Submitted by M.Sc. Aitziber Iturmendi

Submitted at Institute of Chemistry

Supervisor and First Examiner WITH STIMULATED Assoc. Univ.-Prof. Dr. Ian Teasdale

Second Examiner Assoc. Univ.-Prof. Dr. Uwe DEGRADATION Monkowius

Month Year May 2018 PATHWAYS

Doctoral Thesis to obtain the academic degree of Doktorin der Naturwissenschaften in the Doctoral Program Naturwissenschaften

JOHANNES KEPLER UNIVERSITY LINZ Altenberger Str. 69 4040 Linz, Austria www.jku.at DVR 0093696

STATUTORY DECLARATION

I hereby declare that the thesis submitted is my own unaided work, that I have not used other than the sources indicated, and that all direct and indirect sources are acknowledged as references.

This printed thesis is identical with the electronic version submitted.

Linz, May 2018

iii

The Amazing thing about life is that You choose what you allow into it You choose how things affect you You choose how you react Happiness is a Choice Make it It´s up to you

v

ACKNOWLEDGMENTS

First of all, I want to thank Assoc. Univ.-Prof. Dr. Ian Teasdale. Thank you so much for giving me the opportunity to work in your team not only for 6 months, even for more than 5 years! Many thanks also for being my supervisor, my guide, my support (in and out of work) and why not part of my stress ;)

I am grateful to Univ.-Prof. Dr. Oliver Brüggemann for providing me with all equipments I needed, the access to the labs and his support.

I would like to thank also Assoc. Univ.-Prof. Dr. Uwe Monkowius for his help and guidance during my thesis.

Sandra Rothemund, it was a pleasure to work with you. None of this work would have been possible without you, thank you for everything!

I also want to thank all the colleagues at the ICP for your help and our moments together, and particullarly:

- Wolfgang Gnong, for all our long talks, work discussions, and moments full of laughter and craziness! - Helena Henke, Tamara Aigner and Anne Linhardt, because it was a pleasure to work with you all, share lab experiences and discuss about polyphosphazenes. Particularly, thanks Helena for your help and friendship since I came to Linz. - Renate Herbrik and Andreas Schnölzer. Both of you helped me in every moment I needed, at work as well as out of it. Thank you so much for everything!

Thanks also to Sabrina Theis, my project-partner, for your help and support especially with the photochemistry.

I would like to thank also Assoc. Univ.-Prof. Dr. Wolfgang Schöfberger for his help with the NMR measurements, no matter the time they could take.

I cannot forget Filipa Alves and Pascal Scholder. Thanks a lot for all our moments together (breaks, lunch times, beers...) and for being part of my family in Linz!

Last but not least, I want to thank my friends and family. Thanks for your support, patience, love… Definitely, THANK YOU VERY MUCH for everything!

vii

ABSTRACT

Degradable synthetic are of great importance for many applications such as medical applications, as well as environmental reasons. Polymer degradation, based on the cleavage of covalent bonds, is nearly inevitable but it is time limited. Consequently, degradable polymers should be considered as polymers that degrade in specific conditions and within the timescale of the given application. Hence stimulated degradation pathways, in which the degradation process is initiated by an external trigger such as enzymatic, photochemical, and oxidative environments, are attractive tools to control the degradation of polymers.

Poly(organo)phosphazenes are hybrid inorganic-organic polymers with a flexible backbone based on alternation of and atoms. The organic side groups, covalently bonded to the phosphorous atom, can protect the intrinsic hydrolytic backbone and tune the properties and degradation of the polymer.

In this work a variety of degradable poly(organo)phosphazenes are described with different properties to make them water soluble but also porous cross-linked scaffolds. Moreover, pH, oxidation, and photochemical triggering systems can be applied to induce and promote the degradation of the polymers. Particularly, -based poly(organo)phosphazenes are presented with convenient hydrolytic degradation rates, making them attractive for many biomedical applications. Among the studied poly(organo)phosphazenes it has been shown that the degradation was promoted when the amino acid was incorporated between the polymer backbone and the organic . The work described in this thesis focuses on the synthesis of novel poly(organo)phosphazenes with triggered degradation pathways, that is, on stable polymers that degrade upon a certain stimulus. In the first part of the thesis a pH- triggered system is presented, in which degradation rates are increased at lower pH values. Then as potential stimulus also a known reactive oxygen species (ROS), H2O2 has been employed. ROS, generated in the organism as a consequence of aerobic life, can lead to various diseases when it is overproduced. Therefore, H2O2 has been used as an oxidative trigger leading to polymer degradation. As a third stimulus visible has been applied which could be of particular interest also for many biological applications due to its mild and deeply penetrating wavelengths as well as spatial and temporal control. In the last part of the thesis degradable cross-linked scaffolds with

ix highly interconnected pores are presented which provide special attraction for cell growth in tissue engineering applications.

x

KURZFASSUNG

Abbaubare synthetische Polymere spielen auf Grund ihrer Eigenschaften nicht nur in der Umwelt, sondern auch in der Medizin eine bedeutende Rolle. Der Polymerabbau, der auf der Spaltung von kovalenten Bindungen basiert, ist zeitlich limitiert. Infolgedessen sollten abbaubare Polymere als Polymere betrachtet werden, die unter bestimmten Bedingungen und innerhalb des Anwendungszeitrahmens abbauen. Daher sind stimulierte Abbauprozesse von großer Wichtigkeit, wobei externe Triggers als hervorragende Werkzeuge eingesetzt werden können, die entweder enzymatisch, photochemisch oder oxidativ das Polymer kontrolliert abbauen.

Poly(organo)phosphazene sind anorganisch-organische Hybridpolymere mit einem flexiblen Rückgrat, das auf abwechselnden Stickstoff- und Phosphoratomen basiert. Die organischen Seitenketten, die kovalent an das Phosphoratom gebunden sind, können einerseits das hydrolytisch labile Rückgrat schützen, und anderseits die Eigenschaften und den Abbau des Polymers beeinflussen.

Diese Arbeit beschreibt eine Auswahl an abbaubaren Poly(organo)phosphazenen mit unterschiedlichen Eigenschaften, um sie wasserlöslich oder zu porösen vernetzen Gerüsten zu machen. Darüber hinaus können pH-, oxidations- und photochemisch- triggernde Systeme angewandt werden um den Abbau der Polymere zu induzieren und zu fördern. Insbesonders werden aminosäurebasierende Poly(organo)phosphazene mit geeigneten hydrolytischen Abbauraten präsentiert, die dadurch attraktiv für viele biomedizinische Anwendungen sind. Unter den untersuchten Poly(organo)phosphazenen konnte man erkennen, dass der Polymerabbbau durch den Einbau der Aminosäure Glycin zwischen Polymerrückgrat und organischem Substituenten begünstigt wird. Der Fokus dieser Arbeit liegt auf der Synthese von neuartigen Poly(organo)phosphazenen mit getriggerten Abbauprozessen, also auf stabilen Polymeren, die durch einen bestimmten Stimulus abbauen. Im ersten Teil dieser Forschungsarbeit wird ein pH-getriggertes System präsentiert, bei welchem die Abbaurate bei niedrigeren pH Werten gesteigert wird. Im zweiten Ansatz wurde eine bekannte reaktive Sauerstoffspezies (ROS), H2O2, als möglichen Stimulus angewandt. ROS, das im Organismus in Folge von aerobem Leben erzeugt wird, kann durch

Überproduktion zu verschiedenen Krankheiten führen. Daher wurde H2O2 als oxidativer Trigger zum Abbau des Polymers eingesetzt. Als dritter Stimulus wurde Licht im

xi sichtbaren Bereich verwendet. Aufgrund der tiefdringen Wellenlänge des Lichts, sowie die räumlich und zeitliche Kontrolle, kann dieser Stimulus besonders im biologischen Anwendungsbereiche interessant sein. Im letzten Teil dieser Arbeit werden abbaubare quervernetzte Gerüste mit stark miteinander verbundenen Poren präsentiert, welche besonders attraktiv für das Zellwachstum in Tissue Engineering ist.

xii

OVERVIEW

This thesis is divided in the following main chapters:

Chapter 1 outlines the importance of degradable polymers in diverse applications. Furthermore, it introduces the field of polyphosphazenes with a concise summary about common synthesis routes, highlights the particular interest of water-soluble polymers and explains the possible degradation mechanism.

Chapter 2 emphasizes the synthesis of novel poly(organo)phosphazenes with triggered degradation rates, that is, synthesis of polymers with good hydrolytic stability that undergo degradation only upon certain stimuli:

Chapter 2.1 presents the synthesis of water-soluble polyphosphazenes with tunable degradation rates. By incorporation of some amino acid groups between the polymer backbone and the water-solubilizing group the degradation is tailored, as well as increased at lower pH-values. Preliminary cell viability studies prove the biocompatibility and non-toxic nature of these polymers.

Chapter 2.2 shows the selective degradation of some polyphosphazenes upon oxidation. The polyphosphazenes with self-immolative side groups undergo

degradation upon H2O2 exposure while in absence of it are hydrolytically stable. Moreover, the polymers without the self-immolative group show no degradation sign upon oxidative environment.

Chapter 2.3 reveals the light-triggered degradation of polyphosphazenes. Incorporation of photocleavable groups lead to exclusive polymer degradation upon visible light irradiation. Furthermore, mix-substitution with different groups allows the synthesis of diverse polyphosphazenes with tunable properties and degradation rates.

Chapter 3 describes the synthesis of degradable cross-linked scaffolds with porous structure and pore size in the range of 100-200m, being attractive candidates for tissue engineering applications. By thiol-ene chemistry the properties and degradability of the polymer are easily tunable.

Chapter 4 summarizes the work and results presented in this thesis.

xiii LIST OF PUBLICATIONS

This dissertation is consisted of the following papers and manuscripts:

Chapter 1 A. Iturmendi and I. Teasdale. Water Soluble (Bio)degradable Poly(organo)phosphazenes. In Polyphosphazenes in Biomedicine, Engineering, and Pioneering Synthesis ACS Book 2018, accepted.

Chapter 2.1 S. Wilfert, A. Iturmendi, W. Schoefberger, K. Kryeziu, P. Heffeter, W. Berger, O. Brüggemann and I. Teasdale. Water-Soluble, Biocompatible Polyphosphazenes with Controllable and pH- Promoted Degradation Behavior. Journal of Polymer Science, Part A: Polymer Chemistry 2014, 52(2), 287-294. Reproduced with permission

Chapter 2.2 A. Iturmendi, U. Monkowius, I. Teasdale. Oxidation Responsive Polymers with a Triggered Degradation via Arylboronate Self- Immolative Motifs on a Backbone. ACS Macro Letters 2017, 6(2), 150-154. doi: 10.1021/acsmacrolett.7b00015. Reproduced with permission

Chapter 2.3 A. Iturmendi, S. Theis, D. Maderegger, U. Monkowius, I. Teasdale. Coumarin-Caged Polymers with a Visible-light Driven on-Demand Degradation, submitted.

Chapter 3 S. Rothemund, T. B. Aigner, A. Iturmendi, M. Rigau, B. Husár, F. Hildner, E. Oberbauer, M. Prambauer, G. Olawale, R. Forstner, R. Liska, K. R. Schröder, O. Brüggemann, I. Teasdale. Degradable Glycine-Based Photo-Polymerizable Polyphosphazenes for Use as Scaffolds for Tissue Regeneration. Macromolecular Bioscience 2015,15(3), 351-363. Reproduced with permission

xiv

ABBREVIATIONS

ASC adipose derived stem cells ATP adenosine triphosphate ATR-FTIR attenuated total reflectance fourier transform infrared ATRP atom transfer radical BMP morphogenetic protein Boc-Gly-OH N-(tert-butoxycarbonyl)glycine Boc-Val-OH N-(tert-butoxycarbonyl)-L-valine CD -dyclodextrin CMC critical micelle concentration CP-MAS cross polarization-magic angle spinning DBC double bond conversion DCC N,N′-dicyclohexylcarbodiimide DCM dichloromethane DCVC dry column vacuum chromatopraphy DLS dynamic light scattering DMAA N,N-dimethylacrylamide DMAP 4-(dimethylamino) DMF dimethyl formamide DMPA 2,2-dimethoxy- 2-phenylacetophenone D-NP dual interaction of polymeric nanoparticles DOX doxorubicin DPA N,N-diisopropylethylenediamine EO ethylene oxide ESI-MS electrospray ionization mass spectrometry FCS fetal bovine serum FITC fluorescein isocyanate FTIR Fourier-transform infrared spectroscopy GFLG glycine-phenylalanine-leucine-glycine (Gly-Phe-Leu-Gly) sequence GPC permeation chromatography HPLC high performance liquid chromatography HPMA N-(2-hydroxypropyl)methacrylamide Im-PPZ-cyclen poly(imidazole/1,4,7,10-tetraazyclodocane)phosphazene LCST lower critical solution temperature

xvii M-1000 amino-capped poly(ethylene oxide-co-propylene oxide) MEEP poly[bis(2-(2-methoxyethoxy)ethoxy)phosphazene] MES 2-(N-morpholino)ethanesulfonic acid NEt3 triethylamine NIPAm N-isopropylacrylamide NMR nuclear magnetic resonance PBS phosphate buffered saline PC poly(carbonate) PCEP poly[di(sodiumcarboxylatoethylphenoxy)phosphazene] PCL polycaprolactone PCPP poly[di(carboxylatophenoxy)phosphazene] pDMAEA-ppz poly[(2-dimethylaminoethylamino)phosphazene] PEG PET polyethylene terephthalate PHB polyhydroxybutyrate PIDP poly(imidazole/dimethylaminoethylamino)phosphazene PLA polylactic acid PLGA poly(lactic-co-glycolic) acid PmPEG2000 poly(ethylene glycol) methyl ether-based poly(organo)phosphazene PNDGPs poly(NIPAm-co-DMAA)/Glyet-PPPs PO propylene oxide PVA poly(vinyl alcohol) PVP polyvinylpyrrolidone PYRP poly(di[2-(2-oxo-1-pyrrolidinyl)ethoxy]phosphazene) ROP ring-opening polymerization ROS reactive oxygen species SEC size exclusion chromatography SEM scanning electron microscopy TFA trifluoroacetic acid THF tetrahydrofuran TIPS triisopropylnaphthalenesulfonic acid TRIS tris(hydroxymethyl)aminomethane trithiol trimethylolpropane tris(3-mercaptopropionate) UV-Vis -visible VE adipic acid divinyl

xvi

CONTENTS

Acknowledgments ...... vii

Abstract ...... ix

Kurzfassung ...... xi

Overview ...... xiii

List of publications ...... xiv

Abbreviations ...... xv

Contents ...... xvii

1. Introduction ...... 1

1.1. Degradable water soluble polymers ...... 3

1.2. Polyphosphazenes ...... 5

1.3. Water-soluble Poly(organo)phosphazenes...... 7

1.3.1. Anionic poly(organo)phosphazenes ...... 7

1.3.2. Cationic poly(organo)phosphazenes ...... 8

1.3.3. Non-charged water soluble poly(organo)phosphazenes ...... 9

1.3.4. Amphiphilic Poly(organo)phosphazenes ...... 12

1.3.5. Thermoresponsive polyphosphazenes ...... 13

1.3.6. Hydrogels ...... 15

1.4. Degradation mechanisms ...... 18

1.4.1. Stimuli responsive degradation mechanisms ...... 21

1.5. Summary and outlook ...... 23

1.6. References ...... 24

2. Stimuli responsive degradation of polyphosphazenes ...... 33

2.1. pH promoted degradation ...... 33

2.1.1. Introduction ...... 35

2.1.2. Experimental ...... 37

2.1.3. Results and discussion ...... 41

xvii 2.1.4. Conclusions ...... 49

2.1.5. References ...... 50

Supporting information ...... 53

2.2. Oxidative triggered degradation ...... 59

2.2.1. References ...... 71

Supporting information ...... 75

2.3. Photochemically triggered degradation ...... 93

2.3.1. Introduction ...... 95

2.3.2. Results and Discussion ...... 96

2.3.3. Conclusions ...... 102

2.3.4. References ...... 104

Supporting information ...... 107

3. Degradable cross-linked scaffolds ...... 123

3.1. Introduction ...... 125

3.2. Experimental ...... 127

3.3. Results and discussion ...... 134

3.4. Conclusion ...... 147

3.5. References ...... 149

4. Summary and Outlook ...... 153

Publications ...... 157

Conference contributions ...... 158

xviii

1. Introduction

This chapter includes a concise introduction to degradable water soluble polymers and their importance in diverse applications, in particular in the biomedical field, along with the suitability of poly(organo)phosphazenes in this field. Additionally, it summarizes the variety of water soluble poly(organo)phosphazenes such as ionic, amphiphilic, and thermoresponsive. Besides the recent developments on the synthesis of new water soluble polyphosphazenes, the degradability mechanisms of the polymers with particular interest on triggered degradation are also discussed.

My contribution to the paper

I wrote all drafts of this paper. Conceptualization and corrected versions were done together with I. Teasdale.

1

Water Soluble (Bio)degradable

Poly(organo)phosphazenes

Aitziber Iturmendi and Ian Teasdale*

–––––––––

Institute of Polymer Chemistry, Johannes Kepler University Linz, Altenberger

Strasse 69, 4040 Linz, Austria

E-mail: [email protected]

–––––––––

ACS Book “Polyphosphazenes in Biomedicine, Engineering, and Pioneering Synthesis”, accepted

2

Abstract

Water soluble and degradable poly(organo)phosphazenes are promising polymers for a wide range of applications, especially in the biomedical field. This chapter aims to summarize the variety of water soluble polyphosphazenes available ranging from ionic to amphiphilic, and thermoresponsive polymers to hydrogels. The most recent developments in the design and synthesis of water soluble poly(organo)phosphazenes are presented. Furthermore, we focus on the degradation pathways of the polymers with special attention paid to their stimulated degradation mechanisms.

1.1. Degradable water soluble polymers

Water soluble synthetic polymers are widely used in many applications including in the food and pharmaceutical industry, water treatment and electrical applications1. In addition to this, many bioerodible polymers are used, which by definition become water soluble under physiological conditions2. Due to their miscibility with water they are removed from the local site. However, while they “disappear” to the naked eye, being classified as bioerodible does not necessarily imply they are (chemically) degradable to small molecules. For environmental reasons3, 4 as well as medical applications5-7 the degradability of high molecular weight polymers is also an important issue which has to be considered carefully. Degradation is a chemical process based on the cleavage of covalent bonds. Among the possible degradation mechanisms, such as oxidative, photodegradative, or enzymatic8-10, is the most common chemical process in which water is responsible for the covalent bond cleavage. Although the terms are used interchangeably in the scientific literature, the term biodegradation is ideally reserved for processes in which degradation is caused by biological agents, such as enzymes, cells, or microorganisms. Furthermore, since it can be assumed that all polymers degrade somehow, it is only a matter of time, we follow herein the definition of degradable polymers suggested by Göpferich, as polymers that “degrade within the timescale of the application and specific conditions”11.

The environmental impact of non-degradable synthetic polymers has become one of global concern in recent years12, 13. Recently, a number of polymers such as polylactic acid (PLA), polyhydroxybutyrate (PHB), poly(lactic-co-glycolic) acid (PLGA) and polycaprolactone (PCL), which are considered to be, and marketed as,

3 biodegradable polymers, have been studied recently in fresh water and sea water 14. These polymers were stored for one year in vials with holes in their caps for oxygen exchange and immersed in 3 mL fresh water as well as artificial sea water. Thus, under fluorescent light and constant temperature (25°C) the results showed that only PLGA degraded totally in less than one year, whereas PCL, PLA and PET did not degrade at all. In the same timeframe PHB showed only ~8% degradation. Therefore, it is important to define the specific requirements of the materials in order to classify them as degradable or non-degradable polymers. In this sense, many common water-soluble polymers such as polyethylene glycol (PEG), poly(vinyl alcohol) (PVA), and poly[N-(2- hydroxypropyl)methacrylamide] (poly(HPMA)), do not have tendency of the backbone to breakdown so they are regarded as non-degradable polymers15. Semi-degradable polymers that degrade to smaller polymer fragments, can be prepared via copolymerization of persistent blocks with degradable units, as has been shown for example with HPMA copolymers16-18. Some polymers, such as, poly(L-glutamic acid), some polyphosphazenes and poly(phosphoesters) (Figure 1-1), are known to fully degrade to small molecules, an essential feature for some applications6, 19, 20.

Of the many medical applications, drug delivery systems represent one of the most widely researched applications for water-soluble polymers21, 22. Numerous types of polymer-drug conjugates, nanocarriers and polymer therapeutics are under development or entering clinics to improve the effectiveness and biodistribution of small-molecule drugs, as well as biomacromolecular drugs21, 22. Any polymeric carriers which enter the blood-stream for such therapeutic applications should fulfill not only the necessary water- and biocompatibility requirements, but also be proven to undergo complete clearance from the body in order to avoid deleterious long term side effects23-25. Since the essence of macromolecular carriers (long retention times, enhanced permeation and retention effect etc.) is in their high molecular weights, it follows that polymers which degrade in a timely manner to small molecules are essential.

Phosphorous-based polymers are an exceptional possibility for such applications due to the inherent tendency of phosphorus to undergo hydrolysis and hence when incorporated into macromolecules, main chain polymer degradation26. In combination with their facile functionalization this importantly allows the degradation rates to be tailored across a wide range spectrum to suit the specified application.

4

Figure 1-1. Some selected examples of degradable water-soluble synthetic polymers: a) poly(L- glutamic acid), b) poly(diglycerophosphazene) and c) poly(ethylene methylphosphonate).

1.2. Polyphosphazenes

Poly(organo)phosphazenes are polymers with inorganic backbone based on alternation of nitrogen and phosphorous atoms27. They are unique due to their flexible, high thermal stability and intrinsic hydrolytic backbone. Side groups, organic and covalently linked to the phosphorous atom by macromolecular substitution27, can tune the properties of the polymer (including e.g. hydrolysis rates, temperatures)28. Polyphosphazenes are used in a wide variety of applications that range from , lithium conductive polymers to biomedical materials29, 30.

Of the various routes to poly(organo)phosphazenes31, 32, the mostly used is via the inorganic macromolecular precursor poly(dichloro)phosphazene, [NPCl2]n. There are 27, 31 different routes available for the synthesis of [NPCl2]n , but the most used methods are ring-opening polymerization (ROP) of and living cationic polymerization. ROP is the most widely used route to obtain high molecular weight

[NPCl2]n but high purity of monomer, for reproducible results, and high temperatures (250°C) are required27 (Figure 1-2a). Lower polymerization temperatures (210°C) can be used but the addition of some catalysts such as BCl3·OP(OPh)3 or AlCl3 is required. Even if higher temperatures accelerate the , unfortunately the cross- linking effect is also accelerated so, special control over reaction temperature is essential for successful polymerization. Furthermore, with the ROP method chain branching may occur and polymers with high molecular weights and broad polydispersities are obtained.

The alternative method is via living cationic polymerization of trichloro(trimethylsilyl)phosphoranimine (Cl3PNSi(CH3)3) monomer at room temperature33. The most widely used initiator for the polymerization is phosphorus

5 pentachloride (PCl5) which after reacting with Cl3PNSi(CH3)3 leads to the formation of + 34 [Cl3PNPCl3] (Figure 1-2b) . Further equivalents of Cl3PNSi(CH3)3 can react with the cationic species in a living cationic polymerization until the monomer is consumed. Such procedure gives shorter polymer chains (n<75)35 with narrow molecular weight distribution but with living chain ends36.

More recently, polyphosphazenes with defined chain ends have been synthesized 37, 38 utilizing chlorinated tertiary phosphines (R3PCl2) as initiators . It is already known that tertiary phosphines, such as triphenylphosphine dichloride (Ph3PCl2), show ionized + - 39 form, [Ph3PCl] [Cl] , in polar solvents which can be used to initiate the polymerization + of Cl3PNSi(CH3)3. In a similar way as using PCl5, [Ph3PNPCl3] is formed upon addition of Cl3PNSi(CH3)3 which leads to monodirectional growth of the polymer enduring the controlled chain growth and narrow polydispersities (Figure 1-2c). Furthermore, through the use of commercially available monofunctionalized phosphine compounds the length of the polymer can be determined38, as well as modify the possible functional groups at one chain end. It opens the opportunity to achieve block as only + one chain end remains active. Thus, after total reaction of [Ph3PNPCl3] with

Cl3PNSi(CH3)3 different phosphoranimines, such as Cl(PhCH3)PNSi(CH3)3, can be added to achieve block copolymers40.

Figure 1-2. Some synthetic routes for the synthesis of poly(dichloro)phosphazene [NPCl2]n. a) Ring-opening polymerization of hexachlorophosphazene. Living-cationic polymerization of trichloro(trimethylsilyl)phosphoranimine (Cl3PNSi(CH3)3) with b)

(PCl5) and c) triphenylphosphine dichloride (Ph3PCl2) as initiators.

6

An useful property of poly(organo)phosphazenes is the facile synthesis of a wide variety of polymers29, 41-43. Once the backbone is assembled different organic nucleophiles can be introduced through macromolecular substitution of the atoms. Therefore, it makes it easier compared to many other methods of polymer synthesis where the side groups are incorporated into the starting monomer.

1.3. Water-soluble Poly(organo)phosphazenes

In the following sections a selection of some water-soluble poly(organo)phosphazenes will be presented to denote the potential of these polymers for biomedical uses.

1.3.1. Anionic poly(organo)phosphazenes

Acidic/anionic polyphosphazene are the most extensively studied water-soluble polyphosphazenes in nanomedicine. Poly[di(carboxylatophenoxy)phosphazene] (PCPP), first synthesized by Kwon and Allcock42 and later demonstrated by Andrianov and co-workers its outstanding ability to function as an immunoadjuvant43, 44, has been extensively investigated. While the free acid is not soluble in neutral or acidic water, it shows excellent solubility in its sodium salt form (Figure 1-3a)42. Once the backbone is assembled, the chlorine atoms from the [NPCl2]n precursor are replaced with the sodium salt of propyl p-hydroxybenzoate (sodium paraben) nucleophile. The use of protecting groups is required as the free acid groups lead to the breakdown of polymer backbone45.

Different polyphosphazene have been developed with some variations, but among them poly[di(sodiumcarboxylatoethylphenoxy)phosphazene] (PCEP) (Figure 1-3b) have already shown improved immunoadjuvant activity compared to PCPP46, 47. In this case, the polymer backbone is substituted with methyl 3-(4-hydroxyphenyl)propionate groups48. There is also an alternative route to afford anionic poly(organo)phosphazenes via aromatic sulfonation49-51. While at higher temperatures considerable polymer degradation occurs, poly[bis(phenoxyethoxy)phosphazene] could be sulfonated using concentrated H2SO4 at room temperature and treated with sodium carbonate solution to afford a water- soluble poly(organo)phosphazene (Figure 1-3c)50. It is also possible to prepare water-

7 soluble polythionylphosphazene polyelectrolytes32. After ring-opening polymerization of the cyclic thionylphosphazene, subsequent macromolecular substitution with an bearing a protected carboxylic acid and finally upon removal of the protective silyl groups under mild conditions (e.g. tetrabutylammonium fluoride), the anionic polymer is obtained. The triethylammonium counterion was exchanged to sodium and a readily water soluble was obtained.

Figure 1-3. Chemical structures of some anionic poly(organo)phosphazenes: a) Poly[di(sodiumcarboxylatophenoxy)phosphazene] (PCPP), b) Poly[di(sodiumcarboxylatoethylphenoxy)phosphazene] (PCEP), and c) Sulfonated poly[bis(phenoxyethoxy)phosphazene].

1.3.2. Cationic poly(organo)phosphazenes

Cationic polymers can also be prepared. Poly[(2- dimethylaminoethylamino)phosphazene] (pDMAEA-ppz) (Figure 1-4a) is one of the first cationic polyphosphazenes reported by Luten et al52 for its in vitro gene delivery characteristics. It is a polymer bearing a tertiary amine which is positively charged at physiological pH. Furthermore, through the addition of imidazole side groups the degradation rates could be tailored (Figure 1-4b)53. More recently, cationic polyphosphazenes have been reported through cosubstitution of imidazole groups with cyclic polyamine54. Among the polymers under study, Im-PPZ-cyclen (Figure 1-4c) showed promising results due to the tetraamine structure in which the positive charge is delocalized on the heterocyclic ring. Quaternization is also an alternative method to prepare cationic poly(organo)phosphazenes. In this context, some organic-inorganic hybrid brush polymers have been synthesized by grafting [2-(dimethylamino)ethyl

8

methacrylate] into the linear poly(organo)phosphazene via atom transfer (ATRP) and subsequent quaternization with different alkyl chain length iodides55. In a different study, allyl bromide was incorporated into the tertiary amine of pDMAEA-ppz obtaining the cationic polymer which could lead also to cross-linking reactions to obtain different hydrogels (See Hydrogels section)56. Recent synthesis of water soluble amino acid ester polyphosphazenes with free amino functional groups has been also investigated57. Boc-serine ethyl ester was reacted with the precursor poly(dichloro)phosphazene affording an amino-protected O- linked polymer. After deprotection in mild acidic conditions (using trifluoroacetic acid at room temperature) the free pendent hydrophilic amino groups contributed to the water solubility of the poly(organo)phosphazene.

Figure 1-4. Chemical structure of some cationic poly(organo)phosphazenes: a) Poly[(2-dimethylaminoethylamino)phosphazene] (pDMAEA-ppz), b) Poly(imidazole/dimethylaminoethylamino)phosphazene (PIDP), and c) poly(imidazole/1,4,7,10-tetraazyclodocane)phosphazene (Im-PPZ-cyclen).

1.3.3. Non-charged water soluble poly(organo)phosphazenes

Figure 1-5. Water-soluble and fully degradable branched hydroxyl-phosphazene.

9 Polyphosphazenes that bear hydroxyl units show significant features due to the functionality. Poly(diglycerophosphazene) (Figure 1-1b) has excellent water solubility and complete hydrolytic degradation in 720h at 37°C 58. However, investigations into this polymer have been rare, probably, due to the tedious protection and deprotection reactions required, in order to avoid cross-linking during polymer synthesis. In the last few years, the interest in hyperbranched or dendritic polyglycerols has increased significantly as their potential use in biomedical applications59-61 and very recently degradable, dendritic polyols on a branched poly(organo)phosphazene backbone have been reported. Applying a recently developed method to synthesize poly(organo)phosphazenes with star dendritic type structures62, fully degradable hydroxyl-functionalized polymers with exceptionally high functionality and controlled structures have been prepared (Figure 1-5)63. Furthermore, the use of photochemical thiol-yne addition of the glyceryl moieties circumvents the requirement for protection group chemistry. Indeed this type of chemistry has proven highly practical for the preparation of such highly functionalized, hydrophilic polyphosphazenes (see also glycopolymers below), not just due to the orthogonality towards a high number of functional groups, but also the high efficiency of the reaction and also the doubling of the number of functional groups per phosphorus. Importantly the sterically unhindered and reactive propargyl amine allows facile, complete substitution of the [NPCl2]n backbone, before adding bulkier functional groups (Figure 1-6). This is a key feature since the presence of residual Cl atoms is well-known to have a decisive effect on the hydrolytic stability of polyphosphazenes64.

Figure 1-6. Thiol-yne route to functionalized hydrophilic polyphosphazenes.

Polyvinylpyrrolidone (PVP) and PEG polymers are among the most widely used synthetic water-soluble polymers. However, as they are non-degradable their use in some medical applications, such as intravenous application, is restricted. Thus synthesis of new degradable polymers with the well-known pharmaceutical carriers is

10

of great interest. One route to overcome this limitation is the preparation of hybrid polyphosphazenes with a similar chemical functionality to PVP. Poly(di[2-(2-oxo-1- pyrrolidinyl)ethoxy]phosphazene) (PYRP) (Figure 1-7a) was synthesized by Andrianov and co-workers65 using 1-(2-hydroxyethyl)-2-pyrrolidone as substituent. In this way the biological relevant properties of the neutral N-ethylpyrrolidone group were combined with the advantage of the backbone degradability. It was also demonstrated that N- ethylpyrrolidone could modulate the degradation rates of mixed-substituent poly(organo)phosphazenes)65, 66. PYRP has been used as a replacement for non- degradable PVP, for example in the solubilization of the non water-soluble photosensitizer hypericin, providing comparable results to the known hypericin-PVP conjugate but with a degradable polymer67.

Figure 1-7. Chemical structure of some non-charged poly(organo)phosphazenes: a) Poly(di[2- (2-oxo-1-pyrrolidinyl)ethoxy]phosphazene) (PYRP), b) Poly[bis(2-(2- methoxyethoxy)ethoxy)phosphazene] (MEEP), and c) Jeffamine M-1000-substituted polyphosphazene

Poly[bis(2-(2-methoxyethoxy)ethoxy)phosphazene] (MEEP) (Figure 1-7b) synthesized by Allcock and co-workers68 showed also to be water-soluble and used as solid polymer electrolytes68-70. Indeed water soluble polymers with a wide range of side- chain lengths and various branched architectures are readily accessible widely commercially available PEGs71. However, such PEGylated polyphosphazenes tend to show highly stable towards hydrolytic degradation. Cosubstitution with hydrolysis sensitizing moieties can be used to prepare degradable variants72. Alternatively degradable water-soluble poly(organo)phosphazenes have been prepared using the hydrophilic Jeffamine M-1000 (Figure 1-7c)10. The higher lability of the P-N

11 bond compared to the P-O (as in MEEP) increases the susceptibility of the polymer to hydrolytic degradation. Water-soluble poly(organo)phosphazenes with controlled dimensions can be synthesized with varied lengths of Jeffamine polyetheramine grafted to the polymer backbone34. The branching density of the polymer could be increased by incorporation of two more arms per repeat unit. By incorporation of alkyne-functionalized groups into the polymer backbone and subsequent thiol-yne reaction, two thiol groups could be coupled onto a single alkyne group. Furthermore, multi-arm star poly(organo)phosphazenes have been synthesized 62. as well as dendritic molecular brush structures with up to 30 000 functional end groups, water solubility, narrow polydispersities and hydrodynamic diameters ranging from 10 to 70 nm.

