Advances in Bioinspired and Biomedical Materials Volume 1 9780841232198, 9780841232204, 9780841232228, 0841232202

Citation preview

Advances in Bioinspired and Biomedical Materials Volume 1

ACS SYMPOSIUM SERIES 1252

Advances in Bioinspired and Biomedical Materials Volume 1 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. Identifiers: LCCN 2017044797 (print) | LCCN 2017050880 (ebook) | ISBN 9780841232198 (ebook) | ISBN 9780841232204 (volume 1) | ISBN 9780841232228 (volume 2) Subjects: LCSH: Biomimetic materials. | Biomedical materials. Classification: LCC R857.M3 (ebook) | LCC R857.M3 A379 2017 (print) | 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

DNA/Protein/Peptide Self-Assembly 1.

TRAPped Structures: Making Artificial Cages with a Ring Protein .................. 3 Jonathan G. Heddle

2.

Integrated Nanostructures Based on Self-Assembled Amphiphilic Polypeptides ............................................................................................................ 19 Motoki Ueda, Stefan Müller, Siyoong Seo, Md. Mofizur Rahman, and Yoshihiro Ito

3.

Peptides as Smart Biomolecular Tools: Utilization of Their Molecular Recognition for Materials Engineering ................................................................ 31 Toshiki Sawada and Takeshi Serizawa

4.

DNA Condensed Phase and DNA-Inorganic Hybrid Mesostructured Materials ................................................................................................................. 49 Yuanyuan Cao and Shunai Che

Polypeptide and Engineered Proteins 5.

Adhesive Growth Factors Inspired by Underwater Adhesion Proteins ........... 83 Chen Zhang, Hideyuki Miyatake, and Yoshihiro Ito

6.

Polypeptides and Engineered Proteins ................................................................. 93 Xinyu Liu, Jin Hu, Zhuoran Wang, Zhikun Xu, and Weiping Gao

7.

Protein Self-Assembly: From Programming Arrays to Bioinspired Materials ............................................................................................................... 129 Quan Luo, Tiezheng Pan, Yao Liu, and Junqiu Liu

8.

Controlled Syntheses of Functional Polypeptides ............................................. 149 Zhongyu Jiang, Jinjin Chen, Jianxun Ding, Xiuli Zhuang, and Xuesi Chen

Catechols/Dopamine Derivatives 9.

Bioinspired Wear-Protective Coatings for Osteoarthritis ................................ 173 Larry An, Sung Won Ju, Minsoo Park, Jihyung Kim, Haewon Choi, Song Hoe Koo, Jinsoo Ahn, and Kollbe Ahn

vii

10. Catechol Redox Reaction: Reactive Oxygen Species Generation, Regulation, and Biomedical Applications .......................................................... 179 Pegah Kord Forooshani, Hao Meng, and Bruce P. Lee Editors’ Biographies .................................................................................................... 197

Indexes Author Index ................................................................................................................ 201 Subject Index ................................................................................................................ 203

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 1 includes sections (i), (ii), and (iii) and focuses on the bioinspired approaches using biological macromolecules including poly(nucleic acids), polypeptides, and the derivatives. 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

DNA/Protein/Peptide Self-Assembly

Chapter 1

TRAPped Structures: Making Artificial Cages with a Ring Protein Jonathan G. Heddle* Bionanoscience and Biochemistry Laboratory, Malopolska Centre of Biotechnology, Jagiellonian University, Gronostajowa 7, 30-387, Krakow, Poland *E-mail [email protected]

Protein cages are hollow containers typically made from multiple copies of a small number of protein subunits. They occur extensively in nature and morphologies can vary but typically they are close to spherical in shape and have diameters in the nanometers to tens of nanometers range. Because central cavities of protein cages are bounded, they offer possibilities for use as nanoreactors or as transport systems in vivo and in fields as diverse as materials science and medicine. It is now possible to produce artificial cage proteins, thus widening their potential further. Here we consider in brief natural and artificial protein cages before focusing on the development of a ring-shaped protein (TRAP) leading up to and including its use as a component of an artificial protein cage.

Naturally occurring protein cages are well known in bionanoscience and perhaps show greatest potential for medical use. This is illustrated by virus capsids, the protein shells of viruses that protect, transport and mediate delivery of viral genetic material to host cells. Natural viruses are of course superbly adapted for this form of delivery, an ability that has been exploited in the development of gene therapy (1) and they can be further modified genetically or chemically to engineer-in desired properties, with recombinant DNA technology allowing the production of non-infectious virus particles free of genetic material (known as virus-like particles, VLP). It is now even possible to use advanced synthetic biology approaches in which completely artificial protein cages are designed, © 2017 American Chemical Society

typically in silico (albeit with naturally occurring proteins as the basic building blocks). This allows the construction of protein cages that do not exist in nature, thus opening up new potential morphologies and functionalities. The fact that naturally occurring protein cages are easily available and have been widely studied and characterized for decades means that they have benefitted from considerably more research in terms of applications-focused engineering compared to artificial cages. In many cases cages with known high-resolution structures were used, allowing for precise alterations in their structures at well-defined locations. Many excellent reviews in recent years have summarized the advances made in exploiting and developing natural protein cages (2–6). Some highlights include the use of the plant virus cowpea mosaic virus for a range of therapeutic applications (7–9) including, recently as an “in situ vaccine” with promising activity against a number of tumor models (10) wherein simple administration of the unmodified VLP resulted in neutrophil activation. The VLP derived from bacteriophage P22 virus has also shown much promise, particularly as a nano-reactor able to encapsulate functional enzymes (11, 12). Non-virus cages have also been used, most notably ferritin, which has been widely developed. In the cell, ferritin acts as a store of iron, protecting the body from this essential but potentially toxic metal. Ferritin has been shown to be able to mineralize other inorganic materials such as CdSe (13), CoPt (14), TiO2 (15) including some with possible uses in electronics (16). It can also be used to capture potentially useful therapeutic molecules, one example being the encapsulation of doxorubicin and its delivery to tumors (17). In this case the ability of ferritin to be modified with surface peptide sequences specific to cell surface proteins that are up-regulated in some tumors, proved useful (18). Another non-viral cage that has been developed is based on lumazine synthase, an enzyme from Aquifex aeolicus that naturally forms a cage structure. Engineering of this cage has been shown to result in versions able to capture cargo both in vivo (19) and in vitro (20). For example in in vivo work, amino acid residues lining the interior cavity can be mutated to those carrying a negative charge to give a negatively charged interior which is then able to capture a protein (e.g., GFP) that has been modified to contain a short, positively charged amino acid tag (19). Subsequently the protein container’s loading capacity was improved by a directed evolution approach where the cargo (in this case HIV protease) was toxic to the host cell meaning that changes in protein cage sequence resulting in more efficient encapsulation would be favoured. Improved cages were produced after only four rounds of selection (21). More recently lumazine synthase has been shown to be able to act as a nanoreactor, encapsulating an enzyme (ascorbate peroxidase) that is able to polymerise 3,3-diaminobenzidine (22), and it continues to be a promising protein in bionanoscience (23). One step beyond naturally occurring protein cages is the use of designed or artificial protein cages. Here proteins that do not naturally form cages are modified such that cage formation is promoted. As with all examples where designed, artificial versions of natural assemblies are produced, the artificial approach allows the construction of structures and development of capabilities that may not be available in nature and so widens the possibilities beyond those constrained by evolutionary pathways. Such a synthetic biology approach to protein cage 4

production has met with significant success in recent years and the field holds much promise. However, designing protein structures is not easy; the protein folding “paradox” (24) suggested that it is essentially impossible to accurately predict the amino acid sequence that will fold into a desired structure. It is now known that the problem may not be as insurmountable as initially thought (25) and in fact, great progress has been made in protein structure prediction (26). Nevertheless designing protein structures de novo is still extremely challenging and often not practicable. Therefore current approaches make novel structures by utilizing existing proteins that are stable enough to allow further modification to endow them with the ability to assemble into desired cage structures. Note that this approach, utilizing proteins, is distinct from a self-assembled peptide method whereby cages (or other) structures are built up from smaller, peptide subunits. Such a peptide strategy is useful in its own right and it was recently reported that a mixture of two short coiled-coil bundles could self-assemble into a large (~100 nm) diameter peptide cage (27). In general, cages made from proteins may be expected to have greater sequence/structural redundancy compared to peptide structures, allowing a larger number of modifications per monomer building block. In the design of protein cages there have been to date, two main approaches. In the first, naturally occurring proteins with different rotational symmetries are fused together with linker amino acid sequences to produce tandem proteins. In order for the preferred symmetries of the individual domains of these newly formed proteins to be satisfied, they form cage-like assemblies, typically with standard symmetries (icosahedral, octahedral etc.) This fusion approach was first shown in 2001 by Padilla et al. (28) who fused together bromoperoxidase (a homotrimer) and M1 matrix protein from influenza virus (a homodimer). Twelve such subunits assembled into a tetrahedral arrangement with the trimers forming the three-fold axes of the tetrahedron. This approach relies on existing protein-protein interfaces with the designed structures acting essentially as passive “rods” to link together the different proteins. Indeed, the structure of a modified version of this tetrahedron showed the importance of the linker as they demonstrated considerable bending and distortion (29). A second route to achieve assembly of artificial cages is to use computational approaches coupled with protein structural information where, instead of relying on existing protein-protein interactions, new interacting interfaces are designed into proteins of appropriate symmetry. For example the typical soccer-ball (icosahedral) shape would require twelve pentagons and twenty hexagons. Two proteins, one a pentamer with 5-fold rotational symmetry and one a hexamer with 6-fold rotational symmetry could in principle be designed with protein-protein interfaces between monomer edges to promote self-assembly into the icosahedral arrangement. In protein work this usually involves the design of a favourable interface (typically employing hydrophobic and/or electrostatic interactions) to allow stable packing between protein building blocks. This method has been demonstrated with much success by the group of David Baker (30–32). King et al (30) pioneered the approach, first choosing simple octahedral or tetrahedral point group symmetries as a model. Both include three fold symmetry axes, so by choosing homotrimer proteins, the three-fold protein and model axes could be aligned, allowing prediction 5

and subsequent optimization of the resulting protein-protein interfaces between monomers of neighboring timers. The first experiments confirmed production of a 13 nm diameter octahedral cage (made from 24 subunits of modified propanediol utilization polyhedral body protein from Salmonella enterica) and an 11 nm diameter tetrahedral cage (made from 12 copies of putative acetyltransferase SACOL2570 from Staphylococcus aureus) (30). Since then the size and complexity of protein cages has advanced rapidly with highlights including the use of two different proteins (trimers or dimers) that self-assemble to form tetrahedral (31); a 60-subunit icosahedral cage notable for its high thermostability (stable to over 80° C) (32) and the impressive design of a large number of protein cages in silico with the actual production and structural characterization of several of them. These cages range from 24 to 40 nm in diameter (up to 2.8 MDa) with 120 copies of two different protein subunits held together by designed interfaces (33). Some of these cages are notable in that, for the first time, they constitute a tightly packed wall separating the inner cavity of the cage from the outside. This compares to previous cages, which had a significant number of large “holes.” In this respect some of the newest cages more closely resemble virus capsids. In our research with TRAP (trp RNA-binding attenuator protein, Figure 1) we are beginning to develop a third approach in which an inorganic nanoparticle is used to promote interactions between protein rings to form stable protein cages with unusual properties. Although still ongoing, this research suggests a potentially much more limited role for protein-protein interfaces and appears to result in highly stable cage proteins. Our method utilizes metal particles and indeed engineering of proteins so that they interact with metals to assemble into spectacular higher order structures such as crystalline arrays (34, 35) and nanotubes (36) has already been shown as well as a re-engineered ferritin whose ability to form a cage is inducible by copper (37).

Figure 1. Crystal structure of TRAP from G. stearothermophilus (pdb 1qaw (40)). Structure is shown in cartoon format with each monomer numbered individually. Residues discussed in the text are shown as spheres and consist of those involved in nanotube formation, i.e., residues E50 and V69, colored black and dark grey respectively and residue K35 shown in white. Curled arrows indicate a rotation of 90 degrees.

6

TRAP (38, 39) is a toroidal protein found in species of Bacillus and related species. The structure and biochemistry of the protein has been well studied and include a number of crystal structures of the tryptophan-liganded protein alone (40, 41) and in complex with RNA (42–44). It is a homo-11mer with an overall diameter of approximately 8 nm and a central hole of approximately 2 nm in diameter (40, 41) (Figure 1). The outer rim of the ring contains numerous charged residues for binding of a specific sequence of mRNA, including a prominent lysine. Each TRAP monomer can also bind tryptophan in a specific binding pocket (40, 45). In vivo the role of the protein is in the control of tryptophan synthesis: A number of enzymes are involved in the synthetic pathway, primarily located in the trp operon (38) and TRAP regulates their transcription by, when liganded with tryptophan, binding to the trp leader sequence (specifically NAG repeats). This leads to formation of a terminator hairpin and transcription termination (38) (though recent work suggests the terminator hairpin may not be necessary) (46). Binding of tryptophan-liganded TRAP also regulates the translation of trpE by promoting formation of a RNA secondary structure that sequesters the ShineDalgarno sequence, thus disfavoring ribosome binding (38). In the absence of tryptophan, TRAP does not bind to the RNA sequences and tryptophan synthesis is allowed to proceed. Our own work with TRAP has included a number of biochemical and structural studies which can provide useful information and inspiration for subsequent engineering of the protein to make novel, artificial structures: We showed that TRAP from different species demonstrated remarkable differences in cooperativity of ligand (tryptophan) binding (47) and we also succeeded in solving the crystal structure of apo-TRAP (48), i.e., a TRAP ring free of bound tryptophan, something which had proved difficult, presumably due to increased flexibility (49) in the tryptophan free form. The resulting structure was interesting as it showed little difference from the liganded protein, suggesting a control of RNA-binding via dynamic rather than major structural changes and we also showed that the effect of ligand binding could be mimicked by low temperature. In other structural work we solved the structure of TRAP in complex with its inhibitor protein Anti-TRAP. Anti-TRAP is a small homotrimer (50) and in vivo it offers an additional layer of control over tryptophan production: it binds to TRAP and inhibits binding to RNA, with Anti-TRAP production itself controlled by levels of charged tRNAtrp. The structure of Anti-TRAP in complex with TRAP (51) showed that it binds around the outer rim of the ring, blocking the key residues involved in RNA binding. One of the monomers of the trimer is not involved in interacting with TRAP, the other two monomers are, with each one of them binding to one TRAP monomer. As a TRAP ring consists of 11 monomers, this would result in at best, five equally spaced Anti-TRAPs binding to ten of the monomers, with one gap. Indeed this stoichiometry has been confirmed from analytical ultracentrifugation and mass spectrometry (51). How could such a non-symmetrical shape be easily crystallised? The answer appears to be that it did not: a minority of TRAP exists in solution in a 12-membered form (in fact a naturally-occurring 12-membered form is known in B. Haoldurans (52)) and it was such a 12-membered form, with exactly six anti-TRAPs bound that crystallised. 7

It is in the area of bionanoscience that TRAP has proved particularly interesting. One of the goals of bionanoscience is to design and build novel, artificial structures. As with other areas of synthetic biology, building artificial systems allows fine control, bespoke properties and capabilities beyond those found in nature. In early work we showed that TRAP could be modified with addition of a peptide sequence able to bind titanium (and silicon) surfaces (53) and with cysteine in the central hole for binding to gold. This resulted in a TRAP protein able to capture gold particles and localize them on a silicon surface for subsequent integration into a prototype MOS capacitor (54). We have also been able to engineer TRAP so that it forms rings with altered symmetry (55) (“symmetry engineering”) and protein nanotubes (56) as well as hollow cages (57, 58). In the early symmetry engineering work we converted wild type TRAP protein from G. stearothermophilus from its natural 11-fold rotational symmetry form to one with the equivalent of 12 rather than 11 monomers (i.e., 12-fold rotational symmetry) this was achieved by producing tandem gene fusions of the TRAP proteins connected by unstructured amino acid linkers. When three or four such monomers were fused together they were unable to form the equivalent of an 11-membered ring (11 being a prime number). Instead the closest equivalent was 3×4mers or 4×3mers, which assembled to form the equivalent of a 12-monomer ring. These structures were confirmed by X-ray crystallography (55). Artificial protein nanotubes are another area of interest (59, 60) and nanotubes in general, including carbon nanotubes, peptide nanotubes and lipid nanotubes may be useful in electronics, and for delivery of materials (such as therapeutics) that are filled into their internal cavity (59, 61). Protein rings lend themselves to nanotube construction by virtue of the fact that a tube can simply be formed by the stacking of rings on top of each other, with the hole in the ring acting as the central cavity. We stacked TRAP rings in this way as outlined in Figure 2: One residue per monomer on the surface of each of the two flat faces of the ring was changed to a cysteine, resulting in 11 cysteines per face per ring (TRAP has no other cysteine residues). The residues were chosen so that the resulting cysteines on opposing faces would each be approximately the same radial distance from the center of the ring. It was originally envisaged that this would allow cysteines to align and form disulfide bonds: A similar approach was previously used by Ballister et al. utilizing the ring protein hcp (62). We did find that our modified TRAP proteins could form nanotubes with the overall shape resembling bamboo due to the fact that one end of TRAP is narrower then the other and the interface between TRAP rings was always between like faces (i.e., narrow face-narrow face and wide face-wide face) most likely because the cysteines on non-identical faces do not precisely align. Surprisingly we found that these tubes could only form in the presence of reducing agents such as dithiothreitol (DTT) or dimercaptopropanol (DMP) (Figure 2). This was at first counterintuitive as DTT, in particular is used in protein biochemistry as a reducing agent to specifically ensure that disulfide bonds do not form. However, DTT (and DMP) are essentially short carbon chains containing two –SH groups, one at either end of the chain. It is known that adducts can be formed between these –SH groups and those in cysteine side chains and we proposed that in the case of TRAP rings this was occurring at both SH groups of the 8

reducing agent, essentially allowing them to act as cross linkers to link together TRAP rings into the tube form (56). More recent work (63) has allowed us to confirm the details of ring formation and structure in more detail and showed that the wide face-wide face interface absolutely requires the presence of a cross linker due to the fact that other nearby bulky groups prevent the opposing cysteines from approaching close enough for direct disulfide bonds to form. This is not the case at the narrow face-narrow face interface where direct disulfides can form and in fact cysteines can be removed altogether and tubes can still result as long as the cysteines are replaced with residues that promote protein-protein interactions. The protein nanotube work highlights the need for careful use of structural information in designing protein assemblies, particularly at the protein-protein interface.

Figure 2. Formation of a nanotube from modified TRAP. a) Shows crystal structures of four TRAP rings (pdb 1qaw (40)) with residues E50 and V69 shown as black and dark grey spheres respectively. Rings are arranged with like faces opposing each other and double-headed arrows represent interactions between residues of opposing faces. b) Shows molecules of DTT and DMP which are known to be able to promote tube formation by acting as cross-linkers between cysteine groups at position 69 and possibly, at position 50.

We have also used TRAP to build protein cages: This work used a novel method of cage production which, while still under investigation, appears to utilize gold nanoparticles in promoting interactions between rings, resulting in formation of highly stable “spheres”. The discovery of TRAP-cage, complete with low resolution structural data was first reported in 2012 (57): It is constructed from a mutant of TRAP containing cysteine at position 35 in place of the wild type lysine, with the resulting mutant being referred to as TRAP-K35C (Figure 1). We found that upon incubation with 1.4 nm diameter gold nanoparticles (GNP), TRAP-K35C assembled into structures which, under transmission electron microscopy (TEM) resemble spherical virus capsids approximately 21 nm in diameter (Figure 3). The spherical and hollow nature of the assembly was confirmed using cryo-electron tomography (Figure 3). Furthermore if the concentration of GNP was increased, a second, smaller diameter (~15 nm) cage was observed. 9

Figure 3. a) Cartoon showing that TRAP rings (left) in the presence of gold nanoparticles (“Au”) form large cage like “spheres” (right). b) Shows a cryo-EM image of individual cages in vitreous ice. Scale bars in a and b = 40 nm. c) shows two views of five reconstructed individual TRAP-cages. d) Shows tomographic subvolumes of the particles shown in b, demonstrating their hollow nature. e) Shows AFM image of TRAP-cages. f) Shows the AFM height profile taken along the dotted line in d. Panels b-f are reproduced with permission from ref. (57). Copyright 2002 American Chemical Society.

10

Figure 4. Analysis of TRAP-cage structure using AFM. ai) Closely packed TRAP-cages (image area 100 nm x 100 nm). aii) Histogram showing height of 65 TRAP-cages, mean = 19.5 nm ± 0.9. bi) unprocessed high speed AFM image of TRAP-cage; bii) 3D reconstruction image of Gauss filtered (3 Å~ 3 pixel) image of bi) with discernible rings numbered; biii) same image as bi) after the application of Gauss filter (3 Å~ 3 pixel) and high-pass FFT frequency filter (10 nm); biv) same image as biii) with a number assigned to each identifiable TRAP ring. Image area 40 nm x 25 nm. Brightness scales on the left show height information in nanometers. Figure is reproduced with permission from ref. (57). Copyright 2002 American Chemical Society

11

We have subsequently focused research efforts on understanding the larger cage. Intriguingly a convex polyhedron cannot be formed from an 11-sided shape so we initially assumed that the TRAP-cage was formed from TRAP subunits rather than TRAP rings. However, atomic force microscopy (AFM) analysis seemed to suggest that the cage was formed from in tact rings, indicating unusual geometry (Figure 4) (58). It was hypothesized that the cages were held together by disulfide bonds between constituent rings, possibly catalyzed by the presence of the gold nanoparticles which are known to be oxidatively catalytically active (64). The presence of some form of reducible bond (such as a disulfide bond) gained support from spectacular high-speed AFM images which showed the “bursting” of TRAP-cages upon addition of reducing agent (58). These results are interesting – as the interiors of cells are reducing, a stable, inducible cage protein could be envisaged as a potentially useful cell delivery system. Loading of the cage could potentially be achieved simply by cage formation being “switched on” in the presence of cargo molecules (e.g., therapeutics) by the addition of gold particles. These results have raised a number of interesting questions. Firstly the role of gold nanoparticles is intriguing. It was in the 1980s that the discovery was made that small GNPs were highly chemically active (64) a surprising finding given the well-known noble nature of the metal. Initial reactions observed were the catalysis of the oxidation of carbon monoxide to carbon dioxide. The ability of these small gold particles to act as catalysts is now well established although the exact mechanism whereby this is achieved remains the subject of some debate (65). The ability of GNPs to interact with proteins had previously been noted and as the thioaurate bond is well known it is no surprise that some form of interaction with a cysteine residue at the easily accessible position 35 on TRAP could occur. However, typically interactions between GNPs and proteins are non–specific and result in simple aggregation of the protein or formation of a protein “corona” around the GNP (66). The TRAP-cage work represented, to our knowledge, the first example of GNP interaction with proteins in a highly specific manner, resulting in production of a limited variety of highly ordered products with an apparent high yield. The size, geometry and ligand on the gold particle may all be contributing factors to this unique reaction and are currently under investigation. The requirement of the gold nanoparticle in TRAP-cage assembly also offers a challenge to future research and development: As gold particles are not required as part of the final cage structure, then any added particles are likely to be present dispersed throughout the solution or non-specifically interacting with the TRAPcage. Indeed AFM work seems to suggest a preference for binding of the GNPs to the central hole of the TRAP ring presumably through electrostatic interactions (58). For high-resolution X-ray crystal studies of TRAP-cage this could be an obstacle to crystallization of the protein. Furthermore if the TRAP-cage was to be developed therapeutically the presence of these gold particles could be problematic as small (~1.4 nm) GNPs have been shown to be cytotoxic (67, 68). Therefore a way of utilizing gold particles in a different phase from the in-solution TRAP-cage would be advantageous and current experiments are directed towards using GNPs that are tethered by a linker molecules to a surface, over which a solution of cage12

forming TRAP is flowed. It is envisaged that this will allow cage to form and be collected free of GNPs which remain tethered to the surface. Overall, TRAP has proved a very versatile protein building block with the ability to interact with GNPs to form unusual cage structures being the latest manifestation. A deeper understanding of the TRAP-cage assembly mechanism and structure may allow us to advance capabilities for new protein chemistries and assemblies.

