Exploring Amphiphilic PEGMA-Based Architectures as Nanoparticles

for Drug Delivery

Christian Porsch

Doctoral thesis

KTH Royal Institute of Technology, Stockholm 2015 Department of Fibre and Polymer Technology

Akademisk avhandling som med tillstånd av Kungliga Tekniska högskolan i Stockholm framläggs till offentlig granskning för avläggande av teknisk doktorsexamen fredagen den 2 oktober 2015, kl. 09:00 i sal F3, Lindstedtsvägen 26, KTH, Stockholm. Avhandlingen försvaras på engelska. Fakultetsopponent: Professor Steve P. Rannard från University of Liverpool (UK).

Copyright © 2015 Christian Porsch All rights reserved Paper I © 2011 The Royal Society of Chemistry Paper II © 2013 Wiley-VCH Verlag GmbH & Co. KGaA Paper III Paper IV

© 2014 American Chemical Society © 2015 American Chemical Society

Cover page image: © Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission. Porsch, C.; Zhang, Y.; Östlund, Å.; Damberg, P.; Ducani, C.; Malmström, E.; Nyström, A.M., Inside front cover. Drug Delivery: In Vitro Evaluation of Non-Protein Adsorbing Breast Cancer Theranostics Based on 19F-Polymer Containing Nanoparticles, Part. Part. Syst. Charact. 2013, 30 (4), 300. TRITA-CHE Report 2015:32 ISSN 1654-1081 ISBN 978-91-7595-630-5

To Sofia

ABSTRACT Within the last decades, the stated potential of polymer constructs as drug delivery systems have challenged researchers to develop sophisticated polymers with tunable properties. The versatility of polymers makes them highly attractive to tailor nanoparticles (NPs) which fulfill the demands of effective drug delivery systems (DDS). The aim of this work was to design and synthesize amphiphilic ethylene glycol methacrylate-based (EGMA) macromolecules, and explore their potential as NPs for drug delivery.

Initially, a study of the controlled synthesis and solution properties of linear EGMA polymers, as well as the potential to transfer their behavior to amphiphilic comb copolymers, was conducted. Well-controlled polymers with interesting tunable thermo-responsive properties were accomplished by altering the monomer feed ratio. Furthermore, the comb copolymers formed self-assembled core-shell type structures in aqueous solution.

A library of amphiphilic fluorinated polymers was successfully established to explore the potential of EGMA-based polymers in a dual-functional theranostic delivery system. The non-toxic polymers self-assembled into small “stealthy” NPs, and the combination of fluorinated segments with EGMA segments allowed for detection by 19F-MRI with good imaging properties. The hydrophobic core of the NPs was capable to encapsulate and release an anti-cancer therapeutic, and effectively reduced the viability of three different cancer cell lines. The diffusion-controlled release kinetics of the drug from the NPs interestingly depended on the nature of the core moiety.

To reduce issues with instability of self-assembling NP systems the possibility to synthesize amphiphilic hyperbranched dendritic-linear polymers (HBDLPs) was investigated. Their three-dimensional structure was hypothesized to facilitate stabilization as unimolecular micelles. The architecture, hydrophilic/hydrophobic ratio, and high molecular weight showed to be crucial to avoid polymer association and stabilize the HBDLPs individually. In addition, the hyperbranched core of the HBDLPs was readily functionalized with disulfide bonds, either in the backbone or in the pendant groups. Under reductive conditions, selective cleavage of the disulfides thereby enabled either significant molecular weight reduction, or allowed for triggered release of a covalently bound dye, mimicking a drug. Potentially, such HBDLPs could be stable during circulation, while allowing for selective degradation and/or therapeutic release upon delivery to a cancer tissue.

SAMMANFATTNING Under de senaste decennierna har potentialen hos polymerer som bärare för kontrollerad läkemedelsleverans utmanat forskare till att utveckla polymerer med specifika egenskaper. Polymerernas stora variationsmöjligheter gör dem attraktiva för att skräddarsy nanopartiklar (NP) som tillgodoser kraven på ett system för effektiv läkemedelsleverans. Syftet med detta arbete var att designa och syntetisera etylenglykolmetakrylat-baserade (EGMA) makromolekyler, samt att undersöka dess potential som leveranssystem för läkemedel.

Inledningsvis studerades möjligheten att kontrollerat syntetisera amfifila, linjära polymerer och kampolymerer baserade på EGMA-monomerer och dessa polymerers egenskaper i vattenlösning utvärderades. Välkontrollerade polymerer syntetiserades framgångsrikt och genom att variera den satsade monomersammansättningen kunde polymerer med förutsägbara temperatur-responsiva egenskaper erhållas. I vattenlösning bildade de amfifila kampolymererna självorganiserande strukturer med uppdelad morfologi med en hydrofob inre kärna och ett hydrofilt yttre skal.

Ett bibliotek av amfifila, fluorinerade polymerer syntetiserades framgångsrikt för att utvärdera potentialen hos EGMA-baserade polymerer i ett teranostiskt leveranssystem. De icke-toxiska polymererna bildade små självorganiserande NP som uppvisade motstånd mot adsorption av proteiner på ytan. Kombinationen av fluorerade segment med EGMA uppvisade bra diagnostisk potential med 19F-magnetkamera. Dessutom kunde ett cancerläkemedel inkapslas och frisläppas från NPs hydrofoba kärna och den diffusionsberoende frisättningshastigheten berodde på typen av hydrofob komponent.

För att minska problem med instabilitet hos självorganiserande NP undersöktes möjligheten att syntetisera amfifila, hyperförgrenade dendritisk-linjära polymerer (HBDLP) med hypotesen att de kan stabiliseras som unimolekylära miceller. Polymerarkitekturen, hydrofil/hydrofob balans och hög molekylvikt visade sig viktigt för att stabilisera polymererna individuellt. Dessutom funktionaliserades polymerernas hyperförgrenade kärna med disulfidbindningar, antingen i huvudkedjan eller i polymerens sidogrupper. Under reduktiva betingelser bröts dessa disulfider selektivt och gav upphov antingen till betydande minskning i molekylvikt, eller till frisläppning av ett kovalent bundet färgämne som modell för ett läkemedel. Sådana HBDLP kan förbli stabila under cirkulation i kroppen och samtidigt tillåta selektiv nedbrytning och/eller frisättning vid leverans till tumörvävnad.

LIST OF PAPERS This thesis is a summary of the following papers:

I. “Thermo-responsive cellulose-based architectures: tailoring LCST using poly(ethylene glycol) methacrylates”, Christian Porsch, Susanne Hansson, Niklas Nordgren, Eva Malmström, Polymer Chemistry 2011, 2 (5), 1114-1123.

II. “In vitro evaluation of non-protein adsorbing breast cancer theranostics based on 19F-polymer containing nanoparticles”, Christian Porsch*, Yuning Zhang*, Åsa Östlund, Peter Damberg, Cosimo Ducani, Eva Malmström, Andreas M. Nyström, Particle and Particle Systems Characterization 2013, 30 (4), 381-390.

III. “Toward unimolecular micelles with tunable dimensions using hyperbranched dendritic-linear polymers”, Christian Porsch, Yuning Zhang, Cosimo Ducani, Francisco Vilaplana, Lars Nordstierna, Andreas M. Nyström, Eva Malmström, Biomacromolecules 2014, 15 (6), 2235-2245.

IV. “Disulfide-functionalized unimolecular micelles as selective redox-responsive nanocarriers”, Christian Porsch, Yuning Zhang, Maria I. Montañez, Jani-Markus Malho, Mauri A. Kostiainen, Andreas M. Nyström, and Eva Malmström, Biomacromolecules 2015, DOI: 10.1021/acs.biomac.5b00809.

* The authors contributed equally to this work.

The appended papers were collaborations with the co-authors, and I did not contribute to all parts of the papers. My contributions to the appended papers are:

I. All of the synthetic work, most analyses, and writing the major part of the manuscript. Surface tension measurements were performed together with N. Nordgren.

II. All synthetic work, about half of the analyses, and writing approximately half of the manuscript. Cell tests were performed by Y. Zhang. TEM was performed together with C. Ducani. 19F-NMR analysis was performed together with Å. Östlund. 19F-MRI was performed by P. Damberg.

III. All of the synthetic work, major part of the analyses, and writing

the major part of the manuscript. TEM was performed by C. Ducani. SEC-MALLS was performed together with F. Vilaplana. Cell tests were performed by Y. Zhang. 1H-diffusion NMR was performed by L. Nordstierna.

IV. Almost all of the synthetic work, most analyses, and writing the major part of the manuscript. Triple detection SEC was performed by P. Chambon. Cryo-TEM was performed by J-M. Malho.

Scientific contributions not included in this thesis:

V. “Thermo-responsive nanofibrillated cellulose by polyelectrolyte adsorption”, Emma Larsson, Carmen Cobo Sanchez, Christian Porsch, Erdem Karabulut, Lars Wågberg, Anna Carlmark, European Polymer Journal 2013, 49 (9), 2689-2696.

VI. “Histamine-functionalized copolymer micelles as a drug delivery

system in 2D and 3D models of breast cancer”, Yuning Zhang, Pontus Lundberg, Maren Diether, Christian Porsch, Caroline Janson, Nathaniel A. Lynd, Cosimo Ducani, Michael Malkoch, Eva Malmström, Craig Hawker, Andreas M. Nyström, Journal of Materials Chemistry B 2015, 3 (12), 2472-2486.

ABBREVIATIONS ABx-type Monomer with one A-functionality and x B-

functionalities AB*-type Inimer with one vinyl functionality and one initiating

group ARGET Activators regenerated by electron transfer AsAc Ascorbic acid ATRA Atom transfer radical addition ATRP Atom transfer radical polymerization AzSSMA 2-((2-(Methacryloyloxy)ethyl)disulfanyl)ethyl 6-azidohexanoate BBEMA 2-(2-Bromoisobutyryloxy)ethyl methacrylate BiB α-bromoisobutyryl bromide Bipy 2,2-Bipyridyl Bis-MPA 2,2-Bis(methylol)propionic acid BSSMA 2-((2-((2-Bromo-2-methylpropanoyl)oxy)ethyl) disulfanyl)ethyl methacrylate CH2Cl2 Dichloromethane CMC Critical micelle concentration CRP Controlled radical polymerization Cu(I)Br Copper(I) bromine Cu(II)Br2 Copper(II) bromine Cu(I)Cl Copper(I) chloride Cu(II)Cl2 Copper(II) chloride Đ Molar mass dispersity DAPI 4’,6-Diamidino-2-phenylindole DDS Drug delivery system DEGMA Di(ethylene glycol) methyl ether methacrylate Di-TMP Ditrimethylolpropane DLS Dynamic light scattering DMAP 4-(Dimethylamino) pyridine DMEM Dulbecco’s modified eagle medium DMF N,N-Dimethylformamide DOX Doxorubicin DR Disperse red DS Degree of substitution DTT Dithiothreitol EDTA Ethylenediaminetetraacetic acid EGMA Ethylene glycol methacrylate EPR Enhanced permeation and retention EtOAc Ethyl acetate EtOH Ethanol

FACS Flow-activated cell sorting FITC-MA Fluorescein O-methacrylate FRP Free radical polymerization FT-IR Fourier transform infrared (spectroscopy) GPC Gel permeation chromatography GSH Glutathione HA Hexyl acrylate HBDLP Hyperbranched dendritic-linear polymer HBMI Hyperbranched macroinitiator HPC Hydroxypropyl cellulose ICAR Initiators for continuous activator regeneration kact Rate constant of activation kdeact Rate constant of deactivation kp Rate constant of propagation kt Rate constant of termination LCST Lower critical solution temperature Mn Number average molecular weight Mw Weight average molecular weight MALDI-TOF MS Matrix-assisted laser desorption/ionization time of

flight mass spectrometry MALLS Multi-angle laser light scattering MRI Magnetic resonance imaging M Monomer MS Molar substitution MTT 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium

bromide MWCO Molecular weight cut-off NaAsc Sodium ascorbate NMP Nitroxide-mediated polymerization NMR Nuclear magnetic resonance NMR-d Nuclear magnetic resonance self-diffusion NP Nanoparticle OEGMA300 Oligo(ethylene glycol) methyl ether methacrylate (av.

Mn=300 g mol-1) OEGMA475 Oligo(ethylene glycol) methyl ether methacrylate (av.

Mn=475 g mol-1) PBS Phosphate buffered saline PEG Poly(ethylene glycol) PEGMA Poly(ethylene glycol) methacrylate PISA Polymer-induced self-assembly PNIPAAm Poly(N-isopropylacrylamide) PRE Persistent radical effect pyr Pyridine RAFT Reversible addition-fragmentation chain-transfer

RALLS Right angle laser light scattering RDRP Reversible-deactivation radical polymerization RI Refractive index SCVCP Self-condensing vinyl copolymerization SCVP Self-condensing vinyl polymerization SEC Size exclusion chromatography SPION Superparamagnetic iron oxide nanoparticle T1 Longitudinal relaxation time T2 Transversal relaxation time TBBPE 1,1,1–Tris(4-(2 bromoisobutyryloxy)phenyl)ethane TEA Triethylamine TEM Transmission electron microscopy TFA Trifluoroacetic acid TFEMA Trifluoroethyl methacrylate THF Tetrahydrofuran

TABLE OF CONTENTS

1. PURPOSE OF THE STUDY ................................................... 1

2. INTRODUCTION ................................................................. 2

2.1 Polymers ...................................................................... 2

2.2 Reversible-deactivation radical polymerization (RDRP) techniques ........................................................................... 3

2.2.1 Atom transfer radical polymerization (ATRP) ......................... 3

2.2.1.1 The ATRP components and mechanism ............................. 5

2.2.2 Activators regenerated by electron transfer (ARGET) ATRP .. 7

2.3 Dendritic polymers ...................................................... 7

2.3.1 Dendrimers ................................................................................ 8

2.3.2 Hyperbranched polymers ......................................................... 9

2.3.2.1 Self-condensing vinyl polymerization (SCVP) .................... 9

2.3.3 Hyperbranched dendritic-linear polymers............................. 10

2.4 Responsive polymers .................................................. 11

2.4.1 Poly(ethylene glycol) methacrylates (PEGMA) .......................12

2.5 Polymer micelles ........................................................ 13

2.5.1 Polymer micelles as anticancer drug delivery systems ........... 14

2.5.1.1 Solid tumor targeting ......................................................... 16

2.5.1.2 PEGylation .......................................................................... 17

2.5.2 Theranostic systems ................................................................. 17

2.6 Challenges .................................................................. 18

3. EXPERIMENTAL ............................................................... 19

3.1 Materials .................................................................... 19

3.2 Instrumentation and methods .................................... 20

3.2.1 Nanoparticle formation ........................................................... 22

3.2.2 Doxorubicin loading ............................................................... 22

3.2.3 In vitro DOX release ............................................................... 22

3.2.4 Cell viability tests .................................................................... 23

3.2.5 Reductive degradation of HBDLPs……………… . ………………….23

3.2.6 Disperse red release experiments ........................................... 23

3.3 Synthetic procedures ................................................. 24

3.3.1 Synthesis of hydrophobic compounds with ATRP-initiating moieties (papers I and II) .................................................................. 24

3.3.2 ATRP and ARGET ATRP from mono- or multifunctional initiators (papers I-IV) ....................................................................... 25

3.3.3 Synthesis of hyperbranched macroinitiators via SCVCP (papers III and IV) ............................................................................. 25

3.3.4 Post-functionalization via copper(I)-catalyzed azide-alkyne cycloaddition (paper IV) .................................................................... 26

4. RESULTS AND DISCUSSION ............................................. 27

4.1 Thermo-responsive poly(ethylene glycol) methacrylate (PEGMA) architectures (paper I) ....................................... 27

4.1.1 Synthesis of linear and hydroxypropyl cellulose-based comb copolymers of OEGMA and DEGMA ................................................ 27

4.1.2 Solution behavior of amphiphilic polymers ............................ 30

4.2 19F-containing polymer NPs as theranostic agents (paper II) .......................................................................... 32

4.2.1 Synthesis of 19F-containing amphiphilic polymers ................. 33

4.2.2 Nanoparticle formation and solution properties ................... 35

4.2.3 In vitro evaluation of theranostic function ............................ 37

4.2.4 Cell-based assessments ........................................................... 39

4.3 Amphiphilic hyperbranched dendritic-linear polymers (HBDLPs) as unimolecular micelles (papers III and IV) .... 42

4.3.1 Synthesis of HBDLPs (papers III and IV) ............................... 42

4.3.1.1 Synthesis of HBMIs by SCVCP (papers III and IV) ........... 42

4.3.1.2 Chain-extension of HBMIs to form HBDLPs (papers III and IV) ............................................................................................. 46

4.3.2 Evaluation of architectural effects to the NP properties (paper III) ...................................................................................................... 48

4.3.3 Formation and selectively triggered degradation of disulfide-containing NPs (paper IV) ................................................................. 50

4.3.4 Post-conjugation and triggered release from NPs (paper IV) 52

5. CONCLUSIONS ................................................................. 55

6. FUTURE WORK ................................................................ 57

7. ACKNOWLEDGEMENTS ................................................... 59

8. REFERENCES ................................................................... 62

PURPOSE OF THE STUDY

__________________________________________________

1

1. PURPOSE OF THE STUDY The rapid development within nanotechnology has encouraged the scientific community to design and manipulate polymers into nanosized materials with well-defined properties. Polymer formulations at the nanoscale have shown tremendous potential as delivery platforms to effectively treat various types of cancers. The extensive requirements of biocompatibility, non-toxicity, stability, degradability, and therapeutic efficacy of such delivery systems challenge researchers to design sophisticated macromolecules to meet these demands. The general purpose of this study was to design and synthesize amphiphilic polymer architectures based on ethylene glycol methacrylates (EGMA), and to evaluate their potential to serve as nanoparticle (NP) platforms to treat breast cancer. The aim was to investigate if poly(ethylene glycol) methacrylate (PEGMA) could be an alternative to conventional poly(ethylene glycol) (PEG) in drug delivery systems, due to its favorable versatility in terms of macromolecular design. The syntheses of complex polymer architectures were conducted by controlled radical polymerization techniques, and their ability to form stable core-shell type NPs, and treat breast cancer, were examined in vitro. In paper I, the synthesis and solution properties of linear and comb-shaped EGMA-based copolymers were investigated. This know-how was later applied in paper II, where the possibility to introduce fluorine-containing segments into PEGMA-based NPs was explored. The aim was to investigate if such copolymer architectures could enable multifunctional theranostic (therapy and diagnostic) function within the same NP platform. Papers III and IV were devoted to the synthesis of complex hyperbranched dendritic-linear polymers (HBDLP) to afford highly stable NPs. The purpose of these investigations was to reduce the instability of self-assembling systems by enable one high molecular weight polymer amphiphile to stabilize as a unimolecular micelle. Further, the possibility to functionalize such HBDLPs to allow for selective response to changes in reductive environment was investigated.

