In the drug or gene delivery systems, CNTs usually serve as a platform to bind with drugs, genes and peptides for the treatment of disease. CNTs are found to work excellently as scaffold materials when they are conjugated to polymeric materials for nerve and bone tissue regeneration (41). Furthermore, the modification with various chemical functional groups can increase the dispersibility, reduce the cytotoxicity of CNTs and add new characteristics. Thus, the functionalization of CNTs can greatly expand their potential fields of application. Modification of CNTs Due to the inherent insolubility in aqueous solvents, pristine CNTs are incompatible with the biological systems. One efficient way to improve the solubility is to modify the surface of CNTs by hydrophilic groups or polymer chains (35, 51). The functionalization methods mainly include covalent attachment and non-covalent attachment. Covalent functionalization may be described as a firmly chemical grafting of molecules onto the surface of CNTs. The versatile reaction is oxidation under strong acid conditions to form carboxyl groups on the walls of CNTs, which can react with alcohol or amine to link CNTs with various molecules (52). Another abundant covalent modification is based on 1,3-dipolar and [4+2] cycloaddition (53, 54). In our study (55), CNTs were modified with thiol-reactive species by the reaction between oxidized CNTs and S-(2-aminoethylthio)-2-thiopyridine. Then, poly(N-isopropylacrylamide) (PNIPAAm) was grafted to CNTs through thiol-coupling reaction. The prepared PNIPAAm-CNT conjugates had temperature-responsive PNIPAAm chain, and disulfide linkages between PNIPAAm and CNTs which were sensitive to bio-stimuli such as glutathione, therefore these polymer-CNT conjugates exhibited dual-responsiveness. You et al. (56) attached hyperbranched poly(amidoamine) (PAMAM) onto the surface of CNTs via a multi-step Michael addition of 1-(2-aminoethyl) piperazine (AEPZ) and N,N′-methylene bisa-crylamide (MBA) via the “grafting from” 171

strategy (Scheme 1). The amido and amine units linked on CNTs could facilitate the assembly of CNTs with other functional polymers via hydrogen bonds. The external ultrasound-stimulus was found to induce the assembly of functionalized MWNTs with linear PAMAM into a smart carbon nanotube gel at room temperature. The formed carbon nanotube gels are not only responsive to some mechanical stimuli such as vigorous agitating, but also responsive to some chemical stimuli such as water and acids. The sol-gel switching can be easily realized via heating and ultrasonicating. The new multi-responsive carbon nanotube gels may be useful in sensors and nano-medicine. Giambastiani et al. (57) reported the functionalization of MWNTs by the 1,3-dipolar cycloaddition of a cyclic nitrone. It was found that the diffuse structural defects to the sp2 carbon lattice acted as preferential reactive sites for cycloaddition, and this organic functionalization yielded materials with good solubility in DMF (close to 10 mg per mL of DMF).

Scheme 1. Outline of growing hyperbranched poly(amido amine) from carbon nanotubes. Reproduced from (56). Copyright 2009, Royal Society of Chemistry.

Non-covalent functionalization of CNTs depends on the physical interactions like van der Waals interaction, π-π stacking interaction or electrostatic adsorption. In the work by Jo et al. (58), oligothiophene-terminated poly(ethylene glycol) (TN-PEG) was synthesized and its ability to stabilize aqueous CNT dispersions 172

was examined. It was observed that the TN-PEG could be strongly absorbed onto the surface of CNTs via a strong π-π stacking interaction and help to disperse CNTs in aqueous media well. Conveniently, only gentle sonication was needed in this non-covalent coating. Kane and co-workers solubilized SWNTs in water using commercially available proteins through noncovalent interactions. This protein-mediated solubilization of CNTs in water was important for biomedical applications (59). In addition to the above mentioned, Zhang and co-workers (60) proposed the strategy for surface modification of CNTs via combination of mussel inspired chemistry and chain transfer free radical polymerization. Pristine CNTs were coated with polydopamine (PDA), which was formed via self-polymerization of dopamine in alkaline conditions. Due to the strong adhesion of PDA to CNTs and mild reaction condition, this method provided robust surface modification and no damage of CNTs. These PDA functionalized CNTs could be further reacted with amino-terminated polymers through Michael addition reaction. Poly(methacryloxyethyltrimethyl ammonium chloride) (PDMC) terminated with amino-group, which was synthesized via chain transfer free radical polymerization using cysteamine hydrochloride as the chain transfer agent and methacryloxyethyltrimethyl ammonium chloride (DMC) as monomer, was conjugated with surface of functional CNTs with PDA coating.

Application of CNT-Based Hybrid Nanocomposites to Cancer Diagnoses and Treatment Clinical application of CNTs is a very promising field. There are abundant studies about functionalized CNTs used as molecular carriers for applications in cancer treatment. For the early diagnose of cancer, CNTs conjugated to contrast reagent for CT or MRI and antibody with high affinity for cancer cells have been used in highly sensitive detection and imaging of small tumors. The detection of a prostate cancer markers (PSA), colorectal cancer markers (CEA, CA19-9), and a hepatocarcinoma marker (AFP) has been reported (61–64). Furthermore, CNTs have also been used in the photoacoustic imaging of living subjects due to their high photoacoustic signal. For example, SWNTs conjugated with cyclic Arg-GlyAsp (RGD) peptides were used as a contrast agent for photoacoustic imaging of tumor, offering high-resolution images with substantial depth of penetration (65). Drugs delivery systems (DDS) have been vigorously investigated for decades. Among the all kinds of carriers in DDS, CNTs are popular with researchers because of their special characteristics like high surface reactivity, small structure and excellent biocompatibility. Studies show that functionalized-CNTs possess a capacity to be taken up by a wide range of cells and to intracellularly traffic through the different cellular barriers by energy-independent mechanisms. The cylindrical shape and high aspect ratio of functionalized-CNTs can allow their penetration through the plasma membrane (66). CNTs attached with cell-recognizing module can cause the accumulation of drugs within a specific part of the body. The integration of CNTs with biomolecules allows the generation of complex nanostructures with controlled properties and functions (67). 173

Figure 1. Synthesis scheme for NPs of FA-MWCNT@Fe, FA-(FITC)MWCNT@Fe and (FITC)MWCNT@Fe, including the oxidization of MWCNT, loading of Fe(NO3)3 onto o-MWCNT, thermal decomposition of Fe(NO3)3, reduction of Fe2O3 NPs, and diimide-activated amidation reaction between o-MWCNT@Fe and the intermediates. Reproduced from (69). Copyright 2011, Elsevier.

Scheinberg et al. (68) reported the tumor-targeting CNT-based hybrid nanocomposites, which were constructed from sidewall-functionalized, water-soluble CNT platforms by covalently attaching multiple copies of tumor-specific monoclonal antibodies, radiometal-ion chelates, and fluorescent probes. These nanoconstructs were found to be specifically reactive with the human cancer cells that they were designed to target. Zou and co-workers synthesized the folate and iron di-functionalized MWNTs by conjugating folate 174

and iron nanoparticles with oxidized MWNTs, which were used as a dual-targeted drug nanocarrier to deliver doxorubicin into HeLa cells with the assistance of an external magnetic field (Figure 1). This nanocarrier has a sufficient load capacity and a prolonged release property controlled by near infrared radiation. It also demonstrated both biologically (active) and magnetically (passive) targeting capabilities toward HeLa cells in vitro with ca. 6-fold higher delivery efficiency of doxorubicin than free doxorubicin (69). CNTs exhibit strong optical absorption in the near infrared (NIR) regions and can generate heat to induce the thermal destruction of cancer cells containing sufficient CNT concentrations (42). Thus, targeted functionalized-CNTs can be used for the photo-thermal cancer therapy. Xing et al. (70) explored a therapy model in vivo with mitochondria-targeting SWNTs, which was used to convert 980-nm laser energy into heat and selectively destroy the target mitochondria for cancer treatment. It was found that SWNTs could be efficiently accumulated in the mitochondria of cancer cells, afforded remarkable efficacy in suppressing tumor growth in a breast cancer model, and resulted in complete tumor regression in some cases. A number of other studies about the photo-thermal therapy using CNTs have also been reported (71–73). The application of CNTs to gene therapy has also been studied. Prato et al. (74) functionalized CNTs with polyamidoamine (PAMAM) dendrons on their surface (Figure 2), and the functionalized CNTs formed supramolecular complexes with plasmid DNA through ionic interactions. These complexes penetrated within cells and facilitated higher DNA uptake and gene expression in vitro than that could be achieved with DNA alone. In other studies, linear cationic polymers such as poly(3-aminopropyl methacrylamide)-b-poly(2-lactobionamidoethyl methacrylamide) (PAPMA-b-PLAEMA) and polyethylenimine (PEI) were linked to the surface of CNTs to enhance effect of gene transfection and intracellular trafficking, as well as to induce the endosomal release of genes (75, 76).

Application of CNT-Based Hybrid Nanocomposites to Regenerative Medicine

CNTs have shown good blood compatibility, and they are the suitable scaffold material for repair and regeneration of human body tissues and organs. A study reported a type of guided tissue regeneration (GTR) membrane by electrospinning a suspension consisting of poly(L-lactic acid), MWNTs and hydroxyapatite (PLLA/MWNTs/HA) (77). Cytologic research revealed that the PLLA/MWNTs/HA membrane enhanced the adhesion and proliferation of periodontal ligament cells (PDLCs) by 30% and inhibited the adhesion and proliferation of gingival epithelial cells by 30% also, compared with the control group. CNTs functionalized with biocompatible macromolecules like collagen, polylactic acid or calcium phosphates, have been exploited to promote proliferation of various cells in vitro, such as fibroblasts, osteoblast, vascular endothelial cells, ligament cells and so on (78–82). 175

Figure 2. Synthesis of Dendron−MWNTs. Reproduced from (74). Copyright 2009, American Chemical Society.

Saito et al. (83) showed that CNTs promoted bone tissue formation in vivo. The study employed an experimental system that used recombinant bone morphogenetic protein-2 (rhBMP-2) to induce ectopicosteo-genesis in mouse back muscle. Bone formation on a collagen sheet was shown to occur earlier in the presence of rhBMP-2 attached to a scaffold of MWNTs than in the presence of rhBMP-2 alone. The later study (84) reported that CNTs could serve as the seed material for the crystallization of hydroxyapatite, the major component of bone, and that CNTs attracted Ca ions and activated osteoblasts. This activation was accompanied by the deposition of hydroxyapatite around the CNTs, which was catalyzed by alkaline phosphatase (ALP) released from osteoblasts. These findings demonstrated that CNTs, which functioned as a scaffold, could interact 176

with the body to promote osteogenesis and thereby the process of bone tissue regeneration. CNTs have been employed in neuroscience research for the design of neuronal interfaces or used as scaffolds favoring neuronal growth and axonal regeneration (85). Studies support the biocompatibility of functionalized-CNT scaffolds when used in vitro to sustain neuronal growth and axonal elongation and branching. Duan et al. (86) reported the investigation of electro-chemically co-deposited polypyrrole/SWNT (PPy/SWNT) films (Figure 3). This study demonstrated the feasibility and advantages of co-deposited PPy/SWNT films on Pt for improving the electro-de-neural tissue interface. The PPy/SWNT microelectrode exhibited a particularly high capacitance and lower electrode impedance (reduced by 95%) at 1 kHz compared to Pt microelectrodes. The SWNT doped conducting polymer film revealed improved mechanical and electrochemical stabilities than pure conducting polymer films. Furthermore, the PPy/SWNT film showed excellent biocompatibility both in vitro and in vivo.

Figure 3. Schematic illustration of the procedures for electrodeposition of polypyrrole with KCl, poly(styrene sulfonic acid) sodium salt (PSSNa), and carboxyl functionalized single-walled carbon nanotubes (SWCNT–COOH). Reproduced from (86). Copyright 2010, Elsevier.