Acknowledgments The author was funded by the Malopolska Centre of Biotechnology and by the National Science Centre (NCN, Poland) grant No. 2016/20/W/NZ1/00095 (Symfonia-4).

References 1.

Vannucci, L.; Lai, M.; Chiuppesi, F.; Ceccherini-Nelli, L.; Pistello, M. Viral vectors: a look back and ahead on gene transfer technology. New Microbiol. 2013, 36, 1–22. 2. Uchida, M.; Klem, M. T.; Allen, M.; Suci, P.; Flenniken, M.; Gillitzer, E.; Varpness, Z.; Liepold, L. O.; Young, M.; Douglas, T. Biological Containers: Protein Cages as Multifunctional Nanoplatforms. Adv. Mater. 2007, 19, 1025–1042. 3. Cardinale, D.; Carette, N.; Michon, T. Virus scaffolds as enzyme nano-carriers. Trends Biotechnol. 2012, 30, 369–376. 4. Kawano, M.; Xing, L.; Lam, K. S.; Handa, H.; Miyamura, T.; Barnett, S.; Srivastava, I. K.; Cheng, R. H., Design Platforms of Nanocapsules for Human Therapeutics or Vaccines. In Development of Vaccines: From Discovery to Clinical Testing; John Wiley and Sons: 2011; pp 125−139. 5. Yildiz, I.; Shukla, S.; Steinmetz, N. F. Applications of viral nanoparticles in medicine. Curr. Opin. Biotechnol. 2011, 22, 901–908. 6. Wen, A. M.; Steinmetz, N. F. Design of virus-based nanomaterials for medicine, biotechnology, and energy. Chem. Soc. Rev. 2016, 45, 4074–4126. 7. Steinmetz, N. F.; Cho, C. F.; Ablack, A.; Lewis, J. D.; Manchester, M. Cowpea mosaic virus nanoparticles target surface vimentin on cancer cells. Nanomedicine 2011, 6, 351–364. 8. Aljabali, A. A.; Shukla, S.; Lomonossoff, G. P.; Steinmetz, N. F.; Evans, D. J. CPMV-DOX delivers. Mol. Pharm. 2013, 10, 3–10. 9. Yildiz, I.; Lee, K. L.; Chen, K.; Shukla, S.; Steinmetz, N. F. Infusion of imaging and therapeutic molecules into the plant virus-based carrier cowpea mosaic virus: cargo-loading and delivery. J. Controlled Release 2013, 172, 568–578. 10. Lizotte, P. H.; Wen, A. M.; Sheen, M. R.; Fields, J.; Rojanasopondist, P.; Steinmetz, N. F.; Fiering, S. In situ vaccination with cowpea mosaic virus 13

11.

12.

13.

14.

15.

16. 17.

18.

19.

20.

21. 22. 23. 24. 25. 26.

nanoparticles suppresses metastatic cancer. Nat. Nanotechol. 2016, 11, 295–303. Patterson, D. P.; McCoy, K.; Fijen, C.; Douglas, T. Constructing catalytic antimicrobial nanoparticles by encapsulation of hydrogen peroxide producing enzyme inside the P22 VLP. J. Mater. Chem. B 2014, 2, 5948–5951. Patterson, D. P.; Schwarz, B.; Waters, R. S.; Gedeon, T.; Douglas, T. Encapsulation of an enzyme cascade within the bacteriophage P22 virus-like particle. ACS Chem. Biol. 2014, 9, 359–365. Yamashita, I.; Hayashi, J.; Hara, M. Bio-template Synthesis of Uniform CdSe Nanoparticles Using Cage-shaped Protein, Apoferritin. Chem. Lett. 2004, 33, 1158–1159. Mayes, E.; Bewick, A.; Gleeson, D.; Hoinville, J.; Jones, R.; Kasyutich, O.; Nartowski, A.; Warne, B.; Wiggins, J.; Wong, K. K. W. Biologically derived nanomagnets in self-organized patterned media. IEEE Trans. Magn. 2003, 39, 624–627. Klem, M. T.; Mosolf, J.; Young, M.; Douglas, T. Photochemical mineralization of europium, titanium, and iron oxyhydroxide nanoparticles in the ferritin protein cage. Inorg. Chem. 2008, 47, 2237–2239. Yamashita, I.; Iwahori, K.; Kumagai, S. Ferritin in the field of nanodevices. Biochim. Biophys. Acta, Gen. Subj. 2010, 1800, 846–857. Zhen, Z.; Tang, W.; Chen, H.; Lin, X.; Todd, T.; Wang, G.; Cowger, T.; Chen, X.; Xie, J. RGD-modified apoferritin nanoparticles for efficient drug delivery to tumors. ACS Nano 2013, 7, 4830–4837. Uchida, M.; Flenniken, M. L.; Allen, M.; Willits, D. A.; Crowley, B. E.; Brumfield, S.; Willis, A. F.; Jackiw, L.; Jutila, M.; Young, M. J.; Douglas, T. Targeting of cancer cells with ferrimagnetic ferritin cage nanoparticles. J. Am. Chem. Soc. 2006, 128, 16626–16633. Seebeck, F. P.; Woycechowsky, K. J.; Zhuang, W.; Rabe, J. P.; Hilvert, D. A simple tagging system for protein encapsulation. J. Am. Chem. Soc. 2006, 128, 4516–4517. Wörsdörfer, B.; Pianowski, Z.; Hilvert, D. Efficient in Vitro Encapsulation of Protein Cargo by an Engineered Protein Container. J. Am. Chem. Soc. 2012, 134, 909–911. Worsdorfer, B.; Woycechowsky, K. J.; Hilvert, D. Directed evolution of a protein container. Science 2011, 331, 589–592. Frey, R.; Hayashi, T.; Hilvert, D. Enzyme-mediated polymerization inside engineered protein cages. Chem. Commun. 2016, 52, 10423–10426. He, D.; Marles-Wright, J. Ferritin family proteins and their use in bionanotechnology. New Biotechnol. 2015, 32, 651–657. Levinthal, C. Are there pathways for protein folding? J. Med. Phys. 1968, 65, 44–45. Wolynes, P. G. Evolution, energy landscapes and the paradoxes of protein folding. Biochimie 2015, 119, 218–230. Khoury, G. A.; Smadbeck, J.; Kieslich, C. A.; Floudas, C. A. Protein folding and de novo protein design for biotechnological applications. Trends Biotechnol. 2014, 32, 99–109. 14

27. Fletcher, J. M.; Harniman, R. L.; Barnes, F. R.; Boyle, A. L.; Collins, A.; Mantell, J.; Sharp, T. H.; Antognozzi, M.; Booth, P. J.; Linden, N.; Miles, M. J.; Sessions, R. B.; Verkade, P.; Woolfson, D. N. Self-assembling cages from coiled-coil peptide modules. Science 2013, 340, 595–599. 28. Padilla, J. E.; Colovos, C.; Yeates, T. O. Nanohedra: Using symmetry to design self assembling protein cages, layers, crystals, and filaments. Proc. Natl Acad. Sci. U.S.A. 2001, 98, 2217–2221. 29. Lai, Y. T.; Cascio, D.; Yeates, T. O. Structure of a 16-nm cage designed by using protein oligomers. Science 2012, 336, 1129. 30. King, N. P.; Sheffler, W.; Sawaya, M. R.; Vollmar, B. S.; Sumida, J. P.; Andre, I.; Gonen, T.; Yeates, T. O.; Baker, D. Computational design of selfassembling protein nanomaterials with atomic level accuracy. Science 2012, 336, 1171–1174. 31. King, N. P.; Bale, J. B.; Sheffler, W.; McNamara, D. E.; Gonen, S.; Gonen, T.; Yeates, T. O.; Baker, D. Accurate design of co-assembling multi-component protein nanomaterials. Nature 2014, 510, 103–108. 32. Hsia, Y.; Bale, J. B.; Gonen, S.; Shi, D.; Sheffler, W.; Fong, K. K.; Nattermann, U.; Xu, C.; Huang, P. S.; Ravichandran, R.; Yi, S.; Davis, T. N.; Gonen, T.; King, N. P.; Baker, D. Design of a hyperstable 60-subunit protein icosahedron. Nature 2016, 535, 136–139. 33. Bale, J. B.; Gonen, S.; Liu, Y.; Sheffler, W.; Ellis, D.; Thomas, C.; Cascio, D.; Yeates, T. O.; Gonen, T.; King, N. P.; Baker, D. Accurate design of megadalton-scale two-component icosahedral protein complexes. Science 2016, 353, 389–394. 34. Suzuki, Y.; Cardone, G.; Restrepo, D.; Zavattieri, P. D.; Baker, T. S.; Tezcan, F. A. Self-assembly of coherently dynamic, auxetic, two-dimensional protein crystals. Nature 2016, 533, 369–373. 35. Brodin, J. D.; Carr, J. R.; Sontz, P. A.; Tezcan, F. A. Exceptionally stable, redox-active supramolecular protein assemblies with emergent properties. Proc. Natl Acad. Sci. U.S.A. 2014, 111, 2897–2902. 36. Brodin, J. D.; Ambroggio, X.; Tang, C.; Parent, K. N.; Baker, T. S.; Tezcan, F. A. Metal-directed, chemically tunable assembly of one-, two-and three-dimensional crystalline protein arrays. Nat. Chem. 2012, 4, 375–382. 37. Huard, D. J.; Kane, K. M.; Tezcan, F. A. Re-engineering protein interfaces yields copper-inducible ferritin cage assembly. Nat. Chem. Biol. 2013, 9, 169–176. 38. Gollnick, P.; Babitzke, P.; Antson, A.; Yanofsky, C. Complexity in regulation of tryptophan biosynthesis in Bacillus subtilis. Annu. Rev. Genet. 2005, 39, 47–68. 39. Yanofsky, C. RNA-based regulation of genes of tryptophan synthesis and degradation, in bacteria. RNA 2007, 13, 1141–1154. 40. Chen, X.; Antson, A. A.; Yang, M.; Li, P.; Baumann, C.; Dodson, E. J.; Dodson, G. G.; Gollnick, P. Regulatory features of the trp operon and the crystal structure of the trp RNA-binding attenuation protein from Bacillus stearothermophilus. J. Mol. Biol. 1999, 289, 1003–1016.

15

41. Antson, A. A.; Otridge, J.; Brzozowski, A. M.; Dodson, E. J.; Dodson, G. G.; Wilson, K. S.; Smith, T. M.; Yang, M.; Kurecki, T.; Gollnick, P. The structure of trp RNA-binding attenuation protein. Nature 1995, 374, 693–700. 42. Antson, A. A.; Dodson, E. J.; Dodson, G.; Greaves, R. B.; Chen, X.; Gollnick, P. Structure of the trp RNA-binding attenuation protein, TRAP, bound to RNA. Nature 1999, 401, 235–242. 43. Hopcroft, N. H.; Manfredo, A.; Wendt, A. L.; Brzozowski, A. M.; Gollnick, P.; Antson, A. A. The interaction of RNA with TRAP: the role of triplet repeats and separating spacer nucleotides. J. Mol. Biol. 2004, 338, 43–53. 44. Hopcroft, N. H.; Wendt, A. L.; Gollnick, P.; Antson, A. A. Specificity of TRAP-RNA interactions: crystal structures of two complexes with different RNA sequences. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2002, 58, 615–621. 45. Otridge, J.; Gollnick, P. MtrB from Bacillus subtilis binds specifically to trp leader RNA in a tryptophan-dependent manner. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 128–132. 46. McAdams, N. M.; Gollnick, P. The Bacillus subtilis TRAP protein can induce transcription termination in the leader region of the tryptophan biosynthetic (trp) operon independent of the trp attenuator RNA. PLoS One 2014, 9, e88097. 47. Heddle, J. G.; Okajima, T.; Scott, D. J.; Akashi, S.; Park, S. Y.; Tame, J. R. Dynamic allostery in the ring protein TRAP. J. Mol. Biol. 2007, 371, 154–167. 48. Malay, A. D.; Watanabe, M.; Heddle, J. G.; Tame, J. R. H. Crystal structure of unliganded TRAP: implications for dynamic allostery. Biochem. J. 2011, 434, 429–434. 49. McElroy, C.; Manfredo, A.; Wendt, A.; Gollnick, P.; Foster, M. TROSYNMR studies of the 91kDa TRAP protein reveal allosteric control of a gene regulatory protein by ligand-altered flexibility. J. Mol. Biol. 2002, 323, 463–473. 50. Shevtsov, M. B.; Chen, Y.; Gollnick, P.; Antson, A. A. Crystal structure of Bacillus subtilis anti-TRAP protein, an antagonist of TRAP/RNA interaction. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 17600–17605. 51. Watanabe, M.; Heddle, J. G.; Kikuchi, K.; Unzai, S.; Akashi, S.; Park, S. Y.; Tame, J. R. The nature of the TRAP-Anti-TRAP complex. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 2176–2181. 52. Chen, C. S.; Smits, C.; Dodson, G. G.; Shevtsov, M. B.; Merlino, N.; Gollnick, P.; Antson, A. A. How to change the oligomeric state of a circular protein assembly: switch from 11-subunit to 12-subunit TRAP suggests a general mechanism. PLoS One 2011, 6, e25296. 53. Sano, K.; Shiba, K. A hexapeptide motif that electrostatically binds to the surface of titanium. J. Am. Chem. Soc. 2003, 125, 14234–14235. 54. Heddle, J. G.; Fujiwara, I.; Yamadaki, H.; Yoshii, S.; Nishio, K.; Addy, C.; Yamashita, I.; Tame, J. R. Using the ring-shaped protein TRAP to capture and confine gold nanodots on a surface. Small 2007, 3, 1950–1956. 16

55. Heddle, J. G.; Yokoyama, T.; Yamashita, I.; Park, S. Y.; Tame, J. R. Rounding up: Engineering 12-membered rings from the cyclic 11-mer TRAP. Structure 2006, 14, 925–933. 56. Miranda, F. F.; Iwasaki, K.; Akashi, S.; Sumitomo, K.; Kobayashi, M.; Yamashita, I.; Tame, J. R. H.; Heddle, J. G. A Self-Assembled Protein Nanotube with High Aspect Ratio. Small 2009, 5, 2077–2084. 57. Malay, A. D.; Heddle, J. G.; Tomita, S.; Iwasaki, K.; Miyazaki, N.; Sumitomo, K.; Yanagi, H.; Yamashita, I.; Uraoka, Y. Gold NanoparticleInduced Formation of Artificial Protein Capsids. Nano Lett. 2012, 12, 2056–2059. 58. Imamura, M.; Uchihashi, T.; Ando, T.; Leifert, A.; Simon, U.; Malay, A. D.; Heddle, J. G. Probing structural dynamics of an artificial protein cage using high-speed atomic force microscopy. Nano Lett. 2015, 15, 1331–1335. 59. Heddle, J. G. Protein cages, rings and tubes: useful components of future nanodevices? Nanotechnol., Sci. Appl. 2008, 1, 67–78. 60. Heddle, J. G.; Tame, J. R. H. Protein nanotubes, channels and cages. In Amino Acids, Peptides and Proteins; Farkas, E., Ryadnov, M., Eds.; The Royal Society of Chemistry: Cambridge, 2012; Vol. 37, pp 151−189. 61. Mundra, R. V.; Wu, X.; Sauer, J.; Dordick, J. S.; Kane, R. S. Nanotubes in biological applications. Curr. Opin. Biotechnol. 2014, 28, 25–32. 62. Ballister, E. R.; Lai, A. H.; Zuckermann, R. N.; Cheng, Y.; Mougous, J. D. In vitro self-assembly of tailorable nanotubes from a simple protein building block. Proc. Natl Acad. Sci. U.S.A. 2008, 105, 3733–3738. 63. Nagano, S.; Banwell, E. F.; Iwasaki, K.; Michalak, M.; Pałka, R.; Zhang, K. Y.; Voet, A. R.; Heddle, J. G. Understanding the Assembly of an Artificial Protein Nanotube. Adv. Mater. Interfaces 2016, 3. 64. Haruta, M.; Kobayashi, T.; Sano, H.; Yamada, N. Novel gold catalysts for the oxidation of carbon monoxide at a temperature far below 0 °C. Chem. Lett. 1987, 16, 405–408. 65. Stratakis, M.; Garcia, H. Catalysis by supported gold nanoparticles: beyond aerobic oxidative processes. Chem. Rev. 2012, 112, 4469–4506. 66. Cedervall, T.; Lynch, I.; Lindman, S.; Berggard, T.; Thulin, E.; Nilsson, H.; Dawson, K.; Linse, S. Understanding the nanoparticle-protein corona using methods to quntify exchange rates and affinities of proteins for nanoparticles. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 2050–2055. 67. Tsoli, M.; Kuhn, H.; Brandau, W.; Esche, H.; Schmid, G. Cellular uptake and toxicity of Au55 clusters. Small 2005, 1, 841–844. 68. Pan, Y.; Neuss, S.; Leifert, A.; Fischler, M.; Wen, F.; Simon, U.; Schmid, G.; Brandau, W.; Jahnen-Dechent, W. Size-Dependent Cytotoxicity of Gold Nanoparticles. Small 2007, 3, 1941–1949.

17

Chapter 2

Integrated Nanostructures Based on Self-Assembled Amphiphilic Polypeptides Motoki Ueda,1,2,* Stefan Müller,2 Siyoong Seo,1,2 Md. Mofizur Rahman,2 and Yoshihiro Ito1,2 1Nano

Medical Engineering Laboratory, RIKEN, 2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan 2Emergent Bioengineering Materials Research Team, RIKEN Center for Emergent Matter Science, 2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan *E-mail: [email protected]

Nano-ordered materials are requiring on demand shape control in various fields; biomaterials, medicals, pharmaceuticals and so on. Self-assembly of molecules is a typical bottom-up phenomenon observed in nature to form nano-materials. Self-assembly leads to the formation of various structures that impart functions. We found that amphiphilic block polypeptides having helical structure as a hydrophobic block self-assemble to form different structures. First, the hydrophobic segments form helical structures that define specific structures of self-assemblies such as fibers, tubes, vesicles and sheets depending on the diameter and length of helix. Small diameter of 310-helix induced fibers and micelles reasonably due to small critical packing parameter. On the other hand, sheet-like assembly were obtained by peptide having large sectional area of α-helix. Depending on the length of hydrophobic helix, these sheets transformed into various morphology by heat treatment. Length of helix changes the thickness and size of sheet and thus, changes the flexibility of the sheet, resulting that the difference of just two residues induced drastically different shapes of assembly, tube, vesicle and sheet. Next, hydrophilic polymers, polyethylene glycol (PEG) and poly(sarcosine), didn’t affect drastically the shape but did the size of assembly owing to the difference of hydrophilicity and flexibility. Poly(sarcosine) made even small assembly of ~200 nm length and 80 nm © 2017 American Chemical Society

diameter more stable and PEG enhanced the elongation of tubular assembly (more than 1 μm length) with the same diameter of 80 nm. Furthermore, we demonstrate some integration of the structures by mixing with other components such as lipids, cholesterol, or amphiphilic polypeptides having a different hydrophilic chain of polyethylene glycol or a different-length hydrophobic helix. This work shows one of the methods to prepare complex morphology of molecular assembly.

Introduction To prepare molecular assemblies with control over the shape formed, numerous different amphiphiles have been developed. These amphiphiles can be divided into particular groups: lipid-like surfactants (1–3), peptide amphiphiles (4–6) and amphiphilic polymers (7, 8), and other amphiphiles (9). Like nature, some research groups reported phase-separated assemblies by optimization of lipid components and preparation conditions. For example, using the phase-separation property of lipid rafts, a large lateral separation into two domains of cholesterol/sphingomyelin/saturated phospholipids and unsaturated phospholipids in giant unilamellar vesicles was demonstrated to generate Janus vesicles of a conjugate morphology of two vesicles (1–3). This Janus assembly is artificial but is the closest mimic of a natural plasma membrane because this vesicle is composed of a combination of natural components and controlled by only optimizing the mixing ratio and the preparation conditions. Thus, this vesicle is an excellent mimic and model of plasma membranes but is difficult to use in applications owing to instability issues and the inability to control size and shape. However, most of the lipid and lipid-like assemblies are not stable, which limits their applications. In contrast, amphiphilic polymers produce stable molecular assemblies (10) and form various shapes besides the commonly formed sphere, such as tubular structures, mesoporous spheres and sheets. Another instance of a conjugate morphology of self-assemblies was reported where one end of a nanotube was capped with a vesicle to generate a round-bottom flask-type morphology, which was prepared from the mixture of phospholipids and amphiphilic molecules with two hydrophobic legs (11, 12). Here, we focus on polypeptide amphiphiles, which assembly to form stable complexes owing to numerous inter- and intra-interactions, including van der Waals forces, H-bonding, π-π stacking, dipole moments, disulfide bonding, electrostatic interactions, concavo-convexo and steric interactions. The polypeptide assemblies show lipid-like membrane fluidity, morphological diversity and biocompatibility. In addition, it has recently been shown that polypeptide assemblies can create a Janus soft matter consisting of a phase-separated membrane by integration. Using these unique properties, attempts were made to use these polypeptide assemblies for biomedical applications, including bio-matrices, carriers of drug delivery systems and in vivo imaging. 20

In this chapter, as shown in Figure 1, we show the assemblies of amphiphilic polypeptides and derivatives with polyethylene glycol (PEG), and composite formation by integration with other components such as lipids or cholesterol.

Figure 1. Amphiphilic polypeptide assemblies with other components.