INTRODUCTION

__________________________________________________

2

2. INTRODUCTION

2.1 Polymers

Polymers are an essential part of our everyday life. In the early 20th century the first synthetic polymer, Bakelite, was commercialized.1 Ever since, the field of polymer science has been growing, aiding our constantly increasing demands on living standard. One of the major advantages of polymers is their great versatility in terms of designing opportunities. The polymer composition and architecture can be finely tailored, which may result in significantly altered material properties. By covalently linking building blocks, called monomers, in different combinations, several types of either linear or branched polymer architectures can be afforded (Figure 2-1). Monomers of the same type linked together become a homopolymer. By combining two or more monomers in different sequences in the same molecule, random, alternating, or block copolymers can be produced. Both the combination of multiple types of monomers, their sequence in the polymer, as well as the polymer architecture may result in interesting new chemical and/or physical properties.

Figure 2-1. Schematic illustration of different polymer compositions and architectures.

INTRODUCTION

__________________________________________________

3

2.2 Reversible-deactivation radical polymerization (RDRP) techniques

The most widely used technique for industrial production of polymers is free radical polymerization (FRP). FRP facilitates large scale synthesis due to the undemanding reaction conditions required, and is the most cost-efficient way to afford larger quantities of polymers for low-value products. However, unavoidable side reactions, such as radical-radical coupling or back-biting, related to high reactivity and the high amount of radicals present in the reaction, obstruct the use of FRP in more sophisticated products where better control of the final polymer properties are crucial. During the last two decades, controlled radical polymerization techniques which offer comprehensive control of the final polymers have been developed. Reversible-deactivation radical polymerization (RDRP), commonly referred to as controlled radical polymerization, techniques enable the synthesis of complex polymer architectures with a high degree of control over the molecular weight, molecular weight dispersity (Ð), and end-group functionality. The RDRP techniques can be divided into: atom transfer radical polymerization (ATRP)2-4, nitroxide-mediated polymerization (NMP)5, and reversible addition-fragmentation chain transfer (RAFT) polymerization6, 7. These techniques ensure a similar growth for each polymer chain, and the control is achieved by establishing a dynamic equilibrium between dormant and active propagating species (Scheme 2-1). By shifting the equilibrium to the dormant side, the concentration of active radicals is kept in a low and stationary level, and thereby, undesired termination and chain transfer reactions are suppressed.

Scheme 2-1. The general equilibrium present in RDRP techniques.

2.2.1 Atom transfer radical polymerization (ATRP)

ATRP was first developed in 1995, independently by Matyjaszewski and Wang2, 3, and Sawamoto4. The chemistry utilized in ATRP originates from the atom transfer radical addition (ATRA)8 reaction, which in turn

INTRODUCTION

__________________________________________________

4

is a development of the Kharash addition9 reaction. ATRP is widely studied and has had a significant impact on the design of polymeric materials with well-defined compositions, architectures, and functionalities.10 The general mechanism of ATRP (with Cu(I) as catalyst) is schematically described in Scheme 2-2. The control is achieved by a dynamic equilibrium between dormant species (R-X) and propagating radicals (R•). In order to suppress irreversible termination reactions of the highly reactive radical species, the equilibrium should be shifted to the dormant side to keep the concentration of radicals low. The equilibrium in ATRP is maintained by a reversible redox process, in which active radicals are formed when a halogen atom (X) is transferred from the dormant chain to a transition metal/ligand complex (Cu(I)-X/ligand). After this activation step, the produced radical (R•) can either propagate by addition of a monomer (M), or transfer back to the dormant, deactivated, state by reacting with the oxidized metal complex (Cu(II)-X2/ligand).

Scheme 2-2. The general mechanism of a copper-mediated ATRP, as well as of an ARGET ATRP (addition of dashed line) system.

Although the radical concentration is kept low, termination reactions through radical combination or disproportionation cannot be completely avoided, and still occur to a few percent in ATRP.11 Consequently, ATRP is not completely fulfilling the criteria of a true “living” system with complete end-group fidelity, and is usually termed a controlled/”living” polymerization technique. However, the high end-group fidelity obtained by ATRP, available for post-polymerizations, still makes it a powerful and versatile tool in the design of complex macromolecular architectures, such as: block, star, and comb copolymers.

INTRODUCTION

__________________________________________________

5

2.2.1.1 The ATRP components and mechanism As illustrated in Scheme 2-2, the components required in an ATRP system are: an initiator (R-X), a monomer (M), and a catalyst (transition metal + ligand). The true strength of ATRP is its great versatility and tolerance to a variety of different reaction conditions, initiators, monomers, catalysts, and solvents. However, in order to ensure well-controlled and reproducible polymerizations, the reaction parameters need to be optimized for the specific system.12 ATRP initiators are most commonly alkyl halides (R-X, X is usually Cl or Br), which can undergo rapid homolytic cleavage. The main prerequisites of the initiator are that it can stabilize the radical after cleavage, and that it promote a faster rate of initiation compared to the rate of propagation.12 The latter has shown to be favored by utilizing a halide exchange system with an initiator-bromine (R-Br) together with a chlorine-catalyst (Cu(I)Cl/L).13 The reason for this is that the C-Br bond is weaker than the C-Cl bond, which favors a fast initiation in relation to propagation. The catalyst is important in ATRP since it mediates the equilibrium between active and dormant species. Several transition metals (Fe, Ni, Ru, Mo, etc.12, 14) have shown to successfully mediate the ATRP reaction. However, Cu-based catalysts are by far the most utilized owing to their versatility and low cost.12, 15 The ligands used in Cu-catalyzed ATRP, whose role is to solubilize the metal ion, are often nitrogen-based and commercially available.12, 16, 17 Since the equilibrium in ATRP is shifted to the dormant side, termination reactions are highly suppressed, and the reaction mainly proceeds in two steps: initiation and propagation (Scheme 2-3). The initiation occurs by a homolytic cleavage of the carbon-halide bond in the initiator (R-X), followed by an abstraction of the halide by the activator (Cu(I)X/L), thus forming a carbon-centered radical (R•) and the deactivator (Cu(II)X2/L). It is important that the initiation is fast and quantitative in order to achieve good control of the polymerization. Further, it is crucial that the rate of deactivation (kdeact.) is higher than the rate of propagation (kp) to keep the equilibrium shifted toward the dormant species. Active radicals formed in the initiation are then propagating by addition of monomers (M) until they are reversibly deactivated by the Cu(II) complex. However, since termination reactions cannot be completely avoided (typically 5-10 %), the deactivating species (Cu(II)X2/L) will accumulate in the initial, non-stationary stage of the polymerization.11 This will shift the equilibrium toward the dormant side, and thereby lower the concentration of radicals. This phenomenon, known as the persistent radical effect (PRE), will self-regulate the ATRP equilibrium to lower the amount of termination events.18 Consequently,

INTRODUCTION

__________________________________________________

6

Scheme 2-3. (i)-(vi) The suggested ATRP mechanism in aprotic solvents (copper-mediated).

the control in the ATRP system can be improved by addition of a low amount of deactivator, typically 10 mol % with respect to the activator. ATRP can be performed in bulk, or in solution as a homogenous or heterogeneous system. Both polar and non-polar solvents can be utilized (water, alcohols, anisole, toluene, etc.), and the solvent should be chosen considering parameters such as: polymer solubility, chain transfer to solvent, and interactions with the catalytic system.12 ATRP further shows a good tolerance toward many functional groups, and a great variety of vinyl monomers can be polymerized, having in common that they have substituents that effectively aid stabilization of the propagating radical.19 Homopolymers of (meth)acrylates2-4, 20, 21, styrenes3, 22, 23, (meth)acrylamides24, 25, acrylonitriles26, 27, and others28, 29 have been successfully synthesized.

INTRODUCTION

__________________________________________________

7

2.2.2 Activators regenerated by electron transfer (ARGET) ATRP Even though conventional ATRP is a facile synthetic technique as previously discussed, it suffers from two serious drawbacks. The first issue with ATRP is the high sensitivity of the activator towards oxygen. In order to avoid termination of propagating radicals or chain transfer, conventional ATRP systems need to be rigorously deoxygenated, which significantly complicates the reaction procedure. The second disadvantage is the relatively high amount of transition metal catalyst needed. Copper, which is effective and by far the mostly employed catalyst in ATRP, is a heavy metal. From an environmental point of view there is therefore an urge to reduce the required copper amount, or allow for effective regeneration. In addition, residual copper may be an issue in many applications due to problems with toxicity and coloring, and therefore often requires excessive purification. Consequently, developments of the ATRP technique to function with reduced amount of, or even without any, metal catalysts30, have been developed. Proposed methods to reduce the amount of copper in ATRP, by employing continuous regeneration of the activator, include: activators regenerated by electron transfer (ARGET) ATRP31 and initiators for continuous activator regeneration (ICAR) ATRP32. ARGET ATRP, first reported by Matyjaszewski et al.33, is a form of ATRP which employs an alternative approach to establish the equilibrium between dormant and active species. Instead of adding the transition metal complex in its active form (Cu(I)X), only the inactive form (Cu(II)X2) together with a reducing agent is utilized in the reaction (Scheme 2-2). During the reaction, the reducing agent continuously reduces the Cu(II) complex to Cu(I). Therefore, the required amount of copper is decreased to only a few ppm. An advantageous synergistic effect of the reducing agent is the significantly lowered sensitivity to oxygen of ARGET ATRP compared with conventional ATRP. Since the reducing agent continuously scavenges oxygen the reaction can be performed in moderate amounts of air, thus avoiding complicated deoxygenation steps. Commonly used reducing agents are ascorbic acid (AsAc) and tin(II) 2-ethylhexanoate (Sn(EH)2).33, 34

2.3 Dendritic polymers As illustrated in Figure 2-1, polymers are very versatile and can be synthesized in a variety of compositions and architectures. The properties of the polymer will be highly dependent of its structure, and new properties can be attained by altering the polymer architecture.35 The development of sophisticated synthetic protocols and controlled

INTRODUCTION

__________________________________________________

8

polymerization techniques has significantly enhanced the accessibility of advanced macromolecular architectures. Dendritic polymers are a family of highly branched polymers including: dendrimers, hyperbranched polymers, dendronized polymers, and dendrigrafts (Figure 2-2). More recently, additional advanced dendritic architectures like linear-dendritic hybrids36, or the recently proposed hyperbranched polydendrons37, 38 have emerged. In the mid 1950s, Flory theoretically described the polymerization of monomers having more than one functionality, typically ABx-type (A and B are different functionalities, x ≥ 2), to yield highly branched, soluble macromolecules.39 Compared with their linear analogues, the branched structure of dendritic polymers results in interesting differences in chemical and physical properties, such as: low intrinsic viscosity, high solubility, and multivalency.35 Dendrimers are perfectly branched, monodisperse macromolecules, while the other three subgroups to various extents are less controlled in their branching.

Figure 2-2. Schematic illustration of the sub-groups in the dendritic polymer family.40

2.3.1 Dendrimers Structurally, dendrimers are ideal macromolecules, with a globular shape and high number of functional groups.41, 42 Generally, branched layers (denoted generations) of ABx-type building blocks are linked to a multifunctional core by either a divergent43, 44 or a convergent45 synthetic approach. Consequently, the number of functional groups in the outer part of the dendrimer is growing exponentially with each generation, resulting in a beneficial multivalency. However, although the perfect branching of dendrimers yield interesting properties, it also requires that each reaction step is well controlled, and therefore the synthesis is

INTRODUCTION

__________________________________________________

9

complicated and time consuming. The perfect branching in dendrimers further cause steric issues, which restrict their growth at higher generations.

2.3.2 Hyperbranched polymers

Hyperbranched polymers are highly branched, structurally less perfect analogues of dendrimers.46-50 In contrast to the iterative multistep synthesis of dendrimers, hyperbranched polymers are traditionally synthesized in one-pot procedures. Consequently, this limits the control over molecular weight, dispersity, and functionality, thus resulting in an irregular branched structure that consists of both dendritic units (all B functions reacted), linear units (one B function reacted), and terminal units (no B function reacted).48 The ratio of the different units can be described by the degree of branching (DB) as defined by Fréchet et al.51 Despite these structural imperfections, hyperbranched polymers still possess many of the advantages of dendrimers: high solubility, globular structure and a multitude of end-groups. The synthesis of hyperbranched polymers has been widely exploited and can be conducted by many different synthetic strategies, such as: condensation reactions, controlled radical polymerizations, cationic polymerization, and ring-opening polymerization.46, 48, 49, 52-54

2.3.2.1 Self-condensing vinyl polymerization (SCVP) Flory theoretically described the synthesis of highly branched polymers by polycondensation already in the middle of the 20th century.39 However, it was not until 1995 before Fréchet et al. demonstrated the first chain-wise polymerization approach to hyperbranched polymers, called self-condensing vinyl polymerization (SCVP).55 This initial work employed a cationic reaction process, however; similar procedures adapted for radical polymerizations were shortly after reported by Hawker et al.56 and Matyjaszewski et al.57 The general principle of SCVP is to use initiator-functional monomers, so called “inimers”, to yield a branched macromolecule (Scheme 2-4). Inimers are of an AB*-type structure where A is a vinyl group, and B is a pendant group which can be converted into an initiating specie (B*) by external activation (I). Subsequently, such inimers can react both as in a classical radical polymerization and as in a polycondensation reaction to form hyperbranched macromolecules (II).58 Consequently, high molecular weight polymers are mainly obtained at higher conversions in the self-condensing step.55

INTRODUCTION

__________________________________________________

10

Scheme 2-4. Schematic description of the mechanism of self-condensing vinyl polymerization (SCVP) for the synthesis of hyperbranched polymers.55 A represents a vinyl group, B a moiety that can be converted to an initiating specie, and * represents a reactive site which can initiate a vinyl polymerization.

The strength of SCVP is the versatile, one-pot synthesis of hyperbranched polymers with a high amount of reactive chain-ends, starting from commercially available vinyl monomers without severe gelation. Further, copolymerization of the inimer with a suitable vinyl-containing comonomer allows for introduction of functionality to the hyperbranched structure. In addition, efforts in decreasing the molecular weight distributions of hyperbranched polymers synthesized by SCVP have been reported. The molecular weight dispersities have effectively been lowered by the utilization of small amounts of multifunctional initiating moieties59-62, or by conducting the SCVP reaction in microemulsion systems63, 64.

2.3.3 Hyperbranched dendritic-linear polymers Polymers combining dendritic and linear segments are often termed dendritic-linear polymer hybrids. They are interesting since they comprise the unique properties of both the dendritic and the linear polymer. Previously, the most work of dendritic-linear polymers has been performed on the perfectly branched dendrons or dendrimers, in combination with linear chains.36, 65, 66 More recently, dendritic-linear hybrids with hyperbranched cores and linear arms have emerged as interesting alternatives.48, 67, 68 In the literature, such polymers are

INTRODUCTION

__________________________________________________

11

generally denoted as hyperbranched-core star polymers.48, 68 In this thesis they will be referred to as hyperbranched dendritic-linear polymers (HBDLPs). HBDLPs are typically synthesized either by post-functionalization of hyperbranched polymers with initiating moieties for chain extension, or by utilizing a controlled polymerization technique allowing for post-polymerization of the “living” chain-ends remaining from the synthesis of the hyperbranched core. By the design of HBDLPs with different suitable core/arm combinations, advanced macromolecular architectures with amphiphilic properties can be obtained.46, 67

2.4 Responsive polymers Researchers have for a long time turned to nature for inspiration of clever structure-to-property solutions. In this sense, responsive, or “smart”, polymers have attracted attention owing to their ability to change their properties as a response to external stimuli, for example: pH, temperature, ionic strength, redox potential, and light.69-71 Their ability to undergo substantial conformational changes in response to only small environmental alterations has shown to be attractive in many applications, for instance within biomedicine72-77 and for surfaces with switchable wettability78-82. Temperature-responsive polymers exhibit a phase transition in aqueous solution as a response to changes in temperature, switching from a hydrophilic to a hydrophobic character. Depending on the polymer characteristics, this transition occurs either upon heating, at the lower critical solution temperature (LCST), or upon cooling, at the upper critical solution temperature (UCST). Below LCST, water-polymer hydrogen bonds ensure solubility of the polymer (Figure 2-3). However,

Figure 2-3. Schematic illustration of the phase transitions of thermo-responsive polymers, switching from being soluble (below LCST) to insoluble aggregates (above LCST) in aqueous solution.

INTRODUCTION

__________________________________________________

12

at LCST polymer-polymer interactions become more thermodynamically favorable and the polymers collapse, aggregate, and precipitate. The most studied thermo-responsive polymer is poly(N-isopropylacrylamide) (PNIPAAm), mainly since its LCST (32 ˚C) is close to the physiological temperature and relatively insensitive to environmental conditions.83 However, reported issues with cytotoxicity may obstruct the use of PNIPAAm in many biomedical applications.84 Accordingly, other families of thermo-responsive polymers have been proposed.85-87

2.4.1 Poly(ethylene glycol) methacrylates (PEGMA) PEGMAs is a versatile group of polymers which can be synthesized from a multitude of commercially available monomers (Figure 2-4).88 Depending on the choice of monomer, the properties of the final polymer can vary from being insoluble in water, displaying thermo-responsive behavior, to highly hydrophilic macromolecules.89 This is an effect of changing the number of ethylene glycol segments in the side chain of the monomer, which will greatly vary the hydrophilic/hydrophobic ratio of the polymer. With few ethylene glycol units (approx. 2-8) the polymers are thermo-responsive exhibiting a LCST from 26-90 ˚C, while longer side chains renders very water-soluble polymers that may consist of up to 85 wt % of ethylene glycol segments.90

Figure 2-4. Representation of different monomers in the ethylene glycol methacrylate (EGMA) family. 4/5 and 8/9 indicates average numbers of ethylene glycol segments, as stated by the supplier.