In addition to the above-described applications, CNTs are expected to have biomedical application in a wide variety of therapeutic settings. CNT/polyethylene composite for application in artificial joints and CNT/polyether ether ketone composite for spinal fusion cages used in interbody fusion surgery have been developed by researchers (41). Tissue-compliant neural implants from micro-fabricated CNT multilayer composite have been reported (87). CNTs may also serve as muscle actuators or be directly applied to artificial muscles (88). Studies about CNT-based hybrid nanocomposite as biomaterial will be continued and more breakthroughs for disease therapy are expected. 177

MSN-Based Hybrid Nanocomposites as Biomaterials MSNs have been subjected to intense research in recent years for biomedical applications. The MSNs are usually synthesized by a sol-gel method, that is, the hydrolysis of silica precursors (sol) and the condensation on self-assembled surfactant templates (gel), followed by the removal of surfactant templates either by calcination or organic extraction (89). MSNs have the honeycomb-like porous structure with hundreds of empty channels, which lead to the unique properties of high surface area (>900 m2/g) and large pore volume (>0.9 cm3/g). The pore size of MSNs is usually tunable with a narrow distribution (2-10 nm). Combined with the good chemical and thermal stability, MSNs have been explored for various applications, such as catalysis, separation and sensors. Silica materials are known to be biocompatible, and MSNs with defined structures are able to absorb or encapsulate relatively large amounts of bioactive molecules due to the large pore volume. Thus, there are growing reports on the utilization of MSNs as carriers for drug and gene delivery application recently (11). To construct an efficient drug or gene delivery system, the carrier should allow high loading of drug molecules with no leaking of cargo before reaching the destination, possess the cell type or tissue specificity and site directing ability, and control release of drug molecules with a proper rate to achieve an effective local concentration. To meet the requirements, the composite combining MSN and organic molecule or polymer is usually needed for an increased affinity towards targeting molecules, an enhanced control over encapsulated cargoes, or any improved property with regard to application concerns (89).

Functionalization of MSNs Functionalization methods anchored functional groups onto surfaces of mesoporous silica materials generally include co-condensation (one-pot synthesis) and grafting (post-synthesis modification). For the co-condensation approach, functional moiety-containing organoalkoxysilanes are added into reaction solution where template-directed condensation of tetraalkoxysilane occurs. As a result, a direct synthesis of MSNs with functional groups is achieved at the same time as the orderly meso-structures with certain morphology formed. This method is applicable to a wide variety of organoalkoxysilanes, such as thiolate-, carboxylate- and sulfonate-containing organoalkoxysilane (6). The coverage of functional groups is homogeneous, and the loading amount of functional groups is high without affecting the structural ordering of the pores through this method. The grafting is commonly carried out by the reaction between functional moiety and silanol groups on the surface of mesoporous silica materials. Introduction of functional groups to as-made MSNs retains the mesostructure and mesoporosity of parent particles to the largest extent, but the distribution of functional groups on the surfaces of MSNs may be inhomogeneous. The most remarkable feature of the grafting method is that, the functional groups can be selectively modified on the external surfaces of MSNs, when the grafting is conducted before surfactant removal (90). 178

Lin et al. (91) fabricated a series of mesoporous silica materials with MCM-41 type of structure containing a homogeneous layer of organic functional groups inside the pores using co-condensation method. This synthetic approach resulted in high surface coverage with several functional groups such as a primary amine, secondary amine, urea, isocyanate, vinyl and nitrile. Our group (92, 93) anchored atom transfer radical polymerization (ATRP) and reversible addition-fragmentation chain transfer (RAFT) functionalities onto the exterior surface of MSNs through the grafting method. MSNs were modified with 3-aminopropyltriethoxysilane (APTES) to get MSN-NH2, then the ATRP agent-functionalized MSNs were prepared by the reaction of amino groups in MSN-NH2 with 2-bromo-2-methylpropionyl bromide, and it can be used to initiate polymerization of 2-(diethylamino)ethyl methacrylate (DEAEMA) to obtain MSN-PDEAEMA (Scheme 2). For the functionalization with RAFT agent, MSNs reacted with (3-glycidyloxypropyl) trimethoxysilane to obtain MSN-OH. The RAFT functionalized MSNs were prepared by the esterification reaction of MSN-OH with S-1-dodecyl-S′-(α,α′-dimethyl-α′′-acetic acid)trithiocarbonate. Polymer chains could be grafted on the surface of MSNs via surface RAFT polymerization of monomers.

Scheme 2. Synthetic route of PDEAEMA-functionalized MSNs (92).

179

Application of MSN-Based Hybrid Nanocomposites to Drug Delivery MSN-based controlled release systems have been investigated for years. MSNs work as rigid carriers for drug molecules, which effectively protects the cargo from chemical insults and biological degradation. To meet the targeted and controlled release character, various stimuli-responsive MSN-based drug delivery systems, which are sensitive to external or internal stimuli including temperature, redox, pH value, light, magnetism, enzyme, ultrasound, etc., have been developed (94–97). Lin and coworkers (98) synthesized a 2-(propyldisulfanyl)ethylamine functionalized mesoporous silica nanosphere material with an average particle size of 200.0 nm and an average pore diameter of 2.3 nm, which was used as reservoirs to encapsulate various pharmaceutical molecules and neurotransmitters. The open pores of the drug or neurotransmitter-loaded MSN were capped by acid-derivatized cadmium sulfide (CdS) nanocrystals (the average particle diameter is about 2.0 nm) via the amidation reaction, and the CdS nanocrystals could protect the cargos from escape. The disulfide linkages between the MSNs and the CdS nanoparticles can be cleaved by disulfide-reducing agents, such as dithiothreitol (DTT) and mercaptoethanol (ME), resulting in the release of drugs or neurotransmitters. In this system, CdS nanoparticles worked as “gatekeepers” to regulate the encapsulation and release of drug molecules. Subsequently, a series of nano-caps as gatekeepers for MSNs, which would response to the external stimulation by redox, light or magnetism, were developed, such as Au, Fe3O4, coumarin, diethylenetriamine, PAMAM nanoparticles and [2]-pseudorotaxane (99–101). Zink (96) and Fujiwara (102) designed various MSN-based nanomachines, such as smart supramolecular nano-valves ([2]-pseudorotaxane and [2]rotaxane) and nano-impellers (azobenzene and coumarin), to realize the pH-/light-responsive drug release. We constructed (93, 103) the drug delivery system via coating polymers onto MSNs, such as poly(2-(dimethylamino)ethyl methacrylate-co-3-dimethyl(methacryloyloxyethyl)ammonium propanesulfonate), and cross-linked poly(oligo(ethylene glycol) acrylate-co-N,N′-cystaminebismethacrylamide). The flexible polymers help to stabilize the MSNs, and can realize the controlled release of drugs under external stimulation of pH, temperature or redox. Du et al. reported (104) the controlled release of rhodamine 6G (Rh6G) from novel protein-gated carbohydrate-functionalized MSN nanocontainers. Mercaptoterminated mannose derivatives were synthesized, reacted with alkenyl-terminated silanes using a thiol–ene click reaction, and finally the external surfaces of MSNs were functionalized with mannose epitopes. Subsequent capping with Con A encapsulated the cargo within the pores through the multivalent carbohydrate–protein interactions. The carbohydrate-binding activity is abolished by decreasing the pH value of the buffer or by the competitive binding of glucose at high concentrations, releasing the cargo from MSNs on demand. Cai et al. (105) reported the enzyme-responsive drug delivery system based on MSNs for targeted tumor therapy. MSNs were functionalized with polypeptide, which consists of two components: the cell penetrating peptide 180

polyarginine (RRRRRRRRR) and MMP-2 cleavable substrate peptide (PVGLIG). Phenylboronic acid (PBA) was conjugated with human serum albumin (HSA), resulting in PBA-HSA, which was used as the end-capping agent for sealing the mesopores of MSNs. The PSA-HSA was immobilized onto MSNs via the intermediate linker of polypeptides. When the MSN-based drug delivery system reached the tumor site, the over-expressed MMP-2 in the tumor microenvironment could break down the intermediate linker, leading to the release of cargo from MSNs (Scheme 3). Zink et al. (106) designed a light-stimulated, thermally activated system for on-command drug release based on MSNs containing gold nanoparticle cores. Brauchle and co-workers (107) demonstrated a red-light photoactivated drug delivery system by linking the photosensitizer molecules (aluminum phthalocyaninedisulfonate) to the surface of MSNs and encapsulate MSNs with a supported lipid bilayer (SLB) to retain the drug. The activation of the photosensitizer with red light instead of blue light could reduce the phototoxicity and significantly increase the depth of tissue penetration (Scheme 4).

Scheme 3. Schematic illustration of the functionalization routes of an MSN-based drug delivery system and its enzyme mediated biological responses. Reproduced from (105). Copyright 2015, Royal Society of Chemistry.

181

Scheme 4. Synthesis Pathway of Core (Green) Shell (Red) MSN with Covalently Linked AlPcS2a (Red Star) via PEG Linker (Black Chain) and Surrounded by DOPC/DOTAP Supported Lipid Bilayer with Targeting Ligand (Green Star, TL). Reproduced from (107). Copyright 2013, American Chemical Society.

Application of MSN-Based Hybrid Nanocomposites to Bio-Imaging MSNs intrinsically possess many optical features due to the chemical nature of silica and their unique physical structure. Photoluminescence (PL) spectra of MSNs were determined by many structural characteristics, including specific composition, chemical bonding, defect sites, etc. While the application of MSNs in biological system can’t rely on their intrinsic optical patterns, since the PL quantum yield and the PL intensity of MSNs at a reasonable dosage in biological substances are not good enough to visualize MSNs inside cells, especially compared to those of traditional fluorescent dyes (89). The dye-containing MSNs are much more popular in bio-imaging. The fluorescent dye can be encapsulated into pores of MSNs by physical adsorption, or attached to MSNs by covalent linkage via post-synthetic grafting or co-condensation into the silica framework. Mou and co-workers (108) prepared fluorescein isothiocyanate (FITC) functionalized MSNs via co-condensation and treated FITC-MSNs with 3T3-L1 fibroblasts in vitro. The FITC functionalized MSNs with average diameter of 110 nm were demonstrated to be internalized into 3T3-L1 fibroblast cells and to accumulate in cytoplasm. The MSNs appeared to have no apparent cytotoxic effects on the fibroblast cells, revealing FITC-MSNs a promising material for bio-imaging. Similar investigations were employed in various cell cultures. Lo and co-workers (109) encapsulated indocyanine green (ICG) molecules inside the nanochannels of MSNs to probe the biodistribution of MSNs in anesthetized rats via near-infrared (NIR) microscopy. The high dispersion of ICG molecules in the 182

large surface areas of MSNs could efficiently prevent them from aggregation and thus decrease the fluorescence self-quench. In addition, the NIR light source could penetrate centimeters of tissue organic components and minimize the scattering and absorption while traveling through tissues. The fluorescence probe within NIR range was clearly detected after intravenous injection for evaluating particle accumulation inside the liver and kidney and partially in lungs, spleen, and heart. Wiesner et al. (110) reported one-pot synthesis of PEGylated MSNs with sizes around 9 nm that could be labeled with NIR dye Cy5.5, showing promise in future clinical imaging. Zhao’s group (111) loaded NIR dye squaraine inside MSNs, and wrapped the surfaces of MSNs with ultrathin graphene oxide sheets, leading to the formation of a novel hybrid material. The hybrid material exhibited remarkable stability in aqueous solution and could efficiently protect the loaded dye from nucleophilic attack. The in vitro investigation with Hela cells presented significant potential for application in fluorescence imaging (Scheme 5).

Scheme 5. Illustration of enwrapping dye-loaded MSNs with graphene oxide for fluorescence imaging in vitro. Reproduced from (111). Copyright 2012, American Chemical Society.