Amphiphilic Polypeptide Assemblies Ueda and coworkers have reported some unique morphologies prepared from polypeptide amphiphiles (13–18). These amphiphiles have an α-helical forming peptide as a hydrophobic block. An α-helical hydrophobic block is rigid and uniform shape due to the intramolecular hydrogen bonding, resulting that it can enable to form stable assemblies, have lateral fluidity without entangling between hydrophobic blocks and make it easy to control the shape of assembly. Hydrophilic block is composed of poly(sarcosine). Poly(sarcosine) forms polymer brush on the surface of assembly and shows the similar hydrophilicity and stealth ability to polyethylene glycol. In addition, poly(sarcosine) is biodegradable and not toxic in a body, resulting that poly(sarcosine) is very useful tool as a biomaterial for medical application. Furthermore, in previous reports, he also reported that a mixture of two enantiomeric amphiphilic polypeptides with a right- and a lefthanded helical block self-assemble to form large planar and curved sheets. The 21

sheets could be converted to a round-bottom flask-type molecular assembly upon heating in buffer, where the neck part was composed of a single component and the sphere part composed of the stereocomplex (14, 15).

Length of the Hydrophobic Chain

The morphologies of assemblies can be controlled by the design of amphiphilic polypeptides. The design refers to the character of the hydrophilic block and the hydrophobic block and the valence of each block type. In particular, the hydrophobic block is one of the most important blocks to control the shape because the molecular orientation of packed amphiphiles depends on the character of the hydrophobic block. Here, we synthesized some amphiphilic polypeptides with the hydrophilic polypeptide polysarcosine (Sar)n and various lengths of a hydrophobic helix block composed of L-leucine (L-Leu) and α-aminoisobutyryl (Aib); (Sar)16-b-(L-Leu-Aib)4 (L8), (Sar)25-b-(L-Leu-Aib)6 (L12), (Sar)26-b-(LLeu-Aib)7 (L14), (Sar)32-b-(L-Leu-Aib)8 (L16), (Sar)40-b-(L-Leu-Aib)9 (L18) and (Sar)48-b-(L-Leu-Aib)10 (L20). According to the Griffin hydrophilic-lipophilic balance (HLB) (19–21), these polypeptides have similar values of 12.0–13.0. There is no significant difference among these hydrophilic-lipophilic balances. However, these polypeptides assembly to form different shapes in aqueous solvent. These assemblies were categorized into two groups. One group is the fiber-shaped assembly of L8 (Figure 2a) and the other group is the sheet-shaped assembly of L12–L20. Circular dichroism spectra demonstrated that the hydrophobic block of L8 formed a 310-helix and others forming an α-helix (Figure 2e). The packing of α-helices is known to form planer sheets because the α-helix has a thicker diameter than the 310-helix. Thus, the critical packing parameter (22) of L8 was different from the other amphiphiles. The reason why L8 formed a fiber-shaped assembly is due to the difference of the helix structure. We also found another difference among the assemblies of L12–L20. When these sheet assemblies were heated at 90 °C, L12 and L14 sheets rolled up to form a tubular assembly (Figure 2b) (18, 23), the L16 sheet transformed into a vesicular assembly (Figure 2c) (14), and L18 and L20 sheets maintained their shapes (Figure 2d). The edge of sheets is considered to be hydrophobic, and sheets tend to roll up to decrease the area of the edge. L12–L16 species were considered to not have sufficient lengths of hydrophilic chains to maintain the planar sheet shape during heating. On the other hand, L18 and L20 have sufficient lengths of the hydrophilic chain to cover the hydrophobic edge of the sheet. As a result, L18 and L20 sheets were stable and did not change shape during heating. The formations of tubes of L12 and L14, and the sphere formed by L16 were most likely due to the difference of concavo-convexo interactions between neighboring helices. The anisotropic nature of L12 and L14 membranes facilitated tube formation. However, L16 was too long to be packed with a tilt angle; thus, L16 packed in a parallel manner to form an isotropic sphere. 22

Figure 2. Schematic illustration and negative staining images by a transmission electron microscope (TEM). The assemblies were composed of a single component of an amphiphilic polypeptide with various lengths of the hydrophobic block: (Sar)16–48-b-(L-Leu-Aib)n (Ln), (n = 8, 12, ,14, 16, 18, 20). Amphiphilic polypeptides were dissolved in ethanol and the solutions were injected into water and heated at 90 °C for 10 min. The assemblies were negatively stained by 2% samarium acetate for TEM observation. Scale bar: 500 nm (a–c) and 200 nm (d). CD spectra of assemblies from Ln in milliQ water after heating at 90 °C for 1h (e) Helicity of the Hydrophobic Chain The shape of the assembly depends on both the length and the helicity of the hydrophobic block. Here, an amphiphilic polypeptide incorporated with D-leucine, (Sar)25-b-(D-Leu-Aib)6 (dL12), was synthesized and formed a left-handed helix (Figure 1). As expected, dL12 also formed a nanotube assembly. However, an equimolar mixture of enantiomeric peptides, L12 and dL12, demonstrated another shape, a vesicle of 180 nm diameter after heating at 90 °C (Figure 3a). The enantiomeric helix is known to form a stereo complex that is thermodynamically more stable than a single component. The isotropic packing induces the spherical structure; although, the single amphiphilic components form tube structures. 23

Figure 3. Schematic illustration and TEM images of peptide assemblies prepared from an enantiomeric mixture of L12 and dL12 (a) and a single component of PEG750-b-(L-Leu-Aib)6 (PEG750-L12) (b). Each mixture was dissolved in ethanol and the solutions were injected into water and heated at 90 °C for 60 min. The assemblies were negatively stained by 2% samarium acetate for TEM observation. Scale bar: 500 nm.

Conjugation with Polyethylene Glycol (PEG) The hydrophilic block of the amphiphile also affects the shape of the assembly. Since PEG is a typical hydrophilic polymer, it was used as a hydrophilic block of the amphiphile for peptide assembly. A PEG (Mw: 750 and 2000) chain was modified at the N-terminus of the hydrophobic polypeptide block, (L-Leu-Aib)6, which is completely the same as the hydrophobic block of L12. The PEG-incorporated polypeptides, both PEG750-L12 and PEG2000-L12 formed nanotubular assemblies in saline (Figure 3b). The molecular weight of PEG did not affect the nanotube diameter and the diameter is almost 80 nm which is the same as that of L12 nanotubes. As mentioned above, the orientation between hydrophobic helices determines the curvature of the nanotube. On the other hand, although the length of the L12 nanotube was uniformly ~200 nm (Figure 2b), PEG-incorporated nanotubes were longer and the distributions were broader. Considering the length was similar between polysarcosine and PEG amphiphiles, the chemical differences of the hydrophilic chains is important for covering the hydrophobic edge, which dominates the stability of the assembled structure.

24

Integration of Assembled Polypeptides Mismatch of a Hydrophobic Block Assembly can be integrated by two kinds of peptide membrane. As shown above, the packing of an amphiphile, especially the hydrophobic block, defines the morphology of the assembly. An enantiomeric peptide tends to form a stereo complex, but peptides with the same helicity do not mix to form one membrane. Two kinds of amphiphilic polypeptides, which having the same helicity but the different lengths of the hydrophobic block, are divided into two membranes of the same length of the hydrophobic helix. In other words, these amphiphiles phase separate in one assembly. Actually, a mismatch in the length of the hydrophobic block induces phase separation. L16 is composed of polysarcosine 32mers and a hydrophobic leucinebased 16 residue block. As shown above, L16 self-assembles into spheres with approximately 70–100 nm diameters in buffer after heating the planar sheets. This diameter is similar to the diameter of L12 nanotubes. Therefore, L16 and L12 were mixed to obtain an AB-type conjugate morphology. L16 planar sheets were incubated with L12 nanotubes at a ratio of 1:1 (w/w), and the dispersion was heated at 90 °C for 1 h. As shown in Figure 4a, a nano-test-tube-shaped self-assembly with round bottoms were predominantly observed. The sizes of the neck and round-bottom parts of the assemblies correspond to those of the L12 nanotube and L16 sphere, respectively. Furthermore, the membrane thickness of the spherical part was 6 nm, which was the same as that of the L16 sphere. These results suggest that the nano-test-tube-shaped self-assembly was composed of a peptide membrane where L12 and L16 phases separated to form the corresponding neck part and spherical part of the nano-test-tube morphology.

Combination with Stereo Complex Assembly A combination of a stereo complex membrane prepared from L12 and dL12, and a single component membrane of L12 also formed successfully an integrated assembly that formed a round-bottom flask shape (Figure 4b). As mentioned above, the stereo complex membrane of L12 and dL12 formed a sphere shape of 180 nm diameter. Thus, when the stereo complex membrane was incubated with the L12 nanotube in saline and the mixture heated at 90 °C for 1 h, the stereo complex membrane attached to the open mouth of the L12 nanotube through hydrophobic interactions and then rolled-up to form a vesicle that capped the open mouth. The size of spherical part is corresponding to that of stereo complex vesicle from L12 and dL12.

25

Figure 4. Schematic illustration and TEM images of assembly integrations. Assembly integrations were prepared from a combination of some amphiphilic polypeptides: (a) (Sar)25-b-(L-Leu-Aib)6 (L12) and (Sar)32-b-(L-Leu-Aib)8 (L16), (b) L12 and dL12, (c) ((Sar)26)3-b-((L-His)2-(L-Leu-Aib)6) (HL12) and (Sar)25-b-(D-Leu-Aib)6 (dL12), (d) L12 and phospholipid (DPPC), (e) L12, dL12 and DPPC, and (f) L12 and cholesterol. Each mixture was dissolved in ethanol and the solutions were injected into water and heated at 90 °C for 60 min. The assemblies were negatively stained by 2% samarium acetate for TEM observation. Scale bar: 200 nm. Combination with pH-Dependent Assembly Here, we introduce a stimuli-responsive molecule selectively into one of the parts of the self-assemblies. A pH-responsive polypeptide termed HL12, (poly(sarcosine)26)3-b-((L-His)2-(L-Leu-Aib)6), was prepared (12). Since HL12 contains a histidine (His) dipeptide at the junction between the hydrophilic block and the hydrophobic helical block, it was found that disassembly of either the neck or the round-bottom parts of the Janus-type assemblies under acidic conditions with heat treatment would facilitate self-assemble of HL12 into either part. The integration of HL12 enabled preparation of the round-bottom flask-shaped assemblies containing HL12 selectively either in the neck part (12) or the round-bottom part (Figure 4c). The formation of a stereo complex between the right- and the left-handed helices was the basis for creating round-bottom flask-shaped assemblies with dL12 and HL12. The concavo-convex interaction between neighboring α-helical blocks defined precisely the size and shape of the molecular assemblies. Further, the amphiphilic helical peptides were allowed to diffuse laterally in the assemblies at high temperatures (12). The pH-sensitive molecule, HL12, can be confined selectively into one part of the Janus-type assembly of the round-bottom flask-shaped assembly with a suitable combination of other amphiphilic helical peptides. The part containing HL12 can be selectively disrupted by acidification and heat treatment while leaving the other part of the Janus-type assembly intact. By using a combination of some functional amphiphiles, integration of not only structures but also functions also be succeeded. 26

Combination with Lipid Assembly A lipid assembled liposome also be integrated into a nanostructure assembly. The sealing of a nanotube formed by amphiphilic polypeptides was performed using a liposome. The liposome was generated by mixing dipalmitoyl-phosphatidylcholine (DPPC) and cholesterol at a molar ratio of 55/45. The mixture in ethanol was injected into the L12 nanotube suspension (molar ratio of DPPC/L12, 1/1) in buffer and heated at 90 °C for 1 h. As a result, a round-bottom flask assembly was observed (Figure 4d). The sizes of the diameter and length of the neck part of the round-bottom flask were similar to those of the L12 nanotube. The membrane thickness of the neck part was ca. ~10 nm, which was the same as that of the L12 nanotube, indicating clearly that the neck part was composed of L12. On the other hand, the round bottom part of the round-bottom flask showed a smaller morphology than the liposome, and a membrane thickness of ca. 15 nm, which is obviously larger than that of the liposome, suggesting that the round-bottom shaped membrane is composed of a mixture of DPPC, cholesterol and L12.

ABC-Type Phase-Separated Assemblies As shown some combinations, any amphiphiles form a unique assembly on the edge of L12 nanotube, resulting that complex morphologies are obtained by the assembly integration. By using this assembly integration method, ABC-type dumbbell shaped assemblies were prepared and observed. Two different vesicles and one nanotube were prepared from three types of amphiphilic helical peptides by phase separation. First, round-bottom flask assemblies were prepared from a mixture of L12 and dL12 at a molar ratio of 2:8 and purified using a syringe filter to remove assemblies larger than 450 nm. Subsequently, a DPPC solution was added and the mixture was heated at 90 °C for 1 h. By using this twostep method, ABC-type asymmetric dumbbell-shaped assemblies were obtained. As shown in Figure 4e, the asymmetric dumbbell-shaped morphology had two different spherical parts that were each composed of 180-nm stereoscope vesicles of L12 and dL12 and size-uncontrolled DPPC liposomes, and a middle neck part that was the L12 nanotube with a diameter of 80 nm and a length of 180 nm.

Conjugate with Cholesterol In addition to these shape controls, membrane fluidity of the polypeptide assemblies can be controlled by conjugation with cholesterol. Polypeptide nanotube rigidity was increased and displayed higher stability following insertion of cholesterol into the peptide membrane without changing the shape of the assembly. When the membrane fluidity was evaluated by Laurdan reagent, the fluidity increased linearly as the amount of incorporated cholesterol increased. Controlling fluidity is important for further applications of polypeptide assemblies. 27

Conclusion Using amphiphilic polypeptides, it is possible to prepare different nanostructure shapes, such as fibers, sheets, spheres and tubes. In addition, by summation of these nanostructures more complex shapes were prepared. These complex structures are expected to contribute to the fine regulation of various nanosystems. Nanosystem also become to need on demand precise shape- and size-control to improve the function in near future. Development of bottom-up method to prepare various complex morphology is very important challenge. This work in this chapter is one of their advanced methods, show the possibility of molecular assembly integration and give knowledge to establish complex-shaped nanomaterials for researchers in a various fields.

Experimental Section Materials All amino acids were purchased from Watanabe Chemical Industries Ltd. (Japan). Lipid of DPPC and cholesterol were obtained commercially from NOF Co. (Japan) and Wako Pure Chemical Industry (Japan). mPEG was purchased from Sigma-Aldrich Co. Preparation of Molecular Assembly The amphiphilic peptide (12 mg) was dissolved in ethanol (240 μL) to make a stock solution. An aliquot (10 μL) of the stock solution was injected into a saline (1 mL) with stirring. After stirring for 30 min, this dispersion was heated at 90 °C for 1 h. Molecular assemblies of different compositions were prepared similarly with keeping the total volume of the mixed ethanol solution to be 10 μL. Transmission Electron Microscopy (TEM) TEM images were taken using a JEOL JEM-1230 at an accelerating voltage of 80 kV. Peptide aqueous solutions were applied on a carbon-coated Cu grid, and the samples were negatively stained with 2% samarium acetate, followed by suction of the excess fluid with a filter paper. Synthesis of Amphiphilic Peptides The amphiphilic peptides (L8, L12, L14, L16, L18 and L20) were synthesized by conventional procedures in solution following essentially protocols reported previously (13–18, 23). The hydrophobic helical segments (L-Leu-Aib)m (m = 4, 6, 7, 8, 9, 10) were prepared by fragment condensation, whereas the polysarcosine extension at the N-terminal of the helical segments was obtained by the NCA (N-carboxy anhydride) polymerization. MALDI-TOF MS analysis and the area ratios of Sar N-CH3 peaks against those of OCH3 of the C-terminal in 1H NMR spectra were used to determine the degrees of polymerization of 16–48, 28

respectively. Amphiphilic peptide having hydrophobic right-handed helix, dL12, was synthesized by the same way as previous reports (15–18).

Synthesis of PEG750-L12 Synthetic routes of PEGylated peptide PEG750-L12 was illustrated in Schemes 1. The hydrophobic helical peptide (L-Leu-Aib)6 was synthesized as previously reported (13–18, 23). First, mPEG 1 (Mw = 750) (2.7 g, 3.6 mmol) was dissolved in THF (50 mL) and subsequently, (1.45 g, 7.2 mmol) and trimethylamine (1 mL, 7.2 mmol) were added to the solution. After stirring at r.t. for 4 days, the solution was evaporated and purified through gel filtration column (LH-20) with a mixture of chloroform and methanol (1/1, v/v) as an elution solvent. The solid mPEG-pNP 2 (3.12 g) was obtained. Next, mPEG-pNP 2 (220 mg) and (L-Leu-Aib)6 (100 mg) were dissolved in DMF (4 mL) and then the mixture was stirred at r.t. for 4 hours. After evaporation, the compound was purified through LH-20 column with methanol. The white solid of PEG750-L12 3 was isolated. All intermediates were identified by 1H NMR spectroscopy and were further confirmed by MALDI-TOF MS spectrometry.

Scheme 1. Synthetic scheme of PEG750-L12.

mPEG-pNP 2: 1H NMR (400 MHz, DMSO-d6): δ (ppm) 8.32 (d, 2H, OCCHCHCNO2 of phenyl group), 7.57 (d, 2H, OCCHCHCNO2 of phenyl group), 4.36 (t, 2H, CH3OCH2CH2O), 3.59 (t, 2H, CH3OCH2CH2O), 3.52 (m, 71H, OCH2CH2O, CH3O), 3.24 (t, 4H, OCH2CH2OCOO). MALDI-TOF MS; calcd. for C42H75NO22 [M + Na]+: 968.49, found.: 968.43. PEG750-L12 3: 1H NMR (400 MHz, MeOH-d4): δ (ppm) 8.08–7.72 (m, 12H, amide), 4.86–4.02 (m, 6H, LeuCαH), 3.75–3.61 (m, 68H, OCH2CH2O), 3.53 (s, 3H, OCH3), 1.82–1.50 (m, 54H, LeuCH2, LeuCH, AibCH3), 1.00–0.85 (m, 36H, LeuCH3). MALDI-TOF MS; calcd. for C97H182N12O32 [M + Na]+: 2050.31, found.: 2050.13. 29

References 1.

2.

3. 4. 5. 6. 7. 8. 9. 10. 11.

12. 13. 14. 15. 16. 17. 18. 19. 20. 21.

22. 23.

Loew, M.; Springer, R.; Scolari, S.; Altenbrunn, F.; Seitz, O.; Liebscher, J.; Huster, D.; Herrmann, A.; Arbuzova, A. J. Am. Chem. Soc. 2010, 132, 16066–16072. Christian, D. A.; Tian, A.; Ellenbroek, W. G.; Levental, I.; Rajagopal, K.; Janmey, P. A.; Liu, A. J.; Baumgart, T.; Discher, D. E. Nat. Mater. 2009, 8, 843–849. Semrau, S.; Schmidt, T. Soft Matter 2009, 5, 3174–3186. Holowka, E. P.; Sun, V. Z.; Kamei, D. T.; Deming, T. J. Nat. Mater. 2007, 6, 52–57. Harada, A.; Kataoka, K. Science 1999, 283, 65–67. Cornelissen, J. J. L. M.; Fischer, M.; Sommerdijk, N. A. J. M.; Nolte, R. J. M. Science 1998, 280, 1427–1430. Corinna, F.; Jens, G.; Lea, M.; Giuseppe, B.; Robert, L. Sci. Rep. 2016, 6, 33491. Discher, B. M.; Won, Y. Y.; Ege, D. S.; Lee, J. C.; Bates, F. S.; Discher, D. E.; Hammer, D. A. Science 1999, 284, 1143–1146. Mohammed, A. M.; Šulc, P.; Zenk, J.; Schulman, R. Nat. Nanotechnol. 2017, 12, 312–316. Ahmed, F.; Photos, T. J.; Discher, D. E. Drug Dev. Res. 2006, 67, 4–14. Coleman, A. C.; Beierle, J. M.; Stuart, M. C. A.; Maciá, B.; Caroli, G.; Mika, J. T.; Dijken, D. J. V.; Chen, J.; Browne, W. R.; Feringa, B. L. Nat. Nanotechnol. 2011, 6, 547–552. Roux, A.; Cuvelier, D.; Nassoy, P.; Prost, J.; Bassereau, P.; Goud, B. EMBO J. 2005, 24, 1537–1545. Ueda, M.; Uesaka, A.; Kimura, S. Chem. Commun. 2015, 51, 1601–1604. Ueda, M.; Makino, A.; Imai, T.; Sugiyama, J.; Kimura, S. Polym. J. 2013, 45, 509–515. Ueda, M.; Makino, A.; Imai, T.; Sugiyama, J.; Kimura, S. Soft Matter 2011, 7, 4143–4146. Ueda, M.; Makino, A.; Imai, T.; Sugiyama, J.; Kimura, S. Langmuir 2011, 27, 4300–4304. Ueda, M.; Makino, A.; Imai, T.; Sugiyama, J.; Kimura, S. Chem. Commun. 2011, 47, 3204–3206. Ueda, M.; Makino, A.; Imai, T.; Sugiyama, J.; Kimura, S. J. Pept. Sci. 2011, 17, 94–99. Griffin, W. C. J. Soc. Cosmet. Chem. 1949, 1, 311–326. Griffin, W. C. J. Soc. Cosmet. Chem. 1954, 259. Devies, J. T. A Quantitative Kinetic Theory of Emulsion Type. I Physical Chemistry of the Emulsifying Agent. In Gas/Liquid and Liquid/Liquid Interfaces; Proceedings of 2nd International Congress Surface Activity; Butterworths: London, 1957. Israelachvili, J. N. Intermolecular and Surface Forces; 3rd ed.; Academic Press: 2010 Kanzaki, T.; Horikawa, Y.; Makino, A.; Sugiyama, J.; Kimura, S. Macromol. Biosci. 2008, 8, 1026–1033. 30

Chapter 3

Peptides as Smart Biomolecular Tools: Utilization of Their Molecular Recognition for Materials Engineering Toshiki Sawada and Takeshi Serizawa* Department of Chemical Science and Engineering, School of Materials and Chemical Technology, Tokyo Institute of Technology, 2-12-1-H121 Ookayama, Meguro-ku, Tokyo 152-8550, Japan *E-mail: [email protected]

Biomolecules express outstanding properties required for molecular recognition and nanoscale self-assembly. These smart capabilities have been obtained through evolution, and these biomolecules utilized in nature based on a smart function. Recently, their excellent capabilities have been utilized in not only biosystems, but also in materials science and engineering for functional materials. The bioinspired peptide selection technology using phage display systems has been developed based on this natural evolution to generate novel functional peptides. In this chapter, we focused on peptides with a specific affinity for synthetic polymers. The polymer-binding peptides were obtained from a biologically constructed phage-displayed peptide library to recognize polymeric nanostructures, and were utilized for possible applications as novel biomolecular tools for functional polymers. Our approach to utilize peptides as novel biomolecular tools will open excellent opportunities for the next-generation of materials science and engineering.