Post-polymerization functionalization also allows incorporation of diverse saccharides on a poly(organo)phosphazene backbone to give water soluble glycopolymers27. A direct route has also been developed via the conjugation of lactobionic acid onto a primary amine-functionalized poly(organo)phosphazene73. A alternative route is provided by photochemical thiol-yne addition to a propargylamine bearing mixed-poly(organo)phosphazene. The high reactivity towards thioglucopyranose allows incorporation of two saccharide groups per triple bond74, with the thiol-yne reaction again proving useful to prepare highly functionalized, hydrophilic polyphosphazenes without the need for protecting group chemistry.

1.3.4. Amphiphilic Poly(organo)phosphazenes

Figure 1-8. Graphic representation of intra- and intermolecular self-assembly of the amphiphilic poly(organo)phosphazene in aqueous solution. Reproduced with permission from reference 76.

Amphiphilic poly(organo)phosphazenes consisting of hydrophilic and hydrophobic are of great interest. A unique feature compared to the conventional block- polymers usually utilized for self-assembly, is that even a randomly substituted

12

polyphosphazene backbone allows the micellar-like structure formation. This is presumably due to the high rotational freedom of the backbone allowing agglomeration of hydrophobic moieties This behavior for example can be observed in water-soluble polyphosphazenes loaded with hydrophobic drugs75, as well as in peptide based hybrid poly(organo)phosphazenes (Figure 1-8)76. These polymers were synthesized by addition of the Gly-Phe-Leu-Gly (GFLG) sequence in combination with the hydrophilic Jeffamine M-1000 oligomer.

A series of amphiphilic graft poly(organo)phosphazenes, poly(NIPAm-co- DMAA)/Glyet-PPPs (PNDGPs), have been synthesized with ethyl glycinate as hydrophobic side group and a bearing N-isopropylacrylamide (NIPAm) and N,N-dimethylacrylamide (DMAA) as hydrophilic side groups77. By variation of the ratio of hydrophilic to hydrophobic side groups some amphiphilic poly(organo)phosphazenes were investigated with tunable critical micelle concentration (CMC). In this case, the hydrophobic drug doxorubicin (DOX) was trapped in the core of the micelle. Cosubstitution of poly(organo)phosphazenes with hydrophilic poly(ethylene glycol) and hydrophobic ethyl-p-aminobenzoate also lead to amphiphilic polymers with tunable supramolecular structures due to the variation of the hydrophilic ratio78. Furthermore, by exchange of the hydrophobic ethyl-p-aminobenzoate to the pH-sensitive N,N- diisopropylethylenediamine (DPA) group some pH-responsive amphiphilic poly(organo)phosphazenes has been investigated, upon the protonation of the tertiary amine at lower pH values79.

1.3.5. Thermoresponsive polyphosphazenes

Thermoresponsive, sometimes referred as thermosensitive polymers, are polymers that demonstrate changes in their physical properties with temperature, for example a number of amphiphilic polymers show a lower critical solution temperature (LCST) in aqueous solutions, whereas above of this temperature the polymer chains collapse with a consequent precipitation or gelation of the polymer. Indeed some of the amphiphilic polymers described in the previous section also demonstrate thermoresponsive behavior. For example, PNDGPs shows an LCST around 39°C and its temperature sensitivity is utilized to release the encapsulated material77. Thermoresponsive polymers with an LCST around or below body temperature are of particular interest for medical applications.

13 Branched and linear alkylene ether based poly(organo)phosphazenes also present LCST behavior in the range of 40°C-80°C60. More recently, amine-capped Jeffamine with different ratios of ethylene oxide (EO) to propylene oxide (PO) have been used as substituents. These polymers have LCST values in the range of 18°C to 90°C (Figure 1-9a)80. Interestingly, these polymers were observed to agglomerate into colloids in the 100 nm range above their LCST, presumably due the high flexibility of the poly(organo)phosphazenes (Figure 1-9b).

Figure 1-9. a) Jeffamine substituted polyphosphazenes, with increasing hydrophilicity show higher LCST ranges and b) temperature driven self-assembly of amphiphilic polyphosphazenes.Adapted with permission from reference 80.

Another commonly employed method to prepare thermoresponsive poly(organo)phosphazenes is through mixed substitution of hydrophobic amino acid with oligomeric alkoxy PEGs (Figure 1-10). The LCST values can be tailored through varying the hydrophilic/hydrophobic composition. For example, the polymer bearing glycine ethyl ester and 0.31 mol of PEG350/unit has an LCST of 35°C while 1.12 mol/unit shows LCST of 83.7°C. Longer PEG chains also increase the LCST as it is the 81 case of 1.09 mol PEG750/unit in which the LCST is 98.5°C . In contrast, the more hydrophobic amino acid ester group contributes to lower LCST values. In a similar way, thermoresponsive poly(organo)phosphazene bearing L-isoleucine ethyl ester, as hydrophobic substituent, and hydrophilic amino-methoxy-PEG oligomers has been synthesized82. Again, the LCST values also increase with the ratio of PEG. Variations on the ratio between PEG and L-isoleucine ethyl ester allow tailoring the degradation rate and mechanical properties of the polymer.

14

Figure 1-10. Structure of some typical thermoresponsive poly(organo)phosphazenes: a) Jeffamine-substituted polyphosphazenes, and b) Mixed substituted polyphosphazene with hydrophobic amino acid ester and oligomeric alkoxy PEG.

1.3.6. Hydrogels

The ability of thermoresponsive polymers to afford physical and chemical changes as a function of temperature opens the opportunity to synthesize thermoresponsive hydrogels. A significant amount of work related to poly(organo)phosphazene-based injectable hydrogels for drug delivery systems has been developed by S.C. Song and co-workers in which its synthesis, sol-gel properties and degradability have been studied81-83. The gelation properties of the polymers, based on amino acid esters and amino-PEG, can be tailored by variations of PEG chain length, composition of substituents, amino acid esters and pH83. Solution temperatures influence not only sol- gel transition, but also degradation rates84.

Fine-tuning of the systems, including incorporation carboxylic moieties facilitated the dual interaction of polymeric nanoparticles (D-NP) 85. The study showed the effectiveness of the injectable bone morphogenetic protein-2(BMP-2)/D-NP nanocomplex hydrogel system, in which the aqueous solution becomes a hydrogel at body temperature, and the sustained protein release system besides its degradation in vivo and in vitro (Figure 1-11). The effectiveness of the hydrogel to regenerate a vertical bone at the peri-implant site has also been demonstrated86. Moreover, varying the chain-lengths and amounts of the carboxylic acid side group the physical properties, such as dissolution/degradation behavior, water uptake, or BMP-2 release rate could be tailored87.

15

Figure 1-11. a) Structure of the poly(organo)phosphazene for D-NPs. b) Temperature- dependent viscosity changes of D-NPs in aqueous solution. c) Schematic illustration of the nanocomplexes formation between BMP-2 and D-NPs due to the ionic and hydrophobic interactions, followed by hydrogel formation after injection and its sustained release to assist the bone generation. Adapted with persimission from reference 85. Copyright 2015 Elsevier.

Supramolecular-structured hydrogels can be also prepared from water-soluble polyphosphazenes, for example by inclusion complexation between mPEG grafted poly(organo)phosphazenes and cyclodextrin in aqueous solution. Through the addition of cyclodextrin physical cross-linking occurs with mPEG while unreacted mPEG chains assist the hydrophilicity (Figure 1-12). Gelation times, stability and release of a model protein could be tailored with variation of polymer structure and concentration88. Moreover, ionic hydrogels, in which anionic polymer chains form “bridges” with cations, have been synthesized for encapsulation89. For example PCPP polyelectrolytes in presence of di- or trivalent cations, such as calcium, copper, or aluminum, can be cross-linked to form ionic hydrogels42. The salts form “bridges” between the polymer chains leading to cross-linking that can be broken in presence of monovalent cations.

16

Figure 1-12. Combination of poly(ethylene glycol) methyl ether-based poly(organo)phosphazene (PmPEG2000) with -cyclodextrin (-CD) lead to a rapid supramolecular hydrogel formation. Adapted with permission from reference 88.

Of the hydrogels described above, all of them are formed by physical gelation in absence of covalent interactions. Chemical cross-linkers have also been employed in gelation, primarily with the goal of improving the mechanical properties of the gel. Cosubstitution of poly(organo)phosphazenes with methacrylate groups has also been used, leading to photo-crosslinked polymers upon irradiation with UV light90. After physical cross-linking at body temperature, the gel is irradiated producing a cross- linked system with improved mechanical properties. Varying the amount of methacrylate the mechanical gel property and degradation rate could be controlled. However, UV light exposure is not appropriate for use in biological tissues due to its damaging nature and low penetration. In order to overcome these limitations, systems have been developed in which the polymer is UV-pretreated and cross-linked only after injection91. Interestingly, the polymer is exposed to UV light in aqueous solution at low temperatures and it cross-linked only after injection into the body due to the rise in temperature. Allyl groups could be incorporated via quaternization method into a tertiary amine to assist the thiol-ene reaction. After UV light irradiation, hydrogels were easily obtained through the cross-linking reaction between PEG-dithiol and the allyl- functionalized polyphosphazene (Figure 1-13)56. This cationic hydrogel has been studied for its binding and release effects on biomacromolecules.

17

Figure 1-13. a) Schematic illustration of the quaternization of the polyphosphazene by allyl bromide followed by thiol-ene reaction with PEG-dithiol in aqueous solution. b) Polymer solution before (top) and after (bottom) UV irradiation. Adapted with permission from reference56. Copyright 2015 The Royal Society of Chemistry.

1.4. Degradation mechanisms

Figure 1-14. Proposed hydrolytic degradation mechanism of poly(organo)phosphazenes in a) neutral and b) acidic media.

18

The complete backbone degradation of some poly(organo)phosphazenes is of considerable interest in numerous biomedical applications. Poly(organo)phosphazene degradation occurs via hydrolytic cleavage of the organic substituents leading to the instable hydroxyphosphazenes and phosphazenes species triggering further degradation to phosphate and ammonium ions31, 72, 92 (Figure 1-14 route (a)). Importantly, the mixture of phosphates and ammonium salts is well-known to be benign72 so in combination with a suitable choice of the organic substituents, materials with non-toxic degradation products can be synthesized.

The degradation rate is tailored by the characteristics of the organic substituents, which also directly affect a number of different parameters such as mechanical stability and physicochemical properties, which also must be taken into account when the organic substituents are selected. Another aspect to consider is the complete replacement of chlorine atoms in the backbone by the nucleophiles as remaining chlorine atoms influence on the degradation mechanisms45, 48. Uncompleted macromolecular substitution shows accelerated hydrolytic breakdown and hence must be carefully assessed when preparing polymers with controlled degradation profiles.

Poly(organo)phosphazenes bearing amino acid derivatives have attracted much attention due to their biocompatibility and availability and hence intense research efforts have been carried out into the hydrolysis of different poly[(amino acid ester)phosphazenes] in the solid state64, 92-94 with a simultaneous decrease in polymer mass, molecular weight and the release of low molecular degradation products. Although these polymers are generally hydrophobic, their hydrolytic sensitivity has also been applied via cosubstitution to a wide variety of water-soluble polymers. One example could be by incorporation of amino acid moieties as spacer between the polymer backbone and the hydrophilic Jeffamine M-1000 oligomer10. In this way enhanced degradation could be observed compared to the polyphosphazenes without amino acid linkers (Figure 1-7c). The polymer without a spacer in pH 5 solution at 37°C showed 14 % of the backbone degraded to phosphate within 4 weeks, while the polymer bearing valine spacer showed 50 % of the phosphate formation. Using a less bulky amino acid, such as glycine, the polymer underwent almost complete degradation under the same conditions.

Despite the fact that P-N bond is more hydrolytically sensitive compared to the stronger P-O bond, there are some water soluble polyphosphazenes with P-O linkages that show hydrolytic instability with acceptable degradation times. One example is the

19 water soluble PYRP polymer (Figure 1-7a) which showed a decrease in the molecular weight and release of 1-(2-hydroxyethly)-2-pyrrolidone confirming the hydrolytic cleavage of the side group65. The results also showed hydrolytic degradation of the polymer in aqueous solutions at different pH values, where degradation rates increased in the acidic environment (Figure 1-15). PCPP polymer (Figure 1-3a) also showed reduction in the molecular weight over time95 but in this case promoted also by the carboxyl group which can catalyze the degradation (see discussion of self-catalyzed degradation in Stimuli responsive degradation mechanisms).

Figure 1-15. Hydrolytic degradation of PYRP at 55°C and different pH. Representation of molecular weight loss by gel permeation chromatography (GPC) with poly(ethylene oxide) PEO standards.Adapted with permission from reference 65.

Although the degradation rates of different poly(organo)phosphazenes are to some degree predictable from the chemical structures (see Allcock et al.72), a direct comparison of the degradation kinetics of most reported polymers is difficult due to the differences in the experimental conditions (e.g. temperature, solid or solution state and deionized water or buffered systems) between the investigated systems. The PYRP polymer (Figure 1-7a) for example shows ~80 % of Mw loss in ~ 20 days (pH 7.4 tris buffer at 55°C), while the PCPP polymer (Figure 1-3a) in the same conditions shows 65 only ~30 % of Mw loss . DMAEA-ppz (Figure 1-4a) with P-N bond has a half-life of 24 days at 37°C and neutral pH (pH 7.5 tris-HCl buffer), and enhanced degradation at pH 5 with a half-life of 5 days52. By addition of some imidazole groups (Figure 1-4b) the degradation of the polymer has shown to be slightly faster shortening 2 days respectively53.

20

1.4.1. Stimuli responsive degradation mechanisms

Degradable water soluble polymers discussed until now show also pH dependence in which the acidic conditions enhance the degradation kinetics of many poly(organo)phosphazenes10, 64, 65. At lower pH values the protonation of the nitrogen atom occurs followed by a migration of the positive charge to the backbone and therefore, the nucleophilic attack of water is assisted96 (Figure 1-14 route (b)). This pH triggered-response is an interesting mechanism for biomedical applications, as it opens the possibility to design systems which respond to changes in pH, for example for targeting of intracellular environments or diseased or cancerous tissue97.

It is also reported that the presence of acidic groups close to the polymer backbone enhances the hydrolysis rates due to intra-molecular catalysis93, 98. The nitrogen atom is protonated due to the proximity of the acid group and further reaction with water leads to side group cleavage. In this context, poly(organo)phosphazenes bearing a tetrapeptide sequence Gly-Phe-Leu-Gly presented by showed enzymatically triggered degradation by which after enzymatic cleavage of a peptide bond, the free carboxyl group promoted the backbone degradation (Figure 1-16)76.

Figure 1-16. Schematic illustration of the enzymatically triggered and acid-catalyzed degradation of GFLG-peptide based poly(organo)phosphazenes. Reproduced with permission from reference 76.

In early works by Allcock and coworkers, poly(glycine)phosphazene was reported to be difficult to isolate due to its rapid hydrolysis98, presumably due to the proximity of the carboxylic acid moiety to the phosphazene backbone. In a recent study, the inherently instable poly(glycine)phosphazene, has hence been protected (caged) by phenylboronate moieties, enhancing it hydrolytic stability99. A mixed-substituent

21 polymer was then prepared, based on a glycinate arylboronic acid pinacol ester with Jeffamine M-1000 to provide the water solubility to the system. The arylboronate pinacol ester group oxidizes to phenol in presence of the reactive oxygen species 100, 101 (ROS) H2O2 . ROS are important in a wide range of biological functions but it may damage biomolecules and therefore contribute to several diseases when it is overproduced. The polymer showed selective degradation in presence of 10 mM aqueous solution of H2O2 due to the oxidation of the caging group to phenol and thus, its self-immolation to uncage the unstable glycine-substituted polyphosphazene which catalyzes the polymer degradation (Figure 1-17). Analysis by gel permeation chromatography (GPC) showed a decrease in the polymer peak besides an increase in low molecular weight region corresponding to the cleavage of side groups. 31P NMR analysis confirmed also backbone degradation with a visible sharp peak arising with the time due to phosphate formation. Meanwhile, the polymer showed no degradation signal in absence of H2O2 in the same time-frame as it could be confirmed by GPC and 31P NMR spectroscopy. In order to prove the exclusive degradation of the polymer due to the presence of the arylboronate pinacol ester caging group, poly(glycine ethyl ester)phosphazene was exposed even to a higher concentration of H2O2 (100 mM) with no sign of degradation according to 31P NMR spectroscopy.

Figure 1-17. Proposed self-immolation mechanism of the caged poly(glycine)phosphazene 99 upon H2O2 exposure.Adapted with permission from reference .

22

In a similar manner, coumarin-based photocages have also been incorporated into a water-soluble poly(organo)phosphazene that cleaves upon exposure to visible light102. The polymer is shown to be stable in aqueous solution in the dark while fast degradation signs could be observed after irradiation. The sensitiveness to degradation under visible light irradiation opens the opportunity to its safe use in medical applications with longer wavelengths for deeper tissue penetration.

1.5. Summary and outlook

Water-soluble poly(organo)phosphazenes have attracted much attention in a wide variety of applications, especially in medical applications. The facile functionalization and more importantly easily tunable degradation rates allow the synthesis of a vast number of poly(organo)phosphazenes that fully degrade to small molecules. Due to the relative lack of water soluble synthetic organic polymers with suitable and easily tunable degradation rates, this is a highly promising niche market for such polyphosphazenes.

However, while the degradation of some non-water soluble poly(organo)phosphazenes has been systematic studied in solid state (e.g. amino acid- based polymers), investigations in solution state have been much less common. Furthermore, the experimental conditions used have varied widely, hence a more systematic study is required to fully understand the relationship between the chemical structure and the hydrolytic stability in solution.

Particularly promising applications for such water soluble and degradable polymers included in medical applications are drug103, 104, gene52 and vaccine delivery105. Water soluble poly(organo)phosphazenes have been demonstrated to be highly convenient for such applications due to the facile functionalization with a wide variety of nucleophiles, along with the tunable degradability and fully degradation to small molecules, forming pH-buffered system. Moreover, the recent development of polyphoshazenes with responsive degradation that can be switched-on in reply to certain stimuli promises to give further momentum to the field.

23 1.6. References 1. Halake, K.; Birajdar, M.; Kim, B. S.; Bae, H.; Lee, C.; Kim, Y. J.; Kim, S.; Kim, H. J.; Ahn, S.; An, S. Y.; Lee, J., Recent application developments of water-soluble synthetic polymers. J. Indust. Eng. Chem. 2014, 20, 3913-3918. 2. Treiser, M.; Abramson, S.; Langer, R.; Kohn, J., Chapter I.2.6 - Degradable and Resorbable Biomaterials. In Biomaterials Science, 3rd ed.;, Ratner, B. D.; Hoffman, A. S.; Schoen, F. J.; Lemons, J. E., Eds. Academic Press: San Diego, USA, 2013; pp 179-195. 3. Gross, R. A.; Kalra, B., Biodegradable Polymers for the Environment. Science 2002, 297, 803-807. 4. Lambert, S.; Wagner, M., Environmental performance of bio-based and biodegradable plastics: the road ahead. Chem. Soc. Rev. 2017, 46, 6855-6871. 5. Nair, L. S.; Laurencin, C. T., Biodegradable polymers as biomaterials. Progress in Polymer Science 2007, 32, 762-798. 6. Ulery, B. D.; Nair, L. S.; Laurencin, C. T., Biomedical applications of biodegradable polymers. J. Polym. Sci., Part B: Polym. Phys. 2011, 49, 832-864. 7. Delplace, V.; Nicolas, J., Degradable vinyl polymers for biomedical applications. Nat. Chem. 2015, 7, 771-784. 8. Bertrand, O.; Gohy, J.-F., Photo-responsive polymers: synthesis and applications. Polym. Chem. 2017, 8, 52-73. 9. Mutlu, H.; Barner-Kowollik, C., Green chain-shattering polymers based on a self- immolative azobenzene motif. Polym. Chem. 2016, 7, 2272-2279. 10. Wilfert, S.; Iturmendi, A.; Schoefberger, W.; Kryeziu, K.; Heffeter, P.; Berger, W.; Brüggemann, O.; Teasdale, I., Water-soluble, biocompatible polyphosphazenes with controllable and pH-promoted degradation behavior. J. Polym. Sci., Part A: Polym. Chem. 2014, 52, 287-294. 11. Göpferich, A., Mechanisms of polymer degradation and erosion. Biomaterials 1996, 17, 103-114. 12. Van Cauwenberghe, L.; Vanreusel, A.; Mees, J.; Janssen, C. R., Microplastic pollution in deep-sea sediments. Environ. Pollut. 2013, 182, 495-499. 13. Wright, S. L.; Thompson, R. C.; Galloway, T. S., The physical impacts of microplastics on marine organisms: A review. Environ. Pollut. 2013, 178, 483-492. 14. Bagheri, A. R.; Laforsch, C.; Greiner, A.; Agarwal, S., Fate of So-Called Biodegradable Polymers in Seawater and Freshwater. Global Challenges 2017, 1, 1700048.

24

15. Shastri, V. P., Non-Degradable Biocompatible Polymers in Medicine: Past, Present and Future. Curr. Pharm. Biotechnol. 2003, 4, 331-337. 16. Pan, H.; Yang, J.; Kopečková, P.; Kopeček, J., Backbone Degradable Multiblock N-(2-Hydroxypropyl)methacrylamide Copolymer Conjugates via Reversible Addition−Fragmentation Chain Transfer Polymerization and Thiol−ene Coupling Reaction. Biomacromolecules 2011, 12, 247-252. 17. Sponchioni, M.; Morosi, L.; Lupi, M.; Capasso Palmiero, U., Poly(HPMA)-based copolymers with biodegradable side chains able to self assemble into nanoparticles. RSC Adv. 2017, 7, 50981-50992. 18. Tappertzhofen, K.; Beck, S.; Montermann, E.; Huesmann, D.; Barz, M.; Koynov, K.; Bros, M.; Zentel, R., Bioreducible Poly-l-Lysine–Poly HPMA Block Copolymers Obtained by RAFT-Polymerization as Efficient Polyplex-Transfection Reagents. Macromol. Biosci. 2016, 16, 106-120. 19. Teasdale, I.; Brüggemann, O., Polyphosphazenes for medical applications. Smithers RAPRA: Shrewsbury, UK, 2014. 20. Kadajji, V. G.; Betageri, G. V., Water Soluble Polymers for Pharmaceutical Applications. Polymers 2011, 3 (4), 1972-2009. 21. Haag, R.; Kratz, F., Polymer therapeutics: Concepts and applications. Angew. Chem. Int. Edit. 2006, 45, 1198-1215. 22. Duncan, R., Polymer therapeutics: Top 10 selling pharmaceuticals — What next? J. Control. Release 2014, 190, 371-380. 23. Kopeček, J., Polymer–drug conjugates: Origins, progress to date and future directions. Adv. Drug Deliv. Rev. 2013, 65, 49-59. 24. Duncan, R., Polymer therapeutics as nanomedicines: new perspectives. Curr. Opin. Biotechnol. 2011, 22, 492-501. 25. Markovsky, E.; Baabur-Cohen, H.; Eldar-Boock, A.; Omer, L.; Tiram, G.; Ferber, S.; Ofek, P.; Polyak, D.; Scomparin, A.; Satchi-Fainaro, R., Administration, distribution, metabolism and elimination of polymer therapeutics. J. Control. Release 2012, 161, 446-460. 26. Henke, H.; Brüggemann, O.; Teasdale, I., Branched Macromolecular Architectures for Degradable, Multifunctional Phosphorus-Based Polymers. Macromol. Rapid Commun. 2017, 38, 1600644. 27. Allcock, H. R., Chemistry and applications of polyphosphazenes. Wiley: Hoboken, USA, 2003.

25 28. Tian, Z.; Hess, A.; Fellin, C. R.; Nulwala, H.; Allcock, H. R., Phosphazene High Polymers and Models with Cyclic Aliphatic Side Groups: New Structure–Property Relationships. Macromolecules 2015, 48, 4301-4311. 29. Allcock, H. R., The expanding field of polyphosphazene high polymers. Dalton Trans. 2016, 45, 1856-1862. 30. Allcock, H. R., A Perspective of Polyphosphazene Research. J. Inorg. Organomet. Polym. Mater. 2007, 16, 277-294. 31. Rothemund, S.; Teasdale, I., Preparation of polyphosphazenes: a tutorial review. Chem. Soc. Rev. 2016, 45, 5200-5215. 32. Wang, Z.; Manners, I., Synthesis and Characterization of Water-Soluble Cationic and Anionic Polythionylphosphazene Polyelectrolytes. Macromolecules 2005, 38, 5047-5054. 33. Honeyman, C. H.; Manners, I., Ambient Temperature Synthesis of Poly(dichlorophosphazene) with Molecular Weight Control. J. Am. Chem. Soc 1995, 117, 7035-7036. 34. Henke, H.; Wilfert, S.; Iturmendi, A.; Brüggemann, O.; Teasdale, I., Branched Polyphosphazenes with Controlled Dimensions. J. Polym. Sci., Part A: Polym. Chem. 2013, 51, 4467-4473. 35. Blackstone, V.; Lough, A. J.; Murray, M.; Manners, I., Probing the Mechanism of the PCl5−Initiated Living Cationic Polymerization of the Phosphoranimine Cl3P═NSiMe3 using Model Compound Chemistry. J. Am. Chem. Soc. 2009, 131, 3658-3667. 36. Soto, A. P.; Manners, I., Poly(ferrocenylsilane-b-polyphosphazene) (PFS-b-PP): A New Class of Organometallic−Inorganic Block Copolymers. Macromolecules 2009, 42, 40-42. 37. Wilfert, S.; Henke, H.; Schoefberger, W.; Brüggemann, O.; Teasdale, I., Chain- End-Functionalized Polyphosphazenes via a One-Pot Phosphine-Mediated Living Polymerization. Macromol. Rapid Commun. 2014, 35, 1135-1141. 38. Suárez-Suárez, S.; Carriedo, G. A.; Soto, A. P., Tuning the Chirality of Block Copolymers: From Twisted Morphologies to Nanospheres by Self-Assembly. Chem. Eur. J. 2015, 21, 14129-14139. 39. Morozowich, N. L.; Weikel, A. L.; Nichol, J. L.; Chen, C.; Nair, L. S.; Laurencin, C. T.; Allcock, H. R., Polyphosphazenes Containing Vitamin Substituents: Synthesis, Characterization, and Hydrolytic Sensitivity. Macromolecules 2011, 44, 1355-1364.

26

40. Allcock, H. R.; Connolly, M. S.; Sisko, J. T.; Al-Shali, S., Effects of organic side group structures on the properties of poly(organophosphazenes). Macromolecules 1988, 21, 323-334. 41. Allcock, H. R., Polyphosphazene elastomers, , and other soft materials. Soft Matter 2012, 8, 7521-7532. 42. Allcock, H. R.; Kwon, S., An ionically crosslinkable polyphosphazene: poly[bis(carboxylatophenoxy)phosphazene] and its hydrogels and membranes. Macromolecules 1989, 22, 75-79. 43. Payne, L.; Jenkins, S.; Andrianov, A.; Roberts, B., Water-Soluble Phosphazene Polymers for Parenteral and Mucosal Vaccine Delivery. In Vaccine Design, Powell, M.; Newman, M., Eds. Springer US: 1995; Vol. 6, pp 473-493. 44. Andrianov, A. K.; DeCollibus, D. P.; Gillis, H. A.; Kha, H. H.; Marin, A.; Prausnitz, M. R.; Babiuk, L. A.; Townsend, H.; Mutwiri, G., Poly[di(carboxylatophenoxy)phosphazene] is a potent adjuvant for intradermal immunization. Proc. Natl. Acad. Sci. 2009, 106, 18936-18941. 45. Andrianov, A. K.; Svirkin, Y. Y.; LeGolvan, M. P., Synthesis and Biologically Relevant Properties of Polyphosphazene Polyacids. Biomacromolecules 2004, 5, 1999-2006. 46. Eng, N. F.; Garlapati, S.; Gerdts, V.; Potter, A.; Babiuk, L. A.; Mutwiri, G. K., The potential of polyphosphazenes for delivery of vaccine antigens and immunotherapeutic agents. Curr. Drug Deliv. 2010, 7, 13-20. 47. Mutwiri, G.; Benjamin, P.; Soita, H.; Townsend, H.; Yost, R.; Roberts, B.; Andrianov, A. K.; Babiuk, L. A., Poly[di(sodium carboxylatoethylphenoxy)phosphazene] (PCEP) is a potent enhancer of mixed Th1/Th2 immune responses in mice immunized with influenza virus antigens. Vaccine 2007, 25, 1204-1213. 48. Andrianov, A. K.; Marin, A.; Chen, J., Synthesis, Properties, and Biological Activity of Poly[di(sodium carboxylatoethylphenoxy)phosphazene]. Biomacromolecules 2006, 7, 394-399. 49. Wycisk, R.; Pintauro, P. N., Sulfonated polyphosphazene ion-exchange membranes. J. Memb. Sci. 1996, 119, 155-160. 50. Paulsdorf, J.; Burjanadze, M.; Hagelschur, K.; Wiemhöfer, H. D., Ionic conductivity in polyphosphazene polymer electrolytes prepared by the living cationic polymerization. Solid State Ion. 2004, 169, 25-33.

27 51. Allcock, H. R.; Fitzpatrick, R. J.; Salvati, L., Sulfonation of (aryloxy)- and (arylamino)phosphazenes: small-molecule compounds, polymers, and surfaces. Chem. Mater. 1991, 3, 1120-1132. 52. Luten, J.; van Steenis, J. H.; van Someren, R.; Kemmink, J.; Schuurmans- Nieuwenbroek, N. M. E.; Koning, G. A.; Crommelin, D. J. A.; van Nostrum, C. F.; Hennink, W. E., Water-soluble biodegradable cationic polyphosphazenes for gene delivery. J. Control. Release 2003, 89, 483-497. 53. Yang, Y.; Xu, Z.; Jiang, J.; Gao, Y.; Gu, W.; Chen, L.; Tang, X.; Li, Y., Poly(imidazole/DMAEA)phosphazene/DNA self-assembled nanoparticles for gene delivery: Synthesis and in vitro transfection. J. Control. Release 2008, 127, 273-279. 54. Ma, C.; Zhang, X.; Du, C.; Zhao, B.; He, C.; Li, C.; Qiao, R., Water-Soluble Cationic Polyphosphazenes Grafted with Cyclic Polyamine and Imidazole as an Effective Gene Delivery Vector. Bioconjugate Chem. 2016, 27, 1005-1012. 55. Liu, X.; Zhang, H.; Tian, Z.; Sen, A.; Allcock, H. R., Preparation of quaternized organic-inorganic hybrid brush polyphosphazene-co-poly[2-(dimethylamino)ethyl methacrylate] electrospun fibers and their antibacterial properties. Polym. Chem. 2012, 3, 2082-2091. 56. Qian, Y.-C.; Chen, P.-C.; Zhu, X.-Y.; Huang, X.-J., Click synthesis of ionic strength-responsive polyphosphazene hydrogel for reversible binding of enzymes. RSC Adv. 2015, 5, 44031-44040. 57. Morozowich, N. L.; Mondschein, R. J.; Allcock, H. R., Comparison of the Synthesis and Bioerodible Properties of N-Linked Versus O-Linked Amino Acid Substituted Polyphosphazenes. J. Inorg. Organomet. Polym. Mater. 2014, 24, 164-172. 58. Allcock, H. R.; Kwon, S., Glyceryl polyphosphazenes: synthesis, properties, and hydrolysis. Macromolecules 1988, 21, 1980-1985. 59. Steinhilber, D.; Sisson, A. L.; Mangoldt, D.; Welker, P.; Licha, K.; Haag, R., Synthesis, Reductive Cleavage, and Cellular Interaction Studies of Biodegradable, Polyglycerol Nanogels. Adv. Funct. Mater. 2010, 20, 4133-4138. 60. Abbina, S.; Vappala, S.; Kumar, P.; Siren, E. M. J.; La, C. C.; Abbasi, U.; Brooks, D. E.; Kizhakkedathu, J. N., Hyperbranched polyglycerols: recent advances in synthesis, biocompatibility and biomedical applications. J. Mater. Chem. B 2017, 5, 9249-9277. 61. Calderón, M.; Quadir, M. A.; Sharma, S. K.; Haag, R., Dendritic Polyglycerols for Biomedical Applications. Adv. Mater. 2010, 22, 190-218.

