Adhesive complex coacervate inspired by the sandcastle worm as a sealant for fetoscopic defects

Inspired by the Sandcastle Worm, biomimetic of the water-borne adhesive was developed by complex coacervation of the synthetic copolyelectrolytes, mimicking the chemistries of the worm glue. The developed underwater adhesive was designed for sealing fetal membranes after fetoscopic surgery in twin-to-twin transfusion syndrome (TTTS) and sealing neural tissue of a fetus in aminiotic sac for spina bifida condition. Complex coacervate with increased bond strength was created by entrapping polyethylene glycol diacrylate (PEG-dA) monomer within the cross-linked coacervate network. Maximum shear bond strength of ~ 1.2 MPa on aluminum substrates was reached. The monomer-filled coacervate had complex flow behavior, thickening at low shear rates and then thinning suddenly with a 16-fold drop in viscosity at shear rates near 6 s-1. The microscale structure of the complex coacervates resembled a three-dimensional porous network of interconnected tubules. This complex coacervate adhesive was used in vitro studies to mimic the uterine wall-fetal membrane interface using a water column with one end and sealed with human fetal membranes and poultry breast, and a defect was created with an 11 French trocar. The coacervate adhesive in conjunction with the multiphase adhesive was used to seal the defect. The sealant withstood an additional traction of 12 g for 30-60 minutes and turbulence of the water column without leakage of fluid or slippage. The adhesive is nontoxic when in direct contact with human fetal membranes in an organ culture setting.

ADHESIVE COMPLEX COACERVATE INSPIRED BY THE SANDCASTLE WORM AS A SEALANT FOR FETOSCOPIC DEFECTS by Sarbjit Kaur A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Materials Science and Engineering The University of Utah May 2015 Copyright © Sarbjit Kaur 2015 All Rights Reserved The University of Utah Graduate School STATEMENT OF DISSERTATION APPROVAL The dissertation of Sarbjit Kaur has been approved by the following supervisory committee members: Russell J. Stewart Dinesh K. Shetty Vladimir Hlady Jules J. Magda Dmitry Bedrov Taylor Sparks Chair Member Member Member Member Member 7/23/2014 Date Approved 7/23/2014 Date Approved 7/23/2014 Date Approved 7/23/2014 Date Approved 10/21/2014 Date Approved 10/15/2014 Date Approved and by Feng Liu Chair/Dean of the Department/College/School o f ________Materials Science and Engineering and by David B. Kieda, Dean of The Graduate School. ABSTRACT Inspired by the Sandcastle Worm, biomimetic of the water-borne adhesive was developed by complex coacervation of the synthetic copolyelectrolytes, mimicking the chemistries of the worm glue. The developed underwater adhesive was designed for sealing fetal membranes after fetoscopic surgery in twin-to-twin transfusion syndrome (TTTS) and sealing neural tissue of a fetus in aminiotic sac for spina bifida condition. Complex coacervate with increased bond strength was created by entrapping polyethylene glycol diacrylate (PEG-dA) monomer within the cross-linked coacervate network. Maximum shear bond strength of ~ 1.2 MPa on aluminum substrates was reached. The monomer-filled coacervate had complex flow behavior, thickening at low shear rates and then thinning suddenly with a 16-fold drop in viscosity at shear rates near 6 s-1. The microscale structure of the complex coacervates resembled a three-dimensional porous network of interconnected tubules. This complex coacervate adhesive was used in vitro studies to mimic the uterine wall-fetal membrane interface using a water column with one end and sealed with human fetal membranes and poultry breast, and a defect was created with an 11 French trocar. The coacervate adhesive in conjunction with the multiphase adhesive was used to seal the defect. The sealant withstood an additional traction of 12 g for 30-60 minutes and turbulence of the water column without leakage of fluid or slippage. The adhesive is nontoxic when in direct contact with human fetal membranes in an organ culture setting. A stable complex coacervate adhesive for long-term use in TTTS and spina bifida application was developed by methacrylating the copolyelectrolytes. The methacrylated coacervate was crosslinked chemically for TTTS and by photopolymerization for spina bifida. Tunable mechanical properties of the adhesive were achieved by varying the methacrylation of the polymers. Varying the amine to phosphate (A/P) ratio in the coacervate formation generated a range of viscosities. The chemically cured complex coacervate, with sodium (meta) periodate crosslinker, was tested in pig animal studies, showing promising results. The adhesive adhered to the fetal membrane tissue, with maximum strength of 473 ± 82 KPa on aluminum substrates. The elastic modulus increased with increasing methacrylation on both the polyphosphate and polyamine within the coacervate. Photopolymerized complex coacervate adhesive was photocured using Eosin-Y and treiethanolamine photoinitiators, using a green laser diode. Soft substrate bond strength increased with increasing PEG-dA concentration to a maximum of ~90 kPa. The crosslinked complex coacervate adhesives with PEG networks swelled less than 5% over 30 days in physiological conditions. The sterile glue was nontoxic, deliverable through a fine cannula, and stable over a long time period. Preliminary animal studies show a novel innovative method to seal fetal membrane defects in humans, in utero. iv This dissertation is dedicated to my Mother and sisters, Manpreet Kaur and Harjit Kaur, and to all the Women who have inspired me to be here today. ABSTRACT.............................................................................................................................iii ACKNOWLEDGEMENTS...................................................................................................viii CHAPTERS 1 INTRODUCTION..............................................................................................................1 1.1 Adhesives.......................................................................................................................1 1.2 Sandcastle Worm Inspired............................................................................................3 1.3 Biomimetic Complex Coacervate Adhesive.............................................................11 1.4 Fetal Defects................................................................................................................ 18 1.5 Aim of This Research................................................................................................. 20 1.6 Outline of This Thesis..............................................................................................25 1.7 References....................................................................................................................27 2 MULTIPHASE ADHESIVE COACERVATES INSPIRED BY THE SANDCASTLE WORM......................................................................................................... 36 2.1 Abstract....................................................................................................................... 36 2.2 Introduction.................................................................................................................. 37 2.3 Materials and Methods................................................................................................39 2.4 Results and Discussion...............................................................................................43 2.5 Conclusion...................................................................................................................52 2.6 References....................................................................................................................55 3 FETAL MEMBRANE PATCH AND BIOMIMETIC ADHESIVE COACERVATES AS A SEALANT FOR FETOSCOPIC DEFECTS..............................58 3.1 Abstract....................................................................................................................... 58 3.2 Introduction.................................................................................................................. 59 3.3 Materials and Methods................................................................................................61 3.4 Results and Discussion...............................................................................................68 3.5 Conclusion................................................................................................................... 74 3.6 References....................................................................................................................74 4 SYNTHESIS OF METHACRYLATED POLYPHOSPHATE AND POLYAMINE POLYMERS AND ADHESIVE COMPLEX COACERVATE FORMATION...............78 TABLE OF CONTENTS 4.1 Abstract........................................................................................................................78 4.2 Introduction.................................................................................................................. 79 4.3 Materials and Methods................................................................................................82 4.4 Results and Discussion...............................................................................................89 4.5 Conclusion................................................................................................................. 103 4.6 References..................................................................................................................103 5 METHACRYLATED POLYMERS COMPLEX COACERVATE ADHESIVE FOR SEALING FETAL DEFECTS IN TWIN-TO-TWIN TRANSFUSION SYNDROME..107 5.1 Introduction................................................................................................................ 107 5.2 Materials and Methods..............................................................................................110 5.3 Results and Discussion.............................................................................................114 5.4 Conclusion.................................................................................................................129 5.5 References.................................................................................................................. 129 6 PHOTOPOLYMERIZED METHACRYLATED POLYMER COMPLEX COACERVATE ADHESIVE PATCH FOR SPINA BIFIDA......................................... 131 6.1 Introducti on................................................................................................................131 6.2 Materials and Methods..............................................................................................134 6.3 Results and Discussion.............................................................................................139 6.4 Conclusion................................................................................................................. 150 6.5 References.................................................................................................................. 150 7 CONCLUSION.............................................................................................................153 vii ACKNOWLEDGEMENTS The research work of this dissertation was carried out with Dr. Russell J. Stewart with the Department of Biomedical Engineering from September 2009 to June 2014. I want to give my hearty gratitude to my thesis advisor and committee chair Dr. Russell J. Stewart for his tremendous support and guidance throughout this process. He is an outstanding scientist and inspiring mentor. I am fortunate to have the opportunity to work with him on this project. He has continued to push me and has helped me grow academically and personally. I thank him for such significant research. I want to thank our wonderful collaborators at the Texas Fetal Center, University of Texas Houston, Dr. Ramesha Papanna, Dr. Lovepreet K. Mann, and Dr. Kenneth J. Moise, for their kind support and companionship throughout this project. They have been gracious and hosted the animal studies that have been conducted. Thank you for expanding my scope of expertise and making such strong contributions to this work. I also would like to thank Dr. Robert H. Byrd, Dr. Edwina J. Popek, and Dr. Scheffer C. G. Tseng for contributing to the fetal studies. Additionally I want to extend my gratitude to my committee members Dr. Vladimir Hlady, Dr. Dinesh Shetty, Dr. Taylor Sparks, Dr. Dmitry Bedrov, and Dr. Jules J. Magda for their guidance, support, and advice. I would also like to thank my group members: Hui Shao, Ching-shun Wang, Mahika Weerasekare, Achille Mayelle Bivigou-Koumba, and Oscar V. Jasklowski. They have helped me with every small question and provided support, advice, and emotional support and strength. I would like to thank Dr. Raymond A. Cutler for being such an amazing mentor in my life. His firm belief in me and giving me the opportunity to do my first research project has helped me grow tremendously in the field of science. I have learned so much about doing research from him and am obliged and humbled by his work ethics. I am grateful for the knowledge he instilled in me, and it will guide me in the future. Lastly, I want to thank my family and friends, without whom this all would not be possible, my parents, who have shown me that anything is achievable by working hard, my dear sisters Manpreet Kaur and Harjit Kaur, for always standing by my side through thick and thin and never giving up on me. They have helped me get back up after every failure and challenge that I faced throughout my life. My family has been the foundation to what I am today; without them I wouldn't be here. I want to thank my dear friends who have supported me and motivated me to get my PhD done. Thank you to all the inspirational friends and mentors who make me what I am today. ix CHAPTER 1 INTRODUCTION This chapter is an introduction to complex coacervate adhesives, inspired by the sandcastle worm, for the repair of fetal defects. The inspiration behind this work, the Sandcastle Worm, is thoroughly covered. The concepts that are key to this thesis are covered by explaining the complex coacervation. The formation of the synthetic underwater adhesive complex coacervate is discussed. The motivation behind this work comes from lack of adhesives to seal soft tissues, specifically for sealing defects in environments where adhesives have to be applied under aqueous conditions, like in utero. Research for this thesis is based on developing a synthetic complex coacervate adhesive for sealing fetal defects for two applications: Twin-Twin transfusion Syndrome and Spina Bifida, which represents a model for other fetal defect repair. An outline of this thesis is given at the end of this chapter. 1.1 Adhesives Adhesives can be defined as materials that can join two surfaces together upon application and resist its separation [1]. Many terms like glue, cement, paste, etc., can be used interchangeably to define adhesives. Adhesion is a property often used to signify attraction between surfaces or particles. Adhesion forces itself to operate across the interface due to molecular van der Waals forces, dipole-dipole interactions, hydrogen bonding, or covalent bonding [2]. The materials joined by the adhesive are known as substrates. Adhesives have been around for centuries to glue materials together. One of the first adhesives dates back 200,000 years ago, using birch-bark-tar glue to fasten spear stone flake to a wooden shaft in Italy [3]. It is not until the last century, however, that the advances in adhesion and adhesives have been made. A major component playing a key role in these advances is the material developments of synthetic polymers in the 20th century. These materials have balance properties that allow them to adhere to other materials and transmit loads or forces between the two substrates [1]. Increasing demands in applications has spurred the research and development of specific adhesive formulations. 