The developments in RDRP techniques have facilitated well-controlled polymers, and thereby increased the accuracy and reproducibility of the LCST transitions. This has further simplified the controlled copolymerization of different EGMAs, which allows for tailoring of their LCST, for instance to relevant physiological temperatures.91, 92 In combination with their anticipated non-toxicity and biocompatibility, this has resulted in an increasing interest for applications within biomedicine.90, 93, 94

INTRODUCTION

__________________________________________________

13

2.5 Polymer micelles Polymers displaying both a hydrophilic (water-attracting) and a hydrophobic (water-repelling) character are referred to as amphiphilic polymers. Owing to the last decade’s development in RDRP techniques, well-defined, amphiphilic block copolymers of almost infinite variations in composition and architecture (diblock, triblock, star etc.) can be realized. In resemblance to low molecular weight surfactants, amphiphilic block copolymers with immiscible segments, causes a microphase separation upon dissolution in a solvent that is selectively good for one of the blocks. The insoluble blocks will collapse and associate, thus forming physically assembled structures stabilized by the soluble blocks. Depending on the composition, molecular weight, hydrophilic/hydrophobic balance, and architecture of the polymer, several different sophisticated morphologies, i.e. spherical micelles, cylindrical micelles, worms, polymersomes etc. in the nanometer range have been demonstrated.95-99 Polymeric diblock amphiphiles, with a high proportion of the hydrophilic block, typically self-assemble into spherical micelles in aqueous solution, as illustrated in Figure 2-5.95, 97, 100 Due to the two distinct immiscible segments, covalently connected, the self-assembled micelles typically form a so called core-shell architecture with two inherent discrete phases. The insoluble block of the polymers are forming a hydrophobic core, which is solubilized in the solvent by the hydrophilic blocks, forming a protecting, stabilizing corona, of the structure.

Figure 2-5. Schematic illustration of amphiphilic polymers in solution (below CMC) and their spontaneous self-assembly to polymer micelles (above CMC).

INTRODUCTION

__________________________________________________

14

This self-directed association is highly concentration dependent, and the critical concentration at which the first micelle is formed is referred to as the critical micelle concentration (CMC). Above CMC, the associated micellar structure is remarkably stable. However, it is always in equilibrium with its corresponding amphiphile, and there is a constant and rapid chain exchange. Dilution to a concentration below CMC will cause micellar destabilization, and only individual polymer chains will be present. However, owing to the high molecular weight of polymer amphiphiles, compared to conventional low molecular weight surfactants, considerably lower CMC-values is attained, resulting in stable assemblies over a much wider concentration range. Furthermore, the formation of micelles in aqueous solution from amphiphilic polymers can be performed with different procedures depending of the solubility of the polymer.101, 102 Polymers with high water solubility are commonly prepared by direct dissolution. Less soluble polymers are typically dissolved in a good solvent for both blocks, followed by addition of water and slow removal of the good solvent by evaporation or dialysis. Concerns about micellar destabilization, caused by dilution or altered solution conditions, have engaged researchers in designing molecules that are individually stabilized in solution. This concept, of a single molecule with micelle-like properties, was proposed by Newkome et al.44,

103, 104 and is commonly referred to as a “unimolecular micelle”. Typically, star-shaped, highly branched, or post-stabilized polymer amphiphiles of high molecular weight have been proposed to form individual unimolecular micelles in the nanometer range.105-108 Furthermore, a high number of stabilizing arms that form a dense layer around the hydrophobic core of such micelles, have been proposed to reduce multimolecular aggregation.109

2.5.1 Polymer micelles as anticancer drug delivery systems Cancer is one of the leading causes of death world-wide, and the large variety of cancers, as well as their propensity to metastase, complicates treatment. Patients treated with chemotherapy are usually given highly toxic drugs by intravenous injection. One major issue is that the chemotherapeutics available are almost exclusively hydrophobic, low molecular weight compounds, with poor solubility in aqueous solution, which considerably obstructs their distribution in the blood stream. Furthermore, they generally affect all cells in the body, even the healthy ones, thus causing severe toxic response. Consequently, within the last two decades, there has been an increasing interest in utilizing nanotechnology to develop polymer NPs as drug delivery systems (DDS) to effectively assist treatment of numerous cancers.110-115 Polymer

INTRODUCTION

__________________________________________________

15

micelles hold a variety of highly attractive properties as DDS, including: their great versatility, core-shell morphology, and their favorable sizes in the nanometer range. Synthetic progress has further supported the development of polymer micelles, since it endows researchers world-wide with a substantial designer’s freedom in synthesizing well-controlled polymers of a great variety of architectures. To date, several polymer-based DDS have been approved by FDA for clinical use or are in late clinical trials, such as PEGylated liposomal doxorubicin (Doxil116) and targeted docetaxel NPs (BIND-014117). Figure 2-6 illustrates some attractive features of a polymer micelle DDS, of which some are further discussed below.

Figure 2-6. Illustration of some of the components and attractive features of a polymer micelle DDS.

Polymer DDS can effectively overcome many of the disadvantages of conventional chemotherapy. The core-shell architecture of polymer micelles has beneficially shown to solubilize low solubility therapeutics in their hydrophobic core, thus increasing their bioavailability.118 In addition, the encapsulation of the drug within the core of the micelle, prevents undesired toxicity during delivery, as well as minimizes the risk

INTRODUCTION

__________________________________________________

16

of drug degradation in vivo. Furthermore, since the drug is physically loaded in the micelle core, it can be hypothesized that several different hydrophobic drugs can be delivered employing the same polymer platform, ultimately simultaneously. In addition, polymer DDS have shown to favorably modify the biodistribution of a drug. The possibility to tailor the size and surface characteristics of a polymer micelle is highly beneficial to control its fate in vivo. Xenobiotics are typically removed from the body by excretion through the kidneys, having an approximate threshold pore size of 4 x 14 nm.110, 119 Consequently, the excretion of larger objects, like high molecular weight polymer micelles, is significantly retarded. This prolong the circulation time of polymer micelles in the blood stream, which is an advantage for passive solid tumor targeting by the enhanced permeation and retention (EPR) effect (discussed below).120 The rate of elimination of polymers has further been shown to be highly dependent on the polymer chemistry, flexibility, and shape.119 Likewise, large entities, positively charged NPs or hydrophobic particles are mainly cleared via the hepatic system.121, 122 Another advantage of polymer DDS is that the drug pharmacokinetics can be altered. The drug release from the micelle can either be diffusion-controlled or selectively triggered. This results in extended and controlled release kinetics, which means that a higher dose can be administered with reduced risk for toxic side effects. The diffusion-controlled release can for instance be tailored by the interaction between the drug and the hydrophobic core, or by the overall hydrophilic-hydrophobic balance in the micelle. Selectively triggered release utilizes cleavage of covalently bound drugs by changes in tumor-specific conditions, such as: pH or redox potential.74, 123-125

2.5.1.1 Solid tumor targeting One of the key motivations for the interest in polymer micelles as DDS is their passive targeting to solid tumors by the EPR effect. In 1986, Maeda et al.126 reported that macromolecules (proteins of different sizes) favorably accumulate to a higher degree in tumor tissue than in healthy tissue, which since has been a widely investigated tool in cancer therapy.115, 127-131 Fast growing tumors have an irregular and leaky vasculature, which make them more permeable to macromolecules than healthy tissue.132 In addition, they have a partly defective lymphatic drainage, thus obstructing removal of larger particles to a high extent. Larger particles are therefore retained to a greater extent, while low molecular weight substances are returned more effectively to the blood stream.133 Consequently, the strategy to achieve effective DDS is to improve this selective accumulation by developing constructs with

INTRODUCTION

__________________________________________________

17

sufficiently long circulation times, thus resulting in more tumor passages and increased accumulation. Another approach to target solid tumors is by active targeting. One attractive method for this has been to decorate the exterior of the polymer micelle with groups with specific affinities for receptors overexpressed in tumor cells, such as: peptides, antibodies, or folic acid.134-138

2.5.1.2 PEGylation One important barrier to overcome for an effective DDS is to avoid removal from the blood stream by the reticuloendothelial system (RES).121, 139, 140 In the blood stream, unknown particles are subjected to severe adsorption of opsonin proteins, which cause rapid recognition and removal of the particles by macrophages of the RES. This significantly obstructs the particle biodistribution by reducing the blood circulation time. The predominant strategy to avoid this is by conjugation of hydrophilic polymers, typically poly(ethylene glycol) (PEG), to the surface of the particle.121, 141 The hydrophilicity, neutrality, and dynamic nature of PEG has shown to significantly reduce opsonin adsorption, which result in a so called “stealthy” layer of the NP.121 PEG is further known to be non-toxic and biocompatible.142 However, one reported concern is that PEGylated nanoconstructs has shown to cause hypersensitivity reactions in some patients by promoting activation of the complement (C) system.143 During the last decade, oligo(ethylene glycol) methacrylates (OEGMAs) have emerged as an interesting alternative to linear PEGylation.90, 144 Their feasibility to polymerize via controlled radical polymerization techniques offer a high degree of structural freedom as they can be copolymerized with other vinyl monomers to introduce additional functionality to the polymer.

2.5.2 Theranostic systems In the search for effective medicine, DDS are progressively equipped with diagnostic modalities. Such delivery formulations, combining therapy and diagnostics, are referred to as “theranostics”. The high potential of such systems lies within that complimentary to the chemotherapeutic effect, simultaneous information regarding tumor heterogeneity, metastasis, and therapeutic efficacy can be achieved.145, 146 During the last decade, theranostic nanoparticle platforms of various formulations and imaging modalities have been extensively reviewed.145-149 Magnetic resonance imaging (MRI) is a promising diagnostic technique to image

INTRODUCTION

__________________________________________________

18

soft tissue, due to its non-invasive nature and good spatial resolution. The natural 19F isotope has been suggested as a biocompatible alternative to typical MRI contrast agents, such as superparamagnetic iron oxide NPs (SPION) or Gd3+-chelates.150, 151 19F is promising since: it has high relative sensitivity (83 %) compared to 1H; good biocompatibility; and low occurrence in vivo, which results in favorably low background noise for imaging.151, 152 However, high resolution imaging using the 19F isotope requires high fluorine content and sufficient mobility, and efforts have been made to find a suitable chemical design of fluorinated polymer NPs to ensure good relaxation properties.153-158

2.6 Challenges The task to design and synthesize effective polymer-based DDS is as intriguing as it is challenging. It is a wide-spread interdisciplinary research area, which occupies a vast number of researchers every day. The search for effective treatments requires careful balance between polymer structural control, function, and applicability, combined with the risk/reward mentality that pervades medicine. During the last 25 years an extensive know-how has been developed in this field, resulting in a number of successful systems reaching far into clinical trials. However, many challenges still remain in order to realize fully personalized medicine. The increase in demands on sophisticated polymer platforms with a high order of function and complexity requires the development of simple but controlled synthetic protocols. The more complex synthesis sometimes needs compromises in terms of control to successfully screen and evaluate end function. This thesis contributes with understanding of how advanced macromolecular design can be employed to target some of the main prerequisites of polymer nanoparticle DDS.

EXPERIMENTAL

__________________________________________________

19

3. EXPERIMENTAL

This section will give a brief overview of the materials, instrumentation, and synthetic procedures utilized in this thesis. More detailed information can be found in the appended papers.

3.1 Materials Hydroxypropyl cellulose (HPC, Mw = 80 000 g mol-1, Mn = 10 000 g mol-1 according to manufacturer (Sigma-Aldrich); Mw = 74 000 g mol-1, Mn = 34 000 g mol-1 were obtained using SEC with THF as a mobile phase and polystyrene standards; molar substitution of propoxy groups (MS) was determined to be 2.9 using 1H-NMR. 2,2-Bis(methylol)propionic acid (bis-MPA) and ditrimethylolpropane (diTMP) were kindly donated by Perstorp AB, and used as received. The monomers di(ethylene glycol) methyl ether methacrylate (DEGMA, Mw = 188 g mol-1), oligo(ethylene glycol) methyl ether methacrylate (OEGMA, Mw = 300 g mol-1 and Mw = 475 g mol-1), tert-butyl methacrylate (t-BMA, Mw 142 g mol-1), hexyl acrylate (HA, Mw = 130 g mol−1) were purchased from Sigma-Aldrich and were prior to use activated by passing through a column of neutral aluminum oxide. Trifluoroethyl methacrylate (TFEMA, Mw = 168 g mol-1) and fluorescein O-methacrylate (FITC-MA, Mw = 400 g mol-1) were purchased from Sigma-Aldrich and used as received. Human breast cancer cell lines; MDA-MB-231, MDA-MB-468 and MCF-7 cells were purchased from the American Type Culture Collection (ATCC). Mouse macrophage RAW 246.7 cell line was kindly supplied by the group of Professor Agneta Richter-Dahlfors at The Swedish Medical Nanoscience Center, Karolinska Institutet, Sweden. The cell line was maintained in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10 % (v/v) fetal bovine serum (FBS), 100 U/mL penicillin, and 100 mg mL-1 streptomycin, and incubated at 37 °C with 5 % CO2. Fetal bovine serum (FBS), penicillin solution and phosphate buffered saline were obtained from Hyclone Laboratories. 4’,6-Diamidino-2-phenylindole (DAPI), propidium iodide (PI), and Annexin V were obtained from Molecular Probes. All other chemicals and solvents were purchased from conventional commercial suppliers (Sigma-Aldrich, Merck, Acros Organics, VWR, Chemtronica) and used as received.

EXPERIMENTAL

__________________________________________________

20

3.2 Instrumentation and methods 1H and 13C NMR for structure analysis were recorded on a Bruker Avance AM 400 NMR instrument using CDCl3, DMSO-d6, MeOD, or D2O as solvent. 1H NMR was acquired at 400 MHz and 13C NMR at 100 MHz. The residual solvent peak was used as an internal standard. MALDI-TOF MS analyses were conducted on a Bruker UltraFlex MALDI-TOF mass spectrometer with a SCOUT-MTP ion source (Bruker Daltonics, Bremen) equipped with a nitrogen (N2) laser (337 nm), a gridless ion source, and reflector design. The instrument was calibrated using SpheriCal® calibrants obtained from Polymer Factory Sweden AB. THF solutions of either 9-nitroanthracene or dithranol were used as matrices with trifluoroacetic acid sodium salt added. FT-IR absorption spectra were collected by a Perkin-Elmer spectrum 2000 FT-IR, equipped with a MKII Golden Gate, single reflection ATR system from Specec Ltd. The ATR crystal was a MKII heated diamond 45-ATR top plate. Sixteen scans were recorded for each spectrum. SEC with THF (1.0 ml min-1) as the mobile phase was performed at 35 ˚C using a Viscotek TDA model 301 equipped with two GMHHR-M columns with TSK gel (mixed bed, Mw resolving range: 300–100 000 Da) from Tosoh Biosep, a VE 5200 GPC autosampler, a VE 1121 GPC solvent pump, and a VE 5710 GPC degasser (all from Viscotek Corp.). A calibration method was created using linear polystyrene standards. Corrections for the flow rate fluctuations were made using toluene as an internal standard. Viscotek OmniSEC version 4.0 software was used to process the data. SEC with DMF (0.2 mL min-1 with 0.01 M LiBr) as the mobile phase was performed at 50 °C on a TOSOH EcoSEC HLC-8320GPC system equipped with an EcoSEC RI detector and three columns (PSS PFG 5 µm; Microguard, 100 Å, and 300 Å) (MW resolving range: 100-300 000 Da) from PSS GmbH. A conventional calibration method was created using narrow linear poly(methyl methacrylate) standards. Corrections for flow rate fluctuations were made using toluene as an internal standard. PSS WinGPC Unity software version 7.2 was used to process data. SEC-MALLS with H2O (1 mL min-1 with 10 mM NaOH) as the mobile phase was performed at 40 °C by SECcurity 1260, PSS (Mainz, Germany) coupled to a multiple angle laser light scattering detector (MALLS, BIC MwA7000, Brookhaven Instrument Corp., New York) and a refractive index detector (SECcurity 1260, PSS. Mainz, Germany). Separations were performed using a combined column setup with a SUPREMA precolumn, a SUPREMA 1000 Å, and two SUPREMA 3000 Å analytical columns (PSS, Mainz, Germany). Calibration of the detectors and the column setup was performed by injection of pullulan standards (PSS,

EXPERIMENTAL

__________________________________________________

21

Mainz, Germany). dn/dc of the polymers was calculated by five different concentrations between 0.2 and 5 g L−1. Data collection and analysis from SEC separations with light scattering detection was performed using WinGPC software (PSS, Mainz, Germany). SEC triple detection with DMF (1 mL min-1 with 0.01 M LiBr) as the mobile phase was performed at 60 °C on a GPC triple detection equipped with a GPCmax VE2001 auto-sampler, two Viscotek D6000 columns (and a guard column) and a triple detector array TDA305 (refractive index, light scattering and viscometer). DLS was conducted with a Malvern Zetasizer NanoZS operating at 633 nm using deionized water or PBS as solvent. The samples were allowed to equilibrate for a minimum of 2 minutes prior to the measurement and analyzed at 25 °C or 37 °C. Generally, both unfiltered and filtered (0.45 µm nylon filter) samples were analyzed. Fluorescence spectroscopy for assessment of the critical micelle concentration by the pyrene probe technique159 was performed on a Varian Cary Eclipse collecting emission spectra using an excitation wavelength of 332 nm. Measurements were performed in PBS buffer solutions. Surface tension measurements for assessment of the critical micelle concentration were performed on a Thermo Cahn (Radian series 300) Microbalance, using the Wilhelmy plate method. Measurements were performed in PBS buffer solutions. 19F-MRI analyses were performed using an MR-400 scanner (Varian Inc, Yarnton, UK). Plastic syringes (d = 6 mm) containing the 19F contrast agent solutions (10 mg mL-1 in PBS) were placed in a fixed tune surface coil with a curved housing, designed for mouse head 19F MR (Rapid Biomed, Rimpar, Germany), and images were acquired using a gradient echo sequence with tr and te of 200 and 1.66 ms respectively and a flip angle of 20°. The matrix size was 64 x 64 covering a field of view of 48 x 48 mm2 for a slice thickness of 4 mm. TEM images were obtained with a FEI Morgagni 268(D) Transmission Electron Microscope at 80 kV. Samples were prepared by deposition of a polymer solution on a glow-discharged, carbon-coated Formvar grid (Electron Microscopy Sciences) and excess was removed by wicking with filter paper. The samples were then either analyzed directly or after staining with 2 % (w/v) aqueous uranyl formate solution. Cryo-TEM images were obtained using a Jeol’s 3200FSC cryo-transmission electron microscope operating at 300 kV in bright field mode with an Omega-type Zero-loss energy filter. NP dispersions were pipetted on to holey carbon copper grids under 100 % humidity and the excess amount of sample was blotted away with filter paper. Samples were subsequently plunged into -170°C ethane/propane mixture. The

EXPERIMENTAL

__________________________________________________

22

vitrification was done using Fei Vitrobot Mk3. A Gatan Ultrascan 4000 CCD camera was used to acquire the images. 19F NMR and 19F NMR diffusion measurements were performed at 37 (±0.1) C on a Bruker 600 MHz Avance spectrometer operating at 564 MHz for 19F, equipped with a Diff30 diffusion probe. 19F-NMR was employed to determine longitudinal (T1) and transverse (T2) relaxation times, and self-diffusion of the molecules for size determination in different media. For all studies, the different polymers were dissolved at 0.5 % (w/w) in the given solvents. H-NMR diffusion measurements were performed at 25 °C on a Bruker 600 MHz spectrometer, equipped with a diffusion probe with a maximum gradient strength of 1200 G/cm.