28

62. Henke, H.; Posch, S.; Brüggemann, O.; Teasdale, I., Polyphosphazene Based Star-Branched and Dendritic Molecular Brushes. Macromol. Rapid Commun. 2016, 37, 769-774. 63. Linhardt, A.; König, M.; Iturmendi, A.; Henke, H.; Brueggemann, O.; Teasdale, I., Degradable, dendritic polyols on a branched polyphosphazene backbone. Ind. Eng. Chem. Res. 2018, 57, 3602-3609. 64. Andrianov, A. K.; Marin, A., Degradation of Polyaminophosphazenes: Effects of Hydrolytic Environment and Polymer Processing. Biomacromolecules 2006, 7, 1581- 1586. 65. Andrianov, A. K.; Marin, A.; Peterson, P., Water-Soluble Biodegradable Polyphosphazenes Containing N-Ethylpyrrolidone Groups. Macromolecules 2005, 38, 7972-7976. 66. Martinez, A. P.; Qamar, B.; Fuerst, T. R.; Muro, S.; Andrianov, A. K., Biodegradable “Smart” Polyphosphazenes with Intrinsic Multifunctionality as Intracellular Protein Delivery Vehicles. Biomacromolecules 2017, 18, 2000-2011. 67. Feinweber, D.; Verwanger, T.; Brueggemann, O.; Teasdale, I.; Krammer, B., Applicability of new degradable hypericin-polymer-conjugates as photosensitizers: principal mode of action demonstrated by in vitro models. Photochem. Photobiol. Sci. 2014, 13, 1607-1620. 68. Blonsky, P. M.; Shriver, D. F.; Austin, P.; Allcock, H. R., Polyphosphazene solid electrolytes. J. Am. Chem. Soc. 1984, 106, 6854-6855. 69. Allcock, H. R.; O'Connor, S. J. M.; Olmeijer, D. L.; Napierala, M. E.; Cameron, C. G., Polyphosphazenes Bearing Branched and Linear Oligoethyleneoxy Side Groups as Solid Solvents for Ionic Conduction. Macromolecules 1996, 29, 7544-7552. 70. Allcock, H. R.; Sunderland, N. J.; Ravikiran, R.; Nelson, J. M., Polyphosphazenes with Novel Architectures: Influence on Physical Properties and Behavior as Solid Polymer Electrolytes. Macromolecules 1998, 31, 8026-8035. 71. Allcock, H. R.; Dudley, G. K., Lower Critical Solubility Temperature Study of Alkyl Ether Based Polyphosphazenes. Macromolecules 1996, 29, 1313-1319. 72. Allcock, H. R.; Morozowich, N. L., Bioerodible polyphosphazenes and their medical potential. Polym. Chem. 2012, 3, 578-590. 73. Yang, Y.; Zhang, Z.; Chen, L.; Gu, W.; Li, Y., Galactosylated Poly(2-(2- aminoethyoxy)ethoxy)phosphazene/DNA Complex Nanoparticles: In Vitro and In Vivo Evaluation for Gene Delivery. Biomacromolecules 2010, 11, 927-933.

29 74. Chen, C.; Xu, H.; Qian, Y.-C.; Huang, X.-J., Glycosylation of polyphosphazenes by thiol-yne click chemistry for lectin recognition. RSC Adv. 2015, 5, 15909-15915. 75. Aichhorn, S.; Linhardt, A.; Halfmann, A.; Nadlinger, M.; Kirchberger, S.; Stadler, M.; Dillinger, B.; Distel, M.; Dohnal, A.; Teasdale, I.; Schöfberger, W., A pH-sensitive Macromolecular Prodrug as TLR7/8 Targeting Immune Response Modifier. Chem. Eur. J. 2017, 23, 17721-17726. 76. Linhardt, A.; König, M.; Schöfberger, W.; Brüggemann, O.; Andrianov, A.; Teasdale, I., Biodegradable Polyphosphazene Based Peptide-Polymer Hybrids. Polymers 2016, 8, 161-176. 77. Qiu, L. Y.; Wu, X. L.; Jin, Y., Doxorubicin-Loaded Polymeric Micelles Based on Amphiphilic Polyphosphazenes with Poly(N-isopropylacrylamide-co-N,N- dimethylacrylamide) and Ethyl Glycinate as Side Groups: Synthesis, Preparation and In Vitro Evaluation. Pharm. Res. 2009, 26, 946-957. 78. Zheng, C.; Qiu, L.; Zhu, K., Novel polymersomes based on amphiphilic graft polyphosphazenes and their encapsulation of water-soluble anti-cancer drug. Polymer 2009, 50, 1173-1177. 79. Zheng, C.; Yao, X.; Qiu, L., Novel Polymeric Vesicles with pH-Induced Deformation Character for Advanced Drug Delivery. Macromol. Biosci. 2011, 11, 338- 343. 80. Wilfert, S.; Iturmendi, A.; Henke, H.; Brüggemann, O.; Teasdale, I., Thermoresponsive Polyphosphazene-Based Molecular Brushes by Living Cationic Polymerization. Macromol. Symp, 2014, 337, 116-123. 81. Song, S.-C.; Lee, S. B.; Jin, J.-I.; Sohn, Y. S., A New Class of Biodegradable Thermosensitive Polymers. I. Synthesis and Characterization of Poly(organophosphazenes) with Methoxy-Poly(ethylene glycol) and Amino Acid Esters as Side Groups. Macromolecules 1999, 32, 2188-2193. 82. Lee, B. H.; Lee, Y. M.; Sohn, Y. S.; Song, S.-C., A Thermosensitive Poly(organophosphazene) Gel. Macromolecules 2002, 35, 3876-3879. 83. Lee, B. H.; Song, S.-C., Synthesis and Characterization of Biodegradable Thermosensitive Poly(organophosphazene) Gels. Macromolecules 2004, 37, 4533- 4537. 84. Park, M.-R.; Cho, C.-S.; Song, S.-C., In vitro and in vivo degradation behaviors of thermosensitive poly(organophosphazene) hydrogels. Polym. Degrad. Stab. 2010, 95, 935-944.

30

85. Seo, B.-B.; Choi, H.; Koh, J.-T.; Song, S.-C., Sustained BMP-2 delivery and injectable bone regeneration using thermosensitive polymeric nanoparticle hydrogel bearing dual interactions with BMP-2. J. Control. Release 2015, 209, 67-76. 86. Seo, B.-B.; Chang, H.-I.; Choi, H.; Koh, J.-T.; Yun, K.-D.; Lee, J.-Y.; Song, S.-C., New approach for vertical bone regeneration using in situ gelling and sustained BMP-2 releasing poly(phosphazene) hydrogel system on peri-implant site with critical defect in a canine model. J. Biomed. Mater. Res. Part B Appl. Biomater. 2018, 106, 751-759. 87. Seo, B.-B.; Koh, J.-T.; Song, S.-C., Tuning physical properties and BMP-2 release rates of injectable hydrogel systems for an optimal bone regeneration effect. Biomaterials 2017, 122, 91-104. 88. Tian, Z.; Chen, C.; Allcock, H. R., Injectable and Biodegradable Supramolecular Hydrogels by Inclusion Complexation between Poly(organophosphazenes) and α- Cyclodextrin. Macromolecules 2013, 46, 2715-2724. 89. Cohen, S.; Bano, M. C.; Visscher, K. B.; Chow, M.; Allcock, H. R.; Langer, R., Ionically crosslinkable polyphosphazene: a novel polymer for microencapsulation. J. Am. Chem. Soc. 1990, 112, 7832-7833. 90. Potta, T.; Chun, C.; Song, S.-C., Dual Cross-Linking Systems of Functionally Photo-Cross-Linkable and Thermoresponsive Polyphosphazene Hydrogels for Biomedical Applications. Biomacromolecules 2010, 11, 1741-1753. 91. Kim, Y.-M.; Potta, T.; Park, K.-H.; Song, S.-C., Temperature responsive chemical crosslinkable UV pretreated hydrogel for application to injectable tissue regeneration system via differentiations of encapsulated hMSCs. Biomaterials 2017, 112, 248-256. 92. Allcock, H. R.; Pucher, S. R.; Scopelianos, A. G., Poly[(amino acid ester)phosphazenes]: Synthesis, Crystallinity, and Hydrolytic Sensitivity in Solution and the Solid State. Macromolecules 1994, 27, 1071-1075. 93. Allcock, H. R.; Fuller, T. J.; Mack, D. P.; Matsumura, K.; Smeltz, K. M., Synthesis of Poly[(Amino Acid Alkyl Ester)Phosphazenes]. Macromolecules 1977, 10, 824-830. 94. Crommen, J. H. L.; Schacht, E. H.; Mense, E. H. G., Biodegradable polymers: II. Degradation characteristics of hydrolysis-sensitive poly[(organo)phosphazenes]. Biomaterials 1992, 13, 601-611. 95. DeCollibus, D. P.; Marin, A.; Andrianov, A. K., Effect of Environmental Factors on Hydrolytic Degradation of Water-Soluble Polyphosphazene Polyelectrolyte in Aqueous Solutions. Biomacromolecules 2010, 11, 2033-2038.

31 96. Carriedo, G. A.; García Alonso, F. J.; García Álvarez, J. L.; Presa Soto, A.; Tarazona, M. P.; Laguna, M. T. R.; Marcelo, G.; Mendicuti, F.; Saiz, E., Experimental and Theoretical Study of the Acidic Degradation of Poly(2,2′-dioxy-1,1′- biphenyl)phosphazene. Macromolecules 2008, 41, 8483-8490. 97. Binauld, S.; Stenzel, M. H., Acid-degradable polymers for drug delivery: a decade of innovation. Chem. Commun. 2013, 49, 2082-2102. 98. Allcock, H. R.; Fuller, T. J.; Matsumura, K., Hydrolysis pathways for aminophosphazenes. Inorg. Chem. 1982, 21, 515-521. 99. Iturmendi, A.; Monkowius, U.; Teasdale, I., Oxidation Responsive Polymers with a Triggered Degradation via Arylboronate Self-Immolative Motifs on a Polyphosphazene Backbone. ACS Macro Lett. 2017, 6, 150-154. 100. Song, C.-C.; Du, F.-S.; Li, Z.-C., Oxidation-responsive polymers for biomedical applications. J. Mater. Chem. B 2014, 2, 3413-3426. 101. Broaders, K. E.; Grandhe, S.; Fréchet, J. M. J., A Biocompatible Oxidation- Triggered Carrier Polymer with Potential in Therapeutics. J. Am. Chem. Soc. 2011, 133, 756-758. 102. Iturmendi, A.; Theis, S.; Maderegger, D.; Monkowius, U.; Teasdale, I., Unpublished. 103. Teasdale, I.; Wilfert, S.; Nischang, I.; Brüggemann, O., Multifunctional and biodegradable polyphosphazenes for use as macromolecular anti-cancer drug carriers. Polym. Chem. 2011, 2, 828-834. 104. Henke, H.; Kryeziu, K.; Banfić, J.; Theiner, S.; Körner, W.; Brüggemann, O.; Berger, W.; Keppler, B. K.; Heffeter, P.; Teasdale, I., Macromolecular Pt(IV) Prodrugs from Poly(organo)phosphazenes. Macromol. Biosci. 2016, 16, 1239-1249. 105. Cayatte, C.; Marin, A.; Rajani, G. M.; Schneider-Ohrum, K.; Snell Bennett, A.; Marshall, J. D.; Andrianov, A. K., PCPP-Adjuvanted Respiratory Syncytial Virus (RSV) sF Subunit Vaccine: Self-Assembled Supramolecular Complexes Enable Enhanced Immunogenicity and Protection. Mol. Pharm. 2017, 14, 2285-2293.

32

2. Stimuli responsive degradation of polyphosphazenes

2.1. pH promoted degradation

This chapter describes the synthesis of water-soluble poly(organo)phosphazenes with tailored degradation rates. A hydrophilic polyether monoamine, as water- solubilizing group, of approximately 1000 molecular weight (M-1000) is attached to the polyphosphazene backbone and by incorporation of some amino acid groups between the polymer backbone and M-1000, the degradability of the polymers is tuned. These polymers show a pH promoted degradation behavior, which is enhanced at lower pH- values. Preliminary cell viability studies demonstrate the suitability of these polymers to be used in biomedical applications, such as drug delivery, due to their biocompatibility and non-toxic nature.

My contribution to the paper

I conducted the experimental work of this project together with S. Wilfert, including synthesis, characterization and interpretation and degradation studies.

33 Water-Soluble, Biocompatible

Polyphosphazenes with Controllable and

pH-Promoted Degradation Behavior

Sandra Wilfert,a Aitziber Iturmendi,a Wolfgang Schoefberger,b,c Kushtrim Kryeziu,d Petra Heffeter,d,e Walter Berger,d,e Oliver Brüggemann,a and Ian Teasdalea*

––––––––– a Institute of Polymer Chemistry, Johannes Kepler University Linz, Welser Str. 42, 4060 Leonding, Austria. b Institute of Organic Chemistry, Johannes Kepler University Linz, Altenberger Strasse 69, 4040 Linz, Austria. c Faculty of Science, University of South Bohemia, Branišovská 31, 370 05 České Budějovice, Czech Republic. d Institute of Cancer Research and Comprehensive Cancer Center of the Medical University of Vienna, Medical University of Vienna, Borschkegasse 8a, 1090 Vienna, Austria. e Research and Platform “Translational Cancer Therapy Research” Vienna, Austria.

E-mail: [email protected]

–––––––––

J. Polym. Sci., Part A: Polym. Chem. 2014, 52, 287–294

34

Abstract

The synthesis of a series of novel, water-soluble poly(organophosphazenes) prepared via living cationic polymerization is presented. The degradation profiles of the polyphosphazenes prepared are analysed by GPC, 31P NMR spectroscopy and UV- Vis spectroscopy in aqueous media and show tunable degradation rates ranging from days to months, adjusted by subtle changes to the chemical structure of the polyphosphazene. Furthermore, it is observed that these polymers demonstrate a pH- promoted hydrolytic degradation behaviour, with a remarkably faster rate of degradation at lower pH values. These degradable, water soluble polymers with controlled molecular weights and structures could be of significant interest for use in aqueous biomedical applications, such as polymer therapeutics, in which biological clearance is a requirement and in this context cell viability tests are described which show the non-toxic nature of the polymers, as well as their degradation intermediates and products.

KEYWORDS: Polyphosphazenes, water-soluble polymers, biodegradable polymers, biocompatible polymers, polymer therapeutics.

2.1.1. Introduction

Functional polymers with a biodegradable backbone, biocompatible degradation products and controlled degradation profiles are of great promise for biomedical applications.1-3 Although providing eminent water-solubility and biocompatibility, the well-studied polyethylene glycol (PEG)4,5 and poly(N-(2-hydroxypropyl)methacrylamide (HPMA)6,7 lack in degradability, thus leading to accumulation in the body when used as high molecular weight polymers, limiting their applicability to some extent.8,9 In addition to polyacetals and polyketals,10 polyglutamic acid,11 dendritic polyesters,12 and recent work about polyphosphoesters,13 polyphosphazenes certainly represent a promising biodegradable class of polymers for biomedical applications.14,15 Due to their unique, degradable backbone, the properties of polyphosphazenes greatly depend on the choice of organic substituents attached to the phosphorus atoms.16 The major

35 precursor polymer, poly(dichlorophosphazene), can be obtained by thermal ring opening polymerization or living cationic polymerization, followed by macromolecular substitution of the chlorine atoms by a wide range of organic nucleophiles such as and alkoxides.17 Substitution of the polyphosphazene chain with suitable nucleophiles allows the design of poly(organophosphazenes) with versatile properties,17 multifunctionality18 and tunable hydrolytic sensitivity,19 all of which represent crucial features for polymers used in biomedical applications.

The incorporation of various specific substituents remarkably affects the hydrolytic degradability of the polymer.20 In this context, substituents such as amino acid ester,1,21-24 imidazolyl,25 glucosyl,26 glyceryl,27 and N-ethylpyrrolidone28 yield hydrolytically sensitive polyphosphazenes with different degradation rates. The hydrolytic stability can be fine-tuned by adapting the hydrophilicity/ hydrophobicity of the substituent,25 as well as mixed-substitutent polyphosphazenes29,30 with the amount of side groups, that generate hydrolytic sensitivity adjusted. There are several different pathways suggested for the degradation process of polyphosphazenes including the hydrolysis and thus release of the substituted side groups, followed by the attack of water on the phosphazene backbone, leading to the formation of hydroxyphosphazenes and phosphazanes.21 These intermediates then undergo further chain cleavage, generating a pH buffered system of nontoxic phosphate and .31

The purpose of this study was to synthesise a series of water-soluble polyphosphazenes via the living cationic polymerization32,33 with biocompatibility and controlled biodegradability. It is well-established that the rate of degradation of polyphosphazenes can be tailored through the addition of amino acid ester side groups.3,21 However, amino acid esters generally produce hydrophobic water insoluble polyphosphazenes, ideal for tissue engineering34 or implantable biomaterials,35 but not suitable for aqueous applications. Thus, with the aim to extend their applicability to aqueous applications, we sought to prepare water-soluble variants with similarly good control of degradability. Although a mixed substitution with both amino acid esters and hydrophilic groups is also possible,14 this route clearly has its upper incorporation limits (and thereby degradation limit), with increasing hydrophobic portion also leading to amphiphilic polymers and furthermore aggregation.36 The methods described herein show how this can be achieved through the direct incorporation of the prepared pegylated amino acid linkers, significantly broadening the spectrum of water soluble

36

degradable polyphosphazenes achievable, with degradation rates ranging from several days to months. Furthermore, we describe how decreases in pH-value lead to rapid acceleration of the degradation profile, as well as preliminary cell viability tests showing the biocompatibility of the polymers and the benign nature of their degradation products.

2.1.2. Experimental

Materials and methods

All synthetic procedures were carried out air-free either in a glovebox (MBRAUN) under argon or under nitrogen using standard Schlenk line techniques. The glassware was dried in an oven at 120°C prior to use. The amino capped statistical poly(ethylene oxide-co-propylene oxide), sold under the trade name Jeffamine M-1000 (referred to herein as M-1000) with a nominal molecular weight of 1000 g mol-1, was donated by Huntsman Performance Products (Netherlands) and used as received. Solvent were dried using standard laboratory methods. PCl5 was purified by sublimation and stored in the glovebox under argon. NEt3 was dried over molecular sieves and distilled prior to use. All other chemicals were purchased from Sigma Aldrich and used without further purification.

1H NMR spectra were measured on a Bruker Avance 300 spectrometer and a

Bruker DRX 500 spectrometer using CDCl3 or D2O as an internal reference. The 31P NMR experiments were conducted on both magnets at resonance frequencies of 121 and 202 MHz, respectively, using 85 % phosphoric acid as an external standard. Gel permeation chromatography (GPC) was carried out on a Viscothek GPCmax instrument using a PFG column from PSS (Mainz, Germany) (300 mm x 8 mm, 5 µm particle size). The samples were eluted with DMF containing 5 mM LiBr as the mobile phase at a flow rate of 0.75 mL min-1 at 60°C. The molecular weights were calculated relative to polystyrene standards from PSS using a conventional calibration of the refractive index detector. Dynamic light scattering (DLS) was performed on a Malvern Zetasizer Nano ZS instrument with a detection angle of 173° and a 4 mW He-Ne laser operating at a wavelength of 633 nm. The samples were filtered through a nylon microfilter (0.2 µm) prior to the measurement. The hydrodynamic diameter of the -1 polymers in deionized H2O (1 mg mL ) at 25°C was determined from the volume size distribution and is expressed as a mean value. FTIR spectra were obtained with a

37 Perkin Elmer Spectrum 100 FTIR spectrometer equipped with an ATR accessory. UV- Vis spectra were carried out on a Perkin Elmer Lambda 25 UV/VIS spectrophotometer.

Synthetic procedures

Trichlorophosphoranimine (Cl3P=N-SiMe3). The monomer was prepared using a 37,38 method adapted from the literature. 24.22 g LiN(SiMe3)2 (145 mmol) were dissolved in 400 mL anhydrous diethyl ether. The reaction was cooled to 0°C and stirred for 30 minutes. 12.66 mL PCl3 (19.87 g, 145 mmol) were added dropwise at 0°C and the solution was stirred for 30 minutes. 11.70 mL SO2Cl2 (19.53 g, 145 mmol) were added and the mixture was stirred for another hour at 0°C. The reaction was then filtered through celite and the volatiles were removed under vacuum. The product was purified twice by vacuum distillation (50°C, 4 mbar) to yield Cl3P=N-SiMe3 as a colourless liquid which was stored under an inert argon atmosphere at -35°C. Yield 15.00 g (46 %). 1 31 H NMR (300 MHz, CDCl3, δ): 0.17 (s, 9H) ppm. P NMR (121 MHz, CDCl3, δ): -54.37 ppm.

Synthesis of poly(dichlorophosphazene). The polymers were synthesised according to the procedure for the living cationic polymerization of 32 trichlorophosphoranimine. In the glovebox, the monomer Cl3P=N-SiMe3 (0.45 g,

2.01 mmol) and the initiator PCl5 (0.02 g, 0.08 mmol) were dissolved in CH2Cl2 and stirred at room temperature. After 12 hours, the solvent was removed under vacuum. The obtained poly(dichlorophosphazene) was used for macromolecular substitution 31 without further purification. Yield quantitative. P NMR (121 MHz, CDCl3, δ): - 18.16 ppm.

M-1000-gylcine and M-1000-valine. The following representative procedure describes the coupling reactions of M-1000 and the BOC-protected amino acids Boc- Gly-OH and Boc-Val-OH, respectively. Boc-Val-OH (1.63 g, 7.50 mmol), N- hydroxysuccinimde (0.86 g, 7.50 mmol) and N,N’-dimethylaminopyridine (0.09 g,

7.50 mmol) were dissolved in 80 mL CH2Cl2 and cooled to 0°C. In a second flask,

1.55 g N,N’-dicylcohexylcarbodiimide (7.50 mmol) were dissolved in 15 mL CH2Cl2 and transferred to the reaction mixture at 0°C. The mixture was allowed to warm to room temperature and stirred overnight. The formed precipitate was then removed by filtration. The filtrate was added to a solution of 7.50 g M-1000 (7.50 mmol) in CH2Cl2 and stirred for 2 days. The reaction was extracted twice with 10 % , twice with 5 % sodium hydrogen carbonate, saturated sodium chloride and dried over

38

MgSO4. The solvent was removed under vacuum and the product further dried under high vacuum to yield M-1000-Valine-Boc as a white wax-like product. Yield 6.39 g 1 (71 %). H NMR (300 MHz, CDCl3, δ): δ 1.07 (m, 8H), 1.37 (s, 9H), 3.30 (s, 3H), 3.57

(m, 87H) ppm. The M-1000-Valine-Boc was deprotected in TFA/CH2Cl2 (1/3) for 3 hours. The solvent was removed under vacuum, redissolved in CH2Cl2 followed by washing the product with 5 % sodium hydrogen carbonate and saturated sodium chloride. The organic phase was dried over MgSO4 and removed under vacuum. The product was further dried by co-evaporation with toluene and chloroform to obtain M- 1000-Valine. Yield: quantitative.

Synthesis of polymers. The following typical example procedure describes the synthesis of polymer 3 (Scheme 2.1-1). In the glovebox, poly(dichlorophosphazene) (0.15 g, 0.67 mmol, 1 eq) was dissolved in anhydrous THF. A solution of M-1000-

Valine (1.76 g, 1.60 mmol, 2.4 eq) in THF and NEt3 (0.16 g, 0.22 mL, 1.60 mmol, 2.4 eq) was then added to the polymer solution and stirred for 24 hours at room temperature. The suspension was filtered to remove the formed ammonium chloride and the solvent was concentrated under vacuum. The polymer was purified by several precipitations into chilled diethyl ether from THF and dried under high vacuum. Yield -1 1 0.57 g (38 %). FTIR (solid): νmax = 2883 (C-H), 1653 (C=O), 1109 (C-O) cm , H NMR 31 (500 MHz, D2O, δ): 0.99 (br, 9H), 3.21 (s, 3H), 3.53 (s, 82H) ppm. P NMR (202 MHz,

D2O, δ): 1.05 ppm.

All other polymers were synthesized accordingly, with the same molar ratio of monomer to initiator (25/1) and thus with a theoretical number of repeat units n = 50, and the amount of side groups adjusted in order to obtain the desired polymers. For polymers 2 and 4 with two different side groups, the two substituents were mixed in a molar ratio of 1/1 in THF followed by the addition of the polymer precursor and NEt3. All polymers were dried under vacuum to give colorless viscous to wax-like products in yields of 30 to 70 % (see supporting information for further characterization data).

39

Scheme 2.1-1. Synthesis of polymer 3 via the living polymerization of trichlorophosphoranimine followed by macromolecular substitution. Reagents and conditions: (i) PCl5, CH2Cl2, room temperature, 12 hours and (ii) NEt3, THF, room temperature, 24 hours.

Degradation studies

The degradation behavior of the polymers was studied by GPC, 31P NMR and UV- Vis spectroscopy. For 31P NMR degradation studies, 20 mg of the polymer were dissolved in 0.5 mL D2O (pH 7) and D2O acidified with HCl (pH 2, enhanced degradation conditions), respectively. The samples were incubated in the denoted solvents at room temperature and the changes of the phosphorus signals were monitored over a time period of up to two months.

The polymers were incubated in TRIS buffer (pH 7.4), sodium acetate buffer

(pH 5), or acidified H2O (pH 2, enhanced degradation conditions) in a concentration of 4 mg mL-1 at 37°C during the time of analysis. The polymer degradation medium was tested for the presence of inorganic phosphate. Aliquots of the degradation medium were taken in regular time intervals and mixed with a reagent solution consisting of ammonium molybdate, ascorbic acid, sulfuric acid and potassium antimonyl tartrate (method adapted from the literature).23 UV-Vis analysis of the mixtures were performed after 15 minutes of incubation time at 885 nm. The concentration of phosphate was calculated from a calibration curve using potassium dihydrogen phosphate and is given as the percentage compared to the theoretical phosphate amount that can be released from the polymer backbone. For further testing, samples of 0.75 mL were taken and the water was removed followed by GPC analysis as described above.

40

Cell viability studies

Cell culture. The following human cell lines were used in this study: the colon carcinoma cell line HCT116 (a gift from Dr. Vogelstein, John Hopkins University, Baltimore, MD), the ovarian carcinoma cell line A2780 (from Sigma-Aldrich) and the hepatoma model Hep3B (from American Type Culture Collection, Rockville, MD, USA).

All cells were grown in an humidified atmosphere with 5 % CO2 at 37°C in RMPI 1640 medium supplemented with 10 % fetal bovine serum (FCS), with the exception of HCT116 cells which were cultivated in McCoy’s culture medium with 10 % FCS. Cultures were regularly checked for Mycoplasma contamination.

Cytotoxicity assays. Cells were plated (2 x 103 cells/well) in 100 µl cell culture medium per well in 96-well plates. After a recovery period of 24 h, the polymers dissolved in 100 µl growth medium (0.01 to 10 µM) were added and the cells exposed for 72 h (pH 7.4). Polymer stock solutions were either freshly prepared or stored for 4 or 8 weeks at 37°C. The proportion of viable cells was determined by MTT assay following the manufacturer’s recommendations (EZ4U, Biomedica, Vienna, Austria). Given values are normalized to control samples without any treatment. For experiments at pH 6.0, 2-(N-morpholino)ethanesulfonic acid (MES)-buffered medium was used for polymer incubation.

2.1.3. Results and discussion

Synthesis of polymers

Poly(dichlorophosphazenes) were synthesised via the room temperature living cationic polymerization of trichlorophosphoranimine. The chlorine atoms of the polymeric precursor were then replaced with the highly water-soluble oligomer Jeffamine M-1000, an amine-capped polyether belonging to the Jeffamine family, which interestingly have been reported to have better cell uptake than PEG.39 The amino acid spacers, glycine and valine, were incorporated between the polyphosphazene backbone and the hydrophilic side chains to give a series of graft poly(organophosphazenes) (Scheme 2.1-2) with varied hydrolytic stabilities. 31P NMR spectra of the polymers 1-5 showed single, broadened peaks in the range from 0 to 4 ppm indicating the absence of chlorine atoms, at least up to NMR detection limits. A large excess of amine and long reaction times were used to ensure complete removal

41 of the chlorine atoms since remaining P-Cl units would be expected to significantly enhance the rate of degradation.31 Since the stable polymer 5 (no amino acid spacer) was shown to be reproducibly stable under neutral conditions, it is presumed that consistent complete replacement of the chlorine atoms (see degradation studies below) was achieved under the conditions applied. The polymeric structures and purity of the polymers were also confirmed by 1H NMR and ATR-FTIR spectroscopy (see supporting information).

Scheme 2.1-2. Structures of hydrophilic graft poly(organophosphazenes) 1-5. N.B. the structures shown for polymers 2 and 4 are simplified, with the macromolecular substitution resulting in a random, mixed substitution pattern.

The living polymerization provides a route to polyphosphazenes with molecular weights that can be controlled by the initial ratio of monomer to initiator, but is not able to prepare polyphosphazenes with large chain lengths (above n = 75) with good reproducibility.32,40 Through the attachment of larger organic components, higher molecular weights are accessible, in this case the substitution with M-1000 side groups, to yield polyphosphazenes with theoretical molecular weights over 100 kg mol-1

(n = 50) (Table 2.1-1). The apparent molecular weights Mn (exp) measured by GPC were estimated using a conventional calibration versus linear polystyrene standards and were a factor of 5-10 lower than the calculated values Mn (theo), due to the highly branched nature of the polymers. However, the relatively narrow molecular weight distributions Mw / Mn in the range of 1.1 to 1.3 clearly demonstrate the living and controlled nature of the polymerization procedure, especially considering that the attached side chains are not monodisperse. Furthermore, the hydrodynamic diameters of the polymers in aqueous environment were measured by DLS and lie in the range of 5.9 to 7.0 nm.

42

Table 2.1-1. Characterisation of polymers 1-5.

a) theo b) exp c) d) Polymer R1 R2 Mn , Mn , Mw/Mn d (nm) (kg mol-1) (kg mol-1) 1 M-100-Gly - 108 20 1.16 6.5 (± 0.2) 2 M-100-Gly M-1000 105 11 1.34 6.8 (± 0.1) 3 M-100-Val - 112 13 1.34 7.0 (± 0.1) 4 M-100-Val M-1000 107 15 1.20 6.7 (± 0.1) 5 M-1000 - 102 14 1.23 5.9 (± 0.1)

a) b) Molar feed ratio of R1/R2 was 1/1, calculated from the initial molar ratio of monomer to initiator (25/1, n = 50), c) apparent value determined by GPC with conventional calibration d) versus linear polystyrene standards, measured by DLS in deionized H2O.

Degradation studies

The hydrolytic degradation and decline of molecular weight of the synthesized polymers was investigated under enhanced degradation conditions in water at pH 2 using gel permeation chromatography. As an example, GPC chromatographs of polymer 3 reveal peak broadening and a shift to later retention volumes compared to the initial polymer before degradation (Figure 2.1-1). After four days, the polymer peak has nearly completely disappeared suggesting a rapid decrease of the initial molecular weight. Furthermore, a second peak arises at 11.6 mL, corresponding to a molecular weight of around 1000 g mol-1 which can be directly assigned to the cleavage of the M- 1000-Val side groups.

Figure 2.1-1. Degradation studies of polymer 3 with GPC under accelerated conditions in -1 acidified H2O (pH 2) over seven days. The arising peak at around 11.6 mL (ca. 1000 g mol ) confirms the hydrolysis and thus release of the side groups.

43 The breakdown of the polymers was further studied with 31P NMR spectroscopy to monitor intermediate polyphosphazene degradation products. To illustrate this, the changes for polymer 3 observed in the 31P NMR spectra during the hydrolysis, also under enhanced degradation conditions at pH 2, are depicted in Figure 2.1-2. As expected, the degradation proceeded rapidly at pH 2 (see later section on pH effect), as indicated by the appearance of several peaks after four days. These additional 31P resonances at -1.1 ppm, -3.8 ppm and -9.8 ppm stem from the formation of -P-OH, - P=O, and products of geminal hydrolysis.28,41 After seven days, the sharp peak at around 0 ppm dominated the spectrum, which is associated with the formation of inorganic phosphate. These results support the proposed degradation mechanism of polyphosphazenes that involves the hydrolysis and thus hydrolytic cleavage of the side groups from the phosphazene backbone, leading to hydroxyphosphazenes and phosphazanes.31

Figure 2.1-2. Degradation studies of polymer 3 with 31P NMR spectroscopy under accelerated degradation conditions in acidified D2O (pH 2) over seven days. Arising signals can be associated with the formation of various intermediates with -P-OH and –P=O moieties, as well as breakdown of the polyphosphazene backbone. The peak for inorganic phosphate can be observed at around 0 ppm.