1. Introduction Natural biomolecules such as peptides, proteins, nucleic acids, and saccharides exhibit the properties required for molecular recognition and nanoscale self-assembly. Such biomolecules have been constructed through evolutional systems, and functioned based on smart molecular recognition. © 2017 American Chemical Society

Molecular recognition plays an important role in natural systems, and can be observed between antigen-antibody (1, 2), DNA-protein (3, 4), RNA-ribosome (5), and sugar-lectin (6) interactions to sustain biosystems. These molecules have undergone structural-fitting through evolution to appropriately combine numerous weak interactions (noncovalent bonds) such as electrostatic, hydrogen bonding, π-stacking, van der Waals interactions, and hydrophobic effects. The targets of these biomolecules have been considered to naturally occurring ligands present in the biological systems. However, recent studies have revealed certain peptides with specific affinities for artificial materials through bioinspired affinity-based selection procedures from biologically constructed peptide libraries displayed on phages or cell surfaces (7–10). The surfaces of metals, metal oxides, semiconductors, magnets, nanocarbons, synthetic polymers, and artificially designed peptide assemblies were utilized as the specific targets of these novel peptides. These peptides have a regular structure, and therefore might recognize two- or three-dimensional regular distributions of atoms or functional groups on the material surfaces, thus resulting in specific affinities. These studies suggested that biomolecules, particularly peptides, could be evolved toward specific interactions for artificial materials by suitable selection procedures that mimics the natural evolutionary system. Recently, due to limitations in cell surface display systems for the selection against artificial materials, phage display systems have been widely utilized for construction of material-binding peptides. These peptides are known not only as material binders, but also as functional molecular tools for a wide range of applications such as surface modifiers, adsorbents for patterning, catalysts for the preparation of inorganic particles, constructs of polymeric nanoparticles, and switchers for conjugated polymer fluorescence. Since the affinities of the peptides were effectively utilized to achieve all of the above applications, the molecular recognition capabilities of these peptides will be effective for constructing novel classes of peptide-based nanomaterials.

Figure 1. Possible polymeric nanostructures for peptide recognition. Reprinted with permission from ref (11). Copyright 2011 Royal Society of Chemistry. 32

The purpose of this chapter is to overview the advances that have been made using specific interactions through molecular recognition for artificial materials (especially synthetic polymers) for fabricating a novel class of biomolecular tools by our group over the past decade. Indeed, we have screened, characterized, and utilized polymer-binding peptides against polymeric materials using phage displayed peptide libraries (11, 12). We are forcusing on peptides that discriminate against slight structural differences in synthetic polymers. A primary sequence, stereoregularity, amphiphilicity, crystallinity, porosity, a linear/branched structure, and an assembled structure may fit into the three-dimensionally regular nanostructures of certain peptides (Figure 1). Nanomaterials composed of peptides using those excellent capabilities will open attractive opportunities for the science and technology of next-generation biomolecular tools and nanomaterials.

2. Selection of Polymer-Binding Peptides 2.1. Synthetic Polymers as Peptide Targets It is crucial to understand and regulate biological phenomena such as adsorption of proteins and adhesion of cells on the surface of synthetic polymers at the molecular level for developing the novel polymer materials utilized in biomedical fields (13–17). In the case of hydrophobic or charged polymers, proteins can easily adsorb onto polymer surfaces, thereby mediating various biological processes such as cell adhesion, extension, and proliferation. These interfacial phenomena is normally hard to understand, due to multiple and complicated weak interactions at the interfaces. Hence, we originally focused on more reliable interactions of biomolecules against the polymer surfaces. Furthermore, it is also interesting whether biomolecular peptides can recognize nanostructures derived from the characteristic of synthetic polymers, which are typically stereoregularity, amphiphilicity, crystallinity, and so on. In the area of inorganic materials, Belcher and co-workers described peptides discriminating crystal defects in germanium thin films on silicon and germanium wafers, suggesting that the recognition capability of peptides could be utilized for nondestructively probing and identifying the localization of defects in crystalline substrates (18). The result indicated that the molecular recognition capability of artificially evolved peptides could be exploited in materials engineering. Although applications of material-binding peptides have been limited at solid-liquid interfaces due to their selection conditions, peptides with specific affinities for water-soluble polymers would show novel potential utilization in solutions. Thus, peptides that can discriminate slight differences in polymeric nanostructures will open attractive new opportunities for polymer science and technology. In 1985, it was reported that desired peptides could be genetically displayed on the coat proteins of filamentous phages based on the insertion of the corresponding DNA fragment into the phage genome (19). The infectiousness of the resultant genetically engineered phages against a host Escherichia coli (E. coli) was maintained. Thus, the genetically engineered phages could be easily 33

amplified. One of the most commonly utilized phages is M13 filamentous phage, which is composed of five kinds of coat proteins termed pIII, pVI, pII, pVIII, and pIX on the surface (Figure 2). When peptides with different amino acid sequences are displayed on the surface of each different phage, a phage-displayed peptide library can be constructed in principle. Recently, the phage display method is a versatile and important tool for the selection of ligands for biomolecules such as peptides and proteins (20, 21). This approach has also been accepted over the past decade to select material-binding peptides. The selection and characterization of such phage-displayed material-binding peptides has attracted great interest, especially due to their wide utilization in material engineering and nanotechnology. The protocols for the construction of phage-displayed peptide libraries have been established (21, 22). The bioinspired selection process using a peptide library displayed on phages to target-specific phage pools is called biopanning (Figure 3). Biopanning is composed of four fundamental experimental steps as described below. Briefly, step 1 is the interaction in which an aqueous solution of the phage libraries is mounted onto the target material surfaces (solid such as film, particle, and so on) for an adequate time. In step 2, the washing, the target material surfaces are washed with a buffer solution (sometimes containing detergents) several times to remove unbound or weakly bound phages. In step 3, the elution, the strongly bound phages are eluted from the material surfaces by an eluent (acid solutions are usually used). In step 4, the amplification, the eluted phages are then proliferated within E. coli for amplification. After the appropriate number of biopanning cycles, cloning of the selected phages followed by the DNA sequencing of the displayed peptides on each phage clone, which corresponds to each peptide sequence, is performed.

Figure 2. Schematic illustration of M13 filamentous phage displaying peptides on pIII. Reprinted with permission from ref (12). Copyright 2013 Wiley. 34

Figure 3. Schematic illustration of the screening process using a phage-displayed peptide library. Reprinted with permission from ref (12). Copyright 2013 Wiley. 2.2. Recognition of Polymer Stereoregularities by Selected Peptides 2.2.1. Selection and Characterization of Stereoregular Polymer-Binding Peptides For the first target, peptide selection using a phage library with 7-mer random peptides was performed against a stereoregular polymer, isotactic poly(methyl methacrylate) (it-PMMA) (Figure 4a), in which the side chains are almost all on one side of the backbone (Figure 5a) (23). In the case of this target polymer, syndiotactic (st-) PMMA, in which the side chains on alternate sides, was used as a reference polymer. After five rounds of biopanning for it-PMMA spin-cast film, nine phage clones were identified. Enzyme-linked immunosorbent assay (ELISA) using the phage clones revealed that the binding amounts of all phage clones for itPMMA were much greater than those for st-PMMA. The ELISA result suggested that the four amino acid peptide motif (Arg-Pro-Thr-Arg) composed of amino acids with proton-donor hydroxyl and amino lateral groups adjacent to the Pro was an essential for both affinity and specificity. Considering the high molecular weight of the M13 phage (16 300 000) (24), it was surprising that the 7-mer peptides displayed on the filamentous phage termini greatly affected the binding capability of the phage clones. 35

Figure 4. Chemical structures of synthetic polymers applied to the peptide screening. (a) Poly(methyl methacrylate) (PMMA), (b) poly(L-lactide) (PLLA), (c) polystyrene(PS), (d) linear (top) and branched (bottom) poly(p-phenlene vinylene) (PPV), (e) Polyetherimide (PEI), (f) poly(propylene oxide) (PPO), (g) poly(2-methoxy-5-propyloxysulfonate-1,4-phyenylenvinylene), and (h) poly(N-isopropylacrylamide) (PNIPAM).

Figure 5. (a) Recognition of stereoregular it-PMMA films by the c02 peptide with a sequence of Glu-Leu-Trp-Arg-Pro-Thr-Arg. The C-terminal 4-mer peptide (Arg-Pro-Thr-Arg, RPTR) was essential for the specific binding. (b) Comparison of the possible structures of RPTR obtained by Molecular Mechanics and hexa MMA units. Reprinted with permission ref (23). Copyright 2005 ACS and ref (25). Copyright 2007 ACS.

36

The binding of the chemically-synthesized 7-mer peptides for the target itPMMA and the reference st-PMMA were monitored in a real time and kinetically analyzed by surface plasmon resonance (SPR) measurements, and their kinetic parameters were determined (25). The binding constant (Ka) values of the c02 peptide with the sequence Glu-Leu-Trp-Arg-Pro-Thr-Arg approached 2.8 × 105 M-1 for the target, which was 40 times greater than that for the reference polymer (6.9 × 103 M-1). The value for the target was comparable to the Ka values of 12-mer peptides that bind to titanium oxide surfaces (26), suggesting that peptides have the potential to discriminate subtle differences in the stereoregularity of polymeric structures. Detailed analyses by Ala-scanning (binding analyses using substituted peptides, in which each amino acid was changed to Ala) clearly demonstrated that the Ka values of the Ala-substituted peptides decreased significantly. The smallest Ka value was more than 30 times less than that of the original c02 peptide. The magnitude of the decrease of the Ka values was as follows: Pro > Thr > Arg7 > Glu > Arg4, suggesting that these amino acids were essential for the specific binding. These observations from the Ala substitution directly evidenced the importance of the recognition of the polymeric structures by amino acid side chains. A rigid conformation of the peptide derived from a kinked Pro residue should provide three-dimensionally arrange the lateral hydroxyl and amino groups of the Thr and the two Arg residues as proton-donors to successfully recognize the position of the it-PMMA ester groups through hydrogen bonding. To check the importance of the it-PMMA binding motif, the Ka values of the N-terminal (Glu-Leu-Trp-Arg) and C-terminal (Arg-Pro-Thr-Arg) peptides for itand st-PMMA were determined. The Ka of the C-terminal peptide for it-PMMA (4.6 × 104 M-1) was greater than that for st-PMMA (9.4 102 M-1) and that of the N-terminal peptide for it-PMMA (1.6 × 103 M-1). Therefore, it was demonstrated that the C-terminal 4-mer sequence of the c02 peptide, Arg-Pro-Thr-Arg, which is composed of proton-donor hydroxyl and amino lateral groups adjacent to the Pro, was an essential motif for it-PMMA recognition, and that peptides composed of four amino acid residues have the enough potential to recognize polymer stereoregularity (Figure 5a). Possible structures with stable and sterically acceptable conformations of the 4-mer motif were obtained by energy optimization using Molecular Mechanics, and the obtained size of the structural conformation was comparable to the size for 6 units of MMA (Figure 5b). To further characterize these peptide recognition mechanisms, thermodynamic parameters for the c02 peptide against it- and st-PMMA films were determined by analyzing the Ka value at various temperatures (25). The Ka value decreased with increasing temperature, suggesting that the predominant interactions of the peptide affinities for it-PMMA films was derived from hydrogen bonding interactions. The estimated enthalpy (ΔH°) and entropy (ΔS°) changes assuming the van’t Hoff equation suggested that greater conformational changes of the c02 peptide and/or it-PMMA occurred to obtain the specific affinity. In other words, the specific affinities seemd to be derived from an induced fit mechanism. Considering all of the results, it was clearly demonstrated that short peptides had the potential capability to discriminate polymeric nanostructures through specific binding as natural biomolecules did (Figure 5). 37

2.2.2. Generality of Peptide Recognition for Stereoregular Polymers

Previously utilized as a reference polymer (st-PMMA) was applied to the next target of the selection using a phage-displayed linear 7-mer peptide library to determine whether peptides can generally recognize polymeric nanostructures derived from polymer stereoregularity (27). Furthermore, because the surface accumulation of functional groups on st-PMMA films are affected by external environment, st-PMMA is an attractive interesting as a novel peptide target. It is well known that when st-PMMA films are prepared on glass slides in air, hydrophilic ester and hydrophobic alkyl groups are accumulated on the glass and air sides, respectively (28). Therefore, st-PMMA films with variable surface properties are suitable candidates for demonstrating both the potential specificity of peptides against materials and the novel strategy to control interactions between the peptides and softmaterials. Therefore, st-PMMA films were prepared on glass surfaces and immersed in a buffer for 15 h (conditioning). This treatment resulted in the exposure of the ester groups to the air side. After three rounds of biopanning for the conditioned st-PMMA films, several phage clones were identified. DNA sequencing revealed the phage clones displaying peptides composed of several amino acid residues with amino groups and a Pro residue. Therefore, it was anticipated that hydrogen bonding interactions between the peptides and the st-PMMA were essential for the affinity. ELISA result indicated that one phage clone displaying the peptide with a sequence of His-Lys-Pro-Asp-Ala-Asn-Arg had a specific affinity for the target st-PMMA rather than for the reference it-PMMA. The Ka values of chemically-synthesized peptides for st- and it-PMMA films were determined to 9.1 × 104 M-1 and 3.0 × 103 M-1, respectively. Therefore, it was demonstrated that peptides could generally discriminate subtle differences in nanostructures derived from polymer stereoregularities. More interestingly, the binding amounts of the phage clones for the buffer conditioned st-PMMA films as peptide targets were much greater than those for non-conditioned st-PMMA films. In fact, the static contact angles of the st-PMMA films decreased with increasing conditioning time, suggesting that the hydrophilic ester groups of st-PMMA rather than the hydrophobic alkyl and/or methyl groups were gradually exposed into the water side. Therefore, it was confirmed that peptides can recognize nanoscaled arrangements of ester groups due to the syndiotacticities of the methacrylates. As a consequence, peptide motifs that could recognize variability in polymer film surfaces were discovered based on their selection from peptide-displaying phage libraries directed against adequately-treated st-PMMA films. Furthermore, it is well known that it- and st-PMMAs self-assemble into triple-stranded helix structures, stereocomplexes (SCs), in which the it-PMMA double helices are surrounded by a single st-PMMA helix (29), and the SCs could be prepared on substrates through a stepwise layer-by-layer (LbL) assembly (30). Such structurally defined stereocomplex structures were applied to the selection process using a 7-mer peptide displaying phage library (31). The obtained peptide with a sequence of Ser-Thr-Pro-Pro-Arg-Leu-Trp bound specifically to the SC films as compared to single it- or st-PMMA spin-cast films. Therefore, 38

nanostructures prepared from stereoregular PMMAs would generally be definite targets for peptides.

2.3. Recognition of Other Polymer Structures by Peptides Poly(L-lactide) (PLLA) (Figure 4b) is one of the most commonly-used synthetic polymers in biomedical fields due to high thermal stability, mechanical properties, biocompatibility, and biodegradability for non-toxic products (32, 33). However, it is difficult to introduce functionalities onto the surface of PLLA because of its few active groups, thus resulting in limiation of its biomedical applications. Here, α-formed crystalline PLLA was applied to the selection using the 7-mer phage-displayed peptide library toward surface functionalization of PLLA (34). The Ka value of the chemically-synthesized peptide with a sequence of Glu-Leu-Met-His-Asp-Tyr-Arg, which was selected from the library, for the PLLA films determined by SPR kinetic measurements (6.1 × 104 M-1) was 10 times higher than for amorphous PLLA films. Therefore, the obtained peptide clearly recognizes the polymeric structures derived from a simple annealing process that enhances polymer crystallinity. Furthermore, the Ka value for the 103 helical α-form was 4 times higher than that for the 31 helical β-form, indicating that the peptide could discriminate slight differences in the helical pitch delivered from the PLLA morphs. Polystyrene (PS) (Figure 4c) is a universal vinyl polymer frequently utilized as plastic plates for biological experiments. Previously, PS binding peptides were inadvertently obtained (35), even though PS was consisting of simple alkyl main chains and lateral phenyl groups. We focused on specific nanostructures formed by syndiotactic PS (sPS), and applied to the selection procedures using a 7-mer peptide library for the sPS films (36). When the films were prepared from sPS solutions dissolved in a suitable solvent, sPS forms TTGG helical conformations (T and G represent trans and gauche conformations, respectively) (37–39). Evaporation of the free solvent resulted in δ-form films complexed with solvent molecules. Further evaporation under thermal heating conditions fabricated empty δ-form (δe) porous films, and the δe formed sPS films were utilized as peptide targets. The sequence of the selected peptides displayed on the phage clone was Phe-Ser-Trp-Glu-Ala-Phe-Ala composed of aromatic and aliphatic amino acids, suggesting that π-stacking and hydrophobic effects are essential for the sPS binding. In fact, binding amounts of the phage clones determined by ELISA against films composed of target sPS, atactic PS, isotactic PS, and sPS/toluene complexes (δ-form) clearly demonstrated the highly specific affinity of the peptide for the δe form, indicating that peptides have the potential to recognize porous nanostructures composed of synthetic polymers. Conjugated polymers are attractive materials due to their unique optical and electronic properties. Thus, the polymers have been utilized in a number of applications including light emitting diodes (40), photovoltaic cells (41), and biosensing (42). A representative conjugated polymer, poly(phenylene vinylene) 39

(PPV) (Figure 4d), which has a simple chemical structure composed of vinyl groups and phenyl rings, is the most extensively studied conjugated polymer. We focused on peptide recognition for linear and branched polymer structures of PPV, and both PPVs were applied to the peptide selection using a phage library with 12-mer peptides (43). The obtained peptides for each target showed specific affinities, indicating that the peptides recognized the structural differences in the PPV isomers. Because aromatic amino acids were observed in sequences as His-Thr-Asp-Trp-Arg-Leu-Gly-Thr-Trp-His-His-Ser, π-stacking interactions between the peptides and the PPV might be essential. Possible molecular structures determined by molecular mechanics suggested that two Trp residues in the peptide gave suitable interactions with the phenyl ring in the branched PPV. Polyetherimide (PEI) (Figure 4e) is an engineering plastic with excellent physicochemical properties, and has been widely utilized in advanced industries. PEI is a candidate for replacement of metallic materials, and thus has been increasingly employed as a biomaterial (44). Since regulating these biological responses as well as adding another function at the PEI surface is essential for the development of softmaterials, PEI binding peptides were selected from a 7-mer peptide library on phages (45). The Ka value of the chemically synthesized peptide with a sequence of Thr-Gly-Ala-Asp-Leu-Asn-Thr for the target PEI films (5.6 × 108 M-1) was much greater than that for non-target thermally treated PEI films (4.4 × 104 M-1). The aforementioned difference in the Ka values must be attributed to the structural differences in the PEI surfaces before and after thermal treatment. Since thermal treatment of the PEI films strongly affected molecular aggregations or stacking of the monomer units, the identified peptides should discriminate the structural changes of the PEI films. Poly(propylene oxide) (PPO) (Figure 4f) is a hydrophobic component for tribrock copolymer composed of a central PPO chain flanked by two hydrophilic chains of poly(ethylene ooxide) (PEO). The triblock copolymers have been widly utilized as emulsifiers and carrier matrices for a drug delivery system (DDS) because they form micelles, followed by physically closs-linked hydrogels depending on their concentrations. Because it is difficult to functionalize the triblock copolymer due to their few active groups, further applications in biomedical fields are limited. Therefore, we selected peptides with specific affinity for PPO using the 12-mer peptide library (46). After 2 rounds of biopanning, single phage clone displaying peptides with a sequence of Asp-Phe-Asn-Pro-Tyr-Leu-Gly-Val-Thr-Pro-Val-Lys. The Ka value of the peptide for PPO films (1.2 × 106 M-1) is much higher than that for reference PMMA films (3.6 × 104 M-1), indicating specific affinity for PPO. Time dependent release of the peptides from the hydrogels composed of PEO-PPO-PEO was investigated, and the release properties were correlated with the affinities of the peptide, indicating certain binding of the peptide for the target segment PPO even in the hydrogels. Because the aforementioned synthetic polymers used as peptide targets are water-insoluble, the applications for these polymer-binding peptides would be limited to solid surfaces. Therefore, we applied to the peptide selection method to films composed of water-soluble poly(2-methoxy-5-propyloxysulfonate-1,4phenylen vinylen) (mpsPPV) (Figure 4g) using a layer-by-layer (LbL) assembly technique with poly(diallyldimethylammonium chloride) (PDDA). The identified 40

peptide with a sequence of His-Asn-Ala-Tyr-Trp-His-Trp-Pro-Pro-Ser-Met-Thr bound to not only the LbL films, but also to dissolved mpsPPV in aqueous solution (47). These results indicated that a wide variety of water-soluble polymers could be peptide targets. Poly(N-isopropylacrylamide) (PNIPAM) (Figure 4h) is a representative classic thermoresponsive water-soluble polymer and show a reversible coil-grobule transition at the lower critical solution temperature (LCST). The LCST of PNIPAM is near the body temperature; therefore, PNIPAM has been widely used in biomedical fields such as scaffolds to construct cell sheets and thermorespovsive precipitation of desired proteins. Because the interaction of biomolecular ligands into PNIPMA chains through chemical modification is sometimes complicated and time consuming, we selected 12-mer peptides with a specific affinity for water insoluble PNIPAM with high meso diad content (85%) from the peptide library (48). The identified chemically synthesized peptides showed higher Ka value for the PNIPAM films (2.0 × 105 M-1) rather than for it-PMMA films (1.7 × 104 M-1), indicating specific affinity for the PNIPAM. Furthermore, the peptide specifically lowered the LCST of meso-rich (58%) PNIPAM rather than meso-poor PNIPAM, demonstrating the peptide preferentially bound to the mesosequence of PNIPAM chains even in an aqueous phase. As mensioned above, we have successfully screened peptides with spefic affinities for various synthetic polymers (Table 1). The peptides clearly discriminated unique polymeric structures such as stereoregularity, amphiphilic structures, crystallinity, porous structures, linear/branched structures, and so on. Because such artificial nanostructures derived from synthetic polymers were recognized by biomolecular peptides, further various polymers would be applied to the screening procedure to obtain their specific peptides for future applications.

3. Applications of Polymer-Binding Peptides Functionalizaon of material surfaces plays an essential role in material science and engineering. In various surface modification methodorogies, it is usually used to form self-assembled monolayers (SAMs) on inorganic material surfaces for surface modification, both practically and experimentally, because of the simpleness for preparation of functional surfaces with well-defined compositions on substrates (49–51). If other softmaterials could be functionalized by such a concept of SAMs, the range of technical applications would be broaden. Therefore, we anticipitated peptides with specific affinities for the synthetic polymers were used for surface modification towards biomedical applications (Figure 6a). As target substrates, PEI (45) and PMMA SC (31) films were utilized for surface modification through their specific peptide. Biotin with extermely high affinity for streptavidin (SAv) was used as a functional moiety conjugated to the specific peptides. For the immobilization of SAv, the biotin-conjugated 41

peptides were immobilized onto the polymer surfaces. The amount of bound SAv was dependent on the density of the immobilized biotin-conjugated peptides, suggesting that SAv molecules were immobilized onto the substrate through the biotin molecule conjugated to the peptides. The immobilized SAv through the biotin-conjugated peptides could be utilized to further immobilization of probe DNA to efficiently hybridize with its complementary DNA as compared to directly immobilized SAv on the PEI films. The SAv immobilized through the biotinylated specific peptide formed uniform monolayers without denaturing on the PEI films. Therefore, it was demonstarated that polymer-binding peptides had great potential for the functional modification of polymer surfaces.