In addition to the fluorescent imaging, superparamagnetic iron oxide was combined with MSNs to enhance their properties as magnetic resonance imaging (MRI) contrast agents. Qu et al. (112) reported a multistage continuous targeting strategy based on magnetic mesoporous silica nanoparticles (FMSNs) for nuclear drug delivery. In this work, layer of mesoporous silica was coated around the Fe3O4 core and the DNA-toxin anticancer drug camptothecin was loaded into the pores of the silica shell. The nucleus-targeting peptide, trans-activating transcriptional (TAT) activator, was grafted onto the silica surface to enable cellular targeting, and the system was further decorated with a charge-conversional polymer to enhance the stability of FMSNs under physiological conditions. The Fe3O4 nanocore was capable of efficiently accumulating in tumor tissue guided by magnet, and could be simultaneously used as predominant contrast agents for magnetic resonance imaging. Yang and co-workers (113) developed multifunctional theranostic nanoparticles based on mesoporous silica to achieve 183

the combination of MRI and near-infrared fluorescence (NIRF) imaging and the target-specific delivery of hydrophobic drugs. The Fe3O4 nanoparticles grafted with hydrophobic poly(tert-butyl acrylate) (PTBA) were employed as the hydrophobic core, while the mesoporous silica modified with transferrin (Tf) and near-infrared fluorescent (NIRF) dye Cy7 acted as the multifunctional shell. The iron oxide core was used as a MRI agent, and the Cy7 fluorescent dye served as a NIRF imaging probe. The modification of Tf proved to be an effective target ligand endows nanoparticles the ability of tumor-targeting delivery. The hydrophobic antitumor drug paclitaxel (PTX) was used as the model drug, and the drug-loaded nanocarrier showed effectiveness of killing cancer cells in cytotoxicity assay. Lin and co-workers (114) grafted gadolinium (Gd) chelate onto MSNs for designing highly efficient MRI contrast agent (Figure 4). MSN-Gd showed no noticeable toxicity to monocyte cells and it was demonstrated to be a highly efficient T1 contrast agent for intravascular MR imaging and an excellent T2 contrast agent for MR imaging of soft tissues when applied at a higher dosage (31 µmol/kg dose). Subsequently, a series of MRI contrast agents by doping Gd into MSNs were developed (89, 115, 116) to improve the magnetic signals and strengthen the biological applicability (Scheme 6). Furthermore, MSN-coated MnO nanocystals were also applied as T1 contrast agents (117, 118). Shi and co-workers (118) reported the synthesis of manganese oxide-based multifunctionalization of hollow mesoporous silica nanoparticles by in situ redox reaction using mesopores as nanoreactors (Figure 5). This manganese-based mesoporous MRI-T1 contrast agent (CA) could release MnII in acidic condition, significantly enhancing the MRI-T1 performance in acidic tumor microenvironment. The relaxation rate of the Mn-based CA was almost two times higher than commercial GdIII-based complex agents. Meanwhile, the hybrid MSNs could encapsulate and deliver anticancer agents (doxorubicin) intracellularly to cancer cells. Liu et al. (119) developed the fluorescent carbon dot (C-dot) nanoclusters by integrating C-dot with hollow silica spheres or PEG-g-hollow silica. Hollow silica spheres worked as a scaffold for the C-dots, and the C-dot nanoclusters showed a redshifted fluorescence emission with an increased excitation wavelength. The in vitro investigation with live HepG2 cells and live MCF-7 cells presented good cytoplasm imaging, and C-dot nanoclusters are more efficient in providing fluorescence imaging than the free C-dots. In addition to the application in drug delivery and bio-imaging fields mentioned above, MSN-based hybrid nanocomposites are also popular in the application of biosensing and biocatalysis by immobilization of various proteins (e.g., enzymes, antibodies) on mesoporous silicas with a variety of mesostructures (120, 121). Compared to free proteins, confined enzymes on inner or outer surfaces of mesochannels possessed an enhanced stability and activity, mainly because MSNs as the host created a protective microenvironment enclosing the guest proteins, sequestering them from external physical or chemical disturbances, such as abrupt changes in pH, ionic strength or temperature. Therefore, sensing or catalytic efficiency and accuracy are much improved.

184

Figure 4. (a) SEM image of MSN showing the formation of monodisperse, water-dispersible nanoparticles. (b) Schematic showing the Gd-Si-DTTA complexes residing in hexagonally ordered nanochannels of ~2.4 nm in diameter. (c) TGA curves of as-synthesized MSN (black), surfactant-extracted MSN (red), and MSN−Gd (blue). (d) The r1 (solid) and r2 (dashed) relaxivity curves of MSN−Gd at 3 T (black) and 9.4 T (red). Reproduced from (114). Copyright 2008, American Chemical Society.

Scheme 6. Synthesis of NDP2–Gd Complexes within Porous Silica Particles. Reproduced from (115). Copyright 2012, American Chemical Society.

185

Figure 5. (a) Schematic illustration of the microstructure and structure-related theranostic functions of hybrid mesoporous composite nanocapsules (HMCNs); (b) TEM image of HMCNs (inset: the STEM image with scale bar = 100 nm); (c–f) Element mapping of Si (c), O (d) and Mn (e) In HMCNs (f: the merged image of c, d and e). Reproduced from (118). Copyright 2012, Elsevier.

It is worth noting that MSNs with special structure, such as hollow MSNs and rod-shaped MSNs, have been widely applied in biomedical field like the conventional mesoporous silica. Hollow MSNs possess huge hollow interiors and penetrating pores in the shell that endow them with a sustained release property and a higher drug loading capacity than conventional MSNs, which are ideal as drug delivery system (122). The rod-shaped MSNs with large aspect ratio show some superiorities in biomedical applications. For example, the study of tang and co-workers (123) about the shape effect of MSNs on cellar uptake, biodistribution, and clearance, showed that particles with larger aspect ratios had faster internalization rates and longer circulation times. We have designed a novel nanocarrier by attaching pH-sensitive polymer on the exterior surface of silica nanotube for drug delivery (124). This nanocarrier could encapsulate drug molecules into the hollow interior of silica nanotube and realize controlled release of drug by adjusting the pH of medium. Meanwhile, the in vitro cell viability 186

and cellular internalization study presented good biocompatibility and excellent endocytosis property. In general, the quest for intelligent nanomedicine based upon various mesoporous silica materials and the detailed in vitro and in vivo testing is one of the most important areas of biomedical research in recent times as well as in near future.

Conclusion and Perspective Significant and efficient progress has been made employing CNT-based and MSN-based hybrid nanocomposites in various biological applications. The distinguishing properties made them important classes of biologically applicable materials. The number of reports on smart hybrid nanocomposites based on CNTs and MSNs has been increasing. Although these reports have demonstrated the effectiveness of these biomaterials in vitro or in vivo, more effort have to be made to push the hybrid nanocomposites closer to future clinical trials. The prospective studies should focus on two aspects. One is to rationally design and construct nano-platforms, and to study the interactions between such nanoscaled platforms and different biological entities simultaneously. The other is to explore the pharmacokinetics and biocompatibility of these hybrid nanocomposites, and to investigate the toxicity on diseased tissues as well as the side-effect on healthy tissues in vivo. There are still many questions which need to be settled for the practically clinical applications of the hybrid nanomaterials, such as acute and chronic toxicities, genotoxicity, the reproductive toxicity, the long-term in vivo degradation, etc., but we expect that these problems would be solved in the future.

Acknowledgments We thank National Natural Science Foundation of China (No. 21525420 and 21374107) and the Fundamental Research Funds for the Central Universities (WK 2060200012).

References 1. 2. 3. 4. 5. 6. 7. 8.

Greish, K. J. Drug Targeting 2007, 15, 457–464. Khandare, J; Minko, T. Prog. Polym. Sci. 2006, 31, 359–397. Vicent, M. J.; Greco, F.; Nicholson, R. I.; Paul, A.; Griffiths, P. C.; Duncan, R. Angew. Chem., Int. Ed. 2005, 44, 4061–4066. Liu, S.; Maheshwari, R.; Kiick, K. L. Macromolecules 2009, 42, 3–13. Blanazs, A.; Armes, S. P.; Ryan, A. J. Macromol. Rapid Commun. 2009, 30, 267–277. Slowing, I. I; Vivero-Escoto, J. L.; Wu, C. W.; Lin, V. S. Y. Adv. Drug Delivery Rev. 2008, 60, 1278–1288. Xiong, X. B.; Falamarzian, A.; Garg, S. M.; Lavasanifar, A. J. Controlled Release 2011, 155, 248–261. Zhang, Q.; Ko, N. R.; Oh, J. K. Chem. Commun. 2012, 48, 7542–7552. 187

9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

Samarajeewa, S.; Shretha, R.; Li, Y.; Wooley, K. L. J. Am. Chem. Soc. 2012, 134, 1235–1242. Wagner, V.; Dullaart, A.; Bock, A. K.; Zweck, A. Nat. Biotechnol. 2006, 24, 1211–1217. Chen, Y.; Chen, H. R.; Shi, J. L. Adv. Mater. 2013, 25, 3144–3176. Arvizo, R. R.; Bhattacharyya, S.; Kudgus, R. A.; Giri, K.; Bhattacharya, R.; Mukherjee, P. Chem. Soc. Rev. 2012, 41, 2943–2970. Robinson, J. T.; Tabakman, S. M.; Liang, Y. Y.; Wang, H. L.; Casalongue, H. S.; Vinh, D.; Dai, H. J. J. Am. Chem. Soc. 2011, 133, 6825–6831. Li, D. W.; Bu, Y. Z.; Zhang, L. N.; Wang, X.; Yang, Y. Y.; Zhuang, Y. P.; Yang, F.; Shen, H.; Wu, D. C. Biomacromolecules 2016, 17, 291–300. Wu, D. C.; Liu, Y.; Jiang, X.; He, C. B.; Goh, S. H.; Leong, K. W. Biomacromolecules 2006, 7, 1879–1883. Lu, J.; Liong, M.; Li, Z.; Zink, J. I.; Tamanoi, F. Small 2010, 6, 1794–1805. Maeda, H.; Wu, J.; Sawa, T.; Matsumura, Y.; Hori, K. J. Controlled Release 2000, 65, 271–284. Rosenholm, J. M.; Meinander, A.; Peuhu, E.; Niemi, R.; Eriksson, J. E.; Sahlgren, C.; Linden, M. ACS Nano 2009, 3, 197–206. Ferris, D. P.; Lu, J.; Gothard, C.; Yanes, R.; Thomas, C. R.; Olsen, J. C.; Stoddart, J. F.; Tamanoi, F.; Zink, J. I. Small 2011, 7, 1816–1826. Vallet-Regi, M.; Balas, F.; Arcos, D. Angew. Chem., Int. Ed. 2007, 46, 754–7558. Kumar, S.; Ahlawat, W.; Kumar, R.; Dilbaghi, N. Biosens. Bioelectron. 2015, 70, 498–503. Koo, H.; Huh, M. S.; Ryu, J. H.; Lee, D. E.; Sun, I. C.; Choi, K.; Kim, K.; Kwon, I. C. Nano Today 2011, 6, 204–220. Son, K. H.; Hong, J. H.; Lee, J. W. Int. J. Nanomed. 2016, 11, 5163–5185. Yang, N.; Chen, X. P.; Ren, T. L.; Zhang, P.; Yang, D. G. Sens. Actuators, B 2015, 207, 690–715. Liong, M.; Lu, J.; Kovochich, M.; Xia, T.; Ruehm, S. G.; Nel, A. E.; Tamanoi, F.; Zink, J. I. ACS Nano 2008, 5, 889–896. Lee, J. E.; Lee, N.; Kim, T.; Kim, J.; Hyeon, T. Acc. Chem. Res. 2011, 44, 893–902. Gabizon, A.; Goren, D.; Cohen, R.; Barenholz, Y. J. Controlled Release 1998, 53, 275–279. Fernandez-Fernandez, A.; Manchanda, R.; McGoron, A. J. Appl. Biochem. Biotechnol. 2011, 165, 1628–1651. Elsabahy, M.; Wooley, K. L. Chem. Soc. Rev. 2012, 41, 2545–2561. Giri, S.; Trewyn, B. G.; Lin, V. S. Nanomedicine 2007, 2, 99–111. Laurent, S.; Forge, D.; Port, M.; Roch, A.; Robic, C.; Elst, L. V.; Muller, R. N. Chem. Rev. 2008, 108, 2064–2110. Skrabalak, S. E.; Chen, J. Y.; Sun, Y. G.; Lu, X. M.; Au, L.; Cobley, C. M.; Xia, Y. N. Acc. Chem. Res. 2008, 41, 1587–1595. Patra, C. R.; Bhattacharya, R.; Mukhopadhyay, D.; Mukherjee, P. Adv. Drug Delivery Rev. 2010, 62, 346–361. Beg, S.; Rizwan, M.; Sheikh, A. M.; Hasnain, M. S.; Anwer, K.; Kohli, K. J. Pharm. Pharmacol. 2011, 63, 141–163. 188