44

Spacer effect. Various degradation pathways have been proposed, all of which result in hydrolytic chain cleavage yielding low molecular weight fragments up to the formation of phosphates and ammonia.31 For this reason the increase in phosphate formation, a final backbone degradation product, was measured (pH 5, 37°C) for polymers 1, 3 and 5 (Figure 2.1-3). Polymer 5, without an amino acid spacer, degraded at the slowest rate, with only 14 % of the backbone degrading to phosphate after four weeks. However, the incorporation of an amino acid spacer was observed to have a considerable destabilization effect, with polymer 1, incorporating the glycine spacer, degrading almost completely within four weeks, and thus hydrolyzing significantly faster than polymer 5. With the valine spacer (polymer 3), the phosphate formation was retarded in comparison to polymer 1 (glycine spacer) but faster than polymer 5 (no spacer), with half of the polyphosphazene backbone being degraded into phosphate after four weeks. The hydrolytic stabilities thus decreased from polymer 5 > 3 > 1.

Figure 2.1-3. Degradation profiles of polymers 1 (∆), 3 (○) and 5 (□) at pH 5. The increasing amount of phosphate was quantitatively determined by UV-Vis analysis.

The incorporation of amino acid ester units is of particular importance for the synthesis of hydrolytically sensitive polyphosphazenes,21,42 with a proposed hydrolysis of the side group ester linkage leading to the formation of carboxylic acids that promote backbone cleavage and thus enhance the hydrolysis of the polyphosphazene backbone.17,43 However, in this work, the hydrophilic side chains are grafted via amide bonds onto the amino acid spacer. For this reason it would seem reasonable to exclude, as previously proposed,31 hydrolysis of the ester groups on the degradation behaviour of the polymers. As amide degradation is not to be expected, it would

45 appear, therefore, that the adjacent carbonyl groups from the amino acid spacer play a significant role in accelerating the degradation rate of the polyphosphazene backbone, perhaps via a nucleophilic attack from the carbonyl oxygen onto the phosphorus atoms in the polyphosphazene backbone.21 The observation that polymer 3, with a valine spacer, degrades slower than polymer 1, with a glycine spacer, is presumed to be due to shielding of the backbone from hydrolytic attack from the isopropyl groups at the alpha-C-atoms and corresponds well with previous work involving amino acid ester substitutents.22

The hydrolysis process was further monitored by 31P NMR spectroscopy under neutral conditions over nine weeks (Figure 2.1-S5) with the trend corresponding well to that observed for the phosphate testing. Polymer 1 is observed to degrade rapidly, whereas, polymer 2, with 50 % of the glycine-functionalized M-1000 and 50 % of the M- 1000 added, showed a significantly higher aqueous stability. Thus addition of M-1000, with no spacer, has a considerable stabilizing effect on the glycine polymer. However, although with such a mixed macromolecular substitution approach is a simple route to reduce the hydrolytic sensitivity of these polyphosphazenes, there is a distinct lack of control and thus reproducibility associated with the ratio of side-groups attached via a mixed substituent method and even when a step-wise synthesis is carried out, exchange44 at the phosphorus atom is inherently possible. For this reason it was decided to concentrate these studies on polymers 1, 3 and 5, where stoichiometric control can be guaranteed.

pH effect. The influence of the pH-value on the degradation profile was also investigated. The degradation profiles of polymer 1 under acidic and neutral conditions (Figure 2.1-4) revealed the promoting influence of the pH-value on the degradation process, with a slowly increasing formation of phosphate at pH 7 and considerably higher degradation rates observed at pH 5 and 2 for all polymers. Under neutral conditions only 40 % of the polyphosphazene backbone was hydrolyzed within four weeks. In mildly acidic environment, however, a continuous hydrolytic chain cleavage and thus increase of phosphate could be detected, with 90 % of polymer 1 being degraded after four weeks. At pH 2, however, a significantly faster initial degradation rate was observed, with 50 % of the backbone being degraded after six days and complete degradation after four weeks.

46

Figure 2.1-4. Release of inorganic phosphate of polymer 1 under aqueous conditions at pH 2 (∆), pH 5 (○) and pH 7.4 (□), showing the enhanced degradation at reduced pH, as measured by UV-Vis spectroscopy.

Polymer 1 with the glycine spacer degraded completely within four weeks at pH 2, the degradation of the relatively stable polymer 5 was also significantly faster under strong acidic conditions (see supporting information Figure 2.1-S6). Thus the degradation of all studied polymers proceeded substantially faster as the pH-value was decreased, supporting the findings of the degradation studies using UV-Vis spectroscopy to investigate the formation of phosphate. A similar accelerated degradation behaviour at lower pH-values has also been observed for polyphosphazenes containing N-ethylpyrrolidone polyphosphazene substituents,28 as well as for phosphoramidate forming polyphosphoesters.45

These results would appear to confirm the proposed mechanism that a protonation of the polyphosphazene backbone in an acidic environment increases the overall hydrolysis rate of the polymers.15,46-48 Subsequent cleavage of the side chains leads to an exposure of the backbone that is then more accessible to hydrolytic attack.

Cell viability studies

A selection of the polymers underwent preliminary testing for their biocompatibility, in particular for potential use as macromolecular drug carriers. For this purpose, polymer 1 (glycine spacer), polymer 3 (valine spacer) and polymer 5 (no spacer) were tested with HCT116 cells. As shown in Figures 2.1-5A and -S7A, none of the polymers

47 showed a substantial impact on cell viability. Furthermore, the partially degraded polymer (stored for 4 or 8 weeks at 37°C in aqueous solution) did not show any harmful effect on the cell viability. To mimic acidic degradation conditions and thus possible intermediates, additional experiments were performed in MES-buffered medium at pH 6.0 (Figures 2.1-5B and -S7B). Also in these tests no substantial impact on the cell viability was observed compared to controls. Similar results were also observed in A2780 and Hep3B cells (data not shown) confirming the non-toxic nature of these polymers and degradation products within a polymer concentration range relevant for drug delivery applications (0.01 to 10 µM).

Figure 2.1-5. Cell viability of HCT116 cells after incubation with polymer 1. (A) Activity of HCT116 cells was tested by MTT assay after 72 h incubation. Normalized values given are mean values ± S.D. of experiments performed in triplicate. (B) Comparison of standard cell culture conditions (pH 7.4) and mildly acidic milieu (pH 6, MES buffer) on the biocompatibility of polymer 1.

48

2.1.4. Conclusions

A series of novel graft poly(organophosphazenes) were synthesised containing hydrophilic Jeffamine M-1000 side chains that were coupled via an amino acid spacer onto the degradable polyphosphazene backbone. The polymers showed excellent solubility in water and provide a pH-assisted degradation behaviour with a considerably faster degradation at lower pH-values. The hydrolytic stability was tailored by careful choice of the amino acid spacer and could be easily increased by steric shielding of the polymer backbone via the R groups of the alpha-C-atom of the amino acid spacer. The biocompatibility demonstrated by these water soluble polymers, their degradation intermediates and products, in combination with their synthetic structural control and broad spectrum of degradation rates available suggest significant promise as materials for aqueous biomedical applications such as polymer therapeutics.

Acknowledgments

The authors acknowledge financial support of the Austrian Science Fund (FWF), P24659-N28. S. W. acknowledges the Johannes Kepler University Linz for the “JKU Goes Gender dissertation fellowship”. A. I. would like to thank the Basque Government’s Global Training Grant Programme for a scholarship. NMR measurements were carried out at the Austro-Czech RERI-uasb NMR-center which was established with financial support from the European Union through the EFRE INTERREG IV ETC-AT-CZ programme (project M00146, “RERI- uasb”). DLS analysis was carried out at the department of Analytical Chemistry at the University of Vienna (W. Lindner) with the assistance of H. Henke and H. Hinterwirth.

49 2.1.5. References

1. B. D. Ulery, L. S. Nair, C. T. Laurencin, J. Polym. Sci., Part B: Polym. Phys. 2011, 49, 832-864. 2. J. Nicolas, S. Mura, D. Brambilla, N. Mackiewicz, P. Couvreur, Chem. Soc. Rev. 2013, 42, 1147-1235. 3. E. Schacht, J. Vandorpe, S. Dejardin, Y. Lemmouchi, L. Seymour, Biotechnol. Bioeng. 1996, 52, 102-108. 4. R. Haag, F. Kratz, Angew. Chem. Int. Ed. 2006, 45, 1198-1215. 5. R. Duncan, Nat. Rev. Cancer 2006, 6, 688-701. 6. T. Etrych, M. Jelínková, B. Ríhová, K. Ulrbich, K., J. Controlled. Release 2001, 73, 89-102. 7. K. Ulbrich, T. Etrych, P. Chytil, M. Jelínková, B. Ríhová, J. Controlled Release 2003, 87, 33-47. 8. R. Duncan, Nat. Rev. Drug Discovery 2003, 2, 347-360. 9. E. Markovsky, H. Baabur-Cohen, A. Eldar-Boock, L. Omer, G. Tiram, S. Ferber, P. Ofek, D. Polyak, A. Scomparin, R. Satchi-Fainaro, J. Controlled Release 2012, 161, 446-460. 10. R. A. Shenoi, J. K. Narayanannair, J. L. Hamilton, B. F. L., Lai, S. Horte, R. K. Kainthan, J. P. Varghese, K. G. Rajeev, M. Manoharan, J. N. Kizhakkedathu, J. Am. Chem. Soc. 2012, 134, 14945-14957. 11. R. Duncan, M. J. Vicent, Adv. Drug Delivery Rev. 2010, 62, 272-282. 12. D. G. van der Poll, H. M. Kieler-Ferguson, W. C. Floyd, S. J. Guillaudeu, K. Jerger, F. C. Szoka, J. M. Fréchet, Bioconjugate Chem. 2010, 21, 764-773. 13. S. Zhang, A. Li, J. Zou, L. Y. Lin, K. L. Wooley, ACS Macro Lett. 2012, 1, 328- 333. 14. I. Teasdale, O. Brüggemann, Polymers 2013, 5, 161-187. 15. S. Lakshmi, D. S. Katti, C. T. Laurencin, Adv. Drug Delivery Rev. 2003, 55, 467- 482. 16. H. R. Allcock, Soft Matter 2012, 8, 7521-7532. 17. H. R. Allcock, Chemistry and Applications of Polyphosphazenes, 1st edition, John Wiley & Sons, Hoboken, 2003. 18. N. Ren, X.-J. Huang, X. Huang, Y.-C. Qian, C. Wang, Z.-K. Xu, J. Polym. Sci., Part A: Polym. Chem. 2012, 50, 3149-3157.

50

19. I. Teasdale, S. Wilfert, I. Nischang, O. Brüggemann, Polym. Chem. 2011, 2, 828- 834. 20. S. G. Kumbar, S. Bhattacharyya, S. P. Nukavarapu; Y. M. Khan, L. S. Nair, C. T. Laurencin, J. Inorg. Organomet. Polym. Mater. 2006, 16, 365-385. 21. H. R. Allcock, T. J. Fuller, D. P. Mack, K. Matsumara, K. M. Smeltz, Macromolecules 1977, 10, 824-830. 22. H. R. Allcock, S. R. Pucher, A. G. Scopelianos, Macromolecules 1994, 27, 1071- 1075. 23. A. K. Andrianov, A. Marin, Biomacromolecules 2006, 7, 1581-1586. 24. N. L. Morozowich, A. L. Weikel, J. L. Nichol, C. Chen, L. S. Nair, C. T. Laurencin, H. R. Allcock, Macromolecules 2011, 44, 1355-1364. 25. C. T. Laurencin, M. E. Norman, H. M. Elgendy, S. F. El-Amin, H. R. Allcock, S. R. Pucher, A. A. Ambrosio, J. Biomed. Mater. Res. 1993, 27, 963-973. 26. H. R. Allcock, S. R. Pucher, Macromolecules 1991, 24, 23-34. 27. H. R. Allcock, S. Kwon, Macromolecules 1988, 21, 1980-1985. 28. A. K. Andrianov, A. Marin, P. Peterson, Macromolecules 2005, 38, 7972-7976. 29. J. H. L. Crommen, E. H. Schacht, E. H. G. Mense, Biomaterials 1992, 13, 511- 520. 30. J. H. L. Crommen, E. H. Schacht, E. H. G. Mense, Biomaterials 1992, 13, 601- 611. 31. H. R. Allcock, N. L. Morozowich, Polym. Chem. 2012, 3, 578-590. 32. H. R. Allcock, C. A. Crane, C. T. Morrissey, J. M. Nelson, S. D. Reeves, C. H. Honeyman, I. Manners, Macromolecules 1996, 29, 7740-7747. 33. V. Blackstone, A. J. Lough, M. Murray, I. Manners, J. Am. Chem. Soc. 2009, 131, 3658-3667. 34. J. L. Nichol, N. L. Morozowich, H. R. Allcock, Polym. Chem. 2013, 4, 600-606. 35. Z. Tian, Y. Zhang, X. Liu, C. Chen, M. J. Guiltinan, H. R. Allcock, Polym. Chem. 2013, 4, 1826-1835. 36. S. Wilfert, A. Iturmendi, H. Henke, O. Brüggemann, I. Teasdale, Macromol. Symp. in press. 37. B. Wang, E. Rivard, I. Manners, Inorg. Chem. 2002, 41, 1690-1691. 38. J. Paulsdorf, M. Burjanadze, K. Hagelschur, H. Wiemhöfer, Solid State Ionics 2004, 169, 25-33.

51 39. L. Y. Peddada, N. K. Harris, D. I. Devore, C. M. Roth, J. Controlled Release 2009, 140, 134-140. 40. H. Henke, S. Wilfert, A. Iturmendi, O. Brüggemann, I. Teasdale, J. Polym. Sci., Part A: Polym. Chem. 2013, 51, 4467-4473. 41. D. G. Gabler, J. F. Haw, Macromolecules 1991, 24, 4218-4220. 42. A. K. Andrianov, Polyphosphazenes for biomedical applications, 1st edition, John Wiley & Sons, Hoboken, 2009, Chapter 2, pp 25-26. 43. H. R. Allcock, S. R. Pucher, A. G. Scopelianos, Macromolecules 1994, 27, 1-4. 44. S. Zhang, H. Wang, Y. Shen, F. Zhang, K. Seetho, J. Zou, J.-S. A. Taylor, A. P. Dove, K. L. Wooley, Macromolecules 2013, 46, 5141-5149. 45. H. R. Allcock, A. E. Maher, C. M. Ambler, Macromolecules 2003, 36, 5566-5572. 46. D. P. DeCollibus, A. Marin, A. K. Andrianov, Biomacromolecules 2010, 11, 2033- 2038. 47. G. A. Carriedo, F. J. García Alonso, J. L. García Álvarez, A. Presa Soto, M. P. Tarazona, M. T. R. Laguna, G. Marcelo, F. Mendicuti, E. Saiz, Macromolecules 2008, 41, 8483-8490. 48. H. R. Allcock, T. J. Fuller, K. Matsumura, Inorg. Chem. 1982, 21, 515-521.

52

SUPPORTING INFORMATION

Water-soluble, biocompatible polyphosphazenes with controllable and pH- promoted degradation behavior

Sandra Wilfert, Aitziber Iturmendi, Wolfgang Schoefberger, Kushtrim Kryeziu, Petra Heffeter, Walter Berger, Oliver Brüggemann, and Ian Teasdale*

Characterisation of precursors.

1 M-1000-Glycine. Yield 4.84 g (84 %). H NMR (300 MHz, CDCl3): δ 1.07 (m, 8H), 1.37 (s, 9H), 3.30 (s, 3H), 3.57 (s, 87H) ppm.

1 M-1000-Valine. Yield 6.39 g (71 %). H NMR (300 MHz, CDCl3): δ 0.90 (m, 6H), 1.10 (m, 8H), 1.40 (s, 9H), 2.60 (br, 1H), 3.34 (s, 3H), 3.60 (s, 80H), 4.08 (br, 1H) ppm.

1 Figure 2.1-S1. H NMR spectrum of M-1000-Valine in CDCl3.

53 Characterisation of polymers 1 - 5.

Polymer 1. Yield 0.43 g (45 %). FTIR (solid): νmax = 2883 (C-H), 1663 (C=O), 1103 (C- -1 1 O) cm , H NMR (500 MHz, D2O, δ): 0.98 (br, 7H), 3.20 (s, 3H), 3.52 (br, 80H) ppm. 31 -1 P NMR (202 MHz, D2O, δ): -8.04 - 13.23 (very broad) ppm. GPC: Mn = 20 178 g mol , -1 Mw = 23 380 g mol , Mw / Mn = 1.16.

Polymer 2. Yield 0.45 g (32 %). FTIR (solid): νmax = 2883 (C-H), 1660 (C=O), 1108 (C- -1 1 31 O) cm , H NMR (500 MHz, D2O, δ): 0.99 (m, 8H), 3.21 (s, 3H), 3.53 (s, 83H) ppm. P -1 -1 NMR (202 MHz, D2O, δ): 3.41 ppm. GPC: Mn = 11 430 g mol , Mw = 15 314 g mol , Mw

/ Mn = 1.34.

Polymer 3. Yield 0.57 g (38 %). FTIR (solid): νmax = 2883 (C-H), 1653 (C=O), 1109 (C- -1 1 31 O) cm , H NMR (500 MHz, D2O, δ): 0.99 (m, 9H), 3.21 (s, 3H), 3.53 (s, 82H) ppm. P -1 -1 NMR (202 MHz, D2O, δ): 1.05 ppm. GPC: Mn = 13 296 g mol , Mw = 17 849 g mol , Mw

/ Mn = 1.34.

1 Figure 2.1-S2. H NMR spectrum of polymer 3 in D2O. The signals of the valine spacer are overlaid by the large Jeffamine M-1000 signals.

54

Figure 2.1-S3. ATR-FTIR spectrum of polymer 3. Significant bands include the C=O band at 1653 cm-1 stemming from the valine spacer, a large C-H band at 2883 cm-1 and a C-O band at 1109 cm-1 resulting from the M-1000 side chains.

Polymer 4. Yield 0.63 g (66 %). FTIR (solid): νmax 2883 = (C-H), 1663 (C=O), 1107 (C- -1 1 31 O) cm , H NMR (500 MHz, D2O, δ): 0.98 (br, 8H), 3.20 (s, 3H), 3.52 (80H) ppm. P -1 -1 NMR (202 MHz, D2O, δ): 1.53 ppm. GPC: Mn = 15 101 g mol , Mw = 18 143 g mol ,

Mw / Mn = 1.20.

Polymer 5. Yield 1.21 g (67 %). FTIR (solid): νmax = 3305 (N-H), 2884 (C-H), 1109 (C- -1 1 O) cm , H NMR (300 MHz, CDCl3, δ): 1.11 (m, 6H), 3.36 (s, 3H), 3.63 (2, 61H) ppm. 31 -1 -1 P NMR (202 MHz, D2O, δ): 1.02 ppm. GPC: Mn = 14 348 g mol , Mw = 17 709 g mol ,

Mw / Mn = 1.23.

Figure 2.1-S4. Volume size distribution of polymer 3 in deionized H2O with a concentration of 1 mg ml-1 at 25°C. The mean value of the hydrodynamic diameter was calculated from the fitted curve (7.0 nm).

55

31 Figure 2.1-S5. Degradation studies with P NMR spectroscopy in D2O (pH 7) indicating a relatively high stability of polymers 2-5 over 2 months. Changes in the phosphorus peak can be observed, for the glycine polymer 1, therefore degrading significantly faster than polymers 2-5. The rate of degradation of polymer 1 could be decreased by a mixed substitution (polymer 2), i.e. 50 % of the side chains were substituted by the amino acid spacer glycine and 50 % were substituted with M-1000.

56

Figure 2.1-S6. Degradation studies of polymer 1 and 5 with 31P NMR spectroscopy in acidified D2O (pH 2) over 4 weeks. One main phosphoric component resulting in a sharp signal at 1.2 ppm could be observed after four weeks indicating complete degradation of the polyphosphazene backbone into phosphate. The 1H NMR spectrum of the sample after four weeks showed weak signals in the range of 6.9 to 7.2 ppm with a typical splitting, confirming the presence of ammonium, the second degradation product of the polyphosphazene backbone. Degradation studies of polymer 5 confirmed the presence of a considerable amount of phosphate (1.1 ppm) and thus accelerated degradability under harsh acidic conditions.

57

Figure 2.1-S7. Cell viability of HCT116 cells after incubation with (A) polymer 3 (left panel) and polymer 5 (right panel), tested by MTT assay after 72 h incubation. Normalized values given are mean values ± S.D. of experiments performed in triplicate. (B) Comparison of standard cell culture conditions (pH 7.4) and mildly acidic milieu (pH 6.0, MES buffer) of polymer 3 (left panel) and polymer 5 (right panel) confirming the nontoxic nature of the polymers and their degradation products over eight weeks.

58

2.2. Oxidative triggered degradation

In this chapter hydrolytically stable poly(organo)phosphazenes are presented with stimulated degradation pathways selectively in an oxidative environment. Arylboronate pinacol ester group, as a self-immolative caging moiety, is incorporated into the polyphosphazene to control the degradability of the polymer. Upon H2O2 exposure the caging group undergoes self-immolation revealing the unstable poly(glycine)phosphazene which catalyzes the degradation of the polymer. The hydrolytic stability of the poly(organo)phosphazenes in absence of H2O2 evidences the selective degradation of the polymers in the oxidative environment. Furthermore, the polymers without the caging group show no degradation sign upon H2O2 exposure.

My contribution to the paper

In this work, I conducted all the experimental work including synthesis, characterization, data interpretation and degradation studies. I wrote all drafts of the manuscript. The manuscript was conceptualized and corrected together with I. Teasdale and U. Monkowius.

59 Oxidation Responsive Polymers with a

Triggered Degradation via Arylboronate

Self-Immolative Motifs on a

Polyphosphazene Backbone

Aitziber Iturmendi,a Uwe Monkowius,b, Ian Teasdalea*

––––––––– a Institute of Polymer Chemistry, Johannes Kepler University Linz, Altenberger Strasse 69, 4040 Linz, Austria. b Institute of , Johannes Kepler University Linz, Altenberger Strasse 69, 4040 Linz, Austria.

E-mail: [email protected]

–––––––––

ACS Macro Lett., 2017, 6, 150–154

60

Abstract

Oxidation responsive polymers with triggered degradation pathways have been prepared via attachment of self-immolative moieties onto a hydrolytically unstable polyphosphazene backbone. After controlled main-chain growth, post-polymerization functionalization allows the preparation of hydrolytically stable poly(organo)phosphazenes decorated with a phenylboronic ester caging group. In oxidative environments, triggered cleavage of the caging group is followed by self- immolation, exposing the unstable glycine-substituted polyphosphazene which subsequently undergoes to backbone degradation to low-molecular weight molecules. As well as giving mechanistic insights, detailed GPC and 1H and 31P NMR analysis reveal the polymers to be stable in aqueous solutions but show a selective, fast degradation upon exposure to hydrogen peroxide containing solutions. Since the post polymerization functionalization route allows simple access to polymer backbones with a broad range of molecular weights, the approach of using the inorganic backbone as a platform significantly expands the toolbox of polymers capable of stimuli-responsive degradation.

Degradable polymers have ever growing importance for environmental reasons, as well as for use in biomedical applications.1 In recent years there has been a significant progress towards smart responsive polymers that can undergo stimuli controlled degradation, that is remain stable but then undergo spontaneous complete disintegration of the backbone only after activation by a specific stimulus. Such polymers have significant potential in a wide variety of applications, for example in sensoring technologies,2 on demand drug release3 and nanopatterning.4 Spontaneous main-chain disintegration can be achieved by end-to-end backbone depolymerization, so-called “self-immolative polymers.5 Such smart polymers are designed to sequentially disassemble into their respective building blocks in response to a specific triggering event. This property is commonly achieved via incorporation of ortho/para-benzylic amines or alcohols capable of undergoing 1,4-/1,6-eliminations upon deprotection of the amine or alcohol, or alternatively by the design of polymer main chains with urea or carbamate linkages which can undergo intramolecular cyclization.6 Typically self-

61 immolative polymers are prepared via step-growth mechanisms in order to incorporate the aforementioned self-immolating moieties, thus putting constraints in terms of the molecular weight control and architectures available.6 This field has thus also been expanded to chain shattering polymers,3b, 7 in which cleavage of pendant groups along the main chain leads to chain scission into small components. Such a design strategy is potentially open to a wider variety of chemistries and indeed recently ring-opening polymerization8 and olefin metathesis chemistry7b, 9 have been used to prepare poly(caprolactone) and poly(carbonate)s (PCs)10 which undergo a chain-shattering process in response to a variety of stimuli including enzymatic7b photochemical8 and oxidative3b environments.

Scheme 2.2-1. Proposed concept for caged polyphosphazenes with triggered degradation. Upon triggered decaging, the hydrolytically sensitive poly(glycine)phosphazene is produced, which undergoes a rapid self-catalyzed degradation to phosphates and ammonia.

Herein we present an alternative approach towards polymers with stimuli controlled degradation by utilizing the hydrolytically instable inorganic phosphorus nitrogen backbone of polyphosphazenes as a platform.11 Polyphosphazenes are commonly prepared via the highly reactive precursor [NPCl2]n, the facile post- polymerization functionalization of which allows the insertion of a wide range of 12 pendant groups along the polymer backbone. The polymeric precursor [NPCl2]n, can be prepared by ring-opening or living polymerization methods,13 thus allowing high molecular weights and/or controlled Mn with narrow dispersities and potentially a variety of architectures including highly branched structures14 and block copolymers.15 The 11b, 16 precursor [NPCl2]n can be readily substituted with amino acid esters giving rise to poly(amino acid ester)phosphazenes, a family of materials that are of great promise for biomedical applications due to their easily tunable degradation rates.11b, 16a, 17 The backbone degradation mechanisms are well studied16c, 16d, 17b, 18 and known to involve

62

hydrolysis of the backbone phosphorus resulting in cleavage of the amino acid ester16c, 16d with the main chain degradation products shown to be a benign buffered mixture of amino acid (ester), phosphates and ammonium salts.16d It is also well-established that the degradation is acid catalyzed17c, 17d and indeed that the presence of acidic groups in proximity to the backbone phosphorus accelerate hydrolysis rates.17b Indeed, in early studies into poly(amino acid ester)phosphazenes Allcock and coworkers described the inability to isolate the glycine substituted polyphosphazene [NP(NHCH2COOH)2]3 due to its extremely rapid hydrolysis.17b Thus we proposed that through essentially caging poly(glycine)phosphazene via the addition of stimuli responsive protection groups, it should be possible to prepare a stable polymer, which upon removal of the caging moiety produces this hydrolytically unstable glycine-substituted polyphosphazene which would spontaneously and rapidly disintegrate into small molecules (Scheme 2.2- 1), an effect similar to that of a chain-shattering polymer.

Scheme 2.2-2. Synthesis of polymer 1 and polymer 2 with arylboronic acid pinacol ester as a caging groupa.

aReagents and conditions: (i) Glycinate arylboronic acid pinacol ester, THF, rt, 16h; (ii) Excess of R’ (Jeffamine M1000 for polymer 1 and glycine ethyl ester for polymer 2), THF, rt, 16h.

Herein the hydrolytically sensitive [NP(NHCH2COOH)(Rˈ)]n polymer was prepared with arylboronate pincacol ester as a self-immolative caging group (Scheme 2.2-2). The phenylboronic acid ester is known to oxidize to the corresponding phenol in 19 biologically relevant concentrations of the reactive oxygen species (ROS) H2O2. ROS are important in cell signaling but the imbalance of oxidative and reducing species causes oxidative stress which contributes to several diseases such as cancer20

63 cardiovascular disorders21 and Alzheimer’s disease.22 The self-immolative motif was first prepared from the reaction of 4-(hydroxymethyl)benzene boronic acid pinacol ester with Boc-gly-OH (Figure 2.2-S1), followed by selective deprotection of the amine in

CF3CO2H (Figure 2.2-S2). The precursor [NPCl2]n was prepared separately via a recently developed phosphine mediated, living cationic polymerization23 of 24 Cl3PNSiMe3. Glycinate arylboronic acid pinacol ester was then added to partially substitute the backbone. In a second step, an excess of a second amine substituent was added to completely substitute the phosphorus atoms in the backbone. The phosphorus main chain has a quite unique pentavalent nature and thus the second substituent can be used to modulate the chemical and physical properties of the resulting polymer.12a For this work a Jeffamine M-1000 (amino functionalized polyalkylene oxide) was chosen as the secondary substituent, to give the water soluble polymer 1. Jeffamine substituents are known to augment the water solubility, biodegradability and biocompatibility of the polymers.17d The resulting polymer was purified by dialysis and shown by 1H and 31P NMR experiments (Figure 2.2-S3) to have a complete backbone substitution (within the NMR detection range, absence of peaks associated with non and/or partially substituted phosphorus atoms in the 31P NMR spectrum) in a ratio of approximately 50:50, glycinate arylboronic acid pinacol ester to Jeffamine substituents (≈25% wt boronic acid ester according to UV-Vis spectroscopy, Figure 2.2-S4). Thus on average each phosphorus atom bears one boron-containing cleavable unit. The polymer was further characterized by GPC in DMF containing -1 10 mM LiBr (Figure 2.2-S5, Mn,GPC = 68 500 g mol Mw/Mn = 1.5, measured against linear polystyrene standards) and DLS (Figure 2.3-S6, d = 11.23 ± 0.44 nm in H2O).

To investigate the sensitivity of the polymer towards oxidative environments, the polymer was subjected to 10 mM aqueous solution of H2O2 at room temperature. Analysis by size exclusion chromatography (SEC) showed that polymer 1 degrades selectively in the oxidative environments (Figure 2.2-1a) with a decrease in the polymer peak clearly visible. This was accompanied by an increase in the low molecular weight region below approximately 1000 g mol-1 due to the ejection of the Jeffamine oligomers from the hybrid polymer and the formation of low molecular weight compounds. Selective backbone degradation of the polymer was further confirmed by 31P NMR spectroscopy (Figure 2.2-2a) in which a reduction in the broad polymer peak is observed accompanied by the appearance of peaks associated with hydroxyphosphazene (≈-10 ppm) and a sharp peak due to phosphate formation

64

(≈ 0 ppm). Both species are known degradation products for the hydrolysis of the polyphosphazene main-chain.17b, 17d, 18, 25 Meanwhile the polymer could be stored in aqueous solution, that is in the absence of H2O2, at room temperature for the same timeframe (Figure 2.2-1b and -2b), and indeed for several weeks thereafter, before any visible signs of degradation could be detected by 31P NMR spectroscopy (Figure 2.2-S7), thus confirming the selectivity of the degradation to towards the oxidative environment.

Figure 2.2-1. SEC analysis of polymer 1 stored in aqueous solution at room temperature (a) in the presence of 10 mM H2O2 and (b) in the absence of H2O2.

65

31 Figure 2.2-2. (a) P NMR spectroscopy of polymer 1 in D2O in presence of 10 mM H2O2 and

(b) in the absence of H2O2.

After this successful proof-of-principle, a further series of polymers were prepared with glycine ethyl ester co-substituents. Non-water soluble, poly(amino acid ester)phosphazenes belong to the most important polyphosphazenes for biomedical applications.11b Furthermore, the comparatively low of the organic substituent facilitates mechanistic studies of the backbone cleavage mechanism by 1H NMR spectroscopy, for which many relevant peaks are obscured in polymer 1. Polymer 2 (Figure 2.2-3a) was hence prepared with approximately 50 mol% of glycinate arylboronic acid pinacol ester substituent and 50 mol% glycine ethyl ester, as calculated by 1H NMR spectroscopy (Figure 2.2-S8) and ≈ 60% wt by UV-Vis

(Figure 2.2-S4) spectroscopy. Upon exposure to 10 mM acetone solution of H2O2, 1H NMR studies of a sample of polymer 2 (23 mg mL-1) revealed the oxidation of the pinacol ester to be fast, with complete formation of the phenol (still bound to the

66

polymer) in 4 hours (for this particular polymer and under these conditions, see Figure 2.2-3b and -S9).

Figure 2.2-3. (a) Proposed self-immolation mechanism of polymer 2 upon H2O2 exposure and (b) 1H NMR tracking of the self-immolation pathway of polymer 2 in 10 mM acetone solution of

H2O2. Entire 1H NMR spectra are shown in Figure 2.2-S10.

However, the self-immolation proved to be the rate limiting step, with the phenol intermediate remaining stable for some days. Once self-immolation has started, sharp resonance peaks begin to appear, which correspond to low molecular weight degradation products. These observations correlate with the 31P NMR studies of the same sample (Figure 2.2-4b), in which no degradation is observed until a time-frame in which self-immolation has occurred. Thereafter polymer 2 showed a rapid degradation as indicated by the presence of peaks associated with the primary chain degradation products, hydroxphosphazenes and phosphazane, as well as phosphates due to its self-catalyzed degradation (Figure 2.2-4a) in accordance with previous studies into the degradation mechanism of polyaminophosphazenes.17b, 18, 25 Over a longer time period this was seen to degrade fully to phosphates (Figure 2.2-S11a). The kinetics were observed to be slower than for the water soluble polymer 1, which can be

67 explained by the use of an organic solvent for this non-water soluble polymer.26 The nature of the degradation products could further be characterized by ESI-MS, with the detection of glycine ethyl ester and 4-hydroxybenzyl alcohol (Figure 2.2-S12).