Table 1. Summary of characterization of polymer-binding peptides Polymer

Ka/ 105 M-1

Sequence

Specificity

Target

Reference

it-PMMA

ELWRPTR

2.8

0.069 (st-PMMA)

41

st-PMMA

HKPDANR

0.91

0.030 (it-PMMA)

30

PLLA (α-form)

ELMHDYR

0.61

0.057 (Amorphous PLLA)

11

Linear PPV

ELWSIDTSAHRK

0.77

0.36 (Branched PPV)

2.1

Branched PPV

HTDWRLGTWHHS

7.7

0.52 (Liner PPV)

15

mpsPPV

HNAYWHWPPSMT

1.3

-

-

PEI

TGADLNT

5600

0.44 (Thermally treated PEI)

13000

PPO

DFNPYLGVTPVK

12

0.36 (PMMA)

33

PNIPAM

HSFKWLDSPRLR

2.0

0.17 (PMMA)

12

42

Figure 6. (a) Representative applications of polymer-binding peptides: (a) polymer surface modification by proteins through the peptides, (b) polymer-binding peptide-fused proteins, (c) peptide capped gold nanoparticles, and (d) conjugated polymer nanoparticles hybridized to the polymer-binding peptides with and without staining of peptides. Scale bars in (c) and (d) represent 100 nm. (e) Alive and dead cells incubated with the Dox-conjugated peptides or the original peptides released from polymeric hydrogels. Scale bars represent 100 µm. (f) Thermoresponsive precipitation of the peptide-fused HSA in the presence of thermoresponsive polymers. Reprinted with permission from (c) ref (63). Copyright 2009 ACS, (d) ref (70). Copyright 2011 Royal Society of Chemistry, (e) ref (46). Copyright 2016 Royal Society of Chemistry. In the previous paragraph, a functional protein (SAv) was immobilized through the biotin-conjugated polymer-binding peptide. Another protein immobilizaition strategy on surfaces is to fuse the peptide for the functionalization of protein directly. In fact, apoferritin (52, 53), cytokines (54), and green fluorescent protein (55) have been fused with material binding peptides for immobilization through specific interactions. These proteins have the capability to be stably immobilized on inorganic and nanocarbon materials through noncovalent but specific interactions. We also prepared proteins fused with the polymer binding peptides for enhanced adsorption against polymeric substrates. The c02 peptide, which binds specifically to the it-PMMA films (Figure 6b) (56), was conjugated to DnaK419-607, a blocking peptide fragment (BPF) which is part of the substrate binding domain of the molecular chaperon DnaK from E. coli (57, 58), to construct an effective blocking reagent due to its excellent adsorptive 43

properties for plastic substrates through its hydrophobic domain (59). The c02 peptide was fused to the N-terminus of the BPF derivative (dBPF). The Ka of the c02 peptide-fused-dBPF for it-PMMA was 4.9 × 108 M-1, which was 75 times greater than that of the original dBPF. Thus, the fusion protein showed specific affinity for it-PMMA, and the specificity was due to the fused c02 peptide. Furthermore, other functional molecules were also specifically immobilized onto other kinds of substrates. Gold nanoparticles (GNPs) have potential applications in the biological and nanotechnological fields due to their unique surface plasmon properties (60–62). Therefore, the surface functionalization of GNPs is required to widen their applicabilities against various fields. We prepared GNPs with a specific affinity for it-PMMA by capping of peptide with the sequence of Arg-Pro-Thr-Arg, which is an essential motif for the it-PMMA affinity (25). Cys was fused to the N-terminus of the motif as an anchor for the gold surface through strong thiol-Au binding. AuCl4- was reduced in a 2-[4-(2-hydroxyethyl)-1-piperazinyl]ethanesulfonic acid (HEPES) buffer containing the designed 5-mer peptide with a sequence of Cys-Arg-Pro-Thr-Arg, under mild conditions (pH 7.2, ambient temperature), resulting in peptide-capped GNPs, which were stably dispersed in HEPES buffer solutions (Figure 6c) (63). The prepared nanoparticles showed preferentialy affinity for it-PMMA film surface as compared to st-PMMA surfaces, insicating that polymer-binding peptides could be used as anchor molecules for the heterogeneous interface between inorganic compounds and synthetic polymers. Peptide-capped nanoparticles with cores of gold (64, 65), silver (66), platinum (64), silica (67), and polymeric nanogels (68) have been prepared and studied for biomedical and bio-analytical applications. Moreover, conjugated polymer nanoparticles (CPNs) are gaining attention because of their excellent optical and electronic properties (69). Although the desired surface functionalization of CPNs would expand potential capabilities utilized in optical and biomedical fields, only covalent functionalization methods have been reported thus far. Therefore, we have tried to construct peptide-capped CPNs through noncovalent specific interactions. CPNs composed of water-insoluble, branched PPV and its specific peptide were prepared by simply mixing of the PPV dissolved in an organic solvent into an aqueous solution of the specific peptide, and subsequent sonication processes (Figure 6d) (70). Since the peptides were internalized into the CPNs, the CPNs were clearly different from conventionally prepared polymer nanoparticles. This is the first demonstration of utilization of a specific binding peptide for the preparation of polymeric nanoparticles. To demonstrate the biomedical applicability of the polymer-binding peptide, the PPO-binding peptide was modified with an anticancer drug molecule, doxorubicine (Dox), thorugh suitable linker molecules to construct a controlled release system of drugs from the PEO-PPO-PEO tribrock copolymer hydrogels. The release rate of the Dox-conjugated peptides was slightly smaller than that of the original PPO-binding peptide under the same conditions, indicating that the specific affinity was remained after modification with hydrophobic Dox. The release system was applied to cell culture assays to clarify effectiveness of the controlled release system. Numbers of alive HeLa cells were successfully decreased with increasing incubation time in the presence of the Dox-conjugated 44

peptide-containing PEO-PPO-PEO hydrogels, demonstrating sustainable anticancer effects (Figure 6e) (46). Therefore, it was demonstrated that controlled release systems using polymeric hydrogels and specific peptides could be utilized for future biomedical applicaations. The peptide affinity for water-soluble thermoresponsive polymers was utilized to precipitate desired proteins. Human serum albumin, a model protein, was chemically modified with the PNIPAM-binding peptides through a disulfide bond to specifically interact with the meso-rich PNIPAM dissolved in water. The Ka value of the peptide-modified human serum albumin (HSA) for the target PNIPAM was 4.4 × 107 M-1, which was 17 times greater than that of native HSA, indicating that peptide on HSA molecules still showed an affinity for the meso diad sequence of PNIPAM. Thermoresponsive precipitation experiments using meso-rich PNIPAM and HSA modified with or without peptides were performed. As a result, 98% of the peptide-modified HSA was precipitated with the meso-rich PNIPAM, even though approximately 20 % of unmodified HSA was precipitated with the meso-rich PNIPAM (Figure 6f) (48). Hence, the PNIPAM-binding peptide-modification was useful to thermoresponsively precipitate and collect desired biomacromolecules with PNIPAM.

4. Conclusion Recent developments in synthetic polymer-binding peptides that can recognize the nanostructures of synthetic polymers were described. Our observations demonstrated that polymeric nanostructures derived from unique stereoregularity, amphiphilicity, crystallinity, porosity, linear/branching structure, and dynamic comformational changes could be targets for the peptide. The use of these polymer-binding peptides for surface functionalization through noncovalent peptide or fusion protein interaction, metal nanoparticle functionalization with novel synthetic systems, a one-pot synthesis of polymer-peptide conjugated particles, controlled release systems from polymeric hydrogels, and thermoresponsive precipitation of proteins were achieved. In all cases, peptide recognition against the synthetic polymers was effectively utilized. Therefore, it was evidenced that biomolecular peptides could be versatile molecular tools for the development of novel material innovations. Biomolecular peptides have great potential for use in the materials science and engineering field, and are superior to native biological functions. Bioinspired selection technology using phage display systems has exploited the natural process of material evolution to create a new generation of bionanomaterials with novel functions. These excellent characteristics of peptides and strategies for the construction of bioinspired peptidyl molecular tools will open attractive opportunities for the science and engineering of next-generation bionanomaterials.

References 1.

Janeway, C. A.; Travers, P.; Walport, M.; Shlomchik., M. J. Immunobiology; Garland Science: New York, 2001. 45

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. 30. 31. 32.

Litman, G. W.; Rast, J. P.; Shamblott, M. J.; Haire, R. N.; Hulst, M.; Roess, W.; Litman, R. T.; Hinds-Frey, K. R.; Zilch, A.; Amemiya, C. T. Mol. Biol. Evol. 1993, 10, 60–72. Yang, W.; Duyne, G. D. V. Curr. Opin. Struct. Biol. 2006, 16, 1–4. Helwa, R.; Hoheisel, J. D. Anal. Bioanal. Chem. 2010, 398, 2551–2561. Ogle, J.; Carter, A.; Ramakrishnan, V. Trends Biochem. Sci. 2003, 28, 259–266. Lis, H.; Sharon, N. Chem. Rev. 1998, 98, 637–674. Sarikaya, M.; Tamerler, C.; Jen, A.; Schulten, K.; Baneyx, F. Nat. Mater. 2003, 2, 577–585. Sarikaya, M.; Tamerler, C.; Schwartz, T. D.; Baneyx, F. Annu. Rev. Mater. Res. 2004, 34, 373–408. Baneyx, F.; Schwartz, D. Curr. Opin. Biotechnol. 2007, 18, 312–319. Shiba, K. Curr. Opin. Biotechnol. 2010, 21, 412–425. Serizawa, T.; Matsuno, H.; Sawada, T. J. Mater. Chem. 2011, 21, 10252–10260. Sawada, T.; Mihara, H.; Serizawa, T. Chem. Rec. 2013, 13, 172–186. Elbert, D. L.; Hubbell, J. A. Annu. Rev. Mater. Sci. 1996, 26, 365–394. Ha, C.-S.; Gardella, J. Chem. Rev. 2005, 105, 4205–4237. Vogler, E. Adv. Colloid Interface Sci. 1998, 74, 69–186. Kasemo, B. Surf. Sci. 2002, 500, 656–677. Wang, Y.-X.; Robertson, J.; Spillman, W.; Claus, R. Pharm. Res. 2004, 21, 1362–1435. Sinensky, A. K.; Belcher, A. M. Adv. Mater. 2006, 18, 991–996. Smith, G. Science 1985, 228, 1315–1322. Smith, G. P.; Petrenko, V. A. Chem. Rev. 1997, 97, 391–410. Kehoe, J.; Kay, B. Chem. Rev. 2005, 105, 4056–4072. Maassen, C.; Laman, J.; den Bak-Glashouwer, M.; Tielen, F.; van Holten-Neelen, J.; Hoogteijling, L.; Antonissen, C.; Leer, R.; Pouwels, P.; Boersma, W.; Shaw, D. Vaccine 1999, 17, 2117–2128. Serizawa, T.; Sawada, T.; Matsuno, H.; Matsubara, T.; Sato, T. J. Am. Chem. Soc. 2005, 127, 13780–13781. Barbas, C. F.; Burton, D. R.; Scott, J. K.; Silverman, G. J. Phage Display: A Laboratory Manual; Cold Spring Harbor Press: New York, 2001. Serizawa, T.; Sawada, T.; Matsuno, H. Langmuir 2007, 23, 11127–11133. Sano, K.-I.; Sasaki, H.; Shiba, K. Langmuir 2005, 21, 3090–3095. Serizawa, T.; Sawada, T.; Kitayama, T. Angew. Chem., Int. Ed. 2007, 46, 723–726. Tretinnikov, O. J. Adhes. Sci. Technol. 1999, 13, 1085–1102. Kumaki, J.; Kawauchi, T.; Okoshi, K.; Kusanagi, H.; Yashima, E. Angew. Chem., Int. Ed. 2007, 46, 5348–5399. Serizawa, T.; Hamada, K.; Kitayama, T.; Fujimoto, N.; Hatada, K.; Akashi, M. J. Am. Chem. Soc. 2000, 122, 1891–1899. Date, T.; Yoshino, S.; Matsuno, H.; Serizawa, T. Polym. J. 2012, 44, 366–369. Wang, S.; Cui, W.; Bei, J. Anal. Bioanal. Chem. 2005, 381, 547–556. 46

33. Holland, T. A.; Mikos, A. G. Adv. Biochem. Eng. Biotechnol. 2006, 102, 161–185. 34. Matsuno, H.; Sekine, J.; Yajima, H.; Serizawa, T. Langmuir 2008, 24, 6399–6403. 35. Adey, N.; Mataragnon, A.; Rider, J.; Carter, J.; Kay, B. Gene 1995, 156, 27–31. 36. Serizawa, T.; Techawanitchai, P.; Matsuno, H. ChemBioChem 2007, 8, 989–993. 37. Tsutsui, K.; Tsujita, Y.; Yoshimizu, H. Polymer 1998, 5177–5182. 38. Tsutsui, K.; Katsumata, T.; Yamamoto, Y.; Fukatsu, H.; Yoshimizu, H.; Kinoshita, T.; Tsujita, Y. Polymer 1999, 40, 3815–3819. 39. Yamamoto, Y.; Nakai, Y.; Katsumata, T.; Tsutsui, K.; Tsujita, Y.; Yoshimizu, H.; Okamoto, S. J. Mol. Struct. 2005, 739, 13–30. 40. Burroughes, J.; Bradley, D. D. C.; Brown, A.; Marks, R.; Mackay, K.; Friend, R.; Burns, P.; Holmes, A. Nature 1990, 347, 539–541. 41. Yu, G.; Gao, J.; Hummelen, J.; Wudl, F.; Heeger, A. Science 1995, 270, 1789–1791. 42. Feng, F.; He, F.; An, L.; Wang, S.; Li, Y.; Zhu, D. Adv. Mater. 2008, 20, 2959–2969. 43. Ejima, H.; Matsuno, H.; Serizawa, T. Langmuir 2010, 26, 17278–17285. 44. Kurtz, S.; Devine, J. Biomaterials 2007, 28, 4845–4914. 45. Date, T.; Sekine, J.; Matsuno, H.; Serizawa, T. ACS Appl. Mater. Interfaces 2011, 3, 351–360. 46. Serizawa, T.; Fukuta, H.; Date, T.; Sawada, T. Chem. Commun. 2015, 52, 2241–2244. 47. Ejima, H.; Kikuchi, H.; Matsuno, H.; Yajima, H.; Serizawa, T. Chem. Mater. 2010, 22, 6032–6034. 48. Suzuki, S.; Sawada, T.; Ishizone, T.; Serizawa, T. Chem. Commun. 2016, 52, 5670–5673. 49. Love, J.; Estroff, L.; Kriebel, J.; Nuzzo, R.; Whitesides, G. Chem. Rev. 2005, 105, 1103–1169. 50. Onclin, S.; Ravoo, B.; Reinhoudt, D. Angew. Chem., Int. Ed. 2005, 44, 6282–6304. 51. Ulman, A. Chem. Rev. 1996, 96, 1533–1554. 52. Sano, K.-I.; Ajima, K.; Iwahori, K.; Yudasaka, M.; Iijima, S.; Yamashita, I.; Shiba, K. Small 2005, 1, 826–858. 53. Matsui, T.; Matsukawa, N.; Iwahori, K.; Sano, K. I.; Shiba, K.; Yamashita, I. Langmuir 2007, 23, 1615–1618. 54. Kashiwagi, K.; Tsuji, T.; Shiba, K. Biomaterials 2009, 30, 1166–1175. 55. Park, T.; Lee, S.; Lee, S.; Park, J.; Yang, K.; Lee, K.-B.; Ko, S.; Park, J.; Kim, T.; Kim, S.; Shin, Y.; Chung, B.; Ku, S.-J.; Kim, D. H.; Choi, I. Anal. Chem. 2006, 78, 7197–7205. 56. Matsuno, H.; Date, T.; Kubo, Y.; Yoshino, Y.; Tanaka, N.; Sogabe, A.; Kuroita, T.; Serizawa, T. Chem. Lett. 2009, 38, 834–835. 57. Zhu, X.; Zhao, X.; Burkholder, W.; Gragerov, A.; Ogata, C.; Gottesman, M.; Hendrickson, W. Science 1996, 272, 1606–1620. 47

58. Tanaka, N.; Nakao, S.; Wadai, H.; Ikeda, S.; Chatellier, J.; Kunugi, S. Proc. Natl. Acad. Sci. U. S. A. 2002, 99, 15398–15403. 59. Kuroita, T.; Sogabe, A.; Takarada, Y.; Tanaka, N. Protein Achieving Improved Blocking Efficiency. U.S. Patent 10/562,776, July 567, 2004. 60. Pasquato, L.; Pengo, P.; Scrimin, P. J. Mater. Chem. 2004, 14, 3481–3487. 61. Ghosh, S.; Pal, T. Chem. Rev. 2007, 107, 4797–4862. 62. De, M.; Ghosh, P. S.; Rotello, V. M. Adv. Mater. 2008, 20, 4225–4241. 63. Serizawa, T.; Hirai, Y.; Aizawa, M. Langmuir 2009, 25, 12229–12234. 64. Slocik, J.; Wright, D. Biomacromolecules 2003, 4, 1135–1141. 65. Lévy, R.; Thanh, N.; Doty, R.; Hussain, I.; Nichols, R.; Schiffrin, D.; Brust, M.; Fernig, D. J. Am. Chem. Soc. 2004, 126, 10076–10084. 66. Graf, P.; Mantion, A.; Foelske, A.; Shkilnyy, A.; Masić, A.; Thünemann, A.; Taubert, A. Chem. Eur. J. 2009, 15, 5831–5844. 67. Lundqvist, M.; Nygren, P.; Jonsson, B.-H.; Broo, K. Angew. Chem., Int. Ed. 2006, 45, 8169–8173. 68. Zhang, J.-J.; Zheng, T.-T.; Cheng, F.-F.; Zhu, J.-J. Chem. Commun. 2011, 47, 1178–1180. 69. Pecher, J.; Mecking, S. Chem. Rev. 2010, 110, 6260–6279. 70. Ejima, H.; Matsumiya, K.; Sawada, T.; Serizawa, T. Chem. Commun. 2011, 47, 7707–7709.

48

Chapter 4

DNA Condensed Phase and DNA-Inorganic Hybrid Mesostructured Materials Yuanyuan Cao and Shunai Che* School of Chemistry and Chemical Engineering, State Key Laboratory of Metal Matrix Composites, Shanghai Jiao Tong University, 800 Dongchuan Road, Shanghai, 200240, China *E-mail:[email protected]

The compaction and decompaction processes of DNA compose the crucial part in cell life cycle. The research on those compacted phases of DNA derives from both biology and soft matter science, which promotes the development of nucleic acid research and modern nanotechnology, further brings real benefits for human such as the improving of gene therapy. In this chapter, we have first introduced the organic soft matter that originates from the self-assembly of DNA superstructures, as well as several kinds of methods for inducing condensed phase of DNA. Then we have focused on the DNA templated mesostructured assemblies for guiding the formation of inorganic mesoporous materials. The fabrication of those materials main achieved through self-assembly process combining the artificial biomineralization methods, which will promote the designing of novel optical or electric materials.

The “soft matter” field, which mainly concerns polymers, colloids, surfactants, liquid crystals, biomolecules and molecular assemblies, is an exciting and fast-developing field. It is an interdisciplinary subject that utilizes aspects of physics, chemistry and materials science as well as biochemistry or engineering in specific cases. Soft matter has typically been distinguished from traditional “hard materials” through differences between their macroscopic mechanism properties. However, today, the classification is not simply restricted to macroscopic properties but is also considered from a kinetic energy viewpoint regarding the © 2017 American Chemical Society

intermolecular interactions and arrangements of the components, in which the ordering is generally intermediate between that of a crystalline solid and a liquid. Owing to recent developments in the soft matter field and biology, biological soft materials, which have existed in nature over billions of years, have rapidly gained more attention than ever before. Although biological soft matter systems appear more complex than artificial material systems, concepts from polymer physics, physical chemistry and colloidal science as well as surface chemistry have yielded great insights into organism formation mechanisms, such as fibril building, protein crystallization or membrane fluidity. In turn, the concepts from nature have also inspired rapid developments in soft materials, examples include constructing complex structures from assemblies of biopolymers or exploiting nanotechnology to make devices based on the self-organization of polymers (1). Biology related soft matter research has mainly focused on studying comprehensive living systems and the cross-interactions between the different kinds of components, such as proteins, nucleic acids, polysaccharides and lipids, within them. The basic role of DNA, or deoxyribonucleic acid, is to store huge amounts of genetic code and to provide the blueprint for the machinery of life: proteins. When DNA replicates itself, sophisticated information transcription and compaction/decompaction processes occur that are highly efficient and fast and involve cooperative interactions with other biological counterparts that are still not completely understood. The compaction process of DNA in vitro and in vivo can be considered as a process that exists at the cross section between soft matter science and biology, which can promote both the development of nucleic acid research and modern nanotechnology. For instance, the recently popular concept of gene therapy offers a fundamental approach for treating inherited or acquired diseases, which requires introducing specifically engineered genes into a patient’s cells to act as a drug. This technology requires a deep understanding of how to collapse extended DNA chains into compact, orderly particles or to package DNA into vectors for delivery, which can enter the target cell and quickly decompose (2, 3). All of these processes demand understanding the condensed structures of DNA. However, since studying the behavior of DNA in vivo needs rather complex observation methods, research on DNA condensation in vitro using an artificial system is considered to be a smart and reliable substitute. In this chapter, instead of discussing single molecular conformations at the atomistic level, we introduced ordered soft matter that originates from the self-assembly of DNA superstructures and aggregates, and we especially mentioned its chiral features, which are a vital part of DNA. Then, starting from materials science, we will mainly introduce how the DNA mesostructured assemblies work as templates for guiding the formation of inorganic materials through the biomineralization process.