35. Lee, Y.; Geckeler, K. E. Adv. Mater. 2010, 22, 4076–4083. 36. Beg, S.; Rizwan, M.; Sheikh, A. M.; Hasnain, M. S.; Anwer, K.; Kohli, K. J. Pharm. Pharmacol. 2011, 63, 141–163. 37. Hudson, S. P.; Padera, R. F.; Langer, R.; Kohane, D. S. Biomaterials 2008, 29, 4045–4055. 38. Iijima, S. Nature 1991, 354, 56–58. 39. Iijima, S. Phys. B: Condens. Matter 2002, 323, 1–5. 40. Liu, Z; Tabakman, S; Welsher, K; Dai, H. Nano Res. 2009, 2, 85–120. 41. Saito, N.; Haniu, H.; Usui, Y.; Aoki, K.; Hara, K.; Takanashi, S.; Shimizu, M.; Narita, N.; Okamoto, M.; Kobayashi, S.; Nomura, H.; Kato, H.; Nishimura, N.; Taruta, S.; Endo, M. Chem. Rev. 2014, 114, 6040–6079. 42. Peretz, S.; Regev, O. Curr. Opin. Colloid Interface Sci. 2012, 17, 360–368. 43. Boncel, S.; Zajac, P.; Koziol, K. K. K. J. Controlled Release 2013, 169, 126–140. 44. Razzazan, A.; Atyabi, F.; Kazemi, B.; Dinarvand, R. Mater. Sci. Eng. C 2016, 62, 614–625. 45. Zhou, M. Y.; Liu, S. H.; Jiang, Y. Q.; Ma, H. R.; Shi, M.; Wang, Q. S.; Zhong, W.; Liao, W. J.; Xing, M. M. Q. Adv. Funct. Mater. 2015, 25, 4730–4739. 46. Kavosi, B.; Salimi, A.; Hallaj, R.; Amani, K. Biosens. Bioelectron. 2014, 52, 20–28. 47. Liu, J. J.; Wang, C.; Wang, X. J.; Wang, X.; Cheng, L.; Li, Y. G.; Liu, Z. Adv. Funct. Mater. 2015, 25, 384–392. 48. Liang, C.; Diao, S.; Wang, C.; Gong, H.; Liu, T.; Hong, G. S.; Shi, X. Z.; Dai, H. J.; Liu, Z. Adv. Mater. 2014, 26, 5646–5652. 49. Menna, F.; Abdelghani, A.; Menaa, B. J. Tissue Eng. Regen. Med. 2015, 9, 1321–1338. 50. Li, X. M.; Yang, Y.; Fan, Y. B.; Feng, Q. L.; Cui, F. Z.; Watari, F. J. Biomed. Mater. Res. A 2014, 102, 1580–1594. 51. Mehra, N. K.; Mishra, V.; Jain, N. K. Biomaterials 2014, 35, 1267–1283. 52. Kim, S. W.; Kim, T.; Kim, Y. S.; Choi, H. S.; Lim, H. J.; Yang, S. J.; Park, C. R. Carbon 2012, 50, 3–33. 53. Kumar, I.; Rana, S.; Rode, C. V.; Cho, J. W. J. Nanosci. Nanotechnol. 2008, 8, 3351–3356. 54. Georgakilas, V.; Kordatos, K.; Prato, M.; Guldi, D. M.; Holzinger, M.; Hirsch, A. J. Am. Chem. Soc. 2002, 124, 760–761. 55. You, Y. Z.; Hong, C. Y.; Pan, C. Y. Adv. Funct. Mater. 2007, 17, 2470–2477. 56. You, Y. Z.; Yan, J. J.; Yu, Z. Q.; Cui, M. M.; Hong, C. Y.; Qu, B. J. J. Mater. Chem. 2009, 19, 7656–7660. 57. Ghini, G.; Luconi, L.; Rossin, A.; Bianchini, C.; Giambastiani, G.; Cicchi, S.; Lascialfari, L.; Brandi, A.; Giannasi, A. Chem. Commun. 2010, 46, 252–254. 58. Lee, J. U.; Huh, J.; Kim, K. H.; Park, C.; Jo, W. H. Carbon 2007, 45, 1051–1057. 59. Karajanagi, S. S.; Yang, H.; Asuri, P.; Sellitto, E.; Dordick, J. S.; Kane, R. S. Langmuir 2006, 22, 1392–1395. 60. Wan, Q.; Tian, J. W.; Liu, M. Y.; Zeng, G. J.; Huang, Q.; Wang, K.; Zhang, Q. S.; Deng, F. J.; Zhang, X. Y.; Wei, Y. Appl. Surf. Sci. 2015, 346, 335–341. 189

61. Ding, Y.; Liu, J.; Jin, X.; Lu, H.; Shen, G.; Yu, R. Analyst 2008, 133, 184–190. 62. Kim, J. P.; Lee, B. Y.; Lee, J.; Hong, S.; Sim, S. J. Biosens. Bioelectron. 2009, 24, 3372–3378. 63. Lin, J.; He, C.; Zhang, L.; Zhang, S. Anal. Biochem. 2009, 384, 130–135. 64. Li, Q. F.; Tang, D. P.; Tang, J. A.; Su, B. L.; Huang, J. X.; Chen, G. N. Talanta 2011, 84, 538–546. 65. De la Zerda, A.; Zavaleta, C.; Keren, S.; Vaithilingam, S.; Bodapati, S.; Liu, Z.; Levi, J.; Smith, B. R.; Ma, T. J.; Oralkan, O.; Cheng, Z.; Chen, X.; Dai, H.; Khuri-Yakub, B. T.; Gambhir, S. S. Nat. Nanotechnol. 2008, 3, 557–562. 66. Kostarelos, K.; Lacerda, L.; Pastorin, G.; Wu, W.; Wiecklwshi, S.; Luangsivilay, J.; Godefroy, S.; Pantarotto, D.; Briand, J.; Muller, S.; Prato, M.; Bianco, A. Nat. Nanotechnol. 2007, 2, 108–113. 67. Saito, N.; Usui, Y.; Aoki, K.; Narita, N.; Shimizu, M.; Hara, K.; Ogiwara, N.; Nakamura, K.; Ishigaki, N.; Kato, K.; Taruta, S.; Endo, M. Chem. Soc. Rev. 2009, 38, 1897–1903. 68. McDevitt, M. R.; Chattopadhyay, D.; Kappel, B. J.; Jaggi, J. S.; Schiffman, S. R.; Antczak, C.; Njardarson, J. T.; Brentjens, R.; Scheinberg, D. A. J. Nucl. Med. 2007, 48, 1180. 69. Li, R. B.; Wu, R. A.; Zhao, L.; Hu, Z. Y.; Guo, S. J.; Pan, X. L.; Zou, H. F. Carbon 2011, 49, 1797–1805. 70. Zhou, F. F.; Wu, S. N.; Wu, B. Y.; Chen, W. R.; Xing, D. Small 2011, 7, 2727–2735. 71. Fisher, J. W.; Sarkar, S.; Buchanan, C. F.; Szot, C. S.; Whitney, J.; Hatcher, H. C.; Torti, S. V.; Rylander, C. G.; Rylander, M. N. Cancer Res. 2010, 70, 9855–9864. 72. Marches, R.; Mikoryak, C.; Wang, R. H.; Pantano, P.; Draper, R. K.; Vitetta, E. S. Nanotechnology 2011, 22, 095101. 73. Chen, D. Q.; Wang, C.; Nie, X.; Li, S. M.; Li, R. M.; Guan, M. R.; Liu, Z.; Chen, C. Y.; Wang, C. R.; Shu, C. Y.; Wan, L. J. Adv. Funct. Mater. 2014, 24, 6621–6628. 74. Herrero, M. A.; Toma, F. M.; Al-Jamal, K. T.; Kostarelos, K.; Biano, A.; Ros, T. D.; Bano, F.; Casalis, L.; Scoles, G.; Prato, M. J. Am. Chem. Soc. 2009, 131, 9843–9848. 75. Ahmed, M.; Jiang, X. Z.; Deng, Z. C.; Narain, R. Bioconjugate Chem. 2009, 20, 2017. 76. Wang, L.; Shi, J. J.; Zhang, H. L.; Li, H. X.; Gao, Y.; Wang, Z. Z.; Wang, H. H.; Li, L. L.; Zhang, C. F.; Chen, C. Q.; Zhang, Z. Z.; Zhang, Y. Biomaterials 2013, 34, 262–274. 77. Mei, F.; Zhong, J.; Yang, X.; Ouyang, X.; Zhang, S.; Hu, X.; Ma, Q.; Lu, J.; Ryu, S.; Deng, X. Biomacromolecules 2007, 8, 3729. 78. Han, Z.; Kong, H.; Meng, J.; Wang, C. Y.; Xie, S. S.; Xu, H. Y. J. Nanosci. Nanotechnol. 2009, 9, 1400–1402. 79. Meng, J.; Kong, H.; Han, Z. Z.; Wang, C. Y.; Zhu, G.; Xie, S. S.; Xu, H. Y. J. Biomed. Mater. Res., Part A 2009, 88, 105–116. 80. Zhang, R. J.; Metoki, N.; Sharabani-Yosef, O.; Zhu, H. W.; Eliaz, N. Adv. Funct. Mater. 2016, 26, 7965–7974. 190