1.1.1 Soft Tissue Adhesives An advance in adhesive technologies in all fields and applications has led to its use in medicine. Sutures and mechanical fixations were primarily used to bind tissue and bone together and still continue to be used in many areas today. Although a lot of progress has been made over the past decade in medicine, a suture is still the conventional method to close skin incisions. Alternative methods for soft tissue repair over the years have led to the development of soft tissue adhesives. Adhesives for tissue repair can replace sutures in many cases as well as limit their use. Gluing has many advantages because its fast and uncomplicated technique causes very little damage to the surrounding tissue [4]. A homogenous load is distributed between the bonding tissues. They are especially useful in applications where suturing is difficult, whether by location or the type of tissue that is being repaired. Using an adhesive also helps in sealing the 2 3 wound to prevent any leakage of fluid. 1.1.2 Underwater Adhesives Since our body is mostly made up of water, maintaining adhesion underwater is a major challenge. A soft tissue adhesive must be able to bind the tissue together with adequate strength, be nontoxic, and maintain its integrity until wound healing occurs. Getting adequate strength for biological environments is a difficult task to accomplish for many bioadhesives. Synthetic adhesives used today are normally designed for dry applications and perform poorly in the presence of water [5]. They can have high strength to start out with but will eventually fail due to poor interfacial adhesion in watery environments [6-8]. Therefore, a good underwater adhesive must have robust interfacial adhesion to wet surfaces, without any surface preparation, while depositing the adhesive under water, with controlled solidification [5]. A suitable soft tissue adhesive must be able to overcome binding tissue surfaces in wet environments, maintain their strength over time, and be nontoxic. 1.2 Sandcastle Worm Inspired We can learn a lot from Nature. We are inspired by our surroundings and adapt from them. There are many solutions we can derive to our modern day problems by observing natural phenomena. From a material scientist's perspective, a wide variety of materials with different functions serve as a source of inspiration [9]. Taking these materials from observation to in-depth analysis leads to bioinspired materials. To tackle the underwater adhesion problem, marine organisms (like mussels, barnacles, and sandcastle worms) are being studied. These organisms secrete a liquid protein adhesive, which adheres to all types of wet surfaces, whether organic or inorganic [10]. Intrigued by these natural organisms, scientists have developed biomimetic underwater adhesives for strong adhesion to biological tissue [11-13] The sandcastle worm (Phragmatopoma Californica, P-Cal) is important to our research. This organism is unique in that it is able to bind dissimilar materials together under seawater in a single step without much surface preparation, with a self-contained mechanism to trigger a setting reaction, building a protective shell that it lives in [14]. The adhesive is secreted as a colloidal suspension with very low interfacial tension, which allows it to spread easily over the substrate while remaining very cohesive and not dispersing into the ocean [13]. Despite all the turbulent forces, temperature changes, and fluctuating salinities taking place in the ocean, the tubular structure that the worm builds does not fall apart [15]. The sandcastle worm cement represents the simplest permanent bioadhesive investigated to date [16]. Compared to mussels and barnacles, which directly attach themselves to the substrates, the sandcastle worm is mobile within the tube that it builds. Over the years these organisms have evolved workable materials solutions. By gaining knowledge and understanding of these natural phenomena, we are able to design and synthesize adhesives in wet environments [17]. Based on the basic chemistry of the natural worm glue, our synthetic complex coacervate adhesive is made. 1.2.1 Sandcastle Worm (Phragmatopoma Californica) Phragmatopoma Californica (P. californica) is a marine polychaete that belongs to the family of sabellariids. P. californica (also called the sandcastle worm) lives along the coast of California. The sabellariids (known as tube-dwelling polychaetes) build massive reef-like mounds, consisting of tubes held together in honeycomb like 4 arrangements (Figure 1.1A and 1.1B). These tubular homes that the worm lives in almost resemble large-stone masonry [18]. Sabellariids are different than other shell-dwelling marines, such that that they gather mineral phases from the ocean and secrete a proteinaceous glue to bind the sand particles together [20], rather than making a complete mineralized shell [21]. The sandcastle worm has ciliated tentacles (Figure 1.1 C) that capture and transport food and particles from the ocean to its mouth [22]. They collect passing materials like sand grains, calcareous shells, or debris from water to bind. The food is ingested through the mouth, and the particles are evaluated for the right size, shape, composition, and surface chemistry at the building organ near the mouth. The unsuitable particles are cast away, and the suitable particles are pressed onto the existing tube in the best position to minimize any gaps in the structure [19]. Particles at the point of contact are spotted with glue and pressed into place. The worm wriggles the particle until the glue sets, taking approximately 20-30 sec [23]. Each worm builds the tube that it lives in for physical protection. A large colony of reef like structures is built by a coordinated effort, which affects the reef ecology [24]. The exact strength of these tubular structures is difficult to assess, but their location in these turbulent environments and the durability of the structure suggest a robust construction [13]. 1.2.2 Sandcastle Worm Adhesive Structure The sandcastle worm sets into a flexible leathery material with a structure of solidified foam, as shown in Figure 1.2 [20]. In the laboratory, the worm is given glass beads, which it uses to build a glass bead tube (Figure 1.2A). Its structure was analyzed using laser scanning confocal and atomic force microscopies. A porosity gradient was 5 6 (A) Photo courtesy of Dr. Jerome Fournier. Figure 1.1 Reef-building sabellariid tubeworms. A) Lateral growth of isolated dome­shaped colonies of S. alveolota (foreground) leads to fusion of colonies into a continuous tabular surface covering the beach. B) Closer view of a colony of P. californica. Each tube contains an individual worm. C) Left: P. californica removed from tubular shell. White bracket indicates parathorax region that contains the adhesive gland. Right: Zirconium oxide beads have been glued onto the anterior end of the natural tube. Arrow indicates the operculum [19], Reprinted with permission from C. S. Wang, and R. J. Stewart, "Multipart Copolyelectrolyte Adhesive of the Sandcastle Worm, Phragmatopoma Californica (Fewkes): Catechol Oxidase Catalyzed Curing Through Peptidyl-DOPA," Biomac., 14 [5] 1607-1617 (2013). Copyright 2013 American Chemical Society. 7 Figure 1.2 SEM of sandcastle adhesive. A) A tube of partially rebuilt with glass beads. The glue was applied only at four contact points (arrows); B) Sandcastle worms placed on coverslips glue glass beads to the surface. The glue fractured when the bead was pried away; C) The foamy interior in the right box in B; D) A spot of glue left on a glass bead, indicating liquid until set; E) Threads and nonporous skin layer on glue; F) Foamy interior imaged with backscattered electron detector. Distinct layers on the surface (arrow) and linking the pores are visible. Scale bar in (B and D) 50 ^m; (C and E) 15 ^m; F) 5 [26], Springer, New York, and Biological Adhesive Systems, Editors J. V. Byem, I. Grunwalds, 2010, Morphology of the Adhesive System in the Sandcastle Worm, Phragmatopoma Californica, S. Wang, K. K. Svendsen, and R. J. Stewart, is given to the publication in which the material was originally published, by adding; with kind permission from Springer Science and Buisness Media. observed with hardly any porosity on both the outside of the glue and the solid foam within the internal structure (Figure 1.2 B, C, D, E, & F). Showing 50% porosity down the center of the adhesive joint [23]. This was similar to the structure of the byssall adhesive plaques of the mussels. The foamy structure of the adhesive is very advantageous because it increases the adhesives elasticity and toughness (amount of energy that a material can absorb before failing) [25]. A foam packing material is more flexible and absorbs the dissipated energy, which decreases the amount of damage in the tube. The porous material saves the amount of adhesive used to glue the particles. The foam also serves as a cushion between the mismatched modulus of the rigid particles, linking the flexible cement that is key for binding various substrates [17]. These key properties play a big role in strengthening the joint by absorbing and dissipating strain energies. This gives the water-borne adhesive of the sandcastle worm a multiscale energy absorbing system that helps the worm deal with the turbulent environment of the ocean. 1.2.3 Sandcastle Worm Adhesive Composition The composition of P. californica adhesive is known to be a proteinaceous, enriched with amino acids consisting of serine, glycine, lysine, and large amounts of 2,4- dihydroxyphenyl-L-alanine (DOPA) [18, 20, 27]. It is very similar in composition to the byssal adhesive of mussels [28-29], but much less complex. An amino acid and elemental analysis of the adhesive resulted in a mixture of three highly polar proteins Pc1, Pc2, and Pc3 with significant amounts of Ca2+ and Mg2+ [16, 18, 20]. Pc1 and Pc2 are positively charged proteins that are basic with pI > 9. The Pc1 protein consists of three residues, glycine (45 mol%), lysine (14 mol%), and tyrosine (19 mol%), and is highly repetitive and simple. The Pc2 is mostly histidine rich with lysine. Pc3 protein 8 exists in two variants (A, B) containing 4-13 serine residues with a single tyrosine residue. Because 95% of the serines in Pc3 are being phosphorylated, it is highly acidic, pI 0.5-1.5 [20]. This results in positively and negatively charged amino acids, with 30% phosphate sidechains and 10-20% amine sidechains in the worm's glue, as shown in Figure 1.3. Magnesium and calcium were other major components in the glue, with 4-5 times Mg to Ca and total cations to phosphate being a 1500 ppm ratio. 1.2.4 Sandcastle Worm Adhesive Curing Many mechanisms come to play to solidify the worm glue. The glue initially sets in 30 sec followed by a covalent crosslinking that takes up to hours. First the initial set of the glue seems to be triggered by a pH change, from pH <6 inside the sandcastle worm, to pH > 8 when released into the seawater. The insolubility of polyphosphate and divalent cations Mg2+/Ca2+ in seawater seem to suggest a pH triggering mechanism for the initial set [30]. The high amount of DOPA in the adhesive indicates that it readily undergoes oxidative covalent crosslinking. DOPA is known to be an adhesion promoter and facilitates solidification through di-DOPA, crosslinking in the byssal plaques of mussels [12]. The oxidation of DOPA occurs in alkaline sea water, giving rise to quinones that react further to crosslink adhesives proteins via aryl-aryl coupling (di-dopa formation) or possibly via Michael-type addition reaction with amine-containing protein residues [31-34]. This is also apparent when glue changes color, going from a whitish/clear appearance to a brownish coloring over time. DOPA is stable at pH 5 and converts to o-quinone at the pH approaches the pKa (9.4). 9 10 ® PC-3B: MKSFTIFAAILVALCYIQISEAG CCKRYSSSSYSSSSSSSSSSYSSSSSSSSYSSSSSSSS SYSSSSSSSSSSYSSSSSSYSSSSSSSYSSSSSSSSSS YSSSSSSYSSSSSSSSSYSSSSSSSSSYSSSSSSSSSS YSSSSSSYSSSSSSSY 5SSSSSSSSSYSSSSSSYSSSS SSSSSYSSSSSSSSSYSSSSSSSYSSSSSSYSSSSSSS SSSSYSSSSSSSSSSYSSSSSSYSSSSSSSSSYSSSSS SSSSYSSSSSSSSSSYSSSSSSSSSSSYSSSSSSYSSS SSSSYSSSSSSSSSSYSSSSSSSYSSSSSSSSSSSYSS SSSSSSSSSSSYSSS ©Pc-2: MKVLIFLATVAAVYG CGGAGG WRSGSCGG RWGHPAV--------HKALGGYG-G YGAHPAVHAAVHKALGGYGAGAYGAGA 0 E h3n WG-HPAV-- -HKALGGYGAGA WG-HPAV--- HKALGGYG-G YGAHPAVHVAVHKALGGYGAGACGHKTGGYGG YGAHP-- VAV-KA- AY-NHGFNYGANNAIKSTKRFGG YGAHP-- V-VKKAF SRGLS HGAY-AG SKAATGYGYGSGKAAGGYGY HN h 2c. c h 2 h2c \ / \I / ,ch 2 ,ch 2 Figure 1.3 Representative glue protein sequences. A) Sequence of polyacidic Pc3B. B) The serine residues (S) are more than 95% phosphorylated on the hydroxyl sidechain. The tyrosines (Y) are hydroxylated into DOPA residues. C) Sequence of polybasic Pc2. D) Structure of histidine (H) and lysine (K) residues with amine sidechains. 1.3 Biomimetic Complex Coacervate Adhesive 1.3.1 Sandcastle Worm Adhesive Complex Coacervate Model The compositions of the sandcastle worm adhesive, proteins consisting of oppositely charged polyelectrolytes at physiological pH, indicate a model based on complex coacervates [20]. This model explains the foamy structure, fluid character, low interfacial tension, and cohesive properties of the water-borne underwater adhesive of the sandcastle worm [20]. 