3.2.1 Nanoparticle formation In this thesis two different procedures have been employed to form polymer NPs in aqueous solution from the neat polymers. (1) The polymer was dissolved in CH2Cl2 (2 mL, 1.25 mg mL-1). Phosphate buffered saline (PBS, 2 mL) was added by a syringe pump over a time period of 2 hours. CH2Cl2 was evaporated during gentle stirring (100 rpm) over night (papers II and III). (2) The polymer was directly dissolved in PBS (2 mL, 1.25 mg mL-1) on a shaking device (250 rpm) for minimum 24 hours (papers III and IV).

3.2.2 Doxorubicin loading In this thesis doxorubicin (DOX) was used as a model drug (papers II-IV). A DOX stock solution was prepared by dissolving DOX and TEA (1:3 mol eq.) in CH2Cl2 (1 mg mL−1). The stock solution (1 mL) was diluted with CH2Cl2 (1 mL) and mixed with a PBS solution of NPs (4 mL, 1.25 mg mL−1). The mixture was slowly stirred overnight (100 rpm) in order to evaporate the organic solvent. Free DOX was removed by spin filtration with a MWCO of 3 kDa by refilling PBS every 5 min. The concentration of DOX in the NPs was measured by comparing UV absorbance at 490 nm, of samples diluted with DMF:H2O (4:1) to a standard curve (five replicates).

3.2.3 In vitro DOX release For the in vitro drug release, DOX-NPs (or free DOX as control) solutions were transferred into 3 mL dialysis cassettes (MWCO 3500, Slide-A-Lyzer®, Thermo) suspended in 4 L of PBS at 37 °C. Samples inside cassettes were collected at specific time intervals (triplicates of 10 μL each), and transferred into black 96-well plates, containing DMF:H2O

EXPERIMENTAL

__________________________________________________

23

(100 μL, 4:1) in each well. The amount of DOX was determined by the fluorescence intensity at the wavelength 480/600 nm (excitation/emission) compared to a standard curve.

3.2.4 Cell viability tests Cell viability tests were performed by an MTT assay (evaluate mitochondria function). In this thesis four different cell lines were utilized; breast cancer cell lines (MCF-7, MDA-MB-231, MDA-MB-468) and mouse macrophage cell line (RAW 264.7). Breast cancer cells, MCF-7, MDA-MB-231, and MDA-MB-468 were harvested by trypsin and RAW 264.7 via physical scraping. Cells were washed with PBS, resuspended in DMEM, and seeded into 96 well plates at a concentration of 1 × 104 cells per well (1 × 105 cell mL−1, 100 μL) and precultured for 24 h. For a typical test the cells were incubated (24, 48 or 72 h) in fresh medium containing the polymer samples at different concentrations (neat polymer NPs - 0.01, 0.1, 1, 10, and 100 μg mL−1, DOX-loaded NPs - 0.01, 0.1, 1, 5 and 10 μg mL−1 (designed drug concentration)). Five parallel wells were set for each concentration. After incubation, MTT (10 μL, 5 mg mL−1) was introduced into each well, and after additional 4 h, 100 μL of SDS solution (10 %) was added. Cell viability was evaluated by absorbance readings at 570 nm.

3.2.5 Reductive degradation of HBDLPs The HBDLPs were dissolved in PBS (1 mL, 15 mg mL-1) in glass vials for 24 h. To each vial, DTT or GSH stock solution was added to reach one of the following concentrations of reducing agent (DTT: 10 µM or 10 mM; GSH: 10 mM). Further, one control sample in pure PBS was performed. The solutions were shortly subjected to vacuum (1 min), purged with argon (10 min), and allowed to stir at room temperature covered with aluminum foil. 400 µL aliquots were withdrawn after 1 and 24 h, dried under reduced pressure, and analyzed by SEC (DMF).

3.2.6 Disperse red release experiments The in vitro release study was performed at room temperature in Slide-A-Lyzer® MINI dialysis tubes (Nordic Biolabs, MWCO 3.5K). NPs in PBS (0.5 mL, 5 mg mL-1) were dialyzed during shaking (250 rpm) against 14.5 mL of three different reducing media (DTT 10 µM and 10 mM, GSH 10 mM, in PBS) and against pure PBS as control. Aliquots from the sample compartment (triplicates) were collected at specific time intervals and the amount of released dye was determined with a plate reader (Tecan

EXPERIMENTAL

__________________________________________________

24

Infinite® 200 Pro) at 484 nm. The analyzed samples were added back to the dialysis tubes after the measurement and the test continued.

3.3 Synthetic procedures Below, general synthetic procedures employed in this thesis are described. For more experimental details, please approach the appended papers.

3.3.1 Synthesis of hydrophobic compounds with ATRP-initiating moieties (papers I and II)

In this thesis, a number of different hydrophobic compounds functionalized with α-bromoesters, available for ATRP were synthesized, of which some are presented in Scheme 3-1.

Scheme 3-1. Synthetic route to the hydrophobic ATRP initiators utilized in paper II.

EXPERIMENTAL

__________________________________________________

25

The reactions were performed by conventional base-catalyzed esterification reactions by employing either the corresponding anhydride or acid bromide of the bromoisobutyryl ATRP-initiator. Typically, the starting compound was dissolved in solvent together with a base (TEA or pyridine), cooled to 0 °C, and a dilute solution of the acid, or pure anhydride, was added. The reactions were performed in room temperature and purified by extraction with aqueous solutions (acidic/basic) and column chromatography when necessary.

3.3.2 ATRP and ARGET ATRP from mono- or multifunctional initiators (papers I-IV)

ATRP and ARGET ATRP were conducted from either mono- or multifunctional initiators. The molecular weight was targeted by the initial monomer-to-initiator ratio, and polymerizations initiated from multifunctional, high molecular weight compounds were performed in highly diluted systems to avoid undesired coupling. Typically, the initiator was completely dissolved in the solvent before the monomer/s, ligand, and reducing agent (only for ARGET ATRP) were added. The mixture was cooled to 0 °C and subjected to a degassing cycle (5 + 5 min vacuum/argon for ATRP and 5 min argon for ARGET ATRP). The copper catalyst system was added (Cu(I)Br/Cu(II)Br2 or Cu(I)Cl/Cu(II)Cl2) followed by two additional degassing cycles (ATRP: 2 x (5 + 5 min), ARGET: 15 min). Reactions were generally conducted at 40-60 ˚C and allowed to proceed to conversions below 50 %. The products were purified by passing through aluminum oxide, repeated precipitations, and sometimes by dialysis to aqueous EDTA solutions to remove residual traces of copper.

3.3.3 Synthesis of hyperbranched macroinitiators via SCVCP (papers III and IV)

Hyperbranched macroinitiators (HBMIs) were synthesized by employing SCVCP, controlled by an ATRP system. Highly branched structures were realized by utilizing a trifunctional ATRP-initiator in combination with an inimer and a comonomer. Molecular weights were varied by the trifunctional initiator-to-(inimer+comonomer) ratio. Different functionalities were incorporated to the hyperbranched polymers by utilizing inimers and comonomers with disulfide (BSSMA) and/or azide functionalities (AzSSMA), synthesized as described in Scheme 3-2. The polymerizations were conducted by an ATRP system as described above, but allowed to proceed to high conversions (≥95 %) to ensure self-condensation. The reactions were ended by addition of deactivator to maximize end-group fidelity for chain extension.

EXPERIMENTAL

__________________________________________________

26

Br

O

OSSHO

Cl

O

HO SS OH

OBr

O

OSSO

Br

O

Br

TEA, THF

N3OH

O

N N N N

O1. CH2Cl2

OHSSO

O

OSSO

O

ON3

BSSMA

AzSSMA

TEA, THF

2. CH2Cl2

Scheme 3-2. Synthetic routes to the disulfide-containing inimer BSSMA and the comonomer AzSSMA utilized in paper IV.

3.3.4 Post-functionalization via copper(I)-catalyzed azide-alkyne

cycloaddition (paper IV) In paper IV, a HBDLP with available azide moieties in the core was synthesized. The dye, disperse red (DR), was equipped with acetylene functionality and conjugated to the azide in a THF/H2O mixture, catalyzed by NaAsc and CuSO4 at 40 °C for 48 h. The product was passed through a neutral Al2O3 column to remove copper, and unbound dye was removed by precipitation into diethyl ether and subsequent dialysis against ethanol.

RESULTS AND DISCUSSION

__________________________________________________

27

4. RESULTS AND DISCUSSION

The progress within controlled polymerization techniques has attracted researchers world-wide to develop complex macromolecules with controlled composition and architecture to achieve novel materials with intriguing properties. The focus of this thesis has been to design and synthesize sophisticated amphiphilic polymer architectures, based on commercially available EGMAs, by employing controlled radical polymerization (CRP) techniques. In the first part, random copolymerization of EGMAs was evaluated as an approach to linear and comb-shaped polymers with tunable thermo-responsive behavior (paper I). In the second part, the aim was to design and synthesize amphiphilic polymers that form core-shell polymer micelles that effectively could serve as drug delivery systems (papers II - IV). Focus was on controlling the size, stability and functionality of the micelles, and their potential to effectively encapsulate and deliver cancer therapeutics. Throughout this thesis all micellar systems are based on organic amphiphilic polymers. Generally, such micelles are referred to as nanoparticles (NPs) since all studied systems mainly were found to have sizes between 1-1000 nm, and are particle-like in their morphology. However, in paper III and IV the term unimolecular micelle is utilized to denote micelle-like NPs consisting of only one polymer chain.

4.1 Thermo-responsive poly(ethylene glycol) methacrylate (PEGMA) architectures (paper I)

The polymer family of ethylene glycol methacrylates (EGMA) has shown thermo-responsive properties with LCSTs dependent of the number of ethylene glycol segments in the monomer side chain. In this study, the thermo-responsive behavior of polymers from the two monomers di(ethylene glycol) methyl ether methacrylate (DEGMA) and oligo(ethylene glycol) methyl ether methacrylate (OEGMA4/5, average Mn =300 g mol-1) was evaluated.

4.1.1 Synthesis of linear and hydroxypropyl cellulose-based comb copolymers of OEGMA and DEGMA

Linear homo- and copolymers of DEGMA and OEGMA were synthesized by employing ARGET ATRP in anisole (Scheme 4-1a). By varying the monomer feed ratio, polymers with varying compositions were obtained (Table 4-1). Final polymer compositions were assessed with 1H-NMR by comparing the methoxy protons with the methylenes in the ethylene

RESULTS AND DISCUSSION

__________________________________________________

28

glycol segments. Good agreement was found between monomer feed (f) and final (F) composition in the polymers. Conclusively, polymers with well-controlled compositions could be synthesized already at low-to-moderate conversions, thus indicating similar reaction kinetics of DEGMA and OEGMA. The polymerizations proceeded with reproducible kinetics and analysis of the final products by SEC showed reasonably narrow molecular weight distributions (Đ = 1.1-1.6).

Scheme 4-1. Synthetic route to a) linear (Table 4-1) b) comb shaped EGMA-based polymers (HPC0.6-g-PDEGMA, HPC0.6-g-POEGMA, and HPC0.6-g-P(OEGMA52-co-DEGMA48)). Comb copolymers with hydroxypropyl cellulose (HPC) backbone and EGMA-based combs were synthesized in two steps (Scheme 4-1b). Initially, hydroxypropyl cellulose (HPC, molar substitution of propoxy groups determined to 2.9 by 1H NMR) was equipped with ATRP initiators by esterification utilizing the corresponding acid bromide (BiB, in CH2Cl2) or anhydride (in DMF). The successful functionalization by both approaches was confirmed by 1H NMR and FT-IR spectroscopies.

O

O

HO O

O

n

O

Ox

OHx

O

O

HO O

O

n

O

H

OHx

O

O

HO O

O

n

O

Ox

OHx

Br

O

O

O

2

O

O

O

4/5

a

b

Brc

O

OBr Br

O

O

BrBr

(ii)(iii)

O

O

O

4/5

O

O

O

2

(iv)

(i) = AsAc, PMDETA, CuBr2, anisole (40 °C)(ii) = anh. CH2Cl2 (rt)(iii) = DMAP, pyr., DMF (rt)(iv) = AsAc, bipy, CuBr2, MeOH (40 °C)

xb)a)

O

OBr

O

O

O

4/5

O

O

O

2

(i)

O

O

O

4/5

O

O

O

2

O

OBr

cba

Linear polymers

Comb polymers

RESULTS AND DISCUSSION

__________________________________________________

29

Table 4-1. Library of linear PEGMA-based homo- and copolymers.

Polymera OEGMA mol % Mnc

(kDa) Đc LCSTd

(˚C) f Fb

PDEGMA 0 0 30.4 1.2 27 P(OEGMA15-co-DEGMA85) 13 15 25.8 1.2 33 P(OEGMA19-co-DEGMA81) 19 19 33.0 1.3 36 P(OEGMA27-co-DEGMA73) 24 27 53.4 1.6 38 P(OEGMA37-co-DEGMA63) 36 37 58.1 1.3 42 P(OEGMA54-co-DEGMA46) 50 54 21.2 1.1 55 P(OEGMA73-co-DEGMA27) 70 73 30.0 1.3 59 POEGMA 100 100 14.1 1.1 73

a Subscript number denotes mol % of monomer in the final polymer b Assessed by 1H NMR c

Assessed by THF-SEC calibrated with linear PS standards d Assessed by DLS.

FT-IR shows the appearance of carbonyl groups (C=O, stretch at 1729 cm-1)) while the absorption of the hydroxyl groups of cellulose (OH, stretch at 3100-3600 cm-1) is reduced (Figure 4-1). This is more pronounced for the adduct from the anhydride esterification, which is believed to be a solvent effect; DMF dissolves the cellulose more efficiently than CH2Cl2, thus rendering more hydroxyl groups available for substitution.

Figure 4-1. FT-IR of neat HPC (black), HPC-I0.6 (blue), and HPC-I1.4 (green).

RESULTS AND DISCUSSION

__________________________________________________

30

The different conjugation efficiencies were further confirmed by establishing the bromine content by ICP-SMS analysis, and the average degrees of substitution (DS) of the two different HPC macroinitiators were determined to be 1.4 and 0.6 of the 3 available hydroxyl groups of each glucose unit (HPC-I1.4 and HPC-I0.6). In the second step of the synthesis, the HPC macroinitiators were grafted with DEGMA and OEGMA by ARGET ATRP, and characterized by 1H NMR and FT-IR. The polymerizations from HPC-I0.6 were, due to solubility reasons, conducted in methanol, and proceeded reproducibly to form the three comb copolymers; HPC0.6-g-PDEGMA, HPC0.6-g-POEGMA, and HPC0.6-g-P(OEGMA52-co-DEGMA48). Grafting reactions from the more substituted HPC-I1.4 were less reproducible with intermolecular coupling, even at high dilution. Therefore, only HPC-I1.4-PDEGMA was successfully synthesized.