1. Short Introduction of DNA DNA is a typical chain-like polymer composed of polymerized nucleotide monomers. It is composed of two polyester chains made up of alternating sugar 50

(D-2-deoxyribose) and phosphate groups. Each sugar group is attached to four different kinds of nitric bases, two major purine bases (adenine (A) and guanine (G)) and two major pyrimidine bases (cytosine (C) and thymine (T)), which code genetic information through an infinite number of different permutations. Hydrogen bonds between the specifically paired pyrimidines and purines (called base pairs) connect and stabilize the two spiral chains, causing them to twist into a double helix to form the three-dimensional secondary structure of DNA. The typical, well-known, B-conformation DNA, which was discovered in 1953 by Watson and Crick (4), is a right-handed double helix with a diameter of 2.2 nm and an approximately 3.4 nm helical pitch composed of 10 base pairs; a schematic of the double helix is shown in Figure 1 (5). Although the B-type secondary structure is the most stable conformation under physiological conditions, it is not the constant conformation of DNA. The conformation changes from the B-type to the right-handed duplex A-type or the left-handed duplex Z-type depending on variations in the external solution environment, such as the ionic strength, solvent and temperature (6, 7), or the base pair sequences (8).

Figure 1. The structural parameters of the B-type DNA double helix in solution. An important aspect that should be considered when studying the behavior of DNA in solution is its intrinsic polyelectrolyte qualities. The peripheral phosphate groups on the polyester chains endow the backbone with two negative charges per base pair. The pKa of the phosphate groups of DNA is approximately 1; thus, the backbone can be fully charged at pH 7 (However, the phosphomonoester groups on the ends of the chain have a second pKa value of approximately 6) (9). The charges on biomolecules have great physiological significance. They can prevent homologous intermolecular interactions by electrostatic repulsion and form desired complexes with oppositely charged molecules. For example, electrostatic interactions are mainly attributed to the formation the nucleosomes when DNA wraps around histone. The formation of this kind of complex is also affected by the chain flexibility of the DNA molecules. The flexibility parameter that describes the ability of linear 51

molecules to bend locally is the persistence length, which is also known as the orientational correlation length. The persistence length of DNA is approximately 50 nm, as measured from various methods, and depends on the salt concentration in the range of 1-2 mM NaCl (10–12). The flexibility of DNA dramatically changes when its length reaches this threshold value. When DNA is shorter than the persistence length, it is rather stiff and rodlike, while it will behave as a wormlike chain when the length reaches a few persistence lengths. Additionally, the packing of DNA inside a viral capsid is affected by the persistence length.

2. The Mesostructured Packing Phase of DNA 2.1. DNA Assemblies Because DNA plays the role of a carrier for a huge amount of genetic information in a very limited space within a cell, the in vivo packing of DNA, which may be several meters long, is usually very tight. Except for the highly hierarchical DNA storage chromatin in eukaryotic cell nuclei, in most extrachromosomal organisms, such as in sperm heads, virus capsids, and bacterial nucleoids, the volume concentrations of DNA can reach 70% W/V in cells (13–15). The packing is highly efficient and dense but remains accessible for transcription or replication; thus, DNA undergoes alternative condensation and decondensation processes during the cell cycle. The efficient packing ordering requirements and the need for the fluid structure to remain accessible require DNA to under liquid crystal-like organization (16). The formation mechanisms of the liquid-crystalline phases of concentrated solutions of DNA may be related to the tendency of semi-rigid polymers to form liquid-crystalline phases during concentration processes. In 1961, Robinson first reported on the possible liquid crystal phase of DNA in vitro as a supplemental study that was adjunct to a phase study on the liquid crystal behavior of PBLG (poly-γ-benzyl-L-glutamate) and suspected that the phase was of the cholesteric type (17, 18). Later, short DNA fragments with lengths corresponding to the persistence length were applied to study multiple liquid-crystalline phases, as these rigid fragments rule out the entanglement problems that within long DNA molecules (19). Simply by increasing the concentration, isotropic DNA can form multiple liquid-crystalline phases from chiral liquid phases, such as the blue phase (also studied as the precholesteric phase) and the cholesteric phase, to achiral liquid-crystalline phases, such as the 2D columnar hexagonal phase, then to true crystals, such as the 3D hexagonal and 3D orthorhombic phases. The schematic models of these phases are shown in Figure 2 (20, 21). Although it is difficult to obtain well-organized liquid-crystalline phases using long DNA, it appears that the polymer length does not alter the nature of the liquid-crystalline phases, which means that these phases may also appear with long DNA but require time to organize and stabilize. However, the length of the polymer can affect the defects and phase transitions. 52

Figure 2. Schematic representation of the structures of multiple liquid-crystals: (A) Blue phase building blocks with a right-handed double-twist cylinder; (B) Right-handed chiral cholesteric phase; (C) 2D columnar hexagonal structure; (D) 3D hexagonal structure; and (E) 3D orthorhombic phase. Reproduced with permission (21). Copyright 2015, Wiley-VCH Verlag GmbH & Co. KGaA.

The surface charge of DNA strongly contributes to the attraction between these rod-like polyelectrolytes. Kornyshev et al. (22) developed an “electrostatic zipper” model to explain the dense packing and counterion specificity of DNA condensation. Strips composed of positively charged counterions adsorbed in the grooves and negatively charged lines of the phosphates allowed oppositely charged groups to approach to form DNA-DNA contacts, creating a “zipper” that pulls the molecules through electrostatic interactions and fastens the molecules together. The zipper interaction model has given rise to a special packing phase, the 2D square phase, which is rarely found naturally (23). The 2D square phase appears in a narrow concentration range and has a loose packing arrangement with a d-spacing ratio of √2/1, which is different from the spatial confinement packing of noncharged rods. This special mesostructure occurs under subtle charge balance conditions, which we will discuss in detail in the following section. 2.2. Chiral DNA Assemblies 2.2.1. The Cholesteric and Blue Phase The chirality of DNA affects not only the secondary double helix structure but also the packed tertiary structure. Considering the geometry of DNA, the helical shapes affect the packing behavior of rodlike DNA under confinement conditions. Two helices usually prefer to align at an angle instead of parallel to each other because the helices fit well into each other’s grooves. When DNA is not present at a very high concentration, the formation of cholesteric (single-twist plywood) and blue phases (double-twist cylinders) originates from spontaneous twisting. Right-handed B-DNA usually forms the left-handed cholesteric phase in the absence of other disturbances. This is the case for long DNA molecules; however, for ultrashort DNA in the range of 8-20 bp, spontaneous right-handed chiral phases can be obtained (24). In the cholesteric phase, DNA molecules are aligned in parallel, and their orientation rotates continuously along a direction that is perpendicular to the helical axis of the stacking layers. The helical pitch usual varies from 1.5 to 3.5 μm (25). One should notice that the phase is continuous, and the term “layer” is not a real plane but is simply used to clarify the stacking. 53

Although the steric interactions between the chiral shapes favor the formation of the right-hand cholesteric phase, the strong repulsion interactions from the helical-distributed electrostatic charges on the surface of the DNA induce a chirality inversion into a left-handed structure. Thus, extrinsic factors that affect the geometrical parameters and electronic charges of DNA, such as groove binders, intercalators or polycations, greatly affect the cholesteric pitch (26). Even the sign of the cholesteric phase can be changed by adding divalent cations, such as the right-handedness that is induced by the presence of Mg2+ and Ca2+ (27, 28). In higher DNA concentrations, DNA forms either 2D or 3D hexagonal phases. Dense packing prevents spontaneous twisting between the molecules, and the resulting structures originate from the competition between the chiral order and dense packing constraints (29). Within a special given concentration range, the chiral cholesteric and achiral hexagonal phases may coexist because of phase transitions in which slight untwisting occurs in the cholesteric structure. The two-fold symmetry axes of the hexagonal phase become parallel to the helical axes of the cholesteric phase. During the transition from the isotropic phase to the cholesteric phase of DNA, the blue phase has been observed within a narrow concentration range. In this phase, the DNA molecules assume a double-twist structure within small cylindrical domains that are 40-200 nm wide and 800 nm long (30). The helical pitch length in the blue phase, which was calculated as 800 nm, is much smaller than that of the cholesteric phase. The double-twist cylinders can further densely assemble into a long-range BP I phase with body-centered cubic symmetry or a loose random BP III phase.

2.2.2. Circular Dichroism (CD) of Chiral DNA Aggregation An important feature of the chiral condensed phase of DNA is the presence of anomalous circular dichroism (CD) signals, which are tens to thousands of times higher than those of dispersed DNA molecules. The spectrum obtained within the absorption band of DNA is called psi (polymer-and-salt-induced, also written as ψ)-type CD. These signals are produced because of induced oscillating dipole moments that are coupled over all of the particles and the generation of collective excitation modes, which are the eigenmodes of the particle in which the shape depends on the long-range internal organization within the particles (31). In this situation, long-range chirality permits “spatial-resonance” between the appropriate handed circular polarization with very efficient energy exchange, inducing extremely strong circular dichroism signals. In addition, the shape of the spectrum usually has long tails outside the usual absorption band of DNA that extend toward the red wavelengths because of the preferential scattering effect of the long-range chiral aggregates from one circular polarized light (14). Typical CD spectra comparing individual B-DNA prior to and after condensation is shown in Figure 3 (32).

54

Figure 3. CD spectra of B-DNA prior to and after condensation with PEG. Reproduced with permission (32). Copyright, 1991, American Chemical Society.

3. Organic DNA Assemblies There are multiple ordered condensed phases of DNA, that behave as large-scale mesophases or 3D crystals with distinct morphologies. Compaction can occur either as a result of intermolecular aggregation or as a result of self-condensation or intramolecular aggregation of a single DNA molecule. In solution, these phases can be obtained using various types of methods: a) DNA can be confined into a restricted volume by preparing concentrated water solutions of DNA in the presence of monovalent ions and controlling the concentration by the adjusting the amount of added water. b) DNA can interact with multivalent counterions, such as polyamines, cationic polypeptides polylysine or histone H1, biologically derived liposomes or cationic lipids. c) DNA can be condensed by applying osmotic pressure or through the presence of a dehydrating agent, such as a neutral polymer (polyethylene glycol (PEG)) or ethanol. These condensates generally have orderly, toroidal or rodlike shapes and sizes similar to those of DNA that is gently lysed from phase heads. According to the aggregation process, the resulting structures have been referred to as polymer-and-salt-induced DNA (ψ-DNA), which we introduced before. The in vitro condensation of DNA is both biologically and physically interesting. This process can be regarded as a model for DNA packing within intracellular structures from a biological viewpoint and a typical example of a polymer self-assembly process from a physical viewpoint (33). In addition, the in vitro condensation of DNA inspires gene therapy as an alternative to using recombinant virus vectors and provides new structures for materials design.

55

3.1. DNA Condensation and the Theory of Counterion Condensation Since DNA can be regarded as a polyelectrolyte with a charge density of 1 e/1.7 Å (2), DNA condensation prefers to the electrostatic attraction between the negatively charged phosphate groups of the outer polyester chains of DNA with a cationic species. A fraction of counterions is bound to the DNA chains because of strong electrostatic attraction and reduces the electrostatic repulsion between the polyelectrolyte chains, which means that the effective charge becomes smaller than that calculated using stoichiometry. This phenomenon is called counterion condensation (also known as Manning condensation). The Manning theory has been found to be consistent with experiments in that it predicts the approximate constant value of the degree of neutralization required to collapse DNA for various ionic conditions. According to the Manning model, the charge density parameter ξ of DNA is 4.2, and the fraction of charges neutralized can be simply expressed as:

which reveals that, on average, three quarters of the bases are neutralized when condensation occurs. Wilson and Bloomfield developed the Manning theory to determine the extent of DNA charge neutralization using multivalent cations and found that the collapse of DNA occurred at ~89% phosphate charge neutralization using polyamines in aqueous solutions. When the charges become neutralized and counterions are released, the DNA chains begin to partially collapse (1).

3.2. DNA-Polyamine Complex The condensation or aggregation of DNA is a continuous cycle for vital cellular processes and has acquired considerable importance in recent years as a model system to analyze DNA-related metabolism and to promote research on gene uptake by cells for gene therapy. The terms condensation and aggregation describe similar phenomena but have a few differences. Condensation is the formation of individual discrete particles with a distinct morphology, and aggregation is the formation of larger and possibly more amorphous accumulations of molecules (33). According to the theory mentioned above, it is clear that one approach to induce the condensation and aggregation of DNA is to screen negative charges with the phosphate groups of DNA by adding multivalent cations, such as naturally occurring polyamines, cationic polypeptides, proteins (such as histone), inorganic metal salts Co(NH3)63+, polyamino lipids and cationic lipids (34–39). 56

3.2.1. Condensation by Polyamines Polyamines are ubiquitous components of cells and are involved in many fundamental biological processes in organisms, such as cell growth and differentiation (40, 41). These aliphatic polycationic compounds possess multiple positive charges at physiological pH levels because of the protonation of their amine groups, which endows them with a high affinity for the acidic constituents of cells, such as nucleic acids, acidic proteins, phospholipids and ATP (35). Naturally extracted spermidine (3+), spermine (4+) and diamine putrescine are the most commonly used polyamines for compressing DNA (Although from the concepts of polymer science, it is more accurate to refer to them as oligoamines). They bind to the phosphate groups of DNA and reduce the repulsion interactions to induce the collapse (occurring in single DNA molecule) or aggregation (occurring between more than one DNA molecule) of long DNA molecules in the form of rods or toroids. The toroids formed from the collapse of isolated long DNA chains exhibit local hexagonal packing (Figure 4A) (42). For the short DNA fragments with a 146 bp length (within the persistence length range), the aggregates are not restricted to microscopic domains, extend over long-range distances and are liquid crystalline in the cholesteric or hexagonal phase (43). However, the addition of polyamine cations to DNA solutions first precipitates the DNA, but a further addition resolubilizes the DNA aggregates (44). Livolan et al. (45) determined the concentration thresholds between DNA and polyamine for DNA precipitation and resolubilization, using the “ion-bridging” model based on electrostatic bridging. Furthermore, phase diagrams (35, 46) have been developed to clarify these processes, in which three regimes of DNA concentrations are identified (a schematic representation is shown in Figure 4B). Other factors that influence the condensation process are the properties of the amino species. Thomas et al. (47) tested DNA precipitation/resolubilization phenomena using various kinds of natural and synthetic polyamines. Both the increasing of cations and the variation of structure, especially substituting polyamine, would exert remarkable effects on the precipitate and the of polyamines to resolubilize the DNA. PEI (polyethylenimine) is another kind of the classic counterion polymer that is commonly applied for nonviral gene delivery (48). It is a water soluble polymer with highly positive charges, and every third atom nitrogen can be protonated in a broad range of pH values. Both branched and linear PEI can provide strong electrostatic interactions that lead to the partial condensation of the normally large hydrodynamic volume of DNA and are supposed to be among the most transfection efficient nonviral vectors in vitro and in vivo. The intrinsic properties of PEI, such as its molecular weight and conformation, affect the properties of the formed DNA/PEI complexes and the subsequent cell delivery efficiency. When the metal ions are introduced into the polymer system, the produced polymeric metal complexes are effective to promote the DNA condensation and increase the gene transfection efficiency (49). The polymeric metal complexs synthesized from Fe3+ ions or Zn2+ ions chelating with PEI or imidazole group containing polymers show a improvement of the condensation and transfection efficiency (50, 51). 57

Figure 4. (A) Cryo-electron micrograph of λDNA toroids with DNA fringes visible around nearly the entire circumference of the toroid and simulated TEM images of a model toroid constructed from a single continuous path. (B) Schematic representation of the phase diagram for the precipitation of DNA fragments in the presence of spermine. The precipitation domain, in which the dense precipitate separates from the dilute supernatant, is limited by the Cprecip and Credissol curves. In this representation, each experimental point is defined by the DNA concentration (CDNA phosphate) and the spermine salt concentration (Cspermine). Reproduced with permission (35, 42). Copyright 2001 National Academy of Sciences. 2005, the Biophysical Society.

3.2.2. DNA “Beads-on-String” Nucleosome-like Structure In somatic eukaryotic cells, DNA is associated with cationic histone proteins to form the highly periodic structures of nucleosomes. Nucleosomes consists of 147 bp DNA wrapped around a 7 nm cationic octamer of histone proteins, which further hierarchically assemble into chromatin fibers that are confined in a limited volume. Those complex structures are still a matter of debate, and many articles have reported the exploration of their structures using microscopy methods (52–54). Another approach to study the compaction of genomic DNA 58

in chromatin is to artificially wrap DNA around charged hard spheres to yield the so-called “beads-on-string” structures found in chromatin, to investigate the detailed structure and formation mechanisms using an analogous system. Since histone octamers are disc-shaped nanostructures with positively charged binding sites, polymers that have charged hard spherical cores have been considered as protein substitutes. Chen et al. (55–57) have used polyamidoamine (PAMAM) dendrimers to mimic histone proteins. The complexes of DNA with the dendrimers exhibited three distinct nanostructures characterized by different degrees of DNA bending (Figure 5A), which were revealed using synchrotron small angle X-ray scattering (SAXS). When the dendrimers were highly protonated, “chromatin-like fibers” composed of “nucleosome-like” units were formed (Figure 5Aa3). A further higher-ordered assembled “beads-on-string” structure was achieved using block-copolymer polyethylene glycol-b-poly-4-vinylpyridine (PEG-b-PVP) micelles with an inert shell and a positively charged core (58). The micelles preorganized in the “beads-on-string” structure and further self-assembled along the strings into core-shell structured solenoidal nanofibers (Figure 5B). The two-stage assembly process of the DNA/micelles was a high fidelity mimic of the complex assembly mode of chromatin, proving that the synthetic macromolecule was very effective for mimicking the natural process (59).

Figure 5. (A) Schematic illustrations of the three types of nanostructures composed of DNA-PAMAM complexes. The “beads-on-string” structure was formed when the dendrimer was in a highly protonated state (a3). (B) The hierarchical self-assembly process of linear DNA with PEG-b-PVP core-shell micelles. Reproduced with permission (55, 58). Copyright 2010, Royal Society of Chemistry. 2012, Wiley-VCH Verlag GmbH & Co. KGaA. 59

3.3. DNA-Amphiphile Complexes One of the problems using nonviral chemical vectors for gene delivery is to facilitate the uptake of nucleic acids through the cellular membrane. Regarding this problem, investigations have been conducted to induce DNA condensation using membrane analogue systems, such as artificially synthesized surfactant molecules or naturally extracted lipids and their derivations. Mixtures of DNA with these amphiphilic molecules provide an important gene transfection strategy and clues for investigating the interactions between DNA and membranes.

3.3.1. DNA-Surfactant Complexes A surfactant or surface-active agent is an amphiphilic molecule that is active at the surface between two phases. It contains both hydrophilic (head) and hydrophobic (tail) parts that can accumulate at the interface between hydrophilic and hydrophobic phases and modify the surface tension. The hydrophobic part is mainly composed of linear or branched alkyl long chains, and the hydrophilic part, also called the polar part, can comprise cationic, anionic, and zwitterionic functional groups. Structures of some typical surfactants are shown in Figure 6.

Figure 6. Chemical formulas of three kinds of surfactants: a cationic surfactant, cetyltrimethylammonium bromide (CTAB); an anionic surfactant, sodium lauryl sulfate (SDS); and a zwitterionic surfactant, dodecyldimethylamine oxide (DDAO). The formation of DNA-surfactant complexes is achieved through attractive Coulomb interactions, and they are stabilized by hydrophobic interactions between the hydrophobic moieties of the surfactant molecules (60). The structures of the DNA-surfactant complexes mainly depends on the surfactant that is used (61). 60

The packing parameter (g) of the surfactant; which is the relationship between the head area (a), the extended length (l), and the volume of the hydrophobic part (v), g=v/(lmaxa); determines the physicochemical behavior of the surfactant in water. With the increasing of g value, the surfactants form different assembled structures, from micelles, hexagonal and lamellar to cubic and reversed hexagonal phases to reversed micelles (62). Owing to the negatively charged phosphate DNA backbone, it is common to use cationic surfactants as the complex inducing amphiphiles, and cetyltrimethylammonium bromide (CTAB) is mostly used. The binding between the negatively charged phosphate groups of DNA and the positively charged polar head of a surfactant provides the driving force. Depending upon the preferred shape of the surfactant self-assembly, these complexes can exist in a variety of mesoscopic structures including lamellar (LR) structure, in which the DNA molecules are sandwiched between surfactant bilayers, inverse hexagonal (HII) structure, in which the DNA are confined to the aqueous cores of the micelles, and normal topology hexagonal (HI) structure, in which each DNA is surrounded by cylindrical micelles of DNA (63–66). The schematic models of these structures are show in Figure 7. These complexes show complex phase behavior that depends on multiple factors in addition to the surfactant structure. Additionally, phase transfer can occur within the CTAB-DNA complex by changing the concentration of the cosurfactant or DNA (65, 66).

Figure 7. Schematic representation of (A) the lamellar phase (LR), (B) inverted hexagonal phase (HII) and (C) intercalated hexagonal phase (HI) of a DNA-surfactant complex. Reprinted with permission (65). Copyright 2004 American Physical Society.

It is clear that negatively charged anionic surfactants do not have a significant effect on the conformation behavior of DNA, if the DNA concentration is not sufficiently high. In contrast, zwitterionic surfactants, which exist either in a neutral or cationic protonated form at different pH values, have been proven to have remarkable interactions with DNA. The conformational behavior of linear DNA in the presence of a zwitterionic surfactant, dodecyldimethylamine oxide (DDAO), was examined (67). The positively charged DDAO in the vesicular form behaved as a more efficient DNA-condensation agent than that in the micellar form. 61

3.3.2. DNA-Lipid Composition Lipids are one of the most important components in organisms. They are responsible for regulating the passage of biomolecules between cells and within a cell between organelles. Usually, they can be divided into two types: fat-like lipids, which contain ester linkages and long alkyl chains and can be hydrolyzed; and cholesterol- or steroid-like lipids, which do not have ester linkages and cannot be hydrolyzed. The cationic lipids discussed here mainly comprise three basic parts: a charged hydrophilic head group attached to a hydrophobic tail via a linker group, such as an ether, ester, or amide. Their structures are similar to artificially synthesized surfactants but with more complex components as they are often naturally derived. Three kinds of commonly used cationic lipids are listed in Figure 8.

Figure 8. Chemical formulas of three cationic lipids. The co-assembly of DNA and lipids has broadened the investigations on DNA-membrane interactions. Similar to the assembly behavior of surfactants, various structures can be obtained using different lipid structures, including hexagonal (HI), lamellar (LR) and inverse hexagonal (HII) structures. Safinya’s group produced multilamellar-structured CL (cationic liposome)-DNA complexes with alternating lipid bilayers and DNA monolayers by mixing DNA with LCs consisting of DOPC (dioleoyl phosphatidylcholine) and DOTAP (dioleoyl trimethylammonium propane) (68). Then, adding DOPE (dioleoyl phosphatidylethanolamine) lipid or hexanol induced a transition from the lamellar to hexagonal phase owing to the natural curvature (Figure 9a, pathway I) or 62

by reducing the membrane bending rigidity (Figure 9a, pathway II) (69). In this phase, the DNA coated with cationic lipid monolayers was arranged on an inverted 2D hexagonal lattice, which is more efficient for gene delivery.