1.3.2 Complex Coacervate In a colloidal system, separation of a liquid into two phases is called Coacervation [35]. The denser phase in the colloid is called the coacervate, while the other phase is the equilibrium solution. In the aqueous solution of two polymers, phase separation can occur if there is an electrostatic attraction. A complex coacervate is formed when coacervation occurs due to two oppositely charged colloids [36] (Figure 1.4). They could be positively and negatively charged macroions, such as polyelectrolytes, with balanced charges. The two phases coexist and are immiscible in solution. The coacervate phase, or the polymer-rich phase, is an isotropic liquid that contains amorphous particles that move relatively freely to each other. The second phase, known as the supernatant, is a very diluted phase. The two macroions are surrounded by a double layer, a region with increased concentration of counterions, with lower energy (the average distance between positive and negative charges is smaller than that between positives or between negatives), and low entropy (small ions have less translational freedom) [37]. When the two macroions mix, the double layer is destroyed and counterions are released in the form of salt, which shows that both the enthalpy and entropy of the system changes, thus driving the 11 12 A) B) Figure 1.4 Complex Coacervate Formation. A) Mixing solution of polycation and polyanions can lead to associative phase separation and formation of complex coacervate [37], B) Coacervate/supernatant after centrifugation of coacervate system: BSA-F (bovine serum albumin) + Poly(diallyldimethylammonium chloride), at pH 9.5 and I = 0.1 MNaCl [38], coacervation [37]. Tiebackx [39] was the first to notice the coacervation phenomena in 1911. But it was Bungenberg de Jong [40] and Kruyt [41] who first systematically studied it on a gum arabic-gelatin coacervate, and named it complex coacervate. There are many theories and models, like Voorn-Overbeek theory [42-45] (gelatin/acacia coacervate), Veis-Aranyi "dilute phase aggregate model" [45] (albumin/gelatin coacervate), Nakajima-Sato Model [46], and Tainaka model [47-48], which have tried to explain the coacervate process. Burgess tried to compare and resolve a lot of contradictions that exist in coacervates [49]. The complex coacervate formation is dependent on molecular weights, concentrations, and ratio of two interacting polyions and on the ionic strength, pH, and temperature of the media [49]. All of the theories agree on the suppression of coacervation at high ionic strength. The random coil configuration of both polyions plays an important role. The Voorn-Overbeek theory studied the gelatin/acacia coacervates and explained that the electrical attractive forces tend to accumulate on charged polyions and the entropy effects tend to disperse these forces. The bundles of oppositely charged polyelectrolytes associate together due to these electrostatic forces to form a coacervate. The loops between polymers entrap water in the coacervate, which gives rise to entropy, which allows the number of possible macromolecule arrangements to occur. According to Voorn and Overbeek a random coil of the polyions is necessary, if the polymers were completely folded, no water could be entrapped, and coacervation would be unlikely. The distributive nature of electrostatic interaction allows for overall electrical neutrality in the coacervate, yet the molecules are free to move around in the liquid phase [49]. The Veis- Aranyi model considers coacervates a two-step process rather than a spontaneous 13 process. They explain upon mixing of oppositely charged polyions, aggregates of low configurational entropy form, in which "coacervate sols" rearrange to form the coacervate phase [49]. This rearrangement can take from hours to days and is driven by gain in configurational entropy upon the formation of randomly mixed concentrations of coacervate phase and dilution of aggregate phase. Despite the contradictions all these theories could agree to the following about complex coacervates: complexes began to form before the phase separation occurs, even if there is an excess of one polyelectrolyte, the complex are only modestly charged, salt has a dissociating effect on the complexes, salt concentration is equal in the coacervate and supernatant phase, and in coacervate there is clear mobility of both polymers. By one polyelectrolyte carrying a positive and one carrying a negative charge restricts the complex coacervate formation occurring at a finite pH range. The coacervation phenomena are entropically driven. The coacervate and supernatant phases must be neutral or near neutrality, where this neutral complexes resembling gas-liquid separation in colloids. [50-51] There are many reasons that contribute to the stability of complex coacervates. The coulomb attraction (ion pairing) and the entropy increase due to counterion release are major driving forces for the formation, which includes hydrophobic effect, hydrogen bonding, and hydration forces [52]. Coacervates have low interfacial tension in water (~ 0.0005 dynes/cm) and exhibit ~ 0° contact angles [41]. The interfacial tension is important and is directly related to interaction between the macroions. In comlex coacervate core micelles the interfacial tension drives the formation of micelles and can be used to predict the critical aggregation concentration [53]. The interfacial tension is 14 very sensitive to the added salt [54]. There are many examples and applications that the complex coacervates exist in, in nature as well as industry. DNA is packed into small volume of DNA-binding proteins, in both eukaryotic and prokaryotic cells. This DNA compaction is largely due to the electrostatic attraction between oppositely charged macroions [55]. This phase separation can also occur in polyelectrolytes and oppositely charged colloids like micelles [56], proteins [57], and dendrimers [58]. In industry, coacervates have found applications in protein purification [59], drug and enzyme immobilization [60], cosmetic formulations [61], pharmaceutical microencapsulation [62], and in trapping organic plumes [63-64]. The Pc1, Pc2, and Pc3 proteins of the sandcastle worm are water-soluble polyelectrolytes. If the worm secretes them sequentially, it would be risky to loose them by dilution into the surrounding seawater [20]. That is why the complex coacervate method was proposed as a model for the worm. Coacervates can absorb the water from wet surfaces, and with their low interfacial tension they tend to spread easily over wet surfaces [65]. 1.3.3 Synthetic Analog: Complex Coacervate Adhesive Inspired by the sandcastle worm a synthetic analog to the underwater adhesive is made. The P. californica adhesive is of particular interest to us because of its ability to bond to wet surfaces, versatility in bonding to various particle substrates, and its effectiveness at low mortar-to-filler weight ratios [66-67]. This water-borne glue is able to displace surface bound water from the substrates, which is a big insight into its strong interfacial adhesion properties [13]. Its ability to hold together a robust shell capable of holding strong high-energy environments makes it an intriguing model for biomimetic 15 adhesive [13]. The synthetic adhesive was formed by the method of complex coacervation of glue protein analogs of the sandcastle worm [13-14]. The oppositely charged Pc3 and Pc1 protein analog polymers were made, containing the phosphate and amine sidechains, using a similar ratio to that found in the worm (Figure 1.5). These proteins were easily copied with poly(meth)acrylates. The copolymers were water-soluble, inexpensive, and scalable. The Pc3B polymethacrylate also contained the catechol, as shown in Figure 1.5. When mixed under the set conditions, the aqueous polymer solutions condensed into a complex coacervate at neutral pH [13]. The complex coacervate adhesive was chemically crosslinked through the oxidation of DOPA by NaIO4 to convert the catechol to dopaquinone [68]. This biomimetic complex coacervate adhesive was designed for gluing bone together. Our lab group was able to show that using the synthetic complex coacervate adhesive attained 40% strength of commercially available cyanoacrylate glue [13]. This water-borne adhesive has an advantage over other commercially available adhesives; our adhesive can be injectable under water and adheres to wet surfaces, whereas all glues fail under water eventually. 1.3.4 Multipart Copolyelectrolyte Model Sandcastle Worm Adhesive Recent work on the sandcastle worm shows that complex coacervation may not be playing a role in the natural adhesive formation [19]. It would be difficult for the worm to preform and premix the complex coacervate before secretion. The worm glue, instead, is a multipart polyelectrolyte adhesive. The oppositely charged proteins are packaged separately in highly concentrated granules, which are mixed as they leave the building 16 17 CH3 c h 3 B - fCH2- C -H -CH 2- C f c = o c = 0 NH c h 2 1 c h 2 | c h 2 1 c h 2 1 0 1 1 o 1 -CL- O OH NHo I c h 3 c = o r 1 i r 1 i -f-CH2-C-j---f-CH2-CH-]- c= o I NH ICH2 c h 2 Ic h 2 I + n h7 Figure 1.5 Synthetic analogs of glue proteins. A) of the Pc3B polymethacrylate analog copolymer. B) Structure of the polyamine analog copolymer. The analog polymers are random copolymers synthesized by free radical polymerization [30]. Reprinted from Advances in Colloid and Interface Science, 167 (1-2), R. J. Stewart, C. S. Wang, and H. Shao, Complex Coacervates as a Foundation for Synthetic Underwater Adhesives, 85-93, Copyright 2011, with permission from Elsevier. organ and have a "burst" release effect once in contact with seawater. Homogeneous granules contain sulfated macromolecules and Pc2/Pc5 protein. The heterogeneous granules contain Pc3A and Pc3B proteins along with divalent cations, Pc1 and Pc4, with both granules containing DOPA. Once they leave the building organ, the proteins form solid foam and fully set within 30 sec. The sandcastle worm would not have enough time to form a complex coacervate it is more a multipart copolyelectrolytes. 1.4 Fetal Defects The water-borne synthetic complex coacervate adhesive can be used in many fields of medicine. Adhering or binding soft tissues in wet environments is a major challenge. Most soft tissue adhesives are designed for dry applications that ultimately fail under water due to fluid in the joints. Applying the adhesive on wet surfaces and controlling its solidification is difficult. One such field where advanced soft tissue adhesive could be used in is gluing fetal tissue, in utero. 1.4.1 The Need for Fetal Tissue Adhesives Increased use of ultrasound scanning since the 1980s has led to the early detection of fetal defects. Advanced medical diagnostic techniques are able to detect congenital malformations earlier in pregnancy. Early detection has given rise to a large number of treatments and interventions. Fetal surgery is one of those promising therapeutic options for number of congenital malformations [69], where the field has grown from a concept to a medical subspecialty today [70]. In the past 2 decades advances have been made in fetal surgical interventions and fetal therapy by nonoperative means [69]. Fetal imaging, diagnosis, and anesthesia have allowed fetal interventions to be a vital tool for patients 18 who would otherwise face morbidity and mortality [70]. This makes minimal access fetal surgery possible where the fetal condition determines the invasiveness of the surgery. This is possible with laparoscopic surgery and fetoscopy. Fetal surgery, although very successful in a growing number of malformations, is limited in treatment due to conditions like preterm labor, chorionamniontic membrane separation, altered fetal homeostasis, and iatrogenic preterm premature rupture of the fetal membranes (iPPROM) [71]. Even in invasive procedure like fetoscopy, iPPROM is a big complication, resulting in amniotic fluid leakage. Fetoscopy is an endoscopic procedure during pregnancy that gives access to the fetus. Once the patient is diagnosed with iPPROM, the mother can barely carry the fetuses for longer than a few months [72]. In fetoscopic procedures there is a 6-45% rate where iatrogenic preterm premature rupture of the fetal membrane occurs [73]. All these associated risks involved give rise to morbidity and death, which compromises the expected benefits of such methods to begin with [73]. Many attempts have been made to close the ruptured fetal membranes but have been unsuccessful. The natural healing of human fetal tissue appears to be slow, if not absent, even in very small fetoscopic punctures. Histological studies after fetoscopic puncture defect of human membranes show no healing or growth in the tissue [74]. New innovative techniques to plug up the fetoscopic access site are being tested [75-76]. Intraamniotic injection at the puncture site of maternal platelets mixed with fibrin cryoprecipitate (amniopatch) has been successful, but the high platelets in the amniopatch has accounted for otherwise unexplained fetal deaths [77]. Dry collagen and gelatin plugs or liquid blood-derived sealants are being studied [78-79]. Cyanoacrylate adhesives, well 19 known for strong adhesion in surgical and traumatic wound repair [80], damaged the fetal tissue and disrupted the membrane structure [73, 81]. Commercially available Dermabond and Histoacryl adhesives were cytotoxic when in contact with fetal membranes [73]. Other PEG-based hydrogel polymers like SprayGel failed to bond to fetal membranes under wet conditions [73]. Adhesive that can glue in wet conditions, to plug the amniotic sac to prevent amniotic fluid leakage, and is biocompatible is needed. The iPPROM after a fetal surgery or invasive prenatal procedure is an unsolved clinical problem [73]. 1.5 Aim of This Research Adhesives for soft tissue repair, more so for sealing fetal defects, is needed. Bonding tissue is a major challenge in wet environments, and having something that adheres in utero and biocompatible is difficult. The aim of this research is to develop a bioadhesive, inspired by the sandcastle worm, to seal fetal defects, in particular designing the glue for the applications in two fetal conditions: twin-to-twin transfusion syndrome (TTTS) and spina bifida. The twin-to-twin transfusion syndrome is the unequal sharing of maternal blood in twin pregnancies. This syndrome requires fetoscopic laser surgery as one of its treatments, and an adhesive to plug the fetoscope-punctured membrane is needed. The second condition spina bifida is a congenital disorder, where the neural tissue of the fetus is exposed to the amniotic fluid causing neurological defects at birth. An adhesive patch to cover the neural tissue with a minimally invasive procedure, until birth, a more complex closure, can prolong a better outcome when the baby is born [82-83]. The objective of this work is to design a synthetic complex coacervate adhesive that can fit both of these applications, in utero repair. The bioadhesive has to be stable 20 21 and have the ability to be used in the practical application of the adhesive in these conditions. 1.5.1 Twin-to-Twin Transfusion Syndrome Twin-to-twin transfusion syndrome (TTTS) is a condition that is diagnosed during pregnancy by ultrasound. In pregnancies of twins, one-third of twins are monozygotic (MZ), and three-fourths of the MZ twins have presence of monochorionic diamniotic (MCDA) [84]. Twin pregnancies of MCDA placenta are at a high risk of TTTS that affects about 8-10% of pregnancies [85-87]. Currently TTTS occurs in approximately 1-3 per 10,000 births [88]. With TTTS, the two fetuses have unequal sharing of the mothers blood, which leads to asymmetrical fetal growth and fetal mortality, if left untreated. They share a placenta that contains abnormal blood vessels, where the blood supply from one baby to another is disproportional. The donor twin fetus receives less blood, which slows down its growth, and the recipient twin has excess blood, causing too much strain on the heart of the fetus. TTTS is a progressive disease in which sudden deterioration can occur leading to death of the fetuses, risk of miscarriage, brain damage, and morbidity [89-90]. This condition is diagnosed normally in second trimester of the pregnancy. Fetoscopic laser ablation is an effective treatment for TTTS, where a laser through a fetoscope coagulates the blood vessels as shown in Figure 1.6. The survival rates after fetoscopic laser surgery of TTTS are 50-70% [92]. The laser surgery for TTTS is a fetoscopic procedure with insertion of laser into the scope [70]. The risk of iPPROM is 10-30% procedure-associated fetal loss with laser [84, 93]. Sealing the defect site after fetoscopic laser ablation can reduce the perinatal morbidity and mortality. Although we are focusing on TTTS, designing the adhesive to seal the 22 uterus donor twin placenta crossing vessels spine fetoscope with laser amniotic fluid pubic bone *-- vagina cervix Figure 1.6 Fetoscopic laser ablation for twin-to-twin transfusion syndrome treatment [91], Reprinted with permission from so+gi. 23 defect for this application can be a model sealing all fetal tissues for other conditions. 