31 In the absence of H2O2, no main-chain degradation was observed in the P NMR spectrum over same period of time (Figure 2.2-4c) with a slow hydrolytic degradation occurring thereafter (Figure 2.2-S11b). 1H NMR studies meanwhile showed a partial hydrolysis of the boronic acid ester to cleave the pinacol (Figure 2.2-S13).27 In order to exclude the presence of an oxidative reduction pathway for poly(amino acid ester)phosphazenes, polymer 3 (poly(glycine ethyl ester)phosphazene, Figure 2.2- S14), with no boronic acid ester caging groups, was also exposed to oxidative conditions. Studies of the sample by 31P NMR spectroscopy (Figure 2.2-S15) indicated the stability of poly(glycine ethyl ester)phosphazene even to a higher concentration of

H2O2 (100 mM) thus confirming that any H2O2 triggered degradation effect is exclusively due to the presence of the self-immolative boronate ester moiety.

Figure 2.2-4. (a) Probable -cleavage mechanisms of the hydrolytically sensitive glycine substituted polyphosphazene; 31P NMR spectroscopy of polymer 2 (b) in 10 mM acetone solution of H2O2 and (c) in acetone/water solution without H2O2.

68

In summary a new type of polymers based on a polyphosphazene with phenylboronate moieties along the main chain has been prepared. While such unique boron containing polymers may have many interesting properties,28 the linkage of the boronate group via a self-immolative motif allowed the preparation of polymers stable in ambient conditions but with a stimulus-responsive degradation pathway in oxidative environments. The second substituent on the phosphorus atom was used to introduce water solubilizing groups and amino acid ester substituents to the hybrid polymers.

Self-immolation of the boronate upon exposure to H2O2 exposed the hydrolytically sensitive glycine-substituted phosphazene main chain which subsequently underwent a rapid hydrolytic degradation to small molecules. Although the mutually exclusive responsive nature of the degradation in the different environments was clearly shown, degradation rates (hours/days) were slower than may be desired for some applications. Proton NMR studies showed that, while boron oxidation and phosphazene main chain degradation are both rapid, the rate limiting step is the self-immolation of the phenol to present the free acid. Thus future generations of these polymers will look to vary the type of cage and backbone linkage and further enhance the degradation kinetics in the presence of stimuli, without affecting the inherent stability of the polymer. The approach of using the hydrolytically instable inorganic backbone as a platform significantly expands the toolbox of selective degradable polymers, as the post polymerization functionalization allows the preparation of polymer backbones with a broad range of molecular weights and advanced architectures, before insertion of the self-immolative moieties. It is thus envisaged that these novel polymers could be used as the basis of a range of new responsive materials. In particular the unique multivalency of the phosphorus could be used for the addition of functional moieties, to prepare for example sensory materials, while the known biocompatibility of the degradation products also makes them potentially useful polymers for the preparation of biomedical materials, for example in diagnostics or therapeutics.

69 Acknowledgments

The authors acknowledge financial support of the Austrian Science Fund (FWF), P 27410-N28. I.T. and A.I. also extend their appreciation to Prof. Oliver Brüggemann for his support and generous access to laboratory resources. The NMR experiments were performed at the Upper Austrian - South Bohemian Research Infrastructure Center in Linz, co-financed by the European Union in the context of the project "RERI-uasb", EFRE RU2-EU-124/100-2010 (ETC Austria-Czech Republic 2007-2013, project M00146).

70

2.2.1. References

1. Ulery, B. D.; Nair, L. S.; Laurencin, C. T., Biomedical applications of biodegradable polymers. J. Polym. Sci., Part B: Polym. Phys. 2011, 49 (12), 832-864. 2. Sella, E.; Shabat, D., Dendritic Chain Reaction. J. Am. Chem. Soc. 2009, 131 (29), 9934-9936. 3. a) Zhang, Y.; Yin, Q.; Yin, L.; Ma, L.; Tang, L.; Cheng, J., Chain-Shattering Polymeric Therapeutics with On-Demand Drug-Release Capability. Angew. Chem. 2013, 125 (25), 6563-6567; b) de Gracia Lux, C.; Joshi-Barr, S.; Nguyen, T.; Mahmoud, E.; Schopf, E.; Fomina, N.; Almutairi, A., Biocompatible Polymeric Nanoparticles Degrade and Release Cargo in Response to Biologically Relevant Levels of Hydrogen Peroxide. J. Am. Chem. Soc. 2012, 134 (38), 15758-15764. 4. Seo, W.; Phillips, S. T., Patterned Plastics That Change Physical Structure in Response to Applied Chemical Signals. J. Am. Chem. Soc. 2010, 132 (27), 9234-9235. 5. a) Wang, W.; Alexander, C., Self-Immolative Polymers. Angew. Chem., Int. Ed. 2008, 47 (41), 7804-7806; b) Sagi, A.; Weinstain, R.; Karton, N.; Shabat, D., Self- Immolative Polymers. J. Am. Chem. Soc. 2008, 130 (16), 5434-5435. 6. Peterson, G. I.; Larsen, M. B.; Boydston, A. J., Controlled Depolymerization: Stimuli-Responsive Self-Immolative Polymers. Macromolecules 2012, 45 (18), 7317- 7328. 7. a) Zhang, Y.; Ma, L.; Deng, X.; Cheng, J., Trigger-responsive chain-shattering polymers. Polym. Chem. 2013, 4 (2), 224-228; b) Mutlu, H.; Barner-Kowollik, C., Green chain-shattering polymers based on a self-immolative azobenzene motif. Polym. Chem. 2016, 7 (12), 2272-2279. 8. de Gracia Lux, C.; Almutairi, A., Intramolecular Cyclization for Stimuli-Controlled Depolymerization of Polycaprolactone Particles Leading to Disassembly and Payload Release. ACS Macro Lett. 2013, 2 (5), 432-435. 9. Lv, A.; Li, Zi-Long; Du, Fu-Sheng; Li, Zi-Chen, Synthesis, Functionalization, and Controlled Degradation of High Molecular Weight Polyester from Itaconic Acid via ADMET Polymerization. Macromolecules 2014, 47 (22), 7707-7716. 10. Qiu, F.-Y.; Song, C.-C.; Zhang, M.; Du, F.-S.; Li, Z.-C., Oxidation-Promoted Degradation of Aliphatic Poly(carbonate)s via Sequential 1,6-Elimination and Intramolecular Cyclization. ACS Macro Lett. 2015, 4 (11), 1220-1224.

71 11. a) Allcock, H. R., The expanding field of polyphosphazene high polymers. Dalton Transactions 2016, 45 (5), 1856-1862; b) Allcock, H. R.; Morozowich, N. L., Bioerodible polyphosphazenes and their medical potential. Polym. Chem. 2012, 3 (3), 578-590. 12. a) Tian, Z.; Hess, A.; Fellin, C. R.; Nulwala, H.; Allcock, H. R., Phosphazene High Polymers and Models with Cyclic Aliphatic Side Groups: New Structure–Property Relationships. Macromolecules 2015, 48 (13), 4301-4311; b) Allcock, H. R., Polyphosphazene elastomers, gels, and other soft materials. Soft Matter 2012, 8 (29), 7521-7532. 13. Rothemund, S.; Teasdale, I., Preparation of polyphosphazenes: a tutorial review. Chem. Soc. Rev. 2016, 45 (19), 5200-5215. 14. Henke, H.; Brüggemann, O.; Teasdale, I., Branched Macromolecular Architectures for Degradable, Multifunctional Phosphorus-Based Polymers. Macromol. Rapid Commun. 2016, 10.1002/marc.201600644. 15. Soto, A. P.; Gilroy, J. B.; Winnik, M. A.; Manners, I., Pointed-Oval-Shaped Micelles from Crystalline-Coil Block Copolymers by Crystallization-Driven Living Self- Assembly. Angew. Chem., Int. Ed. 2010, 49 (44), 8220-8223. 16. a) Deng, M.; Kumbar, S. G.; Wan, Y.; Toti, U. S.; Allcock, H. R.; Laurencin, C. T., Polyphosphazene polymers for tissue engineering: an analysis of material synthesis, characterization and applications. Soft Matter 2010, 6 (14), 3119-3132; b) Allcock, H. R.; Pucher, S. R.; Scopelianos, A. G., Poly[(amino acid ester)phosphazenes]: Synthesis, Crystallinity, and Hydrolytic Sensitivity in Solution and the Solid State. Macromolecules 1994, 27 (5), 1071-1075; c) Crommen, J. H. L.; Schacht, E. H.; Mense, E. H. G., Biodegradable Polymers .1. Synthesis of Hydrolysis-Sensitive Poly[(Organo)Phosphazenes]. Biomaterials 1992, 13 (8), 511-520; d) Crommen, J. H. L.; Schacht, E. H.; Mense, E. H. G., Biodegradable Polymers .2. Degradation Characteristics of Hydrolysis-Sensitive Poly[(Organo)Phosphazenes]. Biomaterials 1992, 13 (9), 601-611. 17. a) Teasdale, I.; Brüggemann, O., Polyphosphazenes for medical applications. Smithers RAPRA: Shrewsbury, UK, 2014; b) Allcock, H. R.; Fuller, T. J.; Matsumura, K., Hydrolysis pathways for aminophosphazenes. Inorg. Chem. 1982, 21 (2), 515-521; c) Allcock, H. R.; Fuller, T. J.; Mack, D. P.; Matsumura, K.; Smeltz, K. M., Synthesis of Poly[(Amino Acid Alkyl Ester)Phosphazenes]. Macromolecules 1977, 10 (4), 824-830; d) Wilfert, S.; Iturmendi, A.; Schoefberger, W.; Kryeziu, K.; Heffeter, P.; Berger, W.; Brüggemann, O.; Teasdale, I., Water-soluble, biocompatible polyphosphazenes with

72

controllable and pH-promoted degradation behavior. J. Polym. Sci., Part A: Polym. Chem. 2014, 52 (2), 287-294. 18. Andrianov, A. K.; Marin, A., Degradation of polyaminophosphazenes: Effects of hydrolytic environment and polymer processing. Biomacromolecules 2006, 7 (5), 1581- 1586. 19. a) Broaders, K. E.; Grandhe, S.; Fréchet, J. M. J., A Biocompatible Oxidation- Triggered Carrier Polymer with Potential in Therapeutics. J. Am. Chem. Soc. 2011, 133 (4), 756-758; b) Song, C.-C.; Du, F.-S.; Li, Z.-C., Oxidation-responsive polymers for biomedical applications. Journal of Materials Chemistry B 2014, 2 (22), 3413-3426. 20. Trachootham, D.; Alexandre, J.; Huang, P., Targeting cancer cells by ROS- mediated mechanisms: a radical therapeutic approach? Nat Rev Drug Discov 2009, 8 (7), 579-591. 21. Shah, A. M.; Channon, K. M., Free radicals and redox signalling in cardiovascular disease. Heart 2004, 90 (5), 486-487. 22. Butterfield, D. A., Oxidative Stress in Alzheimer Disease: Synergy Between the Butterfield and Markesbery Laboratories. NeuroMol. Med. 2011, 13 (1), 19-22. 23. Wilfert, S.; Henke, H.; Schoefberger, W.; Brüggemann, O.; Teasdale, I., Chain- End-Functionalized Polyphosphazenes via a One-Pot Phosphine-Mediated Living Polymerization. Macromol. Rapid Commun. 2014, 35 (12), 1135-1141. 24. Wang, B.; Rivard, E.; Manners, I., A new high-yield synthesis of Cl3P=NSiMe3, a monomeric precursor for the controlled preparation of high molecular weight polyphosphazenese. Inorg. Chem. 2002, 41 (7), 1690-1691. 25. Linhardt, A.; König, M.; Schöfberger, W.; Brüggemann, O.; Andrianov, A.; Teasdale, I., Biodegradable Polyphosphazene Based Peptide-Polymer Hybrids. Polymers 2016, 8 (4), 161. 26. Qiu, F.-Y.; Zhang, M.; Du, F.-S.; Li, Z.-C., Oxidation Degradable Aliphatic Polycarbonates with Pendent Phenylboronic Ester. Macromolecules 2017, 50 (1), 23- 34. 27. Hall, D. G., Structure, Properties, and Preparation of Boronic Acid Derivatives. In Boronic Acids, Wiley-VCH Verlag GmbH & Co. KGaA: 2011; pp 1-133. 28. Cheng, F.; Jakle, F., Boron-containing polymers as versatile building blocks for functional nanostructured materials. Polym. Chem. 2011, 2 (10), 2122-2132.

73

74

SUPPORTING INFORMATION

Oxidation responsive polymers with a triggered degradation via arylboronate self- immolative motifs on a polyphosphazene backbone

Aitziber Iturmendi, Uwe Monkowius, and Ian Teasdale*

Materials and methods

Syntheses were carried out under argon atmosphere in glovebox (MBRAUN) or under nitrogen using standard Schlenk line techniques. All glassware was dried in an oven overnight at 120°C prior to use. Solvents were purchased from Merck, VWR and Alfa

Aesar and used without further purification. Triethylamine (Et3N) was dried over molecular sieves and distilled prior to use. The polyetheramine copolymer (PEO-PPO-

NH2), sold under the trade name Jeffamine M-1000 and with a nominal molecular weight of 1000 g mol-1, was donated by Huntsman Performance Products

(Netherlands) and used as received. The monomer trichlorophosphoranimine (Cl3P=N- 1 SiMe3) was synthetized according to literature procedure as described previously. All other chemicals were purchased from Sigma Aldrich, TCI chemicals and Fluorochem and used without further purification.

1 H NMR spectroscopy was recorded on a Bruker 300 MHz spectrometer using D2O, 31 DMSO-d6 or acetone-d6 as an internal reference. P NMR (121 MHz) experiments were carried out using 85% phosphoric acid as an external standard. Gel permeation chromatography (GPC) was measured with a Viscothek GPCmax instrument equipped with a PFG column from PSS (Mainz, Germany) (300 mm x 8 mm, 5 m particle size). DMF containing 10 mM LiBr was used as the mobile phase at a flow rate of 0.75 ml min-1 at 60°C. The molecular weights were estimated using conventional calibration of the refractive index detector versus polystyrene standards from PSS. Electrospray ionization mass spectroscopy (ESI-MS) characterization was recorded on

75 Agilent 1100 series HPLC with LC/MSD mass detector in the negative ion mode. 10 mmol methanol was used as eluent at 0.7 mL min-1. UV-Vis spectra were carried out on a Perkin Elmer Lambda 25 UV/Vis spectrophotometer. A Malvern Zetasizer Nano- ZS instrument (Malvern Instruments, UK) was used for dynamic light scattering (DLS) measurements. A 4 mW standard laser was used at a 633 nm wavelength with the -1 detector angle at 173°. The polymer was dissolved in deionized H2O (1 mg mL ) and filtered through a 0.2 m nylon filter and measured at 25°C.

Synthesis of Boc-gly-arylboronic acid pinacol ester

N-(tert-Butoxycarbonyl)glycine (Boc-Gly-OH) (0.90 g, 5.1 mmol) and 4-

(dimethylamino)pyridine (DMAP) (62.6 mg, 0.5 mmol) were dissolved in 40 mL CH2Cl2. Then, 4-(hydroxymethyl)benzeneboronic acid pinacol ester (1.2 g, 5.1 mmol) dissolved in 15 mL CH2Cl2 was added and stirred for 1 hour. N,N′-Dicyclohexylcarbodiimide

(DCC) (1.06 g, 5.1 mmol) was dissolved in 20 mL CH2Cl2 and transferred to the previous reaction which was cooled to 0°C. After stirring the reaction for 2 days at room temperature, precipitated urea was then filtered off. The filtrate was extracted twice with 10% NH4Cl, twice with 5% NaHCO3, twice with saturated sodium chloride and dried over MgSO4. The solvent was removed under vacuum and further dried to yield boc-gly-arylboronic acid pinacol ester as a white viscous oil. Yield = 1.82 g (91%). 1 H NMR (300 MHz, DMSO-d6, ): 1.29 (s, 12H), 1.38 (s, 9H), 3.74 (d, J = 6.2 Hz, 2H), 5.15 (s, 2H), 7.27 (br,s, 1H), 7.37 (d, J = 8.1 Hz, 2H), 7.66 (d, J = 7.9 Hz, 2H).

Synthesis of Polymer 1

Polymers were synthesized according to literature procedure for the phosphine-initiated polymerization of trichlorophosphoranimine.2 The following procedure describes the procedure used for the synthesis of polymer 1. All polymers were synthesized with the same ratio of monomer to chlorinated phosphine (50/1). Briefly, triphenylphosphine

(12.00 mg, 45.7 mol, 1 eq.) and hexachloroethane (C2Cl6) (11.9 mg, 50.3 mol,

1.1 eq.) dissolved in 1 mL anhydrous CH2Cl2 were stirred for 16 hours at room temperature. After this time Cl3P=N-SiMe3 (0.51 g, 2.3 mmol, 50 eq.), also dissolved in

0.5 mL anhydrous CH2Cl2, was added and stirred for 24 hours. Macromolecular

76

substitution of polydichlorophosphazene was carried out as follows: 0.91 g (2.3 mmol,

1 eq.) of boc-gly-arylboronic acid pinacol ester was deprotected in CF3CO2H/CH2Cl2 (1:3) for 3 hours. The solvent was removed under vacuum and further dried by co- evaporation with toluene and chloroform to obtain the glycinate arylboronic acid pinacol ester (Figure 2.2-S2). In the glove box, it was re-dissolved in 15 mL anhydrous THF and an excess of Et3N was added to neutralize TFA residues. The solution of polydichlorophosphazene (0.51 g, 2.3 mmol, 1 eq.) in anhydrous DCM was transferred to the solution of glycinate arylboronic acid pinacol ester and stirred at room temperature for 16 h. Then an excess of Jeffamine 1k (3.20 g, 3.2 mmol, 1.4 eq) dissolved in 15 mL anhydrous THF and Et3N (0.4 mL, 3.2 mmol, 1.4 eq) mixture was added to the partially substituted polymer. The solution was further stirred at room temperature for 16 h. The suspension was filtered in order to remove the insoluble hydrochloride salt and the solvent was concentrated under vacuum. The polymer was purified by dialysis (12 kDa cut-off) in dry EtOH (over molecular sieves) for 5 days. The solvent was removed under vacuum to yield a white wax. Yield = 0.79 g (26 %). 1 H NMR (300 MHz, acetonitrile-d3, ): 1.07 (br, 9H), 1.28 (br, 3H), 3.29 (s, 4H), 3.55 (s, 84H), 3.78 (br, 2H), 5.12 (br, 2H), 7.34 (br, 2H), 7.77 (br, 2H). 31P NMR (121 MHz, -1 -1 acetonitrile-d3, ): 1.03 ppm. GPC: Mn = 68 500 g mol , Mw = 105 600 g mol , Mw/Mn = 1.5. DLS: d = 11.23 ± 0.44 nm.

Synthesis of Polymer 2

Following the same procedure as with polymer 1, 0.96 g (2.5 mmol, 1 eq.) of boc-gly- arylboronic acid pinacol ester was deprotected and added to a solution of polydichlorophosphazene (0.55 g, 2.5 mmol, 1 eq.) in anhydrous THF and Et3N. The reaction mixture was stirred at room temperature for 16 h in the glove box. Meanwhile an excess of glycine ethyl ester hydrochloride (0.62 g, 4.4 mmol, 1.8 eq.) in anhydrous + - THF (20 mL) and Et3N (1 mL) was stirred at room temperature for 24 h. Et3NH Cl was filtered and transferred to the partially substituted polymer. The solution was allowed to react for further 16 h at room temperature. The suspension was filtered and the solvent was concentrated under vacuum. The polymer was purified by three precipitations into

H2O from THF and seven precipitations into n-heptane from DCM. Yield = 0.24 g 1 (22%). H NMR (300 MHz, acetone-d6, ): 1.12 – 1.29 (br, 15H), 4.02 (br, 6H), 5.08 (br, 31 2H), 7.30 (br, 2H), 7.71 (br, 2H). P NMR (121 MHz, acetone-d6, ): 2.03 ppm.

77 Synthesis of Polymer 3

Et3N (2.00 mL, 14.3 mmol) was added to a suspension of glycine ethyl ester hydrochloride (1.10 g, 7.9 mmol) in THF (70 mL) at room temperature. The reaction mixture was stirred for 24 h to form glycine ethyl ester and the precipitated was removed by filtration. Polydichlorophosphazene (0.55 g, 2.5 mmol, 1 eq.) in anhydrous DCM was added to the filtrate and the reaction mixture was stirred at room temperature for 24 h in the glove box. The suspension was filtered and the solvent was concentrated under vacuum. The polymer was purified by three precipitations into H2O from THF and four precipitations into n-heptane from DCM. Yield = 0.22 g (35%). 1 H NMR (300 MHz, acetone-d6, ): 1.27 (t, 3H), 3.89 (br, 2H), 4.14-4.21 (q, 2H). 31 P NMR (121 MHz, acetone-d6, ): 3.01 ppm.

Degradation studies of polymer 1

Polymer 1 (15.05 mg) was dissolved in 968 L of D2O and the solution was transferred 1 31 to a NMR tube. H2O2 was added to make 10 mmol solution and H and P NMR spectra were recorded at various time points. As a control, 15.25 mg of polymer 1 was dissolved in 1013 L of D2O, transferred to a NMR tube and measured at the same time points. GPC studies were done in a similar way. Polymer 1 (15.04 mg) was dissolved in deionized water (968 L) and H2O2 was added to make 10 mmol solution. 253 L aliquots were taken at various time points and water was evaporated. 1.5 mL of DMF (containing 10 mM LiBr) was added and filtered before injection. The same procedure was carried out in absence of H2O2 (15.01 mg of polymer 1 dissolved in 1013 L of deionized water).

Degradation studies of polymer 2

Polymer 2 (15.07 mg) was dissolved in acetone-d6 (600 L) and the solution was 1 31 transferred to a NMR tube. After addition of H2O2 to make 10 mmol solution, H and P NMR spectra were recorded at different time points. As a control, 16.04 mg of polymer

2 was dissolved in acetone-d6 and transferred to a NMR tube. Then 50 L of D2O was added and 1H and 31P NMR were measured at same time points.

78

Degradation studies of polymer 3

Polymer 3 (15.91 mg) was dissolved in acetone-d6 (600 L) and the solution was 1 transferred to a NMR tube. After addition of H2O2 (to make 100 mM solution), H and 31P NMR spectra were recorded at various time points.

References

1. Wang, B.; Rivard, E.; Manners, I., A new high-yield synthesis of Cl3P=NSiMe3, a monomeric precursor for the controlled preparation of high molecular weight polyphosphazenese. Inorganic Chemistry 2002, 41 (7), 1690-1691. 2. Wilfert, S.; Henke, H.; Schoefberger, W.; Brüggemann, O.; Teasdale, I., Chain- End-Functionalized Polyphosphazenes via a One-Pot Phosphine-Mediated Living Polymerization. Macromolecular Rapid Communications 2014, 35 (12), 1135-1141. 3. Linhardt, A.; König, M.; Schöfberger, W.; Brüggemann, O.; Andrianov, A.; Teasdale, I., Biodegradable Polyphosphazene Based Peptide-Polymer Hybrids. Polymers 2016, 8 (4), 161.

79

Figure 2.2-S1. 1H NMR spectrum of boc-gly-arylboronic acid pinacol ester in DMSO (*).

Figure 2.2-S2. 1H NMR spectrum of glycinate arylboronic acid pinacol ester in DMSO (*).

80

Figure 2.2-S3. (a) 1H NMR and (b) 31P NMR spectrum of polymer 1 in acetonitrile- d3 (*). Due to the larger number of protons from the Jeffamine compared to the boronic acid ester group, there is some uncertainity in the integration. Partial hydrolysis of the labile pinacol ester can be assumed because of the purification method.

81

Figure 2.2-S4. UV-Vis spectra of 4-(hydroxymethyl)benzeneboronic acid pinacol ester (0.36 mg mL-1, line a), polymer 1 (0.65 mg mL-1, line b) and polymer 2 (0.41 mg mL-1, line c) in EtOH.

Figure 2.2-S5. GPC chromatograph of polymer 1 with DMF as eluent containing 10 mM LiBr.

82

Figure 2.2-S6. Molecular size distribution by intensity (a) and volume (b) of polymer 1 -1 in deionized H2O with a concentration of 1 mg mL at 25°C. Molecular size distribution by intensity shows a bimodal distribution due to some agglomeration of the amphiphilic polymers.3

83

31 Figure 2.2-S7. P NMR spectrum of polymer 1 in D2O (absence of H2O2). The polymer remains stable for several weeks in aqueous solution.

84

1 31 Figure 2.2-S8. (a) H NMR and (b) P NMR spectrum of polymer 2 in acetone-d6 (*).

85

1 Figure 2.2-S9. H NMR spectrum of polymer 2 in acetone-d6 (*) before H2O2 addition and after 4h in presence of H2O2. The oxidation of arylboronic acid pinacol ester to phenol is complete in 4 h releasing the boronic pinacol ester which is hydrolyzed to boric acid and pinacol.

86

Figure 2.2-S10. Entire 1H NMR spectra of Figure 2.2-4b. 1H NMR tracking of the self- immolation pathway of polymer 2 in 10 mM acetone solution of H2O2.

87

Figure 2.2-S11. 31P NMR spectrum of polymer 2 (a) in 10 mM acetone solution of

H2O2. After 30 days the polymer fully degrades to phosphates; (b) in acetone/water solution without H2O2. In the same time-frame (30 days) no significant sign of phosphate is observed.

88

Figure 2.2-S12. ESI-MS of the H2O2-induced degradation products of Polymer 2.

Figure 2.2-S13. 1H NMR tracking of polymer 2 in acetone/water solution. In absence of

H2O2 arylboronic acid pinacol ester cannot be oxidized, but it can be partially hydrolyzed releasing the pinacol.

89

1 31 Figure 2.2-S14. (a) H NMR and (b) P NMR spectrum of polymer 3 in acetone-d6 (*).

90

31 Figure 2.2-S15. P NMR spectrum of polymer 3 in 100 mM acetone solution of H2O2.

Even at higher concentration of H2O2 no relevant degradation is observed after 21 days.

91

92

2.3. Photochemically triggered degradation

The aim of this study was to synthesize degradable poly(organo)phosphazenes that degrade upon visible light exposure, making them attractive candidates for many biomedical applications due to its spatial and temporal control over the degradation. First of all, the photochemical properties of the photocleavable group, coumarin-caged glycine, are studied and presented. Then, two different poly(organo)phsosphazenes with different properties and degradation rates are synthesized. On one hand, the coumarin derivative group with a water-solubilizing group (M-1000) is included in the polyphosphazene backbone to combine the photodegradabilty and water solubility of the polymer. On the other hand, the same coumarin derivative group is combined with the amino acid glycine ethyl ester to synthesize a photo- and biodegradable polyphosphazene. In both cases stimulated degradation studies with different rates are presented.

My contribution to the paper

I carried out the majority of the experimental work including synthesis, characterization and degradation studies. Small molecule studies were conducted by D. Maderegger and photochemical characterization by S. Theis. I wrote all drafts of the manuscript. The manuscript was conceptualized and corrected together with I. Teasdale and U. Monkowius.

93 Coumarin-Caged Polymers with a Visible-

Light Driven on-Demand Degradation

Aitziber Iturmendi,a Sabrina Theis,b Dominik Maderegger,a Uwe Monkowius,c and Ian Teasdalea*

––––––––– a Institute of Polymer Chemistry, Johannes Kepler University Linz, Altenberger Strasse 69, 4040 Linz, Austria. b Institute of Inorganic Chemistry, Johannes Kepler University Linz, Altenberger Strasse 69, 4040 Linz, Austria. c Linz School of Education, Johannes Kepler University, Linz, Altenberger Strasse 69, 4040 Linz, Austria.

E-mail: [email protected]

–––––––––

Submitted

94

Abstract

Polymers are described which upon photochemical activation with visible light undergo rapid degradation to small molecules. Through functionalization of a polyphosphazene backbone with light sensitive coumarin moieties, polymers which are stable in the dark could be prepared. Upon irradiation, cleavage of the coumarin moieties expose carboxylic acid moieties along the polymer backbone. The subsequent macromolecular photoacid is found to catalyze the rapid hydrolytic degradation of the polyphosphazene backbone. Water soluble and non-water soluble polymers are reported, which due to their sensitivity towards light in the visible region could be significant as photocleavable materials in biological applications

KEYWORDS: on-demand degradation, visible-light, polyphosphazenes, photodegradable, photocages.

2.3.1. Introduction

There is considerable demand for stimuli-responsive polymers that can undergo chemical or physical changes in response to a triggers such as pH,[1] biomolecules,[2] temperature, oxidation[3] and light.[4] Among these stimuli, light is notably attractive due to its ability to exert spatial and temporal control over the desired response. Consequently, several photo-responsive polymers have been prepared that have great promise in diverse applications such as drug delivery, porous membranes and film patterning.[4b, 5]

Macromolecules that undergo photochemical backbone degradation are also of significant interest[6]. Most reported photo-cleavable systems require ultraviolet (UV) light to achieve cleavage, which can be a drawback especially in biomedical environments due to low penetration and biocompatibility, hence a shift to longer wavelengths is required.[7] Coumarin and its analogues are widely employed in various fields such as medicine, polymer science, cosmetics and biology.[8] Coumarin derivatives can be readily synthesized and functionalized to red-shift the light absorption[8], cage the desired molecule as protecting groups[9] and/or to be

95 incorporated on the polymer main chain as photoresponsive units.[10] It is also reported that coumarin photocages can be cleaved via two-photon processes with near-infra-red irradiation[11]. In addition, functionalized aliphatic polycarbonates,[12] photodegradable hydrogels,[13] and drug delivery systems[14] based on coumarin derivatives have been developed. Furthermore, the ability of some coumarin polyesters to undergo chain scission or crosslinking upon certain wavelength irradiation has been demonstrated.[10, 15]

Herein we present a novel approach to photodegradable polymers based on coumarin functionalized polyphosphazenes. Polyphosphazenes are unique due to their hydrolytically instable backbone which can be tailored through the incorporation of different substituents.[16] Furthermore, the polyphosphazene degradation pathway is known to be acid-catalysed,[17] a process which can be intramolecular when acidic substituents are present on the polymer backbone.[18] Hence we proposed that through functionalization of polyphosphazenes with a coumarin-caged amino acid, the sensitivity of the polymers to hydrolysis would be accelerated upon irradiation by effectively producing a macromolecular photoacid which could subsequently catalyse its own degradation.

2.3.2. Results and Discussion

Coumarin-caged gylcine

First 7-(N,N-diethylamino)-4-(hydroxymethyl)coumarin 1 (Figure 2.3-S1) was prepared according to literature procedures.[11] After reaction with N-(tert- butoxycarbonyl)glycine (Boc-gly-OH) and deprotection this gave the coumarin-caged glycine 3 (Figures 2.3-S2 and -S3).[19] Coumarin 1 was chosen as caging group as it is expected to be photochemical active in the visible region due to the electron-donating diethylamino group.[20] The kinetics of the elementary photochemical reaction of 3 has been investigated recently by flash-photolysis[19] but not under steady illumination. Hence we first investigated the photo-cleavage reaction of compound 3 upon irradiation

(100W, ≥ 395 nm) in MeOH/H2O (4:1) solution. UV-Vis spectroscopy (Figure 2.3-S4) showed a decrease in absorption intensity and shifted to lower wavelength (from 382 to 377 nm) after irradiation, indicative for the decaging of the glycine moiety (Figure 2.3- 1a). Interestingly, an increase in intensity and slight hypsochromic shift (from 478 to

96

473 nm) was observed in the emission spectra upon irradiation (Figure 2.3-1b). This ceased after approximately 25 min irradiation, indicating completion of the photosolvolysis reaction. 1H NMR spectroscopy (Figure 2.3-1c) also confirmed the nature of the photocleavage reaction, albeit with a lower reaction rate due to the higher concentration required (10-7 mol L-1 vs 0.082 mol L-1).

Figure 2.3-1. a) Photocleavage reaction of the coumarin-caged glycine 3. b) Change of the emission spectra of 3 in MeOH/H2O (4:1) solution upon irradiation with HBO lamp (100W, cut- off filter at 395 nm, conc. 10-7 mol L-1).(— before irradiation; ··· 5 min irradiation; ▪ 10 min irradiation; •15 min irradiation; -- 25 min irradiation;). c) 1H NMR analysis of the photocleavage -1 1 of 3 in MeOD/D2O (4:1) solution (conc. 0.082 mol L ). Entire H NMR spectra are shown in Figure 2.3-S5.

Coumarin based photo-degradable polyphosphazenes

Our approach was to extend this photodecaging phenomenon to ‘cage’ a hydrolytically instable glycine-substituted phosphazene moiety through incorporation of the coumarin-caged glycine onto a polyphosphazene backbone. Polydichlorophosphazene was first prepared via phosphine-mediated polymerisation of trichlorophosphoranimine[21] (Scheme 2.3-1, see Supporting Information for detailed procedure). One equivalent of the coumarin-cage 3 was then added in order to substitute the majority of chlorine atoms. For related amino acid esters it is known that monosubstitution at the phosphorus atom is strongly favoured due to the higher [22] reactivity of -Cl2PN- in comparison to -ClRPN-. Therefore, we expect a distribution of

97 the coumarin groups along the backbone. Thereafter the remaining chlorine atoms were substituted with either Jeffamine M-1000, a polyether monoamine, or glycine ethyl ester to obtain polymer P1 and polymer P2, respectively.