4.1.2 Solution behavior of amphiphilic polymers The thermo-responsive properties of the linear and comb polymer architectures in aqueous solution were studied by DLS. All synthesized polymers showed good water solubility at room temperature (>50 mg mL-1). In their dissolved state, below LCST, the intrinsic amphiphilic nature of the polymers resulted in self-assembled structures in the nanometer range. DLS confirmed that the linear polymers formed associated structures with diameters of 4-12 nm, while the comb polymers with PEGMA brushes from the HPC backbone were stabilized as larger nanoobjects (20-40 nm). By monitoring the size of the linear polymer assemblies in solution by DLS it was found that all polymers displayed an instant and significant increase in size upon heating (Figure 4-2). Within one degree all linear polymers exhibited a 10 to 35-fold increase in size. This size increase is occurring at the LCST transition of the polymers when hydrogen bonds between water molecules and the ethylene glycol segments in the polymers are broken. At this temperature, polymer-polymer interactions become more thermodynamically favorable, thus causing a structural collapse of the polymer chains. Consequently, the polymer becomes insoluble, aggregates, and precipitates in the aqueous solution (Figure 4-2 insets). It can also be seen that the LCSTs of the polymers are highly dependent of the composition. The PDEGMA homopolymer, the most hydrophobic polymer with the lowest amount of ethylene glycol segments, consequently showed the lowest LCST (27 °C). This correlates well with values from the literature.92 Furthermore, increased molar

RESULTS AND DISCUSSION

__________________________________________________

31

Figure 4-2. Sizes as a function of temperature for the linear homo- and copolymers assessed by DLS. PDEGMA (■), P(OEGMA15-co-DEGMA85) (□), P(OEGMA19-co-DEGMA81) (▲), P(OEGMA27-co-DEGMA73) (Δ), P(OEGMA37-co-DEGMA63) (●), P(OEGMA54-co-DEGMA46) (○), P(OEGMA73-co-DEGMA27) (♦), POEGMA (◊). Insets show aqueous solutions (3 mg mL-1) below (bottom) and above (top) LCST. Numbers above curves represent mol % OEGMA in the polymer.

fraction of OEGMA in the P(OEGMA-co-DEGMA) copolymers increased the LCST almost linearly (Figure 4-3). Conclusively, the transition temperature of EGMAs-based polymers synthesized by ARGET ATRP can not only be tailored by choosing the right monomers, it can further be fine-tuned by copolymerization of different monomers with a specific monomer feed ratio. This has also been found by other controlled radical polymerization techniques.89, 92 The comb copolymer architectures exhibit a thermo-responsive behavior similar to that of the linear polymers, thus displaying a clear phase transition upon heating. However, interestingly their transition temperatures were all lower than for their linear analogues (Figure 4-3), which also has been reported by others for similar systems.160 This is believed to be an effect of a number of parameters: increased hydrophobicity, increased molecular weight, and restricted mobility of the polymer chains in the comb architecture. Most likely the substantially higher molecular weights of the comb polymers is most significant, which correlates well with the less pronounced decrease for HPC0.6-g-PDEGMA than for HPC0.6-g-POEGMA compared to their linear analogues. The architectural difference between HPC0.6-g-PDEGMA, with fewer but longer grafts compared to HPC1.4-g-PDEGMA, resulted in somewhat larger self-assembled structures in solution, however; showed

RESULTS AND DISCUSSION

__________________________________________________

32

no significant impact to the LCST. A striking result is that although the LCST is generally lowered for the comb copolymers, the ability to tune the transition temperature with OEGMA content remained (Figure 4-3). This suggests that also more complex macromolecular architectures can be tailored with the tunable responsive behavior of PEGMAs.

Figure 4-3. LCST as a function of OEGMA content in the linear (♦) and comb-shaped (○) polymers.

4.2 19F-containing polymer NPs as theranostic agents (paper II)

During the last decades, polymer micelles have attracted attention as nanoscale delivery systems to target various cancers.110, 112, 113 Their ability to solubilize low molecular weight therapeutics, as well as their potential to enhance the pharmacokinetics and selective bioaccumulation of the drug is intriguing in the search for effective treatment. Theranostic platforms combine this therapeutic delivery with diagnostic imaging in treatment to enable enhanced synergistic effects. In this study the aim was to design polymer micelles with both therapeutic and diagnostic functions and evaluate their potential to treat breast cancer. Since the micelles have diameters in the nanometer range they will be denoted NPs.

RESULTS AND DISCUSSION

__________________________________________________

33

4.2.1 Synthesis of 19F-containing amphiphilic polymers In this part of the work the aim was to synthesize 19F-containing amphiphilic polymers (Scheme 4-2) that spontaneously self-assemble into core-shell NPs in aqueous solution. Low molecular weight compounds, linear or dendrimers, with initiators suitable for ATRP were readily synthesized by base catalyzed esterification reactions. Linear hydrophilic blocks of the monomers trifluoroethyl methacrylate (TFEMA) and OEGMA475 were polymerized from the hydrophobic initiators, resulting in a library of linear and star-shaped amphiphilic block copolymers (Table 4-2). TFEMA is interesting as contrast agent

Scheme 4-2. Synthetic route to the amphiphilic 19F-containing polymers starting from different hydrophobic initiating moieties (i-v). Inset shows a representative kinetic plot from the polymerizations, (ln([M0]/[M])=f(time), where the dotted line is included to guide the eye.

RESULTS AND DISCUSSION

__________________________________________________

34

due to its commercial availability and chemically equivalent fluorine atoms, which is optimal for NMR imaging since it causes no dilution of the signal. Table 4-2. Library of 19F-containing amphiphilic polymers.

Polymera Typeb

OEGMA:

TFEMAc (mol %)

F-contentd (wt %)

Mne

(kDa) Ðe

EBiB-P(O-co-T) Linear 51:49 8.9 25.3 1.2 Dodecanol-P(O-co-T) Linear 48:52 9.5 21.5 1.1 Cholesterol-P(O-co-T) Linear 44:56 10.6 21.3 1.2 diTMP-[G#0]-P(O-co-T)short Star-4 49:51 8.8 10.3 1.2

diTMP-[G#0]-P(O-co-T)long Star-4 50:50 8.6 27.9 1.1

diTMP-[G#2]-P(O-co-T)short Star-16 44:56 4.4 22.5 1.1

diTMP-[G#2]-P(O-co-T)medium Star-16 47:53 5.4 36.0 1.1

diTMP-[G#2]-P(O-co-T)long Star-16 49:51 6.2 52.0 1.1

a P(O-co-T) denotes P(OEGMA-co-TFEMA) b Number denotes the maximum number of arms c Assessed by 1H NMR d Calculated from molar fraction of TFEMA e Assessed by DMF-SEC calibrated with linear PMMA standards.

The polymerizations were allowed to proceed to 30-50 % monomer conversion, progressed by reproducible kinetics, and resulted in polymers with well-controlled molecular weights and narrow molecular weight dispersities (Figure 4-4). Importantly, since the polymers were designed for biomedical applications, remaining traces of copper were removed by dialysis against aqueous EDTA solution. The composition of the polymers was characterized by 1H NMR, and the introduction of fluorine to the hydrophilic POEGMA segment allowed for polymers with a high fluorine content (4.4-10.6 wt %), while a high water solubility was retained (>15 mg mL-1 in PBS). It was further found that TFEMA is slightly more reactive than OEGMA; thus the resulting polymers are expected to have a tapered composition with more TFEMA closer to the hydrophobic initiator. The incorporation of the chemically equivalent fluorines of TFEMA was further confirmed by 19F NMR of the polymers in D2O, showing a singlet peak at 2.56 ppm (relative TFA).

RESULTS AND DISCUSSION

__________________________________________________

35

Figure 4-4. Size exclusion chromatograms of a) EBiB-P(O-co-T) (· · ·), Dodecanol-P(O-co-T) (─), Cholesterol-P(O-co-T) (- - -) b) diTMP-[G#0]-P(O-co-T)short (· · ·), diTMP-[G#0]-P(O-co-T)long (─) c) diTMP-[G#2] core (- · -), diTMP-[G#2]-P(O-co-T)short (- - -), diTMP-[G#2]-P(O-co-T)medium (─), diTMP-[G#2]-P(O-co-T)long (· · ·).

4.2.2 Nanoparticle formation and solution properties

The polymers were self-assembled into NPs in aqueous solution by a two-phase emulsion method. The polymers were dissolved in a good solvent (CH2Cl2), PBS was slowly added, and CH2Cl2 was allowed to evaporate during gentle stirring overnight. The NP properties were evaluated by DLS, TEM and fluorescence spectroscopy. DLS confirmed that all polymers self-assembled into relatively small NPs (d=6-9 nm), which are in the proximity of the reported threshold for renal clearance.110, 119 The sizes were not significantly affected by the hydrophobic domain, and the small nanoparticle diameters are most likely an effect of the minor hydrophobic component. TEM indicates that the particles formed are circular in shape with a relatively uniform size distribution. The critical micelle concentration (CMC) of the polymers was determined by the pyrene probe technique159 using fluorescence spectroscopy. In Figure 4-5a, a clear change in the fluorescence characteristics of pyrene was observed upon increased concentration, which confirms the onset of micellization. The CMC value was determined as the point of intersection of the two dotted lines. Favorably, very low CMCs were found for all polymers (0.8-4 µg mL-1), and generally the CMCs were lower for the star-shaped polymers. This effect may be attributed to their increased hydrophobic domain and their supposed lower aggregation numbers due to steric limitation. Low CMCs is crucial for self-assembling DDS to avoid disassembly upon injection and dilution in the blood pool.

RESULTS AND DISCUSSION

__________________________________________________

36

Figure 4-5.a) Change in fluorescence characteristics of pyrene (I1/I3) as a function of Cholesterol-P(O-co-T) concentration. Dashed lines are included to guide the eye. b) Size of Cholesterol-P(O-co-T) NPs determined in PBS by DLS and in different media by NMR-d.

PEG has been suggested as an effective antifouling agent to avoid severe protein adsorption during blood circulation.121, 161 Due to the high amount of 19F in the polymers, a unique novel approach was used to study the stealth properties of the PEGMA-based NPs. 19F NMR diffusion allows for the assessment of the NP size in complex protein-rich media such as blood plasma. The diffusivity of the NP is related to its diameter by the Stokes-Einstein equation, and can thereby detect any severe protein adsorption to the NP surface. Nanoparticle diameters assessed by 19F NMR diffusion (NMR-d) in PBS show good correlation to sizes determined by DLS (Figure 4-5b). Further, studies by NMR-d in three different media (PBS, DMEM, and blood plasma) interestingly indicate that the NPs exhibit a stealthy appearance with no significant size change in the complex media compared to PBS. Most often size measurements of NPs in the presence of serum results in some aggregation or protein deposition on the surface which increases the size of the construct.

RESULTS AND DISCUSSION

__________________________________________________

37

4.2.3 In vitro evaluation of theranostic function The drug delivery properties of the polymer NPs were evaluated with the common cancer therapeutic doxorubicin (DOX) as model drug. DOX was effectively loaded to the NPs simultaneously to the self-assembly process, with high loading efficiencies (70-89 wt %). However, DOX loading caused interparticle aggregation to occur to some extent, which is believed to be attributed to some unspecific loading of DOX in the NP corona. It is hypothesized that association between DOX and the hydrophobic fluorinated segments, and strong π-π interactions between DOX molecules, may cause this behavior. It was further shown that this aggregation could be reduced by the incorporation of negatively charged methacrylic acid to the hydrophilic corona of the NPs to induce charge repulsion (further discussed in paper II). The in vitro DOX release from the NPs was evaluated compared to free DOX. All NPs showed a burst release behavior, followed by a diffusion-controlled release profile (Figure 4-6). The release kinetics shows clear core dependence, where the more hydrophobic and larger G#2-core NPs release DOX slower than their G#0 counterparts. Interestingly, the cholesterol-cored NPs show significantly retarded release kinetics compared to the other linear cores. This is believed to be an effect of good assembling properties of cholesterol, as well as favorable π-π interactions between cholesterol and DOX in the NP core. Conclusively, the NPs can successfully be loaded with DOX. Further, encapsulation of DOX to the NPs allows for an extended release, and the release profile can be tailored by the polymer design.

RESULTS AND DISCUSSION

__________________________________________________

38

Figure 4-6. In vitro accumulative DOX release from the 19F-containing polymer NPs as a function of time.

The potential of the NPs as diagnostic agents was evaluated by 19F NMR and 19F MRI in PBS. The relaxation times were evaluated for three chosen NPs: EBiB-P(O-co-T), Cholesterol-P(O-co-T) and diTMP-[G#0]-P(O-co-T)long. No significant differences of the longitudinal, T1 (321-323 ms), and transversal, T2 (5.4-7.9 ms), relaxation times were observed for the different polymers, which confirm that the surroundings of the fluorine atoms are fairly equal in all polymers. The beneficially short T1 relaxations are likely a consequence of the attachment of the fluorine atoms to polymers. The T2 values are somewhat lower than expected, which indicates that the mobility of the fluorine atoms in the dynamic OEGMA segments is not as high as anticipated. This may be an effect of the higher reactivity of TFEMA compared to OEGMA, thus resulting in

RESULTS AND DISCUSSION

__________________________________________________

39

assembled TFEMA sequences, with restricted mobility, closer to the hydrophobic core. Nevertheless, all polymer NPs still exhibited good imaging capabilities by 19F MRI (Figure 4-7). Signal-to-noise ratios of 7-16 were achieved after only 10 minutes scan time (32 scans, 10 mg mL-1), which is believed to be an effect of the high fluorine concentration enabled in the NPs.

Figure 4-7. 19F-MRI phantoms of NP solutions (PBS, 10 mg mL-1) a) EBiB-P(O-co-T) b) Cholesterol-P(O-co-T) c) diTMP-[G#0]-P(O-co-T)long. Images taken as cross-section of plastic syringes (d=6 mm) with NP solution. SNR = signal-to-noise ratio.

4.2.4 Cell-based assessments

The delivery properties of the DOX-loaded NPs were investigated using three different human breast cancer cell lines and one mouse monocyte cell line. The toxicities of both neat polymers and DOX-loaded NPs were evaluated by MTT assay. Cell viabilities of the two cancer cell lines, MCF-7 and MDA-MB-468, after incubation with Chol-P(O-co-T) NPs are reported in Figure 4-8a. All neat polymer NPs showed to be non-toxic to the tested cell lines with cell viabilities above 90 % after 72 hours of incubation. Further, this indicates that most of the copper from the synthesis was successfully removed by treatment with EDTA solution. On the contrary, NPs loaded with DOX, significantly suppressed the cell viability of all tested cell lines to an extent similar to that as of free DOX, indicating promising delivery properties (Figure 4-8b). The DOX-loaded NPs showed both a time- and concentration-dependent toxicity. The observed toxicity effects were further investigated by apoptosis/necrosis staining and quantitatively evaluated by flow activated cell sorting (FACS). No obvious increases of apoptosis or necrosis were found for neat NPs, while DOX-NPs proved to be highly effective in killing the cells.

RESULTS AND DISCUSSION

__________________________________________________

40

Figure 4-8. a) Effect of neat Cholesterol-P(O-co-T) NPs to the cell viability of MCF-7 and MDA-MB-468 cells after 48 h and 72 h of incubation b) Effect of free DOX and DOX-Cholesterol-P(O-co-T) to the cell viability of MDA-MB-468 cells after 48 h and 72 h of incubation. 10 µg mL-1 of DOX equivalents 56 µg mL-1 of polymer NPs.

The drug delivery efficacy was further evaluated by studying the localization of the NPs and DOX-NPs in MDA-MB-231 cells by confocal microscopy. Neat NPs (equipped with fluorescence marker) proved to enter the cells, but were mainly distributed in the cytoplasm, concentrated in the surrounding of the nuclei (see paper II, supporting information). However, as seen in Figure 4-9, DOX-NPs favorably

RESULTS AND DISCUSSION

__________________________________________________

41

showed to deliver DOX mainly to the nucleus of the cells, similarly as for free DOX. Furthermore, the amount of DOX in the nuclei increased with time (4 h and 48 h). Consistently to toxicity results, this indicates that DOX effectively can be released from the NPs and penetrate into the cell nuclei.

Figure 4-9 c) Localization of DOX in the MDA-MB-231 cell line after 4 h and 48 h of incubation with free DOX or DOX-Cholesterol-P(O-co-T). Cell nuclei are stained with DAPI (blue) and DOX (red). Scale bar = 20 µm.

RESULTS AND DISCUSSION

__________________________________________________

42

4.3 Amphiphilic hyperbranched dendritic-linear polymers (HBDLPs) as unimolecular micelles (papers III and IV)

Polymer micelles are, due to their high molecular weight, considered as stable assemblies. However, this is only valid above the CMC, and under constant conditions. Consequently, significant dilution upon injection to the blood pool, or physical effects, such as variations in ionic strength and/or shear forces, may cause micellar destabilization to occur during circulation. This part of the thesis evaluates if high molecular weight polymer amphiphiles with a hyperbranched core and a high number of hydrophilic arms (referred to as HBDLP) can stabilize individually to form unimolecular micelles, i.e. one polymer forms one NP (paper III). Further, the potential to introduce functionality to the unimolecular carriers to allow for selective biodegradation and cargo release is explored (paper IV).

4.3.1 Synthesis of HBDLPs (papers III and IV) A two-step strategy was employed to synthesize micelle-resembling amphiphilic nanostructures consisting of high molecular weight HBDLPs. As illustrated in Scheme 4-3, the proposed synthesis route is very versatile and by changing the starting components, sophisticated and highly functional polymers can be realized.

4.3.1.1 Synthesis of HBMIs by SCVCP (papers III and IV) Generally, hyperbranched macroinitiators (HBMIs) were synthesized by copolymerization of an inimer (2 and/or 3) and a comonomer (4 or 5) by self-condensing vinyl copolymerization (SCVCP), in the presence of a trifunctional initiating core (1) (Scheme 4-3). A multifunctional initiating moiety has shown to narrow the molecular weight dispersities in SCVP.60,

62 SCVCP proceeds both via chain-growth and step-growth kinetics, and therefore the reactions were allowed to proceed to high conversions since higher molecular weights are mainly reached in the later stages of the polymerization. However, it is important to note that this, to some extent, will decrease the end-group fidelity and increase the risk of coupling reactions, and in order to suppress this, the reactions were performed in highly diluted systems.

RESULTS AND DISCUSSION

__________________________________________________

43

Sch

em

e4

-3.O

verv

iew

ofth

e sy

nth

esiz

edh

yper

bra

nch

ed d

end

riti

c-l

inea

rpo

lym

ers

(HB

DL

P)

in p

aper

III

(a)

an

d p

aper

IV

(b

an

d c

).