1.5.2 Spina Bifida Spina Bifida, "split spine," is a developmental congenital disorder where the fetus neural tube is left unclosed, affecting 1,500 babies a year [94]. Myelomeningocele (MMC) is the most severe case of Spina Bifida (Figure 1.7), where the closure defect protrudes and bulges out of the posterior spinal column. The MMC is a severe malformation that can result in disability at birth and be a major challenge to fix in fetal repair [95]. The condition is detected 16 to 20 weeks of pregnancy. The exposed neural tissue undergoes progressive damage with advancing gestation age due to being in contact with amniotic fluid [96]. The fetus also develops a Chiari II condition of the brain, resulting in irreversible neurological impairments at birth, from the pressure disturbance and loss of cerebrospinal fluid [97]. At birth a number of defects result due to MMC: paraplegia, sphincter incontinence hydrocephalus, cranial nerve disturbances, respiratory problems, and death [98]. The treatment for MMC repair is challenging, with high risk of maternal and fetal morbidity and mortality. The surgical procedure could face difficulties like iPPROM, uterine rupture, maternal hemorrhage, and hysterectomy. The first intrauterine surgery repair of fetal myelomeningocele was performed on humans at Vanderbilt University Medical Center in Nashville, Tennessee, USA, in 1994 [100-101]. It was a very difficult and risky procedure, with a lot of research underway to improve the method of treatment [102]. The intrauterine surgery requires the defect to be closed in multilayer fashion, with neural dissection, dural closure, and suturing of the spinal cord, which increases the operating time as well [103]. Animal studies have shown that repair of neural tube defect 24 Figure 1.7 Illustration of a child with Myelomeningocele (MMC). [99] Reprinted with permission from so+gi. in the womb could result in a less severe hindbrain and spinal cord injuries at birth [82, 83, 104]. The multilayer closure has been reported in fetoscopic repair, but the method is technically very demanding and time consuming [102]. Less invasive surgical technique that would just cover up the defect till birth, followed by a more complex MMC repair could be an effective treatment. An adhesive patch to cover up the defect through a minimally invasive method would be an ideal scenario for this application. Currently there is nothing out there that can adhere the patch under aqueous conditions to the spinal column of the fetus. 1.6 Outline of This Thesis This thesis consists of research in developing a synthetic complex coacervate adhesive for sealing fetal defects in utero. The biomimetic adhesive is designed for two applications of TTTS and spina bifida. The research work consists of preliminary studies to prove that the synthetic complex coacervate adhesive can be used to seal fetal membranes, followed by a more in depth approach to making an adhesive composition that is more stable and used in practical applications and taking this adhesive from design and synthesis to animal studies. In Chapter 2 a multiphase adhesive complex coacervate with increased bond strength was developed. Polyethylene glycol diacrylate was entrapped into the coacervate, creating a second polymer network via crosslinking, which helps aid in increased shear bond strength. The rheological flow behavior of the complex coacervate adhesive was extensively studied. Shear-thinning behavior without destructing the coacervate network is an important property for an injectable system. In Chapter 3 the high bond strength adhesive was tested in conjunction with a 25 fetal membrane patch to model the adhesive as a potential sealant for fetoscopic procedures. The adhesive was tested with the in vitro model, mimicking the wall-fetal membrane. The cytotoxicity of the adhesive was tested with direct contact to human fetal membranes. This preliminary study was key in taking the adhesive to the next phase of animal studies. In Chapter 4 a complex coacervate adhesive was developed using methacrylated polyphosphate and polyamine polymers. The synthesis methods used were explained. Aqueous polymerization and grafting methods were evaluated. Reversible Addition- Fragmentation Chain Transfer (RAFT) was used for polyamine polymer. The complex coacervate system was studied to tailor the properties to the application. In Chapter 5 the mechanical properties of the methacrylated complex coacervate adhesive was tailored to the TTTS application. A chemically crosslinked coacervate adhesive was designed for this application. The crosslinking kinetics, bond strengths, and stability were studied. The sterile complex coacervate adhesive packets were prepared for the pig animal studies. In Chapter 6 the mechanical properties of the methacrylated complex coacervate adhesive was tailored for the spina bifida application. A photocrosslinked coacervate adhesive patch was designed for this application. The crosslinking kinetics, bond strengths, and swelling behavior were studied. The sterile complex coacervate adhesive packets were prepared for sheep animal studies. In the concluding chapter findings are summarized. The impact of our study and its contribution to soft tissue adhesives, more specifically fetal tissue adhesive is discussed. The future direction of this study is also discussed. 26 27 1.7 References 1. A. J. Kinloch, Adhesion and Adhesives: Science and Technology. Chapman and Hall, London, 1987. 2. N. M. Bikales, Adhesion and Bonding. Wiley-Interscience, New York, 1971. 3. P. A. Mazza, "A New Palaeolithic Discovery: Tar-hafted Stone Tools in a European Mid-Pleistocene Bone-bearing Bed," J. Archaeol. Sci., 33 [9] 1310-1318 (2006). 4. A. Berg, F. Peters, and M. Schnabelrauch, Biological Adhesive System: From Nature to Technical and Medical Application. Edited by I. Grunwald, J. von Byern, Springer, New York, 2010. 5. R. J. Stewart, T. C. Ransom, and V. Hlady, "Natural Underwater Adhesives," J. Polym. Sci. B Polym. Phys, 49 [11] 757-771 (2011). 6. A. V. Pocius, Adhesion and Adhesive Technology: An Introduction. 2nd ed. Hanser Gardner, Cincinnati, 2002. 7. P. J. Frantzis, "Durability of Adhesive Joints Made Underwater," J. Mater. Civ. Eng., 20 [10] 635-639 (2008). 8. W. E. Cloete, and W. W. Focke, "Fast Underwater Bonding to Polycarbonate Using Photoinitiated Cyanoacrylate," J. Adhes. Adhes., 30 [4] 208-213 (2010). 9. P. Fratzl, "Biomimetic Materials Research: What Can We Really Learn From Nature's Structural Materials?" J. R. Soc. Inter., 4 [15] 637-642 (2007). 10. D. J. Crisp, G. Walker, G. A. Young, and A. B. Yule, "Adhesion and Substrate in Mussels and Barnacles," J. Colloid Interface Sci., 104 [1] 40-50 (1985). 11. J. L. Dalsin, B. H. Hu, B. P. Lee, and P. B. Messersmith, "Mussel Adhesive Protein Mimetic Polymer for the Preparation of Nonfouling Surfaces," J. Am. Chem. Soc., 125 [14] 4253-4258 (2003). 12. J. H. Waite, and X. Qin, "Polyphosphoprotein From the Adhesive Pads of Mytilus Edulis," Biochem, 40 [9] 2887-2893 (2001). 13. H. Shao, K. N. Bachus, and R. J. Stewart, "A Water-borne Adhesive Modeled After the Sandcastle Glue of P. Californica," Macromol. Biosci., 9 [5] 464-471 (2009). 14. H. Shao, and R. J. Stewart, "Biomimetic Underwater Adhesives With Environmentally Triggered Setting Mechanisms," Adv. Mater., 22 [6] 729-733 28 (2010). 15. J. H. Waite, "Adhesion in Byssally Attached Bivalves," Biol. Rev., 58 [2] 209-231 (1983). 16. H. Zhao, C. Sun, R. J. Stewart, and J. H. Waite, "Cement Proteins of the Tube-building Polychaete Phragmatopoma Californica," J. Biol. Chem., 280, 42938-42944 (2005). 17. T. J. Deming, "Mussel Byssus and Biomolecular Materials," Curr. Opi. Chem. Bio., 3 [1] 100-105 (1999). 18. J. H. Waite, R. A. Jensen, and D. E. Morse, "Cement Precursor Proteins of the Reef-building Polychaete Phragmatopoma Californica (Fewkes)," Biochem., 31 [25] 5733-5738 (1992). 19. C. S. Wang, and R. J. Stewart, "Multipart Copolyelectrolyte Adhesive of the Sandcastle Worm, Phragmatopoma Californica (Fewkes): Catechol Oxidase Catalyzed Curing Through Peptidyl-DOPA," Biomac., 14 [5] 1607-1617 (2013). 20. R. J. Stewart, J. C. Weaver, D. E. Morse, and J. H. Waite, "The Tube Cement of Phragmatopoma Californica: A Solid Foam," J. Exp. Bio., 207, 4727-4734 (2004). 21. K. Simkiss, "The Process of Biomineralization in Lower Plants and Animals - An Overview," In Biomineralization in Lower Plants and Animals, 30, 19-37 (1986). 22. B. J. Endrizzi, and R. J. Stewart, "Glueomics: An Expression Survey of the Adhesive Gland of the Sandcastle Worm," J. Adhes., 85, 546-559 (2009). 23. M. J. Steven, R. E. Steren, V. Hlady, and R. J. Stewart "Multiscale Structure of the Underwater Adhesive of Phragmatopoma Californica: A Nanostructured Latex With a Steep Microporosity Gradient," Langmuir, 23 [9] 5045-5049 (2007). 24. J. Vovelle, "The Tube of Salbellaria Alveolata," Arch. Zool. Exp. Gen., 106, 1-187 (1965). 25. L. J. Gibson, and M. F. Ashby, Cellualr Solids: Structure and Properties. Cambridge University Press, Cambridge, UK, 1997. 26. C. S. Wang, K. K. Svendsen, and R. J. Stewart, "Morphology of the Adhesive System in the Sandcastle Worm, Phragmatopoma Californica," Biological Adhesive Systems. Editors J. V. Byern, I. Grunwalds. Springer, New York, 2010. 27. R. A. Jensen, and D. E. Morse, "The Bioadhesive of Phragmatopoma Californica 29 Tubes: a Silk-line Cement Containing L-DOPA," J. Comp. Physiol. B, 158, 317-324 (1988). 28. J. H. Waite, and A. C. Rice-Ficht, "Presclerotized Eggshell Protein From Liver Fluke FasciolaHepatica," Biochem., 26 [24] 7819-7825 (1987). 29. J. H. Waite, "Adhesion a La Moule," Integ. Comp. Biol, 42 [6] 1172-1180 (2002). 30. R. J. Stewart, C. S. Wang, and H. Shao, "Complex Coacervates as a Foundation for Synthetic Underwater Adhesives," Adv. Colloid. Interface Sci., 167 [1-2] 85-93 (2011). 31. M. Yu, J. Hwang, and T. J. Deming, "Role of L-3,4-Dihydroxyphenylalaine in Mussel Adhesive Protein," J. Am. Chem. Soc., 121, 5825-5826 (1999). 32. L. A. Burzio, and J. H. Waite, "Cross-linking in Adhesive Quinoprotein: Studies With Model Decapeptides," Biochem, 39 [36] 11147-11153 (2000). 33. S. Haemers, G. J. M. Koper, and G. Frens, "Effect of Oxidation Rate on Cross­linking Mussel Adhesive Protein," Biomac., 4 [3] 632-640 (2003). 34. J. Monahan, and J. J. Wilker, "Cross-linking the Protein Precursor of Marine Mussel Adhesives: Bulk Measurements and Reagents Curing," Langmuir, 20 [9] 3724-3729 (2004). 35. IUPAC compendium of Chemical Technology 1997. 36. C. G. De Kruif, F. Weinbreck, and R. De Vries, "Complex Coacervation of Proteins and Anionic Polysaccharides," Curr. Opin. in Coll. & Interfa. Sci., 9 [5] 340-349 (2004). 37. J. V. D. Gucht, E. Spruijt, M. Lemmers, and M. A. C. Stuart, "Polyelectrolyte Complexes: Bulk Phases and Colloidal Systems," J. of Coll. and Inter. Sci., 361 [2] 407-422 (2011). 38. C. L. Cooper, P. L. Dubin, A. B. Kayitmazer, and S. Turksen, "Polyelectrolyte-protein Complexes," Curr. Opin. Coll. Inter. Sci., 10 [1-2] 52-78 (2005). 39. F. W. Tiebackx, "Gleichzeitige Ausflockung Zweier Kolloide," Chem. Ind. Kolloide, 8, 198-201 (1911). 40. H. G. Bungenberg de Jong, "Crystallization-coacervation-flocculation," Coll. Sci., Elsevier Publishing Company, Amsterdam, 2, 232-255 (1949). 41. H. J. Bungenberg de Jong, and H. R. Kruyt, "Coacervation (Partial Miscibility in 30 Colloid Systems)," Proc. K. Ned. Akad. Wet., 32, 849-856 (1929). 42. M. J. Voorn, "Complex Coacervation. V Experiments on the Distribution of Salts. Comparison of Theory and Experiments," Rec. Trav. Chim., 75, 317-925 (1956). 43. J. T. Overbeek, and M. J. Voorn, "Phase Separation in Polyelectrolyte Solution; Theory of Complex Coacervation," J. Cell. Comp. Physiol., 49, 7-22 (1957). 44. M. J. Voorn, "Phase Separation in Polymer Solutions," Fortsch Der Hochpoly- Forsch, Springer, Berlin, 192- 233, 1959. 45. A. Veis, and C. Aryani, "Phase Separation in Polyelectrolyte Systems. I. Complex Coacervates of Gelatin," J. Phys. Chem., 64, 1203-1210 (1960). 46. A. Nakajima, and H. Sato, "Phase Relationships of an Equivalent Mixture of Sulfated Polyvinyl Alcohol and Aminoacetalyzed Polyvinyl Alcohol in Microsalt Aqueous Solution," Biopolymers, 10, 1345-1355 (1972). 47. K. Tainaka, "Study of Complex Coacervation in Low Concentration by Viral Expansion Method," J. Phys. Soc. Japan, 46, 1899-1906 (1979). 48. K. Tainaka, "Effect of Counterions on Complex Coacervation," Biopoly., 19, 1289-1298 (1980). 49. D. J. Burgess, "Practical Analysis of Complex Coacervate Systems," J. Coll. Inter. Sci, 140, 227-238 (1990). 50. D. J. Burgess, and J. E. Carless, "Microelectrophoretic Studies of Gelatin and Acacia for the Prediction of Complex Coacervation," J. Coll. Inter. Sci, 98, 1-8 (1984). 51. C. Schmitt, C. Sanchez, F. Thomas, and J. Hardy, "Complex Coacervation Between p-lactoglobulin and Acacia Gum in Aqueous Media," Food Hydrocoll, 13, 483-496 (1999). 52. E. Spruijt, M. A. C. Stuart, and J. van der Gucht, "Dynamic Force Spectroscopy of Oppositely Charged Polyectrolyte Brushes," Macromolecules, 43, 1543-1550 (2010). 53. E. Spruijt, J. Sprakel, M. A. C. Stuart, and J. Van der Gucht, "Interfacial Tension Between a Complex Coacervate Phase and its Coexisting Aqueous Phase," Soft Matter, 6, 172-178 (2010). 54. H. Sato, and A. Nakajima, "Complex Coacervation in Sulfated Polyvinyl Alcohol-aminoacetalyzed Polyvinyl Alcohol System 0.2, Formation of 31 Coacervate Droplets," ColloidPolym. Sci., 252, 944-948 (1974). 55. D. J. Clark, and T. J. Kimura, "Electrostatic Mechanism of Chromatin Folding," J. Mol. Biol, 211 [4] 883-896 (1990). 56. Y. Wang, K. Kimura, and P. L. Dubin, "Polyelectrolyte-Micelle Coacervation: Effects of Micelle Surface Charge Density, Polymer Molecular Weight, and Polymer/Surfactant Ratio," Macromolecules, 33, 3324-3331 (2000). 57. K. Kaibara, T. Okazaki, H. B. Bohidar, and P. L. Dubin, "pH-induced Coacervation in Complexes of Bovine Serum Albumin and Cationic Polyelectrolytes," Biomacromolecules, 1, 100-107 (2000). 58. D. Leisner, and T. Imae, "Interpolyelectrolyte Complex and Coacervate Formation of Poly (Glutamic Acid) With a Dendrimer Studied By Light Scattering and SAXS," J. Phys. Chem. B, 107, 8078-8087 (2003). 59. W. L. Hinze, and D. W. Armstrong, editors, Ordered Media in Chemical Separation, American Chemical Society, Washington DC, 1987. 60. J. Xia, K. Mattison, V. Romano, P. L. Dubin, and B. B. Muhoberac, "Complexation of Trypsin and Alcohol Dehydrogenase With Poly(Diallyldimethylammonium Chloride)," Biopolymers, 41 [4] 359-365 (1997). 61. E. D. Goddard, "Polymer/Surfactant Interaction," J. Soc. Cosmet. Chem, 41, 23-49 (1990). 62. P. B. Deasy, Microencapsulation and Related Drug Process, Marcel Dekker, New York, 1984. 63. Y. Wang, J. Banziger, P. L. Dubin, G. Filipelli, and N. Nuraje, "Adsorptive Partitioning of an Organic Compound Onto Polyelectrolyte-immobilized Micelles on Porous Glass and Sand," Environ. Sci. Technol., 35 [12] 2608-2611 (2001). 64. S. S. Singh, V. K. Aswal, and H. B. Bohidar, "Structural Studies of Agar-gelatin Complex Coacervates by Small Angle Neutron Scattering, Rheology and Differential Scanning Calorimetry," Intern. J. Bio. Macrom., 41 [3] 301-307 (2007). 65. H. G. Bungenberg De Jong, "Morphology of Coacervates," Colloid Science, Elsevier Publishing C, Amsterdam, 2, 433-482 (1949). 66. R. A. Jensen, and D. E. Morse, "Intraspecific Facilitation of Larval Recruitment: Gregarious Settlement of the Polychaete Phragmatopoma Californica (Fewkes)," J. Exp. Mar. Bio. Ecol., 83 [2] 107-126 (1984). 32 67. Y. Gruet, J. Vovelle, and M. Grasset, "Composante Biominerale Du Ciment Du Tube Chez Sabellaria Alveolata (L.), Annelid Polychete," Can. J. Zool., 65, 837-842 (1987). 68. H. Shao, G. M. Weerasekare, and R. J. Stewart, "Controlled Curing of Adhesive Complex Coacervates With Reversible Periodate Carbohydrate Complexes," J. of Biomed. Mat. Res. A, 97 [1] 46-51 (2011). 69. E. Danzer, R. M. Sydorak, M. R. Harrison MR, and C. T. Albanese, "Minimal Access Fetal Surgery," European Journal of Obstetrics & Gynecology and Reproductive Biology, 108 [1] 3-13 (2003). 70. J. Farrell, and L. J. Howell, "An Overview of Surgical Techniques, Research Trials, and Future Directions of Fetal Therapy," Journal of Obstetric, Gynecologic, & Neonatal Nursing, 41 [3] 419-425 (2012). 