Scheme 2.3-1: Synthesis of Polymer P1 and Polymer P2 bearing the coumarin derivative as a caging group of glycine. Reagents and conditions: (i) CH2Cl2, n = 50 and 18 h for polymer P1, and bulk polymerization (solvent free), n = 400 and 96 h for polymer P2; (ii) 1 equivalent of coumarin caged glycine 3, THF, Et3N, rt, 24 h for polymer P1 and 48 h for polymer P2; (iii) excess of R (Jeffamine M-1000 and 24 h for polymer P1, and glycine ethyl ester and 48 h for polymer P2), THF, Et3N, rt.

Polymer P1

Polymer P1 was synthetized using Jeffamine M-1000 as second substituent to give water soluble polymer. M-1000 was also chosen to ensure the hybrid polymer remained hydrolytically stable in the timeframe of the photochemical reactions[17] and hence to be able to induce a photochemical degradation without significant interference from undesired hydrolytic degradation. The polymer was purified by dialysis in the dark and characterized by 1H and 31P NMR spectroscopy (Figure 2.3-S6), size exclusion chromatography (SEC) in DMF containing 10 mM LiBr (Figure 2.3-S7, Mn,GPC = -1 149 000 g mol and Mw/Mn= 1.03, measured using multidetector calibration) and dynamic light scattering (DLS) (Figure 2.3-S8, d = 12.25 nm ± 0.38 nm in H2O at 1 mg mL-1). According to 1H and 31P NMR spectroscopy complete backbone substitution in a ratio of nearly 34:66 (coumarin derivative : Jeffamine) could be

98

observed with no additional peaks in the 31P NMR corresponding to partially substituted phosphorous atoms. UV-Vis spectroscopy (Figure 2.3-S9) showed the loading to be approximately 14 wt%, which corresponds roughly to 35:65 ratio, coumarin derivative to M-1000 substituents.

The polymer was irradiated (100W, ≥ 395 nm) in aqueous solution to investigate its photochemical properties. The photo-reaction was followed by UV-Vis (Figure 2.3-2a) and fluorescence spectra (Figure 2.3-2b). In UV-Vis spectra one prominent long wavelength absorption band with a maximum at 386 nm could be observed, which corresponds to the absorption of coumarin. The maximum of the emission band was approximately 500 nm. Aqueous solutions of the polymer showed no changes during 24 h in the dark (Figure 2.3-S10). However under irradiation with visible light, the sample changed color (from yellow to orange) and significant changes in the UV-Vis and fluorescence spectra could be observed. The intensity of the absorption band at 386 nm gradually decreased and shifted to 381 nm. Four isosbestic points at 457, 359, 266, and 236 nm are visible indicative for a clean photo-reaction. The fluorescence spectra showed both an increase of the emission band and a hypsochromic shift from 508 to 499 nm. This behavior is comparable to the small-molecule studies on compound 3.

Since the photo-cleavage was deemed to be completed after 90 min irradiation in -1 H2O, a further sample was irradiated at the same concentration (0.16 mg mL ) and analysed by SEC (Figure 2.3-2e). Meanwhile, an identical sample held for 24 h (Figure 2.3-2f) and even 7 days (Figure 2.3-S11a) without irradiation showed no signs of degradation. This confirmed the destabilizing effect of glycine after de-caging rendering the polymer hydrolytically unstable.

As the degradation was significant but incomplete, the polymer was kept in the dark for a further period (Figure 2.3-S11b). However, total degradation of the polymer could not be achieved even keeping it for 7 days in aqueous solution after irradiation. Since Jeffamine M-1000 (34:66 ratio) was used in excess, parts of the polymer might consist of blocks constituted only by the M-1000 substituent which are relatively stable against hydrolytic degradation.[17]

99

Figure 2.3-2. Changes of a) UV-Vis absorption spectra and b) emission spectra of polymer P1 -1 in H2O (0.167 mg mL ) upon irradiation with a HBO lamp (cut-off filter at 395 nm). Progress of c) absorbance intensity and d) emission intensity (each measured at the band maxima from spectra in a) and b)) versus time. SEC analysis of polymer P1 e) after irradiation in H2O -1 -1 (0.163 mg mL ) and f) in the dark in H2O (0.173 mg mL ) for 24 h (— before irradiation; -- 90 min irradiation; ͦ 24 h dark).

It is known that some polymers containing coumarin derivatives can undergo 2+2 photodimerization of the 3,4-double bond forming a cyclobutane ring.[23] For the small- molecule caged glycine 3, we could not identify any photodimerisation product. Furthermore, the predominance of the photocleavage in the polymers was also clearly observed (Figure 2.3-S12).

100

Figure 2.3-3a shows the proposed degradation mechanism induced by the photoreaction, in which cleavage of the coumarin cage exposes the free acid groups of the glycine moieties along the polyphosphazene backbone. Thus, once the protecting group was cleaved the polymer underwent chain-shattering type of degradation.[24] This process was further investigated by 31P NMR spectroscopy in which significant depletion of the polyphosphazene backbone peak was observed upon polymer -1 irradiation (90 min irradiation at 0.164 mg mL concentration in H2O and stored for 3 days in the dark, Figure 2.3-S13a). After a longer period, further degradation accompanied by the appearance of the sharp peak associated to phosphate formation could also be observed. The polymer kept in the dark even for 7 days did not show any signs of degradation (Figure 2.3-S13b). Furthermore, 1H NMR spectroscopy revealed the successful cleavage of the caging group upon irradiation with the appearance of sharp peaks corresponding to the coumarin group while no changes were observed in the dark sample (Figure 2.3-S14).

Figure 2.3-3. a) Reaction scheme of the photochemical cleavage of the coumarin and of the rapid self-catalyzed degradation of the hydrolytically sensitive glycine substituted polyphosphazene. b) Progress of emission intensity of polymer P2 versus time upon irradiation in acetonitrile:H2O (4:1) mixture. c) SEC analysis of polymer P2 before and after irradiation in -1 acetonitrile:H2O (4:1) mixture (0.16 mg mL ) (— before irradiation; -- 45 min irradiation).

101 Polymer P2

Amino acid ester substituted polyphosphazenes are known as biodegradable polymers.[16a, 25] Therefore, polymer P2 was synthesized using glycine ethyl ester as second substituent. After purification by dialysis in the dark, polymer P2 was successfully obtained as it could be confirmed by 1H and 31P NMR spectroscopy -1 (Figure 2.3-S15), and SEC (Mn,GPC = 249 000 g mol and Mw/Mn= 1.8, measured using multidetector calibration in DMAc containing 57.6 mM LiBr and 0.1 M acetic acid). A ratio of approximately 33:66 coumarin derivative to glycine ethyl ester could be confirmed by 1H NMR with complete backbone substitution observed in the 31P NMR spectroscopy.

Irradiation was performed in acetonitrile:H2O solution (4:1) (Figure 2.3-S16) upon which the emission spectra showed a predominant increase especially in the first irradiation minutes with an approximate end after 45 minutes (Figure 2.3-3b). As the photocleavage was seemed to have ceased after 45 min, a second sample was irradiated in the same solvent mixture (0.16 mg ml-1) for 45 min and analysed by SEC (Figure 2.3-3c). Significant degradation could be observed after irradiation. Compared to polymer P1, the degradation was more pronounced, presumably due to the more labile substituent chosen. It is known that ethyl glycinate has a pKa value of 7.75 at [26] [27] 25°C, while M-1000 has a pKa value of 9.75. Meanwhile, in the absence of light only slight hydrolytic degradation over 3 days could be observed (Figure 2.3-S17).

2.3.3. Conclusions

The design and synthesis of polyphosphazenes with a light-driven, on-demand degradation mechanism has been described. A coumarin-based photocage, which is sensitive to visible light, was used to protect the carboxylic acid group of a glycine moiety on a polyphosphazene backbone. Upon irradiation, the carboxylic acid was exposed which in turn catalysed the polymer degradation. In the dark, the polymers were hydrolytically stable in the investigated timeframe, i.e., the degradation could be initiated upon exposure to visible light. These results prove the potential to use visible light to switch the hydrolytic stability polyphosphazenes and hence trigger backbone degradation. Whilst complete degradation was observed for P2, an incomplete degradation was observed for P1, indicating the importance of the nature of the secondary substituents on the overall polymer stability. Hence, future work will aim to

102

optimize the balance between the nature and loading of the photocage and further substituents with the aim of maintaining stability in the dark, whilst maintaining a complete on-demand degradation. The general proof-of-principle could also be extended to other photocages to prepare polymers which respond to irradiation with yet longer wavelengths in the red region which would have a considerable impact in biological applications.

Acknowledgments

We thank Wolfgang Gnong and Petra Gründlinger for their assistance. Prof. Oliver Brüggemann and Prof. Günther Knör are kindly acknowledged for generous access to laboratory facilities. Thanks to Melanie Kleindienst for contribution of artistic work. The NMR experiments were performed in part at the Upper Austrian-South Bohemian Research Infrastructure Center in Linz, co-financed by the European Union in the context of the project "RERI-uasb", EFRE RU2-EU-124/100-2010 (ETC Austria-Czech Republic 2007-2013, project M00146). The authors acknowledge financial support of the Austrian Science Fund (FWF), P 27410-N28.

103 2.3.4. References

1. S. Binauld, M. H. Stenzel, Chem. Commun. 2013, 49, 2082. 2. R. J. Amir, S. Zhong, D. J. Pochan, C. J. Hawker, J Am Chem Soc 2009, 131, 13949. 3. a) C.-C. Song, F.-S. Du, Z.-C. Li, Journal of Materials Chemistry B 2014, 2, 3413; b) A. Iturmendi, U. Monkowius, I. Teasdale, ACS Macro Lett. 2017, 6, 150. 4. a) F. D. Jochum, P. Theato, Chem Soc Rev 2013, 42, 7468; b) O. Bertrand, J.-F. Gohy, Polym. Chem. 2017, 8, 52. 5. a) P. Xiao, J. Zhang, J. Zhao, M. H. Stenzel, Prog. Polym. Sci. 2017, 74, 1; b) G. Pasparakis, T. Manouras, P. Argitis, M. Vamvakaki, Macromol. Rapid Commun. 2012, 33, 183. 6. a) F. Moldenhauer, R. Kakuchi, P. Theato, ACS Macro Lett. 2016, 5, 10; b) L. Li, A. Lv, X.-X. Deng, F.-S. Du, Z.-C. Li, Chem. Commun. 2013, 49, 8549; c) A. Nazemi, T. B. Schon, E. R. Gillies, Org. Lett. 2013, 15, 1830. 7. a) M. J. Hansen, W. A. Velema, M. M. Lerch, W. Szymanski, B. L. Feringa, Chem Soc Rev 2015, 44, 3358; b) T. Slanina, P. Shrestha, E. Palao, D. Kand, J. A. Peterson, A. S. Dutton, N. Rubinstein, R. Weinstain, A. H. Winter, P. Klán, J. Am. Chem. Soc. 2017, 139, 15168; c) W. Sun, R. Thiramanas, L. D. Slep, X. Zeng, V. Mailänder, S. Wu, Chemistry – A European Journal 2017, 23, 10832; d) X. Zeng, X. Zhou, S. Wu, Macromol. Rapid Commun., 0, 1800034. 8. S. R. Trenor, A. R. Shultz, B. J. Love, T. E. Long, Chem. Rev. 2004, 104, 3059. 9. P. Klán, T. Šolomek, C. G. Bochet, A. Blanc, R. Givens, M. Rubina, V. Popik, A. Kostikov, J. Wirz, Chem. Rev. 2012, 113, 119. 10. E. A. Chamsaz, S. Sun, M. V. Maddipatla, A. Joy, Photochem Photobiol Sci 2014, 13, 412. 11. J. Babin, M. Pelletier, M. Lepage, J.-F. Allard, D. Morris, Y. Zhao, Angew. Chem. Int. Ed. 2009, 48, 3329. 12. J. P. Chesterman, F. Chen, A. J. Brissenden, B. G. Amsden, Polym. Chem. 2017, 8, 7515. 13. M. A. Azagarsamy, D. D. McKinnon, D. L. Alge, K. S. Anseth, ACS Macro Lett. 2014, 3, 515. 14. a) C. Sinkel, A. Greiner, S. Agarwal, Macromol. Chem. Phys. 2010, 211, 1857; b) Q. Lin, C. Bao, Y. Yang, Q. Liang, D. Zhang, S. Cheng, L. Zhu, Adv. Mater. 2013, 25, 1981.

104

15. a) M. V. S. N. Maddipatla, D. Wehrung, C. Tang, W. Fan, M. O. Oyewumi, T. Miyoshi, A. Joy, Macromolecules 2013, 46, 5133; b) J. Lee, M. V. S. N. Maddipatla, A. Joy, B. D. Vogt, Macromolecules 2014, 47, 2891. 16. a) H. R. Allcock, N. L. Morozowich, Polym. Chem. 2012, 3, 578; b) S. Rothemund, I. Teasdale, Chem. Soc. Rev. 2016, 45, 5200. 17. S. Wilfert, A. Iturmendi, W. Schoefberger, K. Kryeziu, P. Heffeter, W. Berger, O. Brüggemann, I. Teasdale, J. Polym. Sci., Part A: Polym. Chem. 2014, 52, 287. 18. D. P. DeCollibus, A. Marin, A. K. Andrianov, Biomacromolecules 2010, 11, 2033. 19. V. R. Shembekar, Y. Chen, B. K. Carpenter, G. P. Hess, Biochemistry 2007, 46, 5479. 20. A. Gandioso, S. Contreras, I. Melnyk, J. Oliva, S. Nonell, D. Velasco, J. Garcia- Amoros, V. Marchan, J Org Chem 2017, 82, 5398. 21. S. Wilfert, H. Henke, W. Schoefberger, O. Brüggemann, I. Teasdale, Macromol. Rapid Commun. 2014, 35, 1135. 22. Z. Tian, Y. Zhang, X. Liu, C. Chen, M. J. Guiltinan, H. R. Allcock, Polym. Chem. 2013, 4, 1826. 23. a) Y. Chen, J.-L. Geh, Polymer 1996, 37, 4481; b) J. He, L. Tremblay, S. Lacelle, Y. Zhao, Soft Matter 2011, 7, 2380; c) W. Fan, X. Tong, Q. Yan, S. Fu, Y. Zhao, Chem. Commun. 2014, 50, 13492. 24. a) H. Mutlu, C. Barner-Kowollik, Polym. Chem. 2016, 7, 2272; b) Y. Zhang, Q. Yin, L. Yin, L. Ma, L. Tang, J. Cheng, Angew. Chem. 2013, 125, 6563. 25. S. Rothemund, T. B. Aigner, A. Iturmendi, M. Rigau, B. Husár, F. Hildner, E. Oberbauer, M. Prambauer, G. Olawale, R. Forstner, R. Liska, K. R. Schröder, O. Brüggemann, I. Teasdale, Macromolecular Bioscience 2015, 15, 351. 26. W. A. Connor, M. M. Jones, D. L. Tuleen, Inorg. Chem. 1965, 4, 1129. 27. Huntsman Library. Epoxy formulations using jeffamine polyetheramines. www.huntsman.com/performance-products, accessed: March 2018

105

106

SUPPORTING INFORMATION

Coumarin-caged polymers with a visible- light driven on-demand degradation

Aitziber Iturmendi, Sabrina Theis, Dominik Madereger, Uwe Monkowius, and Ian Teasdale*

Materials and methods

All synthetic procedures were carried out either in glovebox (MBRAUN) under argon atmosphere or under nitrogen using standard Schlenk line techniques. All glassware was dried in an oven overnight at 120°C prior to use. Solvents were purchased from

Merck, VWR and Alfa Aesar and used without further purification. Triethylamine (Et3N) was distilled and dried over molecular sieves prior to use. The polyetheramine -1 copolymer (PEO-PPO-NH2) with a nominal molecular weight of 1000 g mol and ethylene oxide / propylene oxide ratio of 19/3, was donated by Huntsman Performance Products (Netherlands) and used as received (sold under the trade name Jeffamine M- 1000). All other chemicals were purchased from Sigma Aldrich, TCI chemicals and Fluorochem and used without further purification.

1 H NMR spectroscopy was recorded on a Bruker 300 MHz spectrometer using D2O, 31 MeOD or acetonitrile-d3 as an internal reference. P NMR (121 MHz) experiments were carried out in the same spectrometer using 85% phosphoric acid as an external standard. Gel permeation chromatography (GPC) measurements were performed on a Viscotek GPCmax and HT-GPC instruments equipped with a PFG column (300 mm x 8 mm, 5 m particle size) and three GRAM columns (300 mm x 8 mm, 10 m particle size) respectively, from PSS (Mainz, Germany). DMF with 10 mM LiBr was used as eluent at a flow rate of 0.75 ml min-1 at 60°C for the PFG column. For the GRAM system, DMAc with 57.6 mM LiBr and 0.1M acetic acid was used as eluent at a flow rate of 1 ml min-1 at 70°C. Multidetector calibration (refractive index detector, viscometer and light scattering detector) was used to estimate the molecular weights of the polymers. UV-Vis spectra were carried out on a Cary 300 Bio and Perkin Elmer

107 Lambda UV/Vis spectrophotometers. Fluorescence measurements were recorded on a Jobin Yvon Fluorolog-3 Steady State spectrofluorometer. The samples were irradiated with a HBO lamp with 100 W and a cut off filter of 395 nm (GG 395 nm from SCHOTT). A Malvern Zetasizer Nano-ZS instrument (Malvern Instruments, UK) was used for dynamic light scattering (DLS) measurements. A 4 mW standard laser was used at a 633 nm wavelength with the detector angle at 173°. The polymer was dissolved in -1 deionized H2O (1 mg mL ) and filtered through a 0.2 m nylon filter and measured at 25°C.

Synthesis of 7-(diethylamino)-4-(hydroxymethyl)-coumarin (1)

7-(N,N-diethylamino)-4-(hydroxymethyl)coumarin (1) was synthetized in a similar procedure according to literature.[1] 4-Methyl-7-diethylaminocoumarin (4.66 g,

20.1 mmol) was stirred in hot (120 mL) with SeO2 (3.35 g, 30.2 mmol) for 24 h under reflux. The insoluble black SeO2 was removed by filtration and the solvent was almost totally removed under vacuum. The residue was dissolved with 105 mL of a dry

THF/EtOH mixture (1 : 0.5) and NaBH4 (762 mg, 20.1 mmol) was added slowly. The solution was stirred overnight under reflux and then, it was carefully hydrolysed with 1 N HCl (40 mL). Solvents were almost totally removed under vacuum; the residue was diluted with CH2Cl2, washed three times with brine and dried over MgSO4. The solvent was removed under vacuum and the product was purified by dry column vacuum chromatopraphy (DCVC) (CH2Cl2 / acetone. 50 mL fractions with 5% increase of 1 acetone). Rf = 0.45 ( / EtOAc 1:2). Yield = 0.62 g (12 %). H NMR

(300 MHz, CDCl3, ): 7.31-7.34 (d, J= 8.8 Hz, 1H, Ar H), 6.55-6.59 (dd, J = 9.0, 2.6 Hz, 1H, Ar H), 6.51-6.52 (d, J = 2.6 Hz, 1H, Ar H), 6.27 (t, J = 1.3 Hz, 1H, Ar H), 4.83-4.85

(d, J = 5.7 Hz, 2H, CH2), 3.38-3.45 (q, J = 7.2 Hz, 4H, CH2), 1.19-1.23 (t, J = 7.1 Hz,

6H, CH3).

108

Synthesis of boc-gly-7-(diethylamino)coumarin (2)

Reaction was made in the dark. N-(tert-Butoxycarbonyl)glycine (Boc-Gly-OH) (0.43 g, 2.5 mmol) and 4-(dimethylamino)pyridine (DMAP) (30 mg, 0.25 mmol) were dissolved in anhydrous CH2Cl2 (20 mL). Then, 1 (0.61 g, 2.47 mmol) dissolved in 10 mL anhydrous CH2Cl2 was added and stirred for 30 min. N,N′-Dicyclohexylcarbodiimide

(DCC) (0.51 g, 2.5 mmol) dissolved in 10 mL anhydrous CH2Cl2 was transferred to the previous reaction which was cooled to 0°C. The reaction was stirred for 2 days at room temperature. Precipitated urea was then filtered off and the solvent was removed under vacuum. The product was purified by dry column vacuum chromatopraphy (DCVC)

(CH2Cl2 / acetone. 50 mL fractions with 5% increase of acetone). Rf = 0.79 1 (cyclohexane / EtOAc 1:2). Yield = 0.79 g (79 %). H NMR (300 MHz, CDCl3, ): 7.26- 7.29 (d, J = 9.0 Hz, 1H, Ar H), 6.56-6.60 (dd, J = 9.0, 2.6 Hz, 1H, Ar H), 6.51 (d, J =

2.6 Hz, 1H, Ar H), 6.11 (s, 1H, Ar H), 5.28 (s, 2H, CH2), 5.07 (br,s, 1H, NH), 4.03-4.05

(d, J = 5.9 Hz, 2H, CH2), 3.38-3.45 (q, J = 7.2 Hz, 4H, CH2), 1.46 (s, 9H, CH3), 1.19-

1.23 (t, J = 7.2 Hz, 6H, CH3).

Synthesis of glycinate-7-(diethylamino)coumarin (3)

3 g of 2 was deprotected in 20 mL CF3CO2H /CH2Cl2 (1:3) overnight in the dark and then washed with 5% NaHCO3 and dried over MgSO4. The solvent was removed under vacuum and further dried by co-evaporation with toluene and chloroform to obtain the

109 glycinate-7-(diethylamino)coumarin. Yield = 1.39g (62 %) . 1H NMR (300 MHz,

MeOD:D2O (4:1), ): 7.65-7.68 (d, J = 9.1 Hz, 1H, Ar H), 7.04-7.08 (dd, J = 9.0, 2.5 Hz, 1H, Ar H), 6.94-6.95 (d, J = 2.5 Hz, 1H, Ar H), 6.31 (s, 1H, Ar H), 5.55 (d, J = 1.1 Hz,

2H, CH2), 4.07 (s, 2H, CH2), 3.52-3.59 (q, J = 7.1 Hz, 4H, CH2), 1.16-1.21 (t, J =

7.2 Hz, 6H, CH3).

Synthesis of Cl3P=N-SiMe3

The monomer trichlorophosphoranimine (Cl3P=N-SiMe3) was synthetized according to [2] literature procedure as described previously. Briefly, LiN(SiMe3)2 (25 g, 149 mmol) was dissolved in 500 mL anhydrous diethylether at 0°C. Then, PCl3 (13.05 mL, 149 mmol) was added dropwise at 0 °C for 30 min and stirred at room temperature for

1 hour. After this time, the solution was cooled down again to 0 °C and SO2Cl2 (12.06 mL, 149 mmol) was added dropwise for 30 min. The solution was filtered after stirring it for 1 hour at 0 °C and the solvent was removed under vacuum. The product was purified by vacuum distillation at 8 mbar and 40-50 °C to yield a colorless liquid. Trichlorophosphoranimine was stored under inert argon atmosphere at -37 °C. Yield:

1 31 20.7 g (62%); H NMR (300 MHz, CDCl3, ): 0.18 (d, 9H) ppm; P NMR (121 MHz,

CDCl3, -54.6 ppm

Synthesis of Polymer P1

Polydichlorophosphazene was synthesized according to a similar literature procedure for the phosphine-initiated polymerization of trichlorophosphoranimine.[3] Dichlorotriphenyl-phosphorane (10.0 mg, 30.0 mol, 1 eq.) dissolved in 0.5 mL anhydrous CH2Cl2 was mixed with Cl3P=N-SiMe3 (0.34 g, 1.5 mmol, 50 eq.) and stirred for 18 hours to get [NPCl2]~50. In the dark, 3 (0.46 g, 1.5 mmol, 1 eq) was re-dissolved in 20 mL anhydrous THF and an excess of Et3N (0.5 mL) was added to neutralize TFA residues. The solution of polydichlorophosphazene (0.34 g, 1.5 mmol, 1 eq.) was transferred to the previous solution and stirred at room temperature for 24 h. Then an excess of Jeffamine 1k (2.10 g, 2.10 mmol, 1.4 eq) dissolved in 15 mL anhydrous THF and Et3N (0.3 mL, 2.1 mmol) was added to the partially substituted polymer. The solution was further stirred at room temperature for 24 h. The suspension was filtered in order to remove the insoluble salt and the solvent was concentrated under vacuum.

110

The polymer was purified by dialysis (12 kDa cut-off) in water for 4 h and EtOH (dried over molecular sieves) for 4 days. The solvent was removed under vacuum to yield a red wax which was kept in the dark. Yield = 1.22 g (61 %). 1H NMR (300 MHz, acetonitrile-d3, ): 7.19 (br, 1H, Ar H), 6.39 (br, 2H, Ar H), 5.92 (br, 1H, Ar H), 5.15 (br, 31 2H, CH2), 3.55 (s, 107H, CH2), 3.29 (s, 6H, OCH3), 1.08 (br, 19H, CH3). P NMR (121 -1 MHz, acetonitrile-d3, ): 0.23 and 10.93 ppm. GPC: Mn = 149 000 g mol , Mw = 153 600 -1 g mol , Mw/Mn = 1.03.

Synthesis of Polymer P2

Polydichlorophosphazene was synthesized according to similar literature procedure for the phosphine-initiated polymerization of trichlorophosphoranimine[3] but solvent free.

Dichlorotriphenylphosphorane (2.50 mg, 7.50 mol, 1 eq.) was mixed with Cl3P=N-

SiMe3 (0.71 g, 3.0 mmol, 400 eq.) and stirred for 96 hour until a viscous solution was obtained and a single peak in the 31 PNM at -18.2 ppm was observed. In the dark, 3 (0.91 g, 3.0 mmol, 1 eq) was re-dissolved in 35 mL anhydrous THF and an excess of

Et3N (0.8 mL) was added to neutralize TFA residues. The solution of polydichlorophosphazene (0.71 g, 3.0 mmol, 1 eq.) was transferred to the previous solution and stirred at room temperature for 48 h. Meantime, an excess of glycine ethyl ester hydrochloride (0.60 g, 4.4 mmol, 1.45 eq) in anhydrous THF (20 mL) and Et3N + - (1.2 mL) was refluxed for 4 h. Et3NH Cl was filtered and transferred to the partially substituted polymer. The solution was further stirred at room temperature for 48 h. The suspension was filtered in order to remove the insoluble salt and the solvent was concentrated under vacuum. The polymer was purified by dialysis (3.5 kDa cut-off) in EtOH (dried over molecular sieves) 1 day and in acetonitrile 2 days. The solvent was removed under vacuum to yield a red film which was kept in the dark. Yield = 0.52 g 1 (39 %). H NMR (300 MHz, CDCl3, ): 7.29 (br, 1H, Ar H), 6.33-6.45 (br, 2H, Ar H),

6.06 (br, 1H, Ar H), 5.21 (br, 2H, CH2), 3.73-4.05 (br, 10H, CH2), 3.33 (br, 4H, CH2), 31 1.78 (br, 2H, NH), 1.14 (br, 12H, CH3). P NMR (121 MHz, CDCl3, ): 0.65 ppm. GPC: -1 -1 Mn = 249 000 g mol , Mw = 442 000 g mol , Mw/Mn = 1.8.

111 References

[1] J. Babin, M. Pelletier, M. Lepage, J.-F. Allard, D. Morris, Y. Zhao, Angew. Chem. Int. Ed. 2009, 48, 3329-3332. [2] B. Wang, E. Rivard, I. Manners, Inorg. Chem. 2002, 41, 1690-1691. [3] S. Wilfert, H. Henke, W. Schoefberger, O. Brüggemann, I. Teasdale, Macromol. Rapid Commun. 2014, 35, 1135-1141.

112

Figure 2.3-S1. 1H NMR spectrum of 7-(diethylamino)-4-(hydroxymethyl)-coumarin (1) in CDCl3 (*).

1 Figure 2.3-S2. H NMR spectrum of boc-gly-7-(diethylamino)coumarin (2) in CDCl3 (*).

113

Figure 2.3-S3. 1H NMR spectrum of glycinate-7-(diethylamino)coumarin (3) in

MeOD:D2O (4:1) (*).

Figure 2.3-S4. a) Change of the absorption spectra of 3 in MeOH/H2O (4:1) solution upon irradiation with HBO lamp (100W, cut-off filter at 395 nm, conc. 10-7 mol L-1). b) Enlarged figure between 300 and 500 nm.

114

Figure 2.3-S5. Complete 1H NMR spectra depicted in Figure 2.3-1c. 1H NMR tracking of the photocleavage of 3 in MeOD/D2O (4:1) solution.

115

Figure 2.3-S6. a) 1H NMR and b) 31P NMR spectrum of polymer P1 in acetonitrile- d3 (*).

Figure 2.3-S7. SEC chromatogram of polymer P1 in DMF containing 10 mM LiBr.

116

Figure 2.3-S8. Dynamic Light Scattering (DLS) of polymer P1 in deionized H2O with a concentration of 1 mg mL-1 at 25°C by a) intensity and b) volume distribution. From the volume distribution no predomination of agglomerates can be observed.

Figure 2.3-S9. a) Calibration curve of 1 in methanol. b) Absorbance spectra of polymer P1 (0.08 mg mL-1, 3 replicates) in methanol. The amount of coumarin bound to the polymer was measured from the absorbance at 380 nm.

Figure 2.3-S10. a) UV-Vis absorption spectra and b) emission spectra of the polymer -1 P1 in H2O (0.163 mg ml ) kept in the dark.

117

Figure 2.3-S11. SEC chromatogram of polymer P1 stored in aqueous solution at room temperature a) without irradiation and b) after irradiation.

Figure 2.3-S12. FTIR spectra of polymer P1 before irradiation (—), after 90 min irradiation (▪), and after 90 min without irradiation (--).If the dimerization would dominate, the carbonyl peak (1718 cm-1) would expect to shift due to the appearance of the new non-conjugated carbonyl.

118

31 Figure 2.3-S13. P NMR spectrum of polymer P1 in acetonitrile-d3 a) upon irradiation and b) in the dark in Milli-Q water (0.164 mg mL-1) at room temperature.

119

1 Figure 2.3-S14. H NMR spectrum of polymer P1 measured in acetonitrile-d3 (*) a) after keeping it for 3 days in the dark in H2O and b) upon 90 min irradiation and stored also for 3 days in the dark at room temperature (0.164 mg mL-1).

120

1 31 Figure 2.3-S15. a) H NMR and b) P NMR spectrum of polymer P2 in CDCl3 (*).

Figure 2.3-S16. Change of the emission spectra of polymer P2 in acetonitrile:H2O mixture (4:1) (0.166 mg mL-1) upon irradiation with HBO lamp (cut-off filter at 395 nm).

121

Figure 2.3-S17. SEC chromatogram of polymer P2 without irradiation in -1 acetonitrile:H2O (4:1) mixture (0.166 mg mL ) (— t = 0; ○ 1 day dark; ► 3 days dark).

122

3. Degradable cross-linked scaffolds

The focus of this chapter was to synthesize degradable cross-linked scaffolds with porous structures to enable cell growth in tissue engineering applications. The incorporation of allyl functionalities onto the glycine substituents facilitates the cross- linking reaction with a trifunctional thiol via thiol-ene chemistry. Furthermore, the properties and degradability of the polymer can be easily tailored by addition of different thiol and vinyl groups. Pore sizes in the range of 100-200 m are readily prepared via a salt leaching technique. Due to the aforementioned results, the non- cytotoxic nature, as well as the cell adhesion and proliferation of adipose-derived stem cells onto the polyphosphazene scaffolds confirm the suitability of this material to be implemented in tissue engineering applications.

My contribution to the paper

I conducted the experimental work together with S. Rothemund and T. Aigner including synthesis, characterization, data interpretation and degradation studies of the scaffolds.

123 Degradable Glycine-Based Photo-

Polymerizable Polyphosphazenes for Use as

Scaffolds for Tissue Regeneration

Sandra Rothemund,a Tamara B. Aigner,a,b Aitziber Iturmendi,a,b Maria Rigau,c Branislav Husár,e Florian Hildner,c Eleni Oberbauer,c Martina Prambauer,a,b Gbenga Olawale,d Reinhard Forstnerb, Robert Liskae, Klaus R. Schröder,d Oliver Brüggemann,a and Ian Teasdalea*

––––––––– a Institute of Polymer Chemistry, Johannes Kepler University Linz, Welser Strasse 42, 4060 Leonding, Austria. b Transfercenter für Kunststofftechnik (TCKT) GmbH, Franz-Fritsch-Strasse 11, A-4600 Wels, Austria. c Red Cross Blood Transfusion Service of Upper Austria, Ludwig Boltzmann Institute for Experimental and Clinical Traumatology, Austrian Cluster for Tissue Regeneration, Krankenhausstraße 7, A-4017 Linz, Austria d BioMed-zet Life Science GmbH, Industriezeile 36, A-4020 Linz, Austria. e Institute of Applied Synthetic Chemistry, Vienna University of Technology, Getreidemarkt 9/163, A-1060 Vienna, Austria.