+

OO

OO

Br

(Pap

er I

II)

(Pap

er I

V)

(Pap

er I

V)

HB

MI(

SS

0), H

BM

I(S

S45

),

HB

MI(

SS

100)

(Va

ryin

gra

tio

so

f2

and

3)

2

32

2

5

(b)

(a)

(c)

HB

MI(

N3)

HB

MI-

1

HB

MI-

2

4

=H

BD

LP

-1 a

nd

-2

HB

DL

P-3

, -4,

an

d -

5

HB

DL

P(S

S0)

, HB

DL

P(S

S45

), H

BD

LP

(SS

100)

HB

DL

P(D

R)

1

4

HB

DL

P(N

3)

Dis

per

se R

ed (

DR

)

RESULTS AND DISCUSSION

__________________________________________________

44

In paper III, the inimer BBEMA (2) was used together with the comonomer hexyl acrylate (HA, 4) to achieve macroinitiators with desired hydrophobicity (Scheme 4-3a). By varying the molar feed ratio, two hyperbranched polymers (HBMI-1 and HBMI-2) of different molecular weight were synthesized and confirmed by 1H NMR and SEC (Table 4-3 and Figure 4-10a). The molecular weight from SEC is most likely underestimated as a consequence of the branched structure in relation to the linear PMMA standards used to calibrate the SEC. The molecular weight determined by NMR is estimated by assuming on average one TBBPE moiety per polymer, which may overestimate the molecular weight since undesired coupling reactions may result in more than one TBBPE moiety per polymer (further discussed in paper III). However, the results clearly suggest that hyperbranched polymers of different molecular weights can be achieved by altering the molar feed ratio. In addition, the theoretical amounts of initiators available for chain extension are significantly different for the two HBMIs. Table 4-3. Properties of the two different HBMIs synthesized in paper III.

a P(TBBPE-co-BBEMA-co-HA) b Average assessed from 1H NMR assuming one TBBPE moiety per HBMI and high end-group fidelity c Assessed by DMF-SEC calibrated with linear PMMA standards.

Figure 4-10. SEC traces of a) the two hyperbranched macroinitiators HBMI-1 and HBMI-2 (paper III) b) HBMI(SS100) before and after treatment with 10 mM DTT (paper IV).

Polymera Compositionb

TBBPE(1):BBEMA(2):HA(4)

Mn,NMRb

(kDa) Mn, SEC

c

(kDa) ĐM

c Br/molb

HBMI-1 1:12:37 10 3.6 1.3 15

HBMI-2 1:51:295 61 6.1 1.5 54

RESULTS AND DISCUSSION

__________________________________________________

45

In paper IV, the possibility to functionalize the HBMIs to allow for selective cleavage under reductive conditions was explored. Disulfide bonds were introduced to the HBMIs by two different approaches; either in the backbone (Scheme 4-3b) or in the pendant side groups (Scheme 4-3c) of the HBMIs. In the first approach, different amounts of disulfide bonds were introduced to the hyperbranched backbone to enable selective biodegradability of the polymers. Disulfide-containing HBMIs were synthesized by copolymerizing varying ratios of the inimers BBEMA (2) and BSSMA (3) with the comonomer, HA (4) (Scheme 4-3b). Consequently, three different HBMIs (HBMI(SS0), HBMI(SS45), and HBMI(SS100)) were synthesized, comprising 0 %, 45 %, and 100 % of disulfides in their potential branching points, respectively (assessed from 1H NMR). The introduction of disulfides was successfully confirmed by 1H-NMR. Furthermore, SEC chromatography before and after treatment with the reducing agent dithiotreitol (DTT) confirmed that the disulfide bonds could be selectively reduced to degrade the hyperbranched polymers (Figure 4-10b). In the second approach (Scheme 4-3c), the highly functional compound AzSSMA (5) was synthesized and utilized as comonomer instead of HA in the copolymerization with BBEMA (no disulfide inimer). The successful formation of HBMI(N3) was confirmed by 1H NMR, SEC, and FT-IR. 1H NMR showed that the disulfide bond (2.8-3.0 ppm) is still present and not cleaved during the polymerization (Figure 4-11a). Further, FT-IR confirmed that the vinyl groups (-C=C- stretch at 1637 cm-1) are almost quantitatively consumed and that the azide moieties (N3 stretch at 2094 cm-1) are preserved in the HBMI (Figure 4-11b).

RESULTS AND DISCUSSION

__________________________________________________

46

Figure 4-11. a) 1H NMR (CDCl3, 400 MHz) of HBMI(N3) b) FT-IR of HBMI(N3) indicating remaining azide moieties.

4.3.1.2 Chain-extension of HBMIs to form HBDLPs (papers III and IV) The preserved end-groups of all synthesized HBMIs were employed for chain-extension with hydrophilic linear polymers of oligo(ethylene glycol) methyl ether methacrylate (OEGMA475, average Mn=475 g mol-1) to form amphiphilic HBDLPs (Scheme 4-3a-c). In paper III, the two HBMIs with different molecular weight, HBMI-1 and HBMI-2, were chain-extended with different lengths of OEGMA segments by varying the initiator-to-monomer ratio, in order to evaluate

RESULTS AND DISCUSSION

__________________________________________________

47

the architectural effect to the NP formation. HBDLPs with varying architectures and molecular weights were successfully synthesized (Table 4-4). Table 4-4. Library of HBDLPs synthesized in paper III.

Samplea Core DP/armb Mn

c (kDa)

ĐMc

Mw, LSd

(kDa) Dh, z

e (nm)

DNMR-dg

(nm)

HBDLP-1 HBMI-1 10 28 1.7 103 21 ± 2f 10 ± 3

HBDLP-2 HBMI-1 116 132 1.5 473 32 ± 1 20 ± 6 HBDLP-3 HBMI-2 10 28 1.3 230 17 ± 2 12 ± 3

HBDLP-4 HBMI-2 49 81 1.5 410 33 ± 6f 20 ± 5

HBDLP-5 HBMI-2 104 153 1.6 605 39 ± 1 24 ± 6

a P((TBBPE-co-BBEMA-co-HA)-b-OEGMA) b Theoretical value assessed by 1H-NMR conversion c SEC (DMF) calibrated with linear PMMA standards d SEC-MALLS in aqueous solution e Assessed by DLS (z-average) in PBS solution f 20-30 intensity % of aggregates of 250-300 nm visible g Assessed by NMR-d in PBS solution. In paper IV, the HMBIs with functional cores (HBMI(SSx) and HBMI(N3)) were only chain-extended with one length of P(OEGMA) since this architecture showed the most potential to form unimolecular micelles (discussed below). All HBMIs were successfully chain-extended to HBDLPs with high molecular weights (500-950 kDa) as determined by triple detection SEC (Figure 4-12 and Table 4-5). The discrepancies between molecular weights from refractive index and light scattering detectors indicates a branched structure of the HBDLPs. As seen from the RI detector, polymers having disulfides in their core, HBDLP(SS45) and HBDLP(SS100), show a small shoulder towards lower molecular weights. This suggests that there is a difference in the reactivity of the initiating group in BSSMA and in BBEMA, affecting the synthesis of the HBMIs. A lower reactivity of BSSMA may result in branching being less probable, and thereby a higher proportion of more linear macroinitiators. Accordingly, HBDLPs with lower molecular weights may be formed as a consequence of macroinitiators with less available initiating groups in the chain extension reaction.

RESULTS AND DISCUSSION

__________________________________________________

48

Table 4-5. Library of disulfide- containing HBDLPs synthesized in paper IV.

Polymera Mn

b

(kDa)

Mwb

(kDa) Ðd

HBDLP(SS0) 450 800 1.9

HBDLP(SS45) 400 950 2.5

HBDLP(SS100) 350 800 2.3

Figure 4-12. SEC traces of disulfide-containing HBDLPs in DMF: HPDLP(SS0) (black), HPDLP(SS45) (red), and HPDLP(SS100) (blue).

a (SSx) where x denotes molar fraction of disulfide-containing brancher in HBMI core b Assessed by triple detection SEC (DMF).

4.3.2 Evaluation of architectural effects to the NP properties

(paper III) The HBDLPs in Table 4-4 were synthesized with different architectures to evaluate how the hydrophilic/hydrophobic ratio and architecture affect the possibility to stabilize as unimolecular micelles. The pure HBDLPs were allowed to form NPs in PBS by the same two-phase method described previously, and their properties in aqueous solution were assessed by DLS, NMR-d, TEM, and SEC MALLS. The molecular weights determined by SEC MALLS in aqueous solution verified that the synthesized HBDLPs are of high molecular weight, and a clear correlation between molecular weight and length of the P(OEGMA) segments was found (Figure 4-13a). DLS performed on PBS solutions confirmed that all HBDLPs formed NPs with mean hydrodynamic diameters between 17-39 nm, which suggests that the NP size can be tuned by the polymer architecture. NP diameters assessed from NMR-d confirmed this trend, however; are generally somewhat lower than results from DLS. This is somewhat expected since DLS to some extent overestimates larger objects of the distribution, while NMR-d rather overestimates smaller specimens.162 In addition, NPs formed by direct dissolution of the HBDLPs in PBS (minimum 24 h) were analysed by DLS, and now significant difference of the NP properties was observed. Further, the stability of the micelles was clearly affected by the polymer architecture, where intermolecular association to some extent occurred for most of the HBDLPs. Bimodality in the distribution curves from DLS and SEC MALLS indicates that NPs originating from the small HBMI

RESULTS AND DISCUSSION

__________________________________________________

49

Figure 4-13. a) Average NP diameters (DLS and NMR-d) and molecular weight (SEC-MALLS) as a function of theoretical DP of OEGMA/arm calculated from 1H NMR of HBDLP-3, HBDLP-4 and HBDLP-5 b) SEC traces of HBDLP-5 in H2O c) TEM of HBDLP-5. Inset shows magnified NP. Scale bars = 100 nm.

RESULTS AND DISCUSSION

__________________________________________________

50

(HBDLP-1 and 2), and with shorter P(OEGMA) segments from the larger HBMI (HBDLP-3 and HBDLP-4) were not individually stabilized. This is hypothesized to be a combined effect of too low molecular weight and unfavourable hydrophilic/hydrophobic ratio. Further, it has been reported that star polymers with few stabilizing arms are more prone to associate, since the density of arms decrease quickly with the radius from the core, thus promoting intermolecular interaction to occur.163 Interestingly, with increased number, and length, of the hydrophilic P(OEGMA) arms in HBDLP-5, uniform size distributions from DLS and SEC-MALLS (Figure 4-13b) were obtained. This suggests that the formation of individually stabilized unimolecular micelles may be governed by HBDLPs with: high molecular weights, an increased hydrophilic/hydrophobic ratio, and a high number of stabilizing arms. In addition, TEM analysis confirmed that the NPs are circular in shape with relatively uniform sizes (Figure 4-13c).

4.3.3 Formation and selectively triggered degradation of disulfide-containing NPs (paper IV)

The high molecular weight of the HBDLP NPs may cause issues with bioaccumulation, why the possibility to cause selective degradation may be an important parameter to control. One attractive approach to degrade the NPs is by introduction of disulfides to the HBDLP backbone. The HBDLPs with disulfides incorporated to the hyperbranched backbone (Table 4-5) were allowed to form NPs in PBS by direct dissolution for 24 h on a shaking device, and analyzed by DLS and TEM. Despite the different amount of disulfides in the NP core the HBDLPs formed uniform NPs (z-average = 35-39 nm) with low DLS polydispersity indices (PdI = 0.11-0.18) for such complex structures with quite broad molecular weight dispersities. The NPs further showed to be relatively insensitive to changes in concentration (Figure 4-14). Varying the HBDLP(SS100) concentration from 0.1 to 10 mg mL-1 only lowered the NP diameter with a few nanometers. This decrease is hypothesized to be caused by the change in solution properties and suggests that the NPs are stabilized individually, rather than being part of a dynamic assembly.

RESULTS AND DISCUSSION

__________________________________________________

51

Figure 4-14. DLS size distributions of HBDLP(SS100) at concentrations 0.1-10 mg mL-1 in PBS.

The selective degradation of HBDLP(SS0), HBDLP(SS45), and HBDLP(SS100) was evaluated by SEC under physiologically relevant reductive conditions (Figure 4-15). The degradation study of the polymers was performed for 24 h, and the degradation was found to be tunable with the amount of incorporated disulfide. HBDLP(SS0), with no disulfides, is unaffected after treatment with 10 mM DTT for 24 h, thus indicating that no unspecific cleavage of the HBDLP occurred. HBDLP(SS100) is significantly degraded by 10 mM DTT (proposed intracellular concentration74, 164), while HBDLP(SS45) kept some structural integrity due to “uncleavable” BBEMA units. Furthermore, at 10 µM DTT (proposed plasma concentration125, 165) the molecular weight of the HBDLPs are almost unaffected, thus suggesting that the NPs may preserve stability during circulation. However, treatment of

Figure 4-15. a) SEC traces of a) HBDLP(SS0) b) HBDLP(SS45) c) HBDLP(SS100) after treatment with different reducing agents and concentrations for 1 h (no further difference was observed after 24 h).

RESULTS AND DISCUSSION

__________________________________________________

52

HBDLP(SS100) with 10 mM of the biologically relevant reducing agent, glutathione (GSH), showed a less effective molecular weight reduction than DTT. This may be an effect of that the equilibrium between GSH and its oxidized form is more easily affected by the presence of air than the DTT analogue, which may cause difficulties in the NP bond reduction. Another possibility is that the NP core needs to be less hydrophobic to ensure accessibility for effective cleavage.

4.3.4 Post-conjugation and triggered release from NPs (paper

IV) The potential of HBDLP(N3) NPs as platform for selectively triggered release was evaluated with a disperse red (DR) dye as a mimic for a drug. Disperse red was equipped with an acetylenic moiety and conjugated to the core of the NPs by a copper(I)-catalyzed azide-alkyne cycloaddition, followed by extensive purification by precipitation and dialysis. 1H NMR confirmed the successful conjugation by the appearance of signals corresponding to DR and the triazole adducts (Figure 4-16), while a new absorption peak at 484 nm was observed by UV-Vis spectroscopy. The conjugation further resulted in enhanced water solubility of the hydrophobic dye when solubilized in the NP core (Figure 4-16 insets).

RESULTS AND DISCUSSION

__________________________________________________

53

Figure 4-16. Illustration of the DR conjugation to HBDLP(N3). 1H NMR of a) pure DR acetylene b) HBDLP(DR) after conjugation. Insets show vials of aqueous solutions before and after conjugation. The selective release of the dye from the NPs was monitored in different reductive conditions in vitro (Figure 4-17). In the control experiment in pure PBS the NPs were fairly stable and only very small amount of the dye was released in 72 h (7 %). The release indicates either some unspecific cleavage of the disulfide bond, or that a small amount of unconjugated DR was trapped in the NP core during purification. Furthermore, a clear difference was observed between the released amount of dye after 72 hours in 10 µM (14 %) and 10 mM (69 %) DTT. This is the proposed difference in reducing concentration between plasma and intracellular space.74, 125 At 10 mM, an instant and distinct release was observed, while at 10 µM the disulfide bonds were significantly more stable and only a minor increase compared to the control was discovered.

RESULTS AND DISCUSSION

__________________________________________________

54

Figure 4-17. Release of DR from HBDLP(DR) as a function of time at different reductive conditions. Insets show vials of aqueous solutions of HBDLP(DR) before (ii) and after 72 h of DR release in 10 mM DTT (i). Consequently, this suggests that the conjugation of a hydrophobic guest molecule to the NP core by disulfide bonds may be a potential strategy to afford selectively triggered delivery. Furthermore, the release in 10 mM of the reducing agent GSH reveals a significantly retarded release profile compared to 10 mM DTT. This indicates that GSH is not as effective as DTT in reducing the disulfide bonds, which correlates well with results for HBDLP degradation discussed previously. However, it is important to note that this needs to be further evaluated in vivo where the reducing environment is far more complex. Nevertheless, the proposed strategy shows interesting potential for selectively triggered release of covalently bound guest molecules from the stable unimolecular NPs.

CONCLUSIONS

__________________________________________________

55

5. CONCLUSIONS

The aim of the work in this thesis was to explore the potential of amphiphilic PEGMA-based architectures as nanoparticle (NP) drug delivery platforms. Initially, the solution and self-assembling properties of linear and comb-shaped EGMA-based polymers were evaluated. Linear homo- and copolymers of the monomers DEGMA and OEGMA with well-controlled molecular weights and molecular weight dispersities were successfully synthesized by ARGET ATRP. The monomers were found to polymerize with similar reactivity. In aqueous solution, all polymers were self-assembled into small NPs (d=4-12 nm) and exhibited very good water solubility (>50 mg mL-1) at room temperature. The polymers further showed thermo-responsive properties, going from soluble to insoluble upon heating above the LCST. This transition was found to increase nearly proportional to the amount of OEGMA in the copolymers. By utilizing hydroxypropyl cellulose macroinitiators, similar polymers of a more complex comb-shaped architecture were successfully synthesized. The comb copolymers also showed self-assembling characteristics, beneficially forming larger self-assembled structures (d=20-40 nm) than their linear analogues, and displaying low CMCs. Interestingly, the comb copolymers retained the thermo-responsive behavior observed for the linear polymers, where LCST was tuned by the OEGMA content. In the second study, OEGMA was copolymerized with TFEMA to form 19F-containing polymer NPs, which simultaneously could allow for therapeutic delivery and diagnostic read-out, so called theranostics. The polymerizations were performed from different hydrophobic (core-forming) initiators to form a library of amphiphilic structures with a high 19F-content. The polymers self-assembled into core-shell type NPs, and showed good water solubility even though about 50 mol % of TFEMA was incorporated in the hydrophilic segments. 19F-NMR diffusion measurements revealed that the NPs were not subjected to any severe protein adsorption in blood plasma, thus suggesting a stealthy corona property. Furthermore, the NPs showed promising theranostic function, allowing for both delivery of the model therapeutic doxorubicin (DOX) to cancer cells in vitro, as well as exhibiting good imaging capabilities by 19F-MRI. The diffusion-controlled release was found to be core-dependent, and the neat NPs were shown to be non-toxic to breast cancer cells, while DOX-loaded NPs showed a time- and concentration-dependent cytotoxicity.