71. C. T. Albanese, and M. R. Harrison, "Surgical Treatment for Fetal Disease. The State of the Art," Annals of the New York Academy of Sciences, 847 [1] 74-85 (1998). 72. R. A. Quintero, W. J. Morales, M. Allen, P. W. Bornick, J. Arroyo, and G. LeParc, "Treatment of Iatrogenic Previable Premature Rupture of Membranes With Intra-amniotic Injection of Platelets and Cryoprecipitate (Amnio-patch): Preliminary Experience," Am. J. Obstet. Gynecol, 181, 744-749 (1999). 73. G. Bilic, C. Brubaker, P. B. Messersmith, A. S. Mallik, T. M. Quinn, C.Haller, E. Done, L. Gucciardo, et al., "Injectable Candidate Sealants for Fetal Membrane Repair: Bonding and Toxicity in Vitro," Am. J. Obstet. Gynecol., 202 [1] 85.e1-9 (2010). 74. E. Gratacos, J. Sanin-Blair, L. Lewi, N. Toran, G. Verbist, L. Cabero, and J. Deprest, "A Histological Study of Fetoscopic Membrane Defects to Document Membrane Healing," Placenta, 27 [4] 452-456 (2006). 75. R. Devlieger, L. K. Millar, G. Bryant-Greenwood, L. Lewi, and J. A. Deprest, "Fetal Membrane Healing After Spontaneous and Iatrogenic Membrane Rupture: A Review of Current Evidence," Am. J. Obstet. Gynecol., 195 [6] 1512-1520 (2006). 76. A. H. Zisch, and R. Zimmermann, "Bioengineering of Foetal Membrane Repair," Swiss Med Wkly, 138 [41-42] 596-601 (2008). 77. R. A. Quintero, "Treatment of Previable Premature Ruptured Membranes," Clinnics in Perinatology, 30 [3] 573-589 (2003). 78. J. Chang, T. F. Tracy Jr, S. R. Carr, D. L. Sorrells Jr, and F. I. Luks, "Port 33 Insertion and Removal Techniques to Minimize Premature Rupture of the Membranes in Endoscopic Fetal Surgery," J. Pediatr. Surg, 41 [5] 905-909 (2006). 79. B. K. Young, A. S. Roman, A. P. MacKenzie, C. D. Stephenson, V. Minior, A. Rebarber, and I. Timor-Tritsch, "The Closure of Iatrogenic Membrane Defects After Amniocentesis and Endoscopic Intrauterine Procedures," Fetal Diagn. Ther., 19 [3] 296-300 (2004). 80. P. A. Leggat, D. R. Smith, and U. Kedjarune, "Surgical Applications of Cyanoacrylate Adhesives: A Review of Toxicity," ANZ J. Surg, 77 [4] 209-213 (2007). 81. M. L. Oyen, S. E. Calvin, and D. V. Landers, "Premature Rupture of the Fetal Membranes: Is the Amnion the Major Determinant?" Am. J. Obstet. Gynecol., 195 [2] 510-515 (2006). 82. B. W. Paek, D. L. Farmer, C. C. Wilkinson, C. T. Albanese, W. Peacock, M. R. Harrison, and R. W. Jennings, "Hindbrain Herniation Develops in Surgically Created Myelomeningocele but is Absent After Repair in Fetal Lambs," Am. J. Obstet. Gynecol, 183 [5] 1119-11123 (2000). 83. S. Bouchard, M. G. Davey, N. E. Rintoul, D. S. Walsh, L. B. Rorke, and N. S. Adzick, "Correction of Hindbrain Herniation and Anatomy of the Vermis after in Utero Repair of Myelomeingocele in Sheep," J. Pediatr. Surg, 38 [3] 451-458 (2003). 84. L. L. Simpson, "Twin-Twin Transfusion Syndrome," Am. J. Obstet. Gyneco., 208 [1] 3-18 (2013). 85. L. Lewi, J. Jani, I. Blickstein, A. Huber, L. Gucciardo, T. V. Mieghem, E. Done, et al., "The Outcome of Monochorionic Diamniotic Twin Gestations in the Era of Invasive Fetal Therapy: A Prospective Cohort Study," Am. J. Obstet. Gynecol., 199 [5] 514-e1 (2008). 86. G. A. Machin, and L. G. Keith, "Can Twin-to-twin Transfusion Syndrome be Explained, and How is it Treated?" Clin. Obstet. Gynecol., 41 [1] 104-113 (1998). 87. R. Acosta-Rojas, J. Becker, B. Munoz-Abellana, C. Ruiz, E. Carreras, and E. Gratacos, "Twin Chorionicity and the Risk of Adverse Perinatal Outcome," Int. J. Gynecol. Obstet, 96 [2] 98-102 (2007). 88. I. Blickstein, "Monochorionicity in Perspective," Ultrasound Obstet. Gynecol., 27 [3] 235-238 (2006). 34 89. F. Haverkamp, C. Lex, C. Hanisch, H. Fahnenstich, and K. Zerres, "Neurodevelopmental Risks in Twin-to-twin Transfusion Syndrome: Preliminary Findings," Europ. J. Paediatr. Neurol., 5 [1] 21-27 (2001). 90. I. Blickstein, "The Twin-twin Transfusion Syndrome," Obstet. Gynecol, 76 [4] 714-722 (1990). 91. http://www.sogi.net.au/mintdigital.net/SOGI.aspx?XmlNode=/Services/Fetal+Tre atment/Laser+surgery 92. D. Roberts, S. Gates, M. Kilby, and J. P. Neilson, "Interventions for Twin-twin Transfusion Syndrome: a Cochrane Review," Ultrasound Obstet. Gynecol., 31 [6] 701-711 (2008). 93. R. A. Quintero, C. Comas, P. W. Bornick, M. H. Allen, and M. Kruger, "Selective Versus Non-selective Laser Photocoagulation of Placental Vessels in Twin-to-twin Transfusion Syndrome," Ultrasound Obstet. Gynecol., 16 [3] 230-236 (2000). 94. Center of Disease Control and Prevention: http://www.cdc.gov/ncbddd/spinabifida/data.html 95. S. Hirose, C. Meuli -Simmen, and M. Meuli, "Fetal Surgery for Myelomeningocele: Panacea or Peril?" World J. Surg, 27 [1] 87-94 (2003). 96. D. S. Heffez, J. Aryanpur, N. A. Rotellini, G. M. Hutchins, and J. M. Freeman, "Intrauterine Repair of Experimental Surgically Created Dysraphism," Neurosurgery, 32 [6] 1105-1110 (1993). 97. D. G. McLone, and P. A. Knepper, "The Cause of Chiari II Malformation: A Unified Theory," Pediatr. Neurosur., 15 [1] 1-12 (1989). 98. R. M. Bowman, D. G. McLone, J. A. Gran, T. Tomita, and J. A. Ito, "Spina Bifida Outcome: a 25 year Prospective," Pediatr. Neurosur., 34 [3] 114-120 (2001). 99. http://www.glogster.com/stevko/spina-bifida-genetics/g- 6l6ma4jtukd42g0jfaa89a0 100. J. P. Bruner, W. O. Richards, N. B. Tulipan, and T. L. Arney, "Endoscopic Coverage of Fetal Myelomeningocele in Utero," Am. J. Obstet. Gynecol, 180 [1] 153-158 (1999). 101. J. P. Bruner, "Intrauterine Surgery in Myelomeningocele," In Seminars in Fetal & Neonatal Medicine, 12 [6] 471-476 (2007). 35 102. D. I. Farmer, C. S. Von Koch, W. J. Peacock, M. Danielpour, N. Gupta, and H. Lee, "In Utero Repair of Myelomeningocele: Experimental Pathophysiology, Initial Clinical Experience, and Outcomes," Arch. Surg., 138 [8] 872-878 (2003). 103. R. Barini, M. W. G. Barreto, K. Cursino, H. Zambelli, A. Prando, and L. Sbragia, "Abruptio Placentae During Fetal Myelomeningocele Repair," Fetal Diagnosis and Therapy, 21 [1] 115-117 (2006). 104. M. Meuli, C. Meuli-Simmen, C. D. Yingling, G. M. Hutchins, K. M. Hoffman, M. R. Harrison, and N. S. Adzick, "Creation of Myelomeningocele in Utero: A Model of Functional Damage From Spinal Cord Exposure in Fetal Sheep," J. Pediatr. Surg, 30 [7] 1028-1033 (1995). CHAPTER 2 MULTIPHASE ADHESIVE COACERVATES INSPIRED BY THE SANDCASTLE WORM 2.1 Abstract Water-borne, underwater adhesives were created by complex coacervation of synthetic copolyelectrolytes that mimic the proteins of the natural underwater adhesive of the sandcastle worm. To increase bond strengths, we created a second polymer network within cross-linked coacervate network by entrapping polyethylene glycol diacrylate (PEG-dA) monomers in the coacervate phase. Simultaneous polymerization of PEG-dA and crosslinking of the coacervate network resulted in maximum shear bond strengths of ~ 1.2 MPa. Approximately 40% of the entrapped PEG-dA polymerized based on attenuated total reflectance-Fourier transform infrared spectroscopy. The monomer-filled coacervate had complex flow behavior, thickening at low shear rates and then thinning suddenly with a 16-fold drop in viscosity at shear rates near 6 s-1. The microscale structure of the complex coacervates resembled a three-dimensional porous network of interconnected tubules. The sharp shear thinning behavior is conceptualized as a structural reorganization between the interspersed phases of the complex coacervate. The Adapted from a pre-peer-reviewed version with permission from S. Kaur, G. M. Weerasekare, R. J. Stewart, "Multiphase adhesive coacervates inspired by the Sandcastle worm," ACSAppl. Mater. Inter., 3 [4] 941-4 (2011). Copyright 2011 American Chemical Society. 37 bond strength and complex fluid behavior of the monomer-filled coacervates have important implications for medical applications of the adhesives. 2.2 Introduction Adhesive bonding in watery environments with common synthetic adhesives is confounded, in general, by poor interfacial adhesion leading to eventual failure by infiltration of water into the joint. Aquatic environments are populated with diverse organisms that have evolved a multitude of workable solutions to the underwater adhesion problem. Natural underwater adhesives have therefore been studied as potential sources of materials or concepts with the goal of creating or improving synthetic adhesives for wet applications, including repair of living tissues. One such model is the underwater adhesive of the Sandcastle worm (Phragmatopoma californica), a marine polychaete [1-3]. The Sandcastle worm employs an ingenious strategy to construct composite mineralized shells; the mineral phase is gathered from its environment preformed as sand grains and shell fragments that are then glued together with small dabs of an underwater adhesive [1]. The sandcastle worm glue is comprised of oppositely charged proteins and divalent cations [2, 3]. Copolyelectrolytes with the same chemical side chains (phosphates and amines) and in the same proportions as the natural proteins were synthesized. When mixed under the right conditions, the synthetic copolyelectrolytes condensed into fluid complex coacervates [4, 5]. As the basis for underwater adhesives, complex coacervates have several ideal properties: the dense, phase-separated fluids sink in water, are sufficiently cohesive that they do not mix with water on a time scale of several minutes, and readily adhere to wet surfaces, all of which allows the adhesive to stay in place where it is applied underwater. The adhesive becomes load-bearing by triggered solidification of the complex coacervate after application. The sandcastle worm glue sets within 30 s through pH-triggered solidification of a polyphosphorylated protein and Ca2+ and Mg2+ [2, 6]. The initial set is followed up over several hours by covalent crosslinking through 3,4-dihydroxy-L-phenylalanine (dopa) residues. Both the pH-triggered set and dopa-mediated crosslinking were replicated in biomimetic adhesive coacervates [4, 5]. Though the natural sandcastle glue is suitable for the dimensions and lifecycle of sandcastle worms, it is not particularly strong, around 300 kPa [7]. Biomimetic adhesives will have to be much stronger than the natural adhesive to find broad utility. Incorporation of micro- or nanophases into the bulk adhesive phase is a well-known strategy for increasing adhesive bond strengths [8]. Our strategy for incorporation of an additional phase into our biomimetic adhesive was to form the coacervate in the presence of a water-soluble neutral monomer, as a first example, polyethylene glycol-diacrylate (PEG-dA). Polymerizable monomers dissolved in the aqueous copolyelectrolyte solutions become incorporated into the dense coacervate phase, which is mostly water by weight. Polymerization created a second polymer network within the coacervated copolyelectrolyte network. The coacervate functioned, in effect, as a container for the polymerizable monomers that could be accurately delivered underwater before polymerization of the second internal polymer network was initiated. 38 39 2.3 Materials and Methods 2.3.1 Materials All reagents were used without further purification unless noted otherwise. Phosphorous oxychloride (POCl3, 98%), 2-hydroxyethylmethacrylate (HEMA, 97%), and triethylamine (99%) were purchased from VWR. The 2,2'-azobisisobutyronitrile (AIBN) was purchased from Polysciences. Ultra filters Pellicon Ultracel Membranes by Millipore were used. N-(3-Aminopropyl) Methacrylamide, Hydrochloride, and Acrylamide (Chemzymes, ultra pure) were purchased from Polysciences. PEG-dA (Polyethylene glycol diacrylate, 760 Da) was purchased from sigma-aldrich. 2.3.2 Monomer Synthesis 2-(methacryloyloxy)ethyl phosphate (MOEP) was synthesized by adding phosphorus oxychloride (16.8g, 110 mmol) under argon to a stirred solution of 2- hydroxyethyl methacrylate (12g, 92 mmol) in toluene (340 mL). The reaction mixture was cooled to 0°C, and triethylamine (39 mL, 276 mmol) was added. The reaction proceeded at 0°C for 30 mins, then at room temperature for 6 hrs. The white solid precipitate was recovered by filtration. Water (240 mL) was added to the filtrate and stirred overnight. The two layers were separated, and the aqueous phase was acidified and then extracted with THF: Ether (1:2, 6x225 mL). The organic phases were combined, dried over Na2SO4, and solvent evaporated to obtain the product as a pale yellow oil (67%, 12.2g). 1H NMR spectroscopy (300 MHz, D2O) d 1.7 (3H, s), 4.0 (2H, m, POCH2), 4.2 (2H, m, POCH2CH2), 5.5 (1H,s) 6.0 (1H, s); 13C NMR (75 MHz, D2O) d 17.4, 64.2 (d, 2Jpoc = 8.3 Hz), 64.4 (d, J p o c c = 5.5Hz), 127.2, 135.6, 169.4; 31P NMR (120 MHz, D2O) d 0.97 (s). Dopamine methacrylamide (DMA) was synthesized as 40 previously described [4]. 2.3.3 Copolyelectrolyte Synthesis Poly(MOEP-co-DMA) was synthesized as previously described [4] by free radical polymerization of MOEP and DMA initiated with AIBN in methanol. The polymerization proceeded at 55oC for 16 hours (Figure 2.1). The copolymer was precipitated with acetone and then washed twice with acetone to remove residual monomers. The polymer was then dissolved in water and ultrafiltered on pellicon ultracel membranes with MWCO 1000 kDa followed by filtration with MWCO of 5 kDa. The concentrations of phosphate and o-DHP side chains were determined by NMR and UV/vis spectroscopy and were 76 and 19 mol%, respectively. The MW (64 kDa) and PDI (2.8) of the copolymer were determined by size exclusion chromatography (SEC) on an AKTA FPLC system with a Superose 6 HR 10/300 column (GE Healthcare) in 0.05 M phosphate and 0.15 M NaCl (pH 7.4). Poly(acrylamide-co-aminopropyl methacrylamide) was synthesized by free radical polymerization of 90 mol% acrylamide and 10 mol% N-(3-amino-propyl) methacrylamide hydrochloride (Figure 2.2), as previously described [4]. The copolymer was purified by dialysis for 3 days and lyophilized. The amine concentration (mol/mg) was determined with ninhydrin using glycine as the standard. The MW and PDI, determined by SEC in 0.5 M NaCl and 0.1 M NH4CH3CO2 on Superdex 200 column (GE Healthcare), were 288 kDa and PDI 1.36. 41 OH OH Figure 2.1 Synthesis schematic of Poly(MOEP-co-DMA) polymer, polymerized by free radical polymerization. n h 3 Cl n h 3ci Figure 2.2 Synthesis schematic of Poly(acrylamide-co-aminopropyl methacrylamide) polymer, polymerized by free radical polymerization. 2.3.4 Coacervate Formation PEG-dA was dissolved in degassed DI water at the desired concentration (0-25 wt%). Poly(acrylamide-co-aminopropyl methacrylamide) and poly(MOEP-co-DMA) were dissolved in separate PEG-dA solutions at a final concentration of 5 wt%. The poly(MOEP-co-DMA) PEG-dA solution also contained a 0.2 molar ratio of Ca2+ to phosphate side chains. The copolymer solutions were adjusted to pH 7.4±0.2 with NaOH. The poly(acrylamide-co-aminopropyl methacrylamide) PEG-dA solution was added dropwise while stirring to the poly(MOEP-co-DMA) PEG-dA solution to a molar ratio of 0.6 amine side chains to phosphate side chains. Within a few minutes a turbid coacervate settled out of solution. 2.3.5 Mechanical Testing The adhesive PEG-dA filled coacervates were cross-linked through the o-DHP side chains of the polyphosphate and/or by polymerizing PEG-dA. o-DHP was oxidatively cross-linked by adding 1 equivalent of NaIO4. To slow the oxidation of o- DHP side chains, in order to allow better control of the setting reaction, a sugar (1,2-O-Isopropylidene- D-glucofuranose, 98%) molecule was used to prepare an aqueous NaIO4/sugar complex solution (100 mg/mL) with a NaIO4:Sugar of 1:1.2 dissolved in water. PEG-dA was polymerized with 3.5 mol% ammonium persulfate (APS) and 5.2 mol% N,N,N',N'-tetramethylethylenediamine (TEMED). Immediately after adding NaIO4, APS, and TEMED, 20 ^l of coacervate was added to a wet 0.5 x 5 cm cleaned and polished Al adherend. A second wet Al was placed on the first with a 14-20 mm overlap, secured with a stainless steel clip, and submerged in water for 20-24 hours at 22-24°C. For each test condition 4-6 specimens were prepared. The shear strength of 42 the bonds were determined on a material testing system (Instron) with a 500 N load cell, crosshead speed 0.2 mm min-1, while fully submerged in a temperature-controlled water bath. 2.3.6 Dynamic Rheology Flow experiments were done on a stress-controlled rheometer (TA Instrument, AR 2000ex) using a 20 mm, 4° cone and plate, gap of 114 ^m, and at 25°C with 150 |iL coacervate samples. All rheology experiments were repeated with three independently prepared coacervate samples. 