Figure 9. (a) Schematic of two distinct pathways for DNA-lipid complex formation from the lamellar phase to the columnar inverted hexagonal phase with the addition of the DOPE lipid (pathway I) or the addition of helper lipids consisting of mixtures of DOPC and a hexanol cosurfactant (pathway II). (b) Model of the coexistence of two packing modes in the DNA-DOTAP-DOPE system. Most of the DNA is embedded within the lipid columnar inverted-hexagonal assembly, and short, unbound segments of the DNA converge into a condensed toroidal Ψ–structure with a left-handed orientation. Reprinted with permission (69, 70). Copyright 1998, Science. 1999, Federation of European Biochemical Societies.

Chiral DNA packaging was also achieved through the interactions between DNA and cationic liposomes. Zuidam (70) found the B-to-C secondary conformational transition of DNA molecules upon binding to DOTAP, as analyzed using CD spectra. DOTAP-DOPE liposomes affected the collapse of DNA into a tightly packed cholesteric-like phase with long-range chiral order, which coexisted with the inverted hexagonal phase in the DNA-liposome complex; the model is represented as a schematic in Figure 8 b. Electronic interactions were not essential for the formation of the DNA-CL assemblies; the nonionic lipids could also have confined the DNA within the 63

reverse hexagonal columnar phases through hydrogen bonds between the polar heads of the lipids and DNA (71). These conditions easily release DNA into excess water. These complexes were able to mimic the gene delivery ability of natural viruses by carrying extracellular DNA across outer cell membranes and nuclear membranes as a nonviral method. The DNA associated within the aqueous channels of the inverse hexagonal (HII) lipids can be actively transcribed by the RNA polymerase (72, 73), and the liquid crystal structure remains during transcription, which demonstrates the potential of using these complexes for transporting genome information. In addition, the DNA-lipid complexes can be prepared as organic-solvent soluble solid films using either cast-stretching methods or the Langmuir-Blodgett (LB) method (74). In the casting method, DNA strands are easily aligned along the stretching direction by casting an organic solution of the DNA-lipid complex onto a substrate or using the hot-press process. In the LB method, the DNA strands are transferred with lipid monolayers at the air-water interface and aligned along the vertical dipping direction. These simple DNA-lipid film preparation methods, which were initially developed by Okahata et al. (75), have promoted the development of DNA-derived functional materials, such as photonic and electronic nanomaterials. 3.4. Dehydrating Agents for Inducing DNA Compaction Aside from the ability of multivalent cations to undergo electrostatic interactions that induce DNA condensation, dehydrating agents that expel DNA single molecules from the water phase have also been proven to efficiently induce the transition of DNA from random coil to compact globule states. Both neutral polymers, such as polyethylene glycol (PEG), poly(2-vinylpyrrolidone) (PVP) and polyacrylamide (PAAm), and poor solvents for DNA, such as ethanol, can be regarded as dehydrating agents.

3.4.1. Neutral-Polymer Induced DNA Compaction In 1971, Lerman (76) discovered that in the presence of over-threshold concentrations of simple neutral polymers (polyethylene glycol (PEG)) and salts, DNA molecules collapse into particles, approaching a compactness similar to that of phage heads. Since then, this strategy, also called crowding-induced DNA condensation, has been studied as a model of DNA condensation in vitro (77–82). The physical mechanism of this phenomenon is caused by depletion interactions between the solution components, usually between different polymers that lack net attractive interactions. Their competition for the solvent space results in phase separation in solution (83). Specific to DNA and PEG, the contacts between these two molecules are considered to be thermodynamically unfavorable. Upon reaching critical concentrations of PEG, the solvent quality for DNA becomes poorer. Therefore, the available free space for the unfolded DNA in solution decreases; thus, DNA undergoes a collapse transition (78). However, while the concentration of PEG is critical for determining the condensation, the degree of 64

polymerization and the salt concentration also have a great influence. In addition, analogous to the behavior of polyamines, the condensed DNA can unfold using higher concentrations of PEG (84). These packing/unpacking transitions can be used as a general method for studying the optical properties of the condensed phase of DNA. Jordan et al. (85) developed a volume-exclusion method by adding PEG dropwise into a dilute solution of DNA. The phase segregation was characterized by the appearance of psi-type circular dichroism, which indicated a tertiary structure transition. Similar to using the PEG neutral polymer, a nonionic surfactant solution of polyoxyethylene octylphenyl ether (Triton X-100) with a 50%-90% concentration could also induce the collapse of DNA, which exhibited a discrete coil-globule transition with the increasing concentration (77). The increase in the osmotic pressure of the Triton X-100 solution was attributed to inducing the compaction of the single DNA molecules. In these model systems, the neutral polymers mimic the intracellular globular proteins that do not directly bind to DNA (86). In a crowded medium, the chain molecule can be compcted in a crowded medium, as the depletion forces between monomers would compete with their excluded-volume interaction. This phenomenon is analogous to the chromosome organization in bacterial cells (87). Recently, a polymeric crowder dextran was induced the continuous compaction of DNA in a confined space and a abrupt tranition in a tube-like space (88, 89).

3.4.2. Poor-Solvent-Induced DNA Compaction DNA condensation using a variety of poor solvent conditions is another way to introduce dehydrating agents to induce DNA compaction. The intramolecular association between DNA segments is greater in a poor solvent than in a good solvent. When the density of DNA is high and the frequency of biomolecular encounters is substantial, DNA tends to associate together to form a more concentrated phase. Ethanol is the most commonly used poor solvent to induced phase transitions in DNA. In 1976, Lang et. al (90) directly observed concentrated particles of DNA after using an ethanol treatment through electron microscopy. This method can produce interesting condensed phases of DNA without requiring the presence of other artificial influences, such as proteins and cationic ions. Therefore, this system provides an idea model for analyzing the optical properties of condensed DNA (79, 91, 92). Minsky (93) analyzed the chiral compaction of long linear DNA after the DNA was exposed to a dehydrating EtOH/water mixture under relatively high ionic strength conditions. In this situation, the DNA molecules collapsed into cholesteric-like rod-shaped aggregates in which the DNA folded into parallel arrays. The formation of ordered tertiary DNA enabled efficient longrange interhelical coupling between the nucleotide chromophores, which exhibited nonconservative CD (Figure 10A). Increasing the ionic strength by modulating the salt concentration resulted in a conformation change in the condensed phase from a right-handed handedness to a long-range left-handed handedness through a nonchiral “nematic-like” mesophase, as shown in Figure 10. 65

Figure 10. (A) CD spectra exhibited by DNA molecules treated with a dehydrating agent (EtOH, 35% v/v) in increasing NaCl concentrations of (1) 0 M; (2) 0.4 M; (3) 0.8 M; (4) 1.2 M; (5) 1.6 M; (6) 2.0 M; and (7) 2.4 M. (B) Schematic representation of chiral rod-like packed DNA molecules and the conformation transition from a right-handed to a left-handed superhelix. Reproduced with persmission (93). Copyright 1998, Wiley-VCH Verlag GmbH & Co. KGaA.

4. Inorganic Mesostructured DNA Assemblies The successful production of mesoporous MCM-41 by Mobil company in Japan in 1992 created a new field for constructing highly ordered mesostructured materials (94, 95). Ordered pore arrangements of mesoporous materials can be achieved using the cooperative assembly of organic template molecules and inorganic sources based on soft-templating synthesis concepts. Although this field started by using artificial soft-templates, such as synthesized surfactants, various kinds of natural biomolecules, such as peptides (96), proteins (97), polysaccharides (98) and even viruses (99, 100), have recently been utilized as templates. Among them, DNA has long been viewed as a soft-template candidate because of its interesting and diverse self-assembly behavior. The fabrication of DNA-templated inorganic materials produces different kinds of mesostructures with various morphologies and structures and provides a better understanding of evolutionary processes, which are difficult to observe directly using organic materials. Although many metal or metal-compound nanomaterials, such as quantum dots, CdS (101) and magnetic Fe3O4 (102), have been constructed using DNA molecules as templates through a biomineralization-like process (103), here, we are mainly concerned with silica inorganic materials with highly ordered mesostructures, in which the production procedure involves both self-assembled condensation and biomineralization processes. 66

4.1. DNA-Silica Nanofibers The silica biomineralization of DNA is achieved through the efficient interactions between DNA molecules and the silica source, followed by the hydrolyzation and condensation of the silica species. However, biomimetic silica replications templated by DNA molecules were largely unsuccessful before 2004. In the pH value range that allowed the DNA to maintain its original secondary conformation, the silicate species maintain negative charge, producing electrostatic repulsion with the negatively charged DNA molecules. To conquer the problem, an ingenious strategy was applied by Shinkais et. Al (104), which induced the biomineralization process of DNA by transforming DNA from an anionic into a cationic species in the presence of a two-headed bridge molecule. Rod-like and circular-like silica nanomaterials were obtained using plasmid DNA as a template during a sol-gel polycondensation, as shown in Figure 11.

Figure 11. A) Schematic representation of the transformation of the DNA surface. B) TEM pictures of the obtained Rod-like (b1-b6) and circular-like silica nanomaterials (b7, b8). Reproduced from persmission (104). Copyright 2004, Wiley-VCH Verlag GmbH & Co. KGaA. Che’s group initiated a co-structure directing route to efficiently achieve the biomineralization of DNA molecules as well as their assemblies (105). Typically, functionalized positively charged organic silicane is applied as the CSDA (co-structure directing agent) to electrostatically interact with the negatively charged phosphate groups of the DNA backbones and to co-condensed with the silica source to form the silica framework (Figure 12A). When using N-trimethoxyl-silylpropyl-N,N,N-trimethylammonium chloride (TMAPS) as the CSDA, pore-structure-tunable mesoporous fibers with chiral, ring and ordered nanochannel arrays were transcribed from flexible long DNA molecules and (Figure 12B). The choices on CSDA had a great influence on the assembly mode of the biomolecule; changing TMAPS into 3-aminopropyltrimethoxysilane (APS) would lead to a highly twisted arrangement of DNA molecules because of the inherent chiral properties of DNA (106). The resultant multi-helical mesoporous fibers exhibited three levels of chirality: a primary DNA double 67

helix, a mesostructured secondary DNA left-handed orientation and a tertiary microscopic right-handed helical orientation with a twisting-thread morphology (Figure 12C).

Figure 12. (A) Schematic representation of the CSDA route for inducing DNA mineralization. The negatively charged DNA molecules electrostatically interact with the positively charged TMAPS. Then, the co-condensation between TMAPS and the silica source TEOS forms the silica wall that wrapping around the DNA molecules. (B) TEM images of the DNA-silica fibers synthesized using TMAPS as the CSDA. The image contrast shows the inner channels casted from parallel (b1) or twisted (b2) long DNA molecules. (C) SEM (c1 and c2) and TEM (c3) images and the model (c4) of the chiral DNA-silica fibers with hierarchical chirality using APS as the CSDA. Reproduced with permission (105, 106). Copyright 2009, 2013, Royal Society of Chemistry. 68

4.2. DNA-Silica Nanoparticles As mentioned above, as a semi-rigid polymer, the stiffness of DNA is quantified by its persistence length (usually 50 nm), which greatly influences the assembly behavior of the molecules. These DNA fragments approach or are shorter than the persistence length, which can be viewed as rigid rods, and will more likely form liquid-crystal phases, which agrees with Onsager’s theory on rigid rod-like polymer assembly (107). The biomineralization of multiple liquid-crystals will highly expand the types of high-ordered structures of inorganic mineralization materials.

4.2.1. DNA-Silica Plates with p4mm Structure Lattices The silica biomineralization of the mesophase of short rigid DNA was first achieved by Che et al. (23) Similar to the strategy used during the formation of DNA-silica fibers, TMAPS was applied as the CSDA to bridge the biomolecular templates and the silica source. The other critical role of TMAPS was to act as a condensation agent to screen the electrostatic repulsions between the negatively charged DNA and to induce the highly ordered packing. The resulting DNA-silica composites exhibited a hexagonal platelet morphology with a thickness corresponding to the length of the template DNA molecules (Figure 13a1 and a3). The silicification of the DNA packing structure produced a mesoporous structure with a rare p4mm supra lattice. The fourfold symmetrical mesochannels with a long axis vertical to the platelets are rarely seen in naturally occurring DNA mesophases (Figure 13a2 and a4). Surprisingly, the morphology of the platelet was not square, which was inconsistent with its crystalline structure. Subsequent precise observations and analyses on the growth process have resulted in the following explanations (108). The formed hexagonal platelet morphologies coincided with the silicatropic DNA liquid crystal 2D hexagonal p6mm mesophase during the initial stage of the reaction. Then, a transition from the p6mm to p4mm structure inside the platelet occurred during the silica condensation, along with a decrease in the distances between the DNA molecules because of the changing charge densities of the reactants. These phenomena caused the DNA molecules to pack densely, which was attributed to the electrostatic interactions or “zipper” that pulled the molecules together (Figure 13b). The p4mm domains, which prefer a specific, smaller interaxial separation, were energetically favored. The flexible properties of the silica framework before the complete copolymerization enabled the transitions while maintaining the hexagonal morphology. This investigation regarding the phase transitions that occur during the solidification of the DNA mesophase provides a feasible observation model for studying the condensed phases of DNA and provides a new general method for the formation of more materials with exceptional structures and morphologies. 69

Figure 13. Morphologies and structures of the DNA-silica platelets. (a1) SEM images showing the hexagonal morphology of the platelets. (a2) and (a3) HRTEM images taken from the top and side of one typical platelet and the corresponding Fourier diffractograms showing the p4mm structure. (a4) Electrostatic-potential maps of the p4mm structure (23). (b) An illustration of the 2D-hexagonal to 2D-square transformation (108). Reproduced with permission (23, 108). Copyright 2009, Wiley-VCH Verlag GmbH & Co. KGaA. 2013, American Chemical Society.

4.2.2. Chiral DNA-Silica Impellers The chiral self-assembly of DNA into mesophases is an important feature of DNA molecules owing to their inherent chiral properties. Several types of interwound chiral conformations during DNA compaction have been found in organisms, such as bacterial plasmids and sperm heads, which are also intriguing for artificial systems in vitro. During the formation of achiral DNA-silica plates, the “zipper” lattice happens under a subtle electrostatic interaction balance. As mentioned above, competition exists between the densely packed constraints and the ability to twist along the chiral order, which determines the properties of the final assembled structure of the DNA. Therefore, a small disturbance in the synthesis system can break the balance and induce structural fluctuations. Chiral DNA-silica assemblies (CDSA) were obtained by introducing divalent alkaline earth ions (Mg2+) into a DNA-silica synthesis system, which produced 4 μm diametrical impeller-like particles with 100-nm thick blades (Figure 14A) (109). 70

Clockwise (marked as R-CDSA) or anticlockwise (marked as L-CDSA) blade stacking in a circle corresponded to the left-handed and right-handed handedness, respectively. Structural analyses revealed that the distorted 2D-square p4mm mesostructure of a typical blade with right-handed CDSAs twisted along the (01) lattice in a left-handed manner, as illustrated in the model in Figure 14B. From the model, one can clearly distinguish the coexistence of hierarchical chirality with opposite handedness. For instance, the R-CDSA with clockwise rotated blades is composed of left-handed chiral distortion p4mm layers and right handed helical array viewed along longitude axis of DNA molecules. Strong CD signal peaks at approximately 230 and 295 nm were attributed to the existence of DNA long-range chiral packing, similar to the nonconservative ellipticities exhibited during chiral cholesteric organization. The final CD signals of the material exhibited overlapping results owing to different responses of circular polarized light from the two opposite chiral structures. Since the presence of water greatly determined the secondary conformations of the DNA, which determined the chromophore coupling in the right-handed helical array, the materials exhibited water-dependent CD signal reversion (Figure 14C), implying their potential to be applied as humidity sensors (110). Interestingly, the synthesized system of CDSAs had a certain flexibility in that the molar ratio of the cationic silanes, temperature and pH values could influence the resulting handedness, which was also reflected in the DRCD spectra (Figure 14B). This flexibility was because that the origin of chirality in the dislocation arrays of the DNA is due to the charge mismatched “ion-bridging” between the DNA molecules and polyamine in the presence of divalent alkaline earth metal ions. Hence, another kind of competition existed during the formation of the CDSAs (111). This competition occurred between two interactions, the electrostatic interactions of the DNA-divalent metal ions and the electrostatic interactions of the DNA-polyamine complexes. Both the Mg2+ and amino species induced chirality but had opposite effects; right-handed handedness was favored by Mg2+, while left-handed chiral phases were favored by the latter (112). Thus, the handedness inversion occurred under proper conditions. 4.3. DNA-Silica Films 4.3.1. Patterned Achiral DNA-Silica Films Patterned DNA-silica films were obtained after the anchoring growth of DNA-silica plates on positively charged silicon substrates (113). This process exhibited interested surface behavior when the DNA-silica complexes were epitaxially grown on a substrate full of electronic charges. During the formation of the DNA-silica film, a strategy was used to align, place and arrange the DNA-silica plates on the patterned silicon substrate. The main method in this strategy was to thermodynamically and kinetically control the ionization degree of the DNA, TMAPS and quaternary ammonium groups at different locations of the lithographically patterned silicon surface, which was influenced by geometry effects (substrate grooves, protuberances and edge areas). As seen in Figure 15, by controlling the pH, TMAPS concentration and the lithographic patterning of 71

the silicon substrate, the alignment (Figure 15a), placement (Figure 15b and c) and arrangement (Figure 15d) of the epitaxially grown DNA-silica plates could be selectively controlled to form different types of patterned films.

Figure 14. Morphology and optical activity of the impeller-like HDSAs. (A) SEM images of R-CDSAs and L-CDSAs synthesized at 0 °C and 25 °C, respectively. (B) Schematic model showing the hierarchical chirality of the opposite handedness orientations from different viewpoints. (C) DRCD spectra of the HDSAs synthesized under different temperatures of (a) 25 °C, (b) 15 °C, (c) 8 °C, (d) 4 °C and (e) 0 °C measured in dry (dotted line) and wet states (solid line). Reproduced with permission (109, 110). Copyright 2011, 2013. Wiley-VCH Verlag GmbH & Co. KGaA.

72

Figure 15. Alignment, placement and arrangement of DNA-silica platelets on a lithographically patterned silicon surface with horizontal (a) or vertical epitaxial growth (b-d). (a) Alignment of DNA-silica plateltes on a silicon surface with horizontal epitaxial growth. (b-d) Positions of the selectively grown platelets influenced by the geometry of the silicon surface, in which the placement was influenced by the grooves (b), protuberances (c) and edges (d), respectively. Reproduced with permission (113). Copyright 2013. Wiley-VCH Verlag GmbH & Co. KGaA.

4.3.2. Chiral DNA-Silica Films The epitaxial growth of impeller-like chiral DNA-silica assemblies on silicon substrates resulted in chiral DNA-silica films that were composed of right-handed, vertically aligned semi-CDSAs (Figure 16a1) (114). The absorption effect of chromophore coupling within the DNA chiral packing and the circular Bragg resonance of the macroscopic impeller architecture induced absorption-based optical activity in the UV wavelength range and scattering-based optical activity that extended to the visible wavelength range, respectively. The CD signals in the scattering wavelength range were tuned by varying the average refractive index and helical pitch, according to the circular Bragg reflection equation (Figure 16a2). Furthermore, in order to meet the requirement of building highly ordered chiral films, a “quartet templating” method was constructed to guide chiral silica nanostructures arranged on crystalline mica surfaces (Figure 16B) (115). Owing to the bridging role of Mg2+, which could anchor to the (100) hexagonal symmetrical crystal surface of mica, an ordering transfer occurred from the “mica-Mg2+-DNA-silica” pathway, which produced densely grown chiral blades with a uniform orientation. This approach may provide a facile strategy for fabricating large-scale functional devices.

73

Figure 16. Morphologies and optical activity of the chiral DNA-silica films. (A1-2) SEM (A1) and DRCD/UV-Vis spectra (A2) of the randomly distributed chiral DNA-silica film that was grown on a silicon substrate. The DRCD spectra present the signals from the as-made film (red line), the calcined film (blue line) and the film after immersion in water (dotted line). B1-2) SEM (B1) and DRCD/UV-Vis spectra (B2) of the uniformly distributed chiral DNA-silica film that was grown on a mica substrate. The DRCD spectra presents the signals from the densely distributed film (black line), sparsely distributed film (red line) and random achiral film (blue line). (C) Photography (C1) and DRCD/UV-Vis spectra (C2) of the freestanding chiral DNA-silica film. The colored lines in the DRCD spectra correspond to the colored backgrounds. Reproduced with permission (114–116). Copyright 2014, Nature Publishing Group. 2016, Wiley-VCH Verlag GmbH & Co. KGaA. 2015, American Chemical Society.

In the absence of substrates, because of the film-forming ability of DNA and the high flexibility of organic silanes with long alkyl chains, pliable free-standing DNA-silica films were obtained by the silicification of the DNA condensed phase in the presence of a packing agent and CSDA, APS, using the evaporation-induced self-assembly process (116). The obtained antipodal films had a semitransparent appearance and exhibited broadband DRCD spectra when measured with a black background because of the long-range coupling and strong scattering effect of the domains of the semi-cholesteric DNA liquid crystals (Figure 16C, black line). The signals that originated from the scattering effect could be easily tuned to various narrow-band signals by changing the absorption bands of the backgrounds or by introducing colorful chromophore dye molecules into the film (Figure 16C, colored line). 74

5. Conclusion Inspired by nature, various kinds of methods for inducing ordered and highly compact phases of DNA have been developed through the use of naturally extracted or artificially synthesized DNA condensation agents. The studies of DNA condensation in vitro have provided simplified models for investigating the complex DNA condensation mechanisms in vivo and have further triggered the formation of various kinds of novel organic and inorganic mesostructured materials. The formation of these materials has enabled great progress in both biology and materials fields. The biggest achievements have been made for improving gene delivery technology, resulting in higher delivery efficiencies, facilitated gene uptake, and the timely release of the genes within targets. In addition, in the materials field, the ordered phases of DNA-related materials have shown many potential applications in electronics and optical fields (117–119). DNA-related optical materials possess better stability, higher optical-damage thresholds and low optical propagation losses, and can be considered as a new generation of photonic materials (120, 121). In the electronics field, DNA molecules can also act as semiconductors with a wide energy gap (122); in addition, the electrical conductivity of doped DNA-based systems exhibits typical ionic characteristics. The combination of these attraction properties of DNA and the ordered arrangements of mesostructured materials could give rise to fascinating high-tech applications, such as electronic chips, biosensors, electron guides and optical polarizers.