E-mail: [email protected]

–––––––––

Macromol. Biosci. 2015, 15, 351–363

124

Abstract

Photo-polymerizable scaffolds with controlled hydrolytic degradability are designed and prepared via short chain poly(organo)phosphazene building blocks bearing (photo)crosslinkable glycine allylester moieties. The synthesized polyphosphazene was combined with a trifunctional thiol and divinylester in various ratios, followed by thiol- ene photo-polymerization to obtain porous matrices for toxicity and cell proliferation studies. The photo-polymerization was monitored by photo-rheology coupled with FTIR spectroscopy. The resulting chemical polymeric structures were thoroughly characterized with solid-state NMR, FTIR spectroscopy and elemental analysis, and showed uniform incorporation of the inorganic polyphosphazene component. Degradation studies under aqueous conditions showed an increasing rate of degradation in correlation with the polyphosphazene content. Preliminary cell studies show the non-cytotoxic nature of the polymers and their degradation products, as well as the cell adhesion and proliferation of adipose derived stem cells to investigate their potential application in osteochondral tissue engineering. The modular synthetic approach using thiol-ene chemistry allows facile modification of the polymer properties and this synthetic flexibility, in combination with the biocompatibility and easily controllable degradation rates, makes these promising materials which with advanced photo-polymerization techniques could be used to prepare advanced future materials for tissue engineering applications.

KEYWORDS: Bioerosion, polyphosphazenes, photo-polymerization, tissue engineering, degradable polymers.

3.1. Introduction

Bioerodible polymers are of major interest for the development of advanced materials for biomedical applications such as implants, delivery systems for controlled drug release or tissue engineering scaffolds.1-3 For polymers to be used successfully as tissue engineering matrices, a number of essential biological and physical characteristics are demanded of the material used including biocompatibility, specific mechanical properties, tunable degradation profiles and non-toxic degradation products, as well as defined morphologies which can promote tissue growth.4-6 Both

125 natural and synthetic hydrolytically degradable polymers have been extensively studied for use in tissue engineering and regeneration, resulting in several approved biomaterials for clinical use.7

Poly(organo)phosphazenes provide interesting inherent features for tissue engineering applications8,9 due to their highly diverse properties, depending on the side group substituents, most importantly their tunable mechanical properties10 and variable rate of degradation of the inorganic backbone.11,12 The degradation products are reported to be a near-neutral pH-buffered mixture of phosphates and ammonia,11,13 in stark contrast to the acidic products of many other degradable polymers widely used in biomedicine, for example, poly(α-esters), that may evoke undesired side reactions such as tissue irritation, inflammatory and allergenic reactions.14 The rate of hydrolytic degradation can be readily tuned, for example the incorporation of various amino acid esters has been proven to favour hydrolytic degradability of polyphosphazenes11,15. Furthermore, it has been shown that blends of polyphosphazenes with polyesters can efficiently buffer the acidic degradation products and help control degradation rates. To this end a number of materials have been formulated and extensively tested, including dipeptide substituted polyphosphazenes blended with poly(lactic acid–glycolic acid) (PLAGA)13 and the fabrication of matrices suitable for load-bearing bone regeneration,16 as well as PLA block copolymers with polyphosphazenes17 and matrices in combination with polycaprolactone (PCL) for soft skeletal tissue regeneration.18

The multifunctionality of polyphosphazenes, with two readily substituted groups per repeat unit can be used for co-substitution with drugs and/or active agents which are released upon bioerosion of the polymer, for example, with vitamins19 and antioxidants.19 The backbone can also be readily substituted with reactive moieties, for subsequent cross-linking, for example incorporating the antioxidant ferulic acid as a co- substituent could be cross-linked via a [2+2] cycloaddition upon UV-irradiation,19 whereas methacrylate substituted variants undergo subsequent cross-linking by free radical polymerization.20 Thermosensitive polyphosphazenes can also be easily prepared by specific co-substitution of the backbone,21,22 for example thermosensitive polyphosphazene-GRGDS conjugates that form hydrogels at body temperature,23 and through the combination with chemical cross-linking thermosensitive injectable hydrogels can be prepared with suitable mechanical properties for potential use as tissue engineering scaffolds.24

126

The preparation of three-dimensional porous structures based on polyphosphazenes are commonly achieved from high molecular weight polymers processed via salt leaching,25 electro-spinning to obtain mats,26 by using sintering techniques to produce polyphosphazene microsphere scaffolds27 or in-situ microsphere self-assembly.28 Thiol-ene chemistry represents a highly efficient and robust method for polymer functionalization29 and cross-linked porous materials30 and since the reaction can be photo-initiated, cross-linked materials can be readily prepared by photochemical methods.31,32 In this work, we present a novel route to prepare porous polyphosphazenes in which a UV-initiated photo-polymerization is combined with a porogen to prepare glycine-based porous scaffolds. The structure, morphology and degradation behaviour of the cross-linked polymers were thoroughly investigated, indicating the promising potential of these scaffolds for tissue engineering applications.

For biocompatibility studies a selection of the polymers were tested using human cells bearing high potential in the field of tissue engineering and regenerative medicine: Adipose derived stem cells (ASC) can be isolated from liposuction material and propagated by in vitro cultivation on plastic surfaces. This results in the accumulation of cells possessing the ability to give rise to at least adipocytic, osteoblastic and chondrocytic lineages.33-36 Furthermore, there is a growing body of evidence that these cells can generate a variety of other cell types including neuronal/glial-like cells,37-40 cardiomyocytes,41 endothelial cells,42-44 hepatocyte-like cells,45,46 various epithelial cell types,47-49 keratinocyte-like cells50 and dental bud structures.51

3.2. Experimental

Materials and methods

1H NMR and 13C NMR spectroscopy were performed on a Bruker 300 MHz spectrometer at 300 MHz and 75 MHz, respectively, and shifts were referenced to residual solvent peaks. 31P NMR spectra were obtained at a resonance frequency of 121 MHz with proton decoupling using 85% phosphoric acid as an external standard. Solid-state 1H-31P and 1H-13C CP-MAS (cross polarization-magic angle spinning) NMR spectra were measured on a Bruker DRX 500 spectrometer at 202 MHz and 125 MHz, respectively. Approximately 25 mg of the sample were packed into a 3.2 mm MAS

127 rotor. The spectra were recorded using a magic angle spinning rate of 10 kHz. A cross polarization time of 1 ms and 2 ms for 1H-31P CP-MAS and 1H-13C CP-MAS, respectively, were employed, using an effective acquisition time of 27.8 ms and a recycling time of 5 s. FTIR spectroscopy was performed on a Perkin Elmer Spectrum 100 FTIR spectrometer equipped with an ATR accessory. Elemental analyses were performed on an Eurovector EA 3000 and were conducted by the Mikroanalytisches Labor of the University of Vienna as a commercial service. Scanning electron microscopy (SEM) was used to examine the morphology of the scaffolds. Samples were coated with gold for 30 seconds using the Sputter Coater 108 (Cressington, Watford, UK). The samples were then subjected to SEM analysis using FEI scanning electron microscopy (Phenom-WorldBV, Eindhoven, The Netherlands) in secondary electron mode using an accelerating voltage of 5 kV and low vacuum setting. The X-ray CT scans were performed at a laboratory system Nanotom 180 NF (GE Phoenix|x-ray, Wunstorf, Germany). Since a matrix detector (Hamamatsu 2300 × 2300 pixels) was used the system works in cone beam geometry. The specimens have to be projected completely onto the detector in horizontal direction. This leads to a possible scan volume that depends on the chosen voxel edge length. For the given specimens a voxel edge length of 2 µm was chosen which allows for a detailed characterization of the foam structures under investigation. Scan parameters: U = 50 kV, I = 170 µA, T = 1000 ms, scantime 173 min, voxel size = 2 µm, target = molybdenum, mode = 0, average= 5, skip= 1, ZD = 250. Data analysis was performed using the software package MAVI 1.4 (Fraunhofer ITWM Kaiserslautern, Germany). Mechanical compression testing was carried out on a Zwick/Roell Z0.5 universal testing machine equipped with a 500N pressure cell in related to ISO 604 standards with a cross beam speed of 1mm/min. Although the sample dimensions did not fit the ISO standard, the settings and calculation were chosen for comparison between samples.

Syntheses were carried out under an inert argon atmosphere in a glovebox

(MBRAUN) or using standard Schlenk line techniques. PCl5 was purified by sublimation and stored under argon. NEt3 was dried over molecular sieves and distilled prior to use. Photo-chemical reactions were carried out in glass vials in a Rayonet Chamber Reactor with a UV lamp from Camag centered at 254 nm (≈ 300 nm glass cut-off). Adipic acid divinyl ester (VE) was purchased from TCI Europe. Polyethylene glycol (PEG-200) has a nominal molecular weight of 200 g mol-1 and was purchased from Sigma-Aldrich. All other chemicals and solvents were obtained from Sigma-Aldrich and

128

used without further purification. Trichlorophosphoranimine was synthesized and purified as reported previously.15

Allyl 2-(tert-butoxycarbonylamino)acetate was prepared according to a literature 52 procedure. Briefly, Boc-Gly-OH (3.00 g, 17.14 mmol, 1 eq.) and K2CO3 (2.37 g, 17.14 mmol, 1 eq.) were mixed in 60 mL DMF and allyl bromide (2.07 g, 17.14 mol, 1.41 mL, 1 eq.) was added at 0°C. The reaction was allowed to warm to room temperature and stirred overnight followed by the removal of DMF at reduced pressure. The residue was dissolved in ethyl acetate, washed with H2O and brine and dried over MgSO4. Allyl 2- (tert-butoxycarbonylamino)-acetate was obtained as a yellowish viscous product. Yield 1 3.31 g (90 %). H NMR (CDCl3): δ = 1.45 (s, 9H), 3.94 (d, 2H), 4.65 (d, 2H), 5.04 (br, 1H), 5.24-5.36 (m, 2H), 5.85-5.98 (m, 1H) ppm.

Polymer synthesis

Polydichlorophosphazene was synthesized via the living polymerization of 53 trichlorophosphoranimine. In the glovebox, 24.5 mg PCl5 (0.12 mmol, 1 eq.) and 0.66 g Cl3P=N-SiMe3 (2.94 mmol, 25 eq.) were dissolved in 10 mL anhydrous CH2Cl2 and stirred for 16 hours at room temperature. The solvent was removed under reduced pressure and the obtained polydichlorophosphazene was used without further 31 purification. Yield quantitative. P NMR (CDCl3): δ = −18.16 ppm.

Macromolecular substitution of polydichlorophosphazene was carried out as follows. 1.52 g 2-(tert-butoxycarbonylamino)acetate (7.06 mmol, 2.4 eq.) was deprotected in TFA/CH2Cl2 (1:3) for 6 hours. The solvents were removed carefully under vacuum to yield allyl glycinate. Allyl glycinate was redissolved in anhydrous THF and an excess of NEt3 was added to neutralize TFA residues. Polydichlorophosphazene (0.66 g, 2.94 mmol, 1 eq.) dissolved in anhydrous THF was then transferred to the solution of allyl glycinate. The reaction was stirred for 24 hours at room temperature. Precipitated salt was removed by filtration and the reaction mixture was concentrated under vacuum. The polymer was purified by precipitation from THF into chilled diethyl ether. The polymer was dissolved in ethyl acetate and further washed with H2O, brine and dried over MgSO4. The solvent was removed under vacuum and the product further dried under high vacuum to yield polymer 1 as a 1 yellowish highly viscous product. Yield 0.66 g (80 %). H NMR (CDCl3): δ = 3.75 (br, 31 2H), 4.55 (br, 2H), 5.19-5.31 (br, m, 2H), 5.84-5.93 (br, m, 1H) ppm. P NMR (CDCl3):

129 13 δ = 1.25 ppm. C NMR (CDCl3): δ = 42.8 (NH−CH2), 65.5 (OCH2), 118.4 (=CH2), 132.2

(−CH=), 172.5 (C=O) ppm. FTIR (solid): νmax = 3341 (N-H), 2938 (C-H), 1737 (C=O), −1 1650 (C=C), 1188 (P=N) cm . Mn ~ 13 KDa, (calculated from M:I ratio).

Thiol-ene photo-polymerization

In a glass vial, polymer 1 (90.0 mg, 0.33 mmol, 1 eq.) and 1 mg 2,2-dimethoxy- 2- phenylacetophenone (DMPA) were dissolved in 0.5 mL CHCl3. Then 0.5 mL PEG-200, trimethylolpropane tris(3-mercaptopropionate) (trithiol, 72 μL, 87.6 mg, 0.22 mmol, 0.67 eq.) and NaCl (approximately 4.2 g, 75 wt% of reaction mixture) were added to obtain a homogenous mixture with the NaCl particles completely dispersed. The mixture was exposed to UV light for 1.5 hours in the UV reactor. The material was removed from the vial and placed repeatedly into an excess of H2O to wash out the salt and PEG-200. The scaffolds were further purified by Soxhlet extraction using EtOH for 16 hours and dried under vacuum to obtain polymer 2 as a porous pellet. 31P NMR (solid): δ = 7.7 13 ppm. C NMR (solid): δ = 7.6 (CH3), 26.8 (CH2), 43.8 (NH−CH2), 65.1 (OCH2), 172.1

(C=O) ppm. FTIR (solid): νmax = 3342 (N-H), 2926 (C-H), 1729 (C=O), 1188 (P=N) cm−1. Elemental analysis: calc., C 44.57%, H 6.23%, N 7.80%, S 11.90%, P 5.75%, found, C 43.97%, H 6.21%, N 7.08%, S 11.23%, P 5.47%.

Polymer 1 was further mixed with the commercially available adipic acid divinyl ester (VE) in different ratios in order to modify the rate of degradation of the resulting scaffolds (polymers 3–5). Thiol-ene photo-polymerization using just VE and trithiol gave polymer 6. Conditions for the thiol-ene cross-linking reaction were similar to the thiol- ene reaction for polymer 1, with the molar ratio of ene groups to thiol groups (1:1) adjusted.

−1 Polymer 3: FTIR (solid): νmax = 3340 (N-H), 2927 (C-H), 1728 (C=O), 1188 (P=N) cm . Elemental analysis: calc., C 45.85%, H 6.34%, N 6.42%, S 12.24%, P 4.73%, found, C 45.60%, H 6.31%, N 5.86%, S 11.46%, P 4.58%.

−1 Polymer 4: FTIR (solid): νmax = 3353 (N-H), 2930 (C-H), 1728 (C=O), 1184 (P=N) cm . Elemental analysis: calc., C 47.91%, H 6.50%, N 4.19%, S 12.79%, P 3.09%, found, C 47.50%, H 6.54%, N 4.17%, S 12.36%, P 3.25%.

130

−1 Polymer 5: FTIR (solid): νmax = 3394 (N-H), 2932 (C-H), 1727 (C=O), 1188 (P=N) cm . Elemental analysis: calc., C 50.16%, H 6.68%, N 1.75%, S 13.39%, P 1.29%, found, C 50.11%, H 6.85%, N 1.68%, S 11.97%, P 1.29%.

−1 Polymer 6: FTIR (solid): νmax = 3394 (N-H), 2961 (C-H), 1727 (C=O) cm . Elemental analysis: calc., C 51.78%, H 6.81%, S 13.82%, found, C 51.51%, H 6.86%, S 12.75%.

Photoreactivity study

All formulations (2-6) were prepared by dissolving polymer 1, VE and trithiol (total weight of 177.5 mg) and DMPA (1 mg) in a mixture of CHCl3 (0.5 mL) and PEG-200 (0.5 mL). Photoreactivity of these formulations was tested with a modular compact rheometer equipped with an optical unit (Anton Paar MCR 302 WESP) coupled with a FTIR spectrometer (Bruker V80) and a UV source (Omnicure S2000). The launching of the rheological and IR measurements, as well as the irradiation was synchronized by a homemade trigger. Viscoelasticity experiments were performed in an oscillatory shear mode at a fixed frequency (f = 10 Hz) in a parallel-plate geometry (d = 25 mm) at

25°C. The upper plate was made of aluminium and the bottom plate was a CaF2 window transparent in mid-IR range. The gap between the plates was set to 100 µm. The values of the strain amplitude were checked to ensure that all measurements were conducted within the linear viscoelastic region. The UV source emitting 320-500 nm light was pre-calibrated to provide effective irradiance of 80 mW cm−2 on the upper surface of the CaF2 window as determined by an Ocean Optics USB 2000+ spectrometer. Curing time for all samples was 300 s. The real time FTIR spectroscopy was performed in transmission mode. Fast-responding MCT detector allowed the acquisition of one spectrum per 0.42 s. IR data were collected from 2800 cm−1 to 1500 cm−1 with scan resolution of 4 cm−1. IR absorbance of the C=C band at 1650 cm−1 was ensured to be less than 1 (A ≈ 0.8). The spectrum of CHCl3 /PEG-200 1:1 (v/v) collected at the gap of 86 µm was used as a background. The double bond conversion was determined from the consumption of the vinyl double bonds. Resulting real time IR curves were smoothed with a Savitzky-Golay filter. All measurements were performed with 5 different samples for each formulation to ensure the reproducibility of the results.

131 Functionalization of polymer 1 using thiol-ene chemistry

Polymer 1 was modified with glutathione, a tripeptide bearing a thiol functionality and fluorescein isocyanate (FITC) using thiol-ene chemistry.

In a first thiol-ene reaction, polymer 1 (50.0 mg, 0.18 mmol, 1 eq.) and 0.6 mg DMPA were dissolved in 0.5 mL THF and transferred to a solution of 28.1 µg glutathione (91.6 µmmol, 0.5 eq.) in 0.5 mL H2O. The reaction was irradiated with UV 1 light for 1 hour to obtain polymer 1G. H NMR (D2O acidified with CF3CO2H): δ = 0.68 (br, 0.5H), 1.06 (br, 0.7H), 1.43 (br, 1.3H), 1.64-1.72 (m, 0.6H), 1.91 (br, 0.1H), 2.44 (br, 0.2H), 2.69-3.07 (m, 4.4H), 3.37 (br, 1.7H), 3.44 (m, 0.4H), 4.01 (br, 2H), 4.62 (br, 31 1H) ppm. P NMR (D2O): δ = 4.11 ppm.

Then, 30 μL trithiol (36.5 µg, 0.09 mmol, 0.5 eq.) were added to the THF solution of polymer 1G and the mixture irradiated for another 2 hours leading to polymer 7G. -1 FTIR (solid): νmax = 3256 (N-H), 2923 (C-H), 1732 (C=O), 1195 (P=N) cm .

A method adapted from the literature54 was used to convert the isothiocyanate of FITC into a thiol group as follows. 2-(tritylthio)ethanamine (164.1 mg, 0.51 mmol, 1 eq.) was dissolved in CH2Cl2 and cooled to 0°C. A solution of FITC (200.0 mg, 0.51 mmol, 1 eq.) in CH2Cl2/DMF (1:1) was slowly added. The reaction was allowed to warm to room temperature and stirred for 16 hours. The solvent was removed under vacuum and the product was used without further purification for deprotection of the trityl protecting group using TFA/CH2Cl2/TIPS (10:9:1) (TIPS = triisopropylnaphthalenesulfonic acid sodium salt). After stirring for 1 hour at room temperature, the solvent was removed at reduced pressure. Diethyl ether was added and the mixture was kept in the freezer overnight. The solvent was decanted and the product further dried under high vacuum to obtain FITC-SH. All reactions were monitored with TLC in ethyl acetate/heptane (3:1). Yield 220.07 mg (91.8%).

Polymer 1 (50.0 mg, 0.18 mmol, 1 eq.), FITC-SH (1.71 mg, 3.66 μmol, 0.02 eq.) and 0.5 mg DMPA were dissolved in 1 mL CHCl3 and irradiated for 15 minutes to yield the labelled polymer. Then 35 μL VE (36.3 mg, 0.18 mmol, 1 eq.), 80 μL trithiol, (97.3 mg, 0.24 mmol, 1.33 eq.), 0.5 mL PEG-200 and NaCl were added and the mixture was irradiated for another hour. The resulting polymer 8FITC was washed with water and EtOH to yield a slightly yellow pellet fluorescent under UV light.

132

Degradation studies

o Degradation studies were carried out in deionized H2O at 37 C for 12 weeks. Samples containing 30 mg of the polymers 2-6 were placed in sealed vials and incubated in 2 mL H2O. Data analysis was carried out in triplicate in regular time intervals over 3 months. The samples were then dried in a vacuum oven at 40°C until the weight was constant. The mass loss was determined gravimetrically and is given as the percentage compared to the initial weight of the degradation sample.

Adipose derived stem cell (ASC) studies

Liposuction material was collected from a female, 64 year old donor in a procedure approved by the local ethical board with patient’s consent. Isolation of ASC was performed as described elsewhere.55 Briefly, adipose tissue was washed with phosphate buffered saline (PBS) to remove most of the blood and tumescence solution. Subsequently, tissue was digested with collagenase (Biochrom, Berlin, Germany) at 37°C for 1 h. To eliminate red blood cells, the isolated fraction was incubated with erythrocyte lysis buffer for 10 min. Remaining cells were filtered through a 100 µm filter and cultured in expansion medium (EGM-2, Lonza) at 37°C and 5%

CO2. Cytotoxicity was measured with the Cytotoxicity detection kit (LDH) (Roche Diagnostics, Germany) by washing the scaffolds (50 mg) twice for 24 hours and testing the cytotoxicity of the conditioned media according to the manufacturer’s protocol. Cells cultured with and without 1% Triton X-100 in assay medium were used as controls for maximum and basal LDH release respectively. To test the cytotoxicity of any degradation products, samples were added to 1 ml of cell culture medium and incubated at 37°C for 42 days. For one of the sets the medium was removed every 7 days and tested for cytotoxicity and new medium added. A separate sample was also prepared in which the medium was not replaced for 42 days before testing.

Cell adhesion and proliferation on the scaffolds were measured by seeding 5×104 cells on polymers 2 and 5 of 4 mm³ in size. TissuFleece E® (Baxter, Heidelberg, Germany), a collagen type I scaffold, was used as a control material. The number of cells adhered on the scaffold was calculated using the CellTiter-Glo® Luminescent Cell Viability Assay (Promega, WI) 4 hours after cell seeding. Cell proliferation was measured 3 days after cell seeding using the same method and luminescence was recorded with the Infinite 200 microplate reader (TECAN, Maenndorf, Switzerland).

133 Three replicates of each polymer were used for all analysis and measurements were performed in duplicates. Statistical analyses were performed with GraphPad Prism 5.01, using OneWay ANOVA and Tukey’s post hoc test; p < 0.05 was considered to be statistically significant.

3.3. Results and discussion

Polyphosphazene synthesis

The synthetic target of this study was to develop porous and degradable scaffolds based on an oligomeric poly(organo)phosphazenes (n = 50) bearing allyl glycine moieties (Figure 3-1). The incorporation of glycine spacers adjacent to the backbone are particularly known to accelerate the hydrolytic degradability of the polyphosphazene backbone11,15 and the allyl ester contributes a large number of reactive carbon-carbon double bonds (two per repeat unit) for subsequent photochemical functionalization and cross-linking of the polymer via thiol-ene addition chemistry.

Figure 3-1: Synthesis of polymer 1 bearing glycine moieties and allyl ester groups along the polymer backbone.

Living cationic polymerization of trichlorophosphoranimine followed by nucleophilic substitution of the chlorine atoms with allyl 2-aminoacetate gave the desired polyphosphazene 1. Any residual P−Cl units may have a significant effect on the degradation behaviour of the polymer56 and thus lead to unreproducible results. For this reason, a large excess of allyl glycinate and a long reaction time were applied for macromolecular substitution to ensure full replacement of the chlorine atoms. Figure 3- 2a depicts the 31P NMR spectrum of polymer 1. A broad peak at 1.25 ppm was observed indicating complete removal of the chlorine atoms, up to 31P NMR detection

134

limits. FTIR (Figure 3-3), 1H NMR and 13C NMR spectroscopy further confirmed the successful synthesis of polymer 1 (Figure 3-2b).

Figure 3-2: Analysis of polymer 1 substituted with allyl glycinate (a) 31P NMR spectroscopy gave a broad signal at 1.25 ppm. (b) 13C NMR spectrum of polymer 1 showing five signals from the substituted allyl glycinate and can be associated with the C=O at 172.5 ppm, the resonances for the double bond at 132.2 ppm and 118.4 ppm, the CH2 group in α-position to the ester group at 65.5 ppm, and the CH2 group of the glycine unit at 42.8 ppm.

Figure 3-3: ATR-FTIR spectrum of polymer 1 before cross-linking with significant bands including the P=N stretching band of the polyphosphazene backbone at 1188 cm−1, the C=C band at 1650 cm−1 and the C=O band at 1737 cm−1, both stemming from the substituted allyl ester side groups (black). ATR-FTIR spectrum of polymer 2 after thiol-ene photo-polymerization of polymer 1 with trithiol showing the P=N stretching band at 1188 cm−1, a strong C=O band at 1729 cm−1, and the absence of the C=C band (blue). The two photographs show polymer 1 before cross-linking and the porous pellet of polymer 2 after cross-linking.

135 Thiol-ene photo-polymerization

Polymer 1 was then cross-linked using thiol-ene photo-polymerization to obtain porous glycine-based scaffolds. The thiol-ene photo-polymerization of the allyl groups of 1 and the thiol trimethylolpropane tris(3-mercaptopropionate) (trithiol) was performed at room temperature in the presence of a porogen (see below) in CHCl3 using DMPA as a photo-initiator in small amounts (ca. 1 wt %) (Figure 3-4). The samples were irradiated with UV light in glass vials (cut off 254 nm (≈ 300 nm) for 1.5 hours in order to achieve a maximum consumption of double bonds and thiol groups, whereby the ene and thiol components were used in stoichiometric amounts. Faster reaction times were possible using shorter wavelengths (254 nm), and hence this would be a viable option in the future as the phosphorus−nitrogen backbone is transparent to mid- and long-wavelength UV radiation57. The reaction could be carried out in air, a useful feature of the thiol-ene radical polymerization29. Solidification of the reaction mixture indicated successful formation of the cross-linked polymeric network around the porogen.

Figure 3-4: Simplified scheme of network formation via UV-initiated thiol-ene cross-linking reaction of polymer 1 (red), VE (black) and trithiol (blue). Flexible and degradable polyphosphazene chains are expected to be homogeneously distributed throughout the network thus promoting degradation of the material.

136

After extensive washing and drying, the structure of 2 and the extent of reaction were studied by FTIR and solid-state NMR spectroscopy. The solid-state 31P CP-MAS NMR spectrum gave one broadened signal around 7.7 ppm (Figure 3-5a) and FTIR spectroscopy confirmed the disappearance of the band for the C=C stretching at 1650 cm−1 (Figure 3-3). This result is in agreement with solid-state 13C CP-MAS NMR spectroscopy (Figure 3-5b) in which no resonances for the C=C bonds in the range of 115–140 ppm could be observed indicating a good conversion of the vinyl groups. Since both FTIR and 13C CP-MAS NMR suggest complete absence of the C=C bonds (within detection limits), polymer 2 would appear to be obtained with an extremely high degree of cross-linking.

Figure 3-5: Solid-state 31P CP-MAS NMR spectroscopy of polymer 2 gave a broad signal around 7.7 ppm (a). The solid-state 13C CP-MAS NMR spectrum of polymer 2 confirms the absence of the C=C signals around 115–140 ppm, indicating high conversion of the double bonds during photo-polymerization and a high degree of cross-linking (b). Further signals correspond to C=O at 172.1 ppm and OCH2 at 65.1 ppm from the ester groups, NH−CH2 of the glycine units at 43.8 ppm and CH3 at 7.6 ppm stemming from the trithiol. The CH2 groups at 26.8 ppm are primarily assigned to the thioether units that are generated in the cross-linking reaction.

Blending with organic polymers is a facile methodology to alter polyphosphazene properties,58 particularly the mechanical properties and overall degradation rate. With this in mind we sought to copolymerize the allyl glycine polyphosphazenes with the adipic acid divinyl ester VE (3-6). Vinyl esters have been shown to have low irritancy and cytotoxicity, as well as improved thiol-ene photo-reactivity thus representing an attractive alternative for methacrylate monomers.59 Polymer 1 was mixed thus with VE

137 in various ratios with equimolar feed ratios of thiol and ene groups being maintained (Table 3-1). Co-polymerization with polymer 1 and various amounts of VE were carried out in an identical manner to with the polymer 2 and elemental analysis was used to confirm the ratios of the various components, with the content of the inorganic polyphosphazene backbone ranging from 7.83% for polymer 2 to 1.85% for polymer 5 (Table 3-1). Furthermore, the calculated and found values of elemental analysis of polymers 2–6 are in good agreement indicating that the initial composition of 1, VE and trithiol remains approximately constant and that all components are homogeneously incorporated into the matrices. Further visual evidence of a uniform distribution of the polyphosphazene in the matrix was also provided by the addition of a fluorescent marker to the polyphosphazene (see section “scaffold functionalization”).

Table 3-1: Composition of polymers 2-6.

Polymer composition (wt%)

a) b) Molar [P=N] [P=N] Polymer Polyphosphazene Vinyl ester Trithiol n n ratio (wt% (wt% matrix 1 VE c) d) 1:VE added) found)

2 51 0 49 100:0 8.35 7.83

3 42 7 51 80:20 6.87 6.53

4 27 20 53 50:50 4.49 4.64

5 11 34 55 20:80 1.87 1.85

6 0 43 57 0:100 - -

a) Vinyl ester = adipic acid divinyl ester b) Trithiol = trimethylolpropane tris(3-mercaptopropionate) c) Calculated [P=N] content, i.e. amount of inorganic chains embedded in the matrix d) Content found by elemental analysis

Scaffolds with interconnected pores and high porosity represent suitable substrates for cell growth, facilitating the transport of nutrients and metabolites which is essential for cell survival. In a simple procedure to prepare preliminary porous matrices for initial biocompatibility and cell proliferation studies, NaCl particles were chosen as a

138

porogen to obtain porous polymeric scaffolds of polymer 1 using thiol-ene photo- polymerization. A solution of polymer 1, the trithiol, VE and the photo-initiator DMPA in 30,60 CHCl3, a known system for thiol-ene polymerizations, were combined with ground NaCl and PEG-200, whereby the PEG-200, although it may slow the polymerization,31 was added as a co-solvent for removal of the cross-linked materials from the glass vials. Irradiation with UV light resulted in solidification and cross-linking of the reaction mixture followed by an extensive washing procedure to dissolve and wash away the embedded salt particles and PEG-200, leaving the insoluble cross-linked polymers 2–6 behind. After drying, the porous scaffolds were obtained as cylindrical pellets with approximate dimensions of 15 mm × 4 mm (diameter × height). The morphology of the polymer pellets was investigated by SEM and x-ray CT and is shown in Figure 3-6 for polymers 2 and 5 as representative examples. The polymers are relatively soft elastic, non-brittle materials, despite the extremely high level of cross-linking, presumably as a consequence of the highly flexible polyphosphazene backbone. The mechanical properties of these materials are clearly dependent on the porous properties, but initial compression testing no discernable trends could be observed with all materials showing the 10% sigma value, (a measure of the stiffness at 10% compression) in the region of 20-40 kPa, a factor of 2-3 higher than the reference material. Further mechanical testing is required as these were only proof-of-principle materials, but it is clear to see that with a more advanced scaffold preparation, these materials have the potential to form engineered porous scaffolds for tissue engineering purposes.

Figure 3-6: Scanning electron microscopy (SEM) images of polymer 2 a) and 5 b) respectively, as well as X-ray CT images of polymer 2 c) §3D imaging and d) in cross section. The highly porous structure with interconnected pores and pore sizes in the range of 100–200 μm (average pore sizes 144 μm and 150 μm for polymer 2 and 5 respectively according to X-ray CT analysis).

139 Photoreactivity

The evolution of the molecular and chemical changes during the thiol-ene photo- polymerization were monitored in situ in a photo-rheometer coupled with a FTIR spectrometer. Double bond conversion (DBC) was determined from the disappearance of the stretching band of the C=C bond at 1650 cm−1. Conversion of thiols determined from the stretching band of the S−H bond at 2560 cm−1 did not give reliable values due to its typical weak absorption and due to uneven baseline caused by solvent absorption in the same region. Since the thiol/ene ratio is 1, all polymerizations should follow first order reaction kinetics.61 The rate constants were determined as the slopes of linear plot ln A/A0 vs. time (data not shown). As expected, the rate of polymerization decreases as the molar ratio of vinyl ester (VE): allyl ester (1) shifts towards allyl ester (Figure 3-7). It is known that the ene reactivity decreases with decreasing electron density of the C=C double bond.61 Hence, sample 6 containing VE as the only ene polymerizes about 40 times faster than sample 2 containing allyl ester groups only (Figure 3-8). Final DBCs after 5 min reached values between 95% (2) and 100% (6). Although it is difficult to determine the traces of unreacted groups, there is good reason to believe that their contribution to irritancy and cytotoxicity is negligible, especially since electron-rich character of vinyl groups of both allyl ester and vinyl ester makes them rather unreactive towards nucleophilic addition of amino and mercapto groups of life-essential molecules, e.g. proteins.59,62 Thiol groups have also been found to be of low cytotoxicity.31

Figure 3-7. Double bond conversion–time plots from real time FTIR of the photo-polymerization of the formulations 2-6.