2.3.7 ATR-FTIR Attenuated Total Reflectance-Fourier Transform Infrared Microscopy (ATR-FTIR, FTS 6000 Spectrometer BioRAD) was used to measure the amount of PEG-dA in coacervates as well as the conversion of polymerization. The scans were made on ZnSe reflective crystal by placing 50 ^l of coacervate and running 30 scans per spectrum. Standard PEG-dA solutions of known concentration in water were scanned to get a standard curve (Figure 2.3A). Peak 1415 cm-1, corresponding to the acrylate group in PEG-dA, was used to fit a linear model of normalized peak area versus known concentration (Figure 2.3B). This standard curve fit was used to analyze the amount of PEG-DA in coacervates. Each sample was measured three times to get an average value. 2.4 Results and Discussion Coacervates were formed with 2-(methacryloyloxy) ethyl phosphate dopamine methacrylamide (poly(MOEP-co-DMA)), poly(acrylamide-co-aminopropyl methacrylamide) and Ca2+, as described in detail previously [4], in solutions containing 43 44 A) 0)O£ (0 n o </> n < o>> ' j _ro Q) ■ 30 wt% ■ 25 wt% 20 wt% 15 wt% 10 wt% 5 wt% Water 1500 1400 1300 1200 1100 1000 B) Wavenumber (crrT^) PEG-dA (wt%) Figure 2.3 ATR-FTIR (A) ATR-FTIR of PEG-dA in water solutions of set concentrations were scanned to generate a standard curve. (B) Peak 1415 cm-1 was used to fit a linear model of normalized peak area vs. known concentration. nominal wt% concentrations of 0, 5, 10, 15, 20, and 25 PEG-dA (MW 700 g/mol). Dense complex coacervates phase separated from the solutions. The concentration of PEG-dA entrapped in the coacervate phase was determined by Attenuated Total Reflectance- Fourier Transform Infrared Microscopy (ATR-FTIR). The absorbance peak at 1415 cm-1 was compared to the 1415 cm-1 acrylate groups of standard solutions of PEG-dA in water (Figure 2.4A) [9, 10]. On average the PEG-dA concentration in the coacervate was 73% of the initial concentration in solution. Above 25 wt% PEG-dA the coacervates were too viscous to work with conveniently. Free radical polymerization of entrapped PEG-dA was initiated by adding ammonium persulfate (APS) and tetramethylethylenediamine (TEMED) to the complex coacervate. The extent of PEG-dA polymerization within the coacervate was determined from the 1415 cm-1 ATR-FTIR peak corresponding to the acrylate functional group (Figure 2.4B). The complex coacervate containing 11.4 wt% PEG-dA reached ~40% conversion after 24 hr. The shear bond strengths of the PEG-dA filled adhesive coacervates were determined in lap shear tests with polished aluminum adherends. Cross-linking of the coacervate network through the o-dihydroxyphenyl sidechains of poly(MOEP-co-DMA) and the amine side chains of poly(acrylamide-co-aminopropyl methacrylamide) was initiated by addition of NaIO4. Free radical polymerization of PEG-dA was initiated simultaneously by addition of APS and TEMED. Immediately after initiation the coacervates were applied to wet Al adherends. The bonds were cured for 24 hours and fully submerged in water (22°C) before loading to failure on a material testing system. The maximum bond strengths increased with increasing PEG-dA (Figure 2.5), nearly doubling from a mean of 512 +/- 208 kPa without PEG-dA to a mean of 973 +/- 263 kPa 45 46 A) Wavenumber (cm^) B) Time (min) Figure 2.4 ATR-FTIR Study of 11.4 wt% PEG-dA complex coacervate. (A) Spectrum of acrylate group at 1415 cm-1 (j indicating peak of decreasing over time after polymerization. (B) Time course of PEG-dA conversion over 24 h. Error bars: ± s.d. (n = 3). 47 PEG-dA (wt%) Figure 2.5 Shear bond strength of PEG-dA filled complex coacervates. The coacervate network was oxidatively cross-linked and PEG-dA polymerized by the simultaneous addition of NaIO4 and APS/TEMED, respectively. Bonds were cured under water for 24 h at 25°C. Gray column: coacervate filled with 15 wt% nonacrylated PEG and cured with NaIO4 and APS/TEMED. Error bars: ± s.d. (n = 5). with 17.7 wt% PEG-dA. Maximum loads were ~1.2 MPa, more than four times higher than estimated bond strengths of the natural sandcastle glue [7] and mussel adhesive plaque byssal thread assemblies [11]. The shear modulus, approximately 25 MPa, was not statistically different between any of the PEG-dA concentrations. To confirm the increased bond strength was due to polymerization of PEG-dA, we formed coacervates with 15 wt% nonacrylated PEG (MW 400 g/mol) and treated with NaIO4 and APS/TEMED. The underwater bond strengths were 211 +/- 39 kPa, less than 25% of the PEG/coacervate bond strengths. The flow behavior is of critical importance for adhesives based on complex fluids such as coacervates. For medical applications, the viscosity should be sufficiently high at low shear rates that the adhesive does not flow away from the application site, yet low enough at high shear rates that it can be conveniently applied through a narrow gauge cannula, or catheter, without high pressure. At the same time, it is imperative to recognize shear-induced, irreversible structural transitions at high shear rates that may compromise cohesive bond formation. The viscosities of the adhesive complex coacervates containing 0 and 11.4 wt% PEG-dA were investigated as a function of shear rate using a cone and plate geometry (Figure 2.6). At low shear rates (0.01 s-1) the viscosity of the 11.4 wt% PEG-dA filled coacervate (13.8 +/- 6.2 Pa s) was substantially higher than the complex coacervate with 0 wt% PEG-dA (0.9+/-0.1 Pa s). With increasing shear rate the PEG-dA-filled coacervates first thickened, then steadily thinned until a sudden 16-fold drop in viscosity occurred at a shear rate of 6 s-1 (Figure 2.6A). The shear thinning behavior was reversible; viscosity recovered with little hysteresis as the shear rate was decreased. The complex coacervate without PEG-dA shear thinned as 48 49 A) Shear Rate(1/s) B) Shear Rate(1/s) Figure 2.6 Flow curves of complex coacervates under steady shear. (A) Filled coacervates of 11.4 wt % PEG-dA: low to high shear rate (closed square) and high to low shear rate (open square). (B) Unfilled coacervate: low to high shear rate (closed circle), and reverse (open circle). Symbols and error bars are the average viscosity and s.d. at each shear rate of three independent coacervate samples. well but did not display a similar sudden sharp drop in viscosity (Figure 2.6B). The flow curves of the complex coacervate containing PEG-dA monomers are similar to other coacervate systems that have been investigated rheologically. Whey protein and gum arabic coacervates displayed a similarly abrupt shear-thinning transition that was accompanied by an increase in turbidity [12]. The transition was reversible. Polycation/mixed micelle coacervates, above a certain temperature, underwent a dramatic shear rate-dependent drop in viscosity before visibly phase separating [13]. In both cases, the visible changes suggest the abrupt shear-thinning events were due to microscale structural reorganizations of the interspersed phases of the complex coacervates. The quiescent nanoscale molecular structure of the complex coacervates were conceptualized as beads on compacted strings: globular whey proteins on gum arabic molecules in the first case, polyanionic mixed micelle beads on polyamine strings in the second case [14]. Abrupt shear thinning was attributed to shear-induced elongation of the beaded string structures, resulting in increased lateral intercomplex associations and coalescence of nanocomplexes into dense microphases. The rich and complex flow behavior of the PEG-dA-filled coacervates suggests that similar shear-induced, reversible structural reorganization may occur within the PEG-dA filled coacervates. Unsheared complex coacervates with and without PEG-dA were frozen, lyophilized, and examined by scanning electron microscopy (SEM). The coacervates had a porous three-dimensional network of tubular structures (Figure 2.7) reminiscent of the sponge-like network of tubules observed in other complex coacervates by cryo-TEM [15]. There were no structural differences apparent between coacervates with and without PEG-dA monomers. Based on the flow behavior and SEM 50 51 Figure 2.7 SEM Images of fractured surfaces of lyophilized complex coacervates. (A) 700x, (B) 5000x. micrographs, a conceptual diagram of the PEG-dA-filled complex coacervate before and after the shear-induced structural transition is shown in Figure 2.8. The quiescent coacervate is pictured as a dense, interconnected, colloid-rich network interspersed within a watery, colloid-depleted network containing PEG-dA (Figure 2.8A). Above a critical shear rate, the interspersed networks may undergo shear-banding (Figure 2.8B), a phenomenon in which the components of a complex fluid phase separate into distinct bands under shear [16-18]. The practical significance of the shear thinning of the PEG-dA-filled coacervates is demonstrated in Figure 2.9, a still image from a supplemental video. A 11.4 wt% PEG-dA-filled coacervate was loaded into a 1 mL syringe fitted with a 27 gauge cannula. Despite the relatively high initial viscosity, it took little manual effort to eject a fine cohesive thread of the PEG-dA-filled coacervate under water. The shear rate during ejection was estimated to be 750 s-1. The water-borne threads were denser than water, maintained their shape, and adhered where they contacted the glass surface. The coacervate also adhered when applied underwater to vertical glass surfaces. In principle, the coacervated threads, or any pattern of threads, can be cross-linked in place by coinjection of polymerization initiators. The ability to accurately deliver adhesive through a fine cannula or catheter will allow precise and less invasive repair of bone fractures [19] and other tissues. 2.5 Conclusion In summary, the bond strength of the biomimetic adhesive coacervates was substantially improved to well above the estimated bond strength of natural bioadhesives by incorporating a second polymer network into the coacervate network. The viscous 52 53 • • • ••• • • • « • • • • high shear rate Figure 2.8 Conceptual diagram of the structure of the PEG-dA filled complex coacervate. (A) In the quiescent state, the electrostatically associated nanocomplexes form a fluid, interconnected, three-dimensional network. An aqueous phase containing PEG-dA is interspersed within the pores of the connected network of nanocomplexes. (B) At a critical shear rate, the nanocomplexes may be elongated leading to additional lateral interactions and a second, reversible macrophase separation. 54 Figure 2.9 Image of a 11.4 wt % PEG-dA filled coacervate loaded into a 1 mL syringe with a 27 gauge cannula, being ejected under water. 55 PEG-dA filled coacervate could be easily ejected through a fine gauge cannula as a result of reversible shear thinning. The threads maintained their form underwater and adhered to wet glass surfaces. The successful incorporation of high concentrations of water-soluble monomers demonstrated that, in principle, almost any water-soluble molecule can be contained in a complex coacervate and precisely delivered in a wet environment, including noninvasive delivery into the body. Such properties merit further evaluation of the filled adhesive coacervates as injectable drug delivery depots in addition to their potential as wet field medical adhesives. Work in progress is focused on incorporating more and different types of nano- and microphases into complex coacervates to further improve underwater bond strengths. 2.6 References 1. R. A. Jensen and D. E. Morse, "The Bioadhesive of Phragmatopoma Californica Tubes: A Silk-like Cement Containing L-DOPA," J. Comp. Physiol. B, 158 [3] 317-324 (1988). 2. R. J. Stewart, J. C. Weaver, D. E. Morse, and J. H. Waite, "The Tube Cement of Phragmatopoma Californica: a Solid Foam," J. Exp. Biol., 207 [26] 4727-4734 (2004). 3. H. Zhao, C. Sun, R. J. Stewart, and J. H. Waite, "Cement Proteins of the Tube-building Polychaete Phragmatopoma Californica," J. Biol. Chem., 280 [52] 42938-42944 (2005). 4. H. Shao, K.N Bachus, and R. J. Stewart, "A Water-borne Adhesive Modeled After the Sandcastle Glue of P. californica," Macromol. Biosci., 9 [5] 464-471 (2009). 5. H. Shao, and R. J. Stewart, "Biomimetic Underwater Adhesives With Environmentally Triggered Setting Mechanisms," Adv. Mater., 22 [6] 729-733 (2010). 6. M. J. Stevens, R. E. Steren, V. Hlady, and R. J. Stewart, "Multiscale Structure of the Underwater Adhesive of Phragmatopoma Californica: A Nanostructured Latex With a Steep Microporosity Gradient," Langmuir, 23 [9] 5045-5049 56 (2007). 7. C. Sun, G. E. Fantner, J. Adams, P. K. Hansma, and J. H. Waite, "The Role of Calcium and Magnesium in the Concrete Tubes of the Sandcastle Worm," J. Exp. Biol., 210 [8]1481-488 (2007). 8. A. J. Kinloch, J. H. Lee, and A. C. Taylor, "Toughening Structural Adhesives via Nano- and Micro-phase Inclusions," J. of Adhesion, 79 [8-9] 867-873 (2003). 9. R. P. Witte, A. J. Blake, C. Palmer, and W. J. Kao, "Analysis of Poly(ethylene glycol)-diacrylate Macromer Polymerization Within a Multicomponent Semi-interpenetrating Polymer Network System," J. Biomed. Mater. Res., 71 [3] 508-518 (2004). 10. G. Kang, Y. Cao, H. Zhao, and Q. Yuan, "Preparation and Characterization of Crosslinked Poly(ethylene glycol) Diacrylate Membranes With Excellent Antifouling and Solvent-resistant Properties," J. Membr. Sci., 318 [1-2] 227-232 (2008). 11. J. R. Burkett, J. L. Woitas, J. L. Cloud, and J. J. Wilker, "A Method for Measuring the Adhesion Strength of Marine Mussels," J. Adhes., 85 [9] 601-615 (2009). 12. F. Weinbreck, R. H. Wientjes, H. Nieuwenhuijse, G. W. Robinjn, and C. G. de Kruif, "Rheological Properties of Whey Protein/Gum Arabic Coacervates," J. Rheol., 48, 1215 (2004). 13. P. L. Dubin, Y. Li, and W. Jaeger, " Mesophase Separation in Polyelectrolyte-mixed Micelle Coacervates," Langmuir, 24 [9] 4544-4549 [2008]. 14. M. W. Liberatore, N. B. Wyatt, M. Henry, P. L. Dubin, and E. Foun, "Shear-induced Phase Separation in Polyelectrolyte/Mixed Micelle Coacervates," Langmuir, 25 [23] 13376-13383 (2009). 15. D. S. Hwang, H. Zeng, A. Srivastava, D. V. Krogstad, M. Tirrell , J. N. Israelachvili JN, and J. H. Waite, "Viscosity and Interfacial Properties in a Mussel-inspired Adhesive Coacervate," Soft Matter, 6, 3232-3236 (2010). 16. R. Makhloufi, J. P. Decruppe, A. Ait-Ali, and R. Creesley, "Rheo-optical Study of Worm-like Micelles Undergoing a Shear Banding Flow," Europhys Lett., 32 [3] 253-258 (1995). 17. P. Butler, "Shear Induced Structures and Transformations in Complex Fluids," Current Opinion in Colloid and Interface Sci., 4 [3] 214-221 (1999). 57 18. H. A. Barnes, J. F. Hutton, and K. Walters, An Introduction to Rheology. Elsevier, Amsterdam, 1989. 