81. Siqueira, I. A. W. B.; Corat, M. A. F.; Cavalcanti, B. das N.; Neto, W. A. R.; Martin, A. A.; Bretas, R. E. S.; Marciano, F. R.; Lobo, A. O. ACS Appl. Mater. Interfaces 2015, 7, 9385–9398. 82. Mikael, P. E.; Amini, A. R.; Basu, J.; Arellano-Jimenez, M. J.; Laurencin, C. T.; Sanders, M. M.; Carter, C. B.; Nukavarapu, S. P. Biomed. Mater. 2014, 9, 1–13. 83. Usui, Y.; Aoki, K.; Narita, N.; Murakami, N.; Nakamura, I.; Nakamura, K.; Ishigaki, N.; Yamazaki, H.; Horiuchi, H.; Kato, H.; Taruta, S.; Kim, Y. A.; Endo, M.; Saito, N. Small 2008, 4, 240–246. 84. Shimizu, M.; Kobayashi, Y.; Mizoguchi, T.; Nakamura, H.; Kawahara, I.; Narita, N.; Usui, Y.; Aoki, K.; Hara, K.; Haniu, H.; Ogihara, N.; Ishigaki, N.; Nakamura, K.; Kato, H.; Kawakubo, M.; Dohi, Y.; Taruta, S.; Kim, Y. A.; Endo, M.; Ozawa, H.; Udagawa, N.; Takahashi, N.; Saito, N. Adv. Mater. 2012, 24, 2176–2185. 85. Fraczek-Szczypta, A. Mater. Sci. Eng., C 2014, 34, 35–49. 86. Lu, Y.; Li, T.; Zhao, X.; Li, M.; Cao, Y.; Yang, H.; Duan, Y. Y. Biomaterials 2010, 31, 5169–5181. 87. Zhang, H. N.; Petel, P. R.; Xie, Z. X.; Swanson, S. D.; Wang, X. D.; Kotov, N. A. ACS Nano 2013, 7, 7619–7629. 88. Lima, M. D.; Li, N.; Jung de Andrade, M.; Fang, S.; Oh, J.; Spinks, G. M.; Kozlov, M. E.; Haines, C. S.; Suh, D.; Foroughi, J.; Kim, S. J.; Chen, Y.; Ware, T.; Shin, M. K.; Machado, L. D.; Fonseca, A. F.; Madden, J. D.; Voit, W. E.; Galvao, D. S.; Baughman, R. H. Science 2012, 338, 928–932. 89. Tao, Z. M. RSC Adv. 2014, 4, 18961–18980. 90. Titinchi, S. J. J.; Singh, M. P.; Abbo, H. S.; Green, I. R. Adv. Healthcare Mater. 2014, 49–86. 91. Huh, S.; Wiench, J. W.; Yoo, J. C.; Pruski, M.; Lin, V. S. Y. Chem. Mater. 2003, 15, 4247–4256. 92. Sun, J. T.; Hong, C. Y.; Pan, C. Y. J. Phys. Chem. C 2010, 114, 12481–12486. 93. Sun, J. T.; Yu, Z. Q.; Hong, C. Y.; Pan, C. Y. Macromol. Rapid Commun. 2012, 33, 811–818. 94. Sauer, A. M.; Schlossbauer, A.; Ruthardt, N.; Cauda, V.; Bein, T.; Brauchle, C. Nano Lett. 2010, 10, 3684–3691. 95. Cauda, V.; Argyo, C.; Schlossbauer, A.; Bein, T. J. Mater. Chem. 2010, 20, 4305–4311. 96. Meng, H.; Xue, M.; Xia, T.; Zhao, Y. L.; Tamanoi, F.; Stoddart, J. F.; Zink, J. I.; Nel, A. E. J. Am. Chem. Soc. 2010, 132, 12690–12697. 97. Lin, Q. N.; Huang, Q.; Li, C. Y.; Bao, C. Y.; Liu, Z. Z.; Li, F. Y.; Zhu, L. Y. J. Am. Chem. Soc. 2010, 132, 10645–10647. 98. Lai, C. Y.; Trewyn, B. G.; Jeftinija, D. M.; Jeftinija, K.; Xu, S.; Jeftinija, S.; Lin, V. S. Y. J. Am. Chem. Soc. 2003, 125, 4451–4459. 99. Torney, F.; Trewyn, B. G.; Lin, V. S. Y.; Wang, K. Nat. Nanotechnol. 2007, 2, 295–300. 100. Park, C.; Oh, K.; Lee, S. C.; Kim, C. Angew. Chem., Int. Ed. 2007, 46, 1455–1457. 101. Hernandez, R.; Tseng, H. R.; Wong, J. W.; Stoddart, J. F.; Zink, J. I. J. Am. Chem. Soc. 2004, 126, 3370–3371. 191

102. Mal, N. K.; Fujiwara, M.; Tanaka, Y. Nature 2003, 421, 350–353. 103. Sun, J. T.; Piao, J. G.; Wang, L. H.; Javed, M.; Hong, C. Y.; Pan, C. Y. Macromol. Rapid Commun. 2013, 34, 1387–1394. 104. Wu, S. S.; Huang, X.; Du, X. Z. Angew. Chem., Int. Ed. 2013, 52, 5580–5584. 105. Liu, J. J.; Zhang, B. L.; Luo, Z.; Ding, X. W.; Li, J. H.; Dai, L. L.; Zhou, J.; Zhao, X. J.; Ye, J. Y.; Cai, K. Y. Nanoscale 2015, 7, 3614–3626. 106. Croissant, J.; Zink, J. I. J. Am. Chem. Soc. 2012, 134, 7628–7631. 107. Mackowiak, S. A.; Schmidt, A.; Weiss, V.; Argyo, C.; Schirnding, C. V.; Bein, T.; Brauchle, C. Nano Lett. 2013, 13, 2576–2583. 108. Lin, Y. S.; Tsai, C. P.; Huang, H. Y.; Kuo, C. T.; Hung, Y.; Huang, D. M.; Chen, C. Y.; Mou, C. Y. Chem. Mater. 2005, 17, 4570–4573. 109. Lee, C. H.; Cheng, S. H.; Wang, Y. J.; Chen, Y. C.; Chen, N. T.; Souris, J.; Chen, C. T.; Mou, C. Y.; Ynag, C. S.; Lo, L. W. Adv. Funct. Mater. 2009, 19, 215–222. 110. Ma, K.; Sai, H.; Wiesner, U. J. Am. Chem. Soc. 2012, 134, 13180–13183. 111. Sreejith, S.; Ma, X.; Zhao, Y. L. J. Am. Chem. Soc. 2012, 134, 17346–17349. 112. Li, Z. H.; Dong, K.; Huang, S.; Ju, E.; Liu, Z.; Yin, M. L.; Ren, J. S.; Qu, X. G. Adv. Funct. Mater. 2014, 24, 3612–3620. 113. Jiao, Y. F.; Sun, Y. F.; Tang, X. L.; Ren, Q. G.; Yang, W. L. Small 2015, 11, 1962–1974. 114. Taylor, K. M. L.; Kim, J. S.; Rieter, W. J.; An, H. Y.; Lin, W. L.; Lin, W. B. J. Am. Chem. Soc. 2008, 130, 2154–2155. 115. Duncan, A. K.; Klemm, P. J.; Raymond, K. N.; Landry, C. C. J. Am. Chem. Soc. 2012, 134, 8046–8049. 116. Huang, W. Y.; Davies, G. L.; Davis, J. J. Chem. Commun. 2013, 49, 60–62. 117. Peng, Y. K.; Lai, C. W.; Liu, C. L.; Chen, H. C.; Hsiao, Y. H.; Liu, W. L.; Tang, K. C.; Chi, Y.; Hsiao, J. K.; Lim, K. E.; Liao, H. E.; Shyue, J. J.; Chou, P. T. ACS Nano 2011, 5, 4177–4187. 118. Chen, Y.; Yin, Q.; Ji, X. F.; Zhang, S. J.; Chen, H. R.; Zheng, Y. Y.; Sun, Y.; Qu, H. Y.; Wang, Z.; Li, Y. P.; Wang, X.; Zhang, K.; Zhang, L. L.; Shi, J. L. Biomaterials 2012, 33, 7126–7137. 119. Wang, G.; Pan, X. Y.; Gu, L. Q.; Ren, W.; Cheng, W. R.; Kumar, J. N.; Liu, Y. Nanotechnology 2014, 25, 375601–375606. 120. Popat, A.; Hartono, S. B.; Stahr, F.; Liu, J.; Qiao, S. Z.; Lu, G. Q. Nanoscale 2011, 3, 2801–2818. 121. Hudson, S.; Cooney, J.; Magner, E. Angew. Chem., Int. Ed. 2008, 47, 8582–8594. 122. Yu, F. Q.; Tang, X. D.; Pei, M. S. Microporous Mesoporous Mater. 2013, 173, 64–69. 123. Huang, X. L.; Teng, X.; Chen, D.; Tang, F. Q.; He, J. Q. Biomaterials 2010, 31, 438–448. 124. Zhou, J. M.; Zhang, W. J.; Hong, C. Y.; Pan, C. Y. ACS Appl. Mater. Interfaces 2015, 7, 3618–3625.

192

Micro/Nano-Fabricated Devices

Chapter 10

In Vitro Design of Nanoparticles Using an Artificial 3D-Blood Vessel Wall Model for Atherosclerosis Treatment Michiya Matsusaki1,2 and Mitsuru Akashi1,3,* 1Department

of Applied Chemistry, Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita 565-0871, Japan 2PRESTO, Japan Science and Technology Agency (JST), 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan 3Graduate School of Frontier Biosciences, Osaka University, 1-3 Yamadaoka, Suita 565-0871, Japan *E-mail: [email protected]

Here, we described a bottom-up approach for the reconstitution of three-dimensional blood vessel model tissues for the application of in vitro design of nanomaterials. Nanomaterials have been widely used for applications in biomedical fields and could become an indispensable in the near future. However since it is difficult to optimize in vivo 3D-biological behavior using a single cell in vitro, there have been many failures in animal models. In vitro prediction systems using 3D-human tissue models reflecting the 3D-location of cell types may be useful to better understand the biological characteristics of nanomaterials for optimization of their function. Here we demonstrate the potential ability of 3D-engineered human arterial models for in vitro prediction of in vivo 3D-behavior of nanoparticles for drug delivery. These models allowed for optimization of the composition and size of the nanoparticles for targeting and treatment efficacy for atherosclerosis. In vivo experiments using atherosclerotic mice suggested the highest biological characteristics and potential treatment effects of the in vitro optimized nanoparticles.

© 2017 American Chemical Society

Introduction As studies on stem cells such as embryonic stem cell (ES cell) and induced pluripotent stem cell (iPS cell) have progressed, their applications for clinically-relevant therapies and drug testing models have been expected (1). Since ES cells and iPS cells have the capability to differentiate into various types of cells such as chondrocytes, cardiac myocytes, hepatocytes and neurocytes, they can be employed for cell transplantation therapies (2). However, the efficiency of cell engraftment is extremely low, resulting in cell death, the formation of tumors, and the diffusion of the transplanted cells into other organs (3). To overcome this low engraftment efficiency and organization with host tissues, the concept of tissue engineering was established by R. Langer and J. Vacanti (4). Tissue engineering basically aims to replace a damaged organ with an engineered tissue, which consists of three components: cells, a scaffold, and growth factors. These living tissues possess many types of cells and extracellular matrices (ECM), such as collagen, fibronectin, laminin, and hyaluronic acid, with oriented and hierarchical higher-order structures (5, 6). To construct them, these tissue structures should mimic the in vivo structure using the three components, and their functions as tissues should be recreated. The in vitro engineering of tissues with structures and functions comparable to living organs would be useful in the field of regenerative medicine. Furthermore these engineered tissues can be used for drug testing models (7). Normally, a cell monolayer has been employed as the in vitro model in pharmaceutical processes. However, it is difficult to obtain tissue responses from monolayered cell models that do not have a three-dimensional (3D) structure like living tissues. Moreover, alternatives to animal testing have attracted increasing attention (8). Basically, pharmaceutical assays have been performed by in vivo animal experiments, but the low reproducibility and species differences remain unsolved issues. In particular, animal testing for cosmetics and chemicals has been prohibited under the 7th Amendment to the Cosmetics Directive (Council Directive 76/768/EEC) and REACH (registration, evaluation, authorization and restriction of chemicals) in the European Union (EU). Taken together, the development of a novel technology for in vitro tissue fabrication is required in the field of tissue engineering and drug assessment.

A Bottom-Up Approach for 3D-Tissue Construction As described above, all tissues and organs in the body are highly organized in a hierarchical manner. For example, the wall of a blood vessel is multilayered, consisting of three parts: endothelial cells, which are the innermost layer and are in contact with the blood; the middle layers of vascular smooth muscle cells; and the outer collagen and elastic fibers extending into the surrounding connective tissue (6). To recreate this hierarchical structure in vitro, a scaffold which can provide a substrate for cell adhesion is essential (9–14). Two of the most important properties for the scaffold are how the cells interact with the constituent materials, and how cellular functions are affected by these materials (15, 16). There are many requirements for the scaffold such as biocompatibility, protein adsorption, stiffness, and porosity. For example, protein adsorption depends on the surface 196

characteristics of the material such as hydrophobicity, charge, roughness and its components (17). Therefore, the design of scaffold surfaces can control cell behaviors and functions such as adhesion, proliferation, and differentiation. With regard to porosity, topologically-controlled scaffolds including nano-meter sized fiber scaffolds constructed by electrospinning (18) or self-assembling amphiphilic peptides (19) have attracted much attention due to their high porosity and the controlled alignment of the fibers. However, 3D-engineered tissues possessing precisely-controlled cell types and cell alignment, but cell-cell interactions have not been developed yet (15). The existing tissue models with biodegradable hydrogels and fiber scaffolds seem to have several limitations in satisfying the above requirements. A bottom-up approach using multiple cell types as pieces of tissue can be expected to solve these problems. Currently, various methodologies such as cell sheet engineering (20–22), the cell beads method (23) and magnetic liposomes (24) have been reported in the development of multilayered tissue constructs. Although these methods are intriguing examples for the bottom-up approach, they have limitations due to the complicated manipulation of the fragile cell sheets or remnant magnetite particles inside the cells. A breakthrough in technology for in vitro tissue construction would attract novel methods for regenerative medicine and drug assessment.