CONCLUSIONS

__________________________________________________

56

To avoid issues with destabilization of self-assembling micelles upon injection, an interesting type of polymer architecture, hyperbranched dendritic-linear polymers (HBDLP), were designed and evaluated as potential unimolecular micelles. Hydrophobic, hyperbranched polymers were synthesized via SCVCP, and subsequently the multitude of available end-groups were chain extended with hydrophilic OEGMA segments. The size of the corresponding NPs in aqueous solution could beneficially be tailored (d=17-39 nm) by the HBDLP architecture and hydrophilic/hydrophobic ratio. However, these parameters also showed to influence the NPs stability, and results from DLS and SEC-MALLS indicated that unimolecular micelles are most likely only formed at sufficiently high molecular weight and a high number of stabilizing arms. Furthermore, the versatility of the unimolecular platform was demonstrated by introducing functionality to the hyperbranched core of the HBDLPs. Incorporation of redox-responsive disulfide bonds to the polymer backbone enabled selective polymer degradation at reductive environment mimicking intracellular conditions. The degree of degradation was conveniently tailored by the amount of incorporated disulfides in the HBDLP. This suggests that the unimolecular micelles can be stable during circulation, while allowing for biodegradation after delivery. In addition, HBDLPs with selectively cleavable reactive handles in the hyperbranched core demonstrated promising potential as platforms for selective delivery. The reactive groups were post-functionalized with a dye, mimicking a drug, and revealed a clear difference in released amount of dye at different reductive conditions. This suggests that undesired systemic toxicity due to premature release during delivery can be avoided, while an instant release may be triggered in tumor cells.

FUTURE WORK

__________________________________________________

57

6. FUTURE WORK

Given the findings in this thesis, the utilization of PEGMA as protecting corona in NP delivery systems is promising. The versatility of employing vinyl-based monomers enables facile synthesis of sophisticated polymer architectures, which allows for screening and understanding of important material-property relations. However, it is important to keep in mind that the NPs potential as drug delivery systems has been evaluated in vitro. Although, this is an effective and convenient approach to screen for suitable NP systems, it is crucial to evaluate promising alternatives in vivo to fully understand their behavior. Generally, the NP platforms developed in this thesis are aimed to passively accumulate in tumor tissue via the EPR effect. However, to ensure a high enough local concentration at the tumor site it would be beneficial to further equip the NPs with active targeting moieties, such as: antibodies, peptides, or small molecule ligands. Reactive sites for such functionalization are already available in most of the synthesized NPs. Such active targeting may be especially crucial for the theranostic delivery platform. The concept of a dual-functional platform by introducing fluorine atoms to the highly mobile, hydrophilic exterior of the polymer NPs is interesting. However, even though our fluorinated NPs showed good imaging capabilities in vitro, this was at a concentration of 10 mg mL-1, which is too high for a single injection. Functionalization with active targeting groups would therefore be essential to allow for effective diagnostic read-out in vivo as the targeting ligand can allow for sufficiently high local concentrations. The concept of utilizing HBDLPs to form unimolecular micelles as delivery systems to treat cancer is interesting. The stability of such NPs, in combination with their potentially selectively triggered biodegradation and therapeutic release, are promising to reduce issues with biopersistence and toxicity caused by premature therapeutic release. However, the following suggestions should be considered to further enhance their efficacy: (1) The two-step synthetic procedure by SCVCP and ATRP is indeed applicable and an effective route to complex polymer architectures. However, the control in the synthesis still need to be optimized since the use in biomedical applications requires a high level of control and batch to batch reproducibility. It would therefore be interesting to study the synthesis of HBDLPs in other polymerization systems, as microemulsion

FUTURE WORK

__________________________________________________

58

or by the polymerization induced self-assembly (PISA) technique, to further control the final polymer properties. (2) In this work the introduction of disulfide bonds to allow for selectively triggered biodegradation and cargo release was only illustrated separately. The natural next step would be to combine the features of biodegradability and therapeutic release within the same NP platform. In addition, the efficacy should be evaluated by the covalent conjugation of a cancer therapeutic, such as Doxorubicin or Paclitaxel, to investigate the selective release to cancer cells. (3) The biodegradability of the HBDLPs needs to be further investigated. The possibilities for more extensive degradation, and the formation of less amphiphilic degradation products, should be examined in detail to ensure efficient clearance after delivery.

ACKNOWLEDGEMENTS

__________________________________________________

59

7. ACKNOWLEDGEMENTS

First of all, none of the work in this thesis would have been possible without the funding. Wilhelm Beckers Jubileumsfond and the Swedish Research Council (Vetenskapsrådet) are greatly acknowledged for the financial support. I would like to express my deepest thanks to my main supervisor Prof. Eva Malmström. Eva, I feel privileged to have had the opportunity to work with you and get to know you during the past six years. You truly inspire me with your endless enthusiasm and your ability to make me leave a meeting full of ideas and motivation. Thank you for always making room in your busy calendar, constantly sacrificing your own time to help and encourage. I have really enjoyed our discussions, both about science and everyday life. I also would like to thank you for your great personal consideration during tough times, it meant a lot to me! My second supervisor Ass. Prof. Andreas Nyström is greatly acknowledged for all help, and for introducing me to the field of nanomedicine. I have truly appreciated your knowledge, patience, and consideration. Thank you for always responding extremely fast to everything I have thrown at you, and for making me confident. I think you belong in academia, and I hope you get the opportunity to return! Prof. Anders Hult is acknowledged for creating the good atmosphere in “Ytgruppen” and for always finding every mistake in an e-mail and taking the opportunity to make fun of the person making it. Prof. Mats Johansson is acknowledged for all nice early morning coffees, “letting” me win in Wordfeud, and introducing me to squash. Thank you also to Dr Linda Fogelström for your sincere kindness and encouragement, and for always being there when needed the most. Ass. Prof. Anna Carlmark Malkoch is greatly thanked for always being positive and for all the help you have given me during the years. Ass. Prof. Michael Malkoch is acknowledged for your caring personality and being the only person seeing me as a “kicker”. I would like to express my sincere gratitude to my co-authors for fruitful collaborations: Yuning Zhang, Niklas Nordgren, Cosimo Ducani, Åsa Östlund, Lars Nordstierna, Peter Damberg, Francisco Vilaplana, Maria I. Montañez, Mauri Kostiainen, and Jani-Markus Malho. I am really grateful for all your help and the support you have given me, as well as for everything I have learnt from you. One special thanks goes to Yuning. Thank you for all nice collaborations during the years, and I am glad for

ACKNOWLEDGEMENTS

__________________________________________________

60

the friendship we have developed. I have really enjoyed working with you and I will remember the trips to KI with joy. All former and present friends at “Ytgruppen” are greatly acknowledged for making my PhD studies unforgettable. Every single one of you has contributed to the unexplainably good atmosphere in the group, and I am so happy that I got the opportunity to get to know you all. Markus and Carl, thank you for being there “all the way”, and for all heavily planned or spontaneous beers (en, max två, kanske tre!). Markus, thanks for all afternoons with training and laughter at KTH-hallen, all trips during the years, for developing your own “technique” at GRIS, and for sacrificing your home to the art of beer brewing. “Psch!” Carl, thank you for your fantastic laughter, all cooking courses, and thinking everything is as “schysst” as I do. Susanne is thanked for being my first supervisor, for teaching me “everything you can”, taking me in at Forskarskolan, and for being Spiderpig. These years would not have been the same without you! Linn (Stinky), thank you for being such a warm person, sharing my lab consideration, and all support during the years. Robert, Pontus and Kim are thanked for all fun during the years, trips to “The Beach” and frisbeegolf. Robert, thank you for all the ”pöh” during volleyball, introducing me to cross-country skiing, the ski trips to Tänndalen, and for always taking the subway to Aspudden. Pontus is thanked for the legendary party night in Santa Barbara and for having Musse Pigg t-shirts; Kim for helping me capture Akropolis, and for being a supportive friend. One special thanks is sent to all my former and present roomies, we really kept the spirit up in the office with no air! Emelie for always having a spare hug; Susana for “taking over” when all the oldies left, GRRRarmar, and being the tough one in the new version of the “Beer fridge mafia”. Also, thanks to both of you for being the perfect secretaries! Sandra is thanked for always having an answer and time to help with the CDI chemistry, and for being a part of the “Zebra Team”; Marie for being so reasonable and always having time to help; Yvonne for making intervals in KTH-hallen fun and being the worst “stekare” at GRIS; Assya for being the best hood neighbor and for all the “distance hugs”; Emma for making the KTH journey together with me; Kristina for all fun both at and outside work; Jocke for bringing so much energy into the group, your good heart, and for taking over frystorken; Janne for always spreading joy and for letting me know my initials are big in Thailand; Camilla for making me feel very welcome in the group; Sam for making me try weird beers, and always taking the scientific discussion; Oliver for taking so much responsibility in the lab and all micelle discussions; Petra for always being so nice; Viktor for being an early lunch eater, and hopefully for taking over as KTH-hallens new posterboy(!?); Carmen for being my first diploma worker, introducing

ACKNOWLEDGEMENTS

__________________________________________________

61

me to delicious Spanish fish balls, and bringing cookies and berries to “ytköket”; Martin for all TEM discussions and for canoeing (and swimming) with a jeans jacket; Fiona for making scones “to die for” and for all the proof reading; Alireza for your version of “Sten-sax-påse”; Danne for making the best lab quarter even better; Ville for your devoted care of the coffee machine, and for always helping out with the chemistry when needed; Samer for always having a series to recommend; Jonas for being awesome at “Radi”; Sara O. for a nice trip to Gothenburg; Yanling for introducing me to the cell lab; Niklas I for covering up after Markus at KTH-hallen (at least for a while…); Marcus for all ”I do not understand this stupid computer thing” help; Tahani for taking over the lab cleaning; Tobias for making ladies walk into fruit stands; Ting for your brutal honesty; Mauro for always being so encouraging. Further, I would like to express my sincere gratitude to all friends and colleagues at the Department of Fibre and Polymer Technology for five fun years. Thank you to the golfers; Andreas, Carl, Emil, Christopher, Anders, and Nicholas for always helping me finding the ball after my sliced driver. Christopher, Dimitri and Ulf are thanked for taking me in at the Tuesday badminton. Thank you also to everyone participating in the Friday floorball. I also want to thank all the administrative staff who constantly helped out: Mia, Bosse, Kjell, Karin, Inga, and Vera. You made life so much easier, thank you! Jag vill uttrycka min stora tacksamhet och kärlek till min familj; Inger, Jan-Peter, Helena och Brittis. Ni har hjälpt mig att bli den jag är, tack för allt stöd och för att ni alltid finns där. Tillsammans klarar vi även svåra stunder, jag älskar er! Sofia, jag är så fantastiskt lycklig att jag har dig! Tack för allt roligt vi har haft hittills och för allt som jag vet kommer. Jag är så glad för ditt oändliga stöd och din förmåga att alltid ge mig värme och trygghet. Utan dig hade den här resan inte varit densamma, jag älskar dig!

REFERENCES

__________________________________________________

62

8. REFERENCES

1. Patterson, G., Prehistory of polymer science, Springer 2012. 2. Wang, J.-S. and Matyjaszewski, K., J. Am. Chem. Soc. 1995, 117

(20), 5614-5615. 3. Wang, J.-S. and Matyjaszewski, K., Macromolecules 1995, 28 (23),

7901-7910. 4. Kato, M.; Kamigaito, M.; Sawamoto, M. and Higashimura, T.,

Macromolecules 1995, 28 (5), 1721-1723. 5. Georges, M. K.; Veregin, R. P. N.; Kazmaier, P. M. and Hamer, G.

K., Macromolecules 1993, 26 (11), 2987-2988. 6. Chiefari, J.; Chong, Y. K.; Ercole, F.; Krstina, J.; Jeffery, J.; Le, T. P.

T.; Mayadunne, R. T. A.; Meijs, G. F.; Moad, C. L.; Moad, G.; Rizzardo, E. and Thang, S. H., Macromolecules 1998, 31 (16), 5559-5562.

7. Mayadunne, R. T. A.; Rizzardo, E.; Chiefari, J.; Chong, Y. K.; Moad, G. and Thang, S. H., Macromolecules 1999, 32 (21), 6977-6980.

8. Curran, D. P., Synthesis 1988, (7), 489-513. 9. Kharasch, M. S.; Jensen, E. V. and Urry, W. H., Science 1945, 102

(2640), 128. 10. Matyjaszewski, K. and Tsarevsky, N. V., J. Am. Chem. Soc. 2014,

136 (18), 6513-6533. 11. Kajiwara, A.; Matyjaszewski, K. and Kamachi, M., Macromolecules

1998, 31 (17), 5695-5701. 12. Matyjaszewski, K. and Xia, J., Chem. Rev. 2001, 101 (9), 2921-

2990. 13. Matyjaszewski, K.; Shipp, D. A.; Wang, J.-L.; Grimaud, T. and

Patten, T. E., Macromolecules 1998, 31 (20), 6836-6840. 14. di Lena, F. and Matyjaszewski, K., Prog. Polym. Sci. 2010, 35 (8),

959-1021. 15. Patten, T. E. and Matyjaszewski, K., Acc. Chem. Res. 1999, 32 (10),

895-903. 16. Ouchi, M.; Terashima, T. and Sawamoto, M., Chem. Rev. 2009, 109

(11), 4963-5050. 17. Matyjaszewski, K., Macromolecules 2012, 45 (10), 4015-4039. 18. Fischer, H., Chem. Rev. 2001, 101 (12), 3581-3610. 19. Patten, T. E. and Matyjaszewski, K., Adv. Mater. 1998, 10 (12), 901-

915. 20. Haddleton, D. M.; Jasieczek, C. B.; Hannon, M. J. and Shooter, A.

J., Macromolecules 1997, 30 (7), 2190-2193. 21. Shipp, D. A.; Wang, J.-L. and Matyjaszewski, K., Macromolecules

1998, 31 (23), 8005-8008.

REFERENCES

__________________________________________________

63

22. Percec, V. and Barboiu, B., Macromolecules 1995, 28 (23), 7970-7972.

23. Matyjaszewski, K.; Patten, T. E. and Xia, J., J. Am. Chem. Soc. 1997, 119 (4), 674-680.

24. Teodorescu, M. and Matyjaszewski, K., Macromol. Rapid Commun. 2000, 21 (4), 190-194.

25. Senoo, M.; Kotani, Y.; Kamigaito, M. and Sawamoto, M., Macromolecules 1999, 32 (24), 8005-8009.

26. Matyjaszewski, K.; Jo, S. M.; Paik, H.-j. and Shipp, D. A., Macromolecules 1999, 32 (20), 6431-6438.

27. Matyjaszewski, K.; Jo, S. M.; Paik, H.-j. and Gaynor, S. G., Macromolecules 1997, 30 (20), 6398-6400.

28. Xia, J.; Paik, H.-j. and Matyjaszewski, K., Macromolecules 1999, 32 (25), 8310-8314.

29. Xia, J.; Zhang, X. and Matyjaszewski, K., Macromolecules 1999, 32 (10), 3531-3533.

30. Treat, N. J.; Sprafke, H.; Kramer, J. W.; Clark, P. G.; Barton, B. E.; Read de Alaniz, J.; Fors, B. P. and Hawker, C. J., J. Am. Chem. Soc. 2014, 136 (45), 16096-16101.

31. Jakubowski, W. and Matyjaszewski, K., Angew. Chem., Int. Ed. 2006, 45 (27), 4482-4486.

32. Matyjaszewski, K.; Jakubowski, W.; Min, K.; Tang, W.; Huang, J.; Braunecker, W. A. and Tsarevsky, N. V., Proc. Natl. Acad. Sci. U. S. A. 2006, 103 (42), 15309-15314.

33. Jakubowski, W.; Min, K. and Matyjaszewski, K., Macromolecules 2006, 39 (1), 39-45.

34. Min, K.; Gao, H. and Matyjaszewski, K., Macromolecules 2007, 40 (6), 1789-1791.

35. Wooley, K. L.; Fréchet, J. M. J. and Hawker, C. J., Polymer 1994, 35 (21), 4489-4495.

36. Gitsov, I., J. Polym. Sci., Part A: Polym. Chem. 2008, 46 (16), 5295-5314.

37. Hatton, F. L.; Chambon, P.; McDonald, T. O.; Owen, A. and Rannard, S. P., Chem. Sci. 2014, 5 (5), 1844-1853.

38. Hatton, F. L.; Tatham, L. M.; Tidbury, L. R.; Chambon, P.; He, T.; Owen, A. and Rannard, S. P., Chem. Sci. 2015, 6 (1), 326-334.

39. Flory, P. J., J. Am. Chem. Soc. 1952, 74, 2718-2723. 40. Tomalia, D. A. and Fréchet, J. M. J., J. Polym. Sci., Part A: Polym.

Chem. 2002, 40 (16), 2719-2728. 41. Tomalia, D. A. and Fréchet, J. M., Prog. Polym. Sci. 2005, 30 (3-4),

217-219. 42. Tomalia, D. A., Mater. Today 2005, 8 (3), 34-46.

REFERENCES

__________________________________________________

64

43. Tomalia, D. A.; Baker, H.; Dewald, J.; Hall, M.; Kallos, G.; Martin, S.; Roeck, J.; Ryder, J. and Smith, P., Polym. J. 1985, 17 (1), 117-132.