19. B. D. Winslow, H. Shao, R. J. Stewart, and P. A. Tresco, "Biocompatibility of Adhesive Complex Coacervates Modeled After the Sandcastle Glue of Phragmatopoma Californica for Craniofacial Reconstruction," Biomat., 31 [36] 9373-9381 (2010). CHAPTER 3 FETAL MEMBRANE PATCH AND BIOMIMETIC ADHESIVE COACERVATES AS A SEALANT FOR FETOSCOPIC DEFECTS 3.1 Abstract Iatrogenic preterm premature rupture of membranes after fetoscopic procedures affects 10-47% of patients, secondary to the nonhealing nature of membranes and the separation of layers during the entry. In this study we developed an in vitro model to mimic the uterine wall-fetal membrane interface using a water column with one end and sealed with human fetal membranes and poultry breast, and a defect was created with an 11 French trocar. In addition, a fetal membrane patch in conjunction with multiphase adhesive coacervates modeled after the sandcastle worm bioadhesive was tested for sealing of an iatrogenic defect. The sealant withstood an additional traction of 12 g for 30-60 minutes and turbulence of the water column without leakage of fluid or slippage. The adhesive is nontoxic when in direct contact with human fetal membranes in an organ culture setting. A fetal membrane patch with multiphase adhesive complex This chapter is reorganized from a pre-peer-reviewed version of the following article: Reprinted from Acta Biomateria., 8(6), L. K. Mann, R. Papanna, K. J. Moise, R. H. Byrd, E. J. Popek, S. Kaur, S. C. G. Tseng, and R. J. Stewart, Fetal Membrane Patch and Biomimetic Adhesive Coacervates as a Sealant for Fetoscopic Defects, 2160-2165, Copyright 2012, with permission from Elsevier. 59 coacervates may help to seal the defect and prevent iatrogenic preterm premature rupture of the membranes. 3.2 Introduction Iatrogenic preterm premature rupture of the membranes (iPPROM) after a fetal intervention procedure is a major complication that affects 10-47% of procedures [1-5]. iPPROM leads to an increased risk of preterm labor and worsens the perinatal mortality, undermining the true benefit of such interventions [6]. There are two possible explanations for the increased risk for iPPROM after invasive fetal procedures. One is the innate nonhealing nature of the fetal membranes, as demonstrated in both in vivo and in vitro studies [7,8]. The other is that separation of the amnion from the chorio-decidual layers that occurs during the introduction of instrumentation into the uterine cavity can cause a persistent parting of membranes with subsequent leakage of amniotic fluid [9]. There have been several attempts to study sealants at the site of the fetal membrane defect, both in vitro and in vivo [10-12]. However, there is no ideal in vitro model to simulate the relationship of the uterine wall, the fetal membranes, and the amniotic fluid environment. There is evidence to suggest that a decelluarized fetal membrane scaffold can promote cellular proliferation at the defect site [13]; however, no method to introduce a fetal membrane patch through a narrow operative cannula and deliver it to the site of the defect has ever been described. Additionally, after the patch has been deployed, the challenge of fixation to the membranes and the uterine wall remains due to the dynamic nature of the amniotic fluid and uterine musculature. An underwater adhesive that would fix a tissue scaffold to the edges of the defect in place for the remainder of the pregnancy would be an ideal solution to the problem iPPROM; however, no adhesive suitable for this task is available. Development of medical adhesives for the wet interior of the body is both chemically and biologically challenging. The adhesive must be delivered, bonded, and cured in the presence of moisture, must be nontoxic, and must not provoke a severe foreign body response. One approach to achieve underwater bonding is to study natural biological underwater adhesives, identify their key chemical features and copy that chemistry using nontoxic, biocompatible, and cost-effective synthetic polymers. Numerous aquatic organisms produce working underwater adhesives as part of their aquatic lifestyle to either position themselves in a suitable environment or to create a protective structure. The sandcastle worm, an intertidal marine polychaete (Phragmatopoma californica), produces a proteinaceous glue with which it joins together sand grains into a protective shell while fully submerged in seawater [14]. The proteins of the natural sandcastle glue are highly charged with opposite charges segregated into different proteins [15]. The polyacidic and polybasic nature of the glue proteins suggested complex coacervates-concentrated, phase-separated, associative polymer fluids-may be intermediates in natural bonding. Copying the side chain chemistry and molar ratios with synthetic poly(meth)acrylate copolymers resulted in adhesive complex coacervates that qualitatively replicated many of the features of the natural underwater adhesive [16]. Biodegradable versions [17] of the synthetic adhesive did not interfere with wound healing in a rat calvarial defect model [18]. Bond strength and other material properties were improved by introducing additional polymer networks into the adhesive coacervates [19]. In this study, we aimed to create an in vitro model to simulate the anatomical 60 relationship of the fetal membranes, uterine wall, and surrounding amniotic fluid. Using such a model, we introduced an iatrogenic defect in a similar fashion to that used in clinical fetal interventions. Furthermore, we tested a technique to introduce a fetal membrane patch through a cannula to the site of a defect and test its sealing capacity and evaluated the use of multiphase adhesive coacervates to adhere the fetal membrane patch to the defect. In addition, we examined the potential tissue cytotoxicity of the adhesive coacervates in an in vitro culture system. 3.3 Materials and Methods 3.3.1 Creating an In Vitro Uterine Model The Institutional Review Board of Baylor College of Medicine, Houston, TX (#H-26110), approved the collection of human fetal membranes for the study. We created an in vitro uterine model using a filleted poultry breast and human fetal membranes. A 100 ml polypro cylinder (VWR International, West Chester PA) was cut at the base and the cut end was lipped using heat. The cylinder was then mounted on a stand. Fresh human fetal membranes were obtained from term vaginal deliveries and were transferred to the laboratory in a balanced salt solution (BSS). The fetal membranes were cut into 6-cm diameter patches and secured to the lipped end of the cylinder with the amnion facing towards the inside of the cylinder. A poultry breast was filleted to 1­cm thickness and pounded gently using a hammer to simulate the uterine wall musculature. A 6-cm diameter patch of poultry breast fillet was then wrapped over the fetal membranes on the cylinder and secured in place with a suture material. The column was filled using BSS. 61 3.3.2 Creating an Iatrogenic Defect A defect in the fetal membrane through the poultry breast and the fetal membranes was created using an 18-gauge needle, followed by a guide wire (Cook® Urological Inc; Bloomington, IN, USA). Subsequently, an 11 French Teflon cannula (Cook® Medical Inc, Bloomington, IN, USA) was introduced over the guide wire using Seldinger's technique [20]. Then, the trocar was removed to leave the cannula in place. This entry method is identical to that used in most fetal intervention centers for fetal surgical procedures. 3.3.3 Technique to Introduce the Fetal Membrane Patch Fetal membrane patches were supplied by Bio-Tissue, Inc. (Miami, FL) and processed in the same manner as described for human amniotic membrane currently used for ocular surface reconstruction [21]. Briefly, fetal membrane patches were placed on a nitrocellulose paper with the amniotic membrane facing up (for ease of handling). After being cut in a circular fashion to the desired size, they were lyophilized to reduce their thickness to facilitate their insertion into the cannula. Upon insertion, one edge of the membrane was removed from the paper and folded in half (Figure 3.1A). The center of the patch was lifted from the paper, and a 4-O Monocryl suture with a tapering needle (Ethicon Inc, San Angelo, TX) was passed through the center of the patch and a noose was tied (Figure 3.1B). The remainder of the patch was removed from the paper (Figure 3.1C). The needle was removed from the suture and the distal end was passed through a 9-French Teflon cannula while the self-check valve on the proximal end was removed using a knife. With gentle traction on the suture, the patch was retracted into the distal tip of the cannula (Figure 3.1D). The original trocar that was an integral part of the 9F 62 63 Figure 3.1 In vitro model for uterine wall and a fetal membrane patch for the defect. (A) The lyophilized amnion-chorion is lifted off of the nitrocellulose paper; (B) a 4-0 Monocryl suture is passed through the center of the patch; (C) a noose is tied to the fetal membrane patch to form a firm knot; (D) the free end of the suture is passed through a 9- French cannula, to align the knot inside the cannula (insert); (E) an in vivo uterine model with an 11-French cannula in place; (F) the 9-French cannula carrying the fetal membrane patch is inserted through the 11-French cannula, and the patch is introduced into the fluid using a blunt plunger; (G) both cannulas are withdrawn and the patch is aligned to the effect, followed by glue is applied around the patch; (H) the patch and the glue are in place. cannula was modified to serve as a blunt introducer. This blunt introducer was advanced from the proximal end of the cannula to abut the patch. Once the 11-French cannula had been introduced through the base of the in vitro model, the 9-French cannula containing the membrane patch was introduced through it. The column was filled to a height of 10 cm with BSS (Figure 3.1E). The patch was introduced into the fluid column advancing the blunt introducer. Once free within the fluid medium, the patch was allowed to swell for 2 minutes - a timescale that had been established for maximum swelling based on prior experiments (Figure 3.1F). Both cannulas were then withdrawn while keeping the suture and the patch in place. The suture was then withdrawn gently to position the patch in the defect so that the amnion faced the fluid medium mimicking the amniotic fluid while the chorion faced the poultry breast mimicking the uterine wall (Figure 3.1G). 3.3.4 Optimization of the Membrane Patch Size for Sealing Triplicates of lyophilized fetal membrane patches were created as mentioned above, with diameters ranging from 1 to 5 cm to determine the minimum size necessary to seal the iatrogenic defect in the above in vitro model. These were used to determine the sealing strength that could withstand the dislodgement of the plug. A 25 cm height of fluid in the column was chosen to mimic the average intrauterine amniotic fluid pressure of 18 mm Hg that we had observed in patients with excess amounts of amniotic fluid (data not shown). We additionally applied 12 g of traction to the plug and created turbulence in the fluid by shaking the column multiple times in all directions to mimic the in vivo fluid dynamics. 64 3.3.5 Adhesive Complex Coacervate Formation Polyethylene glycol diacrylate (PEG-dA, 760 Da, Aldrich) solutions were prepared in degassed, deionized water at the desired final concentration of 15 wt %. Poly(acrylamide-co-aminopropyl methacrylamide) (MW 288kDa, PDI 1.36) and poly (2- (methacryloyloxy)ethyl phosphate dopamine methacrylamide (MOEP-co-DMA, MW 64kDa, PDI 2.8) were then dissolved in separate PEG-dA solutions at final concentrations of 5 wt %. The poly(MOEP-co-DMA)-PEG-dA solution also contained a 0.2 M ratio of Ca2+ to phosphate side chains and 1 wt % nanosilica fillers (10 nm, Aerosil R 7200). The copolymer solutions were separately adjusted to pH 7.4 ± 0.2 with 6 M NaOH. The poly(acrylamide-co-aminopropyl methacrylamide)-PEG-dA solution was added dropwise while stirring to the poly(MOEP-co-DMA)-PEG-dA solution to a molar ratio of 0.6 amine side chains to phosphate side chains. Within a few minutes the complex coacervate settled out. The clear supernatant was removed. The adhesive PEG-dA and nanosilica-filled coacervates were cross-linked through the o-DHP side chains of the polyphosphate with the amine side chains of the polyamine and/or by polymerizing PEG-dA. o-DHP was oxidized to initiate cross­linking by the addition of 1.0 M equivalents of NaIO4 relative to the o-DHP sidechains. The rate of oxidative cross-linking was slowed by forming a reversible 1:1 complex between NaIO4 and 1,2-O-Isopropylidene-D-glucofuranose (IPGF), as described previoulsy [22]. PEG-dA was polymerized by adding 3.5 mol % ammonium persulfate (APS) and 5.2 mol % N,N,N',N'-tetramethylethylenediamine (TEMED) at the same time as NaIO4-IPGF. Bond strengths of nanosilica-filled adhesive coacervates were tested in vitro on 65 aluminum adherends, as described previously for multiphase conplex coacervates [19]. Briefly, NaIO4, APS, and TEMED were added to 20 ^l of coacervates, which was then applied to a wet 0.5 x 5 cm cleaned and polished Al adhered. A wet Al was placed on the first with a 14-20 mm overlap, secured with a stainless steel clip and cured by submergence in water for 20-24 h at 37 °C. Four to six specimens were prepared for each test condition. The load to failure of the bonds was determined on a material testing system (Instron) with a 500 N load cell at a crosshead speed of 0.2 mm min-1, while fully submerged in a temperature-controlled water bath at 37°C. In two separate experiments the loads were 524 ± 182 kPa (n = 5) and 698 ± 42 kPa (n = 3). Bonding of the nanosilica-filled coacervates to aminiotic membranes was evaluated in vitro with fresh tissue cut into 1 cm x 6 cm patches. PEG-filled coacervates were prepared with and without nanosilica fillers. The coacervates (20^1) were applied to a 1 cm2 area, then adhered to a second overlapping patch. The overlapped areas were pressed together under a 20 g weight for 60 min, then manually peeled apart from one end with forceps. The relative bond strengths of the coacervates were graded on a scale of 1-5, with 1 being the lowest bond strength and 5 being the highest. The nanosilica-filled adhesive coacervates formed substantially stronger bonds (4-5) with the amniotic membranes then the unfilled coacervates (1 -2). 3.3.6 Sealing of the Defect with Adhesive Coacervates After we identified the size of the patch that could seal the defect but slipped out at a water column of < 10 cm height, we used that size for the remaining tests in conjunction with the glue. Four sets experiments were conducted with a different source 66 of fetal membrane at each time. The application of the glue to the patch was timed. The sealing strength of the glue was examined by adding the BSS solution to the column to 25 cm height and was further challenged by traction added to the patch with increments of 3 g (maxiumum of 12 g). If the patch held a weight of 12 g, we observed the experiment for 60 minutes (Figure 3.1H). Subsequently, the weights were removed and fluid turbulence was created by tilting the column in multiple directions for 5 min to evaluate for slippage of the glued plug. 3.3.7 In Vitro Cytotoxicity Test The glue was evaluated for direct contact cytotoxicity on fresh term human fetal membranes obtained in a sterile fashion from three elective cesarean deliveries. These membranes were immediately transported to the laboratory in BSS with Pen-Strep (Gemini Bio-Products, West Sacramento, CA) sterilely cut into patches (2 x 2 cm) and placed in a six-well plate. In each test well, 200 |il of freshly prepared glue was applied over the amnion surface followed by the addition of 3 ml of Amniomax C-100 culture medium (Invitrogen Corporations, Carlsbad, CA). Membranes from control (n = 3) and test wells (n = 3) were harvested at 0, 24, and 48 h, fixed in 10% formalin, dehydration with 70% ethanol, embedded in paraffin, and sectioned for histological examination. Slides were dewaxed, rehydrated, and digested in 0.02% trypsin solution and subjected to hematoxylin-eosin staining (H&E). TUNEL (terminal deoxynucleotidyle transferase dUTP nick end labelling) staining was also performed using ApopTag Peroxidase In Situ Apoptosis Detection Kit (Millipore, Billerica, MA). On the H&E slides, the overall morphological condition of the membranes was examined. In the TUNEL staining slides, the amniotic epithelial cells were counted in 10 high power fields 67 or until reaching 500 cell count per slide. The cytotoxicity was calculated by determining the ratio of apoptotic cells to total number of amniotic epithelial cells. 