Hierarchical Cell Manipulation Technique In general cell culture, the cells proliferate on the appropriate substrates in growth media to reach a confluent condition, and the cells stop their growth due to contact inhibition. This is why it has been difficult to obtain 3D-multilayered tissues under normal conditions, except for cell lines which can proliferate indefinitely. Moreover, even if the cells were seeded onto adhered cells, they did not adhere and organize completely because of the lack of an ECM on the cell surfaces and electrostatic repulsion between the cells. Therefore, to make 3D-tissues in vitro, a proper substrate for cellular adhesion should be employed. Given the cellular adhesion mechanism on substrates, the cells recognize and adhere to the substrates directly as well as the components of the ECM through specific interactions between the ECM proteins and membrane proteins on the cell surface. We assumed that if the direct fabrication of nanometer-sized cell adhesive materials like ECM fibrous scaffolds onto the surface of a cell membrane is possible, then the cells seeded as the second layer may be able to recognize the adhesion substrates and adhere onto the first monolayered cells to generate bilayer structures. Repeating this process would yield 3D-multilayered structures in vitro. To prepare this artificial ECM film onto cells, we focused on layer-by-layer (LbL) assembly, which is an appropriate method to prepare nanometer-sized films on a substrate by alternate immersion into interactive polymer solutions (25, 26). Researchers have extended LbL assembly to various fields, including biomedical applications for the control of surface properties at the nano-meter level (27, 28). Initially, electrostatic interactions were mainly utilized for LbL assembly, and subsequently other interactions such as covalent bonding (29), hydrogen bonding 197

(30), charge transfer (31), hydrophobicity (32), host–guest (33), and coordination bond (34) interactions have been investigated to facilitate polymer association for ultrathin film deposition. However, most components of the films formed by LbL assembly crucially affect cellular functions like cell viability. In particular, synthetic polymers show strong cytotoxicity, and even natural polymers can impair cellular functions if they contain cationic polymers which interact strongly with the cells. For example, Rajagopalan et al. constructed a bilayer structure composed of hepatocytes and other cells by preparing a polyelectrolyte (PE) multilayer consisting of chitosan and DNA onto the hepatocyte surface (35). However, chitosan cannot dissolve in neutral buffer, and the use of PE films as a cell adhesive material was limited due to the cytotoxicity of the polycations (36). The appropriate choice of natural ECM components for the preparation of nanofilms is important for biocompatibility and stable cellular adhesion on the cell surface. We selected fibronectin (FN) and gelatin (G) to prepare nano-ECM films on the cell surface. FN is a flexible, multifunctional glycoprotein, and plays an important role in cell adhesion, migration, and differentiation (37). FN can interact not only with a variety of ECM proteins such as collagen (gelatin) and glycosaminoglycan, but also with the α5β1 integrin receptor on the cell surface (38). Even though FN and G have a negative charge under physiological conditions, they interacted with each other through a collagen binding domain (Kd = 0.6 ~ 5 x10-6 M) in FN, which is a different driving force versus PE films using electrostatic interaction. Thus, FN-G nanofilms are expected to serve as an artificial nano-ECM scaffold to fabricate 3D-multilayered tissues by providing a suitable cell-adhesive surface similar to the natural ECM without any cytotoxicity. Using FN-G nanofilms on the cell surface, the in vitro fabrication of 3D-multilayered tissues can be achieved. Fabrication of Multilayered Structure by Nano-ECM Coating A schematic illustration of the hierarchical cell manipulation technique for the construction of 3D-multilayered tissues is shown in Figure 1 (39). The LbL assembly of FN and G onto the cell surface was analyzed quantitatively using a quartz crystal microbalance (QCM) as the assembly substrate, and with a phospholipid bilayer membrane as a model cell membrane (Figure 2a). A phospholipid bilayer composed of 1,2-dipalmitoylsn-glycero-3-phosphatidylcholine (DPPC) and1,2-dipalmitoyl-sn-glycero3-phosphate (DPPA) was prepared onto the base layer, a 4-step assembly of poly(diallyldimethylammonium chloride) (PDDA) and poly(sodium styrenesulfate) (PSS), according to Krishna’s report (40). The mean thickness of the LbL assembly after 1, 7, and 23 steps was calculated to be 2.3, 6.2, and 21.1 nm, respectively. In addition, we have previously reported FN-based LbL multilayers composed of FN and ECM components, such as gelatin, heparin, and elastin (41). This FN-G nanofilm was formed on cultured mouse L929 fibroblasts by the alternate immersion of the culture substrate into FN and G solutions. Top and cross sections of confocal laser scanning microscopic (CLSM) 3D-merged images indicated a homogeneous assembly of fluorescently-labeled 198

FNG nanofilms on the surface of mouse L929 fibroblasts. Quantifying the fluorescent intensity of the FN-G nanofilms formed on the cell surface by line scanning revealed that the fluorescence intensity of the Rh-FN-G nanofilms increased linearly upon increasing the LbL assembly step number, similar to the frequency shift of the QCM analysis, thus indicating a clear increase in the film thickness on the cell surface (Figure 2b). These results demonstrated the fabrication of FN-G nanofilms on the cell surface. To check whether FN-G nanofilms can function as adhesion substrates, the overlap of the first and second layer of cells was confirmed by phase and fluorescent microscopic observation (Figure 2c-i). When the 7-step-assembled FNG nanofilms were prepared on the surface of the first L929 cell layer, the second layer cells were then observed on the first cell layer (Figures 2c,d). However, when the nanofilm was not prepared or the 1-step-assembled nanofilm (only FN) was assembled on the first cell layer, then the bilayer architecture was not observed (Figures 2e-h). These results suggest that at least 6 nm-thick FN-G films were required for stable adhesion on the cultured cells, and 2.3 nm films did not work. This was partly because the amount of FN and G deposited where membrane proteins like integrin can bind through biological recognition was involved in stable cellular adhesion.

Figure 1. Schematic illustration of hierarchical cell manipulation using the LbL assembly technique. Fibronectin (FN) and G were used as components of the LbL assemblies using biological recognition. Reproduced with permission from ref. (27, 28). Copyright 2012 Wiley and Copyright 2012 The Chemical Society of Japan. 199

Figure 2. (a) QCM analysis of the LbL assembly of FN and G onto a phospholipid bilayer (DPPC/DPPA 4:1). The open and closed circles show the assembly steps of G and FN, respectively. The base layer was a 4-step assembly of PDDA and PSS. (b) Relationship between the LbL assembly step of Rh-FN-G on L929 cells and the fluorescence intensity of the Rh-FN. The inset shows the fluorescence intensity of the cells (green) and the 7-step-assembled Rh-FN-G nanofilms (red) by line scanning. The cells were labeled with cell tracker green. (c,e,g) Phase-contrast and (d,f,h) fluorescence microscopic images of L929 fibroblast bilayers (c, d) with (e, f) and without a 7-step-assembled, or (g, h) with a 1-step-assembled FN-G nanofilms on the surface of the first cell layer. (i) Ph image of a L929 cell monolayer as the control. The nuclei were labeled with DAPI. Reproduced with permission from ref. (28). Copyright 2012 The Chemical Society of Japan.

200

Four layered (4L) cellular multilayers were clearly observed after four repetitions of these steps using CLSM and hematoxylin and eosin (HE) staining images (Figures 3a,b). The thickness of the obtained multilayers estimated from 3D-reconstructed CLSM images increased linearly with increasing cell layer number (Figure 3c). This method can be applied to a variety of cells such as myoblast cells (Myo, Figure 3d), cardiac myocytes (CMyo, Figure 3e), smooth muscle cells, hepatocytes, and endothelial cells. The obtained 3D-cell architectures showed sufficient mechanical strength via the production of ECM and intercellular adhesion to peel off from the substrate easily (Figure 3f). Furthermore, the 3D-multilayered fibroblast tissues exhibited high cellular viability and maintained their layered structures even after 28 days of culture (Figure 4) (42). There was no significant difference in the living cell density from the top to the bottom layers during the 28 days of culture, suggesting that some cells proliferated and replaced the empty spaces left by the dead cells to maintain the density of living cells.

Effect of Nanofilms on Cellular Function To understand the effects of nanofilms on the cell surface on cellular functions, we confirmed their morphology and fundamental functions after the preparation of various types of PE and FN films (Figure 5). Although cellular functions on various LbL films have been previously reported with regard to adhesion, proliferation, and differentiation (43–45), it is not appropriate to investigate the direct interactions between PE films and the cell membrane because of the adsorption of serum proteins between them during cellular adhesion. Therefore, we focused on the effects of various LbL nanofilms that were prepared on the cells directly on cell functions, including the charge, thickness, composition, and morphology of the LbL films (46). Table 1 summarizes the nanofilm components such as synthetic polymers, polysaccharides, poly(amino acid)s and proteins, and their effects on cellular functions. As shown in Figure 5a-e, FN-G and FN-dextran sulfate (FN-DS) nanofilms displayed more extended morphologies as compared to PE films. Cells formed on the PE films on the cell membrane had rounded shapes, even though the polymers contained in the PE films did not show any cytotoxicity by themselves (Figure 5f,g). Similarly, PE films drastically affected the cell viability and proliferation upon increasing the thickness of the films. On the other hand, FN films formed through protein-protein recognition exhibited high cytocompatibility and nanometer-sized meshwork fiber structures were observed after 24 hours by scanning electron microscopy (SEM) (Figure 5h). The FN in the films prepared onto the cells transformed their morphologies from films to nano-meshwork fibers, but not the FN in serum, whereas the formation of fiber structures were never observed when PE films were formed (Figure 5i,j). These nano-meshwork morphologies seemed to be similar to the fibrous structure of the natural ECM.

201

202 Figure 3. (a) 3D-reconstructed CLSM cross-sectional image of 4-layered (4L) L929 cells. The cells were labeled with cell tracker green. (b) HE staining image of 4L-L929 cells. (c) Relationship between the L929 cell layer number and the mean thickness estimated from 3D-CLSM images (n = 3). (d) HE staining images of 1-, 3-, and 7L-mouse C2C12 Myo. (e) HE image of 5L-rat CMyo. (f) Photograph of the peeling process of 4L-L929 cells from a coverglass as the substrate. Reproduced with permission from ref. (28). Copyright 2012 The Chemical Society of Japan.

Figure 4. (a) Relationship between the living cell density of each layer in 5L-NHDF tissues and the culture time over 1 month. The error bars represent the standard deviation (n = 4–5). (b) Schematic illustration of the replacement of the empty spaces by the proliferation of living cells. Reproduced with permission from (42). Copyright 2013 Elsevier.

Furthermore, the secretion of interleukin-6 (IL-6) cytokine, which is an inflammatory cytokine, from normal human dermal fibroblasts (NHDFs) with PE films expressed approximately 2.5-fold higher IL-6 per single cell than NHDFs without films (Figure 6). On the other hand, the FN films did not show any effect on IL-6 production. These results suggested that PE films prepared on the cell surface induced strong inflammation in the cells, probably due to cationic cytotoxicity, leading to cell death or growth arrest dependent on the thickness, cationic charge, or cationic species.