44. Newkome, G. R.; Yao, Z.; Baker, G. R. and Gupta, V. K., J. Org. Chem. 1985, 50 (11), 2003-2004.

45. Hawker, C. J. and Fréchet, J. M. J., J. Am. Chem. Soc. 1990, 112 (21), 7638-7647.

46. Voit, B., J. Polym. Sci., Part A: Polym. Chem. 2000, 38 (14), 2505-2525.

47. Kim, Y. H. and Webster, O., J. Macromol. Sci., Polym. Rev. 2002, C42 (1), 55-89.

48. Voit, B. I. and Lederer, A., Chem. Rev. 2009, 109 (11), 5924-5973. 49. Yates, C. R. and Hayes, W., Eur. Polym. J. 2004, 40 (7), 1257-1281. 50. Sunder, A.; Heinemann, J. and Frey, H., Chem. - Eur. J. 2000, 6

(14), 2499-2506. 51. Hawker, C. J.; Lee, R. and Fréchet, J. M. J., J. Am. Chem. Soc. 1991,

113 (12), 4583-4588. 52. Hult, A.; Johansson, M. and Malmström, E., Adv. Polym. Sci. 1999,

143 (1), 1-34. 53. Gao, H. and Matyjaszewski, K., Prog. Polym. Sci. 2009, 34 (4), 317-

350. 54. Gao, C. and Yan, D., Prog. Polym. Sci. 2004, 29 (3), 183-275. 55. Fréchet, J. M. J.; Henmi, M.; Gitsov, I.; Aoshima, S.; Leduc, M. R.

and Grubbs, R. B., Science 1995, 269 (5227), 1080-1083. 56. Hawker, C. J.; Fréchet, J. M. J.; Grubbs, R. B. and Dao, J., J. Am.

Chem. Soc. 1995, 117 (43), 10763-10764. 57. Gaynor, S. G.; Edelman, S. and Matyjaszewski, K., Macromolecules

1996, 29 (3), 1079-1081. 58. Mori, H.; Müller, A. H. E. and Simon, P. F. W., Hyperbranched

Polym. 2011, 139-174. 59. Hong, C.-Y.; Zou, Y.-F. and Pan, C.-Y., Polym. Int. 2003, 52 (2),

257-264. 60. Radke, W.; Litvinenko, G. and Mueller, A. H. E., Macromolecules

1998, 31 (2), 239-248. 61. Cheng, K.-C., Polymer 2002, 44 (3), 877-882. 62. Litvinenko, G. I. and Mueller, A. H. E., Macromolecules 2002, 35

(12), 4577-4583. 63. Min, K. and Gao, H., J. Am. Chem. Soc. 2012, 134 (38), 15680-

15683. 64. Graff, R. W.; Wang, X. and Gao, H., Macromolecules 2015, 48 (7),

2118-2126. 65. Carlmark, A.; Hawker, C.; Hult, A. and Malkoch, M., Chem. Soc.

Rev. 2009, 38 (2), 352-362.

REFERENCES

__________________________________________________

65

66. Aulenta, F.; Hayes, W. and Rannard, S., Eur. Polym. J. 2003, 39 (9), 1741-1771.

67. Wurm, F. and Frey, H., Prog. Polym. Sci. 2011, 36 (1), 1-52. 68. Wang, Y. and Grayson, S. M., Adv. Drug Delivery Rev. 2012, 64

(9), 852-865. 69. Gil, E. S. and Hudson, S. M., Prog. Polym. Sci. 2004, 29 (12), 1173-

1222. 70. Liu, F. and Urban, M. W., Prog. Polym. Sci. 2010, 35 (1-2), 3-23. 71. Roy, D.; Cambre, J. N. and Sumerlin, B. S., Prog. Polym. Sci. 2010,

35 (1-2), 278-301. 72. Galaev, I. Y. and Mattiasson, B., Trends Biotechnol. 1999, 17 (8),

335-340. 73. Kost, J. and Langer, R., Adv. Drug Delivery Rev. 2001, 46 (1-3),

125-148. 74. Fleige, E.; Quadir, M. A. and Haag, R., Adv. Drug Delivery Rev.

2012, 64 (9), 866-884. 75. Alarcon, C. d. l. H.; Pennadam, S. and Alexander, C., Chem. Soc.

Rev. 2005, 34 (3), 276-285. 76. Schmaljohann, D., Adv. Drug Delivery Rev. 2006, 58 (15), 1655-

1670. 77. Hruby, M.; Filippov, S. K. and Stepanek, P., Eur. Polym. J. 2015,

65, 82-97. 78. Cohen Stuart, M. A.; Huck, W. T. S.; Genzer, J.; Mueller, M.; Ober,

C.; Stamm, M.; Sukhorukov, G. B.; Szleifer, I.; Tsukruk, V. V.; Urban, M.; Winnik, F.; Zauscher, S.; Luzinov, I. and Minko, S., Nat. Mater. 2010, 9 (2), 101-113.

79. Xia, F.; Zhu, Y.; Feng, L. and Jiang, L., Soft Matter 2009, 5 (2), 275-281.

80. Lahann, J.; Mitragotri, S.; Tran, T.-N.; Kaido, H.; Sundaram, J.; Choi, I. S.; Hoffer, S.; Somorjai, G. A. and Langer, R., Science 2003, 299 (5605), 371-374.

81. Lindqvist, J.; Nyström, D.; Östmark, E.; Antoni, P.; Carlmark, A.; Johansson, M.; Hult, A. and Malmström, E., Biomacromolecules 2008, 9 (8), 2139-2145.

82. Xu, F. J.; Zhong, S. P.; Yung, L. Y. L.; Tong, Y. W.; Kang, E.-T. and Neoh, K. G., Biomaterials 2006, 27 (8), 1236-1245.

83. Schild, H. G., Prog. Polym. Sci. 1992, 17 (2), 163-249. 84. Vihola, H.; Laukkanen, A.; Valtola, L.; Tenhu, H. and Hirvonen, J.,

Biomaterials 2005, 26 (16), 3055-3064. 85. Lutz, J.-F.; Akdemir, O. and Hoth, A., J. Am. Chem. Soc. 2006, 128

(40), 13046-13047. 86. Hoogenboom, R.; Thijs, H. M. L.; Jochems, M. J. H. C.; van

Lankvelt, B. M.; Fijten, M. W. M. and Schubert, U. S., Chem. Commun. 2008, (44), 5758-5760.

REFERENCES

__________________________________________________

66

87. Mart, R. J.; Osborne, R. D.; Stevens, M. M. and Ulijn, R. V., Soft Matter 2006, 2 (10), 822-835.

88. Neugebauer, D., Polym. Int. 2007, 56 (12), 1469-1498. 89. Becer, C. R.; Hahn, S.; Fijten, M. W. M.; Thijs, H. M. L.;

Hoogenboom, R. and Schubert, U. S., J. Polym. Sci., Part A: Polym. Chem. 2008, 46 (21), 7138-7147.

90. Lutz, J.-F., J. Polym. Sci., Part A Polym. Chem. 2008, 46 (11), 3459-3470.

91. Lutz, J.-F. and Hoth, A., Macromolecules 2006, 39 (2), 893-896. 92. Lutz, J.-F.; Hoth, A. and Schade, K., Des. Monomers Polym. 2009,

12 (4), 343-353. 93. Lutz, J.-F., Adv. Mater. 2011, 23 (19), 2237-2243. 94. Lutz, J.-F.; Stiller, S.; Hoth, A.; Kaufner, L.; Pison, U. and Cartier,

R., Biomacromolecules 2006, 7 (11), 3132-3138. 95. Mai, Y. and Eisenberg, A., Chem. Soc. Rev. 2012, 41 (18), 5969-

5985. 96. Bates, F. S. and Fredrickson, G. H., Phys. Today 1999, 52 (2), 32-

38. 97. Blanazs, A.; Armes, S. P. and Ryan, A. J., Macromol. Rapid

Commun. 2009, 30 (4-5), 267-277. 98. Riess, G., Prog. Polym. Sci. 2003, 28 (7), 1107-1170. 99. Moughton, A. O.; Hillmyer, M. A. and Lodge, T. P., Macromolecules

2012, 45 (1), 2-19. 100. Kataoka, K.; Kwon, G. S.; Yokoyama, M.; Okano, T. and Sakurai, Y.,

J. Controlled Release 1993, 24 (1-3), 119-132. 101. Gaucher, G.; Dufresne, M.-H.; Sant, V. P.; Kang, N.; Maysinger, D.

and Leroux, J.-C., J. Controlled Release 2005, 109 (1-3), 169-188. 102. Letchford, K. and Burt, H., Eur. J. Pharm. Biopharm. 2007, 65 (3),

259-269. 103. Newkome, G. R.; Moorefield, C. N.; Baker, G. R.; Johnson, A. L. and

Behera, R. K., Angew. Chem. 1991, 103 (9), 1205-1207 (See also Angew. Chem., Int. Ed. Engl., 1991, 1230(1209), 1176-1208).

104. Newkome, G. R.; Moorefield, C. N.; Baker, G. R.; Saunders, M. J. and Grossman, S. H., Angew. Chem. 1991, 103 (9), 1207-1209.

105. Moorefield, C. N. and Newkome, G. R., C. R. Chim. 2003, 6 (8-10), 715-724.

106. O'Reilly, R. K.; Hawker, C. J. and Wooley, K. L., Chem. Soc. Rev. 2006, 35 (11), 1068-1083.

107. Moorefield, C. N.; Schultz, A. and Newkome, G. R., Braz. J. Pharm. Sci. 2013, 49, 67-84.

108. Stenzel, M. H., Macromol. Rapid Commun. 2009, 30 (19), 1603-1624.

109. Gao, H., Macromol. Rapid Commun. 2012, 33 (9), 722-734. 110. Duncan, R., Nat. Rev. Drug Discovery 2003, 2 (5), 347-360.

REFERENCES

__________________________________________________

67

111. Ferrari, M., Nat. Rev. Cancer 2005, 5 (3), 161-171. 112. Elsabahy, M. and Wooley, K. L., Chem. Soc. Rev. 2012, 41 (7),

2545-2561. 113. Kataoka, K.; Harada, A. and Nagasaki, Y., Adv. Drug Delivery Rev.

2001, 47 (1), 113-131. 114. Miyata, K.; Christie, R. J. and Kataoka, K., React. Funct. Polym.

2011, 71 (3), 227-234. 115. Minelli, C.; Lowe, S. B. and Stevens, M. M., Small 2010, 6 (21),

2336-2357. 116. Barenholz, Y., J. Controlled Release 2012, 160 (2), 117-134. 117. Hrkach, J.; Von, H. D.; Mukkaram, A. M.; Andrianova, E.; Auer, J.;

Campbell, T.; De, W. D.; Figa, M.; Figueiredo, M.; Horhota, A.; Low, S.; McDonnell, K.; Peeke, E.; Retnarajan, B.; Sabnis, A.; Schnipper, E.; Song, J. J.; Song, Y. H.; Summa, J.; Tompsett, D.; Troiano, G.; Van, G. H. T.; Wright, J.; LoRusso, P.; Kantoff, P. W.; Bander, N. H.; Sweeney, C.; Farokhzad, O. C.; Langer, R. and Zale, S., Sci Transl Med 2012, 4 (128), 128ra139.

118. Torchilin, V. P., J. Controlled Release 2001, 73 (2-3), 137-172. 119. Fox, M. E.; Szoka, F. C. and Fréchet, J. M. J., Acc. Chem. Res.

2009, 42 (8), 1141-1151. 120. Gillies, E. and Fréchet, J., Drug Discovery Today 2005, 10 (1), 35-

43. 121. Owens, D. E., 3rd and Peppas, N. A., Int. J. Pharm. 2006, 307 (1),

93-102. 122. Brigger, I.; Dubernet, C. and Couvreur, P., Adv. Drug Delivery Rev.

2002, 54 (5), 631-651. 123. Rapoport, N., Prog. Polym. Sci. 2007, 32 (8-9), 962-990. 124. Mura, S.; Nicolas, J. and Couvreur, P., Nat. Mater. 2013, 12 (11),

991-1003. 125. Saito, G.; Swanson, J. A. and Lee, K.-D., Adv. Drug Delivery Rev.

2003, 55 (2), 199-215. 126. Matsumura, Y. and Maeda, H., Cancer Res. 1986, 46 (12, Pt. 1),

6387-6392. 127. Maeda, H.; Wu, J.; Sawa, T.; Matsumura, Y. and Hori, K., J.

Controlled Release 2000, 65 (1-2), 271-284. 128. Torchilin, V., Adv. Drug Delivery Rev. 2011, 63 (3), 131-135. 129. Torchilin, V. P., Pharm. Res. 2007, 24 (1), 1-16. 130. Nakamura, H.; Jun, F. and Maeda, H., Expert Opin. Drug Delivery

2015, 12 (1), 53-64. 131. Lammers, T.; Hennink, W. E. and Storm, G., Br. J. Cancer 2008,

99 (3), 392-397. 132. Fang, J.; Nakamura, H. and Maeda, H., Adv. Drug Delivery Rev.

2011, 63 (3), 136-151.

REFERENCES

__________________________________________________

68

133. Noguchi, Y.; Wu, J.; Duncan, R.; Strohalm, J.; Ulbrich, K.; Akaike, T. and Maeda, H., Jpn. J. Cancer Res. 1998, 89 (3), 307-314.

134. Zhong, Y.; Meng, F.; Deng, C. and Zhong, Z., Biomacromolecules 2014, 15 (6), 1955-1969.

135. Steichen, S. D.; Caldorera-Moore, M. and Peppas, N. A., Eur. J. Pharm. Sci. 2013, 48 (3), 416-427.

136. Garcia-Bennett, A.; Nees, M. and Fadeel, B., Biochem. Pharmacol. 2011, 81 (8), 976-984.

137. Byrne, J. D.; Betancourt, T. and Brannon-Peppas, L., Adv. Drug Delivery Rev. 2008, 60 (15), 1615-1626.

138. Sethuraman, V. A. and Bae, Y. H., J. Controlled Release 2007, 118 (2), 216-224.

139. Petros, R. A. and DeSimone, J. M., Nat Rev Drug Discov 2010, 9 (8), 615-627.

140. Storm, G.; Belliot, S. O.; Daemen, T. and Lasic, D. D., Adv. Drug Delivery Rev. 1995, 17 (1), 31-48.

141. Harris, J. M.; Martin, N. E. and Modi, M., Clin. Pharmacokinet. 2001, 40 (7), 539-551.

142. Greenwald, R. B.; Choe, Y. H.; McGuire, J. and Conover, C. D., Adv. Drug Delivery Rev. 2003, 55 (2), 217-250.

143. Szebeni, J., Mol. Immunol. 2014, 61 (2), 163-173. 144. Thomson, D. A. C.; Tee, E. H. L.; Tran, N. T. D.; Monteiro, M. J. and

Cooper, M. A., Biomacromolecules 2012, 13 (6), 1981-1989. 145. Janib, S. M.; Moses, A. S. and MacKay, J. A., Adv. Drug Delivery

Rev. 2010, 62 (11), 1052-1063. 146. Kelkar, S. S. and Reineke, T. M., Bioconjugate Chem. 2011, 22 (10),

1879-1903. 147. Nyström, A. M. and Wooley, K. L., Acc. Chem. Res. 2011, 44 (10),

969-978. 148. Xie, J.; Lee, S. and Chen, X., Adv. Drug Delivery Rev. 2010, 62

(11), 1064-1079. 149. Lammers, T.; Kiessling, F.; Hennink, W. E. and Storm, G., Mol.

Pharmaceutics 2010, 7 (6), 1899-1912. 150. Nelson, T. R.; Newman, F. D.; Schiffer, L. M.; Reith, J. D. and

Cameron, S. L., Magn. Reson. Imaging 1985, 3 (3), 267-273. 151. Srinivas, M.; Heerschap, A.; Ahrens, E. T.; Figdor, C. G. and de

Vries, I. J. M., Trends Biotechnol. 2010, 28 (7), 363-370. 152. Tirotta, I.; Dichiarante, V.; Pigliacelli, C.; Cavallo, G.; Terraneo, G.;

Bombelli, F. B.; Metrangolo, P. and Resnati, G., Chem. Rev. 2015, 115 (2), 1106-1129.

153. Du, W.; Nyström, A. M.; Zhang, L.; Powell, K. T.; Li, Y.; Cheng, C.; Wickline, S. A. and Wooley, K. L., Biomacromolecules 2008, 9 (10), 2826-2833.

REFERENCES

__________________________________________________

69

154. Peng, H.; Blakey, I.; Dargaville, B.; Rasoul, F.; Rose, S. and Whittaker, A. K., Biomacromolecules 2009, 10 (2), 374-381.

155. Nyström, A. M.; Bartels, J. W.; Du, W. and Wooley, K. L., J. Polym. Sci., Part A: Polym. Chem. 2009, 47 (4), 1023-1037.

156. Thurecht, K. J.; Blakey, I.; Peng, H.; Squires, O.; Hsu, S.; Alexander, C. and Whittaker, A. K., J. Am. Chem. Soc. 2010, 132 (15), 5336-5337.

157. Rolfe, B. E.; Blakey, I.; Squires, O.; Peng, H.; Boase, N. R. B.; Alexander, C.; Parsons, P. G.; Boyle, G. M.; Whittaker, A. K. and Thurecht, K. J., J. Am. Chem. Soc. 2014, 136 (6), 2413-2419.

158. Du, W.; Xu, Z.; Nyström, A. M.; Zhang, K.; Leonard, J. R. and Wooley, K. L., Bioconjugate Chem. 2008, 19 (12), 2492-2498.

159. Kalyanasundaram, K. and Thomas, J. K., J. Am. Chem. Soc. 1977, 99 (7), 2039-2044.

160. Li, Y.; Liu, R.; Liu, W.; Kang, H.; Wu, M. and Huang, Y., J. Polym. Sci., Part A: Polym. Chem. 2008, 46 (20), 6907-6915.

161. Otsuka, H.; Nagasaki, Y. and Kataoka, K., Adv. Drug Delivery Rev. 2003, 55 (3), 403-419.

162. Wassenius, H.; Löfroth, J.-E. and Nyden, M., Starch/Staerke 2006, 58 (2), 66-81.

163. Li, Z.; Kesselman, E.; Talmon, Y.; Hillmyer, M. A. and Lodge, T. P., Science 2004, 306 (5693), 98-101.

164. Hassan, S. S. M. and Rechnitz, G. A., Anal. Chem. 1982, 54 (12), 1972-1976.

165. Jones, D. P.; Carlson, J. L.; Samiec, P. S.; Sternberg, P., Jr.; Mody, V. C., Jr.; Reed, R. L. and Brown, L. A. S., Clin. Chim. Acta 1998, 275 (2), 175-184.