3.3.8 Statistical Analyses Statistical analysis was performed using Wilcoxon's sum rank test for continuous variables and Fisher's exact test for categorical variables. A p-value of <0.05 was considered to be significant. 3.4 Results and Discussion 3.4.1 In Vitro Uterine Model With a Fetal Membrane Patch and Its Sealing Capacity Without Adhesive Coacervates The in-vitro uterine model and the iatrogenic defect were created successfully. We created four identical models, with a flow rate of 100 cc over 20 s. A patch size of < 2 cm did not seal the defect from the beginning, and the patches slipped out with a column height of 10 cm or above. A patch size of 3 cm started to leak fluid with a column height of 5-10 cm and failed completely with a water column of 25 cm. A patch size of 4 cm occluded the 11-French defect and was able to withstand a 25 cm column of fluid and 12 g of traction; when creating turbulence, two of the four patches were dislodged into the fluid column. A patch size of 5 cm did not fit into the 9-French cannula tip. Therefore, we chose a 3 cm lyophilized membrane for all subsequent adhesive coacervate experiments. 3.4.2 Sealing Test and Toxicity Testing With Adhesive Coacervates As stated above, all four 3 cm patches began to exhibit leakage at a fluid column height between 5 and 10 cm without glue. The patches were dislodged spontaneously 68 once the height of the fluid column reached 20 cm in two cases and dislodged immediately after applying 3 g of traction in the other two cases. In contrast, with additional glue, none of the cases demonstrated any leakage at a fluid height of 25 cm. Furthermore, no leakage was observed upon challenge with 12 g of traction for 30 minutes in one experiment and 60 minutes in the other three (see Figure 3.2). The latter three cases also held the membrane patch in place even after turbulence created for 5 minutes. They were then harvested and sectioned through the center to examine the junction between the membrane patch and the poultry breast wall. The glue was present in most of the junction between the poultry breast and the fetal membrane defect. In situ examination of the patch showed that the glue was spread 360° between the patch and the defect, including the fetal membrane edges and the muscular wall. Histological examination revealed that the glue-added experimental group did not show any signs of cytotoxicity at any of the three time points compared to controls. At time 0, the control exhibited 2.2% of TUNEL-positive apoptotic cells. At 24 hours, the experimental group demonstrated 2% apoptotic cells while the control showed 4.2% apoptotic cells (p = 0.3). At 48 hours, the experimental group had 0.2% apoptotic cells while the control had 1.6 % apoptotic cells (p = 0.4). 3.4.3 Discussion To investigate the potential efficacy of sealants for an iatrogenic defect created during a fetoscopic procedure, we need a model that simulates the fetal membrane and the uterine wall as well as the fluid dynamics of a pregnant uterus. Additionally, the model should be able to test a sealant's capacity to occlude the defect and bind the fetal membrane to underlying layers to prevent leakage. In this regard, previously reported 69 70 Figure 3.2 Fetal membrane patches sealing the defect with glue as an adhesive. The left image shows the experiment set-up for the in vitro uterine model. The illustration on the right shows the components of the model (dotted yellow line). (A) Fluid in a 100 cc column; (B) uterine wall simulation using fresh fetal membrane with the amnion facing the fluid column and the chicken breast on the outside; (C) traction of weight on the fetal membrane patch; -> fetal membrane patch in the defect with glue between the patch and the defect wall. models have not managed to reproduce these in vivo conditions. For example, Reddy et al. [23] used a 2.5 cm diameter 20 mL syringe with a human fetal membrane attached to the bottom lip. After creating a 20-gauge needle defect in the fetal membrane, various sealants were tested for their abilities to occlude the leakage without describing the height of the fluid column. Their model cannot address the issue of chorioamnion separation. Furthermore, their defect size was smaller than the 2-3 mm diameter that typically occurs after fetoscopic procedures. Suzuki et al. [11] also attached fetal membranes to the bottom of cylinder and applied a gradual pressure up to 100 mm H2O using a water column. The defects created ranged from a pinhead hole to 5 and 10 mm slits. Because photocrosslinkable chitosan was applied as a sealant before adding fluid to the column, their model does not test the efficacy of the sealant in a fluid-filled environment. Bilic et al. [12] used a mechanical stretch device, the Cellerator, to study sealing of a 3.4 mm defect on a wet membrane with a "mussel inspired" PEG-based hydrogel. The efficacy of the glue to seal the defect was tested by stretching the membrane using the Cellerator. Their model also did not test the hydrogel in a fluid filled environment, and it remains unclear whether the mechanical stretching resembles the force caused by hydrostatic pressure. All three models did not consider binding of the fetal membrane to the underlying uterine wall, which is a likely solution to prevent iPPROM. Our in vitro setting is a modified design of our previously published model [24], which was created to test the ability of a chicken ovomucin to seal a fetal membrane defect. The current model included a filleted chicken breast, simulating the uterine muscular layer, over the fetal membrane mimicking the natural anatomical relation. The aim of the model was to test the sealant's capacity to occlude and hold the membranes to 71 the muscular layer at the iatrogenic defect site. The model incorporates the biological and mechanical concerns leading to iPPROM. It is well known that the fetal membrane does not heal after an iatrogenic injury of greater than 2-3 mm in diameter, even up to 12 weeks after the injury [7]. The presence of a chorioamnion separation detected by ultrasound in nearly 25-30% of patients after a fetoscopic procedure increases the risk for iPPROM by 3- to 4-fold [9]. Because the absence of chorioamnion separation reduces the risk for iPPROM, we speculate that the binding of the two layers of the fetal membranes, the amnion to the chorion, is a critical step in preventing the leakage of fluid. This is why we included the filleted poultry breast and used its smoother surface facing the chorion layer of the fetal membrane to simulate the in vivo relationship between the uterine wall and the fetal membranes. An iatrogenic defect in the muscle and fetal membranes was created using Seldinger's technique, and the trocar was introduced in the same manner as it is introduced during clinical fetoscopic procedures. Our model therefore would be expected to produce similar stresses on tissue layers similar to those found in a clinical setting. Our method for introducing the patch through a narrower cannula is readily transferrable to a clinical application. The fetal membrane patch was designed in an "umbrella" shape, with an increasing thickness towards the amnioitc cavity. This helped to occlude the defect through a wedge effect by compressing the fetal membrane edges into the uterine wall to prevent chorioamnion separation. The pressure changes in a contractile uterus were simulated in our model by varying the height of the fluid column, while the traction challenge with fluid turbulence was added to simulate the complexity of uterine environment. The fetal membrane patch used as a scaffold in our study was introduced with the 72 chorion facing the defect site and the amnion facing the fluid environment. Previous studies have shown that an amniotic membrane scaffold fabricated into a plug promoted cellular proliferation at the site of a fetal membrane defect in vivo over a 7-day period [10-13]. Mallik et al. [13] used a surgical plug from a term decellularized fetal membrane for closure of fetal membrane defect in midgestation rabbits, resulting in integration of the scaffold into the fetal membrane and uterine wall in 71% of cases, as evidenced by cellular proliferation. A 4-mm diameter fetal membrane patch was found to seal the defect site without glue, but 50% of the patches were dislodged from the defect site with fluid turbulence. The lyophilization of the fresh human fetal membranes in our study reduced the bulkiness of the scaffold, which helped in the delivery through a 9- French cannula. After being introduced into the fluid environment, the lyophilized fetal membrane took approximately 2-3 minutes to regain its thickness and was secured in place with the suture, giving rise to an "umbrella" shape that enhanced its sealing capability. It remains to be determined whether the use of such a fetal membrane patch as described herein will promote better healing because the chorion facing the defect site might promote local scarring and the amnion facing the amniotic cavity might help reepithelialization. We also noted that the adhesive coacervates helped seal the defect when the fetal membrane was 3 mm in diameter even under weight traction and fluid turbulence. The glue was injected between the membrane patch and the defect in the poultry breast muscle layers directly with a short applicator. In future testing, the glue could be applied with an introducer placed through the main cannula via a percutaneous approach. The glue spread 360° around the patch and the muscular wall-a desirable effect for 73 74 preventing chorioamnion separation. A volume of 200-300 |iL of glue was sufficient to seal the defect even in a water-filled environment. Future studies are needed to determine if the glue itself, or its degradative product, might be released into the amniotic cavity to generate any ill effect - although it did not cause apoptosis in the amniotic epithelium. We chose to assess only the amniotic epithelial damage, as the amnion is considered the most important layer to maintain the integrity of the fetal membrane [10, 25]. 3.5 Conclusion Collectively, our in vitro model has demonstrated that a lyophilized fetal membrane patch effectively occluded a model of an iatrogenic fetal membrane defect in an aqueous environment. The patch was more effective when used in conjunction with a nanosilica-filled adhesive coacervate. Further studies in live animal models are needed to evaluate the efficacy and durability of this fetal membrane patch and the adhesive coacervates to assist in preventing iPPROM. 3.6 References 1. J. C. Jani, K. H. Nicolaides, E. Gratacos, C.M. Valencia, E. Done, J. M. Martinez, L. Gucciardo, R. Cruz, and J. A. Deprest, "Severe Diaphragmatic Hernia Treated by Fetal Endoscopic Tracheal Occlusion," Ultrasound Obstet. Gynecol., 34 [3] 304-310 (2009). 2. R. Papanna, S. Molina, K. Y. Moise, K. J. Moise, and A. Johnson, "Chorioamnion Plugging and the Risk of Preterm Premature Rupture of Membranes After Laser Surgery in Twin-twin Transfusion Syndrome," Ultrasound Obstet. Gynecol., 35 [3] 337-343 (2010). 3. R. Papanna, A. Johnson, R. T. Ivey, O. O. Olutoye, D. Cass, and K. J. Moise, "Laparoscopy-assisted Fetoscopy for Laser Surgery in Twin-twin Transfusion Syndrome With Anterior Placentation," Ultrasound Obstet. Gynecol., 35 [1] 65-70 (2010). 4. M. V. Senat, J. Deprest, M. Boulvain, A. Paupe, N. Winer, and Y. Ville, 75 "Endoscopic Laser Surgery Versus Serial Amnioreduction for Severe Twin-to-twin Transfusion Syndrome," N. Engl. J. Med. 351 [2] 136-144 (2004). 5. A. Roman, R. Papanna, A. Johnson, S. S. Hassan, J. Moldenhauer, S. Molina, and K. J. Jr. Moise, "Selective Reduction in Complicated Monochorionic Pregnancies: Radiofrequency Ablation vs. Bipolar Cord Coagulation," Ultrasound Obstet. Gynecol. 36 [1] 37-41 (2010). 6. R. Papanna, S. Molina, K. Y. Moise, K. J. Moise, and A. Johnson, "Chorioamnion Plugging and the Risk of Preterm Premature Rupture of Membranes after Laser Surgery in Twin-twin Transfusion Syndrome," Ultrasound Obstet. Gynecol., 35 [3] 337-343 (2010). 7. E. Gratacos, J. Sanin-Blair, L. Lewi, N. Toran, G. Verbist, L. Cabero, and J. Deprest, "A Histological Study of Fetoscopic Membrane Defects to Document Membrane Healing," Placenta. 27 [4-5] 452-456 (2006) 8. R. Devlieger, E. Gratacos, J. Wu, L. Verbist, R. Pijnenborg, and J. A. Deprest, "An Organ-culture for in Vitro Evaluation of Fetal Membrane Healing Capacity," Eur. J. Obstet. Gynecol. Reprod. Biol., 92 [1] 145-150 (2000). 9. R. Papanna, L. K. Mann, A. Johnson, H. Sangi-Haghpeykar, and K. J. Jr. Moise, "Chorioamnion Separation As a Risk for Preterm Premature Rupture of Membranes After Laser Therapy for Twin-twin Transfusion Syndrome," Obstet. Gynecol. 115 [4] 771-776 (2010). 10. Devlieger R., L. K. Milar, G. Bryant-Greenwood, L. Lewi, and J. A. Deprest, "Fetal Membrane Healing After Spontaneous and Iatrogenic Membrane Rupture: A Review of Current Evidence," Am. J. Obstet. Gynecol. 195 [6] 1512-1520 (2006). 11. K. Suzuki, M. Shinya, and M. Kitagawa, "Basic Study of Healing Injuries to the Myometrium and Amniotic Membrane Using Photocrosslinkable Chitosan," J. Obstet. Gynaecol. Res., 32 [2] 140-147 (2006). 12. G. Bilic, C. Brubarker, P. B. Messersmith, A. S. Mallik, T. M. Quinn, C. Haller, E. Done, L. Gucciardo, S. M. Zeisberger, R. Zimmermann, J. Deprest, and A. H. Zisch, "Injectable Candidate Selants for Fetal Membrane Repair: Bonding and Toxicity in Vitro," Am. J. Obstet. Gynecol., 202 [1] 85 (2010). 13. A. S. Mallik, M. A. Fichter, S. Rieder, G. Bilic, S. Stergioula, J. Henke, K. T. Schneider, J. Kurmanavicius, E. Biemer, R. Zimmermann, A. H. Zisch, and N. A. Papadopulos, " Fetoscopic Closure of Punctured Fetal Membranes With Acellular Human Amnion Plugs in a Rabbit Model," Obstet. Gynecol., 110 [5] 1121-1129 (2007). 76 14. R. A. Jensen, and D. E. Morse, "The Bioadhesive of Phragmatopoma Californica Tubes: A Silk-like Cement Containing L-DOPA," J. Comp. Physiol. B, 158 [3] 317-324 (1988). 15. H. Zhao, C. Sun, R. J. Stewart, and J. H. Waite, "Cement Proteins of the Tube-building Polychaete Phragmatopoma Californica," J. Biol. Chem., 280 [52] 42938-42944 (2005). 16. H. Shao, K.N Bachus, and R. J. Stewart, "A Water-borne Adhesive Modeled After the Sandcastle Glue of P. Californica," Macromol. Biosci., 9 [5] 464-471 (2009). 17. H. Shao, and R. J. Stewart, "Biomimetic Underwater Adhesives With Environmentally Triggered Setting Mechanisms," Adv. Mater., 22 [6] 729-733 (2010). 18. B. D. Winslow, H. Shao, R. J. Stewart, and P. A. Tresco, "Biocompatibility of Adhesive Complex Coacervates Modeled After the Sandcastle Glue of Phragmatopoma Californica for Craniofacial Reconstruction," Biomat., 31 [36] 9373-9381 (2010). 19. S. Kaur, G. M. Weerasekare, and R. J. Stewart, "Mutliphase Adhesive Coacervates Inspired by the Sandcastle Worm," ACS Appl. Mater. Interfaces., 3 [4] 941-944 (2011). 20. S. I. Seldinger, "Catheter Replacement of the Needle in Percutaneous Arteriography; A New Technique," Acta. Radiol. 39 [5] 368-376 (1953). 21. J. Liu, H. Sheha, Y. Fu, L. Liang, and S. C. Tseng, "Update on Amniotic Membrane Transplantation," Expert Rev. Ophthalmol., 5 [5] 645-661 (2010)