Control of Cellular Function and Activity in 3D-Environments An important advantage of this hierarchical cell manipulation technique is to allow the cells to grow in 3D-environments composed of layered cells and ECM structures. In general, cells are cultured on a plastic dish, which is a completely different condition from living tissue. If the use of 3D-environments which mimic the structures of living tissues can enhance cellular functions and control cell activity, they will be useful tools as compared to cell monolayers, not only for understanding how a 3D environment composed of cells, ECM, and signaling molecules regulates functions similar to natural tissues, but also for creating 3D-artificial tissues resembling natural tissues. The basic properties induced by 3D-cellular structures, such as the layer number or the cell types, have not been clarified yet. 203

Figure 5. (a-e) Phase images of L929 fibroblasts (a) without or with (b) 8 nm thick FN-G, (c) 121 nm thick FN-DS, (d) 17nm thick FN-ε-Lys, and (e) 18 nm thick ε-Lys-PSS nanofilms on the cell surfaces after 24 h of incubation. (f) Relationship between cell viability and the thickness of various nanofilms prepared on the cell surfaces after 24 h of incubation (n = 3). PAH, PAA, and ε-Lys: poly(allylamine hydrochloride), poly(acrylic acid), and ε-poly(lysine hydrochloride), respectively. (g) Cell proliferation versus various nanofilms prepared on L929 cell surfaces during 72 h of incubation. (h) SEM image of the L929 cells on 121 nm thick FN-DS films after 24 h of incubation. Fluorescence microscopic images of: (i) 21nm thick Rh-FN-FITC-G and (j) 17 nm thick Rh-FN-FITC-ε-Lys films prepared on L929 cells after 24 h of incubation. Reproduced with permission from ref. (28). Copyright 2012 The Chemical Society of Japan.

204

Table 1. Summary of the nanofilm components and cell types used in this paper.

205

Figure 6. IL-6 secretion versus amount of DNA in normal human dermal fibroblasts (NHDFs) without or with FN based (FN-G and FN-DS) or polyelectrolyte (PE) films on the cell surfaces after 24 h of incubation (n = 3). Asterisk denotes a statistically significant difference using a two sample t-test (*p < 0.01) for each comparison. Reproduced with permission from ref. (47). Copyright 2012 The Chemical Society of Japan.

206

207 Figure 7. Histological images of 4L-NHDF/1L-HUVEC constructs stained for: (a) HE and (b) factor VIII. Fluorescent immunostaining images of (c, d) 1L-NHDF/1L-HUVEC and (e,f) 4L-NHDF/1L-HUVEC constructs immunostained with an anti-CD31 antibody (g). Production of Hsp70 versus total protein from non-heat shocked and heat shocked layered constructs composed of NHDFs and HUVECs (n = 4). The heat shock conditions were 20 min incubation at 45 °C, followed by a 2 h recovery period. (h) IL-6 secretion versus total protein from the layered structures (n = 3). Asterisk denotes a statistically significant difference using a two sample t-test (*p < 0.01, **p < 0.05) for each comparison. Reproduced with permission from ref. (48). Copyright 2010 Elsevier.

To understand how the 3D-environment influences cell behavior and function, we fabricated a layered structure with a human umbilical vein endothelial cell (HUVEC) layer on top of multilayered NHDF tissue (Figure 7a,b) (48). The two types of layered constructs, 1L-NHDF/1L-HUVEC and 4L-NHDF/1L-HUVEC, were prepared using FN-G nanofilms. To evaluate the general effects of 3D-layered structures on cellular stress and inflammation, we purposely fabricated biologically meaningless layered structures consisting of endothelial cells and fibroblasts. Interestingly, the HUVECs adhered homogeneously on the surface of 4L-NHDF tissue at the centimeter scale, whereas heterogeneous HUVEC domain structures were confirmed on the 1L-NHDF tissue (Figure 7c-f). This result suggests that the multilayered structure was able to provide a suitable adhesion scaffold for HUVECs, and the HUVECs can recognize their microenvironments such as stiffness, ECM deposition, and cell-cell interactions. Moreover, when we investigated protein production in layered tissues as a cell function, there were big differences between 1L- and 4L-NHDF tissues with regard to the production of heat shock protein 70 (Hsp70) and IL-6 (Figure 7g,h). The Hsp70 expression of HUVECs decreased after adhesion onto the 4L-NHDF structure as compared with the HUVEC monolayer. Surprisingly, the Hsp70 production response to heat shock increased drastically by approximately10-fold as compared to non-heat shock by 3D structure formation, whereas the monolayer structures showed no change. Moreover, the production of the inflammatory cytokine IL-6 decreased drastically depending on the layer number of NHDFs. These results suggest that 4L-NHDF provided a more favorable environment for HUVECs as compared to a plastic cell culture dish for the induction of high thermosensitivity, and to suppress inflammatory responses from the substrate. Permeability Assay of Multilayered Fibrous Tissues As described above, we found that three-dimensionally cultured cells expressed higher cellular functions and activities as compared to monolayer-cultured cells. In a similar fashion, the 3D-structures of cells and the ECM allow for 3D-pharmacokinetic analysis in drug testing such as the permeation, distribution, and accumulation of drugs. Therefore, we performed a permeability assay using multilayered fibrous tissues constructed by the hierarchical cell manipulation technique (49). We focused on pancreatic cancer, which is known to have ab extremely low median survival time for advanced pancreatic adenocarcinoma, approximately 6 months. The reason why treatments against pancreatic cancers have failed despite recent progress with conventional chemotherapies is their hypo-vascular structure and the large amount of fibrous tissues. Histologically, human pancreatic tissue, aside from the tumor cells, shows a large volume of fibrotic tissue with occasional vasculature distal to the tumor cells (Figure 8a). Since the amount of drugs that can be delivered is low and drug permeation is poor due to their structure, drug delivery systems using nanometer-sized carrier (nanoDDS) as a novel therapeutic method has attracted increasing attention to achieve efficient drug delivery and to reduce side effects. However, there are no satisfactory models of drug permeability in fibrotic tissues to design better drug carriers. The permeation of drugs, including nanoparticles, 208

in tissues has been investigated in vitro using endothelial cells or epithelial cells cultured in monolayers (50, 51), but the three-dimensional migration of drugs cannot be analyzed in a monolayer system. Many three-dimensional models have been used to mimic cells, including fibroblasts in an in vivo environment (52), but they have not addressed the problem of drug permeability. For example, collagen gel is widely used because it resembles the ECM, but it shrinks after a few days of cell incubation, and hence cannot be used to test permeability.

Figure 8. Histological pattern of human pancreatic cancer and the structure of the model. (a) H&E staining of a pathological specimen of human pancreatic cancer. Most of the tumor tissue is occupied by fibrotic tissue. T, tumor cells (inside blue dotted circles), BV, blood vessels (inside red dotted circles), and F, fibrotic tissue (other parts). (b) Schematic illustration of the multilayered model for the permeation assay. (c) Transmitted light microscopy of the multilayered culture. (d) H&E staining of the multilayered culture. (E) Phalloidin staining of the multilayered culture. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article. Reproduced with permission from ref. (49). Copyright 2012 Elsevier. We analyzed the diffusion of dextran molecules as a model drug across engineered fibrotic tissues composed of fibroblasts derived from a murine spontaneous pancreatic tumor (K643f) (Figure 8b). We confirmed the integrity and homogeneity of the structure by transmission light microscopy and HE staining, and also observed the fibrous structures by phalloidine staining. Using this multilayered fibrous tissue, the permeation of fluorescently-labeled, high molecular weight dextran was measured to investigate the delivery of nanoDDS into tumor fibrotic tissue. First, we tested the relationship between the layer number and permeation using 1L-, 2L-, and 5L-fibrous tissues. As shown in Figure 9a, there were significant differences between the 1L and 2L or 5L tissues, suggesting the permeation of dextran molecules decreased with an increasing number of tissue 209

layers. Next, we confirmed the time-dependent permeation of dextran molecules (Figure 9b). The permeation of 250 kDa and 2000 kDa dextran molecules, 12 and 31 nm in hydrodynamic diameters respectively, decreased with an increasing layer number of tissues. These results indicated that the multilayered fibrous tissues can recreate a microenvironment similar to fibrous tissues in vivo better than existing monolayered cell models. Thus, our three dimensional model will be useful for analyzing the permeation of drugs, including nanoparticles, for diseases such as pancreatic cancer.

Figure 9. (A) Permeation of 20 kDa dextran across 1L-, 2L-, and 5L-K643f tissues after 24 hours of incubation. (Bars: SD, /: p < 0.05, and //: p < 0.01. (B)) Dextrans of 250 kDa or 2000 kDa were tested in 1 or 2 layers of K643f culture. (*p < 0.01, **p < 0.05, n = 3). Reproduced with permission from ref. (49). Copyright 2012 Elsevier.

Blood Vessel Wall Model In physiology and medicine, blood vessels are relevant not only for circulatory diseases and treatments, but also for the biological evaluation of drug diffusion into target tissues, the penetration of cancer cells or pathogens, and the control of blood pressure. A blood vessel is generally composed of three distinct layers: the intima, an inner single layer of endothelial cells (ECs); the media, the middle layers of smooth muscle cells (SMCs); and the adventitia, an outer layer of fibroblast cells (6). Each layer serves specific functions such as anti-thrombogenicity and the selective permeation of molecules in the EC layer and mechanical strength in the SMC layer. To construct such a functional tissue in vitro, a hierarchical layered structure composed of EC and SMC layers should be formed. We employed the hierarchical cell manipulation technique for the construction of a blood vessel wall model of human ECs and SMCs with a layered structure (53). Although long-term culture for several months and additives to promote cell proliferation unnaturally were needed for the construction of multilayered SMC tissues, this method just repeats the FN-G nanofilm formation and cell seeding. Therefore, the obtained blood vessel wall models were expected to display similar structures and functions as living blood vessels. 210

Construction of Blood Vessel Wall Model As shown in Figure 10a, the layered structures composed of 4L-SMC and 1LHUVCEC were confirmed by immunohistological observation, and monolayered HUVECs homogeneously covered the surface of the 4L-SMC layers, as well as the human aorta. Checking whether the HUVEC layer maintained its functions as endothelial cells on the SMC layer by platelet adsorption, the 4L-SMC/1LHUVEC model suppressed the adsorption and activation of platelets in human platelet-rich plasma (PRP) (Figure 10b,c).