References Hamley, I. W. Introduction to soft matter: synthetic and biological self-assembling materials; John Wiley & Sons: Chichester, 2013. 2. Vijayanathan, V.; Thomas, T.; Thomas, T. J. Biochemistry 2002, 41, 14085–14094. 3. Martin, B.; Sainlos, M.; Aissaoui, A. Curr. Pharm. Design 2005, 11, 375–349. 4. Crick, F.; Watson, J. Nature 1953, 171, 737–738. 5. Sinden, R. R. DNA structure and function; Academic Press: Houston, 1994. 6. Filimonova, M.; Gubskaya, V.; Davidov, R.; Garusov, A.; Nuretdinov, I. Int. J. Biol. Macromol. 2008, 43, 289–294. 7. Gray, D. M.; Edmondson, S. P.; Lang, D.; Vaughan, M.; Nave, C. Nucleic Acids Res. 1979, 6, 2089–2107. 8. Kypr, J.; Kejnovska, I.; Renciuk, D.; Vorlickova, M. Nucleic Acids Res. 2009, 37, 1713–1725. 9. Cantor, C. R.; Schimmel, P. R. Biophysical Chemistry: Part III: the Behavior of Biological Macromolecules; W.H. Freeman: New York, 1980. 10. Frontali, C.; Dore, E.; Ferrauto, A.; Gratton, E.; Bettini, A.; Pozzan, M. R.; Valdevit, E. Biopolymers 1979, 18, 1353–1373. 11. Baumann, C. G.; Bloomfield, V. A.; Smith, S. B.; Bustamante, C.; Wang, M. D.; Block, S. M. Biophys. J. 2000, 78, 1965–1978. 1.

75

12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46.

Elias, J. G.; Eden, D. Macromolecules 1981, 14, 410–419. Strzelecka, T. E.; Davidson, M. W.; Rill, R. L. Nature 1988, 331, 457–460. Reich, Z.; Wachtel, E. J.; Minsky, A. Science 1994, 264, 1460–1463. Sartori Blanc, N. J. Struct. Biol. 2001, 134, 76–81. Merchant, K.; Rill, R. L. Biophys. J. 1997, 73, 154–3136. Robinson, C. Tetrahedron 1961, 13, 219–234. Brandes, R.; Kearns, D. R. Biochemistry 1986, 25, 5890–5895. Livolant, F.; Levelut, A. M.; Doucet, J.; Benoit, J. P. Nature 1989, 339, 724–726. Livolant, F.; Leforestier, A. Prog. Polym. Sci. 1996, 21, 1115–1164. Liu, B.; Cao, Y.; Huang, Z.; Duan, Y.; Che, S. Adv. Mater. 2015, 27, 479–497. Kornyshev, A. A.; Leikin, S. Phys. Rev. Lett. 1999, 82, 4138–4141. Jin, C.; Han, L.; Che, S. Angew. Chem., Int. Ed. 2009, 48, 9268–9272. Zanchetta, G.; Giavazzi, F.; Nakata, M.; Buscaglia, M.; Cerbino, R.; Clark, N. A.; Bellini, T. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 17497–17502. Leforestier, A.; Livolant, F. Biophys. J. 1993, 65, 56–72. Tombolato, F.; Ferrarini, A. J. Chem. Phys. 2005, 122, 054908. Reich, Z.; Ghirlando, R.; Arad, T.; Weinberger, S.; Minsky, A. J. Biol. Chem. 1990, 265, 16004–16006. Ziv, R.; Levin-Zaidman, S.; Gutman, S. B.; Arad, T.; Minsky, A. Biochemistry 1994, 33, 14177–14184. Leforestier, A.; Bertin, A.; Dubochet, J.; Richter, K.; Sartori Blanc, N.; Livolant, F. C. R. Chim. 2008, 11, 229–244. Leforstier, A.; Livolant, F. Liq. Cryst. 1994, 17, 651–658. Keller, D.; Bustamante, C. J. Chem. Phys. 1986, 84, 2972–2980. Bustamante, C.; Samorì, B.; Builes, E. Biochemistry 1991, 30, 5661–5666. Bloomfield, V. A. Biopolymers 1991, 31, 1471–1481. Rau, D. C.; Parsegian, V. A. Biophys. J. 1992, 61, 246–259. Raspaud, E.; Durand, D.; Livolant, F. Biophys. J. 2005, 88, 392–403. Bloomfield, V. A. Biopolymers 1997, 44, 269–282. Pelta, J.; Livolant, F.; Sikorav, J. L. J. Biol. Chem. 1996, 271, 5656–5662. Derouchey, J.; Netz, R. R.; R dler, J. O. Soft Matter 2005, 16, 17–28. DeRouchey, J.; Hoover, B.; Rau, D. C. Biochemistry 2013, 52, 3000–3009. Cohen, S. S. A Guide to the Polyamines; Oxford University Press: London, 1998. Hou, M. H.; Lin, S. B.; Yuann, J. M.; Lin, W. C.; Wang, A. H.; Kan Ls, L. Nucleic Acids Res. 2001, 29, 5121–5128. Hud, N. V.; Downing, K. H. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 14925–14930. Sikorav, J. L.; Pelta, J.; Livolant, F. Biophys. J. 1994, 67, 1387–1392. Pelta, J.; Livolant, F.; Sikorav, J. L. J. Biol. Chem. 1996, 271, 5656–5662. Raspaud, E.; de la Cruz, M. O.; Sikorav, J. L.; Livolant, F. Biophys. J. 1998, 74, 381–393. Toma, A. C.; Frutos, M. D.; Livolant, F.; Raspaud, E. Soft Matter 2011, 7, 8847–8855. 76

47. Saminathan, M.; Antony, T.; Shirahata, A.; Sigal, L. H.; Thomas, T.; Thomas, T. J. Biochemistry 1999, 38, 3821–3830. 48. Choosakoonkriang, S.; Lobo, B. A.; Koe, G. S.; Koe, J. G.; Middaugh, C. R. J. Pharm. Sci. 2003, 92, 1710–1722. 49. Li, G.-Y.; Guan, R.-L.; Ji, L.-N.; Chao, H. Coord. Chem. Rev. 2014, 281, 100–113. 50. Jorge, A. F.; Pereira, R. F. P.; Nunes, S. C. C.; Valente, A. J. M.; Dias, R. S.; Pais, A. A. C. C. Biomacromolecules 2014, 15, 478–491. 51. Asayama, S.; Nishinohara, S.; Kawakami, H. Bioconjugate Chem. 2011, 22, 1864–1868. 52. de la Tour, E.; Laemmli, U. K. Cell 1988, 55, 937–944. 53. Cremer, T.; Cremer, C. Nat. Rev. Genet. 2001, 2, 292–301. 54. Scheffer, M. P.; Eltsov, M.; Frangakis, A. S. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 16992–16997. 55. Chen, C. Y.; Su, C. J.; Peng, S. F.; Chen, H. L.; Sung, H. W. Soft Matter 2010, 7, 61. 56. Su, C. J.; Chen, C. Y.; Chen, H. L.; Ivanov, V. A. J. Phys.: Conf. Ser. 2011, 272, 012002. 57. Su, C.-J.; Chen, C.-Y.; Lin, M.-C.; Chen, H.-L.; Iwase, H.; Koizumi, S.; Hashimoto, T. Macromolecules 2012, 45, 5208–5217. 58. Zhang, K.; Jiang, M.; Chen, D. Angew. Chem., Int. Ed. 2012, 51, 8744–8747. 59. Zhao, Y.; Sakai, F.; Su, L.; Liu, Y.; Wei, K.; Chen, G.; Jiang, M. Adv. Mater. 2013, 25, 5215–5256. 60. Hayakawa, K.; Santerre, J. P.; Kwak, J. C. T. Biophys. Chem. 1983, 17, 175–181. 61. Dias, R.; Lindman, B. DNA Interactions with Polymers and Surfactants; John Wiley & Sons: Hoboken, NJ, U.S.A., 2008. 62. Zhao, D.; Wan, Y.; Zhou, W. Ordered mesoporous materials; Wiley-VCH Verlag & Co. KGaA: 2013. 63. Bilalov, A.; Olsson, U.; Lindman, B. Soft Matter 2011, 7, 730–742. 64. Radhakrishnan, A. V.; Ghosh, S. K.; Pabst, G.; Raghunathan, V. A.; Sood, A. K. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 6394–6398. 65. Krishnaswamy, R.; Raghunathan, V. A.; Sood, A. K. Phys. Rev. E 2004, 69, 031905. 66. Krishnaswamy, R.; Pabst, G.; Rappolt, M.; Raghunathan, V. A.; Sood, A. K. Phys. Rev. E 2006, 73, 031904. 67. Mel'nikova, Y. S.; Lindman, B. Langmuir 2000, 16, 5871–5878. 68. Rädler, J. O.; Koltover, I.; Salditt, T.; Safinya, C. R. Science 1997, 275, 810–814. 69. Koltover, I.; Salditt, T.; Rädler, J. O.; Safinya, C. R. Science 1998, 281, 78–81. 70. Zuidam, N. J.; Barenholz, Y.; Minsky, A. FEBS Lett. 1999, 457, 419–422. 71. Amar-Yuli, I.; Adamcik, J.; Blau, S.; Aserin, A.; Garti, N.; Mezzenga, R. Soft Matter 2011, 7, 8162. 72. Corsi, J.; Dymond, M. K.; Ces, O.; Muck, J.; Zink, D.; Attard, G. S. Chem. Commun. 2008, 2307–2309. 77

73. Black, C. F.; Wilson, R. J.; Nylander, T.; Dymond, M. K.; Attard, G. S. J. Am. Chem. Soc. 2010, 132, 9728–9732. 74. Okahata, Y.; Kawasaki, T. Top. Curr. Chem. 2005, 260, 57–75. 75. Tanaka, K.; Okahata, Y. J. Am. Chem. Soc. 1996, 118, 10679–10683. 76. Lerman, L. S. Proc. Natl. Acad. Sci. U.S.A. 1971, 68, 1886–1890. 77. Mel'nikov, S. M.; Yoshikawa, K. Biochem. Biophys. Res. Commun. 1997, 230, 514–517. 78. Vasilevskaya, V. V.; Khokhlov, A. R.; Matsuzawa, Y.; Yoshikawa, K. J. Chem. Phys. 1995, 102, 6595–6602. 79. Post, C. B.; Zimm, B. H. Biopolymers 1979, 18, 1487–1501. 80. Starodoubtsev, S. G.; Yoshikawa, K. Langmuir 1998, 14, 214–217. 81. Mayama, H.; Iwataki, T.; Yoshikawa, K. Chem. Phys. Lett. 2000, 318, 113–117. 82. Starodoubtsev, S. G.; Yoshikawa, K. J. Phys. Chem. 1996, 100, 19702–19705. 83. Asakura, S.; Oosawa, F. J. Chem. Phys. 1954, 22, 1255–1256. 84. Evdokimov, Y. M.; Pyatigorskaya, T. L. Nucleic Acids Res. 1976, 3, 2353–2366. 85. Jmdan, C. F.; Lerman, L. S.; Venable, J. H. Nature 1972, 236, 67–70. 86. Ramos, J. E. B.; Neto, J. R.; de Vries, R. J. Chem. Phys. 2008, 129, 185102. 87. Ha, B.-Y.; Jung, Y. Soft Matter 2015, 11, 2333–2352. 88. Zhang, C.; Shao, P. G.; van Kan, J. A.; van der Maarel, J. R. C. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 16651–16656. 89. Jones, J. J.; van der Maarel, J. R. C.; Doyle, P. S. Nano Lett. 2011, 11, 5047–5053. 90. Lang, D.; Taylor, T. N.; Dobyan, D. C.; Gray, D. M. J. Mol. Biol. 1976, 106, 97–107. 91. Reich, C.; Maestre, M. F.; Edmondson, S.; Gray, D. M. Biochemistry 1980, 19, 5208–5213. 92. Huey, R.; Mohr, S. C. Biopolymers 1981, 20, 2533–2552. 93. Minsky, A. Chirality 1998, 10, 405–414. 94. Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck, J. S. Nature 1992, 359, 710–712. 95. Wan, Y.; Zhao, D. Chem. Rev. 2007, 107, 2821–2860. 96. Tomczak, M. M.; Glawe, D. D.; Drummy, L. F.; Lawrence, C. G.; Stone, M. O.; Perry, C. C.; Pochan, D. J.; Deming, T. J.; Naik, R. R. J. Am. Chem. Soc. 2005, 127, 12577–12582. 97. Soicher, M. A.; Christiansen, B. A.; Stover, S. M.; Leach, J. K.; Yellowley, C. E.; Griffiths, L. G.; Fyhrie, D. P. J. Biomed. Mater. Res. A 2014, 102, 4480–4490. 98. Shopsowitz, K. E.; Qi, H.; Hamad, W. Y.; MacLachlan, M. J. Nature 2010, 468, 422–425. 99. Mao, C.; Wang, F.; Cao, B. Angew. Chem., Int. Ed. 2012, 51, 6411–6415. 100. Fowler, C. E.; Shenton, W.; Stubbs, G.; Mann, S. Adv. Mater. 2001, 13, 1266–1269. 101. Liang, H.; Angelini, T. E.; Ho, J.; Braun, P. V.; Wong, G. C. L. J. Am. Chem. Soc. 2003, 125, 11786–11787. 78

102. Kinsella, J. M.; Ivanisevic, A. Langmuir 2007, 23, 3886–3890. 103. Berti, L.; Burley, G. A. Nat. Nanotechnol. 2008, 3, 81–87. 104. Numata, M.; Sugiyasu, K.; Hasegawa, T.; Shinkai, S. Angew. Chem., Int. Ed. 2004, 43, 3279–3283. 105. Jin, C.; Qiu, H.; Han, L.; Shu, M.; Che, S. Chem. Commun. 2009, 3407–3409. 106. Cao, Y.; Xie, J.; Liu, B.; Han, L.; Che, S. Chem. Commun. 2012, 49, 1097–1099. 107. Onsager, L. Ann. N. Y. Acad. Sci. 1949, 51, 627–659. 108. Han, L.; Jin, C.; Liu, B.; Che, S. Chem. Mater. 2012, 24, 504–511. 109. Liu, B.; Han, L.; Che, S. Angew. Chem., Int. Ed. 2011, 51, 923–927. 110. Liu, B.; Cao, Y.; Duan, Y.; Che, S. Chem. Eur. J. 2013, 19, 16382–16388. 111. Liu, B.; Han, L.; Che, S. Interface Focus 2012, 2, 608–616. 112. Liu, B.; Han, L.; Che, S. J. Mater. Chem. B 2013, 1, 2843–2850. 113. Liu, B.; Yao, Y.; Che, S. Angew. Chem., Int. Ed. 2013, 52, 14186–14190. 114. Liu, B.; Han, L.; Duan, Y.; Cao, Y.; Feng, J.; Yao, Y.; Che, S. Sci. Rep. 2014, 4, 4866. 115. Cao, Y.; Kao, K.; Mou, C.; Han, L.; Che, S. Angew. Chem., Int. Ed. 2016, 55, 2037–2041. 116. Cao, Y.; Che, S. Chem. Mater. 2015, 27, 7844–7851. 117. Steckl, A. J. Nature Photon 2007, 1, 3. 118. Heckman, E. M.; Grote, J. G. Proc. SPIE 2004, 5516, 41–47. 119. Chao, J.; Lin, Y.; Liu, H.; Wang, L.; Fan, C. Biochem. Pharmacol. 2015, 18, 326–335. 120. Heckman, E. M.; Hagen, J. A.; Yaney, P. P.; Grote, J. G.; Hopkins, F. K. Appl. Phys. Lett. 2005, 87, 211115. 121. Moldoveanu, M.; Meghea, A.; Popescu, R.; Grote, J. G.; Kajzar, F.; Rau, I. Mol. Cryst. Liq. Cryst. 2010, 523, 182/[754]–190/[762]. 122. Taniguchi, M.; Otsuka, Y.; Tabata, H.; Kawai, T. Jpn. J. Appl. Phys. 2003, 42, 6629–6630.

79

Polypeptide and Engineered Proteins

Chapter 5

Adhesive Growth Factors Inspired by Underwater Adhesion Proteins Chen Zhang,1,2,3 Hideyuki Miyatake,1 and Yoshihiro Ito1,3,4,* 1Nano

Medical Engineering Laboratory, RIKEN, 2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan 2School of Pharmaceutical Sciences, Jilin University, No. 1266 Fujin Road, Changchun, Jilin 130021, P. R. China 3Key Laboratory of Polymer Ecomaterials, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun, Jilin 130022, P. R. China 4Emergent Bioengineering Materials Research Team, RIKEN Center for Emergent Matter Science, 2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan *E-mail: [email protected]

The generation of material surfaces with biological properties such as cell growth-enhancement and differentiation-inducing abilities could be useful for the development of functional materials for medical applications. Here, by the extension of conventional protein engineering into bio-orthogonal protein engineering using the specific incorporation of non-natural amino acids, we have developed new binding growth factors for the surface modification of materials, to impart biological activity on them.

Introduction A wide variety of different biomaterials are currently employed for medical applications. Research towards the development of new biomaterials has traditionally focused on the preparation of functional materials capable of the simple adhesion of cells or the connection of tissues to metals and ceramics. However, there is a growing interest in the development of biomaterials involving the immobilization of growth factors, which would allow these artificial materials to regulate specific cellular functions, including the gene expression processes © 2017 American Chemical Society

associated with cell growth and differentiation (1, 2). Binding growth factors have been designed and applied for two purposes, as shown in Figure 1. One is delivery to specific sites or enhancement of local concentrations at specific sites using direct injection. Another involves surface modification of scaffolds for implantation.

Figure 1. Applications of binding growth factors. Binding growth factors have been applied using two methods. One involves delivery to specific sites or enhancement of local concentrations at specific sites using direct injection. The other involves surface modification of scaffolds for implantation of bio-functional materials. Reproduced with permission from ref. (4). Copyright [2013] (4) [Elsevier Ltd.] In terms of the latter application, various types of materials have been modified. However, although numerous studies have reported the use of metallic materials in medical devices such as artificial joints, dental implants and stents, there have been very few reports concerning the surface modification of metal or ceramic materials with biological agents (3). In recent years, an increasing number of reports have been published on natural adhesives (5–7). To design such proteins as a strategy for the surface modification of metals and ceramics, biomimetic approaches inspired by underwater adhesive proteins have been employed with biological materials. Figure 2 shows the use of motifs of two underwater adhesive proteins. One of the proteins is the salivary stathelin protein which is a multifunctional molecule that possesses a high affinity for calcium phosphate minerals such as hydroxyapatite (HA), maintains the appropriate mineral solution dynamics of enamel, promotes selective initial bacterial colonization, and functions as a boundary lubricant on the enamel surface (8). Another is mussel foot protein which is involved in a sticky pad at the end of threads which stick firmly to rock, or any other hard surface (9). It is known that the active sites for the adhesion of both proteins consist of post-translationally modified amino acids. 84

Figure 2. Combination of growth factors with the active sites of underwater adhesion proteins such as salivary stathelin and the mussel foot protein (Mfp). One is a phosphorylated serine of salivary statherin and another is the 3,4dihyroxyphenylalanine (DOPA) of mussel foot proteins. Both cannot be directly incorporated into a protein using conventional protein engineering (recombinant DNA ) techniques. Therefore, bio-orthogonal approaches are required to prepare proteins with site-specific incorporation of these amino acids. Here, two approaches were employed. The first is chemical synthesis using a solid phase method. The second is a combination of conventional protein (recombinant gene) engineering and enzymatic modification. We have prepared phosphoserine-incorporated epidermal growth factor (EGF) and bone morphogenetic protein-4 (BMP-4), and DOPA-incorporated insulin-like growth factor-1 (IGF-1) using these bio-orthogonal approaches.

Adhesive EGF To prepare phosphoserine-incorporated EGF, we employed the solid phase synthesis method because EGF is a small protein with 53 amino acids. The length of peptides that can be synthesized using a solid-phase method is generally limited to around 50 amino acids (10). The schematic design of the phosphoserine-incorporated EGF peptide (EGFN8P) is shown in Figure 3A. The minimum portion of bio-active EGF was conjugated with the short peptide with the binding sequence in statherin. Figure 4A shows that the modified EGF bound to HA and titanium (Ti), whereas unmodified EGF did not. This shows that the binding affinity of the stathelin fragment was maintained after conjugation with the EGF fragment. The mitogenic activity of EGF derivatives on HA and Ti is shown in Figure 4B. Smaller amounts of bound EGF had higher effects than soluble EGF. As a conclusion, the bound EGF has higher activity than the soluble form because of its immobilization. 85

Figure 3. Design of binding growth factors. (A) Adhesive EGF using a solid phase method. (B) Adhesive BMP-4 and IGF-1 synthesized using a combination of recombinant DNA technology and enzyme treatments.

Adhesive BMP-4 In the case of human BMP-4 (hBMP4), because its length is 117 amino acids, it is difficult to synthesize it using the solid phase method. Therefore, we combine conventional protein engineering with an enzyme treatment. Here, the sortase A enzyme was used for ligation of modified human BMP (hBMP) with a phosphoserine-containing peptide (11). Figure 3B shows the preparation of the protein derivative. The growth factor portion and the phosphorylated peptide portion were prepared using a gene engineering and the solid phase method, respectively, and they were ligated using the sortase enzyme. Figure 5 shows that the prepared phosphorylated peptide-carrying human BMP-4 (hBMP4-pSpS) has a higher binding affinity to HA than the non-phosphorylated one (hBMP4-SS). In addition, the phosphorylated BMP-4 significantly induced bone formation activity on HA. Considering these results, this novel protein incorporated with phosphoserine is expected to contribute to the preparation of bioactive bone regeneration materials in combination with HA.

86

Figure 4. (A) Binding of EGF and modified EGF (EGFN8P) onto HA (hydroxyapatite) and Ti (titanium). (B) Mitogenic activity of modified EGF (EGFN8P) bound to hydroxyapatite and titanium. Reproduced with permission from ref. (10) Copyright [2013] (10) [Elsevier Ltd.].

Adhesive IGF-1 The enzymatic treatment using tyrosinase can incorporate DOPA into IGF by converting the tyrosine residues to DOPA (Figure 3B) (12). In the procedure, a tyrosine-lysine-tyrosine-lysine-tyrosine sequence was added to the C-terminal of IGF-1 using conventional genetic recombinant technology. The resultant protein is referred to as IGF-Y. Subsequently, IGF-Y was treated with tyrosinase and the tyrosine residues were converted to DOPA residues. The final product is referred to as IGF-X. The binding affinities of the different IGF-1 derivatives towards Ti were investigated using quartz crystal microbalance (QCM) with dissipation monitoring, as shown in Figure 6. The binding affinity of IGF-X was significantly higher than that of IGF-Y at pH 8.5. Furthermore, the bound IGF-X did not dissociate even after it had been washed with phosphate buffered saline. Figure 7A shows the cell growth assay results for NIH 3T3 cells in the presence of soluble IGF-Y, commercial IGF-1 and bound IGF-X on Ti. Compared with soluble IGF proteins, the bound IGF-X produced a significant enhancement of cell growth.

87

Figure 5. (A) Binding of engineered hBMP4 proteins to HA (hydroxyapatite). hBMP4-SS (non-phosphorylated serine-carrying human BMP4) or hBMP4-pSpS (phosphorylated serine-carrying BMP4) was incubated with HA beads. The amount of bound proteins was measured using an anti-BMP4 antibody. Mean (n = 3) ± SD is plotted. * p