140

Rheological measurements reflect changes on a molecular level caused by cross- linking. Prior to curing, the storage modulus G’ is lower than the loss modulus G’’ indicating that the sample is a liquid. At the beginning of the curing, molecular weight increases but the sample still behaves as a liquid as G’ remains constant. Afterwards, G’ starts to grow dramatically. The sample then undergoes liquid-to-solid transition at the gel point, which however cannot be simply determined from a single frequency experiment. The gel point does not occur when G’ crosses G’’ but rather in its vicinity.63 In accordance with the kinetics from IR measurements, the time to reach the G’ – G’’ crossover decreases from 21.6 s (2) to 1.6 s (6). The dependence of this time on molar ratio VE:1 gives a linear plot (Figure 3-8). At the crossover point, the DBC ranges from 51% (2) to 90% (6). Finally, G’ reaches a plateau and a few orders of magnitude higher than G’’ as the sample became more elastic. G’ of the samples 5 and 6 was below 1 kPa while the samples 2-4 with higher cross-linking density reached nearly 30 kPa.

Figure 3-8. Storage modulus of the cured gels, time to reach the G’ G’’ crossover and polymerization rate constant for the formulations 2-6 as a function of molar ratio vinyl ester (VE) : allyl ester (1).

141 Scaffold functionalization

Due to the modular nature of this thiol-ene photo-polymerization, the polymer properties, e.g. cell adhesion and surface wettability, can be readily modified by co- addition of thiol containing reagents. To demonstrate this, polymer 1 was reacted with glutathione, a tripeptide bearing a thiol group, and a thiol-functionalized fluorescein (FITC-SH). UV-induced thiol-ene addition of the respective thiol onto polymer 1 was carried out in advance of the thiol-ene photo-crosslinking, yielding the functionalized polyphosphazenes as depicted in Figure 3-9.

Figure 3-9: Functionalization of polymer 1 with the tripeptide glutathione (a) and the fluorescence marker FITC (b) using thiol-ene chemistry. The residual double bonds can be subsequently cross-linked using the method described above to obtain a fluorescent pellet (photograph).

142

Glutathione was chosen as a model compound for the introduction of thiol containing peptide sequences and is expected to increase the hydrophilicity of the scaffolds. Glutathione was efficiently coupled onto polymer 1G as confirmed by 1H NMR spectroscopy indicating that approximately 25% of the allyl double bonds formed a thioether with a glutathione molecule. Thiol-ene photo-polymerization of the peptide- functionalized polyphosphazenes 1G and trithiol was still efficient enough to produce a cross-linked network with high conversion of the double bonds as revealed by FTIR spectroscopy. The remaining ene bonds of 1G were thus not sterically restricted after conjugation of the glutathione, presumably because of the high flexibility of the polyphosphazene chains, and cross-linking was still possible without any limitations.

In order to insert visible markers into the matrix, polymer 1, was also reacted with FITC-SH (2 mol% with respect to the double bonds of 1) (Figure 3-9), followed by subsequent matrix formation. Extensive washing in EtOH gave polymer 8FITC as a homogeneously fluorescent pellet when illuminated with a low power UV lamp (Figure 9). The FITC modified polyphosphazene building blocks have clearly been incorporated into the matrix and indeed appear equally distributed throughout the entire material. The material can thus be described as a cross-linked organic network with homogenously interdispersed polyphosphazene [P=N] chains.

The synthetic concept of this work can essentially be described as a modular tool box, with polyphosphazene 1 as a flexible building block and versatile platform for modification and/or cross-linking reactions due to the large number of double bonds. Choosing multifunctional thiols with different spacers could be used to adapt properties and to control the hydrophilicity/hydrophobicity of the final scaffold. Functionalization via thiol-ene addition allows covalent attachment of various molecules bearing a thiol group, which was demonstrated herein with two model compounds: (1) glutathione, a tripeptide to increase the hydrophilicity of the scaffolds, and (2) the introduction of a thiol-functionalized fluorescence marker (FITC). This modification method could enable the introduction of, for example, bioactive ligands, growth factors or drugs onto the polyphosphazene component and thus the organic phase of the matrices in order to promote tissue growth. The properties of these porous matrices can thus be easily tuned, whereas only the synthesis of one poly(organo)phosphazene, that provides synthetic flexibility and promotes degradability, was required.

143 Degradation studies

Bioerodible scaffolds with tunable degradation rates and degradation to bioinert small molecules are key requirements for tissue engineering matrices. Amino acid esters are well established to promote the degradation of polyphosphazenes.11

Figure 3-10: Degradation studies on porous matrices in H2O at 37°C, plotted as percentage mass loss versus time. The hydrolytic degradation of the photo-polymerized poly(organo)phosphazene (polymer 2) is observed to be significantly faster than pristine polyester (polymer 6). Upon copolymerization, the hydrolysis rate of polymer 4 is intermediate to that of the pristine polymers.

Studies of the hydrolytic degradation of polymers 2, 4 and 6 were conducted to demonstrate the effect of composition change on the degradation rates (Figure 3-10). For polymer 6, containing no polyphosphazene moieties, only minimal changes were observed under aqueous conditions at 37°C. Polymer 2, however, degrades at a steady rate, with a rapid bioerosion to soluble materials observed within days. Through copolymerization with the more hydrolytically stabile vinyl ester, it was proposed that it should be possible to tailor the degradation rates. To test this, polymer 4, containing an intermediate ratio of polyphosphazene and VE (see table 3-1) was tested under the same conditions and was indeed shown to have a hydrolytic degradation rate between that of the two pristine polymers.

144

Polymer 2, containing the highest proportion of polyphosphazenes showed a clear degradation profile in neutral aqueous solutions at 37°C. As a consequence of the hydrolytic breakdown of the polyphosphazene backbone, the network connections of the cross-linked organic mass are cleaved, thus leading to an enhanced overall degradation rate of the scaffolds. In contrast, no degradation was observed under these conditions for polyester 6, which contained no polyphosphazenes. Copolymerization of the polyphosphazene with VE (polymer 4) was observed to considerably reduce its rate of degradation, presumably due to its higher hydrophobicity (contact angle measurements showed an increase from 41° to 52° and then 60° for polymers 2, 4 and 6 respectively). Degradation of similar structures are known to result in the cleaved organic side groups, phosphates and ammonia.11,15,64

Application studies

One of the potential future applications of the biomaterial developed in this study includes the field of osteochondral tissue engineering. For this reason, cytotoxicity studies were carried out on ASC, with no cytotoxicity being detected for cells tested with medium previously conditioned with polymer 2 or polymer 5 (Figure 3-11a). Furthermore, preliminary studies also showed no cytotoxicity of polymer 2 (Figure 3- 11b), containing the highest proportion of polyphosphazene, up to 42 days at 37°C in cell culture medium, a time point at which 35% of the polymer was shown to have already degraded (see earlier), thus suggesting non-toxicity of the degradation products or their intermediates. Future work will entail detailed studies of the precise nature of the degradation products and their impact on cell growth.

145

Figure 3-11. (a) Cytotoxicity of polymer supernatants for ASC after 1 and 2 days medium conditioning. (b) Cytotoxicity of the partially degraded polymer 2 stored at 37°C in cell culture medium. Fresh medium was added every 7 days, except for sample 1-42, in which the same medium was applied throughout. (c) Adhesion measured by calculating the number of cells on the scaffold 4 hours after seeding of the cells and (d) proliferation of ASC on the scaffolds was measured by quantifying the adenosine triphosphate (ATP) present in the cells, which is proportional to their metabolic activity, after 3 and 6 days. n = 3 biological replicates, * q < 0.05, OneWay ANOVA, Tukey’s post hoc test.

In this field, collagen scaffolds currently play a major role such as BioGide® or ChondroGide® (Geistlich Surgeries, Switzerland); MACI, a collagen I/III fleece (Genzyme, Boston) or the collagen type I gel CaReS® (Arthro Kinetics Biotechnology, Austria)65. Most collagen scaffolds are of bovine or equine origin. In contrast to these materials we aim to develop a synthetic polymer which can be produced in a consistent way and is free of risks which are accompanied with the use of xenogenic biological materials. For this reason, cell adhesion and proliferation studies were also carried out with a TissuFleece E® (Baxter) as a reference material66. TissuFleece E® is a collagen sponge with a pore size in a similar range (188 ± 76 μm)67 as the newly synthesized materials and thus represents a satisfactory reference standard. Furthermore, TissueFleece E has been recently shown to outperform clinically used biomaterials

146

regarding cell proliferation and metabolic activity including BioGide® and is reported as a promising carrier material for chondrogenic MSCs for cartilage tissue engineering attempts.68 In these initial studies it was found that 12.6+5.9 % of ASC adhered to polymer 2, which could be increased to 30.6+11.6 % by copolymerization with VE (polymer 5), possibly due in part to its higher hydrophobicity. Nevertheless, further surface modifications will be necessary to improve cell adhesion. On polymer 2 the number of metabolically active cells decreased over time but significantly increased 2.5-fold within the first three days on polymer 5 (Figure 3-11c). On TissuFleece E® the cell number increased 3.5-fold in our studies, while Gierloff et al. demonstrated a 2-fold induction within 3 days of culture68 (Figure 3-11d).

3.4. Conclusion

In summary, we have presented a modular pathway for the preparation of porous glycine-based highly cross-linked scaffolds with tunable degradability and adaptable properties. Using an efficient one-pot thiol-ene photo-polymerization between the polyphosphazene and a trifunctional thiol resulted in cross-linked scaffolds with pore sizes in the range of 100–200 μm could be prepared. The polymers could be simply functionalized, a tactic which could be used in future as a facile method to introduce specific surface functionalities for the optimization of cell to scaffold interactions, cell adhesion and wettability. The degradation rates can be adjusted by mixing the glycine- substituted polyphosphazene and a divinylester in different ratios, with a more rapid degradation being observed for larger amounts of interdispersed polyphosphazene chains in the organic network. No cytotoxicity was observed for the tested samples or indeed their degradation products, indicating biocompatibility. Furthermore, the samples applicability in osteochondral tissue engineering was investigated, with promising initial results. This work represents a valuable proof of principle for this photo-polymerization route to biocompatible, degradable and easily tunable materials. Future work will entail, as well as further property optimization, the investigation of advanced scaffold preparation techniques to prepare homogeneous porous systems with suitable mechanical properties for tissue engineering applications.

147 Acknowledgements

Solid-state NMR measurements were carried out with the assistance of W. Schoefberger at the Upper Austrian-South Bohemian Research Infrastructure Center in Linz, co-financed by the European Union in the context of the project "RERI-usab", EFRE RU2-EU-124/100-2010 (ETC Austria-Czech Republic 2007-2013, project M00146). Gold sputtering of polymers for scanning electron microscopy was carried out by Markus Gillich, University of Applied Sciences, Wels, Austria and samples were subsequently imaged by Stefan Schwarzmayer (ARS Electronica Center, Linz, Austria). We would like to acknowledge J. Theiner (Mikroanalytisches Labor, Universität Wien) for elemental analysis. Dietmar Salaberger and Christian Hannesschläger are acknowledged for X-ray computed tomography at the University of Applied Sciences, Upper Austria, Wels, Austria. The authors acknowledge financial support of the EU Regio 13 EFRE grant OÖ Wi/225882/2013/FA. B.H. and R.L. wish to thank the European Commission for financial support of the project VINDOBONA (grant no. 297895) in the framework of Marie Curie Intra-European Fellowships.

148

3.5. References

1. Ulery, B. D.; Nair, L. S.; Laurencin, C. T. Journal of Polymer Science Part B: Polymer Physics 2011, 49, 832. 2. Ravichandran, R.; Sundarrajan, S.; Venugopal, J. R.; Mukherjee, S.; Ramakrishna, S. Macromol. Biosci. 2012, 12, 286. 3. Nair, L. S.; Laurencin, C. T. In Tissue Engineering I: Scaffold Systems for Tissue Engineering; Lee, K., Kaplan, D., Eds. 2006; Vol. 102, p 47. 4. Hollister, S. J. Nat. Mater. 2006, 5, 590. 5. Lutolf, M. P.; Hubbell, J. A. Nature Biotechnology 2005, 23, 47. 6. Chen, G.; Ushida, T.; Tateishi, T. Macromol. Biosci. 2002, 2, 67. 7. Place, E. S.; George, J. H.; Williams, C. K.; Stevens, M. M. Chemical Society Reviews 2009, 38, 1139. 8. Deng, M.; Kumbar, S. G.; Wan, Y.; Toti, U. S.; Allcock, H. R.; Laurencin, C. T. Soft Matter 2010, 6, 3119. 9. Kumbar, S. G.; Bhattacharyya, S.; Nukavarapu, S. P.; Khan, Y. M.; Nair, L. S.; Laurencin, C. T. J. Inorg. Organomet. Polym. Mater. 2006, 16, 365. 10. Allcock, H. R. Soft Matter 2012, 8, 7521. 11. Allcock, H. R.; Morozowich, N. L. Polym. Chem. 2012, 3, 578. 12. Tian, Z.; Zhang, Y.; Liu, X.; Chen, C.; Guiltinan, M. J.; Allcock, H. R. Polym. Chem. 2013, 4, 1826. 13. Deng, M.; Nair, L. S.; Nukavarapu, S. P.; Jiang, T.; Kanner, W. A.; Li, X.; Kumbar, S. G.; Weikel, A. L.; Krogman, N. R.; Allcock, H. R.; Laurencin, C. T. Biomaterials 2010, 31, 4898. 14. Sung, H.-J.; Meredith, C.; Johnson, C.; Galis, Z. S. Biomaterials 2004, 25, 5735. 15. Wilfert, S.; Iturmendi, A.; Schoefberger, W.; Kryeziu, K.; Heffeter, P.; Berger, W.; Brüggemann, O.; Teasdale, I. Journal of Polymer Science Part A: Polymer Chemistry 2014, 52, 287. 16. Deng, M.; Kumbar, S. G.; Nair, L. S.; Weikel, A. L.; Allcock, H. R.; Laurencin, C. T. Adv. Funct. Mater. 2011, 21, 2601. 17. Weikel, A. L.; Cho, S. Y.; Morozowich, N. L.; Nair, L. S.; Laurencin, C. T.; Allcock, H. R. Polym. Chem. 2010, 1, 1459. 18. Peach, M. S.; Kumbar, S. G.; James, R.; Toti, U. S.; Balasubramaniam, D.; Deng, M.; Ulery, B.; Mazzocca, A. D.; McCarthy, M. B.; Morozowich, N. L.; Allcock, H. R.; Laurencin, C. T. J. Biomed. Nanotechnol. 2012, 8, 107.

149 19. Morozowich, N. L.; Nichol, J. L.; Mondschein, R. J.; Allcock, H. R. Polym. Chem. 2012, 3, 778. 20. Cui, Y.; Zhao, X.; Tang, X.; Luo, Y. Biomaterials 2004, 25, 451. 21. Wilfert, S.; Iturmendi, A.; Henke, H.; Brüggemann, O.; Teasdale, I. Macromol. Symp. 2014, 337, 116. 22. Song, S. C.; Lee, S. B.; Jin, J. I.; Sohn, Y. S. Macromolecules 1999, 32, 2188. 23. Chun, C.; Lim, H. J.; Hong, K. Y.; Park, K. H.; Song, S. C. Biomaterials 2009, 30, 6295. 24. Potta, T.; Chun, C.; Song, S. C. Biomaterials 2009, 30, 6178. 25. Laurencin, C. T.; El-Amin, S. F.; Ibim, S. E.; Willoughby, D. A.; Attawia, M.; Allcock, H. R.; Ambrosio, A. A. Journal of Biomedical Materials Research 1996, 30, 133. 26. Nair, L. S.; Bhattacharyya, S.; Bender, J. D.; Greish, Y. E.; Brown, P. W.; Allcock, H. R.; Laurencin, C. T. Biomacromolecules 2004, 5, 2212. 27. Brown, J. L.; Nair, L. S.; Laurencin, C. T. Journal of Biomedical Materials Research Part B: Applied Biomaterials 2008, 86B, 396. 28. Deng, M.; Nair, L. S.; Nukavarapu, S. P.; Kumbar, S. G.; Jiang, T.; Weikel, A. L.; Krogman, N. R.; Allcock, H. R.; Laurencin, C. T. Adv. Funct. Mater. 2010, 20, 2794. 29. Lowe, A. B. Polym. Chem. 2010, 1, 17. 30. Caldwell, S.; Johnson, D. W.; Didsbury, M. P.; Murray, B. A.; Wu, J. J.; Przyborski, S. A.; Cameron, N. R. Soft Matter 2012, 8, 10344. 31. Mautner, A.; Qin, X.; Wutzel, H.; Ligon, S. C.; Kapeller, B.; Moser, D.; Russmueller, G.; Stampfl, J.; Liska, R. Journal of Polymer Science Part A: Polymer Chemistry 2013, 51, 203. 32. Zavada, S. R.; McHardy, N. R.; Scott, T. F. Journal of Materials Chemistry B 2014. 33. Bourin, P.; Bunnell, B. A.; Casteilla, L.; Dominici, M.; Katz, A. J.; March, K. L.; Redl, H.; Rubin, J. P.; Yoshimura, K.; Gimble, J. M. Cytotherapy 2013, 15, 641. 34. Hildner, F.; Eder, M. J.; Hofer, K.; Aberl, J.; Redl, H.; van Griensven, M.; Gabriel, C.; Peterbauer-Scherb, A. Journal of Tissue Engineering and Regenerative Medicine 2013, n/a. 35. Malafaya, P. B.; Pedro, A. J.; Peterbauer, A.; Gabriel, C.; Redl, H.; Reis, R. L. J. Mater. Sci.: Mater. Med. 2005, 16, 1077.

150

36. Peterbauer-Scherb, A.; Danzer, M.; Gabriel, C.; van Griensven, M.; Redl, H.; Wolbank, S. Journal of Tissue Engineering and Regenerative Medicine 2012, 6, 434. 37. Ashjian, P. H.; Elbarbary, A. S.; Edmonds, B.; DeUgarte, D.; Zhu, M.; Zuk, P. A.; Lorenz, P. H.; Benhaim, P.; Hedrick, M. H. Plast. Reconstr. Surg. 2003, 111, 1922. 38. Banerjee, A.; Nürnberger, S.; Hennerbichler, S.; Riedl, S.; Schuh, C. A. P.; Hacobian, A.; Teuschl, A.; Eibl, J.; Redl, H.; Wolbank, S. Cell Tissue Bank 2013, 1. 39. Boquest, A. C.; Shahdadfar, A.; Frønsdal, K.; Sigurjonsson, O.; Tunheim, S. H.; Collas, P.; Brinchmann, J. E. Mol. Biol. Cell 2005, 16, 1131. 40. Safford, K. M.; Hicok, K. C.; Safford, S. D.; Halvorsen, Y. D. C.; Wilkison, W. O.; Gimble, J. M.; Rice, H. E. Biochem. Biophys. Res. Commun. 2002, 294, 371. 41. Rangappa, S.; Fen, C.; Lee, E. H.; Bongso, A.; Wei, E. S. K. Annals of Thoracic Surgery 2003, 75, 775. 42. Miranville, A.; Heeschen, C.; Sengenes, C.; Curat, C. A.; Busse, R.; Bouloumie, A. Circulation 2004, 110, 349. 43. Planat-Benard, V.; Silvestre, J. S.; Cousin, B.; Andre, M.; Nibbelink, M.; Tamarat, R.; Clergue, M.; Manneville, C.; Saillan-Barreau, C.; Duriez, M.; Tedgui, A.; Levy, B.; Penicaud, L.; Casteilla, L. Circulation 2004, 109, 656. 44. Rehman, J.; Traktuev, D.; Li, J. L.; Merfeld-Clauss, S.; Temm-Grove, C. J.; Bovenkerk, J. E.; Pell, C. L.; Johnstone, B. H.; Considine, R. V.; March, K. L. Circulation 2004, 109, 1292. 45. Al Battah, F.; De Kock, J.; Vanhaecke, T.; Rogiers, V. Thescientificworldjournal 2011, 11, 1568. 46. Seo, M. J.; Suh, S. Y.; Bae, Y. C.; Jung, J. S. Biochem. Biophys. Res. Commun. 2005, 328, 258. 47. Baer, P. C. Stem Cells and Development 2011, 20, 1805. 48. Brzoska, M.; Geiger, H.; Gauer, S.; Baer, P. Biochem. Biophys. Res. Commun. 2005, 330, 142. 49. Vossmerbaeumer, U.; Ohnesorge, S.; Kuehl, S.; Haapalahti, M.; Kluter, H.; Jonas, J.; Thierse, H. J.; Bieback, K. Cytotherapy 2009, 11, 177. 50. Du, Y.; Roh, D. S.; Funderburgh, M. L.; Mann, M. M.; Marra, K. G.; Rubin, J. P.; Li, X.; Funderburgh, J. L. Mol. Vis. 2010, 16, 2680. 51. Ferro, F.; Spelat, R.; Falini, G.; Gallelli, A.; D'Aurizio, F.; Puppato, E.; Pandolfi, M.; Beltrami, A. P.; Cesselli, D.; Beltrami, C. A.; Ambesi-Impiombato, F. S.; Curcio, F. Am. J. Pathol. 2011, 178, 2299.

151 52. Hu, M.; Li, L.; Wu, H.; Su, Y.; Yang, P.-Y.; Uttamchandani, M.; Xu, Q.-H.; Yao, S. Q. J. Am. Chem.Soc., 2011, 133, 12009. 53. Honeyman, C. H.; Manners, I.; Morrissey, C. T.; Allcock, H. R. J. Am. Chem.Soc. , 1995, 117, 7035. 54. Kaufmann, T.; Gokmen, M. T.; Rinnen, S.; Arlinghaus, H. F.; Du Prez, F.; Ravoo, B. J. Journal of Materials Chemistry 2012, 22, 6190. 55. Wolbank, S.; Peterbauer, A.; Fahrner, M.; Hennerbichler, S.; Van Griensven, M.; Stadler, G.; Redl, H.; Gabriel, C. Tissue Engineering 2007, 13, 1173. 56. Andrianov, A. K.; Svirkin, Y. Y.; LeGolvan, M. P. Biomacromolecules 2004, 5, 1999. 57. Allcock, H. R.; Phelps, M. V. B.; Barrett, E. W.; Pishko, M. V.; Koh, W.-G. Chem. Mater. 2006, 18, 609. 58. Ibim, S. E. M.; Ambrosio, A. M. A.; Kwon, M. S.; El-Amin, S. F.; Allcock, H. R.; Laurencin, C. T. Biomaterials 1997, 18, 1565. 59. Husár, B.; Heller, C.; Schwentenwein, M.; Mautner, A.; Varga, F.; Koch, T.; Stampfl, J.; Liska, R. Journal of Polymer Science Part A: Polymer Chemistry 2011, 49, 4927. 60. Claudino, M.; Jonsson, M.; Johansson, M. RSC Advances 2013, 3, 11021. 61. Hoyle, C. E.; Lee, T. Y.; Roper, T. Journal of Polymer Science Part A: Polymer Chemistry 2004, 42, 5301. 62. Heller, C.; Schwentenwein, M.; Russmueller, G.; Varga, F.; Stampfl, J.; Liska, R. Journal of Polymer Science Part A: Polymer Chemistry 2009, 47, 6941. 63. Winter, H. H. Polym. Eng. Sci. 1987, 27, 1698. 64. Andrianov, A. K.; Marin, A.; Peterson, P. Macromolecules 2005, 38, 7972. 65. Nuernberger, S.; Cyran, N.; Albrecht, C.; Redl, H.; Vecsei, V.; Marlovits, S. Biomaterials 2011, 32, 1032. 66. Hildner, F.; Concaro, S.; Peterbauer, A.; Wolbank, S.; Danzer, M.; Lindahl, A.; Gatenholm, P.; Redl, H.; van Griensven, M. Tissue Engineering Part A 2009, 15, 3961. 67. Büttner, O., J., PhD Thesis, Albert-Ludwigs-Universität Freiburg im Breisgau 2006. 68. Gierloff, M.; Nitsche, T.; Adam-Klages, S.; Liebs, K.; Hedderich, J.; Gassling, V.; Wiltfang, J.; Kabelitz, D.; Acil, Y. Clinical Oral Investigations 2014, 18, 247.

152

4. Summary and Outlook

In this thesis, diverse poly(organo)phosphazenes have been presented with tailored properties and degradation rates.

First, water-soluble poly(organo)phosphazenes with pH-promoted degradation behavior could be prepared (Chapter 2.1). The hydrolytic degradation of each polymer was carefully studied in which faster degradation rates could be achieved at lower pH values or by incorporation of some amino acid spacers between the polymer backbone and the organic substituents. Furthermore, the cytotoxicity of selected polymers, as well as their degradation products and intermediates was investigated with no observed detrimental effect on the cell viability studies. Therefore, these degradable and good water-soluble poly(organo)phosphazenes could be of great importance for many biomedical applications. Considering the need for appropriate or controlled degradation rates depending on the application, the degradation rates might be further tuned by varying the amounts of the side groups or by addition of other hydrophilic substituents with different leaving ability or stimulated behavior.

Then, oxidation responsive polymers that degrade in presence of H2O2 have been presented (Chapter 2.2). In this case, the degradability of the polymers in the oxidative environment could be observed while no degradation signs in absence of H2O2 over same period of time were shown. After H2O2 addition, the arylboronate pinacol ester group is quickly oxidized to a phenol, which undergoes self-immolation through 1,4- elimination, to uncage the poly(glycine)phosphazene that catalyzes the polymer degradation. As a rate-limiting step was observed between the phenol formation and the self-immolation, modification of the caging group or the backbone linkage might be useful to enhance the degradation rates. Partial hydrolysis of the boronic acid ester, that is, cleavage of the pinacol group, in aqueous media is also perceived which opens the opportunity to further chemistries, such as incorporation of a fluorescence marker or cross-linking reactions with certain diols. Therefore, it would be of great interest to synthesize some nanoparticles loaded with a possible drug and investigate the nanoparticle degradation and its cargo release upon H2O2 exposure. Furthermore, biocompatibility studies of these polymers would be necessary to obtain further information about the suitability and applicability of them in biomedical applications.

153 As the final stimulus, visible light has been introduced to trigger the poly(organo)phosphazenes degradation (Chapter 2.3). A coumarin derivative with diethylamino electron-donating groups was chosen to cage the unstable poly(glycine)phosphazene due to its bathochromically shifted absorption that provides its photo-cleavage upon irradiation with light even around 400 nm. After the detailed study and successful photo-cleavage reaction of the coumarin-caged glycine derivative, it was incorporated onto the poly(organo)phosphazene backbone together with Jeffamine M-1000 or glycine ethyl ester. Both polymers show photo-cleavage reactivity upon irradiation with the subsequent polymer degradation but with different rates. The diversities of the polymer degradation rates may be due to the differences in the leaving ability of the second substituent, that show faster degradation with glycine ethyl ester (pKa 7.75) than with M-1000 (pKa 9.75). Moreover, some parts of the polymer might consist only of M-1000 which may shield the polymer backbone from a possible hydrolytic attack due to its long chains. The coumarin derivative studied is photochemically active around 400 nm, hence other compounds with longer wavelengths could be studied while hopefully keeping the rapid degradation of the polymers. Water soluble derivatives would be also appealing for several medical applications, thus it would be of great importance to test the cytotoxicity of these polymers as well as their degradation products to confirm their suitability in medical applications. Last but not least, degradable porous scaffolds to allow cell growth in tissue engineering applications have been presented (Chapter 3). The incorporation of allyl functionalities on the glycine group allows the cross-linking reaction as well as further functionalization via thiol-ene chemistry. Furthermore, mechanical properties along with degradation rates can be readily tailored by mixture of divinyl adipate with the polymer. Although highly interconnected pores, in the range of 100-200 m, via the salt leaching technique have been achieved, improved methods might be required in order to obtain homogenous and controlled pore sizes. The biocompatibility of selected scaffolds has been confirmed with no cytotoxic sign even in the degradation products. Albeit cell proliferation studies have been performed, it might be useful to investigate the cell proliferation for a longer period of time, while the scaffolds degrade, to confirm the suitability of such materials in tissue engineering applications.

154

The presented polyphosphazenes, with tunable and stimulated degradation rates, are attractive candidates for the design of water-soluble drug-delivery systems and biomaterials for tissue engineering applications. The effectiveness of diverse stimuli to trigger the degradation of the polymers opens the possibility to design systems to be used in medical applications that respond to different target environments.

155

156

PUBLICATIONS

 A. Iturmendi, S. Theis, D. Maderegger, U. Monkowius, I. Teasdale. Coumarin- Caged Polymers with a Visible-Light Driven on-Demand Degradation 2018, submitted.

 A. Iturmendi and I. Teasdale. Water Soluble (Bio)degradable Poly(organo)phosphazenes. In Polyphosphazenes in Biomedicine, Engineering, and Pioneering Synthesis ACS Book 2018, accepted.

 A. Linhardt, M. König, A. Iturmendi, H. Henke, O. Brüggemann and I. Teasdale Degradable, Dendritic Polyols on a Branched Polyphosphazene Backbone. Ind. Eng. Chem. Res. 2018, 57 (10), 3602-3609. doi: 10.1021/acs.iecr.7b05301

 S. Theis, A. Iturmendi, C. Gorsche, M. Orthofer, M. Lunzer, S. Baudis, A. Ovsianikov, R. Liska, U. Monkowius and I. Teasdale. Metallo-Supramolecular Gels that are Photocleavable with Visible and Near-Infrared Irradiation. Angew. Chem. Int. Ed. 2017, 56, 1-5. doi.org/10.1002/anie.201707321

 A. Iturmendi, U. Monkowius, I. Teasdale. Oxidation Responsive Polymers with a Triggered Degradation via Arylboronate Self-Immolative Motifs on a Polyphosphazene Backbone. ACS Macro Lett. 2017, 6(2), 150-154. doi: 10.1021/acsmacrolett.7b00015

 I Teasdale, S. Wilfert, T. Aigner, A. Iturmendi, O. Brüggemann, K. Schröder, G. Olawale, F. Hildner, M. Rigau de Llobet, Polymers for Tissue Engineering, Patent application 2017, EP3180390A1

 I Teasdale, S. Wilfert, T. Aigner, A. Iturmendi, O. Brüggemann, K. Schröder, G. Olawale, F. Hildner, M. Rigau de Llobet, Polymerstruktur und dreidimensionales Gerüst für die Gewebezüchtung Austrian Granted Patent 2016, AT 515 955 B1

157  S. Rothemund, T. B. Aigner, A. Iturmendi, M. Rigau, B. Husár, F. Hildner, E. Oberbauer, M. Prambauer, G. Olawale, R. Forstner, R. Liska, K. R. Schröder, O. Brüggemann, I. Teasdale. Degradable Glycine-Based Photo-Polymerizable Polyphosphazenes for Use as Scaffolds for Tissue Regeneration. Macromol. Biosci. 2015,15(3), 351-363. doi: 10.1002/mabi.201400390

 S. Wilfert, A. Iturmendi, W. Schoefberger, K. Kryeziu, P. Heffeter, W. Berger, O. Brüggemann and I. Teasdale. Water-Soluble, Biocompatible Polyphosphazenes with Controllable and pH-Promoted Degradation Behavior. J. Polym. Sci. A Polym. Chem. 2014, 52(2), 287-294. doi: 10.1002/pola.27002

 S. Wilfert, A. Iturmendi, H. Henke, O. Brüggemann and I. Teasdale. Thermoresponsive Polyphosphazene-Based Molecular Brushes by Living Cationic Polymerization. Macromol. Symp. 2014, 337(1): 116-123. doi: 10.1002/masy.201450314

 Henke, H., S. Wilfert, A. Iturmendi, O. Brueggemann and I. Teasdale. Branched Polyphosphazenes with Controlled Dimensions. J. Polym. Sci. A Polym. Chem. 2013, 51(20), 4467–4473. doi: 10.1002/pola.26865

CONFERENCE CONTRIBUTIONS

 A. Iturmendi, S. Theis, O. Brüggemann, U. Monkowius, I. Teasdale. Triggered degradation of inorganic polymers, oral presentation at Danube Vltava Sava Polymer Meeting (DVSPM), Vienna, September 2017

 A. Iturmendi, D. Maderegger, O. Brüggemann, U. Monkowius, I. Teasdale. Polyphosphazenes with controlled degradation pathways, oral presentation at 3rd European Conference on Smart Inorganic Polymers (EUSIPs), Porto, September 2016

 A. Iturmendi, U. Monkowius, O. Brüggemann, I. Teasdale. Triggered degradation of polyphosphazenes, oral presentation at Career in Polymers VIII, Prague, July 2016

158

 A. Iturmendi, H. Henke, O. Brüggemann, I. Teasdale. Polyphosphazenes with novel architectures and self-immolative degradation, oral presentation at 14th Pacific Polymer Conference, Kauai, December 2015

 A. Iturmendi, S. Wilfert, T. B. Aigner, O. Brüggemann, I. Teasdale. Glycicne- based photo-crosslinkable polyphosphazenes as biodegradable scaffolds for tissue engineering, oral presentation at Bratislava Young Polymer Scientist Workshop Bypos, Zázrivá, June 2014

 A. Iturmendi, S. Wilfert, H. Henke, O. Brüggemann, I. Teasdale. Thermoresponsive molecular brushes based on polyphosphazenes, poster presentation at the conference Advances in Polymer Sciences and Technology, Linz, September 2013

159

160