Figure 10. (a) Immunohistochemical evaluations of 4L-SMC/1L-HUVEC constructs (left), and the human aorta (right). The histological image of the human aorta was focused on the inner HUVEC layer for comparison against engineered structures. (b) SEM images of the surface of the 4L-SMC and 4L-SMC/1L-HUVEC constructs after immersion into PRP for 15min at 37 °C. The arrows indicate adhered and activated platelets on the surface of the 4L-SMC constructs. (c) The number of adhered platelets on 4L-SMC/1L-EC and 4L-SMC counted from SEM images at 120 μm x 90 μm areas (n = 3). The asterisk denotes a statistically significant difference using a two sample t-test (*p < 0.01). Reproduced with permission from ref. (28). Copyright 2012 The Chemical Society of Japan. Quantitative 3D Analysis of Nitric Oxide Using Blood Vessel Wall Model We applied this artificial 3D-blood vessel wall model to the biological evaluation of nitric oxide (NO) production from ECs (54). NO produced from the ECs diffuses into the SMCs through their cell membrane, and activates guanylate cyclase to produce intracellular cyclic guanosine monophosphate (cGMP), which induces a signaling pathway mediated by kinase proteins leading 211

to SMC relaxation (55). Therefore, quantitative, kinetic, and spatial analyses of the extracellular delivery of NO molecules in the blood vessel wall are crucial for pharmaceutical evaluations of hypertension and diabetes. Recently, a significant correlation between NO production and diabetes mellitus was reported; reduced NO production in both type 1 and 2 diabetes (56, 57) and also increased NO production related to improvements in type 2 diabetes were reported (58). Thus, the development of a convenient and versatile method for the in vitro quantitative and spatial analysis of NO diffusion in the artery wall is important for biological and pharmaceutical applications and as an alternative to animal experiments. To quantify NO production three dimensionally, NO sensor particles of mesoporous silica were employed in the 3D-blood vessel wall model. We reported biocompatible and highly-sensitive NO sensor particles using LbL assembly. The mesoporous silica micro-particles encapsulated 4,5-diaminofluorescein (DAF-2), which is a NO fluorescent indicator dye, by covering the porous surfaces with PE multilayered films composed of chitosan (CT) and DS. The CT-DS films allowed the NO sensor particles (SPs) to retain the dye inside the particles and to exhibit fluorescence in response to the NO molecules. The SPs showed high NO-sensitivity at 5500 nM, which is sufficient to detect NO at concentrations of hundreds of nM (EC production level (59)) (60). We combined 3D-blood vessel wall models with the SP technique for the 3D-detection of NO production from ECs (Figure 11). As a model of the blood vessel wall, 3D-artery models composed of 4L-human aortic smooth muscle cells (AoSMCs) and 1L-human aortic endothelial cells (HAECs) were constructed by Nano-ECM coating as described above. The SPs were seeded into each layer in the tissues at the same time as the cell seeding to obtain layered artery models with homogeneously-distributed SPs in the tissues. The amount of NO production from the ECs was calculated from the fluorescent intensity of SPs obtained by CLSM observation. To evaluate the 3D-structural effect of the cellular alignment on localized NO production, the NO concentration at the apical and basal membranes of the HAECs in monolayers or bilayers after 48 h of incubation was analyzed using the SPs (Figure 12a). The NO concentrations at the basal membrane of HAECs in the monolayer and bilayer were 4.5-fold and 2-fold higher, respectively, than those on the HAEC surfaces, suggesting that the NO molecule was more readily produced at the basal membrane than at the apical membrane, independent of the 3D-layered structure. This is in agreement with a report by Ortiz and coworkers (61). Finally, we performed 3D-analyses of NO diffusion using 5L-artery wall models by CLSM observation (Figure 12b-d). The first layer is HAECs, and the second to fifth layers were AoSMCs. The fluorescent intensity of SPs with DAF-2 increased in response to NO molecules (green), and the nuclei were stained with DAPI to distinguish each layer. The NO concentrations in first (HAEC) and second (AoSMC) layers were about 530-550 nM, and then gradually decreased upon increasing the layer number. The NO concentration reached about 50% (290 nM) in the fifth layer (32 nm in depth). Although NO diffusion was limited to approximately 60 μm in 3D-artery wall models and this value is slightly lower than the reported in vivo NO diffusion distance (approximately 100-200 nm), this might be reasonable considering the NO consumption (approximately 37% 212

consumption) during the chemical reactions (62). These results suggest that this technique will be a useful method for in vitro artery assays to replace in vivo animal experiments.

In Vitro Optimization of Nanoparticles Using Blood Vessel Wall Models The estimated cost of research and development (R&D) for a single new drug molecule has been reported to exceed $1.8 billion (63), and improvements in R&D productivity to reduce costs is a key challenge. One potential solution is the acceleration of the preclinical development and assessment processes. Although preclinical animal experiments are absolutely needed to test pharmacological and pharmacokinetic responses, they are usually complex, less repeatable, time consuming, and produce different results than shown in humans. These issues have also affected the development of nanomaterials. Recently, nanomaterials have been more widely applied in biomedical applications and they could become indispensable for use in medical and diagnostic tools in the near future. However since it is generally difficult to optimize the in vivo three dimensional (3D)-biological and physicochemical behaviors using in vitro cell assays, there have been many failures using animal models. Accordingly, if the in vivo biological properties and behaviors of drugs and nanomaterials (e.g., diffusion, accumulation, and toxicity inside the target tissues) can be predicted partly at an initial in vitro assessment during drug discovery, it will serve to dramatically improve the R&D process. To this end, we focused on in vitro reconstructed 3D-tissues consisting of multiple cell types. 3D-spheroids and the other engineered tissues have been studied when conducting toxicity and activity assays of pro-drugs during drug discovery, especially for hepatic tissues (64, 65). However, engineered tissues might also have promising for the “in vitro prediction of the in vivo 3D-behavior of drugs and drug carriers”, since 3D-engineered tissues can partly replicate the 3Dproperties and environment of in vivo living tissues. Although physicochemical properties are usually evaluated by in vitro analysis, biological properties and 3D-behavior inside tissue cannot be evaluated until in vivo animal experiments are performed. Here we demonstrate the “in vitro prediction of the in vivo behavior of drug delivery nanocarriers”. We have previously demonstrated the 3D-diffusion of nitric oxide (NO) molecules inside 3D-arterial wall models (3D-AWMs) in response to a hormone peptide, where 3D-AWMs were composed of endothelial cell (ECs) layers and underlying multilayered smooth muscle cells (SMCs) (54). This result encouraged us to apply 3D-AWMs to the in vitro prediction of the biological properties of nanoparticles (NPs) for drug delivery. In this study, atherosclerosis and paclitaxel (PTX) were selected as a target disease and drug, respectively. Atherosclerosis, which affects the vasculature systemically, poses a serious risk to the coronary artery because rupture may lead to myocardial infarction (66). Migration and overgrowth of SMCs occurs as a cascade of events and results in a buildup of plaques which thicken the arterial wall 213

(67). 3D-structures consisting of multilayered SMCs and ECs are suitable for reproducing the coronary arterial environment. Generally, monolayer cultures of vascular ECs, or co-culture models of vascular ECs and SMCs which are cultured on both sides of the substrate membrane, are used as an in vitro model to study the response of cells to drugs or NPs (68, 69). For effective treatment, the drug carriers not only need to target the diseased tissue, but also to penetrate the tissue and remain long enough before being removed from the body through the blood clearance system (70, 71). We describe the in vitro prediction of 3D-penetration, accumulation, cytotoxicity, and treatment effect of NPs using 3D-AWMs and conduct a proof of concept (POC) using a mouse model of atherosclerosis (Figure 13) (72). 3D-AWMs consisting of 1L-EC grown on 4L-SMCs (5L-AWMs) were used in this study. We previously reported PTX-loaded L-phenylalanine ethyl estergrafted-poly(γ-glutamic acid) NPs (γ-PGA-NPs-PTX) with a diameter of 180 nm and evaluated their anti-proliferative effect on the SMC monolayer (73). The NPs were able to lower cell proliferation for at least 6 days in culture after incubation for only 1 h. Nevertheless, the effect of the drug in the cell monolayer was different from that in the 3D-culture, and is even more distinct in vivo (74). Accordingly, it is necessary to evaluate the effect of the NPs in the 3D-AWMs to obtain more accurate results for the treatment of atherosclerosis.

Figure 11. Schematic illustration of the in vitro spatial and quantitative analyses of NO diffusion from the uppermost HAEC layer to the AoSMC layers in a 3D-artery model using SPs. The SPs were homogeneously distributed within the 3D-layered blood vessel wall model. Reproduced with permission from ref. (54). Copyright 2011 Wiley.

214

Figure 12. (a) Localized NO diffusion in relation to 3D-structural effects. The localized NO concentrations were analyzed by SPs on and under the HAECs in monolayers and bilayers after 48 h of incubation. The mean fluorescence intensity of the SPs was measured from CLSM images (n = 3, over 20 SPs per image), and the NO concentration was estimated from a calibration curve. If: fluorescence intensity. (b) 3D-reconstructed CLSM image of a 5L-artery model containing SPs (green) in each layer after 48 h of incubation. The 1st layer is HAECs, and the 2nd to 5th layers are AoSMCs. (c) Cross-sectional CLSM image of a) in the white arrow direction. The dashed lines indicate the brief interfaces of each layer. (d) The localized NO concentrations in each cell layer were analyzed using the SPs. The fluorescence intensities of the SPs were measured from each layer in (c) (n = 3, over 20 SPs per image, **p < 0.05, *p < 0.01)). Reproduced with permission from ref. (54). Copyright 2011 Wiley.

215

Figure 13. Overview of this study. (a) Evaluation of the transport of NPs with various factors across the 5L-AWM (1L-EC4L-SMC) in vitro, and the effects of the NPs EC barrier. (b) The NPs with properties optimized for use as drug carriers were loaded with PTX and evaluated in vivo using a mouse model of atherosclerosis. Reproduced with permission from ref. (72). Copyright 2016 Wiley. In Vitro Optimization of NPs Using 5L-AWMs SMCs are known to become a proliferative phenotype when cultured in vitro in the presence of serum protein (69). We therefore used 5L-AWMs as a model of early stage atherosclerosis due to their growth during culture. The growth of 5L-AWMs during these 2 weeks was effectively suppressed when the NPs were applied for 24 h on day 1. A dose-dependent effect of the NPs on growth suppression was observed when the concentration of PTX was lower than 50 µg/mL. On the other hand, free PTX did not suppress growth more than γ-PGA-NPs-PTX. A single application of the NPs for 24 h was able to effectively suppress the growth of the 5L-AWMs, for at least two weeks, suggesting accumulation of sufficient amounts of the NPs in the tissue even after washing. However, most of the free PTX was removed during washing, therefore resulting in a lower suppressive effect. In our previous report, the NPs still remained in the SMCs after incubating for 1 h and culturing for 6 days, suggesting rapid cellular uptake and long-term intracellular stability (73). PTX was spontaneously released by desorption from the NPs in the physiological environment, suppressing cell growth. PTX is known to have high cytotoxicity to SMCs in a monolayer culture even at a concentration of 10 ng/mL (75). In this study, the 5L-AWMs were incubated with free PTX at a concentration of 100 µg/mL, the concentration showing the highest suppressive effect. However, the surviving cells proliferated, resulting in thicker 5L-AWMs over the 2 weeks of culture. This demonstrated the tolerance to the drug in 3D-culture. The 3D-AWMs should give more accurate results on the dose-dependent effect of the drug for the treatment of atherosclerosis. Histological sections of the 3D-AWMs stained with anti-CD31 antibody showed the ECs inside the SMC layers after 2 weeks, due to the overgrowth of SMCs as occurs in atherosclerosis. The stable presence of ECs on the top surface was only found in the 5L-AWMs treated with NPs. We also applied the NPs to examine the treatment effect on the over-grown 4L-SMCs. The 4L-SMCs had been cultured for 5 days before the application of the NPs to allow for the overgrowth of the SMCs as occurs in atherosclerosis. When 5 µg/mL γ-PGA-NPs-PTX was applied, the growth of the structure was effectively 216

suppressed and the situation at day 19 seemed similar to what was observed at day 11 without treatment. The therapeutic effect of the drug carriers could be estimated in vitro using 3D-AWMs. To understand the effect of the properties of the substance on cytocompatibility, we evaluated the permeability of the NPs and the barrier functions of the EC layers after applying various NPs with a similar size of approximately 50 nm, as used in our previous permeability assays (Figures 14a,b) (76). Fluorescein-labeled γ-PGA-NPs, fluorescein isothiocyanate (FITC)-labeled polystyrene (PS) NPs, and FITC-labeled silica (Si) NPs which represent examples of biodegradable NPs, non-biodegradable NPs, and inorganic NPs, respectively, were used. All NPs exhibited negative surface charges (zeta potentials of γ-PGA-NPs, PS-NPs, and Si-NPs were -30.6, -50.2, and -18.7 mV, respectively). Interestingly, even though all NPs were larger than the tight-junctions of the EC layers (< 2 nm) (77), they showed high permeability greater than 40%. In particular, the PS-NPs exhibited the highest permeability and the lowest tendency to accumulate. Kim and co-workers reported higher permeability of EC monolayers by treating with tumor necrosis factor-α (TNF-α) which induces disruption of intercellular junctions (e.g., adherens junctions) same condition as inflammation (atherosclerosis) (78). Cellular uptake of γ-PGA-NPs and PS-NPs could be observed after 24 h of incubation. Si-NPs did not show significant cellular uptake, but became a cluster (